VOLUME 27, NUMBER 3 225 different lengths, tipped with darker gray. Hindwing pale yellowish-white, darkened near apex and, in the female, along outer margin. Beneath: forewing yellowish-gray, hindwing yellowish-white. Wing expanse: 21 to 24 mm. Holotype: Male, Big Bend National Park, Green Gulch, Texas, 28 March 1971, gen. prep. A.B. 3030, deposited in the National Museum of Natural History, (No. 72380). Paratypes: All from Big Bend Nat. Park, Green Gulch: 9 October 1969, 1, (A.B. 2011); 28 March 1971, 5, (A.B. 3030, 2649, 2650); 3 May 1972, l<j>, (A.B. 3029); 12 May 1972, 1, 1<;l, (A.B. 3028). ACKNOWLEDGMENTS I am deeply grateful to Dr. D. C. Ferguson of the Entomolgy Research Division, U.S.D.A., for revising the manuscript and making several very helpful suggestions. The authorizations given me by Mr. C. D. Laughlin to collect on the McDonald Observatory grounds, and by the Administration of Big Bend National Park to collect around and in the Chis os Mountains are also gratefully acknowledged. LITERATURE CITED HEINRICH, C. 1956. American Moths of the subfamily Phycitinae. U.S. Nat. Mus. Bull. 207. NOTES ON THE TAXONOMIC STATUS OF HYALOPHORA COLUMBIA (SATURNIIDAE) MICHAEL M. COLLINS 924 Mendocino Avenue, Berkeley, California 94707 f I The work of many authors (Sweadner, 1937; Weast, 1959; Collins & Weast, 1961; Wright, 1971) has shown that the various forms of Hyalophora are not reproductively isolated from one another. Females of any form of Hyalophora (in the restricted sense of Ferguson, 1972) will attract and mate indiscriminately with males of any other form. Generally ova laid by cross-mated females are viable and usually produce fertile F 1 males and sterile females. Backcrossing the F 1 male with a female of either parental form or even a third form again produces fertile males and generally sterile females. Occasionally these females may lay some fertile ova. Intergrades and hybrids occur in nature. A population currently designated "kasloensis" (Cockerell) exists between H. gloveri (Strecker) and H. euryalus (Boisduval) in Idaho, western Montana, and British
226 JOURNAL OF THE LEPIDOPTERISTS' SOCIETY Columbia and is believed to be an intergrade between these two forms. H. cecropia (Linnaeus) produces occasional hybrids with gloveri. I obtained such an individual when a hybrid-like cocoon collected along with typical gloveri in the Black Hills of South Dakota produced a female very similar to laboratory hybrids. Mr. Duke Downey (pel's. comm.) has collected supposed hybrids in Sheridan, Wyoming, but only rarely. The contact between gloveri and cecropia may be the recent result of a westward invasion of the latter species (Cockerell, 1929; Peterson & Worden, 1962). Wild hybrids between H. columbia (Smith) and cecropia have been cited by Sweadner (1937). I collected a hybrid male along with typical columbia one mile west of Whiteshell Provincial Park along Highway One in Manitoba on 18 June 1963. This specimen will be described in a later paper. No true blending occurs between cecropia and the other forms but introgression is indicated by columbia and gloveri specimens with red scales in the extradiscal band. About half the specimens of gloveri and columbia sampled by Sweadner in areas of contact with cecropia exhibited this trait. The density of recognizable hybrids in nature is lower than predicted from the results of laboratory crosses. In addition, the mobility of the males during the mating flight and of the females during oviposition assures a potentially high rate of gene flow. Selection must act strongly against hybrids in some way. Hybrids may be ill-adapted such as in oviposition habits and consequent foodplant acceptance. H. columbia spins a small cryptically colored cocoon on exposed tamarack branches. W'hen cecropia breeds with this form the larger non-cryptic hybrid cocoon may be more easily found by predators. Given the ability of the Hyalophora forms to intergrade, the origin and present status of columbia seemed to be an excellent subject for study. In the United States gloveri is quite distinct phenotypically and is separated geographically from the small dark form in Maine, Wisconsin, and Michigan designated columbia. In western Canada the gloveri population exhibits a phenotype nearer to columbia in size but often even more brightly colored than the Rocky Mountain phenotype. Consequently, gloveri and columbia have up to now been considered separate species. On the basis of geographic distribution, wing pattern, and identical genitalic structure, Sweadner believed columbia arose from gloveri at the end of the last period of glaciation. To account for the apparent lack of intergradation, he believed the two became geographically isolated by foodplant preference. The gloveri population in the prairies of Alberta, Saskatchewan, and Manitoba, which has been given the subspecies name nokomis, feeds on wolf willow (Shepherdia) and to a lesser extent on willow (Salix). The Hyalophora in the swamps east of Winnipeg,
