Interacting effect of ozone with pesticide exposure
As expected (Hayes et al. 2012; Leisner & Ainsworth 2012; Millset al. 2013), ozone levels were negatively correlated to crop pollination. Recent studies have estimated that global agricultural losses due to high ozone levels totalled 79–121 million metric tons in 2000 with global economic losses ranging from $11 to $26 billion (Van Dingenen et al. 2009; Avnery et al. 2011a) and predicted increases of between $17 and $35 billion annually by 2030 (Van Dingenen et al. 2009; Avnery et al. 2011a). Such effects may be partly related to a reduction in pollen germination (Leisner & Ainsworth 2012; Taia et al. 2013; Gillespie et al. 2015). Our results suggest that changes in pollination by insects (due to changes in flower visitation patterns) may also play an important role.
The fact that increasing ozone levels modified the response of crop pollination to pesticide exposure (which turns from positive to negative) may be related with pest control. Farmers widely use pesticides to minimize infestations by pests and protect crops from potential reduction of crop production, both in quality and quantity (Damalas 2009), and hence positive effects of pesticide use on production are expected if pests are more limiting than pollinators to production.
It is however possible, that in more degraded environments, i.e., with a higher level of ozone pollution, the cost/benefit ratio of pesticides on crop production changes. In less intensive landscapes with higher pollinator pool, the negative impact of pesticides on pollinators and these consequences on crop pollination can be compensate by the benefit of pest regulation by pesticide use. However, in highly intensive landscapes, due to scarcity of pollinators which limits pollination and crop production, the negative effects of pesticides on crop pollinators (which are more accentuated under high ozone levels, Fig 3) may outweigh the positive effects on pest reduction on crop production.
The negative relationship between ozone pollution and flower visitor abundance could be due to changes in plant-pollinator communication and flower attractiveness that may affect crop pollinator foraging behaviour. Previous studies have showed that ozone induces changes in availability of floral resources by modifying flowering timing and number of flowers, some plant species being particularly sensitive (Hayes et al. 2012; Leisner & Ainsworth 2012; Mills et al. 2013). Ozone also alters pollinator decision‐making, modifying and reducing the volatile floral scents (Farré‐Armengol et al. 2016; Fuentes et al. 2016; Saunier & Blande 2019; Vanderplancket al. 2021) and damaging pollinators olfactory organs (Dötterlet al. 2016; Vanderplanck et al. 2021).
The fact that the negative effect of pesticide exposure on non-Apis pollinators (Mancini et al. 2019; Walker & Wu 2017; Woodcock et al. 2017) was more accentuated under high ozone concentration (Table S1) could be due to communities being less diverse and/or abundant in regions with high ozone, but also to changes in pollinator assemblages. In more degraded areas (high pesticide exposure, high ozone concentration), crop pollinator communities are dominated by a handful of very dominant widespread species that are more resilient to intensive land use (Kleijn et al. 2015), which often have a more generalist diet and may be more mobile (Biesmeijer et al. 2006; Goulson et al. 2008; Connop et al. 2010). Consequently, in such regions the negative effect of ozone on non-Apis crop pollinators might be less detectable, only under more degraded environment, i.e., under high level of pesticide exposure.
Although the negative impact of pesticides on honey bees is well known (e.g. Mancini et al., 2019; Walker and Wu, 2017; Woodcock et al., 2017; Park et al., 2015; Tosi et al., 2017), we found that pesticide exposure was positively related to honey bee density in crops. This result is probably due to beekeeping management strategies that are likely more frequent in intensive agricultural areas where the demand for colony supply to ensure efficient pollination is high (Garibaldi et al.2017; Rollin & Garibaldi 2019), masking (and even compensating) the negative effects of pesticides. However, the positive relationship between abundance of honey bees in crops and pesticide exposure was lower when ozone concentration increased. This can reflect the negative effect of pesticides on honey bees, decreasing the pollination efficiency and survival of honey bees (Prado et al. 2019), despite the local increase of individuals due to the import of colonies by beekeepers in intensive farming systems.