We found some evidence that riparian reserves increase arthropod foraging activity in oil palm plantations, but this did not correspond to a change in herbivory on palm fronds. However, our data suggest that herbivory rates may be lower on oil palm adjacent to larger riparian reserves. Our results suggest that retaining riparian reserves increases the foraging activity of arthropods that bite or chew prey (e.g. ants, centipedes and beetles) on Tyrosine Kinase Inhibitor Library solubility dmso oil palms. This is likely to be the result of spillover from populations

in the riparian reserves (Lucey and Hill, 2012 and Lucey et al., 2014). However, our methodological study (see below) calls into question the extent to which the higher proportion of attack marks from arthropods reflects a higher level of predation on real pests. It may be that the increase in arthropod attacks results from an overall increase in arthropod foraging activity, but not of pest predators in particular. We found that the proportion of artificial pest mimics attacked by birds was not elevated in the vicinity of riparian reserves. This may be because forest fragments do not increase bird abundance or diversity in surrounding areas of oil palm (Edwards et al., 2010), and/or because populations of birds existing exclusively within oil palm plantations provide adequate pest control services. The results of our methodological PI3K inhibitor study (see below) indicate that attack rates on mimics by birds are more likely to reflect

real predation on living pests than data on mimic attack rates by arthropods. too We can therefore be more confident that the data on bird attack rates reflects the role of riparian reserves in provisioning of ecosystem services. The results from our assessment of herbivory rates provide the strongest evidence that riparian reserves characteristic of oil palm landscapes in our study area do not provide a pest control service; there was no significant difference in herbivore activity between sites with and without riparian reserves. However, we were not

able to collect data during a pest outbreak. Outbreaks occur infrequently and are economically much more consequential than background herbivory rates (Basri et al., 1995 and Kamarudin and Wahid, 2010). It is possible that service provision from riparian reserves is only apparent under such conditions, when the population of predators of pests supported by pure oil palm stands becomes saturated with prey. In addition, we were only measuring the impact of defoliating herbivores, and it is possible that the presence of natural habitat in oil palm reserves has a different effect on other pest guilds such as seed predators and stem or root pests. Previous studies have found that increasing the width of riparian reserves in oil palm can increase the species richness or diversity of some taxa (Gray et al., 2014 and Viegas et al., 2014) and that spillover increases with forest fragment size (Lucey et al., 2014).