General Enquiries on the form should be made to:
Defra, Procurements and Commercial Function (Evidence Procurement Team)
E-mail: [email protected]
Evidence Project Final Report

Note
In line with the Freedom of Information
Act 2000, Defra aims to place the results
of its completed research projects in the
public domain wherever possible.
The Evidence Project Final Report is
designed to capture the information on
the results and outputs of Defra-funded
research in a format that is easily
publishable through the Defra website
An Evidence Project Final Report must
be completed for all projects.

This form is in Word format and the
boxes may be expanded, as appropriate.

ACCESS TO INFORMATION
The information collected on this form will
be stored electronically and may be sent
to any part of Defra, or to individual
researchers or organisations outside
Defra for the purposes of reviewing the
project. Defra may also disclose the
information to any outside organisation
acting as an agent authorised by Defra to
process final research reports on its
behalf. Defra intends to publish this form
on its website, unless there are strong
reasons not to, which fully comply with
exemptions under the Environmental
Information Regulations or the Freedom
of Information Act 2000.
Defra may be required to release
information, including personal data and
commercial information, on request under
the Environmental Information
Regulations or the Freedom of
Information Act 2000. However, Defra will
not permit any unwarranted breach of
confidentiality or act in contravention of
its obligations under the Data Protection
Act 1998. Defra or its appointed agents
may use the name, address or other
details on your form to contact you in
connection with occasional customer
research aimed at improving the
processes through which Defra works
with its contractors.
Project identification
1.
Defra Project code
2.
Project title
MF1220
Developing the scientific basis for using real-time
closures as a fishery management measure
3.
Contractor
organisation(s)
Christopher Lynam
Cefas,
Lowestoft Laboratory
Suffolk
NR33 0HT
54. Total Defra project costs
(agreed fixed price)
5. Project:
EVID4 Evidence Project Final Report (Rev. 06/11) Page 1 of 27
£
75231.00
start date ................
01-11-2010
end date .................
30-01-2012
6. It is Defra’s intention to publish this form.
Please confirm your agreement to do so. ....................................................................................YES
X
NO
(a) When preparing Evidence Project Final Reports contractors should bear in mind that Defra intends that
they be made public. They should be written in a clear and concise manner and represent a full account
of the research project which someone not closely associated with the project can follow.
Defra recognises that in a small minority of cases there may be information, such as intellectual property
or commercially confidential data, used in or generated by the research project, which should not be
disclosed. In these cases, such information should be detailed in a separate annex (not to be published)
so that the Evidence Project Final Report can be placed in the public domain. Where it is impossible to
complete the Final Report without including references to any sensitive or confidential data, the
information should be included and section (b) completed. NB: only in exceptional circumstances will
Defra expect contractors to give a "No" answer.
In all cases, reasons for withholding information must be fully in line with exemptions under the
Environmental Information Regulations or the Freedom of Information Act 2000.
(b) If you have answered NO, please explain why the Final report should not be released into public domain
Executive Summary
7.
The executive summary must not exceed 2 sides in total of A4 and should be understandable to the
intelligent non-scientist. It should cover the main objectives, methods and findings of the research, together
with any other significant events and options for new work.
Real-time closures (RTCs) involve the closure of a relatively small area to fishing for a limited period in order
to provide a measure of protection for one or more fish stocks within that area. In contrast to traditional
closures that are fixed in space and time, RTCs are adaptive measures that attempt to track the spatial
distribution of fish, using real-time information from the fishery or other sources. RTCs have recently been
introduced as a cod recovery measure in response to suggestions from the fishing industry. As a result, the
implementation of RTCs was without detailed scientific analysis of those issues pertaining to the suitability of
methodologies to identify areas for closures or choose ‘trigger’ levels (i.e. thresholds past which action is
taken) for sampling to achieve a given effect. Although a number of member states have begun national
programmes of measures that include RTCs, no common system for RTCs within the EU has been established
(Bailey et al 2010). As part of the “Conservation Credits” scheme, and in response to Regulation 1342/2008,
Scottish RTCs were implemented in 2008 in the North Sea (ICES subarea IV) north of 56°N. In 2009, an English
scheme was implemented to the south of 56°N (including the eastern Channel, ICES division VIId) with
closures covering 64 nm2 for 1 month or 23 nm2 for 1 fortnight if in coastal waters. A notable difference in the
Scottish scheme is the use of buffer zones around the closures. A number of gears are prohibited in English
RTCs, south of 56°N, and seasonal closures for cod, they are as follows: demersal trawls including otter and
seines (TR1 with ≥ 100 mm mesh, TR2 70-99 mm, TR3 16 - 31 mm); beam trawls (BT1 with ≥ 120 mm mesh,
BT2 80-119 mm); long lines (LL1); gill nets (GN1) and trammel nets (GT1). In addition to the English and
Scottish RTCs, an EU RTC scheme in agreement with Norway is also in place with the aim to reduce discards of
juvenile whiting (<27 cm), cod(<35 cm), haddock (<30 cm) and saithe(<35 cm). The EU scheme was amended
on 01 August 2011 to reduce the trigger level and sampling levels for which the trigger must be assessed
(section 2.3.3). Also in September 2011, the Dutch agreed to participate in the English RTC scheme. As part of
this agreement, Dutch effort and landings data were included with English data during the analysis to identify
EVID4 Evidence Project Final Report (Rev. 06/11) Page 2 of 27
candidate areas for closure and this has altered the spatial distribution of the resultant candidate closures
based on historic landings per unit effort (LPUE) information (section 2.6).
This project collated extensive data on cod commercial catch (landings of cod from fishermen’s logbooks and
vessel tracks from the satellite-based Vessel Monitoring System VMS) alongside information on movements of
fish from tagged cod (see Deliverable 1). Statistical analyses were made to demonstrate the theoretical basis
for RTCs (see Deliverable 2). Detailed spatial and biological modelling was conducted in order to develop
approaches to select closures (Deliverables 3 and 4). Subsequent simulation studies have been used to
evaluate the current RTC schemes, suitable trigger levels and possible modifications (see below).
Information on the movement of cod from tagging studies was also used to address the following questions:
do cod reside in an area for any length of time such that spatial closures may be effective?; is there a pattern in
the migration of cod that can help to inform managers of the likely location of high catch rates of cod? Fishing
effort derived from the VMS was combined with cod landings information from logbooks and simulation
studies were used to investigate a number of questions, namely: what is the theoretical basis for RTCs?; what
impact might English RTCs have on the English trawl fleets?; what level of effort ‘buyback’ (additional time
allocated to comply with the scheme) for the TR1-2 fleets is appropriate?; what trigger levels are suitable for
trigger closures based on live sampling?; are current sampling levels suitable to detect catches over the trigger
level?; are closures based on historic LPUE information effective?; does inclusion of the historic Dutch data
improve the effectiveness of the English RTC scheme?
The fundamental reasoning behind the RTC scheme is sound and trigger levels at which closures should be
implemented based on real-time catches are suitable. However, the English RTC scheme is hampered by an
over-reliance on the largely ineffectual historic LPUE-based closures that were intended to complement the
real time scheme. Further work should focus on improving real-time information on cod catch-rates through
either improving monitoring of the fleets or stimulating communication with the fishermen. The potential for
fully documented catch-quota vessels using Remote Electronic Monitoring (REM) to inform on the current
abundance of cod in real-time should not be overlooked. Given the high movement rates of cod shown by
tagging studies (section 2.2), real-time closures are likely to become redundant within a month. If catch-quota
vessels were exempt from the RTC scheme then they would serve as useful agents to monitor the continued
relevance of closures.
Project Report to Defra
8.
As a guide this report should be no longer than 20 sides of A4. This report is to provide Defra with details of
the outputs of the research project for internal purposes; to meet the terms of the contract; and to allow Defra
to publish details of the outputs to meet Environmental Information Regulation or Freedom of Information
obligations. This short report to Defra does not preclude contractors from also seeking to publish a full,
formal scientific report/paper in an appropriate scientific or other journal/publication. Indeed, Defra actively
encourages such publications as part of the contract terms. The report to Defra should include:
 the objectives as set out in the contract;
 the extent to which the objectives set out in the contract have been met;
 details of methods used and the results obtained, including statistical analysis (if appropriate);
 a discussion of the results and their reliability;
 the main implications of the findings;
 possible future work; and
 any action resulting from the research (e.g. IP, Knowledge Exchange).
EVID4 Evidence Project Final Report (Rev. 06/11) Page 3 of 27
1 Introduction
The main focus of the project was to consider critically the recent use of RTCs in the context of the recovery of
North Sea cod. However, the theoretical basis developed is a general framework for which conclusions about
the implementation of RTC schemes can be drawn. The main operational requirements of an RTC scheme are to
be able to track a fishery spatially (i.e. using VMS), to identify areas with significant aggregations of the species
of interest (from the catch, see section 2.1) and the ability to close an area of a suitable size and for a suitable
duration in order to protect the aggregation. Both of these dimensions (size and duration) are rather difficult to
estimate in practice, but we have gained insight from data on fish movements from tagged cod (section 2.2). In
order to identify possible areas for closure, there are a number of sources of information which can be used,
namely:
1.
Information on current catch rates provided voluntarily by the fishing industry
2.
Information on current catch rates obtained by control authorities
3.
Information on recent or historic catch rates obtained through statutory reporting and monitoring
and/or scientific sampling
4.
Fishery-independent information on stock distribution and life history.
Of these data sources, 1 is potentially the most useful and might include additional information from catchquota vessels. Since the RTC scheme began in 2009, there have been no closures in the North Sea (south of 56
°N) or eastern Channel based on catch rates provided voluntarily by the English industry. Ideally, the industry
would report every time the catch rates of a particular species exceeded a specified trigger level. However, such
a scenario is unlikely for all vessels at present. Similarly, information from data source 2 can identify
aggregations in real time, but the low coverage means that not all aggregations can be identified and this limits
the likely impact of live ‘sampling’ (section 2.3.2). Such approaches based on real-time sampling typically
involve a specified minimum catch rate, which, if exceeded, will trigger a closure. We investigated this through
theoretical and statistical approaches (section 2.3.1-2.3.3 and Deliverables 2 and 4).
Currently, the English RTC scheme uses information on the earlier distribution of cod aggregations, i.e. landings
per unit effort (LPUE) information in the corresponding month in the previous two years, to predict future
aggregations and we have investigated the potential benefit of this approach through data analyses (section
2.5.2). The identification of areas holding concentrations of undersized fish poses particular problems. We use
the data from the Observer programme to attempt to identify areas where high rates of discarding occur
(section 2.1).
