Marine debris survey manual

[From the U.S. Government Printing Office, www.gpo.gov]

NOAA Technical Report NMIFS 108 April 1992

Marine Debris Survey Manual

Christine A. Ribic
Trevor R. Dixon
Ivan Vining

SHII U.S. Department of Commerce
.A44672
no.108

NOAA Technical Reports NMFS

The major responsibilities of the National Marine Fish- continuing programs of NMFS; intensive scientific reports
eries Service (NMFS) are to monitor and assess the abun- on studies of restricted scope; papers on applied fishery
dance and geographic distribution of fishery resources, to problems; technical reports of general interest intended to
understand and predict fluctuations in the quantity and aid conservation and management; reports that review, in
distribution of these resources, and to establish levels for considerable detail and at a high technical level, certain
their optimum use. NMFS is also charged with the devel- broad areas of research; and technical papers originating
opment and implementation of policies for managing na- in economics studies and in management investigations.
tional fishing grounds, with the development and Since this is a formal series, all submitted papers, except
enforcement of domestic fisheries regulations, with the sur- those of the U.S.-Japan series on aquaculture, receive peer
veillance of foreign fishing off U.S. coastal waters, and review and all papers, once accepted, receive professional
with the development and enforcement of international editing before publication.
fishery agreements and policies. NMFS also assists the
fishing industry through marketing service and economic Copies of NOAA Technical Reports NMFS are avail-
analysis programs and through mortage insurance and able free in limited numbers to government agencies, both
vessel construction subsidies. It collects, analyzes, and federal and state. They are also available in exchange for
publishes statistics on various phases of the industry. other scientific and technical publications in the marine
sciences. Individual copies may be obtained from the
The NOAA Technical Report NMFS series was estab- U.S. Department of Commerce, National Technical Infor-
lished in 1983 to replace two subcategories of the Tech- mation Service, 5285 Port Royal Road, Springfield, VA
nical Report series: "Special Scientific Report -Fisheries" 22161. Although the contents of these reports have not
and "Circular." The series contains the following types of been copyrighted and may be reprinted entirely, reference
reports: scientific investigations that document long-term to source is appreciated.

Recently Published NOAA Technical Reports NMFS

96. Marine flora and fauna of the eastern United 102. Marine ranching: proceedings of the seventeenth
States--Copepoda, Cyclopoida: Archinotodelphy- U.S.-Japan meeting on aquaculture; Ise, Mie
idae, Notodelphyidae, and Ascidicolidae, by Patricia Prefecture, Japan, 16-18 October 1988, edited by
L. Dudley and Paul L. 111g. January 1991, 40 p. Ralph S. Svrjcek. May 199 1, 180 p.

97. Catalog of osteological collections of aquatic mam- 103. Benthic macrofauna of the New York Bight,
mals from Mexico, by Omar Vidal. January 1991, 1979-89, by Robert N. Reid, David J. Radosh, Ann B.
36 p. Frame, and Steven A. Fromm. December 1991, 50 p.

98. Marine mammal strandings in the United States. 104. Incidental catch of marine mammals by foreign
proceedings of the second marine mammal and joint venture trawl vessels in the U.S. EEZ of
stranding workshop; Miami, Florida, 3-5 Decem- the North Pacific, 1973-88, by Michael A. Perez and
ber, 1987, edited by John E. Reynolds III and Daniel Thomas R. Loughlin. December 1991, 57 p.
K. Odell. January 1991, 157 p.
105. Biology, oceanography, and fisheries of the North
99. Marine flora and fauna of the northeastern United Pacific transition zone and subarctic frontal zone,
States: erect Bryozoa, by John S. Ryland and Peter edited byJcrry A. Wetherall. December 1991, 92 p.
J. Hayward. February 1991, 48 p.
106. Marine ranching: proceedings of the eighteenth
100. Marine flora and fituna of the eastern United States: U.S.-Japan meeting on aquaculture; Port Ludlow,
Dicyemida, by Robert B. Short, February 1991, 16 p. Washington, 18-19 September 1989, edited by Ralph
S. Svdcek. February 1992, 136 p.
101. Larvae of nearshore fishes in oceanic waters near
Oahu, Hawaii, by Thomas A. Clarke. March 1991, 107. Field guide to the searobins (P@ionotus and Bel-
19 P. lator) in the western North Atlantic, by Mike Russell,
Mark Grace, and Elmerj. Gutherz. March 1992, 26 p.

SH11.A44672 NO.108
FEB 19, 1996

NOAA Technical Report NMFS 108

Marine Debris Survey Manual

Christine A. Ribic
U.S. EPA Environmental Research Laboratory
Corvallis, OR

Trevor R. Dixon
The Tidy Britain Group
United Kingdom
Ivan Vining
Center for Quantitative Science
University of Washington
Seattle, WA

April 1992

UNITED STATES OF AMERICA
DEPARTMENT OF COMMERCE

U.S. DEPARTMENT OF COMMERCE
Barbara Hackman Franklin, Secretary
National Oceanic and Atmospheric Administration
John A. Knauss, Under Secretary for Oceans and Atmosphere
National Marine Fisheries Service
William W. Fox Jr., Assistant Administrator for Fisheries

LIBRARY
NOAA/CCEH
1990 HOBSON AVE.
SC 29408-2623

The National Marine Fisheries Service (NMFS) does not approve,
recommend or endorse any proprietary product or proprietary
material mentioned in this publication. No reference shall be made
to NMFS, or to this publication furnished by NMFS, in any adver-
tising or sales promotion which would indicate or imply that NMFS
approves, recommends or endorses any proprietary product or pro-
prietary material mentioned herein, or which has as its purpose
an intent to cause directly or indirectly the advertised product to
be used or purchased because of this NMFS publication.

Contents

Introduction v

How to Use This Manual vi

Chapter 1: Methodologv I
Definition and Categories of Marine Debris 4
General Monitoring Guidelines 8
Chapter 2: Shipboard Sighting Surveysfor Large Debris Item 13
General Description 13
Objectives and Purpose 13
Population of Interest 13
Field Measurement 13
Quality Assurance Program 17
Field Sampling Designs 18
Analytical Procedures 18
Summary 19
Appendix 20
Chapter 3: Shipboard Trawling Surveysfor Small Debris Item 25
General Description 25
Objectives and Purpose 25
Population of Interest 25
Field Measurement 25
Quality Assurance Program 30
Field Sampling Designs 31
Analytical Procedures 31
Summary 32
Appendix 33
Chapter 4: Beach Surveysfor Small to Large Debris Item 37
General Description 37
Objectives and Purpose 37
Population of Interest 37
Historical Information 38
Field Measurement 39
Pilot Studies 40
Quality Assurance Program 42
Field Sampling Designs 42
Analytical Procedures 47
Case Studies 47
Alaska Beach Surveys 48
The Tidy Britain Group 49
Summary 53
Appendix 54

Chapter 5.- Benthic Surveysfor Large Submerged Debris Items 61
General Description 61
Objectives and Purpose 61
Population of Interest 61
Historical Information 61
Trawl Surveys 62
Field Measurement 62
Quality Assurance Program 63
Field Sampling Designs 65
Analytical Procedures 66
Submersible Surveys 67
Field Measurement 67
Quality Assurance Program 68
Field Sampling Designs 69
Analytical Procedures 69
Diving Surveys 69
Field Measurement 69
Quality Assurance Program 71
Analytical Procedures 73
Summary 73
Appendix 74
Chapter 6. Aerial Surveys 77
General Considerations 77
Aerial Photography 77
Conclusion 78

Glossary 79
List of Acronyms 81

Citations 83

Introduction

Over the last several years, concern has increased about and four anonymous reviewers. Any remaining errors or
the amount of man-made materials lost or discarded at omissions are ours. For the technical editing of this
sea and the potential impacts to the envifonment. The manual, we wish to thank Marcus Duke of the School of
scope of the problem depends on the amounts and types Fisheries, University of Washington.
of debris. One problem in making a regional compari- For the Tidy Britain Group case study, the survey data
son of debris is the lack of a standard methodology. The and other information were compiled as part of the Tidy
objective of this manual is to discuss designs and method- Britain Group's Marine Utter Research Programme and
ologies for assessment studies of marine debris. from joint studies with the Advisory Committee on Pollu-
This manual has been written for managers, research- tion of the Sea. Tim Dixon was a co-researcher, with addi-
ers, and others who are just entering this area of study tional support from the Keep Wales Tidy Campaign and
and who seek guidance in designing marine debris sur- the University of Keele. Technical assistance was provided
veys. Active researchers will be able to use this manual by the British Plastics Federation.
along with applicable references herein as a source for We thank Cindy Helfrich for typing the many drafts of
design improvement. To this end, the authors have syn- this manual and illustrator Sandrajohnson for the cover
thesized their work and reviewed survey techniques that drawing. We thank the National Marine Fisheries Service
have been used in the past for assessing marine debris, Marine Entanglement Research Program for the use of
such as sighting surveys, beach surveys, and trawl surveys, their slides of marine debris, which served as inspiration
and have considered new methods (e.g., aerial photogra- for the drawing.
phy). All techniques have been put into a general survey The senior author's salary was provided by the U.S.
planning framework to assist in developing different ma- Environmental Protection Agency. The manual has been
rine debris surveys. subjected to the Agency's peer and administrative review,
We thank the following people who discussed the dif- and has been approved for publication.
ferent survey methodologies with us: Jay Brueggeman, This manual was the result of discussions held at the
Ray Highsmith, Scott Johnson, Barry Troutman, and Sixth Session of the Intergovernmental Oceanographic
Heather Trulli. We thank David Redford for allowing us Commission's Working Group on the Global Investiga-
to use EPA SOP No. 4-35. We thank the Center for Ma- tion of Pollution in the Marine Environment (25 Sep-
rine Conservation, the Environmental Protection-Agency, tember-I October 1986). The Marine Mammal Commis-
the Tidy Britain Group, Scott Johnson, Linda Jones, Jef- sion recommended that the National Marine Fisheries
frey June and Heather Trulli for the use of their data Service Marine Entanglement Research Program take on
forms. the effort of producing a procedures manual and drafted
We thank the following people for reviewing various the original scope of work for this project.
drafts and parts of this manual: James Coe, E. David Major funding was provided by the National Marine
Ford, Donald Gunderson, James Herkelrath, Scott John- Fisheries Service Marine Entanglement Research Program
son, Linda Jones, Jeffrey June, Theodore Merrell, Tho- to the University of Washington (Contract No. 52ABNF-
mas W. Miller, David Redford, David Rugg, and Heather 0-00071).
Trulli.
We thank the following people for reviewing the entire Christine A- Ribic, Research Ecologist
manual: Marcia Bollman (editorial review), Deborah Corvallis, August 31, 1991
Coffey (quality assur-ance review), David Laist, Tony Olsen,

How to Use This Man'ual

We do not expect the user to read this manual from scope of this manual and discussion of the general ar-
cover to cover. We expect specific chapters will be of rangement of this manual are explained in Chapter 1.
more interest to the user than others..Thus, we have Therefore, we reco'mmend that the user read Chapter 1 before
made the chapters self-contained. However, the general reading any other chapark.

Chapter I

Methodologv

In the late 1960s and 1970s, trash and other debris flects the total amount of debris entering over a
of human origin in the ocean began to be recognized longer time period. Second, synthetic materials often
as a widespread problem (Risebrough 1969; are less expensive than the natural fibers they re-
Heyerdahl 1971; Colton 1974; NAS 1975). Various place, thereby decreasing incentives to reuse or
solid materials of human origin were becoming in- recycle items. Third, and most obvious, there are sim-
creasingly apparent both on beaches and floating at ply more ships and coastal residents that can lose or
sea. The debris then, as now, typically included der- discard materials.
elict fishing gear, plastic bags or sheeting, paper The impacts from marine debris include
products, strapping bands, rope, line, cans, bottles, 0 aesthetic degradation (Heyerdahl 1971; NAS
balloons, plastic pellets, wood planks, clothing, light 1975);
bulbs, rubber tubes, and gloves. THese items were 0 hazards to wildlife (Laist 1987; Bourne 1990; Ryan
either discarded or lost at sea or carried to the ocean 1990a; Sileo 1990);
from land by rivers, domestic and industrial outfalls, 0 economic losses (e.g., damage to boats and fish-
shoreline runoff, offshore winds, or other means of ing gear and decreased tourism; Heneman and
transport. Although scattered records of seats and the CEE 1988); and
other marine life entangled in such debris have been 0 human health hazards (e.g., physical injury to
reported for at least several decades before 1970 (see bathers, exposure to chemical packaging and pos-
for example Scheffer 1950), such occurrences were sible spread of contagious disease [Dixon and
considered isolated events and the growing amount Dixon 1981a, 1986; High 1985; Wallace 1985;
of debris was characterized as a litter problem (NAS
Prute@ 1.987a]).
1975).
Since the mid-1980s, however, many articles, pa- Because of the visual aspect of beach litter, beach
pers, and reports documenting marine debris and its cleanups by volunteers have been organized to both
effects have appeared in the popular, scientific, and educate the public about the extent of the problem
technical literature (Duerr 1980; Horsman 1982, and to help mitigate aesthetic effects (Neilson 1985,
1985; Bourne 1983; Wehle and Coleman 1983; Dixon 1987, HMEPA 1991). In the United States,
Coleman and Wehle 1984; Dean 1985; Gosliner 1985; state-wide cleanups are now coordinated by the Cen-
Shomura and Yoshida 1985; Wallace 1985; Carr 1986, ter for Marine Conservation (O'Hara 1989, CMC
1987; CEE 1986, 1987a; 1988, Clark 1986; Coe 1986; 1991). Hazards to wildlife have been detailed in
FoN@_" and Merrell 1986; Azzarello and Van Vleet many studies. Entanglement- in discarded net frag-
1987; Bean 1987; FACI 1989; Laist 1987; Lentz 1987; ments has been of primary concern for impacts to
Pruter 1987, a and b; Wilber 1987; Wolfe 1987; marine mammals (Table 1). Ingestion of debris has
Augerot 1988; Gramentz 1988; Heneman and the been reported most frequently for sea birds although
CEE 1988; MPDTF 1988; O'Hara et al. 1988; ingestion by sea turtles, economically important fish,
Cawthorn 1989; Croxall 1990; Klemm and Wendt and cetaceans (Walker and Coe 1990) is of growing
1990; Parker 1990). This new information describes concern (Table 1). Impacts on the population level
problems that are far more widespread and signifi- have been difficult to document (Laist 1987; Pruter
cant than previously recognized, and it established 1987a; Ryan 1987a; 1988a; Ryan and Jackson 1987;
marine debris as another major form of ocean pollu- Ryan et al. 1988). The most frequently cited and con-
tion. troversial (e.g., Scordino 1985) case is the decline of
The increase in amounts of marine debris over the the northern fur seal (CalloAinus ursinus) population
past several decades can be attributed to at least because of entanglement in discarded fishing nets
three factors (MMC 1991). First, synthetic materials (e.g., Fowler 1987; Fowler et al. 1990). Economic
replaced natural fibers in the manufacture of more losses (Meade et al. 1990; Takehama 1990) and pub-
and more everyday items. Because these materials lic health problems (Dixon 1981, 1987; Dixon and
tend to degrade more slowly in seawater, the total Dixon 1981b, 1986; Wagner 1990) have been less
amount of debris in the ocean at any given time re- publicized.

2 NOAA Technical Report NWS 108: Marine Debris Survey Manual

Table I
Impacts of marine debris on marine animaIs.

Impact Animal/Taxon Reference

Entanglement Monk seal Andre and Itmer (1980)
(Monachus schauinslandi) Balazs (1979)
Henderson (1984; 1985; 1990)

Northern fur seal Bigg (1979)
(Callorhinus ursinus) Fowler (1982; 1984; 1985; 1987; 1988)
--------F6w]er and Ragen (1990)
Fowler et al. (1989)
Sanger(1974)
Scheffer (1950)
Scordino (1985)
Scordino and Fisher (1983)
Scordino et al. (1984; 1988)

Other marine mammals Bonner and McCann (1982)
Calkins (1985)
Cawthorn (1985)
Croxall et al. (1990)
Jones and Ferrero (1985)
Ryan (1990b)
Shaughnessy (1980)
Stewart and Yochem (1985; 1987; 1990)

Seabirds Conant(1984)
Dean (1985)
DeGrange and Newby (1980)
Ryan (1990b)
Schrey and Vauk (1987)

Turtles Balazs (1985)

Ingestion Seabirds Ainley et al. (1990, a and b)
Baltz and Morejohn (1976)
Bayer and Olson (1988)
Bond (1971)
Bourne and Imber (1982)
Connors and Smith (1982)
Day (1980)
Day et al. (1985)
Dickerman and Goelet (1987)
Fry et al. (1987)
Furness (1983; 1985, a and b)
Harper and Fowler (1987)
Hays and Cormons (1974)
Kenyon and Kridler (1969)
Ogi (1990)
Parslow andjefferies (1972)
Pettit et al. (1981)
Randall et al. (1983)
Rothstein (1973)
Ryan (1985; 1986; 1987, a and b; 1988c; 1990b)
Sileo et al. (1990)
Slip et al. (1990)
van Franeker (1983; 1985)
van Franeker and Bell (1988)
Zonfrillo (1985)

CHAPTER 1: Methodolog 3

Table I (continued)

Impact Animal/Taxon Reference

Fishes Anonymous (1981)
Carpenter et al. (1972)
Hjelmeland et al. (1988)
Hoss and Settle (1990)
Kartar et al. (1973, 1976)
Kubota (1990)
Ryan (1990b)

Turtles Balazs (1985)
Bourne (1985)
Carr (1987)
Cawthorn (1985)
Duronslet et al. (1991)
Lutz (1990)
Plotkin and Amos (1990)
Ryan (1990b)
Sadove and Morreale (1990)

Marine mammals Ryan (1990b)
Sadove and Morreale (1990)
Walker and Coe (1990)

In response to growing concern over marine de- groups as well as the public at large. The programs
bris, actions have been taken by governments advise the groups about debris-related problems, new
nationally as well as internationally to reduce dis- regulatory requirements, and proper garbage han-
charges at their source (Bean 1984). For example, dling and disposal practices.
intentional at-sea dumping of garbage generated on As these efforts were implemented, it became ap-
land became subject to international control as of parent that monitoring studies would be needed to
1972 through the Convention on the Prevention of assess the effectiveness of actions in reducing the
Marine Pollution by Dumping of Wastes and Other overall amount of marine debris as well as certain
Matter (commonly called the London Dumping Con- types of debris of particular concern, such as plastics
vention). Similarly, at-sea disposal of garbage generally, medical wastes, and fishing gear (MMC
generated during the routine operation of ships 1987). To address marine pollution monitoring
(e.g., garbage not deliberately carried to sea for the needs generally, the Intergovernmental Oceano-
purpose of disposal) was addressed through a 1978 graphic Commission (IOC), part of UNESCO,
Protocol to the 1973 International Convention for initiated a program for the Global Investigation of
the Prevention of Pollution by Ships (commonly Pollution in the Marine Environment (GIPME) in
called the MARPOL Convention). Specifically, the 1976 (Andersen et al. 1986). The GIPME program is
1978 Protocol to the MARPOL Convention added guided by a comprehensive plan consisting of four
five Annexes, each dealing with a different form of stages (Kullenberg 1986): a mass balance determina-
pollution from ships. Of these, Annex V established tion (including baseline measurements); a
regulations on discharging ship-generated garbage, contamination assessment (including an evaluation
including a prohibition of discharging all plastics at of the distribution, movement, and trends in levels of
sea.. pollutants in the marine environment); a pollution
National efforts to implement programs consistent assessment (involving an evaluation of a pollutant's
with these conventions may go beyond the specific effecton marine life); and regulatory action.
measures required by the international regimes. For To provide direction for its GIPME program, the
example, in the United States, substantial education IOC established a Working Group to oversee interna-
efforts have been mounted through the Marine En- tional efforts. Among other things, the GIPME
tanglement Research Program of the National Working Group 1) develops manuals on procedures
Marine Fisheries Service (NMFS). These programs for collecting, recording, and archiving data on spe-
are directed toward mariners, beach users, and other cific marine pollutants; 2) supports training exercises

4 NOAA Technical Report NMFS 108: Marine Debris Survey Manual

in the use of those procedures; and 3) conducts in- rectly assess debris in the open ocean, usually on ves-
ter-calibration exercises to ensure that data collected sels of opportunity, have increased (e.g., Gregory et
by one country or group can be statistically compa- al. 1984; Dahlberg and Day 1985; Jones and Ferrero
rable with those collected by others. As an example 1985; Yoshida and Baba 1985, a and b; Ignell and
of its efforts, the Working Group adopted a manual Dahlberg 1986; Mio and Takebama 1988; Yagi and
describing a standard methodology for monitoring Nomura 1988; Nasu and Hiramatsu 1990). In most
tar balls and dissolved oil in seawater and on beaches cases, the individual monitoring studies have had dif-
(IOC 1984) and subsequently assisted efforts to test ferent ob ectives and different sampling designs, thus
procedures in the manual in the-Caribbean Sea area making comparisons and broad assessments question-
(Corredor et al. 1987). The success of this approach able (Ribic and Bledsoe 1986). A case illustrating
is evident from publications that have successfully problems with non-standardized methodology is that
documented and described the extent of tar pollu- of assessing roadside litter in the U.S.; areas were not
tion in the Caribbean (Atwood et al. 1987, a through comparable because different survey techniques were
c). used (Marquez and Zandi 1985).
Prior to 1986, the GIPME Working Group had not Therefore, the purpose of this manual is to review
addressed monitoring needs for marine debris. the sampling designs used to measure marine debris
Therefore, at its Sixth Session in Paris, France, on 25 and to put them into a framework useful to others in
September-1 October 1986, it reviewed debris-re- planning and executing surveys to assess the types,
lated information and agreed that a procedures amount, distribution, and movement of marine de-
manual for monitoring debris on beaches and at sea bris in open water and on beaches. By doing so, we
warranted consideration. To assist in developing the hope that scientists, resource managers, and others
manual and to help in encouraging and guiding ma- who can collect useful data will be encouraged to do
rine debris monitoring efforts in the U.S., the Marine so in a manner that will be useful and help contrib-
Mammal Commission recommended to the NMFS ute to a broader understanding of the status and
that a manual be developed as part of the Marine trends of marine debris pollution.
Entanglement Research Program to encourage OP7 This manual is divided into chapters according to
portunistic monitoring efforts in the U.S. and that type and location of debris survey. Chapter 2 deals
this manual be provided to the GIPME Working with shipboard sighting surveys for larger debris in
Group for consideration in its series of pollution open water, Chapter 3 addresses shipboard surface
manuals (MMC 1987). sampling for smaller debris in open water, Chapter 4
Following some preliminary work on the manual, reviews beach surveys, Chapter 5 addresses benthic
the matter was examined further at the Second Inter- surveys for larger debris in open water, and Chapter
national Conference on Marine Debris held in 6 deals with the experimental approach of aerial sur-
Honolulu, Hawaii, on 2-7 April 1989, at which time a veys. Even though the manual is divided into separate
Conference Working Group to Assess the Amount chapters, investigators may use two or more survey
and Types of Marine Debris (hereafter referred to as methodologies from the separate chapters to esti-
the Assessment Working Group) was formed (Ribic mate the magnitude of the marine debris problem in
1990). Its participants agreed that work on a proce- a particular area. The rest of Chapter I discusses ter-
dures manual should proceed as a matter of high minology and categories of marine debris, the
priority and that the work should focus on describing importance of defining objectives prior to starting a
study methodologies to meet the first two stages of survey, and general monitoring guidelines. A glossary
the above-mentioned GIPME program plan for pollu- is provided at the end of the manual.
tion monitoring (i.e., baseline studies and
contaminant assessment) (Ribic 1990).
In this regard, the Assessment Working Group Definition and Categories of
noted that two basic approaches have been used to Marine Debris
assess marine debris: open-water surveys (including
both visual sighting surveys, surface trawls, and As note previously, the marine debris problem was
benthic trawls); and beach surveys. Initially, studies initially characterized as a marine litter problem. The
most often involved beach surveys (e.g., Gregory National Academy of Sciences (1975) defined marine
1977; 1978, a and b, 1987; Dixon and Dixon 1980, litter as solid materials of human origiri that are dis-
1981b, 1983; Merrell 1984, 1985; Henderson and carded at sea or reach the sea through waterways or
Pillos 1985), which Dixon and Dixon (1981a) sug- domestic and industrial outfalls. While the definition
gested were the most cost-effective monitoring is broad and remains useful, we prefer the term "ma-
strategy for debris. However, recent attempts to di- rine debris" because it does not suggest impacts are

CHAPTER 1: Methodology 5
primarily aesthetic. The Academy's definition prop- medium debris @!2.5 cm and 5;10 cm, e.g.,
erly distinguished between sources of debris that styrofoam cups, tampon applicators
originated at sea and those that originated on land. large debris >10 cm and !@l in, e.g.,
In this manual, the term "vessel-source" debris will bleach bottles, gillnet floats
refer to material of human origin discarded in open
water. MARPOL Annex V established regulations gov- verylarge debris >1 in, e.g., derelict fishing
erning the discharge of garbage during the normal net
operations of ships, including a prohibition of at-sea The boundary of 2.5 cm can be justifie .d because
disposal of any plastics (Augerot 1988). An example MARPOL Annex V regulations Istate that material re-
of vessel-source debris is fishing-related debris such leased from ships will be ground to <2.5 cm. The
as trawl net fra Fments. "Landbased" debris, in this distinction among the other categ .ories, while more
manual, will refer to material of human origin that arbitrary, is based on sizes of the major debris items
reaches the sea through waterways or domestic and often found on beaches.
industrial outfalls. Included in "landbased" debris is There are many categories used for medium to
litter left by beach users, material lost from coastal very large debris by studies done in op-en water
landfill sites, and items such as tampon applicators (Table 2). Researchers usually identified individual
discharged through sewage outfalls. This distinction items to produce @an exhaustive list and then grouped
is important because MARPOL Annex V addresses the debris into major categories. The general group-
the problem of vessel-source debris only, whereas the ings used have been similar. The debris items are
Clean Water Act addresses landbased debris (e.g., usually organized -according to what the items are
ocean dumping of landbased garbage and combined made from (e.g., paper, rubber, plastic, wood, glass,
sewer overflow systems) (U.S. EPA 1990b). Any ma- metal), the manufacturer's intended use (e.g., fish-
rine debris sampling scheme, especially surveys done ing gear, ropes, bottles), or a combination of the two.
on land, must recognize that sampled material may The major categories have typically been fishing gear,
originate from both sources. plastic, styrofoam, glass, wood, metal, paper, and mis-
Marine debris may be classified based on size. This cellaneous. Fishing gear was usually subdivided into
type of distinction is important because size will influ- nets and other gear (including plastic floats). Some-
ence the way debris is dispersed and deposited, the times plastic and styrofoam were put into one
wildlife impacts that may occur, and what type of sur- category. Wood was mostly divided into natural (e.g.,
vey approaches may be practical. The Assessment logs) and man-made (e.g., boxes). In some studies,
Working Group (Ribic 1990) proposed the following cloth, cardboard, and rubber were separated into ma-
debris categorization by size: jor categories. Most of the open-water studies did not
state that particular debris items were of interest,
mega-debris-debris >2-3 cm though some of the categories used tended to reflect
macro-debris-5 mm to 2-3 cm that interest (e.g., the emphasis on fishing gear in
meso-debris-<5 min the studies of Mio and Takehama 1988 and Nasu and
micro-debris-powdered Hiramatsu 1990; [Table 2]). Beach debris surveys
tended to use categories that reflected specific study
"Macro-debris," "megalitter" (McCoy 1988; Gre- objectives (Table 3). For example, because Merrell
gory 1990), and "large plastic" (Day and Shaw 1987) (1985) emphasized entangling debris, his list reflects
are terms used to describe marine debris visible to that interest (Table 3). Willoughby (1986) empha-
the eye or with binoculars during the course of vessel sized man-made materials with long degradation
sighting surveys and beach surveys. The lower size times; thus, his list did not include paper or card-
limit of this type of debris varied, ranging from <0.5 board (Table 3). One of the most general lists is that
cm (McCoy 1988) to 1.5 cm (Morris 1980a), and up used by the Center for Marine Conservation (CMC)
to 2.5 cm in length (Dahlberg and Day 1985; Day and (formerly the Center for Environmental Education),
Shaw 1987). McCoy (1988) used 7 X 50 binoculars to with 59 individual items (CEE 1988 [Table 31; form
make observations on a stationary ship in calm seas. can be found in Appendix 4). The CMC's list paral-
Alternately, Day and Shaw (1987) used IOX binocu- lels the major categories used in the open-water
lars on a moving ship in variable seas. In this manual, studies with one exception: fishing gear is put into
the following size categories for debris are used. the plastic category. Classifying fishing gear can be
problematical because the category is one of func-
small debris <2.5 cm (not visible by eye tion rather than material from which the object is
in water), e.g., polystyrene pel- made. The U.S. Environmental Protection Agency
lets, fragmented plastic (1990b) categorized fishing gear by material, so net-

6 NOAA Technical Report NMIFS 108: Marine Debris Survey Manual

Table 2
Debris categories used by open-water sighting surveys for medium to very large debris (pieces >2.5 cm).

Reference Categories Comments

Venrick et a]. (19731 plastic bottles List of items found; categories not
plastic fragments set up in advance
glass fishing floats
glass bottles
rope
balloon
finished wood
shoebrush
rubbersandal
paper items
coffee can

Morris (1980a) plastic bags List of items found; categories set up in
cups advance
sheets
packing material
bottles
fragments
timber
rubber
nylon rope
feathers
glass bottles
paper items

Dixon (T.J.) and Dixon (1983) man-made wood items
paper
cardboard
nylon rope
netting
plastics and styrofoam
metal
glass

Dahlberg and Day (1985) plastic 48 individual items listed
styrofoam
metal
glass
paper
cloth
wood

Mio and Takehama (1988) net gear
plastic bands
other fishing gear
styrofoam
other plastic articles
pieces of wood/drifting logs
seaweed
other

Yagi and Nomura (1988) styrofoam
buoys
plastic sheets/bags
fishing net fragments
rope
wood
glass
metal
other

CHAPTER 1: Methodolqy 7

Table 2 (continued)

Reference Categories Comments

McCoy (1988) plastic
wood
other (paper, cloth, or unidentified)

Nasu and Hiramatsu (1990) artificial objects:
fishing gear
net
fishing gear other than nets
other than fishing gear
pieces of wood
petrochemical products
styrofoam
glass and metal products

natural objects:
seaweed
logs

other:
unknown

Day et a]. (1990a) glass
metal
paper/fiber
rubber
wood
plastic

ting is found under plastic while buoys are found un- categories is not possible owing to specific study ob-
der the polystyrene category. The National Park jectives and debris unique to the particular area. An
Service (Cole et al. 1990) used the category "Plastic example of this was the expansion of the CMC form
Fishing Gear," which is a combined material and from 59 to 200 items by the harbor studies program
function category. (U.S. EPA 1990a).
Most categories are self-explanatory and require For small debris, all the studies in Table 4 occurred
no specific knowledge for use (e.g., rope, styrofoam in open water and used similar categories. The items
food containers); however, some others such as buoy identified by the Assessment Working Group (Ribic
bags (Merrell 1985) and crustacean pot floats (Cole 1990) as being most important to record are re-
et al. 1990) may require more definition. Few studies flected in the studies in Table 4.
conclusively define each debris category. The most Primary categories by material can be developed
detailed list of definitions was found in Appendix B similar to medium-size and larger debris categories.
of Cole et al. (1990), where over 50 debris categories For small debris, color and size may become impor-
were listed; this appendix was used to define debris tant in determining the likelihood that different
categories in the glossary of this manual. species may detect and ingest debris. Additional sub-
For medium and large debris categories, we recom- categories based on size and color would be useful
mend that researchers organize lists of items first by for small debris. Size categories could be based on
material type (i.e., plastic including foamed plastics the most common size for pellets. Because most pel-
[styrofoam], glass, metal, rubber, fiber [cloth], wood, let sizes fall between 1 mm and 6 mm (Carpenter and
paper) and then, under these primary categories, de- Smith 1972; Carpenter et al. 1972; Gregory 1977,
velop secondary categories according to function or 1978a, 1983, 1990), three categories-<I mm, 1-6
manufactured product (e.g., fishing gear, bottles, mm, and >6 mm-can be used. Color categories
medical use). All individual items would be listed un- could be based on Day et al. (1985, 1990b), who used
der one of these primary or secondary categories, or 11 colors (transparent, red/pink, blue, yellow, white,
both. We recognize that complete standardization of tan, green, brown, black/gray, orange, miscellaneous).

8 NOAA Technical Report NMFS 108: Marine Debris Survey Manual

Table 3
Debris categories used by beach surveys for medium to very large debris (pieces >2.5 cm).

Reference Categories Comments

Dixon and Cooke (1977) plastic Counting containers; 21 individual items
glass listed
metal

paper

Merrell (1985) trawl web
straps-open
straps-closed
trawl floats
synthetic line
bait containers
gillnet floats
bottles
caps and lids
fragments-hard
fragments-soft
buoy bags
six-pack yokes
other

Willoughby (1986) plastic bags Only man-made materials with long
footwear degradation times counted
styrofoam
bottles
tins
ropes and netting
lamp bulbs

Vauk and Schrey (1987) plastic Individual items enumerated; then put into
paper categories
metal
glass
fishing gear
clothing
foodstuff
wood

Center for Environmental plastic 59 subcategories
Education (1988) glass
styrofoam
metal

paper
man-made wood
rubber

Cole et al. (1990) plastic fishing gear 51 items listed
plastic packaging material
personal effects
miscellaneous plastics

General Monitoring Guidelines line studies for marine debris are generally car-
ried out over large geographical areas with low sam-
The Assessment Working Group (Ribic 1990) divided pling frequency. Assessment studies, however, were
marine debris studies into baseline and assessment considered to be more focused in their objectives.
categories. By definition, baseline studies were de- Two examples of assessment studies are measuring
signed to determine the characteristics of the debris density of debris in certain areas and measuring
problem (e.g., what type of material is found). Base- changes over time (i.e., trend monitoring). Assess-

CHAPTER 1: Mediodology 9

Table 4
Debris categories used by open-water sampling studies for small debris (pieces <2.5 cm).

