No abstract

Abstract. Measurements of egg number, egg size and total egg volume in relation to post‐stripped weight, made on the eggs stripped from 12 separately maintained farm stocks of 4‐year‐old rainbow trout. Oncorhynchus mykiss Richardson, revealed highly significant differences between populations when analysed by regression and covariance statistical techniques. All of the stocks showed increases in fecundity with increasing fish size, with 61% (coefficient of determination or r2) of the variation in fecundity of the 12 stocks explained by the common regression on parental fish size. Comparison of the regressions of the different stocks revealed that the rate of increase of fecundity with increasing fish weight was constant throughout the weight range (i.e. the regressions had similar slopes) although there were highly significant differences in elevation, with some stocks producing almost twice as many eggs as some of the others. By contrast, the relationship of egg size with fish weight was much more variable with three of the stocks showing no significant regression of egg size on fish weight and a further two stocks with data poorly fitted to the regressions. Of the remaining seven stocks only one showed a significant difference in slope of the regression with modest increases in fish weight being associated with large increases in egg size. Only four of the remaining stocks of fish showed significant differences in the elevation of their regressions, i.e. the egg sizes were significantly different after the potentially conflicting influence of fish size was partitioned using covariance techniques. Assessments of total egg volume produced far improved regressions on post‐strip weight with 75% (r2) of the variation in the pooled data of the 12 stocks being described by the line of the common regression. In general, there appeared to be a “trade‐off” between egg number and size with the most and least fecund fish producing somewhat smaller and larger eggs respectively. All 12 regressions of totai egg volume on fish weight showed a common slope. However, there were significant differences in elevation of the regressions, with some of the stocks of trout producing up to a 55% greater volume of eggs. There appeared to be no relationship between the differences in egg number, size or total egg volume and the type or geograpfiical location of the farms on which the stocks were held. Although it remains to be determined to what extent these differences in reproductive performance are Inheritable, it is recommended that these characteristics should be taken into account by hatcheries in tbe design of long‐term broodstock selection programmes for rainbow trout.