Temperature stress induces mites to help their carrion beetle hosts by eliminating rival blowflies

Syuan-Jyun Sun; Rebecca M Kilner

doi:10.7554/eLife.55649

. 2020 Aug 5;9:e55649. doi: 10.7554/eLife.55649

Temperature stress induces mites to help their carrion beetle hosts by eliminating rival blowflies

Syuan-Jyun Sun ^1,^2,^✉, Rebecca M Kilner ¹

Editors: Dieter Ebert³, Christian Rutz⁴

PMCID: PMC7431131 PMID: 32755542

Abstract

Ecological conditions are known to change the expression of mutualisms though the causal agents driving such changes remain poorly understood. Here we show that temperature stress modulates the harm threatened by a common enemy, and thereby induces a phoretic mite to become a protective mutualist. Our experiments focus on the interactions between the burying beetle Nicrophorus vespilloides, an associated mite species Poecilochirus carabi and their common enemy, blowflies, when all three species reproduce on the same small vertebrate carrion. We show that mites compete with beetle larvae for food in the absence of blowflies, and reduce beetle reproductive success. However, when blowflies breed on the carrion too, mites enhance beetle reproductive success by eating blowfly eggs. High densities of mites are especially effective at promoting beetle reproductive success at higher and lower natural ranges in temperature, when blowfly larvae are more potent rivals for the limited resources on the carcass.

Research organism: Other

Introduction

Protective mutualisms among macro-organisms are both widespread and well-known (Clay, 2014; Palmer et al., 2015; Hopkins et al., 2017). They involve one species defending another species from attack by a third party species, in exchange for some form of reward (Clay, 2014; Palmer et al., 2015; Hopkins et al., 2017). Theoretical analyses predict that mutualisms like this can evolve when a commensal or mildly parasitic species, that lives in or upon its host, is induced to become a protective mutualist upon exposure to an environmental stressor (Fellous and Salvaudon, 2009; Lively et al., 2005; Hopkins et al., 2017; Rafaluk-Mohr et al., 2018). The stressor can be biotic (Ashby and King, 2017; Clay, 2014; Ewald, 1987; Lively et al., 2005; Schwarz and Müller, 1992) or abiotic (Corbin et al., 2017; Engl et al., 2018; Hoang et al., 2019).

Although the adaptive evolution of mutualisms has been studied in detail, the contextual factors that drive equivalent variation in the expression of mutualisms on an ecological timescale are relatively less well understood (Chamberlain et al., 2014; Jaenike et al., 2010; Hoeksema and Bruna, 2015), especially for protective mutualisms (Hopkins et al., 2017; Palmer et al., 2015). In particular, it is unclear how different biotic and abiotic factors combine to influence the expression of a mutualism, especially when conditions vary locally. Nor is it well understood whether the extent of mutualism is density-dependent (Hoeksema and Bruna, 2015; Palmer et al., 2015). Here we investigate how biotic and abiotic stressors combine to induce the context-dependent expression of a protective mutualism. Specifically, we determine how temperature and partner density interact with the presence of a third party enemy species to influence the likelihood that a phoretic organism can be induced within a single generation to become a protective mutualist.

Our experiments focus on burying beetles (Nicrophorus vespilloides), which use the dead body of a small vertebrate to breed upon (Scott, 1998). A pair of beetles works together to convert the carcass into an edible carrion nest for their larvae by removing any fur or feathers, and rolling the meat into a ball. The beetles also reduce competition with rival species for the resources on the dead body by smearing the flesh in antimicrobial exudates, consuming eggs laid by rival insects and concealing the body below ground (Chen et al., 2020; Duarte et al., 2018; Scott, 1998). During carcass preparation, beetle eggs are laid in the surrounding soil and then hatch within 3–4 days. The larvae crawl to the carcass and feed themselves on the edible nest, where they are also fed and defended by both parents. Within a week of hatching, the larvae disperse away from the scant remains of the carcass to pupate, while adults fly off – often to breed again.

Adult burying beetles carry up to 14 species of mites, which also breed on carrion and which use the burying beetle as a means of transport between breeding opportunities. The Poecilochirus carabi species complex is the most salient and common of these mite species (e.g. Wilson, 1983; Schwarz et al., 1998), and it is the focus of this study. P. carabi travels as sexually immature deutonymphs on the burying beetle, and derives no nourishment directly from its host while it is on board (Wilson and Knollenberg, 1987). Upon arrival at a carcass, the deutonymphs alight and moult into adults, which then reproduce. The next generation of mite deutonymphs is ready to disperse by the time the adult burying beetles cease caring for larvae and leave the breeding event. Roughly 90% of deutonymphs disperse on the departing adults rather than on the burying beetle’s larvae (Schwarz and Müller, 1992).

P. carabi is often described as a phoretic mite because it uses burying beetles (Nicrophorus spp.) to travel between breeding opportunities on carrion, and seemingly imposes few costs on its hosts during transportation. Phoretic interactions are thought to pave the way for further interactions between host and phoront that have more positive or negative effects on host fitness. This is especially likely when interactions between host and phoront endure beyond the transport phase (White et al., 2017). For example, female Trichogramma parasitoid wasps hitch a relatively cost-free ride to their butterfly hosts’ egg-laying site, but upon arrival are easily able to locate butterfly eggs to parasitise (Fatouros and Huigens, 2012). Likewise, the phoretic mite Ensliniella parasitica travels on female mason wasps Allodynerus delphinalis. Female wasps lay a single egg in a brood cell within a dead plant, and provision the cell with paralysed caterpillars and a few phoretic mites. The mites are mildly parasitic because they feed on the developing wasp’s haemolymph (Okabe and Shun'ichi, 2008). However, if the wasp pupae are threatened by parasitoid wasps, the mite protects them from attack, thus switching from parasite to mutualist (Okabe and Shun'ichi, 2008). Nevertheless phoretic interactions are generally under-studied and their capacity to extend into further interactions that influence host fitness remains poorly understood (White et al., 2017).

For burying beetles, their phoretic relationship with P. carabi mites changes once the beetle has located the dead body. This study focuses entirely on the interactions that take place from that point onwards, during reproduction. The intimate association between beetles and mites continues through frequent contact as the two species breed alongside each other on the small dead body, and this enables each party to influence the other’s fitness. We characterize the changing relationship between the mite and the beetle by measuring the fitness outcome for each of them (Figure 1—figure supplement 1).

The beetle has a net positive effect on mite fitness. Without the beetle, the mite would not be able to breed at all. Furthermore, mites have greater reproductive success on beetle-prepared carrion than on other dead meat (Sun and Kilner, 2019). However, in some contexts, the mite reduces burying beetle fitness. Mite offspring compete with burying beetle larvae for resources on the carcass, and can directly predate upon beetle eggs and newly-hatched larvae (Wilson, 1983; Beninger, 1993; De Gasperin and Kilner, 2015). Thus, in some contexts the mites are harmful for the burying beetle.

In other contexts, though, the mite can potentially become a protective mutualist by defending burying beetle reproductive success when it is threatened by an enemy species (Wilson, 1983). Blowflies (Calliphoridae) are a particular competitive threat for burying beetles (Scott, 1994; Sun et al., 2014) because they can locate the newly dead more rapidly than burying beetles (within a few hours: Shelomi et al., 2012); personal observations) and start to lay eggs within minutes of arriving on the dead body (Bornemissza, 1957; Matuszewski et al., 2010; Payne, 1965). Mites can potentially prevent burying beetles from losing fitness to rival blowflies by eating blowfly eggs (Springett, 1968). As an indirect effect of the mites’ predatory actions, the net fitness outcome of the mite-beetle interaction becomes positive-positive. Since the mite is only able to feed upon blowflies because it was transported to the carrion by the burying beetle, the mite becomes a mutualist.

Two other factors additionally seem likely to determine whether mites have negative or positive effects on the fitness of their burying beetle hosts: temperature and mite density per host. Previous work has shown that at higher temperatures blowflies pose a greater threat to burying beetle and mite fitness. Blowflies are more abundant on carrion at higher temperatures, develop more rapidly and have higher reproductive success (Sun et al., 2014; Wall et al., 1992). High densities of mites might be more effective at protecting from blowflies under these conditions (Okabe and Shun'ichi, 2008). Yet high densities of phoretic mites and phoretic nematodes are also known to reduce the number and quality of burying beetle larvae produced, potentially making mites more harmful (De Gasperin and Kilner, 2016; Wang and Rozen, 2019). Therefore it is unclear how these three factors (temperature, mite density, and the presence of blowflies) interact to determine whether interactions between mites and their burying beetles are harmful to beetles or more mutualistic.

We used field and laboratory experiments on burying beetles and their P. carabi mites to determine how the effects of blowflies, temperature and mite density combine to influence the expression of a protective mutualism. Our experiments were designed specifically to investigate whether: 1) the presence of blowflies causes mites to switch from being harmful to becoming protective mutualists; 2) whether any transition to and from mutualism is modulated by temperature; and 3) whether any transition is additionally mediated by the density of mites on the carrion.

Results

Complementary patterns of reproductive success in burying beetles and blowflies, in the field

We found that the reproductive success of burying beetles and blowflies varied with temperature, though in a complementary pattern (Figure 1A and B). Whereas burying beetle reproductive success peaked at intermediate temperatures, and dipped at lower and higher temperatures (Figure 1A and Supplementary file 1a), blowflies had greatest reproductive success at lower and higher temperatures and much less success at intermediate temperatures (Figure 1B and Supplementary file 1a).

Figure 1—figure supplement 1. — Shaded regions represent 95% confidence intervals, and solid and dashed lines represent statistically significant and non-significant regression lines from GLMM, respectively. Each datapoint represents one breeding event.

Mites enhance burying beetle fitness in the field when there are blowflies present, but the effect depends on temperature and mite density

Adding mites to the breeding event changed these relationships, for both beetles and blowflies, though in different ways at different mite densities. When we added 10 mites, there was little effect on the overall reproductive success of beetles (Figure 1C; Supplementary file 1a), though mites significantly reduced the reproductive success of the blowflies at lower and higher temperature ranges (Figure 1D; Supplementary file 1a). When we added 20 mites, however, mites were especially effective at promoting beetle reproductive success at these same lower and higher temperatures (Figure 1E; Supplementary file 1a). Once again, they caused a corresponding decline in the success of blowflies breeding at lower and higher temperatures (Figure 1F; Supplementary file 1a).

Turning to the mites’ perspective, we found that variation in their reproductive success could not be explained by temperature (Supplementary file 1a). From these initial results, we conclude that mites act as protective mutualists for burying beetles against blowflies in natural breeding conditions, matching results obtained previously for a different burying beetle species (Wilson, 1983), and that their effects are contingent on mite density per breeding event. Our results extend the findings of previous work by showing that mites promote burying beetle reproductive success specifically at lower and higher temperatures.

Complementary patterns of reproductive success in burying beetles and blowflies are induced by each other in the lab

Next, we analysed data from Laboratory Experiment 1, focusing first on the effects of blowflies on burying beetle reproductive success, when there were no mites present (Figure 2A v. 2D). We found that blowflies reduced burying beetle reproductive success at lower and higher temperatures (interaction blowfly treatment x temperature treatment, χ² = 25.85, d.f. = 2, p<0.001), and that blowflies caused greater reduction at higher temperatures than at lower temperatures (post-hoc comparison high v. low, z = −2.47, p=0.036).

Figure 2. — Sample sizes are shown above each boxplot. Boxplots show median (solid line), first quartile (bottom of box), third quartile (top of box), values that fall within 1.5 times of the interquartile range (dotted lines), and outliers (open circles). Each datapoint represents one breeding event.

Figure 2—figure supplement 1. — Sample sizes are shown above each boxplot. Boxplots show median (solid line), first quartile (bottom of box), third quartile (top of box), values that fall within 1.5 times of the interquartile range (dotted lines), and outliers (open circles). Each datapoint represents one breeding event.

To determine whether beetles likewise influenced blowfly reproductive success, we compared the number of blowfly larvae produced in Laboratory Experiment 1 with the number of blowfly larvae produced in Laboratory Experiment 2, when there were no beetles present. We found that burying beetles substantially reduced blowfly reproductive success but that the effect was temperature-dependent (interaction beetle x temperature treatments: χ² = 38.32, d.f. = 2, p<0.001). Blowfly reproductive success was most strongly reduced by beetles at intermediate temperatures (z = 10.59, p<0.001), with a less pronounced decrease at lower temperatures (z = 9.40, p<0.001), and the least change of all at higher temperatures (z = 7.04, p<0.001).

Blowflies are enemies to mites

Further analyses of Laboratory Experiment 1 revealed that blowflies reduced mite reproductive success (Figure 2—figure supplement 2; Supplementary file 1b) and that the extent of mite fitness loss was modulated by temperature (Supplementary file 1b). We found that blowflies reduced mite reproductive success at mid and higher temperatures (mid temperatures: post-hoc comparison without blowflies v. with blowflies, z = 2.24, p=0.025; higher temperatures: post-hoc comparison without blowflies v. with blowflies, z = 3.29, p=0.001). However, blowflies had no effect on mite reproductive success at lower temperatures (post-hoc comparison without blowflies v. with blowflies, z = 0.30, p=0.766). Temperature thus modulates the negative effects of the blowfly on both burying beetle and mite fitness (Supplementary file 1b).

In the lab, mites reduce burying beetle fitness at high densities when blowflies are absent

Adding mites generally reduced burying beetle reproductive success, though to different degrees at different mite densities (Figure 2A-C; Supplementary file 1b). Across all temperatures, mites had no effect on beetle reproductive success in groups of 10 (post-hoc comparison 0 v. 10 mites, z = 1.49, p=0.298). However, adding 20 mites significantly reduced beetle reproductive success (post-hoc comparison 0 v. 20 mites, z = 3.20, p=0.004). Therefore, mites have mildly negative effects on burying beetle fitness, as has been reported before in previous work on N. vespilloides (Beninger, 1993; De Gasperin and Kilner, 2015; Nehring et al., 2017; Sun et al., 2019) and other Nicrophorus species (Wilson and Knollenberg, 1987).

