Appl. Entomol. Zool. 37 (2): 257–262 (2002)

Temporal differences in mating behavior between - and water-oats-populations of the striped stem borer, suppressalis (Walker) (: )

I Made Samudra,1 Kaoru Emura,2 Sugihiko Hoshizaki, Yukio Ishikawa and Sadahiro Tatsuki* Laboratory of Applied Entomology, Graduate School of Agricultural and Life Sciences, The University of Tokyo, Tokyo 113–8657, 2 Saitama Prefectural Agriculture and Forestry Research Center, Saitama 360–0831, Japan (Received 3 September 2001; Accepted 11 January 2002)

Abstract Rice-feeding and water-oats-feeding populations of Chilo suppressalis usually occur over areas around paddy fields in Japan. Laboratory observations showed temporal differences in mating and female calling between the populations. The rice-population was sexually active in the first half of the scotophase whereas the water-oats-population was sexu- ally active in the second half of the scotophase. Under a no-choice condition in the laboratory, adults of the rice- and water-oats-populations could hybridize in both reciprocal crosses. However, mating frequency was significantly lower in crossings between females of the rice-population and males of the water-oats-population than in the opposite crosses. The peak time of mating in the F1 generation was intermediate between the times of their parental popula- tions. Similar patterns of mating time were observed in the F2 generation. These results indicate that the time of mat- ing in C. suppressalis is genetically controlled, most likely by a polygenic system.

Key words: Chilo suppressalis, rice-population, water-oats-population, mating time, inter-populational crossing

and CSW have been considered as a single species. INTRODUCTION Nevertheless, there have been controversial The striped stem borer, Chilo suppressalis, is statements on the possibility of mutual migration one of the most important pests of rice or host change between CSR and CSW under natu- plants in Asian countries (Kalshoven, 1981; Kiri- ral conditions. Takai (1970) showed that bimodal tani, 1988). In Japan, this species has been known distribution of the head width in each gender of to feed on two major host plants, rice (Oryza field-collected was a combination of the two sativa) and water-oats (), aquatic normal distributions, which suggested no host weeds that belong to the same tribe, Oryzeae, as change between CSR and CSW. Similar results rice. with male moths caught by pheromone traps were Although the size of the collected from obtained by Tsuchida and Ichihashi (1995). In con- water-oats (hereafter designated as CSW) is signifi- trast, it has also been suggested that CSW migrated cantly larger than that from rice (hereafter desig- between rice fields and water-oats vegetation based nated as CSR) (e.g. Maki and Yamashita, 1956; on the extremely low population density of CSW Takasaki et al., 1969), no detectable differences in larvae in the first generation in water-oats (Miya- other morphological traits (Marumo, 1930) nor in hara and Abe, 1969; Koike et al., 1981). the karyotypes (Kurihara, 1930) were found. In ad- Recently, Konno and Tanaka (1996b) observed a dition, each population could mature on either of clear difference in the mating time between CSR the plants (Maki and Yamashita, 1956; Takasaki et and CSW in the laboratory using insects collected al., 1969), hybridize with the opposite population from Okayama Pref., i.e. mating of CSR was ob- and produce F1 and F2 progenies under laboratory served during the first half of the scotophase as re- conditions (Koike et al., 1981). Consequently, CSR ported in several previous studies (Tatsuki and

* To whom correspondence should be addressed at: E-mail: [email protected] 1 Present address: Bogor Research Institute for Food Crops, Bogor 16114,

257 258 I Made Samudra et al.

