Technical ReportPDF Available

Abandoned, Lost and Discarded Gillnets and Trammel Nets. Methods to Estimate Ghost Fishing Mortality, and Status of Regional Monitoring and Management

Authors:

Abstract and Figures

The ecological and socioeconomic problems caused by abandoned, lost and discarded fishing gear (ALDFG) are increasingly of concern. Marine gillnets and trammel nets, which have relatively high ghost fishing potential, are globally important gear types. Used primarily by coastal, artisanal, small-scale fisheries worldwide, gillnet and trammel net fisheries supply about a fifth of global marine fisheries landings. The Food and Agriculture Organization of the United Nations and the Global Programme of Action for the Protection of the Marine Environment from Land-based Activities, hosted by the United Nations Environment Programme, as Secretariat for the Global Partnership on Marine Litter, commissioned this study to identify best practices to estimate ghost fishing mortality rates and levels, priority research needs, and the status of international monitoring and management of ALDFG and ghost fishing by marine gillnet and trammel net fisheries. Accurate estimates of total ghost fishing mortality levels can be made given quality data on the density of ALDFG retaining fishing efficiency, duration of ghost fishing efficiency, and total ghost fishing mortality level of a unit-of-effort of ALDFG over the full period that the derelict gear retains fishing efficiency. Recommendations to improve estimates of regional and global rates and levels of ghost fishing from ALDFG from marine gillnet and trammel net fisheries were made. An assessment was made and opportunities were identified to improve intergovernmental organizations’ data collection protocols and management measures to prevent and remediate ALDFG and ghost fishing by marine gillnets and trammel nets.
Content may be subject to copyright.
Problems resulting from abandoned, lost and discarded fishing gear (ALDFG) from marine
gillnet and trammel net fisheries is increasingly of concern. Marine gillnets and trammel
nets, which have relatively high ghost fishing potential, are globally important gear types,
supplying about a fifth of global marine fisheries landings. The study describes and
evaluates approaches to estimate ghost fishing mortality rates and levels and reviews the
status of international monitoring and management of ALDFG and ghost fishing by
marine gillnet and trammel net fisheries. The report recommends methods to estimate
ghost fishing rates and levels, identifies research priorities, and recommends future action
to enhance data collection and management to prevent and remediate ALDFG and ghost
fishing by marine gillnets and trammel nets.
600
FAO
FISHERIES AND
AQUACULTURE
TECHNICAL
PAPER
Abandoned, lost and
discarded gillnets and
trammel nets
Methods to estimate ghost fishing mortality, and the status of regional
monitoring and management
600
FAO
ISSN 2070-7010
Abandoned, lost and discarded gillnets and trammel nets – Methods to estimate ghost fishing mortality, and the status of regional monitoring and management
I5051E/1/10.15
ISBN 978-92-5-108917-0
9 789 2 5 1 0 8 917 0
ISSN 2070-7010
Cover photograph:
Top row, left to right: sperm whale entangled in a drift gillnet (Alberto Romero/Marine Photobank); artisanal gillnet
fishing vessel, Solomon Islands (Wolcott Henry 2005/Marine Photobank); decomposed trevally caught in a ghost net,
Muscat, Damaniyat Islands, Oman (Sijmon de Waal/Marine Photobank).
Bottom row, left to right: removing salmon from a gillnet, Bristol Bay, Alaska, the United States of America (Karen
Ducey/NMFS/NOAA Photo Library); derelict gillnet, Oahu, Hawaii, the United States of America (Frank Baersch/Marine
Photobank); seabird caught in derelict net (Dave Peake/Marine Photobank).
Abandoned, lost and
discarded gillnets and
trammel nets
Methods to estimate ghost fishing mortality, and the status of
regional monitoring and management
Eric Gilman
FAO Consultant
Honolulu, the United States of America
Francis Chopin
Fishing Operations and Technology Branch
FAO Fisheries and Aquaculture Department
Rome, Italy
Petri Suuronen
Fishing Operations and Technology Branch
FAO Fisheries and Aquaculture Department
Rome, Italy
and
Blaise Kuemlangan
Development Law Branch
FAO Legal Office
Rome, Italy
FOOD AND AGRICULTURE ORGANIZATION OF THE UNITED NATIONS
Rome, 2016
FAO
FISHERIES AND
AQUACULTURE
TECHNICAL
PAPER
600
The designations employed and the presentation of material in this information product do not imply the
expression of any opinion whatsoever on the part of the Food and Agriculture Organization of the United
Nations (FAO) concerning the legal or development status of any country, territory, city or area or of its
authorities, or concerning the delimitation of its frontiers or boundaries. The mention of specific companies
or products of manufacturers, whether or not these have been patented, does not imply that these have
been endorsed or recommended by FAO in preference to others of a similar nature that are not
mentioned.
The views expressed in this information product are those of the author(s) and do not necessarily reflect the
views or policies of FAO.
ISBN 978-92-5-108917-0
ISSN 2070-7010
© FAO, 2016
FAO encourages the use, reproduction and dissemination of material in this information product. Except
where otherwise indicated, material may be copied, downloaded and printed for private study, research and
teaching purposes, or for use in non-commercial products or services, provided that appropriate
acknowledgement of FAO as the source and copyright holder is given and that FAOs endorsement of
users views, products or services is not implied in any way.
All requests for translation and adaptation rights, and for resale and other commercial use rights should be
made via www.fao.org/contact-us/licence-request or addressed to copyright@fao.org.
FAO information products are available on the FAO website (www.fao.org/publications) and can be
purchased through publications-sales@fao.org.
iii
Preparation of this document
This publication was produced under a 2014 Letter of Agreement between FAO and the
Global Programme of Action for the Protection of the Marine Environment from Land-
based Activities, hosted by the United Nations Environment Programme, as Secretariat
for the Global Partnership on Marine Litter. The study scope covered abandoned, lost
or discarded fishing gear (ALDFG) from marine gillnet and trammel net fisheries. It
describes methods to estimate ghost fishing mortality rates and levels, and synthesizes
estimates of loss rates, density, duration of fishing efficiency and ghost fishing
mortality rates. It also assesses related measures of regional bodies and arrangements
for monitoring and managing ALDFG and ghost fishing. Information for the study
was obtained through a review of published and grey literature, and consultations with
relevant intergovernmental organizations.
iv
Abstract
The ecological and socio-economic problems caused by abandoned, lost and discarded
fishing gear (ALDFG) are increasingly of concern. Used primarily by coastal, artisanal,
small-scale fisheries worldwide, marine gillnets and trammel nets, which have relatively
high ghost fishing potential, account for about one-fifth of global marine fisheries
landings. FAO and the Global Programme of Action for the Protection of the Marine
Environment from Land-based Activities, hosted by the United Nations Environment
Programme, as Secretariat for the Global Partnership on Marine Litter, commissioned
this study to identify best practices to estimate ghost fishing mortality rates and levels,
priority research needs, and the status of international monitoring and management
of ALDFG and ghost fishing by marine gillnet and trammel net fisheries. Accurate
estimates of total ghost fishing mortality levels can be made given quality data on the
density of ALDFG retaining fishing efficiency, duration of ghost fishing efficiency, and
total ghost fishing mortality level of a unit of effort of ALDFG over the full period that
the derelict gear retains fishing efficiency. Recommendations to improve estimates of
regional and global rates and levels of ghost fishing from ALDFG from marine gillnet
and trammel net fisheries were made. An assessment was made and opportunities were
identified to improve intergovernmental organizations’ data collection protocols and
management measures to prevent and remediate ALDFG and ghost fishing by marine
gillnets and trammel nets.
FAO. 2016. Abandoned, lost and discarded gillnets and trammel nets: methods to
estimate ghost fishing mortality, and the status of regional monitoring and management,
by Eric Gilman, Francis Chopin, Petri Suuronen and Blaise Kuemlangan. FAO Fisheries
and Aquaculture Technical Paper No. 600. Rome. Italy.
v
Contents
Preparation of this document iii
Abstract iv
Acknowledgements vii
Acronyms and abbreviations viii
Executive summary x
1. Introduction 1
1.1 ALDFG causes and effects 1
1.2 International efforts to address ALDFG 4
1.3 Methods to mitigate ALDFG 6
1.4 Factors affecting ghost fishing efficiency and duration 9
1.5 Study scope and aims 11
2. Catching process, design and operation of marine drift and set
gillnets and trammel nets 13
2.1 Gillnets 15
2.2 Trammel nets 17
2.3 Combination gillnets – trammel nets 18
3. Methods and estimates 19
3.1 Methods and results 19
3.2 Discussion 26
3.2.1 Methods and estimates of rates of abandonment, loss and discarding
and use in estimating ALDFG density 26
3.2.2 Methods and estimates of the density of ALDFG 28
3.2.3 Methods and estimates of ghost fishing mortality rates and duration
of fishing efficiency 30
3.2.4 Research priorities for robust regional and global estimates of gillnet
and trammel net ghost fishing mortality levels 36
4. Monitoring and management by regional fishery bodies and
arrangements 39
4.1 Introduction, study scope and methods 39
4.2 Results: RFB/A monitoring and management of ALDFG and ghost
fishing 41
4.2.1 General Fisheries Commission for the Mediterranean (GFCM) 43
4.2.2 Indian Ocean Tuna Commission (IOTC) 44
4.2.3 International Commission for the Conservation of Atlantic Tunas
(ICCAT) 44
4.2.4 Joint Norwegian-Russian Fisheries Commission (JNRFC) 45
4.2.5 North Atlantic Salmon Conservation Organization (NASCO) 46
4.2.6 North Pacific Anadromous Fish Commission (NPAFC) 47
4.2.7 Pacific Salmon Commission (PSC) 48
4.2.8 Regional Commission for Fisheries (RECOFI) 48
4.2.9 Southern Indian Ocean Fisheries Agreement (SIOFA) 49
4.2.10 Western and Central Pacific Fisheries Commission (WCPFC) 49
vi
4.3 Discussion and conclusions 50
4.3.1 RFB/A mandate to monitor and control ALDFG 50
4.3.2 RFB/A logbook and observer data collection protocols on ALDFG 50
4.3.3 RFB/A controls of ALDFG and ghost fishing 51
4.3.4 Monitoring and controlling ALDFG and ghost fishing by bilateral and
multilateral arrangements and bodies 53
5. Summary and recommendations 55
6. References 61
vii
Acknowledgements
The authors are grateful for assistance and information provided by: Robin Allen,
South Pacific Regional Fisheries Management Organisation; Anthony Beeching,
Western and Central Pacific Fisheries Commission; Simon Brockington and David
Mattila, International Whaling Commission; Ricardo Federizon, Northwest Atlantic
Fisheries Organization; Martin Hall, Inter-American Tropical Tuna Commission;
Pilar Hernandez, General Fisheries Commission for the Mediterranean; Miguel
Herrera, Indian Ocean Tuna Commission; Peter Hutchinson, North Atlantic Salmon
Conservation Organization; Bob Kennedy, Commission for the Conservation of
Southern Bluefin Tuna; Bruce Leaman, International Pacific Halibut Commission;
Loh-Lee Low, Convention on the Conservation and Management of the Pollock
Resources in the Central Bering Sea; Vladimir Radchenko, North Pacific Anadromous
Fish Commission; Keith Reid, Commission for the Conservation of Antarctic Marine
Living Resources; and Ben van Zyl, South East Atlantic Fisheries Organisation. The
report benefited considerably from formal expert peer reviews by: Dr Milani Chaloupka
(Ecological Modelling Services and University of Queensland), Joan Drinkwin and Kyle
Antonelis (Northwest Straits Foundation), David Laist (Marine Mammal Commission),
Graeme Macfadyen (Poseidon), and Heidi Savelli (United Nations Environment
Programme).
viii
Acronyms and abbreviations
ALDFG abandoned, lost or discarded fishing gear
BRD bycatch reduction device
CCAMLR Commission for the Conservation of Antarctic Marine Living
Resources
CCBSP Convention on the Conservation and Management of the Pollock
Resources in the Central Bering Sea
CCSBT Commission for the Conservation of Southern Bluefin Tuna
CMM conservation and management measure
Code FAO Code of Conduct for Responsible Fisheries
COFI FAO Committee on Fisheries
CPUE catch per unit of effort
CV coefficient of variation
EEZ exclusive economic zone
FAD fish aggregating device
FAO Food and Agriculture Organization of the United Nations
GFCM General Fisheries Commission for the Mediterranean
GPA Global Programme of Action for the Protection of the Marine
Environment from Land-Based Activities
IATTC Inter-American Tropical Tuna Commission
IBSFC International Baltic Sea Fishery Commission
ICCAT International Convention for the Conservation of Atlantic Tunas
ICES International Council for the Exploration of the Sea
IGO intergovernmental organization
IOTC Indian Ocean Tuna Commission
IMO International Maritime Organization
IPHC International Pacific Halibut Commission
IUU illegal, unreported and unregulated (fishing)
IWC International Whaling Commission
JNRFC Joint Norwegian–Russian Fisheries Commission
MARPOL International Convention for the Prevention of Pollution from Ships
NAFO Northwest Atlantic Fisheries Organization
NASCO North Atlantic Salmon Conservation Organization
NEAFC North East Atlantic Fisheries Commission
NPAFC North Pacific Anadromous Fish Commission
PSC Pacific Salmon Commission
PSMA Agreement on Port State Measures to Prevent, Deter and Eliminate
Illegal, Unreported and Unregulated Fishing
RECOFI Regional Commission for Fisheries
ix
RFB/A regional fishery body and/or arrangement
RFMO regional fisheries management organization
ROV remotely operated vehicle
SEM standard error of the mean
SEAFO Southeast Atlantic Fisheries Organisation
SIOFA Southern Indian Ocean Fisheries Agreement
SPRFMO South Pacific Regional Fisheries Management Organisation
UNCLOS United Nations Convention on the Law of the Sea of 10 December 1982
UNEP United Nations Environment Programme
UNFSA Agreement for the Implementation of the Provisions of the United
Nations Convention on the Law of the Sea of 10 December 1982
Relating to the Conservation and Management of Straddling Fish
Stocks and Highly Migratory Fish Stocks
UNGA United Nations General Assembly
WCPFC Western and Central Pacific Fisheries Commission
x
Executive summary
Abandoned, lost and discarded fishing gear (ALDFG) causes substantial ecological and
socio-economic problems. Ghost fishing, one problem resulting from ALDFG, has
received increasing international attention in the past decade. Ghost fishing mortality
is infrequently accounted for in fisheries management, potentially compromising
the accuracy of population and stock assessment models and the efficacy of harvest
strategies. Ghost fishing by ALDFG removes both target and non-target species.
Species with relatively low fecundity and other life-history characteristics that make
them particularly sensitive to anthropogenic mortality sources are also subject to ghost
fishing mortality. These include species of seabirds, sea turtles, marine mammals and
elasmobranchs, some of which are endangered, threatened or protected. Ghost fishing
mortalities are also a source of wastage, and reduce the sustainable production of fishery
resources and economic opportunities for the marine capture sector. There are also
social welfare issues relating to ghost fishing mortality of flagship megafauna, as well as
the time some organisms caught in ALDFG take to die relative to captures in in-use gear.
Ghost fishing is most problematic in gillnet, entangling trammel net and other passive
fishing gear types, where the capture process relies on the movement of organisms into
the gear. Used worldwide primarily by coastal, artisanal, small-scale fisheries, about one-
fifth of global marine fisheries landings comes from gillnet and trammel net fisheries.
Fishing mortality caused by ALDFG has substantial adverse ecological and socio-
economic effects. Marine gillnets and trammel nets have relatively high ghost fishing
potential. Recognizing this, FAO and the Global Programme of Action for the
Protection of the Marine Environment from Land-based Activities, hosted by the
United Nations Environment Programme, as Secretariat for the Global Partnership on
Marine Litter, commissioned this study. Its purpose is to: identify best practice methods
for estimating ghost fishing rates and levels; determine priority research needs; and
assess the status of international monitoring and management of ALDFG and ghost
fishing by marine gillnet and trammel net fisheries.
1
METHODS AND ESTIMATES
This study reviews the methods employed to estimate: rates of gear abandonment, loss
and discarding; density of derelict gear in spatially explicit sites; duration of fishing
efficiency; and ghost fishing mortality rates by ALDFG. It identifies best practice
methods for reducing uncertainty. It synthesizes findings from past studies, providing
an understanding of the degree of dispersion in estimates and of the severity of ALDFG
and ghost fishing. The study identifies priority information gaps to provide robust
estimates of regional and global ghost fishing mortality rates and levels by ALDFG.
At some sites and under certain conditions, ALDFG can result in substantial
ghost fishing removals of both market and non-market species, including endangered,
threatened and protected species and other species of conservation concern. Relative
to some other gear types and to other collateral indirect sources of fishing mortality,
there has been good progress in developing methods to estimate the duration of fishing
efficiency and ghost fishing mortality rates in gillnets and trammel nets. However, a
wide variety of assessment methods and units for reporting results have been employed
in the sparse number of relevant studies. This precludes meaningful comparisons of
1
Hereinafter, to avoid repetition and unless otherwise stated, the text refers to ALDFG and ghost fishing
solely by gillnets and trammel nets.
xi
findings between most studies and prevents pooling of datasets to support large-scale
temporal and spatial analyses. Studies are largely dated and may not characterize
ALDFG and ghost fishing in contemporary fisheries. Studies have been spatially and
temporally patchy, with very large dispersion in estimates.
An accurate estimate of the total level of ghost fishing mortality in a spatially explicit
site over a selected period can be made given data on: the density of ALDFG (unit
amount per unit of area of seafloor and/or per unit volume of water column) retaining
fishing efficiency; area of the site; duration of ghost fishing efficiency; and total ghost
fishing mortality level of a unit of effort of ALDFG over the full period that the
derelict gear continues to retain fishing efficiency. When combined with information
on the amount and spatial distribution of fishing effort, information on the rate of
abandonment, loss and discarding of fishing gear can be used to estimate the density of
ALDFG by gear type in a selected area. Study methods designed to be representative
across sites and periods that have variability in potentially significant explanatory
factors affecting ALDFG and ghost fishing mortality rates and levels are more likely to
characterize a site accurately.
Rates of loss, abandonment and discarding
Fisher surveys have been the most common method to estimate rates of gear loss.
They provide a critically important rough order-of-magnitude approximation where
previously little or no information was available. Based on a small number (n=10) of
comparable estimates from fisher surveys, marine gillnet and trammel net fishers lose
an average of 1percent of their gear (e.g. 1percent of gear is lost per vessel per year),
but with very high dispersion in estimates (38percent coefficient of variation, CV). Few
studies have estimated rates of gear abandonment and discarding. Estimates have not
been based on data from experiments, observer programmes or logbook programmes,
which could validate qualitative estimates derived from fisher surveys and provide
more certain findings. Findings are of higher certainty when methods account for the
proportion of initially lost gear that was subsequently recovered, the generation of
ALDFG from illegal fishing, and the transport of ALDFG into and out of a spatially
explicit site.
Density
Fisher surveys have also been used to estimate the density of ALDFG in a spatially
explicit area. However, surveys of fishing grounds are a more accurate method,
including via towing “creeper” grappling devices and various in situ survey methods
such as observations by divers, and by sonar, video and photography deployed from
marine vessels, towed structures, crewed submersibles and underwater remotely
operated vehicles. Estimates obtained from surveys of a subset of a fishing ground can
be extrapolated to a larger area of interest, such as an entire fishing ground. Some studies
have explicitly accounted for an estimate of error of the survey method (the proportion
of derelict gear present in a study site that the survey method did not identify). There
was high variability in estimates of the density of ALDFG from gillnet and trammel net
fisheries. Four studies that used comparable units, of unit length of nets per unit area of
surveyed fishing grounds, found a mean of 4.4km of nets per square kilometre of fishing
grounds, with extremely high variability (86percent CV).
Ghost fishing mortality rates, duration of ghost fishing efficiency
A wide variety of methods have been employed to estimate ghost fishing mortality rates
and the duration of ghost fishing efficiency in ALDFG, or in experimental nets deployed
to simulate derelict gear. Study designs using simulated derelict gear will be more likely
to characterize ghost fishing rates and levels in the commercial fishery if using typical
gear designs and methods, including selecting study sites within typical fishing grounds.
xii
Short-period ghost fishing mortality rates have been estimated by counting the number
of organisms that became newly captured since a previous observation. Monitoring
has been conducted in situ and via repeated net retrieval. There are several sources
of uncertainty associated with the each of the methods used to estimate short-term
catch rates in derelict nets, which should be accounted for to improve the accuracy of
estimates. For example, organisms caught in between two monitoring events that are
completely removed from the net by predators or scavengers, or that escape and later die
due to injuries resulting from the interaction before the net is subsequently monitored,
may not be accounted for in ghost fishing mortality estimates. Estimates made through
periodic retrieval of a subset of derelict nets may also underestimate ghost fishing
mortality rates, as a proportion of catch may drop out during net retrieval.
Exponential regression decay models have been fitted to time series of short-period
ghost fishing catch rate data to estimate the duration of fishing efficiency and the
total ghost fishing mortality level for the estimated duration of fishing efficiency. The
duration of fishing efficiency has been estimated via periodic monitoring of derelict
or simulated derelict gear until the gear is observed to no longer retain any catching
capacity, no longer catches main market species, or retains a small proportion of species-
specific or total catch capacity relative to its initial fishing efficiency or relative to in-use
gear deployed in the same area and time. There was high dispersion in estimates of
ghost fishing mortality rates. The mean number of ghost-caught fishes per unit area
of nets for the full duration of fishing efficiency or to reach 5percent of initial catch
efficiency was 92.8fish per 100m
2
of net (51percent CV) based on a small number of
studies using comparable units (n=5). There was moderate variability in estimates of the
duration of ghost fishing efficiency that used units of time to cease catch efficiency of all
organisms, or to decline to a small percentage of fishing efficiency: The mean of 11study
findings was 35.0weeks (18percent CV). Very few studies have assessed ghost fishing
by driftnets. Estimates of the duration of driftnet fishing efficiency have ranged from
less than a day for small, 50–100m length nets, to three months for 2km length nets.
This suggests that anchored gillnets and trammel nets pose larger problems from ghost
fishing relative to driftnets. Most studies designed to estimate ghost fishing mortality
rates and the duration of fishing efficiency by ALDFG have studied simulated derelict
demersal nets set at coastal sites within commercial fishing grounds at relatively shallow
depths.
The large dispersion in estimates of ghost fishing mortality rates and duration of
ghost fishing efficiency is probably a result of extremely small sample sizes as well
as from pooling data from studies employing variable methods, studying ALDFG
with variable gear designs and materials, and at sites with variable environmental and
physiographic conditions (e.g. flat substrate in shallow water with strong currents and
abundant biofouling organisms, debris and particulate matter vs entangled on three-
dimensional objects in deep water with weak current and limited biofoulers, debris and
particulate matter; site with active towed fishing gear vs site lacking active gear fishing
effort).
Research needs and recommendations
Several information gaps were identified that are critical to producing robust estimates
of regional and global rates and levels of ghost fishing from ALDFG. There were small
sample sizes in available estimates of rates of producing derelict gear, density of derelict
gear, ghost fishing mortality rates and levels, with under-representation by region and
gear type. Many estimates are dated and may not characterize ALDFG and ghost
fishing in contemporary fisheries. Large sources of uncertainty were introduced in
some of these studies. The use of variable units to report estimates prevents pooling of
some records, reducing sample sizes available to estimate means. There are no available
databases estimating regional and global levels of gillnet and trammel net fishing effort.
xiii
As a result of these deficiencies, there would be very high uncertainty in estimates of
regionally and globally averaged ghost fishing mortality rates and levels, especially for
taxa that are rare-event captures, such as marine megafauna. Recommendations were
made to address these deficiencies:
Domestic and regional authorities should harmonize logbook and observer
programme data fields, data collection protocols, and database formats on ALDFG
where they are in place, and fill gaps where currently programmes to monitor
ALDFG are not in place. This would produce larger sample sizes of records of
rates of generating ALDFG, and rates of fishing vessel encounters with ALDFG,
collected using standardized methods. Standardizing data fields, data collection
protocols and database formats facilitates comparisons between regions, enables
pooling of data necessary to support large spatial scale analyses within and across
regions, and enables global standardization of training materials and courses.
More research is needed, using identified best practice methods to minimize
uncertainty, to estimate ghost fishing mortality rates and levels. These studies
should be balanced spatially, temporally and by type of gillnet and trammel net
fishing gear and method.
Studies designed to estimate ghost fishing mortality rates and levels should employ
standardized units to report estimates in order to facilitate pooling.
Meta-analyses of data from relevant compiled studies should be conducted to
produce estimates of generating ALDFG, density of ALDFG, and ghost fishing
mortality rates with increased precision, accuracy and statistical power over
estimates from individual studies.
Robust estimates of regional and global gillnet and trammel net fishing effort
should be developed.
MONITORING AND MANAGEMENT BY BILATERAL AND MULTILATERAL FISHERY
BODIES AND ARRANGEMENTS
The past decade has seen increasing international recognition of the need for multilateral
efforts to address effectively the transboundary problems resulting from ALDFG,
including ghost fishing. To benchmark regional measures for monitoring and mitigating
ALDFG and ghost fishing, this study assessed the data collection protocols and
management measures to prevent and remediate ALDFG and ghost fishing of ten
bilateral and regional bodies and arrangements with the competence to establish binding
management measures for regional marine capture fisheries, and that have competence
over fishery resources that are captured in an active gillnet or trammel net fishery.