Nevertheless, the loss in beetle reproductive success caused by mites at high temperatures was much less than that induced by blowflies (post-hoc comparison 0 mites, with blowflies v. 10 mites, without blowflies, z = −3.61, p=0.002; post-hoc comparison 0 mites, with blowflies v. 20 mites, without blowflies, z = −2.85, p=0.023).

Mites switch from being harmful to mutualistic at lower and higher temperatures

We found that the presence of blowflies caused mites to switch to becoming more mutualistic. Furthermore, the extent of mutualism was dependent both on temperature and mite density, matching our findings in the field. At lower temperatures, neither density of mites affected beetle reproductive success when blowflies were present (post-hoc comparison 0 v. 10 mites, z = −0.77, p=0.720; Figure 2E; post-hoc comparison 0 v. 20 mites, z = −0.60, p=0.822; Figure 2F). At higher temperatures, 10 mites had no effect on burying beetle reproductive success either (post-hoc comparison 0 v. 10 mites, z = −1.03, p=0.560; Figure 2E). However, when 20 mites were added to the breeding event, they increased beetle reproductive success but only at higher temperatures (post-hoc comparison 0 v. 20 mites, z = −3.04, p=0.007; Figure 2F).

The increase in beetle reproductive success was matched by a corresponding mite-induced decline in blowfly reproductive success (Figure 3), with the pattern of decline again matching the results of our field experiment (Figure 1B). When there were no mites present, blowflies breeding alongside burying beetles had much greater reproductive success at higher temperatures and lower temperatures than at intermediate temperatures (post-hoc comparison high v. mid temperature, z = 5.61, p<0.001; post-hoc comparison low v. mid temperature, z = 3.21, p=0.004; Figure 3A).

In summary, the field and lab experimental results each suggest that burying beetles can manage singlehandedly to defend their reproductive success against blowflies at intermediate temperatures, but that they struggle to produce as many larvae at higher and lower temperatures (Figure 1B, Figure 2D). These are the temperatures at which blowflies have highest reproductive success when there are no mites present. Although adding 10 mites did not cause a significant reduction in the number of blowfly larvae produced (lower temperatures: post-hoc comparison 0 v. 10 mites, z = 1.76, p=0.183; higher temperatures: post-hoc comparison 0 v. 10 mites, z = −0.65, p=0.792; Figure 3B), adding 20 mites to the breeding event caused blowflies to perform badly at all temperatures (Figure 3C).

How are burying beetles (at intermediate temperatures) and mites (at lower and higher temperatures) able to cause such a reduction in blowfly reproductive success? Both species wander all over the carrion nest, especially during carcass preparation before the burying beetle larvae hatch (Smiseth et al., 2003). They graze on the surface of the carrion as they go, and have been observed to consume blowflies when they are eggs or newly hatched 1^st instar blowfly larvae (Wilson, 1983; Wilson and Knollenberg, 1987). The likelihood that blowfly eggs will be eaten therefore depends partly on the duration of these vulnerable early life stages during blowfly development, and partly on the extent to which beetles and mites prey upon blowflies. We tested whether each is temperature dependent.

At higher temperatures, blowflies evade attack through more rapid development

We found that temperature could not explain any variation in either blowfly reproductive success (Figure 4—figure supplement 1; Supplementary file 1c), or the extent to which blowfly larvae consumed the carcass (Figure 4—figure supplement 1; Supplementary file 1c). However, blowfly development was greatly accelerated at higher temperatures (Figure 4A; Supplementary file 1c), with blowflies spending significantly less time as eggs and 1^st instar larvae at higher temperatures than at lower temperatures (eggs: t = −3.76, p<0.001; 1^st: t = −4.89, p<0.001).

At lower temperatures, beetle defences against blowflies are weaker

When we compared the number of blowfly larvae produced in Laboratory Experiment 2, when beetles were able to prepare a carcass, and Laboratory Experiment 3, when beetles were absent, we found that carcass preparation by beetles reduced the number of blowfly larvae produced and but that its effectiveness was sensitive to temperature (interaction carcass preparation x temperature treatments: χ² = 19.67, d.f. = 2, p<0.001). Blowflies showed the greatest loss in fitness at intermediate temperatures (z = 9.84, p<0.001) with a less marked reduction in fitness at lower (z = 5.16, p<0.001) and higher temperatures (z = 6.25, p<0.001).

We found that the effectiveness of carcass preparation by beetles varied with temperature (Figure 4B; Supplementary file 1d). Specifically, beetles converted a dead body into a rounder nest for their larvae at both higher and mid temperatures than at lower temperatures (post-hoc comparison high v. low, z = 4.68, p<0.001; post-hoc comparison low v. mid, z = −4.83, p<0.001). The rounder the prepared carcass was, the fewer the blowfly larvae that survived (χ² = 13.78, d.f. = 1, p<0.001; Figure 4C).

The combined effects of temperature on both carcass preparation by beetles and blowfly development, explain why blowflies are able to produce more larvae at higher and lower temperatures than at mid temperatures - and therefore why they pose more of a threat to burying beetle and mite fitness at these temperatures. Burying beetles can singlehandedly defend themselves against blowflies at intermediate temperatures through their activities during carcass preparation. At higher temperatures, blowflies develop sufficiently rapidly that they can evade these beetle defences. At lower temperatures, burying beetles are less able to defend themselves against blowflies during carcass preparation.

Discussion

The aim of this study was to determine how biotic and abiotic factors combine to influence the context-dependent expression of a protective mutualism, using the changeable interactions between burying beetles and their mites as a model system. Our experiments reveal a web of direct and indirect ecological interactions between burying beetles, P. carabi mites and blowflies as they breed alongside each other on small carrion (see Figure 5). The web is partly constructed by the burying beetles themselves, because they alone transport mites to the carrion. However, the interaction between burying beetles and their P. carabi mites depends on whether blowflies are present too - because predation by mites on blowfly eggs then indirectly enhances burying beetle reproductive success. The extent of mutualism also varies with increasing temperature stress, and with increasing mite density. All three factors cause a corresponding change in the net fitness outcome for burying beetles and this determines whether the mite harms burying beetle fitness or is more mutualistic (Figure 5).

Figure 5. — Direct interactions between species are shown with solid lines while indirect interactions are shown with dashed lines. The arrow points to the species whose fitness is affected by the focal species. The signs (+/-) indicate positive or negative effects on fitness. Our overall conclusion is that a temperature-enhanced threat from blowflies causes mites to become protective mutualists of their burying beetle hosts.

(1) Do blowflies cause mites to switch from being harmful to becoming protective mutualists?

Consistent with previous work on other burying beetle species (Wilson, 1983), we found that mites were antagonistic to beetles at all temperatures in the absence of blowflies (Figure 2). A similar decrease in the extent of mutualism has been detected in other protective mutualisms when the third-party enemy species is absent or removed (Hopkins et al., 2017). Then, it is common for the host to reduce the rewards it offers its protective mutualist (Palmer et al., 2015; Palmer et al., 2008). It is unclear whether this happens in burying beetles too. However, the main service that beetles offer to mites is transport to carrion. This means that the beetles’ payment to the mites would have to be modulated either in advance of their protection service, when mites are transported to carrion, or retrospectively, when the adult beetles fly off carrying the mites’ offspring with them at the end of reproduction. Either way, since the prevalence of blowflies is likely to vary locally from one breeding attempt to the next, it is hard to see how beetles could accurately modulate the transport service they offer to mites in relation to the prevalence of blowflies. An alternative possibility is that some of the other mite species carried by burying beetles in nature (which we excluded from our experiments), or the phoretic nematodes that are also present upon the beetle (Wang and Rozen, 2019) modulate the harm inflicted by P. carabi on its burying beetle host. Whether this actually happens, however, remains to be determined in future work.

(2) Is the expression of the protective mutualism modulated by temperature?

Previous studies have emphasised the significance of the abiotic environment in shifting the outcome of species interactions (Chamberlain et al., 2014; Gorter et al., 2016; Hoeksema and Bruna, 2015; Hopkins et al., 2017). Protective mutualisms sometimes break down at higher temperatures because the protecting partner is more vulnerable to heat stress when temperatures rise (Barton and Ives, 2014; Doremus and Oliver, 2017; Fitzpatrick et al., 2014). However, we found no evidence that mites were more vulnerable to higher temperatures, whether in field or laboratory conditions. Instead, the main driver of change in the protective mutualism came from the response of enemy blowflies, and the behaviour of the burying beetles themselves, to variation in temperature (Figure 4). We suggest that similar effects might be found in other protective mutualisms where enemy species are more likely to thrive at high temperatures, providing that both partners can tolerate some thermal stress. Predicting how populations might respond to more variable temperatures thus involves understanding its interactions within the natural ecological community as well as some knowledge of the intrinsic variation in the thermal tolerance of the mutualistic partner (Early and Keith, 2019).

(3) Is the expression of the protective mutualism modulated by the density of mites?

The mites’ capacity to defend burying beetles against competition from blowflies was both temperature-dependent and density-dependent. In the field and in the lab, blowflies posed a greater threat to burying beetle fitness at higher temperatures and then it took a high density of mites to neutralize this danger. Increased mite density has been found to influence the effectiveness of defences against enemy species in other protective mutualisms as well (e.g. Okabe and Shun'ichi, 2008). Our experiments captured the likely variation in mite density at natural breeding events. However, we have no evidence to suggest that beetles can regulate the density of mites they carry in anticipation of the threats they face to their reproductive success (Sun et al., 2019).

In conclusion, we have shown how the expression of a protective mutualism between burying beetles and their P. carabi mites is context-dependent and depends on a complex interplay of biotic and abiotic factors. In common with other facultatively expressed mutualisms (Afkhami et al., 2014; Johnson, 2015; Peay, 2016), short-term variation in the expression of this protective mutualism may influence the capacity of its host burying beetle to persist in adverse environments.

Materials and methods

Burying beetles and phoretic mites in Madingley Wood

Fieldwork was carried out at Madingley Woods in Cambridgeshire UK, an ancient woodland (Goldberg et al., 2007) of mixed deciduous trees near the Sub-Department of Animal Behaviour, University of Cambridge, (Latitude: 52.22730°; Longitude: 0.04442°). We trapped N. vespilloides carrying the mite P. carabi by setting Japanese beetle traps, baited with ~30 g fresh mice, from June to October, 2016–2017. Ambient air temperature was recorded locally at 1 hr intervals using an iButton temperature data logger (n = 8; DS1922L-F5#, Maxim Integrated Products, Inc), which was suspended alongside each trap at 1 m above the ground, and shielded from direct exposure to sunlight. Traps were checked daily to determine when the beetles first located the dead body within. The mean ± S.E.M. time to discovery was 3.42 ± 0.77 days. Each trap was emptied every two weeks, and re-baited with a fresh mouse carcass. At this point, we took the contents back to the lab and counted the total number of N. vespilloides caught in the trap and the number of P. carabi carried by each individual beetle. Beetles were temporarily anaesthetized using CO₂ and mites were then detached with a fine brush and tweezers. Field-caught burying beetles naturally carried a mean ± S.E.M. of 10.82 ± 0.45 mites (see Figure 2—figure supplement 1 from Sun et al., 2019 for frequency distribution of mite density), while 70% of them carried 1–20 mites (n = 1369 beetles). Field-caught beetles, mites, and blowfly larvae collected from the traps were used to establish laboratory colonies (see below).

Field experiment: how does burying beetle reproductive success covary with blowflies, mite density and ambient air temperature?

Experimental breeding events were staged in Madingley Woods. Breeding events were established at 20 different sites (see Figure 1—figure supplement 1), separated from each other by approx. 30 m. Each site was used more than once during the course of the burying beetle’s breeding season. We recorded ambient temperature during each experiment by using iButton temperature data loggers placed at 1 m above ground at 1 hr intervals throughout. The set-up for each breeding event is shown in Figure 1—figure supplement 2. A 8–16 g (12.40 ± 0.15 g) mouse carcass was placed on the compost and left for three days, to simulate the average time taken by beetles to locate a carcass in the field (see above). Blowflies that were naturally present in the woodland were able to lay their eggs opportunistically on the mouse corpse too, while it remained above ground. We then added a pair of burying beetles from the laboratory colony. We also added mites from the lab colony at one of three different densities: 0 (n = 66), 10 (n = 68), or 20 mite (n = 61) deutonymphs. We staged 195 breeding events in all. Each experiment was terminated either when the beetle larvae dispersed or when the dead body was completely consumed by blowfly larvae. At this point we measured components of beetle fitness (number of beetle larvae; see below), blowfly fitness (number of blowfly larvae), and mite fitness (number of dispersing mite deutonymphs on adult beetles).

Maintenance of laboratory colonies of beetles, mites, blowflies

Burying beetles

We bred burying beetles by introducing pairs of unrelated males and females to a mouse carcass (7–15 g) in a plastic container (17 × 12 × 6 cm filled with 2 cm of moist soil). All larvae were counted and collected at dispersal, and transferred to eclosion boxes (10 × 10 × 2 cm, 25 compartments) filled with damp soil. Once they had developed into adults, beetles were kept individually in plastic containers (12 × 8 × 2 cm) filled with moist soil, and were fed twice a week with small pieces of minced beef.

Mites

We maintained mite colonies in plastic containers (17 × 12 × 6 cm filled with 2 cm of moist soil). Each container was provided with an adult beetle and fed with pieces of minced beef twice a week. We bred mites once a month by introducing 15 mite deutonymphs to a pair of beetles and a mouse carcass in plastic containers (17 × 12 × 6 cm filled with 2 cm of moist soil; n = 10). When the burying beetle larvae had completed their development, we collected mite deutonymphs that were dispersing on adult beetles. Newly-emerged mites were reintroduced to the containers holding the mite colony.