Fukami, 1972; Tatsuki et al., 1975; Kanno, 1979), host plant leaves, similar to what was described in whereas mating of CSW was during the late sco- the previous section. Since most moths of both tophase. Furthermore, Konno and Tanaka (1996a) CSR and CSW emerged within 1–2 h into the sco- found differences between the populations in the tophase, calling behavior was recorded every 30 sensitivity to organophosphorus insecticides and in min during the scotophases of three successive the activity and zymogram of carboxylesterases re- days (0 to 2 days old). The calling behavior was sponsible for the insecticide sensitivity. These find- designated as females extruding the ovipositors ings suggest that CSR and CSW are reproductively (Tatsuki et al., 1975). The data for each population isolated populations and that the difference in the from three successive scotophases were pooled and mating time is a possible mechanism for the repro- analyzed. In our preliminary observations, no fe- ductive isolation. It is, thus, important to clarify males showed calling during the photophase. whether the late mating time found in CSW is a Crossing experiments between populations. local trait and what factor(s) controls the mating Crossings were conducted using plastic cups con- time of CSR and CSW. taining pieces of water-oats leaves, because leaves In this study, by using insects from different lo- of this plant equally facilitated the mating of both calities, we confirm the findings of Konno and populations (Samudra, unpublished data). Tanaka (1996b) and show that the difference in the The F1 offspring between CSW females and mating time may be affected by the time of calling CSR males were noted as F1A, and those of the behavior of females and controlled by a polygenic opposite crossing were noted as F1B. The F2 in- system. sects were obtained by pairing 0-day-old females and males of F1A (noted as F2A) or F1B (noted as F2B). MATERIALS AND METHODS The incidence of mating was observed every 30 Insects. CSR and CSW were collected at Kuma- min as previously described and the findings of gaya (60 km north-west of Tokyo), Saitama Pref. each crossing were treated as in the observation of and Moriya (35 km north-east of Tokyo), Ibaraki mating. Pref., respectively. The two sites are about 60 km apart. Larvae from both populations were cultured RESULTS on rice seedlings under laboratory conditions based on the methods previously reported (Uchiumi, Mating behavior 1974; Kamano and Sato, 1985). Room conditions Mating times of CSR and CSW are illustrated in for both rearing and experiments were maintained Fig. 1. There was a significant difference between at 25Ϯ1°C and 60–80% RH. Illumination was ob- the populations in the onset time of mating (Mann- tained from 6ϫ40-W fluorescent lamps during the Whitney U-test, pϽ0.01). In CSR, mating began at photophase with an L15 : D9 photo-regime. Red 3.7Ϯ0.1 h (meanϮSE, nϭ130) into the scotophase, lamps were used for observation during the sco- ranging from 2 to 6 h. In contrast, in CSW, mating tophase. began at 7.4Ϯ0.1 h, ranging from 5 to 9 h (nϭ129). Observations of mating. Pieces of host plant leaves (rice for CSR or water-oats for CSW, length 8.5 cm) with wet cotton wrapped at the bottom were placed in clear plastic cups (200 ml). Pairs of newly emerged females and males of each popula- tion were individually put in the cups. The mating behavior of each pair was checked every 30 min during the scotophase on successive days until the moths died. The data for each population from the several scotophases in which mating pair(s) were observed were pooled and analysed. Observations of calling. Newly emerged fe- Fig. 1. Onset time of mating in rice- (CSR; nϭ130) and males were individually placed in cups containing water-oats- (CSW; nϭ129) populations of C. suppressalis. Differences in Mating Time of Chilo 259

Fig. 4. Hourly changes in calling frequency of females of ϭ ϭ Fig. 2. Duration of mating in rice- (CSR; nϭ31) and rice- (CSR; n 235) and water-oats- (CSW; n 134) popula- water-oats- (CSW; nϭ43) populations of C. suppressalis. tions of C. suppressalis.

Fig. 3. Onset time of female calling in rice- (CSR; nϭ46) and water-oats- (CSW; nϭ272) populations of C. suppressalis. Fig. 5. Duration of female calling in rice- (CSR; nϭ46) and water-oats- (CSW; nϭ94) populations of C. suppressalis. Thus, there was a slight overlap of the populations in the onset time of mating from 5 to 6 h. The over- all mating activity also showed a different pattern 0.1 h into the scotophase, ranging from 5 to 9 h between the populations, i.e. in the first half of the (nϭ272). Almost all CSW females began calling in scotophase for CSR and in the second half for the last half of the scotophase (Fig. 3). Between the CSW. two populations, there was a significant difference Mating durations between the populations were in the time they began calling (Mann-Whitney U- also significantly different (Mann-Whitney U-test, test, pϽ0.01), although a slight overlap was ob- pϽ0.01), 2.6Ϯ0.1 h (nϭ31) and 2.0Ϯ0.1 h (nϭ43) served. for CSR and CSW, respectively (Fig. 2). However, The overall calling activity of the two popula- the mating frequencies of the two populations were tions is shown in Fig. 4. The calling activity of not significantly different (c 2-test, pϾ0.05), being CSR was observed from 2 to 9 h with the mean at 74.5% (105/141) for CSR and 73.5% (83/113) for 5.6Ϯ0.1 h into the scotophase (nϭ235), while for CSW. CSW it was from 6 to 9 h with the mean at 8.0Ϯ 0.1 h (nϭ134). The peak calling times of both pop- Female calling behavior ulations were also significantly different (Mann- Virgin females of CSR began calling at 3.7Ϯ Whitney U-test, pϽ0.01). 0.2 h (meanϮSE, nϭ46) into the scotophase, rang- Individual calling duration is shown in Fig. 5. ing from 2 to 6 h (Fig. 3). Most began calling in the The mean calling duration of CSR was 4.9Ϯ0.1 h first half of the scotophase. In contrast, virgin fe- (nϭ46), while it was 1.9Ϯ0.2 h for CSW (nϭ94). males of CSW showed a different temporal pattern There was a significant difference between the call- of calling behavior. They began calling at 6.8Ϯ ing durations of the two populations (Mann-Whit- 260 I Made Samudra et al.

