Ants, pests and natural enemies in Mediterranean citrus. Ecological interactions and practical implications for biological control

Abstract

Ants constitute an important component of the citrus agroecosystem fauna acting simultaneously as predators and as hemipteran mutualists. Thus, ants in citrus are in the center of a complex food web affecting the composition and the population dynamics of a wide arthropod community including honeydew and non-honeydew producing herbivores as well as their natural enemies. In eastern Spain the most abundant and widespread ant species are the natives Lasius grandis Forel and Pheidole pallidula (Nylander), whereas the invasive Linepithema humile (Mayr) is also present but not widespread. We have conducted ant-exclusion experiments in three commercial citrus orchards, each one dominated by one ant species (Pheidole pallidula, Lasius grandis or Linepithema humile) with the aim to disentangle the ecological interactions with honeydew and non-honeydew producing pests and with natural enemies at the community-level. We quantified the effect of the ant-exclusion on the infestation levels and parasitism of three of the most important citrus pests in the area, the honeydew producer Aleurothrixus floccosus (Maskell) and the non-honeydew producers Aonidiella aurantii (Maskell) and Phyllocnistis citrella (Stainton). California red scale densities on fruits were significantly lower in the two seasons and in the three orchards in the ant-excluded treatment. Similarly, the percentage of shoots occupied by A. floccosus was significantly lower in the ant-excluded plots in the orchards dominated by P. pallidula and L. humile. Interestingly, no significant differences were found in the percent parasitism between ant-allowed and ant-excluded treatments. These results suggest that factors other than parasitoid disruption might explain the increased pest populations observed in the presence of ants. Thus, in the same orchards we compared the abundance, species richness, diversity and community structure of predators and parasitoids between the ant-allowed and ant excluded treatments. A total of 176,000 arthropods belonging to 81 taxa were captured and identified. Regarding abundance, the response of natural enemies to ant-exclusion was species specific. When examining functional groups, parasitoids showed higher abundances in the ant-allowed treatment whereas most generalist predators were less abundant. Similarly, the species richness and the Shannon diversity index of parasitoids were higher in the ant-allowed treatment, whereas the species richness of predators was significantly lower. The community structure of predators and parasitoids was not significantly different between treatments. Thus, ants were not associated with a dramatic and/or general decrease in natural enemy abundance or biodiversity. On the other hand, the negative impact of ants on generalist predators may have important implications for the regulation of pest populations. Finally we tested a novel hypothesis examining the potential competition between ants and natural enemies for honeydew produced by Hemiptera. Through the use of high performance liquid chromatography (HPLC) we related the level of ant activity with the energy reserves and feeding history of individual specimens collected in the field. A significant negative correlation between ant activity and the total sugar content and honeydew feeding incidence by A. chrysomphali was found in summer, when ant activity peaked. Ant activity was negatively correlated with the sugar feeding incidence by C. carnea in spring. This is a previously undocumented indirect interaction in food webs in which ants interfere with the physiological state of the natural enemies. Given that the absence of sugar feeding is detrimental for the fitness of many species of predators and parasitoids, this interaction may have important consequences for the arthropod community composition and practical implications for biological control.