The text below is grayed out because it is not intended to be read. It is a necessarily imperfect OCR of the original and is only used by a search engine.
Journal of the Lepidopterists' Society 32(4), 1978, 282-288
OVER-WINTERING BEHAVIOR IN EUPHYDRYAS PHAETON (NYMPHALIDAE)
M. Deane Bowers Department of Zoology, University of Massachusetts, Amherst, Massachusetts 01003
ABSTRACT. The behavior of the pre-diapause larvae of Euphydryas phaeton (Nymphalidae) associated with over-wintering is described. Observations are based on studies of five wild populations in central Massachusetts over a three-year period, as well as on larvae reared in the laboratory. The functional significance of this larval behavior is discussed.
Most Lepidoptera over-winter as either an egg or pupa; some nym-phalid butterflies (e.g. the Mourning Cloak, Nymphalis antiopa L., and Compton's Tortoiseshell, N. vau-album Boisduval and Leconte), and certain groups of moths [e.g. the Lithophanini (Schweitzer, 1977)] overwinter as adults. In only a relatively few cases do Lepidoptera spend the winter in the larval stage. The Viceroy, Limenitis archippus Cramer (Nymphalidae) for example, is multivoltine and may enter a facultative diapause in the third instar during late summer and fall (Hong and Piatt, 1975). Unusually, however, the Baltimore Checker-spot, Euphydryas phaeton Drury (Nymphalidae), and other members of the genus Euphydryas, are univoltine and exhibit an obligatory diapause in the fourth instar.
Although some descriptive work has been done on the development of the early, pre-diapause instars of E. phaeton, little has been reported of the actual behavior of the larvae during over-wintering. Observations on several wild colonies of E. phaeton and on larvae reared in the laboratory have revealed some interesting behavioral aspects of the over-wintering process in this species.
Methods
Observations were made on five colonies of E. phaeton in Hampshire, Franklin, and Hamden Counties in central Massachusetts from 1974-1977. The site of each colony was wet and marshy, typical habitat for E. phaeton and its primary foodplant, Turtlehead (Chelone glabra L., Scrophulariaceae). The five sites differed in elevation, area, amount of Turtlehead present, and size of the E. phaeton population; but these differences were not factors in the present study.
At each of the colonies, I determined the absolute numbers of egg masses, and (later) larval webs. At the three largest colonies, heights of the pre-hibernation webs above the ground were measured. All
Volume 32, Number 4
283
colonies were usually observed at least once a week. In addition I reared larvae from eggs in the laboratory at 22-25°C under continuous light. Developmental times for each instar, dates of larvae entering the pre-hibernation web and ceasing feeding, and time to the third molt were recorded for these larvae.
Observations
Euphydryas phaeton females usually oviposit two to three large masses of from 100-600 eggs each (Scudder, 1889; Bowers, pers. obs.) from late June through July in Massachusetts. The eggs hatch in about 20 days. When the larvae hatch they may construct a small web on the leaf next to the egg shell but move to the growing tip of the plant within 24 hours and begin web construction and feeding. This feeding web is extended down the length of a stalk of Turtlehead as the larvae devour the leaves. The larvae are gregarious and develop to the end of the third instar in approximately three weeks.
About mid-August in central Massachusetts, the larvae stop feeding, and thicken and compact a section of the web (this date may vary widely, depending on the hatch date of a particular cohort of larvae, elevation, and current climatic conditions). Upon entering this "pre-hibernation web" larvae become quiescent and molt to the fourth instar in about five days (from 3-7 days in the laboratory) (Edwards, 1875; Scudder, 1889; Bowers, pers. obs.). Edwards (1875) reported that larvae enter this web about 15 July in West Virginia, but at higher elevations in Massachusetts this web entrance can occur as late as the second week in September.
Previous authors (e.g. Scudder, 1889; Clark, 1927) have referred to this strengthened and compacted web as the "hibernating web." However, larvae do not spend the winter in this web at all, but leave it and move into the litter; this web is thus better referred to as the "pre-hibernation web/' which is the term used throughout this paper. This designation also serves to differentiate it from the feeding web which encloses plant parts on which the larvae are feeding.
