general enquiries on this form should be made...

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

NoteIn 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 websiteAn 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 INFORMATIONThe 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

MF1104

2. Project title

Spatial and temporal patterns in scallop recruitment and their implications for management

3. Contractororganisation(s)

The Centre for Environment, Fisheries and Aquaculture Science

Lowestoft Laboratory

Pakefield Road

Lowestoft

NR33 0HT

54. Total Defra project costs £ 677,500

(agreed fixed price)

5. Project: start date 01/04/2007

EVID4 Evidence Project Final Report (Rev. 06/11) of 29

mailto:[email protected]

end date 31/3/2012


6. It is Defra’s intention to publish this form.

Please confirm your agreement to do so. YES 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 Summary7.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.

Scallops (Pecten maximus) are one of the most valuable fisheries to England and Wales generating ~£27.4m in first sale value in 2010 (MMO statistics). The use of dredges for scallop capture is, however, controversial as they have the potential to inflict significant damage to the sea bed and conflicts between various stakeholders often arise. There is a pressing need for marine spatial planning in relation to scallop fishery management and in 2008 an area of Lyme Bay was eventually closed to all mobile fishing gears following concern about the activities of some scallop dredgers in the vicinity of features considered sensitive and of conservation value. In order to evaluate the potential costs and benefits of differing styles of management before they are imposed it is desirable to have a modelling framework which can test various management plans. Previous Defra funded projects have focussed on modelling the dynamics of the fishery and adult scallops in the English Channel, so this study was to enhance our understanding of the physical and biological processes driving scallop recruitment and settlement.

The project had two distinct components. The first component was a series of field-based trials and data gathering exercises. The objectives, were designed to determine the timing of spawning events and look for spatial differences in timing, to develop and refine survey methodologies for scallops, particularly juveniles and to try and establish factors affecting settlement and subsequent recruitment to the fishery.

The second component of the project centred around mathematical modelling and would see the further development of the management scenario model alongside new hydrodynamic models of the English Channel with which to drive the larval re-distribution process. We also planned to see if spatial management scenarios developed for current climatic conditions would be robust to future scenarios of climate change.

Over the course of 4.5 years we refined a process for using samples purchased from fishers to determine the spawning cycles of scallops in the western channel. Direct chartering of a fishing vessel gave the most reliable source of material albeit on a small spatial scale, whereas purchasing from vessels/merchants gave a wider spatial coverage but with much larger temporal gaps. The results of this exercise showed that spawning activity from the area exploited by the chartered vessel was synchronous at the scale of ~30km but beyond this the patterns appeared to be different. There were also substantial differences between years with one or two main spawning events occurring over a protracted period. Such variation makes predicting the potential effects of management on population dynamics more problematical but may well provide a buffer to local depletion as larval transport is likely to be more diverse. Examination of oceanographic data could not provide any clues as to why such large inter-annual


differences in spawning cycles was observed.

Sea-going trials were undertaken to seek ways of capturing 1 year old scallops. A number of different techniques were deployed and a 2m beam trawl was the most effective, however the capture rate of even this gear was so low, and the staffing required to adequately sort through the catch that it does not represent a realistic method of sampling for juvenile scallops.

We also tested the theory that pulses of electricity used in conjunction with an underwater TV system would enhance the detection rate of both adult and possibly juvenile scallops. Scallops are often semi-recessed in the sediment and not easy to spot on a TV survey leading to a low detection rate. There is also the possibility of mis-identification of dead-shell. It was anticipated that scallops would, if not swim away from the current at least react to it sufficiently strongly that the motion (or sediment disturbance caused by their movement) would betray their location. A system which delivered pulses of DC electricity was constructed and tested in laboratory and field conditions and the optimal voltage and pulse length established which would induce a reaction in scallops. The delivery vehicle for this system started on a wooden-drop frame and was subsequently transferred to a plastic-coated camera sledge. Despite the confirmed reaction in experiments on captive individuals, when deployed on the sledge and towed, no Pecten maximus were observed to react to the field. Many other species were observed to react and the system may yet prove to be useful in the detection of and surveying for other species.

The project resulted in the construction of the finest scale hydrodynamic model of the English Channel which was then run in tandem with a spatial management scenario testing model. Using results emanating from the field programme we simulated the scallop populations and their fisheries in the western English Channel. The concept of combining spatial management evaluation with hydrodynamic models was demonstrated to work, however some additional refinement is required before the model can be used to test real scenarios.

Project Report to Defra8. 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).


Introduction.This project was designed to provide the opportunity to fill some of the gaps in our understanding of processes governing the recruitment of scallops (Pecten maximus). Previous work in MF0229 had resulted in the development of a management scenario evaluation platform in which the impact of different spatial management regimes for scallop fisheries could be tested. One of the key difficulties in transforming this hypothetical testing framework into something that could actually be used to evaluate "real-world" fisheries, was an understanding of the processes governing the production, distribution and development of newly-spawned scallop larvae.

The project had two distinct components. The first component was a series of field-based trials and data gathering exercises. The objectives, formally listed below, were designed to determine the timing of spawning events and look for spatial differences in timing, to develop and refine survey methodologies for scallops, particularly juveniles and to try and establish factors affecting settlement and subsequent recruitment to the fishery.

1. Fieldwork1.1 To gain a complete seasonal cycle of spawning activity of scallops in the western English Channel from years 2007/2008 and 2008/2009. 1.2 Test response of scallops to electrical fields in tank experiments prior to field tests. 1.3 Develop optical identification technique for determination of Pecten maximus larvae prior to field work.

DROPPED IN AGREEMENT WITH DEFRA.

Changes to the costs of operating the RV Endeavour meant that the planned 2 research trips per year had to be reduced to one longer trip which would not coincide with times considered to be suitable for sampling of scallop larvae. The costs which would have been incurred in this objective were therefore re-invested by intensifying the shore based sampling scheme.

1.4 To identify key locations for spawning, larvae, spatfall and settlement of scallops.1.5 To quantify settlement rates of larvae and to identify and quantify the factors affecting survival of post-settled spat. The second component of the project centred around mathematical modelling and would see the further development of the management scenario model alongside new hydrodynamic models of the English Channel with which to drive the larval re-distribution process. We also planned to see if spatial management scenarios developed for current climatic conditions would be robust to future scenarios of climate change.

2. Modelling2.1 Use climate prediction model data to force Cefas hydrographic models of the English Channel.2.2 To construct a model linking hydrographic processes with adult populations, transport of larvae and subsequent settlement of juveniles. 2.3 To incorporate stock assessment methodologies into the existing spatial management model.2.4 To construct a new management and recruitment model for the Western English Channel. 2.5 Perform a series of test runs with the model to explore the effects of different spatial management regimes and the potential for disruption due to climate-forced changes in hydrography.

1. Fieldwork.1.1. Establishing the spawning timing in the Western English Channel.

This objective was met by obtaining regular, monthly samples from specific areas within the western channel and using these to establish the spawning condition of scallops through the year. Using the findings from a previous Defra-funded project (MF0229), we were aware of several different fishing grounds within the western channel and the presence of different biological characteristics on these grounds. The sampling scheme therefore targeted five different areas within the western channel.

1. Cornish waters west of the Eddystone2. Cornish/Devon waters east of the Eddystone (to start point)3. Scillies4. Offshore & Western Approaches5. Lyme Bay.


3

1

4

25

Figure 1.1 Location of sample areas overlayed on VMS records associated with scallop activity - taken from MF0229 final report.

Area 1 was later subdivided into Inshore and Offshore due to the location and character of the samples, the Inshore region being within 3 miles of the coast.

Method.Although initially it was planned to run this scheme for two years it was eventually run from June 2006 to December 2011 for two reasons. Sample collection commenced in June 2006 and was fully implemented by September 2006. The collection of samples was arranged using both processors and individual fishing skippers. In the original plan sampling was designed to be essentially reactive to where the fleet were fishing as this would enable the widest coverage of samples for the minimum cost. The alternative, of chartering a specific vessel (which was also suggested by an initial reviewer), was considered to be too expensive for a limited spatial coverage. In the end we were able to use both sampling schemes in order to maximise the sampling opportunities.

The first 12 months of sampling encountered difficulties in obtaining sufficient samples for a full spatial and temporal coverage as there was relatively little scalloping activity in area VIIe (a large level of fishing effort was transferred to Cardigan Bay). In addition to this the poor weather conditions prevented many of the smaller boats from fishing.

Between November 2008 and December 2011 a small vessel was chartered to collect monthly samples from four set locations between the Lizard and Mevagissey spanning around 30km. These locations are referred to as "Manacles", "South Old Wall", "The Bizzies" and "Veryan Bay" after the fishing ground names. This proved to be the most valuable source of regular samples, albeit from a smaller geographical area, and provided complete coverage for 2009-2011 (missing only April 2010). Samples were obtained on a calendar month basis and we aimed for a 4 week space between samples but this would vary between 3 and 5 weeks depending upon the weather forecast (smaller vessels are highly weather dependent). Between 2009-20011 the frequency of sampling trips was increased to fortnightly between July and September in order to provide better definition of potential spawning activity.

