204 Game Survey of the North Central States 



between the northern prairie lakes and those of the Forest Belt. Even when the 

 woods lakes look similar to those of the prairie, they produce but a fraction of 

 the birds. 



In company with Alfred Peterson, of Pipestone, an ornithologist, I censused 

 eight lakes, aggregating 15 sections, or 9,600 acres, in southwestern Minnesota in 

 August, 1928. These lakes are only average in quality. We estimated an output 

 of 9,500 ducks, which is an average of 600 per square mile, or one duck per 

 acre. There was no migratory flight yet all these ducks were raised in situ. 



For comparison I obtained from local sportsmen estimates of the usual annual 

 output of 18 lakes and marshes constituting the larger remaining breeding 

 grounds in the forest belt of Wisconsin (see Map 17). The estimates total 

 23,700 ducks. I could not obtain all the acreages, but the total would be in six 

 figures, and one of them alone was determined to be 24,000 acres. If its 

 capacity were equal to that of the Minnesota lakes (one per acre) it alone would 

 be producing as much as all the 18 Wisconsin lakes actually produce. 



Another comparison: Lake Winnebegoshish in Minnesota is estimated by 

 local sportsmen to raise 100,000 ducks per year. This is four times the output of 

 the eighteen Wisconsin lakes, and possibly not much smaller than the present 

 output of the whole State of Wisconsin (but not her former output. Horicon 

 Marsh, now drained but being restored, is estimated by local sportsmen as having 

 once produced a million ducks) . 



The moral is that losses of breeding grounds from drainage, or programs 

 for restoring such losses, are questions not of acreage but of quality. The loss 

 of one lake on the Minnesota, Dakota, or Canadian prairie may equal a whole 

 eastern State. 



Clyde Terrell, of Oshkosh, Wisconsin, estimates the 1930 output of eight 

 lakes and marshes in and near Winnebago County, Wisconsin, as 17,880 ducks 

 and 9,250 coots or mudhens. Four of the eight were included in the 18 Wis- 

 consin breeding grounds censused by the game survey. In three of these, his 

 census checks mine within 10 per cent; in the fourth his census is double mine. 



The composition of the breeding population in Terrell's Wisconsin census 

 (by per cent) and the game survey's Minnesota census (by order of abundance) is 

 given in Table 42. (Breeding composition is of course different from the com- 

 position of the kill in Table 46, the latter including mostly migrating, not local, 

 birds.) 



In Terrell's census, mudhens were one-third more abundant than mallards, 

 the most abundant duck. In my census, mudhens were not enumerated, hence 

 they are omitted from the table. 



Breeding grounds in the southern parts of the north central region have not 

 only been radically restricted by drainage, but ducks, except woodducks, show a 

 curious reluctance to use the ones that remain. This is true as far north as the 

 Upper Mississippi Refuge. There is no sharp line of demarcation between the 

 acceptable and unacceptable breeding grounds. 



