192 
BULLETIN OF THE BUREAU OF FISHERIES 
eries biological station at Fairport, Iowa, conducted the experiment mentioned in a 
following paragraph. Meantime, however, unsuccessful attempts had been made at 
various places to propagate the buffalo fish artificially by the mixing of eggs and sperm 
and the application of indoor hatchery methods. The difficulties entailed in this sort 
of propagation were solved by Supt. H. L. Canfield at the Fairport station in the spring 
of 1915, and since that time the buffalo fish has been extensively propagated by the 
Bureau of Fisheries in various places. The methods, which have been fully described 
by Canfield (1918), involve the use of special methods to avoid agglutination of the 
eggs, as by the use of corn starch and by brushing the eggs through bobbinet screens at 
different times. He tells of the taking of spawning fish in fyke nets set over inundated 
lands, and of the great diminution of catch whenever the water recedes, suggesting 
that the fish return to the river with falling water. This is a point of interest, as we 
shall see. 
The Fairport station had also attempted to promote natural propagation of 
buffalo fish in ponds, but met with no success until the spring of 1917, when the 
conditions were varied by keeping the pond about half full of water in the early part 
of the season and allowing it to fill gradually early in May. A few days after the 
production of this artificial flood stage the splashing of buffalo fish was observed 
(May 11 and again May 17, 18, and 19) in overflow regions along the margins of the 
pond, and propagation was found to have been successful. (Shira, 1917.) The 
methods have been more fully described by Canfield (1922), who found that in 
artificial ponds it was necessary nicely to time the artificial rise of water level to the 
rising temperature of the water. The rise should begin when the temperature of 
the water is 56° F. and should be so controlled that it is completed in 10 to 15 days, 
with the water at 62 to 64° F. Spawning begins at 56 to 58° but is more active at 
60 to 62°, so that the fish have spawned out when the rise is concluded. The large- 
mouthed buffalo fish ( cyprinella ) may be bred without the artificial rise, but the 
smallmouth ( bubalus ) does not do well without it. The rise is found desirable for 
both species, as the weathered grounds seem to offer a more favorable environment 
for the eggs during incubation. It has also been the experience of fish culturists 
propagating buffalo fish in the field that a normal rise in the river is beneficial (Can- 
field, 1918: Fisheries Service Bulletin [Culler], 1922), although it may lead to scatter- 
ing of the fish and thus hamper the collectors who are seeking the spawners. 
This account of the conditions of propagation has been given because the facts 
are not readily available and also because, in appraising the effect of the formation 
of a lake on the reproduction of buffalo fishes, it is important to know whether or 
not a normal spring flood, the effect of which would be diminished under lake con- 
ditions, is favorable or unfavorable to the multiplication of these important fishes. 
Apparently a rise of the river in spring, causing the water to flow out over previously 
exposed ground, is a positively favorable factor; but it may be that there will still 
be, in a pool of the nature of Lake Keokuk, a rise sufficient to meet the needs of the 
buffalo fishes. 
From the fact that buffalo fishes are frequently taken in the current, it has 
been thought by some that they are migratory; a prominent fisherman in Lake 
Pepin spoke of “southern buffalo” as being less common than before the dam was 
built at Keokuk. Although buffalo fish move upstream at times, and doubtless 
drop back at others, and the upstream movements may normally be more vigorous 
when the fish are seeking spawning grounds, there seems to be no evidence at all 