Blowflies

Colonies of blowflies Calliphora vomitoria (n = 5 colonies) were reared in screened cages (32.5 × 32.5 × 32.5 cm). They were continuously supplied with a mixture of powdered milk and dry granulated sugar, and ad lib. water. We fed newly emerged blowflies with pig liver to induce maturation of the flies’ ovaries. After a week, these blowflies were then given mouse carcasses to breed upon. All beetle, mite, and blowfly colonies were kept at 21 ± 2°C with a photoperiod of 16:8 light:dark.

Laboratory Experiment 1: manipulations of blowflies, mites and temperature

To understand how temperature and mite density together mediate blowfly competition with burying beetles, we repeated the field experiment in a lab setting so that we could manipulate temperature and the presence of blowflies as well as mite density.

Manipulating the presence/absence of blowflies

We placed 30 mg (30.22 ± 0.07 mg) newly-laid blowfly eggs onto a 7–16 g (11.13 ± 0.15 g) mouse carcass before giving it to beetles to breed upon, to mimic the rapid oviposition by blowflies in nature on a freshly dead vertebrate (Wilson, 1983). As a control, dead mice of similar size (10.64 ± 0.15 g) were kept free of blowflies. In both blowfly treatments, the dead mouse was placed on the soil in a breeding box in a temperature-regulated breeding chamber for 3 days before adding the beetles, simulating the later arrival time of the beetle at the carcass that is seen in nature (see above). During this time, the fly eggs were able to hatch and the blowfly larvae started to consume the carcass.

Manipulations of mite density

We used the same treatment as in the field experiment: 0, 10, or 20 mites. Mite deutonymphs were introduced to the dead mouse at the same time as the burying beetles.

Manipulations of temperature

The six treatments described above were each staged in temperature-regulated breeding chambers (Panasonic MLR-352-PE). Each temperature treatment mimicked the 8°C diurnal temperature fluctuation that is typical for Madingley Woods, during the burying beetle’s breeding season (Figure 2—figure supplement 1). The mean temperature for each manipulation was 11, 15, and 19°C, which matches the mean seasonal low, intermediate, and high temperatures, respectively, in Madingley Woods (Figure 2—figure supplement 1). Each of the six treatments was carried at these three temperatures, generating a fully factorial experiment with 18 treatments in all (three mite treatments (0, 10 or 20 mites) x two blowfly treatments (blowfly or no blowfly) x three temperature treatments (11, 15, and 19°C). At the end of each breeding bout, indicated by either the beetle larvae starting to disperse away or carcass consumption by blowfly larvae, whichever came sooner, we measured the fitness components of beetles, mites, and blowflies using the methods described above in the field experiments. For logistical reasons, replicates of all 18 treatments were evenly spread over four blocks, carried out in succession.

Laboratory Experiment 2: effect of temperature on blowfly development

To examine how blowflies respond to temperature, in the absence of the mites and the burying beetles, we counted the number of dispersing blowfly larvae, and the rate of carcass consumption, at the three different temperatures used in laboratory experiment 1 (11, 15, and 19°C; n = 13 carcasses for each temperature treatment). Once again, we placed blowfly eggs (30.22 ± 0.09 mg) on a mouse carcass (10.74 ± 0.30 g) that sat on soil in a plastic breeding box, and put the box in a temperature-controlled breeding chamber. (No burying beetles or mites were added this time). Every 12 hr we checked the boxes and determined the stage of blowfly larval development attained, namely 1^st, 2^nd, 3^rd instars and post-feeding. In addition, we recorded when the carcass entered the bloating stage (indicated by swelling and putrefaction). When the larvae entered the post-feeding stage, we counted them, and recorded their total mass. From these data, we determined the proportion of carcass consumed, calculated as total mass of larvae divided by initial carcass mass.

Laboratory Experiment 3: effect of temperature on beetle defences against blowflies during carcass preparation

To understand the effect of temperature on the effectiveness of carcass preparation by burying beetles in defending against infestation by blowflies, we placed blowfly eggs (30.05 ± 0.09 mg) on a mouse carcass (13.25 ± 0.24 g) prior to introducing pairs of beetles at three different temperatures (11, 15, and 19°C; n = 23, 23, 22 carcasses for each temperature treatment, respectively). This time, each carcass was transferred to a new plastic breeding box once the beetles had completed carcass preparation but before their eggs had hatched. Once the carcass had been moved, it was kept at the same intermediate temperature regardless of the temperature treatment previously experienced during carcass preparation. This allowed us to isolate the effects of temperature on beetle carcass preparation, and its relation to subsequent blowfly fitness.

We quantified the extent of carcass preparation by measuring the sphericity of each prepared carcass, using previously established methods (De Gasperin et al., 2016), calculating roundness from a two-dimensional proxy. Each carcass was photographed against a white background from the top and the side using two identical digital cameras (Fuji lm av200), each kept at a constant distance of 30 cm to the carcass. We processed the images with white circle to remove legs, tails, and large pieces of soil in GIMP (version 2.6.11), prior to roundness analysis. We estimated the roundness from each image using a boundary tracing routine, bwboundaries, in Matlab (The Mathworks, USA). Each image was separated from the white background with a filter of 5 pixels to remove the smallest details, such as hairs and soils smaller than 1 mm (the photographs taken from the top and side were 6.4 and 6.36 pixels per mm, respectively). The roundness was then determined by calculating a metric, 4π∗area/perimeter2, in which a score of 1 denotes a perfect circle. An overall roundness score was derived by averaging roundness of the top and the side images of each carcass.

Statistical analyses

Generalised linear mixed model (GLMM) analyses were carried out in the statistical programme R 3.4.3 using the package lme4 (Bates et al., 2015). Model formulae are given in the tables of results (see Supplementary file 1). Non-significant interaction terms were dropped from the analyses before deriving the final model. As is common statistical practice (e.g. Gelman and Hill, 2007), if we found a significant interaction term, we split the dataset accordingly to determine how the interaction arose. Power analyses were performed based on 1000 Monte Carlo simulations, with the function powerSim in the package SIMR (Green and MacLeod, 2016).

Field experiment

We sought correlates of beetle brood size, the number of blowfly larvae, and the number of mite offspring number at the end of each trial, using separate GLMMs each with negative binomial distributions. For the models with beetle brood size and the number of blowfly larvae as independent variables, we included the variables carcass mass, mite treatment (0, 10, 20 mites), temperature, and the interaction between mite treatment and temperature. Mite treatment was a categorical factor, whereas carcass mass and temperature were continuous variables. Temperature was calculated as the average daily mean temperature, from carcass presentation to larval dispersal (or carcass consumption by blowfly larvae). We also included a squared measure of temperature in the model because we found that the non-linear effects of temperature explained more variation than any linear effects. (We compared the performance of different models using the Akaike Information Criterion (AIC), using the function model.sel in the package MuMIn, and obtained the following results. Models of burying beetle reproductive success: with temperature as a non-linear variable: AICc = 802.2, Akaike weight = 0.93 v. with temperature as a linear variable: AICc = 807.4, Akaike weight = 0.07. Models of blowfly reproductive success: with temperature as a non-linear variable: AICc = 1541, Akaike weight = 0.99 v. with temperature as a linear variable: AICc = 1550.2, Akaike weight = 0.01).

The model analysing mite reproductive success included data from the treatments with 10 and 20 mites and included carcass mass and temperature as covariates. In all three models, experimental site and year were included as random factors.

Laboratory experiments

We analysed the reproductive success of beetles, blowflies, and mites using GLMMs with a negative binomial distribution to account for data overdispersion. We also included block as a random factor. Post-hoc pairwise comparisons were performed using the package lsmeans (Lenth, 2016) if an interaction was detected; p value for post-hoc comparisons were adjusted using Tukey’s honestly significant difference (HSD) method. The data from the field experiment revealed a non-linear relationship between temperature and measures of reproductive success (see Figure 1). Therefore, we conservatively analysed the effect of the three different temperature (11, 15, 19°C) by treating temperature as a categorical factor in all these models.

Analyses of beetle reproductive success

We tested for the interacting effects of blowfly (yes/no), mite (0, 10, 20), and temperature (11, 15, 19°C) treatments on the reproductive success of beetles by including all three treatments as categorical factors. Separate GLMMs were used to make further comparisons between blowfly and mite treatments to determine how any significant interactions arose.

Analyses of blowfly reproductive success

We tested for the interacting effects of mites (0, 10, 20) and temperature (11, 15, 19°C) treatments on the reproductive success of blowflies, and again by including them as categorical factors.

Analyses of mite reproductive success

We tested for the interacting effects of blowfly (yes/no), mite (0, 10, 20) and temperature (11, 15, 19°C) treatments on the reproductive success of beetles. All three were included as categorical factors.

Effect of temperature on blowfly larval development

We analysed the number of blowfly larvae in a negative binomial regression model with the function glm.nb in the MASS package to account for overdispersion. We analysed carcass consumption rate in a beta regression model in the betareg package. In both analyses, we included temperature treatment (11, 15, 19°C) as a categorical factor and blowfly egg mass and carcass mass as continuous variables. To analyse the effect of temperature on the developmental rate of blowfly larvae, we used a GLMM with Gaussian error structure and included the interaction between temperature treatment and developmental stage (both as categorical factors), blowfly egg mass, and carcass mass as continuous variables. In this analysis, we also included the ID of each carcass as a random factor, since carcasses were sampled repeatedly across different developmental stages.

Effect of temperature on beetle’s carcass preparation

We analysed the roundness of carcasses in a GLM and the number of blowfly larvae in a negative binomial regression model. In both analyses, temperature treatment (11, 15, 19°C) was included as a categorical factor, whereas blowfly egg mass and carcass mass were included as continuous variables. To further investigate the effects of carcass roundness on the number of blowfly larvae that developed, we analysed the number of blowfly in a separate negative binomial regression model by additionally including roundness as a continuous variable.

Acknowledgements

We are very grateful to the Editor and three anonymous referees for their constructive and insightful comments on earlier drafts of this paper. We would also like to thank E Turner, R Mashoodh and members of the Kilner Group for comments, and S Aspinall and C Swannack for substantial logistical support. This work was funded by a Rosemary Grant Award from the Society for the Study of Evolution. S-JS was supported by the Taiwan Cambridge Scholarship from the Cambridge Commonwealth, European and International Trust. RMK was supported by a European Research Council Consolidator grant 301785 BALDWINIAN_BEETLES and a Wolfson Merit Award from the Royal Society.

Funding Statement

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Contributor Information

Syuan-Jyun Sun, Email: sjs243@ntu.edu.tw.

Dieter Ebert, University of Basel, Switzerland.

Christian Rutz, University of St Andrews, United Kingdom.

Funding Information

This paper was supported by the following grants:

Society for the Study of Evolution Rosemary Grant Award to Syuan-Jyun Sun.
Cambridge Commonwealth, European and International Trust Taiwan Cambridge Scholarship to Syuan-Jyun Sun.
European Research Council Consolidator grant 301785 BALDWINIAN_BEETLES to Rebecca M Kilner.
Royal Society Wolfson Merit Award to Rebecca M Kilner.

Additional information

Competing interests

No competing interests declared.

Author contributions

Conceptualization, Data curation, Formal analysis, Funding acquisition, Validation, Investigation, Visualization, Methodology, Writing - original draft, Project administration, Writing - review and editing.

Conceptualization, Resources, Supervision, Funding acquisition, Investigation, Methodology, Writing - original draft, Writing - review and editing.

Ethics

Animal experimentation: All of the animals were handled according to approved institutional animal care of the University of Cambridge. The protocol for field experimentation was approved by the Sub-Department of Animal Behaviour, University of Cambridge. During our experiments we handled our animals with care and they were not harmed at any stage. None of the animals that we used showed any signs of stress before, after or during the experiments.

Additional files

Supplementary file 1. Results from the final models for each variable analysed.

(a) Results from the final models for the reproductive success of beetles, blowflies, and mites in the field experiment. The final models used were: glmer.nb(Number of larvae ~ Mite treatment*(poly(temperature,degree = 2)[,2]+ poly(temperature,degree = 2)[,1])+Carcass mass+(1|site)+(1|year)). Models analyzing burying beetle larvae and blowfly larvae were both sufficient to reject the null hypotheses, with 81.3% and 98.6% power, respectively, whereas the model analyzing mite offspring was not, with a power of 36.9%. (b) Results from the final models for the reproductive success of beetles, blowflies, and mites in the Laboratory Experiment 1. For beetles, the final model used was: glmer.nb(Number of larvae ~ Mite treatment*Temperature treatment*Blowfly treatment+Carcass mass+(1|block)); for blowflies, the final model used was: glmer.nb(Number of larvae ~ Mite treatment*Temperature treatment+Carcass mass+(1|block)); and for mites, the final model used was: glmer.nb(Number of larvae ~ Blowfly treatment*Temperature treatment+Mite treatment+Carcass mass+(1|block)). All these models were sufficient to reject the null hypotheses, with the 97%, 97%, and 98.2% power, for analyses of burying beetle larvae, blowfly larvae, and mite offspring, respectively. (c) Results from the final models for the development of blowfly larvae in the Laboratory Experiment 2. For number of blowfly larvae, the final model used was: glm.nb(Number of larvae ~ Temperature treatment+Carcass mass+Blowfly egg mass); for carcass consumption rate, the final model used was: betareg(Consumption rate ~Temperature treatment+Carcass mass+Blowfly egg mass); and for development rate, the final model used was: glmer(Days ~ Temperature treatment*Developmental stage+Carcass mass+Blowfly egg mass+(1|carcass ID)). Models analyzing number of blowfly larvae and carcass consumption rate were both not sufficient to reject the null hypotheses, with 12.9% and 22.8% power, respectively, whereas the model analyzing development rate of blowfly larvae was highly sufficient, with a power of 100%. (d) Results from the final models for beetle's carcass preparation in the Laboratory Experiment 3. For number of blowfly larvae, the final model used was: glm.nb(Number of larvae ~ Temperature treatment+Carcass mass+Blowfly egg mass); and for carcass roundness, the final model used was: glm.nb(Roundness ~Temperature treatment+Carcass mass+Blowfly egg mass). Models analyzing number of blowfly larvae and carcass roundness were both sufficient to reject the null hypotheses, with 96.4% and 99.5% power, respectively.

elife-55649-supp1.docx^{(37.7KB, docx)}

Transparent reporting form

elife-55649-transrepform.docx^{(66.7KB, docx)}

Data availability

All data generated or analysed during this study are included in the manuscript and supporting files. The data has also been deposited on Dryad (Sun, Syuan-Jyun; Kilner, Rebecca (2020), A temperature-enhanced threat from a common enemy converts a parasite into a mutualist, Dryad, Dataset, https://doi.org/10.5061/dryad.sj3tx961z).

