Monarchs in the Classroom

Environmental Factors Influencing Postdiapause Reproductive Development in Monarch Butterflies

We studied diapause development in female monarchs collected near the end of the overwintering period in Mexico. At this time, most females had undeveloped ovaries, but many had mated. Females collected in copula were more likely to have oocytes present, and to have mated previously. When exposed to warm temperatures, females developed mature oocytes rapidly, with most becoming fully reproductive in 7 to 12 days, suggesting that they were in a state of postdiapause quiescence at the time of collection. Exposure to an increasing day length treatment did not affect the rate of reproductive development, but access to host plants and mating speeded development. In addition, females exposed to host plants were more likely to mate. This response to environmental cues may serve to synchronize oogenesis with oviposition opportunities.

Many researchers have noted a lack of reproductive activity and undeveloped reproductive organs in overwintering monarchs, and referred to this as reproductive diapause (Urquhart and Urquhart 1976; Brower et al. 1977). Herman (1981) reported a period of minimal reproductive tract development in monarchs collected in California overwintering sites from September to December even after they were incubated under conditions that are normally favorable to reproductive development (16h:8h light:dark; 25°C), clearly demonstrating the refractory period necessarily associated with diapause. Subsequent work comparing California and Mexico overwintering butterflies revealed qualitatively similar diapause patterns, with longer diapause durations in the eastern population (Herman et al. 1989). More recent work in Mexican overwintering sites followed male reproductive development and mating strategies (Van Hook 1993, 1996). Van Hook identified significant variation in the degree of male diapause and suggested that this leads to multiple mating strategies in the overwintering colonies. In Australia, monarch ovarian dormancy is more accurately classified as oligopause, an intermediate condition that lacks the refractory period associated with diapause (James 1982). Goehring and Oberhauser (2002) reported on environmental factors influencing monarch diapause induction. Here, we report studies of the effects of photoperiod, access to milkweed, and mating on diapause completion and postdiapause ovarian development in eastern population females.

We collected adult monarchs from the Sierra Chincua colony, located in the transvolcanic mountains of Central Mexico, during the 1996 and 1997 overwintering seasons. In both years, we collected during a time of increased activity (e.g. flying, drinking, nectaring and mating), but before mass dispersal for spring migration. We collected 324 females and 72 males on 28 February 1996, and 300 females and 100 males on 5 March 1997. These dates roughly correspond to the average start of the mass mating period (Van Hook 1996).

We sampled butterflies from three hypothesized sub-populations - roosting, mating and active - following Van Hook (1996). Roosting butterflies were collected from trees in the center of the colony. These butterflies were hanging immobile on tree trunks or branches 3.5 to 6 m from the ground. Mating butterflies were collected in copula, also in the colony center. Active butterflies were collected in flight in a meadow or near a stream downhill from the colony. Collections from the three populations were made at the same time. In 1996, equal numbers of adults were collected from each sub-population. In 1997, to ensure male mating readiness, all males were collected in copula. In 1997, the nucleus of the colony moved downhill on the day prior to collection, possibly mixing sub-populations.

We marked butterflies at collection and then placed them in glassine envelopes. Mating pairs were stored together in the same envelope and allowed to separate on their own. Butterflies were kept on ice in a cooler and transported to St. Paul, Minnesota within 3 days of collection.

We measured butterfly mass, wing length and wing condition (scale loss and tatter along the edges) immediately upon arrival in Minnesota. We assessed scale loss on an ordinal scale from 1 to 5 (no loss to severe loss) and wing edge tatter on an ordinal scale from 0 to 4 (no wing tatter to all 4 wings tattered). Scale loss may approximate age and wing tatter may indicate distance traveled, damage from mating struggles, or encounters with predators (Frey and Leong 1995; Van Hook 1996).

Females in reproductive diapause have small ovarioles with no ovarian development (Herman 1973). We examined ovaries for the presence of unyolked, yolked and mature (chorionated) oocytes, and counted mature oocytes, if present (for details on dissection methods see Oberhauser and Hampton 1995). Although we measured and report all of the stages in diapause development, we used the presence of mature oocytes as the indication that diapause was completed.

