Background: Plants use floral displays to attract mutualists, but simultaneously need to prevent attacks by antagonists. Chemical displays detectable from a distance include attractive or repellent floral volatile organic compounds (FVOCs). Post-landing, visitors perceive contact chemicals including nutrients, but also deterrent or toxic constituents in pollen and nectar, protecting flowers from overexploitation. The composition of FVOCs and pollen chemistry is well known to vary among and within species. However, we lack knowledge about differences and similarities in the detectability of and behavioural responses towards these compounds for insect flower visitor groups of key importance, i.e., mutualistic pollinators versus antagonistic florivores, as well as the correlation between FVOCs and pollen chemodiversity. Scope: We reviewed how FVOCs and non-volatile floral chemical displays, i.e., nutrients and toxins of pollen, vary in composition and how they affect the detection and behaviour of insect flower visitors. Moreover, we used a meta-analytic approach to evaluate the detection of and responses towards FVOCs by pollinators vs. florivores within the same plant genera. Furthermore, we tested whether the chemodiversity of FVOCs as well as nutrients and toxins stored in pollen are correlated and hence informative about each other. Key Results: According to the available data, florivores are more likely to detect FVOCs than pollinators. Common FVOCs such as linalool and methyl salicylate were often reported as attractive to pollinators and repellent towards florivores. A higher number of FVOCs was found to be attractive to both mutualists and antagonists compared to shared repellent compounds. Furthermore, a negative correlation between FVOC richness and the number of pollen toxin classes was revealed, besides a trend towards a positive correlation between pollen protein amount and the number of pollen toxins. Conclusions: Plants face critical trade-offs when producing floral chemicals, as these partly mediate the same information, particularly attraction but also repellence or toxicity, to both mutualists and antagonists. Moreover, chemodiversity of different floral parts is partly correlated and thus highly relevant for investigations of flower-insect interactions. Further research is needed on more different wild and cultivated plant species and mutualistic and antagonistic interaction partners to test for generalisation of these patterns.

Background Plants often use floral displays to attract mutualists and prevent antagonist attacks. Chemical displays detectable from a distance include attractive or repellent floral volatile organic compounds (FVOCs). Locally, visitors perceive contact chemicals including nutrients but also deterrent or toxic constituents of pollen and nectar. FVOC and pollen chemical composition may vary intra- and interspecifically. For certain pollinator and florivore species, responses towards these compounds are studied in specific plant systems; yet we lack a synthesis of general patterns comparing these two groups and insights into potential correlations between FVOC and pollen chemodiversity. Scope We reviewed how FVOCs and non-volatile floral chemical displays, i.e., pollen nutrients and toxins, vary in composition and affect the detection and behaviour of insect visitors. Moreover, we used meta-analyses to evaluate the detection of and responses towards FVOCs by pollinators vs. florivores within the same plant genera. We also tested whether the chemodiversity of FVOCs, pollen nutrients and toxins are correlated and hence mutually informative. Key Results According to available data, florivores could detect more FVOCs than pollinators. Frequently tested FVOCs were often reported as pollinator-attractive and florivore-repellent. Among FVOCs tested on both visitor groups, there were a higher number of attractive than repellent compounds. FVOC and pollen toxin richness correlated negatively, indicating trade-offs, while a marginal positive correlation between pollen protein amount and toxin richness was observed. Conclusions Plants face critical trade-offs, as floral chemicals mediate similar information to both mutualists and antagonists, particularly through attractive, and fewer repellent, FVOCs. Furthermore, florivores may detect more FVOCs, whose richness is correlated to reward chemical richness. FVOC chemodiversity is potentially informative of reward traits. To better understand the ecological processes shaping floral chemical displays, more research is needed on floral antagonists of diverse plant species, and on the role of floral chemodiversity towards visitor responses.

