Estimating pollination potential in orchards: looking for a simple field method
Océane Bartholomée  1, *@  , Amandine Aullo  2@  , Juliette Becquet  3@  , Lavorel Sandra  4@  
1 : Laboratoire dÉcologie Alpine  (LECA)  -  Website
Université Joseph Fourier - Grenoble 1, Université Savoie Mont Blanc, Centre National de la Recherche Scientifique : UMR5553
bat. D - Biologie 2233 Rue de la piscine - BP 53 38041 GRENOBLE CEDEX 9 -  France
2 : Université de Montpellier  (UM)  -  Website
Faculté des Sciences
163 rue Auguste Broussonnet - 34090 Montpellier -  France
3 : Université Savoie Mont Blanc  (USMB [Université de Savoie] [Université de Chambéry])  -  Website
Université Savoie Mont Blanc, LCME, F-73000, Chambéry-France
27, rue Marcoz - 73000 Chambéry -  France
4 : LECA, Univ. Grenoble Alpes, Univ. Savoie Mont-Blanc, CNRS, F-38000 Grenoble, France
Université Grenoble Alpes, CNRS : UMR5553, Université Savoie Mont Blanc, LCME, F-73000, Chambéry-France
* : Corresponding author

Pollination is a critical ecosystem service due to its essential role in sustaining food production, and the recent observed decline of pollinating insects worldwide. However estimating the pollination service is complicated by the existence of multiple definitions and associated estimation methods. Here, based on a systematic bibliographic review we focused on methods for estimating the potential pollination allowed by the presence of pollinators. We tested a protocol including direct estimation of pollinator abundance and richness, and indirect proxies based on landscape context and orchard feeding and nesting resources. We aimed to identify the subset of variables accounting for potential pollination from insect presence, from which a simplified and standardized method could be designed for environmental impact studies. The method comprised 3 steps: a land cover analysis, field observations of plot-scale resources and insect focal observations. We worked in orchards of the Grenoble region during two consecutive years, sampling 8 orchards in 2017 and 31 in 2018. Mixed linear models with orchard identity as random effect showed that pollinator abundance was significantly impacted by landscape composition within a 2-km radius. For 2017, it decreased with increasing built-up land. For 2018, it increased with increasing grassland area. For both years, floral resources of orchard ground layer, and specifically richness in flower shapes, favoured pollinator abundance. Conversely orchard edge features had only weak marginal impacts on pollinator abundance or richness. For 2017, a combined model with built-up area and an indicator of orchard management intensity explained 42% of variation in pollinator abundance and 50% of variation in pollinator richness. For 2018, a combined model with grassland area and flower shape richness explained only 23% of variation in pollinator abundance. Consequently, a protocol simplification excluding direct observations of pollinators does not seem yet possible. Further work could explore potential influences of orchard management practices.


Online user: 2