Designing sampling protocols for plant-pollinator interactions - timing, meteorology, flowering variations and failed captures matter

ABSTRACT

Plant-pollinator interactions are key components of ecosystem functioning and are therefore increasingly studied. Of all the approaches used to estimate these interactions, the capture of pollinators along transects is a widely used and recognized method. However, specific choices of sampling design can strongly influence observations of insect visits and bias ecological interpretations. Yet, there is no agreement on the best transect design. Sampling intensity (number and length of transects) is an important element of these choices, but not the only one. Here we investigate the influence of three other facets of protocol choices that commonly arise when designing pollination transects: (i) the influence of sampling conditions that may interact with ecological variables of interest or bias observations, (ii) the measurement of floral availability frequency, and (iii) the management of insects observed but not captured. We quantified the importance of these three protocol choices using a large dataset of 720 plant-pollinator transects in protected wet meadows in France. Our results demonstrate the need to (i) cover a wide range of temporal and meteorological conditions for each site, and (ii) repeat the assessment of plot attractiveness for pollinators (a major covariate, usually simply derived from one-off vegetation surveys). In addition, we show that (iii) for analyses of visitation density among insect groups, failed insect captures should not be discarded but incorporated into the analyses. Overall, this research identifies three key choices in transect design and highlights their influence in our understanding of plant-pollinator interactions.

KEYWORDS:

• Meteorological conditions
• patch attractiveness
• failed captures
• pollination
• insect sampling
• flower visitation
• transect
• phenology

Acknowledgments

We thank D. Lopez-Pinot, C. Birck, R. Perin, B. Bal, and C. Dubosson from ASTER for their help in choosing the study sites and giving us access to them. We are grateful to M. Cario, C. Grange, E. Mesquida, C. Tournier, S. Weil, C. Martinez, F.C. Boucher, and F. Dommanget for their joyful help in the field. Finally, we thank D. Inouye and another anonymous reviewer for their helpful comments that helped improving the manuscript.

Data availability statement

Raw data and derived data supporting the findings of this study are available from the corresponding author MCG on request.

Disclosure statement

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

Author contributions

MCG, LG and FM conceived the study. All authors contributed to the field work. JR created the database, MCG performed the statistical analyses, and MG helped with the figures. MCG wrote the first draft of the manuscript, which was significantly improved by all authors.

Supplementary material

Supplemental data for this article can be accessed here

Additional information

Funding

This research was funded through: the 2017-2018 Belmont Forum and BiodivERsA joint call for research proposals, under the BiodivScen ERA-Net COFUND programme, and with the funding organisation: the French National Research Agency (ANR-18-EBI4000106); the French National Research Agency in the framework of the “Investissements d’avenir” program (ANR-15-IDEX-02); the DIPEE Grenoble-Chambery and the FREE-Alpes Federation (FR n°2001-CNRS); and the CNRS-INEE with the PEPS ECOMOB 2019 program