Fig. 6. Onset time of cross-mating between rice- (CSR) Fig. 7. Onset time of mating in the F1 of rice- (CSR) and and water-oats- (CSW) populations of C. suppressalis the water-oats- (CSW) populations of C. suppressalis (F1A; (CSW/ϫCSR?; nϭ29, CSR/ϫCSW?; nϭ13). nϭ17, F1B; nϭ19). For a detailed explanation of F1A and F1B, see Materials and Methods. ney U-test, pϽ0.01).

Crossing between populations Although cross-mating between CSR and CSW could occur, mating frequencies were largely dif- ferent between reciprocal crossings: CSW females ϫCSR males (group A), 72.5% (29/40) and CSR femalesϫCSW males (group B), 15.1% (13/86) (c 2-test, pϽ0.05). The mating frequency of group A was not significantly different from that of CSR or CSW, while the mating frequency of group B was significantly lower than the frequencies of both Fig. 8. Onset time of mating in the F2 of rice- (CSR) and parental populations (c 2-test, pϽ0.05). water-oats- (CSW) populations of C. suppressalis (F2A; ϭ ϭ In group A, the onset time of mating occurred 5 n 52, F2B; n 49). For a detailed explanation of F2A and F2B, see Materials and Methods. to 9 h into the scotophase (7.0Ϯ0.2 h; meanϮSE, nϭ29; Fig. 6), which was relatively closer to the onset time of mating in CSW (7.4Ϯ0.1 h). How- significantly between F1A (65.4%; 17/26) and F1B ever, it was significantly different from that of ei- (76.0%; 19/25) (c 2-test, pϾ0.05). ther of the parental populations (Mann-Whitney U- The onset time of mating of the F2 generation test, pϽ0.05). On the other hand, in group B, the showed a very similar temporal pattern to those of onset time of mating showed an irregular pattern, the F1 generation (Fig. 8). Mating in F2A occurred probably because the mating frequencies were too from 3 to 7 h into the scotophase with the mean at low (7.6Ϯ0.4 h, nϭ13) (Fig. 6). 5.2Ϯ0.1 h (meanϮSE, nϭ52). Mating in F2B The F1 insects initiated mating at a time inter- began at 5.2Ϯ0.2 h, ranging from 2 to 8 h into the mediate between CSR and CSW (Fig. 7). Mating scotophase (nϭ49). No significant differences were of F1A began at 5.8Ϯ0.1 h (meanϮSE, nϭ17), shown in the onset time of mating between F2A ranging from 4 to 7 h into the scotophase. Simi- and F2B (Mann-Whitney U-test, pϾ0.05). How- larly, for F1B, mating began at 5.3Ϯ1.4 h ranging ever, significant differences were found in the onset from 2 to 8 h (nϭ19). There was no significant dif- time of mating between F2 and both CSR and ference in the onset times of mating between F1A CSW (Mann-Whitney U-test, pϽ0.05). Mating fre- and F1B (Mann-Whitney U-test, pϾ0.05). The quencies were not significantly different between onset of mating in both F1 groups occurred at sig- F2A (51.5%; 52/101) and F2B (62.0%; 49/79) (c 2- nificantly different times from those of the parental test, pϾ0.05). populations, CSR and CSW (Mann-Whitney U- test, pϽ0.05). Mating frequencies did not differ Differences in Mating Time of Chilo 261

