Impact of a synthetic fungicide (fosetyl-Al and propamocarb-hydrochloride) and a biopesticide (Clonostachys rosea) on soil bacterial, fungal, and protist communities

Highlights

• The impact of a synthetic pesticide and a biopesticide on soil microbes was tested.

• Impact on microbial α-diversity was weak, transient and compound-specific.

• Both pesticides altered co-occurrences of microbial taxa and community composition.

• Both the biopesticide and the synthetic pesticide can affect soil microorganisms.

• Measures of α-diversity are insufficient to evaluate the impact of pesticide use.

Abstract

The use of synthetic pesticides in agriculture is increasingly debated. However, few studies have compared the impact of synthetic pesticides and alternative biopesticides on non-target soil microorganisms playing a central role in soil functioning.

We conducted a mesocosm experiment and used high-throughput amplicon sequencing to test the impact of a fungal biopesticide and a synthetic fungicide on the diversity, the taxonomic and functional compositions, and co-occurrence patterns of soil bacterial, fungal and protist communities.

Neither the synthetic pesticide nor the biopesticide had a significant effect on microbial α-diversity. However, both types of pesticides decreased the complexity of the soil microbial network. The two pesticides had contrasting impacts on the composition of microbial communities and the identity of key taxa as revealed by microbial network analyses. The biopesticide impacted keystone taxa that structured the soil microbial network. The synthetic pesticide modified biotic interactions favouring taxa that are less efficient at degrading organic compounds. This suggests that the biopesticides and the synthetic pesticide have different impact on soil functioning.

Altogether, our study shows that pest management products may have functionally significant impacts on the soil microbiome even if microbial α-diversity is unaffected. It also illustrates the potential of high-throughput sequencing analyses to improve the ecotoxicological risk assessment of pesticides on non-target soil microorganisms.

Introduction

Despite a worldwide drive toward sustainable agricultural production, pest management strongly relies on the use of different types of pesticides. For example, in Europe, around 380,000 tons of synthetic and inorganic pesticides (average between 2011 and 2017 for 28 European countries) are sold per year (Chiaia-Hernandez et al., 2017; Eurostat). While pesticides are often presented as essential for food safety (Cooper and Dobson, 2007), their use remains a controversial issue at the regulatory forefront in most countries (Guedes et al., 2016). Indeed, pesticides may also affect non-target organisms (Imfeld and Vuilleumier, 2012; Pisa et al., 2015), contaminate soil and water (Vryzas, 2018), and constitute a risk for human health (Alavanja et al., 2004; Bonmatin et al., 2015; Cimino et al., 2017). Pesticides and their transformation products may persist in the environment (Prashar and Shah, 2016) and reach nearby ecosystems (Bonmatin et al., 2015; Daly et al., 2007). Pesticides may impact local biodiversity including wild plants and animals (Geiger et al., 2010) as well as aquatic organisms (Beketov et al., 2013). In addition, pesticide degradation can lead to the production of harmful metabolites (Machado et al., 2017). Among pesticides, biocontrol agents (one of several categories of biopesticides) rely on the use of living organisms to suppress directly or indirectly the population density - and thus the impact - of a specific pest organism (Eilenberg, 2006). Because of their natural origin and their specificity, they are often assumed to be safer for humans and the environment than chemical control products. Yet, manipulating microbial biodiversity can affect the population size of particular taxa, its interactions with other taxa, and therefore the structure of microbial communities. Although biopesticides may therefore also have adverse impact on soils, very few studies have compared the impact of synthetic pesticides vs biopesticides on soil microbial communities.

