922 ICES HAWG REPORT 2018 Annex 7: Request on management areas for sandeel in the North Sea Request code (client): 1707_sandeel_NS Background: During the sandeel benchmark in 2016 the management areas for sandeel were amended. This included amendments to management areas 2 and 3, whereby management area 3 now predominantly covers Norwegian waters, whereas management area 2 predominantly covers EU-waters. However due to the breakdown into statistical rectangles, there are areas of management area 3 in EU-waters, where sandeel banks are present. ICES is requested to advice on a way forward for taking this into account. Request: ICES is requested to assess, in order of descending priority, whether: it is possible to include the areas of management area 3, located in EU waters, where sandeel banks are present, to the advice given for management area 2 it is possible to assess the historical level of EU catches (as % of total catches) taken on the sandeel banks in the amended management area 3, located in EU waters it is seen as precautionary to set a joint TAC for EU waters of management areas 2 and 3, allowing the EU TAC for area 2 to be fished also in management area 3 Interpretation of the request If the stock units between management area 3r and area 2r are not well defined, one could lump banks in EU waters in area 3r to area 2r. Can we use VMS and logbook data to find out how much sandeel landings originate from sandeel banks in area 3r in EU waters Would lumping the EU waters 3r TAC and 2r TAC result in overexploitation of either the sandeels in area 3r (EU waters) or area 2r (EU waters). Material and method The amount of Danish catches taken in the EU part of management area 3r and in the Norwegian EEZ of area 2r were estimated and subsequently compared to total catches in area 2r and 3r. This involved analysing Danish fishing activities within ICES rectangles 42F5, 43F6 and 44F8 (see Figure 1), and comparing catches from these areas to the total international catches reported for management area 2r and area 3r. From the Danish logbooks, trips with sandeel fishery are identified, and merged with VMS data by vessel-id and date for the years 2005 2016. The VMS ping rate is one position per hour. Since 2005 VMS has been mandatory for vessels larger than 15 meter, and since 2012 for vessels larger than 12 meter. The positions are filtered by speeds to include those with speeds between 2 and 4 knots. The sandeel catches reported for a catch date are distributed out on the VMS positions with estimated fishery by the time difference between the VMS pings. Logbooks and VMS for other countries were not considered here.denmark realised on average 88% of all EU sandeel catches over the period considered and can therefore be considered a representative sample.
ICES HAWG REPORT 2018 923 The distribution of catches between rectangles has depended on both area-based advice (from 2011 onwards) and implementation of restrictions related to EEZ (from 2006 only Norway has been allowed to fish in Norwegian EEZ). E.g. 2005 and 2006 were the last years of a string of 7 years with very low stock size of sandeel in area 3r and resulted in a fishery which was substantially more widespread than in the years before. Area 2r has had low stock size from 2004 to 2010. In 2005, 2006 and 2009, Norwegian catches were highly restricted in the Norwegian EEZ. In 2016, the fishery along the coast of Denmark (the previous area 2) was restricted to a monitoring fishery while a large TAC was given for area 3, now the northern part of area 2. In 2014 and 2015, substantial area misreporting has been detected. Years where one of the stocks is at low stock size and the area specific TACs are implemented are likely to show more fishing along the border of the area with the lowest TAC. All in all, there are no years in the period for which data are available, where the distribution of catches can be seen as unaffected by management measures or very low stock sizes. For the present analysis, 2014 and 2015 were excluded from the analysis due to known misreporting (ICES 2015; ICES 2016) and 2016 can be considered an anomaly due to large differences in TACs between the previous areas 2 and 3. This hinders answering the request. Results Danish catches taken in the EU-part of area 3 (where area 3r meets area 2r) correspond to 5% of the total international catches in area 2r on average and 8% of international catches in 3r. Excluding the years in which Norwegian catches in 3r were highly restricted, this percentage decreases to 3.7% on average. The percentages are highly variable from year to year ranging from 0 to 16% (0 to 14% if excluding 2016) of area 2r catches and 0 to 42% of catches in area 3r (0 to 13% if excluding years with very restricted fishery in Norwegian EEZ). Table 1 shows total catches in area 2r and total estimated catches within the EU part of area 3r (as % of total catches in area 2r and 3r). Discussion Biological data and modelling exercises presented at the ICES sandeel benchmark in 2016 were in favour of having both area 2r and 3r (instead of fusing them into one area). However, it was not possible to judge, based on biology, exactly where the line should be drawn. Instead, the ICES benchmark group chose the pragmatic solution to make the division between area 2r and area 3r follow the ICES rectangles in such a way that the surface area of EU sandeel fishing grounds in 3r was as small as possible. The decision to use ICES rectangles was made to accommodate the historical data, which are all collected on ICES rectangles. For example, catches cannot be allocated to other than ICES rectangles prior to 2005 due to the lack of VMS data, and even after 2005 the allocation of catches to each VMS ping is an estimate with underlying assumptions. Agedata is only resolved on ICES rectangles.