The pre-hibernation web may be constructed at the base of the Turtlehead stalk on which the larvae have been feeding, but is sometimes found a short distance away. This web, as well as the feeding web constructed throughout the development of the pre-hibernation larvae, includes the stalk and leaves of Turtlehead as well as adjacent plants such as ferns, grasses and herbs (Figs. 1-3). The pre-hibernation webs are always constructed above ground level; most are found at a height of more than 50 cm. The mean height for 43 pre-hibernation
284
Journal of the Lepidopterists' Society
W * ■■ ^0 P
Volume 32, Number 4
285
webs measured in the summer of 1976 was 73 cm (range: 30-127 cm; standard deviation: ±25.50). I have never observed a pre-hibernation web directly on the ground.
Once the larvae enter the pre-hibernation web they do not feed at all. There may be some movement and occasionally a larva or two will leave and wander over the web surface. If the larvae are disturbed by movement of the web, they will move around a bit, but are usually inactive. An exception to this occurs when the web is damaged by storms or predators. In order to observe larval reactions to such damage, I broke open several webs. When the web is broken, the larvae begin moving around almost immediately and one or two larvae begin to crawl out of the web and over the surface. Other larvae soon follow. Within a day or two the web has been repaired and the larvae return to their inactive state. Thus, although the larvae are quiescent and not feeding, they can still react to stimuli such as disruption of the web.
The larvae do not spend the winter inside this web as has been suggested by previous authors (e.g. Edwards, 1875; Scudder, 1889; Clark, 1927). Rather, about the end of October, the larvae move out of the web en masse to the base of a plant on which the pre-hibernation web was constructed. Here they form a large contiguous aggregation among the dead grass and litter on the ground. This movement occurred during the weeks 27 October-4 November 1975, 22 October-29 October 1976, and 20 October-27 October 1977. While in this aggregation, individual larvae are active on warm days, moving from the primary mass as far as 25 cm. On cool, cloudy days or early in the morning, the larvae are usually found in a tight group, but individuals will quickly move around and out of the mass when disturbed. After approximately a week (from two days to two weeks; different for each aggregation of larvae), groups of about 10 to 100 individuals from this large mass move distances of 5-100 cm away and roll up leaves and bits of debris and fasten them with silk. It is here that they spend the winter. From 1974 to 1977 over 100 pre-hibernation webs were examined and in all webs the larvae moved into the litter. All the larvae remain near the pre-hibernation web and thus close to the Turtlehead stalk on which
<r
Figs. 1-3. Development of the feeding web of E. phaeton on C. glabra: 1, beginning web formation at top of plant; 2, further development of web with more leaves enclosed; 3, extension of web to encompass several stalks of C. glabra and adjacent plants.
286
Journal of the Lepidopterists' Society
they originally fed. Because Turtlehead is perennial this behavior ensures an adequate food supply for larvae emerging the following spring.
On warm, sunny days during the winter, larvae may become active, and groups may shift position by a few cm, rolling up new leaves in which to spend the remainder of the winter.
When the larvae are dug up from under the snow in January and brought to temperatures of 20-25°C, they quickly become active and crawl around the container in which they are confined. Usually they will not feed at this time. Clark (1927) brought hibernating larvae indoors in February and tried feeding them forced shoots of Lonicera japonica Thunb. (Caprif oliaceae), an alternate foodplant for this species (Clark, 1927; Scudder, 1889). After wandering around for a week, they began to feed, but all eventually died before reaching pupation. If larvae are brought indoors in this manner, allowed to be active and then returned to outside conditions, they resume their resting state with no apparent ill effects, emerging from diapause in the normal manner in the spring.
I attempted to break diapause in E. phaeton by subjecting several egg masses and their emerging larvae to 24 hours of light throughout development and hibernation. Although little is known about breaking diapause in univoltine insects, Beck (1968) suggests this method. These lab-reared larvae formed a pre-hibernation web in the same way as larvae in the field, yet were occasionally active. The larvae were given fresh food when plants seemed wilted or were consumed, and they fed occasionally through the end of October, with approximately 40% mortality. Between 25 October to 1 November there was a fourth molt (which does not occur in natural populations) that was synchronized within any single group of larvae but not among the groups. After this time, most of the larvae were inactive, and by mid-January all the larvae had died. Although attempts to break diapause in other insects with an obligatory diapause have been successful (e.g. Beck, 1968), this attempt was not. None of the larvae survived and none exceeded the size attained by hibernating larvae in the field. Thus diapause in this univoltine insect is truly obligatory and probably under genetic rather than photoperiodic control; the latter is probably the case in most multivoltine species (Beck, 1968; Hong and Piatt, 1975).
