Investigating pollination strategies in a changing world

Background and context

Studying reproductive strategies of plants is an important topic of research in evolutionary ecology, bringing essential knowledge on their adaptation and diversification (Barrett 2010), as well as on the ecosystem functioning, with potential applications in biodiversity conservation and crop breeding. Pollination, which allows the transfer of pollen from the anther to the stigma of flowering plants, is a central evolutionary process given its importance in the dispersal of genes and their assortment among individuals of a species. Pollen may be dispersed by abiotic factors (i.e. wind or water), but biotic dispersal through interactions with insects (in at least two-thirds of cases) is more commonly involved (Ollerton et al. 2011; Willmer 2011). A variety of questions have thus been tackled in pollination biology, such as floral traits as signals for pollinators, the evolution of pollination syndromes, the specificity of plant–pollinator interactions, and the reconstruction of pollination networks (Baguette et al. 2020; Johnson and Steiner 2000; Willmer 2011; Phillips et al. 2020).

An extensive literature has been devoted to the study of plant resource allocation to floral display and rewards for attracting pollinators (e.g. flower shape and colour, emission of volatile compounds, nectar production, and thermogenesis), to record flower visitation by putative pollinators (e.g. visit duration and frequency, or feeding behaviour in flowers), and to measure pollination efficiency (Schiestl and Schlüter 2009; Willmer 2011; King et al. 2013; Parachnowitsch et al. 2019). Such studies regularly point out that, before investigating the pollination strategy of a plant species, the description of all partners involved in the interaction is a necessary first – albeit sometimes complex – step. Beyond the fundamental interest of these questions to understand the functioning of plant–pollinator interactions, it is now clear that these studies are becoming more and more important from an applied point of view. Indeed, in a context of global change and pollinator decline, research on pollination has become critical to assess the impact of environmental modifications on reproductive strategies of plant species and on the functioning of terrestrial ecosystems (Potts et al. 2010; Burkle and Alarcón 2011; IPBES 2016).

In this special issue, we present a series of eight papers on pollination biology, representative of major questions tackled in this field of research. Various topics are covered, including the description of pollinators, plant traits involved in pollinator interaction, and the impact of global changes on strategies used by plants to face pollen limitation.

Identifying partners involved in insect-mediated pollination

When documenting plant–insect pollinator interactions, one of the first questions that arise concerns the exact identity of pollinators. Behind this apparently simple conceptual question, a methodological challenge consists of adapting approaches to biological models, which either involves specialist or generalist strategies (Johnson and Steiner 2000). The implementation of these actions has certainly been hindered due to the difficult identification of both pollinators and plants in natural habitats. Several web tools were thus recently developed to facilitate this task (e.g. www.IDmyBEE.com for wild bees, the “Syrph the net” database for hoverflies, and https://plantnet.org/ for plants), which allow the work of taxonomic experts to be progressively concentrated on the more difficult groups. However, in a context of global changes, conducting actions to raise the general public awareness and protect plant–pollinator interactions should certainly not wait for complete knowledge of such interactions.

The entomofauna interacting with plants has to be first characterized to identify the efficient pollinators within the high diversity of visiting insects, as 40% of visitors on average are not effectively involved in pollination (King et al. 2013). In this special issue, several studies have tested the role of various insects in plant pollination, revealing specialist or generalist strategies in temperate or tropical plants involving Hymenoptera, Diptera, or Coleoptera partners (Barriault et al. 2021; Carreño-Barrera et al. 2021; Gibernau et al. 2021; Larue et al. 2021; Schatz et al. 2021b). In particular, Larue et al. (2021) experimentally demonstrate the contribution of beetles to chestnut pollination, a tree species that has been long considered as wind-pollinated. In a study focusing on rewardless orchids, for which the frequency of pollinator visits is low, Schatz et al. (2021b) show the necessity to look for the plant-specific pollinator by considering the surrounding community of flowering plants.

Identifying the exact identity of partners within plant/pollinator systems is also a prerequisite before investigating interaction networks at the community level. However, estimating the specialization of plant–pollinator interactions is far from being trivial and depends on the spatial and temporal scales at which these interactions are observed (e.g. Dupont et al. 2009; Ollerton 2017; Metelmann et al. 2020; Schwarz et al. 2020; Resasco et al. 2021). The specific choices of the sampling design may also strongly influence observations of insect visits, biasing in turn ecological interpretations, as shown here by Grange et al. (2021). In the perspective of obtaining an accurate, global picture of these interactions over various habitats, methodological studies that investigate and test fieldwork protocols become a major component of experimental research in pollination.

Evolutionary ecology of plant reproductive strategies

Another major issue addressed by pollination ecologists concerns the selective pressures mediated by pollinators on floral traits (Stebbins 1970; Schiestl and Johnson 2013). Variation in behaviour and morphology among pollinator taxa has been historically proposed as an explaining factor for the astonishing diversity of flowers in angiosperms, starting with the seminal contributions of Charles Darwin in its Origin of Species (Darwin 1859). Since then, studies in pollination biology have repeatedly investigated the various functions of floral traits, which include attractive signals, resources sought by pollinators, usually available within flowers, and co-adaptation with the pollinator(s) body to optimize pollen dispersal (Harder and Johnson 2009). Because different insect orders can be involved in pollination (i.e. mostly Hymenoptera, Lepidoptera, Diptera, and Coleoptera in Europe), and because insects interact with plants in different ways due to different perception of environmental signals (e.g. scent, colour, shape) depending on the order, plant strategies may result in contrasting pollination syndrome according to the partner(s) involved. Although the concept of pollination syndrome (i.e. the suite of floral traits that may be used to predict which insects pollinate a given plant species) has been largely discussed and questioned (Ollerton et al. 2009; Schiestl and Schlüter 2009; Rosas-Guerrero et al. 2014; Phillips et al. 2020), this general view of pollinators partly involved in shaping the diversity of flowers remains a major paradigm in pollination biology.