Monitoring and management
Of the ten assessed fishery bodies and arrangements, three collect data via logbook
or observer programmes related to ALDFG. Only one of the assessed bodies/
arrangements is explicitly mandated by its convention or agreement text to monitor
and control ALDFG and ghost fishing. More than half of assessed bodies/arrangements
have adopted binding measures that directly or indirectly contribute to avoiding or
remediating ALDFG. However, the six bodies/arrangements with controls in place do
not take advantage of the full range of available tools. Five of 18categories of methods
identified as being of potential use in preventing and remediating ALDFG and ghost
fishing were used by the 10 bodies/arrangements. Five methods used exclusively
to mitigate ALDFG and ghost fishing are not implemented by any of the bodies/
arrangements. Prohibiting the use of gillnet and trammel net gear in part or all of the
area of competence of a body/arrangement, which contributes to reduced ALDFG and
ghost fishing, was the most commonly employed measure. Gear marking to identify
ownership and to increase passive surface gear visibility was the second-most commonly
employed method. Both forms of gear marking contribute to reducing ALDFG.
xiv
Recommendations
Findings identify opportunities to improve regional monitoring and management of
ALDFG and ghost fishing:
Bilateral and multilateral fishery bodies and arrangements can harmonize ALDFG
logbook and observer data collection and reporting protocols where they are in
place, and fill gaps in bodies and arrangements lacking procedures to collect and
report this information.
For bodies and arrangements lacking binding measures to manage ALDFG and
ghost fishing, members can raise awareness of the impacts of ALDFG and ghost
fishing, and learn from the experiences of bodies and arrangements that have made
progress in adopting relevant measures, with an aim to harmonize management
systems to achieve consistency and compatibility.
Through consideration of the full suite of complementary methods, members of
fishery bodies and arrangements should consider adopting management measures
that directly and indirectly prevent and remediate ALDFG and associated ghost
fishing. These methods include, inter alia:
− Preventive methods:
- Gear marking systems to identify ownership and to increase surface gear
visibility, where adoption of a global standard for gear marking would facilitate
consistent implementation regionally and nationally.
- Technology to avoid unwanted gear contact with the sea bed.
- Technology to track gear position and gear used to mark passive gear location
that is designed to minimize the risk of loss owing to contact by passing
vessels.
- Gear designs and materials that reduce the risk of gear loss.
- Input controls, including limits on gear soak time.
- Periodic or constant attendance by fishers while the gear is soaking.
- Marine spatial and temporal planning, including to separate passive and mobile
gears to avoid gear conflicts and concomitant gear loss, and to phase out gillnet
and trammel net fishing at sites with high risk of gear snagging on submerged
features.
- Deterrents of illegal, unreported and unregulated (IUU) fishing.
- Raised member awareness and incentives for compliance with the prohibition
on intentional abandonment and discarding of fishing gear at sea under the
International Convention for the Prevention of Pollution from Ships.
- Economic incentives for proper disposal of unwanted gear and disincentives
for fishers to generate ALDFG.
- Raised member awareness that adequate (affordable and accessible) port
reception and recycling facilities for unwanted “retired” fishing gear contributes
to preventing ALDFG.
- Programmes to train new fishery entrants to minimize the likelihood of gear
loss and augment capacity to recover lost and abandoned gear.
− Remedial methods:
- Raised member awareness that adequate port reception and recycling facilities
incentives the reporting, retrieval and delivery to these facilities of ALDFG
encountered at sea.
- Programmes for ALDFG detection, reporting and safe retrieval.
- Programmes to disable the ghost fishing efficiency of ALDFG.
- Gear designs and fishing practices that reduce ghost fishing catch and mortality
rates of species of conservation concern.
- Less-durable and degradable gear to reduce ghost fishing duration, if determined
to outweigh costs of increased introduction of synthetic compounds into
marine ecosystems and increased rates of gear loss and retirement.
1
1. Introduction
1.1 ALDFG CAUSES AND EFFECTS
Abandoned, lost and discarded fishing gear (ALDFG), also called derelict fishing gear,
cause substantial ecological and socio-economic problems. An estimated 6.4 million
tonnes of marine debris are added to global seas annually (UNEP, 2005a). ALDFG is
estimated to compose less than 10percent of total marine debris by volume at a global
scale but the composition of marine debris and density of ALDFG is highly variable at
small spatial scales (Macfadyen, Huntington and Cappel, 2009; Pham et al., 2014). The
amount, distribution and effects of ALDFG have risen substantially in past decades
with the rapid expansion of fishing effort and fishing grounds, and the transition to
synthetic, more durable and more buoyant materials used for fishing gear (Derraik,
2002; Macfadyen, Huntington and Cappel, 2009; Gilardi et al., 2010).
There are numerous causes, both intentional and unintentional, for fishing gear to
be abandoned, lost or discarded. Losses occur when gear that has been set for fishing is
unintentionally left at sea. All or a portion of lost gear may later be found and retrieved.
Fishers may lose gear when there is contact with passing vessels or with active gear. For
example, gear conflicts occur when passive gear is inadvertently, or intentionally, towed
away or marker buoy moorings are cut by trawlers or dredgers (Laist, 1995, 1997;
Santos et al., 2003a; Hareide et al., 2005; Antonelis, 2012, 2013). For example, gillnet
loss has been documented to be frequent at fishing grounds that also have high bottom
trawling effort, where trawl gear moves or cuts the nets or buoy lines (MacMullen et
al., 2003; Suuronen et al., 2012). There is also evidence of damage to anchored gillnets
and trammel nets from entanglement with passive gear, including demersal longlines
and traps (Erzini et al., 1997; MacMullen et al., 2003). Loss of fish aggregating devices
(FADs) due to interactions with gillnet fisheries has also been documented (Atapattu,
1991). Fishers may lose gear due to the removal of marker buoys, which can occur
when surface gear is cut by passing vessels and by sea ice in poleward regions, and due
to the breakage or malfunction of tracking systems. Considerable gear loss can occur
during natural hazard events, such as hurricanes (e.g. O’Hare, 2001). High wind and
strong currents can push marker buoys under water, causing fishers to lose the gear.
Gear can also be lost and abandoned when the gear becomes snagged on wrecks and
natural submerged features (Breen, 1990; Pawson, 2003; Cho, 2009; Ayaz et al., 2010;
FAO, 2010a). Damage by marine organisms can also lead to gear loss (Vanderlaan,
Smedbol and Taggart, 2011). Improper designs and materials can lead to gear loss,
such as from not properly maintaining gear and not replacing worn components used
to locate the gear. Improper fishing methods can also lead to gear loss, such as new
entrants setting passive gear in areas where it is likely to snag submerged features,
setting gear at grounds where there is a high probability of interaction with mobile
gear, long soak times during which anchored gear moves from its original position, and
where strong currents are prevalent (MacMullen et al., 2003; Antonelis, 2012, 2013).
Gear can also be lost owing to inclement weather or strong currents.
Abandoned fishing gear results when gear that has been set for fishing is intentionally
left at sea and not retrieved. In addition to causing gear to be lost, bad weather may
also result in gear abandonment, if it becomes too dangerous to retrieve the gear.
Fishers may abandon gear when operating illegally and a risk of detection occurs (e.g.
Imamura, 2011). For example, since the United Nations moratorium on large-scale
pelagic drift-nets over 2.5 km long in international waters, illegal, unreported and
unregulated (IUU) high seas driftnetting has been an ongoing problem in some regions
Abandoned, lost and discarded gillnets and trammel nets
2
(UNGA, 1991; Pramod et al., 2008). Fishers may opt to abandon gear, or to refrain
from attempting to locate and retrieve lost gear, when there is insufficient time, or
when it would be too difficult to retrieve the gear, such as when the gear is snagged on
submerged features (MacMullen et al., 2003; Santos et al., 2003a).
Discarded fishing gear is produced when fishers intentionally throw unwanted gear
overboard at sea. Crew may discard unwanted components of gear at sea when deemed
more practical or economical than disposal on shore, especially where port reception
facilities are unavailable. Setting excessive gear can also result in discarding gear. For
example, there may be insufficient room on board for all of the gear, such as when
the space used to store nets when starting a trip is subsequently used as the fish hold
(Hareide et al., 2005; Macfadyen, Huntington and Cappel, 2009).
Achieving sustainable marine fisheries requires reliably estimating and accounting
for all main sources of fishing mortality (FAO, 1995, 2003, 2011a). However, several
components of fishing mortality, including that caused by ALDFG and other indirect
collateral sources that are largely not manifest or readily detectable, are not routinely
accounted for in fisheries management owing to a lack of both adequate data and
accurate estimation methods (Chopin and Arimoto, 1995; ICES, 1995, 2005; Chopin
et al. 1996; Chopin, Inoue and Arimoto, 1996; Gilman et al., 2013). Errors can result
when population and stock assessment models do not account for total fishing
mortality, including from direct and collateral removals caused by ALDFG. This
unaccounted fishing mortality has the potential to compromise the efficacy of fisheries
harvest strategies and the sustainable production of fishery resources, irreparably harm
affected populations and stocks, and cause broader community- and ecosystem-level
effects (Hall, 1996; Punt et al., 2006; Coggins et al., 2007). Mortalities from ghost
fishing by ALDFG are also a source of wastage and reduce economic opportunities
for the marine capture sector (Goñi, 1998; Gilardi et al., 2010; Antonelis et al.,
2011). For example, ghost fishing has been estimated to remove between 0.5percent
and 30 percent of landed catches of market species in various European and North
American fisheries (Laist, 1995, 1997; Sancho et al., 2003; Santos et al., 2003b; Brown
and Macfadyen, 2007). There are also social welfare issues related to ghost fishing
mortality of flagship charismatic marine megafauna, and related to the duration that it
takes organisms caught in ALDFG to die relative to in-use gear (Akiyama, Saito and
Watanabe, 2007; IWC, 2013a; WSPA, 2014).
Ghost fishing occurs when ALDFG continues to catch and kill organisms (Kaiser
etal., 1996; FAO, 2010b; Uhlmann and Broadhurst, 2013; Wilcox et al., 2013). Long-
term “automated re-baiting” occurs when moribund and decomposing organisms
caught in the derelict gear attract scavengers. Feeding by scavengers, in turn, releases
odours that augment attraction to the ALDFG. Some of these scavengers become
caught and eventually decompose, providing a continual source of “bait” until the
ALDFG loses its fishing efficiency. This automated re-baiting can augment the ghost
fishing efficiency of some gear types for some species (Breen, 1990; Kaiser et al., 1996;
Matsuoka, Nakashima and Nagasawa, 2005; FAO, 2010a; Gilman et al., 2013). While
some species are repelled by dead conspecifics (organisms of the same species), others
are attracted to live conspecifics caught in the derelict gear. Moreover, the structure of
ALDFG acts as a FAD, increasing the local abundance of organisms, which increases
ghost fishing efficiency (Breen, 1990; MacMullen et al., 2003).
Ghost fishing mortality is of particular concern for marine megafauna with
K-selected life-history strategies (long life span, slow growth, late sexual maturity, low
fecundity, and low natural mortality rates of subadult and adult age classes), including
seabirds, sea turtles, marine mammals, elasmobranchs (sharks and their relatives) and
some bony fishes (Laist, 1995, 1997; Kaiser et al., 1996; Donohue et al., 2001; Good
et al., 2009, 2010; IWC, 2013a). Mortality in marine capture fisheries, including from
ghost fishing, can contribute to compromising the viability of some populations in
3
Introduction
these groups. Owing to their life-history characteristics, they can decline over short
temporal scales (decades and shorter) and are slow to recover from large declines
(Musick, 1999a, 1999b; Hall, Alverson and Metuzal, 2000; Lewison et al., 2004;
Gilman, Owens and Kraft, 2013c). However, even populations of fecund, widespread
species can be at risk of extirpation (local extinction) owing to unsustainable fishing
mortality and other anthropogenic stressors (Casey and Myers, 1998; Dulvy et al.,
2004; Gilman et al., 2011).
Floating ALDFG and other marine debris can obstruct navigation. Floating derelict
gear can entangle or clog a vessel’s water intake valves, propellers, propeller shaft and
rudder, so stranding vessels and placing vessels and crew in danger. ALDFG can also
foul in-use fishing gear, which may require costly repairs and lost fishing time, and
can result in additional gear loss (Macfadyen et al., 2009; FAO, 2010a). In addition,
floating marine debris can accumulate as mass concentrations in pelagic ecosystems at
ocean convergence zones for extended periods of time, potentially altering community
structure and processes (Derraik, 2002; Macfadyen, Huntington and Cappel, 2009;
FAO, 2010a).
Derelict fishing gear that sinks to the sea bed can adversely affect benthic habitats
of deep-water and shallow coastal areas, including altering microhabitats such as by
obstructing reef crevices.
1
For example, ALDFG can entrap fine sediment, smothering
benthic communities, and can obstruct water flow, creating anoxic areas, which, if
prolonged, can cause substantial mortalities (Parker, 1990; Hall, Alverson and Metuzal,
2000; Levin et al., 2009; Macfadyen, Huntington and Cappel, 2009). When dragged by
currents and wind or during retrieval, it can scour and abrade the sea bed and associated
communities, including vulnerable and ecologically and biologically significant marine
areas such as coastal seagrass beds and coral reefs, and benthic cold water coral reefs
and sponge fields (Rose et al., 2000; Donohue et al., 2001; FAO, 2010a).
Collateral effects of fishing are those that are indirectly caused by various ecological
effects of fishing (ICES, 2005; Gilman et al., 2013). Collateral indirect effects of fishing,
including from ALDFG, are complex and difficult to quantify, in part, because there
is high uncertainty in inferring what factors significantly explain mortalities (Gilman
et al., 2013). For example, ALDFG and other marine debris may transport invasive
alien species, which can disrupt community structure and processes, including causing
niche contraction, declines in abundance and local extirpations of native species (Mack
et al., 2000; Longpierre et al., 2005; Galil, 2007; FAO, 2010a). Organisms entangled
in ALDFG may experience sublethal effects, including reduced mobility. This can
compromise foraging ability and lead to starvation, reduce their ability to avoid
predators, increasing the probability of mortality, and cause lacerations and subsequent
infection, which through synergistic and cumulative effects might eventually lead to the
organism’s demise (Chopin and Arimoto 1995; Suuronen and Erickson, 2010; Gilman
etal., 2013; Uhlmann and Broadhurst, 2013). The ALDFG may be located in critical
habitat where it poses a hazard to wildlife, including in foraging areas, fish spawning
grounds, turtle nesting areas and migration routes (Gilman et al., 2010). It may provide
an unnatural food source for species that remove catch from the derelict gear that are
not typical components of their diet (Gilman et al., 2010). Another collateral effect of
fishing is that synthetic compounds, including microscopic plastic material and toxic
chemicals derived from fishing gear and other marine debris, and from lead in fishing
weights, accumulate in marine food webs (Derraik, 2002; Moore, 2008; Arthur, Baker
and Bamford, 2009; Hammer, Kraak and Parsons, 2012).
Based on preliminary findings of research on artificial drifting FADs (Marsac,
Fonteneau and Ménard, 2000; Hallier and Gaertner, 2008; Dagorn et al., 2013),
1
An estimated 70 percent of total marine debris sinks to the sea bed (FAO, 2010a). Estimates of the
proportion of ALDFG from different gear types that eventually sink are unavailable.
Abandoned, lost and discarded gillnets and trammel nets
4
drifting rafts of marine debris, a proportion of which is ALDFG, might aggregate
marine organisms from a surrounding area and alter their survival probability. The
drifting debris may alter associated organisms’ spatial distributions over hundreds of
kilometres, transporting them to areas outside of their normal distribution, modifying
their diet composition, and changing their behaviour, such as horizontal movements,
vertical habitat use and diel vertical migration cycles. In some regions, FADs also have
the potential to bring organisms to areas of low productivity (Dagorn et al., 2013). The
lack of regional monitoring and controls on the number, density, locations and designs
of FADs (the fishing gear is largely unregulated and unreported) contributes to this
potential collateral effect of drifting rafts of debris (Gilman, 2011).
1.2 INTERNATIONAL EFFORTS TO ADDRESS ALDFG
In the past decade, there has been increasing international recognition of the need
for multilateral efforts to address effectively the transboundary problems resulting
from marine debris including ALDFG (Macfadyen, Huntington and Cappel, 2009;
Kuemlangan, Chopin and d’Offay, 2011). Since 2004, several United Nations General
Assembly (UNGA) resolutions have explicitly recognized problems resulting from
ALDFG, and called upon States and international organizations to take steps to mitigate
these problems. Through its resolutions, UNGA provides mandates for ALDFG
to be addressed globally and by specific intergovernmental organizations (IGOs).
International organizations identified in these UNGA resolutions have included the
International Maritime Organization (IMO), Food and Agriculture Organization of
the United Nations (FAO), FAO Committee on Fisheries (COFI), Regional Seas
Programme of the United Nations Environment Programme (UNEP) and regional
fisheries management organizations (RFMOs). The UNGAs recommended actions
have included: collecting data on gear loss, economic costs, and effects on marine
ecosystems; taking preventive and remedial measures; integrating the issue into relevant
national and regional strategies; and having relevant international organizations adopt
and improve existing measures (UNGA, 2004, 2006a, 2006b, 2007a, 2007b, 2009a,
2009b, 2010a, 2010b, 2011, 2012a, 2012b, 2013a, 2013b, 2014a, 2014b).
In 2011, COFI endorsed FAO’s International Guidelines on Bycatch Management
and Reduction of Discards (FAO, 2011a). The Guidelines included recommendations
for member States to identify, quantify and reduce impacts of mortality from ghost
fishing by identifying this as an objective in fisheries management plans, improving
scientific information on the magnitude and causes of these mortality sources, and
developing technology for assessment and mitigation (FAO, 2011a). COFI had
previously highlighted the importance of ALDFG, including fishing gear marking,
reporting and recovering lost gear, and a regulatory framework to deal with violators
(FAO, 2007a).
In July 2011, IMO, a specialized agency of the United Nations that sets global
standards for the safety, security and environmental performance of international
shipping, adopted amendments to the International Convention for the Prevention of
Pollution from Ships (MARPOL) Annex V Regulations for the Prevention of Pollution
by Garbage from Ships. The amendments entered into force on 1January 2013 (IMO,
2011). Annex V Regulation 3 prohibits the disposal of all plastics, including fishing
gear, into the sea in all locations (IMO, 2011, 2012). Regulation 7 defines exceptions to
the prohibitions during emergency and non-routine situations, where exceptions are
allowed when discharging is necessary to secure the safety of a ship and those on board,
or to save life at sea. Fishing vessel operators are required to record the discharge or
loss of fishing gear in a garbage record book or ship’s log as specified within regulations
7.1 and 10.3.4 of MARPOL Annex V. Moreover, under Annex V Regulation 10.6,
fishing vessel operators are required to report the accidental loss or discharge of fishing
gear, “which poses a significant threat to the marine environment and navigation,” as
5
Introduction
determined by each member State’s government,
2
and provide these reports to the
flag State and, if relevant, also to the coastal State in whose jurisdiction the ALDFG
occurred (IMO, 2011, 2012). AnnexV Regulation 8 obligates governments to provide
adequate port reception facilities for garbage from ships and to facilitate and promote
their use (IMO, 2012).
Members of FAO and the UNGA have mandated FAO to conduct research,
provide technical advice, support regional fishery bodies (RFBs) including RFMOs,
and provide advocacy to address ALDFG. Article 8 of the FAO Code of Conduct
for Responsible Fishing (the Code) specifically addresses MARPOL requirements,
encouraging States to minimize ghost fishing mortality, mark fishing gear to enable
identification of the owner and provide waste reception facilities for fishing gear (FAO,
1995).
The Regional Seas Programme’s UNEP Global Initiative on Marine Litter assisted
regional seas conventions and action plans to organize and implement regional activities
on marine litter, including ALDFG (UNEP, 2012). UNEP launched its Regional Seas
Programme in 1974 by encouraging groups of countries sharing common seas to find
regional solutions to their particular problems. Currently, 143countries participate in
13regional seas conventions and action plans and in 5partner programmes, making
it one of the most globally comprehensive initiatives for the protection of marine and
coastal environments.
Under the United Nations Fish Stocks Agreement (UNFSA), a set of recommended
minimum criteria were produced for RFMO performance assessment. One criterion
includes assessing whether measures are in place to minimize ghost fishing: “Extent to
which the RFMO has adopted measures to minimize … catch by lost or abandoned
gear” (United Nations, 2007 [AnnexII]).
The Fifth International Marine Debris Conference, held in 2011, resulted in
the Honolulu Commitment and the Honolulu Strategy: A Global Framework for
Prevention and Management of Marine Debris. The Honolulu Strategy defined
strategies and actions to reduce adverse effects of marine debris, including to reduce
the amount and impacts of ALDFG (UNEP and NOAA, 2012). In 2012, the Manila
Declaration on Furthering the Implementation of the Global Programme of Action
for the Protection of the Marine Environment from Land-based Activities was
adopted at the Third Intergovernmental Review of the Implementation of the Global
Programme of Action for the Protection of the Marine Environment from Land-Based
Activities (GPA). The Manila Declaration recommended the establishment of the
Global Partnership on Marine Litter, which was launched later in 2012, and focuses
on implementing the Honolulu Strategy by preventing impacts from both land- and
sea-based sources of marine litter (UNEP, 2012). Chapter 4 synthesizes conservation
and management measures (CMMs) related to preventing, remediating and monitoring
ALDFG and ghost fishing from gillnets and trammel nets that have been adopted by
bilateral and regional fishery bodies and arrangements (RFB/As), including RFMOs.
Despite the existence of strong international policy and legal frameworks on
ALDFG and binding measures to mitigate ALDFG and ghost fishing, past performance
2
The IMO 2012 Guidelines for the Implementation of MARPOL Annex V clarify that Party’s
governments should determine if accidentally lost and discharged fishing gear is required to be reported
under AnnexV Regulation 10.6 by considering factors including: (i) the amount of lost and discharged
gear; (ii) the conditions of the marine environment where it was lost or discharged; (iii) the characteristics
of the lost gear, including types, weight and/or length, quantity, material, and buoyancy; and (iv) the
vulnerability of habitat and protected species to gear interactions in the location where the gear was lost/
discharged, e.g. was the gear lost or discharged in a sensitive area such as coral reefs, or in a protected
species’ foraging or breeding area (IMO, 2012). The IMO guidelines use the example of “whole or nearly
whole large fishing gear or other large portions of gear” as lost or abandoned fishing gear that could be
considered to meet the Annex V Regulation 10.6 definition of posing “a significant threat to the marine
environment and navigation” (IMO, 2012).
Abandoned, lost and discarded gillnets and trammel nets
6
assessments have concluded that there are substantial governance deficits, including in
monitoring, surveillance and enforcement, indicating low compliance with international
mandates to prevent and remediate ALDFG (Macfadyen, Huntington and Cappel,
2009; Gilman, Passfield and Nakamura, 2013b; Gilman, 2015).
1.3 METHODS TO MITIGATE ALDFG
Table1 summarizes preventive methods to reduce the incidence of fishing gear from
becoming abandoned, lost and discarded, and remedial methods to reduce the duration
that ALDFG remains in the marine environment. Preventive methods are understood
to be more cost-effective than remedial methods (Macfadyen, Huntington and Cappel,
2009). Moreover, because some remedial methods can compromise economic viability
and practicality, such as the use of biodegradable materials and other methods to disable
ALDFG, efforts focusing on preventive methods and quick recovery of ALDFG are
likely to be more effective (Matsushita et al., 2008; Suuronen et al., 2012). Many of
these methods have broad fisheries management purposes unrelated to ALDFG and
ghost fishing, such as managing fishing mortality of market species and mitigating
bycatch of vulnerable taxa in in-use gear (Figure1). However, their implementation
nonetheless contributes to mitigating ALDFG and ghost fishing.
TABLE1
Preventive methods to avoid and minimize fishing gear from becoming abandoned, lost and discarded, and
curative or remedial methods to reduce the longevity of ALDFG
Method Description
Preventive
Gear marking to
identify ownership,
and to increase
visibility
1
Requirements for marking fishing gear enable competent authorities to identify the owner or
user of ALDFG, which can create a disincentive for the deliberate abandonment and discarding
of unwanted gear and incentive to report abandoned and lost gear (Convention on Conduct of
Fishing Operations in the North Atlantic, 1967; IMO, 1978; Caddy, 1996; FAO, 1991, 1993, 2011a).
Gear marking to increase the visibility of passive gear has also been prescribed by management
authorities in order to reduce the navigational risk to vessel operators, which could contribute
to avoiding accidental gear loss when damaged by passing vessels or active gear (FAO, 1993). A
global standard for marking fishing gear to determine ownership and indicate position has been
proposed but to date has not been endorsed (FAO, 1991, 1993).
Technology to avoid
unwanted gear
contact with seabed
1
Global Positioning System (GPS) and seabed mapping technology help to reduce the likelihood
of gear loss from to unintended contact with the seabed, reducing the probability of accidental
gear loss (FAO, 2010a).
Technology to track
gear position
1
Attaching radar reflectors and radio buoys to fishing gear reduces the risk of losing gear, avoids
interactions with towed gear, and aids in locating lost gear (MacMullen et al., 2003; FAO, 2010a).
Setting gear used to mark the location of passive gear below the sea surface can reduce the risk
of loss from being cut by passing vessels (Macfadyen, Huntington and Cappel, 2009).
