- Email: [email protected]
Egg-dumping lace bugs preferentially oviposit with kin
ANIMAL BEHAVIOUR, 2000, 59, 379–383 doi: 10.1006/anbe.1999.1328, available online at http://www.idealibrary.com on
Egg-dumping lace bugs preferentially oviposit with kin MICHAEL L. G. LOEB*, LYNN M. DIENER† & DAVID W. PFENNIG*
*Department of Biology, University of North Carolina at Chapel Hill †Bard College, Annandale-on-Hudson, New York (Received 17 May 1998; initial acceptance 30 July 1999; final acceptance 21 September 1999; MS. number: A8481)
Egg dumping, or abandoning eggs and young to the care of other conspecifics, results in an extreme form of alloparental care. It is unclear, however, if egg dumpers discriminate among kin and nonkin egg recipients. In the lace bug Gargaphia solani (Heteroptera: Tingidae), some females with eggs (guards) also accept and defend eggs of conspecifics. Other females (egg dumpers) abandon their offspring after oviposition, leaving a single guard as the caregiver. We asked if egg dumpers preferentially dump their eggs among unguarded eggs of kin or nonkin. When given a choice between dumping among eggs of full siblings and eggs of nonsiblings, most eggs (67%) were dumped with full siblings’ eggs. Furthermore, egg dumpers were just as likely to oviposit among eggs of kin with whom they had interacted on a shared host plant during juvenile development as they were to oviposit with kin reared on different host plants. Thus, egg dumpers discriminate kin by using cues associated with eggs, and such cues are not likely to be acquired through interaction on a common host plant environment.
dumpers might have evolved an ability to identify kin. In particular, kin discrimination may evolve so that benefits of receiving eggs, if any, are directed towards kin, or so that the costs of receiving eggs, if any, are directed away from kin. However, it is not clear for any species whether egg dumpers preferentially avoid or seek out kin. The subsocial lace bug Gargaphia solani (Heteroptera: Tingidae) is ideal for testing the importance of kinship in mediating egg dumping. In this species, a single adult, or ‘guard’, female may simultaneously care for her own offspring as well as young of one or more conspecific females that have laid eggs within a common egg mass (Tallamy 1985). ‘Egg dumper’ females, by contrast, abandon their brood after oviposition among guard eggs and do not provide direct care to their young (Tallamy 1985). Our study tested whether egg-dumping G. solani preferentially avoid or seek out kin. We show that kinship can have important consequences for egg-dumping behaviour.
Alloparental care, or extending parental effort towards offspring of others, is taxonomically widespread. An especially striking form of alloparental care is seen in species where female reproductives deposit young among offspring of conspecifics, and then abandon parental care to a single adult or pair of adult conspecifics. This behaviour, known as egg dumping, occurs in birds (Yom-Tov 1980; Eadie et al. 1988; Petrie & Moller 1991; Bertram 1992), insects (Field 1992; Brockmann 1993) and salamanders (Harris et al. 1995), and egg dumpers may conduct it either overtly or covertly. Egg-dumping behaviour and alloparental care are difficult to explain from the perspective of individual-level selection. Egg dumpers, freed of parental care, are able to pursue additional reproductive options. Alloparents, on the other hand, assume costs of caring for offspring of egg dumpers (Andersson 1984; Eadie et al. 1988). Thus, unless alloparents have enhanced offspring production (e.g. through predator dilution or selfish herd effects: Hamilton 1971) or compensatory indirect reproductive benefits through helping kin, alloparental care should not be favoured by natural selection (Hamilton 1964; West-Eberhard 1975). If egg-dumping behaviour is maintained by kin selection (Andersson 1984; Eadie et al. 1988), then egg
METHODS
Study System Gargaphia solani lay eggs in discrete groups on leaves of its host plant, the horsenettle, Solanum carolinense (Tallamy 1985, 1986). Guards defend brood against arthropod predators from egg stage to adulthood, a period of about 19 days (Tallamy & Denno 1981). Individual females lay up to 100 eggs per clutch and a mixed
Correspondence: M. Loeb, Department of Biology, CB 3280, University of North Carolina, Chapel Hill, NC 27599-3280, U.S.A. (email: [email protected]). L. M. Diener is at the Environmental Toxicology Center, University of Wisconsin, B157 Steenbock Library, 550 Babcock Drive, Madison, WI 53706-1293, U.S.A. 0003–3472/00/020379+05 $35.00/0
2000 The Association for the Study of Animal Behaviour
379
2000 The Association for the Study of Animal Behaviour
380
ANIMAL BEHAVIOUR, 59, 2
sibship brood may contain several hundred eggs, depending on the number of egg dumpers that visit the egg mass (Tallamy & Horton 1990). Egg dumpers, which tend to lay eggs along the perimeter of the egg mass (Tallamy & Horton 1990), are attracted by a pheromone secreted by ovipositing females onto their newly laid eggs (Monaco et al. 1998). The pheromone in the absence of eggs is sufficient to induce egg dumping for up to 72 h after its initial secretion (Monaco et al. 1998). Females denied opportunities to egg dump eventually oviposit and guard their own clutch (Tallamy 1986; M. L. G. Loeb, unpublished data). Thus, reproductive tactics in G. solani are adopted facultatively as a function of the social and chemical environment.