In order to establish the spawning condition of scallops it is necessary to open the shells to inspect the gonads. As opening the shell is fatal to scallops, samples were purchased specifically for the project. Each sample consisted of 30 scallops covering the range of sizes (above Minimum Landing Size) caught at that event. Ideally we would have purchased a random, unsorted sample which would have enabled the additional investigation of size composition of catches, but in order to ensure that we had individuals from the largest size classes in the haul a very large sample (~200 individuals) would have required purchasing and that was prohibitively expensive.

For each sample we recorded the positional information from the haul, the date of capture (and where possible the time of capture), the date of landing, how the sample was stored prior to collection. The handling and storage of scallops prior to collection was largely beyond our control. Some samples were collected directly from the vessel upon arrival at the quayside, others were left at ambient temperature in sacks for a number of hours, others chilled at quayside storage and some even held on ice. There was also a spread of the time difference from capture to the time of eventual processing. Figure 1.2 shows the distribution of times from capture to final processing, where time of capture was not recorded it has been assumed that the sample was removed from the sea at 12:00. Upon arrival at the laboratory, scallops were either re-immersed


in aerated sea water, or placed in a chilled, controlled temperature facility (~6 degrees). Re-immersion was found to result in high mortality rates, particularly in the summer months and so was discontinued after the first year. Samples were processed either at the Cefas Weymouth facility, or at the Marine Biological Association in Plymouth. The effects of different storage conditions between capture and processing were explored in a separate exercise.

Figure 1.2 Histogram of time difference in hours between capture and processing

For each scallop we recorded the height and width of the whole animal using transmitting, digital callipers. The shell was then opened and the muscle, gonad and viscera (mantle and other organs) weighed separately. The maturity state of the gonad was noted using a 7 point scale (Mason, 1958). Both valves from each shell were cleaned and labelled using a unique identifier for that individual and a photographic record of the inside and outside of both shells made. This was a labour intensive process taking between 4-6 minutes per individual.

Oceanographic data directly concurrent with the samples were not available as data such as sea temperature are not routinely recorded by fishers. The Western Channel Observatory have a moored buoy in 50m of water at a point roughly equidistant from Plymouth harbour and the Eddystone Lighthouse (http://www.westernchannelobservatory.org.uk/buoys.php). The data logging includes physical environmental factors such as temperature and salinity as well as biological factors including chlorophyll. The time series of data does not completely cover the sampling period due to servicing requirements in the winter but provides sufficient data to look for physical cues for spawning of scallops in the wild.

The experiment to determine the potential effect of storage facilities upon the condition of the scallop was performed during one of the RV cruises and the full write-up of this experiment can be found in appendix 1. Scallops were stored at either ambient temperature, chilled or on ice for a period of up to 36 hours between capture and processing, with samples taken at regular intervals. The same processing protocol was followed as for the shore-based sampling with the addition of placing the dissected tissues into a drying oven. The weights of the dried tissues were then compared to their wet-weights and the level of water loss for each treatment and time period was established. A more simple analysis was undertaken for the data from the shore-based sampling in which the weight of gonad and muscle (adjusted for shell size) was compared to the time out of water for storage in chilled and at ambient conditions.

Results.Results: Effect of Storage Conditions.Data from the ship-based experiment showed that there were no significant differences in moisture content induced by either the storage conditions nor the time delay between capture and processing (figure 1.6). There was also no significant effect of the individual performing the dissection. There were, however, significant positive relationships between the percentage of water and the shell size.

In the comparison of shore-based data, there were significant effects of time out of water in terms of the shell-height adjusted weight of the adductor muscle and the gonad. Scallops kept in chilled conditions increased their muscle weight whilst their gonad weight decreased. Scallops kept in ambient conditions increased their muscle weight but the gonad weight did not change over time.


a)

0 24 36 48 60

5060

7080

90

On ice (A)

Time (hrs)

moi

stur

e (%

)

0 24 36 48 60

4050

6070

8090

Chilled (B)

Time (hrs)

moi

stur

e (%

)

0 24 36 48 60

5060

7080

90

Ambient (C)

Time (hrs)

moi

stur

e (%

)

0 24 36 48 60

4050

6070

8090

All treatments

Time (hrs)

moi

stur

e (%

)

b)

0 24 36 48 60

4050

6070

8090

100

On ice (A)

Time (hrs)

moi

stur

e (%

)

0 24 36 48 60

4050

6070

8090

100

Chilled (B)

Time (hrs)

moi

stur

e (%

)

0 24 36 48 60

4050

6070

8090

100

Ambient (C)

Time (hrs)

moi

stur

e (%

)

0 24 36 48 60

2040

6080

100

All treatments

Time (hrs)

moi

stur

e (%

)

Figure 1.6. Effect of storage conditions and time on the percentage moisture in a) adductor muscle and b)gonad.

Results: Shore-based sampling.The number of scallops measured per month per area is given in table 1.1. Area 1 inshore was the most consistently sampled, followed by area 2. Only one sample was obtained for areas 3 and 4 which is insufficient for any form of analysis so these areas shall be ignored.

As the individual scallops were not randomly selected from the hauls it is impossible to say anything about the size frequency within the areas.


Area Date 1 Offshore 1 Inshore 2 3 4 5

Jun-07 - - - - - 76 Jul-07 31 - - - - -

Aug-07 59 30 - - - - Sep-07 86 - - - - 30 Oct-07 60 - - - - 27 Nov-07 30 - - - - 95 Feb-08 - - 41 - - 65 Mar-08 - - 27 - 30 31 Apr-08 40 114 58 28 - 30

May-08 30 - - - - - Jun-08 30 - 59 - - - Jul-08 115 - 66 - - -

Aug-08 100 - 36 - - - Sep-08 60 - 81 - - - Oct-08 30 30 - - - 30 Nov-08 - 123 47 - - 36 Dec-08 - 125 60 - - - Jan-09 - 123 - - - - Feb-09 30 124 60 - - - Mar-09 59 126 59 - - - Apr-09 30 127 60 - - -

May-09 - 119 59 - - - Jun-09 60 124 59 - - - Jul-09 89 250 - - - -

Aug-09 59 252 - - - - Sep-09 60 243 149 - - - Oct-09 30 119 - - - 30 Nov-09 - 121 - - - - Dec-09 29 113 90 - - - Jan-10 - 119 - - - - Feb-10 - 123 - - - - Mar-10 - 121 117 - - - Apr-10 - - 60 - - -

May-10 30 120 59 - - - Jun-10 - 119 60 - - - Jul-10 88 123 31 - - -

Aug-10 150 216 118 - - - Sep-10 59 247 119 - - - Oct-10 - 123 120 - - - Nov-10 - 119 - - - - Dec-10 - 122 115 - - - Jan-11 - 122 - - - - Feb-11 - 121 - - - - Mar-11 - 249 118 - - - Apr-11 - 124 119 - - -

May-11 61 128 - - - - Jun-11 60 124 - - - - Jul-11 161 247 149 - - -

Aug-11 89 123 149 - - - Sep-11 57 240 - - - - Oct-11 - 120 118 - - - Nov-11 - 121 120 - - - Dec-11 - 119 - - - -

Table 1.1 Scallop numbers sampled per area by month.

Results: Shore-based sampling morphologyThere are small but significant differences in the gross morphology of scallops from the different areas. Scallops from area 1.Inshore and 5 are slightly taller for a given length than those in area 1.Offshore and scallops from area 2 are slightly smaller, but these differences although statistically significant are only in the order of less than 1mm.lm(formula = ShHgt ~ ShLen + HArea)Residuals: Min 1Q Median 3Q Max -16.301 -1.346 0.008 1.323 15.864 Coefficients: Estimate Std. Error t value Pr(>|t|) (Intercept) 7.241008 0.326760 22.160 < 2e-16 ***ShLen 0.819704 0.003031 270.459 < 2e-16 ***HArea1 Inshore 0.509935 0.056064 9.096 < 2e-16 ***HArea2 -0.155478 0.062419 -2.491 0.012759 * HArea5 0.377055 0.108513 3.475 0.000513 ***---Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

Residual standard error: 2.055 on 10542 degrees of freedomMultiple R-squared: 0.8877, Adjusted R-squared: 0.8876 F-statistic: 2.082e+04 on 4 and 10542 DF, p-value: < 2.2e-16

Differences in the relationship between shell height and shell weight would demonstrate differences in the ability of the scallops to lay down shell. As weight will be predominantly a function of the area of the shell we use the square of the height as the explanatory variable. Again there are slight regional differences, with area 1.offshore having the lightest shell at length and area 5 having the heaviest shell at length.lm(formula = ShWgt ~ (ShHgt^2) + HArea)


Residuals: Min 1Q Median 3Q Max -40.039 -5.318 -0.392 4.782 53.386

Coefficients: Estimate Std. Error t value Pr(>|t|) (Intercept) -143.87408 1.30759 -110.03 <2e-16 ***ShHgt 2.40112 0.01367 175.68 <2e-16 ***HArea1 Inshore 2.66464 0.22254 11.97 <2e-16 ***HArea2 3.10445 0.24686 12.58 <2e-16 ***HArea5 9.30992 0.42930 21.69 <2e-16 ***---Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