Gear technology to
reduce the incidence
of gear loss
1
Changes in fishing gear designs or materials might reduce the incidence of loss. For example, an
alternative design was developed for drift and anchored gillnet floats in an attempt to reduce
the probability of loss of buoys from the nets (Chaves and Silveira, 2014).
Input controls,
including limit on
soak time
1
Limiting the amount of fishing effort or capacity can reduce the quantity of ALDFG. Input
controls include limiting vessel numbers of a specified size, prohibiting new entrants, instituting
buy-back schemes, capping the number of fishing days or sets per year, and eliminating subsidies
that contribute to overcapacity (Gilman, Passfield and Nakamura, 2012). Limiting the length
of gear soak time and requiring the retrieval of gear during closed periods can reduce the
probability of gear loss (Macfadyen, Huntington and Cappel, 2009; FAO, 2011a).
Periodic or constant
observation of
passive gear
1
Requiring passive gear to be under periodic or constant observation, a recommended practice to
increase the probability of sea turtles caught in gear being released alive (Gilman, 2009), can also
reduce the probability of gear loss.
Spatial and temporal
restrictions on
fishing
1
Separating passive and mobile gear sectors temporally and/or spatially to avoid gear conflicts
(i.e. passive gear is towed away by mobile gear), and not using fishing methods in areas where
there is a high probability of loss on submerged features, can reduce the probability of gear
loss (MacMullen et al., 2003; Macfadyen, Huntington and Cappel, 2009; FAO, 2010a). Some
intergovernmental bodies and agreements have adopted measures banning the use of gillnet
and trammel net gear, in some cases for the explicit purpose of avoiding ghost fishing (Chapter 4;
Gilman, 2015).
7
Introduction
Method Description
IUU deterrents
1
Effective deterrents of illegal, unreported and unregulated (IUU) fishing can reduce incentives for
abandoning gear.
Ban intentional
abandonment and
discarding of fishing
gear at sea
2
The international prohibition on intentional discarding and abandonment of fishing gear at sea
under MARPOL Annex V (IMO, 2011, 2012 [MARPOL Annex V Regulation 3, see Section 1.2]) can
be effective if surveillance and enforcement systems elicit strong compliance.
Economic incentives
and disincentives
2
Economic incentives to reduce the incidence of gear becoming abandoned, lost or discarded
include creating a mandatory deposit on new gear, which is returned when unwanted gear is
delivered to an appropriate port reception facility, and not subsidizing the cost for fishers to
replace ALDFG (MacMullen et al., 2003). Or, given sufficient resources for effective monitoring,
penalties that are sufficiently onerous to create an incentive to avoid and reduce the incidence of
ALDFG could be instituted for abandoned, lost or discarded gear.
Port reception
facilities for
unwanted gear
2
Providing accessible and affordable port reception facilities for unwanted fishing gear can reduce
the incentive for at-sea discarding (IMO, 1978; FAO, 2010a).
Training for new
entrants
1
Providing training opportunities for new entrants to fisheries with a high probability of gear
loss can increase skipper capacity to employ best practice gear designs and fishing methods and
minimize the likelihood of gear loss and augment their capacity to recover gear when it is lost or
abandoned (MacMullen et al., 2003).
Remedial
ALDFG port
reception and
recycling facilities
2
In addition to reducing the incentive to discarding old “retired” fishing gear at sea, accessible
and affordable port reception facilities can encourage the reporting, retrieval and delivery of
ALDFG (IMO, 1978; FAO, 2010a; NFWF, 2013). Management authorities have created systems to
report ALDFG and have developed incentives for fishing vessels to retrieve derelict gear at sea
when it has tangled in their propellers or fishing gear, or that they encounter at their fishing
grounds or in transit, deliver it to port reception facilities (Gilman, 2005; Yates, 2007; FAO, 2010a,
2010b, 2011a). There are several programmes designed to create incentives for port disposal
of unwanted gear and of ALDFG retrieved at sea. For example, government agencies in the
Republic of Korea manage an incentive programme, paying fishers to retrieve marine debris
from coastal fishing grounds and deliver it to designated seaports (Cho, 2009). Several ALDFG
reception programmes provide opportunities for reuse by the fishing industry, recycling, and
conversion to energy (Yates, 2007; FAO, 2010b; Recht, 2010; NFWF, 2013). The National Fish and
Wildlife Foundation’s Fishing for Energy programme, for example, has established port reception
facilities for old, retired and derelict fishing gear, recycling metals and converting non-recyclable
materials into energy, and has supported removal and assessment programmes in 41seaports in
ten states in the United States of America (NFWF, 2013). The fishery management authority of the
Unites States of America has established a port reception facility for Hawaii longline fishers’ own
unwanted fishing gear and ALDFG that they collect at sea, which is then incinerated to generate
electricity (Gilman, 2005; Yates, 2007; FAO, 2010b).
Detection and
removal of ALDFG
2
Some intergovernmental bodies and agreements have adopted measures requiring fishing
vessels to have onboard equipment to retrieve ALDFG, for captains to attempt to retrieve
ALDFG generated from their vessel or ALDFG from other vessels that they encounter while at
sea, and to report information on lost gear that could not be retrieved (Chapter 4; Gilman,
2015). Mechanisms for fishers to report ALDFG generated from their vessels or that they
encounter at sea, including regulatory frameworks that allow “no fault” reporting, eliminating
the assessment of penalties for losing gear that would present a disincentive for self-reporting
lost gear, can lead to quick identification and retrieval of ALDFG, where derelict gear retrieval
programmes exist (Good et al., 2009, 2010). Several government agencies and organizations
implement programmes to survey periodically fishing grounds and sensitive marine habitats
along coastlines, coastal benthic habitats and pelagic areas in order to locate and then remove
ALDFG and other marine debris (e.g. Northwest Straits Foundation, 2007; McElwee and
Morishige, 2010; GhostNets Australia, 2012; Suuronen et al., 2012). Methods to search for ALDFG
include aerial surveillance, side-scan sonar, remotely operated vehicles (ROVs) and diver surveys
and tows (Northwest Straits Foundation, 2007; Murphy, 2011; Natural Resources Consultants,
2013). In shallow waters, inflatable lift bags, winches on vessels, and human-powered collection
by scuba diving, hookah and snorkelling have been used to remove ALDFG and other marine
debris (Manuel and Koyanagi, 2011; Natural Resources Consultants, 2013). Offshore ALDFG can
be recovered by dragging grappling devices along the seabed that are designed to snag the
derelict gear (“creepers”), using trawl nets or other fishing gear, using ROVs and removal by
divers (MacMullen et al., 2003; FAO, 2005; Good et al., 2009; Jung et al., 2010; Natural Resources
Consultants, 2013).
Disablement of ghost
fishing efficiency of
ALDFG
2
Programmes periodically use a trawl net or other fishing gear to sweep fishing grounds with
unobstructed low-relief substrate with known ALDFG in order to remove or otherwise damage
derelict gear sufficiently to discontinue its ghost fishing efficiency (MacMullen et al., 2003; FAO,
2005).
TABLE1 (CONTINUED)
Abandoned, lost and discarded gillnets and trammel nets
8
Method Description
Gear technology
designed for
bycatch mitigation
in in-use gear that
also increases ghost
fishing selectivity in
ALDFG
1
Modifications to fishing gear designs to reduce problematic bycatch (bycatch reduction devices
[BRDs]) in in-use gear can also reduce ghost fishing rates of vulnerable species in ALDFG from
these fisheries. For example, reducing net mesh size, reducing gillnet profile (vertical height),
and eliminating or reducing the length of anchored gillnet tiedowns (Figure1) have been used
to reduce turtle capture (Gearhart and Eckert, 2007; Price and Van Salisbury, 2007; Eckert et al.,
2008; Gilman et al., 2010). For example, reducing the length or eliminating the use of tiedowns
and the amount of webbing in demersal gillnets reduces or eliminates the bag of slack webbing,
and this has been found to reduce the incidence of sea turtle entanglement (Gilman et al., 2010),
and might reduce sea turtle ghost fishing mortality in ALDFG from set gillnet fisheries (Figure1).
Increasing gillnet filament diameter, modifying the weaves (e.g. using multimonofilament instead
of single monofilament), using larger floats on the top rope and heavier weights or lead-core
on the bottom rope, and infusing compounds can make the net stiffer (increase net tension),
reducing the likelihood of entangling large organisms (Werner et al., 2006; Larsen, Eigaard and
Tougaard, 2007; Thorpe and Frierson, 2009). Deploying driftnets 2m below instead of at the sea
surface, and using highly visible netting in the upper portion of a surface drift gillnet reduced
seabird catch rates (Hayase and Yatsu, 1993; Melvin, Parrish and Conquest, 2001). Making nets
more visible, such as through net colour, thicker twine diameter, and attaching corks or other
visual markers within the net, has also been shown to reduce bycatch rates of marine mammals
and turtles, but can also reduce target species catch rates (Barlow and Cameron, 2003; Gilman
etal., 2010). Attaching materials such as thick polyester rope and chains to fishing nets, and
infusing nylon nets with metal compounds such as barium sulphate and iron oxide can reduce
cetacean captures. This might occur because the materials increase acoustic reflectivity, increase
the net’s visibility or increase twine stiffness (Trippel et al., 2003; Koschinski et al., 2006; Werner
et al., 2006; Larsen, Eigaard and Tougaard, 2007). Acoustic pingers and alarms are also used to
reduce marine mammal bycatch in gillnets and other fishing gear (e.g. Koschinski et al., 2006;
Werner et al., 2006), and illuminating nets with chemical or battery-operated lightsticks might
reduce bycatch of sea turtles and other vulnerable taxa (Wang, Fisler and Swimmer, 2010), but
would be ineffective in reducing ghost fishing mortality once the energy source has drained.
Less-durable and
degradable gear to
reduce ghost fishing
duration
1
Using less-durable materials (e.g. thinner net twine diameter and weaker material) to produce
a breaking strength that allows large organisms to break free of the gear and escape might
reduce ghost fishing mortality (Gilman et al., 2010). Gear technology has been developed to
reduce the duration of the fishing power of derelict gear via designs that employ degradable
materials. For example, degradable FADs have been designed; degradable cotton fibre is still used
in some gillnet fisheries; attaching floats using biodegradable materials has been trialled in a
demersal gillnet fishery; and degradable escape mechanisms are required in some trap fisheries
(Breen, 1990; Carr, Blott and Caruso, 1992; Chopin et al., 1996; Chanrachkij, Siriraksophon and
Loog-on, 2008; Matsushita et al., 2008; Antonelis et al., 2011). Degradable escape panels and
cords can be used to reduce ghost fishing by traps, which are required in some fisheries (FAO,
2010a). Synthetic gear materials have been developed that can be broken down by microbes and
ultraviolet light (Tabata and Kanehiro, 2004). Simulated derelict demersal gillnets constructed
of multifilament twine have been observed to have a shorter duration of fishing efficiency than
gillnets constructed of monofilament twine (Ayaz et al., 2006). The likelihood that weaker and
degradable fishing gear would increase the frequency that gear components require replacement
and increase gear loss requires consideration.
1
Measure might be adopted for a range of fishery management purposes but contributes to avoiding and remediating ALDFG and
ghost fishing.
2
Measure is implemented specifically to mitigate ALDFG and ghost fishing.
Source: Adapted from Macfadyen, Huntington and Cappel, 2009; FAO, 2010a; Suuronen et al., 2012; Gilman, Passfield and
Nakamura, 2013.
TABLE1 (CONTINUED)
FIGURE 1
Conventional demersal gillnet (left) and modified net without tiedowns (right)
Source: Original drawing by Jeff Gearhart, re-designed by Manuela D’Antoni (FAO).
9
Introduction
1.4 FACTORS AFFECTING GHOST FISHING EFFICIENCY AND DURATION
Gear type is a large, significant explanatory factor in determining species- and size-
specific ghost fishing catch rates and levels. Ghost fishing is thought to be most
problematic in passive fishing gear that was set and subsequently lost or abandoned.
Passive gear types include gillnets, trammel nets, pelagic and demersal longlines, pots
and other trap gear, where the capture process relies on the movement of organisms
into the gear. The catching process of active gear (e.g. purse seines, trawls) typically
ceases when te gear collapses upon detachment from the vessel (Matsuoka, Nakashima
and Nagasawa, 2005; SEAFO, 2009; Gilman et al., 2013). However, there is minimal
empirical information on ghost fishing in ALDFG from active gear, where effective
methods to estimate the frequency of gear loss and the duration and efficiency of ghost
fishing by such gear have not been developed (Gilman et al., 2013). Nevertheless, ghost
fishing has also been observed in ALDFG from active gear, including in seine nets, and
there are observations of marine mammal entanglement in trawl net fragments (Jones,
1995; Donohue et al., 2001; Matsuoka, Nakashima and Nagasawa, 2005). Ghost fishing
in derelict FADs used in purse seine and other fisheries has also been documented
(Chanrachkij, Siriraksophon and Loog-on, 2008; Filmalter et al., 2013). Ghost fishing
mortalities can also occur from discarded bait containing hooks used in both passive
and active gear types (Weimerskirch and Jouventin, 1987).
Various factors affect the ability, efficiency and duration of derelict gear to ghost
fish. The condition of the gear upon being lost/abandoned/discarded is an important
explanatory variable, including whether it was set for fishing or otherwise discarded
and therefore less likely to ghost fish. The location of ALDFG, including the depth,
substrate material, degree of protection from wave energy, presence of features upon
which the gear can become entangled, relative abundance of organisms that are
susceptible to capture in the ALDFG, and relative abundance of biofouling organisms,
debris and particulate matter can all be significant explanatory factors determining
ghost fishing mortality rate and duration. For example, these factors can determine
whether derelict gear is subsequently disabled by a passing vessel or fishing gear, and
the degree of exposure to environmental forces (storms, currents) that eventually
disable the derelict gear (Erzini et al., 1997; Pawson, 2003; Revill and Dunlin, 2003;
Sancho et al., 2003; Akiyama, Saito and Watanabe, 2007; FAO, 2010a).
For example, studies have observed that when gillnet and entangling nets are
deployed on a flat substrate in relatively shallow water, their ghost fishing catching
efficiency and longevity declines rapidly over the initial few days of release, and
declines to within about 5percent of initial catching efficiency within weeks to months.
While exhibiting a rapid reduction in fishing efficiency, a derelict net that retains
5percent of its original catching efficiency that persists for years results in a large level
of ghost fishing mortalities. The gear quickly loses its profile owing to the weight of
captured organisms. Also, as meshes become obstructed owing to the accumulation of
debris and particulate matter and from biofouling (encrusting by biological organisms),
this increases the net’s visibility, and reduces its profile and surface area (Kaiser et al.,
1996; Erzini et al., 1997; Nakashima and Matsuoka, 2004, 2005; Akiyama, Saito and
Watanabe, 2007). In some studies, physical damage to derelict gear occurred from
interactions with other fishing gear, usually trawlers and other towed gear, but in some
cases by passive gear (e.g. Erzini et al., 1997; MacMullen et al., 2003). When a derelict
net occurs on open, flat substrate, some studies have found that the gear initially causes
ghost fishing mortality primarily of demersal fishes over the first few days to weeks
of deployment. Then, once most of the net area has collapsed, the remaining loose
horizontal sheets of netting primarily catch scavenging crustaceans and molluscs until
the net structure is no longer intact or is buried (Kaiser et al., 1996; Pawson, 2003;
Revill and Dunlin, 2003; Akiyama, Saito and Watanabe, 2007).
Abandoned, lost and discarded gillnets and trammel nets
10
However, when derelict nets become entangled on three-dimensional objects and/
or are in a location where environmental conditions, such as currents and weather,
and interactions with other fishing gear do not damage the gear, including in very
deep water, gill and entangling nets can maintain high ghost fishing catch rates for
relatively long periods (several months to several years) (Kaiser et al., 1996; Erzini et
al., 1997; Matsuoka, Nakashima and Nagasawa, 2005; Brown and Macfadyen, 2007;
Good et al., 2010). In one study, a FAD entangled with a gillnet was observed to
have a higher number of fish aggregated around the device relative to an aggregation
device of the same design but without an entangled gillnet (Nakashima and Matsuoka,
2005), suggesting that when entangled on three-dimensional objects, derelict nets can
augment the fish aggregating capacity of the submerged feature, enhancing the ghost
fishing catch rate. Organisms caught in derelict fishing gear can attract scavengers,
which subsequently are caught, contributing to long-term ghost fishing efficiency
owing to this self-baiting (Kaiser et al., 1996; Matsuoka, Nakashima and Nagasawa,
2005; FAO, 2010a; Gilman et al., 2013).
The depth at which the derelict gear occurs affects its species-specific ghost fishing
selectivity (Brown and Macfadyen, 2007). For example, to catch seabirds, a derelict net
must be at or near the sea surface (Hayase and Yatsu, 1993; Kaiser et al., 1996).
Gear design and materials are other potentially significant explanatory factors of
ghost fishing. Gear designs that affect detection and mechanisms for escape will affect
ghost fishing ability, efficiency and duration. A few examples follow.
Gillnet mesh size has been shown to affect gear selectivity (Price and Van
Salisbury, 2007).
The height or profile of demersal and surface gillnets has been shown to affect sea
turtle bycatch rates significantly, perhaps because of the effect on the net stiffness
and proportion of the water column that is fished (Gearhart and Eckert, 2007;
Price and Van Salisbury, 2007; Eckert et al., 2008; Gilman et al., 2010).
In demersal gillnet fisheries, tiedowns are typically used to maximize the catch of
demersal fish species, discussed in Section 2.1. The shorter the length of tiedowns,
the deeper the webbing pocket is, and the higher the probability of capturing non-
target species including sea turtles (Figure 1) (Price and Van Salisbury, 2007).
Mesh, floats, and float and lead line characteristics will also have effects on species-
specific catch rates and ghost fishing mortality. For example, mesh characteristics,
including embedded luminescent materials and infusing nylon nets with metal
compounds, affect a net’s species-specific detectability or affects species-specific
susceptibility to capture in the net, and concomitant catch and ghost fishing
mortality rates (Melvin, Parrish and Conquest, 2001; Bjordal, 2002; Trippel et al.,
2003; Werner et al., 2006; Larsen et al., 2007).
Using alternative net materials (thinner twine and weaker material) to produce a
breaking strength that allows large organisms to break free of the gear and escape
might reduce ghost fishing mortality (Gilman et al., 2010).
Gear materials, in combination with local mechanical action and the rate of
physical abrasion, water chemistry, and light penetration, affect ALDFG duration
of ghost fishing. For example, discussed in Table 1, degradable materials have
been tested and are now in use in a few fisheries (Carr, Blott and Caruso, 1992;
Chanrachkij, Siriraksophon and Loog-on, 2008; FAO, 2010a, 2010b, 2011a).
Gear components that affect the depth and weight of the gear affect the ability and
energy required for caught air-breathing organisms to reach the surface and the
time period for possible escape (Gilman et al., 2013).
Characteristics of derelict fishing gear that become attached to organisms affect
the probability of mortality (e.g. the length of trailing line, Swimmer and Gilman,
2012; Gilman et al., 2013).
11
Introduction
1.5 STUDY SCOPE AND AIMS
Recognizing that substantial ecological and socio-economic adverse effects result
from fishing mortality caused by ALDFG, that marine gillnets and trammel nets are
globally important gear types, and that gillnets and entangling nets have relatively high
ghost fishing potential (Pawson, 2003; Matsuoka, Nakashima and Nagasawa, 2005;
Macfadyen, Huntington and Cappel, 2009), FAO and UNEP commissioned this study
to identify best practice methods for estimating ghost fishing rates and levels, priority
research needs, and the status of international monitoring and management of ALDFG
and ghost fishing by marine gillnet and trammel net fisheries. This report supplements
the broad overview of the magnitude, composition, impacts, causes and approaches
to reduce ALDFG provided in UNEP Regional Seas Reports and Studies No.185/
FAO Fisheries and Aquaculture Technical Paper No.523 (Macfadyen, Huntington and
Cappel, 2009).
Study aims were to: (i) identify best practice methods to estimate ghost fishing
mortality levels by ALDFG from marine gillnet and trammel net fisheries; (ii) identify
priority information gaps, including research priorities to support robust estimates of
global ghost fishing mortality levels by gillnets and trammel nets; and (iii) benchmark
international monitoring and management of ALDFG and ghost fishing from marine
gillnet and trammel net fisheries.
3
To accomplish the first two aims, a review was
conducted of methods and findings of terms to estimate total ghost fishing mortality
in an explicit spatial area over a fixed time period. This included reviewing methods
and findings on: rates of abandonment, loss and discarding of fishing gear; density of
ALDFG (unit amount per unit of area of seafloor and/or per unit volume of water
column); and the duration of ghost fishing efficiency and ghost fishing mortality rates
of ALDFG. To implement the third aim, an assessment was conducted of the data
collection protocols and management measures to prevent and remediate ALDFG and
ghost fishing of ten regional bodies and arrangements with the competence to establish
binding controls for regional marine capture fisheries, and that have competence over
fishery resources that are captured in an active gillnet or trammel net fishery.
3
Hereinafter, to avoid repetition and unless otherwise stated, the text refers to ALDFG and ghost fishing
solely by gillnets and trammel nets.
13
2. Catching process, design and
operation of marine drift and set
gillnets and trammel nets
Now used worldwide, gillnets and trammel nets are types of passive fishing gear
where the capture process relies on the movement of organisms into the gear, where
it becomes gilled, enmeshed or entangled (Bjordal, 2002). An estimated 19percent of
global marine fisheries landings comes from gillnet and trammel net fisheries (SAUP,
2011). There are both extensive industrial and artisanal, small-scale, household-based
gillnet and trammel net fisheries (Hubert, 1983; Shester and Micheli, 2011). Artisanal,
small-scale fisheries typically use passive gear (Bjordal, 2002; MacMullen et al., 2003). In
2006, the Russian Federation (Pacific coast), Myanmar, India, VietNam, and Indonesia
(eastern) had the five highest reported landings from marine gillnet fisheries that
occurred in their exclusive economic zones (EEZs) (SAUP, 2011). The five countries
for which landings from marine gillnet fisheries accounted for the largest proportion of
total marine fisheries landings from their EEZs in 2006 were Timor-Leste (94percent),
Bangladesh (91percent), Somalia (88percent), Myanmar (86percent) and VietNam
(79percent) (SAUP, 2011).
Explained in Section 1.4, various gillnet and trammel net gear designs and fishing
methods can have significant effects on ghost fishing ability, catch rates and duration
of fishing efficiency. This chapter provides an overview of the catching process, designs
and methods of operation of different types of marine gillnets and trammel nets, which
are factors that have the potential to influence ghost fishing catch rates.
Gillnets and trammel nets can be anchored and stationary on a variety of substrates
of flat open ground and on the edges and slopes of canyons, and on reefs, wrecks and
other structures. Gillnets can also be drifting at the sea surface, mid-water or near or on
the sea bed (MacMullen et al., 2003). Because they rely on target species swimming into
the net, gillnets and trammel nets are designed to minimize their visibility and might be
used in areas with low light levels or high turbidity (Bjordal, 2002).
Since the 1960s, nets have been made with synthetic twine usually of nylon
(polyamide), either as multifilament thread or monofilament or mutlimonofilament
line, which is much less visible, more durable and requires less maintenance than nets
made of non-synthetic fibre (Moore, 2008; Macfadyen, Huntington and Cappel, 2009).
Mesh size is typically measured as the length of a whole stretched mesh (Figure2), or
FIGURE 2
Measuring mesh size as the length of a whole stretched mesh
Source: SEAFDEC.
Abandoned, lost and discarded gillnets and trammel nets
14
as the half-length (bar-length). Figure3 illustrates the method for counting the number
of meshes in a net.
To maintain a roughly vertical profile, gillnets, trammel nets and combination set
gillnet/trammel nets, described below, have floats, typically made of plastic or cork,
attached along the top rope (float line, cork line, headline or headrope), and weights
attached along the bottom rope (sinker, lead line, footline or groundrope), or a bottom
rope made with a lead core (Hubert, 1983; Nedelec and Prado, 1990; Bjordal, 2002).
Wreck nets might include metal rings to the leadline to avoid snagging on the wreck
(Rose et al., 2000). Commercial vessels might use hydraulic-driven haulers and net
drums to set and haul driftnets and to haul set nets, while artisanal vessels typically set
and haul drift and set gillnets, trammel nets and combination gillnets/trammel nets by
hand (Nedelec and Prado, 1990). After hauling, the catch is removed from the net by
hand.
Depending in part on the target species and size, and environmental characteristics
of the fishing grounds, there can be large variability in the designs and methods of
deployment of gillnets and trammel nets. Parameters that can vary include (Hubert,
1983; Bjordal, 2002; Hall et al., 2009):
whether a net is drifting, anchored or sweeping;
whether a net is at the sea surface, midwater, slightly above the substrate, or at the
substrate;
the hanging ratio, a measure of how tightly the net is stretched, is the length of a
rope on which a panel is mounted divided by the length of stretched netting on
the rope. Low hanging ratios have meshes with narrow openings with a tall mesh
height and narrow lateral opening, while large hanging ratios have a low mesh
height and wide lateral opening;
the length of each panel, number of panels (a series of connected panels is referred
to as a “fleet”), and total fleet length;
mesh size;
twine diameter and material, and concomitant breaking strength;
total surface area;
mesh characteristics (mesh size, twine material, diameter, colour);
float line characteristics (material, diameter, colour);
float characteristics (number per unit length of float line, material, colour,
dimensions, shape);
whether floats are set at or below the surface;
lead line characteristics (material, diameter, colour, weight per unit length of lead
line, mass of each weight);
whether bait is placed in net;
FIGURE 3
Net with a mesh count of ten, determined by counting the number of meshes
between the float and lead lines
Source: SEAFDEC.
