Results
Breeding and Egg development. The proportion breeding was significantly affected by pairing (Fisher’s Exact Test, p < 0.001). Pairings with A. texanum females bred 100% of the time and pairings with A. barbouri mothers bred 33% of the time (Figure 3). Egg laying locations or oviposition locations were also significantly affected by maternal species (Fisher’s exact test, p=0.0004). The paternal species did not have a significant effect on oviposition location (Fisher’s exact test, p=0.1906). Only A. barbouri laid their eggs underneath rocks, meanwhile only A. texanum scattered their eggs loosely throughout the aquarium (Figure 4.). Both readily laid eggs attached to vegetation. Oviposition began within 24 hours of pairing and all pairs that oviposited completed oviposition within 72 hours. Time of embryonic period was significantly affected by genotype (one-way ANOVA, F = 25.09, p < 0.001), however, only A. barbouri significantly differed from the other genotypes in post-hoc t-tests (Supp. Table 1.). A. barbouri had a much slower rate of egg development, nearly doubling the time of the other three genotypes (Figure 3.). The number of larvae per clutch was affected by genotype (one-way ANOVA, F = 5.453, p = 0.009, Table 1.), however, genotypes with the same maternal species did not significantly differ. Pairings with A. texanum mothers had nearly double the number of eggs than those genotypes with A. barbouri mothers (Figure 3.). Differences in initial larval size between genotype pairs were significant (one-way ANOVA, F = 13.08, p = 0.001, Table 1.), withA. barbouri♀ x A. barbouri♂ and A. barbouri♀ x A. texanum ♂ larvae being significantly larger than A. texanum♀ x A. texanum♂ and A. texanum♀ x A. barbouri♂ larvae. No differences in initial larval size were seen between hybrid and non-hybrid genotypes with the same maternal species (Figure 3.). Pigmentation was significantly affected by genotype (one-way ANOVA, F = 49.82, p < 0.001, Table 1.), however, the effect of clutch was significant, indicating significant differentiation between clutches (F = 7.01, p < 0.001, Table 1.). Only A. barbouri♀ x A. barbouri♂ larvae significantly differed between the genotypes, being much darker than the other genotypes. The hybrids were intermediate in pigmentation between A. barbouri♀ x A. barbouri♂ and A. texanum♀ x A. texanum♂ but were slightly more similar to their maternal parents.
Laboratory and Mesocosm Fitness Experiments. Size at metamorphosis in the laboratory was significantly affected by genotype (one-way ANOVA, F = 12.593, p < 0.001, Table 2.), and results were not influenced by the effect of clutch (p-value 0.203, Table 2.). The largest genotype at metamorphosis was A. texanum♀ x A. texanum♂ , followed by A. barbouri♀ x A. texanum ♂ , and the smallest genotype was A. barbouri♀ x A. barbouri♂ (Figure 3.). The effect of genotype on time to metamorphosis of laboratory raised larvae was significant (one-way ANOVA, F = 29.841, p < 0.001, Table 2.). On average A. texanum♀ x A. texanum♂ metamorphosed 40-60 days later than the other genotypes, which all metamorphosed at similar times (Figure 5.). Laboratory survival to metamorphosis was not significantly affected by genotype (one-way ANOVA, F = 0.853, p = 0.477, Table 2.), and results again were not influenced by the effect of clutch (F = 1.758, p = 0.165, Table 2.). The lowest survival was in A. texanum♀ x A. texanum♂ , which significantly differed from the other genotypes, all which had similar survival (Figure 5.). Time to metamorphosis, size at metamorphosis and survivorship to metamorphosis were all correlated, with increasing time leading to increased size and lower survivorship.
In the mesocosm experiment, size at metamorphosis was significantly affected by genotype (one-way ANOVA, F = 5.514, p = 0.008, Table 2.), and clutch effect did not influence these results (p = 0.132, Table 2.). The largest genotype was A. texanum♀ x A. texanum♂ followed byA. barbouri♀ x A. texanum ♂ , with the smallest genotype beingA. barbouri♀ x A. barbouri♂ (Figure 5.). Time to metamorphosis of mesocosm raised larvae was significant (one-way ANOVA, F = 23.116, p < 0.001, Table 2.). The A. texanum♀ x A. texanum♂ andA. texanum♀ x A. barbouri♂ genotypes took a much longer time to metamorphose than A. barbouri♀ x A. barbouri♂ and A. barbouri♀ x A. texanum ♂ (Figure 5.). Clutch effect on time to metamorphosis was also statistically significant (one-way ANOVA, F = 5.042, p = 0.007, Table 2.), indicating variation between clutches within genotypes, albeit the effect of clutch was much weaker than that of genotype. Additionally, survival to metamorphosis was significantly affected by genotype (one-way ANOVA, F = 8.154, p < 0.001, Table 2.), with no statistically significant effect from clutch variation influencing this result (F = 1.25, p = 0.312, Table 2.).Ambystoma texanum♀ x A. texanum♂ and A. texanum♀ x A. barbouri♂ both had significantly fewer individuals surviving to metamorphosis then the A. barbouri maternal parent genotypes. The survival differences were extreme, with the A. barbouri maternal genotypes having survivorship more than twice the amount of A. texanum maternal genotypes (Figure 5.).