We determined mating status by examining the contents of the bursa copulatrix for spermatophores and spermatophore remains (Van Hook 1996). Thus we had two indicators of mating; mating activity at collection (i.e. "mating" if collected in copula) and mating status at dissection (i.e. "mated" if female had one or more spermatophores).

To establish a baseline on degree of ovarian development and mating status at the time of collection, we dissected a random sample of females from each activity category (active, mating and roosting) (1996 n = 20 each category; 1997 n = 16 each). These butterflies were dissected on day 0 of the experiments. Remaining butterflies were used in environmental chamber experiments designed to test the hypothesis that exposure to certain environmental cues (host plant access, increasing day length and mating) will stimulate postdiapause ovarian development.

1996 Experiment: Host plant and daylength. In 1996, we tested the effects of photoperiod and access to milkweed on postdiapause reproductive development. Butterflies in the overwintering colony experience a day length of 11 hours and 45 minutes at the end of February; this increases with time and northward movement. By early April, monarchs in the southern US (~30°N latitude) experience 12 hours and 30 minutes of daylight, an increase of 45 minutes or a gain of over a minute per day. To simulate these changes, we used programmable timers on standard shop florescent fixtures with 40W fluorescent tubes to control daily light:dark cycles, increasing the amount of day length 2 minutes per day. To test the effect of access to milkweed on ovarian development, we used potted Asclepias curassavica in host plant treatments and potted Cassia fasciculata (Partridge Pea) in non-host plant treatments.

We used a 2 X 2 factorial design, testing increasing day length (12:12 L:D hours +2 minute / day) vs. constant day length (12:12 L:D hours), and access to milkweed vs. no access. We used two controlled-environment growth chambers (4 x 3 x 2 m), each sub-divided with black opaque (6 mil) plastic sheeting to allow two light treatments in the same chamber (2 x 3 x 2 m each). Milkweed and non-milkweed treatments were in separate chambers.

Sixty females (20 randomly selected from each activity category) and twelve males were put in each treatment, divided evenly into two net cages (0.75 m³). We kept temperatures at 24 ± 1°C, which is suitable for ovarian development (Barker and Herman 1976, Malcolm et al. 1987). Adults were fed ad libitum from sponges soaking in honey-water (20% honey). Relative humidity was 59 ± 3.3% in the milkweed chamber and 60 ± 2.7% in the non-milkweed chamber.

To assess ovarian development, we randomly sampled one fourth of the females from each activity category on days 1, 3, 7 and 12 for dissection.

1997 Experiment: Host plant and mating. After observing an ad hoc effect of mating on ovarian development in 1996, we explicitly tested effects of mating in 1997. We omitted the photoperiod treatment after observing no effect of this independent variable, but included a host plant access variable. We conducted a 2 X 2 factorial experiment, with access to males and vs. no access to males, and potted A. physocarpa for host plant treatments vs. potted Leucanthemum x superbum (chrysanthemum) in non-host plant treatments. Temperature and humidity were the same as in 1996, and the photoperiod was 12h:12h L:D.

We used 63 females in each treatment (21 randomly selected from each activity category) and 40 males in mating treatments. To keep cage densities relatively even, we divided mating treatments into two cages. We recorded mating daily. Butterflies were fed as before, and temperature and humidity were similar to the previous experiment. On days 2, 4, 6 and 8, we dissected one fourth of the females from each activity category. We dissected these as before with the exception of more precise examination of the bursa copulatrix to count the number of spermatophores present.

We assessed postdiapause development by presence or absence of mature oocytes (a yes / no, or binary, variable) and used stepwise analysis of deviance and logistic regression models to examine the relationship between oocyte presence and the independent variables of sub-population, butterfly size and wing condition, mating status, and experimental treatment (Hardy and Field 1998; K. Chaloner, pers. comm.). Probabilities of 0.05 or less indicate statistical significance.