Reference Categories Comments

Colton et al. white opaque polystyrene spherules
(1974) translucent to clear polystyrene spherules containing
gaseous voids
opaque to translucent polyethylene cylinders or discs
styrofoam
sheets of thin, flexible, wrapping material
pieces of plastic

Day et aL (1985) fragments 10 color categories used
monofilament line fragments
pellets
polypropylene line fragments
styrofoam
miscellaneous/unidentified

Ryan (1988b) industrial pellets 9 color categories used (color when
pieces of manufactured items wet); 3 categories of wear used
fibers
styrofoam/other foamed plastics

Day et al. (1990b) pellets I I color categories used
fragments
styrofoam
polypropylene line fragments
miscellaneous line/threads

ment studies tend to be more limited in geographical 3 Collect information on the geographic areas
area and have more intense sampling efforts. This of interest to develop the sampling plan (e.g.,
split in the types of studies reflects the ideas of physical features, weather patterns, historical
Barnard et al. (1985), who divided monitoring stud- information).
ies into two types: descriptive monitoring and 4 Define the field measurement to be made.
location-specific monitoring. Other researchers (Gil- 5 Examine data from previous studies or conduct
bert 1987) do not distinguish between the two types pilot studies to approximate the likely variability
of studies and refer to the baseline and assessment in the field measurements.
studies collectively as monitoring studies in contrast 6 Develop a quality assurance program plan to en-
to research studies (e.g., research studies to deter- sure that the data collected will be of high
mine the transport of pollutants through the quality, verifiable, and defensible.
environment). In this manual, we use the framework 7 Develop field sampling designs and measurement
in Table 5, which presents the survey types relevant procedures that will yield representative data
to marine debris surveys. In terms of the base- from the defined population, along with a speci-
line and assessment study definitions made by fied variance or confidence limit. If necessary,
the Assessment Working Group (Ribic 1990), surveys make decisions on identifying the source of de-
with objectives 1-4 (Table 5) are assessment bris (vessel-source versus landbased).
studies, whereas studies with objective 5 (Table 5) 8 Determine the statistical analyses to be used.
are baseline. 9 Conduct the study according to the written pro-
Whether or not the studies are baseline or assess- tocol.
ment, good planning is essential to collect useful 10 Analyze the data.
information. Advice in this manual is organized by 11 Evaluate the study (e.g., were the objectives
the planning guidelines of Gilbert (1987): met? Were the collected data adequate to meet
I State the objectives clearly. the stated objectives? Should the design be
2 Define the population of interest. modified?).

10 NOAA Technical Report NMB.. 108: Marine Debris Survey Manual

Table 5
Marine debris study objectives adapted from Lettenmaier et al. (1982).

Objective Description

(1) Surveillance Detect impact of known pollution source or detection of point source pollution (e.g., spills).
Need frequent sampling, possibly continuous monitoring.

Example:
Cargo spill (Dixon and Dixon 1981b)
Polystyrene spherules (Kartar et al. 1973, 1976)

(2) Model parameterization Provide data on time and space scales to allow identification of input decay rates; linked to
prediction. Sampling is frequent and intense in a small area for a short time period.

Example:
Population dynamics of marine debris (Gerrodette 1985)

(3) Cause-effect Identify functions relating input and output; assess presentconditions. Sampling is frequent and
intense in a small area for a short time period.

Example:
Turnover rates for debris on beaches (Dixon and Cooke 1977; Johnson 1988, 1990a)

(4) Trend detection Analyze time series for evidence of changes in a statistical sense. Long sequences of observations
at frequencies on the order of weekly or monthly at stations scattered throughout a large area.

Example:
Assessment of changes of trawl web over time (Ribic andjohnson 1990)
Assessment of changes in total debris over time (Ribic 1991)
Assessment of changes in total plastic over time (Johnson 1990a)

(5) Baseline For little or no pre-existing data; establish the level of the problem. Low sampling frequencies.
Spatial density depends on problem (local problem will result irl more concentrated stations;
problems covering a large area can result in dispersed stations).

Example:
Nationwide beach surveys (O'Hara 1989, 1990)

Clearly defining and stating the objectives cannot Texas beaches in the fall or net fragments in the
be overemphasized. The objectives can be as simple North Pacific during the fishing season. It may be
as determining the kinds of debris occurring on a that only certain areas are available for study. For
particular beach to as complex as determining a de- example, only beaches with good public access or
crease in the amount of entangling debris seen on ocean areas visited regularly by vessels of opportunity
Alaska beaches as a consequence of NLAPPOL Annex may be available for study. Then the population of
V. Explicitly stated objectives guide the development interest (e.g., debris on all beaches) is not available,
of the sampling design from defining the population and only a subset (e.g., debris on public beaches) is
of interest to what data are collected and how they available. Given such restrictions, the target popula-
are analyzed. Stating objectives also sets the scope of tion (population of interest minus the restricted
the study and data analyses, which is important given areas) may not be truly representative of the larger
that no single sampling design can answer all ques- population of interest, unless certain assumptions
tions of interest, as is evident from Table 5. can be made and tested (e.g., debris composition on
The population of interest in marine debris studies private beaches is not different from that on public
is often associated with a geographical area. Ex- beaches). Deciding on the target population is im-
amples of populations of interest include net portant because that population will be used to
fragments in the North Pacific Ocean, domestic waste define the sampling frame from which representative
in the North Sea and tar in the Caribbean. Medium- units for measurement are chosen. A "representative
or larger-sized debris on Texas beaches can also be a unit" is a unit selected from the sampling frame in
population of interest. Populations of interest can such a way that it, in combination with the other rep-
have a temporal component, such as large debris on resentative units, will give an accurate picture of the

CHAPTER I., Methodology I I
phenomenon being studied (Gilbert 1987). Another researchers to help design their surveys. In addition,
word for representative unit is "sampling unit" (the if many people are involved in a study that is con-
term used in this manual). Typical sampling units for ducted over a long time period, the quality assurance
marine debris studies are beaches or transects. on program plan provides a unified set of directions to
beaches, areas of ocean scanned, area swept by sur- follow. Quality assurance program plans address mea-
face sampler, and area swept by demersal trawl. sures that will describe data quality. They should
Information on the physical environment, weather contain a training element such that the accuracy
patterns, and site history may be useful in planning of data collected by volunteers can be assured. On
the sampling design. For example, wind direction a more personal level, the plan should incor-
plays a large part in the deposition of debris on porate procedures that maintain a safe working
beaches in the United Kingdom (Dixon and Dixon environment.
1981a). Therefore, in this situation, information re- Developing field sampling designs and measure-
garding offshore and onshore winds is important for ment procedures that give representative data
determining sampling times. For open-water studies, requires a statistical assessment. Sample size require-
information on curr -ents or areas where debris are ments are determined, a procedure for choosing
concentrated can be used to decide where the sam- sampling units is developed, and precision and statis-
pling should be concentrated (i.e., stratification tical power are. discussed. Developing a sampling
variables). In addition, surface drift experiments may design is the most problematical task for a study.
be used to identify sampling conditions in relation to There may be many sampling designs to choose
oceanographic conditions (e.g., onshore currents). from, but each design is likely to have different ef-
This aspect is considered further in the individual fects on results and require different effort for the
chapters. survey. Field sampling designs will be addressed in
Defining the field measurements is important and the individual chapters.
is discussed in more detail in the individual chapters. Certain statistical analyses are appropriate for most
Issues that must be resolved 'are defining what will be studies of marine debris. These are called descriptive
measured, what the sampling unit is, how the mea- statistics (e.g., means, variances, plots of data) and
surements will be taken on the sampling unit, and exploratory data analyses (e.g., box plots). All are
what field methodology will be used. These are criti- available in a wide variety of computer packages. The
cal elements of geographically-based surveys such as more difficult decisions are made when there are sta-
those for marine debris. The sampling unit may be a tistical hypotheses to be tested. How to test for
fixed area of ocean with the measurement being changes in trend over time or changes in mean level
counts of all items >2.5 cm. The field methodology over time are questions that can be addressed in a
may be a strip transect. All of these issues should be variety of ways. In general, there are parametric and
resolved before the start of the field work and noted nonparametric statistical methods. Both approaches
in the quality assurance program plan discussed have strengths and weaknesses. The scope of this
below. manual precludes discussion of these in detail. Stan-
Pilot studies play a critical role in survey design, dard statistical texts (e.g., Conover 1980; Sokal and
Particularly for large-scale or long-term surveys (or Rohlf 1981) should be consulted; Gilbert (1987) pre-
both). Pilot studies play a key role in training for sents a discussion on detecting changes in trend.
field measurement techniques, preliminary assess- Evaluating the study design is critical. Sometimes
ment of debris sources (which may change the changes in design are made owing to problems in the
design), and assessment of debris variability in the field as they arise. In addition, as more is learned
sampling units over the geographical area of interest. about the problem of interest, objectives and study
Pilot studies also are invaluable for determining cost designs change. For example, the beach surveys in
and effort to complete the survey. Then, study objec- Alaska started by Merrell (1985) have changed from
tives can be modified or new objectives stated a general survey of trawl webbing on beaches to focus
because of sampling constraints found during the on detectable changes that may be due to MARPOL
pilot study (e.g., available resources, type of equip- Annex V. The effect of this particular change is dis-
ment on hand). cussed in more detail in Chapter 4.
Quality assurance program plans have not been de- Individual chapters follow the I I guidelines of Gil-
scribed for many published surveys, and we are bert (1987) with an emphasis on guidelines 1, 2, 4,
uncertain whether many studies have had one. and 7. The other guidelines are discussed when ap-
Quality assurance activities help to make -studies re- propriate. In all cases, investigators are encouraged
peatable. Detailed descriptions of what was done to discuss design and data analysis procedures with a
(e.g., standardized procedures) can be used by other statistician.

Oiapter 2

Shipboard Sighting Surveys
for Large Debris Items

General Description Population of Interest

This chapter discusses open-water sighting surveys In planned studies, debris in specific oceanic areas
whereby all floating debris are identified and comprises the population of interest, which must be
counted from an elevated platform on a moving ship. defined by the researcher. For example, fur seal re-
Transect Width may vary from .100 m to the visual searchers were interested in the amount of floating
horizon depending on the type of debris being stud- net debris around the Pribilof Islands, the breeding
ied. Surveys are usually conducted from the rookeries for northern fur seals. Therefore, floating
glare-free side of the ship, and objects are sighted net debris in a specific area around the islands dur-
visually either unaided or with binoculars. ing the fur seal breeding season was defined as the
Sighting surveys collect information on the distri- population of interest, and surveys were conducted
bution and amounts of floating, medium to very in that area (Yoshida and Baba 1985b; Baba et al.
large debris in areas of the open water during spe- 1988; 1990). Alternatively, the population of interest
cific time periods. Baseline surveys have been done can be as large as all the debris in the North Pacific
in the North Pacific Ocean using vessels of opportu- Ocean (Mio and Takehama 1988; Mio et al. 1990;
nity (Dahlberg and Day 1985; Ignell 1985; Jones and Nasu and Hiramatsu 1990). Defining the population
Ferrero 1985; Ignell and Dahlberg 1986; Mio and of interest is determined by the objectives of the
Takeharna 1988; Day et al. 1990a; Mio et al. 1990; study.
Nasu and Hiramatsu 1990; Shaw 1990). Vessels of op- Restrictions to the population of interest are likely
portunity also have been used to survey areas of the when surveys rely on vessels of opportunity. In such
North Sea (Dixon (T.J.) and Dixon 1983) and the cases, areas of the ocean are surveyed not because of
Mediterranean Sea (Morris 1980a). Dedicated vessels any particular sampling plan but because that is the
have been used in assessment studies to detect tem- ship's destination. This restriction often is not stated
poral trends in the North Pacific along 137'E by authors, who may generalize the debris in the
longitude between 0' and WN latitude (Yagi and sampled area to an entire oceanic area without any
Nomura 1988). Other studies have resurveyed areas justification for doing so. We basically consider it in-
to look for temporal changes; for example, Day and appropriate to generalize results from vessels of
Shaw (1987) surveyed the Gulf of Alaska along 155' opportunity studies.
longitude in 1984 and 1985.

Field Measurement
Objectives and Purpose
The most common variables of interest for open-
Typical objectives for open-water sighting surveys are water sighting surveys are density (number/kml) and
as follows: types of medium to very large debris.
to identify types of floating marine debris;
to estimate densities of floating marine debris; Description
to identify areas of low or high concentrations of
floating marine debris relative to either oceano- Observer(s) on a moving ship stand on the flying
graphic features (e.g., currents, convergence bridge or other elevated section. Observer heights
zones) or man-made structures (e.g., offshore oil above the water line and speed of ship will vary ac-
platforms); cording to- the type of ship. Using the glare-free side
to relate floating debris to entanglement or other of the ship for observation, .observer(s) visually scan
effects on animals; and for objects floating on the ocean as the ship moves
to detect temporal and spatial chang es in the o Ic- through the area. Binoculars are generally not used
currence of marine debris.

14 NOAA Technical Report NMFS 108: Marine Debris Survey Manual

to sight objects; instead, they are used only to con- seen within that strip and any objects seen outside
firm the identity or to help estimate sizes of objects. the specified distance are not counted (e.g., object 1
The number of observers on a survey varies (1-10). in Figure 1 is counted; object 2 is not). Common
The Assessment Working Group (Ribic 1990) recom- strip widths are 50 m (Day and Shaw 1987; Day et al.
mended that a minimum of two observers be em- 1990a) and 100 m (Dixon (T. R.) and Dixon 1983).
plu)yed in any survey. Survey transects must be de- The actual strip width used will depend on the study
fined in terms of width and length: the survey width objectives. Other researchers have counted all debris
(i.e., the maximum distance from the ship's side in seen without regard to a specified strip width and
which debris will be censused) may vary, in part, de- then have truncated.the data at certain distances for
pending on the number of observers; length of analysis (e.g., 50 m, Dahlberg and Day 1985; 10 m,
transect is defined as the straight-line distance cov- Mio and Takehama 1988). Appendix 2A contains a
ered by the ship during an observation period. protocol by the Tidy Britain Group for using strip
Observation periods may be defined in terms of time transects to estimate debris density.
periods of constant sighting conditions, vessel speed, Line Transect-All objects are counted regardless of
and direction. Length is then calculated by recording the distance from ship, and the perpendicular dis-
beginning and ending location (latitude/longitude) tance from the object to the ship is measured
or by calculating the distance traveled using the (Fig. 2). Two other variables-the distance of the ob-
ship's speed at the start of the transect, or both (the ject to the ship at the time of first sighting and the
latter method assumes the ship is moving at a con- angle of observation (Fig. 2)-can be measured and
stant speed). If vessel's speed or course changes, then converted to a perpendicular distance. While these
new coordinates, speed, and time must be recorded. latter two variables have been recorded most fre-
quently (Dahlberg and Day 19,85; Mio and Takehama
1988; Nasu and Hiramatsu 1990), perpendicular dis-
Options tance (Mio et al. 1990) is preferred (Burnham et al.
1980). When the latter two variables are measured,
Strip Transect-In a strip transect, only objects the measurement errors inherent in both variables
within a specified distance from the side of the ship
are counted (Fig. 1). All objects are assumed to be

00
P,

el
P2
02
W r2
900 ................................
0
0 0
0 0
SHIP
0
SHIP
Figure I
Schematic diagram of a strip transect. W
specified strip width. The dark circle indi- Figure 2
@P,

cates the observer, and the strip is between 0 Schematic diagram of a line transect. Pi = perpendicular
and 90'. Object 1 is inside the strip and is distance of object i to ship, ri = distance from object i to
recorded while Object 2 is outside the strip observer at time of sighting, and Oi = angle between
and is not recorded (even if observed). object i and observer at sighting.

CHAPTER 2: Shipboard Sighting Surveysfor Large Debns IL-im 15

result in errors in perpendicular distances that are when observations should take place. For example,
difficult to correct. When perpendicular distance is Dixon (T.J.) and Dixon (1983) restricted observa-
measured, errors can be dealt with by using distance tions to sea states (based on a combination of wind,
classes (Burnham and Anderson 1984). For example, waves, and swell height) of 3 or less, whereas Shaw
instead of using all distances as recorded, distances (1990) used se 'a states of 4 or less, and Day et al.
can be grouped into distance classes (e.g., 0-5 m, (1990a) did not sample when high waves could have
5-10 m). Errors made in estimating an object 6 m affected "sightability."
from the ship, versus 7 m, are then unimportant be- Characteristics of Marine Debris-Various authors
cause the data are analyzed in terms of distance (Dahlberg and Day 1985; Jones and Ferrero 1985;
classes rather than individual distances. Mio and Takehama 1988) have noted that color,
It is beyond the scope of this manual to present a size, shape, and buoyancy of objects affect their
detailed discussion of the advantages and disadvan- sightability. Currently, no data are available with
tages of the line- versus strip-transect method. Both which the problem can be evaluated. When using a
have strengths and weaknesses (Burnham and Ander- strip. transect approach, trials could be undertaken
son 1984). The strip transect method requires that all with materials of known characteristics deliberately
objects be seen in the strip of width w, but distance or placed at different distances from a vessel or in vary-
angle measurements are unnecessary. Bias in the ing weather conditions to determine sighting
density estimate results from objects being missed probabilities.
within the strip, from objects along the strip perim- Vessel Variability@Ship's speed and observer's
eter mistakenly being included or excluded from the height above the water will affect marine debris
ship, from observer differences (e.g., experience dif- sightings (Mio and Takehama 1988). While the im-
ferences, fatigue), from the physical setting (e.g., portance of these variables has been noted, no data
weather, speed of travel), and from variability in the are available to determine optimal height of ship
objects (e.g., color, size, shape). speed for marine debris surveys. If all data are com-
The line transect method requires four assump- bined into one set, the sighting differences due to
tions: 1) objects on the line are detected with vessel variability add to the variability in the data.
certainty; 2) objects do not move in response to the Measurement Variability-Exact measurements are
observer before detection; 3) perpendicular distance critical for the data analysis stage (Burnham et al.
data are accurate; and 4) detections are indepen- 1980). Most studies have estimated angle and dis-
dent. For sighting marine debris, the first and third tance from the ship to the o@ject at first sighting with
assumptions are the two most likely to be violated. no indication that the accuracy of the data has been
The first assumption can be handled by having one checked. If the perpendicular distances are discov-
observer watch the center line. The third assumption ered to be inaccurate, the analysis can still proceed
will be difficult to fulfill given the problems of esti- by putting the distances into distance classes
mating distances at sea; however, training observers (Burnham and Anderson 1984). Distances can be
to estimate distances at sea would be an important measured with a range finder or binoculars with
part of the quality assurance program plan. Biases reticles. Distance classes (e.g., 0-10 m, 10-20 m, etc.)
due to observer differences, the physical setting, and can be set up prior to the study and used instead of
object variability are incorporated into the analysis of measured distances. Accuracy of distance classes will
the perpendicular distances. still be important because the boundaries between
Of the two methods, theoretical studies by Burn- classes must be identified.
ham et al. (1985) suggest that, in terms of increased
efficiency and lower bias, the line transect method
gives better density estimates than the strip transect Data Collection
method. If an observer at sea can accurately measure
angles and distances or accurately use distance class- Researchers should collect the following data.
es, the line transect method is preferred (Ribic 1990).
* date
0 time at start of transect
Variables to Consider * duration of transect (time elapsed)
e location (latitude/longitude) at start of transect
Weather-Avoid making observations when condi- o distance traveled during transect:
tions restrict visibility. For example, Yoshida and Baba ship's speed at start of transect (the transect
(1985b) made no observations when visibility fell be- should be stopped if the vessel changes speed)
low 200 m. Sea state has also been used to determine location at end of transect (latitude/longitude)

16 NOAA Technical Report NMFS 108: Marine Debris Survey Manual
esighting condition-visibility, wind speed and direc- number of transects made, effort, and area surveyed.
tion, cloud cover, sea state, direction of sun relative A transect is defined as the straight-line distance be-
to ship's course tween the starting and ending locations of the ship or
*a list of all items sighted by classification (see Chap- the straight-line distance traveled for the duration of
ter 1), location at sighting, and distance/angle the transect. based on the ship's speed at the start of
measurements (if line transect) the transect. In both cases, the ship's course must not
*comments change during the observation period.
Note that the first five items in the above list are Forms for collecting debris sighting data have not
necessary whether or not debris is sighted on a been standardized, and few researchers have pub-
transect. This information is used to calculate the total lished their data forms. Figure 3 is a suggested data

Observer Name(s) Starting location - lat/long
Vessel Name Ending location - lat/long

Date (Yr/Mo/Day) Time: Start transect
Sea state (Beaufort) Time: End transect

Visibility Ship's speed at start of transect

Weather Wind. Speed

Direction

Time Classification Object Perpendicular Size Comments
(plastic, wood, etc.) Distance (m)

Figure 3
Suggested data form for open-water mega-debris sightings.

CffAPTER 2. Shipboard Sighting Surveysfor Lage Debns Items 17
form,, and Appendix Figure B (this chapter) is an for room and board, the cost for a I-month cruise
example of a data form used by the Marine Mammal will be less than $1,000 for two people for the use of
Observer Program to collect marine debris informa- a ship compared to $30,000 to $200,000+ when char-
tion. Debris information is substituted for species tering a vessel. The trade-offs are in the great
information (L.L. Jones, NOAA, National Marine restriction of the population of interest and the
Mammal Laboratory, Seattle, WA, pers. commun. availability of such cruises. Additional expense can be
February 1991). saved if biologists or oceanographers (volunteers)
make debris sightings between their primary research
Material and Personnel tasks. For volunteer-collected data, quality assurance
of the collected data will be an important issue
The basic equipment needed for sighting debris is a that needs to be resolved before such a program is
pair of binoculars (e.g., 8X40, 1OX50) to identify implemented.
sighted objects. Typical costs for high quality bi- Required personnel includes at least two ob-
nocul .ars are $300-1,000 (1991 U.S. dollars). Perpen- servers experienced or trained in sighting objects
dicular distances can be measured with a. range floating at sea and who also have training or experi-
ence in the use of equipment to measure distances
finder or binoculars with reticles and compass and angles. Because experienced observers see more
($500-1,000). Clipboards, pencils, and data forms than inexperienced observers, an inexperienced ob-
will be necessary to record the data (approximately server should be paired, with an experienced
$200). A tape recorder may be useful for document- observer. Whenever possible, for different' cruises
ing and verifying observations. during the study, the same observers should. be in-
The major cost of a dedicated debris survey at sea volved in the field work to help control observer
is the ship. Ship costs are related to vessel size, equip- variability (Day and Shaw 1987; Dixon (T.R.) and
ment, and trip duration. For example, a vessel about Dixon 1983).
38 in (125 ft) long for high seas travel and that fulfills
the NOAA requirements for doing small cetacean
sighting work will probably cost in excess of $6,600/ QualityAssuranceProgram
day, not including fuel cost. Therefore, for a I-month
cruise, the ship alone would cost approximately Unfortunately, publications describing results of
$198,000 (excluding fuel costs). An observer's travel
expenses and salary are additional. The salary of the open-water sighting surveys have not described or ac-
observer needs to be a minimum of $3,000/mo to knowledged a quality assurance plan. Some details of
importance for ensuring quality data from open-wa-
compete with other observer programs (L.L. Jones,
ter sighting surveys are as follows:
NOAA, National Marine Mammal Laboratory, Se-
attle, WA, pem commun. February 1991). Other * What is the population of interest, and what is the
expenses might be necessary, such as extra equip- population actually available for sampling (i.e., re-
ment (e.g., a wrench and hook to retrieve nets). strictions to the sampling frame; Guideline 7,
Thus, for two observers, the 338 in (125 ft) ship, and Chapter 1)?
travel expenses (e.g., $1,000 per observer), a & What is the justification for coricluding that the tar-
I-month cruise could cost over $216,500. Larger get population represents the broader population
vessels (e.g., NOAA research vessels like the Oceanog- of interest?
rapher and the Miller Freeman), are much more 0 How is a transect defined (e.g.; strip or line
expensive (for estimates, write Director, Pacific Ma- transect, how wide, how long)? How is the transect
rine Center, NOAA, 1801 Fairview, Seattle, WA selected for sampling?
98102). Surveys near the coast will be able to use 9 How is the sighting survey to be carried out. (e.g.,
smaller ships, which may rent for as low as $1,000/ how many people, how is debris sighted and veri-
day (L.L. Jones, NOAA, National Marine Mammal fied [with or without binoculars], what side of the
Laboratory, Seattle, WA, pers. commun. February ship will be used [one or both], how high above
1991). the water will the observer be, what are the restric-
Given the high cost of a dedicated research cruise, tions on sighting conditions, how. fast will the vessel
researchers usually use research vessels scheduled for travel)?
other purposes. This "vessel of opportunity" research How experienced are the observers in sighting ob-
considerably lowers the cost (to $8,000 in the preced- jects at sea?
ing example) because the researcher will only have to e-Were the observers trained to estimate angles and
pay for travel, room, and board. Assuming $10/day distances at sea? How were they trained?

18 NOAA Technical Report NMFS 108: Marine Debris Survey Manual
Were devices or procedures used to aid in making guidelines can be stated. More published informa-
and verifying these measurements? tion is available regarding open-water distribution of
Of critical importance to the sighting survey is esti- marine debrisin the North Pacific than in any other
mating the distances (and angles, if necessary) from ocean body (Mio and Takehama 1988; Mio et al.
the ship to the debris object. Most ships have a com- 1990; Nasu and Hiramatsu 1990). There is probably
pass on the flying bridge that can be used to measure enough information available to establish survey
angles. Distance or distance classes can be measured lines for regular monitoring (as suggested by Nasu
by visual estimation, or with either a range-finder and Hiramatsu 1990). How one picks the areas or
(which can be as simple as a set of calipers or a card- survey lines depends on the objectives of the survey.
board triangle) or binoculars with reticles. A If a survey of general debris trends over time is the
range-finder or binoculars with reticles can only be goal, transect lines in areas known to concentrate de-
used when the horizon is clear. Poor visibility due to bris may be desirable (e.g., northeast to northwest of
fog or low clouds and sea state (i.e., the horizon is the Hawaiian Islands [Nasu and Hiramatsu 1990]). If
obscured by the swells and the ship's pitch is ex- the goal is to monitor conditions in an area of special
treme) will affect the use of these items. Appendix interest (e.g., near the Pribilof Islands [Yoshida and
Figure A (this chapter) contains an example of a Baba 1985b; Baba et al. 1988]), then transect lines
range-finder to estimate the outer boundary of the obviously should be located in that area regardless of
strip and contains a figure explaining the use of the possible surrounding concentration points.
range-finder. Practical advice and further references The number of surveys needed also depends on
concerning the estimation of distances and the use of the objectives. Ribic and Bledsoe (1986, 1990) pro-
range-finders at sea can be found in Gould and duced some sample size estimates based on a specific
Forsell (1989). Along with consultation with experi- definition of a transect for estimating density of
enced observers, this information should be used to floating debris. The study did mot consider stratifica-
prepare a detailed training session for observers to tion because of insufficient information. Because
ensure the quality of the collected distance measure- previous information indicated that debris was sparse
ments. in the open ocean, estimates of sample sizes needed
In addition to carrying out the survey, the ap- were large (and probably impractical on a large
proach for data entry and analysis is outlined. For scale) based on the nonparametric approach of
example, who will check the data to ensure correct Burnham et al. (1980).
entry? What analysis techniques will be used and Given the constraints of open-water surveys done
on vessels of opportunity, the typical sampling design
why? Obviously,, all such details need not be put into is a series of systematic transects made along the ves-
a published paper; however, the quality assurance sel track with the number of transects determined by
program plan should be referenced with key features the number of observers and sighting conditions.
noted in publications to the extent possible, and de- Even with a dedicated cruise, logistical constraints
tails should be available to interested parties on will usually preclude a completely randomized de-
request. Appendix Figure A (this chapter) contains sign. For large oceanic areas, a systematic survey
detailed instructions used by the Tidy Britain Group design would be the alternative in most cases, as seen
to carry out strip transects. in the design of many NOAA National Marine Fisher-
As a practical matter, funding agencies should re- ies Service surveys (Rice and Wolman 1982; Bakkala
quire and review the quality assurance program plan. and Wakabayashi 1985).
If a government agency is funding the study, they
may wish to codify the study protocols as Standard
Operating Procedures (SOP), (e.g., see EPA SOP in Analytical Procedures
Appendix Figure B, chap. 3).
Burnham et al. (1980) give detailed procedures for
analyzing line transect data, including an available
Field Sampling Designs computer program named TRANSECT. Most re-
searchers used strip transect estimates even when line
The actual sampling design depends on the objec- transect information was collected (Dahlberg and
tives of the project. Is the survey a baseline study or Day 1985; Mio and Takehama 1988). For the sighting
an assessment study? If it is an assessment study, is it function model used in line transect estimation, both
to develop model parameters or to detect changes in Nasu and Hiramatsu (1990) and Mio et al. (1990)
space or time (see Table 5 in Chapter 1)? Given the used a hazard rate model that was developed for
large areas of ocean to be considered, some general whale sighting surveys.

CHAPTER 2: Shipboard Sighting Surveysfor Large Debris Item 19

A
For a strip transect, density (D) (number/kml) is Confidence intervals can be calculated (Burnham et
estimated by al. 1980) * Testing hypotheses regarding density esti-
h = n mates between areas and between years can be done
2Lw with a variety of parametric and nonparametric statis-
tical methods. Avoid combining data sets from
where n = number of objects counted, different areas when observations from each area
L = total length of all the transects were made in different years because area differences
(km), and will not be distinguishable from yearly differences.
w = 1/2 the width of the strip transect Analysis of the data set depends on the objectives of
the study, and a statistician should be consulted,
(km).
For a line transect, density b (number/kM2) is esti- Summary
mated by
A
nf (0,0) * Density of marine debris is the preferred field mea-
2L surement.
* Strip and line transect methodologies are available
where A(0,0) the estimated probability distribu- for density estimation. More discussion is necessary
tion function at zero distance before the preferred methodology can be chosen
A for marine debris surveys.
(based on the function f (xO) fit to e At least two observers should be used.
the perpendicular sighting dis- 9 Debris should be sighted by visual observation; bin-
tances) (units are 1/km), n = oculars should be used to identify or verify item
number of objects counted, L = to- classification and to estimate distances.
tal length of all the transects (km). a Cost considerations make dedicated surveys for
Each individual transect can be used to estimate baseline studies unlikely. Vessels of opportunity (see
the variance of A Glossary) are the usual method. Inherent restric-
R A )2 tions on the population of interest should be
Di(A-D recognized and considered.
A i-I
@ar(D)= - L(R-1) e If the same cruise track is resampled over time,
temporal trends can be assessed using vessels of
opportunity.
where -bi= densi ty- estimate for transect i (number/ * Because of the relative sparseness of debris in
kM2), the open ocean, large numbers of transects are
li = length of transect i (km), needed.
R = total number of transects, and o A quality assurance program plan should be devel-
L = total length of all the transects (km). oped describing sampling design, field methods,
See Burnham et al. (1980) for further details. and data analysis.
*.A statistician should be consulted at the survey
Exploratory data analysis can be done to compare planning stage and should be involved throughout
densities between areas or latitude/ longitude bins. the study.

20 NOAA Technical Report NMFS 108: Marine Debris Survey Manual
Appendix

METHOD FOR MEASURING 100 METERS OUT FROM THE SIDE OF THE
SHIP OF-PENDING ON THE HeGHT OF THE OBSERVER ANVE WATER
LEVEL
The litter at sea survey is designed to produce information on the 40
distribution of all floating, man-made material which can be seen
floating on the water from a ship.
The system depends on several factors which will affect its
accuracy. Hortzon
1. How rough is the surface of the sea? Is it too rough to see every- 30 sup-8
thing floating? A wind of force 3 or more usually causes too -E sue Water lavd
many waves to allow the observer to see everything within a band 100 m
100 m from the side of the ship.
2. Ile density of floating litter, or how much litter there is in an 20
area can only be measured accurately by 'sampling.' This means
that everything seen within a known area should represent the 0
amount of litter floating in an area much larger all around. By
Isampling,' this amount can be estimated from the small area 10
being measured.

0
Ile litter at sea survey works on loom
the following idea:
Record the time every 5 minutes. Using the guide above, cut out a paper or cardboard triangle that
All items of litter recorded in each 5- is at the required angle given for your height above the water. This
Smin-
mmute period have been seen in the angle is given by a line drawn between point A and your height above
area shaded. This area is 100 m wide the water. Keeping the right angle of the triangle towards you, look
and 5 minutes long. If the ship's along the remaining edge of the triangle out from the side of the ship.
speed is recorded, then this area can Where this meets the water, that point will be 100 meters out from the
be easily worked out. From this the side of the ship.
number of litter items/km2 can be
estimated. This figure can then be WHAT Do I RECORD ABOUT THE FLOATING LITTER SEEN?
compared with any other measured
0 min-
loom anywhere else, if the same method Any item of floating litter seen within the 100-meter band can be
has been used. described, if it cannot be fully identified, e.g.:
I egg box, polystyrene, undamaged
1 paper cup
WHAT Do I RECORD? 1 plastic sheet about 6 feet long and 2 feet wide
2 wooden planks about 3 feet by 2 feet
1 . The position of the ship at the start of recording (lat. and long.). 8 cardboard sections all about 2 feet square and corrugated
2. 'Me ship's speed in knots. cardboard
1 plastic container, washing up liquid
3. Your height above the water in meters.
4. The approximate wind speed and direction. If a particular item cahnot be identified, then simply record it as
5. Ile time ever 5 minutes. one item made of one of the following:
6. The ship's position at the end of your recording Oat. and long.). 1. wow 6. plastic
7. Number of litter items seen within each 5-minute period of time. 2. fishbox 7. polystyrene
8. Identify as closely as you can every item of litter you see. What 3. all paper and 8. metal drums of all sizes
it is made out of is especially important. If you cannot tell, do not cardboard items 9. glass bottles, containers, fishing
guess!! 4. fishing net floats, etc.
9. Can I clearly see the whole of the width of the sampling area, 5. rope
from the side of the ship out to 100 meters? As a general rule, items which cannot be described at an should
be listed as unidentified. Tlie purpose of the survey is to collect
N=. It is important that only the items of litter seen within the 100- information on the materials making up the litter as well as the num-
meter band are recorded. Others will make all the results wrong. ber of items seen.
T.J. Dixon
St
W
at a_

Appendix Figure A
Offshore marine litter survey (The Tidy Britain Group).

CHAPTER I'Shipboard Sighting Surveysfor Large Debris Itents 21

SmEm RECORDiNG HEADwGs FoR EAcH Dup

Date Ship's name Arind speed
Ship's route (destination) Mind direction
Ships speed
Ship's position: Start Finish

Number of
Time items Description

Appendix Figure A (Continued)

22 NOAA Technical. Report NMFS 108: Marine Debris Survey Manual

RECORD OF SHIPBOARD OBSERVATIONS OF DERELICT FISH NETS
AND DISABLEMENT OF VESSELS BY MARINE LITTER

This form should be included in your ship's Marine Mammal Report which
is forwarded to: Platforms of rtunity Promm, NMML. 7600 Sand Pt.
Way N.E.. Bldg. 4 Seattle, WA 9=5 (F/NWC3)

DERELICT NET OBSERVATION:

WE OF VESSEL
LOCATION OF NET (LORAN or coordinates)
DATE
REPORTER (name, address, phone)
DESCRIPTION OF NET: ell
Stretch-mesh size
Material (monofil nylon, polyprop, etc.)
Color

Twine diameter
Attached floats (number. description)
Estimated size (length. volume, etc.)
Number and tyne of marine mammals, birds, fish in net

If possible, take photographs and forward small representative. sample of
net and floats to above address.

