[From the U.S. Government Printing Office, www.gpo.gov]
Biological Report 90 (18) December 1990 Federal Coastal Wetland Mapping Programs A T '7 I N"FY.,; R. i:1 I A 71"10 N C-*,,-,, NAM '71 QH 104 ce F43 sh and Wildlife Servi 19 90 S. Department of the Interior AND S. DEPT OF COMMERCE/NOAA/NOS -.'fice of the Chief Scientist Biological Report This publication series of the Fish and Wildlife Service co.mprises reports on the results of research, developments in technology, and ecological surveys and inventories of effects of land-use changes on fishery and wildlife resources. They may include proceedings of workshops, technical conferences, or symposia; and interpretive bibliographies. Copies of this publication may be obtained from the Publications Unit, U.S. Fish and Wildlife Service, 1849 C Street, N.W, Mail Stop 130-ARLSQ, Washington, DC 20240, or may be purchased from the National Technical Information Service (NTIS), 5285 Port Royal Road, Springfield, VA 22161. The publication of this report was :ftmded, in part, by the National Oceanic and Atmospheric Administration@s Coastal Ocean Program, U.S. Department of Commerce. ISSN 0895-1926 V . S . DEPARTMENT OF COMMERCE NOAA Biological ]Report 90 (18) COASTAL SERVICES CENTER December 1990 2234 SOUTH HOBSON AVENUE CHARLESTON , SC 29405-2413 Federal Coastal Wetland Mapping Programs A Report by the National Ocean Pollution Policy Board's Habitat Loss and Modification Working Group Edited by PrOPertY Of CSC 4ibr=7 Sari J. Kiraly National Oceanic and Atmospheric Administration Office of the Chief Scientist National Ocean Pollution Program Office 1825 Connecticut Avenue, N. W. Washington, D.C. 20235 Ford A. Cross National Oceanic and Atmospheric Administration National Marine Fisheries Service Southeast Fisheries Science Center Beaufort Laboratory Beaufort, North Carolina 28516 and John D. Buffmgton U.S. Fish and Wildlife Service Region 8, Research and Development zj- 1849 C Street, N. W. Washington, D.C. 20240 U.S. Department of the Interior Fish and Wildlife Service Washington, D.C. 20240 Contents Page Preface ....................................................................... iv Federal Coastal Wetland Mapping Programs: Overview and Recommendations Sari J. Kiraly, Ford A. Cross, and John D. Buffington ................ 1 National Programs The U.S. Fish and Wildlife Service's National Wetlands Inventory. Bill 0. Wilen ........ 9 Coastal Barrier Resources System Mapping' Process. Mary C. Watzin .................. 21 National Oceanic and Atmospheric Administration's Habitat Mapping Under the Coastal Ocean Program. James P. Thomas and Randolph L. Ferguson .............. 27 National Oceanic and Atmospheric Administration's National Coastal Wetlands Inventory. Don W. Field, Anthony J, Reyer, Charles E. Alexander, Beth D. Shearer, and Paul V. Genovese ....................................................... 39 Overview of the Land-Sea Interface Research Program. Armond T Joyce, Richard L. Miller, and Ramona E. Pelletier ..................................... 51 Enhanced Environmental Sensitivity Index Mapping Using Remote Sensing and Geographic Information System Technology. Bruce.A. Davis, John R. Jensen, Elijah W. Ramsey, III, and Jacqueline Michel ................................... 59 Wetland Mapping Supported by the U.S, Environmental Protection Agency. John R. Maxted ............................................................ 61 U.S. Environmental Protection Agency's Environmental Monitoring and Assessment Program, an Ecological Status and 7@rends Program. John F. Paul, A. F. Holland, Steven C. Schimmel, J. Kevin Summers, and K. John Scott ........................ 72 Importance of Hydrologic Data for Interpreting Wetland Maps and Assessing Wetland Loss and Mitigation. Virginia Carter .......................................... 80 The U.S. Geological Survey's National Mapping Division P@-ograms, Products, and Services that can Support Wetlands Mapping. Franklin S. Baxter .................. 88 Soil Conservation Service's Wetland Inventory. Billy M. Teels ....................... 94 Regional and Federal-State Cooperative Programs Coastal Mapping Programs at the U.S. Fish and Wildlife Service's National Wetlands Research Center. James B. Johnston and Lawrence R. Handley ................... 106 Monitoring Seagrass Distribution and Abundance Patterns: A Case Study from the Chesapeake Bay. Robert J. Orth, Kenneth A. Moore, and Judith R Nowak ........... 112 Mapping Submerged Aquatic Vegetation in North Carolina with Conventional Aerial Photography. Randolph L. Ferguson and Lisa L. Wood ........................... 126 Page Project Plan for Mapping and Geographic Information System implementation of Land Use and Land Cover Categories for the Albemarle-Pamlico Estuarine Study. H. M. Cheshire and Siamak Khorram .......................................... 135 Loss of Coastal Wetlands in Louisiana: Cooperative Research to Assess the Critical Processes. S. Jeffress Williams and Asbury H. Sallenger, Jr . ...................... 139 Marine Wetland Mapping and Monitoring in Florida. Kenneth Haddad ............... 145 Satellite Data and Geographic Information Systems Technology Applications to Wetlands Mapping. Richard H. Sinclair, Jr., Mark R. Graves, and Jack K. Stoll ...... 151 The Digital Wetlands Data Base for the U.S. Great Lakes Shoreline. Michael Scieszka ... 159 Appendix. Habitat Loss and Modification Working Group ..................... 173 Preface This report was prepared by the National Ocean Pollution Policy Board's Habitat Loss and Modification Working Group, which is an interagency technical committee established by the National Ocean Pollution Policy Board pursuant to recommendations contained in the current National Marine Pollution Program Federal Plan for Ocean Pollution Research, Development, and Monitoring. Fiscal Years 1988-1992 (Federal Plan). The working group is jointly chaired by the National Oceanic and Atmospheric Administration's (NOAA) National Marine Fisheries Service and the U.S. Department of the Interior's Fish and Wildlife Service. The activities of the working group are coordinated through NOAA's National Ocean Pollution Program Office,which also directed preparation of the Federal Plan. Understanding the effects of losing or modifying marine habitats as a result of human activities is one of six goals identified in the Federal Plan. The working group was charged with undertaking projects that would address recommendations outlined in the Federal Plan for achieving this goal at the Federal level, and to arrive at products that would be useful for Federal agencies planning and conducting habitat programs. Three study areas were selected: coastal wetlands mapping, coastal habitat loss, and wetland mitigation. Examining the Federal effort in mapping the Nation@s coastal wetlands was selected as the initial project because determining the current areal extent of these wetlands is fundamental to determining the actual rates and locations of loss. For this project, a workshop was conducted that included persons representing federally funded coastal wetlands mapping programs. The workshop took place in December 1989 at the U.S. Fish and Wildlife Service's National Wetlands Research Center in Slidell, Louisiana. The papers presented at the workshop are contained in this report. They are preceded by an overview of the major federally funded programs and the working group's conclusions and recommendations as to how the overall Federal effort in coastal wetlands mapping could be improve, so that the status and trends of the Nation's coastal wetlands are documented in a timely fashion. iv Federal Coastal Wetland Mapping Programs Overview and Recommendations by Sari J. Kiraly National Oceanic and Atmospheric Administration Office of the Chief Scientist National Ocean Pollution Program Office 1825 Connecticut Avenue, N. W. Washington, D.C. 20235 Ford A. Cross National Oceanic and Atmospheric Administration National Marine F@isheries Service Southeast F'isheries Center Beaufort Laboratory Beaufort, North Carolina 28516-9722 and John D. Buffington U.S. Ksh and Wildlife Service Region 8, Research and Development 184 9 C Street, N. W, Washington, D.C. 20240 Introduction areas. Fundamental to appropriate management is the development of a comprehensive data base New legislative mandates and the increasing that documents the extent, actual locations, and awareness of the value of wetlands have caused rates of loss of the Nation's remaining wetlands. various government and private agencies to in@ Wetlands mapping provides an important basis crease their efforts to study and manage these for such a data base. 1 2 BiowricAL REPoRT 90(18) The manuscripts contained in this report de- U.S. Geological Survey. Because the existing data scribe what the Federal goverriment is doing to base has been developed with user-pays fanding, map the Nation's coastal wetlands. Various as- a completion date for digitizing all NWI maps pects of a series of Federally funded programs are cannot be set. described, including the purpose and intent of the The NWI status and trends analysis was de- programs, technologies used, the type of data and signed to document losses and gains in the other mapping products generated, and how the Nation's wetlands. The national sampling grid for information is used. In this paper, we summarize this analysis consists of stratified random samples the major programs and present the Habitat Loss of 3,635 4-square mile plots distributed within and Modification Working Group's conclusions strata formed by State boundaries, physical and recommendations for actions that could be boundaries, the coastal zone, and the Great Lakes. taken to improve the effectiveness of Federal ac- Plots are allocated to strata in proportion to the tivities. We hope that this assessment of the vari- expected amount of wetland acreage. As legisla- ous Federally funded coastal wetland mapping tively mandated by the Emergency Wetlands Re- programs will reveal strengths, weaknesses, areas sources Act of 1986, a national status and trends for improvement, and opportunities for better co- report for the mid-1960's to the mid-1970's has ordination among the Federal agencies and be- been updated recently. Future updates are to be tween Federal and State agencies as well. prepared every 10 years. National Coastal Wetland The National Oceanic and Atmospheric Administration Mapping Programs As part of its Coastal Ocean Program, NOAA is Two Federal programs are designed to map developing a comprehensive, nationally standard- coastal wetlands on a comprehensive, nationwide ized information system for land cover and habitat basis. These programs are conducted by the change in the coastal region of the United States. U.S. Fish and Wildlife Service (FWS) and the Satellite imagery, aerial photography, and surface National Oceanic and Atmospheric Administra- geographic data will be interpreted, classified, an- tion (NOAA). alyzed, and integrated within a geographic infor- mation system (GIS). The program will delineate The U.S. Fish and Wildlife Service coastal wetland habitats and adjacent uplands, and will monitor changes in these habitats on a The U.S. Fish and Wildlife Service's (FWS) Na- cycle of 1-5 years. Because maps will be spatially tional Wetlands Inventory (NWI) is the most ex- registered digital images, land cover change will tensive national wetlands mapping program. In be detected in a pixel by pixel (30- x 30-m pixels) addition to providing the most comprehensive in- comparison of different time periods, rather than ventory of the Nation's inland and coastal wet- by a comparison of stratified random samples. In lands, it provides the basis for many other Federal addition, maps for a given period will be synoptic, and State mapping efforts. The NWI was initiated based on satellite images or aerial photographs by FWS in 1975 to generate detailed wetland maps collected over short (days or weeks) time intervals. (based on Cowardin et al. [1979]), and reports on This type and frequency of information is required wetland status and trends. to determine the linkages between wetlands and By using conventional aerial photography, the the distribution, abundance, and health of living NWI has produced over 30,000 wetland maps, marine resources. Monitoring changes on a fre- including over 5,300 detailed 1:24,000-scale maps quent basis will also allow appropriate manage- covering 100% of the coastal wetlands in the lower ment steps to be taken in a timely manner. 48 States. The program is scheduled to complete The Coastal Ocean Program mapping effort wetland mapping of the conterminous United will build upon and complement ongoing mapping States by 1998, and mapping of Alaska will be programs of other Federal and State agencies by completed as soon as practicable thereafter. In using existing data to supplement field verifica- addition, 1% of the completed coastal wetland tion. Current efforts include a change analysis for maps have been digitized for inclusion as a na- 1984-1988 and 1989 for emergent coastal wet- tional mapping data-base category in the National lands and adjacent uplands of Chesapeake Bay by Digital Data Base under the supervision of the using satellite imagery. Submerged aquatic vege- WFTLAND MAPPING PRoGRAms 3 tation in North Carolina is also being mapped by mation pertaining to tidal and nontidal wetlands. using aerial photography at scales of 1: 12,000 and This information complements the two-dimen- 1:24,000. The North Carolina study is being con- sional information provided by wetlands maps. ducted cooperatively with the U.S. Environmental The USGS's Geologic Division collects, inter- Protection Agency's (EPA) Albemarle-Pamlico prets, and disseminates basic geological informa- National Estuary Program, and all maps are tion on inland and coastal wetlands. Much of this being digitized and placed in the State of North information is displayed on thematic maps and Carolina's GIS. The intent of the program is to includes data on the three-dimensional structure of eventually map all coastal regions of the United wetlands, as well as how wetlands evolve and States. change through time. Operational protocols for delimiting emergent and submergent coastal wetlands are being devel- The National Oceanic and oped through a series of interagency workshops Atmospheric Administration and meetings. Remote determination of biomass, productivity, and functional status of wetlands will In 1989, NOAA's National Ocean Service and be tested, as will new platforms and sensors as National Marine Fisheries Service completed a they become available. comprehensive Coastal Wetlands Inventory of es- tuarine drainage areas of the United States. The project used a 45-acre grid-sampling technique to Other Federal Mapping quantify existing NWI wetlands maps that were Programs based on aerial photographs from 1971 to 1985. Data were entered into a GIS data base that can The U.S. Geological Survey display and calculate acreage summaries by NWI map, county, State, and estuary. The data base, The U.S. Geological Survey (USGS) performs which contains 5,290 NWI maps and presents data numerous mapping and mapping-related activi- on 507 counties and 92 estuaries, has been useful ties. The major base mapping effort is conducted in providing summaries of wetland distribution by the National Mapping Division. Through its and abundance across large geographic areas. National Mapping Program, the division provides These data will be incorporated into NOAA's Na- standard topographic maps at specified scales, as tional Estuarine Inventory, a comprehensive data well as a diversity of cartographic, geographic, and base useful for evaluating the health and status of remotely sensed data, products, and services, in- the Nation's estuaries. eluding wetlands information. Many Federal and State programs rely on the USGS's primary map The U.S. Environmental series as a basis for site-specific wetland and other Protection Agency environmental studies. The program also provides technical assistance to Federal agencies in estab- Wetland mapping has been supported by the lishing their GIS capabilities for the development U.S. Environmental Protection Agency through of wetlands data bases. the Clean Water Act Section 404 and Superfund The National Mapping Division has prepared programs. There are two basic types of wetland 1:24,000-scale topographic maps covering most of mapping activities under these programs: (1) com- the Nation. Program emphasis has been shifted to prehensive planning under the Section 404 pro- revising the inaccurate and out-of-date maps of gram, referred to as "advance identification" this series. In addition, development of a new se- (ADID), and (2) specific studies of certain identi- ries of land use and land cover maps at the fied Superfund sites. Site-specific mapping in the 1:100,000 scale is being considered. Cooperative second context focuses on wetland boundary efforts with the U.S. Department of Agriculture changes over time, generally as part of a criminal Soil Conservation Service and the NWI should prosecution, and historical data often provide the result in additional products to aid in the study of goal for restoration of the site to its original condi- wetlands, including image base maps and state-of- tion. Mapping conducted under the ADID program the-art GIS's. is intended to steer development away from the The USGS's Water Resources Division collects most valuable wetlands. and disseminates, in written and digital formats, The EPA is initiating an Enviromnental Moni- groundwater and surface water hydrological infor- toring and Assessment Program (EMAP) to char- 4 BiowGicAL RFPoRT 90(18) acterize the condition of the Nation's ecological gram is currently focused on the severe loss of resources on regional and national scales and over wetlands in Louisiana. In cooperation with the long periods. The wetland resource component of State of Louisiana and FWS, USGS is conducting EMAP will develop a program to assess the status field investigations on wetlands loss to identify and trends of wetland condition and extent. The natural and human-made causes. The USGS is also proposed EMAP sampling design calls for selection establishing a GIS network of providers and users of 30 representative 40-km 2 sites within each of of wetlands data in Louisiana. This system will 11 near-coastal geographic regions. Each year, probably be expanded to include the entire Gulf of 25% of these sites will be visited; samples will be Mexico region. taken from plots within each site to determine At the request of Congress, USGS recently pre- habitat condition. pared a study plan for conducting coastal and EPA's wetland mapping activities rely, to a wetlands research to address gaps and needs in large extent, on the mapping conventions devel- geologic information on wetlands evolution. The oped by the NWI program, and in most instances plan was prepared in close coordination with other use NWI maps and NWI mapping capabilities. For Federal agencies and coastal States, and was sub- example, the EPA and FWS collaborate to produce mitted to Congress in June 1990. For fiscal year reports describing the status and trends of wetland .1991, wetland studies are planned for Louisiana, acreage (NWI) and condition (EMAP). Florida, the Great Lakes, and San Francisco Bay. All of these studies will be done in cooperation with State agencies. Major Regional and The Chesapeake Bay Program Federal-State Cooperative Mapping Programs The Chesapeake Bay Program is a joint effort among a number of Federal agencies and the The U.S. Fish and Wildlife Service States bordering the bay. Under this program, submerged aquatic vegetation (SAV) has been The FWSs National Wetlands Research Center surveyed by the Virginia Institute of Marine Sci- has an ongoing program in habitat mapping of ence. The Virginia Institute of Marine Science has wetlands, seagrasses, and uplands. Projects are mapped SAV on a baywide basis five times be- developed in cooperation with other Federal and tween 1978 and 1987, with standard aerial photo- State agencies, such as the U.S. Army Corps of graphic techniques at a scale of 1:24,000. In addi- Engineers, EPA, and the Louisiana Department of tion, data from photointerpretation of the imagery Natural Resources. The wetlands center cooper- have been entered and stored on a Virginia Insti- ates with the NWI and uses NWI procedures for tute of Marine Sciences's GIS. The result of these photointerpretation, quality control, and quality efforts is a temporal delineation of SAV that pro- assurance, and produces maps at several scales. In vides the basis for long-term trends analysis on addition to wetland classification, these maps de- the distribution and abundance of this resource in pict upland classification so that habitat change Chesapeake Bay. analyses can determine what type of uplands re- placed wetland areas. The center also is developing North Carolina wetland maps to include selected indicator species. Information gathered under the program has been Under the Albemarle-Pamlico Estuarine used to develop digital data bases for various Study (funded by EPA and the State of North coastal areas. These data bases can be entered into Carolina), North Carolina State University's the center's GIS to implement natural resource Computer Graphics Center is conducting a land inventories, habitat trend analyses, and carto- use inventory of Albemarle and Pamlico sounds graphic modeling projects. and their tributary basins. This inventory will include over two-thirds of North Carolina's The U.S. Geological Survey coastal wetlands and will be prepared from re- motely sensed satellite data. SAV data generated The USGS has an active National Coastal Geol- by NOAA under its Coastal Ocean Program also ogy Program that includes a number of research will be included in the inventory. The goal is to field investigations related to wetlands. The pro- develop a digital land use and land cover inven- WETLAND MAPPING PRMRAMS 5 tory of the entire Albemarle-Pamlico drainage base that provides information on wetland habitat area that can be maintained and updated as changes in a variety of forms (e.g., statistical and needed as part of the State's GIS. mapped) has been identified by Federal agencies and others as an important tool for decision makers Florida in administering coastal programs. A standardized and centralized data base will allow data collected In 1983, the Florida Department of Natural by different program under varying legislative Resources, Marine Research Institute began build- mandates to be incorporated into individual GIS's ing a digital ecosystem data base through NOAA!s to suit user needs, and the data can be readily Coastal Zone Management Program. Habitat map- updated to reflect current information. ping and trend analysis are key components of the Because of the value of FWS's National Wet- effort. An efficient, cost-effective mapping pro- land Inventory, it is important that it con- gram has been developed based on a combination tinue and that an effort be made to digitize of conventional aerial photography and satellite the available coastal wetland information. At images. State-of-the-art techniques are used for the Federal level, the NWI is the most comprehen- image analysis, resulting in highly accurate maps. sive nationwide mapping program, providing de- A data base for trend analyses also is being created tailed maps of wetland distribution, including those by incorporating other contemporary and histori- in the coastal zone. The NWI is a valuable resource cal data with data collected under the program. All that serves not only as a useful data base both of the data will be incorporated into a GIS to use within the Federal and private sectors, but also as in implementing an ecosystem approach to coastal the basis of many other Federal and State mapping resource management. programs. The status and trends analysis compo- nent, which is based on a stratified random sample, Michigan may not be suitable for assessing trends at the local The Michigan Resource Inventory Program has level, but is appropriate for assessing trends on a prepared a detailed land cover and land use inven- national scale. tory that includes a set of wetland maps. The inven- Because it is critical that changes in coastal tory used conventional infrared aerial photogra- wetland acreage be monitored on a timely phy, primarily oil a 1:24,000 scale, for its mapping basis so that appropriate management strat- effort. In addition, digital products have been pre- egies can be implemented or existing strate- pared from the data and incorporated into the gies modified, particularly in areas of rapid Michigan Resource Information System. The data habitat loss, the national mapping effort collection and digital processing methodology, as needs to be accelerated. It is also essential that well as the products generated, are being used b the implications of change on coastal ecosystems, the International Joint Commission. y including living marine resources, be evaluated while documenting the location and acreage of the Nation!s coastal wetlands. Documentation should be done both at the national level, to assess the Conclusions and overall status and trends of the Nation's coastal Recommendations wetlands, and at the regional or local level so that more detailed assessments can be made. The Federal Effort More focused research is needed to support the development of cost-effective, state-of- Although. many mapping programs are the-art mapping technologies with detailed under way, a centralized and standardized digital satellite images and aerial photo- national digital mapping data base of coastal graphs. We anticipate these newer technologies wetlands is not available and needs to be de- will make it possible to map coastal areas more veloped. Various Federal agencies conduct pro- frequently and accurately, which will provide more grams to document coastal wetland acreage. Some up-to-date information for decision makers. of the programs are on a nationwide scale; others It is particularly important that high-res- are regional. Methodology, frequency, and degree olution, georeferenced digital data bases for of resolution may vary, primarily based on purpose, critical habitat types, based on standard pro- technology availability, and intended use of the tocols and synoptic images, be developed. products. A georeferenced and computerized data Such data bases would allow comparison of chron- 6 Biowr_icAL REPoRT 90(18) ological digital data to assess both national and C'urrently, considerable coordination among Fed- local trends in wetland coverage. Developing a eral agencies exists. One example of coordination standardized set of protocols for extracting digital activities is the cost-sharing agreements between information on wetlands coverage from satellite the NWI and many Federal and State agencies. In images and aerial photographs is fundamental to addition, interagency coordination was important such an effort. NOAA!s Coastal Ocean Program is in preparing NOANs Coastal Wetlands Inventory, developing standardized protocols to produce which was based on maps prepared as part of the georeferenced, digital data bases and digital maps NWI. Habitat mapping under NOAA!s Coastal from satellite images. The NOAA program is con- Ocean Program is another example of an integrated fined to the coastal zone and, in part, relies on NWI 'Federal agency effort that will result in a compre- maps for ground truthing of satellite images. This hensive data base for the Nation's coastal wetlands effort, which complements rather than duplicates and will provide timely information for document- the NWI effort, should continue. ing trends in wetlands and SAV acreage. Because of the relative importance of sub- Coordination among Federal, State, re- merged aquatic vegetation (SAV) to coastal gional, and local levels needs to be improved. ecosystems, a Federal initiative should be de- A number of Federal mapping programs are al- veloped to standardize SAV mapping and to ready coordinating efforts with State agencies (e.g., provide a national data base. Currently, map- FWS's National Wetlands Research Center coastal ping is being conducted by the State of Florida, the mapping projects, USGS coastal erosion and wet- Chesapeake Bay Program, and NOAA"s Coastal land loss projects, EPXs Albemarle-Pamlico Estu- ocean Program for the North Carolina coast. In ary Program, and NOAA!s Coastal Ocean Pro- 1991, EPA will begin mapping SAV in the Gulf of gram). However, Federal cooperation with State Mexico. However, SAV is not being mapped consis- mapping projects should be increased and national tently on a national scale. In addition, standard protocols developed so that State mapping efforts protocols for mapping SAV, such as those being can be integrated with, and complement, other developed by NOAA!s Coastal Ocean Program, are regional and national projects. Ultimately, coordi- needed and should be instituted for future mapping nation and cooperation at this level will allow the programs. development of a comprehensive data base of Fundamental cartographic information, coastal wetland habitats and could provide the such as that developed by USGS and NOAA, model for continued coordination efforts for other needs to be updated, particularly for areas habitat types. An example of such an effort is the where shorelines have eroded or accreted Michigan Resource Inventory Program, which is substantially. Also, coordination and mainte- coordinating with the U.S. Army Corps of Engi- nance of the national data bases, which provide neers and the International Joint Commission in standardized and uniform quality photographic making its data available. A possible mechanism coverage of the 48 conterminous States on a 5-year for improved coordination of programs at the State acquisition cycle, should be continued. In addition, level is the Coastal Zone Management Program existing cartographic, geologic, and hydrographic administered by NOAA. Also, the coastal mapping information should be digitized and collated into a and change analysis protocols being developed by coastal wetland GIS. NOAA!s Coastal Ocean Program could provide the vehicle by which standardization of methodology Federal and State Coordination and data generation could occur. Existing mechanisms for coordinating Coordination among the Federal agencies agency programs could be used for develop- should continue, and efforts should be made ing a consensus on how Federal agencies to identify additional opportunities for coop- should be planning and promoting research erative mapping programs. Coastal wetland on state-of-the-art wetlands mapping tech- mapping programs are being conducted by Federal nology. In addition, these coordination mech- and State agencies for a variety of purposes, and at anisms should be used to aid in the identifi- different levels of resolution. It is important that cation of additional mapping efforts that are these efforts be coordinated, not only because no needed, and to promote even closer coordina- single program can meet all user needs, but also so tion and interaction among the coastal map- that duplication of effort, wasteful resource alloca- ping programs. One existing mechanism is the tion, and incompatibility of data can be minimized. interagency National Ocean Pollution Policy " U. S. DEPARTMENT OF COMMERCE NOAA, National Marine Fisheries Serv. Southeast Fisheries Science Center @Nrls Of Beaufort Laboratory 101 Pivers Island Road Beaufort, NC 28516 To - Addressee From: Dr. Ford A. Cross, Director Enclosed are recent publications of the Beaufort Laboratory - for your information. :NSWTTAL PORM C 0-02 A 110-471 SC WORD OV DAO 21-2 WETLAND MAPPING PRoGRAms 7 Board, which addresses Federal agency coordina- gies, built upon existing programs, legislative man- tion for marine (including coastal) pollution re- dates, and management expertise, should provide search and monitoring. Another mechanism is the a framework for identifying and implementing co- Office of Management and Budget's Circular A-16, ordinated mapping efforts at the national and re- which, as revised, provides for the establishment of gional levels. an interagency committee to promote the coordi- nated development, use, sharing, and dissemina.- tion of surveying, mapping, and related spatial data. The President's Domestic Policy Council Task Reference Force on Wetlands, which is charged with develop- ing a national policy for attaining no net loss of Cowardin, L. M., V Carter, R C. Golet, and E. T. LaRoe. wetlands, is another potential mechanism for inter- 1979. Classification of wetlands and deepwater agency cooperation. Finally, multiagency initia- habitats of the United States. U.S. Fish Wildl. Serv., tives for developing coastal management strate- FWSVOBS-79/31. 103 pp. NATioNAL PRwpAms 9 National Programs The U.S. Fish and Wildlife Service's National Wetlands Inventory by Bill 0. Wilen U.S. Fish and Wildlife Service National Wetlands Inventory 400 Arlington Square Department of the Interior 18th and C Streets, N. W. Washington, D.C. 20240 ABSTRACT.-In 1974, the U.S. Fish and Wildlife Service directed its Office of Biological Services to design and conduct an inventory of the Nation's wetlands. The mandate was to develop and disseminate a technically sound, comprehensive data base concerning the characteristics and extent of the Nation's wetlands. This data base should be used to foster wise use of the Nation's wetlands and to expedite decisions that may affect this important resource. To accomplish this, state-of-the-art principles and methodologies pertaining to all aspects of wetland inventory were assimilated and developed by the newly formed project. By 1979, when the National Wetlands Inventory (NWI) Project became operational, it was clear that two very different kinds of information were needed. First, detailed wetland maps were needed for site-specific decisions. Second, national statistics developed through statistical sampling on the current status and trends of wetlands were needed to provide information to support the development or alteration of Federal programs and policies. Authorization soon as practicable thereafter for Alaska and non- contiguous portions of the United States. The Emergency Wetlands Resources Act of 1986 directs the Secretary of the Interior, through the Introduction Director of the U.S. Fish and Wildlife Service, to produce by 30 September 1990, and at 10-year intervals thereafter, reports to update and im- The Fish and Wildlife Service (FWS) has always prove the information in the report Status and recognized the importance of wetlands to water- Trends of Wetlands and Deepwater Habitats in the fowl and other migratory birds, in part because Conterminous United States, 1950's to 1970's 10-12 million ducks breed annually in the United (Frayer et al. 1983). This act also requires the Fish States, and millions more overwinter here. Conse- and Wildlife Service to produce, by 30 September quently, FWS has a direct interest in protecting 1998, National Wetlands Inventory maps for the wetlands, especially wetlands where waterfowl remainder of the contiguous United States and, as breed and overwinter. 10 BioLoGicAL REPoRT 90(18) We know that wetlands also play an integral gency plans, natural resource inventories, and role in maintaining the quality of human life habitat surveys. through material contributions to the national National estimates of the current status and economy (through the food supply, water supply trends (i.e., losses and gains) of wetlands, devel- and water quality, flood control, and fish, wildlife, oped through statistical sampling, also are needed. and plant resources) and thus to the health, safety, These estimates will be used to evaluate the effec- recreation, and economic well-being of all United tiveness of existing Federal programs and policies, States citizens. identify national or regional problems, and in- crease general public awareness of wetlands. Need for a National Wetlands Inventory In 1954, the FWS conducted a nationwide wet- Pre-operational Phase lands survey that focused on important waterfowl wetlands. This survey covered roughly 40% of the Before actually beginning wetland mapping in lower 48 States. Although this survey was not a 1979, the National Wetlands Inventory Project comprehensive wetlands inventory by today's reviewed existing State and local wetland inven- standards, it was instrumental in stimulating pub- tories and existing classification schemes to deter- lic interest in the conservation of waterfowl wet- mine the best way to inventory wetlands. lands. These findings were published in a well- Researchers determined that a remote sensing known FWS report Wetlands of the United States, technique would be the best method to inventory commonly referred to as Circular 39 (Shaw and wetlands. Fredine 1956). Since this survey, however, wetlands have un- Review of Existing Wetland Surveys dergone many changes, both natural and human- The first step of the pre-operational phase was induced. These changes, coupled with an increased to review existing wetland inventories. The NWI understanding of wetland values, led FWS to es- consulted with Federal and State agencies to learn tablish the National Wetlands Inventory (NWI) where and when wetland surveys had previously Project. The NWI goal is to generate and dissemi- been completed, what inventory techniques were nate scientific information on the characteristics employed, where to obtain copies of any wetland and extent of the Nation's wetlands. We hope this maps that may have been produced, and the status information will foster wise use of the Nation's of State wetland protection. Only a handful of wetlands and provide data for making quick and States had inventoried their wetlands, and most of accurate resource decisions. Decision makers are these had only mapped coastal wetlands. These not able to make informed decisions about wet- results were published in a 1976 FWS report (U.S. lands without knowing how many wetlands, and Department of the Interior 1976). of what type, are where. Needed Wetland Information Review of Existing Classification Systems Before the inventory could begin, NWI research- Two different kinds of information are needed: ers had to decide how to classify wetlands. Thus, detailed maps and status and trends reports. De- in 1975, FWS brought together 15 of the Nation's tailed wetland maps are needed for assessing the top wetland scientists to evaluate the usefulness effects of site-specific projects. These maps serve a of existing wetland classification schemes for the purpose similar to the Soil Conservation Service's National Wetlands Inventory. These scientists de- soil survey maps, the National Oceanic and Atmo- termined that none of the existing systems could spheric Administration's coastal geodetic survey be used or modified for the NWI and that a new maps, and the U.S. Geological Survey's topo- classification system should be developed. graphic maps. Detailed wetland maps are used by local, State, and Federal agencies-as well as by Development of a New Wetlands private industry and organizations-for many Classification System purposes, including comprehensive resource man- agement plans, environmental impact assess- The FWS's wetlands classification system (Cow- ments, facility and corridor siting, oil spill contin- ardin et al. 1979) was developed by a team of four NATioNAL PRoGRAms 11 wetland ecologists, one each from the U.S. Fish sile animal species, or dominant plant and animal and Wildlife Service, U.S. Geological Survey, U.S. species. Cowardin et al. only provide examples of National Oceanic and Atmospheric Administra- the many dominance types possible. Those using tion, and the University of Rhode Island, with the this classification system are expected to identify assistance of local, State, and Federal agencies, as these dominance types and use them as part of the well as many private groups and individuals. The hierarchical classification system. It is also proba- new system went though four major revisions and ble that as the system is used in more detail to meet extensive field testing before its official adoption a user's site-specific needs, the need for additional by FWS on 1 October 1980. subclasses and special modifiers will become clear. This classification system describes ecological This classification system has been adopted by units having certain common natural attributes, many national and international organizations. arranges these units in a system that aids resource Illinois, Michigan, and Oregon have passed State management decisions, furnishes units for inven- wetlands legislation that relies heavily on NWI tory and mapping, and provides uniformity in wet- wetland information for implementation. The NWI land concepts and terminology throughout the was the first phase of a long-range State wetland United States. Although it is not an evaluation Plan for Illinois. Indiana and Minnesota have wet- system, it does provide the information upon which lands legislation currently under consideration evaluations can be made. that will also depend heavily on NWI maps and Wetlands are extremely diverse and complex digital data. The first International Wetlands Con- * ference met in New Delhi, India, on 10-17 Septem- The FWS classification system defines the limits ber 1980. Conference participants passed a motion of wetlands according to ecological characteristics to adopt this classification system. and not according to administrative or regulatory programs. In general terms, wetlands are defined in Cowardin et al. (1979) as lands where satura- Selecting a Remote Sensing Tool tion with water is the dominant factor determining Because of the magnitude of the NWI, remote the nature of soil development and the types of plant and animal communities living in the soil sensing was the obvious technique for inventory of and on its surface. the Nation's wetlands. The basic choice was be- This state-of-the-art wetland classification sys- tween high-altitude photography and satellite im- tem presents a method for grouping ecologically agery (Landsat). After comparing Landsat's capa- similar wetlands. The system is hierarchical, with bilities with FWS's and other agencies' needs for wetlands divided among five major systems at the wetland information, it was evident that Landsat broadest level: Marine, Estuarine, Riverine, La- could not provide the needed data for classification custrine, and Palustrine. Each system is further detail and wetland determinations within the de- subdivided by subsystems that reflect hydrologic sired accuracy requirements. conditions, such as Subtidal versus Intertidal in The National Wetlands Inventory Project has the Marine and Estuarine systems. Below subsys- continued testing satellite technologies. Along tem is the class level, which describes the appear- with the National Aeronautic and Space ance of the wetland in terms of vegetation (e.g., Administration's Jet Propulsion Laboratory, NWI Emergent, Aquatic Bed, Forested), or substrate if conducted a year-long test of the multispectral vegetation is inconspicuous or absent (e.g., Un- scanner to detect and map wetlands in Alaska. With Ducks Unlimited, NWI also tested Thematic consolidated Shore, Rocky Shore, Streambed). Mapper data, as well as data from the French Each class is further divided into subclasses. The satellite called SPOT. A year-long test is now being classification system also includes modifiers to conducted by the Earth Observation Satellite describe hydrology (water regime), water chemis- Company to test the feasibility of using Thematic try (PH, salinity, and halinity), and special modifi- Mapper satellite data to detect wetlands, map wet- ers relating to human activities (e.g., impounded, lands, or update existing wetlands maps. None of partly drained, farmed, artificial). these tests has provided any hope that present Below the class level, the classification system is satellite configurations can provide the needed open-ended and incomplete. The dominance type is data for classification detail and wetness determi- the taxonomic category subordinate to subclass. nations within desired accuracy requirements of Dominance types are determined on the basis of the National Wetlands Inventory Project and its dominant plant species, dominant sedentary or ses- State and Federal cooperators. 12 BiowrxcAL REPoRT W18) Organizational Structure of reports, and textual and geographic computerized the National Wetland data bases. Inventory Project National Wetlands Inventory Maps The Fish and Wildlife employs a small staff of The primary product of the NWI is large-scale biologists assembled into two basic groups; Na- (1:24,000) maps that show the location, shape, and tional Wetlands Inventory central control group characteristics of wetlands and deepwater habi- and regional wetland coordinators. The NWI proj- tats on U.S. Geological Survey base maps. These ect leader works out of the Washington, D.C., office detailed maps are excellent for site-specific project and coordinates the budget, annual work plans, evaluation. and strategic planning. To produce final National Wetlands Inventory The NWI Central Control Group in St. Peters- maps, seven major steps must be completed: burg, Florida, is the focal point for all operational (1) preliminary field investigations, (2) interpreta- activities of the NWI. This group acquires all ma- tion of high-altitude photographs, (3) review of terials necessary for performing an inventory, pro- existing wetland information, (4) regional and na- vides technical assistance and work materials to tional consistency quality control of interpreted the regional coordinators, and produces the wet- photos, (5) draft map production, (6) interagency lands maps. A private service support contractor review of draft maps, and (7) final map production. is responsible for map production, and provides Swartwout (1982) and Crowley et al. (1988) evalu- needed personnel (about 100 technicians and pro- ated NWI maps and determined that the maps fessionals). were 95 and 91% accurate, respectively. Accuracy Regional wetland coordinators at FWS's seven determinations included errors of omission as well Regional Offices are responsible for the inventor as commission. This high accuracy was achieved y because of the NWI technique, which involves a of wetlands within their regions and ensuring that combination of field studies, photointerpretation, all NWI products meet regional needs. They man- use of existing information, and interagency re- age contracts for wetland photointerpretation, co- view of draft maps. ordinate interagency review of draft maps, secure cooperative funding from other agencies, produce regional wetland reports, and disseminate NWI Wetland Status and Trends Reports products. Their addresses, phone numbers, and The national wetlands status and trends analy- areas of responsibility are listed 'in the Appendix. sis study originated from the need for national Photointerpretation and fieldwork are per- estimates on the present extent of our Nation's formed by contractors hired by FWS. These con- wetland resources in the lower 48 States and on tractors photointerpret wetlands with stereo- corresponding losses and gains over the past 20 scopes. In addition, they review soil maps, conduct years. A statistical survey of U.S. wetlands in the field checks, and examine existing information on mid- 1950's and mid- 1970's was conducted through an area's wetlands to ensure accurate identifica- conventional air photointerpretation techniques. tion of wetlands. The status of wetlands in the mid-50's and mid- 70's was determined, and estimates of losses and gains during that interval were computed. The Operational Phase national sampling grid consists of a stratified ran- dom sample of 3,635 4-square-mile plots distrib- uted within strata being formed by State bound- The operational phase of the NWI, initiated on aries, and the 35 physical subdivisions described 1 October 1979, involves two main efforts: wet- by Hammond (1970). Additional strata were added lands mapping and wetlands status and trends to include (1) a coastal zone stratum encompassing analysis. In addition to the wetlands maps and those wetlands near coastal influence, and (2) the trends reports produced through statistical analy- area immediately adjacent to the Great Lakes. sis, NWI produces other products that complement S ample units were allocated to strata in proportion the mapping effort, including a preliminary list of to the expected amount of wetland and deepwater hydric soils, the National List ofPlant Species that habitat acreage estimated as determined by the Occur in Wetlands (Reed 1988), numerous wetland earlier work of Shaw and Fredine (1956). The NATioNAL PfiomAA@ 13 results of this study were published in four major with the Soil Conservation Service, developed the reports by Prayer et al. (1983); Tiner (1984); U.S. first list of the Nation's hydric soils. Since then, the Congress, Office of Technology Assessment (1984); Soil Conservatiolf Service has chaired the Inter- and Goldstein (1988). agency National Technical Committee for Hydric About 215 million acres of wetlands existed in Soils. The National List of Hydric Soils of the the conterminous United States (i.e., lower 48 United States, December 1987 (U.S. Department of States) at the time of the Nation's settlement. In Agriculture 1987) is available from the Soil Con- the mid- 1970's, only 99 million acres, or 46% of the servation Service. This soils list is useful for mak- original wetland acreage remained; these 99 mil- ing wetland determinations in the field, or in the lion acres included 93.7 million acres of palustrine office through use of soil survey maps. wetlands and 5.2 million acres of estuarine wet- lands. Wetlands now cover about 5% of the land List of Plants that Occur in Wetlands surface of the lower 48 States. Between the mid-1950's and mid-1970's, about The U.S. Fish and Wildlife Service published 11 million acres of wetlands were lost, while the National List of Plants Species that Occur in 2 million acres of new wetlands were created. Wetlands: 1988 National Summary (Reed 1988). Thus, in that 20-year interval alone, there was a The plants in this list' are divided into four indi- net loss of 9 million acres of wetlands or an aver- cator categories based on the plants' frequency of age annual net loss of 458,000 acres. This annual occurrence in wetlands: (1) obligate-always loss equals an area about half the size of Rhode found in wetlands (greater than 99% of the time); Island. Agricultural development was responsible (2) facultative wet-usually found in wetlands for 87% of recent national wetland losses, urban (66-99% of the time); (3) facultative sometimes development caused 8%, and other development found in wetlands (33-66% of the time); and caused 5%. (4) facultative upland-seldom found in wetlands The most extensive wetland losses were in Ar' (less than 33% of the time). kansas, Florida, Louisiana, Mississippi, Ne- The wetland plant list data base is a listing of braska, North Carolina, North Dakota, South plants associated with wetlands, as defined by the Dakota, and Texas. Greatest losses of forested U.S. Fish and Wildlife Service's wetland definition wetlands were in the lower Mississippi Valley, and classification system (Cowardin et al. 1979). It with the conversion of bottomland hardwood for- lists scientific and common names of plants, distri- ests to farmland. Shrub wetlands were hardest hit bution, and regional wetland indicator status of in North Carolina, where pocosins in wetlands almost 6,700 plant species. It can be accessed by were converted to cropland or pine plantations, or plant name, region, State, and wetland indicator mined for peat. Inland marsh drainage for agri- status. The data base is updated as additional culture was most significant in the Prairie Pothole information is received. region of the Dakotas and Minnesota, Nebraska's The wetland plant species data base 2 has two Sandhills and the Rainwater Basin, and Florida's parts. The first, PLANTS, contains detailed Everglades. Between the mid-1950's and mid- 1970's, estuarine wetland losses were heaviest in the Gulf States of Louisiana, Florida, and Texas, 1 This listis available from the Superintendent ofDocuments, Most of Louisiana's coastal marsh losses were U.S. Government Printing Office, Washington, D.C., 20402, attributed to submergence by coastal waters. In phone (202)783-3235, Stock Number 024-010-00682-0; cost is $12.00. Thirteen regional subdivisions of the list are other areas, urban development was the major available from the National Technical Information Service, direct human-induced cause of coastal wetland 5285 Port Royal Rd., Springfield, Va. 22161, phone loss. Dredge and fill for residential development (703)487-4650. State lists are also available. in coastal areas was most significant in Califor, 2 Regional subdivisions ofthewetland plant list data base are nia, Florida, New Jersey, New York, and Texas. available on floppy disks in ASCII format for use on IBM XT/AT-compatible machines running the equivalent or MS-DOS 2.0 or higher. Contact the Office or Conference Services, Colorado State University, Fort Collins, Colo. Hydric Soils List (Wetland Soils) 80523, phone (303)491-7767. State subdivisions of the wetland plant list data base are available in a wide variety Hydric soils are defined by soil saturation for a of formats on floppy disks for use on IBM PC significant period or by frequent flooding for long XTIAT-compatible machines running the equivalent of MS-DOS 2.0 or higher. Contact 13IO-DATA, INC., PO. Box periods during the growing season. To clarify the 280605, 331 Wright Street, 7-107, Lakewood, Colo. 80228, meaning of "hydric soils," the NWI, in cooperation phone (303)987-2557. 14 BioLoGicAL REPoRT 90(18) taxonomic, distributional, and habitat informa- major rivers, and other areas that reflect the goals tion on more than 6,200 wetland plants found in of the North American Waterfowl Plan. the United States and its territories. The second The actual priority of mapping depends on the part, BOOKS, contains bibliographic citations for availability of funds and the existence of high- more than 280 sources such as floras, checklists, quality aerial photography. Obtaining acceptable and botanical manuals used to compile PLANTS. photographs for the Prairie Pothole region has been particularly difficult because of the need to capture optimum water conditions. Consequently, Wetland Reports the NWI has established a special agreement with NASA to obtain this photography. The NWI pro- Two basic wetland reports are developed by the duces wetland maps at a rate of 5% per year in the NWI; map reports and State wetland reports. The lower 48 States and at 2% annually in Alaska. This map reports briefly outline NWI procedures and is the equivalent of 3,200 1:24,000-scale quads a findings (e.g., list of wetland plant communities, year in the lower 48 States and 60 1:63,360-scale photointerpretation problems). Map reports are quads in Alaska. available for all mapped areas. By contrast, the State wetland report is a comprehensive publica- tion on the results of the NWI in a given State. It Map Dissemination is prepared on completion of the wetlands inven- More than one million copies of draft and final tory in a State. The State report includes wetland wetlands maps have been distributed by the NWI. 4 statistics and detailed discussions of NWI tech- niques, wetland plant communities, hydric soils, This figure does not include the secondary distri- and wetland values. To date, State reports have bution made through the State-run distribution been produced for Delaware and New Jersey. NWI centers in Alabama, Connecticut, Delaware, expects to prepare reports for Connecticut, Flor- Guam, Hawaii, Illinois, Kentucky, Louisiana, ida, Hawaii, Pennsylvania, Rhode Island, and Maine, Maryland, Massachusetts, Mississippi, Washington when statistics become available. Nebraska, New Hampshire, New Jersey, New York, North Carolina, Ohio, Pennsylvania, Rhode Island, South Carolina, Texas, Vermont, Washing- The Wetlands Values Citation Data Base ton, West Virginia, and Wyoming. The wetlands values citation data base 3 is a National Wetlands Inventory Digital bibliographic listing of more than 12,000 scientific articles about the functions and values of wet- Data Base lands. Field names include author, year, sequence, The NWI is constructing a georeferenced wet- title source, and subject. land data base 5 with geographic information sys- tem (GIS) technologies. Digitizing is done in arc- node format, with attributes assigned to the left, center, and right sides of each arc. Wetland attri- Status of Mapping butes are coded according to Cowardin et al. (1979). As digitization occurs, points are converted The National Wetlands Inventory has produced to latitude/longitude coordinates. As a result, all wetland maps for 68% of the lower 48 States and map data are stored in a common, ground-based 21% of Alaska (Figs. 1 and 2). Mapping priorities geographic reference system. are based principally on the needs of FWS andother Federal and State agencies. Priorities include the coastal zone (including the coastline of the Great Information on the availability of the National Wetlands Lakes), prairie wetlands, playa lakes, floodplains of Inventory maps and ordering information may be obtained by calling (toll-free) 1-800-USA-MAPS. Copies of data-base files can be purchased at cost from the 3 The information is available on floppy disks in ASCII format NWI Office in St. Petersburg, Florida, phone (813)893-3873. for use on IBM PC/XT/AT-compatible machines running the The data are provided on magnetic tape in MOSS export, equivalent of MS-DOS 2.0 or higher. Contact the Office of DLG3 optional, and ELAS, ICES, GRASS formats. Other Conference Services, Colorado State University, Fort products available at cost include acreage statistics by quad, Collins, Colo. 80523, phone (303)491-7767. county, or study area, and color-coded wetland maps. DISTRIBUTION MAPS AVAILABLE ... DRAFT MAPS AVAILABLE WORK UNDER CONTRACT L N WATERFOWL AREAS OF CONCERN TELEPHONE 1-800-USA-MAPS FOR MAP INFORMATION JANUARY 1, 1 -- Fig. 1. Status of the National Wetlands Inventory in relation to waterfowl habitat areas. DISTRIBUTION MAPS AVAILABLE TELEPHONE 1-800-USA-MAPS FOR m MAP INFORMATION DRAFT MAPS AVAILABLE WORK UNDER CONTRACT WATERFOWL AREAS OF CONCERN JANUARY 1, 1990 Fig. 2. Status of the National Wetlands Inventory in Alaska in relation to waterfowl habitat areas. DICITIZATION COMPLETE DIGITIZATION FUNDED TELEPHONE 1-800-USA-MAPS FOR MAP INFORMATION JANUARY 1, 1 Fig. 3. Status of National Wetlands Inventory digital data. 18 BiowmcAL RFPoRT 90(18) To date, more than 5,700 NWI maps, represent- United States, 1950's to 1970's by 30 September ing 10.5% of the continental United States, have 1990, and thereafter at 10-year intervals. Other been digitized (Fig. 3). Statewide data bases have activities will produce National Wetlands Inven- been built for Delaware, Illinois, Maryland, New tory maps for the remainder of the contiguous Jersey, and Washington, and are in progress for United States by 1988 and, as soon as practicable Indiana and Virginia. NWI digital data also are thereafter for Alaska and noncontiguous portions available for portions of 20 other States. of the United States. These digital data are being used for applica- The top priority activity for funding increase is tions such as resource management planning, im- to intensify the national sampling grid used to pact assessment, wetland trend analysis, and in- determine national wetlands status and trends. formation retrieval. This would allow accurate regional wetland acre- age change and loss information to be developed Map and Digital Data: Users and Uses for the Atlantic and Gulf coasts, Great Lakes watershed, lower Mississippi River alluvial plain, The number of map users has grown steadily and lower Prairie Pothole region in 1992, 1993, since the maps were first introduced. Requests are and 1994, respectively. In future years, this fund- common from individuals, private organizations, ing increase will allow the development of accu- industry, consulting firms, developers, agencies rate regional data for the southeastern United from all levels of goverm-nent, and educational and States and the Playa Lakes region in the South- research groups. User surveys have documented west. Regional intensifications of the wetlands over 100 different uses of the wetland maps. Re- inventory will support other resource efforts un- source managers in FWS and in the States are derway within the States and the Department of provided with information on wetland location and the Interior, and will fill a significant wetland type, which is essential for effective habitat man- information data gap identified by the National agement and acquisition of important wetland Wetlands Policy Forum. The intensifications will areas. These areas are needed to perpetuate wa- provide the information needed to develop or alter terfowl populations and other migratory bird pop- management programs to ensure sound steward- ulations as called for in the North American Wa- ship and protection. The coastal intensification terfowl Management Plan. will provide the wetlands data needed to support The Department of Agriculture uses the maps the Secretarial Initiative on Global Change. as a major tool in the identification of wetlands for A second priority will be to operate the current the administration of the "Swampbuster" provis- wetlands Status and trends effort in a continuous ions of the 1985 Food Security Act. Copies of all mode, with reporting done at 5-year intervals, and draft and final maps are sent to the Soil Conserva- with interim estimates as necessary. A continuous tion Service's county offices. wetland trends inventory would involve updating Regulatory agencies use the maps to help in a percentage of the plots each year; for example, advanced identification, determining wetland val- 10% of the plots would be updated each year on a ues, and mitigation requirements. Private sector 10-year cycle. Advantages gained from a continu- planners use the maps to determine the location ous trend process include better coordination with and nature of wetlands to aid in framing alterna- resource priorities, better responsiveness to State tive plans to meet regulatory requirements. These and regional needs, more accurate and current maps are instrumental in preventing problems assessment of wetland resources, and better use that arise because the maps eliminate confusion of existing data. over whether an area is a wetland. They are also instumental because they provide facts that allow sound business decisions to be made quickly, accu- rately, and efficiently. Acknowledgments I thank the staff and contract personnel of the ]Future Activities National Wetlands Inventory for their work over the last decade. The accomplishments presented Future activities of the FWS include updating in this paper could not have been achieved without the report entitled Status and Trends of Wetlands the financial support of many Federal, State, and and Deepwater Habitats in the Conterminous local cooperators. The accuracy of the maps, in NATioNAL PRoGRAms 19 good part, is the result of the voluntary map review Hammond, E. H. 1970. Physical subdivisions of by many Federal, State, local, and private sector the United States. Pages 61---(A in National atlas of agencies and organizations, as well as persons the United States. U.S. Geological Survey, such as G. Fore of Texas. I thank C. Lee for typing Washington, D.C. the manuscript and M. Bates for her review. Reed, P B., Jr. 1988. National list of plant species that occur in wetlands: national summary. U.S. Fish Wildl. Serv., Biol. Rep. W2A). 2M pp. Shaw, S. P, and C. G. Fredine. 1956. Wetlands of the References United States. U.S. Fish Wild. Serv., Circ. 39.67 pp. Swartwout, D. J. 1982. An evaluation of National Cowardin, L. M., V Carter, F C. Golet, and E. T. LaRoe. Wetlands Inventory in Massachusetts. M.S. thesis, 1979. Classification of wetlands and deepwater University of Massachusetts, Amherst. 123 pp. habitats of the United States. U.S. Fish Wildl. Serv., Tiner, R W, Jr. 1984. Wetlands of the United States: FWSVOBS-79/31.130 pp. current status and recent trends. U.S. Fish and Crowley, S., C. O'Brien, and S. Shea. 1988. Results ofthe Wildlife Service, National Wetlands Inventory, wetland study and the 1988 draft wetland rules. Washington, D.C. 59 pp. Report by the Agency of Natural Resources Divisions U.S. Congress, Office of Technology Assessment. 1984. of Water Quality, Waterbury, Vt. 33 pp. Wetlands: their use and regulation. OTA-0-026. Frayer, W, E., T. J. Monahan, D. C. Bowden, and R A. 208 pp. Graybill. 1983. Status and trends of wetlands and U.S. Department of Agriculture, Soil Conservation deepwater habitats in the conterminous United Service. 1987. Hydric soils of the United States. In States, 1950's to 1970's. Department of Forest and cooperation with the National Technical Committee for Wood Sciences, Colorado State University, Hydric Soils. U.S. Department of Agriculture, Soil Fort Collins. 32 pp. Conservation Service, Washington, D.C. Goldstein, J. H. 1988. The impact of Federal programs on wetlan ds. Vol. L The lower Mississippi alluvial U.S. Department ofthe Interior. 1976. Existing State and plain and the Prairie Pothole region. Report to local wetlands surveys 1965--75. U.S. Fish and Wildlife Congress by the Secretary of the Interior. 