Dolphin bycatch: observations and mitigation work
in the UK bass pair trawl fishery 2004-2005 season
Occasional Report to DEFRA
Simon Northridge, Alice Mackay & Terry Cross
Sea Mammal Research Unit
October 2005
Introduction.
As in four previous seasons, observers from the Sea Mammal Research Unit, at the invitation of the skippers involved in the UK element of the bass pair trawl fishery, have monitored operations for a substantial proportion of the total UK fishery during the 2004-2005 winter fishing season. At the same time we have continued to work with the same skippers to try to minimise the bycatch of common dolphins in this fishery.
Previous work had suggested that exclusion devices could be effective in lowering bycatch rates. In 2003 (late February to April) we had shown that one boat using a steel exclusion grid and equipped with an escape hatch had a substantially lower dolphin bycatch rate than other boats working in the same vicinity. Among 60 observed tows using a grid there were no recorded dolphin mortalities, while 26 dolphins were recorded in 53 observed tows in other boats. (Two dolphins were reported by the skipper as caught among about 20 unobserved tows by the boat equipped with the grid). In the winter season of 2003-2004 we compared three different grid types and escape hatches and found that one combination (a steel grid with a relatively large escape hole and a small meshed cover net) was associated with a 60% lower bycatch rate than other devices. It therefore seemed that exclusion devices were associated with lowered bycatch rates in previous seasons. However, we had not observed any dolphins actually using the escape hatch, largely due to the absence of an effective video monitoring system in the 2003-2004 season.
The main objectives of this season were to monitor and document bycatch rates, to continue to develop an underwater monitoring system to enable us to monitor escape hatches in the net, and to modify escape hatch designs developed over the past two seasons to improve their effectiveness. The mitigation measures and video monitoring developments were pursued under the project NECESSITY, funded by the European Commission with additional funding from the US National Marine Fisheries Service.
One major change between previous seasons and the 2004-5 season was the introduction of a regulation banning the use of pelagic pair trawls directed at bass within the UK 12-mile limit. This restriction came into force in January 2005. The geographical distribution of fishing effort by the two pairs of vessels involved in the fishery this season may have been altered compared with previous seasons, at least after the introduction of this regulation, but this is difficult to assess.
Summary of bycatch observations in the 2004-2005 winter season.
Observers joined the boats in October and, with a few periods of absence, remained with the boats until the end of the season in April 2005. On 35 occasions observers monitored a trawl operation in which one or more dolphin was observed dead. A total of 164 tows were considered adequately observed (for 12 of which the primary target was anchovy and a different net was used), and 95 dolphins were recorded as having died in these fishing operations. No cetacean bycatch was recorded in the anchovy tows. The overall bycatch rate was therefore 0.579 dolphins per tow, or 0.625 excluding anchovy tows. This compares with an overall rate of 1.29 in the 2004-2005 winter season.
Table 1: Summary of observations of dolphin bycatch
in the bass fishery during 2000-2005
Season / Mortalities / Hauls / Bycatch rate2000-2001 / 52 / 91 / 0.57
2001-2002 / 9 / 91 / 0.10
2002-2003 / 26 / 113 / 0.23
2003-2004 / 169 / 131 / 1.29
2004-2005 / 95 / 152 / 0.63
Weighted average / 0.56
Totals / 351 / 578 / 0.61
Figure 1: Seasonal changes in bycatch rate
Bycatch rates were highest in late January and early February, and decreased with increasing distance from shore. In the 2003-2004 season bycatch rates peaked earlier in the season, in December. This was in contrast to the preceding years (2001-2003) during which observed bycatch had been almost restricted to late February and March. This season’s observations might therefore be seen as tending towards a return to the previously observed seasonal pattern of bycatch. It is hard to draw any firm conclusions about bycatch rate in relation to distance from shore, as fishing tends to move further offshore as the season progresses and bycatch rates by distance zone are therefore confounded with seasonal patterns, and vice versa.
Figure 2: Distance from shore
The 95 recorded mortalities were observed among 35 tows, and the average group size in these tows was therefore 2.71. This is lower than in previous seasons (average 4.4). The reason for this is not known.
Estimates of total mortality
The total amount of fishing effort in the UK offshore pelagic pair trawl fishery for bass can be estimated from logbook records in the Fishing Activity Database by identifying fishing trips by vessels using pelagic trawl gear and which make landings that are dominated by bass. Two potentially useful measures of fishing effort are available from these data, these are the number of days spent at sea and the number of fishing operations.
As bycatch rates are generally described in terms of the number of animals per fishing operation, in order to multiply up the observed number of bycatches to an estimate of the total bycatch, it would be most useful to have a count of the number of fishing operations. This statistic is recorded in fishing logbooks, but it is not a mandatory field, and when it is left blank it is always estimated by fishery officers in the port of landing when logbook entries are entered into the Fishing Activity Database, based on their experience of how many tows boats are likely to make per day.
We have identified some discrepancies between those data on tows recorded in the Fishing Activity Database and tows noted by observers on board vessels. We have therefore examined the FAD logbook data and our own observations on a trip by trip basis throughout the bass fishing season, and have tried to estimate the actual number of tows made by the fleet in the 2004-5 season, based on our own observations in the majority of cases, and our estimates of the number of tows per day at sea likely by each of the pairs involved, again based on our observations. In this way we estimate that there were about 223 fishing operations made by this fishery in 2004-2005, of which 152 or 68% were adequately monitored for dolphin bycatch. Effort in the 2004-5 season was clearly very much lower than in the previous season, when the FAD records indicated there were in excess of 400 fishing operations.
