The mortality that nearly exterminated the huge stocks of smelt in Lakes Huron and Michigan during the fall and winter of 1942–1943 appears to have originated in central Lake Huron in the Saginaw Bay area in late September or early October 1942. The mortality spread rapidly northward reaching the Drummond Island area about the latter part of October and the St. Ignace region of the Straits of Mackinac near the end of the month. In the latter part of October smelt died also in the Canadian waters of Lake Huron including North Channel and Georgian Bay but exact details as to time and course are lacking. There is some evidence that the epidemic had not reached the Ontario shore of central Lake Huron by late May 1943. Spreading through northern Lake Michigan the mortality had penetrated as far south as Grand Traverse Bay by November 19 and as far west as Point Aux Barques, Michigan, by November 26, 1942. Smelt were reported to be dying in Lake Charlevoix, Michigan, in early February 1943, and in Green Bay toward the middle of that month. The mortality did not reach Crystal Lake where in contrast to Lake Charlevoix a dam barred the passage of fish from Lake Michigan. At the time of the 1943 spring spawning run (April) only a few scattered survivors remained from the vast populations.

After consideration of possible causes, it was concluded that the mortality could be explained only as resulting from a communicable disease (bacteria or virus). This explanation alone is in harmony with the following facts: the mortality was progressive, spreading from one area into adjacent areas over a period of at least 4 1/2 months and under a great diversity of habitat conditions; only smelt were affected but within the species death overtook fish of all sizes from 2 inches on and all ages of both sexes, mature and immature; the mortality penetrated Lake Charlevoix where the passage of fish to and from Lake Michigan was possible but did not reach Crystal Lake where the passage from Lake Michigan was barred by a dam; the epidemic did not reach other inland lakes where free and easy access from the Great Lakes was impossible nor did it extend to Lakes Superior, Erie, and Ontario.

Considered at first as a nuisance and a threat to the native fishes of the Great Lakes, the smelt ultimately became a fish of primary importance to commercial fishermen, sportsmen, and others. In Green Bay, the center of the commercial fishery, smelt became the dominant commercial species, yielding more than 4 million pounds in some years. Almost all of the commercial production was from nets set under the ice. The take by amateurs and others who dipped smelt from streams during the spawning run was even greater, amounting to as much as 5 1/2 million pounds in a single year in the State of Michigan alone (the yield in Wisconsin may have been nearly as great).

The mortality of smelt was a severe blow to the nation's war-time food-production program. It is estimated that in 1943, in which year elaborate preparations had been made for the efficient utilization of the spawning-run production, the mortality reduced the output of smelt by about 13 million pounds. The total loss through the present (1946) season can be set in the neighborhood of 50 million pounds.

The first indication of a recovery of the smelt came in 1945 when a small amount was produced commercially in Green Bay and numerous light runs occurred in streams tributary to Lakes Huron and Michigan. The general level of abundance in 1945, however, is believed to have been less than 10 per cent of that of “pre-mortality” years. Such information as is available for 1946 suggests considerable further improvement in this year. Given good survival of young, it is anticipated that a large rise in the abundance of smelt can occur in 1947 and that by 1948 or 1949 the size of the populations should no longer be influenced by the number of spawners available in preceding years.

The smelt from the Escanaba area of Green Bay were without exception significantly longer and heavier in 1944 and 1945 than were fish of corresponding age captured in the same region in 1941. Furthermore, three of four comparisons indicated significantly greater size in 1945 than in 1944. This improvement in growth rate is believed to have been associated with the reduction in the smelt population brought about by the 1942–1943 mortality.