15
Catching process, design and operation of marine drift and set gillnets and trammel nets
the angle of the net in relation to the coastline;
the time of day and season of gear soak;
for demersal nets, whether tie downs are used and their height;
whether nets are patrolled or otherwise left unattended during the soak.
2.1 GILLNETS
A fleet of gillnets is constructed of a series of connected single panels of meshes made
from fine thread, with reinforcing ropes along the sides. Gillnets typically have a hanging
ratio ≥ 0.5 (Uhlmann and Broadhurst, 2013). Gillnets are relatively size-selective for
finfish, but can have poor species selectivity, depending on the species composition at
individual fishing grounds (Valdemarsen and Suuronen, 2003; Suuronen et al., 2012).
The catching process in gillnets entails fish being caught in one of the meshes in the gill
region of its body (between its head and body) (Figure4) (Bjordal, 2002). A fish can
become caught in a gillnet when it swims part way through a mesh, struggles to free
itself, and the twine of the mesh slips under the fish’s opercula (gill covers) preventing
escape, i.e. the fish becomes “gilled” (Bjordal, 2002) (Figure4). Less frequently, a fish
may also become wedged around its body inside a gillnet mesh, or parts of its body
(fins, teeth, or other projection) may become entangled in the twine (Murphy and
Willis, 1996; Price and Van Salisbury, 2007).
FIGURE 4
Catching process in gillnets where a fish becomes caught in one of the meshes in
the gill region of its body
Source: SEAFDEC.
FIGURE 5
Drift gillnet (driftnet) set at the surface, connected to a fishing vessel and drifting
with the current
Source: SEAFDEC.
Abandoned, lost and discarded gillnets and trammel nets
16
Drift gillnets (driftnets), set at
or below the sea surface, can be
connected to a fishing vessel and
drift with the current (Figures5
and 6).
Set or anchored gillnets are a
type of static gear deployed at
or near the sea bed. Set gillnets
can be kept stationary using
anchors or weights at both ends
(Figure7). Tiedowns may be used
in set gillnets. Tiedowns are lines
that are shorter than the fishing
height of the net and connect
the float and lead lines at regular
intervals along the entire length
of the net. Tiedowns create a bag
of slack webbing which aids in
entangling, rather than gilling,
FIGURE 6
Drift gillnet (driftnet) deployed subsurface/midwater, attached to a vessel and
drifting with the current
Source: Nedelec and Prado (1990); redrawn by SEAFDEC.
Source: SEAFDEC.
FIGURE 7
Bottom set (anchored) gillnet
FIGURE 8
Staked set gillnet
Source: Nedelec and Prado (1990); redrawn by SEAFDEC.
17
Catching process, design and operation of marine drift and set gillnets and trammel nets
demersal fish species (Figure 1)
(Price and Van Salisbury, 2007).
Gillnets with panels stretched
between two or more stakes are
also deployed primarily in the
intertidal zone in locations with
a large tidal range (Figure 8)
(Nedelec and Prado, 1990).
Staked gillnets are left to soak
for several days, with the catch
collected at low tide (Nedelec
and Prado, 1990).
Encircling gillnets are used
in shallow, nearshore waters
(Figure 9) (Nedelec and Prado,
1990). Typically, groups of
small-scale fishers, or in some
cases one fisher, using small open boats or canoes will encircle a school of fish in
shallow water, and will then use various methods to disturb and scare the fish to swim
into the net. Methods to scare the fish include making noise by slapping the surface
of the water with sticks or a paddle, and having a group of people move towards the
school (Nedelec and Prado, 1990).
2.2 TRAMMEL NETS
Trammel nets usually have three panels of meshes, but occasionally can have two
layers (Figure10) (Bjordal, 2002; Hall et al., 2009). The middle panel is slack and has
small-sized meshes. The two outer panels have larger meshes. When a fish comes into
contact with a trammel net, it pushes the small mesh through an adjacent larger mesh
and becomes entangled (Figure11). As a result of this catching process, trammel nets
are less size-selective relative to gillnets (Bjordal, 2002), and as gillnets are not species
selective (e.g. Goncalves et al., 2008). Trammel nets typically have a hanging ratio <0.5
(Uhlmann and Broadhurst, 2013).
Like set gillnets, trammel nets are usually used to target benthic and demersal
species. They are typically deployed anchored on the substrate in shallow nearshore
areas (Figure 12). Drift trammel nets are also used near the substrate (Nedelec and
Prado, 1990; Bjordal, 2002). They may be set as a single panel or occasionally in fleets
of a string of connected panels (Nedelec and Prado, 1990).
FIGURE 9
Encircling gillnet
Source: SEAFDEC.
FIGURE 10
Trammel net
Source: SEAFDEC.
Abandoned, lost and discarded gillnets and trammel nets
18
2.3 COMBINATION GILLNETS – TRAMMEL NETS
Originating in the Mediterranean and now used in many parts of the world, gear
employing a gillnet in the upper portion and trammel net in the lower portion are
typically set on the substrate as set gillnets and trammel nets (Figure13) (Nedelec and
Prado, 1990). The gillnet portion might target pelagic and/or semi-demersal species,
while the trammel net portion targets demersal species.
Source: SEAFDEC.
FIGURE 11
Catching process in a trammel
net with three panels of meshes,
where a fish pushes the small
mesh of a centre panel through
an adjacent larger mesh of
an outer panel and becomes
entangled
FIGURE 12
Anchored trammel net
Source: SEAFDEC.
FIGURE 13
Combined gillnet – trammel net, with gillnet in the upper section and trammel
net at the bottom
Source: SEAFDEC.
19
3. Methods and estimates
3.1 METHODS AND RESULTS
A sample of studies were reviewed to document the range of methods and findings on:
(i) rates of abandonment, loss and discarding of fishing gear; (ii) the density of ALDFG
(unit amount per unit of area of seafloor and/or per unit volume of water column); and
(iii) ghost fishing mortality rates and duration of ghost fishing efficiency of ALDFG
(Table 2). Both structured and unstructured literature searches were conducted
to compile relevant literature. The structured search was conducted using various
combinations of the following Boolean search terms in Google Scholar: gillnet, gill-
net, gill, trammel, entangle, net, ghost, ALDFG, abandoned, lost, discarded, derelict,
fishing, and gear. An unstructured literature search was conducted by reviewing
reference lists of relevant publications and reports and then searching for identified
relevant citations, posting a query on ResearchGate.net and via an informal network
of fisheries professionals requesting suggestions of relevant publications. Literature
compilation was conducted from February to June 2014. The aim of the literature
search was to obtain adequate sample sizes for each of the three study categories to
characterize the range of methods employed and findings, and to include studies across
regions. For each record included in Table2, the UNEP Regional Seas Convention and
Action Plan region and FAO Major Marine Fishing Areas in which the study area was
located was identified (FAO, 2014; UNEP, 2014).
These three categories of studies were included in the review because they provide
information on factors needed to estimate total ghost fishing mortality in an explicit
spatial area over a fixed time period. To produce accurate estimates of the total level
of ghost fishing removals that occurs in a fishing ground during a fixed time period,
accurate information is required on the amount of derelict gear present in the area
and the mean ghost fishing catch rate of the derelict gear in the area at a point in time
(Matsuoka, Nakashima and Nagasawa, 2005). Robust estimates of the total ghost
fishing mortality level that an individual derelict net will cause require information on
the duration of ghost fishing efficiency of that net, and the change (likely exponential
decay) in species-specific catch rate during the period when the net retains some fishing
efficiency (Kaiser et al., 1996; Erzini et al., 1997; Sancho et al., 2003; Nakashima and
Matsuoka, 2004; Ayaz et al., 2006).
The purposes for reviewing this sample of relevant studies were to: provide an
understanding of the dispersion in methods and findings; document the potentially
significant explanatory effect of identified factors; and determine the state of
understanding of the severity of adverse ecological effects from gillnet and trammel
net ALDFG. The assessment also enabled the identification of general best practice
estimation methods to reduce uncertainty and identification of priority gaps in
information needed to produce robust estimates of regional and global ghost fishing
mortality rates and levels by ALDFG.
Abandoned, lost and discarded gillnets and trammel nets
20
TABLE2
Synthesis of methods and estimates of: (a) rates of abandonment, loss and discarding of fishing gear; (b)
density of ALDFG; and (c) ghost fishing mortality rates and duration of ghost fishing efficiency of ALDFG
TABLE2A
Methods and estimates of rates of abandonment, loss and discarding of fishing gear from gillnet and
trammel net fisheries
UNEP
Regional
Seas
1
FAO Major
Marine
Fishing
Area
2
Fishery or study site location Method
3
Rate of abandonment, loss and/or
discarding
Citation
Baltic 27 Swedish Baltic Sea south
coast, Hano Bay and east
coast demersal gillnet cod
and turbot fisheries
a 0.08% of nets set were lost and not
retrieved (25.5km of 28021km length of
nets set per year were lost, of which 3.35km
was retrieved by the vessel that temporarily
lost it). This is equivalent to a mean of 3.7
nets (108 m average length per net) per
vessel per year.
MacMullen
etal., 2003
East Asian
Seas and
North-West
Pacific
61 Coastal gillnet fisheries of
the Republic of Korea
a,b 38535 tonnes gillnets per year abandoned,
lost or discarded in coastal waters of the
Republic of Korea. On average, a gillnet
vessel abandons, loses or discards 9.64 units
of gillnets per year; one unit weighs an
average of 478.4kg.
Kim, Lee and
Moon, 2014
Mediter-
ranean
37 Section of the Gokova
Special Environmental
Protection Area off Turkey,
eastern Mediterranean
Sea, demersal gillnet and
trammel net fisheries
a 0.8% and 3.4% of demersal gillnets and
trammel nets, respectively, are lost per year
Ayaz et al.,
2010
Mediter-
ranean
37 Spanish Cantabrian region
demersal gillnet red mullet
and hake fisheries and
coastal demersal trammel
net mixed species fishery
a 12.7 nets per vessel per year are lost and
not retrieved (13.3 nets per vessel per year
were lost of which 0.58 nets per vessel were
retrieved by the vessel that temporarily
lost it).
Information was not presented to determine
the percentage of total gear set that was
lost and not recovered.
MacMullen
et al., 2003
North-East
Atlantic
27 United Kingdom, German
and Panamanian deep-water
anchored gillnet monkfish
and shark fishery, United
Kingdom and Ireland EEZs
and adjacent high seas,
west of the British Isles,
north of Shetland, on the
continental slopes from
south of Porcupine Bank to
Tampen, and at Rockall and
the Hatton Bank
c 600 50m-long panels of gillnet (30km total
length) per vessel per trip are discarded.
Hareide
etal., 2005
North-East
Atlantic
27 Southern Norwegian coastal
anchored gillnet cod fishery
a 0% of nets set were lost and not retrieved
(10 of 170000 nets were lost, and all were
retrieved by the vessel that temporarily
lost it).
MacMullen
et al., 2003
North-East
Atlantic
27 Southern Norwegian coastal
anchored gillnet Greenland
halibut fishery
a 0.09% of nets set were lost and not
retrieved (5 of 5350 nets set were lost
and none was retrieved by the vessel that
temporarily lost it).
MacMullen
et al., 2003
North-East
Atlantic
27 Southern Norwegian coastal
anchored deep-sea gillnet
saithe fishery
a 0.18% of nets set were lost and not
retrieved (275 of 152550 nets set were lost
and none was retrieved by the vessel that
temporarily lost it).
MacMullen
et al., 2003
North-East
Atlantic
27 Southern Norwegian
anchored gillnet blue ling
and ling fishery
a 0.5% of nets set were lost and not retrieved
(159 of 12135 nets set were lost of which
97were retrieved by the vessel that
temporarily lost it).
MacMullen
et al., 2003
North-East
Atlantic
27 Algarve, Portugal, demersal
trammel net and gillnet
fisheries
a Local fishery: 17.3 panels lost/vessel/year, of
which 3.2 panels are not recovered.
Coastal fishery: 27.0 panels lost/vessel/year,
of which 6.0 panels are not recovered.
Hake fishery: 33.6 panels lost/vessel/year, of
which 7.4 panels are not recovered.
MacMullen
et al., 2003
21
Methods and estimates
UNEP
Regional
Seas
1
FAO Major
Marine
Fishing
Area
2
Fishery or study site location Method
3
Rate of abandonment, loss and/or
discarding
Citation
North-East
Atlantic
27 United Kingdom demersal
trammel net and demersal
gillnet hake fisheries
a Trammel net fishery: 845 m length of
nets per vessel per year are lost and not
retrieved (1.3km length of nets per vessel
per year were lost of which 0.455km per
vessel was retrieved by the vessel that
temporarily lost it).
Gillnet fishery: 500 m of net per vessel per
year are lost and not retrieved (1km length
of nets per vessel per year were lost of
which 0.5km per vessel was retrieved by the
vessel that temporarily lost it).
Information was not presented to determine
the percent of total gear set that was lost
and not recovered.
MacMullen
et al., 2003
North-East
Atlantic
27 Algarve, Portugal, local,
coastal and gillnet hake and
trammel net fisheries
a The average annual number of panels lost
per boat was 3.2, 5.1 and 7.4 for the local,
coastal and hake fisheries, respectively.
Santos et al.,
2003a
North-East
Atlantic
and
Mediter-
ranean
27 and 37 French demersal gillnet and
trammel net fisheries of the
(i) East Channel and North
sea coasts, (ii) North and
West Brittany coasts, and (iii)
Mediterranean coast.
a East Channel and North sea coasts: 1.6km/
vessel/year lost; 1% of the length of nets set
per year are lost.
North and West Brittany coasts: 2.6km/
vessel/year lost; 1% of the length of nets set
per year are lost.
Mediterranean coast: 0.95km/vessel/year
lost; 1.7% of the length of nets set per year
are lost.
MacMullen
et al., 2003
North-East
Pacific
67 United States of America
Puget Sound, Washington
salmon driftnet fishery.
a,b 2–10% of vessels lose a fragment or entire
fleet. Extrapolating this loss rate estimate to
the entire fishery, 18–42 driftnet fragments
are lost per year.
Antonelis,
2012
1
UNEP, 2005b, 2014. “None” indicates there is no UNEP Regional Sea Convention or Action Plan in the region for this study.
2
FAO, 2014.
3
(a) Survey (remote and/or in-person) of fishers.
(b) Survey (remote and/or in-person) of non-fishers experts.
(c) Method not specified (“anecdotal evidence”).
TABLE2A (CONTINUED)
TABLE2B
Methods and estimates of the density of ALDFG from gillnet and trammel net fisheries
UNEP
Regional
Seas
1
FAO Major
Marine
Fishing
Area
2
Fishery or study site
location
Method
3
Density of ALDFG
(unit amount per unit of area of seafloor and/or
per unit volume of water column)
Citation
Baltic 27 Swedish demersal gillnet
fishery, south of Gotland,
Baltic Sea
a,g 7.5 kg of gillnet/km
2
(35.1 m length of gillnet/
km
2
)
MacMullen
et al., 2003
East Asian
Seas and
North-West
Pacific
61 Eastern Yellow Sea,
Republic of Korea
a,b 1570 kg of fishing nets /km
2
Kang, 2003
Mediter-
ranean
37 Section of the Gokova
Special Environmental
Protection Area
off Turkey, eastern
Mediterranean Sea,
demersal gillnet and
trammel net fisheries
c 15700 m of combined gillnets and trammel
nets perkm
2
Ayaz et al.,
2010
None 21 United States of America
Gulf of Maine, Jeffries
Ledge and Stellwagen
Bank demersal gillnet
fishery
d,e,f Jeffries Ledge: 54 m
2
gillnet /km
2
fishing
ground (27 m length gillnet/km
2
fishing
ground).
Stellwagen Bank: 921 m
2
gillnet /km
2
fishing
ground (1842 m length gillnet/km
2
fishing
ground)
Extrapolation to total fishing ground: 2240
gillnets (1.02km
2
of netting; 1 net = 91 m ×
5m)/ 14042km
2
fishing ground.
Carr and
Cooper,
1987
Abandoned, lost and discarded gillnets and trammel nets
22
UNEP
Regional
Seas
1
FAO Major
Marine
Fishing
Area
2
Fishery or study site
location
Method
3
Density of ALDFG
(unit amount per unit of area of seafloor and/or
per unit volume of water column)
Citation
None 21 Demersal gillnet cod
fishery, Newfoundland,
Canada
a 148 demersal gillnet fragments per 48.3 hours
of trawling/grappling effort; 167 demersal
gillnet fragments per 53.5 hours of trawling/
grappling effort.
Way, 1977
North-East
Atlantic
27 Deepwater demersal
gillnet anglerfish, hake
and shark fisheries,
Rosemary, Porcupine and
Rockall Banks, off the
United Kingdom and
Ireland
a CEFAS 2005 survey at Rosemary Bank: 0.011 m
2
of gillnet perkm of towed transect.
BIM 2005 and 2006 surveys at Rockall and
Porcupine Banks: 0.12km and 0.014km of
gillnet perkm of towed transect, respectively.
CEFAS 2006 survey at Porcupine Bank: 0.032km
of gillnet perkm of towed transect.
Large et al.,
2009
North-East
Atlantic
27 Norwegian demersal
gillnet Greenland halibut
fishery
a 0.11km of gillnet perkm of towed grapnel.
(The total distance of towed transects was
not reported, and was therefore estimated
by assuming a mean tow speed of 1.5 knots,
and mean haul duration of 2.25 hours was
conducted, such that the total distance covered
by the survey was 116.3km, the distance
covered by each haul was 1.88km, and 0.21km
of gillnet was retrieved per haul).
Misund et
al., 2006
North-East
Atlantic
27 Northeastern United
Kingdom wreck gillnet
fishery
c 7 of 11 wrecks had some ALDFG from gillnet
fisheries. None of the wrecks had ALDFG from
gillnets that retained fishing efficiency (0 of 27
gillnets and gillnet fragments retained fishing
efficiency).
Revill and
Dunlin,
2003
North-East
Pacific
67 United States of America
Puget Sound, Washington
salmon driftnet fishery
b,c,f Between 3550 and 6442 lost fragments or
fleets of gillnets were estimated to be in the
Washington State waters of the Salish Sea (area
of the fishing grounds not reported).
Antonelis,
2013
North-East
Pacific
67 United States of America
Puget Sound, Washington
salmon driftnet fishery
c,f 4518 fragments or fleets of gillnet was
estimated to occur in the Washington State
waters of Puget Sound (area of the fishing
grounds not reported).
Northwest
Straights
Foundation,
2007
1
UNEP, 2005b, 2014. “None” indicates there is no UNEP Regional Sea Convention or Action Plan in the region for this study.
2
FAO, 2014.
3
(a) Tow ‘creeper’ grappling device.
(b) Side scan sonar.
(c) Divers (scuba, diver tows).
(d) Manned submersible.
(e) Underwater ROV.
(f) Raised estimate from sampled area to entire fishing grounds.
(g) Raised estimate explicitly accounted for proportion of ALDFG that was estimated not to have been observed.
TABLE2B (CONTINUED)
23
Methods and estimates
TABLE2C
Methods and estimates of ghost fishing mortality rates and duration of ghost fishing efficiency of ALDFG
from gillnet and trammel net fisheries
UNEP
Regional
Seas
1
FAO
Major
Marine
Fishing
Area
2
Fishery or study
site location
Method
3
Ghost fishing mortality
rates
Ghost fishing duration Citation
Baltic 27 Simulated derelict
demersal cod
gillnets, Hano Bay,
Swedish
a,c,d,e,h,i NA Fishing efficiency declined
to 5–7% of the initial level
after 3 months. Retained
some fishing efficiency at 27
months.
Tschernij
and Larsson,
2003
Mediter-
ranean
37 Simulated derelict
demersal gillnets,
Izmir Bay, eastern
Aegean Sea, Turkey
a,d,f,g,k,l Multifilament gillnets: 62
fish in three 33 m × 2.8 m
gillnets for duration of fish
fishing efficiency.
Monofilament gillnets: 115
fish in three 33 m × 2.8 m
gillnets for duration of fish
fishing efficiency.
Multifilament and
monofilament gillnets
ceased to catch fish at
106 and 112 days after
deployment, respectively.
Ayaz et al.,
2006
Mediter-
ranean
37 Simulated derelict
demersal crawfish
trammel net and
hake and seabass
demersal gillnet,
St. Tropez Canyon
and Cassis harbour,
coastal France
a,c,d,e,k Gillnet open ground: 46
hake and 36 crawfish per
5400 m
2
of net per year.
Trammel net open ground:
46.25 crawfish per 2100m
2
of net per year.
Nets on open ground
retained some catching
efficiency at 18 months after
deployment. Gillnet and
trammel nets set on wrecks
no longer retained catch
efficiency by 6 months after
deployment.
MacMullen
et al., 2003
None 21 United States of
America Gulf of
Maine, Jeffries
Ledge and
Stellwagen Bank
demersal gillnet
fishery
b,f,i 15% of fish catch rate of
in-use gear.
NA Carr and
Cooper, 1987
None 21 Simulated derelict
demersal gillnet,
Cape Cod Bay, Gulf
of Maine, United
States of America
a,c,f NA Retained catching efficiency
after 74 days.
Carr et al.,
1985
None 21 Simulated derelict
demersal gillnet,
Buzzards Bay, New
England, United
States of America
a,c,f NA Catch efficiency of the
control and all experimental
treatment nets continued
after 2 years.
Carr, Blott
and Caruso,
1992
None 71 and
77
Simulated derelict
Japanese high seas
squid drift gillnets,
central Pacific
Ocean near the
Hawaii archipelago
a,c,d,f,k NA Lengths of 50 m and 100 m
length nets reduced to <5%
of original in less than 0.5
day. The 350 m length net
reduced to <5% of original
at 2 days. The 1km length
net reduced to <5% of
original length at 10 days.
Gerrodette,
Choy and
Hiruki, 1987,
1990
North-
East
Atlantic
27 Simulated derelict
demersal trammel
nets, central coast
of Portugal
a,c,d,f,i,k Rocky substrate: 541
organisms per 100 m of net
during study period.
Sandy substrate: 257
organisms per 100 m of net
during study period.
Fishing efficiency <1% from
an in-use net at 10.5 months
at the site with rocky
substrate, and at 8 months
at the site with sandy
substrate.
Baeta, Costa
and Cabral,
2009
North-
East
Atlantic
27 Simulated derelict
demersal gillnet
and trammel net,
Algarve, Faro,
southern Portugal
a,c,d,f,g,k,l Gillnet: 314 fish, 0 seabirds,
0 reptiles, 0 mammals per
240 m
2
net for duration of
finfish fishing efficiency.
Trammel net: 221 fish,
0 seabirds, 0 reptiles, 0
mammals per 190 m
2
net
for duration of finfish
fishing efficiency.
Duration of fishing
efficiency for finfish: 15-20
weeks.
Erzini et al.,
1997
Abandoned, lost and discarded gillnets and trammel nets
24
UNEP
Regional
Seas
1
FAO
Major
Marine
Fishing
Area
2
Fishery or study
site location
Method
3
Ghost fishing mortality
rates
Ghost fishing duration Citation
North-
East
Atlantic
27 Simulated
derelict deep
water demersal
Greenland halibut
gillnets, Norway
a,c,d,e,h,j Experiment 1: 67–100 kg
halibut per 4207.5m
2
net
per day once net fishing
efficiency declined to
20–30% of original.
Experiment 2: 28–43 kg
halibut per 4207.5 m
2
net
per day once net fishing
efficiency declined to
20–30% of original.
Retained catch efficiency
after 68 days. Catch rate
reached 20–30% of initial
efficiency between 21 and
45 days after setting and
remained at that level
through the remainder of
the study period to 68 days
after setting.
Humborstad
et al., 2003
North-
East
Atlantic
27 Simulated derelict
demersal gillnet
and trammel net,
St. Bride’s Bay,
southwest Wales,
United Kingdom
a,c,d,f,g,l Gillnet: 226 fish, 839
crustaceans per 243 m
2
net for duration of fishing
efficiency.
Trammel net: 78 fish, 754
crustaceans per 243m
2
net for duration of fishing
efficiency.
The ghost fishing catch
rate of number of fish per
24-hour period approached
0 at 70 and 22days after
deployment for the
gillnet and trammel net,
respectively. Crustaceans
continued to be observed
to be caught at low rates
at 9 months after initial
deployment.
Kaiser et al.,
1996
North-
East
Atlantic
27 Simulated derelict
demersal hake
gillnet, southwest
England, United
Kingdom
a,c,d Fleet 1: 39 crustaceans and
2 fish per 400 m net during
study period.
Fleet 2: 30 crustaceans and
6 fish per 400 m length of
net during study period.
Not known; the
experimental fleets were
lost when checked at 14
weeks after deployment.
MacMullen
et al., 2003
North-
East
Atlantic
27 Simulated derelict
demersal gillnet,
Bay of Biscay, Spain
a,c,e,h,l 7.38 kg of monkfish per
180 m
2
net for duration of
fishing efficiency
Still maintained some
demersal fish and
invertebrate catch efficiency
after 12 months of
deployment.
