Seasonal changes in the reproductive and life-history traits of Chrysomela populi L. (Coleoptera: Chrysomelidae)

INTRODUCTION

The timing of diapause induction in insects affects fitness traits, and is an important factor in the evolution of life histories (Taylor 1980; Bradshaw et al. 1998). The effects of season on survival and reproduction in relation to timing of diapause induction have been investigated in several insects (Tauber et al. 1986; Danks 1987). Individuals that enter diapause early in the growing season suffer high mortality in the spruce budworm, Choristoneura fumiferana (Han & Bauce 1997, 1998), and low reproductive output in the fall webworm, Hyphantria cunea (Gomi 2000). In some insects, fecundity is reduced in postdiapause individuals compared with non-diapause individuals (e.g. Fujiie 1980; Akkawi & Scot 1984), but is the same as or higher than that of the non-diapause individuals in others (e.g. Tauber et al. 1970; Soni 1976; Ishihara & Shimada 1995). Reproductive performance is influenced by the duration of diapause in the cabbage beetle, Colaphellus bowringi (Wang et al. 2006), but not in the southern green stink bug, Nezara viridula (Musolin et al. 2007).

The red poplar leaf beetle, Chrysomela populi L. (Coleoptera: Chrysomelidae), is widely distributed in Asia, and damages the young leaves of its host plants, Populus spp. and Salix spp., in the larval and adult stages (Kimoto & Takizawa 1994). Adults of C. populi are observed from early April to late October in Shobara (34°52′N, 133°01′E), Honshu, Japan (Gomi et al. 2006). In the spring, after feeding for a few weeks, adult females that have overwintered deposit eggs in a clutch on the undersides of leaves. In a laboratory experiment, one female that averted diapause at 25°C produced an average of 14 egg clutches, each containing approximately 50 eggs (Gomi & Suzuki 2003). The mean longevity of adults has been found to be longer than 50 days and to not differ significantly between sexes at 25°C (Gomi & Suzuki 2003). Larvae feed in a group during the early instars, and then gradually disperse. Such group feeding behavior is commonly observed in chrysomelid beetles (Breden & Wade 1985). Mature larvae of the third instar pupate on the undersides of the leaves of the host or neighboring plants. Adults that emerge in autumn enter diapause and overwinter under litter around the host plant after feeding (Okuda 1994).

The life cycle of this species in Japan is not yet completely understood. Kimoto and Takizawa (1994) inferred univoltine or bivoltine life cycles, but Okuda (1994) suggested trivoltine or tetravoltine life cycles. Our previous studies have shown that the lower threshold temperature for development and the thermal constant for one generation are 9.0°C and 614 degree-days, respectively (Gomi et al. 2005), and that the critical photoperiod for diapause induction at 25°C is approximately 14 h (Gomi et al. 2006). These life-history traits and climate data suggest that C. populi enters diapause in mid-August or later and has two or three generations per year in Shobara (Gomi et al. 2005, 2006). The effects of temperature on survival and reproduction were clear in non-diapause adults of the Shobara population of C. populi, and temperatures above 25°C reduced these traits in the laboratory (Gomi et al. 2005). Therefore, the higher temperatures in summer may affect the survival and reproduction of this beetle in the field.

In the present study, the timing of diapause induction in C. populi under natural photoperiod and temperature conditions was studied in Shobara. In addition, the effects of season on survival and reproduction were investigated in relation to diapause induction. To investigate these reproductive and life-history traits, C. populi hatching at various times in the 2004 season were reared under natural photoperiod and temperature conditions in Shobara.

MATERIALS AND METHODS

Eight female and 11 male adult C. populi were collected in late September on the campus of Hiroshima Prefectural University in Shobara in 2003. The adults were reared on young leaves of Populus nigra L. in transparent plastic cups (500 mL). These adults entered diapause shortly after collection and were placed outdoors under a trestle bridge. The area was 4 m × 12 m in size, bounded to the east and west by 7 m walls, and with the north and south aspects open. To the south, there was a building approximately 40 m away from where the beetles were kept. Therefore, the beetles were not directly exposed to sunlight. This area was also used for experiments. All adults except one male overwintered successfully. These individuals started feeding in early April and oviposited in mid-April in 2004. The progeny of these adults were reared at 25 ± 0.5°C under conditions of 16 h light : 8 h dark (LD 16:8). Hatchlings from egg clutches kept at 25°C under this photoperiod in the laboratory were used for experiments.

