Abortions
The probability of abortion increased with age across all treatments
(models without age: control ω = 0.000; mating delay ω =0.001; and nutritional stress ω = 0.00) (Tables S6 1-3). The
increase in the odds of abortion with age were similar across treatments
(from the model with lowest AIC: control 1.072, C.I. 1.054 – 1.094;
mating delay 1.053, CI 1.025 – 1.086; and nutritional stress 1.047, CI
1.034 – 1.061) (Fig. 2, S6 4-6). Females in the nutritional stress
treatment had higher probability of abortion at any age compared with
the other two treatments (Fig. 2).
For all treatments, including a random intercept and slope in models
resulted in a singular fit indicating that models were overfitted. We
did not therefore assess evidence for individual variation in the effect
of maternal age on probability of abortion. There was, however, strong
evidence for variation among individual mothers in the probability of
abortion for the nutritional stress treatment (model with random
intercept ω = 0.985; model without ω = 0.015). This was
not the case for the control (model with random intercept ω =0.289; model without ω = 0.711) or mating delay treatments (model
with random intercept ω = 0.368; model without ω = 0.631)
(Tables S6 1–3). Plots of the raw data for this analysis, and analyses
described below, can be viewed in S7 File.