| dc.description.abstracteng | Relationships between adult attributes and lifetime fitness have been previously explored in numerous studies, but although many adult attributes diversify during childhood or even prenatally, and although lifetime fitness is driven by survival until maturation, the pattern of early development and its consequences for lifetime fitness have been studied to a much lesser extent. Consequently there are several fundamental knowledge gaps regarding the ultimate and proximate mechanism behind variation in phenotypes and survival during the developmental period. Play activity, for example, was repeatedly emphasized to be the behavioural hallmark of mammalian development, and early theories about the evolution and function of play date to the 1870s. However, our current knowledge about the costs and benefits and the evolution of play is still marginal, and current theories about the evolution of play are based on the largely untested assumptions that play benefits motor skill acquisition, or that the developmental costs and benefits of play are rather negligible. Similarly, epigenetic prenatal maternal stress effects on offspring phenotype with considerable long-term consequences on adult phenotype have been well known for several decades even in humans, but the debate whether and how these effects may be evolutionary adaptive or maladaptive is still in its infancy, with the most prominent theory basing on largely untested assumptions that may hardly apply to natural conditions. Also, the existence of differentiated male-immature relationships even in some promiscuous mammals was shown a long time ago, but whether and how these relationships may benefit the development of the immature is still not well understood.
This thesis aims to fill some fundamental knowledge gaps on these topics by investigating the causes and consequences of variation in immature development in wild Assamese macaques (Macaca assamensis) with particular focus on immature growth. Growth is a central aspect of early development and the entire life history of an organism, and variation in growth rate may have strong consequences for life history and lifetime fitness. Growth is, therefore, assumed to be under strong positive selection and of high ontogenetic priority. Growth rates are highly plastic even within species, which may basically be caused by starvation avoidance in reaction to reduced energy availability, resource allocation to other, size-independent fitness traits and/or time constraints on maturation caused by a predictable increase in adult mortality. This thesis will address all of these possibilities. For this purpose, I measured monthly body sizes of all immatures of the group noninvasively via photogrammetry, collected behavioural data including detailed records of play behaviour on 17 immatures from birth onwards, recorded the first occurrences of 38 gross motor skills for each of these immatures, collected pre- and postnatal maternal faecal samples of all mothers for faecal glucocorticoid metabolite analysis and roughly assessed immature immune function from daily records of visible signs of eye infection during a two month outbreak of conjunctivitis. Additionally, monthly food availability data which are part of the log-term field project were integrated into the analysis.
First I tested the two most prominent hypotheses about the costs and benefits of play. The “surplus resource”-hypothesis bases on the largely untested assumption that play is never performed at the expense of growth but uses only surplus energy which remains after investment in growth such that the developmental costs of play are negligible. Previous studies have shown positive correlations between play and growth rates, but since these studies failed to control for energy intake, these positive relationships may be entirely mediated by energy intake. The “motor skill”-hypothesis postulates that locomotor play benefits the individual by accelerating motor skill acquisition, but also this assumption was never adequately tested since previous correlational studies did not control for reversed causality, that is, better motor skills may enable higher play rates. My results on Assamese macaque immatures contradict the “surplus resource”-hypothesis and support the “motor skill”-hypothesis. I show first-time evidence that after controlling for food availability, investment in locomotor play is strongly at the expense of growth and probably increases age at maturation. Additionally, locomotor play entails benefits by accelerating motor skill acquisition as revealed by a time-series analysis. Within this pattern, growth rates and the proportion of time spent in locomotor play were also constrained by food availability and influenced by the sex of the offspring. Males invested increasing food availability primarily in play in order to achieve faster motor skill acquisition and females in growth in order to achieve faster maturation, which may correspond to their adult sex roles.