Lace Bug Pedigrees To create lace bug families of known relatedness, we collected 100 second- to third-instar nymphs on 29 May 1998 from each of two natural populations in central North Carolina, U.S.A. (NC1 and NC2 populations, respectively). Each field sample contained nymphs from two to three different guarded clutches. A third group consisted of 100 nymphs whose parents had been reared on eggplant, S. melongena, in a greenhouse at the University of Delaware, Newark (DE population). Ancestors of DE greenhouse animals were originally collected in northern Virginia, U.S.A., and were reared for many generations in a greenhouse before we used them in the experiment reported here. We reared all three population samples on potted horsenettle in a greenhouse under natural light levels where temperatures fluctuated between 25–32C. After nymphs eclosed to adulthood, we separated sexes prior to mating and reared adults in single-sex groups of 10–20 adults on potted horsenettle enclosed in fine mesh. We paired mature, virgin males and females from different source populations and confined one pair each to a potted horsenettle plant and allowed them to mate. Thus, all offspring of field-collected lace bugs were hybrids of two source populations (i.e. NC1NC2; NC1DE; NC2DE). We created hybrids to increase genetic diversity among descendents of the original, likely inbred DE samples (D. W. Tallamy, personal communication). Thus, hybridization increased the chances of observing kin discrimination mediated by genetically derived cues. We permitted each male–female pair of field-collected lace bugs to lay a first clutch of hybrid young on one horsenettle plant for 36 h, after which we moved each pair to a fresh plant to lay a second clutch. Second clutches were laid within 3 days after we moved pairs to fresh plants. We removed adult mated pairs from the second clutch after 2–3 days. Thus, each male–female pair produced two groups of offspring for use as experimental subjects: full siblings that developed and interacted on the same host plant and were therefore familiar to each other; and full siblings that developed on different host plants and could not interact and were therefore unfamiliar to each other. Additionally, offspring from different families were necessarily raised on different host plants
and therefore constituted a third experimental group of unfamiliar nonkin. A fourth group of familiar nonkin was not created because it is not possible at present to identify individuals within a mixed sibship brood according to their sibship.
Egg-dumper Choice Tests We conducted egg-dumper choice tests to determine whether labels emanating from a clutch of eggs (1) are sufficient to permit kin discrimination by egg dumpers, and if so, (2) whether labels are acquired from social or host plant environments present from the egg stage to adulthood. Thus, we presented egg dumpers with one of two types of choices. In test A, egg dumpers could oviposit among unguarded eggs of a familiar female full sibling (familiar sister) and among unguarded eggs of an unfamiliar female nonsibling (unfamiliar nonsister). In choice test B, we presented egg dumpers with unguarded eggs of an unfamiliar female full sibling (unfamiliar sister) and unguarded eggs of an unfamiliar nonsister. Similar levels of preference for kin in tests A and B would suggest that kin discrimination is mediated by cues associated with eggs and that there is no effect of familiarity on kin discrimination (i.e. kinship labels are not transmitted between former interactants or through sharing of a common host plant during juvenile development). To create recipient clutches for use as egg-dump sites in both choice tests, we first confined one pair each of a virgin hybrid male and a virgin hybrid female to a single leaf (75–100 mm long) on potted horsenettle plants. Each member of a male–female pair was from a different family and therefore no more closely related to each other than first cousins. Twenty-four hours after egg laying began, we removed male–female pairs from their clutches, clipped clutch-bearing leaves from plants and inserted one leaf stem each into a plastic tube filled with tap water. Within 1.5 h of clipping clutch-bearing leaves from the host plant, one of us used a stereomicroscope to count the number of eggs that had been laid in each recipient clutch. We then placed each of two leaves and their counted recipient clutches in opposite ends of a high translucent polystyrene box, 198 cm and 32 cm high, lined with paper towel. We refilled water tubes as needed. Thirty minutes after we placed recipient clutches in test boxes, we put one mated, gravid hybrid female as an egg dumper on a small (4 cm2) horsenettle leaf located midway between each leaf bearing a recipient clutch. We hydrated the central leaf by inserting the stem into a damp cotton ball wrapped in parafilm. We chose mated females as egg dumpers according to the criteria for choice test A (eggs of familiar sisters versus eggs of unfamiliar nonsisters) or choice test B (eggs of unfamiliar sisters versus eggs of unfamiliar nonsisters). We only used females as egg dumpers if they had not yet started their own clutch. Each egg dumper was used only once during the experiment. To control for position effects, we evenly represented each type of recipient clutch on each side of the test boxes. We placed covered test boxes in an environmental chamber (25C, 16:8 h light:dark cycle) and
LOEB ET AL.: KIN DISCRIMINATION BY EGG DUMPERS
RESULTS
Egg-dumper Responses Egg dumpers from 11 different lace bug families were tested for kin discrimination. Of the 66 egg dumpers used (27 from test A; 39 from test B), 24 egg dumpers (36%) did not lay eggs, one egg dumper (1.5%) laid eggs on the small central leaf that had no recipient eggs, and three egg dumpers (4.5%) died before the end of the 72-h observation period. Thus, these 28 observations were excluded from further analysis. Of the remaining 38 replicates, 16 (42.1%) laid eggs within the perimeter of recipient clutches and 24 (57.9%) laid eggs adjacent to recipient clutches. The frequency of clutches laid within the perimeter and adjacent to the perimeter was not influenced by the proportion of eggs dumped with sisters’ eggs; that is, individual egg-dumper preference did not predict where a female would place her eggs (logistic regression: R2 =0.004, 21 =0.205, N=38, NS). Additionally, familiarity among egg dumpers and recipient egg layers did not influence placement of eggs (likelihood ratio chi-square test: 21 =0.001, N=38, NS). Thus, for all
25
20
Frequency
evenly distributed replicates of choice tests A and B between two shelves within the chamber. We noted movement patterns between recipient clutches by checking locations of egg dumpers 1–2 h after release into the test boxes, and then rechecking egg dumper locations every 12 h thereafter. After 72 h, we removed egg dumpers from the test boxes and the same person that initially counted eggs of recipient clutches recounted the eggs to determine whether and where (i.e. on which leaf) the introduced egg dumper had laid eggs. Egg dumpers make one of several oviposition decisions when under similar experimental protocols (M. L. G. Loeb, unpublished data): (1) no eggs laid; (2) eggs laid on the small central leaf not bearing a recipient clutch; (3) eggs laid within the perimeter of the recipient clutch; or (4) eggs laid adjacent to the recipient clutch. Responses (1) and (2) may imply reproductive inactivity or no preference for either recipient clutch. Because our design did not permit discrimination between these two possibilities we excluded both responses (1) and (2) from statistical analyses. We also excluded observations on egg dumpers that died before termination of the 72-h observation period. We treated responses (3) and (4) as equivalent after finding no significant relationship between the frequency of each type of egg-dumper clutch and egg-dumper preference (see Results). Thus, number of dumped eggs was the difference between the initial number of eggs in a recipient clutch and the final number of eggs in and adjacent to a recipient clutch. We a priori defined egg-dumper preference as the clutch type in each choice test that received the largest proportion of dumper eggs. To normalize proportion data we transformed proportions to their arcsine square root prior to conducting parametric analyses. All statistical tests reported are two tailed. Untransformed means and standard errors are reported.
