Genseeeeeaenecectatetetateceeseotosseseoate'atetsteatetetaotereteanete et erecete neesecesretatatatatatetetetatatotetotetetetatetetetatetatetetetetes PO a OT Rg THE UNIVERSITY OF KANSAS SCIENCE BULLETIN & ‘J ‘e*e = me i ‘ee vee ‘eve *% OSS & = 5 oe Res PARASITID MITES ASSOCIATED WITH EES BUMBLEBEES IN ALBERTA, CANADA PEE (ACARINA: PARASITIDAE; HYMENOPTERA: APIDAE). II. BIOLOGY os OR SS ees By LYNN A. RICHARDS and K. W. RICHARDS oacadatatotetetesecsetetetctehea’e"s*etetee'e’st eres SM ey i eoe2ecne a oD acaaheetateteratetetasetetatotasetenenetocanetone Cm ne ececececeeteets m 2 Seaieeesees Vol. 51, No. 1, pp. 1-18 February 4, 1976 “ eeseenesoeseneE ete Re Ee Se ERE EEE EERE EE EEE ANNOUNCEMENT The University of Kansas Science Bulletin (continuation of the Kansas Uni- versity Quarterly) is an outlet for scholarly scientific investigations carried out at the University of Kansas or by University faculty and students. Since its incep- tion, volumes of the Bulletin have been variously issued as single bound volumes, as two or three multi-paper parts or as series of individual papers. Issuance is at irregular intervals, with each volume prior to volume 50 approximately 1000 pages in length. Effective with volume 50, page size has been enlarged, reducing the length of each volume to about 750 pages. The supply of all volumes of the Kansas University Quarterly is now ex- hausted. However, most volumes of the University of Kansas Science Bulletin are still available and are offered, in exchange for similar publications, to learned societies, colleges and universities and other institutions, or may be purchased at $20.00 per volume. Where some of these volumes were issued in parts, individual parts are priced at the rate of 2 cents per page. Current policy, initiated with volume 46, is to issue individual papers as published. Such separata may be purchased individually at the rate of 3 cents per page, with a minimum charge of $1.00 per separate. Subscriptions for forthcoming volumes may be entered at the rate of $20.00 per volume. All communications regarding exchanges, sales and subscriptions should be addressed to the ExcHance LisrariAN, UNIVERSITY oF Kansas Lipraries, LAWRENCE, Kansas 66045. Reprints of individual papers for personal use by investigators are available gratis for most recent and many older issues of the Bulletin. Such requests should be directed to the author. The International Standard Serial Number of this publication is US ISSN 0022-8850. Editor Eugene C. Bovee Editorial Board William L. Bloom Philip W. Hedrick Rudolf Jander Harvey Lillywhite Charles D. Michener Norman A. Slade Henry D. Stone George W. Byers, Chairman THE UNIVERSITY OF KANSAS SCIENCE BULLETIN Vol. 51, No. 1, pp. 1-18 February 4, 1976 Parasitid Mites Associated with Bumblebees in Alberta, Canada (Acarina: Parasitidae; Hymenoptera: Apidae). II. Biology’ Lynn A. RicHarps AND K. W. RiIcHarps TABLE OF CONTENTS LUD GTUNOTE® oP eg cer bale eee coe Pe eas a Se ey sR Bo ee ee Oe eel ee Oe Mean eet eee 1 OD REGALO Nae tr ewe CORSE” a! gh) eter iirsce, be 4 Gees se WR 1 PRCRE OSI DGE NUP NG Sareea eee we Sen Oe Se Oy eee et eee 2 STE TEEYS DS AUNT) AIM VAGRSIS UN EG SM ates ee ee a a eR SR ee ee ee 2 aC CES HUIDNE Stee ee ee sop hd, be hee neti et cen Ye eat Pee ee 3 IMESTSOTUDIES <....-.-.--2 a ee tc is eo Pe ae en i SPECIBKCIAN? "Se es i ok a Se Ce ee i ey Cee 2 ee eee eee ee 10 EaRESTEAN DE IDISPERSAU: Smet Mit: (Wier 2 watmerge Lt Eyl eeg Bs ae ye 4 ae 13 MOU IIMONT LALSULE Se ee ee ek ee me a Eee i7/ eRe GTia RIDIN OTE Dee ee era a ne ee a eee tas be oe eh U7, ABS DRAGI Mites of five species of Parasitus, associated with Bombus in Alberta, Canada, were studied. Life cycle is typically egg, larva, protonymph, deutonymph, and adult, with a duration of about 20-26 days. Cultures of these mites indicate the most effective food to be pollen and wax mixtures. However, protonymphs and deutonymphs failed to molt in culture on any food. Populations of P. favus, P. perthecatus and P. fimetorum have 1n- creasing deutonymphal populations with a lower more constant adult mite population. Deutonymph increase is synchronous with queen production. Deutonymphs are phoretic on queens, except P. fimetorum which is not phoretic on bumblebees. Parasitus is not restricted to one species of Bombus, but is more commonly found with B. bifarius, B. occidentalis and B. frigidus. The percentage of queens with only one species of Parasitus was 59.8°,, while the percent of nests with two or more species was 78°. Mechanisms of dispersion other than establishing queens includes superseding queens, Psithyrus, foragers, and males. Attached to some Parasitus deutonymphs were Scutacarus acarorum. INTRODUCTION The purpose of this study was to inves- the Genus Parasitus and the bumblebees. tigate the relationships between mites of Field work was conducted in southwest- ——— F c av 2) ‘Contribution number 1530, Department of Ento- ern Alberta, Canada, from May 26 mology, University of Kansas, Lawrence, Kansas. through August 25, 1970; and April 25 2 Tue University oF Kansas ScIENCE BULLETIN through August 25, 1971, on the east side of Prairie Bluff Mountain, 49° 20’N, 114° 06’ W, at about 5200 feet elevation. Tax- onomy of these Parasitus is reported by Richards (in press). The biological data associating Para- situs mites with Bombus have been re- ported by Frison (1926), Plath (1934), Sellnick (1938), Kuhnelt (1950), Hughes (1959), Free and Butler (1959), Skou, Holm and Haas (1963), Husband (1968), and Chemielewski (1971). Bumblebees of the following species of Bombus and Psithyrus were encountered in the study area: B. appositus Cresson; B. bifarius nearcticus Handlirsch; B. califor- nicus F. Smith; B. centralis Cresson; B. flavifrons Cresson; B. frigidus F. Smith; B. hunt Greene; B. melanopygus Nylan- der; B. mixtus Cresson; B. nevadensis Cockerell; B. occidentalis Greene; B. rufo- cinctus Cresson; B. ternarius Say; B. terrt- cola Kirby; B. vagans F. Smith; P. insu- laris (F. Smith); P. suckley: (Greene); and P. fernaldae Franklin. The following is an abbreviated ac- count of the life cycle of Bombus (for more detail the works of Hobbs (1964- 1968) and Hasselrot (1960) should be con- sulted). The bumblebee queens appear in the spring after hibernating all winter. When they emerge in early May they start feeding on nectar and pollen and begin looking for nesting sites. If a suitable site is found, the nest material is arranged by the queen and she begins to build an egg cup out of wax and pollen. A honey pot is built either before, after, or at the same time as the first brood eggs are laid. When the queen is incubating the brood she keeps her brood temperature about 37- 38 C (98 F) (Hasselrot 1960). The tem- perature drops when she leaves the nest to forage and rises rapidly when she re- turns, When the first workers emerge they assist the queen in expanding the nest, car- ing for additional brood (feeding and food foraging), and maintaining a con- stant nest-temperature. More workers are usually produced as the season progresses until queen and male eggs are laid. At the end of July or beginning of August, the colonies produce males and queens and begin to break up. After mating, the queens hibernate in chambers in the soil, and the following spring they initiate new colonies. The males and all the workers have died by the time winter arrives. Queens differ from workers most conspic- uously in their larger sizes and greater ovarian development; intermediates be- tween the castes are known in some spe- cies. ACKNOWLEDGEMENTS To Dr. G. A. Hobbs, John Virostek and Evan Gushel and to the Canadian Depart- ment of Agriculture Lethbridge, Alberta, Canada, we extend special thanks for permis- sion to use equipment and for help in setting up a field laboratory. Thanks go, too, to Dr. R. W. Husband, Adrian College, Adrian Michigan, who sent us refrence materials, ideas and advice, and for his helpful review of the manuscript. We wish also to thank Drs.. R. E. Beer, J. H. Camin ande@agme Michener, University of Kansas, for assistance and advice in this project. Our thanks go also to Mr. and Mrs. William Richards and Deborah Richards for their help. METHODS AND MATERIALS To obtain a suitable number of Bom- bus nests for study, 1048 artificial domiciles were placed in natural settings in 1970, and 1080 in 1971. These domiciles were wooden boxes with upholsterer’s cotton used as nesting material (Hobbs, Virostek, Nummi 1960; Hobbs, Nummi, Virostek 1962; Hobbs 1967a). They were placed in groups of four: one above ground, one on the surface, one false underground (ice., on the surface but with an entrance tube covered by a piece of sod), and one under- Parasitip Mires AssociATED WITH BUMBLEBEES 3 ground. Natural nests found in the area were also collected. Field collected bumblebees — were checked for number, genera, and where possible species of mites. In late summer, new queens, workers and males were col- lected from nests and checked for mites. Nest materials were sorted for mites. The upholsterer’s cotton was pulled apart and mites collected into 70°% alcohol with a small brush. Similarly, the bee broods were carefully cleaned of mites. When natural nests were collected, the nesting materials were put into plastic bags, and dirt surrounding the nests was also gath- ered. The broods were carefully picked over for mites, while the nesting material and dirt were placed on white paper and sorted in a manner similar to that for the artificial domiciles. The deutonymphs from field collected bumblebees and from bee nests, as well as adults, protonymphs and larvae from nests were collected and observed alive in small culture dishes. Culture dishes were con- structed from plastic containers used to store zippers, 25 mm deep and 30 mm in diameter. A plaster of Paris and charcoal mixture (Huber, 1958) was poured into the bottom, 5-7 mm deep. Small culture dishes were made from plastic vials 14 mm high filled with the plaster and charcoal mixture to within about 7 mm of the top and covered with a 15 mm cover slip. The small culture dishes were used to observe eggs, larvae and protonymphs. To retain adequate humidity a few drops of water were added to the plaster charcoal mixture when needed. Mating pairs of Parasitus initiated cul- tures used for the life history studies of P. favus and P. perthecatus. Three pairs of P. perthecatus, four pairs of P. favus and two pairs of P. hobbsi were collected in a mating attitude. All of these pairs, except one of P. hobbsi and one of P. favus, were cultured. The eggs laid by the females were removed from the culture dishes con- taining the parents and placed in the smaller dishes. The information gathered gave an indication of lengths of certain life stages, but low numbers and lack of temperature control in the field laboratory were variables that could not be controlled. Temperatures of the cultures followed the outside air temperatures except on cold days when the field laboratory was heated. Mating pairs were collected on June 22, 24, 29" july 9, Aucuse 7) 10, a5; anda20: 1971. All pairs collected in June and July were placed in individual cultures. REARING STUDIES The life histories of the Parasitus spe- cies associated with bumblebees are typi- cal of most Parasitidae. The mites pass through the stages of egg, larva, proto- nymph, deutonymph, and adult. Mating: In each case observed the fe- male was on top and the male underneath, their ventral sides together. The male used the apophyses on his second pair of legs to grasp the female’s fourth pair of legs’ The ‘male's ‘third’ pair “or legs wrapped around the female’s opisthosoma, and the fourth pair curled up around the posterior end of the female’s body but were not clasped tightly. The female car- ried the male around in this position seem- ingly unhindered by his presence. If dis- turbed, the pair quickly separated. Five days after a pair was placed in the culture, a male of P. favus was seen trying to copulate with the female, indicating that mating may occur more than once. Costa (1964) noted for P. copridis Costa that males mate several times with the same female. On one occasion a male was seen trying to mate with a deutonymph. Possibly this was a female deutonymph near the time of final molting. The males of P. coleoptratorum molt first and pur- sue the female deutonymphs, which upon molting are immediately mated (Rapp 4 Tue University oF Kansas ScrENCE BULLETIN 1959). Newman (1943) states that males of Poecilochirus necrophori Vitzthum, a closely related genus, copulate successfully only with newly molted females. Eggs: In this study 0-8 eggs were laid per day by females of P. perthecatus, and 0-1 egg were laid per day by P. favus. The eggs were laid under the pieces of pollen and wax, in the cotton nesting material from bumblebee nests, on the plaster of Paris surface, and rarely on the side of the plaster container. Rapp (1959) reports P. coleoptratorum females to lay one egg every one to two hours, with each female laying between 100 to 130 eggs. Eggs of P. favus and P. perthecatus were oval in shape, 0.39-0.44 mm long, and no differ- ence was noted between the eggs of these species. In three cultures of P. perthecatus, 26 eggs were seen; 10 hatched into larvae an average of 3.9 days (range 3-6) after being laid. In one P. favus culture, four eggs were laid; one hatched after five days. Mating pairs collected in August did not lay eggs, and lived an average of only eight days after being collected. At this time in mite phenology, few deuto- nymphs are molting since they quickly be- come phoretic in the nests. Larvae and protonymphs: Two P. per- thecatus larvae, one three days and the other six days old, molted into proto- nymphs and each lived an additional six days. The single P. favus larva molted after six days into a protonymph, which lived an additional two days. Deutonymphs: Deutonymphs overwin- ter on bumblebee queens and molt, becom- ing adults after the queen establishes a nest. Thirty attempts were made to in- duce deutonymphs collected from the spring queens to molt, without success. The deutonymphs lived in culture up to 61 days. Possibly these deutonymphs did not molt because of the type of food of- fered, the temperature fluctuations, the hu- midity, the lack of some stimulus provided by the bees themselves, or combinations of these factors. Adults: Since no deutonymphs molted, the only information on adult longevity was obtained from adults collected from nests (Table 2). They survived from 3 to 12 days. This gives an indication of mini- mum longevity, if the mites’ survivorship was the same in the cultures and in the nests, and if the mites collected were on the average the same age. No difference was found between the longevities of males and females for P. hobbst, favus, and perthecatus. For comparison of known life history see Table 1. Lengevity on Different Foods: Larvae and protonymphs in culture dishes were offered the pollen-wax mixture used for construction by bumblebees. One larva and five protonymphs were collected from bumblebee nests, eleven larvae were reared from eggs laid by females of mating pairs, and 13 larvae from single females. Larvae and protonymphs were never observed feeding. Larvae lived an average of 3.8 (range 1-8) days. Six of twenty-five larvae molted into protonymphs. Protonymphs lived an average of 4.3 (range 1-10) days before dying. Taste 1. Durations of life stages in days for Parasitus species found on the beetle Copridis hispanus (L.), on other beetles, and on bumblebees. Parasitus species references egg larvae protonymph deutonymph EE PUCO PIVOTS ees ec ercee ss Costa, 1964 1-2 1-2 79 + P. coleoptratorum ........ Rapp, 1959 2 3 5 10+ Pit Rerihecatus eA 3-6 3-6 ? Te LOGS ae 2 She he 5 6 P Parasitip Mires AssociATED WITH BUMBLEBEES 5) Taste 2. Longevity in culture of males and females of P. hobbsi, P. perthecatus, and P. favus taken from Bombus nests. avg. no. days Parasitus survived in species no. cultured culture range in days 2 3 Q 3 Q é ee ODOST 2222 ee 1 if 10 WOME ULES: SAN. I es ve. 4 i 6.75 6.75 3-11 2-12 PrePETiReCCULUs |... 1.2. 6 3 To 4-10 3-10 Deutonymphs were offered a number of possible foods with the results indicated below: Pollen and wax—lived 5-61 days (x= is; u—20)- Pollen of Erythronium grandiflorum Pursh—lived 6-17 days (x=13.3, =) Yeast—lived 6-25 days (x=13.0, n= 13). Once a deutonymph was seen feeding on yeast. Acarid mites and yeast—lived 7-24 days (x=14.3, n=6); four P. favus and two P. perthecatus. No response to- wards acarids. Dr. R. Husband ob- served a Parasitus deutonymph with an acarid mite in its chelicerae (personal communication, 1972). Adult bumblebee pieces—lived 1-46 days (X=22.0, n=8). The mites congregated around the fresh mus- cle. Bumblebee larvae—no response. Honey—no response. In two cases, deutonymphs were seen eating another deutonymph, and once eat- ing eggs laid by a female Parasitus. Adults were also supplied with several possible foods with the following results: Pollen and wax—lived 2-36 days (x= 9.0, n=36). Several times, mites were seen with pieces of pollen and wax in their chelicerae. The pieces of pollen-wax became pock-marked after being in the cultures for a few days. The mites did not feed on moldy pollen-wax. Costa (1964) noted that P. copridis would stay in mold-free areas in a culture, and that the copris beetles inhibited growth of mold. If mold grew in his cultures uninhibited, the mites died. Acarid mites and yeast—lived 9-12 days C—1005n—35). Bumblebee larvae—no response. Females were twice seen eating their own eggs. Females of P. coleoptratorum were reported by Rapp (1959) to fre- quently eat their own unfertilized eggs. In my study one male was seen eating an- other male; one female was seen eating another female, and one female was seen eating a Parasitus larva. Rapp (1959) states that younger stages of P. coleoptra- torum are easy prey for older stages. A female was observed feeding on a chironomid larva as she was being col- lected from a nest. However, chironomid larvae were not common in nests, and are not important as a food source. When these mites approach the possi- ble food, the first pair of legs are waved in front and contact the food first. Experi- ments conducted by Rapp (1959) with P. coleoptratorum on the Haller’s organ of tarsus I and on sense organs of the last segment of the palps show both to have a positive response to odors. Parasitus use a variety of food in the bumblebee nests, as indicated from the above observations. Although mites fed on bumblebee muscle in culture, it is not known if the mites actually feed on the 6 Tue University oF Kansas ScrENCE BULLETIN adult bees in the nest. Frequently mites crowd on a bee, especially the queen, early in the season after she has returned from a foraging trip. The mites might be feeding on nectar (Stebbing 1965) or on pollen. However, none congregated anywhere in the nest, not even around honey or pollen stores and there was no response to honey placed in the cultures. Therefore, it seems unlikely that they would be attracted to a bumblebee because of honey stuck to her pile. A number of other stimuli might at- tract the mites to the bee, such as increased body temperature, pollen, COz, and move- ments. Plath (1934), Skou, Holm, and Haas (1963), and Chemlewski (1971) sug- gest that Parasitus species feed on excre- ment from bumblebees and their larvae. P. coleoptratorum, a strict carnivore, feeds invertebrates on other mites and small number of nests 4 7 3 14 24 18 11 100 pa o [os] (=) fo) ° mean number of mites i) °o age of nests in days Fic. |. Population fluctuations of P. favus from all Bombus nests. o—o. found in dung (Rapp 1959). Parasttus copridis is reported (Costa 1964) to feed on nematodes and possibly collembolans and other small arthropods. Kuhnelt (1950) reports the Parasitus species in soil will feed on nematodes and small insect larvae. Since a variety of small arthropods are found in bumblebee nests, the associ- ated Parasitus might use many of them for food, although only two direct observa- tions verifying this have been made. All life stages of parasitiform mites usually use the same type of food (Hirschman 1959), so it is possible these species of Para- situs also use the same food in all life stages. Longevity varies among mites fed the same foods. The averages indicate that deutonymphs may do better on muscle, or on pollen and wax, but none did well enough to molt. Some mites may not have fed, but survived on food reserves left over from winter or built up before being col- lected. Perhaps some stimulus from the number of nests 24 27 31 30) 19) 308 35 i7a56 Ti me, ol 100 a o te) °o °o °o mean number of mites nN ° 20 40 60 100 age of nest in days deutonymph ——-, male -—-, female Parasitip Mires AssociATED WITH BUMBLEBEES 7 bumblebee causes the deutonymphs to start feeding and to molt. The deuto- nymphs were able to survive longer in cul- ture than any other stage of Parasitus. Since the deutonymph is the _phoretic stage, it must be able to survive on the bees, presumably with no food, from the time the new queens leave the nest, through hibernation, until the following spring when the queens have established colonies. No significant difference in life spans between adults given wax and pol- len and adults given yeast and acarids was found. Since adults were collected from nests, food reserves obtained in the nest may have been adequate to produce the eggs laid in culture. An inadequate food source may have been the reason these mites laid eggs irregularly and in such low numbers as compared to P. coleoptrato- rum. No life cycle was ever completed on any of the foods tried in rearing tests in this study. number of nests 12 22 15 9 4 2 3 30 NEST STUDIES Noting the ages of the Bombus nests and the condition of the deutonymphs in them (assuming a fully engorged deuto- nymph is near the time of final molting), studies of populations of Parasitus suggest that P. favus and P. perthecatus have more than one generation per year. A deuto- nymph collected 22 July 1970, had the adult structures of a male P. perthecatus visible through the deutonymph cuticle. The collection site, a B. occidentalis nest, contained 103 workers, 19 cell eggs, 677 larvae, and 190 pupae. From a second nest inspected on 17 July 1970, fully en- gorged deutonymphs and a female of P. favus with the deutonymph’s dorsal shield still on her were collected. In another nest which had already produced bumblebee sexuals, a deutonymph of P. perthecatus was collected while molting to become a female. Thus new adults are produced during the summer as well as spring when the overwintered deutonymphs mature. Studying the age composition of the Parasitus populations from nests of differ- number of nests 20 25 29 27 16 29 28 16 5 eel alle ec plea | Oo 30 1971 ! = | £ | oe 20 © i | ° = : AM = € | CaN / 3 « 20 / | 1e) - \ j E ¢ 7 hoa =) © / ! = ie} L / \1 [S 10 € - aie \ ® c / 1Ots fh E 5 vo ‘e 10) age of nests in days age of nests in days Fic. 2. Population fluctuations of P. perthecatus from all Bombus nests. deutonymph ——-, male -—:, female o—o. 8 Tue Universiry oF Kansas ScreENcE BULLETIN ent ages gives some additional information on the duration of the different life history stages. The youngest bumblebee nest col- lected was 11 days old (from establishment by the queen to date of collection). Adults of both sexes of P. perthecatus were pres- ent at this time. A 12 day old nest con- tained adults of P. favus. Larvae of Para- situs were encountered in a 14 day nest but not in the two younger ones; however, lack of larvae observed from younger nests may be due to the small sample size. Protonymphs were found in 16 day nests, and occurred in 62.5°% of the nests (24) 16 to 22 days old. Thus Parasitus deuto- nymphs must molt, become adults, mate, and lay eggs within at least 14 days. Since a minimum of six days is required (Table 1) in culture from egg deposition to proto- nymphs of P. favus and P. perthecatus and number of nests 2 SD? 18)\+8 mean number of mites 20 40 60 80 age of nests in days as much as twelve days may be needed, it is likely that the overwintered deuto- nymphs molt before the Bombus nests are 11 days old. Such would be necessary in order to have protonymphs in the nests as early as 16 days after their establishment. One might speculate that deutonymphs molt into adults within 7 days after nest- establishment. By this time the bumblebee queen has rearranged the nest material, constructed the first brood cells, laid a full complement of eggs, and constructed and partly filled a honey pot. The length of time from nest establishment to deuto- nymphs’ molting may be similar to the duration of the summer deutonymph stage. Thus we can estimate the total length of the summer life cycle from egg to adult to be 20-26 days for these two spe- cies of Parasitus. Population-curve analyses are difficult because the exact life cycle data are not available, the movement of deutonymphs in and out of the nest on workers and number of nests 1 6 15 100-9) S192 Seel 2a 10 b 1971 iF ° 4 \ \ \ \ : \ 2 NY 7 _—- Se / mean number of mites — ae - ue Jee 20 40 60 80. 100 age of nests in days Fic. 3. Population fluctuations of P. hobbsi from all Bombus nests, deutonymph ——-, and -—-, female Parasitip Mires AssociATED WITH BUMBLEBEES 9 queens was not measured, and the number of mites initiating the populations is not known. The low numbers of P. hobbsi and P. inguilinobombus make the inter- pretations of their curves less meaningful, since the sample size is smaller. The data are based on nests up to 80 days old in 1970 but were continued to 100 days in 1971. The population curves for P. favus (Fig. 1) show an increase in deutonymphs at least up to 80 days after nest establish- ment in 1970 and 100 days after nest estab- lishment in 1971. This species may be lim- ited by the length of the season. Some colonies of B. californicus, B. appositus, and B. rufocinctus, the last three bee spe- cies to emerge and establish nests in 1970, were producing queens as early as 43 to 49 number of nests mean number of mites 10 20 30 40 50 60 70 80 age of nests in days 1Ei 3 days after nest establishment. Since the increase in deutonymphs occurs at the same time as the early queen production, most of the deutonymphs must become phoretic at this time. Therefore, the deu- tonymph population increases while the adult population remains more or less con- stant. The population curves for P. perthe- catus (Fig. 2) are similar to those of P. favus. The population fluctuation curves for the relatively rare P. hobbsi (Fig. 3) and P. inquilinobombus (Fig. 4) in 1971 both show a decline in males and deutonymphs near the end of the summer (late popula- tions were not sampled in 1970). This de- cline seems strange since overwintering is in the deutonymphal stage. P. fimetorum was not found in nests less than 40 days old in 1970, but was col- lected after 20 days in 1971 (Fig. 5). The lack of observation of this species earlier in the season in 1970 may have been sam- number of nests 24 3 7 10 14 7 6 mean number of mites 20 40 60 80 100 age of nests in days . 