It is also desirable to be able to evaluate the effects of any closures. However, due to the nature of closed areas,
real-time data from those areas are lacking. To circumvent this issue we have used past data on cod landings
and effort with simulation studies to evaluate the potential impact of the English RTC scheme (section 2.4-2.5).
Unfortunately due to limitations of the available data we are not able to assess directly the potential impact on
discarding of the closures.
1.1 Objectives and Deliverables
The overarching aim of this project was to establish a strong scientific basis for the use of real-time closures as
a fishery management measure. In order to achieve the overall objective, the project had the following specific
objectives and associated outputs:
1, To compile and document data relevant to the study of real time closures. Output 1: VMS/logbook data,
EVID4 Evidence Project Final Report (Rev. 06/11) Page 4 of 27
discard data, enforcement sources, survey/tagging data were described in Report “Data Sources” (Deliverable 1,
submitted 01 January 2011)
2, To develop the theoretical basis for using real-time closures as a fishery management measure. Output 2
described in Report “Species Avoidance as a Fisheries Management Measure; A Theoretical Perspective”
(Deliverable 2, submitted 31 March 2011)
3, To quantify annual and seasonal variation in the distribution of cod catches Outputs described in two reports
“Results From Initial Data Analysis” (Deliverables 3, submitted 31 March 2011) and “Results From Detailed Data
Analysis” (Deliverable 4, submitted 31 August 2011)
4, To evaluate different approaches to selecting areas for closure based on past catch data Output 4 described in
Report submitted 31 August 2011 (Deliverable 4 and final report section 2.5)
5, To outline an adaptive approach to implementing real-time closures as a fishery management measure
described in this document
This final report will summarise the main findings as follows: 2.1 Spatial analysis of fishery data (landings and
discards); 2.2 Spatial analysis of cod tagging data; 2.3 Detection rate of high catch per unit effort (CPUE) values
and suitable trigger levels; 2.4 Estimated impact of RTCs in 2010; 2.5 Simulated maximum effect of RTC scheme
with 9 or 18 closures per month based on real-time LPUE of English fleets; 2.6 Evaluation of the effect of
including the Dutch LPUE information into the English RTC scheme; 2.7 Modifications to the scheme to improve
protection for cod; 2.8 Synthesis of findings
2. Methodology
2.1 Spatial analysis of fishery data (landings and discards of cod)
To provide a clear representation of where and when regulated vessels are fishing in the North Sea, south of
56°N we present maps of effort and landings data from VMS and logbook data (as combined by the Marine
Management Organisation) and discard/retained catch from the Observer programme.
For the period 2006-2011, the total effort, cod landings and overall average LPUE by the English fleets south of
56°N is presented in Fig. 1. The proportion of effort and landings by each gear category is also shown and this
illustrates the differing fishing grounds that the fleets occupy. For instance, the TR2 fleet clearly dominates
(>90%) the effort and cod landings in much of the Channel (east of 1°W and south of 50.5°N) and the remainder
is made up of BT2 effort. TR2 also dominate in the area off the NE coast of England to north of 54.5°N and west
of 1°W. While to the south and east of this ‘TR2 area’, the TR1 fleet dominates the effort and cod landings in a
diagonal strip from 54°N, 1°W to 56.5°N, 3°E. The small meshed beam trawl fleet (BT2) is the major fishery
along the European coastline in the southern and eastern North Sea and catches a great percentage of the cod
landed from these areas, except in areas where the gill-net fishery operates. However, the total cod landings by
English fleets from the southern and eastern North Sea are small. On average, the TR1 fleet lands the greatest
tonnage of cod in a month, while the TR2 fleet spends the most time fishing, largely for Nephrops per month.
Maps of monthly fishery effort (hours fished) and landings of cod for the major fleets combined (i.e. otter, beam
and gillnet fleets) in the period 2006-2011 are given in Deliverable 4 (Figs. 5.1.1 and 5.1.7). Similar monthly
plots by major gear category in 2010 (TR1, TR2, BT1, BT2, GN) and showing the RTC implemented are given in
Deliverable 4 (Effort is shown in Figs. 5.1.2-6 and tonnage landed in Figs. 5.1.8-12).
The gill-netters, followed by the long-liners, have the highest cod LPUE associated with their operations
(Deliverable 4 Fig. 5.1.14). However, for establishing candidate RTCs based on LPUEs these data are excluded.
Gill-netters are quota restricted and management by this measure is considered sufficient. Of the towed gears,
EVID4 Evidence Project Final Report (Rev. 06/11) Page 5 of 27
TR1 has by far the greatest LPUE, on average 5 to 10 times that of the TR2 and BT fleets. There is a well-defined
seasonal pattern in LPUE for the otter trawl gears (TR1-2) and for small meshed beam trawlers (BT2, see
Deliverable 4 Fig. 5.1.15). High LPUE scores are found in the TR1 and BT2 gears during the first quarter of the
year followed by a decrease in LPUE during April and May. For the TR1 gear, LPUE scores increase from July
and may be high in November and December. In contrast, the TR2 fleet typically has highest cod LPUE during
April-September and the gill-netters the highest LPUE in July-September. This difference in the two groups TR1
and BT2 versus TR2 and GN is most likely due to the change in cod distribution (section 2.2) and in fishing
opportunities during the summer months.
Total landings by sub-rectangle for the period 2006-2011 were linearly positively related (all p < 0.001,
Deliverable 4, Fig. 5.1.22) to hours fishing by sub-rectangle for gears TR1, TR2 and BT1 (percentage of
variability explained by linear model: 47%, 67% and 52% respectively). Although more variable (45%, 34%
and 18% respectively), these relationships held when landings of cod only were regressed on fishing effort
indicating that effort management has the potential to control the landings. Nevertheless, we should not forget
that high LPUE scores can result from sub-rectangles with very low effort and moderate catches of cod.
Sampling of catches during the Observer Programme (1994-2010) shows the seasonal change in fishing
patterns by otter trawlers by quarter: moving into the central North Sea during quarter 3 (July-September)
(Deliverable 4 Fig. 5.1.16). Descriptive linear regression models indicate that the biomass of adult (≥35 cm) or
juvenile cod discarded by the TR1 fleet is related linearly and positively to the total biomass of cod retained
(Deliverable 4 Table 5.1.1). No relationship between discarded biomass and effort was evident for the TR1 fleet.
However, for the TR2 and BT2 fleets fishing effort (hours fished) was significantly (p<0.05) but only weakly
positively related to discarding of adult cod. Similarly, effort was weakly related to the discarded biomass of
juveniles by BT2 in Q4 (7%, p<0.01). Not enough data were available to assess discarding patterns for BT1 or
TR3 vessels. Given the relationships between juvenile cod discards and retained biomass of cod in the main
fleet targeting cod, TR1, the closures based on historic LPUEs will go some way to predicting where juvenile cod
measures might be required, but real time sampling measures are particularly important to address this issue
fully. The statistical models developed here are not of sufficient quality to be used to predict accurately
where/when discarding is likely to occur. Nevertheless, the quarterly maps of discarded and retained cod by
gear group for pooled 1994-2010 data (Deliverable 4 Figs 5.1.16-19) do give an indication of the spatial
patterns in the data and show that discarding is possible in all areas where cod are fished. However, it is clear
that large discard rates of cod per haul have occurred off NE England where the majority of English landings,
south of 56°N, are reported.
Summary
Fleets clearly occupy distinct fishing grounds: the TR1 fleet lands the greatest tonnage of cod in a month, while
the TR2 fleet spends the most time fishing, largely for Nephrops per month. Relative to the TR2 and BT gear
categories, TR1 vessels have by far the greatest LPUE, on average 5 to 10 times. Given the relationships
between retained and discarded tonnages by TR1 vessels, closures based on historic LPUEs will go some way to
predicting where juvenile cod measures might be required. Nevertheless, real time sampling measures are
particularly important to address this issue fully. It is clear that large discard rates of cod per haul by English
fleets can occur off NE England where the majority of English landings, south of 56°N, are reported. Thus
increased levels of real-time information regarding catch-rates from this area would be beneficial.
EVID4 Evidence Project Final Report (Rev. 06/11) Page 6 of 27
Fig. 1
Summary maps (2006-2011), where top row: overall (TR1-3, GN1, BT1-2) cod landings by subrectangle followed by % all cod landed within a sub-rectangle due to 3 major gear category; middle
row: overall effort (TR1-3, GN1, BT1-2) followed by % of all effort expended within a sub-rectangle due
to major gear category; bottom row overall cod LPUE (tonnes per hour, TR1-2, BT1-2) followed by
LPUE by major gear category.
2.2 Spatial analysis of cod tagging data with estimates of movement rates
Information from data storage tags (DSTs, Deliverable 3) was input to a tidal-based geolocation model to
determine the best estimates of cod location (Pedersen et al., 2008). Analyses were extended to map the likely
location and movement between months using probability surfaces (Deliverable 4 section 5.2). Daily movement
distances and rates were also calculated to indicate the potential for cod to move out of implemented RTCs.
This information is then compared to the spatial distribution of catch rates determined from the observer
programme and to maps of landings, effort and LPUE from the fleets catching cod.
The maps of cod location (Deliverable 4 Fig 5.2.1.5-7) show clearly that cod can make long-range migrations,
for example cod released in the southern North Sea were found in the western Channel. Cod tagged in the
eastern Channel frequently move into the western Channel and to a more limited extent the southern North
Sea. Cod tagged in the southern North Sea show many movements occurring during the summer months (MayAugust) into the central North Sea (IVb) and eastern Channel (VIId). Although based on few data, cod tagged in
the central North Sea appear to belong to one of two separate stock units in the east or west and only small
movements, greater during the summer, are evident in the dataset. During the winter, September-December,
there are fewer tag returns available for the model. However, the data suggest that during this period the cod
exert a relatively stable pattern with aggregations in the eastern Channel, eastern-central North Sea (near
Jutland) and western-central North Sea (near Whitby). The ICES stock assessment of North Sea cod includes
both the eastern Channel and Skagerrak and it is likely that the western central North Sea unit move in and out
EVID4 Evidence Project Final Report (Rev. 06/11) Page 7 of 27
of the Skagerrak. An attempt was made to use the monthly difference in probability scores to forecast monthly
changes in LPUE in order to make projections for candidate RTCs based on data that was nearer in time to the
month in which RTC are to be implemented (Deliverable 4 section 5.2). However, the spatial pattern in fishing
effort and LPUE can change between months to a much greater extent than the change in the cod probability
surface. Therefore, the spatial variability in the seasonal pattern of LPUE is greater than that in the modelled
distribution of cod. Given this limitation, the tagging data were not able to improve predictions of spatial
patterns in LPUE of cod. A higher resolution in the cod distribution by the tagging studies could be attained if
the numbers of tags released and/or returned were increased.
