Surrounding pathogens shape maternal egg care but not egg production in the European earwig

Surrounding pathogens shape maternal egg care but not egg production in the European earwig Abstract Pathogens are ubiquitous in nature and typically entail major fitness costs in their hosts. These costs can be particularly important when individuals exhibit poor immune defenses, as it is often the case during early developmental stages. Hence, selection should favor parental strategies limiting the risks of pathogen exposure and infection in their offspring. In this study, we investigated 1) whether females of the European earwig Forficula auricularia avoid areas contaminated with spores of the entomopathogenic fungus Metarhizium brunneum prior to and at egg laying, as well as 2) whether spore presence entails an increase in females’ investment into both pre-hatching forms of care and clutch quantity and quality. Our results first show that females did not avoid contaminated areas prior to and at egg laying. However, females returned to their eggs faster in presence of living spores compared to UV-killed or no spores. They were also more likely to construct a nest when in presence of both living and UV-killed spores (but only in one studied population). Finally, we found that spore presence did not influence maternal investment into egg grooming, egg gathering and egg defense, as well as into clutch quantity and quality. Overall, our results demonstrate that earwig females do not avoid contaminated environments, but could mitigate the associated costs of pathogen exposure by adjusting their level of egg care. These findings emphasize the importance of pathogens in the evolution of pre-hatching parental care and, more generally, in the emergence and maintenance of family life in nature. INTRODUCTION Pathogens are considered as a major threat in nature, because they often entail large fitness costs in their hosts (Schmid-Hempel 2014). To limit the risks of pathogen exposure and infection, individuals from both vertebrate and invertebrate species have developed a broad spectrum of defenses (Siva-jothy et al. 2005; Schmid-Hempel 2014). The most studied defenses rely on the immune system of the potential host, which typically detects the presence of pathogens in the organism, prevents their development and ultimately kill and purge them from the host (Siva-jothy et al. 2005; Beckage 2008). However, many non-immunological defenses may also help hosts to fight against pathogens (Parker et al. 2011). Among them, hosts have been shown to change some of their behaviors to prevent direct contacts to pathogens (de Roode and Lefèvre 2012). This can be done, for instance, by avoiding infected areas and limiting the consumption of pathogen-infected food (Villani et al. 1994; Meyling and Pell 2006; Ormond et al. 2011), and by prophylactically or curatively consuming natural components with antimicrobial properties, a process called self-medication (de Roode et al. 2013; Bos et al. 2015). Interestingly, defenses against pathogens are not only expressed by the hosts themselves (thereafter called personal immune defenses) but can also come from surrounding conspecifics and group members (Schmid-Hempel 1998). Over the last decades, these external defenses have been mostly studied in colonies of eusocial insects, such as bees, ants and termites (Cremer et al. 2007; Meunier 2015). In these species, this so-called social immunity is often illustrated by allogrooming or hygienic behaviors, during which workers remove parasites from the cuticle of another group member (Reber et al. 2011) or clean the nest from waste materials, such as corpses, left-over food and feces (Bot et al. 2001; Jackson and Hart 2009; Diez et al. 2012). The expression of social immunity—which can complement personal immune defenses – is considered as a central process in the evolution of eusocial species (Meunier 2015), since the obligatory and frequent contacts between their related group members can dramatically increase the risks of pathogen exposure and transmission (Pie et al. 2004; Stroeymeyt et al. 2014). Social immunity, however, is not exclusive to eusocial systems and can be found in less derived forms of group living, such as simple family units (Cotter and Kilner 2010a; Meunier 2015). In these units, parents generally provide multiple forms of care to their eggs and/or juveniles (Smiseth et al. 2012) and, because eggs and juveniles often exhibit an underdeveloped immune competence (DeVeale et al. 2004; Vogelweith et al. 2017), parental care often includes protection against pathogens. For instance, mothers can protect their eggs against microbial infections by applying antimicrobial secretions to their surface, as reported in blennies, grass gobies and fringed darters (Knouft et al. 2003; Giacomello et al. 2006; Giacomello et al. 2008). Similarly, parents may cover the nest with feces exhibiting antibacterial activity to prevent the development of pathogens next to the juveniles, a phenomenon found in the burying beetle and the European earwig (Rozen et al. 2008; Cotter and Kilner 2010b; Diehl et al. 2015). Finally, females of the beewolf digger wasp Philanthus triangulum have been shown to transfer symbiotic bacteria to the brood cell cover (Kaltenpoth et al. 2005) and these bacteria to be taken up by the larvae and incorporated into their cocoon to serve as antimicrobial protection (Kroiss et al. 2010). Whereas these maternal behaviors typically protect offspring against pathogens, it remains unclear whether these behaviors are curative/therapeutic (i.e. help fighting against pathogens already in contact with offspring) or prophylactic (i.e. help preventing potential contacts between pathogens and offspring). Disentangling between these two processes is however crucial to better understand on which level pathogen-induced selection operates in the evolution of parental care. In this study, we investigated whether maternal egg care is a prophylactic process in the European earwig Forficula auricularia. Specifically, we tested whether the presence of pathogens in the environment triggers changes in females’ investment into pre-hatching egg care (i.e. social immunity) and/or egg production. In this insect species, mothers guard their eggs for several months over winter (Lamb 1976). During egg development (but also during the subsequent development of juveniles), females express a wide range of protective behaviors, such as grooming, clutch displacement and rearrangement, as well as aggressions against predators (Lamb 1976; Meunier and Kölliker 2012; Boos et al. 2014; Koch and Meunier 2014; Thesing et al. 2015). Whether mothers can detect the presence of living pathogens in the environment (and not only on the eggs, see Boos et al. 2014) and prophylactically adapt their behaviors remains, however, unknown. Moreover, a recent study revealed that some life history traits could be population-specific in earwigs (Ratz et al. 2016), begging the question of whether prophylactic defenses against pathogens might differ between populations. We conducted two experiments to test the expression of three prophylactic defenses in earwig females originating from three independent populations. First, we tested whether females avoid areas infected with an entomopathogenic fungus to lay their eggs. Second, we investigated whether females modify their investment into egg quantity and quality (i.e. egg size and weight) when the fungus is present in their direct environment. Finally, we tested whether females change their level of pre-hatching care when the nesting area (but not necessarily the eggs) contains this fungus. We used spores of the generalist entomopathogenic fungus Metarhizium brunneum (formerly M. anisopliae), a natural and potentially lethal pathogen of F. auricularia (Günther and Herter 1974; Kohlmeier et al. 2016) also known to infect and reduce the viability of eggs in many arthropods (Samuels et al. 2002; Gindin et al. 2009; Nozad-Bonab et al. 2017). In the first experiment, we recorded the movement and egg laying behaviors of females preliminary set up in areas covered with a spore solution on one side and with a spore-free solution on the other side. If mothers prophylactically limit the exposure of their eggs to pathogens, we expected females to avoid contact with fungal spores both before and at egg laying. Similarly, females laying their eggs on the spore-exposed side should be more likely to relocate them afterwards. In the second experiment, we measured the level of five forms of egg care behaviors (nest construction, egg gathering, egg defense, egg abandonment and egg grooming) in females previously set up in areas fully covered with either living spores, UV-killed spores or no spores. We predicted that females increase their investment into egg care only in presence of living spores. Finally, we measured the number, mean volume, mean weight and hatching success of the eggs produced in the two experiments. If females can assess the risks of pathogen infection for their eggs (and subsequent juveniles), we expected that females reduce the risks and/or costs of this infection by prophylactically increasing their investment into egg quantity and/or quality when spores were present in the environment. If females manage to mitigate the costs of pathogen exposure, we also predicted egg-hatching success to be independent of pathogen presence. MATERIAL AND METHODS Animals and pathogen All the females used in the following experiments were the first laboratory born generation of individuals collected in 2014 in 3 independent populations: Vincennes (France; 48° 51′ N, 2° 26′ O), Girona (Spain; 41° 59′ N, 2° 49′ O), and Montblanc (Spain; 41° 23′ N, 1° 10′ O). At adult emergence, these females were maintained with unrelated males in groups of 50 individuals (balanced sex-ratio) to allow them to mate freely (Meunier et al. 2012). Each group was set up in a plastic container (37 × 22 × 25 cm) grounded with moist sand, containing egg carton as shelters, and maintained under standard conditions (constant 20 °C, 14:10 h light:dark cycle, 60% humidity). Three months later, each female was isolated in a Petri dish (9 cm diameter) grounded with moist sand, and subsequently maintained under winter conditions to allow egg production (gradual temperature change from 15 °C to 5 °C during 3 weeks, with constant 60% humidity and complete darkness). Ten days later, the temperature was increased to 10 °C for 2 more weeks and finally to 15 °C for the rest of the experiments to allow egg development. After egg hatching, each family was kept at room temperature (23 °C) until the end of the following experiments. Groups and isolated females were fed twice a week with an ad libitum amount of a standard diet mainly composed of pollen, cat food and carrots (see detailed composition in Kramer et al., 2015). The conidia (spores) of the entomopathogenic fungus M. brunneum used in the following experiments were obtained from a genotyped strain isolated from soil samples in Switzerland (Reber and Chapuisat 2011). To ensure the virulence of this strain to earwigs, this fungus has been grown on F. auricularia for 3 generations prior to their use in this experiment. Specifically, earwigs were infected in a 1-ml spore solution of M. brunneum (108/ml) conidia solved in 0.05% Tween 20 and held in a climate cabinet at 25 °C in small Petri dishes (5.5 cm diameter) filled with moist sand. After the death of the exposed individuals, corpses were rinsed 3 times with 1 ml of concentrated bleach and distilled water to prevent contamination by other microbes. Afterwards, the surface sterilized insects were kept in new Petri dishes sealed with parafilm in the climate cabinet (25 °C) until growth of the fungus appeared. The resulting spores were rinsed from the body with distilled water and applied on potato extract glucose growth medium (Roth). Growing fruit bodies were finally collected with a sterile scalpel, diluted in 0.05% Tween 20 solution and then used to infect new individuals. This process has been repeated 3 times to obtain the spores used in the following experiments. Question 1: Do females avoid infected sites to lay their eggs? To investigate whether females avoid laying their eggs in a pathogenic environment, we transferred a random sample of 37 females (n = 11 Girona, n = 18 Montblanc and n = 8 Vincennes) to individual Petri dishes (9 cm diameter, grounded with humid sand) divided in two parts of equal size. One side was previously sprinkled with 0.55 ml of a spore solution of M. brunneum (107/ml) solved in Tween 20 (0.05%, Roth), whereas the other side (control) was previously sprinkled with the same amount of a spore-free Tween 20 solution. A direct exposure to a solution of 107 spores per ml is known to entail a 20% adult mortality after 25 days (Kohlmeier et al. 2016). The 2 sides were separated by a cardboard barrier (8.7 × 1 cm) covered with aluminum foil to avoid any passive diffusion of the applied solutions between the two sides, while allowing females to easily move from one side to the other. Each experimental Petri dish was then checked on a daily basis in order to 1) count the number of days each female stayed on each side until egg laying, 2) record the side on which females deposited theirs eggs and finally to 3) count the number of times the clutches were moved to the other side during egg development. All these measurements were done blind regarding the treatments (i.e. the experimenter did not know which side of the experimental Petri dish contained the pathogen or the control). Each female was provided with an ad libitum amount of standard food until egg laying (detailed food composition in Kramer et al. 2015), as females stop feeding once they produce eggs (Kölliker 2007). Note that the pathogen and control solutions were applied twice (once just before the start of the experiment and once 1 month later) to make sure that each treatment lasts for the entire experiment. We statistically analyzed whether females avoided pathogenic environments for egg laying using 2 successive steps. In the