Twig to Root: Egg-Nest Density and Underground Nymph Distribution in a Periodical Cicada (Hemiptera: Magicicada Septendecim (L.))

Entomologica Americana 116(1/2):73–77, 2010

TWIG TO ROOT: EGG-NEST DENSITY AND UNDERGROUND NYMPH DISTRIBUTION IN A PERIODICAL CICADA (HEMIPTERA: MAGICICADA SEPTENDECIM (L.))

1 ADRIANNE SMITS ,JOHN COOLEY AND ERICA WESTERMAN Department of Ecology and Evolutionary Biology, Yale University, 165 Prospect Street, New Haven CT 06520

Abstract.—The egg-laying behavior of female insects could directly benefit the mother, her offspring, or a combination of both. Periodical cicada (Magicicada septendecim (L.)) females oviposit in twigs in the forest canopy, and newly hatched nymphs fall to the ground, where they spend 17 years feeding on tree roots. If nymph dispersal from their mother’s oviposition site is limited, then female oviposition site selection could influence offspring fitness and survival. Here, we show that there is no correlation between egg-nest density in the forest canopy and nymph density directly below. Although the extent of nymph post-hatching dispersal remains unknown, our findings cast doubt on direct benefits to offspring as an explanation for female oviposition site choice. Key words: Egg-laying behavior, post-hatching dispersal, oviposition site choice.

INTRODUCTION AND BACKGROUND already containing cicadas (Simon et al., 1981), either because they are all making the same site-choice Periodical cicadas (Magicicada spp.) exhibit decisions, or because they are actively aggregating. some of the most extreme life histories of any Explanations for these behaviors include direct insect, with large populations of adults appearing benefits to females over their lifetime of reproduction only once every 13 or 17 years. Mating occurs (e.g., high sunlight levels are advantageous for within dense, noisy choruses, and after mating, ovipositing female cicadas, or high adult densities female Magicicada cut shallow, v-shaped egg-nests reduce predation risk), or direct benefits to offspring in twigs of 3–11 cm diameter and insert up to thirty in a given oviposition site (e.g., trees with more access eggs in each (White, 1980). Magicicada nymphs to sunlight can grow more vigorously and provide spend 13 or 17 years underground feeding on xylem more nutrition to the root-feeding nymphs). Al- from roots (White and Strehl, 1978), and are though some predation may occur as nymphs hatch reported to be fairly sessile, moving only several and move below ground, these losses must be centimeters during their development (White and negligible or randomly distributed if females gain Lloyd, 1975). An adult female’s choice of oviposi- benefits by manipulating the underground distribution sites could substantially affect nymph survival tion of their offspring via oviposition behavior. Thus, if her decisions determine the environment and any hypothesis that females manipulate the under- resources available to her developing nymphs. ground distribution and density of their offspring Magicicada Several patterns in oviposition through their oviposition activities requires at least a behavior have emerged. Ovipositing females show strong initial correlation between aboveground some preference for certain tree species. Ovipos- oviposition density and belowground patterns of M. septendecim iting (L.) females prefer sugar nymph density. A lack of such correlation severely maples and oaks, while generally avoiding coni- discounts the hypothesis that female behaviors can fers (White, 1980). Oviposition density correlates substantially influence the belowground distribution positively with sunlight availability (Yang, 2006), of nymphs. This study examines the question of and females tend to concentrate their oviposition whether the density of first instar nymphs reflects the activities along sunny forest edges (White, 1980). density of egg-nests immediately above them. Females also appear to be attracted to areas

MATERIALS AND METHODS

1 E-mail address for correspondence: adriennesmits@ We sampled a forest edge near the offices of the msn.com American Physical Society on the west side of the 74 ENTOMOLOGICA AMERICANA Vol. 116(1/2)

Fig. 1. There was no significant correlation between the number of eggnests found in a 10 centimeter twig segment and the number of nymphs found in a 0.25 square meter soil area below when all data points were used (A), or when only data points under oak trees, with more than one nymph, were used (B).