15
10
5
0
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 Proportion of eggs dumped with sister’s clutch
1
Figure 1. Frequency distribution of the proportion of dumper eggs oviposited with sisters’ eggs. Data shown were pooled across both choice tests; N=38. Arrow points to mean.
subsequent analyses we treat clutches laid within the perimeter and adjacent to the perimeter as equivalent egg-dumper responses.
Do Egg Dumpers Prefer Kin? When presented with clutches from familiar sisters and clutches from unfamiliar nonsisters (test A), egg dumpers laid a meanSE of 7013.1% (N=12) of their eggs with familiar sisters’ eggs. Similarly, when presented with clutches from unfamiliar sisters and clutches from unfamiliar nonsisters (test B), egg dumpers laid on average 659.0% (N=26) of their eggs with unfamiliar sisters. Differences between tests A and B in the percentage of eggs laid with full siblings was not significant (t test: t36 =0.346, NS), therefore we combined data from tests A and B to determine whether there was an overall preference for sisters’ eggs. Significantly more eggs were dumped with full siblings (677.3%) than expected if equal proportions of eggs were dumped on sisters’ and nonsisters’ clutches alike (one-sample t test: t36 =2.25, N=38, P<0.035). Egg dumping followed an ‘all-or-none’ pattern; that is, only four of 38 egg dumpers laid eggs in both recipient clutches; all others dumped eggs in only a single recipient clutch (Fig. 1). Additionally, egg dumpers that preferred sisters laid as many eggs total as dumpers that preferred nonsisters (respectively, 71.49.25 eggs, N=26; and 74.56.28 eggs, N=12; t test: t36 =0.275, NS). Initial choice of recipient clutch by egg dumpers was nonrandom. In 29 of 38 egg dumpers (76%), the first clutch visited received the most dumper eggs (likelihood ratio chi-square test: 21 =10.23, P<0.002). Furthermore, most egg dumpers (61%) remained on the first leaf visited and did not inspect the alternative recipient clutch.
381
382
ANIMAL BEHAVIOUR, 59, 2
DISCUSSION This study demonstrates three points regarding proximate control of kin discrimination in egg dumping G. solani. First, cues associated with conspecific eggs are sufficient to facilitate discrimination of kin by egg dumpers. These kin discrimination cues are either visual or, more likely, chemical in nature. Kinship-specific labels may be released directly by eggs, or labels may originate as maternally derived substances. For instance, pheromones secreted directly onto eggs by mothers (Monaco et al. 1998) may contain information about kinship. Chemical mediation of kin discrimination in G. solani is consistent with the sorts of chemical recognition labels used by other arthropods (Gamboa et al. 1986; Pfennig & Sherman 1995). Second, the kinship-specific labels egg dumpers use to recognize kin are probably either acquired before or during oviposition (e.g. applied by the mother to the egg) or are genetically specified. This is based on the finding that egg dumpers were just as likely to oviposit among eggs of full sisters with which they had interacted and shared a common host plant as they were to oviposit among eggs of unfamiliar full sisters that had developed on different host plants. Furthermore, because no test animals were guarded beyond the early egg stage, we can rule out use of cues passed directly from adults to nymphs as kin discriminators. Use of gene products may be favoured in G. solani because egg dumping among nonkin may make environmentally acquired labels less reliable indicators of kinship than genetically specified labels. Third, recipient egg masses that received a majority of dumper eggs were most often the first clutches visited by egg dumpers. This suggests that recognition labels contain sufficient information to permit kin discrimination from at least 11 cm away (i.e. the distance from the egg dumper release site and recipient clutches). The extent to which recognition labels function at greater distances is unclear. However, pheromones used by egg dumpers to locate egg masses are attractive up to 150 cm away (Monaco et al. 1998), suggesting that kinship labels, if any, contained in pheromones may also operate at this distance. Other processes, such as host plant philopatry, may work in concert with chemically mediated kin recognition labels to foster kin association in the wild. A previous study (Tallamy & Tallamy 1993) showed that kinship does not affect the frequency of aggression towards egg dumpers by guards. Indeed, vigilant aggression by guards is rare regardless of relatedness among interactants (Tallamy & Tallamy 1993). Why do egg dumpers discriminate kin but guards do not? There are at least two evolutionary reasons why egg dumpers (but not guards) discriminate kin. First, aggression by guards towards nonkin may be energetically costly and increase guard exposure to predators. These costs of discrimination are not likely to be as large for egg dumpers, given that egg dumpers can reject nonkin egg masses at a distance rather than through direct physical contact. Thus, if costs of discrimination for guards are high relative to benefits, then selection may favour
passive acceptance of egg dumpers, regardless of their relatedness to the guard. Second, receiving dumper eggs may have direct fitness benefits to guards. For instance, egg dumping G. solani lay proportionately more eggs along the vulnerable perimeter of the egg mass where predators are more likely to strike (Tallamy & Horton 1990). Consequently, egg dumpers may indirectly promote the spread of their genes by laying their survivalenhancing eggs with kin. Thus, for either or both of the above reasons, selection may favour facultative expression of kin discrimination in individual G. solani, depending upon whether the individual is acting as a guard or a dumper. Such morph-specific plasticity in expression of kin discrimination has been demonstrated recently in facultatively cannibalistic tadpoles (Pfennig 1999). Preference for dumping eggs with kin does not necessarily imply that egg dumping is beneficial to guards. Indeed, egg dumping may be deleterious, or even neutral, in its fitness effects on recipients of eggs (Eadie et al. 1988). However, the implication of our study is that egg dumping may have effects on the indirect component of inclusive fitness. Future studies using G. solani will focus on relatedness among interactants in the wild and on inclusive fitness consequences of egg dumping. Acknowledgments We thank K. Pfennig and R. Podolsky for helpful criticism and comments on the manuscript. We also thank D. Tallamy for graciously providing lace bugs. This work was made possible by support provided to M. Loeb and L. Diener from the Blandy Experimental Farm, University of Virginia. References Andersson, M. 1984. Brood parasitism within species. In: Producers and Scroungers. Strategies of Exploitation and Parasitism (Ed. by C. J. Barnard), pp. 195–228. London: Chapman & Hall. Bertram, B. C. R. 1992. The Ostrich Communal Nesting System. Princeton, New Jersey: Princeton University Press. Brockmann, H. J. 1993. Parasitizing conspecifics: comparisons between hymenoptera and birds. Trends in Ecology and Evolution, 8, 2–3. Eadie, J. M., Kehoe, F. P. & Nudds, T. H. 1988. Pre-hatch and post-hatch brood amalgamation in North American Anatidae: a review of hypotheses. Canadian Journal of Zoology, 66, 1709– 1721. Field, J. 1992. Intraspecific parasitism as an alternative reproductive tactic in nest-building wasps and bees. Biological Review, 67, 79–126. Gamboa, G. J., Reeve, H. K. & Pfennig, D. W. 1986. The evolution and ontogeny of nestmate recognition in social wasps. Annual Review of Entomology, 31, 431–454. Hamilton, W. D. 1964. The genetical evolution of social behaviour I, II. Journal of Theoretical Biology, 7, 1–52. Hamilton, W. D. 1971. Geometry for the selfish herd. Journal of Theoretical Biology, 31, 295–311. Harris, R. N., Hames, W. W., Knight, I. T., Carreno, C. A. & Vess, T. J. 1995. An experimental analysis of joint nesting in the salamander Hemidactylium scutatum (Caudata: Plethodontidae):
LOEB ET AL.: KIN DISCRIMINATION BY EGG DUMPERS
the effects of population density. Animal Behaviour, 50, 1309– 1316. Monaco, E. L., Tallamy, D. W. & Johnson, R. K. 1998. Chemical mediation of egg dumping in Gargaphia solani Heidemann (Hemiptera: Tingidae). Animal Behaviour, 56, 1491–1495. Petrie, M. & Moller, A. P. 1991. Laying eggs in others’ nests: intraspecific brood parasitism in birds. Trends in Ecology and Evolution, 6, 315–320. Pfennig, D. W. 1999. Cannibalistic tadpoles that pose the greatest threat to kin are most likely to discriminate kin. Proceedings of the Royal Society of London, Series B, 266, 57–61. Pfennig, D. W. & Sherman, P. W. 1995. Kin recognition. Scientific American, 272, 98–103. Tallamy, D. W. 1985. ‘Egg dumping’ in lace bugs (Gargaphia solani, Hemiptera: Tingidae). Behavioral Ecology and Sociobiology, 17, 357–362.
Tallamy, D. W. 1986. Age specificity of ‘egg dumping’ in Gargaphia solani (Hemiptera: Tingidae). Animal Behaviour, 34, 599–603. Tallamy, D. W. & Denno, R. F. 1981. Maternal care in Gargaphia solani (Hemiptera: Tingidae). Animal Behaviour, 29, 771–778. Tallamy, D. W. & Horton, L. A. 1990. Costs and benefits of the egg-dumping alternative in Gargaphia lace bugs (Hemiptera: Tingidae). Animal Behaviour, 39, 352–359. Tallamy, C. D. & Tallamy, D. W. 1993. The effect of relatedness on Gargaphia solani egg dumping behavior. Animal Behaviour, 45, 1239–1241. West-Eberhard, M. J. 1975. The evolution of social behavior by kin selection. Quarterly Review of Biology, 50, 1–33. Yom-Tov, T. 1980. Intraspecific nest parasitism in birds. Biological Review, 55, 93–108.