114 pp. Service, Washington, D.C. 453 pp. 20 BioLoGim REPoRT W18) Appendix. Regional Wetland Coordinators Region Geographic Area Regional Wetland Coordinator 1 California, Guam, Hawaii, Idaho, Nevada, Regional Wetland Coordinator Oregon, Samoa, Washington U.S. Fish and Wildlife Service Eastside Federal Complex 911 N.E. 11th Avenue Portland, Oregon 97232-4181 Comm: (503) 281-6154 FTS: 429-6154 2 Arizona, New Mexico, Oklahoma, Texas Regional Wetland Coordinator U.S. Fish and Wildlife Service 500 Gold Avenue, S.W., Room 4012 Albuquerque, New Mexico 87103 Comm: (505) 766-2914 FTS: 474-2914 3 Illinois, Indiana, Iowa, Michigan, Minne- Regional Wetland Coordinator sota, Missouri, Ohio, Wisconsin U.S. Fish and Wildlife Service 401 East 80th Street Bloomington, Minnesota 55425-1600 Comm: (612) 725-3417 ITS: 725-3417 4 Alabama, Arkansas, Florida, Georgia, Ken- Regional Wetland Coordinator tucky, Louisiana, Mississippi, North U.S. Fish and Wildlife Service Carolina, South Carolina, Tennessee, R.B. Russell Federal Building Virgin Islands 75 Spring Street, S.W., Suite 1276 Atlanta, Georgia 30303 Comm: (404) 331-6343 FTS: 841-6343 5 Connecticut, Delaware, Maine, Maryland, Regional Wetland Coordinator Massachusetts, New Hampshire, New U.S. Fish and Wildlife Service Jersey, New York, Pennsylvania, Rhode One Gateway Center, Suite 700 Island, Vermont, Virginia, West Vir- Newton Corner, Massachusetts 02158 ginia Comm: (617) 965-5100 FTS: 829-9379 6 Colorado, Kansas, Montana, Nebraska, Regional Wetland Coordinator North Dakota, South Dakota, Utah, Wy- U.S. Fish and Wildlife Service oming P.O. Box 25486 Denver Federal Center Denver, Colorado 80225 Comm: (303) 236-2985 FTS: 776-2985 7 Alaska Regional Wetland Coordinator U.S. Fish and Wildlife Service 1011 East Tudor Road Anchorage, Alaska 99503 Comm: (907) 786-3471 FTS: 869-3471 NATIONAL PRoGRAms 21 Coastal Barrier Resources System Mapping Process by Mary C. Watzini U.S. Fish and Wildlife Service National Wetlands Research Center 1010 Gause Boulevard Slidell, Louisiana 70458 ABSTRACT.-The Coastal Barrier Resources Act of 1982 (PL. 97--M) established the Coastal Barrier Resources System (system), a452,834-acre system ofundeveloped, unprotected coastal barriers along 666 shoreline miles of the Atlantic Ocean and Gulf of Mexico coasts. Within the 186 units of the Coastal Barrier Resources System, most Federal expenditures that encourage development are prohibited. Section 10 of the act directed the Department of the Interior (DOI) to conduct a study and prepare a report to Congress on the Coastal Barrier Resources System. The report, delivered to Congress in December 1988, recommended additions to, or deletions from, the Coastal Barrier Resources Systern, and listed modifications to the boundaries of system units. The DOI's recommendations, if passed by Congress, would add about 790,884 acres and 423 miles of shoreline to the Coastal Barrier Resources System. The coastal barriers included in the Coastal Barrier Resources System by Congress in 1982 were designated based on definitions and delineation criteria developed by the DOI in 1981-82. The criteria used by the DOI to delineate barriers included in the 1988 recommendations to Congress differed from those used in 1981 in several respects, reflecting advances in the scientific understanding of coastal barriers, and the functional requirements of a good dermition. I outline the mapping criteria used in 1981-82 and in 1984-87 during the Section 10 study. I also discuss some of the problems encountered in consistently identifying and delineating features across a heterogeneous national coastline, and I comment on future reinventories of coastal barriers. Background on the Coastal Federal expenditures that encourage develop- Barrier Resources Act and the ment are prohibited. The CBRA was the end result of several years Section 10 Study of study by Congress and the Department of the Interior (DOI) of Federal programs and their ef- fects on the development of coastal barriers. In The Coastal Barrier Resources Act (CBRA 1977, the DOI initiated studies to assess options or act) of 1982 (P.L. 97-348) established the for modif@@ing about 40 Federal programs that af- Coastal Barrier Resources System (CBRS or sys- fect coastal barriers, including the National Flood tem), a 452,834-acre system of undeveloped, un- Insurance Program. The results of these studies protected coastal barriers along 666 shoreline were released in a draft Environmental Impact miles of the Atlantic Ocean and Gulf of Mexico Statement in January 1980. Congressional action coasts. Within the 186 units of the CBRS, most followed in 1981 with the enactment of Section 341 ofthe Omnibus Budget Reconciliation Act (OBRA). Present address: Vermont Cooperative Fish and Wilcuife Section 341 of OBRA amended the National Research Unit, School of Natural Resources, Aiken Center, Flood Insurance Act of 1968 to prohibit the issu- University of Vermont, Burlington, Vt. 05405 ance of new Federal flood insurance policies after 22 BioLoGicAL Rm@oRT 90(18) 1 October 1983 for any new construction or sub- acted by Congress, would add about 790,884 acres stantial improvement of structures on undevel- and 423 miles of shoreline to the CBRS. oped coastal barriers. The OBRA directed the DOI to (1) designate coastal barriers based on a defini- tion provided in the OBRA, and (2) report to Con- Definition and Delineation of gress with recommendations relating to the term "coastal barrier." On 13 August 1982, the DOI Coastal Barriers submitted a report to Congress endorsing the gen- eral definitions and delineation criteria for coastal Section 3(lXA) of the Coastal Barrier Resources barriers contained in the OBRA, and designating Act defines a coastal barrier as a depositional 188 coastal barriers that met those criteria. After feature that "(i) consists of unconsolidated sedi- the DOI delivered its report to Congress, but before mentary materials, (ii) is subject to wave, tidal, final OBRA implementation, Congress enacted the and wind energies, and (iii) protects landward Coastal Barrier Resources Act. aquatic habitats from direct wave attack." This In addition to a ban on Federal flood insurance, definition is essentially identical to that included the act also prohibited Federal assistance for a wide in the Omnibus Budget Reconciliation Act in 1981. range of other programs that encourage develop- The definition includes barrier islands, barrier ment of coastal barriers, such as U.S. Army Corps spits, bay barriers, and tombolos. of Engineers structural development projects and Although the definition of coastal barriers is cost-sharing programs for the construction of based on structural and compositional character- roads, bridges, water supply systems, and sewers. istics and is scientifically sound, it was not suffi- Section 6 of the act does allow certain Federal cient for delineating the boundaries of individual activities such as the maintenance of shipping barrier units designated for inclusion in the CBRS. channels, essential military activities, emergency Delineation criteria that were pragmatic, concise, related to features in the ground, and applicable disaster relief, research, and recreational projects. over the full range of coastal barrier variation The purposes of the CBRA are to minimize the needed to be developed. The original criteria devel- loss of human life, reduce damage to fish and wild- oped in 1981-82 relied primarily on features that life habitat and other valuable natural resources of were observable on the ground, maps, and aerial coastal barriers, and reduce the wasteful expendi- photographs. These were modified somewhat in ture of Federal revenues. The effect of the act is to 1984-87 during the Section 10 study. In the follow- place the financial risk associated with develop- ing sections, I discuss the major criteria used for ment on those who choose to live on, or who invest both the 1981-82 and the 1984-87 delineations. in, the coastal barriers. Details appear in the report to Congress (U.S. Section 10 of the act directed the DOI to conduct Department of the Interior 1988). a study and prepare a report to Congress that Minimum size. Because the CBRA allowed for contained recommendations for additions to, or de- portions of undeveloped barriers to be included in letions from, the CBRS, and for modifications to the the CBRS, criteria for minimum size were neces- boundaries of system units. The studies to prepare sary. The major concern was not to include pieces the Section 10 report to Congress began in late 1983 that were so small that they could not function as and culminated in 1988 with a final 22-volume natural geological and ecological units. Two cri- report with 4 appendixes.-The study was conducted teria were established: the unit must include at by a task force representing the U.S. Fish and least 1/4 mile of shoreline on the ocean side and the Wildlife Service (FWS), National Park Service, unit must extend across the barrier from the beach U.S. Geological Survey (USGS), and other Depart- to the bayside aquatic habitats. ment of Interior offices, under the direction of the Developmental status. The difficulty in distin- Assistant Secretary for Fish and Wildlife and guishing developed from undeveloped coastal bar- Parks. The final report includes a description of the riers lies in the fact that relatively few coastal CBRS, background information about the coastal barriers are pristine. Many barriers have been barriers and coastal programs of each State or visibly altered, and others that may seem undevel- Territory, an assessment of conservation alterna- oped contain structures such as highways or recre- tives for the CBRS, and recommendations for spe- ational facilities. The Coastal Barrier Resources cific additions to, or deletions from, the boundaries Act states only that few artificial structures should of the CBRS. The DOI's recommendations, if en- be present and that the areas should function NATioNAL PRoGRAms 23 naturally. Hence, determining whether a coastal helped distinguish linear sand shoals from true barrier was developed required the establishment barriers. of several thresholds and criteria: Associated aquatic habitat. The CBRA defines � The unit contains fewer than one roofed and an "undeveloped coastal barrier" to include all walled building for every 5 acres of fastland associated aquatic habitats: "adjacent wetlands, (nonwetland). These buildings must be marshes, estuaries, inlets, and nearshore waters." constructed according to Federal, State, or local This definition reflects the specific conservation legal requirements and have a projected ground purposes of the CBRA to protect the fish, wildlife, area exceeding 200 square feet. and other natural resources of coastal barriers. All � No structures or human activities significantly such associated aquatic habitats are inseparable impede geological and ecological processes (i.e., parts ofthe coastal barrier ecosystem. The original the area functions naturally). units of the CBRS, however, only include mini- � For units making up only a portion of a barrier, mum aquatic habitat because the 1982 congres- the boundary line is drawn along the 'break' in sional designations were based on DOI delinea- development. tions for a prohibition on just the sale of Federal flood insurance as required by the Omnibus Bud- Composition. In 1982, only coastal barriers get Reconciliation Act. Those delineations focused composed of unconsolidated sand and gravel were on the undeveloped fastland portion of the barriers included in the CBRS. Although coastal barriers where residential development might occur. generally consist entirely of unconsolidated sedi- Coastal barriers protect the aquatic habitats ment composed of sand or gravel, sediments do between the barrier and the mainland. These hab- sometimes include silt, clay, cobbles, or larger itats are critically important for many fish and rocks, or can be consolidated. Areas can be identi- wildlife species, including most species harvested fied that contain carbonate-cemented deposits in the Nation's commercial fish and shellfish in- (such as the Florida Keys), that consist primarily dustries. The barrier and its associated habitats of silt and clay (such as fringing mangroves and are one ecological system, and the health and cheniers), or that contain discontinuous outcrops productivity of the entire system depend on the of bedrock or coarse glacial deposits that never- rational use of all the component parts. theless function as coastal barriers. In the Sec- "Associated aquatic habitat" includes all wet- tion 10 study, the DOI did not require barriers to lands (e.g., tidal flats, swamps, mangroves, and be composed only of sandy, unconsolidated sedi- marshes), lagoons, estuaries, coves between the ments, and recommended to Congress that the barrier and the mainland, inlets, the nearshore definition of coastal barriers in Section 3(l)(A) be waters seaward of the sand-sharing system, and, amended by deleting subparagraph (i) (see previ- in some tropical areas, the coral reefs associated ous definition). with the nearshore mangroves. Under normal Wind, wave, and tidal energies. Winds, waves, weather conditions, only aquatic habitats imme- and tides are the immediate forces that maintain diately adjacent to coastal barriers are exposed to or modify coastal barriers. Criteria were needed direct wave attack. Major storms, however, rou- to determine whether sufficient wind, wave, and tinely affect the entire landward aquatic habitat. tidal energies were present for a landform to be Such habitat survives major storms because considered a coastal barrier. Two criteria were coastal barriers receive the brunt of the ocean's developed: energies. Storm waves break on the barrier beach, leaving a diminished wave to travel into the wet- � A linear or curvilinear beach is present. This land. At the same time, the wetland stores storm kind of beach indicates the existence of floodwaters, easing the flood pressure on the sufficient physical energy as well as an mainland. In the Section 10 study, the associated adequate supply of sediment to satisfy the aquatic habitat was considered as the entire area statutory definition of a barrier. By using this subject to diminished wind, wave, and tidal en- criterion, mud flats, exposed marshes, and other ergy during a major storm because ofthe presence emergent coastal features lacking this linearity of the coastal barrier. This is a considerably more are clearly distinguished from barriers. expanded area than was included in 1981-82. � A beach ridge or berm is located along the Delineation ofthe landward boundary. Once an seaward side of the barrier. This criterion undeveloped coastal barrier and its associated 24 BioLomcAL REPoRT 90(18) aquatic habitat were identified, boundary delin- ocean coast. Nonetheless, these secondary barriers eation was made in the manner presented next. are formed of unconsolidated sediments just like 1. General case: more oceanic coastal barriers, and more impor- The landward boundary is a continuous line tantly, they also protect critical fish and wildlife that follows the interface between the aquatic habitat and provide substantial protection for the habitat and the mainland, as defined on mainland during major storms. In the Section 10 topographic maps or aerial photographs by a study, secondary barriers were delineated and rec- change in vegetation. ommended for addition to the CBRS. Geographic scope. When the Coastal Barrier 2. Special conditions: a. Open water body greater than 1-mile wide Resources Act was enacted in 1982, Congress only landward of the coastal barrier. The bound- included barriers on the Atlantic Ocean and Gulf ary is drawn through the open water about 1 of Mexico coastlines in the CBRS. Although the mile landward of the farthest landward ex- DOI compiled information about coastal barriers tent of wetlands on the protected side of the along the Great Lakes, Pacific Coast, Alaska, Ha- coastal barrier. If a discernible natural chan- waii, and American Samoa, it did not include nel, artificial channel, or political boundary these areas in its final recommendations because exists in the open water about 1 mile land- the legislative history of the act did not clearly ward of the coastal barrier, it is used as the indicate that Congress intended this act to include landward boundary. The boundary is drawn other coastlines. along the channel side nearest the coastal Undeveloped andunprotected coastalbarriers in barrier. the Florida Keys, Puerto Rico, and the Virgin Is- b. Continuous wetlands that extend for more lands also were not included in the CBRS in 1982. than 5 miles landward of the coastal barrier. These barriers border the Atlantic Ocean and are The boundary is generally drawn through the subject to the same dynamic forces and develop- wetlands along an identifiable natural chan- mental pressures as other Atlantic coastal barriers. nel, artificial channel, or political boundary These coastal barriers fully qualify for addition to nearest to the 5-mile limit. If such features the CBRS under the DOI's definitions and were are lacking, the boundary is drawn through delineated and recommended for inclusion in the the wetland generally parallel to, and 5 miles system in the Section 10 report to Congress. landward of, the mean high waterline on the Otherwise-protected coastal barriers. Congress unprotected side of the barrier. excluded from CBRS undeveloped coastal barriers Delineation on the seaward side'. Each unit con- that are "included within the boundaries of an area tains the entire sand-sharing system, including established under Federal, State, or local law, or the beach, shoreface, and offshore bars. In 1982, held by a qualified organization as defined in Sec- the seaward boundary of the CBRS units was not tio.n 170(h)(3) of the Internal Revenue Code of 1954, delineated. In the Section 10 study, the sand-shar- primarily for wildlife refuge, sanctuary, recrea- ing system of open coast barriers was defined as tional, or natural resource conservation purposes" the 30-foot bathymetric contour. In large coastal (hereafter referred to as "otherwise-protected" embayments, the sand-sharing system is more areas). About one-third of the Atlantic and Gulf limited in extent. It was defined as the 20-foot coasts fall into this otherwise-protected category. bathymetric contour or a line about 1 mile sea- Although these barriers were not recommended for ward of the shoreline, whichever is nearer the addition to the CBRS, they were delineated in the secondary barrier. Section 10 study for informational purposes. Secondary barriers. Secondary barriers are Summary. All CBRS units were delineated located in large, well-defined bays (e.g., Nar- on United States Geological Survey (USGS) ragansett Bay, Chesapeake Bay) or in lagoons on 1:24,000-scale topographic quadrangles. Table 1 the mainland side of coastal barrier systems if a lists the number of coastal barrier units, shoreline suitable sediment source and sufficient wind, lengths, and acreages included in the CBRS in 1982 wave, and tidal energies exist. These barriers are and in the final 1988 Section 10 report to Congress. maintained primarily by waves generated inter- Note that the delineations included in the report to nally by wind, rather than open ocean waves. Con- Congress are recommendations only; no changes sequently, these barriers are generally smaller can be made in the CBRS without congressional and more ephemeral than barriers along the open action. NATioNAL PRocRAw 25 Table 1. Summary of the existing Coastal Barrier barrier in 1982. From these photographs, maps Resources System (1982) and the Department of were prepared delineating fastland (any nonwet- the Interior's recommendations to Congress land), wetland, open water, developed areas, and under Section 10 (1988). selected cultural features (Table 2). These maps were prepared as mylar overlays to 1:24,000-scale Existing Coastal Coastal Barrier USGS topographic quadrangles. The center digi- Barrier Resources Resources System tized these maps using a Geographic Information System as recommended System, and compiled acreage statistics about Number of units 186 461 each CBRS unit. These maps and statistics were Shoreline length intended to serve as baseline data accurately char- (miles) 666.4 1,088.9 acterizing the CBRS. For details concerning this Total acreage 452,834 1,243,678 effort, see Appendix A of the report to Congress Fastland acreage 100,934 139,703 (Watzin and Baumann 1988). The center also studied shoreline change and habitat loss in 19 of the 186 original CBRS units. National Wetlands Research We compared historic maps with the 1982 maps to quantify changes, and we examined the processes Center Mapping of the Coastal and human activities in and around the barriers to Barrier Resources System understand the causes of change. We found that in addition to natural changes, all but 1 of the 19 CBRS units had experienced culturally related im- The FWS's National Wetlands Research Center pacts. Dredging has occurred in or near 17 barrier (center) has constructed a digital data base of the units, 15 units have shoreline stabilization struc- existing 1982 Coastal Barrier Resources System. tures in or near them, and 8 units have dams This data base was prepared from large-scale upstream, reducing the sediment supply to the (1: 12,000 or 1: 24,0000) color-infrared photographs coast. Most areas that have experienced human taken of each Atlantic and Gulf of Mexico coastal influences are eroding (Table 3). Table 2. Interpretation classes used in a 1982 inventory of the Coastal Barrier Resources System. Symbol Legend Symbol Legend Coastal Barrier Interpretation FL Fastland M Structure and associated developed area WL Landward wetlands including tidal F @3 Concentrated structures and associated de- flats (between fastland and open veloped areas (number represents total waters) count) F/W5 Fastland with interior wetland Jetties, docks, groins (number represents approximate percentage of the interior wetland-in this example, 5%) low Interior open water, water totally en- --- Road closed within the barrier fastland or wetland Ow Open water Study area boundary Peripheral Land Use Interpretation 1 Developed (includes residential, indus- 5 Open water trial, recreational) 2 Undeveloped (includes open space) --- Road 3 Agriculture Limit of interpretation 4 Wetland 26 BioLomcAL REPorrr 90(18) Table 3. Summary of human perturbations in the 19 Coastal Barrier Resources System units in the study. (Columns do not total because each unit can experience more than one kind ofperturbation.) Number of units by condition Human perturbation Eroding Accreting Stable Total Dredging 15 2 0 17 Structures-updrift 7 1 0 8 Structures-downdrift 5 2 0 7 Structures-within 7 2 0 9 Dams upstream 8 0 0 8 None 1 0 0 1 Number of study units per condition 16 3 0 19 Periodic Reinventory of the Coastal Barrier Resources System. The acreage of Coastal Barrier Resources fastland in each unit gives a general indication of the developable land in each unit, and the count System of structures indicates the development status of those units. At the same time, however, the clas- The CBRA mandates that the CBRS be re- sification system used in the inventory was so inventoried at least every 5 years to update the simplistic that the inventory does not provide official system maps. Section 4(c)(3) states: much useful habitat data for fish and wildlife management, or sufficiently accurate shoreline The Secretary shall conduct, at least once data (because of discrepancies in classifying tidal every five years, a review of the [CBRS] maps flats). Future inventories should correct these . . . and make, in consultation with problems. ne CBRA Section 10 study, however, appropriate officers . . . such minor was a once-only event. There is no mechanism for modifications to the boundaries of system an ongoing inventory or for change analyses of units as are necessary solely to reflect these coastal habitats. changes that have occurred in the size or location of any system units as a result of natural forces. References The first such review was conducted in 1988. High-altitude infrared photographs (1:65,000 U.S. Department of the Interior, Coastal Barriers Study scale) of the CBRS were collected during 1986-88 Group. 1988. Report to Congress: Coastal Barrier and visually compared with the 1982 photographs Resources System. Vols. 1-22. Executive summary. to determine if units had migrated out of their Appendixes A-D. Washington, D.C. existing boundaries. Only one unit, Cedar Island Watzin, M. C., and R. H. Baumann. 1988. Shoreline in Virginia, needed a boundary modification. change and wetland loss in the Coastal Barrier There is no indication that the DOI intends to Resources System: a case study analysis. Appendix A in Report to Congress: Coastal Barrier Resources collect new photographs for this reinventory in the System. U.S. Department of the Interior, future. Washington, D.C. 146 pp. Future Needs for the 1 1 On 16 November 1990, Congress passed the Coastal Barrier Inventory of Coastal Barriers Improvement Act of 1990 (PL. 101-591). This actreauthorizes the CBRA and codifies many of the recommendations in the 1988 report to Congress. The act adds 384 units and 819,666 The 1982 inventory of the CBRS provides a acres to the CBRS. These additions include 80 new units and retrievable data base on the entire existing about 30,000 acres along the Great Lakes coast. NATioNAL PwxmAms 27 National Oceanic and Atmospheric Administration's Habitat Mapping Under the Coastal Ocean Program by James P Thomas National Oceanic and Atmospheric Administration's National Marine Fisheries Service Office of Research and Environmental Information 1335 East West Highway Silver Spring, Maryland 20910 and Randolph L. Ferguson National Oceanic and Atmospheric Administration's National Marine Fisheries Service Southeast Fisheries Center Beaufort Laboratory Beaufort, North Carolina 28516 ABSTRACT.-Timely documentation of the location, abundance, and change in coastal wetlands is critical to their conservation and to effective management of marine fisheries. The rapid changes occurring in these valuable wetlands require monitoring on a 1- to 5- year cycle. Therefore, National Oceanic and Atmospheric Administration, within its Coastal Ocean Program, is initiating a cooperative interagency and State and Federal effort to map coastal wetlands and adjacent upland cover and change in the coastal region of the United States every 2 to 5 years, and to annually monitor areas of significant change. In the first year, fiscal year 1990, the program concentrated on protocol development and prototype studies in Chesapeake Bay and coastal North Carolina. Through a series of workshops and working group meetings, a documented standard protocol will be developed for classifying and mapping habitat location, abundance, and change in the coastal zone of the United States. The Chesapeake Bay prototype study will use Landsat Thematic Mapper imagery and collateral data to map emergent coastal wetlands and adjacent upland cover and change. The coastal North Carolina study will use aerial photography to map and determine change in submerged aquatic vegetation. In outyears, coastal wetlands and adjacent upland cover and change maps will be generated for various coastal regions of the United States, beginning in the Gulf of Mexico. Extant land use and habitat mapping data bases in other Federal and State agencies will be used, where appropriate, to minimize data acquisition cost, supplement ground truth, and assist in verification. This program is cooperative with other Federal and State agencies. Coastal wetlands are being destroyed by dredge and shellfish resources in the coastal United and fill operations, impoundments, toxic pollu- States. Continued loss of these wetlands may lead tants, eutrophication, and, for submergents, exces- to a collapse of coastal ecosystems and associated sive turbidity. Coastal wetlands with emergent fisheries. Documentation of loss or gain of coastal and submergent vegetation (salt marshes, man- wetlands is critical to their conservation and to groves, macroalgae, and submerged aquatic vege- effective management of marine fisheries (Haddad tation [SAV]) support a majority of marine finfilsh and Ekberg 1987; Haddad and McGarry 1989). 28 BiowGicAL REPoRT 90(18) Such information is necessary to respond to Pres- The 1- to 5-year monitoring cycle will provide ident Bush's call for no net loss of wetlands. feedback to habitat managers on the success or Timely quantification of wetland area, location, failure of habitat management policies and pro- and rate and cause of loss is needed now (Kean et al. grams. Frequent feedback to managers will help 1988). Management decisions can then be proactive ensure the continued integrity or recovery of and based on fact, rather than supposition of the coastal ecosystems and the attendant productivity effects of coastal development on coastal wetlands and health of fish and other living marine re- and wetland-dependent fisheries (Fig. 1). Current sources at minimal cost. In addition, the geograph- projections for U.S. population growth in the ical data base developed under the program will coastal zone suggest accelerating losses of wetlands allow both the manager and the researcher to and adjacent habitats as waste loads and competi- evaluate, and ultimately to predict, cumulative tion for limited space and resources increase (U.S. direct and indirect effects of coastal development, Congress 1989). Agencies responsible for coastal on wetland habitats and living marine resources. management must be kept current with regard to Remote sensing (from satellites and aircraft) the extent and status of wetlands and adjacent and other techniques will be used to quantify and uplands. Changes in wetlands are occurring too map coastal wetlands and adjacent uplands. The fast and too pervasively to be monitored once a first cycle will document status and change (retro- decade. Therefore, National Oceanic and Atmo- spectively). The data base, increasing with each spheric Administration (NOAA), within its Coastal subsequent monitoring cycle, will be an invaluable Ocean ]Program, is initiating a cooperative inter- resource for research; evaluation of local, State agency and State and Federal effort to map coastal and Federal wetland management strategies; and wetlands and adjacent upland cover and change in construction of predictive models. As such, this the U.S. coastal region every 2 to 5 years, and to program directly supports NOAA's legislated re- annually monitor areas of significant change. sponsibilities in estuarine and marine science, Monitoring and Research for Understanding, Prediction and Management Human Management, Economic Change 4W Regulations, Change Planning Land use Land Cover Habitat Fisheries Change Change Change Change Fig. 1. An overview of the NOAA habitat mapping paradigm. Long-term and self-sustaining economic, ecological, and fisheries productivity of the coastal zone requires planning and management of human activity. Habitat classification and change analysis, the initial focus of this program, provide a basis for such planning and management. NATIONAL PROGRAms 29 monitoring, and management contained in the mapping SAV and other habitats in selected lim- Fish and Wildlife Coordination Act; the Magnuson ited areas, as suggested by Patterson (1986) and Fishery Conservation and Management Act; the Lade et al. (1988). This combination should be the Coastal Zone Management Act; the Clean Water most effective for accomplishing stated objectives Act; the Marine Protection, Research and Sanctu- at minimal cost. aries Act; the National Environmental Policy Act; and others. Relation to Other Programs NOAA!s habitat mapping effort will -work with Approach and use data from other Federal and State agen- Based on a standard protocol, habitat-classified cies during all program phases. It will build on and maps will be generated from remotely sensed data complement existing coastal habitat mapping pro- " grams and provide essential timeliness, synoptic- including satellite imagery (Landsat Thematic ity, and frequency of repetitive cycles not currently Mapper UK, Multispectral Scanner [MSS]' or available. The 1- to 5-year monitoring cycle is SPOT) and conventional aerial photography. critical to NOAA for effective coastal habitat man- Coastal uplands and wetlands within NOAA!s de- agement and research on a local, regional, and fined "estuarine drainage areas" (NOAA 1985), or national scale. modifications of them, will be mapped retrospec- Extant land use and habitat mapping data tively, and monitored every 2 to 5 years (annually bases in other Federal and State agencies will be in locations of rapid development or significant used, where feasible, to minimize data acquisition change). Habitat-classified maps and historical cost, supplement ground truth, and assist in veri- data from other programs, as well as from wetland fication. Current Federal land use and habitat and estuarine ecologists and fishery biologists, will mapping programs within the U.S. Departments supplement the surface-level verification compo- of Interior, Agriculture, Defense, and Commerce, nent of the program. This approach is intended to as well as the U.S. Environmental Protection build upon and complement ongoing coastal zone Agency and the National Aeronautics and Space sampling and mapping programs carried out by Administration, can provide valuable historical other Federal and State agencies. It will provide and collateral data for this program (Fig. 2). Por- timely and synoptic habitat maps, including SAV, tions of habitat mapping programs, ongoing in and maps of habitat change in the coastal region many States, will be incorporated, where appropri- of the United States. These maps will complement ate, into the overall program to reduce redundancy and augment the more geographically comprehen- in data acquisition, ground truthing, and field ver- sive National Wetlands Inventory maps produced ification. A number of geographically limited TM- by the U.S. Fish and Wildlife Service (FWS). and SPOT-based land use and habitat mapping MSS, TM, and SPOT imagery have been used programs are in their early stages (e.g., in Florida, successfully to detect all types of wetlands Chesapeake Bay, and Albemarle and Pamlico (Haddad and Harris 1985; Lade et al. 1988). Sat- sounds in North Carolina). It is our intent to en- ellite imagery, however, has neither been tested courage such programs through cooperation and nor applied to mapping wetlands on a regional or joint or supplemental funding to provide compara- national scale. The use of satellite imagery for ble and compatible data for mutual use. mapping of wetlands promises a number of advan- tages over conventional aerial photography includ- ing timeliness, synopticity, and reduced costs. Program Development While aerial photography is appropriate for con- struction of habitat-classified maps, satellite imag- NOAA's habitat mapping effort under the ery is better suited and less costly for rapid, re- Coastal Ocean P!rograin is part of CoastWatch- peated observations over broad regions (Haddad Land Applications and Estuarine Habitat Studies, and Harris 1985; Bartlett 1987; Klemas and both components of the Coastal Ocean Program. Hardisky 1987). Although the program will stress As such, the habitat mapping effort will involve the use of satellite imagery, particularly for emer- (internally) four of NOAA!s line organizations: the gent coastal wetlands and adjacent uplands, aerial National Marine Fisheries Service (NMFS); the photography or a combination of photography and Office of Oceanic and Atmospheric Research satellite imagery (TM or SPOT) will be used for (OAR); the National Ocean Service (NOS); and the 30 BiowmcAL REPoRT W18) NOAA- Environmental, NOS,NESDIS, NOAA-NOS NMFS Demographic, Economic & Census NMFS, NW@ Historical & FWS USGS G Fisheries Data EPA Collateral Data uUSDA FWS DOD COE EPA Scs NASA states states bib.. Habitat Classification Mapping & Change Land Use, Habitat Change & Analysis Data Archival Fish: GIS Development, r and Modelling, Prediction NOAA- NIVIFS, NOS Dissemination NESDIS, NOS NOAA - NMFS, OAR, NOS Protocol Other - Federal, State, Academia -st -ve 4Z is M e Ii I---, Team Habitat Function Habitat/Fisheries Federal State Linkage Research Academia Research Team Team NOAA - NMFS, OAR, NOS NOAA - NIVIFS, OAR, NOS Other - Federal, State, Academia Other - Federal, State, Academia Fig. 2. Programmatic relations of NOAA habitat mapping. Relationships of programmatic elements (emphasized blocks), external data and research needs, and program output to management. The text beside each box indicates the responsible or contributing NOAA and external agencies. National Environmental Satellite, Data and Infor- graphical information systems (GIS), or wet- mation Service (NESDIS). Externally, the effort lands in each coastal region of the United will involve other Federal and State programs, as States. The published document resulting previously mentioned. The program will (Fig. 2): from these meetings will be the operational 9 Establish a national operational protocol. protocol to be followed by those generating This protocol will provide a uniform basis for the classified images within, or cooperatively with, the program. classification (i.e., identification of land and 0 Generate, summarize, and distribute habi- habitat cover types) from scene to scene and tat-classified data and maps. SAV, emergent time to time, thereby allowing comparison of coastalwetlands and adjacent uplands in the two or more scenes or times. Methods will be United States co'astal region will be mapped selected that will be valid for all regions and every 1 to 5 years. The derived products will will be able to accommodate various types of be applicable to research and management. remote sensing data. The protocol for gener- They will include (1) classified, color-en- ating, ground truthing and controlling the hanced, geocorrected, and registered images quality of the habitat classifications, and the in digital format; (2) hard-copy inventory use of historical and collateral data, will be maps showing habitat classifications with L"' w at. produced through a series of workshops and geographic references; and (3) reports de- working group meetings. These meetings will scribing and tabulating the classification have input from Federal and State agencies slimm by State, county, and hydrologic producing or using habitat-classified data in unit. Maps will be produced at a 1:250,000 the coastal zone, and from professionals with scale. However, data for all areas will be related expertise in remote sensing, geo- retained at full mapping resolution. Thus, NATioNAL PRwRAms 31 finer-scale maps (1:24,000) could be produced through the CoastWatch-Land Applications for selected areas on a cost-recoverable basis. manager by the group responsible for the 0 Determine, summarize, and map change in processing of the imagery. Such copies will areal coverage of each habitat classification. meet the following programmatic needs: Change in land and habitat cover in the U.S. quality control, integration and archiving of coastal region will be mapped every 1 to 5 data, guidance, oversight, and planning. years. To delineate change, pixel by pixel (or 0 Apply habitat classification and change anal- equivalent) comparison between scenes from ysis to habitat and fisheries research and two different times will be done. Such change management. This program will integrate its will be presented as geocorrected and regis- products with other spatial data (e.g., demo- tered maps (1:250,000 scale), with State, graphics, land use, pollution, living marine county, or hydrological boundaries added. resources, fisheries, and economics) to gener- Mapped data for all areas will be retained at ate a data base with a depth and scale pre- full mapping resolution. Thus, finer-scale viously unavailable to researchers and man- maps (1:24,000) could be produced for se- agers. The data base will be researched by a lected areas on a cost-recoverable basis. Ad- multidisciplinary team to generate guidance ditionally, synthesis reports, with text and and feedback to habitat management and tables listing change (hectares) for each hab- research personnel and to aid in Federal and itat classification by State, county, and hy. State long-range regional planning for habi- drologic unit, will be produced. tats and fisheries. Remotely determine biomass, productivity, and health status of habitats. This is a re- search activity to develop methodology and Discussion of Methods algorithms for image processing that will allow the determination of biomass, produc- The spatial and temporal resolution of data and tivity, and functional health status of coastal landward and seaward extent of the study area are wetlands habitat by remote sensing. The use critical issues for the program. Imagery must re- of such algorithms would allow large areas of solve and detect changes in coastal wetlands af- wetlands to be surveyed and assessed by sat- fecting living marine resources at scales suitable ellite or aircraft much more rapidly and eas- to support habitat research and conservation; ily (Bartlett 1987). By comparing two or more allow modeling of the link between coastal habi- time periods, change in biomass, productiv- tats, coastal ecosystem functioning, and fisheries; ity, or some other as yet undefined observable and support strategic habitat and fisheries man- factor affecting or correlated with spectral agement. Change or loss of coastal wetlands must reflectance could be used to index functional be documented on time scales necessary to avoid health. This activity requires ground-based or decrease future losses and to assist in focusing research to relate remotely sensed spectral efforts to address the no net loss of wetlands policy radiances to biomass, productivity, and, po- for the benefit of living marine resources. Areas of tentially, other factors indicative of the func- significant change will be mapped more frequently tional health of coastal wetlands. (i.e., every 1 or 2 years), and areas of relatively little change will be mapped less frequently (i.e., Archive and disseminate data. The digital every 3 to 5 years). data base will be archived and disseminated The landward and seaward extent of the area in standard exchange formats as either an mapped must include the coastal waters and the optical disc or data tape. NOAA/NESDIS/Na- adjacent uplands that most directly affect coastal tional Oceanographic Data Center in Wash- wetland habitats. The NOAA/NOS "estuarine ington, D.C., will distribute the data base on drainage area" (i.e., the land/water component of a cost-recoverable basis to outside users. an entire watershed that most directly affects an Hard-copy maps will be produced, archived, estuary) or a modification of that area will be used and distributed to outside users by the to defme the landward boundary (NOAA 1985). NOAA/National Ocean Service on a cost-re- The seaward boundary, beyond the estuary, will be coverable basis. For those participating in the defined by the 10-in isobath in most cases. How- program, both the digital data and prelimi- ever, in special circumstances it may need to be nary habitat-classified maps will be provided defined to 50 in or deeper (Orth et al. 1990). 32 BioLoGicAL REPoRT 90(18) Compared with aerial photography, digital re- verification of the preliminary habitat-classified mote sensing from satellites offers the advantages maps will be accomplished by local, State, and of synoptic, large-area coverage and frequent repe- Federal experts and NOAA-supported surveys. tition (Dobson 1987).Translation of digital spectral The program also will make use of historical and data to habitat classifications can be done objec- collateral data as much as possible for ground tively, rapidly, and reproducibly. The major advan- truthing and verification. These collateral data tage is that starting with digital input data, remote- sources potentially will include NOAA-supported sensing specialists, with existing software and field surveys, individual State surveys, the FWS's equipment, can analyze and communicate far more National Wetlands Inventory, the U.S. Geological information about much larger areas faster and at Survey's (USGS) Land Use Data Analyses, the far lower cost than is possible with aerial photogra- county soils and wetlands surveys of the U.S. Soil phy (Haddad and Harris 1985; Bartlett 1987). In Conservation Service (SCS), the National High fact, the spatial resolution available from TM and Altitude Aerial Photography Program, and the SPOT imagery is more precise than that available wetland and coastal surveys of the U.S. Army on many analog habitat maps (Dobson 1987). Corps of Engineers, the U.S. Environmental Pro- Three satellite sensors provide a diverse capa- tection Agency, and the National Aeronautics and bility. MSS data provide a spatial resolution (pixel Space Administration. size) of 79 by 56 in. Each scene covers an area on The program will produce a documented, com- the ground about 180 by 180 kin. TM data are puterized digital data base of all habitat classifica- collected for pixels of 30 by 30 m, with the same tions, including those (e.g., SAV) obtained by aerial size scene as MSS. SPOT pixels are 20 by 20 in for photography. These data are to be incorporated spectral data (10 by 10 in for panchromatic data) into a user-friendly GIS capable of iritercomparing and cover an area on the ground of 60 by 60 kin. habitat-classified data and surface-level data, in- TM offers a greater spectral range (seven intermit- cluding biological, demographic, edaphic, eco- tent bands from 0.45 to 12.5 pm than MSS (four nomic, fisheries, and physical data, in statistical contiguous bands from 0.5 to 1.1 pm) or SPOT ways and in graphics. (three intermittent bands from 0.5 to 0.89 pm). The higher spatial resolution and greater spec- Proposed Activities tral discrimination of TM and the higher spatial resolution of SPOT relative to MSS are valuable Proposed activities for NOAA habitat mapping characteristics, but they come at a higher cost (full are depicted in Fig. 3. scenes are $4,900 for geocoded- TM, $3,200 for Task 1. During fiscal year (FY) 90-91, through SPOT Level II, and $200 for MSS data before a series of workshops and working group meetings, 1988). Because they are more data intensive, TM a documented standard protocol for classifying and and SPOT images also require greater data pro- mapping habitat location, abundance, and change cessing and storage space per unit of surface area in the coastal region of the United States is being on the ground. Both the MSS and the TM data are developed (Fig. 4). A preliminary protocol devel- capable of providing classified results with 80 to oped for the Chesapeake Bay area is being evalu- 90% accuracy. TM can provide better discrimina- ated and modified as necessary. Habitat classifica- tion of wetland types, whereas TM and SPOT data tions will allow SAV, mangroves, marshes, and are better able to resolve small features. At this other coastal wetlands and adjacent upland habi- point we favor TM or SPOT data for routine imag- tats, and potentially subdivisions of these habitat ery of the coastal zone. MSS data, however, have types, to be determined. The protocol will be flexi- value for retrospective analyses because image ble, yet allow comparison of data such that statis- acquisition began in 1972, compared with 1982 for tical analyses of areal coverage, location, and rate TM, and 1986 for SPOT. In areas of special inter- of change can be computed for specified habitat est, where even greater resolution and discrin-Ana- classifications. Flexibility is needed to take into tion may be needed (e.g., SAV), low-altitude verti- account data of different spatial and temporal cal aerial photography will be used. scales, and from many regions and sources. Data Clustering analysis techniques will be used to input will include that from satellite sensors (TM, generate the habitat types. Classification methods MSS, and SPOT), aircraft sensors (color, infrared, will be evaluated to determine which best account and black and white photography), and on-foot for latitudinal/longitudinal and tidal effects. Field surveys. Each has different spatial, spectral, and NATioNAL PRoGRAms 33 l"Y90 FY9 1, FY92 FY93 FY94, FY95 A. PROTOCOL DEVELOPMENT Workshops Ches. Bay Wetl, Uplands Prototype N.C. Seagrass Prototype Regional Projects B/C. LAND COVER/ HABITAT CHANGE Gulf of Mexico Atlantic West Coast Seagrass surveys NC FL TX A SUPPORTING RESEARCH Biomass, productivity, health E. INTEGRATION & MODELING Land, habitat, fish, and economics Fig. 3. Activities schedule for NOAA habitat mapping. temporal resolutions. Likewise, the protocol will Participants at the workshops and working allow for use of data from other major sources (e.g., group meetings include technical experts in the FWS's National Wetlands Inventory, USGS's collection, processing, and use of habitat-classifi- Land Use Data Analysis, SCS's soil survey maps, cation data (e.g., ecologists, fishery biologists, sat- the National ffigh Altitude Photography Program, ellite image processors, geographic data-base spe- State surveys, and others). The protocol topics will cialists, modelers, and habitat and fisheries include: managers). � habitat classifications (e.g., Cowardin et al. Task 2. In FY90, a prototype effort classifying [1979] for wetlands, Anderson et al. [1976] and merging four TM scenes from two different for uplands, or some modification of the two); periods (1984 and 1988/89) for the estuarine drain- � selection and surface-level sampling of train- age area of Chesapeake Bay will be accomplished. ing sites; The classified, merged scenes will be produced as color-enhanced maps (scale 1:100,000), one for � image selection from satellites, mission au- each period. The two periods will be compared thorization for aerial photography, and qual- pixel by pixel (or equivalent), and a change analy- ity control of images; sis map (scale 1:100,000) will be produced. Finer- � image processing, use of historical and collat- scale maps (1:24,000) covering any portion of the eral data, photointerpretation and digitiza- area can be produced for each period or for change tion, and change analysis; analysis on a cost-recoverable basis. Tables will � map projection and scale, and other graphic summarize the areal extent (in hectares) of habitat and statistical products; and classifications at each period and the changes that � quality control of image acquisition, occurred from one period to the next. We have georeferencing, habitat classification, data access to, and will incorporate into this task, a processing, digitization, and map projection previous habitat classification of the Chesapeake and scale. Bay area for 1978 (Dobson and Bright 1989), and 34 BioLoGicAL REPoRT 90(18) Protocol Functional Development Health Research Digital Data FAerial Photography Tape Purchase Acquisition Z GroundTruthing for Image Classification Field Cliecls Fessing Imagery & Photography A I Use of Collateral Data .771 First Time Period (Second Time Period) Tables - Co., State, Water Unit Maps - 1:24,000 1:250,000 Classification Field V rification e Change Analyses Tables & Maps (as above) Digital Data Archive MAPS GIs Input NODC NOS NMFS I NOSOAR Fig. 4. Framework for protocol development. NA,NONAL PRoaR"s 35 change analyses for Metonikin Inlet, Virginia, and coasts, or analyses of new coastal areas, could vicinity (1974-1982). During FY91, a statistical begin in FY94. evaluation of the results, and additional protocol In FY92 and beyond, SAV mapping will be ex- testing will take place. tended to include Florida, Texas, and other Gulf of Task 3. During FY90-91, researchers will con- Mexico and east and west coast States, and moni- tinue a prototype effort to map SAY in the sounds toring will be established for North Carolina. High- of eastern North Carolina (Ferguson et al. 1989; quality, baywide, digitized information about the Ferguson and Wood 1990). This task will use ex- annual distribution of SAV in Chesapeake Bay tant (1985 and 1988) and supplemental aerial pho- (Orth et al. 1990) already is available through the tography at scales of 1:12,000 to 1:50,000 and will combined FederaVState Chesapeake Bay Pro- produce photographic scale tracings and 1:24,000- gram, with supplemental funding from the NOAA scale maps to document location, area, and change Coastal Zone Management P@rogram to the States in area of SAY coverage. Photointerpretation ofthe Of Maryland and Virginia. Once the SAY mapping presently available 1:12,000- and 1:20,000-scale effort is in place in North Carolina and other photography (1985) will be completed for SAV in States, repeat surveys for change analyses should the area from Bogue Inlet to Drum Inlet. Pho- take place every 2 to 4 years. These surveys will be tointerpretation of photography at the 1:24,000 merged with surveys of emergent wetlands and scale (1988) will be completed in the area from adjacent uplands to form an integrated mosaic. Cape Lookout to Oregon Inlet. A portion of this Task 6. In FY91, depending on funding, re- photography (Cape Lookout to Drum Inlet, 1985, searchers will begin to relate the functional health and Drum Inlet to Ocracoke Inlet, 1988) already (biomass, primary and secondary productivity) of has been interpreted, digitized, and published as a tidal wetlands to the spectral radiances observed NOAA chart (Ferguson et al. 1989; Ferguson et al. through remote sensing. Our initial effort will be 1991). The photography from 1988 overlaps this a literature survey and report describing previous area and will allow change analysis between 1985 research, status oftechnology and knowledge, and and 1988. directions for future research. The literature sur- Task 4. In FY91, the protocol will be tested vey serves two purposes: it will evaluate the po- through a number of regional prototype demon- tential for remote discernment of the functional strations around the country (location and number health of emergent wetland types, and it will determined in FY91). These regional prototypes determine what additional field studies are will include regional representatives of each habi- needed to achieve this goal. The ultimate goal is tat-classified type and will be distributed along the to determine the conditions in which remote sens- coastal zone of the United States such that latitu- ing can subclassify functional states of major hab- dinal, longitudinal, and tidal differences (i.e., veg- itat classifications. etation type, height, biomass, season, and degree Beyond FY91, based on the literature survey, of inundation) will be considered. Special tests will wetland types of diverse functional health would be be conducted to determine where and when water observed to seek spectral definitions of health and level changes affect the data from one period to the productivity. This effort would involve fieldwork next. NOAA/State Estuarine Research Reserves, such as that described by Crouse (1987) and Gross EPA/State National Estuary Program areas, and et al. (1987) who used hand-held radiometers to examine both affected and -unaffected marsh areas. FWS refuges will be used, where appropriate, as Hand-held radiometric measurements would be sites for these demonstrations. compared with concurrent spectral radiance mea- Task 5. In FY91, habitat classification and surements obtained from satellites -(and perhaps change analyses for emergent coastal wetlands aircraft) passing over the experimental sites. The and adjacent uplands will begin via the purchase plan is to study a series of experimental sites rep- of remotely sensed data tapes for the Gulf of Mex- resentative of different areas of the coastal zone of ico or Atlantic coasts ofthe United States, depend- the United States. These areas will be selected in ing on funding. Processing will begin in FY92. concert with the regional prototype demonstrations Purchase and processing of data tapes for the mentioned previously, as well as with sites used by Pacific coast of the United States could occur as NOAA (i.e., Coastal Ocean Program-Estuarine early as FY92-93, depending on funding. Repeat Habitat Studies) and other Federal and State pro- surveys of the Gulf of Mexico, Atlantic, or Pacific grams (e.g., Mendelssohn and McKee 1985, 1987) 36 BioLoGicAL REPoRT 90(18) to develop informatiQn on natural versus anthropo- sensing of estuaries. U.S. Department of genic stress-related alterations in biological pro- Commerce, the National Oceanic and Atmospheric ductivity. Such information will be integrated with Administration, and the U.S. Government Printing this effort in order to develop a more complete Office, Washington, D.C. picture of the extent of estuarine habitat degrada- Costanza, R., E H. Sklar, M. L. White, and J. W Day, Jr. tion, the processes active in bringing about such 1988. A dynamic spatial simulation model of land loss and marsh succession in coastal Louisiana. Pages changes, and potential areas for subsequent field 99-114 in W J. Mitsch, M. Straskraba, and S. E. investigations. Jorgensen, eds. Wetland modeling, developments in Task Z In FY91, depending on funding, data environmental modeling, 12. Elsevier Science integration and analysis will be initiated to begin Publishers B.V, Amsterdam. linking demographic patterns and habitat man- Cowardin, L. M., V Carter, F C. Golet, and E. T LaRoe. agement practices to wetland stability or loss on 1979. Classification of wetlands and deepwater habitats of the United States. U.S. Fish Wildl. Serv., an area-specific basis. Beyond FY91, studies are FWEVOBS-79/31. 103 pp. planned to relate wetland changes and fishery Crouse, V 1987. Gauging the health of a salt marsh management practices to the success of estuarine through remote sensing. Univ. Del. Sea Grant Rep. and coastal ocean fisheries. Economic assessments 6(l):1-5. of alternative management strategies will be in- Dobson, J. E. 1987. Geographic analysis for wetlands cluded in these studies. change detection. Pages 157-161 in V Klemas, J. P Thomas, and J. B. Zaitzeff, eds. Proceedings of a This task will require the cooperative efforts of workshop on remote sensing of estuaries. U.S. a multidisciplinary team (i.e., demographers, wet- Department of Commerce, the National Oceanic and land ecologists, fishery biologists, and economists) Atmospheric Administration, and the U.S. within and outside of NOAA, in order to integrate Government Printing Office, Washington, D.C. the diverse data generated by this program into a Dobson, J. E., and E. A. Bright. 1989. Cover photo: comprehensive geographic information system. Chesapeake Bay. Photogram. Eng. Remote The output ultimately would be used to develop F Sens, 55(6). erguson, R. L., J. A. Rivera, and L. L. Wood. 1988. models (e.g., Costanza et al. 1988) to assess pres- Seagrasses in southern Core Sound, North ent status and to predict future trends in coastal Carolina. National Oceanic and Atmospheric uplands, wetlands, and fisheries resources. Re- Administration-Fisheries submerged aquatic gional Fishery Management Councils, land-use vegetation study, southern Core Sound, North planners, economists, and environmental manag- Carolina. NOAA-Fisheries, Beaufort Laboratory, ers require this information (Haddad and Ekberg Beaufort, N.C. 28516. 1987; Haddad and McGarry 1989) and will be Ferguson, R. L., J. A. Rivera, and L. L.Wood. 1989. encouraged to help plan for its generation. Submerged aquatic vegetation in the Albemarle- Pamlico estuarine system. Albemarle-Pamlico Estuarine Study, Project 88-10. North Carolina Department of Natural Resources and Community Acknowledgments Development, Raleigh, N. C., and the U.S. Environmental Protection Agency, National Estuary We thank R. Edwards, R. Lippson, and J. Pearce Program, Washington, D.C. 68 pp. for ideas used in this paper. J. Dobson, K. Haddad, Ferguson, R. L., and L. L. Wood. 1990. Mapping V. Klemas, P. Lade, R. Orth, and others provided submerged aquatic vegetation in North Carolina technical advice. NOAA's habitat mapping working with conventional aerial photography. Pages 126-134 group, composed of R. Ferguson (NMFS), P. Grose in S. J. Kiraly, F A. Cross, and J. D. Buffington, Eds. (NOS), R. Lippson (NMFS), G. Mayer (OAR), Federal coastal wetland mapping programs. U.S. Fish Wildl. Serv., Bio. Rep. 90(18). R. Stumpf (NESDIS), and J. Thomas (NMFS), pro- Gross, M. E, M. A. Hardisky, V Klemas, and P L. Wolf. vided programmatic guidance. 1987. Quantification of biomass of the marsh grass (Spartina alterniflora Loisel) using Landsat Thematic Mapper imagery. Photogram. Eng. Remote References Sens. 53(11):1577-1583. Haddad, K. D., and D. R. Ekberg. 1987. Potential of Anderson J. R., E. E. Hardy, J. T Roach, and R. E. Landsat TM imagery for assessing the national Witmer. 1976. A land use and land cover classification status and trends of coastal wetlands. Pages system for use with remote sensor data. U.S. Geol. 5192-5201 in 0. T. Magoon, H. Converse, D. Miner, Surv. Prof. Pap. 964. 28 pp. L. T. Tobin, D. Clark and G. Dormurat, eds. Coastal Bartlett, D. S. 1987. Remote sensing of tidal wetlands. Zone '87. Proceedings of the fifth symposium on Pages 145-156 in V Klemas, J. P Thomas, and J. B. coastal and ocean management. American Society of Zaitzeff, eds. Proceedings of a workshop on remote Civil Engineers, New York. NATioNAL Pw)GRAms 37 Haddad, K. D., and B. A. HarTis. 1985. Use of remote Mendelssohn, 1. A., and K. L. McKee. 1985. The effect of sensing to assess estuarine habitats. Pages 662-675 nutrients on adenine nucleotide levels and the in 0. T. Magoon, H. Converse, D. Miner, D. Clark, and adenylate energy charge ratio in Spartina alterniflora L. T Tobin, eds. Coastal zone'85. Proceedings of the and S. patens. Plant Cell Environ. 8:213-218. fourth symposium on coastal and ocean management. Mendelssohn, 1. A., and K. L. McKee. 1987. Effect of American Society of Civil Engineers, New York. salinity on proline accumulation in three Spartina Haddad, K. D., and G. McGarry. 1989. Basin-wide management: a remote sensing/GIS approach. Pages species. XIV International Botanical Congress, 1822-1836 in 0. T. Magoon, H. Converse, D. Miner, Berlin, Germany, 24 July-1 August. L. T. Tobin and D. Clark, eds. Coastal zone '89. National Oceanic and Atmospheric Administration. Proceedings of the sixth symposium on coastal and (NOAA) 1985. National estuarine inventory data atlas. ocean management. American Society of Civil Vol. 1. Physical and hydrologic characteristics. U.S. Engineers, New York. Department of Commerce, Washington, D.C. Kean, T H., C. Campbell, B. Gardner, and W K. Reilly. Orth, Robert J., K. A. Moore, and J. F Nowak. 1990. 1988. Protecting America's wetlands: an action Monitoring seagrass distribution and abundance agenda. The final report of the national wetlands patterns: a case study from the Chesapeake Bay. policy forum. The Conservation Foundation, Pages 111-123 in S. J. Kiraly, F A. Cross, and J. D. Washington, D.C. 69 pp. Buffmgton, eds. Federal coastal wetland mapping Klemas, V, and M. A. Hardisky, 1987. Remote sensing programs. U.S. Fish Wildl. Serv., Biol Rep. 90(18) of estuaries: an overview. Pages 91-120 in V Klemas, J. P Thomas, and J. B. Zaitzeff, eds. Proceedings of a Patterson, S. G. 1986. Mangrove community boundary workshop on remote sensing of estuaries. U.S. interpretation and detection of areal changes on Department of Commerce, the National Oceanic and Marco Island, Florida: application of digital image Atmospheric Administration, and the U.S. processing and remote sensing techniques. U.S. Government Printing Office, Washington, D.C. Fish Wildl. Serv., Biol. Rep. 86(10). 87 pp. Lade, P K., D. Case, J. French, and H. Reed. 1988. U.S. Congress. 1989. Coastal waters in jeopardy: Delineation and classification of submerged aquatic reversing the decline and protecting America's vegetation using SPOT satellite multispectral digital data. Final report (14 September) to the Maryland coastal resources. Oversight Report of the Department of Natural Resources, Tidewater Committee on Merchant Marine and Fisheries. Administration, Coastal Resources Division, Serial 100-E. U.S. Government Printing Office, Annapolis, Md. 53 pp. + Appendixes. Washington, D.C. 47 pp. NAnONAL PROGRAMS 39 National Oceanic and Atmospheric Administration's National Coastal Wetlands Inventory by Don W Field, Anthony J. Reyer, Charles E. Alexander, Beth D. Shearer, and Paul V Genovese National Oceanic and Atmospheric Administration National Ocean Service Office of Oceanography and Marine Assessment Ocean Assessments Division Strategic Assessment Branch 6001 Executive Boulevard Rockville, Maryland 20852 ABSTRACT.-A comprehensive and consistently derived data base describing the areal extent and distribution of coastal wetlands in the conterminous United States does not presently exist. We discuss efforts of the National Oceanic and Atmospheric Administration (NOAA) to develop such a data base using a systematic grid-sampling procedure on wetland maps produced by the National Wetlands Inventory (NWI) of the U.S. Fish and Wildlife Service. These maps, developed using areal photography, are generally based on 1:24,000-scale U.S. Geological Survey quadrangles; the maps identify wetland habitats classified using the Cowardin et al. (1979) system. Ile grid-sampling technique offers a reasonable alternative to more expensive and time-consuming techniques for quantifying NWI map information with a reasonable degree of accuracy and detail. Grid-sampled data are entered into the Spatial Analysis System, a microcomputer-based geographic information system published byrydac Technologies Incorporated, for processing and manipulation. Digitized estuary boundaries and other study area boundaries can be intersected with grid-sampled data to produce acreage summaries and color maps for specific units of interest. Grid sampling of all 5,290 NWI maps available in coastal areas was completed in October 1989. The coastal wetlands inventory is one of several habitat elements in NOAKs program to develop a national estuarine assessment capability. These habitat elements will be incorporated into NOANs National Estuarine Inventory Program along with other physical, hydrological, biological, and economic information to provide a more comprehensive understanding of estuarine environments. The coastal wetlands project was initiated by Ocean Service, and the Beaufort Laboratory of the the National Oceanic and Atmospheric Adminis- Southeast Fisheries Center, National Marine tration (NOAA) in June 1986 and is being con- Fisheries Service (NMFS). ducted jointly by the Strategic Assessment Branch The coastal wetlands project is developing a of the Ocean Assessment Division of the Office of comprehensive and consistent national coastal Oceanography and Marine Assessment, National wetlands data base to increase our knowledge of 40 BioLoGicAL REPoirr 90(18) the distribution and areal extent of wetlands and The data base and assessment capability under to improve our understanding and management of development for the NEI are part of a dynamic and this vital resource. The data developed from this evolving process. Other estuaries and subestuar- project eventually will be incorporated into ies have been added to the NEI from around the NOAA's National Estuarine Inventory (NEI) and country. Refinements are being made to physical will be used with other information, such as land and hydrological data estimated in Volume 1. At- use, coastal pollution, distribution of estuarine tributes such as volume and flushing rates have fishes and invertebrates, and the status of classi- been added to the data base. Other NOAA projects fied shellfish waters, to develop a national estua- for which data and information will be included in rine assessment capability. the NEI are the distribution of estuarine-depen- The cornerstone of the NEI is the National dent living marine resources; characterization of Estuarine Inventory Data Atlas. Volume 1, com_ estuarine shoreline modification, navigational pleted in 1985, identifies 92 of the most important channels, and dredged material disposal areas; the quality of shellfish-growing waters and related estuaries of the conterminous United States, and projects; the National Coastal Pollutant Discharge presents information through maps and tables on Inventory; and the Inventory of Outdoor Coastal physical and hydrological characteristics of each. Recreation facilities These estuaries represent about 90% of the estu- arine water surface area and 90% of the freshwa- ter inflow to estuaries of the east coast, west coast, Introduction and Gulf of Mexico. Volume 2, Land Use, presents area estimates for 7 categories and 24 subcategor- The Nation@s coastal wetlands are an important ies of land use, as well as 1970 and 1980 popula- natural resource. They provide critical habitat for tion estimates. Volume 3, Coastal Wetlands-New fish, shellfish, and wildlife (Shaw and Fredine England, is the first atlas in the wetlands series. 