Previous analyses of the behaviour of individual cod, based on the distances that cod travelled between
successive days, does contain useful information relating to RTCs (Righton et al 2008 and Deliverable 4 section
5.2.3). At the time of analysis, 263 cod (24%) had been recaptured, yielding over 31,000 days of data. Analyses
were restricted to tags from cod that were recaptured more than 3 months after release i.e. those that had
opportunity to migrate. On average, these cod were recaptured 96 km (± 127, n = 80, where the range is ±1
standard deviation) from their point of release (maximum = 680 km). Average time at liberty was 187 days (±
134 d, max = 908 d). Recapture positions of these cod are shown in Deliverable 4 Fig. 5.2.3.1. It was only
possible to reconstruct (in sufficient detail and with sufficient certainty of location), the daily movements of 35
(6935 locations) cod from the datasets available (Table 1). From these 35 fish, extended directed movement
(EDM, when cod accomplished a movement of >15 km within a day) was identified to varying degrees
dependent on the region (ICES sub-division) in which they were located (Table 1). When cod did not move
more than 2 km, their behaviour was classed as resident (RES). All other periods were defined as extended
localised movement (ELM) and were identified when cod moved > 2km but < 15 km.
Table 1 Summary statistics including: total number of days of data; average time at liberty; daily movement
rates and relative duration of each behavioural type. RES, resident; ELM, extended localised
movements; ELM, extended directed movements.
Region
(ICES
subdivision)
NNS (IVa)
CNS (IVb)
SNS (IVc)
CHA (VIId)
Number
of
Releases
447
191
303
158
Number
of
Recaptures
76
26
81
80
Total
data
(days)
8890
4638
8627
8865
Average
days at
liberty
177
187
153
249
Distance
moved
(km per day)
6±6
8.5 ± 14.9
10.6 ± 13.3
16.4 ± 13.4
RES
%
days
93%
71%
22%
37%
ELM
%
days
5%
26%
64%
57%
EDM
%
days
2%
3%
15%
6%
On average, cod moved a distance of 12.5 (± 12.4) km per day indicating that in less than two days of sustained
travelling, in one direction, cod could cross an entire RTC closure of 64 nmi2 (approx 15 by 15 km). However,
there was considerable variation in behaviour between individuals and regions. Many individuals undertook
little to no movement, while other individuals were highly vagile, sometimes moving up to 70 km in a single
day. Cod tagged in the southern North Sea and Channel exhibited the greatest daily rate of movement, while cod
in the northern North Sea exhibited the lowest values. Cod in the northern North Sea and central North Sea
generally spent the majority of their time in the area close to their release position (Deliverable 4 Fig. 5.2.3.1ab), with low rates of movement throughout the year. In contrast, cod in the southern North Sea moved rapidly
in spring from presumed spawning areas in the southern and eastern North Sea to feeding grounds in the
central and western North Sea, including the Nephrops grounds at ‘Botney Gut’ (54°N, 2°E). Cod tagged at their
presumed spawning grounds in the eastern Channel either moved to feeding grounds in the western Channel
during spring, or moved locally in the eastern Channel throughout the year.
With the exception of the cod tagged in the northern North Sea, where resident behaviour occurred almost
EVID4 Evidence Project Final Report (Rev. 06/11) Page 8 of 27
exclusively, cod elsewhere appeared to show resident behaviour in particular seasons (Deliverable 4 Fig.
5.2.3.2). Resident behaviour occurred in the first half of the year in the southern and south eastern North Sea. In
the English Channel, resident behaviour tended to occur in deeper water during the latter half of the year.
Extended directed movement occurred predominantly in the southern North Sea and English Channel and was
associated with offshore channels. Extended localised movement (ELM) occurred throughout the utilised area,
and was more prevalent in winter and spring in the northern North Sea and eastern Channel. Cod in the central
North Sea spent a high proportion of their time making ELM during the feeding period, April-September, which
is when fishermen target them here.
Summary
Individual cod have the potential to move over large distances rapidly and across a typical RTC in as little as a
couple of days, if they chose to move a sustained direction, as they might during spawning and feeding
migrations. Nevertheless, in the central North Sea (IVb), where the majority of landings from the English fleets
are caught, 71% of daily movements were found to be <2 km indicating that small closures have the potential to
protect a substantial proportion of cod. Seasonally, a clear picture emerges of movement between spawning
and feeding grounds, such that the location of RTCs will be stable during the winter months. Thus, month-long
closures would be most effective when cod are at their spawning grounds and shorter period may be more
suitable during migration periods. For RTCs to be effective during the summer, closures will need to be
adaptive and real-time sampling data is crucial. On the basis of this study, we would suggest that sampling
efforts be increased during the summer.
2.3 Detection rate of high CPUEs and suitable trigger levels
2.3.1 Theoretical approach
Based on theoretical principles, RTCs based on live sampling were shown (Deliverable 2) to potentially reduce
the catchability of cod and thus fishing mortality, assuming no change in total effort. A range of reductions in
fishing mortality, of between 2 and 38% were estimated (Table 2) based on differing assumptions for the
parameters of catch rate distribution, detection rate, trigger level for closures and effort reallocation.
Three gamma distributions (A, B, and C, Fig. 2) for catch rates were modelled based on observations of catch
(retained and discards component) from the Observer Programme and also the Fisheries Science Partnership
‘Codwatch’ project (see Deliverable 1). Distribution A was based on the distribution of cod catch-rates in hauls
sampled in TR1 gears during onboard discard sampling by Cefas scientific observers over 2002-2010.
Distribution, B was based on the observed distribution of cod catch-rates in hauls sampled during the
‘Codwatch’ project (2007-2009). An alternative third distribution, C, was considered as an example distribution
for a fishery with a more even catch rate distribution than either distribution A or B.
The maximum effect of a 38% decrease in catchability was based on a scheme that excluded the highest 10% of
catch rates, from an assumed highly-skewed distribution (distribution A), with 100% detection of high catch
rates and random redistribution of the affected effort. If the detection rate were lowered to 25% then the
maximum effect of such a scheme would lower similarly and equate to a 9% drop in fishing mortality assuming
random reallocation of fishing effort. However, if the vessels were assumed to move to areas with high LPUE
without loss of effort then the maximum effect would fall to a 5% drop in fishing mortality.
Summary
RTCs based on live sampling can potentially reduce the catchability of a cod by a fishery and thus the fishing
mortality. However, the importance of a high detection rate is clear, and for cod RTC schemes to be effective
EVID4 Evidence Project Final Report (Rev. 06/11) Page 9 of 27
they require great levels of engagement and information from commercial fishermen and control authorities.
Table 2 Relative catchability achieved by a theoretical cod avoidance scheme based on avoiding the upper 10%
of LPUE. The results are presented for 3 distributions of catch rate (A-C), 4 detection rates for hauls
above the trigger level, and 3 scenarios for the effort reallocation
Distribution A
‘Observer
programme’
Distribution B
‘Codwatch
programme’
Distribution C
‘Less skewed
distribution’
Fig. 2
Detection rate for
hauls above
trigger level
100%
75%
50%
25%
100%
75%
50%
25%
100%
75%
50%
25%
Relative catchability by effort reallocation method
Biased to upper
Biased to upper
100% at random
50% of LPUE
25% of LPUE
0.62
0.76
0.80
0.72
0.82
0.85
0.81
0.88
0.90
0.91
0.94
0.95
0.76
0.84
0.88
0.82
0.88
0.91
0.88
0.92
0.94
0.94
0.96
0.97
0.83
0.88
0.91
0.87
0.91
0.93
0.91
0.94
0.96
0.96
0.97
0.98
The 4 example gamma distributions (A-C, see text) used to represent relative frequency of catch rates
in the model runs.
2.3.2 English RTCs for spawning and juvenile cod
The current North Sea RTC live sampling rule (i.e. sampling at sea by enforcement officers) for English vessels
has two parts:
1.
Spawning cod measures between 1 January to 30 April 2011 where a closure is triggered if 10 or more
mature cod (>50 cm) are caught per hour.
2.
Juvenile cod measures between 1 May to 31 December 2011 to avoid juvenile discarding where a
closure is triggered if 80 or more cod of all sizes are caught per hour.
In this section we analyse data on boardings by control authorities and logbook-VMS from the MMO along with
Observer Programme data to investigate the values of current detection rate in English RTC scheme and likely
effect of the trigger level on catch rates. Both detection rate and trigger level were shown to be important
parameters in the theoretical approach (section 2.3.1 above).
EVID4 Evidence Project Final Report (Rev. 06/11) Page 10 of 27
Catch rates and trigger levels
To assess the coverage of the Observer data and provide context for the enforcement ‘boardings’ data, the
number of fishing trips for vessels >10 m and the number of hauls were extracted from fisheries databases. A
list of boardings from January 2009 to mid-2011 was provided by the MMO. To assess the catch rate
distribution, cod catch data recorded at the level of individual hauls were taken from the Observer programme.
Records with suitable effort data were selected for ICES divisions IVb, IVc and VIId, i.e. the central and southern
North Sea and eastern English Channel, to match the RTC scheme and all years from 1994 to 2010 were used to
reflect a range of stock sizes.
The distribution of cod biomass by haul in the Observer data analysed was highly skewed (as shown in
Distribution A section 2.3.1) with the top 10% of hauls (those with catch ≥253.6 kg) accounting for 57.5% of
total cod biomass caught (156.5 out of 272.3 tonnes). This pattern was repeated within each month and ICES
division. The distribution of catch rates in numbers per hour was also skewed, a pattern seen for each gear
category in both the spawning measures and juvenile measures periods of the year (Deliverable 4 Figs. 6.2.1
and 6.2.2). The current trigger levels are in the tail of the catch distributions, as required to prevent the
majority of hauls triggering closures. If all gears and both periods are combined then the trigger level of ≥ 80
cod per hour represents the top 10.5% of hauls, indicating these data are consistent with the process used to
set the trigger level.
