Causes of adaptive differences in age-dependent reproductive effort

Abstract

Sexually selected ornaments are among the most spectacular traits in nature. Indeed, the extreme costs associated with producing sexual traits seem to play a crucial role in their evolution by enforcing honest levels of advertisement: only males with high levels of acquired resources (or high ‘condition’, as it is known in the literature) can afford to produce extravagant signals, a phenomenon which maintains signal reliability in a constant environment. In my thesis I examine many implications of this condition-dependent model of ornament and preference evolution for variation in age-dependent allocation to sexual signals and other life history traits. In Chapter 1, I review theoretical implications of condition-dependent signalling for life history and sexual selection theory. I note that a universal cost of expenditure in sexual advertisement is metabolic in nature: metabolites used to fund ornament expression are by definition unavailable to other life history traits that compete for a limited resource pool. This universal constraint on expenditure does more than maintain honesty (as noted above), however: the reliance of sexual displays on high levels of nutrient acquisition may help maintain genetic variation in sexual signals that would otherwise be eroded by strong mate choice, and without which the selective basis for good-genes choice would disappear. Three mechanisms in particular probably help to maintain genetic variation in acquisition. 1) Because acquiring resources and converting them efficiently to useful forms depends on the high function of many biochemical pathways, condition is undoubtedly highly polygenic, which slows the erosion of genetic variation under strong directional selection by females (especially in the presence of epistatic interactions). 2) The highly polygenic nature of condition also presents a large target for mutation, which continually restores variation at the loci under selection. 3) The many loci underlying condition may also be particularly sensitive to environmental heterogeneity in time or space. By favouring the most ornate males, females acquire high performing genes for their offspring, regardless of the precise allele combinations that have conferred the ability to acquire resources. Selection on specific alleles is liable to fluctuate over time or space whenever allelic performance is strongly context-specific. I close by noting the considerable challenges in advancing research on sexual selection and life history allocation, including the fact that two key processes central to life history (acquisition and allocation) are latent variables that interact in complexways and are intrinsically difficult to measure empirically. In the remainder of my thesis I conduct a series of experiments involving decorated crickets, Gryllodes sigillatus, which are useful models for studying life history because they enable precise measurement of male reproductive effort. Male G. sigillatus face important allocation decisions owing to the highly polyandrous nature of females, and the substantial costs involved in signalling and mating. Chapter 2 examines sex differences in age-dependent reproductive effort as a function of diet and development stage. I reared outbred crickets using four combinations of diet nutritional quality, and studied the effects of these combinations on male and female reproductive effort (calling effort in males and fecundity in females) and longevity. While I expected males to be more sensitive than females to variation in diet and developmental changes in its quality, I actually observed the opposite: males in all treatments increased calling effort over time, exhibiting consistently positive covariance between calling effort and longevity across treatments. By contrast, the relationships between female reproductive effort and longevity changed dramatically across treatments, and females who lived to intermediate ages had the highest fecundity. Although my results support sex-specific selection on life history allocation over time, a compelling additional explanation for my findings relates to the strategic role of calling for achieving male fitness. In the absence of positive feedback from potential mates, perhaps male allocation to sexual advertisement is careful and only increases gradually as a function of accumulating metabolic resources and increasing risk of intrinsic mortality. Alleles underlying condition are expected to be particularly sensitive to environmental heterogeneity. While this sensitivity may help maintain additive variation in male quality (which is essential for the sustenance of adaptive good-genes mate choice, as noted in Chapter 1), too much environmental sensitivity could also underiii mine the signal value of the male trait. For example, if there are strong genotypeby- environment interactions (GEIs) for sexual advertisement, in a rapidly changing environment females risk favouring a male whose alleles are no longer best suited to current conditions. This problem is particularly pressing for animals like crickets where males exhibit a behaviourally plastic sexual display (such as calling), and so may dynamically adjust signalling effort over time. In Chapter 3, I used inbred lines of decorated crickets to quantify age and diet dependent genetic variation in male signalling. I demonstrate that while genetic correlations across diets were quite strong for morphological traits, correlations between measures of the male sexual trait rapidly approached zero as I increased the distance in time (i.e., across widely spaced ages) or diet (i.e., comparing more dissimilar dietary histories) between samples. While extrapolating from my laboratory experiments to nature is difficult, my findings nevertheless cast doubt on the value of behaviourally dynamic signals (such as cricket calls) for reliably indicating genetic quality in realistically complex environments. In Chapter 4 I used physiological assays to evaluate factors affecting metabolite storage and use over time in decorated crickets. I manipulated the acquisition ability of all males using artificial diets that varied linearly in nutrient quality, and manipulated access to female mates over the course of the second week of adult life. By sacrificing crickets at key stages before and after manipulating the diet and social environment, I was able to estimate changes in stored metabolites, and relate these changes to calling effort and longevity. During the first week of adulthood (in the absence of females), higher diet quality significantly increased calling effort and storage of lipid, glycogen, and carbohydrate (but not protein). The presence of females increased both the probability of calling and the amount of calling during the second week, whereas diet quality only improved calling effort. By the end of the second week, calling effort had decreased, even by high quality males in the presence of females, suggesting a depletion of resources. Furthermore, the loss of condition during week 2 covaried with calling effort during the previous week irrespective of diet. Males who started the second week in high condition lost more glycogen and carbohydrate than rivals; meanwhile, lipid accumulation covaried positively with calling effort during week 2. The contrasting patterns of storage and use for lipids compared to the ‘quick-release’ metabolites (glycogen and carbohydrates) affirms starkly distinct functions for the different storage components, and underlines the importance of specific physiological measures in life history research. Finally, in the general discussion, I attempt to synthesise my thesis’s contributions to the study of life history trade-offs involving behavioural sexual displays. I argue that my work may have strong implications for the general honesty of male advertisements, and invite further research focused on quantifying interactions between the genotype, age, and environment in natural systems. I also question the prevalence of adaptive age-dependent plasticity in sexual advertisement, arguing for the parsimony of a more mechanistic and non-adaptive explanation for variation among populations and taxa in age-dependent signalling: males vary in signalling effort primarily as a function of constraints on energy expenditure, rather than because they are carefully saving resources for future use.