Figure 2

Outcrossing is limited at high NaCl concentrations. (A) Development from the L1 larval stage (24 ± 2 hours) to the time of reproduction (96 ± 2 hours) in varying NaCl concentrations did not affect male frequencies in the lab-adapted population (parental generation, P). Sex ratios in the following F1 generation were affected, however, with a high NaCl concentration reducing population male frequencies when compared to a low NaCl concentration (generalised linear mixed effects model (GLMM): z-ratio test for a significant difference at F1 among salt environments: |z| =12.2, P-value <0.001, residual d.f. =91; see Methods for justification of statistical modelling]. (B) Despite the large impact of high NaCl concentrations in hermaphrodite fertility from 72 ± 2 hours to 96 ± 2 hours (GLMM: main effect NaCl |z| =27.5, P-value <0.001, residual d.f. =123), outcrossed hermaphrodites had the same fertility as selfed hermaphrodites in high or low salt (post hoc Tukey comparisons at 305 mM NaCl: |z| =0, P-value =0.99; at 25 mM NaCl |z| =0.3, P-value =0.78). (C) Development to maturity since the L1 larval stage was retarded at 305 mM NaCl since there was a reduced proportion of fertile hermaphrodites in the lab-adapted population at this concentration (generalised linear model (GLM): main effect NaCl |z| =2.97, P-value =0.003, residual d.f. =6). After 72 ± 2 hours of development in 305 mM NaCl only 10% of the hermaphrodites were fertile when compared to the 40% in 25 mM NaCl (|z| =3, P-value =0.003). However, after 96 ± 2 hours differences among NaCl environments were no longer significant (|z| =0.001, P-value =0.99). For all panels, triangles, bars and standard errors depict ordinary least-square estimates; ***indicates P-values <0.001. d.f., degrees of freedom.