To assess the impact of a specific trigger level, the next step is to calculate the amount of cod biomass from
hauls with catch rates above the trigger level. The current mature cod trigger level of 10 fish per hour was
exceeded in 8.6% of hauls that contained cod in the Observer data, these hauls accounted for 10 tonnes of
biomass, which was 43.8% of the total. Similarly the ‘all cod’ trigger level of 80 fish per hour was exceeded in
13.8% of hauls, accounting for 128 tonnes, or 59% of biomass contained within the hauls. Therefore, the
current trigger levels have the potential to have a significant impact on fishing mortality, depending on
detection rates and effort reallocation.
Further examination of the data by gear (Deliverable 4 Table 6.2.4) shows that the majority of the cod biomass
in the Observer data was from the larger meshed otter trawl gear category, TR1, although this majority was
larger for the ‘all cod' biomass than the mature cod biomass (TR1: 87% of all cod biomass, May-Dec, TR1: 65%
of mature cod biomass, Jan-Apr). Consequentially, 96% of the biomass above the 80 cod per hour trigger rule
would come from gear TR1, with only small changes in the percentage for other trigger levels (Deliverable 4
Table 6.2.5).
Boardings and detection rates
The North Sea RTC live sampling is based on sampling at sea by enforcement officers, therefore to investigate
inspection rates we analyse data on the number of inspection boardings by the Royal Navy (HMS) and Other
Member States’ (OMS) enforcement officers.
A dataset of boardings records for January 2009-July 2011 was provided by the MMO. From this dataset,
boardings in ICES divisions IVb, IVc, VIId were selected. The majority of the 1089 records (89%) were from
Royal Navy enforcement officers, with 501 boardings of UK vessels and most of the remaining 588 boardings
being of Belgian (230), French (161) and Dutch (123) vessels. Boardings of non-UK vessels could trigger
closures, but the focus here is on UK vessels as matching effort data are available.
The number of inspections per month was broadly consistent throughout 2009 and 2010, with some increase
in the first part of 2011 (Deliverable 4 Table 6.2.6). The boardings of UK vessels covered a range of vessel
types, both those permitted and not permitted within RTCs (Deliverable 4 Table 6.2.7). Of those not permitted
EVID4 Evidence Project Final Report (Rev. 06/11) Page 11 of 27
within RTCs the most commonly boarded vessel types were demersal stern trawler, beam trawler, gill netter
and stern trawler (pelagic/demersal).
For estimating detection rates the boardings dataset was split into the seasons when the spawning and juvenile
measures of the RTC scheme apply. The dataset was narrowed down to gears not permitted in RTCs and to
years with complete coverage: 2009 and 2010. The location of the boardings (Deliverable 4 Fig. 6.2.3) reflected
the distribution of fishing effort with the most boardings occurring in the waters close to the ports in the
northeast and southeast of England and during the second part of the year on the fishing grounds in the central
North Sea.
Some caution needs to be applied to linking boardings with effort data to estimate detection rates since the
National Control Action Programme states that “British fishery limits encompass not only the cod management
zones but also areas designated under the multi-annual plans for western Channel sole, northern hake and North
Sea sole and plaice. The UK undertakes risk-based, intelligence-led enforcement in relation to all these plans and
patrols are deployed between these areas accordingly. Therefore it is not possible separately to identify the
number of days patrolling the cod recovery zone. Neither is it possible to separately identify the total number of
patrol days spent on cod monitoring activity.” As a result specific areas at times may have high detection rates if
they relate to enforcement strategy. We cannot represent the intelligence-led enforcement scheme accurately
and this caveat should be kept in mind when interpreting the results that follow since they are based on
estimates of average sampling rates by ICES division and closure season.
Enforcement officers generally sample one haul per boarding, so the proportion of trips sampled (pUK) and
proportion of hauls sampled (qUK) can give a guide to the probability of a haul with a high catch rate being
sampled (Deliverable 4 section 6.2.5). The final estimate of the probability of detection will be dependent on the
sampling strategy assumed. Using the proportion of hauls sampled, qUK, provides an estimate of the probability
of a haul with a high catch rate being sampled assuming the sampled haul is selected randomly from the hauls
on a trip. While using the proportion of UK trips sampled, pUK, results in a much higher estimate of the detection
rate, since here we assume that all hauls on a trip were sampled: this might be a more suitable estimate if
boardings were targeted on those parts of a trip where high catch rates of cod are found.
The number of closures generated by boarding UK vessels (c UK) can be estimated if we make the strong
assumption that the sampling estimates (pUK and qUK) can be applied to the proportion of hauls with cod catch
and the proportion of these ‘cod hauls’ that were above trigger levels (Deliverable 4 Tables 6.2.2 and 6.2.3).
Summing the estimated numbers of closures across the two seasons and three ICES divisions gives an
estimated 3.2 closures across 2009 and 2010, based on live sampling of UK vessels with current trigger levels
(Deliverable 4 Table 6.2.8). However this estimate would rise to an optimistic 24, if the sampling that occurred
was equivalent to inspecting all hauls on a sampled trip. These values should be considered guideline estimates
only given the caveats and assumptions made. Over 2009 and 2010, a closure was most likely in division IVb
during May-December, which is the period of juvenile cod measures.
Summary
The range of detection rates as estimated above (3.2 to 24 over 2 years) for the current English RTC scheme
based on live-sampling are well below the levels required to generate the 9 closures per month expected by the
scheme. If possible, sampling levels must be increased, either by a greater number of boardings particularly of
TR1 vessels (which we consider unlikely) or by greater engagement in the scheme by fishermen volunteering
information to the MMO regarding catch rates. A potential means for the communication of such information
could be via the footage received from catch-quota vessels.
EVID4 Evidence Project Final Report (Rev. 06/11) Page 12 of 27
2.3.3
Likely impact of the amendment to the EU-Norway regulations concerning the juvenile RTC Scheme
The regulations concerning the EU RTC scheme for juvenile RTCs were altered in August 2011 such that the
following sampling and trigger levels changed:

Reduce the size of haul needed to commence a sample from 300 kg combined of the four species, cod,
haddock, whiting and saithe to 200 kg

Reduce the threshold for a closure from the presence by weight of 15% of the four species that are
undersize fish to 10% undersize.

Reduce the threshold for a closure, in the case that >75% of biomass is cod, from the presence by
weight of 10% of the four species that are undersize fish to 7.5% undersize.
Juveniles by species are classified based on their length as follows: whiting <27 cm, cod<35 cm, haddock <30
cm and saithe<35 cm. To investigate the likely effect that this change in the regulation would have we require
data on the entire catch (not simply the landed component of the catch). Therefore, to address this question we
focus on the Observer data, which includes estimates of the discarded and retained proportion of the catch.
Sampling levels in the Observer programme are designed to capture the fleets with a level of effort proportional
to the effort by the fishery in the previous year and as such the number of trips and thus hauls observed differ
between fleets and areas (Table 3a). Sampling in the northern (IVa) and central (IVb) North Sea is dominated
by otter trawlers with mesh ≥100 mm (TR1), In the central (IVb) and southern (IVc) North Sea smaller meshed
beam trawlers (80-119 mm, BT2) and otter trawlers (70-99 mm, TR2) are also well sampled. Gillnets (GN) are
well sampled in the southern North Sea effort. Sampling in the eastern Channel (VIId) is dominated by beam
trawlers (BT2). Large meshed beam trawlers (mesh ≥120 mm BT1), longliners (LL) and very small meshed
otter trawlers (TR2, 16-31 mm) are not well sampled.
The quantity of target species caught in each haul is typically <100 kg and the proposed change from 300 kg to
200 kg, in minimum catch of the target species for which sampling will occur during a boarding, will increase
the probability of a haul being sampled during boardings from 31% to 40%.
Hauls by TR1, TR2, BT2 and GN typically (≥70%) contain some (nonzero) of the target species (cod, haddock,
whiting or saithe) in the catch (Table 3b). Similarly, juveniles of the target species are typically caught by: TR1
gears in the central and southern North Sea; TR2 in the northern and central North Sea; BT2 in the southern
North Sea (Table 3c). Of these hauls with juveniles, recorded in the Observer programme, cod comprise ≥75%
by weight in the majority of TR1 hauls in the southern North Sea (78%) and eastern Channel (83%) and for gill
nets in the central North Sea (72%, Table 3d).
If the trigger level alone were increased to 10%, the overall effect (data for all years, gears and areas pooled)
would be to increase the proportion of hauls over the trigger level by 3% i.e. from 11% to 14% (Table 3).
However when coupled with the proposed shift in the sampling threshold to 200 kg, the proportion of hauls
over the trigger level would increase by 4% i.e. from 15% to 19% (bottom right). Given that the Observer
programme samples the fleet in proportion to the effort expended by the fleets in the area, these estimates
should be robust overall predictions of the likely effect of the proposed change in the regulation for juvenile
real time closures. Nevertheless, in terms of the proportion of hauls over the trigger level, these estimates do
mask the differences between fleets and areas. Gear-category specific calculations indicate that any increase in
the rate at which live-samples are found to be over the trigger level will most likely be due to hauls by otter
trawlers (TR1 and TR2) in the central (IVb) North Sea and to a lesser extent to TR2 hauls in the northern North
Sea (IVa) (Table 4). In VIId and IVc and also for other gears (BT2 and GN) there would most likely be no effect
EVID4 Evidence Project Final Report (Rev. 06/11) Page 13 of 27
since our Observer data indicates that for these gear/area combinations the trigger is unlikely to be met under
the proposed regulations. There is not enough data for BT1, TR3 or LL to say what the effect would be but it is
likely negligible or small.
It is important to note that this analysis is based only on English Observer Programme data that covers a small
percentage of the total UK TR trips that have taken place (and therefore decreases as the focus moves
northward). For example, as a percentage of the total UK otter trawler (TR1-3) trips taken place in 2009 and
2010 the English observations cover: 1% in VIId, 0.4% IVc, 0.3% IVb and only 0.1% in IVa. Given that the
Scottish fleet may not behave in the same way as the English fleet, the results above may not accurately reflect
the entire UK fleet. Also the results do not take into account any biases due to some areas/gears being targeted
more often by HMS/OMS (section 2.3.2) if indeed that is the case.
Summary
The strengthening of the trigger level and reduction in the sampling level at which the trigger is assessed will
increase the frequency at which TR1-2 vessels are likely to have catch-rates over the trigger, particularly in the
central North Sea and to a lesser extent English TR2 vessels in the northern North Sea Table 4). However, there
is likely to be no significant effect on vessels fishing for cod with towed gears in IVc and VIId or in any area with
gears BT1, BT2 and GN. The potential total number of closures will be limited still by low detection rates due to
limited opportunities for sampling by control authorities (as indicated in section 2.3.2).
