Original article Ants use odour cues to exploit figefig wasp interactions Bertrand Schatz * , Martine Hossaert-McKey Centre d'Ecologie Fonctionnelle et Evolutive (CEFE), UMR 5175 CNRS,1919 route de Mende, 34293 Montpellier, France article info Article history: Received 12 December 2008 Accepted 19 October 2009 Published online 19 November 2009 Keywords: Fig phenology Predation Chemical mediation Community ecology abstract Fig wasps may constitute a relatively abundant food source for ants associated with the figefig wasp nursery pollination mutualism. We found previously that a Mediterranean ant species detects fig wasps by chemical signals. In this paper we want to test the generality of this finding by studying two tropical ants, Oecophylla smaragdina and Crematogaster sp., preying on fig wasps on the dioecious Ficus fistulosa in Brunei (Borneo). Behavioural tests in a Y-tube olfactometer showed that these two ants were attracted both to odours emitted by receptive figs and to those emitted by fig wasps (male and female of the pollinator, and a non-pollinating fig wasp) used here as a kairomone. Naïve workers were not attracted to fig wasps, suggesting that olfactory learning may play a role in prey detection. We also found that O. smaragdina was much more likely to be present on figs of male trees (where fig wasps are more abundant), and that the abundance of this ant species varied strongly with developmental phase of figs on individual trees. Moreover, its aggressiveness was also strongly influenced by the nature of the object presented in our behavioural tests, the site of the test and the developmental phase of the fig tested. Investigation on the chemical and behavioural ecology of the different interacting species provides important insights into the intricate relationships supported by the figefig wasp mutualism. Ó 2009 Elsevier Masson SAS. All rights reserved. 1. Introduction Searching for food constitutes the major part of the time-budget and energy expenditure of most animals (Bell, 1991; H€ olldobler and Wilson, 1990). In social insects, the energy cost of foraging may be minimized by adjusting foraging activities to food availability in time and space, according to optimal foraging theory (Stephen and Krebs, 1986; Traniello, 1989; Bell, 1991; H€ olldobler and Wilson, 1990). Among social insects, ants are well known to have developed a large variety of foraging strategies depending on their species- specific behavioural repertoire and communication systems, and on various parameters related to the availability of local food sources (Traniello, 1989; H€ olldobler and Wilson, 1990; Schatz et al., 1997). Moreover, in addition to an intense food searching activity displayed by several workers, ants are also able not only to establish chemical trails between nest and food site. They are also able to learn one or more characteristics of prey such as their distribution in time and space (Traniello, 1989; H€ olldobler and Wilson, 1990; Schatz et al., 1999a), or their distinctive odour (H€ olldobler and Wilson, 1990; Schatz et al., 2003), thereby increasing their foraging efficiency. However, patterns of presence for some insects are sometimes quite complex, and their exploitation as prey by foraging ants requires precise adjustment in time and space (Schatz et al., 1999a, 2003). An example is the case of fig wasps involved in obligatory pollination mutualisms with fig species. Fig wasps are important prey of several ant species in numerous geographical and ecological conditions (reviewed in Schatz et al., 2008). However, Zachariades (1994) pointed out that detailed studies of interactions between ants and fig wasps are lacking, notably at the level of behavioural and functional aspects. Fig wasps are a diverse set of different chalcidoid wasp radiations, including the pollinators (Agaonidae) and several groups of non-pollinating wasps (Anstett et al., 1997). Both pollinating and non-pollinating wasps may constitute a rela- tively abundant food source for ants, as several tens of wasps may be found on an individual fig in receptive phase (pollinating wasps) and interfloral phase (non-pollinating fig wasps), and even several hundreds of fig wasps per fig at fig maturity; in contrast, fig wasps are always absent during the phase of fig buds (Kerdelhu e and Rasplus, 1996; Weiblen, 2002; Kjellberg et al., 2005; Schatz et al., 2008). Moreover, the ant presence is clearly dependent on the fig species (at least 700 cases of species-specific interactions, but see also Cook and Rasplus, 2003), the sex of the fig in the case of dioecious species, where on the tree figs are borne (axillary, geo- carpic, cauliflorous), and size of the fig when mature (Schatz et al., 2008). How do predatory ants adjust their predatory activity to * Corresponding author. Tel.: þ33 4 67 61 33 00; fax: þ33 4 67 41 21 38. E-mail address: bertrand.schatz@cefe.cnrs.fr (B. Schatz). Contents lists available at ScienceDirect Acta Oecologica journal homepage: www.elsevier.com/locate/actoec 1146-609X/$ e see front matter Ó 2009 Elsevier Masson SAS. All rights reserved. doi:10.1016/j.actao.2009.10.008 Acta Oecologica 36 (2010) 107e113