4

Sigh as ‘nee “ay ; ee

tHE OB So TPs Pag teas i

S <b. By oe & By tosh Le aS be wes KES RS ‘ ORE .

< .

SoG ae Sa AR toh aor 8s) Be

tii

Mi he

ye!

BA is ee \

Pat a

aS

a

Volume 42 1988 Number 1

ISSN 0024-0966

JOURNAL

of the

LEPIDOPTERISTS’ SOCIETY

Published quarterly by THE LEPIDOPTERISTS’ SOCIETY

Publié par LA SOCIETE DES LEPIDOPTERISTES Herausgegeben von DER GESELLSCHAFT DER LEPIDOPTEROLOGEN Publicado por LA SOCIEDAD DE LOS LEPIDOPTERISTAS

16 March 1988

THE LEPIDOPTERISTS’ SOCIETY

EXECUTIVE COUNCIL

JERRY A. POWELL, President JEAN-FRANCOIS LANDRY, Vice DoucLas C. FERGUSON, Immediate Past President

President ATUHIRO SIBATANI, Vice JACQUELINE Y. MILLER, Vice President President RICHARD A. ARNOLD, Secretary JAMES P. TUTTLE, Treasurer

Members at large:

MIRNA M. CASAGRANDE M. DEANE BOWERS JULIAN P. DONAHUE EDWARD C. KNUDSON RICHARD L. BROWN JOHN E. RAWLINS FREDERICK W. STEHR PAUL A. OPLER Jo BREWER

The object of the Lepidopterists’ Society, which was formed in May 1947 and for- mally constituted in December 1950, is “to promote the science of lepidopterology in all its branches, .... to issue a periodical and other publications on Lepidoptera, to fa- cilitate the exchange of specimens and ideas by both the professional worker and the amateur in the field; to secure cooperation in all measures” directed towards these aims.

Membership in the Society is open to all persons interested in the study of Lepi- doptera. All members receive the Journal and the News of the Lepidopterists’ Society. Institutions may subscribe to the Journal but may not become members. Prospective members should send to the Treasurer full dues for the current year, together with their full name, address, and special lepidopterological interests. In alternate years a list of members of the Society is issued, with addresses and special interests. There are four numbers in each volume of the Journal, scheduled for February, May, August and November, and six numbers of the News each year.

Active members—annual dues $25.00 Student members—annual dues $15.00 Sustaining members—annual dues $35.00 Life members—single sum $350.00 Institutional subscriptions—annual $40.00

Send remittances, payable to The Lepidopterists’ Society, to: James P. Tuttle, Treasurer, 3838 Fernleigh Ave., Troy, Michigan 48083-5715, U.S.A.; and address changes to: Julian P. Donahue, Natural History Museum, 900 Exposition Blvd., Los Angeles, California 90007-4057 U.S.A. For information about the Society, contact: Richard A. Arnold, Sec- retary, 50 Cleaveland Rd., #8, Pleasant Hill, California 94523-3765, U.S.A.

To obtain:

Back issues of the Journal and News (write for availability and prices); The Com- memorative Volume ($10.00; $6.00 to members, postpaid); A Catalogue/Checklist of the Butterflies of America North of Mexico (clothbound $17.00, $10.00 to members; paperbound $8.50, $5.00 to members): order (make remittance payable to “The Lepi- dopterists’ Society’) from the Publications Coordinator, Ronald Leuschner, 1900 John St., Manhattan Beach, California 90266-2608, U.S.A.

Journal of the Lepidopterists’ Society (ISSN 0024-0966) is published quarterly for $40.00 (institutional subscription) and $25.00 (active member rate) by the Lepidopterists’ Society, % Los Angeles County Museum of Natural History, 900 Exposition Blvd., Los Angeles, California 90007-4057. Second-class postage paid at Los Angeles, California and additional mailing offices. POSTMASTER: Send address changes to the Lepidopterists’ Society, % Natural History Museum, 900 Exposition Blvd., Los Angeles, California 90007- 4057.

Cover illustration: Computer-drawn figure of Dakota skipper, Hesperia dacotae (Skin. ), nectaring on narrow-leaved purple coneflower, Echinacea angustifolia DC. Drawing originally executed freehand in the Macintosh application Fullpaint and embellished with FatBits, exported to SuperPaint and further embellished with LaserBits, then produced on a LaserWriter. Cover version is 50% width of LaserWriter version. Submitted by Ronald A. Royer, Division of Science, Minot State University, Minot, North Dakota 58701.

JouRNAL OF Tue LeErPiIpoprTreERIstTs’ SOCIETY

Volume 42 1988 Number 1

Journal of the Lepidopterists’ Society 42(1), 1988, 1-13

SPEYERIA ATLANTIS IN COLORADO: REARING STUDIES CONCERNING THE RELATION BETWEEN SILVERED AND UNSILVERED FORMS

JAMES A. SCOTT 60 Estes Street, Lakewood, Colorado 80226

ABSTRACT. Speyeria atlantis in the SE Rocky Mts. occurs in two forms, silvered and unsilvered, that could be mere forms or separate species. Nine wild females laid eggs and produced adults in the laboratory. Offspring resembled mothers in most cases, except for two mothers about half silvered and one mother about one-third silvered that produced nearly unsilvered offspring. The two forms have the same courtship, without obvious courtship barriers between them, and male pheromones smell the same. Silvered and unsilvered adults have differently colored larvae. The two forms can differ in habitat, and adults actively select different habitats. The two are probably forms of the same species.

Additional key words: Nymphalidae, habitat selection, polymorphism, courtship.

The relation between silvered and unsilvered forms of Speyeria at- lantis (Edw.) has puzzled many people (Scott 1986b). Thus Grey et al. (1963) discussed the two forms in the Black Hills of South Dakota, where the silvered form with chocolate ventral hindwing (a. atlantis) predominates in wet meadow areas, and the unsilvered form with reddish-brown ventral hindwing (a. hesperis Edw. = a. lurana dosP. & G.) prevails in drier areas. From a locality with 44% silvered adults, W. Evans (in Grey et al. 1963:146) reared 8 silvered offspring with chocolate ventral hindwing from silvered mothers with chocolate ven- tral hindwing, and 26 unsilvered plus at least 1 silvered offspring with a reddish-brown ventral hindwing from unsilvered mothers with a reddish-brown ventral hindwing. The exact number of mothers con- tributing was not known, but was probably one or two for each form. Evans noted that the double dorsal stripes were light brown on atlantis larvae, grayish white on hesperis larvae, and that hesperis pupae have more light-brown shading on the wing case than do atlantis. Grey et al. (1968) suggested that the two could be treated as separate species,

2 JOURNAL OF THE LEPIDOPTERISTS SOCIETY

though they retained them in one species because they seem to inter- grade in other western U.S. regions.

A similar situation occurs in SW Manitoba where a dark variety of silvered a. atlantis (a. hollandi F. & R. Cherm.) with chocolate ventral hindwing flies in mountains and forest, whereas a very pale silvered a. dennisi dosP. & G. usually with light-brown ventral hindwing flies on tallgrass prairie. They occur near each other. At Duck Mountain, ad- jacent populations show no intergradation (J. Troubridge pers. comm.). In this area, they behave as separate species, although westward they intergrade at Meadow Lake Park, Saskatchewan (Hooper 1973).

In Colorado E of the continental divide, the unsilvered form (a. hesperis) prevails in the mountain foothills, and as one goes higher in the mountains the silvered form (a. atlantis, = a. electa Edw.) gradually increases in frequency until it predominates in the Canadian Zone. Silvered forms in Colorado’s Front Range usually have a chocolate- brown ventral hindwing, whereas unsilvered forms usually have a red- dish-brown ventral hindwing, although this association sometimes breaks down; thus some silvered adults have a red-brown ventral hindwing, and some unsilvered ventral hindwing adults have a darker reddish- brown ventral hindwing. Females have a slightly darker ventral hind- wing than males; a bilateral gynandromorph from Critchell, for in- stance, has a very red-brown ventral hindwing on the male side, a darker red-brown ventral hindwing on the female side.

The silvered or unsilvered color is due to light reflection from in- dividual scales. Silver scales appear transparent through a microscope, but their surfaces reflect a white sheen (evidently due to structural interference of light) which causes the silver appearance. Unsilvered scales are cream in color because they appear to be filled with cream pigment, and their surfaces do not reflect light; their scale structure could be the same as silvered scales if the internal pigment blocks transmission of light through the scale to prevent light interference. So the difference between silvered and unsiivered scales could result solely from absence or presence of internal cream pigment. A given wing spot can be entirely cream (unsilvered), or it can be cream with a few silver scales, or the entire spot can be covered with silver scales. Po- tentially silvered spots occur in four series on the ventral hindwing: basal, postbasal, postmedian, and submarginal. In the basal series, the dot in the discal cell is more likely to have silver scales than the other spots. The postbasal series of spots is less likely to be silvered than the other series, and the basal and marginal series are most likely to be silvered in the mostly unsilvered forms.

To determine the relation between the forms in Colorado, I reared the eggs oi selected females, especially those females of a form rare in

VOLUME 42, NUMBER 1 3

their population because these would have the greatest likelihood of mating with a male of the opposite form.

REARING METHODS

To obtain eggs, females were collected from Colorado Front Range localities, brought to the laboratory, and placed in jars with Viola nephrophylla Green leaves and fed honey-water once per day. Most females lived about a week and laid several dozen eggs. Eggs hatch readily, but first-stage larvae diapause in nature, so to prevent diapause they were placed under constant light in tiny vials with a slice of green violet leaf. After a few days or weeks some larvae ended diapause and started to feed; these fed steadily until pupation on V. nephrophylla leaves. Three months were required to raise offspring of one female. Voucher specimens including larvae, pupal shells, and reared silvered and unsilvered adults are in the National Museum of Natural History, Washington, D.C.

RESULTS Silvering of Mothers and Offspring

A total of 104 adult offspring were reared from 9 mothers from 6 Colorado sites. Each site is described below.

Tinytown (2120 m), Jefferson Co., isa Transition Zone foothills valley bottom with ponderosa pine, douglasfir, willow, alder, honeysuckle, etc., along the creek; the hostplants Viola canadensis L. and V. adunca Smith (Scott 1986a) are common on the shaded gulch bottom and the base of the N-facing slope. Here 92% of adults had a reddish-brown ventral hindwing with mostly unsilvered spots, 6% were partly silvered (N = 6 half silvered, N = 1 mostly silvered), and 2% were fully silvered with a chocolate-brown ventral hindwing (N = 117). If the fully silvered mother mated at random, the father was probably unsilvered; yet all offspring were silvered (Table 1).

Corwina Park (2120 m), Jefferson Co., is a Transition Zone foothills wooded gulch draining N; the hostplants V. adunca and probably V. canadensis are in gulch-bottom shade and E-facing shaded slopes. Here 91% of adults were unsilvered with a red-brown ventral hindwing, 9% silvered with a chocolate-brown ventral hindwing (N = 21). If the completely silvered mother mated at random, the father was probably unsilvered; yet all offspring were fully silvered (Table 1).

O’Fallon Park (2100 m), Jefferson Co., is near Corwina Park, and is also a Transition Zone foothills wooded gulch draining N with the hostplants V. adunca and V. canadensis in gulch-bottom shade and E-facing shaded slopes. Here 83% were unsilvered with a red-brown ventral hindwing, 18% silvered with a chocolate-brown ventral hind-

TABLE 1. Extent of silvering on ventral hindwing spots, and color of basal two-thirds of ventral hindwing, of mothers and offspring. Numbers are proportions: for example, “1” under “base” means all scales on wing base spots are silvered, “1/5” under “post- median” means 20% of scales of postmedian spots are silvered, “0” under “‘submarginal”’ means no scales of submarginal spots are silvered, etc., “gyn’’ is bilateral gynandromorph, “fis female, and “m”’ is male.

Material Sex Ventral hindwing Base Postbasal Postmed. Submarg.

Tinytown, Jefferson Co., mother caught 20 July 1984

Mother liek chocolate 1 1 1 1 Offspring 27 m chocolate I 1 il il Offspring 19 f chocolate 1 1 1 1 Corwina Park, Jefferson Co., mother caught 13 July 1985 Mother et dark choc-brown 1 1 1 i. Offspring 2m dark red-brown i 1 1 i Offspring Pt choc-brown 1 1 I 1 Offspring lf dark choc-brown 1 Il 1 1 O'Fallon Park, Jefferson Co., mother caught 12 August 1985 Mother Ib it red-brown 2/38 Wis 12) 2 Offspring lm very red-brown 0 0 1/4 1/3 Offspring Gh very red-brown 0 0 0 1/5 Offspring Ike very red-brown 2/3 0 0 1/2 Critchell, Jefferson Co., mother caught 3 August 1985 Mother lf red-brown We 1/10 V2, 2 Offspring 8 f very red-brown 0 0 0 0 Offspring 6 f very red-brown 0 0 0 0 Offspring l gyn very red-brown 0 0 0 0 Mt. Judge female B, Clear Creek Co., mother caught 8 August 1985 Mother i at red-brown 28 1/8 is 1/8 Offspring lm very red-brown 0 0 0 1/6 Offspring 6m very red-brown 0 0 0 1/10 Offspring lm very red-brown 0 0 0 Vs Offspring lm very red-brown 1/10 0 0 1/5 Offspring IL sem very red-brown 1/5 0 0 1/10 Offspring 2a very red-brown 0 0 0 1/10 Offspring Si very red-brown 0 0 0 0 Cherry Gulch, Jefferson Co., mother caught 17 July 1984 Mother ef red-brown 2/3 1/5 0 1/3 Offspring lef very red-brown 1/4 0 0 1/3 Mt. Judge female D, Clear Creek Co., mother caught 8 August 1985 Mother Lak dark red-brown yd 0 0 1/2 Offspring 2m very red-brown 0 0 0 0 Mt. Judge female F, Clear Creek Co., mother caught 8 August 1985 Mother at red-brown LS 0 0 1/10 Offspring 1m very red-brown 0 0 0 1/10 Offspring 1m very red-brown 1/10 0 0 1/5 Mt. Judge female A, Clear Creek Co., mother caught 8 August 1985 Mother hag red-brown 0) 0) 0 IAs Offspring 9m very red-brown 0 0 0 0 Offspring 3f dark red-brown ) 0 0 1/10 Offspring lg dark red-brown 0 0 0 WG Offspring ie) red-brown 0 0 0 0 Offspring I red-brown 0 0 0 1/6 Offspring Zit very red-brown 0 0 0 0

VOLUME 42, NUMBER 1 5

wing, and 4% intermediate (N = 19). If the nearly half-silvered mother mated at random, the father was probably unsilvered; all offspring were nearly unsilvered (Table 1).

Critchell (2370 m), Jefferson Co., is a shaded E—W streamside in the upper Transition Zone foothills, with ponderosa pine, douglasfir, various shrubs, grassy glades, and V. canadensis and V. adunca. Here 88% were unsilvered with a reddish brown ventral hindwing, 7% fully sil- vered, and 5% intermediate (N = 2 half silvered, N = 1 mostly silvered) (N = 53). If the nearly half-silvered mother mated at random, the father probably was unsilvered; all offspring were completely unsilvered (Ta- ble 1).

Cherry Gulch (2100 m), Jefferson Co., is a Transition Zone foothills gulch at the base of a N-facing slope covered with douglasfir, Holo- discus, Physocarpus, other shrubs, and Viola canadensis. Here 97% were unsilvered with a reddish brown ventral hindwing, 3% silvered with a brown ventral hindwing (N = 69). If the mostly unsilvered mother mated at random, the father was probably unsilvered; the single off- spring was less silvered than the mother (Table 1).

Mt. Judge (2 km NE, 2770 m), Clear Creek Co., is a Canadian Zone valley bottom, with forest (spruce, pine, douglasfir, some aspen) beside grassy meadows, a tiny creek on the valley bottom, and V. canadensis and V. nephrophylla. Silvered adults with a chocolate ventral hindwing were most common, with a few silvered adults with a reddish brown ventral hindwing; but unsilvered adults with a red-brown ventral hindwing were also found, a few unsilvered adults with a brown ventral hindwing, and a few variably silvered intermediates. The upperside black lines vary from narrow to wide independent of ventral hindwing variation. Shape of silver spots varies between individuals, as does amount of black at the base of each silver spot, but this variation is also inde- pendent of degree of silvering. Four females from this site labeled A, B, D, and F, produced offspring (Table 1). If the Mt. Judge mothers mated at random, they probably mated with silvered males because 74% of males here were silvered (Table 2). However, because of habitat selection at this site (described in next section), and because all four mothers were found in mixed woods away from the creek where only 38% of males were silvered (Table 2), the mothers probably mated with unsilvered fathers. Mother B was about one-third silvered; her offspring were almost completely unsilvered. Mothers A, D, and F, and their offspring, were almost completely unsilvered.

Habitat Selection and Movements

The Mt. Judge site displayed habitat selection by the forms (Table 2). In several meadows along the tiny creek 90% of adults were silvered,

6 JOURNAL OF THE LEPIDOPTERISTS SOCIETY

TABLE 2. Frequency of color forms at Mt. Judge site, based on nine visits 1984 to 1987.

Mixed woods away from creek Meadows along creek

Silvering No. male No. female No. male No. female Mostly unsilvered 24 13 let 2 Half silvered 1 0 1 0 Completely silvered 15 8 2 53

10% unsilvered. In contrast, at the habitat edge near the head of the valley, away from the creek in mixed woods—tiny meadows edging the large meadow and in the adjacent meadow-edge, one-third (38%) of adults were silvered, and two-thirds (62%) unsilvered.

A small mark-recapture study was conducted at Mt. Judge in 1987 (31 July, 5, 9 Aug.), in which 33 adults were marked and 16 recaptured. Six unsilvered adults were marked (2 male, 4 female), and 3 females recaptured, all in the mixed woods, one after 5 days. Twenty-seven silvered adults were marked (14 male, 13 female), and 13 recaptured (7 male, 6 female), after up to 9 days, including 5 moves completely across the habitat, and 6 halfway across it. I conclude that silvered adults move completely about the habitat, and females probably oviposit in the mixed woods where host violets grow under conifers. But judging from the restricted distribution of unsilvered adults (Table 2), these are more local, and their restricted movement causes the habitat selection difference. In general, unsilvered Colorado adults prefer open woods with violets (N-facing slopes and gulch bottoms in the foothills), whereas silvered adults also occupy more open wet valley bottoms.

Pheromones

Male odor of both forms from Mt. Judge was compared by the author. Males of silvered and unsilvered forms smelled the same: the odor is sweet but has a “hot” or “peppery” sensation, sweet but slightly peppery pungent. Virtually every male had this odor, a few weaker than others. Females lacked an odor. The description of odor is subjective, and different observers might use different words to describe it, but it was the same for both forms. Thus, the male pheromone is probably the same in both forms, although the human nose certainly cannot match the precision of laboratory instruments.

The pheromone system is complex. Males have androconial scales on dorsal wing veins (Scott 1986b:fig. 37) which evidently produce the pheromone odor; pheromone from these scales in the closely related European Argynnis paphia L. causes the female to land and accept the male (Magnus 1958). Females have a dorsal gland between abdomen segments 7 and 8 (Scott 1986b:fig. 37). This gland in A. paphia produces

VOLUME 42, NUMBER 1 a

a pheromone that attracts males: femalelike dummies attract males but do not elicit complete courtship, and freshly killed females are more attractive to males than dried females (Magnus i958); virgins respond to nearby males by exposing the abdomen gland and aiming the ab- domen tip toward the male (Treusch 1967). Males have a paired gland on the abdomen tip (Arnold & Fischer 1977, Scott 1986b) which, by comparison with Heliconiini (Scott 1986b), could possibly transfer pher- omone to the female during mating to enable mated females to produce a third pheromone that repels males.

Courtship

Courtship of Speyeria atlantis, which is nearly identical to that of Argynnis paphia (Magnus 1950), was described by Scott (1986b) based mainly on unsilvered form courtships in Jefferson Co., Colorado. In addition, a completed courtship between silvered male and female forms was seen at Mt. Judge: female on flower when male sighted her and landed; she fluttered her mostly spread wings with small amplitude for 1 s, he flicked his nearly closed wings behind her for 1-2 s; she rotated around flower top | revolution with her wings still spread while he rotated after her and flicked his nearly closed wings once during turn; she stopped, closed her wings, tilted forward so that her abdomen was raised slightly but lowered from between hindwings; he spread his wings partway; they joined.

Four courtships were seen at Mt. Judge between unsilvered males and silvered females, as follows.

1) Male patrolled near female (prior mating status unknown) on flower, landed, flicked wings (wingtips vibrating 0 to 1 cm apart about twice per s) for 10 s, curved abdomen laterally to attempt joining (meanwhile female, wings closed, leaned forward with ab- domen lowered from between closed hindwings and abdomen raised above horizontal about 60°); wind blew them and he flew, fluttered over her for 1 s, landed, flicked beside her 10 s, curved his abdomen but was too close and his abdomen tip missed (during his bending she kept abdomen exposed), then he flew away. Female was evidently receptive because she exposed her abdomen and did not perform rejection dance (fluttering wings vigorously).

2) Male patrolled near silvered virgin (later found to have no spermatophores) on flower, landed, flicked his wings, she crawled away with closed wings, he crawled after her for 5 min while flicking and bending his abdomen, she stopped and spread wings partly while he flicked and curved abdomen to attempt mating for 5 min, he flew away (evidently she did not extrude genitalia, so he could not join). She was unreceptive even though she did not flutter her wings, perhaps because, as judged from weak flight, she was too young.

3) He pursued her in flight, they landed, she fluttered slightly and crawled away while he flicked his wings and crawled behind, she got farther away, he flew up a short distance but did not find her and flew away.

4) She raised her wings and slightly lowered and partly extruded her abdomen while he flicked his nearly closed wings behind her, he flew away after about 30 s.

8 JOURNAL OF THE LEPIDOPTERISTS SOCIETY

Data on courtship between forms are too few to be conclusive, but no obvious courtship barriers to mating occur. Releases of reared virgins are needed. Grey et al. (1963) inconclusively report abnormal courtship of a few laboratory adults.

Larval Differences (Figs. 1-14)

Color photographic slides were made of larvae and pupae from each study site except larvae from Corwina, and some larvae and pupae were preserved to correlate their color pattern with adult appearance.

From a distance, older larvae producing silvered adults (Figs. 5, 7-9) appear mottled black with orangish tan spines and two middorsal white lines, whereas older larvae producing unsilvered adults (Figs. 10-14) appear solid black with orange spines. Viewed more closely, larvae of both forms are basically black, with a pair of middorsal whitish lines 1 mm apart, and three rows of scoli (lateral to middorsal lines, supraspiracular, and subspiracular) which are tan or orange with black tips. The head of both forms is black with the dorsal half of the rear half of the head orangish.

Larvae of the silvered form (based on larvae from Tinytown, Figs. 5, 7-9) have the middorsal whitish lines conspicuous and mostly continuous, though alternately wider and narrower. Because Corwina pupae had less conspicuous lines than Tinytown pupae, Corwina larvae may not have had the lines this conspicuous. Scoli of the silvered form are orangish tan with black tips. Ground color is not as black as in the unsilvered form so three rows of black bands with very sinuous narrowly white edges are recognizable: along the dorsalmost scoli (edging middorsal white lines), along the supraspiracular scoli, and in between these (Figs. 8, 9). A light gray-brown transverse band circles the rear of each segment except middorsally, a remnant of the pale transverse stripes of Speyeria nokomis (Edw.) larvae (Scott & Mattoon 1981).

Larvae of the unsilvered form (from O'Fallon, Critchell, Cherry Gulch, Mt. Judge) are a little darker black and the pattern is obscured, so the black sinuous bands are unrec- ognizable without a microscope, and the middorsal two lines are fainter and broken into two dashed lines (Figs. 10-14). Scoli are orange with black tips. The only variation between localities among unsilvered larvae involves the single Cherry Creek larva which had slightly less orangish scoli. Edwards’ (1888b) description of the unsilvered form is very similar.

The above descriptions of larvae do not correspond with descriptions of larvae of the silvered and unsilvered forms in the Black Hills of South Dakota (Grey et al. 1963). Both are described as identically black with orange spine shafts, the two middorsal lines grayish white in the unsilvered form, light brown in the silvered form. Thus the two middorsal lines are described as whiter in the unsilvered form in South Dakota, whereas they are whiter in the silvered form in Colorado. My descriptions are based on 104 larvae and dozens of color slides from many sites, whereas the South Dakota data are fewer.

Width of the two pale middorsal lines of the larva is apparently not closely linked to degree of silvering of the adult; among larvae producing silvered adults, the whiteness differed somewhat between the Tinytown and Corwina sites in Colorado as noted above, and differed between Colorado and South Dakota adults.

Thus, both larvae and adults of the unsilvered form have more pig- ment—more cream in adult scales, more orange on larval spines, more black on larval body—so one can guess that the gene responsible for the unsilvered form causes an increased deposition of some dark pig- ment such as melanin.

Larvae and pupae of silvered ventral-hindwing S. atlantis from NE

VOLUME 42, NUMBER 1 9

Fics. 1-20. 1, First-stage larva, silvered form, Tinytown; 2, Second-stage larva, sil- vered form, Tinytown; 3, Third-stage larva, silvered form, Tinytown; 4, Fourth-stage larva, silvered form, Tinytown; 5, Fourth-stage larva, silvered form, Tinytown; 6, Third- stage larva, silvered form, Tinytown; 7, Mature larva, silvered form, Tinytown; 8, Mature larva, silvered form, Tinytown; 9, Mature larva, silvered form, Tinytown; 10, Third- stage larva, unsilvered form, O'Fallon female C; 11, Mature larva, unsilvered form, O'Fallon female C; 12, Mature larva, unsilvered form, Cherry Gulch; 13, Mature larva, unsilvered form, Mt. Judge female F; 14, Mature larva, unsilvered form, Mt. Judge female A; 15, Pupa (orange-brown wings), silvered form, Tinytown; 16, Pupa (orange- brown wings), silvered form, Tinytown; 17, Pupa (orange-brown wings), silvered form, Tinytown; 18, Pupa (orange-brown wings), silvered form, Tinytown; 19, Pupa (partly orange-brown wings), unsilvered form, O'Fallon female C; 20, Pupa (black wings), un- silvered form, Mt. Judge female F.

10 JOURNAL OF THE LEPIDOPTERISTS SOCIETY

U.S. (Edwards 1888a) are grayer than Colorado-South Dakota S. at- lantis; larvae and pupae evidently show geographic variation as do adults.

Pupal Differences (Figs. 15-20)

Pupae from localities with sufficient numbers show great individual variation, but there is no obvious important difference between silvered and unsilvered forms. The pupa resembles S. nokomis (Scott & Mattoon 1981) in general, but is darker (orange-brown), and the posterior half of each abdominal segment is darker because it is mottled with tiny black dots and dashes. The anterior half of each abdominal segment is not uniformly black as in nokomis: some pupae have a broad black irregular band, but most have the black areas broken into spots, in- cluding triangular spots just beside the anterior-pointing orange-brown middorsal triangles on segments 5-7. Pupae from Tinytown have two sinuous tan middorsal abdominal lines, but pupae from Corwina (both sites produced silvered adults) and the other sites had weak tan mid- dorsal lines. Pupal wing color varies from mostly black to almost wholly orange-brown, but most are mostly orange-brown, a few black-winged.