The following dataset was generated:

Syuan-Jyun S, Kilner RM. 2020. Data from: Temperature stress induces mites to help their carrion beetle hosts by eliminating rival blowflies. Dryad Digital Repository.

References

Afkhami ME, McIntyre PJ, Strauss SY. Mutualist-mediated effects on species' range limits across large geographic scales. Ecology Letters. 2014;17:1265–1273. doi: 10.1111/ele.12332. [DOI] [PubMed] [Google Scholar]
Ashby B, King KC. Friendly foes: the evolution of host protection by a parasite. Evolution Letters. 2017;1:211–221. doi: 10.1002/evl3.19. [DOI] [PMC free article] [PubMed] [Google Scholar]
Barton BT, Ives AR. Direct and indirect effects of warming on aphids, their predators, and ant mutualists. Ecology. 2014;95:1479–1484. doi: 10.1890/13-1977.1. [DOI] [PubMed] [Google Scholar]
Bates D, Maechler M, Bolker B, Walker S. R source package; 2015. https://www.jstatsoft.org/article/view/v067i01 [Google Scholar]
Beninger CW. Egg predation by Poecilochirus carabi (Mesostigmata: Parasitidae) and its Effect on Reproduction of Nicrophorus vespilloides (Coleoptera: Silphidae) Environmental Entomology. 1993;22:766–769. doi: 10.1093/ee/22.4.766. [DOI] [Google Scholar]
Bornemissza GF. An analysis of arthropod succession in Carrion and the effect of its decomposiion on the soil fauna. Australian Journal of Zoology. 1957;5:1. doi: 10.1071/ZO9570001. [DOI] [Google Scholar]
Chamberlain SA, Bronstein JL, Rudgers JA. How context dependent are species interactions? Ecology Letters. 2014;17:881–890. doi: 10.1111/ele.12279. [DOI] [PubMed] [Google Scholar]
Chen BF, Liu M, Rubenstein DR, Sun SJ, Liu JN, Lin YH, Shen SF. A chemically triggered transition from conflict to cooperation in burying beetles. Ecology Letters. 2020;23:467–475. doi: 10.1111/ele.13445. [DOI] [PubMed] [Google Scholar]
Clay K. Defensive symbiosis: a microbial perspective. Functional Ecology. 2014;28:293–298. doi: 10.1111/1365-2435.12258. [DOI] [Google Scholar]
Corbin C, Heyworth ER, Ferrari J, Hurst GD. Heritable symbionts in a world of varying temperature. Heredity. 2017;118:10–20. doi: 10.1038/hdy.2016.71. [DOI] [PMC free article] [PubMed] [Google Scholar]
De Gasperin O, Duarte A, Troscianko J, Kilner RM. Fitness costs associated with building and maintaining the burying beetle's carrion nest. Scientific Reports. 2016;6:35293. doi: 10.1038/srep35293. [DOI] [PMC free article] [PubMed] [Google Scholar]
De Gasperin O, Kilner RM. Friend or foe: inter-specific interactions and conflicts of interest within the family. Ecological Entomology. 2015;40:787–795. doi: 10.1111/een.12259. [DOI] [PMC free article] [PubMed] [Google Scholar]
De Gasperin O, Kilner RM. Interspecific interactions and the scope for Parent-Offspring conflict: high mite density temporarily changes the Trade-Off between offspring size and number in the burying beetle, Nicrophorus vespilloides. PLOS ONE. 2016;11:e0150969. doi: 10.1371/journal.pone.0150969. [DOI] [PMC free article] [PubMed] [Google Scholar]
Doremus MR, Oliver KM. Aphid heritable symbiont exploits defensive mutualism. Applied and Environmental Microbiology. 2017;83:e03276-16. doi: 10.1128/AEM.03276-16. [DOI] [PMC free article] [PubMed] [Google Scholar]
Duarte A, Welch M, Swannack C, Wagner J, Kilner RM. Strategies for managing rival bacterial communities: lessons from burying beetles. Journal of Animal Ecology. 2018;87:414–427. doi: 10.1111/1365-2656.12725. [DOI] [PMC free article] [PubMed] [Google Scholar]
Early R, Keith SA. Geographically variable biotic interactions and implications for species ranges. Global Ecology and Biogeography. 2019;28:42–53. doi: 10.1111/geb.12861. [DOI] [Google Scholar]
Engl T, Eberl N, Gorse C, Krüger T, Schmidt THP, Plarre R, Adler C, Kaltenpoth M. Ancient symbiosis confers desiccation resistance to stored grain pest beetles. Molecular Ecology. 2018;27:2095–2108. doi: 10.1111/mec.14418. [DOI] [PubMed] [Google Scholar]
Ewald PW. Transmission modes and evolution of the parasitism-mutualism continuum. Annals of the New York Academy of Sciences. 1987;503:295–306. doi: 10.1111/j.1749-6632.1987.tb40616.x. [DOI] [PubMed] [Google Scholar]
Fatouros NE, Huigens ME. Phoresy in the field: natural occurrence of Trichogramma egg parasitoids on butterflies and moths. BioControl. 2012;57:493–502. doi: 10.1007/s10526-011-9427-x. [DOI] [Google Scholar]
Fellous S, Salvaudon L. How can your parasites become your allies? Trends in Parasitology. 2009;25:62–66. doi: 10.1016/j.pt.2008.11.010. [DOI] [PubMed] [Google Scholar]
Fitzpatrick G, Lanan MC, Bronstein JL. Thermal tolerance affects mutualist attendance in an ant-plant protection mutualism. Oecologia. 2014;176:129–138. doi: 10.1007/s00442-014-3005-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
Gelman A, Hill J. Data Analysis Using Regression and Multilevel/Hierarchical Models, Data Analysis Using Regression and Multilevel/hierarchical Models. Cambridge University Press; 2007. [Google Scholar]
Goldberg E, Kirby K, Hall J, Latham J. The ancient woodland concept as a practical conservation tool in great britain. Journal for Nature Conservation. 2007;15:109–119. doi: 10.1016/j.jnc.2007.04.001. [DOI] [Google Scholar]
Gorter FA, Scanlan PD, Buckling A. Adaptation to abiotic conditions drives local adaptation in Bacteria and viruses coevolving in heterogeneous environments. Biology Letters. 2016;12:20150879. doi: 10.1098/rsbl.2015.0879. [DOI] [PMC free article] [PubMed] [Google Scholar]
Green P, MacLeod CJ. SIMR : an R package for power analysis of generalized linear mixed models by simulation. Methods in Ecology and Evolution. 2016;7:493–498. doi: 10.1111/2041-210X.12504. [DOI] [Google Scholar]
Hoang KL, Gerardo NM, Morran LT. The effects of Bacillus subtilis on Caenorhabditis elegans fitness after heat stress. Ecology and Evolution. 2019;9:3491–3499. doi: 10.1002/ece3.4983. [DOI] [PMC free article] [PubMed] [Google Scholar]
Hoeksema JD, Bruna EM. Context-dependent outcomes of mutualistic interactions. In: Bronstein JL, editor. Mutualism. Oxford University Press; 2015. pp. 181–202. [Google Scholar]
Hopkins SR, Wojdak JM, Belden LK. Defensive symbionts mediate Host-Parasite interactions at multiple scales. Trends in Parasitology. 2017;33:53–64. doi: 10.1016/j.pt.2016.10.003. [DOI] [PubMed] [Google Scholar]
Jaenike J, Unckless R, Cockburn SN, Boelio LM, Perlman SJ. Adaptation via symbiosis: recent spread of a Drosophila defensive symbiont. Science. 2010;329:212–215. doi: 10.1126/science.1188235. [DOI] [PubMed] [Google Scholar]
Johnson NC. Mutualisms and ecosystem-level processes. In: Bronstein JL, editor. Mutualism. Oxford University Press; 2015. pp. 221–238. [Google Scholar]
Lenth RV. Least-squares means: the R package lsmeans. Journal of Statistical Software. 2016;69:i01. doi: 10.18637/jss.v069.i01. [DOI] [Google Scholar]
Lively CM, Clay K, Wade MJ, Fuqua C. Competitive co-existence of vertically and horizontally transmitted parasites. Evolutionary Ecology Research. 2005;7:1183–1190. [Google Scholar]
Matuszewski S, Bajerlein D, Konwerski S, Szpila K. Insect succession and carrion decomposition in selected forests of central Europe part 2: composition and residency patterns of carrion fauna. Forensic Science International. 2010;195:42–51. doi: 10.1016/j.forsciint.2009.11.007. [DOI] [PubMed] [Google Scholar]
Nehring V, Müller JK, Steinmetz N. Phoretic Poecilochirus mites specialize on their burying beetle hosts. Ecology and Evolution. 2017;7:10743–10751. doi: 10.1002/ece3.3591. [DOI] [PMC free article] [PubMed] [Google Scholar]
Okabe K, Shun'ichi M. Parasitic mites as part-time bodyguards of a host wasp. Proceedings of the Royal Society B: Biological Sciences. 2008;275:2293–2297. doi: 10.1098/rspb.2008.0586. [DOI] [PMC free article] [PubMed] [Google Scholar]
Palmer TM, Stanton ML, Young TP, Goheen JR, Pringle RM, Karban R. Breakdown of an ant-plant mutualism follows the loss of large herbivores from an african savanna. Science. 2008;319:192–195. doi: 10.1126/science.1151579. [DOI] [PubMed] [Google Scholar]
Palmer T, Pringle E, Stier A, Holt R. Mutualism in a Community contextMutualism. Oxford University Press; 2015. [Google Scholar]
Payne JA. A summer carrion study of the baby pig Sus scrofa linnaeus. Ecology. 1965;46:592–602. doi: 10.2307/1934999. [DOI] [Google Scholar]
Peay KG. The mutualistic niche: mycorrhizal symbiosis and community dynamics. Annual Review of Ecology, Evolution, and Systematics. 2016;47:143–164. doi: 10.1146/annurev-ecolsys-121415-032100. [DOI] [Google Scholar]
Rafaluk-Mohr C, Ashby B, Dahan DA, King KC. Mutual fitness benefits arise during coevolution in a nematode-defensive microbe model. Evolution Letters. 2018;2:246–256. doi: 10.1002/evl3.58. [DOI] [PMC free article] [PubMed] [Google Scholar]
Schwarz HH, Starrach M, Koulianos S. Host specificity and permanence of associations between mesostigmatic mites (Acari: Anactinotrichida) and burying beetles (Coleoptera: Silphidae: Nicrophorus ) Journal of Natural History. 1998;32:159–172. doi: 10.1080/00222939800770101. [DOI] [Google Scholar]
Schwarz HH, Müller JK. The dispersal behaviour of the phoretic mite Poecilochirus carabi (Mesostigmata, Parasitidae): adaptation to the breeding biology of its carrier Necrophorus vespilloides (Coleoptera, Silphidae) Oecologia. 1992;89:487–493. doi: 10.1007/BF00317154. [DOI] [PubMed] [Google Scholar]
Scott MP. Competition with flies promotes communal breeding in the burying beetle, Nicrophorus tomentosus. Behavioral Ecology and Sociobiology. 1994;34:367–373. doi: 10.1007/BF00197007. [DOI] [Google Scholar]
Scott MP. The ecology and behavior of burying beetles. Annual Review of Entomology. 1998;43:595–618. doi: 10.1146/annurev.ento.43.1.595. [DOI] [PubMed] [Google Scholar]
Shelomi M, Matern LM, Dinstell JM, Harris DW, Kimsey RB. DEET (N,N-diethyl-meta-toluamide) induced delay of blowfly landing and oviposition rates on treated pig carrion (Sus scrofa L.) Journal of Forensic Sciences. 2012;57:1507–1511. doi: 10.1111/j.1556-4029.2012.02159.x. [DOI] [PubMed] [Google Scholar]
Smiseth PT, Darwell CT, Moore AJ. Partial begging: an empirical model for the early evolution of offspring signalling. Proceedings of the Royal Society of London. Series B: Biological Sciences. 2003;270:1773–1777. doi: 10.1098/rspb.2003.2444. [DOI] [PMC free article] [PubMed] [Google Scholar]
Springett BP. Aspects of the relationship between burying beetles, Necrophorus spp and the mite, Poecilochirus necrophori vitz. The Journal of Animal Ecology. 1968;37:417. doi: 10.2307/2957. [DOI] [Google Scholar]
Sun SJ, Rubenstein DR, Chen BF, Chan SF, Liu JN, Liu M, Hwang W, Yang PS, Shen SF. Climate-mediated cooperation promotes niche expansion in burying beetles. eLife. 2014;3:e02440. doi: 10.7554/eLife.02440. [DOI] [PMC free article] [PubMed] [Google Scholar]
Sun SJ, Horrocks NPC, Kilner RM. Conflict within species determines the value of a mutualism between species. Evolution Letters. 2019;3:185–197. doi: 10.1002/evl3.109. [DOI] [PMC free article] [PubMed] [Google Scholar]
Sun S-J, Kilner RM. Cryptic host specialisation within Poecilochirus carabi mites explains population differences in the extent of co-adaptation with their burying beetle Nicrophorus vespilloides hosts. bioRxiv. 2019 doi: 10.1101/641936. [DOI] [PMC free article] [PubMed]
Wall R, French N, Morgan KL. Effects of temperature on the development and abundance of the sheep blowfly Lucilia sericata (Diptera: Calliphoridae) Bulletin of Entomological Research. 1992;82:125–131. doi: 10.1017/S0007485300051531. [DOI] [Google Scholar]
Wang Y, Rozen DE. Fitness costs of phoretic Nematodes in the burying beetle, Nicrophorus vespilloides. Ecology and Evolution. 2019;9:26–35. doi: 10.1002/ece3.4570. [DOI] [PMC free article] [PubMed] [Google Scholar]
White PS, Morran L, de Roode J. Phoresy. Current Biology. 2017;27:R578–R580. doi: 10.1016/j.cub.2017.03.073. [DOI] [PMC free article] [PubMed] [Google Scholar]
Wilson DS. The effect of population structure on the evolution of mutualism: a field test involving burying beetles and their phoretic mites. The American Naturalist. 1983;121:851–870. doi: 10.1086/284108. [DOI] [Google Scholar]
Wilson DS, Knollenberg WG. Adaptive indirect effects: the fitness of burying beetles with and without their phoretic mites. Evolutionary Ecology. 1987;1:139–159. doi: 10.1007/BF02067397. [DOI] [Google Scholar]

eLife. doi: 10.7554/eLife.55649.sa1

Decision letter

Editor: Dieter Ebert¹

In the interests of transparency, eLife publishes the most substantive revision requests and the accompanying author responses.