1956; McHugh 1966; Turner 1977; Flake 1979; This volume presents acreage estimates for 15 Lindal and Thayer 1982; Sather and Smith 1984), habitat types in 16 estuaries and 42 counties from filter and process agricultural and industrial Maine to Connecticut. Volume 4, Public Recre- wastes (Kadlec and Kadlec 1979; Tchobanoglous ation Facilities in Coastal Areas, presents data for and Culp 1980; Benner et al. 1982), and buffer Federal, State, and locally owned recreation facil- coastal areas against storm and wave damage ities in 327 counties and 25 estuary groups. (Knutson 1982). They also generate large revenues The goal of the NEI is to build a comprehensive from recreational activities such as fishing and framework for evaluating the health and status of hunting (National Marine Fisheries Service 1981; the Nation!s estuaries, and to bring estuaries into U.S. Fish and Wildlife Service 1982). focus as a national resource base. The principal Rapid loss of wetlands is occurring in many spatial unit for which all data are organized is the areas because of urbanization, agriculture, hydro- "estuarine drainage area," which is defined as carbon exploration, sea level rise, shoreline ero- "that land and water component of an entire wa- sion, and other factors. More than 11 million acres tershed that most directly affects an estuary" (Na- of wetlands have been lost over the past 25 years tional Oceanic and Atmospheric Administration (Frayer et al. 1983) because of human activity and 1985). These data will be used to make comp i natural processes. Although most of the losses arl- have occurred in inland areas, coastal wetlands sons, rankings, statistical correlations, and other have also declined at an alarmi rate over this analyses related to resource use, environmental 11, period (about 20,000 acres [31 mi ] per year). Re- quality, and economic values among estuaries. The cent rates of wetland loss may be even higher in main tool for performing these analyses will be some States. For example, in coastal Louisiana, NOAA!s GeoCOAST facility. GeoCOAST is a hard- losses are estimated at 32,000 acres (50 ini 2) per ware and software information facility developed year (Day et al. 1981). by NOAA. GeoCOAST software consists of both commercial geographic information systems and NOAA-written packages. GeoCOAST provides the 'Additional information on NOAA!s National Estuarine resources for developing and supporting systems Inventory is available from the Strategic Assessment Branch, Ocean Assessments Division, National Oceanic and used to store and analyze the spatial and temporal Atmospheric Administration, 6001 Executive Boulevard, relation of data in coastal areas. Rockville, Md. 20852, phone (301) 443-8843. NATioNAL PpocRAms 41 Despite increased awareness in the public and tive was to use wetland maps produced by the scientific sectors of the importance of coastal wet- National Wetlands Inventory (NWI) program of lands, no data base exists to document the current the U.S. Fish and Wildlife Service (FWS). distribution and abundance of coastal wetlands. Recognizing this gap in wetlands information, the National Marine Fisheries Service and the Stra- The National Wetlands tegic Assessment Branch undertook a cooperative Inventory Program effort to compile the first comprehensive and con- sistent coastal wetlands data base. We describe The NWI program was established in 1975 to these efforts and summarize the data compiled to generate scientific information on the characteris- date. tics and extent of the Nation's wetlands, and to provide data for making quick and accurate re- source decisions (Tiner 1984). This information Methods was developed in two stages: (1) the creation of detailed wetland maps, and (2) research on histor- Preliminary Investigations ical status and trends. Since 1975, the FWS has produced thousands of detailed wetland maps, cov- As a first step in establishing a coastal wetlands ering more than 60% of the conterminous United data base, NOAA examined and compiled existing States and over 98% of the coastal zone. The maps data on the areal extent and distribution of coastal are developed from aerial photography and are wetlands. Twenty-three sources were consulted in generally based on 1:24,000-scale U.S. Geological order to compile acreage figures for 242 counties Survey (USGS) maps. They illustrate wetland hab- in 22 coastal States (Alexander et al. 1986). These itats classified using the FWS's wetland classifica- data indicated that more than 11 million acres of tion system (Cowardin et al. 1979). wetlands exist along the coastline of the contermi- The NWI wetland maps represent the most re- nous United States. About 5 million acres were liable source of consistently derived coastal wet- identified as swamp, 4.4 million acres as salt land information available. However, fewer than marsh, 1.5 million acres as fresh marsh, and 0.2 2,000 of the more than 5,500 maps required for million acres as tidal flats. The Gulf of Mexico had complete coverage of the Natiori!s estuaries and the most wetlands (5.2 million acres), followed by other coastal areas have been converted to digital the Southeast (4.2 million acres), the Northeast data for computer processing and mapping. There- (1.7 million acres), and the west coast (0.2 million fore, only a fraction of the wetlands data required acres). Detailed information on data sources and a is available. Further, a complete digital data base complete table of wetland types and acreages by of NWI coastal maps is not anticipated by FWS. coastal county are presented in two appendixes to Since the current procedure for digitizing is expen- the inventory. sive and time-consuming, FWS presently digitizes While the compilation and evaluation of exist- maps primarily on a user-pays basis (Dahl 1987). ing data were necessary first steps in establishing NWI maps remained, however, the preferred a national coastal wetland data base, much exist- data source for this project because of their avail- ing information is incomplete or outdated. Vari- ability across broad coastal regions. For example, ability in data quality and consistency, and lack in the Gulf of Mexico region, 1,543 of about 1,850 of a unifying theme or purpose, also contributed maps needed for complete coverage of all coastal to the difficulty of consolidating data into a single, counties and 23 different estuarine systems are comprehensive data base. Therefore, our next step currently available from the FWS. Most maps not was to evaluate alternative sources of informa- yet available are from areas not generally consid- tion. A key consideration was the ability to de- ered coastal. velop a data base in a timely and cost-effective manner. Evaluating Grid-sampling Techniques Some investigators have successfully used multispectral scanner (MS S) and thematic mapper Preliminary tests that used a grid-sampling (TM) Landsat satellite imagery to inventory wet- technique on NWI maps indicated that this proce- land habitats (Haddad and Harris 1985; May dure could offer a reasonable alternative to more 1986). However, these techniques are beyond the expensive and time-consuming techniques for resources of the project. A more realistic alterna- quantifying NWI map information with a reason- 42 BioLoGicAL REPoRT 90(18) able degree of accuracy and detail (Field et al. and recommendations from the workshop partic- 1988). To test this procedure, a simple grid-sam- ipants on NOAA!s proposed grid-sampling project. pling technique was used to quantify habitat types In general, workshop participants supported for 16 previously digitized, 1:24,000-scale NWI NOAA!s proposal to grid-sample NWI maps (Na- maps. For the preliminary tests, the numerous tional Oceanic and Atmospheric Administration habitat types designated on the NWI maps were 1986). Participants suggested, however, that the aggregated into six general habitat categories: technique should be modified to improve the qual- salt marsh, fresh marsh, tidal flats, swamp, open ity and usefulness of the data being developed. water, and uplands. After some testing, we deter- Two key reconimendations were proposed: mined that a 45-acre grid cell size was both effi- Expand the number of habitat types re- cient and accurate for estimating these six habitat corded. Participants said they believed that types at this scale. We sampled each map sepa- the six habitat types identified in the prelim- rately by mounting a mylar grid sheet over the inary tests were inadequate and suggested a map and systematically recording the habitat list of 11 habitat categories (Table 2). Since type at each sampling point. The sampling took the workshop, 15 habitats have been incorpo- about 1 h. Based on the results (Table 1), it ap- rated into the project. peared that grid sampling could provide a time- ID Conduct a more complete statistical evalua- and cost-effective technique for compiling a rea- tion of the grid-sampling procedure. sonably accurate coastal wetlands data base. These recommendations were examined by NOAA and incorporated into the operational NOANs Coastal Wetlands Workshop phase of the project. The current grid-sampling technique is explained in detail in the following Before embarking on a national grid-sampling section. Grid sampling of available NWI maps effort, the Strategic Assessment Branch and the began in June 1986. National Marine Fisheries Service organized a workshop for professionals with experience in wetlands mapping and management to discuss Table 2. The 15 habitat types identified in the grid- NOAA's proposal to compile a national coastal sampling procedure. wetlands data base. Sixteen professionals from six Federal organizations participated: U.S. Envi- Salt marsh ronmental Protection Agency, U.S. Army Corps of Brackish Engineers, USGS, FWS, National Marine Fisher- High ies Service, and the National Ocean Service. Spe- Low cific objectives of the workshop were to review Unspecified' current information on the distribution and ex- Fresh marsh tent of coastal wetlands and to solicit comments Nontidal Tidal Unspecified' Table 1. Grid-sampling results for two test areas in Forested and scrub-shrub coastal Louisiana and Texas (acres x 100.) Estuarine Nontidal fresh Difference Tidal fresh Habitat Digital Grid % Unspecified fresh' Salt marsh 976 972 _<1 Tidal flats Fresh marsh 176 179 +2 Forested and Nonfresh open water scrub-shrub 12 11 -8 Fresh open water Tidal flats 80 79 -1 Open Water 4,349 4,320 -1 Upland Upland 1,092 1,084 -1 'The unspecified categories were added to accommodate areas Total 6,685 6,645 -1 for which more specific information on salinity and water regime was not available. NATiONAL PRwRAms 43 The Grid-sampling Procedure and the acreage of wetland types as identified in the grid-sampling process. The SPANS map index- The grid-sampling technique used to quantify ing module will be a valuable tool in data-base coastal wetlands involves the placement of a trans- management, and for many modeling functions to parent grid over an NWI map, and the identifica- look at the distribution and abundance of coastal tion of the wetland type on which each sampling wetlands on a national scale. point falls. The grid cells used in this procedure are 0.7 inches on a side, corresponding to about 45 Inte7preting the Data acres when used on a 1:24,000-scale map. A small dot in the center of each grid cell is used as the Although the data used to compile this volume sampling point. The exact number of sampling are the most complete and up-to-date available for points varies with latitude and may contain be- the Nation's coastal regions, two major factors tween 800 and 1,000 sampling points. must be considered when interpreting the data: Before sampling, the map name, State, scale, (1) the limitations of the sampling technique, and date of aerial photography, latitude and longitude (2) the age of the photography used to produce the of the lower right and upper left comers, and the NWI maps. number of columns and rows of grid cells are re- corded. For this technique, the numerous wetland Technique Limitations types identified on NWI maps were aggregated As a result of discussions at NOAXs Coastal into 15 habitat types (Table 2). Each cell is re- Wetlands Workshop, representatives from the corded as the habitat type on which its center dot USGS's National Mapping Division aided NOAA!s falls. A quality-control procedure is used to mini- wetlands team in determining if the 45-acre resolu- mize the types of errors inherent in this technique. tion was adequate for capturing coastal wetlands Coastal counties were grid-sampled to the extent acreage with a reasonable degree of accuracy. Equa- of NWI map availability. Noncoastal counties were tions to determine acceptable sample size were cal- grid-sampled to the extent of NWI map availability culated at several levels of acceptable error and for that portion of the county intersecting estua- degrees of confidence. These calculations indicated rine drainage area boundaries as defined in Vol- that the 45-acre cell size and subsequent 800 plus ume 1 of NOAA's National Estuarine Inventory sampling points per 1:24,000-scale map were ade- (National Oceanic and Atmospheric Administra- quate for the development of wetlands data at the tion 1985). national, regional, and estuarine level of analysis. Grid-sampled data, however, are not intended to Geographic Information System be sufficiently accurate for making decisions at the Framework site-specific level. In addition, the data are not intended to accurately estimate rare habitat types. Grid-sampled data are entered into the Spatial But when these data are aggregated across a large Analysis System (SPANS), a microcomputer- geographic area, such as an estuary, they do pro- based geographic information system (GIS) pub- vide an accurate summary of the general distribu- lished by Tydac Technologies, Inc. SPANS allows tion and abundance of major wetland types. for easy loading and manipulation of grid-sampled Age of Photography data, and displays and calculates acreage totals for the habitats found on each map. Hard copies are The date of aerial photography for the maps produced using a color ink-jet printer or a color wax used in this study ranged from 1971 to 1985. The transfer printer. Wetland acreage and map sum- photography age also varied between regions. In maries can be produced by NWI map, county, New England, about 60% of the maps were pro- State, or estuary. duced from photography taken from 1975 to 1977, The newest and one of the most useful aspects while about 20% of the maps were produced from of the wetlands GIS capability is the SPANS map 1980 and 1981 photography. The mid-Atlantic indexing module. The SPANS map indexing mod- dates were slightly more recent, with 32% of the ule is a GIS that has a level of resolution based on maps produced from photography taken from 1:24,000-scale maps as identified in the USGS 1975 to 1978, and 43% from photography taken topographic series. A multitude of information can from 1979, to 1985. The photography for the Gulf be entered for each map, including location Oati- of Mexico was generally the most recent, with 28% tude and longitude), date of aerial photography, taken from 1979, and 42% from 1980 to 1984. 44 BioLoGicAL RFPoRT 90(18) Analysis of these data is difficult because of the that Louisiana contains more salt marsh than all photography date range and because of a lack of States in New England, the n-iid-Atlantic, and the regional data of comprehensive trends. As men- Gulf of Mexico combined, accounting for 53% of the tioned previously, losses in coastal Louisiana may salt marsh in these three regions. The Gulf coast of be as high as 32,000 acres (50 mi 2) per year (Day Florida is also extremely important, especially in et al. 1981). However, in New Jersey, after 1970, terms of forested wetlands. About 5,032,100 acres about 50 acres of tidal wetlands were lost annu- of forested wetlands are on Florida's Gulf coast, ally (Tiner 1985a). Likewise, in Delaware, from accounting for 47% of the forested wetlands and 1973 to 1979, about 20 acres of tidal wetlands 28% of the total wetlands in the three regions were lost annually (Tiner 1985b). Because na- inventoried. Florida's Gulf coast also contains tional trends indicate that the abundance of most 1,405,600 acres of fresh marsh, accounting for 50% wetland types is still declining (F!rayer et al. of the fresh marsh in the three regions inventoried. 1983), the wetlands data presented in this report Table 4 summarizes data for 47 estuarine drain- may represent more than the current amount of age areas in New England, the mid-Atlantic, and coastal wetlands. the Gulf of Mexico. As with the State data given previously, it is impossible to have a complete discussion on the distribution of coastal wetlands Results and Discussion in estuaries without data from the west coast and the southeastern coast. Once again, however, cer- Grid sampling of all 5,290 NWI maps available tain estuaries stand out, particularly the Missis- in coastal areas was completed in October 1989. sippi Delta region and the Ten Thousand Island The figure illustrates the extent ofNWI map ava- estuary on the southwest coast of Florida, ranked ability for coastal areas in the conterminous number one and two, respectively, in amount of United States. To date, data have been compiled total wetlands. With the exception of the Missis- by coastal county and estuarine drainage areas for sippi Delta region estuary, the Ten Thousand Is- 412 maps in New England (Maine to Connecticut), land estuary contains nearly one million acres 735 maps in the mid-Atlantic (New York to Vir- more of wetlands than any other estuary. This ginia), and 1,543 maps in the Gulf of Mexico (Flor- estuary is also ranked number one in amount of ida to Texas). Data for the west coast (Washington forested wetlands and fresh marsh. The Missis- to California) and the southeastern coast (North sippi Delta region estuary contains over one mil- Carolina to Florida) have been processed and lion acres of salt marsh, which is three times the should be available by fall 1990. amount of the Ten Thousand Island estuary. It is Table 3 summa izes data by State for New also ranked second in amount of fresh marsh. England, the mid-Atlantic, and the Gulf of Mex- ico. In these regions, 92.2 million acres of land were inventoried using the grid-sampling process. Comparisons with Fish and Wildlife Of this land, about 19%, or 17.7 million acres, was Service Data ,identified as wetlands. Forested wetlands were the dominant wetland type, accounting for 60% To monitor the effectiveness of the grid-sam- (10.7 million acres) of the total wetlands, followed pling technique, grid-sampled data are compared by salt marsh (18%, 3.3 million acres), fresh with NWI digital data whenever these data are marsh (16%,2.8 million acres), and tidal Rats (5%, available. While there are no complete digital data 0.9 million acres). bases available for any Gulf coast state, the NWI A complete discussion on the distribution and has digitized an area approximately two to three abundance of coastal wetlands on a national scale maps in from the coast for most of the region. will not be possible until data for the west coast and Digital data were obtained for five areas and com- the southeastern coast have been processed. How- pared to grid-sampled data (Table 5). These data ever, simple analysis of data for New England, the were developed by the FWS by using the Map mid-Atlantic, and the Gulf of Mexico reveals the Overlay Statistical System (MOSS). usefulness of these data as an indicator of the These data indicate that abundant wetland distribution and abundance of coastal wetlands on types, such as salt marsh in Galveston Bay and a regional scale. For example, the importance of Laguna Madre, are estimated extremely well, Louisiana@s extensive salt marshes has been recog- while estimates for rare wetland types, such as nized for a long time. Grid-sampled data indicate forested wetlands in Galveston Bay, are some- Figure. National Wetlands Inventory map availability in coastal areas of the conterminous United States. 46 BioLocicAL REPoRT 90(18) times close to digital estimates but are generally to be published in the fall of 1990. A report de- more variable. scribing the wetlands in 127 counties and 8 estu- aries in the mid-Atlantic region (New York to Product Schedule Virginia) was distributed in May 1990, and a report on wetlands of the west coast was distrib- An atlas describing the distribution and abun- uted in summer 1990. Data for the Southeast dance of coastal wetlands in 42 counties and 16 region (North Carolina to Florida) will be included estuaries of the Northeast region (Maine to Con- in a national summary report (being done in coop- necticut) has been published, and another atlas eration with the FWS's National Wetlands Re- containing data from 157 counties and 23 estuar- search Center and the NWI) scheduled for comple- ies in the Gulf of Mexico (Texas to Florida) is due tion in April 1991. Table 3. Coastal wetlands by state (Acres x 100) Wetlands Non-Wetlands State Soft Fresh Forested & Tidal Total Open Total Marsh Marsh Scrub-Shrub Flats Wetlands Water Upland Subtotal Acreage Maine 215 (3) 251 (3) 6,085 (80) 1,038 (14) 7,589 (15)b 6,246 37,349 43,598 69,103 New Hampshire 56 (7) 53 (6) 694 (81) 57 (6) 860 (10) 258 7.160 7,418 8,278 Massachusetts 471 (10) 315 (7) 3,521 (75) 382 (8) 4,689 (12) 2,291 32,394 34.685 39,374 Rhode Island 38 (6) 15 (2) 564 (85) 43 (7) 660 (8) 1,069 6,209 7.278 7,938 Connecticut 126 (8) 106 (7) 1,236 (81) 57 (6) 1,525 (5) 1.196 28,558 29,754 31,279 New York 297 (27) 59 (5) 345 (32) 392 (36) 1,092 (5) 2,280 17,152 19,433 20,525 New Jersey 1,936 (24) 325 (4) 5,283 (65) 641 (8) 8,186 (18) 1,755 36.182 37,938 46.123 Pennsylvania 0 (0) 48 (24) 155 (76) <1 (<1) 204 (2) 271 10,849 11,120 11.324 Delaware 867 (39) 96 (4) 1.240 (55) 43 (2) 2,245 (17) 636 10,405 11,041 13,286 Maryland 2,042 (38) 158 (3) 2,896 (54) 299 (6) 5,394 (11) 4,074 40,820 44,894 50.288 District of Columbia 0 (0) 2 (50) 1 (50) 0(0) 3 (<I) 44 470 514 517 Virginia 1,685 (23) 370 (5) 4,243 (58) 1,052 (14) 7,352 (8) 4.954 79.441 84,395 91747 Georgia 0 (0) 23 (6) 334 (94) 0(0) 357 (13) 50 2,288 2,338 2,695 Florida 2,542 (4) 14,056 (20) 50,321 (73) 1.930 (3) 68,849 (35) 11,638 118.199 129.837 198,686 Alabama 255 (2) 144 (1) 10,276 (96) 41 (1) 10,716 (24) 1.021 32.585 33,606 44,322 Mississippi 588 (8) 105 (1) 6,481 (90) 23 (1) 7.197 (25) 626 20.703 21,329 28.526 Louisiana 17,228 (52) 6,770 (20) 9,142 (27) 319 (1) 33.459 (47) 20,171 18,284 38,455 71,914 Texas 4,320 (26) 5,343 (32) 4,211 (25) 2,751 (17) 16.625 (8) 15,638 172,222 187,860 204.485 a Values in parentheses represent the percent of total wetlands b Values in parentheses represent the percent of total acreage NA,nONAL PRoGRAms 47 Table 4. Coastal wetlands by estuarine drainage area (Acres x 100) Wetlands Non-Wetlands Estuary Salt Fresh Forested & Tidal Total Open Total Marsh Marsh Scrub Shrub Flats Wetlands Water Upland Subtotal Acreage Passamaquoddy Bay 10 (1)a 51 (3) 1386(90) 80 (6) 1527(18) b 968 5,921 6,889 8,416 Englishman Bay 15 (1) 37 (3) 981 (87) 104 (9) 1,137(18) 747 4,436 5,183 6,320 Narraguagus Bay 23 (4) 4 (1) 451 (79) 93(16) 572(20) 453 1,898 2,351 2,923 Blue Hill Bay 2 (<1) 16 (3) 486(84) 73(13) 577(11) 1,073 3,611 4,684 5,261 Penobscott Bay 10 (1) 28 (3) 775(79) 166(17) 979(11) 2,250 5,418 7,669 8,647 Muscongus Bay 2 (2) 1 (2) 58(64) 29(32) 90(16) 68 422 490 580 Sheepscot Bay 50(16) 28 (9) 117(37) 119(38) 314(11) 581 2,003 2,584 2,898 Casco Bay 24 (6) 15 (4) 167(43) 186(47) 393 (8) 1,287 3,211 4,498 4,891 Saco Bay 29 (6) 40 (8) 413(82) 18 (4) 500 (9) 327 4,606 4,934 5,433 Great Bay 27 (5) 19 (4) 396(78) 67(13) 509(13) 168 3,165 3,333 3,842 Merrimack River 23 (4) 48 (8) 535(86) 11 (2) 617 (9) 214 5,809 6,024 6,640 Boston Bay 18 (4) 37 (9) 305(69) 79(18) 439(10) 522 3,547 4,069 4,508 Cape Cod Bay 106(23) 25 (5) 91 (20) 241 (52) 463(19) 832 1,153 1,985 2,448 Buzzards Bay 41 (9) 82(17) 312(64) 48(10) 483(13) 1,468 1,871 3,339 3,822 Narrangansett Bay 38 (4) 62 (6) 864(88) 24 (2) 988(11) 1,290 6,424 7,714 8,702 Connecticut River 31 (8) 43 (12) 289(79) 4 (1) 367 (5) 236 6,523 6,759 7,126 Gardiners Bay 33(24) 3 (2) 30(21) 74(53) 141 (5) 1,139 1,633 2,771 2,912 Long Island Sound 161 (8) 116 (6) 1,586(79) 153 (8) 2,016 (5) 5,793 35,498 41,291 43,307 Great South Bay 183(41) 2 (0) 44(10) 219(49) 447 (8) 893 4,021 4,914 5,362 Hudson River 168(10) 147 (8) 1,243 (72) 162 (9) 1,719 (7) 1,791 21,302 23,093 24,812 Barnegat Bay 416(17) 35 (1) 1,710(70) 299(12) 2,460(29) 547 5,524 6,070 8,530 Delaware Bay 1,472(36) 241 (6) 2,202(54) 187 (4) 4,102(14) 3,561 20,920 24,481 28,583 Chincoteagae Bay 249(68) 2 (1) 73(20) 44(12) 368(18) 800 901 1,700 2,068 Chesapeake Bay 2,779(28) 508 (5) 5,685(57) 990(10) 9,962 (7) 23,116 105,290 128,410 138,368 Ten Thousand Islands 548 (2) 8,076(37) 12,616(58) 409 (2) 21,650(76) 1,004 5,644 6,688 28,338 Charlotte Harbor 68 (1) 2,896(46) 2,713(44) 562 (9) 6,240(20) 2,259 22,181 24,440 30,680 Tampa Bay 31 (1) 466(18) 1,647(65) 376(15) 2,520(16) 2,323 10,817 13,140 15,660 Suwanee Bay 209 (9) 176 (8) 1,902(83) 3 (<1) 2,290(20) 455 8,419 8,874 11,164 Apalachee Bay 244 (4) 254 (4) 6,368(92) 88 (1) 6,954(32) 1,300 13,553 14,853 21,807 Apalachicola Bay 170 (3) 87 (2) 5,585(94) 75 (1) 5,917(50) 1,581 4,168 5,749 11,666 St. Andrew Bay 85 (3) 28 (1) 2,362(94) 35 (1) 2,511 (33) 679 4,318 4,997 7,508 Choctawhatchee Bay 27 (1) 37 (1) 2,679(96) 58 (2) 2,801 (21) 975 9,384 10,359 13,160 Pensacola Bay 67 (3) 61 (2) 2,297(94) 20 (1) 2,445(19) 1,001 9,199 10,200 12,645 Perdido Bay 19 (1) 18 (1) 1,657(97) 7 (<1) 1,702(22) 324 5,671 5,995 7,697 Mobile Bay 170 (3) 72 (1) 6,273(96) 30 (<1) 6,545(23) 2,882 19,122 22,004 28,549 Mississippi Sound 1,706(16) 432 (4) 8,477(79) 74 (1) 10,689(22) 11,057 21,112 32,169 42,858 Mississippi Delta Region 10,429 (59) 3,325(19) 3,788(21) 151 (1) 17,693(40) 21,256 2,564 23,820 41,513 Atchafalaya and 1,265(27) 1,026(22) 2,304(50) 19 (<1) 4,614(30) 4,312 4,037 8,349 12,963 Vermillion Bays Calcasieu Lake 826(68) 329(27) <1 (<1) 65 (5) 1,220(33) 1,375 1,127 2,502 3,722 Sabine Lake 1,100(28) 852(22) 1,871 (47) 114 (3) 3,937(19) 1,374 15,710 17,084 21,021 Galveston Bay 949(40) 589(25) 744(31) 110 (5) 2,392(11) 4,258 15,495 19,753 22,145 Brazos River 3 (2) 68(34) 126(64) <1 (<1) 197 (3) 234 5,661 5,895 6,092 Matagorda Bay 435(51) 289(34) 70 (8) 64 (8) 858 (5) 2,181 15,105 17,286 18,144 San Antonio Bay 328(49) 283(42) 22 (3) 35 (5) 668(20) 1,484 1,118 2,602 3,270 Aransas Bay 307(32) 420(43) 108(11) 139(14) 974 (6) 1,623 12,992 14,615 15,589 Corpus Christi Say 122(41) 73(25) 14 (5) 87(29) 296 (4) 1,533 5,376 6,909 7,205 Laguna Madre 678(15) 1,933(43) 226 (5) 1,668(37) 4,506 (6) 3,776 60,821 64,597 69,103 a Values in parentheses represent the percent of total wetlands grid sampled by NOAA b Values in parentheses represent the percent of total estuarine drainage area grid sampled by NOAA that is wetlands Table 5. NOAA grid sampled data vs U.S. Fish and Wildlife Service digital data for five estuaries in the Gulf of Mexico. P;@. Aggregates of ten to 15 maps were compared in each estuary. 00 W Region Salt Marsh Fresh Marsh Forested Tidal Flats Total Wetlands Upland Open Water 5 Mobile Bay NOAA 11,075 1,382 43,244 3,594 59,295 59,655 239,074 Fish & Wildlife 11,047 1,340 44,585 3,211 60,183 60,699 238,786 % Difference 0.3 3.1 -3.0 11.9 -1.5 -1.7 0.1 Tampa Bay NOAA 1,839 5,348 14,259 29,445 50,901 65,166 156,208 Fish & Wildlife 1,580 4,577 15,341 28,361 49,859 65,247 157,584 % Difference 16.4 16.8 -7.1 3.8 2.1 -0.1 -0.9 Mississippi Delta NOAA 29,512 65,198 5,169 765 100,644 100,734 511,884 Fish & Wildlife 29,930 65,666 5,326 727 101,649 101,870 510,727 % Difference -1.4 -0.7 -2.9 5.2 -1.0 -1.1 0.2 Galveston Bay NOAA 78,557 9,592 315 8,139 96,603 96,783 275,158 Fish & Wildlife 77,644 9,296 488 7,801 95,229 95,402 267,040 % Difference -1.2 -3.2 35.5 -4.3 1.4 -1.4 -3.0 Laguna Madre NOAA 31,762 22,922 193 97,469 152,346 152,751 143,907 Fish & Wildlife 31,204 23,508 229 97,588 152,529 152,756 143,876 % Difference -1.8 2.5 15.7 0.1 -0.1 0.0 0.0 NATIONAL Pfior_mms 49 References Lindall, W. N., Jr., and G. W. Thayer. 1982. Quantification of National Marine Fisheries Service Alexander, C.E., M. A. Broutman, and D. W Field. 1986. habitat conservation efforts in the Southeast region An inventory of coastal wetlands of the USA. of the United States. Mar. Fish. Rev. 44:18-22. Strategic Assessment Branch, National Oceanic and May, L. N., Jr. 1986. An evaluation of Landsat MSS Atmospheric Administration, Rockville, Md. 25 pp. digital data for updating habitat maps of the [unpublished manuscript] Louisiana coastal zone. Photogram. Eng. Remote Benner, C. S., R L. Knutson, R. A. Brochu, and A. K. Sens. (52)8:1147-1159. Hurme. 1982. Vegetative erosion control in an McHugh, J. L. 1966. Management of estuarine fisheries. obligolialine environment, Currituck Sound, North Am. Fish. Soc. Spec. Publ. 3:133-154. Washington, Carolina. In Third annual meeting of the Society of D.C. Wetland Scientists, Wrightsville Beach, N.C. National Marine Fisheries Service. 1981. Fisheries of Cowardin, L. M., V Carter, F C. Golet, and E. T La. Roe. the United States, 1980. Current fishery statistics 1979. Classification of wetlands and deepwater 8100, U.S. Department of Commerce, National habitats of the United States. U.S. Fish Wildl. Serv., Oceanic and Atmospheric Administration, FWS/OBS-79/31.103 pp. Washington, D.C. Washington, D.C. U.S. Government Printing Office. Dahl, T. E. 1987. Wetlands mapping in the coastal 132 pp. zone-progress towards creating a national data National Oceanic and Atmospheric Administration. base. Pages 465-473 in Proceedings of coastal zone 1985. National estuarine inventory data atlas: '87. Vol. 1. Seattle, Wash. physical and hydrologic characteristics. Office of Day, J. W, Jr., and N. J. Craig. 1981. Comparison of Oceanography and Marine Assessment, Strategic effectiveness of management options for wetlands Assessment Branch. Rockville, Md. 103 pp. loss in the coastal zone of Louisiana. Pages 232-239 National Oceanic and Atmospheric Administration. in Proceedings, conference on coastal erosion and 1986. Summary of proceedings: NOAA coastal wetland modification in Louisiana: causes, wetlands workshop. Office of Oceanography and consequences, and options. U.S. Fish Wildl. Serv., Marine Assessment, Strategic Assessment Branch, FWSVOBS-82159. Rockville, Md. 11 pp. [unpublished manuscript] Field, D. W, C. E. Alexander, and M. A. Broutman. 1988. Sather, J. H., and R. D. Smith. 1984. An overview of Towards developing an inventory of coastal wetlands major wetland values and functions. U.S. Fish Wildl. of the USA. Mar. Fish. Rev. 50(l):40-46. Serv., FWSVOBS-84/18.68 pp. Flake, L. D. 1979. Wetland diversity and waterfowl. Shaw, S. P, and C. G. Fredine. 1956. Wetlands of the Pages 312-319 in P E. Greeson, J. R. Clark, and J. E. United States: their extent and their value to Clark, eds. Wetland functions and values: the state of waterfowl and other wildlife. U.S. Fish Wildl. Serv., our understanding. American Water Resources Circ. 39.67 pp. Association. Minneapolis, Minn. Tchobanoglous, G., and G. L. Culp. 1980. Wetland Frayer, W E., T. J. Monahan, D. C. Bowden, and F A. systems of wastewater treatment: an engineering Graybi,l. 1983. Status and trends of wetlands and assessment. University of California, Davis. deepwater habitats in the conterminous United Tiner, R. W , Jr. 1984. Wetlands of the United States: States, 1950's to 1970's. Department of Forest and current status and recent trends. U.S. Fish and Wood Sciences, Colorado State University, Fort Wildlife Service, National Wetlands Inventory, Collins. 32 pp. Washington, D.C. 59 pp. Haddad, K. D., and B. A. Harris. 1985. Use of remote Tiner, R. W, Jr. 1985a. Wetlands ofNewJersey. U.S. Fish sensing to assess estuarine habitats. Pages 662-675 and Wildlife Service, National Wetlands Inventory, in 0. T Magoon, ed. Coastal zone '85. Proceedings of Newton Corner, Mass. 117 pp. the fifth symposium on coastal and ocean Tiner, R. W, Jr. 1985b. Wetlands of Delaware. U.S. Fish management Vol 1. American Society of Civil and Wildlife Service, National Wetlands Inventory, Engineers, New York. Newton Corner, Mass., and Delaware Department of Kadlec, R. H., and J. A. Kadlec. 1979. Wetlands and Natural Resources and Environmental Control, water quality. Pages 436-456 in P E. Greeson, J. R. Wetlands Section, Dover, Del. Coop. Publ. 77 pp. Clark, and J. E. Clark, eds. Wetland functions and Turner, R. E. 1977. Intertidal vegetation and values: the state of our understanding. American commercial yield ofpenaeid shrimp. Trans. Am. Fish. Resources Association. Minneapolis, Minn. Soc. 106:411-416. Knutson, R L., and W N. Selig. 1982. Wave damping in U.S. Fish and Wildlife Service. 1982. The national Spartina alternaflora marches. In Third annual survey of fish, hunting, and wildlife-associated meeting of wetlands scientists. Wrightsville recreation. U.S. Government Printing Office, Beach, N.C. Washington, D.C. 132 pp. NATiONAL PRwmms 51 Overview of the Land-Sea Interface Research Program by Armond T Joyce, Richard L. Miller, and Ramona E. Pelletier Science and Technology Laboratory Building 1100 National Aeronautic and Space Administration Stennis Space Center, Mississippi 39529 ABSTRACT.-The general goal of the Earth Sciences Research Program at the NASA/Stennis Space Center Science and Technology Laboratory is to provide a better understanding of the state and dynamics of global biological, chemical, and physical processes under natural and anthropogenic perturbations. Our research is conducted by using remotely sensed data acquired by a variety of sensors operated on a truck boom, aircraft, and spacecraft. Although some studies are site-specific, our overall objective is to gain information and knowledge that would allow modeling from a global perspective. Research is conducted through a team approach with a multidisciplinary staff. We give preference to developing joint research projects with university faculty or other external investigators in order to form an appropriate team for each particular research objective. Collaborative research with external investigators is also aided through the Resident Research Associateship Program administered through the National Research Council, the Summer Faculty Program, and a Summer Visiting Scientist-Lecturer ]Program. The Stennis Space Center Science and Technology Laboratory (SSQ/STL) Research Program has three main focuses: forest ecosystems research, land-sea interface, and soils and geological research. This research takes place in irnany separate projects; however, each of these interrelated projects is a component of a research program aimed at understanding the functioning of the ecosystem being addressed. The Land-Sea Interface Research Program at SSQ/STL views the land-sea interface in an ecosystem context, including both the terrestrial component and the nearshore waters. Wetlands Productive Capacity trolled burning results in a watershed NAPP de- Modeling crease of 2 to 20% (6% average between 1978 and 1985). This work is described in further detail in This project combines remote-sensing analysis, the proceedings of the 1987 symposium of the field studies in the Grand Bayou, Louisiana, wa- Society of Wetland Scientists (Dow et al. 1989). tershed, and mathematical modeling to examine the coupling between the production of detritus (dead organic matter) in wetlands and the yield of Spectral Studies coastal fisheries. Thematic Mapper (TAD data Our past research efforts characterized the have been used successfully to estimate the stand- spectral curves exhibited by different species of ing crop biomass in the marsh through the use of marsh plants in the field, and measured the spec- regression equations. Two types of regression bml response of Spartina alterniflora grown in equations were employed; those that used raw hy droponic culture in the laboratory to salinity and digital counts and others that used ratioed data waterlogging stresses. The field studies of marsh from different bands (reflectance ratio and vegeta- plant spectra measured 6 replicates of live and tion index). Our remote-sensing analysis, com- dead leaves for 11 different species characteristic bined with field studies, suggests that wetland loss ofthe saline, brackish, intermediate, or freshwater results in a watershed net aboveground primary marshes at Grand Bayou, Louisiana. Spartina pa- production (NAPP) decrease of 4%, whereas con- tens was collected from the brackish and interme- 52 Biou)GicAL REPoRT 90(18) diate marsh salinity zones, while S. alterniflora Our results so far indicate that the analysis of was collected from the brackish and saline marsh TM data provides the most accurate classification zones. The marsh salinity zone does not seem to of pertinent vegetation types that are correlated change the characteristic spectra of S. alterniflora with methane emission (Dow et al. 1987). How- or S. patens. ever, the combination of cloudy weather and infre- In the laboratory studies of S. alterniflora grown quent TM coverage has led to an evaluation of the in hydroponic culture for 6 weeks at 5 salinity lev- suitability of Advanced Very High Resolution Ra- els (0, 6, 12, 24, and 35 ppt) and 3 different levels diometer (AVIIM) data to classify vegetation be- of waterlogging (drained, saturated, and sub- cause of its frequent coverage, despite its coarser merged), the rate of leaf expansion and photosyn- spatial resolution (1 km). Daytime AVHRR data thetic rate manifested stress responses at salinity had demonstrated some promise as an indicator of levels above 24 ppt. I. A. Mendelssohn and cowork- inundation extent (Pelletier and Dow 1989). We ers of the Laboratory for Wetland Soils and Sedi- are now evaluating both day and night acquisi- ments at Louisiana State University (LSU) made tions to better understand seasonal inundation the growth measurements and conducted assays variation, which is directly related to the anaerobic for leaf proline (a general indicator of plant salinity conditions necessary to methanogenesis and the or water stress). These data are being statistically production of other "greenhouse" gases. analyzed on an individual replicate basis in com- This project also addresses the following objec- parison with the moisture stress index (0/6 reflec- tives: (1) estimating the vertical extent of inunda- tance at 1650 ruilm/o reflectance at 1260 run). The tion by incorporating detailed topographic data wealth of growth, biochemical assay (proline levels, and ancillary precipitation and stage data; root alcoholic dehydrogenase levels), and elemental (2) monitoring dynamics of seasonal variation by concentrations measured for each treatment by modeling temporal changes in water flow patterns LSU researchers should provide a basis for inter- during a typical year; and (3) estimating potential preting the spectral response curves. minimum and maximum inundation states for selected years exhibiting extreme ranges between wet and dry seasons. In our initial modeling efforts, we developed Biogeochemical Flux spatial distributions of five methane flux classes Estimation for a typical wet season period and another for a typical dry season period. TM data were acquired Wetlands are the focus of our research because from a December overpass and classified into 13 of their production of trace gases of carbon, nitro- distinct vegetation classes. In situ methane flux gen, and sulfur, which may modify the earth's measurements were collected by using these cate- climate through the "Greenhouse Effect", and be- gories as a sampling guide. Based on these flux cause of the potential role of some trace gases in readings, the 13 vegetation classes were recom- controlling the stratospheric ozone layer, which bined into 5 statistically significant vegeta- protects the earth from ultraviolet radiation. The tion/flux classes. The inundation status was deter- wetlands biogeochemical flux project is a joint ef- mined from the vegetation/flux classification to fort being conducted by the Stennis Space Center represent dry season and wet season periods, and Science and Technology Laboratory, Langley Re- new methane flux distribution maps were devel- search Center, and the South Florida Research oped. These results are reported in ajournal article Laboratory at the Everglades National Park. (Bartlett et al. 1989). While our efforts to date have The objectives of this project are to (1) examine been focused in the Everglades, we plan to expand the capabilities of current remote-sensing instru- our research to larger geographic areas. ments to delineate certain wetland vegetation types, and (2) develop and test a geographic infor- mation system (GIS) for estimating trace gas emis- Airborne Electromagnetic sions from wetland ecosystems. Our past efforts Profiles Research focused on methane estimates; however, because the data-base requirements are quite similar for Many of today's typical airborne or spaceborne other trace gas estimation models, our future ef- remote-sensing devices are relatively surface- forts will include GIS model development for nitro- oriented. Because soils have a large component gen and sulfur trace gases as well. beneath the surface, many of these surface- NATioNAL PRoGRAms 53 oriented sensors are notably limited in character- strongly suggested that a much more significant izing properties of the whole soil. Surface condi- driving force in controlling the coastal sediment tions can often be determined and some subsur- budget is the passage of intense cold fronts during face conditions inferred, but the subsurface the winter months. While a hurricane is intense, condition inferences are not always reliable. it is relatively localized, only affecting a length of These concerns are exacerbated if the soil or sed- coast about 100 miles. Furthermore, for any given iment material is submerged, as is true in many piece of shoreline, the mean time between hurri- wetland and shallow coastal areas. In these situ- canes is about 33 years. In comparison, winter ations, not only is the subsurface inaccessible to storms occur 30 to 40 times every year, and sweep most remote sensors, but the soil surface is also over almost the entire coastal area. Also, the pro- obscured by a layer of water. cesses that occur during the passage of winter In this research, NASA is cooperating with the storms seem to maximize their effectiveness in Naval Oceanographic and Atmospheric Research moving sediment. Laboratory in evaluating the capabilities of the The purpose of our research is to examine the new Airborne Electromagnetic (AEM) Profiler. responses of the sediment, water, and atmosphere The AEM Profiler is an active helicopter-borne at and near the coast to the passage of these cold sensor working in the 90-4500 Hz range. The fronts. In cooperation with members of the Coastal system was primarily developed for mapping ba- Studies Institute, we hope to determine how im- thymetry in relatively shallow marine environ- portant these fronts are as "engines" for sediment ments, but it also is useful in assessing a number transport and deposition (Roberts et al. 1987). of physical properties for water and underlying Geometric registration of remotely sensed data sediment. Because the AEM Profiler has multi- is a major concern, as the features being studied frequency capability, these physical properties can have several spatial scales, change as a function be differentially determined at multiple depths of time, and are not shown on maps. This concern within the soil or sediment profile. Water proper- has led to the development of a fundamentally ties, such as salinity and temperature, and soil new georeferencing software system inside the information, such as bulk density, porosity, or- Earth Resources Laboratory Applications (ELAS) ganic matter content, and generalized mineralogy, software package. The design and the results of can be determined. testing this software are presented in Rickman Submerged sediment data can be valuable for et al. (1988). both commerical (e.g., monitoring sediment dy- namics in shipping channels) and scientific appli- cations (e.g., soil and sediment mapping, monitor- Wetlands Change Detection ing coastal geomorphology, and modeling marsh processes dependent on soil conditions and water Many of Louisiana@s wetlands are rapidly de- depth). This research program is trying to merge grading. Multitemporal remotely sensed data can AEM Profiler data with more traditional surface- be quite useful in identifying the location, amount, oriented sensor data. This will give three- and type of change taking place. The pattern of dimensional depictions of shallow, submerged change that often occurs in wetland environments, coastal areas. We have acquired and are analyzing such as those in Louisiana, consists of small scat- data from a barrier island environment at Cape tered spots that gradually grow in size, unlike the Lookout, North Carolina. Preliminary analysis concentric or corridor growth patterns generally from that data set is reported by Pelletier and Wu associated with urban change. Many traditional (1989). Additional AEM Profiler missions are means of detecting change in digital data have planned for the Louisiana coast, and an inland required that the data sets be geographically rec- wetland area in northeastern Canada. tified to a high degree of accuracy. In wetland environments undergoing significant change, high-confidence control points for adequate Coastal Geomorphology georegistration precision may be quite limited. The Science and Technology Laboratory has Previous work by the Coastal Studies Institute developed a change detection technique that does at LSU has raised questions about the relative not require the data sets to exhibit a strict degree significance of hurricanes in controlling the of geographic coregistration. Instead, our tech- coastal sediment budget. The institute's work has nique uses a gridding approach to partition the 54 BioLoricAL REPoRT 90(18) data into segments; these segments are classified rate of rise in eustatic sea level, the model predicts by land cover and then compared with geographi- that marshes previously considered "stable" are in cally similar segments in the second data set. danger of significant submergence during the next Although slight misalignment might cause signif- 50 to 100 years, even if the present rate of sea level icant overestimation of change when compared on rise is only doubled. If sea level rise quadruples a pixel by pixel basis, when grouped into grids or from the present rate (well within the range sug- multiple pixels the error is averaged out so that a gested by Hansen et al. 1985), sedimentation- more accurate estimate of real change can be cal- dominated aggrading marshes would be in danger culated. of submergence. We need more fieldwork and mod- The test data chosen for a preliminary evalua- eling activities to better understand the coping tion of this gridding technique were two Landsat mechanisms marshes have to counteract sea level Multispectral Scanner (MSS) data sets (1972 and fluctuations. 1981) for the Cameron-Creole watershed in Loui- Our research continues to improve the model siana. Test grid sizes ranged from 1 x 1 pixels to and to provide for a three-dimensional perspective 50 x 50 pixels. Grids from 5 x 5 pixels up to perhaps of multiple data transects in the model. We are 10 x 10 pixels provided calculated change esti- modifying the model to accept data from AEM mates from slightly to moderately misregistered Profiler studies as one source of input data. Ulti- data sets that were comparable in accuracy with mately, this model will be incorporated into geo- pixel by pixel calculations from almost perfectly graphical information systems of coastal regions, registered data sets. Grid sizes larger than along with remotely sensed and ancillary data 10 x 10 pixels tended to begin canceling out the sources for many biological and physical land-sea influence of relatively small spots of change. De- interface models. tails from this initial study are reported by Pellet- ier and Dow (1987). Land-Sea Interface Wetlands Landscape Modeling In 1986, Stennis Space Center Science and Technology Laboratory and the University of The Louisiana coast has many marsh condi- Puerto Rico (UPR) began a multiyear cooperative tions, including rapidly subsiding, aggrading, and research project to improve the understanding of those that seem to be in relative equilibrium. Be- exchange processes between terrestrial and ma- cause of Louisiana's variety of marsh conditions, rine ecosystems. Participating investigators are its coast has been an excellent location for model- faculty and graduate students from the Depart- ing the impact of changing sea level on a variety of ments of Marine Science and Engineering at the marsh conditions. We take theoretical and actual UPR Mayaguez Campus, and the staff of the Divi- cross sections of marsh landscapes from different sions of Terrestrial and Marine Ecology of the marsh types and modeled them for sea level effect Center for Energy and Environment Research. over various time intervals and rates of sea level During fiscal year (FY) 1988, much of our work rise. The key variables being monitored are hori- was devoted to enhancing the project's ability to zontal and vertical marsh topographic conditions, process and analyze ocean color imagery derived sedimentation rate, organic accretion rate, subsi- from the Coastal Zone Color Scanner (CZCS), the dence rate, toxic sulfide species concentrations, Airborne Ocean Color Imager (AOCI), and the and above- and belowground plant biomass. A Calibrated Airborne Multispectral Scanner spatial perspective for three-dimensional analyses (CAMS). A series of software modules was devel- of landscape change is permitted by extrapolating oped to process these data within the ELAS oper- between a series of marsh topographic transects. ating environment. In particular, an integrated set Results from the models provide an innovative of interactive modules was developed to compute means for visualizing how streamside segments near-surface chlorophyll concentrations with are capable of persisting for many years because CZCS data; these modules apply the clear water of higher sediment load and lower accumulation of radiance method for atmospheric correction, cor- toxic sulfide species when compared with the more rections for orbital and radiometric decay, and rapidly degrading back-marsh segments (Pelletier bio-optical algorithms. Atmospheric correction al- 1987). While subsidence-dominated marshes are gorithms and optical algorithms to compute chlo- in immediate danger of degrading at the present rophyll and suspended sediment concentrations NATioNAL Pw)GRAms 55 for AOCI and CAMS data are now being developed. investigators designed a program of measure- In addition, all ocean color software models are ments leading to the development of material bal- being transported to operate in the personal com- ances for water, salt, carbon, nitrogen, and phos- puter environment. phorous. All processes that contribute or remove Our FY88 activities focused on the large these materials will be evaluated. Automated tidal Guanajibo watershed (which discharges into wa- and steam gauging is being conducted in a joint ters along the west coast of Puerto Rico) and an project with the U.S. Geological Survey (USGS). intensive study site encompassing the Joyuda La- Groundwater flows from drilled wells are being goon within the Guanajibo watershed. We used assessed, and automated meteorological observa- if* are Landsat TM data to generate a-land cover class i- tiOns being made. Investigators@ took water cation for the entire Guanaj ibo watershed, and use samples to measure water mass chlorophyll con- CAMS data acquired during 1987 to generate a tent while acquiring data with the CAMS in March more detailed land cover classification of the 1987. Spectral data for mangrove leaves have been Joyuda Lagoon watershed. Soils maps and contour acquired with a ground-operated imaging spec- lines on topographic maps were digitized, and the trometr. This instrument measures reflectance resulting data were assembled, together with the from 0.38 to 2.5 microns in very narrow land cover data, in a geographically referenced bandwidths. The investigators made measure- data base. ments for mangrove leaves across salinity gradi- We used the soils, land cover, and topographic ents and at a site in the lagoon that has high data to develop a method for implementing the concentrations of nickel. The preliminary results Universal Soil Loss Equation (LJSLE) on a regional of these analyses were reported at the American scale for the Guanajibo area of western Puerto Institute of Biological Sciences Symposium in Rico. This task not only evaluated erosion condi- Davis, California (Lawrence 1988). tions of the land itself, but it also developed a baseline for assessing sediment effects on the coastal environment. Soil erosion from the moun- Phytoplankton Modeling tainous and agricultural regions within the This project integrated remotely sensed digital Guanajibo has contributed greatly to the sediment . influx in adjacent coastal regions. These soils are imagery of S O'uth San Francisco Bay (SSFB), Cal- inherently highly erodible, and the area's high ifornia, into a numerical model of seasonal and rainfall and increased agricultural pressure on the spatial phytoplankton dynamics. The model was land magnifies the erosion problem. Although the initially developed during a joint project with the resulting sediments delivered to the coastal waters USGS Water Resources Division in Menlo Park, bring some nutrients, they also tend to screen California. The specific objectives of this project much of the life-giving light from the area's phyto- are to (1) modify and refine model coefficients of plankton and coral. A good understanding transport and phytoplankton production in SSFB Of Po- by using both historical shipboard data and re- tential sediment load due to erosion from the ter- motely sensed ocean color data; (2) validate model restrial environment would be helpful to models of output with digital maps of near-surface chloro- coastal marine ecology. phyll concentrations derived from remotely sensed In order to address future land-sea interface data; and (3) develop ELAS modules to process and issues, we will continue to transform soil erosion analyze remotely sensed ocean color imagery. values from models such as the USLE into more The numerical model of SSFI3 follows the finite- likely values of actual sediment influx to the marine difference box model approach described by Officer environment on a variety of temporal scales. (1980) and Officer and Nichols (1980). The SSFB Another aspect of this coastal ecosystem study box model is a three-dimensional model containing focused on the amount and movement of organic both two-layer and lateral flow; the geometry of the carbon from terrestrial sources through estuaries boxes represents the average bathymetry at mean and lagoons. The principal study site for this proj- lower low water. Simulation parameters were cal- ect is the Joyuda Lagoon, a mangrove-fringed la- ibrated using shipboard data acquired by the goon on the west coast of Puerto Rico; the lagoon USGS during 1980 (Cloern 1984; Alpine and that is fed by a small watershed and exchanges Cloern 1988). During 1988 the SSFB box model with the sea through a narrow canal. The UPR and was transported to the IBM PC environment and Center for Energy and Environmental Research enhanced to provide an efficient user-interface 56 Biow=AL RF.Poirr 90(18) base on a windowing environment, fast execution, An October 1989 overflig lit of CAMS over the rapid modification of simiilation parameters, effi- Mississippi River Delta, Terrebonne Bay, and the cient storage and analysis of model output, and Atchafalaya Bay in Louisiana, provided the initial incorporation of different aquatic systems as sim- data set in which to meet the project objectives. A ulation environments. These improvements have coordinated field-sampling program consisted of established this model for the Science and Tech- ship surveys at all three sites. Continuous surface nology Laboratory as a generic modeling tool that profiles of in vivo fluorescence, suspended sedi- can be applied to various aquatic systems. ments ' temperature, and salinity were obtained Nine AOCI, six CZCS, and six TM Simulator using a flow-through system aboard the IW digital images of South San Francisco Bay have Pelican. In addition to all flow-through instru- been acquired. All images have been reformatted ments, other instruments interfaced to a micro- for processing within ELAS. Because the spectral computer collected continuous samples of ship and spatial characteristics of the AOCI were de- position and solar irradiance for continuous data signed specifically for ocean color analysis, our collection and archiving. The digital imagery has major effort has been to process and analyze the been georeferenced and coregistered to produced AOCI imagery. To date, each scene has been cali- large-scale mosaics of the study area. Data anal- brated and georeferenced for co-location with field ysis of both field and remotely sensed data is samples. We are still developing algorithms for underway. atmospheric correction. In large part, these algo- rithn-is are based on the clear water concept used for affecting atmospheric correction of CZCS data. Mississippi River Plume and Our analysis of remotely sensed data has provided Oceanographic Processes estimates of horizontal transport (vector displace- ment) and has indicated potential areas for the The primary objectives of this research are to initiation and development of phytoplankton (1) evaluate the information content of ocean color blooms. imagery acquired from CAMS and assess its po- The synoptic data of the CZCS and the AOCI tential for estimating surface chlorophyll concen- suggest that the blooms originate over the south- trations, suspended sediment concentrations, and eastern shoals and migrate northward along the elements of water quality (and develop atmo- eastern shoals. We are developing programs to spheric correction and bio-optical algorithms); estimate horizontal transport through visible and (2) investigate on a large spatial scale the biolog- thermal AOCI data. Model coefficients are being ical responses to riverine inputs of organic mate- modified based on these results. rials, dissolved nutrients, sediments, and fresh water associated with the Mississippi River plume during both high and low river discharge; Sediment Transport and Land (3) investigate on small spatial scales, both hori- Loss Processes zontally and vertically, in a cross-plume direction, the roles of oceanographic fronts, discontinuities, This project began in January 1989 and is a and boundaries; and 4) examine the biological cooperative effort between the Stennis Space Cen- responses to the passages of meteorological fronts. ter Space and Technology Laboratory, the This research will provide remotely sensed ocean Louisiana Geological Survey, and the Coastal color imagery to complement the Louisiana Stim- Studies Institute of Louisiana State University. ulus for Excellence in Research Project (LaSER). The project is designed to develop strategies and On 9 September 1989, the LaSER project flew a procedures for monitoring processes and responses successful CAMS mission. Five flightlines pro- associated with coastal zone land loss. Specific vided complete coverage of the Mississippi plume. objectives are to (1) provide a synoptic monitoring This imagery offers large gradients of chlorophyll capability; (2) develop a data base with remotely pigments, suspended sediments, and dissolved or- sensed data, together with analyses procedures ganic and inorganic constituents for developing suitable for long-range planning in the coastal comprehensive algorithms. During the overflight, zone; and (3) develop an understanding of the links the P./V Pelican collected continuous surface pro- between process and response, particularly with files of in vivo fluorescence, nephelometry, temper- regard to hydrology/sediment transport, so that a ature, salinity, solar irradiance, and plant nutri- set of predictive models can be generated. ents with a near-surface flow-through system NATioNAL PROmAw 57 interfaced to an on-board computer. Numerous western United States (EEZ-SCAN 84). Four discrete samples were collected for sensor calibra- sonographs acquired during the EEZ-SCAN 84 tion. These data are being processed and analyzed survey off central California were used to develop for incorporation into algorithm development of a series of computer program modules to process the CAMS digital imagery. Several programs were GLORIA II data within NASA!s ELAS image pro- developed for NASA!s ELAS image processing en- cessing environment. These modules provide vironment to calibrate and georeference CAMS multibyte preprocessing techniques to reformat data. The georeference software is of special note GLORIA images and to correct for geometric and in that it provides for efficient and accurate radiometric distortions, including water column georeferencing of aircraft data without the need for offset, slant range geometry, cross-track power numerous control or "tie" points. The CAMS data drop off, multiple returns, speckle noise, striping, collected have been reformatted for processing and anamorphic ratio. In addition to these mod- within ELAS, calibrated to yield spectral radiance, ules, which are specific to GLORIA II data, ELAS and georeferenced to latitude and longitude earth contains a comprehensive set of general image coordinates. The data are currently being pro- processing procedures to provide an investigator cessed to yield spatial maps of near-surface chlo- with a consistent and powerful environment in rophyll pigments and suspended sediments. which to fully process and analyze GLORIA 11 data. Side-scan Sonar Acknowledgments Geologists have greatly benefited from the par- allel development of earth-viewing remote-sens- The research discussed in this paper was made ing instruments and comprehensive image pro- possible through funds provided by the Earth Sci- cessing techniques. Airborne platforms or ence and Applications Division and the Life satellite platforms routinely provide data to inves- Sciences Division of the NASA Office of Space tigators to formulate complex spectral analyses Sciences and Applications, the NASA Office of over large spatial scales. Until recently, this tech- Equal Opportunity Programs, and the NASA-S SC nology was unavailable to marine geoscientists. Director's Discretionary Program. The Geological Long-Range Inclined ASDIC (GLORIA) II side-scan sonar system is an acoustic imaging system capable of mapping the sea floor References and providing data for geophysical, geological, and oceanographic investigations. However, as a Bartlett, D. S., K. B. Bartlett, J. M. Hartman, R C. prerequisite to extracting information from an Harriss, D. I. Sebacher, R. R Travis, D. D. Dow, and image for data analysis, various geometric and D. P Brannon. 1989. Methane emissions from the Florida Everglades: patterns of variability in a radiometric distortions must be corrected. A col- regional wetland ecosystem. Global Biochem. Cycles laborative effort exists between NASA/Science 3(4);363-374. Technology Laboratory and the Geodynamics Re- Cloern, J. E. 1984. Temporal dynamics and ecological search Institute at Texas A&M University to de- significance of salinity stratification in an estuary velop image processing software for processing (South San Francisco Bay, USA). Oceanol. Acta, Vol. 7, pp. 137-141. and analyzing digital images acquired with long- Dow, D. D., J. A. Browder, and A. L. Prick. 1989. range side-scan sonars. This project will focus Modeling the effects of coastal wetland change on primarily on data obtained from TAMU2, a state- marine resources. Pages 221-227 in K. Mutz, ed. of-the-art multifrequency side-scan sonar system Proceedings of the eighth annual Society of Wetland under development at the Geodynamics Research Scientists meeting, 26-29 May 1989, Seattle, Wash. Institute. Presently, software is being developed Dow, D., R. Pelletier, C. Clark, D. Brannon, and L. Gunderson. 1987. Remote sensing of the Florida to preprocess and analyze data acquired from the Everglades conducted in support of the methane flux GLORIA II and SeaMARC II systems. study. Pages 326-M in Proceedings of the space life In 1983, the United States declared sovereign sciences symposium on three decades of life science rights over 200 nautical miles seaward from its research in space, 21-26 June 1987, Universities shore. In 1984, the Institute of Oceanographic S i Space Research Association, Washington, D.C. cl- Hansen, J., G. Russell, A. Laces, 1. Fung, and D. Rind. ences and the USGS conducted surveys of the 1985. Climate response times: dependence on climate so-called Exclusive Economic Zone (EEZ) off the sensitivity and ocean mixing. Science 229:857-859. 