4. Population fluctuations of P. inquilinobombus from all Bombus nests. deutonymph ——-, male -—:, female o—o. 10 Tue Universiry oF Kansas ScreENcE BULLETIN pling error. However, P. fimetorum is not phoretic on bumblebees. Production of deutonymphs increased at the same time as that of P. favus and P. perthecatus. Per- haps high deutonymph production is a function of time rather than of bumblebee life cycle; or the phoretic host if any of P. fimetorum may be available at the same time as bumblebee queens. P. fimetorum was more common than P. hobbsi and P. inquilinobombus (Table 3). Nests occupied by bumblebees the pre- vious season had no Parasitus even though old cells and food pots still contained po- tential food and protection. SPECIBICILY All five species of Parasitus occurred in number of nests See 1a 77, nests of all the species of Bombus in the study area, except for those species of Bom- bus for which only one nest was available for inspection (Table 3). Parasitus favus and P. perthecatus were identified from all bumblebee species collected in the field, except for those Bombus species of which five or fewer individuals were collected with Parasitus on them. Thus, P. favus and P. perthecatus show no precise restric- tions as to Bombus species. Because P. hobbsi and P. inquilinobombus are not as common but do occur in nests of all Bom- bus species, and have been identified from more than one species of field-collected queens, I assume that they also are not re- stricted by Bombus species (Table 4). Randomness of distribution using the Poisson formula was calculated for the genus Parasitus, using only field-collected number of nests 2 4 10 9 S§ Ohi 16a0225 40 wn ® = Ss O » 30 L 2 © € 2 20 <= Ss ° =) S o Q = = 20 5 =) o = £ = 5 10 a = 10 : Oh eee: = oO Oo 40 t ee 20 40 60 80 100 age of nests in days age of nests in days Fic. 5. Population fluctuations of P. fimetorum from all Bombus nests. deutonymph ——-, male ——-, female (O—.0F, 1] Parasitip Mires AssociaTED witH BUMBLEBEES LE} eks ha | Sr Ol CUSE CE 0179 6'°S6 OUiea eee ISLIIAY Pepe cae ears [PIO], 00'T 0¢0 00°F €C'6I 00'T 99°CZ O01 Coy See suvdva “g 00'1 WEE O8'ST 67 09 18°C £978 OOL | SRUMDULI °F Ol O10 OFZ bS'81 060 8627 6 $8 VO Sone SNIIUIIO{NA “J Orb OTC OS'8I €TSS £8°8 O68 16 Lo” ae SUVvIUNAPIIIO “FT 0) 0 002 00'0S 002 00'F9 OOT ree ee SISUgpvAIU “T SS'vI 08°0 £19 62°87 L6o'I 9€¢°7S 76 0G O Seraa Sngxiul “g 0) 0 00'¢ 00'E () 00'9 O0T I > snd Adounjau +g 00°T 00°8 0S'7 OO Eb 0S 9 00°19 OOT a a. eae SHUNY Ors I by y cZ'8 OS'ZS GLE, bO'ZOL G'S6 Shih ee ae ae snpisiaf “gq €8' 07'0 EL bl 06°6€ bE'7 Eb'6S 001 Scape rs suodfiavy “g ) 0) 00'T 0 ) 00'T O0I l 2 eee SyvAquad “g 88'S 070 OF TT v9'CC 00°C IC cr S96 eS SNIMAOfIVI “J b0'F 06°C 88°F 9¢° CP b0'S C769 S'b6 Cie Sa smirfig “gq S6'r cS'0) Sb'6 OS TE €9'| ZE8b C98 a snyisoddy “g wnsojauy snguogouyimbut snyivoay.sad snav{ isqqoy [e303 SNISDAD porios sotoads ‘d ‘d ‘d ‘d ‘d yim 9% [e101 snquiog ysou sod SNISVAD jo Joquunu ues sqsou “\sou snqQuog Jad satoads SNUISVADE jo Joquinu UvIT “¢ ATV [ Tue Universiry oF Kansas ScrENcE BULLETIN 0S0 (ENV) 601 C 99 BOY? COR Cee eS at [vI0 |, 0 cr 0 € 0 ¢ 8] () 9°77 (et Ea gre Wa]yons “dq £0 aa) 0 I II OZ I t'6l C6 Soe ae Ipppusrl “dq JO S1ipjnsut *¢d C0 seal () () I € 0) 9°09 Fo. CS DIO * a0) S70 0) 0 0 ¢ 0) $'SZ Cl Be ree SMIADUAIL “FY (G10) 7a) b 0) I 9 () (07 Chime wr ear SNIIUIIO{NA “J OS0 82Z'1 Al 0) eT] 9L I OTE OO her sypquap120 *g (610) Or 0 0) 0) I | () 0'0Z C, rr ae Sisuapvaau *g () () 0 0 0) ) () 0) i li ir a ia Snqxtu “g 1-0 0S’) () () () I () 0'0S Ce ae eae sndidouvjauw *g C0 S20 () 0) c 0) I 00S Vin tae pe ee muny “g 60 9F'0 8 () 17 b () V'Z1 Sie) ei aaaT od snpisis{ “g 10 600 () () I () 0) 16 lil =. heen suowfiavyy “g 80 £90 Il 0 ¢ at 0) Cie | FS aes SNIUAO{NVI “J c1-0 ale] ZL [ 17 9¢ 0 LEZ Pe ate oe sniavfig “gy 10 ITO b 0 | 0) 0) +07 1 anaes oe aa snqyisoddy -q asuri usonb sod peyn snquiog SNIDI snav{ isqgoy syduikuoynap pardaqjoo saioads SNIISDAD -uaprun = -ounimbur:g —-ayssad “gq ma : yam 9%, ae snquog 10 Jaquinu wesw sydurduoynap Jo raquinyy susonb suguog pur snadysisg snahyIs ‘susonb saqayquing Sutids wiosy payayjoo sydurAuo nap sugisving *} ATAV J, Parasitip Mires AssociATED WITH BUMBLEBEES 13 queens not carrying a pollen load and field collected workers. When queens and workers were combined the C. D. (Coefh- cient of Dispersion) was 3.798, and when only queens were used the C. D. was 3.372. Thus, there appears to be a tendency for the deutonymphs to congregate on queens. Even among the queens some clumping is indicated. The numbers in Table 4 indi- cate a tendency for P. perthecatus and P. favus to be more numerous on B. occt- dentalis and B. bifarius queens, while in the nests Parasitus seems more abundant with B. frigidus, B. occidentalis and B. bifarius. PHORESY AND DISPERSAL The deutonymph is the phoretic stage in Parasitus. Deutonymphs were found on the bee’s metathorax and propodeum. They attach themselves to the base of the pile with their chelicerae (as do other phoretic deutonymphs of Parasitus on other insects (Kihnelt 1950)). If their numbers are great, they will also crowd around the neck. The mites are easily ex- cited, and if disturbed will run over the entire body, but seldom leave the bee. The importance of phoresy to acarine evolution is indicated by its appearance several times in three major groups: Meso- stigmata, Prostigmata, and Astigmata (Cross and Bohart 1969). Phoresy allows mites a means of attaining new resources when such resources are widely dispersed. Dispersal occurs before mating. As Mitchell (1970) determined for any bisex- ual mite dispersing before mating (assum- ing equally numerous sexes and equal probabilities of dispersal), four mites of a species must meet 87.5% of the time to include one of each sex, assuming that males and females are necessary to start a population. The sex ratio for the Parasitus species was determined for adults in Bom- bus nests (Table 5). In none of the species was a 1:1 ratio shown. When both years are considered, all species of Parasitus show more females than males. A higher percentage of females would increase the number of eggs laid, and thus increase the number of deutonymphs produced for dis- persal. With females more numerous than males, the number of four mites necessary to include both sexes 87.5°% of the time would have to be raised. Only 19 of 112 (17°%,) queens caught in the spring which were carrying Parasitus deutonymphs had four or more mites, 59.894, of these spring queens carried only one Parasitus deuto- nymph. Seventy-eight percent of the 434 nests sorted had two or more species of Parasitus. These two facts indicate that the mites do not depend entirely on the establishing queen for dispersal. The highest number of Parasitis deuto- nymphs found on one bee in the study area was 50, but there is a report of 106 on a single bumblebee queen (Skou, Holm, and Haas 1963). A B. sylvicola Kirby queen sent to me from McConnell River Northwest Territories, Canada, collected by K. MacInnes on June 18, 1970, had 104 P. favus and 10 P. perthecatus. A B. po- Tase 5. Sex ratio for adult Parasitus species in Bombus nests. Parasitus 1970 1971 total species no-mutes “ ¢ no. mites Ye ae no. mites % ODD Ste 3 es ee 337 54 842 42 1179 45 A INCLORUT. oie = oe 337 48 489 32 826 39 Pe GUA es WOT Ae 1388 54 2462 40) 3850 44 LE RPETERCCALUS Pion ots! 353 Sy) 675 36 1008 40 P. inquilinobombus .......... 110 32 298 34 407 33 14 Tue Universiry oF Kansas ScrENcE BULLETIN Tasce 6. Parasitus on Bombus and Psithyrus females removed from established Bombus nests in 1970, 1971. °4 of Bees with Total No. of Parasitus Species No. of Bees Parasitus from all queens PS OPI TES Se tet ee eee ae 34 2.65 1 2. LASTS ee 246 9.76 37 PEORSECR CVE? Wee eae, Sat: 43 9.30 - A erpGldae® x:..42 5500 oes es 21 0 0 laris Curtis queen from Devon Island, Northwest Territories, Canada, collected by J. R. Ryan, had 211 P. perthecatus and 210 P. favus. In southern Alberta the aver- age number of deutonymphs found on bumblebees having Parasitus is 3.2 (range 1-50) for spring collected queens and 3.0 (range 1-10) for field collected workers. Only one of the 99 Bombus males col- lected in the field had Parasitus deuto- nymphs. Clearly one must seek alternative means of introduction of mites into Bom- bus nests. Deutonymphs occur on the bees and in the nests throughout the season. Generally they do not stay on a queen long once she enters a prospective nest and _ starts to rearrange nesting material. Twenty-four queens carrying pollen were collected (in- dicating they had already established a nest), but only two of these (8.3394) had phoretic Parasitus deutonymphs. Also con- specific queens and Psithyrus removed from nests had few mites (Table 6). However, deutonymphs do leave the nest after establishment. Not only were queens with pollen loads, Psithyrus and conspe- cific queens from nests sometimes carrying deutonymphs, but also 8.05°% of the work- ers collected in the field had attached deu- tonymphs. A deutonymph on a bumblebee that is actively gathering pollen and _nec- tar could be left on a flower and picked up later by another bumblebee. Parasitus deu- tonymphs were collected from flowers of Hydrophyllum capitatum Douglas on the fourth and sixth of June 1970, and Arcto- staphyllus uva-urst (Linnaeus) Sprengel on the 27 May 1970, both in the area for- aged by bumblebees. Stebbing (1965) re- ports a Parasitus deutonymph being acci- dentally collected on flowers which were to be used for pollen slides. Banks (1919) reports P. bomborum (syn. of P. fucorum) from Bombus and from catkin of Salix in the Northwest Territories, Canada. Other flowers commonly visited by bumblebees, 100 Taraxacum officinale Weber, Ane- mone patens Linnaeus, and 130 Dodeca- theon pauciflorum (Durand) Greene, were collected and searched for deutonymphs, but none were found. Possibly the length of time a mite remains on a flower, before a bumblebee revisitation, is short. Thus mites not on the original nest establishing queen could be introduced into the nest. Worker bees in the study area were observed to enter nests other than their own on several occasions. If these workers were carrying deutonymphs from their original nest, they could be added to the population in the new nest. Psithyrus, an obligate parasite of Bom- bus, carries the same species of mites. Pst- thyrus females invade the nests of the bumblebees and lay their eggs, later to be reared by the bumblebee workers. Psithy- rus emerges from hibernation in late May and June and seeks out bumblebee colo- nies. They often enter and later leave nests with few or no workers, but they may pick up mites and carry them to another nest, or may leave mites. A Psithyrus fe- male attempting to enter a Bombus nest with workers must be accepted by the col- ony before she can stay. She may enter Parasitiy Mires AssociaTED WITH BUMBLEBEES iS several times for brief periods of crawling around the nest or taking nectar from the honey pot. Such a Psithyrus has been ob- served going into a nest and remaining as long as 10 minutes, or may even spend the night in a nest which she does not parasi- tize (Richards 1928). If she remains in the nest, her load of mites will be added to those already in the nest. Even if she leaves the nest, the number of brief visits allow the deutonymphs time to get on or off her. More than one Psithyrus female may find the same Bombus colony. In two cases 14 Psithyrus females were taken from single bumblebee colonies in 1971. At the end of the season deutonymphs may leave the nest both on new Bombus queens and Psithyrus females. Also probably important in introduc- ing mites are the excess of Bombus queens. Hobbs (1965, 1967b) describes incidences of superseding queens. As in the case of Psithyrus, mites might leave these queens and become part of the nest’s population, attach to a queen and be carried away if she leaves. Some bumblebee queens have been ob- served to enter a domicile, camouflage the entrance, rearrange the nesting material and drop their mites in the nest. Shortly thereafter such a queen may abandon the nest. Abandoned nests, containing mites, can later be reoccupied by other bumblebee queens which then may be parasitized. During May and June, 1971, twelve such abandoned nests were investigated. The number of Parasitus in these nests ranged from one to fourteen and averaged four per nest. Four of these nests were later oc- cupied by other bumblebee queens, which established broods. Nests are also some- times abandoned after some brood devel- opment and in some cases these same nests were used by another queen to rear her brood. If the time period is not too great between successive brood establish- ments, the deutonymphs from the first queen could survive to become part of the population of mites in the colony of the second queen. During brood-decline, some unknown factors (i.e., decreased worker population, increased temperature, humidity fluctua- tions, appearance of sexuals, type and amount of food being foraged, or combina- tions of the above) cause the deutonymphs to attach enmass to the bees which leave the nest. In declining nests, mites are seen crawling over the cotton through which the bees will crawl in leaving the nest. Deutonymphs use the first pair of legs to locate the bumblebees. The deutonymphs crawl quickly up the leg of any bee con- tacting them. They more or less indis- criminately attach to workers, males, and queens, but there is some tendency to con- gregate in greater numbers on queens. The percent of the total number of bum- blebees collected from fall nests with Para- situs is 15.9 for queens, 8.66 for workers and 12.45 for males. In contrast, 22.16% of the spring queens collected were infested and only one in 99 field-collected males had deutonymphs. Possibly, queens col- lected directly from nests had not yet ob- tained deutonymphs. When a queen with no mites was placed in nests where deuto- nymphs were actively searching for hosts, seven to ten mites crawled onto the queen within 30 seconds. The difference between numbers of mites on field collected and nest males indicates that deutonymphs at- taching to males are repelled by some stimulus and leave, either while the males are still in the nest or when visiting flow- ers. Deutonymphs that remain on workers and males die when these bees die at the end of the flight season. Mites remaining in the nest also die. Thus selection would favor a mechanism by which the deuto- nymphs recognize queens. Many insect associates are found in bumblebee nests. These insects may be eaten by other mites in the nest and may ANSAS SCIENCE BULLETIN THe UNIVERSITY OF 16 SGLINW HIIM LSAN NI SdSITAVLSA N4aNG GNOOUS N@Q — HO + LSAN YALNA SNHANSB SOHWOH HO SNHAHLISd Sa, gonaouddau GNV ———t;>-- 440 L49 NG NQ@ + LS4N YHAHLONY WOU WANHOM ONIONGOud AY SALIN qaondgdoud SYaWuOomM NG — LSAN INGYaAAIC OL HANHOM on LSAN WOUd/OL SaINUVO a ISN SHAHSITAVLISA Naan SHAVAT NAGANG ae Nad — JLSaAN GFAILIAdSOUd YELNA NAANS YAMOTA LV NG — ¥O+ SNaqNG GASHa WA NG HLIM SNAAND NV SHAMOTA LY Na — HO + SUaADNVHOU ee HLIM SNAAND NOILYNUAGIN qgaondoud SaIVW qqao0nqaoud SNA qT010 Nd + SNAANG qid S¥awyvom uO SATIVW NO LSAN NI SOLIW ) by a Bombus nest. Various mechanisms by which mites may be gained (--) or lost ( IGaGe E Parasitip Mires AssociaTED WITH BUMBLEBEES 17 influence the population of all the mites by providing a food source or introducing predators or competitors. A summary of the various mechanisms by which mites may be gained or lost by the population in a nest is given in Fig. 6. P. fimetorum was found in 26.194 of the nests sorted. It has not to my knowl- edge been reported previously from bum- blebee nests. Holzmann (1969) studied the deutonymphs and notes that they can live up to six months on little food, are quick moving, and have been found on dung-beetles of the genera Aphodius, Geo- trupes, and Hister. ‘This species is known to be common in Europe. It was collected in the soil in Poland by Micherdzinski (1969) and in the Swiss Alps by Schweizer (1949). Costa (1963) reports collecting it from nests of voles and on Copris his- panus (Linnaeus) (109 deutonymphs on one beetle) in Israel. It is not known how P. fimetorum gets into the bumblebee nests in Alberta. Perhaps they wander in from the soil or are carried in by beetles which enter the nests. They were never collected on bumblebees, and were not found in six soil samples. They have been collected in above ground domiciles, which suggests that at least some are carried into nests by some insect. They were more common in surface nests than in subterranean or above ground nests, and were not common in any of the five natural nests (four of which were underground and one on the sur- ace) SCUTACARUS ACARORUM ON PARASITUS During field observations and later dur- ing studies of mite species composition from nests, a small mite, Scutacarus acaro- rum (Goeze, 1780), was often seen at- tached to Parasitus. This species has been previously reported from Bombus by Cross and Bohart (1969), Husband (1968), Kara- fiat (1959), and Kihnelt (1950). Hughes (1959) shows a picture of a Parasitus fucorum deutonymph with a scutacarid attached to its leg. S. acarorum is known from Germany, Italy, England, Mongolia, North America and South America. Cross and Bohart (1969) state that they become progressively less frequent on all Bombus species at lower latitudes, but Husband (personal communication, 1974) says this is due to a paucity of bumblebees at lower altitudes, both in numbers and species of the bees. These mites fasten themselves to their host with the claw complex on the first pair of legs. In southern Alberta these mites were commonly found attached to the deutonymphs of Parasitus favus and P. perthecatus, but a few were also found on males of P. favus, females of P. perthe- catus, deutonymphs of P. hobdst, and fe- males of P. inquilinobombus and one on a macrochelid male. An S. acarorum was also found on a field-collected B. occiden- talis queen. S. acarorum was usually found attached to the setae of the legs, but also attaches to setae on the dorsum of the deutonymphs. S. acarorum was found on Parasitus spp. in 179% of 424 of the bum- blebee inests: In these nests, 247, (62 of 258) of the deutonymphs of P. favus and 33.99% (21 of 62) of the deutonymphs of P. perthecatus had S. acarorum attached to them. The number of S. acarorum on deutonymphs ranged from 1 to 13. A female of S. acarorum attached to a deuto- nymph of Parasitus in culture remained attached until the deutonymph died. It then crawled off and remained in the cul- ture eight additional days. Scutacarus uses the Parasitus and the bumblebees phoreti- cally. If they use the host for anything else, it is not known. Only females of S. acarorum were encountered. LITERATURE CilED Banks, N. 1919. Report of the Canadian Arctic Ex- pedition, 1913-1918. vol. Ill. Insects, Part H. 18 Tue University oF Kansas ScrENCE BULLETIN Spiders, Mites and Myriapods. Ottawa, Kings Printers. H: 12-13. CHEMIELEWsKI, W. 1971. The mites (Acarina) found on bumblebees (Bombus Latr.) and in their nests. Ekol. pol. 19 (4): 57-71. Costa, M. 1963. The mesostigmatic mites associated with Copris hispanus (L.) (Coleoprera, Scara- baeidae) in J. Linn. Soc. Lond. Zool. 45: 24-45. 1964. Description of the hitherto unknown stages of Parasitus copridis Costa (Acari-Meso- stigmata) with notes on its biology. J. Linn. Soc. Lond. Zool. 45: 209-222. Cross, E. A. AND G. E. Bounart. 1969. Phoretic be- havior of four species of Alkali bee mites as influenced by season and host sex. J. Kans. ent. Soc. 42: 195-218. Free, J. B. anp C. G. Butter. 1959. Bumblebees. Collins, London. 208 p. Frison, T. H. 1926. Contribution to the knowledge of the interrelationships of the bumblebees of Illinois with their animate environment. Ann. ent. Soc. Am. 19: 203-233. HaAsseLrot, T. B. 1960. Studies on Swedish Bum- blebees (Genus Bombus Latr.) their domesti- cation and biology. Opusc. ent Supp. 17. 192 p. HirscuMann, W. 1959. Gangsystematik der Para- sitiformes, Teil II, Hypostom bestimmungsta- feln. Furth/Bayrn. 23 p. Hosss, G. A. 1964. Phylogeny of bumblebees based on brood-rearing behavior. Can. Ent. 96: 115-116. ——. 1965. Ecology of species of Bombus Latr. (Hymenoptera: Apidae) in southern Alberta. III subgenus Cullumanobombus Vogt. Can. Ent. 97: 1293-1302. ——. 1966a. Ecology of species of Bombus Latr. (Hymenoptera: Apidae) in southern Alberta. IV subgenus Fervidobombus Skorikoy. Can. Ent. 98: 33-39. 1966b. Ecology or species of Bombus Latr. (Hymenoptera: Apidae) in southern Alberta. V subgenus Subterraneobombus Vogt. Can. Ent. 98: 288-294. 1967a. Obtaining and protecting red clover pollinating species of Bombus (Hymenoptera: Apidae). Can. Ent. 99: 943-951. 1967b. Ecology of species of Bombus (Hy- menoptera: Apidae) in southern Alberta. VI subgenus Pyrobombus. Can. Ent. 99: 1271- 1292. 1968. Ecology of species of Bombus (Hy- menoptera: Apidae) in southern Alberta. VIII subgenus Bombus. Can. Ent. 100: 156- 164. , J. F. VirostEx, AND W. O. Numi. 1960. Establishment of Bombus spp. (Hymenoptera: Apidae) in artificial domiciles in southern Al- berta. Can. Ent. 92: 868-872. , W. O. Numi, AnD J. F. VirosteK. 1962. Managing colonies of bumblebees (Hymenop- tera: Apidae) for pollinating purposes. Can. Ent. 94: 1121-1132. HotzMann, C. 1969. Die Familie Parasitidae Oude- mans 1901. Acarologie 13: 3-54. Huser, I. 1958. Color as an index to the relative humidity of plaster of Paris culture jars. J. Ent. Soc. Wash. 60: 289-291. Hucues, T. W. 1959. Mites or the Acari. Athlone Press, University of London. 225 p. Huspanp, R. 1968. Acarina associated with Michi- gan Bombinae. Papers of Mich. Acad. Sci.. Arts, and Letters 53: 109-112. Karariat, H. 1959. Systematik und Okologie der Scutacaridae. Beitr. Syst. Okol. mitteleur., Acarina | (2): 627-712. KUHNELT, W. 1950. Bodenbiologie. Herold Verlag, Wien. 368 p. MICHERDZINSKI, W. 1969. Die Familie Parasitidae Oudemans 1901 (Acarina, Mesostigmata). Panstowowe Wydawnictwo Naukowe, Kra- kow. 690 p. MitcHetr, R. 1970. An analysis of dispersal in Mites. Am. Nat. 104: 425-431. NewMann, k. W. 1943. Die Lebensgeschichte der Kafermilbe Poecilochirus necrophori Vitzth. nebst Beschreibung aller Entwicklungsstufen. Zool. Anz. 142: 1-21. PratH, O. E. 1934. Bumblebees and their ways. Macmillan, New York. 201 p. Rapp, A. 1959. Zur biologie und Ethologie der Kafermilbe Parasitus coleoptratorum L. Zool. Jb. Syst. Okol., Geog. 86: 303-366. Ricuarps, L. A. 1974. Parasitid mites associated with bumblebees in Alberta, Canada (Aca- rina: Parasitidae; Hymenoptera: Apidae). I. Taxonomy. Kansas Sci. Bull. 50(13): = RicHarps, O. W. 1928. A revision of the European bees allied to Psithyrus quadricolor, Lepeletier (Hymenoptera, Bombidae). Trans. R. ent. Soc. Lond. 11: 345-365. ScuHweizer, J. 1949. Die Landmilben des Schwei- zerischen Nationalparkes. I. Teil: Parasiti- formes Reuter 1909. Ergebn. wiss. Unters. schweiz. Nat. Parks. 2: 99. SELLNicK, M. 1938. Milben als Parasiten von Insek- ten. Verh. II Internat. Kongr. Entomologie, Berlin. 2: 1300-1307. Sxou, J. P., S. N. Horm, H. Haas. 1963. Prelimi- nary investigations on diseases in bumblebees (Bombus Latr.). Royal Vet. and Ag. College Yearbook, Copenhagen. 27-41. SrepBInc, P. D. 1965. A study of the mite Parasitus bomborum. Essex Nat. 3 (4): 284-287. dr. ead e é ee. ye fos = “i 4 Yue ‘ ne ge = ee ‘ , r > ~ ot - - - . = PRR ar eter eane alee ae aeavatetecptecectateseroseressotehataha eters "THE UNIVERSITY OF KANSAS SCIENCE BULLETIN THE-MECOPTERA OF INDIA AND ADJACENT REGIONS oe Pee") 07s" 00"0"0"9 eee ™ one ON Pata eee” OX vate", ™ * Po" 0" 0" aga a"a ea" ae" reer pene, ” , os (eee Ceaueeeugea ei a°e°n eta aa me =m CetesetcRetoneSHinseceecrcecs cat tatesetscasecoscceareeass Se WSR By MICHAEL K. RUST and GEORGE W. BYERS Vol. 51, No. 2, pp. 19-90 September 17, 1976 RRS nA Re 9 RS a J wate a’ a ae ae ea? Bs x” »*, = AA e, ae’ 5 P.°, xy fe » onan * o,4 at Gr et ase Ge", ANNOUNCEMENT The University of Kansas Science Bulletin (continuation of the Kansas Uni- versity Quarterly) is an outlet for scholarly scientific investigations carried out at ~ the University of Kansas or by University faculty and students, Since its incep- tion, volumes of the Bulletin have been variously issued as single bound volumes, as two or three multi-paper parts or as series of individual papers. Issuance is at irregular intervals, with each volume prior to volume 50 approximately 1000 — pages in length. Effective with volume 50, page size has been enlarged, reducing the length of each volume to about 750 pages. The supply of all volumes of the Kansas University Quarterly is now ex- — hausted. However, most volumes of the University of Kansas Science Bulletin — are still available and are offered, in exchange for similar publications, to learned societies, colleges and universities and other institutions, or may be purchased at $20.00 per volume. Where some of these volumes were issued in parts, individual parts are priced at the rate of 2 cents per page. Current policy, initiated with volume 46, is to issue individual papers as published. Such separata may be purchased individually at the rate of 3 cents per page, with a minimum charge of $1.00 per separate. Subscriptions for forthcoming volumes may be entered at the rate of $20.00 per volume. All communications regarding exchanges, sales and subscriptions should be addressed to the ExcHaNnce Liprarian, UNIVERSITY oF Kansas Lipraries, LAwRENCE, Kansas 66045. Reprints of individual papers for personal use by investigators are available gratis for most recent and many older issues of the Bulletin. Such requests should - be directed to the author. The International Standard Serial Number of this publication is US ISSN 0022-8850. Editor Eugene C. Bovee Ed:torial Board William L. Bloom Philip W. Hedrick Rudolf Jander Harvey Lillywhite Charles D. Michener Norman A. Slade Henry D. Stone George W. Byers, Chairman THE UNIVERSITY OF KANSAS SCIENCE BULLETIN Vol. 51, No. 2, pp. 19-90 September 17, 1976 The Mecoptera of India and Adjacent Regions’ By MicHatEt K. Rust and Grorce W. Byers TABLE OF CONTENTS 115 PETRY DUB TEINS ~ acetal eo 20 |S INSTROTRIYERVL.y [RUSAWIIAN AE oo oo aes Oe oe a ee ee ee ee eee ee ee 20 |e As SP PBIRar UE aS RII) LY (TDS OTD SS Ee Se i Ee 21 licens lOPOGRAPEY. AND, ©rINCAiby |= -- Soe Sey ee 22 IDISTRIBUSION- OF MECOPTERA IN THE INDIAN REGION) -222--2.c<20<-0..2---22.---ccee0-22eseeecee-ehbeet-ece 24 SWAK@ NONUVAOE IEIE GENUS: BUGCUS, 2s fcc. onazzcecsde 20st le Ae ces 26 IK e57 (EO) Ces Min ev oyts jofere Cee Oe 317.777 1617 1a ere eee ne cn 27 Descuptionor tie Species: Genus Bitjacus <4 ee ee ee D7, SIPAON @NEVAOHEREIES GENUS INGOPANOTD A: xa coon wane cneceen ape eeenen cece cetennetsatececsteceepeeceda-noceecen oe Key to Malesot Indian’Speciesof the Genus Neopamorpa: <2. 222s2 2522s. cccsececteetec ese aetes 38 Key torBemales of Indian Species of the Genus Neopanorpa _........--.-:-:2:--2c-sceeee es ay Descriptionpok Species: (Genus Neopanorpa 2. ee aie 40 IDESCRIBMONBOFLOPECIES a GENUSEQ1OF Pd. a. oo eae sack cose ste oe sencnstece sot gegstoch sen sebzaee esestonnaeces 84 LocaLiTiEs AND HasitTAts oF MECOPTERA IN THE INDIAN SUBCONTINENT .............----------- 88 | Lantiemaieinis (Gin csey: ee e e ee e e nn ee e 9() ABSTRACT Six species of Bittacus, 21 of Neopanorpa and 3 of Panorpa are described, compared, illustrated, and differentiated in taxonomic keys. The following new species are described: Bittacus nodosus, Neopanorpa denticulata, N. echinata, N. gibbosa, N. indica and N. ochrura. Previously unknown sexes of 5 species based on single holotypes are described and illustrated. N. effusa (Navas) and N. furcata (Hardwicke) are removed from Lepto- panorpa to Neopanorpa. The geographical distribution of Mecoptera in the Indian sub- continent is correlated with elevation, rainfall, forest cover and other ecological factors, and zoogeographical relationships are discussed. ‘Contribution no. 1607 from the Department of Entomology, The University of Kansas, Lawrence, Kansas 66045. 