Grey et al. (1963) describe the pupal wing cases of silvered forms as darker with less light brown mottling than those of unsilvered forms in the Black Hills. However, they reared only three silvered adults, so the difference is probably due to small sample size because all Colorado sites with large samples show considerable variation in pupal wing color. Pupae producing silvered adults are not darker in Colorado.

DISCUSSION

There are several reasons why S. a. atlantis and S. a. hesperis could be treated as distinct species: they often fly together, they prefer dif- ferent microhabitats, amount of silvering seems usually linked with ventral hindwing color, mothers usually produce offspring resembling themselves, and their larvae differ. If scientists were aware only of Black Hills populations, the two would certainly be treated as separate species because they are so distinct there. Some anecdotes (coinci- dences?) also fit the two-species theory. For instance, six unsilvered males and one si!vered pair were found in the Mt. Judge mixed woods 28 July 1987, the silvered pair in copula.

There are several reasons why S. a. atlantis and S. a. hesperis could be treated as one species:

|) Silvered and unsilvered forms are linked by a complete series of intermediate adults, from slightly to partly to half to mostly silvered, although only slightly silvered inter- mediates are common.

VOLUME 42, NUMBER 1 iat

2) Unsilvered mothers sometimes produce silvered offspring (Grey et al. 1963:146), and half-silvered mothers often produce unsilvered offspring (Table 1).

3) In many populations, silvered forms are rare (<5%) as in the lower foothills of the Colorado Front Range, rarity a true species might have difficulty surviving. The reverse is also true, in which unsilvered forms are rare within silvered populations, as in the wet center of the Black Hiils (Grey et al. 1963). However, S. coronis (Behr) is just as rare and it survives.

4) Frequencies of the forms show clinal trends, both altitudinally in the Colorado Front Range, and along habitat gradients. For instance, in the Black Hills (Grey et al. 1963), atlantis is common in wet meadow habitats on poorly drained granite, and is rarer away from these areas. Similarly, in S Colorado (Scott & Scott 1980) hesperis predominates in the lower foothills, both forms occur in dry areas at higher altitude, and atlantis pre- dominates in three wet meadow enclave habitats at middle altitudes: Coaldale in Arkansas Canyon, Fremont Co.; SW of Westcliffe on Wet Mountain Valley floor, Custer Co.; Stonewall in upper Purgatoire River valley, Las Animas Co. Such enclaves have not been found in the Front Range W of Denver, where silvered forms are rare in the foothills and increase in frequency with altitude until they predominate in the upper Canadian Zone.

5) When attempts are made to divide S. atlantis into silvered and unsilvered “species ’’, their distributions are incongruous because unsilvered forms cut an E-W swath through the range of silvered forms, replacing them in the process (Scott 1986b).

The silvered-unsilvered division also fails to solve the problem of sympatry of S. a. dennisi and S. a. atlantis (hollandi), both of which are silvered, in Manitoba. A species S. dennisi could include S. atlantis ratonensis Scott from NE New Mexico and S. a. greyi from NE Nevada, but dennisi is said to intergrade W to atlantis in Saskatchewan-Alberta, and greyi intergrades with dodgei in S Idaho (P. C. Hammond pers. comm.), and at least greyi seems independently evolved toward similar pallidity.

6) Other S. atlantis subspecies have polymorphisms of silvered-unsilvered adults: wa- satchia dosP. & G. (=tetonia dosP. & G.) in W Wyoming-Utah is usually unsilvered, chitone (Edw.) in S Utah and schellbachi Garth in N Arizona are usually silvered.

7) Other species of Speyeria have silvered-unsilvered polymorphisms: zerene (Bdv.) in California and S Oregon, callippe (Bdv.) in N California and the Sierra Nevada, egleis (Behr) in the Sierra and Utah, hydaspe (Bdv.) in British Columbia. These polymorphisms are accepted by lepidopterists. Boggs (1987) hypothesized that rare unsilvered S. mor- monia are homozygous recessives that fail to reproduce, which is dubious because S. mormonia artonis (Edw.) are nearly always unsilvered.

8) Association between ventral hindwing color and silvering and larval color pattern breaks down geographically. In the Black Hills and E of the continental divide in the Colorado mountains, silvered adults have a chocolate-brown ventral hindwing (darker in the Black Hills), and unsilvered adults usually have a reddish brown ventral hindwing. However, in N-central New Mexico, 98% of adults (N = 60) are silvered but the ventral hindwing varies from chocolate- to reddish brown. In SW Manitoba S. atlantis dennisi and S. a. atlantis (a. hollandi) are 100% silvered but the ventral hindwing is usually light brown in the former and chocolate-brown in the latter. And silvered adults have the ventral hindwing browner in the Black Hills than in the Colorado Front Range. Larval differences in Colorado are partially reversed in the Black Hills, and larvae are grayer in E North America.

The conclusion that silvered and unsilvered adults are polymorphic forms of one species seems preferable.

Paleogeography

The current geographic distribution of wing characters suggests that the dark silvered form (S. a. atlantis) occupied the coniferous forest in N U.S. and the Rocky Mountain foothills during the Ice Age; afterwards

12 JOURNAL OF THE LEPIDOPTERISTS SOCIETY

it moved higher in altitude and latitude. The unsilvered form with reddish brown ventral hindwing (now S. a. hesperis, a. wasatchia, a. irene [Bdv.]) occupied open forest in the southern Great Basin lowlands; after the Ice Age it spread N into the mountains, E through lowland S Wyoming to the Black Hills, and S along the Colorado mountain foot- hills. The silvered form with narrower black lines and a pale brown ventral hindwing (now S. atlantis dennisi and a. ratonensis) occupied aspen parkland in the current S Great Plains or central Texas; after the Ice Age it spread, respectively, N to Canada, and upward to a mountain mesa. The forms became sympatric after the Ice Age.

Mechanism of Inheritance

The inheritance mechanism of silvering is unknown. Rarity of half- silvered adults suggests dominance, but some broods with half the offspring silvered and half unsilvered should occur but did not. If half- silvered adults were heterozygotes, they would not produce all-unsil- vered broods as at Critchell. If silvered is dominant, rare silvered moth- ers would be likely to produce silvered offspring, as at Tinytown and Corwina. O’Fallon and Cherry Gulch broods perhaps suggest modifier genes that cause part-silvering.

Maternal inheritance seems the best guess now, and fits all the reared broods; offspring would resemble the mother, the father having no effect or perhaps merely modifying partly silvered offspring. Sterling O. Mattoon (pers. comm.) states that Speyeria offspring generally re- semble their mother very closely, although some silvered offspring have been reared from unsilvered mothers and vice versa.

ACKNOWLEDGMENTS

I thank L. P. Grey, T. C. Emmel, and C. L. Boggs for reviewing the manuscript.

LITERATURE CITED

ARNOLD, R. A. & R. L. FISCHER. 1977. Operational mechanisms of copulation and oviposition in Speyeria (Lepid.: Nymphalidae). Ann. Entomol. Soc. Amer. 70:455- 468. See Fig. 6.

Boccs, C. L. 1987. Demography of the unsilvered morph of Speyeria mormonia in Colorado. J. Lepid. Soc. 41:94-97.

Epwarpbs, W. H. 1888a. Description of the preparatory stages of Argynnis atlantis, Edw. Can. Entomol. 20:1-3.

1888b. Description of the preparatory stages of Argynnis hesperis, Edw. Can. Entomol. 20:67-69.

GREY, L. P., A. H. MoEcK & W. H. Evans. 1963. Notes on overlapping subspecies. II. Segregation in the Speyeria atlantis of the Black Hills (Nymphalidae). J. Lepid. Soc. 17:129-147.

Hooper, R. 1973. Butterflies of Saskatchewan. Saskatchewan Mus. Nat. Hist., Regina, Saskatchewan. 216 pp.

Macnus, D. 1950. Beobachtungen zur Balz und Eiablage des Kaisermantels Argynnis paphia L. (Lep., Nymphalidae). Z. Tierpsychol. 7:435-449.

VOLUME 42, NUMBER 1 13

1958. Experimentelle Untersuchungen zur Bionomie und Ethologie des Kai- sermantels Argynnis paphia L. (Lep. Nymph.) I. Ueber optische Ausloeser von An- fliegereaktionen und ihre Bedeutung fuer das Sichfinden der Geschlechter. Z. Tier- psychol. 15:397—426.

ScoTT, J. A. 1986a. Larval hostplant records for butterflies and skippers (mainly from western U.S.), with notes on their natural history. Papilio (New Series) #4. 37 pp.

1986b. The butterflies of North America, a natural history and field guide. Stanford Univ. Press, Stanford, California. 583 pp., 64 pls.

Scott, J. A. & S.O. MATTOON. 1981. Early stages of Speyeria nokomis. J. Res. Lepid. 20:12-15.

ScoTT, J. A. & G.R. Scott. 1980. Ecology and distribution of the butterflies of southern central Colorado. J. Res. Lepid. 17:73-128 (corrections 19:240).

TREUSCH, H. W. 1967. Bisher unbekanntes gezieltes Duftanbietin paarungsbereiter Argynnis paphia-Weibchen. Naturwissenschaften 54:592.

Received for publication 22 April 1987; accepted 30 November 1987.

Journal of the Lepidopterists’ Society 42(1), 1988, 14-18

POPULATION FLUCTUATIONS OF AZETA VERSICOLOR (FABRICIUS) (NOCTUIDAE) ON GLIRICIDIA SEPIUM (JACQ.) (FABACEAE) IN NORTHEASTERN COSTA RICA

ALLEN M. YOUNG

Invertebrate Zoology Section, Milwaukee Public Museum, Milwaukee, Wisconsin 53233

ABSTRACT. Counts of early stages, especially caterpillars, of Azeta versicolor on the host tree Gliricidia sepium planted as shade cover in a vanilla plantation were made intermittently during five years. Based on field observations and rearings, the mature caterpillar and pupa were described, noting two distinct color morphs in the former. Tachinid parasites were also noted. Caterpillar abundance was analyzed and interpreted in relation to monthly rainfall and leaf-flushing in the host tree, since caterpillars feed preferentially on new (flush) leaves. Numbers of caterpillars were highly correlated with monthly rainfall. It is concluded that population cycles of the moth are regulated by seasonal patterns of leaf-flushing in the host.

Additional key words: immature stages, leaf flushing, population dynamics.

Impact of seasonal fluctuations in rainfall on leaf-flushing of semi or fully deciduous host trees is a major environmental factor molding population dynamics of noctuids and other Lepidoptera in the tropics (Vaishampayan & Veda 1980, Blair 1982, Tucker & Pedgley 1983). Fabaceous legume crops in the tropics are especially preferred hosts of noctuid and pyralid defoliators, with seasonal patterns of population outbreaks typical for several of these host species (Bradley & Carter 1982, Panchabhavi & Holihosur 1982). In many species, caterpillars preferentially defoliate immature leaves or other most nutritious tissues of the host, which are often only seasonally available (Futuyma & Wasserman 1980, Bracken 1984). Here I report seasonal abundance pattern of immature stages for the noctuid moth Azeta versicolor (Fa- bricius) on leaves of the fabaceous legume tree Gliricidia sepium (Jacq.) planted as shade cover in a vanilla plantation.

METHODS

Counts of life stages of Azeta versicolor were obtained on 16 dates between March 1982 and June 1987 at “Finca La Tirimbina,” near La Virgen (10°23'N; 84°07”W; 200 m elev.), Sarapiqui District, Heredia Province. Within a ca. 1600 m? plot containing about 900 trees of Gliricidia sepium planted a few years earlier to shade vanilla plants, 30 arbitrarily selected trees (canopy height ca. 3 m) were censused for Azeta versicolor caterpillars at various times. The medium-sized (40 mm wingspan) adults and caterpillars were readily recognizable in field censuses: adult moths are drab greenish brown with striking red ab-

VOLUME 42, NUMBER 1 15

dominal coloration, and yellowish mature caterpillars usually rest close to the base of host trees, typically on stems and leaves of vanilla orchid vines and other epiphytes under the trees.

On a given caterpillar census, as many as 100 samples of both mature or immature leaves and stems on each tree (usually up to height of 1.5 m) were searched for “young” caterpillars (mixed early instars) and eggs. Condition of canopy foliage of Gliricidia sepium was also noted (such as presence or absence of flush leaves), providing a qualitative picture of local timing of peak flushing periods in relation to seasonality. A total of 80 caterpillars (later instars) were placed in clear-plastic bags containing fresh cuttings of G. sepium and kept tightly shut for rearing. Parasitism of caterpillars and pupae was noted from this sample.

RESULTS

Natural history. In both of two color morphs of the final stage caterpillar, roughly equal in abundance and not sexual dimorphism, the head is pinkish white with black dots. Thoracic and abdominal regions of the mature caterpillar (40 mm long by 5 mm wide) have eight lengthwise narrow bands, which, in the dark form are as follows, dorso medial to latero ventral: (1) deep yellow; (2) faintly yellow edged in black; (3) pale bluish streaked with tiny black lines and a single round black dot on each segment; (4) pale bluish yellow; (5) wide pale blue; (6) lateral (spiracular) stripe pale blue with thin black line medially and reddish spiracle openings, each with a black dot dorsoanteriorly and yellow dot ventroposteriorly; (7) yellow with black edging ventrally; (8) grayish with raised black dot, one per segment. Prolegs pinkish, each with yellow dot laterally, ringed with black. Glossy black elongate setae on profuse raised areas of cuticle. Anal clasper faintly pinkish; true legs reddish. In the light form, there are no black stripes bordering other stripes.

The reddish brown pupa (20-22 mm long by 5-6 mm wide) occurs in a loosely constructed cocoon of host leaves pulled together and an- chored with light brown silk. Both caterpillar and pupa thrash about vigorously when picked up. Adults are active throughout the day, and are skittish and difficult to capture with an insect net. The spherical, glossy yellow eggs are placed singly on the undersides of G. sepium leaves. Of 257 eggs discovered in the field, ca. 70% were on immature (meristem) leaves. As noted above, mature caterpillars rest on vanilla vines and other epiphytic debris on host trunks during daytime, and are chiefly nocturnal feeders, crawling into the G. sepium canopy to feed. Each of 3 pupae (out of 30 reared from collected caterpillars) yielded 1 tachinid parasite.

16 JOURNAL OF THE LEPIDOPTERISTS SOCIETY

RAINFALL (MM)

1100

1000-

800

TOTAL CATERPILLARS

700

600 300

1 71 1 m1

400 200

100

100 #

N \ | Cae 5 res J FMAMJJ ASONDJ FMAMJ J ASONDJ FMAMJ J ASONDJ FMAMJJ ASONDJ FMAMJ J ASONDJ FMAMJ JASOND 1982 1983 1984 1985 1986 1987 MONTHS AND YEARS

Fic. 1. Monthly total rainfall (line), total numbers of Azeta versicolor caterpillars (vertical bars, 5th instars represented by hatching), timing of leaf-flushing (brackets), and periods of caterpillar absenses (x), in the 36-tree subplot of Gliricidia sepium at “Finca La Tirimbina.” Rainfall data courtesy of Finca La Tirimbina.

Seasonal population fluctuations and leaf-flushing. Abundance of Azeta versicolor caterpillars on sampled Gliridia sepium trees varied greatly among census dates (Fig. 1). Aside from an occasional hesperiid and limacodid caterpillar, I did not observe other herbivores abundant on these trees. When the data are examined relative to rainy and dry season periods at La Tirimbina, two patterns become apparent: (1) the highest numbers of mature and partly grown caterpillars occurred in the rainy season, especially June-August, approximately during the first half of the lengthy rainy season characteristic of this locality; (2) cat- erpillars are absent during the dry season (February—March) (Fig. 1). A high positive correlation resulted between numbers of caterpillars and monthly rainfall (r = 0.81, P < 0.01).

Also during July-August, as many as 500 adults were counted within a 600 m? strip of low vegetation bordering one side of the vanilla grove during a 2 morning census (0800-1000 h). As many as 100 eggs were counted within the 36-tree subplot on a single day in July or August,

VOLUME 42, NUMBER 1 t7

and none were found in February or March. During dry months, host trees are partly deciduous, and only mature leaves are present. Flow- ering in G. sepium at La Tirimbina is most intense during March and early April. During the first three months of the rainy season, G. sepium exhibits intensive leaf flushing (Fig. 1).

The highest population density of Azeta versicolor at La Tirimbina follows intense flushing of new leaves on larval host trees. The increased availability of immature (flush) leaves during the beginning of the rainy season provides an abundant food resource for larvae. Population build- up can be so intense in the rainy season as to result in 80-100% defo- liation of G. sepium on some plots. I conclude that the breeding pop- ulation of this Neotropical noctuid fluctuates in size throughout the year at La Tirimbina in a consistent manner, and in response to the seasonal leaf-flushing cycle of G. sepium.

DISCUSSION

Some tropical legume crops attacked by host-specific noctuids and other moths undergo severe defoliation at certain times of year (Singh & Budhraja 1980). Legume tree species typically planted as a permanent shade over perennial crops in the tropics such as cacao, coffee, and vanilla, including G. sepium (Inostrosa & Fournier 1982), and others such as Erythrina (Borchert 1980) undergo pronounced seasonal cycles in leaf-flushing in direct response to water-stress and rehydration (Reich & Borchert 1982). The complete absence of Azeta versicolor caterpillars on Gliricidia sepium in the dry season at La Tirimbina is due to absence of immature (newly flushed) leaves. Thus, availability of edible leaf tissues, a consequence of seasonally regulated hostplant leaf-flushing, determines temporal pattern of population build-up in this noctuid. The degree to which A. versicolor exploits other larval host plants at La Tirimbina is unknown.

Skittish behavior of the diurnally active adults, and their vivid red abdominal colors, suggest aposematism, perhaps a consequence of larval feeding on G. sepium, a species well known for high concentrations of courmarin compounds in its leaves (Allen & Allen 1981). Marked build- up of the adult population in the first half of the rainy season at La Tirimbina suggests a population structure in which biotic regulation of the herbivore may be minimal.

Gliricidia sepium is capable of producing a new flush of leaves following a period of intense herbivory by (J. R. Hunter & A. M. Young pers. obs.). The ability of G. sepium to recover rapidly from intense defoliation may be mediated in large part by the tree’s capacity to fix nitrogen in the soil.

18 JOURNAL OF THE LEPIDOPTERISTS SOCIETY

ACKNOWLEDGMENTS

Fieldwork was supported in part by grants from the American Cocoa Research Institute. I thank J. R. Hunter for allowing me to conduct research at his farm, Finca La Tirimbina. I thank Jorge Mejias of La Tirimbina for field assistance, and R. W. Poole of the Systematic Entomology Laboratory, U.S. Department of Agriculture, for identifying the moth. Voucher specimens of the moth are deposited in the collections of the Milwaukee Public Museum. Comments of two reviewers, and statistical advice froin R. E. Spieler were helpful in revising the manuscript.

LITERATURE CITED

ALLEN, O. N. & E. K. ALLEN. 1981. The Leguminosae: A source book of characteristics, uses and nodulation. Univ. Wisconsin Press, Madison, Wisconsin. 812 pp.

Bair, B. W. 1982. Seasonal abundance of Agrotis segetum (Denis & Schiff.) and A. ipsilon (Huefn.) (Lepidoptera: Noctuidae) in Zimbabwe, and a method of forecasting post-winter population densities. J. Entomol. Soc. South Africa 45:201-215.

BORCHERT, R. 1980. Phenology and ecophysiology of tropical trees: Erythrina peop- pigiana O. F. Cook. Ecology 61:1065-1074.

BRACKEN, G. K. 1984. Within plant references of larvae of Mamestra configurata (Lepidoptera: Noctuidae) feeding on oilseed rape. Can. Entomol. 116:45-—49.

BRADLEY, J. D. & D. J. CARTER. 1982. A new lyonetiid moth, a pest of winged-bean. Syst. Entomol. 7:1-9

FutuyMa, D. J. & S. S. WASSERMAN. 1980. Resource concentration and herbivory in oak forests. Science 210:920-922.

INostrosa, I. & L. A. FOURNIER. 1980. Efecto alelopatico de Gliricidia sepium (Jacq.) (Madero Negro). Revista Biol. Trop. 30:35-39.

PANCHABHAVI, K. S. & S. N. HoOLrHosuR. 1982. Notes on groundnut as a new host of Achaea janata Linn. (Lepidoptera: Noctuidae) at Dharwad, Karnataka. Ind. J. Agric. Sci. 52:43.

REICH, P. B. & R. BORCHERT. 1982. Phenology and ecophysiology of the tropical tree, Tabebuia neochrysantha (Bignoniaceae). Ecology 63:294-299.

SINGH, O. P. & K. BuDHRajJA. 1980. Zur Biologie des indischen Sojabohnen-Schadlings Plusia acuta Walker (Lep., Noctuidae). Anz. Schadlingsk. Pflanzenschutz Umwelt- schutz 53:184-185.

TUCKER, M. R. & D. E. PEDGLEy. 1983. Rainfall and outbreaks of the African army- worm, Spodoptera exempta (Walker) (Lepidoptera: Noctuidae). Bull. Entomol. Res. 73:195-199.

VAISHAMPAYAN, S. M. & O. P. VEDA. 1980. Population dynamics of gram podborer, Helicoverpa armigera (Hiibner) and its outbreak situation on Gram, Cicer arietinum L. at Jabalpur. Ind. J. Entomol. 42:453-459.

Received for publication 6 August 1987; accepted 16 October 1987.

Journal of the Lepidopterists’ Society 42(1), 1988, 19-31

BUTTERFLIES OF NORTHEAST TENNESSEE

CHARLES N. WATSON JR. Department of Entomology, Clemson University, Clemson, South Carolina 29634

AND

JOHN A. HYATT 439 Forest Hills Drive, Kingsport, Tennessee 37663

ABSTRACT. Here we give results of a 10-year survey of butterflies in a seven-county, 7000 km? area of NE Tennessee. Ninety-one species are listed and their seasonal occurrence tabulated on a 10-day basis. Twenty-seven species are judged to be univoltine, twenty- nine bivoltine, and twenty-one multivoltine. The remainder are thought to be migrants or strays that do not overwinter in NE Tennessee. Comparison of our species list with that of SW Virginia and N Georgia indicates the fauna lacks a number of lowland species that occur in N Georgia, and some typically northern species in SW Virginia. Ten species known to occur in both comparison areas, but not recorded here, will probably be found in the future.

Additional key words: Appalachians, biogeography, survey, Georgia, Virginia.

There is little published information on the butterfly fauna of Ten- nessee (Field et al. 1974). Osburn (1895a, 1895b) lists 70 species oc- curring around Nashville. Richards (1932) provides some Tennessee records. Watson (1946) and Snyder (1957) list some species occurring in the Smoky Mountains. The best source for the State as a whole is Opler (1983) which contains county distribution maps for all species occurring in the eastern U.S.

We have collected extensively in NE Tennessee for more than 10 years. Here we summarize results of our collecting, make comparisons with other areas in the S Appalachian region, and list additional species likely to occur in NE Tennessee.

STUDY AREA

The area encompasses seven counties in NE Tennessee with a total area of 7000 km? (Fig. 1). Two physiographic subdivisions of the S Appalachian region are represented. The SE portion of the area lies within the Blue Ridge Province, the remainder in the Ridge and Valley Province.

The peaks of the Blue Ridge are known locally as the Unaka Mountains. They are characterized by rugged terrain and heavily forested slopes. Elevations vary from 450- 600 m in the narrow valleys to 750-1900 m on the peaks. Underlying sedimentary and metamorphic rocks are Cambrian and Pre-Cambrian in age. Soils tend to be sandy and acidic. Most of this portion of the area lies within the Cherokee National Forest (Miller 1974, USDA 1958, 1956, 1985).

The Ridge and Valley portion is underlain by strongly folded sedimentary rocks of Ordovician and Cambrian age. Differential weathering has resulted in long, narrow sandstone ridges trending NE to SW, alternating with valleys developed on less resistant limestone and shale. The easternmost valley is broad and part of a series of connecting valleys extending from Pennsylvania to Alabama commonly called the Great Valley

20 JOURNAL OF THE LEPIDOPTERISTS’ SOCIETY

VIRGINIA

> i

SULLIVAN

-~ CARTER

15 MI 20 KM

Fic. 1. Study area in NE Tennessee, showing county boundaries. Dashed line is approximate boundary separating Blue Ridge Province (Unaka Mts., SE) from Ridge and Valley Province (NW).

Within it there is local relief in the form of shale knobs and entrenched streams. To the NW, straddling the border of Greene and Hawkins counties, and extending into SW Sullivan Co., are a group of ridges collectively called Bays Mountain. Another prominent feature, Clinch Mountain, runs through NW Hawkins Co. Elevations average lower in the Ridge and Valley Province, ranging from 300 m in the valleys to 600-900 m on ridges. Ridge soils are generally sandy, shallow, and unproductive while valley soils developed on limestone are rich and fertile (Fenneman 1938, Miller 1974, U.S. Dep. Agric. 1953b, 1958a, 1958b, 1979, 1985).

The entire area is drained by the Holston River and its tributaries, part of the Tennessee River drainage system. The rivers have been extensively impounded for flood control and power generation (Hunt 1967).

Climate is characterized by mild winters and warm summers. Average annual precip- itation is 100-150 cm except at highest elevations where it may exceed 200 cm. Topog- raphy and altitudinal differences cause much local variation in climate. As a rule, S- and W-facing slopes are drier than those facing N and E. Average frost-free season varies from 190 days in NW valleys to 150 days in the Unaka mountains (Walker 1969, U.S. Dep. Agric. 1953, 1979).

Before European settlement, the area was covered with oak-chestnut forest. Clearing of valleys for agriculture, logging in the mountains, and chestnut blight decimated primary forests, especially in the Ridge and Valley. Today forests are concentrated in the Unaka Mountains and on the NW ridges. At lower elevations, oaks (Quercus spp.), hickories (Carya spp.), yellow poplar (Liriodendron tulipfera L.) and other hardwoods are common, often mixed with hemlock (Tsuga spp.), and several pines (Pinus spp.). The Unaka Mountains are high enough to show altitudinal zonation. Above 900 m, northern forest types such as sugar maple (Acer saccharum Marsh.), beech (Fagus grandifolia Ehrh.), and yellow birch (Betula alleghaniensis Britton) are common. Above 1500 m, red spruce (Picea rubens Sarg.), and fraser fir (Abies fraseri [Pursh.] Poir) predominate. Treeless, dome-shaped summits called balds occur on some peaks. In the Ridge and Valley, stands of red cedar (Juniperus virginiana L.) are common in old fields on limestone soils. Marshes and canebrakes are rare throughout, most having been drained, cleared, or inundated by reservoirs (Braun 1950, Walker 1969, U.S. Dep. Agric. 1953a, 1953b, 1956, 1958a, 1958b, 1979, 1985).