Understanding the impact of pesticides and biopesticides on soil microbial communities is of primary importance given the central role of microbes in the functioning of soil ecosystems. Soil microbial communities are highly diverse. They are key actors of biogeochemical cycling (van der Heijden and Wagg, 2013), recycling of wastes, and detoxification of environmental pollutants (Aislabie et al., 2013). In agricultural soils, microbes play a fundamental role for fertility and productivity (Barrios, 2007; de Vries et al., 2013) through their contribution to plant productivity and diversity (van der Heijden et al., 2008), plant-plant interactions (Bever et al., 2010), nitrogen cycling (Hayatsu et al., 2008), carbon cycling (Schimel and Schaeffer, 2012), and soil formation (Rillig and Mummey, 2006). Negative impact of pesticides on soil microbes can thus decrease soil fertility and crop production (Muñoz-Leoz et al., 2011).

However, we currently lack an understanding of how pest management alters soil microbial diversity and community composition, how this impact differs among taxonomic groups and taxa, and how such changes might affect biotic interactions. The effects of pest management on soil microbial communities have mostly been investigated by indirect or culture-dependent methods including measurements of microbial carbon and nitrogen, counting of colony forming units, general microbial activity measurements such as enzymatic activity or soil respiration, and fingerprinting methods such as PLFA or DGGE (Bünemann et al., 2006). In general, the results obtained with such approaches suggest a minor, although dose or exposure dependent, impact of pesticides on soil microbial communities, followed by fast community recovery (Rousidou et al., 2013; Wainwright, 1978; Wardle and Parkinson, 1990). However, a major limitation of these approaches is that they only cover a minor fraction of the total soil microbial diversity and/or do not consider the response of individual taxa and the impacts on biotic interactions. As such, they likely miss fundamental aspects of microbial responses and adaptation to pesticides, including key specific soil functions mediated by some microbial groups.

Methodological developments within the past few years have opened novel opportunities to evaluate pesticide effects in complex soil environments (Jacobsen and Hjelmsø, 2014). Indeed, the developments of DNA sequencing technologies such as metabarcoding allows the automated identification of multiple microbial taxa from a single soil sample (Taberlet et al., 2012). This has dramatically increased the amount and precision of data about soil microbes in general. It also unveiled a large diversity of mostly unknown soil microbial taxa (Bates et al., 2013; Buée et al., 2009). Such methods are particularly well-suited to assess the response of individual taxa to low-level exposure to pesticides which likely represents the most common type of soil contamination by pesticides (Imfeld and Vuilleumier, 2012). For instance, pesticide applications can affect taxa playing a central role in the microbial food web (hereafter keystone taxa) and/or favour some taxa that would otherwise be absent or found in low abundance in the soil ecosystem (indicator taxa). To date, relatively few studies have exploited the potential of soil microbial metabarcoding data to assess the impact of pesticides on soil diversity. For example, Gallego et al. (2019) showed an effect of oxamyl on bacterial community diversity and composition that could only be detected with the RNA-based High-Throughput Sequencing (HTS) analysis. Overall, HTS studies reveal more pronounced effects of pesticides than those based on classical methods, illustrating the higher sensitivity of this method. However, few HTS-based studies have focused on the impact of biopesticides on soil microbial diversity. And, studies comparing the effect of biopesticides and chemical pesticides on the whole microbial diversity are missing so far. This is particularly important as biopesticides are assumed to reduce the environmental impact of agricultural pest management and contribute to more sustainable agricultural practices.

Here, we used a HTS/metabarcoding approach to compare the impact of a microbial biopesticide (PRESTOP® a concentrate of Clonostachys rosea f. catenulata) and a synthetic fungicide (PREVICUR® containing fosetyl-Al and propamocarb-hydrochloride) on the diversity, taxonomic and functional composition, and co-occurrence patterns of soil microbial communities (bacteria, fungi, and protists) in soil mesocosms. We also identified, for each treatment, indicator taxa based on occurrence and frequency, and keystone taxa which had a disproportionate importance for the structure of the microbial co-occurrence network. Indicator and keystone taxa can reveal changes induced by the biopesticide and/or the synthetic pesticide in functions fulfilled by microbial communities (Banerjee et al., 2018; Berry and Widder, 2014; Shi et al., 2016).