Residual standard error: 8.124 on 10542 degrees of freedomMultiple R-squared: 0.773, Adjusted R-squared: 0.7729 F-statistic: 8975 on 4 and 10542 DF, p-value: < 2.2e-16

Results: Shore-based gonad staging.The visual staging of gonads is a subjective exercise and so an alternative quantitative gonad index was also developed from the tissue weights. Given that a significant relationship between water content of the gonad with shell size described above, all gonad weights from the shore-based sampling were transformed to dry-weight equivalents using coefficients from a linear model of percentage dry weight to shell height. Call:lm(formula = PCdryG ~ (ShHgt))

Residuals: Min 1Q Median 3Q Max -16.983 -8.668 -4.160 3.063 69.885

Coefficients: Estimate Std. Error t value Pr(>|t|) (Intercept) 53.5661 11.9929 4.466 1.13e-05 ***ShHgt -0.3853 0.1255 -3.069 0.00234 ** ---Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

Residual standard error: 14.39 on 296 degrees of freedomMultiple R-squared: 0.03085, Adjusted R-squared: 0.02757 F-statistic: 9.421 on 1 and 296 DF, p-value: 0.002344

The overall size of an individual will have an effect upon the size of gonad for a given maturity stage, so this was accounted for by modelling dry gonad weight equivalents (GW) as a function of shell height (SH). Given the significant differences in morphology described above as well as previous data showing different growth rates by area, sample location was also used as an explanatory variable.. A full scallop gonad is strongly three dimensional shape and therefore the relationship between GW and SH is expected to be non-linear. Several models of GW to SH were explored, GW~SH, GW~SH^2 and GW~SH^3 and the suitability of the third model (SH cubed) was confirmed by comparing the Akaike Information Criterion scores from the different model fits. Provided that the dataset contains a wide range of gonad stages and scallop sizes the residuals (expected value - observed value) from this model fitting exercise indicates a level of gonad fullness where negative values are gonads lighter than expected for that size and vice versa. Temporal changes in these model residuals (termed Relative Gonad Index or RGI) can then be used to see when gonads are filling (an increase in RGI), or when spawning occurs (decrease in RGI). The temporal trend of RGI was summarised using the median value for each combination of location and sampling trip.

Comparison of the proportion of gonads scored 5&6 (just pre-spawning) from the visual staging technique with the RGI shows broadly similar patterns in spawning activity. Of the two approaches the RGI is easier to interpret. There were distinct differences in spawning pattern over the three years of sampling with the most detailed coverage (Inshore sector of area 1, 2009-2011). 2009 saw gradual build up of GSI before a drop in both GSI and the visual staging around week 20, followed by another decline after week 22. 2010 by contrast saw a much stronger build up of GSI until week 26 whereupon there was a very sharp drop indicating a strong spawning event, followed by a second, smaller event in week 36. The 2011 spawning pattern was for less pronounced gonad filling and two spawning events, one around week 26 and the second around week 35.

There is a high degree of similarity in the GSI trajectories between the four inshore locations showing a strong synchrony in spawning times over a distance of around 30km. One common feature across these three years is that the scallops from the Manacles have larger model residuals implying that there is more reproductive development and output per scallop from this site than in other sites.


Figure 1.3 Comparison of the visual gonad staging technique (left) to the objective Gonad Somatic Index (right). Panels are the different years, each line represents a sample site from within inshore Cornish waters.

Although the temporal coverage is not as extensive with areas 1, 2 and 5, it can be seen that there are definite differences in the spawning pattern in terms of strength and timing between the years. Area 2 in 2010 for example appears to have a single spawning event around week 33 whereas there is evidence of several small spawning events over a more protracted period in 2008.

Figure 1.4. Gonad Somatic Indices for the sample areas over all years.

Results: Oceanographic correlation.No physical oceanographic data were collected with the samples as they came from commercial fishers. The Western Channel Observatory (WCO) is an oceanographic time-series and marine biodiversity reference site in the Western English Channel. As part of their data collection stream there is a moored buoy (L4) lying in ~50m of water equidistant between Plymouth and the Eddystone Rock which captures physical and biological information at hourly intervals.


a) b)

c)

Figure 1.5 Time series of GSI (red line) from area 1 (inshore) and oceanographic measurements from the Western Channel Observatory. a) Temperature at 40m, b) Surface temperature, c) Chlorophyll.

DiscussionThe lack of significant effect of storage conditions and time delay between capture and processing in the controlled experiment gives us confidence that the signals observed in the GSI are more likely to reflect spawning activity than being artefacts of the storage conditions. The indication of significant effects of storage in the shore-based sampling may indicate that the parameter set explored in the controlled experiment was not broad enough to capture the range of real-world conditions (although the time component is of course the same). The shore-based sampling was conducted year-round and therefore a much greater range of external conditions would have been experienced by the scallops, including the increased possibility of being exposed to rain-water which is highly likely to affect the water content of living tissues.

The GSI appears to clearly show patterns of spawning activity that vary both temporally and spatially. The Inshore Cornish samples have a high degree of synchronicity operating over the range of ~30km. It is not clear however if this is in response to environmental variables or whether the scallops themselves are able to trigger spawning activity in others. Had the temporal resolution of the data been even finer (weekly or even less) then it may have been possible to test for waves of spawning activity passing through the region. The Manacles ground may represent a region of higher reproductive output due to the wider range of GSI values and the tendency for more scallops to be classed as stage 5 or 6. Data collected during the RV surveys of this project and previous studies have demonstrated that whilst scallops from Start Point to the Lizard are considerably slower growing their counterparts in Lyme bay, the individuals towards the Falmouth end have a faster growth rate than at the Plymouth end. It is therefore of interest to know what contribution scallops from small patches as the Manacles might make to the wider sustainability of the local populations.

The inter-annual difference in spawning timing is a striking feature of this dataset and has only become apparent with the re-allocation of resources to this particular part of the project. It highlights the need for research in the natural environment to be able to plan for longer term studies. Had this been a three year exercise, given the weather difficulties experienced in the first two years of the project we might easily never have collected such a comprehensive dataset. That said, the intensive sampling has only covered a relatively small area and it would be of great interest to determine if such variation and synchrony occur in other areas.

The presence of such a potentially protracted spawning period has immediate implications for management. The use of closed seasons to protect a spawning population would result in a very short fishing period which would be unpopular and quite possible uneconomical to the industry. In addition to the small time window in which spawning is unlikely to be occurring, the marked value of scallops is highest when the roes are well developed and the meat yield/quality immediately post spawning is particularly poor as the scallops use up reserves of energy during spawning. The process of modelling the generation and transport of scallop larvae was made more complicated by the possibility of such protracted and multiple spawning seasons. One issue we were unable to test, as the sampling process was destructive to the individual scallops, is whether the total gamete production in a year is constant, or whether years with multiple spawning events permit a higher overall output.

1.2. Test response of scallops to electrical fields in tank experiments prior to field tests. The use of video footage to determine scallop abundance is widely used, including for direct stock assessment in the USA. The species assessed using these techniques tend not to be as cryptic as Pecten and tend to rest on the sea-bed rather than slightly burrowing down. It can, therefore, be difficult to tell if an observed Pecten shell is necessarily part of a live individual and what the absolute level of detection rate is. We hypothesised that the use of electrical currents as part of the TV survey apparatus would induce a reaction that would enhance the detection rate of scallops. It was anticipated that scallops would, if not swim


away from the current at least react to it sufficiently strongly that the motion (or sediment disturbance caused by their movement) would betray their location.

Poll and Carr (2002) tested electrical stimuli on Placopecten magellanicus and Argopecten irradians with the aim of inducing a escape (swimming) response. Such an effect would mean that scallop fisheries could use gears with minimal sea-bed contact and therefore be more "environmentally friendly". Poll and Carr tested the effects of both voltage and pulse length and their results guided the voltage range used in our tests. We had direct discussion with Poll and Carr regarding the equipment they used for their tests. Commercial suppliers of marine electronics were approached to construct the electronic systems required to safely deliver DC current of variable voltage and pulse length but these proved to be prohibitively expensive (>£16,000). The approach eventually taken was to construct a pulse generator using in-house expertise which proved adequate for the task. The final configuration arrived at consisted of an array of 9 electrolytic capacitors (22 micro-farad), arranged in 3 banks within which the capacitors were arranged in parallel. Each bank was then switched by a solid-state relay and the whole assemblage was mounted in a pressure housing on the camera frame. The relays were triggered by a frequency generator which controlled the of each electrical pulse. The capacitors were trickle-fed through one core of the towing cable whilst the trigger signal was passed down a separate core. The towing cable also contained cores for powering lighting and the video camera, and a co-axial cable for returning the video feed to the surface. The voltage to the system was controlled using a series of power transformers which, along with the frequency generator were kept on the ship-side. A commercial system would have provided a 1-box product which would have been transferable between different operating platforms, whereas the system here comprised numerous separate units and was therefore only fit for use on the RV Endeavour. The system used in the 2009 and 2010 surveys used capacitors rated to 100v and these were up-rated to 200v for the 2011 survey and the relays were up-graded to match.