VOLUME 42, NUMBER 1 ya

METHODS

Most records come from collections and fields notes made by the authors from 1975 through 1986. Additional records were obtained from participants in a Southern Lepi- dopterist Society field meeting in the area in 1980, and from collections made by students at Sullivan (County) High School during fall 1977 and 1978. Collections at the U.S. National Museum (USNM) and the Carnegie Museum of Natural History (CMNH) were examined, but no additional records were found. Most specimens are retained in the authors’ collections; others have been placed in USNM and CMNH. Some identifications were confirmed by C. V. Covell Jr., University of Louisville, and by R. K. Robbins and J. M. Burns (USNM). Butterfly nomenclature follows Hodges et al. (1983).

To facilitate comparison with NE Tennessee, we define SW Virginia as Giles, Mont- gomery, and Floyd counties and those counties to the SW entirely or predominantly within the transition zone of Clark and Clark (1951). North Georgia is defined as those counties entirely or predominantly within the mountain region of the State as defined by Harris (1972). Species records for these regions were obtained from Opler (1983), Clark and Clark (1951), and Harris (1972).

RESULTS AND DISCUSSION

We recorded 91 species of butterflies and skippers from NE Tennessee (Table 1). In addition, specimens of Celastrina ladon form neglecta- major Tutt, considered by some to be a distinct species (Opler & Krizek 1984), have been collected in May and early June. An old sight record for Anaea andria Scudder for which we do not have a precise date is not included in the table but is discussed below.

The species found in NE Tennessee can be considered as falling into two categories: residents, which overwinter in the area; and migrants or strays, which do not normally overwinter in the area, although many regularly occur in summer and fall.

A number of resident species are rare or local in distribution, but only one appears limited to a particular part of the study area. Speyeria aphrodite (F.) has been collected only in the Blue Ridge, where it is often common at elevations above 600 m.

Analysis of flight-period data in Table 1 to determine number of broods for resident species is complicated by the fact that the flight period of a species at any particular locality may vary from year to year due to climatic and biological factors. Flight periods are also affected by elevation, beginning and ending one to three weeks later at high elevations in the Blue Ridge than in the Ridge and Valley. For example, summer brood Erynnis horatius (Scudder & Burgess) has been collected at Bays Mountain Park (600 m) in Sullivan Co. from late June through mid-August, but a fresh specimen was collected in Carter Co. at 1200 m on 8 September.

We believe the following residents are univoltine in NE Tennessee:

Thorybes bathyllus (J. E. Smith) E. brizo (Bdv. & Leconte) T. pylades (Scudder) E. juvenalis (F.) Erynnis icelus (Scudder & Burgess) Wallengrenia egremet (Scudder)

22 JOURNAL OF THE LEPIDOPTERISTS SOCIETY

TABLE 1. Temporal distribution of butterfly species adults in NE Tennessee.

Species 20 31 10 20 30 10 20 31

Hesperiidae Epargyreus clarus X xX xX xX 4 Autochton cellus xX Achalarus lyciades xX Thorybes bathyllus T. pylades xX

Staphylus hayhurstii

Erynnis icelus

E. brizo

E. juvenalis

E. horatius

E. baptisiae

Pyrgus communis x

Pholisora catullus

Nastra lherminier

Ancyloxypha numitor

Thymelicus lineola

Hylephila phyleus

Polites coras x

P. themistocles

P. origenes xX

Wallengrenia egeremet

Pompeius verna x

Atalopedes campestris

Atrytone delaware

Poanes hobomok

P. zabulon xX xX

Euphyes ruricola metacomet

Amblyscirtes hegon xX

A. aesculapius

A. vialis xX X

Papilionidae

rm

~~ KX

rm xX PX rm XK rm XK

Pr

mx MX

Battus philenor

Papilio polyxenes asterius

P. cresphontes

P. glaucus xX P. troilus

Eurytides marcellus xX

pd dx >> > > x >> > D> Dd be Dd > dd ><

Pieridae

Pontia protodice

Artogeia virginiensis

A. rapae x Euchloe olympia

Falcapica midea

Colias philodice xX C. eurytheme xX Phoebis sennae eubule

Eurema lisa

E. nicippe

ee eeEeeeeeeeeeSFSFeee

~ KKM KKK KM KK KKM KKK x Kw MM ras

> >< >

23 ce

20

Nov 1

10

21- 3]

Oct. 1l- 20

jz

10

21- 30

Sept. 1l- 20

Continued. = 10

21- 31

Aug. ll- 20

i=

PABEE I 10

21- 31

July ee Se a Oe eo Une Se Le paoIe Ie A 10 20

21- 30

June 1l- 20

1 10

VOLUME 42, NUMBER 1

» ro ba bd be * x“ pe > bd bd x oe * * ree pee 5d » ce Pa bd x Dr >< re x x >< ex x OX <> x DM Od re ba mx x «Mx ex x bd > >< >< rp Mx xx OM <r OX ra x x rr OO KO xx xX x x << bd bd ><! MX bd » x < x x >< r <>< Mx. <x < » be < > >< >< od ><! < » x y < «xx » es * rd Od » >< rd >< y x <x xxx xX re rd >< re >< >

Species

Lycaenidae

Feniseca tarquinius Lycaena phlaeas americana Harkenclenus titus mopsus Satyrium calanus falacer S. caryaevorum

S. liparops strigosum Calycopis cecrops

Mitoura grynea

Incisalia augustus croesioides I. henrici

I. niphon

Parrhasius m-album Strymon melinus

Erora laeta

Everes comyntas Celastrina ladon

C. ebenina

Glaucopsyche lygdamus Libytheidae

Libytheana bachmanii Nymphalidae

Polygonia interrogationis P. comma

Nymphalis antiopa Vanessa virginiensis

V. cardui

V. atalanta

Junonia evarete

Euptoieta claudia

Speyeria diana

S. cybele

S. aphrodite

Clossiana bellona toddi Phyciodes tharos Charidryas nycteis Euphydryas phaeton Basilarchia arthemis astyanax B. archippus

Apaturidae Asterocampa celtis A. clyton Satyridae

Enodia anthedon E. creola

Cyllopsis gemma Hermeuptychia sosybius Megisto cymela Cercyonis pegala Danaidae

Danaus plexippus

TABLE l. Mar. ie ee 20 31 x xX xX x x x xX x

Continued.

xx KR MM

rr

KKM KM

mx

KKM MMM RK

mx

rm xX

10

rm xX

~MK MMM RK OK

mx

xx PM

rs XK

PK XS

mx KKM MX

rs XK

1]-

20

Nov. te

10

2\- 31

Oct. (i= 20

1 10

21- 30

Sept. ll- 20

Continued. l= 10

2\- 31

Aug. ll- 20

ie

TABLE 1. 10

21- 31

July nr ine Serre aol Ola Reo OLS. ee eas Ole.) hee don 10 20

21- 30

June ll- 20

1 10

KKK x va * * * a ad ~ %X ~~ * va ~ va

KK RRM KK KK KM ~ xX

~ xX x «KK KK ~ xX

Kr mK x KK KK ~ xX KA KM KR KR KR KK ~ X

KKK ~ x< x «Kx

mK MK RK RRR KM Km KKK KK KK

* x KK KKK

Kee MM KR KK KK

~ * * *

mx x Ke KR KKK KK XK

26 JOURNAL OF THE LEPIDOPTERISTS SOCIETY

Poanes hobomok (Harr.) Incisalia augustus croesioides (Scudder) Euphyes ruricola metacomet (Harr.) I. henrici (G. & R.)

Amblyscirtes hegon (Scudder) I. niphon (G. & R.)

Artogeia virginiensis (Edw.) Celastrina ebenina Clench

Euchloe olympia (Edw.) Glaucopsyche lygdamus (Doubleday) Falcapica midea (Hbn.) Speyeria diana (Cram.)

Harkenclenus titus mopsus (Hbn.) S. aphrodite

Satyrium calanus falacer (Godt.) Euphydryas phaeton (Drury)

S. caryaevorum (McD.) Megisto cymela (Cram.)

S. liparops strigosum (Harr.) Cercyonis pegala (F.)

Speyeria cybele (F.) flies from May through September and would

appear to be multivoltine, but the long flight period is caused by stag-

gered emergence of a single brood (Opler & Krizek 1984, Scott 1986). The following are bivoltine:

Autochon cellus (Bdv. & Leconte) Mitoura grynea (Hbn.)

Achalarus lyciades (Gey.) Nymphalis antiopa (L.)

Nastra lherminier (Latr.) Charidryas nycteis (Doubleday) Polites coras (Cram.) Basilarchia arthemis astyanax (F.) P. themistocles (Latr.) B. archippus (Cram.)

P. origenes (F.) Asterocampa celtis (Bdv. & Leconte) Pompeius verna (Edw.) A. clyton (Bdv. & Leconte) Atrytone delaware (Edw.) Enodia anthedon A. H. Clark Poanes zabulon (Bdv. & Leconte) E. creola (Skin.)

Lycaena phleas americana (Harr.) Cyllopsis gemma (Hbn.) Calycopis cecrops (F.) Hermeuptychia sosybius (F.)

Fresh Basilarchia archippus and B. arthemis astyanax taken in October and early November indicate that partial third broods are produced when mild weather persists well into fall.

Additional species are probably bivoltine, though not apparent from our data. Erynnis horatius (Scudder & Burgess) and E. baptisae (Fbs.) should have spring broods on the wing in April and May. They have likely been overlooked amid large numbers of E. juvenalis flying at that time. Pholisora catullus (F.) is also likely to have a spring brood, and is probably more common than our records suggest. Erora laeta (Edw.), Amblyscirtes aesculapius (F.), A. vialis (Edw.), and Staphylus hayhurstii (Edw.) have been taken only in spring or early summer. All four species probably have second broods in summer overlooked due to very local occurrence.

Another group of resident species are multivoltine, with three or more broods per year:

Epargyreus clarus (Cram.) Artogeia rapae (L:) Ancyloxypha numitor (F.) Colias philodice Godt. Battus philenor (L.) C. eurytheme Bdv. Papilio polyxenes asterius Stoll Feniseca tarquinius (F.) P. glaucus L. Strymon melinus Hbn. P. troilus L. Everes comyntas (Godt.)

Eurytides marcellus (Cram.) Celastrina ladon (Cram.)

VOLUME 42, NUMBER 1 OF

Polygonia interrogationis (F.) V. atalanta (L.) P. comma (Harr.) Clossiana bellona toddi (Holl.) Vanessa virginiensis (Drury) Phyciodes tharos (Drury)

One additional species, Parrhasius m-album (Bdyv. & Leconte), is probably multiple brooded. We have taken a worn specimen in SW Virginia near the Tennessee line in early May, and sources indicate that a third brood in late August-September is likely (Opler & Krizek 1984, Scott 1986).

We consider the following species to be migrants or strays:

Pyrgus communis (Grt.) E. nicippe (Cram.)

Hylephila phyleus (Drury) Libytheana bachmanii (Kirtland) Atalopedes campestris (Bdv.) Vanessa cardui (L.)

Papilio cresphontes (Cram.) Junonia coenia (Hbn.)

Pontia protodice (Bdv. & Leconte) Euptoieta claudia (Cram.) Phoebis sennae eubule (L.) Danaus plexippus (L.)

Eurema lisa (Bdv. & Leconte)

Most of these species overwinter in the SE coastal plain where they are multivoltine. As their populations expand during the summer, they move N and W,, often penetrating into the Appalachians. Although they may reproduce during summer and fall, they generally cannot survive winter in NE Tennessee. There are exceptions, as evidenced by an April record for Pyrgus communis. In NE Tennessee, migrants are most likely to be found from mid-August through October. During this period Atalopedes campestris is one of the most common butterflies in gardens and disturbed areas. At the other extreme, Papilio cresphontes, Pontia protodice, and Hylephila phyleus are known from only one or two records. Remaining species are usually present every year in varying numbers. Libytheana bachmanii differs from the usual migrant pattern of occurrence in that it has been found from mid-June through mid- August. It is regularly present, but usually only as one or two individuals at a given time and place. We include it as a migrant because we have never collected overwintered individuals in spring.

We are not certain of the status of Thymelicus lineola (Ochs.) in NE Tennessee. It has been taken only once, near a campground in Sullivan Co. adjacent to a N-S interstate highway. This European species has spread rapidly southward since it was accidently introduced into Can- ada around 1910 (Scott 1986), and there are records from SE Kentucky and SW Virginia (Opler 1983). If not already a resident, it is likely to become one soon.

While walking in the late 50’s or early 60’s, the senior author saw a single Anaea andria flying in a clover field in Sullivan Co. Without a net he could not capture it, but followed it for a distance and was certain of the identification. This species is resident around Center Hill

28 JOURNAL OF THE LEPIDOPTERISTS SOCIETY

Lake, 130 km E of Nashville, and the junior author recently captured several overwintered individuals in Lee Co., SW Virginia. While we have not seen A. andria in NE Tennessee during the past 10 years, it is somewhat migratory (Scott 1986), and should be expected on occasion.

Southwest Virginia and N Georgia have more species than NE Ten- nessee, 120 and 108, respectively. This disparity is at least partly due to the fact that Virginia and Georgia have been collected longer than NE Tennessee.

Amblyscirtes aesculapius was the only species found in NE Tennessee that has not been recorded from SW Virginia. The Clarks (1951) re- corded it only from the coastal plain of Virginia, but there are records from E Kentucky, and it probably occurs locally along rivers in SW Virginia. Euchloe olympia and Clossiana bellona toddi are resident in NE Tennessee, but are not known to occur in N Georgia. These species are at or near the limits of their ranges in NE Tennessee.

The 39 species recorded from SW Virginia and/or N Georgia not collected in NE Tennessee are listed in Table 2. Sixteen of these species are known only from SW Virginia, nine from N Georgia only, and fourteen occur in both regions.

Many species recorded from SW Virginia but not from NE Tennessee are northern species whose ranges extend southward in the Appalachian region. Southwest Virginia includes the entire breadth of the moun- tainous Blue Ridge Province, and elevations in the Valley and Ridge Province exceed 1200 m in places (Fenneman 1938). More extensive areas of high elevation coupled with higher latitude make SW Virginia more hospitible for some northern species than NE Tennessee.

Species recorded from N Georgia but not NE Tennessee include Satyrium kingi (Klots & Clench), Amblyscirtes carolina (Skin.), Agrau- lis vanillae (L.), and other species more typical of the lowland Piedmont and Coastal Plain provinces. Relative to NE Tennessee, the Appalachian region of N Georgia is lower in elevation and has a milder climate. In particular, the prominent ridges that characterize the Ridge and Valley further N are absent (Fenneman 1938). Broad valleys open onto the Piedmont, while the oak-pine forest association and red-yellow podzolic soils characteristic of the Piedmont extend into the Georgia portion of the Ridge and Valley (Braun 1950, Walker 1969). These climatic and topographic factors create favorable habitats for some lowland species, and provide easy access for migrants.

We predict that the following species in SW Virginia and N Georgia will eventually be found resident in NE Tennessee:

Thorybes confusis Bell Wallengrenia otho (J. E. Smith) Erynnis martialis (Scudder) Atrytonopsis hianna (Scudder) Hesperia metea (Scudder) Satyrium edwardsii (G. & R.)

VOLUME 42, NUMBER Il 29

TABLE 2. Butterfly species occurring in SW Virginia (VA) and N Georgia (GA) but not recorded from NE Tennessee.

Species State Thorybes confusis VA, GA Erynnis martialis VA, GA E. zarucco VA, GA E. lucilius VA E. persius VA Pyrgus centaurae VA Lerema accius VA, GA Hesperia metea VA, GA H. leonardus VA H. sassacus VA Polites mystic VA P. vibex VA Wallengrenia otho VA, GA Atrytone arogos VA Euphyes conspicua VA E. bimacula VA Atrytonopsis hianna VA, GA Panoquina ocola VA, GA Amblyscirtes carolina GA A. alternata GA Megathymus yuccae GA M. harrisi GA Zerene caesonia GA Eurema daira VA, GA Atlides halesus VA Satyrium edwardsii VA, GA S. kingi GA Incisalia irus VA, GA Fixenia ontario VA Calephelis borealis VA Agraulis vanillae GA Charidryas gorgone GA Speyeria idalia VA Clossiana selene VA Phyciodes batesii VA, GA Polygonia progne VA P. faunus VA, GA Enodia portlandia GA Satyrodes appalachia VA, GA

Incisalia irus (Godt.) Polygonia faunus (F.) Phyciodes batesii (Reak.) Satyrodes appalachia (R. Chermock)

Hesperia leonardus (Harr.), H. sassacus (Harr.), Speyeria idalia (Drury), and Polygonia progne (Cram.) have been recorded from bor- dering counties in Virginia and North Carolina (Opler 1983) and also seem likely to be found in NE Tennessee eventually.

It is possible that Amblyscirtes celia belli H. A. Freeman occurs in NE Tennessee. We have taken it flying with Wallengrenia otho in

30 JOURNAL OF THE LEPIDOPTERISTS SOCIETY

moist woods beside an arm of Loudon Reservoir near Knoxville, Ten- nessee, about 65 air km SW of our study area boundary. Similar habitats should occur around reservoirs in NE Tennessee.

Additional migratory species such as Erynnis zarucco (Luc.) and Panoquina ocola (Edw.) may eventually turn up also, but lack of direct access from the Piedmont is a hindrance to the movement of such species; to enter NE Tennessee, they must first pass through the rugged North Carolina portion of the Blue Ridge, or travel a considerable distance up valleys from Georgia.

Concentration of collecting efforts on species listed above should increase the number of butterfly species known from NE Tennessee to between 100 and 110.

ACKNOWLEDGMENTS

We thank P. A. Opler and an anonymous reviewer for suggesting useful changes in a draft of this paper; also Tom Bowman of Bays Mountain Park, Sullivan Co., Tennessee, for permitting us to collect within the Park.

LITERATURE CITED

BRAUN, L. E. 1950. Deciduous forests of eastern North America. Blakiston, Philadelphia. 596 pp.

CiarK, A. H. & L. F. CLARK. 1951. The butterflies of Virginia. Smiths. Misc. Coll. 116: 1-239.

FENNEMAN, N. M. 1938. Physiography of the eastern United States. McGraw Hill, New York. 714 pp.

FIELD, W. D., C. F. Dos Passos & J. H. MASTERS. 1974. A bibliography of the catalogues, lists, faunal and other papers on the butterflies of North America arranged by state and province. Smiths. Contrib. Zool. 157. 104 pp.

Harris, L. 1972. Butterflies of Georgia. University of Oklahoma Press, Norman. 326 pp.

HopcEs, R. E. (ed.). 1983. Check list of the Lepidoptera of America north of Mexico. E. W. Classey Ltd., London. 284 pp.

Hunt, C. B. 1967. Physiography of the United States. Freeman, San Francisco. 480 pp.

MILLER, R. A. 1974. The geologic history of Tennessee. Tenn. Dep. Cons. Div. Geol. 63 pp.

Op_LeER, P. A. 1983. County atlas of eastern United States butterflies (1840-1982). US. Fish. Wild]. Serv. Div. Biol. Serv., Washington, D.C. 86 pp.

OpLeER, P. A. & G. O. KRIZEK. 1984. Butterflies east of the Great Plains. Johns Hopkins, Baltimore. 294 pp.

OsBURN, W. 1895a. Rhopalocera of Tennessee. Entomol. News 6:245-248.

1895b. Rhopalocera of Tennessee—II. Entomol. News 6:281-284.

RICHARDS, A. G. 1932. Distributional studies on southeastern Rhopalocera. Bull. Brook- lyn Entomol. Soc. 26:234-253.

ScorT, J. A. 1986. The butterflies of North America. Stanford University Press, Stanford, California. 583 pp.

SNYDER, K. D. 1957. Checklist of insects of Great Smoky Mountains National Park. Privately printed. 78 pp.

U.S. Dep. AGRIC. 1953a. Soil survey of Carter County, Tennessee. Series 1942, no. 4. 199 pp.

1953b. Soil survey of Sullivan County, Tennessee. Series 1944, no. 2. 199 pp.

—— 1956. Soil survey of Johnson County, Tennessee. Series 1946, no. 2. 150 pp.

1958a. Soil survey of Greene County, Tennessee. Series 1947, no. 7. 89 pp.

VOLUME 42, NUMBER 1 3]

1958b. Soil survey of Washington County, Tennessee. Series 1948, no. 5. 91 pp.

1979. Soil survey of Hawkins and Hancock counties, Tennessee. 84 pp.

1985. Soil survey of Unicoi County, Tennessee. 99 pp.

WALKER, L. C. 1969. Geography of the southern forest region. Division of Forestry, Stephen F. Austin State University, Austin, Texas. 68 pp.

WATSON, J. R. 1946. Some August skippers of the Great Smoky Mountain National Park and vicinity. Florida Entomol. 28:50-53.

Received for publication 2 January 1987; accepted 19 October 1987.

Journal of the Lepidopterists’ Society 42(1), 1988, 31

GENERAL NOTE

GLASSBERG, LEHMAN, AND PELLMYR COLLECTIONS TO THE SMITHSONIAN INSTITUTION

Dr. Jeffrey S. Glassberg has donated his collection of New World butterflies to the National Museum of Natural History (Smithsonian Institution). It consists of more than 2000 specimens, primarily Neotropical Theclinae (approximately 350 species). Dr. Glass- berg is a molecular geneticist who lives in Chappaqua, New York, and is Vice President for Research of Lifecodes Corp. He has a strong interest in conservation and butterfly watching, and is currently President of the Xerces Society.

The Smithsonian Institution has received Mr. Robert Lehman’s collection of Honduran Lepidoptera. There are 4222 meticulously spread specimens representing 1852 species, plus about 5000 papered specimens. The Macrolepidoptera are well represented, and there are many Pyralidae, Tortricidae, and Oecophoridae. Most of the specimens were collected along the wet Atlantic coast of Honduras, an area that is poorly represented in collections, and which augments the Smithsonian’s strong holdings from Mexico, Gua- temala, Costa Rica, and Panama. Mr. Lehman has been teaching elementary school science and, more recently, computer science, at the Mazapan School in La Ceiba, Honduras, for 9 years, and has been collecting in Honduras since 1968.

Dr. Olle Pellmyr has donated his collection of Fennoscandian (primarily Swedish) Lepidoptera to the Smithsonian Institution. It includes 6907 specimens of approximately 1200 species, and is rich in both Macro- and Microlepidoptera. Because so many Swedish species are close relatives of North American ones, this collection provides important comparative material. Dr. Pellmyr is an evolutionary biologist who works on chemical and ecological aspects of plant-pollinator mutualism and lepidopteran courtship behavior. He is a Swedish national, and currently a research scientist at the State University of New York at Stony Brook.

None of the collections contains primary type specimens.

ROBERT K. ROBBINS AND GARY F. HEVEL, Department of Entomology, NHB Stop 127, Smithsonian Institution, Washington, D.C. 20560.

Journal of the Lepidopterists’ Society 42(1), 1988, 32-36

BODY WEIGHT AND WING LENGTH CHANGES IN MINNESOTA POPULATIONS OF THE MONARCH BUTTERFLY

WILLIAM S. HERMAN

Department of Genetics and Cell Biology, University of Minnesota, St. Paul, Minnesota 55108-1095

ABSTRACT. Body weights and rear wing lengths were obtained from about 1900 monarch butterflies captured near Minneapolis, Minnesota, during the past decade. Mean values for both were lowest in immigrants and highest in subsequent generations. Mean wing length was highest in males. Mean body weights of immigrant females were higher than those of males, but mean male body weights were higher than those of females in subsequent generations. The data argue against the return to Minnesota of emigrants from the previous year, and suggest that attainment of large adult size could be one reason for monarch migration to northern regions.

Additional key words: Nymphalidae, Danaus plexippus, migration, sexual differ- ences.

During the past several years workers in my laboratory have ex- amined various aspects of the biology of the monarch butterfly, Danaus plexippus L. Our studies have impressed us with the great variation exhibited by monarch populations in our locality with respect to re- productive status, hormone titers, behavior, and other variables (Her- man 1985). Monarch butterflies of both sexes also exhibit predictable changes in body weights and wing lengths during their residence in our area, and such changes are the topic of this report.

MATERIALS AND METHODS

Animals used for this study were captured near Minneapolis, Min- nesota, between 1976 and 1986. They were taken to the laboratory for measurement soon after capture, usually within a few h. Whole-body wet weights were determined to the nearest 1 mg using an analytical balance, and rear wing maximal lengths were measured to the nearest 0.5 mm with a ruler. Immigrant butterflies rarely arrive in our locality before 15 May, and most local monarchs emigrate by late September. The results are therefore for animals captured 16 May to 15 September, and data in Fig. 1 are summarized for 2-wk and 2-mo intervals during that period. All data are presented as mean + standard error; statistical analysis was done using Student’s t-test.

RESULTS

Mean wing lengths for both sexes were smallest during the 2-mo period 16 May-15 July (Fig. 1). Most of these animals were presumably immigrants from southern regions, since large numbers of monarchs

VOLUME 42, NUMBER 1 33

REAR WING LENGTH (mm) 39

$

38

ANIMAL WET WEIGHT (mg) 940 ¢

l

Ce 8s ne aaal Ol 7715. 8/1 OSA

CAPTURE DATE

Fic. 1. Wing lengths and body weights of monarch butterflies captured near Min- neapolis, Minnesota, 16 May-15 September. Data are summarized for 2-wk and 2-mo periods. Means for 2-mo periods are shown numerically, and number of individuals given in parentheses. Vertical lines indicate SE.

34 JOURNAL OF THE LEPIDOPTERISTS SOCIETY

do not typically emerge in our area until early July. Mean wing length in both sexes increased significantly (P = 0.001 for both sexes) in the 2nd 2-mo period 16 July-15 September. Presumably, most of the latter animals emerged in our locality. Mean female wing length increased 3.8% in the second 2-mo period, that of males, 3.0%. Mean male wing lengths were significantly larger than those of females in both the Ist (P = 0.001) and 2nd (P = 0.05) 2-mo periods. Mean wing lengths recorded 16 July-15 September were indistinguishable from those ob- served at emergence in monarchs reared on milkweed, Asclepias syriaca L., in our area in July and August: 37.8 + 0.1 mm (n = 100) and 38.0 + 0.1 mm (n = 83) for females and males, respectively, on the day of eclosion. Rear wing length varied from 29.5 to 42.0 mm in this study, and both extremes were observed in males.

Body weights of both sexes changed in a manner similar to that of wing lengths, with low mean values characterizing the mainly immi- grant populations of 16 May-15 July, and significantly higher mean values observed in monarchs that had presumably emerged in our area 16 July—15 September (Fig. 1). Mean body weights for females were elevated 6.7% in the 2nd 2-mo period, those of males, 20.8%. Mean female body weights were significantly larger (P = 0.001) than those of males 16 May-15 July, principally due to higher female weights of 16 May through 15 June. Male values were significantly higher (P = 0.005) than those of females during the final 2-mo period. The lowest mean values for both sexes were recorded in late June, when senescence and death of immigrants is most pronounced, and the highest were recorded in late August, when reproduction generally ceases in our area. The increasing mean weights for both sexes from 1 July to 15 August were recorded for populations consisting principally of actively reproducing monarchs of various ages. Mean body weights of wild- caught butterflies never reached the mean values (680 + 32 mg [n = 26] and 652 + 11 mg [n = 109], respectively) measured on day of eclosion for females and males reared in our area. Body weights ranged from 195 to 836 mg during this study, and both extremes were again found in males.