VOLUME 27, NUMBER 3 227 recognized as columbia, is thought to feed entirely on tamarark (Larix). These food plants are not isolated, however. The Shepherdia of the plains meets to the north a broad band of spruce-tamarack forests ranging from British Columbia across Alberta and Saskatchewan to Manitoba where the conifers dip south to the east of Winnipeg (Harlow & Harrar, 1949). Thus, a foodplant bridge connects the two populations but at a more northern point than was sampled by Sweadner. I Field Work I conducted field research in Canada in 1963, 1964, and 1966 to better sample these populations. Tied females and females in specially designed moth traps were placed at intervals throughout selected areas. The moth trap consists of a nine inch metal funnel mounted at one end of a screen cylinder two to three feet high. A small cage is suspended above the mouth of the funnel and confines a H yalophora female. Attracted males, in their frenzy to reach the female, eventually collide with the funnel and slide through its enlarged opening. Thus, several males can be taken at each site and the female can be used as bait for up to 4 or 5 days. Three population areas were sampled. The general trap line was along Highway One east and west of Winnipeg. The small dark phenotype was first taken just outside of the tamarack bog 30 miles east of Winnipeg near Richer. Thirty-two specimens were taken at regular intervals up to the Manitoba-Ontario Border beyond which no traps were placed. No males of any form were collected beyond the western extent of the bog 30 miles east of Winnipeg to a point 90 miles west of Winnipeg. Shepherdia does not seem to occur in this immediate area and extensive farming must further isolate the two Hyalophora forms. Stands of Shepherdia occur west of here along fence rows and in untillable terrain. To the west the first male of the brightly colored phenotype was taken 26 miles east of Brandon, Manitoba. A total of 17 specimens was taken westward along Highway One to Swift Current, Saskatchewan. The only cecropia collected was trapped 5 miles west of Virden, Manitoba on 21 June, 1963. In an attempt to collect a predicted intermediate form between gloveri and columbia, I collected north along Highway Ten to Flin Flon, Manitoba and then west to Prince Albert Park, Saskatchewan. The terrain near Flin Flon is unusual in that the tamarack bog is interrupted by granite outcroppings and the flora is consequently more varied. Shepherdia was not seen near Flin Flon. The possible implications of this environmental change was discussed below. Frequent cold weather limited collecting but 12 males were taken from 17 to 22 June, 1964.
228 JOURNAL OF THE LEPIDOPTERISTS' SOCIETY
VOLUME 27, NUMBER 3 229 Analysis of Specimens When the population samples collected were compared superficially, it was clear that not all specimens could be placed into one of two distinct phenotypes. Furthermore, the light and dark moths were not always associated with the expected foodplant community. One of the males from the tamarack bogs east of Winnipeg was as brightly colored as the average Shepherdia-feeding phenotype and at least one of the specimens from the latter population was quite as dark as those from east of Winnipeg. Sweadner reported similar specimens from the large samples he secured. The moths taken from the north were quite variable; some were as dark as those from east of Winnipeg and others were as bright as the prairie phenotype. Most were intermediate in coloration (Fig. 1). To present a more objective comparison, I quantified this color variance and analyzed it statistically. While the more northern sample was somewhat small, I believe the results are significant. The ground color of the dark phenotype is usually accompanied by a darkening of the color band distal to the white wing band. Under the microscope one sees that this darkening is the result of an increase in the number of black scales relative to the number of white scales. Using a magnification of 40x with a grid reticle I measured the relative scale density in this band contained in the cell formed by veins CUl and M 3 The total numbers were converted to percentages and then plotted as a distribution with the mean and standard deviation. Fig. 2 shows that the specimens from northern Manitoba are more variable than those from east of Winnipeg and are very intermediate for this character. Surprisingly, the moths from northern locales were larger on the whole in both average and maximum size than specimens taken further south. Males collected east of Winnipeg (32 specimens) averaged 50.3 mm measured from point of attachment of the forewing to the apex. These specimens ranged from 45.0 to 54.0 mm. Moths collected from 250 to 375 miles NW of Winnipeg (12 specimens) averaged 55.8 mm and ranged from 50.0 to 61.0 mm. The prairie sample (17 specimens) averaged 52.2 mm and ranged from 48.0 to 55.0 mm. Fig. 1. Extreme light and dark specimens for each population area: H. columbia: 1, Manitoba, 8 mi. e. Richer, 18 June 1966; 2, Manitoba, 17 mi. e. Richer, 18 June 1966. Intermediate Hyalophora: 3, Manitoba, 3 mi. s. Baker's Narrows, 21 June 1964; 4, Saskatchewan, 80 mi. w. Flin Flon, Ballantyne Bay, 22 June 1964. H. gloveri nokomis: 5, Manitoba, 3 mi. e. Brandon, 17 June 1966; 6, Manitoba, 5 mi. w. Virden, 21 June 1963.