first step, we conducted 2 Generalized Linear Models (GLM) with binomial error distribution to determine whether females’ choice for each side (i.e. preference for or avoidance of the contaminated side before and at egg laying) depended on clutch size and/or on their population of origin. In the first GLM, the proportion of days each female spent on the contaminated side before egg laying was entered as response variable (using the cbind function in R), while egg number, population (Girona, Montblanc, Vincennes) and the interaction between these 2 variables were used as explanatory variables. In the second GLM, the side chosen for egg laying (1 = contaminated side, 0 = control side) was entered as response variable, while the 2 factors described above were also used as explanatory variables. Because neither egg number nor population was significant in these 2 models (see Results), we then pooled the data across populations to conduct the second step. In this second step, we used a 1-sample student’s t-test to analyze whether females spent overall more days on the control compared to contaminated side (i.e. ratio of time on the control side compared to the value 0.5), as well as a binomial test to analyze whether the number of females that had laid eggs on the contaminated side was overall lower than the number of females that had laid eggs on the control side. All statistical analyses (here and below) were done with R v3.4.0 (http://R-project.org) loaded with the packages car, lmerTest, survival and exactRankTests. Question 2: Do females adjust their pre-hatching care to the presence of pathogens? To investigate whether females adapt their investment into pre-hatching egg care to the presence of pathogens in the environment, we used a random sample of 123 females (n = 32 Girona, n = 65 Montblanc, n = 26 Vincennes; all different from the ones used to address question 1) distributed among three treatments. The experiment started by isolating these females in Petri dishes grounded with moist sand and entirely sprinkled with 1.1 ml of one of the 3 following solutions: 1) M. brunneum spore solution (107/ml 0.05% Tween 20; n = 11 Girona, n = 22 Montblanc, and n = 9 Vincennes), 2) UV-killed M. brunneum spore solution (107/ml 0.05% Tween 20; n = 11 Girona, n = 21 Montblanc and n = 10 Vincennes) or 3) spore-free 0.05% Tween 20 solution (n = 10 Girona, n = 22 Montblanc and n = 7 Vincennes). We used both the UV- killed spores and Tween solution as controls to disentangle the effects of pathogen activity (UV-killed versus living spores) from the effects of pathogen presence (spore-free versus living spores). The UV-killed spore solution was exposed to UV light (254 nm) for three minutes in the UV irradiation system Bio-Link 254 (Vilber). The UV treatment was efficient enough to inhibit the growth of M. brunneum spores on growth medium (results not shown). To mimic the process conducted in Experiment 1, the initial application of each solution in the Petri dish was done right before female isolation and a second application one month later. Note that the presence of spores in the environment did not affect females survival rate, as 6 (12.0%), 7 (13.5%), and 6 (12.2%) females died before egg laying in the treatments with living spores, UV-killed spores and no spores, respectively. No female died after egg laying. We measured 5 forms of pre-hatching maternal care: egg gathering, egg defense, clutch abandonment, nest construction, and egg grooming. 1) Egg gathering indicates the propensity of a female to gather its eggs after these were experimentally spread out in the nest. This measurement was done 5 days after egg laying. At that time, we confined each female within its Petri dish, then spread a random selection of 30 eggs (or all, if fewer were available) evenly within a circle (~3 cm diameter, including the site where the eggs were originally laid) and subsequently recorded the seconds between the mother’s first movement out of the confinement area and the re-assembly of the eggs within one body length of the female. Note that the eggs not involved in this measurement were kept in a small closed plastic dish in the meantime. 2) Egg defense revealed females’ reaction to a simulated predator attack and was measured ten days after egg laying. This measurement was done following a standard method (Thesing et al. 2015), in which females were poked with a glass capillary on the thorax (1 poke / 2 s) until they abandoned their clutch (i.e. moved more than 2 body lengths away from it). The values of egg defense were the number of pokes the females sustained before leaving the eggs. 3) Clutch abandonment was measured subsequently to egg defense and was used to quantify how long a mother abandons her clutch of eggs after a simulated attack. The value of egg abandonment was the time (in seconds) between the moment where each female left its nest due to a simulated predator attack (see above) and the moment where females returned to their clutch and touched at least one egg. 4) Nest construction reflected whether a female invested into the construction of a nest at day 15 after egg laying. Following a method already developed in this species (Meunier et al. 2012), we considered females to have constructed a nest when they dug a pit to the ground of the Petri dish and put their eggs into the pit. Finally, 5) egg grooming was recorded 15 days after egg laying, just after checking for females nest construction. At that time, each female was observed for 46 min in its original Petri dish using a scan sampling method (one scan every 2 min, i.e. 23 scans per female). To increase the chance of observing egg grooming (Boos et al. 2014), females involved in this measurement were separated from their clutch during 30 min before the measurements. Because earwigs are nocturnal, all behavioral tests were performed under red-light. All behavioral measurements were also done blindly regarding the treatment. We used one GLM with binomial error distribution, 2 General Linear Models (LM) and 2 Cox proportional hazard regression models to analyze the 5 measured forms of egg care. In these models, either egg gathering (Cox model), egg defense (LM), clutch abandonment (Cox model), nest construction (1 or 0; GLM) or egg grooming (LM) was used as a response variable, whereas treatment (living spores, UV-killed spores or control), population and their interaction were entered as explanatory factors. Note that the 2 Cox proportional hazard regression models allowed for censored data—that is, females that did not gather all their eggs or did not return to their clutch at the end of the observation time. When applicable, pair-wise comparisons between populations or treatments were tested using Tukey contrasts (for LM and GLM) and the survdiff function in R (for Cox models). Question 3: Do females adjust clutch quality and quantity to pathogen presence? Finally, we investigated whether females adjust the number, volume, and weight (proxies of egg quality, Koch and Meunier, 2014) of their eggs to the presence (enforced or not) of living pathogens in the environment. To this end, we measured these three traits in the clutches of the 36 females used to address question 1 (n = 11 Girona, n = 18 Montblanc and n = 8 Vincennes) and the 86 females that were in the living spores and spore-free control solution treatments in the second experiment (Question 2; n = 22 Girona, n = 45 Montblanc and n = 19 Vincennes). The number of eggs produced was counted 3 days after the first egg laying, because egg production typically takes three days in this species (Koch and Meunier, 2014). At that time, we also weighed a random subset of 10 eggs per female to the nearest of 10–3 mg using a micro scale (Pescale MYA 5), as well as measured the mean volume of the same 10 eggs (or all, if fewer were available). The mean egg volume per clutch was obtained by measuring the length and width of each of the 10 eggs to the nearest 10–3 mm and then using these values with the following formula (Meunier and Chapuisat 2009): egg volume = ((4 × pi)/3) × (egg width/2)2 × (egg length/2). All size measurements were done through a binocular microscope (Leica, Type DFC425, Leica Microsystems) with ×20.0 magnification that was run using the Leica Application Suite software (Version 4.5.0, Leica Microsystems). We finally estimated the hatching success of each of these 122 clutches by counting the total number of nymphs that eventually hatched. To control for the non-independence of egg number, volume and weight, we first conducted a Principal Component Analyses (PCA) to obtain non-correlated principal components (PC) reflecting single or combinations of egg traits. The resulting and selected PCs were then analyzed separately using LMs in which population, treatment (presence or absence of living spores on the area receiving the eggs) and type of arena (partly or fully covered with spores, i.e. Question 1 or 2, respectively) were entered as fixed factors. Independently of clutch properties, we finally analyzed the egg-hatching success using a GLM with binomial error distribution corrected for over dispersion. In this model, the hatching success was used as a response variable (entered with the cbind function in R). Moreover, population, whether females laid their eggs in Petri dishes half-covered (i.e. Experiment 1) or fully covered (i.e. Experiment 2) with living spores, and where the eggs have been laid (spores exposed or control side/Petri dish) were entered as explanatory factors. Each of these models first included interactions between all factors and was then simplified by removing the non-significant interactions (all P values > 0.073). When applicable, pair-wise comparisons between populations were tested using Tukey contrasts. RESULTS Question 1: Do females avoid infected sites to lay their eggs? When given a choice, females did not avoid the side covered with pathogens both before oviposition (Mean ± SE proportion of time spent by a female on the control side: 54.0 ± 18.0%; t48 = 1.52, P = 0.136) and at egg laying (22 and 15 females laid their eggs in the control and infected sides, respectively; exact binomial test: P = 0.324). These results were independent of female’s population, egg number and of an interaction between these 2 variables (Table 1). Overall, only 1 of the 15 females who laid their eggs on the pathogen side relocated its clutch to the control side, which is comparable to the 1 of the 22 females who relocated its eggs after having laid them on the control side. Note that this general tolerance of females for the spores’ side is unlikely to result from the unintentional spread of spores by moving females, since they were equally distributed between the 2 sides already on the first day (28 and 21 females on the Tween and spores sides, respectively; exact binomial test: P = 0.392) and the second day (22 and 27 females on the Tween and spores sides, respectively; exact binomial test: P = 0.568) of the experiment. Table 1 Effects of population and egg number on females’ likelihood (a) to avoid the pathogen-exposed side of the experimental arena before oviposition or (b) to lay eggs on the control side of the experimental arena.   (a) Before egg laying  (b) At egg laying    LR χ2  df  P  LR χ2  df  P  Population  0.92  2  0.633  2.36  2  0.308  Egg number  2.10  1  0.147  0.85  1  0.357  Population: Egg number  0.11  2  0.945  2.68  2  0.262    (a) Before egg laying  (b) At egg laying    LR χ2  df  P  LR χ2  df  P  Population  0.92  2  0.633  2.36  2  0.308  Egg number  2.10  1  0.147  0.85  1  0.357  Population: Egg number  0.11  2  0.945  2.68  2  0.262  View Large Question 2: Do females adapt their pre-hatching care to the presence of pathogens? In absence of a choice, the presence of pathogens in the environment altered two of the five measured forms of maternal care: the duration of egg abandonment and nest construction (Table 2, Figure 1). The duration of egg abandonment was overall shorter in females occupying Petri dishes covered with living spores compared to females in Petri dishes with UV-killed spores (Survdiff; χ12 = 6.5, P = 0.011) or without spores (Survdiff; χ12 = 3.4, P = 0.065), even if this last effect was marginally non-significant. The duration of egg abandonment was, however, comparable between Petri dishes covered with UV-killed spores and not covered by any spores (Survdiff; χ12 = 1.1, P = 0.287). On the other hand, the effect of spore presence on nest construction depended on females’ population (Interaction in Table 2b). In particular, females from Vincennes were more likely to build a nest in the Petri dishes covered with living spores (Figure 1b; Tukey contrasts; t = 3.63, P = 0.004) and UV-killed spores (t = 3.71, P = 0.003) compared to females in Petri dishes without any spores, with no difference between the two latter treatments (t < 0.001, P = 1.000). The presence of pathogens had, however, no effect on nest construction in females from Girona (Figure 1b; F2,28 = 0.82, P = 0.451) and Montblanc (F2,62 =2 .51, P = 0.090). Finally, the speed of egg gathering, the level of egg defense and the occurrence of egg grooming were all independent of treatment and population (Table 2c–e and Figure 2). Table 2 Effects of population and treatment on (a) the duration of clutch abandonment, (b) nest construction, (c) egg gathering, (d) egg defense, and (e) egg grooming   (a) Clutch abandonment  (b) Nest construction  (c) Egg gathering    χ2  df  P  χ2  df  P  χ2  df  P  Population  0.68  2  0.7129  17.25  2  0.0002  1.96  2  0.3744  Treatment  7.38  2  0.0250  0.78  2  0.6768  0.41  2  0.8143  Population: Treatment  3.73  4  0.4431  16.09  4  0.0029  3.85  4  0.4267  Statistical model  Cox model  GLM (binomial)  Cox model    (d) Egg defense  (e) Egg grooming      F  df  P  F  df  P    Population  1.20  2  0.3061  4.415  2  0.0143    Treatment  0.71  2  0.4949  0.163  2  0.8496    Population: Treatment  1.71  4  0.1533  0.884  4  0.4763    Statistical model  LM  LM      (a) Clutch abandonment  (b) Nest construction  (c) Egg gathering    χ2  df  P  χ2  df  P  χ2  df  P  Population  