383
Egg-dumping lace bugs preferentially oviposit with kin MICHAEL L. G. LOEB*, LYNN M. DIENER† & DAVID W. PFENNIG*
*Department of Biology, University of North Carolina at Chapel Hill †Bard College, Annandale-on-Hudson, New York (Received 17 May 1998; initial acceptance 30 July 1999; final acceptance 21 September 1999; MS. number: A8481)
Egg dumping, or abandoning eggs and young to the care of other conspecifics, results in an extreme form of alloparental care. It is unclear, however, if egg dumpers discriminate among kin and nonkin egg recipients. In the lace bug Gargaphia solani (Heteroptera: Tingidae), some females with eggs (guards) also accept and defend eggs of conspecifics. Other females (egg dumpers) abandon their offspring after oviposition, leaving a single guard as the caregiver. We asked if egg dumpers preferentially dump their eggs among unguarded eggs of kin or nonkin. When given a choice between dumping among eggs of full siblings and eggs of nonsiblings, most eggs (67%) were dumped with full siblings’ eggs. Furthermore, egg dumpers were just as likely to oviposit among eggs of kin with whom they had interacted on a shared host plant during juvenile development as they were to oviposit with kin reared on different host plants. Thus, egg dumpers discriminate kin by using cues associated with eggs, and such cues are not likely to be acquired through interaction on a common host plant environment.
dumpers might have evolved an ability to identify kin. In particular, kin discrimination may evolve so that benefits of receiving eggs, if any, are directed towards kin, or so that the costs of receiving eggs, if any, are directed away from kin. However, it is not clear for any species whether egg dumpers preferentially avoid or seek out kin. The subsocial lace bug Gargaphia solani (Heteroptera: Tingidae) is ideal for testing the importance of kinship in mediating egg dumping. In this species, a single adult, or ‘guard’, female may simultaneously care for her own offspring as well as young of one or more conspecific females that have laid eggs within a common egg mass (Tallamy 1985). ‘Egg dumper’ females, by contrast, abandon their brood after oviposition among guard eggs and do not provide direct care to their young (Tallamy 1985). Our study tested whether egg-dumping G. solani preferentially avoid or seek out kin. We show that kinship can have important consequences for egg-dumping behaviour.
Alloparental care, or extending parental effort towards offspring of others, is taxonomically widespread. An especially striking form of alloparental care is seen in species where female reproductives deposit young among offspring of conspecifics, and then abandon parental care to a single adult or pair of adult conspecifics. This behaviour, known as egg dumping, occurs in birds (Yom-Tov 1980; Eadie et al. 1988; Petrie & Moller 1991; Bertram 1992), insects (Field 1992; Brockmann 1993) and salamanders (Harris et al. 1995), and egg dumpers may conduct it either overtly or covertly. Egg-dumping behaviour and alloparental care are difficult to explain from the perspective of individual-level selection. Egg dumpers, freed of parental care, are able to pursue additional reproductive options. Alloparents, on the other hand, assume costs of caring for offspring of egg dumpers (Andersson 1984; Eadie et al. 1988). Thus, unless alloparents have enhanced offspring production (e.g. through predator dilution or selfish herd effects: Hamilton 1971) or compensatory indirect reproductive benefits through helping kin, alloparental care should not be favoured by natural selection (Hamilton 1964; West-Eberhard 1975). If egg-dumping behaviour is maintained by kin selection (Andersson 1984; Eadie et al. 1988), then egg
METHODS
Study System Gargaphia solani lay eggs in discrete groups on leaves of its host plant, the horsenettle, Solanum carolinense (Tallamy 1985, 1986). Guards defend brood against arthropod predators from egg stage to adulthood, a period of about 19 days (Tallamy & Denno 1981). Individual females lay up to 100 eggs per clutch and a mixed
Correspondence: M. Loeb, Department of Biology, CB 3280, University of North Carolina, Chapel Hill, NC 27599-3280, U.S.A. (email: [email protected]). L. M. Diener is at the Environmental Toxicology Center, University of Wisconsin, B157 Steenbock Library, 550 Babcock Drive, Madison, WI 53706-1293, U.S.A. 0003–3472/00/020379+05 $35.00/0
2000 The Association for the Study of Animal Behaviour
379
2000 The Association for the Study of Animal Behaviour
380
ANIMAL BEHAVIOUR, 59, 2
sibship brood may contain several hundred eggs, depending on the number of egg dumpers that visit the egg mass (Tallamy & Horton 1990). Egg dumpers, which tend to lay eggs along the perimeter of the egg mass (Tallamy & Horton 1990), are attracted by a pheromone secreted by ovipositing females onto their newly laid eggs (Monaco et al. 1998). The pheromone in the absence of eggs is sufficient to induce egg dumping for up to 72 h after its initial secretion (Monaco et al. 1998). Females denied opportunities to egg dump eventually oviposit and guard their own clutch (Tallamy 1986; M. L. G. Loeb, unpublished data). Thus, reproductive tactics in G. solani are adopted facultatively as a function of the social and chemical environment.
