Report on razor clam surveys on Tarbert Bank

Executive summary

This report describes a survey carried out on Tarbert Bank in 2023 to estimate the densities and sizes of razor clams, Ensis siliqua and Ensis magnus. The survey was conducted as part of the Scottish Government’s electrofishing scientific trial (Scottish Government, 2017). A combination of commercial electrofishing rig and towed video was deployed using the fishing vessel ‘Skye’. The video recordings were subsequently analysed for the number and sizes of razor clams observed on the surface of the seabed following passage of the electrofishing equipment. The count data were converted to area densities (numbers of razor clams m^-2) in three size classes (small < 100; medium ≥ 100 and < 150; large ≥ 150 mm shell length) based on estimates of the swept area.

Ninety-five tows were completed down to 22.5 m water depth (below sea surface). The maximum depth which could be surveyed was determined by the lengths of the electrofishing cables on the vessel, rather than the video cameras which are equipped with 100 m of cable. Fifty-five of the tows were conducted to the north of latitude 56ºN and forty tows to the south. The importance of this division is that under the Scottish Government’s electrofishing scientific trial, fishing for razors is only permitted north of this division so the southern sector has not been fished for around five years. It was hoped that surveying the fished and presently unfished sectors would provide an interesting comparison. Surveying commenced on 16 February 2023, but sea conditions then deteriorated. Surveying re-commenced on 25 February and continued until 3 March 2023. In addition, a day was spent conducting a catchability experiment on 4 March 2023.

The equipment worked well with up to 15, but typically around 12 video tows completed each day. Image quality on the recorded video was clear and detailed, allowing identification and measurement of objects on the seabed. Fishing vessel ‘Skye’ has been built with electrofishing in mind and several features, such as the aft derrick and raised platform, made deploying and recovering the electrofishing and video rigs especially easy and safe. The cabin is spacious, providing plenty of room for the recording equipment and good visual access to the chart and depth plotters.

During the survey water temperatures were between 8.4 and 8.9°C and salinities between 33.7 and 34.2. Average towing speed was 3.7 ± 0.9 m min^-1 (mean ± std dev) so that most of the exposure times were longer than 30 seconds. Estimated swept areas averaged 93 ± 27 m² (mean ± std dev). From the video recordings 14,563 individual E. siliqua were identified and 12,350 measured. Low numbers of E. magnus (formerly Ensis arcuatus) were also noted in areas of coarser sediment with 570 being measured.

Considering all sizes of E. siliqua, the average density was 1.57 m^‑2 ± 0.11 (mean ± SE) with a maximum of 6.68 m^‑2. The average density in the presently fished area (north of latitude 56ºN) was 1.87 m^‑2 (± 0.17 SE) and south of 56ºN was 1.16 m^‑2 (± 0.11 SE). The commercial fishers tend to target E. siliqua which are longer than 150 mm as these fetch the highest price and mean densities for this size group were 0.82 ± 0.05 m^-2 (mean ± SE) in the presently fished sector and 0.84 ± 0.10 m^-2 (mean ± err) in the presently unfished sector. The size distribution of the razor clams north of 56ºN suggested the presence of three length modes while in the southern sector two modes were apparent. The size distributions in the southern and northern sectors were statistically different with larger E. siliqua being dominant in the presently unfished area, although at a lower average density. The spatial distribution of medium (≥100 mm and < 150 mm shell length) and small (< 100 mm shell length) E. siliqua was notably different between the areas north and south of 56°N. These size groups were recorded at higher densities in the presently fished area. For medium sized E. siliqua, the average density was 0.80 ± 0.12 m^‑2 (mean ± SE) to the north of 56°N and 0.26 ± 0.04 m^‑2 (mean ± SE) to the south. For small sized E. siliqua, the average density was 0.26 ± 0.06 m^‑2 (mean ± SE) to the north and 0.06 ± 0.02 m^‑2 (mean ± SE) to the south of latitude 56°N. Small sized E. siliqua were predominantly found at higher densities around the northern periphery of the bank. This notable spatial pattern could be related to local hydrodynamics, although a possible fishery effect cannot be ruled out.

There were no very clear trends relating E. siliqua densities to water depth although there was a tendency for higher densities of small and medium sized razors to be found on tows deeper than 18 m. This reflects their spatial distribution mentioned above. In contrast, E. magnus tended to be occur on deeper tows reflecting their preference for coarser shelly sediment which is found around the edges of the bank.

Other organisms seen on the videos included crabs (likely Carcinus maenas), hermit crabs (unidentified species), common starfish (Asterias rubens) and brittlestars (likely Ophiura ophiura). A few juvenile flatfish (probably dab or plaice) were observed and several sandeels (Ammodytes spp.) were seen especially on tows over finer sand. In contrast to the survey conducted in Largo Bay (Fox 2021), eider ducks (Somateria mollissima) were not observed in the area. The survey vessel skipper suggested that although this species was occasionally seen on Tarbert Bank, they were generally in low numbers.

Although in previous surveys the efficiency of the electrofishing equipment has been assumed to be high, this has not been scientifically confirmed under field conditions. A series of depletion experiments were conducted on the last working day to measure the proportion of razor clams which emerge with increasing electrical exposure time. For three of the replicates, the proportion of razor clams which emerged following the first 30 seconds stimulation was as low as 0.5 but for the remaining 15 replicates was above 0.75. On average 82.0% of the total razor clams emerged during the first 30 seconds of electrical stimulation, with 11.9% following the second stimulation and 0.1% following the third stimulation. Based on the estimated exposure times during the survey tows, average catchability should be at least 82%. Future surveys could use slightly slower towing speeds which should ensure catchability closer to 100%, although this would be at the cost of slightly less ground covered per unit time. Furthermore, based on fishers’ anecdotal evidence, catchability is thought to be lower in cold water and repeating the depletion study during the summer months might show higher efficiency at the speeds used in the present survey.

The data collected on Tarbert Bank provides a baseline with which to compare Ensis densities and sizes from future surveys. Comparison between the presently fished and unfished sectors of the bank suggests that the fishery has resulted in a statistically significant reduction in the proportion of larger razors north of latitude 56°N. However, there may also be differences in the extent of optimal habitat and growth conditions between the two areas because both the maximum and average densities of E. siliqua seem to be lower in the unfished area. The finding that higher densities of small and medium size E. siliqua were mainly confined to the northern part of the bank may suggest an important role for local hydrography in determining recruitment dynamics.