Summary• Here, we evaluate the role of pollen limitation and selfing in the maintenance of labile sex expression in subdioecious plant species.• We used a literature survey to explore which factors correlated with a significant occurrence of hermaphrodites in dioecious species. We developed models to explore the selective maintenance of labile sex expression. The models had similar ecological assumptions but differed in the genetic basis of sex lability.• We found that a significant frequency of hermaphrodites was associated with animal pollination, and that hermaphrodites were 'inconstant' males with perfect flowers, suggesting evolution through the gynodioecious pathway. Models showed that a modifier converting pure males into inconstant males could be maintained under a wide range of reduction in both male and female fitness. Pollen limitation and self-fertilization facilitated invasion of the modifier. Depending on the genetics of sex determination, we found pure dioecy, stable subdioecy (trioecy), and situations where inconstant males coexisted with either pure females or pure males. Under selfing and pollen limitation, certain conditions selected for inconstant males which will drive populations to extinction.• We discuss our results in relation to the evolution towards, and the breakdown of, dioecy, and the ecological and evolutionary implications of labile sex expression.

Chemical communication is ubiquitous. The identification of conserved structural elements in visual and acoustic communication is well established, but comparable information on chemical communication displays (CCDs) is lacking. We assessed the phenotypic integration of CCDs in a meta-analysis to characterize patterns of covariation in CCDs and identified functional or biosynthetically constrained modules. Poorly integrated plant CCDs (i.e. low covariation between scent compounds) support the notion that plants often utilize one or few key compounds to repel antagonists or to attract pollinators and enemies of herbivores. Animal CCDs (mostly insect pheromones) were usually more integrated than those of plants (i.e. stronger covariation), suggesting that animals communicate via fixed proportions among compounds. Both plant and animal CCDs were composed of modules, which are groups of strongly covarying compounds. Biosynthetic similarity of compounds revealed biosynthetic constraints in the covariation patterns of plant CCDs. We provide a novel perspective on chemical communication and a basis for future investigations on structural properties of CCDs. This will facilitate identifying modules and biosynthetic constraints that may affect the outcome of selection and thus provide a predictive framework for evolutionary trajectories of CCDs in plants and animals.

Local modification of the soil environment by individual plants may affect the performance and composition of associated plant species. The aromatic plant Thymus vulgaris has the potential to modify the soil through leaching of water-soluble compounds from leaves and litter decomposition. In southern France, six different thyme chemotypes can be distinguished based on the dominant monoterpene in the essential oil, which is either phenolic or non-phenolic in structure. We examine how soils from within and away from thyme patches in sites dominated by either phenolic or non-phenolic chemotypes affect germination, growth and reproduction of the associated grass species Bromus erectus. To do so, we collected seeds of B. erectus from three phenolic and three non-phenolic sites. Seeds and seedlings were grown on soils from these sites in a reciprocal transplant type experiment in the glasshouse. Brome of non-phenolic origin performed significantly better on its home soil than on soil from a different non-phenolic or a phenolic site. This response to local chemotypes was only observed on soil collected directly underneath thyme plants and not on soil in the same site (<5 m away) but where no thyme plants were present. This is preliminary evidence that brome plants show an adaptive response to soil modifications mediated by the local thyme chemotypes. Reproductive effort was consistently higher in brome of phenolic origin than in brome of non-phenolic origin (on both thyme- and grass-soil), indicating that life-history variation may be related to environmental factors which also contribute to the spatial differentiation of thyme chemotypes. Moreover, we found that brome growing on thyme-soil in general was heavier than when growing on grass-soil, regardless of the origin of the brome plants. This is concordant with thyme-soil containing higher amounts of organic matter and nitrogen than grass-soil. Our results indicate that patterns of genetic differentiation and local adaptation may modify competitive interactions and possible facilitation effects in natural communities.

Using a sample of 1,403 flowering plant species, we tested the hypothesis that flower openness and flower‐visitor generalization level of a plant species correlate positively. The “flower‐visitor generalization level” Ln of a flowering plant species n, here defined as number of flower‐visiting animal species attracted to the flowers of n in a given study site, varied enormously among plant species. Its frequency distribution was extremely skewed. Within a study site, L also increased with number of flower‐visitor species A. In order to correct for this, we expressed L relatively, as the proportion of the total flower‐visitor fauna in a study site that visited a given plant species (relative generalization level, L/A). We listed the top‐10 most generalized species (both according to L and L/A) in the “world”, i.e., out of our sample of 1,403 plant species. Flower openness is defined as accessibility to the interior of the flower. We placed the blossom classes of Fægri & van der Pijl along a gradient, albeit not very well defined, of decreasing flower openness (dish‐bowl, bell‐funnel, head‐brush, tube, gullet, flag) and tested for any relationship to their generalization level. The classes differed slightly but significantly in their level of L/A. Tube, bell‐funnel, and dish‐bowl had the highest generalization level and flag, gullet, and head‐brush the lowest. Thus, flower openness and generalization level were not correlated. We discuss other factors influencing generalization level such as accessibility to pollen and nectar, morphology and behavior of visitor, and species diversity of the different functional types of visitors.

Thymus vulgaris has a chemical polymorphism with six different chemotypes that show marked spatial segregation in nature. Although some populations have a single chemotype in majority, many have two or three chemotypes. In this study we analyze the quantitative variation among T. vulgaris populations in the percentage of oil composed of the dominant monoterpene(s) for each chemotype. In general, phenolic chemotypes (thymol and carvacrol), which occur at the end of the biosynthetic chain, have a significantly lower proportion of their oil composed of their dominant monoterpene than nonphenolic chemotypes (geraniol, alpha-terpineol, and linalool). This is due to the presence of high amounts of precursors (gamma-terpinene and paracymene) in the oil of phenolic chemotypes. The essential oil of the nonphenolic thuyanol chemotype has four characteristic monoterpenes that together make up a lower proportion of the oil than the single dominant monoterpene of the other nonphenolic chemotypes. For all chemotypes, the percentage composition of the dominant monoterpene decreased significantly at sites where the chemotype is not the majority type. This decrease is correlated with a significant increase in either the proportion of the two precursors for the thymol chemotype or the monoterpenes characteristic of the other chemotypes at the site. The latter result suggests that a plant with dominant genes is responsible for the production of different monoterpenes can produce several molecules.