Survival of adult females
Of the 96 females in each treatment group, 94% in the control, 92% in the mating delay, and 65% in the nutritional stress group survived the 100 days of the experiment (Fig. 2a). The hazard of death for adult females in the nutritional stress group was 6.7 times greater than that of the control (95% C.I. 2.82 – 16), according to Cox proportional hazards analysis. The hazard of death for the mating delay group was 1.3 times that of the control (95% C.I. 0.46 – 3.8) (Tables S3.1, S3.2). The smoothed hazard function indicated increasing risk of death with age for the nutritional stress group (Fig. 2b) and this was supported by a better fit to the data using a Weibull survival model in the parametric analysis, compared with exponential (Weibull ω = 1). Parametric analysis also provided evidence that the risk of death increased with age in the mating delay group (Weibull ω = 0.996), but not the control group (Weibull ω = 0.348) (Tables S4.3, S4.4).