VESSEL DISABLEMENT:

NAME OF VESSEL
LOCATION OF DISABLEMENT (LORAN or coordinates)
DATE
CONTACT FOR FURTHER INFCP.IATICN

CAUSE OF DISABLEMENT:
Net Rope Sheet plastic Other
DESCRIPTION OF DEBRIS:
Material (monofil nylon, polyprop. etc.)
Color
Size (length, volume. wt., etc.)
Twine diameter
Floats attached (number, description)
Stretch-mesh size
Corrective Action (tow to port, cleared without assistance, divers)

ADD ANY ADDITIONAL REMARKS ON REVERSE OF 'THIS FORM.

(NMML- will forward this form to NMFS1 Auke Bay La@boratory)

Appendix Figure B
Offshore marine litter data form used by NOAA/National Marine Mammal Laboratory, Marine Mammal
Sighting Program.

CHAPTER 2: Shipboard Sighting Surveysfor Large Debns Item 23

MARINE MAMMAL SIGHTING FORM
* DO NOT FILL IN BOXES PRECEDED BY AN ASTERISK

1.,NAME RECORD ID
VESSEL YR DAY 1 2 3 4 5 6
2. DATE (Yr./Mo./Day) & TIME (local) OF SIGHTING I @ "10 1 1 1 1 1
7 8 9 10 11 12 13 14 15 16
3. LATITUDE (degrees/minutes/lOths)-N/S 7
18 19 20 21 22 23
4. LONGITUDE (degrees/minutes/lOths)-E/W I 1 1 7
24 25 26 27 28 29 30
5. SPECIES Common name Scientific name FF TENTATIVE -F@
33 34 35
6. NUMBER SIGHTED + C-1. - F-1 I I 1 1:1
36 37 38 39 40
7. BEHAVIOR * 1=
45 46

8.ANGLE FROM 13OW 9. INITIAL SIGHTING DISTANCE
(10's of degrees) 47 48
10's of meters F]
49 50 51
10.VISIBILITY - 11 - SEA STATE (Beaufort) 12. VIS CODE F@
52
13. WEATHER 14. SEA SURFACE TEMP (' C) F] =
53 54 55
15.PLATFORMCODE *F17M 16.TIMEZONE �[--] =
56 57 58 59 60 61 62

17. How did you identify animal (s)? Sketch and describe animal; associated organisms;
behavior (include closest approach); comments.

64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80

Appendix Figure B (Continued)

Cha ter 3
P

Shipboard Trawling Surveys
For Small Debris Items

General Description lematical. In this case, small debris surveys are typi-
cally conducted in conjunction with other oceano-
To sample floating small (and some medium) debris graphic surveys (i.e., vessels of opportunity). Restric-
from a moving vessel, a surface sampler (e.g., neus- tions to the population of interest can arise. For
ton net) is pulled along the sea surface to collect example, in studies done in conjunction with large
debris. The lowest size of the small debris sampled surveys of fish, larvae, the areas chosen and the sur-
depends on mesh of the net used. The upper limits vey design used are appropriate for the objectives of
of medium debris sampled depend on the size of the the fish larvae survey, and not necessarily for the de-
opening (mouth). The density of debris is calculated, bris survey. Whether or not a restriction of the
and the distribution of the debris is mapped. population of interest exists and whether or not it is
Most studies of small floating debris have been significant in relation. to the objectives of the debris
conducted in conjunction with large surveys of fish survey have to be evaluated on a case-by-case basis.
larvae or plankton; such surveys typically cover large
oceanic areas (e.g., North Atlantic-Colton et al.
1974; southeastern continental shelf and slope water Field Measurement
of the United States-van Dolah et al. 1980; sea sur-
face off the southwestern Cape Province of South The number, type, density, and weight of small debris
Africa-Ryan 1988b). Some surveys have been made in a strip transect are common variables of interest.
in harbors and bays (Trulli et al. 1990; U.S. EPA
1990b; Yukinawa and Mio 1990). Most, if not all, of
the studies to date concerning floating small debris Description
would be considered to be baseline studies (see Table
6 for sources). A surface sampler (e.g., a neuston net) is deployed by
I or 2 people to sample the surface water while the
ship is moving at speeds usually <5 knots. The net is
Ob ectives and Purpose placed on a boom so that it can sample water sur- I
faces outside the ship's wake, either to the side of the
Typical objectives are as follows: ship or forward of the bow. The vessels used by a
1. Estimate types and quantities of small debris, harbor studies program are shown in Figure 4. The
2. determine distribution of small debris, and amount of time the net is sampling (or alternately,
3. assess changes in the types and amounts of not sampling) is calculated by one person watching
small debris in given areas over time. the net during the tow. The net may be equipped
with a flowmeter to measure the volume of water
sampled directly (Carpenter 1976), although debris
Population of Interest may foul the flowmeter, causing it to fail. The width
of the strip is determined by the width of the open-
As with sighting surveys, various types of debris in a ing (mouth)'of the net used (Table 6). Length of the
large oceanic area are usually the populations of in- transect is the straight-line distance traveled, deter-
terest. The population of interest can be as small as mined by the ship's speed during the transect and
the floating debris in harbors (Trulli et al. 1990; U.S. duration of tow. Lengths of transects from previous
EPA 1990b; Yukinawa and Mio 1990) and as large as studies varied from 0.065 nmi to 13 nmi (Table 6). In
floating debris in coastal and oceanic waters between areas of high debris density, lengths will be shorter
Cape Cod and the Caribbean Sea (Colton et al. because the net can be towed only until it is full.
1974). While smaller areas can usually be surveyed Longer tows are typically made in areas of low debris
easily, surveys of large oceanic areas are more prob- density; short tows (e.g., 2 min in length) may miss

26 NOAA Technical Report NMFS 108: Marine Debris Survey Manual

A OSV Peter W. Anderson

Capstan

Deployment
Aft sta and
retrieval line

lx2x4-m neuston
net, 0.33-mm rnesh
Whaler Forward stay

"*--Telescopin boom

B Whaler

K-A
0.5xlx2-m Neuston net, ""@Sleeve
0.33-mm mesh
Figure 4
rc Fixed boom Example of sampling
vessel equipped with
telescoping boom
(from the U.S. En-
vironmental Protec-
tion Agency [ 1990b]).

the small debris entirely. The size of debris sampled the net. Tows should not be made in strong winds
is determined by the mesh size of the net (Table 6). because the net will sail out of the water.
Larger mesh sizes do not collect all of the smallest- Net mesh-The size of debris collected by the surface
sized debris. sampler depends on the size of the net mesh. Ulti-
mately, the particular debris of interest will influence
the net mesh size. Carpenter (1976) and the U.S.
Variables to Consider EPA (1990b) recommend using a net mesh of 0.333
mm to sample polystyrene spherules. We therefore
Weather-If the sea is too choppy, tows should not be recommend that 0.333-mm mesh net be used in most
made because the net will be submerged or be com- cases because it will catch most sizes of small debris.
pletely out of the water, and it will not sample the Ship's Speed-Most tows are made at speeds of 5
surface waters. In addition, debris may become knots or less (Table 6), although some larger mesh
Capsta

De ment
TA ta pl eonyd
retrieval line

x4
net
ad stay

resuspended in the, water column below the level of nets may be used at speeds up to 7 knots. The ship

CHAPTER 3: Shipboard. Trawling Surveysfor Small Debris Itons 27

Table 6
Examples of nets used to sample floating small debris.

Net Description Length of
Reference Net mesh of tow tow (nmi)

Austin and Stoop- 3/4 opening plankton net Not 5-min surface, Not stated
Glas (1977) stated tow

Environmental Neuston-type net with dimensions: 0.33 m 30-min tow at Variable
Protection Agency I X 2 X 4 m, 0.5 X I X 4 m, 0.5 X I X 2 knots depending
(1990b); Trulli et al. 2 m on slick
(1990) size and
debris
density

Carpenter et al. Oblique plantkton tows with 0.5-m 0.333 mm Not stated Not stated
diameter, at mouth

Carpenter and 1-m diameter neuston net 0.33 mm 30-min. to 1-3
Smith (1972) 6-1 / 2-hr tows
at 2 knots

Colton et al. (1974) 2 X I m rectangular neuston net 0.947 mm 10-min. tow at 0.83
5 knots

Day and Shaw Ring net (1.3-m mouth diameter, 0.333 mm 10-min tow at Q.50
(1987) 4.5-m length) 5.6 km/h

Day et al. (1990b) Sameoto neuston sampler: 0.5 m 0.500 mm 10-min tow at 0.33
wide X 0.3 m high X 0.6 m long 2 knots

FAO (1989) No details given 1.4 cm 20-min tow at 0.83-1
2.5-3 knots

Gregory et al. Sameoto andjarozynski aluminum- 0.475, -1 -hr tow at 4-6.5
(1984) framed otter-style neuston net (60-cm 0.860 mm 4-6-'/2 knots
wide mouth), 0.860 mm or
0.475 mm net mesh (40 cm X
40 cm or 60 cm X 40 cm)

Morris (1980a) Neuston sledge 0.32 mm 20 to 45-min 0.67-3
tows at 2-4
knots

Morris and Hamilton Lowestoft plankton sampler 0.270 mm Not stated Not stated
(1974)

Ryan (1988a) 1.57 m X 0.42 m rectangular 0.90 mm 2-min tow 0.65
neuston net at I m/sec

Shaw (1977) Sameoto andjarozynski seston 0.363 mm 15-min tow 1
tow: 0.4 in wide mouth at 4 knots

Shaw and Mapes Sameoto andjarozynski seston 0.363 mm I nmi 1
(1979) tow

van Dolah et al. I m X 2 m Boothbay neuston net 0.947 mm 10-min tow at 0.84-0.87
(1980) 2.5 m/sec or
15"min tow at
1.8 m/sec

28 NOAA Technical Report NMFS 108: Marine Debris Survey Manual

Table 6 (continued)

Net Description Length of
Reference Net mesh of tow tow (nmi)

Yukinawa and Mio Ring net with 1.4 m diameter 1.7 min at 10 min at 0.5
(1990) mouth mouth, 3 knots
0.5 mm
at cod end

Wilber (1987) 1 X 0.5 m neuston net 0.333, 1 nmi
0.500 mm

Wong et al. (1974) 80 cm X 30.5 cm Kohl scientific 0.150 mm 10 to 15-min 0.67-1.25
model neuston net tows at 4-5
knots

must be going faster than the tide in bays and sounds. contains a suggested data form for use with surface
For nets with a 0.333-mm mesh, we recommend a tow- sampling, and Figure 6 contains a suggested data
ing speed of approximately 2 knots. The duration of form for sorting samples.
the tow depends on the amount of debris in the water;
however, a tow should be made for at least 10 min. Material and Personnel
The duration of the tow also will be influenced by
other ship operations and will have to be determined Basic equipment needed for sampling small debris is
on a case-by-case basis. The speed of the ship should as follows:
be such that there is no backwash.
0 To collect the samples:
neuston net (surface sampler) and all equipment
needed to deploy the net
Data Collection log forms (sample collection) (see Figure 5)
The following information should be recorded each * To analyze the samples:
time a sampler is -deployed. pre-weighed vials
bottles to store samples/labels for the vials
time of day trays for sorting and drying samples
location forceps (to sort samples)
date magnifying glass (to identify samples)
volume of water sampled (if the net has a flow- data sheets (for sarnple analysis) (see Figure 6)
meter) or time the net is not sampling rounded to gloves
the nearest 0.25 min (if net does not have a 30-L bags for medium debris
flowmeter); without a flowmeter, duration of sam- No standard surface sampling net is available; Car-
pling equals duration of tow minus duration of penter (1976) describes a variety of nets, although no
net not sampling prices are given. The Sameoto and jarozynski (1969)
distance traveled: starting and ending position net will cost about $75 (1991 U.S. dollars), while
(latitude/longitude) or duration of transect and other neuston nets may cost $100-$300 (1991 U.S.
ship's speed dollars) depending on mesh size and type of frame
width of sampler (fi)Ced by choice of net) used Ua Halstead, Research Nets, Inc., Bothell, WA,
sea state and tidal conditions pers. commun. August 1991). Square-mouthed nets
presence/size of debris concentrations (if sam- will give a less variable area sampled than ring nets
pled) /current features (H. Trulli, Battelle Memorial Institute, Duxbury, MA,
Few forms have been published for recording pers. commun. January 1991); therefore, square-
sample collection. Appendix Figure A (this chapter) mouthed nets are preferred over ring nets.
contains the sample collection form used by the U.S, Budgeting $1,500 for supplies should cover the inci-
Environmental Protection Agency (1990b). Figure 5 dental items.

CIMPTER 3: Shipboard Trawling Surveysfor Small Debris Item 29

Observers: Date (Yr/Month/Day):
Ship: Sea State or Tide:

Tow/Sample No.
Starting Location: lat Ending Location: ]at

long long

Time: Start End Total

Time Not Sampling Time Sampling

Ship's Speed at Start of Tow-

Width of Sampler

Comments: e.g., (a) unusual weather conditions, (b) problems encountered while sampling.

Figure 5
Suggested data form for collecting surface samples of small debris.

The major cost of the project is ship time, as dis- must have a retractable boom for net deployment or
cussed in sighting surveys (Chapter 2). For sampling must have the capacity for a fixed boom (Figure 4).
inshore waters and harbors, smaller boats may be Required personnel include researchers who are
used, thus reducing costs as noted in Chapter 2. For experienced in or trained in the use of a surface sam-
example, the U.S. Environmental Protection Agency pler. Training in safety issues is essential, including
(1990b) used 5 in (17 ft) Boston Whaler skiffs in potential health risks from handling landbased
some of the harbor surveys. In all cases, the ship wastes such as hypodermic needles.

30 NOAA Technical Report NNOFS 108: Marine Debris Survey Manual

Tow/Sample Number Wet/Dry Weight

Presence of
Item Color Wear Shape Length Width Encrusting
Biota

Figure 6
Suggested data form for analysis of surface samples of small debris.

.Quality Assurance Program 0 choice of the population of interest and any re-
strictions to it
Most published papers give enough information to 0 details on sampling design
understand how the study was done. Putting the in- 0 description of sampler including dimensions,
formation into a standard format would facilitate mesh size, and net maker
comparison between studies. Suggested information 0 definition of tow length
in a quality assurance program plan includes the fol- 0 how the tows were carried out (ship's speed,
lowing: transect duration, net deployment)

CHAPTER 3: Shipboard Trawling Surveysfor Small Debris Item 31
sorting/ handling of materials (sorting and stor- egg and larvae surveys (Colton et a]. 1974; Carpenter
age), including proper handling of hazardous et al. 1982; Ryan 1988a), the design used for the
wastes small debris surveys was the same as the fish egg and
key for identifying material larvae surveys: systematic sampling in a grid. Sam-
storage of material (who is responsible and the pling both large debris and small debris would be
location of material) possible using a systematic sample on a grid. An im-
data handling (checking and storage of data, portant design consideration would be the distance
sample tracking) between the sampling stations (i.e., the possibility of
data analysis procedures sighting the same piece of large debris from two dif-
The EPA's Office of Marine and Estuarine Protec- ferent stations should be zero). But one potential
tion has a Standard Operating Procedure for the sampling design would be to sample the small debris
Collection and At-Sea Processing of Neuston Samples at the points on the'grid and carry out sighting sur-
(SOP No. 4-35) (U.S. EPA 1987). We modified this veys as the ship is in transit between points. A second
SOP to reflect marine debris. Appendix Figure B possibility would be concurrent sampling, which may
(this chapter) contains a suggested SOP potentially be possible depending on the amount of debris in
applicable to collecting small floating debris. the area. For example, if the small debris surface
sampler is towed for 1 hour, then sighting surveys
may be done concurrently. Alternatively, if the sur-
Field Sampling Designs face tow is only made for 10 min, concurrent sighting
surveys would bF questionable, particularly if the
Ribic and Bledsoe (1986, 1990) addressed the ques- density of larger debris was low (i.e., few larger debris
tion of sample sizes for small debris surveys. They items would be seen in 10 min).
assumed that the small debris was randomly distrib-
uted in the ocean and was not concentrated in any Analytical Procedures
areas. However, some evidence suggests that debris
can be concentrated in certain areas by currents and
other oceanographic and weather conditions. Ran- Handling of Material
dom samples from nonuniformly distributed debris
will yield tows with varying amounts of material, and After collecting a sample, most researchers washed it
the variance estimate from the tows will be large. Be- and put it into a bottle for further analysis (see Table
cause sample size calculations depend on an estimate 6 for references). A few researchers froze the sample
of variability, large variance estimates increase the or fixed it in a seawater-formalin mixture or in alco-
number of required samples, a problem that has hol (e.g., 70% ethanol.). Preservation is preferable to
been noted by researchers studying tar balls (Butler freezing because freezing may crack the debris,
and Morris 1974). For obvious debris concentrations changing its size. The material in each tow was classi-
such as drift lines, Carpenter (1976) recommended fied as to category and measured. Table 3 (Chapter
sampling perpendicular to the concentration or do- 1) lists some categories that could be used in sample
ing circular transects. An alternative is to collect analysis. We recommend that the material be classi-
more samples in the concentrated areas to better de- fied as to color and, if possible, wear. Encrusting
fine the area of concentration, and then to use biota (which indicate length of time at sea) also
declustering techniques (Isaaks and Srivastava 1989) should be noted. The material in each tow should be
to make an overall assessment. If the ob ective of the weighed and counted. Most studies have not speci-
study is just to find out what is in an area to deter- fied wet or dry weight. We recommend using dry
mine types of debris, sampling the concentrations weight measurements, but in all cases investigators
will be useful (e.g., U.S. EPA 1990b). should.indicate whether wet or dry weights are used.
Because most researchers rarely do both sighting When dry weight was specified, material was dried at
surveys for large debris and surface sampling for room temperature for I day (van Dolah et al. 1980)
small debris (exceptions are,Day and Shaw 1987; and to I week (Day et al. 1985). Figure 6 contains a sug-
Day et al. 1990, a and b), the sampling designs for gested sample analysis form.
the two techniques are often considered separately.
However, the sampling designs discussed in Chapter
2 for sighting surveys can be used for surface sam- Analysis of Data
pling of small debris. Since many of the vessels of
opportunity used for small debris sampling were gov- Data are transformed into estimates of density (num-
ernmental research ships carrying out large-scale fish ber or weight/kM2) . The equation for density follows

32 NOAA Technical Report NMIFS 108: Marine Debris Survey Manual

the strip transect estimate given in Chapter 2. There nonparametric analyses, depending on the objectives
are two ways to determine length of the tow (km): of the study. Data collected from different areas in
1. (duration of sampling) (speed of vessel), or different years should not be compared because area
2. straight-line distance between the starting and differences cannot be distinguished from yearly dif-
ending coordinates (adjusted for distance the net ferences. For data collected on a grid, geostatistical
was not sampling). techniques may be useful for estimating density over
A large areas (Isaaks and Srivastava 1989).
The density of (D) (amount/kml) is as follows:

Summary
A _ n
D - Lw 0 Density of small debris (number/weight per kM2)
is the preferred field measurement.
where w = width of the mouth of the net (km), 0 The mesh of the sampling unit should be 0.333
L = total length of all the tows (km), and mm, which will sample most small debris of inter-
n = number or weight (g) of collected debris est.
in all trawls. 0 Tows should be made at speeds of 2 knots for at
least 10 min. The duration of the tow will be in-
of (A fluenced by the amount of debris and ship
The variability D) (Burnham et al. 1980) can be
calculated as: operations.
0 Cost considerations make dedicated surveys for
R A A baseline studies unlikely.
Y-liDi - D)2
A A 0 If the restriction to the population of interest is
Var (D) acceptable, vessels of opportunity that sample the
L (R 1) same large oceanic area over multiple years
where 1. = length of tow i (km), should be useful for assessment studies.
1 0 Variability in amounts of debris per tow can be
D. = density estimate for tow i (in km2), large, thus increasing the required sample size
@ = total number of tows, and (number of tows) for a given level of confidence.
L = total length of all the tows (km). * The data are collected with a strip transect meth-
odology, with the width of the net defining the
The data can be analyzed in a variety of ways de- width of the strip.
pending on the study objectives. General descriptive 0 A quality assurance program plan should be pre-
statistics (such as percent composition of the debris pared that details sample design, sample col-
from each sample or percent composition of all de- lection, sample p rocessing, and data analysis.
.bris from a harbor) are useful as are maps; 0 A statistician should be consulted at the survey
exploratory data analysis can be used to look at dif- planning stage and should be involved through
ferences between areas and over time. Confidence the completion of the study. Experienced re-
intervals for the density estimate can be calculated searchers should be consulted concerning the use
(Burnham et al. 1980) or various hypotheses con- of equipment (e.g., surface samplers, boat han-
cerning density can be tested with parametric or dling).

CHAPTER 3: Shipboard Trawling Surveysfor Small Debris Item 33
A"pendix

BATTELLE
Harbor Studies Program

Neustan Sampling Log

Date Harbor
Do MM YY (NY,B0,GA,LA,PS)

Slick No. Slick Location

Replicate/Tow No. Sampling Platform

Start TDI TD2 Tine
Coordinatt-s: Lat N Lon W (24 h)

Finish TD1 T02 Time
Coordinates: Lat N Lon W (24 h)

FIELD MEASUREMENTS Collector's Initials:
Weather: TE Sea State: Wind: Speed kt
ode) (coze-) Direction 0

Net Frame Size: 1.0m x 2.Om Mesh Size: 0.1mm
0.52 X 1.0m 0.3mm
0.5mm

Tow Area: Area of Tow - Duration of Tow h x
Speed of Net Through the Water t X.
Width of Net (Tow) m

Slick Type: Slick Area: sq m
(Type 0-4 from Survey Plan) (visual estimaCe)

LABORATORY SAMPLES COLLECTED 11111111111111 @
Sample Types Collected for Sample Number *AAXS88*

(initial Here and on Labels)

Large Debris Small@DebrLs

COMMENTS:

Scientist:

(White - DATA MGR Yellow - PRGM MGR Pink FIELD COORD GOLD P.I.)

Appendix Figure A
Example of a sampling log used during EPA Harbor Studies Program survey (U.S. Environmental
Protection Agency 1990b)

34 NOAA Technical Report NMFS 108: Marine Debris Survey Manual

5. Shackles (0.63 cm or 1/4 in)
6. Kevlar cable-0.95 cm (3/8 in), or 0.95 cm
(3/8 in) nylon line
Preparation 7. Nylon braided line (0.32 cm or 1/8 in)
8. Hose clamps (various sizes)
Supplies and Equipment 9. Plasticiar (1 L)
10. Sorting trays
I .Two aluminum neuston net frames, 0.83 m X 11. Forceps
1.85 m X 0.04 m 12. Squirt bottles (for rinsing nets)
2. Two nylon neuston nets, 0.333-mm mesh size 13. Clear tape
(must fit aluminum frame) 14. Labels
3. Laundry detergent 15. Log sheets
4. Stainless steel bridle (for each neuston net 16. Cod-end bottle (i.e., 500 ml teflon) (also
and frame) called a "net bucket")

0.33 mm
Mesh net

Neu3ton frame
and not

0. mm

.<-33"FL

Frame 1-112' Se ring 10"tlap
cu
aluminumiubing One (passes inside
the frame)

b
Neuston not Akirninum tabs
with attached
bridles

1/6'wire rope

Eye with themble
1/6"wire rope

Swivel

112" trawl cable

Appendix Figure B
Preliminary standard operations procedures for sampling surface small debris. Adapted from the plan described in the U.S.
Environmental Protection Agency (1990b) Harbor Studies Program.

CHAPTER 3: Shipboard Trawling Surveysfor Small Debris Items 35

The registered trademarks and materials suppli- ft) section of Kevlar cable or nylon line with
ers are referenced for reader convenience in an eye splice.
replicating experiments and do not represent en- 9. With a 0.79 cm (5/16 in) shackle, attach one
dorsement by the U.S. Environmental Protection end of the Kevlar cable to the 0.63 cm (1/4
Agency. in) shackle thatjoins the two bridles.

Preparations for Deploying the
Cleaning Net and Components to be Used
No special cleaning procedures are required for for Collecting Samples
collecting neuston samples for small debris analy- 1. Request that the Captain slow the ship to tow-
sis. Equipment can be washed with detergent at ing speed (2 knots) before arriving on station.
the end of the sampling period. Netting may also 2. Rig the main towing sheave to the last segment
be solvent-cleaned if soiled with grease or tar. of the telescoping boom.
3. Carry the assembled net, bridle, and tow cable
(or nylon line) to the area of deployment.
Assembly of Net and Frame 4. Thread the tow cable or nylon line through
the sheave mounted on the boom.
1. Match each pair of the four flaps bordered 5. Take up the slack (by hand) and tie off the
with grommets at the mouth of the neuston tow cable (or nylon line) to the bits, forward
net (2 opposed flaps = 0.9 m and 2 opposed of the deployment area.
flaps = 1.9 m) with the appropriate sides of the
neuston frame as shown in Figure a. Net Deployment
2. By threading the 0.32 cm (1/8 in) braid
through the grommets and around the frame, 1. With the boom fully extended, signal the
secure the flaps of the net to the aluminum winch operator to lower the boom over the
frame. Ensure that the frame attaches to the side of the ship. This procedure will lower the
outer surface of the net flaps. The net passes net into the water and extend it well beyond
through the frame as shown in Figure a. the wake of the ship.
3. After securing the net to the frame, remove 2. After the boom has been completely lowered,
the two 0.63 cm (1/4 in) stainless steel bridles deploy the Kevlar or nylon tow cable until half
(Fig. b) from the case. of the neuston frame (0.42 m) remains sub-
4. Arrange the two towing bridles so that the merged below the surface of the ocean.
longest cables (one for each bridle) are di- 3. Inform the bridge to mark the time for the
rected away from the net and so that the short start of the neuston tow.
cables (two for each bridle) are directed to- 4. Record the requested information on a Neu-
ward the net. ston Sample Log form.
5. With 0.63 cm (1/4 in) shackles, connect the 5. Tow the net for 10-min at a speed of 2 knots or
two short cables of each bridle to the alumi- until full. In rough seas, estimate the time the
num tabs provided on each corner of the net completely leaves the water or completely
neuston net. Shackle the short cables of one submerges while being towed.
bridle to the left side of the frame and the
short cables of the other bridle to the right Net Retrieval
side of the frame.
6. With a 0.63 cm (1/4 in) shackle, connect the At the end of the 10 min towing period, retrieve
free ends of the bridles together (if not al- the net using the tow cable (or line).
ready done).
7. With a stainless steel hose clamp, attach a 1-L
precleaned glass jar to the cod end of the ny- Processing Samples
lon neuston net.
8. With 0.79 cm (5/16 in) cable clamps and 1. Discard all water retained in the jar at the cod
thimbles, terminate each end of the 45 m (150 end of the net.

Appendix Figure B (continued)

36 NOAA Technical Report NMFS 108: Marine Debris Survey Manual

2. Loosen the hose clamp that secures the cod necessary to alter the procedures outlined in the
end to the net. SOP or to deviate from the guidelines appearing
3. Support the cod end of the net over a tray and in the quality assurance program plan, make sure
separate the cod end from the net. that a Sample Alteration Form (U.S. EPA 1990b) is
4. Empty the contents of the jar and net into the completed for that particular tow.
tray. Larger material may have to be removed
from the mouth end because it cannot pass
through the cod end. Training
5. Rinse all net surfaces into the tray.
6. Transfer contents of the net into prelabeled All personnel responsible for the collection and
jars (I L) or bags. preservation of samples must perform satisfacto-
7. Precautions: rily under the direct supervision of a qualified
a. Small and lightweight items are easily lost in supervisor. This includes the proper implementa-
windy conditions and extra care, must be tion of safety guidelines.
taken. The proficiency of the trainee will be observed
b. Rubber gloves and protective clothing throughout the entire survey. The supervisor will
should be worn because of possible disease- judge the ability of the trainee to process debris
bearing debris (e.g., from combined sewer samples.
overflow systems).

Documentation and Labeling

Record all required information on a Neuston
Sample Log. If, during a neuston tow, it becomes

Appendix Figure B (continued)

Chapter 4

Beach Surveys
for Small to Large Debris Items

General Description have been used in this chapter for recommendations
and guidance. Information from these two programs
On-land or beach surveys can be classified into two is summarized here because much of the written ma-
types: 1) where debris on a particular beach is of terial is difficult to acquire (e.g., . government
intrinsic interest (beach-focused studies); and 2) reports) and spans some 20 years. For people plan-
where debris on the beach is an indicator of oceanic ning a survey, this information can aid in project
conditions (ocean-focused studies) (Ribic and John- development, show how initial study designs are in-
son 1990). Beach surveys are known to give a fluenced by the debris types, and explain how
distorted picture of the composition of marine debris programs may change over time.
owing to different fates of materials at sea (Dixon
and Dixon 1981a). To date, no attempts have been
made to assess what proportion of debris discharged Objectives and Purpose
from ships at sea later washes ashore. Some surface
drift experiments using plastic and glass bottles (i.e., There are two general objectives of on-land (beach)
the release of bottles at sea and their recovery on surveys for marine debris:
land) have achieved high recovery rates in the North 1. To determine the types and amounts of debris on
Sea (Dixon and Cooke 1977). Beach surveys inte- beaches in a specified geographical area at a cer-
grated with at-sea surveys are a potentially powerful tain time. This objective is generally associated
tool. with baseline studiesand is beach-focused.
To assess marine debris on a beach, surveyors 2. To determine how types and amounts (or both)
count and classify individual debris items or record of debris on beaches change over time. This ob-
them as present or absent. Debris may or may not be jective is associated with assessment studies
removed from the beaches depending on study ob- (usually trend assessment), and can consider ves-
jectives. Entire beaches or smaller sections sel-source debris or specific landbased debris
(transects) may be surveyed. Individual pieces of such as sewage items and medical wastes (i.e.,
small debris are usually counted within randomly ocean-focused studies).
placed or predetermined transects; small debris gen-
erally is not removed. These two objectives require different field de-
This chapter addresses both beach-focused and signs, as noted by the Assessment Working Group
ocean-focused studies considering vessel-source de- (Ribic 1990). For the first objective, standard survey
bris and landbased debris. New international and sampling techniques can be applied (Gilbert 1987).
national disposal restrictions have increased interest For the second objective, the Assessment Working
in using ocean-focused studies to detect changes in Group (Ribic 1990) recommended that selected
vessel-source debris on beaches. Therefore, addi- beaches be monitored over time. The selection of the
tional attention is given to the use of beach surveys to beaches for monitoring should be guided by statisti-
monitor vessel-source debris. This will be done by cal sampling techniques such as stratification and
presenting two case studies using beach surveys to randomness (Gilbert 1987). This is similar to a time-
detect changes in vessel-source debris. The first is a series approach used in water quality monitoring
program developed by Theodore Merrell and Scott (Lettenmaier 1978).
Johnson (NOAA, National Marine Fisheries Service,
Auke Bay, Alaska) to study very large debris (i.e.,
trawl web net fragments). The second is the program Population of Interest
developed by the Tidy Britain Group in England to
study large debris (i.e., plastic containers). The devel- For baseline, beach-focused studies, the population
opers of the two programs have published extensively of interest is all or some subset of debris on all, the
on their study designs and methods, and their results beaches in a defined geographic area at a particular

38 NOAA Technical Report NMFS 108: Marine Debris Survey Manual
time. Debris can be landbased or vessel-source. Re- promontory). Oceanic influences may be docu-
strictions to the population of interest are generally mented from recent hydrographic studies, such
due to access problems (e.g., private beaches, remote as surface drift experiments;
areas). Consideration of how debris types or amounts 2. collect vessel-source debris (due to physical char-
may be different on the restricted beaches compared acteristics of the beach, such as substrate type
to the unrestricted beaches is necessary to decide if and slope); and
problems will occur. For trend assessment, ocean-fo- 3. be remote from populated areas (urban sources
cused studies, the population of interest is the of pollution, recreation areas) and estuarine ar-
amount of marine debris in an unspecified area of eas (e.g., no known or limited landbased sources
the adjoining water mass that is swept onto beaches of debris), unless methods are available for
after a certain time interval. Vessel-source or specific clearly identifying each source of debris.
landbased debris on beaches is being used as an indi-
cator of the oceanic debris condition. The degree to Information on oceanic current patterns and ves-
which beach debris reflects marine debris conditions sel traffic information (Requirement 1) can be used
in the open water is not commonly evaluated. For to identify beaches that are more likely than others to
both types of studies, beach dynamics, such as depo- collect vessel-source debris. Information on currents
sition rates and current influences, need to be is often used after a study has been done to interpret
understood. the results.
If fulfilling the third requirement is impossible,
some a p7iori decision must be made about differenti-
Historical Information ating between vessel-source and landbased debris.
For example, O'Hara (1989) used indicator items
such as galley wastes, fishing/boating gear, and other
characteristic operational wastes for vessel-source de-
Baseline Studies bris information. The Food and Agricultural Or-
ganization (FAO 1989) decided that metal and glass
Because baseline studies focus on beaches, informa- debris are probably landbased because, given their
tion on the numbers and types of beaches in the weight, they likely would sink soon after being dis-
population of interest is necessary. Other variables charged from ships. However, on the shores of the
such as distance to urban centers, coastal dumps and North Sea, it was clear that metal and glass did wash
landfills, recreational centers (e.g., campgrounds), ashore from ships (Dixon and Dixon 1981). Further-
and other debris sources would be useful to identify more, metal and glass containers were found to be
possible stratification variables to be used in study prominent components of ships' garbage (Horsman
design and to help in interpreting the data. Previous 1982). Any approaches and assumptions made to dis-
records of types of debris may be useful in determin- tinguish landbased and vessel-source debris should
ing whether the majority of debris is landbased or be described explicitly.
vessel-source. The conditions for using beaches to indicate
landbased marine debris are the same as the first two
conditions for vessel-source debris indicator beaches.
Trend Assessment Studies Oceanic influences will be important to predict
which areas of the coast may be vulnerable to
Many studies have commented on factors that restrict landbased debris (e.g., landbased debris coming
the choice of beaches when the objective is trend from harbors, river mouths, or offshore dumping ar-
assessment. The debris of interest may be either ves- eas). This information would be directly relevant as a
sel-source debris or landbased debris (specifically stratifying variable for the survey design. The poten-
sewage-related items or medical waste). tial beaches to be used for landbased marine debris
To use beach surveys as indicators of oceanic con- do not have the remoteness condition necessary for
ditions, indicator beaches must meet the following vessel-source debris indicator beaches (Condition 3).
conditions: Since landbased debris is generated by humans in
populated areas, the potential beaches should not be
1. have known oceanic influences (e.g., currents isolated from areas of known human influence, such
running past or converging in the area) or be as urban sources (e.g., sewage and industrial outfalls,
open to the marine environment (e.g., not shel- harbors) or estuaries. The potential indicator
tered by a breakwater, island, or land beaches must collect landbased debris, so variables

CHAPTER 4: Beach Surveysfor SmaH to Large Debrislierm 39
affecting the choice of beaches would be the same Accumulation rate is the amount of debris that
for both types of surveys. washes and stays ashore on the sampling unit over a
Variables to consider when doing beach surveys are certain time period. In order to measure this, the
the physical characteristics of the beach (slope, sub- sampling unit must be cleared of all debris at the
strate, composition, uniformity), prevailing weather start of the sampling time period. This variable is
factors (onshore winds, frequency of storms), beach actually the accumulation rate of visible materials.
accessibility (private, public; roads and parking Many items become buried on beaches soon after
nearby), and composition of debris on beach deposition, particularly the heavier items. In an ideal
(landbased and vessel-source categories). The physi- situation, observations should be made daily in order
cal characteristics and accessibility of beaches to produce the most accurate data.
influence the number of potential beaches that can Standing stock is the amount of material on the
be used for sampling. For example, beaches with low sampling unit at a given point in time. In general, the
or very steep gradients, or beaches consisting of boul- sampling unit is not cleared of debris (this is dis-
ders, should not be candidates for sampling. cussed further under Field Sampling Designs).
Low-gradient beaches are especially unsuitable be- In some surveys, the variable that was measured
cause. storm winds and surf scatter debris inland, was unclear (FA0 1989). This becomes an important
where it becomes hidden in vegetation. Boulder as consideration because of the increased popularity of
well as bedrock beaches also are unsuitable: debris beach cleanups (CEE 1987b, 1988; O'Hara and
between boulders is difficult to see; bedrock beaches Debenham 1989; O'Hara and Younger 1990).
are often too steep for walking and do not acCUMU7 Beaches with a cleanup history will only be suitable
late debris. In the United States, preferred beaches for measuring accumulation rate since the last
for marine debris surveys have moderate-to-steep cleanup.
slopes, have sand or gravel substrate, and are ex- For small debris, standing stock usually is mea-
posed to the open ocean. In addition, beaches sured because of the difficulty involved in removing
should not be cleaned during annual "beach clean- such items from the sampling unit. Accumulation
ups" (unless the beach cleanup matches the sampling rates are more easily measured for larger debris. Ac-
frequency of the planned study). cumulation rate is the preferred measurement
Physical characteristics of beaches affect debris variable, particularly for trend assessment studies, be-
turnover time (i.e., the rate of disappearance of de- cause it will be more sensitive to changing oceanic
bris from the beach), which is important for conditions (see Golik [1982] for discussion in relation
determining sampling frequency, especially when to tar balls).
measuring standing stock. For frequent sampling, Choices must be made concerning what particular
weather factors influence the timing of sampling debris items to study. In some cases, focusing on a
(e.g., sampling should not be done during periods of particular debris type will be as informative as enu-
offshore winds or during storms). Beach accessibility merating all debris types. This is particularly
is important because private beaches usually have re- important for ocean-focused studies where studying a
stricted access; beaches in remote areas, as in Alaska, known vessel-source or landbased debris type is im-
may be prohibitively expensive to reach regularly. De- portant. Fishing gear such as trawl web (e.g., Merrell
bris composition will affect what is measured 1985) and plastic containers (e.g., Dixon and Cooke
(standing stock or accumulation rate) as well as what 1977) have been successfully used in ocean-focused,
the sampling unit should be (e.g., the entire beach vessel-source debris studies. The advantage of study-
or transects). Sampling multiple high tide lines over ing plastic containers is that they can be aged by
� reasonably narrow beach profile is ideal for getting codes molded into the plastic body. Also, in some
� good sample of the debris composition in a sam- cases, origin can be identified by codes on the con-
pling unit. tainer. Information on manufacturing dates and
country of origin can help in assessing changing ac-
cumulation rates.