Plants produce a great number of phytochemical compounds mediating a variety of different functions. Recently, phytochemical diversity (chemodiversity), a way which to quantify the complex phenotype formed by sets of phytochemicals, has been suggested to be important for function. However, no study has systematically examined the potential (in)direct functional importance of chemodiversity on a general level, partly due to a lack of an agreement on how to quantify this aspect of the plant phenotype. This paper has four aims: 1) We discuss how chemodiversity (deconstructed into components of richness, evenness and disparity) may quantify different aspects of the phenotype that are ecologically relevant. 2) We systematically review the literature on chemodiversity to examine methodological practices, explore ecological patterns of variability in diversity across different levels of biological organization, and investigate the functional role of this diversity in interactions between plants and other organisms. 3) We provide a framework facilitating decisions on which measure of chemodiversity is best used in different contexts. 4) We outline open questions and avenues for future research in this area. A more thorough understanding of phytochemical diversity will increase our knowledge on the functional role phytochemical compounds, and how they shape ecological interactions between plants and their environment.

Introduction: Assessing intraspecific variation in plant volatile organic compounds (VOCs) involves pitfalls that may bias biological interpretation, particularly when several laboratories collaborate on joint projects. Comparative, inter-laboratory ring trials can inform on the reproducibility of such analyses. Objectives: In a ring trial involving five laboratories, we investigated the reproducibility of VOC collections with polydimethylsiloxane (PDMS) and analyses by thermal desorption-gas chromatography-mass spectrometry (TD-GC-MS). As model plant we used Tanacetum vulgare, which shows a remarkable diversity in terpenoids, forming so-called chemotypes. We performed our ring-trial with two chemotypes to examine the sources of technical variation in plant VOC measurements during pre-analytical, analytical, and post-analytical steps. Methods: Monoclonal root cuttings were generated in one laboratory and distributed to five laboratories, in which plants were grown under laboratory-specific conditions. VOCs were collected on PDMS tubes from all plants before and after a jasmonic acid (JA) treatment. Thereafter, each laboratory (donors) sent a subset of tubes to four of the other laboratories (recipients), which performed TD-GC-MS with their own established procedures. Results: Chemotype-specific differences in VOC profiles were detected but with an overall high variation both across donor and recipient laboratories. JA-induced changes in VOC profiles were not reproducible. Laboratory-specific growth conditions led to phenotypic variation that affected the resulting VOC profiles. Conclusion: Our ring trial shows that despite large efforts to standardise each VOC measurement step, the outcomes differed both qualitatively and quantitatively. Our results reveal sources of variation in plant VOC research and may help to avoid systematic errors in similar experiments.

The chemical composition of plant individuals can vary, leading to high intraspecific chemodiversity. Diversity of floral chemistry may impact the responses of flower-feeding insect visitors. Plants of Tanacetum vulgare vary significantly in their leaf terpenoid composition, resulting in distinct chemotypes. We investigated the composition of terpenoids and nutritents of flower heads and pollen in plants belonging to three chemotypes, dominated either by β-thujone (BThu), artemisia ketone (Keto) or a mixture of (Z)-myroxide, santolina triene and artemisyl acetate (Myrox). Moreover, we tested the effects of these differences on preferences, weight gain and performance of adults of the shining flower beetle, Olibrus aeneus. The terpenoid composition and diversity of flower heads and pollen significantly differed among individuals belonging to these chemotypes, while total concentrations of terpenoids, sugars, amino acids or lipids of the pollen did not differ. Beetles preferred the BThu over the Myrox chemotype in both olfactory and contact choice assays, while the Keto chemotype was marginally repellent in olfactory assays. The beetles gained the least weight within 48 h and their initial mortality was highest when feeding exclusively on floral tissues of the Myrox chemotype. Short-term weight gain and long-term performance were highest on the BThu chemotype. In conclusion, the beetles showed chemotype-specific responses towards different T. vulgare chemotypes, which may be attributed to the terpenoid composition in flower heads and pollen rather than to differences in nutrient profiles. Both richness and overall diversity are important factors when determining chemodiversity of individual plants and their consequences on interacting insects.