The oviposition of unfertilized eggs depends on the presence of potential mates in a kissing bug

In most species, females face the risk of losing their oocytes if they remain unmated for a prolonged period. However, it remains unclear whether they have developed mechanisms to prevent the loss of fertilizable eggs according to the chances of finding a mate. We predict that virgin females will lay a lower proportion of their produced eggs when they perceive the presence of a male conspecific. We tested this prediction in the kissing bug Rhodnius prolixus, comparing the oviposition of virgin females randomly exposed to the presence of a potential mate or isolation over 10 days. We found that isolated females laid three times as many eggs as females exposed to the presence of a male. The dissection of females showed that both groups produced similar numbers of developed eggs, suggesting that differences in oviposition are due to changes in egg retention, rather than on egg production. We discuss the possible benefits of egg retention in virgin females, and how this process may contribute to their reproductive decision-making.

Methods
Treatments

Virgin females were individually placed in one of the compartments of the experimental arena and then randomly assigned to one of the following treatments: M+, in which a male was present in the other compartment of the arena (n = 22); and M-, in which the other compartment was empty (n = 22). The transparent acrylic barrier separating the compartments allowed the female to perceive the male through visual and olfactory cues, although physical contact between the insects was prevented .

Data analysis

To analyze differences in the probability of laying of produced eggs, we used a generalized linear mixed model (GLMM) with a logit link and a binomial distribution, with the treatment (M+/M-) as the explanatory variable and the number of laid and retained eggs for each female as the dependent variable. As overdispersion in the data was detected using the DHARMa R package (Hartig 2024), we included observation-level random effects (OLRE) to account for it. The model was parameterized in R using the cbind() function, grouping the number of laid (successes) and retained (failures) eggs.

For the study of the oviposition dynamics, differences in the number of laid eggs over time within and between treatments were analyzed. Daily oviposition data were grouped into four temporal intervals (days 1, 2–4, 5–7, and 8–10).  We used a GLMM with a repeated-measures design (RMD), with a log link and a Poisson distribution, with the treatment (M+/M-) and the oviposition days (1, 2-4, 5-7 and 8-10) as explanatory variables, the female as a random (intercept) variable and the number of laid eggs as a dependent variable. When the interaction between the main factors (treatment and time) was significant in the GLMM with repeated measures design (RMD), we analyzed the simple effects separately. Post-hoc Tukey’s tests were then used to compare treatments within each time interval and time intervals within each treatment that showed significant effects.

Differences in the total number of eggs produced between treatments were analyzed using a GLM with a log link and a negative binomial distribution, with treatment (M+/M-) as the explanatory variable and the total eggs produced as the response variable.

Additionally, to ascertain that females in both treatments had similar characteristics, we recorded their weight at the beginning of the experiment. No significant differences were found for the females’ weight between the treatments .The data were analyzed by means of parametric statistics after verification of the assumptions, i.e., normal distribution of residuals (Shapiro–Wilk’s test) and homogeneity of variances (Levene’s test). The distribution of female weights met the assumptions required for ANOVA. The weights of males in the M+ treatment followed a normal distribution.