Bird stomach contents analysis - final report: Goosander and Cormorant diet on four Scottish rivers 2019 to 2020

7. Future research requirements

1. Current information on the dietary composition and sizes of Salmon taken by Goosanders and Cormorants is limited by the small sample sizes of birds and rivers. Research involving larger samples of birds at various times of year, and from a larger number of rivers, is required to quantify (temporal and spatial) dietary variation and to improve levels of confidence in the evidence for the possible effects of predation by fish-eating birds in relation to the protection of Salmon stocks in Scottish rivers.

2. The possibility of 'mismatch' between the timing of bird sampling and the timing of biological events, such as the smolt run, raises issues over the comparability of samples in relation to the timing of their collection. This is currently poorly understood and so the implications of such temporal mismatches in sampling dates require investigation. So too do the likely levels of inter-annual variation in diet assessments from bird samples from the same location taken in different years.

3. The implications of bird sampling regimes (including their timing) to assess fish-eating bird diet requires further investigation. This should include fine-scale analysis of temporal assessments of dietary composition to better understand over what time-periods dietary assessments are representative, to examine the influence of sampling location on these assessments, and the likely levels of inter-annual variation in diet assessments from bird samples collected from the same location in different years. Such information could then be used, for instance, to plan a more intensive sampling programme for evidence gathering of bird diet in relation to the smolt run on Scottish rivers (including new information provided from fish trapping and tagging studies).

4. Assessments of general diet from samples of bird stomach contents are currently hard to compare statistically. For each sample, such assessments are the sum of the mass of each fish species recorded, which will either be zero or a positive value. As zero values contribute to the total sample size for statistical tests, simulation studies are needed to understand what sample size, and what number of positive values within a sample, might be most appropriate for robust statistical comparisons.

5. Fine-scale data analysis is required to better understand the influence of both sample location and sample size of birds on subsequent length frequency distributions for prey fishes, particularly Salmon, recorded in stomachs.

6. Any individual bird may opportunistically capture smaller (or larger) than average fish when foraging or, alternatively, be feeding in a location where such fish occur. Fish recorded in stomach contents are therefore not necessarily independent samples. Larger datasets could be used to explore whether statistical analyses should take account of the stomach as a 'sampling unit' – for both assessments of general diet and for estimates of length frequency distributions for particular prey species.

7. It is vital to know where birds are actually foraging, rather than where they are sampled for subsequent diet assessment. Further research (including the tagging and tracking of birds) is thus needed to better understand the spatial and temporal distribution of fish-eating birds within river systems, their (seasonal) movements and behaviour, and levels of fidelity to different locations (e.g. foraging sites, roosts) at different times of year.

8. An analysis of all available data on Goosander and Cormorant numbers and distribution on Scottish Salmon rivers, coupled with a programme of more standardised counts, would help determine the spatial and temporal distribution of birds in relation to both sampling regimes to assess bird diet and future management actions to protect fisheries.

9. Further dietary data are needed for Goosanders and Cormorants from times of year outside the spring smolt run period. For both bird species, an improved understanding is needed of seasonal variations in the relative proportions (and sizes) of juvenile Salmon consumed throughout the year. This would help to put predation during the spring into broader context and improve our understanding of fine-scale temporal variation in diet and any implications this has for the degree of mismatch between comparative samples (see 2).

10. An improved understanding is needed of the relationship between the length-at-age distributions of juvenile Salmon (e.g. 'resident' parr, 'mobile' pre-smolt parr, smolts) in river systems (mainstem and tributaries) and the length frequency distributions of those individuals consumed by Goosanders and Cormorants. Inevitably, this is linked to improved understanding of both the timing and length of the smolt run period, the efficiency and wider representativeness of smolt trap data, and to the temporal dynamics and spatial distribution of juvenile Salmon in the mainstem and tributaries over this period. This understanding should also be improved for other times of year, for instance the period when larger parr move downstream in autumn. Overall, further information on fish abundance, behaviour, and vulnerability to predation (mainstem and tributaries) is needed, along with year-round population modelling at the catchment scale to determine whether (and to what degree) bird predation is a contributory factor to the demographic rates of fish.

Acknowledgements

Many thanks to Simon Dryden for initial discussions about this work. Samples were collected and delivered by staff from the River Dee Trust, Nith District Salmon Fishery Board, the Nith Catchment Fishery Trust, The Tweed Foundation, and the Spey Fishery Board. Jamie Urquhart, Jim Henderson, Debbie Parke, James Hunt, Roger Knight, Brian Shaw, and Jason Hysert provided local background information throughout this work. Rob Brooker, Duncan White, and Lucinda Robinson helped with temporary lab access at The James Hutton Institute to enable the timely processing of some of the samples. The four river catchment maps (Figures 3-7) were kindly produced by James Hunt. James also provided information (and interpretation) from the Gala Water smolt trap in the Tweed catchment. Dina Sadykova, environmental statistician at UKCEH, helped with data analysis, provided statistical advice and computation, and produced the boxplots and ecdf plots presented in this report. Kate Randall, graphic designer at UKCEH, reformatted the full content of this report to ensure it complied with the Scottish Government's accessibility standards for publishing. Earlier drafts of this report benefitted from comments from Marine Scotland and Francis Daunt and Richard Pywell at UKCEH. Heartfelt thanks to all.