DISCUSSION

The data show that predictable variations occur in rear wing lengths and body weights during the period that monarch butterflies reside near Minneapolis. Small wings and low weights characterize the im- migrant population, and both parameters increase significantly in both sexes when monarchs that have apparently emerged in our area pre- dominate in the local population, as they normally do after 1 July. Causes of these variations, and their possible adaptive value, are un-

VOLUME 42, NUMBER 1 35

determined. However, the data suggest that local environmental factors (nutrient value of foodplant, temperature, or photoperiod) during June, July and August may provide optimal conditions for larval growth, and thereby result in larger adults with longer wings. If so, suboptimal conditions for larval development of the presumed immigrant gener- ation in southern areas could account for reduced size in immigrant butterflies. This line of reasoning implies that northward migration in spring could be, to at least some extent, an adaptation for locating regions that optimize adult size. Larger adults may have a greater probability of successful southward migration, survival in the overwin- tering colonies, or remigration.

The smaller wings of immigrants might somehow facilitate north- ward migration, while the larger wings of animals emerging in late August and September may be more advantageous for southward mi- gratory flights. Perhaps larger wings are more efficient for soaring and gliding, phenomena reported only for monarchs migrating south (Gibo 1981). Immigrant males with smaller wings might also be more suc- cessful at mating, as reported for males in Mexican overwintering col- onies (Van Hook 1986). James (1984) noted no significant differences in wing lengths of Australian monarchs observed during a full year.

The data on monarch body weight generally agree with those in other reports (Cenedella 1971, Brown & Chippendale 1974, Brower & Glazier 1975). Other studies have reported significantly higher body weights of males in southward migrating and overwintering monarch populations (Tuskes & Brower 1978, Chaplin & Wells 1982). However, others have apparently not observed periods in the monarch annual cycle when females are significantly heavier than males, as Fig. 1 records for immigrants to our area.

Data in Fig. 1 argue against the return to our locality of monarchs that emigrated the previous year. Our immigrants, especially females, have intermediate weights, and, based on body weight and external appearance, most appear to be young or middle-aged, certainly not old. Immigrants to our area also exhibit both senescence and precipitous weight loss (Fig. 1) within 2—4 wk after arrival, and it seems reasonable to assume that comparable rates of aging and weight loss occur after monarchs leave Mexican overwintering colonies. In view of these ob- servations, it is unlikely that overwintering monarchs could leave Mex- ican colonies in mid-March (Norman 1986), fly northward for 8-10 wk while actively breeding, and arrive in our area with body weight and external appearance comparable to young populations of July. Similarly, smaller wings of our immigrants suggest they are not members of the emigrant generation of the previous year, since emigrants have signif- icantly larger wings. In addition, monarchs captured in Mexican col-

36 JOURNAL OF THE LEPIDOPTERISTS SOCIETY

onies in February and March 1984 had wings comparable in length to our emigrants, and significantly larger than those of our immigrants (Herman unpubl.). For these reasons, the data support an earlier con- clusion (Herman 1985), and that of Malcolm et al. (1986), that most immigrants to the northern United States are probably one generation removed from individuals forming Mexican overwintering colonies.

ACKNOWLEDGMENTS

Some of this research was supported by the University of Minnesota Graduate School. Ann and Alden Mikkelsen generously assisted with the capture of many animals used in this study.

LITERATURE CITED

BROWER, L. P. & S. C. GLAZIER. 1975. Localization of heart poisons in the monarch butterfly. Science 188:19-25.

BROWN, J. J. & G. M. CHIPPENDALE. 1974. Migration of the monarch butterfly, Danaus plexippus: Energy sources. J. Insect Physiol. 20:1117-1130.

CHAPLIN, S. B. & P. H. WELLS. 1982. Energy reserves and metabolic expenditures of monarch butterflies overwintering in southern California. Ecol. Entomol. 7:249-256.

CENEDELLA, R. J. 1971. The lipids of the female monarch butterfly, Danaus plexippus, during fall migration. Insect Biochem. 1:244-247.

Giso, D. L. 1981. Altitudes attained by migrating monarch butterflies, Danaus p. plexippus (Lepidoptera: Danaidae), as reported by glider pilots. Can. J. Zool. 59:571- o72.

HERMAN, W. S. 1985. Hormonally mediated events in adult monarch butterflies. In Rankin, M. A. (ed.), Migration: Mechanisms and adaptive significance. Contrib. Mar. Sci. 27:799-815.

JAMEs, D. G. 1984. Phenology of weight, moisture and energy reserves of Australian monarch butterflies, Danaus plexippus. Ecol. Entomol. 9:421—428.

MALCOLM, S. B., B. J. COCKRELL & L. P. BROWER. 1986. Milkweed cardenolides as labels of the monarch’s spring remigration strategy. Abstracts from MONCON II, Second International Conference on the Monarch Butterfly, Natural History Museum of Los Angeles County, Los Angeles, California. P. 9.

NORMAN, C. 1986. Mexico acts to protect overwintering monarchs. Science 233:1252- 1253.

TuskEs, P. M. & L. P. BRoweErR. 1978. Overwintering ecology of the monarch butterfly, Danaus plexippus, in California. Ecol. Entomol. 3:141-154.

VAN Hook, T. 1986. Sexual selection in monarch butterflies (Danaus plexippus L.) overwintering in Mexico: Nonrandom mating via male choice and alternate male mating strategies. Abstracts from MONCON II, Second International Conference on the Monarch Butterfly, Natural History Museum of Los Angeles County, Los Angeles, California, ‘P77,

Received for publication 4 May 1987; accepted 21 October 1987.

Journal of the Lepidopterists’ Society 42(1), 1988, 37-45

HABITAT AND RANGE OF EUPHYDRYAS GILLETTI (NYMPHALIDAE)

ERNEST H. WILLIAMS Department of Biology, Hamilton College, Clinton, New York 13323

ABSTRACT. Fifteen sites occupied by Euphydryas gillettii are compared according to 10 characteristics. All sites are moist, open, mostly montane meadows, many with a history of disturbance, commonly fire. Population size correlates with relative availability of nectar but not with overall abundance of the usual hostplant, Lonicera involucrata. Habitats at higher latitudes often have a southerly exposure. Reduction in hostplant size at higher latitudes contributes to the northern range limit. Three populations likely have become extinct since 1960, but the species range does not appear to be changing.

Additional key words: nectar, Lonicera involucrata, biogeography, extinction.

Euphydryas gillettii (Barnes), a checkerspot butterfly, occurs in dis- crete, isolated populations (Williams et al. 1984) in the central and northern Rocky Mountains (Ferris & Brown 1981). It is attractive and easily caught but uncommon and not often collected. Though usually considered a montane species (Williams et al. 1984), variation in sites occupied by E. gillettii has not been studied, and lack of knowledge about its habitats has led to uncertainty about its range.

Here I report characteristics of sites occupied by E. gillettii, present range of the species, and factors influencing its distributional pattern. This study is based on direct observation of the habitats of 15 populations throughout the range, thus affording an uncommon view of habitat variability in a single insect species.

METHODS

Populations of E. gillettii were located through correspondence with collectors and researchers listed in Acknowledgments, examination of specimen labels in collections listed in Acknowledgments, and a survey of published reports (News Lepid. Soc., Seasonal Summaries 1960- 1986). When directions were sufficient to pinpoint locations on a to- pographic map, I visited the sites, and assessed relative population size and habitat characteristics.

Population size was determined by a one-day count of adults, egg masses, and larval webs. Egg masses of E. gillettii are distinctive, easily found, and readily counted, thus permitting quantitative comparisons of colony size even after the flight season; in fact, egg mass counts are better indicators of population size than adult counts because the former are independent of weather. Eggs do not begin hatching until late in the flight season (Williams et al. 1984), so developmental state of egg masses at each site indicated timing of the count relative to flight season,

38 JOURNAL OF THE LEPIDOPTERISTS SOCIETY

Females average one to two egg masses per individual (unpubl.); thus, relative population size can be estimated from sum of egg masses and adults.

‘In addition to population size, I recorded nine site characteristics, and searched for evidence of disturbance. Observations were quantified as much as possible for later analysis. Each site is marked on U.S. Geological Survey and Canada Department of Energy, Mines & Re- sources topographic maps in my possession, and latitude and elevation were measured directly from these maps. I used a compass as well as contour lines on the maps to determine exposure. I recorded number of distinct shrubs or clumps of the usual hostplant, Lonicera involucrata (Rich.) Banks (Caprifoliaceae) (Comstock 1940, Williams et al. 1984), in open areas where egg masses and adults were found. Nectar sources were identified (Hitchcock & Cronquist 19783), and relative nectar avail- ability was determined by site comparison. Nearby trees were identified and cored with a 5 mm diam increment borer for age determination. Presence and distance to standing water and streams were recorded. I inferred source and history of disturbance from characteristics such as tree species and age, charring, stems gnawed by beavers, and location in a flood plain.

RESULTS Populations

I visited 29 localities reported as sites for Euphydryas gillettii and found populations at 18. With my 2 previous study sites (Williams et al. 1984), I had a total of 15 colonies throughout the geographic distribution of the species for comparison. More than 15 egg masses and adults were found at 7 sites (“‘large”’ populations), while fewer than 15 were found at 8 sites (“‘small” populations) (Table 1).

Habitat Characteristics

All occupied sites are wet (Table 1). Most have a small stream passing through, though several are marshy without obvious flowing water; E. gillettii occurs infrequently near rivers, perhaps because of flood dis- turbance to hostplants, nectar sources, larvae, and adults. In habitat characteristics, E. gillettii is similar to its congener E. phaeton (Drury) (Scudder 1889). There is no observable relation between population size and type of water present.

There appears to be a correlation between colony size and nectar abundance (x? = 3.2, df = 1, P = 0.07). Only two sites have large populations with low nectar availability, but these populations are mar- ginally “large” (sites 7 & 9, Table 1). Total amount of nectar is also important in Ewphydryas editha (Boisduval), influencing its population

VOLUME 42, NUMBER 1

39

TABLE 1. Characteristics of 15 sites occupied by Euphydryas gillettii. Loni Heol Nectar Site Colony abun- availa- Nearby trees (age of largest Water no. size! ance? bility to nearest 5 yr) (stream width) Disturbance 1 >380 >380 High Lodgepole pine (75) Stream (<1 m) Fire® Engelmann spruce (65) 2 >380 >80 High Quaking aspen (60) Streams (<1 m) None; meadow Subalpine fir (75) edge 3 7 10 Low Engelmann spruce (150) Stream (1-3 m) None; meadow edge 4 2 10 Low Lodgepole pine (55) Marshy Fire; wet soil 5 4 20 Low Lodgepole pine (90) Stream (<1 m) Fire; logging Quaking aspen (65) 6 >830 20 High Subalpine fir (155) Streams (<1 m) _ Fire’; logging Lodgepole pine (15) 7 18 10 Low Cottonwood (40) Stream (1-3 m) Beaver activity Lodgepole pine Co eal 10 High Lodgepole pine (65) Stream (>5m) Flooding 9 22 20 Low Lodgepole pine (95) Marshy, Wet soil Engelmann spruce (70) stream (<1 m) 10 8 >80 High Lodgepole pine (55) Stream (<1 m) Fire?; meadow Engelmann spruce (50) edge Subalpine fir (40) die fl 5 Low. Subalpine fir (95) Marshy Fire® _ Engelmann spruce 12 3 >80 Low Lodgepole pine Stream (1-3 m) Flooding; fire? 1s if 20 Low Engelmann spruce (195) Stream (1-3 m) Fire® Lodgepole pine (40) 14 2 20 Low Willow (no trees) Marshy, Wet soil; graz- stream (<1 m) ing 15 >80 5 High Lodgepole pine (75) Marshy None; meadow

edge

‘ Total number eggs and adults. 2 Approximate number Lonicera clumps in 30 x 30 m quadrat. 8 Charred tree trunks.

dynamics (Murphy et al. 1983). Nectar is supplied by a number of genera (Table 2), mostly commonly Aster, Senecio, and Agoseris, but each occurs conspicuously at no more than 9 of the 15 sites. Williams et al. (1984) found the butterflies to switch nectar sources readily when an early source senesces. Total amount of nectar thus appears more important than particular sources.

Hostplants were considered highly abundant when there were more than 15 distinct shrubs or clumps. In contrast to nectar availability, hostplant abundance does not correlate directly with population size (x2 = 0.1, df = 1, P > 0.5). Reasons are considered later.

Most sites have been disturbed (Table 1), with fire being the com- monest natural source. Lodgepole pine, Pinus contorta Dougl., is com-

40 JOURNAL OF THE LEPIDOPTERISTS SOCIETY

TABLE 2. Common nectar sources for Euphydryas gillettii at 15 study sites.

Genus Number of sites where present Genus Number of sites where present Aster 9 Polygonum 2 Senecio 8 Antennaria It Agoseris 7 Chrysanthemum IL Geranium 6 Cirsium ] Achillea 5 Geum 1 Heracleum 5 Helianthella il Potentilla 4 Saxifraga 1 Valeriana 8 Solidago 1

mon near colonies (Table 1), indicating common disturbance history in these areas (Pfister et al. 1977). Whatever the cause, disturbance opens a site for growth by more hostplants and nectar sources. The few sites not clearly showing disturbance are on edges of permanent wet mead- ows of grasses and sedges.

At higher latitudes, occupied sites occur at lower elevations (Fig. 1, r? = 0.49, P < 0.005). This result likely reflects colder climates and reduced height of mountains at higher latitudes. Furthermore, impor- tance of a minimum growing season length is shown in frequent south- erly exposure of sites at higher latitudes, in contrast to the variable exposure of sites at lower latitudes (Fig. 1). All large northern popu- lations occupy sites with southern exposure, while southern sites show no observable relation between population size and exposure. Williams (1981) demonstrated the importance of within-habitat exposure effects; current results suggest larger-scale influences as well.

Range

Available records of E. gillettii are mapped in Fig. 2. Sightings are concentrated in the mountainous regions of W Wyoming, central Idaho, NW Montana, and SW Alberta. Some regions for which there are only older records, such as Yellowstone National Park and SW Montana, undoubtedly support populations, but their inaccessibility makes col- lecting sporadic. Continued existence of E. gillettii in extreme SW Wyoming is questionable because extensive search has failed to uncover specimens (C. F. Gillette pers. comm.). A reported record from central Montana may be erroneous. There is also a single museum specimen from Ontario, but improbable date as well as location suggest misla- beling.

Sites in Alberta have smaller populations of butterflies than do those farther south, and all northern sites have one characteristic in common: Lonicera involucrata does not reach the large size and luxuriant growth characteristic of Wyoming and Montana sites. In moist areas at higher

VOLUME 42, NUMBER | 4]

ELEVATION (1000 FT- M)

42 43 44 45 46 47 48 49 50

LATITUDE (°)

Fic. 1. Elevation and latitude of fifteen Euphydryas gillettii sites. Large circles represent “large” populations. Arrows pointing down indicate sites with southerly ex- posure; those pointing right, easterly exposure; etc. Absence of arrow indicates site has no obvious slope.

latitudes, willows (Salix spp.) are often taller than L. involucrata, shad- ing them and making them less accessible to searching females; this rarely occurs at lower latitudes. Oviposition sites are therefore scarcer than at lower latitudes, because oviposition occurs on the highest leaves of hostplants that are fully exposed to sunlight (Williams 1981, Williams et al. 1984).

DISCUSSION

There appear to be four reasons for lack of correlation between population size and abundance of L. involucrata. First, and most im- portantly, this plant grows in moist areas regardless of amount of sun- light, while the butterfly requires sunlit hostplants (Williams 1981). In fact, the most luxuriant hostplants often grow in shade of conifers, but are not used as oviposition sites. Second, an extension of the first, much L. involucrata is over-shaded by willows at high latitudes, thus provid- ing fewer potential oviposition sites in such areas. Third, some Euphy-

JOURNAL OF THE LEPIDOPTERISTS SOCIETY

42

ome we ——— = —

ALBERTA

| | | Pcie pat ee | whoa U

en E>

SG

de eee

wy ee

:

/

)

“AN \

Fic. 2. Range of Euphydryas gillettii. Closed circles are sites described in this study; open circles are locations of populations believed extinct; closed triangles are locations

VOLUME 42, NUMBER 1 43

dryas gillettii populations are mostly biennial (Williams et al. 1984), and so may fluctuate greatly in abundance from year to year. While most E. gillettii sites are characterized by abundant Lonicera involu- crata, these three factors limit the size of an observed butterfly popu- lation to less than might be expected given the total amount of Lonicera. The fourth reason is butterfly use of alternative hostplants.

Only at one site was the colony larger than would seem possible given the amount of nearby L. involucrata. That population (site 15, Table 1) lives where L. involucrata is uncommon, and the butterflies oviposit extensively on two other plants, Pedicularis and another Lonicera (in prep.). There are several possible reasons for dietary expansion in but- terflies (Singer 1971, 1983); but whatever they may be for this popu- lation, other study populations have not followed suit, even though all known alternative hostplants grow throughout the Euphydryas gillettii range. Except for site 1, where an alternative hostplant was chosen at low frequency (less than 4% of egg masses, Williams & Bowers 1987), I did not find eggs on or see ovipositional behavior near other plants at the other 14 sites. Because of the known use of alternative hostplants, I expect other E. gillettii populations use alternative hostplants as well. The relation between population size and Lonicera involucrata abun- dance is thus weaker than has been widely accepted.

Because its hostplants and nectar sources require wet sites, and be- cause adults and larvae require sunlit areas for warmth, Euphydryas gillettii most often occurs in open montane meadows. The one study population that is not montane occupies a permanently wet, grazed seepage area in the transition zone. Several populations were observed along forested edges of seemingly permanent montane meadows; such meadows may change little through time because of allelopathic in- teractions of meadow vegetation or soil instability. More commonly, open sites are created temporarily through disturbance. The most fre- quent disturbance is fire, and most study sites have clearly been affected by it. Other forms of disturbance, such as flooding, beaver activity, or human activities like grazing and logging, also serve to open forested areas.

Vegetational succession in disturbed areas leads to changes that make sites less suitable through time. In particular, encroachment by sur- rounding forest leads to greater evapotranspiration, producing a drier site and thereby limiting growth of hostplants and nectar sources. Fur- thermore, invasion by trees reduces the sunlight that reaches the shrub

—

where E. gillettii has been seen since 1960; open triangles are records before 1960; question mark denotes uncertain record.

44 JOURNAL OF THE LEPIDOPTERISTS SOCIETY

and herb layer, thus eliminating warmer microsites preferred for ovi- position (Williams 1981).

_ Life in disturbed sites suggests that E. gillettii populations are subject to periodic extinction like E. editha (Singer & Ehrlich 1979), and such appears to be the case. I identified with precision one site where E. gillettii was collected in the 1960’s, but by 1983 vegetational succession had taken place, most remaining Lonicera involucrata was shaded, and no sign of butterflies could be found. Furthermore, human development of recreational areas has led to loss of additional populations, one known and one suspected.

Habitat requirements of E. gillettii, including moisture for hostplants and nectar, and sunlight for larvae and ovipositing females, produce the limits of its geographic distribution. Thus, plains east of the Rockies and arid basins westward form effective biogeographic barriers to dis- persal in either direction because of lack of water. Holdren and Ehrlich (1981) have shown that another arid region, the Red Desert of S Wy- oming, is the southern barrier since they successfully transplanted in- dividuals across the barrier to central Colorado where one colony has survived since 1977. Their transplant locales are similar to natural habitats farther north in being wet and having an abundance of nectar and Lonicera involucrata.

The northern range limit has been assumed to result from lower temperatures and shorter growing season. However, all the Alberta sites have much smaller L. involucrata, and willows dominate northern wet sites by growing taller than other shrubs. All populations of the butterfly at higher latitudes are smaller as well. Although no northern populations have been found to use hostplants other than L. involucrata, alternative hostplants used elsewhere also decline in abundance at higher latitudes. It seems likely that competition by willows reduces size and perhaps density of potential hostplants. Thus, fewer oviposition sites and poorer (more shaded) ones would be found during normal hostplant searching by females (Williams et al. 1984). I suggest that loss of oviposition sites contributes, along with shorter growing season, to the northern limit.

Euphydryas gillettii is uacommon, but there is no evidence that its range has been changing in recent decades. The greatest conservation advantage this species has compared to other uncommon species is that its habitat lies largely in mountainous areas that are not readily acces- sible and in which there is little immediate potential for human mod- ification. Its greatest conservation disadvantage is its occurrence through a limited range in discrete, localized populations, which are individually susceptible to disturbance and extinction.

VOLUME 42, NUMBER 1 45

ACKNOWLEDGMENTS

For collection records, detailed directions, maps, and correspondence, I acknowledge contributions from the following: Karolis Bagdonas, P. J. Conway, N. S. Curtis, G. R. DeFoliart, C. J. Durden, J. D. Eff, P. R. Ehrlich, C. D. Ferris, C. F. Gillette, L. P. Grey, R. L. Hardesty, C. E. Holdren, F. E. Holley, W. E. Knoshaug, S. J. Kohler, N. G. Kondla, D. D. Lawrie, J. A. Legge Jr., S. O. Mattoon, Adam Peters, F. W. Preston, J. D. Preston, John Reichel, J. R. Slotten, R. E. Stanford, Kenneth Tidwell, W. H. Wagner Jr., and W. D. Winter Jr. For access to collections, I thank F. H. Rindge, American Museum of Natural History; J. D. LaFontaine, Canadian National Collection; Chen Young, Carnegie Museum of Natural History; Richard Hoebeke, Cornell University; Leo Marnell, Glacier National Park; R. K. Robbins, National Museum of Natural History; J. P. Donahue, Natural History Museum of Los Angeles County; and C. A. Triplehorn, Ohio State University. I thank T. L. Marsh, N. E. Stamp, and two anonymous reviewers for comments on a manuscript draft, and my wife Sharon for assistance and companionship while searching in many out-of-the-way locations. This research was funded by grants from Hamilton College and the Brachman-Hoffman Foundation.

LITERATURE CITED

ComsTOckK, J. A. 1940. Notes on the early stages of Euphydryas gillettii Barnes. Bull. S. Calif. Acad. Sci. 39:111-118.

FERRIS, C. D. & F. M. BROWN. 1981. Butterflies of the Rocky Mountain States. Univ. Oklahoma Press, Norman. 442 pp.

Hitcucock, C. L. & A. CRONQUIST. 1973. Flora of the Pacific Northwest. Univ. Washington Press, Seattle. 730 pp.

HOLDREN, C. E. & P. R. EHRLICH. 1981. Long range dispersal in checkerspot butterflies: Transplant experiments with Euphydryas gillettii. Oecologia 50:125-129.

Murpny, D. D., A. E. LAUNER & P. R. EHRLICH. 1983. The role of adult feeding in egg production and population dynamics of the checkerspot butterfly Euphydryas editha. Oecologia 56:257-263.

PFISTER, R. D., B. L. KOVALCHIK, S. F. ARNO & R. C. PResBy. 1977. Forest habitat types of Montana. U.S. Dep. Agric. Forest Service, Gen. Tech. Rep. INT-34. 178 pp.

SCUDDER, S. H. 1889. The butterflies of the eastern United States and Canada. Vol. I. Introduction, Nymphalidae. Publ. by author, Cambridge, Massachusetts. 766 pp.

SINGER, M. C. 1971. Evolution of food plant preference in the butterfly Euphydryas editha. Evolution 25:383-389.

1983. Determinants of multiple host use by a phytophagous insect population. Evolution 37:389-—408.

SINGER, M. C. & P. R. EHRLICH. 1979. Population dynamics of the checkerspot butterfly Euphydryas editha. Fortschr. Zool. 25:53-60.

WILLIAMS, E. H. 198]. Thermal influences on oviposition in the montane butterfly Euphydryas gillettii. Oecologia 50:342-346.

WILLIAMS, E. H. & M. D. Bowers. 1987. Factors affecting host-plant use by the montane butterfly Euphydryas gillettii (Nymphalidae). Am. Mid]. Nat. 118:153-161.

WILLIAMS, E. H., C. E. HOLDREN & P. R. EHRLICH. 1984. The life history and ecology of Euphydryas gillettii Barnes (Nymphalidae). J. Lepid. Soc. 38:1-12.

Received for publication 27 June 1986; accepted 6 November 1987.

Journal of the Lepidopterists’ Society 42(1), 1988, 46-56

BIOLOGY OF POLYGONIA PROGNE NIGROZEPHYRUS AND RELATED TAXA (NYMPHALIDAE)

JAMES A. SCOTT 60 Estes Street, Lakewood, Colorado 80226

ABSTRACT. The life history of Polygonia progne nigrozephyrus is compared with that of P. gracilis zephyrus, P. faunus hylas, and P. satyrus in Colorado. Adult predator deterrent behaviors occur: adults resemble leaves as they rest on twigs showing leaflike undersides, roost with forewings drawn forward with antennae resting between them, and feign death when handled. Larvae also have predator-avoidance strategies: scoli presumably act as a physical deterrence, small larvae can drop using a silk thread, a ventral neck gland possibly repels predators, larvae vomit on an attacker, older larvae resemble twigs as they rest in a three-dimensional twisted-S shape, pupae resemble a dried curled leaf or short twig. Larval host plants differ between species, with some overlap. Identification features for the four species are presented for each stage. Despite adult similarity of P. progne nigrozephyrus and P. gracilis zephyrus, P. g. zephyrus larvae most resemble those of P. faunus.

Additional key words: Polygonia gracilis, P. faunus, P. satyrus, predator deterrence, chaetotaxy.

Scott (1984) described P. progne nigrozephyrus which occurs in Colorado-S Wyoming—Utah-SE Idaho-NE Nevada. It resembles P. gracilis zephyrus (Edw.) on the upperside, P. p. progne (Cram.) on the underside and in male abdominal structure, and was long confused with zephyrus. Polygonia p. nigrozephyrus is certainly the same species as oreas (Edw.), but some may question whether it and oreas belong to P. progne. Early stages of nigrozephyrus are similar to those of oreas and progne, and are distinct from zephyrus and other Polygonia; wing undersides and abdominal structures resemble those of progne. There- fore, nigrozephyrus does seem to be a subspecies of progne.

Since 1984, minor differences between populations of nigrozephyrus in Colorado have been found. Adults from the E slope of the continental divide in the Front Range usually have the dorsal hindwing darker because the submarginal spots are the same size as those of P. g. zephyrus, whereas adults from the W slope usually have the spots larger like those of P. satyrus (Edw.). However, the difference is not great enough to warrant a new name for W slope populations, and some adults from each area resemble those from the other. The Front Range populations may have slightly darker dorsal hindwings because of occasional im- migration of subspecies progne, which has a very dark dorsal hindwing margin.

