Immune-mediated change in the expression of a sexual trait predicts offspring survival in the wild

Abstract

Background: The "good genes" theory of sexual selection postulates that females choose mates that will improve their offspring's fitness through the inheritance of paternal genes. In spite of the attention that this hypothesis has given rise to, the empirical evidence remains sparse, mostly because of the difficulties of controlling for the many environmental factors that may covary with both the paternal phenotype and offspring fitness. Here, we tested the hypothesis that offspring sired by males of a preferred phenotype should have better survival in an endangered bird, the houbara bustard (Chlamydotis undulata undulata).

Methodology/principal findings: We tested if natural and experimentally-induced variation in courtship display (following an inflammatory challenge) predicts the survival of offspring. Chicks were produced by artificial insemination of females, ensuring that any effect on survival could only arise from the transfer of paternal genes. One hundred and twenty offspring were equipped with radio transmitters, and their survival monitored in the wild for a year. This allowed assessment of the potential benefits of paternal genes in a natural setting, where birds experience the whole range of environmental hazards. Although natural variation in sire courtship display did not predict offspring survival, sires that withstood the inflammatory insult and maintained their courtship activity sired offspring with the best survival upon release.

Conclusions: This finding is relevant both to enlighten the debate on "good genes" sexual selection and the management of supportive breeding programs.

Figure 1. Effect of the LPS injection on the number of days with courtship display in sire houbara bustards in 2007.

Values on the y-axis represent the pre-injection values (time 0) and the weekly post-injection values. Bars represent standard errors.

Figure 2. Positive correlation between offspring survival and sire traits.

(a) Change in sire courtship behaviour induced by the inflammatory insult. (b) Sire age. Both plots report the percentage of offspring survival per sire with the adjustment based on the predicted values of a generalized linear mixed model where sire and dam identities were fitted as random factors.

Figure 3. Timeline of the experiment.

The figure describes the different steps of the study, for sires (daily behavioural scans, LPS injection) and for offspring (birth, release, post-release monitoring).