© 2003 by Institute of Chartered Foresters
Using cohort analysis to reconstruct the size and structure of deer populations in forestry with special reference to roe deer (Capreolus capreolus L.)
1 Centre for Ecology and Hydrology, Merlewood Research Station, Grange-over-Sands, Cumbria LA11 6JU, England
This paper describes an attempt to assess retrospectively the numbers and demography of a population of roe deer (Capreolus capreolus L.) on an estate and its contiguous area of forestry in North Cumbria, England, amounting to 968 ha, of which 285 ha was woodland. After ageing all the jaws of deer which had been culled or found dead over a period of 29 years (359 deer), populations were reconstructed for each of the earlier years using cohort analysis. Excessive immigration of bucks onto the area of newly planted woodland and the absence of resident does for the first 8 years limited the analysis to one part of the estate (Kingfield) with assessments for the years 1972–1997. As mortality data for Kingfield were limited (222 deer), they also magnified the effects of gains and losses through migration. As a result, and because the area was surrounded by forestry with only lightly culled deer populations, immigration by non-territorial bucks caused some difficulty in assessing the real population in the earlier years. At first sight therefore, these results would appear to negate the usefulness of cohort analysis to determine the numbers of deer in forestry. They do, however, demonstrate quite clearly that, as long as virtually complete mortality data are collected, even with small numbers reasonably accurate assessments of previous populations can be obtained. The most reliable analyses for the whole period were those of the does and female kids, the only real residents. Using these data and those of the bucks for recent years, the probable levels of stocking latterly amounted to 48.8 deer km–2 in plantations of all ages, or 60 deer km–2 in woodland >13 years after planting. The density of deer was also shown to have been positively and significantly (P 
0.001) related to the area of woodland >13 years after planting (<40 ha in 1972, 70 ha after 1993). Moreover, it is now possible to set more realistic targets for the culls on the estate at 
40 per cent of the reconstructed populations, amounting to an annual cull of 19.5 deer km–2 of the total woodland area or 24 deer km–2 of woodland >13 years after planting. This annual cull amounts to five bucks, six does and six kids. A similar exercise on a much larger scale is now required if whole deer populations are ever to be effectively managed in woodland.
Received 24 July 2002.