20 Tue UNIVERSITY OF KANSAS SCIENCE BULLETIN INTRODUCTION Accounts of insects of the order Mecop- tera of the Indian subcontinent,” primarily species descriptions, have been published oc- casionally over the past 150 years. Of the 21 species so far described, many are still repre- sented by single holotypes or a few syntypes, and little or nothing is known about their dis- tribution. It is possible that the ranges of several Indo-Chinese and Chinese species ex- tend into the Indian region. Five previously unknown species of Neo- panorpa and one of bittacus are described in this paper. In addition, the previously unde- scribed sexes (one female, four males) for five species heretofore known only from sin- gle holotypes are described and illustrated. Two species previously assigned to Leptopan- orpa have been transferred to Neopanorpa. It is our hope that this paper will serve pri- marily as a basis for future research on Mecoptera in the Indian region, especially since our survey has been so limited. In addi- tion to those already mentioned, we have ex- amined several specimens of undescribed species of Neopanorpa and Bittacus that have not been included because the males were badly damaged or only females were avail- able. Three species of Panorpa reported from Tibet, but probably from western China are discussed and illustrated to clarify the orig- inal locality data and to provide more useful illustrations of the types. We have examined an undescribed species of Panorpa from northern Burma, and it seems probable that the range of Panorpa extends into the hills of eastern Assam. On the basis of wing pattern and genitalic characters of both males and females, three of the species of Neopanorpa form a compact group. We have called this the appendi- culata group, after the oldest named species included. The members are: appendiculata, sala, and indica n. sp. Similarly, another four southern Indian species form a closely related group, which we have called the den- ticulata group, after the most abundant in- cluded species. Its members are: denticulata * The Indian subcontinent, as used here, includes Assam, Bangladesh, Bhutan, Sri Lanka (Ceylon), India, Nepal, Pakistan and Sikkim. n. sp., Airsuta, echinata n. sp., and zebrata. The other species of Neopanorpa cannot yet be arranged in useful taxonomic groups. The most extensive collection of Indian Mecoptera is in the Snow Entomological Museum, University of Kansas. Three addi- tional collections from the Indian region have been examined. They are from the British Museum (Natural History), London, loaned by Mr. Peter H. Ward, the Illinois Natura! History Survey, Urbana, Illinois, loaned by Dr. L. J. Stannard, and the Naturhistorisches Museum, Basel, loaned by Dr. W. Wittmer. We also wish to thank Mr. P. Susai Nathan for his help in locating several of the col- lection sites. Unfortunately, we have not been able to obtain on loan any specimens from the vari- ous institutions in India, Pakistan or adjacent countries. We are not aware of the extent to which this has limited our investigation. Support for this study, through grants GB-30837 and GB-7045X from the National Science Foundation, is gratefully acknowl- edged. HISTORICAL REVIEW The first record of Mecoptera from the Indian subcontinent was the description of Neopanorpa furcata by Hardwicke (as Panorpa) in 1825. Westwood’s description of Neopanorpa appendiculata (1846, as Panorpa) was followed by the description of Bittacus indicus by Walker in 1853. Thirty-two years later Gerstaecker (1885) described Bittacus latipennis. Between 1908 and 1935, Navas de- scribed nine additional Indian species, of which seven are still considered valid. He used the generic names Aulops Enderlein and Campodotecnum Enderlein for the species saffusa and effusum respectively. Aulops suffusa Navas was synonymized by Esben-Petersen (1915) with Neopa- norpa nipalica Navas; and Campodotec- num effusum Navas was transferred to Leptopanorpa effusa by Esben-Petersen (1915) (but we consider it a species of Neopanorpa). The other species were Neopanorpa ocellaris (1908, as Panorpa), Tue MecopTera oF INDIA AND ADJACENT REGIONS 21 Panorpa stigmalis (1908), Panorpa davidi (1908), Neopanorpa nipalica (1910, as Panorpa), Neopanorpa salai (1929), and Neopanorpa benacit (1935). Needham (1909) added two new spe- cies, Neopanorpa sordida and N. fenestrata (both as Panorpa). In 1915, Esben-Peter- sen transferred most of the Indian species described as Panorpa into the genera Neo- panorpa and Leptopanorpa. He presented no evidence justifying the synonymy of Aulops suffusa Navas with Neopanorpa nipalica Navas, and he agreed with Navas’ earlier expectation that the female type of P. davidi was in fact a new species, P. guttata. In addition, he described Neo- panorpa cornuta, N. flava, and N. zebrata. In his world monograph of the Mecoptera (1921), Esben-Petersen redescribed and il- lustrated most of these species. Since Esben-Petersen’s monograph, six further new species have been described. These are Bittacus henry: (Kimmins, 1928), Neopanorpa hirsuta (Crampton, 1931, as Panorpa), N. contracta (Cheng, 1954), Bittacus taraiensis (Penny, 1969), Neopanorpa chillcotti (Byers, 1971) and Neopanorpa ramulata (Byers, 1975). The following species are recognized in this investigation. BITTACIDAE: Bittacus henryi Kimmins indicus Walker insularis Esben-Petersen latipennis Gerstaecker nodosus Nn. sp. taraiensis Penny PANORPIDAE: Neopanorpa appendiculata (Westwood ) benact Navas chillcotti Byers contracta Cheng cornuta Esben-Petersen denticulata n. sp. echinata n. sp. effusa (Navas) fenestrata (Needham) flava Esben-Petersen furcata (Hardwicke) gibbosa n. sp. hirsuta (Crampton) indica Nn. sp. nipalica (Navas) syn. suffusa (Navas) ocellaris (Navas) ochrura Nn. sp. ramulata Byers salat Navas sordida (Needham) zebrata Esben-Petersen Panorpa davidi Navas gutatta Navas stigmalis Navas MATERIALS AND METHODS Dissections of male and female geni- talia were preceded by removing the pos- terior abdominal segments from the speci- men and gently boiling them for one to two minutes in water. In several cases, it was necessary to remove the male’s hypo- valves in order to examine the aedeagus. This was accomplished by carefully cut- ting the hypovalves and ninth tergum apart from the basistyles. To remove the genital plate of the female, it was neces- sary to make an incision through the pleural membrane on each side of seg- ments 7-8. The genital plate is usually secured by tough membranous tissue, and care must be taken to prevent damage to it when this tissue is cut. To soften pinned specimens for dis- section or to be placed in fluids, a dilute solution of trisodium phosphate was used. Specimens should not remain in the triso- dium phosphate solution over 24 hours, or the membranes disintegrate and only the sclerotized portions remain. Softened spec- imens should be thoroughly rinsed in water and placed in alcohol. To photograph the wings, it was often necessary to remove them. A relaxing jar was used to soften the specimens, and the in) in) wings were removed with a_ scalpel, mounted between glass slides and photo- graphed. Drawings and body measurements were made with the aid of a camera lu- cida. Considering the variable length of the abdomen due to the amount of ab- dominal contents at preservation, the method of preservation, and position, the recorded measurements of body length are at best approximations. In descriptions of the genital plates of females, total length is measured in ven- tral aspect. In many species, total length is from the anterior apices of the apo- demes to the posterior apices of the arms of the distal plate. The length of the arms is measured from the apices to a line per- pendicular to the axial portion at its junc- tion with the arms’ inner margin (Fig. 70). Length of the axial portion is from posterior apex to apices of the apodemes. The length of the apodemes is measured along the mid-line from the anterior edge of the basal plate to the apices of the apo- demes. The total length of the male genital bulb in ventral aspect is the distance from the base of the bulb to an imaginary line extending through the most distal points on each dististyle (in a closed position). Hypovalves are occasionally compared to length of basistyles, which is measured from the base of the basistyles to a line connecting the most distal points on both basistyles. Approximately 450 specimens were ex- amined. Most were from South India (Nilgiri Hills), and collections from other parts of the Indian region were small and from scattered localities. Of these 450 specimens, 284, 24, and 39 represented Neopanorpa denticulata, N. hirsuta, and N. appendiculata, respectively. The re- maining approximately 100 specimens rep- resented about 20 species. Tue UNIVERSITY OF KANSAS SCIENCE BULLETIN THE INDIAN TOPOGRAPHY AND CLIMATE The Indian subcontinent has a tre mendously diverse topography, ranging from sea level to the highest mountain ranges in the world, and climatic pattern, varying from deserts to tropical rain for- ests. The central anchor-shaped mountain range of Ceylon ascends from the coastal plain to summits over 5000 feet. Extend- ing north from Cape Comorin to Bombay along the western coast of India, the West- ern Ghats rise sharply and descend to the Deccan Plateau. Separating the Deccan Plateau and some outlying uplands (Vindhya Range, etc.) from the foothills of the Himalaya is a broad lowland area (the Indo-Gangetic Plain) extending from the valley of the Indus, across the Indian Desert and eastward to the Khasi Hills of Assam. Vast areas of Nepal, Sikkim, and Bhutan are in the foothills of the Hima- laya. The northeast monsoonal air mass flows southward and southeastward from a continuous high pressure region near the Hindu Kush. As the air mass flows over India, it is continuously diverted from the Bay of Bengal across southern India by the Coriolis effect. Beginning in December in the northern provinces, the monsoon continues until February. Com- monly referred to as the cold season, this period is characterized by generally dry, cool air resulting in lower maximal and minimal temperatures. With the termination of the northeast monsoon, a warming trend occurs from March until May in most of the Indian subcontinent. This period is characterized by clear skies and warm temperatures. In the northeastern provinces of India, vio- lent storms result when dry land air meets humid sea air; these storms provide sub- stantial rainfall that enhances agriculture. Originating over the Indian Ocean in a zone of high pressure, the wet south- Tue MecopTerA oF INDIA AND ADJACENT REGIONS western monsoonal air mass flows north- eastward. The beginning of the wet mon- soon in Ceylon is around May; the air mass progresses northward slowly and reaches Karachi, Pakistan, about 15 July. The physical relief and plant cover of India and Burma and the upper wind currents are influential in diverting the winds through Assam, the northeastern provinces of India, and the Himalayan foothills. These regions receive about 25- 30°% of their annual rain in July, when it normally rains at least 20 days. As the air iS) W mass slowly withdraws to the sea, there is a transitional period of cool, dry weather during October and November. Along the western slopes of the West- ern Ghats, there is a substantial amount of rain (80 or more inches annually), sup- porting tropical evergreen forests (fig. 1). The eastern slopes and plateau receive only about 40 inches of rain annually and are covered predominantly by tropical dry deciduous and thorn forests. The eastern coast of southern India consists of many fertile delta areas now under heavy culti- \ ari 70 80° 90° | ice tte 30/4 Si EN of, ia LL | 1 Le i (ei oe ta a alee! LY Sel pam f) \ 20°4 log” YOO m1: a 100 YoOkm ical | Vegetational Zones Cultivated Scrub Arid to semi-arid Deciduous monsoon 10: Evergreen forest Z Alpine-Himalayan 90 Fic 1. Vegetational zones of India and adjacent regions. 24 Tue UNIverRsITy oF KANSAS SCIENCE BULLETIN vation. In a very similar pattern, Ceylon is divided from west to east into climatic and vegetational zones. The lowlands of western India and Pakistan do not receive much rain (0-10 inches in desert areas, up to 20 inches per year in western India) and are covered with semi-arid and arid vegetation. In sharp contrast, mid-central and mid-east- ern India receive from 30 to 75 inches of rain annually and the area is intensely cultivated, especially the Ganges River basin. The higher plateaus are covered primarily by dry, tropical deciduous for- ests. The lowlands and river basins of As- sam and Bangladesh receive about 60 inches of rain annually, whereas the pla- teaus of Assam receive between 100 and 200 inches annually. Some areas receive considerably more rain, for example, Cherrapunji, Assam, about 428 inches. The forests in these areas are primarily tropical evergreen. As elevation increases from the foot- hills into the Himalaya, the temperate forests change to coniferous forests at about 12,000 feet, and these in turn give way to alpine meadows and tundra on the southern slopes of the mountains at about 15,000 feet. DISTRIBUTION OF MECOPTERA IN THE INDIAN REGION The climatic zones, topography, and natural vegetational zones (figs. 1, 2) can be considered together in forming a gen- eral picture of India with respect to Mecoptera and their distribution. The Shillong Plateau of Assam and the West- ern Ghats are similar in that they are covered primarily by tropical evergreen forests, and characterized by large seasonal rainfalls and cool temperatures. More than half the described species of Indian Neo- panorpa are from these areas. The other areas in which Neopanorpa has been col- lected are the deciduous monsoonal and dry tropical forest zones of the Himalayan foothills. The vegetation of these areas is stratified according to elevation. The ex- act environmental parameters important for the survival of Neopanorpa are not known. There are no records of Neopanorpa from central India. Most of that area is agricultural and moreover lacks the com- bination of climatic, topographical, and vegetational conditions found in either northern or southern India. Along the northern edge of the Deccan Plateau, the forests are primarily of the tropical de- ciduous type, and possibly Neopanorpa occurs in these areas. In southern India, the majority of the specimens of Neopanorpa were collected in early May, during the transitional pe- riod between monsoons. The remaining specimens were collected from late August through October, which is slightly prior to the termination of the southwest mon- soon. In sharp contrast, the majority of specimens collected in northern India were taken from late July through Sep- tember. In both of these areas and _ sea- sons, Neopanorpa was collected from 700 feet to about 7,000 feet. Only Bittacus has been reported from Sri Lanka (Ceylon). From the few avail- able specimens, it appears that Brttacus also emerges prior to the termination of the wet monsoon. All four Singhalese specimens were taken at elevations under 500 feet. Ceylon has a dense human pop- ulation and the lowland areas are intensely cultivated. Destruction of suitable habitat has probably severely limited the recent distribution of Bzttacus on Ceylon. The apparent distribution of Buttaczs is very different from that of Neopanorpa. The elevations of areas (fig. 2) in which Bittacus has been collected in India range from nearly sea level to 1400 feet. In gen- eral, these areas are warmer and drier Tue Mecoprera oF INDIA AND ADJACENT REGIONS than those in which Neopanorpa has been collected. There is, however, a lowland species of Neopanorpa (undescribed) from Nepal, and the type locality of B. latz- pennis is Darjeeling, India. The ranges of some North American species, such as B. strigosus Hagen and B. stigmaterus Say, are over a thousand miles wide, so it is possible that some of the western Chinese species may extend into Assam and along the Himalayan foothills. There are no records of Panorpa in the + 20 ey 70° Sredeal twice MET CAN }- Nilgiri Hills e Anaimalai Hills iS) Yr Indian subcontinent. Within a hundred miles of the northeastern border of Assam, however, in the mountains of Sikang, China, there are ten species of Panorpa (Cheng, 1957). Only a single species is known from Yunnan Province, and there is an undescribed species in northern Burma. Since some species of Panorpa, such as the North American P. helena Byers, are wide-ranging, it seems that Panorpa may extend westward into the mountains of Assam and the Himalayan Darjeelins fe Se aN Q Sj Cle t Khasi Hills; i) SS J] 200 4g0km Elevation Pe beeee (0-1640 ft.) 19 Elan one m. (1640-6562 ft.) 2000 + m. (65:62. tft.) 90° » ee Fic. 2. Elevational map of India and adjacent regions. Black circles indicate localities where Mecoptera have been collected. See accompanying list for details. 26 Tue UNIversiry oF KANsAS SCIENCE BULLETIN foothills. Panorpa has never been collected there, however, due possibly to the lack of collecting or to its absence for ecological reasons. TAXONOMY OF THE GENUS BITTACUS Since the head and thorax other than the wings are rarely used for taxonomic purposes, they have been omitted here. Setty (1940) and Hepburn (1969, 1970) discussed and illustrated these body re- gions in detail. Otanes (1922) and Mat- suda (1965) provided a somewhat more generalized discussion of the head in Mecoptera. Storch and Chadwick (1968) investigated the thorax of Bittacus strigo- sus Hagen of North America. Wincs: The color, shape, and venation of the wings (fig. 3) are extensively used in the taxonomy of Bittacus. The veins include the costa, a two-branched subcosta, a five-branched radius (Rez some- times further branched), a four-branched media, a two-branched cubitus, and three anal veins (3A very short). Both fore and hind wings have similar shape, but the fore wings are longer (by one or more mm.). The only significant differences in venation between the fore and hind wings are that the basal fusion of veins Cu and M is longer in the hind wings, and Cuz and 1A are partly fused in the hind wings, not independent as in the fore wings. The color of the wing membrane varies from clear to yellowish or tinged with brown. The apical wing margin may be bluntly rounded or more pointed, often a taxo- nomic character. In many species, the position of certain cross-veins may be diagnostic, especially the apical and subcostal cross-veins. The number of pterostigmal cross-veins (fig. 3, Pcv) between veins Ri and R2+3 has often been used for taxonomic purposes, but since the number of cross-veins may vary within an individual specimen, this character is not always reliable. The posi- tion of the subcostal cross-vein (Scv) in relation to the origin of the radial sector (ORs) and the first fork of the radial sector (FRs) is a taxonomically useful character, but the position of the Scv may vary. In many species, there is an apical cross-vein (Av) between Cuz and the dis- tal end of 1A near the hind margin of the wing; the presence or absence of this cross-vein appears to be a reliable taxo- nomic character. ABDOMEN OF THE Mae: Many taxonomic characters are found on segments 9-11 (Tjeder, 1956). The general shape and the projections of the epiandrial lobes of the ninth tergum (figs. 5, 8) vary considerably between species and are usu- ally diagnostic. The ninth sternum is not noticeably modified and is rarely included in descriptions. The basistyles are fused basally and extend posteriorly beneath the ninth tergum; their shape, size, and hairi- ness may be taxonomically useful. The dististyles, one on the posterodorsal apex of each basistyle, are small compared to the basistyles. Lobes (penunci) at the base of the aedeagus and the shape of the apical portion of the aedeagus are taxo- nomically useful in bittacids generally. However, in the Indian species, the base of the aedeagus is not modified and lobed as it is, for example, in some of the Chi- nese species. The proctiger, or combined tenth and eleventh segments, is usually small and inconspicuous, but in some spe- cies it is enlarged and extremely modified and extends dorsad between the epiandrial lobes. The single-segmented cerci are rarely important in taxonomy. ABDOMEN OF THE Femare: In earlier de- scriptions, the female genitalia were al- most completely ignored, but Tjeder (1956) based several descriptions of Af- rican species on female holotypes, using such characters as degree of sclerotiza- tion of the antecosta of the posterior ab- Tue Mecoprera oF INDIA AND ADJACENT REGIONS 27 dominal segments, subgenital plate, sper- matheca and its duct, and cerci. These characters have not been thoroughly in- vestigated, and nothing is known concern- ing their variation in Indian_ bittacids. When possible, these structures have been included in the descriptions in this paper. The subgenital plate is below the eighth and ninth terga. The shape, hairi- ness, and degree of fusion of the sclerites of the subgenital plate appear to be re- liable taxonomic characters. The supra- anale and subanale (tergum and sternum of the eleventh segment) often have char- acteristic posterior margins. The tricho- bothria of the cerci may also be used for taxonomic purposes (Tjeder, 1956). The spermatheca, which may be diagnostic of the species, is usually within segment 7 a little anterior to the subgenital plate. The width and degree of sclerotization of the posterior abdominal antecostae have also been used in taxonomy. Key To THE INDIAN SPECIES oF Birracus The males of B. indicus and insularis and the females of B. taratensis and no- dosus are unknown. The sex of the holo- type of latipennis is unknown. 1 Vein 1A extends to or beyond level it [ERG Se an ee eee Soe ere oD Vein 1A extends at most to mid- point between levels of ORs and TFRs Ses ee eee ey ee 3 2(1) Scv one-fourth distance from ORs to FRs; proctiger of male large, directed dorsad between epiandrial eves es Se eee taraiensis Scv midway between ORs and [EVR G: Jvc le 2 6 Sane eee? indicus 3(1) Vein 1A joins hind margin oppo- site level of ORs; Scv before FRs; foremwaneg 4-6 mamas es 2s 4 Vein 1A joins hind margin mid- way between ORs and FRs; Scv variable; fore wing over 17 mm _.. 5 4(3) Scv midway between ORs and FRs; Cuz joins hind margin oppo- site level of fork of M344; epian- drial lobes of male stout, about twice as long as wide, with numer- ous spines at apex of each __ insularis Scv slightly before FRs; Cuz joins hind margin slightly proximal to level of fork of Ms+4; (male un- Knit) Weteas Uns eevee Sh henryt 5(3) Scv midway between ORs and FRs to slightly distal of FRs; Sc joins costal margin proximal to level of Cur; epiandrial lobes of male stout, about twice as long as wide, with four or five spines on small dorsal protuberance at apex ORC AC NN tN SO ee era nodosus Scv distal to FRs; Sc joins costal margin before level of Cui, oppo- site) level) (ob Cus.) (Gnale) une known) 2 ee latipennis DESCRIPTIONS OF SPECIES: GENUS Bittacus Bittacus henryt KimMMINs Bittacus henryt Kimmins, 1928: 395-396, fig. 1. Unfortunately the original description was based on a single female. This species closely resembles B. insularis but, on the basis of the holotype, differs from it in the positions of the Scv and Sc, the length of the fore wings (henry: about 16.1 mm. and insularis about 14.8 mm.), and the more basal position of Cuz with respect to the fork of M314. In znsularts, the Scv is midway between the ORs and the FRs, and the Sc merges with the costal margin slightly distal to the FRs. In henryz, the Scv is nearly opposite the FRs, and the Sc merges with the costal margin barely past the FRs. Kimmins (1928) stated that the position of the FRs in relation to the fork of vein M is different in henry: and in- sularis. However, the position of the FRs 28 Tue UNIVERSITY OF KANSAS SCIENCE BULLETIN appears to be the same in the holotypes of both species. Another difference in the wing venation is the additional cross-vein between Re and Rs in Aenry:, but it 1s present in only three of four specimens we consider to be henryi. Very little is known concerning variation in Indian bittacids, and there is a strong possibility that despite the described venational dif- ferences henryi and insularis may be syn- onyms. The venation, particularly the position of the Scv, of males from southern India is similar to that of henry; therefore, these males are included in the following redescription, which is otherwise based on notes and drawings of the female holo- type: Heap: Vertex, genae, frons, and ros- trum sordid yellowish brown; two lateral ocelli slightly larger than medium ocellus; a row of long yellowish hairs on ridge across frons. Antennal scape and pedicel sordid dark brown, flagellum blackish brown with short pilosity. TuHorax: One large black spine at each side on anterior margin of pronotum. Entire dorsum, pleural regions, and coxae sordid yellowish brown. Legs yellowish brown, femora and tibiae with blackish apices; hind femora slightly enlarged in male. Wing membrane with light yellow- ish brown tinge, outer margins. slightly darker; longitudinal veins brown, cross- veins in basal portion of wing brown, several in distal portion pale. Av absent; Scv slightly basal to FRs; two Pev. Sc merges with costal margin slightly beyond FRs; vein 1A merges with hind margin opposite ORs. ABDOMEN oF Mate: ‘Terga and _ sterna 2-9 yellowish brown. Base of tergum 9 with transverse ridge extending distally onto epiandrial lobes. Lobes stout, with numerous spines at apices, hairs long- est at bases of lobes; mesal protuberance on lower inner surface with many small spines (fig. 8). Angle of hind margin slightly acute. Dististyles short (figs. 6, 7), apices curved mesad. Cerci short, 0.8 length of basistyles. Aedeagus stout at base, tapering to a fine point. ABDOMEN OF FemaLe: Terga 2-11 and corresponding sterna and subgenital plate yellowish brown. Subgenital plate bluntly rounded at apex. Cerci sub-equal in length to segment 9. Lenctu or Fore Wine: Female holotype, 16 mm.; male, 14.6-15.0 mm., hind wing, 13.3-13.6 mm. Hototyre: Female, Vavuniya, N. P., Ceylon, 16 December 1923; in the Brit- ish Museum (Natural History), London. In addition, we have examined the fol- lowing specimens: 1 male, Pondicherry, India (no date), Maindron, in the Mu- seum National d'Histoire Naturelle, Paris; 1 female, Yala, Ceylon, 18 December 1931, in Colombo Museum, Ceylon; 1 male, Coimbatore, Madras State, India, 1400 feet, November 1962, P. S. Nathan (figs. 5-8), and 5 males, 4 females, Coimbatore, June 1974, S. Bharathi, in the Snow Entomo- logical Museum. On the basis of male genital structures, this species resembles B. nodosus from Pakistan and certain African species, espe- cially B. chevaliert Navas. The epiandrial lobes have a striking resemblance to those of nodosus and chevalieri, particularly the protuberance on the lower, mesal margin. In lateral view, the epiandrial lobes and basistyles of henry: are somewhat smaller but in shape resemble those of B. sjostedt Weele from Africa. The male genitalia suggest a closer relationship of henryz to several northeastern African species than to the western Chinese bittacids. Bittacus indicus WALKER Bittacus indicus Walker, 1853: 469. This species is known only from the holotype, from which the abdomen has been largely eaten away by dermestids or Tue MecopTerA oF INDIA AND ApJACENT REGIONS psocids. There are two eggs clinging to the remaining portion of the abdomen, which indicate that the species description There has been some confusion concerning the type local- ity; that on the label is “E. Ind.” which Walker interpreted as East Indies. Esben- Petersen (1921) interpreted it to be East was based on a female. Rs Scv FRs 2» India, which is the more probable locality, because no other Bittacus has ever been collected in the East Indies. Esben-Petersen (1921) makes reference to a type series. A second specimen with the type is labeled “East Ind.” and was also part of the Saunders collection, but it does not have a type label. Vein 1A ends Fic. 3. Right fore wing of Bittacus henry: (length about 13.5 mm.); A—anal vein, Cu—cubitus, FRs—first fork of radial sector, M—media, ORs—origin of radial sector, Pcv—pterostigmal cross-vein, R—radius, Rs—radial sector, Sc—subcosta, Scv—subcostal cross-vein. (@) 0.5mm [eS ee eee Fics. 4-8. Buittacus henry: Kimmins. 4, terminal abdominal segments, male, left lateral aspect (aed— aedeagus, bs—basistyle, c—cercus, ds—dististyle, s—sternum, t—tergum); 5, epiandrial lobes, dorsal aspect; 6, right dististyle, posterior aspect; 7, right dististyle, anterodorsal aspect; 8, left epiandrial lobe, dorsal aspect. Upper scale figs. 4-5, lower scale figs. 6-8. 