Growth reaction norms of domesticated, wild and hybrid Atlantic salmon families in response to differing social and physical environments

Table 5 Parameter estimates of the selected GLMM best explaining variation in survival in the restricted treatments, experiment II

Variable	Estimate	± SE	Z	P
Intercept (Farm, Restricted hatchery)	-10.51	2.63	-3.99	0.0001
Hybrid	-0.31	0.18	-1.68	0.09
Wild¹	-1.11	0.21	-5.37	<0.0001
Restricted semi-natural	-3.79	0.39	-9.72	<0.0001
Egg size²	20.31	4.39	4.62	<0.0001

Wald Z-statistics of the selected GLMM fitted using Laplace approximation, with tank nested within treatment as a random intercept and 2 x 2 (co) variance matrix allowing for heterogeneity of variance between families across treatments, i.e., random intercept and slope. SE; standard error of estimates. Differences in probability of survival between salmon of farmed, hybrid or wild origin were investigated by re-running the final model while excluding one of the genetic origins at a time, while multiple comparisons were corrected for by the Bonferroni correction, giving an adjusted significance level of 0.017. ¹ Wild relative to hybrid salmon: estimated value = -0.81 ± 0.22, Z = -3.68, P = 0.0002. ² The effect of egg size upon survival was no longer significant in the farmed and hybrid salmon after the Bonferroni correction: estimated value = 15.58 ± 6.66, Z = 2.34, P = 0.019.

ISSN: 2730-7182