0.68  2  0.7129  17.25  2  0.0002  1.96  2  0.3744  Treatment  7.38  2  0.0250  0.78  2  0.6768  0.41  2  0.8143  Population: Treatment  3.73  4  0.4431  16.09  4  0.0029  3.85  4  0.4267  Statistical model  Cox model  GLM (binomial)  Cox model    (d) Egg defense  (e) Egg grooming      F  df  P  F  df  P    Population  1.20  2  0.3061  4.415  2  0.0143    Treatment  0.71  2  0.4949  0.163  2  0.8496    Population: Treatment  1.71  4  0.1533  0.884  4  0.4763    Statistical model  LM  LM    Significant P values are in bold. View Large Figure 1 View largeDownload slide The duration of clutch abandonment and the likelihood of nest building reflected the presence of either living spores, UV-killed spores or no spores in the nesting area. In particular, (a) Mothers abandoned their clutch of eggs less long in presence of living spores as compared to UV-killed spores and no spores at all. (b) Mothers were also more likely to build a nest in presence of UV-killed and living spores compared to no spore at all, but only when they were originating from the Vincennes population. Sample sizes are at the bottom of each bar. Different letters correspond to P values < 0.065. Figure 1 View largeDownload slide The duration of clutch abandonment and the likelihood of nest building reflected the presence of either living spores, UV-killed spores or no spores in the nesting area. In particular, (a) Mothers abandoned their clutch of eggs less long in presence of living spores as compared to UV-killed spores and no spores at all. (b) Mothers were also more likely to build a nest in presence of UV-killed and living spores compared to no spore at all, but only when they were originating from the Vincennes population. Sample sizes are at the bottom of each bar. Different letters correspond to P values < 0.065. Figure 2 View largeDownload slide The presence of either living spores, UV-killed spores or no spores in the nesting area did not influence three measured forms of maternal care: (a) the speed of egg gathering after an experimental egg spread, (b) the level of egg defense against a simulated predator attack and (c) the frequency of egg grooming. Figure 2 View largeDownload slide The presence of either living spores, UV-killed spores or no spores in the nesting area did not influence three measured forms of maternal care: (a) the speed of egg gathering after an experimental egg spread, (b) the level of egg defense against a simulated predator attack and (c) the frequency of egg grooming. Question 3: Do females adjust clutch quality and quantity to pathogen presence? The PCA conducted on the three egg traits provided three orthogonal principal components (PCs), of which we extracted the 2 first (total variance explained = 97.6 %). The first component (PC1) was highly and positively loaded with egg volume and egg weight (0.959 and 0.964, respectively; loading of egg number = −0.42), therefore reflecting egg quality. Conversely, the second component (PC2) was highly and positively loaded with egg number (0.91; loading of egg volume = 0.21; loading of egg weight = 0.19), therefore reflecting egg quantity. Overall, egg quality (PC1) and egg quantity (PC2) were independent of whether the eggs were laid with or without spores in the nesting area, or whether females laid their eggs in Petri dishes either half-covered or fully covered with living spores (i.e. the type of experimental arena; Table 3, Figure 3). By contrast, PC1 and PC2 were population-specific (Table 3, Figure 3). Egg quality (PC1) was larger in clutches produced by females from Montblanc compared to Girona, with an intermediate level in females from Vincennes (Figure 3). Conversely, egg quantity (PC2) was larger in clutches produced by females from both Girona and Montblanc compared to Vincennes, with no difference between the two first populations (Figure 3). Table 3 Effects of the type of experimental arena (females had either the choice or no choice to lay on a spore-covered area), population, and spore presence (if the eggs where laid in the spore-covered area or not) on egg quality and egg quantity   Egg quality (PC1)  Egg quantity (PC2)    F  df  P  F  df  P  Experiment  1.58  1  0.212  0.14  1  0.710  Population  6.70  2  0.008  9.61  2  <0.001  Spores presence  <0.01  1  0.966  0.03  1  0.856    Egg quality (PC1)  Egg quantity (PC2)    F  df  P  F  df  P  Experiment  1.58  1  0.212  0.14  1  0.710  Population  6.70  2  0.008  9.61  2  <0.001  Spores presence  <0.01  1  0.966  0.03  1  0.856  All interactions have been tested and were then removed from the statistical models because non-significant. Significant P values are in bold. View Large Figure 3 View largeDownload slide The population, but not the presence of spores in the nesting area shaped the quality (PC1) and quantity (PC2) of eggs produced. Pair-wise comparisons based on Tukey contrasts. Different letters P < 0.005. Figure 3 View largeDownload slide The population, but not the presence of spores in the nesting area shaped the quality (PC1) and quantity (PC2) of eggs produced. Pair-wise comparisons based on Tukey contrasts. Different letters P < 0.005. Hatching success was overall higher when females laid their eggs in Petri dishes half-covered (i.e. Experiment 1) compared to fully covered (i.e. Experiment 2) with living spores (F1,115 = 4.13, P = 0.044; Figure 4). Hatching success also depended on females’ population (F2,115 = 3.35, P = 0.039). The hatching success was higher in clutches produced by females from Girona compared to Vincennes (Z = −2.53, P = 0.030), but comparable between Girona and Montblanc (Z = −1.06, P = 0.242) and Vincennes and Montblanc (Z = −1.41, P = 0.331). The hatching success was, however, independent of the presence of living spores in the nesting area (F1,115 = 0.034, P = 0.854). Figure 4 View largeDownload slide Effect of population, possibility for females to avoid locations covered with spores (choice or no choice) and of the presence of living spores in their vicinity on hatching success. Pair-wise comparisons based on Tukey contrasts. Different letters P < 0.030. Figure 4 View largeDownload slide Effect of population, possibility for females to avoid locations covered with spores (choice or no choice) and of the presence of living spores in their vicinity on hatching success. Pair-wise comparisons based on Tukey contrasts. Different letters P < 0.030. DISCUSSION Shedding light on which maternal strategies can protect offspring against infection is of central importance to better understand the evolutionary drivers of parental care and more generally, the emergence and maintenance of family life in nature (Royle et al. 2012; Klug and Bonsall 2014). In this study, we demonstrated that the presence of pathogens in the environment influenced maternal investment into egg care, but not into egg production in the European earwig F. auricularia. Specifically, we found that the duration of egg abandonment (after an experimental disturbance) was shorter when the nesting area was covered with living spores compared to covered with UV-killed spores or no spores at all. Females were also more likely to build a nest in presence compared to absence of both living and UV-killed spores, even if this effect was only present in females from one of the three studied populations. By contrast, the presence of living spores had no effect on egg grooming, egg gathering and egg defense. We also showed that mothers did not avoid contaminated areas at egg laying, and that the presence of living spores did not influence maternal investment into egg number, volume and weight, and did not shape hatching success. Finally, our data confirm that egg properties and hatching success are population-specific in this species (see also Ratz et al. 2016). Altogether, our results show that mothers can detect the presence of spores in the environment and consequently increase their investment into certain forms of egg care. In particular, the presence of pathogens shortened the duration of egg abandonment, favored the construction of a nest (in females from Vincennes), but had no effect on maternal investment into egg gathering, egg defense and egg grooming. The effect of pathogens on the first 2 behaviors is not surprising, as both can help mothers to protect their eggs against pathogen exposure. Returning to the eggs quickly after disturbance is indeed essential to ensure the frequent physical contact between eggs and mothers that are typically required to remove pathogens from egg surface (Boos et al. 2014). Similarly, nest construction is a form of hygienic behavior that may help cleaning the nest from pathogen spores by shifting the sands around (Arathi et al. 1999). Conversely, egg gathering and egg defense are 2 maternal behaviors triggered by predator attacks (Thesing et al. 2015) and therefore unlikely to depend on pathogens presence. What was more surprising, however, is that egg grooming was independent of pathogen presence. This behavior is typically known to help individuals cleaning external parasites and pathogens in insects (Reber et al. 2011; Meunier 2015) and has been previously show to help earwig mothers cleaning their eggs from non-pathogenic fungal spores (Boos et al. 2014). The apparent discrepancy between our results and the ones from Boos et al. (2014) likely relies on the different quantity of spores covering the eggs on the days of measurement. Boos et al. (2014) directly covered the eggs with a high quantity of fungal spores and then immediately measured egg care—resulting in a high concentration of spores on the tested eggs. In our experiment, the eggs were laid in an area contaminated with fungal spores (i.e. indirect spore exposure) and were then groomed by their mothers for 15 days before we measured egg care—therefore resulting in a comparatively lower concentration of spores (if any) on the tested eggs. Overall, these studies thus suggest that egg grooming depends on the quantity of spores present on the eggs and hence, that its anti-pathogenic function is curative rather than prophylactic. Somewhat surprisingly, females exhibited different reactions in presence of UV-treated and living spores: the UV treatment inhibited the effect of spore presence on the duration of egg abandonment, but not on the likelihood of nest construction. It is generally known that individuals can change their expression of social behaviors when they encounter conspecifics that are infected or non-infected by Metarhizium spores, as reported in several eusocial insect species (Walker and Hughes 2009; Reber et al. 2011; Leclerc and Detrain 2016). However, the fact that females can change their level of care to the presence of infectious or non-infectious spores in the environment was—to the best of our knowledge—unknown. The capability to discriminate between infectious and non-infectious spores can be crucial for mothers. It may allow females to prevent their investment into costly protections against non-active threats, and instead favor investment in other fitness related processes, such as the provisioning of food to current offspring or the accumulation of energy for future reproduction (Royle et al. 2012). On a proximal level, the mediators of this apparent discrimination are unclear. M. brunneum spores typically infect hosts by adhering to their cuticle and then germinating to penetrate into their body (Vestergaard et al. 1999; Thomas and Read 2007). We hypothesize that the damages of UV on the spores may not only affect their DNA (as generally known, see Braga et al. 2015), but also their shape, chemical signature and/or adherence capability, which all could either help females discriminating against dead spores or simply make dead spores undetectable to females. Further studies should investigate which of these two scenarios explain the present finding and more generally, whether this apparent discrimination capability applies to other pathogens and hosts. Contrary to the previous spore-dependent effects, we demonstrated that earwig females showed no preference for either pathogen-exposed or pathogen-free environments for oviposition, nor did they adjust their egg properties to the presence of pathogens in the nesting area. The reported absence of pathogen-dependent effects in earwigs is uncommon among invertebrates, where most studies show that females either avoid (Meyling and Pell 2006; Lam et al. 2010) or prefer (Brütsch et al. 2014; Pontieri et al. 2014) areas covered with entomopathogenic spores for oviposition, as well as either increase (Villani et al. 1994) or reduce (Machtinger et al. 2016) their egg production in presence of entomopathogenic spores. In earwigs, this effect is unlikely to reflect females’ inability to detect spore presence (see above). Our results therefore suggest that females detected pathogens but neither chose their egg laying location nor adjusted their investment into egg production accordingly. Two non-mutually exclusive hypotheses could explain this pattern. First, the presence of spores in the environment might not be a major threat for the eggs. This is an unlikely hypothesis, as this generalist entomopathogenic fungus is a well-known threat for eggs in insects (Samuels et al. 2002; Nozad-Bonab et al. 2017) and our data demonstrate that earwig females react to spore presence by increasing the expression of the egg care specifically directed against pathogens. The second hypothesis is that maternal egg care is efficient enough to prevent the infection of eggs by fungal spores, and thus to relax selection for avoidance behavior and pathogen-dependent change in egg properties. In line with this hypothesis, a previous study demonstrated that maternal presence mitigates the costs of a direct exposure to mould on hatching success in several earwig species (Klostermeyer 1942; Miller and Zink 2012; Boos et al. 2014) and our results demonstrate that the presence of living spores had no effect on hatching success (in presence of a tending mother). Disentangling the effects of temperature and maternal care efficiency on egg production will be the goal of future studies. To conclude, our results reveal that mothers can express prophylactic mechanisms to defend their eggs against surrounding pathogens and that these defenses occur after egg production. Interestingly, these prophylactic defenses come together with other types of anti-pathogenic defenses exhibited by earwig mothers, such as the removal of fungal spores from egg surfaces (Boos et al. 2014), the transfer of chemical compounds with antimicrobial properties on the eggs (Boos et al. 2014) and the maintenance of feces with antimicrobial functions into the nest (Diehl et al. 2015; Körner et al. 2016). These results thus overall reveal that both prophylactic and curative/therapeutic defenses against microbes can co-occur in a species where nesting material is not subject to a direct competition with microbes—as compared, for instance, to carcasses in burying beetles (Rozen et al. 2008) or feces in dung beetles (Byrne et al. 2013). Active protection against microbial pathogens could thus play a central role in the emergence and maintenance of parental care in nature, as well as explain the taxonomically widespread occurrence of egg attendance in insects and arthropods (Wong et al. 2013). FUNDING This work was supported by the German Science Foundation (DFG; ME4179/3-1 to J.M.). We thank Maximilian Körner and Fanny Vogelweith for their comments on this manuscript. Data accessibility: Analyses reported in this article can be reproduced using the data provided by Meunier and Diehl (2017). REFERENCES Arathi HS, Burns I, Spivak M. 1999. Ethology of hygienic behavior in the honey bee Apis mellifera L. (Hymenoptera: Apidae): behavioral repertoire of hygienic bees. Ethology . 