Acceptance summary:

This work on the interactions of burying beetles, mites and blowflies with temperature is a nice example of how multi-trophic relationships can change in response to an environmental factor and turn the sign of interactions around (from negative to positive). The combination of field work and experimental confirmation is strongly supporting the authors conclusions. Most interesting is the observation that at extreme temperatures the mites enhance beetle breeding success by eating blowfly eggs, thus turning the negative effect of mites in the absence of blowflies, into a positive effect.

Decision letter after peer review:

Thank you for submitting your article "A temperature-enhanced threat from a common enemy converts a parasite into a mutualist" for consideration by eLife. Your article has been reviewed by three peer reviewers, one of whom is a member of our Board of Reviewing Editors, and the evaluation has been overseen by Christian Rutz as the Senior Editor. The reviewers have opted to remain anonymous.

The reviewers have discussed their reviews with one another, and the Reviewing Editor has drafted this decision letter to help you prepare a revised submission.

We would like to draw your attention to changes in our revision policy that we have made in response to COVID-19 (https://elifesciences.org/articles/57162). Specifically, we are asking editors to accept without delay manuscripts, like yours, that they judge can stand as eLife papers without additional data, even if they feel that they would make the manuscript stronger. Thus the revisions requested below only address clarity and presentation.

This paper studies a complex trophic web between carrion beetles, phoretic mites and blowflies. The mites and beetles share carrion located by the beetles. The third carrion feeder, blowflies, compete with the other two carrion consumers. Both beetles and mites eat fly larvae as well as carrion. Flies are more common when it is warm and do better because they spend time in less vulnerable stages. They do better when it is cold because beetles are less able to roll the carcass into a ball to interfere with blowfly success. Beetles compete poorly with flies when it is too cold or too warm. Mites can eat blowfly eggs, thereby reducing the beetle's main competitor. The authors further consider if the net effect of mites on beetles is affected by flies and whether this net effect changes with temperature. It does, and the mechanism for this environmental interaction is mostly well explained.

The three reviewers have raised a variety of issues concerning the presentation of the study. Here is a short summary, and further details are provided in the three separate reports appended below.

1) Terminology. The general presentation of the work needs some revision. The reviewers made suggestions regarding the title and various terminological issues.

2) Framework. The placement of the work in a parasitism-mutualism framework was strongly criticised. The study would not lose any appeal if it was instead placed in a more conventional ecological framework but would gain in clarity.

3) Statistical issues. The Akaike Information Criterion is central for distinguishing between different statistical models. It is mentioned in the text but apparently not applied.

4) Literature cited. In connection with point 2 (above), it will be necessary to cover relevant literature when placing the work in a food-web theoretical framework.

These issues, including those listed below, should be taken care of in a revised version.

Reviewer #1:

This manuscript describes the breeding success of burying beetles in the presence and absence of two competing species, mites and blow flies. A combination of field and laboratory experiments shows that in the absence of blow flies, mites are harmful, while in the presence of blow flies, mites are-at extreme temperatures-beneficial to the beetles. The study is placed in a framework of parasitism – mutualism continuum, with the take home message that temperature can shift the form of the interaction from parasitism to mutualism.

I like the topic, the study and the way the data where produced. I have however problems with some of the approaches taken by the authors to analyse and present their data.

1) The title starts with "A changing climate…". This is far-fetched. The term climate shows only once up in the entire text (at the very end of the Discussion section). This title seems more aiming to catch attention than to give an adequate title for the study. I find this misleading and suggest changing this.

2) The parasitism – mutualism framework is a far fetch. The authors justify their claim by citing a dictionary entry (supplement) about the term parasitism. Even so this entry is ok, it is not how it is perceived in the general community and I think therefore it is misleading. I don't understand why it has to be parasitism. The message of the study remains the same if one states that species interactions change sign (from positive to negative) when the environmental conditions change. The main interaction is that mites compete with the beetles for food, which is competition. Adding the blow flies to the picture turns the mites into predators of blow fly eggs. This is helpful for the host. So, it is a change from a competitor to a mutualist.

3) Statistics: The authors show the models in the legends to the tables. But this is not complete. What are the different items in the model? Which of the items are factors, which are continuous variables? In the text some info is given, but is not very clear. What is a "categorical covariates". Is this a factor (like presence/absence of blow flies) or is this a continuous covariable (like temperature and carcass mass)? Please clarify. Temperature should be used throughout as a contiguous covariable, not as a factor. Furthermore, it would be helpful if the error distribution is given for each model. Given that the data have rather special error distributions (see figures) this is very important.

4) The authors state that they used "Akaike Information Criterion (AIC) by evaluating AICc, δ AICc and Akaike weights". I did not find any evidence in the statistics that this was actually done. This is important in particular for the fitting of the non-linear relationships.

Reviewer #2:

This study is very timely and unique, as it utilizes a combination of both field and lab work to examine protective mutualism using eukaryotes. A few comments:

1) "Transition", which is used throughout the manuscript can refer to evolutionary history/trajectory or an ecological-time, context-dependent switch. Given the combination of references used herein on evolutionary and ecological transitions, we strongly recommend clarifying throughout and choosing references/wording more carefully. There are numerous references to condition-dependent transitions along the parasite-mutualist continuum across the tree of life which are not mentioned.

2) There was also a lack of discussion of protective symbioses against abiotic stresses (like temperature) in the Introduction, which are different than the biotic stresses.

Here are some references on protective microbes against abiotic stresses:

Corbin.et al., 2017 - This reference was cited, but not sure the point of this paper was how it was used in the manuscript (–Introduction, "Yet theoretical analyses that consider how such interactions evolve and persist derive mainly from recent interest, with a particular focus on the microbial endosymbionts that can be induced to defend their hosts from attack"). This is a good source for examples of symbionts that protect hosts from temperature stresses and how temperature affects the association, not about biotic factors, which is what the sentence in the manuscript implies.

Engl et al., 2018.

Feldhaar, 2011.

Hoang, Gerardo and Morran, 2019.

3) Introduction is a bit sparse. For example, before the last paragraph, a brief sentence or two to summarize the big picture and tie to what is known to what is not known. Moreover, there was no mention of mutualist/symbiont density or mite density in the Introduction, even though that was what was driving the results

4) A diagram could be used to show how each player interacts with one another, especially because there are several definitions being used and different environmental contexts involved. For example, in the Introduction: being a mutualist is not mutually exclusive of being predatory. Mites are mutualistic towards the beetles and predatory towards the blowflies. A diagram can represent this. Moreover, doesn't the definition of parasitism have to be in a supplement (seems like an odd place) as opposed to just mentioned when the system is introduced?

5) Another point to mention in the Discussion section would be how the association becomes more mutualistic when there are more mites: the results indicate that more mites increase beetle fitness, which in turn benefit the mites (at the least, in order for beetle fitness to increase there has to be more mites. I'm not saying that the beetles directly promote mite fitness, but there is an association there. Also, perhaps more beetles could mean more mites can hitchhike?).

6) How common is it for phoresy to be parasitic? Worth mentioning how your consideration of the costs of mites squares with phoresy as currently understood in the literature.

7) Supplementary file 1, Supplementary file 2, Supplementary file 3 and Supplementary file 4, but particularly Supplementary file 3. Given the size of these models, do the authors have enough power to detect significant differences?

Reviewer #3:

This paper studies a complex trophic web between carrion beetles and phoretic mites. The mites and beetles share carrion located by the beetle, third carrion feeder, blowflies, competes with the other two carrion consumers. Both beetles and mites eat fly larvae as well as carrion. Flies are more common when it is warm and do better because they spend time in less vulnerable stages. They do better when it is cold because beetles are less able to roll the carcass into a ball to interfere with blowfly success. Beetles compete poorly with flies when it is too cold or too warm. Mites can eat blowfly eggs, thereby reducing the beetle's main competitor. The authors further consider if the net effect of mites on beetles is affected by flies and whether this net effect changes with temperature. It does, and the mechanism for this environmental interaction is well explained.

The paper is relatively well written and the authors are to be commended for their use of manipulative field and laboratory experiments. I enjoyed reading the paper and learned a lot. I do, however, have several suggestions that I think will improve the paper.

The title is vague. Authors sometimes assume that this will help the paper appear general and therefore capture broader readership. It doesn't. At least I would not bother to read it based on the title. And I think you want people like me to read it. A good title should state the key results and identify the system. I would read a paper titled something like: Phoretic mites help their carrion beetle hosts under temperature stress by eating competing blowfly larvae.

A general point about the Introduction: the paper will flow better (and read more like a scientific paper) if the authors clearly state the series of predictions they are going to test an how these predictions flow from their hypotheses. Each P-value in the result should be an evaluation of either an apriori prediction or the assessment of an assumption, or a controlling variable. And it should be clear how the hypotheses derive from the literature, theory or field observations.

The paper is missing some key information on what happens in nature. In particular, what is the natural distribution of mites on beetles in the field (this will help readers understand whether the treatments in the experiments are reasonable).

This gets me to my point about whether the mite is a parasite or a mutualist or a whatever. This might be semantics, but I put a high bar on assigning parasitism. And the beetle-mite interaction does not cross that bar for me. Namely, the negative trophic interaction does not appear to occur during intimacy. Parasitism must be defined at the stage level rather than the species level. I also put a particularly high bar for deciding that parasitism is context dependent when simpler explanations via indirect effects are possible. To me it is clear that the phoretic mite stage is not a parasite because it does not feed on the beetle and the free-living mite stage is not a parasite because it neither feeds on the beetle or is intimate with the beetle. This does not mean the interaction is not interesting. In fact, I think it is far more interesting than the way the authors have packaged it. The mite is a phoretic commensal that uses its host to be transported to a common food source. Once on that common food source, the relationship can become competitive. But that competition can lead to a net indirect benefit under situations where another competitor (blowflies) might dominate. This interpretation is better couched in food-web theory rather than evolutionary theory.

As an example, here is a paper I remembered from decades ago that pondered over the relationship between mites and hosts and third-party interactions. I include the citation for the authors' interest. This author had similar observations but did not conclude a plastic parasitic-mutualistic relationship.

Rigby, (1996).

[Editors' note: further revisions were suggested prior to acceptance, as described below.]

Thank you for submitting your article "Temperature stress induces mites to help their carrion beetle hosts by eliminating rival blowflies" for consideration by eLife. Your article has been reviewed again by a Senior Editor, a Reviewing Editor, and two reviewers.

Following a consultation discussion, the Reviewing Editor has drafted this decision letter to help you prepare a revised submission.

The editors have judged that your manuscript is of interest, but as described below, feel that additional revision is required before it can be published. We are also offering, if you choose, to post your manuscript to bioRxiv (if it is not already there) along with this decision letter and a formal designation that the manuscript is "in revision at eLife". Please let us know if you would like to pursue this option.

The reviewers were mostly satisfied with your responses to their comments from the first round of review. However, there are still points that require further attention. In particular, both reviewers remained unconvinced by your assertion that this is a host-parasite system. We suggest the following course of action for addressing this concern: We feel there is no need to stress the host-parasite framework in the abstract, introduction, and result section – the work is very interesting without this. If you wish, you can add a short paragraph to the Discussion section that outlines the reasons why you think this system can also be viewed productively within a host-parasite framework. In this section, if you choose to include it, please contrast your interpretation with alternative views, as presented by the reviewers – that way, readers can hear both sides of the argument.

Further comments from reviewers #1 and #3 follow.

Reviewer #1:

The revised manuscript fixed many of the issues I had with the earlier version. The authors did a nice job. There are still two points I feel very uncomfortable with.

1) The work is still largely placed in a host – parasite framework. While I can follow the reasoning for doing so (positive – negative interactions), I think it is misleading and distorts the picture. It needs a lot of explanation to understand why this system might be explained by a host – parasite metaphor. Reducing this to positive – negative interactions is rather simple and not informative. Furthermore, host – parasite interactions and mutualism are concepts invented to describe pairwise interactions. As soon as three or more players act together these concepts do not work anymore (see the debate on this in the host – microbiota literature). While I see that other worker in the field have also used the host-parasite framework, it is not commonly done so.

2) The treatment of temperature in the experiment is odd. Temperature is clearly a continuous variable, which in the experiment is broken into 3 categories. But the idea behind it is not an idea of categories, but rather of a continuum. This it is how it is presented in the interpretation of the study.

Reviewer #3:

The authors have done a nice job with the revision. We had a key disagreement, which they supported admirably, but I try to provide a bit more explanation so that this concern can be distinguished from picky semantics. The mite is not a parasite.