58 BiowrmcAL REPoRT 90(18) Lawrence, W T 1988. Spectral characteristics of the American Society of Programmatic Remote Sensing, foliage of tropical vegetation. Bull. Ecol. Soc. Am. 2-7 April 1989, Baltimore, Md. 69(2):204. Pelletier, R. E., and S. T Wu. 1989. A preliminary Officer, C. B. 1980. Box models revisited. Pages 65-113 evaluation of the Airborne Electromagnetic in P Hamilton, ed. Wetlands and estuarine processes Bathymetr7 System for characterization of coastal and water quality modeling. Plenum Publishing sediments and marsh soils. Pages 366-375 in Corporation, New York. Proceedings of the technical papers of the American Officer, C. B., and M. M. Nichols. 1980. Pages 329-340 Society of Programmatic Remote Sensing, 2-7 April in Box model application to a study of suspended 1989, Baltimore, Md. sediment distributions and fluxes in partially mixed Rickman, D. L., M. C. Ochoa, and K. W Holladay. 1988. estuaries. Estuarine Perspectives. Multitemporal CAMS over coastal Louisiana: a Pelletier, R. E. 1987. A predictive model to monitor problem in automated georeferencing. Pages temporal and spatial changes in marsh landscape 258-267 in Proceedings of the second Forest Service features in the Barataria Basin. Proceedings of the remote sensing conference, American Society of ninth biennial International Estuarine Research Photogrammatic Remote Sensing, 11-15 April Federation conference, 25-29 October 1987, New 1988, Stennis Space Center, Miss. Orleans, La. [abstract] Roberts, H. H., 0. K. Huh, S. A. Husu, L. J. Rouse, and Pelletier, R. E., and D. D. Dow. 1987. A gridding approach D. L. Rickman. 1987. Impact of cold-front passages to detect patterns of change in coastal wetlands from on geomorphic evolution and sediment dynamics of digital data. Pages 119-128 in Proceedings of the the complex Louisiana coast. Coastal sediments'87, technical papers of the American Society of WW Division of the American Society of Civil Programmatic Remote Sensing, 4-9 October 1987, Engineers, New Orleans, La. 12-14. May 1987, Reno, Nev. Tech. Rep. 463. Pelletier, R. E., and D. D. Dow. 1989. Monitoring the Walters, R. A., R. T Cheng, and T J. Conomos. 1985. inundation of the Florida Everglades with AVHRR Time scales of circulation and mixing processes data in a geographic information system. Pages of San Francisco Bay waters. Hydrobiolo- 266-275 in Proceedings of the technical papers of the gia 129:13-36. WETLANDs MAPPiNG FRoGRAms 59 Enhanced-Environmental Sensitivity Index Mapping Using Remote Sensing and Geographic Information System Technology by Bruce A. Davis Science and Technology Laboratory National Aeronautic and Space Administration Stennis Space Center, Mississippi 39529 John R. Jensen and Elijah W. Ramsey, III Department of Geography University of South Carolina Columbia, South Carolina 29208 and Jacqueline Michel RPI International, Inc. 1200 Park Street Columbia, South Carolina 29201 ABSTRACT.-Environmental Sensitivity Index (ESI) maps are used to support oil spill response teams by providing information about the biological diversity of shorelines. In the event of an oil spill, these maps are taken into the field and used to determine where limited resources will be deployed to mitigate the effects of such a spill. RPI International, Inc. has been producing ESI maps for various geographic areas since 1979, and the company has produced more than 40 atlases. ESI maps are based on information gathered from several large oil spills throughout the world, including the Amoco Cadiz, Burnish Agate, Urquiola, and Metula. As a response tool, ESI maps must contain current information and convey that information in a meaningfid manner to the response team. Four types of information are associated with each ESI map sheet: planimetric base map, shoreline sensitivity index, oil-sensitive wildlife, and access and protection features. Base map construction for ESI maps typically relies on the United States Geological Survey's (USGS) 7.5- or 15-min topographic quadrangle map series. Biological data and shoreline type are manually drawn on mylar overlays, and the product is photographically reproduced. Shoreline sensitivity index data describe environment types that have varying degrees of sensitivity to oil or other pollutants. Oil-sensitive wildlife data are indicated by a symbol representing the species and a line transecting the extent of the species habitat. These symbols carry a wealth of wildlife information, including seasonal patterns, special status (endangered or threatened), and species name. Access and protection features are noted through the use of icons that identify existing marinas, boat ramM booms, oil skimmers, and so forth, used during and after an oil spill. Through the Earth Observation and Cornmercial Application Program, the Science and Technology Laboratory at the National Aeronautic and Space Administration's (NASA) Stennis Space Center and 60 BiowmcAL REPoRT 90(18) the Department of Geography at the University of South Carolina are investigating the development of ESI maps through the use of remote sensing and geographic information system (GIS) technologies. The incorporation of these technologies will enhance ESI through the solution of three major problems. First, adequate base map information is not always available for coastal areas covered by mangroves and other vegetation. When maps do exist they generally are not current and may not be at the desired scale. Remote sensing offers the advantage of routine data acquisition on a temporal basis adequate for most mapping needs in dynamic environments. Second, land cover analysis in tropical areas carried out by boat or plane can lead to inaccurate results. Remote sensing can provide detailed information of land cover through the use of digital image processing and manual image analysis of satellite acquired imagery. Third, portrayal of oil-sensitive wildlife information is difficult in ESI map development. This information can be understood more easily if presented as a single layer in a multivariate data base within a GIS. This would allow spill response managers to query complex data and derive clear and concise information in map form. The focus to date in this project has been to replicate the proven ESI map product using remotely sensed data. SPOT Image Corporation panchromatic data have been used to develop a current base map. Geometric rectification of these data resulted in a base map product that has a t5 m root mean square error. This meets most national mapping accuracies for mapping at the 1:24,000 scale. The updated base map adds substantially to the value of the ESI map because of the improvement in the description of the transportation network Classification of land cover using remotely sensed data meets or exceeds ESI requirements. Red, black, and mixed mangrove classes were mapped in a region surrounding Marco Island, Florida, with both SPOT panchromatic and multispectral data in a merged format (10 x 10 spatial resolution). Furthermore, tidal flats, sand, water, and urban areas were classified to a Level I description with SPOT data. Classification accuracies for all land cover classes exceeded 85% with the satellite digital data. For areas in which confusion of multispectral data resulted in poor classification, the image analyst used interactive on-screen digitizing to classify the imagery. This was incorporated into the overall classification for use in the shoreline sensitivity rating. From the classified satellite digital data it was possible to develop a shoreline sensitivity index by using a spatial search technique to construct a two-pixels-wide ribbon around each land use category. The results of the spatial search were then overlaid on the digital panchromatic base map. This process resulted in a color-coded symbol that was placed adjacent to the shoreline feature shown on the base map. The final product closely resembles the ESI map developed with conventional methods. However, the information in the digital product is current, and the ability to update to meet changing conditions in a timely and cost-effective manner is built into the map. Future work in this project will concentrate on developing the data base aspect of the ESI map. The use of icons and "hot keys" to query the data base will improve the usability of ESI maps. These icons link to a data base containing important information, such as the number of skimmers or type of launch ramps at a marina. Much more work is required before the ESI map is fully automated. However, current results show that incorporation of remote sensing and GIS technologies can produce accurate and current ESI maps showing shoreline sensitivity. NATioNAL PRwRAms 61 Wetland Mapping Supported by the U.S. Environmental Protection Agency by John R. Maxtedi Office of Wetlancls Protection U.S. Environmental Protection Agency A-104F 401 M Street, S.W. Washington, D.C. 20460 ABSTRACT.-Wetland mapping is supported by the U.S. Environmental Protection Agency (EPA) through the Section 404 and Superfurid programs. There are two basic types ofwetlands mapping activities under these programs; comprehensive planning activities under the 404 program referred to as "advance identification" (ADID) and specific studies of 404 enforcement and Superfund sites. ADID projects assess the locations, finictions and values, and potential threats to wetlands within a prescribed area. ADID projects are generally conducted at the 1:24,000 scale or smaller (up to 1:250,000), over areas generally greater than 1,000 acres (and up to millions of acres), and use information sources ranging from high-resolution aerial photography to satellite imagery. Section 404 enforcement and Superfund mapping activities of specific sites are also supported by EPA. This mapping is conducted generally at scales of 1:24,000 or larger (down to 1:3,000), over areas generally less than 1,000 acres, and uses aerial photography as the information source. Technical capability for EPA wetland mapping is available primarily through the Office of Research and Development; limited capability is available through EPA:s regional offices. EPA wetlands mapping activities rely, to a large extent, on the mapping conventions developed by the U.S. Fish and Wildlife Service's National Wetlands Inventory (NWI) program, and in most cases directly use NWI maps and NWI mapping capabilities. Our understanding of the importance of wet- tion has been given to the need to protect the lands and the effects of both natural and anthro- remaining wetland resources in the United States pogenic influences on the Nation@s wetlands re- through the completion of the final report of the sources has increased tremendously over just the National Wetlands Policy Forum (The Conserva- last 30 years. The U.S. Fish and Wildlife Service tion Foundation and the National Wetlands Policy (FWS) has played a leadership role in the classifi- Forum 1988) and the adoption of the Forum's cation of wetlands (Cowardin et al. 1979), the as- recommendation ofno net loss ofwetlands by Pres- sessment of fLmctions and values (Sather 1984) ident Bush. The U.S. Environmental Protection and the assessment of the causes and rate of wet' Agency's (EPA) role in wetlands mapping within land losses Mner 1984). In just the last 5 years, the Section 404 and Superfund programs has in- creased over the last several years as our scientific EPA has developed a fully operational wetlands understanding of wetlands and public support for research program to complement research pro- wetlands protection have increased. grams within the FWS and the U.S. Army Corps of Engineers (COE). Most recently, added atten- Section 4 04 Present address: State of Delaware, Department of Natural Since its introduction in the Federal Water Pol- Resources and Environinental Control, Division of Water lution Control Act of 1972 (Clean Water Act), Sec- Resources, Watershed Assessment Branch, 89 Kings tion 404 has grown to be a major program within Highway, P 0. Box 1401, Dover, Del. 19903. EPA; the program presently includes about 120 62 BioLorxcAL REPoRT 90(18) full-time employees in the regions and headquar- based on the evaluation of 291 NPL sites, 62.5% are ters. EPA!s primary role under Section 404 is to within 2 miles of wetlands, and 32.6% are in wet- review permits issued by COE for the discharge of lands themselves. The EPA Office of Emergency dredge and fill material into waters of the United and Remedial Response is responsible for adminis- States. The scope of the program has evolved from tering the SuperfLmd program and has also devel- one that covered only navigable waters in 1972 to oped similar information documenting the proxim- the current program that applies to all waters of ity of NPL sites to wetlands. This information the United States. The current Federal definition supports the continued development of wetlands of "waters of the United States" is contained in mapping capability by EPA!s Superfund program. several Federal regulations, including those devel- oped for the National Pollution Discharge Elimi- nation System (40 CFR, Part 122.2) and the Sec- Objectives tion 404 program (33 CFR, Part 328; 40 CFR, Parts 230.3 and 232.2). These regulations also include I describe two types of wetland mapping sup- specific definitions ofwetlands. Wetlands mapping ported by EPA: mapping to support the compre- supported by EPA has grown as the scope Of the hensive planning of wetland resources, referred to Section 404 program has grown. as "advance identification" (ADID); and mapping The 1987 Amendments to the Clean Water Act of specific Section 404 enforcement and Super- give EPA and COE joint authority to enforce the fund sites. requirements of Section 404. Section 309 provides a variety of enforcement mechanisms, including the authority to require violators to stop discharge Advance Identification activities and to seek civil and monetary penalties Mapping and prison sentences for violators. This enforce- ment authority requires EPA to generate evidence of violations to be presented in court. Violations Section 230.80 of the Section 404(b)(1) guide- often are detected from aerial photographs. lines (45 Federal Register 85336, 24 December Therefore, as EPA's authority to enforce the pro- 1980) provides for EPA and COE to jointly evalu- visions of Section 404 has increased, so too has ate potential disposal sites within a prescribed EPA's capability to map wetlands subject to illegal area, a process referred to as ADID. EPA has fill activities. EPA recently developed a general prepared draft guidance on the methods of con- overview document that describes the enforce- ducting ADID projects (EPA 1989b); the draft in- ment and other elements of the -Section 404 pro- cludes a list and description of all ADID projects gram (EPA 1989a). completed or proposed to date. ADID identifies, in advance of activities (i.e., development), wetlands suitable for fill and wetlands unsuitable for fill. Superfund This planning approach is designed to direct devel- The Comprehensive Environmental Response, opment away from the most valuable wetlands, Compensation and Liability Act (CERCLA) of 1980 thereby reducing conflicts between affected par- and the Superfund Amendments and Reauthoriza- ties. The following lists the basic characteristics of tion Act of 1986 give EPA the responsibility for ADID's: managing the cleanup of hazardous waste sites in 0 Jointly administered by EPA and COE. The the United States. Mapping of a site, including the project must have the involvement and delineation of wetlands, is often used in remedia- endorsement of both organizations to be tion to document the extent of contamination and called ADID. the parties responsible for the cleanup. If a respon- 0 Provides regulatory predictability to a broad sible party cannot be identified, wetlands mapping range of interests-government, develop- may be completed to define cleanup goals. Such ment, environment, and the general public. delineations are done by using present and histor- * ADID provides information and advice. ical aerial photography. ADID results are not regulatory; that is, Currently, 1,165 sites are on EPA!s national wetlands suitable for fill will not necessarily priority list (NPL). This list is used to rank the receive a Section 404 permit and wetlands expenditure of Federal fimds appropriated under unsuitable for fill will not necessarily be CERCLA. Magistro et al. (1989) reported that, denied a Section 404 permit. NATioNAL PRoGRAms 63 ADID's are used to support many regulatory vide the basis for the mapping conducted under activities under the Section 404 program ADID. Most ADID projects use the mapping con- (e.g., permits, enforcement, and mitigation.) ventions developed for the NWI program. In many Many other benefits can result from completed instances, ADID funding is used directly to gener- ADID's. For example, ADID's provide a basis for ate new NWI maps or to update existing maps. the development of wetlands protection programs Consequently, map scales are generally 1:24,000, at the Federal, State, or local levels of government, and wetland types are classified according to the the acquisition of priority wetlands by government Cowardin et al. (1979) system used in the NWI and private organizations, and the development of program. public education programs. The map products developed in ADID projects There are several steps in the ADID process; consist of two basic types: NWI maps and aerial photographs. In both instances, the wetland types these steps are described in detail in the EPA developed with NWI mapping conventions are guidance draft mentioned previously. The first used to designate wetland areas either suitable or step is the selection of the site to be evaluated. unsuitable for fill. In some instances, the final map Ecological, threat potential, and political factors product omits the detailed wetland classification are used to select the size and specific boundaries information and only includes the designations of of the ADID project. Next, the goals of the ADID wetlands suitable or unsuitable for fill. are selected. The goal may be to support regulatory activities (i.e., Section 404 permitting), State and local program development, or public outreach and Summary of Projects: EPA Region 4 education. In most cases, ADID projects include a (Atlanta) regulatory-based goal. Interagency coordination Detailed information on all of the ongoing or and public participation are the next steps in the planned ADID projects could not be collected in process. In this step, EPA and COE issue a public time for presentation at the NOAA workgroup notice before beginning the ADID. (Government symposium. The following summary of ADID pro- agencies and the public are also informed of in- jects in Region 4 is presented to illustrate the terim and final results.) Finally, the wetlands variety ofIwetland mapping activities supported by within the site are mapped, their functions and the ADID process. Appendix A is a map of the nine values defined, and suitable or unsuitable deter- ADID projects currently underway or proposed minations completed. in EPA Region 4 (southeastern United States). No hard and fast criteria exist for determining Appendix B provides a breakdown of information whether a particular wetland is suitable or unsuit- on each project with regard to funding, the group able for fill. General criteria are included in the responsible for the mapping, the status of NWI EPA guidance draft, and include those criteria contained in the Section 404 program guidelines mapping, and the type of photography used. (45 Federal Register 85336, 24 December 1980). As shown in Appendix B, most projects use NWI The delineation of wetland areas and the assess- maps as the baseline map product, with the excep- ment of their functions and values provide the tion of the Carolina Bays project, which will use wetland maps produced by the State. In some cases, technical basis for this determination. existing NWI maps are used directly, whereas at As of March 1989, there were 58 ADID projects other times, new or updated NWI maps are gener- either completed, underway, or proposed for start- ated. Two basic mechanisms exist for producing up. Table 1 provides a breakdown of these projects w or updated maps: (1) the use of interagency by EPA regional office. A size distribution for ne 36 ADID projects summarized in Appendix C of agreements or grants to other agencies, including the ADID guidance draft is listed in Table 2. the FWS, State agency, and one local government; and (2) the use of in-house technical staff within EPA. Two of the largest ADID's, Pocosins in North Mapping Carolina and Carolina Bays in South Carolina, will use existing wetland maps. Most new mapping The mapping of wetlands for ADID projects activities will use National Iiigh Altitude Photog- provides the basis for assessing wetland functions raphy as the base photos. Two projects include the and values and the determination of areas suitable digitization of the wetland mapping information. or unsuitable for fill. National Wetlands Inventory The Mobile Bay Area ADID is unique in that the (NWI) maps from the FWS, where available, pro- study includes the comparison between NWI maps 64 BioLoGicAL RF.PoRT 90(18) Table 1. Number and status of advance identifications completed, currently under way, or proposed as of March 1989. Map shows geographic area of each eegion. REGION COMPLETED CURRENT PROPOSED TOTAL 2 0 3 11 0 1 1 2 111 5 4 0 9 IV 0 3 6 9 V 5 6 1 12 VI 1 3 2 6 VII 0 2 1 3 VIII 0 2 2 4 Ix 0 0 1 1 x 2 7 0 9 TOTALS 14 30 14 58 EPA REGIONAL OFFICES 2. .0.. ct-. 1. 1. N1- 01. .11 San F-@,. 0 .en,e, K amas C, ty K@ 0.1. TEX .Iss -A 0 6 10 1-co NATioNAL PB)wRAms 65 Table 2. Size distribution for 36 advance identifi- required. If a smaller scale is necessary, the trans- cation (ADID) projects. lation of the NWI information to the smaller scale is not labor- or time-intensive when compared Number with the generation of maps using other sources of ADID of information. Projects Percent Size 4 11 More than 1 million acres (largest 3.5 million) Site-specific 10 28 100,WO to 1 million acres Mapping-Section 404 13 36 10,000 to 100,000 acres Enforcement and Superfund 7 20 1,000 to 10,000 acres 2 5 Fewer than 1,000 acres EPA supports the mapping of wetlands on spe- (smallest 50 acres) cific project sites identified as part of the Section 404 enforcement and Superfund program activi- ties. Stokely (1987) summarized EPA remote- sensingsupport for Section 404 enforcement activ- and computer-generated maps based on aerial pho- ities, and Norton and Prince (1985) summarized tography and ERDAS software. the use of remote sensing for wetlands assessment EPA funding supporting ADID projects in fiscal at Superfund sites. I do not discuss these two year 1990 is about $250,000. It is not possible to programs in detail. The reader is encouraged to estimate the proportion of this funding that specif- review these documents and to contact the EPA ically supports wetland mapping. program offices responsible for these two pro- grams. The following is a general description of the Perspectives for the Workgroup procedures for conducting site-specific wetlands mapping under these two programs. The level of activity involving wetland mapping The deternAnation of wetland boundary changes under ADID has increased over the last few years over time is the most important characteristic that and is expected to increase in the future. Before distinguishes wetland mapping activities under 1987, five ADID projects covered hundreds to thou- these two programs from mapping conducted under sands of acres. Between 1988 and 1989,58 projects ADID. In Section 404 enforcement, it is not as covered tens of thousands to millions of acres. This important to understand the functions and values growth is due to additional EPA funding for ADID of the wetlands as it is to precisely define the extent activities from EPA's Section 404 program. (acreage) of wetland loss, and wetland condition ADID relies on NWI maps for documentation, before an illegal activity occurred. This historical and in some instances directly supports the devel- baseline condition is used in criminal prosecution opment of new or updated NWI maps. NWI maps and provides the goal for restoration of the site back and the wetland classification system used in to its original condition. Restoration often involves NWI provide a consistent basis for evaluations. removal of the fill and some replanting of native Because the detailed classification is often not wetland plants. In Superfund, historical informa- necessary in the final determination of wetlands tion also is important to define the extent of impact, suitable and unsuitable for fill, the final product support the identification of parties responsible for often does not include such detail. This experience the cleanup, and to define cleanup goals. The delin- illustrates the utility of a standardized mapping eation of wetland boundaries and how these bound- protocol and map products developed by the Fed- aries have changed over time because of human eral government. This is particularly important activity is the primary objective for wetland map- for assessing large areas that may cross State ping conducted under both programs. boundaries, where different protocols would cre- The present and historical data collected in- ate compatibility problems. clude wetland boundary delineation, vegetation Mapping that is more detailed than NWI is cover type, and physical parameters. This infor- usually not required. The 1:24,000 scale of NWI mation is often easily detected from aerial photo- maps is an appropriate scale for most ADID's. For graphs. Because aerial photographs are often large areas covering parts of entire States, map- available as far back as the 1930's, they provide ping at smaller scales (e.g., 1:100,000) is usually the necessary historical baseline information. 66 BioLk)GicAL REPoRT 90(18) Consequently, aerial photography has been a grams relies heavily on the standardized mapping powerful tool for both the Section 404 enforcement provided by the Federal NWI program. and Superfund programs. Mapping Support Services from the Generally, the sites evaluated under these two Office of Research and programs are have fewer than 1,000 acres, and Development most often fewer than 100 acres. Because the area is relatively small, and the need for both accurate EPA's primary source of original wetlands and precise information is great because of legal mapping services is located within the Office of actions and liability determinations, mapping done Research and Development. The Environmental under these programs is often at larger scales than Photographic Interpretation Center (EPIC) of the the 1:24,000 scale used by the NWI. In instances Environmental Monitoring Systems Lab-Las where up-to-date NWI mapping is available, the Vegas (a branch of the Advanced Monitoring Sys- NWI maps are often used directly, although even tems Division), supports many of EPXs regional then new photography is often taken on lower-alti- wetlands mapping needs. For example, EPIC pro- tude flights to more precisely define wetland vided the wetlands mapping for EPA's pilot ADID boundaries. for Chincoteague, Virginia, by using large-scale The mapping conducted under these programs aerial photography (Norton 1986). EPIC supports relies on the mapping conventions developed by Section 404 wetlands enforcement with special NWI. Legal actions and liability determination re- overflights, before and after documentation of il- lated to these programs benefit from the widely legal actions, chronological change analysis, accepted system of delineation and classification courtroom displays, and expert testimony. For provided by NWI. The map products developed for several years, EPIC has also produced chronolog- court cases under both the Section 404 enforcement ical assessment map series of the wetlands and Superfund programs are generally aerial pho- around selected Superfurid sites. tographs with overlays. Photos provide an element of reality that is needed in the courtroom proceed- ings, and the small site area often lends itself to presentation on an aerial photograph. Map or photo Conclusions scales are often in the range of 1:3,000 to 1:24,000. The classification of wetlands is based on Cowardin et al. (1979) or a more simplified scheme derived EPA provides limited support for wetlands map- from Cowardin et al. ping under the Section 404 and Superfund pro- No detailed estimate has been developed of EPA gram . Most of these activities rely on the map funding for wetland mapping under Section 404 conventions developed by FWS for the NWI pro- enforcement and Superfund; a rough estimate for gram. The activities range from comprehensive the Section 404 program is under $100,000 per planning covering thousands to millions of acres to year. site-specific assessments of areas less than 1,000 acres where mapping is needed to support legal actions. Consequently, abroad range of map scales Perspectives for the Workgroup is used. Wherever possible, the information avail- able on the NWI 1:24,000-scale maps is directly The level of enforcement and Superfund activity used in EPA-supported wetland mapping. In many related to wetland mapping will increase as these instances, particularly under the comprehensive two program continue to develop. planning program known as advance identifica- Map scales are often larger than NWI scales, tion, EPA funding directly supports the develop- and NWI provides a consistent protocol for devel- ment of NWI maps. EPA is one of many users of the oping these more detailed wetland maps. However, FWS classification system, the NWI mapping pro- because these assessments cover relatively small tocol, and NWI maps. The close association in these areas, these maps are often not used to directly activities between FWS and EPA illustrates the support the development of NWI maps. Similar to value of a coordinated Federal approach toward ADID, the mapping done under these two pro- wetland mapping in the United States. NATioNAL PRoGRAms 67 Acknowledgments symposium on remote sensing of the environment. Vol. Il. 11 pp. Sather, J. H. 1984. An overview of major wetland I thank D. Norton of EPIC, and G. Vanderhoogt functions and values. U.S. Fish Wildl. Serv., of EPA Region 4 for their assistance and guidance FWS/OBS-84/18. 68 pp. in preparing this paper. Stokely, P M. 1987. The Environmental Protection Agency's remote sensing support of the Clean Water Act's Section 404 enforcement activity. 8 pp. The Conservation Foundation and the National References Wetlands Pblicy Forum. 1988. Protecting America's wetlands: an action agenda. 69 pp. Cowardin, L. M., V Carter, R C. Golet, and E. T. LaRoe. Tiner, R. W, Jr. 1984. Wetlands of the United States: 1979. Classification of wetlands and deepwater current status and recent trends. U.S. Fish and habitats of the United States. U. S. Fish Wildl. Serv., Wildlife Service, National Wetlands Inventory, FWS/OBS-79/31. 103 pp. Washington, D.C. 59 pp. Magistro, John L., L. C. Lee, and W Ives. 1989. The U.S. Environmental Protection Agency. 1989a. relationship of Superfund sites to wetlands. 40 pp. Highlights of Section 404. U.S. Environmental [unpublished report] Protection Agency, Office of Wetlands Norton, D. J. 1986. Suitability of Chincoteague wetlands Protection. 12 pp. for Section 404 activities. U.S. Environmental U.S. Environmental Protection Agency. 1989b. Protection Agency, Rep. TS-TIC-85037. Warrenton, Advance identification-guidance to EPA regional Va. 19 pp. offices on the use of advance identification Norton, D. J., and J. Prince. 1985. Use of remote authorities under Section 404 of the Clean Water sensing for wetland assessment in hazardous waste Act. U.S. Environmental Protection Agency, Office sites. Proceedings of the nineteenth international of Wetlands Protection. 60 pp. [draft report] 68 Biour_icAL REPogr W18) Appendix A. Advanced Identification (ADID) Studies in Region IV WEST KENTUCKY COAL FIELD WETLANDS POCOSINS A140 ASSOCIAIEL) WEII@NNI)S N.C. COASTAL PLAIN (PROIP06ED WOLF RIVER WETLANDS MEMPHIS AREA CAEOLINA DAY WETLANDS S.C., COASTAL PLAIN f SWAMP OF TOA &BANY AREA @PROPO.50) [!EARL RIVER FLOODPLAIN JACKSON AREA MOBIL!!AY A EA WEST BROWARD COUNTY N,E. SHARK RIVER SLO GH- -5011THWEST BISCAYNE BAY ADVANCED IDENTIFICATION DADE COUNTY GOAStAL AREA i,@_) -FLORIDA KEYS STUDIES IN REGION IV pIELD wCk)K uM1)pkWA\j OR COMPLE-rED OF zt4pi,F_tiWTA-r10N "ODPLAIN @A@YRF A fieldwork under way or completed). NATioNAL Ppo(,,R"s 69 Appendix B. Costs Associated with Wetlands Mapping in Advanced Identification (ADID) Studies Costs Associated with Wetlands Mapping in ADTDqtudjeq: ADIDs using 1AGs or Grants to other Agencies to perform mapping: Pearl River/Jackson, MS: (No NWI maps for this study area) 1AG w/FWS = $37,000; Assume 2/3 for mapping $24,666 NHAP photography 845 Digitization of maps 25,000 Total to map & digitize 50,511 Mobile Bay Area: (Existing NWI maps out-of-date.) IAG w/FWS to create new NWI maps for 24 quads $36,235/EPA & $15,000/State 51,235 Digitization 20,000 Aerial Photos for ERDAS mapping 7,500 EPA personnel time to process in ERDAS 6,000 Kentucky Coalfield: (Using NWI maps as basis for mapping; NWI maps are based on most recent NHAPs) Grant to State = $35,000; V2 for field verification/updating of NWI 17,500 USGS Topo maps 450 Southwest Biscayne Bay: Grant to Local Agency = 20,000; mapping 13,333 NHAP 2,000 (NWI maps out-of-date) ADIDs using in-house technical staff for mapping: Northeast Shark River Slough: NHAP 1,100 staff time-4 weeks/GS 11 2,441 travel to study area 1,500 (NW1 maps out-of-date) West Broward County: NHAP 1,100 staff time-6 week4/GS 11 3,362 travel to study area 2,400 (NW1 maps out-of-date) Swamp of Tow NHAP 1,215 staff tim" week4/GS 11 4,882 travel to study area 2,000 (NWI maps out-of-date) ADIDs using existing wetlands maps (no new mapping): Pocosine/N.C.-using existing NWI maps Carolina Bays of S.C.-using existing maps produced by the state NATIONAL PwGi?Ams 71 U.S. Environmental Protection Agency's Environmental Monitoring and Assessment Program, an Ecological Status and Trends Program by John F. Paul U.S. Environmental Protection Agency Environmental Research Laboratory 27 Tarzwell Drive Narragansett, Rhode Island 02882 A. F Holland Versar, Inc. 9200 Rumsey Road Columbia, Maryland 20145 Steven C. Schimmel U.S. Environmental Protection Agency Environmental Research Laboratory 27TarzwellDrive Narragansett, Rhode Island 02882 J. Kevin Summers U.S. Environmental Protection Agency Sabine Island GulfBreeze, Florida 32561 and K. John Scott Science Application International Corporation 27 Tarzwell Drive Narragansett, Rhode Island 02882 ABSTRACT.-The U.S. Environmental Protection Agency (EPA) is initiating an Environmental Monitoring and Assessment Program (EMAP) to monitor the status and trends of the Nation's near-coastal waters, forests, freshwater wetlands, surface waters, agroecosystems, deserts, and rangelands. This program is also intended to evaluate the effectiveness of EPA policies in protecting the ecological resources of these systems. The monitoring data collected for all ecosystems will be integrated for national status and trends assessments. The near-coastal component of EMAP consists of four ecosystem categories: estuaries, wetlands, coastal waters, and the Great Lakes. The near-coastal ecosystems have 72 BIOLOGICAL REPORT 90(18) been regionalized and classified, an integrated sampling strategy has been designed, and quality-control procedures and data-base management designs will be implemented. A demonstration project will be conducted in the Virginian biogeographic province in 1990, followed by a full-scale national implementation. EMAP will characterize national ecological resources to establish a baseline for monitoring and assessment. The characterization strategy involves the application of remote-sensing technology to obtain high-resolution data on selected sample sites and lower resolution data over broad geographic areas. The cost of environmental regulatory programs developing EMAP to determine the current status has been estimated at more than $70 billion annu- extent, changes, and trends in the condition of the ally, yet the means to assess the long-range effects Nation's ecological resources. When fully imple- of these programs on the environment do not exist. mented, EMAP will be able to respond to the While regulatory programs are based on our best following questions: understanding of the environment at the time Of * What proportion of the Nation's ecological program development, it is critical to use long-term resources are degrading or improving, and monitoring to confirm the effectiveness of these where and at what rate? programs in achieving environmental goals, and to 9 What are the likely causes of the observed corroborate the science on which they are based. degraded conditions? The U.S. Environmental Protection Agency * What is the current status, extent, and (EPA), the U.S. Congress, and private environ- geographic distribution of our ecological mental organizations have long recognized the resources? need to improve our ability to document the condi- 0 Are control and mitigation programs effective in tion of the environment. Congressional hearings in maintaining or improving the quality of the 1984 on the Monitoring Improvement Act con- resources? cluded that despite considerable expenditures on monitoring, Federal agencies could assess neither the status of ecological resources nor the overall 0 Jectives of the progress toward legally mandated goals of mitigat- Environmental Monitoring ing or preventing adverse ecological effects. In the last decade, articles and editorials in professional and Assessment Program journals of the environmental sciences have re- To provide the information necessary to address peatedly called for the collection of more relevant the previous questions and the goal ofthe program, and comparable ecological data and easy access to EMAP has the following objectives: those data for the research community. The most commonly suggested tools for accomplishing these 1. Estimate the current status, extent, changes, goals include a national ecological survey and a and trends in indicators of the Nation's ecological bureau of environmental statistics. resources on a regional and national basis, with In 1988 the EPA Science Advisory Board, af- known confidence limits. firming the existence of a major gap in environ- 2. Monitor indicators of pollutant exposure and mental data and recognizing the broad base of habitat condition, and seek correlative relations support for better environmental monitoring, rec- between anthropogenic stresses and ecological ommended that EPA start a program that would conditions that identify possible causes of adverse monitor ecological status and trends and develop effects. innovative methods for anticipating emerging 3. Provide periodic statistical summaries and problems before they reach crisis proportions. EPA interpretive reports on ecological status and was encouraged to become more active in ecologi- trends to the EPA administrator and the public. cal monitoring because its regulatory responsibil- ities require quantitative, scientific assessments of ENUP Approach the complex effects of anthropogenic activities on ecosystems. The Environmental Monitoring and Assessing whether the condition of the nation's Assessment Program (ENLAP) is EPA's response to ecological resources is improving or degrading re- these recommendations. quires ecological data on large geographic scales EPA's Office of Research and Development, in and over a long time. EMAP represents a different concert with several other Federal agencies, is approach to monitoring than has been used by NA"ONAL PRWRAW 73 EPA in the past. Specifically, the program is dif- design the monitoring program for the near- ferent in five aspects: coastal component: 1. EMAP proposes to use a top-down ecosystem * Review and evaluate existing data on approach in determining appropriate parameters near-coastal ecosystems with respect to EMAPs in the environment. The program proposes to mon- objectives. It is not EPA's intention to develop a itor those things that relate most directly with new program that disregards historical and ecosystem-level responses. ongoing monitoring activities. EMAP will use 2. An integrated approach is used in the sense the wealth of information that has already been of being able to look across ecosystem types. For obtained on near-coastal ecosystems. example, researchers not only want to determine 9 Determine the pollution endpoints of concern the condition of estuaries, but also would like to that the program is to address, and then determine the possible causes of adverse condi- develop, evaluate, and standardize measures tions that are observed. Scientists recognize that a and indicators of conditions that relate to these lot of problems in estuaries are not due to activities endpoints. Some of these indicators can be that occur directly in the estuaries, but are due to implemented directly into the program, anthropogenic activities in the terrestrial systems whereas others will require further research of our country. Estuaries are the downstream re- and evaluation before they can be implemented. pository for the products of human activities, Regionalize and classify near-coastal whether these products get there by aquatic routes ecosystems as an objective way of grouping or through the atmosphere. systems with similar attributes. Useful 3. A number of indicators will be used to deter- groupings are those that provide within-group mine ecological condition. A lot more information variations that are less than those among about the systems can be obtained by monitoring groups. Because all ecosystems cannot be a group of parameters than by measuring a single sampled, classification should aid extrapolation parameter. among systems within a class. 4. EMAP is envisioned as a long-term program 0 Design a statistically unbiased, flexible, within EPA. The concept of long-term within some integrated sampling strategy for all near- parts of the agency is 3-6 months. Typical research coastal ecosystems that will be compatible with programs in EPA generally last 3 years. In EMAP, the EMAP inland ecosystem strategy however, long-term means decades. For example, (U.S.Environmental Protection Agency 1989). to study the responses of ecological systems to For the program to provide diagnostic regional-scale pollution control strategies (which capabilities (e.g., relate pollution problems with in themselves take years to implement over an potential causes), the overall sampling strategy entire region) would take as much as a decade. must be integrated with a compatible statistical 5. EPA is proposing to implement EMAP as a design. multiagency endeavor. For example, researchers 0 Implement logistics, quality assurance and in the near-coastal component of the program are quality control, and data-base operations. These working with NOAA's National Status and Trends elements are the core of the program, and are Program to merge the two programs into a single necessary to meet EMAPs objectives. Federal marine monitoring program. Also, EMAP 0 Conduct a demonstration project for estuaries is closely coordinating its wetlands activities with in the Virginian biogeographic province (Cape the U.S. Fish and Wildlife Service's (FWS) Na- Hatteras to Cape Cod) in 1990. This demon- tional Wetlands Inventory to ensure compatibility stration project will be used to design the and to use the extensive expertise that already full-scale national implementation portion of exists. the program. 0 Implement the full-scale national program. ENUP Near-coastal Component While the EPA's goal is to establish EMAP in Near-coastal Endpoints of all ecosystem types, its initial emphasis is on Concern testing and implementing the program in the near-coastal estuarine and wetlands systems. The EPA clearly does not have resources to monitor following approach is being used to develop and all attributes of all near-coastal resources, or to 74 BioLoGicAL REPoRT 90(18) conduct research on all specific pollution problems These endpoints of ecological values may be that are likely to be identified as being of concern. affected by any number of anthropogenic or natu- Therefore, EMAP activities must focus on ecosys- ral forces. For example, wetland loss and the tem attributes that are of utmost concern to soci- subsequent declines of fishery nursery areas may ety. These attributes are termed endpoints of con- be affected more severely by hurricanes and sea cern, and each endpoint selected for monitoring level rise than by shoreline development. EMAP should have a direct and easily recognized value researchers find it challenging to discriminate to society. It may not always be possible to take a this type of effect. To do this, we selected indica- direct measurement of each of the selected end- tors of endpoint condition that, when used in points. In some cases, it is necessary to measure concert, would broadly identify the environmental variables, here termed indicators, that have char- impact. The major environmental problems ad- acteristics reflective of the endpoints, but for dressed are: which field data are more easily collected and * eutrophication, to include both primary and interpreted. Measurements of indicators will be secondary productivity imbalances in the water used as estimates of endpoints only if they are column and benthos; directly comparable to endpoint responses and are 0 toxic and pathogenic contamination ofbiological also typical of systemwide responses. tissue, water column, and sediments; The endpoints of ecological status are clearly 0 habitat modification, primarily oriented at related to the public's use of the near-coastal eco- wetlands and submerged aquatic vegetation; systems for commercial, recreational, and aes- 0 cumulative impacts resulting from the thetic purposes. A primary endpoint is the health integrated effects of various categories of of fish and shellfish populations. In other words, environmental stress; and are fish and shellfish populations present in den- 0 emerging environmental problems, such as sities sufficient to make commercial and recrea- global climate change, unknown contaminants, tional harvesting feasible? Also, if the populations overharvesting, and declining biodiversity. are abundant, are they free of disease and other manifestations of stress, and are they safe to eat? In short, a major endpoint of concern is the ability Indicators Selected for ENLAP of the near-coastal water to support harvestable and contaminant-free fishery populations. While Near-coastal Component this endpoint refers to species of commercial or The core indicators for the near-coastal compo- recreational importance, it is directly related to nent of EMAP are: regulatory mandates to maintain naturally repro- Dissolved oxygen. Hypoxic or anoxic conditions ducing populations and communities of resource are aftmetional response of the system to primary value or otherwise (Federal Water Pollution Con- production imbalances, which can result from nu- trol Act of 1972 and subsequent amendments, and trient and biochemical oxygen demand (BOD) Marine Protection, Research and Sanctuaries Act loadings. Associated indicators are nutrient dis- of 1972). charge and loadings data. The second major endpoint of concern is the Water clarity. Algal blooms and high suspended maintenance ofnear-coastal habitat structure. An loads can have significant effects on other system example is the public concern for wetland loss and components. Transmissometry and fluorometry its subsequent effects on species and the func- measurements will be made at least two times tional value of wetlands as physical and chemical during the index period at each station. These buffers between terrestrial and aquatic systems. measurements will be used in conjunction with the Changes in the distribution and abundance Of dissolved oxygen (DO) indicator. submerged aquatic vegetation (SAV) also have Benthic abundance, biomass, and species com- dramatic effects on the public's perception of en- position. This indicator reflects the ability of the vironmental health. Any modification in habitat benthos to support bottom fish populations and the structure, whether it be the filling of a wetland, ability of the benthos to maintain the natural diversion of freshwater inflow, or the presence of sediment processing features important to nutri- noxious algal blooms, is correctly perceived as an ent and contaminant flux. The condition of the environmental health problem. benthic community is also an integrator of the NATjoNAL PRwRAms 75 overall condition of the water body, and may re- Shellfish contaminants. These contaminant spond to contaminants or to eutrophic conditions. measures are the same as for sediments and will Sediment toxicity. The sediment toxicity indica- serve as a direct measure of contaminant expo- tor that uses amphipods is also an integrated mea- sure. They will, however, only be used to explain sure that, in this case, is specific to contaminant changes in growth and survival. exposure and effects. Sediment mixing depth. This measurement is Sediment contaminants. The selected suite of proposed because it is an indication of the func- contaminants is a direct measure of exposure to tional activity of the benthos as related to sedi- this form of input, and will be related to responses ment processing. The implication is that shallowly of thebenthic community and sediment toxicity. mixed sediments have less potential for contami- The number and abundance offish species. This nant flux than deeply mixed sediments. indicator is a cumulative effect response indicator, Biomarker responses. Several biomarkers are which will respond to a host of anthropogenic and proposed for testing, including DNA unwinding, natural factors. phagocytic killing ability, micronucleus forma- Fish gross pathology. This indicator is a re- tion, and stress protein concentration. sponse to contaminant exposure, and reflects on In addition to the indicators just described, there the marketability of the subject fish populations, are a number of stressor indicators used to enhance the interpretation of the indicator responses pre- The core indicators just listed form the basic viously mentioned and to help describe possible measurements we have proposed for the near- causes of adverse conditions. These variables will coastal component. All of these indicators will be be provided by other EMAP groups and Federal, evaluated during a 2-month, late summer index regional, and State agencies. The stressor indica- period, when biological responses to environmen- tors include nutrient and contaminant loadings, tal perturbations are expected to be enhanced. land use patterns, incidence of fish kills and beach We propose to test a number of additional indi- closures, loadings via atmospheric deposition, inci- cators during the 1990 demonstration project, dence and extent of fishery closures, census data, These indicators are allocated to the research and commercial fishery landings. category. Wetlands and SAV acreage. These indicators provide a direct measure of habitat modification Regionalization and and loss, and include wetland fLmctional measures Classification such as shape and boundary variables. The current measurement methods to describe these indicators The near-coastal waters of the United States have not yet demonstrated their utility. We also contain hundreds, perhaps thousands, of estuarine, propose to test the feasibility of using wetlands data tidal wetland, and coastal water ecosystems. It collected from satellite imagery. would be impractical for ENIAP to measure the Remotely sensed chlorophyll and suspended ecological conditions of all of these ecosystems. solids. Responses in the dissolved oxygen would EPA!s available resources allow only a subset of indicate the need to identify prior imbalances in these ecosystems to be sampled. Extrapolation of primary productivity. A posteriori examination of monitoring results to unsampled systems will be satellite images for selected systems would be difficult because the characteristics, functions, pol- assessed for sensitivity in selected low and high lution exposure, and human uses of near-coastal susceptibility classes. environments vary among and within regions. Re- Water column toxicity. The proposed chronic gionalization and classification paradigms provide toxicity tests are integrated measures of water an objective method for grouping ecosystems into column exposure to contaminants. The tests will categories based on sets of similar attributes (e.g., climate, geology, hydrology, currents, and be related to shellfish growth and survival. These biota). The regionalization scheme used in EMAP are our only indicators of water column contami- for near-coastal ecosystems is based on the primary nant exposure. climate provinces and major offshore ocean cur- Shellfish growth and survival. The shellfish rents. We are using the 12 biogeographic provinces indicator represents a nonspecific response indi- shown in the Figure, which are consistent with cator that integrates the ability of the water body those published by NOAA (1990) and used by FWS to support shellfish growth. (Cowardin et al. 1979). GREAT LAKES A!, COLUMBIAN ACAD VIRGINI CALIFORNIAN CAROLINIA ARCTIC MASKA BERING LOUISIANIAN WEST INDIA C21 EMAP NEAR COASTAL REGIONS COLUMBIAN ca ALEUTIAN Figure. Biogeographical province used for near-coastal regionalization scheme in the Environmental Monitoring and Assessment Pro NATioNAL PRorRAms 77 EMAP's basic classification system will group The proposed schedule for implementation of the estuaries into three categories: large estuaries demonstration projects was 1990 for the Virginian large tidal rivers, and small estuaries. A 280-lan province, 1991 for the Louisianian province, 1993 grid network is located over the region, by using a for the Carolinian province, 1995 for the Califor- random start point to provide a systematic grid for nian and Columbian provinces, and 1994 for the selection of sampling locations in the large estuar- Acadian and West Indian provinces. ies. The large tidal rivers are sampled by a linear Implementing the EMAP Near-coastal Project analog of the design for the large estuaries. A into Alaska, coastal waters, and the Great Lakes systematic grid (or more appropriately, a line) is could proceed on a similar parallel track as re- used to characterize the spine of these tidal rivers. sources become available. A list frame of all of the remaining estuarine sys- tems is created and used to select systems for sampling. EMAP Landscape Characterization Near-coastal Demonstration National assessments of status and trends of the Pro'ec condition of ecosystems require knowing not only what percentage of a particular resource is in de- An EMAP Near-coastal Demonstration Project sirable or acceptable condition, but also how much was conducted in the Virginian biogeographic of that resource exists. Some types of wetlands are province during summer 1990. The project is de- being lost at an alarming rate; conversion and loss signed to serve as a model for the implementation of other types of ecosystems are also occurring. of EMAP in other ecosystems. The project has Such changes may be of particular concern if caus- these major goals: ally correlated with pollutant exposure or other � Test and validate the utility of the EMAP anthropogenic stresses. For most ecosystems, few Near-coastal Project indicators for making national data bases can currently be used to derive regional assessments. quantitative estimates of ecosystem extent with � Assess the effectiveness of the sampling design known confidence. for making regional estimates of ecological Landscape characterization within EMAP is a conditions. description of landscape features (e.g., wetlands, � Identify and resolve logistical issues associated forests, soils, land use, and urban areas) in areas with implementing a national-scale monitoring associated with EMAP sampling sites. The charac- program. terization provides some of the stressor indicator � Provide regional-scale information to refine the information for the ENLkP Near-coastal Project. sampling design for full-scale implementation. Characterization uses remote-sensing technology � Select indicators for use in full-scale imple- (satellite imagery and aerial photography) and mentation. other techniques (e.g., cartographic analysis and A report on the results of the demonstration analysis of census data) to quantify the extent and project should be available in September 1991. distribution of ecosystems. Over time, periodic ae- rial and satellite imagery will permit quantitative estimates of changes in landscape features that Proposed Implementation mightberelated to anthropogenic activities and Schedule pollutants. The characterization strategy involves the appli- The proposed full-scale implementation of the cation of remote-sensing technology to obtain high- EMAP Near-coastal Project will be a phased ap- resolution data on selected sample sites and lower- proach into all of the regions of the conterminous resolution data over broad geographic areas. Other United States. The year prior to monitoring activi- data sources will be used to supplement remotely ties in any province will be used to plan and design sensed data. the specifics for implementation. The first year of EMAP will assemble, manage, and update these monitoring will be treated as a regional demonstra- data in geographic information system format. A tion project to test and validate the indicators and standardized characterization approach and a design. The operational monitoring would start in landscape information network common to all eco- the second year. systems will be used to optimize cost and data 78 BloLcGicAL REPoRT 90(18) sharing, and to ensure common format and consis- views of the agency; no official endorsement tency. Through close work with other agencies, should be inferred. EMAP will establish design requirements for the integrated characterization, including acceptance criteria for baseline data, consistent classification References detail and accuracy, and suitable spatial and tem- Cowardin, L. M., V Carter, F C. Golet, and E. T. LaRoe. poral resolution to detect landscape features of 1979. Classification of wetlands and deepwater particular interest. habitats ofthe United States. U.S. Fish Wildl. Serv., The design of the characterization plan and the FWS/OBS-79/31. 103 pp. evaluation of potential characterization techniques National Oceanic and Atmospheric Administration. are in progress. A prototype methodology for high- 1990. Title 15, Code of Federal Regulations, resolution characterization has been developed Chapter IX, Part 921-National Estuarine (Norton et al. 1989). EMAP characterization began Sanctuary Program regulations. in 1990 at about 100 sites. Norton, D. J., D. M. Muchoney, E. T. Slonecker, and J. H. Montanari, 1989. Landscape characterization for ecological monitoring. Rep. TS-PIC-89301. Environmental Monitoring Systems Laboratory, Acknowledgments U.S. Environmental Protection Agency, Las Vegas, Nev. 42 pp. U.S. Environmental Protection Agency. 1989. Design Although the work described in this article was report for the Environmental Monitoring and supported by the U.S. Environmental Protection Assessment Program. U.S. Environmental Agency, it has not been subjected to agency review Protection Agency, Environmental Research and therefore does not necessarily reflect the Laboratory, Corvallis, Oreg. 125 pp. [Draft report] NATioNAL PRwRAms 79 Importance of Hydrologic Data for Interpreting Wetland Maps and Assessing Wetland Loss and Mitigation by Virginia Carter U.S. Geological Survey 430 National Center Reston, Virginia 22092 ABSTRACT.-The U.S. Geological Survey collects and disseminates, in written and digital formats, groundwater and surface-water information related to the tidal and nontidal wetlands of the United States. This information includes quantity, quality, and availability of groundwater and surface water; groundwater and surface-water interactions (recharge-discharge); groundwater flow; and the basic surface-water characteristics of streams, rivers, lakes, and wetlands. Water resources information in digital format can be used in geographic information systems (GIS's) for many purposes related to wetlands. U.S. Geological Survey wetland-related activities include collection of information important for assessing and mitigating coastal wetland loss and modification, hydrologic data collection and interpretation, GIS activities, identification of national trends in water quality and quantity, and process-oriented wetland research. Wetlands are dynamic ecosystems whose exis- tion systems (GIS's) for many purposes related to tence, persistence, and function are controlled by wetlands. hydrology. Hydrologic data are important for wet- Wetland studies are an integral part of the land map interpretation, trend -assessment, im. USGS's hydrologic activities, and USGS wetland pact prediction, site selection for mitigation and projects have been conducted in a wide range of wetland research, and data-collection network de@ hydrogeologic settings throughout the United sign for both small- and large-scale studies. Wet, States. These projects include problem-oriented land maps provide two-dimensional information field investigations, data collection activities, and on the location and classification of wetlands; this process-oriented research. The broad research and information needs to be supplemented with geo- data collection capabilities of the USGS have re- logic and hydrologic information to provide the sulted in significant contributions to the under- third dimension needed for wetland assessment. standing of the role of hydrology in wetland func- As part of its mission, the U.S. Geological Sur- tions and processes. The USGS Federal-State vey (USGS) collects and disseminates, in reports cooperative programs with Federal, State, and local and digital formats, groundwater and surface- agencies in all 50 States aid in maximum use of water information related to the tidal and nontidal research results. In the following sections, I de- wetlands of the United States. This information scribe some of the wetland-related activities of the USGS. includes quantity, quality, and availability of groundwater and surface water; groundwater and surface-water interactions (recharg"ischarge); groundwater flow; and the basic surface-water Coastal Wetland Loss or characteristics of streams, rivers, lakes, and wet- Modification lands (e.g., sediment load, transport, and flood- plain geomorphology). Hydrologic information in Louisiana has experienced significant recession digital format can be used in geographic informa- of its shorelines accompanied by losses in coastal 80 BIOLOGICAL REPORT 90(18) wetlands in the Mississippi Delta during the last of the entire annual U.S. atrazine production, is few decades. Under natural conditions and in the estimated to be carried by the Mississippi River historic past, water overflowing from the Missis- to the Gulf of Mexico (Periera et al. 1989). sippi River at times of high river stage replenished Coastal wetlands throughout the United States the sediment on the wetland surface, resulting in are threatened by rising sea levels. In particular, a balance between accretion and subsidence. Re- fresh and brackish tidal wetlands may be affected cently constructed dikes along the lower Missis- by intrusion of salt water. In addition to rising sea sippi River now prevent the natural overflow of level, other factors that cause such intrusions may sediment-laden water onto coastal wetlands and include storm surges; diversions of freshwater for thus contribute to the current wetland loss. An- industrial, agricultural, and municipal use; and other strong contributory factor in the rapid reces- extensive pumping of groundwater in the coastal sion of shorelines in the Mississippi Delta is the plain. In the case of Hurricane Hugo, which struck decrease in the supply of river sediment. Histori- the South Carolina coast on 21-22 September 1989, cally, the Missouri River basin has been the great- the USGS, in cooperation with the Federal Emer- est supplier of sediment to the lower Mississippi gency Management Agency and the U.S. Army River. Following the completion of five major dams Corps of Engineers, is mapping the height and for irrigation and hydroelectric power on the Mis- extent of the storm surge in South Carolina. souri River from 1953 to 1963, the flow of sediment Within 2 days after the storm hit, USGS staff were from the Missouri River basin virtually stopped in the field locating and surveying the high-water (Meade and Parker 1985), decreasing the sediment marks. Maps showing the altitude of high water load of the lower Mississippi River. caused by the storm surge are being drawn and will be published early in 1990. A similar effort is One of the suggested solutions for stopping, or at least slowing, the loss of coastal wetlands in occurring in Puerto Rico. Through the use of satel- Louisiana is the rediversion of water from the lite telemetry, the USGS also monitors the extent lower Mississippi River. However, the diminished of saltwater intrusion in South Carolina estuaries sediment load of the Mississippi River, as well as (Carswell et al. 1988), and even followed the intru- the river's burden of pollutants, must be consid- sion resulting from Hurricane Hugo as it occurred. ered in any program designed to divert water for Information of this kind can be used to identify stabilization of remaining wetlands in Louisiana. wetlands threatened by rising sea level or saltwa- ter intrusion in order to assist agencies responsible The USGS has an ongoing study on the Missis- for wetland protection and mitigation of storm- or sippi River (Meade 1989) to determine sediment salinity-related damage. loads and to characterize dissolved and sediment- transported contaminants. The portion of the river being studied begins near St Louis, Mis- Hydrologic Data Collection souri, near the confluence of the Mississippi, the Missouri, and the Illinois rivers. The sampling and Interpretation plan is designed to represent variable levels of The USGS summarizes water-resource infor- river stage and discharge. Fortuitously, the sam- mation on many topics in a series of reports known pling period has included the drought years of as the National Water Summary (NWS). Present 1987-88. Depth-integrated samples from 10 to 40 plans are for the 1992-93 NWS to be devoted to verticals in each cross section are composited and wetlands. This NWS will be the first to draw ex- subjected to detailed analysis for metals and se- tensively on information from ongoing programs in lected organic compounds. Preliminary results other agencies doing major work in wetlands, in- show the presence of a variety of trace metals at cludingthe U.S. Fish andWildlife Service, the U.S. nominal concentrations, and computations Of Environmental Protection Agency, the U.S. Army loadings show that most metals are conservative Corps of Engineers, the National Oceanic and At- throughout the system (Taylor et al. 1989). These mospheric Administration, and the U.S. Soil Con- findings are consistent with other studies on the servation Service. 