590 "U. S. BUREAU OF FISHERIES 



been reduced on that account. In such a case, the increased produc- 

 tion of 1922 may be quite as well explained by assuming possible 

 fluctuations in the five variables mentioned in the following para- 

 graph, as by assuming that fish have become more numerous. 

 Finally, it is not entirely probable that fish can increase under condi- 

 tions of pollution, virtually unrestricted and most intensive fishing, 

 and little propagation. 



It has been stated elsewhere in this paper that the available statis- 

 tics of the Great Lakes fisheries can not be statistically treated and 

 in these two tables the reason can be shown. 



1. None of the figures show the mesh of the gear employed. The 

 size of the mesh determines not only the kind of species that will 

 be taken but also the abundance of the species in the catch. The 

 data given on page 611 indicate that a fishery for deep-water Leu- 

 cichthys in Lake Ontario, which might soon fail with a 23^-inch mini- 

 mum mesh, M'ould flourish if the mesh were reduced 34 inch. Of 

 course, in Lake Erie the size of meshes used has not varied greatly 

 in the last few years, but there have always been two classes of gill 

 nets, one chiefly for whitefish and one for smaller species. Thus, 

 the increased catch may be due chiefly to the catch of smaller fish 

 by nets with smaller mesh. 



2. The length of the gill nets is given in Canadian waters, but not 

 the depth, which is an extremely important factor. For American 

 waters no dimensions whatever are given. The effectiveness of such 

 apparatus depends also to a very great degree on its state of repair, 

 the material of which it is spun, the fineness of its threads, etc. 



3. The method of employing the apparatus is nowhere reflected 

 in the statistics. On other pages it has been stated that the floating 

 of nets has greatly increased the catches. 



4. It is not known whether market conditions favored the capture 

 of all species throughout the season of each year for which statistics 

 are recorded. In 1920 thousands of pounds of blue pike and herring 

 could not be sold because the markets were glutted, and the fisheries 

 reacted accordingly. Such conditions have obtained at other times 

 and their eft'ects, of course, determine production. 



5. Even if gear, market conditions, etc., were stabilized and the 

 number of fish in the lake remained unchanged, the yield would 

 certainly vary from year to year as a result of other conditions, 

 especially those influen'ied by the weather. Being taken at about 

 5-year intervals, the American statistics do not permit evaluation of 

 this annual fluctuation, even if the data were given in sufficient detail. 



If, then, fish are not more abundant now than formerly, the in- 

 crease in or the maintenance of the general level of production must 

 be determined by some other factor, and the only other factor likely 

 to influence production is increased demand, which is reflected in 

 higher prices. 



Since there are no cities on the Canadian sliore of Lake Erie, and 

 since the populaton of the Province of Ontario may obtain its fish 

 supply from other lakes that are nearer, the Canadian fishermen are 

 chiefly dependent on American markets. Therefore the following 

 table of comparative prices for American waters will apply to the 

 Canadian side also. There are statements of value for each year in 

 the Ontario statistical bulletins, but the values given were constant 

 over long periods, and hence must have been fixed arbitrarily. 