All deference to Tara Stewart's paper aside, figure 1 supplement does not solve my concern. I still don't agree that this study can be easily summarized as a host parasite relationship. I say this as someone that has explicitly defined such relationships in papers (see Lafferty and Kuris, 2002) and am weirdly troubled when other authors have not done so carefully. So, unfortunately, I have a strong opinion about this. And although the semantics might not be particularly important to most people, let me explain why they are to me in this particular case. It is a fascinating question how species interactions can change from positive to negative based on environmental conditions. And one often reads the assumption that it is the case that parasitism often changes over to mutualism based on the environmental context. The first sentence of your abstract does so with force (Ecological transitions between parasitism and mutualism are relatively commonplace). However, when reading carefully, these examples are usually not describing parasites, by which I mean consumers that have an intimate, non-lethal, dependent long-term feeding relationship on a single host individual during a particular parasitic life stage. When defined explicitly as such, transitions to mutualism is not a general rule about the nature of parasitism. It is far more common in other types of species associations like the one described here. When we refer to other parasite-like interactions, we usually take care to preface this as brood parasitism, or parasitoidism, or epiphyte, etc. By which I mean, describing this interaction as parasitism muddies our view about parasitism rather than clarifies our understanding of species interactions. The solution is to simply not rely on parasite as a shorthand to describe this particular species interaction. Rather, just describe how the species interact. It is quite interesting on its own, and calling it parasitism does not do it service. Below is an example of the edited Abstract that accomplishes this.

"Ecological transitions in and out of mutualism are relatively commonplace though the causal agents driving such change remain poorly understood. Here we show that temperature stress modulates the harm threatened by a common enemy, and thereby induces a phoretic mite to become a protective mutualist. Our experiments focus on the interactions between the burying beetle Nicrophorus vespilloides, an associated mite species Poecilochirus carabi and their common enemy, blowflies, when all three species reproduce on the same small vertebrate carrion. We show that mites compete with burying beetles for food in the absence of blowflies, because they reduce beetle reproductive success. However, when blowflies breed on the carrion too, mites enhance beetle reproductive success by eating blowfly eggs. High densities of mites are especially effective at promoting beetle reproductive success at higher and lower natural ranges in temperature, when blowfly larvae are more potent rivals for the limited resources on the carcass."

Similar care in terminology make it easy to get rid of parasite and specify the actual biology of the system.

Lafferty and Kuris, (2002).

[Editors' note: further revisions were suggested prior to acceptance, as described below.]

Thank you for submitting your article "Temperature stress induces mites to help their carrion beetle hosts by eliminating rival blowflies" for consideration by eLife.

Your revised article has been evaluated by a Reviewing Editor in consultation with a statistical adviser, and we would like to ask you to make some additional revisions before final acceptance. For details, please see below.

The Reviewing Editor commented on your response to a point that had been raised during the previous round of review, namely the treatment of temperature in statistical analyses. This is a lightly edited version of their feedback:

There are three types of independent variables: continuous variables (you know the order and the distance between the values: e.g., 1, 2, 3, 4…); ordered variables (you know the order (e.g., A > B > C), but you may not know the distance between the values); and factors (where the order does not play any role).

Temperature is continuous. The difference between 10C and 15C is the same as the difference between 15C and 20C. The authors prefer not to treat it as continuous, because they argue that, if you have treatments of different temperatures, you cannot know what happens in between these temperatures. This is right in a world where even very small temperature differences make big differences in the outcome. I don't think this is likely here, but this is not worth arguing about.

The next option would be to use ordered values for the temperature treatments (12{degree sign}C < 16{degree sign}C < 20{degree sign}C, disregarding how big the differences are). This allows you to say that 16{degree sign}C is in between 12{degree sign}C and 20{degree sign}C, and that 12{degree sign}C and 20{degree sign}C are low and high, respectively. This is what the authors use when they talk about "low", "intermediate" and "high" temperatures. Using these attributes makes only sense if the treatments are at least ordered, as otherwise you use a statistical model assuming no order, but construct arguments based on an order.

Why does this matter?

Statistically: Using temperature as a factor means that the variation of the independent variable is totally unconstrained. Treatments can vary in any way from each other. Using temperature as an ordered or continuous variable imposes a constraint on the way how the variation of the independent variable is structured. This reduces the power of the statistical test, but gives you more information as you understand more about the structure of the variance.

Biologically: A factor does not assume that the different treatments have anything to do with each other. An ordered or continuous variable assumes that treatments are not independent. This is what we can use in the argumentation. For example, when we talk about "intermediate" temperature, we can make a biological statement in relation to the more extreme treatments. We gain this freedom by a penalty in the statistical power.

The authors also argue that having just three values is not enough for a regression, which is incorrect. There are many independent replicates for each of the three temperatures, so it is not three but many. It is in fact possible to do a regression with just two temperatures (but multiple measurements per temperature). The aim of a regression is to estimate the change in the dependent variable in relation to the independent variable. This is exactly what was aimed for here.

Following further discussion, we decided to consult a statistical adviser on this matter. This is a lightly edited version of their feedback:

It is of course possible to treat the temperature variable as continuous even though the experiment has constrained it to be one of three specific values. The only reason that would be justified for treating it as a factor would be because they suspect there is a non-linear relationship between temperature and the outcome. This appears to be the case in some of their field experiments. So, while some of their defence is incorrect, their point about suspecting a U-shaped or non-linear relationship is justification for treating temperature as a categorical factor. If they have specific hypotheses (after the field observations and before the lab experiments) about low and high temperatures that would also make sense to treat as a factor.

We would like to ask you to address these considerations in your final revisions, which we feel can be achieved by clarifying the rationale of your approach to statistical analyses, and by using more nuanced wording for describing the findings from the experiments with three temperature treatments.

eLife. 2020 Aug 5;9:e55649. doi: 10.7554/eLife.55649.sa2

Author response

The three reviewers have raised a variety of issues concerning the presentation of the study. Here is a short summary, and further details are provided in the three separate reports appended below.

1) Terminology. The general presentation of the work needs some revision. The reviewers made suggestions regarding the title and various terminological issues.

We have amended the terminology where it caused problems, throughout.

2) Framework. The placement of the work in a parasitism-mutualism framework was strongly criticised. The study would not lose any appeal if it was instead placed in a more conventional ecological framework but would gain in clarity.

We have now explained in great detail why we chose to use a parasitism-mutualism framework to analyse this study system. We have also incorporated the excellent ecological concepts raised by the reviewers into this way of thinking. In our view this strengthens the study by retaining the emphasis on fitness outcomes (required by the parasitism-mutualism framework) and blending it with a strong ecological perspective that explains how these interactions fit into a wider web of direct and indirect interactions.

3) Statistical issues. The Akaike Information Criterion is central for distinguishing between different statistical models. It is mentioned in the text but apparently not applied.

We have provided these details in the text.

4) Literature cited. In connection with point 2 (above), it will be necessary to cover relevant literature when placing the work in a food-web theoretical framework.

We have expanded our citation of the literature.

These issues, including those listed below, should be taken care of in a revised version.

Reviewer #1:

1) The title starts with "A changing climate…". This is far-fetched. The term climate shows only once up in the entire text (at the very end of the Discussion section). This title seems more aiming to catch attention than to give an adequate title for the study. I find this misleading and suggest changing this.

We agree and we have rewritten the impact statement (we assume this is what the reviewer is referring to here, although we have also rewritten the title).

2) The parasitism – mutualism framework is a far fetch. The authors justify their claim by citing a dictionary entry (supplement) about the term parasitism. Even so this entry is ok, it is not how it is perceived in the general community and I think therefore it is misleading. I don't understand why it has to be parasitism. The message of the study remains the same if one states that species interactions change sign (from positive to negative) when the environmental conditions change. The main interaction is that mites compete with the beetles for food, which is competition. Adding the blow flies to the picture turns the mites into predators of blow fly eggs. This is helpful for the host. So, it is a change from a competitor to a mutualist.

Thank you for your thoughts on this. As you have realised, it is not easy to define the interactions between burying beetles and their mites. Nevertheless, we have given this considerable thought. Furthermore, we are not the first to describe these mites as parasitic. However, since the terminology we have used has caused problems with more than one reviewer we decided to go back to the literature to more clearly explain the reasoning of our thinking.

Stewart and Schnitzer, (2017) set out the key differences between competition and parasitism, specifically focusing on interactions where the difference between them has become somewhat blurred. We follow their framework for defining mites as parasites rather than competitors and explain that using their conceptual roadmap in Figure 1—figure supplement 1.

The main reason for defining mites as parasites and not competitors lies in the effect that the interaction has on the fitness of mites versus the fitness of burying beetles: parasitism is a positive-negative interaction, whereas competition is a negative-negative interaction.

Mite fitness is positively affected by beetles because mites depend solely on beetles to be transported to their breeding resource (Schwarz and Müller, 1992), and because they then produce more offspring in the presence of beetles than in the absence of beetles (Sun and Kilner, 2019). Burying beetles, on the other hand, lose fitness to mites because mites compete with beetle adults and larvae for carrion resource (Nehring et al., 2017), and because mites can predate directly upon beetle eggs and larvae (Beninger, 1993; De Gasperin and Kilner, 2015). Therefore, mites are parasites of burying beetles.

We have now explained this thinking in the paper (Results section). It is also evident in Figure 1—figure supplement 1, which was also based on Stewart and Schnitzer, (2017) definitions.

3) Statistics: The authors show the models in the legends to the tables. But this is not complete. What are the different items in the model? Which of the items are factors, which are continuous variables? In the text some info is given, but is not very clear. What is a "categorical covariates". Is this a factor (like presence/absence of blow flies) or is this a continuous covariable (like temperature and carcass mass)? Please clarify. Temperature should be used throughout as a contiguous covariable, not as a factor. Furthermore, it would be helpful if the error distribution is given for each model. Given that the data have rather special error distributions (see figures) this is very important.

Thank you for this suggestion. We agree that our use of categorical covariates is misleading. We have now clarified variables in all statistical analyses sections, either as categorical factors or as continuous variables.

We included temperature firstly as a continuous variable for field experiment because we measured continuous natural variation in temperature. However, for laboratory experiments, we manipulated temperature by establishing three temperature treatment categories (i.e. 11, 15, and 19°C). This is why temperature is a categorical factor in the analyses of laboratory-collected data.

We describe the error distribution we used with these models (negative binomial distributions).

4) The authors state that they used "Akaike Information Criterion (AIC) by evaluating AICc, δ AICc and Akaike weights". I did not find any evidence in the statistics that this was actually done. This is important in particular for the fitting of the non-linear relationships.

We have now incorporated this information.

Reviewer #2:

This study is very timely and unique, as it utilizes a combination of both field and lab work to examine protective mutualism using eukaryotes. A few comments:

1) "Transition", which is used throughout the manuscript can refer to evolutionary history/trajectory or an ecological-time, context-dependent switch. Given the combination of references used herein on evolutionary and ecological transitions, we strongly recommend clarifying throughout and choosing references/wording more carefully. There are numerous references to condition-dependent transitions along the parasite-mutualist continuum across the tree of life which are not mentioned.

Thank you – this is a truly helpful suggestion. We have now made it clear throughout that transition in this study specifically refers to ecological transition throughout the manuscript.

2) There was also a lack of discussion of protective symbioses against abiotic stresses (like temperature) in the Introduction, which are different than the biotic stresses.

Here are some references on protective microbes against abiotic stresses:

Corbin et al., 2017 This reference was cited, but not sure the point of this paper was how it was used in the manuscript (–Introduction, "Yet theoretical analyses that consider how such interactions evolve and persist derive mainly from recent interest, with a particular focus on the microbial endosymbionts that can be induced to defend their hosts from attack"). This is a good source for examples of symbionts that protect hosts from temperature stresses and how temperature affects the association, not about biotic factors, which is what the sentence in the manuscript implies.

Engl et al., 2018.

Feldhaar, 2011.

Hoang, Gerardo and Morran, 2019.

We have made the distinction between abiotic and biotic factors clearer throughout and also cited Engl et al., 2018 and Hoang et al., 2019 (Introduction).

3) Introduction is a bit sparse. For example, before the last paragraph, a brief sentence or two to summarize the big picture and tie to what is known to what is not known. Moreover, there was no mention of mutualist/symbiont density or mite density in the Introduction, even though that was what was driving the results

We agree with you that more details are needed to briefly summarise the bigger picture etc. We have re-written the Introduction to make that clearer.

We have also included a paragraph in the Discussion section to highlight how mite densities are linked to the transition from parasitism to mutualism.

4) A diagram could be used to show how each player interacts with one another, especially because there are several definitions being used and different environmental contexts involved. For example, in the Introduction: being a mutualist is not mutually exclusive of being predatory. Mites are mutualistic towards the beetles and predatory towards the blowflies. A diagram can represent this. Moreover, doesn't the definition of parasitism have to be in a supplement (seems like an odd place) as opposed to just mentioned when the system is introduced?

This is another great suggestion. We have made a summary diagram of our results, following the referee’s suggestions here (Figure 5). Our use of the term ‘parasitism’ has turned out to be more controversial than we expected. We now explain the logic behind using more fully in the Introduction and in Figure 1—figure supplement 1.

5) Another point to mention in the Discussion section would be how the association becomes more mutualistic when there are more mites: the results indicate that more mites increase beetle fitness, which in turn benefit the mites (at the least, in order for beetle fitness to increase there has to be more mites. I'm not saying that the beetles directly promote mite fitness, but there is an association there. Also, perhaps more beetles could mean more mites can hitchhike?).

We agree that we should have mentioned in the Discussion section the effect of mite density on any movement along the parasitism-mutualism continuum. We have now made that clear.

The effects of beetle success on mite success are more complicated to discern than perhaps the referee realizes. In the field, most mite offspring disperse away on the adult beetles (Schwarz and Müller, 1992), rather than on beetle larvae produced during reproduction (as explained in the Results section). Mites are mostly transmitted horizontally between adults. So, while it is true that more beetles means more hosts for mites, it hard to link reproductive success at each breeding event to the subsequent reproductive success of the mites produced from the same carcass.