Sea-water has a high impedance to electrical currents, whilst the majority of frames for the deployment of marine camera systems are constructed of stainless steel. The use of such a steel frame would provide a pathway of considerably lower resistance to any electrical field and therefore a non-conductive frame was required. The first frame was constructed of predominantly of wood (100mm square fence posts), with metal strengthening at the joints (ensuring that the metalwork did not provide a complete circuit). Given the fragility of a wooden structure it was necessary to mount the cameras on a metal frame which was situated on top of the wooden frame (see figure 1.6). The closest continuous metalwork was 1.2 m above the electrodes which proved to be sufficient distance to prevent shorting through the frame. Upon deployment it was discovered that the natural buoyancy of the wooden feet combined with the denser metal top structure caused stability issues, and further weights were added to the feet. The electrodes were two bars of threaded steel positioned just inside the frame on opposite sides, giving 1.2m between them. The safe operation of the equipment was paramount and, in addition to strict protocols governing the powering-up, discharging and monitoring of the capacitor bank, an additional fail-safe system was added in the form of floating short-bars. These comprised a pivoting arm with floats on one side such that when the frame was in the water the floats lifted the arm away from the electrodes, but when emerging into air the bar would fall down and create a short circuit which ensured that the capacitor bank was fully discharged.

Figure 1.6. Drop-frame with electrical stimulator apparatus being deployed. A row of scallops to be tested can be seen attached to the netting at the bottom of the frame.

Initial testing of the concept was performed on land using a 2m circular tank filled with seawater to a depth of 50cm. 40 commercially caught, live scallops were purchased and introduced to the tank in batches of 5 where the temperature was 5.5oC ±0.2 . Scallops were placed in a line perpendicular to the electrodes across the centre of the tank and were left to acclimatise until all scallops had opened their shells and the mantle was visible. Three voltages (40, 70 and 95v DC) and a range of pulse lengths (5, 20, 30, 40, 50, 100 and


150) were tested. The 100 and 150ms pulses were only tested at 95 volts, otherwise the full combination was explored. Scallop reactions were scored from 0-3, with 0 = no response, 1= Slight Reaction (i.e. slight close, bubbles), 2=Moderate reaction (i.e. slow closure), 3= Strong reaction (i.e. "clap" shut, spinning). Once tested, the scallops were removed from the tank and a new batch introduced for acclimatisation prior to testing.

No escape response (swimming via clapping) was ever observed and in all tests there was at least one scallop which did not react to the stimulus. The proportion of scallops exhibiting one of the reaction levels under each combination of voltage and pulse length is shown in figure 1.7.

Reaction

Pro

porti

on

0.0

0.2

0.4

0.6

0.8

1.0

1 2 3 4

5

1 2 3 4

20

1 2 3 4

30

1 2 3 4

40

50 100

0.0

0.2

0.4

0.6

0.8

1.0

150

40 70 95

Figure 1.7 Reaction strength (1-4) of scallops to DC electrical fields, different colour bars represent voltage, panels are pulse length in milliseconds.

Using proportional-odds modelling (fitted in R), voltage was the best descriptor of response but there was no difference in response over 40v. Pulse length had no significant effect overall, however very short pulses (5ms) at low voltage (40vDC) generated no response from any animal.

Although the wooden drop-frame was successfully deployed, it was felt that the ability of the gear to cover sufficient ground and general fragility of such a structure in the marine environment meant that it was unsuitable for extended deployment. A marine camera sledge was therefore modified for use in these trials. Previously used for the underwater TV survey of Nephrops stocks, the 2.15m long * 1.2m wide * 1.6m high steel sledge required electrical insulation prior to use as a vehicle for electrical fields. This was achieved by heating the frame and then dipping in nylon-polymer resin (Leyfos Plastics, Shropshire, UK) which gave a coating of ~1mm thickness. The tubular structure had previously been drilled to ensure that it did not crush under pressure but the plastic coating process required that the drain holes were filled before dipping. To ensure that the frame did not crush under pressure the voids were subsequently filled with vegetable oil. The electrodes took the form of metal shoes attached to the underside of the runners, and the integrity of the electrical insulation was protected by attaching polypropylene pipes to the runners. This was deemed necessary as the plastic coating was likely to be abraded on the sea-bed and by the action of clamping the electrode-shoes to the frame. A digital video camera (Kongsberg) and 4 LED light arrays were positioned on the sledge which was then towed with an armoured umbilical cable. Video footage including an overlay of the date, time and GPS position was recorded to DVD on board the RV.


Figure 1.8 Plastic-coated underwater TV sledge. Metal skids (A) at front act as the electrodes, capacitors are housed in the grey-plastic pressure housing (B), camera mount (C), lights (D), Hipap beacon for gear monitoring and recovery (E). Flotation devices are to ensure the frame lands upright and to aid with recovery.

Figure 1.9 Plastic-coated underwater TV sledge being deployed from RV Endeavour


1.3. Dropped, see introduction.

RV surveys.1.4. To identify key locations for spawning, larvae, spatfall and settlement of scallops1.5. To quantify settlement rates of larvae and to identify and quantify the factors affecting survival of

post-settled spat Objectives 1.4 and 1.5 were tackled jointly during the RV surveys of 2009, 2010 and 2011. Along with the testing of the electrical stimulator device, these surveys were also looking to determine the most effective method of sampling juvenile scallops in order to be able to determine factors affecting settlement and recruitment. We also sampled for adult scallops to look for the proximity of juvenile and adult areas and finally we sampled for newly settled Pecten spat on epifauna.

Survey design.Given that the primary purpose of the surveys was to test different sampling methodologies and test for linkages between the different life stages rather than track relative abundance through the years, the location and design of the survey stations were adapted throughout the series. During the 2009 and 2010 survey a local scallop fisher joined the survey whilst the most coastal stations were fished. This was considered important as they had an intimate knowledge of the area in relation to the location of fixed gears (i.e. pots and fixed nets) and would be able to direct us towards clear tows as well as giving an insight into the precise areas used by scallop fishers. The intention was to have such an industry observer on-board for the 2011 survey but this fell through at the last minute. As a result a number of stations which we had intended to sample close inshore in the Falmouth region could not be reached without the possibility of damage to commercial gear so these plans were abandoned.

Each station sampled required a pass over the ground with the multi-beam sonar running to check for obstacles likely to snag towed gear. The site was then sampled with a minimum of two different gear types. A set of 4 commercial scallop dredges and the 2m beam trawl were always deployed and on the majority of stations a sandeel dredge was also deployed. During the hours of darkness the sites were re-visited and the underwater TV system was deployed. During the 2009 cruise this comprised the drop-frame, for the 2010 and 2011 cruises this was the modified sledge. The RV Endeavour's Dynamic GPS system enables highly precise position holding for the drop-frame and permits a steady course to be maintained at the slow speeds required for reliable TV footage.

For each sample obtained all Pecten were measured (shell height to the nearest mm). All commercially exploited fish were identified to species and measured to the nearest cm. All other species of benthic fauna were identified to species where possible (otherwise to order level was deemed acceptable) and a total count and weight were recorded by species (or order). Qualitative abundance scores were also used for colonial animals or very small animals where detection rates may have been low.

Adult Scallops.At each station a set of 4 heavy-duty commercial scallop dredges (manufactured by BeeJay Ltd) was deployed. The dredges had belly rings of 10mm steel with a nominal internal diameter of 75mm and back rings of the same nominal internal diameter made from 8mm steel. The tooth-bars had nine round teeth of 25mm diameter and 80mm effective length. The four dredges were mounted on a single towing bar with rubber bobbins. This was deployed from the trawl winch.

The dredges were towed for a nominal 30 minutes from the point of reaching the seabed to the start, of hauling. Shoot and haul positions were recorded and the distance travelled for each tow was determined. After tipping the dredges, the entire catch was sorted on deck and the total weight of each species recorded, together with the weight of stones and shell retained by the dredges. All commercial fish and shellfish were measured.

Size frequency compositions, for 5mm size categories, were been constructed for each station in each survey for the catches in commercial scallop dredges. It appears that the dredges retain scallops in significant numbers down to a size of about 70mm shell height, although dredge efficiency at size (a combination of gear selectivity and catchability) increases with size up to around 100mm in dredges of this specification (Dare et al 1993) so scallops will not be fully selected by the gear until they are significantly larger than the minimum landing size (100mm shell length, approximately equivalent to 88mm shell height) . Smaller scallops are retained down to 40mm but this depends to a large extent on the nature and amount of material in the dredge. For purposes of comparison, scallops below 65mm shell height have been excluded from the analysis.

There was a significant east-west relationship in both overall catch rate (higher in the west) and mean size (lower in the west) (figure 1.8). This relationship with size was also observed during previous scallop survey programmes in the area and is largely driven by significantly slower growth rates from the Lizard round to the Eddystone. Despite the same gear being used, catch rates in 2011 for were higher than in previous years


across a range of sizes (figure 1.9)

a Size class

Pro

porti

on

0.0

0.1

0.2

0.3

2009

0.0

0.1

0.2

0.3

2010

0.0

0.1

0.2

0.3

67.5

72.5

77.5

82.5

87.5

92.5

97.5

102.

