Fig. 4: Phenotypic and genetic relationships between short- and long-term heat stress tolerance. | Nature Communications

Fig. 4: Phenotypic and genetic relationships between short- and long-term heat stress tolerance.

From: Selective breeding enhances coral heat tolerance to marine heatwaves

Fig. 4

Offspring selected for short-term heat stress tolerance (F1-2018) and subsequently subjected to both short- and long-term heat stress exposures. a Phenotypic trait correlation between short- and long-stress tolerance for the F1 cohort (in terms of DHW50), showing a significant positive correlation between these two traits (Linear Mixed Model fit shown on plot with 95% confidence intervals, Z = 2.4, and bivariate animal model residual trait correlation, rR = 0.48 (0.18 SE), Z = 2.650/42, P > 0.05). b Selection for short-stress tolerance did not yield changes in long-stress tolerance. This is shown as distributions of offspring long-stress tolerance, grouped by their parents short-stress tolerance (low or high), with no significant difference in ΔDHW50-long between LL and HH families based on Generalised Linear Mixed Model (Z = −0.650/46, P > 0.05). c Relationship between parent midpoint short-stress tolerance and their offspring’s long-stress tolerance based on z-scored data (points). The linear regression slope (β1, the mean and 95% credible intervals of slope posterior distribution), and the predicted regression (mean and 95% credible intervals) show the trend is no different than zero (i.e., a flat line). A 1:1 relationship (dashed line) would suggest that the same genetic controls are present for both traits. The lack of a significant genetic correlation was also shown from a bivariate animal model (rG = 0.06 (0.66 SE), Z = 0.09, P = 0.90). d Progression of BSI for each F1 offspring (faint lines and points) throughout the long-term heat stress exposure (degree heating weeks – DHW) corroborated the non-significant genetic correlation between short- and long-stress tolerance. Cohort-level sigmoidal dose-response curves are grouped by parental short-term heat stress tolerance showing no significant difference between LL and HH families based on Generalised Linear Mixed Model (Z = −0.71500/1495, P > 0.05, error bands are 95% confidence intervals). Source data are provided as a Source Data file.

Back to article page