Table 3a Number (N) of hauls sampled in the observer programme 1996-2010
N
BT1
BT2
GN1
TR1
TR2
All gears
IVa
0
74
0
274
68
416
IVb
69
449
41
1000
796
2355
IVc
0
140
262
51
289
742
VIId
0
397
50
6
76
529
All areas
69
1060
353
1331
1229
Table 3b Mean yearly % of hauls with one of the four target species (cod, haddock, whiting or saithe) present
in the catch. Values ≥70% bold if ≥50 hauls observed
%
BT1
IVa
IVb
BT2
GN1
94
46
83
95
TR1
TR2
100
99
99
98
IVc
91
67
95
78
VIId
61
81
100
69
Table 3c Mean yearly % hauls with juveniles of target species (see text). Bold if ≥70% and ≥50 hauls observed
%
BT1
IVa
IVb
BT2
GN
60
23
TR1
TR2
52
96
53
47
97
97
IVc
70
45
87
68
VIId
22
69
17
55
Table 3d Mean yearly % of hauls with a target species in the catch, where 75% by weight of catch of target
species is cod. Bold if ≥70% and ≥50 hauls observed
%
BT1
BT2
GN1
EVID4 Evidence Project Final Report (Rev. 06/11) Page 14 of 27
TR1
TR2
IVa
0
29
0
20
20
IVb
4
33
72
24
1
IVc
0
49
41
78
16
VIId
0
30
36
83
15
Table 4 Potential effects on each English fleet targeting cod in IV and VIId based on observer data 1996-2010
(all data pooled by gear category and ICES division) Note: no effect in the areas IVc and VIId. No effect
in any area for Gears BT1, BT2 and GN. Also this table shows minor differences from that in Deliverable 4
as the table presented here includes the new sub-trigger level (7.5% juvenile) if >75% of biomass is cod
ICES Division
IVa
IVb
Gear cat
TR1
TR2
TR1
TR2
% of hauls over the
trigger sampling level at
300 kg trigger 15% juvenile,
or
10% juvenile if >75% cod
1%
12 %
24 %
18 %
% of hauls over the
trigger sampling level at
200 kg trigger 10% juvenile,
or
7.5% juvenile if >75% cod
1%
21 %
42 %
35 %
increase
9%
18 %
17 %
2. 4. Estimated impact of the 2010 RTCs
Advice on real-time closures (RTC) for the cod recovery scheme has been provided since 2009. At the start of a
given month closures are set based on previous catches and then supplemented by closures based on real-time
information. Closures based on historic landing per unit of effort (LPUE) information have been identified in a
two-step process, quantitatively using data from the vessel monitoring system combined with landings data,
and then qualitatively with some input from industry. The first step in identifying the potential sub-rectangles
(1/16 of an ICES statistical rectangle) for closure is based on assigning the five highest ranking sub-rectangles
in each RTC area using landings per unit of effort from all towed gears in the previous two years. Of these, three
were selected qualitatively in each area.
We attempt to evaluate the RTCs effectiveness in 2010 in reducing landings of cod by UK vessels south of 56 °N.
To do so precisely, we would require an estimate of the landings foregone from the closed areas. Since this data
is lacking, historic data for the years 2008 and 2009 were considered and landings and effort for the areas and
months in which closures were made in 2010 were taken as best estimates of the missing information (Fig. 3).
These data were extracted from the vessel monitoring system and log book databases using an algorithm
developed by Cefas to produce fishing activity and effort information to a fine scale resolution, for details of the
methods used see Lee et al. 2010. Combined with these data landings of cod for each trip, area, vessel and gear
are then split across the trip where a vessel record is identified as fishing, each trip will have multiple records
giving the vessel’s location and speed. A value for LPUE (kg/hr) is then calculated for each record.
Seasonal closures during the months of February to April and RTCs in 2010 for each month were also applied to
2009’s fine scale resolution data. Vessels in December 2009 and January 2010 that were recorded as being
present in the seasonal closure areas were subsequently found in those areas in May 2010 once the seasonal
closure was lifted. Similarly, the data also show vessel compliance with the RTC scheme with vessels recorded
in areas prior to the real-time closures being implemented returning again once re-opened.
Effort for each gear in the previous years (2008 and 2009), within the area of the 2010 closures, was
redistributed to other areas of activity by the gear on each trip. The redistribution of effort was allocated in
proportion to the effort the vessel spent in each sub-rectangle prior to the closure. Gear was taken into account
as some vessels operated multiple gears and each gear type displayed slightly different spatial patterns.
EVID4 Evidence Project Final Report (Rev. 06/11) Page 15 of 27
Following this assumed redistribution of effort, landings of cod would have been reduced in the majority of
months for those vessels that fished in the areas closed in 2010 resulting in a simulated average decline in
landings of 10 to 17% for the vessels affected (Table 5). However, in some months the effect would have been
opposite to the intention with potential landings estimated to increase. At the fleet level, the estimated change
in total cod landings is small and in some cases negligible (Table 6). In addition, no account was taken of the
additional time allocated (‘buyback’) for otter trawlers (TR1-2) to comply with the scheme. Thus, this
conclusion is similar to that found by Needle and Caterino (2011) who suggested that the Scottish RTCs in 2009
did reduce overall cod mortality while they were closed, but that they may not have had any longer-lasting
effect on cod exploitation patterns.
There are a number of assumptions when applying this method, one being that the spatial pattern and
distribution of cod and effort remains constant across years. Also the selection of a fixed number of closures per
area does not take full account of the migration patterns of cod and effort, showing in some months that
closures are applied where there are very little landings of cod (Fig. 3). To address this problem, the process for
selecting the closures was developed and from June 2011 a new method was introduced, in which the number
of candidate sub-rectangles to be chosen in each sector (IVb_E, IVb_W, IVc, VIId) is apportioned based on the
proportion of average landings (all gears and TR1 and 2 gears) over the last three years in each sector and we
include this approach in further simulation work below (section 2.5).
Summary
Substantial reductions in landings of cod are suggested for individual vessels displaced by real-time closures. At
the fleet level however, the estimated change in total cod landings due to RTCs in 2010 is small and in some
cases negligible.
Table 5
Effect on landings of cod by vessels with effort reallocated only, when applying 2010 RTCs
retrospectively. Values in red show increases in landings of cod after displacement
January
February
March
April
May
June
July
August
September
October
November
December
Total % change
% change in 2008
-19
-25
-10
25
-45
-27
-13
-14
-9
-1
17
-18
-17
% change in 2009
92
17
-66
-14
-0.5
37
32
-16
-13
-31
15
-49
-10
Table 6 Effect (%) on total landings for all towed gears after redistribution of effort when applying 2010 RTCs
retrospectively. Values in red show increases in landings of cod after displacement
January
February
March
April
May
June
July
August
September
% change in 2008
-0.5
-0.8
-0.1
0.7
-4.1
-1.9
-1.3
-0.8
-1.3
EVID4 Evidence Project Final Report (Rev. 06/11) Page 16 of 27
% change in 2009
0.4
0.3
-2.8
-0.1
0
0
1.4
-1.3
-2.4
October
November
December
Total % change
Fig. 3
-0.1
0.5
-1.0
-1.0
-3.0
0.5
-2.5
-0.6
Real time closures (squares) and spawning closures (circles) in 2010 and estimated fishing location of
vessels with associated cod landings in 2008 (green) and 2009 (orange) by month (January top left,
March top right, October bottom left, December bottom right).
EVID4 Evidence Project Final Report (Rev. 06/11) Page 17 of 27
2. 5 Simulated effect of RTC scheme with 9 or 18 closures per month, reallocation and buyback of effort
2.5.1 Effects based on real-time catch rates of English fleets
As described previously, it is notoriously difficult to identify the actual effect of closures on a fishery due to the
lack of data from the closed areas. Here, we use monthly VMS-logbook data from the MMO collected previously
for the English fleets from Jan 2006 to Jun 2011 to simulate the potential effect of closing sub-rectangles in realtime. The maximum effect that the English RTC scheme could have is determined simply by removing
altogether the effort and landings recorded from the simulated closed areas. Such a reduction of effort is, of
course, not likely to happen in reality, but it does provide an estimate of the level of effort that would require
reallocation by the fishermen due to the scheme and give an upper limit to the potential cod catch that might be
protected by the scheme. To mitigate the impact on their catches fishermen will have many options to consider.
Should they simply fish nearby to the closure or should they move to areas further away where higher catches
might be found? To simulate these options, we explored the following four mitigation scenarios:
M1 The effort displaced by closures is evenly distributed among all areas fished by the fleet with that gear
category in that month
M2 Within the ICES rectangle containing the closure, the effort displaced is reallocated to an alternative
sub-rectangle with highest fishing effort
M3 Within the ICES rectangle containing the closure, the effort displaced is reallocated to an alternative
sub-rectangle with above average effort by the fleet and high landings of cod reported
M4 Within the ICES rectangle containing the closure, the effort displaced is reallocated to an alternative
sub-rectangle with above average effort by the fleet and high landings per unit effort of cod
The effect of each mitigation scenario M1-4 above is to reduce the impact of the closures on the fleet. Scenario
M1 is the weakest option for the fishermen (least benefit derived by the action), while scenario M4 is the
strongest (most benefit). We suggest that the most likely scenario is that fishermen will fish nearby in areas
where others in the fleet are fishing (M2) or where cod are caught and landed in high abundance (M3).
Therefore, we also average the results of the simulations based on M2 and M3 and consider this the most likely
model reality. Although scenario M1 may allow some effort to be reallocated over large distances, the
simulation does not rely upon strong and potentially poor assumptions about how fishermen behave.
Mitigation scenario M4 is particularly unlikely and certainly would not apply in every instance of reallocation of
effort. Although M4 may result in an overly positive scenario for alleviating the impact of the scheme on
fishermen we investigate this option to determine the lower bound of the scheme’s effectiveness to contrast to
the upper bound resulting from the complete removal of effort. No allowance is made for effort lost in practice
due to steaming to an alternative sub-rectangle. However, in M2-3 all reallocations of effort are within an ICES
rectangle and therefore lost time due to steaming is minimal. Nevertheless, we note that there will be some fuel
costs associated with additional steaming for closures that are based on sampling in real-time.