5

107.

5

112.

5

117.

5

122.

5

127.

52011

East West

)b) Longitude

Mea

n le

ngth

in m

m

90

100

110

120

2009

90

100

110

120

2010

90

100

110

120

-5.0 -4.5 -4.0 -3.5 -3.0 -2.5

2011

c) Longitude

Tota

l cat

ch ra

te

0

50

100

150

2009

0

50

100

150

2010

0

50

100

150

-5.0 -4.5 -4.0 -3.5 -3.0 -2.5

2011

Figure 1.8 a) Size frequency from scallop dredge catch east and west of Eddystone, b) Catch rate vs longitude c) mean size vs longitude

Figure 1.9. Catch rates in the scallop dredge for sizes <90mm and >=90mm,

There is, however, even finer spatial structuring in the population. For those sites off Cornwall (west of 4oW), the mean size in the catches (for scallops 70mm and above) showed those sites to the west, off Falmouth Bay and down toward the Lizard, had a significantly higher mean size than those further to the east (figure 1.10). The more comprehensive coverage of the sampling in 2011 suggests not only that scallops are larger to the west, but also in the more inshore sampling.

This division could result from differences in exploitation ( e.g removal of larger scallops through higher fishing pressure to the east); faster growth rates in the west; or differences in the size at first maturity since scallops, like many bivalve switch resources from shell growth to reproductive output at maturity leading to a marked reduction in growth rate. This difference in population mean size was also apparent in surveys conducted off Cornwall in the 1990s and growth curves rates show a significantly faster growth rate to the west of the area suggesting that the difference is growth related. We were unable to use the shore-based sampling to test the hypothesis that size at first maturity might affect the growth rates as the commercial size classes used meant that <0.3% scallops were classed as immature.


Figure 1.10 Catch rates of scallops (>=70mm height) in dredge catches split by those samples with a mean length <90mm and those >=90mm.

Juvenile Scallops.Juvenile scallops have historically proved difficult to sample, leading to uncertainty about the relation between where they settle and the distribution of the adult population. Attempts to employ epibenthic sledges and small mesh scallop dredges in Cefas surveys in the 1990s were largely unsuccessful, although patches of one-year-old scallops 25-35mm in size were found in 1990, and re-sampled over the following 4 years (Palmer 1998). In addition numbers of scallops between 17 and 25mm where found at the Manacles, south of Falmouth in 1994. Both of these areas were identified using scallop dredges rigged with 40mm belly rings and 14 teeth on the tooth-bar. However these dredges were only effective in these two very limited areas. It seems most likely that the retention of scallops at sizes below 40mm depend as much on the nature of the seabed and the action of the gear on it as the abundance of the scallops themselves.

For this project, several different types of sampling gear were deployed in order to test their effectiveness at capturing small (<40mm) scallops.

Beam trawl. A 2m Jennings beam trawl was rigged with a 20mm mesh net and 4mm cod end liner. The net was deployed on a wire from Endeavour’s net drum through a block on the A-frame. It was towed for a nominal 10 minutes and on recovery the entire catch was tipped into fish boxes and carried to the sorting table. The quantity of bulk catch was recorded as number of fish-boxes. Each box in turn was tipped out and all organisms picked out. Where the catch was very large this was sub-sampled. Where the catch consisted of a large proportion of sand, the catch was first washed through a 5mm sieve to reduce the bulk and make the organisms easier to spot and remove. Sorting the catch was almost invariably a labour intensive process with typically five staff searching a sample for a period in excess of 30 minutes. Several different ways of organising the sorting were trialled over the three cruises but the method which coincided with the most juvenile scallops being found was a conveyor type system in which the sample was picked over at least three times as the bulk was moved down the sorting table. Even with such a rigorous sorting technique detection rates are probably below 100% as small scallops often have very similar pigmentation to the other benthos and their dual concave shape means they can readily adhere to pebbles with surface tension. In contrast, small queen scallops (down to 7mm shell height) were relatively easy to spot, partly due to their higher abundance, but also due to their dual convex shape which reduces adhesion to other benthos. The beam trawl was used as a sampling tool at each station over all three years. Figure 1.11 shows the length distribution of scallops captured using the 2m beam trawl although the very low numbers caught in 2009 mean the frequency is relatively meaningless . Of the smaller size scallops (<40mm) the yearly totals for this gear were 3, 12 and 50 for the three cruises respectively, dropping to 1, 0 and 14 for scallops <25mm.

Sandeel dredge. A modified 1.5m scallop dredge with a fixed-tooth design has been effectively deployed as a sampling tool for sandeels by a number of European laboratories including Cefas. The bag consists of a 4mm mesh liner inside a chain belly. Tow duration was limited to 15 minutes due to the tendency of the gear to rapidly fill. The sorting protocol used for the sandeel dredge was the same as for the beam trawl. The sandeel dredge was used quite extensively through the surveys as it was initially felt that the toothed bar should increase the chance of picking up any small scallops that were recessed into the sediment as well as digging more sediment out which might reveal evidence of previous settlement (i.e. dead shell). These theories proved to


be unsupported by the capture rates of live juvenile Pecten and despite meticulous fingertip searches, fragments of dead juvenile Pecten shell were never observed. We ceased using this particular piece of gear part-way through the 2011 survey and increased the staffing level on the sorting of the beam trawl catches.

Anchor dredge. This simple piece of kit is essentially a rectangular metal bucket with an angled scoop. It was only used briefly during the 2009 cruise as it proved to be very ineffective at sampling for scallop larvae due to its tendency to rapidly fill up.

Baird Oyster dredge.

This was trialled in 2010 and though it was effective for catching larger scallops when unmodified, it would not fish when a small mesh liner was fitted

Size class

Pro

porti

on

0.00

0.05

0.10

0.15

0.20

0.25

2009

0.00

0.05

0.10

0.15

0.20

0.25

2010

0.00

0.05

0.10

0.15

0.20

0.25

12.5

17.5

22.5

27.5

32.5

37.5

42.5

47.5

52.5

57.5

62.5

67.5

72.5

77.5

82.5

87.5

92.5

97.5

102.

510

7.5

112.

511

7.5

2011

Figure 1.11 Length frequencies for Pecten from the 2m beam trawl.

Scallop SpatAs scallop larvae metamorphose and settle out of the pelagic phase, they seek a primary settlement site to which they attach with byssus thread. These sites will usually be sessile plants or animals standing clear of the sea-bed. In inshore waters such material will include macro algae but in deeper waters such as the survey area covered in this project, hydrozoans and erect bryozoans provide the most abundant sites. Scallops remain attached to these sites to a size of around 5mm shell height, but then detach and settle onto the sea-bed, typically choosing sand and gravel substrates although they also settle onto soft silty sand.

There is evidence (Beaumont & Barnes 1992) that scallop spat can use their byssus thread as a drogue to kite in the tide possibly as a means of sweeping them over the seabed until they encounter a suitable primary settlement site, although it may be that they can use this mechanism to move on from the primary site if it proves unsuitable This secondary dispersal phase, called bysso-pelagic drift, has also been demonstrated for post-settlement mussels (Lane et al 1985).

Investigations In the mid 1980s into the primary settlement of scallop spat in the Western Channel (Dare & Bannister 1987) identified four common hydroid species and two erect bryozoans as widespread although of very variable abundance. However scallop spat were not found associated with the latter two hydroid types and numbers were generally low. On all samples spat of the queen, Aequipecten opercularis were greatly more abundant than were Pecten.

MethodDuring the three cruises completed for this project, material that could potentially form primary settlement sites was picked out for examination. Not all stations were sampled, due to the logistics of space and


chemical requirements, but generally at least two stations per day were sampled. Of the various gears employed during these surveys, the 2m beam-trawl caught the largest quantities of hydroid and bryozoan material. Time did not allow for the separation of species, often they formed a tangled mat, although frequently one species was dominant. The most abundant species were the bryozoan Cellaria and the hydrozoans Abietina , Nemertesia and to a lesser extent Sertularella. None of the bryozoan Eucratia, with which scallop spat were associated in Lyme Bay during the 1980s study, was observed.

After sorting out from the rest of the catch, the material was put into a solution of 5% sodium hypochlorite for 20mins to dissolve the byssus. The material was then thoroughly washed through a 345μ sieve. The sieved material was then examined under a low power binocular microscope for pectinid spat.

ResultsTable 1.2 gives details of the material caught over the three surveys, and the numbers of Pecten spat identified for stations where Pecten were identified.

Year Station Description of mixed material

No. Pecten

No. Aequipecten

2009 45 Cellaria & Sertularella 1 97

48 Cellaria & Sertularella 1 57

2010 65 Cellaria 11 175

69 Sertularella 4 102

2011 5 Artificial fibre 1 3

58 Cellaria & Diphasia 3 19

Table 1.2 Numbers of scallop spat identified by benthos type and year.

DiscussionThe numbers of scallop spat identified during the surveys were low, and of the same order as those found during investigations in the 1980s. A significantly higher number were found during the 2010 survey than in 2009 and 2011, but there is no way to tell if this results from a higher level of settlement or is due to varying spawning times. In any case the higher numbers in that year were almost all from two samples. The higher number of spat in 2010 was followed by increased numbers of 1 year old scallops in beam-trawl catches in 2011, but there is no match between the spatial distributions.

