280 The Ruffed Grouse 



lation left a breeding stock for 1935 which was thirty per cent 

 higher than that of the year before. In the ensuing summer, brood 

 survival again fell disastrously as it had in 1933, giving a recovery 

 of only nineteen per cent for 1935. This time the abnormal loss was 

 apparently largely caused by extreme precipitation, ten inches of 

 rain in forty-eight hours, which also provided the worst floods in 

 the history of the region. With the fall population thus lowered to 

 two hundred eighty-seven birds, we would expect a reduction in 

 winter loss. The number of birds lost was lower than in the pre- 

 vious winter, but the percentage lost remained high. In fact the 

 forty-six per cent reduction was the highest in the first seven years. 

 So also was the net loss in breeding stock— thirty-six per cent from 



1935. Again we find the immediate explanation in local environ- 

 mental conditions— adverse winter weather that resulted in abnor- 

 mally high vulnerability to predation. And I pause to note that any 

 excellence of habitat would not prevent an increase of predation 

 under those extreme conditions of snow and low temperatures. 



From the relatively low population carrying into the spring of 



1936, recovery was efi^ected for the third time in the period of the 

 survey. Productivity was good although not as high as in some years. 



Survival through the winter of 1937 proved to be the lowest yet 

 observed, sixty-one per cent of the September, 1936, population 

 being lost by the next April. This reduced the breeding population 

 to the lowest level since 1930. Once more, however, a productivity 

 of one hundred twenty-five per cent brought a quick recovery which 

 continued with increasing fall and spring populations until the 

 summer of 1940. Then, as had occurred twice before, the produc- 

 tivity fell to a low level, only fifty-two per cent, this time the result 

 of an abnormally high nesting loss. 



From this record, we may conclude that during periods of grouse 

 increase, the net summer productivity is normally about one hun- 

 dred twenty per cent of the breeding population. At the peak, occa- 

 sional disasters, affecting brood mortality primarily, may reduce this 

 expectation practically to the vanishing point. The summary of the 

 actual productivity rates on the Connecticut Hill area, together 

 with the subsequent changes in breeding populations as compared 

 with the preceding year, are given in Table 10. 



There is a direct relationship between yearly recovery, or pro- 

 ductivity rate, and population density. This same relationship car- 