Field Measurement

Two basic variables can be measured: accumulation Material and Personnel
rate and standing stock. A choice must be made be-
tween the two because both cannot be measured on Following is a list of suggested basic equipment for
the same sampling unit. an on-land (beach) survey.

40 NOAA Technical Report NMFS 108: Maxine Debris Survey Manual
� Metric measurement To measure length of nel costs, but more supervision and training will be
tapes beach surveyed and necessary to ensure quality of collected data. Beach
measure very large surveys are considerably less costly than open-water
debris too big to lift. A sighting surveys.
100-m tape is usually
adequate Data Collection
� Metric ruler For measuring mesh
sizes Suggested information to collect includes the
� Stakes, flagging tape, To mark ends of the following:
PVC (polyvinyl chloride) survey area, to mark I . Date
pipe transects 2. Time (start and end)
� Topographical map To mark survey area 3. Location
or photographs 4. Weather conditions
(or both) 5. List of debris items by type and number (see Ap-
� Tags and paint for For accumulation rate pendix Figures A-D [this chapter])
marking very objects: marking pens 6. Volume or approximate weight of large objects
large debris for medium and large 7. Containers-geographical origins, bottlemaker
debris imprints
� Waterproof data forms 8. Beach conditions (slope, substrate, etc.)
and clipboard The specific data to collect will depend on the ob-
� Random number table or To choose random jectives of the beach survey. Most studies list all
calculator with random transects; for standing debris items and then, during the data analysis phase,
number generator stock put the debris items into function or material-type
Knife/scissors categories or both. Some studies focusing on indica-
Heavy work gloves tor items or particular problems (e.g., entanglement)
Small and large To collect specimens may list the debris categories before the data are col-
garbage bags or debris, if feasible lected. Some general categories used by researchers
Scales: To weigh debris: were listed in Table 2 (Chapter 1). Few researchers
Spring scales of three 0-300 g, 0-2 kg, 0-20 kg had defined landbased and vessel-source debris cat-
sizes are adequate egories a priori.
jars with labels To sample small debris Examples of forms used to collect data are in Ap-
Camera and film To photograph study pendix Figures, A-D (this chapter). Because of the
area; photograph differences in survey objectives, no generic form can
entangled organisms be suggested for all studies. Figure 7 contains a sug-
Prismatic compass To fix locations in gested template that can be adapted for specific
remote sites objectives. As noted in Chapter 1, we recommend the
A budget of $500-1,000 should cover the basic approach used by the Center for Marine Conserva-
equipment. tion whereby subcategories describing function/
manufactured use are grouped under a set of mate-
At least two trained people should participate in rial-type categories.
each survey; one to process debris and the other to
record data. The alternative is to use one person to
process debris and record the data into a tape re-
corder. In Alaska, $25,000 per year is the total budget Pilot Studies
for beach surveys run by the National Marine Fisher-
ies Service Auke Bay Laboratory (S. Johnson, NMFS Many of the factors listed in the third guideline (His-
Auke Bay Laboratory, AK, pers. commun. April torical Information) are often not known before a
1991). That budget allows two people to carry out a study starts. But if the study is long-term, this infor-
total of 10 surveys on two or three islands in south- mation is necessary to determine a successful design.
east Alaska. Some information, such as location of dumping sites,
The major expenses of landbased surveys will be may be available from governmental agencies or uni-
transportation to remote sites and personnel ex- versities, but other information may not. Pilot studies
penses (including salary, lodging, and meals). The will be important if nothing is known about the
use of volunteers to collect data will reduce person- oceanographic influences to the beaches as well as

CHAPTER 4: Beach Surveysfor SmaU to Large Debris Items 41

Observers: Date: Year/Month/Date

Sampling Unit e.g., a beach or transect. Time: Start

End
Location (lat/long): This is important for beaches not Tide:
permanently marked.

Sampling Unit Width Both are important for Weather.
transects:

Length Length is important for
beaches. Beginning and
ending points should be
well-marked with flagging
and stakes.

Beach Condition: Slope

Substrate

Comments:

Item Number Weight Dimensions (length/width)

This can either be a
list like that of the
Center for Marine
Conservation (1991)
and Cole et at.
(1990) or just
general categories
like those of FAO
(1989). [See forms
in Appendix 41

Figure 7
Suggested template for beach survey forms.

beach characteristics. For example, Johnson (1989) plastic and glass bottles. Pilot studies will be neces-
marked trawl net fragments to investigate seasonal sary to develop techniques to differentiate between
turnover rates on Alaska beaches. As another ex- vessel-source and landbased debris for specific geo-
ample, the Tidy Britain Group measured turnover graphical areas, particularly if plastic containers are
rate (shore retention rate) of containers by marking the debris type of interest (Dixon and Cooke 1977).

42 NOAA Technical Report NMFS 108: Marine Debris Survey Manual
Quality Assurance Program sampling units are chosen; how sampling units are
marked (if permanent units); how surveys are carried
Information concerning quality assurance program out in detail; how data are recorded, stored, and
plans for beach surveys is scarce. For studies in which checked for errors; and how analyses are performed.
data are collected by volunteers, a training session to
guarantee the quality of the data is imperafive. Com-
mon problems with data collected by volunteers are Field Sampling Designs
an overrepresentation of plastic and-an underestima-
tion of vessel-source debris (T. Dixon, unpubl. data). Beach surveys have been developed featuring many
Merrell's (1985) detailed beach survey is the closest different objectives and field sampling designs. Key
thing to instructions for a quality assurance program features of selected studies are listed in Table 7. Few
plan that has been published to date. surveys have been designed for trend assessment
Generally, the following details should be included (Merrell 1985, FA0 1989, Cole et al. 1990). The field
in a quality assurance program plan for on-land sur- designs for three trend assessment studies are pre-
veys: objective of study; choice of population of sented in Table 8.
interest and any restrictions; choice of debris items to Differences in field sampling designs are attribut-
be studied; details of the sampling design; definition able to differences in objectives and the types of
of the sampling unit; sample size calculations; how debris common to the areas. Merrell (1985) empha-

Table 7
Key features of selected beach surveys for marine debris.

Sampling

Reference Purpose Variable Unit Interval Notes

Vauk and (1) Characterize debris Accumulation 60-m length Every Sandy beach;
Schrey (1987) on beach; (2) Use as rate of beach 3 days area not used for
an indication of (high tide recreation; wind
problems at sea line) direction monitored.

Willoughby Characterize litter on Standing 50 M of Once Systematic sampling
(1986) islands stock high tide of entire high
line tide line of
representative islands
in archipelago: .
50-250 m between transects.

Cole et al. (1) Characterize debris Standing 3-5 sections 3-4 Variable substrate;
(1990); Manski on beaches; (2) monitor stock in some atloo-1000 times a recreational use;
et al. (1991) at-sea debris areas;accu- m of 42 year different debris
mulation rate beachesin typesrecorded
in others 8 parks

Merrell (1985) Monitor for entangling Standing 11 beaches Annually Beaches remote
debris stock (I km each) from populated
surveyed areas; sand and
boulder/cobble
substrate, moderate
to steep slopes

Henderson et Characterize fishing net Not stated Allbeaches Regularly
al. (1987); washed ashore on on six atolls patrolled
Henderson beaches
(1988)

Manville Il Characterize plastic Standing 25 beaches Once Surveys opportunistic,
(1990) debris stock beaches randomly
chosen; outer Aleutian
Islands surveyed

CHAPTER 4: Beach Surveysfor Small to Large Debris Items 43

Table 7 (continued)

Sampling

Reference Purpose Variable Unit Interval Notes

Duronslet et al. Document types and Standing 3-4 transects Monthly Part of a study
(1991) amounts of man-made stock in 3.3 m wide in dealing with strandings
debris some areas; 6 zones of sea turtles; some
accumulation transects randomly
rate in others chosen; other, fixed;
length was variable,
depending on
first storm line and
tide stage

FACI (1989) Pilot monitoring Standing 1-6 beaches Monthly All transects run
program for marine- stock in 4 of in 5 coun- in 4 of 5 from low water line
based litter 5 countries; tries; 2-11 countries; to back end of
accumulation transects per weekly/ beach; substrate
rate in I of 5 beach;I-to biweekly variable; varying
countries I 00-m wide in I of 5 recreational use;
transects countries choice of beaches
restricted by funding
and availability of
labs to carry out
monitoring

Lindstedt and Characterize debris Standing 6 beaches; Quarterly 4 beaches had
Holmes (1989) on beaches stock 3 established high recreational
transects; 50 use; 2 had low use
m in length,
10-77 m in
width

Gregory et al. Characterize debris Not stated All accessible Not
(1984) on beaches beaches and stated
low rocky
shores
searched

Golik and 1. Evaluate the quan- Standing 6 beaches; Monthly Counted all litter
Gertner (1990) tity of coastal litter, stock 5-8 random greater than 2 cm;
2. determine the relation- transects; beaches differ in
ship between beach each transect morphology, substrate,
morphology or use 5 m wide, and use
and litter, 3. identify length from
litter as landbased or water line to
sea-based back of beach
(start of
vegetation)

Center for Characterize debris on Accumulation Beaches Yearly Volunteer program,
Environmental beaches rate along Texas (some had return of data cards
Education shore additional is voluntary
(1988) cleanups)

Cundell (1973) Determine whether Accumulation I beach Once Beach selected
debris on beaches is rate of due to location at
marine-based or plastics entrance of bay and
landbased its northerly aspect

44 NOAA Technical Report NMFS 108: Marine Debris Survey Manual

Table 7 (continued)

Sampling

Reference Purpose Variable Unit Interval Notes

Dixon and Monitor discarded Accumulation 3 sectors Periods Variable beach
Cooke(1977) plastic containers in rate (1.6 km ea.) of on- types, high fre-
the marine environment of shoreline shore quency of onshore
with N/S winds winds, lack of
orientation human disturbance
during winter months,
in close proximity to
Straits of Dover

Slip and Burton Identify types and ori- Standing Entire coast- Once First survey-litter
(1990) gins of litter on beaches stock line (94 km) of removed, wooden
Macquarie objects not included
Island owing to presence
of old shipwrecks and
past sealing activities

Caulton and Monitor maritime litter Standing 300 m length Systematic; Surveyed just after
Mocogni (1987) stock of beach weekly high tide; heavy
divided into for 6 recreational use
three 100 rn months
areas; 5
parallel-strip
transects, 1 m
wide at 5 m
intervals for
each area

Wong et al. Baseline measurements Not stated Selected Annual
(1976) of pollution in marine stretches of
environment beach to
collect plastic

Scott Determine whether Standing 2 areas 100 yds Used inaccessible
(1972,1975) plastic debris was land- stock (1 -100 yds, once;50 stretches of rocky
based or marine-based 1-50 yds) yds twice shore
(3 years
apart)

Wilber (1987) Monitor amount of Standing 30 cm X Not
plastic in marine stock 30 cm stated
environment quadrats

Gregory (1977, Monitor amount of Standing I-m-wide Not
1978, a and b, plastic pellets on stock transects stated
1983) beaches along high
tide line

Shiber Monitor amount of Standing Hand Not
(1979,1982) plastic pellets on stock collections stated
beaches on multiple
beaches

sized very large debris, especially trawl web, which terested in plastic containers (large debris), sub- sam-
would be difficult to subsample in transects. FAO pling was used along with surveys of the high tide
(1989), however, was interested in medium and large line for plastic containers. Differences in the designs
debris; thus, subsampling was used. Because the Tidy are numerous; Merrell (1985) emphasized the actual
Britain Group (see Case Studies, this chapter) was in- field work, FAO (1989) gave more guidance as to sta-

CHAPTER 4: Beach Surveysfor SmaU to Large Debris Items 45

Table 8
Outline of three field surveys for trend assessment of small to large beach debris items.

Alaska Mediterranean North Sea
(Merrell 1985) Sea (FAO 1989) (Tidy Britain Group Case Study)

Total number of beaches to survey not At least 2 beaches that differ in Study site chosen according to following characteristics:
indicated; each beach should be as far morphology, sedimentology, 9 Typical examples of open coastal locations or relation
as possible from urban areas; each beach and type of use. Beaches should to semi-enclosed or oceanic water masses.
should have at least I km of similar not be regularly cleaned. * Presence of sandy beaches with shallow slope, well-
substrate and slope. Beaches should defined backshore zone, and accumulated debris.
have moderate-to-steep slope, sand or * Surface currents that run toward or parallel to the
gravel substrate, and be exposed to the study site.
open ocean. Beaches should have - Situated in close proximity to major shipping routes
accumulated debris present. or fishing grounds or both.

Within the area of the population of interest, a minimum
of 40 sampling units (beaches) chosen in the following
way:

* Divide the area of the population of interest into at
least 8 subareas that contain beaches with the required
characteristics; the 8 subareas should be distributed
throughout the area of the population of interest.
e Within each subarea, identify all potential beaches
from maps. Delete beaches that are inaccessible and that
have major sources of obviously land-generated debris
(camping, bathing areas).
- Randomly select 5 beaches from all the potential
beaches.

1 km of beach surveyed from water's Per beach: At least 4 transects Per beach: Identify and photograph permanent features.
edge to seaward limit of terrestrial of 5-m width; each transect Three transects of 5-m randomly chosen. Length of
vegetation at the upper limit of normal randomly chosen. Length of transect: low water line to high water marks plus 30-m into
high tides. Beach permanently marked transect: low water line to back the foredunes of the backshore zone when present. Walk
and photographed. end of the beach. I km parallel to strandline, tabulate all containers, and
collect representative specimens.
Standing stock (#/km). Standing stock (# or weight/ Standing stock (g/ml). Per transect: Tabulate and weigh
Count all items greater than 5 mm in meter). Collect and weigh all all materials <15 kg, excluding timber or driftwood. Age
size; tabulate and estimate weights of visible persistent litter greater and note geographical origin and original contents of
partially buried net fragments. than 1-2 cm. all containers. Note the distribution of containers along
the transect line.

Sampling frequency not Sampling frequency: at least Sampling frequency: Between October and April during
stated. monthly. or after spells of persistent onshore winds (>96 h) with
velocities >8 m/s.

Each item is listed. Nine categories used. Each item listed.
Recommended log form. Recommended log form. Recommended log form.

tistical details, and the Tidy Britain Group gave ex- Baseline surveys with the objective of estimating
treme detail in choosing beaches. The studies of the amount of total debris on the beaches of a geo-
Merrell (1985) and the Tidy Britain Group are dis- graphical area at a particular time pose a classical
cussed further in the Case Studies portion of this sampling problem. With this objective, a new set of
chapter. beaches to sample that fit the conditions listed previ-
The framework proposed by Ribic and Johnson ously should be randomly chosen each time a survey
(1990) for developing beach survey sampling designs is made. The sample size estimation (i.e., number of
is recommended with some modifications to address beaches to survey) is based on equations taking into
baseline and trend assessment studies (Table 9). This account spatial correlation (correlation between
suggested framework can be used for any type of ma- beaches). More complicated approaches are available
rine debris. Design differences are due to variable that take into account time and space correlations
study objectives. (Gilbert 1987).

46 NOAA Technical Report NMFS 108: Marine Debris Survey Manual

Table 9
Framework detailing alternative sampling design for beach surveys.

Study design
component Baseline Trend assessment
Objective Estimate the amount of debris on the beaches of a Estimate the trend in the amount of marine-
specified geographical area during a specified time. based debris on beaches of a specified area
during a specified time period as an indicator
of oceanic trends.

Variable Standing stock (mean #/km) Accumulation rate (#, weight/time interval) or
standing stock (#/km)

Sampling un it Beach Beach

Field design Stratified random sampling of beaches: Random choice of indicator beaches:
To ensure geographic representation, follow a
Beaches: Appropriate beach topography and "stratified" approach (Dixon and Dixon 1981;
morphology, no access restrictions. Ribic 1991.)
* Stratify on amount of debris on the beaches (low,
medium, high) * Indicator beaches: accumulators of large
e Number of beaches dependent on variability amounts of marine-based debris, no access
of debris on beaches, specified precision of estimate, restrictions, appropriate beach topography and
correlation between beaches. morphology.
Cost of survey built into some sample size equations. - Number of beaches and sampling interval
depend on detection of specified change over
time periods using specified model, also
influenced by resource allocation.

Without with Without
With subsampling subsampling subsampling subsampling
of beach of beach of beach of beach

FAO (1989): Merrell (1985): Golik (1982): Clean entire beach
e At least 4 randomly 9 Count all debris on entire Systematically placed
placed transecLs per beach from water's edge to transects: 100-m in
beach. mean high tide line. length, from water's
* Each transect 5-m in edge to mean high tide
width, length is from Tidy Britain Group Case line, cleaned of all
water's edge to mean Study: debris.
high tide. * Count all debris on entire
beach from water's edge to
30-m into vegetation.

Alternative:
Gregory (1978a): Sample
high tide line; transects:
3-m width, 25-m length
Measure variable of choice (e.g., counts, weights) Resurvey at established time periods; clean
each time; count and weight chosen
Estimate mean, variance, and confidence intervals marine-based debris, age containers
for standing stock
Look at changes over time
Generalize to specific geographical area for
specified time frame Generalize to oceanic conditions

Common survey sampling techniques like stratifi- having low, medium, -_ or high amounts of debris. The
cation (Gilbert 1987) can be used to obtain more number of sampling units to survey from each of
precise debris estimates. For example, beaches that these categories could be based on the proportion of
have the characteristics defined earlier in this chap- beaches falling into each of the categories. A discus-
ter (Historical Information) could be classified as sion of the usefulness of stratification can be found

ULAITER 4: Beach Surveysfor SmaU to Large Debris Iterm 47

in Gilbert (1987). In addition, differences in beach four quarterly surveys will not equal the amount
lengths can be accounted for using probability pro- found during one annual survey. This is because the
portional to length of the beach in the choice of amount of debris found on the beach at any one time
survey units. - is the result of the dynamic process of beach deposi-
With trend assessment, individual beaches are tion, burial, and loss (e.g., a piece of debris washed
monitored over time. Sample frequency per beach up on the beach in April may or may not be on the
will be high, particularly if a time series approach is beach in September).
taken (Ribic 1991). The effects of environmental For baseline studies, confidence intervals can be
variables (such as wind, tide, etc.) can be controlled used for comparisons between areas and years. Data
through selection of candidate beaches and sampling are commonly summarized with pie charts and histo-
period. The actual beaches that would be followed grams of debris types. For studies where sampling
over time would be chosen randomly from the group units are randomly chosen, hypotheses can be tested
of candidate beaches. Accumulation rate is the more with a variety of nonparametric and parametric tests.
important variable of interest, though, of course, Properly planned trend assessment studies can be
standing stock can be measured. For medium to very analyzed in a time-series framework. Lettenmaier
large debris surveys, all debris items can be counted (1978) suggested that monthly samples taken for 5
or, alternatively, the survey may focus on certain indi- years will give a minimum sample size for time-series
cator items. If vessel-source debris is of interest, care analysis. All the assumptions behind this approach
must be taken to determine any landbased fraction of will have to be checked (Brockwell'and Davis 1989).
such debris. An important difference between sam- An alternative to time-series analysis is the nonpara-
pling large debris and small debris is the use of metric analysis for trend (Hirsch et al. 1982; Gilbert
subsampling; typically, there is too much small and 1987). However, for a short time series (on the order
medium debris to count or collect in any reasonable of 5 years), these techniques may have low power
amount of time. (Hirsch and Slack 1984). Another alternative is to
use a within-subjects analysis of variance model
(Keppel 1982) where beaches are the "subjects" and
Analytical Procedures the time period when measurements were taken are
the "treatments." Various adjustments can be made
Caution must be used in comparing sites if standard- for violations of the typical analysis of variance as-
ized field techniques are not used. Debris sumptions (Keppel 1982). This is basically the
accumulation rate and standing stock measurements alternative considered by Ribic and Johnson (1990)
are not comparable; that is, if one study measures the although they use a randomized block in time design
accumulation rate and another measures standing where beaches were blocks. The simplest approach-
stock, then the two sites cannot be compared. If plas- a "before and after" comparison of specific
tic containers are used as indicators of vessel-source beaches-is also possible (see Case Study: The Tidy
debris, analysis of container ages may directly mea- Britain Group, this chapter). Because of the potential
sure accumulation rate in the marine environment, complications in the analysis of trend assessment
avoiding the problems of comparing indirect beach- data, a statistician should be involved from the onset
based measures. of planning the study through the analysis stage.
To compare accumulation rates between different
studies, the sampling frequency must be the same.
For example, if one study measures accumulation Case Studies
rate for 3 months and another for 6 months, the re-
sults are not comparable. In order to compare the Two case studies are presented to demonstrate the
sites, the rates have to be put on the same time scale different ways the marine debris problem can be
(e.g., the amount of debris per month). This makes studied. In both cases, concerns about aesthetics ini-
the assumption that the rate of accumulation within tiated the work. The field survey designs were
the sampling interval is uniform in all the studies influenced by geographical variables and the major
(this assumption may not be justified). debris items found in areas. Both studies standard-
In measuring both the accumulation rate and ized sampling protocols early on and continued to
standing stock, it is inappropriate to add together the follow basic protocols throughout the studies. Finally,
amounts of debris found at a site during several sam- both studies have changed and refined their objec-
pling efforts at short intervals to determine total tives as more was learned about their particular
amount of debris over a longer time interval. As an problems and as control measures (e.g., MARPOL
example, the sum of all debris found during each of Annex V) were implemented.

48 NOAA Technical Report NMFS, 108: Marine Debris Survey Manual

Table 10
Outline and chronology for Alaska beach surveys.

1972-1974 Surveys started at Amchitka Island during course of an unrelated project
Predominant debris was related to fishing industry
Standardized methodology established (outlined in Table 8)
Standing stock primary variable

1974-1980s Decrease in fishing effort off Alaska

1982 Original Amchitka Island beaches resurveyed to assess prediction of decrease in amount of debris on Amchitka
Island
Focus on debris related to fishing industry
Standing stock primary variable

Early 1980s Entanglement in fishing gear identified as possible causes of northern fur seal decline

1984 Seven islands in southeastern Alaska surveyed for entangling debris; prediction is that there is less fishing in
southeastern Alaska and, therefore, there should be less debris
Standing stock primary variable

1985 Sampling design emphasizes entangling debris
Two islands of original seven in southeastern Alaska the focus
Aleutian Islands (including Amchitka Island) surveyed

Standing stock still primary variable
Design modified to assess accumulation rate
Beach dynamics of plastic debris investigated

1986 Focus on two islands in southeastern Alaska
Standing stock and accumulation rate being measured on different beaches
Beach dynamics--emphasis on trawl web

Mid-1980s Education efforts begun to discourage disposal of fishing nets in North Pacific

1989 Two islands in southeastern Alaska the focus
First island: surveyed twice, accumulation rate measured
Second island: surveyed annually, standing stock measured

Emphasis on trawl web dynamics on second island to assess impact of MARPOL Annex V

XIAUOL Annex V enters into force 31 December 1988

Alaska Beach Surveys Gulf of Alaska, the other the Bering Sea) away from
major sources of landbased debris. Most debris
Alaska beach surveys of marine debris are being done proved to be vessel-source (e.g., fishing nets, floats,
by researchers within the National Marine Fisheries cargo-related ropes, strapping bands). In the absence
Service. The chronology of the Alaska beach survey of a standard survey methodology, Merrell developed
studies is presented in Table 10. The following infor- an approach that is still appropriate to this day (see
mation was derived from Merrell (1980, 1984, 1985), Table 8 and Merrell 1985). The sampling unit was an
Merrell and Johnson (1987), Johnson (1988; 1989; entire beach because the beaches on Amchitka Island
1990, a and b), andjohnson and Merrell (1988). are relatively small and discrete (i.e., bounded by
The initial work on beach surveys in 1972 was rocky headlands.
started by Theodore Merrell (NOAA, National Ma- In 1984, a temporary shift was made to sample
rine Fisheries Service, Auke Bay, AK) on Amchitka southeast Alaska beaches. Merrell (1985) did not pro-
Island in the western Aleutian Islands of Alaska. This vide the rationale for selecting those beaches;
work was started as an opportunistic study during however, they were generally known to collect debris.
Merrell's stay on the island for another project. The Because many outer coast beaches in southeast
choice was fortuitous because Amchitka Island is situ- Alaska are steep bedrock and not suitable for debris
ated in the North Pacific Ocean (one side facing the surveys, the selected southeast Alaska survey beaches

CILIPTER 4: Beach Surveysfor Small to Large Debris Item 49

were among the few beaches with suitable substrate, ways of handling debris and -waste. This has been
slope, and access characteristics (S. Johnson, National achieved by a series of educational and public in-
Marine Fisheries Service, Auke Bay Laboratory, AK, volvement programs directed toward the entire
pers. commun. April 1991). In 1985, the emphasis community. A chronology of the beach survey studies
shifted to entanglement debris (fishing nets and strap- undertaken by the Tidy Britain Group is presented in
ping bands) as concern increased about the impact of Table 11.
entanglement of marine animals in such debris. In I The Tidy Britain Group supports a Marine Litter
1985, focus on trawl web started and, in 1989, the re- Research Program as an integral part of its approach
searchers shifted their studies to use trawl web as an towards litter abatement. The Program was estab-
indicator of the impact of MARPOL Annex V. lished in 1973 in response to numerous complaints
Thus, there has been a change in measurement about increasing quantities of all types of litter on the
variables as the survey has evolved. Standing stock United Kingdom's beaches, frequent reports of inju-
(both type and and amount of debris) was of initial ries sustained by bathers owing to encounters with
interest. Beach dynamics of debris (primarily trawl broken glass and sharp pieces of metal, and general
web) and measurements of accumulation rates have dissatisfaction with declining aesthetic conditions of
become increasingly important. Studies of the move- primary bathing beaches.
ment and fate of trawl web and other plastics have-
been conducted by mark-recapture studies. Tagging
or removal of trawl web from the beaches is now rou- The Development of Beach Survey Techniques and
tine, as is the removal of rope, gillnets, and strapping Analytical Methods for the Surveillance of Marine
(S. Johnson, National Marine Fisheries Service, Auke Litter-In the early 1970s, few references in the lit-
Bay Laboratory, AK, pers. commun. April 1991). In erature described systematic studies of marine litter.
terms of sizes of debris, the surveys emphasized large Consequently, the research program first sought to
and very large debris. develop suitable beach survey techniques and analyti-
During this series of surveys, objectives have been cal methods. Preliminary observations at several
refined and the sampling design, field measure- locations around the United Kingdom coastline sug-
ments, and data analyses altered accordingly. Despite gested that glass, paperboard, metal, and plastic
refinements, however, the basic field methodology containers were the main components of marine lit-
has remained unchanged. The data collection by one ter. Plastic containers were therefore selected as the
person and his training of other surveyors reduced debris type of primary interest. The initial studies
the problems of inter-observer variability and helped had the following objectives:
insure continuity between studies. This will make 0 De Itermine the main kinds of containers and their
comparisons between previous surveys valid. The in- relative abundance;
terest in a specific type of debris-entangling 0 identify the range of geographical origins;
debris-is also a strong point because of the way this 0 quantify container retention rates on different
interest was used to focus the research. beach types; and
0 assess the persistence of plastic debris in the ma-
The Tidy Britain Group rine environment.
The initial observations were undertaken on a 4.8-
The following study is based on Dixon and Cooke kin stretch of Dover Strait coastline. This was chosen
(1977), Dixon and Dixon (1980, 1981a; 1983), and as a long-term reference area because it included a
Dixon and Hawksley (1980). variety of beach types, had a relatively high frequency
The Tidy Britain Group is the United Kingdom's of onshore winds during all seasons, was in close
national litter abatement agency. It has a broad mem- proximity to major shipping lanes, and recreational
bership including national and local governmental boats operated offshore throughout the year. In addi-
agencies, industry, commerce, and voluntary organi- tion, there was evidence of litter originating from
zations. The Group functions primarily as an advisory landbased sources, including beach users, sewage
body, but also offers a wide range of practical pro- outfalls, rivers, and a nearby coastal landfill site.
grams dealing with litter problems on land, or more Use of Containers as Indicators-The observations in
recently, at sea. the reference area suggested that plastic containers
The Group's approach to deal with litter problems were the most common type of debris. They were
involves tackling two main causes: first, the attitudes deposited along high tide lines at a rate up to 80 km-1
and behavior of people towards littering and the en- day-', and were therefore suitable indicators, identify-
vironment; and second, the correct and incorrect ing the major sources and subsequent movements of

50 NOAA Technical Report NNOS 108: Marine Debris Survey Manual

Table 11
Chronology for Tidy Britain Group surveys.

1973-1977 Marine Litter Research Programme established in response to apparent increasing quantities of litter on the United
Kingdom's beaches

Purpose-Generate systematic data showing qualitatively and quantitatively the nature and scope of the problem

Primary components of marine debris identified as containers

Single sampling unit chosen as a long-term reference area; it consists of varying beach types, high frequency of
onshore winds during all seasons, known high density of shipping operating offshore; also accumulates landbased
debris

Containers collected, contents, ages, and manufacturers identified

Retention rates assessed, beach dynamic studies undertaken with mark-recapture experiments

1978-1980 Population of interest redefined to a large area; to assess large-scale trends, survey design further developed and
reassessed to identify limitations (outlined in Table 8)

MARPOL Annex V becoming an issue

Field sampling design changed to include the following:
� single subsample per site
� increased container samples
� distribution of containers on beach omitted

1980 on New program to assess impact of MARPOL Annex V

Two studies developed:
I .Source-specific beach debris surveys to detect changes in at-sea waste disposal due to MARPOL Annex V; arranged
surveys as "before" and "after" entry into force of the Annex at same sampling units and times of year:
- Emphasis on plastic containers (geographical origin, ages, original contents)
- 185 sampling units (not all fulfill previous substrate requirements)
2. Assessments of the provision and use of reception facilities in ports and marinas for disposal of ships' garbage
ashore.

1988 MARPOL Annex V enters into force on 31 December

vessel-source debris. Consequently, technical support posited by the wind. Glass and metal items, in con-
networks previously established with packaging and trast, were more likely to become buried "in situ," in
product manufacturers, especially plastic bottle mak- close proximity to high tide lines, and later exposed
ers, were considerably extended on a global basis. A by wave action. All types of litter were found mixed
detailed database was compiled for the most fre- with algal materials. Longshore movements by wind,
quently observed containers, incorporating data on currents, and wave action were also evident, often
packaging histories. The data base continues to be causing litter to be washed from beaches, transported
updated regularly. seawards, and deposited elsewhere.
Beach Type and Litter Assessments-Given the Optimum Sampling Period and Frequency-An opti-
marked relationship between beach form and reten- mum sampling period was apparent for assessing
tion rates of containers found in the reference area, litter originating from sources other than beach us '-
sandy beaches were identified for more detailed ob- ers. This period was during or immediately after
servations. Further mark-recapture studies were spells of persistent onshore winds (>96 h), with ve-
undertaken, providing a clear understanding of the locities >8 m s-1, between October and April. Tidal
processes by which materials were removed from the stage did not affect this optimum sampling period.
beaches and later deposited elsewhere. Containers Daily inspections of foreshore high tide lines and
and other types of plastic litter were often recovered backshore zones enabled collection of accurate data
from the backshore zone, where they had been de- on the types and quantities of deposited litter.

