Environmental induction and phenotypic retention of adaptive maternal effects

Abstract

Background: The origin of complex adaptations is one of the most controversial questions in biology. Environmental induction of novel phenotypes, where phenotypic retention of adaptive developmental variation is enabled by organismal complexity and homeostasis, can be a starting point in the evolution of some adaptations, but empirical examples are rare. Comparisons of populations that differ in historical recurrence of environmental induction can offer insight into its evolutionary significance, and recent colonization of North America by the house finch (Carpodacus mexicanus) provides such an opportunity.

Results: In both native (southern Arizona) and newly established (northern Montana, 18 generations) populations, breeding female finches exhibit the same complex adaptation - a sex-bias in ovulation sequence - in response to population-specific environmental stimulus of differing recurrence. We document that, in the new population, the adaptation is induced by a novel environment during females' first breeding and is subsequently retained across breeding attempts. In the native population, first-breeding females expressed a precise adaptive response to a recurrent environmental stimulus without environmental induction. We document strong selection on environmental cue recognition in both populations and find that rearrangement of the same proximate mechanism - clustering of oocytes that become males and females - can enable an adaptive response to distinct environmental stimuli.

Conclusion: The results show that developmental plasticity induced by novel environmental conditions confers significant fitness advantages to both maternal and offspring generations and might play an important role not only in the successful establishment of this invasive species across the widest ecological range of extant birds, but also can link environmental induction and genetic inheritance in the evolution of novel adaptations.

Figure 1

Short-term sex-bias in ovulation sequence in house finch females. Sex-bias in response to A) ambient temperature below egg-viability threshold during early breeding season in Montana (n = 86 nests), and B) late season nest mite infestation in Arizona (n = 110 nests). In both populations, there was no sex-bias in ovulation sequence in other parts of the breeding season. Asterisks show sex-ratios significantly deviating from parity. Coefficient of variation (cv) indicates variability in relative ovulation sequence of male and female eggs (see Methods).

Figure 2

The relationship between environmental stimulus (critical temperature days during oogenesis in Montana and number of mites at nest sites in Arizona) and response to the stimulus (number of biases in ovulation sequence) in A) first-breeding females in Montana (n = 93 females), B) first-breeding females in Arizona (n = 131 females). C) Estimated number (mean ± s.e.m.) of critical temperature days during oogenesis required to exert full response (three biases in ovulation sequence) across female's lifetime in Montana (n = 51 females), D) Estimated number (mean ± s.e.m.) of nest mites during oogenesis required to exert full response (four biases in ovulation sequence) across female lifetime in Arizona (n = 29 females). Note that the ordinate axes in C) and D) are scaled identically to the abscissa axes in A) and B) to show the full range of the stimulus. E) Biases (mean deviations ± s.e.m.) in ovulation sequence across three breeding episodes of the same females in Montana, and F) in Arizona. Bubble radius is proportional to the number of overlapping data points. Lines connect means that are not significantly different.

Figure 3

Estimated contour plots of offspring survival as a function of number of deviations (biases) in ovulation sequence. Response to A) critical temperature days in Montana population (n = 128 nests), and B) number of mites at nest site in Arizona population (n = 96 nests infested with mites). Note that the ordinate axis shows number of mites at the stage of egg-laying and this number increases greatly by the time nestlings hatch and mite-induced mortality occurs. Numbers show proportion of nestlings fledged out of the number of eggs laid.

Figure 4

Groups of oocytes similar in yolk (x-axis – Ward's minimum distance) in relation to oocyte' sex and ovulation order in first-breeding females. A) Arizona population under mite infestation conditions (n = 72 nests), B) Montana population with > 5 critical days during oogenesis (n = 63 nests), C) Arizona population under mite free conditions (n = 99 nests), D) Montana population with ≤ 1 critical days during oogenesis (n = 34 nests). Drawings show hypothetical arrangement of oocytes in the ovary that would correspond to sex-specific clusters in A) and B) or non-sex specific hierarchical arrangement in C) and D). Vertical bars on the left side delineate significantly distinct clusters.