MacMullen
et al., 2003
North-
East
Atlantic
27 Simulated derelict
wreck gillnet and
demersal trammel
net, North Sea off
northeast United
Kingdom
a,c,d,f,g,k NA Wreck gillnet ceased
finfish fishing efficiency at
45 weeks and crustacean
fishing efficiency at 2years
after being set. Open
ground trammel net ceased
fishing efficiency at 58 days
after being set.
Revill and
Dunlin, 2003
North-
East
Atlantic
27 Simulated derelict
demersal monkfish
gillnet, Bay of
Biscay, Basque
Region, Cantabrian
Sea, northern Spain
a,c,d,e,h,
k,l
4.7 monkfish (17.7 kg) per
360 m
2
net for duration of
fishing efficiency.
224 days until ceased to
catch monkfish.
Sancho et
al., 2003
North-
East
Atlantic
27 Simulated derelict
demersal hake
gillnets, Faro,
Algarve, Portugal
a,c,d,e,h,i,
k,l
May-deployed fleets: 116
organisms (29.8 kg) / 9
hake (20.6 kg) per 620 m
2
net for duration of fishing
efficiency.
Sept.-deployed fleets: 413
organisms (90.1kg)/ 88
hake (29.9 kg) per 620 m
2
net for duration of fishing
efficiency.
Ghost fishing maximum
duration was estimated to
be 248 days; negligible catch
was predicted to be reached
after 3 months.
Santos et al.,
2003b
North-
East
Atlantic
27 Simulated derelict
demersal hake
gillnets, Algarve,
Faro, southern
Portugal
a,c,d,e,h,i,l 249.9 non-hake organisms
(64.4 kg) per 620m
2
net
for duration of fishing
efficiency
Retained catching efficiency
after 12 months. Estimated
fishing capacity would end
at 430days after setting.
Santos,
Gaspar and
Monteiro,
2009
North-
East
Pacific
67 United States of
America Puget
Sound, Washington
salmon driftnet
fishery
b,f,g 2.119 invertebrates, 0.196
seabirds, 0.275 fish per
3610 m
2
net per day
NA Gilardi et al.,
2010
TABLE2C (CONTINUED)
25
Methods and estimates
UNEP
Regional
Seas
1
FAO
Major
Marine
Fishing
Area
2
Fishery or study
site location
Method
3
Ghost fishing mortality
rates
Ghost fishing duration Citation
North-
East
Pacific
67 United States of
America Puget
Sound, Washington
salmon driftnet
fishery
b,f,k NA Fish and diving seabirds
ceased to be caught after
about 3 years. Crabs
continued to be caught
after 6 years.
High, 1985
North-
West
Pacific
61 Simulated derelict
demersal Japanese
spiny lobster
gillnets, Tateyama
Bay, Chiba
Prefecture, Japan
a,c,d,f,g,l Artificial reef experiment
1: 44 crustaceans,
11gastropods, 2 bony
fishes, 2 sand dollars per
9.4 m
2
net during study
period.
Artificial reef experiment
2: 33 crustaceans,
5gastropods, 5 bony
fishes, 1 sea cucumber per
9.4 m
2
net during study
period.
Sandy sea bed experiment
1: 8 crustaceans,
4gastropods, 1 bony fish
per 9.4 m
2
net during study
period.
Sandy sea bed experiment
2: 7 crustaceans,
1gastropod per 9.4 m
2
net
per during study period.
Duration of fishing
efficiency derelict gillnet in
an artificial reef, experiment
1: 561 days.
Duration of fishing
efficiency derelict gillnet in
an artificial reef, experiment
2: 284 days.
Duration of fishing
efficiency derelict gillnet on
sandy sea bed, experiment
1: 200 days.
(Sandy sea bed experiment
2, no significant correlation
between soak time
and number of caught
organisms).
Akiyama,
Saito and
Watanabe,
2007
North-
West
Pacific
61 Simulated derelict
salmon drift
gillnets, northwest
Pacific Ocean east
of Japan
a,c,d,f,k NA < 3 months for nets to form
a solid mass.
Mio et al.,
1990
North-
West
Pacific
61 Simulated derelict
demersal gillnet,
coastal Japan
a,f,g,j,l 455 fish per 165.6 m
2
net
until net reached 5% of
original fishing efficiency.
142 days to reach 5% of
initial fishing efficiency.
Nakashima
and
Matsuoka,
2004
North-
West
Pacific
61 Simulated derelict
demersal gillnet
wrapped on a fish
aggregation device,
and control fish
aggregation device
with no tangled
gillnet, coastal
Japan
a,f,g,j 191 fish per 2.25 m
2
net
per year.
No declining trend in ghost
fishing catch rate observed
during the 1149 day study
period.
Nakashima
and
Matsuoka,
2005
1
UNEP, 2005b, 2014. “None” indicates there is no UNEP Regional Sea Convention or Action Plan in the region for this study.
2
FAO, 2014.
3
(a) Deployed simulated derelict gillnets and/or trammel nets.
(b) Observed ALDFG from gillnet and/or trammel net fisheries.
(c) Simulated derelict gear used commercial gear design and fishing methods, in some cases modified to simulate derelict
conditions.
(d) Simulated derelict gear set at conventional fishing grounds, including cases where the study site was selected in a subset of
grounds to avoid disturbance, e.g. from conflict with mobile gear.
(e) Monitored catch and/or changes to gear condition via periodic retrieval of subset of gear.
(f) Monitored catch and/or changes to gear condition via in situ monitoring.
(g) Estimated short-term (hours to weeks) ghost fishing mortalities by counting the number of organisms that became newly
captured since a previous observation. Marked catch to enable the identification of new catch in subsequent monitoring event.
(h) Estimated short-term (hours to weeks) ghost fishing mortalities by counting the number of recently captured organisms in
‘good condition’ observed present at the time of monitoring.
(i) Fishing efficiency of derelict gear/simulated derelict gear at end of study period compared to that of in-use gear during the
same period and area as the study gear.
(j) Fishing efficiency of derelict gear/simulated derelict gear at end of study period compared to its initial fishing efficiency.
(k) Monitored ALDFG until cessation of ghost fishing, until cessation of fishing efficiency for target species, or until retained small
proportion of initial species-specific or total catch capacity based either on observations of ghost fishing catch rates or on net
condition factors that indicate catch capacity.
(l) Fit decay model to short-term ghost fishing catch rate data to: (i) estimate total ghost fishing mortality level over a study period
that ended before derelict gear ceased to ghost fish, or for the estimated duration of fishing efficiency; and/or (ii) estimate the
duration of fishing efficiency.
TABLE2C (CONTINUED)
Abandoned, lost and discarded gillnets and trammel nets
26
3.2 DISCUSSION
3.2.1 Methods and estimates of rates of abandonment, loss and discarding
and use in estimating ALDFG density
Information on the rate of abandonment, loss and discarding of fishing gear can
provide a requisite input to understand total ghost fishing mortality levels in a
spatially explicit area over a given time period. The most common method employed
to estimate the rate of abandonment, loss and discarding of fishing gear is to conduct
in-person interviews and remote surveys, e.g. via telephone, mail and e-mail, of
captains and crew of a fishery and of other experts (Santos et al., 2003a; Antonelis,
2012, 2013; Kim, Lee and Moon, 2014). Surveys of fishers and other experts
can provide a critically important first-order qualitative understanding of basic
characteristics of a fishery, including rates and density of ALDFG, when previously
little or no information was available (Gilman et al., 2010). Data from expert surveys
can then be validated through methods that provide more certain results (MacMullen
etal., 2003; Ayaz et al., 2010).
Experiments and analyses of observer and logbook programme data can provide
more robust estimates of rates of the abandonment, loss and discarding of fishing gear.
Long-time series of records of such incidents may be needed to account for potentially
high interannual variability as documented to occur in some fisheries (e.g. documented
in southern Norwegian coastal anchored gillnet fisheries, MacMullen et al., 2003). One
source of error, rates of ALDFG generated from IUU fishing may be substantially
different from estimates obtained from legal fisheries if a main cause of ALDFG by
the IUU fishers is abandonment of gear when operating illegally and a risk of detection
occurs (Imamura, 2011).
Estimates of gear loss rates should be explicit in indicating if they are for initial
“gross” gear loss or otherwise if the estimate is for “net” gear loss after accounting
for the proportion of the initially lost gear that was recovered (MacMullen et al.,
2003; Santos et al., 2003a). Vessels may recover a proportion or all of their lost gear
depending on the ability of the captain to locate the exact position where they set
the gear, to track the location of gear that moved from its original position, weather
conditions, the cause of the gear loss, etc. Lost gear resulting from cut float lines is
likely to have a high rate of recovery. If the gear loss was due to gear conflict (e.g. a
trawler towed over a set gillnet), then it is possible that only a section of the lost fleet
of set gear was towed away, and a large portion of the fleet might be recovered by the
vessel. If a storm or currents caused the gear to move position, then the probability of
finding the lost gear can be high if tracking technology is used, and otherwise the gear
is less likely to be retrieved.
Equation 1 provides a model using information on the rate of abandonment, loss
and discarding of fishing gear to estimate the density of ALDFG at a fixed point in
time and spatially explicit area.
Equation 1: D
ALDFG
= E*(A
f
+ L
f
+ D
f
) + ALDFG
c
– (ALDFG
r
+ ALDFG
d
)
The terms of Equation 1 are defined as follows:
D
ALDFG
: The density of ALDFG of a particular gear type, or from a specific
fishery, at a single point in time, and explicit spatial area such as the grounds where
a fishery operates, where units are: the unit of effort (e.g. length of net, area of net,
weight of nets) per area under assessment.
E: Effort, the mean number of active vessels per year in the fishery in question
times the number of years that the fishery in question has been operating.
A
f
, L
f
and D
f
: The rates of abandonment, loss and discarding of fishing gear by the
fishery in question into the study area, respectively, in units of: unit of effort per
vessel per selected time period.
27
Methods and estimates
ALDFG
c
: The rate of inputs of ALDFG into the study area from currents and
storms, in the unit of: unit of fishing effort per selected time period as used in the
A
f
, L
f
and D
f
terms.
ALDFG
r
: The rate of all sources of removals from the study area (transported
out of the area by currents and storms, recovered by the vessels of the fishery
in question, by vessels of other fisheries, by ALDFG collection programmes,
entangled on large marine organisms that migrate out of the study area, etc.) of
ALDFG that were introduced by the fishery in question, in the unit of: unit of
fishing effort per selected time period as used in the A
f
, L
f
and D
f
terms.
ALDFG
d
: The rate of complete decomposition of ALDFG in the study area, in
units of: unit of fishing effort per selected time period as used in the A
f
, L
f
and D
f
terms.
When combined with information on amount and spatial distribution of fishing
effort, information on the rate of abandonment, loss and discarding of fishing gear
can be used to estimate the density of ALDFG by gear type. However, especially in
open bodies of water, ALDFG can be carried by currents out of the area where it was
abandoned, lost or discarded, and ALDFG can similarly be carried into an area from
afar (e.g. Ebbesmeyer et al., 2012). Therefore, information on gear abandonment/loss/
discarding rates probably provide a less accurate basis to estimate the current amount
of ALDFG in a defined spatial area, such as the fishing grounds for a fishery, relative
to other methods for estimating the density of derelict gear, discussed in the following
section.
Of the 14 studies reviewed in Table2a, 13 surveyed a sample of fishers from a fishery,
one included marine pollution experts in its survey sample, and one study did not
explain the method employed to estimate the rate of abandonment, loss or discarding
of fishing gear from gillnet and trammel net fisheries. No studies were found that
estimated ALDFG based on data from experiments, observer programmes or logbook
programmes. Most of the studied fisheries employed coastal demersal gillnets and
trammel nets. Rates of gear loss were reported in units of length or area of nets lost per
vessel per year, and percentage of set gear lost. Nine of the 14studies reporting estimated
rates of producing derelict gear were from the Northeast Atlantic, resulting from the
European Commission’s FANTARED project (Table2a; MacMullen et al., 2003).
Several units were used to report the rates of abandonment, loss and discarding of
gillnets and trammel nets. Of the studies reviewed in Table2a, the following eight units
were used: (i) percentage of total gear set that was lost and not recovered (n=10); (ii)
number of net panels per vessel per year that were lost and not recovered (n=8); (iii)
number of discarded net panels per vessel per trip (n=1); (iv) length of nets per vessel
per year that were lost and not recovered (n=6); (v) length of nets per vessel per set that
were either abandoned, lost or discarded (n=1); (vi) weight of discarded nets per vessel
per year (n=1); (vii) percentage of vessels in a fishery that lose a fragment of set gear or
the entire set gear (n=1); and (viii) number of net fragments that are lost per year by all
vessels in a fishery (n=1).
The mean estimated percentage of lost gear from gillnet and trammel net fisheries,
where nets were set, lost and not subsequently retrieved by the vessel, was 0.9percent
(± 0.3 SEM, range 0–3.4percent, n=10) (e.g. 1 percent of gear is lost per vessel per
year) (Table2a). A 38percent CV (coefficient of variation, the standard deviation of
the mean was 38percent of the mean), with 7of the 10estimates falling outside ± one
SD from the mean, indicates that there was relatively high variability/ low consistency
in the 10 study findings. Similar CVs resulted when employing a unit of the number
of lost net panels per vessel per year (mean of 11.7 ± 3.0 SEM, 26percent CV, 6of
8estimates falling outside ± one SD from the mean) and length of lost nets per vessel
per year (mean of 1.1km/vessel/year ± 0.4SEM, 33percent CV, 4of 6estimates falling
outside ± one SD from the mean).
Abandoned, lost and discarded gillnets and trammel nets
28
In summary, there is high variability in the few available estimates of rates of
abandonment, loss and discarding of gear from gillnet and trammel net fisheries,
consistent with Macfadyen, Huntington and Cappel (2009). Studies have primarily
estimated rates of abandonment, loss and discarding of gear through fishers surveys;
estimates have not been based on data from experiments, observer programmes or
logbook programmes, which would have higher certainty findings and could be used
to validate the first-order estimates from the fishers surveys. Some study findings did
not specify whether estimates of loss rates accounted for the proportion of initially
lost gear that was recovered by the vessels. There has been inconsistent use of units to
report rates of abandonment, loss and discarding, using the length or area of nets lost
per vessel per year, and percentage of set gear that was lost. This precludes meaningful
comparisons of findings between studies and prevents the pooling of records. There
have been some relevant studies, few studies conducted outside of Europe, with most
studies having been conducted on coastal demersal gillnets and trammel nets; data
sources are dated and patchy spatially, temporally, and by gear type.
3.2.2 Methods and estimates of the density of ALDFG
Two common methods to estimate the density of ALDFG, the unit amount per unit
area of seafloor and/or per unit volume of water column, in a spatially explicit site,
are: (i) to survey the sea bed of a subset of the area of the total fishing grounds and
extrapolate to the total area; and (ii) to conduct interviews to obtain expert opinion
of estimates of the amount of derelict gear present in a designated area (e.g. Carr
and Cooper, 1987; Northwest Straits Foundation, 2007; Antonelis, 2013). In some
studies, survey sites were randomly selected to attempt to characterize the density of
ALDFG across the fishing grounds, while in others survey sites were selected based on
information from fishers identifying sites where they lost gear or observed ALDFG,
in some cases, to attempt to maximize the quantity of ALDFG retrieval (Misund et al.,
2006; Large et al., 2009). For example, identifying wrecks as having high concentrations
of ALDFG, Revill and Dunlin (2003) surveyed wrecks along a 100 km section of
coastline off northeast England, the United Kingdom of Great Britain and Northern
Ireland, to observe the number of ALDFG from gillnet fisheries. Northwest Straits
Foundation (2007) used a combination of observations of removed ALDFG, reports
of ALDFG received through a reporting system, and diver surveys focused on heavily
fished areas.
Observations by divers, and by sonar, video and photography deployed from
marine vessels, towed structures, manned submersibles and underwater ROVs, have
been used to survey for derelict gear (Carr and Cooper, 1987; MacMullen et al., 2003;
Revill and Dunlin, 2003; Ayaz et al., 2010). For example, Matsuoka, Nakashima
and Nagasawa (2005) estimated the amount of derelict gear in a study area through
information derived from a combination of interviews of fishers and from a seabed
survey using sidescan sonar. One trial where demersal gillnets were deployed to
simulate derelict gear found that towed sidescan sonar equipment was unreliable in
detecting the gear when high sea swell prevented maintaining the towed device at a
constant depth, the device was too distant from the sea bed or when the vessel speed
was too fast (MacMullen et al., 2003). Towing “creeper” grappling devices is another
method to estimate the density of ALDFG in a sampled area of a fishing ground (Kang,
2003; Misund et al., 2006; Large et al., 2009).
Estimates of the density of ALDFG at a point in time obtained from surveys of
a subset of a fishing ground can then be extrapolated to the entire fishing ground.
For example, in an early study to estimate the density of ALDFG at a commercial
demersal gillnet fishing ground, Way (1977) trawled with a grappling device and
reported the number of derelict gillnet fragments retrieved per number of hours of
trawling. Subsequent studies have reported findings in units of the number, length
29
Methods and estimates
or area of retrieved ALDFG from net fisheries per linear length surveyed (crept),
and extrapolated this to the length or area of derelict nets per area of fishing ground
(Misund et al., 2006; Large et al., 2009; Ayaz et al., 2010).
Models estimating the density of ALDFG in a total fishing ground based on
observations from a sample of the area can include a factor to account explicitly
for an estimate of error of the survey method, by accounting for an estimate of the
proportion of derelict gear present in the study site that was not observed during a
survey (MacMullen et al., 2003). In one experiment testing a new design of creeper,
27percent of the length of deployed simulated derelict demersal gillnets were retrieved,
and some or all of simulated derelict nets were retrieved in half of the tracks in which
derelict nets had been set (MacMullen et al., 2003).
Of the 10studies reviewed in Table2b, 5towed a grappling device, 2used sidescan
sonar, 4used divers, 1a crewed submersible, and 1an underwater ROV. Four of the
studies raised estimates from a sampled area to a larger fishing ground, one of which
accounted for an estimate of error in the sampling method (the proportion of total
ALDFG present along tow lines that a towed creeper did not retrieve). There were
5studies of coastal demersal gillnet and trammel net fisheries, 2of a coastal driftnet
fishery, 1of a coastal gillnet wreck fishery, 1of a deep-water gillnet fishery, and 1study
did not determine the source fisheries of retrieved derelict fishing nets (Table 2b).
Three of the 10 studies were from the Northeast Atlantic, 2 from the Northwest
Atlantic, 2from the Northeast Pacific, and 1each from the Baltic, Mediterranean and
Northwest Pacific (Table2b).
Variable units were used to report the density of ALDFG from gillnet and trammel
net fisheries: (i) length of derelict net per unit area (n=4); (ii) length of derelict net per
unit length of survey transects (n=4); (iii) number of net fragments per unit of sampling
effort (n=2); (iv) area of derelict nets per unit area (n=2); (v) number of net fragments
per area of a fishing ground (n=2); (vi) weight of derelict nets per unit area (n=1); (vii)
area of derelict nets per unit length of survey transects (n=1); and (viii) proportion of
surveyed wreck sites with derelict nets present (n=1).
The mean estimated density of ALDFG from gillnet and trammel net fisheries,
in units of length of nets per area of surveyed fishing grounds, was 4.4km net/km
2
fishing grounds (± 3.8 SEM, range 0.027–15.7, n=4) (Table 2b). An 86percent CV,
with all 4estimates falling outside ± one SD from the mean, indicates extremely high
variability. Similar variability was found when using results presented in units of length
of derelict gillnets per unit length of survey transects, with a mean density of 0.07km
gillnet/km survey transect (± 0.03 SEM, range 0.01–0.1, n=4), 39 percent CV, with
all 4 estimates falling outside ± one SD from the mean. One of the reviewed studies,
which reported the observed density of fishing nets without estimating the proportion
that was gillnet/trammel net gear (Kang, 2003), was excluded from the records used in
producing these summary statistics.
Studies have employed a mix of towing “creeper” grappling devices and various in
situ survey methods to estimate the density of ALDFG from gillnet and trammel net
fisheries. In Puget Sound, Washington, the United States of America, two studies used
a combination of interviews, surveys and direct observations from ALDFG removal
operations. Few studies towing creepers accounted for the estimated proportion of
ALDFG that the creeper did not recover. Main units for reporting ALDG density have
been the length of derelict nets per unit area of survey fishing grounds, or length per
unit length of survey transects. Consistent with Macfadyen, Huntington and Cappel
(2009), the sparse number of relevant studies are primarily from Europe, are largely
dated and spatially and temporally patchy, with large dispersion in estimates of the
density of ALDFG.
Abandoned, lost and discarded gillnets and trammel nets
30
3.2.3 Methods and estimates of ghost fishing mortality rates and duration
of fishing efficiency
Various methods have been employed to estimate ghost fishing mortality rates and
the duration of ghost fishing efficiency in ALDFG from gillnet and trammel net
fisheries, or in experimental nets deployed to simulate derelict gear (Table2c). In some
experimental studies, nets were deployed at conventional fishing grounds. In others,
study areas were selected away from conventional fishing grounds in order to avoid
disturbance by commercial fishing vessels using the same gear as well conflicts with
mobile gear, and to facilitate monitoring (shallow depth to aid observations by divers,
close to seaport to facilitate access). Most studies employing simulated derelict gear
employed commercial gear designs and fishing methods (Table2c). Study designs that
deviate from typical commercial operations might not characterize ghost fishing in the
commercial fishery (e.g. Kaiser et al., 1996).
In some studies, one end of simulated derelict demersal nets was loose, either by
not anchoring it or cutting the float line, and in some cases the net was dragged for a
certain period (e.g. 2minutes) or distance (e.g. 150m) to simulate a net that was lost
from interacting with towed gear, such as a trawl net dragging the gillnet or trammel
net along the bottom until eventually cutting the net, dragging away one portion and
leaving the remainder with one end still anchored (Kaiser et al., 1996; Erzini et al.,
1997; MacMullen et al., 2003; Revill and Dunlin, 2003; Santos et al., 2003b; Santos,
Gaspar and Monteiro, 2009). To simulate the loss of end markers or movement of a
fleet due to currents, which are common causes of gear becoming lost at deep-water
fishing grounds, studies simulating derelict gear at deep-water sites anchored both ends
of a fleet (Humborstad et al., 2003).
Duration of fishing efficiency has been estimated via periodic monitoring of derelict
or simulated derelict gear until the gear is observed to no longer retain any catching
capacity, no longer catches main market species, or retains a small proportion (e.g.
1–5 percent) of species-specific or total catch capacity relative to its initial fishing
efficiency or relative to in-use gear deployed in the same area and time (Kaiser et al.,
1996; Erzini et al., 1997; Revill and Dunlin, 2003; Sancho et al., 2003; Tschernij and
Larsson, 2003; Ayaz et al., 2006; Baeta, Costa and Cabral, 2009). The percentage of
retained fishing efficiency of ALDFG has been estimated by comparing the short-
period catch rate when first set to that after a period of soaking, or to that of similar
commercial in-use nets used in the same area and time.
Short-period (hours to weeks) ghost fishing mortality rates, in units of number or
biomass of catch per time period per unit of fishing effort by species or group, have
been estimated by counting the number of organisms that became newly captured since
a previous observation. To make this estimate, tags have been affixed to organisms
caught in derelict gear to enable the identification of new organisms caught in the net
since a previous monitoring event (Kaiser et al., 1996; Akiyama, Saito and Watanabe,
2007; Gilardi et al., 2010). Other studies assumed that catch in good condition observed
present in a net at the time of monitoring were caught within an estimated time period
prior to the monitoring event, based on previous observations made in that region.
For example, Santos et al. (2003b) and Santos, Gaspar and Monteiro (2009) and a
study in MacMullen et al. (2003) assumed “good condition” catch had been caught
in the previous 24hours. Another study in MacMullen et al. (2003) assumed that fish
degraded within 3days and shellfish within 15days. Sancho et al. (2003) assumed that
monkfish in “fresh” condition had been caught in the previous four days. Similarly,
Tschernij and Larsson (2003), upon retrieving simulated derelict demersal gillnets at
multiple-month intervals, categorized catch as marketable, decomposing and skeletons.
To estimate short-term catch rates in a simulated derelict deep-water demersal gillnet,
Humborstad et al. (2003) assumed that live fish and fish that were dead but with no
or minor damage had been recently captured, and excluded more degraded dead fish.
31
Methods and estimates
Use of catch condition to estimate the short-term catch rate is necessary for studies
of derelict gear at deep-water sites where in situ monitoring to tag organisms is not
possible (Humborstad et al., 2003). Monitoring has been conducted in situ by divers
and using ROVs, in some cases aided with the use of still photography and video,
attaching still and video cameras in one case to a balloon, and via repeated retrieval of
the nets (Kaiser et al., 1996; Erzini et al., 1997; Pawson, 2003; Sancho et al., 2003; Baeta,
Costa and Cabral, 2009; Gilardi et al., 2010).