'Me NWS will also report on lower Mississippi River. Although some organic USGS national and State activities in wetlands. pollutants are found only in certain tributaries, Although the report is still in its early planning atrazine and other herbicides are ubiquitous stages, the 1992-93 NWS will contain maps show- throughout the entire Mississippi River system. ing the distribution of wetlands on a State-by- Currently, 100 tons of atrazine per year, or 0.3% State basis. The USGS also plans to work closely NAMNAL PRWRAW 81 with the U.S. Fish and Wildlife Service in produc- ity or quantity changes in wetlands or loss of ing and digitizing a national base map and wetland wetlands on water quality and flood-flow charac- maps for each State and for Puerto Rico and the teristics; and to make management and siting de- Trust Territories. A variety of wetland-related hy- cisions. The USGS has installed GIS's at more drologic data and the locations of data-location than 50 sites, including 3 USGS regional GIS lab- sites will be overlaid on the State and national base oratories, so that nearly every hydrologist has maps for the report. access to a GIS. The USGS is one of the many Federal, State, and local agencies cooperating with the Chesa- peake Bay Program. The USGS is providing data Toxics on quantity and quality of water entering the Chesapeake Bay, and data on submersed The USGS is studying the distribution of toxic macrophyte distribution and abundance in the materials and pollutants in groundwater and sur- tidal Potomac River and its estuary. These data face water in many parts of the United States; GIS will be incorporated into the Chesapeake Bay GIS technology has many potential uses in the study of and into models being developed for the bay and such contamination in wetlands. One recent study its tributaries. One of the major concerns of the involving wetlands is determining the effects of Chesapeake Bay Program is the baywide decline contaminants (specifically selenium) in irrigation in aquatic plant beds (wetlands containing sub- drainage on wetland areas of the middle Green mersed aquatic vegetation [SAV]) in recent years. River basin in Utah (Stephens and Waddell 1989). Data on distribution and abundance of SAV in the Elevated concentrations of selenium in biological Potomac River are being supplied to the Virginia tissue are known to be harmful to wildlife; irnpair- Institute of Marine Science, which is presently ment of reproduction in waterfowl can occur at conducting the baywide SAV monitoring and map- selenium concentrations as low as 1 to 5 pglg (wet ping program. These data are particularly impor- weight) in bird eggs. Elevated concentrations of tant in light of the recent recovery of SAV in the selenium have been found in Utah's water in the freshwater tidal reach of the Potomac River. Since Stewart Lake Waterfowl Management Area, in 1983, the USGS has also supplied maps showing Ashley Creek, and in Ouray National Wildlife Ref- the general distribution of SAV, as well as the uge. The USGS study showed that the sources of location and amount of the exotic submersed selenium at the Stewart Lake Waterfowl Manage- macrophyte Hydrilla verticillata, to the Metropol- ment Area are runoff from irrigated land and the itan Washington Council of Governments to assist discharge of shallow groundwater from sedimen- in its management program (Carter et al. 1985; tary deposits of marine and norimarine origin. Rybicki et al. 1985, 1986, 1987, 1988). Concentrations of selenium in irrigation drainage entering Stewart Lake ranged from 14 to 140 pg/L. Tissues of American coot (17ulica americana) and Geographic Information carp from the lake contained concentrations as System Activities high as 26 pglg (dry weight). Contamination of Ashley Creek is from springs, seeps, and subsur- A GIS combines two computer software technol- face drains that discharge water containing as ogies: data-base management and digital mapping much as 16,000 pg/L of selenium. Selenium con- (Lanfear 1989). Data-base management is a sys- centrations in a pond in Ouray National Wildlife tematic way of organizing and accessing tabular Refuge, which also receives irrigation runoff and data and interfacing the data with maps. The key shallow groundwater, were as high as 93 pg/L; the feature of a GIS is that the digital map elements principal source of contamination is shallow are linked with the tabular information in such a groundwater containing as much as 9,300 pg/L of way that when the map or the tabular data are selenium. Concentrations of selenium in coot em- manipulated, both sets of data are updated and bryos and eggs from the refuge ranged from 6.5 to adjusted to maintain the relation between them. 15 peg (wet weight). GIS technology is a promising tool for combining GIS technology has been used to support the wetland maps, such as those produced by the U.S. Utah study. The ARC/INFO system produced Fish and Wildlife Service, with hydrologic and maps showing the distribution of seleniferous for- geochemical information to detect and analyze mations, waterfowl (wetland) areas, and associ- stresses; to analyze potential effects of water qual- ated concentrations of selenium in source waters. 82 BioLomcAL RFPoirr 90(18) Currently being developed are large-scale maps Program. The goals of this program (Hirsh et al.) showing the distribution of waterfowl nests in are to (1) provide a nationally consistent descrip- contaminated areas, as well as small-scale maps tion of current water-quality conditions for a large showing the numbers and types of banded and part of the Nation!s water resources, (2) define captured waterfowl within selenium-contami- long-term trends (or lack of trends) in water qual- nated areas of Utah. A time-series of maps such ity, and (3) identify, describe, and explain the as these could be used to show trends in the major factors that affect observed water-quality concentrations of selenium over time. conditions and trends. The USGS is conducting seven NAWQA pilot Spatial Analysis of Statewide projects of surface water and groundwater sys- Water Quality tems. One study is of groundwater on the Delmarva Peninsula (Maryland, Delaware, and Federal law requires State governments to as- Virginia). In the initial phase of this project, all sess water quality to aid in the design and assess historic data on groundwater quality through 1987 the effectiveness of pollution-control programs were collected and analyzed (Hamilton et al. dealing with both point and nonpoint sources of 1989). The peninsula was divided into six subre- pollution. In New Jersey, the USGS is using re- gions, each having a distinctive combination of gression techniques to correlate a variety Of hydrogeologic and landscape features, such as sur- water-quality characteristics with spatially de- ficial geology, geomorphology, soils, and land use tailed information on land use and pollution patterns. Each of the regions (referred to as sources. Ambient water-quality data and pollu- hydrogeomorphic regions) represents a landscape tant-loading rates for individual municipal and with differing hydrologic characteristics, which industrial point sources are provided by the New presumably reflect differing water-quality pat- Jersey Department of Environmental Protection. terns. One ofthese regions is a central upland that Information for estimating nonpoint-source con- accounts for more than 25% of the study area. This taminant loads is being derived from spatially region has hummocky topography, is poorly detailed population data and digital land use and drained, and contains hundreds of seasonally land cover data. Land use and land cover classifi- ponded wetlands. 'Me study team is observing cation is based on existing 1973 Geographic Infor- local water-quality patterns around a few of the mation Retrieval and Analysis System (GIRAS) wetlands, but it is also investigating the effects of coverage, updated to 1985 conditions through the wetlands on water-quality patterns throughout use of Landsat Thematic Mapper data. Overland the region, thus underscoring the importance of flow paths and channel networks are identified recognizing and mapping wetlands terrains as through the use of digital elevation data. Once well as individual wetlands. Inclusion of water- constructed, the regression model is applied to a quality and landscape-feature data in GIS data large and representative sample of stream bases will permit the USGS to determine water- reaches to obtain unbiased estimates of water- quality patterns influenced by wetlands at local quality conditions. Potential uses for the method- and regional scales. ology developed in the project include comparison Historical data are important for predicting na- of water quality in basins that have numerous wetlands with basins without wetlands, and iden- tional water-quality trends; these data make it tification of wetlands that might be threatened or possible to compare water quality in basins as a affected by contaminants (e.g., bioaccumulation function of numbers and types of numerous wet- of toxicants in plant and animal tissue). The qual- lands, as well as to examine the effect of changes ity of water discharging from wetlands could also in wetland type or acreage on basin water quality. be included in the assessment of water quality The National Stream Quality Accounting Network statewide. (NASQAN) provides a continuous record of water quality at 441 active stations throughout the coun- try for use in assessing water-quality trends Identification of National (Briggs 1978). Of these, 150 stations are in coastal Trends areas (USGS accounting units) and provide long- term (> than 10 years) data for trend analysis. The USGS is testing and refining concepts for a Coastal accounting units, including those along National Water-Quality Assessment (NAWQA) the Great Lakes, usually contain numerous small NATioNAL PRoGRAms 83 streams with roughly parallel drainage's flowing Water Resources Information into the oceans or Great Lakes. Streams in each of the coastal units reflect similar geographic, geo- The Office of Water Data Coordination is the logic, and hydrologic conditions, although cultural focal point for interagency coordination of ongoing features may differ. and planned water data-acquisition activities of all Water-quality constituents currently are mea- Federal agencies and many nonFederal organiza- sured at NASQAN stations bimonthly or quar- tions. The National Handbook of Recommended Methods for Water-data Acquisition and other pub- terly. Determinations resulting from each site visit lications are available from this office.1 include field measurements of temperature, pH, The National Water 12ata &change (NAWDEX) specific conductance, dissolved oxygen, and bacte- maintains a computerized data system that iden- ria; common constituents; major nutrients; and tifies sources of water data and indexes informa- suspended sediment. Quarterly samples are used tion on the water data available from the sources. to determine concentration of trace elements in The NAWDEX Program Office and Local Assis- addition to the constituents previously mentioned. tance Centers assist data users in locating sources Data collected from NASQAN are available from of water data, identifying sites at which data have the USGS WATSTORE (Water Storage and IL--- been collected, and obtaining specific data.2 trieval) computer storage and retrieval system. Questions about water resources in general, and Data also are published in the series Water Re- about the water resources of specific areas of the source Data for (State), Water Year (date). These United States can be directed to the USGS Hydro- data are being used for trend analysis (Smith et al. logic Information Unit. This office also will answer 1987) and load estimates for nutrients and com- inquiries about the availability of reports of water- mon constituents. Load and trend estimates can be resource investigations.3 related to basin characteristics, including the number and types of basin wetlands. Summary U.S. Geological Survey Data The USGS collects, interprets, and supplies hy- Sources drologic data to supplement and enhance the use of wetland maps. Wetland maps supply the user with The USGS provides many types of information essential spatial information about wetland loca- for wetland managers and data users. Earth Sci- tion, size, and relation to other basin and landscape ence Information Centers (ESIC) offer nationwide features. Hydrologic data are needed to increase information and sales service for USGS map prod- the usefulness of these maps for planning, manag- ucts and earth science publications. This network ing, evaluating, and mitigating loss or degradation of ESIC's provides information about geologic, hy- of wetlands. Long-term information about trends in drologic, topographic, and land use maps, books, water quality and quantity can assist the manager and reports; aerial, satellite, and radar images and in detecting or predicting the effects of various management practices on wetlands or the effects of related products; earth science and map data in wetlands on local or regional water quality. The digital format, anarelatecl applications software; inclusion of hydrologic information in a GIS that and geodetic data. For further information, contact uses wetland maps as a significant layer of infor- any of the ESIC's listed in the Appendix. mation can greatly expand the analysis capabilities ESIC offices assist users in securing publica- of the GIS for wetland planners and managers. tions and associated products in the earth science Federal agencies are responding to the increas- disciplines and use many computerized informa- ing national emphasis on slowing the rate of wet- tion systems to research inquiries. These systems land loss and improving wetland evaluation and include the Geographic Information and Retrieval System for land use and land cover maps and associated overlays (e.g., political and demo- 1 Office of Water Data Coordination, U.S. Geological Survey, graphic), and the Earth Science Data Directory for 2 417 National Center, Reston, Va. 22092. information about earth science and natural re- National Water Data Exchange, U.S. Geological Survey, 421 National Center, Reston, Va. 22092. source data bases maintained by government ' Hydrologic Information Unit, U.S. Geological Survey, 420 agencies and other sources. National Center, Reston, Va. 22092. 84 Biou)GicAL REPoRT 90(18) mitigation. The USGS will be reassessing the Virginia-analysis of available water-quality data ability of traditional data-collection and dissemi- through 1987. U.S. Geol. Surv. Open-File Rep. nation programs to provide information needed by 89-34. 72 pp. wetland managers. The Federal-State Coopera- Hirsh, R. M., W M. Alley, and W G. Wilber. 1988. Concepts for a national water-quality assessment tive Program is refocusing its activities as a result program. U.S. Geol. Surv. Circ. 1021. 42 pp. of changing demands for basic hydrology informa- Lanfear, K. J. 1989. Geographic information systems tion for wetlands. The USGS has already proposed and water resources applications. Water Resour. Bull. a process-oriented interdisciplinary research pro- 25:v-vi. gram focusing on the hydrologic, geologic, and Meade, R. H., and R. S. Parker. 1985. Sediment in rivers of the United States. U.S. Geol. Surv. Water-Supply geochemical processes in wetlands. The objectives Pap. 2275:49-61. of this research are to improve understanding of Meade, R. H. 1989. Sediment-transported pollutants in the integrated hydrologic, geologic, and geochem- the Mississippi River. U.S. Geol. Surv. Yearb. Fiscal ical:ftmctions of wetlands, and to develop predic- Year 1988:20-23. tive capabilities for the evaluation of stresses on Pbriera, W E., C. E. Rostad, and T J. Leiker. 1989. Preliminary assessment of the fate and transport of wetland environments. This will be accomplished synthetic organic agrochemicals in the lower through research on (1) the current hydrologic Mississippi River and its tributaries. Pages 453-464 and geologic processes that create and maintain in U.S. Geol. Surv. Water-Resour. Invest. Rep. wetlands and lakes, including the movement of 88-4220. atmospheric, surface water, and groundwater, Rybicki, N., R. T Anderson, and V Carter. 1988. Data on the distribution and abundance of submersed aquatic and the associated transport of sediment and vegetation n the tidal Potomac River and transition chemicals; and (2) the geomorphic and hydrologic zone of the Potomac estuary, Maryland, Virginia and processes that control the evolution of wetlands. the District of Columbia, 1987. U.S. Geol. Surv. A major component of the research would be sup- Open-File Rep. 88-307. 31 pp. port of long-term studies at several selected wet- Rybicki, N., R. T Anderson, J. M. Shapiro, C. L. Jones, and V Carter. 1986. Data on the distribution and land systems that represent a variety of regional abundance of submersed aquatic vegetation in the hydrologic, geologic, and climatic environments in tidal Potomac River, Maryland, Virginia and the the United States. District of Columbia, 1985. U.S. Geol. Surv. Open-File Rep. 86-126.49 pp. Rybicki, N. B., R. T Anderson, J. M. Shapiro, K. L. References Johnson, and C. L. Schulman. 1987. Data on the distribution and abundance of submersed aquatic Briggs, J. C. 1978. Nationwide surface water quality vegetation in the tidal Pbtomac River and estuary, monitoring networks of the U.S. Geological Survey- Maryland, Virginia and the District of Columbia, Pages 49-57 in Establishment of water quality 1986. U.S. Geol. Surv. Open-File Rep. 87-575.82 pp. monitoring programs. Proceeding of the American Rybicki, N. B., V Carter, R. T Anderson, and T J. Water Resources Association. 12-14 June 1978. Trombley, 1985. Hydrilla verticillata in the tidal Carswell, W J., Jr., C. L. Sanders, Jr., and D. M. Pbtomac River, Maryland, Virginia and the District Johnson. 1988. Freshwater supply potential of the of Columbia, 1983 and 1984. U.S. Geol. Surv. Open-File Rep. 85-77. 26 pp. Atlantic Intracoastal Waterway near Myrtle Beach, Smith, R. A., R. B. Alexander, and M. G. Wolman. 1987. South Carolina. U.S. Geol. Surv. Water-Resour. Water-quality trends in the Nation's rivers. Science Invest. Rep. 88-4066.45 pp. 235:1607-1615. Carter, V, N. B. Rybicki, R. T Anderson T J. Trombley, Stephens, D. W, and B. Waddell. 1989. Selenium and G. L. Zynjuk. 1985. Data on the istribution and contamination from irrigation drainage in the abundance of submersed aquatic vegetation in the western United States with emphasis on Utah. tidal Potomac River and transition zone of the Geology and hydrology of hazardous waste, mining Pbtomac estuary, Maryland, Virginia and the District waste, waste water, and repository sites in Utah. of Columbia, 1983 and 1984. U.S. Geol. Surv. Utah Geol. Assoc. Publ. 17:165-181. Open-File Rep. 85-82.61 pp. Taylor, H. E., J. R. Garborino, and T I. Brenton. 1989. Hamilton, P A., R. J. Shedlock, and P J. Phillips. The occurrence and distribution oftrace metals in the 1989. Groundwater- quality assessment of the Mississippi River and its tributaries. Sci. Total Delmarva Peninsula, Delaware, Maryland, and Environ. In press. NATIONAL PRoGRAms 85 Appendix. Earth Science Information Centers (ESIC's) Anchorage-ESIC Rolla-ESIC 4230 University Drive, Room 101 1400 Independence Road, MS 231 Anchorage, Alaska 99508-4664 Rolla, Missouri 65401 (907)561-5555 (314)341-0851 Anchorage-ESIC Salt Lake City-ESIC U.S. Courthouse, Room 113 8105 Federal Building 222 W. 7th Avenue, #53 125 South State Street Anchorage, Alaska 99513-7546 Salt Lake City, Utah 84138 (801)524-5652 Denver-ESIC San Francisco-ESIC 169 Federal Building 504 Custom House 1961 Stout Street 555 Battery Street Denver, Colorado 80294 San Francisco, California 94111 (303)844-4169 Lakewood-ESIC Spokane-ESIC Box 25046, Federal Center, MS 504 678 U.S. Courthouse Denver, Colorado 80225-0046 W. 920 Riverside Avenue (303)236-5829 Spokane, Washington 99201 (509)353-2524 Los Angeles-ESIC Stennis Space Center-ESIC Federal Building, Room 7638 Building 3101 300 N. Los Angeles Street Stennis Space Center, Mississippi 39529 Los Angeles, California 90012 (601)688-3544 (213)894-2850 Washington, D.C.-ESIC Menlo Park-ESIC Department of Interior Building Building 3, Room 122, Mail Stop 33 18th and C Streets, N.W., Room 2650 345 Middlefield Road Washington, D.C. 20240 Menlo Park, California 94025 (202)343-8073 (415)329-4309 Reston-ESIC 507 National Center 12201 Sunrise Valley Drive Reston, Virginia 22092 (703)860-6045 NATIONAL PROGRAMS 87 The U.S. Geological Survey's National Mapping Division Programs, Products, and Services that can Support Wetlands Mapping by F'ranklin S. Baxter U.S. Geological Survey National Mapping Division National Center, MS-590 Reston, Virginia 22092 ABSTRACT.-The U.S. Geological Survey (USGS) programs can play an important role in support of President Bush's policy of no net loss of wetlands. A principal goal of USGS is to provide cartographic information that contributes to the wise management of the Nation's natural resources. This information consists of maps, cartographic data bases (graphic and digital), remotely sensed imagery, and information services. These products are used by Federal, State, and local governments, the private sector, and individual citizens in making decisions on the existence and use of land and water resources. The identification and classification of wetlands and the activities that affect the quantity, fate, and character of wetlands are described, analyzed, and monitored through the use of cartographic data. There are several specific areas where USGS's National Mapping Division can support the study of wetlands. These include supplying a cartographer to the U.S. Fish and Wildlife Service's St. Petersburg facility to review National Wetlands Inventory (NWI) program procedures and to identify cost- and time-efficient methods for accelerating the inventory; assisting the NWI in using National Aerial Photography Program products for interpretation of wetlands; assisting in research to standardize scanning procedures, to train NWI personnel, and to incorporate data into the National Digital Spatial Data Base System; and integrating data from the NWI program with the National Mapping Division's land use and land cover data and topographic map data. I discuss the programs, products, and information services of the National Mapping Division, the tools available to determine where wetlands exist, and the capability of periodic measurement of wetlands to help in assessing compliance with the concept of no net loss of wetlands. President Bush's policy of no net loss of wet- Programs, Products, and lands is resulting in a refocus of priorities for the Services of the National collection, processing, and publishing of carto- graphic data. The National Mapping Division of Mapping Division the United States Geological Survey (USGS) has been collecting wetlands information as part of The National Mapping Division provides a di- its National Mapping Program for a number of versity of cartographic, geographic, and remotely years. The refocusing of priorities will ensure sensed data, products, and services in support of that data collection will directly support the Federal, State, and public interests through the president's initiative. National Mapping Program. These products and 88 BiowwcAL REPoRT 90(18) services include cartographic and geographic in- The use of graphic maps is being supplanted formation about the earth's natural and cultural rapidly by the use of base digital cartographic data features, basic and special maps in several scales, because such data are more useful and are cost-ef- digital cartographic data, and remotely sensed ficient to maintain and apply. The National Map- data. The division prepares standard topographic ping Division is collecting digital cartographic data maps at specified scales and revises existing maps to meet the needs of a wide variety of users, and is to provide current and accurate cartographic data. producing data in both digital line graph and dig- The cartographic data needs of Federal and ital elevation model formats. Digital data revision State programs are identified and ranked by pri- methods aid in the recording of changes to the ority under the Office of Management and natural and cultural environment. Also, digital Budget's Circular A-16 process. Circular A-16, re- cartographic data are essential in analyzing the vised in 1967, names the Department of the Inte- impact of environmental problems in a geographic rior (delegated to the U.S. Geological Survey) as information system (GIS) context. The National "responsible for the National Topographic Map- Mapping Division has devoted considerable effort ping Series of the United States of America and to developing and promoting the standards and outlying areas of sovereignty and jurisdiction' and specifications necessary to ensure accessibility for exercising "governmentwide leadership in as- and usability of base cartographic data throughout suring coordinated planning and execution" of car- the Federal government. Working through the tographic activities that are funded in whole or in Federal Interagency Coordinating Committee on part with Federal funds. This directive was ex- Digital Cartography and the Interior Digital Car- panded to include digital cartography in 1983. tography Coordinating Committee, other Federal The primary map series provides the largest- agencies have been encouraged to develop their scale information available on a nationwide basis own digital capabilities and have supported the This series includes the 7.5-min topographic quad* development of a National Wetlands Data Base rangle maps of Hawaii, Puerto Rico, and the con- and a National Soils Data Base. terminous United States. In Alaska, the series Other map products that have been useful in provides 15-min topographic quadrangle map cov- Federal and State programs are the intermediate- erage. Many Federal and State programs rely on scale maps at 1:50,000 scale and 1: 100 000 scale in this map series as a base for site-specific environ- quadrangle and county formats. TL National mental studies or as the primary series for record- Mapping Division plans to complete the 1: 100,000- ing information relative to their program needs. A scale topographic map series in FY 95. In 1989, major goal of the National Mapping Divisi .on is to there were about 650 maps available as planimet- achieve initial once-over national coverage in this ric editions, about 950 available as Bureau of Land map series by the end of fiscal year (FY) 1990. Management editions (surface and subsurface Maps covering about 95% of the United States mineral overlays), and another 25 available as have been published, and advance manuscript cop- advance manuscript copies. The 1:50,000-scale quadrangle maps are produced to meet a Defense ies are available for an additional 2%. Ortho- Mapping Agency requirement but are made avail- photoquads at 1:24,000 scale also are available for able to the general public. The county-formatted about two-thirds of the United States. maps are produced as needed on a cooperative The currentness, accuracy, and usefulness of the basis with individual States. primary map series will be maintained through an The intermediate-scale maps also are being expanded map revision program. The National digitized to support the planning needs of Federal Mapping Division has begun a comprehensive Plan and State agencies. Currently, all hydrographic for the identification and scheduling of map revi- and transportation data at the 1: 100,000 scale are sions that would be most beneficial to Federal and available. Other categories of data, such as hyp- State agencies and the general public. The most sography (contours), public land information, efficient methods for revising the primary scale boundaries, and digital elevation models, are pro- maps are being tested, and procedures for incorpo- duced in response to Federal and State agency rating user requirements identified through the requirements. Circular A-16 process are being devised. Projects Many Federal and State agencies use a combi- that reflect the most urgent needs of the user are nation of intermediate-scale products and data for being designed for short-term production.. planning purposes and larger-scale data for more NATioNAL PRocRAms 89 detailed analyses, whereas some agencies are sat- Data Center in Sioux Falls, South Dakota, and the isfied with the level of information at an interme- U.S. Department of Agriculture's Aerial Photogra- diate scale for land management and environmen- phy Field Office in Salt Lake City, Utah. tal studies. For example, time and cost benefits can Image maps, primarily orthophotoquads, are be realized by locating study areas from a regional prepared in response to specific requirements of perspective and obtaining source material (photo- Federal and State agencies. Orthophotoquads are graphic coverage, base maps, or appropriate digital scale-rectified image bases that meet national map data) for only those areas where more detailed accuracy standards, are produced from NAPP pho- analyses are needed. tographs, and are prepared at 1:24,000, 1:63,360, The National Mapping Division is conducting or 1:12,000 scales. These image bases can be pro- pilot projects to investigate the benefits of produc- duced in about one-third the time required for ing a large-scale orthophotoiniage product and of topographic maps; however, they contain no con- revising the land use and land cover map series at tours and only a limited number of feature names. a larger scale. In both instances, the division is During FY 89, the National Mapping Division pre- responding to specific requirements expressed by pared 1,178 orthophotoquads at 1:24,000 scale, 140 Federal and State agencies through the Circular at 1:12,000 scale, and, in Alaska, 175 at 1:63,360 A-16 process. scale. Presently, about 40,000 orthophotoquads are The National Mapping Division is assessing the available, covering more than two-thirds of the value of land use and land cover map revisions at conterminous United States, all of Hawaii, and a the 1:100,000 scale. Now that completion of the portion of Alaska. Much of the work is produced topographic editions at this scale is within sight, directly from requests from the Bureau of Land the use of these maps as a base for land use and Management and the Soil Conservation Service. land cover mapping based on an enhanced classi- The cornerstone of the National Mapping fication system seems quite promising. Many Fed- Division's information delivery network is the Of- eral and State agencies have expressed interest in fice of Information and Data Services. This office designing the classification system for use in a GIS manages the Earth Science Information Center environment. Currently, the land use and land (ESIC) network composed of 13 ESIC offices, and cover mapping program results in maps and asso- 1 Federal and 61 State ESIC affiliates. Developed ciated information (political boundaries, hydro- through the merging of National Cartographic In- logic units, and census county subdivisions) at the formation Centers and the Public Inquiries Of- 1:250,000 scale. The lower 49 States are covered at fices, ESIC's responded to about 567,000 inquiries this scale; about 85% of these maps have been I ot yeox, E0IC offices maintain data records in digitized using the Geographic Information Ra- a h publications as the Cartographic Catalog, the trieval and Analysis System. The maps and data suc are becoming out-of-date, and use of the Geo- Map and Chart Information System, and the Ae- graphic Information Retrieval and Analysis Sys- rial Photography Summary Record System. tem is not widespread. In addition to the benefits The EROS Data Center produces high-quality to State users, a larger-scale mapping and digital map products from satellite data for a variety of data program will benefit wetland analyses and Federal and international organizations. The other studies, such as global change research. EROS Data Center archives more than 800,000 The National Aerial Photography Program Landsat scenes and, in 1990, will have more than (NAPP) provides standardized and uniform quality 150,000 Thematic Mapper scenes and Advanced photographic coverage of the 48 conterminous Very High Resolution Radiometer data for the States on a planned 5-year acquisition cycle. Color- entire country. The EROS Data Center has estab- infrared photographs, at a scale of 1:40,000, are lished agreements with the commercial compa- centered on quarter sections of each standard nies EOSAT and SPOT Image Corporation to 7.5-min USGS quadrangle. NAPP contracts serve as a single point of contact to purchase awarded in 1989 cover all or part of Arkansas, Landsat and SPOT data for Federal agencies. In southern California, Louisiana, South Carolina, the mid-1990's, the EROS Data Center will pro- Texas, eastern Virginia, and Wyoming. Those Fed- cess, archive, and distribute remotely sensed land eral agencies or States that participate in the NAPP data acquired by selected sensors flown by the program receive a discount on all NAPP products. National Aeronautic and Space Administration's NAPP products are available from USGS's EROS Earth Observing System. 90 Biou)GicAL RFPoRT W18) Applications to Wetlands Digital Data Studies (Digital line graphs and digital elevation models The following are several examples of National at several scales) Mapping Division products and how these prod- ucts are used by other Federal agencies in wet- U.S. Fish and Wildlife Service lands studies. Coastal areas of the southern and eastern United States-to be used as ancillary data Primary Map Series bases for wetlands habitat data. (1:24,000-scale maps in the conterminous Lake Okeechobee-to support wetlands studies United States and Hawaii, and 1:63,360-scale being conducted by FWS's Region 8 Florida maps in Alaska) Cooperative Fish and Wildlife Research Unit. �U.S. Fish and Wildlife Service National Park Service Conterminous United States-to serve as abase Cape Cod National Seashore-to support use of map of the National Wetlands Inventory a GIS in the monitoring of land use changes, conducted by the U.S. Fish and Wildlife Service. including wetlands. Coastal Louisiana-to revise maps to better reflect the loss of coastal wetlands to support Cumberland Island National Seashore-to studies on the effects of habitat loss. study, with several other agencies, a variety of activities along the coast, including emergency �National Park Service preparedness and habitat studies. Cape Cod National Seashore-to update and Other areas where digital data are to be used to correct topographic maps to better portray support wetlands studies: coastal wetland areas. Wrangell-St. Elias National Park and Pre- Okefenokee National Wildlife Refuge, serve-to update maps to portray the current Georgia-Florida topographic situation, including development of The Everglades, Florida wetlands because of rapid glacial retreat. Galveston Bay, Texas �Environmental Protection Agency San Joaquin Valley, California Malheur National Wildlife Refuge, Oregon Horry County, South Carolina-to revise maps Charles M. Russell National Wildlife to assist in the study of the creation, Refuge, Montana maintenance, and impact of environmental Hawaiian Islands problems on Carolina Bays. Mobile Bay, Alabama Intermediate-scale Maps Environmental Protection Agency (1:50,000-scale and 1:100,000-scale maps in Areas where hydrographic data are needed to quadrangle and county formats, and 1:250,000- support related habitat and wetlands studies: scale quadrangle maps) Albemarle/Pamlico Environmental Study, �U.S. Fish and Wildlife Service North Carolina Coastal areas of the southern United States-to Merrimack River, Massachusetts-New update maps to reflect the rapid loss of coastal Hampshire wetlands in Alabama, Louisiana, Mississippi, Massachusetts Bays, Massachusetts and Texas. Slidell, Louisiana Edisto River arid Horry County, South � National Park Service Carolina Kenai Fjords National Park-to update maps Chesapeake Bay Study, Virginia, for recording the creation of wetlands and other Maryland, Pennsylvania conditions due to glacial recession. Narragansett Bay, Rhode Island NATIONAL PROGRAMS 91 Land Use and Land Cover Coordination Efforts in (The National Mapping Division is considering Wetlands Research and the creation of a new series of land use and land Technical Assistance cover maps at the 1:100,000 scale. The wetlands classifications would be developed in coordination Several research studies, program initiatives, with the U.S. Fish and Wildlife Service.) and coordination ventures that relate to wetlands are being pursued in cooperation with Federal and Areas where land use and land cover data are State agencies. Examples of these include pro- needed include: grams in Mystic, Connecticut; Elizabeth River, National Park Service Virginia; James River, Virginia; and the Prairie Pothole region in the Midwest. Big Thicket National Preserve, Texas As the use and acceptance of GIS technologies Saratoga National Historic Park, New York become more widespread at all levels of govern- Environmental Protection Agency ment, the reliance on computer-based environ- mental studies in the USGS will increase accord- Chesapeake Bay Study, Virginia, ingly. Several studies are being conducted by the Maryland, Pennsylvania National Mapping, Water Resources, and Geologic Albemarle/Pamlico Study Area, North divisions that investigate the quality of water in a Carolina wetlands environment, the creation and mainte- Pearl River Basin, Louisiana nance of wetlands, and the effect of human activi- Savannah River Basin, Georgia ties on wetlands. GIS projects that are underway Georgetown and Beaufort, South Carolina include management of hazardous waste sites, Delaware Bay, Delaware, New Jersey some of which have direct effects on nearby wet- lands; the movement of toxins through groundwa- Image Maps ter and surface water; and continuing research on the environmental health of the Chesapeake Bay The National Mapping Division is investigating drainage area. The wetlands ecosystem provides the usefulness of new maps that could enhance the investigators with a natural laboratory in which study of wetlands. This includes the development complex environmental processes can be investi- of anew series of image-based maps at the 1: 12,000 gated. GIS modeling allows scientists to further scale, called quarter-quad orthophotos. The use of expand the horizon of scientific inquiry by permit- these maps could result in the more precise record- ting effective visualization and the interaction of ing of the existence and extent of wetlands. Be- the many complex data sets involved, while simul- cause the orthophotoquad production process is taneously providing the capability for the quanti- much shorter than the production of standard to- fied investigation of spatial and temporal patterns pographic map revisions and the positional accu- in the data. racy is comparable with the revised map, the or- The National Mapping Division and the Na- thophotoquad could become an essential tool in the tional Wetlands Inventory staffs are pursuing the study of wetlands. development of formal agreements to conduct The Soil Conservation Service, Department of apping activities that will result in mutually Agriculture, has been working with the National in Mapping Division on designing an image base map beneficial data production and use. The three prin- at the 1: 12,000 scale in support of the national soils cipal objectives of these cooperative ventures are inventory. These image maps are used by Soil sharing of personnel, technology, and data. Conservation Service field personnel; they are pro- With regard to personnel, the National Mapping duced under a joint funding arrangement and use Division proposes to make a cartographer avail- NAPP photographs as source materials. As the use able to NWI for a maximum of 2 years to review of 1:12,000-scale orthophotoquads increases, the the production processes for the generation of na- division will assess its position on standardizing tional wetlands maps. A remote-sensing specialist the compilation and final design to reflect the most also will be available on an as-needed basis to advantageous use of this product. At present we identify the most effective use of NAPP photo- believe that the 1: 12,000-scale orthophotoquad can graphs, the procedures for handling large amounts be an integral part of wetlands research. of new source materials, and the conventions re- 92 BimiowcAL REPoRT 90(18) quired to classify wetlands from NAPP source ma- graphic and digital form. For project planning, the terials. intermediate-scale maps and data provide a re- With regard to technology, the National Map- gional perspective. Some Federal agencies are now ping Division proposes to identify state-of-the-art increasing their support of even larger-scale maps software and hardware systems to assist in stan- and data, primarily in an image format. The Na- dardizing scanning procedures, to assist NWI per- tional Mapping Division is investigating the use- sonnel in developing techniques to convert NWI fulness of quarter-quad orthophotographic prod- graphic products to digital products, and to assist ucts to respond to this growing need. Data in quality-assurance procedures so that wetlands dissemination networks are in place and accessible data can be incorporated into the National Digital to Federal and State agencies nationwide. Spatial Data Base System. This technical assis- Currently, the National Mapping Division is tance will reduce duplicative efforts ensure data providing support to the National Park Service, collection meets national standar&, maximize the U.S. Fish and Wildlife Service, the U.S. Envi- program efficiencies, and encourage technology ronmental Protection Agency, the Soil Conserva- transfer. tion Service, the U.S. Army Corps of Engineers, With regard to data, a cooperative effort is nec- the National Oceanic and Atmospheric Adminis- essary for transfer of wetlands data from the NWI directly to the National Digital Spatial Data Base tration, and the U.S. Forest Service in a number of System. These data will be important for the effi- studies related to wetlands research. One of the cient conduct of the National Map Revision Pro- most effective research tools in the study,of wet- gram and the land use and land cover mapping lands is GIS, a technology in which the National effort. Procedures could be developed to assist in Mapping Division has valuable expertise. The use the revision and updating of wetlands data cur- of division maps and data in GIS's is increasing rently recorded on the 1:24,000-scale topographic and is expected to continue. One of the major map series and to update the land use and land initiatives of the National Mapping Division is to cover maps that require current wetlands classifi- provide technical assistance to any Federal or cation and mapping. This effort also will ensure State agency that seeks cooperative development that the most current wetlands data are available of wetland research projects. to the general public through the National Digital The need for a coordinated approach to support Spatial Data Base System. the president's Wetland Initiative is being ad- dressed by the National Mapping Division and the Summary National Wetlands Inventory through the estab- lishment of formal cooperative agreements. These The National Mapping Division produces and agreements will involve the sharing of expertise, disseminates a variety of cartographic, image, and the development of a wetlands component in the digital maps and data that are useful to Federal National Digital Spatial Data Base System, and and State agencies involved in wetlands research. the exchange of wetland thematic and base carto- The primary map series is most often used, in both graphic data between the agencies. NATioNAL PwGRAms 93 Soil Conservation Service's Wetland Inventory by Billy M. Teels Soil Conservation Service U.S. Department of Agrriculture South Agriculture Building Room 6144 P.O. Box 2890 Washington, D.C. 20013 ABSTRACT@The Soil Conservation Service (SCS) conducts its wetland inventory under the auspices of the Food Security Act (FSA) of 1985. Through the wetland conservation (Swampbuster) provision of FSA, agricultural producers are denied United States Department of Agriculture (USDA) program benefits for converting wetlands for agricultural production. The SCS has the technical responsibility for identifying: FSA wetlands and converted wetlands. The USDA program agencies (Agricultural Stabilization and Conservation Service, Farmers Home Administration, and Federal Crop Insurance Corporation) determine producer eligibility for their respective programs once wetlands and converted wetlands have been identified. Critical to the effective implementation of Swampbuster is the accurate and timely identification of wetlands for affected persons and agencies.The SCS Wetland Inventory focuses on inland h7eshwater wetlands that have a high potential for agricultural conversion. The conversion of wetlands to agricultural land has accounted for more than 80% of the Nation's wetland loss. The SCS has set a goal of 31. December 1991 to complete wetland determinations for all USDA program participant croplands and other lands identified as having a high potential for conversion. The Soil Conservation Service's (SCS) Wetland to occur without detection by the self-certification Inventory began in the Red River Valley of the process. Also, producers with lands in different north, in North Dakota and Minnesota, January States or counties would get different answers on 1988. Initially, SCS had not planned to conduct a what constituted a wetland, depending on where wetland inventory for the Food Security Act (FSA). their land was located. It soon became evident that Swampbuster (the wetland conservation provision an inventory was necessary to avoid confusion on of the FSA) was to have worked based solely on the location of wetlands and to clarify the extent to producer certification, a process whereby the U.S. which maintenance could be performed. This con- Department of Agriculture (USDA) program par- ditiOn was particularly true in the Red River Val- ticipants would certify annually on form AD-1026 ley ofthe north, where most producers annually as to their intent to modify wetlands. However, perform some maintenance of their drainage sys- that process proved inadequate because producers tems. If producers were going to comply with did not know what was considered a wetland under Swampbuster, wetlands had to be identified. FSA, nor did they know to what extent Swampbus- The inventory was a success in the Red River ter would allow for maintenance of existing drain- Valley. For the most part, producers agreed with age systems that involved wetlands. Therefore, on the wetlands identified by the inventory, and con- nearly all self-certification forms, producers had servation agencies and environmental groups were checked "no" in the blocks that asked if modifica- pleased with the results. Because of the variety of tions were to be made in wetlands. Nevertheless, the information from which wetland interpreta- documented modifications in wetlands continued tions were made and the large scale of the mapping 94 BiowGicAL REPoRT 90(18) tools employed, the inventory identified more wet- Scope of the Project and lands than had been identified in previous invento- Project Period ries. Even though more wetlands were identified, it was clear that the vast majority of delineations met As previously discussed, the scope of the inven- the wetland definition under FSA, and that the tory includes all lands that have a high potential inventory did not include land that was outside that for agricultural conversion. The 31 December 1991 definition. Producers in the Red River Valley aP- date sets an ambitious goal for completed determi- preciated that wetland determinations were made nations, a goal difficult to achieve without the aid promptly and that they were made before the 1988 of an inventory. Inventory funds have been made growing season when decisions had to be made on available in fiscal years arY) 1989, 1990, and 1991 drainage maintenance and planting. The inventory to SCS State offices with an interest in conducting ended the confusion, speculation, and conjecture an inventory. that was plaguing Swampbuster at the time. Based on the success in the Red River Valley, SCS decided to expand the inventory elsewhere. Description of Map Products Wetlands and converted wetlands are identi- Inventory Coverage to Date fied on a variety of map products. There is no standard scale or map on which the inventory is Because the potential for agricultural conver- produced. However, wetland determinations are sion is low in much of the Nation's wetlands, SCS usually made on photocopies of black and white does not intend to conduct an inventory that covers aerial photographs provided by the Agricultural the entire Nation. Only those regions considered Stabilization and Conversion Service (ASCS). to have high potential for conversion of wetlands There is a difference between wetlands identified to annual crops have been identified, and funds for the inventory and wetlands identified during have been made available to States within those the determination process. The wetland inventory regions for inventory purposes (Fig. 1). The inven- identifies more or less all the wetlands in a county tory is about 25% complete to date, with the great- or major land resource area, and is used as a tool est progress-occurring in the Midwest (Fig. 2). FSA from which wetland determinations are made. wetland determinations, for which the inventory Wetland determinations are made on form CPA- was developed, are complete for about 25% of the 026 by SCS. The producer and the program agency USDA program participants. are provided a photocopy of an aerial photograph VVA @g` ME VT AK OR. WY ............ ............. P IL N ......... NJ T W .......... k' NM V 5W" $1-99,000 CH-1 k -199,000 100,DOO 200,000 - 299,000 300,000 - and up Fig., 1L. 1989 allocations for wetland inventories. Ulm I r ow 3OURM DM PPIVAM rr 9M nELD PVt5ONNIEL kVP CO&PLO U" AM06"TO 1?0 2?0 300 400 sw w WAP COMMUCTON VAM TW FVC4S EQU@%10ff. O"MOD"t CARfDGPAPW cewro row WORTK VIAS I*". Pig. 2. Completed wetland inventories (September 1989) of the Soil Conservation Service. 96 BIOLOGICAL REPORT 90(18) of the producer's property delineating the wet- tory is an office process that shortcuts the need to lands and converted wetlands. The scale of the make on-site wetland investigations and without a photography on which determinations are made significant loss in accuracy. varies; however, 8 inches per mile or 1:660 is the Mapping conventions have been developed as a most common. Because this scale is used for most guide to interpret office information and provide determinations, the inventory commonlyuses this consistency to the inventory. Conventions are gen- scale as its base. In most instances, SCS will order erally developed for each SCS field office and are the latest black and white aerial photography tailored to the information that is available locally. (prints) at the same scale that ASCS uses for its The SCS State offices have developed broad con- program on which to produce the inventory. In- ventions for -major land resource areas" that serve ventory delineations are made directly on the as a framework through which field office mapping black and white prints with the appropriate FSA conventions are developed. Likewise, regional map- designations. Photocopies can then be conve- niently made of the prints and provided to the ping conventions have been developed by SCS's National Technical Centers to serve as a guide for program agency and program participants as part the development of State conventions. The appen- of the determination process. Some inventories dix provides an example of a regionally developed use soil survey maps as a base; these vary in scale mapping convention for the Prairie Pothole region. (1:10,000-1:12,000 for detailed surveys or The mapping conventions are, in effect, the control 1:24,000-1:64,000 for extensive surveys). Other over the inventory, and are usually codeveloped inventories have used satellite data to interpret with the FWS. The SCS's National Technical Cen- wetlands and produce delineations on mylar over- ters must concur with State mapping conventions lays at a scale of 1:24,000. before they can be used in the inventory. Although there is no standard map product at Mapping is performed by teams of SCS techni- this time, we anticipate that SCS wetland deter- cians, SCS district conservationists, or by teams minations eventually will be incorporated into a working with technical consultants. In much of the standardized county map system and a digital Midwest, SCS technicians are assigned to field county data base that will be adopted by all USDA offices to conduct the inventory. Once all the office agencies. information and collateral data (e.g., weather re- cords, river gauge data) have been assembled, a Inventory Methods three- or four-person team interprets the informa- tion based on the local mapping conventions, and The FSA!s definition of wetland is as follows: makes the delineations on the base maps. The team Lands that have a predominance of hydric soils that can conduct the inventory for the field office area are inundated or saturated at a frequency and (usually a county) within about 2 weeks. Then the duration sufficient to support, and under normal team moves on to the next field office and repeats circumstances do support, a prevalence of hydro- the process. Before the team begins mapping, team phytic vegetation typically adapted to life in satu- members spend time at the field office familiarizing rated soil conditions. This definition contains the themselves with available mapping tools and con- three wetland parameters that have been used to ventions and becoming acquainted with the identify wetlands under Section 404 of the Clean county's landscape and ecology. The local district Water Act, and by the U.S. Fish and Wildlife Ser- conservationist contributes information based on his or her experience in the county. To ensure vice (FVS) in the National Wetlands Inventory. acy, the team members and the district con- Those same parameters are now recognized by the accur new Federal Manual for Identifying and Delineat- servationist make periodic field checks of the delin- ing Jurisdictional Wetlands (Federal Interagency eated wetlands during the mapping process. Committee for Wetland Delineation 1989). The The SCS area office staff annually reviews all of SCS's Wetland Inventory uses office information, the field offices involved in the inventory, providing such as ASCS compliance slides, other aerial pho- first-line quality control for the inventory. The SCS tography, FWS National Wetlands Inventory State offices annually spot-check a minimum of maps, SCS soil surveys, -local weather records, 109/6 of the field offices involved in the inventory and stream gauge data, and other locally available data, all of the area offices performing quality control. as the basis for determining if wetland soils, hydrol- The SCS National Technical Centers annually re- ogy, and vegetation are present. Thus, the inven- view the performance of all State offices involved in NATioNAL PRoc.RAms 97 the inventory. Reviews focus on whether the teams requires to qualify as a farmed wetland. The team are making accurate determinations as compared and representatives from the Stennis Space Center with determinations that would be made on site, corTelated river gauge data from the Mississippi and whether the teams are accurately applying the and Yazoo rivers with available satellite imagery mapping conventions to the inventory tools. showing flooding or ponding that equaled or ex- The SCS State offices must approve each field ceeded 15 days. FSA wetland and converted wet- office inventory before it becomes final and is re- lands were then color-coded on mylar overlays at a leased to the public. The inventory is made avail- scale of 1:24,000. able to land users, first on an informal basis, either though direct mailing or by conducting public meetings to review the inventory. The land users Cartographic Procedures are asked if they agree with the delineations. If Wetland delineations are hand-drawn on exist- they do, the SCS makes final determinations from ing maps or photographs. Interpretations are made the inventory and transmits them to the producer according to the mapping conventions to determine and the program agency through form SCS CPA- an area is a wetland, and then the technician 026. The form includes a represeantation of the F producer's farm with wetlands and converted wet- interprets the extent of the wetland boundary lands delineated, including information on the re- based on the signatures produced from the various strictions associated with the delineations. If the imagery or lines drawn on other wetland maps (e.g., land users object to the determination, they can National Wetland Inventory maps). Lines are nor- mally drawn without the aid of transfer scopes and meet with the district conservationist to reconsider without rules to ensure consistency of the delinea- the determination. Many times the differences are tions. However, land users seldom appeal the resolved as a result of the meeting. At other times, boundary lines that SCS produces. the district conservationist may have to make an on-site determination, which then becomes the final determination. The FSA determination pro- Availability of Map Products cess thus provides a means for determinations to be made either from the office or on site, based on The SCS's primary responsibility under FSA is complexity of the determination and agreement of to make wetland and converted wetland determi- the land users. Where inventories have been com- nations for USDA program participants and USDA pleted, most final determinations have been made program agencies. Because of an intense focus on from the inventory with the land users' concur- providing wetland information to primary users in rence. However, if land users disagree with final a short period, there has been no concentrated wetland determinations, they can appeal the de- effort to make the SCS Wetland Inventory available terminations through a formal appeals process. to the general public. However, interested persons In other instances (e.g., where the workload is can request photocopies of specific wetlands from relatively low or where the cost of establishing SCS State conservationists. In addition to the in- teams is prohibitive), the district conservationist ventory, other information is available from the conducts the inventor SCS to aid others in making wetland determina- ,y with the same mapping tions (e.g., county lists of hydric soils, soil surveys, conventions and subject to the same quality control and aerial photography). That information can also as a team. However, consistency usually suffers be obtained from SCS State conservationists. when the inventory is performed in this manner. In the lower Mississippi Valley, satellite imagery has been used to conduct an inventory. In Missis- Estimated Funding sippi, a State wetland inventory team has worked with the Stennis Space Center Institute for Tech- The SCS spent $6,143,000 on the wetland inven- nology Development to produce inventory maps for tory in FY 89, and $8,175,000 in FY 90. Inventory the Mississippi Delta with Landstat Thematic funds were allocated to SCS State offices based on Mapper and Landstat Multispectral Scanner data the amount of wetlands identified as having a high from 1984 to 1989. The major reason for relying on potential for conversion and the State's expressed satellite imagery in the Mississippi Delta is to interest in conducting an inventory (Fig. 1). Some verify seasonal flooding (inundation for 15 consec- States supplemented the inventory funds provided utive days during the growing season), which FSA nationally with general funding from FSA. 98 BIOLoGicAL RF.PoRT W18) Anticipated Future Act* ffies tifying only those areas that have high potential for IVIL conversion to annual crops. The SCS plans to continue the inventory for FY 91 but not beyond unless it becomes necessary to identify additional wetlands. For example, the User Perspective next farm bill may call for the conversion of wet- As previously discussed, the primary users of lands as the trigger for Swampbuster penalty, the SCS Wetland Inventory are USDA program rather than planting a crop on converted wetland. participants and USDA program agencies. The Such a clause would require a more complete inven- inventory is produced with the various tory because conversions for other purposes may Swampbuster designations marked within each include citrus, pasture, hayland, or other agricul- delineation to signify the restriction that is placed tural resources. The inventory now focuses on iden- on the identified wetland or converted wetland Table. Summary of use, maintenance@ and improvements of various wetland designations. Wetland designation Use Maintenance Improvement Prior Conversion (PC@- Produce agricultural Yes Yes converted before 23 commodities December 1985, but not abandoned Farmed wetland (FW)- May be farmed as it was May maintain the de- None still meets the wet- before 23 December gree of drainage that land criteria, includ- 1985 existed before 23 De- ing seasonally cember 1985 ponded wetlands, sea- sonally flooded wet- lands, potholes, and playas Wetland (W@-includes May be used to produce None None natural conditions agricultural commodi- and abandoned wet- ties when weather lands permits without re- moving woody vegeta- tion Commenced conversion Same as prior conver- Yes Yes (CC) sion when completed Third party Produce agricultural May maintain the de- None, unless determined commodities gree of drainage that by the Soil Conservation existed as of date of Service to have minimal third party action effects Converted wetland Production of agricul- None None (CW)--converted tural commodities after 23 December will cause a person to 1985 be ineligible for USDA benefits Minimal Effect UAW) Produce agricultural As per minimal effect As per minimal effect agree- commodities agreement ment Artificial Wetland Produce agricultural Yes Yes (AW)-including irri- commodities gation-induced wet- land NATjoNAL PRwRAw 99 (Table). To date, the SCS has provided inventory FSA exemptions. Also, some wetlands are in- information to very few other users. However, the cluded in lands designated as "prior converted inventory and related products (e.g., hydric soils cropland." Such lands may be subject to other lists, soil maps, and ASCS slides) are valuable wetland laws or authorities even though they are tools for other agencies or interested parties who exempt from FSA. Users should consult with SCS wish to make wetland determinations. State conservationists before using the SCS Wet- The SCS Wetland Inventory is designed pri- land Inventory to become aware of limitations of marily for FSA; therefore, users should be aware the inventory and to understand how FSA wet- of its advantages and limitations. Because the land and converted wetland designations were SCS Wetland Inventory uses various tools and applied. works from information available at a compara- tively large scale, it is very detailed in the wet- Reference lands and converted wetlands it identifies. Gener- ally, the SCS Wetland Inventory identifies more Federal Interagency Committee for Wetland Delineation. wetlands than other inventories. However, in 1989. Federal manual for identifying and delineating most instances, the SCS Wetland Inventory iden- jurisdictional wetlands. Cooperative technical tifies only FSA wetlands and converted wetlands. publication of the U.S. Army Corps of Engineers, U.S. Environmental Protection Agency, U.S. Fish and Other lands meeting wetlands criteria (e.g., arti- Wildlife Service, and USDA Soil Conservation Service. ficial wetlands) may not be identified because of Washington, D.C. 76 pp. 100 BioLoGicAL REPoRT 90(18) Appendix. Prairie Soils Regions'Wetland Mapping Conventions for the 1985 Food Security Act (FSA) POTHOLES AND SATURATED - PRAIRIE SOILS (WISCONSIN GLACIATED REGION) Wetlands will be inventoried using the following procedure which was developed to maintain consistency between field offices. This will be used as the basis for making office determinations of wetlands in the Prairie Pothole soils. It takes into consideration above normal and below normal precipitation years. The principal tools used to make the wetland inventory are: soil surveys, National Wetland Inventory (NWI) maps, black and white aerial photos, and ASCS color slides. Step l.--Review NWI maps where available. NWI maps will give an excellent overview of the wetlands in the area. All wetlands on the NWI maps will be considered wetlands for these conventions unless review of the ASCS slides fails to confirm the area as meeting vetland criteria. This could happen for the following reasons: 1. Review of the slides for all the years does not show pothole basins as having water, hydrophytic vegetation, drowned out crops. or crop color during abnormally dry or wet years. 2. The vetland has been drained since the WWI photos were taken. Look for manipulation such as ditches, new tile lines, dikes or levees. RM: Many wetlands are excluded on NWI maps because of the Fish and Wildlife Service's Farmed Wetland Policy. The SCS state office may wish to contact the FWS regional wetland coordinator to get an overview of the KWI mapping conventions. Ste2 2.