6) How common is it for phoresy to be parasitic? Worth mentioning how your consideration of the costs of mites squares with phoresy as currently understood in the literature.

Since phoretic interactions remain relatively understudied, it is hard to give an accurate answer to this question (White et al., 2017). We have followed your advice, though, and given two examples in the Introduction to show that burying beetle mites are not unusual in this respect (Results section). The main observation is that the ‘cost-free’ phoretic transport phase paves the way for a continued intimate association, after the host and phoront arrive at their destination. At this point, the relationship can become more parasitic (or more mutualistic).

7) Supplementary file 1, Supplementary file 2, Supplementary file 3 and Supplementary file 4, but particularly Supplementary file 3. Given the size of these models, do the authors have enough power to detect significant differences?

This is a good point. We have now performed power analyses for the mixed models presented in Supplementary file 3. We found all of these models were sufficient to reject the null hypotheses, with the 97%, 97%, and 98.2% power, for analyses of burying beetle larvae, blowfly larvae, and mite offspring, respectively. We have also performed power analyses for Supplementary file 1, Supplementary file 2, and Supplementary file 4. All these results are now included in the table captions. These powers were calculated using simulations based on 1000 Monte Carlo simulations, with the function powerSim in the package SIMR.

Reviewer #3:

[…] The paper is relatively well written and the authors are to be commended for their use of manipulative field and laboratory experiments. I enjoyed reading the paper and learned a lot. I do, however, have several suggestions that I think will improve the paper.

The title is vague. Authors sometimes assume that this will help the paper appear general and therefore capture broader readership. It doesn't. At least I would not bother to read it based on the title. And I think you want people like me to read it. A good title should state the key results and identify the system. I would read a paper titled something like: Phoretic mites help their carrion beetle hosts under temperature stress by eating competing blowfly larvae.

We have changed the title to make it more specific:

Temperature stress induces mites to help their carrion beetle hosts by eliminating rival blowflies

A general point about the Introduction: the paper will flow better (and read more like a scientific paper) if the authors clearly state the series of predictions they are going to test an how these predictions flow from their hypotheses. Each Pvalue in the result should be an evaluation of either an apriori prediction or the assessment of an assumption, or a controlling variable. And it should be clear how the hypotheses derive from the literature, theory or field observations.

We have rewritten the Introduction to make it clearer how we arrived at the questions we investigated in this study. As is evident from the experimental design, we expected that blowflies, temperature and mites – either alone or in combination – might influence burying beetle reproductive success. We have also restructured the Discussion section about the same set of questions.

The paper is missing some key information on what happens in nature. In particular, what is the natural distribution of mites on beetles in the field (this will help readers understand whether the treatments in the experiments are reasonable).

We have now provided this information by referring to our previously published paper (Sun et al., 2019), which describes the frequency distribution of mite density per beetle in their natural habitats.

This gets me to my point about whether the mite is a parasite or a mutualist or a whatever. This might be semantics, but I put a high bar on assigning parasitism. And the beetle-mite interaction does not cross that bar for me. Namely, the negative trophic interaction does not appear to occur during intimacy.

We can see we should have explained the natural history more clearly.

For example, we disagree on the point of about ‘intimacy’. The burying beetles and mites remain intimately associated during reproduction on the carcass. Mite offspring and larvae are in sufficiently close proximity that they bump into each other frequently during this period. We explain this more clearly now in the text (Results section).

Parasitism must be defined at the stage level rather than the species level.

Again we disagree – mainly because we don’t understand what additional conceptual insights we can gain by doing this, especially as we are working with a definition of parasitism that is centred upon fitness (and this makes it hard to separate parents from offspring conceptually). It’s also impossible to disentangle the stages in a practical sense. The parents and offspring generations of both mites and beetles mix together on the carcass, for example. The mite offspring are also dependent on the beetle parents to disperse away from the breeding site at the end of reproduction.

I also put a particularly high bar for deciding that parasitism is context dependent when simpler explanations via indirect effects are possible. To me it is clear that the phoretic mite stage is not a parasite because it does not feed on the beetle and the free-living mite stage is not a parasite because it neither feeds on the beetle or is intimate with the beetle.

We explain why we disagree with this characterization of the ‘free-living stage’ above. We now clarify that mites are mildly parasitic on beetles because they can directly predate on beetle eggs and larvae (Beninger, 1993; De Gasperin and Kilner, 2015; Nehring et al., 2017; Sun et al., 2019) (Results section). Therefore, mites can reduce beetle fitness both by direct predation as well as by indirect competition for carrion resource.

This does not mean the interaction is not interesting. In fact, I think it is far more interesting than the way the authors have packaged it. The mite is a phoretic commensal that uses its host to be transported to a common food source. Once on that common food source, the relationship can become competitive. But that competition can lead to a net indirect benefit under situations where another competitor (blowflies) might dominate. This interpretation is better couched in food-web theory rather than evolutionary theory.

We are grateful to the referee for pointing out a food web way of thinking about the results. We have integrated this style of thinking into the Discussion now and in a new figure to summarise the results (Figure 5).

As an example, here is a paper I remembered from decades ago that pondered over the relationship between mites and hosts and third-party interactions. I include the citation for the authors' interest. This author had similar observations but did not conclude a plastic parasitic-mutualistic relationship.

Rigby, (1996).

We now cite a paper by Stewart and Schnitzer, (2017), which clarifies the difference between competition and parasitism (summarized in Figure 1—figure supplement 1), and upon which we base the framing of the Introduction.

[Editors' note: further revisions were suggested prior to acceptance, as described below.]

The reviewers were mostly satisfied with your responses to their comments from the first round of review. However, there are still points that require further attention. In particular, both reviewers remained unconvinced by your assertion that this is a host-parasite system. We suggest the following course of action for addressing this concern: We feel there is no need to stress the host-parasite framework in the abstract, introduction, and result section – the work is very interesting without this. If you wish, you can add a short paragraph to the Discussion section that outlines the reasons why you think this system can also be viewed productively within a host-parasite framework. In this section, if you choose to include it, please contrast your interpretation with alternative views, as presented by the reviewers – that way, readers can hear both sides of the argument.

Further comments from reviewers #1 and #3 follow.

Reviewer #1:

The revised manuscript fixed many of the issues I had with the earlier version. The authors did a nice job. There are still two points I feel very uncomfortable with.

1) The work is still largely placed in a host – parasite framework. While I can follow the reasoning for doing so (positive – negative interactions), I think it is misleading and distorts the picture. It needs a lot of explanation to understand why this system might be explained by a host – parasite metaphor. Reducing this to positive – negative interactions is rather simple and not informative. Furthermore, host – parasite interactions and mutualism are concepts invented to describe pairwise interactions. As soon as three or more players act together these concepts do not work anymore (see the debate on this in the host – microbiota literature). While I see that other worker in the field have also used the host-parasite framework, it is not commonly done so.

We accept your point and we have now framed mite interactions as negative relationships instead of parasitism, wherever is relevant.

2) The treatment of temperature in the experiment is odd. Temperature is clearly a continuous variable, which in the experiment is broken into 3 categories. But the idea behind it is not an idea of categories, but rather of a continuum. This it is how it is presented in the interpretation of the study.

We agree with you that temperature is a continuously varying trait in nature, and this is why we analysed the data by including temperature as a continuous variable in our field experiment. However, in the lab we did not allow temperature to vary continuously in our experimental design, because it was easier to manipulate temperature by creating experimental categories of low, medium and high – within the natural range.

We use this approach commonly in our experimental designs. We take traits that vary continuously in nature and we subject them to experimental analysis by creating treatments in which variation (within the natural range) is cast into categories, so that we can more easily disentangle cause and effect. Here we have used the same approach to understand temperature by creating three treatments 11, 15, and 19°C. We did not measure temperature within each incubator so have no further information about temperature within each treatment.

In our experimental datasets, these variables of interest no longer vary continuously, but instead are made to resemble factors by our experimental design. Thus 13°C and 17°C do not exist in our experiments. This is why we have analysed temperature as a factor. We cannot draw inferences about effects at e.g. 13°C and 17°C in the lab because we did not measure them (unlike with the field dataset).

Furthermore, since there are only three different quantitative variables (11, 15, and 19°C), fitting a line or curve through the data is not appropriate.

Finally, we have good reasons (from our field data) for supposing that the effects of temperature are non-linear, which makes the business of drawing inferences about effects at temperatures we did not evaluate even trickier.

Reviewer #3:

The authors have done a nice job with the revision. We had a key disagreement, which they supported admirably, but I try to provide a bit more explanation so that this concern can be distinguished from picky semantics. The mite is not a parasite.

All deference to Tara Stewart's paper, figure 1 supplement does not solve my concern. I still don't agree that this study can be easily summarized as a host parasite relationship […] Below is an example of the edited Abstract that accomplishes this.

"Ecological transitions in and out of mutualism are relatively commonplace though the causal agents driving such change remain poorly understood. […] High densities of mites are especially effective at promoting beetle reproductive success at higher and lower natural ranges in temperature, when blowfly larvae are more potent rivals for the limited resources on the carcass."

Similar care in terminology make it easy to get rid of parasite and specify the actual biology of the system.

Lafferty and Kuris, (2002).

Thank you – this is very helpful indeed. We have now rephrased interactions between burying beetles and mites as negative relationships instead of parasitisms throughout the manuscript. We would also like to thank you for updating our abstract.

[Editors' note: further revisions were suggested prior to acceptance, as described below.]

The Reviewing Editor commented on your response to a point that had been raised during the previous round of review, namely the treatment of temperature in statistical analyses. This is a lightly edited version of their feedback: […]

We would like to ask you to address these considerations in your final revisions, which we feel can be achieved by clarifying the rationale of your approach to statistical analyses, and by using more nuanced wording for describing the findings from the experiments with three temperature treatments.

Thank you very much again for your advice on the use of temperature as a categorical factor. We had a good a priori reason for suspecting a non-linear effect of temperature, from the data we collected with our field experiment (Figure 1). This is (partly) why we adopted the conservative approach of including temperature as a categorical variable. We have now made that clear in the description of the statistical models in the Materials and methods section.

We have also changed the description of the results to add nuance (Results section, Discussion section).

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Citations

Syuan-Jyun S, Kilner RM. 2020. Data from: Temperature stress induces mites to help their carrion beetle hosts by eliminating rival blowflies. Dryad Digital Repository. [DOI] [PMC free article] [PubMed]

Supplementary Materials

Supplementary file 1. Results from the final models for each variable analysed.

elife-55649-supp1.docx^{(37.7KB, docx)}

Transparent reporting form

elife-55649-transrepform.docx^{(66.7KB, docx)}

Data Availability Statement

The following dataset was generated:

Syuan-Jyun S, Kilner RM. 2020. Data from: Temperature stress induces mites to help their carrion beetle hosts by eliminating rival blowflies. Dryad Digital Repository.

[bib1] Afkhami ME, McIntyre PJ, Strauss SY. Mutualist-mediated effects on species' range limits across large geographic scales. Ecology Letters. 2014;17:1265–1273. doi: 10.1111/ele.12332. [DOI] [PubMed] [Google Scholar]

[bib2] Ashby B, King KC. Friendly foes: the evolution of host protection by a parasite. Evolution Letters. 2017;1:211–221. doi: 10.1002/evl3.19. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib3] Barton BT, Ives AR. Direct and indirect effects of warming on aphids, their predators, and ant mutualists. Ecology. 2014;95:1479–1484. doi: 10.1890/13-1977.1. [DOI] [PubMed] [Google Scholar]

[bib4] Bates D, Maechler M, Bolker B, Walker S. R source package; 2015. https://www.jstatsoft.org/article/view/v067i01 [Google Scholar]

[bib5] Beninger CW. Egg predation by Poecilochirus carabi (Mesostigmata: Parasitidae) and its Effect on Reproduction of Nicrophorus vespilloides (Coleoptera: Silphidae) Environmental Entomology. 1993;22:766–769. doi: 10.1093/ee/22.4.766. [DOI] [Google Scholar]

[bib6] Bornemissza GF. An analysis of arthropod succession in Carrion and the effect of its decomposiion on the soil fauna. Australian Journal of Zoology. 1957;5:1. doi: 10.1071/ZO9570001. [DOI] [Google Scholar]

[bib7] Chamberlain SA, Bronstein JL, Rudgers JA. How context dependent are species interactions? Ecology Letters. 2014;17:881–890. doi: 10.1111/ele.12279. [DOI] [PubMed] [Google Scholar]

[bib8] Chen BF, Liu M, Rubenstein DR, Sun SJ, Liu JN, Lin YH, Shen SF. A chemically triggered transition from conflict to cooperation in burying beetles. Ecology Letters. 2020;23:467–475. doi: 10.1111/ele.13445. [DOI] [PubMed] [Google Scholar]

[bib9] Clay K. Defensive symbiosis: a microbial perspective. Functional Ecology. 2014;28:293–298. doi: 10.1111/1365-2435.12258. [DOI] [Google Scholar]

[bib10] Corbin C, Heyworth ER, Ferrari J, Hurst GD. Heritable symbionts in a world of varying temperature. Heredity. 2017;118:10–20. doi: 10.1038/hdy.2016.71. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib11] De Gasperin O, Duarte A, Troscianko J, Kilner RM. Fitness costs associated with building and maintaining the burying beetle's carrion nest. Scientific Reports. 2016;6:35293. doi: 10.1038/srep35293. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib12] De Gasperin O, Kilner RM. Friend or foe: inter-specific interactions and conflicts of interest within the family. Ecological Entomology. 2015;40:787–795. doi: 10.1111/een.12259. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib13] De Gasperin O, Kilner RM. Interspecific interactions and the scope for Parent-Offspring conflict: high mite density temporarily changes the Trade-Off between offspring size and number in the burying beetle, Nicrophorus vespilloides. PLOS ONE. 2016;11:e0150969. doi: 10.1371/journal.pone.0150969. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib14] Doremus MR, Oliver KM. Aphid heritable symbiont exploits defensive mutualism. Applied and Environmental Microbiology. 2017;83:e03276-16. doi: 10.1128/AEM.03276-16. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib15] Duarte A, Welch M, Swannack C, Wagner J, Kilner RM. Strategies for managing rival bacterial communities: lessons from burying beetles. Journal of Animal Ecology. 2018;87:414–427. doi: 10.1111/1365-2656.12725. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib16] Early R, Keith SA. Geographically variable biotic interactions and implications for species ranges. Global Ecology and Biogeography. 2019;28:42–53. doi: 10.1111/geb.12861. [DOI] [Google Scholar]

