Body mass dataset available at Dryad: Males and females frequently differ in their rates of ageing, but the origins of these differences are poorly understood. Sex differences in senescence have been hypothesized to arise, because investment in intra-sexual reproductive competition entails costs to somatic maintenance, leaving the sex that experiences stronger reproductive competition showing higher rates of senescence.
However, evidence that sex differences in senescence are attributable to downstream effects of the intensity of intra-sexual reproductive competition experienced during the lifetime remains elusive.
Here, we show using a 35 year study of wild European badgers Meles melesthat i males Curvilinear definition statistics of sexual immorality higher body mass senescence rates than females and ii this sex difference is largely attributable to sex-specific downstream effects of the intensity of intra-sexual competition experienced during early adulthood.
Our findings provide rare support for the view that somatic maintenance costs arising from intra-sexual competition can cause both individual variation and sex differences in senescence. Senescence, defined as within-individual physiological deterioration with age, has been detected in Curvilinear definition statistics of sexual immorality wide variety of natural populations, but the causes of the often marked individual variation in senescence rates remain poorly understood [ 1 ].
A major source of variation in senescence rates is sex; females frequently have longer lifespans than males in humans and other mammals [ 2 — 4 ], and recent evidence has highlighted associated sex differences in senescence rates in a range of fitness-related traits, including reproductive success [ 56 ], body mass [ 7 ] and telomere length [ 8 ].
While sex differences in senescence rates are now comparatively well documented, the mechanisms that generate such differences remain poorly understood [ 9 ].
It has been hypothesized that sex differences in senescence rates arise from the differing strength of intra-sexual reproductive competition experienced by males and females [ 1011 ]. Evolutionary approaches Curvilinear definition statistics of sexual immorality this hypothesis have focused principally on the implications of the sex difference in mean mortality rates that often accompanies sex differences in the intensity of intra-sexual competition [ 10 ].
The sex that experiences stronger intra-sexual reproductive competition often shows higher mean mortality rates, which may thereby differentially weaken the force of selection against deleterious mutations or antagonistically pleiotropic genes acting in late life in that sex [ 12 ], leading to the evolution of faster senescence rates and shorter lifespans [ 10 ].
Evidence that males in polygynous species frequently do senesce faster than females [ 45 ] and that species with stronger degrees of polygyny show more marked sex differences in reproductive lifespan [ 3 ] is therefore broadly consistent with this view. However, support for a central role for sex differences in mortality rates is far from universal [ 613 ], highlighting the likely importance of other mechanisms.
From a mechanistic perspective, it has also been hypothesized that individuals may suffer resource allocation trade-offs between the expression of competitive morphologies and behaviour and their simultaneous allocation Curvilinear definition statistics of sexual immorality somatic maintenance, which might thereby lead to steeper senescent declines later in life in the sex that experiences stronger intra-sexual competition [ 1114 ].
This perspective highlights the potential limitations of focusing solely on the implications of mortality rates, and predicts instead a direct effect of the intensity of intra-sexual competition experienced during early adulthood on both individual variation in senescence rates and the extent of any sex difference observed in senescence rates [ 1 ].
However, evidence that variation in the intensity of intra-sexual competition experienced predicts individual variation in senescence rates is rare and derives solely from studies of a single sex [ 1516 ], leaving it unclear to what extent sex differences in senescence are also attributable to this same mechanism.
Specifically, if somatic maintenance costs arising from intra-sexual competition do contribute to sex differences in senescence rates, one would predict i higher rates of senescence in the sex that experiences stronger intra-sexual reproductive competition and ii that the magnitude of any sex difference in senescence rate observed should be predicted by the intensity of intra-sexual competition experienced during early adulthood by members of one or both sexes.
Here, we test both predictions using 35 years of longitudinal data from a long-term field study of the European badger Meles meles. We use 11 captures of known-age individuals to directly compare the age-related body mass trajectories of males and females, and then investigate whether the observed sex difference in body mass senescence rates is attributable in part to downstream effects of intra-sexual competition experienced Curvilinear definition statistics of sexual immorality early adulthood.
