Collection of Data to Inform the Implementation of a Discards Ban: Scottish Marine and Freshwater Science Vol 7 No 12

3 Methods

This project used self-sampling by fishermen to collect samples of the fish being discarded by vessels in the Shetland whitefish fleet. The self-sampling methodology was based on that developed and used in Dutch fisheries (van Helmond et al., 2012; Uhlmann et al., 2013).

3.1 Data Collection

3.1.1 Vessel Self-Sampling

Fishing crews willing to participate in the self-sampling programme were identified by the Shetland Fishermen's Association. Participating crews were asked to retain a representative sample of the fish that they were discarding from some of their tows; ideally two catches each week.

Members of each crew were briefed on the objectives of the project and about what they were being asked to do. Each vessel was issued with sample record sheets (an example record sheet is included in the Appendix) on which to record information about each discard sample and the tow from which it was taken. Each sheet bore a random number between 1 and 10 and the crew were asked to sample the tow that corresponded to that number. (For example, if the first sheet bore the number '7' they should sample the 7 ^th tow of their trip. If the next sheet bore the number '3' they should then sample the 3 ^rd tow after that). Despite this protocol being laid out, the randomised sampling methodology was not followed by the fishermen in a high proportion of cases.

For each sampled tow the crew were asked to collect two boxes of the fish that were being discarded (stress was laid on the importance of the sample being as representative of the discards as possible). These boxes were to be iced, labelled and tagged and stored in the vessel's fish hold.

Derogations were provided by Marine Scotland Compliance ( MSC) for each participating vessel to cover any undersized or other fish, the retention of which on-board would normally be prohibited. MSC also provided numbered tags to mark the boxes of samples.

Fishermen were asked to record relevant information on the sample record sheet, including the date and fishing ground, and the number of boxes retained and discarded from the sampled tow.

The samples were landed with the rest of the vessels' catches to the fish markets in either Lerwick or Scalloway.

A provisional target of 72 samples was set at the beginning of the programme, with 24 each to come from seine net, single trawl and twin trawl vessels.

3.1.2 Sample Processing

Samples were processed at the fish market where they had been landed. Each sample was sorted by species; the principal commercial species were individually measured (overall length); and each species was weighed (Table 1).

Following processing the material from the samples was bagged and disposed of at the Energy Recovery Plant (incinerator) in Lerwick. During periods when this plant was closed they had instead to be disposed of at the Gremista Waste Management Facility landfill site, also in Lerwick.

Table 1: Summary of sampling protocol for fish species in the fishermen's discard samples.

Species		Sampling Protocol
Cod Haddock Hake Lemon Sole Ling Megrim Monk Plaice Saithe Whiting Witch Rays Common Skate		Total weight Lengths
Other Species	-	Total weight

3.2 Data Analysis

All data collected were entered into Excel spreadsheets for collation and analysis. Statistical analyses were carried out in Excel using the Real Statistics Resource Pack add-in ^[3] .

The Kruskall-Wallis test was used to compare the discard composition between fishing gear types.

3.2.1 Estimation of Discard Rates

Overall discard rates for the sampled catches (for all species) were determined using the information recorded by the fishermen on each sample record sheet. This included the number of boxes that were retained (landed) and discarded from the sampled catch. The discard rate was calculated as the quantity of fish discarded divided by the total quantity caught (retained + discarded).

It was not possible to directly determine discard rates for individual species as the quantities of each species caught in the sampled catches was not known. The total quantity of each species discarded could be estimated by raising the weights in each discard sample to the level of the catch (the size of the sample and the total quantity discarded from the catch were known). However, the composition of the retained portion of the sampled catches was not recorded. (Whilst this information would have been desirable it was decided not to risk overloading fishermen by asking them to record too much information. This was information that observers would have collected had they been available).

