Seasonal patterns of egg production in field colonies of the termite Reticulitermes speratus (Isoptera: Rhinotermitidae)

Popul Ecol (27) 49:179 183 DOI 1.17/s1144-6-3-4 NOTES AND COMMENTS Seasonal patterns of egg production in field colonies of the termite Reticulitermes speratus (Isoptera: Rhinotermitidae) Kenji Matsuura Æ Norimasa Kobayashi Æ Toshihisa Yashiro Received: 7 July 26 / Accepted: 16 October 26 / Published online: December 26 Ó The Society of Population Ecology and Springer 26 Abstract This is the first report on the annual egg production patterns in mature termite colonies in the field. Data on the seasonal patterns of egg production in field colonies are very important for understanding the annual colony growth schedule, resource allocation, and population dynamics of the termites. However, collecting the eggs from a sufficient number of colonies is extremely difficult in Reticulitermes termites because their multiple-site nesting makes it difficult to find the reproductive center of the colonies. Here, we first show the seasonal pattern of egg production in the subterranean termite Reticulitermes speratus by collecting the reproductive center of ten colonies each month from April through October. We had to destructively examine dozens of nests to find eggs from enough field colonies each month. Mature field colonies began to produce eggs in late May, soon after the swarming season, and the egg production rate (EPR) reached its maximum in early July. The eggs hatched until late October. The EPR was significantly correlated with the average monthly temperature. Additional investigation of the egg distributions in the nests showed that most eggs were kept around the royal cell, which contained the reproductives. The largest colony had 19 supplemental queens and 94,23 eggs, suggesting that each queen produced an average of 24.7 eggs per day, based on the known mean hatching period of an inseminated egg of 34.9±.12 (SE) days. K. Matsuura (&) N. Kobayashi T. Yashiro Laboratory of Insect Ecology, Graduate School of Environmental Science, Okayama University, 1-1-1 Tsushima-naka, Okayama 7-83, Japan e-mail: kenjijpn@cc.okayama-u.ac.jp Keywords Annual life cycle Colony growth Reproductive schedule Sociogenesis Sociometry Introduction The demography of colonies of perennial social insects involves matters of both annual life cycle and wholecolony life cycle. Termite colonies allocate resources to various functions, e.g., egg production, rearing larvae, colony structure, soldiers, and alate production. The optimal resource allocation is likely to change with colony growth and should vary among seasons (Oster and Wilson 1978). Egg production is one of the most important factors in determining the demography of colonies. Little is known, however, about when and how many eggs are produced in the field colonies of Reticulitermes termites. Their nesting and feeding habits make it extremely difficult to examine the seasonal patterns of egg production. Reticulitermes termites are classified as multiple-site nesters (also called multiple-piece type and intermediate type), based on their nesting and feeding habits (Abe 1987; Shellman-Reeve 1997). Multiple-site nesters are defined as species whose mature colony life is spent in more than one nest. Nests of a single colony are often interconnected to form a large, loosely defined feeding territory (Shellman-Reeve 1997). Because of this nesting habit, it is difficult to find the reproductive-center cells among the multiple nests, where reproductives, eggs, and larvae are protected. The nursery cells, including eggs and larvae, are sometimes located underground, such as in roots or soil, making it harder to collect eggs. To examine the total number of eggs in a colony, it was necessary to