106: 365– 379. Google Scholar CrossRef Search ADS   Beckage NE. 2008. Insect immunology . Oxford: Elsevier Inc. Boos S, Meunier J, Pichon S, Kölliker M. 2014. Maternal care provides antifungal protection to eggs in the European earwig. Behav Ecol . 25: 754– 761. Google Scholar CrossRef Search ADS   Bos N, Sundström L, Fuchs S, Freitak D. 2015. Ants medicate to fight disease. Evolution . 69: 2979– 2984. Google Scholar CrossRef Search ADS PubMed  Bot ANM, Currie CR, Hart AG, Boomsma JJ. 2001. Waste management in leaf-cutting ants. Ethol Ecol Evol . 13: 225– 237. Google Scholar CrossRef Search ADS   Braga GUL, Rangel DEN, Fernandes EKK, Flint SD, Roberts DW. 2015. Molecular and physiological effects of environmental UV radiation on fungal conidia. Curr Genet . 61: 405– 425. Google Scholar CrossRef Search ADS PubMed  Brütsch T, Felden A, Reber A, Chapuisat M. 2014. Ant queens (Hymenoptera: Formicidae) are attracted to fungal pathogens during the initial stage of colony founding. Myrmecological news . 20: 71– 76. Byrne MJ, Watkins B, Bouwer G. 2013. Do dung beetle larvae need microbial symbionts from their parents to feed on dung? Ecol Entomol . 38: 250– 257. Google Scholar CrossRef Search ADS   Cotter SC, Kilner RM. 2010a. Personal immunity versus social immunity. Behav Ecol . 21: 663– 668. Google Scholar CrossRef Search ADS   Cotter SC, Kilner RM. 2010b. Sexual division of antibacterial resource defense in breeding burying beetles, Nicrophorus vespilloides. J Anim Ecol . 79: 35– 43. Google Scholar CrossRef Search ADS   Cremer S, Armitage SAO, Schmid-Hempel P. 2007. Social immunity. Curr Biol . 17: R693– 702. Google Scholar CrossRef Search ADS PubMed  DeVeale B, Brummel T, Seroude L. 2004. Immunity and aging: The enemy within? Aging Cell . 3: 195– 208. Google Scholar CrossRef Search ADS PubMed  Diehl JM, Körner M, Pietsch M, Meunier J. 2015. Feces production as a form of social immunity in an insect with facultative maternal care. BMC Evol Biol . 15: 40. Google Scholar CrossRef Search ADS PubMed  Diez L, Deneubourg J-L, Detrain C. 2012. Social prophylaxis through distant corpse removal in ants. Naturwissenschaften . 99: 833– 42. Google Scholar CrossRef Search ADS PubMed  Giacomello E, Marchini D, Rasotto MB. 2006. A male sexually dimorphic trait provides antimicrobials to eggs in blenny fish. Biol Lett . 2: 330– 333. Google Scholar CrossRef Search ADS PubMed  Giacomello E, Marri L, Marchini D, Mazzoldi C, Rasotto MB. 2008. Sperm-duct gland secretion of the grass goby Zosterisessor ophiocephalus exhibits antimicrobial activity. J Fish Biol . 73: 1823– 1828. Google Scholar CrossRef Search ADS   Gindin G, Ment D, Rot A, Glazer I, Samish M. 2009. Pathogenicity of Metarhizium anisopliae (Hypocreales: Clavicipitaceae) to tick eggs and the effect of egg cuticular lipids on conidia development. J Med Entomol . 46: 531– 538. Google Scholar CrossRef Search ADS PubMed  Günther K, Herter K. 1974. Dermaptera (Ohrwürmer). In: Helmcke JG, Starck D, Wermuth H, editors. Handbuch der Zoologie . Berlin: Walter de. Jackson DE, Hart AG. 2009. Does sanitation facilitate sociality? Anim Behav . 77: e1– e5. Google Scholar CrossRef Search ADS   Kaltenpoth M, Göttler W, Herzner G, Strohm E. 2005. Symbiotic bacteria protect wasp larvae from fungal infestation. Curr Biol . 15: 475– 479. Google Scholar CrossRef Search ADS PubMed  Klostermeyer EC. 1942. The life history and habits of the ringlegged earwig, Euborellia annulipes (Order Dermaptera). J Kansas Entomol Soc . 15: 13– 18. Klug H, Bonsall MB. 2014. What are the benefits of parental care? The importance of parental effects on developmental rate. Ecol Evol . 4: 2330– 2351. Google Scholar CrossRef Search ADS PubMed  Knouft JH, Page LM, Plewa MJ. 2003. Antimicrobial egg cleaning by the fringed darter (Perciformes: Percidae: Etheostoma crossopterum): implications of a novel component of parental care in fishes. Proc R Soc B Biol Sci . 270: 2405– 2411. Google Scholar CrossRef Search ADS   Koch LK, Meunier J. 2014. Mother and offspring fitness in an insect with maternal care: phenotypic trade-offs between egg number, egg mass and egg care. BMC Evol Biol . 14: 125. Google Scholar CrossRef Search ADS PubMed  Kohlmeier P, Holländer K, Meunier J. 2016. Survival after pathogen exposure in group-living insects: don’t forget the stress of social isolation!. J Evol Biol . 29: 1867– 1872. Google Scholar CrossRef Search ADS PubMed  Kölliker M. 2007. Benefits and costs of earwig (Forficula auricularia) family life. Behav Ecol Soc . 61: 1489– 1497. Google Scholar CrossRef Search ADS   Körner M, Diehl JM, Meunier J. 2016. Growing up with feces: benefits of allo-coprophagy in families of the European earwig. Behav Ecol . 27: 1775– 1781. Kramer J, Thesing J, Meunier J. 2015. Negative association between parental care and sibling cooperation in earwigs: a new perspective on the early evolution of family life? J Evol Biol . 28: 1299– 1308. Google Scholar CrossRef Search ADS PubMed  Kroiss J, Kaltenpoth M, Schneider B, Schwinger M-G, Hertweck C, Maddula RK, Strohm E, Svatoš A. 2010. Symbiotic streptomycetes provide antibiotic combination prophylaxis for wasp offspring. Nat Chem Biol . 6: 261– 263. Google Scholar CrossRef Search ADS PubMed  Lam K, Tsang M, Labrie A, Gries R, Gries G. 2010. Semiochemical-mediated oviposition avoidance by female house flies, Musca domestica, on animal feces colonized with harmful fungi. J Chem Ecol . 36: 141– 147. Google Scholar CrossRef Search ADS PubMed  Lamb RJ. 1976. Parental behavior in the dermaptera with special reference to Forficula auricularia (Dermaptera: Forficulidae). Can J Entomol . 108: 609– 619. Google Scholar CrossRef Search ADS   Leclerc J-B, Detrain C. 2016. Ants detect but do not discriminate diseased workers within their nest. Sci Nat . 103: 70. Google Scholar CrossRef Search ADS   Machtinger ET, Weeks ENI, Geden CJ. 2016. Oviposition deterrence and immature survival of filth flies (Diptera: Muscidae) when exposed to commercial fungal products. J Insect Sci . 16 (1):54, 1–6. Meunier J. 2015. Social immunity and the evolution of group living in insects. Philos Trans R Soc B Biol Sci . 370: 20140102. Google Scholar CrossRef Search ADS   Meunier J, Chapuisat M. 2009. The determinants of queen size in a socially polymorphic ant. J Evol Biol . 22: 1906– 1913. Google Scholar CrossRef Search ADS PubMed  Meunier J, Kölliker M. 2012. Parental antagonism and parent-offspring co-adaptation interact to shape family life. Proc R Soc B Biol Sci . 279: 3981– 3988. Google Scholar CrossRef Search ADS   Meunier J, Wong JWY, Gómez Y, Kuttler S, Röllin L, Stucki D, Kölliker M. 2012. One clutch or two clutches? Fitness correlates of coexisting alternative female life-histories in the European earwig. Evol Ecol . 26: 669– 682. Google Scholar CrossRef Search ADS   Meunier J, Diehl JMC. 2017. Data from: surrounding pathogens shape maternal egg care but not egg production in the European earwig. Dryad Digital Repository . http://dx.doi.org/10.5061/dryad.pv50q. Meyling NV, Pell JK. 2006. Detection and avoidance of an entomopathogenic fungus by a generalist insect predator. Ecol Entomol . 31: 162– 171. Google Scholar CrossRef Search ADS   Miller JS, Zink AG. 2012. Parental care trade-offs and the role of filial cannibalism in the maritime earwig, Anisolabis maritima. Anim Behav . 83: 1387– 1394. Google Scholar CrossRef Search ADS   Nozad-Bonab Z, Hejazi MJ, Iranipour S, Arzanlou M. 2017. Lethal and sublethal effects of some chemical and biological insecticides on Tuta absoluta (Lepidoptera: Gelechiidae) eggs and neonates. J Econ Entomol . 110: 1138– 1144. Google Scholar CrossRef Search ADS PubMed  Ormond EL, Thomas APM, Pell JK, Freeman SN, Roy HE. 2011. Avoidance of a generalist entomopathogenic fungus by the ladybird, Coccinella septempunctata. FEMS Microb Ecol . 77: 229– 37. Google Scholar CrossRef Search ADS   Parker BJ, Barribeau SM, Laughton AM, de Roode JC, Gerardo NM. 2011. Non-immunological defense in an evolutionary framework. Trends Ecol Evol . 26: 242– 248. Google Scholar CrossRef Search ADS PubMed  Pie MR, Rosengaus RB, Traniello JFA. 2004. Nest architecture, activity pattern, worker density and the dynamics of disease transmission in social insects. J Theor Biol . 226: 45– 51. Google Scholar CrossRef Search ADS PubMed  Pontieri L, Vojvodic S, Graham R, Pedersen JS, Linksvayer TA. 2014. Ant colonies prefer infected over uninfected nest sites. PLoS One . 9: e111961. Google Scholar CrossRef Search ADS PubMed  Ratz T, Kramer J, Veuille M, Meunier J. 2016. The population determines whether and how life-history traits vary between reproductive events in an insect with maternal care. Oecologia . 182: 443– 452. Google Scholar CrossRef Search ADS PubMed  Reber A, Chapuisat M. 2011. Diversity, prevalence and virulence of fungal entomopathogens in colonies of the ant Formica selysi. Insectes Soc . 59: 231– 239. Google Scholar CrossRef Search ADS   Reber A, Purcell J, Buechel SD, Buri P, Chapuisat M. 2011. The expression and impact of antifungal grooming in ants. J Evol Biol . 24: 954– 964. Google Scholar CrossRef Search ADS PubMed  de Roode JC, Lefèvre T. 2012. Behavioral immunity in insects. Insects . 3: 789– 820. Google Scholar CrossRef Search ADS PubMed  de Roode JC, Lefèvre T, Hunter MD. 2013. Self-medication in animals. Science . 340: 150– 1. Google Scholar CrossRef Search ADS PubMed  Royle NJ, Smiseth PT, Kölliker M. 2012. The evolution of parental care . Oxford: Oxford University Press. Google Scholar CrossRef Search ADS   Rozen DE, Engelmoer DJP, Smiseth PT. 2008. Antimicrobial strategies in burying beetles breeding on carrion. PNAS . 105: 17890– 5. Google Scholar CrossRef Search ADS PubMed  Samuels R, Coracini DL, Martins dos Santos C, Gava Ca. 2002. Infection of Blissus antillus (Hemiptera: Lygaeidae) eggs by the entomopathogenic fungi Metarhizium anisopliae and Beauveria bassiana. Biol Control . 23: 269– 273. Google Scholar CrossRef Search ADS   Schmid-Hempel P. 1998. Parasites in social insects . Princeton (NJ): Princeton University Press. Schmid-Hempel P. 2014. Evolutionary parasitology . Oxford: Oxford University Press. Siva-jothy MT, Moret Y, Rolff J. 2005. Insect immunity: An evolutionary ecology perspective. Adv In Insect Phys . 32: 1– 48. Google Scholar CrossRef Search ADS   Smiseth PT, Kölliker M, Royle NJ. 2012. What is parental care? In: Smiseth PT, Kölliker M, Royle NJ. The evolution of parental care . Oxford: Oxford University Press. p. 1– 17. Google Scholar CrossRef Search ADS   Stroeymeyt N, Casillas-Pérez B, Cremer S. 2014. Organisational immunity in social insects. Curr Opin Insect Sci . 5: 1– 15. Google Scholar CrossRef Search ADS   Thesing J, Kramer J, Koch LK, Meunier J. 2015. Short-term benefits, but transgenerational costs of maternal loss in an insect with facultative maternal care. Proc R Soc B Biol Sci . 282: 20151617. Google Scholar CrossRef Search ADS   Thomas MB, Read AF. 2007. Can fungal biopesticides control malaria? Nat Rev Microbiol . 5: 377– 383. Google Scholar CrossRef Search ADS PubMed  Vestergaard S, Butt TM, Bresciani J, Gillespie AT, Eilenberg J. 1999. Light and Electron Microscopy Studies of the Infection of the Western Flower Thrips Frankliniella occidentalis (Thysanoptera: Thripidae) by the Entomopathogenic Fungus Metarhizium anisopliae. J Invertebr Pathol . 73: 25– 33. Google Scholar CrossRef Search ADS PubMed  Villani MG, Krueger SR, Schroeder PC, Consolie F, Consolie NH, Preston-Wilsey LM, Donald RW. 1994. Soil application effects of Metarhizium anisopliae on Japanese Beetle (Coleoptera: Scarabaeidae) behavior and survival in turfgrass microcosms. Environ Entomol . 23: 502– 513. Google Scholar CrossRef Search ADS   Vogelweith F, Körner M, Foitzik S, Meunier J. 2017. Age, pathogen exposure, but not maternal care shape offspring immunity in an insect with facultative family life. BMC Evol Biol . 17: 69. Google Scholar CrossRef Search ADS PubMed  Walker TN, Hughes WOH. 2009. Adaptive social immunity in leaf-cutting ants. Biol Lett . 5: 446– 448. Google Scholar CrossRef Search ADS PubMed  Wong JWY, Meunier J, Kölliker M. 2013. The evolution of parental care in insects: the roles of ecology, life history and the social environment. Ecol Entomol . 38: 123– 137. Google Scholar CrossRef Search ADS   © The Author(s) 2017. Published by Oxford University Press on behalf of the International Society for Behavioral Ecology. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Behavioral Ecology Oxford University Press

