4. Discussion
Using molecular markers, we could identify captive-bred juveniles of 56 experimentally produced families and compare them to wild-born juveniles of the same cohort. The captive-bred 0+ dominated in number (by a factor 5.3), had a significantly more balanced sex ratio, and were on average larger than wild-born 0+. The reason for these differences remains unclear but could be linked, for example, to different stress levels during embryogenesis, the timing of stocking relative to the timing of emergence of wild-born, or size differences at the time when exogenous feeding starts. Given these many possible reasons, it may even be a general rule that captive-bred and wild-born fish of the same cohort differ in growth and survival because of differences in early life history (Palejowski et al. 2022).
We used a panel of > 1 million SNPs to calculate the kinship coefficients of 47 parental combinations that resulted from our experimental breeding. These 47 kinship coefficients (minus one outlier) may well reflect the average inbreeding coefficient per experimental full-sib family in a population that does not seem to suffer from elevated levels of inbreeding, as concluded from measurements of hybrid vigour that included our study population (Clark et al. 2013; Stelkens et al. 2014) in crosses of populations that are genetically distinct (Stelkens et al. 2012). Nevertheless, the variation in kinship coefficients that we observed could be used to predict female growth in the wild. Females grew slower with increasing inbreeding coefficients. No such effects could be observed among males. We conclude that the susceptibility to inbreeding depression is sex-specific at this juvenile stage.
We sampled the juveniles about 6 months after release into the wild. Sexual maturity and first breeding are expected at the end of their second or third year of life. Therefore, the sex-specific effects of inbreeding that we observed cannot be explained by the sex-specific stress that is expected during the mating season. Moreover, because sex chromosomes of brown trout are largely homomorphic (Guiguen et al. 2019), as is typical for lower vertebrates (Beukeboom & Perrin 2014), they are not expected to contribute significantly to sex-specific inbreeding depression (Vega-Trejo et al. 2022). The effects we found here are therefore best explained by sex differences in life histories. In the case of brown trout, these sex differences are rather cryptic. Little is known about sex differences in morphometry or behaviour at such early life-history stages even though the brown trout is a common and well-studied species. However, recent studies on brown trout and grayling revealed that the sexes differ in the timing of gonad development. Females generally develop their gonads earlier than males while males in turn grow faster than females during that time (Maitreet al. 2017; Palejowski et al. 2022). The size difference that was found before in other populations (Palejowski et al.2022) could be confirmed in the present study but was overall small (around 3%). This difference was linked to the variance in kinship coefficients and may even be caused by the variance in kinship coefficients because the sex difference in size was not apparent in families with small inbreeding coefficients. We therefore predict that increased inbreeding within a population will accentuate the sex difference in growth.
The overall sex ratio in our re-captured sample was about equal, and there was no significant effect of inbreeding on recapture rates for the different families. Based on the rate of experimentally produced fish among the sampled ones (80.3%), the total number of fish that could be sampled, and the number of fertilized eggs that were used for this study, the overall mortality of the experimentally produced fish over their first 9 months of their life was 78.4% or less if some fish had escaped sampling. Because embryo mortality was low in a parallel study on the same families (Wilkins et al. 2017), the mortality in the present study reflects the acute stress during stocking and/or the selection during the fish’s first spring and summer in the wild. Somewhat comparable levels of mortality during these first months have been observed in nearby populations of brown trout (Palejowski et al. 2022), i.e., the level of selection in our study system seems not extra-ordinary.
As is typical for studies on wild populations, quantifying likely effects of emigration remained difficult. In our study system, upstream emigration was not possible. Downstream emigration into the larger stream (Rotache) was possible. However, if migration happens in this species, it typically starts at later developmental stages and is then often sex-biased, with females being more likely to migrate than males (Forseth et al. 1999; Nevoux et al. 2019). The most parsimonious explanation for the observed overall equal sex ratio in the captive-born fish, and the non-significant correlations between recapture rates and inbreeding, is therefore that there was no sex-specific mortality and no sex-specific emigration, and that inbreeding depression only affected growth but did not lead to increased mortality in the hatchery-produce fish. The pattern was different in the wild-born 0+ who grew smaller than the hatchery-produced 0+ and had a male-bias sex ratio, suggesting that wild-born females suffered from a higher mortality than wild-born males during their first spring and summer. It is possible that a combination of sex-specific inbreeding and larger hatchery-born competitors led to sex-specific mortality among the wild-born.
Inbreeding coefficients can show significant heritability in small and structured populations (Neff & Pitcher 2008; Nietlisbach et al.2016). This prediction is supported by the significant correlation that we found between parental inbreeding coefficients and the average kinship coefficients, i.e., the expected inbreeding coefficients of their offspring. If parental inbreeding coefficients predict offspring inbreeding coefficients, and if inbreeding depression during the spawning season affects intra- and inter-sexual selection (i.e., giving inbred individuals a selective advantage), natural spawning would be expected to reduce the average inbreeding coefficient of the next generation. However, the one extreme kinship coefficient that we recorded would not be avoid through effects of inbreeding on health and vigour because the parents of this sib group had average and very similar inbreeding coefficients, suggesting that they could have been brother and sister that would not be able to avoid each other without some form in kin recognition.
Our experimental breeding also allowed to test for general maternal and paternal effects on juvenile growth. We found significant maternal but not paternal effects, suggesting that juvenile growth is affected by maternal environmental effects such as egg size and egg content. The absence of significant paternal effects is either due to limited statistical power or suggests that heritability of growth is small when measured in juveniles recaptured from the wild. Paternal effects on offspring growth are, however, frequently observed in brown trout larvae when studied under controlled laboratory conditions, revealing significant heritability of growth in this species (Marques da Cunhaet al. 2019; Nusbaumer et al. 2019). The limited number of recaptured juveniles per each of the 60 full-sib families did not allow to test for possible effects of dam x sire interactions on growth.
In conclusion, inbreeding did not significantly affect mortality of juvenile brown trout that had been stocked into the wild as larvae. However, female growth during their first spring and summer in the wild was reduced with increasing inbreeding coefficients. No such effect could be observed in males who even could grew on average about 3% larger than females during that time. Inbreeding depression is hence sex-specific around the time of gonad formation and long before intra- or inter-sexual selection are expected to cause sex-specific (male-biased) inbreeding depression.