If the RTC scheme were based solely on English data and 9 closures per month were distributed across the
North Sea (south of 56N) and eastern Channel in proportion to the landings of the otter trawl fleets (TR1-2), on
average 3.8% (Table 7) of the annual effort by beam and otter trawl fleets combined would be affected by the
EVID4 Evidence Project Final Report (Rev. 06/11) Page 18 of 27
closures. If the number of closures were increased to 18 per month, on average 7.8% (range, 5.5% to 10.2%) of
towed gear’s annual effort would require reallocating. However, TR1-2 groups have been able to ‘buyback’
some effort from the cod recovery plan (i.e. they are given more effort than they would otherwise be allowed)
for adhering to this particular conservation measure: in 2011 this ‘buyback’ amounted to 11.25% of the annual
effort of these groups. When we consider the likely impact of the RTC scheme in terms of effort that would be
reallocated by gear (Table 7), the effort buyback has been higher than the level of effort likely displaced by 9
closures per month as are typically implemented. For this situation to be effective, the RTC must move vessels
from areas of high cod densities, particularly of undersized fish, to areas of lower abundance and the difference
must be sufficient to result in a reduction in catch despite the increased effort.
Assuming 9 closures per month (the typical number implemented currently), 16.4% of the potential annual
landings of cod prior to the reallocation of effort and 3.2% of potential total annual landings (all species) would
be protected within the closed areas from vessels with towed gears. These estimates vary greatly by gear group
as would be expected given the differing cod LPUE by gear (Deliverable 4 Figs. 5.1.15-15) with the greatest
impact on TR1 followed by TR2 and BT2 gears (Table 7). Similarly, the scheme is likely to be most effective in
the western central North Sea (IVb, west of 2 degrees longitude east) and least effective in the eastern central
North Sea (Table 7). The scheme is only likely to impact greatly on fishing in the southern North Sea and
eastern Channel during the winter period, when cod are at their spawning grounds (section 2.2).
Including any of the scenarios for the reallocation of effort displaced by closures in the simulation clearly
reduces the potential effectiveness of the scheme and the effect on total landings becomes negligible (Table 8).
By spreading the displaced effort throughout the fleet, simulations suggest that the potential reduction in
landings of cod would be 11.9% of the typical annual landings. If fishermen always knew where best to target
for high catch rates of cod (option M4) the RTC scheme would likely result in a smaller reduction in annual
landings (5.7%, Table 8). If we raise the landings for these fleets in proportion to the effort increase allowed
through ‘buyback’ we can see that there is potential to remove the impact on cod landings and, dependent on
the mitigation option, potentially reverse the effect. Although we cannot say what the actual effect on the total
catch (including discards) might be, we argue in section 2.1 (Deliverable 4) that discards by otter trawlers
(particularly TR1 vessels) are linearly related to landings, therefore reducing the expected landed catch should
also reduce the expected discards. Reducing the effort buyback to 5% from the current 11.25% for TR12 would
likely result in a decrease in the landings of cod and thus the total catch, irregardless of the mitigation option
taken by the fishermen.
Increasing the number of closures from 9 to 18 would result in some interesting consequences since the effect
is to magnify the results above. The simulation suggests that, prior to the reallocation of effort, 25% of potential
annual landings of cod and 6.6% of potential total landings would be protected within the closed areas. The
percentage decrease in cod landings would be expected to be 16.9% of annual landings if fishermen simply
spread their effort over the area fished by the respective gears. However, if all vessels moved to where they
would see the highest LPUE of cod (assuming perfect knowledge) the decrease would only be 3.4% and this is a
smaller reduction than the value obtained from simulations with 9 closures (i.e. 5.7%). There are two
explanations for this effect: 1. the mitigation strategy assumes that all vessels displaced are targeting cod, which
may not always be the case, 2. the raising of the landings in the target sub-rectangle linearly and in direct
proportion by the increase in effort will not always be correct, particularly when the high catch rate in the
target area is based on a large outlying point or on a small effort value as is often the case. Therefore, the results
from options M3-4 are likely overly optimistic in terms of simulated potential landings by the fishermen. As we
EVID4 Evidence Project Final Report (Rev. 06/11) Page 19 of 27
move across the buyback scenarios in Table 9 the effect of each scenario (5% or 11.25% increase in effort for
TR1-2) relative to zero buyback is very similar to that found in Table 8 for nine closures: i.e. the % decreases in
cod landings are 3.3% smaller with a 5% effort buyback and 8.3 % smaller with a 11.25% buyback (which then
results in a potential increase under either option M3 or M4).
Summary
The English RTC can potential deliver significant reductions in mortality. However, any reduction in landings
and thus the likely decrease in discarding can be offset by the behaviour of fishermen. If the buyback level is set
to high the gain from the RTC scheme can be reversed. The findings in this section assume all monthly closures
were identified in real-time. However, at present the majority of closures are based on historic LPUE
information and therefore the scheme is less effective than it might be (see following section).
Table 7 Simulated annual effect of real-time closures scheme in terms of average annual values 2006-2010
with max and min annual values in brackets. Analyses assume perfect sampling at sea (all high LPUE
values detected) and were based solely on English data with 9 closures per month distributed across
the scheme area in proportion to the landings of the TR1-2 fleets
Simulated maximum impact of
RTCs, overall, by gear and by
sector
% cod landings
protected by
closures prior to
effort reallocation
(min, max)
% effort
reallocated by 9
closures
(min, max)
% total landings
protected by
closures prior to
effort reallocation
(min, max)
TR1-3 and BT1-2 (area IV south of
56N & eastern Channel VIId)
3.8 (2.3, 5.5)
16.4 (14.4, 18.4)
3.2 (2.8, 3.6)
TR1, all sectors
6.6 (5.7, 7.1)
20.4 (16.9, 23.0)
7.5 (6.5, 8.7)
TR2, all sectors
4.8 (0.7, 10.1)
9.5 (3.0, 17.9)
5.0 (1.6, 10.4)
BT1, all sectors
2.6 (0.0, 5.0)
9.4 (0.0, 29.4)
3.6 (0.0, 10.6)
BT2, all sectors
1.1 (0.9, 2.0)
5.8 (3.3, 8.9)
1.5 (1.3, 1.6)
IVbW (central-west North Sea)
5.9 (3.6, 8.4)
20.1 (15.7, 24.2)
6.6 (5.2, 8.2)
IVbE (central-east North Sea)
0.6 (0.4, 0.7)
6.0 (3.2, 8.2)
1.0 (0.9,1.2)
IVc (southern North Sea)
3.8 (1.8, 6.5)
14.2 (7.9, 24.6)
7.3 (4.9, 10.8)
VIId (eastern English Channel)
2.2 (0.7, 4.9)
20.7 (7.2, 35.3)
9.9 (4.8, 19.7)
Table 8
Simulated annual effect of real-time (sampling) closures scheme in terms of average annual values
2006-2010 with each mitigation option implemented (M1-4 see text). Analyses assume perfect
sampling at sea (all high LPUE values detected) and correspond to maximum effects. Closures were
based on English data with 9 closures per month distributed across the scheme area in proportion to
the landings of the TR1-2 fleets. Red values show potential increases (>1%) in landings.
Simulated impact of
RTCs on TR1-3 and
BT1-2 in area IV
(south of 56N) & the
eastern Channel
(VIId)
% decrease in cod
landings following
closures and
mitigation
without effort
buyback
(min, max)
% decrease in cod
landings following
closures and
mitigation with
effort buyback of
5% for TR1-2
(min, max)
% decrease in cod
landings following
closures and
mitigation with
effort buyback of
11.25% for TR1-2
(min, max)
-
-
% decrease in
total landings
following closures
and mitigation
without effort
buyback
(min, max)
no reallocation
16.4 (14.4, 18.4)
M1 – ‘spread effort’
11.9 (10.2, 13.7)
8.3 (6.3, 10.0)
3.8 (1.5, 5.5)
0.3 (-0.5, 1.0)
M2 – ‘to high effort’
9.7 (7.8, 12.5)
6.0 (3.9, 8.7)
1.4 (-1.0, 4.1)
0.7 (0.1, 1.1)
EVID4 Evidence Project Final Report (Rev. 06/11) Page 20 of 27
3.2 (2.8, 3.6)
M3– ‘to high cod
landings’
M4– ‘to high cod
LPUE by gear’
Mean M2-M3
6.6 (3.8, 9.5)
2.8 (-0.1, 5.6)
-1.9 (-4.4, 0.8)
0.5 (-1.0, 0.9)
5.7 (2.6, 9.0)
1.7 (-1.6, 5.1)
-2.9 (-5.7, 0.2)
0.3 (-1.0, 0.7)
8.2 (5.8, 11.0)
4.4 (1.9, 7.2)
-0.3 (-2.7, 2.5)
0.6 (-0.5, 1.0
Table 9 Simulated effect of RTC (sampling) scheme in terms of average annual values 2006-2010 with each
mitigation option implemented (M1-4 see text). Analyses assume perfect sampling at sea (all high
LPUE values detected) and correspond to maximum effects. Closures were based solely on English
data with 18 closures per month distributed across the scheme area in proportion to the landings of
the TR1-2 fleets. Red values show potential increases (>1%) in landings.