Surrounding pathogens shape maternal egg care but not egg production in the European earwig

Loading next page...
 
/lp/ou_press/surrounding-pathogens-shape-maternal-egg-care-but-not-egg-production-VWQPGwhZn5
Publisher
Oxford University Press
Copyright
© The Author(s) 2017. Published by Oxford University Press on behalf of the International Society for Behavioral Ecology. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com
ISSN
1045-2249
eISSN
1465-7279
D.O.I.
10.1093/beheco/arx140
Publisher site
See Article on Publisher Site

Abstract

Abstract Pathogens are ubiquitous in nature and typically entail major fitness costs in their hosts. These costs can be particularly important when individuals exhibit poor immune defenses, as it is often the case during early developmental stages. Hence, selection should favor parental strategies limiting the risks of pathogen exposure and infection in their offspring. In this study, we investigated 1) whether females of the European earwig Forficula auricularia avoid areas contaminated with spores of the entomopathogenic fungus Metarhizium brunneum prior to and at egg laying, as well as 2) whether spore presence entails an increase in females’ investment into both pre-hatching forms of care and clutch quantity and quality. Our results first show that females did not avoid contaminated areas prior to and at egg laying. However, females returned to their eggs faster in presence of living spores compared to UV-killed or no spores. They were also more likely to construct a nest when in presence of both living and UV-killed spores (but only in one studied population). Finally, we found that spore presence did not influence maternal investment into egg grooming, egg gathering and egg defense, as well as into clutch quantity and quality. Overall, our results demonstrate that earwig females do not avoid contaminated environments, but could mitigate the associated costs of pathogen exposure by adjusting their level of egg care. These findings emphasize the importance of pathogens in the evolution of pre-hatching parental care and, more generally, in the emergence and maintenance of family life in nature. INTRODUCTION Pathogens are considered as a major threat in nature, because they often entail large fitness costs in their hosts (Schmid-Hempel 2014). To limit the risks of pathogen exposure and infection, individuals from both vertebrate and invertebrate species have developed a broad spectrum of defenses (Siva-jothy et al. 2005; Schmid-Hempel 2014). The most studied defenses rely on the immune system of the potential host, which typically detects the presence of pathogens in the organism, prevents their development and ultimately kill and purge them from the host (Siva-jothy et al. 2005; Beckage 2008). However, many non-immunological defenses may also help hosts to fight against pathogens (Parker et al. 2011). Among them, hosts have been shown to change some of their behaviors to prevent direct contacts to pathogens (de Roode and Lefèvre 2012). This can be done, for instance, by avoiding infected areas and limiting the consumption of pathogen-infected food (Villani et al. 1994; Meyling and Pell 2006; Ormond et al. 2011), and by prophylactically or curatively consuming natural components with antimicrobial properties, a process called self-medication (de Roode et al. 2013; Bos et al. 2015). Interestingly, defenses against pathogens are not only expressed by the hosts themselves (thereafter called personal immune defenses) but can also come from surrounding conspecifics and group members (Schmid-Hempel 1998). Over the last decades, these external defenses have been mostly studied in colonies of eusocial insects, such as bees, ants and termites (Cremer et al. 2007; Meunier 2015). In these species, this so-called social immunity is often illustrated by allogrooming or hygienic behaviors, during which workers remove parasites from the cuticle of another group member (Reber et al. 2011) or clean the nest from waste materials, such as corpses, left-over food and feces (Bot et al. 2001; Jackson and Hart 2009; Diez et al. 2012). The expression of social immunity—which can complement personal immune defenses – is considered as a central process in the evolution of eusocial species (Meunier 2015), since the obligatory and frequent contacts between their related group members can dramatically increase the risks of pathogen exposure and transmission (Pie et al. 2004; Stroeymeyt et al. 2014). Social immunity, however, is not exclusive to eusocial systems and can be found in less derived forms of group living, such as simple family units (Cotter and Kilner 2010a; Meunier 2015). In these units, parents generally provide multiple forms of care to their eggs and/or juveniles (Smiseth et al. 2012) and, because eggs and juveniles often exhibit an underdeveloped immune competence (DeVeale et al. 2004; Vogelweith et al. 2017), parental care often includes protection against pathogens. For instance, mothers can protect their eggs against microbial infections by applying antimicrobial secretions to their surface, as reported in blennies, grass gobies and fringed darters (Knouft et al. 2003; Giacomello et al. 2006; Giacomello et al. 2008). Similarly, parents may cover the nest with feces exhibiting antibacterial activity to prevent the development of pathogens next to the juveniles, a phenomenon found in the burying beetle and the European earwig (Rozen et al. 2008; Cotter and Kilner 2010b; Diehl et al. 2015). Finally, females of the beewolf digger wasp Philanthus triangulum have been shown to transfer symbiotic bacteria to the brood cell cover (Kaltenpoth et al. 2005) and these bacteria to be taken up by the larvae and incorporated into their cocoon to serve as antimicrobial protection (Kroiss et al. 2010). Whereas these maternal behaviors typically protect offspring against pathogens, it remains unclear whether these behaviors are curative/therapeutic (i.e. help fighting against pathogens already in contact with offspring) or prophylactic (i.e. help preventing potential contacts between pathogens and offspring). Disentangling between these two processes is however crucial to better understand on which level pathogen-induced selection operates in the evolution of parental care. In this study, we investigated whether maternal egg care is a prophylactic process in the European earwig Forficula auricularia. Specifically, we tested whether the presence of pathogens in the environment triggers changes in females’ investment into pre-hatching egg care (i.e. social immunity) and/or egg production. In this insect species, mothers guard their eggs for several months over winter (Lamb 1976). During egg development (but also during the subsequent development of juveniles), females express a wide range of protective behaviors, such as grooming, clutch displacement and rearrangement, as well as aggressions against predators (Lamb 1976; Meunier and Kölliker 2012; Boos et al. 2014; Koch and Meunier 2014; Thesing et al. 2015). Whether mothers can detect the presence of living pathogens in the environment (and not only on the eggs, see Boos et al. 2014) and prophylactically adapt their behaviors remains, however, unknown. Moreover, a recent study revealed that some life history traits could be population-specific in earwigs (Ratz et al. 2016), begging the question of whether prophylactic defenses against pathogens might differ between populations. We conducted two experiments to test the expression of three prophylactic defenses in earwig females originating from three independent populations. First, we tested whether females avoid areas infected with an entomopathogenic fungus to lay their eggs. Second, we investigated whether females modify their investment into egg quantity and quality (i.e. egg size and weight) when the fungus is present in their direct environment. Finally, we tested whether females change their level of pre-hatching care when the nesting area (but not necessarily the eggs) contains this fungus. We used spores of the generalist entomopathogenic fungus Metarhizium brunneum (formerly M. anisopliae), a natural and potentially lethal pathogen of F. auricularia (Günther and Herter 1974; Kohlmeier et al. 2016) also known to infect and reduce the viability of eggs in many arthropods (Samuels et al. 2002; Gindin et al. 2009; Nozad-Bonab et al. 2017). In the first experiment, we recorded the movement and egg laying behaviors of females preliminary set up in areas covered with a spore solution on one side and with a spore-free solution on the other side. If mothers prophylactically limit the exposure of their eggs to pathogens, we expected females to avoid contact with fungal spores both before and at egg laying. Similarly, females laying their eggs on the spore-exposed side should be more likely to relocate them afterwards. In the second experiment, we measured the level of five forms of egg care behaviors (nest construction, egg gathering, egg defense, egg abandonment and egg grooming) in females previously set up in areas fully covered with either living spores, UV-killed spores or no spores. We predicted that females increase their investment into egg care only in presence of living spores. Finally, we measured the number, mean volume, mean weight and hatching success of the eggs produced in the two experiments. If females can assess the risks of pathogen infection for their eggs (and subsequent juveniles), we expected that females reduce the risks and/or costs of this infection by prophylactically increasing their investment into egg quantity and/or quality when spores were present in the environment. If females manage to mitigate the costs of pathogen exposure, we also predicted egg-hatching success to be independent of pathogen presence. MATERIAL AND METHODS Animals and pathogen All the females used in the following experiments were the first laboratory born generation of individuals collected in 2014 in 3 independent populations: Vincennes (France; 48° 51′ N, 2° 26′ O), Girona (Spain; 41° 59′ N, 2° 49′ O), and Montblanc (Spain; 41° 23′ N, 1° 10′ O). At adult emergence, these females were maintained with unrelated males in groups of 50 individuals (balanced sex-ratio) to allow them to mate freely (Meunier et al. 2012). Each group was set up in a plastic container (37 × 22 × 25 cm) grounded with moist sand, containing egg carton as shelters, and maintained under standard conditions (constant 20 °C, 14:10 h light:dark cycle, 60% humidity). Three months later, each female was isolated in a Petri dish (9 cm diameter) grounded with moist sand, and subsequently maintained under winter conditions to allow egg production (gradual temperature change from 15 °C to 5 °C during 3 weeks, with constant 60% humidity and complete darkness). Ten days later, the temperature was increased to 10 °C for 2 more weeks and finally to 15 °C for the rest of the experiments to allow egg development. After egg hatching, each family was kept at room temperature (23 °C) until the end of the following experiments. Groups and isolated females were fed twice a week with an ad libitum amount of a standard diet mainly composed of pollen, cat food and carrots (see detailed composition in Kramer et al., 2015). The conidia (spores) of the entomopathogenic fungus M. brunneum used in the following experiments were obtained from a genotyped strain isolated from soil samples in Switzerland (Reber and Chapuisat 2011). To ensure the virulence of this strain to earwigs, this fungus has been grown on F. auricularia for 3 generations prior to their use in this experiment. Specifically, earwigs were infected in a 1-ml spore solution of M. brunneum (108/ml) conidia solved in 0.05% Tween 20 and held in a climate cabinet at 25 °C in small Petri dishes (5.5 cm diameter) filled with moist sand. After the death of the exposed individuals, corpses were rinsed 3 times with 1 ml of concentrated bleach and distilled water to prevent contamination by other microbes. Afterwards, the surface sterilized insects were kept in new Petri dishes sealed with parafilm in the climate cabinet (25 °C) until growth of the fungus appeared. The resulting spores were rinsed from the body with distilled water and applied on potato extract glucose growth medium (Roth). Growing fruit bodies were finally collected with a sterile scalpel, diluted in 0.05% Tween 20 solution and then used to infect new individuals. This process has been repeated 3 times to obtain the spores used in the following experiments. Question 1: Do females avoid infected sites to lay their eggs? To investigate whether females avoid laying their eggs in a pathogenic environment, we transferred a random sample of 37 females (n = 11 Girona, n = 18 Montblanc and n = 8 Vincennes) to individual Petri dishes (9 cm diameter, grounded with humid sand) divided in two parts of equal size. One side was previously sprinkled with 0.55 ml of a spore solution of M. brunneum (107/ml) solved in Tween 20 (0.05%, Roth), whereas the other side (control) was previously sprinkled with the same amount of a spore-free Tween 20 solution. A direct exposure to a solution of 107 spores per ml is known to entail a 20% adult mortality after 25 days (Kohlmeier et al. 2016). The 2 sides were separated by a cardboard barrier (8.7 × 1 cm) covered with aluminum foil to avoid any passive diffusion of the applied solutions between the two sides, while allowing females to easily move from one side to the other. Each experimental Petri dish was then checked on a daily basis in order to 1) count the number of days each female stayed on each side until egg laying, 2) record the side on which females deposited theirs eggs and finally to 3) count the number of times the clutches were moved to the other side during egg development. All these measurements were done blind regarding the treatments (i.e. the experimenter did not know which side of the experimental Petri dish contained the pathogen or the control). Each female was provided with an ad libitum amount of standard food until egg laying (detailed food composition in Kramer et al. 2015), as females stop feeding once they produce eggs (Kölliker 2007). Note that the pathogen and control solutions were applied twice (once just before the start of the experiment and once 1 month later) to make sure that each treatment lasts for the entire experiment. We statistically analyzed whether females avoided pathogenic environments for egg laying using 2 successive steps. In the first step, we