Figure 4 illustrates the progression of ovarian development for 1996 females. Across all treatments, 59% (36 of 61) of the females had some degree of ovarian development after 1 day, compared with 30% (Figure 1) of the females from the baseline group. By days 7 and 12, over 77 and 100% had mature oocytes, respectively. Although our experimental design did not explicitly test for an effect of mating, we determined whether females had mated when we dissected them and included their mating status as a predictor in our model. Table 1 summarizes the analysis of postdiapause ovarian development, as measured by the presence of mature oocytes.

Table 1. Effects of milkweed access and photoperiod on mature oocyte development

In addition to the experimental variables of photoperiod and host plant access, we tested the effects of activity at collection, mating status, mass and wing length. The most parsimonious model for predicting mature oocyte presence includes days in treatment, mating status and access to milkweed. The odds of full reproductive development in mated females, given access to milkweed, are eleven times the odds for unmated females. The odds of reproductive development in females with access to milkweed, given having mated, are five times the odds for females without access to milkweed. There were no significant effects of photoperiod, activity at collection, massor wing length .

Host plant access and mating

Figure 5 shows the progression of ovarian development in 1997. A greater proportion of females with access to milkweed and males developed mature oocytes than those without access to milkweed or males. A large proportion those in treatments without milkweed or males were still undeveloped 6 days into the experiment and the majority contained only yolked oocytes by the end of the experiment. Other treatments reveal intermediate patterns of ovarian development.

Table 2 summarizes the analysis of post-diapause ovarian development in the 1997 experiment. We tested the effects of access to milkweed and males, activity at collection, mating status, mass and wing length. The resulting model includes days in treatment, access to milkweed, mating status, and activity at collection. Access to milkweed had the greatest effect on postdiapause ovarian development, followed by mating status, and activity at collection. Figure 6 shows the proportion of females with mature oocytes over time by milkweed treatment and mating status. Most unmated females without access to milkweed remained undeveloped.

Table 2. Effects milkweed and mating access on mature oocyte development

Predictor	Coefficient	SE	p	Log-odds	95% CI
Constant	-26213.0	3895.9	< 0.001
Days in treatment	.7383	.110	<0.001	2.18	1.74 - 2.73
Access to host plant	2.919	.459	<0.001	19.60	7.79 - 49.35
Mated	1.826	0.399	<0.001	6.42	2.88 - 14.3
Mating at collection	1.155	0.418	0.006	3.17	1.40 - 7.21

Deviance	178.29
d.f.	237

Summary of binomial regression model for 1997 experiment

These variables are not independent. Access to milkweed has an effect on the probability of mating within the experiment. In treatments with males, more females mated when milkweed was present than when it was not present: 55 out of 62 (89%) and 43 out of 60 (72%) respectively, *2=5.6, p=0.018. They also mated more frequently (Table 3, KW statistic H=20.66, p<0.001). Although some matings occurred prior to collection, the frequency of mating prior to collection was not different between milkweed and non-milkweed treatments (25% and 23% respectively, *2 = 0.11, p=0.741).

Table 3. Mating Frequency in milkweed vs. non-milkweed treatments

Number of Spermatophores	Milkweed Treatments	Non-Milkweed Treatments
0	7	17
1	11	22
2	19	16
3	10	4
>3	15	1

Only females in treatments with males are included.

Discussion

Diapause development is a progressive physiological process. While diapause induction is initiated in response to particular cues (e.g., Goehring and Oberhauser 2002), diapause development typically follows a predetermined course not necessarily terminated by specific cues (Tauber et al. 1986). The progression may be influenced by external factors that speed completion of diapause, and response to these factors often varies as a function of diapause intensity. Given the wide geographic range and environmental conditions experienced by North American monarchs, diapause intensity is likely to be variable, making characterization of diapause development complex.

