Changes in growth and in gonadal hormone status in the sea-water life of farmed Atlantic salmon (Salmo salar L.)
Sixty one year old (S1) and sixty two year old (S2) Atlantic salmon smolts derived from comparable egg stocks, were dye-marked and their growth and development individually monitored in a seawater cage for a period of 26 months. During that time, blood samples were withdrawn at regular intervals and their gonadal hormone content determined by radioimmunoassays. The growth of S1 fish, in both weight and length, was distinguished from that of S2s not only by its higher magnitude but also by its mode, S1s growing relatively more in length than in weight. These observations are discussed in relation to previous studies of growth bi-modality in freshwater. Growth was found to be highly seasonal. Although the period of maximum growth corresponded with maximum water temperature, sharp changes occurred in growth when the temperature was relatively steady. In spring, non-maturing fish grew relatively more in length than in weight and in the autumn more in weight than in length; the time of inflection was close to that of maximum day-length. Fish maturing after one sea-winter (grilse) were not recruited from the larger or faster growing members of the population. Grilse were larger than non-maturing fish of the same sea-age because they had experienced a surge of growth, postulated to be hormone induced, in the early stages of gonadal development. In the summer preceding spawning, maturing fish suffered a sharp decrease in specific growth rate. Changes in the digestive tract were rapidly reversed after spawning and are considered to be the result of inanition rather than its cause. Recovering fish (kelts) experienced a growth "rebound" in spring, showing higher specific growth rates than salmon of the same sea-age. A highly specific and sensitive radioimmunoassay technique was developed for the simultaneous determination of the three principal gonadal steroids oestradiol-17beta, testosterone and 11-oxotestosterone, together with rapid, but less specific individual assays for the first two of these, steroids. Use of these assays showed that in maturing female fish, blood serum levels of oestradiol-17beta reached maximum values of 1-3 microg/100 ml approximately one month before spawning, and testosterone titres peaked, at spawning time, at 4-8 microg/100 ml. In male grilse, serum titres of testosterone and 11-oxotestosterone reached their maximum values of 1-3 microg/100 ml and microg/100 ml in October and November respectively. Levels of the latter steroid remained high until March, long after the normal period of spawning. In salmon, early spring peaks in the serum concentrations of oestradiol-17beta and testosterone (in females) and of testosterone and 11-oxotestosterone (in males) were invariably associated with sexual maturation in the same year. Early peaces in serum oestradiol-17beta and 11-oxotestosterone but not in testosterone were noted in female and male grilse respectively. These observations suggest that the physiological decision to mature is taken during the late winter or spring preceding spawning. The incidence of grilse maturity was markedly higher in the experimental fish than in the control farm populations from which they were drawn. This is interpreted as the effect of stress on the decision to mature. Groups of fish from a population which had not matured as grilse were subjected to periods of food deprivation during winter and spring. Starvation and stress associated with handling in January or in mid February-mid March resulted in a significant increase in the proportion of fish maturing in the same year. Handling stress alone during mid March-mid April caused a significant increase in grilse maturation in a one sea winter stock. Stress supplied after that time did not significantly affect the percentage maturing as grilse. Administration of oral oestradiol for any monthly period from November to February induced a significant decrease in the proportions of sea-water smolts maturing as grilse, but not, apparently by a predominant affect on one sex.