conducted 2 Generalized Linear Models (GLM) with binomial error distribution to determine whether females’ choice for each side (i.e. preference for or avoidance of the contaminated side before and at egg laying) depended on clutch size and/or on their population of origin. In the first GLM, the proportion of days each female spent on the contaminated side before egg laying was entered as response variable (using the cbind function in R), while egg number, population (Girona, Montblanc, Vincennes) and the interaction between these 2 variables were used as explanatory variables. In the second GLM, the side chosen for egg laying (1 = contaminated side, 0 = control side) was entered as response variable, while the 2 factors described above were also used as explanatory variables. Because neither egg number nor population was significant in these 2 models (see Results), we then pooled the data across populations to conduct the second step. In this second step, we used a 1-sample student’s t-test to analyze whether females spent overall more days on the control compared to contaminated side (i.e. ratio of time on the control side compared to the value 0.5), as well as a binomial test to analyze whether the number of females that had laid eggs on the contaminated side was overall lower than the number of females that had laid eggs on the control side. All statistical analyses (here and below) were done with R v3.4.0 (http://R-project.org) loaded with the packages car, lmerTest, survival and exactRankTests. Question 2: Do females adjust their pre-hatching care to the presence of pathogens? To investigate whether females adapt their investment into pre-hatching egg care to the presence of pathogens in the environment, we used a random sample of 123 females (n = 32 Girona, n = 65 Montblanc, n = 26 Vincennes; all different from the ones used to address question 1) distributed among three treatments. The experiment started by isolating these females in Petri dishes grounded with moist sand and entirely sprinkled with 1.1 ml of one of the 3 following solutions: 1) M. brunneum spore solution (107/ml 0.05% Tween 20; n = 11 Girona, n = 22 Montblanc, and n = 9 Vincennes), 2) UV-killed M. brunneum spore solution (107/ml 0.05% Tween 20; n = 11 Girona, n = 21 Montblanc and n = 10 Vincennes) or 3) spore-free 0.05% Tween 20 solution (n = 10 Girona, n = 22 Montblanc and n = 7 Vincennes). We used both the UV- killed spores and Tween solution as controls to disentangle the effects of pathogen activity (UV-killed versus living spores) from the effects of pathogen presence (spore-free versus living spores). The UV-killed spore solution was exposed to UV light (254 nm) for three minutes in the UV irradiation system Bio-Link 254 (Vilber). The UV treatment was efficient enough to inhibit the growth of M. brunneum spores on growth medium (results not shown). To mimic the process conducted in Experiment 1, the initial application of each solution in the Petri dish was done right before female isolation and a second application one month later. Note that the presence of spores in the environment did not affect females survival rate, as 6 (12.0%), 7 (13.5%), and 6 (12.2%) females died before egg laying in the treatments with living spores, UV-killed spores and no spores, respectively. No female died after egg laying. We measured 5 forms of pre-hatching maternal care: egg gathering, egg defense, clutch abandonment, nest construction, and egg grooming. 1) Egg gathering indicates the propensity of a female to gather its eggs after these were experimentally spread out in the nest. This measurement was done 5 days after egg laying. At that time, we confined each female within its Petri dish, then spread a random selection of 30 eggs (or all, if fewer were available) evenly within a circle (~3 cm diameter, including the site where the eggs were originally laid) and subsequently recorded the seconds between the mother’s first movement out of the confinement area and the re-assembly of the eggs within one body length of the female. Note that the eggs not involved in this measurement were kept in a small closed plastic dish in the meantime. 2) Egg defense revealed females’ reaction to a simulated predator attack and was measured ten days after egg laying. This measurement was done following a standard method (Thesing et al. 2015), in which females were poked with a glass capillary on the thorax (1 poke / 2 s) until they abandoned their clutch (i.e. moved more than 2 body lengths away from it). The values of egg defense were the number of pokes the females sustained before leaving the eggs. 3) Clutch abandonment was measured subsequently to egg defense and was used to quantify how long a mother abandons her clutch of eggs after a simulated attack. The value of egg abandonment was the time (in seconds) between the moment where each female left its nest due to a simulated predator attack (see above) and the moment where females returned to their clutch and touched at least one egg. 4) Nest construction reflected whether a female invested into the construction of a nest at day 15 after egg laying. Following a method already developed in this species (Meunier et al. 2012), we considered females to have constructed a nest when they dug a pit to the ground of the Petri dish and put their eggs into the pit. Finally, 5) egg grooming was recorded 15 days after egg laying, just after checking for females nest construction. At that time, each female was observed for 46 min in its original Petri dish using a scan sampling method (one scan every 2 min, i.e. 23 scans per female). To increase the chance of observing egg grooming (Boos et al. 2014), females involved in this measurement were separated from their clutch during 30 min before the measurements. Because earwigs are nocturnal, all behavioral tests were performed under red-light. All behavioral measurements were also done blindly regarding the treatment. We used one GLM with binomial error distribution, 2 General Linear Models (LM) and 2 Cox proportional hazard regression models to analyze the 5 measured forms of egg care. In these models, either egg gathering (Cox model), egg defense (LM), clutch abandonment (Cox model), nest construction (1 or 0; GLM) or egg grooming (LM) was used as a response variable, whereas treatment (living spores, UV-killed spores or control), population and their interaction were entered as explanatory factors. Note that the 2 Cox proportional hazard regression models allowed for censored data—that is, females that did not gather all their eggs or did not return to their clutch at the end of the observation time. When applicable, pair-wise comparisons between populations or treatments were tested using Tukey contrasts (for LM and GLM) and the survdiff function in R (for Cox models). Question 3: Do females adjust clutch quality and quantity to pathogen presence? Finally, we investigated whether females adjust the number, volume, and weight (proxies of egg quality, Koch and Meunier, 2014) of their eggs to the presence (enforced or not) of living pathogens in the environment. To this end, we measured these three traits in the clutches of the 36 females used to address question 1 (n = 11 Girona, n = 18 Montblanc and n = 8 Vincennes) and the 86 females that were in the living spores and spore-free control solution treatments in the second experiment (Question 2; n = 22 Girona, n = 45 Montblanc and n = 19 Vincennes). The number of eggs produced was counted 3 days after the first egg laying, because egg production typically takes three days in this species (Koch and Meunier, 2014). At that time, we also weighed a random subset of 10 eggs per female to the nearest of 10–3 mg using a micro scale (Pescale MYA 5), as well as measured the mean volume of the same 10 eggs (or all, if fewer were available). The mean egg volume per clutch was obtained by measuring the length and width of each of the 10 eggs to the nearest 10–3 mm and then using these values with the following formula (Meunier and Chapuisat 2009): egg volume = ((4 × pi)/3) × (egg width/2)2 × (egg length/2). All size measurements were done through a binocular microscope (Leica, Type DFC425, Leica Microsystems) with ×20.0 magnification that was run using the Leica Application Suite software (Version 4.5.0, Leica Microsystems). We finally estimated the hatching success of each of these 122 clutches by counting the total number of nymphs that eventually hatched. To control for the non-independence of egg number, volume and weight, we first conducted a Principal Component Analyses (PCA) to obtain non-correlated principal components (PC) reflecting single or combinations of egg traits. The resulting and selected PCs were then analyzed separately using LMs in which population, treatment (presence or absence of living spores on the area receiving the eggs) and type of arena (partly or fully covered with spores, i.e. Question 1 or 2, respectively) were entered as fixed factors. Independently of clutch properties, we finally analyzed the egg-hatching success using a GLM with binomial error distribution corrected for over dispersion. In this model, the hatching success was used as a response variable (entered with the cbind function in R). Moreover, population, whether females laid their eggs in Petri dishes half-covered (i.e. Experiment 1) or fully covered (i.e. Experiment 2) with living spores, and where the eggs have been laid (spores exposed or control side/Petri dish) were entered as explanatory factors. Each of these models first included interactions between all factors and was then simplified by removing the non-significant interactions (all P values > 0.073). When applicable, pair-wise comparisons between populations were tested using Tukey contrasts. RESULTS Question 1: Do females avoid infected sites to lay their eggs? When given a choice, females did not avoid the side covered with pathogens both before oviposition (Mean ± SE proportion of time spent by a female on the control side: 54.0 ± 18.0%; t48 = 1.52, P = 0.136) and at egg laying (22 and 15 females laid their eggs in the control and infected sides, respectively; exact binomial test: P = 0.324). These results were independent of female’s population, egg number and of an interaction between these 2 variables (Table 1). Overall, only 1 of the 15 females who laid their eggs on the pathogen side relocated its clutch to the control side, which is comparable to the 1 of the 22 females who relocated its eggs after having laid them on the control side. Note that this general tolerance of females for the spores’ side is unlikely to result from the unintentional spread of spores by moving females, since they were equally distributed between the 2 sides already on the first day (28 and 21 females on the Tween and spores sides, respectively; exact binomial test: P = 0.392) and the second day (22 and 27 females on the Tween and spores sides, respectively; exact binomial test: P = 0.568) of the experiment. Table 1 Effects of population and egg number on females’ likelihood (a) to avoid the pathogen-exposed side of the experimental arena before oviposition or (b) to lay eggs on the control side of the experimental arena.   (a) Before egg laying  (b) At egg laying    LR χ2  df  P  LR χ2  df  P  Population  0.92  2  0.633  2.36  2  0.308  Egg number  2.10  1  0.147  0.85  1  0.357  Population: Egg number  0.11  2  0.945  2.68  2  0.262    (a) Before egg laying  (b) At egg laying    LR χ2  df  P  LR χ2  df  P  Population  0.92  2  0.633  2.36  2  0.308  Egg number  2.10  1  0.147  0.85  1  0.357  Population: Egg number  0.11  2  0.945  2.68  2  0.262  View Large Question 2: Do females adapt their pre-hatching care to the presence of pathogens? In absence of a choice, the presence of pathogens in the environment altered two of the five measured forms of maternal care: the duration of egg abandonment and nest construction (Table 2, Figure 1). The duration of egg abandonment was overall shorter in females occupying Petri dishes covered with living spores compared to females in Petri dishes with UV-killed spores (Survdiff; χ12 = 6.5, P = 0.011) or without spores (Survdiff; χ12 = 3.4, P = 0.065), even if this last effect was marginally non-significant. The duration of egg abandonment was, however, comparable between Petri dishes covered with UV-killed spores and not covered by any spores (Survdiff; χ12 = 1.1, P = 0.287). On the other hand, the effect of spore presence on nest construction depended on females’ population (Interaction in Table 2b). In particular, females from Vincennes were more likely to build a nest in the Petri dishes covered with living spores (Figure 1b; Tukey contrasts; t = 3.63, P = 0.004) and UV-killed spores (t = 3.71, P = 0.003) compared to females in Petri dishes without any spores, with no difference between the two latter treatments (t < 0.001, P = 1.000). The presence of pathogens had, however, no effect on nest construction in females from Girona (Figure 1b; F2,28 = 0.82, P = 0.451) and Montblanc (F2,62 =2 .51, P = 0.090). Finally, the speed of egg gathering, the level of egg defense and the occurrence of egg grooming were all independent of treatment and population (Table 2c–e and Figure 2). Table 2 Effects of population and treatment on (a) the duration of clutch abandonment, (b) nest construction, (c) egg gathering, (d) egg defense, and (e) egg grooming   (a) Clutch abandonment  (b) Nest construction  (c) Egg gathering    χ2  df  P  χ2  df  P  χ2  df  P  Population  0.68  2  0.7129  17.25  2  0.0002  1.96  2  0.3744  Treatment  7.38  2  0.0250  0.78  2  0.6768  0.41  2  0.8143  Population: Treatment  3.73  4  0.4431  16.09  4  0.0029  3.85  4  0.4267  Statistical model  Cox model  GLM (binomial)  Cox model    (d) Egg defense  (e) Egg grooming      F  df  P  F  df  P    Population  1.20  2  0.3061  4.415  2  0.0143    Treatment  0.71  2  0.4949  0.163  2  0.8496    Population: Treatment  1.71  4  0.1533  0.884  4  0.4763    Statistical model  LM  LM      (a) Clutch abandonment  (b) Nest construction  (c) Egg gathering    χ2  df  P  χ2  df  P  χ2  df  P  Population  0.68  2  0.7129  17.25  2  0.0002  1.96  2  0.3744  Treatment  7.38  2  0.0250  0.78  2  0.6768  0.41  2  0.8143  Population: Treatment  3.73  4  0.4431  16.09  4  0.0029  3.85  4  0.4267  Statistical model  Cox model  GLM (binomial)  Cox model    (d) Egg defense  (e) Egg grooming      F  df  P  F  df  P    Population  1.20  2  0.3061  4.415  2  0.0143    Treatment  0.71  2  0.4949  0.163  2  