Information
Fertility is an important prerequisite directly affecting the success of selective breeding programs. Unlike livestock species, where unfavourable genetic correlations between production and reproduction have been reported, this relationship has been understudied in aquatic animals. In this study we investigate genetic parameters and correlations for a series of milt quality traits and growth in Arctic charr (Salvelinus alpinus) from the closed nucleus of the Swedish national breeding program. This population has been subject to selection for approximately 40 years (nine non-overlapping generations), with notable genetic gains for growth rate but also low and variable fertilization success rates (Palaiokostas et al., 2021) that make breeding practice challenging .
Materials and Methods
Using state-of-the-art, fluorescence-based cell counting and a standardized protocol for Computer Assisted Sperm Analysis (CASA) we recorded sperm concentration and kinematic parameters from 1,106 male charrs over two spawning seasons. Variance components estimation was performed in a multi-trait fashion with REML using either pedigree information alone or a combination of pedigree and genomic information from ~1/4 of the individuals using single step methodology. Our models adjusted for effects of year and water temperature at the day of sampling when applicable. A similar multivariate analysis was also applied to assess covariance components across two sex-limited traits, namely egg-count and sperm concentration, together with female and male growth proxies.
In addition, using retrospective data from 576 crosses we developed a statistical framework to disentangle latent fertility traits and their respective genetic parameters under two Bayesian hierarchical models. Unobserved male and female fertility factors were considered in non-linear functions describing the fraction of spawned eggs that were fertilized and reached the eyed stage. Animal models for the underlying factors were used as sub-models and genetic variances were partitioned utilizing additive genetic relationships.
Results
The phenotypic variation in the studied traits was extensive and could explain part of the variation in fertilization success rates. Heritability estimates for milt phenotypes appear relatively low to moderate spanning from ~0.1 to ~0.3. At the same time, fork length, a phenotype used as a proxy of growth rate, had a relatively moderate to high heritability of ~0.6 and revealed negative genetic covariance with male fertility proxies. More specifically, genetic correlations of -0. 4 to -0.5 (depending on model) were yielded between sperm concentration and fork length while estimates around -0.2 were computed for between the latter and most sperm motility/velocity parameters. Lastly, fecundity, which is moderately heritable (h2 of ~0.4) show ed positive genetic covariance with female and male growth but weak negative genetic correlation with sperm concentration.
Genetic variance posterior estimates corresponding to the female and male fertility variables converged at high levels explaining from around half and up to ~90% of the total variance (though characterized by relatively large standard errors ).
Discussion
Our analysis suggests low to moderate heritability estimates for sperm quality traits. While this is in line with preliminary studies in the same populations (Kurta et al., 2022), here we extend by examining additional parameters to obtain insights on the genetic architecture of reproductive success in Swedish Arctic charr. Such parameters are egg count, latent variables for female and male fertility and the genetic covariance of growth rate with fertility. Opposing genetic correlation signs between fertility and growth for females and males reveal a likely complex genetic architecture governed by trade-offs and sex-specific genetic interactions with possible implications for the future of selective breeding practice in the closed nucleus.
Latent fertility variables seem to have a strong heritable component, with estimates notably higher than the heritabilities for fertility proxies. That said, the respective estimated breeding values for male fertility were positively correlated with those for key sperm phenotypes. Despite a certain level of uncertainty due to limited data dimensions, this might reflect that fertility is a non-linear composite of underlying factors, with only a subset of which having been used as proxies in this study.
References
Kurta K, Jeuthe H, de Koning DJ, Palaiokostas C (2022) Evaluating the potential of improving sperm quality traits in farmed Arctic charr (Salvelinus alpinus) using selective breeding. Aquac Rep 25:101234
Palaiokostas C, Jeuthe H, De Koning D (2021) Assessing the potential of improving growth and survival to the eyed stage in selectively bred Arctic charr (Salvelinus alpinus). J Anim Breed Genet 138:326–337