The aggregation of parents with offspring is generally associated with different forms of care that improve offspring survival at potential costs to parents. Under poor environments, the limited amount of resources available can increase the level of competition among family members and consequently lead to adaptive changes in parental investment. However, it remains unclear as to what extent such changes modify offspring fitness, particularly when offspring can survive without parents such as in the European earwig, Forficula auricularia. Here, we show that under food restriction, earwig maternal presence decreased offspring survival until adulthood by 43 per cent. This effect was independent of sibling competition and was expressed after separation from the female, indicating lasting detrimental effects. The reduced benefits of maternal presence on offspring survival were not associated with higher investment in future reproduction, suggesting a condition-dependent effect of food restriction on mothers and local mother–offspring competition for food. Overall, these findings demonstrate for the first time a long-term negative effect of maternal presence on offspring survival in a species with maternal care, and highlight the importance of food availability in the early evolution of family life.
Parental care has been described in a variety of species across the animal kingdom, wherein one or two parents attend their brood. Parental presence generally increases offspring survival through protection against predators and pathogens or through food provisioning. But such parental care often comes at costs to parents (see [1,2] for a review). For instance, it may reduce their capability to invest in future reproduction, delay successive broods or decrease their likelihood to survive until the next reproductive period [3,4]. To maximize fitness returns on investment, parents are expected to adjust their level of care in relation to variation in benefits to offspring and in costs to themselves . Hence, understanding which factors affect benefits and costs of parental care provides key insight into the conditions under which parental care evolved .
Whereas a growing number of studies have investigated how genetic conflicts and variation in relatedness among family members influence the costs/benefits ratio of parental care [2,7], little is known about the effects of external environmental conditions and their influence on the early evolution of parental care . A traditional hypothesis is that harsh environments favour the evolution of parental care  owing to the enhanced benefits to offspring survival [1,2]. However, such environments are also expected to exacerbate the cost of care and reduce the value of current offspring for parents, who may consequently favour investments in future reproduction to the detriment of current offspring [5,9,10]. In line with this idea, positive associations between resource availability and parental care have been reported in a few altricial species where offspring cannot survive without care (reviewed in ). While such patterns provide information about when parents should stop caring owing to enhanced costs, these studies are of limited relevance to understand how environmental quality shapes the early evolution of parental care and family life [4,11,12].
Here, we tested whether maternal presence under restricted food conditions influenced offspring survival and female investment in future reproduction in the European earwig, Forficula auricularia. This species is ideal for our question since mothers protect their clutches against natural enemies, provide food to their young and their presence improves offspring (nymph) survival under ad libitum food conditions. Furthermore, nymphs can survive and feed for themselves in the absence of mothers, and females can lay a second and final clutch a few weeks after having tended first clutch nymphs [4,13,14]. To assess whether maternal presence possibly decreases offspring survival by exacerbating the effects of sibling competition (partly mediated by sibling cannibalism ), we also manipulated brood size  and siblings’ relatedness  in addition to maternal presence.
2. Material and methods
We set up a total of 132 experimental groups of nymphs under restricted food availability, in which we manipulated the presence of mothers, the number of nymphs (brood size) and their relatedness. Maternal presence and brood size were manipulated by splitting 44 clutches of one-day-old nymphs (= clutch of origin) into three experimental groups: two groups of 10 (small group) and 20 (large group) nymphs without mother, and one group of either 10 or 20 nymphs with their own mother (see the electronic supplementary material, table S1). Nymph relatedness was manipulated by controlling the number of mating partners their mothers previously had access to: 23 females were singly mated and 21 females were multiply mated each with a different male (i.e. successively mated with four unrelated males) . Adult earwigs used in this experiment were the first laboratory-born generation produced by individuals sampled in May 2009 in Dolcedo, Italy.
At set up, the 132 groups were maintained in small Petri dishes for 16 days, which corresponds to the approximate duration of family life in earwigs . On day 16, we determined the proportion of nymphs still alive and the proportion of surviving nymphs that had reached the second juvenile instar (= developmental rate). All nymphs were subsequently transferred to medium Petri dishes for 15 days, and then to large Petri dishes from which the proportion of nymphs surviving until adulthood was later determined. Simultaneously, tending mothers were isolated in medium Petri dishes on day 16 to quantify the occurrence and the size of second clutches . The fresh weights of mothers were measured at egg hatching and on day 16.
To ensure food restriction, we provided each group with a standardized amount of food every six days (diet composition in ) and removed the left-over food three days after supply (food was generally finished in two days). The quantity of food was adjusted according to the age of the nymphs, with 60, 120 and 240 mg while in small, medium and large Petri dishes, respectively. Females had access to ad libitum food once isolated for second clutch production .
Survival and developmental rates were tested using general linear mixed models (GLMMs) with binomial error distributions, while the absolute number of emerging adults was tested using GLMMs with a Gaussian error distribution. Each model was computed using maternal presence, brood size and mating treatment as fixed factors, and the clutch of origin as the random factor. To assess whether variation in the benefits of maternal presence (BMP) on offspring survival until adulthood was related to female food intake and future reproduction, analyses of covariance were used with BMP as a covariate, brood size and mating treatment as fixed factors, and either female relative weight gain on day 16, number of second clutch eggs produced or relative investment of females into second clutch (= number of second clutch eggs divided by the number of total eggs ) as the response variable. BMP was calculated by subtracting the survival rates of nymphs without mothers from the ones with mothers, between groups of same size and same clutch of origin. All statistical models were conducted using R v. 2.14.0.