It is possible that this method underestimates spat numbers because spat are lost from the material during the catching process, particularly when there is a lot of stones and abrasive shell debris in the catch. However, embyssed bivalves are very strongly attached and the beam trawl was towed for only ten minutes so it is likely that such losses are not significant.

Distinguishing spat of Pecten from those of the much more abundant Aequipecten is straightforward. The former tend to be strongly D shaped and smooth in appearance at sizes below 5mm, while queen spat are always more extended dorso-ventrally and are invariably ribbed or at very small sizes (below 1mm) exhibit a pocked surface sculpture (Figure.1.12)

Figure 1.12 Spat of P. Maximus (left) and A. Opercularis, 1.6mm and 1.8mm respectively.


Deployment of the underwater TV system with electrical stimulatorThe drop-frame design of UWTV system was successfully tested during the 2009 cruise. At-sea modifications were required to the frame as the natural buoyancy of the timber was such that it tended to invert when deployed. There was concern that the controlled tests in the plastic tank might artificially distort and concentrate the electric field and that deployment at sea may result in the electric field dissapating more readily through the sea-bed. In order to test for this we anchored a line of scallops within the frame and lowered the frame to the sea-bed before testing. A piece of 5mm nylon netting (non-conducting) was strung across the base of the frame with line of scallops ~90mm shell height attached at regular intervals so that the full width of the frame was sampled. Attachment proved to be problematic in that the upper shell needed to be free to open and close and the process of attaching the scallops to the netting needed to be as non-invasive as possible. A first attempt to glue a cable-tie to the round shell (using a hot-glue gun) failed as the glue failed to bond adequately. Through trial and error we discovered that small strips of self-adhesive industrial strength velcro (one stuck to the underneath of the scallop, the other stuck to a cable tie though the netting) provided sufficient adhesion for the tests. The frame was lowered to the sea-bed and after a period of acclimatisation (4 out of 5 scallops opened after ~15 minutes), the wave-form generator was triggered and the scallops observed to snap shut. Although the scallops were held on netting rather than semi-rececessed into the seabed, the results of this test gave us reasonable confident that field dissipation through the sea-bed was not likely to be a significant issue.

In further testing the following observations were made regarding the response of different types of organism. Fish were observed to be temporarily stunned and then usually make a fast escape once the pulse switched off. Crustaceans exhibited a twitch each time the system pulsed, but appeared to continue moving at the same rate after the pulse ceased. Echinoderms were also observed to twitch with the pulses. Queen scallops, disturbed by the approach of the sledge would take flight with their clapping response, only to be stunned by the field and fall back to the sea-bed. Upon cessation of the pulse the queen scallops appeared to recover over a period of some minutes. Queen scallops which had not taken flight appeared to be temporarily paralysed by the pulse with their shells open, although a twitching response could still be seen in part of the mantle. Buried fauna, whilst not exiting from the burrow were observed to emit a column of sediment once pulsing had ceased.

The wooden drop-frame, whilst functioning correctly, was not able to cover enough ground quickly enough to make it a viable option for surveying scallops, particularly given the low density of scallops typically observed in this region. The move was therefore made to the modified sledge. During the 2010 survey the pulse generator, having been moved to the sledge encountered technical difficulties which resulted in several burnt-out components, however the system was observed to be functioning as it elicited a range of motility responses in different organisms. Unfortunately, the one species the system was designed for, Pecten maximus was not observed to respond to the passing of the sledge. Following this disappointment and the technical difficulties with the circitury, up-rated components were installed which would take a higher voltage. The revised pulse generator was then tested on the final 2011 survey and a range of voltages used, but again whilst a wide variety of organisms exhibited a response, no Pecten were observed to respond to the field.

RV survey conclusions.It was anticipated that using the wealth of data collected regarding the abundance and distribution of different species recorded on the scallop grounds that we would be able to describe the characteristics of scallop nursery grounds even though the capture rates of juvenile scallops might be low. The reality is that the capture rates of small scallops were too low to effect any meaningful analysis. The refinement of the sorting technique for the 2m beam trawl catches in the 2011 survey did result in significantly more numbers of juvenile Pecten being found although it is also possible that this coincided with a higher level of recruitment for that year-class. If an index of potential Pecten recruitment were required for fishery management more than 2 years prior to them entering the fishery then this is the only method which would deliver such a metric. Spat collection from natural sources (rather than the placing of spat-collectors) relies upon fairly precise timing of surveying due to the short time that Pecten spat remain adhered to the emergent epifauna.

The lack of response from Pecten in the wild to the electrical field stimulation was disappointing. Whilst the methodology was proved to be successful in the tank and in the field using tethered individuals, the act of placing the electrodes on the sledge appeared to illicit no response. This may due to the scallops reacting in advance of the arrival of the sledge either through physical disturbance, or the effective electrical field extending beyond the effective field of view of the video camera. We therefore conclude that the use of electrical stimulation to enhance the detection rate of Pecten is not viable, however the reaction of several other organisms, particularly crustaceans may be of use in other situations. Female brown crab (Cancer paguras) are thought to bury themselves and become both cryptic and static when brooding eggs. This may be occurring on grounds used for other purposes such as dredging or aggregate extraction. The use of the electrical stimulator in conjunction with a TV survey may therefore help to detect the presence of such crabs which would otherwise be very hard to detect and sample.


2. Modelling2.1. Hydrodynamic modelling (objectives 2.1, 2.2 and 2.5)

In order to establish larval re-distribution patterns, a high resolution hydrodynamic model of the English Channel was developed. Previous models of the channel had been developed within Cefas but operated at the level of ~2 nautical miles which was not felt to be adequate to capture the dynamics from some of the smaller spatial management measures in force for scallops. The General Estuarine Transport Model (GETM) was used to simulate the hydrodynamics, and is freely available from www.getm.eu and described in Burchard and Bolding (2002) and Stips et al. (2004). The GETM model uses the General Ocean Turbulence Model (GOTM) for description of the vertical processes (see www.gotm.net). GETM is a fully baroclinic model including sea surface elevations, currents, drying and flooding, temperature and salinity. Forcings include meteorological conditions, riverine fresh water input and boundary conditions for elevation, temperature and salinity.

The GETM model was set up and run for the English Channel for this project, with a spatial resolution of 0.00833 degrees (~ 0.6 x 1.0 km). Initial plans for the model were to only run for western end of the Channel, but the first model runs were confounded by so many issues surrounding the boundary conditions that the only realistic solution was to extend the domain of the model to encompass the whole channel. Fortunately this coincided with the arrival of a new computing cluster that made the computation time for this larger domain possible. One key difficulty in getting the model to run at this resolution was obtaining appropriately scaled bathymetric data, particularly in coastal regions. Bathymetry for the model domain was created using GEBCO_08 and NOOS information. The GEBCO_08 Grid (released: January 2009, updated November 2009) is a global 30 arc-second grid generated by combining quality-controlled ship depth sounding with interpolation between sounding points guided by satellite-derived gravity data (Figure 5). The NOOS bathymetry is an enhancement for shallow coastal areas of the original GEBCO 1 minute grid through local knowledge and datasets from MEES institutes (POL, DMI, BSH, SHOM, RIKZ) , and is therefore at a lower resolution than that desired for the English Channel model. In-house data (collected on RV sruveys) relating to bathymetry for the region were therefore used to enhance and improve resolution, particularly in coastal regions. One section was even surveyed overnight during the 2010 survey, collecting data to within 500m of the shoreline. The resulting high-resolution hydrodynamic model was validated and calibrated using tidal height data. The model outputs required to run the particle tracking model were the hourly values of the three dimensional flow fields. Given the resolution of the model and the size of the overall domain used, the flow field files were very large and only ~5 years of output could be stored. It is not feasible to run the GETM model each time a new year's data are required as run time is around 4 days per year. GETM was run for a full year before results were used for particle tracking, to allow for proper spin up of the hydrodynamics. The used GETM domain was larger than that used by the management model or the particle tracking model (which use roughly the same domain), in order to describe the flow conditions in the entire Channel and to avoid boundary issues.

The particle tracking code GITM (General Individuals Tracking Model) is based on a passive particle tracking model by Wolk (2003), which was further developed by Hide Yamazaki (version 0.9.0). Internal development has focussed on different modules to describe biological development and behaviour based on observations for different species. The model allows for multiple egg and larvae stages, temperature-based development, mortality and several types of vertical migration behaviour (e.g. constant, diurnal or ebb-flood based migration). For this project, three stages were included for scallops: an egg stage (lasting two days) and two larvae stages (transition based on size). Larvae settled as soon as they reached a size within the settling size range (0.23 to 0.27 mm). Growth for both eggs and larvae was linear in time, but larvae were able to migrate upwards in the water column with a constant velocity of 0.001 m/s. Mortality was not imposed, so scallops either settled or were advected outside the applied model domain. Development of all stages was time dependent only, so temperature effects were not included.

