| Summary: | This study aimed to assess the impact of pine marten (<I>Martes martes</I>) predation on red squirrels (<I>Sciurus vulgaris</I>) in Scotland. Red squirrel density, range use and habitat selection were compared at two study sites: Novar, where pine martens were present, and Learney, where they were absent. At Novar there was a lower density of squirrels than at Learney. Habitat selection at Novar was greater than at Learney with squirrels preferring Scot pine (<I>Pinus sylvestris</I>) and avoiding sitka spruce (<I>Picea sitchenis</I>). At Learney, no selection was shown and with squirrels using all habitats, including sitka spruce, in proportion to their availability. Pine marten diet was assessed from the analysis of 1938 scats at the Novar study site. Red squirrels comprised 0.6% of estimated weight intake in pine marten diet; in relation to the density of squirrels at the study site this represented a high proportion of the population. The field vole (<I>Microtus agrestis</I>) was the most important species in the diet. Pine marten habitat use was assessed by radio tracking. This revealed that martens avoided non-forest habitats and used broad-leaved woodland in greater proportion than it was available. There was overlap in the home ranges of pine martens and red squirrels and martens hunted in all habitat classes used by squirrels. In addition, the habitat used most frequently by pine martens (pole stage coniferous woodland) also had the greatest proportion of squirrel dreys present. Analysis of drey use at the two study sites did not show any evidence of squirrels adjusting their behaviour in the presence of pine martens. In contrast to what was expected, squirrels at Learney used a significantly greater number of dreys, which were more concealed than at Novar. At Learney drey characteristics could not explain drey use, whereas at Nova drey use was positively related to dreys that were in smaller trees with high canopy contact, and a high density of surrounding trees. It was suggested that this was due to fewer suitable trees being present at Novar whereas at Learney there were more spruces present which provided better cover and concealment for dreys. The density of red squirrels was assessed at several forests throughout Scotland, using counts of cones eaten by squirrels, and was then related to the length of time pine martens had been present. Forests where martens had been present for more than 10 years had a significantly lower density of red squirrels than either areas where martens had been present for less than 10 years or where martens were absent. There differences were independent of any variation in the availability of cones and the habitat of the forests surveyed. Using records from the Forestry Commission squirrel questionnaires, changes in the distribution of red squirrels over a 20 year period were related to the presence and absence of pine martens, and also to changes in the distribution of martens. Red squirrels were present in a smaller proportion of 10 km squares in areas where martens were present than where they were absent. There was also a greater proportion of losses and fewer colonisation of red squirrels in areas where martens were present.
|
|---|