Monarchs used in this study were collected during the last third of the overwintering period, a time of increasing mating activity. Mating dynamics change dramatically during the overwintering period, from very little mating in early November to mass mating during the last 6 weeks in the Mexican colonies (Brower et al. 1977; Brower 1985; Calvert and Brower 1986;Van Hook 1996). The onset of mating is quite consistent from year-year; mass-mating begins in mid February and increases until monarchs disperse from the overwintering sites (Van Hook 1996). Although mating marks the end of diapause in males, it does not necessarily signify the same in females because they can be forced to mate. Females with undeveloped ovaries during this period may be either nearing completion of diapause, or are in a postdiapause quiescence awaiting suitable conditions for reproduction.

Status of diapause development and mating at the time of collection

In both years, most females showed no ovarian development at the time of collection. Herman et al. (1989) reported similar findings for monarchs obtained in March 1983 and 1984 from two Mexican sites. However, we found that oogenesis had commenced in some females in late February and early March. Females with ovarian development were most likely to be collected while mating (Figure 1) or to have mated prior to collection (Figure 3), suggesting an association between ovarian development and mating. However, 15% of unmated females in 1996 showed ovarian development, indicating that mating is not required for postdiapause development.

The fact that more females had produced oocytes in 1996 than in 1997 suggests that females were further along in post-diapause development in 1996, even though they were collected earlier. Previous work has shown that the course of diapause is primarily influenced by temperature (Hodek 1983; Tauber et al. 1986; Danks 1987); unfortunately temperature data were not available for comparison.

Postdiapause ovarian development in experimental treatments

Ovarian development progressed rapidly under experimental conditions. While a significant proportion of females without access to milkweed or males remained undeveloped, most of those exposed to milkweed and allowed to mate developed mature oocytes within 3 days (Figures 4-6). This rapid development suggests that females were in a state of postdiapause quiescence (Tauber et al. 1986) awaiting stimuli to activate and accelerate postdiapause morphogenesis. Mating and access to milkweed appear to serve as such stimuli.

Given the importance of synchronizing reproduction with host plant availability, it is not surprising that milkweed is an external stimulus for postdiapause development in monarchs. Similarly, the presence of host plant Vigna unguiculata pods stimulates diapause termination in the seed beetle Bruchidius atrolineatus (Tran et al. 1993; Lenga et al. 1993; Glitho et al. 1996), as does host plant presence for the leek moth, Acrolepiopsis assectella, (Abo-Ghalia and Thibout 1983).

Although mating has been shown to stimulate the rate of non-diapause oocyte production in monarchs, it is not required for oogenesis (Herman and Barker 1977; Oberhauser and Hampton 1995; for other insects see review in Barth and Lester 1973). Few studies have assessed the effect of mating on postdiapause ovarian development (review in Danks 1987). In A. assectella, Abo-Ghalia and Thibout (1983) found that along with host plant, mating significantly increases the postdiapause mean number of oocytes, and that these factors have a synergistic effect. In our study, mating was associated with postdiapause oogenesis, both in the baseline analysis and in the experimental treatments. However, our results do not allow us to distinguish whether mating stimulates diapause completion or if it just results in faster postdiapause development.

Photoperiod

Although photoperiod has been shown to affect diapause completion, it does not usually affect postdiapause development (e.g., McNeil and Fields 1985; Tanaka and Sadoyama 1997). An increasing daylength, simulating that experienced by northward migrants, did not affect monarch postdiapause ovarian development in our 1996 experiment. Photoperiod may affect diapause termination, but our study did not address that question.

Variation in diapause development

Given the effect of mating on postdiapause ovarian development, it is instructive to consider male diapause development. While the course of insect diapause is influenced by environmental conditions experienced by both sexes, it does not always proceed identically in males and females. In general, it has been postulated that given the greater energy requirement to maintain eggs vs. sperm and the smaller metabolic change required to enter diapause in males, male diapause is usually less intense and of shorter duration (Danks 1987). In monarchs, diapause appears to last longer in females than males (Herman 1981; Herman et al. 1989; Lessman and Herman 1983). Wiklund et al. (1992) proposed that the difference in the timing of diapause between sexes is an example of protandry, where males are expected to benefit by emerging or being ready to mate before females; Nylin et al. (1995) suggested this is the case in overwintering monarchs. Males may complete diapause earlier and begin mating at the overwintering sites to maximize their number of matings.

