Research article Protein marking reveals predation on termites by the woodland ant, Aphaenogaster rudis G. Buczkowski and G. Bennett Departmentof Entomology, 901 W. State St., Purdue University, West Lafayette, IN 47907, USA, e-mail : gbuczkow@purdue.edu Received 19 January 2007; revised 23 March 2007; accepted 26 March 2007. Published Online First 20 April 2007 Abstract. Subterranean termites provide a major poten- tial food source for forest-dwelling ants, yet the inter- actions between ants and termites are seldom investigat- ed largely due to the cryptic nature of both the predator and the prey. We used protein marking (rabbit immuno- globin protein, IgG) and double antibody sandwich enzyme-linked immunosorbent assay (DAS-ELISA) to examine the trophic interactions between the woodland ant, Aphaenogaster rudis (Emery) and the eastern sub- terranean termite, Reticulitermes flavipes (Kollar). We marked the prey by feeding the termites paper treated with a solution of rabbit immunoglobin protein (IgG). Subsequently, we offered live, IgG-fed termites to ant colonies and monitored the intracolony distribution of IgG-marked prey. Laboratory experiments on the distri- bution of protein-marked termite prey in colonies of A. rudis revealed that all castes and developmental stages receive termite prey within 24 h. In field experiments, live, protein-marked termites were offered to foraging ants. Following predation, the marker was recovered from the ants, demonstrating that A. rudis preys on R. flavipes under field conditions. Our results provide a unique picture of the trophic-level interactions between preda- tory ants and subterranean termites. Furthermore, we show that protein markers are highly suitable to track trophic interactions between predators and prey, espe- cially when observing elusive animals with cryptic food- web ecology. Keywords : Aphaenogaster rudis , Reticulitermes flavipes, predation, prey-predator interactions, protein marking. Introduction Animal ethology and ecology studies often involve experiments in conditions that may preclude visual observations. This is especially true when observing the trophic ecology of small and/or elusive animals with cryptic behavior. While vertebrate food webs often involve visible predator-prey interactions, which makes field observations relatively straightforward, invertebrate food webs are often substantially more difficult to document. To accurately asses the trophic interactions between invertebrate predators and prey, an efficient marker is required. Current approaches to study preda- tor-prey interactions in invertebrates may involve radio- isotope labeling (Showler et al., 1988; Breene and Sterling, 1988), stable isotopes (e.g. 15 N, Nienstedt and Poehling, 2004), visual identification of gut and/or fecal samples (Sunderland, 1988; Ingerson-Mahar, 2002), se- rological tests that utilize either protein markers (Hagler and Durand, 1994; reviewed in Hagler and Jackson, 2001; Hagler, 2006) or monoclonal antibodies (Hagler et al., 1992; Harwood et al., 2004; reviewed in Sheppard and Harwood, 2005), or various DNA-based techniques (Augusti et al., 2003; reviewed in Symondson, 2002; Harper et al. , 2005 ; reviewed in Sheppard and Harwood, 2005). Despite the availability of these techniques, few studies have examined trophic interactions in inverte- brate food-webs, especially under field conditions. Here, we demonstrate the utility of protein marking to track and quantify the trophic connection between a generalist insect predator and its insect prey. Specifically, we examine predation by the woodland ant, Aphaenogaster rudis (Emery) on the eastern subterranean termite, Reticulitermes flavipes (Kollar). Ants (Hymenoptera) and termites (Isoptera) have been co-evolving for over 100 million years (Hçlldobler and Wilson, 1990) and the great majority of ant-termite Insect. Soc. 54 (2007) 219 – 224 0020-1812/07/030219-6 DOI 10.1007/s00040-007-0933-x  Birkhäuser Verlag, Basel, 2007 Insectes Sociaux