Objective 2.1 was to use climate change scenarios to force the hydrodynamic model and Milestone 2.2.1 was to obtain relevant forcing data from the Hadley climate model. Forcing data from the Met Office Hadley Centre Regional Model Perturbed Physics Ensemble UK domain run (HadRM3-PPE-UK) was been obtained through the British Atmospheric Data Centre (BADC). The HadRM3-PPE-UK experiment was designed to simulate the regional climate for the UK in the period 1950-2100 for an historical and a medium (SRESA1B) emissions scenario. The challenges of using these data were such that it was not considered practical within the limits of the project and a simpler approach was envisaged in which the boundary temperatures of the GETM model would be increased in line with the changed predicted by the HadRM3-PPE-UK run. Unfortunately the final runs of the GETM and GITM models using current climatology were significantly delayed which precluded the experimentation with temperature forcing.

2.2. Management modelling "Spatman" (objectives 2.2, 2.3 and 2.4)The model developed for the tracking of scallop populations and their fisheries is built upon the model designed under the previous Defra contract MF0229. The model is a scenario testing platform in which


different, spatially resolved management regimes can be tested for their impact upon the long term performance of the fishery and the stock. By setting the spatial structure of the model to match, as closely as possible, the physical characteristics of a real environment and using locally derived parameters for the populations and the fisheries, the model should be able to evaluate the impact of specific management actions for local populations and fisheries.

The model is written in C++ and is highly object-orientated, using both individual based approaches (for following the activities of individual vessels) and more collective approaches (modelling scallop populations as matrices of numbers at age and size). Flexibility in the ability to set up particular instances of the model was a key consideration, so the spatial arena can be defined with any desired size or spatial resolution, time is flexibly handled from steps of 1 day up to 1 year and the model can even follow multiple species should the need arise.

Spatial structuring.

Space is handled as a two dimensional grid of rectangular cells (termed "locations") and each cell holds a list of population units (one population unit per species in the model). In the implementation described here only one species (Pecten) has been used. Locations and the populations within them are essentially independent from neighbouring locations. There is no flow of individuals from one location to another except when larval production is re-distributed and all biological parameters (e.g. those controlling growth, maturity, natural mortality etc) can be determined independently for each population. There is an explicit assumption that within a Location, individuals have a uniform-random distribution.

The flexibility of the model is such that a location can be any dimensions required and as such individual locations may be on too fine a scale for realistic management. Locations are therefore grouped into different Management Units which then represent the level at which Management would collect fishery information and set management actions accordingly. Management Units are defined on a species basis, so if the model is run in multi-species mode, then Management Units for the different species can have different spatial arrangements. The only exception to this is the Locations grouped into a Management Unit called "LAND". This is a special grouping which applies to all species and the Locations obviously have no populations residing within them.

The location of Ports is important to the model as any vessels classed as day boats must return there for each sub-season, effectively limiting their ability to search the exploit the entire arena (as vessels have a maximum range for a sub-season).

Temporal structuring.There are three levels of temporal structure to the model, annual, seasonal, and sub-seasonal. Again with flexibility in mind, the number of seasons and sub-seasons can be set to any desired length, but the parameterisation of fishing vessels as "day boats" effectively forces the sub-seasons to be steps on one day. Seasons are used to allow biological parameters (e.g. growth rates) to vary within a year and are the time steps at which biological process such as growth and maturity are inflicted.

Fishing operations and the measurement against Management plans occurs on the sub-seasonal basis. So after each sub-season the fishery statistics in terms of landings, effort and/or catch rate are compared with any limits imposed by the Management regime (i.e. uptake of landings or effort quota, catch rates in relation to a threshold etc).

The spatial redistribution of larval production which the hydrodynamic model predicts is imported at the start of the year (January 1st). Whilst this is acknowledged to be an arbitrary date, larval production is such that there should be no larvae in the water column at this time of year and therefore the use of January 1st as a hard cut off date allows the effective following of cohort development.

The realised length of a "year" in the management model is the product of the number of seasons and the number of sub-seasons defined for that implementation. In the implementation used here we have used 12 seasons and 30 sub-seasons resulting in a "year" of 360 sub-seasons. The implications of this when integrating with the hydrodynamics are discussed later.

Biological operations.The development of each scallop population is controlled by parameters specific to the location it resides in. The heart of each "population" unit is an array of numbers at size and age but the range of functions and associated parameters which can operate on this array are quite complex. Parameters governing natural mortality, maturity and commercial value are held for each size class and season. In the current implementation these values are therefore fixed between years.

Following the findings of the shore-based sampling, two different spawning events can take place within a year. A sigmoid function, controlled by the day number, provides a probability that spawning will take place on a given day. All spawning activity for that population is considered to take place on a single day, so the spatial resolution of the "Locations" will play a key part in the temporal distribution of spawning releases.


Spawning is controlled by a stock-recruit function (again which can be parameterised for individual locations but in this instance used the same value for all locations) and dictates the number of larval packets released. Each packet is moved around by the hydrodynamic model as a single unit but can represent any number of larvae - see the particle tracking model description for more information. Given the assumption that the distribution of individual spawning scallops is uniform-random, each packet is given a random location from within the boundaries of the Location. Each spawning event is recorded in a file with the location, packet size and day number so that the hydrodynamic modelling can transport the packet to its final location.

Upon recieving the results of the hydrodynamic transport, the packets of larvae are integrated into the population matrix residing in the new location. The number of larvae entering the population will depend upon two factors. The particle tracking module has the facility to model mortality events based upon the oceanographic conditions encountered by the packet (such as adverse temperature or salinity conditions). Each location is also defined with a settlement suitability index (one for each species in the model). This is to allow for the different benthic communities / physical structure which affects the survivability of the settling larvae.

Fishery operations.Unlike the matrix-style of modelling used for the populations, vessels are treated as individual units with parameters governing its ability to search for scallops and its ability to catch scallops. Searching ability is controlled by the time at sea and trip duration as well as how far it will travel in a given sub-season. It is assumed that for each sub-season a vessel is at sea it completes a full fishing operation with one Location, therefore a full day's steaming to/from distant grounds is not currently possible. The actual fishing event is controlled by the spread of the gear carried, the towing speed and the duration of fishing activity in a given sub-season. Trip duration and tow duration can both be varied by season, so weather-based influences can be incorporated.

The multispecies nature of the model requires differentiation in the efficiency of a vessel at capturing the different species, so for each vessel a set of parameters is held for each species governing the selectivity curve, the retention curve, discard mortality and incidental mortality. Incidental mortality measures the probability of death for an individual which encountered the gear but was not selected. Given the assumption of uniform-random distribution within a Location, the probability of encounter between individuals and the gear was simply the ratio of swept-area to total area of the Location.

For each fishing event (a species at a Location in a sub-season) the vessel creates a log-book record. This details when, where and what was caught, allowing each vessel to use its track-record to inform which Location to target for the next sub-season. The decision process regarding which location to move to started by determining the exact Locations which lay in its search radius (i.e. how far can I travel in one sub-season and still fish? The second step was a decision as to whether to choose somewhere from that range at random or whether to use the track-record (where no. The probability of opting to fish at random was set to a fairly low level and is specified independently for each vessel. When opting to use the track record a vessel searches through all of its logbooks looking for those which

a) originate from Locations within the search radius b) which are currently open to fishing c) are for the same season as the current seasond) are no older than X years, where X is a parameter defined at the start of the program.

From this sub-set of log-books, the Location which generated the highest value landings is selected as the one to move to for the next sub-season.

Program flow.The program first defines the spatial and temporal resolution of the model and then seeds each Location with populations as appropriate. The initialisation files specify a number of recruits at each location which is then propagated through the size classes to give numbers at equilibrium for a virgin (i.e. no fishing activity) population. These initial numbers are only used the first time the model is started in a given run and are discarded if the model is being used in a loop with the hydrodynamic model as explained later.

Following initialisation, the program then loops through the seasons and sub-seasons, updating the populations with growth on a seasonal basis and inflicting the mortality on a sub-seasonal basis. Within a sub-season the list of vessels is worked through sequentially, allowing each vessel to fish in turn. This does mean that a vessel fishing first on particular location would have a larger population to exploit than subsequent vessels in that sub-season. In order to minimise this impact the order in which the vessels are handled is randomised at each step. Fishery statistics are checked against management limits for each sub-season and the Management Unit is closed as soon as a limit is exceeded. Each newly spawned packed of larvae is recorded in a file which at the end of the year is picked up by the hydrodynamic model.

At the end of the year, the entire population structure and vessel structure (including the logbooks) are written to file and the program closes. Once the hydrodynamic model finishes and returns the new larval positions,


the Management model re-starts but rather than using the initialisation files to seed the population numbers, it picks up the populations and fishery from the last year and runs for another year.

