Appendix Figure 1. Rationale for mesh size used for sievingDelia platura pupaePupal weight (left panel) and pupal diameter of ‘normal’ weight pupae
for the H- and N- lines of D. platura. ‘Normal’ pupal weight was
calculated as the inter-quartile range (salmon portion of distribution,
left panel). Pupae with weight corresponding to this range were measured
at the wider-most region of pupal casing to obtain a distribution of
pupal diameters (right panel). The green line represents the mesh size
for the sieve used to remove the smaller individuals from the
experiments.
As small individuals tend to have decreased mating success (Benelli et
al., 2016; Rhainds, 2019), we aimed for removing smaller individuals
from both D. platura lines in an attempt to control for size
effects on mating probability. To determine the mesh size required to
sieve out the smaller pupae of both lines, we measured the diameter of
pupae of ‘normal’ weight (interquartile range; Sup. Fig. 1 left panel;
pupal weight data supplied by F. Fournier). We then chose an available
mesh size (1.7mm) closest to the minimal pupal diameter of both lines
(Sup. Fig. 1 right panel). While this mesh size was successful in
retaining all ‘normal’ size pupae for the N-line, it eliminated some of
the smaller H-line pupae of ‘normal’ size. As such, by retaining only a
larger proportion large H-line pupae, we would expect a higher and less
variable mating success within this line. However, we observed the
opposite trend; H-line females had a lower mating probability than
N-line females. Thus, sieving a greater proportion of small-sized H-line
individuals further enforces the amplitude of the effect of sex ratio
and density measured in our study.
Appendix Figure 2. Arena used in
the group sex ratio and density experiment
Arena used to evaluate the effect of group sex ratio and density on
mating probability, pre-oviposition period and egg hatchability for the
H- and N- lines of D. platura.
Appendix
Figure 3. Average (± SD) number of dead males replaced for each
treatment of the group sex ratio and density experiment
Average (± SD) number of dead males replaced for each treatment
evaluating the effect of group sex ratio and density on mating
probability, pre-oviposition period and egg hatchability for the H- and
N- lines of D. platura. A higher number of substituted males did
not correspond proportionally to the size of the effects observed in the
mating probability and pre-oviposition period analyses.
Appendix
Table 1. Output of statistical analyses for sex ratio and density
experiment
Estimates/Hazard ratios and corresponding 95% confidence intervals (CI)
of the effects of D. platura line (H and N), group sex ratio,
group density and their interactions on mating probability (Generalized
Linear Model) and pre-oviposition period (Cox Proportional Hazards). 8
treatments x 2 lines x 10 replicates = 160 experimental units