The estimated total mortality, based on these figures, and with 95% log-normal confidence limits is between 90 and 207 animals (point estimate 139). These figures are summarised below.
Table 2: Estimated total mortality of common dolphins
in the UK offshore pair trawl fishery for bass 2004-2005 season
No of observed hauls / 152Observed bycatches / 95
Mean bycatch rate / 0.625
Variance of bycatch rate / 2.75
Estimated total bycatch / 139
CV of estimate / 0.22
Upper 95% CL / 207
Lower 95% CL / 90
This figure compares with estimates of more than 400 animals for last season (though we have not gone back to analyse thoroughly the effort data for the 2003-2004 season which may also have been over-estimated to some extent). What is nevertheless clear is that both bycatch rate (per tow) and overall bycatch numbers – which is a function of both the rate and the fleet effort - were substantially down on the previous season.
Mitigation work
Work began by testing a 20cm mesh net panel inserted into the mid section of the trawl, which was intended to keep dolphins out of the small meshed part of the net where they are vulnerable to getting caught. This proved unsatisfactory because, although no dolphins were caught (none was expected either so early in the season), the extra netting increased drag on the net considerably, and fish catches were very low. Once the net panel was removed fish catches picked up considerably.
Figure 4: Potential barrier for dolphins –200mm dyneema netting
Following on from proposals from the previous season’s work, we devoted considerable time to the development of a robust video monitoring system. Whereas a useful underwater video system had been developed in the 2002-2003 season, this had relied upon the availability of a 38m vessel with a suitable deck winch to run a steel re-inforced cable to a camera situated at the rear of the net. Following the withdrawal of this vessel from the fishery, we began work with a pair of 15m vessels in 2003, neither of which had a suitable winch or sufficient deck space to manage the system that we had developed. We made several attempts to overcome this in the 2003-2004 winter season, but were left without a functional monitoring system at the end of that season.
We planned this season to make use of the spare net winches on the lower deck of the two vessels, which would not normally be considered suitable for an instrumentation cable. However, by purchasing suitably rugged video cable and developing appropriate cable junctions we were able to develop and run a video monitoring system. The monitoring system worked well for the duration of the season, though we needed to make repairs to parts of the cable during the season, and by the end of the season one of the steel video cables showed considerable wear. Observers and crew were able to monitor the grid and escape hatch from the wheelhouse.
By late November, after testing the dyneema exclusion panel, we reverted to using the stainless steel tubular grid with small-meshed sandeel netting as a cover for the escape hatch. This was placed towards the front end of the extension piece as before, with a camera monitoring the escape hatch about 3 metres in front of the grid. Thirteen dolphins became caught using the grid in this place, but none of them approached the grid. We concluded that the grid was still too far back in the trawl and moved it forwards by some 30m to a section of the net where the expected width of the net is about 9m. Previous observations of 152 dead dolphins for which we had recorded their location in the net showed that only 7 such animals (4.6%) had been recovered forwards of this point.
We installed a flexible grid at this point using the same sandeel netting as a cover hatch. The grid was much narrower than the inside diameter of the net and consequently some distortion of the net resulted at this point, with the net ‘ballooning’ around the grid.
This grid and cover combination did not perform well at this point of the net, as the flexible grid became distorted, and fish became entrapped around it. More significantly, because we had now implemented an escape hole in an area of the trawl in which the top part of the net slopes down (i.e. the net is tapering at this point), the cover net tends to rise up towards the horizontal plane leaving a gap around the mouth of the exit. This allowed fish to escape, and required weights (chain) to be added to the trailing edge of the cover net to keep it down and ensure the escape hole is covered.
Despite these drawbacks we observed 5 animals escaping through the escape hatch. However, the crew and observers reported that some effort was clearly required to lift the cover net in each case, apparently due to the weight of the netting itself and the attached chains. Over the same time period a further 18 animals died and were recovered from the net; most of these were well in front of the grid and escape hatch in larger meshed sections of the net (mesh sizes 100 & 200 mm).
The flexible grid and the heavy, small meshed sandeel net were then replaced with the steel grid at the next net section, about 10m further forward again, with a piece of much lighter 20mm mesh purse seine netting that only required a length of light leadline to keep it closed. This arrangement fished better than the preceding one but was still not ideal, as the net was clearly ballooning in front of the grid, which constricts the width from the expected 15m to about 1.5m.
A further 4 dolphins were observed to swim through the escape hole, and it was reported that their exits appeared to be much easier than in the preceding set-up. Nevertheless, 15 further animals were also caught, most of them again some distance in front of the grid, and yet further forward in the net
The observations of animals at the escape hatch made by our observer and by the crew of the boat concerned show that the animals are alive and freely able to push through an escape hatch. Although only 9 animals were seen to escape, this is a minimum number as some escapes were reported to happen very quickly and could have been missed and because the camera was not always fixed with all of the escape hatch in view due to movements of the housing.
A further 32 animals did not escape, but were caught much further forward in the net than normal, suggesting that they had either stopped before going too far into the net, presumably because they detected the grid, or that they had reached the grid and turned around but had not found their way back to the shark’s teeth section of the net where they might escape. We did not see any animals approach the grid and then turn back, but the camera is so close to the grid and the escape hole that this is perhaps not surprising. It was clear that, unlike the previous grid tests in 2003, when the grid is moved so far forward, fish losses are substantial and will need to be reduced by modifying the escape hatch. Several ideas are under currently development.