Predation of grape berry moths by harvestmen depends on landscape composition

Highlights

• A high proportion of harvestmen consumed grape berry moths, springtails and grape phylloxera.

• The presence of alternative prey did not explain harvestmen predation rates on grape berry moth.

• Harvestmen predation rates on grape berry moths varied with landscape composition.

• Conservation of semi-natural habitats in vineyards may increase biological pest control services delivered by harvestmen.

Abstract

Landscape complexity can benefit natural enemy communities and the biological pest control services they provide in agricultural landscapes. Harvestmen are generalist predators consuming a large range of prey in terrestrial ecosystems including agroecosystems. However, their ecology and their role in controlling pest populations in such ecosystems remain poorly studied. In this study, we examined predator–prey interactions between the European harvestmen (Phalangium opilio L.) and several potential prey species found in a vineyard agroecosystem. We sampled 20 populations of harvestmen in vineyards selected along a gradient of proportion of semi-natural habitats and used gut-content molecular analyses to quantify interaction strength between harvestmen and the grape berry moth, the main insect pest of grape, and two alternative prey species, springtails and the grape phylloxera. We found a high proportion of harvestmen positive to each type of prey with, on average, half of the individuals collected that had consumed grape berry moths. Increasing the proportion of semi-natural habitats in the landscape enhanced the proportion of harvestmen preying upon grape berry moths. Despite a significant number of harvestmen preying on springtails and grape phylloxera, the strength of the feeding interaction between harvestmen and these alternative prey species never significantly explained predation rates of the grape berry moth. Our results indicate that conserving semi-natural habitats increases biological pest control services delivered by harvestmen in vineyard landscapes and highlight the potentially important role of harvestmen in those agricultural systems.

Introduction

Agricultural intensification is a main driver of biodiversity loss in human-modified landscapes (Stoate et al., 2001, Tscharntke et al., 2005). Future agricultural landscapes will need to better balance productivity with minimising negative impacts on the environment (Doré et al., 2011). One promising way to achieve this balance is to design farming systems that replace external inputs, such as agrochemicals, by ecosystem functions and services, such as biological control, generated by biodiversity (Bommarco et al., 2013). To design farming systems that efficiently rely on services provided by beneficial species, we need to considerably improve our understanding of their ecology, particularly on two key aspects: (i) the role of species and functional groups key to functions and services; (ii) the consequences of environmental changes at several spatio-temporal scales on populations and community dynamics (Bommarco et al., 2013, Schellhorn et al., 2015).

Trophic interactions in agroecosystems are affected by several variables acting at different spatio-temporal scales, such as crop management at the field scale or landscape context (Rusch et al., 2010, Tscharntkea et al., 2007). Two comprehensive reviews demonstrated that the proportion of semi-natural habitats in the landscape (i.e., landscape complexity) enhances the abundance and the diversity of natural enemies (Bianchi et al., 2006, Tscharntkea et al., 2007). Moreover, there is some evidence that this positive effect of landscape complexity on natural enemies led to increased predation or parasitism rates of insect pests in agroecosystems (Rusch et al., 2016, Dainese et al., 2019). This positive effect of landscape complexity is due to semi-natural habitats providing several key resources for natural enemies such as alternative host and prey species, nectar, overwintering sites or favourable microclimatic conditions (Rusch et al., 2010, Sarthou et al., 2014). However, very few studies have examined the indirect effects of landscape context on lower trophic levels (i.e., herbivores and plants) mediated by trophic cascades. Moreover, a large majority of studies considers spiders, ground beetles or parasitoids, yet the effects of landscape complexity remain largely unexplored for several other groups of arthropods, such as staphylinids, earwigs or harvestmen.

Recent works have shown that the taxonomic richness of service-providing organisms contribute to supporting biological pest control service in agriculture (Dainese et al., 2019). Species-rich communities tend to increase the level of predation owing to niche complementarity between natural enemies, sampling effect or even facilitation (Cardinale et al., 2003, Letourneau et al., 2009, Greenop et al., 2018). However, negative or neutral relationships between more diverse natural enemy communities and the level of pest control emerge in a non-negligible number of cases (approximately 30% in the meta-analysis of Letourneau et al., 2009). These relationships may arise from negative interactions between species (e.g., intraguild predation, behavioral interference) or sampling effect (Letourneau et al., 2009). Much of the unexplained variation in studies investigating how the structure of natural enemy communities affect pest control services comes from the lack of basic knowledge about the ecology, behavior and diet of a large number of species groups. Producing such knowledge about the role of key predator species in real-world agroecosystems and their impact on pest populations and crop damage is of major importance if we are to develop farming systems integrating natural pest control services (Welch et al., 2014, Schellhorn et al., 2015).

Arthropod predation is generally difficult to observe and estimate, especially in real-world ecosystems. One way to study trophic interactions is to analyze gut contents of field-collected predators (Kuusk et al., 2008, Sheppard and Harwood, 2005, Birkhofer et al., 2017). Several studies have demonstrated the added value of using molecular tools to detect prey-specific DNA sequences within the gut contents of predators and to reveal feeding links in complex food webs (Davey et al., 2013, Kuusk et al., 2008). However, these approaches are not widely used in agroecosystems and a very limited number of studies have used them to study how trophic interactions respond to environmental changes, such as changes in landscape context or predator diversity (but see Roubinet et al., 2015). Moreover, Harwood and Obrycki (2005) revealed that in over 100 studies that have used some sort of gut content analysis, more than 70 of them studied carabid beetles. This clearly highlights the need for more studies on other taxa.

Harvestmen (Opiliones), the third largest order of the class Arachnida that encompasses more than 6000 species are considered as important predators in many terrestrial ecosystems (Curtis and Machado, 2007, Pinto-da-Rocha et al., 2007 and references therein). Several studies have shown that harvestmen consume a large range of prey including collembolans, aphids, lepidopteran larvae, dipterans, ants, spiders, mites, earthworms or gastropods in arable land (Wolff et al., 2014, Leathwick and Winterbourn, 1984, Dixon and McKinlay, 1989, Clark et al., 1994, Halaj and Cady, 2000, Acosta and Machado, 2007). However, even if they are usually found in arable land at high abundance, their ecology and their role in controlling pest populations in such habitats remains poorly studied (Acosta and Machado, 2007).

In this study, we examined predator-prey interactions between the European harvestmen (Phalangium opilio L.) and several potential prey species found in a vineyard agroecosystem. We decided to study the European harvestmen in vineyards because they are found at high level of abundance and that no basic ecological understanding of their role in this agroecosystem exists to our knowledge (Muneret et al., 2019). Moreover, developing conservation biological control of pests is a major applied issue in vineyards because vineyards are submitted to very high level of pesticide use (Muneret et al., 2018). In this study, our aims were (i) to quantify the strength of trophic links between the European harvestmen and potential prey species, including grape berry moth, using molecular gut content analysis, (ii) to examine how population density of harvestmen and potential alternative prey species (grape phylloxera and springtails) affect interaction strengths between harvestmen and grape berry moth populations, (iii) and how the proportion of semi-natural habitats in the landscape context modify these trophic interactions. We particularly hypothesized that the proportion of semi-natural habitats in the landscape would have a direct positive effect on the level of predation of grape berry moth by harvestmen and an indirect negative effect on the level of pest infestation. Moreover, we hypothesized a negative effect of the interaction strengths between harvestmen and their alternative prey species on the level of predation of grape berry moths.