Table 1 summarizes the experimental schedule and the abbreviations used for the experimental groups. The number of hatchlings in each group and the dates of hatching and adult emergence are also given in Table 1. For each group, hatchlings that originated from three to five egg clutches that hatched on the same day were transferred to young leaves of P. nigra in a transparent plastic cup (500 mL) on the day of hatching, and were placed in the outdoor experimental area. This procedure was repeated eight times from late April to early September in 2004 (Table 1). The larvae were reared on young leaves of P. nigra, which were cut at the leafstalk, bound with cotton and put in a glass vial (2 mL) filled with water. The number of adults was counted and survival rates from hatching to adult emergence were calculated in each experimental group. Each adult was weighed on the day of emergence before commencement of feeding. A male and a female that emerged within 3 days of each other were confined in a small transparent plastic cup (200 mL) with young P. nigra leaves that were replenished every 2 days. The number of pairs used in the experiment was 30 for each experimental group. If either of the pair died before oviposition, the pair was excluded from the analyses of incidence of diapause and reproductive traits mentioned below.

For each pair, the day of oviposition was recorded, and the number of deposited and viable eggs in each egg clutch and the number of egg clutches with viable eggs were counted in all experimental groups. Oviposition was observed daily until one individual of the pair died. If the female of a pair laid eggs, new leaves were provided for the pair and either the eggs or the adults were transferred to another cup. All egg clutches were kept at 25 ± 0.5°C under conditions of LD 16:8 and hatching was observed. Hatchlings occasionally showed sib-cannibalism, where they ate eggs that remained unhatched, as known in some other leaf beetles (Jolivet & Verma 2002). Therefore, the number of viable eggs was taken as the number of larvae that hatched on the first day of hatching. The reproductive traits of pairs that overwintered at the experimental site were then investigated in 2005.

Overwintering adults of C. populi were first observed on 15 April 2005 in the field (Gomi et al. 2006). In the present study, survival of the overwintering female adults was checked on 26 April 2005 in each group. The females that survived to this date and deposited eggs thereafter were regarded as diapause females. The females that deposited eggs in 2004 but did not survive until the above-mentioned date were regarded as non-diapause females. The incidence of diapause females was calculated for each group. The oviposition period, defined as the number of days from the first to the last oviposition, and the total numbers of deposited and viable eggs were calculated for each female. To investigate the effects of season and diapause, these reproductive characteristics were compared among experimental groups and between diapause and non-diapause females.

The statistical analyses used in the present study were the Tukey-Kramer test for adult weight, oviposition periods and the number of egg clutches; the Steel-Dwass procedure for the number of viable eggs per female (Dwass 1960; Steel 1960); and the Mann–Whitney U-test for several reproductive traits in diapause and non-diapause individuals. Some data analyzed using the last two methods did not follow a normal distribution according to the Kolmogorov-Smirnov goodness of fit procedure (P < 0.05). The proportional data, such as percentage diapause, were analyzed by Tukey-type multiple comparisons for proportions (Zar 1999). The females that deposited unhatched eggs exclusively were removed from all statistical analyses except incidence of diapause. The climate data for Shobara (natural day length and air temperature) were obtained from the Japan Meteorological Agency.

RESULTS

Incidence of diapause

The incidence of diapause in each experimental group increased with progression of the growing season in 2004 (Fig. 1). All adults in 5L and 6M averted diapause, and the incidence of diapause in 7E and 7M was not significantly different from that in 5L and 5M (Tukey–type multiple comparisons for proportions, P > 0.05). The incidence of diapause in 8M did not differ significantly from that in 8L and 9M, in which all individuals entered diapause (P > 0.05). Based on this result, the critical timing of diapause induction, as defined by 50% of females entering diapause, is estimated to be between late July and early August.

One female in 7L, which emerged on 28 July, produced seven egg-clutches between 17 and 27 August 2004, then entered diapause, and deposited 15 egg-clutches between 2 May and 13 June after overwintering in 2005. This female was regarded as a diapause female, but was removed from comparisons of reproductive traits between diapause and non-diapause females.

DISCUSSION

The critical photoperiod for diapause induction in C. populi, as defined by 50% of individuals entering diapause, was found to be 13 h 59 min at 25°C in the laboratory (Gomi et al. 2006). This critical photoperiod and the natural day length in Shobara suggest that the diapause of this beetle is induced in mid-August, when 0.5 h is added to the time from sunrise to sunset for twilight. In the present study, the critical time by which 50% of individuals had entered diapause was estimated to be between late July and early August. This critical time of diapause induction was very close to mid-August, as suggested using the laboratory data (Gomi et al. 2006), although the present result was approximately 2 weeks earlier than the previous estimate. These results clearly indicate that adults emerging in the first half of August and later enter diapause in Shobara.