[bib17] Engl T, Eberl N, Gorse C, Krüger T, Schmidt THP, Plarre R, Adler C, Kaltenpoth M. Ancient symbiosis confers desiccation resistance to stored grain pest beetles. Molecular Ecology. 2018;27:2095–2108. doi: 10.1111/mec.14418. [DOI] [PubMed] [Google Scholar]

[bib18] Ewald PW. Transmission modes and evolution of the parasitism-mutualism continuum. Annals of the New York Academy of Sciences. 1987;503:295–306. doi: 10.1111/j.1749-6632.1987.tb40616.x. [DOI] [PubMed] [Google Scholar]

[bib19] Fatouros NE, Huigens ME. Phoresy in the field: natural occurrence of Trichogramma egg parasitoids on butterflies and moths. BioControl. 2012;57:493–502. doi: 10.1007/s10526-011-9427-x. [DOI] [Google Scholar]

[bib20] Fellous S, Salvaudon L. How can your parasites become your allies? Trends in Parasitology. 2009;25:62–66. doi: 10.1016/j.pt.2008.11.010. [DOI] [PubMed] [Google Scholar]

[bib21] Fitzpatrick G, Lanan MC, Bronstein JL. Thermal tolerance affects mutualist attendance in an ant-plant protection mutualism. Oecologia. 2014;176:129–138. doi: 10.1007/s00442-014-3005-8. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib22] Gelman A, Hill J. Data Analysis Using Regression and Multilevel/Hierarchical Models, Data Analysis Using Regression and Multilevel/hierarchical Models. Cambridge University Press; 2007. [Google Scholar]

[bib23] Goldberg E, Kirby K, Hall J, Latham J. The ancient woodland concept as a practical conservation tool in great britain. Journal for Nature Conservation. 2007;15:109–119. doi: 10.1016/j.jnc.2007.04.001. [DOI] [Google Scholar]

[bib24] Gorter FA, Scanlan PD, Buckling A. Adaptation to abiotic conditions drives local adaptation in Bacteria and viruses coevolving in heterogeneous environments. Biology Letters. 2016;12:20150879. doi: 10.1098/rsbl.2015.0879. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib25] Green P, MacLeod CJ. SIMR : an R package for power analysis of generalized linear mixed models by simulation. Methods in Ecology and Evolution. 2016;7:493–498. doi: 10.1111/2041-210X.12504. [DOI] [Google Scholar]

[bib26] Hoang KL, Gerardo NM, Morran LT. The effects of Bacillus subtilis on Caenorhabditis elegans fitness after heat stress. Ecology and Evolution. 2019;9:3491–3499. doi: 10.1002/ece3.4983. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib27] Hoeksema JD, Bruna EM. Context-dependent outcomes of mutualistic interactions. In: Bronstein JL, editor. Mutualism. Oxford University Press; 2015. pp. 181–202. [Google Scholar]

[bib28] Hopkins SR, Wojdak JM, Belden LK. Defensive symbionts mediate Host-Parasite interactions at multiple scales. Trends in Parasitology. 2017;33:53–64. doi: 10.1016/j.pt.2016.10.003. [DOI] [PubMed] [Google Scholar]

[bib29] Jaenike J, Unckless R, Cockburn SN, Boelio LM, Perlman SJ. Adaptation via symbiosis: recent spread of a Drosophila defensive symbiont. Science. 2010;329:212–215. doi: 10.1126/science.1188235. [DOI] [PubMed] [Google Scholar]

[bib30] Johnson NC. Mutualisms and ecosystem-level processes. In: Bronstein JL, editor. Mutualism. Oxford University Press; 2015. pp. 221–238. [Google Scholar]

[bib31] Lenth RV. Least-squares means: the R package lsmeans. Journal of Statistical Software. 2016;69:i01. doi: 10.18637/jss.v069.i01. [DOI] [Google Scholar]

[bib32] Lively CM, Clay K, Wade MJ, Fuqua C. Competitive co-existence of vertically and horizontally transmitted parasites. Evolutionary Ecology Research. 2005;7:1183–1190. [Google Scholar]

[bib33] Matuszewski S, Bajerlein D, Konwerski S, Szpila K. Insect succession and carrion decomposition in selected forests of central Europe part 2: composition and residency patterns of carrion fauna. Forensic Science International. 2010;195:42–51. doi: 10.1016/j.forsciint.2009.11.007. [DOI] [PubMed] [Google Scholar]

[bib34] Nehring V, Müller JK, Steinmetz N. Phoretic Poecilochirus mites specialize on their burying beetle hosts. Ecology and Evolution. 2017;7:10743–10751. doi: 10.1002/ece3.3591. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib35] Okabe K, Shun'ichi M. Parasitic mites as part-time bodyguards of a host wasp. Proceedings of the Royal Society B: Biological Sciences. 2008;275:2293–2297. doi: 10.1098/rspb.2008.0586. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib36] Palmer TM, Stanton ML, Young TP, Goheen JR, Pringle RM, Karban R. Breakdown of an ant-plant mutualism follows the loss of large herbivores from an african savanna. Science. 2008;319:192–195. doi: 10.1126/science.1151579. [DOI] [PubMed] [Google Scholar]

[bib37] Palmer T, Pringle E, Stier A, Holt R. Mutualism in a Community contextMutualism. Oxford University Press; 2015. [Google Scholar]

[bib38] Payne JA. A summer carrion study of the baby pig Sus scrofa linnaeus. Ecology. 1965;46:592–602. doi: 10.2307/1934999. [DOI] [Google Scholar]

[bib39] Peay KG. The mutualistic niche: mycorrhizal symbiosis and community dynamics. Annual Review of Ecology, Evolution, and Systematics. 2016;47:143–164. doi: 10.1146/annurev-ecolsys-121415-032100. [DOI] [Google Scholar]

[bib40] Rafaluk-Mohr C, Ashby B, Dahan DA, King KC. Mutual fitness benefits arise during coevolution in a nematode-defensive microbe model. Evolution Letters. 2018;2:246–256. doi: 10.1002/evl3.58. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib41] Schwarz HH, Starrach M, Koulianos S. Host specificity and permanence of associations between mesostigmatic mites (Acari: Anactinotrichida) and burying beetles (Coleoptera: Silphidae: Nicrophorus ) Journal of Natural History. 1998;32:159–172. doi: 10.1080/00222939800770101. [DOI] [Google Scholar]

[bib42] Schwarz HH, Müller JK. The dispersal behaviour of the phoretic mite Poecilochirus carabi (Mesostigmata, Parasitidae): adaptation to the breeding biology of its carrier Necrophorus vespilloides (Coleoptera, Silphidae) Oecologia. 1992;89:487–493. doi: 10.1007/BF00317154. [DOI] [PubMed] [Google Scholar]

[bib43] Scott MP. Competition with flies promotes communal breeding in the burying beetle, Nicrophorus tomentosus. Behavioral Ecology and Sociobiology. 1994;34:367–373. doi: 10.1007/BF00197007. [DOI] [Google Scholar]

[bib44] Scott MP. The ecology and behavior of burying beetles. Annual Review of Entomology. 1998;43:595–618. doi: 10.1146/annurev.ento.43.1.595. [DOI] [PubMed] [Google Scholar]

[bib45] Shelomi M, Matern LM, Dinstell JM, Harris DW, Kimsey RB. DEET (N,N-diethyl-meta-toluamide) induced delay of blowfly landing and oviposition rates on treated pig carrion (Sus scrofa L.) Journal of Forensic Sciences. 2012;57:1507–1511. doi: 10.1111/j.1556-4029.2012.02159.x. [DOI] [PubMed] [Google Scholar]

[bib46] Smiseth PT, Darwell CT, Moore AJ. Partial begging: an empirical model for the early evolution of offspring signalling. Proceedings of the Royal Society of London. Series B: Biological Sciences. 2003;270:1773–1777. doi: 10.1098/rspb.2003.2444. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib47] Springett BP. Aspects of the relationship between burying beetles, Necrophorus spp and the mite, Poecilochirus necrophori vitz. The Journal of Animal Ecology. 1968;37:417. doi: 10.2307/2957. [DOI] [Google Scholar]

[bib48] Sun SJ, Rubenstein DR, Chen BF, Chan SF, Liu JN, Liu M, Hwang W, Yang PS, Shen SF. Climate-mediated cooperation promotes niche expansion in burying beetles. eLife. 2014;3:e02440. doi: 10.7554/eLife.02440. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib49] Sun SJ, Horrocks NPC, Kilner RM. Conflict within species determines the value of a mutualism between species. Evolution Letters. 2019;3:185–197. doi: 10.1002/evl3.109. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib50] Sun S-J, Kilner RM. Cryptic host specialisation within Poecilochirus carabi mites explains population differences in the extent of co-adaptation with their burying beetle Nicrophorus vespilloides hosts. bioRxiv. 2019 doi: 10.1101/641936. [DOI] [PMC free article] [PubMed]

[bib51] Wall R, French N, Morgan KL. Effects of temperature on the development and abundance of the sheep blowfly Lucilia sericata (Diptera: Calliphoridae) Bulletin of Entomological Research. 1992;82:125–131. doi: 10.1017/S0007485300051531. [DOI] [Google Scholar]

[bib52] Wang Y, Rozen DE. Fitness costs of phoretic Nematodes in the burying beetle, Nicrophorus vespilloides. Ecology and Evolution. 2019;9:26–35. doi: 10.1002/ece3.4570. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib53] White PS, Morran L, de Roode J. Phoresy. Current Biology. 2017;27:R578–R580. doi: 10.1016/j.cub.2017.03.073. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib54] Wilson DS. The effect of population structure on the evolution of mutualism: a field test involving burying beetles and their phoretic mites. The American Naturalist. 1983;121:851–870. doi: 10.1086/284108. [DOI] [Google Scholar]

[bib55] Wilson DS, Knollenberg WG. Adaptive indirect effects: the fitness of burying beetles with and without their phoretic mites. Evolutionary Ecology. 1987;1:139–159. doi: 10.1007/BF02067397. [DOI] [Google Scholar]

PERMALINK

Temperature stress induces mites to help their carrion beetle hosts by eliminating rival blowflies

Syuan-Jyun Sun

Rebecca M Kilner

Roles

Abstract

Introduction

Results

Complementary patterns of reproductive success in burying beetles and blowflies, in the field

Figure 1. Reproductive success of burying beetles and blowflies under field conditions in relation to ambient air temperature, across the three different mite treatments.

Figure 1—figure supplement 1. Spatial distribution of breeding sites (yellow dots) used in the field experiment at the study in Madingley Wood, Cambridge, UK (Latitude: 52.22730°; Longitude: 0.04442°).

Figure 1—figure supplement 2. Schematic side-view representation of the experimental setup used for each breeding event in the field (dimensions are in cm).

Mites enhance burying beetle fitness in the field when there are blowflies present, but the effect depends on temperature and mite density

Complementary patterns of reproductive success in burying beetles and blowflies are induced by each other in the lab

Figure 2. Burying beetle reproductive success under lab conditions in relation to ambient air temperature in the incubator, without and with blowflies, and across three different mite treatments.

Figure 2—figure supplement 1. The daily mean, maximum, and minimum ambient air temperature in Madingley Woods during the field experiments conducted in 2016 and 2017.

Figure 2—figure supplement 2. Reproductive success of mites in relation to temperature, without and with blowflies and across the temperature treatments.

Blowflies are enemies to mites

In the lab, mites reduce burying beetle fitness at high densities when blowflies are absent

Mites switch from being harmful to mutualistic at lower and higher temperatures

Figure 3. Blowfly reproductive success in relation to temperature in the presence of (A) 0 mites, (B) 10 mites and (C) 20 mites.

At higher temperatures, blowflies evade attack through more rapid development

Figure 4. Effect of temperature on blowfly and burying beetle performance during carcass preparation.

Figure 4—figure supplement 1. Effect of temperature on blowfly reproductive performance.

At lower temperatures, beetle defences against blowflies are weaker

Discussion

Figure 5. A summary of the experimental results, showing how the interactions between burying beetles, mites, and blowflies change in response to an increase in temperature stress (caused by temperatures that are higher or lower than average).

(1) Do blowflies cause mites to switch from being harmful to becoming protective mutualists?

(2) Is the expression of the protective mutualism modulated by temperature?

(3) Is the expression of the protective mutualism modulated by the density of mites?

Materials and methods

Burying beetles and phoretic mites in Madingley Wood

Field experiment: how does burying beetle reproductive success covary with blowflies, mite density and ambient air temperature?

Maintenance of laboratory colonies of beetles, mites, blowflies

Burying beetles

Mites

Blowflies

Laboratory Experiment 1: manipulations of blowflies, mites and temperature

Manipulating the presence/absence of blowflies

Manipulations of mite density

Manipulations of temperature

Laboratory Experiment 2: effect of temperature on blowfly development

Laboratory Experiment 3: effect of temperature on beetle defences against blowflies during carcass preparation

Statistical analyses

Field experiment

Laboratory experiments

Analyses of beetle reproductive success

Analyses of blowfly reproductive success

Analyses of mite reproductive success

Effect of temperature on blowfly larval development

Effect of temperature on beetle’s carcass preparation

Acknowledgements

Funding Statement

Contributor Information

Funding Information

Additional information

Competing interests

Author contributions

Ethics

Additional files

Data availability

References

Decision letter

Roles

Author response

Associated Data

Data Citations

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases