ABSENCE OF KIN DISCRIMINATION BEHAVIOR IN A SOLDIER-PRODUCING APHID, CERATOVACUNA JAPONICA (HEMIPTERA: PEMPHIGIDAE; CERATAPHIDINI)

-Certain aphid species produce sterile soldiers, morphologically and behaviorally specialized individuals that defend fertile colony members, some or all of which are clonemates. If the soldier morph is maintained by inclusive fitness advantages, its altruism should prefer entially benefit relatives, suggesting a potential role for kin discrimination. We performed a field experiment on spatial segregation and two laboratory experiments on agonistic behavior among non-soldiers and by soldiers of the cerataphidine aphid, Ceratovacunajaponica. For the test of spatial segregation, we introduce a new method of nearest-neighbor analysis, by con structing a minimum spanning tree from the map of individual locations and comparing the numbers of within-group and between-group connections. The results provide no evidence of kin recognition abilities in this species. Members of different clones showed no tendency to segregate spatially, nor to direct displacement attempts against non-kin when competing for feeding sites. Soldiers were indiscriminately aggressive toward early instar reproductives from their own and other colonies. We discuss the implications of these findings for several evolu tionary hypotheses on the maintenance of the soldier morph in aphids. The discovery of sterile aphid soldiers (Aoki, 1977) has been widely interpreted as confirming a prediction of inclusive fitness theory, that altruism can evolve within clones of cyclically parthenogenetic organisms (Dawkins, 1979; Aoki, 1987; Ham ilton, 1987; Ito, 1989). Soldiers are first or second instar individuals which are morphologically and behaviorally specialized for colony defense and which, in failing to molt to adulthood, do not reproduce. Soldier morphs have evolved independently at least four times among pemphigid aphids, and soldier-like behavior by mono morphic first instars, which do become reproductive adults, also occurs in some species (Aoki and Kurosu, 1987; Kurosu and Aoki, 1988; Moran, 1993). Attacks by soldiers and monomorphic defenders on vertebrates and the eggs and larvae of predatory insects have been documented in a number of studies (Aoki, 1979; Ohara, 1985; Aoki and Kurosu, 1987; Kurosu and Aoki, 1988; Foster, 1990). If the colony originated from one parthenogenetic foundress, soldiers protecting fertile colony members would promote the reproduction of their clone, a role analogous to that of somatic cells in a metazoan body (Dawkins, 1979; Hamilton, 1987). However, were some females to join a colony but produce fewer or no soldiers, the "cheaters" could enjoy enhanced reproductive success at the expense of the soldier-producers (Aoki, 1980; Hamilton, 1987). The proportion of soldiers is highly labile both between colonies and, over time, within colonies (Sunose et al., 1991; *to whom correspondence should be addressed This content downloaded from 157.55.39.211 on Sat, 25 Jun 2016 06:33:27 UTC All use subject to http://about.jstor.org/terms 288 JOURNAL OF THE NEW YORK ENTOMOLOGICAL SOCIETY Vol. 102(3) Sakata et al., 1991), suggesting that there may be some variation in soldier-producing propensity, which could facilitate the generation of cheating mutants. Ito (1989), summarizing suggestions of previous authors (e.g. Aoki, 1987; Hamilton, 1987), offered three hypotheses for the evolution and maintenance of soldier production under these circumstances: (1) soldiers recognize and expel unrelated genotypes, preserving clonal integrity; (2) dispersing females join established colonies so infre quently that genetic mixing is unlikely; or (3) predation pressure requires the main tenance of a certain proportion of soldiers even if colonies are genetically diverse, while failure to produce soldiers decreases colony fitness. The experiments reported here were designed to test an expanded version of the kin recognition hypothesis. Kin discrimination by soldiers, and/or by reproductives, could preserve effective clonal integrity by the exclusion of most or all foreign intruders, as in Ito's hypothesis 1 (henceforth referred to as hypothesis la). Alternatively, discrimination could enable soldiers preferentially to defend clonemates within a genetically diverse colony (hy pothesis lb). E.g., if the usually sedentary reproductives form spatially segregated, clonal subgroups, kin recognition could enable the mobile soldiers to remain in the vicinity of their clonemates. In addition, apparently agonistic interactions within colonies have been observed in the form of "butting" behavior, by which aphids attempt to displace fellow colony members from feeding sites (Aoki et al., 1981; Aoki and Kurosu, 1985). Aoki (1987) reported intracolonial butting in all species studied in the tribe Cerataphidini, in cluding soldier-producing and monomorphic species, and speculated that the defen sive behavior and frontal horns of soldiers derived from butting by monomorphic precursors. Noting that conflict is unexpected within pure clones, Aoki suggested that such colonies might indeed be genetically diverse, in contradiction to hypotheses (la) and (2), but consistent with (I b) or (3). (Other interpretations of butting are possible; see Discussion.) Unfortunately, no data on the genetic composition of colonies are yet available for any soldier-producing aphid species. However, the hypothesis that kin discrimination is used to direct butting toward non-clonemates can be tested. Ceratovacunajaponica (Hemiptera: Pemphigidae; Cerataphidini) produces "pseu doscorpion-like" first instar soldiers, with prolonged, sharp frontal horns and thick ened, grasping forelegs (Aoki, 1977, 1987). Dense clusters of these woolly aphids are commonly found on the underside of bamboo grass leaves (Pleioblastus spp.), the secondary host (Aoki and Kurosu, 1991). For the purpose of this study, a "colony" is functionally defined as an aggregation of aphids on one leaf, which can benefit from defense by the soldiers present on that leaf. The inhabitants of multiple leaves on the same or neighboring plants may constitute one genetic population, however. Production of an alate sexual phase, migration to a primary host and gall formation are uncommon in this species; in fact, its gall generation and primary host (Styrax japonica) have been discovered only recently (Aoki and Kurosu, 1991). Most colonies are founded asexually by non-soldier first instars, dispersing on the wind to new plants (Pierce and Berry, unpubl.). In dense populations, over the lifetimes of colonies, it seems possible that a dispersing female may arrive on an already occupied host plant, but the frequency of this event remains unknown. We tested C. japonica aphids for recognition of colonymates and putative clone mates in the contexts of spatial segregation, butting and soldier aggression. Incipient colonies separated by large distances were used in the first two experiments because This content downloaded from 157.55.39.211 on Sat, 25 Jun 2016 06:33:27 UTC All use subject to http://about.jstor.org/terms 1994 NO KIN DISCRIMINATION IN A SOLDIER-PRODUCING APHID 289 they were more likely to be pure clones, still comprised only of offspring of the first female to arrive on the plant. Since incipient colonies contain few soldiers, larger colonies provided soldiers for the third experiment, though an unknown number of subsequently arriving females may have contributed to these colonies. Were these aphids capable of recognizing kin as required by hypotheses (la) or (lb), we expected to observe discrimination against members of foreign colonies in at least one of these three behavioral contexts.