104 DAVID JENKINS AND ADAM WATSON 



the first half, and the result suggests that we should really have repeated the 

 whole count. On III the first two runs of figures were broken by counts of 

 131 and 109 respectively. In both cases there was an explanation — the 131 

 total was obtained by a student ill-acquainted with the ground and un- 

 accustomed to handling dogs and the figure was probable shghtly inaccurate. 

 On the 109 count eight birds rose just outside the study area and were 

 discounted and another flock flew at extreme range. This group was noted 

 as *at least 20, not more than 30' and was counted as twenty birds. Had 

 the extra 18 grouse been added to the 109 total, there would have been seven 

 similar totals in a row, from 27th November 1957 to 30th January 1958. 

 The method, however, does not depend on the detailed accuracy of single 

 counts and the fact that this one count registered 109 is immaterial. Clearly 

 the population changed between count 4 and count 5 and then remained 

 more or less constant until after count 11. This is the sort of conclusion in 

 which we are interested and for which we claim our method of counting to 



be vahd. 



Between August and April we do counts fortnightly if the weather allows. 

 Each count over the whole study area usually takes one observer about five 

 days or two people about three days. Counting is impossible in summer-time 

 when grouse are nesting, because the disturbance would be too great. We 

 then visit different parts of the area daily and, by summing the results of our 

 observations, build a mosaic picture of the population. Sometimes a count 

 has been possible before shooting starts but the figures for 12th August 1957 

 and 1958 are derived from several visits on different days. Through extensive 

 ringing of young birds we are confident that movement of family parties 

 was negligible in these years, but difficulties have arisen with non-breeding 

 adults, particularly in 1958. Such birds apparently move into and through 

 the study area during the summer and may cause some errors in our popula- 

 tion estimates, particularly with regard to young-to-old ratios (Jenkins, 1957, 

 igsSb, i960). Breeding success is not considered here but we think our study 

 area, and the numbers of breeding grouse involved, are sufficiently large to 

 make such inaccuracies small and unimportant. 



(2) MARKING 



In addition to ringing young grouse, we catch full-grown birds on the 

 study area stubbles in the autumn, mostly in October and November, in 

 funnel-entrance traps made of wire netting and baited with oat stooks. The 

 grouse are ringed and individually marked with coloured plastic tabs (Blank 

 & Ash, 1956). These tabs are 11-4 cm (4! in.) long by 2-5 cm (i in.) wide 

 and he on the birds' backs, secured by soft leather thongs looped round the 

 base of the wings. Apparently they cause no discomfort or disability and the 