William Floyd Parkway, Suffolk County, New By August, the area had notable ‘‘flagging’’ York, on September 21, 2008. This forest consist- damage from female oviposition. Only Magici- ed of densely spaced, uniform trees (, 50 cm dbh, cada septendecim has been found on Long Island , 10 m height) in a narrow strip overhanging (unpublished data). lawn on either side. Soils were thin and sandy. We demarcated the forest edge using a 0.25 m During the peak of the 2008 emergence of 3 0.25 m sampling frame, placed so that the area Magicicada Brood XIV, adult cicadas were enclosed within the frame showed 50% lawn and concentrated mainly on the edges of this strip. 50% leaf litter; we designated the centerline of the 2010 DISTRIBUTION IN A PERIODICAL CICADA 75

Table 1. Results of regression analyses with eggnest number as independent variable.

Nymph number regressed on eggnest number n coefficient p all data 32 0.03 0.817 nymphs . 1 10 0.137 0.619 nymphs under oaks 24 0.268 0.09 nymphs . 0, under oaks 14 0.365 0.103 nymphs . 1, under oaks 8 0.403 0.054 frame as the forest edge. Along this edge, we RESULTS established a transect starting 34.3 m north of A regression analysis of egg-nest number and Research Road and extending northward. We dug nymph number revealed no significant relation- holes along the transect every 2 m, starting at zero ship between these two variables (see Table 1, meters and ending at 80 m, for a total of 40 holes. Fig. 1A). After removing all the data points Each hole had a diameter of 0.3 m and a depth of containing one or fewer nymphs and looking only 0.08 m. We inspected and sifted the dirt from each at data points under oak trees, regression analysis hole, and we removed all periodical cicada nymphs revealed only a weak relationship between egg- and counted them. Because of their ivory color and nest number and nymph number (p # 0.054, see movement, though small, they stood out in the Table 1, Fig. 1B). A regression analysis of branch fine-grained soil. All nymphs were first instar, and height and nymph number also showed no no nymphs of other cicada species were found. significant relationship between these two vari- We estimated egg-nest density above each hole ables (p # 0.32, see Table 2, Fig. 2A). Removing by sampling the egg-nests above. Using a ‘plumb data points did not result in a relationship bob’, we identified the lowest woody twig or between nymph number and egg-nest height (see branch of approximately pencil diameter (preferred Fig. 2B). We found no egg-nests on pine branch- oviposition substrate; (Williams and Simon, 1995)) es, but we did find nymphs directly beneath the directly above each hole and within 2 m of the branches of pine trees. ground. The forest at our study site was charac- terized by densely spaced, thick branches near to forest edge and low to the ground, with only DISCUSSION thinner branches above and behind the edge: We found no strong association between the nymphs falling from higher than 2 m in the canopy densities of egg-nests and the nymphs under them. would face an obstructed path to the ground. After Our goal was to identify those branches that had identifying the lowest branch overhanging each of the most direct and unimpeded access to the our plots, we counted egg-nests on these branches ground below; nymphs falling from higher within 10 cm of the starting location. If a plot had branches would have hit the lower branches no suitable twigs above it lower than 2 m, we before hitting the ground, and thus egg-nest eliminated that plot from further consideration. densities on upper branches would not be We documented the height from the ground and expected to correlate strongly with nymph density species of each twig. We performed a series of in our plots. We sampled only this thicker, dense regression analyses on our data set to determine subset, where we would expect the maximum whether nymph number was dependent on egg-nest correlation between egg-nest density in the number or the height of the egg-nests. branches and nymph density below ground. The

Table 2. Results of regression analyses with height of eggnest-bearing branch as independent variable.