30 THe UNIVERSITY OF KANSAS SCIENCE BULLETIN opposite the ORs in the front wing of this second individual, which would indicate that it is not B. indicus. This condition is found in B. insularis and B. henryi, but the fore wing of this second specimen is over 3 mm. longer than that of either insularis or henryi. This specimen prob- ably represents an undescribed species, but since it is in very poor condition, the spe- cies should not be described until addi- tional specimens are available. The following brief redescription is based partly on notes and drawings of the female holotype: Heap: Dorsum pale brown, except black ocellar prominence. Rostrum shiny brown. Antennae yellowish brown, short haired, 10 mm long. Tuorax: Dorsum and _ pleural areas brown. Legs brown, except tips of femora and tibiae narrowly banded blackish brown. Wing membrane tinged with pale grayish brown; longitudinal veins yellow- ish brown in distal portion. No Av; two Pev; Scv equidistant between ORs and FRs; end of Sc opposite end of Cuz; end of 1A opposite FRs. AppoMEN oF FEMALE: Basal segments brown (terminal segments broken off). Mate: Unknown. Lenctu oF Fort Wine: Female holotype, 19 mm.; hind wing, female holotype, 17 mm. Hototyre: Female, “E. Ind.” (probably eastern India); in the British Museum (Natural History), London. On the basis of wing venation, this species closely resembles B. tarazensis, but it differs slightly in that the Scv is mid- way between the ORs and FRs, not as proximally situated as in taraiensis. An- other slight difference is that the end of vein LA of indicus is nearly opposite FRs, not beyond it as in tararensis, but this may result from differences in wing size (fore wing of taraiensts, 175-18.0 mm.; indicus, 19 mm.). The basal cross-vein between veins LA and 2A of taraiensis is at mid- length of 2A but is in the distal one-sixth of 2A in indicus. In two North American species of which numerous specimens were examined, B. pilicornis Westwood and B. strigosus Hagen, the position of this second cross-vein appears to be con- stant. In B. stigmaterus Say, also from North America, it is absent. Some vari- ation occurs in B. punctiger Westwood from North America. The position of this cross-vein may prove to be taxonomically useful, but its reliability is still somewhat doubtful. This species is distinct from the other Indian species, except taratensis, in that vein 1A extends distally beyond the level of FRs. It seems probable that zndicus and taraiensis are synonyms. Bittacus insularis EsBEN-PETERSEN Bittacus insularis Esben-Petersen, 1915: Sa 3 This Singhalese species was originally described from a single female. As stated in the redescription of B. henryz, there is a likelihood that these nominal species are synonyms. But several pairs of Chinese species, for example, B. vexilliferus Byers and B. carpenteri Cheng, have nearly identical wing patterns yet are readily dif- ferentiated by the male genitalia. Further species descriptions based on females only should therefore probably be avoided, at least until reliable diagnostic characters can be demonstrated. The following brief redescription 1s based on drawings and notes of the female holotype: Heap: Vertex brown, except ocellar prom- inence black; rostrum brown grading to blackish brown towards apex. Antennae dark brown. Tuorax: Prothorax sordid brown. Meso- and metathorax reddish brown. Femora, tibiae, and tarsi brown grading to black- ish brown at apices. Wing membrane Tue MEcoprera oF INDIA AND ADJACENT REGIONS 31 with yellowish brown tinge, with smoky brown apical margin. No Av. End of Sc extends distally beyond FRs. Scv mid- way between ORs and FRs. Vein la joins hind margin opposite ORS. End of Cu: extends distally beyond first fork of Mi +2. Two Pcv. ABDOMEN OF FemaAe: Terga and sterna brown. Subgenital plate slightly concave near mid-length; apex bluntly rounded with numerous long hairs. Cerci about 25 mm. long. Mate: Unknown. Lenetu oF Fore Wine: Female holotype, 15 mm.; hind wing, 13 mm. Hovotyre: Female, Hambantota, Cey- lon, 29 November 1908, T. B. Fletcher; in the British Museum (Natural History), London. This small species differs from the larger Indian species in having the end of vein 1A opposite the level of the ORs. In B. latipennis and B. nodosus, the Scv is distal to the FRs, not midway between the ORs and FRs as in znsularis (and in- dicus and henry). Bittacus latipennis GERSTAECKER Bittacus latipennis Gerstaecker, 1885: 120. Since the species was described from a single specimen from which the abdomen was absent, it is impossible to determine its sex. However, the position of the Scv distal to the FRs, the merging of vein 1A with the hind margin beyond the level of ORs, and the width of the wing should permit recognition of either sex, when latipennis is again discovered. The posi- tion of 1A with respect to the ORs is similar in B. henryi and B. insularis, but both of these species are smaller than latipennis and have the Scv located be- tween the ORs and FRs. Esben-Petersen (1921) adequately dealt with the general body coloration. His ac- count of the wing venation is, however, somewhat incomplete. The following re- description is based on his photographic illustration of the holotype of latipennis: Wines: Membrane with yellowish brown tinge, except for several hyaline spots; longitudinal veins brown; cross-veins in basal portion brown, paler in distal por- tion. Av absent. Scv dstal to FRs in fore wing, near ORs in hind wing. One Pcv. Sc extends about 2 mm. beyond end of FRs; 1A merges with hind margin slightly distal to ORs. LencTH oF Fore Wine: holotype, 19 mm.; hind wing 17 mm. Hototyre: Darjeeling, India, date un- known; in Zoologisches Institut und Mu- seum, Ernst Moritz Arndt Universitat, Greifswald, D. D. R. (East Germany). This species can be easily recognized by the long, broad wings. In B. indicus and B. taratensis, vein 1A joins the hind margin of the wing distal to the FRs, not slightly beyond the level of ORs as in dati- pennis. There is a broad-winged, unde- scribed species from Nepal (only one female available), but it differs from J/ati- pennis in having vein 1A opposite the level of the FRs, and the Scv far beyond the FRs, at the level of the end of the Cus. Bittacus taraiensis PENNY Bittacus taraiensis Penny, 1969: 161-164, figs. 1-3. This species is very similar to B. in- dicus. The only significant difference is the position of the Scv. In another Indian Bittacus, B. nodosus, the position of the Scv varies, and this would support syn- onymizing indicus and taraiensis. But as emphasized in the redescription of B. in- sularis, some Chinese species that have very similar wings can be recognized only from the male genitalia. Our knowledge of Indian bittacids is very incomplete. Penny (1969) stated that taratensts also differed from indicus in the coloration of the abdomen. Coloration is a somewhat unreliable character, often affected by Sy Tue University oF Kansas SciENCE BULLETIN method of preservation, age of the speci- men when preserved, etc. The types of taraiensis are preserved in alcohol and the holotype of indicus is a pinned specimen. The following brief redescription is de- rived from Penny’s (1969) original de- scription: Wines: Membrane hyaline, stigma and ad- jacent cells tawny brown; distal two rows of cross-veins nearly white. Av absent, 1A extending distally beyond first fork of M. Two Pcv. Sc extends beyond FRs. Scv one-fourth distance from ORs to FRs. AspoMEN oF Mate: Epiandrial lobes with connecting transverse ridge at base, apices converging and curved dorsad; short, many-spined protuberance on basal, mesal margin directed mesad. Basi-styles with bluntly rounded apices. Aedeagus stout at base, tapering abruptly to fine point. Proctiger projecting dorsally be- tween epiandrial lobes, with long hairs at apex. Femare: Unknown. Lenctu oF Fore Wine: 17.5-18.0 mm. 3opy Lenctu: 14-16 mm. Hototyre: Male, Pant Nagar, Nainital Dist. India, 5 August 1968, Lb. J. Stan- nard; in the Illinois Natural History Sur- vey collection, Urbana, Illinois. Paratype, 1 male, same locality and collector, 26 July 1968. On the basis of the proctiger, this spe- cies is similar to several western Chinese species, especially B. carpenteri Cheng and B. pielt Navas, in having the 10th segment elongated and extended dorsally between the epiandrial lobes (Penny, 1969). But in faraiensis, the aedeagus is not lobed at the base (i.e., has no penunci) as in these species but is similar to those of B. henry: and B. nodosus. The epiandrial lobes of taraiensis are similar to those of some In- dian and African species, especially B. chevalieri Navas, in having a many-spined protuberance on the inner, basal margin. The genitalia of taraiensis represent an intermediate form between certain west- ern Chinese and African species. In wing venation, taraiensis differs from the other Indian species, except in- dicus and an undescribed species (female) from Nepal, in having 1A extending dis- tally beyond FRs. Bittacus nodosus, NEw SPECIES Description based on two males and one specimen of unknown sex, pinned. Heap: Vertex, frons, and genae light yel- lowish brown; rostrum light yellowish brown grading into dark yellowish brown at apex; maxillary palpi yellowish brown, except apical segment dark yellowish brown. Ocelli extremely large, on con- spicuous black prominence with two spines posterior to median ocellus. Epi- stomal suture with row of hairs of uni- form length. Antennal scape and pedicel light yellowish brown; flagellum light yel- lowish brown with 14 segments. Tuorax: Pronotum with one long black spine at each side, on both anterior and posterior margins. Several small black setae on meso- and metanotum; two short, black setae on posterior margin of metepi- Dorsum, pleural region, and coxae yellowish brown. Anterior surfaces of coxae with long yellowish hairs. Legs yellowish brown, except apices of tarso- meres blackish brown; hind femora not especially enlarged. Wing membrane hya- line, outer margin with faint brownish tinge; longitudinal veins and cross-veins brown. Av absent. One Pev, except right fore wing (two), in holotype; two Pecv in all wings in paratypes. Scv opposite FRs, in left fore wing, slightly distal to FRs in right fore wing in holotype, midway be- tween FRs and ORs in paratype; Scv basal to FRs in hind wings of both specimens. Vein 1A extends slightly beyond level of ORs. Three cross-veins between Rs and Ry in left fore wing of holotype, two in other wings. meron. Tue Mecoprera oF [NDIA AND ADJACENT REGIONS 33 Fics. 9-12. Bittacus nodosus n. sp. 9, terminal abdominal segments, male holotype, leit lateral aspect; 10, epiandrial lobes, dorsal aspect; 11, tip of left epiandrial lobe, lateral aspect; 12, left dististyle, an- terodorsal aspect. Scale figs. 9-10. ABDOMEN oF Mate: Terga and sterna 2-9 sordid yellowish brown. Base of ter- gum 9 with transverse ridge extending postero-laterally onto epiandrial lobes. Lobes stout, each with 4-5 spines on small protuberance at apex (figs. 10, 11); pro- tuberance on lower mesal margin with many spines (fig. 10). Hind margin of basistyles squarely blunt, with long hairs. Dististyles small, apex abruptly truncate forming a somewhat concave surface (fig. 12). Cerci short, about half as long as epiandrial lobe. Aedeagus stout at base, tapering to fine point. FemaLe: Unknown. LencTH oF Fore Wine: 16.0-17.3 mm. (holotype 17.3 mm.); hind wing, 14.5-16 mm. (hototype, 16 mm.). Ho.otyre: Male, Karachi, Pakistan, 8 October 1959, R. I. Sailer; in Snow En- tomological Museum. Paratypes, 1 male, 1 without abdomen, Deesa, India (about 330 mi. E. of Karachi), 8 January 1935, G. G. Nurse; in British Museum (Nat- ural History), London. In wing venation, this species resem- bles B. latipennis in that the Scy is oppo- site or slightly distal to the FRs and vein 1A extends slightly beyond ORs. But nodosus differs from latipennis in having narrower wings (greatest width, nodosus, about 2.6 mm.; latipennis, about 4.9 mm.) and a shorter Sc. In addition, there is a second basal cross-vein between veins 1A and 2A in latipennis, not present in no- dosus, and it appears from Esben-Peter- sen’s (1921) photograph of the wings of lattipenmis that the longitudinal veins are slightly darkened at their bases. The male genitalia of nodosus closely 34 Tue UNIversity oF KANsAs SCIENCE BULLETIN resemble those of B. chevalieri Navas from Africa, and vaguely those of B. henry, especially in the epiandrial lobes. The absence of penunci from the base of the aedeagus, the small, inconspicuous proctiger, and the spined protuberance on the inner, basal margin of the epiandrial lobes are characters shared by all three of these species. B. nodosus differs from the other two in having a short, spined, dorsal protuberance at the apex of each epian- drial lobe. The posterior margins of the basistyles are abruptly truncate, not rounded as in the African and Singhalese species. On the basis of the genitalia, nodosus resembles the African species more than it does either taratensis or henry. Species OF Bittacus NOT DESCRIBED. Two bittacids that probably represent new species were examined but not de- scribed. The following brief discussion of these specimens is primarily to note their existence, their affinities with other spe- cies, and their location. Since the male genitalia provide the most reliable taxo- nomic characters, we have declined to base descriptions of new species on these in- dividuals. In the British Museum, there is a sin- gle Bittacus (abdomen missing) from Hadibu, Socotra Island (12°36’ N, 53°59’ E), in the Gulf of Aden about 125 miles east of Capo Guardfui, Republic of So- mali. The wing venation is similar to that of B. henryi and B. tinsularis, but this species differs in having a clear membrane and slightly larger wings (about 16 mm. long). Another undescribed Biéttacus, a single female from Simra, Nepal, has distinctive wing venation. The Scv is distal to the FRs and 1A extends almost to the level of the FRs (fore wing). Since no other known species from the Indian subconti- nent has such a venational pattern, this may allow recognition of the male when discovered. This specimen is in the col- lection of the Entomology Research Insti- tute, Canada Department of Agriculture, Ottawa. TAXONOMY OF THE GENUS NEOPANORPA Heap: The head of Neopanorpa has the large lateral compound eyes, long promi- nent rostrum, and three ocelli on a promi- nence, as commonly found in Panorpi- dae. In the adult, the ecdysial cleavage line is present but confined to the ocellar region (Hepburn, 1969). Two major components of the rostrum are the cly- peus (anterior) and the subgenae (lat- eral); the elongated maxillary stipites and membranous submentum comprise nearly all of the posterior aspect. The absence of the occipital suture makes it impossible to define the exact extent of the genae and postgenae, but the subgenal suture clearly outlines the postgenal process. In Neopanorpa, the antennae are long and conspicuous. The scape is subcylin- drical and the pedicel nearly globular. The flagellum is usually blackish brown to black and consists of numerous (usually 35-45) finely pubescent flagellomeres. There is a whorl of six to eight slightly longer bristles near the apex of each fla- gellomere. Tuorax: In addition to the short hairs that cover most of the thoracic sclerites, the pronotum may possess several long setae along the anterior margin. In the meso- and metathorax, the meron is en- larged and extends nearly the length of the adjacent coxa. The legs are similar on all three thoracic segments, each hav- ing the femur and tibia of approximately equal length, and two long spurs at the apex of the tibia. The basitarsus is ap- proximately equal in length to the other four tarsomeres combined. Usually amber colored, the pre-tarsal claws have five pectinations, Tue Mecoprera oF INDIA AND ADJACENT REGIONS 35 Fic. 13. Left lateral aspect of male of Neopanorpa denticulata n. sp. A—anal vein, AB—apical band, BB—basal band, BS—basal spot, bs—basistyle, Cu—cubitus, ds—dististyle, hv—hypovalves, M—media, MS—marginal spot, pnp—posterior notal process, pp3t—posterior process of third tergum, PTB—ptero- stigmal band, R—radius, Sc—subcosta, t—tergum. Note that the basal band is divided, or double, in the fore wing in many individuals of this species. 36 Tue UNIVERSITY OF KANSAS SCIENCE BULLETIN Wincs: The fore wings are slightly (1 or more mm.) longer than the hind wings. On the proximal hind margin of the fore wing, there are four to eight setae which together with two long setae on the proximal costal margin of the hind wing serve to unite the wings during flight. The venation of the wing (fig. 13) consists of a costa, a subcosta, a five- branched radius, a four-branched media, a two-branched cubitus, and three anal veins. The only significant departure in the venation of the hind wing from that of the fore wing is a consolidation of anal and cubital veins, in which the cubitus forks and converges with both the media and first anal veins; the anal and cubital veins are not independent as in the fore wing. At the first fork of the media, the veins are always pale within a clear spot (thyridium). Vein 1A merges with the hind margin before the origin of the Rs, a generic character differentiating Neo- panorpa from Panorpa. The wing membrane varies from clear to yellowish brown and may be slightly to extensively marked with spots or bands. The presence or absence of these bands can often be used to differentiate species, although there is some variation within species in their size and shape. Generally, within a species of Neopanorpa, the back- ground color and wing markings are somewhat reduced in the hind wings, and the dark coloration is usually more ex- tensive in the female. If present, the apical band is in the ex- treme distal portion of the wing; its prox- imal margin may be irregular or notched. Usually the broadest band, the pterostig- mal is proximal to the apical band and has a variety of shapes depending on the species. It may be unbranched or may fork posteriorly. In several species, it is partially fused with adjacent bands. Near the costal margin in the midsection of the wing, the marginal spot may partially fuse to the pterostigmal band, basal band, or both. The basal band, in the proximal half of the wing, is often incomplete and may appear as a series of spots. In several species, the basal band is complete and may be partially fused to the marginal spot, basal spot or proximal branch of the pterostigmal band. Occasionally there is a distinct basal spot near the base of the wing; but in many heavily marked species, individual spots are difficult to discern. ABpoMEN OF Mate: In segments 2-5, there is a large membranous pleural area between the terga and sterna. Seg- ments 6-8 are completely sclerotized and roughly cylindrical. On the posterior mar- gin of the third tergum, there is a sclero- tized median projection (fig. 13) extend- ing over the fourth tergum for which the terms “notoorganus”, “notal organ”, or “notorgan” were proposed by Crampton (1931). A small protuberance on the fourth tergum, which Crampton called the “postorganus”, is a functional part of the notal organ. Mickoleit (1971) reported that in Panorpa communis Linnaeus from Europe the combined structure served as a mechanism for securing the female’s wings during mating. This function was also recorded by Issiki (1933) and_ has been observed by us in several species of North American Panorpa. Often ne- glected in older species descriptions, the shape and length of the notal organ differ from species to species. In a few species of Neopanorpa, the fifth and sixth terga are modified with projections or protuberances that are ex- cellent taxonomic characters. In several large Himalayan species, the seventh and eighth segments are elongated but are thicker than those of Leptopanorpa. The constricted base of the ninth segment 1s cylindrical; the basally fused ventral and dorsal sclerites diverge to form the hypo- valves and ninth tergum respectively. Seg- ments 10-11 are small and weakly sclero- Tue MEeEcopTerA oF INDIA AND ADJACENT REGIONS 37 tized in Panorpidae. The single-seg- mented cerci are the only portion of the last two segments not normally concealed by the ninth tergum. The genital bulb (figs. 64, 65) con- sists of the ninth tergum (epiandrium), ninth sternum (including the hypovalves), basally fused basistyles, dististyles, and aedeagus. At the ventromesal base of each dististyle, there is a single lobe that is usually directed mesad, but occasionally more ventrad. The shape, size, and hairi- ness of this lobe are useful in the recog- nition of many species. The basal lobe is almost always single in Panorpa and Neo- panorpa, while in Leptopanorpa it is usu- ally divided into two parts. Located between the basistyles, the aedeagus (figs. 68, 69) has several struc- tures that are useful for taxonomic pur- poses. The dorsal and ventral valves are on the distal margin of the axial portion of the aedeagus. In Neopanorpa, the dor- sal valves are generally small, rounded, darkly sclerotized protuberances and rela- tively unimportant in taxonomy. The shape of the ventral valves varies and may be useful for taxonomic purposes. The ventral and dorsal parameres often have characteristic shapes diagnostic for the species. Arising on the proximal, ventral margin of the aedeagus beside the ventral valves, the ventral parameres extend dis- tally, with the terminal portion having a variety of shapes. In several species, they are reduced to small knobs or hooks; in others, they are blade-shaped or filamen- tous. If present, the dorsal parameres are above the dorsal valves and are usually short and stout, rarely attaining the length Fics. 14-15. Abdomen of female Neopanorpa denticulata n. sp. 14, left lateral aspect of entire abdomen; 15, left lateral aspect of terminal segments (c—cercus, gp genital plate, s—sternum, sgp— subgenital plate, t—tergum). Scale a— fig. 14; b— fig. 15. 38 Tue UNIVERSITY OF KANSAS SCIENCE BULLETIN and bizarre shapes of the ventral para- meres. Extending laterally from the mid- section of the aedeagus, the lateral proc- esses are generally short and somewhat triangular, in Neopanorpa. In several species, they are fused with the dorsal parameres, the two structures together modified into long blades. ABDOMEN OF FEMALE: In segments 2-8 (fig. 14), there is a large membranous pleural area separating the terga and sterna. A wedge-shaped sclerite (latero- tergite) is situated on each side in the pleuron of the distal half of segments 7-8 in many species. In other species, these sclerites are absent, but hairs outline their position. The subgenital plate is consid- ered to be formed of extensions of the eighth sternum, and the genital plate is the highly modified ninth sternum (Byers, 1954). Rarely visible from the outside, the genital plate (fig. 70) consists of three parts that are variously developed in dif- ferent species. The distal and basal plates extend laterally from a central body or axial portion. The arms of the distal plate extend posteriorly and are occasionally visible without dissection. There are in most species two divergent apodemes at the anterior end of the axial portion, serv- ing as sites of muscle attachment. The shapes of the subgenital plate and, more importantly, the composite genital plate are useful in the recognition of species. Seg- ments 10-11 are greatly reduced in size as compared to the more anterior segments. Usually blackish brown to black, the two- segmented cerci are covered with many short hairs. The eleventh segment, situ- ated below the cerci, has only weakly de- veloped sclerites. KEY TO MALES OF INDIAN SPECIES OF THE Genus Neopanor pa The males of flava, ocellaris, and benact are unknown, and not enough is known about the males of fenestrata and sordida to determine accurately where they should be placed in the key. Both fenestrata and sordida should be located somewhere after couplet six. 1 Posterior half of sixth abdominal segment greatly modified, either con- spicuously thickened and humped dorsally, or with slender projections 2 Posterior half of sixth segment not so’ modified 222 0 ee 4 2(1) — Sixth abdominal segment with thick dorsal hump; seventh segment bear- ing long dorsal hairs; ninth tergum with median, setose projection near mid-lensthi (ies 018)) gibbosa Sixth abdominal segment not thick- ened and humped dorsally; seventh segment without long hairs; ninth tergum without median projection Neary muid-length ==) 3 3(2) Projections from sixth abdominal segment more than half length of segment (fig. 43); dististyles shorter than basistyles (fig. 40); wings 12- 13 emimielon gy tee ee cornuta Projections from sixth abdominal segment much shorter than half length of segment (fig. 57); disti- styles much longer than _basistyles (fig. 58); wings 15-17 mm. long ..... furcata 4(1) Posterior process of third abdominal tergum extends beyond posterior Marpin Of filth tereum) = 5 Posterior process of third tergum ex- tends at most to posterior margin of fourthetergum) ee eese ee 6 5(4) | Hypovalves each with single mesal projection (figs. 17, 19) ........ appendiculata Hypovalves each with one mesal and one ventral, subapical projec- tions (Gigs.982,,(86)) salai 6(4) Wing membrane with strong yellow- Sh? CNG 2. sc eccic ee 7 Wing membrane not yellowish, with hyaline to brownish gray tinge ...... 8 7(6) | Hypovalves widely divergent at base, converging somewhat at tips (fig. 28); basal lobe of dististyles broad, thick blade with convex ventral sur- face. 7 ett. ot 2... en chillcotti Hypovalves not widely divergent at base, apical half of each deflected ventrad (figs. 48, 49); basal lobe Narrow .°..2.2e.33) ee ee effusa 8(6) Wings unmarked except for stigmal spot, or only weakly patterned __...... 9 Wings distinctly marked with bands, SPOtS;rOIgy WO Chee cee ee 1] 9(8) 10(9) 11(8) Tue MecoprerA oF INDIA AND ADJACENT REGIONS 39 Wing membrane hyaline; hypovalves sinuate with spiny, hairy apices curved abruptly mesad (figs. 110, 2) ieee eee ee ee echinata Wing membrane with diffuse brown- ish gray pattern; sinuate, their apices not spiny —_.... 10 hypovalves not Hypovalves narrowly separated at base (fig. 136); aedeagus short, not extending between tral parameres vergent apically (fig. 138) Hypovalves separated by subcircular space at base (fig. 71); aedeagus long, extending between dististyles; ventral parameres long, beyond hypovalves, divergent api- fealliyam (iverson lee /))) eee nipalica dististyles; ven- inconspicuous, con- ochrura projecting Posterior process of third abdominal tergum extending entire length of fourth broad near base, slender subapically, widened at apex (figs. 36, 39) ... contracta Posterior third tergum short, rarely extending beyond mid- length of fourth tergum, but never entire length; ventral parameres broader subapically than near base .. 12 tergum; ventral parameres process of 12(11) Dististyles largely covered with long, especially near apices glabrous; dense black hairs, mid-length, _ their hypovalves slender, sinuate (fig. 64) Airsuta Dististyles without long dense black hairs near mid-length, with hairs 14(13) 15(14) basal lobe; hypovalves a Bt eS BS oA OE lo 13 thickest at level of tapering toward apices, widely separated basally (as fig. 129) Hypovalves wider near mid-length longest on not sinuate Hypovalves basal separation, LO ie ene antl Wi RAT IS Lee indica than at level of basal separation, not widely separated basally —_............... 14 Mesal margins of hypovalves entire, without lobes or spines (fig. 147); ventral parameres two-branched Gites, HGS aly 1153) eee ramulata Mesal margins of hypovalves uneven, with spinous lobes; ventral parameres unbranched _...... 15 setiferous or Mesal margins of hypovalves each with two. setiferous lobes before Be< (Grea, SN, SiS) See zebrata Mesal margins of hypovalves each with two spinous points before apex, another at apex (figs. 101, LOE ited cette ccd Pee. oa trea denticulata KEY TO FEMALES OF INDIAN SPECIES OF THE GENus Neopanor pa The females of echinata n. sp., fenes- trata, gibbosa n. sp., and nipalica are un- known. Not enough is known about the female of sordida to include this species in the key. The genital plates of benaci Fic. 16. Neopanorpa appendiculata (Westwood), male, right wings (fore wing length 13.8 mm.). 40 Tue UNIVERSITY oF KANsAs SCIENCE BULLETIN have never been examined and the posi- tion of this species in the key is entirely on external characters. 1 Wing membrane with strong yel- lowship tinge Wing membrane not yellowish, hya- line sto: brownish orays ee Proximal margin of apical band oblique to costal edge of wing; proximal and distal branches of pterostigmal band subequal in width; pronotum and anterior half of mesonotum black, posterior half of mesontum and all metanotum yellowishwbrowny se Proximal margin of apical band either perpendicular to costal edge of wing or uneven and_ notched; branches of pterostigmal band of unequal width Genital plate broad, nearly as wide as long, its axial portion poorly de- veloped (fig. 60); pronotum black with narrow yellowish band along posterior margin; anterior half of mesonotum black, posterior half and all metanotum yellowish brown __.. Genital plate not nearly as broad as long, axial portion short but dis- tinct; thoracic nota not as described above Subgenital plate deeply notched api- cally (fig. 32); mesonotum and metanotum sordid yellowish brown throughout, without distinct pattern; fore wing 14-16 mm. (fig. 27) _... Subgenital plate only shallowly notched apically (fig. 55); anterior two-fifths of mesonotum and _ante- rior one-third of metanotum black, rest of both nota yellowish brown; fore wing about 18 mm. (fig. 52) Distal half of fore wing heavily pat- terned, apical band broadly joined to pterostigmal band along hind margin Distal half of fore wing less heavily patterned, apical band not connected to pterostigmal band Apical and pterostigmal bands _ of fore with distinct included pale areas; cross-veins in apical por- tion of wing pale (fig. 26) Apical and pterostigmal bands with- out included pale areas; outermost cross-veins not pale, same color as veins in basal portion of wing ...... Axial portion of genital plate less than one-half total length: basal plate and apodemes of genital plate absent 2(1) 3(2) 4(3) 5(1) 6(5) wings based furcata chillcotti flava “I benaci ocellaris 8 8(7) 9(7) 10(9) 11(10) 12(11) 13(9) 14(13) Axial portion long, over one-half to- tal length; basal plate and apodemes welll developedy= 3. = ee 9 Anterior half of mesonotum dark brown, posterior half yellowish; metanotum yellowish with dark brown semicircular spot on anterior margin; genital plate small, less than 5 mm. lone (fis: 45)) =e cornuta Both mesonotum and metanotum black on anterior half and broadly along mid-line across posterior half; genital plate longer than .5 mm. .... ochrura Apex of subgenital plate notched, plates ventrally kceled en 10 Apex rounded, plate not keeled ___. 