Next I explored the effects of prenatal maternal stress (PreGC, in terms of faecal glucocorticoid metabolite level) on offspring phenotype. Two prominent theories were proposed to explain the long-term effects of PreGC or early adversity on offspring phenotype, which both predict accelerated reproduction and thus growth in reaction to predicted reduced life expectancy (“predictive adaptive responses”, PAR). Both theories differ in the mechanism of how offspring may predict its life expectancy. The “external PAR”-hypothesis assumes PreGC to inform the offspring about adverse prenatal environmental conditions which predict similar conditions during its adult life, but the applicability of this assumption to long-lived species and/or unpredictable environments has been questioned. The “internal PAR”-hypothesis postulates that early life adversity may result in adverse somatic states throughout life, which enables the offspring to predict its life expectancy from its current somatic state without reference to an external environmental forecast. Studies which test between these hypotheses are currently lacking which may be due to the difficulties that emerge from the largely identical predictions. My analyses reveal that in my study population, PreGC is negatively correlated to prenatal food availability and postnatal immune function and positively correlated to offspring growth and body size. The effects on offspring phenotype remained also after controlling for potentially confounding postnatal variables like postnatal maternal physiological stress, maternal caretaking style, food availability and investment in play. Since Assamese macaques are long-lived and evolved and live in a highly unpredictable environment, the PreGC-effects on offspring phenotype are not only in support of a PAR but in particular of the “internal PAR”-hypothesis because external PARs would be highly dysfunctional and maladaptive under such conditions and can therefore be excluded.
Collectively the current evidence on the relationship between PreGC and offspring growth is highly inconsistent and ranges from positive to negative effects even within the same species which has hampered broader generalizations. Hence in a next step, I developed a novel parsimonious framework that integrates different competing hypotheses and explains the entire range of observed effects. This framework postulates that prenatal maternal exposure to environmental stress results in a) reduced maternal investment and thus reduced extrinsic offspring growth during gestation and lactation (but not after lactation), and b) elevated PreGC-levels which epigenetically increase the intrinsic offspring growth rate. Thus if both processes coincide, they may largely cancel each other out during the period of maternal dependence while the growth rate after lactation is unaffected by maternal investment and may thus reflect the prenatally increased intrinsic growth rate only, ultimately resulting in overall accelerated growth.
I tested these predictions in a large-scale comparative analysis of published data across mammals (88 studies on 11 species ranging from rodents to ungulates and primates) and by analysing data from my field project. The results strongly supported my predictions but also revealed an unpredicted additional effect of gestational timing of maternal stress exposure. In the comparative analysis, early-gestational stress exposure resulted in unaffected offspring growth rates during gestation and lactation and increased growth rates after lactation, which was in accordance with my predictions. Prenatal stress exposure during the second half of gestation, however, leads to reduced offspring growth rates during gestation and lactation and unaffected growth rates after lactation, which resembles my predictions of reduced maternal investment effects and suggests that PreGC-effects on offspring growth are limited to the first half of gestation. Interestingly, all these results were independent of the species. The results of early-gestational maternal stress are also reflected in my results on Assamese macaques. Further analyses suggest that the “unaffected” growth rate during lactation is most likely the result of the combined, counterbalancing effects of increased PreGC and reduced maternal investment.
In my fourth study I investigated the pattern of male-immature relationships in Assamese macaques and whether and how they may benefit the immature. Assamese macaques are highly promiscuous, but it was previously shown that they still form differentiated male-immature bonds which partly reflect paternity. Building on that, I found that male-immature bonds are maintained by the immature and last beyond weaning and thus beyond the period of high infanticide risk. Immatures seek the proximity of their preferred male in particular if their mother is absent and if other unpreferred males are present, suggesting that male-immature bonds provide protection for the immature. The rate of mild harassment the immature received from other group members is not affected by the presence of the preferred male though, but the strength of the male-immature relationship is positively correlated to the agonistic support the immature receives from the male. How this support benefits immature development and/or fitness remains unresolved and is open to future research.
In summary, my results provide first time evidence for several fundamental questions on offspring development. I tested basic assumptions of prominent hypotheses on the evolution of play and confirmed the “motor skill”-hypothesis by showing time-series causality but rejected the basic assumption of the “surplus resource”-hypothesis by showing a resource allocation trade-off between investment in play and investment in growth. These novel findings may bring about a new hypothesis of the evolution of play which may coherently explain both within- and between-species variation in play rates via the relationship between the respective costs and benefits of play. I tested for the first time between competing hypotheses on epigenetic prenatal stress effects and provided results in support of an internal, somatic state-based PAR. I further provided and tested a novel integrative framework for prenatal stress effects on offspring growth which explains the previously highly inconsistent results and provides an entirely new perspective on the adaptive value of PreGC-effects. In particular, it contrasts the current perspective by suggesting that the short-term adaptive PreGC-effect is reflected in unaffected offspring growth rates during the period of (reduced) maternal investment whereas negative correlations between PreGC and offspring growth result from the absence of PreGC-effects on offspring growth. | de |