Finally, we did not find any significant covariance between mean egg mass, carotenoid concentration and lysozyme concentration (P>0.5).
The mean egg mass was 12.70, 0.41, 0.46 and 20.36 g for C. serpentina, S. undulatus, B. duperreyi, and E. taeniura, respectively.
We calculated mean egg mass for each clutch and performed T-tests to examine the effect of female treatment on this variable.
Overall, there was no association between egg number, mean egg mass and the relative weight loss by mothers during the period of egg care (Table 1A).
The second GLM was run using the same explanatory variables, but with the mean egg mass of the 2nd clutch eggs as response variable.
Conversely, decreases in the weight loss by mothers entailed a negative association between mean egg mass and nymph number (Table 2A, Figure 2C).
Here, we show that correcting egg number and mean egg mass by female weight at egg laying allowed the detection of negative associations between egg quantity and quality and between egg quantity and egg care.
However, when correcting for variation in female weight at egg laying, the residuals of egg number were negatively correlated with both the residuals of mean egg mass (Table 1B) and the relative weight loss during egg care (Table 1B).
In particular, decreases in mean egg mass or in relative weight loss by mothers during egg care cancelled the otherwise positive association between egg and nymph numbers (Table 2A, Figures 2A and 2B).
To determine whether female condition possibly masks trade-offs in the measured traits, we then re-ran the above correlation tests using the egg number and mean egg mass corrected for female weight at egg laying.
