Does alternative food for predatory arthropods improve biological pest control? A meta-analysis

The use of alternative food for arthropod predators in augmentative and conservation biological pest control has increased considerably during the last two decades. Dozens of studies on this topic have been published, showing mixed effects. There are theoretical arguments to expect positive as well as negative effects of alternative food on pest control. In this review, we first discuss these arguments, resulting in a list of factors that potentially affect augmentative and conservation biocontrol with alternative food. We subsequently perform a meta-analysis of selected literature, which is mainly on control of various species of spider mites, thrips, whiteflies and chrysomelids, but also includes some studies on aphids, psyllids and mealybugs. The analysis reveals three factors that have significant effects on the way alternative food affects pest control. First, supplying alternative food should result in increases in the predator densities. Hence, there should be sufficient time for the predators to produce future generations during the cropping season. In the absence of such a numerical response, supplying alternative food can result in lower predation due to satiation of the predators, resulting in decreased pest control. Second, biocontrol is better when predators are released and provided with alternative food before pest invasion, allowing them to build up a population of predators that protects the crop. Third, when the alternative food consists of prey, it should not be of higher quality than the pest, otherwise predators may preferentially feed on the alternative prey, releasing the pest from predation. We furthermore analysed data on biocontrol of thrips, spider mite pests, whiteflies and chrysomelids for which enough data were available to disentangle effects of the various factors mentioned above. These analyses confirmed general results, from which guidelines for improvement of biocontrol of these specific pest groups can be derived. We also argue that the effects of alternative food on biocontrol should be studied by repeatedly assessing densities of pests and predators through time, preferably throughout a growing season.