Exponential regression decay models (or their inverse logarithmic function) have
been fitted to time series of records of short-period (24-hour to biweekly) catch rate
data in order to estimate a total ghost fishing mortality level over a study period that
ended before derelict gear ceased to ghost fish, or for the estimated duration of fishing
efficiency of an individual derelict net or group of nets, and to estimate duration of
fishing efficiency (e.g. Santos et al., 2003b; Santos, Gaspar and Monteiro, 2009). Catch
rate data can be soak time vs the number or length or weight of organisms caught
since the previous observation, or otherwise estimated to have been caught within a
certain time period prior to the current monitoring event per unit of effort. A decay
model could be fitted to catch rate data for a single species group (e.g. demersal fish,
crustaceans) and for a specific type of ALDFG (e.g. gillnet ALDFG on wreck, trammel
net ALDFG on flat sandy bottom). For example, fitting daily ghost fishing catch
rate data, in numbers of caught organisms, to Equation 2, an exponential regression
function:
Equation 2: N
t
= N
0
e
-rt
The terms of Equation 2 are defined as follows:
N
t
: The total number of organisms that an individual derelict net will catch from
time 0, the point in time when the net is deployed, to time t, the time period that
the derelict net continues to have catch efficiency.
N
0
: The y-intercept.
r: The rate of decrease in the daily catch rate.
To clarify, t is the duration of ghost fishing efficiency (the time between setting the
net and when the derelict net no longer catches organisms). Thus, Equation 2 provides
an estimate of total ghost fishing mortality level, in weight or number of organisms, for
an individual derelict net (Kaiser et al., 1996; Erzini et al., 1997; MacMullen et al., 2003;
Ayaz et al., 2006; Akiyama, Saito and Watanabe, 2007; Santos et al., 2003b; Santos,
Gaspar and Monteiro, 2009). Alternatively, Sancho et al. (2003) developed a model to
estimate the total level of target species ghost fishing catch in a derelict gillnet during
the period that the derelict net maintains fishing efficiency that assumed that there was
no trend in the ghost fishing mortality rate over an initial soak period, followed by a
period with a linearly decreasing trend in the catch rate until reaching t, cessation of
fishing efficiency.
Affixing tags to organisms caught in derelict gear, and to captive organisms placed
in the gear, enables monitoring of the change in condition of catch between monitoring
events and over a full study period. This enables the complete removal of ghost caught
organisms from a derelict net due to decomposition and scavenging to be accounted for
in models that estimate total ghost fishing mortality levels (Kaiser et al., 1996; Erzini
et al., 1997; Gilardi et al., 2010).
Given temporal variability in factors that affect the duration of fishing efficiency
at a fixed study site where an anchored net is deployed or over the area that a driftnet
covers during a study period, the season (or seasons) during which a study is conducted
can have a large effect on findings. For example, a study conducted during a season
with higher probability of storms might find lower ghost fishing mortality rates and
Abandoned, lost and discarded gillnets and trammel nets
32
levels than one conducted in a season with mild weather (Carr, Blott and Caruso, 1992;
Erzini et al., 1997; MacMullen et al., 2003; Sancho et al., 2003; Santos et al., 2003b).
Temporal (seasonal, annual) variability in the abundance of species caught by gillnets
and trammel nets, and in seasonal variability in degree of biofouling, also means that
the timing and duration of a study will affect observed ghost fishing mortality rates
and levels (e.g. Sancho et al., 2003; MacMullen et al., 2003). To obtain robust estimates
of the duration of fishing efficiency of ALDFG, relatively long study periods of years
can be required at fishing grounds with conditions that result in a long duration of
ghost fishing efficiency (Section1.4) (e.g. Kaiser et al., 1996; Nakashima and Matsuoka,
2004).
Ghost fishing mortality rates based on data from periodic monitoring of derelict
nets can result in large underestimates when organisms are caught in between
two monitoring events and are completely removed from the net before the net
is subsequently monitored (Kaiser et al., 1996; Erzini et al., 1997). This source of
uncertainty applies to very short intervals in between monitoring periods of less than
a day. For example, an organism can become captured in ALDFG, escape, and later
die as a direct or indirect result of the interaction, and predators can rapidly remove
recently caught live organisms (Carr, Blott and Caruso, 1992; Kaiser et al., 1996; Erzini
et al., 1997; Gilardi et al., 2010; Gilman et al., 2013). It also applies to intervals of days
to months. For example, scavengers have been observed to completely remove catch
over a period of days, and catch can completely decompose over periods of days to
weeks, depending on the location of the site and species of catch (Kaiser et al., 1996;
Erzini et al., 1997; Gilardi et al., 2010).
There is also uncertainty in estimating the fate of an observed new captured
organism that is not present upon a subsequent monitoring event. The organism may
have: escaped alive and survived; escaped injured and subsequently died from the
interaction; been completely removed by predators; died in the gear and fallen out from
mechanical action or been completely removed by scavengers; or died and completely
decomposed (Kawamura, 1993; Kaiser et al., 1996; Erzini et al., 1997; Akiyama, Saito
and Watanabe, 2007; Gilardi et al., 2010; Gilman et al., 2013).
Estimating the number of caught organisms in nets through periodic retrieval of
a subset of derelict nets may provide a large underestimate of ghost fishing mortality
rates as there is evidence that a large proportion of the caught organisms can drop out
of the nets during retrieval. For example, Gilardi et al. (2010) observed that 13percent
of invertebrates, 32percent of fish and 21percent of seabirds caught in derelict nets fell
out of the gear during the process of retrieving the net.
Matsuoka, Nakashima and Nagasawa (2005) developed a model of ghost fishing
removals per unit of time by extrapolating from experimentally derived estimates of
ghost fishing. Model inputs included estimates of: average species-specific mortality
rates per derelict gear per unit of time; the area of the fishing grounds; the amount
of derelict gear in the fishing ground; and the proportion of total derelict gear in the
area that continues to maintain some catching efficiency (Matsuoka, Nakashima and
Nagasawa, 2005).
Equation 3 identifies factors needed to estimate the total level of ghost fishing
mortality in a spatially explicit area over a selected time period:
Equation 3: G
p
= (D
FE-ALDFG
* A) * (N
t
* p*t
-1
)
The terms of Equation 3 are defined as follows:
G
p
: Total ghost fishing mortality level, of specified species or all catch, in a selected
study area over a specified time period p.
D
FE-ALDFG
: Estimate of the density of derelict gear retaining some ghost fishing
efficiency (i.e. it has been soaking for <t).
33
Methods and estimates
A: Area of the selected study area, such that the factor D
FE-ALDFG
*A produces the
total quantity of the unit of fishing effort of ALDFG with fishing efficiency in the
study area at a point in time.
N
t
: As in Equation 2, the total ghost fishing mortality level of a single derelict net
over the full period that the derelict gear continues to have fishing efficiency.
p*t
-1
: Referring to the definition of t provided in Equation 2, the proportion or
factor of duration of fishing efficiency that the selected study period p covers.
The model assumes that the mean age of ALDFG retaining fishing efficiency in the
site is half of t; i.e. that the rate of input of ALDFG is constant over p. The term p needs
to be sufficiently long to account for temporal variability in the various terms included
in the equation. For example, there might be cyclical intra-annual (seasonal) uneven
distribution of fishing effort, which causes temporal variability in the mean age of
derelict gear in the study area over a small time series. Moreover, there can be seasonal
variability in the loss of gear owing to seasonal variability in the frequency of conflicts
with towed gear sectors, or in inclement weather, which causes temporal variability in
the mean density of derelict gear in the study area over a short time series.
Fishing mortality by gear that was temporarily lost but soon retrieved by the vessel
might not be considered ghost fishing mortality, as the fisher did not permanently lose
control over the gear (Matsuoka, Nakashima and Nagasawa, 2005). While it can be
helpful to have agreed terminology for the various components of fishing mortality,
what is ultimately important is to ensure that all mortality sources are accurately
estimated and accounted for (ICES, 2005).
Duration of fishing efficiency and total ghost fishing mortality removals from a
derelict net has also been estimated for some taxa with time series data on the effective
fishing area and visibility of gillnets and trammel nets (Kaiser et al., 1996; Erzini et al.,
1997; Revill and Dunlin, 2003). For example, Kaiser et al. (1996) found a significant
relationship between headline height (an indicator of effective fishing area) and number
of fish caught in a simulated derelict gillnet, but not for crustaceans.
Given the paucity of data on ghost fishing mortality rates by gillnets and trammel
nets derived from experiments and from monitoring ALDFG in some regions, data
on catch rates from monitoring in-use gear is a more readily available source of
information that could be used in models to predict ghost fishing mortality rates. For
example, Carr and Cooper (1987) estimated that the average catch rates of observed
derelict demersal gillnets found in the Gulf of Maine were about 15percent of active
commercial groundfish gillnets, based on observations of the average profile, degree
of fouling and overall integrity of the derelict nets. While this method can be applied
to ALDFG from demersal fisheries, it may not be as appropriate in driftnet fisheries
where ghost fishing from derelict nets shifts to non-target species (often demersal) once
deposited on the seafloor.
There have been inconsistent definitions of fishing effort in gillnet and trammel net
fisheries (FAO, 2007; Gilman, 2009). It would be beneficial to standardize units for
the reporting of catch per unit of effort (CPUE) in gillnet and trammel net fisheries in
order to enable more meaningful comparisons between experiments and fisheries, and
to support the pooling of data for broad spatial-scale studies, including meta-analyses.
Alternative CPUE units for passive net fisheries identified in Gilman (2009) were the
number of caught organisms per: (i) trip, (ii) set, (iii) unit length of net, (iv) unit area of
net, (v) unit area per soak time, and (vi) the weight of the net. Many of these methods
do not provide effective ways to compare catch rates between fisheries, vessels within
a fishery, or even of catch rates by an individual vessel. For example, reporting catch
per horizontal length of a net can be a misleading measure of CPUE for comparisons
of different net designs if the net heights are dissimilar, and if organisms are not caught
in the same vertical portion of the net (Gilman et al., 2010). Fishing effort is not
Abandoned, lost and discarded gillnets and trammel nets
34
effectively characterized by identifying the number of vessels in a fleet or number of
fishers participating in a fishery, as this is not a reliable indicator to compare fishing
efficiency between fisheries. Moreover, comparisons of effort by gear with different
designs can be problematic because of inherent differences in the catching process
of the different gear (Chapter 2). There is also a need for gillnet and trammel net
ghost fishing catch rate units to be standardized to account for the duration that the
gear has been derelict, as there can be significant reductions in catching efficiency of
derelict gear over time (Kaiser et al., 1996; Erzini et al., 1997; Pawson, 2003; Matsuoka,
Nakashima and Nagasawa, 2005). The lack of use of standardized ghost fishing catch
rate units, this in addition to variable methods employed in studies that estimated ghost
fishing mortality rates (Table2c), has made it problematic to compare findings among
these studies (Chopin et al., 1996).
Of the 23 studies reviewed in Table2c, 56percent were from Europe (10Northeast
Atlantic Ocean, 2 Mediterranean Sea, 1 Baltic Sea), 22percent from the United States of
America (3Atlantic, 2Pacific), 13percent from Japan, and 9percent on the Pacific high
seas. Seventeen of the studies observed coastal demersal gillnets and/or trammel nets,
one was of a demersal wreck gillnet, one was of a coastal demersal gillnet entangled on
a FAD, one was of deep water demersal gillnets, two of high seas driftnets, and two
were of coastal driftnets deposited on the seafloor (Table2c).
Three of the 23 studies were based on observations of ALDFG from gillnet and
trammel net fisheries, the other 20having deployed simulated derelict gear (Table2c).
Of the 20using simulated derelict gear, most (14, 70percent) employed commercial
gear designs and fishing methods, and selected sites within commercial fishing grounds.
Fifteen of the 23studies (65percent) monitored the catch and condition of the gear
via in situ methods, 7(30percent) via periodic retrieval of subset of gear, and 1study
hauled the gear after a period of soaking without conducting periodic monitoring
during the soak (Table2c). Fourteen of the 23studies estimated short-term (hours to
weeks) ghost fishing mortalities. Eight of these 14(57percent) counted the number
of organisms that became newly captured since a previous observation, marking
catch to enable the identification of new catch in a subsequent monitoring event. The
remainder (6of 14, 43percent) counted the number of recently captured organisms in
“good condition” observed present at the time of monitoring. The fishing efficiency
of derelict gear/ simulated derelict gear at the end of the study period was compared
with its initial fishing efficiency in 3studies, and was compared to in-use gear during
the same period and area as the study gear in 5studies. In ten studies, derelict gear
was monitored until it ceased to retain any fishing efficiency or a small proportion of
initial efficiency, by either monitoring temporal changes in ghost fishing catch rates or
in net condition factors that are an indicator of catch capacity. Nine studies fitted an
ALDFG catch efficiency decay model to short-term ghost fishing catch rate data to:
(i) estimate a total ghost fishing mortality level over a study period that ended before
derelict gear ceased to ghost fish, or for the estimated duration of fishing efficiency;
and/or (ii) estimate the duration of fishing efficiency (Table2c).
The two studies of simulated derelict driftnets monitored net condition over time.
Mio et al. (1990) observed five 2km long driftnets had formed a single large tangled
mass after 3months of soaking. Gerrodette, Choy and Hiruki (1987, 1990) observed
four driftnets 50m, 100m, 350m and 1km long had been reduced to <5percent of
their original lengths after between 0.5and 10days of soaking.
Numerous units were used to report ghost fishing mortality rates: (i) number (n=4)
or weight (n=2) of organisms caught per unit area of net for the full duration of fishing
efficiency of fish; (ii) number of fishes per unit area of net for the full duration of
fishing efficiency (n=4) or to reach 5percent of initial efficiency (n=1); (iii) number
of crustaceans per unit area of net for the full duration of fishing efficiency (n=1);
(iv)number of fishes per unit length of net for the full duration of fishing efficiency
35
Methods and estimates
(n=2); (v)number of market species per unit area of net (n=1), weight of target species
per unit area of net (n=2), number of market species per unit length of net (n=1), and
weight of target species per unit length of net (n=1) for the full duration of fishing
efficiency; (vi) number and weight of non-market species per unit area of net for the
full duration of fishing efficiency (n=1); (vii) number of organisms, market species or
fish per unit area of net (n=4, 2 and 1, respectively) or unit length of net (n=4) per unit
time based on the observed level of catch during the study period; (viii) percentage of
fish catch efficiency of derelict nets surveyed in an area relative to that of in-use gear
(n=1); and (ix) weight of target species per unit area of net per day once catch efficiency
declined to 20–30percent of the initial fishing efficiency (Table2c).
Similarly, various units were likewise used for the duration of fishing efficiency:
(i)time for fishing efficiency to cease for all species (n=7), fishes (n=7), target species
(n=1), crustaceans (n=1), (ii) time for fishing efficiency to decline to <1percent of in-use
gear (n=2); (iii) time for fishing efficiency to decline to <5percent of initial efficiency
(n=1); (iv) time for fishing efficiency to decline to <7percent of initial efficiency (n=1);
and (v) time for fishing efficiency to decline to <30percent of initial efficiency.
The mean of ghost fishing mortality rate estimates, in units of the number of fishes
per unit area of gillnets and trammel nets for the full duration of fishing efficiency
(n=4), and to reach 5percent of initial catch efficiency (n=1), was 92.8fish per 100m
2
of net (±47.2SEM, range 22.4–275, n=5) (Table2c). There was very high variability,
with 51percent CV, with 4of 5of the records falling outside of ± one SD from the
mean. There was similar high variability when using results using units of the number
of organisms per unit area of net for the full duration of finfish fishing efficiency,
with a mean of 83.1organisms per 100m
2
of net (±25.5SEM, range 18.7–131.0, n=4),
31percent CV, and 3of the 4records falling outside ± one SD from the mean.
The mean of duration of ghost fishing efficiency estimates in units of the number
of weeks for ALDFG to cease catch efficiency of all organisms (n=7), decline to
<1percent of the catch rate of in-use gear (n=2), decline to <5percent of initial ghost
fishing efficiency (n=1) and decline to < 7percent of initial ghost fishing efficiency
(n=1) was 35.0weeks (±6.4SEM, range 8.3–80.1, n=11). An 18.2percent CV, with 7of
the 11records falling outside ± one SD from the mean, indicates moderate dispersion
in estimates. Relatively higher variability was found when using results presented
in units of the number of weeks to cease fishing efficiency of fishes, with a mean of
37.5weeks (±20.4SEM, range 3.1–156.0, n=7), 54percent CV, and 5of the 7records
falling outside ± one SD from the mean.
The large dispersion in estimates of ghost fishing mortality rates and duration of
ghost fishing efficiency is probably a result of extremely small sample sizes as well as
from the pooling of data from studies employing variable methods, studying ALDFG
with variable gear designs and materials, and at sites with variable environmental and
physiographic conditions (Section1.4).
More than half of the sparse number of relevant studies were conducted in Europe.
Most are dated, so that results might not characterize current fisheries. Studies were
spatially and temporally patchy, and there was large dispersion in estimates. A wide
variety of units were used to report ghost fishing catch rates and duration of fishing
efficiency, precluding the pooling of records and comparing findings. Most studies
designed to estimate ghost fishing mortality rates and the duration of fishing efficiency
by ALDFG have used simulated derelict gear. Most studied demersal nets set at coastal
sites within commercial fishing grounds at relatively shallow depths. Most studies
monitored the catch and condition of the gear via in situ methods, others periodically
retrieved a subset of gear. Many of the reviewed studies fit a ghost fishing efficiency
decay model to short-term ghost fishing catch rate data to estimate the duration of
fishing efficiency and the total ghost fishing mortality level for the estimated duration
of fishing efficiency.
Abandoned, lost and discarded gillnets and trammel nets
36
3.2.4 Research priorities for robust regional and global estimates of gillnet
and trammel net ghost fishing mortality levels
Findings highlight fundamental gaps in information to support robust estimates of
regional and global rates and levels of ghost fishing mortality by individual species and
higher taxonomic groups. Each of the terms of Equation3 represent basic information
needed to estimate the total level of ghost fishing mortality in a spatially explicit area
over a selected time period. There are large information deficits for each of the terms.
The following is a summary of the key information gaps:
There are small sample sizes for rates of producing derelict gear, density of derelict
gear, ghost fishing mortality rates and levels, by region and gear type. This reduces
the certainty in the estimation of means.
A portion of the small number of studies employed methods that reduced the
certainty of findings, such as introducing sampling bias in selecting study sites
and not accounting for organisms that are caught in a ghost net but are completely
removed in between two monitoring events. This increases the margin of error of
estimated means.
There was under-representation by region and gear type for estimates of rates
of producing derelict gear, of the density of derelict gear, and of ghost fishing
mortality rates and levels. This lack of balance in sample sizes by region and types
of gillnets and trammel nets would reduce the accuracy of estimates if data across
regions and fishing methods were pooled and then raised to produce a global
estimate.
Variable units have been used to report estimated rates of abandonment, loss and
discarding gear, density of ALDFG, and ghost fishing mortality rates and levels.
This prevents the pooling of some records, reducing sample sizes available to
estimate means.
There was wide dispersion in estimates of rates of producing derelict gear, density
of ALDFG, and of ghost fishing mortality rates. Raised estimates will similarly
have large estimates of error.
Many estimates are dated. They may not characterize ALDFG and ghost fishing
in contemporary fisheries.
There are no available databases providing estimates of levels of global gillnet and
trammel net fishing effort (Luca Garibaldi, FAO, personal communication, 2015;
e.g. the Global Capture Production database does not contain information on
fishing effort or catch levels by gear type, FAO, 2015). This information could,
in theory, be used to raise estimates of rates of producing ALDFG, the density of
ALDFG, and ghost fishing mortality levels.
In summary, these information gaps would result in very high uncertainty in an
estimate of a globally averaged ghost fishing mortality level, especially for taxa that are
rare-event captures, such as marine megafauna. Four priorities to fill these identified
information gaps are to:
Harmonize data collection protocols on ALDFG from gillnet and trammel net
fisheries in logbook and observer programmes where they are in place, and fill
gaps in ALDFG data collection protocols where they are not currently in place.
Producing larger logbook and observer programme datasets of records of rates
and amounts of abandoned, lost and discarded fishing gear, and rates of fishing
vessel encounters with ALDFG produced by other vessels, using standardized
data collection protocols, provides a priority resource for research on ALDFG
and ghost fishing (discussed further in Chapter4).
Conduct additional research, using best practice methods identified in this study
to reduce sources of uncertainty, to estimate ghost fishing mortality rates and
levels, balanced spatially, temporally and by type of gillnet and trammel net
37
Methods and estimates
fishing gear and method. Employ standardized units to report estimated rates of
abandonment, loss and discarding of gear, density of ALDFG, and ghost fishing
mortality rates and levels.
Conduct meta-analyses of data from relevant compiled studies to produce more
precise and accurate estimates of rates of producing ALDFG, density of ALDFG,
and ghost fishing mortality rates. Owing to the larger sample size plus the number
of studies, correctly designed meta-analyses can provide estimates with increased
precision and accuracy over estimates from individual studies, with increased
statistical power (e.g. Borenstein et al., 2009; Musyl et al., 2011).
Develop robust estimates of regional and global gillnet and trammel net fishing
effort. This activity is prioritized both to identify regions where managing
ALDFG by these gear types is most important, and to support estimates of ghost
fishing mortality levels.
39
4. Monitoring and management
by regional fishery bodies and
arrangements
4.1 INTRODUCTION, STUDY SCOPE AND METHODS
An assessment was conducted to benchmark international monitoring and management
of ALDFG and ghost fishing from marine gillnet and trammel net fisheries.
Regional fishery bodies (RFBs) and arrangements (RFAs), including regional fisheries
management organizations (RFMOs), were selected for inclusion in the study sample
if they: (i) have the competence to establish binding measures for marine capture
fisheries; (ii) have competence over fishery resources that are captured in an active
gillnet or trammel net fishery; and (iii) the agreement that formed the RFB or RFA is
in force (Table3) (FAO, 2011b; Gilman, Passfield and Nakamura, 2012). A regional
or global “body” is an organization that has established a secretariat that operates
under a governing body of member States, while a regional or global “agreement”
does not have such a secretariat (FAO, 2013). Hereafter, the bilateral (two parties) and
multilateral (three or more parties) bodies and arrangements included in the study
sample are collectively referred to as RFB/As.
The RFMOs are a type of RFB/A with the competence to establish binding
conservation and management measures. They provide a formal mechanism for
fishing States and States in whose jurisdiction fishery resources occur to meet their
international obligation to cooperate to sustainably govern shared living marine
resources throughout their distributions (UNCLOS Articles 63, 66(5), 118; Code
Articles 7.1.5, 6.12 [FAO. 1995]; PSMA Article 4(1)(b) [FAO, 2009]). Since the first
was established in 1923, RFMOs have played a critical role in multilateral fisheries
governance of stocks that straddle or occur beyond national jurisdiction and of highly
migratory stocks. While spatial, fishery and taxonomic gaps remain, a large proportion
of global marine fisheries are now managed by one or multiple RFMOs, and most
areas of the high seas are now covered by at least one RFMO (Lodge et al., 2007; FAO,
2011b; Gilman, Passfield and Nakamura, 2012).
The assessment identified the gear types employed in active fisheries that catch
covered species/stocks (fishery resources over which the RFB/A is mandated to
manage) in order to determine whether an RFB/A manages active gillnet or trammel
net fisheries. Convention and agreement texts typically identify covered species or
stocks over which the RFB/A has a mandate, and do not identify fisheries or gear
types that the convention or agreement covers. Therefore, this study determined
which RFB/As to include by identifying whether an RFB/A’s covered fishery
resources are caught in an active gillnet or trammel net fishery. This was determined
by reviewing: (i) relevant databases identifying catch by gear type; (ii) conservation
and management measures to identify what gear types they apply to; (iii) authorized
vessel lists to identify gear types of vessels authorized to fish in the convention area
for covered resources; (iv) member and cooperating non-member reports to the
RFB/A that identify the gear types used by their flag vessel to catch covered fishery
resources; (v) IUU fishing vessel lists to identify what gear types were employed
by vessels presumed to have carried out IUU fishing for resources covered by the
RFB/A; and (vi) input from secretariat staff on gear types of active fisheries over
which the RFB/A has competence.
Abandoned, lost and discarded gillnets and trammel nets
40
The convention texts or agreements that established the RFB/As were reviewed to
determine whether they include an explicit mandate to monitor and/or control ALDFG
and ghost fishing. Binding CMMs and logbook and observer data collection protocols
designed to monitor and mitigate (prevent and remediate) ALDFG and ghost fishing
adopted by the RFB/As included in the study were summarized. The CMMs were
included if they had a direct connection to monitoring or mitigating ALDFG and ghost
fishing, reviewed in Table1. Measures designed to deter IUU fishing can contribute
to reducing intentional abandonment and discarding of fishing gear (Table1). Only
measures designed to curtail IUU fishing directly related to deterring non-compliance
with requirements to monitor and control ALDFG and ghost fishing were included
in the assessment. Those CMMs related to deterring IUU fishing with broader, more
general aims indirectly related to ALDFG and ghost fishing were not included in
the assessment. The assessment included only binding measures, as the adoption of
binding measures, as opposed to voluntary measures, potentially demonstrates a
stronger political will by parties of an RFB/A to address issues resulting from ALDFG.
Moreover, binding measures may be more likely to be implemented in the domestic
legal and regulatory systems of member States (e.g. Barth and Dette, 2001).
