Table 3.

Results of the linear mixed effects model on the ratio of males per female. There was no correlation structure necessary, but a varExp correction for heterogeneity was improving the model. The data (n = 5) were not transformed and “Day” was used as fixed effect. The ratios of males per female were significantly higher in the high-salt population. Over time the ratios declined in both treatments (Fig. 2B). In the model Ma/Fem is the comparison between high-salt and food-limited populations to normal-COMCO populations

	Estimate	SE	df	t-Value	P-value
Ma/FemCP	0.1151	0.0244	102	4.7146	0.0000***
Day × Ma/FemCP	−0.0018	0.0006	102	−3.1128	0.0024**
Ma/FemFP	0.0640	0.0717	12	0.8932	0.3893
Ma/FemSP	0.8022	0.1147	12	6.9913	0.0000***
Day × Ma/FemFP	−0.0002	0.0019	102	−0.0972	0.9227
Day × Ma/FemSP	−0.0057	0.0034	102	−1.6625	0.0995

Ma/Fem = ratio of males per female.

SP, high-salt population; FP, food-limited population; CP, Normal-COMBO populations.

***0, **0.001, *0.01, 0.05, 0.1, 1.