Vincent de Leijster

31 Agroecological management improves ecosystem services in almond orchards within one year 3 3.1 INTRODUCTION Intensification and land use homogenization in the agricultural sector have been identified as important drivers for land degradation (Foley et al., 2005; Mirzabaev et al., 2016; Tscharntke et al., 2005). Land degradation is an anthropogenic process causing terrestrial ecosystem services and biodiversity to decline (IPBES, 2018). Worldwide, land degradation is estimated to affect over 29% of the land surface and 25% of cropland (Le et al., 2015). The 2018 assessment of the Intergovernmental Panel on Biodiversity and Ecosystem Services (IPBES) reports that, although food provisioning in Europe is increasing, at least seven of the sixteen ecosystem services it examined are declining (IPBES, 2018). This suggests that food production and economic growth might be provided at the cost of other services (Raudsepp- Hearne et al., 2010; Scherr and McNeely, 2008). In the long term, land degradation is even expected to negatively affect crop productivity itself (Nkonya and Mirzabaev, 2016). This trend can already be observed in European woody fruit-crop systems, such as olive, citrus, vine and almond, which declined in productivity by 12% between 1982 and 2010 as a result of land degradation processes (Cherlet et al., 2013). Thus, rehabilitating degraded ecosystem services within agricultural landscapes and simultaneously meeting global food demands is a major challenge (Foley et al., 2005; Gaba et al., 2015). Successful rehabilitation of individual ecosystem services has been demonstrated for pollination (Saunders et al., 2013), pest control (Eilers and Klein, 2009), carbon storage (Lal, 2006) and soil fertility (Ramos et al., 2011). The simultaneous rehabilitation of multiple ecosystem services in productive agricultural landscapes is more challenging and is an emerging field of research (Bennett et al., 2009). When multiple ecosystem services are positively related to each other across space and time (Raudsepp-Hearne et al., 2010), bundles of ecosystem services emerge. For example, Gamfeldt et al. (2013) found a bundle including bilberry production, game hunting potential, understory plant species richness and dead wood across multiple forests in Sweden. On the contrary, when ecosystem services are negatively correlated, trade-offs may occur (Rodríguez et al., 2006). For example, trade- offs between carbon sequestration and biodiversity conservation have been described when comparing monoculture plantings for carbon credits and ecological restoration projects in Australia (Bryan, 2013; Crossman et al., 2011), andbetweenbiodiversity conservation and food production in Indonesian oil palm plantations (Teuscher et al., 2015). ‘No-effect’ relationships are more difficult to detect or less reported than trade-off and bundle relationships as there are less methods available that can effectively measure them (Lee and Lautenbach 2016). Lee and Lautenbach (2016) did a meta-analysis on a wide range of land-use systems and