--Review the soil survey. Review of the soil survey will help identify which areas of the field have potential for wetlands. Is the site on a hydric soil map unit or on a map unit with hydric inclusions, or on any wet miscellaneous areas or spots symbols such as depressional areas, riverwash, and beaches, or on water areas that meet hydric water table, ponding, or flooding criteria? See Appendix for hydric soil criteria or KFSAM 512.10-512.12. Step 3.-Review ASCS color slides (and color infrared if available) for the years 1981 to 1988 (when available). In most cases, 5-7 years will be available in most counties. Use Geological Survey or weather service climatological data in conjuction with the ASCS slides. Review the climatological data to determine those years which were above or below normal precipitation 2 to 3 months prior to the date of the slide. The slides were taken in late June or July. In most cases, flights were flown in July. NATioNAL PRorRAms lol When reviewing slides, the folloving criteria are considered indicators of a vetland and will be marked. 1. Hydrophytic vegetation in the area. 2. Water or drowned out crop (mud flat). 3. Stressed crop production due to wetness (yellow). 4. Color of crop in dry or vet years (greener or yellow). 5. Differences in color due to different planting dates. When viewing the slides, place a clear overlay on the Kodak caramate screen. Circle the wetlands with a dry erasable marker for the first year (view vettest year first) reviewed. Go to the next year slide, circle new wetlands, and place a checkmark by those wetlands that have reoccured. Repeat the process for all the years. The clear overlay is a good way to being the process. After using the conventions for a period of time, experience may allow the clear overlay to be dropped from the process. Always check for manipulation of the wetlands. Document manipulations! (See exhibit - Wetness History . . . - as an example of documentation.) For 5 or more years of slides (see exhibit 1): 1 circle, no checks, and vetland is verified by NWI map, possible vetland, review weather data to make a determination. The NWI must be reviewed. If the area with 1 circle and no checks cannot be verified by NWI, the area is not a wetland. I circle and 1 check and verified by NWI area is a vetland. If area is not verified by NWI, area is a probable vetland, review weather records to help make the determination. I circle and 2 or more checks, area is vetland whether or not verified by NWI. If area shows up on NWI map but does not show on any years of the ASCS slides, area is not a vetland. Check for vetland manipulations. For 4 or less years of slides: 1 circle, no checks, and verified by NWI, area is wetland 1 circle and no checks, and not on NWI area is possible wetland, check weather records and prior manipulations to help make a decision. I circle and 1 or more checks, area is a wetland whether or not verified by NWI. No circles or checks from ASCS slides and on NWI, area is a possible wetland, check weather records. A field check may be necessary. Step 4.--The vetland boundaries will then be transferred to in ASCS 8 inch/mile map or other suitable base map (aerial photo). This transfer is 102 BioLoGicAL REPoRT 90(18) more accurately done by projecting the ASCS slide on the ASCS map and outlining the wetlands. The wetlands will be delineated and labeled with a "W.11 Converted Wetlands will be recorded with a "CW." Those potholes located In cropfielde,vhere drainage activities are evident before December 230 1985, but have not completely drained the potholes and they still meet vetland criteria but are farmed, will be recorded on map as a Famed Wetland 1OFW' " Undrained potholes in prairie soils with herbaceous vetland plants or wetlands farmed under natural conditions will be shown as a vetland "W. Saturated Prairie Soils that meet vetland criteria, but have not been manipulated (except farmed under natural conditions), are wetlands "W." Artificial wetlands "AW" may be difficult to determine with this process. Farmer information or an onsite visit may be necessary. Step S.-The district conservationiat will review the wetlands inventory and any other pertinent information available. A field trip will be taken only if necessary to check questionable wetlands. The appropriate FSA wetlands determination will be documented on the official ASCS map (photo) and SCS-CPA-026. Pertinent supporting data will be added to the case file. Scope and effect of the existing drainage on famed wetlands "EV" will be documented. NATioNAL Programs 103 GUIDELINES FOR qwEn6qm 2qULTNqEATIONS POTHOLES AND SATURATED - PRAIRIE SOILS yes or no - 8qHydric soil 2. yes or no - Does vetla0qnd show up on NWI? 3. 0 = Circle vetland first year observed 4. 2qV- Checkmqark for each subsequent year observed S. Outline boundaries of wetland and enter symbol EXAMPLE: Five (5) or more years of ASCS slides ASCS HYDRIC q811-qmqi SLIDES q0 SOILS 0qNWI B q& W q8q1-q8q8 STATUS q1 Yes Yes No 0 or X Posaible-Check weather records* No No No 0 None 2 Yes No Yes 0 Probable-Check weather records* Yes Yes 0 Wetland 3 Yes Yes Yes 0 Wetland or or or or .No No No more X = An X is used when team member has a question on a call. District conservationist needs to make decision. Not used often. *Field checks may be needed. REGIONAL,kND FEDERm-STATE COOPMUTIVE PROGrUMS 105 Regional and Federal-State Cooperative Programs Coastal Mapping Programs at the U.S. Fish and Wildlife Service's National Wetlands Research Center by James B. Johnston and Lawrence R. Handley U.S. Fish and Wildlife Service National Wetlands Research Center 1010 Gause Boulevard Slidell, Louisiana 70458 ABSTRACT-Over the past 10 years, the U.S. Fish and Wildlife Service's (FWS) National Wetlands Research Center (center; formerly the National Coastal Ecosystems Team) has been continuously involved in the production of maps for use by coastal decision makers. The types of maps produced by the center have been national, regional, or local in scope depending on user needs. Map scales have ranged from 1:24,000 to 1:250,000. Themes depicted have included biological resources, including wetlands and seagrasses; upland habitat or land use; water resources such as water quality, bathymetry, and salinity; cultural features such as ownership, archaeological sites, and dredge-spoil disposal areas; and soils and landforms. We present overviews on the various mapping programs of the center. We highlight efforts such as the ecological inventories of the Atlantic, Gulf, and Pacific coasts; the ecological characterization atlases of the Gulf of Mexico; and the large scale (1:24,000) habitat maps of various coastal regions of the United States. Center methods and techniques are discussed, including the collaborative efforts between the center and FWS's National Wetlands Inventory for updating wetland maps and adding upland and seagrass bed delineations to inventory maps. We also make recommendations for future coastal ecosystem mapping programs that use conventional and automated mapping methodologies, such as geographic information systems and image processing. The National Wetlands Research Center (cen- All of our mapping projects are developed as ter) of the U.S. Fish and Wildlife Service (FWS) has special interest programs (e.g., Louisiana land loss an ongoing program in habitat mapping of wet- or seagrass mapping) in cooperation with other lands and uplands. We cooperate with the National Federal and State agencies, such as the U.S. Army Wetlands Inventory (NWI), using its processes of Corps of Engineers (COE), the U.S. Environmental photointerpretation, quality control and assurance, Protection Agency (EPA), and the Louisiana De- and distribution. We differ from NWI in mapping partment of Natural Resources. Technical assis- biological data and resources at other scales of tance is also provided from within FWS. We have 1: 100,000 and 1:250,000, in adding upland habitat added uplands to wetland maps and developed to the wetland maps, and in developing time-se- criteria for the incorporation of the additional up- quenced mapping for habitat trend analysis. land categories. We have provided updates of hab- 106 BjowrsicAL REPoRT 90(18) itat maps that the center completed previously. as well as additional subcategory identifiers such From the sequential dates of mapping we can look as rice fields, parks, cemeteries, golf courses, spoil at a trend analysis of habitat loss and gain. We use areas, and transportation corridors. The North the Cowardin et al. (1979) classification system as American Waterfowl Management Plan suggests the primary criteria for wetland delineation looking at rice fields as habitat for wintering through the various systems, subsystems, classes, ducks, and FWS is interested in the breakdown of and subclasses. However, if additional information forested land into scrub-shrub or evergreen, and is available, we are able to add modifiers, which deciduous forests as habitats suitable for red- NWI is usually not able to do because data are not cockaded woodpecker (Picoides borealis) nesting. available. For example, for coastal Louisiana we One of the great needs in habitat mapping is the are adding a salinity modifier on the habitat maps. addition of the upland classification to properly Through additional coordination at the local level, analyze and assess habitat changes and processes. we are able to gather this type of information for In assessing the habitat changes for an area, it is many of our special projects. We have completed a difficult to understand "Where did the wetlands number of projects related to habitat mapping at go?" if uplands are the only category. To under- the 1:24,000 scale. Coastal Louisiana was mapped stand the processes of change in the landscape, for for 1956, 1978, and 1988, with an update in 1983 example, it is necessary to know what type of for the lower Mississippi River Delta and the Ter- uplands replaced the wetlands. Can one assume reborme Marsh area. In 1985, we coordinated with that wetlands were filled in or drained for urban the National Aeronautics and Space Administra- development or could they have gone into upland tion (NASA) to produce aerial photography of agricultural land, rangeland, or forest? Around coastal Louisiana. At that time, only 11 habitat San Francisco Bay, with its wholesale develop- maps were developed for the State of Louisiana to ment, it is assumed that any loss of wetlands went compare data with pre-Hurricane Juan satellite into the uplands category, but this is not always imagery. We mapped coastal Texas for 1956 and true. Finally, wetlands lose acreage to other wet- 1979, and mapped 10 quadrangles for 1983. We land categories as the water regimes change. mapped habitats in coastal Mississippi in 1956 and The addition of the upland categories helps us 1979. For Alabama, 32 quads for the Mobile Bay understand the overall picture of habitat change area were completed for 1956 and 1979. For the in a particular area. In the San Francisco Bay, for west coast of Florida, we completed selected quads example, wetlands certainly lost many acres, but of the Panhandle area as well as 26 quads of Tampa the greatest loss of all occurred in upland agricul- Bay in the mid-1950's, 1962, and 1982. For San tural land. The land of the market gardens, truck Francisco Bay, we completed 20 quads of the south farming, and alfalfa around the San Francisco Bay bay for the mid-1950's, 1976, and 1985. For the lost almost four times as many acres as in the north bay, an additional 24 quads were completed wetlands. The upland categories are very impor- for 1976 and 1985; another 63 quads were com- tant, and the need is certainly present for the pleted for only 1985 for a surrounding area of the development of a comprehensive and systematic bay. These are most of the habitat mapping projects uplands classification system that will comple- that we have completed with multiple year up- ment the existing wetlands classification. dates. We have developed projects to analyze habitat The upland classification that we use is pat- trends and changes. For example, in San Francisco terned after Anderson et al.'s (1976) classification, Bay, we have put together a habitat change map which is used by the U.S. Geological Survey for the mid-1950's, 1976, and 1985. We have done (USGS) in its land use mapping. However, the the same thing for the lower Mississippi Delta for upland classification is gradually evolving, just as 4 years (1956, 1978, 1983, and 1988). In addition, the wetland classification has changed over time. in several of our mapping projects we are develop- The upland classification we use is now commonly ing the wetland maps to include selected indicator referred to as "the Handley upland classification" species. For example, in the lower Mississi i ppi because we have gradually added more identifiers River Delta, we are adding a habitat modifier for to the uplands, as the special projects dictate two particular species-spartina alterniflora greater detail and varying needs. (smooth cordgrass) and Phragmites australis Examples of these identifiers include urban (common reed). Spartina is primarily an indicator forest, rangeland, agricultural, and barren lands, of salinity, and Phragmites an indicator of fresh or REGioNAL AND FEDERAL-STATE CooPERATrm PROI,-RAms 107 brackish water. We are also interested in deter- bay's waterfowl management plan, by EPA's Es- mining how much loss or gain has taken place for tuarine Program assessments, in two court cases, each species between 1978 and 1983, and between by the California attorney general's office, and in 1983 and 1987. at least a dozen other projects, programs, and The mapping of the Chandeleur Islands is pri- studies. marily seagrass mapping. We have photography Special projects completed or ongoing at the from three dates (1978, 1982, and 1987) that has center, are generally done in the interest of FWS. been interpreted for habitat. In addition, in 1990 In particular, we provide technical assistance to we will interpret photos from April 1969, October Fish and Wildlife Enhancement Offices, regional 1969, November 1988, and June 1989. Also, we are offices, or national wildlife refuges. For example, acquiring aerial photography of the Chandeleur we are mapping seagrasses in Perdido Bay, in Islands on a quarterly basis to study seasonal Florida and Alabama, for 1940, 1978, and 1987, for variations in the seagrass cover. Although this the Panama City Enhancement Office. On Eglin project was undertaken originally in conjunction Air Force Base in Florida, we are developing eco- with other center studies on the redhead (Aythya logical community maps for the U.S. Air Force and americana) population that winters at the FWS to use in surveying red-cockaded woodpecker Chandeleur Islands, it has evolved into a seagrass habitats for active colonies. photointerpretation study of its own. The information we have collected has been The Louisiana Coastal Zone Project was per- used to develop digital data bases that can be formed for the State of Louisiana to update data entered into the center's geographic information following Hurricane Juan. Louisiana is using system (GIS) to implement natural resource in- these habitat maps to compare with Landsat The- ventories, habitat trend analyses, and carto- matic Mapper Simulator digital data acquired graphic modeling projects. We work with other before Hurricane Juan to analyze the hurricane's Federal and State agencies in need of the habitat effects on the breakup of marshes. In coastal Lou- maps and the digital data to conduct their work. isiana, we are also photointerpreting and map- These other agencies include the National Park ping uplands and wetlands in 330 quads. This Service, COE, EPA, and Louisiana's Department project will take about 3 years to complete. At of Natural Resources. We have developed a digital present, we are in Phase I, which is the photo- data base ofthe habitat maps for coastal Alabama, interpretation of 110 quads. Phase II and Phase Louisiana, Mississippi, Texas, and portions of the III will be the completion of the photointerpreta- Gulf Coast of Florida. In addition, we have digital tion of the remaining 220 quads over the next data for other selected areas ofthe country includ- 2 years. This project will provide an update using ing New Jersey, the lower Chesapeake Bay, the 1988 photography to add to our existing 1956 and St. Lawrence Seaway, and the San Francisco Bay. 1978 data bases of coastal Louisiana. We have also been involved in the development In Mobile Bay, Alabama, we are mapping 26 quads of maps for atlases and inventories at scales of to update the 1956 and 1979 wetland maps. In the 1:100,000 and 1:250,000. In 1978, we began devel- San Joaquin Valley, we are mapping 83 quads; 26 of oping the first of the ecological atlases for regions these quads focus on uplands, and the other 57 are an of the Gulf Coast. In all, four atlases were devel- update of wetlands and uplands. This mapping is for oped: the Mississippi Deltaic Plain Atlas, the the San Joaquin Valley Drainage Program and will Texas Barrier Islands Regional Atlas, the Coastal be used in analyzing the Kesterson National Wildlife Alabama Ecological Atlas, and the Florida Ecolog- Refuge selenium problem. ical Atlas. A fifth atlas, the Chenier Plain Ecolog- To show the overall aspects of some of these ical Atlas, is in progress; it will fill in the final gap projects, we not only had the habitat maps for the along the Gulf Coast. The mapping for each of San Francisco Bay Project for several dates, but these atlases is completed on 1:100,000 USGS we also developed two reports that provided an base maps. Five topics per map are displayed: analysis of the habitat trends in the south bay and biological resources, socioeconomic features, soils the north bay, a report on the comparison of fish and landforms, oil and gas infrastructure, and and wildlife use of a natural marsh with an arti- climatology and hydrology. For each topic we ac- ficial marsh, and two large-format habitat maps cumulated a great deal of information from many of the bay area. The information produced in our resources in mapped form, text format, site visits, trend analysis is being extensively used in the meetings with regional experts, and reviewers' 108 BIOLOGICAL REPORT 90(18) comments. The reports that were produced as part rookeries and nesting sites have either disap- of these overall projects include bibliographies of peared or changed locations. biological and socioeconomic literature, informa- Representatives from oil companies and State tion synthesis, map narratives, and some special and Federal agencies met on 7 December 1989 to reports on modeling efforts, ecological community discuss the need for a comprehensive, updated profiles, and seagrass atlases. The Minerals Man- biological resources mapping program. Nation- agement Service, EPA, and various State agencies wide, the greatest need in thematic mapping is to were instrumental in funding, collecting data, update FWSs ecological inventory. One sugges- writing reports, and reviewing the atlases and tion made by the center is that the 1:250,000-scale reports. maps do not lend themselves well to detail for The ecological inventories were completed by site-specific analysis, oil spill risk assessment, or the center in 1984; they cover the Atlantic, Gulf digitizing. We propose that the ecological inven- and Pacific coasts, and the lower Mississippi Val- tory be updated using the 1:100,000-scale USGS ley. The scale of the maps we used was 1:250,000; maps as the mapping base. The USGS 1:100,000 the maps included an inventory of a single digital line graphs are completed for the country. topic-biological resources. Some resources we By doing this we would provide additional theme mapped are fish spawning areas, bird rookeries, overlays of political boundaries, hydrology, and of bird nesting areas, endangered species habitats, the transportation network. This scale of maps major natural waterways, turtle nesting areas, would provide a manageable and usable product and State and Federal wildlife refuges and man- that would be more meaningful to planners, envi- agement areas. The ecological inventory maps ronmental consultants, and analysts, and is more were conceived as aids to site planning of thermal specific and detailed for oil spill cleanup, risk power plants along the Atlantic Coast; their scale, assessment, site planning, and permit analysis. however, made specific site planning difficult. Another aspect of mapping the center provides Overall use has far overshadowed the deficit-, is the coordination and organization of flights to these maps have become extremely valuable aids acquire aerial photography and digital data over for regional environmental impact assessment many areas. We organized a consortium of Fed- and environmental analysis, oil spill risk assess- eral and State agencies to provide the funding for ment, oil spill sensitivity, and oil'spill cleanup a flight of coastal Louisiana, Mississippi, Ala- planning. bama, and a portion of the western Florida Pan- Several entfites have developed products based handle. These groups included FWS, EPA (At- on these maps. Resource Planning Institute of lanta Region and Dallas Region), COE (New Columbia, South Carolina, an oil industry consul- Orleans District and Mobile District), and the tant, has developed a set of maps of the coastal States of Alabama and Mississippi. Nearly 3,000 United States; these maps deal with the sensitiv- line-miles were flown resulting in the collection of ity of particular coastal segments to oil spill 1,000 colorinfrared photographs at 1:65,000 scale, cleanup activities. MERG, an oil industry consor- tium working through consultants such as and airborne Thematic Mapper Simulator (TMS) Coastal Environmental, Inc., has developed sets digital data. The coastal Louisiana, Mississippi, of maps that delineate segments of the coastal and Alabama flights were completed between 6 United States that should be protected from oil November 1988 and 30 March 1989. Because of spill impact on a priority basis. S.L. Ross of Can- the success of this flight, we were asked to orga- ada has developed a computerized data base that nize a similar group to fund a flight of coastal many oil companies are using on microcomputers Texas for the fall of 1989. This consortium in- for oil spill risk assessment and oil spill cleanup. cluded FWS, EPA (Dallas Region), COE (Galves- All of these products have one major flaw-the ton District), and the Soil Conservation Service. data used to develop the maps. In particular, the The coastal Texas Right, flown by NASA out of biological resource information taken from the Ames Research Center at Moffett Field, Califor- ecological inventory maps and ecological atlases nia, encompassed about 3,000 flight-line miles, is outdated and in some cases highly generalized. took a thousand 1:65,000-scale col6r-infrared pho- For example, the priority resources to be protected tographs, and collected TMS digital data. The or cleaned may not be in those locations any Texas coast was flown between 27 November and longer. For instance, 35% of the Gulf Coast bird 15 December 1989. REGIONAL AND FEDEML-STATE COOPEFUTIVE PROGMMS 109 References data. U.S. Geol. Surv. Prof Pap. 964. 28 pp. Cowardin, L. M., V Carter, F C. Golet, and E. T. LaRoe. Anderson, J. R., E. E. Hardy, J. T. Roach, and R. E. 1979. Classifications of wetlands and deepwater Witmer. 1976. A land use and land cover habitats of the United States. U.S. Fish Wildl. Serv., classification system for use with remote sensor FWS/OBS-79/31.103 pp. REmoNAL AND FEDERAL-STATE CooPERATivE PRoGRAms ill Monitoring Seagrass Distribution and Abundance Patterns: A Case Study from the Chesapeake Bayl by Robert J. Orth, Kenneth A. Moore, and Judith F Nowak Virginia Institute of Marine Science School of Marine Science College of William and Mary Gloucester Point, Virginia 23062 ABSTRACT.-Seagrasses, or submerged aquatic vegetation (SAV), have been mapped in the Chesapeake Bay five times between 1978 and 1987 with standard aerial photographic techniques, resulting in annual reports on SAV distribution. Acquisition of the vertical photography at a scale of 1:24,000, adhering to strict quality-assurance guidelines based on sun angle, tidal stage, cloud cover, wind speed, and season, has produced excellent, high-contrast imagery delineating beds of SAV from adjacent, unvegetated areas. Ground-truthing data from various State, Federal, and public organizations have corroborated the photographic data base. Digitized bed outlines resulting from photointerpretation of the imagery onto 1:24,000-U.S. Geological Survey topographic quadrangles have been stored on a Virginia Institute of Marine Science geographic information system (GIS). A report summarizing the photographic. and ground survey data is produced each year. Results from these surveys have shown distinct changes in the distribution and abundance of SAV in different areas in the bay over the last 10 years. The amount of SAV has increased 21% from 1978 to 1987 with some areas showing rapid increases in less than 5 years. The success of these annual surveys in the Chesapeake Bay indicates that aerial photographic techniques can be used to delineate spatial and temporal patterns of seagrass communities, as well as those communities comprised of brackish-water species. Appropriate GIS systems can be employed to assess historical trends at any location. Seagrasses are submersed vascular plants dense assemblages of vertebrates and inverte- found in shallow-water coastal and estuarine en- brates and often serve as nursery areas for many vironments throughout the world. There are about commercially important species, such as the bay 50 species growing in a wide variety of sediments scallop, Aequipectin irradians. Seagrass meadows from the intertidal zone to depths of 10 m. In turbid are important in nutrient cycling between sedi- estuarine environments, such as the Chesapeake ments and the overlying water, and they contrib- Bay, seagrasses are not found at depths below 2 In ute to the detrital food chain. Only a few groups of at mean low water (MLW), whereas in less turbid animals (e.g., geese, dugongs, manatees) actually areas, such as the Caribbean Sea, seagrasses can consume seagrassses; however, the attached epi- be found at depths of 50 m or more. phytes are food for invertebrates (e.g., gastropods, Seagrasses, like their emergent wetland coun- amphipods), which in turn are food for secondary consumers. terparts, serve many different functions. Because In the continental United States, seagrasses are they baffle currents and stabilize sediments, ex- present in every coastal State except Delaware, tensive seagrass beds adjacent to shorelines can Georgia, and South Carolina, although quantita- reduce shoreline erosion. Seagrass beds support tive estimates on distribution and abundance in many States are generally lacking. Table 1 pres- ents a summary of data currently available on the Contribution No. 1576 from the Virginia Institute of Marine abundance of seagrasses as compared with total Science, Gloucester Fbint, Virginia 23062. area of salt marsh. Seagrass coverage in many 112 BioLoGicAL RFPoRT 90(18) Table 1. Salt marsh and seagrass coverage ing seagrass distribution and abundance is criti- (hectares) by Stateo (modifted from Orth and cal for making quantitative assessments of losses, van Montfrans 1990). No data are available for thereby increasing our understanding of factors seagrasses in those coastal States not listed. controlling growth and distribution. Salt marsh Seagrass State (reference@ (referencel) Development of a Seagrass New York 10,810, 78100 10 Monitoring Program: A Case New Jersey 83,9892 12:62A 1,11 Study of Chesapeake Bay Delaware 26,183 3 0 Virginia-Maryland 145,813 3,4 1735312 1 13 A decline of seagrass and brackish-water spe- North Carolina 64)291 5 80,972 cies throughout Chesapeake Bay in the late 1960's South Carolina 149,580 1 0 Georgia 151,538 1 0 14 and 1970's (Kemp et al. 1983; Orth and Moore Florida-Atlantic Coast 38,826 6,7c 2800 14 1983b, 1984) led the U.S. Environmental Protec- Florida-Gulf Coast 137)455 8 9131700 14 tion Agency to initiate a major research program Alabama 11,855 9 12,300 14 in 1978. This program determined the distribution Mississippi 24,919 9 2' 000 14 and abundance of submersed bay grasses and the Louisiana 720,648 6 4100 14 factors that contributed to their decline. The great- Texas 174,899 68:500 est loss of vegetation occurred in the upper and a Wetland areas identified as containing salt-tolerant middle sections of the bay and tributaries (Fig. 1). vegetation (categorized as "salt marsh" or "nonfresh" in data The results of the studies indicated that nutrient reports or published papers) were used and listed in the totals enrichment and high levels of turbidity were asso- above. ciated with the declines in a number of areas b 1, Field et al. 1988; 2, Tiner 1985a; 3, Tiner 1985b; 4, Silberhorn, Virginia Institute of Marine Science, personal (Kemp et al. 1983). communication; 5, '17iner 1977; 6, Reyer et al. 1988; 7, Perry A 1987 agreement signed by the governors of 1984; 8, Roach et al. 1987; 9, E. C. Pendleton, U.S. Fish and Maryland, Pennsylvania, and Virginia, and the Wildlife Service, personal communication; 10, Macomber and mayor of Washington, D.C., committed the States Allen 1979; 11, Dennison, et al. In press; 12, Orth et al. 1989; 13, Ferguson et al. 1988; 14, Iverson and Bittaker 1986. to develop management policies for the living re- Includes 34,540 ha of mangroves listed in Perry 1984. sources of the bay. A committee of Federal, State, and university scientists and managers developed a management policy to protect, enhance, and re- States may be underestimated because of the lack store seagrass and brackish-water species (collec- of quantitative mapping studies. Seagrass moni- tively referred to as submerged aquatic vegetation toring programs are rare because of the inherent or SAV) in the bay. This policy was approved and technical difficulties and cost in censusing these signed in July 1989. An implementation plan for underwater populations (Orth and Moore 1983a). the SAV management policy is being developed by Some seagrass beds have been mapped success- the committee. fully with remote-sensing techniques such as low- Surveys of SAV and brackish-water species level or satellite photography, or through field have revealed several large changes in distribu- surveys including transects or randomized sam- tion and abundance over a short time. Therefore, pling (Orth and Moore 1983a; Walker 1989). How- one requirement of the SAV management policy is ever, most State and Federal agencies have fo- to develop a monitoring program that will annu- cused their efforts on emergent wetlands. The U.S. ally determine the distribution and abundance of SAV. This program will be implemented by using Fish and Wildlife Service's National Wetlands In- low-level, vertical aerial photographs and ground ventory is one such mapping effort. surveys. This survey methodology was developed In recent decades, seagrass declines have oc- over a 10-year period in Chesapeake Bay. In aerial curred worldwide (Kemp et al. 1983; Orth and photographs, seagrasses-under appropriate en- Moore 1983b; Cambridge and McComb 1984; vironmental conditions-generally have a signa- Neverauskas 1987). The magnitude of these ture distinct from adjacent, unvegetated areas. losses, in many cases, has been difficult to assess Aerial photographs also provide a synoptic view of because of inadequate data on distribution and baywide patterns for future analysis. The first abundance patterns before the decline. Monitor- baywide survey to use low-level, vertical aerial RFGioNAL AND FEDERAL--STATE CooPERATivE PRoGRAms 113 77- 100' 14 - SUSOVENANNA 760105 U FLATS KILOMETERS 0 10 20 30 40 BALTIMORE cr_ W a- a. WASHINGTON 13 390 00, W -J 10 9 12 7 180 8 00 3 ku QL 5 2 W 4 0 14 cw I., X-11 370 00 Fig. 1. Chesapeake Bay and tributaries showing major declines of submerged aquatic vegetation (SAV-, crosshatched area) during the 1960's and 1970's, and showing areas where SAV was still abundant (stippled area; reprinted with permission of Science; see Orth and Moore 1983b). 114 BiowmcAL REPoRT 90(18) photography was conducted in 1978 (Orth et al. elodea), Ceratophyllum demersum (coontail), Na- 1979; Anderson and Macomber 1980). Subsequent jas guadalupensis (southern naiad), and Vallisne- baywide surveys were conducted in 1984-87 and ria americana (wild celery) are less tolerant of high 1989 with the same methodology (Orth et al. 1985, salinities and are found in the middle and upper 1986, 1987, 1989). Additional aerial surveys were sections of the bay and tributaries. Ruppia mari- conducted in the lower bay in 1974, 1980, and tima (widgeon grass) is tolerant of a wide range of 1981, and historical aerial photographs were used salinities and is found throughout the bay. About to map the lower western shore in 1971 (Orth and 11 other species are occasionally found in the mid- Gordon 1975). dle and upper reaches of the bay and tidal rivers (Table 2). Hydrilla verticillata (hydrilla) was intro- duced into the Potomac River in 1981 and rapidly Submerged Aquatic became abundant in the tidal freshwater section. Vegetation Species Aerial Photography and Ground Ten SAV species are commonly found in the Truthing Chesapeake Bay and its tributaries. The limits of a species' distribution are determined by its salin- SAV photographs are obtained by using stan- ity tolerance (Orth and Moore 1981). Zostera ma- dard aerial mapping cameras, with either black rina (eelgrass), tolerant of salinities as low as and white or color film (both film types have been 10 o/oo, is abundant in the lower portion of the bay. used effectively in the monitoring program). Pho- Myriophyllum spicatum (water milfbil), Potamo- tographs are taken at an altitude of about 12,000 geton pectinatus (sago pondweed), Potamogeton feet, yielding a 1:24,000 photographic scale. Cov- perfoliatus (redhead grass), Zannichellia palustris erage includes all areas known to have SAV and (horned pondweed), Elodea canadensis (common areas that could potentially support SAV (i.e., Table 2. Species of submerged aquatic plants found in Chesapeake Bay and tributaries (from Orth et al. 1989). Family Species Common name Characeae (muskgrass) Chara braunii Muskgrass Chara zeylanica Nitella flexilis Potamogetonaceae (pondweed) Potamogeton perfoliatus bupleuroides Redhead grass Potamogeton pectinatus Sago pondweed Potamogeton crispus Curly pondweed Potamogeton pusillus Slender pondweed Ruppia maritima Widgeon grass Zannichellia palustris Horned pondweed Zostera marina Eelgrass NaJadaceae Najas guadalupensis Southern naiad Najas gmcillima Naiad Najas minor Naiad Hydrocharitaceae (frogbit) Vallisneria americana Wild celery Elodea canadensis Common elodea Egeria densa Water-weed Hydrilla verticillata Hydrilla Pontedariaceae (pickerelweed) Heteranthera dubia (= Zosterell dubia) Water stargrass Ceratophyllaceae (coontail) Ceratophyllum demersum Coontail Trapaceae Trapa natans Water chestnut Haloragaceae (water milfoil) Myriophyllum spicatum Eurasian water milfoil REGIONAL AND FEDERAL-STATE CooPERATivE PRoGRAms 115 generally all areas where water depths are less Table 3. Guidelines followed during acquisition of than 2 in at MLW), as well as land control points. aerial photographs. Survey flight lines are prioritized by area and are flown when the standing crop for the dominant Tidal stage--Photography is acquired at low tide, species is at its peak. General guidelines govern- � 0-1.5 feet, depending on overall water clarity and ing mission planning and execution have been tidal regime of the area, as predicted by the Na- established; these guidelines address tidal stage, tional Ocean Survey tables. plant growth, turbidity, sun elevation, wind, Plant growth-Growth stages must ensure maxi- water and atmospheric transparency, sensor op- mum delineation of SAV, and when phenologic eration, and plotting (Table 3). These guidelines stage overlap should be greatest. ensure that photographs will be obtained during Sun angle-Surface reflection from sun glint must optimal conditions for detecting SAV, thus aiding not cover more than 30% of frame. Sun angle accurate photointerpretation. should be between 20' and 40* to minimize water Field surveys of SAV communities are done by surface glitter. At least 60% line overlap and 20% a number of State and Federal agencies and per- side lap are used to minimize image degradation sons in Maryland and Virginia, including the U.S. due to sun glint. Geological Survey (USGS), Maryland Department Turbidity--Clarity of water must ensure complete de- of Natural Resources, and Chesapeake Bay Foun- lineation of grass beds. This is visually determined dation. Some surveys are conducted independent from the airplane to ensure that SAV could be seen of the aerial mapping program; these include by the observer. those surveys associated with SAV restoration Wind-Photography is acquired during periods of no programs in Maryland and Virginia, whereas wind or low wind. Offshore winds are preferred other surveys support the aerial survey by check- over onshore winds when wind conditions cannot ing SAV beds that were mapped the previous year. be avoided. All data are synthesized in a report of the annual Atmospherics-Photography is acquired during peri- mapping program. ods of no haze or low haze or clouds below aircraft. There should be no more than scattered or thin bro- Mapping Process ken clouds, or thin overcast above aircraft, to en- sure maximum SAV-to-bottom contrast. The USGS's 7.5-min topographic quadrangles Sensor operation-Photography is acquired in the are used as a basis for mapping SAV beds from vertical mode with 5* tilt. Scale/altitude/film/focal aerial photography, digitizing SAV beds, and com- length combination must permit resolution and piling SAV bed-area measurements (Fig. 2). Pho- identification of about 1 m2 area of SAV (surface). tointerpretation of SAV beds requires all avail- Plotting-Each flight line includes sufficient identifi- able information, including knowledge of distinct able land area to ensure accurate plotting of grass aquatic grass signatures on film, ground surveys, beds. and low-level aerial reconnaissance surveys. De- lineation of boundaries of SAV beds onto topo- graphic quadrangles is done by superimposing the ber (1 = very sparse or <10% coverage; 2 = sparse appropriate mylar quadrangle onto the appropri- or 10-40% coverage; 3 = moderate or 40-70% ate photograph. A best fit is obtained where minor coverage; 4 = dense or 70-100% coverage) corre- scale differences are evident between the photo- sponding to the density categories. Additionally, graph and the mylar quadrangle. Shoreline each distinct SAV unit is assigned a two-letter changes are noted on the quadrangle if significant designation unique to the map. Subsections of shoreline erosion or accretion has occurred since beds are further identified as being part of a USGS publication of a map. contiguous bed by the addition of a code unique to In addition to delineating the boundaries of the that bed. SAV bed, the percent of cover within each bed is estimated by using an enlarged crown-density scale similar to that developed for estimating for- SAV Perimeter Digitization and Area est crown cover. Bed density is classified into one Calculation of four categories based on a subjective compari- son with the density scale. Either the entire bed, The perimeters of all SAV beds mapped from or subsections within the bed, are assigned a num- aerial photographs are digitized using a Numonics 116 BiowwAL REPoRT 90(18) 4 @5 _WSOJA NANNA ArS B IMORE, 14 S, 1. 17 .1 '0 v2ql 25 5 \-@ /8- "' ASHIN N N 158 31@@ Ln 159 38 '60 35 I 7A 40 @41 14"3' 1 46 C@,- 0 4 61 411@ (X 51,,1 ' @3 so 52- 53 54 58 '162 _60 65 to '69 02 87 89 90 91 92 t@293 17 96 94 x 101 103 ",105 107 108 119 121 120 x 122 L.19 124 121 126 1271 128 129\' '30 133 111 139 141 135 -/fn 14 146 .4 N 2 6 @40. 0 20 148 149 @ KILOMETER f 155 15! 156 157 I Fig. 2. Chesapeake Bay-locations of topographic quadrangles used in submerged aquatic vegetation monitoring program. REGioNAL AND FEDERmSTATE CooPERAnvE PRwRAms 117 Model 2400/2200 Digitablet Graphics Analysis amount of SAV that had been lost in the Chesa- System with a resolution of 0.00254 cm and an peake Bay up to that time. Qualitative assess- accuracy of 0.0127 cm. Coordinates are transmit- ments indicated that there may have been in ted to a PRIME 9955 computer for area calculation excess of 50,000 ha, at peak levels (Bayly et al. and data manipulation with a software program 1978). Thus, current SAV populations may be less developed at the Virginia Institute of Marine Sci- than half of those that existed 20 years ago. Sev- ence. The area of each bed is reported as a mean of eral areas exemplify the changes described pre- three trials. The range of these three trials is not viously and are discussed in more detail to provide to deviate from the mean by more than 5%. an additional perspective on the changes that The perimeter of each SAV bed is defined by a have occurred in the bay. polygon with a linear point density of 60 per chart The lower eastern shore (section 14) has had centimeter (5 m ground resolution). The total abundant seagrass since 1978 (Fig. 6). Zostera ma- number of points defming any SAV bed is depen- rina and Ruppia, maritima are the dominant spe- dent on overall bed size. The SAV bed perimeter is cies in this area. Because this area is close to the stored as X and Y coordinates in centimeters from mouth of the Chesapeake Bay, the generally less the quadrangle origin. Perimeters are later con- turbid water apparently allows for a much greater verted to latitude and longitude. depth penetration of light and thus a greater depth A standard operating procedure was developed distribution of SAV as compared with western to aid orderly and efficient processing of data, and shore areas (Orth and Moore 1988a). to comply with the need for consistency, quality Seagrass in the Rappahannock River (section assurance, and quality control. These standard 16), which consists of Zostera marina and Ruppia. operating procedures include a detailed procedure maritima, was abundant along both shores in outlining 46 steps for digitization of SAV maps; a 1971. There was a rapid decline in seagrass be- 47-step checklist for editing SAV perimeter com- tween 1971 and 1974, with continued absence of puter files; a digitizer log in which all operations SAV through 1986. However, since 1987 there has are recorded and dated, and which is used to guide been a rapid increase of R. maritima in some and record editing operations; and a flowchart downriver areas (Fig. 7). This change has paral- used to track progress of all computer operations, leled similar increases observed with this species including all changes in file names. in other mid-bay areas. Submerged vegetation in the upper Potomac River was absent in 1978. However, a rapid in- Vegetation Trends in crease was observed in 1984, with continuing ex- Chesapeake Bay pansion through 1987 (Fig. 8). The abundance in 1987 was the most recorded since the early 1900's The distribution of SAV in the Chesapeake Bay and was largely due to the rapid spread ofHydrilla and tributaries has been organized into 3 zones verticillata, after its accidental introduction in and 21 sections (Fig. 3). In 1978, the first baywide 1981. Although H. verticillata is by far the most survey of seagrasses delineated 16,894 ha with dominant species in this region, 12 other species 17.8, 44.0, and 38.2% in the upper, middle, and have been reported. The reason for their reoccur- lower bay zones, respectively (Fig. 4). By 1987, rence is unknown, but may be associated with the there were 20,230 ha, a 21% increase from 1978, increase in water clarity created by the dense mats with 14.6, 45.9, and 39.2% in the upper, middle, of H. verticillata in inshore areas. The increase in and lower bay zones, respectively. From 1978 to submerged vegetation in the upper Potomac River 1987, there were relatively small changes in most may have been accelerated because of the reduc- sections of the lower bay zone, and both increases tion in the discharge of nutrients by the Blue and decreases in sections of the middle and upper Plains Sewage Treatment Plant in Washington, bay zones (Fig. 5). The increases were primarily D.C. Total suspended solids and phosphate loading in the upper Potomac River (section 11) and the have declined. Nitrification began in 1983, chang- middle reaches of the bay along the eastern shore ing the main nitrogen input from ammonia to (sections 12 and 13). Decreases were in the upper nitrate. Although no defitiite links between nutri- reaches of the bay (sections 3, 4, 5, 6, and 7). Data ent reductions and seagrass regrowth in this re- are not available for seagrass abundance in the gion have been made, these changes in discharge bay before 1978, making it difficult to estimate the could only have had positive effects. 118 BioLoGicAL REPoRT 90(18) 3940. 0 1 SUSQUEHANNA R N CHESAPEAKE Cc BAY Uj a_ a_ PATAPSCO D 3910. 0 4 5 6 3840. 0 - 8 Ljj R. 0 7 0 POT 3810. 0 - 13@ 15 3740. 0 - 14 16 LLJ 0 3710@ 0 - 20 3640@ 0 21 0 0 0 0 C 0 0 0 C) 0 0 fn (0 Fig. 3. Chesapeake Bay and tributaries showing delineation of zones (3) and sections (21) developed for discussion of trends of submerged aquatic vegetation. RmioNAL AND FEDERAI-STATE COOPERATIVE PROGRAMS 119 SAV Abundance in the Chesapeake Bay 25000 20000' Ni f-'@Wzgagv [I UPPER Fig. 4. Abundance of submerged aquatic vegetation by zone for the Chesapeake 15000. Bay and tributaries for 1978, and 1984 El MIDDLE through 1987. 10000. 0 LOWER 5000- 0- 1" 4P I" 1978 1984 1985 1986 1987 Summary and sources and groundwater inputs as well as reduc- Recommendations tion in sediment inputs, must be expanded if seagrasses are to remain a part of the Chesapeake Bay's important living resources (Orth and Moore Submerged vegetation in the Chesapeake Bay 1988b). and its tributaries has been an abundant natural Because of the importance of seagrasses to resource and, in some sections, it still is. Popula- coastal estuaries and lagoons of the United States, tions that experienced rapid declines in the 1970's and because of their vulnerability to changes in have had some recovery in the 1980's. The recov- water quality, we recommend that a major initia- ery in some sections has been substantial and may tive be undertaken to census this resource on a be due to the improved water quality from reduced nationwide basis, as is ongoing in the Chesapeake upland input of nutrients and sediments. How- Bay. For most areas we recommend that a combi- ever, large areas of the bay still have the potential nation of low-level aerial photography, flown to support seagrass populations. Thus, nutrient under strict guidelines, and ground-truth studies, reduction strategies, including point and nonpoint including permanent transects, be established to 4500- 4000- 3500- 3000- 0 1978 2500- M 1987 2000- 1500- 1000 500 0 00/1' 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 Upper I Middle I Lower Fig. 5. Abundance of submerged aquatic vegetation for the 21 bay and tributary sections for 1978 and 1987. 120 BioLoGicAL REPoirr 90(18) EASTERN SHORE HISTORICAL WINDOW 1000- PLOT OF TOTAL SAV AREA FOR ALL DENSITY CLASSES S A 800- V A R >00< E 600- >00< A H E 400 C T A R E S 200- 0 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 YEAR Fig. 6. Abundance of submerged aquatic vegetation (SAV) for a portion of the lower eastern shore of Virginia (section 14), 1978-87. examine long-term changes in species density and the cooperation and financial support of many composition. Some regions (e.g., Florida), because State and Federal organizations over the past 10 of the extent of the seagrass beds, may require years. These organizations include the U.S. Envi- high-altitude or satellite photography. However, ronmental Protection Agency, National Oceanic these baseline data are critical for the proper management of this resource, regionally as well and Atmospheric Administration!s Coastal Zone as nationally. A coordinated, cooperative program Management Program, Maryland Department of between Federal and State agencies, in which Natural Resources, U.S. Army Corps of Engi- standardized methods are used, will not only neers, U.S. Fish and Wildlife Service, the Commit- allow an assessment of the changes in distribution tee to Preserve Assateague Island, and Allied-Sig- and abundance at these different levels, but also nal, Inc. We especially thank R. Batiuk of the U.S. will protect existing resources. Environmental Protection Agency's Chesapeake Bay Liaison Office, for his continuing encourage- Acknowledgments ment, and A. Frisch, who has made significant contributions to ensure the quality of the data The monitoring program for SAV in the Ches- management. H. Neckles provided valuable com- apeake Bay would not have been possible without ments on the initial draft. REGioNAL AND PEDERmSTATE CooPFRATivE PRocmms 121 RAPPAHANNOCK RIVER HISTORICAL WINDOW PLOT OF TOTAL SAV AREA FOR ALL DENSITY CLASSES 800- S A V 600- A R E A 400- H E C T A R 200- E S 0 1970 1972 1974, 1976 1978 1980 1982 1984 1986 1988 YEAR Fig. 7. Abundance of submerged aquatic vegetation (SAV) for the lower Rappahannock River (section 6), 1971-87. 2000 1500- 1000- 500 0-- 1978 1984 1985 1986 1987 Fig. 8. Abundance of submerged aquatic vegetation for the upper Potomac River area (section 11), 197"7. 122 BiowcmcAL REPoRT 90(18) References Orth, R. J., and K. A. Moore. 1983a. Submersed vascular plants: techniques for analyzing their distribution and abundance. Mar. Tech. Soc. J. 17:38-52. Anderson, R. R., and R. T. Macomber. 1980. Orth, R. J., and K. A. Moore. 1983b. Chesapeake Bay: Distribution of submersed vascular plants in an unprecedented decline in submerged aquatic Chesapeake Bay, Maryland. Final report. U.S. vegetation. Science 222:51-53. Environmental Protection Agency, Chesapeake Bay Orth, R. J., and K. A. Moore. 1984. Distribution and Program, Grant R805970. 126 pp. abundance of submerged aquatic vegetation in Chesapeake Bay: an historical perspective. Estuaries Bayly, S., V D. Stotts, P F Springer, and J. Stennis. 7:531--540. 1978. Changes in submerged aquatic macrophyte Orth, R. J., and K. A. Moore. 1988a. Distribution of populations at the head of the Chesapeake Zostera marinaL. andRuppia maritimaL. sensu lato Bay-1958-1975. Estuaries 1:74-85. along depth gradients in the lower Chesapeake Bay, Cambridge, M. L., and A. J. McComb. 1984. The loss of U.S.A. Aquat. Bot. 32:291-305. seagrass from Cockburn Sound, western Austra- Orth, R. J., and K. A. Moore. 1988b. Submerged aquatic lia. I. The time course and magnitude of seagrass vegetation in the Chesapeake Bay: a barometer ofbay decline in relation to industrial development. Aquat. health. Pages 619-629 in M. Lynch, ed. Bot. 20:229-243. Understanding the estuary: advances in Chesapeake Dennison, W C., G. Marshall, and C. Wigand. 1990. Bay Research. Chesapeake Res. Consort. Publ. 129. Effect of 'brown tide" shading on eelgrass (Zostera CBP/I`R,5@2@. marina L.) distribution. Lecture notes in coastal Orth, R. J., K. A. Moore, and H. H. Gordon. 1979. estuarine studies. In press. Distribution and abundance of submerged aquatic Ferguson, R. L., J. A. Rivera, and L. L. Wood. 1988. vegetation in the lower Chesapeake Bay, Virginia. Submerged aquatic vegetation in the Albermarle- Final report. U.S. Environmental Protection Agency, Pamlico estuarine system. Final report. National Chesapeake Bay Program. EPA- 600/8-79-029/SAV1. Marine Fisheries Service, National Oceanic and 219 pp. Atmospheric Administration, Beaufort Laboratory, Orth, R. J., J. Simons, R. Allaire, V Carter, L. Hindman, Beaufort, N. C. 68 pp. K. Moore, and N. Rybicki. 1985. Distribution of Field, D. W, C. E. Alexander, and M. Broutman. 1988. submerged aquatic vegetation in the Chesapeake Toward developing an inventory of U.S. coastal Bay and tributaries- 1984. Final report. wetlands. Mar. Fish. Rev. 50:40-46. Environmental Protection Agency, Cooperative Iverson, R. L., and H. R Bittaker. 1986. Seagrass Agreement X-003301-01. 155 pp. distribution and abundance in eastern Gulf of Mexico Orth, R. J., J. Simons, J. Capelli, V Carter, A. Frisch, coastal waters. Estuarine Coastal Shelf Sci. L. Hindman, S. Hodges, K. Moore, and N. Rybicki. 22:577--602. 1987. Distribution of submerged aquatic vegetation Kemp, W M., W. R. Boynton, R. R. Twilley, J. C. in the Chesapeake Bay and tributaries and Stevenson, and J. C. Means. 1983. The decline of Chincoteague Bay-1986. Final report. U.S. submerged vascular plants in upper Chesapeake Environmental Protection Agency. 180 pp. Bay: summary of results concerning possible causes. Orth, R. J., J. Simons, J. Capelli, V Carter, L. Hindman, Mar. Tech. Soc. J. 17:78-89. S. Hodges, K. Moore, and N. Rybicki. 1986. Macomber, R. T, and D. Allen. 1979. The New Jersey Distribution of submerged aquatic vegetation in the submerged aquatic vegetation distribution atlas. Chesapeake Bay and tributaries-1985. Final report. Final report. New Jersey Department of U.S. Environmental Protection Agency. 296 pp. Environmental Protection. 25 pp. Orth, R. J., and J. van Montfrans. 1990. Utilization of Neverauskas, V P 1987. Monitoring seagrass beds marsh and seagrass habitats by early stages of around a sewage sludge outfall in South Australia. Callinectes sapidus: a latitudinal perspective. Bull. Mar. Pollut. Bull. 18:158-164. Mar. Sci. 46:126-144. Orth, R. J., A. A. Frisch, J. F Nowak, and K. A. Moore. Perry, H. M., editor. 1984. A profile of the blue crab 1989. Distribution of submerged aquatic vegetation fishery of the Gulf of Mexico. Gulf States Mar. Fish. in the Chesapeake Bay and tributaries and Comm. 9.80 pp. Ch-incoteague Bay-1987. Final report- U.S- Reyer, A. J., D. W Field, J. E. Cassells, C. E. Alexander, Environmental Protection Agency. 260 pp. and C. L. Holland. 1988. The distribution and areal Orth, R. J., and H. H. Gordon. 1975. Remote sensing of extent of coastal wetlands in estuaries of the Gulf of submerged aquatic vegetation in the lower Mexico. National Oceanic and Atmospheric Chesapeake Bay. Final report. National Aeronautics Administration, Strategic Assessment Branch, and Space Administration. Contract NAS1-10720. Rockville, Md. 18 pp. 62 pp. Roach, E. R., M. C. Watzin, J. D. Scurry, and J. B. Orth, R. J., and K. A. Moore. 1981. Submerged aquatic Johnston. 1987. Wetland changes in Mobile Bay. vegetation of the Chesapeake Bay: past, present Pages 92-101 in T A. Lowery, ed. Symposium on the and future. Pages 271-283 in Transactions of the natural resources of the Mobile Bay estuary, Mobile, 46th North American wildlife natural resources Alabama. Alabama Sea Grant Extension Service conferences. MASGP-87-007. RF,GioNAL AND FEDERAL-STATF CoopERATivF PRWRAMs 123 Tiner, R. W, Jr. 1977. An inventory of South Carolina's Burke, H. A. Groman, T R. Henderson, J. A. Kusler, coastal marshes. S. C. Mar. Resour. Cent. Tech. and E. J. Meyers, eds. Wetlands of the Chesapeake Rep. 23. 33 pp. Bay. Environmental Law Institute. Washington, Tiner, R. W, Jr. 1985a. Wetlands ofNewJersey. U.S. Fish D.C. 389 pp. and Wildlife Service, National Wetlands Inventory, Newton Corner, Mass. 117 pp. Walker, D. 1. 1989. Methods for monitoring seagrass Tiner, R. W, Jr. 1985b. Wetlands of the Chesapeake habitat. Victoria Institute of Marine Sciences, Bay watershed: an overview. Pages 16-24 in D. M. Victoria, Australia. Working Paper 18. 32 pp. REGIONAL AND FE I)ERAL-STATE COOPERATIVE PROGRAMS 125 Mapping Submerged Aquatic Vegetation in North Carolina with Conventional Aerial Photography by Randolph L. Ferguson and Lisa L. Wood National Oceanic and Atmospheric Administration National Marine Fisheries Service Southeast Fisheries Center Beaufort Laboratory Beaufort, North Carolina 28516 ABSTRACT.-Mapping submerged aquatic vegetation (SAV) directly supports the National Oceanic and Atmospheric Administration's legislated responsibilities in estuarine and marine science, and it supports President Bush's no net loss of wetlands policy. Marine SAV includes some of the most productive primary producers in the marine environment. SAV functions as an underwater nursery area forjuveniles or adults ofmany estuarine-dependent, commercially and recreationally exploited fish and shellfish. SAV is a valuable resource to monitor, conserve, and enhance, as is being done, for example, in Chesapeake Bay. We are mapping SAV in coastal waters of North Carolina with conventional aerial photography. The immediate objective is to complete initial photographic coverage, photointerpretation, and mapping of SAV habitat in this State. We also will evaluate remote sensing of SAV by digital sensors on a variety of platforms, including satellites and aircraft. The long-term objective of this work is to monitor SAV in North Carolina and to develop a protocol to monitor SAV nationwide in coastal waters on a 2- to 4-year cycle. Digital remote sensing of emergent and submergent wetlands may offer significant advantages for frequent monitoring of large study areas relative to aerial photography, the present standard forremote sensing and mapping of SAV. Because ofits submergent existence, however, SAV is more difficult to detect and map than emergent wetland vegetation. Authorization of SAV from satellites and aircraft, and (3) develop a protocol for monitoring of SAV in coastal marine This project directly supports the National Oce- waters nationwide on a 2- to 4-year cycle, as part anic and Atmospheric Administration's (NOAA) of NOAA's Coastal Ocean Program (Thomas and legislated responsibilities in estuarine and marine Ferguson 1990). science, monitoring, and management contained in the Fish and Wildlife Coordination Act, the Coastal Introduction Zone Management Act, the Clean Water Act, the Marine Sanctuaries Research and Protection Act, SAV-vegetation adapted to growing under the Magnuson Fisheries and Conservation Act, the water-includes some of the most productive pri- National Environmental Policy Act, and President mary producers in the marine environment (Fer- Bush's no net loss of wetlands policy. guson et al. 1980). SAV provides habitat rich in food and cover for juveniles and adults of many 0 ectives estuarine- epen ent, commercially and recrea- tionally exploited fish and shellfish, but SAV also The goals of this project are to (1) map and is vulnerable to adverse effects from anthropogenic monitor submerged aquatic vegetation (SAV) in activities (Zeiman 1982; Thayer et al. 1984; Zei- coastal North Carolina with conventional aerial man and Zeiman 1989). SAV is too valuable a photography, (2) evaluate digital remote sensing national resource not to monitor, conserve, and 126 BIOLOGICAL REPORT 90(18) develop on a timely basis, as is being done, for areal extent of SAV (Orth et al. 1990). In 1987, example, in Chesapeake Bay (Orth et al. 1990). North Carolina had about four times more SAV At present, we are mapping SAV in coastal than Chesapeake Bay (Orth et al. 1989). In North Carolina with conventional aerial photogra- coastal North Carolina, the area of SAV exceeds phy. We will complete photographic mapping and the area of salt marshes and is 81% of the total then evaluate digital remote sensing of SAV with area (101,000 ha) of salt water plus freshwater airborne and satellite digital sensors. marshes (Field et al. 1988). In North Carolina, the Digital remote sensing may provide significant total estuarine area is 890,000 ha, while the estu- advantages in cost and timeliness over aerial pho- arine shoalwater area, (that area less than 6 feet tography, the present standard for remote sensing deep at mean low water [MLW]), is about and mapping of SAV. Costs of photographic and 320,000 ha (U. S. Fish and Wildlife Service 1970). digital approaches are generally similar for small Bottoms less than 6 feet deep at MLW are poten- study areas, but costs increase rapidly for photo- tially habitable by SAV in coastal North Carolina graphic approaches when subject areas exceed (Ferguson et al. 1989b). Therefore, about 9% of 50,000 ha (Klemas and Hardesky 1987). SAV is the total estuarine area and 25% of the potentially difficult to detect and map relative to terrestrial or habitable estuarine shoal area in the sounds of emergent habitats, however, because it occurs un- North Carolina currently support SAV. derwater. Although satellite imagery has been ap- The Beaufort Laboratory of the National Ma- plied to detection of SAV with limited success rine Fisheries Service (NMFS) has aerial photog- (Ackleson and Klemas 1987), SAV mapping is raphy of most known SAV in coastal North Caro- problematic with available satellite imagery. New lina (Figure) at scales of 1: 12,000 to 1: 50,000. The airborne digital sensors, such as multispectral, photography covers the sounds from Bogue Inlet solid-state video cameras (McKim et al. 1985), to Drum Inlet in 1985 at a scale of 1:12,000 or may ultimately provide the necessary combination 1:20,000, and the sounds from Cape Lookout to of spectral and spatial resolution and flexibility of Oregon Inlet, northern Core Sound, and southern timing for image acquisition required for detection and eastern Pamlico Sound in 1988 at scales of and mapping of SAV. 1:24,000 and 1:50,000. The 1:50,000 photography Conventional aerial photography is the stan- was taken to provide horizontal control for the dard approach for mapping SAV at the present soundward leg of two parallel flight lines of time. SAV generally is mapped from photographs 1:24,000-scale photography required to span the taken at low tide at scales of 1:24,000 or larger entire width ofthe extensive shoal area in eastern (Ferguson et al. 1989b; Orth et al. 1989). Emer- Pamlico Sound. gent wetlands are routinely mapped from aerial Photointerpretation and photography is ongo- photography taken at scales of 1:40,000 or smaller ing (Figure), but the extent of photography inter- (e.g., Wilen 1990). This difference in photographic preted and mapped has been limited. Only about scale for the two types of wetlands is necessary 5% of total SAV habitat-seagrass habitat in because the contrast between SAV and un- southern Core Sound between Cape Lookout vegetated bottoms can be inherently low, as it is in and Drum Inlet-has been delineated in a pub- areas where submerged sediments are dark or lished chart (Ferguson et al. 1989b). The 1988 where water has high concentrations of dissolved 1:50,000-scale photography (but not the 1:24,000) organic matter or suspended particulate material. from Ocracoke Inlet to Oregon Inlet has been Visualization canbe improvedbyuse oflarge-scale interpreted, and preliminary seagrass maps have photography taken when dissolved and particulate been digitized for that area. A preliminary map of materials are at a minimum. seagrasses in northern Core Sound also has been digitized based on interpretation of the 1988, Current Inventory Coverage 1:24,000-scale photography, and a chart of SAV that includes this area will be published in 1991. For a number of reasons, North Carolina is a Seagrass occurs in western Pamlico Sound, in particularly appropriate location to map SAV and southern Roanoke Sound, and west of Bogue Inlet test digital remote sensing of high salinity and to the border with South Carolina, but these brackish SAV, (Ferguson et al. 1989b). Of the areas, as well as SAV in brackish-water areas of contiguous 48 States, North Carolina, with about northern Roanoke, Albemarle, and Currituck 81,000 ba of SAV, ranks second after Florida for sounds have not yet been suitably photographed. REGIONAL AND FEDERAL-STATE CoopERATivE PRoGRAms 127 .AVIRGINA STATE LINE.4, CURRITUCK SOUND SOU140 P, A CHOWAN RIVER B. ROANOKE RIVER C. ROANOKE SOUND D. OREGON INLET czoo E. PAMLICO RIVER F. NEUSE RIVER G. DRUM INLET H BACK SOUND vl@ I BOGUE INLET 0\1 lp J. SOUTH CAROLINA lb STATE LINE 0 IS OCRACOKEINLET CORE SOUND E30GUE SOUNDA CAPE LOOKOUT 50 MILES Figure. Coastal North Carolina. 