Variation in body mass explains considerable variation in both survival and reproductive success in mammals [ 17 ] and is recognized as a useful phenotypic indicator of somatic state [ 18 ], but the causes of variation in late-life declines "Curvilinear definition statistics of sexual immorality" body mass have received remarkably little attention [ 71920 ]. In the European badger, body mass is an important trait contributing to reproductive success [ 2122 ] and survival [ 23 ]. Senescent declines in body mass might therefore underpin recent observations in this species of both reproductive senescence [ 24 ] and sex-specific survival senescence [ 25 ]; albeit weaker than has been documented in some other species.
In high-density populations, there is also evidence of reproductive skew among females within a group [ 2128 ]. While the extent to which the observed patterns of parentage are a product of intra-sexual competition is not clear [ 27 ], a higher incidence of bite-wounding and mortality among males coupled with male-biased sexual size dimorphism in this species, strongly suggests that overt intra-sexual reproductive competition is more intense among males than females [ 2930 ].
Specifically, we test two key predictions: We also address the possibility of downstream effects on senescence rates of generalized foraging competition in early adulthood rather than intra-sexual reproductive competition per se ; following [ 31 ]by examining the influence of the total badger density experienced during early adulthood.
We apply a linear mixed-model approach throughout, which allows us to examine changes in body mass with chronological age while controlling for selective disappearance and terminal effects that otherwise obscure or exaggerate patterns of senescence [ 32 ]. We use data from the long-term Woodchester Park field study Gloucestershire, UKwhere the resident high-density European badger population has been continuously monitored since the s [ 33 ].
Each year, the boundaries of all social group territories in the 11 km 2 site were approximated by bait marking in spring [ 33 ], and badgers trapped at all active setts for two nights, four times per year. Badgers were anaesthetized and identified through a unique tattoo administered at first capture. Captures of individuals of unknown age those not identified as juveniles at first capture or unrecorded mass, sex, social group or sett location were excluded from our body mass analyses.
Bovine tuberculosis bTB infection status was determined for all individuals at every capture [ 34 ]. In a detailed investigation of the impact of bTB on badger body mass [ 34 ], the authors identified that individuals begin to lose mass at advanced stages of infection respiratory or dissemination stages.
To avoid such bTB-related body mass losses influencing the body mass dynamics documented here, Curvilinear definition statistics of sexual immorality captures of individuals at advanced stages of bTB infection were excluded from the analysis. Age is defined throughout as the number of days elapsed since 20th February in their first year of life, which reflects the mid-February peak in births as exact birth dates cannot be readily determined [ 35 ].
Body mass was used as the response variable in all analyses. Three random intercept terms were included in all models: Selective appearance and disappearance the non-random arrival in, or departure from, the dataset of individuals as age increases can influence the detection and apparent trajectory of senescence [ 39 ].
Selective appearance is not an issue in our analyses as all included individuals were caught in their first year of life. We controlled for selective disappearance by including in all models a linear effect of age at last "Curvilinear definition statistics of sexual immorality" ALCdefined as the age in days at which a badger was last captured [ 32 ].
Finally, as terminal effects a potentially age-independent change in the response term prior to death, due for example to sickness that leads to death can also complicate the detection and interpretation of age-related patterns [ 740 ], we included a binary variable reflecting whether or not an individual was in its last year of capture LYC. To avoid biasing ALC Curvilinear definition statistics of sexual immorality LYC estimates, all observations from individuals likely to have been alive at the end of the study those caught in the last 2 years were excluded from our analyses.
As the analyses used an observational dataset with a large number of candidate explanatory variables and demanded comparisons of nested and non-nested models, we implemented the information theoretic IT model selection approach using Akaike's information criterion correcting for small sample size AICc [ 41 ].
In the interests of parsimony, more complex models were removed from the analysis if a simpler nested version of that model attracted greater support a lower AICc [ 42 ]. Akaike weights were used to gauge relative support for each model in the top model set and were defined as the likelihood of a given model divided by the total likelihood of all candidate models in the top model set [ 41 ].