0.8496    Population: Treatment  1.71  4  0.1533  0.884  4  0.4763    Statistical model  LM  LM    Significant P values are in bold. View Large Figure 1 View largeDownload slide The duration of clutch abandonment and the likelihood of nest building reflected the presence of either living spores, UV-killed spores or no spores in the nesting area. In particular, (a) Mothers abandoned their clutch of eggs less long in presence of living spores as compared to UV-killed spores and no spores at all. (b) Mothers were also more likely to build a nest in presence of UV-killed and living spores compared to no spore at all, but only when they were originating from the Vincennes population. Sample sizes are at the bottom of each bar. Different letters correspond to P values < 0.065. Figure 1 View largeDownload slide The duration of clutch abandonment and the likelihood of nest building reflected the presence of either living spores, UV-killed spores or no spores in the nesting area. In particular, (a) Mothers abandoned their clutch of eggs less long in presence of living spores as compared to UV-killed spores and no spores at all. (b) Mothers were also more likely to build a nest in presence of UV-killed and living spores compared to no spore at all, but only when they were originating from the Vincennes population. Sample sizes are at the bottom of each bar. Different letters correspond to P values < 0.065. Figure 2 View largeDownload slide The presence of either living spores, UV-killed spores or no spores in the nesting area did not influence three measured forms of maternal care: (a) the speed of egg gathering after an experimental egg spread, (b) the level of egg defense against a simulated predator attack and (c) the frequency of egg grooming. Figure 2 View largeDownload slide The presence of either living spores, UV-killed spores or no spores in the nesting area did not influence three measured forms of maternal care: (a) the speed of egg gathering after an experimental egg spread, (b) the level of egg defense against a simulated predator attack and (c) the frequency of egg grooming. Question 3: Do females adjust clutch quality and quantity to pathogen presence? The PCA conducted on the three egg traits provided three orthogonal principal components (PCs), of which we extracted the 2 first (total variance explained = 97.6 %). The first component (PC1) was highly and positively loaded with egg volume and egg weight (0.959 and 0.964, respectively; loading of egg number = −0.42), therefore reflecting egg quality. Conversely, the second component (PC2) was highly and positively loaded with egg number (0.91; loading of egg volume = 0.21; loading of egg weight = 0.19), therefore reflecting egg quantity. Overall, egg quality (PC1) and egg quantity (PC2) were independent of whether the eggs were laid with or without spores in the nesting area, or whether females laid their eggs in Petri dishes either half-covered or fully covered with living spores (i.e. the type of experimental arena; Table 3, Figure 3). By contrast, PC1 and PC2 were population-specific (Table 3, Figure 3). Egg quality (PC1) was larger in clutches produced by females from Montblanc compared to Girona, with an intermediate level in females from Vincennes (Figure 3). Conversely, egg quantity (PC2) was larger in clutches produced by females from both Girona and Montblanc compared to Vincennes, with no difference between the two first populations (Figure 3). Table 3 Effects of the type of experimental arena (females had either the choice or no choice to lay on a spore-covered area), population, and spore presence (if the eggs where laid in the spore-covered area or not) on egg quality and egg quantity   Egg quality (PC1)  Egg quantity (PC2)    F  df  P  F  df  P  Experiment  1.58  1  0.212  0.14  1  0.710  Population  6.70  2  0.008  9.61  2  <0.001  Spores presence  <0.01  1  0.966  0.03  1  0.856    Egg quality (PC1)  Egg quantity (PC2)    F  df  P  F  df  P  Experiment  1.58  1  0.212  0.14  1  0.710  Population  6.70  2  0.008  9.61  2  <0.001  Spores presence  <0.01  1  0.966  0.03  1  0.856  All interactions have been tested and were then removed from the statistical models because non-significant. Significant P values are in bold. View Large Figure 3 View largeDownload slide The population, but not the presence of spores in the nesting area shaped the quality (PC1) and quantity (PC2) of eggs produced. Pair-wise comparisons based on Tukey contrasts. Different letters P < 0.005. Figure 3 View largeDownload slide The population, but not the presence of spores in the nesting area shaped the quality (PC1) and quantity (PC2) of eggs produced. Pair-wise comparisons based on Tukey contrasts. Different letters P < 0.005. Hatching success was overall higher when females laid their eggs in Petri dishes half-covered (i.e. Experiment 1) compared to fully covered (i.e. Experiment 2) with living spores (F1,115 = 4.13, P = 0.044; Figure 4). Hatching success also depended on females’ population (F2,115 = 3.35, P = 0.039). The hatching success was higher in clutches produced by females from Girona compared to Vincennes (Z = −2.53, P = 0.030), but comparable between Girona and Montblanc (Z = −1.06, P = 0.242) and Vincennes and Montblanc (Z = −1.41, P = 0.331). The hatching success was, however, independent of the presence of living spores in the nesting area (F1,115 = 0.034, P = 0.854). Figure 4 View largeDownload slide Effect of population, possibility for females to avoid locations covered with spores (choice or no choice) and of the presence of living spores in their vicinity on hatching success. Pair-wise comparisons based on Tukey contrasts. Different letters P < 0.030. Figure 4 View largeDownload slide Effect of population, possibility for females to avoid locations covered with spores (choice or no choice) and of the presence of living spores in their vicinity on hatching success. Pair-wise comparisons based on Tukey contrasts. Different letters P < 0.030. DISCUSSION Shedding light on which maternal strategies can protect offspring against infection is of central importance to better understand the evolutionary drivers of parental care and more generally, the emergence and maintenance of family life in nature (Royle et al. 2012; Klug and Bonsall 2014). In this study, we demonstrated that the presence of pathogens in the environment influenced maternal investment into egg care, but not into egg production in the European earwig F. auricularia. Specifically, we found that the duration of egg abandonment (after an experimental disturbance) was shorter when the nesting area was covered with living spores compared to covered with UV-killed spores or no spores at all. Females were also more likely to build a nest in presence compared to absence of both living and UV-killed spores, even if this effect was only present in females from one of the three studied populations. By contrast, the presence of living spores had no effect on egg grooming, egg gathering and egg defense. We also showed that mothers did not avoid contaminated areas at egg laying, and that the presence of living spores did not influence maternal investment into egg number, volume and weight, and did not shape hatching success. Finally, our data confirm that egg properties and hatching success are population-specific in this species (see also Ratz et al. 2016). Altogether, our results show that mothers can detect the presence of spores in the environment and consequently increase their investment into certain forms of egg care. In particular, the presence of pathogens shortened the duration of egg abandonment, favored the construction of a nest (in females from Vincennes), but had no effect on maternal investment into egg gathering, egg defense and egg grooming. The effect of pathogens on the first 2 behaviors is not surprising, as both can help mothers to protect their eggs against pathogen exposure. Returning to the eggs quickly after disturbance is indeed essential to ensure the frequent physical contact between eggs and mothers that are typically required to remove pathogens from egg surface (Boos et al. 2014). Similarly, nest construction is a form of hygienic behavior that may help cleaning the nest from pathogen spores by shifting the sands around (Arathi et al. 1999). Conversely, egg gathering and egg defense are 2 maternal behaviors triggered by predator attacks (Thesing et al. 2015) and therefore unlikely to depend on pathogens presence. What was more surprising, however, is that egg grooming was independent of pathogen presence. This behavior is typically known to help individuals cleaning external parasites and pathogens in insects (Reber et al. 2011; Meunier 2015) and has been previously show to help earwig mothers cleaning their eggs from non-pathogenic fungal spores (Boos et al. 2014). The apparent discrepancy between our results and the ones from Boos et al. (2014) likely relies on the different quantity of spores covering the eggs on the days of measurement. Boos et al. (2014) directly covered the eggs with a high quantity of fungal spores and then immediately measured egg care—resulting in a high concentration of spores on the tested eggs. In our experiment, the eggs were laid in an area contaminated with fungal spores (i.e. indirect spore exposure) and were then groomed by their mothers for 15 days before we measured egg care—therefore resulting in a comparatively lower concentration of spores (if any) on the tested eggs. Overall, these studies thus suggest that egg grooming depends on the quantity of spores present on the eggs and hence, that its anti-pathogenic function is curative rather than prophylactic. Somewhat surprisingly, females exhibited different reactions in presence of UV-treated and living spores: the UV treatment inhibited the effect of spore presence on the duration of egg abandonment, but not on the likelihood of nest construction. It is generally known that individuals can change their expression of social behaviors when they encounter conspecifics that are infected or non-infected by Metarhizium spores, as reported in several eusocial insect species (Walker and Hughes 2009; Reber et al. 2011; Leclerc and Detrain 2016). However, the fact that females can change their level of care to the presence of infectious or non-infectious spores in the environment was—to the best of our knowledge—unknown. The capability to discriminate between infectious and non-infectious spores can be crucial for mothers. It may allow females to prevent their investment into costly protections against non-active threats, and instead favor investment in other fitness related processes, such as the provisioning of food to current offspring or the accumulation of energy for future reproduction (Royle et al. 2012). On a proximal level, the mediators of this apparent discrimination are unclear. M. brunneum spores typically infect hosts by adhering to their cuticle and then germinating to penetrate into their body (Vestergaard et al. 1999; Thomas and Read 2007). We hypothesize that the damages of UV on the spores may not only affect their DNA (as generally known, see Braga et al. 2015), but also their shape, chemical signature and/or adherence capability, which all could either help females discriminating against dead spores or simply make dead spores undetectable to females. Further studies should investigate which of these two scenarios explain the present finding and more generally, whether this apparent discrimination capability applies to other pathogens and hosts. Contrary to the previous spore-dependent effects, we demonstrated that earwig females showed no preference for either pathogen-exposed or pathogen-free environments for oviposition, nor did they adjust their egg properties to the presence of pathogens in the nesting area. The reported absence of pathogen-dependent effects in earwigs is uncommon among invertebrates, where most studies show that females either avoid (Meyling and Pell 2006; Lam et al. 2010) or prefer (Brütsch et al. 2014; Pontieri et al. 2014) areas covered with entomopathogenic spores for oviposition, as well as either increase (Villani et al. 1994) or reduce (Machtinger et al. 2016) their egg production in presence of entomopathogenic spores. In earwigs, this effect is unlikely to reflect females’ inability to detect spore presence (see above). Our results therefore suggest that females detected pathogens but neither chose their egg laying location nor adjusted their investment into egg production accordingly. Two non-mutually exclusive hypotheses could explain this pattern. First, the presence of spores in the environment might not be a major threat for the eggs. This is an unlikely hypothesis, as this generalist entomopathogenic fungus is a well-known threat for eggs in insects (Samuels et al. 2002; Nozad-Bonab et al. 2017) and our data demonstrate that earwig females react to spore presence by increasing the expression of the egg care specifically directed against pathogens. The second hypothesis is that maternal egg care is efficient enough to prevent the infection of eggs by fungal spores, and thus to relax selection for avoidance behavior and pathogen-dependent change in egg properties. In line with this hypothesis, a previous study demonstrated that maternal presence mitigates the costs of a direct exposure to mould on hatching success in several earwig species (Klostermeyer 1942; Miller and Zink 2012; Boos et al. 2014) and our results demonstrate that the presence of living spores had no effect on hatching success (in presence of a tending mother). Disentangling the effects of temperature and maternal care efficiency on egg production will be the goal of future studies. To conclude, our results reveal that mothers can express prophylactic mechanisms to defend their eggs against surrounding pathogens and that these defenses occur after egg production. Interestingly, these prophylactic defenses come together with other types of anti-pathogenic defenses exhibited by earwig mothers, such as the removal of fungal spores from egg surfaces (Boos et al. 2014), the transfer of chemical compounds with antimicrobial properties on the eggs (Boos et al. 2014) and the maintenance of feces with antimicrobial functions into the nest (Diehl et al. 2015; Körner et al. 2016). These results thus overall reveal that both prophylactic and curative/therapeutic defenses against microbes can co-occur in a species where nesting material is not subject to a direct competition with microbes—as compared, for instance, to carcasses in burying beetles (Rozen et al. 2008) or feces in dung beetles (Byrne et al. 2013). Active protection against microbial pathogens could thus play a central role in the emergence and maintenance of parental care in nature, as well as explain the taxonomically widespread occurrence of egg attendance in insects and arthropods (Wong et al. 2013). FUNDING This work was supported by the German Science Foundation (DFG; ME4179/3-1 to J.M.). We thank Maximilian Körner and Fanny Vogelweith for their comments on this manuscript. Data accessibility: Analyses reported in this article can be reproduced using the data provided by Meunier and Diehl (2017). REFERENCES Arathi HS, Burns I, Spivak M. 1999. Ethology of hygienic behavior in the honey bee Apis mellifera L. (Hymenoptera: Apidae): behavioral repertoire of hygienic bees. Ethology . 