FLUCTUATIONS IN A RED GROUSE POPULATION 99 



desiccated owing to high evaporation and transpiration and minimum or nil 

 water uptake from the frozen soil. Heather was spoiled in Glen Esk in anti- 

 cyclonic weather both in April 1958 and, more severely, in February 1959 

 when much heather less than about five years old was killed and some older 

 plants growing in dry places were severely damaged. This effect was local 

 and seems to be prevented by a covering of snow. Similar damage to young 

 plants occurred on some shallow soils in the drought of summer 1959, but 

 this was insufficiently widespread to be important to grouse. 



The first winter of the study, 1956-7, was mild without extremes of frost 

 or snow and the subsequent spring and summer were wet and rather cool. It 

 was the most successful grouse season so far. In 1957-8 there was a prolonged 

 cold spell in the late winter, with deep snow on the hills till April and severe 

 frost at low altitudes. The summer was essentially similar to 1957, with 

 periods of both rain and sun when the chicks were small. 1958-9 was a 

 season of extremes — the winter was open, with little snow in Glen Esk but 

 prolonged black frosts for weeks in the late winter, freezing the ground over 

 2 ft deep, while the summer that followed was one of the warmest and driest 

 this century. 



METHODS 



(l) COUNTING 



We use trained dogs to find both live and dead grouse. Without dogs it 

 would often be difficult to count the birds. We do this by walking backwards 

 and forwards across the study area at right angles to the direction of the 

 wind. The dogs run about within 50-70 m on either side, finding and 

 flushing the grouse. We walk the transects in different directions according 

 to the wind at the time, so that the wind carries the flushed birds out of the 

 area altogether or on to ground already covered. Grouse usually return 

 within about 10-30 mins., when flushed, and our transects are never more 

 than half a mile long so that we can see most of the ground all the time and 

 return on the next transect before the birds come back again. The method 

 is an attempt to count all the grouse that are present, and is suitable because 

 the terrain is very open, the highest vegetation being shorter than o • 5 m. 



The size of a piece of ground to be counted on any one occasion depends 

 to some extent on the topography and on the urgency of the occasion (we 

 might count more if a storm was forecast). We find that it takes about 45- 

 60 mins. to count grouse satisfactorily on 40 hectares (100 acres) and we 

 rarely cover more than 120-160 hectares at a time. We have divided our 

 study area into parts and the method is standardized in that we always count 

 at least one whole part (50-110 hectares) at once on roughly the same 

 transects. These are not straight lines ; we zig-zag to give the best chance of 