TABLE3
RFB/As with the competence to establish binding measures for marine capture fisheries, and
the subset with competence over fishery resources captured in an active gillnet or trammel net
fishery, where the agreement that formed the RFB/A is in force
Body/Arrangement Acronym Active gillnet or trammel
net fishery catch covered
resources
Global, transocean and mandate broader than managing fishing
International Whaling Commission IWC
Commission for the Conservation of Antarctic Marine Living
Resources
CCAMLR
North Atlantic Salmon Conservation Organization NASCO X
North Pacific Anadromous Fish Commission NPAFC X
Tuna RFMOs
Commission for the Conservation of Southern Bluefin Tuna CCSBT
Indian Ocean Tuna Commission IOTC X
Inter-American Tropical Tuna Commission IATTC
International Commission for the Conservation of Atlantic Tunas ICCAT X
Western and Central Pacific Fisheries Commission WCPFC X
Other RFB/As
General Fisheries Commission for the Mediterranean GFCM X
International Pacific Halibut Commission IPHC
Joint Norwegian-Russian Fisheries Commission JNRFC X
Northwest Atlantic Fisheries Organization NAFO
North East Atlantic Fisheries Commission NEAFC
Pacific Salmon Commission PSC X
Regional Commission for Fisheries RECOFI X
South East Atlantic Fisheries Organisation SEAFO
Southern Indian Ocean Fisheries Agreement SIOFA X
South Pacific Regional Fisheries Management Organisation SPRFMO
This assessment adapted methods employed in two previous studies. Gilman,
Passfield and Nakamura (2012) assessed ecosystem-based governance of bycatch
by RFMOs having three or more parties, and included a criterion that considered
whether binding conservation and management measures to mitigate ghost fishing
41
Monitoring and management by regional fishery bodies and arrangements
from ALDFG had been adopted. Gilman (2015) assessed RFB/As’ data collection
protocols and management measures to mitigate ALDFG and ghost fishing in marine
capture fisheries. Consistent with Gilman, Passfield and Nakamura (2012) and Gilman
(2015), this study excluded the Convention on the Conservation and Management
of the Pollock Resources in the Central Bering Sea (CCBSP) and the International
Baltic Sea Fishery Commission (IBSFC). There are no active CCBSP-managed
fisheries because, since its convention came into effect in 1995, owing to low pollock
biomass, the annual harvest level has been set at zero and no individual national
quotas have been established (CCBSP, no date, 2012). The IBSFC was dissolved and
ceased activities on 31December 2005 (Aps et al., 2007; FAO, 2011c). Three bilateral
bodies and arrangements that, as RFMOs, are multilateral and thus have at least three
parties (e.g. Keohane, 1990), also make binding decisions (IPHC, JNRFC, PSC) were
considered for inclusion in this study. However, they were excluded by Gilman,
Passfield and Nakamura (2012) because their convention areas were exclusively or
predominately under national jurisdiction, and as a result, were presumed to probably
have governance structures that differ from RFMOs with at least three parties, whose
convention areas are predominantly on the high seas (Lugten, 2010). Moreover, two
RFMOs whose agreements entered into force since Gilman, Passfield and Nakamura
(2012) were considered for inclusion in the study sample (Southern Indian Ocean
Fisheries Agreement, and Convention on the Conservation and Management of High
Seas Fishery Resources in the South Pacific Ocean).
Four of the RFB/As considered for inclusion in this assessment are global or
transoceanic, and three of these have mandates that are broader than managing
marine fisheries (CCAMLR, NASCO, NPAFC). These are regional bodies with
the competence to establish binding fisheries CMMs, including measures related to
ALDFG, and thus their inclusion was deemed relevant.
Gilman, Passfield and Nakamura (2012) excluded the International Whaling
Commission (IWC) from their sample because it does not manage fisheries for
marine fish or shellfish species (IWC, 1946). The present study did consider the IWC
for inclusion in the study sample, consistent with Gilman (2015), because there are
active fisheries for whales for scientific research permissible under Article VIII of
the convention, active indigenous subsistence whaling permitted under current IWC
regulations, and, while a moratorium has been in effect since the 1985/86 season,
Norway and Iceland conduct commercial whaling under objection or reservation to
the moratorium (IWC, 1946, 2013b).
Of the 19 assessed RFB/As (Table 3), the following were excluded from the
study sample based on the determination that covered resources are not caught in
an active gillnet or trammel net fishery: CCAMLR, CCSBT, IATTC, IPHC, IWC,
NAFO, NEAFC, SEAFO and SPRFMO (NEAFC, 2008; Lopez-Abellan et al., 2010;
CCAMLR, 2012, 2013a, 2013b; IWC, 2011, 2013b; CCSBT, 2013; Bob Kennedy,
CCSBT Secretariat, personal communication, 2014; IATTC, 2013, 2014a, 2014b; IPHC,
2013, 2014; NAFO, 2013; Ricardo Federizon, NAFO, personal communication, 2014;
SEAFO, 2010, 2014; SPRFMO, 2013, 2014).
4.2 RESULTS: RFB/A MONITORING AND MANAGEMENT OF ALDFG AND
GHOST FISHING
For each of the ten RFB/As included in the study, the following sections provide
information on evidence for active gillnet and/or trammel net fisheries that take fishery
resources covered by the RFB/A, and whether the RFB/As mandate explicitly calls for
monitoring and controlling ALDFG and ghost fishing. Summaries are also provided
of data collection protocols related to ALDFG, and CMMs related to preventing
or remediating ALDFG and ghost fishing (Table4). For information on the area of
competence of these RFB/As, see FAO (2011b).
Abandoned, lost and discarded gillnets and trammel nets
42
TABLE4
RFB/As with the competence to establish binding CMMs for marine capture fisheries that have
competence over fishery resources captured in an active gillnet or trammel net fishery, whether
the mandate explicitly calls for monitoring and controlling ALDFG and ghost fishing, and
CMMs and observer and logbook data collection protocols related to monitoring, preventing or
remediating ALDFG and ghost fishing
Regional
fishery body/
arrangement
1
Types of gillnets and/or
trammel nets employed in
active fisheries that catch
covered species/stocks
Convention
specifically
mandates
monitoring
and/or
controlling
ALDFG and/or
ghost fishing?
CMMs and data collection protocols on
monitoring and controlling ALDFG and/or
ghost fishing by gillnets or trammel nets
Mandate broader than managing fisheries
NASCO Surface gillnet (NASCO,
2011).
N
Closed areas to targeted salmon fishing,
including on the high seas and in areas
beyond 12nautical miles from the baseline
of coastal States, excluding an area around
West Greenland and within the area of
fisheries jurisdiction of Faroe Islands (NASCO,
1983).
NPAFC Driftnet (NPAFC, 2010;
NFMS, 2012; Vladimir
Radchenko, NPAFC
Secretariat, personal
communication, 2014).
N
Prohibition on directed high-seas fishing for
North Pacific anadromous fish (NPAFC, 1992).
Tuna RFMOs
IOTC Driftnet, set gillnet
(IOTC, no date [Table11],
2009, 2010, 2014a,
2014b).
N
High seas large-scale driftnets are banned in
the IOTC area of competence (IOTC, 2012).
Gear marking of flag or radar reflector buoys
by day and light buoys by night is required
for the ends of nets, lines and gear in the
sea sufficient to indicate their position and
extent (IOTC, 2001, 2013a).
Marker buoys and similar objects intended
to indicate the location of fixed fishing gear
(IOTC, 2001, 2013a).
Onboard observers are to collect information
on the amount of ALDFG (IOTC, 2013c).
ICCAT Anchored gillnet (ICCAT,
2014).
N
Requirement to mark fishing gear to identify
ownership (ICCAT, 2003a).
Driftnets for fisheries of large pelagics are
banned in the Mediterranean (ICCAT, 2003c).
WCPFC Gillnet (WCPFC, 2013a). Y
High seas large-scale driftnets are banned on
the high seas in the WCPFC Convention Area
(WCPFC, 2008).
Observers of the WCPFC Regional Observer
Programme (currently placed in purse seine,
longline and pole-and-line fisheries) are to
record whether the vessel found ALDFG from
another vessel (WCPFC, no date; SPC, 2009a,
2009b).
Other RFMOs
GFCM Encircling gillnet,
driftnet, staked gillnet,
anchored gillnet,
combination gillnet –
trammel net, trammel
net (GFCM, 2014).
N
Driftnets for fisheries of large pelagics are
banned in the Mediterranean (GFCM, 2005).
Driftnets > 2.5km are banned, nets >1km
are required to remain attached to the vessel
unless the vessel is within the 12mile coastal
band, in which case a vessel may detach itself
from the net provided it keeps the net under
constant observation (GFCM, 1997b).
The use of gillnets with monofilament with a
twine diameter > 0.5mm is banned (GFCM,
2012).
JNRFC
2
Gillnet (JNRFC, no date
a; FCI, 2010; Lockwood
etal., 2010).
N A Norwegian regulation requires vessel
operators to record in catch reports whether
or not gear was lost during each fishing
operation (Directorate of Fisheries, 2012).
43
Monitoring and management by regional fishery bodies and arrangements
Regional
fishery body/
arrangement
1
Types of gillnets and/or
trammel nets employed in
active fisheries that catch
covered species/stocks
Convention
specifically
mandates
monitoring
and/or
controlling
ALDFG and/or
ghost fishing?
CMMs and data collection protocols on
monitoring and controlling ALDFG and/or
ghost fishing by gillnets or trammel nets
PSC
2
Driftnet, gillnet, net,
reefnet (PSC, 2013a,
2013b).
N None
3
RECOFI Anchored gillnet,
driftnet (RECOFI, 2008
[Appendix G], 2013a,
2013b).
N None
SIOFA Deep sea gillnet, driftnet
(SIOFA, 2013).
N None
1
Acronyms defined in Table3, and listed under “Abbreviations and acronyms” at front of publication.
2
Bilateral bodies and agreements.
3
PSC does not adopt fisheries regulations. Instead, PSC recommends fishery management measures that are adopted
and implemented by the two PSC parties (PSC, 2013a [Article IV]).
4.2.1 General Fisheries Commission for the Mediterranean (GFCM)
Active gillnet or trammel net fisheries under GFCM’s area of competence
Established by the Agreement for the Establishment of the General Fisheries Commission
for the Mediterranean, an agreement under Article XIV of the FAO Constitution, the
agreement does not identify specific fisheries or gear types that are to be managed
by the GFCM (GFCM, 1997a). The agreement mandates GFCM to, “promote the
development, conservation, rational management and best utilization,” of all living
marine resources in the area covered by GFCM (Article III(1), GFCM, 1997a).
The GFCM R  V   M A  O 
 GFCM A identified the following gillnet or trammel net gear types: encircling
gillnet, driftnet, staked gillnet, anchored gillnet, combination gillnet – trammel net,
and trammel net (GFCM, 2014). The GFCM manages fisheries for small pelagics,
bottom fisheries for mixed demersal fish and invertebrates, as well as fisheries for large
pelagics (GFCM, 2011a). Assessments and advice for fisheries for large pelagic species,
however, are provided by ICCAT (Pilar Hernandez, GFCM Secretariat, personal
communication, 2014).
Mandate Includes monitoring and/or controlling ALDFG and ghost fishing?
The Agreement does not explicitly mandate the GFCM to monitor, prevent or
remediate ALDFG or ghost fishing (GFCM, 1997a).
Relevant conservation and management measures
A GFCM recommendation bans driftnets > 2.5km, requires nets >1km to remain
attached to the vessel unless the vessel is within the 12 mile coastal band, in which
case a vessel may detach itself from the net provided it keeps the net under constant
observation (GFCM, 1997b). To reduce fishing mortality of juvenile swordfish, a GFCM
recommendation bans driftnets for fisheries of large pelagics in the Mediterranean
(GFCM, 2005). To mitigate cetacean bycatch, a GFCM recommendation bans the
use of gillnet fisheries using monofilament with a twine diameter greater than 0.5mm
starting 1January 2015 (GFCM, 2012).
Relevant observer and logbook data collection protocols
The GFCM does not implement a regional observer programme (GFCM, 2011a).
The GFCM logbook data collection protocols, required for use by vessel >15m in
TABLE4 (CONTINUED)
Abandoned, lost and discarded gillnets and trammel nets
44
length overall, do not call for vessel operators to report whether they abandoned, lost
or discarded any fishing gear or encountered any ALDFG during fishing operation
(GFCM, 2011b [Annex 1]).
4.2.2 Indian Ocean Tuna Commission (IOTC)
Active gillnet or trammel net fisheries under IOTC’s area of competence
The IOTC agreement identifies 16 species of tuna and tuna-like species and billfishes
that are covered by the agreement but does not identify covered fisheries or gear
types (IOTC, 1993). The IOTC fisheries data reporting requirements apply to any
contracting party and cooperating non-contracting party vessel under its flag that
catches covered species in the IOTC area, regardless of what gear type is employed
by the vessel (IOTC, no date). Gillnet and trammel net fishing gear types for which
statistics on the capture of IOTC covered species have been reported to date include:
driftnet and set gillnet (IOTC, no date [Table11], 2009, 2010, 2014a). The IOTC record
of currently authorized vessels includes vessels that use gillnets (IOTC, 2014b).
Mandate includes monitoring and/or controlling ALDFG and ghost fishing?
The IOTC agreement does not specifically mandate the IOTC to monitor or manage
ALDFG or ghost fishing (IOTC, 1993).
Relevant conservation and management measures
The IOTC (2012) bans large-scale (>2.5km in length) high seas driftnets within the
IOTC area of competence.
The IOTC (2001, 2013a) requires that gear be “marked appropriately, e.g. the ends
of nets, lines and gear in the sea, shall be fitted with flag or radar reflector buoys by day
and light buoys by night sufficient to indicate their position and extent”, which could
contribute to avoiding accidental gear loss when damaged by passing vessels or active
gear (Table1). In addition, these measures require “Marker buoys and similar objects
floating and on the surface, and intended to indicate the location of fixed fishing gear,
shall be clearly marked at all time with the letter(s) and/or number(s) of the vessel to
which they belong,” which might prevent ALDFG (Table1).
Relevant observer and logbook data collection protocols
The IOTC (2011) requires a minimum of 5percent observer coverage “for each gear
type by the fleet of each [contracting party and cooperating non-contracting party]
while fishing in the IOTC area of competence of 24metres overall length and over,
and under 24metres if they fish outside their Exclusive Economic Zone (EEZ).” A
template observer trip report adopted by the IOTC Scientific Committee and endorsed
by the IOTC, calls for the collection of information on, “lost fishing gear, such as
length of line lost, amount of net, and other gear such as floats,” (IOTC, 2013c).
The IOTC (2013a) requires vessels > 24 m in length overall and authorized to fish
in the IOTC convention area to maintain a logbook. The IOTC (2013b [Annex II and
III]), which identifies information to be recorded in logbooks by operators of longline,
purse seine, gillnet, pole-and-line, handline and trolling vessels, does not require the
recording of information on the vessel’s abandonment, loss or discarding of fishing
gear, or encounters with ALDFG from other vessels.
4.2.3 International Commission for the Conservation of Atlantic Tunas (ICCAT)
Active gillnet or trammel net fisheries under ICCAT’s area of competence
The International Convention for the Conservation of Atlantic Tunas (ICCAT) does
not identify specific fisheries or gear types covered by its convention but does state that
45
Monitoring and management by regional fishery bodies and arrangements
the convention applies to “populations of tuna and tuna-like fishes (the Scombriformes
with the exception of the families Trichiuridae and Gempylidae and the genus Scomber)
and such other species of fishes exploited in tuna fishing in the Convention area as
are not under investigation by another international fishery organization” (Article IV,
ICCAT, 1992). The ICCAT record of vessels (includes fishing vessels authorized to
fish in the ICCAT convention area that are longer than 20m, fish actively for bluefin
tuna, catch swordfish in the Mediterranean Sea, and fish for bigeye and yellowfin tuna)
includes anchored gillnet (ICCAT, 2014).
Mandate includes monitoring and/or controlling ALDFG and ghost fishing?
Its convention does not explicitly mandate ICCAT to monitor, prevent or remediate
ALDFG or ghost fishing (ICCAT, 1992).
Relevant conservation and management measures
An ICCAT recommendation requires parties to mark fishing gear “in such a way that
they can be readily identified in accordance with generally accepted standards such
as the FAO standard specification for the marking and the identification of fishing
vessels” (ICCAT, 2003a). Gear marking to identify ownership might prevent ALDFG
(Table1).
An ICCAT recommendation bans driftnets for fisheries of large pelagics in the
Mediterranean (ICCAT, 2003c).
Relevant observer and logbook data collection protocols
Data collection protocols for parties observer programmes, for the ICCAT Regional
Observer Programme for Eastern Atlantic and Mediterranean Bluefin Tuna, and
for logbook data collection requirements for vessels catching bluefin tuna in the
eastern Atlantic and Mediterranean do not call for collecting information on a vessel’s
abandonment, loss or discarding of fishing gear or encounters with ALDFG from
other vessels (ICCAT, 2012a [Paragraphs 90 and 91, and Annex 2 and 7]; 2012b).
Data collection protocols under the ICCAT Regional Observer Programme for
Bigeye and Yellowfin Tuna and for logbook data collection requirements for vessels
catching bigeye and yellowfin tunas do not call for the recording of information on a
vessel’s abandonment, loss or discarding of fishing gear or encounters with ALDFG
from other vessels (ICCAT, 2003b, 2011 [Annex 1 and 3]; 2012b).
4.2.4 Joint Norwegian-Russian Fisheries Commission (JNRFC)
Active gillnet or trammel net fisheries under JNRFC’s area of competence
The 1975 Framework Agreement (Agreement between the Government of the
Kingdom of Norway and the Government of the Union of Soviet Socialist Republics
on Co-operation in the Fishing Industry), which established the JNRFC, and is
complemented by the Mutual Access Agreement of 1976 and the 2010 Grey Zone
Agreement (the latter was not renewed and is no longer in force), mandates the JNRFC
to negotiate total allowable catches, allocation of these fishery resources between
Norway, the Russian Federation and third States, and to establish reciprocal access to
fisheries in national zones and quota exchanges for joint and national stocks (JNRFC,
1975, 2010). Decisions of the JNRFC are based on advice provided by the International
Council for the Exploration of the Sea (ICES) in response to requests by Norway
and the Russian Federation (JNRFC, no date a). The framework agreement does
not identify specific fisheries or gear types that fall under the JNRFC’s competence
area (JNRFC, 1975; Molenaar, Iferink and Rothwell, 2013). Measures adopted by the
JNRFC have been for shrimp demersal trawl fisheries and fisheries that target joint
stocks of demersal species of northeast Arctic cod, haddock, capelin and Greenland
Abandoned, lost and discarded gillnets and trammel nets
46
halibut (JNRFC, no date a, no date b). Gillnet is one of the fishing gear types identified
as being used by fisheries to target these demersal stocks managed by the JNRFC
(JNRFC, no date a; FCI, 2010; Lockwood et al., 2010).
Mandate includes monitoring and/or controlling ALDFG and ghost fishing?
The framework agreement does not mandate the JNRFC to monitor, prevent or
remediate ALDFG or ghost fishing (JNRFC, 1975).
Relevant conservation and management measures
No JNRFC measures relate to monitoring, preventing or remediating ALDFG or
ghost fishing (JNRFC, no date a).
Relevant observer and logbook data collection protocols
A Norwegian regulation requires vessel operators to record in catch reports whether
gear was lost during each fishing operation (Directorate of Fisheries, 2012). No JNRFC
requirements for observer or logbook data collection on Norwegian or Russian flagged
vessel loss, abandonment or discarding of fishing gear, or encounters with ALFG from
other vessels, have been adopted (JNRFC, no date a).
4.2.5 North Atlantic Salmon Conservation Organization (NASCO)
Active gillnet or trammel net fisheries under NASCO’s area of competence
The Convention for the Conservation of Salmon in the North Atlantic Ocean mandates
NASCO to propose regulatory measures for salmon fisheries in the area of fisheries
jurisdiction of one party that takes salmon originating in the rivers of another party
(Article7[1], NASCO, 1983). Regulatory measures or decisions have been adopted by
NASCO for the salmon fisheries at Faroe Islands and West Greenland in most years
since NASCO’s establishment in 1984. Currently, a relatively small West Greenland
nearshore surface gillnet internal-use fishery is the one active NASCO-managed
fishery. The reported catch in 2012 was 33 tonnes (9 900 salmon) (NASCO, 2013).
While there is one salmon river in Greenland, stocks fished in Greenland fisheries are
from North America and Southern European rivers (NASCO, 2008). Under NASCO
decisions there has been no commercial salmon fishery at Faroe Islands since the
early 1990s, and no non-commercial salmon fishing since 2000 (NASCO, 2010, 2011).
NASCO-member States of origin retain their management of salmon fisheries in
national homewaters (NASCO, 1983; Crozier et al., 2004).
Mandate includes monitoring and/or controlling ALDFG and ghost fishing?
Its convention does not explicitly mandate NASCO to monitor, prevent or remediate
ALDFG or ghost fishing (NASCO, 1983). The convention obligates NASCO
members of the North American Commission “with respect to its vessels and the area
under its fisheries jurisdiction, take the measures necessary to minimize by-catches of
salmon originating in the rivers of the other member” (Article 7[2], NASCO, 1983),
which might be interpreted to include salmon ghost fishing mortality in ALDFG. The
ICES has advised that the current salmon fisheries probably have nominal influence
on the marine ecosystem, but may affect species composition in riverine ecosystems
(NASCO, 2013a). Throughout the NASCO convention area, an increasing proportion
of salmon catch is taken in rivers or estuaries rather than coastal fisheries (NASCO,
2013a). NASCO management of distant-water salmon-targeted fisheries represents
but one small component of NASCO’s broad comprehensive North Atlantic salmon
conservation activities (NASCO, 2012).
47
Monitoring and management by regional fishery bodies and arrangements
Relevant conservation and management measures
The convention created a large area closed to targeted fisheries for Atlantic salmon,
including the high seas, and in areas beyond 12 nautical miles from the baseline of
coastal States, excluding an area around West Greenland (up to 40 nautical miles from
the baseline) and within the area of fisheries jurisdiction of Faroe Islands (NASCO,
1983).
Relevant Observer and Logbook Data Collection Protocols
NASCO’s Minimum Standard for Catch Statistics does not call for reporting information
on ALDFG (NASCO, 1993). Data collection protocols of a voluntary logbook system
employed in the Greenland fishery do not include reporting information on vessels
abandonment, loss or discarding of gear or encounters with ALDFG from other fisheries
(NASCO, 2013a). There is no NASCO regional or domestic observer programme for
the Greenland fishery (NASCO, 2013b [paragraph 7.3]).
4.2.6 North Pacific Anadromous Fish Commission (NPAFC)
Active gillnet or trammel net fisheries under NPAFC’s area of competence
The Convention of Anadromous Stocks in the North Pacific Ocean has a broad
aim of promoting the conservation of anadromous stocks in the convention area.
The convention mandates NPAFC to manage specific anadromous fish stocks in
the convention area, and does not identify fisheries or gear types managed by the
Commission (NPAFC, 1992, 2012a). The convention prohibits directed fisheries for
anadromous fish on the high seas, calls for minimized incidental catch of anadromous
fish, and prohibits the retention of anadromous fish taken incidentally during fishing
directed at non-anadromous fish (NPAFC, 1992). Incidental capture of anadromous
species in fisheries that occur in the convention area occurs in driftnet and other
fisheries (NPAFC, 2010; NMFS, 2012; Vladimir Radchenko, NPAFC Secretariat,
personal communication, 2014).
Mandate includes monitoring and/or controlling ALDFG and ghost fishing?
Its convention does not explicitly mandate NPAFC to monitor, prevent or remediate
ALDFG or ghost fishing (NPAFC, 1992). The convention obligates NPAFC members
to minimize the “incidental taking” of anadromous fish, defined by the Convention as
the “catching, taking or harvesting a species or stock of fish while conducting directed
fishing for another species or stock of fish”, and as such does not include ghost fishing
mortality in ALDFG (Article II(5) and Article III(1)(b), NPAFC, 1992).
Relevant conservation and management measures
The convention prohibits directed high-seas fishing for anadromous fish in the
convention area (NPAFC, 1992). Soon after the convention was adopted, NPAFC
successfully ended directed high seas salmon fishing in the North Pacific by the parties,
which had been mainly conducted using large-scale drift gillnets, and there has been a
large decrease in illegal high seas driftnet fishing in the convention area (NPAFC, 2010,
2012a).
Relevant observer and logbook data collection protocols
The NPAFC does not require vessels operating in the convention area to use logbooks
and does not manage a regional fisheries observer programme. Instead, activities focus
on enforcing the ban on high-seas driftnetting for anadromous fish in the NPAFC
convention area (NPAFC, 2012b).
Abandoned, lost and discarded gillnets and trammel nets
48
4.2.7 Pacific Salmon Commission (PSC)
Active gillnet or trammel net fisheries under PSC’s area of competence
The Treaty between the Government of Canada and the Government of the United
States of America Concerning Pacific Salmon, which established the Pacific Salmon
Commission, covers fisheries that harvest or seek to harvest salmon. Fishery resources
subject to the treaty are Pacific salmon stocks that originate in the waters of one party
(either Canada or the United States of America) and: (i) are subject to interception by
the other party; (ii) affect the management of stocks of the other party; or (iii) affect
biologically the stocks of the other party (PSC, 2013a [Article 1]). The treaty, and
PSC regulatory advice, identify several specific fisheries that are covered by the treaty,
including those using driftnet, gillnet, and unspecified net gears (PSC, 2013a, 2013b).
Mandate includes monitoring and/or controlling ALDFG and ghost fishing?
The treaty does not specifically mandate the PSC to monitor, prevent or remediate
ALDFG or ghost fishing (PSC, 2013a).