230 JOURNAL OF THE LEPIDOPTERISTS' SOCIETY B R I 100 90 80 70 60 50 S.D.=11.0 x=71 i INTERMEDIATE HYALOPHORA 250-375 MILES NORTH-WEST OF WINNIPEG l n en o Fig. 2. Comparative scale counts of specimens collected in the three population areas. Scales were counted in a cell of the grey band of the FW formed by veins Cu, and M3 S.D.= 11.0 i=57 H. GLOVERI NOKOMIS 90-500 MILES WEST OF WINNIPEG After a thorough examination, I am in agreement with Ferguson ( 1972) that there are no significant structural differences, including genitalic characters, between gloveri nokomis and columbia adults. As with the adults, the larvae of these two forms differ mainly in size and coloration. I have reared over 100 gloveri larvae from north central Montana and found 25% to exhibit up to the fifth instar the same black dorsal and lateral scoli seen in immature columbia larvae. In the last instar the columbia larva differs from gloveri mainly in the color of the enlarged dorsal tubercules on segments 2, 3, and 4; these scoli vary from dull red-brown to a brighter red and are encircled at their bases by black. In gloveri these same scoli are usually yellow with a reduced black basal ring. I have seen a larva from Sheridan, Wyoming which possessed dull orange thoracic dorsal scoli. The remaining dorsal and lateral scoli are nearly identical in both columbia and gloveri; those on the sides vary from light blue tipped with white to nearly all white while the dorsal scoli on the abdominal segments are pale yellow. I have reared a few Montana (Toole Co.) gloveri which possessed yellowish orange dorsal scoli. Published descriptions of fifth instar gloveri larvae are varied. Sweadner (1937) states that all the dorsal scoli are straw yellow, "but in a few exceptions those on segments two, three, and four are slightly deeper in shade." He further mentions a larva from Glacier National
VOLUME 27, NUMBER 3 231 Park, Montana with the dorsal scoli on segments 2 and 3 colored a "dull burnt orange." Cooley (1908) reared gloveri larvae from Bozeman (Gallatin Co.), Montana and described the dorsal scoli on segments 2 and 3 as "coral red" in color. Few descriptions of the gloveri nokomis larva have been published but Freedley (1908) describes the dorsal scoli as all yellow in Alberta specimens. Thus it appears that while columbia larvae are rather constant in coloration, a certain percentage of fifth ins tar gloveri larvae possess reddish dorsal scoli on segments 2 and 3 and the immature larvae of both forms may be nearly identical except for size. Mature hybrid larvae of columbia X gloveri may exhibit either all yellow or all red dorsal scoli on segments 2, 3, and 4. The cocoons of gloveri nokomis and columbia are said to be indistinguishable. Laboratory Breeding The results of interbreeding were limited by the high rate of disease present in all my Hyalophora larvae, including pure strains. Nevertheless, two female adults were reared on Larix from a mating between a female from Montana (Toole Co.), and a male taken 58 miles east of Winnipeg. Both of these females appeared fully fertile and laid a normal compliment of ova when bred to Montana gloveri. No exact count was made of the number of ova laid or the percentage of hatch but approximately 90% of the ova produced larvae. The resulting larvae were unfortunately lost to disease. More recently, hybrids between columbia (Livingston Co., Michigan) and gloveri from western Utah were reared by Mr. James Tuttle and Robert Weast. Two pairings of a male columbia X female gloveri yielded 92% and 90% hatch of ova laid. The reciprocal cross produced an 89% and 61% hatch (James Tuttle, pers. comm.). Robert Weast (pers. comm.) succeeded in inbreeding the F 1 hybrids of gloveri X columbia; the three females laid an average of 150 ova each. This is a normal number for the smaller Hyalophora forms. Fertility was over 90%. The results of this and future crossbreeding should be quantitatively compared to the degree of fertility seen in crosses involving other forms, such as gloveri X euryalus. Furthermore, it should be determined if the fertility between gloveri and columbia changes as a function of geographic (and ecologic) separation of the populations studied. Some new information is available on the conifer feeding habits in the Hyalophora. In 1963 I reared to maturity on Larix two normal gloveri from a Russian Olive feeding population in Montana. Two dozen larvae were initially started successfully on Larix but again disease disrupted the experiment. In England gloveri is reported to be commonly reared on European Larch (Crotch, 1956). Dr. Thomas Koerber of the U.S.