CHAPTER 4: Beach Surveysfor Smail to Large Debris Items 51
In contrast, assessments during the remainder of * the presence of sandy beaches, with shallowly in-
the year, particularly during or at the end of the bath- clined beach face gradients and well-defined
ing season, generally identified a small residual backshore zones, on which marine debris was
proportion of litter deposited earlier. This was com- known to accumulate in the short term
posed largely of plastics mixed with the more recent 0 exposure to surface currents running towards or
discards of beach users. parallel with the coastline, and the relatively high
frequencies of onshore winds, favoring the strand-
Large-Scale Beach Litter Surveys on the Shores of the ing of debris
English Channel, North Sea, and North Atlantic 0 close proximity to major shipping routes or fish-
Ocean-Between August 1978 and July 1980, new ing grounds or both
sampling units were chosen from the beaches of Careful consideration was given to the spatial dis-
Cherbourg Peninsula, France, west Jutland in Den- tribution of the sampling units throughout the
mark, Portugal, and the Western Isles of Scotland. A population of interest. However, from a practical
total of 170 sampling units were examined. viewpoint, the duration of each survey was largely de-
The goals of this expanded program of beach litter termined by the level of funding available.
surveys were as follows: Therefore, it was decided that a minimum of 40 sam-
e Assess large-scale trends in the composition, geo- pling units should be examined with a team of
graphical origin, distribution, and persistence of between two and four observers. A method of multi-
marine debris in the coastal and oceanic waters of stage stratified random sampling was employed in
western Europe; and selecting the sampling units. The selection process is
0 develop a standardized method for assessing ves- detailed in Table 8.
sel-source debris from beach surveys on the On arrival at each sampling unit, permanent fea-
shores of different water masses, and subsequently tures or fixed points were identified on or near the
identify any limitations. shoreline and photographed. The locations of
The survey program sought to produce data of suffi- transect lines were fixed using two-digit random
cient accuracy drawn from a sufficiently large numbers (from tables) that represented a linear dis-
geographical area such that results could alert na- tance down the beach from the fixed points.
tional authorities and intergovernmental organ- Each survey on a sampling unit lasted between 14
izations of the consequences to uncontrolled solid and 18 days following spells of onshore winds; two
waste discharges into the marine environment. were completed in March or April, and the remain-
Accordingly, specific survey objectives were as fol- der in July or August. In the case of the latter surveys,
lows: sites where substantial numbers of beach users were
likely to congregate were avoided.
� Identify the major sources and relative distribu- At each sampling unit, subsampling was done using
tions of marine debris in the semi-enclosed and three 5-m strip transects established at right angles to
open ocean waters of western Europe; the shore using measuring tapes and markers. Each
� assess the persistence of plastics and other types transect extended from the water line across the fore-
of solid wastes in these waters; shore to include all visible high tide marks, and an
� document the environmental impacts of marine additional distance of up to 30 in into the foredunes
debris; of the backshore zone, when present.
� adapt the methodologies and techniques devel- The following data were recorded within each
oped in earlier pilot studies for use on a larger transect.
scale; 0 total wet weights of the main fabrication materials
� determine the most appropriate statistical meth- and the density of all litter in each foreshore
ods for analyzing data from large-scale beach transect, excluding items >15 kg, and timber or
debris surveys; and driftwood
� improve as possible the survey design including 0 frequency, fabrication materials, geographical ori-
spatial sampling considerations. gins, ages, and original contents of containers*
To achieve these objectives, the sampling units 0 the distribution of containers within the transects
were selected on the basis of the following biophysi- * the distribution of plastic fragments -by their pres-
cal and anthropogenic factors. ence or absence in I in' plots along the transect
their relative geographical positions on the shores line
of , semi-enclosed and oceanic water masses, as More extensive searches beyond the transect
typical examples of open coastal locations boundaries were employed to collect samples of con-

52 NOAA Technical Report NWS 108: Marine Debris Survey Manual

tainers for dating. For this purpose, at each sampling 0 Generate statistics to detect any major reductions
unit two to four observers walked along the foreshore in the overall quantities or types of litter, notably
parallel to high tide lines for a distance of up to I plastics;
km. All containers were examined and samples col- 0 detect changes in at-sea disposal practices from a
lected. The procedure was then repeated in the detailed study of the container proportion of
foredunes of the backshore zone of each site. beach litter; and
0 relate the data obtained in the first two objectives
The Use of Large-scale Litter Surveys in Programs to trends in the availability on use of port recep-
Designed to Assess the Effectiveness of XLARPOL An- tion facilities.
nex V-In anticipation of the expected entry into
force of MARPOL Annex V by the mid-1980s, the The debris in the respective water masses around
program strategy was revised. In order of priority, the the coastline of the United Kingdom was the popula-
new program strategy had the following objectives. tion of interest. One-hundred eighty-five sampling
units were chosen using a multi-stage stratified ran-
� assist in implementing the Annex by raising pub- dom sampling scheme.
lic awareness of the need to protect the marine Northeast Atlantic Ocean 66 sampling units in
environment;
� evaluate the effectiveness of the various regula- the Western Isles of
tory and other measures designed to reduce Scotland and Cornwall
garbage discharge at sea; Irish Sea 22 sampling units in
� determine the types and amounts of marine litter Cardigan Bay
entering the marine environment from sources North Sea 65 sampling units in
other than ships' discharges. northeast Scotland,
Cleveland, Yorkshire,
At the planning stage of this new program, it was Humberside, and
recognized that at least two complementary data sets Norfolk
were necessary to fulfill the revised objectives: First, English Channel 32 sampling units on
data documenting the availability and use of recep- the Isle of Wight
tion facilities in ports and marinas for the disposal The timing and duration of surveys at each sam-
ashore of shipboard generated wastes, notably plas- pling unit used the procedures described earlier. The
tics; second, beach survey data of specific debris types baseline surveys (i.e., before MARPOL Annex V en-
organized in an appropriate time series to document tered into force) were undertaken between 1980 and
improved waste disposal practices at sea. 1987, usually in March or April, following spells of
In order for a port and marina reception facilities onshore winds and before the commencement of
study (which is presently underway) to be supported, regular summer beach cleaning operations.
a national beach litter survey commenced in 1980. The survey methods were the same as those de-
The aim was to identify any significant long-term scribed previously, with the following exceptions:
changes in the quantities and types of beach litter 0 Total debris wei ht and that of each of the main
originating from ships' discharges on varying geo- 9
graphical scales. items were derived from a single transect at each
Determining the most suitable spatial sampling de- site, and
sign was necessary at the outset for comparative * observations relating to the distributions of con-
purposes. A number of different approaches were tainers and other types of debris across beach
considered; for example, "before" and "after" surveys faces were omitted.
organized separately using different sampling units These minor adjustments were made to increase the
(where "before" and "after" refers to the date number of sampling units that could be surveyed and
MARPOL Annex V went into effect). to ensure the optimum use of resources.
An alternative approach that was chosen was the Data on debris weight within transects will be ana-
use of paired observations, "before" and "after," on lyzed by comparisons of means (before and after
the same sampling units. This sampling design would MARPOL Annex V entered into force), with particu-
control for influences from variables such as beach lar reference to plastic debris categories, and com-
type or topography, hydrographic features of water parisons of spatial distributions and associated rela-
masses, and different amenity values of sampling tive weight differences will be evaluated based on
units. Consequently, the following specific survey ob- data collected before and after MARPOL Annex V
jectives were identified on regional and national entered into force. The major parameters used in
geographical scales: comparing the container data are listed below.

CHAPTER 4: Beach Surveysfor Small to Large Debris Items 53
0 ori ginal contents classifications, the common hazardous materials, which are being found in in-
products in the baseline surveys being typical of creasing quantities on beaches.
those used on ships Over the next 3 years, inputs of debris to the ma-
0 geographical origins, with particular reference to rine environment from nonship sources will be
changes in the relative proportions of samples examined in more detail. Particular attention will be
that are foreign in origin given to landbased debris, such as sewage items, and
0 age classes, with an emphasis on changes in distri- the development of suitable methodologies for mea-
butions following the entry into force of suring riverine discharges of debris to the marine
MARPOL Annex V environment. In addition, joint studies with local au-
0 varying combinations of the above, analyzed re- thorities and other interested groups will examine
gionally and nationally the most effective means of preventing beach users
from discarding debris.
The remaining non-container debris observed in
transects will be compared on the basis of frequen- Summary
cies of occurrence for each type, notably fishing gear
debris. The "after" surveys (surveys made after There are two types of objectives commonly en-
MARPOL Annex V entered into force) will be com- countered in beach surveys: 1) baseline (studies
pleted over the next 5 years, and the results documenting types and amount of debris on
published periodically on a survey-by-survey basis beaches); and 2) trend assessment (long-term
followed by a national review. Reports will be sub- studies to detect changes in overall amounts of
mitted to the appropriate regulatory authori- specific debris types).
ties, governmental departments, and the shipping Different survey designs are necessary to address
industry. the two objectives. Remote sites are ideal for trend
No major difficulties have been encountered to assessment of vessel-source debris because the
date in the organization of surveys or subsequent source of debris is readily identified as vessel-
analysis of data. However, in numerous instances, source.
ideal beaches have not been located and, therefore, Baseline studies with multi-source debris must be
other beaches, primarily sand backed by shingle, able to discriminate between landbased and ves-
cobbles, rocks, and cliffs, have been used. In identify- sel-source debris. Beaches meeting basic criteria
ing long-term trends, allowances have been made for should be randomly chosen each time a survey is
changes in the patterns and densities of offshore made if the focus is on the beach rather than the
shipping operations. beach as an indicator of oceanic conditions.
Following the broad guidelines contained in the Trends in oceanic conditions are best assessed by
revised beach survey program, the future work pro- using indicator beaches of interest and measuring
gram will focus primarily upon compliance the same beaches over time. The indicator
monitoring in connection with MARPOL 73/78 An- beaches are randomly chosen from a list of
nexes III and V, and assessing inputs of debris into beaches meeting basic criteria. Indicator items,
the marine environment from sources other than such as plastic containers or sewage items, can be
ships. As noted above, an examination of the effi- used to assess accumulation rates and long-term
ciency and availability of port reception facilities for trends in occurrence.
ships' garbage is already underway, and findings Well-conceived field sampling designs are impera-
from the beach surveys will be considered with the tive in all cases. Studies with multiple and con-
port reception facility surveys to facilitate an over- flicting objectives, if not recognized as such, can
all appraisal of the effectiveness of the Annex V cause design problems.
regulations. * The cost of a beach survey for trend assessment is
In spring 1991, an extensive study of packaged dan- generally low compared to directly assessing ma-
gerous and hazardous goods recovered on beaches rine debris conditions in the open water (at-sea
commenced. The analysis of the data will generate programs).
statistics concerning the efficacy of ship reporting 0 The two case studies demonstrate different ways
systems for lost cargoes, changes in substance identi- of studying the marine debris problem with beach
fication markings used on packages, and other surveys. The case studies illustrate how geography
aspects of the. regulations contained in MARPOL An- and differences in major marine debris types
nex III and the recently revised IMDG (International greatly influenced the study designs.
Maritime Dangerous goods) Code. Provision will also A statistician should be consulted at the onset of
be made within the survey objectives to compile in- survey planning and be involved through the
formation on medical wastes and other potentially completion of the study.

Surveyors Location FTI Beach No. Length (m) E-ET-T-1
1 2 3 4 S 6 7 8 0
Transect substrate F-I Slope 0 Month Day Year = Cleared
9 10 it 12 1516 1718 19
Fishing Gear Packaging

Trawl web Bottles
Rope 2223 Caps/lids
Gillnet floats Bags
252627
open straps <lml
Closed straps 28 >lm'
31 71 72 73
Othd oil containers cups
3
Trawl floats 35 38 Plastic @=475
Crab pot floats = Styrofoam CD
37 38 76 77 0
Buoy bags = Bowls/utensils/straws Cr
9144 718147 00
Other floats Small pails So
41 4
Crab bait containers = Six-pack yokes
43 44
Troll bait containers 0 Beverage crates
45 84 85
Monofilament gillnet Styrofoam food
4=7 containers El
q 86
multifilament gillnet 4 Bulk liqjuid containers
Monofilament fishing line E3 Be 89
49 Styrofoam packaging
Fish baskets 9
so 51 misc. MY
Troll plugs 5 92 9a
5-Gal oil containers W 5]
53 5
Chemical ampules W55 6 NMFS-Plastic Debris Survey Form
Loops (Rope, etc.) 0
7
E
misc. 5;05 Page of
6061
31

Appendix Figure A
Data form for beach surveys of debris. National Marine Fisheries Service-plastic debris survey form.

Personal Effects Misc. Plastics

Hats/helmets Visqueen/Plastic sheets

Footwear <lm2
Gloves 00 1 >1m, 11511
Smoking accessories 1 Shotgun wads M711
Toys 041 Pipe/tubing PQ 9
Combs/brushes/eyeglasses = Fragments/Pellets
108107 121122123124
Beach whistles Gaskets P25@
Medical Brushes/brooms

Misc. Garbage can/lids
111MI12 Foam/insulation
Misc. PM 19

f33134

CODES

Location Beach No. Transect (m) Substrate slope Cleared

1 Yakutat 1-99 i = 0-100 1 Sand 1 LOW 1 Yes
2 Middleton Is. 2 = 1-200 2 Gravel 2 Moderate 2 No
3 Amchitka Is. 3 = 2-300 3 Boulder 3 Steep 3 Trawl web I
Kruzof Is.
4 4 = 3-400 4 Combination 4 4
5 Kuiu Is. 5 = 4-500 5 5 5
6 Noyes Is. 6 = 5-600 6
7 Suemez Is. 7 = 6-700
8 Admiralty Is. 8 = 7-800
9 Lincoln Is. 9 = 8-900
10 Ralston Is. 10 = 9-1000
11 Kayak Is. 11 = 1000-1100
12 12 = 1100-1200
13 13 = 1200-1300
14 14
15 15

Appendix Figure A (continued)
P3 ;11@4

W1
W

NMFS-PLASTIC MEASUREMENT FORM M 1 4 1 1 1? 1 14 1 16 17 18 19 20
z
Surveyors Location [:1:3 B ... h No. M Length (.j month Day E= Year E= cleared
1 2 34 5 6 7 a 910 1112 1314 1516 Dead ani Is 7 18
Netting Rope
r Fishing li a lClosed st apsl Floats ch erv d
0
0 u m r
0 A
0 -4 0 4) ", 0 0 0
D, Is
0 0
V t) -Y 0 - z :3 Remarks
41e Q 0 43 Id
0 mW V r. a 0 U 0 L)
ri 0 r 0 H
2C P. C: N 9 Q J@ o? @0
@'o 11 '0 i @ 4'
0 ZOI21122123 2425126127 28 2030 1323 35]36]3A38 3940141]4 43 444SI4047I48I49150 S1 52153 Sj55]S65A58 69 8061 62163164 65ISO 67 68 69 70171172 73174 75 76 77 78 9 80 81 82 83 84 85 a was-9790

2
3
4

190 L
11 tv
13
14
is- -H- +PV

18 LIL - - -
19
20
17 :LL
16

Location Position Hattina Type Transact iml i- Animals (entangled/deA& obseryedl
IYaku :t I Buried 1 Trawl web 10-100 1 y a
t /4: 3 1 Trawl 1 Sao bir I Gunshot
2Hiddl ton Is. 2 Snarled on loge 2 monofilamnt qMnst 21-200 2 No l/ 6 2 Spongex 2 Seal/sea lion 2 Si-pack yoke
3lachitka Is 3 1... 3 No Itifilament qillnot 32-300 3 Trawl web only 3/8 3 Buoy bag 3 Porpoise 3 oil
2: 20
4K-..f Is. 4 Buried and snarled 4 Cargo net 43-400 Tr 1 web, gillnet, 13 4 Solid foam 4 Whale 4 Ropelfishinq line
5 ad r
, Pi
5Kuiu 18, 5 5 4-SOO clos trap, rope 5/:: 16 5 C11h 5 Unknowa
6N-ye re 6 6 66-600 5 3/ 19 6 6 staccan 6
1., 25 7 Fish seabirds
7 76-700 6 14" 32 a Fish a bird. A marine mammals
aAdmi alty in. 7-200 1- a a ri
..r, 43 .6 Nation 9 Be.
9Li . n 1:: color B-900 1:1/2. birds 4 me n. mammals
10 Ralston I TwiBted/knotted 10 io So ott.r
19-10000 1.3/4
11 Kayak is 2 Br-idd/kott :d I Clnr 1Z30 -1100 2 1 Japan 11 Bird b....
12 3 Tit'd/kotl as2 Gr 11 11OU 2 u. s' 12 Me= I bones
3 0:1200
13 4 Braidd/k, tl 3 lack 2120 1300 3 Russia 13 Fur saal
14 5 Double trend 4 B us 14 4 Poland 14
is 6 Nylon 5 Yellow is 5 Korea 15
7 As:orted mash size. 6 Red S Taiwan
Beach No. nd color. 7 orange 7 Canada
8 whit.
1-99 9 B@ wn
ixed
1 Act al
Race 12 2 Estimated Page of
I y
2 NO
0
@3 13..

Appendix Figure B
Data form for beach surveys of debris. National Marine Fisheries Service-plastic measurement form.

CHAPTER 4: Beach Surveysfor Small to Large Debris Items 57

BEACH CLEANUP DATA CARD

Thank you for completing this data card. Answer the questions and return to your area coordinator or to the address at the
bottom of this card. This information will be used in the Center for Marine Conservation's National Marine Debris Data Base
and Report to help develop solutions to stopping marine debris.

Name Affiliation

Address Occupation Phone
City State -Zip - M F -Age:
Today's Date: Month: - Day -Year - Name of Coordinator
Location of beach cleaned Nearest city-

How did you hear about the cleanup?

SAFETY TIPS
I .Do not go near any large drurns.
2. Be careful vwith sharp objects.
3. Wear gloves.
4. Stay out of the dune areas.
5. Watch out for snakes.
6. Don't fift anything too heavy.
WE WANT YOU TO BE SAFE

Number of people woriung together on this data card - Estimated distance of beach cleaned Number of bags filled

SOURCES OF DEBRIS. Please list all items with foreign labels (such as plastic bleach bottles from Me>dco) or other marlmgs that indicate the item's
origin (such as cruise line names@ military idendfication or debris with names and/or address of shipping/fteighdng or fishing companies, or oil/gas
e)ploration activities).

SOURCE rTEM FOUND
e--Plw -s-trappim,
A-Mc- -A*teP114q ComP-4

STRANDED AND/OR ENTANGLED ANIMALS (Please describe type of anirnal and type of entangling debris. Be as specific as you can.)

What was the most peculiar item you collected?

Comments

Thank you! PILEASE RE`rURN THIS CARD TO
VOUR AREA COORDINATOR
OR MAIL rr TO:
Center fbir Marine Conservadon
1725 DeSales Stneet, NW
Washington, DC 20036
A Membership Orpnizatkm

AWE Appendix Figure C
t@r N'ftvEPA Data form for beach
Unked S@
Marim E-@ surveys of debris.
Ag@ Ifto
llv
Cmiservation F/ W Center for Marine
ftmxdy Center for Emronmenml E&Lmm, Ea. 19M Printed on recycled paper @ LOA Coatcr far Muim ConwxyWon Conservation-beach
cleanup data card.

58 NOAA Technical Report NMFS 108: Marine Debris Survey Manual

ITEMS COLLEUED
You may find it helpful to work with a buddy as you clean the beach, one of you picking up trash and the other taWng notes. An easyway
to keep track of the items you find is by making tick marks. The box is for total items; see sample below.
E-riVile: TMAL MTAL
M cartons F/-67 cups FzY]
PLASTIC
lbtd Tatm
ork-,` afkMl,
bags: fishiq nets
food bapWWp-- hard hats
trash light Sticks
salt PC-
c#- bag. pipe thread prolector
bottles rope
be,verage. soda sheeting.
bleach, cleaner lorger than 2 feet
rratk(water gal. q& 2 feet or sInciner
oil, lube &pack holders
other bottles strapping bands
buckets straws
caps, lids synnges
cigarette filters Lampon applicators
cigarette lighteis toys
cups, utensils vegetable sacks
clapers "wrice protectiorf'rings
fishiN line other plasiric (speoM
fishing lures, floats
STYROFOAMO
(or oil- 0-tstic fbarn)
buoys packong
Cups pieces
egg cartons plates
last food containers other StyrobarrP (speaFA
mear Days

F= MONO TM UNE
GLASS
botoes/Ws: ftiorescent IOU tubes
be@ tionles light bulbs
food jars pieces
other botdes/jars other gl- ('peom
RUBBER
balloons fires
conclorns other rubber (Specify)

METAL
bottle caps 55 gallon drums:
cans: R&Y
aerosol rm
beverage pieces
food pull tabs
other wre
crab/hsh traps other rnetal Ispecify)
PAPER
b,V newspapersInwazines
cardbwrd peces
cartons plates
cups other paper (specify)
WOOD
(leave ctriftwood on the beach]
crab/lobster traps pal"
crates other wood (spedfy)
lumber pieces

CLOTH

dothiVpieces
Renflernber tO tum the Card over and fig out your name and address and to recOrd soumes and entar4ped willldflfe!

Appendix Figure C
(continued)

CHAPTER 4: Beach Surveysfor SmaU to Large Debris Itents 59

Name Address Ile Tidy Britain Group: blarine Litter Research Progranune
SURVEY FORM

LOCATION OF BEACH (please TYPE OF BEACH (sand, shingle or mud, etc) Office Use Only
state county and nearest town)

DATE OF SURVEY IS THIS PART OF A REGULAR
SURVEY?
YES/No (please underline your answer)

PART 1: Containers Found Please only record containers, If any, from within your survey area.

A. TYPE OF CONTAINER, B. ORIGINAL CONTENTS C. COUNTRY OF ORIGIN D. OTHER MARKINGS SUCH AS
MATERIAL & COLOUR OR NAME OF PRODUCT IMPRINTS AND DATE
CODES, AND C014MENTS

Example
VV T+ aAsc@ELLOIC> 4172

PART 2 Other Litter Tick each time you find an item. If very common,
write in "WIDESPREAD". Please add any general
(Excluding Containers) comments to the reverse of this sheet.

A. PAPER Office J. RAW SEWAGE Office
Use Us
Only Only

B. CARDBOARD K. FISHING NET

C. PLASTIC FRAGMENTS L. FISHING LINE

D. PLASTIC BAGS OR M. ROPE
SHEETING

E. GLASS N. WIRE

F. METAL 0. CLOTHING

G. WOOD P. PAPER OR
PLASTIC CUPS

H. OIL 0. CONFECTIONERY
WRAPPINGS
(Inc. crisp packets)

1. SHOTGUN CARTRIDGES R. OTHER FINDS
(please describe)
(a) plastic case
(b) Paper case

Appendix Figure D
Data form for beach surveys of debis. The Tidy Britain Group---survey form.

60 NOAA Technical Report NNIFS 108: Marine Debris Survey Manual

When Was It Made ? WHERE WAS
Plastic bottles are dated in several IT MADE?
different ways. The moat common is
known as a CLOCK COtE. The year in
which the bottle was made is shown
by the number in the centre of th
circle . 0 - 1980 1 . 1981 , - 1979 0 FRANCE
and so on. The ;i;nth of bott9le r,@q
0 production is shown by the number
of dots on the radiating lines.
In this example, there are seven
dots, the month of production was FRANCE
the 7th month - July. The number
in the centre of the circle is 0, WF FRANCE
so the contsiner was produced in
July 1980.
XDIN
12 This is another type of clock code, %ZUP IRELAND
frequently used outside the United
10 2 Kingdom. The last 2 digits of the
9 78 1@ 3 year of production are found in the STAR IRELAND
centre and an arrow points to a
4 single number on the outside of the
cir le, representing the month. In
7 6 5 thics example, the container was made UPLA PORTUGAL
in February 1978.

SOUTH
Dots and a number in a row are another AFRICA
type of date code. The number at
the end or beginning of the row
represents the last digit of the year
of production. For example, 7 = 1977,
and so on. The dots show the month BELGIUM
in the year when the container was
made. One dot Is removed for each
month, so 12 dots is January, 9 dots
April, and so on. In the example,
the date of production would be IN BELGIUM
November 1980.
BarCodes A
These consist of a series of numbers and parallel vertical P B CANADA
lines. Each product sold in a supermarket will eventually
have its owr code. The first two numbers show the nationality U
of the 'number bank'. The next five numbers show the manu-
facturer and the last five show the product.
00-09 USA + Canada 70 Norway I HOLMIA DENMARK
30-37 France 73 Sweden
40-43 West Germany 76 Switzerland
49 Japan 77 Australia HAUSTRUP DENMARK
PLASTIC
50 UK 80-83 Italy
5 54 Belgium 84 Spain VANGLIA"tA SPAIN
57 Denmark 89 Netherlands UNITED
JEYES KINGDOM
Example: 64 Finland 90-91 Austria
50 00317 00201 3 Paperboard carton. Longlife Kilk. 1 pint. UK.
Please record all numbers on the survey form. UNITED
0 KINGDOM
STOP PRESS!
Please keep a special watch for any pink plastic cylinders UNITED
in your study area. Their shape and dimensions are shown KINGDOM
in the diagram. If you find any, please record
the number found in your study area in Section R (other finds)
of part 2 (other litter). CASCELLOID UNITED
KINGDOM

MONSANTO USA

i
ch
n
rc USA
3 Inches @I I
Appendix Figure D (continued)

Chapter 5

Benthic Surveys
for Large Submerged Debris Items

General Description Population of Interest

Benthic surveys for medium to very large debris The population of interest is the amount of debris on
items involve counting, classifying, and, in some the bottom of a specific area of ocean at a specified
cases, collecting items that have sunk to the'sea floor. time. The population of interest can be as small as
Only small sections of the population of interest will the bottom of a bay or cove, or as large as an ocean
be surveyed, but results can be extrapolated to the basin. Restrictions to the population of interest may
total area. Collection and disposal of debris will de- occur because of floor composition (i.e., substrate
pend on the technique used for collection and the and topography), depth, fauna, flora, and. use pat-
size of the debris. terns of the area (e.g., cargo shipping lanes, fishing
Three survey techniques are discussed in this grounds, or recreational areas).
chapter. Restrictions to the population of interest may be a
1. trawl surveys result of the survey technique more than any other
2. submersible surveys reason. For example, if the survey technique must be
3. diving surveys a trawl survey, then even if the population of interest
was benthic debris in the Great Barrier Reef of Aus-
Trawl surveys, which have been used most often for tralia, the population of interest would have to be
assessing types and amounts of benthic marine debris changed because trawl surveys are not possible in
(Holmstr6m 1975; Jewett 1976; Feder et al. 1978; that area.
Berger and Armistead 1987; Bingel et al. 1987; FA0
1989; June 1990), are the main focus of this chapter.
Because of their high costs and limited availability, Historical Information
remotely operated vehicles (ROVs) and manned
submersibles have been used only in a few studies on The factors that may effect the trawl survey tech-
benthic marine debris (Carr et al. 1985; High 1985). nique are as follows:
Divers using scuba equipment have assessed the ef- 9 depth (e.g., this affects the amount of cable
fects of "ghost fishing" by lost nets (Carr et al. 1985) needed for trawls)
in the North Atlantic Ocean and benthic debris in * slope (e.g., a bottom with a steep slope cannot be
McMurdo Sound, Antarctica (Lenihan et al. 1990). sampled by a trawl)
Note that all of these techniques are still in an ex- * substrate composi Ition (e.g., areas with large rocks
perimental stage. Further, repeated and repeatable and pinnacles can damage the trawl net)
studies are needed. 0 currents (e.g., cross-currents can cause fouling of
the trawl net)
Objectives and Purpose a dump sites (e.g., areas for oceanic dumping by cit-
ies should be avoided)
Benthic surveys can be used to provide information * local fishing practices (e.g., areas with crab or lob-
on the distribution and amount of debris on the ster traps may pose a hazard to trawls)
ocean floor for a specific area. The debris of interest For the other two survey techniques, bottom depth
is usually large to very large and often of a particular and ocean currents will be the most important fac-
type (e.g., lost fishing nets). Few baseline surveys tors. In addition, visibility will be an important factor
have been done for benthic debris (FA0 1989, June (e.g., high plankton density will make underwater
1990). Other studies noted marine debris found navigation and debris counts impractical).
while collecting other benthic samples (Holmstr6m From Guideline 4 to Guideline 8 (see page 62) the
1975; Jewett 1976; Feder et al. 1978; Bingel et al. three survey techniques are quite different. The
1987). manual will first look at the guidelines in relation to

62 NOAA Technical Report NMFS 108: Marine Debris Survey Manual

trawl surveys, since trawls are most often used. Sub- Variables to Consider-
mersible surveys will be second, followed by diving
surveys. Vessel Variabilit)@-Different ships will tow more ef-
fectively than others. This variability is due to the
Trawl Surveys ability of the ship and crew to identically reproduce
tows, stay on track, and handle variability (such as a
crab pot encountered in the middle of a tow).
Field Measurement Net Variability-The size of mesh is the most im-
portant net variable to consider. Tows cannot be com-
Trawl surveys usually seek to measure benthic debris pared unless the meshes are the same, owing to dif-
(number of items, density, or wet/dry weight per ferential escapement of debris (unless all debris
trawl). below a certain size is not considered). Other net-
related variables are the sweep area and volume of
Description-The ship slows to a "trawl speed" (this the net. Acoustic mensuration systems are available
varies with the type of boat, type of net and sea condi- that measure both horizontal and vertical net open-
tions), at which time the trawl is released from the ings constantly during the tow, so differences
stern of the ship, either over a stern wheel or down a between nets used can be quantified. However,
ramp. The net is extended such that the footrope is changes in these variables may cause problems in
on the bottom. The footrope should be a "hard standardizing tows.
bottom" type with very small bobbins or a metal bar. Foo"e Pariabilii@--"Hard bottom" nets should be
Once the trawl net reaches the bottom, the time is used with only the smallest bobbin size. Many bottom
recorded. The ship maintains the trawl speed for the trawl nets are designed to scare fish and do not "dig"
duration of the trawl. After a prespecified time has sufficiently (if at all) into the sea floor (due to large
elapsed, the trawl net is retrieved and the time re- bobbins) to "catch" many kinds of marine debris
corded (Fig. 8). From the length of time on the sea (i.e., pipes, heavy cans, large plastic .sheets).
bottom and average ship speed, the distance trawled Depth Variabilii@-If minor increases in depth oc-
can be measured. Most trawling vessels have LORAN cur during a tow (e.g., 5-10 in on a 100-200 m depth
C or GPS navigation equipment, which will make dis- tow), the net may list off the bottom. Compensations
tance an easy measurement. should be made so that the bottom is followed at all
The same trawling procedure should be repeated times. As with net variability, acoustic and mechanical
at each predetermined sampling point. The speed devices are available that detect contact of the foot-
and time towed should be the same for all trawls. rope with the bottom during the tow.

(6) TraW lawing (E) Releasirigindependeritpice;rable
@ Heaving In @ Danlenoe moving of chute
(D Cable (Z) Taldng in body of travA (diagonal 11M
Q Otter board @ vertical Int
(2) Warp and lead to VAnch EMPLYIngoodend
(D Cable and lead

Stem travAlng

Figure 8
Method of heaving gear aboard a stern trawler (adapted from Garner 1967).

CffAPTER 5. Benthic Surveysfor Large Submerged Debris Items 63

Weather-Rough sea conditions can cause net en- Fisheries Service), with only the shipping and insur-
tanglement, loss of debris from the net, or lift the net ance fees to pay (which will depend on the location
off the sea floor. of the trawl survey).
Measurement Vdriabilil)-The above factors all con- The cost of the ship will depend on the population
tribute to measurement variability. In addition, errors of interest and the size of the net. To charter a ship
in sorting large tows (especially when tows are done about 30-m long to perform a bottom trawl survey in
in cooperation with fishery trawl surveys) can be a the eastern Bering Sea will cost around $3,500-
problem when measuring debris. $4,000/day, excluding fuel. Fuel costs will be about
$300-$350/day when trawling. While in transit, fuel
Data Collection-Data collected with respect to cat- costs will be about $500-$600/day. A 30-day cruise
egories of debris are similar to those in the previous could therefore cost around $110,000, excluding net
chapters (see Chapter 1, Categories of Marine De- and the personnel costs. The cost will increase dra-
bris). Specifically, the following information should matically if the ship is larger. The size of the vessel
be collected at each tow. needed will vary depending upon the following:
date o the size of the net-the larger the net, the more
time tow started horsepower needed to pull the net;
time tow stopped o depth of the area to be trawled-the greater the
exact location (either latitude/longitude of each depth, the more cable needed to reach the bottom.
trawl [e.g., 60'N by 175'W] or distance traveled The ratio of cable to depth is usually considered to
between each trawl (e.g., 5.2 mi from trawl 3)) be 3:1, i.e., a trawl depth of 100 meters needs 300
speed of ship during trawl
weather conditions throughout trawl meters of cable. A 30-in ship can only trawl to
any holes in the net at the end of the trawl about 200 in owing to the amount of cable it can
any trawl period during which the net is not drag- hold; and
ging the bottom distance from shore-a large vessel, 60 in or better,
is needed for high-seas trawling.
Few data forms are available for benthic debris sur-
veys (see Appendix Figures A-C (this chapter). Forms The crew will handle the deploying and retrieving
for other trawl surveys may be adapted for debris of the net in most cases, but the sorting of debris
purposes. Appendix Figures A and B (this chapter) once on deck must be handled by the investigator or
comprises forms used by the National Marine Fish- assisting personnel. For assisting personnel, the cost
eries Service for demersal trawl surveys for bottom is similar to the National Marine Fisheries Service
fish. Figure 9 depicts a suggested data form for trawl marine mammal observer program: the salary will
surveys. run around $3,000/mo plus travel expenses (room
and board may be additional, if not included in the
Material and Personnel-The basic equipment price of the ship charter).
needed for trawl surveys is categorized, by two On the basis of the preceding information,
objectives: the typical price of an entire 30-day cruise, including
the cost of the net, may well be $150,000 or more.
� Sample collection Though the net can be reused, the cost for
a bottom trawl net and all equipment needed for continued trawling still will be quite high. Using a
the deployment of the net log forms for sample vessel of opportunity can reduce costs to those associ-
collection ated with the personnel required for sorting the
� Trawl analysis debris, travel to and from the ship, and room and
large bags (@!103 L) to pile and store debris board.
scales for weighing debris (0-100 kg)
protective gloves for handling samples
data sheets (sample analysis) Quality Assurance Program
The greatest expense is the cost for the ship and Because of the number of trawls likely to be per-
the net. A large net designed for marine debris stud- formed, a quality assurance/ quality control program
ies is about 4 in high and 12 in across and will cost Dlan is one of the most important aspects of the trawl
about $20,000-$30,000 (Net Systems, Inc., Bain-
'Survey. For example, the National Marine Fisheries
bridge Island, WA). It may be possible to obtain nets Service has been performing trawl surveys in the
through an interagency loan (e.g., National Marine Eastern Bering Sea since 1973 and has a detailed

64 NOAA Technical Report NNIFS 108: Marine Debris Survey Manual

Vessel Name Starting Location (lat/long)
Date (Yr/Mo/Day) Ending Location (tat/long)
Gear Depth (in m) Loran Start

Bottom Depth (in m) Loran End
Sea State (Beaufort) Time: Start Trawl (24-hr time)
Time: End Trawl (24-hr time)

Bottom Type (hard, rocky, mud, silt)
Area Use Type (commercial fishing, shipping lane, pleasure craft, etc.)