An additional difference not mentioned by Scott (1984) between some P. p. progne adults and other Polygonia, first noticed by W. H. Edwards, involves one of the dark stripes in the ventral forewing discal cell: in most P. p. progne the anterior stripe is unbroken, whereas in some of

VOLUME 42, NUMBER 1 47

them and in other subspecies and species the stripe is broken into two parts.

For oviposition and larval rearing, cut host-plant sprigs were put into water-filled vials, cotton-plugged so the water would not drain when vials were on their sides. For older larvae, large host branches were cut and placed in wet sand.

Adult Stage

Adults bask with wings spread (dorsal basking). In the laboratory, some nigrozephyrus females closed the wings above the thorax and vibrated them rapidly (up to 2 mm apart at the tips) when lights were turned on in the morning; this is shivering behavior to raise the thorax temperature prior to flight.

Adults of nigrozephyrus, zephyrus, and faunus (Edw.), as well as Nymphalis milberti (God.), roost on twigs with wings closed, forewings drawn far forward (nearly out of hindwings) and covering the head and antennae which rest between the forewings. This posture perfects the resemblance to a leaf on the twig by elongating the “leaf”, breaking up its margin, and hiding antennae to avoid predation during fall, winter, and spring. Adults frequently feign death when handled, which would also signal a predator that the butterfly is a dead leaf.

There is evidently a circadian rhythm of oviposition, because females laid eggs in the laboratory only during daytime, and even when lights remained lit females begar. roosting in late afternoon. For obtaining oviposition, fluorescent bulbs were superior to incandescent bulbs, prob- ably because the former produce a greater and more natural amount of ultraviolet light.

Immature Stages

Host plants. Polygonia progne nigrozephyrus feeds on gooseberry: Ribes inerme Rydb., in Delta and Douglas counties, Colorado, R. lep- tanthum Gray at Wiiliams Canyon, El Paso Co., Colorado. In the laboratory, nigrozephyrus larvae accepted leaves of Ribes inerme, but refused wax currant, R. cereum Dougl., and ate very little golden currant, R. aureum Pursh. They ate only leaves. Additional host records for P. p. progne, based on preserved larvae in the Smithsonian, are gooseberry (Si. Albans, West Virginia, Monticello, New York) and cur- rant (Centreville, Rhode Island).

Polygonia gracilis zephyrus usually eats Ribes cereum in Colorado. However, I found an egg on R. inerme at Tinytown, Jefferson Co., on 2 June 1984, and reared it to a mature larva; and a larva under a R. inerme leaf 5 km W Idledale, Jefferson Co., on 12 June 1984, which | reared to an adult. In the laboratory, zephyrus larvae eat R. cereum,

48 JOURNAL OF THE LEPIDOPTERISTS SOCIETY

and do not move off its leaves to eat adjacent R. inerme leaves, although larvae will also accept inerme. Additional host plants of zephyrus are Ribes sanguineum (Jones 1951), and Rhododendron occidentale (larvae reared to adults, Big Trees Park, Calaveras Co., California, 4 June 1983, J. F. Emmel & S. O. Mattoon).

Polygonia faunus hylas (Edw.) usually eats Salix, but I found three first-stage larvae and five eggshells under leaves of Ribes inerme at Tinytown on 26 May 1984, and an adult emerged 20 June. In the laboratory, fawnus larvae refused Ribes aureum leaves, but ate R. inerme and preferred it to R. cereum.

Thus all three Polygonia will eat Ribes inerme occasionally.

The only known Colorado host of P. satyrus is Urtica dioica gracilis (Ait.) Sel., though Humulus lupulus L. is eaten elsewhere. In the lab- oratory, satyrus larvae accepted Humulus and Urtica leaves equally well.

Life Cycle

Five larval instars have the following approximate head widths, re- spectively: 0.4, 0.7, 1.2, 1.7, 2.6 mm. Stage | is easily recognized by its black head without scoli; stage 2 has head scoli but is still black; stage 3 has head scoli but is black usually with an ochre pattern tending toward the pattern of stages 4-5. Usual laboratory durations of nigro- zephyrus stages at 19°C were: egg, 5-6 days; larval stages, 3, 2.5, 2, 2, 4 days, respectively; and pupa, 9-10 days; totalling 27-30 days. In the cooler and more variable temperatures of nature, these periods are probably nearly doubled, so that adults should appear by late July— early August, although eggs laid in late April might produce the few fresh late-June adults known in nature. A faunus stage 1 larva found 26 May emerged as an adult 20 June in the laboratory, even though faunus emerges in nature only in late July and August. The laboratory life cycle of P. p. progne is 31-32 days (Edwards 1880), of P. inter- rogationis 28-40 days (Edwards 1882b), and of P. comma 27-88 days (Edwards 1882a). Thus all Polygonia have similar developmental rates indoors, and all have five larval stages. However, in Colorado P. faunus and P. progne nigrozephyrus have only one generation per year, while P. satyrus and P. gracilis zephyrus have two generations at low altitude and one at high altitude; and P. interrogationis has two or three gen- erations.

Predator-Avoidance Structures and Behavior

Stinkbugs and ants were found on R. inerme host plants and may prey on immatures. The scoli of stage 2-5 larvae presumably physically deter predators.

VOLUME 42, NUMBER 1 49

They slightly hurt the human skin when touched, evidently a physical puncturing rather than an urticating chemical.

A ventral neck gland occurs on stage 2-5 larvae of all 4 Polygonia species; it contains 2 internal transverse dark secretory pads which perhaps produce repellent chemicals.

When grasped, the larva often bends its head arcund and vomits green fluid onto the attacker.

Fourth- and fifth-stage larvae of nigrozephyrus grasp a twig with the prolegs, bend the front part of the body right or left, and raise the end of the abdomen. This “corkscrew” posture may make the larva resemble a dead leaf or twisted twig, perhaps lessening predation by birds. This posture also occurs in ssp. progne (Edwards 1880) and in satyrus (C. F. Gillette pers. comm.).

Young larvae of all four species rest on the underside of a leaf, and when older may also rest on a twig. Only older larvae of P. satyrus, also P. comma, live in a nest. It is made by chewing the base of the leaf on each side, thus making it droop, and silking Urtica leaf edges down and together below the enclosed larva, which rests on the leaf underside.

Disturbed young larvae can extrude a silk thread as they fall, then crawl up the thread to return to the plant.

Pupae are constricted in the middle where silver spots also visually break up the outline, making the pupa resemble a dead, shriveled leaf or twig.

Gooseberry hosts are armed with sharp spines which act as physical protection against vertebrates. A punctured pupa recovered completely.

Descriptions of Early Stages

Colors are based on live individuals. Immatures have been deposited in the Smithsonian Institution. Many dozen individuals of Polygonia p. nigrozephyrus were reared from eggs laid by females from NE of Cedaredge, Delta Co., and Nighthawk, Douglas Co. Each stage is described, and is followed by comparisons with the other three Polygonia species and subspecies, each of which were represented by less than 10 individuals. Segments are named T1 for prothorax, A8 for abdominal segment 3, etc. (Fig. 3). Scoli are named with the letter B followed by name of nearest primary seta. They are not preceded by S because of confusion with primary seta SD1, etc.; sp is spiracle; VNG is ventral neck gland on older larvae. Names of setae are from Hinton (1946) and Scott (1986), with slight modifications (Scott 1988) that improve homology and make head and body setal nomenclature different to avoid confusion.

Egg. Green, averaging 8.6 vertical ribs (Table 1), each rib steep-walled, increasing in height to maximum at top, then disappearing; 40-50 horizontal ribs forming ladder between adjacent vertical ribs; the day before hatching turning blackish with transparent silvery-reflecting shell as larva becomes partly visible.

Comparison. All Polygonia eggs green. Polygonia p. progne has 8-9 ribs, P. g. zephyrus averages 9.8, other Polygonia average 10.4-11.5 (Table 1).

First-stage larva (Figs. 1, 4, 5, 9, 11, 12, 16). Head black without pattern or horns Body dark brown with long black setae, bumplike bases of which are chitin brown; with

50 JOURNAL OF THE LEPIDOPTERISTS SOCIETY

TABLE 1. Number of vertical ribs on eggs.

Taxon Mean SD Range N Source Polygonia p. nigrozephyrus 8.6 0.55 8-10 40 _ this paper P. p. progne — — 8-9 — Edwards (1880) P. g. zephyrus 9.8 0.59 9-11 27 this paper P. satyrus 10.4 0.54 10-12 43 this paper P. faunus 10.5 0.63 10-12 31 __ this paper P. interrogationis — — 8-10 — Edwards (1882b),

Pyle (1981)

P. comma (Colo.) IS 0.50 11-12 58 _ this paper P. comma (Minn.) 10.6 0.54 10-12 388 __ this paper Nymphalis vau-album (D. & S.) 11.0 — 9-12 — CF Gillette

pers. comm.

cream spots (Fig. 1) as follows: front half of Tl cream except for small supralateral brown patch on some larvae; Tl cream in front of, behind, just beside black prothoracic shield; rest of Tl brown except for 2 cream dashes extending rearward above, below spiracle. T2-3 brown, large yellow-cream patch around D2, smaller cream patch around L1-2. A2, A4, A6 brown, with 4 pale patches: broad cream mid-dorsal V aimed posteriorly on anterior part of each segment; broad yellow-cream patch below D1; narrow supralateral light brown dash; long cream sublateral dash. Al, A3, A5, A7 brown, with 4 light brown patches on each side corresponding to pale patches on A2, A4, A6; sublateral dash cream on A3, A5, A7, A8. A8 same as A7 but 3 upper patches slightly creamier. A9 brown, subdorsal cream patch twice as long vertically as horizontally. A10 brown, suranal plate black, proleg cream, proleg plate brown, large circular cream supralateral patch.

Comparison. Other Polygonia larvae very similar, with black hornless head and similar body pattern. Polygonia satyrus same as nigrozephyrus, pale bumps cream-white, a few creamy sublateral dashes. Polygonia g. zephyrus same as nigrozephyrus, except pale bumps cream-white instead of yellow-white, seta D1 on T3 on whiter bump as is seta D2, no supralateral brown patch on front of Tl though it appears on some second-stage larvae so may be individual trait, supralateral dash on A2, A4, A6 cream, Al, A8, A5, A7, A8 all brown except for lateral cream dash. Polygonia interrogationis similar (Ed- wards 1882b), but P. comma “whitish-green” (Edwards 1882a). Polygonia faunus larvae differ from all other Polygonia in having white areas expanded away from bumps: for instance, white patch on T2, T3 includes both D1, D2 setae; on A2, A4, A6 white V lengthened anteriorly, subdorsal white patches below D1 extend posteriorly.

Second-stage larva (Figs. 2, 6-8, 17). Head black with 2 short black spiny horns (BPA2 scoli) each with 1 long seta on tip, 5 setae on crown just below, no setae on long stalks; bases of PA1, AG3, LH1, O2 pale, membranous; very narrow short pale line along middorsal groove. Body reddish brown, brownish orange toward rear, similar to 1st stage in pattern, prothorax mostly orangish yellow; orange V’s on top of A2, A4, A6, yellow- cream areas of first stage now orange, scoli present with bases orangish. Scoli BD2 on A2, A4, A6 ochre on some larvae, mostly brown on most, other scoli black. BD2 scoli on T2, T3, A2, A4, A6 rest on large orange bumps making segments conspicuously paler, other scoli rest on small orangish bumps. Body has weak cream mid-dorsal, subdorsal spots which help form abdominal V’s; remaining segments have thin wavy lateral cream line between BL1 scoli, thin wavy supralateral cream line between BSD1 scoli. Tiny pale subdorsal transverse dashes present. Ventral neck gland present.

Comparison. Polygonia g. zephyrus has slightly shorter horns, body undergoes less color change from first stage: color pattern the same, pale patches still white, though BD2 on A2, A4, A6 yellow-cream, in some larvae blackish, making segments still paler on top, other scoli black. Only BD2 on T2-8, A2, A4, A6 rest on yellow-cream bumps; other scoli rest on small whitish bumps. Tiny cream transverse dashes occur behind, before BD2 on A2, A4, A6 to help form V’s as in P. satyrus; middorsal, subdorsal, supralateral, lateral

VOLUME 42, NUMBER 1 51

D1 D2

XD1 aN FI RST a STAGE vt ih DL

iJ

S)

I

ACVAS TEA TAST AG Ar “AG AS AiO

Fics. 1-8. Setal maps of Polygonia progne nigrozephyrus larvae. 1, First stage. Color pattern shown on some segments, except that plates at base of setae, including prothoracic shield and suranal plate, are dark brown; T2 and T8 patterns similar; Al, A8, A5, A7 patterns similar; A2, A4, A6, A8 patterns similar except that A8 darker; 2, Second stage; 3, Fifth (mature) stage. L inside circle is true leg; P inside circle is proleg; S inside circle is scolus. Hundreds of small setae not shown. See text for further explanation.

white spots present. Polygonia satyrus resembles |st-stage zephyrus, thus head black, T1 mostly white except for black prothoracic shield. BD2 on A2, A4, A6 also yellow-cream; other scoli black, except BD1 on A6 whitish, BD1 on A4 partly whitish, BL1 on A4, A6, A7, A8 mostly white. BD2 on T2, T3, A2, A4, A6 rest on large yellow-cream bumps; other scoli rest on small tan hills, though BD1 on A2, A4, A6, BL1’s rest on fairly white bumps. Polygonia faunus has enlarged white areas compared to other species, on at least 1 larva BD1 and BD2 on A2, A4, A6 pale. Ventral neck gland occurs in all 3 Polygonia.

Third-stage larva (Figs. 8, 13, 18). Head black with black scoli, following structures ochre: Mid-dorsal notch, adfrontal cleavage line (lateral to frontoclypeus), lower fronto- clypeus, head just above antennae, bases of all major setae except black horns; but some individuals have head mostly black, nearly devoid of pattern. Head setae AG3, PA], LH1, O2 on long ochre stalks. Body dark brown, with long mostly orange scoli: BL1, BSD1 mostly black; BD2 mostly orange; BD2 on T3, A2, A4, A6 strongly orange; scoli on T2, A10 mostly black. Body pattern similar to stages 4—5.

Comparison. Polygonia g. zephyrus larvae have BD1, BD2 more whitish cream on abdomen. Polygonia satyrus differs greatly: head black with cream notch on top running

52 JOURNAL OF THE LEPIDOPTERISTS SOCIETY

iA OSS

Fics. 4-15. Leg and cranial setae of Polygonia progne nigrozephyrus. 4, Ventral- medial view of first-stage larval thoracic leg, showing setae typical of butterflies; 5, Setae and olfactory pores of first-stage larval head; 6, Head of second-stage larva. Head horn derived from, or incorporates, seta PA2. Setae PA], AG8, LH1, O2 arise from small cones on transparent circles of exoskeleton. X’s show positions of setae present on some larvae; 7, Head horn of second-stage larva, includes PA2 seta of first stage; 8, Ventral neck gland of larval stages 2-5 partly everted. It appears slitlike when retracted, is fully everted in some preserved larvae; 9, Mandible of first-stage larva; 10, Head setae of fifth-stage (mature) larva with primary setae whose origin is traceable to first-stage larval seta lettered; 11, Labrum of first-stage larval head, anterior view showing one olfactory pore; 12, Labrum of first-stage larval head, posterior view showing two olfactory pores, three spatulate setae; 13, Labrum of third-stage larval head, anterior view; 14, Labrum of fourth-stage larval head, anterior view; 15, Labrum of fifth-stage (mature) larval head, anterior view.

forward to inverted cream V on face, head horns, setae mostly black, some setae on sides and lower face white; body has lateral cream band with cream BL1; top of body cream with cream scoli, black dashes in shape of V without point angling forward from each BD1.

Fourth-stage larva (Figs. 8, 14). Head as in mature larva. Body similar to mature larva, but scoli more orangish, BD1, BD2 on A8, A5, A7 with dark brown ring around each above base, whereas other scoli and all scoli on mature larva, lack brown ring.

Comparison. The other species also resemble mature larva.

Mature larva (Figs. 3, 8, 10, 15, 19, 20, color photo on pl. 3 of Scott 1986). Head black, horns dark brown, orangish cream notch on top, orange-red W on front consisting of streak along upper part of each adfrontal cleavage line plus streak angling down from base of each horn, lower 3rd of frontoclypeus orange-brown, orange-red patch surrounding eye cluster, orangish mottling beside neck. Some setae everywhere on head including AG3, PA1, LH1, O2 orange-red, on long orange stalks; AG2, some dorsal setae beside neck, about 3 lateral setae beside neck on smaller orange stalks. Body scoli ochre, only needle tips orange, except: BD2, BSD1 on T2 black with some orange branches; BD2 on T3 mostly black, orangish on basal 5th, BSD1 ochre; BD2 on A8 partly black, BSD1 mostly black, BD1, BL1 ochre; BD2 on AQ partly black; BD2 on A10 black. Body blackish brown in ground color, with complex pattern. Tl brown with mid-dorsal, subdorsal, supraspiracular, subspiracular orangish lines, some small mostly orange spinelike setae; mid-dorsal ochre band extending from head to T1, narrowing on T2, very narrow on TS. A few ochre transverse dorsal lines between T1, T2, between A8-10. Body joints between

VOLUME 42, NUMBER 1 53

T2, A8 have 5 ochre joint lines, line 2 grayish, lines 1, 3 widest, separated by 4 black joint lines, most posterior very narrow. Segments T2, T3 ochre on top, with paired short black grooves on either side of black mid-dorsal line. Segments Al, A2 similar but paired dark grooves form brown transverse streak behind BD1. A1, especially A2, begin to show dorsal black rearward-aimed V’s characteristic of all Polygonia on A3-8. Tip of V blunt, wide, corresponding to brown transverse streak on Al-—2 just behind BD1, each arm of V thickest in middle anterodorsal to BD2 where V becomes orangish black, outlined by ochre bands as thick as V itself. Three more black spots posterior to point of each V that continue point: black transverse mid-dorsal dash formed by 2 interruptions in 1st black joint line circling segment, narrower dash formed by narrower interruptions in next joint line, mid-dorsal black triangular spot on anterior edge of posterior segment. Ochre joint lines stop at 2 wavy lateral lines characteristic of all Polygonia. Upper wavy lateral line orange, on each segment obliquely extending from BSD1, which is ochre with orange base, up, forward then down; behind BSD1 obliquely extending down, backward then down, forward, resembling orange staple aimed down, forward, centered on BSD1. Upper line interrupted between segments by last 3 ochre joint lines which splinter into about 5 ochre wavy narrow lines that stop just above lower wavy lateral line. Lower wavy lateral line ochre, extending from each BL1 obliquely up, forward, then straight forward, then angling down toward BL1 of preceding segment. Beneath this line a vague ochre line above prolegs. Prolegs, underside blackish brown, ochre ventral bands running along abdomen on each side of mid-ventral line. Ventral neck gland present.

Of more than 50 larvae, a few slightly paler (dorsal areas yellow anteriorly, cream behind). Early 5th stage slightly more pinkish violet as orange-red scoli of 4th stage change to ochre.

Comparison. California P. p. oreas, based on preserved larvae, photos, same as ni- grozephyrus, except that top front of former oranger, yellowish orange vs. orangish yellow on top of thorax, Al-2; BSD1 on orange upper wavy lateral band more orangish than nigrozephyrus, ochre in latter with only base orangish. Based on 50-year-old preserved larvae in Smithsonian, ssp. progne similar to nigrozephyrus in structures, all pattern elements seem present, though impossibie to discern true colors; dorsal V-marks, transverse lines between segments present. Edwards (1880) described progne color as buff (ochre), dorsal area “reddish” (probably orangish ochre) around black V’s; he described T2-3, A9-10 scoli as black, others ochre as in nigrozephyrus; described BSD1 as black, but contradicted on the preserved larvae, these being pale also. T1 collar described as yellow in progne, and is pale in the preserved larvae, whereas it is black except for mid-dorsal line in nigrozephyrus, other 3 species. Head seems to have larger black areas in nigro- zephyrus than ssp. progne. Evidently ssp. progne larva does not change color from front to rear as much as western subspecies, and dark brown areas of former are smaller.

Mature larvae of other Polygonia species differ greatly. All 3 have black V's on top of abdomen slightly narrower than nigrozephyrus, point of each V less strongly connected. All 3 have wavy lower lateral lines as in nigrozephyrus, but these are slightly reddish cream in zephyrus, red-orange in faunus, orangish cream in satyrus. Polygonia g. zephyrus (Fig. 27, color photo on pl. 2 of Scott 1986) much more 2-toned, top of segments T2-3, Al-2 red-orange, especially T3, A2); top of A3-8 whitish, especially A4, A6 which are yellowish white. Basic pattern elements of zephyrus same as in nigrozephyrus, but wavy lateral lines weak, slightly reddish, scoli black except BL1 along lower wavy line whitish in some larvae, BD1 orange within orange areas, white within white areas. Head of zephyrus also mostly black, except for white mid-dorsal notch, sometimes thin orange inverted V on front, scattered small white seta bases. Some zephyrus larvae have T3, A2 orangest on top, A4, A6 whitest, whereas in others T2-3, Al-2 equally red-orange on top, A3-8 equally white on top. Latter characteristic of P. fawnus, which has top of body orange in front, white behind as in photo 14 of faunus (=silvius) in Pyle (1981). Polygonia faunus has both wavy lateral lines red-orange, BL1 on lower line white, head black with some cream setae, cream dorsal notch, orange W on front. Thus mature larvae of P. g. zephyrus, P. faunus are similar. Polygonia c-album L. larvae resemble faunus closely (photos in Pyle 1981, Whalley 1979:19, Brooks & Knight 1985:79).

Polygonia satyrus mature larvae differ greatly from other Polygonia (Fig. 25, color

54 JOURNAL OF THE LEPIDOPTERISTS SOCIETY

photo on pl. 2 of Scott 1986). Top near-uniform yellow, same pattern elements present: head black, inverted cream V on front, mid-dorsal cream notch, some small cream setae; middorsal line cream on thorax, transverse rings between segments, dorsal V’s present. However, entire top of body greenish yellow, T2-3, Al-2 ochre-yellow in some larvae, lower wavy line thick, pale yellow, orangish between segments in some larvae, yellow BL1’s; upper wavy line nearly absent, with black BSD1’s or line thin, orange, with cream BSD1’s in some larvae.

Polygonia interrogationis (Fab.) mature larvae are also very different from other Polygonia (Pyle 1981:photo 15, Edwards 1882b). Polygonia comma (Harr.) mature larvae vary (Edwards 1882a).

Ventral neck gland present in stages 2-5 of P. g. zephyrus, P. faunus, P. satyrus.

Chaetotaxy (Figs. 1-7, 9-15). Head of lst-stage larva has only primary setae. Second- stage head has many secondary setae, scoli BPA2; each horn incorporates seta PA2 of Ist stage because scolus in same position as lst-stage PA2, other dorsal primary setae rec- ognizable on 2nd-stage head by large size, position. Each 2nd-stage horn has long PA2 seta on tip, 5 setae on crown below tip. Head setae, horns on stages 3-5 like those of stage 2 except for proliferation of small setae, primary setae recognizable on mature larval head by larger size, horn still including only 1 primary seta, PA2. Setae on labrum constant at 6 on each side, 3 spatulate setae on posterior oral surface, during larval stages 1-5, setae on other mouthparts also constant, except mandible setae which rise from 2 on each stage 1 mandible to about 10 on stage 5.

Proprioceptor setae, those that detect cuticular folds telescoping over, on head, body same as present in other Lepidoptera.

Body of Ist-stage larvae has mainly primary setae, also some secondary L2 setae, present on all individuals examined, on T3, Al—8; on A3-6 of some larvae 4th L seta present near L2. On 2nd-stage body, many secondary setae, scoli appear. Body scoli of 2nd-stage larva not homologous with Ist-stage larval primary setae because primary setae of Ist stage occur on 2nd stage, sometimes slightly moved in position, with scoli. Thus A10 of lst stage has paler spot where BSD1 appears on 2nd stage, yet both stages have same dorsal primary setae on A10; on T2-3, 2nd stage retains same SD, L setae of stage 1 adds BLI1; on A8, 2nd stage retains same D1-2, SD1 setae of stage 1, adds BD1, BD2. Body scoli add small setae between stages 2-5, otherwise change little. Small SD plate on T2- 3 of stage 2 disappears, only 1 or 2 setae remain on stage 5. Body setae multiply between stages, hundreds of which are not shown on stage 5 setal map (Fig. 3). Crochets typical of butterflies: 14 of anterior 8 prolegs forming circle in stage 1, medial crescent in mature larvae; 12 anal crochets form anteromedial crescent in all stages. Each true leg has 5, 2, 6, 2 tactile setae plus 3, 1, 0, 2 proprioceptor setae on Ist 4 leg segments of stage 1, the usual number in lst-stage butterflies, additional setae joining these on mature larvae, Ist segment having about 8 setae, for instance, on mature larvae. No anal comb present on any stage.

Comparison. Setae, scoli of all larval stages same in 4 Polygonia compared, also in mature P. interrogationis larvae based on preserved specimens, Petersen (1965) showing drawing of mature interrogationis larva: thus secondary Ist-stage L2 seta occurs in all species, L1 on A3-6 in some zephyrus individuals splitting into 3 instead of 2 setae, making 4 L’s instead of the normal 8, head horn on stages 2-5 incorporating primary seta PA2, consisting of 1 terminal setae, crown of 5 main setae below. Secondary L2 seta on |st-stage T3-A8 distinguishes Polygonia from Nymphalis, Vanessa.

Pupa (Figs. 21-24, color photo on pl. 5 of Scott 1986). Usually pinkish tan, sometimes paler, rarely blackish gray. Segments T3, Al, A2 have silver or gold subdorsal spot, usually silver on T3, Al, often gold on A2 because of reddish tan A2 top, making 6 in all, mid- dorsal silver streak sometimes on Al. Segment A2, to lesser extent A3, reddish tan on top. Four abdominal bands: lateral tan-edged brown band, mid-ventral tan-edged brown band, mid-dorsal brown-edged tan line. Basal half of each tibia brown. Sliver of hindwing just above forewing brown. Light-brown V’s on A4-7, weakly on A8, on both sides of tan mid-dorsal line, 1 arm of each V ending at each subdorsal cone. Broad brown, often greenish brown, band crosses wing from tornus to mid-costa, short brown subapical band parallel to it. Many cones, bumps usually at larval scoli positions: very small mid-dorsal

VOLUME 42, NUMBER 1 ap

Fics. 16-28. Polygonia larvae and pupae. 16-24 P. progne nigrozephyrus from Delta Co., Colorado; 25-28 other taxa as noted from Jefferson Co., Colorado. 16, First-stage larva, dorsal view; 17, Second-stage larva, dorsolateral view; 18, Third-stage larva, dorsal view; 19, Fifth-stage larva, dorsal view; 20, Fifth-stage larva, lateral view; 21, Pupa, dorsal view; 22, Pupa, lateral view; 23, 24, Pupae, lateral views showing variation; 25, P. satyrus mature larva, lateral view; 26, P. satyrus pupa, lateral view; 27, P. gracilis zephyrus mature larva, dorsal view; 28, P. faunus hylas pupa, lateral view.

bump on A2-8; large subdorsal cone on T2-3, Al—8; supralateral bump on A3-7; lateral bump on A4-8, lateral bump on each head horn; large bump on wing base; bump on lower basal corner of wing; subventral bump on A5-6, another on head, 1 on each tibia; 2 stout cones (horns) projecting forward from each side of head; mid-dorsal keel on T2.