The survival rate of nymphs until adulthood was significantly lower in groups tended by mothers (figure 1a; t = −5.53, d.f. = 86, p < 0.0001) and in large groups (t = −7.82, d.f. = 86, p < 0.0001). The above effects on survival rate until adulthood translated into a smaller number of adult produced in groups tended by mothers (figure 1b, t = −6.21, d.f. = 86, p < 0.0001) but not in large groups (t = −0.33, d.f. = 86, p = 0.75). Mating treatment neither significantly influenced the survival rate of nymphs until adulthood (figure 1a, t = −0.18, d.f. = 42, p = 0.86) nor adult numbers (figure 1b, t = 0.21, d.f. = 42, p = 0.83). No interaction significantly influenced the two above traits (all p > 0.26).
The survival rate of nymphs until day 16 (i.e. when tending mothers were removed) was not significantly influenced by maternal presence (t = −1.34, d.f. = 86, p = 0.19), brood size (t = 0.81, d.f. = 86, p = 0.42), mating treatment (t = 0.96, d.f. = 42, p = 0.35) nor by any interaction (all p > 0.27). By contrast, the developmental rate of nymphs measured at day 16 was lower both in presence of mothers (figure 1c; t = 9.08, d.f. = 86, p < 0.0001) and in large groups (t = −10.14, d.f. = 86, p < 0.0001). There was no significant effect of the mating treatment on developmental rate (figure 1c; t = 0.39, d.f. = 42, p = 0.70) and none of the tested interactions was significant (all p > 0.70).
BMP was significantly and negatively correlated with the female weight gain until day 16 (figure 2; r = −0.36; likelihood ratio, χ2 = 6.23, p = 0.013; other factors and interactions: all p > 0.15). However, BMP did not predict maternal investment in future reproduction, as it was not significantly associated with the number of second clutch eggs (r = 0.08, LR, χ2 = 0.29, p = 0.59; other effects and interactions: all p > 0.10) or with the relative investment of females in second clutch (r = 0.14, LR, χ2 = 0.81, p = 0.37; other effects and interactions: all p > 0.09).
Models for the evolution of parental care and family life generally assume that parents pay a fitness cost for tending their brood, whereas their presence provides fitness benefits to offspring [5,6]. Here, we demonstrate that under food restriction, the usual BMP for offspring (approx. 15% increase in the survival rate ) are erased: females reduced offspring survival by 43 per cent. This result is at odds with the traditional hypothesis that harsh environments should favour the evolution of parental care . Furthermore, the above effect emerged after the period of family life, which excludes the immediate influence of females on offspring survival (e.g. through filial cannibalism ), but emphasizes their critical action on developmental processes affecting long-term survival. This interpretation is further supported by the negative effect of maternal presence observed on nymph developmental rate.
Parents are predicted to adaptively reduce offspring numbers when (i) it limits sibling competition over resources and consequently improves the survivorship of the remaining offspring, and/or when (ii) it allows parents to invest the resulting gain of energy into future reproduction [1,10,16,17]. Our results support neither of these two hypotheses. First, the presence of mothers did not modify the level of sibling competition, as there was no interaction effect between maternal presence and brood size on nymph survival. Secondly, we found that BMP reduction, albeit associated with an increased gain of weight in females, did not translate into larger maternal investment in second clutches. One reason for the latter result may be that the ad libitum access to food after day 16 in the experiment allowed females to invest in future reproduction without being strongly affected by earlier food restriction. Another non-mutually exclusive explanation could be that BMP reflects a condition-dependent effect of food restriction on females, where those exhibiting low intrinsic condition compete more intensely with their offspring for food. In line with this hypothesis, we found considerable variation in how maternal presence influenced the offspring survival rate (figure 2), with nine females (20%) enhancing offspring survival despite food restriction.
Sibling competition influenced offspring survival independently of maternal presence, a result not fully in line with burying beetles where maternal presence exacerbates sibling competition . Living in large broods reduced survival rate of nymphs until adulthood but did not affect the absolute number of adults produced. Because all groups had access to a similar amount of food, this finding suggests a negative association between per capita food availability and sibling cannibalism in earwig families . Conversely, variation in sibling relatedness had no significant effects on nymph survival. Provided that multiple mating translated into multiple paternity, this result suggests that preferential siblicide between nymphs from different clutches  does not reflect genetic kin recognition mediated by competition between siblings from different patrilines .
To conclude, this study demonstrates that harsh environments erase the benefits of parental care in a species with facultative care, possibly through mother–offspring competition for the limited resources. In such species, and at an early evolutionary stage, resource availability therefore affects the nature and extent of selection on parents versus offspring and can generate conditions where parental presence become detrimental to offspring fitness and parental care not adaptive.
We thank J. W. Y. Wong, H. Klug and an anonymous reviewer for constructive comments on this manuscript. This study was financed by the Swiss National Science Foundation (grant no. PP00-119190 to M.K.).
- Received February 20, 2012.
- Accepted April 3, 2012.
- This journal is © 2012 The Royal Society