rence of the two populations at the same time in DISCUSSION the same area and competitive attractiveness of the Mating of CSR was mostly observed in the first female sex pheromone with that of CSW. Further half of the scotophase, whereas that of CSW was investigations are needed to examine the possibility observed in the second half of the scotophase. This of hybridization between these two populations pattern was similar to that reported by Konno and under field conditions. Tanaka (1996b) for C. suppressalis collected in Mating frequency between CSW females and Okayama Pref. (ca. 500 km west from Tokyo). CSR males was significantly higher than that be- Moreover, a similar pattern was observed between tween CSR females and CSW males. This differ- CSR and CSW in Gifu Pref., located just between ence may be related to the temporal difference in Tokyo and Okayama (Tsuchida, personal commu- calling activity between CSR and CSW females nication). The delayed mating time in CSW, there- (Fig. 3). However, based on the calling period of fore, is suggested to be a common feature. In C. CSR females (Fig. 4) and the mating activity of suppressalis, it is possible that this species has split CSW (Fig. 1), the mating frequency between CSR or is on the way to split into two distinct popula- females and CSW males during 6–8 h into the sco- tions that are reproductively isolated according tophase in the present study (Fig. 6) appeared to be to differences in their mating times. A similar lower than expected. Because no major differences reproductive isolation between host strains in were observed between the populations in the fe- Spodoptera frugiperda has been reported (Pashley male pheromone composition and their blend ratios et al., 1992). Temporal partitioning was shown in (Samudra, unpublished data), other factors, e.g. the mating activities of a corn strain and a rice physiological changes in the females leading to the strain, the corn strain mated in the first two thirds decrease in male receptivity during the late calling of the scotophase, while the rice strain mated in the period, may be involved in the lower mating fre- last one third. The authors suggested that the quency of CSR females and CSW males. Similar strongest barrier for interbreeding between the two circumstances have been reported for two pairs of strains was the temporal difference in nocturnal sibling species. Crossing between Spodoptera lati- mating activities. fascia males and S. descoinsi females showed a Observations of the onset of female calling be- lower mating frequency than that of the opposite havior in CSR and CSW showed the same pattern cross (Monti et al., 1995). The mating activity of S. of temporal differences found in mating between latifascia was observed in the last half of the sco- the populations. This suggests that the major cause tophase, while that of S. descoinsi occurred in the of the temporal change in mating is calling activity, first half. In the case of two plume species, since mating behavior in males could never occur the mating activity of Platyptilia carduidactyla was without pheromone release by the female. In ad- observed in the first half of the scotophase and that dition, the sexually active period of CSR females of P. williamsii in the second half (Haynes and was longer than that of CSW females, because a Birch, 1986). Interspecific copulation between P. smaller, but significant, number of virgin females williamsii females and P. carduidactyla males oc- of CSR showed calling behavior even in the second curred, but not in the reciprocal direction. We sug- half of the scotophase. The mating times of the two gest, therefore, that for male moths, in general, it is populations also overlapped. Between the popula- rather difficult to mate earlier than in the normal tions, therefore, there was a time overlap in calling time, whereas it may be easy to delay their mating activity, suggesting that there is a chance for CSR time. females to stimulate the CSW males, even if only The peak mating times of F1s were mid-way be- for a short period. This suggests that between the tween those of CSR and CSW. We also found that populations there is a little chance to hybridize. In- the peak of mating times of F2s were not signifi- deed, cross-mating occurred under laboratory no- cantly different from those of F1s. These results choice conditions and hybrids were produced. show that the onset time of mating is genetically Under natural conditions, however, other possible controlled, most likely by a polygenic system. factors may be necessary for CSR females to at- Monti et al. (1997) reported that in S. latifascia and tract and mate with CSW males, e.g. the