Nymph number regressed on height of eggnests n coefficient p all data 32 20.036 0.32 nymphs . 11020.118 0.08 nymphs under oaks 24 20.028 0.517 nymphs . 0, under oaks 14 20.096 0.103 nymphs . 1, under oaks 8 20.036 0.604 76 ENTOMOLOGICA AMERICANA Vol. 116(1/2)

Fig. 2. There was no significant correlation between the height of the eggnest-bearing twig from the ground and the number of nymphs found in a 0.25 square meter soil area below when all data points were used (A), or when only data points under oak trees, with more than one nymph, were used (B). lack of any such correlation would suggest that density effects may operate over the long term, the either cicada nymphs do not fall straight down cicadas in our study had been underground for from their egg-nests into the ground, or that once less than two months. Karban (1984) observed on the ground or in the soil, they move. Although high nymph mortality (up to 98%), but over a it is possible that the initial distribution of nymphs two-year period; more research must be done to did match egg-nest distribution and that subse- pinpoint the exact timescale of nymph mortality quent mortality reduced variance in density, we underground. The lack of any discernable pat- find this explanation implausible; although such terns at our study site, as well as the presence of 2010 DISTRIBUTION IN A PERIODICAL CICADA 77 cicada nymphs underneath pine trees without American Physical Society for facilitating this study. obvious egg-nests in them, suggests that nymphs Mike Neckermann and Elias Bonaros provided do not fall straight to the ground. Thus, females assistance in the field. probably cannot control the exact underground distributions of their offspring. While female LITERATURE CITED oviposition site choice may determine the broad Harris, M. O., M. Sandanyaka and W. Griffin. 2001. area in which nymphs ultimately establish, for Oviposition preferences of the Hessian fly and example on a forest edge rather than in a forest their consequences for the survival and repro- interior, at a smaller scale the female’s decision may ductive potential of offspring. Ecological Ento- have little effect on nymph success. Thus, our mology 26: 473–486. results cast doubt on hypotheses that explain female Henry, L. M., D. R. Gillespie and B. D. Roitberg. cicada oviposition behavior as an adaptation for 2005. Does mother really know best? Oviposi- increasing nymph survival (Yang, 2006). Other tion preference reduced reproductive perfor- studies have failed to demonstrate that female mance in the generalist parasitoid Aphidius ervi. oviposition behavior optimizes the fitness of Entomologia Experimentalis et Applicata 116: 167–174. hatching offspring; Hessian fly (Mayetiola destruc- Karban, R. 1984. Opposite density effects of nymphal tor (Say)) revealed that females fail to choose and adult mortality for periodical cicadas. Ecol- optimal plant hosts for oviposition and larval ogy 65: 1656–1661. feeding despite a complete lack of larval dispersion Lloyd, M. and H. S. Dybas. 1966. The periodical cicada (Harris et al., 2001), and parasitoid wasp females problem. I. Population Ecology. Evolution 20: (Aphidius ervi Haliday) likewise choose sub-optimal 133–149. aphid hosts for oviposition and larval feeding Simon, C., R. Karban and M. Lloyd. 1981. Patchiness, (Henry et al., 2005). Considering that oviposition density, and aggregative behavior in sympatric site choice seems suboptimal in these examples, in allochronic populations of 17-year cicadas. Ecol- which there is no offspring dispersal, our results ogy 62: 1525–1535. White, J. 1980. Resource partitioning by ovipositing with periodical cicadas, in which nymphs do appear cicadas. American Naturalist 115: 1–28. to disperse, seem even more plausible. Thus, in White, J. and C. E. Strehl. 1978. Xylem feeding by periodical cicadas, the hypothesis that females periodical cicada nymphs on tree roots. Ecolog- choose oviposition sites to optimize their offspring ical Entomology 3: 323–327. distribution appears to have little support. Future White, J. A. and M. Lloyd. 1975. Growth rates of 17 and studies should investigate how far, and by what 13-year periodical cicadas. American Midland means, first-instar nymphs travel to the ground, Naturalist 94: 127–143. how long they remain aboveground after hatching, Williams, K. S. and C. Simon. 1995. The ecology, and if they travel either aboveground or below- behavior, and evolution of periodical Cicadas. ground before feeding. Annual Review of Entomology 40: 269–295. Yang, L. H. 2006. Periodical cicadas use light for oviposition site selection. Proceedings of the ACKNOWLEDGMENTS Royal Society of London Series B-Biological This study was conducted as part of the E&EB Sciences 273: 2993–3000. 230a Methods in Field Ecology course at Yale University. The authors would like to thank the Received 16 June 2009; accepted 27 January 2010