Lace Bug Pedigrees To create lace bug families of known relatedness, we collected 100 second- to third-instar nymphs on 29 May 1998 from each of two natural populations in central North Carolina, U.S.A. (NC1 and NC2 populations, respectively). Each field sample contained nymphs from two to three different guarded clutches. A third group consisted of 100 nymphs whose parents had been reared on eggplant, S. melongena, in a greenhouse at the University of Delaware, Newark (DE population). Ancestors of DE greenhouse animals were originally collected in northern Virginia, U.S.A., and were reared for many generations in a greenhouse before we used them in the experiment reported here. We reared all three population samples on potted horsenettle in a greenhouse under natural light levels where temperatures fluctuated between 25–32C. After nymphs eclosed to adulthood, we separated sexes prior to mating and reared adults in single-sex groups of 10–20 adults on potted horsenettle enclosed in fine mesh. We paired mature, virgin males and females from different source populations and confined one pair each to a potted horsenettle plant and allowed them to mate. Thus, all offspring of field-collected lace bugs were hybrids of two source populations (i.e. NC1NC2; NC1DE; NC2DE). We created hybrids to increase genetic diversity among descendents of the original, likely inbred DE samples (D. W. Tallamy, personal communication). Thus, hybridization increased the chances of observing kin discrimination mediated by genetically derived cues. We permitted each male–female pair of field-collected lace bugs to lay a first clutch of hybrid young on one horsenettle plant for 36 h, after which we moved each pair to a fresh plant to lay a second clutch. Second clutches were laid within 3 days after we moved pairs to fresh plants. We removed adult mated pairs from the second clutch after 2–3 days. Thus, each male–female pair produced two groups of offspring for use as experimental subjects: full siblings that developed and interacted on the same host plant and were therefore familiar to each other; and full siblings that developed on different host plants and could not interact and were therefore unfamiliar to each other. Additionally, offspring from different families were necessarily raised on different host plants
and therefore constituted a third experimental group of unfamiliar nonkin. A fourth group of familiar nonkin was not created because it is not possible at present to identify individuals within a mixed sibship brood according to their sibship.
Egg-dumper Choice Tests We conducted egg-dumper choice tests to determine whether labels emanating from a clutch of eggs (1) are sufficient to permit kin discrimination by egg dumpers, and if so, (2) whether labels are acquired from social or host plant environments present from the egg stage to adulthood. Thus, we presented egg dumpers with one of two types of choices. In test A, egg dumpers could oviposit among unguarded eggs of a familiar female full sibling (familiar sister) and among unguarded eggs of an unfamiliar female nonsibling (unfamiliar nonsister). In choice test B, we presented egg dumpers with unguarded eggs of an unfamiliar female full sibling (unfamiliar sister) and unguarded eggs of an unfamiliar nonsister. Similar levels of preference for kin in tests A and B would suggest that kin discrimination is mediated by cues associated with eggs and that there is no effect of familiarity on kin discrimination (i.e. kinship labels are not transmitted between former interactants or through sharing of a common host plant during juvenile development). To create recipient clutches for use as egg-dump sites in both choice tests, we first confined one pair each of a virgin hybrid male and a virgin hybrid female to a single leaf (75–100 mm long) on potted horsenettle plants. Each member of a male–female pair was from a different family and therefore no more closely related to each other than first cousins. Twenty-four hours after egg laying began, we removed male–female pairs from their clutches, clipped clutch-bearing leaves from plants and inserted one leaf stem each into a plastic tube filled with tap water. Within 1.5 h of clipping clutch-bearing leaves from the host plant, one of us used a stereomicroscope to count the number of eggs that had been laid in each recipient clutch. We then placed each of two leaves and their counted recipient clutches in opposite ends of a high translucent polystyrene box, 198 cm and 32 cm high, lined with paper towel. We refilled water tubes as needed. Thirty minutes after we placed recipient clutches in test boxes, we put one mated, gravid hybrid female as an egg dumper on a small (4 cm2) horsenettle leaf located midway between each leaf bearing a recipient clutch. We hydrated the central leaf by inserting the stem into a damp cotton ball wrapped in parafilm. We chose mated females as egg dumpers according to the criteria for choice test A (eggs of familiar sisters versus eggs of unfamiliar nonsisters) or choice test B (eggs of unfamiliar sisters versus eggs of unfamiliar nonsisters). We only used females as egg dumpers if they had not yet started their own clutch. Each egg dumper was used only once during the experiment. To control for position effects, we evenly represented each type of recipient clutch on each side of the test boxes. We placed covered test boxes in an environmental chamber (25C, 16:8 h light:dark cycle) and
LOEB ET AL.: KIN DISCRIMINATION BY EGG DUMPERS
RESULTS
Egg-dumper Responses Egg dumpers from 11 different lace bug families were tested for kin discrimination. Of the 66 egg dumpers used (27 from test A; 39 from test B), 24 egg dumpers (36%) did not lay eggs, one egg dumper (1.5%) laid eggs on the small central leaf that had no recipient eggs, and three egg dumpers (4.5%) died before the end of the 72-h observation period. Thus, these 28 observations were excluded from further analysis. Of the remaining 38 replicates, 16 (42.1%) laid eggs within the perimeter of recipient clutches and 24 (57.9%) laid eggs adjacent to recipient clutches. The frequency of clutches laid within the perimeter and adjacent to the perimeter was not influenced by the proportion of eggs dumped with sisters’ eggs; that is, individual egg-dumper preference did not predict where a female would place her eggs (logistic regression: R2 =0.004, 21 =0.205, N=38, NS). Additionally, familiarity among egg dumpers and recipient egg layers did not influence placement of eggs (likelihood ratio chi-square test: 21 =0.001, N=38, NS). Thus, for all
25
20
Frequency
evenly distributed replicates of choice tests A and B between two shelves within the chamber. We noted movement patterns between recipient clutches by checking locations of egg dumpers 1–2 h after release into the test boxes, and then rechecking egg dumper locations every 12 h thereafter. After 72 h, we removed egg dumpers from the test boxes and the same person that initially counted eggs of recipient clutches recounted the eggs to determine whether and where (i.e. on which leaf) the introduced egg dumper had laid eggs. Egg dumpers make one of several oviposition decisions when under similar experimental protocols (M. L. G. Loeb, unpublished data): (1) no eggs laid; (2) eggs laid on the small central leaf not bearing a recipient clutch; (3) eggs laid within the perimeter of the recipient clutch; or (4) eggs laid adjacent to the recipient clutch. Responses (1) and (2) may imply reproductive inactivity or no preference for either recipient clutch. Because our design did not permit discrimination between these two possibilities we excluded both responses (1) and (2) from statistical analyses. We also excluded observations on egg dumpers that died before termination of the 72-h observation period. We treated responses (3) and (4) as equivalent after finding no significant relationship between the frequency of each type of egg-dumper clutch and egg-dumper preference (see Results). Thus, number of dumped eggs was the difference between the initial number of eggs in a recipient clutch and the final number of eggs in and adjacent to a recipient clutch. We a priori defined egg-dumper preference as the clutch type in each choice test that received the largest proportion of dumper eggs. To normalize proportion data we transformed proportions to their arcsine square root prior to conducting parametric analyses. All statistical tests reported are two tailed. Untransformed means and standard errors are reported.