Diapause termination typically occurs over a considerable time span subject to individual and environmental variation, even in species with synchronous cohorts. Monarch overwintering populations are comprised of individuals coming from a wide geographic area, subjected to a wide range of environmental conditions. The range of physiological ages and diapause intensities could lead to an even greater span of time for diapause termination. In addition, there are probably complicated relationships among postdiapause development and mating. The fact that females with access to milkweed mated more suggests that females may be more amenable to male mating attempts when they contain mature oocytes.

While the factors that trigger monarch diapause development remain to be determined, our results suggest that postdiapause reproductive development are strongly influenced by external factors such as host plant availability and mating, serving to synchronize reproduction with the seasonal availability of breeding habitat.

Acknowledgements

We thank Alfonso Alonso-M., Eduardo Rendon-S. and Eneida Montesinos-P. for the many ways they helped with this work; Tonya Van Hook for important insights on sampling; Michelle Prysby, Kari Geurtz and Sonia Altizer for help in collection; Amy Alstad and Leah Alstad for help measuring and recording data; and Imants Pone for laboratory assistance. This work was supported in part by a National Science Foundation grant to KO (DEB-9220829), Monarchs in the Classroom and a James W. Wilkie Award to EG.

References

Abo-Ghalia, A. and E.Thibout. 1983. The influence of the host plant (Allium porrum) on the reproductive activity of the female leek moth, Acrolepiopsis assectella, after the termination of its reproductive diapause. Entomol. Exp. Appl. (French) 33:188-194.

Andrewatha, H. C. 1952. Diapause in relation to the ecology of insects. Biol. Rev. 27:50-107.

Barker, J. F. and W. S. Herman. 1976. Effect of photoperiod and temperature on reproduction of the Monarch butterfly, Danaus plexippus. J. Insect Physiol. 22:1565-1568.

Barth, R. H. and L. J. Lester. 1973. Neuro-hormonal control of sexual behavior in insects, Annu. Rev. Entomol. 18:445-472.

Brower, L. P. 1985. New perspectives on the migration biology of the monarch butterfly, Danaus plexippus L. In M. A. Rankin, ed., Migration: Mechanisms and Adaptive Significance. pp. 748-785. Contributions in Marine Science, 27 Suppl. Port Aransas, Texas: Marine Science Institute, The University of Texas at Austin.

Brower, L. P., W. H. Calvert, L. E. Hendrick and J. Christian. 1977. Biological observations of an overwintering colony of monarch butterflies (Danaus plexippus, Danaidae) in Mexico. J. Lepid. Soc. 31(4):232-242.

Calvert, W. H. and L. P. Brower. 1986. The location of monarch butterfly (Danaus plexippus L.) overwintering colonies in Mexico in relation to topography and climate. J. Lepid. Soc. 40(3):164-187.

Danks, H. V. 1987. Insect Dormancy: An Ecological Perspective. Ottawa, Canada: Entomological Society of Canada.

Frey, D. F. and K. L. H. Leong. 1995. Reply to Nylin, Wickman and Wiklund regarding sex ratios of California overwintering monarch butterflies. Anim. Behav. 49:515-518.

Glitho, I. A., A. Lenga, D. Pierre, and J. Huignard. 1996. Changes in the responsiveness during two phases of diapause termination in Bruchidius atrolineatus Pic (Coleoptera: Bruchidae) J. Insect Physiol. 42(10):953-960.

Goehring, L. and Oberhauser, K. S. 2002. Effects of photoperiod, temperature and host plant age on induction of reproductive diapause and development time in Danaus plexippus. Ecol. Entomol. In press.

Hardy, I. C. and S. A. Field. 1998. Logistic analysis of animal contests. Anim. Behav. 56:787-792.

Herman, W. S. 1973. The endocrine basis of reproductive inactivity in monarch butterflies overwintering in central California. J. Insect Physiol. 19:1883-1887.

