On the relevance of molecular tools for taxonomic revision in Malvales, Malvaceae s.l., and Dombeyoideae.

In this article, we present an overview of changes to the taxonomy of Malvales. In traditional classifications, this order was variously circumscribed as including four main families (i.e., Malvaceae, Bombacaceae, Sterculiaceae, and Tiliaceae, also known now as "Core Malvales"), but major disagreements existed between different taxonomic treatments. Contributions from molecular data, new morpho-anatomical data, and progress in methodological approaches have recently led to a new broader concept of this order (namely, "expanded Malvales"). Now, expanded Malvales includes ten families (Neuradaceae, Thymelaeaceae, Sphaerosepalaceae, Bixaceae, Cistaceae, Sarcolaenaceae, Dipterocarpaceae, Cytinaceae, Muntingiaceae, Malvaceae s.l.) distributed among seven monophyletic lineages. All these families were previously considered to have malvalean affinities in some traditional treatments, except the holoparasitic and highly modified Cytinaceae. Although molecular evidence has clarified the Malvales taxonomy, the phylogenetic positions of Sarcolaenaceae, Thymelaeaceae, and Sphaerosepalaceae are still controversial and need new analyses focusing specifically on these families to assess their phylogenetic placement and their morphological evolution.In a phylogenetic context, molecular data combined with recent examination of morphological characters supported the hypothesis of a common origin of "core Malvales." However, these analyses also showed that the former families but Malvaceae s.s. were paraphyletic or polyphyletic. As a consequence, recent taxonomic treatments grouped taxa formerly included in "Core Malvales" in a broader concept of Malvaceae s.l. Additionally, the intrafamilial taxonomy has been deeply modified, and in its present circumscription, Malvaceae includes nine subfamilies (Grewioideae, Byttnerioideae, Sterculioideae, Dombeyoideae, Brownlowioideae, Tilioideae, Bombacoideae, Malvoideae, Helicteroideae) in two main lineages. Phylogenetic studies on subfamilial rearrangements have focused on the relationships between emblematic taxa such as Bombacoideae and Malvoideae (which form together the /Malvatheca lineage). However, our understanding of the phylogenetic relationships among and within taxa of the other subfamilies (e.g., Dombeyoideae, Tilioideae, and Sterculioideae) has not followed at the same pace. Despite recent investigations, the relationships between the subfamilies of Malvaceae s.l. remain controversial. As an example of these taxonomic issues, we review the systematic studies on Dombeyoideae, with special emphasis on taxa endemic to the Mascarene archipelago (Indian Ocean). Recent investigations have shown that several island endemic genera such as Trochetia, Ruizia, and Astiria (endemic to the Mascarenes) are nested within the mega-genus Dombeya. Consequently, the current taxonomy of this genus does not match the phylogeny and should be modified. Therefore, we propose three possible taxonomic schemes as part of an ongoing revision of the Mascarene Dombeyoideae. However, these taxonomic rearrangements should only be made after a broader study of the diversity in Madagascar and adjacent areas. This broader approach shall avoid possibly multiple and contradictory taxonomic revisions of restricted regions if they were each studied in isolation.

Multiple colonizations from Madagascar and converged acquisition of dioecy in the Mascarene Dombeyoideae (Malvaceae) as inferred from chloroplast and nuclear DNA sequence analyses.

BACKGROUND AND AIMS: In the Mascarenes, a young oceanic archipelago composed of three main islands, the Dombeyoideae (Malvaceae) have diversified extensively with a high endemism rate. With the exception of the genus Trochetia, Mascarene Dombeyoideae are described as dioecious whereas Malagasy and African species are considered to be monocline, species with individuals bearing hermaphrodite/perfect flowers. In this study, the phylogenetic relationships were reconstructed to clarify the taxonomy, understand the phylogeographic pattern of relationships and infer the evolution of the breeding systems for the Mascarenes Dombeyoideae. METHODS: Parsimony and Bayesian analysis of four DNA markers (ITS, rpl16 intron and two intergenic spacers trnQ-rsp16 and psbM-trnD) was used. The molecular matrix comprised 2985 characters and 48 taxa. The Bayesian phylogeny was used to infer phylogeographical hypotheses and the evolution of breeding systems. KEY RESULTS: Parsimony and Bayesian trees produced similar results. The Dombeyoideae from the Mascarenes are polyphyletic and distributed among four clades. Species of Dombeya, Trochetia and Ruizia are nested in the same clade, which implies the paraphyly of Dombeya. Additionally, it is shown that each of the four clades has an independent Malagasy origin. Two adaptive radiation events have occurred within two endemic lineages of the Mascarenes. The polyphyly of the Mascarene Dombeyoideae suggests at least three independent acquisitions of dioecy. CONCLUSIONS: This molecular phylogeny highlights the taxonomic issues within the Dombeyoideae. Indeed, the limits and distinctions of the genera Dombeya, Trochetia and Ruizia should be reconsidered. The close phylogeographic relationships between the flora of the Mascarenes and Madagascar are confirmed. Despite their independent origins and a distinct evolutionary history, each endemic clade has developed a different breeding systems (dioecy) compared with the Malagasy Dombeyoideae. Sex separation appears as an evolutionary convergence and may be the consequence of selective pressures particular to insular environments.

