What does not kill you makes you stronger? Effects of paternal age at conception on fathers and sons

Abstract

Advancing male age is often hypothesized to reduce both male fertility and offspring quality due to reproductive senescence. However, the effects of advancing male age on reproductive output and offspring quality are not always deleterious. For example, older fathers might buffer the effects of reproductive senescence by terminally investing in reproduction. Similarly, males that survive to reproduce at an old age might carry alleles that confer high viability (viability selection), which are then inherited by offspring, or might have high reproductive potential (selective disappearance). Differentiating these mechanisms requires an integrated experimental study of paternal survival and reproductive performance, as well as offspring quality, which is currently lacking. Using a cross-sectional study in Drosophila melanogaster, we test the effects of paternal age at conception (PAC) on paternal survival and reproductive success, and on the lifespans of sons. We discover that mating at an old age is linked with decreased future male survival, suggesting that mating-induced mortality is possibly due to old fathers being frail. We find no evidence for terminal investment and show that reproductive senescence in fathers does not onset until their late-adult life. Additionally, we find that as a father's lifespan increases, his probability of siring offspring increases for older PAC treatments only. Lastly, we show that sons born to older fathers live longer than those born to younger fathers due to viability selection. Collectively, our results suggest that advancing paternal age is not necessarily associated with deleterious effects for offspring and may even lead to older fathers producing longer-lived offspring.

Keywords: Lansing effect; life-history trade-offs; pleiotropy; senescence; terminal investment; viability selection.

Figure 1. Males who conceive at older ages have longer lifespans on average.

Violin plots show the smoothed distribution of data, while dark lines and shaded areas show regression lines and 95% CI. Lower Y-axis limit set to 25 because no deaths occurred before 25 days of age. The sample size of fathers across the six PAC treatments (from 4 to 74 days) is 25, 25, 26, 26, 37, 5, respectively.

Figure 2. Males who mated at 4 days of age (N = 25) did not have a significantly different survival probability compared to the unmated experimental males (N = 300).

“+” indicates censored individuals. The shaded area shows 95% CI. Dotted lines show age at median survival probability.

Figure 3

(A) Males who mated at 60 days of age (N = 37) had a lower survival probability than males who mated at 4 days of age but survived beyond 60 days of age (N = 14). (B) Males who mated at 74 days of age (N = 5) had a lower survival probability than males who mated at 4 days of age but survived beyond 74 days of age (N = 6). The dotted lines show age at median survival probability.

Figure 4

(A) PAC interacted with paternal lifespan to affect the probability of siring an offspring (p_s). (B) PAC affected the number of offspring sired (N_s: after excluding fathers who did not produce offspring) in a quadratic way. Shaded areas represent 95% CI. Sample size of fathers across the six PAC treatments (from 4 to 74 days) is 25, 25, 26, 26, 37, and 5, respectively.

Figure 5. An increase in PAC led to an increase in the lifespans of sons.

Each dot represents the lifespan of a single son; shaded areas show 95% CI. See Supplementary Figure S5 for means and SE. The sample size of fathers across the six PAC treatments (from 4 to 74 days) is 25, 25, 26, 26, 37, and 5, respectively.