1. Insect pollination and biological pest control simultaneously influence crop yield, but are often investigated individually. This can lead to under-or over-estimation of the importance of individual services when they interact to affect yield. Recent, limited evidence from field studies showed contrasting results with both additive and non-additive positive and negative effects.2. To disentangle the mechanisms underlying these responses, we conducted a greenhouse experiment and a field study. We tested the potential and realized contribution of insect pollination to cotton boll retention and yield under various pest pressures and biocontrol levels.3. We found both additive and interactive effects of insect pollination and biocontrol within a single crop system depending on the level of pest pressure. In the greenhouse experiment, pollination did not contribute to cotton boll retention and final yield at low pest pressure. At high pest abundances, boll retention and final yield were higher when pollinators were present. In the field study, pollination was sufficient to alter the negative effect of pests on boll retention. Thus, interactive effect between the two ecosystem services on boll retention was present at high pest pressure in the greenhouse and at natural levels of pest pressure in the field, but not at lower pest abundances in controlled conditions. 4. Although cotton plants partly compensated for bolls shedding by increasing their weight in the greenhouse experiment, this effect was not detected in the field study, likely due to higher environmental variation. Similarly, interactive effect of pollination and biocontrol on the final yield was present only in the greenhouse study. Synthesis and applications.We conclude that the contrasting findings of additive versus non-additive effects between ecosystem services may be due to the levels of services and disservices tested and environmental variation. Further, this study shows that even when an ecosystem service does not appear to limit crop yield, it can make a substantial contribution to yield and act as insurance when the other service is reduced. For achieving food and fibre security, it is essential that future studies test interactive effects between these ecosystem services in different systems and environmental conditions.

The aphelinid parasitoid Eretmocerus hayati Zolnerowich & Rose (Hymenoptera: Aphelinidae) was recently released in Australia as a biocontrol agent against the crop pest Bemisia tabaci Gennadius (Hemiptera: Aleyrodidae). It was found that the parasitoid can spread over several kilometers in a single generation and continue laying eggs for over a fortnight. A simple wind-advection model was fitted to emergence data from a first release between Fassifern and Kalbar, Queensland, and its predictive ability was tested against the second release near Carnarvon, Western Australia. The fitting of the model was used to develop several hypotheses about the dispersal of E. hayati, which were validated by the second release: E. hayati flies in the same direction as the wind to a distance proportional to the wind speed; this wind-borne flight takes place at any time during daylight hours; a flight is attempted every day after emergence unless there are high wind conditions during that day; and the high wind condition that will delay flight is wind speeds in excess of ≍2 m/s. This model of E. hayati dispersal may be contrasted with previous models fitted for Eretmocerus species, for which dispersal was dominated by diffusion processes, and parasitoid spread was constrained to the scales of tens and hundreds of meters.