Silk pad spun by pupating larva bright pink.

Comparison. All Polygonia pupae have similar silver or gold spots in saddle, similar cones, keels, horns, dark bands on abdomen, wings. Species differ in overall color, shape, size of cones, horns. Polygonia p. oreas resembles nigrozephyrus, but 2 oreas pupae seen were brown, not pinkish tan. Polygonia p. progne pupa (Edwards 1880) also pinkish brown like nigrozephyrus, with similar markings; head, thorax sometimes greenish brown. Polygonia g. zephyrus like nigrozephyrus in shape, but most individuals light brown, some creamy gray or tinged with green, rarely blackish gray, abdomen more mottled, subdorsal area on A4 lighter than on other segments, on A5—A7 a paler streak angling forward, down from each subdorsal cone. Few zephyrus pupae resemble nigrozephyrus in overall color, yet reddish tan top of A2 of nigrozephyrus identifies most. P. faunus pupa (Fig. 28, color photo 14 of Pyle 1981, as silvius) light brown (often with reddish

56 JOURNAL OF THE LEPIDOPTERISTS SOCIETY

flush on top of A2-8 as in nigrozephyrus) or dark gray, easily identified by elongate shape, >10% longer, long head horns, twice as long as other Polygonia. P. satyrus pupa (Fig. 26, color photo on pl. 5 of Scott 1986) paler, tan or straw, sometimes yellowish dorsally, rarely brown all over, easily identified by mid-dorsal T2 keel being twice as high as other species, subdorsal abdomen cones about twice as large. P. interrogationis similar in color to some nigrozephyrus, faunus, with similar sized bumps, but its T2 keel very large (color photo 15 of Pyle 1981, Edwards 1882b). P. comma pupa quite variable (Edwards 1882a). Polygonia c-album pupa brown, resembling nigrozephyrus in shape but T2 keel larger as in satyrus (Brooks & Knight 1985:79).

Oddly, silk cremaster pad spun by pupating larvae colored differently in other species: bright pink in nigrozephyrus, also interrogationis (photo in Pyle 1981); pale pink in zephyrus, faunus; yellowish white, rarely faintly pink, in satyrus.

ACKNOWLEDGMENTS

I thank J. R. Heitzman for larvae of P. interrogationis, R. K. Robbins for loans of Polygonia immatures from the Smithsonian Institution, P. A. Opler for photos of P. progne oreas immatures, and C. F. Gillette for reviewing the manuscript.

LITERATURE CITED

Brooks, M. & C. KNIGHT. 1985. A complete pocket guide to British butterflies. Jonathan Cape, London. 159 pp.

EDWARDS, W. H. 1880. Description of the preparatory stages of Grapta progne. Can. Entomol. 12:9-16.

1882a. Description of the preparatory stages of Grapta comma. Can. Entomol.

14:189-196.

1882b. Description of the preparatory stages of Grapta interrogationis. Can. Entomol. 14:200-208.

HINTON, H. E. 1946. On the homology and nomenclature of the setae of lepidopterous larvae, with some notes on the phylogeny of the Lepidoptera. Trans. Roy. Entomol. Soc. London 97:1-37.

JONEs, L. 1951. An annotated checklist of the Macrolepidoptera of British Columbia. Entomol. Soc. British Columbia Occ. Pap. 1:1-148.

PETERSEN, A. 1965. Larvae of insects. Part I. Lepidoptera and plant-infesting Hyme- noptera. Publ. by author, Columbus, Ohio. 315 pp.

PyLE, R. M. 1981. The Audubon Society field guide to North American butterflies. Knopf, New York. 916 pp.

ScCoTT, J. A. 1984. A review of Polygonia progne (oreas) and P. gracilis (zephyrus) (Nymphalidae), including a new subspecies from the southern Rocky Mountains. J. Res. Lepid. 23:197-210.

1986. The butterflies of North America. A natural history and field guide.

Stanford Univ. Press, Stanford, California. 583 pp.

1988. The small forest: Chaetotaxy of first-stage butterfly larvae. In preparation.

WHALLEY, P. 1979. Butterflies. Hamlyn nature guides. Hamlyn Pub. Group, London. 128 pp.

Received for publication 22 April 1987; accepted 2 October 1987.

Journal of the Lepidopterists’ Society 42(1), 1988, 57-58

GENERAL NOTE

DIFFERING OVIPOSITION AND LARVAL FEEDING STRATEGIES IN TWO COLOTIS BUTTERFLIES SHARING THE SAME FOOD PLANT

Additional key words: Pieridae, Colotis amatus, C. vestalis, eggs, Salvadoraceae.

There is much interest in the habit of certain butterfly species laying eggs in clusters. It is generally agreed that cluster-laying is a derived trait, the ancestral butterfly having laid single eggs. Cluster-laying has evolved independently several times in all butterfly families. Its significance has been subject to a variety of interpretations. The purpose of this paper is to present oviposition data for two closely related species of Colotis in New Delhi, India.

The species in question are Colotis amatus F., whose geographic distribution covers most of Africa, Arabia, India, and Sri Lanka; and C. vestalis Butler, found in NW India, Pakistan, and East Africa, but unaccountably absent from Arabia (Larsen, T. B. 1983, Fauna of Saudi Arabia 5:333-478). Together with C. phisadia Godart, C. amatus and C. vestalis form a small section of the genus that feed on Salvadoraceae rather than on the more usual Capparidaceae.

In New Delhi both butterflies feed on Salvadora persica L. and S. oleoides Decaisne. Usually both are phenologically synchronous, and occur on the same trees or bushes. In size and behavior they are very similar and were not the ground colours salmon and white, respectively, they would be difficult to tell apart on the wing. M. A. Wynther- Blyth (1957, Butterflies of the Indian Region, Bombay Natural History Society, Bombay, 523 pp.) even suggests they interbreed, interspecific copula having been observed.

Given the overall similarity, the difference in oviposition behavior is startling. Colotis amatus lays clusters averaging ca. 30 eggs on upper surfaces of fresh leaves at outer extremities of the host plant (Table 1). Eggs are evenly spaced within each clutch. Colotis vestalis lays single eggs deep inside the host plant, usually on a twig or a branch, rarely on an old leaf. I observed eggs being laid as far as 90 cm from the nearest leaf, a considerable distance for a small, freshly hatched larva to travel. Larvae of C. amatus feed gregariously on fresh foliage, but group cohesion weakens in final instars. Those of C. vestalis feed singly on old leaves, usually deep inside the bush or tree. I never found both species on the same leaf.

The egg of C. vestalis is chalk white with 20-22 keels extending from the micropyle to the base. It is covered in fine hairs, best visible when the egg is submerged in fluid. Egg volume appeared 15-20% greater than that of C. amatus. The latter’s eggs are yellow, have only 14-16 keels, lack hairs, and unlike those of C. vestalis are covered with a sticky substance. Midges and mosquitoes were often found trapped on egg clutches.

S. Courtney (1984, Am. Nat. 123:276-281) mentions that Aporia crataegi L. in Morocco may adjust egg-clutch size to food plant quality. The data are given in more detail by S. Courtney (1986, Adv. Ecol. Res. 15:51-131). Colotis amatus clutch-size on the broad- leaved Salvadora persica averaged 28.7 eggs (n = 106), and on the narrow-leaved S. oleoides, 22.7 (n = 17) in my Delhi sample; the difference is not statistically significant.

Although these two common butterflies are synchronous and share foodplants, they seem to be noncompetitive. I never saw complete defoliation of food plants. There are a number of potential pathways for two such butterflies to evolve different ovipositing strategies, but data to support any specific hypothesis are not available. Probably no single causal factor underlies all egg clustering. However, available data do not support the hypothesis of R. A. Fisher (1930, The genetical theory of natural selection, Clarendon Press, Oxford, 272 pp.) that egg clustering leads to aposematism; if anything C. vestalis, which feeds on old leaves, should be the more aposematic of the two. I masticated a number of specimens without finding the least pungency or emetic response, although I found other aposematic butterflies emetic (Larsen, T. B. 1983, Entomol. Ree. J. Var. 95: 66-67).

The closest parallel I have seen to the two Colotis species is that of Eurema hecabe L. and E. blanda Boisduval in Papua New Guinea and S India. The former lays single eggs,

58 JOURNAL OF THE LEPIDOPTERISTS SOCIETY

TABLE 1. Number of eggs in 123 clutches laid in the wild by Colotis amatus in New Delhi, India (autumn 1986).

No. eggs in clutch No. clutches No. eggs in clutch No. clutches

1-5 0 41-45 3) 6-10 2 46-50 3 11-15 10 51-55 0 16-20 24 56-60 2 21-25 25 61-65 il 26-30 19 66-70 1 31-35 17 71-75 1 36-40 13 76+ 0

Average 27.9 eggs per clutch.

the latter clutches. However, in both places the two show more ecological and spatial segregation than Colotis; they can feed on the same plants but usually do not do so in the same locality. In Yemen I noticed that Capparidaceae-feeding Colotis tend towards local food plant specialisation.

The Urtica feeding members of the Vanessini in the Palaearctic fall into two groups. Vanessa lay single eggs, Aglais lay clutches. Members of both genera are often found on the same batch of nettles, but as in Colotis complete defoliation is rare.

This paper was prepared under a general research grant for 1987 kindly provided by the Carlsberg Foundation of Denmark.

TORBEN B. LARSEN, Snoghoj alle 29C, DK 2770 Kastrup, Denmark.

Received for publication 23 july 1987; accepted 26 October 1987.

Journal of the Lepidopterists’ Society 42(1), 1988, 59

BOOK REVIEW

THE LIVES OF BUTTERFLIES, by Matthew M. Douglas. 1986. xv + 241 pp. 16 pp. color photographs. University of Michigan Press, Ann Arbor, Michigan, U.S.A. Hard cover. $45.00.

This attractive book is the product of a scientist and teacher whose enthusiasm is contagious. Its strengths include substantive explanations of many aspects of work on butterfly biology, its discussion of experimental and other evidence for scientific conclu- sions, and its emphasis on scientific literature. The book is rich in clear, often detailed, explanations of work in several major areas: anatomy, development, and evolution of morphological features of life stages; biophysical, physiological, and ecological constraints on life stages and community structure; behavioral, biochemical and ecological aspects of speciation and coevolution with plants. This exposition is accompanied by many black- and-white diagrams (often from published original drawings or photographs), a section of color photographs illustrating activities and morphological characteristics of life stages, a glossary, several appendices, and a useful index. This combination makes the book an engaging, accessible, self-contained store of information.

In addition, the author enhances the book’s informational content in two ways. First, he places specific examples in a conceptual context by discussing considerations that underlie specific hypotheses. Explanations of how observations and experimental data are collected contribute to a clear sense of how scientific questions are raised and examined, and why “answers may be open to alternative interpretations. This aspect of the book reflects the author’s experience as a university professor; many of his discussions would make good lecture notes for an advanced undergraduate course. This bold focus on processes of scientific research may be the book’s most important contribution to its educational goals. Second, the book’s emphasis on recent research literature provides a resource for further study.

The question of readership presents problems for the book. While ostensibly written for an audience that has some background in biology, its attempt to appeal to both lay and professional audiences sometimes creates disconcerting inconsistencies. For example, the author describes “‘sex-limited mimicry’ as a special case of Batesian mimicry in which one sex mimics unpalatable models; he includes a definition of this term in the glossary. Ten pages later, however, “sex-limited”’ is used colloquially to describe distribution of a trait whose pattern of inheritance is sex-linked. This colloquial use of a term that has specific meaning in genetics is confusing. Similarly, the author emphasizes his personal research experience in a way likely to engage the interest of lay readers. To a professional readership, however, such emphasis is likely to seem egotistical and annoying.

This book thus attempts the dual challenges of engaging and educating a lay readership as well as concisely reviewing recent literature for a professional audience. This is a rarely attempted goal, and the author presents us with a unique solution. The book’s value to its potential professional audience lies in its conciseness and timely review of much recent literature. Its appeal to this audience is uncertain, because professors whose students study these research topics in class may assign the original literature rather than this book. However, the author’s contribution to explicating butterfly biology and scientific research for a lay audience is a noteworthy success.

F. S. CHEW, Department of Biology, Tufts University, Medford, Massachusetts 02155.

Journal of the Lepidopterists’ Society 42(1), 1988, 60-61

OBITUARY

ABNER ALEXANDER TOWERS (1916-1987): A Tribute

The Lepidopterists’ Society lost one of its charter members with the passing of Abner A. Towers, who was well-known to collectors in the Southeast and to participants in the first of the collecting expeditions to Ecuador organized by Thomas C. Emmel and Gio- vanna Holbrook.

Abner Alexander Towers

Born in Gadsden, Alabama, 28 January 1916, Abner Towers developed an interest in wildlife as a boy, particularly his lifelong fascination with butterflies and moths, birds, and other flying creatures. He grew up in Gadsden, completing his primary schooling there, then attended the Kent School, in Kent, Connecticut, during which time he began to collect and study Lepidoptera seriously. At the age of eighteen, in 1934, he took his first trip to the Florida Keys expressly to observe and collect butterflies and moths. He attended the Massachusetts Institute of Technology as a general science major, and, after earning the Bachelor of Science in 1939, served as an officer in the U.S. Army Corps of Engineers. He spent most of World War II in the Aleutian and Philippine Islands. Following the war he settled in Georgia, the state he would call home for the rest of his life. Abner Towers married, raised a family, and built a career as an engineer and chemist, and was often described in both capacities by co-workers and peers as “brilliant.” In August 1972, he cofounded A-Jay Chemical Company, in Powder Springs, Georgia, an industrial chemical firm he continued to administer until his terminal illness.

In the 1950’s Abner resumed his study of the Lepidoptera of the region, focusing his attention almost entirely on the butterflies of Georgia and Florida, and he steadily built an impressive collection containing substantial series of virtually all the species recorded from the two states. He established a strong friendship with Lucien Harris Jr., and his contributions to Harris's The Butterflies of Georgia (University of Oklahoma Press, 1972) were significant, and included numerous state records and field observations. Abner’s

VOLUME 42, NUMBER 1 61

persistent and dedicated collecting subsequently added several species to the Georgia butterfly fauna, including Mitoura hesseli Rawson & Ziegler, and, in 1981, he participated in the discovery of a new geometrid, described as Narraga georgiana Covell, Finkelstein & Towers (J. Res. Lepid. 23:161-168, 1984). Occasionally, when his other responsibilities allowed, Abner traveled and collected outside the country; most notable were his collecting trips to Ecuador in 1980 and Jamaica in 1982.

Abner Towers died 18 March 1987, after a bravely fought three-year battle with leukemia. He is survived by his wife, Margaret Le Craw Towers, his children John A. Towers, Marsha Towers Endictor, and Andrea Towers Rohaly, and his sister Harriet Towers Bjelouvucic. His friends and co-workers remember him as “a man of warmth... who always took time to inquire of people’s families, discuss their hobbies, jobs or personal interests and give advice, if asked, with a sincerity derived from a love of people. He was totally unselfish with his time, his knowledge and his abilities.” (From a eulogy by Alan Shipp and Polly Buford.)

His collection was donated to the University of Florida in 1985 and deposited in the Florida State Collection of Arthropods, Gainesville. In addition to the Lepidopterists’ Society, Abner was a charter member of the Southern Lepidopterists, a group he served since its founding in 1978 as Georgia zone coordinator.

“Abner Towers was a gentle man, and a gentleman. He will be missed.” (Shipp and

Buford. )

IRVING L. FINKELSTEIN, 425 Springdale Drive N.E., Atlanta, Georgia 30305.

Date of Issue (Vol. 42, No. 1): 16 March 1988

iar bries wot’ TROL yal tae ae huts ene om

wil an ee

.

97 ae. Se >

get) vit geting ‘owt bch gE Soa

oe ee oe We Vier ni Ome tirpeeite bh buts Pic chek Cetera npc BP: | barat? PAL ghee yap ee o* een 'y pred abba’ tpi sy abel Ra + leila age oath

ays

n¥ A 2 : r ra ha 4) ; j ’ f , EE t 5 4) * ao 5 2 t : f, Ba ise , Pan Es ‘i : f 7 st wa 5 E Le i 7 a Sy ~~»! ;, a y) i ~ ; th

by ake | 2 a } re h { \ y 7 nl , ' il 4 it ae f i \ f ; | sf it Ni ‘ \ bad | H A f y A I } j ‘ i { i | i 7 vy IR, ron i 3 ‘ t i ae ih: ih a" iu +0 ‘ yt vil i i ’ t / Ll) t a a why , ? :

es,

PhS sai’

‘ f iA é f "+ " | ; ; “4 ne Ee : ~~, rs _ Lad i i :

EDITORIAL STAFF OF THE JOURNAL WILLIAM E. MILLER, Editor

Dept. of Entomology University of Minnesota St. Paul, Minnesota 55108 U.S.A.

Associate Editors: M. DEANE BOWERS, BOYCE A. DRUMMOND III, DOUGLAS C. FERGUSON, ROBERT C. LEDERHOUSE, THEODORE D. SARGENT, ROBERT K. ROBBINS

NOTICE TO CONTRIBUTORS

Contributions to the Journal may deal with any aspect of Lepidoptera study. Categories are Articles, General Notes, Technical Comments, Book Reviews, Obituaries, and Cover Illustrations. Journal submissions should be sent to the editor at the-above address. Short manuscripts concerning new state records, current events, and notices should be sent to the News, June Preston, Editor, 832 Sunset Drive, Lawrence, Kansas 66044 U.S.A. Journal contributors should prepare manuscripts according to the following instructions, and submit them flat, not folded.

Abstract: An informative abstract should precede the text of Articles.

Key Words: Up to five key words or terms not in the title should accompany Articles, General Notes, and Technical Comments.

Text: Manuscripts should be submitted in triplicate, and must be typewritten, entirely double-spaced, with wide margins, on one side only of white, letter-sized paper. Titles should be explicit, descriptive, and as short as possible. The first mention of a plant or animal in the text should include the full scientific name with author, and family. Measurements should be given in metric units; times in terms of the 24-hour clock (0930 h, not 9:30 AM). Underline only where italics are intended.

Literature Cited: References in the text of Articles should be given as Sheppard (1959) or (Sheppard 1959, 1961a, 1961b) and listed alphabetically under the heading LITERATURE CITED, in the following format without underlining:

SHEPPARD, P. M. 1959. Natural selection and heredity. 2nd ed. Hutchinson, London. 209 pp.

196la. Some contributions to population genetics resulting from the study of

the Lepidoptera. Adv. Genet. 10:165-216.

In General Notes and Technical Comments, references should be shortened and given entirely in the text as P. M. Sheppard (1961, Adv. Genet. 10:165-216) or (Sheppard, P. M., 1961, Sym. R. Entomol. Soc. London 1:23-30) without underlining.

Illustrations: Only half of symmetrical objects such as adults with wings spread should be illustrated, unless whole illustration is crucial. Photographs and drawings should be mounted on stiff, white backing, arranged in the desired format, allowing (with particular regard to lettering) for reduction to fit a Journal page. Illustrations larger than letter-size are not acceptable and should be reduced photographically to that size or smaller. The author's name and figure numbers as cited in the text should be printed on the back of each illustration. Figures, both line drawings and photographs, should be numbered con- secutively in Arabic numerals; “plate” should not be employed. Figure legends must be typewritten, double-spaced, on a separate sheet (not attached to illustrations), headed EXPLANATION OF FIGURES, with a separate paragraph devoted to each page of illustrations.

Tables: Tables should be numbered consecutively in Arabic numerals. Headings for tables should not be capitalized. Tabular material must be typed on separate sheets, and placed following the main text, with the approximate desired position indicated in the text. Vertical lines as well as vertical writing should be avoided.

Voucher specimens: When appropriate, manuscripts must name a public repository where specimens documenting identity of organisms can be found. Kinds of reports that require vouchering include life histories, host associations, immature morphology, and experimental enquiries.

Proofs: The edited manuscript and galley proofs will be mailed to the author for correction of printer’s errors. Excessive author’s changes at this time will be charged to authors at the rate of $2 per line. A purchase order for reprints will accompany proofs.

Page charges: For authors supported by institutions, page charges are $20 per Journal page. For unsupported authors, page charges are $10 per Journal page with a $50 maximum. Authors of Book Reviews and Obituaries are exempt from page charges. Authors unable to pay page charges for reasons such as difficulties with foreign exchange should apply to the editor for free publication. Four free publication slots are available annually in the Journal.

PRINTED BY THE ALLEN PRESS, INC., LAWRENCE, KANSAS 66044 U.S.A.

CONTENTS

SPEYERIA ATLANTIS IN COLORADO: REARING STUDIES CONCERNING THE RELATION BETWEEN SILVERED AND UNSILVERED FORMS. James A. Scott: 0.0 ee

POPULATION FLUCTUATIONS OF AZETA VERSICOLOR (FABRICIUS) (NOCTUIDAE) ON GLIRICIDIA SEPIUM (JACQ.) (FABACEAE) IN

NORTHEASTERN Costa Rica. Allen M. Young WW BUTTERFLIES OF NORTHEAST TENNESSEE. Charles N. Watson Jr. i> John A. Hyatt, 200

BODY WEIGHT AND WING LENGTH CHANGES IN MINNESOTA POPULATIONS OF THE MONARCH BUTTERFLY. William S. Herman .cc

HABITAT AND RANGE OF EUPHYDRYAS GILLETTI (NYMPHALIDAE). Ernest H, Williams uu ee

BIOLOGY OF POLYGONIA PROGNE NIGROZEPHYRUS AND RELATED TAXA (NYMPHALIDAE). James A. Scott 23a :

GENERAL NOTES

Glassberg, Lehman, and Pellmyr collections to the Smithsonian Institution. Robert K. Robbins <> Gary F. Hevel 2. ee

Differing oviposition and larval feeding strategies in two Colotis butterflies sharing the same food plant... Torben B. Larsen _.____.__._

Book REVIEW The Lives of Butterflies. °F.S. Chew i220) ee

OBITUARY Abner Alexander Towers (1916-1987): a tribute. Irving L. Finkelstein

46

31

57

59

Volume 42 1988 Number 2

ISSN 0024-0966

JOURNAL

of the

LEPIDOPTERISTS’ SOCIETY

Published quarterly by THE LEPIDOPTERISTS’ SOCIETY Publié par LA SOCIETE DES LEPIDOPTERISTES - Herausgegeben von DER GESELLSCHAFT DER LEPIDOPTEROLOGEN Publicado por LA SOCIEDAD DE LOS LEPIDOPTERISTAS

24 May 1988

THE LEPIDOPTERISTS’ SOCIETY

EXECUTIVE COUNCIL

JERRY A. POWELL, President JEAN-FRANCOIS LANDRY, Vice DouGLas C. FERGUSON, Immediate Past President

President ATUHIRO SIBATANI, Vice JACQUELINE Y. MILLER, Vice President President RICHARD A. ARNOLD, Secretary JAMES P. TUTTLE, Treasurer

Members at large:

MIRNA M. CASAGRANDE M. DEANE BOWERS JULIAN P. DONAHUE EDWARD C. KNUDSON RICHARD L. BROWN JOHN E. RAWLINS FREDERICK W. STEHR PAUL A. OPLER Jo BREWER

The object of the Lepidopterists’ Society, which was formed in May 1947 and for- mally constituted in December 1950, is “to promote the science of lepidopterology in all its branches, .... to issue a periodical and other publications on Lepidoptera, to fa- cilitate the exchange of specimens and ideas by both the professional worker and the amateur in the field; to secure cooperation in all measures’ directed towards these aims.

Membership in the Society is open to all persons interested in the study of Lepi- doptera. All members receive the Journal and the News of the Lepidopterists Society. Institutions may subscribe to the Journal but may not become members. Prospective members should send to the Treasurer full dues for the current year, together with their full name, address, and special lepidopterological interests. In alternate years a list of members of the Society is issued, with addresses and special interests. There are four numbers in each volume of the Journal, scheduled for February, May, August and November, and six numbers of the News each year.

Active members—annual dues $25.00 Student members—annual dues $15.00 Sustaining members—annual dues $35.00 Life members—single sum $350.00 Institutional subscriptions—annual $40.00

Send remittances, payable to The Lepidopterists’ Society, to: James P. Tuttle, Treasurer, 3838 Fernleigh Ave., Troy, Michigan 48083-5715, U.S.A.; and address changes to: Julian P. Donahue, Natural History Museum, 900 Exposition Blvd., Los Angeles, California 90007-4057 U.S.A. For information about the Society, contact: Richard A. Arnold, Sec- retary, 50 Cleaveland Rd., #3, Pleasant Hill, California 94523-3765, U.S.A.

To obtain:

Back issues of the Journal and News (write for availability and prices); The Com- memorative Volume ($10.00; $6.00 to members, postpaid); A Catalogue/Checklist of the Butterflies of America North of Mexico (clothbound $17.00, $10.00 to members; paperbound $8.50, $5.00 to members): order (make remittance payable to “The Lepi- dopterists’ Society”) from the Publications Coordinator, Ronald Leuschner, 1900 John St., Manhattan Beach, California 90266-2608, U.S.A.

Journal of the Lepidopterists’ Society (ISSN 0024-0966) is published quarterly for $40.00 (institutional subscription) and $25.00 (active member rate) by the Lepidopterists’ Society, % Los Angeles County Museum of Natural History, 900 Exposition Blvd., Los Angeles, California 90007-4057. Second-class postage paid at Los Angeles, California and additional mailing offices. POSTMASTER: Send address changes to the Lepidopterists’ Society, % Natural History Museum, 900 Exposition Blvd., Los Angeles, California 90007- 4057.

Cover illustration: Mature larva of Papilio polyxenes asterius Stoll on wild carrot, Daucus carota L. Submitted by John V. Calhoun.

