vol. 175, no. 3 the american naturalist march 2010 Natural History Note Spiteful Interactions in a Natural Population of the Bacterium Xenorhabdus bovienii Hadas Hawlena, * Farrah Bashey, Helena Mendes-Soares, and Curtis M. Lively Department of Biology, Indiana University, Bloomington, Indiana 47405 Submitted October 6, 2009; Accepted November 24, 2009; Electronically published January 22, 2010 abstract: An individual behaves spitefully when it harms itself in the act of harming other individuals. One of the clearest potential examples of spite is the costly production and release of toxins called bacteriocins. Bacteriocins are toxins produced by bacteria that can kill closely related strains of the same species. Theoretical work has predicted that bacteriocin-mediated interactions could play an im- portant role in maintaining local genetic and/or species diversity, but these interactions have not been studied at biologically relevant scales in nature. Here we studied toxin production and among-strain in- hibitions in a natural population of Xenorhabdus bovienii. We found genetic differences and inhibitions between colonies that were col- lected only a few meters apart. These results suggest that spite exists in natural populations of bacteria. Keywords: bacteriocins, spatial scale, genotypic variability, coinfec- tion, entomopathogenic bacterium. Introduction Spiteful behaviors are those that are harmful to both the actor and the recipient. It has generally been assumed that the conditions required for spite to evolve are too restric- tive (reviewed in Hamilton 1970, 1996; Keller et al. 1994), but this assumption has been challenged in recent years by a handful of examples of spite from social insects (re- viewed in Foster et al. 2001; Gardner and West 2004b, 2006). One such example involves worker policing in hy- menopteran societies. Although typically unable to mate, workers in most hymenopteran societies have functioning ovaries that allow them to produce males from unfertilized eggs. However, some nonreproductive workers prevent other workers from producing male progeny through ag- gression or by eating their eggs. The workers’ behavior is considered to be spiteful because they invest time and energy to kill other workers’ sons with no direct repro- ductive benefit to themselves (Ratnieks 1988; Foster et al. * Corresponding author; e-mail: hadashaw@gmail.com. Am. Nat. 2010. Vol. 175, pp. 374–381.  2010 by The University of Chicago. 0003-0147/2010/17503-51637$15.00. All rights reserved. DOI: 10.1086/650375 2001). As in this example, most examples in social insects are cases of “weak spite,” because the spiteful individuals usually have little to lose with respect to direct fitness (Hamilton 1971). “Strong spite,” a behavior that demands high costs for the actor, is more difficult to explain (Ham- ilton 1971). The best-documented example for strong spite comes from the sterile soldier caste of polyembryonic par- asitic wasps in the hymenopteran family Encrytidae. In some (but not all) species, some of the wasp larvae develop precociously as soldiers, giving up their own future re- production in order to kill competitors, including oppo- site-sex siblings. Soldiers are capable of recognizing their kin, and they preferentially attack nonkin larvae (Gardner and West 2004a; Gardner et al. 2007). The production and release of bacteriocins provides an- other potential example of strong spite (West et al. 2006). Bacteriocins are extracellular toxins, produced by almost all bacteria, that can kill other strains of bacteria. Bacte- riocin-mediated interactions can be considered spiteful be- cause the actor cells die or pay substantial costs to produce the toxins (Riley and Wertz 2002). Theoretical work has suggested that bacteriocin-mediated interactions could play an important role in nature because they promote local species (or genetic) diversity (Kerr 2007). For ex- ample, bacteriocins might promote local diversity when producer, sensitive, and resistant strains are engaged in a version of the rock-paper-scissors game (i.e., the producer can kill the sensitive strain, the resistant strain outcompetes the producer, and the sensitive strain outcompetes the resistant strain) in a spatially structured environment (e.g., Kerr et al. 2002; Kirkup and Riley 2004). Alternatively, if there are multiple producers, each producer might occupy a certain area and prevent the invasion of the other pro- ducer (e.g., due to inherent bistabilities in the population dynamics; Pagie and Hogeweg 1999; Czaran et al. 2002; Johnson and Seinen 2002). At present, most of the empirical evidence for bacte- riocin-mediated interactions has been documented from inhibition assays between strains that were isolated from