In many insects with adult diapause, the stage that is sensitive to photoperiod is the early phase of the adult stage (Danks 1987). Adults of Chrysomela fatuosa, a congener of C. populi, are able to respond to the photoperiod (Danilevsky 1965). In the present study, one female C. populi emerged in late July and deposited egg clutches without any delay in ovarian development in August, and then entered diapause and resumed oviposition after overwintering. This result suggests that the period that is sensitive to photoperiod also occurs during adulthood in C. populi.

The body size of many insects negatively correlates with the temperature at which they developed (Atkinson 1994). Adult body weight at emergence in C. populi also decreases with an increase in the constant temperature from 18 to 29°C in the laboratory (Gomi et al. 2005). In the present study, adults emerging in late May and mid-June were heavier than those emerging in late July and mid-August, and those emerging in the other seasons were intermediate between them. This pattern of weight change agrees well with seasonal fluctuations in the air temperature in Shobara in 2004 (Fig. 2). This agreement suggests that the response of body size to temperature also operates in the field. Photoperiod is another possible environmental factor affecting the body size of insects (Danks 1987). The effect of photoperiod on body size has been reported for the milkweed leaf beetle, Labidomera clivicollis, in relation to diapause (Palmer 1982, 1983), and temperature has no influence on body size in this species (Palmer 1984). In C. populi, adult body weight at emergence became greater with shorter photoperiods under which diapause was induced than under longer photoperiods at 25°C in the laboratory (T. Gomi, unpublished data, 2005). Therefore, the seasonal change in adult weight observed in the present study is caused by temperature rather than photoperiod. Another possible reason for the seasonal change in adult body size of C. populi is changes in the leaf quality of the host plant. In many plants, the quality of leaves decreases over the growing season, because of decreases in protein and water content, and increases in tannin and other toxic chemicals (Slansky 1993; Price 1997). This reduction in food quality may partly contribute to the decrease in body size of the beetle in the summer.

The adverse effects of temperatures above 25°C on the proportion of ovipositing females and the survival rate from hatching to adult emergence have been demonstrated in our previous studies in the laboratory (Gomi et al. 2005). However, the seasonal effects of temperature on these traits were not clearly observed in the present study. The reason for these obscure seasonal effects is probably that the average air temperature rarely exceeded 25°C in 2004 (Fig. 2).

The reproductive output of C. populi was greater for diapause females than for non-diapause females in the present study. A similar result has been found for the Colorado potato beetle, Leptinotarsa decemliniata, in which daily fecundity is higher for postdiapause females than for non-diapause females (Jansson et al. 1989). In C. populi, the average numbers of deposited and viable eggs per egg clutch did not differ significantly between diapause and non-diapause females. Therefore, these traits did not contribute to the increase in the reproductive output of diapause females. In diapause females, the mean oviposition period was twice as long as that of non-diapause females, and the numbers of egg clutches, deposited eggs and viable eggs were approximately 1.5 times the corresponding numbers for non-diapause females. The diapause females commenced oviposition from the beginning of May in 2005, when the air temperature was relatively low compared with the other reproductive periods (Fig. 2). It is likely that the oviposition period in diapause females is extended, probably due to the moderate temperatures in late spring and early summer. Consequently, the number of egg clutches is higher than for non-diapause females, and thus the reproductive output became higher. In a bruchid beetle, Kytorhinus sharpianus, diapause females live significantly longer than non-diapause females, and diapause females deposit more eggs, although the difference was not significant (Ishihara & Shimada 1995). In C. populi, there is the possibility that diapause and non-diapause females differ physiologically, as in the case of K. sharpianus, which may lead to differences in longevity and reproductive output.

Insects that enter diapause too early in the growing season would reduce their reproductive potential (Taylor 1980). When diapause is induced in C. populi in early August, as observed in the present results, a substantial amount of cumulative heat above the lower threshold temperature for development remains in the growing season after diapause induction. In fact, larvae have been observed until mid-October in the field in Shobara (Gomi et al. 2006). However, in one study, the total length of the pre-oviposition and oviposition periods was found to sometimes exceed 60 days after adult emergence at 25°C in the laboratory (Gomi & Suzuki 2003). If female adults emerging in early August and later avert diapause, there is a risk that offspring, especially in the later period of their life, may fail to develop into adults or may not have sufficient energy reserves for overwintering due to shortages of suitable host leaves and cumulative heat for development. The reproductive output of diapause females was greater than that of non-diapause females in the present study. This increase in reproductive output would favor females entering diapause relatively early in the season. The mortality after entering diapause, excluding that from natural enemies, was 7.8% (8/103) of the overwintering pairs in the present study. Thus, the seasonal timing of diapause induction in C. populi would be affected by various risks and benefits to reproduction and survival, in addition to the physiological restrictions on development.