Net

Net Type
Door Description (width, height in m)
Mesh Size (mm stretched):

Wings Body

Liner Cod End

Footrope:

Bobbin Size (in cm)

Chain Size (in mm)
Bar Size (in meters)

Wireout (in meters)
Average Trawl Net Horizontal Opening (in meters)

Percent of Time with Bottom Contact

Figure 9
Suggested template for trawl survey forms.

document describing exactly how trawl surveys The items listed below should be included in the
should be performed. quality assurance plan.
As noted previously, surveys should use the same 0 specific boundary of the population of interest
trawling procedure at each station. A set procedure is 9 all aspects of the net (e.g., door width, mesh size,
not only good on a day-to-day basis but on a year-to- volume)
year basis as well. For comparisons to be meaningful, * specific points trawled (e.g., accurate location coor-
the trawls must be as similar as possible, and any dif- dinates)
ferences must be noted. A set sampling procedure 0 categories used to sort debris
cannot be done without a strict quality assurance pro- 0 data analysis details
gram plan in effect.

CHAPTER 5. Benthic Surveysfor Large Submerged Debris Items 65

Percent Subsampled
Items Disposal Comments
(Categories) Number Weight (Probable Origin,
Material Composition)

Figure 9 (Continued)

ship and accessory gear description out the population of interest. Certain large areas
crew efficiency at handling trawl will have virtually no debris while other areas will
subsampling and sorting procedures have significant amounts, which is why the historical
information on the population of interest is particu-
larly important.
Field Sampling Designs If a study is meant to give a baseline assessment or
a year-to-year assessment of the type and amount of
As with floating small debris, one cannot assume that benthic debris found in the population.of interest,
the materials will be randomly distributed through- then a regular systematic or random sample will give

66 NOAA Technical Report NMIS 108: Marine Debris Survey Manual
a good assessment of the amount. If, however, a study Note: The "fishing" power correction factor has not
is meant to assess changes in benthic debris over been computed for any debris type, so in most cases
space and time due to legislation, then one should it will be assumed to be 1.0.
concentrate samples in areas where a change is most For an overall mean CPUE and its variance for the
likely to be detected. population of interest of a specific strata:
In either case, multiple techniques for benthic ma- N
rine debris assessment would increase the f(CPUEj)
information about the population of interest. For ex- Mean CPUE= EP-UE i@j
ample, a study might use a trawl survey of an area N
followed up by an ROV survey. N
Y.(CPUEj-CPUE)1
i=1
Analytical Procedures Variance CPUE = SICPUE - N (N- 1)
Many standardized techniques for analyzing biologi- where N = the number of hauls in the area.
cal data can be obtained from systematic or stratified
trawl surveys or both (Doubleday and Rivard 1981)@ To determine the total weight of debris'in an. area
These techniques may be adapted for analyzing or strata, perform the following calculations:
benthic debris survey data. One difference will be the
need to disregard factors correcting for "animal Total weight = WtT = A (CPUE)
movement." Otherwise, the statistical techniques will C
be the same when assessing standing stock. 2
Variance weight = S1Wt @ X2 WME),
There are no "standard" techniques available to C
perform analyses dealing with changing accumula- where A = specified area of interest (e.g., eastern
tion of debris. Research is being conducted on how Bering Sea), and
to evaluate this kind of data and the effects associ-
ated with it Uune 1990). The problem of clustering C = vulnerability; the fraction caught ver-
samples and assessing changes over time must be sus the fraction missed during a trawl
taken into account: sweep.
To assessing the amount of benthic debris.in an Note: As with "fishing" power, vulnerability is assumed
area of a random sample or a strata of a sample, the to be 1.0 for most debris items.
estimate is made using standard fisheries trawl survey If using a stratified random sample, the overall
procedures Uones 1990) in the following manner: mean would be calculated as follows:
For each haul, catch per unit effort (CPUE) is n
measured, usually the amount (either number or Y(Ak CPUEk)
weight of pieces) per area swept (usually in ha or Overall CPUE n
nmi2, e.g., 100 kg/ha). The area swept is the distance Y(Ah)
trawled multiplied by net width. CPUE is calculated
from a trawl survey i for a debris type j as follows: n
S2E-
PUE)
S2 k=1
CPUE Wij Overall CPUE n
Di Fij Pi X(Ah)2
k-1
where Wij = weight (or number) of debris type j on
trawl i (kg), where n = number of strata, and
Di = distance trawled on trawl i (km) Ak = area of each strata.
Pi = effective trawl width for trawl i (km),
and To find the total estimated weight of the area of a
Fij = relative "fishing" power correction fac- stratified sample, the following is performed:
tor (which is how well one ship's n
efficiency at "catching" debris is com- Total weight =WtT= WtTi
pared to another ship's in the same
area) for trawl i in respect to debris 12
type j. Overall S ,t - S 2Wt

CHAPTER 5. Benthic Surveysfor Large Submerged Debris Item 67
Submersible Surveys Data Collection-It may be difficult to categorize
benthic debris accurately. Broad categories will usu-
Field Measurement ally be used. When actual collection by use of
manipulators found on some submersibles is used,
The field measurements of importance are the num- categories can be more narrow. The following data
ber and, if possible, the type and size of the benthic should be collected during each dive.
debris observed in a strip transect, or per dive. o date
Description-The submersible is deployed from a o time the bottom or predetermined depth is
mothership (the exact procedure depends on the r.eached
type of submersible). Upon reaching the bottom, or 0 time the ascent is started (i.e., end of the transect
or course)
more specifically, just above the bottom, the survey 0 exact location (either latitude/ longitude [e.g.,
will start. The submersible should follow a predeter-
mined transect as closely as possible, although this is 60'N by 175'W] at each dive or distance between
often difficult to do (Caddy 1976). The debris is ob- each dive [5.2 miles N from last trawfl)
served, counted, and classified, if possible, although 0 speed of submersible
rarely collected with the submersible manipulators. 0 any changes of distance off bottom
Debris may be observed directly from manned 0 depth
submersibles or via a camera. Unmanned sub- e bottom topography and substrate type
mersibles with camera systems are also available. As * estimated range of visibility
with the trawl surveys, to calculate the area surveyed, 0 number, type, and size of debris observed
the starting and ending time of the transect are re- Noting any biological growth on the debris may
corded and the speed is held as constant as possible. help to determine the age of the debris (Carr et al.
If multiple transects are made to assess the piopula- 1985). Figure 10 is a suggested data form for use with
tion of interest, the procedures should be repeated as a submersible survey.
consistently as possible during each transect.
Material and Personnel -Besides the submersible
Variables to Consider- and,support vessel, the only materials necessary are
the data forms for tabulating the survey. The sub-
Weather-Launching and retrieving of manned mersible and support vessel are usually v Iery
submersibles and large ROVs require a low sea state. expensive. The cost of a submersible will depend a
State 3 is usually the upper limit for a safe launch great deal on the maximum working depth and
(Keller 1977). whether a manned or unmanned submersible i's
Vessel Variability-Different submersibles will have needed. A manned submersible with a maximum
different areas of visibility and different degrees of working depth of 330 in and the neces ary support
mobility. Visibility will depend on the submersible s
lights, the size of the view port, and on the type of vessel will likely cost more than $7 ,000/day excluding
camera lens. Mobility will depend on whether the fuel. An ROV with a maximum working depth of
submersible is tethered or untethered, the type of 2,000 in will run around $2,000/day in addition to
servo propellers used, and7the size of the vehicle. the support vessel at $10,000/day excluding fuel. For
Characteristics of marine debris-The color, size, very deep dives (e.g., 5,000 in) the price could be
shape, extent of encrustation, and degree of burial over $10,000/day for the ROV and $10,000/day for
will affect the sightability of the debris. the support vessel. To have a manned submersible
Turbidity-Very turbid waters will reduce visibility capable of operating at depths over 2,000 in would
to a few cm (Palmer 1977). At great depths (>1000 cost much more than the ROVs. All these previous
in), however, visibility will be fairly clear (about 60 prices include the necessary crews but not necessarily
in), which is due, in part, to the lack of life (Keller an obs.erver. The prices also do not include the trans-
1977). portation of the submersible to the dive location
Measurement Variabilit)-As with trawl surveys, the (shipping will vary greatly with the size of the sub-
variables listed above will limit an investigator's abil- mersible). Obviously, a submersible survey can be
very expensive, unless a submersible is already owned
ity to classify debris as to type or size, or even limit by the institution or can be borrowed through an
the ability to detect debris. Errors in identifying what interagency loan (e.g., NOANs West Coast National
is debris and what is not will, however, depend greatly Underwater Research Centers).
on the training and experience of the observer.

68 NOAA Technical Report NAM 108: Marine Debris Survey Manual

Observer Name Starting Location (lat/long)
Pilot(s) Name(s) Ending Location (lat/long)
Support Vessel Name Submersible's Speed (km/hr)
Submersible Name Time of Descent (24-hr time)
Date (Yr/Mo/Day) Time Observing Begins (24-hr time)
Substrate Composition (mud, rock, etc.) Time Observing Ends (24-hr time)

Time of Ascent (24-hr time)

Bottom Depth (in meters) Estimated Visibility at Bottom (in m)

Depth of Submersible (in m) Estimated Search Width (in m)

Time Object Category Comments
(if identifiable)

Figure 10
Suggested data form for submersible surveys.

Quality Assurance Program Beyond safety, the points noted for the trawl
survey quality assurance program plan also apply
Quality assurance/ quality control program plans go to submersible surveys. Some items to be included
beyond repeatability when dealing with submersibles. in the quality assurance plan (not including safety,
Because of danger to personnel, safety is a prime see Pritzlaff [1979], and excluding the specific guide-
consideration. A good guide to safety in submersible- lines set for the submersible to be used) are as
oriented research can be found in Pritzlaff (1979). follows:

CHAPTER 5. Benthic Surveysfor Large Submerged Debris Items 69
specific boundary of the population of interest When using transects, procedures in sighting sur-
large physical features and substrata on the ocean veys for floating debris (Chapter 2) can be employed.
floor Strip transects will most likely be used. The reason
predetermined course for the submersible for using strip transects instead of a line transect is
actual course followed by the submersible twofold: first, determining distance underwater can
means of determining visibility be difficult at best; and second, the width of view may
categories used for describing (sorting) debris be a meter or less owing to the size of the observa-
Details of data analysis tion port or the size of the lens on the camera. To
observer and pilot experience calculate the density of debris in an area, see Chapter
specifics about the submersible: 2, Analytical Procedures. Strip transects are designed
-manned or unmanned for smaller areas and fewer dives compared to the
-tethered or untethered CPUE method.
-size of observation port
-type. of camera (s) used (including lens)
-manufacturer name of vessel Diving Surveys
-any specific modifications used

Field Measurement
Field Sampling Designs
The field measurement for diving surveys is the
As stated earlier in Trawl Surveys (this chapter) amount (number of items or weight) of benthic de-
benthic debris is not likely to be randomly distrib- bris per quadrat or transect. This method is
uted throughout the population of interest. particularly suitable for assessing medium to very
Depending on the type of study being performed, large debris or when the population of interest is
different sampling schemes should be used. small (e.g., <1 ha) or both.
Submersibles are very expensive, and thus, dive time
must be used cost effectively. Limited search times Description-Two methods used for assessing popula-
may bias results, so declustering techniques (as de- tions of benthic organisms are the quadrat method
scribed by Isaaks and Srivastava 1989) may be and the transect method (Dart and Rainbow 1976;
needed. Hiscock 1987,1989). While these techniques have not
Transect methods will provide accurate data on de- been tested for assessing marine debris, they merit
bris for a small area and will be more appropriate discussion because neither trawl nor submersible sur-
than the procedure in the following discussion on veys can be used in very shallow waters (<5 in).
accumulation studies. The transect method also can
be used to give a continuous description of debris Options-
type and changes in composition throughout the Quadrat-A map of the chosen area has a uniform
population of interest. grid placed over it, often generated by computer. A
CPUE methods will yield an overall estimate of how sample of the squares (blocks) of the grid or the ver-
much debris is on the bottom but will yield little spe- tices of the grid are randomly selected, using a
cific information on the composition of the benthic random number table or a random number genera-
debris over the population of interest. A stratified tor on a computer or calculator. A team of at least
sample would be best, using depth and area usage as two experienced divers with scuba or snorkeling
the stratification variables. equipment sample at the selected sites. A sample grid
is marked on the bottom, using stakes and cord or a
large, fabricated metal or PVC pipe square. The area
Analytical Procedures within the square then is searched meticulously for
all debris. Small pieces are collected for sorting and
Procedures for data analysis are similar to the pro- weighing on shore or on the deck of the boat. The
cedures used for the trawl surveys. The difference is larger pieces are recorded with a@ underwater pad
that instead of strip width being the net width, it is and grease pencil or photographed with an underwa-
the estimated width of the field of vision. Also, in- ter camera or both. Their type and estimated size
stead of the CPUE being kg/ha, the CPUE will be the should be recorded.
amount of debris observed per area (e.g., 50.7.nets/ Transect-To conduct this type of survey, a
ha). This analytical procedure generally will be used weighted line is placed, as taut as possible, on the
when performing many dives in a large area. bottom. A pole (usually made of metal) is taken

70 NOAA Technical Report NMFS 108: Marine Debris Survey Manual
down to the line. As the diver moves along the line indicates the necessary gear for diving surveys in cold
holding the pole parallel to the bottom, one end of water <1.5 kin offshore.
the pole is always touching the line and the pole is 0data sheets
held at at 90" angle to the line. As in the quadrat 0underwater pad and grease pencil
technique, the diver monitors benthic debris occur- enet bag with plastic liner
ring within the area swept by the pole. The 0cord (usually nylon) with weights
procedure may then be repeated along the other side 9stakes and hammer
of the line. I. opole (for transect)
The quadrat or transect procedures are repeated at oflippers
each of the selected blocks, vertices, or points. Again, 0mask
any new quadrats or transects must be the same size esnorkel
and configuration to ensure conformity. ediver knife
Variables to Consider- 0protective gloves
eweights and weight belt
Weather-Getting in and out of the water from 0buoyancy compensator
shore or a boat can be difficult in rough weather. 0regulator
Also, surges can make it difficult to run a transect or 9depth gauge
collect debris. oair tank
Characteristics of marine debris--As with submersible 0camera
surveys, the color, size, shape, extent of encrustation, 0wet or dry suit
and degree of burial of objects will affect their sight- 9boat or, ship (depending on the distance from
ability. shore)
Turbidity---In highly turbid water, visibility will be As with other surveys, the most expensive item will
reduced, thus increasing the likelihood that objects likely be the ship or boat. Vessel costs will increase
will be overlooked.
Equipment Variability--Generally, the larger the air with the distance from shore. In dives close to shore,
tanks, the longer a p .erson can stay submerged and generally <1.5 kin, or in bays or coves, a small boat,
thus a longer transect or a larger quadrat can be such as a double or triple hull Boston Whaler, may be
sampled. suitable. A boat such as this may cost $50/day to rent;
Measurement Variabilit however, it may be possible to borrow one through
.@-Training and experience an interagency loan. A complete set of scuba equip-
will affect the amount of debris seen in any given ment, excluding a wet or dry suit, will be about $50/
survey. day to rent or $1,500 to purchase. A wet suit can be
Data Collection-It should be possible to categorize rented for about $10/day and be purchased for
and measure collected debris with a high degree of about $400. A dry suit can be rented for about $25/
accuracy; however, the larger debris not brought up day and can be purchased for $500-1,000. Owing to
may be difficult to categorize due to algal growth and the.variety of equipment, purchasing all the scuba
time allotted for each dive. Along with the sorting of equipment and suits is best to ensure proper func-
the data, the following information should be col- tion and fit.
lected at each dive. The scuba divers should be experienced and have
at least an open water certification.. Commercial diver
date wages typically are $1,500/week; given that safety
exact location of dive (often distance from shore) rules generally require no less than two divers, a
sampling quadrat(s) or line segment(s) sampled week-long survey will cost about $3,000. Noncommer-
visibility cial divers are available for considerably less.
bottom topography (e.g., rocky or sandy) The total cost for a diving survey in cold water that
number, type, size, and condition of debris not col- uses commercial divers and is <1.5 km from shore
lected will run $3,600/week based on daily equipment
depth rental rates. For the same conditions and if all the
Suggested data forms for use in quadrat and equipment is purchased (except the boat, which is
transect techniques are presented in Figures 11 and still rented), the cost will be about $5,700/week (the
12, respectively. equipment which may be reused will be about $2,400
total). Diving is obviously the least expensive survey
Material and Personnel-The necessary equipment technique discussed in this chapter, but it is also the
will depend on the survey location. The following list most limited.

CHAPTER 5. Benthic Surveysfor Large Submerged Debris Items 71

Divers' Names Substrate Composition

Date (Yr/Mo/Day) Quadrat Area (in sq m)

Depth (in m)

Estimated Visibility (in m)

Small Debris Items Large Debris Items
(Collected) (From Underwater Pad)

Object Weight Category Comments Object Estimated Category Marked Comments
size

(e.g., amount of
algae growth on
item)

J-

Figure 11
Suggested template for quadrat survey forms.

Quality Assurance Program Certified divers have safety training, but specific
hazards and risks associated with each dive should
As with the quality assurance program plan for also be considered. For reasons of safety and
research in submersibles, the quality assurance efficiency, a detailed pre-dive plan should be
program plan for diving surveys must consider made. The plan should consider the following
the safety of the divers as well as the accuracy information:
and repeatability of the data collection procedure.

72 NOAA Technical Report NMFS 108: Marine Debris Survey Manual

Divers' Names Substrate Composition
Date (Yr/Mo/Day) Length of Line (in m)

Depth (in m) Length of Pole (in m)

Estimated Visibility (in m)

Small Debris Items Large Debris Items
(Collected) (From Underwater Pad)

Object Weight Category Comments Object Estimated Category Marked Conunents
size

(e.g., amount of
algae growth on
item)

Figure 12
Suggested template for transect survey forms.

� How much time is needed to sample predeter- 0 How will emergency situations be handled?
mined transects, and quadrats? * How much time is needed at each dive site?
� How much time is needed in transit between shore -How long will it take to decompress?
and site? Between sites? -How long will it take to get to and from the bot-
� How much time is needed to set up quadrat or tom?
transect lines? -How long will it take to survey the transect or
quadrat?

CHAPTER 5. Benthic Surveysfor Large Submerged Debris Item 73
-How much time can each,diver spend in the wa- n = to tal number of quadrats,
ter per day? A = total area of study (ha),
-How much time does each tank give each diver? a = area of each quadrat (ha),
Items that might be included in the quality assur- R = ratio of total area to quadrat area,
ance program plan for the survey itself are as follows: S' = variance between quadrats (#I/hal or
kgl/hal),
categories used for describing (sorting) debris 11@= estimate of total amount or weight of
maximum size of debris to collect debris in the area of study (# or kg) or
length of each transect or size of each quadrat A both, and.
jj2 =
measurement of visibility variance associated with the estimate of
details of the data analysis N debris (#2/hal or kgl/hal)
diver experience
Repeatability should be ensured between the two The only difference in calculations for transect sur-
plans listed previously. The importance of this repeat- veys is that the average number or weight of benthic
ability cannot be overstressed. debris per quadrat is now defined as the average
number or weight of benthic debris per transect, and
the area per transect is defined by the length of the
Analytical Procedures cord multiplied by twice the width of the pole.
The two techniques-quadrat and transect-are simi- Summary
lar. In each survey, the total number or weight (or
both) of benthic debris in the population of interest 1. Three survey techniques are discussed: trawl, sub-
is observed and collected if possible. The average mersible, and diving surveys. Because of limited
number or weight (or both) of debris is calculated experience in using these three approaches when
from all the samples and extrapolated to the total assessing benthic debris, all techniques should be
area. The analysis is very similar to that described considered experimental at this point, especially
for trawl surveys, except that each quadrat or transect submersible and diving surveys.
should cover the same area, a, or bias can be 2. The number or, weight of benthic debris is the
introduced. standard field measurement. Sampling results can
I The following are calculations for quadrat surveys be extrapolated to estimate the amount of benthic
(Seber 1982): n debris in the population of interest.
'X Y, A 3. Of the three techniques, trawl surveys have been
X= '= ], R used the most often.
n
4. Trawl surveys are the least expensive technique to
perform in deep water (>50 m), especially when
taking advantage of vessels of opportunity.
n 5. Quality assurance/quality control program plans
X (Xi - are extremely important for all aspects of each sur-
S2 n - vey technique, with safety and repeatability being
of highest priority.
S2(I 6. A statistician, who has dealt with similar types of
o-2 = R 2
studies, should be consulted at the outset of survey
N n R2 planning and be involved through the completion
of the study.
where @C = average amount or weight per quadrat
(#/ha or kg/ha) or both,
X, = amount or weight in a specific quadrat
(#/ha or kg/ha) or both,

74 NOAA Technical Report NMFS 108: Marine Debris Survey Manual
Appendix

HAUL-POSITION FORM
2 3 4 6 7
VESSEL I I I I - CRUISE MAUL
14 13 16 IF is 19
YOUR
YEAR E= MONTH [= DAY LLJ NAME

LATITUDE ON 0
20 21 22 23 24 26 21 28 29 32 33 35 36
POSITION 1+1 JEJ
START ... I I I I I I I I I _j
X DEGREE MIN. MR. 45
C)
U_ 37 38 39 40 41 43 44 1 4514614714@ 149 30 32 53
= END .....
C)
RATE FIRST READING RATE SECOND READING
55 56 So 59 60 61 62 64 65 67 60 70 71 72 73 74 7
L?RAN = I I I I I I I = I
S ART
C)
0- LORAN 79 so 82- 83 84 83 as as so 91 92 94 95 96 97 98 100 101
END ... = I I I I I FT@ I I I I I I I I I I I
---------- DUP. COL. 1-36 FROM ABOVE
R
G AR DEPT" 37 38 39 40
. AVG. (FM) MIN AX
0
LL BOTTOM DEPTH 41 42 43 44
_j TO. AVG. (FM) FT-M MIN AX
TIME STANDARD USED
EQUIL. 6 7 DURATION 48 49 51 52 TIME START OUT
HOUR @n (MRS.) FT-M EQUILIBRIUM
54 55 57 58 so at HAUL
DISTANCE AUL r T__1
F YYPE I IN
(h?A!? I I I 1 1
62 63 64 WTD- AVG. 65 66 67 so MIN -MAX
TRACE DEPTH
STRATUM 4F?0 DESCR.

WATER 70 71 73 73 76 78
TEMPERATURLS SUK-1 I
(DEGREES C.) FACE GEARE771 METHOD

83 84
@OTTOM
0 YPE . = DESCR.
as 86 87 as 94 93 96 97 98 99
WE OUT GEAR DOOR
o ( M) TYPE ACCES
AU7SAMPLE
00 E HOD F]
103
MJjFOR WEATH@R. SEA
CE CONDI IONS

CL
c3 REMARKS

Appendix Figure A
Haul-position form used by the National Marine Fisheries Service during bottom trawl surveys in the Bering Sea.
L ITU E
28 29 30 31

. . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . . . . . . . .
141,
v
A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . .

aQ . . . . . . . . . . . . . . . . . . . . . . . ... . . . .
cr
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

aq
cn

76 NOAA Technical Report NAM 108: Marine Debris Survey Manual

MARINE DEBRIS DATA FORM

Name: Date:

Vessel:-(- Cruise:- Haul:
name @oda

Percent of Catch Sampled For Marine Debris:

ITEMS CAUGHT

TOW Twi
PLASTIC IN nunuw
014- STYR0F0AAMAA jor other plastic foaml old-
bags War
bottles: CUPS
green egg canons
Soda fast-food containers
other meat trays
caps. lids pieces
cups, spoons, forks, straws (specify)
diapers
disposable lighters METAL
fishing line beverage cam
fishing net battle caps
ficats & lures containers
hardhats crabilish traps
fight sticks 55 gallon drums
milk As Ar-Y
rope new
sheeting. large P@---
69adc holdlers Pull tabs
Scrapping bands wife
tampon applicators other Ispecif)l
toys PAPER
vegetable sacks bags
..write protection- cardboar
rings
other (specify) CLAPS
GLASS newspaper
bottles pieces
thForescer-4 light tubes other (specify)
light bulbs
pieces WOW (Leave driftwood an the beachl
Other IsPecify) crab or lobster traps
RUBBER cram
balloons panels
gloves pieces
tires cow Isplecify)
otheir Ispecify) CLOTHFRAGS
pieces

MARINE DEBRIS SOURCES

Percent foreign Percent U.S. Percent Unknown

ENTANGLED ANIMALS

Marine Debris Item Type of Animal Number of Animals

Consiente:

Appendix Figure C
Marine debris data form for bottom trawls Uune 1990).

Chapter 6

Aerial Surveys

Techniques used in other assessment problems two engines, is safer to use in the same areas as the
have a potential for use in assessing marine debris Cessna as well as over offshore areas and along rocky
problems. The Assessment Working Group men- coasts.
tioned two techniques (Ribic 1990): aerial surveys For safety reasons, both planes should be flown at
and remotely operated vehicles (ROVs). ROVs were 60 m or higher to reduce the effects of engine noise
described in Chapter 5. This chapter discusses aerial on the behavior of nearby animals. For marine mam-
surveys. mal and marine bird surveys, typical flying elevations
are 120-140 m. Assessments regarding optimal flying
elevation and speed for sighting marine debris have
General Considerations not been done and will likely vary depending on sur-
vey objectives and sighting conditions.
Researchers havejust started to investigate the use of The Cessna 172 seats a pilot and two observers; the
aircraft to collect data on large to very large debris. Twin Otter seats up to six people. In either plane two
Ryan (1988a) sighted large plastic objects (>200 mm people would be needed- to make observations: one
in diameter) from an elevation of 130 m in a light person to observe and the other to record data. A
aircraft during calm seas. S. Johnson (NMFS, Auke tape recorder may be used as a backup data recorder.
Bay, AK, pers. commun. August 1991) has used In the Twin Otter, there could be three people with
aircraft to assess quantities of trawl web on some two making observations out of opposite windows
Alaska beaches. The major advantage of aerial sur- while the third records the data.
veys is that large areas can be covered by aircraft in a Rental costs for a plane and pilot are $70-85/hour
relatively short period of time. The disadvantage for for the Cessna 172 and $800-900/hour for the Twin
general use is the lack of consensus on the field Otter (3-h minimum). If the survey takes more than
methods and the cost of renting the aircraft. Because I day, the pilot receives $200 per diem.
the technique is experimental, more research on
field design will be needed before its usefulness can
be established. For example, there is no consensus Aerial Photography
on the type of aircraft to use, at what elevation or
speed to fly, what types of debris can be seen from Aerial photography would be useful primarily for
the air (we expect that very large debris will be seen), concentrations of large to very large debris.
or what the sampling design should be. Other areas Primarily, a photograph can be taken so that the
to investigate include the use of aerial photography concentration of debris can be analyzed in detail
(see Golik and Rosenberg [1987] for tar balls) and later. This procedure is typically done for con-
aerial reconnaissance as a tool specifically for pilot centrations of marine mammals such as dolphin
studies for selecting beach or nearshore sites for species.
other sampling techniques discussed in Chapters 2 For aerial photography, a Twin Otter should be
and 4. used because typically the survey planes have the
Typical aircraft used for marine mammal surveys camera mounts in place. A 9X9 T-11 camera can be
are the Cessna 172 and the Twin Otter. Both are rented for about $1,000/mo, and film for this camera
above-wing aircraft (an airplane with the wing lo- will cost about $1,000/roll. This type of camera is
cated above the fuselage) with good visibility. The suggested because of its ability to take photographs
Cessna 172 should be used within 15 km of shore. of large areas, with high resolution (which is in
The Twin Otter can be used up to 320 km offshore. part due to the negatives being 22.9 CM2)
The Twin Otter travels at faster speeds (140-160 km/ (R. Grotefendt, Ebasco Environmental, Bellevue, WA,
h) than does the Cessna 172 (120-140 km/h). The pers. commun. August 1991). Similar cameras of
Cessna has one engine and is safe to use in bays and newer design may be used but will cost much more.
sounds and near flat beaches. The Twin Otter, with The camera weighs 25 kg or more and will have to be

78 NOAA Technical Report NMf`S 108: Marine Debris Survey Manual
shipped to the place of use. The cost of shipping is Conclusion
not covered in the monthly rental fee. The plane, More work must be done to assess the usefulness of
which travels at speeds greater than 140 km/h, must aerial surveys for monitoring marine debris. Aerial
fly higher than 60 m to get useful pictures (to pre- photography may be particularly useful for surveying
vent blurring of images and to cover a large enough debris in coastal areas and on beaches where debris is
area). The film development will be an additional concentrated.
cost.

Glossary

Accumulation rate. Amount of debris added to a Marine debris (marine litter). Solid materials of hu-
sampling unit during a specified time period man origin that are discarded at sea or reach the
(usually measured after cleaning of all debris).- sea through waterways or through domestic and
Aerial survey. A survey made using an airplane or industrial outfalls (National Academy of Sciences
helicopter. 1975).
Assessment studies. Studies that seek to quantify Medium debris. solid waste of human origin or
distribution, movement, and/or trends in the manufacture that is @:2.5 cm and !@10 cm.
type and amount of marine debris over space Nautical mile. A measure of distance based on lati-
and/or time. tude and longitude; equal to 1.9 kin (abbrev-
Backshore. Zone extending above normal high tide iated as "nmi").
level, but innundated by exceptionally high tides Nonparametric statistical method. A statistical
or large waves during storms. method is nonparametric if it may be used on
Baseline studies. Studies that describe the types and data with a nominal or ordinal scale of measure-
amount of debris over space and/or time to ment and if no assumptions are made about the
identify the magnitude of the marine debris data (i.e., distribution-free methods). Non-
problem. parametric methods are often used in situations
Beach. Whole of,the area affected by normal wave when data describe populations that are not
action, extending from a depth of 10 in below normally distributed. Some methods are ana-
water level at the lowest tide to the edge of the logues of parametric tests. Examples include. the
permanent coast; beaches may be composed of Mann-Whitney U test and the Kruskall-Wallis
mud, sand, gravel, boulders, and/or rock ledges. test.
Composite. A container made with a cardboard Non-reusable. A manufactured product designed
body and metal or plastic ends. for one-time use (e.g., bottle, can) on which no
Diving survey. A survey made underwater by per- deposit is normally charged.
sonnel using scuba or snorkeling equipment. Oceanic influences. A factor affecting a marine de-
Fiberboard. Thick brown cardboard that may be bris survey that is related. to physical oceano-
used to package cases of cans. graphic processes such as tides and currents.
Fishing gear. Any physical item or combination of Open top cans. Completely sealed cans with no
items that is placed in the water for the intended reclosable lid that may be opened with a can
purpose of capturing or controlling for subse- opener or have pull-off ends.
quent capture, living marine or aquatic organ- Parametric statistical method. A statistical method
isms (Coe 1986). derived by assuming a specified theoretical
Foreshore. A zone that includes that part of the model for the the data. The most common theo-
beach regularly covered and uncovered by high retical model used is the normal distribution.
tides. Many parametric tests have nonparametric ana-
Gill net. Lightweight singlestrand or multistrand logues. Examples include the t-test and F-test.
filament netting (Cole et al. 1990). Population of interest. All or a selected type of de-
Gillnet floats. Small elongated rigid foamed floats bris within an area defined in space and time
that are grooved and have four holes for attach- and about which an inference is to be made.
ment to the "cork line" of a gill net (Cole et al. Primary packaging. Any package that is in direct
1990). contact with the product and without which the
Knot. A measure of speed equal to nautical miles product normally would not be sold.
per hour. Quality assurance program plan. An orderly collec-
Laminate. - A material made up of two or more com- tion of detailed and specific operational pro-
ponents. cedures that delineate how a project will be
Landbased debris. Solid materials of human origin implemented and what quality control @ proce-
that reach the sea through waterways, domestic dures will be employed to ensure that data of
and industrial outfalls, or improper disposal on known and acceptable quality will be generated;
beaches. it further specifies how data will be evaluated to
Large debris. Solid waste of human origin or manu- ensure that it meets specified project goals
facture that is >10 cm and!@I in. (Verner 1990).

80 NOAA Technical Report NWS 108: Marine Debris Survey Manual
Returnable. The general term used for a beverage Small debris. Solid waste of human origin or manu-
container that is intended to be reused or re- facture that is <2.5 cm.
cycled. Standing stock. Amount of material found in the
Sample mean. A measure of central tendency; the sampling unit at a specified time.
average of the measurements. Statistical hypothesis. A statement about a popula-
tion parameter that a researcher is interested in
testing; the most usual parameters of interest are
X=-
n population means or changes in population
means.
Sample variance. A measure of spread; the average Statistical power. The probability of rejecting a null
squared deviation of the observations from their hypothesis when that hypothesis is false. Used in
mean: trend assessment studies to determine the num-
(unbiased ber of survey units to be used. Typically, a power
S2 (Xi - formula) of 75% or more is used in sample size calcula-
n - tions.
Stratification. The use of additional information to
Sampling frame. A listing of all possible sampling divide the sampling frame into non-overlapping
units that can be defined in the target popula- groups and then selecting a simple random
tion; the sampling units used in the survey are sample from each group. Use of stratification
randomly or systematically chosen from the sam- may produce a gain in precision in the param-
pling frame. eter estimates.
Sampling unit. A defined area on which a measure- Systematic survey. A survey design that follows a
ment will be taken. rule for choosing the sampling units. For ex-
Sea state. A code combining information on wind, ample, after a random starting point is chosen,
waves, and swell height to describe oceanic con- every kth sampling unit is used. Often used with
ditions; numbered from 0 to 9. The most useful a grid to sample a large geographic area.
in marine debris surveys are sea states 0 through Target population. The difference between the
4, with sea states of 5 or greater being used for population of interest and any restricted-access
severe conditions (i.e., gales). Sea states (SS) 0 areas within the population of interest. If there
through 4 are as follows: are no restrictions, the target population and the
SS 0 = sea like a mirror; winds <1 nmi; average population of interest are the same.
wave height is 0 m. Transect. The linear sampling unit on a beach or in
SS I = a smooth sea; ripples, very light winds; the open water of known length, Width may or
average wave h .eight 0-0.3 m. may not be defined. For strip transects, width is
SS 2 = a slight sea; small wavelets; winds light to fixed; for line transects, width is not fixed.
gentle; average wave height 0.3-0.6 m. Trawl/seine web. Twisted or braided fishing net
SS 3 = a moderate sea; large wavelets, crests be- (Cole et a]. 1990).
ginning to break; winds gentle to moderate; Trawl survey. A survey made using a boat that pulls
average wave height 0.6-1.2 m. a net at set depths in the water column (e.g.,
SS 4 = a rough sea; moderate waves; whitecaps; surface, mid-water, bottom).
winds moderate to strong breeze; average wave Very large debris. Solid waste of human origin or
manufacture that is >1 m.
height 1.2-2.4 M. Vessels of opportunity. A ship dedicated to a pur-
(from Duxbury and Duxbury 1984). pose not related to studying debris but which
Secondary packaging. Packaging used to collate allows researchers to conduct debris studies that
multiples of other containers, usually used while do not conflict with the ship's primary purpose.
transporting goods. Vessel-source debris. Solid materials of human ori-
Shore. The zone between the water's edge at nor- gin which were discarded at sea.
mal low tide and the shoreward limit of effective < Less than.
wave action. > Greater than.
Shoreward limit of wave action. The landward limit <_ Less than or equal to.
of effective wave action that generally occurs on Greater than or equal to.
the upper foreshore and usually is identified by a
pronounced concentration of debris.