18 Popul Ecol (27) 49:179 183 completely dismantle the nest. Once a colony is dismantled, it is impossible to assess the egg production of the colony in the following months. Periodical random sampling of eggs from a number of colonies is, therefore, the only way to examine seasonal egg production patterns in field colonies, although the colony size should also affect the number of eggs. In this study, we first examined the spatial distribution of eggs in large Reticulitermes speratus (Kolbe) nests in the wild. Reticulitermes termites often harbor sclerotia of a termite-egg mimicking fungus termiteballs along with their eggs (Matsuura et al. 2; Matsuura 2, 26). Because differences in the percentage of termite-balls among egg piles would make it difficult to accurately estimate the total number of eggs, we therefore compared the percentage of termite-balls among egg piles in the nest. We then conducted destructive sampling of the field colonies to examine the seasonal pattern of egg production. Materials and methods Distribution of eggs in the nests We collected the nests of two large mature colonies of R. speratus in Kamigamo (colony A) and Iwakura (colony B), Kyoto, western Japan, in July 21. The nest of colony A was found in a 3 12-cm (diameter length) pine log, and that of colony B in a 3 18- cm pine log. We cut the logs into 2-cm lengths (colony A: segments A F; colony B: segments A I) using a chainsaw, carefully dismantled each piece, and extracted the egg piles and reproductives using an aspirator. Before cutting the log, we removed the bark and collected the eggs in the space between the bark and the wood to avoid crushing any eggs. These eggs were counted separately as the eggs just under the bark (segment UB). The eggs and reproductives were preserved in 1% ethyl alcohol. The eggs in a cell formed an egg pile, and the total number of eggs was counted for each pile. All of the eggs and termite-balls were counted individually under a stereoscope. Egg collection from field colonies Nests of mature colonies of R. speratus were collected in a forest mainly consisting of pine and oak from April to October 2 in the City of Okayama, Okayama, western Japan. Our preliminary research conducted in Kyoto City in 2 and 21 and in the City of Okayama in 24 showed that R. speratus colonies have no eggs in the field from November to March. In total, 1 colonies collected in November and 22 colonies in March had only larvae but no eggs in their nursery calls. We therefore began the egg collection in April and continued through to the end of October. When we found termites in wood, the nest wood was completely dismantled to find the reproductive-center cells, where reproductives, eggs, and larvae were harbored. The nursery cells always contained larvae, even in early May and late October, when there were no eggs in the colony. When we could not find any larvae or eggs, but, rather, only workers, i.e., we could not find the reproductive-center cell, the colony was not counted. When we could not find eggs, but could find larvae, the colony was counted as a colony having no eggs. When we found eggs, we very carefully dismantled the nest wood around the eggs and extracted as many eggs as possible using an aspirator. We continued to collect colonies until we had found the reproductive-center cells of ten colonies each month. Five colonies were collected in the first half and another five colonies were collected in the second half of each month. The data of the colonies collected in the same month were pooled for statistical analysis. The eggs were placed in 1.-ml microtubes, kept in a cooler box with a cooling gel, and then returned to the laboratory. We efficiently removed wood dust and soil from eggs using the egg protection behavior of termites (Matsuura et al. 2; Matsuura 2, 26) as follows: the eggs and approximately 1 workers were placed in a 9-mm Petri dish. The workers carried the eggs and piled them in the Petri dish, after which, they removed saw dust and soil from the egg pile. We then collected the clean eggs from the Petri dish and weighed the egg pile using an electronic scale accurate to.1 mg. One-hundred eggs were randomly chosen from each colony and weighed. We estimated the total number of eggs in the colony from the weight of 1 eggs and the total egg weight. The number of termite-balls was discounted from the estimate. Our earlier study showed that the mean hatching period of sexual eggs was 34.9 (±.12 SE) days (Matsuura and Kobayashi 26). The number of eggs at any given time in the nest is, therefore, the accumulated number of eggs laid in the last 34.9 days. The number of eggs present divided by 34.9 can, therefore, be used as the average daily egg production rate (EPR) in each colony. We analyzed the seasonal egg production patterns and the correlation between temperature and EPR. The average temperature of each month was obtained from the annual report for 2 of the Okayama Local Meteorological Observatory (http://www.osaka-jma. go.jp/okayama/okayam1.htm).