128 Biow=AL REPoRT 90(18) Albemarle and Currituck sounds will be photo- phyllum spicatum), bushy pondweed (Najas qua- graphed in 1990 with funding from the Albe- dalupensis), sago pondweed (Potamogeton pecti- marle-Pamlico Estuarine Study. natus), redhead grass (P. perfoliatus), widgeon- Surface-level sampling has progressed beyond grass (Ruppia maritima), wild celery (Vallisneria the area presently photographed and includes americana), and horned pondweed (Zannichellia both high-salinity and brackish-water areas (Fer- palustris). guson et al. 1989b). High- and intermediate-salin- ity waters in North Carolina are populated by the Pro .ect Period and Scope temperate species, eelgrass (Zostera marina); the tropical species, shoalgrass (Halodule wrightii); This project began in 1985 in North Carolina and the panlatitudinal species, widgeon grass under base (NMFS) funding from the Beaufort (Ruppia maritima). Unlike eelgrass and shoal- Laboratory, Southeast Fisheries Center. Subse- grass, widgeon grass is eurylialine and occurs in quently, project activities have been supported both high-salinity waters and brackish waters. with additional funding from NOAA's Coastal North Carolina is the southern limit for eelgrass Ocean Program and the U.S. Environmental Pro- and the northern limit for shoalgrass on the east tection Agency's (EPA) National Estuarine Pro- coast (Thayer et al. 1984). The northern limit of gram for Albemarle and Pamlico sounds. The first shoalgrass has been extended to near Oregon aerial photography (both color and infrared) was Inlet, North Carolina (Ferguson et al. 1989b). done using basefunds for Bogue, Back, and south- SAV includes seagrasses that require moder- ern Core sounds in 1985. Additional support was ate- to high-salinity seawater and it includes obtained in 1987-1988 and in 1990-1991 from freshwater species that tolerate low-salinity EPA's Albemarle-Pamlico Program. Under this brackish water. An exception is widgeongrass, funding, a visual aerial survey (December 1987) of which thrives in fresh water and in seawater. Core Sound and eastern Albemarle and Pamlico The SAV species of eelgrass, shoalgrass, and sounds, and photography (April 1988) of Core widgeongrass occur to a limited extent in southern Sound and eastern Panilico Sound (both color and Roanoke Sound, but are abundant to the south in infrared at scales of 1:24,000 and 1:50,000), were Pamlico, Core, Back, and Bogue sounds (Figure). completed. SAV samples from Core, eastern These species also occur west of Bogue Sound and Pamlico, Roanoke, and eastern Albemarle sounds to the border with South Carolina. In Pamlico, (March 1988) and Currituck Sound (October 1987) Core, Back, and Bogue sounds, SAV habitats tend also were collected to provide ground-level verifica- to be large and luxuriant on the extensive shoals tion for interpretation of current and anticipated along the inside of the Outer Banks where salinity photography, and to provide regional data on spe- tends to be highest and bottoms are sandy. Along cies composition of SAV. the Outer Banks, muddy shoals also support lux- uriant seagrass meadows, but these areas are Project activities are ongoing. F\mding from restricted to small protected bays associated with NOAA!s Coastal Ocean Program will allow (1) com- emergent marshes. In contrast, the relatively soft pletion of interpretation of photography and anal- bottom of the mainland shore in Core Sound is ysis of SAV and sediment samples in hand, (2) con- characterized by thin shoreline beds. Although struction and publication of SAV charts, (3) testing SAV occurs in western Pamlico Sound, eelgrass ofremote sensing of SAV by digital sensors, and (4) and shoalgrass are displaced in the low-salinity development of a protocol for nationwide monitor- areas of the estuaries of lower Neuse and Pamlico ing of SAV. Cooperative funding from NOAA and rivers by species tolerant to brackish waters. EPA will provide for complete photographic cover- These two species also have not been reported to age and initial mapping of SAV in coastal North occur in Albemarle or Currituck sounds. Carolina. Subsequently, fimding will be sought to SAV does occur in brackish waters of the estuar- implement a SAV habitat monitoring program. We ies of the Neuse and Pamlico rivers and in the anticipate that the initial mapping of North Caro- Albemarle, Currituck, and northern Roanoke lina will be completed in 1993 and that the moni- sounds (Beal 1977; Ferguson et al. 1989b), where toring will continue indefinitely. Testing of digital salinities remain low and waters tend to be turbid. sensors has not begun, but will be proposed as Species common in brackish water in coastal North technological advances and funding permit Carolina include Eurasian water milfbil (Myrio- (Thomas and Ferguson 1990). REGioNAL AND FEDERAL-STATE CooPERATivE PRoGRAms 129 Product Description line have 60% endlap, and photographs along par- allel flight lines have 20% sidelap. Development We have two types of products. Published and copying of film are done commercially by a charts are on 3- by 4-foot chart paper at a scale of contractor for CGS and follow CGS guidelines and 1:36,000. The subject area's of the two charts quality-control procedures. published to date are southern Core Sound be- Surface-level information is acquired in two tween Cape Lookout and Drum Inlet (Ferguson phases. The first phase of sampling collects re- et al. 1989a), and northern Core Sound and gional information about SAV and environmental sotheastern Panilico Sound between Drum Inlet conditions, especially turbidity, in the study area. and Osracoke Inlet (Ferguson, et al. 1991). The Stations for sampling of SAV are selected by a next chart to be published (with a similar size and dot-matrix approach. A matrix of rectangularly scale) will document change in SAV habitat (1985 arranged dots of appropriate dimensions and spa- to 1988) in southern Core Sound. The second type tial density (e.g., 1.3-scaled nautical miles from of product is a digital data base. The SAV habitat center to center) is placed over a NOAA nautical data are stored in a geographic information chart. The latitude and longitude of dots occurring system (GIS). at water depths of 0 to 10 feet are determined. These locations are then visited with the aid of Inventory Methods LORAN C, and they are examined for the presence of SAV to a radius of about 0.2 nautical miles. Any Inventory methods include acquisition of pho- SAV present is identified to species or is sampled tography and surface-level information and pho- along with surface sediment and returned to the tointerpretation. Cartographic products can be no laboratory for analysis. Salinity and Secchi disc better than the quality of the source data, which depth are recorded. Activities affecting water qual- are the aerial photography and surface-level sam- ity (e.g., dredging, commercial fishing activity, or ples. Aerial photography for the project has been local drainage of turbid water) are noted along conducted by NOAA, National Ocean Survey, and with general environmental observations at the Coastal and Geodetic Services (CGS) in Rockville, site. This sampling phase is done just before and Maryland. We select seasonally optimal times during the photographic mission. At this time, the based on biological considerations and potential surface party is in periodic (at least daily) contact for clear-air days. These times are April and May with the flight crew by telephone or radio to discuss or late August through early October for moder- the mission decisions for a given day. ate- to high-salinity areas in coastal North Caro- SAV habitat noted in the first phase of surface- lina. Different species of SAV in moderate- to level sampling is key to the recognition of the high-salinity waters in North Carolina achieve variety of SAV habitat areas visible in the photog- maximum biomass at different times; eelgrass in raphy. Interpreting photos of SAV habitat is based late spring, and shoalgrass and widgeongrass in on the photointerpreter's experience and often in- late summer to early fall (Ferguson et al. volves subjective judgment. Visualization is best 1989a,b). Brackish-water SAV is best photo- achieved stereoscopically at low magnification graphed during maximum biomass, which occurs (e.g., x 8), viewing pairs of the 9- by 9-inch color between August and October for most species in transparencies illuminated with high-, uniform-, North Carolina. Within these seasonal windows, and variable-intensity light. Appearance of SAV the decision to fly a particular photographic mis- habitat can vary considerably and may riot be sion is dependent on time of day (sun angle less consistent from place to place. Experience is re- than 25' to minimize glint from reflected sun- quired to identify and delineate SAV habitat with light), tidal stage (low for minimum amount of accuracy and reproducibility. That experience is water to penetrate), and other local conditions increased by feedback from the second phase of (absence of cloud cover, minimal haze, low water surface-level sampling. turbidity, and absence of surface waves). During The second phase of surface-level sampling oc- photography, location along flight line, yaw, pitch, curs subsequent to the acquisition and initial ex- and altitude are controlled within CGS guidelines amination of the photography, and it is an essen- by the pilot and navigator. The photographer de- tial training activity for photointerpreters. Specific termines exposure, focus, and overlap of adjacent areas of SAV habitat not sampled during phase exposures. Sequential photographs along a flight one-in particular, unusual, potential, or ques- 130 BioLomcAL REPoirr 90(18) tionable SAV habitat observed in the photogra- Cartographic Procedures phy-are located and visited. SAV habitat is circumscribed by tracing a pen- Cartographic procedures include georeferenc- cil line around continuous meadows of SAV or ing, scaling, compiling, production of chart prod- clusters of "patches" of SAV onto a stable base ucts, and digitization of SAV habitat data. The overlay of the photograph. The growth form of base maps are 1:24,000-scale, 7.5-min United SAV beds is a combination of historical and pres- States Geological Survey (USGS) topographic ent physical and biological interactions (Fonseca maps or 1:20,000-scale CGS shoreline manu- et al. 1983; Fonseca and Kenworthy 1987). Thus, scripts. Four cultural (e.g., road intersections and areas with clusters of SAV patches, as well as buildings) and, if necessary, natural (e.g., shore- areas with apparently continuous cover of SAV, lines) features visible in the photograph and also constitute SAV habitat. We have made no attempt present in the base map are traced, along with the yet to categorize polygons of SAV habitat accord- polygons of SAV habitat, to provide horizontal ing to the two gradients of patchy to continuous control for the photography. The SAV polygons are and thin to dense. Distinctions tend to be arbi- compiled on referenced stable base overlay of the trary and problematic. The appearance of SAV base map by tracing, if scales are consistent, or bed form can change, for example, as a function of with a zoom transfer scope that also allows for scale and overall quality of the photography. Beds scaling and (if necessary) correction for distortion. of SAV, moreover, often intergrade from one Su .bsequently, the SAV habitat tracings are inked growth form to another. The causative factors of using a 0.3-mm permanent ink pen. the different bed forms and their significance to The inked SAV overlays and theirbase maps are secondary productivity and habitat management the source materials for charts produced by stan- issues, in any case, remain the subject of further dard photographic and printing techniques and for research. digitizing and plotting computer-generated maps. In our experience, estimates of SAV habitat Printed chart products of SAV habitat are pro- tend to be conservative. For example, thinly duced by CGS. For the chart that was published in grassed areas (e.g., shoreline beds with a contin- 1989 (Ferguson et al. 1989a), SAV habitat infor- uous cover of small plants on dark-ap'pearing bot- mation was superimposed on a base map compiled toms, or widely dispersed patches of SAV) can be from edited stable bases of USGS 7.5-min quad- virtually undetectable in the photography, de rangles for shoreline and land information, and - NOAA nautical charts for navigational aids (chan- pending on water clarity, substrate darkness, and nel markers) and bathymetric data. The chart photography scale. SAV habitat discovered by published in 1991, for the area between Drum Inlet surface-level sampling is added to the photo- and Ocracoke Inlet, is based on CGS shoreline graphic tracings after in situ measurement of the manuscripts generated from photography col- habitat and reexamination of the photography. lected at the same time as that for SAV (1988). This Under ideal conditions, individual SAV patches as is being done because changes in shoreline be- small as 1 m in diameter are detectable (with tween the most recently published USGS maps magnification) in photography at a scale Of and our 1988 photography were substantial. The 1:24,000, but in practice, minimum habitat sizes SAV overlays are digitized on a cooperative basis, recorded are:5 0.3 ha. and maps are generated as computer plots by the The project maintains all commissioned pho- State of North Carolina, Department of Environ- tography, photographic-scale stable base tracings ment, Health and Natural Resources, Center for of SAV habitat, stable base reference maps, and Geographic Information and Analysis (CGIA), in stable base SAV overlays. We are also seeking Raleigh. For map products or information about historical photography and examining it for refer- obtaining map products, see Appendix A. ence and retroactive change analysis of SAV dis- tribution and extent. Unfortunately, historical Estimated Funding photography often is of limited value for estimat- ing total SAV habitat because of the absence of Estimated funding requirements for aerial pho- surface-level verification and because of inappro- tographic mapping of SAV in marine and brackish priate scale, season, tidal stage, water turbidity, waters of coastal North Carolina are $190,000 per sun glint, or areal coverage. year for 3 years. This amount includes photogra- REGIONAL AND FEDERA1,STATE COOPERATIVE PROGRAMS 131 phy, photointerpretation, surface-level sampling, 0 justification for classification of areas -as scaling and compiling of SAV on 1:24,000-scale outstandmig resource Iwaters, and USGS maps, digitizing into a GIS system, and 0 public interest reports in newspapers and on construction and publication of SAV charts. It does television. not include funding required to construct current CGS shoreline manuscripts. Requirements for References mapping SAV in other States would be similar, dependent on the extent of estuarine shoal areas, Ackleson, S. G., and V Klemas. 1987. Remote sensing of growth characteristics and extent of SAV, and submerged aquatic vegetation in lower Chesapeake water- and bottom-quality considerations. Addi- Bay: a comparison of Landsat MSS to TM imagery. tional funds, dependent on sensor and platform, Remote Sens. Environ. 22:235-248. are required to conduct evaluations of digital re- Beal, E. 0. 1977. A manual of marsh and aquatic vascular plants of North Carolina with habitat data. mote sensors. North Carolina Agricultural Experiment Station, N.C. State Univ. Raleigh. Tech. Bull. 247.298 pp. Anticipated Future Activities Ferguson, R. L., J. A. Rivera, and L. L. Wood. 1989a. Seagrasses in southern Core Sound, North Anticipated future activities include (1) comple- Carolina. NOAA-Fisheries Submerged Aquatic tion of initial aerial photographic mapping of SAV Vegetation study, southern Core Sound, North in coastal North Carolina, (2) initiation of monitor- Carolina. National Oceanic and Atmospheric ing of SAV in coastal North Carolina on a cycle of Administration-Fisheries, Beaufort Laboratory, 2 to 4 years, (3) evaluation of digital sensors Beaufort, N.C. 3- x 4-foot chart, with text and (e.g., Landsat Thematic Mapper, SPOT, and ai illustrations. ir- Ferguson, R. L., J. A. Rivera, and L. L. Wood. 1989b. borne multispectral solid-state video camera) for Submerged aquatic vegetation in the detection and mapping of SAV habitat, (4) devel- Albemarle-Pamlico estuarine system. North opment of a protocol including photographic and Carolina Department of Natural Resources and digital imagery for mapping SAV, and (5) coordi- Community Development, Raleigh, N.C., and U.S. Environ- mental Protection Agency, National nation with or initiation of mapping of SAV in the Estuary Program. Albemarle-Pamlico Estuarine coastal areas of other States. Study, Project 88-10. 68 pp. Ferguson, R. L., G. W Thayer, and T R. Rice. 1980. Marine primary producers. Pages 9-69 in F. J. User Perspective Vernberg and W. Vernberg, eds. Functional State and Federal environmental managers and adaptations of marine organisms. Academic Press, New York. researchers and private citizens have been Partic- Ferguson, R. L., L. L. Wood, and B. T Pawlak. 199 1. SAV ularly interested in information related to location habitat from Drum Inlet to Ocracoke Inlet, North and extent of SAV. A blue-ribbon panel on environ- Carolina. NOAA-Coastal Ocean Program mental indicators recently reported to the gover- Submerged Aquatic Vegetation Study. National nor of North Carolina a high-priority need for Oceanic and Atmospheric Administration/National Marine Fisheries Service, Beaufort Laboratory, assessment of SAV as an indicator of coastal water Beaufort, N.C. [Three- by four-foot chart with text quality and of the well-being of the State's living and illustrations] marine resources. Throughout this project, the Field D. W, C. E. Alexander, and M. Broutman. 1988. following information requests concerning SAV Toward developing an inventory of U.S. coastal have been received: wetlands. Mar. Fish. Rev. 50(l):40-46. Fonseca, M. S., and W J. Kenworthy. 1987. Effects of � habitat measurement methodology, current on photosynthesis and distribution of � importance to fisheries, seagrasses. Aquat. Bot. 27:59-78. � importance to waterfowl, Fonseca, M. S., J. C. Zieman, G. W Thayer, and J. S. � distribution of endemic eelgrass wasting Fisher. 1983. The role of current velocity in disease, structuring eelgrass (Zostera marina L.) meadows. Estuarine Coastal Shelf Sci. 17:367-WO. � regulation of inshore fishing activities, Klemas, V, and M. A. Hardisky 1987. Remote sensing � distribution as an oil-sensitive habitat, of estuaries: an overview. Pages 91-120 in V Klemas, � site-specific occurrence and species composition J. P Thomas, and J. B. Zaitzeff, eds. Remote sensing in areas proposed for dredge and fill operations of estuaries, proceedings of a workshop. U.S. or water-related construction, Department of Commerce, National Oceanic and Atmospheric Administration, Estuarine Programs � index for monitoring water quality and health Office and National Environmental Satellite Data of living marine resources, Service, Washington, D.C. 132 Biow=AL REPoRT 90(18) McKim, H. L., V Klemas, L. W, Gatto, and C. J. Merry. Thomas, J., and R. L. Ferguson. 1990. National 1985. Potential of remote sensing in the Corps of Oceanic and Atmospheric Administration@s habitat Engineers dredging program, U.S. Army Water mapping under the Coastal Ocean Program. Pages Resources Support Center, U.S. Army Corps of 27-37 in S. J. Kiraly, F. A. Cross, and J. D. Engineers Cold Regions Research and Engineering Buffington, eds. Federal coastal wetland mapping Laboratory, Hanover, N.H. Spec. Rep. 85-20. 42 pp. programs. U.S. Fish Wildl. Serv., Biol. Rep. 90(18). Orth, R. J., A. A. Fi-isch, J. F Nowak, and K. A. Moore. 1989. U.S. Fish and Wildlife Service. 1970. Technological Distribution of submerged aquatic vegetation in the impacts on estuary resource use. Pages 1-122 in Chesapeake Bay and tributaries and Chincoteague Bay National estuary study. Vol. 5. U.S. Government 1987. Chesapeake Bay Program, U.S. Environmental Printing Office, Washington, D.C. Protection Agenc@, Annapolis, Md. 247 pp. Wilen, W. 1990. The U.S. Fish and Wildlife Service's Orth, R. J., ,K. A. Moore, and J. C. Nowak. 1990. National Wetlands Inventory. Pages 9-20 in S. J. Monitoring seagrass distribution and abundance Kiraly, F. A. Cross, and J. D. Buffington, eds. patterns: a case study from the Chesapeake Bay. Pages Federal coastal wetland mapping programs. U.S. 112-124 in S. J. Kiraly, R A. Cross, and J. D. Fish Wildl. Serv., Biol. Rep. 90(18). Buffington, tech. coords. Federal coastal wetland Zeiman, J. C. 1982. The ecology of the seagrasses of mapping programs. U.S. Fish Wild]. Serv., Biol. south Florida: a community profile. U.S. Fish Wildl. Rep. 90(18). Serv., FWS/OBS-82/25.185 pp. Thayer, G. W, W, J. Kenworthy, and M. S. Fonseca. Zeiman, J. C., and R. T. Zeiman. 1989. The ecology of 1984, The ecology of eelgrass meadows of the the seagrass meadows of the west coast of Florida: Atlantic coast: a community profile. U.S. Fish Wildl. a community profile. U.S. Fish Wildl. Serv., Biol. Serv., FWS/OBS-84/02. 147 pp. Rep. 85(7.25). 155 pp. REGioNAL tND FmmmSTATE CoopmunvE PRwmms 133 Appendix. Availability of Map Products and Contacts for More Information For a free copy of published charts or more information on SAV mapping, write to: Randolph L. Ferguson NOAA, National Marine Fisheries Service Beaufort Laboratory, Southeast Fisheries Center Beaufort, North Carolina 28516 For a customized computer plot of SAV (not free) or more information on digitization of SAV data and the North Carolina GIS data base, write to: Karen Siderelis, Director Center for Geographic Information and Analysis North Carolina Department of Environment, Health and Natural Resources 512 North Salisbury Street, Room 1193 P. 0. Box 27687 Raleigh, North Carolina 27611 For more information on aerial photographic and cartographic procedures, write to: Rear Admiral J. Austin Yeager NICG Director, Charting and Geodetic Services NOAA, National Ocean Service Room 1006, Rockwall Building 6001 Executive Boulevard Rockville, Maryland 20852 REGIONAL AND FEDERAL-STATE CooPEP.ATivE FRoGmms 135 Project Plan for Mapping and Geographic Information System Implementation of Land Use and Land Cover Categories for the Albemarle-Pamlico Estuarine Study by H. M. Cheshire and Siamak Khorram Computer Graphics Center North Carolina State University Raleigh, North Carolina 27695-7106 ABSTRACT.-The Albemarle-Pamlico (A/P) system in North Carolina is one of 12 estuaries in the U.S. Environmental Protection Agency's National Estuary Program. The lack of a current land use inventory for the Albemarle-Pamlico estuarine drainage area has been identified as a critical gap in the A/P Study resource data base. At an A/P Study workshop late in 1987, Landsat Thematic Mapper UND digital data were recommended as the most cost-effective and practical source for developing an inventory for the 12 million acres of A/P drainage basin. The Computer Graphics Center (North Carolina State University) and the Center for Geographic Information and Analysis (formerly Land Resources Information Service; North Carolina Department of Environment, Health and Natural Resources) are cooperating in the development, storage, and dissemination of the inventory. The study area includes a portion of Virginia and nearly one-third of North Carolina including almost all of the Tidewater region. The project will result in: 1) a current digital land use and land cover inventory based on Landsat TM data classified, verified, and registered to the A/P Study geographic information system data base; 2) digital files in a standard data exchange format available to investigators and resource managers; 3) a capability within the A/P Data Management Center to maintain, analyze, and make future updates to the inventory; and 4) land use and land cover maps summarized by geopolitical boundaries. Objectives and Background Albemarle and Pamlico sounds, encompassing about 3,000 square miles of protected inshore wa- The Albemarle-Pamlico Estuarine Study (A/P ters and 20,000 square miles of land. It includes Study) is a joint project of the U.S. Environmental over two-thirds of North Carolina's coastal wet- Protection Agency (EPA) and the State of North lands and extends west into the Piedmont region. Carolina. The estuary is 1 of 12 in the Federal In 1989, the Computer Graphics Center (CGC) National Estuary Program. The ultimate goal of at North Carolina State University and the Center the A/P Study is to aid in effective management of for Geographic Information and Analysis (CGIA) the important estuarine resources in northeastern were funded to conduct such an inventory. The North Carolina through scientific research and CGC is a university-wide research unit that con- public awareness. The North Carolina A/P Study ducts research and training in the areas of remote Program Office determined that lack of a land use sensing, image processing, geographic information inventory for the Albemarle-Pamlico estuary was system (GIS) design and applications, and inte- a critical gap in the A/P resource data base. As a grated relational data-base design and manage- result of an A/P Study workshop in late 1987, ment systems. The CGIA operates a GIS and Landsat Thematic Mapper (TM) data were identi- serves as the official repository of digital geo- fied as the most practical and cost-effective data graphic data for the State of North Carolina. The source for developing a land use inventory for the CGIA is a receipt-funded agency in the Division of more than 12 million acres of drainage basin. The Information Services, North Carolina Department study area includes all of the tributary basins of of Environment, Health, and Natural Resources. 136 BioLoGicAL REPoRT 90(18) The goal of this mapping project is to provide National Aeronautics and Space Administration- baseline data on the Albemarle-Pamlico drainage Goddard Space Flight Center. A number of specific basin resources in a form usable by scientists and applications have also been developed by CGC decision makers to aid in research and manage- staff and implemented under the Land Analysis ment activities. The objectives of the project are to System. Before analyses, TM data will be con- (1) develop a current digital land use and land verted to a Lambert Conformal projection to im- cover inventory of the entire Albemarle-Pamlico prove compatibility with the CGIA A/P data base. drainage area, (2) integrate these data into the A/P The drainage area encompasses several physio- Study data base at CGIA, and (3) develop mecha- graphic provinces including Tidewater, Middle nisms for maintaining and updating the land use and Upper Coastal Plains, and Piedmont. Win- and land cover data. Up-to-date, accurate land use dows roughly corresponding to the physiographic and land cover data are not currently available for provinces will be created from each TM scene so North Carolina, but would serve a critical need in that an area under consideration at any one time the user community. will be fairly uniform with respect to topography, soils, moisture, and other physical characteristics. Methodology A guided clustering algorithm will be used for an initial separation of each area into broad, Level I, categories. Each broad category will then be bro- General Approach ken down into more detailed categories. Training Landsat TM digital data will be used to map sites in the TM scenes will be used for testing land use and land cover over the entire Albe- which combination of bands is best suited for dis- marle-Pamlico Study area. Data will be partly criminating more detail within each broad cate- classified based on, U.S. Geological Survey gory. Spectral signatures of the more detailed (USGS) Standard Level II categories with a min- cover types will also be determined by interactive imum mapping unit of 5 to 10 acres. Land use and guided clustering of digital data for training sites. land cover information will then be integrated The clustering algorithm developed at CGC pro- with theA/P Study data base being maintained at vides the analyst with an interactive display ofthe CGIA, and procedures for updating the informa- spatial distributions of clusters at each iteration tion will be outlined. and with final cluster statistics. The clusters will be compared with maps or photographs ofeach site Remote Sensing Data Acquisition to determine if the clustering process has ade- quately characterized a training site. We expect Five Landsat TM scenes cover all but a very that several clusters or spectral signatures will be small portion of the Albemarle-Pamlico drainage found representing each cover type. Cluster statis- basin. The study area encompasses nearly one- tics will be compiled, nonunique clusters will be third the land area of North Carolina and a por- deleted or merged, and confusion areas will be tion of southeastern Virginia. Cloud cover made identified. the most recent (1989) TM scenes unsuitable for At this point, the list of land use and land cover use over much of the area, but five scenes from categories may be revised to show categories not winter of 1987 and 1988 (November, December, previously included but which are distinct on the and January) have been acquired from Earth Ob- imagery, or to merge categories that cannot be servation Satellite Company. Aerial photography adequately separated. Cluster statistics will then required for location of training sites or verifica- be used to categorize an entire window by using a tion of classification accuracy will be obtained K-means minimum distance classifier. Data from from existing sources. the various windows (and categorical levels) will be recombined and classification accuracies will be Data Analysis evaluated before transfer to CGIA. Remotely sensed data will initially be digitally Registration and Vertical Integration analyzed by North Carolina State University per- of Data sonnel at CGC facilities. The major-image process- ing software package at CGC is the Land Analysis CGIA has aquired the Earth Resources Data System running under the T!ransportable Applica- Analysis Systems (ERDAS) software, which is tions Executive, both of which were developed by compatible with the ARQ/INFO GIS. Classified REGioNAL AND FFDERmSTATE CooPERATPvE PPOGRAms 137 image data from the Land Analysis System at the to 10-acre mapping unit and recommended a clas- Computer Graphics Center is being converted to sification scheme compatible with the USGS stan- ERDAS format and transferred to CGIA as classi- dard hierarchical land use and land cover classifi- fications are completed. Personnel from CGIA will cation scheme. This classification scheme would complete the transfer from ERDAS to the provide a framework for the identification of broad ARC/INFO Albemarle-Pamlico Study data base. categories, but is flexible enough to permit aggre- This transfer initially consists of vectorizing the gation or greater separation at lower levels. For land use data and entering them into ARC/INFO. example, while the project is committed only to Land use data will have to be registered to the differentiating between forested and nonforested existing AIP data base to ensure geometric accu- wetlands (Level II), it is fully expected that greater racy and data continuity. separation will be possible. For instance, the user community would benefit from information on the Final Results relative distributions of salt marshes versus fresh- water coastal marshes. Particular interest has The inventory will provide complete coverage been expressed in determining if stands of Atlantic for all but about 3% of the A/P drainage area. The white-cedar (Chamaecyparis thyoides) can be iden- data will be georeferenced to the North Carolina tified from the digital data. These possibilities will State Plane Coordinate System and will be inte- be investigated as the project progresses. grated with the existing A/P data base. In October 1989, a half-day introductory training Classified color-coded image data will be repro- session at the Computer Graphics Center intro- duced in photographic format at an approximate duced CGIA personnel to the basics of remote-sens- scale of 1:250,000. Results of the classification may ing technology. Topics included terminology, char- also be plotted in map format at variable scales. acteristics of Landsat TM data, and a discussion of CGIA plans to produce a series of acre summary the approach to be used in completing the project. reports of land use by county and subbasin. CGIA Two people from CGIA have been designated to will also produce digital files of land use data in a work on the A/P Study and are being trained in standard data exchange format that can be distrib- digital image processing at the Computer Graphics uted on a cost-recovery basis for use in GIS's in- Center. CGIA has just received ERDAS, and com- stalled in county, local, and regional agencies, and pleted installation in January 1990. ne total fand- to other A/P cooperating agencies, such as EPA, ing for this portion of the project, including GIS USGS, and the U.S. Army Corps of Engineers. implementation, is $139,622. In addition, procedures for maintaining and up- dating the information will be in place at the com- Time Schedule pletion of this project. CGC and CGIA will produce a report that describes the techniques used to Image classification was completed on the first develop the land use and land cover inventory, image by the end of March (1990). This also defines the classification scheme, documents the marked completion of formal training of key CGIA limitations of the satellite data, and describes the personnel. Work on conversion of the Land Analy- data available at CGIA for the project area. sis System data to ERDAS format began as soon as CGIA received the first scene. By early spring Project Status 1990, procedures for classifying and transferring the data will have been tested and verified. The In summer 1989, an advisory committee met to target date for completing the raster-to-vector con- review the project and to discuss a proposed clas- version and integration with ARC/INFO is 30 Sep- sification scheme. The committee consisted of rep- tember 1990. Final results of the image classifica- resentatives from Federal, State, and local agen- tion are expected to be available through CGIA by cies, including the U.S. Fish and Wildlife Service, October 1990 and are currently available for se- EPA, the North Carolina Department of Agricul- lected areas. ture, North Carolina Divisions of Coastal Manage- ment and Environmental Management, Univer- Relevance of Project Results sity of North Carolina, North Carolina State University, CGIA, and city, county, and regional The need for land use and land cover data has planning agencies. The committee approved the 5- been clearly expressed by managers and research- 138 BiowwcAL RFPoRT 90(18) era concerned with the Albemarle and Pamlico cies, representatives of county and local govern- sounds. No accurate assessments of the contribu- ments in the A/P Study area, and university per- tions of nonpoint sources to instream. water-quality sonnel. The committee represents the data needs problems can occur without up-to-date information of the user community. The commi e'sobjectives on land use. Assessing the effects of nonpoint are to (1) assist in refining the land use classifica- source activities on eutrophication will be critical tion scheme, (2) identify critical areas for which for developing effective management strategies. more detailed resource data are needed, (3) rec- Resource analysts may also use acreage estimates ommend output products, and (4) plan for future of land uses to sensitive areas to estimate loading data needs beyond the time frame of this project. values for sediments, nutrients, or toxic substances The State of North Carolina is considering the for use in water-quality or groundwater models. use of Landsat TM or other remotely sensed data Researchers will also use the information for wild- for developing land use and land cover informa- life habitat analyses or multistage sampling. Re- tion for the entire State on a regular 2- to 5-year source managers require the information for eval- basis. The advisory committee will help evaluate uating proposed development, determining the the results of this AT Study to determine if the proximity of a particular land use to water intake approach produces products that will meet the locations, point source discharges, or other critical needs of the user community. point locations,,or for generating acre summary reports for land use categories. Acknowledgments An advisory committee has been established by the Computer Graphics Center and CGIA to over- We wish to acknowledge the cooperation and see the project. This committee consists of re- participation of the Center for Geographic Infor- source managers from Federal and State agen- mation and Analysis. REGioNAL AND FEDERAL-STATE COOPERATIVE PW)GRAMS 139 Loss of Coastal Wetlands in Louisiana: Cooperative Research to Assess the Critical Processes by S. Jeffress Williams U.S. Geological Survey 914 National Center Reston, Virginia 22092 and Asbury H. Sallenger, Jr. U.S. Geological Survey 600 Fourth Street, South St. Petersburg, Florida 33701 ABSTRACT.-Erosion of the Nation's shorelines and loss and deterioration of our coastal wetlands are widespread and serious problerns that affect all regions of the United States. As a result of natural and human-induced factors, the coastal plain of Louisiana, which contains 40% of the tidal wetlands in the conterminous 48 States, is undergoing the greatest amount of coastal erosion and wetlands lose of any State in the Nation. The barrier islands that provide a natural buffer for Louisiana's deltaic plain environments are experiencing erosion rates of 20 m/year, while wetlands losses are about 100 km2/year. In response to these problems and the lack of scientific understanding of the processes causing erosion and land loss, the U.S. Geological Survey has, since 1986, conducted field investigations in Louisiana, working closely with the U.S. Fish and Wildlife Service and other Federal and State agencies. Research elements included in the studies of Louisiana's coastal barriers and wetlands are (1) the shallow geologic framework, (2) documentation by maps and aerial photographs of the physical changes that have occurred during the past 135 years, (3) measurements of several critical processes in the coastal zone and in a typical sediment-starved or sediment-rich basin, and (4) transfer of the results and findings to coastal resource managers. Studies of a similar nature are also under way in Lake Michigan and along the Alabama-Mississippi coast. More than one-half of the U.S. population lives bordering a coast are experiencing erosion and within a 1-hour drive of the Nation's marine or wetland deterioration, and 26 of these States suf- Great Lakes coasts, and the density of population fer from an overall net erosion of their shorelines. and development in the coastal zone is predicted The National Academy of Sciences forecasts an to increase into the 21st century. At present, de- increase in sea level rise; this would accelerate veloped coastal areas face potential loss of life and coastal erosion and wetland degradation. billions of dollars in property damage because of The physical processes causing wetlands loss long-term coastal erosion and storm effects. In and barrier island erosion are complex and varied, addition, valuable coastal wetlands and estuarine and many are not well understood. In addition, the habitats are being rapidly altered as a result of technical and academic community debates about natural and human-induced factors. All 30 States which of the many contributing processes, both 140 BIOLOGICAL REPORT 90(18) natural and human-induced, are most significant. Federal agencies and State geological surveys as Controversy also surrounds some of the measures well as academic researchers. that are being proposed to mitigate erosion and reduce wetlands loss. Much of the debate is focused Louisiana Barrier Island on the reliability of predicted results of a given management, restoration, or erosion mitigation Erosion Study technique. With better understanding of the phys- As shown in Fig. 1, much of the territory bor- ical processes of wetlands loss, such predictions dering the Gulf of Mexico is undergoing shoreline will become more accurate, and a clearer consen- erosion. Louisiana, however, has the greatest rate sus should appear on how to reduce erosion and of erosion compared with other Gulf region States, land loss. and also with other coastal States. Much of this erosion occurs along the barrier islands, which act Role of the U.S. Geological as buffers, protecting the wetlands and estuaries Survey in Coastal Erosion and landward from the effects of storms, ocean waves, Wetlands Loss Research and currents. In 1986, the USGS and the Louisiana Geological As the primary Federal agency for conducting Survey (LGS) began a 5-year study that focuses on research and information gathering on all earth the processes causing barrier island erosion. The science topics, the U.S. Geological Survey (USGS) study areas (Fig. 2) extend from the Isles is engaged in studies focused on improving scien- Dernieres to Sandy Point and to the Chandeleur tific understanding of the physical processes affect- Islands east of the Mississippi River Delta. Be- ing coastal environments. USGS's Coastal Geology cause long-term erosion of Louisianars barrier is- Program consists of four major studies: (1) Louisi- lands is due to both sea level rise, relative to the ana Barrier Island Erosion Study, (2) Louisiana land, and diminishing sand supply, the primary Wetlands Loss Study, (3) Southern Lake Michigan objectives of this study are to quantify processes Coastal Erosion Study, and (4) Alabama/Missis- related to sea level rise and sand supply, and to sippi Coastal Erosion and Pollution Study. Each present the results in a form that can be applied to study is being done in close cooperation with other practical problems such as predicting future f ANNUAL SHORELINE CHANGE OKLAR014A ARKANSAS I Erosion Accretion .- - - M 5 3.0- 4.9 - 1.0- 2.9 - L_ @. I -- C= LOLTISLANA ED 77-XAS A:A Al0'A M.bi@e_ P- City Lake cld. New Houston Orleans Tampa Gulf 0 MeXicO ?f C@ Chn.d F- COASTAL EROSION and ACCRETION in the GULF OF MEXICO 0 so 00 300 WES 1@ 0 100 .1.0 .0. KILOMETERS Fig. 1. Map of shoreline erosion and accretion around the Gulf of Mexico. REGIONAL AND FEDERAIL-STATE COOPERATIVE PROGRAMS 141 LOUISIANA NEW ORLEANS HOUMA 0 Atchafalaya IWef Baralaiia BaytS a' Brelon Sound V Teffebonne Bay N&I, 4 e Sandy @Point Wes Derri@es GULF OF MEXICO Fig. 2. Map of the southern Louisiana deltaic plain. The U.S. Geological Survey investigations of barrier island erosion and wetland loss cover regions east and west of the Mississippi River Delta. changes. The study is divided into three main wash, net offshore sediment transport, and gra- parts: dients of sediment transport along the length of III Investigate the geologic framework of the Mis- the shoreline. Careful analyses of tide gauge sissippi River deltaic plain (Fig. 3) where the records show a progressive rise in relative sea barrier islands have formed and migrated land- level over the entire region, with local rates ward. This involves using sediment cores and exceeding 1 cnVyear (Penland et al. 1987, 1989). geophysical profiles to provide a broad regional Most of this rise is due to compaction and sub- understanding of the historical development of sidence of the recent deltaic sediments. A series the barrier islands and a conceptual view of the of field experiments and modeling efforts is processes of barrier island erosion. Compari- being undertaken (e.g., direct measurements of sons of archival maps and photographs of the the waves that wash over the Isles Dernieres coast (from the past 135 years) are yielding barrier islands during winter storms and hurri- accurate measurements of the geomorphic canes). changes taking place (Fig. 4). 0 Assemble the research results as digital data Develop a better quantitative understanding of sets, atlases, and technical reports for use by the processes responsible for erosion. The focus coastal scientists, planners, and engineers. Ap- has been on only a few of the many physical plications of the study results include develop- processes, including relative sea level rise, over- ing better techniques for determining the rate ox': EH,NA Fig. 3. A succession of six Mississippi River deltaic complexes has been de- posited over the past 7,000 years be- cause of channel switching by the LAFOURCHE- -MODERN river. (Adapted from Prazier 1967.) Plaque'----., ya 4 Trinn Shor I" Shoal Bayou Lafour Che Shis) Shoreline Isles Pleistocene Shoal Dernieres Holocene R Miles 50 Barrier Shoreline Subaqueous Gul of Mexico !11111111116E@!@ -- - 0 Kilometers 100 Sand Bodies 142 BIOLOGICAL REPORT 90(18) 4 Terrebonne Bav caiI1014 Bay elln ig Pelto BaY BaY q 'C@q6' -7 . .. .......... ... . . Fig. 4. Widespread erosion and deterio- ration of the Isles Dernieres barrier S DO, island arc since 1853 resulted from 1853 e-r,. co I qLqE rapid rise in relative sea level, lack of sediment, and frequent storm effects on the coast. Terrebonne Ba.v Caillou Ba@' q_@2qW 'qiqv Lake Pelto 0 mi 3 'q@'O4qf qAqfe,-Co ISLES DE6qRq1q4q11E2qV-'qS 0 km 5 at which artificially nourished beaches should port renewable natural resources estimated at a be replenished and predicting future shoreline value of $1 billion per year. However, an estimated erosion so coastal planners can plan construc- 80% of the Nation's tidal wetlands area loss has tion at a safe distance landward qfq@rom the erod- occurred in Louisiana. The areas of greatest loss ing shoreline. are in the modem Mississippi River Delta and the Barataria and Terrebonne basins to the west Louisiana Wetlands Loss Study (Fig. q5q)_ Map comparisons by several scientists have been used to show that wetlands loss has Of the 48 conterminous States, Louisiana has steadily increased during the 20th century to an 2 25% of the vegetated wetlands and 40% of the tidal estimated 100 q1k:m /year by 1978, the latest year for wetlands. These coastal wetlands, including the which detailed measurements are available. If this associated bay and estuarine environments, sup- rate of wetland loss continues, the U.S. Army Corps Coastal Zone q@O 1q@0 k. Baton Boundary Rouge ake Lafayette 0q/,L, 0qtl 1, New Charles Nw. 8q-40 39.4 mi2/yr 30 - 28.1 2qMjq2q/yr LOSS GAIN .q2 20q- 15.8 mi2/yr Very Severe _qj 10 q- Severe .7 mi2q/yr Moderate 1900 1940 1980 Low Years Fig. 5. Map showing the areal distribution and wetland loss rates for coastal Louisiana (adapted from Gagliano et al. 1981). REGioNAL AND FFDFRAi__STATE CooPERATivE PRoGRAms 143 of Engineers estimates that in the next 50 years Southern Lake Michigan nearly 1 n-,dllion acres of Louisiana wetlands will be Erosion Study converted to open water. Conceived as a natural extension of the Barrier Over the past several years, fluctuating water Island Erosion Study, the USGS studybegan in late levels in the Great Lakes, combined with storm 1988 in cooperation with the U.S. Fish and Wildlife waves and surge flooding, have caused significant Service (FWS) and Louisiana State agencies. Em- and widespread coastal erosion and damage, par- phasis is on understanding the critical physical ticularly in urban areas such as Chicago. The processes that cause the extreme rate of wetlands USGS, working closely with the State geological loss in coastal Louisiana and identifying the best surveys of Illinois and Indiana, recently completed management practices to address those losses. the second year of a planned 5-year investigation of This USGS and FWS wetlands study includes the shoreline of southern Lake Michigan. This four parts: (1) baseline data is being compiled and study included surveying the coast and nearshore entered into a computer-based geographic infor- areas to (1) assess the extent of historic erosion, mation system; (2) research is being conducted on (2) investigate the geologic factors controlling the a basin scale to understand some of the critical magnitude and range of water level fluctuations in processes causing wetlands loss; (3) at specific the recent geologic past, (3) locate offshore sand sites, research is being conducted on the effects bodies for use as fill to rebuild beaches and dam- aged portions of the shore, and (4) measure sedi- and utility of various wetlands management ac, ment transport processes throughout all seasons of tivities on the processes; and (4) the information the year. and results from these studies will be relayed to the user community by means of reports, maps, Alabama-Mississippi Erosion and workshops. The wetlands study elements dealing with re- and Pollution Study search on some of the critical physical processes are As in much of Louisiana, the Alabama- being undertaken by USGS scientists as well as Mississippi coastal region is a dynamic system of scientists at the Louisiana Geological Survey and coastal barriers, tidal inlets, wetlands, and large Louisiana State University under contract with the bays and estuaries that are undergoing environ- USGS. Field studies will be conducted in two sepa- mental change due to natural and human activi- rate hydrologic basins, one sediment-rich and the ties. In response to the physical changes taking other sediment-poor, in order to compare and con- place, the USGS, in cooperation with the two State trast the dominant processes in each. Investiga- geological surveys, is undertaking a 5-year study tions are now under way in the sediment-poor Ter- focused on understanding the geologic processes rebonne basin-Timbalier Bay and parts of the that cause erosion and movement of fine-grained Barataria basin (Fig. 2); field studies in the sedi- sediments and pollutants in the coastal zone. The ment-rich Atchafalaya basin will start in 1991. first year of effort, fiscal year 1990, will concen- Research elements under investigation for each trate on deciphering the geologic framework of the basin include: Alabama-Mississippi coastal region. � meteorological forcing events, Summary � fine-grained sediment dispersal, � saltwater and freshwater dispersal, In addition to the four studies currently under � physical processes of marsh deterioration, way in USGS's Coastal Geology Program, several � wetlands soil development, and other activities are in progress. As directed by � subsidence-soil compaction. Public Law 100-220, USGS and the National Oce- In addition, a study contracted to Coastal Envi- anic and Atmospheric Administration have devel- oped a plan for conducting geologic studies along, ronments, Inc., is examining the effects of small- and remapping the coastal zone of, the U.S. portion scale freshwater diversions from the Mississippi of the Great Lakes. This plan, submitted to Con- River on brackish marshes adjacent to the levees. gress in December 1989, recommends a 10-year The duration of the USGS-FWS wetlands study is effort of phased surveys and would include research expected to be 6 years. contributions by agencies in each of the affected 144 BiowwcAL REPoRT 90(18) States. To date, Congress has not provided funds chronology. Pages 287-315 in Transactions of the for implementing this study. Gulf Coast Association of Geological Societies, Congress has also directed USGS to formulate a Vol. 17. plan to extend and expand the present regional Gagliano, S. M., K. J. Meyer-Arendt, and K. M. Wicker. coastal studies into a research program of national 1981. Pages 295-300 in Land loss in the Mississippi scope. This effort is under way and includes obtain- River deltaic plain: Transactions of the Gulf Coast ing recommendations from other Federal agencies Association of Geological Societies. Vol. 31. as well as the appropriate agencies in each of the Penland, S., K. E. Ramsey, R. A. McBride, T. F Moslow, coastal States. This plan was submitted to Con- and K. A. Westphal. 1989. Relative sea level rise and gress in June 1990. subsidence in Louisiana and the Gulf of Mexico. La. Geol. Surv. Tech. Rep. 3. 65 pp. Penland, S., J. R. Suter, and R. A. McBride. 1987. Delta References plain development and sea level history in the Terrebonne coastal region, Louisiana. Pages Frazier, D. E. 1967. Recent deltaic deposits of 1689-1705 in N. C. Kraus, ed. Coastal Sediments the Mississippi River-their development and '87. American Society of Civil Engineers. REGioNAL AND nDER"TATE CoopmuTrm PRormAms 145 Marine Wetland Mapping and Monitoring in Florida by Kenneth Haddad Marine Research Institute Florida Department of Natural Resources 100 Eighth Avenue, S.E. St. Petersburg, Florida 33701 ABSTRACT.-The Department of Natural Resources, Florida Marine Research Institute, has, implemented a program of mapping and monitoring Florida's coastal marine wetland habitat. Because of Florida's extensive coastline and the need for timely monitoring, Landsat Thematic Mapper CM satellite data have been used as the base for the mapping effort. Aerial photography is used for seagrass mapping; the photointerpreted results are digitized into the USGS quad-rectified TM base map. The TM data are processed to distinguish the marine and estuarine emergent vegetation. Although the protocol and techniques for the mapping effort have begun and an initial mapping effort has been completed, a fully established monitoring effort is still in a developmental stage. The success of this program is predicated on the flexibility of using multiple sources of data with a resultant digital product. The State of Florida has one of the most exten- technical data on the status and trends of coastal sive coastlines in the United States and climati- and marine resources have become available, it cally ranges from tropical and subtropical to tem- has become evident that this targeted approach to perate. This has resulted in a complex and diverse management is inadequate over the long term. assemblage of species and habitats that are often Habitat has been lost, species abundance has de- unique and fragile. Florida@s population growth is clined, polluted waters have reduced shellfish har. one of the highest in the Nation, with more than vest areas, and fisheries have been closed. 80% of State inhabitants living within 16 1km of the This realization has stimulated the evolution of coast. The resultant effects on marine and estua- an ecosystem approach to resource management. rine resources, although at times obvious, have This kind of approach is based on the fact that been poorly understood, rarely quantified, and as- without an understanding of species' interactions, sumed to be far-reaching. communities 'community interactions, and cumu- lative environmental impacts (natural and System Analyses and Management human-induced), management actions will often be reactive rather than preventive or corrective. With such a diverse richness of Florida's marine resources and a resultant diverse group of users, management of the State's marine resources is not Habitat Mapping and Trend an easy task. The difficulty is compounded by the Analyses State's rapid growth and the currently unquantifi- able effect of this growth on marine resources. A first step in building a digital ecosystem data A primary goal of the Florida Department of base is the determination of the extent and loca- Natural Resources, Marine Research Institute tion of critical habitat. In 1983, FMRI, through the (FMRI), is to conduct research and synthesize that National Oceanic and Atmospheric Administra- research into information that can be used to make tion (NOAA) Office of Ocean and Coastal Resource sound resource management decisions. Most ma- Management and Florida's Department of Envi- rine resource management strategies and actions ronmental Regulation, initiated a program to map in Florida have been oriented to single species. As and monitor coastal wetlands and submerged 146 BioLoGicAL REPoRT 90(18) habitat, including salt marshes, mangroves, sub- Geographic Referencing merged aquatics, oyster reefs, and unconsolidated bottom. With such an expansive coastline in Flor- TM data consist of six spectral layers of infor- ida, we analyzed unconventional methods for the mation for each V4 acre (30 x 30 m) on the ground mapping effort. and a thermal band with 4-acre resolution. Each Initially, we evaluated mapping techniques to spectral band is rectified to 7.5-min U.S. Geologi- determine cost, accuracy, and production-time cal Survey (USGS) quadrangles in a UTM projec- comparisons between digital image processing of tion by using a bilinear interpolation technique. Landsat Thematic Mapper CM data and carto- Welch et al. (1985) determined that this type of graphic aerial photography methods. A 69% cost process can achieve accuracy standards for saving and 83% production-time reduction was 1:50,000-scale maps and approach the standards realized with TM data (Haddad and Harris 1985a). for 1:24,000-scale maps. Rectification of the indi- We also determined that aerial photography was vidual spectral bands, rather than the finished often needed for photointerpretation and digitiza- product, is standard because of the need to con- tion into the resource map when submerged habi- tinually return to the raw data for additional tats were being mapped (Haddad and Harris analyses. 1985b). In marine wetlands, classification accu- racy for both aerial photographs and TM data was Image Analyses >90%. Based on these results, FMRI began sys- tematically mapping Florida's estuarine and ma- We have not developed a rigid protocol for sta- rine wetlands, excluding the Everglades National tistical analysis of the satellite imagery data, but Park and Biscayne Bay. That effort began in 1984, workable techniques have been standardized. Nu- was completed with updates in 1986, and required merous types of statistics have been tested for about 2 years of effort (1 year = 2080 hours). their ability to classify marine and estuarine wet- lands and for computer processing times. Stan- Trend Analyses dard classifiers, such as the maximum likelihood, Habitat trend analyses also have been com- which can use either supervised or unsupervised pleted for selected areas of the State from the approaches to generate statistical clusters, are 1940's to the present. A major conclusion from the processing-intensive and cumbersome in a pro- trend analyses is that submerged aquatics have duction operation. This observation is based on often experienced the greatest loss, and this loss is our specific needs relative to coastal wetlands and no longer due to mechanical effects, but rather to does not consider the use of this approach for changes in water quality. This conclusion is sup- general mapping needs. With this type of algo- ported by the fact that submerged aquatic losses rithm, and most algorithms in use, the higher the often occur in deeper waters within estuaries, sug- spatial resolution the more difficult it is to resolve gesting insufficient light penetration as a caus- confusion within and among classes. At some ative factor. Uss of marsh and mangrove has point, human intervention with a photoiziterpre- substantially decreased in Florida, and where suf- tive-like process is necessary. ficient protective measures have been established, Our approach has been to use a rapid parallel- increases in aerial extent have been observed piped type of classifier to initially process the data (Haddad and Hoff-man 1985b). into 256 classes. The classifier is run on the green, red, and near-infrared, and the red, near-infrared, Mapping Techniques and mid-irS3rared TM spectral bands, respectively, to generate two statistical images. The first image We needed to decide on a base map (the digital is pictorially similar to a color-infrared photograph map to which all data are referenced) early in the and can be image-interpreted by identifying those program. As is common in many areas, base maps clusters that represent the wetland categories of were not available in digital form on a statewide interest. We found that it is often advantageous to basis, and the cost of digitization was prohibitive. use the second image because of its accentuation Therefore, the only reasonable approach was to on the infrared bands. In particular, we have found make the TM data the base map, and any addi- that the mid-infrared band enhances our ability to tional map layers (i.e., seagrasses, oysters) would differentiate wetlands. In many cases, we use both be digitally rectified to that base. images to selectively differentiate categories of REGioNAL AND FEDERAL-STATE CooPERATwE PRWRAms 147 interest, with the results being a third image com- We have not found any statistical analyses that posed of the best clusters from each image. adequately define seagrasses, although we have Although this approach is rapid and effective it had success in limited cases. Variations in water still does not meet accuracy standards expected for clarity, water depth, and sediment type preclude wetlands mapping when compared with interpre- the use of standard spectral analyses. The image tation of photographs at similar spatial resolu- must be manually photointerpreted in either the tions. The associative and subjective analyses per- blue, green, or red spectral bands. Because of these formed by a photointerpreter are not yet obstacles we commonly use aerial photography (ei- reproducible statistically. On the other hand, use ther existing or contractually flown) to map of the TM mid-infrared band can have advantages seagrasses. The photographs are photointerpreted in certain analyses where identification of differ- and rectified to the Landsat base map, and the ent levels of moisture content enhance the ability seagrass coverage is conventionally digitized as to differentiate wetland types beyond those observ- wetland types into the wetlands data base. able in an infrared photograph. Once the images are clustered as best as can be Habitat Trend Analyses statistically accomplished, National High Alti- tude Mapping Program aerial photographs, exist- Teelmiques ing National Wetlands Inventory (NWI) maps, Trend analyses for coastal wetlands can be con- ground truthing, and many other data sources are used to identify or confirm clusters that are not ducted with numerous techniques. The creation of pure to a given wetland type. For example, some data for actual analyses must be done with caution clusters representing mangroves may be confused because in most cases it is difficult to separate with a wet orange grove or a freshwater wetland, errors in classification from actual habitat resulting in a 70% identification accuracy. The changes. Trend analyses cannot be conducted on remote-sensing literature has many examples of data that use different classification systems that this type of confusion, and it reports the statistical have not been normalized. In fact, it is very diffi- inaccuracies of this type of analysis. The litera- cult to compare data that have been interpreted by ture reflects an academic approach to,image anal- different investigators that use the same classifi- yses and not a production approach. We routinely cation system if tedious interpretive calibrations "fix" the confused clusters by using simple digital are not conducted. If done properly, habitat trend manipulations based on the interpreter's assess- analyses can provide valuable insights on the ef- ment of the data. Orange groves and fi-eshwater fects of habitat management regulations and the wetlands are reclassified into appropriate catego- changes in the resources that use those habitats. ries, often increasing identification accuracies for mangroves >95%. Historical Data This flexible and rapid approach to wetlands Historical analyses have been accomplished for mapping results in a highly accurate product, but many areas in Florida by photointerpreting ar- only for wetlands. We routinely produce a final chived photographs from the 1930's to 1970's. We map product that merges the wetland types with rectify the interpreted data to the Landsat base the original color-infraredlike image. By providing map and table-digitize them into a separate data this pictorial image for the background data, the layer. When we use aerial photography, the inter- user is able to orient to the image and eliminate pretations often must be transferred to a USGS the need for a summary presentation of data not quadrangle to geo-correct the data for spatial in- classified as wetlands. consistencies before digitization. We can often by- Seagrasses pass this step by using a three-point triangulation method when digitizing off the photographs. When Seagrass mapping presents special problems for positional deviations are observed, new points are satellite image analyses. Landsat only collects an picked and the digitization process is continued. If image over a given area once every 16 days. nis the interpretation of the historical photographs is means that conditions conducive to mapping must compatible with the TM analyses, then trend anal- all coincide on that given day. If the water is clear yses can be conducted. We have not attempted to and clouds do not obscure the area, there is a good compare historical Landsat Multipectral Scanner potential for using imagery for seagrass mapping. (MSS) data with the recent TM data because of the 148 BioLoGicAL RFxoRT 90(18) uncertainties introduced by spectral and spatial show both the process and, if using disparate data resolution differences. sources, the problems. The observed areas of change represent differences in final product res- Contemporary Data olution, habitat classifications, and real changes When building a data base for trend analyses, in habitat. Figure 1 a is a general map of a coastal it is important to create an accurate habitat data area of Tampa Bay, Florida. The data have been layer with which historical and future data will be consolidated to three classes and are a digital compared. We concluded that contemporary data representation of the 1982 NWI map. Figure lb should be that layer. Contemporary data can be represents the statistically clustered 1987 TM ground-truthed and corrected for errors in clas i data for the area of mangroves delineated in the B1_ 1982 data. Figure 1c shows those areas that were fication, which cannot be done for historical photo- labeled as mangrove in 1982, but not classified as graphs. This also gives the investigator a "feel' for mangrove in 1987. Quantitatively, the area was the area and increases the potential for accurate reduced from 2,952 ha of mangrove to 2,564 ha, a interpretation of historical photos. By expending 13% loss. However, when investigating the initial efforts in the creation of the contemporary changes, it becomes obvious that a large portion data, a considerable reduction in effort is realized of that change is not real and represents differ- when developing the historical data base and con- ences in interpretation techniques and classifica- ducting future map updates. tion systems. Many of the smaller areas of change are actually uplands within the mangrove com- Data-base Updates plex. These types of features are averaged by the One approach to updating the habitat data base photointerpreter to become mangroves, even is to remap a given area to compare with the though the photography was at the 1:24,000 scale. original maps. That process is time-consuming. In the photointerpretation and digitization pro- We have developed a technique that takes advan- cess it becomes impractical and costly to try to tage of the fact that TM data are digital. When delineate these features at that scale. The pho- working with a focused data base, such as coastal t0interPreter makes a conscious decision to delin- wetlands, we process the new TM data into 256 eate them or they are lumped into the mangrove classes, as previously described. This produces an classes; digital processing automatically rnain- image, rectified to the base map, that can be ma- tains their separation. nipulated to update the original map. The original The use of classification systems also contrib- data are used to mask a given habitat, which is utes to discrepancies in updating data. The NWI then compared in a very rudimentary way with the maps are based on the Cowardin et al. (1979) sys- new TM image. For example, when updating man- tem, whereas the State of Florida uses a modified groves, we would use the original coverage of man- Anderson (1976) system tailored to State needs. In groves to locate those areas in the new TM image Fig. 1c, a 162-ha area defined by NWI as man- that should contain inangroves. Mangroves, in the grove, falls outside the spectral clusters we con- new image, can be expected to fall within a specific Bider mangrove. In fact, this is a salt flat that has range of statistical clusters, and those clusters :530% mangrove and would never be classified as that fall outside that range are identified as poten- mangrove. To confuse the process further, this tial areas of change. These areas can then be same area was called the equivalent of a salt flat visually assessed for changes. In theory, an inverse in the 1950 NWI analyses, and thus shows a mis- process can be used to identify areas of mangrove leading increase of 162 ha of mangroves within the growth, but we have not tested this approach be- same classification system. cause of insignificant amounts of growth in wet- The point to be made is this-trend analyses lands since our initial mapping effort with TM must be conducted with caution and with a full data. evaluation and understanding of the data being compared. In fact, of the 388-ha change between Problems with Disparate Data 1982 and 1987, less than 17 ha are due to real change (<l% change). If the original image used Figures 1 a-1c depict the results of the updating was TM rather than NWI, then the data updating process, except that we have used mangroves dig- would not have the problems that have been iden- itized from a 1982 NWI aerial photographic map- tified. This does not indicate that one process is ping effort as the mask to a 1987 TM image to better than the other, just that they are different. REGioNAL AND FFDERmSTATE C@@M"W @@s 149 A M ....... .. FIGURE IR FIGURE !]