List q cronyms

EPA Environmental Protection Agency
FAO Food and Agricultural Organization (of the United Nations)
10C Intergovernmental Oceanographic Commission (of UNESCO)
GIPME Global Investigation of Pollution in the Marine Environment
CMC Center for Marine Conservation
MARIPOL International Convention for the Prevention of Pollution from Ships
NMML National Marine Mammal Laboratory
IM0 International Maritime Organization
NOAA National Oceanic and Atmospheric Administration
ROV remote operating vehicle
CPLTE catch per unit effort

Citaftom

Ainley, D. G., W. R. Fraser, and L. B. Spear. Bakkala, R. G., and K. Wakabayashi, eds.
1990a. The incidence of plastic in the diets of Antarctic sea- 1985. Results of the cooperative U.S.-Japan groundfish inves-
birds. In Proceedings of the second international confer- tigations in the Bering Sea during May@August 1979. Int.
ence on,marine debris, 2-7 April 1989, Honolulu, HI (R. S. North Pacific Fisheries Comm., Vancouver, Canada, 249 p.
Sbomuia and M. L. Godfrey, eds.), p. 682-691. NOAA Balazs, G. H.
Tech. Memo. NMFS, NOAA-TM-NMFS-SWFSC-154. 1979. Synthetic debris observed on a Hawaiian monk seal.
Ainley, D. G., L. B. Spear, and C. A. Ribic. 'Elepaio U. Hawaiian Audubon Soc.) 40:43-44.
1990b. The incidence of plastics in the diets of pelagic sea- 1985. Impact of ocean debris on marine turtles: entangle-
birds in the eastern equatorial Pacific region. In Proceed- ment and ingestion. In Proceedings of the workshop on the
ings of the second international conference on marine de- fate and impact of marine debris; 27-29 November 1984,
bris; 2-7 April 1989, Honolulu, HI (R. S. Shomura and M. Honolulu, HI. NOAA Tech. Memo., NMFS, NOAA-TM-
L. Godfrey, eds.), p. 653-664. NOAA Tech. Memo. NMFS, NMFS-SWFC-54.
NOAA-TM-NMFS-SWFSC-154. Baltz, D. M., and G. V. Morejohn.
Andersen, N.R., R. Dawson, and G. Kullenberg. 1976. Evidence from seabirds of plastic particle pollution off
1986. The program of Global Investigation of Pollution in central California. Western Birds 7:111-112.
the Marine Environment (GIPME) of the Intergovern- Barnard, J .., W. Myers, J. Pearce, F. Ramsey, M. Sissenwine, and
mental Oceanographic Commission (IOC). Mar. Tech. Soc. W. Smith.
J. 20:21-28. 1985. Surveys for monitoring changes and trends in renew-
Andre, J. B., and R. Ittner. able resources: forests and marine fisheries. The American
1980. Hawaiian monk seal entangled in fishing net. Statistician 39:363-373.
'Elepaio U. Hawaiian Audubon Soc.) 41:51. Bayer, R. D., and R. E. Olson.
Anonymous.
1981. Nylon thread pollution. Mar. Poll. Bull. 12:397. 1988. Plastic particles in 3 Oregon fulmars. Oregon Birds
Atwood, D. K., F. J. Burton, J. E. Corredor, G. R. Harvey, A. J. Mata- / 14:155-156.
Jimenez, A. Vasquez-Botello, and B. A. Wade. Bean, M. J.
1987a. Results of the CARIPOL petroleum pollution moni- 1984. United States and international authorities applicable
toring project in the wider Caribbean. Mar. Pollut. Bull. to entanglement of marine mammals and other organisms
18:540-548. in lost or discarded fishing gear and other debris. Rep. to
Atwood, D. K, H. H. Cummings, W. J. Nodal, and R. C. Cul- Mar. Mamm. Comm. Washington, D.C., 56 p.
bertson. 1987. Legal strategies for reducing persistent plastics
1987b. The CARIPOL petroleum pollution monitoring in the marine environment. Mar. Pollut. Bull. 18(6B):
project and the CARIPOL petroleum pollution data- 357-360.
base. Caribbean Jour. Sci. 23:1-3. Berger, J. D., and C. E. Armistead.
Atwood, D. IC, F.J. Burton,J. E. Corredor, G. R. Harvey, A. J. Mata- 1987. Discarded net material in Alaskan waters, 1982-84.
Jimenez, A. Vasquez-Botello, and B. A. Wade. NOAA Tech. Memo., NMFS F/NWC-1 10.
1987c. Petroleum pollution in the Caribbean. Oceanus Bigg, M. A.
30:25-32. 1979. Incidence of adult northern fur seals entangled in de-
Augerot, X. bris on St. Paul Island, 1978. Background paper submitted
1988. Plastic in the ocean: what are we doing to clean it to the 22nd Annual Meeting of the Standing Scientific Com-
up? Washington Sea Grant Marine Advisory Services, Univ. mittee, North Pacific Fur Seal Commission, 9-13 April 1979,
Washington, Seattle, WA, 8 p. Washington, D.C., Pacific Biological Station, Nanaimo, Brit-
Austin, H. M., and P. M. Stoop-Glas. ish Columbia V9R SK6, Canada, 6 p.
1977. The distribution of polystyrene spheres and nibs in 1982. Sizes of scrap fishnet and plastic packing bands from
Block Island Sound during 1972-1973. Chesapeake Sci. western Vancouver Island during August-September
18:89-92. 1982. Background paper submitted to the 26th Annual
Azzarello, M. Y., and E. S. Van Vleet. Meeting of the Standing Scientific Committee, North Pa-
1987. Marine birds and plastic pollution. Mar. Ecol. Prog. cific Fur Seal Commission, 28-29 March 1983, Washington
Ser. 37:295-303. D.C. Pacific Biological Station, Nanaimo,, British Columbia
Baba, N., K Yoshida, M. Onoda, N. Nagai, and S. Toishi. V9R SK6, Canada, 4 p.
1988. Results of research on floating fishing gear and fish Bingel, F., D. Avsar, and M. Onsal.
net fragments in the area southwest of the Pribilof Islands
and off southern coasts of the Aleutian Islands, July-August 1987. A note on plastic materials in trawl catches in the
1985. In Proceedings of the North Pacific Rim fisherman's north-eastern Mediterr anean. Meeresforsch 31:227-2.33.
conference on marine debris; 13-16 October 1987, Kailua- Bond, S. L
Kona, HI (D. L. Alverson and J. A. June, eds.), p. 143- 1971. Red phalarope mortality in Southern California.
164. Natural Resources Consultants, Seattle, WA. Calif. Birds 2:97.
Baba, N., M. Kiyota, and K. Yoshida. Bonner, W. N., and T. S. McCann.
1990. Distribution of marine debris and northern fur seals in 1982. Neck collars on fur seals, Arctocephalus gazella, at South
the eastern Bering Sea. In Proceedings of the second inter- Georgia. Br. Antarct. Surv. Bull..57:73-77.
national conference on marine debris; 2-7 April 1989, Bourne, W. R. P.
Honolulu, HI (R. S. Shomura and M. L. Godfrey eds.), 1983. Reappraisal of the threats to seabirds. Mar., Pollut.
p. 419-430. NOAA Tech. Memo. NMFS, NOAA-TM-NMFS- Bull. 14:1-2.
SWFSC-154. 1985. Turtles and pollution. Mar. Pollut. Bull. 16:177-178.

84 NOAA Technical Report NMFN 108: Marine Debris Survey Manual

Bourne, W. R. P. (chair). Cawthorn, M. W.
1990. Report of the working group on entanglement of ma-' 1985. Entanglement in and ingestion of plastic litter by ma-
rine life. In Proceedings of the second international con- rine mammals, sharks, and turtles in New Zealand
ference on marine debris; 2-7 April 1989, Honolulu, HI, waters. In Proceedings of the workshop on the fate and
(R. S. Shomura and M. L. Godfrey, ed,-?.), p. 1207-1215. impact of marine debris, 27-29 November 1984, Honolulu
NOAA Tech. Memo. NMFS, NOAA-TM-NMFS-SWFSC-154. HI (R.S. Shomura and H.O. Yoshida, eds.), p. 336-
.Bourne I W. R. P., and M. J. Imber. 343. NOAA Tech Memo. NOAA-TM-NMFS-SWFS-54.
1982. Plastic pellets collected by a prion on Gough Island, 1989. Impacts of marine debris on wildlife in New Zealand
central South Atlantic Ocean. Mar. Pollut. Bull. 13:20-21. coastal waters: Marine debris in New Zealand's coastal wa-
Brockwell, P.J., and R. A. Davis. ters; proceedings of the national workshop on marine de-
1989. Time series: theory and methods. Springer-Verlag, bris, Department of Conservation, 9 March 1989,
New York. 519 p. Wellington.
Burnham, K_ P., and D. R. Anderson. CEE (Center for Environmental Education).
1984. The need for distance data in transect counts. J. 1986. Marine wildlife entanglement in North Amer-
Wildl. Manage. 48:1248-1254. ica. Center for Environmental Education, Washington
Burnham, K. P., D. R. Anderson, andj. L. Laake. D.C. 219 pp.
1980. Estimation of density from line transect sampling of 1987a. Plastics in the ocean: more than a litter prob-
biological populations. Wildl. Monogr. 72. lem. Center for Environmental Education, Washington
1985. Efficiency and bias in strip and line transect sampling. D.C. 128 pp.
J. Wildl. Manage. 49:1012-1018. 1987b. 1986 Texas coastal cleanup report. Center for Envi-
Butler,J. N., and B. F. Morris. ronmental Education, Washington, D.C. 56 p.
1974. Quantitative monitoring and variability of pelagi`c tar 1988. Texas coastal cleanup report. Center for Environ-
in the North Atlantic. In Marine pollution monitoring (pe- mental Education, Washington, D.C. 105 pp.
troleum), proceedings of a symposium and workshop held CMC (Center for Marine Conservation).
at NBS, 13-17 May 1974, Gaithersburg, Maryland. NBS 1991. Cleaning North America's beaches, 199.0 beach
Spec. publ. 409. cleanup results. Center for Marine Conservation, Wash-
Caddy, J. R. ington, D.C. 291 p.
1976. Practical considerations for quantitative estimation of Clark, R. B.
benthos from a submersible. In Underwater research (E.A. 1986. Marine pollution. Clarendon Press, Oxford. 215 pp.
Drew, J. T@. Lythgoe, and J. D. Woods, eds.), p. @85- Coe, J. M.
298. Academic Press Inc., N-Y. 1986. Derelict fishing gear: disaster or nuisance? M.A. thesis,
Calkins, D. G. Univ. of Wash., Seattle, WA. 79 p.
1985. Stellar sea lion entanglement in marine debris. In Pro- Cole, C. Andrew, John P. Kumer, David A. Manski, and Daniel V.
ceedings of the workshop on the fate and impact of marine Richards (eds).
debris; 27-29 November 1984, Honolulu, HI (R. S. 1990. Annual report of National Park Marine Debris Moni-
Shomura and H. 0. Yoshida, eds.), p. 308-314. NOAA toring Program: 1989 marine debris Survey. Tech. Report
Tech. Memo. NMFS, NOAA-TM-NMFS-SWFC-54. NPS/NRV%rV/NRTR-90/04.
Carpenter, E.J. Coleman, F. C., and D. H. S. Wehle.
1976. Plastics, pelagic tar, and other litter. In Strategies for 1984. Plastic pollution: a world-wide problem. Parks 9:9-
marine pollution monitoring (E.D. Goldberg, ed.) p. 77- 12.
89. John Wiley and Sons. NY. Colton,]. B., Jr.
Carpenter, E.J., and K. L. Smith,jr. 1974. Plastics in the ocean. Oceanus 18:61-64.
-.1972. Plastics on the Sargasso Sea surface. Science Colton, J. B., Jr., F. D. Knapp, and B. R. Burns.
175:1240-1241. 1974. Plastic particles in surface waters of the northwestern
Carpenter, E. J., S. J. Anderson, G. R. Harvey, H. P. Miklas, and B. Atlantic. Science 185:491-497.
B. Peck. Conant, S.
1972. Polystyrene spherules in coastal waters. Science 1984. Man-made debris and marine wildlife in the North-
178:749-750. western Hawaiian Islands. 'Elepaio I (J. Hawaiian Audubon
Carr, A. Soc.) 44:87-88.
1986. Rips, FADS, and little loggerheads. BioScience Connors, P. G., and K. G. Smith.
36:92-100. 1982. Oceanic plastic particle pollution: suspected effect on
1987. Impact of non-degradable marine debris on the ecol- fat deposition in red phalaropes. Mar. Pollut. Bull. 13:
ogy and survival outlook of sea turtles. Mar. Pollut. Bull. 18-20.
18(6B):352-356. Conover, W. J.
Carr, A. H., E. H. Amaral, A. W. Hulbert, and R. Cooper. 1980. Practical nonparametric statistics. John Wiley and
Sons, NY, 493 p.
1985. Underwater survey of simulated lost demersal and lost Corredor, J. E., D. K_ Atwood, A. Mata, and A. Vasquez-Botello
commercial gill nets off New England. In Proceedings of (eds.).
the workshop on the fate and impact of marine debris, 1987. Proceedings of the CARIPOL symposium on research
27-29 November 1984, Honolulu, HI (R.S. Shomura and and monitoring of petroleum pollution in the Caribbean
H.O. Yoshida, eds.), p. 439-447. NOAA Tech. Memo. Sea and Adjacent Regions convened in La Parguera, Puerto
NMFS, NOAA-TM-NMFS-SWFC-54. Rico, 2-6 December 1985. Caribbeanjour. Sci. 23(l).
Caulton, E., and M. Mocogni. Crittenden, R. N.
1987. Preliminary studies of man-made litter in the Firth of 1989. Abundance estimation based on echo counts. Ph.D.
Forth, Scotland. Mar. Pollut. Bull. 18(6B):446-450. dissertation, Univ. Washington, Seattle, 169 p.

-citatiow 85

Croxall, J. Dixon, T. R., and A. J. Cooke.
.1990. Impact of incidental mortality on Antarctic marine 1977. Discarded containers on a Kent beach. Mai. Pollut.
vertebrates. Antarctic Science 2: 1. Bull. 8:105-109.
Croxall,J, P., S. Rodwell, and I. L. Boyd. Dixon, T. R., and A. J. Dixon.
1990. Entanglement in man-made debris of Antarctic fur 1980. Marine litter surveillance at two sites on the western
seals at Bird Island, South Georgia. Marine Mammal Sci- Cherbourg Peninsula and westjutland shores of the English
ence 6:221-233. Channel and southern North Sea. Marine Litter Research
Cundell, A. M. Programme, Stage 2. The Tidy Britain Group, The Pier
- 1973. Plastic materials accumulating in Narragansett Bay. Wagen, Great Britain, 80 p.
Mar. Pollut. Bull. 4:187-188. Dixon, T.R., and T.J. Dixon
Dahlberg, M. L., and R. H. Day. 1981a. Marine litter surveillance. Mar. Pollut. Bull. 12:289-295.
1985. Observations of man-made objects on the surface of 1981b. Aeolian Sky. packaged chemicals pollution inci-
the North Pacific Ocean. In Proceedings of the workshop dent. Mar. Pollut. Bull. 12:53-56.
on the fate and impact of marine debris, V-29 November 1983. Marine litter surveillance on the North Atlantic Ocean
1984, Honolulu, HI, (R.S. Shomura and H.O. Yoshida, eds.) shores of Portugal and the Western Isles of Scotland. Ma-
p. 198-212. NOAA Tech. Memo. NMFS, NOAA-TM-NMFS- rine Litter Research Programme, Stage 5. The Tidy Britain
SWFC-54. Group, the Pier Wagen, Great Britain, 70 p.
Dart, J. K G., and P. S. Rainbow. 1986. Packaged dangerous goods washed on to beaches of
1976. Some underwater techniques for estimating echino- England and Wales. The Environmentalist 6:209-218.
derm populations. In Underwater research (E. A. Drew, J. Dixon, T. R. and C. Hawksley.
N. Lythgoe, and J. D. Woods, eds.), p. 302-312. Academic 1980. Litter on the beaches of the British Isles. Report of the
Press Inc., NY. First National Shoreline Litter Survey sponsored by The Sun-
Day, R. H. day Times. Marine Litter Research Programme, Stage 3, The
Tidy Britain Group, 70 P.
1980. The occurrence and characteristics of plastic pollution Doubleday, W. G. and D. Rivard (eds.).
in Alaska's marine birds. M.S. thesis, Univ. Alaska, Fair- 1981. Bottom trawl surveys: proceedings of a workshop; 12-
banks. I I I p. 14 November 1980, Ottawa, Canada. Can. Spec. Publ. Fish.
Day, R. H., and D. G. Shaw. Aquat. Sci. 58, 273 p.
1987. Patterns in the abundance of pelagic plastic and tar in Duerr, C.
the North Pacific Ocean, 1976-1985. Mar. Pollut. Bull. 1980. Plastic is forever: our nondegradable treasures.
18(6B):31 1-316. Oceans, November 1980:59-60.
Day, R. H., D. G. Shaw, and S. E. Ignell. Duronslet, M.J., D. B. Revera, and K M. Stanley.
1990a. The quantitative distribution and characteristics of 1991. Marine debris and sea turtle strandings on beaches of the
marine debris in the North Pacific Ocean, 1984-88. In upper Texas and southwestern Louisiana coasts, June 1987
Proceedings of the second international conference on ma- through September 1989. NOAA Tech. Memo NMFS-SEFC.
rine debris; 2-7 April 1989, Honolulu, HI (R. S. Shomura Duxbury, A.C., and A. Duxbury.
and M. L. Godfrey, eds.), p. 182-211. NOAA Tech. Memo. 1984. An introduction to the world's ocean. Addison
NMFS, NOAA-TM-NMFS-SWFSC-154. Wesley Pub]. Co., Menlo Park, CA; Reading, MA, 408 p.
1990b. The quantitative distribution and characteristics of FAO.
neuston plastic in the North Pacific Ocean, 1984-88. In 1989. Report of the IOC/FAO/UNEP review meeting on the
Proceedings of the second international conference on ma- persistent synthetic materials pilot survey; 12-14junq 1989,
rine debris; 2-7 April 1989, Honolulu, HI (R. S.- Shomura Haifa, Israel, 46 p.
and M. L. Godfrey, eds.), p. 247-266. NOAA Tech. Memo. Feder, H. M., S. C. Jewett, and J. R. Hilsinger.
NMFS, NOAA-TM-NMFS-SWFSC-154. 1978. Man-made debris on the, Bering Sea fioor. Mar.
Day, R. H.', D@ H. S. Wehte, and F. C. Coleman. Pollut. Bull. 9:52-53.
1985. Ingestion of plastic pollutants by marine birds. In Fowler, C. W.
Proceedings of the workshop on the fate and impact of ma- 1982. Entanglement as an explanation for the decline in
rine debris; 27-29 November 1984, HI (R. S. Shomura and northern fur seals of the Pribilof Islands.' Background pa-
H. 0. Yoshida, eds.), p. 344-386. NOAA Tech. memo. per submitted to the 25th Annual Meeting of the Standing
NMFS, NOAA-TM-NMFS-SWFC-54. Scientific Committee of the North Pacific Fur Seal Commis-
Dean, T. sion, 9-13 April 1984, Moscow, U.S.S.R. NOAA, National
1985. Plastic hazards to birds. Br. Birds. 78:661-662. Marine Mammal Lab., 7600 Sand Point Way N,E., Seattle,
DeGange, A. R., and T. C. Newby. WA 98115, 24 p.
1980. Mortality of seabirds and fish in a lost salmon drift- 1984. Entanglement in fishing debris as a contributing factor
in the decline of northern fur seals on the Pribilof Islands.
net. Mar. Pollut. Bull. 11:322-323. Background Ipaper submitted to the 27th Annual Meeting of
Dickerman, R. W., and R. G. Goelet.' the Standing Scientific Committee of the North Pacific
1987. Northern Gannet starvation after swallowing styro- Fur Seal Commission, 9-13 April 1984, Moscow, U.S.S.R.
foam. Mar. Pollut. Bull. 18:293. NOA,4@, National Marine Mammal Lab., 7600 Sand Point
Dixon, T.J., and T. R. Dixon. Way N.E., Seattle, WA 98115, 33 p.
1983. Marine litter distribution and composition in the 1985. An evaluation of the role of entanglement in the
North Sea. Mar. Pollut. Bull. 14:145-148. population dynamics of northern fur seals on the Pribilof
Dixon, T. R. Islands. In Proceedings of the workshop on the fate and
1981. Danger on the beach. Mar. Pollut. Bull. 12:3. impact of marine debris; 27-29 November 1984, Honolulu,
1997. More reports of dangerous packages and muni- HI (R.S. Shomura and H. 0. Yoshida, eds.), 'p. 291-
tions. Mar. Pollut. Bull. 18:146. 307. NOAA Tech. Memo. NMFS, NOAA-TM-NMFS-SWFC-54.

86 NCIAA Technical Report NMIFS 108: Marine Debris Survey Manual

1987. Marine debris and northern fur seals: a case study. Gosliner, M.
Mar. Pollut. Bull. 18(6B):326--335. 1985. Legal authorities pertinent to entanglement by marine
1988. A review of seal and sea lion entanglement in marine debris. In Proceedings of the workshop on the fate and
fishing debris. In Proceedings of the North Pacific Rim impact of marine debris; 27-29 November 1984, Honolulu,
fisherman's conference on marine debris;.13-16 October HI (R. S. Shomura and H. 0. Yoshida), p. 15-33. NOAA
1987, Kailua-Kona, HI (D. L. Alverson and J. A. June, eds.), Tech. Memo. NMFS, NOAA-NMFS-SWFC-54.
p. 16-63. Natural Resources Consultants, Seattle, WA. Gould, P. J., and D. J. Forsell.
Fowler, C. W., and T. R. Merrell. 1989. Techniques for shipboard surveys of marine birds.
1986. Victims of plastic technology. Alaska Fish and Game U.S. Fish Wild]. Serv. Tech. Rep. 25. Washington, D.C.,
18:34-37. 22 p.
Fowler, C. W. and T. J. Ragen. Gramentz, D.
1990. Entanglement studies, St. Paul Island, 1989 juvenile 1988. Involvement of loggerhead turtle with the plastic,
male northern fur seals. NMFS, Northwest and Alaska metal, and hydrocarbon pollution in the Central Med-
Fisheries Centers NWAFC Processed Report 90-06. iterranean. Mar. Poll. Bull. 19:11-13.
Fowler, C. W., R. Merrick, and N. Baba. Gregory, M. R.
1989. Entanglement studies, St. Paul Island, 1988 juvenile 1977. Plastic pellets on New Zealand beaches. Mar. Pollut.
male roundups. NMFS, Northwest and Alaska Fisheries Bull. 8:82-84.
Centers NWAFC Processed Report 89-01. .1978a. Accumulation and distribution of virgin plastic gran-
Fowler, C. W., R. Merrick, andj. D. Baker. ules on New Zealand beaches. N.Z.J. of Marine and Fresh-
1990. Studies of the population level effects of entanglement water Research 12:399-414.
on northern fur seals. In Proceedings of the second inter-- 1978b. Virgin plastic granules on southwest Pacific beaches
national conference on marine debris; 2-7 April 1989, Ho- and their possible environmental implications. Tenth Int.
nolulu, HI (R.S. Shomura and H. 0. Yoshida, eds.), p. 291- Congress on Sedimentology 1:270-271.
307. NOAA Tech. Memo. NMFS, NOAA-TM-NMFS- 1983. Virgin plastic granules on some beaches of eastern
SWFSC-154. Canada and Bermuda. Marine Environ. Res. 10:73-92.
Fry, D. M., S. I. Fefer, and L. Sileo. 1987. Plastics and other seaborne litter on the shores of New
1987. Ingestion of plastic debris by Laysan albatrosses and Zealand's subantarctic islands. New Zealand Antarctic
wedge-tailed shearwaters in the Hawaiian Islands. Mar. Record 7:32-47.
Pollut. Bull. 18(6B):339-343. 1990. Plastics: accumulation, distribution, and environmen-
Furness, B. L. tal effects of meso-, macro-, and megalitter in surface waters
1983. Plastic particles in three Procellarifform seabirds from and on shores of the southwest Pacific. In Proceedings of
the Benguela Current, South Africa. Mar. Pollut. Bull. the second international conference on marine debris; 2-7
April 1989, Honolulu, HI (R. S. Sbomura and M. L.
14:307-308.
Furness, R. W. Godfrey), p. 55-84. NOAA Tech. Memo. NMFS, NOAA-
1985a. Plastic particle pollution: accumulation by Pro- TM-N.MFS-SWFSC-154.
cellariiform seabirds at Scottish colonies. Mar. Pollut. Bull. Gregory, M. R., R. M. Kirk, and M. C. G. Mabin.
16:103-106. 1984. Pelagic tar, oil, plastics and other litter in the surface
1985b. Ingestion of plastic particles by seabirds at Gough Is- waters of the New Zealand Sector of the Southern Ocean,
land, South Atlantic Ocean. Environ. Poll. (Series A) and on Ross Dependency shores. N.Z.J. Ant. Res. 6:12-28.
38:261-272. Harper, P. C., andj. A. Fowler.
Garner,J. 1987. Plastic pellets in New Zealand storm-killed prions
1967. Modern deep sea trawling gear. Fishing News (Pachyptila spp.). Notornis 34:65-70.
(Books) Ltd., London. Hays, H. and G. Cormons.
Gerrodette, T. 1974. Plastic particles found in tern pellets, on coastal
1985. Toward a population dynamics of marine debris. In beaches and at factory sites. Mar. Pollut. Bull. 5:44-46.
Proceedings of the workshop on the fate and impact of HMEPA (Hellenic Marine Environment Protection Association).
marine debris; 27-29 November 1984, Honolulu, HI (R. S. 1991. Public awareness campaign to limit garbage pollution
Shomura and H. 0. Yoshida, eds.), p. 508-518. NOAA of the Greek seas and beaches. Final rep. MEDSPA-89-1/
Tech. Memo. NMFS, NOAA-NMFS-SWFC-54. GR/008/GR/S, Mediterrean Special Action Programme for
Gilbert, R. 0. the Environment of the European Communities.
1987. Statistical methods for environmental pollution mon- Henderson, J. R.
itoring. Van Nostrand Reinhold Company. NY. 1984. Encounters of Hawaiian monk seals with fishing gear
Golik, A. at Lisianski Island, 1982. Mar. Fish. Rev. 46:59-61.
1982. The distribution and behaviour of tar balls along the 1985. A review of Hawaiian monk seal. entanglements in ma-
Israeli coast. Estuarine Coastal and Shelf Sci. 15:267-276. rine debris. In Proceedings of the workshop on the fate
Golik, A., and Y. Gertner. and impact of marine debris; 27-29 November 1984, Hono-
1990. Solid waste on the Israeli coast - composition, lulu, HI (R. S. Shomura and H. 0. Yoshida, eds.), p. 326-
335. NOAA Tech Memo. NOAA-TM-NMFS-SWFS-54.
sources, and management. In Proceedings of the second 1988. Marine debris in Hawaii. In Proceedings of the
international conference on marine debris; 2-7 April 1989, North Pacific Rim fisherman's conference on marine de-
Honolulu, HI (R.S. Shomura and M. L. Godfrey), p. 369- bris; 13-16 October 1987, Kailua-Kona, HI. Natural Re-
378. NOAA Tech. Memo. NMFS, NOAA-TM-NMFS- sources Consultants, Seattle, WA.
SWFSC-154. 1990. Recent entanglements of Hawaiian monk seals in ma-
Golik, A. and N. Rosenberg. rine debris. In Proceedings of the second international
1987. Quantitative evaluation of beach stranded tar conference on marine debris; 2-7 April 1989, Honolulu, HI,
balls by means of air photographs. Mar. Pollut. Bull.
p. 540-553. NOAA Tech. Memo. NMFS, NOAA-TM-NMFS-
18:289-293.- SWFSC-154.

Citations 87

Henderson, J. R., and M. B. Pillos. mitted to the International North Pacific Fisheries Commis-
1985. Accumulation of net fragments and other marine de- sion, Anchorage, AK, November 1986, NWAFC, NMFS,
bris in the Northwestern Hawaiian Islands (Abstract only). NOAA, Auke Bay Lab., P. 0. Box 210155, Auke Bay, AK_
In Proceedings of the workshop on the fate and impact of 99821, 15 p.
marine debris; 27-29 November 1984, Honolulu, HI (R. S. IOC (Intergovernmental Oceanographic Commission)
Shomura and H. 0. Yoshida, eds.), p. 326-335. NOAA 1984. Manual for monitoring oil and dissolved/dispersed pe-
Tech Memo. NOAA-TM-NMFS-SWFS-54. troleum hydrocarbons in marine waters and on beaches;
Henderson, J. R., S. L. Austin, and M, B. Pillos. Procedures for the petroleum component of the IOC Ma-
1987. Summary of webbing and net fragments found on rine Pollution Monitoring System (MARPOLMON - P).
northwestern Hawaiian Islands beaches, 1982-1986. UNESCO Manuals and Guides No. 13, 35 p.
NOAA, NMFS, SWFC Admin. Rep. H-87-11,15 p. Isaaks E. H., and R. M. Srivastava.
Heneman, B., and the Centerfor Environmental Education. 1;89. An introduction to applied geostatistics. Oxford
1988. Persistent marine debris in the North Sea, northwest Univ. Press. NY.
Atlantic ocean, wider Caribbean area, and the west coast of Jewett, S. C.
Baja California. Unpubl. rep. to the Marine Mammal 1976. Pollutants of the Northeast Gulf of Alaska. Mar.
Commission and National Ocean Pollution Program Office, Pollut. Bull. 7:169.
NOAA, USDC, Contr. Rep. MM3309598-5,. Johnson, S. W.
Heyerdahl, T. 1988. Deposition of entanglement debris on Alaskan
1971. Atlantic Ocean pollution and biota observed by the beaches. In Proceedings of the North Pacific Rim
'Ra' Expeditions. Biol. Conservation 3:164-167. fisherman's conference on marine debris, 13-16 October
High, W. L. 1987, Kailua-Kona, HI (D. L. Averson and J. A. June, eds.),
1985. Some consequences of lost fish *ing gear. In Proceed-
ings of the workshop on the fate and impact of marine de- p. 207-231. Natural Resources Consultants, Seattle, WA.
bris: 27-29 November 1984, Honolulu, HI (R. S. Shomura 1989. Deposition, fate, and characteristics of derelict trawl
and H. 0. Yoshida, eds.), p. 430-437. NOAA Tech Memo. web on an Alaskan beach. Mar. Pollut. Bull. 20:164-168.
NOAA-TM-NMFS-SWFS-54. 1990a. Distribution, abundance, and source of entanglement
Hirsch, R. M., andj. R. Slack. debris and other plastics on Alaskan beaches, 1982-88. In
1984. A nonparametric trend test for seasonal data with se- Proceedings of the second international conference on ma-
rial dependence. Water Resources Research 20:727-732. rine debrisj 2-7 April 1989, Honolulu, HI (R. S. Shomura
Hirsch, R. M.j R. Slack, and R. A. Smith. and M. L. Godfrey, eds.), p. 331-348. NOAA Tech. Memo.
1982. Techniques of trend analysis for monthly water quality NMFS, NOAA-TM-NMFS-SWFSC-154.
data. Water Resources Research 18:107-121. 1990b. Entanglement debris on Alaskan beaches, 1989.
Hiscock, K. NWAFC Processed Report 90-10. 16 p.
1987. Subtidal rock and shallow sediments using diving. In Johnson, S. W., and T. R. Merrell.
Biological surveys of estuaries and coasts U. M. Baker and 1988. Entanglement debris on Alaskan beaches, 1986.
W. J. Wolfe, eds.), p. 198-237. Cambridge United Publish- NOAA Tech. Memo., NMFS, F/NWC-126.26 p.
ing, Cambridge. Jones, L. L., and R. C. Ferrero.
1989. Development of methods for surveys and monitoring 1985. Observations of net debris and associated entangle-
using diving. Progress in Underwater Science 13:57-64. ments in the North Pacific Ocean and Bering Sea, 1978-
Hjelmeland, K., B. H. Pedersen, and E. M. Nilssen. 84. In Proceedings of the workshop on the fate and impact
1988. Trypsin content in intestines of herring larvae, Clupea of marine debris; 27-29 November 1984, Honolulu, HI
harengus, ingesting inert polystyrene spheres or live crusta- (R. S. Shomura and M. L. Godfrey, eds.), p. 183-
cea prey. Mar. Bio. 98:331-335. 196. NbAA Tech Memo. NOAA-TM-NMFS-SWFS-54.
Holmstr6m, A. June, J. A.
1975. Plastic films on the bottom of the Skagerack. Nature 1990. Type, source, and abundance of trawl-caught marine
(London) 255:622-623. debris off Oregon, in the eastern Bering Sea, and in Norton
Horsman, P. V. Sound in 1988. In Proceedings of the second international
1982. The amount of garbage pollution from merchant conference on marine debris, 2-7 April 1989, Honolulu,
ships. Mar. Pollut. Bull. 13:167-169. HI (R. S. Shomura and H. 0. Yoshida, eds.), p. 279-
1985. Garbage kills. BBC Wildlife, August, 391-303. 301. NOAA Tech. Memo. NMFS, NOAA-TM-NMFS-SWFSC-
Hoss, D. E., and L. R. Settle. 154.
1990.. Ingestion of plastics by teleost fishes. In. Proceedings
Kartar, S.,R. A. Milne, and M. Sainsbury.
of the second international conference on marine debris, 1973. Polystyrene waste in the Severn estu .ary. Mar. Pollut.
2-7 April 1989, Honolulu, HI (R. S. Shornura and M. L. Bull. 4:144.
Godfrey), p. 693-709. NOAA Tech. Memo. NMFS, NOAA-
TM-NMFS-SWFSC-154. Kartar, S., F. Abou-Seedo, and M. Sainsbury.
Ignell, S. E. 1976. Polystyrene spherules ih the Severn estuary-A
1985. Results of the 1985 research on the highseas squid progress report. Mar. Pollut. Bull. 7:52.
driftnet fisheries of the North Pacific Ocean. Document Keller, G. H.
submitted to by the International North Pacific Fisheries 1977. 'The submersible-a unique tool for marine ge-
Commission, Tokyo, Japan, November 1985. Northwest and ology. In Submersibles and their use in oceanography and
Alaska Fisheries Center, NMFS, NOAA, Auke Bay Labora- ocean engineering, Elsevier Scientific Pub]. Co.,
tory, Auke Bay, Alaska. Amsterdam, Netherlands.
Ignell, S. E., and M. L. Dahlberg. Kenyon, K. W., and E. Kridler.
1986. Results of -cooperative research on the distribution of 1969. Laysan albatrosses swallow indigestible matter. Auk
marine debris in the North Pacific Ocean. Document sub- 86:339-343.

