The ability of an insect to survive attack by natural enemies can be modulated by the presence of defensive symbionts. Study of aphid-symbiont-enemy interactions has indicated that protection may depend on the interplay of symbiont, host and attacking parasite genotypes. However, the importance of these interactions is poorly understood outside of this model system. Here, we study interactions within a Drosophila model system, in which Spiroplasma protect their host against parasitoid wasps and nematodes. We examine whether the strength of protection conferred by Spiroplasma to its host, Drosophila melanogaster varies with strain of attacking Leptopilina heterotoma wasp. We perform this analysis in the presence and absence of ethanol, an environmental factor that also impacts the outcome of parasitism. We observed that Spiroplasma killed all strains of wasp. However, the protection produced by Spiroplasma following wasp attack depended on wasp strain. A composite measure of protection, including both the chance of the fly surviving attack and the relative fecundity/fertility of the survivors, varied from a <4% positive effect of the symbiont following attack of the fly host by the Lh14 strain of wasp to 21% for the Lh-Fr strain in the absence of ethanol. We also observed that environmental ethanol altered the pattern of protection against wasp strains. These data indicate that the dynamics of the Spiroplasma-Drosophila-wasp tripartite interaction depend upon the genetic diversity within the attacking wasp population, and that prediction of symbiont dynamics in natural systems will thus require analysis across natural enemy genotypes and levels of environmental ethanol.

The outcome of natural enemy attack in insects is commonly influenced by the presence of protective symbionts in the host. The degree to which protection functions in natural populations, however, will depend on the robustness of the phenotype and symbiosis to variation in the abiotic environment. We studied the impact of a key environmental parameter—temperature—on the efficacy of the protective effect of the symbiont Spiroplasma on its host Drosophila hydei, against attack by the parasitoid wasp Leptopilina heterotoma. In addition, we investigated the thermal sensitivity of the symbiont's vertical transmission, which may be a key determinant of the ability of the symbiont to persist. We found that vertical transmission was more robust than previously considered, with Spiroplasma being maintained at 25°C, at 18°C and with 18/15°C diurnal cycles, with rates of segregational loss only increasing at 15°C. Protection against wasp attack was ablated before symbiont transmission was lost, with the symbiont failing to rescue the fly host at 18°C. We conclude that the presence of a protective symbiosis in natural populations cannot be simply inferred from the presence of a symbiont whose protective capacity has been tested under narrow controlled conditions. More broadly, we argue that the thermal environment is likely to represent an important determinant of the evolutionary ecology of defensive symbioses in natural environments, potentially driving seasonal, latitudinal and altitudinal variation in symbiont frequency.

The outcome of natural enemy attack in insects has commonly been found to be influenced by the presence of protective symbionts in the host. The degree to which protection functions in natural populations, however, will depend on the robustness of the phenotype to variation in the abiotic environment. We studied the impact of a key environmental parameter – temperature – on the efficacy of the protective effect of the symbiont Spiroplasma on its host Drosophila hydei, against attack by the parasitoid wasp Leptopilina heterotoma. In addition, we investigated the thermal sensitivity of the symbiont’s vertical transmission, which may be a key determinant of the ability of the symbiont to persist. We found that vertical transmission was more robust than previously considered, with Spiroplasma being maintained at 25 °C, 18 °C and with 18/15 °C diurnal cycles, with rates of segregational loss only increasing at 15 °C. Protection against wasp attack was ablated before symbiont transmission was lost, with the symbiont failing to rescue the fly host at 18 °C. We conclude that the presence of a protective symbiosis in natural populations cannot be simply inferred from presence of a symbiont whose protective capacity has been tested under narrow controlled conditions. More broadly, we argue that the thermal environment is likely to represent an important determinant of the evolutionary ecology of defensive symbioses in natural environments, potentially driving seasonal, latitudinal and altitudinal variation in symbiont frequency, and modulating the strength of selection for symbiotic protective systems compared to defensive systems encoded in the nuclear genomes.

AbstractWhen a parasite attacks an insect, the outcome is commonly modulated by the presence of defensive heritable symbionts residing within the insect host. Previous studies noted markedly different strengths of Spiroplasma-mediated fly survival following attack by the same strain of wasp. One difference between the two studies was the strain of Spiroplasma used. We therefore performed a common garden laboratory experiment to assess whether Spiroplasma-mediated protection depends upon the strain of Spiroplasma. We perform this analysis using the two strains of male-killing Spiroplasma used previously, and examined response to challenge by two strains of Leptopilina boulardi and two strains of Leptopilina heterotoma wasp. We found no evidence Spiroplasma strain affected fly survival following wasp attack. In contrast, analysis of the overall level of protection, including the fecundity of survivors of wasp attack, did indicate the two Spiroplasma strains tested varied in protective efficiency against three of the four wasp strains tested. These data highlight the sensitivity of symbiont-mediated protection phenotypes to laboratory conditions, and the importance of common garden comparison. Our results also indicate that Spiroplasma strains can vary in protective capacity in Drosophila, but these differences may exist in the relative performance of survivors of wasp attack, rather than in survival of attack per se.

When a parasite attacks an insect, the outcome is commonly modulated by the presence of defensive heritable symbionts residing within the insect host. Previous studies noted markedly different strengths of Spiroplasma‐mediated fly survival following attack by the same strain of wasp. One difference between the two studies was the strain of Spiroplasma used. We therefore performed a laboratory experiment to assess whether Spiroplasma‐mediated protection depends upon the strain of Spiroplasma. We perform this analysis using the two strains of male‐killing Spiroplasma used previously, and examined response to challenge by two strains of Leptopilina boulardi and two strains of Leptopilina heterotoma wasp. We found no evidence Spiroplasma strain affected fly survival following wasp attack. In contrast, analysis of the overall level of protection, including the fecundity of survivors of wasp attack, did indicate the two Spiroplasma strains tested varied in protective efficiency against three of the four wasp strains tested. These data highlight the sensitivity of symbiont‐mediated protection phenotypes to laboratory conditions, and the importance of common garden comparison. Our results also indicate that Spiroplasma strains can vary in protective capacity in Drosophila, but these differences may exist in the relative performance of survivors of wasp attack, rather than in survival of attack per se.