Environmental variation often induces plastic responses in organisms that can trigger changes in subsequent generations through non-genetic inheritance mechanisms. Such transgenerational plasticity thus consists of environmentally-induced non-random phenotypic modifications that are transmitted through generations. Transgenerational effects may vary according to the sex of the organism experiencing the environmental perturbation, the sex of their descendants, or both, but whether they are affected by past sexual selection is unknown. Here we use experimental evolution on an insect model system to conduct a first test of the involvement of sexual selection history in shaping transgenerational plasticity in the face of rapid environmental change (exposure to pesticides). We manipulated evolutionary history in terms of the intensity of sexual selection for over 80 generations before exposing individuals to the toxicant. We found that sexual selection history constrained adaptation under rapid environmental change. We also detected intergenerational and transgenerational effects of pesticide exposure in the form of increased fitness and longevity. These cross-generational influences of toxicants were sex-dependent (they affected only male descendants), and intergenerational, but not transgenerational, plasticity was modulated by sexual selection history. Our results highlight the complexity of intragenerational, intergenerational, and transgenerational influences of past selection and environmental stress on phenotypic expression.

We tested, using the seed beetle Callosobruchus maculatus, if divergent regimes of sexual selection (after 84 generations of continued selection) modulate transgenerational plasticity in response to environmental stress (pesticide exposure). The experimental design hence consisted of a 2 x 2 design in which we crossed the two selection regimes (monogamous, henceforth Mono, or polygamous, henceforth Poly) with exposure/lack thereof of non-lethal concentration of pesticide. Before beetles were subjected to environmental stress treatment, the different selection lines underwent two generations of common garden breeding. Common garden breeding was carried out, first by duplicating the number of inoculated beans collected from each line to give rise to the next generation, and then by establishing a spare set of eight populations that were all bred under polygamous conditions for two generations, while maintaining the selection experiment with the original populations. After those two common garden generations, and to obtain focal virgin individuals of known age, we isolated three times as many infested beans as beetles were needed in our pesticide experiment. The isolation procedure was identical to that employed by the selection experiment and consisted of extracting inoculated beans with only one egg on them from the containers 11 days after oviposition, and reallocating them individually in the perforated Eppendorf tubes to ensure the virginity of the adults. This isolation protocol also allowed us to know the exact age of the focal adult beetles since adult emergences from the bean were checked daily.

Five virgin males and five virgin females were randomly selected within each of the eight selection lines that underwent common garden breeding as the focal individuals to be allocated to each of the pesticide exposure levels (exposed and control) in our experiment to non-lethal concentrations of the pesticide deltamethrin (COMBO Deltamethrin 2.5% w/v, Sarabia). Pesticide application was restricted to these beetles (the parental generation, F0), but subsequent effects were tracked over three generations (F0, F1, F2). Thus, 160 virgin beetles of C. maculatus (80 females and 80 males), with different evolutionary histories regarding sexual selection regimes (Mono, Poly), were exposed to different levels of previously tested non-lethal concentrations of the pesticide, inside a flow cabinet for 24 h, at 23 ± 1 °C, 12:12 L:D and 35% RH. Two different pesticide treatments were given: 1) control treatment (C), without pesticide, exposed to distilled water; 2) pesticide treatment, 2 g/L deltamethrin in aqueous solution (P). In short, the exposure was done randomly assigning 5 individuals from each sex from each of the eight lines (4 Mono and 4 Poly) to each treatment, having a total sample size of 160 parental exposed beetles in our experiment. Individuals at the moment of pesticide/control exposure were between 1 and 4 days old.

For the pesticide exposure, the ends of cotton swabs were impregnated with 30 uL of the pesticide or water and put individually into 2 mL opened Eppendorf tubes, which would be filled with the evaporation of the pesticide solution shortly after closing the tubes and throughout most of the 24 h of exposure. Next, beetles were introduced to a second perforated Eppendorf tube with the bottom removed and stacked onto the first tube, with a mesh barrier between them, keeping the cotton swab in the lower compartment and avoiding direct contact of the beetles with the pesticide. After the exposure phase, beetles were individually allocated to 26 mL perforated plastic containers with ad libitum beans (around 60 beans per vial) to be used as an oviposition substrate. After the exposure phase, and until death, beetles were kept in walk-in climate chambers (Fitoclima 10000 EHF, Aralab) at a constant 29 °C temperature with 40% humidity and a 12 h/12 h light/dark cycle, which are the conditions at which all animals from the experimental evolution program had been cultured since the start of the selection experiment. Each individual shared the recipient with a non-exposed tester mate (male or female), all sourced from a standardized heterozygote tester line that was generated by crossing two near-isogenic lines that had been generated following 15 generations of full-sib mating (all individuals used in the generation of the near-isogenic lines were drawn from the stock population, i.e., they were individuals from outside the selection experiment). Mating and oviposition were allowed for 48 hours. Afterwards, focal males were individually reallocated to Eppendorf tubes to track their longevity, whereas mated females (focal females that were mated to tester males, or tester females who were mated to focal males) were kept in their container to determine longevity, fecundity, and lifetime reproductive success for each of them.

We recorded life-history traits (longevity, fecundity, and lifetime reproductive success) on the generation exposed to pesticide treatment (F0), and on their offspring (F1) and grand offspring (F2). The F1 and F2 generations were not exposed to pesticides, but conditions for isolation, housing, and mating were similar to the parental generation. The sample size doubled in each generation because we selected both a son and a daughter from each cross in each generation (160 focal beetles in F0, 320 focal beetles in F1, and 640 focal beetles in F2), hence allowing us to test for differences between the sexes, in the F0, F1, and F2 generations.

We recorded the longevity of all beetles by individually monitoring survival on a daily basis from adult emergence until their death. We estimated total fecundity by carefully checking the presence of eggshells on each bean within every vial, on day 11^th after oviposition. Fecundity was measured for the F0 generation only as our focus was primarily on total number of adult offspring produced along life (lifetime reproductive success, LRS), and fecundity measures could not be attained for F1 and F2 generations due to the large sample sizes for individuals assessed in those generations. LRS could be measured as viable eggs hatched in a 28-day interval. As mating and oviposition times were controlled in our experiment, containers were frozen 28 days after the death of the female (the last possible day for oviposition). For those females who lived longer than 14 days, containers were frozen always before the predicted emergence date of the next generation (corresponding to day 42 from the first oviposition date). LRS was measured for all beetles in all generations by accounting for the lifetime number of adults produced by each female (either focal females or tester females mated to focal males). Dry body weight was measured after death for all individuals, focal or tester, and in all generations, with a Sartorius Cubis MSA 6.6S microbalance (accuracy 0.001mg, Sartorius, Goettingen, Germany). Beetles were frozen at -20 ºC when found dead, for a period of several weeks until the end of the experiments. The animals were then thawed and dried for 1 week at 40 ºC (so as to remove variation in weight due to time elapsed between death and freezing) before being weighed.