Contents lists available at ScienceDirect Agriculture, Ecosystems and Environment journal homepage: www.elsevier.com/locate/agee Overwintering of pollen beetles and their predators in oilseed rape and semi- natural habitats Louis Sutter a, ⁎ , Michael Amato b , Philippe Jeanneret a , Matthias Albrecht a a Agroecology and Environment, Agroscope, Reckenholzstrasse 191, CH-8046, Zürich, Switzerland b Center for Environmental Management Military Lands, Colorado State University, 631 South Mason St., Fort Collins, CO 80523, United States ARTICLE INFO Keywords: Conservation Biological control Meligethes Brassica napus Carabidae Functional biodiversity Habitat management Predators Semi-Natural habitat ABSTRACT Semi-natural habitats (SNH) maintain high levels of biodiversity and ecosystem functioning within agroeco- systems. While management recommendations for SNH to promote biodiversity and optimize natural pest control, have mostly focused on improving food resources for predators, overwintering conditions have hitherto received less attention. The success of conservation biological control is often hampered by a lack of knowledge on how different habitats and their associated traits drive the overwintering of both, crop pests and their natural enemies. Moreover, there are concerns that SNH may act as reservoirs for crop pests. We investigated the overwintering of pollen beetles (Meligethes spp.) and their predators (predatory ground beetles) across 40 habitats of all major types of SNH as well as 8 winter oilseed rape (WOSR) fields, to identify the importance of type and key traits of habitats. Overwintering of pollen beetles was higher in WOSR fields (7 individuals per 0.25 m 2 ) and forest edges (4) compared to flower strips (2) or forest interiors (1). Similarly, overwintering of predatory ground beetles was higher in forest edges and WOSR fields (11 and 10 individuals, respectively) than in flower strips, meadows or hedgerows (4, 4 and 2 individuals). The predator-prey ratio (number of predatory ground beetles / number of pollen beetles) was higher in forest edges than in flower strips, meadows or hedgerows, and intermediate in WOSR fields. Pollen beetle numbers were negatively associated with increasing litter cover and positively with increasing bare soil, while predatory ground beetles were po- sitively associated with soil bulk density and bare soil cover. The proportion of SNH at the landscape level (1 km radius) did not affect overwintering pollen beetles, predatory ground beetles or predator-prey ratio. We conclude that WOSR fields, rather than SNH are the major overwintering habitat and source of spring colonisation of WOSR by pollen beetles. Forest edges promote predatory ground beetles rather than pollen beetle populations according to their disproportionally high value of overwintering individuals. Agri-environmental measures to promote predatory ground beetles rather than pollen beetles should promote overwintering habitats with compact soils and high litter cover. 1. Introduction Agricultural landscapes are typically composed of crops and various types of uncropped, semi-natural habitats (SNH). SNH play a vital role for sustaining biodiversity in agricultural landscapes and the func- tioning of agroecosystems (Landis et al., 2000; Tscharntke et al., 2012). They provide essential resources for beneficial organisms, e.g. natural enemies of crop pests, such as food resources, shelter and overwintering habitat (Holland et al., 2016; Sarthou et al., 2014; Thorbek and Bilde, 2004). During the growing season of crops, SNH can enhance popula- tion densities and the diversity of natural enemies, thereby improving natural pest control services (Rusch et al., 2016; Sutter et al., 2018; Tschumi et al., 2015). Outside the crop growing season, SNH can pro- vide valuable overwintering habitats from which natural enemies can recolonize crops (Landis et al., 2000; Pfiffner and Luka, 2000; Sarthou et al., 2014). The availability of suitable overwintering habitats in agroecosystems is therefore a key component of successful conservation biological control (Griffiths et al., 2008; Landis et al., 2000; Leather et al., 1995). In fact, the proportion and composition of SNH at the landscape scale may be an important factor shaping communities of crop pests and their natural enemies (Bianchi et al., 2006). For example, the abundance of predatory ground beetles and predation of pollen beetles in WOSR fields can be positively related to the proportion of SNH in the agricultural landscape (Sutter et al., 2018). But much less is https://doi.org/10.1016/j.agee.2018.06.030 Received 13 November 2017; Received in revised form 22 June 2018; Accepted 27 June 2018 ⁎ Corresponding author at: Agroecology and Environment, Agroscope, Reckenholzstrasse 191, CH-8046, Zürich, Switzerland. E-mail addresses: louis.sutter@agroscope.admin.ch (L. Sutter), mjamato@gmail.com (M. Amato), philippe.jeanneret@agroscope.admin.ch (P. Jeanneret), matthias.albrecht@agroscope.admin.ch (M. Albrecht). Agriculture, Ecosystems and Environment 265 (2018) 275–281 0167-8809/ © 2018 Elsevier B.V. All rights reserved. T