Herman, W. S. 1981. Studies on the adult reproductive diapause of the monarch butterfly, Danaus plexippus. Biol. Bull. 160:89-106.

Herman, W. S. and J. F. Barker. 1977. Effect of mating on monarch butterfly oogenesis. Experientia. 33:688-689.

Herman, W. S., L. P. Brower and W. H. Calvert. 1989. Reproductive tract development in monarch butterflies overwintering in California and Mexico. J. Lepid. Soc. 43(1):50-58.

Hodek, I. 1983. Role of environmental factors and endogenous mechanisms in the seasonality of reproduction in insects diapausing as adults. In K. V. Brown and I. Hodek, eds.), Diapause and life cycle strategies in insects. pp. 9-33. Dr W. Junk Publishers, The Hague.

James, D. G. 1982. Ovarian dormancy in Danaus plexippus (L.) (Lepidoptera Nymphalidae)-Oligopause not diapause. J. Aust. Entomol. Soc. 21:31-35.

Leather, S. R., K. F. A. Walters and J. S. Bale. 1993. The Ecology of insect overwintering. Cambridge: University Press.

Lenga, A., I. Glitho, and J. Huignard. 1993. Interactions between photoperiod, relative humidity and host-plant cues on the reproductive diapause termination in Bruchidius atrolineatus Pic (Coleoptera Bruchidae). Inverteb. Reprod. Dev. 24(2):87-96.

Lessman, C. A. and W. S. Herman. 1983. Seasonal variation in hemolymph juvenile hormone of adult monarchs Danaus p. plexippus (Lepidoptera). Can. J. Zool. 61:88-94.

Malcolm, S. B., B. J. Cockrell, and L. P. Brower. 1987. Monarch butterfly voltinism: Effects of temperature constraints at different latitudes. Oikos. 49:77-82.

McNeil, J. N. and P. G. Fields. 1985. Seasonal diapause development and diapause termination in the European skipper, Thymelicus lineola Ochs. J. Insect Physiol. 31(6):467-470.

Nijhout, H. F. 1994. Insect Hormones. Princeton, New Jersey: Princeton University Press.

Nylin, S., P. Wickman, and C. Wiklund. 1995. An adaptive explanation for male-biased sex ratios in overwintering monarch butterflies. Anim. Behav. 49:511-514.

Oberhauser, K. S. and R. Hampton. 1995. The relationship between mating and oogenesis in monarch butterflies (Lepidoptera: Danainae). J. Insect Behav. 8(5):701-713.

Tanaka, S. and Y. Sadoyama. 1997. Photoperiodic termination of diapause in field-collected adults of the Bombay locust, Nomadacris succincta (Orthoptera: Acrididae) in southern Japan. Bull. Entomol. Res. 87:533-539.

Tauber, M. J., C. A. Tauber, and S. Masaki. 1986. Seasonal Adaptations of Insects. New York: Oxford University Press.

Tran, B., J. Darquenne, and J. Huignard. 1993. Changes in responsiveness to factors inducing diapause termination in Bruchus rufimanus Boh. (Coleoptera: Bruchidae). J. Insect Physiol. 38(9):769-774.

Urquhart, F. A. and N. R. Urquhart. 1976. The overwintering site of the eastern population of the monarch butterfly (Danaus p. plexippus; Danaidae) in southern Mexico. J. Lepid. Soc.30:153-158.

Van Hook, T. 1996. Monarch butterfly mating ecology at a Mexican overwintering site: Proximate causes of non-random mating. Ph.D. Thesis, University of Florida.

Van Hook, T. 1993. Non-random mating in monarch butterflies overwintering in Mexico. In Malcolm, S. B. and M. P. Zalucki, eds., Biology and conservation of the monarch butterfly. pp. 49-60. Los Angeles: Natural History Museum of Los Angeles County.

Wiklund, C., P. Wickman, and S. Nylin. 1992. A sex difference in the propensity to enter direct/diapause development: a result of selection for protandry. Evolution 46(2):519-528.

Host plant access and mating

Number of Spermatophores

Acknowledgements