Contrasting impacts of pesticides on butterflies and bumblebees in private gardens in France

Highlights

• Pesticide use in private gardens has a measurable impact on flower-visiting insects.

• Insecticide and herbicide have a negative impact on butterflies and bumblebees.

• Fungicide, snail pellets and Bordeaux mixture have a positive impact.

• The impact varies with landscape characteristics, i.e. urbanization.

• Citizen science based on data provided by non-specialists offers a powerful tool to monitor management impact at nationwide scale.

Abstract

Private gardens are an important food source and refuge for animals in urban areas because they represent a large part of the green space. It has been shown that garden management regime (water use, floral composition) may impact the species they shelter. However, due to access restrictions, lack of regulations and the difficulty of data collection on private property, the impact of management practices and in particular pesticide use has seldom been assessed in private gardens. Using data collected in the framework of a nationwide participatory monitoring scheme in France, we assess here, for the first time, the effect of private garden management on two important groups of flower-visiting insects, i.e. butterflies and bumblebees, at a large scale. We show that the correlation between butterfly and bumblebee abundance and use of insecticides and herbicides is negative, whereas the use of Bordeaux mixture (fungicide approved for organic use), fungicides and anti-slugs is positively correlated with butterfly and bumblebee abundance. We hypothesize that herbicides have an indirect negative impact on insects by limiting the amount of available resources, and that the Bordeaux mixture, fungicides and slug repellants have an indirect positive impact on these insects by fostering healthier plants, probably offering higher level of resources to pollinators. Moreover, we show that the impact of pesticides varies according to the landscape, the negative effect of insecticides being more important in highly urbanized areas. Overall, our results show that gardener practices can have a positive impact on flower-visiting insects, even in a highly anthropized, urban landscape.

Introduction

Private gardens represent an important part of green spaces in cities, e.g. 23% in Sheffield (UK), (Gaston et al., 2005), or 36% in Dunedin, New-Zealand (Mathieu et al., 2007). Representing nature oases in cities, green spaces are known to positively influence human health and wellbeing (Fuller et al., 2007, Gross and Lane, 2007, Gaston et al., 2007). Furthermore, it has been suggested that private gardens might mitigate the impact of urbanization on biodiversity (Goddard et al., 2010). Even if each garden taken individually is too small to be of biological importance, gardens taken as a whole can be an important component of urban floristic diversity (Thompson et al., 2003, Smith et al., 2006b, Loram et al., 2008, Stewart et al., 2009) and provide important sources of food and shelter for birds (Cannon et al., 2005, Davies et al., 2009), wild bees (Fetridge et al., 2008, Samnegård et al., 2011) and amphibians (Gaston et al., 2005). Private gardens can also provide landscape connectivity for plants and animals (Rudd et al., 2002, Sperling and Lortie, 2010, Vergnes et al., 2012, Vergnes et al., 2013). However, they may also have a negative impact on the environment: for instance, Dehnen-Schmutz et al. (2007) and Marco et al. (2010) have shown that ornamental plants cultivated in private gardens could be an important vector of plant invasions. Assessing the role of private gardens in maintaining urban biodiversity still requires an understanding of the factors driving the biodiversity hosted within these private areas.

Landscape and local scale factors may impact urban biodiversity. Pardee and Philpott (2014) showed that presence of native plants in gardens but also landscape characteristics, such as amount of semi-natural area in the landscape, influence urban bee diversity. Similar results were shown for British moths (Bates et al., 2013). Furthermore, Bergerot et al. (2011) showed that the level of urbanization in the landscape surrounding private gardens was a strong driver of the diversity and composition of butterfly communities in gardens, with lower species richness and lower occurrence of feeding specialists in strongly urbanized sites. On the other hand, Smith et al. (2006a), found that the extent of green space around gardens only occasionally explained the abundance of 22 invertebrate groups, and that most variables correlated with abundance occurred at the scale of the garden itself. These seemingly contradictory results might arise from temporal and spatial variability, or from a lack of power of the analysis performed. Untangling the local and landscape effects on insect diversity in gardens might require larger datasets encompassing various garden types and levels of urbanization.

Another difficulty of studying private gardens is that they are unregulated habitats with various water and chemical use intensity and vegetation structure. Moreover, these characteristics are generally unknown, depending on each gardener’s own decisions (Mathieu et al., 2007). Although the effect of management practices on private gardens has been little studied, it has been shown that increased pesticide use on residential yards may negatively impact the environment (Robbins et al., 2001). Direct effects on species abundance in private gardens have seldom been studied, but available results suggest it could be important, especially because their use in gardens is unregulated and the amount private gardeners use may be significant. Smith et al. (2006a) included pesticide use in their study of invertebrates in urban gardens at a city scale, but this factor was pooled in a global management intensity index including several variables, such as weeding, pruning, watering or bird feeding. Such an aggregated index of management intensity makes it difficult to identify the components that most affect biodiversity. More specifically, Byrne and Bruns (2004) and Cheng et al. (2008) have revealed the negative impact of pesticides on non-target soil microfauna, whereas Politi Bertoncini et al. (2012) have shown it on floristic composition, and Stewart et al. (2009) found a negative correlation between lawn management intensity, including use of phytochemicals, and the presence of various plant species in urban lawns. There are a few citizen-science studies that have investigated bumblebees or Lepidoptera in private gardens (e.g. Lye et al., 2012, Bates et al., 2013); however, to our knowledge, the impact of pesticide use on biodiversity, and especially flower-visiting insects, has never been studied in private gardens at a large scale and in different landscape contexts.

Restricted access to private gardens and the difficulty of data collection on biodiversity and management in this habitat probably accounts for the paucity of research on this topic. When collecting data in private gardens, citizen science is an efficient tool because garden owners can directly provide the data (Cooper et al., 2007). Based on a nationwide citizen survey on private gardens in France, we assess here the relative impact of local scale factors (i.e. garden structure and management) and landscape composition (i.e. proportion of urban area) on two groups of pollinating insects, butterflies and bumblebees. We specifically measured the impact of pesticides on these insects, depending on the type of pesticide (e.g. herbicide, molluscicide, insecticide), and quantified this impact relatively to other factors, such as garden characteristics and urbanization level. We hypothesized that gardening practices would have a larger impact on insect abundance in densely urbanized districts than in more rural districts.