no 



DAVID JENKINS AND ADAM WATSON 



In 1957 the total autumn decrease was similar to that caused by shooting 

 (48 per cent), but in the two subsequent shooting seasons only 17 per cent 

 and 15 per cent respectively of the birds available were shot (c£ Table II). 

 Hence there was no clear relation between the bag and the total losses at 

 this season, and the close agreement in 1957 may well have been due to 

 chance. This is in fact suggested by the data; counts between September 

 and November 1957 showed that the population increased at the end of 

 shooting and fell again later (Fig. i), probably owing to some process quite 

 unconnected with shooting. 



N D J F r 

 1956 1957 



Fig. I.— Changes in numbers of grouse on the study area (1,135 acres) from November 1956 to 

 January i960. Note. The first three points, A-B, are based on empirical data from 830 acres 

 — see Table II. 



In 1958 and 1959, when the losses through shooting were considerably less 

 than the total decrease in numbers, the autumn declines apparently occurred 

 all in one stage in August or September. The decreases coincided with the 

 shooting season (when we do no counts), but were not necessarily attribut- 

 able to exploitation. In 1959 some of the losses were probably due to move- 

 ments, and the very low count figure for September was associated with an 

 exodus to a temporary food supply provided by berries on some of the hills. 

 Some grouse apparently returned later (Fig. i). 



A period of relative stability of population density during each mid- 

 winter period followed this late summer or autumn decrease, and later there 

 was a further fall in numbers of 24-43 per cent in February or March. Then 

 numbers stayed level again through the spring. In the first two seasons 

 (1956-7 and 1957-8) the mid-winter and spring populations were each 

 similar, but the levels were different in 1958-9 and subsequently. None the 



