FLUCTUATIONS IN A PARTRIDGE POPULATION 121 



shooting depends upon high autumn populations (high young/old ratio) it 

 may be argued that the high early winter loss is not necessarily connected 

 with shooting. However, in 1958 when the autumn population was low 

 (resulting from a low rate of chick-survival) shooting pressure was deliber- 

 ately increased out of proportion to the September population. The result 

 was that in spite of the relatively low autumn stock (with a low young/old 

 ratio) the rate of loss (17*6 per cent) was much higher than in years with 

 similar autumn populations but much reduced shooting pressure (average 

 I953» I954j 1956 = 3*3 per cent). Observation of marked coveys on an 

 unshot area confirms that loss during this period is relatively small, although 

 some local movement occurs. This movement is really an extension of the 

 daily range of the coveys, and is initiated by the ploughing up of the autumn 

 stubbles. Coveys generally extend their daily range so as to include a favourite 

 feeding ground (i.e. on unploughed or undersown stubble). Since a large area 

 is censused, it is unlikely that chance crop-distribution has seriously affected 

 the vahdity of the censuses, but some part of the early winter loss may be 

 due to the redistribution of feeding grounds after the autumn ploughing is 

 completed. Early winter loss has averaged 1 1 • i per cent, while the range is 

 from o per cent to 30 per cent of the September population. 



Partridge shooting on the area censused has always ended in November. 

 The count which has enabled us to distinguish between the early and late 

 winter losses takes place in mid-December — just before the earliest date 

 that pair-formation is likely to occur. From late December until early 

 February the rate of pairing increases, and by the end of February all the 

 coveys have broken up. Over this period, losses from accidents etc. occur, 

 but towards the end of this period there is some evidence of shghtly increased 

 losses to predators. This loss is apparent even on the covey range, and may 

 well be more marked in birds which emigrate. By late March a considerable 

 reduction in partridge numbers has occurred, and observation of marked 

 birds has shown that the major part of this reduction results from the move- 

 ment of some juveniles away from the winter ranges of the parent coveys. 

 Over the observed range in December population (1,126 to i,9i7)» late 

 winter loss has been a fairly constant proportion, although modified to some 

 extent by the crop distribution. A high proportion of arable to grass tends to 

 reduce breeding pair density and to increase the rate of late winter loss. The 

 relation is not necessarily an exact one, however, since the pattern of the 

 crop distribution is also of some importance. Large areas of contiguous 

 ploughed ground usually produce low pair densities, while ploughed land 

 interspersed with grass or undersown stubbles, encourages relatively high 

 breeding densities. In 1949 and 1950 the total acreage of ploughed land during 

 the winter months was high and late winter loss well above the average. 