Popul Ecol (27) 49:179 183 181 Results We collected 94,23 eggs and 1,688 termite-balls from colony A, which contained one primary king and 19 supplementary queens (Figs. 1, 2). In colony B, we found 2,788 eggs and 84 termite-balls, and one primary king and 1 supplementary queens (Figs. 1, 2). The average daily EPR divided by the number of supplementary queens in each colony is the number of eggs produced by each queen per day. Thus, it can be estimated that each queen produced 24.7 eggs per day in colony A and 39.7 eggs in colony B. All supplementary queens were nymphoid, i.e., neotenic reproductives derived from nymphs. In both colonies, the largest number of eggs was found in the segment that contained royals (kings and queens). The percentage of termite-balls in colony A (1.76%) was significantly greater than that in colony B (.4%; v 2 =212.11, P<.1). The eggs were kept in 7-egg piles in colony A and in 29-egg piles in colony B (Figs. 1, 2). The number of eggs and termite-balls in each egg pile were significantly linearly related in both colony A (linear regression, r=.93, F 1,73 =43.4, P<.1) and in colony B (r=.74, F 1,27 =33.2, P<.1) (Fig. 2). The intercept was not significantly different from in colony A (t-test, t=.8, P=.6) or in colony B (t=.2, P=.61). Mature field colonies began to produce eggs in late May, soon after the swarming season, and the EPR reached its maximum in early July. Eggs continued to No. of termite eggs No. of termite eggs No. of egg piles No. of egg piles 2 2 1 1 4 3 2 1 1 1 12 8 4 Colony A A B C D E F UB* Segment Colony B egg termite ball 8 6 4 2 4 3 2 1 No. of termite-balls No. of termite-balls No. of termite-ball No. of termite-ball 1 8 6 4 2 1 1 Colony A 1 2 3 4 No. of termite egg Colony B A B C D E F G H I UB* Segment 1 2 3 No. of termite egg Fig. 1 The number of egg piles and the total number of eggs (closed bars) and termite balls (open bars) in each segment of the nest. The segment containing royals (kings and queens) is indicated by an arrow. *UB=eggs found just under the bark Fig. 2 Significant correlations between the number of termite eggs and the number of termite-balls in each egg pile in colony A (Y=.78+.19X, r=.93, F 1,73 =43.4, P<.1) and in colony B(Y=.29+.36X, r=.74, F 1,27 =33.2, P<.1)

182 Popul Ecol (27) 49:179 183 No. of eggs collected Egg production rate (eggs/day) (a) 1 1 1 1 1 1 (b) 1 8 6 4 2 SW J F M A M J J A S O N D J F M A M J J A S O N D Fig. 3a, b Annual pattern of egg production. a Number of eggs collected from each colony. Five colonies were collected in the first half and another five colonies were collected in the second half of each month. SW=swarming occurred in 2. The dashed line indicates the average temperature for each month. b Estimated daily egg production rate (EPR); the bars indicate standard errors Egg production rate (eggs/day) 2 2 1 1 hatch until late October (Fig. 3a). The EPR showed significant seasonal differences (one-way ANOVA, month: F 6,63 =2.99, P=.12; Fig. 3b), and was significantly correlated with the average monthly temperature (Kendall s correlation, s=., P<.; Fig. 4). 4 3 2 1 Temperature ( C) 1 2 2 3 Temperature ( C) Fig. 4 Significant correlation between the average monthly temperature and the EPR (Kendall s correlation, s=., P<.) Discussion In this paper, we illustrate the first clear seasonal pattern of egg production in R. speratus colonies in the field, although the variation in the number of eggs among colonies was large. For the colony foundation, the females (foundresses) of R. speratus begin to lay eggs in mid- to late May in western Japan (Matsuura and Nishida 21; Matsuura et al. 22, 24). This is almost simultaneous to the beginning of egg production in mature colonies. This timing seems adaptive for mature colonies in terms of resource allocation, since they need to allocate an abundance of resources to alate production, rather than to egg production, until swarming. For termites in temperate regions, temperature is an important limiting factor in determining feeding efficiency. It has been reported that the optimum temperature for the feeding activity of R. speratus is 3 C, in comparison with 2 C, 2 C, 3 C, and 4 C (Nakayama et al. 24). The seasonality of egg production was partially explained by the temperature itself; however, temperature alone is not sufficient to explain the variation. We demonstrated, for the first time, that the EPR is at its maximum in early July (Fig. 3b), even though the monthly average temperature reaches its maximum in August (28.3 C in 2). This can be reasonably explained by the fact that the hatching period is 3 days (Matsuura and Kobayashi 26), and, thus, the number of immature larvae soon after hatching is at a maximum in August, when termites have the best feeding efficiency. The first and second instar larvae cannot feed by themselves, and, therefore, need to be fed by workers. Feeding and tending to immature larvae requires as much nutritive resources as egg production. Tsunoda et al. (1999) estimated the foraging population of three field colonies of R. speratus using a mark release recapture method, and reported that the three foraging populations ranged from 19,4 to 466,4 termites per colony. These estimated foraging populations were much larger than those previously reported. We agree with these estimates, rather than previous reports, although the mark release recapture data must be interpreted with caution (Thorne et al. 1996). It seems likely that earlier studies underestimated the colony size of R. speratus, as we found that the number of eggs alone exceeded 9, in the large field colony. This report on the seasonal patterns of termite egg production should be beneficial for the study of termite life history, and can be applied to the study of effective termite control.