924 ICES HAWG REPORT 2018 Concluding Request #1: Is it possible to include the areas of management area 3, located in EU waters, where sandeel banks are present, to the advice given for management area 2? Since no exact stock boundaries could be determined, it cannot be stated that sandeel banks in area 3r contribute only to the 3r sandeel stock, hereby opening up the potential to merge banks in EU waters in area 3r to EU waters of area 2r. Providing management advice would however require a redistribution of catch and survey information which is not deemed possible at this point in time (as only reliable data from 2005 onwards is available). Given that catch fractions taken from the EU waters in area 3r are rather variable over time, using a fixed reallocation method would not be appropriate. In addition, catches taken by Norway in Norwegian waters in sandeel area 2r need to be subtracted for this method should to be consistent. This is interpreted as whether a fixed proportion of the 3r TAC can be added to the 2r TAC without the risk of being more unprecautionary. The average proportion of 3.7% could be reallocated to the EU area 2r TAC, but the present analysis does not include Norwegian VMS data and hence cannot reveal the percentage of area 2r TAC which is located in Norwegian EEZ, and which should be subtracted if this method should be consistent. Request #2: Is it possible to assess the historical level of EU catches (as % of total catches) taken on the sandeel banks in the amended management area 3, located in EU waters? It is currently possible to assess the Danish historical level back to 2005 (see table 1). However, it should be noted that this is based on estimated catches, where the catch for a given vessel and date is distributed evenly among VMS pings from that vessel. Therefore any historical (and future) record is associated with uncertainty. Data from other countries are not currently available increasing the uncertainty. Further, there are no years where the distribution of catches can be considered largely unaffected by spatial management measures. Request #3: Is it seen as precautionary to set a joint TAC for EU waters of management areas 2 and 3, allowing the EU TAC for area 2 to be fished also in management area 3? As shown in table 1, while the percentage taken in the EU part of area 3 by the Danish fishery is less than 5% in most years, it was as much as 14% in 2008 and 16% in 2016. The percentage of the 3r catches which are taken in EU zone is up to 13% when excluding the years with highly restricted Norwegian catches. This illustrates that it is very well possible to overexploit either sandeel in area 2r or 3r when a joint TAC would be applied. Conclusion A pragmatic solution to the problem could be to allow that a maximum of X% of the quota for area 2r is taken in the EU part of area 3r within ICES rectangle 42F5, 43F6, 44F8 and that Y% of the 3r TAC is assumed to be derived from EU waters. Such fixed percentages can be taken into account when providing EU TAC advice resulting in precautionary TAC advice.
ICES HAWG REPORT 2018 925 References: Jensen, H., Rindorf, A., Wright, P. J., & Mosegaard, H. (2010). Inferring the location and scale of mixing between habitat areas of lesser sandeel through information from the fishery. ICES Journal of Marine Science, 68(1), 43-51. ICES. 2015: http://www.ices.dk/sites/pub/publication%20reports/advice/2015/2015/sanns3.pdf ICES. 2016: http://www.ices.dk/sites/pub/publication%20reports/advice/2016/2016/san-ns3.pdf
926 ICES HAWG REPORT 2018 Table 1. Estimated Danish catches taken in the EU part of area 3r (inside 42F5, 43F6, and 44F8) presented as % of total international catches in area 2r. Management area Year Total international catch Catch in EU area 3r (% of area 2) Catch in EU area 3r (% of area 3) 2r 2005 34569 0.06 0.15 2r 2006 37952 0.00 0.01 2r 2007 43403 0.02 0.01 2r 2008 35123 0.14 0.07 2r 2009 36709 0.07 0.42 2r 2010 51640 0.02 0.02 2r 2011 24897 0.01 0.00 2r 2012 12552 0.00 0.00 2r 2013 47847 0.03 0.13 2r 2016 9238 0.16 0.03
ICES HAWG REPORT 2018 927 Figure 1. Map of total Danish sandeel catches distributed using VMS pings from 2005 2016. Landing estimates are plotted on a map showing ICES rectangles, sandeel management areas, sandeel fishing grounds (Jensen et al., 2010), and EEZ (thin grey line).