13 Apodemes of genital plate long, widely divergent anteriorly (fig. 70); no basal spot in fore wings (ig. 63) 42.22.22 ee arsaiee Apodemes shorter, only slightly di- vergent and only shallowly separated anteriorly; basal spot in fore wings present or absent Apodemes of genital plate deflected abruptly ventrad from anterior edge of plate (fig. 155); basal spot absent FLOM COLE Wiles 2 ee ramulata Apodemes of genital plate deflected gradually anteriorly; basal spot in fore wings large _...W.. 12 Large clear spot in apical band; dis- tal plate of genital plate dorso-ven- ventrally flattened (figs. 98, 99) _. zebrata Apical band without large clear spot, with notch on proximal margin; lateral margins of genital plate strongly curved ventrad (figs. 108, 109) denticulata Apodemes of genital plate about one- third of total length (fig. 132) _.. Apodemes about one-fourth of total length Subgenital plate about 1.5 times as long as greatest width, without setae on triangular anteromedian area (fig. 23); apodemes of genital plate separated from edge of plate for- ward (fig. 24) appendiculata Subgenital plate only slightly longer than its greatest width, covered with setae anteromedially (fig. 89); apo- demes diverging from axis well for- ward of edge of genital plate (fig. Si) i oo soe ot 11 dorsad indica 14 salat DEscRIPTIONS OF SPECIES: GENUS Neopanor pa Neopanorpa appendiculata (Werstwoop) Panorpa appendiculata Westwood, 1846: 186. Tue Mecoptera oF INDIA AND ADJACENT REGIONS 4] Neopanorpa appendiculata; Esben-Peter- sen, 1915: 229. Originally described as a Panorpa, this species was transferred to Neopanorpa by Esben-Petersen (1915), who pointed out that the vein 1A does not extend beyond the ORs. The redescription and photo- graph in Esben-Petersen’s world mono- GWwB Ue L 0.5 mm i graph (1921) are based on a female syn- type. This syntype, however, probably represents the female of N. salat, which species differs from appendiculata in hav- ing more extensive wing pigmentation, and in the male an additional projection on each hypovalve and a shorter posterior process of the third abdominal tergum. Fics. 17-22. Neopanorpa appendiculata (Westwood). 17, genital bulb, male lectotype, ventral aspect; 18, genital bulb, lectotype, right lateral aspect; 19, tip of left hypovalve, male, mesal aspect; 20, aedeagus, male, ventral aspect (part of right side omitted); 21, 3-6, lectotype, dorsal aspect. terga 3-6, lectotype, left lateral aspect; 22, terga 42 Tue UNIVERSITY OF KANSAS SCIENCE BULLETIN On the basis of wing pattern and body coloration, appendiculata closely resembles N. indica, n. sp., but the male of appendi- culata differs in having a much longer posterior process of the third tergum and a larger mesal projection on the hypo- valves. The females are not, however, as easily differentiated. Males of both species have been collected at the same locality and on the same date, together with eight females, probably all N. appendiculata. In the type series of appendiculata, there are two males of indica and two females with wing patterns identical to that of these males. The genital plates of these two females differ from those of the eight mentioned females in having longer apo- demes and narrower distal plate. These two female syntypes of appendiculata are therefore tentatively included in indica. The following composite redescription is based on notes and drawings of the male lectotype, 23 additional males, and 17 females: Heap: Dorsum, frons, and rostrum sordid yellowish brown. Scape, pedicel and basal half of first flagellomere sordid yellowish brown, antennal color then abruptly changing to black, remaining 37 flagello- meres black. Tuorax: Dorsum dark sordid yellowish brown (lectotype) to dark grayish brown. Pleural area and coxae sordid yellow- ish brown to grayish brown; femora and tibiae sordid yellowish brown; tarsi brownish grading to black on segments 45. Wing (fig. 16) membrane hyaline; markings dark smoky brown; cross-veins in distal portion pale, more basal ones brown. Apical band broad. Pterostigmal band forked posteriorly, with narrow dis- tal branch. Marginal spot extends from costal margin to base of vein Mi +2. Basal band large anteriorly, constricted at vein M. Basal spot large. ABDOMEN OF Mate: Terga 2-9 and cor- responding sterna black. Posterior proc- ess of third tergum (figs. 21, 22) ex- tending to hind margin of fifth tergum or beyond, its basal one-third broad and flattened, a broad ventral thickening near mid-length, remaining portion slender; entire process yellowish brown. Process of fourth tergum large dorsal protuberance with a narrow hooked projection directed anteriorly; protuberance covered with long black hairs. Hypovalves (fig. 17) stout, somewhat cylindrical basally; short finger- like protuberance with long hairs on inner margin in distal half, posterior margin straight, apical part deflected into interior of bulb, blade-like, with marginal row of thick black spines. Lateral margins of tergum 9 abruptly convergent in distal half, posterior margin straight. Broad hook-shaped protuberance on anterior margin of sternum 10, with long hairs on posterior margin. Cerci short, extending only to edge of ninth tergum. Dististyles about one-third length of entire bulb. Ba- sal lobe of dististyles (figs. 17, 18) wide, bluntly rounded, directed ventrad, with long hairs on inner margin. Ventral parameres (fig. 20) thin, curved laterally in apical one-fourth, with short hairs on dorsal surface. Ventral valves of aedeagus broad, with thin lateral wings. Lateral processes broad at base, tapering to bluntly rounded point. Dorsal parameres curved ventrad, apex abruptly truncate. ABDOMEN OF FemaLe: Terga 2-11 and corresponding sterna black. Subgenital plate (fig. 23) broad with bluntly rounded apex, with many long hairs on distal side and apex. Genital platem(aesme. 25) about 1.1 mm. long; posterior arms of distal plate long, about half of total length, slender, somewhat twisted at base; distal and basal plates broad, about .25 of total length, with lateral margins abruptly curved dorsad; apodemes short, about .25 of total length, with apices twice as broad as base and distinct notch on an- terior margin. Tue MEcoprTera oF INDIA AND ADJACENT REGIONS 43 LenctH oF Fore Winc: Male, 11.0-13.0 mm. (lectotype, 11.6 mm.); female, 11.3- 12.6 mm. Lecrotyre: Male, Madras, date unknown; in British Museum (Natural History), London. In addition to the types mentioned above, we have examined the following specimens; 9 males, 8 females, Gudalur, Nilgiri Hills, South India, 3,500 feet, Sep- tember 1960, P. S. Nathan, in the Snow Entomological Museum; 14 males, 9 fe- males, Cherangoda (1200 m.) and Singara (1100 m.), Nilgiri Hills, October 1950 and May 1954, in the Naturhistorisches Mu- seum, Basel, Switzerland. The similarity of hypovalves, aedeagus, and dististyles in males and the subgenital and genital plates in females are the pri- mary characters used in grouping appen- diculata, indica, and salai into the ap- pendiculata group. It was called the appendiculata group after the first de- scribed, included species. Neopanorpa benaci Navas Neopanorpa benaci Navas, 1935: 97-98. This species was originally described from the female holotype only, from Kurseong, India. The abdomen has been broken off beyond the seventh segment since being described, so nothing is known I 1 mm } 25 G6.wW. 6. Osama een Fics. 23-25. Neopanorpa appendiculata (Westwood). 23, subgenital plate, female, ventral aspect (upper scale); 24, genital plate, female, ventral aspect (lower scale); 25, genital plate, right lateral aspect. ae Tue UNIvERSITY OF KANSAS SCIENCE BULLETIN about the genitalia. On the basis of wing pattern, benaci is very similar to N. nr- palica and may be a synonym. Another male, holotype of N. ochrura n. sp., has a wing pattern similar to that of benact but differs in having paler markings and smaller wings (about 11 mm.). The following redescription is based on notes and drawings of the holotype and the original description: Heap: Vertex, frons, and rostrum black. Antennae black, about 8.5 mm. long. Tuorax: Pronotum black; anterior half of mesonotum blackish brown, posterior half brownish; metanotum blackish brown. Pleural areas and coxae pale grayish. Legs brownish. Wing pattern complex (fig. 26); membrane with a faint brown- ish tinge, becoming pale whitish in distal portion of wing; markings light to dark smoky brown; cross-veins in distal portion of wing pale. Apical band widely fused to pterostigmal band along hind margin, with six pale to clear spots sur- rounding cross-veins. Pterostigmal band complete, forked posteriorly, with a pale spot between Rs and Rs. Marginal spot large, extending from costal margin to base of Mi+2. Basal band complete, con- stricted at vein M. ABDOMEN OF FEMALE: Anterior segments black, posterior segments dark brown to black. Mae: Unknown. LencTH oF Fore Wine: Female holotype, 14.4 mm. Fic. 26. Neopanorpa benaci Navas, female holotype, right fore wing (length about 13.9 mm.). Fic. 27. Neopanorpa chillcotti Byers, male paratype, right wings (fore wing length 15.0 mm.). Tue MeEcopTera oF INDIA AND ADJACENT REGIONS 45 Hovoryre: Female, Kurseong, India, 1932; in the Museum National d’Histoire Naturelle, Paris. Neopanorpa chillcotti Byers Neopanorpa chillcottt, Byers, 1971: 534- 539, figs. 1-10. Only certain diagnostic characters of this recently described Nepalese species are included in this brief redescription, to pro- vide a basis for comparisons with other species. The following is based on the original description and reexamination of three male and one female paratypes: G.W.B, 28 0.5 mm 31 Wine: (fig. 27): Membrane with strong dark yellowish tinge, markings brown. Apical band complete, its anterior proxi- mal margin with a small projection di- rected posteriorly. Pterostigmal band vari- able, usually forked posteriorly. Marginal spot not attaining costal margin. Basal band incomplete, consisting of small spot near hind margin. Basal spot absent. ABDOMEN OF Mate: Posterior process of third tergum extending to midlength of fourth tergum. Hypovalves (fig. 28) widely divergent at bases, convergent near apices; slender, somewhat wider in apical half. Dorsal parameres elongate, Fics. 28-34. Neopanorpa chillcotti Byers. 28, genital bulb, male holotype, ventral aspect; 29, aedeagus, ventral aspect; 30, ninth tergum, ventral aspect; 31, ninth tergum, left lateral aspect; 32, subgenital plate, female, dorsal aspect; 33, genital plate, ventral aspect; 34, genital plate, right lateral aspect. Scale a— figs. 28, 30-32; b— figs. 29, 33, 34. 46 Tue UNIVERSITY OF KANSAS SCIENCE BULLETIN slender, wedge-shaped. Ventral parameres slender, with minute hairs near apices. AspoMEN oF FemMate: Subgenital plate deeply notched at apex (fig. 32). Arms of genital plate (figs. 34, 35) broad and stout with bluntly rounded apices, sub- equal in length to axial portion. Axial portion broad, rounded anteriorly. Lenctu oF Fore Winc: Male, 14-16 mm. (holotype, 14.8 mm.). Hototyre: Male, Godavari, 10 mi SE of Katmandu, Nepal, 5,000 feet, 31 July 1967, Canadian Nepal Expedition. Holo- type, allotype, and 13 paratopotypes are in the Entomology Research Institute, Canada Department of Agriculture, Ot- tawa, Ontario, Canada; 4 paratypes are in the Snow Entomological Museum. Neopanorpa contracta CHENG Neopanorpa contracta Cheng, 1953: 122, {18S 4-55 ple D: Cheng’s original description was based on two females from Darjeeling, India. It deals adequately with the genitalia and wing pattern, but the drawings lack pre- cise detail. Two males, determined to be contracta on the basis of wing pattern and body coloration, are also included in the following brief redescription. The two males were badly damaged in transit from the Illinois Natural History Survey, but fortunately each specimen was pinned in its own sealed container so that genitalia could be associated with certainty. Heap: Dorsum shiny black, frons gray- ish brown, with a large median, shiny black spot; rostrum sordid grayish brown. Antennal scape blackish; base of pedicel blackish brown, changing abruptly to dark yellowish brown; flagellum blackish brown. TuHorax: Pronotum deep brown (hol- otype) to shiny blackish brown (males). Anterior half of mesoscutum and narrow band along median line to mesoscutellum blackish brown; posterolateral areas of mesoscutum light grayish brown. Meta- notum similar to mesonotum. Pleural area and coxae pale grayish brown, except vertical brownish streaks on meso- and metamera. Femora and tibiae grayish; basitarsi grayish grading to black on seg- ments 4-5, apices of tarsomeres 1-3 black. Wing membrane hyaline, markings light smoky brown. Apical band wide, with two hyaline spots in cells Rs and Mi. Pterostigmal band complete, forked pos- teriorly. Marginal spot large, extending from costal margin to base of vein Mi +2. Basal band incomplete, reduced to small spot in males, to large spot extending from Cui to hind margin in female (hol- otype). AppoMEN oF Mate: Terga 2-3 blackish brown, 46 sordid grayish brown (seg- ments 7-11 missing in one male, seg- ment 7 in other), 8-9 sordid yellowish brown. Sterna 2-9 sordid brown. Posterior process of third tergum broadly triangular in basal half, slender in distal half, ex- tending full length of fourth tergum, shiny black. Hypovalves (figs. 35, 36) extending to bases of dististyles, nearly uniform in width with bluntly rounded apices; distal half of each faintly sclero- tized, with pale whitish tinge. Dististyles (figs. 35, 39) slender, blackish brown, outer margins concave in basal half; basal lobes directed mesad, dorsal surface shal- lowly concave, with long yellowish hairs at apices. Ventral parameres (fig. 39) broad at base, slender near mid-length, with apices flattened into blades; distal half of each curved ventrad. Lateral proc- esses complex, mesal margin curved ven- trad over ventral parameres, with spines along distal edge. Dorsal parameres long, contiguous, converging to a fine point. ABDOMEN OF FEMALE: Terga 2-7 deep brown, 8-10 light brown. plate deeply notched. Arms of genital plate long, 0.6 of total length of genital Subgenital Tue MeEcopTera oF INDIA AND ADJACENT REGIONS Ae Axial portion slightly plate, twisted at base. slightly broadened anteriorly, notched. Lencru oF Fore Wino: Female holotype, 145 mm.; hind wing, 13.2 mm. Length of fore wing: Male, 14.0 mm.; hind wing, 13.2 mm. Hovotyre: Female, Darjeeling, India, May 1939, 'T. C. Maa; in the Maa col- lection, Taipei, Taiwan, China. Paratype, G.W.B. 1 female, same collection data (Cheng collection). We have examined the following individuals: 1 male, Rangiroon, Darjeeling, India, 6,250 feet, 25 May 1966, Kamath; 1 male, same locality, 26 May 1966, Gupta; both males are in the Illinois Natural History Survey collection, Ur- bana, Illinois. On the basis of the female genitalia, contracta superficially resembles several Fics. 35-39. Neopanorpa contracta Cheng. 35, genital bulb, male, ventral aspect; 36, genital bulb, left lateral aspect; 37, ninth tergum, dorsal aspect; 38, ninth tergum, left lateral aspect; 39, aedeagus and basal lobe of dististyle, right side, ventral aspect. Scale a— figs. 35-38; b— fig. 39. 48 Tue UNIVERSITY OF KANSAS SCIENCE BULLETIN regional species of Neopanorpa in having a deeply notched subgenital plate and a genital plate in which the posterior arms are longer than the axial portion. The genital plate of contracta is similar to that of N. appendiculata, especially the poste- rior arms, but lacks the long, anterior axial portion found in appendiculata. The genital bulb of contracta resem- bles that of N. nipalica in having long hypovalves of uniform width, long, con- tiguous dorsal parameres, and curved ven- tral parameres. However, contracta differs from nipalica in having a slightly darker wing pigmentation. Neopanorpa cornuta EspEN-PETERSEN Neopanorpa cornuta Esben-Petersen, 1915: 227. This species was originally described from eight males and ten females from the Khasi Hills, Assam. Illustrations of the male abdomen and wings (Esben- Petersen, 1921) show the sixth and seventh abdominal segments apparently in an un- natural position. In the illustration of the genital bulb, there are two forked struc- tures (titillators), which are the ventral parameres. N. cornuta can be recognized by its thoracic coloration and wing pattern. In addition, there are in the male two long projections on the enlarged posterodorsal margin of the sixth abdominal segment. In N. furcata, there are two claw-like pro- jections, but the posterior half of the sixth segment is not enlarged in lateral aspect as in cornuta. The posterior half of the sixth segment of N. gibbosa n. sp. is en- larged, but there are no projections. The following redescription is based on examination of the male lectotype, one male syntype from which the genital bulb is missing, two female syntypes and the original description: Heap: Vertex blackish brown, frons and rostrum yellowish brown. Antennal scape, pedicel, and proximal half of basal flagel- lomere yellowish brown, remaining 43 segments blackish brown. Tuorax: Pronotum black. Anterior half of mesoscutum dark brown, posterior half and mesoscutellum yellowish. Meta- notum yellowish with semicircular dark brown spot on anterior margin. Pleural areas and coxae yellowish to yellowish brown. Legs yellowish, apices of tarsi dark brown to black. Wing membrane with faint yellowish tinge, markings pale smoky brown. Cross-veins in distal por- tion of wing whitish, in proximal portion brown. Apical band incomplete, abruptly truncate posteriorly at Mi. Pterostigmal band narrowly connected to apical band along costal margin, greatly constricted at vein Re, forked posteriorly. Marginal spot slender, extending from Ri to Mi. Basal band incomplete, comprising faint spot near base of Rs and wide spot extending from hind margin to Cus. ABpoMEN oF Mate: ‘Terga 2-5 dark brown; anterior half of tergum 6 dark brown, posterior half yellowish brown; segments 7-9 yellowish brown. Sterna 2-9 yellowish brown. Posterior process of third tergum (fig. 43) extends to basal one-third of fifth tergum, slender at base, cylindrical in cross-section entire length. A small protuberance on each side of posterolateral margin of fifth tergum. Pos- terior half of sixth tergum greatly ele- vated (lateral aspect), with two diverging, finger-like, yellowish brown projections abruptly blackened at apices. Hypovalves with a large dorsal protuberance at apex, smaller dorsomedian protuberance; inner margin concave in basal half (figs. 40, 41). Lateral margins of dististyles slightly con- cave in basal half; basal lobes directed mesad, each with a blunt, flattened tooth on anterodorsal margin and a more rounded posterodorsal tooth; adjacent ven- Tue MEcopTeErA oF INDIA AND ADJACENT REGIONS 49 tral protuberance with stiff hairs (fig. 42). yellowish brown. Subgenital plate deeply Tergum 9 terminating at base of disti- notched, with several long spines at apex styles. (fig. 44). Arms of genital plate broad, ABDOMEN OF FEMALE: Terga brown, sterna 0.4 of total length. Anterior margin of e.w.8. 0.5 mm Fics. 40-46. Neopanorpa cornuta Esben-Petersen. 40, genital bulb, male lectotype, ventral aspect; 41, genital bulb, left lateral aspect; 42, base of right dististyle, ventral aspect; 43, third to sixth abdominal segments, male lectotype, dorsal aspect; 44, subgenital plate, female syntype, ventral aspect (lower scale); 45, genital plate, ventral aspect (lower scale); 46, genital plate, right lateral aspect. 50 Tue UNIVERSITY OF KANSAS SCIENCE BULLETIN genital plate shallowly emarginate (figs. 45, 46). LenctH oF Fore Wine: Both male and female syntypes, 125-15 mm.; hind wings: 11.5-13.5 mm. SyntypEs: 4 males, 6 females, N. Khasia Hills (N. Khasi Hills), Assam date un- known; in the British Museum (Nat- ural History), London; from the same locality, 2 females in the Museo de Ciencias Naturales, Barcelona, and 1 male, 1 female in the Universitetets Zoologiske Museum, Copenhagen. There is an addi- tional male in the British Museum, the label of which reads “Assam, 1901-262.” This may be one of the missing male syn- types. To prevent confusion in the future, the male syntype in the British Museum (Natural History) bearing a white label with a red circle around the word “type” is hereby designated as the lectotype. In wing pattern, this species somewhat resembles N. chillcotti Byers from Nepal and N. burmana Byers from Burma, and closely resembles N. gibbosa n. sp. from Simla, India. But the wing membrane of cornuta is clear, not with a strong yellow- ish tinge as in chillcotti, and the wings are slightly smaller. The basal band in bur- mana is nearly complete, not consisting of two spots as in cornuta. The wing pattern of gibbosa is nearly identical to that of cornuta but slightly lighter in color. Two Himalayan species, N. gibbosa and N. furcata, have the posterior part of abdominal segment 6 in the male modi- fied, but in each the shape is diagnostic of the species and quite different from that in cornuta. Some undescribed species from northern Burma have hypovalves similar to those of cornuta but these can easily be recognized by the absence of prolongations of the sixth segment and by their distinctly yellowish seventh and eighth abdominal segments. Two Chinese species, N. pielina Navas and N. mutabilis Cheng, have deeply cleft ventral para- meres similar to those of cornuta but are otherwise unlike cornuta in genitalia. The genital plate in cornuta closely re- sembles that of several species from north- ern Burma (undescribed), western China, and Indo-China in having long, twisted spatulate arms and a short, stout axial portion. The deeply notched subgenital plate of cornuta also resembles that of many of these species. Neopanorpa effusa (Navas) Campodotecnum effusum Navas, 1914: 429, fig. 7. Leptopanorpa effusa; 1921: 91-92 fie. OIE Esben-Petersen (1921) referred to a male type specimen, but mentioned noth- ing of the female syntype. He was some- what reluctant to transfer this species to Leptopanorpa because of the relatively short, stout seventh and eighth abdominal segments, as compared to those in other species of Leptopanorpa, the sessile genital segment, and the long, setose ventral parameres, not found in Leptopanorpa. In addition, the single basal lobe of the disti- style and the width of the wings indicate that this species is a Neopanorpa. Navas’ (1914) original illustrations are small and lack detail. Esben-Petersen (1921) illustrated only the fore wing of the male syntype. The following com- posite redescription is based in part on notes and drawings of the male syntype: Heap: Vertex brownish black; frons and rostrum brown; maxillary palpi yellow- ish brown. Antennal scape yellowish; pedicel yellowish with dark brown apex; proximal half of basal flagellomere black- ish, remaining segments dark brown. Tuorax: Pronotum black. Anterior half of mesoscutum black, posterior half of mesoscutum and scutellum yellowish brown. Metanotum yellowish brown. Pleural regions yellowish brown. Legs yellowish, except distal tips blackish. Esben-Petersen, Tue Mecoprera oF [np1A AND ApjAceNnr REcIoNs 51 Wing membrane with strong yellowish tinge; longitudinal veins — yellowish, slightly darker than membrane; markings dark brown. Apical band complete, its proximal margin oblique in posterior half. Pterostigmal band complete, forked _pos- teriorly, with branches of sub-equal width. Basal band complete from Ri to hind margin. ABDOMEN OF Mate: Terga and sterna 2-6 brown, 7-9 yellowish brown. Pos- terior process of third tergum (figs. 49a, 49b) somewhat triangular, extending almost to mid-length of fourth tergum. Two small, hair-covered protuberances on anterior margin of fourth tergum. Seventh and eighth segments (fig. 47) each about 3.5 mm. long. Lateral and inner margins G.W.,B. Fics. 47-51. Neopanorpa effusa (Navas). 47, terminal abdominal segments, male syntype, right lateral aspect (scale a); 48a, genital bulb, ventral aspect; 48b, genital bulb, right lateral aspect; 49a, third and fourth terga, dorsal aspect; 49b, third and fourth terga, left lateral aspect; 50, subgenital plate, female ventral aspect; 51a, genital plate, ventral aspect; 51b, genital plate, right lateral aspect. Sy Tue UNIVERSITY oF KANsAS SCIENCE BULLETIN of proximal two-thirds of hypovalves straight, distal one-third curved and arched ventrad with long hairs on ventral sur- face (figs. 48a, 48b). Dististyles (fig. 48a) long, nearly equal to length of basistyles; long, dense hairs on proximal two-thirds of lateral margin; basal lobes directed ven- trad, not cupped, with long hairs on apices. Ventral parameres setose, slender, long, extending between bases of disti- styles. Apices of ventral valves slightly divergent. ABDOMEN OF Femate: Subgenital plate (fig. 50) roughly oval, its apex shallowly indented. Genital plate (figs. 51a, 51b) with short, stout axial portion, long broad arms bent angularly dorsad near mid-length, and with short, thick anterior apodemes connected by a_ sclerotized lamella. Bopy LenetrH: Male syntype, 21 mm.; female syntype, 12 mm. LeNcTH oF Fore Wine: Male and female syntypes, 16 mm. Lenctu oF Hinp Wine: Male and female syntypes, 14.5 mm. Syntypres: Male and female, Sikkim, col- lected by Felder, date unknown; in the Naturhistorisches Museum, Wien. In addition, the following specimen has been identified: 1 female, 87 km. N. of Phuntsholing, Bhutan, 22 May 1972, in Naturhistorisches Museum, Basel, Switz- erland. By mistake, the original identification labels of both the male and female syn- types read “Campodotecnum falcatum Nav., Navas S. J. det.” Additional labels reading “Campodotecnum effusum Na- vas” and “Leptopanorpa effusa Navas, Es- ben-Petersen” appear to have been written by Esben-Petersen. In wing pattern, this species resembles the other large Himalayan species in hav- ing broad apical and pterostigmal bands and the membrane strongly yellowish- tinged. The basal and distal branches of the pterostigmal band of effwsa are sub- equal in width, while the distal branch is narrower in furcata. The basal band is nearly complete in effwsa, not consisting of a large spot along the hind margin as in furcata and flava. In effusa, the long, slender, setose ven- tral parameres are very diagnostic and have somewhat the appearance of those found in many species of Panorpa. The basal lobes of the dististyles are also un- usual in that they are not cupped on either the ventral or mesal surface. Mesonotal- metanotal color pattern will differentiate effusa from flava (Byers, 1975). Neopanorpa fenestrata (NEEDHAM) Panorpa fenestrata Needham, 1909: 195- 196, figs. 13-15. Neopanorpa fenestrata; Esben-Petersen, 1915: 230. Unfortunately, the male holotype could not be obtained from the Indian Museum, and Needham’s original description is not very informative. Esben-Petersen (1915) transferred this species to Neopanorpa, but it seems that he did not examine the holotype and in his world monograph (1921) only quoted Needham’s descrip- tion. The genital bulb resembles that of Neopanorpa, but the wing venation has never been confirmed. Needham’s (1909) illustrations of the genital bulb are small and lack detail. The following is derived entirely from the original description: Heap: Dorsum black, lateral margins of rostrum and basal segment of antennae somewhat reddish. Antennae 11 mm. long. Tuorax: Dorsum black; pleural areas paler, somewhat reddish. Wing membrane hyaline. Apical band present. Pterostig- mal band forked posteriorly. Basal band fused with marginal spot, enclosing a hya- line spot on costal margin. Basal spot present. ABpoMEN oF Mate: Terga and _ sterna 26 black, 7-9 reddish. Posterior process Tue Mecoprtera or [NDIA AND ADJACENT REGIONS 33) of third tergum short, extending about one-third length of fourth tergum. Hy- povalves long, extending beyond bases of dististyles, their inner margins contig- uous in distal half; inner margins folded under lateral margins at base, forming a dorsal protuberance. Aedeagus extends between dististyles. Lenctu oF Bopy: Male holotype, 15 mm.; expanse of fore wings 28 mm. Hototyre: Male, Assam, date unknown; in the Indian Museum, Calcutta. Neopanorpa flava EsBEN-PETERSEN Neopanorpa flava Esben-Petersen, 1915: 225-226. This species was originally described from the female holotype only. It is one of the largest Neopanorpas known, its fore wings being 18.5 mm. long. This was probably the reason that Esben-Petersen felt that when the male was discovered the species might be transferred to Lepto- panorpa. But flava has not only large but also broad wings, a long but stout abdo- men, and a genital plate and bi-colored thorax that are similar to those of N. fur- cata and N. effusa. The original description dealt fairly adequately with coloration. The following composite redescription is based on notes and drawings of the female holotype and one additional female specimen: Heap: Dorsum black; frons and rostrum yellowish brown. Antennal scape, pedicel, and first flagellomere yellowish brown; remaining flagellomeres blackish brown. Tuorax: Pronotum black. Anterior two- fifths of mesonotum black, changing abruptly to yellowish brown posteriorly. Metanotum similar, except only anterior one-third to one-half black. Pleural areas, coxae, and legs yellowish brown. Wing (fig. 52) membrane with strong yellowish tinge, markings dark brown, longitudinal veins brownish through spots and bands, yellowish in uncolored membrane, cross- veins in distal portion of wing pale. Api- cal band complete but uneven, posterior edge deeply and broadly emarginate. Pterostigmal band complete, widely sepa- rated along costal and hind margins from apical band, greatly constricted at first fork of Re+3, forked posteriorly. Marginal spot small, not in contact with costal margin (holotype). Basal band incomplete, com- prising small spot near base of Rs and wide band extending from hind margin of wing almost to M. Basal spot absent. ABDOMEN OF FEMALE: Terga 2-3 black- Fic. 52. Neopanorpa flava Esben-Petersen, female, right wings (fore wing length 17.8 mm.). 