to W to HYALOPHORA COWMBIA e INTERMEDIATE HYALOPHORA o H. GLOVER I NOKOMIS \. II 1IIIllll!. IJ ' ':::::::::::lln... jf:::... B ' ::::::::::::::::::::::!I!i, "1.:::),111111111111111111...=nmlllHHmm :111111111111 }.,',';;;;;;,;;,;; :"i ::..:m::: ::::::::::::::::::::::::;;:':'::-i-i:+:f::ti:::::::::w " :---. ' :TIIT1HfuP.: :TI::mmm::mm:::::::::::::!tm:::::" ;;;;;;;;;;'::::.'. 1949) ::::::::::.. :. :::::... tl Harlow & Harrar,. 0,::.... C C i plants (based par y on Fig. 3. Collection sites, with the approximate distribution or IOU' _ '-< g t" o "l t'l t"" t'l >tl >tl. rjl g t;;
VOLUME 27, NUMBER 3 233 Forest Service has collected wild euryalus larvae and cocoons on Douglas Fir (Pseudotsuga) at several locales in the Klamath and Cascade Mountain ranges of northern California. 'Ve have both reared gloveri and euryalus on Douglas Fir in captivity. A more detailed discussion of this work will be published at a later time. I am at present rearing thirteen columbia larvae which I have found will switch in the fifth instar from Larix to Pseudotsuga after 30 to 45 minutes hesitation. During the first instar five larvae of a total of 24 initially began feeding on Pseudotsuga but eventually wandered to nearby Larix twigs. I did not force any of the first instar larvae to feed on Pseudotsuga due to the present scarcity of columbia stock. I feel that the ability of the various Hyalophora to feed in captivity on conifers, even reluctantly, is especially significant when one considers that many rather monotypic species, such as Callosamia promethea (Drury), may be regionally quite food plant specific (Collins & Weast, 1961). One would expect the acceptance of conifers to decrease outside the normal range of these food plants. Conclusion Gene flow between the brightly colored prame population and the population of dark moths in the tamarack bogs east of Winnipeg produces in each population occasional phenotypes similar to the opposite form. The variability in the northern population of intermediates is also the result of gene flow from the two extreme populations. This clinal variation seems not to be any more sudden in the sense of a "step cline" than is the transition from the Rocky Mt. phenotype into the bright prairie form. Both clines involve phenotypic and foodplant changes. Selection must favor the dark phenotype in the tamarack bogs and the lighter form in the prairie region since on the average the two populations have rather distinct phenotypes. North of Winnipeg the environment undergoes a transition from plains to conifer forest. Perhaps here where the intermediate form occurs, selective forces do not clearly favor either the light or dark phenotype, thus promoting the rate of gene exchange. See Fig. 3 for food plant distribution and collecting sites. Melanism in other Lepidoptera has been shown to be controlled by a relatively few alleles. Thus, while the dark phenotype described as typical of columbia may appear to be qualitatively different from gloveri, the genetic basis for the apparent differences is probably minor. Whether this melanism is directly adaptive or is somehow linked with other adaptive gene systems is unknown. The conifer feeding habit has been shown to not be restricted to columbia and may even prove to be widespread in the western forms. Workers should attempt to determine if "kasloensis" and gloveri feed
234 JOURNAL OF THE LEPIDOPTERISTS' SOCIETY naturally on either Douglas Fir (Pseudotsuga) or Western Larch (Larix occidentalis) in the Bitteroot and Rocky Mountain ranges. The selective basis for the melanism seen in columbia may operate during the larval stage. Perhaps the darker columbia larva is more cryptically colored in comparison to the lighter gloveri larva when feeding on the sparsely needled Larix. The immature columbia are especially difficult to locate when reared in sleeves on larch. If the genes controlling melanin production in the larva are somehow linked with those controlling scale pigmentation in the adult, predator selection would quickly produce dark adults while acting upon the larval stage. Admittedly, such a gene system would be exceptional. Larvae are subject to much greater