Executive Summary

• Aims. These were to:

(a)develop protocols to assess and validate known methods of surveying and monitoring that differentiate between red and grey squirrels and where possible reveal population trends;

(b) investigate possible new survey and monitoring methods and, if feasible and effective, develop protocols as above;

(c) make recommendations on the design of survey and monitoring programmes in the UK.

• Study period and sites. The study was carried out over two years: Year 1 - 2005, Year 2 - 2006 at 3 field sites in each of Northern Ireland, Wales, northern England and Scotland by experienced field researchers.
• Methods. In Year 1, a standard programme of work was carried out between April and August in each field site involving replicated visual counts and hair tube surveys with standard live trapping surveys carried out as a means to validate the results from the indirect methods. The power of the sampling methods was considered at the outset. In Year 2 a new technique was trialled between April and November, namely baited visual transects, again with standard live trapping surveys to validate the results from the indirect method. The method of using visual baited transects was carried out by walking each line, stopping every 50 m (0 m, 50 m, 100 m etc), and throwing (scattering) 50g mixed squirrel food over a distance of approximately 10 m behind, 10 m in front and 10 m to either side of the stopping point. This procedure was carried out on a Monday and repeated on the following Friday, before starting the visual counts the following Monday.
• Volunteer surveys. In Year 2, volunteers surveyed 26 baited visual transects between June and August; these were carried out in close collaboration with the squirrel officers in Cumbria, Northumberland and the Borders (Southern Uplands Partnership).
• Results:

(a) Overall, detection rates using visual transects (baited or unbaited), hair tubes and trapping were low, variable and uncorrelated, and the data, although providing information on presence of red or grey squirrels, were insufficient to provide confidence that they represented estimates of abundance indices.

(b) Trapping. Trapping success was low with less than half the sites trapped yielding 3 or more trapped squirrels over the two years,

(c) Hair tubes - Year 1. These were not deployed in red squirrel/grey squirrels sites because of the potential for transmitting squirrelpox virus (SQPV). 20 tubes placed in a grid layout were examined twice a week for two weeks at an interval of 4 weeks in 8 sites. The proportion of tubes visited varied between 0% and 50% and improved through time. On average, 2% of tubes were visited on visit 1-week 1, 14% visit 2-week 1, 12% visit 2-week 2 and 23% visit 2-week 2. The method of using hair tubes is more labour intensive than visual counts and requires laboratory equipment and expertise at identifying hairs to species.

(d) Visual transects - Year 1. At each site, 500 m transects were walked at dawn on 5 consecutive days in each of 2 weeks, 4 weeks apart. Sightings were few with no squirrels seen at 7 out of the 12 sites on either of the two weeks. The return on effort was low.

(e) Baited visual transects - Year 2. There was a very slight increase in numbers of sightings following baiting, but numbers were very low and of no significance (2 sightings on 11 lines before baiting, 6 sightings after).

(f) Visual transects - both years. There were significant negative relationships between the number of squirrels seen and the proportion of Sitka spruce and between the number of squirrels seen and the nearest neighbour (NN) distance between trees. Thus, Sitka spruce, especially at high density, is particularly poor habitat to carry out visual surveys.

(g) Volunteer surveys. Provided clear and significant evidence that baiting transects increased the number of squirrel sightings (mean numbers of sightings per transect +/- SD: unbaited 0.73+/-0.72, baited 1.73+/-1.82. Most squirrels were seen in deciduous woodland and least in Sitka spruce, although this was not significant.

• Surveying and monitoring squirrels through time and space. Red squirrel numbers vary widely through space and time as they track tree seed availability. Densities, especially in conifer and especially in Sitka spruce plantation forest, can be very low, <<1 ha-1. This makes it very difficult survey and monitor squirrels with any confidence. As a result, we suggest that managers can only 'manage' squirrel carrying capacity and not red squirrel populations; grey squirrel populations, of course, can be controlled. Furthermore, target forests and woods should be initially surveyed using a simple broadscale survey to get a crude relative estimate of squirrel carrying capacity, and crude estimates of current squirrel activity and hence numbers of squirrels.
• The effectiveness of baiting visual transects, and indeed hair tube lines or grids, requires further study, as do the power of counts and hair tubes to detect trends in population indices. Drey counts, feeding transects, whole maize bait and nest boxes can be useful in certain situations. Of the range of other direct and indirect techniques that have been used to survey and monitor mammals, with our present state of knowledge there are none that instil confidence that they might be beneficial in detecting trends in numbers, and can distinguish between red and grey squirrels, sufficient to invest large amounts of time and money.
• Recommendations are based on acquiring knowledge about the carrying capacity and squirrel activity in target forests, using baited transects and/or whole maize (for grey squirrels) in spring to detect presence. Monitoring trends in numbers is problematic and can probably only be attempted using intensive studies. In these cases, visual counts and hair tubes should be used, perhaps as part of an integrated monitoring programme incorporating drey counts, cone feeding transects and nest boxes.

Download

• Download Developing a monitoring strategy for red squirrels (Sciurus vulgaris) across the UK (PDF, 489 kb)

You will need the free Adobe Acrobat Reader to view this document.