| Summary: | Thesis (M.Sc.)--Memorial University of Newfoundland, 1997. Biology Bibliography: leaves 78-87. The population ecology, home range size, and caching behaviour of red squirrels was studied in black spruce habitat in Terra Nova National Park, Newfoundland. Density was almost twice as high during summer 1995 (a year of heavy cone crop) than in 1994, but densities in both years were similar to those reported in other studies. Densities ranged from 0.4 - 1.3 squirrels /ha in 1994 and 1.0 - 2.7 /ha and 0.8 - 3.7 /ha on grids 95-1 and 95-2 respectively in 1995. Transients (animals captured only once) were a minor component of the population in both years. The larger number of spring recruits and higher summer body weight in 1995 indicate that the greater density was due to higher overwinter survival during 1994/1995. Total adult sex ratios (based on all animals captured) were significantly biased in favour of males in 1994 and for the combined adults in 1995. -- Summer home range size in 1995 ranged from 0.34 - 4.1 ha (95-1) and 1.8-2.4 ha (95-2) using the minimum convex polygon method and from 0.34 - 4.8 ha (95-1) and 4.3 - 8.0 ha (95-2) using the adaptive kernal method. These values are generally larger than home range and territory sizes reported in other studies. The presence of deciduous food and the breeding season, may have contributed to the larger home range size. Home ranges were not exclusive and all showed signs of overlap. -- North American red squirrels exhibit variation in caching behaviour across their range. Larderhoarding dominates in the west and scatterhoarding dominates in the east. One proposed explanation for this variation is predation risk. If safe feeding and caching sites are unavailable, then moving to, and eating at scattered caches may elevate the risk of predation for the caching individual. It was predicted that in areas of low understory shrub cover, predation risk is higher and safe caching sites are rare. Therefore, creating larger caches at one or a few low-risk sites would be the better strategy. -- The caching behaviour of red squirrels was documented in 1995 in grids 95-1 (high cover) and 95-2 (low cover). In both grids, both single cone and large caches (11+ cones) occurred more frequently than expected, but cache size frequency distributions did not differ significantly between grids. The mean (± SE) number of cones in large caches (30.9 ± 2.5) was significantly greater in low cover than in high cover (18.3 ± 2.6). Large caches accounted for 65% and 20% of all cones stored in the low and high cover grids respectively. The size of the cache (single or multiple cones) was not dependent on the distance from the nearest tree in low cover, but in high cover, single cone caches were found farther from a tree (mean ± SD: 1.5 m ± 1.3) than multiple cone caches (1.0 m ± 0.67). It appears that the amount of cover may be one component of red squirrel caching behaviour.
|
|---|