Implementation.The spatial resolution was set to 0.1 degree of latitude and 0.2 degrees of longitude and covered the western English Channel with the LAND Management area approximating the coastline of Cornwall, Devon and Dorset. The remaining Locations were aggregated into 6 different Management Areas which approximate current delineation of some management authority. (B) covered the Bristol Channel, (C) represented the inshore waters of Devon, (D) the inshore waters of Cornwall, (E) UK waters beyond the 12 mile limit, (F) represented closed areas in Lyme Bay and (G) inshore waters of Dorset (figure 2.1)

-6.00 -5.80 -5.60 -5.40 -5.20 -5.00 -4.80 -4.60 -4.40 -4.20 -4.00 -3.80 -3.60 -3.40 -3.20 -3.00 -2.80 -2.60 -2.40 -2.20

50.90 B B B B B B B B A A A A A A A A A A A A

50.80 B B B B B B B A A A A A A A A A A A A A

50.70 B B B B B B B A A A A A A A A F A A A A

50.60 B B B B B B B A A A A A A C C F G A A A

50.50 B B B B B B A A A A A A C F C G F A G G

50.40 B B B B B A A A A A A A C C C G G G G G

50.30 B B B B A A A A D A A A C C C G G G G G

50.20 B B B B A A D D D C C A C C C G G G G G

50.10 B A A A A D D D D C C C C C C G G G G G

50.00 B A D D A D D D D C C C C C C G G G G G

49.90 B B D D D D D D D E E E E E E E E E E E

49.80 B B D D D D D D D E E E E E E E E E E E

49.70 E E D D D D E E E E E E E E E E E E E E

49.60 E E E E E E E E E E E E E E E E E E E E







Figure 2.1 spatial arrangement of the SpatMan Locations and Management Areas.

Scallop populations were given biological parameters governing growth, weight-length relationships and maturity determined from historic sources, spatially resolved where possible. Natural mortality was set at 0.2 per annum. The discard and incidental mortality were set at 10%. The results of the shore-based sampling demonstrated the existence of potentially more than one spawning event in a given year but the time series of years is not complete or long enough to know if there is any particular periodicity in these changes from single to multiple events. For the purposes of demonstrating the system we have used alternating years of one or two spawning events.

Thirty vessels were defined operating out of four ports (Brixham, Newly, Plymouth and Falmouth) offering a variety of size, spatial range and trip-length characteristics. In order to start the model running with a fully age structured population, SpatMan was run for six years without redistribution of recruits before the hydrodynamic phase was introduced.

To combine the information of the fisheries activities, spawning locations, currents and settling locations the particle tracking code (GITM) and the management model (SpatMan) were linked via a bash script and run in a loop over consecutive years. GETM hydrodynamic results were generated beforehand and used by the particle tracking code when needed. SpatMan provided the initial spawning locations for 2001, using a spin up period of 6 years. The particle tracking code then transported and developed the eggs until all larvae were settled. The settling locations (if within the applied domain) were then provided to SpatMan to determine spawning locations for 2002. This loop was repeated 6 times, up to 2006. Spawning times varied between one and two spawning periods, with twice as many particles released during a double release year. Due to the size of GETM hourly flow field files the GETM model was only run for 2001-2005 as the system could not hold any more.

There then followed a sequence of alternating particle-tracking and SpatMan model runs. Whilst it only too ~29 seconds to run SpatMan for one year of fishery the particle tracking took considerably longer (several hours even when running in parallel on multiple nodes).

Due to logistical constraints the two modelling components were only brought together at the very end of this project and therefore we did not have the time to fully evaluate and refine the settings of SpatMan in order to make best use of the fine scale hydrodynamics. The planned testing of climate change scenarios therefore were not undertaken.

Results.

Although the number of years used in the projection presented here is short, the models have been demonstrated to run as expected and we are able to track not only the development of the stocks and the re-distribution of larvae (figure 2.2 but also see how the fishery might respond to these patterns and changes in them (figure 2.3) . There are some consistent patterns in larval distribution emerging from these preliminary runs. Scallop larvae spawned around the eastern boundary of the domain are either advected westwards or, more likely, outside the domain. This leaves a general lack of settlement east of Portland which ties in with a general lack of scallops in this region (in fact there are few scallops found between Portland and the eastern


edge of the Isle of Wight). Concentrations of larvae are appearing off the south east Devon peninsular with also ties in with known scalloping grounds. There are, however some anomalies appearing in the release sites which warrant further exploration.

a) b)

c) d)

Figure 2.2 Before and after plots of the particle tracking model, colour coded by the Management Areas the particles originated from. a) and b) are the start and end points from a year with two spawning events, c) and d) are from a year with just one spawning event.

a)

Day trip vessels

Year

land

ings

0

2000

4000

6000

2001 2002 2003 2004 2005

vess_1 vess_10

2001 2002 2003 2004 2005

vess_11 vess_12

vess_16 vess_17 vess_18

0

2000

4000

6000

vess_190

2000

4000

6000

vess_2 vess_20 vess_23 vess_24


0

2000

4000

6000

vess_30

2000

4000

6000

vess_8

2001 2002 2003 2004 2005

vess_9

C D E G b)

Multi-Day vessels

Year

land

ings

0

20000

40000

60000

80000

2001 2002 2003 2004 2005

vess_13 vess_14

2001 2002 2003 2004 2005

vess_15

vess_21 vess_22

0

20000

40000

60000

80000

vess_280

20000

40000

60000

80000


vess_5

2001 2002 2003 2004 2005

vess_6

0

20000

40000

60000

80000

vess_7

C D E G

Figure 2.3 Annual landings by vessel, the different colours showing the management areas the landings were made from. a) shows those vessels restricted to day-trips and b) shows landings from those able to undertake multi-day trips.

Discussion.

It must be emphasised that the results shown here simply demonstrate that the dual modelling concept works, but some issues regarding the parameterisation of the two models only became apparent when the two models began to work in tandem. For instance there are a lack of release sites visible in figure 2.1a in the Falmouth region although we know that populations of scallops are definitely in existence there. This turned out to be a function of the relatively coarse grid of locations and the definition of land compared to the high-resolution scale of the hydrodynamics.

The assumption within SpatMan that a vessel can only fish one Location in a sub-season means that, at present, the Location resolution has to match that of typical fishery operations, however a number of alternative options for changing the structure of the program to enable finer Location resolution have already been considered. Finer Location resolution would also make better use of the fine scale larval re-distribution outputs and enable the fishery model to better reflect the patterns observed with monitoring systems such as VMS.

There should, however, be some caution exercised with increasing model resolution. In order for the higher resolution to be meaningful, more spatially explicit parameter estimates would be required. Without


appropriate parameterisation, the increased confidence in the ability of the model to resemble reality may well be unfounded.

The modelling of vessels as individual units means that in theory it should be possible to make the model more applicable to, and more accepted by fishers as they could directly assist with the parameterisation of their fishing vessel. Having said that any model is always a simplification and generalisation of a process and it could be argued that attempting to forecast changes in the operational mode of genuine individual vessels is too complex a task.

3. Implications for Policy/Management Scallop recruitment indices.

Fishery management in which output measures are determined on an annual basis (such as has been in operation for the CFP) requires a prediction of the likely level of recruitment coming into the fishery over the next 12-24 months. At present there is no survey which generates a suitable recruitment index for scallops. This project has demonstrated that surveying for 1 year old scallops is unlikely to generate data of sufficient robustness for management purposes and would be a costly, labour intensive exercise.

Obtaining data regarding spawning patterns.

Spawning patterns for scallops have been demonstrated to vary considerably in time and space. Accurate and realistic modelling of larval transport means that data regarding spawning patterns are required for a wide area and over a long period of time. One or two year snapshots of data are unlikely to be sufficient as they may miss key seasons. Detailed scientific work (including dissection of animals) delivered the most consistent results, but the use of a simple visual key gave broadly similar results. There may be scope for involving industry in some self-sampling provided sufficient training and validation are under taken.

Modelling of larval transport and management.

The concept of combining spatial management evaluation with hydrodynamic models has been demonstrated to work, however some additional refinement is required before the model can be used to test real scenarios.


References to published material9.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.

A.R. Beaumont & D.A. Barnes. Aspects of veliger larval growth and byssus drifting of the spat of Pecten maximus and Aequipecten opercularis. ICES J. Mar. Sci. (1992) 49 (4):417-423

Burchard H, Bolding K (2002) Getm: a general estuarine transport model. Technical Report EUR 20253 EN, European Commission

Lane, D.J.W., Beaumont, A.R., Hunter, J.R. (1985). Byssus drifting and the drifting threads of the young post-larval mussel Mytilis edulis. Mar Biol. 84:301-308

Mason, J., 1958. The breeding of the scallop, Pectan maximus(L), in Manx waters. J. Mar. Biol Ass UK 37, 653-671

P.J. Dare & R.C.A. Bannister. Settlement of scallop, Pecten maximus (L.) spat on natural substrates off south-west England: The hydroid connection. 6th Int. Pectinid workshop, Menai Bridge, Wales 9-14 April 1987.

Stips A, Bolding K, Pohlman T, Burchard H (2004) Simulating the temporal and spatial dynamics of the north sea using the new model getm (general estuarine transport model). Ocean Dyn. 54:266–283. ISSN 1616-7341

Wolk, F., 2003 Three-dimensional Lagrangian tracer modelling in Wadden Sea areas. Diploma thesis, Carl von Ossietzky University Oldenburg, Hamburg, Germany, 85 pp.


general enquiries on this form should be made...

Documents