Similarly, some robust annual populations were able to produce enough flowers over the extended growth season in this mesic habitat to outperform the focal population. Conversely, fast-cycling annuals appeared developmentally unable to exploit the longer growth season in this habitat, flowering earlier at a small size and producing few flowers . The relatively low performance of the montane perennial ecotype could reflect inbreeding depression and/or a bet-hedging strategy that prioritizes adult persistence through below-ground rhizomes at the expense of reproductive output in any given year . Together, the relative performance of both immigrants and hybrids was quadratic with phenotypic divergence, suggesting that ecologically intermediate genotypes enjoy a fitness advantage whereas increasing phenotypic divergence results in immigrant inviability or, potentially, reduced hybrid performance. This pattern may be common during the early stages of ecological divergence due interactions between inbreeding depression and divergent adaptation . I found strong support for this hypothesis by manipulating the inbreeding history prior to hybridization, suggesting that, in many systems, selection against immigrants and hybrids may not be the earliest or strongest barriers to evolve. Indeed, procona system several studies have failed to find selection against immigrants or hybrids even when other ecological barriers are present.
Several lines of evidence suggest that the traits underlying reproductive isolation in this study are the result of divergent selection between annual and perennial habitats. The trade-off between rapid flowering and vegetative growth is a major axis of phenotypic divergence in this species complex that is determined by multiple pleiotropic QTL . Early flowering is under strong selection in fast-drying annual habitats, whereas later flowering and/or greater vegetative growth are favored in coastal and montane perennial habitats . Correspondingly, flowering time isolation in immigrant populations was greater for annuals than inland perennials, whereas immigrant and hybrid fitness was greatest for inland perennials that invest in vegetative growth. Reduced viable seed set could be due to pollen-pistil incompatibilities preventing fertilization or early-acting genetic incompatibilities in developing hybrid seeds. Although the traits driving these barriers are as yet unknown, the relationship between crossing barrier strength and phenotypic divergence suggests a role for natural selection. Crossing barriers with the focal population were stronger in inland perennial populations than annual populations, and this pattern could reflect differences in life history strategy. Fecundity selection is strong in annual populations but weaker in perennial populations, where other components of fitness including survival and growth also shape the pattern of selection . Relaxed fecundity selection in perennial populations could allow more rapid genetic divergence in loci related to pollen-pistil interactions or seed development, potentially increasing the likelihood of incompatibilities.
Future research should identify the mechanisms underlying crossing barriers to test this hypothesis.What determines whether partially isolated taxa will evolve into good species, remain partially isolated ecotypes, or collapse into a single population?Parallel ecological contrasts have yielded variable progress toward genetic isolation among White Sands lizards , lake-stream stickleback , insect host-races , and mimetic butterflies . Several hypotheses have been developed to explain variation in progress toward speciation, including the strength and consistency of selection, the number of traits or loci mediating adaptation or isolation, and opportunities for gene flow and reinforcement . I found no evidence that reproductive isolation was related to geographic distance, a proxy for opportunities for gene flow. Alternatively, individual barriers and total reproductive isolation with a focal montane perennial population were correlated with phenotypic divergence. Yet, the strength of total isolation was still incomplete and dependent on environmental variation between years as well as the history of inbreeding. Reproductive isolation with a montane perennial population is weaker than reproductive isolation between annual and coastal perennial ecotypes in this species . In annual populations, late-flowering coastal perennial genotypes face nearly complete immigrant inviability due to their inability to reproduce prior to the onset of drought. In coastal habitats, annual genotypes have reduced fitness due to salt spray and flower earlier than the native coastal perennial genotypes, while F1 hybrids exhibit strong heterosis. Together, reproductive isolation in these habitats ranges from 0.93 – 1.00 , exceeding estimates from this study which ranged from 0.60 – 0.76 for annual populations including hybrid fitness. The differences in reproductive isolation among these life history ecotypes suggests that strong viability selection early in the life-cycle, due to drought or salt spray, is an important factor in the rapid evolution of strong reproductive isolation.
Drought avoidance traits are highly correlated among populations of M. guttatus , providing support for the view that strong selection on few traits, such as rapid development or salt tolerance, promotes speciation . However, quantitative genetic studies have demonstrated a polygenic basis for both flowering time and salt tolerance in this species , such that selection on individual loci may be weaker. Negative pleiotropy between early flowering and vegetative rosette production could generate fitness trade-offs between annual and perennial habitats , whereas Lowry et al. found no fitness costs for salt tolerance QTL in an annual habitat. Thus, although conceptually useful, distinguishing between the stronger selection and multifarious selection hypotheses may be difficult in practice.The broader community of microbes within a host, the microbiome, can determine the health status of an individual. Many microbes provide beneficial functions for the host including metabolism and immunity. In honey bees, certain Lactobacillus strains offered protection against a microsporidian and bacterial pathogen. Similarly, in bumble bees, increased microbiome diversity was linked to reduced infection by the trypanosomatid parasite Crithidia. In Osmia ribofloris, the pollen provision microbiome is crucial for larval development. Therefore, it is important to characterize and understand the microbiome to understand bee health. Our current knowledge on the microbiome of bees is predominantly based on honey bees , and to a lesser degree bumble bees. Both of these are highly social and closely related members of the corbiculate apid bees, and as such they share a very similar core microbiome. Outside of these genera, the bee microbiomes sequenced so far do not conform to the Apis and Bombus models. Even within the corbiculates, the stingless bees and the orchid bees lack some of the most common symbionts of Apis and Bombus, although several related symbionts are shared amongst the corbiculates. Looking more broadly, bacteria thatwere previously classified as Lactobacillus but have been recently split into the genera Apilactobacillus, Bombilactobacillus and Lactobacillus sensu strictu are some of the few symbionts common to multiple bee taxa including Apis, Bombus, the small carpenter bee Ceratina, megachilid and halictid bees. Microbe acquisition in Apis and Bombus occurs within the hive, facilitated by nestmate interactions or transfer from feces. While honey and bumble bees live in large groups, this level of sociality is rare among bees, the vast majority of bee species being solitary.