54 Tue UNIversiry oF KANsAs SCIENCE BULLETIN Fics. 53-55. Neopanorpa flava Esben-Petersen. I 0.5 mm 54 53, genital plate, ventral aspect; 54, genital plate, right lateral aspect; 55, subgenital plate, ventral aspect (upper scale). ish brown, 3-4 blackish brown laterally, sordid reddish brown medially, 5-9 sor- did reddish brown. Sterna 2-8 sordid reddish brown. Apex of subgenital plate (fig. 55) shallowly emarginate, with long brownish setae in distal one-third. Arms of genital plate (figs. 53, 54) broad, their bases arched dorsad and darkly sclerotized ventrally. Axial portion stout, its anterior apodemes short, thick, rounded at apex, the notch between them mostly closed by a sclerotized lamella. Mate: Unknown. Lenetu oF Fore Wine: Female holotype, 18.5 mm. Hotoryre: Female, Sikkim, September 1885, Bingham; in the Zoologisches Mu- seum, Humboldt University, Berlin. Ad- ditional specimen examined: 1 female, Sikkim, R. P. Bretaudeau, 1894; in the British Museum (Natural History), Lon- don. This species differs from the other two large Himalayan species, N. furcata and N. effusa, in having a complete apical band and wide pterostigmal band. The very striking color pattern on the meso- notum and metanotum should also prove useful in the recognition of the male when discovered. There are five species from China and one (undescribed) from northern Burma that have very large wings (fore wings over 15 mm.) Of these species, N. caveata Cheng, N. cavaleriei Navas, and the Bur- mese species have markings on the wings. The wing markings of cavaleriet are nearly identical in shape to those of flava Tue MEeEcopTerA oF INpIA AND ApJACENT REGIONS 55 Fic. 56. Neopanorpa furcata (Hardwicke), male, right wings (fore wing length 15.4 mm.). but slightly lighter in color. The fore wings of cavaleriei are 16 mm. long, not as long as those of flava (18.5 mm.), and the membrane is nearly clear, not strongly tinged with yellow as in flava. On the basis of the genital plate, flava resembles N. translucida Cheng from China in having long arms and broad axial portion, but it differs from translu- cida in not having the arms of the distal plate twisted. Neopanorpa furcata (HarpwickeE) Panorpa furcata Hardwicke, 1825: 132, figs. 2-6. Leptopanorpa ISIS ZI Probably due to its wing length (19.5 mm.) and the elongate seventh and eighth abdominal segments of the male, Esben- Petersen (1915) transferred furcata from the genus Panorpa to Leptopanorpa. The undivided basal lobe of the dististyles, the slightly emarginate ninth tergum, the width of the wings, and the shape of the genital plate of the female, however, war- rant transferring farcata to Neopanorpa. Several inaccurate statements exist in Hardwicke’s original description. Origi- nally reported to extend the entire length of the fourth tergum, the posterior process furcata; Esben-Petersen, of the third tergum is only half that length. The antennal flagella were also reported to consist of “about 52 segments.” Unfortunately, the flagella are now miss- ing from both types, but there are only 46 flagellomeres in other male and female specimens examined. Since little mention had been made of the female by Hard- wicke, Esben-Petersen (1921) redescribed the female syntype and gave a brief ac- count of the male syntype from the orig- inal description, repeating several of Hard- wicke’s inaccuracies. The following redescription is a com- posite from the original description, Es- ben-Petersen’s redescription, drawings of the male syntype, and study of two addi- tional males and seven females: Heap: Vertex black with narrow band of yellowish brown along anterior and lateral margins; frons, genae, and rostrum yellowish brown. Antennal scape yellow- ish brown, pedicel blackish brown, basal flagellomere brown at base grading to black at apex, remaining segments black. Tuorax: Pronotum black, except nar- row yellowish brown band along hind margin. Anterior half of mesonotum black, posterior half yellowish brown to black (syntypes). Metanotum yellowish brown to grayish brown (syntypes). Legs 56 Tue UNIVERSITY oF KANsAS SCIENCE BULLETIN Fics. 57-59. Neopanorpa furcata (Hardwicke). 57, terminal abdominal segments, left lateral aspect; 58, genital bulb, ventral aspect (upper right scale); 59, third and fourth terga, dorsal aspect. sordid yellowish. Wing (fig. 56) mem- brane with a strong yellowish tinge; longi- tudinal veins brownish black, except basal portion of subcosta and distal portion of radius yellowish; cross-veins paler, espe- cially in distal portion of wing; markings dark smoky brown. Apical band com- plete, proximal margin nearly straight. Pterostigmal band wide, forked posteri- orly, with distal branch narrow, proximal branch sub-equal in width to anterior part of band. Marginal spot small, between Ro+3 and Rais. Basal band incomplete, a large spot extending from hind margin to cell M. Basal spot along hind margin. ABDOMEN OF Mare: ‘Terga and_ sterna Tue MecopTerA oF INDIA AND ADJACENT REGIONS Sy 2-9 yellowish brown. Yellowish hairs along posterodorsal margin of third ter- gum longer than other abdominal pi- losity. Posterior process of third tergum extending half length of fourth tergum, apex bluntly rounded, sides of process sub- parallel (fig. 59). Hairs along mid-line of fourth tergum slightly longer than other pilosity, directed cephalad. Conspicuous rounded, dorsal protuberance on posterior margin of tergum 4. Posterior margin of sixth tergum (fig. 57) with two claw-like projections with long, yellowish hairs on lateral margins; small protuberance at base of each projection densely covered with long hairs. Segment 7 about 7.5 mm. long; segment 8 about 7 mm. long; both segments abruptly more slender than pre- ceding segments, with long hairs on an- terodorsal margin. Genital bulb (fig. 58) large, about 7.5 mm. long, exceeding usual proportion to body size and wing length. Outer (lateral) margins of hypovalves (fig. 58) slightly convex; inner margins overlapped and arched ventrad in distal half; large subtriangular projection di- rected dorsad from mesal margin. Disti- styles long, slightly over half length of bulb, slender, with outer margins concave in basal one-third; margins of basal lobe subparallel, apex truncate, lobe directed mesad, covered with long hairs. ABDOMEN OF Femate: Terga 2-9 dark yellowish brown to blackish brown (syn- type). Sterna 2-8 yellowish brown. Sub- genital plate (fig. 62) slightly emargi- nate at apex, with long yellowish hairs on distal sides and apex. Arms of genital plate (figs. 60, 61) long, two-thirds of to- tal length, twisted at base, spatulate at apex, with inner basal surface darkly scle- rotized. Lateral margins of distal plate projecting, thickened, curved cephalad; anterior margin shallowly emarginate. LenctH oF Fore Wine: Male, 17.0-19.5 mm. (syntype, 19.5 mm.). LeNncTH oF Fore Wine: Female, 15.6-19.0 mm. (syntype, 19.0 mm.). SyntypEs: One male and one female, 61 I 0.5 mm Fics. 60-62. Neopanorpa furcata (Hardwicke). 60, genital plate, female, ventral aspect (lower scale); 61, genital plate, right lateral aspect; 62, subgenital plate, ventral aspect. 58 Tue UNIVERSITY OF KANSAS SCIENCE BULLETIN Fic. 63. Neopanorpa hirsuta (Crampton), male, right wings (fore wing length 12.4 mm.). Nepal, no other data; in the British Mu- seum (Natural History), London. In addition, we have examined the fol- lowing specimens: 1 male and 2 females, 1 mi. S of Ulleri, Nepal, 5,000-7,000 feet, 16 May 1954, J. Quinlan; 1 male and 1 female, S of Baron River, East Nepal, 11 June 1954, 7,000 feet, L. Swan; all of these are in British Museum (Natural History), London; 1 female, Namdu, Nepal, 1450 m., 18 June 1961; 2 females, Manga Deo- rati, Nepal, 2300-2500 m., 19 June 1961; 1 female, Solu, Nepal, 2700-2900 m., 11 June 1961; all are in the collection of Dr. H. Janetshek, Innsbruck, Austria. On the basis of the female genitalia, this species resembles several Chinese spe- cies, especially N. claripennis Carpenter, in having a broad genital plate, darkly sclerotized inner, basal surface of the arms of the genital plate, and no apodemal processes. Body size and wing pattern of furcata are similar to those of another large Himalayan species, N. flava, but the genital plate of flava lacks the large trans- verse distal plate and the darkened basal surface of the arms and has a somewhat elongate axial portion. The male of furcata can easily be rec- ognized by the modifications of the sixth abdominal tergum and by its genital bulb, which is over twice the size of that of most species of Neopanorpa. The seventh and eighth abdominal segments are ex- tremely elongated, but they are of greater diameter than their similarly elongate counterparts in some species of Lepto- panorpa, and the entire body of N. frcata is more robust than that of any typical Leptopanorpa. Neopanorpa hirsuta (CRAMPTON) Panorpa hirsuta Crampton, 1931: 10, fig. 23. Neopanorpa hirsuta; Carpenter, 1931: 184- Io); safer, IE In a paper concerning the morphology and phylogeny of Notiothauma_ reed, Crampton published a sketch of the geni- tal bulb of an Indian species, suggesting that if it proved to be undescribed it should be called Panorpa hirsuta. Carpen- ter’s (1931) redescription of this single specimen is somewhat brief, and his draw- ing of the genital bulb lacks certain de- Tue MeEcopTerA oF INDIA AND ADJACENT REGIONS Fics. 64-67. Neopanorpa hirsuta (Crampton). 64, genital bulb, ventral aspect; 65, terminal ab- dominal segments, male, left lateral aspect; 66, ninth and tenth segments (basistyles and dististyles re- moved); 67, subgenital plate, female, ventral aspect; bs valve, t—tergum. tails; in addition, the notal process is not described. The following redescription is based on 12 males and 13 females. Eight of the females were collected together with ten males at Naduvatam in the Nilgiri Hills, basistyle, c—cercus, ds—dististyle, hv—hypo- in May 1958. In wing pattern and body coloration, the females agree with all known males of /zrsuta. Heap: Dorsum, rostrum, and genae light reddish brown to reddish brown. Anten- nal scape light reddish brown; pedicel 60 Tue UNIVERSITY OF KANSAS SCIENCE BULLETIN blackish brown; flagellum black, with 42- 43 segments in females, 44-45 segments in males. Tuorax: Anterior edge of pronotum with short pilosity, no setae. Dorsum, pleural areas and coxae light to dark reddish brown. Femora and tibiae light red- dish brown; basitarsi blackish brown grading to black on tarsomeres 4-5. Wing (fig. 63) membrane hyaline, markings light brown (holotype) to dark smoky brown; cross-veins pale in distal portion of wing, brown in proximal portion. Api- cal band ending posteriorly at tip of Mi in male, at tip of Ms in female; clear notch extending from proximal margin almost to tip of Me in female. Pterostig- GC.w.B. 68 mal band wide, forked posteriorly. Mar- ginal spot present, ending near base of Mz+5. Basal band incomplete. Basal spot absent. AspoMEN oF Mate: Terga 25 blackish brown to black, 6 blackish brown fad- ing posteriorly and laterally to light red- dish brown. Posterior process of third tergum (notal organ) half length of fourth tergum, bent downward over proc- ess of fourth tergum. Basistyles deeply separated. Hypovalves (figs. 65, 66) ex- tend full length of basistyles, with distinct transverse ridge at base, lateral margins slightly convex at base and near mid- length; hook-shaped dorsal protuberance at apex lightly sclerotized, nearly trans- L 0.5 mm Fics. 68-70. Neopanorpa hirsuta (Crampton). 68, aedeagus, ventral aspect; 69, aedeagus, left lateral aspect; 70, genital plate, female, ventral aspect; ap—apodemal process, ax—axial portion, da—distal arms, dp-lp—fused dorsal parameres and lateral processes, dv—dorsal valve, vp—ventral paramere, vv—ventral valve. Tue MecopTera oF INDIA AND ADJACENT REGIONS 61 parent. Tergum 9 extends nearly entire length of bulb, some hairs perpendicular to surface (fig. 64). Tenth sternum with long hairs. Cerci pale at base gradually darkening apically. Dististyles densely set with long black hairs, only tips exposed; basal lobe directed mesad; prominence at dorsal base of lobe bearing tuft of stout black hairs. Aedeagus (figs. 68, 69) elon- gate, lateral processes and ventral para- meres becoming wide, inwardly concave blades, toward apex. Outer lateral mar- gins of ventral parameres with short, thick pale hairs. Lateral processes and dorsal parameres fused, longer than ventral para- meres. ABDOMEN OF Female: Terga 2-6 black- ish brown to black, 7-10 dark reddish brown, cerci black. Sterna 2-6 blackish brown to black, 7-8 dark reddish brown. Subgenital plate (fig. 67) deeply notched and ventrally keeled. Arms of genital plate (fig. 70) long, blade-like, twisted 180°, with tips narrowly rounded. Ante- rior axial portion of genital plate forming two strongly divergent, large apodemes. LenctH oF Fore Winc: Males, 11.0-14.0 mm. (holotype, 14.0 mm.). LeNcTH oF Fore Wine: Females, 11.4- 12.1 mm. Hototyre: Male, Mysore, India, date un- known; in the Museum of Comparative Zoology, Harvard University. In addition to the holotype, we have seen the following specimens, all collected in South India by P. S. Nathan: 2 males, Chembra Peak area, Kerala State, 3500 feet, May 1970; 10 males, 8 females, Nadu- vatum, Nilgiri Hills, 6000 feet, May 1958; 1 female, Devala, Nilgiri Hills, 3200 feet, October 1960; 2 females, Gudalur, Nilgiri Hills, 3500 feet, September 1960; all in the Snow Entomological Museum; 1 female, Cherangoda, 1200 m., Nilgiri Hills, No- vember 1950, 1 female, same locality but without date, both in Naturhistorisches Museum, Basel. This species is aptly named. Males may be recognized from all other species of Neopanorpa by the long, dense hairs on the dististyles. The long, narrow, sinuate hypovalves with the dorsal protuberance at the apex are also diagnostic. On the basis of wing pattern, this species closely resembles N. cantonensis Cheng from Kwantung, China, but differs in not hav- ing a basal spot. The genital plate of the female holotype of cantonensis differs from that of Azrswta in not having anterior apodemes or a large basal plate. The aedeagus of /Airsuta superficially resembles that of other southern Indian species in having long, blade-like ventral parameres and fused dorsal parameres and lateral processes. The short, thick, pale hairs on the outer margin of the apically lobed ventral parameres are diagnostic of hirsuta. The aedeagus of hirsuta also dif- fers from those of N. zebrata and N. den- ticulata in lacking the large mesal exten- sions of the fused dorsal parameres and lateral processes. Neopanorpa nipalica (Navas) Panorpa nipalica Navas, 1910: 288, fig. 1. Aulops suffusa Navas, 1914: 427, fig. 5. Neopanorpa nipalica; Esben-Petersen, 1153230! Esben-Petersen (1915) transferred this species from Panorpa to Neopanorpa and synonymized Aulops suffusa with it with- out stating the reasons. Navas had de- scribed Aulops suffusa from a single male that was somewhat teneral and_ badly mashed, especially part of the genital bulb. The wing coloration and processes on the third and fourth abdominal terga of these two male holotypes are nearly identical (figs. 77, 78). Even though the genital bulb of the type of saffusa is badly dam- aged, the shapes of the hypovalves, disti- styles, and parts of the aedeagus are rec- ognizable and are very similar to the corresponding structures in nzpalica. 62 Tue UNIversitTy oF KANSAS SCIENCE BULLETIN Esben-Petersen’s (1921) redescription of nipalica deals adequately with body col- oration and wing pattern; however, the process of the third abdominal tergum and the genital bulb were not described in detail. The genital bulb of the male holotype of nzpalica has not been dissected, and the hypovalves conceal most of the aedeagus. The following redescription is based on examination of the holotypes of nipalica and suffusa and one additional male specimen: Heap: Antennae about as long as fore wing, comprising scape, pedicel, and 45- 46 flagellomeres. Tuorax: Dorsum blackish brown. Wing membrane tinged with brownish gray, unmarked except for light brownish stigma. ABDOMEN OF Mate: Terga 2-9 black, cor- responding sterna yellowish brown. Pos- terior process of third tergum (figs. 76, 77, 78) slender, with apex bent downward, extending nearly entire length of fourth tergum. Dorsal process of fourth tergum mid-length of segment in type of suffusa, slightly further caudad in type of nipalica. Hypovalves (figs. 71, 72) extend to base of dististyles, lateral margins straight, apices bluntly rounded; inner margins concave in basal one-fourth, sep- arating gradually in distal half. Basal por- tions of dististyles blackish; basal lobe di- rected mesad, cupped on dorsal surface; tips of dististyles directed slightly dorsad (lateral view). Aedeagus (fig. 74) with long conspicuous dorsal parameres; ven- tral parameres slender, somewhat bulbous at apices (fig. 75), which project from genital chamber (fig. 72). Femate: Unknown. LenctuH oF Fore Wine: Male, 12.8-13.8 mm. (holotype, nipalica, 12.8 mm.). Hototyre: Male, Sikkim, 1890, Har- mand; in the Museum National d’Histoire Naturelle, Paris. Additional specimens ex- amined: 1 male (holotype of suffusa), Sikkim, H. Fruhstorfer, in the Natur- historisches Museum, Wien; 1 male, Dar- jeeling, Rangiroon, India, 6250 feet, 25 May 1966, Kamath, in the Illinois Natural History Survey collection, Urbana, Illinois. If the female is similar to the holotype in color, the unmarked wings, black dor- sum of the thorax and abdomen and yel- lowish brown sterna should permit its recognition when discovered. There are several species from the Himalayan region known only from female holotypes, but these differ from the holotype of nzpalica in having distinct wing markings. On the basis of wing markings, the male of m7- palica from Darjeeling is closely similar to the female holotype of N. benaci. There is a possibility that benacz is a synonym. The genital bulb of mnzpalica super- ficially resembles that of N. clartpennts Carpenter from China, but the hypovalves of claripennis are slightly wider and the notal organ of claripennis is short and semicircular in shape, not long and _ nar- row as in mipalica. The dorsal parameres of claripennis are not as long or narrow as in ntpalica, and the ventral parameres of claripennis are short and filamentous. Several clear-winged species (undescribed) from northern Burma have hypovalves similar to those of nipalica but differ in having distinctly yellowish seventh and eighth abdominal segments. Neopanorpa ocellaris (Navas) Panorpa ocellaris Navas, 1908: 417. Neopanorpa ocellaris; _ Esben-Petersen, 1921: 82-83, fig. 92. This species is known only from the female holotype from Sikkim. In his re- description, Esben-Petersen (1921) dealt adequately with general body coloration and improved the illustration of the wing characters. The terminal segments of the female holotype were dissected and the following Tue MEcopTerA oF INDIA AND ApJACENT REGIONS 63 OlSa mim eens, Fics. 71-78. Neopanorpa nipalica (Navas). 71, genital bulb, holotype of nipalica, ventral aspect; 72, genital bulb, holotype of mzpalica, right lateral aspect; 73, left dististyle, holotype of suffusa, ventral aspect; 74, aedeagus, male, ventral aspect (lower scale); 75, tip of right ventral paramere, male, lateral aspect (lower scale); 76, third and fourth terga, holotype of suffusa, left lateral aspect; 77, third and fourth terga, holotype of suffusa, dorsal aspect; 78, third and fourth terga, holotype of nipalica, dorsal aspect. brief redescription is based on the notes and drawing this specimen: Heap: Dorsum shiny black; rostrum shiny black, except lateral margins red- dish brown. Antennal flagella missing. Tuorax: Dorsum shiny black. Pleural areas and coxae pale grayish brown, ex- cept a lateral black stripe on each coxa. Femora and tibiae yellowish brown, ex- cept apices of tibiae brown; tarsi light yellowish brown at base, grading to brown at apex. Wing membrane hyaline; mark- ings dark smoky brown. Apical band complete, unusually wide. Pterostigmal band wide, forked posteriorly, distinctly fused to apical band along costal and hind margins. Marginal spot broadly fused to proximal fork of pterostigmal band. Basal band complete, unusually wide. Basal spot large, extending from costal margin to Cup. ABDOMEN OF Femate: Terga 2-11 shiny black. Sterna 2-8 pale grayish brown. Apex of subgenital plate slightly notched 64 Tue UNIversity oF KANsAS SCIENCE BULLETIN 72 je. Me Fics. 79-80. Neopanorpa ocellaris (Navas). 79, genital plate, female holotype, ventral aspect (lower scale); 80, subgenital plate, ventral aspect. (fig. 80). Arms of genital plate twisted, spatulate (fig. 79). Axial portion of geni- tal plate short, thick, with large rounded lobes at base on each side, giving it a broadly triangular shape. Mate: Unknown. LENcTH oF Fore Wine: Female holotype, 14 mm.; hind wing 13 mm. Hotoryre: Female, Sikkim, 1890, Har- mand; in the Museum National d’His- toire Naturelle, Paris. When the male is discovered, the pat- tern of the wing markings, the uniformly shiny black dorsum of the thorax, and the black abdomen should permit recognition. On the basis of broad, interconnected Fic. 81. Neopanorpa salat Navas, male, right wings (fore wing length 12.0 mm.). Tue MeEcopTerA oF INDIA AND ApJACENT REGIONS 65 bands of the wing pattern, this species somewhat resembles N. ornata Byers from Vietnam and superficially N. harmandi Navas from Vietnam and Thailand (see Byers, 1965). But ornata differs from ocel- laris in having a distinctly bi-colored thorax and brownish abdominal terga. The notched apical band and the un- branched pterostigmal band of harmandi differ from those of ocellarts. In ocellaris, the gential plate resembles those of some Indo-Chinese species in hav- Fics. 82-89. Neopanorpa salai Navas. 82, genital bulb, male holotype, ventral aspect; 83, genital bulb, male holotype, right lateral aspect; 84, third and fourth terga, male, left lateral aspect; 85, ninth tergum, male, dorsal aspect (cerci, shown by dashed lines, do not extend beyond margin of ninth ter- gum); 86, tip of right hypovalve, male, right lateral aspect; 87, genital plate, female, ventral aspect; 88, genital plate, female, right lateral aspect; 89, subgenital plate, female, ventral aspect. Scale a— figs. &2- 85, 89; b—figs. 86-88. 66 Tue UNIVERSITY OF KANSAS SCIENCE BULLETIN ing long spatulate arms and rounded lat- eral lobes, and in lacking basal (anterior) apodemes. In several Chinese species of Neopanorpa, the anterior apodemes are absent, but the shape and size of the basal plate differ from those of ocellaris. Neopanorpa salai NavAs Neopanorpa salai Navas, 1929: 196. Navas described salai from the male holotype only. This species very closely resembles N. appendiculata and N. indica in general appearance. Until another male and a female of salat were discovered among specimens identified as N. appen- diculata, few characters were known that separated them. There are noticeable dif- ferences in the shape of the basal lobes of the dististyles, hypovalves, and_ posterior process of the third tergum between males of salai and appendiculata. This composite redescription is based on notes and drawings of the holotype, one other male, and two females: Heap: Vertex, frons, genae, and_ ros- trum dark yellowish brown. Antennal scape and pedicel dark brown, flagellum blackish brown with 37-38 segments (hol- otype, 38). THorax: Dorsum dark sordid brown. Pleural areas and legs sordid yellowish brown. Wing (fig. 81) membrane hyaline; longitudinal veins brownish, except in dis- tal hyaline spot pale; cross-veins brownish; markings dark brown. Apical band com- plete, notched in proximal margin or mar- gin straight and hyaline spot included (holotype). Pterostigmal band wide, forked posteriorly with branches sub-equal in width. Marginal spot large, extending from costal margin to first fork of media, fused to pterostigmal and basal bands. Ba- sal band wide, complete. Basal spot large. ABpoMEN OF Mate: Terga 2-9 and cor- responding sterna blackish brown. Pos- terior process of third tergum (fig. 84) long, extending one-third distance across sixth tergum; basal one-third wide, dorso- ventrally flattened with ventral tuft of hairs; abruptly becoming cylindrical and tapering to a point. A low sharp ridge across base of hypovalves (figs. 82, 83) with distinct notch at midpoint. Hypo- valves somewhat rounded with two long projections (of length sub-equal to width of hypovalves at base) in distal one-half; both proximal, mesal projection and distal ventromesal projection covered with black hairs longest at apex; tips of hypovalves flattened, tapered, directed posterodorsad (into interior of genital bulb), with row of flattened black spines along outer (pos- teroventral) margin (fig. 86). Entire dis- tistyle covered with blackish hairs longest on basal lobe; basal lobe broad, its bluntly rounded apex directed ventrad. ABDOMEN OF FEMALE: Terga and sterna 2-6 blackish brown, 7-9 dark sordid red- dish brown. Subgenital plate (fig. 89) broad, about three-fourths as wide as long, with long hairs along margin. Genital plate (figs. 87, 88) about 1.25 mm. long; arms of genital plate twisted near base, about one-half total length; anterior apo- demal process short, about one-fourth of total length. LenctH oF Fore Wine: Male, 11.3-13.0 mm. (holotype, 13.0 mm.); female, 12.8 mm. Ho.otyre: Male, Khandala, India, date unknown; in the Museum National d’His- toire Naturelle, Paris. In addition, the following specimens were examined: 1 male, Bombay, India, and 1 female, In- dia, place and date unknown, both in the British Museum (Natural History), Lon- don; 1 female, Coimbatore, Madras, South India, October 1951, in the Naturhisto- risches Museum, Basel. Neopanorpa sordida (NEEDHAM) Panorpa sordida Needham, 1909: 196, figs. L607 Tue MecopTErA oF INDIA AND ADJACENT REGIONS 67 Neopanorpa sordida; Esben-Petersen, 1915: 230. This species is still known only from one male and two female syntypes from Assam. The original description is some- what vague, and unfortunately the syn- types could not be obtained from the In- dian Museum for further study. Accord- ing to Needham, the male and one of the female syntypes have unmarked wings. The other female has diffuse bands near the middle of the wing, posterior to the stigma, and distal to the stigma. It seems likely that the more darkly marked female represents a second species. The following is based entirely on the original description: Heap: Dorsum pale brown. Antennae about 10 mm. long (female), their bases tawny yellow. Tuorax: Pronotum pale brown. Meso- and metanotum reddish yellow. Femora, tibiae, and tarsi tawny yellow with brown- ish apices. Wing membrane smoky-tinged, hyaline, unmarked (1 male and 1 female), longitudinal veins brown; or diffuse pterostigmal band forked _ posteriorly, oblique apical spot beyond stigma (1 fe- male). AspoMEN OF Mate: Basal segments pale brown, terminal segments tawny. Pos- terior process of third tergum extends half length of fourth tergum. Hypovalves extend to bases of dististyles (apices bro- ken off). LenctH oF Bopy: Male, 11 mm.; female, 10 mm. Expanse of wings 22 mm. Syntyres: 1 male and 2 females, Khasi Hills, Assam, 5000 feet, May (year un- known); in the Indian Museum, Calcutta. Neopanorpa zebrata EspeN-PETERSEN Neopanorpa zebrata Esben-Petersen, 1915: 228. This species was originally described from a single female from Travancore, India. A photograph of the wings of the holotype appears in Esben-Petersen’s world monograph (1921). With permission of Mr. D. E. Kimmins of the British Mu- seum, the terminal abdominal segments of the holotype were dissected. Several male specimens in the British Museum, tentatively determined to be zebrata, are from Periyar Dam, Travancore State, India. The distal wing markings of the male examined are somewhat smaller and paler but clearly correspond in position to those of the female. In addition, the body color- ation of this male is nearly identical to that of the female holotype. This male was collected together with several females of zebrata. The following redescription is based on the original description, notes on the female holotype, and one male: Heap: Dorsum, frons, and rostrum