Discussion
Early authors (e.g. Scudder (1889) and Edwards (1884)) believed that the larvae of E. phaeton spent the winter in the pre-hibernation web (which they called the "hibernation web"). Observation of these
I
Volume 32, Number 4 287
webs throughout the fall in Massachusetts shows that it would not be feasible for larvae to remain in the web: although the pre-hibernation web has many more layers of silk and is smaller and more compact than the feeding web, wind and rain begin to damage it by the end of October. Before winter has advanced very far there is little left of the web, and the larvae cannot repair the web during the winter.
One of the harshest winter microclimates is just above the snow level where most of the pre-hibernation webs are found. Air temperature fluctuations here are much more dramatic than those at the soil surface (MacKinney, 1929; Mail, 1932) where larvae aggregate. Vegetation and litter reduce these temperature fluctuations, and snow provides further insulation (MacKinney, 1929; Mail, 1932). In Massachusetts there is snow cover for most of the winter and thus the larvae are quite well protected. Wind is also an important agent above the snow or ground level, but would have little impact on the larvae under snow or litter cover. Thus, regardless of snow cover, by moving out of the pre-hibernation web into the litter, larvae escape extreme temperature fluctuations and the desiccating effects of wind characteristic of winter conditions.
The question, then, is why should larvae construct a pre-hibernation web at all; why not move into the litter in August when feeding stops? Perhaps the groups of hibernating larvae would be easy prey for ground predators such as spiders and beetles which are abundant at the end of the summer. By the end of October, however, most of these predators are absent or inactive and the larvae could safely move into the litter.
In summary, pre-hibernation larvae of E. phaeton exhibit three major behavioral sequences: first, construction of the feeding web during the first three instars; second, abandonment of this web and construction of a small, compact pre-hibernation web in which the larvae remain quiescent; third, departure from this web and movement of smaller groups into the litter where they form an over-wintering site by rolling up leaves and bits of debris and fastening them with silk. The last two sequences require appreciable expenditures of energy (for movement and silk-making) while no food is being eaten. This expenditure suggests that these behaviors are necessary to ensure larval survival over the fall and winter.
Acknowledgments
I am very grateful to Dr. Theodore D. Sargent for his encouragement and editorial comments. This research was supported in part by a Sigma Xi Grant-in-Aid of Research.
288
Journal of the Lepidopterists' Society
Literature Cited
Beck, S. D. 1968. Insect photoperiodism. Academic Press, New York. 288 pp. Clark, A. H. 1927. Notes on the melitaeid butterfly Euphydryas phaeton (Drury)
with descriptions of a new subspecies and a new variety. Proc. U.S. Nat. Mus.
Wash. 71, article 11, #2683. pp. 1-22. Clark, S. H. & A. P. Platt. 1969. Influence of photoperiod on development and
larval diapause in the Viceroy butterfly, Limenitis archippus. J. Insect Physiol.
15: 1951-1957. Edwards, W. H. 1875. Notes on butterflies. Canad. Entomol. 7: 150-151. ----------. 1884. Description of the preparatory stages of Melitaea chalcedon, Bois.,
with some notes on larvae of M. phaeton. Papilio 4: 63-70. Hong, J. W. & A. P. Platt. 1975. Critical photoperiod and day-length threshold
differences between northern and southern populations of the butterfly, Limenitis
archippus. J. Insect Physiol. 21: 1159-1165. MacKinney, A. L. 1929. Effects of forest litter on soil temperature and soil freezing in autumn and winter. Ecology 10: 312-321. Mail, G. A. 1932. Winter temperature gradients as a factor in insect survival. J.
Econ. Entomol. 25: 1049-1053. Scudder, S. 1889. The Butterflies of the Eastern United States and Canada. W.
H. Wheeler, Cambridge, Mass. Schweitzer, D. F. 1977. Life history strategies of the Lithophanini (Lepidoptera:
Noctuidae, Cuculliinae), the winter moths. Ph.D. dissertation. Univ. of Mass.,
Amherst, Mass. 304 pp.