928 ICES HAWG REPORT 2018 Annex 7.1: Reviewers Comments A.7.1.1 Review of the answer to the request on management areas for sandeel in the North Sea Review conducted by Dr. Youen Vermard (Ifremer, Nantes, France) Given the available information, I think the advice answers the specific questions. During Benchmark management area 2 and 3 were redefined but without being able to really determine the real stock boundaries. For simplification purposes boundaries were set using ICES squares, allocating the ICES squares with most of its surface in EU waters in area 2r and ICES squares in Norwegian waters in area 3r. As this delineation was set arbitrarily, it would a priori be possible to include the fraction of ICES squares located in area 3r but in EU waters in the management area 2r. However, giving advice based on these new areas is conditioned to the possibility of redistributing the assessment data to these new areas. As the actual assessment models are based on survey and catch/effort data, changes in the stock distribution is conditioned to the spatial redistribution of these data. The spatial distribution of catch/effort data will allow checking the contribution of these areas compared to the total catches of areas 2r and 3r. Given the available information, it is possible to make hypothesis to spatially reallocate Danish catches and then estimate the percentage of Danish catches made in the EU waters of the 3r area. These catches made in the EU waters of the 3r area are big enough not to be neglected when doing the assessment of 3r and 2r areas and providing advice (catches made EU waters of the 3r area may represent up to 16% of the 2r catches in 2016 only considering the Danish fleets (will be even more if adding the other countries), and up to 42% of the catches made in 3r in 2009). Even if these hypotheses in spatial reallocation are considered accurate, they will not allow building new assessment models following the EEZ as it will only be possible to re-evaluate the catch/effort data from 2005 onward and the actual assessment model need among others catch and effort data on the assessed area. If at some point the EU waters of area 3r are somehow included in the advice given for area 2r, they should then be removed from area 3r and taken into account in the 3r advice. Suggestion: As not familiar with sandeel management, I had difficulties to understand the area delineation, I suggest adding two maps (Figure 1 referred to in the method section is not available in the advice sheet). One representing the previous assessment areas (as the one extracted from the 2016 advice) and another with the new areas (the one provided with the report).
ICES HAWG REPORT 2018 929
930 ICES HAWG REPORT 2018 - I did not understand why ICES square 42F5 was included in the analyses? A large part of its surface is in EU waters and it contains sandeel banks but it is totally included in management area 3. Are there reasons to include this rectangle in area 2r more than 1? - I think that a crucial point is not mentioned in the document; on top of the great proportion of the catches realized in the EU water of the 3r area, it should be mentioned somewhere that the way area 2r and 3r are defined at the moment make the TAC between both areas quite different. In 2017, the advice for area 2r is 2.5 times the advice for area 3r. - I do not understand the first sentence of the advice sheet ICES considers it precautionary, without changing the current TAC procedure, to include areas of management area 3r, located in EU waters, in the advice given for management area 2r. I think it is contradictory with the other conclusions? - A table with the percentage of the Danish catches over the total catches of area 2r and 3r may be interesting to see their contribution over time? Specific comments: Elaboration of the advice - Point 1: It is stated that survey data will likely be affected to a minor extend. I m not sure to understand why. Are survey data defined in a spatial resolution that can cope with a new area definition? If think the main problem is coming from the fact that the assessment model uses international catch and effort data that cannot be disaggregated at the EEZ scale (well described in point 2). Remove the sentence on survey or explain why? - Thorough the text catch is used for landings. Just mentioned at the beginning that discards are considered negligible? - Point 2: does spatial management measures need a small description (area/country specific)? - Point 3: While the percentage taken in the EU part of area 3 by the Danish fishery is less than 5% in most years, it was as much as 14% in 2008 and 16% in 2016. should be rephrased to tell that the percentages are referring to the 2r catches. While the percentage taken in the EU part of area 3 by the Danish fishery is less than 5% of the total catches of the 2r area in most years, it increased to 14% in 2008 and 16% in 2016. Method section - How were trips with sandeel fishery defined? In percentage of the catches? Need for short clarification? - What is the length repartition of the sandeel fishery? Are boats larger than 15 meters in general? Would it be possible to extend the analysis to 2012 (VMS coverage might not be good enough though) - In 2014 and 2015, substantial area misreporting has been detected WG reference?