JOURNAL OF

THe LEPIDOPTERISTS’ SOCIETY

Volume 42 1988 Number 2

Journal of the Lepidopterists’ Society 42(2), 1988, 63-98

IMPACT OF OUTDOOR LIGHTING ON MOTHS: AN ASSESSMENT

KENNETH D. FRANK 2508 Pine St., Philadelphia, Pennsylvania 19103

ABSTRACT. Outdoor lighting has sharply increased over the last four decades. Lep- idopterists have blamed it for causing declines in populations of moths. How outdoor lighting affects moths, however, has never been comprehensively assessed. The current study makes such an assessment on the basis of published literature. Outdoor lighting disturbs flight, navigation, vision, migration, dispersal, oviposition, mating, feeding and crypsis in some moths. In addition it may disturb circadian rhythms and photoperiodism. It exposes moths to inereased predation by birds, bats, spiders, and other predators. However, destruction of vast numbers of moths in light traps has not eradicated moth populations. Diverse species of moths have been found in illuminated urban environments, and extinctions due to electric lighting have not been documented. Outdoor lighting does not appear to affect flight or other activities of many moths, and counterbalancing eco- logical forces may reduce or negate those disturbances which do occur. Despite these observations outdoor lighting may influence some populations of moths. The result may be evolutionary modification of moth behavior, or disruption or elimination of moth populations. The impact of lighting may increase in the future as outdoor lighting expands into new areas and illuminates moth populations threatened by other disturbances. Re- ducing exposure to lighting may help protect moths in small, endangered habitats. Low- pressure sodium lamps are less likely than are other lamps to elicit flight-to-light behavior, and to shift circadian rhythms. They may be used to reduce adverse effects of lighting.

Additional key words: conservation, evolution, flight, urban ecology, light pollution.

Since the invention of the incandescent lamp over a hundred years ago, outdoor lighting has progressively increased. The growth has been characterized by expansion into new geographic areas, development of new lamps with new spectral characteristics, and increases in total amount of light and radiant energy (Riegel 1973, Hendry 1984, Sullivan 1984). Outdoor lighting has transformed the nocturnal face of the earth (Croft 1978). However, despite universal awareness that electric light disturbs behavior of nocturnal insects, the ecological impact of outdoor lighting has never been comprehensively assessed.

The possibility that outdoor lighting may adversely affect our fauna

64 JOURNAL OF THE LEPIDOPTERISTS SOCIETY

is well recognized. Lepidopterists have blamed outdoor lighting for declines in populations of North American moths, especially saturniids in the northeastern United States (Holland 1908, Ferguson 1971, Hessel 1976, Muller 1979, Worth & Muller 1979, Krivda 1980, Pyle et al. 1981). This view assumes a direct causal link between lamps and faunal change. Fundamental questions about such a link, however, have never been closely examined: What mechanisms might link lamps with changes in populations of moths? If lamps cause populations of moths to change, specifically what might the changes be? How important are effects of lighting compared to effects of other environmental disturbances? This study examines each of these questions. It investigates the hypothesis that outdoor lighting influences populations of moths.

The investigation is based on a review of literature. The presentation is organized into three sections. The first section describes distribution, growth, energy, and spectral composition of outdoor lighting. The sec- ond describes how lamps affect behavior, life functions and survival of individual moths. The third explores how such effects may disturb moth populations; it also discusses measures to reduce disturbances caused by lighting. Citations are deliberately extensive to facilitate retrieval of source material which is widely scattered among different disciplines.

LIGHTING

Nocturnal images of earth viewed from orbiting satellites show the distribution of outdoor lighting (Fig. 1). In the United States this dis- tribution coincides with that of the country’s population (Croft 1978). Nocturnal illumination is clustered around all large metropolitan areas, with greatest concentration in the Northeast corridor. Viewed from an airplane, nocturnal lighting delineates a web of interconnecting road- ways lined with illumination from houses, parking lots, billboards, and other landmarks. Such aerial observation suggests that lighting forms an illuminated web that envelops the nocturnal environment of Lepi- doptera. The web’s density varies with human population density, and its distribution is continental.

The magnitude of lighting in a major metropolitan area is illustrated by Philadelphia’s streetlighting (Table 1). Philadelphia has 100,000 high- pressure sodium streetlamps at a density of almost 300 lamps/km?. The energy they radiate equals more than 10 kilowatts/km?, an order of magnitude greater than the energy density of moonlight at full moon (Agee 1969). During the last 4 decades, lamp size (lumens) increased 7-fold, number of lamps tripled, and type of lamp changed from tung- sten filament and mercury to high-pressure sodium (Figs. 2 & 3) (Wain- wright 1961, C. A. Oerkvitz pers. comm.). Nationwide per capita con- sumption of electrical power for streetlighting is similar to that of

< «

‘60808 OpesojoD ‘leplnog ‘1aplnog ye ope10jor jo AjIss9AtUy ‘6PP XOG sndurery ‘1ayue_) vyeq V<o_ puke Moug [eUOT}eNY WIOIZ YdeAsOjOYg ‘(WIPIBOIG aj[[A}eG D1BO[O1Oa\ aj asuajod]) G86T [[eF UL [faves SuT}IG1O WLOIZ paMOTA se ‘s9}e}G pou) [BUINOOU Jo oseUII ayIsoduIOy)y “Tf “OL

ox 6c, ca = a =) a of —~ ea = a = | e) >

66 JOURNAL OF THE LEPIDOPTERISTS SOCIETY

TABLE 1. Streetlamps in Philadelphia, 1983. Total lamps, lumens, and demand (watts) from C. A. Oerkvitz (pers. comm.). Radiant energy calculated from GTE Products Corp. (Sylvania) (1977b). Demographic data from World Almanac (1986).

Number Streetlamp parameter Total Per capita Per km? Lamps 1.0 x 10° 3.8 X 107 2.8 x 10? Lumens 1.8 x 10° Le <Oe 5.0 x 10° Radiant energy (watts) emitted for wavelengths 350-700 nm 5.6 x 10° 3.3 1.6 x 10! Electric power demand (watts) 2.2 x 10° 3y x NO 6.1 x 10!

Philadelphia, and growth in lumens has been comparable or higher (Riegel 1973, Edison Electric Institute 1971, 1985, Sullivan 1984).

Conversion from mercury to high-pressure sodium lamps reduces radiant energy at the short-wavelength end of the spectrum. However, high-pressure sodium light is spectrally broad and does include radiant energy in the blue spectral region (Fig. 2B).

In contrast to high-pressure sodium light, low-pressure sodium light is spectrally narrow. It excludes practically all energy in the ultraviolet, blue, and green regions of the spectrum (Fig. 2A). Viewed through a spectroscope, its spectrum contains a bright yellow-orange line (actually 2 spectral lines very close together) near 589 nm. Because the human eye is particularly sensitive to light in the 589 nm region, low-pressure sodium lamps can provide bright illumination with comparatively little radiant energy (Finch 1978). Compared to other lamps used for outdoor lighting, low-pressure sodium lamps minimize environmental exposure to radiant energy both in number of wavelengths and number of watts. These lamps are used for streetlighting and other outdoor lighting, but much less frequently than are high-pressure sodium lamps.

Conversion of streetlamps from mercury to high-pressure sodium has changed the spectral distribution of outdoor lighting, but it has not changed it as much or as clearly as one might suppose. Mercury lamps, for example, are still used for residential and commercial lighting in Philadelphia, and for streetlighting in neighboring areas. Tungsten fil- ament (Fig. 3), low-pressure sodium, metal halide (Fig. 2C) and flu- orescent lamps (Sorcar 1982) all contribute to spectral diversity of out- door lighting in the city. While density and distribution of outdoor lighting have increased, spectral composition has diversified.

EFFECTS ON INDIVIDUAL MOTHS Vision Bright light can lower sensitivity of moth eyes 1000-fold (Bernhard & Ottoson 1960a, Hoglund & Struwe 1970, Agee 1972, 1973, Eguchi

VOLUME 42, NUMBER 2

Energy emitted (watts per 10 nm)

A

350 400 450 500 550 600 650 700 Wavelength (nm)

Electrical input = 135 watts Luminous output = 20,000 lumens Manufacturer: North American Philips Lighting Corp.

Low-pressure sodium

Energy emitted (watts per 10 nm)

Wavelength (nm)

Electrical input = 400 watts Luminous output = 34,000 lumens Manufacturer: GTE Products Corp. (Sylvania)

Metal halide

Ultraviolet Violet

350 400 500

Energy emitted (watts per 10 nm)

350 400 450 500 550 600 650

Wavelength (nm) Electrical input = 400 watts

Luminous output = 50,000 lumens Manufacturer: GTE Products Corp. (Sylvania)

High-pressure sodium

Energy emitted (watts per 10 nm)

D

350 400 450 500 550 600 Wavelength (nm)

Electrical input = 400 watts Luminous output = 20,500 lumens Manufacturer: GTE Products Corp. (Sylvania)

Clear mercury

550

Wavelength (nm)

ie 42:

67

700

Spectral energy distribution of vapor discharge lamps. Sources for A: Judd

1951, Finch 1978, Illuminating Engineering Society 1981, North American Philips Light- ing Corp. 1982. Sources for B, C, and D: GTE Products Corporation (Sylvania) 1977a,

1977b, 1979.

& Horikoshi 1984). Electroretinographic studies suggest what happens to the visual sensitivity of a moth that flies to a lamp. If the moth remains at the lamp and then flies away, full visual sensitivity may not return for 30 min or longer (Bernhard & Ottoson 1960a, 1960b, Agee 1972). This effect requires exposure to the lamp over a period of time, probably 10 min or longer (Day 1941, Héglund 1963, Yagi & Koyama

68 JOURNAL OF THE LEPIDOPTERISTS SOCIETY

Relative energy

100

Wavelength (nm)

Electrical input: 1000 watts Luminous output: 23,100 lumens Color temperature: 3030° Kelvin

Fic. 3. Spectral energy distribution of tungsten filament (“incandescent ’’) lamp. Sources: GTE Products Corporation 1972, 1974.

1963). A moth flying away from a lamp into relative darkness on a cloudy, moonless night may be functionally blind until enough time has elapsed for it to become fully dark-adapted.

Continuous exposure to bright electric lamps could in theory “dazzle” moths. This means it could stimulate the moth retina so intensely that the retina could not respond to additional increases in light. The result would be functional blindness so long as the moth remained exposed close to the lamp. Electroretinographic evidence, however, suggests that lamps do not dazzle moths (Eguchi & Horikoshi 1984).

Net effects of electric lighting on moth vision may vary according to local conditions as well as moth behavior. Urban lighting increases background illumination which in turn may help moths see. Electric lighting in some areas has increased nocturnal sky brightness as much as 20-fold (Hendry 1984). However, the spectral composition, polariza- tion and spatial distribution of outdoor lighting varies widely in different settings. In some locations they may differ so much from that of natural nocturnal light that they create visual artifacts and distortions. One

VOLUME 42, NUMBER 2 69

outcome of disturbed vision is flight to outdoor lamps, but many dis- turbances in visual function and behavior are possible.

The suggestion that urban lighting influences nocturnal vision of moths may appear paradoxical. Municipal light sources have shifted away from mercury lamps and toward high-pressure sodium lamps. One might suppose that moth retinas are insensitive to the relatively long wavelengths which characterize most of the energy contained in high-pressure sodium light (Fig. 2B). Moths, for example, do not fly to the 589 nm light of low-pressure sodium lamps (Fig. 2A), or do so rarely (Robinson 1952). Such a supposition, however, is incorrect: electroreti- nograms of moths consistently demonstrate sensitivity to light in the 589 nm region, and most studies have found maximum sensitivity in the green rather than ultraviolet part of the spectrum (Jahn & Crescitelli 1939, Héglund & Struwe 1970, Hsiao 1972, Mikkola 1972, Agee 1978, MacFarlane & Eaton 1978, Langer et al. 1979, Mitchell & Agee 1981, Eguchi et al. 1982). Retinal sensitivity extends farther into the long- wavelength end of the spectrum than flight-to-light behavior typically would suggest (Mikkola 1972, MacFarlane & Eaton 1973, Mitchell & Agee 1981).

Navigation

Diversion to lamps. Three hundred fifty-six species of Macrolepi- doptera, or about a third of those species found in all of Great Britain, were collected at a single light trap in England (Williams 1939). Com- parable findings have been reported in Britain and North America (Dirks 1937, Robinson & Robinson 1950a, Beebe 1958, Bretherton 1954. Moore 1955, Langmaid 1959, Hosny 1959, Holzman 1961, Moulding & Madenjian 1979). Tens of thousands of moths have flown to a single lamp in a single evening (Robinson & Robinson 1950a), and huge swarms of moths have aggregated around urban light sources (Howe 1959). On the other hand, some species of nocturnal moths rarely fly to lamps even though large populations of them may be flying nearby (Bretherton 1954, Taylor & Carter 1961, Janzen 1983). A variety of physiologic, behavioral and environmental factors may determine which species of moths fly to light and when (Geier 1960, Gehring & Madsen 19638, Milyanovskii 1975, Mazokhin-Porshnyakov 1975, Janzen 1983, 1984).

Large numbers of moths flying to lamps may give a false impression that lamps divert moths from great distances. Effective radius of a 125- watt mercury vapor light trap was initially reported to be 91 m, but later estimates reduced the figure to 17 m, and the most recent analysis cut the distance to 3 m (Robinson & Robinson 1950a, Robinson 1960, Baker & Sadovy 1978). Other studies have shown flight-to-light dis-

70 JOURNAL OF THE LEPIDOPTERISTS SOCIETY

tances of 10 m or less (Stanley 1932, Hamilton & Steiner 1939, Hartstack et al. 1971, Plaut 1971). Long-distance estimates ranging up to half a kilometer represent either extrapolation, artificial conditions or both (Graham et al. 1961, Hsiao 1972, Agee 1972, Stewart et al. 1969, Plaut 1971, Bowden & Morris 1975).

If the mechanism by which a lamp disturbs moths depends on di- version of flight paths to the lamp, then the moths disturbed must be limited to those flying in the geographic area immediately adjacent to the lamp. In this sense any direct effects of a particular lamp would tend to be local, except when topography (Beebe 1949, Beebe & Fleming 1951), foodplants, pheromones, or other factors concentrate moths near the lamp. Only in urban regions would density and distribution of lamps be great enough to influence large populations of moths over broad geographical areas.

Effects of electric lamps in urban areas, however, may be much smaller than one might expect. Robinson & Robinson (1950a) noted that lamps in isolated phone booths appear to be much more effective in eliciting flight-to-light behavior than are clusters of bright urban lamps located immediately adjacent to areas with large populations of moths. They demonstrated that lamps interfere with each other’s ca- pacity to elicit flight-to-light behavior, and the closer together the lamps, the greater the interference. The high density which characterizes dis- tribution of urban lamps suppresses flight-to-light behavior.

Urban lighting may suppress flight to light for a number of reasons. Light trap collections vary with the lunar cycle and are lowest at full moon (Williams et al. 1956, Agee et al. 1972, Nemec 1971, Dufay 1964, Bowden & Church 1978, Janzen 1983, Stradling et al. 1983). A similar correlation with moonlight cannot be demonstrated when nocturnal flight is measured by suction traps (Williams et al. 1956, Danthana- rayana 1986), pheromone-baited traps (Saario et al. 1970, Janzen 1984) or radar (Schaefer 1976). Moths active at dusk typically appear in suction traps before they appear in light traps (Taylor & Carter 1961). Eye pigment must be in a position of dark adaptation before moths will fly to light (Collins 1934), and even relatively dim background light can cause the pigment to move away from this position (Bernhard & Ottoson 1964). Diffuse urban light, like moonlight and twilight, reduces the darkness essential for flight-to-light behavior.

The moon not only increases background lighting but also constitutes a concentrated source of light by which insects may be able to orient (Sotthibandhu & Baker 1979). Moths flying by lunar navigation may bypass lamps (Baker & Sadovy 1978). Lamps may provide navigational cues which suppress flight to other lamps.

Light sources that emit large amounts of ultraviolet energy are gen-

VOLUME 42, NUMBER 2 7

erally most effective in eliciting flight-to-light behavior (Williams et al. 1955, Glick & Hollingsworth 1955, Klyuchko 1957, Deay et al. 1965, Mazokhin-Porshnyakov 1969, 1975, Mikkola 1972, Sargent 1976, Mitch- ell & Agee 1981). Conversion of mercury streetlamps to high-pressure sodium and metal halide streetlamps has undoubtedly tended to reduce flight to streetlamps. On the other hand, moths do fly to high-pressure sodium and metal halide lamps, and a small minority of species may fly preferentially to lamps with little or no ultraviolet emission (Klyuch- ko 1957, Mikkola 1972). Unlike high-pressure sodium lamps, however, low-pressure sodium lamps rarely elicit flight-to-light behavior (Rob- inson 1952).

In summary, increases in electric lighting do not necessarily impair nocturnal vision and navigation. Under some conditions they may im- prove moths’ nocturnal vision and suppress flight-to-light behavior.

Diversion away from lamps. Electric lamps may also divert moths away from them (Robinson & Robinson 1950a, Robinson 1952, Herms 1929, 19382, Nomura 1969, Nemec 1969, Hsiao 1972). These effects may depend in part on spectral output of the lamp (Mazokhin-Porsh- nyakov 1969, 1975, Nomura 1969). Several theories attempt to explain this behavior (Hsiao 1972), but none accounts for diversity of flight paths at lamps (Janzen 1984): while some moths make spiral or circular flights around lamps and land several meters away, others make a beeline straight to lamps and crash into them. Flight paths approaching lamps may zig-zag or be totally chaotic (Holzman 1961, Mazokhin- Porshnyakov 1969, Janzen 1984). Diversion away from lamps has been debated (Bretherton 1950, Robinson & Robinson 1950b). Evidence that moths avoid large illuminated areas (Herms 1929, 1932, Nomura 1969, Nemec 1969) is inconclusive, but this behavior is more difficult to demonstrate than flight to lamps.

Lamps suppress flight of moths that fly to them. Moths approaching lamps may land near them and remain quiescent for a moment or for the entire night. Lamps suppress flight of some species more than others (Blest 1963, Graham et al. 1964). In some cases lamps do not appear to suppress flight; in other cases they excite quiescent moths into flight (Collins 1984, Hsiao 1972). Diurnal moths occasionally fly at night to lamps (Engelhardt 1946, Janzen 1983), but here it is unclear whether the lamps help to initiate nocturnal flight.

Diversion and suppression of flight may impair orientation and nav- igation based on lunar, stellar or other visual celestial cues (Mazokhin- Porshnyakov 1969, Sotthibandhu & Baker 1979, Wehner 1984) includ- ing polarization of celestial light (Danthanarayana & Dashper 1986). It also may impair navigation and orientation based on geomagnetic, gravitational, barometric, aerodynamic, inertial, olfactory, acoustic or

12 JOURNAL OF THE LEPIDOPTERISTS SOCIETY

visual terrestrial cues (Baker & Kuenen 1982, Baker & Mather 1982, Janzen 1984, Schone 1984, Riley & Reynolds 1986). How much electric lighting disturbs use of particular cues may be expected to vary in part according to which cues the moth happens to be using at the moment it encounters the lamp.

Migration and Dispersal

Light sources divert moths engaged in migratory or dispersal flights (Cockerell 1914, Williams 1937, Beebe & Fleming 1951, Wolf et al. 1986). Urban lighting surrounds habitats isolated by urban sprawl, so that moths may have to traverse dozens of kilometers of densely illu- minated territory to arrive at potential breeding sites. Moths flying high (Glick 1965) may fly to urban light sources on tall buildings (Stanley 1932, Glick 1961). Because location of natural flyways is poorly doc- umented for North American moths, one cannot determine the extent urban lighting may intersect long-range natural migration routes here. In Venezuela, vast numbers of migrating moths aggregated around lamps near a narrow mountain pass which functions as a natural flyway (Beebe 1949, Beebe & Fleming 1951). Lighting along roads following topographical features such as valleys, rivers, and coastlines might se- lectively interfere with North American moth migrations (Fig. 1).

Oviposition

Electric lighting can disturb oviposition. Light-trap surveys have shown that the vast majority of females collected at lamps are gravid (Dirks 1937, Ficht et al. 1940, Glick & Hollingsworth 1954, Geier 1960, Gehring & Madsen 1963) although males usually outnumber them (Dirks 1937, Williams 1939, Sargent 1976, Worth & Muller 1979, Janzen 1984). Flight to light can shift oviposition to sites located near the lamp (Ficht et al. 1940, Martin & Houser 1941, Pfrimmer & Lukefahr 1955, Beaty et al. 1951, Nemec 1969, Brown 1984). Eggs may be deposited on lampposts, window screens, buildings, and other unsuitable sites near lamps. Egg densities may be several-fold higher on plants near lamps (Martin & Houser 1941). The result may be larval overcrowding and increased susceptibility to starvation, microbial infection, and preda- tion.

Lamps shift the distribution of oviposition sites toward them probably by diverting ovipositing females and not by stimulating oviposition. In cornfields, Ostrinia nubilalis (Hbn.) (Pyralidae) tends to oviposit near lamps (Ficht et al. 1940, Beaty et al. 1951), but in the laboratory nocturnal illumination suppresses O. nubilalis oviposition (Skopik & Takeda 1980). Similar observations have been reported in Pectinophora gossypiella (Saund.) (Gelechiidae) (Pfrimmer & Lukefahr 1955, Lu-

VOLUME 42, NUMBER 2 Fle

kefahr & Griffin 1957, Henneberry and Leal 1979). Outdoor lighting may decrease oviposition by Cydia pomonella (L.) (Tortricidae) and Heliothis spp. (Noctuidae), although the mechanism is unclear (Herms 1929, 1982, Nemec 1969).

Mating

Outdoor lighting does not prevent mating in certain Saturniidae: male Hyalophora cecropia (L.) and Samia cynthia (Drury) complete long-distance mating flights to virgin females at night across illuminated urban territory, and breed in urban habitats (Rau & Rau 1929, Pyle 1975, Sternburg et al. 1981, Waldbauer & Sternburg 1982). Most freshly emerged female saturniids do not fly at all until they have emitted pheromone and mated (Blest 1963, Nassig & Peigler 1984, Waldbauer & Sternburg 1979). Male sphingids and saturniids fly to virgin females before they fly to nearby electric lamps (Allen & Hodge 1955, Worth & Muller 1979, Janzen 1984). Almost all female Cydia pomonella col- lected at black lights have already mated (Gehring & Madsen 1963). Although more males than females typically fly to lamps, the capacity of males to mate with more than one female (Rau & Rau 1929, Allen & Hodge 1955, Lukefahr & Griffin 1957, Vail et al. 1968) may moderate the reproductive impact of disproportionate harm to males.

In contrast, electric lighting may have a major effect on mating in certain Noctuidae. Heliothis zea (Boddie) is an example. The peak time of night during which H. zea flies to light traps coincides with the period of copulation (Graham et al. 1964, Stewart et al. 1967). Only a third to a half of female H. zea collected at light sources have mated (Gentry et al. 1971, Vail et al. 1968). In the laboratory, H. zea will not mate unless its eyes are in a state of dark adaptation, as indicated by the presence of eye glow. Light intensity must be below 0.015 wW/ cm?, the intensity of light of a quarter-moon (Agee 1969). The suggestion is that H. zea females fly to light sources whose radiant energy suppresses mating.

A criticism of this scenario is that unmated H. zea females that fly to light may be migrating (Raulston et al. 1986) and therefore sexually immature (Johnson 1969). Female H. zea in the laboratory do not mate for 30-60 h after eclosion (Agee 1969). However, even if unmated females at lamps were sexually immature migrants, the lamps could disrupt reproductively important behavior, such as flight to locations where courtship and mating would be likely to occur. Furthermore, outdoor lighting may interfere with H. zea mating regardless of flight to light. Levels of light that suppress mating in the laboratory (Agee 1969) are well below ambient levels of light in electrically illuminated environments outdoors. Low levels of incandescent light (Nemec 1969)

74 JOURNAL OF THE LEPIDOPTERISTS SOCIETY

and moonlight (Nemec 1971) have influenced activities of Heliothis spp. in the field.

Other evidence suggests that lighting may interfere with mating. Unmated females of four other noctuid species fly to lamps (Vail et al. 1968). Male sphingids caught in light traps baited with virgin females do not seek out the females (Hoffman et al. 1966). In the laboratory, even dim electric light (0.3 lux) suppresses female Trichoplusia ni (Hbn.) (Noctuidae) pheromone release and male response to pheromone (Shorey & Gaston 1964, 1965, Sower et al. 1970). Electric light also suppresses female pheromone release and male response to pheromone in Dioryctria abietivorella (Grt.) (Pyralidae) (Fatzinger 1979). Mating by Pectinophora gossypiella requires a period of relative darkness last- ing at least 7 h (Lukefahr & Griffin 1957).

Feeding

Moths may feed in illuminated environments. Sphingids and noctuids visit food sources in full view of electric lamps located sometimes less than a few meters away, or they fly to electric light sources after they have completed feeding (Bretherton 1954, Milyanovskii 1975, Mazo- khin-Porshnyakov 1975, Janzen 1983, 1984). I have observed Buddleja (Gentianaceae) blossoms covered with noctuids at night (2300 h) vir- tually directly under a tungsten filament street lamp illuminating a heavily traveled road in Quisset, Massachusetts. Light from automobile headlamps and from a flashlight did not alter the moths’ activities.

Electric lamps, however, may interfere with feeding. Orchard illu- mination has reduced the number of Cydia pomonella feeding at bait (Herms 1932). In Japan, orchard illumination has been used to protect fruit from damage by fruit-piercing noctuids (Nomura 1969). Light has disturbed nectaring sphingids (Brown 1976). Diversion of moths away from light may explain why lamps interfere with feeding. Suppression in feeding is moot for the large number of moth adults that never feed (Norris 1936).

Electric lighting theoretically could injure larval foodplants. Sodium vapor lighting may harm plants by disrupting photoperiodic regulation of growth and development (Sinnadurai 1981, Cathey & Campbell 1975, Shropshire 1977), but such effects are apparently greater indoors in greenhouses than outdoors on the street (Andresen 1978).

Time Keeping Electric lighting can delay or advance vital activities of moths and their larvae, and these shifts could affect the insects as much as changes in the activities themselves (Beck 1980, Saunders 1982). This possibility has been the basis for proposals to exploit biological clocks for purposes

VOLUME 42, NUMBER 2 vis)

of pest control (Barker et al. 1964, Nelson 1967). In a field trial, however, light exposure failed to prevent diapause in larvae of Adoxophyes orana (F.R.) (Tortricidae) (Berlinger & Ankersmit 1976). The trial suggests that it is easier to manipulate biological clocks indoors than outdoors where temperature and other factors cannot be controlled.

Biological clocks of flying insects, however, may be much more sus- ceptible to outdoor electric lighting than those of larvae. This is because flight to light increases exposure to radiant energy. Exposure to a pulse of light lasting only 15 min is sufficient to attenuate a circadian rhythm in Drosophila; light 10° times more intense produces the same effect after only 10 sec; light 10° times more intense does it after an exposure of less than 0.1 sec (Chandrashekaran & Engelmann 1976). Energy for even a minute fraction of a second (photoflash) can disturb photope- riodic clocks in larvae of Lepidoptera (Barker et al. 1964). The an- thropomorphic observation that quiescent moths adjacent to a lamp are “asleep because they think it is daytime’? may be close to the truth.

Shifts in timing of nocturnal behavior of moths at lamps do not necessarily imply shifts in phase of endogenous rhythms. Changes in timing of behavior could represent other responses to light, or they could represent complex mixtures of responses. Regardless of these possibilities, magnitude and character of responses may vary according to when in the circadian cycle exposure to light occurs (Pittendrigh & Minis 1971, Skopik & Takeda 1980). Responses may also vary depending on spectral output of the lamp. For example, Pectinophora gossypiella has two light-sensitive clocks, only one of which responds to the 589 nm light emitted by low-pressure sodium lamps (Bruce & Minis 1969, Pittendrigh et al. 1970).