Simulated impact of
RTCs on TR1-3 and
BT1-2 in area IV
(south of 56N) & the
eastern Channel
(VIId)
% decrease in cod
landings following
closures and
mitigation
without effort
buyback
(min, max)
% decrease in cod
landings following
closures and
mitigation with
effort buyback of
5% for TR1-2
(min, max)
% decrease in cod
landings following
closures and
mitigation with
effort buyback of
11.25% for TR1-2
(min, max)
-
-
% decrease in
total landings
following closures
and mitigation
without effort
buyback
(min, max)
no reallocation
25.0 (19.3, 29.8)
6.6 (5.9, 8.2)
M1 – ‘spread effort’
16.9 (12.0, 22.2)
13.6 (8.6, 18.7)
9.3 (4.3, 14.8)
0.2 (-1.5, 1.5)
M2 – ‘to high effort’
14.4 (10.8, 19.6)
11.1 (7.4, 16.4)
6.5 (2.6, 12.3)
1.7 (0.4, 2.8)
M3– ‘to high cod
landings’
M4– ‘to high cod
LPUE by gear’
5.5 (-1.5, 11.3)
2.2 (-5.9, 7.3)
-3.3 (-11.4, 2.7)
0.6 (-1.8, 1.6)
3.4 (-3.5, 9.2)
0.1 (-8.0, 5.5)
-5.6 (-13.7, 0.9)
-0.1 (-2.3, 0.8)
Mean M2-M3
10.0 (4.7, 15.5)
6.7 (0.8, 11.9)
1.6 (-4.4, 7.5)
1.2 (-0.7, 2.2)
2.5.2 Simulated effects for closures based on the historic LPUE (previous two years) of English fleets
If the RTC scheme were based solely on English historic LPUE data and 9 closures per year were distributed
across the North Sea and eastern Channel in proportion to the landings of the otter trawl fleets (TR1-2), on
average 4.1% of beam and otter trawl fleet’s annual effort would be affected by the closures and 9.5% of the
potential annual landings of cod by towed gears would be protected within the closed areas. The otter trawl
fleets would be affected the most greatly with on average 6% of the TR1 and TR2 effort being redistributed due
to closures. A small effect (2% of annual effort) on beam trawlers, BT2, is also predicted by simulations. The
large meshed otter trawlers, TR3, and larger meshed beam trawlers BT1 would be typically unaffected by
closures. The percentage of effort affected by closures is similar when the 9 closures were distributed across
the area due to the combined landings of beam and otter trawlers (BT1-2 and TR1-3 groups). If we assume no
reallocation of effort by vessels affected by closures, then average declines of 11% of the annual landings of cod
by TR1 vessels, 8% of TR2 landings and 3.5% of BT2 landings of cod would ensue (all simulated values are
smaller than those simulated to be attained by real-time sampling, Table 7). Similarly, the closures would
inhibit up to 4% of the expected annual landings of other species by otter trawlers (TR1-2) and 2% of the
expected landings of other species by BT2. The greatest impact by closures is expected in the western central
North Sea (IVb_W), followed by the southern North Sea (IVc), where 7% and 5% of effort would be displaced
respectively and 13% and 6% of cod landings protected respectively. Very little impact is detectable in the
eastern Channel.
EVID4 Evidence Project Final Report (Rev. 06/11) Page 21 of 27
By spreading the displaced effort throughout the fleet, simulations suggest that the potential 8.2% reduction in
landings would be reduced to only 3.2% of annual landings, while option M2 would result in only a 1.0%
reduction in annual landings of cod (Table 10). In contrast, a 1.1% increase in cod landings could potentially
result if fishermen chose to follow option M3; while option M4 would result in 1.8% higher landings of cod on
average than prior to the closures. It is clear that the effect of closures based on LPUE effort is marginal and can
be mitigated by action by fishermen. Of course, fishermen will not have the data available that we have used
retrospectively to indicate where the highest effort, landings or LPUE of cod were reported but it is well known
that many fishermen do communicate to eachother where/when good catches are found so that to some degree
they will be able to make informed choices about their fishing opportunities. Nevertheless, to suggest that
option M4 is taken by all those displaced by closures is highly unlikely. The exact behaviour of the fishermen is
not possible to simulate but we suggest that a mixture of option M2 and M3 is most likely, such that the effect of
9 closures per month in the RTC area is likely to have no effect on cod landings annually. In each of the above
simulations the total catch, including other species, was approx. equal following closures as before.
Assuming that effort displaced throughout the fleet (the weakest mitigation effect by fishermen), a buyback
(extra time for otter trawlers) of 11.25% of annual effort by TR1 and TR2 fleets for adhering to the scheme
would result in a potential 7.5% increase in total cod landings by all fleets annually (assuming that their quota
allowed it, Table 10). Any other mitigation action by fishermen would increase this percentage of landings
further.
If the number of closures per month were increased to 18 the percentage of effort to be reallocated would
increase accordingly (Table 11). However following reallocation of effort, the scheme is not likely to reduce
landings of cod further. No real change is seen when comparing mitigation option M1 between Table 10 and
Table 11. There would be a small effect if the vessel followed option M2 and reallocated to areas of high effort
rather than high landings or LPUE of cod. However, if the fishermen chose to target cod specifically there is a
danger that closing areas based on historic LPUE will drive the fishermen to areas of current high LPUE and this
is reflected in the increases in landings following reallocation of effort by options M3 and M4.
Summary
Although LPUE closures have the potential to decrease cod landings fishermen will be likely to offset the effect
easily if they are aware of areas were cod are typically caught (M3) or areas with particularly high cod catch
rates (M4) and any buyback of effort will exacerbate the situation. If closures are based solely on historic
information, buyback of effort can lead to potential increases in landings of cod irrespective of whether there
are 9 or 18 closures or effort buybacks for TR1-2 of 5% or 11.25%. Of course, the fleet are quota restricted so
such potential increases would translate to increases in the rate at which cod are caught within the year rather
than in the overall landings of cod.
EVID4 Evidence Project Final Report (Rev. 06/11) Page 22 of 27
Table 10 Simulated annual effect of RTC scheme in terms of average annual values 2008-2010 with each
mitigation option implemented (M1-4 see text, note the high LPUE targeted for effort reallocation in
M4 is based on data for the month in which closure is made, not historic LPUE). Simulated RTCs
were based on historic English monthly LPUE values (2006-2008) using average LPUE by subrectangle for 1 and 2 years previous to the month in which the closure was made. 9 closures per
month were distributed across the scheme area in proportion to the landings of TR1-2 trawlers. Red
values show potential increases (>1%) in landings
Simulated impact
of RTCs based on
historic LPUE, on
TR1-3 and BT1-2 in
area IV (south of
56N) & the eastern
Channel (VIId)
% decrease in
cod landings
following
closures and
mitigation
without effort
buyback
(min, max)
% decrease in
cod landings
following
closures and
mitigation with
effort buyback
of 5% for TR1-2
(min, max)
% decrease in cod
landings
following closures
and mitigation
with effort
buyback of
11.25% for TR1-2
(min, max)
% decrease in
total landings
following
closures and
mitigation
without effort
buyback
(min, max)
no reallocation
8.2 (7.7, 8.6)
-
-
2.6 (2.3, 2.8)
M1 – ‘spread effort’
3.2 (2.1, 3.7)
-0.9 (-2.3, -0.1)
-7.5 (-7.8, -7.2)
-0.6 (-1.0, -0.3)
M2 – ‘to high effort’
1.0 (0.3, 1.5)
-3.4 (-3.5, -3.2)
-8.4 (-8.8, -7.6)
0.1 (0.0, 0.3)
M3– ‘to high cod
landings’
M4– ‘to high LPUE
by gear’
-1.1 (-2.5, 0.9)
-5.4 (-7.2, -3.2)
-10.7 (-12.8, -8.3)
0.0 (0.0, 0.1)
-1.8 (-3.5, 0.2)
-6.1 (-8.2, -3.9)
-11.5 (-13.8, -9.0)
-0.3 (-0.5, 0.0)
Mean M2-M3
-0.1 (-1.1, 1.2)
-4.4 (-5.4, -3.2)
-9.6 (-10.8, -8.0)
0.1 (0.0, 0.2)
Table 11 Simulated annual effect of RTC scheme in terms of average annual values 2008-2010 as in Table 10,
but with 18 closures per month were distributed across the scheme area in proportion to the
landings of TR1-2 trawlers. Red values show potential increases (>1%) in landings.
Simulated impact
of RTCs, based on
historic LPUE, on
TR1-3 and BT1-2 in
area IV (south of
56N) & the eastern
Channel (VIId)
no reallocation
% decrease in
cod landings
following
closures and
mitigation
without effort
buyback
(min, max)
15.6 (14.8, 16.4)
% decrease in
cod landings
following
closures and
mitigation with
effort buyback
of 5% for TR1-2
(min, max)
% decrease in cod
landings
following closures
and mitigation
with effort
buyback of
11.25% for TR1-2
(min, max)
-
-
% decrease in
total landings
following
closures and
mitigation
without effort
buyback
(min, max)
6.8 (6.5, 7.2)
M1 – ‘spread effort’
3.9 (0.6, 6.7)
-0.1 (-3.8, 2.9)
-7.5 (-7.8, -7.2)
-2.3 (-3.2, -1.4)
M2 – ‘to high effort’
3.2 (2.7, 3.9)
-0.8 (-1.5, 0.0)
-5.8 (-6.8, -4.9)
0.1 (-0.9, 0.7)
-4.4 (-8.9, 0.6)
-8.8 (-13.4, -3.6)
-14.3 (-19.0, -8.7)
-1.5 (-2.5, -0.5)
-8.0 (-13.2, -2.6)
-12.6 (-17.9, -6.9)
-18.3 (-23.9, -12.2)
-2.4 (-3.7, -1.1)
-0.6 (-3.1, 2.3)
-4.8 (-7.5, -1.8)
-10.1 (-12.9, -6.8)
-0.7 (-1.7, 0.1)
M3– ‘to high cod
landings’
M4– ‘to high LPUE
by gear’
Mean M2-M3
2.6 Evaluation of the effect of including the Dutch LPUE information into the English RTC scheme
In contrast to the English trawl fleets, the majority of the Dutch fishing effort is concentrated in the southern
EVID4 Evidence Project Final Report (Rev. 06/11) Page 23 of 27
and eastern North Sea (Deliverable 4). The Dutch fleet is dominated by beam trawlers (BT2) fishing for flatfish
(sole and plaice) with lower LPUE of cod than similar English trawlers, indicating the difference in targeting
behaviour of the fishermen and spatial coverage of the fleets (Deliverable 4). In November 2011, the Dutch
fishing industry agreed to abide by RTCs suggested by the MMO and chosen with consultation with industry
and the relevant authorities. Dutch LPUE data has subsequently been included into the algorithm that
determines the most suitable ICES sub-rectangles in which to allocate candidate RTC areas. Dutch landings
information has similarly been included into the calculation of the number of candidate areas to suggest for
each sector of the North Sea and the eastern Channel. Unfortunately, an unintended implication of this change
in the scheme is a potential reduction in the effectiveness of the English RTC scheme on English trawl vessels
fishing for cod.
