This paper reports upon the use of nest boxes in coniferous woodland by the red squirrel Sciurus vulgaris, great tit Parus major and the jackdaw Corvus monedula over a 6-year period from 1994 to 1999. Initially 30 nest boxes were erected with eight additional boxes being put up in 1995. In the period 1994 to 1997, red squirrels were found to use nest boxes regularly as nesting and breeding sites (7.8 ± 1.0 se litters found annually). Box use was greatest in the summer and autumn months (70% of boxes being used at least once) and declined through the winter (50%) into the spring (45%). The presence of great tit clutches and broods in the spring did not prevent red squirrels from occupying boxes. Only 28% (n = 12) of all the recorded great tit clutches produced fledgling young in 1994 to 1997 and red squirrels were found to have consumed some of the chicks and eggs. The establishment of a jackdaw colony within the woodland led to 86% of the nest boxes being used by the birds in the spring of 1999. Red squirrel occupation of boxes in the spring declined to 7% and only three litters were recorded that year. The study showed that the jackdaw has the capacity to dominate nest boxes during the spring months despite the high density of red squirrels within the reserve.

In non-native conifer plantations characterized by strong spatial and temporal variations in the availability of tree seeds in Spadeadam Forest, northern England, the home range and habitat use of red squirrels Sciurus vulgaris was very flexible. Males tended to have much larger home ranges than females and core-areas of most breeding females seemed mutually exclusive. Adult female red squirrels were found to increase their home range and core-area size in forest patches where food was less abundant. Home-range size was significantly related to home-range quality and the extent of overlap by other females. In contrast with high-quality continuous conifer forests: (1) a considerable proportion of adult males and females at Spadeadam shifted home range, (2) both sexes had much larger home ranges than reported from other habitats in Britain or Belgium. Many ranges were multinuclear, particularly from January onwards, when supplies of seeds become depleted through consumption and seed shed. Squirrels tracked the availability of conifer seeds (lodgepole pine cones throughout the study, Norway spruce cones in spring 1992 and Sitka spruce cones in autumn 1993) and intensively used several non-adjacent activity centres in temporally food-rich patches. Consequently, habitat preference changed markedly with time. The squirrels seemed to maximize nitrogen intake and to avoid the smaller seeds when possible. This resulted in an overall preference for a mixed diet of lodgepole pine and spruce seeds and avoidance of Sitka spruce seeds when Norway spruce seeds were available. These results lend support to the hypothesis of Ostfeld (1985) that when food is sparse and patchily distributed, females should develop intrasexual territoriality, concentrating their activity in food-rich patches, while males should be non-territorial and adapt their space use to the distribution of females.

Fourteen radio-tagged red squirrels were released in pine woodland containing grey squirrels. Movements of the squirrels were related to the tree species of the donor site. Survival after release was lower than for the grey squirrels: of 11 red squirrels that survived at least a week, only three survived more than three months and none for four months. More than half were eaten or cached by predators, mainly foxes; an experiment with grey squirrel carcasses indicated that they had been killed, not scavenged after death. Hypertrophied adrenals, disease and loss of weight indicated stress as another factor in the deaths. Data on overlap of core ranges, and reluctance of red squirrels to enter traps used by grey squirrels in the mixed population, indicated interference competition between the two species, with grey squirrels possibly dominant. We recommend: (i) that care should be taken to release translocated animals in similar habitat to their origin; (ii) that grey squirrels should be excluded from future release areas until red squirrels have settled and, before biodiversity is reduced by landscape management for red squirrels; (iii) more research to determine whether interactions with grey squirrels or differential predation will ultimately displace red squirrels in conifers.

The demography of red and grey squirrels was studied by live-trapping and radio-tagging at 14 deciduous and conifer sites in southern Britain and at eight conifer sites for one year in northern England. Densities and productivity correlated with tree seed crops for both squirrel species in deciduous and conifer habitats. Productivity was reduced by high density of full-grown squirrels relative to seed abundance. In oak-hazel woods, demography of grey squirrels correlated with abundance of acorns but not of hazel-nuts, whereas density and productivity of red squirrels correlated with hazel-nut abundance. Correlations of female density and productivity with pine-cone crops did not differ between red and grey squirrels. Predators ate many radio-tagged grey squirrels in conifers, and annual survival was only 50% compared with 80-82% for both species in other habitats. Grey squirrel populations in southern conifer sites were sustained by immigration, and at northern sites female density correlated with oak abundance within 500m. Failure to exploit acorn crops puts red squirrels at a competitive disadvantage in deciduous woodland. Red squirrels had higher survival than grey squirrels in conifers, which may give them an advantage in that habitat, but could also have been explained by a lack of predators on their island study site.