sor- did light yellowish brown. Antennal scape and pedicel sordid yellowish brown; flagel- lum (broken in holotype) blackish brown, with 38 flagellomeres (male). Tuorax: Short pilosity and four long, black setae on anterior margin of pro- notum. Dorsum, pleural areas, and coxae sordid yellowish brown. Femora, tibiae, and tarsi yellowish, with apices of tarsi brown. Wing membrane with light yel- lowish tinge; markings light brown (male) to smoky brown (holotype) ; cross- veins pale in distal portion of wing, brown in proximal portion. Apical band complete, with a large clear spot extend- ing from Ri to Mi in female, absent in male; large notch on proximal posterior margin (holotype), absent in male. Ptero- stigmal band complete, forked posteriorly, with basal fork constricted at base of Mi and Me (holotype), or incomplete, with posterior branches reduced to two spots (male). Marginal spot large, fused with pterostigmal band in holotype, but not so fused and extending to vein M in male. Basal band complete, constricted at vein 68 Tue UNIVERSITY OF KANSAS SCIENCE BULLETIN M (holotype). Basal spot large (holo- type), absent in male. AspoMEN oF Mate: Terga 2-4 blackish brown, 5-6 brown, 7-9 sordid yellow- ish brown. Sterna 2-9 sordid yellowish brown. Posterior process of third tergum (figs. 93, 94) approximately half length of fourth tergum, subquadrate in dorsal aspect, somewhat bulbous at apex; base blackish brown fading to yellowish brown posteriorly. Lateral margins of hypovalves (figs. 90, 96) convex in distal one-fourth; distal half of inner margin with two mesal protuberances, one near mid-length with two setae at apex and one more distal with minute nodules on surface, both with long hairs on mesal surface (fig. 96). Tergum 9 (fig. 92) abruptly narrower in distal portion, apex truncate. Dististyles slightly concave on outer surface near mid-length; basal lobe directed ventro- mesad, its outer surface with long hairs; prominence at inner base of dististyle (dor- sal aspect, fig. 95) with tuft of stout, black setae. Aedeagus (fig. 95) with mesally concave blades extending caudad from lat- eral processes; blades broad basally, taper- ing to bluntly rounded apex. Ventral parameres extend between lateral processes and ventral valves, apex of ventral para- mere truncate. Dorsal parameres com- pressed oval blades, narrowly connected at base to lateral processes, posterior edges divergent (fig. 95). ABDOMEN OF FeEMate: ‘Terga_ blackish brown, except apical segments brown. Sterna brown. Subgenital plate (fig. 97) strongly keeled ventrally and notched at apex. Arms of genital plate (figs. 98, 99) short, tips bluntly rounded; basal plate somewhat depressed. Apodemes long, nearly straight, slightly divergent near anterior apex. LenctH oF Fore Wino: Female holotype, 13 mm.; male, 11.6 mm. LenctH oF Hinp Winc: Female, 12 mm. Hotoryre: Female, Travancore, date un- known, Hampson; in the British Mu- seum (Natural History), London. In addition to the holotype, the follow- ing specimens are in the British Museum: 3 males, 7 females, Thekkadi, Periyar Dam, Travancore, India, 6-10 May 1937; 1 male, 1 female, Naraikadu, Tinnevelly District, South India, 2500-3000 feet, 3-8 May 1938; 1 male, 1 female, Pirmed, Tra- vancore, 3400 feet, 4-6 May 1937. The il- lustrations of the male were made from one of the series from Thekkadi, Periyar Dam, now in the Snow Entomological Museum, by exchange. This species can easily be recognized by the large, clear spot in the apical band, especially evident in the female, and by the mesal projections of the inner margins of the hypovalves in the male. The distal wing markings of two Chinese species, N. chelata Carpenter (1938) and N. pitelina Navas (1936), superficially resemble those of zebrata in having a large, clear spot in the apical band, but the spot is not nearly as large or centrally situated as in zebrata. In addition, these Chinese species lack the extensive basal spots that are present in the holotype of zebrata. On the basis of male and female geni- talia, zebrata closely resembles the other southern Indian species, especially N. den- ticulata n. sp. In both of these species, the aedeagus bears long, blade-like ventral parameres and fused lateral processes and dorsal parameres. Aside from the south- ern Indian species, N. burmana Byers (1965) from southern Burma is the only known species in which the lateral proc- esses and dorsal parameres have fused, but the ventral parameres of burmana (Byers, 1965) are not evident, while they are large in the southern Indian species. Usually the basal lobes of the dististyles are directed mesad, in Neopanorpa, but in zebrata, appendiculata, and indica, they are directed ventrad; but those of appen- Tue Mecoprera oF INDIA AND ADJACENT REGIONS 69 diculata and indica are nearly as broad as long, not slender and bluntly rounded as in zebrata. In zebrata, the female genital plate superficially resembles that of denticulata, oe 0.5mm Loe. 5 eS) i Eee ee ee | but differs in having shorter distal arms and in being more dorsoventrally flat- tened. The apodemal processes of zebrata are not as long or divergent as those of appendiculata and N. hirsuta. 99 Fics. 90-99. Neopanorpa zebrata Esben-Petersen. 90, genital bulb, male, ventral aspect; 91, genital bulb, left lateral aspect; 92, ninth tergum, dorsal aspect; 93, third and fourth terga, left lateral aspect; 94, third and fourth terga, dorsal aspect; 95, aedeagus and basal lobe of left dististyle, ventral aspect; 96, tip of left hypovalve, ventral aspect; 97, subgenital plate, female holotype, dorsal aspect; 98, genital plate, holotype, ventral aspect; 99, genital plate, right lateral aspect. Scale a— figs. 90-94, 97; b— figs. 959659985) 99): 70 Tue UNIVERSITY OF KANSAS SCIENCE BULLETIN Fic. 100. Neopanorpa denticulata n. sp., male paratype, right wings (fore wing length 10.8 mm.). eopanorpa denticulata ’ SPECIES Neopanorpa denticulata NEw SPECIE Description based on 150 males and 121 females pinned, 1 male and 13 females in alcohol. Heap: Dorsum, rostrum, frons, and genae sordid yellowish brown (holotype) to dark yellowish brown. Antennal scape yellowish brown; pedicel yellowish brown to dark yellowish brown (holotype); basal portion of first flagellomere dark yellowish brown, remaining flagellum black, with 39-42 flagellomeres (holotype, 42). Tuorax: Short pilosity but no setae on anterior edge of pronotum. Dorsum sor- did yellowish brown (holotype) to sor- did dark brown; pleural region and coxae yellowish brown (holotype) to dark yel- lowish brown. Femora and tibiae sordid yellowish brown; tarsi sordid yellowish brown grading to black on segments 4-5. Wing (fig. 100) membrane hyaline, mark- ings brown (holotype) to dark smoky brown. Apical band complete with clear spot or notch at Mi and Mb on proximal hind margin (holotype). Pterostigmal band complete, forked posteriorly. Mar- ginal spot variable, sometimes fused with pterostigmal band or with basal band. Basal band variable, partially or com- pletely divided (holotype) into two trans- verse bands in fore wing (figs. 13, 100). Basal spot present. ABDOMEN OF Mate: Terga 25 black; anterior one-third of sixth tergum black, fading posteriorly to dark yellowish brown; terga 7-9 sordid yellowish brown. Sterna 2-8 sordid yellowish brown. Pos- terior process of third tergum (fig. 105) somewhat triangular, rounded apically, extending half length of fourth tergum. Hypovalves (figs. 101, 104) slightly wid- ened in apical half, then narrowed and abruptly thickened at apex; a mesal thick brown spine on apical enlargement, an- other on mesal margin just before apex, and a third on dorsomesal margin just beyond mid-length of hypovalves; hypo- valves with uniform brown pilosity. Ninth tergum (fig. 106) abruptly narrowed in distal one-fourth, terminating in shallowly emarginate apex with numerous long, brownish hairs. Basal lobes of dististyles directed ventrad (fig. 102), shallowly cupped on dorsomesal surface, with bluntly rounded apices; ventral posterior surface with long blackish hairs; tuft of stout black setae at the dorsal base of lobe. Aedeagus complex (fig. 103), with blade- Tue Mecoprera oF INDIA AND ApJAcENT REGIONS 71 103 105 106 Fics. 101-106. Neopanorpa denticulata n. sp. 101, genital bulb, male paratype, ventral aspect; 102, genital bulb, right lateral aspect; 103, aedeagus, ventral aspect (vertical scale); 104, tip of left hypovalve, ventral aspect (vertical scale); 105, third and fourth terga, dorsal aspect; 106, ninth tergum, dorsal aspect. like ventral parameres and fused lateral processes and dorsal parameres extending between bases of dististyles. Dorsal para- meres enlarged, blade-like, directed dorsad. ABDOMEN OF FEMALE: Terga 2-6 black, 7-9 dark yellowish brown. Sterna 2-8 yellowish brown. Subgenital plate (fig. 107) notched apically, with elongate hairs along margins and at apex. Arms of geni- tal plate (figs. 108, 109) short, about one- fifth of total length, slightly curved in- ward. Apodemes about .3 total length of plate, fused for about half their length, diverging anteriorly. Basal and distal plates broad, with lateral margins of distal plate folded ventrad. LENcTH oF ForeE Winc: Male, 10.0-12.0 mm. (holotype, 11.0 mm.); female, 10.3- 12.6 mm. (allotype, 12.4 mm.). Lenctu oF Hinp Wine: Male, 9.3-10.6 mm (holotype, 10.0 mm.); female, 10.0- 11.0 mm. Hototyre: Male and allotype, female, Devala, Nilgiri Hills, India, 3200 feet, May 1961, P. S. Nathan. Paratypes, 150 males, 133 females: at type locality, May 1961 (120 males, 63 females); type locality, October 1960 (5 males, 6 fe- males); Walayar Forest, Kerala, India, 700 feet, P. S. Nathan, September 1959 (2 males, 9 females), 15 September 1959 (1 male), October 1959 (4 males, 13 females), i Tue UNIVERSITY oF KANSAS SCIENCE BULLETIN August 1960 (2 males), September 1965 (1 female); at Chembra Peak area, India, 3500! feet, i. R. S: (Nathan, May 1970°G males, 6 females), October 1970 (1 fe- male); all in the Snow Entomological Museum; Walayar Forest, S. Malabar, In- dia, 1000 feet, P. S. Nathan, August 1957 (1 male, 9 females), November 1957 (4 females), in the Illinois Natural History Survey collection, Urbana, Illinois; Muthi- kolam, Coimbatore Dist., India, 3000 feet, 23-26 September 1938 (1 male, 1 female); Siruvani, 1700 feet, 23-26 September 1938 (1 male), in the British Museum (Nat- ural History), London; Cherangoda, Nil- girt Hills, 1200 m., November 1954 (7 males, 14 females), October 1954 (1 male, 3 females), October 1950 (1 male), no Fics. 107-109. Neopanorpa denticulata n. sp. 108, genital plate, ventral aspect; 109, genital plate, right lateral aspect. Scale a— fig. 107; b— figs. 108, 109. date (1 male); Coimbatore, Madras, South India, October 1951 (1 male), in the Naturhistorisches Museum, Basel, Switz- erland. This species superficially resembles in general appearance two regional species, N. zebrata and hirsuta; and the aedeagus resembles those of hirsuta, echinata n. sp., and especially zebrata, in having long, blade-like ventral parameres and fused dorsal parameres and _ lateral processes. The lateral processes and dorsal parameres are also fused in N. burmana Byers (1965) from Indo-China, but burmana_ differs from these Indian species in lacking the long, blade-like ventral parameres. In den- ticulata, the dorsal parameres are large, slightly divergent, vertical plates similar E E ° ° 2 oO 107, subgenital plate, female paratype, ventral aspect; Tue MecopTera oF INDIA AND ADJACENT REGIONS 73 to those of zebrata, but these plates have a broader connection to the body of the aedeagus than in zebrata. The structure of the aedeagus is the primary character used in grouping /ursuta, echinata n. sp., denticulata, and zebrata into the denticu- lata group. It is called the denticulata group after the commonest included spe- cies. The basal markings of the wing, the solidly-colored, complete apical band, and the three tooth-like projections on the in- ner margin of the hypovalves permit easy recognition of denticulata. The specific name refers to the three projections of the hypovalves (Latin denticulus—=small tooth). Some variation in wing pattern, especially the fusion of the two basal bands, was observed in the paratypes from the Walayar Forest. On the basis of female genitalia, den- ticulata superficially resembles N. zebrata in having large, broad basal and distal plates and apodemes subequal in length to the basal and distal plates; but it differs from zebrata in having longer arms of the distal plate and a somewhat flattened subgenital plate, not strongly keeled as in zebrata. Neopanorpa echinata New SPEcIEs Description based on 2 males, pinned. Heap: Dorsum sordid yellowish brown; frons sordid yellowish brown, with dark brown medial spot; rostrum sordid yel- lowish brown, edges becoming translucent yellowish brown. Antennal scape sordid yellowish brown; pedicel dark yellowish brown; flagellum blackish brown, with 41-42 flagellomeres (holotype, 42). TuHorax: Short pilosity but no setae on anterior margin of pronotum. Pro- notum blackish brown. Medial portion of mesoscutum sordid grayish brown, color extending along midline to scutellum; lat- eral margins of scutum dark brown. Meta- notum similar to mesonotum. Pleural areas and coxae light grayish brown. Fe- mora and tibiae sordid light brown. Basi- tarsi sordid brown grading to dark brown on segments 4-5. Wings hyaline, except light yellowish brown stigmal spot. Longi- tudinal veins brownish; cross-veins brown in proximal portion of wing, pale in distal portion. ABpoMEN OF Mate: Terga 2-5 blackish brown, 6 dark yellowish brown, 7-9 sor- did grayish brown. Sterna 2-9 sordid grayish brown. Posterior process of third tergum (figs. 116, 117), triangular, ex- tending across one-third to one-half length of fourth tergum. Notal process of ter- gum 4 with short black hairs on an- terior two-thirds. Segments 7 and 8 short (fig. 111), each with a dorso-medial trough in its posterior one-half to one- third. Base of hypovalves broad, forming a U-shaped ridge, bearing blackish hairs; hypovalves sinuous (figs. 110, 112), apices with numerous small black spines and long black hairs. Basal lobe of dististyles with concave dorsal surface and long hairs on ventral, posterior surface. Ventral parameres (fig. 113) slender, wedge-shaped apices directed ventrad. Dorsal parameres short, rounded protuberances. Lateral processes of aedeagus long, concave, blade- shaped. FemaLe: Unknown. LENGTH OF ForE Winc: Male, 11.0 mm. (holotype); hind wings, 9.7-10.0 mm. (hol- otype, 10.0 mm.). Hototyre: Male, Naduvatam, Nilgiri Hills, India, 6000 feet, May 1958, P. S. Nathan. Holotype and 1 male paratopo- type are in the Snow Entomological Mu- seum. This species can be easily recognized from other regional species by the absence of wing markings, except for a faint stig- mal spot, and the narrow, sinuous hypo- valves with spiny apices. The specific name (Latin echinatus=spiny, prickly) 74 THe UNIVERSITY OF KANSAS SCIENCE BULLETIN t= 0.5 mm fis Fics. 110-117. Neopanorpa echinata n. sp. 116 110, genital bulb, male holotype, ventral aspect; 111, terminal abdominal segments, right lateral aspect; 112, tip of left hypovalve, ventral aspect (lower scale); 113, aedeagus, ventral aspect, part of right side omitted (lower scale); 114, tip of left ventral paramere, lateral aspect (lower scale); 115, ninth tergum, ventral aspect; 116, third and fourth terga, dorsal aspect; 117, third and fourth terga, left lateral aspect. refers to the spiny apices of the hypo- valves. On the basis of the aedeagus, this spe- cies was placed in the denticulata species group. The lateral processes are elongated as in all the other species in this group, but the dorsal parameres of echinata are not fused with the lateral processes as in denticulata and zebrata. The dorsal para- meres are not distinguishable in Azrszta. The posterior process of the third tergum of echinata is triangular, not narrow with subparallel sides as in the other species of the group. The broad, thick base of the hypovalves and the ventral parameres of echinata are similar in appearance to those found in several species of the appendicu- lata group. Neopanorpa gibbosa New SPECIES Description is based on one male, pinned. Heap: Vertex black, rostrum dark sor- did yellowish brown, frons and antennal scape and pedicel sordid yellowish brown; antennal flagellum blackish brown, with Tue MecopTera oF [Np1A AND ADJACENT REGIONS 75 43 flagellomeres (left flagellum missing). TuHorax: Pronotum blackish brown. Meso- and metanotum blackish brown (metanotum partly obscured by pin), pleural areas grayish brown. Wing mem- brane hyaline; longitudinal veins brown- ish, cross-veins brown in proximal half and pale in distal half of wing; markings light brown. Apical band complete. Ptero- stigmal band constricted near mid-length, forked posteriorly. Marginal spot a nar- row band extending from costal margin to basal branch of pterostigmal band. Ba- sal band a small spot near the base of Rs. Basal spot absent. ABDOMEN OF Mate: Terga 2-3 blackish brown, posterior margin of third ter- gum and base of posterior process of third tergum with long yellowish hairs, 4-9 yellowish brown. Sterna 2-9 yellowish brown. Posterior process of third tergum (fig. 119) blackish brown, curved down- ward on fourth tergum, extending across entire length of fourth tergum. Posterior \ \\ == y) | NY Wane se z Q ne 119 j | ih) 120 Ate I 1 mm J Fics. 118-120. Neopanorpa gibbosa n. sp. 118, terminal abdominal segments, male holotype, left lateral aspect; 119, terga 3-5, dorsal aspect; 120, posterior portion of sixth abdominal segment, dorsal aspect. 76 Tue UNIVERSITY oF KANsAs SCIENCE BULLETIN Gwe. 121 124 126 Fics. 121-126. Neopanorpa gibbosa n. sp. 121, genital bulb, male holotype, ventral aspect; 122, distal portion of ninth tergum, left lateral aspect; 123, left hypovalve, left lateral aspect, and apical portion (mesal aspect); 124, aedeagus, left half, ventral aspect (right scale); 125, base of left dististyle, dorsal aspect (right scale); 126, base of right dististyle, ventral aspect. half of sixth tergum enlarged, curved dor- sad into large, broad hump (figs. 118, 120). Posterior half of seventh segment somewhat enlarged with long hairs near posterior margin. Ninth tergum with a median projection extending dorsad near mid-length, with short black pilosity; apex bluntly truncate (figs. 118, 122). Basal petiole of each hypovalve with heavily sclerotized, rounded mesal projection and narrowly triangular dorsal projection; hy- povalves broad, thin, laterally flattened lobes. Inner ventral margin of basistyle forms a distinct ridge with a branch lead- ing into interior of bulb. Basal lobe of dististyles cupped mesally, with hair-coy- ered protuberance on posteroventral sur- face (fig. 126) and 3-4 thick black, dorsal spines (fig. 125). Ventral parameres of aedeagus short, slender, sinuate, conver- gent at apices. Ventral valves large, thin blades. Dorsal parameres, long, narrow, bluntly pointed blades. Tue MeEcoprera oF INDIA AND ADJACENT REGIONS 77 Fic. 127. Neopanorpa indica n. sp., male paratype, right wings (fore wing length 11.3 mm.). LenctH oF Fore Winc: Male holotype, 12.3 mm.; hind wing, 11.0 mm. Ho.otyre: Male, Simla, India, from the McLachlan collection in the British Museum (Natural History), London. On the basis of wing pattern and body coloration, the female should be recog- nized when discovered. The modification of the sixth tergum and the dorsal pro- jection of the ninth tergum of the male are diagnostic of this species. The specific name (Latin gibber, humped, hump- backed, protuberant) refers to this modi- fication. The projection on the ninth ter- gum is unique among known species of Panorpidae. N. cornuta and N. furcata have a modified sixth tergum, but in both species it is extremely different from that of gibbosa. The aedeagus of gibbosa, es- pecially the ventral valves, is similar to that of other northern Indian and Indo- Chinese species. Neopanorpa indica New Species Description is based on 14 males and 2 females, pinned. Heap: Vertex, rostrum, genae, and frons sordid yellowish brown. Dorsum of an- tennal scape yellowish brown, venter brown; pedicel blackish brown; flagellum black with 39 flagellomeres. TuHorax: Short pilosity, no setae on anterior margin of pronotum. Dorsum sordid dark brown. Pleural areas black. Coxae and femora blackish brown, tibiae pale yellowish brown, tarsi yellowish brown grading to black on segments 3-5. Wing (fig. 127) membrane hyaline; mark- ings smoky brown. Apical band complete with small notch on proximal margin. Pterostigmal band complete, forked pos- teriorly. Marginal spot large, extending from costal margin to behind Ri+5. Basal band complex and variable. Basal spot large. ABDOMEN OF Mate: Terga 2-9 and cor- responding sterna black. Posterior proc- ess of third tergum narrow (figs. 130, 131), extending half to three-fourths length of fourth tergum, sordid yellowish brown to dark brown, with bluntly rounded apex. Notal process of fourth tergum (fig. 130) with a tuft of short black hairs. Sternum 9 broad at base (fig. 129), forming interrupted transverse ridge. Hypovalves (fig. 129) stout, rounded ba- sally becoming compressed laterally in dis- tal one-third; small protuberance on mesal 78 THe UNIVERSITY OF KANSAS SCIENCE BULLETIN margin sparsely covered with long black hairs; posteroventral edge of apex with a row of flattened black setae. Dististyles (fig. 129) about .3 of total length of geni- tal bulb, with large basal lobes directed ventrad; apices of basal lobes bluntly rounded, dorsal surface strongly convex, covered with long, dense black hairs. Ven- tral parameres of aedeagus narrow, darkly sclerotized, diverging and becoming some- what compressed laterally in distal one- fourth; apices with short white pilosity. Posterior margin of lateral process form- ing a thin, finger-like process. Dorsal parameres stout, apices curved mesad. Cerci short, extending to posterior margin of ninth tergum. ABDOMEN OF FEMALE: Terga 2-9 and cor- responding sterna black. Subgenital plate (fig. 134) broad, nearly as wide as long, with shallowly notched apex, ventral keel, and long hairs along periphery. Genital plate (figs. 132, 133) long, about 13 mm.; arms of distal plate twisted, slightly narrowed basally; apodemes about one-third total length of plate, diverging anteriorly, with thick interconnecting lamella. LENGTH OF FORE WING: Male, 12.2-13.4 tama a(nolotype, 134): 7% female, 13.7-149 mm. (allotype, 13.7). Hototyre: Male, Gudalur, Nilgiri Hills, 3500 feet, September 1960, P. S. Nathan. Allotype, 3 male and 1 female paratypes. Bombay, India, date and collector un- known; in the British Museum (Nat- ural History), London. Eight male para- types: Nilgiri Hills, May 1954 (5 males), October 1950 (1 male), Cherangoda, 1200 m., no date (2 males), also 26 females not paratypes (same dates), in Naturhis- torisches Museum, Basel. Holotype and 3 male paratopotypes in the Snow Ento- mological Museum. In wing pattern, this species closely re- sembles N. appendiculata and N. salat, but it differs from those species in the shape of the hypovalves and the posterior process of the third tergum. The mesal projection of the hypovalves is short in indica, not long as in appendiculata and salai. The posterior process of the third tergum in indica extends midway or a little more across the fourth tergum, not beyond the posterior margin of the fifth tergum as in appendiculata and sala. The position and shape of the dorsal parameres of indica and appendiculata are very sim- ilar to those of N. burmana Byers (1965) from Burma, especially in dorsal aspect. The genital plates of the females of appendiculata, indica, and salai are simi- lar, and these females have been identified in most cases by their association in col- lections with males. The heavily pig- mented wing of sala: distinguishes that species from appendiculata and indica. The relative length of the apodemes to the total genital plate length in zndica 1s longer than in appendiculata and _ salat and may be used to separate the species. Neopanorpa ochrura New SPECIES Description based on 1 male and 1 female, pinned. Heap: Vertex black; frons black me- dially, sordid yellowish brown beneath antennal sockets; rostrum blackish brown with lateral margins brown (holotype) to dark sordid yellowish brown (allotype). Antennal scape and_ pedicel brownish black (flagella missing from both types). Tuorax: Pronotum black. Anterior half of mesoscutum and wide median streak, including mesoscutellum, black; postero- lateral areas of mesoscutum brown. Me- tanotum similar to mesonotum. Pleural areas grayish brown to blackish brown. Femora, tibiae, and tarsi brownish with apices dark brown. Wing (fig. 135) mem- brane hyaline, markings light brown (holotype) to brown (allotype), stigma slightly darker; longitudinal veins brown, cross-veins brown in basal portion of Tue MecopTera oF INDIA AND ADJACENT REGIONS 79 G.ws&. 129 130 Wey 131 134 0.5 mm Fics. 128-134. Neopanorpa indica n. sp. 128, terminal abdominal segments, male holotype, left lateral aspect; 129, genital bulb, holotype, ventral aspect; 130, third and fourth terga, left lateral aspect; 131, third and fourth terga, dorsal aspect; 132, genital plate, female allotype, ventral aspect (lower scale); 133, genital plate, right lateral aspect; 134, subgenital plate, ventral aspect (lower scale). wing, pale in distal portion. Apical band complete, witth several hyaline spots sur- rounding cross-veins, broadly joined to distal branch of pterostigmal band along hind margin. Pterostigmal band wide, forked posteriorly. Marginal spot not at- taining costal margin. Basal band nearly complete. AppoMEN oF Mate: Terga 2-6 blackish brown, 7-9 yellow ocher. Sterna 25 brown, 6 blackish brown, 7-9 yellow ocher. Pos- terior process of third tergum (fig. 141) short, extending across one-third length of fourth tergum; pale area on fourth tergum beneath process of third. Basistyles not deeply divided. Hypovalves 80 THe UNIVERSITY OF KANsAs SCIENCE BULLETIN Fic. 135. Neopanorpa ochrura n. sp., male holotype, right wings (fore wing length 11.3 mm.). (figs. 136, 139) narrow at base, broadest near mid-length, apices narrowed and arched ventrad. Basal lobe of dististyles (fig. 138), strongly convex on dorsomesal surface, covered with long hairs ventrally. Basal one-half of ventral parameres (fig. 138) concealed under basistyles, tips bent mesad and converging, apical one-fourth contiguous. Ventral aedeagal valves small, projected ventrad. Lateral processes short. Dorsal parameres absent. ABDOMEN OF FEMALE: Terga 2-7 blackish brown, 8-10 yellow ocher, cerci black. Sterna 2-6 grayish brown, 7-8 light ocher. Apex of subgenital plate (fig. 142) deeply notched, with many long yellowish hairs along margin and at apex. Axial por- tion of genital plate (figs. 143, 144) about 4 of total length of genital plate, stout, about .5 width of genital plate. Distal arms of genital plate broad, spatulate, twisted at base. LenctH oF Fore Winc: Male holotype, 11.3 mm.; female allotype, 11.6 mm. LenctH oF Hinp Wine: Male holotype, 10.0 mm.; female allotype, 10.6 mm. Ho.otyre: Male, Simla, India, date and collector unknown. Allotype, female, same data as for holotype; both types in the British Museum (Natural History), Lon- don. The wing pattern of this species closely resembles that of N. cornuta, but ochrura has a different color pattern on the nota and no modification of the sixth tergum in the males. The nota are especially use- ful in distinguishing the females of these two species. The anterior half of the meso- notum of cornuta is dark brown and the posterior half is yellowish, without a me- dian blackish stripe as in ochrura. In cornuta, there is a small blackish semi- circular spot along the anterior margin of the metanotum and no median stripe as in ochrura. In ochrura, the female genitalia, espe- cially the genital plate, resemble those of several Indo-Chinese and northern Indian species in having large spatulate distal arms and lacking elongate apodemes. The genital plate of ochrura resembles that of cornuta but differs in being larger (about 0.7 mm.) and having a_ proportionately longer axial portion. The aedeagus of ochrura is similar to that of many Indo-Chinese species, but the ventral parameres are much longer. The specific name (from Greek ochra, a yellowish brown or ochre color + ouwra, tail) refers to the color of the posterior abdominal segments in both the female and male. Tue MecopTera oF INpIA AND ApJACENT REGIONS 8] Gw.8. 136 bu 0.5 mm 143 144 Fics. 136-144. Neopanorna ochrura n. sp. 139 140 \ i) . 