A multiplicity of approaches is needed to assess such a question satisfactorily. The first logical step obviously involves a detailed description of the floral traits, and a characterisation of their exact function. While some traits are a priori relatively simple to characterise (number of flowers, petal size or surface), others require complex methodologies. This is the case for floral odours, a major object of study in chemical ecology, which can be characterised by Gas Chromatography/Mass Spectrometry (GC-MS) approaches that allow both the identification of the volatile compounds emitted and their relative abundance (Knudsen et al. 1993; Delle-Vedove et al. 2017). For such complex traits, characterisation of function, i.e. identification of the part of the signal that is actually detected by pollinators, also requires fine approaches (e.g. Gas Chromatography/Electroantennographic Detection or GC-EAD; Schneider 1957), which are still relatively little used.

Beyond the description of traits and their functionality, the study of selection pressures on these traits, including those mediated by pollinators, can be carried out using selection gradient approaches (Lande and Arnold 1983; reviewed in Harder and Johnson 2009). The literature in pollination biology is rich in such studies, and more recently, analyses inferring selection pressures not only through the female function but also through the male function of plants have started to be developed (e.g. Wright and Meagher 2004; Sahli and Conner 2011; Austen and Weis 2016). Such studies, requiring the use of molecular markers for paternity analyses, are likely to provide a better picture of the pollinator-mediated evolutionary forces potentially at play in angiosperms. Although such approaches are extremely promising, purely descriptive works are still needed. Compiling floral traits in databases, which are useful in both ecology and evolution for comparative analyses, requires not only a precise description of the traits but also of their potential variation, within individuals (e.g. in correlation with phenology), between individuals and between populations. Several studies in this special issue have focused on characterising different types of traits [i.e. scent (Barriault et al. 2021; Gibernau et al. 2021); colour (Imbert 2021); floral morphology (Carreño-Barrera et al. 2021; Larue et al. 2021; Schatz et al. 2021b)], as well as certain levels of variation [i.e. temporal variation (Gibernau et al. 2021); within and among population variation (Imbert 2021)].

Ecological and evolutionary implications of pollinator decline in a context of global change

Finally, a huge gap in our knowledge of plant and pollinator systems lies in the difficulty to study pollinator decline over the long term (Schatz et al. 2016). Recently developed tools dedicated to the analysis of interaction networks allow investigating their resilience to the decline of involved partners. In such context, Burkle et al. (2013) described the variation of a large plant-pollinator network over a 120-year period and found a considerable decline of both plant and pollinator species involved in this network and degradation of the mutualistic network structure. The current decline in abundance and diversity of pollinator communities is leading to a homogenisation and hence a simplification of plant–pollinator interaction networks that may mainly affect plant species engaged in specialised interactions due to pollen limitation (Zattara and Aizen 2021). Specialised pollination strategies are therefore directly threatened by this homogenisation (Weiner et al. 2014; IPBES 2016; Schatz et al. 2021a). Combining field studies that focus on identifying pollinators associated with given plant species to increase knowledge on pollination strategies and evolution (see above) with global analyses of pollination networks should thus be crucial to define effective conservation strategies.

Potential threat on insect-pollinated plant populations is addressed by two studies of this special issue. First, Cheptou (2021) questions how global change may induce short-term evolution of pollination strategies for evolutionary rescue of some plant species, potentially breaking major trophic links in ecosystems. Second, Schatz et al. (2021b) report field observations on pollinators of the supposedly highly specialized Ophrys genus, and argue how opportunistic interactions may limit the disturbance of natural pollination. The impact of global changes on plant–pollinator interactions is thus still an open question that will undoubtedly be the central matter of many future studies in pollination ecology.

Future prospects

In the context of global change affecting pollinator abundance and diversity as well as homogenising interactions at the expense of specialised ones, there is an urgent need to go beyond species lists and atlases. To enable effective understanding and conservation of plant–pollinator interactions, there is an urgent need to document them in detail through the creation of extensive databases at least on a national scale. The participatory science operation Spipoll has raised a strong awareness of the diversity of these interactions among the different actors involved, and has allowed the development of a database including thousands of observations, although it faces the problem of photograph-based species identification. Databases have been initiated for plants, such as orchids, for which the currently known network of orchid-pollinator interactions has been recorded on a European scale (Joffard et al. 2019). In this context, the GDR Pollinéco research group, federating working on this theme in France, Walloon Belgium and western Switzerland (Romandie), constitutes a unique place in Western Europe where this type of centralisation of all plant–pollinator interaction networks is possible and would be useful to the whole scientific community (Drossart and Gérard 2020; Schatz et al. 2021a). This centralisation of plant–pollinator interaction networks is often one of the first axes of the regional action plans for pollinators currently being deployed, and the first axis of the future governmental plan for pollinators (planned for 2022). All of this knowledge will provide solid grounds to increase our understanding of pollination strategies and the effective conservation of plant–pollinator interactions.

Authors contribution

All authors equally contributed to writing this editorial.

Acknowledgments

We thank Sophie Nadot and Florian Jabbour for their assistance during the preparation of this Special Issue, Michel Baguette for helpful exchanges and the GDR Pollinéco for general scientific exchanges and collaborations.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

BS and MD are supported by the LabEx CeMEB [ANR-10-LABX-04-01]; GB is supported by  the LabEx TULIP [ANR-10-LABX-0041].