Do rewardless orchids show a positive relationship between phenotypic diversity and reproductive success?

Among rewardless orchids, pollinator sampling behavior has been suggested to drive a positive relationship between population phenotypic variability and absolute reproductive success, and hence population fitness. We tested this hypothesis by constructing experimental arrays using the rewardless orchid Dactylorhiza sambucina, which is dimorphic for corolla color. We found no evidence that polymorphic arrays had higher mean reproductive success than monomorphic arrays for pollinia removal, pollen deposition, or fruit set. For pollinia removal, monomorphic yellow arrays had significantly greater reproductive success, and monomorphic red the least. A tendency for yellow arrays to have higher pollen deposition was also found. We argue that differential population fitness was most likely to reflect differential numbers of pollinators attracted to arrays, through preferential long-distance attraction to arrays with yellow inflorescences. Correlative studies of absolute reproductive success in 52 populations of D. sambucina supported our experimental results. To our knowledge this is the first study to suggest that attraction of a greater number of pollinators to rewardless orchids may be of greater functional importance to population fitness, and thus ecology and conservation, than are the behavioral sequences of individual pollinators.

Experimental investigation of the effect of spatial aggregation on reproductive success in a rewardless orchid.

Plant reproductive success within a patch may depend on plant aggregation through pollinator attraction. For rewardless plants that lack rewards for pollinators, reproductive success may rely strongly on the learning abilities of pollinators. These abilities depend on relative co-flowering rewarding and rewardless plant species spatial distributions. We investigated the effect of aggregation on the reproductive success of a rewardless orchid by setting up 16 arrays in a factorial design with two levels of intraspecific aggregation for both a rewardless orchid and a rewarding co-flowering species. Our results show that increasing aggregation of both species negatively influenced the reproductive success of the rewardless plants. To our knowledge, this is the first experimental study demonstrating negative effects of aggregation on reproductive success of a rewardless species due both to its own spatial aggregation and that of a co-flowering rewarding species. We argue that pollinator learning behaviour is the key driver behind this result.

The evolution of empty flowers revisited.

The evolution of plants that provide no form of reward for their pollinators is puzzling because they receive low numbers of pollinator visits and so have low reproductive success. To predict the evolutionary dynamics of empty morphs within a plant population, we modeled different foraging strategies that pollinators could use to avoid them. We predicted that the optimal strategy was to visit empty inflorescences randomly when these were infrequent but to use strategies such as visiting fewer flowers per inflorescence to avoid wasting time on them. As the frequencies of empty inflorescences increased, discriminating directly against empty morphs was more likely to be an optimal strategy than was avoiding the species altogether and switching to an alternative one. An experimental test of this model using artificial inflorescences showed that bumblebees used a variety of strategies to minimize time wasted on empty inflorescences. They showed weak discrimination against empty inflorescences but switched to an alternative type of inflorescence as the frequency of empty inflorescences increased. We predicted that empty morphs would be at a visitation rate disadvantage even when at low frequencies in a plant population. Differences in outcrossing rates, or male function, may explain how rewardlessness spreads in a plant population.

The potential for floral mimicry in rewardless orchids: an experimental study.

More than one-third of orchid species do not provide their pollinators with either pollen or nectar rewards. Floral mimicry could explain the maintenance of these rewardless orchid species, but most rewardless orchids do not appear to have a rewarding plant that they mimic specifically. We tested the hypothesis that floral mimicry can occur through similarity based on corolla colour alone, using naive bumble-bees foraging on arrays of plants with one rewarding model species, and one rewardless putative mimic species (Dactylorhiza sambucina) which had two colour morphs. We found that when bees were inexperienced, they visited both rewardless morphs randomly. However, after bees had gained experience with the rewarding model, and it was removed from the experiment, bees resampled preferentially the rewardless morph most similar to it in corolla colour. This is the first clear evidence, to our knowledge, that pollinators could select for floral mimicry. We suggest that floral mimicry can be a selective force acting on rewardless orchids, but only under some ecological conditions. In particular, we argue that selection on early-flowering rewardless orchids that receive visits from a large pool of naive pollinators will be weakly influenced by mimicry.