ROE-DEER CENSUS AND POPULATION ANALYSIS 79 



and the 154 observations made of them and considering all other deer 

 unmarked (and n = 462) a population estimate of ninety-six is arrived at (as 

 opposed to 100 when the entire material is used); although the standard 

 deviation goes up to 6 the estimate is only slightly inferior to the one 

 mentioned earlier. 



In 1959 the weather was not favourable for trapping, since it was mild and 

 practically without snow; nevertheless we succeeded in trapping sixty-eight 

 deer (thirty-seven fawns and thirty-one older individuals). It has already 

 been mentioned that in all probabihty the percentage of fawns was too high. 

 Four individuals were removed from the population and only sixty-four 

 were released, among them were thirty-six fawns (born 1958) which received 

 an orange collar. All older animals received the usual red-green collar, and 

 this means that the number of red-green deer is, by now, unknown. The 

 same appHes to the blue deer of the previous year, and both these classes 

 may have emigrated to some extent during the previous summer, the actual 

 amount of emigration being unknown since we cannot be certain to receive 

 report of all emigrants. 



A total of 650 observations are available from 1959: 



Without collar 183 (28 per cent), blue (born 1957) 133 (20 per cent), 

 orange (born 1958) 155 (24 per cent), red-green 136 (21 per cent), white 

 (born 1955) 44 (7 per cent). 



It is not yet possible to set up a Hfe table or to consider the age classes 

 individually, primarily because a considerable number of deer were killed 

 selectively in 1955. In general one should not expose the population to 

 shooting as long as experiments of this kind are being conducted; the 

 removal of a few individuals on the other hand is probably inconsequential 

 as long as the annual emigration is of the order of one year's production. In 

 1955 we made the mistake of shooting eighteen deer during the autumn, a 

 few months prior to the trapping. Under the Danish Game Act there are 

 two open seasons for roe-deer; bucks may be killed from 15th May to 15th 

 July and bucks and does from ist October to 31st December. I have long 

 been of opinion that the shooting of bucks during early summer was 

 unlikely to affect the autumn population, but to lead primarily to decreased 

 summer emigration. However, I am now inclined to believe that, after all, 

 summer shooting may affect the autumn population. 



Calculations based on the data from 1959 can only be made by considering 

 the orange deer as opposed to aU others. The result indicates a winter-spring 

 population of about 150 deer, the standard deviation being rather high 

 (5 = 10). This would imply that the 1959 population is a good deal larger 

 than found for previous years and that the population is not of constant size 

 as we have believed up till now. 