predation pressure than the short-lived nocturnal adults, however. The "ecological image" of the adults is the underside of their wings as they rest with their wings folded over their backs. Furthermore, euryalus, gloveri, and columbia are much more similar when viewed in this aspect. Thus, the visual variance in dorsal wing coloration may not be the result of selection for this characteristic per Sq. Selective breeding for light and dark larvae and adults may reveal a correlation. In summary, the obvious similarity of the two forms in all stages, the elinal variation of the melanization of the adults, and the demonstrated lack of isolation mechanisms coupled with the fertility seen in hybrid females all suggest columbia is not a separate species but is rather a melanic subspecies of gloveri. It is hoped that this paper will encourage and aid further investigation. ACKNOWLEDGMENTS I would like to sincerely thank Dr. C. Don MacNeill for continued help in the conception and analysis of the research project and in preparing this paper. I am also grateful for the living material and breeding information kindly provided by James Tuttle and Robert Weast. LITERATURE CITED COCKERELL, T. D. A. 1929. The spread of Samia cecropia. J. Econ. Entomoi. 22: 97-98. COLLINS, M. M. & R. D. WEAST. 1961. Wild Silk Moths of the United States. Collins Radio Corp. 138 p. COOLEY, R. A. 1908. Sixth annual report of the state entomologist of Montana. Bull. 75, Mont. Agric. College Exp. Sta. CROTCH, W. F. B. 1956. A Silk Moth Rearer's Handbook. The Amateur Entomoi. Soc., England. p. 143. FERGUSON, D. C. in DOMINICK, R. B. et al. 1972. The Moths of Amelica North of Mexico. Fasc. 20.2B, Bombycoidea (in part). p. 252, 256, 259. FREEDLEY, W. J. 1908. The early stages of Samia columbia nokomis. Can. Entomoi. 40: 350-354.
VOLUME 27, NUMBER 3 235 HARLOW, W. M. & E. S. HARRAR. 1949. Textbook of Dendrology. McGraw-Hill. p. 120-125. PETERSON, L. O. T. & H. A. WORDEN. 1962. The cecropia moth in the prairie provinces. Contr. 20 Forest Nursery Sta. Research branch, Can. Agric., Indian Head, Sask. SWEADNER, W. R. 1937. Hybridization and the phylogeny of the genus Platysamia. Ann. Carnegie Mus. 25: 163-242. WEAST, R. D. 1959. Isolation mechanisms in populations of Hyalophora (Saturniidae). J. Lepid. Soc. 13: 212-216. WRIGHT, D. A. 1971. Hybrids among species of Hyalophora. J. Lepid. Soc. 25: 66-73. NOTES ON SIPROET A AND MET AMORPHA WITH FIGURES OF SIPROETA EPAPHUS GADOUI MASTERS (NYMPHALIDAE) JOHN H. MASTERS 5211 Southern Avenue, South Gate, California 90280 Siproeta epaphus gadoui Masters (1967) was described from a series of specimens collected by Albert and Mary Lou Gadou at El Pao, Bolivar, Venezuela. The subspecies is of particular interest because it is intermediate in many respects between S. epaphus (Latreille) and S. trayja Hubner, which have been considered to be distinct species in the past. S. gadoui appears to be derived from trayja stock rather than from the geographically adjacent epaphus, although the entire Amazon Valley now separates trayja and gadoui. The description of gadoui was hurried in order to make the name available for a genus revision by Richard M. Fox and Alden C. Forbes and the new subspecies was not figured. The original description did promise figures in the coming revision of Fox & Forbes. The senior author of this revision, R. M. Fox, died on 25 April 1968 with the manuscript partially completed. F. Martin Brown completed the manuscript which was finally published on 24 December 1971, albeit without figuring gadoui. I feel that it is important to figure this interesting butterfly and figures are hereby presented (Fig. 1). The Fox & Forbes revision (1971) may not be readily available to all lepidopterists and a summary may be useful. The genus Metamorpha was divided into two genera: Metamorpha Hubner (type species elissa Hubner), and Siproeta Hubner (type species trayja Hubner). Differences were cited in venation (in M etamorpha a vestige of the posterior tip of the third disco cell ular is present on the forewing), male forelegs (in Siproeta the tarsus is little more than half the length of the tibia, in Metamorpha