COMMON FISHES OF MISSISSIPPI RIVER 
193 
that any of the species are strongly migratory or that conditions at Keokuk could 
have any effect whatever upon the abundance of buffalo fishes at a place several 
hundred miles distant. 
The statistics of the commercial fisheries for the years 1914, 1917, 1922, and 
1927 (Coker, 1929) show a steady increase in the catch of buffalo fish in Lake Pepin— 
an increase, however, that is not so marked as to require any special explanation. 
In 1926 there was general complaint of scarcity of buffalo fishes at all places visited 
in Minnesota and Wisconsin, but the years 1925 and 1926 had been poor for buffalo 
fishes in all parts of the upper river, and apparently the same condition prevailed 
in 1927. The catches for the several years of report were as follows in pounds: 
261,000 (1914), 301,800 (1917), 340,000 (1922), and 33,000 (1927). 
The story as regards Lake Keokuk is quite different. The yield of buffalo 
fishes took an enormous upward jump in 1917 and showed a tremendous slump 
in 1922 and 1927. The figures in pounds in round numbers, are as follows: 250,000 
(1914), 697,000 (1917), 114,000 (1922), and 68,000 (1927). Comment on this 
remarkable series of facts has already been made in a previous paper. (Coker, 1929.) 
We had supposed after 1917 either that the creation of the lake had made most 
favorable conditions for the reproduction and growth of buffalo fishes or that fish, 
dropping down from the upper river, collected in the lake because of the slackened 
current. It was naturally presumed that a permanent high level of catch in the 
lake was assured. It has not turned out so. It may be instructive to record some 
of the physical changes in the lake that occurred between 1917 and 1922. 
In 1917 the lake extended out over the broad Green Bay bottoms on the western 
side between Montrose and Fort Madison, and these were known to be most pro- 
ductive fishing grounds and reputed to be the spawning ground of buffalo and other 
fishes. Later this whole region was formed into a drainage district and leveed, 
the levees, 21 miles in length from near Fort Madison to Skunk River, being closed 
in the winter of 1918-19. It does not appear that there are, along the whole course 
of the lake, any other outlying breeding grounds of adequate area. We may well 
question if the recovery of the overflowed lands and the old Green Bay for purposes 
of agriculture can yield any public benefit sufficient to compensate for the sacrifice 
of its value to the fisheries of the river as a whole. 
In the body of the lake in 1917 there were many submerged islands still covered 
with the trees they had grown prior to the formation of the lake. By 1922 nearly all 
of the trees had disappeared, partly, perhaps, from decay, but chiefly, it is said, from 
being cut away in winter when they could be made into firewood and carried over the 
ice to the near-by villages. The flats corresponding to the islands are now almost 
fully exposed and open to wave action. Local informants say that buffalo fish and 
carp are seen abundantly in spawning activities over some of the “islands,” and give 
it as their opinion that the nests are destroyed by the strong wave action that prevails 
when high winds sweep across the broad open expanses of the lake. 
It may be significant that it is the catfishes that seem consistently on the increase 
in the lake. These are fish that can dig into banks and find or make protected places 
for breeding. 
If a later growth of trees that endure submergence of the bases or the further 
growth of strongly rooted aquatic plants should give these regions protection from 
too vigorous wave action, probably two good purposes would be served — favorable 
breeding places for fish might be provided and a reduction might be effected in the 