Theoretical Effects

To what extent nocturnal flight to light affects timing of nocturnal behavior has never been formally investigated. For example, if a moth flies to a light source, receives intense irradiation for 15 min, and flies away, how will its activities during the rest of the night be affected? If a male, will its mating period still coincide with that of females not exposed to light? If a female, will pheromone release still occur during the flight period of males? Shifts in mating times could cause sympatric, closely related species to attempt to mate with each other; such species normally do not mate with each other in part because their different mating periods keep them temporally segregated (Tuttle 1985).

Synchronization of activities with lunar rhythms may help moths navigate, mate, and avoid predators (Danthanarayana 1986). Lamps may disturb oviposition synchronized to lunar rhythms (Nemec 1969, 1971). To what extent moth activity synchronizes with lunar rhythms,

76 JOURNAL OF THE LEPIDOPTERISTS SOCIETY

and to what extent electric lighting may disturb such synchrony war- rants investigation.

Predation

Bats, birds, skunks, toads, and spiders hunt moths flying to lamps (Stanley 1932, Thaxter 1957, Holzman 1961, Krivda 1980, Covell 1985, Brower 1986). Lamps increase predation by clumping prey, and directly exposing them to attack (Turnbull 1964). Concentrated experience with particular species may help birds learn to defeat defenses based on surprise, novelty, or deceit (Blest 1957, Wickler 1968, Coppinger 1970, Sargent 1973b, Pietrewicz & Kamil 1979). Lamps also can destroy defensive behavior, such as that required for crypsis (Sargent & Keiper 1969, Sargent 1973a, 1976). The outcome is exemplified by a dark, bark-colored moth conspicuously resting on a white wall near a lamp at dawn. Lamps may help birds learn to recognize unpalatable species, but moths unpalatable to some birds may be acceptable to others (L6hrl 1979). Lamps may enable different birds to pick and choose among different possible prey. Because moths often land before they arrive at lamps, lamps may provide predators with far more prey than one might expect from the moths immediately adjacent to the lamp (Hartstack et al. 1968).

Parasitoids of Lepidoptera fly to electric light sources (Collins & Nixon 1930, Cline et al. 1983). Electric lighting could reduce predation on Lepidoptera by suppressing populations of parasitoids (Worth & Muller 1979). It may divert parasitoids used for biological control of pest Lepidoptera in warehouses (Cline et al. 1983). Even brief exposure to intense sources of radiant energy (photoflash) may sterilize minute hymenopterous parasites which survive the radiation (Riordan 1964). Theoretically, lighting could affect secondary parasites, thus potentially disturbing the food chain at three levels, and producing changes in populations which would be difficult to predict (Frank 1986).

EFFECTS ON MOTH POPULATIONS Evidence Against Effects

Migration and dispersal. Even though lamps may contribute to the destruction of vast numbers of moths, the impact on moth populations may be negligible. For example, more than 10000 Autographa (Plusia) gamma (L.) (Noctuidae) were collected in a light trap in one season in England (Robinson & Robinson 1950a). In England the population of A. gamma is maintained almost entirely by immigration in spring from southern Europe (Ford 1972). A particular light source in England should have a negligible influence on the breeding stock which annually

VOLUME 42, NUMBER 2 rol

replenishes the population of A. gamma around it. Seasonal movement of moths over long distances is not rare (Williams et al. 1942, Williams 1958, Johnson 1969, Ford 1972) and may be sustained by wind trans- porting moths at altitudes sometimes hundreds of meters above most electric light sources (Glick 1965, Mikkola 1986, Raulston et al. 1986, Wolf et al. 1986).

Failure to suppress agricultural pests and other species. One might expect that light traps could substantially reduce or eliminate some moth populations. However, elaborate efforts to exploit such traps for pest control have failed, and successes could not be consistently repli- cated (Cantelo 1974, Hienton 1974). The failure has been attributed to influx of moths from outlying areas, but light trapping may fail to control insect populations even on small islands. On St. Croix, United States Virgin Islands, 250 black-light traps were deployed during a period of 43 months. The island is 208 km? in area. Although decreases in light-trap collections suggested that traps were depleting the island’s sphingids (Cantelo et al. 1972a, 1972b), other studies using the same traps at the same time found similar decreases in collections of Heliothis zea even though traps collected only a minute fraction of the island’s H. zea population (Cantelo et al. 1973, 1974, Snow et al. 1969). Fur- thermore, light-trap collections of sphingids were beginning to increase at the time the study was terminated. Meteorologic and density-de- pendent ecological forces may determine the size of moth populations exposed to lighting, even on isolated islands.

Failure of light traps to reduce insect populations extends beyond species of agricultural interest. Williams (1939) examined 150 species of Noctuidae and Geometridae collected in his stationary light trap during a 4-year period in Rothamsted. Comparison of numbers of individuals of each species collected from year to year provided no evidence of any consistent declines in populations, except possibly in the case of one geometrid. More recent observations at Rothamsted extended Williams’ studies. Taylor et al. (1978) tabulated annual num- ber of species and number of specimens of each trapped at Rothamsted from 1966 to 1975, and also calculated an index of diversity for each year. No downward trends are apparent, despite wide fluctuations from year to year.

Prevalence of urban moths. The above studies did not simulate urban conditions where lighting is dense and widespread. However, large numbers of species have been collected in urban areas in Britain and the United States (Langmaid 1959, Lutz 1941). Collections based on a nationwide network of 172 light traps in Britain suggest that moth populations in areas undergoing urban changes can substantially recover despite electric lighting (Taylor et al. 1978). In North America, some

78 JOURNAL OF THE LEPIDOPTERISTS’ SOCIETY

saturniid species not only tolerate urban lighting but may actually thrive better in urban than in rural habitats. Hyalophora cecropia and Samia cynthia are two examples. The ecology of both species is complex, and numerous factors other than lighting can account for changes in their abundance in illuminated environments (Sternburg et al. 1981, Frank 1986). In New England, eight species of Catocala (Noctuidae) thrive in illuminated urban or suburban areas. Seven of these species can be found within a mile of downtown New Haven, and one occurs in downtown Boston. Several depend almost entirely on urban-suburban shade trees (D. F. Schweitzer pers. comm.).

Extinctions unrelated to lighting. Most declines and extinctions in moth populations can be linked to specific circumstances unrelated to lighting (Bretherton 1951, Ford 1972, Heath 1974). These include de- forestation, agriculture, and draining of fens. Destruction of habitats as a cause of widespread declines in Lepidoptera populations has been described in detail for European butterflies (Kudrna 1986). In Britain, many species of moths became scarce around the middle of the last century, but after World War I the situation reversed, probably because of favorable climatic changes (Heath 1974). Declines in numbers of Malacosoma americanum (F.) (Lasiocampidae) in Winnipeg, Mani- toba, have been attributed to English sparrows (Passer domesticus L., Passeridae) eating the moths at lamps (Krivda 1980), but M. ameri- canum populations fluctuate at intervals independent of changes in lighting. Interval duration is about 10 years (Johnson & Lyon 1976). Attacks by microbial and parasitic agents probably account for periodic reductions in populations of this species (Lutz 1941).

Saturniid populations in the northeastern United States declined in the 1950’s. This observation is supported by dates of last capture for species represented in regional collections, and by surveys of collectors (Ferguson 1971, Hessel 1976, D. F. Schweitzer pers. comm.). Popula- tions of some saturniid species have since shown signs of recovery, whereas other saturniids, especially the two Citheronia species native to the area, have failed to recover in several states (D. F. Schweitzer pers. comm.). Declines that occurred in the 1950’s coincided with wide- spread aerial spraying against gypsy moth, and recoveries coincided with drastic curtailment of this spraying (D. F. Schweitzer pers. comm., Gerardi & Grimm 1979). Whether pesticides can account for changes in saturniid populations is unclear. However, changes in populations of saturniids as a group correlate poorly with changes in outdoor lighting.

Evidence for Effects

Small colonies exposed to lighting. Evidence that outdoor electric lighting has the capacity to affect populations of moths is illustrated by

VOLUME 42, NUMBER 2 79

Hydraecia petasitis Doubleday (Noctuidae) in Finland. Only three or four isolated colonies are known to exist in the country. The isolation is not due to urbanization but rather to the fact that the species in Finland is at the extreme tip of its range. Two small colonies were studied, one covering 700 m?, the other 800 m?. A mark-recapture experiment conducted during 48 days in one colony demonstrated that a trap equipped with an 80-watt mercury lamp captured 53% of males in the colony and 30% of females at least once. The colony was estimated to consist of 218 individuals. These and other observations suggest that continuous light trapping could destroy this population. The authors point out that this species is only mildly attracted to light, and that the effect of light trapping might be more severe for other Lepidoptera (Vaisanen & Hublin 1983). The number of moths the authors trapped probably underestimated the number that flew to the lamps (Hartstack et al. 1968).

The Finnish light-trap study demonstrates that a substantial propor- tion of individual moths within a geographically small colony may fly to an electric lamp. It is conceivable that disturbances in oviposition, mating, feeding, vision, navigation, dispersal, crypsis, circadian rhythms or photoperiodism would be sufficient to disrupt an already shaky population or to impede establishment of a new one. Disruptive effects would be even greater when caused by lamps in special conditions. These include lamps in traps equipped with electrocuting grids (“bug zappers’) and lamps near bird feeders and bird houses. Lamps may incinerate or desiccate moths trapped inside poorly constructed or bro- ken luminaires. Lamps near hostplants may disturb females attracted to the plants, or they may disturb males attracted to the females. Lamps in open garages and pavilions may direct moths into areas from which they cannot escape. Automobile headlamps and streetlamps divert moths into the paths of moving vehicles.

Urbanization and fragmentation of habitats. The same urban changes that increase outdoor electric lighting also tend to fragment habitats (MacArthur & Wilson 1967). The result is creation of small colonies exposed to electric illumination. Man has made many species of British moths in effect relict faunas, remnants of a bygone era when their habitats were much more widespread (Bretherton 1951, Ford 1972). Three species of noctuids once plentiful in southern California have been reduced to small, isolated colonies, in one instance in the vicinity of the Los Angeles International Airport (Hessel 1976). Urban gardens and parks now function as important faunal reservoirs (Frankie & Ehler 1978, Davis 1978, 1982, Owen 1978, Schaefer 1982). Urbanization in- creases both vulnerability and exposure of moth populations to lamps.

Lighting as a selective force. Outdoor lighting may act as a selective

80 JOURNAL OF THE LEPIDOPTERISTS SOCIETY

force against particular individuals within a population. For example, it may select against individuals that tend most strongly to exhibit flight- to-light behavior. In the Finnish light-trap study, such individuals would include those that flew into traps most frequently. Industrial melanism demonstrates that urban change may cause evolutionary change in populations of moths, and that disturbances in crypsis can generate the selective forces needed to produce such evolution (Kettlewell 1973, Cook et al. 1986). Electric lighting disturbs crypsis, but also a multitude of other functions. That some species of noctuids and other nocturnal moths do not fly to nearby light sources, or do so only rarely (Bretherton 1954, Taylor & Carter 1961, Janzen 1983), suggests that evolutionary modification of flight-to-light behavior has already occurred, although the causes are unknown.

Responses to selective pressures produced by lighting may be diverse. For species active at dusk, natural selection could favor individuals that fly at the beginning of the population’s flight period, rather than at the end when flight to light occurs. The evolutionary response would be a shift in flight period rather than a specific change in flight-to-light behavior. Biological clocks are in part genetically controlled, and clock mutants affecting time of eclosion and locomotor activity have been identified in Drosophila (Konopka & Benzer 1971, Yu et al. 1987). In moths, different races or strains of a single species exhibit different photoperiodic behavior (Gardiner 1982, Ankersmit & Adkisson 1967), and selective pressures can account for such differences (Tauber & Tauber 1978, Hoy 1978, Waldbauer 1978). On the other hand, ad- vancing or delaying flight times could disturb species segregation me- diated through allochronic flight periods (Tuttle 1985), or it could expose moths to increased predation by birds or bats that fly only at certain times. Any evolutionary response to selective pressures generated by electric lighting would have to represent a net response to opposing selective pressures.

The diversity of moth behavior around lamps suggests a multitude of possible mechanisms for reducing adverse effects of electric light. The degree to which moths of different species fly to lamps may depend on the degree to which they respond to alternative navigational cues that compete with the lamps (Janzen 1984). Suppression of flight-to- light behavior could take the form of increasing responsiveness to com- peting stimuli such as olfactory, geomagnetic, aerodynamic, gravita- tional and inertial cues, plus alternative visual cues (Baker & Kuenen 1982, Baker & Mather 1982, Schone 1984, Janzen 1984, Riley & Reyn- olds 1986). Within a population of moths, variation exists not only in tendency of different individuals to fly to light, but also in tendency to linger at the light or fly past it. Variation may also exist in tendencies

VOLUME 42, NUMBER 2 81

to avoid lamps or oviposit near them. Evolutionary changes in response to electric lighting may be complex.

Forces opposing evolutionary reduction of flight-to-light behavior, however, are difficult to understand and assess in individual cases. Studies have employed suction traps to measure aerial densities of moth populations and at the same time light traps to measure flight to light. These studies suggest that Xestia (Amathes) c-nigrum (L.) (Noctuidae) is 5000 times as likely to fly to light as Amphipyra tragopoginis (Cl.) (Noctuidae) (Taylor & Carter 1961). Why these two noctuids behave so differently around lamps is a mystery. Failure to evolve seemingly advantageous adaptations has been well described in Lepidoptera (Ehr- lich 1984). Populations of moths may resist strong selective pressures to evolve defenses against adverse effects of electric light.

Fewer moths at urban lamps. Evolutionary changes in wing color- ation can be documented by inspection of collections of moths obtained over a period of time (Kettlewell 1973). Evolutionary changes in flight- to-light behavior cannot be documented in this way. Observations a century ago, however, are worth noting. Riley (1892: 51) advises col- lectors where to look for moths: “‘. .. nowadays the electric lights in all large cities furnish the best collecting places, and hundreds of species may be taken in almost any desired quantity.”’ Denton (1900:35) was more explicit:

While employed in Washington, D.C., I made a splendid collection of the moths of that region simply by going the rounds of a number of electric lights every evening. The lamps about the Treasury Building were sometimes very productive of fine spec- imens and the broad stone steps and pillars were frequently littered with moths, May flies beetles, etc., where one could stand and pick out his desiderata with little difficulty. I captured several of the Regal Walnut moths (Citheronia regalis) and a number of our largest and handsomest sphinxes. Besides making the acquaintance of a number of insects new to me, I met several entomologists who, like myself, had been attracted to the lights by the abundance of specimens.

Today lamps in big cities such as Washington, D.C., Philadelphia, and Boston rank among the worst places to collect moths or meet ento- mologists. Reductions in numbers of moths flying to lamps have been noted in other locations (Hessel 1976, Muller 1979, Janzen 1983). De- creases in moths at urban lamps can be explained by many factors, including declines in moth populations, dilution of moths among thou- sands of city light sources, and suppression of flight-to-light behavior as a result of diffuse background light. However, reductions in numbers of moths flying to urban lamps are what one would expect if urban moths today were genetically less inclined to fly to lamps than were those a century ago.

In densely illuminated urban environments, lighting may have fa- vored species that either fly during the day, do not fly to lamps, or do

82 JOURNAL OF THE LEPIDOPTERISTS SOCIETY

not fly at all. Urban pests exemplify such species. These include sesiids (Engelhardt 1946) and domestic tineids (Ebeling 1978). Species with flightless females include the bagworm moth, Thyridopteryx ephem- eraeformis (Haw.) (Psychidae), gypsy moth, Lymantria dispar (L.), (Lymantriidae), white-marked tussock moth, Orgyia leucostigma (J. E. Smith) (Lymantriidae), and fall cankerworm, Alsophila pometaria (Harris) (Geometridae) (Lutz 1941, Drooz 1985). The two urban sa- turniids, Hyalophora cecropia and Samia cynthia, do not commonly fly to urban light sources (G. P. Waldbauer pers. comm., Covell 1984). The extent to which lighting may have influenced the kinds of moths inhabiting densely illuminated urban environments is unclear.

METHODS TO REDUCE DISTURBANCES

Low-pressure sodium lamps may be used to reduce disturbances caused by lighting. Low-pressure sodium lamps elicit flight-to-light behavior less frequently than do other lamps (Robinson 1952). They do not disturb certain circadian rhythms of Lepidoptera and other insects (Frank & Zimmerman 1969, Bruce & Minis 1969, Pittendrigh et al. 1970, Truman 1976). The low-pressure sodium lamp radiates less energy than does any other kind of lamp of equal illuminance (Finch 1978).

A variety of measures may protect moths from adverse effects of outdoor lighting. Lamp-free reserves such as sheltered hollows shielded from lighting have been suggested to save the glow worm, Lampyris noctiluca L.. (Coleoptera: Lampyridae), a species whose survival in Britain may be threatened by outdoor lighting (Crowson 1981). To reduce lighting impact in habitats already exposed to lamps, the most effective action is to turn off the lamps. Low-pressure sodium lamps may replace other lamps when illumination is essential. Filters to block ultraviolet light may be installed over mercury vapor lamps, and shields may be placed around lamps to block stray light. Low-watt orange- colored incandescent lamps (“bug lights’) may replace ordinary in- candescent lamps, but some moths fly to these lamps. Bird feeders may be removed from windowsills, lampposts, and other sites close to light sources. “Bug zappers” should be turned off. Natural light-traps such as open garages may be closed to prevent entry of insects. Operators of nearby commercial light sources such as illuminated billboards may be contacted and invited to save money and moths by turning lamps off during those hours of night and early morning when billboards are rarely seen.

Although the feasibility of such changes may be questioned, several North American cities have taken similar steps to reduce light pollution. Light pollution interferes with astronomical work at observatories (Hen-

VOLUME 42, NUMBER 2 83

dry 1984). These cities have converted streetlamps to low-pressure so- dium, required ultraviolet-blocking filters over mercury lamps, imposed curfews on the use of commercial lighting, and mandated shielding of luminaires (Hendry 1984). Low-pressure sodium lighting, however, has provoked political controversy on aesthetic and other grounds (San Jose Committee of the Whole 1980).

CONCLUSION

Effects of outdoor lighting may be divergent. They vary according to species, lamps, and habitats. Improved levels of illumination may increase nocturnal vision, but creation of visual artifacts may disturb vision. Increased numbers of lamps may promote flight-to-light behav- ior, but high levels of background light may suppress this behavior. Expansion of streetlighting may increase flight to streetlamps, but shifts from mercury to sodium lamps may decrease it. Diversion of moths to lamps may increase numbers of moths in illuminated areas, but diver- sion of moths away from lamps may decrease numbers. Lamps may suppress oviposition in the laboratory, but oviposition may increase or decrease near lamps in the field. Clumping of moths near lamps may increase predation by birds and bats, but destruction of parasitic wasps and flies at lamps may decrease predation. Disturbances such as habitat destruction and urbanization may further confound effects of outdoor lighting.

Several conclusions emerge from the observations on lighting. Out- door lighting may destroy vast numbers of individual moths without apparently suppressing populations of moths. However, it disturbs some populations more than others, and it disturbs some individuals more than others in the same population. It generates selective pressures favoring adaptations for protection against adverse effects of lamps. The result may be evolutionary changes in behavior, or changes in the kinds of moths inhabiting illuminated environments. These changes may increase through time as urban expansion fragments habitats and exposes smaller moth populations to electric illumination.

Conservation efforts need to consider adverse effects of outdoor light- ing. If one wishes to protect Lepidoptera in small, endangered habitats exposed to outdoor lighting, reducing or changing exposure may be helpful. In such habitats light traps including “bug zappers’” may de- plete populations of moths. Some cities have attempted to reduce light pollution to protect astronomical observatories. Whether similar large- scale restrictions on lighting might help to conserve Lepidoptera has yet to be demonstrated.

Future research could help clarify lighting impact. Despite abundant evidence that outdoor lighting affects individual moths, few studies

84 JOURNAL OF THE LEPIDOPTERISTS SOCIETY

have attempted to quantify lighting effects on moth populations. Evi- dence that lighting has suppressed populations of particular moths such as saturniids is weak. Studies similar to those on the effects of illumi- nation of orchards and cotton fields (Herms 1929, 1932, Nomura 1969, Nemec 1969) could be extended to other settings and species. Faunal surveys, life history studies, and ecological studies could examine Lep- idoptera in differently illuminated environments. Behavioral and phys- iological studies could investigate the possible evolution of tolerance to adverse effects of lighting. The method might include comparison of Lepidoptera sampled from large geographic regions that possess dif- ferent levels or kinds of outdoor illumination.

ACKNOWLEDGMENTS

D. F. Schweitzer, D. C. Ferguson, and anonymous reviewers provided valuable infor- mation and criticism. C. A. Oerkvitz of the City of Philadelphia Department of Streets provided data on street lighting in Philadelphia. R. L. Edwards introduced me to astro- nomical literature on light pollution. The National Oceanic and Atmospheric Adminis- tration provided nocturnal satellite images of the United States. GTE Corporation granted permission to reproduce spectral distribution graphs for Sylvania lamps. The Illuminating Engineering Society granted permission to reproduce the spectral distribution graph for the low-pressure sodium lamp. R. H. Frank and S. E. Frank edited the manuscript. E. F. Hoeber, T. R. Hoeber, and C. A. Baylor provided special assistance. This paper was presented in part at a meeting of the American Entomological Society on 18 February 1987.

LITERATURE CITED

AGEE, H. R. 1969. Mating behavior of bollworm moths. Ann. Entomol. Soc. Am. 62: 1120-1122.

1972. Sensory response of the compound eye of adult Heliothis zea and H.

virescens to ultraviolet stimuli. Ann. Entomol. Soc. Am. 65:701—705.

1973. Spectral sensitivity of the compound eyes of field-collected adult bollworms and tobacco budworms. Ann. Entomol. Soc. Am. 66:613-615.

AGEE, H. R., J. C. WEBB & H. M. Tarr. 1972. Activity of bollworm moths influenced by the full moon. Environ. Entomol. 1:384-385.

ALLEN, N. & C. R. HODGE. 1955. Mating habits of the tobacco hornworm. J. Econ. Entomol. 48:526-528.

ANDRESEN, J. W. 1978. Do streetlights turn city trees into late-growing insomniacs? Am. For. 84:12.

ANKERSMIT, G. W. & P. L. ADKISSON. 1967. Photoperiodic responses of certain geo- graphic strains of Pectinophora gossypiella (Lepidoptera). J. Insect Physiol. 13:553- 064.

BAKER, R. R. & J. G. MATHER. 1982. Magnetic compass sense in the large yellow underwing moth, Noctua pronuba L. Anim. Behav. 30:543-548.

BAKER, R. R. & Y. SAaDOvy. 1978. The distance and nature of the light trap response of moths. Nature 276:818-821.

BAKER, T. C. & L. P.S. KUENEN. 1982. Pheromone source location by flying moths: A supplementary non-anemotactic mechanism. Science 216:424—427.

BARKER, R. J., C. F. COHEN & A. MAYER. 1964. Photoflashes: A potential new tool for control of insect populations. Science 145:1195-1197.

BEATY, H. H., J. H. Litty & D. L. CALDERWOOD. 1951. Use of radiant energy for corn borer control. Agr. Engin. 32:421-422, 426, 429.

VOLUME 42, NUMBER 2 85

BEcK, S. D. 1980. Insect photoperiodism. 2nd ed. Academic Press, New York. 387 pp.

BEEBE, R. 1953. Sampling Michigan Lepidoptera by the fixed light trap. Lepid. News 20:28.

BEEBE, W. 1949. Insect migration at Rancho Grande in north-central Venezuela. General account. Zoologica 34:107-145.

BEEBE, W. & H. FLEMING. 1951. Migration of day-flying moths through Portachuelo Pass, Rancho Grande, north-central Venezuela. Zoologica 36:243-266.

BERLINGER, M. J. & G. W. ANKERSMIT. 1976. Manipulation with the photoperiod as a method of control of Adoxophyes orana (Lepidoptera: Tortricidae). Entomol. Exp. Appl. 19:96-107.

BERNHARD, C. G. & D. OTTOSON. 1960a. Comparative studies on dark adaptation in the compound eyes of nocturnal and diurnal Lepidoptera. J. Gen. Physiol. 44:195- 208.

1960b. Studies on the relation between the pigment migration and the sensitivity

changes during dark adaptation in diurnal and nocturnal Lepidoptera. J. Gen. Physiol.

44:205-215.

1964. Quantitative studies on pigment migration and light sensitivity in the compound eye at different light intensitities. J. Gen. Physiol. 47:465-—478.

BLEsT, A. D. 1957. The function of eyespot patterns in the Lepidoptera. Behavior 11: 209-256.

1968. Longevity, palatability and natural selection in five species of New World saturniid moth. Nature 197:1183-1186.

BOWDEN, J. & B. M. CHURCH. 1973. The influence of moonlight on catches of insects in light-traps in Africa. Part I]. The effect of moon phase on light-trap catches. Bull. Entomol. Res. 63:129-142.

BOWDEN, J. & M. G. Morris. 1975. The influence of moonlight on catches of insects in light traps in Africa. III. The effective radius of a mercury-vapor light-trap and the analysis of catches using effective radius. Bull. Entomol. Res. 63:303-348.

BRETHERTON, R. F. 1950. The behavior of moths at light traps: A comment. Entomol. Gaz. 1:102-104.

1951. Our lost butterflies and moths. Entomol. Gaz. 2:211-240.

1954. Moth traps and their lamps: An attempt at comparative analysis. Entomol. Gaz. 5:145-154.

BROWER, A. E. 1986. Predation on Catocala moths (Noctuidae). J. Lepid. Soc. 39:280- 283.

Brown, C. H. 1976. A survey of the Sphingidae of Sanibel Island, Florida. J. Lepid. Soc. 30:230-233.

Brown, L. N. 1984. Population outbreak of pandora moths (Coloradia pandora Blake) on the Kaibab Plateau, Arizona (Saturniidae). J. Lepid. Soc. 38:65.

BRUCE, V. C. & D. H. MINIs. 1969. Circadian clock action spectrum in a photoperiodic moth. Science 163:183-1835.

CANTELO, W. W. 1974. Blacklight traps as control agents: An appraisal. Bull. Entomol. Soc. Am. 20:279-282.

CANTELO, W. W., J. L. GOODENOUGH, A. H. BAUMHOVER, J. S. SMITH JR., J. M. STANLEY & T. J. HENNEBERRY. 1974. Mass trapping with blacklight: Effects on isolated populations of insects. Environ. Entomol. 3:389-395.

CANTELO, W. W., J. S. SMITH JR., A. H. BAUMHOVER, J. M. STANLEY & T. J. HENNEBERRY. 1972a. Suppression of an isolated population of the tobacco hornworm with blacklight traps unbaited or baited with