ICES HAWG REPORT 2018 931 A.7.1.2 Review of the answer to the request on management areas for sandeel in the North Sea Review conducted by Dr. Eoghan Kelly (Marine Institute, Galway, Ireland) Advice Summary ICES considers it precautionary, without changing the current TAC procedure, to include areas of management area 3, located in EU waters, in the advice given for management area 2. This seems reasonable given that this change would add a relatively small amount of catch allowance from area 3 to area 2. It would be helpful to know the corresponding TAC figures for the areas in question. From 2005 and onwards the level of Danish historical catches taken on sandeel banks in the EU waters of 3 range between zero and 16% relative to the area 2 catches. The relatively small proportion of catches made by the Danish fleet in the EU waters of area 3 is encouraging for new advice, although these data need to be checked for misreporting. Also the catches from the other participating states needs to be included in the analysis. ICES is at present not in position to evaluate if, and if so, to what degree, any amendment of the border between area 2 and area 3 could be accompanied by a redistribution of fishing opportunities and therefore calculate whether such a change results in precautionary management. For management to be precautionary it relies on good stock identity. If a joint TAC were applied to areas 2 and 3 it is possible that redistribution of fishing effort could result in the overexploitation of the sandeel stocks in either management areas. Elaboration on the advice 1. The spatial resolution of available biological data and historical catches does not allow the exact stock boundaries to be determined between area 2 and 3. Therefore it cannot be stated that sandeel banks in area 3 contribute only to the 3 sandeel stock, hereby opening up the potential to merge banks in EU waters in area 3 to EU waters of area 2 while area specific TACs are maintained. Such a change in area border would likely affect historical catch and survey data to a minor extent. Providing management advice would however require an assessment in which historical catch and survey information is redistributed. This is not deemed possible at this point in time as the available catch data from 2005 onwards shows substantial interannual variation. This is a reasonable approach until stock identity is confirmed. 2. It is currently only possible to assess the Danish historical catch by EEZ from 2005 onwards. However, this is based on estimated catches, where the catch for a given vessel and date is distributed evenly among VMS pings from that vessel. Therefore any historical (and future) record is associated with uncertainty. Data from other countries are not currently available, thereby increasing uncertainty. Further, there are no years where the distribution of catches can be considered largely unaffected by spatial management measures. It would be useful to look at the location of suspicious misreporting of catches and reallocate during analysis. If vessels are misreporting by logging dummy VMS pings
932 ICES HAWG REPORT 2018 these records should be identifiable by suspiciously high catch rates. VMS and logbook data from other participating states needs to be added to the analysis. 3. While the percentage taken in the EU part of area 3 by the Danish fishery is less than 5% in most years, it was as much as 14% in 2008 and 16% in 2016. The percentage of the 3 catches which are taken in the EU zone is up to 13% when excluding the years with highly restricted Norwegian catches. Including all years, this percentages was as high as 42%, indicating that there may be substantial changes to the estimated stock in some years if these catches were to be moved from area 3 to area 2. This further illustrates that it is possible to exert a high fishing pressure on one stock when large differences exist between the catch opportunities in the two areas unless restrictions are placed to limit the catch which can be taken in the EU part of area 2 or 3 in years where the advised TAC is low in one of the two areas. If the TAC setting procedure had to be amended to reflect the change in habitat using a fixed relocation scheme (i.e a certain percentage of the 3 TAC would be moved to the 2 TAC and vv), sandeel area and member state state specific TACs can be set taking the relocation into account and hereby securing precautionary exploitation levels. Estimating the quantity of quota that should be moved from 3 to 2 is at present not possible and neither is it at present possible to estimate what quantity of quota that should be moved form 2 to 3. Such an analyse should ideally be evaluated by running alternative assessments for area 2 and 3 using amended catch data as well. However since catch data can only be split from 2005 this is not possible. Given the discrete nature of the sandeel banks the ideal management units should be at the functional unit level as is used for Nephrops. Such an approach would resolve the issues regarding TAC allocation between areas 2 and 3. Suggestions This request only concerns EU fishery in EU waters in sandeel area 3 while for a precautionary approach also the Norwegian fishery in Norwegian waters in sandeel area 2 needs to be investigated. This depends on the linkage between the stocks in each management area. From the Danish VMS analysis it seems that the sandeel banks on the EEZ border are connected but there are other banks in the western part of area 3 that may be separate. As mentioned above stock identity needs to be confirmed to insure a precautionary approach.