Relevant conservation and management measures
Regulatory advice from the PSC for the most current (2013) salmon season did not
address monitoring, preventing or remediating ALDFG (PSC, 2013b).
Relevant observer and logbook data collection protocols
The PSC does not manage a logbook or regional fisheries observer programme (PSC,
2013a). It does not adopt regulations for salmon fisheries but instead recommends
fishery regimes that are transmitted to the two member States, which then adopt the
fishery regimes, and establish and enforce regulations to implement the adopted fishery
regimes (PSC, 2013a [Article IV]).
4.2.8 Regional Commission for Fisheries (RECOFI)
Active gillnet or trammel net fisheries under RECOFI’s area of competence
The Agreement for the Establishment of the Regional Commission for Fisheries
established the Regional Commission for Fisheries (RECOFI) as an international
agreement under Article XIV of the FAO Constitution (FAO, 1999; RECOFI, 2009).
RECOFI is mandated to conserve and manage living marine resources in the RECOFI
agreement area (FAO, 1999 [Article III (1)(b)(i)]). RECOFI (2008 [Appendix G],
2013a, 2013b) has identified anchored and drift gillnet for finfish as included as some of
the RECOFI-managed marine capture fisheries that are conducted by member States.
Mandate includes monitoring and/or controlling ALDFG and ghost fishing?
The agreement does not explicitly mandate RECOFI to monitor, prevent or remediate
ALDFG (FAO, 1999).
Relevant conservation and management measures
There are no RECOFI binding measures related to monitoring, preventing or
remediating ALDFG (RECOFI has adopted one binding measure, on minimum data
reporting [RECOFI, 2011]).
Relevant observer and logbook data collection protocols
There is no regional observer coverage of RECOFI-managed fisheries (RECOFI,
2009, 2013b). Logbooks are not explicitly identified as being employed by RECOFI
members to meet RECOFI reporting requirements for national catch and effort data
(RECOFI, 2013a). Data reporting requirements of RECOFI members do not include
49
Monitoring and management by regional fishery bodies and arrangements
reporting information on vessel abandonment, loss or discarding of fishing gear, or
encounters with ALDFG of other vessels (RECOFI, 2011).
4.2.9 Southern Indian Ocean Fisheries Agreement (SIOFA)
Active gillnet or trammel net fisheries under SIOFAs area of competence
SIOFA, a regional fishery arrangement established by the Southern Indian Ocean
Fisheries Agreement, does not identify specific fisheries or gear types covered by the
agreement, but identifies covered fishery resources. SIOFA-covered fishery resources
are fish, molluscs, crustaceans and other sedentary species within the agreement area,
excluding sedentary species under the jurisdiction of coastal States, and excluding
highly migratory species listed in AnnexI of the United Nations Convention on the
Law of the Sea of 10 December 1982 (UNCLOS) (SIOFA, 2006 [Article 1(f)]). The
need for conservation and management measures for deep-sea gillnet and large-scale
driftnet fisheries was discussed at the first SIOFA meeting (SIOFA, 2013), indicating
that these gear types are employed in active fisheries for SIOFA-covered resources.
Mandate includes monitoring and/or controlling ALDFG and ghost fishing?
The agreement does not explicitly identify a mandate to monitor, prevent or remediate
ALDFG or ghost fishing (SIOFA, 2006).
Relevant conservation and management measures
No relevant measures have been adopted by SIOFA (European Commission, 2014).
SIOFA entered into force in 2012. It held its first meeting in October 2013, where
decisions focused on establishing rules of procedure, with a second meeting scheduled
for March 2015 (SIOFA, 2013; European Commission, 2014). As a regional fishery
agreement, unlike RFBs, SIOFA does not automatically establish a commission.
Instead, issues are to be discussed at an annual SIOFA meeting of parties.
Relevant observer and logbook data collection protocols
A SIOFA resolution that identifies recommended information to be collected from
high seas non-tuna fisheries does not call for the collection of information on vessel
abandonment, loss or discarding of fishing gear or on encounters with ALDFG of
other vessels (SIOFA, 2004).
4.2.10 Western and Central Pacific Fisheries Commission (WCPFC)
Active gillnet or trammel net fisheries under WCPFC’s area of competence
The convention of the WCPFC applies to all stocks of highly migratory fish (defined as
all fish stocks of the species listed in Annex I of UNCLOS occurring in the convention
area and such other species of fish as the commission may determine) within the
convention area, except sauries (United Nations, 1982; WCPFC, 2000). However, the
Convention does not identify specific fisheries or gear types covered by the convention
(WCPFC, 2000).
Fishing gear types reported used by member and cooperating non-member fisheries
operating in the WCPFC convention area in calendar year 2012 to catch WCPFC-
covered stocks included gillnet (WCPFC, 2013a). Of the 30countries, territories and
entities that reported having one or more fishery active in the WCPFC convention area
in 2012, only one reported catch with gillnet (VietNam) (WCPFC, 2013a).
The main industrial fishing methods employed in the WCPFC convention area to
catch covered stocks are purse seine, longline, pole-and-line and troll (Miyake et al.,
2010; SPC, 2013; Anthony Beeching, WCPFC Secretariat, personal communication,
2014). However, a number of other gear types, including gillnets, which are used
Abandoned, lost and discarded gillnets and trammel nets
50
primarily in domestic waters, also contribute to fishing mortality of WCPFC-covered
stocks (e.g. 20percent of reported landings of yellowfin tuna in the WCPFC statistical
area in 2012 was by gear types other than these four main industrial methods) (Miyake
et al., 2010; SPC, 2013). While some WCPFC CMMs are explicit in not being binding
in archipelagic waters and territorial seas (e.g. WCPFC, 2013b [Paragraph I(12)]),
there is a lack of consensus among members and cooperating non-members regarding
whether the convention area includes “domestic” archipelagic waters, territorial seas
and EEZs. Moreover, it is unclear whether, under the convention, members and
cooperating non-members are obligated to minimize adverse effects of ALDFG in
their “domestic” waters and fisheries (WCPFC, 2000).
Mandate includes monitoring and/or controlling ALDFG and ghost fishing?
The convention explicitly mandates the Commission to “adopt measures to minimize 
catch by lost or abandoned gear [and] pollution originating from fishing vessels”
(WCPFC, 2000 [Article 5(e)]).
Relevant conservation and management measures
Use of large scale drift gillnets (>2.5km in length) is prohibited on the high seas in the
WCPFC convention area, in part, to avoid ghost fishing (WCPFC, 2008).
Relevant observer and logbook data collection protocols
Observers of the WCPFC regional observer programme are to record whether the
vessel abandoned, lost or discarded any fishing gear, whether the vessel found ALDFG
from another vessel, and whether the vessel failed to report any lost or abandoned gear
if required by the country in whose waters the vessel was fishing (WCPFC, no date;
SPC, 2009a, 2009b). The WCPFC regional observer programme currently provides
coverage of purse seine, longline and pole-and-line fisheries (SPC, 2009c; Gilman,
Passfield and Nakamura, 2012). Therefore, of the observer data collection protocols,
the observer data collection protocol of recording vessel encounters with ALDFG
produced by other vessels, including ALDFG from gillnet and trammel net fisheries, is
of relevance to this assessment. Logbook forms for longline, purse seine, pole-and-line,
handline, troll and artisanal vessels do not have vessel operators report whether they
abandoned, lost or discarded any fishing gear or encountered any ALDFG from other
vessels (SPC, 2007a, 2007b, 2007c, 2009d, 2009e, 2010, 2011).
4.3 Discussion and conclusions
4.3.1 RFB/A mandate to monitor and control ALDFG
Of the ten RFB/As included in the assessment, one, WCPFC, has a convention that
explicitly mandates mitigating ALDFG and ghost fishing (Table 4). Members might
consider whether modifying mandates is necessary to enable RFB/As to adopt needed
measures to effectively monitor and manage ALDFG and ghost fishing by gillnet and
trammel net fisheries.
4.3.2 RFB/A logbook and observer data collection protocols on ALDFG
Of the ten assessed RFB/As, three have logbook and/or observer data collection
protocols that call for reporting abandonment, loss and discarding of fishing gear
from gillnet or trammel net fisheries (IOTC, JNRFC and WCPFC; Tables4 and 5).
Relevant observer data collection protocols require observers to collect information
on the amount of ALDFG (IOTC, no date), and whether the vessel found ALDFG
generated by another vessel (including gillnets and trammel nets) (WCPFC, no date;
SPC, 2009a, 2009b). One relevant logbook data collection protocol requires vessel
operators to record whether gear was lost during each fishing operation (Directorate
of Fisheries, 2012). There is a need to harmonize ALDFG data fields, data collection
51
Monitoring and management by regional fishery bodies and arrangements
protocols, and database formats where they are in place, and to fill gaps for those
lacking procedures to collect and report this information. Standardizing data fields,
data collection protocols and database formats facilitates comparisons between regions,
enables pooling of data necessary to support large spatial scale analyses within and
across regions, and enables standardization of training materials and courses within and
across regions (Gilman and Hall, 2015).
Collecting and reporting accurate information on the abandonment, loss and
discarding of gillnets and trammel nets can contribute to more accurate estimates of
the ecological and socio-economic effects of ALDFG, including from ghost fishing
mortality. However, reporting systems on ALDFG only contribute to mitigating
ALDFG and ghost fishing if they are implemented in combination with a derelict gear
retrieval programme, where retrieval responses ideally are conducted as close to the
time of loss as possible to maximize the likelihood of finding and then removing the
lost gear.
TABLE5
RFB/As with the competence to establish binding measures
Body or agreement
1
Bilateral or
multilateral
Observer or logbook data
collection on ALDFG from
gillnet or trammel net
fisheries
≥1 binding measure related to
monitoring or controlling ALDFG
or ghost fishing in gillnet or
trammel net fisheries
Mandate broader than managing fisheries
NASCO Multilateral X
NPAFC Multilateral X
Tuna RFMOs
IOTC Multilateral X X
ICCAT Multilateral X
WCPFC Multilateral X X
Other RFB/As
GFCM Multilateral X
JNRFC Bilateral X
PSC Bilateral
RECOFI Multilateral
SIOFA Multilateral
1
Acronyms defined in Table3, and listed under Abbreviations and acronyms.
4.3.3 RFB/A controls of ALDFG and ghost fishing
Of the ten assessed RFB/As, six had one or more binding measure that contribute to
controlling ALDFG or ghost fishing from gillnet or trammel net fisheries (Table5).
Table6 identifies which of the 18categories of measures for preventing and remediating
ALDFG and ghost fishing described in Table1 that each RFB/A employs. The six RFB/
As with relevant measures are making use of a small proportion of available approaches
to prevent and remediate ALDFG and ghost fishing. Of 18 identified categories of
measures to avoid and prevent ALDFG and ghost fishing, 13are not used by any of
the 10assessed RFB/As. As explained in Chapter1, while many of these 18categories
of methods have broad fisheries management purposes that do not have a primary
purpose of managing ALDFG and ghost fishing, their implementation contributes to
mitigating ALDFG and ghost fishing. Of the 18categories of methods, the 5with a
primary purpose of avoiding and remediating ALDFG and ghost fishing are: (i)raised
awareness and compliance with the international ban on intentional discarding and
abandonment of fishing gear at sea under MARPOL Annex V; (ii)economic incentives
and disincentives; (iii) port reception facilities for both unwanted fishing gear and
ALDFG; (iv) removal of ALDFG detected inadvertently or via programmes that
Abandoned, lost and discarded gillnets and trammel nets
52
search for derelict gear; and (v) programmes to disable the ghost fishing efficiency
of ALDFG (Table 1). None of the ten RFB/As had binding measures in place to
implement any of these five methods (Table6).
Spatial and temporal planning and management measures were used by six of the
RFB/As to control ALDFG and ghost fishing (Tables4 and 6). Most of these measures
prohibit the use of gillnet and trammel net gear in part or all of the RFB/A’s area
of competence, which contributes to reducing ALDFG and ghost fishing (Table4).
However, none of the measures separates passive and mobile gear sectors to avoid gear
conflicts and concomitant gear loss, or prohibits fishing in areas where there is a high
probability of gear loss and abandonment owing to contact with submerged features,
which are both common causes of ALDFG (MacMullen et al., 2003; Macfadyen,
Huntington and Cappel, 2009; FAO, 2010a; Antonelis, 2012, 2013).
Gear marking to identify ownership and to increase passive surface gear visibility
were used by 2 and 1 of the RFB/As, respectively (Tables 4 and 6). Gear marking
to identify ownership can create a disincentive for the deliberate abandonment and
discarding of unwanted gear, an incentive to retrieve lost gear, and facilitate enforcement
actions of violations of rules on monitoring and controlling ALDFG (Table1). Gear
marking to increase the visibility of passive gear can contribute to avoiding gear loss
through interactions with passing vessels or active gear (FAO, 1993; Table1).
A GFCM measure, which established a maximum gillnet twine diameter (GFCM,
2012), was the one measure requiring the use of a gear technology method with the
TABLE6
Methods for preventing and remediating ALDFG and ghost fishing in marine gillnet and
trammel net fisheries, and which are required by the RFB/As assessed in this study
Method Body/agreement with ≥ 1 relevant binding
measure in effect
1
Preventive
Gear marking to identify owner ICCAT, IOTC
Gear marking to increase visibility of passive gear IOTC
Technology to avoid unwanted gear contact with sea bed
Technology to track gear position
Gear technology to reduce the incidence of gear loss
Input controls, including limit on soak time
Periodic or constant observation of passive gear GFCM
Spatial and temporal planning (including measures
banning gillnets and trammel nets gears in part of or in
the entire area of competence of an RFB/A)
GFCM, ICCAT, IOTC, NASCO, NPAFC, WCPFC
Deter IUU fishing
Prohibition of intentional abandonment and discarding
of fishing gear at sea
Economic incentives and disincentives
Port reception facilities for unwanted gear
Training for new entrants
Remedial
ALDFG port reception and recycling facilities
Detect and remove ALDFG
Disable ghost fishing efficiency of ALDFG
Gear technology that increases ghost fishing selectivity in
passive gear ALDFG
Less durable and degradable gear to reduce ghost fishing
duration
GFCM
1
Acronyms defined in Table3,and listed under Abbreviations and acronyms.
53
Monitoring and management by regional fishery bodies and arrangements
potential to reduce ghost fishing mortality of cetaceans and other taxa that are strong
enough to escape from the less durable gear. The GFCM was also the only RFB/A to
have a measure in place requiring fishing practices (requires large nets to be attached
to the vessel or be under constant observation [GFCM, 1997b]) that could reduce the
incidence of ALDFG. The RFB/As could consider additional measures prescribing
the use of gear technology methods and fishing practices that prevent and remediate
ALDFG and ghost fishing by gillnets and trammel nets (Table 1). None of the ten
RFB/As required gillnet and trammel net fishing vessels to attempt to detect and
attempt to remove ALDFG and to report ALDFG that they could not retrieve, an
additional approach to remediate ALDFG (Table1).
4.3.4 Monitoring and controlling ALDFG and ghost fishing by bilateral and
multilateral arrangements and bodies
Two of the ten arrangements and bodies in the study sample are bilateral arrangements
(Table 5). In general, bilateral arrangements and bodies have convention areas that
are exclusively or predominately under national jurisdiction, while multilateral
arrangements and bodies (≥3parties) have convention areas that fall predominantly in
the high seas (Lugten, 2010; Gilman, Passfield and Nakamura, 2012). The observation
that two of the three identified bilateral arrangements/bodies have competence over
fishery resources captured in an active gillnet or trammel net fishery (Table 3) is
consistent with the understanding that gillnet and trammel net fisheries largely occur
in coastal areas (Bjordal, 2002; MacMullen et al., 2003). The observation that neither
the JNRFC nor the PSC, the two bilateral arrangements included in the study sample,
had binding measures in place to avoid or mitigate ALDFG or ghost fishing from
gillnet and trammel net fisheries, while 6of the 8multilateral arrangements and bodies
did have relevant measures in place, was unexpected given that these gear types are
predominant in nearshore areas.
55
5. Summary and recommendations
Abandoned, lost or otherwise discarded fishing gear causes substantial ecological
and socio-economic problems. Ghost fishing is one problem resulting from ALDFG
that has received increasing international attention over the past decade. Ghost
fishing mortality is infrequently accounted for in fisheries management, potentially
compromising the accuracy of population and stock assessment models and efficacy
of harvest strategies. Ghost fishing by ALDFG removes both target and non-target
species. Species with relatively low fecundity and other life-history characteristics that
make them particularly sensitive to anthropogenic mortality sources are also subject to
ghost fishing mortality. These include species of seabirds, sea turtles, marine mammals
and elasmobranchs, some of which are endangered, threatened and protected.
Mortalities from ghost fishing by ALDFG are also a source of wastage and reduce
the sustainable production of fishery resources and economic opportunities for the
marine capture sector. Social welfare issues are also raised over ghost fishing mortality
of flagship megafauna, as well as the extensive duration for some organisms caught in
ALDFG to succumb relative to captures in in-use gear.
Marine gillnets and trammel nets have relatively high ghost fishing potential. These
gear types are used worldwide primarily in coastal, artisanal fisheries, and supply
about a fifth of global marine fisheries landings. Recognizing this, FAO and UNEP
commissioned this study to identify best practice methods for estimating ghost fishing
rates and levels, priority research needs, and the status of international monitoring and
management of ALDFG and ghost fishing by marine gillnet and trammel net fisheries.
A sample of studies were compiled and synthesized in order to document methods
and estimates of: rates of abandonment, loss and discarding of gillnets and trammel
nets; the density of ALDFG; the duration of fishing efficiency of ALDFG; and ghost
fishing mortality rates of ALDFG. This provided a basis for understanding the degree
of dispersion in methods and findings, and augmented the state of understanding of
the severity of ALDFG and ghost fishing by gillnet and trammel net fisheries. In
addition, general best practice estimation methods to reduce uncertainty and priority
information gaps were identified to provide robust estimates of regional and global
ghost fishing mortality rates and levels by ALDFG.
Relative to some other gear types, and relative to other sources of collateral
fishing mortality, there has been good progress in the development of methods
to estimate ghost fishing mortality rates and the duration of fishing efficiency in
gillnets and trammel nets. However, a wide variety of units have been employed for
rates of abandonment, loss and discarding, density, ghost fishing catch rates, and
duration of fishing efficiency of ALDFG from gillnet and trammel net fisheries. This
precludes meaningful comparisons of findings between most studies, and prevents
pooling data. The few relevant studies are primarily from Europe, are largely dated
and spatially and temporally patchy, with large dispersion in estimates. Potentially
significant explanatory factors were highly variable amongst the compiled studies. This
heterogeneity of the sampled studies, small sample size, and high dispersion in rates of
production and density of ALDFG and ghost fishing mortality rates and duration of
fishing efficiency in gillnet and trammel net fisheries means that the observed means
are not generalizable regionally or globally.
A mean of 0.9 percent (± 0.3 SEM, 38 percent CV, n=10) of gear was lost from
gillnet and trammel net fisheries and was not subsequently retrieved. Similar high
dispersion in results were found for studies that reported findings as the number of lost
Abandoned, lost and discarded gillnets and trammel nets
56
net panels per vessel per year (26percent CV, n=8) and length of lost nets per vessel
per year (33percent CV, n=6). Few studies estimated rates of gear abandonment and
discarding. Studies have primarily estimated gear loss rates through fishers surveys;
estimates have not been based on data from experiments, observer programmes or
logbook programmes, which could validate the first-order estimates from the fishers
surveys.
There was extremely high variability in estimates of the density of ALDFG from
gillnet and trammel net fisheries reported in units of length of nets per area of surveyed
fishing grounds (mean of 4.4km of net/km
2
fishing grounds ± 3.8 SEM, 86percent
CV, n=4). There was also low consistency in findings using a unit of length of
derelict gillnets per unit length of survey transects (39percent CV, n=4). Studies have
employed a mix of fisher interviews, towing “creeper” grappling devices and various
in situ survey methods to estimate the density of ALDFG from gillnet and trammel
net fisheries. Few studies towing creepers accounted for the estimated proportion of
ALDFG that the creeper did not recover. Main units for reporting ALDG density have
been the length of derelict nets per unit area of survey fishing grounds, or length per
unit length of survey transects.
There was similarly high dispersion in estimates of ghost fishing mortality rates
and duration of ghost fishing efficiency. The mean number of ghost caught fishes per
unit area of gillnets and trammel nets for the full duration of fishing efficiency or to
reach 5percent of initial catch efficiency was 92.8fish per 100m
2
of net (±47.2SEM,
51percent CV, n=5). Similar variability was found with results presented in units of
the number of organisms per unit area of net for the full duration of finfish fishing
efficiency (mean of 83.1organisms per 100m
2
of net, ±25.5SEM, 31percent CV, n=4).
There was moderate variability in estimates of the duration of ghost fishing efficiency
in units of the unit of time to cease catch efficiency of all organisms, or to decline to a
small percentage of in-use gear or of initial ghost fishing efficiency (mean of 35.0weeks,
±6.4SEM, 18.2percent CV, n=11). There was higher variability when using results
presented in units of the unit of time to cease fishing efficiency of fishes (mean of
37.5weeks, ±20.4 SEM, 54percent CV, n=7). Estimates of the duration of driftnet
fishing efficiency have ranged from less than a day for small, 50–100m length nets,
to three months for 2km length nets. Most studies designed to estimate ghost fishing
mortality rates and the duration of fishing efficiency by ALDFG have used simulated
derelict gear. Most studied demersal nets set at coastal sites within commercial fishing
grounds at relatively shallow depths. Most studies monitored the catch and condition
of the gear via in situ methods, others periodically retrieved a subset of gear. Many of
the reviewed studies fit a ghost fishing efficiency exponential regression decay model to
short-term ghost fishing catch rate data to estimate the duration of fishing efficiency and
the total ghost fishing mortality level for the estimated duration of fishing efficiency.
The large dispersion in estimates of ghost fishing mortality rates and duration of ghost
fishing efficiency is probably a result of extremely small sample sizes as well as from
pooling data from studies employing variable methods, studying ALDFG with variable
gear designs and materials, and at sites with variable environmental and physiographic
conditions (e.g. flat substrate in shallow water with strong currents and abundant
biofouling organisms, debris and particulate matter vs entangled on three-dimensional
objects in deep water with weak current and limited biofoulers, debris and particulate
matter; site with active towed fishing gear vs site lacking active gear fishing effort).
There were small sample sizes in available estimates of rates of producing derelict
gear, density of derelict gear, ghost fishing mortality rates and levels, with under-
representation by region and gear type. Many estimates are dated, and may not
characterize ALDFG and ghost fishing in contemporary fisheries. Large sources of
uncertainty were introduced in some of these studies, such as owing to sampling
bias in selecting study sites. The use of variable units to report estimates prevents the
57
Summary and recommendations
pooling of some records, reducing sample sizes available to estimate means. There are
no available databases estimating regional and global levels of gillnet and trammel net
fishing effort. As a result of these deficiencies, there would be very high uncertainty
in estimates of regionally and globally averaged ghost fishing mortality rates and
levels, especially for taxa that are rare-event captures, such as marine megafauna. Four
priorities were identified to fill these key information deficits. Recommendations to
obtain robust estimates of regional and global rates and levels of ghost fishing from
ALDFG from marine gillnet and trammel net fisheries are, inter alia:
Domestic and regional authorities should harmonize logbook and observer
programme data fields, data collection protocols and database formats on ALDFG
from gillnet and trammel net fisheries where they are in place, and fill gaps in
ALDFG monitoring. Standardizing data fields, data collection protocols and
database formats facilitates comparisons between regions, enables the pooling of
data necessary to support large spatial scale analyses within and across regions,
and enables standardization of training materials and courses within and across
regions. This would also produce larger logbook and observer programme datasets
of records of rates of generating ALDFG from gillnet and trammel net fisheries,
and rates of fishing vessel encounters with ALDFG, with broader spatial coverage.
More research should be conducted, using best practice methods identified
through this study to minimize sources of uncertainty, to estimate ghost fishing
mortality rates and levels. These studies should be balanced spatially, temporally
and by type of gillnet and trammel net fishing gear and method.
Studies designed to estimate ghost fishing mortality rates and levels should
employ standardized units to report estimates in order to facilitate pooling.
Meta-analyses of data from relevant compiled studies should be conducted to
produce estimates of generating ALDFG, density of ALDFG, and ghost fishing
mortality rates with increased precision, accuracy and statistical power over
estimates from individual studies.
Robust estimates of regional and global gillnet and trammel net fishing effort
should be developed.
Despite increasing international recognition of the need for multilateral efforts to
address effectively the transboundary problems resulting from ALDFG, including
ghost fishing, there has been limited progress in international monitoring and
management of ALDFG by gillnet and trammel net fisheries. To benchmark regional
measures for monitoring and mitigating ALDFG and ghost fishing, an assessment
was made of the data collection protocols and management measures to prevent and
remediate ALDFG and ghost fishing of then regional bodies and arrangements with
the competence to establish binding controls for marine capture fisheries, and that have
competence over fishery resources captured in an active gillnet or trammel net fishery.
Of the ten RFB/As, three collect data via logbook or observer programmes
related to ALDFG from gillnet or trammel net fisheries. Harmonizing ALDFG data
collection protocols where they are in place, and filling gaps for those RFB/As lacking
procedures to collect this information, would contribute to improved monitoring
of ALDFG in regional marine capture fisheries. Only one of the assesse