15
10
5
0
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 Proportion of eggs dumped with sister’s clutch
1
Figure 1. Frequency distribution of the proportion of dumper eggs oviposited with sisters’ eggs. Data shown were pooled across both choice tests; N=38. Arrow points to mean.
subsequent analyses we treat clutches laid within the perimeter and adjacent to the perimeter as equivalent egg-dumper responses.
Do Egg Dumpers Prefer Kin? When presented with clutches from familiar sisters and clutches from unfamiliar nonsisters (test A), egg dumpers laid a meanSE of 7013.1% (N=12) of their eggs with familiar sisters’ eggs. Similarly, when presented with clutches from unfamiliar sisters and clutches from unfamiliar nonsisters (test B), egg dumpers laid on average 659.0% (N=26) of their eggs with unfamiliar sisters. Differences between tests A and B in the percentage of eggs laid with full siblings was not significant (t test: t36 =0.346, NS), therefore we combined data from tests A and B to determine whether there was an overall preference for sisters’ eggs. Significantly more eggs were dumped with full siblings (677.3%) than expected if equal proportions of eggs were dumped on sisters’ and nonsisters’ clutches alike (one-sample t test: t36 =2.25, N=38, P<0.035). Egg dumping followed an ‘all-or-none’ pattern; that is, only four of 38 egg dumpers laid eggs in both recipient clutches; all others dumped eggs in only a single recipient clutch (Fig. 1). Additionally, egg dumpers that preferred sisters laid as many eggs total as dumpers that preferred nonsisters (respectively, 71.49.25 eggs, N=26; and 74.56.28 eggs, N=12; t test: t36 =0.275, NS). Initial choice of recipient clutch by egg dumpers was nonrandom. In 29 of 38 egg dumpers (76%), the first clutch visited received the most dumper eggs (likelihood ratio chi-square test: 21 =10.23, P<0.002). Furthermore, most egg dumpers (61%) remained on the first leaf visited and did not inspect the alternative recipient clutch.
381
382
ANIMAL BEHAVIOUR, 59, 2
DISCUSSION This study demonstrates three points regarding proximate control of kin discrimination in egg dumping G. solani. First, cues associated with conspecific eggs are sufficient to facilitate discrimination of kin by egg dumpers. These kin discrimination cues are either visual or, more likely, chemical in nature. Kinship-specific labels may be released directly by eggs, or labels may originate as maternally derived substances. For instance, pheromones secreted directly onto eggs by mothers (Monaco et al. 1998) may contain information about kinship. Chemical mediation of kin discrimination in G. solani is consistent with the sorts of chemical recognition labels used by other arthropods (Gamboa et al. 1986; Pfennig & Sherman 1995). Second, the kinship-specific labels egg dumpers use to recognize kin are probably either acquired before or during oviposition (e.g. applied by the mother to the egg) or are genetically specified. This is based on the finding that egg dumpers were just as likely to oviposit among eggs of full sisters with which they had interacted and shared a common host plant as they were to oviposit among eggs of unfamiliar full sisters that had developed on different host plants. Furthermore, because no test animals were guarded beyond the early egg stage, we can rule out use of cues passed directly from adults to nymphs as kin discriminators. Use of gene products may be favoured in G. solani because egg dumping among nonkin may make environmentally acquired labels less reliable indicators of kinship than genetically specified labels. Third, recipient egg masses that received a majority of dumper eggs were most often the first clutches visited by egg dumpers. This suggests that recognition labels contain sufficient information to permit kin discrimination from at least 11 cm away (i.e. the distance from the egg dumper release site and recipient clutches). The extent to which recognition labels function at greater distances is unclear. However, pheromones used by egg dumpers to locate egg masses are attractive up to 150 cm away (Monaco et al. 1998), suggesting that kinship labels, if any, contained in pheromones may also operate at this distance. Other processes, such as host plant philopatry, may work in concert with chemically mediated kin recognition labels to foster kin association in the wild. A previous study (Tallamy & Tallamy 1993) showed that kinship does not affect the frequency of aggression towards egg dumpers by guards. Indeed, vigilant aggression by guards is rare regardless of relatedness among interactants (Tallamy & Tallamy 1993). Why do egg dumpers discriminate kin but guards do not? There are at least two evolutionary reasons why egg dumpers (but not guards) discriminate kin. First, aggression by guards towards nonkin may be energetically costly and increase guard exposure to predators. These costs of discrimination are not likely to be as large for egg dumpers, given that egg dumpers can reject nonkin egg masses at a distance rather than through direct physical contact. Thus, if costs of discrimination for guards are high relative to benefits, then selection may favour