88 NOAA Technical Report NMFS 108: Marine Debris Survey Manual

Keppel, G. MPDTF (Marine Plastic Debris Task Force).
1982. Design and analysis, a researcher's handbook, 2nd 1988. Marine plastic debris action plan for Washington
edition. PrenticerHall, Inc. Englewood Cliffs, NJ. State. Washington State Dept. of Natural Resources. 45 pp.
Klemm, B., and D. Wendt. Marquez, J. A., and 1. Zandi.
1990. Beach confetti. Sea Frontiers 36:28-29. 1985. Litter management-measurement aspects. J. Res.
Kubota, T. Manage. Tech. 14:67-75.
1990. Synthetic materials found in the stomachs of longnose McConnell, K. E. (chair).
lancetfish collected from Suruga Bay, central Japan. In 1990. Report of the working group on economic aspects of
Proceedings of the second international conference on ma- marine debris. In Proceedings of the second international
rine debris, 2-7 April 1989, Honolulu, HI (R. S.'Shomura conference 'on @marine debris; 2-7 April 1989, Honolulu,
and M. L. Godfrey, eds.), p. 710-717. NOAA Tech. Memo. HI (R. S. Shomura and M. L. Godfrey, eds.), p. 1235-
NMFS, NOAA-TM-NMFS-SWFSC-154. 1239. NOAA Tech. Memo. NMFS, NOAA-TM-NMFS-
Kullenberg, G. SWFSC-154.
1986 *The IOC programme on marine pollution. Mar. McCoy, F. W.
Pollut. Bull. 17:341-352. 1988. Floating megalitter in the eastern Mediter-
Laist, D. W. ranean. Mar. Pollut. Bull. 19:25-28.
1987. Overview of the biological effects of lost and discarded Meade, N. F., K. M. Drazek, and V. R. Leeworithy.
plastic debris in the marine environment. Mar. Pollut. 1990. An economic perspective on the problem of marine
Bull. 18(6B):319-326. debris. In Proceedings of the second international confer-
Lenihan, H. S.,J. S. OliverJ. M. Oakden, and M. D. Stephenson. ence on marine debris, 2-7 April 1989, Honolulu, HI (R. S.
1990. Intense and localized benthic,,marine pollution Shomura and M. L. Godfrey, ects.), p. 777-791. NOAA
around McMurdo Sound, Antarctica. Mar. Pol. Bull. Tech. Memo. NkFS, NOAA-TM- N*MFS-SWFSC-I 54.
21:422-430. Merrell, T. R.
Lentz, S. A. 1980. Accumulation of plastic litter on beaches of Amchitka
1987. Plastics in the marine environment: legal approaches Island, Alaska. Mar. Environ. Res. 3:171-184.
for international action. Mar. Pollut, Bull. 18(6B):361- 1984. A decade of change in nets and plastic litter from
365. fisheries off Alaska. Mar. Pollut. Bull. 15 :378-384.
Lettenmaier, D. P. 1985. Fishnets and other plastic litter on Alaska beaches. In
1978. Design considerations for ambient stream quality Proceedings of the workshop on the fate and impact of ma-
monitoring. Water Resources Bulletin 14:884-902. rine debris; 27-29 November 1984, Honolulu, HI (R. S.
Lettenmaier, D. P., L. L. Conquest, andJ. P. Hughes. Shomura and H. 0. Yoshida, eds.), p. 160-182. Memo.
1982. Routine streams and rivers water quality trend moni- NOAA-TM-NMFS-SWFS-54.
toring review. C. W. Harris Hydraulics Laboratory, Dept. Merrell, T. R., and S. W. Johnson.
of Civil Engineering, Univ. of Washington, Seattle, WA, 1987. Surveys of plastic litter on Alaskan beaches, 1985.
Technical Report No. 75. NOAA Tech. Memo. NMFS F/NWC-1 16. 21 p.
Lindstedt, D. M., andJ. C. Holmes. Mio, S., and S. Takeharna.
1989. Debris is not a cheese: litter in coastal Louisiana. pro- 1988. Estimation of distribution of marine debris based on
ceedings of the sixth symposium on coastal and ocean man- the 1986 sighting survey. In Proceedings of the North Pa-
agement, p. 1297-1310. cific Rim fisherman's conference on marine debris;' 13-16
Lutz, P. L. October 1987, Kailua-Kona, HI (D. L. Alverson and J. A.
1990. Studies on the ingestion of plastic and latex by June, eds.), p. 64-94. Natural Resources Consultants, Se-
sea turtles. In Proceedings of the second international attle, WA.
conference on marine debris, 2-7 April 1989, Honolulu, HI Mio, S., S. Takehama, and S. Matsumura.
(R. S. Shomura and M. L. Godfrey, eds.), p. 719- 1990. Distribution and density of fioating objects in the
735. NOAA Tech. Memo. NMFS, NOAA-TM-NMFS- North Pacific based on 1987 sighting survey. In Proceed-
SWFSC-154. ings of the second international conference on marine de-
Manski, D. A., W. P. Gregg, C. A. Cole, and D. V. Richards. bris; 2-7 April 1989, Honolulu, H1 (R. S. Shomura and M.
1991. Annual report of the national park marine debris L. Godfrey, eds.), p. 212-246. NOAA Tech. Memo. NMFS,
monitoring program, 1990 marine debris surveys. USDI NOAA-TM-NMFS-SWFSC-154.
NPS Technical Report NPS/NRWV/NRTR-91. In press. Morris, A. W., and E. 1. Hamilton.
Manville, II, A. M. 1974. Polystyrene spherules in the Bristol Channel. Mar.
1990. . A survey of plastics on western Aleutian Island beaches Pollut. Bull. 5:26-27.
and related wildlife entanglement. In Proceedings of the Morris, R.J.
second international conference on marine debris; 2-7 1980a. Floating plastic debris in the Mediterranean. Mar.
April 1989, -Honolulu, HI (R. S. Shomura and M. L. Pollut. Bull. 11:125.
Godfrey), p. 349-363. NOAA Tech. Memo. NMFS, NOAA- 1980b. Plastic debris in the surface waters of the South
TM-NMFS-SWFSC-154. Atlantic. Mar. Pollut. But.J. 11:164-166.
MMC (Marine Mammal Commission). Nasu,.K., and K. Hiramatsu.
.1987. Annual report of the Marine Mammal Commission cal- 1990. Distribution and density of marine debris in the North
endar year 1986. Rep. to Congress. NTIS PB98-154092, Pacific based on sighting surveys in 1989. Document sub-
193 p. mitted to the Annual Meeting of. the International North
1991. Annual report of the Marine Mammal Commission cal- Pacific Fisheries Commission, Vancouver, Canada, 1990.0c-
endar year 1990. Rep. to Congress. NTIS. 270 p. tober. Fisheries Agency of Japan, National Research Insti-

Citations 89

tute of Far Seas Fisheries, 7-1 Orido 5 chome, Shimizu, Pritzlaff, J. A. (ed.).
Shizuokajapan 424. 28 p., 1979. International safety standard guidelines for the opera-
NAS (National Academy of Sciences). tion of undersea vehicles. Marine Technology Society,
1975. Marine litter. In Assessing potential ocean Washington, D.C.
pollutants. A report of the study panel on assessing poten- Pruter, A. T. -
tial ocean pollutants to the Ocean Affairs Board, Commis- 1987a. Sources, quantities and distribution of persistent
sion on Natural Resources, National Research Council, Na- plastics in the marine environment. Mar. PolluL. Bull.
tional Academy of Sciences, Washington, D.C. 18(6B):305-310.
Neilsonj 1987b. Plastics in the marine environment. Fisheries
1985. The Oregon experience. In Proceedings of the work- 12:16-17.
shop on the fate and impact of marine debris; 27@29 No- Randall, B. M., R. M. Randall, and G. J. Rossouw.
vember 1984, Honolulu, HI (R. S. Shomura and H. 0. 1983. Plastic particle pollution in great Shearwater (Puffinius
Yoshida, eds.), p. 154-159. NOAA Tech Memo. NOAA-TM- gravis) from Gough Island. South African journal of An t-
NMFS-SWFS-54. arctic Research 13:49-50.
Ogi, H. Ribic, C. A. (chair).
1990. Ingestion of plastic particles by sooty and short-tailed 1990. Report of the working group on methods to assess the
shearwaters in the North Pacific. In Proceedings of the amount and types of marine debris. In Proceedings of the
second international conference on marine debris, 2-7 second international conference on marine debris; 2-7
April 1989, Honolulu, HI (R. S. Shomura and M. L. April 1989, Honolulu, HI (R. S. Shomura and M. L.
Godfrey, eds.), p. 635-652. NOAA Tech. Memo. NMFS, Godfrey, eds.), p. 1201-1206. NOAA Tech. Memo. NMFS,
NOAA-TM-NMFS-SWFSC-154. NOAA-TM-NMFS-SWFSC-154.
O'Hara, V_ J. 1991. Design of shoreline surveys for aquatic litter pollution.
1989. Trash on America's beaches: a national Final report to U.S. EPA Office of Marine and Estuarine
assessment. Center for Marine Conservation, Washington Protection. EPA/600/3-91/026. NTISPB91179051/AS.
D.C. Ribic, C. A., and L. J. Bledsoe.
1990. National marine debris data base: findings on beach 1986. Design of surveys for density of surface marine debris
debris reported by citizens. In Proceedings of the second in the North Pacific. NOAA, NMFS, NWAFC Processed
international conference on marine debris; 2-7 April 1989, Rept. No. 267. 69 p.
Honolulu, HI (R. S. Shomura and M. L. Godfrey, eds.), p. 1990. Estimating the density of floating marine debris: De-
379-391. NOAA Tech. Memo. NMFS, NOAA-TM-NMFS- sign considerations. In Proceedings of the second interna-
SWFSC-154. tional conference on marine debris; 2-7 April 1989, Hono-
O'Hara, Y, J., and P. Debenham. lulu, HI (R. S. Shomura and M. L. Godfrey, eds.), p.
1989. Cleaning America's beaches: 1988 national beach 302-308. NOAA Tech. Memo. NMFS,.NOAA-TM-NMFS-
cleanup results. Center for Marine Conservation, Wash- SWFSC-154.
ington, D.C. 202 p. Ribic, C. A., and S. W. Johnson.
O'Hara, K-J., and L. K. Younger. 1990. Guidelines for the design of beach debris surveys. In
1990. Cleaning North America's beaches, 1989 beach Proceedings of the second international conference on ma-
cleanup results. Center for Marine Conservation, Wash- rine debris; 2-7 April 1989, Honolulu, HI (R. S. Shomura
ington, D.C. 310 p. and M. L. Godfrey, eds.), p. 392-402. NOAA Tech. Memo.
O'Hara, K_ J., S. Iudicello, and R. Bierce. NMFS, NOAA-TM-NMFS-SWFSC-154.
1988. A citizens guide to plastic in the ocean: more than Rice, W. D., and A. A. Wolman.
a litter problem. Center for Environmental Education, 1982. Whale census in the Gulf of Alaska, June to August
Washington, D.C. 131 p 1980. Rep. to the Int. Whaling Comm. 32:491-497.
Palmer, H. D. Risebrough, R. W.
1977. The use of manned submersibles in the study of ocean 1969. The sea: should we write it off as the future garbage
waste disposal. In Submersibles and their use in oceanog- pit? From background book supplement, 13th National
raphy and ocean engineering (R. A. Geyer, ed.), p. 317-334. Conference, U.S. National Comm. for UNESCO, San Fran-
Elsevier Scientific Publ. Co., Amsterdam, Netherlands. cisco, 23-25 November. 23 p.
Parker, P. A. Rothstein, S. 1.
1990. Clearing the oceans of plastic. Sea Frontiers 36:18- 1973. Plastic particle pollution of the surface of the
27. Atlantic Ocean: evidence from a seabird. Condor 75:
Parslow, J. L. F., and D. J. Jefferies. 344-345.
1972. Elastic thread pollution in puffins. Mar. Pollut. Bull. Ryan, P. C.
3:43-45. 1985. Plastic pollution at sea and in seabirds off southern
Pettit, T. N., G. S. Grant, and G. C. Whittow. Africa (Abstract only). In Proceedings of the workshop on
1981. Ingestion of plastics by Laysan albatross. Auk 98:839- the fate and impact of marine debris; 27-29 November
841. 1984, Honolulu, HI (R. S. Shomura and H. 0. Yoshida,
Plotkin, P., and A. F. Amos. eds.), p. 523. NOAA Tech. Memo. NMFS, NOAA-NMFS-
1990. Effects of anthropogenic debris on sea turtles in the SWFG-5 4.
northwestern Gulf of Mexico. In Proceedings of the sec- 1986. The incidence and effects of ingested plastic in sea-
ond international conference on marine debris; 2-7 April birds. Ms.C. thesis, Univ. Cape Town, South Africa.
1989, Honolulu, HI (R. S. Shomura,and M. L. Godfrey, 1987a. The effects of ingested plastic on seabirds: correla-
eds.), p. 736-743. NOAA Tech. Memo. NMFS, NOAA-TM- tions between plastic load and body condition. Environ.
NMFS-SWFSC-154. Pollut. 46:119-125.

90 NOAA Tecbnical Report NAHS 108: Marine Debris Survey Manual
1987b. The incidence and characteristics of plastic particles the North Pacific Fur Seal Commission, 28 March-8 April
ingested by seabirds. Mar. Environ. Res. 23:175-206. 1983, Washington, D.C., NOAA, National Marine Mammal
1988a. Effects of.ingested plastic on seabird feeding: evi- Lab., 7600 Sand PointWay NE, Seattle, WA 98115. 90 p.
dence from chickens. Mar. Pollut. Bull. 19:125-128. Scordino, J., G. Beekman, H. Kajimura, K. Yoshida, Y Fujimaka,
1988b. The characteristics and distribution of plastic par- and M. Tomita.
ticles at the sea-surface off the Southwestern Cape Province, 1984. Investigations on fur seal entanglement in 1983 and
South Africa. Mar. Environ. Res. 25:249-273. comparisons with 1981 and 1982 entanglement data, St.
1988c. Intraspecific variation in plastic ingestion by sea- Paul Island, Alaska. Document submitted to the 27th An-
birds and the fiux of plastic through seabird popula- nual meeting of the Standing Scientific Committee, North
tions. Condor 90:446-452. Pacific Fur Seal Commission, 29 March-6 April 1984,
1990a. The effects of ingested plastic and other marine de- Moscow. NOAA, National Marine Mammal Lab., 7600
bris on seabirds. In Proceedings of the second interna- Sand Point Way NE, Seattle, WA 98115. 2_6 p.
tional conference on marine debris; 2-7 April 1989, Hono- Scordino,J., H. Kajimura, N. Baba, and A. Furuta.
lulu, HI (R. S. Shomura and M. L. Godfrey, eds.), p. 1988. Fur seal entanglement studies in 1984, St. Paul Island,
623-634. NOAA Tech. Memo. NMFS, NOAA-TM-NMFS- Alaska. In Fur seal investigations, 1985 (P. Kozloff and
SWFSC-154. H. Kajimura, eds.). p. 70-78. NOAA Tech. Memo., NMFS,
1990b. The marine plastic debris problem off southern Af- F/NWC-146.
rica: types of debris, their environmental effects, and con- Scott, G.
trol measures. In Proceedings of the second international 1972. Plastics packaging and coastal pollution. Int. journal
conference on marine debris, 2-7 April 1989, Honolulu, HI of Env. Studies 3:35-36.
(R. S. Shomura and M. L. Godfrey, eds.), p. 85- 1975. The growth of plastic packaging litter. Int. journal of
102. NOAA Tech. Memo. NMFS, NOAA-TM-NMFS- Env. Studies 7:131-132.
SWFSC-154. Seber, G. A. F.
Ryan, P. G., and S. Jackson. 1982. The estimation of animal abundance and related param-
1987. The life span of ingested plastic particles in seabirds eters. Charles Griffin and Company Ltd., London, 654 p.
and their effect on digestive efficiency. Mar. Pollut. Bull. Shaughnessy, P. D.
18:217-219. 1980. Entanglement of Cape fur seats with man made
Ryan, P. G., A. D. Connell, and B. D., Gardner. objects. Mar. Pollut. Bull. 11:332-336.
1988. Plastic ingestion and PCBs in seabirds: is,there a rela- Shaw, D. G.
tionship? Mar. Pollut. Bull. 19:174-176. 1977. Pelagic tar and plastic in the Gulf of Alaska and Bering
Sadove, S. S., and S.J. Morreale. Sea 1975. Sci. Total Environ. 8:13-20.
1990. Marine mammal and sea turtle encounters with ma- Shaw, D. G., and G. A. Mapes.
rine debris in the New York Bight and the 'northeast 1979. Surface circulation and the distribution of pelagic tar
Atlantic. In Proceedings of the second international con- andplastic. Mar. Pollut. Bull. 10:160-162.
ference on marine debris; 2-7 April 1989, Honolulu, HI (R. Shaw, W.
S. Shomura and M. L. Godfrey, eds.), p. 560-570. NOAA 1990. Summary of marine debris sightings during Canadian
Tech. Memo. NMFS, NOAA-TM-NMFS-SWFSC-1 54. high seas research surveys, 1989-1990. Document pre-
Sameoto, D. D., and L. 0. Jaroszynski. sented at the Annual Meeting of the International North
1969. Otter surface sampler: A new neuston net. J. Fish. Pacific Fisheries Commission, Vancouver, British Columbia,
Res. Bd. Canada 26:2240-2244. Canada, November 1990. Department of Fisheries and
Sanger, G. A. Oceans, Biological Sciences Branch, Pacific Biological Sta-
1974. On the effect of fish net scraps and other oceanic de- tion, Nanaimo, B.C., Canada V9R 5K6.
bris in northern fur seals. Background paper submitted to ShiberJ. G.
the 17th Annual Meeting of the Standing Scientific Commit- 1979. Plastic pellets on the coast of Lebanon. Mar. Pollut.
tee of the North Pacific Fur Seat Commission, 11-25 March Bull. 10:28-30.
1974, Ottawa, Canada, National Marine Mammal Lab., 7600 1982. Plastic pellets on Spain's 'Costa del Sol' beaches.
Sand Point Way N.E., Seattle, WA 98115. 9 p. Mar. Pollut. BullA 3:409-412.
Scheffer, V.B. Shomura, R. S., and H. 0. Yoshida, eds.
1950. Experiments in the marking of seals and sea lions. 1985. Proceedings of the workshop on the fate and impact of
U.S. Fish. Wildl. Serv., Spec. Sci. Rep. Wildl. 4. 33 p. marine debris; 27-19 March 1984, Honolulu, HL NOAA
Schrey, E., and G. J. M. Vauk. Tech. Memo, NMFS NOAA-YM-NMFS-SWFC-54.
1987. Records of entangled Gannets (Sula bassana) at Helgoland, Sileo, L. (chair).
German Bight. Mar. Pollut. Bull. 18(6B):350-352. 1990. Report of the working group on ingestion. In Pro-
Scordino,J. ceedings of the second international conference on marine
1985. Studies on fur seal entanglement, 1981-84, St. Paul debris, 2-7 April 1989, Honolulu, HI (R. S. Shomura and M.
Island, Alaska. In Proceedings of the workshop on the fate L. Godfrey, eds.), p. 1226-1234. NOAA Tech. Memo.
and impact of marine debris; 27-29 November 1984, Hono- NMFS, NOAA-TM-NMFS-SWFSC-154.
lulu, HI (k. S. Shomura and H. 0. Yoshida, eds.), p. 278- Sileo, L., P. R. Sievert, M. D. Samuel, and S. I. Fefer.
290. NOAA Tech Memo. NOAA-TM-NMFS-SWFS-54. 1990. Prevalence and characteristics of plastic ingested by
Scordino,J., and R. Fisher. Hawaiian seabirds. In Proceedings of the second interna-
1983. Investigations on fur seal entanglement in net frag- tional conference on marine debris, 2-7 April 1989, Hono-
ments, plastic bands and other debris in 1981 and 1982, St. lulu, HI (R. S. Shomura and M. L. Godfrey, eds.), p. 665-
Paul Island, Alaska. Background paper submitted to the 681. NOAA Tech. Memo. NMFS, NOAA-TM-NMFS-
26th Annual Meeting of the Standing Scientific Committee SWFSC-154.

Citations 91

Slip, D.J., and H. R. Burton. 1985. Plastic ingestion in the North Atlantic fulmar. Mar.
1990. The composition and origin of marine debris stranded Pollut. Bull. 16:367-369.
on the shores of subantarctic Macquarie Island. In Pro- van Franeker, J. A. and P. J. Bell.
ceedings of the Second International Conference on Ma- 1988. Plastic ingestion by petrels breeding in Ant-
rine Debris, 2-7 April 1989, Honolulu, HI (R. S. Shomura arctica. Mar. Poll. Bull. 19:672-674.
and M. L. Godfrey, eds.), p. 403-416. NOAA Tech. Vauk, G.J. M., and E. Schrey.
Memo. NMFS, NOAA-TM-NMFS-SWFSC-154. 1987. Litter pollution from ships in the German Bight.
Slip, D. J., K. Green, and E. J. Woehler. Mar. Pollut. Bull. 18(6B):316-318.
1990. Ingestion of anthropogenic articles by seabirds at Venrick, E. L., T. W. Backman, W. C. Bartram, C. J. Platt, M. S.
Macquarie Island. Marine Ornithology 18:74-77. Thornhill, and R. E. Yates.
Sokal, R. R., and F. J. Rohlf. 1973. Man-made objects on the surface of the North Pacific
1981. Biometry, 2nd edition. W. H. Freeman and Co., San Ocean. Nature 241:271.
Francisco, Verner, S. S.
Stewart, B. S., and P. K. Yochem. 1990. Handbook for preparing quality assurance project
1985. Entanglement of pinnipeds in net and line fragments plans for environmental measurements, Technical Re-,
and other debris in the Southern California Bight. In Pro- sources, Inc., Rockville, MD.
ceedings of the workshop on the fate and impact of marine Wagner, K. D.
debris, 27-29 November 1984, Honolulu, HI (R. S. 1990. Medical wastes and the beach washups of 1988: Issues
Shomura and H. 0. Yoshida, eds.), p. 315-325. NOAA and impacts. In Proceedings of the second international
Tech. Memo. NMFS, NOAA-TM-NMFS-SWFC-54. conference on marine debris; 2-7 April 1989, Honolulu, HI
1987. Entanglement of pinnipeds in synthetic debris and (R. S. Shomura and M. L. Godfrey, eds.), p. 811-
fishing net and fishing line fragments at San Nicolas and 824. NOAA Tech. Memo. NMFS, NOAA-TM-NMFS-
San Miguel Islands, California, 1978-1986. Mar. Pollut. SWFSC-154.
Bull. 18(6B):336-339. Walker, W. A., and J. M. Coe.
1990. Pinniped entanglement in synthetic materials in the 1990. Survey of marine debris ingestion by odontocete
Southern California Bight. In Proceedings of the second cetaceans. In Proceedings of the second international
international conference on marine debris; 2-7 April 1989, conference on marine debris; 2-7 April 1989, Honolulu, HI
Honolulu, HI (R. S. Shomura and M. L. Godfrey, eds.), p. (R. S. Shomura and M. L. Godfrey, eds.), p. 747-
792-809. NOAA Tech. Memo. NMFS, NOAA-TM-NMFS- 774. NOAA Tech. Memo. NMFS, NOAA-TM-NMFS-
SWFSC-154. SWFSC,154.
Takehama, S. Wallace, N.
1990. Estimation of damages to fishing vessels caused by ma- 1985. Debris entanglement in the marine environment:
rine debris, based on insurance statistics. In Proceedings a review. In Proceedings of the workshop on the fate
of the second international conference on marine debris; and impact of marine Debris; 27-29 November 1984,
2-7 April 1989, Honolulu, HI (R. S. Shomura and M. L. Honolulu, HI (R. S. Shomura and H. 0. Yoshida, eds.),
Godfrey, eds.), p. 792-809. NOAA Tech. Memo. NMFS, p. 259-277. NOAA Tech Memo. NOAA-TM-NMFS-SWFS-
NOAA-TM-NMFS-SWFSC-154. 54.
Trulli, W. R., H. K. Trulli, and D. P. Redford. Wehle, D. H. S., and F. C. Coleman.
1990. Characterization of marine debris in selected harbors 1983. Plastics at sea. Nat. Hist. (Feb.) :20-26.
of the United States. In Proceedings of the second interna- Wilber, R.J.
tional conference on marine debris; 2-7 April 1989, Hono- 1987. Plastics in the north Atlantic. Oceanus 30:61-68.
lulu, HI (R. S. Shomura and M. L. Godfrey, eds.), p. S04- Willoughby, N. G.
324. NOAA Tech. Memo. NMFS, NOAA-TM-NMFS- 1986. Man-made litter on the shores of the Thousand Island
SWFSC-154. Archipelago, Java. Mar. Pollut. Bull. 17:224-228.
U.S. EPA (U.S. Environmental Protection Agency). Wolfe, D. A.
1987. Standard operating procedures for the collection and 1987. Persistent plastics and debris in the ocean: an inter-
at-sea processing of neuston samples. SOP No. 4-35, Revi- national problem of ocean disposal. Mar. Pollut. Bull.
sion No. 01, U.S. EPA Office of Marine and Estuarine Pro- 18(6B):303-305.
tection, Incineration-at-Sea Program. Wong, C. S., D. R. Green, and W.J. Cremey.
1990a. Methods to manage and control plastic wastes, report 1974. Quantitative tar and plastic waste distributions in the
to Congress. Office of Solid Waste/Office of Marine and Pacific Ocean. Nature 247:30-32.
Estuarine Protection, EPA/530-SW-89-051. Washington, Wong, C. S., D. MacDonald, and W. J. Cretney.
D.C. 1976. Distribution of tar and other particulate pollutants
1990b. Final data re.p;rt for the study of floatables in U.S. along the Beaufort Sea coast. Beaufort Sea Project, Victoria,
waters (Harbbr Studies Program) November 1988 through British Columbia, Canada. Beaufort Sea Technical Report
February 1989. Office of Marine and Estuarine Protection, No. 13.
EPA 503/4-90-003. Washington, D.C. 139 pp. + Appendices. Yagi, N., and M. Nomura.
van Dolah, R. F., V. G. Burrell, Jr., and S. B. West. 1988. Sighting survey on marine debris in the mid-western
1980. The distribution of pelagic tar and plastics in the Pacific from 1977 through 1986. In Proceedings of the
South Atlantic Bight. Mar. Pollut. Bull. 11:352-356. North Pacific Rim fisherman's conference on marine de-
van FranekerJ. A. bris; 13-16 October 1987, Kailua-Kona, HI (D. L. Alverson
1983. Plastics-een bedreiging voor zeevogels (with English and J. A. June, eds.), p. 130-142. Natural Resources Con-
summary). Nbr. NSO 4:41-61. sultants, Seattle, WA.

92 NOAA Technical Report NMFS 108: Marine Debris Survey Manual

Yoshida, K, and N. Baba. international conference on marine debris; 2-7 April 1989,
1985a. A survey of drifting stray fishing net fragments in the Honolulu, HI (R. S. Shomura and A L. Godfrey, eds.), p.
northern Sea of Japan (Western Pacific Ocean). 325-330. NOAA Tech. Memo. NMFS, NOAA-TM-NMFS-
Document submitted to the 28th Meeting of the Standing SWFSC-154.
Scientific Committee of the North Pacific Fur Seal Commis- Zonfrillo, B.
sion, Tokyo, 4-12 April 1985,13 pp. 1985. Petrels eating contraceptives, polythene and plastic
1985b. Results of a survey on drifting fishing gear or fish net beads. Br. Birds 78:360-351.
pieces in the Bering Sea. Document submitted to the 28th Zsolnay, A.
Meeting of the Standing Scientific Committee of the North 1987. Spatial and temporal variation of pelagic tar in the
Pacific Fur Seal Commission, 4-12 April 1985, Tokyo, 13 pp. Mediterranean Sea. Chemosphere 16:399-404.
Yukinawa, M., and S. Min.
1990. Preliminary report on the distribution of small-sized
marine debris in Suruga Bay. In Proceedings of the second

Guide for Authors
NOAA Technical Report NMFS

PREPARA TION footnoted with italic lower case letters (a, b, c) for general com-
ment, with asterisks for probability in statistical data. Because
Title of manuscript should be as brief as possible (between 6 and tables are typeset, they need only be submitted typed and
12 words) and should not include the Latin binomial except formatted.
when it is needed for clarification, i.e. to distinquish between spe-
cies sharing the same common name or between species known Figures include line illustrations and photographs. Unless photo-
internationally by different common names. graphs are submitted on glossy paper with good contrast, we can-
not guarantee a good final printed copy. Figures must be labeled
Abstract should be one paragraph of about 200 words or less. with author's name and number of figure. Lettering within the
It should state the main scope of the paper but emphasize its con- figure should not be so heavy and large as to upstage the impact
clusions and relevant findings. Because abstracts are circulated of the entire figure. Avoid "outlier" words in a figure which take
by abstracting agencies, it is important that they represent the up needless space. Each figure must include a legend that explains
research clearly and concisely. all symbols and abbreviations. These figure legends should
be typewritten on a separate piece of paper at the end of the
Text must be typed double-spaced throughout and should be manuscript.
divided into the following sections: Introduction, Materials and
Methods, Results, Discussion (or Conclusions), and Acknowl- Copies of Published Reports are available free of charge to the
edgments. Headings within each section must be short, reflect senior author (50 copies) and to his or her laboratory (50). If the
a logical sequence, and follow the rules of multiple subdivision article is part of a proceedings or collection of articles, the senior
(i.e., there can be no subdivision without at least two items of author receives 50 free copies of the individual article and one
subdivision). The entire text should be intelligible to interdisci- copy of the collective work. Additional copies (reprints or oflprints)
plinary readers; therefore, all acronyms, abbreviations, and of an article may be purchased in lots of 100 at a cost of $1.40
technical terms should be defined the first time they are used. per page at the time of first printing. Additional copies of a
The scientific names of species must be written in full the first collective work must be sought from the editor assigned to coor-
time they are mentioned and abbreviated in subsequent refer- dinate the research papers for submission as a NOAA Technical
ences. Footnotes should be avoided as much as possible. We follow Report.
the U.S. Government Printing Office, SyleManual (1984 ed.) and the
CBE Style Manual (5th ed.) for general format and style; the
American Fisheries Society's most recent edition of Common and SUBMISSION
Scientific Names of Fishes from the United States and Canada for fish
nomenclature. Dates should be written as 11 November 1991. Send manuscript (original and two copies) to the Scientific Editor:
Measurements should be expressed in metric units, e.g., tons as Dr. Linda Jones, Scientific Editor
metric tons (t), but if the work is in British long tonnes, please National Marine Maninial Laboratory, F/AKC3
make this fact explicit to the reader. The numeral one (1) should National Marine Fisheries Service, NOAA
be typed as a one, not a lower-case el (1). 7600 Sand Point Way NE
Citations comprise both unpublished and published works. Seattle, WA 98115-0070.
Authors are advised to avoid references to nonstandard material Once the manuscript has been accepted for publication, you will
such as internal and project reports wherever possible. For these be asked to submit a software copy of your manuscript to the
works, include whether they are available from NTIS (National Scientifc Editor. The software copy should be submitted in ASCII
Technical Information Service) or from some other public format (i.e., in a MS-DOS "print" or "nondocument" file) and
depository. Personal communications and unpublished data must should be placed on a 5.25-inch (preferably) or 3.5-inch disk that
be cited in parentheses in the text with full address of com- is double-sided, double or high density, and.that is compatible
municator or author. Follow the name-and-year system for cita- with either IBM or Apple Macintosh systems.
tion format. In the text, cite Smith and Jones (1977) or (Smith
and Jones 197 7). If there is a sequence of citations, list chron- Once the manuscript is being typeset and prepared for publica-
ologically: Smith 1932; Green 1947; Smith and Jones 1985. tion, all inquiries should be made to the Managing Editor:
Abbreviations of serials should conform to abbreviations given
in Serial Sourcesfor Biosis Data Base". AUTHORS ARE RESPONSIBLE Sharyn Matriotti, Managing Editor
FOR THE ACCURACY AND COMPLETENESS OF ALL CITATIONS. NOAA Technical Reports NMFS
National Marine Fisheries Service
Tables should not be excessive in size and must be cited in Scientific Publications Office, F/NWR1
numerical order in the text. All unusual symbols must be ex- 7600 Sand Point Way NE
plained in the table heading. Other incidental comments can be Seattle, WA 98115-0070.

UNITED STATES
DEPARTMENT OF COMMERCE
NATIONAL OCEANIC AND ATMOSPHERIC BULK RATE
ADMINISTRATION POSTAGE & FEES PAID
NATIONAL MARINE FISHERIES SERVICE
SCIENTIFIC PUBLICATIONS OFFICE U.S. Department of Commerce
BIN C15700 Permit No. G-19
SEATTLE, WA 98115
OFFICIAL BUSINESS
Penalty for Private Use, $300

NOAA SCIIENTIFTC AND TECIINICAL PUBLICATIONS

7he National Oceanic and Atmospheric Acbriinistration was established as part of the Department of Commerce
oil October 13, 1970. The mission responsibilities of NOAA are to assess the socioeconomic impact of natural and
technological changes in the environment and to monitor and predict the state of the solid Earth, the oceans and their
living resources, the atmosphere, and the space environment of the Earth.

The major components of NOAA regularly produce various types of scientific and technical information in the
following kinds of publications:

PROFESSIONAL PAPERS-Important definitive research TECHNICAL SERVICE PUBLICATIONS-Reports ron-
results, major techniques, and special investigations. taining data, observations, instructions, etc. A partial
listing includes data serials; predictions and outlook
CONTRACT AND GRANT REPORTS-Reports prepared periodicals; technical manuals, training papers, planning
by contractors or grantees under NOAA sponsorship. reports, and information serials; and miscellaneous
technical publications.
ATLAS-Presentation of analyzed data generally in the TECHNICAL REPORTS-Journal quality with extensive
form of maps showing distribution of rainfall, chemical details, mathematical developments, or data listings.
and physical conditions of oceans and atmosphere, distribu- TECHNICAL MEMORANDUMS-Reports of prelim-
tion of fishes and marine mammals, ionospheric condi- inary, partial, or negative research or technology results,
tions, etc. interim instructions, and the like.

4011 Information on availability ofNO4A publications can be obtainedfrom:
A U.S. Department of Commerce
R IV,
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
IVI I @11 1111 1@ 1111
3 6668 00003 7319