Simulations suggest that the detrimental effect will have been caused due to a number of candidate RTCs been
allocated away from the English fleet’s fishing grounds in waters fished by the Dutch. The simulations were
conducted based on historic LPUE information, as in section 2.5.2, but with combined English-Dutch LPUE
information. Following the current implementation of the scheme with Dutch data included and the number of
closures in each sector of the scheme (IVb_E, IVb_W, IVc and VIId) based on combined landings of all towed
gears (TR1-2), indicates that for all English gears much less fishing effort will fall within candidate closures (for
TR1 36% less, TR2 80%, BT1 68% and BT2 35%) and thus much lower tonnages of cod will be protected by
closures from English vessels. When the numbers of candidate areas per sector were based on landings of all
towed gears (TR1-3 and BT1-2), the effect was slightly greater for TR1 vessels (48% less effort in closed areas)
but similar for the other gear types (TR2 79%, BT1 70% and BT2 35%). The % decrease in effort that would be
displaced by RTCs once Dutch data is included in the analyses was greatest for the English BT1 fleet since the
total effort expended by the fleet was lowest of all the English fleets. Average LPUE of cod by the TR1 fleet is
particularly high relative to other towed gears (BT1-2. TR2-3), therefore the decline in number of closures
affecting this fleet is smaller than for other English gears.
Summary
Due to the differences in spatial coverage by the English and Dutch trawl fleets, the location of RTCs based on
historic LPUE information changes greatly. A direct result of the inclusion of Dutch data into the English RTC
scheme is an allocation of more closures in the southern North Sea and fewer closures in the western central
North Sea, where the majority of cod landed by the English fleets is caught. What the effect on the whole catch
would be is not possible to determine at present since the current data on discards is not of sufficient resolution
spatially or temporally.
2. 7 Modifications to scheme to improve protection for cod
The current English RTC scheme relies heavily on closures identified by historic average LPUEs by subrectangle for the month in question over the two years previously (see section2.5.2 and 2.6). For simplicity, the
calculation of LPUE in the current algorithm takes the total landings and effort by all towed gears for each subrectangle to determine the average LPUE. However, it is known the average LPUE differs greatly by gear. One
might argue that the closure of a sub-rectangle would affect all towed gears within that sub-rectangle and
therefore the average value is appropriate. However, since the scheme is aimed at reducing the highest catch
rates, to avoid very large catches that will likely result in discards, it would be more appropriate to calculate the
gear specific LPUE values. A simple step to improve the current algorithm would therefore be to base the LPUE
score for each sub-rectangle on the maximum of each gear-specific LPUE within the sub-rectangle. For those
areas where a single fleet with a single gear operates the score will be unchanged. However in areas with mixed
EVID4 Evidence Project Final Report (Rev. 06/11) Page 24 of 27
fisheries, such as where TR1 and TR2 groups operate, the higher LPUE score of TR1 vessels will not be downweighted by the low LPUE of TR2 vessels.
Similarly, more recent information might be more suitable such as LPUE information for the month or two
months preceding the proposed closure. Here we simulate the potential effect of the RTC scheme based on
changing the calculation from mean LPUE from all gear data pooled (option C1, Table 12) to max LPUE by gear
(C2) and from changing the basis to be the two months (C3-4) preceding rather than the previous two years.
We also compare this to the simulated effect using real-time data (C5). However, we should note that the ‘realtime’ calculation assumes that the max LPUE would be reported at the beginning of month and therefore the
reduction in landings and reallocation of effort is the total recorded for the month and therefore an
overestimate.
The findings from these simulations indicate that a further 1-2% reduction in typical annual landings of cod
could be attained (assuming scenario M1 or M2) by using the maximum LPUE by gear (C2) rather than the
average over gears (C1) to identify areas for closure (Table 12). However, if those vessels displaced choose to
target areas of high landings (M3) or with high catch rates of cod (M4) the further reduction will be lost. A
change in the LPUE calculation from data from the last two years (C1) to the last two months (C3), would likely
draw into the closed areas 5% more of the typical annual cod landings prior to reallocation of effort. However,
there is also a greater risk of adverse effects if fishermen are able to reallocate their effort to areas of abundant
cod (scenario M3 or M4). Interestingly, this risk is reduced if the LPUE calculation using the preceding two
months is based on the maximum LPUE score by gear (C4). The risk of adverse effects would be none existent if
the LPUE computation were based on real-time gear-specific LPUE data (C5). The most likely scenario, a
mixture of M2 and M3 responses, suggests that historic LPUE based closures are not likely effective (but also
not adverse) unless determined by gear with a much more recent historic period than is currently the case, i.e.
the preceding two months rather than two years.
Summary
Slight improvements to the RTC scheme, in terms of reductions to overall landings of cod, could be made by
altering the calculation of historic LPUE. However, the overall effect of such historic-LPUE closures would still
be small compared to the benefits that might be drawn from effective real-time sampling.
Table 12 Simulated effect of RTC scheme in terms of average annual values 2008-2010 with each mitigation
option implemented (M1-4 see text, note the high LPUE in M4 is based on LPUE data for the month in
which closure is made, not historic LPUE). Simulated closed areas based on historic English monthly
LPUE values (2006-2008) using the average LPUE by sub-rectangle for 1 and 2 years (columns 1 and
2) or months (columns 3 and 4) previous to the month in which the closure was to be made or using
max LPUE in real time (column 5). 9 closures per month were distributed across the scheme area in
proportion to the landings of TR1-2 trawlers. Red values show potential increases (>1%) in landings.
% decrease in cod landings following closures and mitigation
without effort buyback (min, max)
Simulated
impact of
RTCs
C1. Mean
LPUE (current
algorithm) for
2 years
preceding
C2. Max LPUE
by gear by
sub-rectangle
for 2 years
preceding
C3. Mean
LPUE by subrectangle for
2 months
preceding
C4. Max LPUE
by gear by
sub-rectangle
for 2 months
preceding
C5. Max LPUE
by gear by subrectangle in
real-time
no reallocation
8.2 (7.7, 8.6)
10.3 (7.8, 12.5)
13.2 (8.8, 16.2)
13.7 (9.3, 16.3)
17.7 (17.3, 18.4)
M1 – ‘spread
effort’
M2 – ‘to high
effort’
3.2 (2.1, 3.7)
4.1 (1.1, 6.4)
5.6 (4.3, 8.0)
6.3 (4.5,9.4)
13.0 (12.2, 13.7)
1.0 (0.3, 1.5)
2.5 (0.8, 3.4)
3.6 (1.1, 6.2)
3.6 (1.1, 5.8)
10.4 (8.4, 12.5)
EVID4 Evidence Project Final Report (Rev. 06/11) Page 25 of 27
M3– ‘to high
cod landings’
M4– ‘to high
LPUE by gear’
-1.1 (-2.5, 0.9)
-1.1 (-2.9, 2.5)
-3.3 (-5.0, -0.4)
-0.7 (-2.8, 2.4)
6.6 (3.8, 9.5)
-1.8 (-3.5, 0.2)
-1.9 (-3.9, 1.9)
-3.8 (-5.7, -0.3)
-1.3 (-3.9, 1.6)
5.7 (2.6, 9.0)
Mean M2-M3
0.0 (-1.1,1 .2)
0.7 (-1.0, 3.0)
0.2 (-2.0, 2.9)
1.5 (-0.9, 4.1)
8.5 (6.5, 11.0)
2.8 Synthesis of findings
The fundamental reasoning behind the RTC scheme is sound and trigger levels at which closures should be
implemented based on real-time catches are suitable. Unfortunately, the current detection levels of catches
over the trigger points are low and insufficient to generate the desired nine closures per month by the scheme.
Simulations suggest that the real-time closures scheme, with nine closures per month, can potentially
incorporate a buyback of 5% and still reduce landings of cod. However, for this to be the case closures would
have to be based on real-time sampling. For this to happen, detection rates of high catch rates must be
increased through active monitoring by control authorities or by increased information direct from the fishery.
In this latter respect catch-quota vessels could be utilised as pilot vessels to indicate where high catch rates and
potentially high discards might occur. No buyback is appropriate for adhering to LPUE closures since they do
not appear to have any overall effect on landings of cod, assuming that vessels continue to fish within an ICES
rectangle of the closures.
Evaluations made during this project have been based on the change in landings, by the English fleets fishing
south of 56°N, that are considered likely as a result of real-time closures. The benefit of the scheme should
ideally be considered in terms of levels of discarding reduced. However, current data are not of sufficient
resolution, spatially or temporally, to assess the potential changes in catch in undersize fish. One should also
note that a proportion of cod (10 to 20%) landed by the English fleets is caught north of 56°N and this
proportion has been increasing since 2007.
Additionally, tagging studies of cod indicate that real-time closures are likely to become redundant within a
month and real-time data are required to monitor such movement and make RTC responsive.
To improve the scheme, further work should focus on improving real-time information on cod catch-rates
through either improving monitoring of the fleets or stimulating communication with the fishermen.
EVID4 Evidence Project Final Report (Rev. 06/11) Page 26 of 27
References to published material
9.
This section should be used to record links (hypertext links where possible) or references to other
published material generated by, or relating to this project.
Bailey N, Campbell N., Holmes S., Needle C., Wright P. (2010) Note for the Directorate General For Internal
Policies. Policy Department B Structural and Cohesion Policies. Real Time Closures of Fisheries
IP/B/PECH/IC/2009-091. 15 June 2010. 56 pp http://www.europarl.europa.eu/studies
Lee, J., South, A.B., Jennings, S. (2010) Developing reliable, repeatable and accessible methods to provide highresolution estimates of fishing-effort distributions from vessel monitoring system (VMS) data. ICES
Journal of Marine Science, 67, 1260-1271. doi:10.1093/icesjms/fsq010
Needle C. and Caterino R. (2011). Evaluating the effect of real-time closures on cod targeting. ICES Journal of
Marine Science, 68, 1647–1655. doi:10.1093/icesjms/fsr092
Pedersen, M.W., Righton, D., Thygesen U.H., Andersen K.H., Madsen H. (2008) Geolocation of North Sea cod
using Hidden Markov Models and behavioural switching. Canadian Journal of Fisheries and Aquatic
Sciences, 65: 2367–2377
Righton D., Quayle V.A., Hetherington S., Burt G. (2007) Movements and distribution of cod (Gadus morhua) in
the southern North Sea and English Channel: results from conventional and electronic tagging
experiments. Journal of the Marine Biological Association of the United Kingdom, 87: 599-613. doi:
10.1017/S0025315407054641
Righton D., Quayle V.A., Neat F., Pedersen M., Wright P., Armstrong M., Svedang H., Hobson V., Metcalfe J.
(2008) ICES CM 2008/P:04
Codwatch Fisheries Science Partnership www.cefas.defra.gov.uk/our-science/fisheries-information/fisheriesscience-partnership/north-sea-codwatch.aspx
EVID4 Evidence Project Final Report (Rev. 06/11) Page 27 of 27