ROE-DEER CENSUS AND POPULATION ANALYSIS 77 



the sex ratio was seventy-six females to forty-six males in 1950, i.e. 1-7:1 

 and thirty-four females to seventeen males in 1958, i.e. 2:1; and these 

 differences are not considered significant. 



The trapping results in 1956 support the assumption that fawns are more 

 easily trapped than adults ; a total of thirty-eight deer were trapped among 

 which were 53 per cent fawns; the corresponding data from 1959 were: 

 sixty-eight deer trapped and 54 per cent fawns. 



On the basis of the 1950 population analysis and the good trapping data 

 from 1958 I had foreseen that in the end we would arrive at a percentage of 

 fawns of about 40 ; a hypothetical diagram showing fawn percentage as a 

 function of duration of trapping period had already been constructed, the 

 percentage starting at about 60 and declining, first slowly and when about 

 half the population had been trapped more rapidly, to stop at about 40 per 

 cent. I even thought myself able to provide a detailed explanation but this 

 year's results, coming in while this paper was being written, have shown that 

 such an action would be unwise. Up till now (2nd February i960) we have 

 trapped seventy-five deer; displayed as before, the data look as follows: 



The first 25 captures included 8 fawns, i.e. 32 per cent 



„ „ 50 ,, ,, 18 ,, 36 



All 75 „ „ 30 „ „ 40 



It is reassuring that the fmal percentage seems to be stable, but I have given 

 up the idea of drawing a curve to show how one arrives at this figure. My 

 conclusions may be summarized by stating that it is necessary to capture a 

 very large proportion of the population in order to get representative data. 

 Our preliminary results are only mentioned here because they show how 

 easily one is led to make erroneous conclusions. In 1950, during the kill, we 

 had a similar experience: the sex ratio changed with time, and not until 

 75 per cent of the population had been killed was stability attained. 



I shall now return to the marking procedure used in 1958. All fawns (bom 

 during the summer of 1957) received a blue collar, all older ones a green-red, 

 i.e. a collar with green and red buttons alternating. Among the older 

 individuals were some which had been captured in 1956 and received the 

 collar used for fawns at that time (white) ; their collar was replaced by another 

 one but the white colour was retained. Not all deer captured were released, 

 a few died and some were removed to enclosures to be used for other 

 experiments. The individuals actually released were seventy-four in number: 

 nine white (12 per cent), thirty-two blue (43 per cent), thirty-three green-red 

 (45 per cent), i.e. forty-two white + green-red (57 per cent). 



During the subsequent period (19th March to ist May) the observer made 

 462 observations distributed as follows: 



