100 DAVID JENKINS AND ADAM WATSON 



finding corpses. Detailed results given in the progress reports suggest that 

 although there is sometimes movement from one part of the area to another, 

 upwards of 40 hectares (100 acres) is a reasonable minimum unit for counting. 

 Three hundred acres is the standard size for our other study areas but we 

 do not count on these in deep snow, when grouse are moving widely. Nearly 

 all counts are done between 10.00 and 14.00 hours, since it is mostly during 

 the afternoons that grouse desert their territories, and counting by this 

 method would then be less reliable; exceptionally counts are done later 

 when a bhzzard is imminent. We do not count grouse in winds stronger 

 than force 4 (Beaufort scale) since the birds then tend to become very wild 

 and to flush far away from us. Rain makes counting impossible. 



Grouse may sometimes be counted twice if they fly in front of an observer, 

 particularly if they are flushed at a distance. There is also a chance that some 

 birds are missed. Figures obtained for counts are therefore often presented 

 as a range consisting of a 'minimum', i.e. birds which flew away behind the 

 observer and are known to have been seen only once, and a 'maximum', i.e. 

 the minimum plus those birds which flew in front and which may possibly 

 have been counted twice. It is usually possible to see where birds land and if 

 known individuals are flushed a second time they are not counted again. 



It is important that the area to be counted is not too large because one 

 might then be unable to see where birds landed that flew far away; with a 

 relatively small area it does not matter where such birds land because the 

 place will be outside the ground to be counted. The maximum includes only 

 birds that were not seen to land, but which from their behaviour seemed 

 hkely to land on the area yet to be counted. Few birds fly in front when 

 there is a moderate breeze blowing, but some counts have to be done in 

 calm weather. For convenience in deahng with figures (as in Fig. i), we use 

 the mid-point between the maximum and the minimum. We also count all 

 the dead birds we find, marking them with a stick and recording the probable 

 immediate cause of death (Jenkins, 1958^, p. 57 ^^ s^^-)- 



We have checked the accuracy of count totals by recounting with the 

 same or a different observer and also by counting from a Land-Rover 

 (Table I). Counting from the Land-Rover is not much use except in dry 

 places, because of ditches and bogs which prevent one covering the whole 

 of a piece of ground (but it is useful for checking the identities of marked 

 birds). We do a check-count whenever time allows and as routine whenever 

 a count total shows a large change from a previous figure or there is too 

 large a discrepancy, more than about 15 per cent, between the maximum 

 and minimum figures. Counts are also repeated when it seems that the grouse 

 may have been disturbed during a count, perhaps by a shepherd or predator; 

 an advantage of working gently sloping ground is that one can see whether 



