Temporal and spatial variability in density, relative condition, gender composition and maturity status of Atlantic salmon (Salmo salar L.) parr in the Harry's River drainage system, insular Newfoundland

Spatial and temporal variability in population densities and relative (length- specific) condition of Adantic salmon parr in Harry's River were examined over the period 1987-1997. Apparent differences in growth allometry (slopes from log weight: log forklength regression) among sites and years were attributable to variability in stomach contents of the smaller fish and to sex and maturity status of the largest fish. Pooling of all weight: length data permitted identification of significant differences in relative condition factors (weight: length regression intercepts for specific sites/years) among all sites and years. No significant correlations were found between site-specific relative condition and parr density, food availability or physical habitat factors. The lack of correlation may be due to differences in the scale of the prey measurements versus parr feeding ranges. -- A large proportion of males in the tributaries were sexually mature or "precocious" which corresponded with a preponderance of females (71%) in returning anadromous adults. There was a clear trend for increasing percentage male composition in the streams with age and even at age 1+ more than half of the males were precociously mature. The declining proportion of females with increasing age class in the streams suggests that females made greater use of pond habitat than males. Precociously maturing males had higher mean fork lengths and mean weights than immature males at age 1+ but growth rates decreased in the older age classes. The relative length-specific condition of precocious male parr was significantly higher than that of immature male and female parr at ages 1+ and 2+ while relative length-specific condition of 3+ female parr was very close to that of precocious males, indicating increased weight gain for females prior to smoltification. -- Discriminant analysis was employed to construct a model that would predict sex and maturity status from the easily measured variables length and weight, and from length-specific condition which is calculated from these two parameters. The model was used to predict group membership for all unknown fish in 1997 to examine the relative distribution of precocious males throughout the watershed. The predictions of sexual maturity status provided by this model suggested that much of the historical variation observed in mean relative condition of parr populations among tributaries of Harry's River could be due to differences among sex and maturity status among sites.There was a positive correlation between predicted % precocity and observed site-specific condition. Those populations comprised of over 40% precocious males were mostly concentrated at sites above George's Lake, while the sites with the lowest (≤ 25.0%) proportion of mature male parr were located mostly in tributaries below George's Lake, suggesting that different life history strategies were being employed in the two regions. Such diversification should enable the population to adapt to a wider range of freshwater and marine habitat variation