Hybrid larch heterosis: for which traits and under which genetic control?

Abstract

Despite the interest foresters have for interspecic hybrid trees, still little is known about their quantitative genetics. This is especially true for the hybrid (HL) between Larix decidua (EL) and L. kaempferi (JL). Long-term, well-designed, multi-site experiments are necessary to estimate the parameters required for HL breeding programs. This paper presents the results from a diallel mating trial between 9 EL and 9 JL, set up in 3 contrasted sites. Growth traits (height, circumference), quality traits (wood density, stem form, heartwood proportion), and bud ush were measured from plantation to up to 18 years after plantation. Wood density and heartwood proportion were assessed using increment cores. We did a spatial analysis to take into account environmental heterogeneity at the tree level, and we tted a multi-trait, Bayesian MCMC (Markov Chain Monte Carlo) genetic model. Our study conrmed, in most situations, that HL expressed heterosis over its best parent for growth traits taking advantage of an early faster growth, with no loss in wood density. However, growth traits showed low levels of heritability. On the other hand, bud ush and stem exuosity had high heritabilities, and wood density was clearly under JL control. Site-dependent heritabilities were expressed by EL. Additive genetic correlations were presented. The traits with high heritabilities showed high correlation between their performances in pure species and in hybridization, as well as high across-site correlations. The discussion focused on the interest of these genetic parameters for the hybrid larch breeding programs.

Deciphering hybrid larch reaction norms using random regression

“…Besides the site effect, a substantial part of the individual variation was shown to occur between taxa, with hybrid larch being more plastic and at a higher mean level than its parental counterparts. Therefore, this increase in longitudinal plasticity appears as a broader picture of the heterosis that was observed on an integrative scale across years, namely in the total circumference (Marchal et al 2017).…”

“…Progenies were produced by control-crossing in the frame of a diallel mating design between 9 European larch ( Larix decidua , EL) parents and 9 Japanese larch ( L. kaempferi , JL) parents, producing pure species and HL full-sib progenies. The set-up is described further in Marchal et al (2017).…”

“…The model was fitted by Markov chain Monte Carlo (MCMC) with the same priors as in Marchal et al (2017) (Hadfield 2010; R Core Team 2017): parameter expansion was used on the genetic variance-covariance matrices, and flat improper priors were set on the permanent environment variance-covariance matrices as well as on the residual variance. All chains were 5.5 × 10 6 iterations long, with 5 × 10 5 iterations burn-in and a thinning of 5 × 10 3 .…”

“…The contrasted areas were assessed using tree height (at the last age available) spatial effect maps. The spatial effects were predicted as best linear unbiased predictions (BLUPs) from model ‘M1’ in Marchal et al (2017). In each pit, we measured soil horizon thickness, stone content, and we collected samples to assess the soil texture.…”

“…× L. kaempferi (Lamb.) Carr., and the reciprocal cross), a highly productive conifer cultivated for wood in Western Europe and North America (Marchal et al 2017). For this reason, phenotypic plasticity is suspected to play a key role in the construction of hybrid larch (HL) heterosis.…”