However, although the composition of the retained portions of each catch were not known, the composition of the landings from each sampled trip were available (from data provided by the Shetland Fish Producer's Organisation, covering to the end of December 2014). Overall discard rates for individual species were therefore estimated for each sampled trip by raising the sample data to the level of the trip, as follows:

The total quantity of fish landed at the end of the trip was known (from the SFPO data), as was the total quantity of fish retained (landed) from the catches sampled during that trip (from the sample record sheets). From these, the percentage of the total landings that came from the sampled catches could be calculated, as well as a raising factor.

This raising factor was then used to raise the estimated total quantities of each species discarded from the sampled catches (see above), to an estimate of the total quantities discarded during the trip. The total quantity of each species caught during the trip could then be determined as the sum of the total quantity landed (known) and the quantity discarded (estimated). The discard rates for each species were then estimated from these values.

These estimates of the discard rates for individual species are based on the assumption that the catches sampled, and the discards from them, are representative of all the catches made during the trip. In the absence of observers, or of any other sources of information, it is not possible to verify this assumption of the estimated discard rates for individual species. These estimates therefore need to be treated with some caution, but remain the best possible from the available data.

3.2.2 Comparison with Other Discards Data

In the absence of observers, no data were available to allow for direct verification of the discard sample data. Instead these data were compared to data from two other programmes:

3.2.2.1 Discard Tally Book Data

Between June 2013 and July 2014 the Shetland Fishermen's Association ( SFA) in conjunction with the NAFC Marine Centre used tally books to collect information from Shetland whitefish vessels on the nature and scale of their discards (Napier, 2014). Participating vessels were asked to record, for each catch, estimates of the total quantities of each species retained and discarded. This information was used to estimate the composition of the discards and the discard rates for each species.

The tally book scheme recorded information on 1,513 catches over 422 days of fishing, by eight fishing vessels (2 seine net, 3 single trawl & 3 twin trawl) over varying periods between June 2013 and July 2014, during which about 46,000 boxes of fish of more than 24 different species were caught. Several of the vessels that participated in the tally book scheme also collected discard samples in this project.

The composition of the fishermen's discard samples were compared to those estimated through the discard tally book scheme using the Spearman Rank Correlation test, while the discard rates were compared using the Mann-Whitney U test.

3.2.2.2 Observer Data

Observer data on discards were available from an ongoing project to collect information on data-limited species in the northern North Sea ^[4] . This project used fisheries observers on commercial fishing vessels to collect fisheries and biological data on selected whitefish species (hake, lemon sole, ling, megrim, monk and plaice). Of relevance to this study, these data included the quantities of each species retained and discarded (from which discard rates could be estimated), and the lengths of discarded fish.

Data collected during the period from June to December 2014 were used for comparison with the results of the analyses of the fishermen's discard samples. These included data on 571 catches during 18 trips (usually only one or two species were sampled from each catch). To avoid the effects of any possible seasonal variations in the discards data comparisons between the fishermen's discard sample and observer data were made by quarter.

The size distributions of the fish in the fishermen's discard samples were compared with those measured by the observers using the Kolmogorov-Smirnov test.

3.3 Implications of the Discards Ban

Two potential implications of the discards for fishermen were considered.

Firstly, the potential costs of handling and disposing of fish that fishermen would be required to land under a discards ban, but which they could not sell (because it fell below minimum landing sizes). The discards data collected through this study together with landings data for the Shetland whitefish fleet ^[5] were used to estimate the total quantity of unmarketable discards caught by the fleet in 2014. (The quantities estimated from the fishermen's discard samples were raised to the level of the fleet for the year). The potential costs of disposing of this quantity of material were then calculated.

Secondly, the potential impact of the discards ban on quota uptake; in particular how much quota for other species might remain uncaught if fishing had to stop when one quota runs out. The potential magnitude of such uncaught quota was assessed by estimating the dates that quotas available to the Shetland whitefish fleet for selected species might have run out in 2014 under a discards ban, and the total quantities of quota that would have remained uncaught on those dates (based on actual landings data and quota availability for the fleet ^[5] ).