Timing of first feeding and life‐history strategies in salmon: genetic data

摘要

The change in the size-frequency distribution from normal to bimodal over time is a well-known phenomenon in both natural (Heggenes and Metcalfe 1991; Nicieza et al. 1991) and cultivated (Metcalfe et al. 1989; Blanco et al. 1998) populations of Atlantic salmon. It is important due to the influence of bimodality in the subsequent smolting rates and in the timing of seaward migration (Bailey et al. 1980; Bailey and Friars 1994). Smolting rate is a trait of particular interest for both cultured and in wild populations and much of the interest has been focussed on studying the behaviour and growth of the morphotypes in the upper and lower modal groups in the bimodal size-frequency distribution (Thorpe et al. 1992). Although the factors that control life-history variation in Atlantic salmon are not fully understood, different studies indicate that the physiological state, the size and the growth rate of an individual will determine whether the animal adopts the early maturing or early migrant strategy the following year (Thorpe et al. 1998). Some studies have suggested a threshold size beyond smolting can occur in a given populations (reviewed by Thorpe et al. 1992). The developmental rate in the alevin phase and in the relative timing of first feeding can, on the other hand, influence the social status and the subsequent life-history strategy (Metcalfe and Thorpe 1992). Temporal variation of up to two weeks has been reported in the timing of first feeding within groups of Atlantic salmon in wild populations (Gustavson-Marjanen and Dowse 1983; Beall et al. 1994) and in cultivated groups (Metcalfe and Thorpe 1992; Brannas 1995). It has been reported that fish that has completed their alevin phase more quickly and were earlier to active first-feed, tended to be dominant over their later feeding siblings, grew faster and were more likely to smolt at an earlier age (Metcalfe and Thorpe 1992).Previous work on salmonid fish has shown a family effect on developmental rate during the alevin phase (Beacham et al. 1985; de March 1995) and on growth rate in freshwater (Thorpe and Morgan 1978). There are also evidence of isozyme pattern differences between fish adopting different life-history strategies (Carl and Healey 1984; Sánchez et al. 1994, Blanco et al. 1998, Presa et al. 1996).However, Atlantic salmon populations usually show lower levels of genetic variation at protein-coding loci and, in general, six loci (MDH-3,4*, mMEP-2*, IDHP-3*, sAAT-4*, IDDH-1* and IDDH-2*) account for more than 95 % of the total gene diversity in the species (Cross and Ward 1980; Stahl 1987; Davidson et al. 1989; Bourke et al. 1997). Allozymes have therefore a limited value as genetic markers and thus alternative methods are required in order to make more detailed analyses of genetic variation in this species. Currently, new molecular genetic markers, like RFLPs, AFLPs, RAPDs, minisatellite, microsatellite, etc., are being developed and between them, microsatellite markers exhibit attributes that make them particularly suitable for genetic characterisation; for example, they are very abundant, exhibit usually high levels of allelic variation, are codominant markers inherited in a Mendelian fashion and, in addition, only small amounts of tissue are required for analysis (Wright and Bentzen 1995; Jarne and Lagoda 1996).In this work, a characterization of genetic diversity of different groups into timing of first active feeding and life-history strategies in Atlantic salmon was carried out using 6 allozymes and 8 microsatellites markers.