passive acceptance of egg dumpers, regardless of their relatedness to the guard. Second, receiving dumper eggs may have direct fitness benefits to guards. For instance, egg dumping G. solani lay proportionately more eggs along the vulnerable perimeter of the egg mass where predators are more likely to strike (Tallamy & Horton 1990). Consequently, egg dumpers may indirectly promote the spread of their genes by laying their survivalenhancing eggs with kin. Thus, for either or both of the above reasons, selection may favour facultative expression of kin discrimination in individual G. solani, depending upon whether the individual is acting as a guard or a dumper. Such morph-specific plasticity in expression of kin discrimination has been demonstrated recently in facultatively cannibalistic tadpoles (Pfennig 1999). Preference for dumping eggs with kin does not necessarily imply that egg dumping is beneficial to guards. Indeed, egg dumping may be deleterious, or even neutral, in its fitness effects on recipients of eggs (Eadie et al. 1988). However, the implication of our study is that egg dumping may have effects on the indirect component of inclusive fitness. Future studies using G. solani will focus on relatedness among interactants in the wild and on inclusive fitness consequences of egg dumping. Acknowledgments We thank K. Pfennig and R. Podolsky for helpful criticism and comments on the manuscript. We also thank D. Tallamy for graciously providing lace bugs. This work was made possible by support provided to M. Loeb and L. Diener from the Blandy Experimental Farm, University of Virginia. References Andersson, M. 1984. Brood parasitism within species. In: Producers and Scroungers. Strategies of Exploitation and Parasitism (Ed. by C. J. Barnard), pp. 195–228. London: Chapman & Hall. Bertram, B. C. R. 1992. The Ostrich Communal Nesting System. Princeton, New Jersey: Princeton University Press. Brockmann, H. J. 1993. Parasitizing conspecifics: comparisons between hymenoptera and birds. Trends in Ecology and Evolution, 8, 2–3. Eadie, J. M., Kehoe, F. P. & Nudds, T. H. 1988. Pre-hatch and post-hatch brood amalgamation in North American Anatidae: a review of hypotheses. Canadian Journal of Zoology, 66, 1709– 1721. Field, J. 1992. Intraspecific parasitism as an alternative reproductive tactic in nest-building wasps and bees. Biological Review, 67, 79–126. Gamboa, G. J., Reeve, H. K. & Pfennig, D. W. 1986. The evolution and ontogeny of nestmate recognition in social wasps. Annual Review of Entomology, 31, 431–454. Hamilton, W. D. 1964. The genetical evolution of social behaviour I, II. Journal of Theoretical Biology, 7, 1–52. Hamilton, W. D. 1971. Geometry for the selfish herd. Journal of Theoretical Biology, 31, 295–311. Harris, R. N., Hames, W. W., Knight, I. T., Carreno, C. A. & Vess, T. J. 1995. An experimental analysis of joint nesting in the salamander Hemidactylium scutatum (Caudata: Plethodontidae):
LOEB ET AL.: KIN DISCRIMINATION BY EGG DUMPERS
the effects of population density. Animal Behaviour, 50, 1309– 1316. Monaco, E. L., Tallamy, D. W. & Johnson, R. K. 1998. Chemical mediation of egg dumping in Gargaphia solani Heidemann (Hemiptera: Tingidae). Animal Behaviour, 56, 1491–1495. Petrie, M. & Moller, A. P. 1991. Laying eggs in others’ nests: intraspecific brood parasitism in birds. Trends in Ecology and Evolution, 6, 315–320. Pfennig, D. W. 1999. Cannibalistic tadpoles that pose the greatest threat to kin are most likely to discriminate kin. Proceedings of the Royal Society of London, Series B, 266, 57–61. Pfennig, D. W. & Sherman, P. W. 1995. Kin recognition. Scientific American, 272, 98–103. Tallamy, D. W. 1985. ‘Egg dumping’ in lace bugs (Gargaphia solani, Hemiptera: Tingidae). Behavioral Ecology and Sociobiology, 17, 357–362.
Tallamy, D. W. 1986. Age specificity of ‘egg dumping’ in Gargaphia solani (Hemiptera: Tingidae). Animal Behaviour, 34, 599–603. Tallamy, D. W. & Denno, R. F. 1981. Maternal care in Gargaphia solani (Hemiptera: Tingidae). Animal Behaviour, 29, 771–778. Tallamy, D. W. & Horton, L. A. 1990. Costs and benefits of the egg-dumping alternative in Gargaphia lace bugs (Hemiptera: Tingidae). Animal Behaviour, 39, 352–359. Tallamy, C. D. & Tallamy, D. W. 1993. The effect of relatedness on Gargaphia solani egg dumping behavior. Animal Behaviour, 45, 1239–1241. West-Eberhard, M. J. 1975. The evolution of social behavior by kin selection. Quarterly Review of Biology, 50, 1–33. Yom-Tov, T. 1980. Intraspecific nest parasitism in birds. Biological Review, 55, 93–108.
383