Indeed, in the other bee species studied so far, much of their microbiota appears to be gained from the environment rather than through social transmission. Therefore, differences in environmental and pollen-associated bacteria may have larger impacts on wild bee development than for the highly social corbiculate bee species. Microbial acquisition from the environment may be influenced by the diet of the bee. As for bees, flowers harbor a variety of microbes, which can potentially be passed to foraging bees. Crithidia can be transmitted between foraging Bombus at flowers and communities of pollinators have been found to share microbes. For bees that use foliage to line their nests, procona valencia buckets both flower and foliar source affect their pollen provision microbiome. However, there are also many more complex factors to consider such as flower morphology, volatiles and even the secondary compounds produced by the microbes themselves, that can alter floral bacterial communities and transmission to pollinators. Therefore, diet may be an important factor to consider when looking at the wild bee microbiome, which is thought to be largely environmentally sourced. To conserve wild bees, we need to understand their health, and their associated microbial symbionts. It seems likely that the microbes present in the environment, and therefore those gained environmentally by bees, will vary geographically with changes in climate, interacting insect species and floral communities. Ceratina calcarata Robertson, 1900 is a small carpenter bee species that is a widespread and prominent pollinator across eastern North America. This species nests in the dead stems of various plants, commonly raspberry and sumac. The plants it nests in also produce flowers, potentially biasing pollen collection and thereby microbial acquisition. This bee constructs separate brood cells within the stem nest, each provisioned with a single pollen ball on which an egg is laid . This brood provision is the only source of food given to the offspring until it reaches maturity. Study of these brood provisions from nests at the northern extent of its range found they contain multiple pollen species and a diversity of microbes dominated by Lactobacillus, Wolbachia, Acinetobacter and Sodalis. However, C. calcarata is found across a broad geographic range in eastern North America and acquires at least part of its microbiome from the environment, therefore its microbiome may vary geographically with corresponding changes in climate and floral landscape. The aim of this project was to investigate whether the microbiome of C. calcarata varies geographically by sequencing brood provisions spanning this species’ range across the eastern United States. Specifically, we asked whether pollen or bacterial species vary in composition or diversity among sites and if there are identifiable plant/microbe associations. We also asked whether foraging was biased by the proximity of nest plant flowers.The floral resources utilized by C. calcarata differed between regions, brood provisions being dominated by different pollen genera in each state, showing this generalist bee’s local adaptation to regional floral communities . Foraging preference was not biased towards flowers present on the host nest plant, indicating that spatial assortment of floral resources alone does not determine foraging preferences. Despite changes in floral resources, the same core microbes dominated brood provisions across all states, although the relative abundance of these groups did vary between region . Our overall network analyses identified some correlations in plant and bacterial occurrences, however the broad changes in floral diet between states did not correspond to large changes in the bacterial community, suggesting that these floral–bacterial associations are transient or non-obligate . There were a number of core bacterial genera found across all sites but the relative abundance of these varied strongly . Comparison to McFrederick and Rehan’s study in New Hampshire suggests there may also be annual variation within sites. In 2016, Lactobacillus was the most common genus, while we recovered Sodalis and Wolbachia as the top two genera in the current study . Being obligate facultative endosymbionts in many insect species, the occurrence of Wolbachia and Sodalis is likely due to contamination from mites or other parasites, or transfer directly from the mother’s crop rather than through any specific floral sources. A major limitation of amplicon sequencing studies is that they can only determine the presence or absence of an organism’s DNA, not whether that organism is metabolically active, and this holds true for all bacteria recovered in our samples. This aside, it is interesting that Wolbachia contamination is so prevalent, and future microscopic examination of pollen material for mite infestation and tests of possible vertical transmission via pollen are needed. Lactobacillus was only present in 2.2% of reads in the New Hampshire samples for this study. Within Georgia, the abundance of bacterial ESVs differed between nests in Rubus and Rhus but significant differences disappeared when the phylogenetic similarity of bacteria was considered. This suggests there could be some differences in the abundance of microbial species or strains between nesting substrates but overall the taxonomic distribution of bacteria in the brood provisions is similar. Nesting substrate, therefore, has a smaller or negligible effect on bacterial abundance compared to differences among states. Ceratina calcarata foraged from a slightly greater phylogenetic richness of floral genera in New Hampshire than in Georgia. The phylogenetic richness of floral genera in Missouri did not significantly differ from either Georgia or New Hampshire, perhaps because of its mid-lying geographic position and climate. As expected, our reanalysis of the New Hampshire data with 99% ESV matching recovered fewer genera than in McFrederick and Rehan, who found 110 genera compared to this study with 65. The genera not recovered in our reanalysis were all present in less than 1% of reads in the original study. We identified the same five genera as being the most abundant , these genera accounting for 92% of the reads .