Microbial Ecology Temperature Determines Symbiont Abundance in a Multipartite Bark Beetle-fungus Ectosymbiosis

In this study, we report evidence that temperature plays a key role in determining the relative abundance of two mutualistic fungi associated with an economically and ecologically important bark beetle, Dendroctonus ponderosae. The symbiotic fungi possess different optimal temperature ranges. These differences determine which fungus is vectored by dispersing host beetles as temperatures fluctuate over a season. Grosmannia clavigera is the predominant fungus carried by dispersing beetles during cool periods but decreases in prevalence as daily maximum temperatures approach 25-C, and becomes extremely rare when temperatures reach or exceed 32-C. In contrast, Ophiostoma montium increases in prevalence as temperatures approach 25-C, and becomes the predominant symbiont dispersed when temperatures reach or exceed 32-C. The possession of different optimal growth temperatures may facilitate the stable coexistence of the two fungi by supporting growth of each fungus at different times, minimizing direct competition. Furthermore, the beetle may reduce its risk of being left aposymbiotic by exploiting not one, but two symbionts, whose combined growth optima span a wide range of environmental conditions. The possession of multiple symbionts with different temperature tolerances may allow the beetle to occupy highly variable habitats over a wide geographic range. Such temperature-driven symbiont shifts are likely to have major consequences for both the host and its symbionts under current temperature regimes and those predicted to occur because of climate change. Introduction Mutualisms are predicted to be inherently unstable and prone to erosion because of cheating by established symbionts or invasion by exploiters [7]. Obligate mutualisms are also viewed as risky given that if one associate is lost, the other cannot survive [13]. However, despite these predictions, many mutualisms, including many obligate associations, appear to have existed in a relatively stable state over long evolutionary periods. Most studies attempting to detect mechanisms of stability in mutualisms have focused on pairwise interactions (one host–one symbiont) ([8] and references therein). However, many mutualisms involve multiple symbionts, several of which may fulfill similar roles with the host (symbiont redundancy). Whereas redundant symbionts impart the same type of benefit to a host, it is highly unlikely that any two will provide exactly the same degree of benefit. That Bredundant^ symbionts may differ in the degree of benefit they confer raises a number of questions unique to multipartite symbioses. For example, do such symbionts differ in their relative prevalence among host populations, and if so, what factors influence this variability? Do differences in the prevalence of symbionts within and among populations affect host fitness and host population dynamics? If the answer to the last question is yes, then why and how apparently inferior symbionts are maintained in stable symbiosis becomes an especially intriguing question. To address these questions, we are investigating a multipartite mutualistic ectosymbiosis occurring among a tree-killing bark beetle, Dendroctonus ponderosae, and two fungi, Grosmannia clavigera (formerly Ophiostoma clavigerum [31]) and Ophiostoma montium. Phylogenetic studies indicate that G. clavigera has a long, shared evolutionary history with the host beetle, whereas O. montium appears to be a relative newcomer [27; D.L. Six et al., unpublished observation]. Although O. montium is Correspondence to: D.L. Six; E-mail: [email protected] DOI: 10.1007/s00248-006-9178-x & * Springer Science+Business Media, Inc. 2006 a more recent associate, it is widespread in its occurrence and is likely to have also been associated with the host over a relatively long period of evolutionary history. The symbiosis between the beetle and the two fungi is considered mutualistic. The fungi receive benefit from association with the beetle through reliable transport to food resources and habitat [24]. Before emergence from the host tree, new adult beetles feed on spore layers produced by the fungi on the walls of beetle pupal chambers [24]. While feeding, the beetles pack their mycangia (structures of the integument specialized for dissemination of fungi) with spores, and then disperse to colonize new trees where they inoculate the fungi into tree tissues and lay eggs [24]. In return, the beetles gain nutritional benefit from feeding on the fungi, both as larvae and adults. The larvae of the beetles develop and feed in the phloem layer of trees [1]. Phloem is a nutrient-poor food, and symbiotic fungi are believed to supplement the tree-based diet of Dendroctonus beetles by concentrating nitrogen [2] and/or providing sterols required for molting and reproduction [4]. The two fungi associated with D. ponderosae appear to differ in their relative contributions to host nutrition. D. ponderosae reared in logs with G. clavigera develop faster and produce more brood than when associated with O. montium [26]. In addition, feeding on fungi by new adult beetles before dispersal appears to be critical for beetle reproduction and both fungi appear to be adequate for this purpose. New adults allowed to feed on spores of either fungus, will enter logs, construct galleries and lay eggs. In contrast, beetles that do not feed on spores, will not enter logs, and do not produce galleries or eggs [26]. In addition to sharing the same host, the two fungi exploit the same habitat and the same nutritional resources within the tree. Such a broad overlap of niches should lead to strong direct competition between the two fungi for space, nutrients, and hosts, destabilizing the multipartite symbiosis. In addition, differential benefits conferred to the host by the two symbionts should also act to destabilize the symbiosis by increasing selection for the symbiont that confers superior benefits and selecting against the symbiont that confers inferior benefits. Despite these predictions, this multipartite symbiosis appears to be relatively stable, indicating the existence of a mechanism that allows both fungi to coexist with the host. In this research, we investigated whether the relative proportions of the two fungal symbionts associated with D. ponderosae vary seasonally among and within populations, and from year to year within a population, by monitoring the fungi carried in mycangia of dispersing new adults at several sites in Montana and Idaho (Table 1). We then investigated whether the environmental variable, ambient temperature, influenced the patterns of symbiont prevalence that we detected. By investigating patterns of symbiont prevalence and the factors that affect variability in prevalence, we can gain insight in to the proximate mechanisms by which multipartite associations maintain stability.