106: 365– 379. Google Scholar CrossRef Search ADS   Beckage NE. 2008. Insect immunology . Oxford: Elsevier Inc. Boos S, Meunier J, Pichon S, Kölliker M. 2014. Maternal care provides antifungal protection to eggs in the European earwig. Behav Ecol . 25: 754– 761. Google Scholar CrossRef Search ADS   Bos N, Sundström L, Fuchs S, Freitak D. 2015. Ants medicate to fight disease. Evolution . 69: 2979– 2984. Google Scholar CrossRef Search ADS PubMed  Bot ANM, Currie CR, Hart AG, Boomsma JJ. 2001. Waste management in leaf-cutting ants. Ethol Ecol Evol . 13: 225– 237. Google Scholar CrossRef Search ADS   Braga GUL, Rangel DEN, Fernandes EKK, Flint SD, Roberts DW. 2015. Molecular and physiological effects of environmental UV radiation on fungal conidia. Curr Genet . 61: 405– 425. Google Scholar CrossRef Search ADS PubMed  Brütsch T, Felden A, Reber A, Chapuisat M. 2014. Ant queens (Hymenoptera: Formicidae) are attracted to fungal pathogens during the initial stage of colony founding. Myrmecological news . 20: 71– 76. Byrne MJ, Watkins B, Bouwer G. 2013. Do dung beetle larvae need microbial symbionts from their parents to feed on dung? Ecol Entomol . 38: 250– 257. Google Scholar CrossRef Search ADS   Cotter SC, Kilner RM. 2010a. Personal immunity versus social immunity. Behav Ecol . 21: 663– 668. Google Scholar CrossRef Search ADS   Cotter SC, Kilner RM. 2010b. Sexual division of antibacterial resource defense in breeding burying beetles, Nicrophorus vespilloides. J Anim Ecol . 79: 35– 43. Google Scholar CrossRef Search ADS   Cremer S, Armitage SAO, Schmid-Hempel P. 2007. Social immunity. Curr Biol . 17: R693– 702. Google Scholar CrossRef Search ADS PubMed  DeVeale B, Brummel T, Seroude L. 2004. Immunity and aging: The enemy within? Aging Cell . 3: 195– 208. Google Scholar CrossRef Search ADS PubMed  Diehl JM, Körner M, Pietsch M, Meunier J. 2015. Feces production as a form of social immunity in an insect with facultative maternal care. BMC Evol Biol . 15: 40. Google Scholar CrossRef Search ADS PubMed  Diez L, Deneubourg J-L, Detrain C. 2012. Social prophylaxis through distant corpse removal in ants. Naturwissenschaften . 99: 833– 42. Google Scholar CrossRef Search ADS PubMed  Giacomello E, Marchini D, Rasotto MB. 2006. A male sexually dimorphic trait provides antimicrobials to eggs in blenny fish. Biol Lett . 2: 330– 333. Google Scholar CrossRef Search ADS PubMed  Giacomello E, Marri L, Marchini D, Mazzoldi C, Rasotto MB. 2008. Sperm-duct gland secretion of the grass goby Zosterisessor ophiocephalus exhibits antimicrobial activity. J Fish Biol . 73: 1823– 1828. Google Scholar CrossRef Search ADS   Gindin G, Ment D, Rot A, Glazer I, Samish M. 2009. Pathogenicity of Metarhizium anisopliae (Hypocreales: Clavicipitaceae) to tick eggs and the effect of egg cuticular lipids on conidia development. J Med Entomol . 46: 531– 538. Google Scholar CrossRef Search ADS PubMed  Günther K, Herter K. 1974. Dermaptera (Ohrwürmer). In: Helmcke JG, Starck D, Wermuth H, editors. Handbuch der Zoologie . Berlin: Walter de. Jackson DE, Hart AG. 2009. Does sanitation facilitate sociality? Anim Behav . 77: e1– e5. Google Scholar CrossRef Search ADS   Kaltenpoth M, Göttler W, Herzner G, Strohm E. 2005. Symbiotic bacteria protect wasp larvae from fungal infestation. Curr Biol . 15: 475– 479. Google Scholar CrossRef Search ADS PubMed  Klostermeyer EC. 1942. The life history and habits of the ringlegged earwig, Euborellia annulipes (Order Dermaptera). J Kansas Entomol Soc . 15: 13– 18. Klug H, Bonsall MB. 2014. What are the benefits of parental care? The importance of parental effects on developmental rate. Ecol Evol . 4: 2330– 2351. Google Scholar CrossRef Search ADS PubMed  Knouft JH, Page LM, Plewa MJ. 2003. Antimicrobial egg cleaning by the fringed darter (Perciformes: Percidae: Etheostoma crossopterum): implications of a novel component of parental care in fishes. Proc R Soc B Biol Sci . 270: 2405– 2411. Google Scholar CrossRef Search ADS   Koch LK, Meunier J. 2014. Mother and offspring fitness in an insect with maternal care: phenotypic trade-offs between egg number, egg mass and egg care. BMC Evol Biol . 14: 125. Google Scholar CrossRef Search ADS PubMed  Kohlmeier P, Holländer K, Meunier J. 2016. Survival after pathogen exposure in group-living insects: don’t forget the stress of social isolation!. J Evol Biol . 29: 1867– 1872. Google Scholar CrossRef Search ADS PubMed  Kölliker M. 2007. Benefits and costs of earwig (Forficula auricularia) family life. Behav Ecol Soc . 61: 1489– 1497. Google Scholar CrossRef Search ADS   Körner M, Diehl JM, Meunier J. 2016. Growing up with feces: benefits of allo-coprophagy in families of the European earwig. Behav Ecol . 27: 1775– 1781. Kramer J, Thesing J, Meunier J. 2015. Negative association between parental care and sibling cooperation in earwigs: a new perspective on the early evolution of family life? J Evol Biol . 28: 1299– 1308. Google Scholar CrossRef Search ADS PubMed  Kroiss J, Kaltenpoth M, Schneider B, Schwinger M-G, Hertweck C, Maddula RK, Strohm E, Svatoš A. 2010. Symbiotic streptomycetes provide antibiotic combination prophylaxis for wasp offspring. Nat Chem Biol . 6: 261– 263. Google Scholar CrossRef Search ADS PubMed  Lam K, Tsang M, Labrie A, Gries R, Gries G. 2010. Semiochemical-mediated oviposition avoidance by female house flies, Musca domestica, on animal feces colonized with harmful fungi. J Chem Ecol . 36: 141– 147. Google Scholar CrossRef Search ADS PubMed  Lamb RJ. 1976. Parental behavior in the dermaptera with special reference to Forficula auricularia (Dermaptera: Forficulidae). Can J Entomol . 108: 609– 619. Google Scholar CrossRef Search ADS   Leclerc J-B, Detrain C. 2016. Ants detect but do not discriminate diseased workers within their nest. Sci Nat . 103: 70. Google Scholar CrossRef Search ADS   Machtinger ET, Weeks ENI, Geden CJ. 2016. Oviposition deterrence and immature survival of filth flies (Diptera: Muscidae) when exposed to commercial fungal products. J Insect Sci . 16 (1):54, 1–6. Meunier J. 2015. Social immunity and the evolution of group living in insects. Philos Trans R Soc B Biol Sci . 370: 20140102. Google Scholar CrossRef Search ADS   Meunier J, Chapuisat M. 2009. The determinants of queen size in a socially polymorphic ant. J Evol Biol . 22: 1906– 1913. Google Scholar CrossRef Search ADS PubMed  Meunier J, Kölliker M. 2012. Parental antagonism and parent-offspring co-adaptation interact to shape family life. Proc R Soc B Biol Sci . 279: 3981– 3988. Google Scholar CrossRef Search ADS   Meunier J, Wong JWY, Gómez Y, Kuttler S, Röllin L, Stucki D, Kölliker M. 2012. One clutch or two clutches? Fitness correlates of coexisting alternative female life-histories in the European earwig. Evol Ecol . 26: 669– 682. Google Scholar CrossRef Search ADS   Meunier J, Diehl JMC. 2017. Data from: surrounding pathogens shape maternal egg care but not egg production in the European earwig. Dryad Digital Repository . http://dx.doi.org/10.5061/dryad.pv50q. Meyling NV, Pell JK. 2006. Detection and avoidance of an entomopathogenic fungus by a generalist insect predator. Ecol Entomol . 31: 162– 171. Google Scholar CrossRef Search ADS   Miller JS, Zink AG. 2012. Parental care trade-offs and the role of filial cannibalism in the maritime earwig, Anisolabis maritima. Anim Behav . 83: 1387– 1394. Google Scholar CrossRef Search ADS   Nozad-Bonab Z, Hejazi MJ, Iranipour S, Arzanlou M. 2017. Lethal and sublethal effects of some chemical and biological insecticides on Tuta absoluta (Lepidoptera: Gelechiidae) eggs and neonates. J Econ Entomol . 110: 1138– 1144. Google Scholar CrossRef Search ADS PubMed  Ormond EL, Thomas APM, Pell JK, Freeman SN, Roy HE. 2011. Avoidance of a generalist entomopathogenic fungus by the ladybird, Coccinella septempunctata. FEMS Microb Ecol . 77: 229– 37. Google Scholar CrossRef Search ADS   Parker BJ, Barribeau SM, Laughton AM, de Roode JC, Gerardo NM. 2011. Non-immunological defense in an evolutionary framework. Trends Ecol Evol . 26: 242– 248. Google Scholar CrossRef Search ADS PubMed  Pie MR, Rosengaus RB, Traniello JFA. 2004. Nest architecture, activity pattern, worker density and the dynamics of disease transmission in social insects. J Theor Biol . 226: 45– 51. Google Scholar CrossRef Search ADS PubMed  Pontieri L, Vojvodic S, Graham R, Pedersen JS, Linksvayer TA. 2014. Ant colonies prefer infected over uninfected nest sites. PLoS One . 9: e111961. Google Scholar CrossRef Search ADS PubMed  Ratz T, Kramer J, Veuille M, Meunier J. 2016. The population determines whether and how life-history traits vary between reproductive events in an insect with maternal care. Oecologia . 182: 443– 452. Google Scholar CrossRef Search ADS PubMed  Reber A, Chapuisat M. 2011. Diversity, prevalence and virulence of fungal entomopathogens in colonies of the ant Formica selysi. Insectes Soc . 59: 231– 239. Google Scholar CrossRef Search ADS   Reber A, Purcell J, Buechel SD, Buri P, Chapuisat M. 2011. The expression and impact of antifungal grooming in ants. J Evol Biol . 24: 954– 964. Google Scholar CrossRef Search ADS PubMed  de Roode JC, Lefèvre T. 2012. Behavioral immunity in insects. Insects . 3: 789– 820. Google Scholar CrossRef Search ADS PubMed  de Roode JC, Lefèvre T, Hunter MD. 2013. Self-medication in animals. Science . 340: 150– 1. Google Scholar CrossRef Search ADS PubMed  Royle NJ, Smiseth PT, Kölliker M. 2012. The evolution of parental care . Oxford: Oxford University Press. Google Scholar CrossRef Search ADS   Rozen DE, Engelmoer DJP, Smiseth PT. 2008. Antimicrobial strategies in burying beetles breeding on carrion. PNAS . 105: 17890– 5. Google Scholar CrossRef Search ADS PubMed  Samuels R, Coracini DL, Martins dos Santos C, Gava Ca. 2002. Infection of Blissus antillus (Hemiptera: Lygaeidae) eggs by the entomopathogenic fungi Metarhizium anisopliae and Beauveria bassiana. Biol Control . 23: 269– 273. Google Scholar CrossRef Search ADS   Schmid-Hempel P. 1998. Parasites in social insects . Princeton (NJ): Princeton University Press. Schmid-Hempel P. 2014. Evolutionary parasitology . Oxford: Oxford University Press. Siva-jothy MT, Moret Y, Rolff J. 2005. Insect immunity: An evolutionary ecology perspective. Adv In Insect Phys . 32: 1– 48. Google Scholar CrossRef Search ADS   Smiseth PT, Kölliker M, Royle NJ. 2012. What is parental care? In: Smiseth PT, Kölliker M, Royle NJ. The evolution of parental care . Oxford: Oxford University Press. p. 1– 17. Google Scholar CrossRef Search ADS   Stroeymeyt N, Casillas-Pérez B, Cremer S. 2014. Organisational immunity in social insects. Curr Opin Insect Sci . 5: 1– 15. Google Scholar CrossRef Search ADS   Thesing J, Kramer J, Koch LK, Meunier J. 2015. Short-term benefits, but transgenerational costs of maternal loss in an insect with facultative maternal care. Proc R Soc B Biol Sci . 282: 20151617. Google Scholar CrossRef Search ADS   Thomas MB, Read AF. 2007. Can fungal biopesticides control malaria? Nat Rev Microbiol . 5: 377– 383. Google Scholar CrossRef Search ADS PubMed  Vestergaard S, Butt TM, Bresciani J, Gillespie AT, Eilenberg J. 1999. Light and Electron Microscopy Studies of the Infection of the Western Flower Thrips Frankliniella occidentalis (Thysanoptera: Thripidae) by the Entomopathogenic Fungus Metarhizium anisopliae. J Invertebr Pathol . 73: 25– 33. Google Scholar CrossRef Search ADS PubMed  Villani MG, Krueger SR, Schroeder PC, Consolie F, Consolie NH, Preston-Wilsey LM, Donald RW. 1994. Soil application effects of Metarhizium anisopliae on Japanese Beetle (Coleoptera: Scarabaeidae) behavior and survival in turfgrass microcosms. Environ Entomol . 23: 502– 513. Google Scholar CrossRef Search ADS   Vogelweith F, Körner M, Foitzik S, Meunier J. 2017. Age, pathogen exposure, but not maternal care shape offspring immunity in an insect with facultative family life. BMC Evol Biol . 17: 69. Google Scholar CrossRef Search ADS PubMed  Walker TN, Hughes WOH. 2009. Adaptive social immunity in leaf-cutting ants. Biol Lett . 5: 446– 448. Google Scholar CrossRef Search ADS PubMed  Wong JWY, Meunier J, Kölliker M. 2013. The evolution of parental care in insects: the roles of ecology, life history and the social environment. Ecol Entomol . 38: 123– 137. Google Scholar CrossRef Search ADS   © The Author(s) 2017. Published by Oxford University Press on behalf of the International Society for Behavioral Ecology. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com

Journal

Behavioral EcologyOxford University Press

Published: Jan 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 12 million articles from more than
10,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Unlimited reading

Read as many articles as you need. Full articles with original layout, charts and figures. Read online, from anywhere.

Stay up to date

Keep up with your field with Personalized Recommendations and Follow Journals to get automatic updates.

Organize your research

It’s easy to organize your research with our built-in tools.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

Monthly Plan

  • Read unlimited articles
  • Personalized recommendations
  • No expiration
  • Print 20 pages per month
  • 20% off on PDF purchases
  • Organize your research
  • Get updates on your journals and topic searches

$49/month

Start Free Trial

14-day Free Trial

Best Deal — 39% off

Annual Plan

  • All the features of the Professional Plan, but for 39% off!
  • Billed annually
  • No expiration
  • For the normal price of 10 articles elsewhere, you get one full year of unlimited access to articles.

$588

$360/year

billed annually
Start Free Trial

14-day Free Trial