!@@11 W A T E R M MRNGROVE.`..@,.@@ M LAND Ni Q $ .. ... . ..... IF I I I rg r-M K L. M.. A- ............... . . FIGURE Ic C Fig. la. 1982 NWI map depicting the location of man- groves used to mask the 1989 TM data. Fig. 1b. Classified 1987 TM data of the areas defaied as NDS mangrove in 1982. Fig. lic. Areas depicted as mangrove in 1982 but not as mangrove in 1987. SALT FLATS (162 HR.) Classification Systems by name. Thus, we name a salt marsh complex a salt marsh, and if we go to the next level of delin- The importance of the classification system can- eation we would name Juncus and Spartina as not be underestimated when using satellite image components of that complex. Our classification, at processing for habitat delineation. This is some- that point, could be cross-referenced with either thing that must be addressed in the initial stages the NWI Cowardin et al. (1979) system or the of the mapping program. Because we have been Anderson (1976) system. Because we are working primarily mapping coastal wetlands, we have cho- with raster data at 30-m spatial resolution, we sen to tailor our classification to Florida wetlands have categories that consist of marsh and water. 150 BIOLOGICAL REPORT 90(18) These areas are often presented as a marsWwater is not considered as part of an ecosystem. The category, which is not used in most classification wetlands are just one layer of information, out of systems. many, that we are building into the Marine Re- In Florida, we have observed that the TM anal- source Geographic Information System. Linkage yses can be better tuned to the Anderson system to dredge and fill permits and other types of per- and can have major discrepancies with the Cow- mits, which will allow us to reconstruct permitted ardin system. It is best to determine the limits of habitat losses that cannot be mapped, is being the classification systems relative to TM process- investigated. Concurrent with our mapping ef- ing and develop a hybrid system. If this is not done, forts, we are conducting field research to assess much effort can be spent attempting to force a species utilization and production within the dif- classification of the data, thus reducing the ability ferent habitats. All of these efforts will eventually to efficiently conduct trend analyses. provide the information necessary to implement an ecosystem approach to coastal resource man- Conclusions agement. The Florida Marine Research Institute has de- References veloped and implemented a coastal mapping effort designed for efficient and cost-effective mapping Anderson, J. R., E. E. Hardy, J. T Roach, and R. E. and monitoring of Florida's geographically expan- Witmer. 1976. A land use and land cover classification sive coastal wetlands. A combination of Landsat system for use with remote sensor data. U.S. Geol. imagery, aerial photography, ground truthing, and Surv. Prof. Pap. 964. 28 pp. ancillary map data is used to produce digital maps Cowardin, L. M., V Carter, F C. Golet, and E. T LaRoe, from a Landsat TM map base. I have described, in 1979. Classifications of wetlands and deepwater habitats of the United States. U.S. Fish Wildl. Serv., a very general presentation, the techniques and FWEVOBS-79/31.103 pp. concepts we employ in the map-making and subse- Haddad, K. D., and B. A. Harris. 1985a. Use of remote quent habitat trend analyses. The success of this sensing to assess estuarine habitats. Pages 662-675 effort has been based on the flexibility built into in 0. T Magoon, H. Converse, D. Minor, D. Clark, and the standardization of the mapping process. L.T. Tobin, eds. Coastal zone '85. Proceedings of the fourth symposium on coastal and ocean Many issues, such as ground truthing and digi- management. Vol. 1. American Society of Civil tal and hard-copy data distribution, have not been Engineers, New York. discussed. All require substantial planning and Haddad, K. D., and B. A. Harris. 1985b. Assessment can become major operational components of an and trends of Florida's marine fisheries habitat: an effective program. We also have evaluated SPOT integration of aerial photography and thematic satellite data for mapping efficiency, and we use mapper imagery. Pages 130-138 in S. K. Mengel and SPOT data when higher resolution mapping is D. B. Morrison, eds. Machine processing of remotely required. The spectral superiority (particular sensed data. Purdue University, West Lafayette, the mid-infrared bands) and lower costs of Land Ind. 370 pp. TM data make its use more advantageous for I e Welch, R., T. R. Jordon, and M. Ehlers. 1985. Comparative evaluation of geodetic accuracy and geographic areas. cartographic potential of Landsat-4 and Landsat-5 Although our habitat mapping effort is impor- thematic mapper image data. Photogram. Eng. tant, it has little long-term meaning if the habitat Remote Sens. 51(11):1799-1812. REGioNAL AND FEDERAL-STATE CooPERATivE PFoGRAms 151 Satellite Data and Geographic Information Systems Technology Applications to Wetlands Mapping by Richard H. Sinclair, Jr., Mark R. Graves, and Jack K. Stoll U.S. Army Engineer Waterways Experiment Station Environmental Laboratory (EN-B) 3909 Halls Ferry Road Vicksburg, Mississippi 39180 ABSTRACT.-Satellite digital images and geographic spatial data-base technology are well suited for mapping and analysis of the vast, complex wetland environments of the lower Mississippi River Valley. We applied these technological resources on two wetlands mapping projects. The first mapping project, completed for the U.S. Fish and Wildlife Service (FWS), illustrates the quantitative value of spatial seasonal data for mallard (Anas platyrhynchos) habitat analysis. Crops and forested areas mapped from a summer Landsat Thematic Mapper scene were combined with flooded areas mapped from a winter scene to derme wintering habitat types and distributions. Results ofthis analysis were used by FWS's National Wetlands Research Center's regional office in Vicksburg, Mississippi, to determine the reliability of a Habitat Suitability Index model for mallards wintering in the lower Mississippi River Valley. Satellite resources are often dominantly competitive or, in fact, the only affordable solution to reliable seasonal data for habitat analysis of large geographic areas. The second project illustrates the efficiency of digital map data-base development to satisfy multiple requirements within the U.S. Army Corps of Engineers for wetlands regulation and impact analysis. The specific area involved is the Yazoo River basin floodplain, which is parallel to the Mississippi River in west-central Mississippi. Tlie basin covers all or parts of a 20-county region. Hydric (wetland) and nonhydric (nonwetland) soils are being digitized fi-om Soil Conservation Service soil survey photo map sheets. Results of this effort will be a high resolution, georeferenced digital data base and accompanying acreage statistics for wetland and nonwetland areas. Mallard Wintering lard (Artas platyrhynclws) wintering habitat vari- Habitat Study ables in the lower Mississippi River Valley (LMV) from Landsat digital images. The scope of work completed by WES-EL for Landsat Thematic Mapper Land Cover the FWS included the analysis of two seasonal Mapping and Habitat Analysis in the Landsat Thematic Mapper (TM) scenes. A sum- Lower Mississippi Valley mer scene was selected in August 1988 for map- ping forests and agricultural cropland classes. A This study was conducted by the U.S. Army winter scene was selected in January 1989 for engineer, Waterways Experiment Station, Envi- Mapping typical surface hydrology (permanent ronmental Laboratory (WES-EL) for the U.S. water bodies and seasonal flooding) that exists Fish and Wildlife Service (FWS). Our objective during the mallard's southern migration. Statis- was to derive georeferenced spatial data on mal- tical land cover information was developed as 152 BioLoGicAL RFPoRT 90(18) input for a Habitat Suitability Index (HSI) model compatible tapes in 6,250 bits per inch, band- of mallard wintering habitat in the LMV (Allen sequential format. The tapes were generated by 1986). The FWS's objective was to use Landsat the Thematic Mapper Image Processing System TM data as a means to validate the mallard HSI at the National Aeronautics and Space model. The FWS scientists required land cover Administration's (NASA) Goddard Space Flight statistics for 16-km2 sample areas within the proj- Center in Greenbelt, Maryland. ect boundaries, as the HSI model specifies an U.S. Geological Survey (USGS) 1:24,000-scale 8-km mallard foraging radius (Allen 1986). FWS maps were used to geometrically rectify the raw personnel selected 25 sample areas from the sum- Landsat TM data. The scene was centered near mer scene based on statistics from 49 sample Greenville in western Mississippi, and covered areas that define the study area. Two of the 25 parts of Mississippi, Louisiana, and Arkansas. 2 sample areas were selected from a region border- The scene size was about 2,000 km . Eighteen ing the 7 by 7 matrix making up the original 49 USGS map sheets were required to locate about sample areas. These two areas were added so that 100 reference (control) points distributed uni- densely forested land cover types would be in- formly throughout the scene. cluded in the analyses. Color-coded maps of land All Landsat data were georeferenced to the Uni- cover classes were plotted for each of the 25 se- versal Transverse Mercator (UTM) r ection lected sample areas. These maps were used to 2P 0j (zone 15), with a grid resolution of 30 in . The final record a FWS census (aerial transect) of the exist- data file represents a matrix of 5,534 lines (verti- ing mallard population near the time of data ac- cal) by 5,368 elements (horizontal). The array ex- quisition. Statistics, including area calculations tracted for the 7- by 7-matrix analysis has the for each class, were developed for the same 25 following UTM coordinates: upper left corner East- sample areas (measuring 16 km2) in both scenes. ing 605970, Northing 3751030; and the lower right Our final results consisted of acreage data for comer Easting 767010, Northing 3585010. This flooded forest, flooded rice fields, and other coordinate space does not include all the area flooded cropland. These results were derived by covered by the two additional sample sites selected overlaying the classified data for the two Landsat at a later date. scenes and extracting acreage"where forests and Color aerial photography, obtained in winter crop classes from the summer scene coincided 1984, and information gathered from subsequent spatially with flooded areas in the winter scene. low-altitude aerial reconnaissance Rights were Three sets of plots were produced for the 25 sam- compared with the sunimer Landsat scene classi- ple areas to illustrate the results for all stages in fication. The photographs were particularly useful the analyses. in differentiating a small percentage of rice fields confused with adjacent forestland in the unsuper- Data Sources vised Landsat data analysis. The United States' Landsat 4 and 5 satellites Computer Hardware and Software Assets carry the TM sensor package as the principal data- gathering instrument. Radiant energy is recorded With the exception of reformatting in the raw in seven wavelength bands. Six of the bands Occur data tapes, all image data analyses for this project in the 0.45 pm (ultraviolet radiation) to 2.35 pm were completed on one 386-based personal com- (reflected infrared radiation) portion of the electro- puter workstation. The personal computer work- magnetic spectrum. Recorded data from these station is enhanced for image processing and geo- channels have a spatial resolution of 30 in. The graphic information system (GIS) applications by remaining band (band 6) is in the 10.4 to 12.5 Pm a number of specific peripheral devices and add-on (thermal infrared) portion of the spectrum and is computer cards, including the following: an Opus recorded with a spatial resolution of 120 m. Systems CLIPPER 32-bit microprocessor board The project study area was located within the with four megabytes of random access memory, correct Landsat TM frame by using the Landsat operating in a UNIX environment at a clock speed Worldwide Reference Systems Map at Path/Row in excess of 30 MHz and executing 4 to 5 million 23,37 (U.S. Geological Survey 1982). FWS person- instructions @per second; a Revolution Number nel acquired both Landsat TM scenes from the Nine 512 by'5-12 by 32-bit image board and 19-inch Earth Observation Satellite Company in Lanham, RGB graphics display monitor, an Archive V4 -inch Maryland. The data were received on computer- (60-megabyte cartridge) streaming tape backup REGioNAL AND FEDERA"TATE CooPERAnvE PnoGRAms 153 system; a GTCO 24- x 36-inch digitizing tablet; and colder temperatures. This tends to lessen the scat- dual Maxtor 320-megabyte hard disk drives. Other tering effect of the atmosphere on visible wave- peripherals were used in generating output prod- lengths. ucts, including a Versatec 36-inch color electro- Imageprocessingand analysis. After the individ- static plotter, Matrix Instruments digital and an- ual channels were viewed on an image display alog cameras, and a Toyo thermal screen dump device, we made decisions as to which data chan- 11 -inch plotter. nels would be used for gathering spectral cluster All image analyses and GIS operations were statistics over the images. The blue band was dis- conducted with the Earth Resources Laboratory carded from both image analyses because of the (now the Science and Technology Laboratory) Ap- excessive haze in the summer scene and-because it plications Software OMAS) developed by NASA. did not contribute significantly to surface water Staff at the Waterways Experiment Station Envi- delineation in the winter scene. After selection of ronmental Laboratory completed the first port Of the proper channels, spectral cluster statistics for ELAS to the personal computer environment, first the summer image were developed with both super- executing under the MS/DOS operating system vised and unsupervised algorithms. Unsupervised and later under the UNIX operating system on the statistics were gathered with the ELAS modules CLIPPER microprocessor. ELAS is a geobased in- Normal Variation and TM M-amer. The Normal formation system originally designed for process- Variation module is designed to compute the nor- ing and -analyzing digital imagery acquired by mal variation of digital count data (reflectance val- multispectral scanners on aircraft or spacecraft, ues from 0-255) found within selected channels of and data digitized fi-om maps. Digitized map data the raster image data. Resulting parabolic coeffi- include polygon data digitized from thematic maps cients for each channel are stored in an ELAS (e.g., soils, forest) and digitized topographic data subfile for use by the ELAS module TM Trainer. such as those distributed by the National Carto- TM Trainer uses a 3- by 3-pixel window to search graphic Information Center. the raw data for homogeneous training fields. The coefficients computed by Normal Variation are Landsat Thematic Mapper Data Classifi- used to model expected variations within the data. cation and Analysis of Results A 3 by 3 field in each channel is considered to be The scope of the image classification and analy- homogeneous if its variance falls below the parab- sis includes the procedures used to input the raw ola for that channel (National Aeronautics and satellite digital image data, converting the image Space Administration 1989). If that field is then data into meaningful terrestrial classes, and anal- determined to be homogeneous in every channel, it ysis ofthe classified data fi-om both Landsat scenes. is stored as one of the preliminary statistics. Once Reformatting Landsat Thematic Mapper and 60 fields have been collected, the two with the computer compatible tape data. Raw Landsat TM smallest-scaled distance are merged, opening one data were read from the tapes into the ELAS of the temporary statistic bins for collection of an- operating environment by using the module Refor- other field. Once this has happened, the process of mat Thematic Mapper Image Processing System, searching for another statistic and merging the two The individual channels ofthe data were displayed most similar statistics is continued throughout the in black and white and in true and false color-com- remainder of the input data. Once all input data posite format to assess the quality of the digital have been processed, the final processing command imagery. The summer scene was judged to be of is used to merge all of the statistics that remain in marginal quality, as scattered clouds were present the 60 temporary bins until no two statistics have throughout most of the scene and the blue and a scaled distance less than 4. green bands (TM bands 1 and 2) exhibited a blur- The supervised statistics for the summer scene ring effect attributed to high humidity during the were derived from field data gathered by FWS's time of satellite data acquisition. The winter scene National Wetlands Research Center Field Station was of relatively high quality except for a single in Vicksburg, and FWS's Patuxent Wildlife Re- band of very thin, wispy, high-altitude clouds ori- search Center. Polygons were digitized that ented east-west across the northern portion of the bounded spatial locations in the data correspond- data. Better-quality data usually are obtained dur- ing to known land cover types. Statistics for the ing the winter months (if cloud cover is absent) pixel values within these polygons were computed because the humidity is normally lower during by using the EI-AS module Supervised Training. 154 BiowmcAL RF.Poirr 90(18) Because the analysis of the winter scene was lim- classes, percent of area covered, and acreage calcu- ited to discriminating water bodies and flooded lations for the summer scene are presented in Table areas (surface hydrology classes), we decided that 1. Classes developed, percent of area covered, and unsupervised statistics-gathering methods would acreage calculations for the winter flooding condi- be the most expedient and reliable. tions are presented in Table 2, and habitat classes We analyzed the final statistics (unsupervised developed, percent of area covered, and acreage and supervised) by comparing statistical distance calculations for each class in the 7- by 7- matrix measures. Specifically, we used the transformed study area are presented in Table 3. divergence measurements, in conjunction with vi- sual display analysis, as a basis for merging or Discussion deleting particular statistical clusters. Trans- formed divergence is a saturating function of di- We encountered considerable difficulty in pro- vergence that has been demonstrated as helpful in cessing the summer Landsat imagery. Atmo- measuring the average difference between two- spheric conditions at the time of scene capture, class density functions (Swain and King 1973). We coupled with poor crop conditions caused by a also used several ELAS modules that produced drought, made it difficult to distinguish between visual representations of multivariate statistics land cover types that should have been spectrally during examination of the statistics. discrete. Relative humidity was excessive at the After confi i the final set of spectral cluster time of image capture; therefore, single and mul- gurmg statistics, we used a minimum distance classifica- tiple scattering in the visible spectrum severely tion algorithm to assign each digital count (pixel diminished the quality of the three visible bands reflectance value between 0 and 255) within the (Landsat TM channels 1, 2, and 3). Because of study area to one of the clusters. We used the ELAS severe drought conditions, most field crops were module Classifier Minimum Distance to obtain the severely stressed. Only those fields that were well image spectral classification for the study area. irrigated displayed a closed canopy condition at Georeferencing the Landsat images. We did geo- the time of image acquisition. metric rectification to the LJTM coordinate system Optimal Landsat classifications are derived after spectral clustering of the images to avoid any when researchers acquire satellite and ground- degradation of the computer-compatible tape data truth data at the same time. When high-quality before statistical analysis. The georeferencing procedure consisted of finding easily recognizable surface features (such as road intersections) on Table 1. Summer land cover types and acreage the image data that were present on the USGS calculations for the complete study area. 1:24,000-scale quadrangle maps. We obtained the Summer UTM coordinates of these points by manual digi- Land cover area calculations tizing procedures, while the image data coordi- Class Description Percent Acres nates (line and element values) were gathered by 1 Agriculture, 2.7 84,173.5 positioning the cursor over the feature and using bare ground EILAS module Common Display"read target"com- 2 Water 3.7 116,064.9 mand. The 100 control points were evenly distrib- 3 Forest 13.5 419,817.8 uted through the entire study area. Once a rela- 4 Sand 0.8 25,736.3 tion was established between the image data and 5 Cloud 3.6 110,984.6 UTM coordinates, we derived a transformation 6 Cloud shadow 1.4 42,339.7 equation through using the ELAS module Com- 7 Grass, shrub, scrub 4.0 122,754.6 pute Mapping Coefficients. We used the resulting 8 Water edge 2.3 70,386.9 9 Agriculture, unknown 1.9 58,701.6 equation to transform the entire study area image 10 Agriculture, 39.8 1,231,307.1 file to the LJTM coordinate projection. These pro- predominantly cotton cedures were applied separately to the summer 11 Agriculture, 8.1 251,987.4 and winter scenes. predominantly soybean Merging Landsat Thematic Mapper classifica- 12 Agriculture, fallow 16.3 515,141.3 tions to obtain a habitat map. We constructed the 13 Agriculture, rice 1.5 48,012.9 final habitat map by overlaying the classified sum- Total mer and winter scene spatial data files. Land cover REGioNAL AND FEDERm-STATE CooPERATrm RtOGRAms 155 Table 2. Winter land cover types and acreage Table 3. Knal habitat types and acreage cal- calculations for the complete study area. culations for the complete study area. Winter and summer scene analyses are combined. Winter Landsat Thematic Mapper land cover types Winter and summer and area calculations final habitat classification Class Description Percent Acres Class Percent Description Acres 1 Nonwater 86.9 2,693,000.8 1 93.3 Nonfl t-d 2,889,185.5 2 Water 13.1 404,408.2 2 0.3 Flooded rice 8,269.8 Total 3 4.7 Flooded agriculture 146,661.6 (other) 4 1.7 Flooded forest 53,292.0 Subtotal 208,223.4 Landsat imagery does not coincide with ground- Total 3,097,409.0 truth data, more accurate results may require the use of archive data from a previous year. Also, if separation of crop types is especially important for a satellite data study, crop calendars should be These boundaries are digitized, displayed, and consulted in conjunction with local weather condi- edited on a color video monitor, converted to grid tions when selecting imagery, so that spectral dif- cell format (raster), and put into the proper coor- ferences among land cover classes can be maxi- dinate space in the GIS. Final operations are mized. performed to adjust and edit data along map sheet boundaries in the GIS. County boundaries and Yazoo Basin Wetlands Mapping project river reach boundaries are also digitized so that retrieval of wetland locations and size The WES-EL, Environmental Systems Division (acres) can be done by county or river reach. remote-sensing applications team is involved in a wetlands mapping project for the U.S. Army Corps Source Data of Engineers in the Vicksburg District. The objec- The mosaics prepared for digitizing are com- tive of this project is to create a georeferenced posed of four SCS Soil Survey photo map sheets digital wetlands data base for the Yazoo River at a scale of 1:15,840 or 1:20,000, joined together. basin in west-central Mississippi. The mapping's Georeferencing procedures are accomplished with being completed to aid regulatory personnel in USGS 1:62,500-scale quadrangles. All soils data addressing requirements set forth in Section 404 are rectified to the UTM coordinate system and of the Clean Water Act of 1977. gridded at a resolution of 20 m. The project area covers about 4.5 million acres of predominantly agricultural land. The database Computer Hardware and Software Assets under development is a 20-county area in the Yazoo River basin, a major tributary to the Mis- The personal computer workstation used in sissippi River. The work involves the acquisition developing the digital map data base of Yazoo of Soil Conservation Service (SCS) photomosaic River basin wetlands has an identical configura- soils maps for each of the 20 counties. An exten- tion to that previously described for waterfowl sive reconnaissance of SCS soil types presented on habitat mapping. The remote sensing applications these maps was made in the field by the authors team within WES-EL has four such workstations. and staff from SCS, the U.S. Environmental Pro- Operators of three digitizing workstations help tection Agency, and Vicksburg District personnel. complete work on 151 mosaics. The digitizing soft- Soil types were categorized as hydric or nonhyd- ware used is the commercial software package ric. A consultant, who was formerly an SCS em- from Earth Resources Data Analysis System. ployee and the principal developer of the SCS Data-base Development Methodology hydric and nonhydric: classification methodology, also accompanied personnel in the field. Based on A generalized description of the sequential steps this review, the hydric or nonhydric soils bound- required to develop the digital map data base and aries were traced on the photomosaic soils maps. to calculate wetland acreage follows. 156 BiowaicAL REPoRT 90(18) 1. Aggregate hychic and nonhydrie soils on individ- to adjust all the control points, including the two ual SCS soil survey photo map sheets and as- initial points, in the digitized file in the X and Y semble the mosaics. directions. These adjustments are sometimes 2. Select georeferencing control points on mosaics made to get a better geometric fit of each map and 1:62,500-scale USGS maps. sheet data set as it is gridded into the GIS master 3. Digitize all hydric and nonhydric soil bound- data base. Immediately after each mosaic digi- aries. tizer file is gridded, it is displayed and its spatial a. Map data are digitized at any one of the three relation to the surrounding hydric and nonhydric personal computer workstations. Resulting soils is carefully examined. Gridding of the indi- digitizer files are copied onto 5.25-inch floppy vidual mosaics into the larger digital data-base disks and delivered to the data-base integra- file is analogous to fitting very small pieces into a tion administrator. large puzzle. However, the gridding process is b. Digitizer data files for individual maps (mo- mathematically controlled; therefore, any offsets saics) are gridded into the GIS master data- (greater than 40 in) must be corrected by transla- base file containing a UTM coordinate space tion. The previously selected control points may covering the entire project area. Some editing be used by shifting the gridded mosaic in relation may be necessary where map sheet bound- to the average difference, in northing and easting aries join because of photomosaic distortions (UT1@D control point locations. This reduces the or other cartographic irregularities encoun- amount of editing required to fill in small data tered in the data integration process. gaps, and it allows smooth transitions for data 4. Wetland acreages are calculated by county and overlaps at the edges. reach and displayed on the color video monitor or plotted as hard-copy maps and transparent Conclusions USGS map overlays. The application of satellite digital image data To date, 4 of 20 counties have been digitized and and GIS technology is a highly effective technique gridded into the GIS master data-base file. Defin- for rapid and accurate wetlands mapping and ing accurate coordinate reference points on the analysis, especially for large inaccessible wetland photomosaic map sheets is a critically important complexes. Satellite data resources as a national step before the digitizing operation. The process asset are grossly underused for inventorying wet- for selecting two diagonally opposed reference lands and monitoring changes over time. GIS ca- points begins by locating road L-Aersections or pabilities offer tremendous advantages for quan- other permanent landmarks that are readily iden- titative analysis and visualization of spatial tifiable on the mosaic map sheet and the corre- relation that are so important to regional wet- sponding 1:62,500-scale USGS quadrangle. Once lands analyses. This ability to investigate spatial these two points are located, the USGS quadran- relations challenges scientists to exploit analyti- gle is placed on the digitizing table and set up. The cal modeling techniques for experimenting with upper left and lower right map corner coordinates new concepts that will increase the scientific are read from the map in latitude and coordinates, knowledge of wetland processes. The national keyed into the computer, and related to a digitizer goal of no net loss in wetlands provides the impe- file by digitizing each corresponding reference tus to apply these advanced technologies rou- point. Next, the two specific mosaic digitizer setup tinely and effectively in meeting or exceeding this points are digitized off the USGS quadrangle *in order to identify those coordinates for later use in objective. digitizing the hydric and nonhydric soil bound- aries. An average of six additional control points Acknowledgments per quadrangle also are located, and their UTM coordinates are recorded. These points must be Funding for the Mallard Wintering Habitat selected at landmarks that are visible on the mo- Study was provided by the U.S. Fish and Wildlife saic map sheet. Differences are calculated be- Service's National Wetlands Research Center. tween common point coordinates from USGS Our work was assigned and coordinated with J. R. maps and the mosaic map for the six additional Nassar and M. W. Brown of the National Wet- control points. An arithmetic average is calcu- lands Research Center. Their cooperation and ex- lated for the difference and, if necessary, is used cellent technical support throughout the study RF,GioNAL AND FEDERAi-STATE CooPERATrm PRoGRAms 157 was greatly appreciated. M. R. Graves, WES-EL, technical coordination and supervision of work was responsible for the Landsat image analysis, conducted in both studies. development of statistical habitat data, and pro- duction of high-resolution color maps. References The Yazoo River Basin Wetlands Mapping Task was funded by the U.S. Army Corps of Engi- Allen, A. W. 1986. Habitat Suitability Index models: neers, Vicksburg District. This work was con- mallard (winter habitat, lower Mississippi Valley). U.S. ducted under the guidance of K. D. Parrish, the Fish Wildl. Serv., Biol. Rep. 82(10). 37 pp. project manager of the upper Yazoo River Basin National Aeronautics and Space Administration. 1989. Project. E. J. Clairain, Jr., WES-EI!s research ELAS, Science and Technology Laboratory team leader, and his staff were responsible for Applications Software, programmer reference, Vol. 1, delineating wetland boundaries on maps. J. Tin- user manuals Vol. II and III. Report 183. John C. gle of WES-EL was responsible for quality control Stennis Space Center, Miss. of map and mosaic preparation to ensure accuracy Swain, P H. and R C. King. 1973. Two effective feabire and compliance with cartographic standards for selection criteria for multispectral remote sensing. digitizing. J. S. Hutto, also of WES-EL, was re- Pages 5W--540 m Rvoeedings ofthe first international joint conference on pattern recognition, IIEEE. Cat. No. sponsible for generating a composite of all digi- 73 CHO 82 1-9 C. Piscataway, N. J. tized maps into a single geographic data base; she U.S. Geological Surey. 1982. Index to Landsat worldwide also was the coordinator of all computer-based reference systems (WRS) Landsat 1,2,3, and 4, Sheet operations required to produce the final product. 9. National Oceanic and Atmospheric Administration, J. K. Stoll of WES-EL was responsible for overall USGS-EROS Data Center, Sioux Falls, S. Dak. RFr,ioNAL Ai-m FMERAL-STATE CooPERAnvE PRoaRws 159 The Digital Wetlands Data Base for the U.S. Great Lakes Shoreline by Michael Scieszka Land and Water Management Division Michigan Department of Natural Resources Stevens T Mason Building P.O. Box 30028 Lansing, Michigan 48909 ABSTRACT.-Michigan has mapped and digitized a detailed land cover and land use inventory. The program processes the land cover and land use files into various theme maps, including a set of wetland maps. The wetlands map set is used to implement inventory and public information requirements of the State's wetland protection act. The data collection methodology and digital processing environment are being used by the International Joint Commission to map the remainder of the United States' shoreline. I present an overview of the Michigan Resource Inventory Program, how the data were collected, and how to access the data. The Michigan Resource Inventory Act (1979 PA for a land cover and land use data base for the 204) authorized the Michigan Department of Nat- Great Lakes shoreline. The DNR, in close cooper- ural Resources (DNR) to conduct a statewide land ation with the U.S. Army Corps of Engineers cover and land use inventory. The land cover and (COE), was assigned the authority to acquire ae- land use inventory mapped seven main catego- rial photography for the shoreline, interpret land ries: urban land, agricultural land, openland, cover and land use data, digitize the maps, and forestland, water, wetlands, and barrens. This act deliver various data sets for IJC use. The IJC required DNR to digitize the inventory and to intends to use this data to quantify the effects of distribute the data in a format that maximizes various shoreline and water level management its use in local planning and zoning. During the scenarios it is considering recommending to the same legislative session, Michigan adopted the governments within the Great Lakes basin. Goemaere-Anderson Wetland Protection Act (1979 PA 203), which required DNR to provide for Proiect Scope and Inventory the preservation, management, protection, and Methods use of wetlands. PA 203 also required DNR to make a wetland inventory of the State, file it with The land cover and land use inventory is a local governments, and use the inventory data as component of the Michigan Resource Inventory one of the identifiers of wetlands protected under Program (MRIP) created by PA 204. This program: the statute and administrative rules. The land cover and land use inventory required through PA 0 manages and distributes the results of a 204 is being used to meet wetland inventory re- statewide 1:24,000 color-infrared aerial quirements of PA 203. photography flight made in 1978-79; The International Joint Commission (IJQ, 0 manages and distributes the results of a through its Great Lakes Water Level Reference of 1986-87 1:15,840 black and white infrared 1986 (U.S. State Department and Canadian Min- reflight of the northern two-thirds of istry of External Affairs 1986), identified the need Michigan, and a 1:24,000 black and white 160 BIOLOGICAL REPORT 90(18) panchromatic flight from 1988, which cover and land use classifications have been covered the remainder of the state; mapped. The minimum-size mapping unit is 1.5 to � operates the Michigan Resource Information 5 acres. Our system is hierarchical to resolve ques- System (MIRIS), which is the umbrella for tions of double or multiple category classifications. geographic information system (GIS) The system allows for further classification refme- processing in Michigan. MIRIS contains a ments to enable those users with specific needs to digital base map for the State, the land cover inventory smaller areas with greater detail and and land use inventory, and, in selected exactness. areas, soils data, prime lands information, To assist the photointerpreters, the inventory and other thematic overlays (see Appendix A, program contracted with the Michigan State Uni- B, C, and D); versity Center for Remote Sensing to develop a � provides mapping and GIS services to the report entitled A Photo Interpretation Key to Mich- Great Lakes research community through igan Land Coverl Use. This report lists each cate- the Great Lakes Information System. The gory to be mapped and provides the definition and geographic focus of these products is along interpretive characteristics, such as tone and the shoreline and into the Great Lakes, color, texture, pattern, and shape. Stereo appear- where DNR is encoding such information as ance and commonly associated land cover and land fish spawning sites, bathymetry, bottom use activities were also presented when applicable. sediments, sensitive shoreline features, and An example of this report follows in a section on wetlands; Emergent Wetlands. � implements a statewide groundwater data Description: base that verifies public and private water These are areas dominated (30 percent well locations and digitizes the verified or more cover) by erect, rooted herbaceous location along with the water well log to hydrophytic plants which are growing out create a data-base record containing such of standing water or waterlogged soils. information as the well's static water level, Typical emergent plants are cattails, bul- geological formation encountered, and depth; rushes, rushes, reed grass, bur reed, arrow and � provides con Itractual servicesto public and arum, arrowhead, pickerelweed and private organizations in need of mapping and sedges. GIS services. Marsh areas containing emergent types of aquatic plants can be differentiated The land cover and land use inventory, which from aquatic beds and open water by the was completed between 1981-86, is an important magenta hues indicative of denser vegeta- component of all the activities and services pro- tive cover and by a coarser texture. Sepa- vided by the Michigan Resource Inventory Pro- rating emergent marshes from shrub gram. To-complete this inventory in an orderly and swamp usually poses little problem be- consistent manner, the Michigan Resource Inven- cause of differences in texture and pattern, tory Program established standards, a data collec- and when viewed stereoscopically, height. tion methodology, a classification system, and Some emergent types have very distinc- training and quality-control procedures. tive signature characteristics. For exam- The land cover and land use inventory used ple, hybrid cattail (7@ypha glauca), when 1978-79 aerial photography. The aerials are at a canopy is homogenous and completely scale of 1 inch to 2,000 feet (1:24,000) color-infra- pure, is a bright green hue on CIR photog- red photography. The photography mission was raphy. A midseason shift from crimson or flown between 1 June and 30 September to ensure magenta to green for non-hybrid cattails maximum leaf-on condition of trees. Sixty percent indicates a decrease in IR reflectance asso- overlap was shot for stereo viewing. ciated with dehydration of mesophyll de- The Michigan Resource Inventory Program generation accompanying early senes- adopted a land cover and land use classification cence. system that was designed to make the best use of Muskrat houses may be detectable in information from our aerial photography. The pro- emergent wetlands as small, distinctive gram started with the system by Anderson (1972), white dots. Some may be ringed by a nar- and expanded it to a third level. About 60 land row dark band of water. Since cattails and REGioNAL AND FEDERAL-STATE CooPERATivE PRoGRAms 161 bulrushes are the principal vegetation rected overlay and the base were reviewed and comprising muskrat habitats, the photo corrected by the inventory program's chief cartog- interpreter can be certain that one or both rapher before digitization. In all, more than 1,100 plants are present. land cover and land use overlays were prepared Reed grass varies in color from greens to between 1981 and 1986. On the average, each reds to pinks, depending indirectly upon overlay contained 2,000 distinct land cover and water quality and soil moisture. However, land use polygons. A little more than $1 million it has a characteristically smooth, velvety were invested in the photointerpretation effort. texture due to lack of leaf bending and The 1978-79 aerial flight cost nearly $350,000. large distinct heads. It is often found in The same basic procedure is being used for IJC disturbance areas (e.g., dredge spoil work, although we did add one substantial wetland deposits). category called Coastal Submergents, which are Interpretive Characteristics: defined as areas contiguous to the shorelines of the Great Lakes where rooted submerged aquatic Color: Red, deep red-brown, blue green, plants are dominant. dull green or mottled white patches of bleached stalks (cattail, bur reed), medium Description of Mapped and red-brown, dark gray red-brown, browns, Digital Products olive drabs, dark greens (bulrush, rush), strong pink, purplish pink (pigweed, The land cover and land use inventory overlay smartweed), pinks (sedges), light pink, was either hand-digitized at the Michigan Re- gray pink, gray blue gray, gray blue source Inventory Program or subcontracted to dig- (grasses), brilliant green blue, dark green itizing service bureaus. The digital data are struc- blue, white (dead vegetation). tured as line strings/text file. Each boundary line Texture: Normally medium but may be between different land cover and land use areas smoother or fine if stands are pure; emer- was digitized, and a single text per polygon area gents may have a slightly granular tex- was inserted. The internal coordinate system used ture. for georeferencing is the Michigan State Plane Shape: Irregular. Coordinate System for the State of Michigan files and is in latitude and longitude internal coordi- Pattern: May be concentric or banded nates for IJC work. around lake. After the data were encoded, they were avail- Site: Occurs in depressions in moraine, till able in four basic forms: plain and outwash and frequently borders 0 line/text file, which can be plotted in scales open water in such depressions, shallow ranging from 1 inch to 1,320 feet to 1 inch to shoreline areas of lakes. 2,000 feet. This product will contain all land With the aerials and classification system, the cover and land use including wetlands; photointerpreters prepared the land cover and * digital version of the line/text file can be land use inventory through relatively standard produced in Intergraph Design File Format, procedures. Clear sheets of acetate were placed Standard Interchange Format, or Data over a photo being viewed under a stereoscope, Exchange Format; Homogenous land cover and land use polygons 0 an acreage report, which quantifies land were delineated, interpreted, and coded. Supple- cover and land use by governmental units, is mentary source materials, such as older invento- published;and ries, soils data, and topographic maps, were used 0 a patterned theme map, which selectively when available. The acetate overlays were trans- displays various land cover and land use ferred to a stable mylar overlay registered to a polygons. screened mylar of the U.S. Geological Survey topo- Appendix E shows a line/text example and a graphic base map. During this transfer, photo dis- wetland theme map example. The theme maps are tortion was corTected by "rubber sheeting" the ac- one of our most popular products. To generate a etate overlay to roads, property lines, and woodlot theme map, we process the line,/text file through features on the topographic base. The final cor- a series of routines built with IntergrapYs Spatial 162 BioLcGicAL REPoRT 90(18) Editor/Spatial Analyst program. The routines includes shrub and small or stunted trees. This search the linEAext files, pull out the polygon class includes both stable shrub wetlands and wanted, and pattern the resultant file to highlight areas in a successional stage leading to wooded the theme. wetlands. Some of the predominant species To generate a wetland theme map, the follow- include alder, dogwood, sweetgale, leatherleaf, ing categories are searched for and extracted. willow-buttonbush associations, and water (Note: The following definitions are from the clas- willow. Any standing dead trees, shrubs, and sification system adopted for the Michigan Re- stumps should be in the 612 category. source Inventory Program; this system is the one 621 Aquatic Bed Wetland being used for IJC effort.) The 621 category is used to map an area that 414 Lowland Hardwoods generally has 300/6 or more vegetation cover of Ash, elm, and soft maple, along with cottonwood, submerged, floating-leaved or floating plants, and balm of Gilead, and other lowland hardwoods. is less than 2 m deep. Typical plant species are 423 Lowland Conifers yellow water, lily, duckweed, and pond weeds. Lowland species category, including areas of 622 Emergent Wetlands predominantly cedar, tamarack, black and white These are wetland areas dominated (30% or more spruce, and balsam fir stands. cover) by erect, rooted, herbaceous hydrophytic 51 Streams and Waterways plants, which are present for most of the growing This category includes rivers, streams, creeks, season in most years. These areas are usually canals, drains, and other linear bodies of water. dominated by perennial plants, although annuals Where the water course is interrupted by a control are often present too. Typical species include structure that creates an impoundment, the cattail, bulrush, sedges, reeds, wild rice, impounded area should be classified as a pickerelweed, arrowhead, and so forth. reservoir. The boundary between streams and 623 Flats lakes, or reservoirs, is the straight line across the These are level or nearly level deposits of mouth of the stream. unconsolidated sand, mud, or organic sediments, 52 Lakes with less than 75% aerial coverage of stones, Lakes are nonlinear water bodies, excluding boulders, or bedrock, and less than 30% aerial reservoirs. A water body should be classified as a coverage of vegetation other than pioneering lake if a structure has been installed primarily to Plants- regulate or stabilize lake levels without significantly increasing the water area. The Map Product Availability delineation of a lake will be based on the areal extent of water at the time the data are collected. The Michigan Resource Inventory Program 53 Reservoirs maps and digital data are available through three Reservoirs are artificial impoundments of water, methods. First, the program has all files in a whether for irrigation, flood control, municipal or readily retrievable form. Appendix F is a sche- industrial water supply, hydroelectric power, or matic of the overall system used to digitize, store, recreation. process, and output digital data for Michigan. People can call or write to acquire either plots or 611 Wooded Wetland digital versions of the inventory program's data. 1 This class applies to wetlands dominated by trees A second method of obtaining maps and digital more than 6.1 m tall. The soil surface is seasonally data is through the "local holders of map sets." flooded with up to 30.5 cm of water. Several levels When a county's land cover and land use base map of vegetation are usually present, including trees, and other data sets are processed, that county shrubs, and herbaceous plants. Some of the pre- dominant tree species include ash, elm, red maple, cedar, black spruce, tamarack, and balsam fir. A noininal fee is charged. For instance, if a user wanted a single quadrangle ofland coverdata, plots would cost$45 and 612 ShrublScrub Wetland digital $60. Base maps for the same area would cost $35 for a plot and would cost $50 for digital. The Michigan Resource This class applies to wetlands dominated by Inventory Program is authorized to use these fees during the woody vegetation less than 6 in tall. Vegetation fiscal year to cover operating and staff costs. RwioNAL AND FEDERmSTATE CoopiERATivE PRWRAms 163 receives a set of mylars or a digital version of these Resource Inventory Program and the U.S. Depart- data. This is required through PA 204, and it ment of Agriculture's Soil Conservation Service of follows the legislative intent of the program, Michigan have jointly worked on encoding modern which is to assist local governments in making soil surveys. We will continue providing the abil- land use decisions by providing them with accu- ity to identify hydric soils in relation to wetlands rate land resource data. Counties ' also in- and other vegetation. The Michigan Inventory formed that the wetlands data is to be considered Resource Program is also working on a proposal as a preliminary wetland inventory of their area to integrate SPOT Image Corporation satellite as required by PA 203. imagery with existing land cover and land use Land cover and land use data can also be ob- files to identify where land cover and land use tained through COE. The COE is establishing GIS changes are occurring. This imagery, along with processing capabilities in its Detroit District of- recent reflights, gives us the ability to update land fice. As part of our working relationship with COE, Michigan's shoreline land cover and land cover and land use date for Michigan. use files, and the remainder of the U.S. shoreline files we are developing for IJC, will be delivered to the COE. References U.S. State Department and the Canadian Ministry of ]Future Activities External Affairs. 1986. Great Lakes water level reference of 1986. U.S. State Department, Our staff will be focusing its future wetlands Washington, D.C., and the Canadian Ministry of activities in two areas. Since 1985, the Michigan External Affairs, Ottawa, Canada. 10 pp. REGioNAL AND FEDERMSTATE COOPERAnVE PRWRAMS 165 Appendix A. Current Use Inventory Status Map of Bangor Township, Bay County MAP I KEWEENA CURRENT USE INVENTORY STATUS ONTONAGON 2 BIC LUCE 2 �tjjjOICKINSON 11 ALGER MACKINAC HIPPEWA CELIA SCHOOLCRAFT 2 MENOMINEE MME F CHEBOYGAN PRESGUE ISLE HARLEVO X F 2 2 2 F F F MON - ANT IM OTSEGO MORENCY ALPENA DELIVERED TO COUNTY (40) LEELANAU 2 F 2 (n F 2 ORAN PATTERNED MAPS BENZIE RAVER KALKASKA CRAWFORD OSCOCA ALCONA 2 1 LINE & TEXT MAPS MANISTEE WEXFORD HISSAUKEE ROSCC OGEMAW 10SCO ARENAC DETAILED FOREST INVENTORY. (D I I CURRENT USE DATA IS AVAILA MASON LAKE OSCEOLA CLARE ADW IN Rim AND IS BEING READIED FOR BLE ER I Ef 1 2 DELIVERY TO COUNTIES, k46CE;@@ MECOSTA SABELLA MIDLAND BAY CHECK WITH MIRIS REGARDING STATUS AND COPIES. I NEWAYOO TUSCOL CN7CALM GRATIOT NESEE SANILAC m 2 ST CLAIR APEER MACOMB I AWA N IONIA I I I I NOTE: DIGITAL COPIES OF ALLEGAN BARRY -EATON INGHAM VING, OAKLAND CURRENT USE MAPS CAN BE OBTAINED I I I FROM MIRIS,SEE PRICE INFORMATION VAN N KAL AZOO CALHOUN WAS WAYNE AT THE BACK OF THIS CATALOG. 2 MARCH 1990 ST JOSEPH BRANCH SOAL ENAVEE MONRO 166 BIOLOGicAL REPoRT W18) EXAMPLE PATTERNED MAP V, .2 PATTERNED MAPS ARE PRODUCED BY SELECTIVELY RETRIEVING FROM LINE & TEXT FILES SPECIFIC THEMES. THE EXAMPLE MAP IS A WETLAND THEME MAP PLOTTED ON THE BASE FEATURES. OTHER THEMES INCLUDE URBAN, AGRICULTURE AND OPEN. FORESTLAND, AND EXTRACTIVE. 6@ LEGEND LOWLAND HARDWOOD FRI (414,611) SHRUB WETLAND 1612) EMEMERGENT WETLAND (622) AOUATIC BED (623) EXAMPLE LINE & TEXT MAP 113 5 41 m LINE & TEXT MAPS REPRODUCE THE 13 24 ORIGINAL PHOTO INTERPRETATION. 7 3 Iq3 THE MAPS CONTAIN THE CLASSIFICATIONS 612 414 612 13 414 146 LISTED ON THE LEGEND BELOW. 113 31 146 113 5 32 31 113 126 414 115 113 31 126 126 nr 113 412 31 414 CURRENT LAND CCIVER/USE LEGEND 31 113 URBAN NONFORESTED 11 RESIDENTIAL 31 HERBACEOUS 11 Ill MULTI-FAMILY.HIGH RISE 32 SHRUB 412 414 414 32 112 MULTI-FAMILY.LOW RISE 113 SINGLE FAM LY.OUPLEX FORESTED 21 115 MOBILE HOME PARK 41 DECIDUOUS 411 NORTHERN HARDWOOD 14 12 COMMERCIAL. SERVICES. INSTITUTIONAL 412 CENTRAL HARDWOOD 32 121 PRIMARY/CENTRAL BUSINESS DISTRICT 413 ASPEN/WHI7E BIRCH 113 122 SHOPPING CENTER/MALL 414 LOWLAND HARDWOOD 124 SECONDARY BUSINESS/STRIP COMMERCIAL 42 CONIFEROUS 3 126 INSTITUTIONAL 421 PINE 13 INDUSTRIAL 422 OTHER UPLAND CONIFER 1318 INDUSTRIAL PARK 423 LOWLAND CONIFER EXAMPLES ARE FROM A SMALL PART OF 14 TRANSPORTATION, COMMUNICATIONS. UTILITIES 429 CHRISTMAS TREE 141 AIR TRANSPORTATION WATER 142 RAIL TRANSPORTATION 1 51 STREAM BANGOR TOWNSHIP,BAY COUNTY 43 WATER TRANSPORTATION 52 LAKE 144 ROAD TRANSPORTATION 53 RESERVOIR MILES 145 COMMUNICATIONS 54 GREAT LAKES 146 UTILITIES WETLANDS 0 0.5 1 2 17 EXTRACTIVE 61 FORESTED 171 OPEN PIT 611 WOODED 172 UNDERGROUND 612 SHRUB. SCRUB 173 WELLS 62 NONFORESTED 19 OPEN LANO,OTHIER 621 AGUATIC BED SOURCE. 1978-79 1:24,000 COLOR-INFRAR@D 193 OUTDOOR RECREATION S22 EMERGENT 194 CEMETERIES Q3 FLATS PHOTOGRAPHY ILIVINGSTON, MACOMB, ST.CLAIRI AGRICULTURE BARREN 21 CROPLAND 72 BEACH. RIVERBANK WASHTENAW AND WAYNE COUNTIES HAVE BEEN 22 ORCHARDS 73 SAND DUNE 23 CONFINED FEEDING 74 EXPOSED ROCK UPDATED TO 1985) 24 PERMANENT PASTURE 29 OTHER REGioNAL AND F@' DERMSTATE COOPERATIVE PROGRAMS 167 Appendix B. Base Map Status by County and Example Base Map for Bangor Township, Bay County 4=1 KEWEEMAW HOUGHTON ONTONAGON BARAGA 2 LUCE MACKINAC CHIPPEWA CRAFT MENOMINEE MME CHEBOYGAN PRESOLIE ISLE 0Q IX ANTRIM 0 sg MONT. I MOREb[C LEEL0" 2 2 2 CRAND BASE MAP DELIVERED TO COUNTY; BENZIE TRA@ERS@ KALKASKA WFORD OSCODA COMA ALSO AVAILABLE THROUGH MIRIS 2 1 MANIST E WEXFORD MISSkJKEE ROSCOMM OrEMAW I JOSCO 1'=4000- SCALE MAPS I AREMAC 1'=2000' SCALE MAPS MASON LAKE OSCEOLA L W N 1 2 HLRON AVAILABLE THROUGH MIRIS OCEANA MECOSTA ISABELLA IDLAN SAY BASE MAPS COMPLETED AND NEW TOO TUSCOLA MONTCALM GRATIOT SAGINAW OEWSEE SANILAC 2 LAI NOTE; DIGITAL COPIES OF 1 2 LAPEER BASE MAPS CAN BE OBTAINED 0 TAW KENT IONIA CLINTON IAWASSEE ACOMB FROM MIRIS, SEE PRICE INFORMATION 2 1 1 1 AT THE BACK OF THIS CATALOG. ALLEGAN 13ARRY @ EATON Ivi T OAKLAND I I I SWUREN KALAMAZOO CALM" VASHTENA WAY MARCH 1990 BERRIEN JOSEPH BRANCH LE14AWEE No 168 BioLoGicAL RFPoRT 90(18) INTERSTATE HIGHWAYS U.S. HIGHWAYS STATE HIGHWAYS COUNTY ROADS MINOR AND LOCAL ROADS ------------- TWO-TRACK ROADS 19 AIRPORTS GRASS AIRSTRIPS RAILROADS t ABANDONDED RAILROADS RIVERS, STREAMS, AND LAKES 30 DRAINS AND INTERMITTENT STREAMS TRANSMISSION LINES BEAVERI OIL AND GAS PIPELINES POLITICAL BOUNDARIES ? SECTION LINES + SECTION CORNERS 31 32 -A5N,R5E 3 I 80YISCOUT RD. Z IMMER W,4VLjV 4 5 m WHEELER RD. - --------- 7 9 10 8 It r-1 RDA _j ------- MILES 0,5 1 2 SOURCE: USGS TOPOGRAPHIC BASE MAP SERIES REGIONAL AND FEDERAL-STATE CoopEmnw PRoGRAms 169 Appendix C. Michigan Resource Information System (MIRIS) and Great Lakes Information System (GLIS) Summary Status of Digital Map Products CHART I MICHIGAN RESOURCE INFORMATION SYSTEM RM GREAT LAKES INFORMATION SYSTEM Ob SUMMARY STATUS OF DIGITAL MAP PRODUCTS COUNTY ALCONA 9000 101 ALeONA ALGER 0.0 1 1 0 0 ALGER ALLEGAN 0 INIII 00 0 ALLEGAN ALPENA 9000 101 1 ALPENA ANTRIM goo I I 1 0 --0 0 ANTRIM ARENAC -00,z, INI 1v XOOO ARENAC BARAGA goo I 1 1 K BARAGA BARRY 0 INI I BAY 0 0 C7 INI I'v 000 BAY 0 BENZIE 00 1 101 101 BENZIE BERRIEN -0 10 BERRIEN BRANCH - 2 CALHOUN - CASS 29200 CHARLEVOIX 2920 11 10 0 0 CHARLEVOIX CHEBOYGAN 00 1 1 1 J-j CHEBOYGAN CHIPPEWA -2 0 101 CHIPPEWA CLARE 2 0 CLINTON 0000 0 CRAWFORD -- 0 0 101010 0 DELTA -- 0 0 010101 .0 10, DELTA DICKINSON 9220 101 F 10 0 EATON -00 0 EMMET goo o EMMET GENESEE o U GLADWIN 9du GOCEBIC 9-00 1 1 1 1 1 GOGEBic GRAND TRAVERSE 92 0 0101 1 101 1 1 101 1 GRAND TRAVERSE GRATIOT -22 IN I HILLSDALE X22 1 1 HOUGHTON moo @5 HOUGHTON HURON @5 -V -IC7 +,v 1XI010101 HURON INGHAM -00 IN ..I-oo IONIA 0 -REF-1 10SCO -auvl IN IvI IOSCO IRON 9 0 0 0 101- 1 10, ISABELLA 922-INEE-1: JACKSON 9 INIJ -1 KALAMAZOO 0 INF7 -E KALKASKA 00 1 1 I-A KENT ' 000 @5 -0, 0COMPLETE PARTIAL OR IN PRODUCTION 0 INL I -J- 0 too 14 NOT APPLICABLE 9MAPS DISTRIBUTED TO COUNTIES NOVEMBER 1989 170 BiowmcAL REPoyrr W18) Appendix D. Michigan Resource Information System (MIRIS) and Great Lakes Information System (GLIS) Summary Status of Digital Map Products CHART 2 MICHIGAN RESOURCE INFORMATION SYSTEM GREAT LAKES INFORMATION SYSTEM 0&ft SUMMARY STATUS OF DIGITAL MAP PRODUCTS, COUNTY KEWEENAW 10101 1 1 1 0 0 KEWEENAW LAKE 0101 1 INI LAPEER 9 0101010INI LEELANAU 010i I I I LEELANAU LENAWEE 0 0 LIVINGSTON X 010i I INI - - - LUCE -001 1 0 --0 LUCE MACKINAC 001 1 0 - 0 MACKINAC MACOMB X00 - - - MACOMB MANISTEE X -q 2 @q 2 - 0 MANISTEE MARQUETTE X000 0 MARQUETTE MASON -00 10-61 101 MASON MECOSTA 22- MENOMINEE 0 'z' 0 MENOMINEE MIDLAND X 0 01010 N, MISSAUKEE 001 1 1 MONROE X001 I NI 01 1 1 1 MONROE MONTCALM 0 1 1 INI MONTMORENCY X 0 01 1 1 1 MUSKEGON 0 01 1 141 - - - r 1 101 MUSKEGON NEWAYGO 0 OAKLAND X00 NI OCEANA 00 101 101 OCEANA OGEMAW 6-@5--u ONTONAGON 0 0 1 10, ONTONAGON OSCEOLA 0, OSCODA 001 1 OTSEGO 900101 0100 OTTAWA 001 1 INI OTTAWA PRESQUE ISLE -001 1 1 1 - - - PRESQUE ISLE ROSCOMMON 0 0 SAGINAW -001 1 INI SANILAC 0 1 1 INI I I I T-Tj SANILAC SCHOOLCRAFT 001 1 1 1 SCHOOLCRAFT SHIAWASSEE 001 1 ItAl ST. CLAIR 9 0 01':'1 INI ST. CLAIR ST.JOSEPH -0 1 1 INI TUSCOLA 0 TUSCOLA VAN BUREN 2 0- F-F0101 10 VAN BUREN WASHTENAW 9 0 0 WAYNE 9 0 WAYNE 0 0 0 0 WEXFORD -5 -5 0 COMPLETE -v PARTIAL OR IN PRODUCTION N NOT APPLICABLE MAPS DISTRIBUTED TO COUNTIES NOVEMBER 1989 REGioNAL AND FEDERAL-STATE Coopmmw PRoGRAms 171 Appendix E. Example of a Patterned Map and Example of a Line and Text Map of Bangor Township, Bay County MAP 3 EXAMPLE PATTERNED MAP PATTERNED MAPS ARE PRODUCED BY SELECTIVELY RETRIEVING FROM LINE & TEXT FILES SPECIFIC THEMES. THE EXAMPLE MAP IS A WETLAND THEME MAP PLOTTED ON THE BASE FEATURES. "Irk- 'umm OTHER THEMES INCLUDE URBAN, AGRICULTURE AND, OPEN, FORESTLAND, AND EXTRACTIVE, UEGEND M LOWLAND HARDWOOD (414,611) 0 SHRUB WETLAND (612) MEMERGENT WETLANO (622) MAOUATIC BED (623) EXAMPLE LINE & TEXT MAP 1 41 193 LINE & TEXT MAPS REPROOUCE THE 113 121 ORIGINAL PHOTO INTERPRETATION. 32 414 193 THE MAPS CONTAIN THE CLASSIFICATIONS 612 612 414 146 LISTED ON THE LEGEND BELOW. 113 31 146 113 5 32 113 12 115 113 31 31 126 1 ." 126 412 414 113 31 31 CURRENT LAND COVER/USE LEGEND 113 URBAN NONFORESTED It RESIDENTIAL 31 HERB C GUS @@t, MULTI-FAMILY.HIOH RISE 2 S RABE M 3 H U 414 414 32 ULTI-FAMILY.LDW RISE FORESTED @:3 SINGLE FAMILY.OUPLEX 41 DECIDUOUS 414 211 5 MOBILE HOME PARK 411 NORTHERN HARDWOOD 32 12 COMMERCIAL. SERV ICES, INSTITUTIONAL 412 CENTRAL HARDWOOD 113 121 PRIMARY/CENTRAL BUSINESS DISTRICT 413 ASPEN/WHITE BIRCH @22 S"OPPING CENTER/MALL 414 LOWLAND HARDWOOD 24 SECONDARY BUSINESS/STRIP COMMERCIAL 42 CONIFEROUS 31 126 INSTITUTIONAL 421 PINE 13 INDUSTRIAL 422 OTHER UPLAND CONIFER 138 INDUSTRIAL PARK 423 LOWLAND CONIFER 14 TRANSPORTATION. COMMUNICATIONS, UTILITIES 429 CHRISTMAS TREE 141 AIR TRANSPORTATION WATER @42 RAIL TRANSPORTATION 51, STREA 43 WATER TRANSPORTATION 5 LA EM 144 R AD TRANS ORTATION SERVOIR EXAMPLES ARE FROM A SMALL PART OF 54 OR ' 45 COOMMUN[CATFIGNS 53 REEAT LAKES 146 RUTILITIES WETLANDS 17 EXT AC IVE 61 FORESTED BANGOR TOWNSHIP,BAY COUNTY @721 OPEN PIT 611 WOODED 7 UNDERGROUND 612 SHRUB. SCRUB MILES 173 WELLS 62 NONFORESTED PEN LAND,OTHER 621 AQUATIC BED 19 0 0 0.5 1 2 ;q3 OUTDOOR RECREATION 622 EMERGENT 94 CEMETERIES 623 FLATS AGRICULTURE 2 1 BARREt CR OPLAND 2 BEACH, RIVERBANK 0 C 7 22 CR HARDS 73 SAND DUNE 23 ONFINED FEEDING 74 EXPOSED ROCK 24 PERMANENT PASTURE 21 OTHER 172 BioLa,=AL REPoRT W18) Appendix F. Michigan Resource Information System (MIRIS) Configuration DIAGRAM 3 MIRIS SYSTEM CONFIGURATION 7717-@ 68K BW I&NTYL COrTY 68K COLOR SON VTAYL S IVA W RK 68K COLOR 4129 USER WORK MID 3@ A INTERPRO 220 120 MG HD 340 1:1 120 G HD-- 340 120 IAG HD WILD 32C UCM,GPPU,LIB 80 MG HD ZFI 748 ElEl 340 20 MG HD 559MG MUX _f@@ SYSTEM ETC VAX 11/785 ETHERNET INTERGRAPH 250 .. 16MB MEMORY TAPEI Y TEM A 965 ETC SPAN/SPED/DMRS IGDS TAPE2 T ,@F@l L@E VT "VE E @M 0' RIH-D GPPU MOOT r3 PLANNING VT VT VT v GEOLOGICAL PUBLIC USER COMMANDS. LWM 2ND FLOOR SURVEY HEALTH LIBRARIES.FONTS,ETC 68K BW REAL ESTATE DIV. 990 68K BW G 4VT@@ APPENDix 173 Appendix. Habitat Loss and Modification Working Group Co-chairs: John D. Buffington Region 8, Research and Development U.S. Fish and Wildlife Service U.S. Department of the Interior 1849 C Street, N.W., Room 3245 Washington, D.C. 20240 Ford A. Cross Beaufort Laboratory, Southeast Fisheries Center National Marine Fisheries Service National Oceanic and Atmospheric Administration Beaufort, North Carolina 28516-9722 Coordinator: Sari J. Kiraly National Ocean Pollution Program Office, CSIOP Office of the Chief Scientist National Oceanic and Atmospheric Administration Universal Building, Room 625 1825 Connecticut Avenue, N.W. Washington, D.C. 20235 Franklin S. Baxter National Mapping Division U.S. Geological Survey 590 National Center Reston, Virginia 22097 John W. Bellinger Office of Environmental Policy, CECW-RE U.S. Army, Corps of Engineers 20 Massachusetts Avenue, N.W. Washington, D.C. 20314-1000 Virginia Carter Water Resources Division U.S. Geological Service 430 National Center Reston, Virginia 22092 Don W. Field Ocean Assessments Division National Ocean Service, N/OMA31 National Oceanic and Atmospheric Administration 6001 Executive Boulevard, Room 305 Rockville, Maryland 20852 174 BIOLOGICAL REPORT 90(18) Robert Middleton Branch of Environmental Operations Mineral Management Service U.S. Department of the Interior Mail Stop 644 381 Elden Street Herndon, Virginia 22070 Dorene Robb Office of Wetlands Protection U.S. Environmental Protection Agency A-104F 401 M. Street, S.W. Washington, D.C. 20460 Michael Slimak and Steve Cordle Office of Environmental Processes and Effects Research U.S. Environmental Protection Agency RD-682 West Tower, Room 609 401 M. Street, S.W. Washington D.C. 20460 John Sutherland Office of Oceanic Research Progr s National Sea Grant College Program Oceanic and Atmospheric Research National Oceanic and Atmospheric Administration 1335 East West Highway, Room 5226 Silver Spring, Maryland 20910 Billy M. Teels Soil Conservation Service U.S. Department of Agriculture South Agriculture Building, Room 6144 P.O. Box 2890 Washington D.C. 20013 Bill 0. Wilen National Wetlands Inventory U.S. Fish and Wildlife Service 400-ARLSQ U.S. Department of the Interior 18th and C Street, N.W. Washington D.C. 20240 Kiraly, SariJ., Ford A. Cross, andJohnD. Buffington. 1990. Federal Coastal Kiraly, SariJ., Ford A. Cross, andJohnD. Buffington. 1990. Federal Coastal Wetland Mapping Programs. U.S. Fish Wildl. Serv., BioL Rep. 90(18). Wetland Mapping Programs. U.S. Fish Wildl. Serv., MOL Rep. 90(18). 174 pp. 174 pp. A workshop was held in December 1989 to examine the Federal effort in A workshop was held in December 1989 to examine the Federal effort in mapping the Nation@s coastal wetlands. The workshop took place at the U . S. mapping the Nation@s coastal wetlands. The workshop took place at the U.S. Fish and Wildlife Service's National Wetlands Research Center in Slidell, Fish and Wildlife Service's National Wetlands Research Center in Slidell, Louisiana. The papers presented at the workshop and recommendations for Louisiana. The papers presented at the workshop and recommendations for improving the Federal effort are contained in this report. improving the Federal effort are contained in this report. Key words: Coastal wetlands mapping, mapping programs, wetlands. Key words: Coastal wetlands mapping, mapping programs, wetlands. Kiraly, Sari J., Ford A. Cross, and John D. Buffington. 1990. Federal Coastal Wettand Mapping Kiraly, Sari J., Ford A. Cross, and John D. Buffington. 1990. Federal Coastal Wetland Mapping Programs. U.S. Fish Wildl. Sem, M. RepL 90(18).174 pp. Programs. U.S. Fish Wildl. Serv., Biol. Rep. 90(18).174 pp. A workshop was held in December 1989 to examine the Federal effort in mapping the Nation's A workshop was held in December 1989 to examine the Federal effort in mapping the Nation's coastal wetlands. Ile workshop took place atthe U.S. Fish and Wildlife Service's National Wetlands coastal wetlands. The workshop took place atthe U.S. Fish and Wildlife Service's National Wetlands Research Centerin Slidell, Louisiana. The papers presented atthe workshop and recommendations Research Center in S lidell, Louisiana. The papers presented at the workshop and recommendations for improving the Federal effort are contained in this reporL for improving the Federal effort are contained in this report. Key words. Coastal wetlands mappmg, mapping programs, wetlands. Key words: Coastal wetlands mapping, mapping programs, wetlands. NOTE: The mention of trade names does not constitute endorsement or recommendation for use by the Federal Government. TAKE PRIDE rt T o'Fr in America SFXWCE 6 U.S. DEPARTMENT OF THE INTERIOR FISH AND WILDLIFE SERVICE As the Nation's principal conservation agency, the Department of the Interior has responsibility for most of our nationally owned public lands and natural resources. This includes fostering the wisest use of our land and water resources, protecting our fish and wildlife, and providing for the enjoyment of life through outdoor recreation. The Department assesses our energy and mineral resources and works to assure that their development-is- in- the -best-interests of all our people. The Departmnet also has a major hdian reservation communities and for people who , 11111111111 11111 S. administration. 3 E668 00001 8103