Local adaptation to hummingbirds and bees in Salvia stachydifolia: insights into pollinator shifts in a Southern Andean sage.

BACKGROUND AND AIMS: Differences among populations in pollinator assemblages can lead to local adaptation mosaics in which plants evolve different floral morphologies and attractive traits. Mountain habitats may promote local adaptation because of differences in environmental conditions with altitude, causing changes in pollinators, and because mountaintops can act as isolated habitats. We studied if the differences in floral shape, size and nectar traits in Salvia stachydifolia can be attributed to variations in the relative contribution of hummingbirds and insects. METHODS: We studied eight populations of S. stachydifolia in natural and under common garden conditions, to assess whether population differences have a genetic component. We recorded pollinators, their behaviour and visitation rates, and characterized pollinator assemblages. In addition, we measured nectar volume and concentration, and collected flowers to describe floral shape and size variation using geometric morphometric methods. We then applied an unsupervised learning algorithm to identify ecotypes based on morphometric traits. Finally, we explored whether populations with different pollinator assemblages had different climatic and/or elevation preferences. KEY RESULTS: We found that variation in the identity of the main pollinators was associated with differences among populations in all traits, as expected under a local adaptation scenario. These differences persisted in the common garden, suggesting that they were not due to phenotypic plasticity. We found S. stachydifolia populations were pollinated either by bees, by hummingbirds or had mixed pollination. We identified two ecotypes that correspond to the identity of the main pollinator guilds, irrespective of climate or altitude. CONCLUSIONS: Variation in S. stachydifolia floral traits did not follow any evident association with bioclimatic factors, suggesting that populations may have diverged as the product of historical isolation on mountaintops. We suggest that differences among populations point to incipient speciation and an ongoing pollinator shift.

Did early shifts to bird pollination impose constraints on Salvia flower evolution?

A recent article by Kriebel et al. (2020) examines the relationship between floral shape evolution and pollination shifts in Salvia, a plant genus almost worldwide distributed, but particularly diverse in the New World. Kriebel et al. (2020) argue that a major shift to bird pollination at the origin of the subg. Calosphace (~20 million years ago [Mya]) imposed a legacy of constraints, resulting in significant differences in flower morphology between New World and Old World Salvia. However, reanalyses of the data using hidden states to account for the heterogeneity in evolutionary rates do not support an early origin of bird pollination in this group. Instead, bird pollination may have appeared after the arrival of modern hummingbirds to North America (15.5-12 Mya), as in other North American plant clades. The use of more complex models of ancestral state reconstruction into comparative analyses provides a different perspective to explain morphological differences within Salvia. Our results indicate that bird pollination did not impose constraints on corolla shape evolution. Evolutionary constraints in anther connective and style shapes may have arisen at the origin of Calosphace but they were not associated with shifts to hummingbird pollination, being more likely the product of contingent evolution.

Evolutionary transition between bee pollination and hummingbird pollination in Salvia: Comparing means, variances and covariances of corolla traits.

Covariation among traits can modify the evolutionary trajectory of complex structures. This process is thought to operate at a microevolutionary scale, but its long-term effects remain controversial because trait covariation can itself evolve. Flower morphology, and particularly floral trait (co)variation, has been envisioned as the product of pollinator-mediated selection. Available evidence suggests that major changes in pollinator assemblages may affect the joint expression of floral traits and their phenotypic integration. We expect species within a monophyletic lineage sharing the same pollinator type will show not only similarity in trait means but also similar phenotypic variance-covariance structures. Here, we tested this expectation using eighteen Salvia species pollinated either by bees or by hummingbirds. Our findings indicated a nonsignificant multivariate phylogenetic signal and a decoupling between means and variance-covariance phenotypic matrices of floral traits during the evolution to hummingbird pollination. Mean trait value analyses revealed significant differences between bee- and hummingbird-pollinated Salvia species although fewer differences were detected in the covariance structure between groups. Variance-covariance matrices were much more similar among bee- than hummingbird-pollinated species. This pattern is consistent with the expectation that, unlike hummingbirds, bees physically manipulate the flower, presumably exerting stronger selection pressures favouring morphological convergence among species. Overall, we conclude that the evolution of hummingbird pollination proceeded through different independent transitions. Thus, although the evolution of hummingbird pollination led to a new phenotypic optimum, the process involved the diversification of the covariance structure.

A pilot study applying the plant Anchored Hybrid Enrichment method to New World sages (Salvia subgenus Calosphace; Lamiaceae).

We conducted a pilot study using Anchored Hybrid Enrichment to resolve relationships among a mostly Neotropical sage lineage that may have undergone a recent evolutionary radiation. Conventional markers (ITS, trnL-trnF and trnH-psbA) have not been able to resolve the relationships among species nor within portions of the backbone of the lineage. We sampled 12 representative species of subgenus Calosphace and included one species of Salvia's s.l. closest relative, Lepechinia, as outgroup. Hybrid enrichment and sequencing were successful, yielding 448 alignments of individual loci with an average length of 704bp. The performance of the phylogenomic data in phylogenetic reconstruction was superior to that of conventional markers, increasing both support and resolution. Because the captured loci vary in the amount of net phylogenetic informativeness at different phylogenetic depths, these data are promising in phylogenetic reconstruction of this group and likely other lineages within Lamiales. However, special attention should be placed on the amount of phylogenetic noise that the data could potentially contain. A prior exploration step using phylogenetic informativeness profiles to detect loci with sites with disproportionately high substitution rates (showing "phantom" spikes) and, if required, the ensuing filtering of the problematic data is recommended. In our dataset, filtering resulted in increased support and resolution for the shallow nodes in maximum likelihood phylogenetic trees resulting from concatenated analyses of all the loci. Additionally, it is expected that an increase in sampling (loci and taxa) will aid in resolving weakly supported, short deep internal branches.

Beyond neutral and forbidden links: morphological matches and the assembly of mutualistic hawkmoth-plant networks.

A major challenge in evolutionary ecology is to understand how co-evolutionary processes shape patterns of interactions between species at community level. Pollination of flowers with long corolla tubes by long-tongued hawkmoths has been invoked as a showcase model of co-evolution. Recently, optimal foraging models have predicted that there might be a close association between mouthparts' length and the corolla depth of the visited flowers, thus favouring trait convergence and specialization at community level. Here, we assessed whether hawkmoths more frequently pollinate plants with floral tube lengths similar to their proboscis lengths (morphological match hypothesis) against abundance-based processes (neutral hypothesis) and ecological trait mismatches constraints (forbidden links hypothesis), and how these processes structure hawkmoth-plant mutualistic networks from five communities in four biogeographical regions of South America. We found convergence in morphological traits across the five communities and that the distribution of morphological differences between hawkmoths and plants is consistent with expectations under the morphological match hypothesis in three of the five communities. In the two remaining communities, which are ecotones between two distinct biogeographical areas, interactions are better predicted by the neutral hypothesis. Our findings are consistent with the idea that diffuse co-evolution drives the evolution of extremely long proboscises and flower tubes, and highlight the importance of morphological traits, beyond the forbidden links hypothesis, in structuring interactions between mutualistic partners, revealing that the role of niche-based processes can be much more complex than previously known.