Popul Ecol (27) 49:179 183 183 Acknowledgments We thank S. Tatsumi and A. Sato for their assistance in the course of this study, and K. Okada, F. Nakasuji, and T. Miyatake for their useful discussions. This work was funded by the Japan Society for the Promotion of Science (JSPS) and the Program for Promotion of Basic Research Activities for Innovative Biosciences (PROBRAIN) to K.M. References Abe T (1987) Evolution of life types in termites. In: Kawano S, Connell JH, Hidaka T (eds) Evolution and coadaptation in biotic communities. University of Tokyo Press, Tokyo, Japan, pp 12 148 Matsuura K (2) Distribution of termite egg-mimicking fungi ( termite balls ) in Reticulitermes spp. (Isoptera: Rhinotermitidae) nests in Japan and the United States. Appl Entomol Zool 4:3 61 Matsuura K (26) Termite-egg mimicry by a sclerotium-forming fungus. Proc R Soc Lond B 273: 129 Matsuura K, Fujimoto M, Goka K, Nishida T (22) Cooperative colony foundation by termite female pairs: altruism for survivorship in incipient colonies. Anim Behav 64:167 173 Matsuura K, Fujimoto M, Goka K (24) Sexual and asexual colony foundation and the mechanism of facultative parthenogenesis in the termite Reticulitermes speratus (Isoptera: Rhinotermitidae). Insectes Soc 1:32 332 Matsuura K, Kobayashi N (26) Size, hatching rate, and hatching period of sexually and asexually produced eggs in the facultatively parthenogenetic termite Reticulitermes speratus (Isoptera: Rhinotermitidae). Appl Entomol Zool (in press) Matsuura K, Nishida T (21) Comparison of colony foundation success between sexual pairs and female asexual units in the termite Reticulitermes speratus (Isoptera: Rhinotermitidae). Popul Ecol 43:119 124 Matsuura K, Tanaka C, Nishida T (2) Symbiosis of a termite and a sclerotium-forming fungus: Sclerotia mimic termite eggs. Ecol Res 1:4 414 Nakayama T, Yoshimura T, Imamura Y (24) The optimum temperature humidity combination for the feeding activities of Japanese subterranean termites. J Wood Sci :3 34 Oster GF, Wilson EO (1978) Caste and ecology in the social insects. Princeton University Press, Princeton, New Jersey Shellman-Reeve JS (1997) The spectrum of eusociality in termites. In: Choe JC, Crespi BJ (eds) Evolution of social behavior in insects and arachnids. Cambridge University Press, Cambridge, UK, pp 2 93 Thorne BL, Russek-Cohen E, Forschler BT, Breisch NL, Traniello JFA (1996) Evaluation of mark release recapture methods for estimating forager population size of subterranean termite (Isoptera: Rhinotermitidae) colonies. Environ Entomol 2:938 91 Tsunoda K, Matsuoka H, Yoshimura T, Tokoro M (1999) Foraging populations and territories of Reticulitermes speratus (Isoptera : Rhinotermitidae). J Econ Entomol 92:64 69