occur- S. discoinsi the timing of calling initiation in fe- 262 I Made Samudra et al. males was likely to be under polygenic control. Maki, Y. and M. Yamashita (1956) Ecological difference of Further studies on differences between CSR and rice stem borer, Chilo suppressalis Walker in the various CSW, such as in oviposition preference, female host plant. Bull. Hyogo Pref. Agric. Exp. Stn. 3: 47–50 (in Japanese). pheromone system, diapause physiology and natu- Marumo, N. (1930) On the water-oats stem borer. Oyo-Dobu- ral enemies, will contribute to a better under- tsugaku-Zasshi 2: 91–95 (in Japanese). standing of the speciation process and the selection Miyahara, K. and Y. Abe (1969) Occurrence of the rice stem pressure leading to it. borer on the water-oats in creeks. Proc. Assoc. Plant Prot. Kyusyu 15: 123–124 (in Japanese). ACKNOWLEDGEMENTS Monti, L., J. Genermont, C. Malosse and B. Lalanne-Cassou (1997) A genetic analysis of some components of repro- We thank Dr. Yasuhiko Konno, Tohoku University, for his ductive isolation between two closely related species, kind guidance on insect rearing and Dr. Kei Kawazu for his Spodoptera latifascia (Walker) and S. descoinsi (Lalanne- kind help editing the figures. Thanks are also due to Dr. Atsu- Cassou and Silvain) (Lepidoptera: Noctuidae). J. Evol. shi Naito, a previous JICA expert, for his encouragement to Biol. 10: 121–134. I M.S. Monti, L., B. Lalanne-Cassou, P. Lucas, C. Malosse and J. F. REFERENCES Silvain (1995) Differences in sex pheromone communi- cation systems of closely related species: Spodoptera lati- Haynes, K. F. and M. C. Birch (1986) Temporal reproductive fascia (Walker) and S. descoinsi (Lalanne-Cassou and isolation between two species of plume moths (Lepi- Silvain) (Lepidoptera: Noctuidae). J. Chem. Ecol. 21: doptera: Pterophoridae). Ann. Entomol. Soc. Am. 79: 641–660. 210–215. Pashley, D. P., A. M. Hammond and T. N. Hardy (1992) Re- Kalshoven, L. G. E. (1981) Pests of Crops in Indonesia (re- productive isolating mechanism in fall armyworm host vised by P. A. Van der Land). P. T. Ichtiar Baru-Van strain (Lepidoptera: Noctuidae). Ann. Entomol. Soc. Am. Hoeve, Jakarta. 701 pp. 85: 400–405. Kamano, S. and Y. Sato (1985) Chilo suppressalis. In Hand- Takai, A. (1970) Variation of the head width in adult of the book of Insect Rearing II (P. Singh and R. F. Moore eds.). rice stem borer captured by light trap. Bull. Ibaraki Agric. Elsevier, Tokyo, pp. 219–227. Exp. Stn. 11: 53–56 (in Japanese). Kanno, H. (1979) Effects of age on calling behaviour of the Takasaki, T., M. Noda and Z. Murata (1969) Ecology of rice rice stem borer moth, Chilo suppressalis (Walker) (Lepi- stem borer moth on Zizania latifolia. I. Occurrence of doptera: Pyralidae). Bull. Entomol. Res. 69: 331–335. rice stem borer on Zizania latifolia. Proc. Assoc. Plant Kiritani, K. (1988) What has happened to the rice borers dur- Prot. Kyushu 15: 118–121 (in Japanese). ing the past 40 years in Japan? JARQ 21: 264–268. Tatsuki, S., S. Atsusawa, K. Uchiumi, M. Kurihara and J. Koike, K., K. Tamura and Y. Saito (1981) Potential of the Fukami (1975) Sex pheromone of the rice stem borer water-oats, Zizania latifolia Turcz, as an alternative host moth, Chilo suppressalis Walker (Lepidoptera: Pyrali- plant of the rice stem borer, Chilo suppressalis Walker, in dae). I. Laboratory mating behavior. Botyu-Kagaku 40: Niigata Prefecture. Proc. Assoc. Plant Prot. Hokuriku 29: 143–150. 28–31 (in Japanese). Tatsuki, S. and J. Fukami (1972) Field mating behavior of the Konno, Y. and F. Tanaka (1996a) Aliesterase isozymes and rice stem borer, Chilo suppressalis Walker (Lepidoptera: insecticide susceptibility in rice-feeding and water-oat- Pyralidae). Kontyû 40: 143–150 (in Japanese with Eng- feeding strains of the rice stem borer, Chilo suppressalis lish summary). Walker (Lepidoptera: Pyralidae). Appl. Entomol. Zool. Tsuchida, K. and H. Ichihashi (1995) Estimation of moni- 31: 326–329. toring range of sex pheromone traps for the rice stem Konno, Y. and F. Tanaka (1996b) Mating time of the rice-feed- borer moth, Chilo suppressalis (Walker) (Lepidoptera: ing and water-oat-feeding strains of the rice stem borer, Pyralidae) by male head width variation in relation to Chilo suppressalis (Walker) (Lepidoptera: Pyralidae). two host plants, rice and water oats. Appl. Entomol. Zool. Jpn. J. Appl. Entomol. Zool. 40: 245–247 (in Japanese 30: 407–414. with English summary). Uchiumi, K. (1974) Mass rearing method of the rice stem Kurihara, S. (1930) The water-oats stem borer examined borer, Chilo suppressalis W. (Lepidoptera: Pyralidae). through the chromosome. Oyo-Dobutsugaku-Zasshi 2: Bull. Inst. Physic. Chem. Res. 50: 70–74. 144–145 (in Japanese).