The ecology of gestational growth in a wild cooperative mammal

Abstract

In wild mammals, early postnatal growth strongly affects offspring survival and fitness, but little is known about the causes and consequences of variation in prenatal growth. We investigated whether gestational weight gains vary according to maternal traits and social and environmental conditions, and how prenatal growth affects the fates of the resulting offspring, using an exceptionally large sample of repeated pregnant body weight records from individually recognizable wild meerkats (Suricata suricatta). Pregnant meerkats' body weights remained stable during the first half of gestation and then increased linearly until they gave birth. Gestational weight gains were more rapid under favourable environmental conditions and when mothers were experimentally food-supplemented, suggesting that nutrition strongly determines prenatal growth. While social conditions and reproductive competition shape postnatal growth in many social vertebrates (including meerkats), these factors had a limited effect on prenatal growth, and adjustment to gestation lengths were modest and unrelated to social factors. Pups that grew faster in utero were heavier when they emerged from the birth burrow yet this rapid growth was not associated with shortened leukocyte telomeres, and they were consequently more likely to survive to adulthood. Broadly, we identified pronounced variation in gestational weight gains, which is largely driven by food availability and strongly predicts offspring birth weights and survival. Our findings also highlight constraints in the flexibility of prenatal growth and gestation lengths in this species, which may limit adjustments in response to prevailing social conditions, and enhance selection for flexibility in postnatal growth.

Keywords: gestation; gestational weight gain; maternal investment; meerkat; pregnancy; prenatal growth; telomeres.

FIGURE 1

Changes in body mass in pregnant wild meerkats. The body mass data for all pregnancies is plotted in grey (n = 381), with three representative pregnancies highlighted in dark grey to show among‐pregnancy variation. The predicted population‐level mean body mass (Model 1) is plotted in red, with shaded bands indicating the uncertainty in predictions when accounting for population‐level ‘fixed’ effects and the observation‐level ‘residual’ variance (red band; ±95% credible intervals [CrI]).

FIGURE 2

The effect of maternal, social, and environmental factors on gestational weight gain parameters in wild meerkats. Factors were modelled against three components of gestational weight gains: the average body mass at the start of gestation (Supporting Information S12), the timing of the onset weight gain (presented as the duration of gestational weight gain), and the rate of gestational weight gain. The full posterior distribution for each model estimate is presented, with point intervals providing the mean, 89% and 95% credible intervals. Squared predictors refer to those fit as quadratic terms, and parentheses refer to factor reference levels. The probability of direction (positive: p ₊; negative: p ₋) is indicated to the right of each distribution.

FIGURE 3

The effect of litter size on gestational weight gains in dominant and subordinate meerkats. Dominant females accelerated their gestational weight gain when gestating larger litters, while subordinate females did not (dominants: N = 309 pregnancies from 85 females; subordinates: N = 74 pregnancies from 63 females). Lines display the predicted body mass of mothers with litter sizes of 2, 4 and 6, with shaded bands indicating the uncertainty after accounting for population‐level ‘fixed’ effects only (±95% credible intervals). All other continuous predictors were held at their status‐specific population means.

FIGURE 4

The effect of food supplementation on the rate of gestational weight gain in wild meerkats. Pregnant females given additional food gained more body weight than control (unfed) females. The full posterior distributions are shown in red for each treatment (n = 14 females) with point intervals providing the mean, 89% (thick error bars) and 95% (thin error bars) credible intervals. The posteriors incorporate the population‐level ‘fixed’ effects only and reflect the predicted gestational weight gain at the mean litter size of 3.24.

FIGURE 5

Constraints on meerkat pregnancy lengths before and after the onset of gestational weight gains. For known‐length pregnancies (in which conception and birth were observed, n = 76 pregnancies from 40 females), longer pregnancies were achieved primarily by mothers extending the period from conception to the onset of gestational weight gains (‘inflection’), rather than by delaying giving birth (i.e. extension of the period from inflection to birth). Lines represent the predicted duration of the two phases of gestation (±89% (thick) and 95% (thin) credible intervals).

FIGURE 6

The fitness consequences of variation in prenatal growth on offspring development. (a) A meerkat pup emerging from the burrow at 3 weeks old. (b) Increases in pups' prenatal growth rates (estimated as the per capita gestational growth rate) were associated with increased body mass at emergence from the birth burrow, but (c) did not lead to pups having shorter leukocyte relative telomere lengths. (d) Pups that emerged heavier displayed higher survival to adulthood. Solid lines display the predicted mean response for dominant females when all other model covariates were held at their mean, with shading highlighting the 95% credible intervals conditional on population‐level ‘fixed’ effects. Points display the raw data (d) or the mean posterior estimates (b, c).