141 137 136, genital bulb, male holotype, ventral aspect; 137, genital bulb, right lateral aspect; 138, aedeagus and base of left dististyle, ventral aspect; 139, tip of left hypovalve, ventral aspect; 140, ninth tergum, dorsal aspect; 141, third and fourth terga, dorsal aspect; 142, subgenital plate, female, ventral aspect; 143, genital plate, ventral aspect; 144, genital plate, right lateral aspect. Scale a— figs. 136, 137, 140-142; b— figs. 138, 139, 143, 144. Neopanorpa ramulata Byers Neopanorpa ramulata Byers, 1975: 87-91, figs. 1-11. Since this species from Bhutan was re- cently described, only its more important and diagnostic characteristics are rede- scribed here, to permit comparison with other regional species. The following paragraphs are based upon the original description: Heap: Dorsum shiny black, rostrum pate brown with two diffuse longitudinal stripes. Antennae about 14 mm. long in male, 12 mm. in female, with 47-48 flagel- lomeres. Tuorax: Pronotum dark brownish black throughout, with only very short hairs 82 Tue UNIversITy oF KANSAS SCIENCE BULLETIN Fics. 145-151. Neopanorpa ramulata Byers. valves removed; 146, genital bulb, male holotype, left lateral aspect; 147, hypovalves, male paratype, ventral aspect; 148, apex of ninth tergum, paratype, dorsal aspect; 149, third and fourth terga, dorsal 145, genital bulb, male paratype, ventral aspect, hypo- aspect; 150, third and fourth terga, left lateral aspect; 151, subgenital plate, female allotype, ventral aspect. on anterior margin. Mesonotum and metanotum dark brownish black on an- terior half, along mid-line and on scutel- lum; dark yellowish brown spot adjacent to each wing base. Wines: Lightly tinged with brown; bands and spots light to dark smoky brown. Apical band darkest anteriorly, fading or absent behind vein Rs. Pterostigmal band nearly black over stigma, entire and forked, the branches fading toward pos- terior edge of wing. Marginal spot from near C to slightly behind Riis. Basal band absent, or rarely faintly indicated by diffuse spots. ABpoMEN OF Mate: ‘Terga 1-5 black, segment 6 dark brown anteriorly, yel- lowish brown on posterior one-third, seg- ments 6-9 yellowish brown except dis- tistyles dark brown. Posterior process of third tergum (figs. 149, 150) broad at base, slender and curved downward apically. Hypovalves (figs. 146, 147) somewhat rolled, convex ventrolaterally, concave Toe Mecoptera oF INDIA AND ADJACENT REGIONS 8 Fics. 152-155. Neopanorpa ramulata Byers. Us 152, aedeagus, male paratype, ventral aspect; 153, aedeagus, right lateral aspect; 154, genital plate, female allotype, ventral aspect; 155, genital plate, right lateral aspect. mesally, pale at tips, with lateral apical corner slightly extended and curled over ventral apex (fig. 146). Tergum 9 (fig. 148) narrowed and rounded at apex. Outer margins of dististyles slightly con- cave near mid-length (fig. 145); basal lobe of each dististyle prolonged ventrad, rounded and somewhat flattened, bearing numerous long hairs. Ventral parameres slender, darkly — sclerotized, forked near apex, one branch curving ven- trad, the other caudad, both covered with fine, short hairs (figs. 152, 153). Dorsal parameres greatly enlarged, expanded dor- sad and caudad as subparallel pale blades. ABDOMEN OF FEMALE: Terga 1-5 black, 6 dark brown, 7-9 yellowish brown, cerci black. Subgenital plate (fig. 151) slightly narrowed toward apex, broadly and shallowly notched at tip. Genital plate (figs. 154, 155) weakly sclerotized, with elongate anterior apodemes strongly deflected ventrad. elongate, LenetH oF Fore Wine: Male, 15.2-15.8 mm., female 14.2-15.0 mm. Hotortyre: Male, Tongsa (Tongsa Dzong, 27°33’N, 90°30’E), Bhutan, 2150 m., 24 June 1972, Naturhistorisches Museum Basel Bhutan Expedition. Allotype and 1 female paratype, same data as for holotype. One male, 1 female paratypes, Changra, 18 km. south of Tongsa, 1900 m., 22 June 1972. Holotype, allotype and 1 female paratype in Naturhistorisches Museum, Basel; 1 male, 1 female para- types in Snow Entomological Museum, University of Kansas. Neopanorpa ramulata resembles N. chillcotti Byers and N. ntpalica (Navas) in having slender ventral parameres and conspicuously developed dorsal parameres, in the male, and in the female a large genital plate, the axial portion of which is strongly sclerotized and bears divergent anterior apodemes. From these species, males of ramulata may be differentiated 84 Tue UNIVERSITY oF KANSAS SCIENCE BULLETIN by the peculiar, antler-like branched ven- tral parameres, by the shape of the hypo- valves, and shape of the basal lobe of the dististyles. In aedeagal structure of males and genital plate of females, ramulata approaches some species of Panorpa. SPECIES OF NEOPANORPA Nor DEscriBED Since most of the reliable taxonomic characters in females are on the subgenital and genital plates, dissections are usually necessary to identify species, especially to differentiate those having similar wing patterns. On the other hand, males have several excellent external taxonomic struc- tures. Association of males and females of a species has been based mainly on body coloration, wing pattern, body size, and date and place of collection. Six female specimens, probably repre- senting four new species, were not de- scribed because of lack of associated males. Two males, probably representing two ad- ditional new species, were not described because they were so badly damaged. Fol- lowing is a summary of information con- cerning these specimens: Two large-winged female specimens from Nepal externally resemble N. fur- cata, but the genital plates and color pat- tern of the thorax differ from those of furcata. These two females are in the collection of Canada Department of Ag- riculture. One badly damaged male specimen (missing genital bulb) from Darjeeling, India, generally resembles N. nzpalica. Another male from Darjeeling (missing hypovalves) has distinctly shaped disti- styles. Both males are in the Illinois Nat- ural History Survey collection. One female from India (no more de- tailed locality) resembles N. effusa but is smaller than effzsa (wings about 14 mm.). It is in the British Museum (Natural History). One female specimen from Chinchona, Anaimalai Hills, southern India, resem- bles N. appendiculata, but the basal half of the wing is much paler. This specimen is in Snow Entomological Museum. Two additional females from Changra, Bhutan, somewhat resemble N. ramulata but have yellow-tinged wings with more extensive markings and reduced genital plate more typical of other Neopanorpas of the Himalayan region. For further de- tails, see Byers (1975). DEscRIPTIONS OF SPECIES: Genus Panorpa Panorpa davidi Navas Panorpa davidi Navas, 1908: 415. Originally P. davidi was described from one male and one female. The fe- male syntype is now the holotype of P. guttata. The distinctive apical, pterostig- mal, and incomplete basal bands may al- low recognition of the female when dis- covered. Rarely is body coloration helpful in the recognition of a species, but the blackish head with reddish brown rostrum in davidi, and the blackish thorax with a reddish brown median streak, may be diagnostic. The original label on the holotype reads “David, Moupin, Thibet,” and a later one “Museum Paris, Mou-Pin, A. David 1870.” Since 1908, the borders of Tibet have shifted westward from about 102° E to their present position, about 95° E. No such locality as Mou-Pin appears in detailed gazetteers of Tibet proper, to- day, but in southwestern Szechwan, at 28° 48’ N, 103° 39’ E, about 50 kilometers east of the 1908 Tibetan border, there is the town of Mapien, whose pronunciation would resemble that of “Mou-Pin” in French. At an elevation of 2000 feet, it is in a more likely habitat for Panorpa than the higher, colder regions in modern Tibet. Esben-Petersen (1921), illustrated in the ventral aspect of the genital bulb, two separate appendages between the ventral Tue Mecoptera oF INpIA AND ApJACENT REGIONS 85 parameres. These, however, are merely basal thickenings of the ventral parameres. The following redescription is based in part on drawings of the holotype: Heap: Dorsum black, rostrum reddish brown. Antennal scape and pedicel red- dish brown; both flagella missing. Tuorax: Dorsum black, with longitu- 197 dinal median reddish streak; pleural areas black. Legs testaceous. Wing membrane with a faint yellowish tinge; longitudinal veins brownish; cross-veins in apical por- tion pale. Apical band incomplete, ending at margin between Rs and Mj; large spot extending from hind margin to cell Rs (fore wing only). Pterostigmal band con- 0.5 mm 158 159 Fic. 156. Panorpa davidi Navas, male holotype, genital bulb, ventral aspect. Fics. 157-159. Panorpa guttata Navas. 157, subgenital plate, female holotype, ventral aspect; 158, genital plate, ventral aspect (vertical scale); 159, genital plate, right lateral aspect. Fic. 160. Panorpa stigmalis Navas, male holo- type, genital bulb, ventral aspect. 86 Tue UNIVERSITY OF KANSAS SCIENCE BULLETIN tinuous, diagonal, not forked; distal branch not present. Marginal spot absent. Basal band incomplete, comprising three spots. AspoMEN oF Mate: Terga and sterna 2-6 black, 7-9 testaceous. Virtually no modification of terga 3-4 (i.e., notal or- gan poorly developed). No anal horn on sixth tergum. Basistyles (fig. 156) deeply separated ventrally; distal mesal margin of each with dense fringe of hairs. Dististyles slightly more than half length of basistyles; basal “lobe” shallowly cup- shaped, one-half length of inner margin of dististyle. Hypovalves narrow, strap- like, with rounded apices; extending to about two-thirds length of basistyles. Ven- tral parameres long, thick at base but tapering to sharp tips, setose on outer margins, extending to distal ridge of basal concavity of dististyle, sinuate, crossing near mid-length. Dorsal parameres thin, flattened dorsoventrally, strap-like, extend- ing beyond bases of dististyles. FemaALeE: Unknown. Lenctu or Fore Wine: Male holotype, 13 mm.; hind wing 11.5 mm. Hototryre: Male, Mou-Pin, Thibet (Ti- bet), 1870, A. David; in the Museum National d’Histoire Naturelle, Paris. Cheng (1957) based the davidi group, in which there is no anal horn on the sixth abdominal segment, on this species. The structure of the genital bulb, how- ever, is not similar to that of any other species assigned to this group. According to Cheng (1957), the wing markings re- semble those of P. chent Cheng from Chekiang, China; but the genital bulbs of these two species are quite different. Sev- eral Chinese species have extremely elon- gate dorsal parameres, especially P. wa- ongkehzengi Cheng from Kiangsi, China. The dorsal parameres in that species are somewhat similar to those of davidi, but the ventral parameres and hypovalves are much shorter in waonghehzengi. Unfortunately, 12 Chinese species are known only from their female holotypes. Eight of these can be eliminated from consideration as the possible female of davidi because they have forked pterostig- mal bands, and several of them also have basal spots. The pattern of the wings of P. leet Cheng and P. semtfasciata Cheng is similar to that of david: but differs in having the apical band somewhat reduced and the basal band absent. From the orig- inal description (Cheng, 1957), it seems that P. grahamana Cheng and P. carpen- tert Cheng have wing patterns nearly identical to that of davidi. The wing membrane in these two females is hyaline, not with a faint yellowish tinge as in davidt. Due to a somewhat vague descrip- tion of overall body coloration of these two females, it is impossible to compare them accurately with the male type of davidi. Cheng (1957) states that the only dif- ferences between grahamana and carpen- tert are the coloration of the rostrum and the bent apices of the two distinct apo- demes of the genital plate of carpentert. Future investigation may indicate that these two nominal species, both collected in Szechwan, are synonymous. And we think there is some likelihood that both are synonyms of davidi. Panorpa guttata NavaAs Panorpa guttata Navas, 1908: 416. Panorpa davidi Navas, 1908: 415 (in part). Navas originally described the holotype of guttata as the female of P. david (1908). In his closing statements, he sug- gested that the female be called gzttata it the differences in the wing markings were significant; therefore, Esben-Petersen (1915) recognized the female as the holo- type of guttata. The indistinct apical and pterostigmal bands of gwttata differ from those of davidi in being paler in color and much smaller. In gwttata, the thorax and abdomen are grayish brown, not reddish brown and black as in david. Tue MEcopTera oF INDIA AND ADJACENT REGIONS 87 The locality data on the original label are the same as for P. davidi: Mou-Pin, Tibet (see P. davidi for discussion). The following redescription is based on previ- ous descriptions and on notes and draw- ings of the holotype: Heap: Dorsum and rostrum grayish brown, except ocelli on blackened promi- nence. Antennal scape grayish brown, pedicel brownish black: flagellum pale grayish brown grading to dark brown towards apex. Tuorax: Dorsum pale brown, pleural re- gions pale grayish yellow. Legs brown- ish yellow, except apices of tibiae and tarsomeres brown. Wing membrane hya- line with a faint yellowish tinge, longi- tudinal veins brownish. Apical band in- complete, reduced to several spots in front wings. Pterostigma yellowish; pterostig- mal band incomplete. ABDOMEN OF FemaLe: Terga and sterna pale yellowish brown. Blackish pigment- ed spot on each side of basal part of eighth sternum. Subgenital plate evenly rounded at apex, with numerous long hairs on margin (fig. 157). Arms of geni- tal plate (figs. 158, 159) short, slightly longer than one-sixth of total length; ba- sal plate large; apodemes long, projecting cephalad about two-fifths of total length of plate, divergent from anterior edge of basal plate but connected by pale, sclero- tized lamella. Mate: Unknown. LenctH oF Fore Wino: Female holotype, 12 mm.; hind wing 11 mm. Hototyre: Female, Mou-Pin, Thibet, 1870, A. David; in the Museum National d’Histoire Naturelle, Paris. Nine species of Chinese Panorpa are known only from male holotypes. They differ from gwttata in having darkly col- ored thoracic and abdominal sclerites and more extensive markings on the wings, except for P. obtusa Cheng. The color pattern of the wings of obtusa resembles that of gzttata, but the stigma is grayish brown, not with a yellowish tinge as in guttata. Cheng (1957) was unable to assign guttata to any one of his groups of Chi- nese Panorpa. He noted similarities in the pterostigmal band and differences in the apical band between guttata and P. tyederi Carpenter from Yunnan. Comparison of the subgenital and genital plates of these two species indicates no particular rela- tionship of guttata with the diceras group to which t7ederi belongs. The genitalia of guttata resemble those of the centralis group, especially P. Jeez Cheng and P. emarginata Cheng. But the genital plate of guttata also resembles that of some east- ern North American species in having a sclerotized lamella between the apodemes and the basal and distal plates broad. Panorpa stigmalis NavAs Panorpa stigmalis Navas, 1908: 416, fig. 20. When Byers examined the probable male holotype in 1964, it did not have a type label. The labels read “Museum Paris, Mou-Pin, A. David 1870,” “Panorpa stigmalis Navas,” and “Panorpa stigmalts Navas, Longin Navas det. 1907.” Since there was no mention of any other speci- men in the original description, this male may be regarded as the holotype, and a label reading “probably holotype, G. W. Byers °64” was added. Cheng (1957) stated the type locality of stzgmalis to be Mou-Pin, Sikang. The collecting labels are the same as those of P. davidi and P. guttata, except in the latter two species the type locality is Mou-Pin, Tibet (see P. davidi for complete discussion). Esben-Petersen’s redescription (1921) is adequate; however, he did not present any additional information concerning the genitalial characters. The following re- description is based in part on notes and a drawing of the holotype: Heap: Dorsum and rostrum brown, ocelli enclosed by blackish spot. Antennal scape and pedicel brown; flagellum brown grad- 88 Tue UNIVERSITY OF KANSAS SCIENCE BULLETIN ing to dark brown at apex, with 42 fla- gellomeres. Tuorax: Dorsum, pleural areas, and coxae black. Femora, tibiae, and _ tarsi brown; apices of tarsi blackish. Wings long and narrow, membrane with sordid yellowish tinge; bands and spot smoky brown, stigma red. Longitudinal veins yellowish brown. Apical band complete but ending on outer margin at M1. Ptero- stigmal band narrowly connected to apical band along costal margin. Marginal spot absent. Faint spot extending from Ms to hind margin (fore wings only). A narrow band extending from Cu: to hind margin (fore wings only). AsBpoMEN oF Mate: Terga 23 black, 4-9 reddish brown. Sternum 2 black, sterna 3-9 reddish brown. Posterior process of third tergum two-thirds length of fourth tergum. Hairs on dorsal and_ ventral! posterior margins of sixth segment twice as long as other pilosity. Apex of ter- gum 9 bearing a pair of lobes. Basi- styles (fig. 160) separated only to about mid-length, with sharp ridge along ven- tral mesal margins of each. Hypovalves short, narrow, strap-like, borne on slender pedicel. Basal lobes of dististyles (fig. 160) rounded, with thickened edge, not noticeably cupped on inner surfaces. Ven- tral parameres slender, slightly curved mesad at apices. Dorsal parameres elon- gate, extending between dististyles; lateral processes short, fully exposed in ventral aspect. Femare: Unknown. Lenctn oF Fore Wine: Male holotype, 17.4 mm.; hind wing 16.0 mm. Hovotyre: Male, Mou-Pin, 1870, A. David; in the Museum National d’His- toire Naturelle, Paris. The sordid yellowish wing membrane and distinctly red stigmal spot should al- low recognition of the female when dis- covered. In addition, such body coloration as the solidly black thorax and the black and reddish brown abdomen may prove helpful. The absence of the anal horn from the sixth abdominal segment of the male places stigmalis in the davidi group as defined by Cheng (1957). The narrow hypovalves, simple aedeagus, and the hairs on the sixth segment separate stigmalis from the other members of this group. Cheng (1957) noted the similarity of the genital bulb (hypovalves) of stzgmalis to those of the kongosana group of Korea but also the differences in the wing mark- ings. The illustrations of P. kongosana by Okamoto (1925) are not usefully detailed; the hypovalves of kongosana are similar to those of stigmalis but somewhat longer. LocaLiries AND HasiTats OF MECOPTERA IN THE INDIAN SUBCONTINENT Previously published distributional in- formation has usually consisted of locality name, date and occasionally elevation. To aid in the location of collection localities, we have included the name of the state, district, and geographic coordinates. Since we found no detailed gazetteers of India, it was necessary to estimate the coordinates of several localities. Names of localities are in alphabetical order, and other data appear in the following sequence: 1. Name of locality from pin label. 2. Synonyms or variant spellings of name, in parentheses. State, district, and country if all pertain to locality. 4. Geographic coordinates of locality. There may be a slight difference between the site used for identifi- cation and actual collecting site. 5. Elevation, if known. Elevation de- rived from maps are stated as ranges. 6. Reference for locality data, if previ- ously published. Anaimalai Hills (Anamalai_ Hills), Kerala and Madras, India; about 10°15’ N, 77°00’ E; 1950-5750 feet (600-1750 m.). io) Tue MecopTera oF INDIA AND ADJACENT REGIONS 89 Bombay, Maharashtra, India; about 18°56’ N, 72°51’ E; 0-330 feet (0-100 m.). Changra (village 18 km. S of Tongsa), Bhutan; 1900 m.; (Byers, 1975). Chembra Peak, Kerala State, India; summit 3500 feet (1070 m.). Chinchona, Kerala, Anaimalai Hills, India; 10°22’ N, 77°55’ E; 3500 feet (1070 m.). Coimbatore, Madras, India; 11°00’ N, 76°58’ E; 1400 feet (430 m.). Darjeeling, West Bengal, India; 27°05’ N, 88°16’ E; 6500 feet (1980 m.); (Cheng, 1953). Deesa, Gujarat, India; 24°14’ N, 72°13’ E; 330-660 feet (100-200 m.). Devala, Madras, India; 11°29’ N, 76° 22 B 3500 feet (1070 m.). Godavari (a village 10 mi. SE of Kat- mandu), Nepal; about 27°45’ N, 85°19’ E; 5000 feet (1530 m.); (Byers, 1971). Gudalur (in Nilgiri Hills), Madras, India; 11°30’ N, 76°30’ E; 3200 feet (975 ils) Hambantota, Hambantota Dist., Cey- lon: 6°15’ N, 81°16’ E; 100-150 feet (30- 45 m.), numerous forests in surrounding area. Igatpuri, Maharashtra, India; 19°41’ N, 73°38’ E; about 1640-3280 feet (500-1000 m.). Karachi, Pakistan; 24°52’ N, 67°03’ E; 0-330 feet (0-100 m.). Katmandu, Nepal; 27°45’ N, 85°19’ E; 5000-6000 feet (1520-1830 m.). Khandala, Maharashtra, India; 18°45’ N, 73°25’ E; 660-3280 feet (200-1000 m.); (Navas, 1928). Khasi Hills (N. Khasia), Assam, In- dia; about 25°30’ N, 93°00’ E; 5000 feet (1520 m.); (Needham, 1909). Kurseong, Bihar, India; 26°52’ N, 88° 17’ E; 4860 feet (1480 m.). Ma-pien (Mou-Pin), Szechwan, China; 28°48’ N, 103°39’ E; 3280-4920 feet (1000- 1500 m.); there are a river and a town by the same name; (Navas, 1908). Muthikolam, Coimbatore Dist., India; no additional data. Naduvatam (in Nilgiri Hills), Ma- dras, India; 11°30’ N, 76°34’ E; 6000 feet (1830 m). Naga Hills, Assam, India; about 26° 00’ N, 94°30’ E; 3000-12000 feet (910- 3660 m.). Nagarkot, Nepal; 27°42’ N, 85°31’ E; 7000 feet (2130 m.). Nainital, Uttar Pradesh, India; 29°23’ N. 79°30’ E; 4920-6560 feet (2000-3000 m.); (Penny, 1969), Naraikadu, Tinnevelly Dist. India; 2500-3000 feet (760-910 m.). Nilgiri Hills, Madras, India; 11°30’ N, 76°30’ E; summits up to 8650 feet (2640 m.). Peermade (Pirmed), Kerala, India; 9°31’ N, 77°02’ E; 3400 feet (1040 m.). Periyar Dam, Kerala india 9°30 © Ne 72°20’ E; approximately 1075 m. Pondicherry, Madras, India; 11°59’ N, 79°50’ E; 0-330 feet (0-100 m.). Shillong, Assam, India; 25°34’ N, 91° 53’ E; 5000 feet (1520 m.). Simla, Himachal Pradesh, India; 31° 07’ N, 77°09 E; 6560-9840 feet (2000-3000 m.). Simra, Nepal; 27°37’ N, 84°16’ E; 330- 660 feet (100-200 m.). Thanikudi, Kerala, India; 9°30’ N, 77°16’ E; town in valley, surrounding hills 4000 feet. Tongsa (Tongsa Dzong), Bhutan; 27°33’ N, 90°30’ E; 2150 m.; (Byers, 1975). Vavuniya, Vavuniya Dist., Ceylon; 8° 45’ N, 80°30’ E; 200-600 feet (60-180 m.). Walayar Forest, Kerala, India; about 10°48’ N, 76°48’ E; about 700 feet (215 mi): Wattegama, Monaragala Dist., Ceylon; 6°48’ N, 81°30’ E; 0-330 feet (0-100 m.). Yala; Yala Wists) ‘Ceylon2 46-22 INE 81°31’ E; town is on coast, many small forest reserves nearby inland. 90 Tue University oF KANsAs ScrENCE BULLETIN LIFERATURE(GI PED Byers, G. W. 1954. Notes on North American Mecoptera. Ann. Entomol. Soc. Amer., 47: 484-510. 1965. The Mecoptera of Indo-China. Pa- cific Insects, 7: 705-748. 1970. New and little known Chinese Mecoptera. J. Kansas Entomol. Soc., 43: 383- 394. 1971. A new Neopanorpa from Nepal. J. Kansas Entomol. Soc., 44: 534-539. . 1971. An illustrated, annotated catalogue of African Mecoptera. Univ. Kansas Sci. Bull., 49: 389-436. ——. 1975. Ergebnisse der Bhutan Expedition 1973 des Naturhistorischen Museums in Basel. Mecoptera: Fam. Panorpidae. Entomologica Basiliensia 1: 87-91. CarPeNnrer, F. M. 1931. Neopanorpa hirsuta Crampton. Psyche, 28: 184-185, fig. 1. 1938. Mecoptera from China, with descrip- tions of new species. Proc. Entomol. Soc. Washington 40: 267-281. CHATTERJEE, S. B. 1954. Indian Climatology. Cal- cutta-6. CuHenc, F. Y. 1953. Three new species of Pa- norpidae (Mecoptera). Psyche, 60: 119-122. ———. 1957. Revision of the Chinese Mecoptera. Bull. Mus. Compar. Zool., Harvard Coll., 116: 1-118, 23 plates. Crampton, G. C. 1931. The genitalia and terminal structures of the male of the archaic Me- copteran Notiothauma reedi, compared with related Holometabola from the standpoint of phylogeny. Psyche, 38: 1-21. EspeN-PETERSEN, P. 1915. A synonymic list of the Order Mecoptera. Entomolgiske Meddelelser, 10: 216-242. ——. 1921. Mecoptera. Collections Zoologiques du Baron Edm. de Selys Longchamps. Fasc. 5. 172 p. Bruxelles. GeERSTAECKER, A. 1885. Ueber einige Arten der Gattung Bittacus Latr. Mitt. Nat. Verein von Neu-Vorpommern und Rigen, p. 117-121. Harpwickr, T. 1825. Description of Cermatia longicornis and of three new insects from Nepaul. Trans. Linnean Soc. London, 14: 131-136. Hergurn, H. R. 1969. The skeleto-muscular system of Mecoptera: the head. Univ. Kansas Sci. Bull., 48:721-765. —. 1970. The skeleto-muscular system of Me- coptera: the thorax. Univ. Kansas Sci. Bull., 48: 801-844. Isstk1, S. 1933. Morphological studies on the Pa- norpidae of Japan and adjoining countries and comparisons with American and Euro- pean forms. Japanese Journal of Zoology, 4: 316-416. Kimmins, D. E. 1928. Two new Bittacidae. Ann. Mag. Nat. Hist. (Ser. 10), 1: 395-396. Martsupa, R. 1965. Morphology and evolution of the insect head. Mem. American Entomol. Inst, 4; 16334: MicxoteiT, G. 1971. Zur phylogenetischen und funktionellen Bedeutung der sogenannten Notalorgane der Mecoptera (Insecta, Me- coptera). Zeitschr. Morphol. Tiere, 69: 1-8. Navas, L. 1908. Neurdpteros nuevos. Mem. Real. Acad. Cienc. Art. Barcelona, 6: 406-417. 1910. Description d’une nouvelle espéce de Panorpidae. Deutsche Ent. Zeitschr., 1: 288. 1914 (1913). Neuroptera asiatica. Revue Russe d’Ent., 13: 271-284, 424-430. —. 1929. Communicaciones entomologicas, in- sectos de la India. Rev. Acad. Ciencias Zara- goza, 12: 177-197, 8 figs. 1935. lecadas de insectos nuevos. Decada 27. Brotéria: Série trimestral, Ciencias natu- rais, 4: 97-107 (vol. 31 of whole series). 1936. Neévroptéres et insectes voisins— Chine et pays environnants. Notes d’Ent. Chinoise, Mus. Heude 3: 37-62, 117-132. NEEDHAM, J. G. 1909. Notes on the Neuroptera in the collection of the Indian Museum. Records of the Indian Museum, 3(3): 185-210, plates 19-21. Oxamorto, H. 1925. The Mecoptera of Korea. Bull. Agric. Expt. Sta. Gov. Chosen, Suigen, Korea. 2(1): 1-8. Oranes, F. Q. 1922. Head and mouth-parts of Mecoptera. Ann. Entomol. Soc. Amer., 25: 310-323. Penny, N. D. 1969. A new species of Bittacus (Mecoptera: Bittacidae) from India. Oriental Insects, 3: 161-164. Serty, L. R. 1940. Biology and morphology of some North American Bittacidae (Order Me- coptera). Amer. Midland Naturalist, 23(2): 257-353. Srorcu, R. H. and L. E. CHapwicx. 1968. Thoracic structure of the adult mecopteran, Bzttacus strigosus Hagen (Mecoptera: Bittacidae). J. Morph., 126: 199-210. Tyeper, B. 1956. Mecoptera. South African Ani- mal Life. Results of the Lund University Expedition in 1950-1951. Vol. 3: 344-390. ———. 1970. Mecoptera. In: Tuxen, S. L., Tax- onomust’s Glossary of Genitalia in Insects, p. 84-88. E. Munksgaard, Copenhagen. Wacker, F. 1853. List of the specimens of Neu- ropterous insects in the collection of the Brit- ish Museum. Part II. Sialidae-Nemopterides. 469 p. London. Westwoop, J. O. 1846. Monograph of the genus Panorpa with descriptions of some species be- longing to other allied genera. Trans. Ent. Soc. London, 4: 184-197. September 17, 1976 LETIN De RR ' 4 OF KANSAS pp AO : ae | a Ss a ae 2, 2. CE (RA DOES Ee. sacs ta at & [IEp. yep swva1s M9 BUNISIG (CAD) G51) uMmolq 601 O01 (1)Z8 (2) 882 [I43sou 0) 949 (TE) 61Z (81)9°08 (+7Z)S"F8 aniq dato yep snqonuayy = (c) spurq pol $ < 8°Z “ag (81)L9S aka (8) S91 (OL)Z6S (ZDE79 aniq faduvs0-ysIMmoy[o4 snayonbo (0Z) (+2) spurgq ayed LS 9 (ZL)Es (91)0Z9 aka (SL)O6L (IZ0'LZS (Sb) BIZ yarp Sper ystdand 140] 404 (Z) (Sb) aspo aed 61 8? (9)001 (TE)Z1¢ [Tasou 0) af9 (Zh) SLT (61) 8'8¢ (6F)OTS YstmoyaA “paz stapjnzoqns (1) eats qyods ystpper HC 7a (1) 0+ ey [Ensou 0} 949 (1) 8ZI (1)0';0S (1) 09+ yep [e.Quad fosurso0 pyppnouiaand (2) are aspa ayed Ee (6 Il Se, aa ypou (Z) SLI (Z)O'8b (T)O It yarp Spor ystydund snjopidajosiud spurq asol Z oz ipo (FI)EZ (9) IFT Ivo (€1)EZI (FI) Feb (L)Z br SEN ‘98uvI0 YsIppot snjopyoydnsau syeons (I COI) ystdind yep g Cy nic (s)bZI (¢)0S9 949 0} A]TBOU (OL) EZT (Z1)9°29 (01)9'0Z ep ‘yurd ystjdind uaopys Gb) a Ce — (1) 0€¢ 249 (1) 261 (QL bb (g) FES S[TEp pes QIN (Z) (TD) aspa vo Ob (ID)IS (7%z)909 989 (FI)9ZT (cI)o9r (@zEE9 yiep uv} {9801 snutdno Gye (G4) : aspa Ghemme cre re = Puosaq IO Iva (Z) 981 (109s (zZ)o'Es yrep MOTJ94 {pat stdajosiup 6 ) 6 Ne quay pury A-S O41 SYJ8UI] 6 P SU dpjmap SHIOAdS (suuvis ) (wu) papuayxa papnsof SOU Ih (wu) ¥73ua] {o 10]0) fo 10]0) 14319 va.w dpjnaq Ag payrvas 1U10g quan-jnous YNpy ‘dds S17J0Uf’ AO SOLLSIUALOVUVAC) TVOISOTOHAOJY T AtdVviL 10.5 133(7) 797(3) nostril 190(13) 258(3) 43.7(23) 80.0(3) 43.3(15) 82.2(7) dark blue orange yellow i ~ a rodrigue beyond snout checkered light aequatorialis (3) and dark brown 72 (3) ear to eye 402(2) 188(3) 58.0(1) 3-513) blue chloris dull yellowish white; pale blue anteriorly; dark scales Soff (27) (31) 1.48 (1) eye to nostril 433 (27) 4.1 268 58.5(28) 62.5(38) blue yellow; streaked w/blue, white, green gemmosus 187(1) ear to eye 194(1) 46.0(1) pale golden rusty red maculiventris yellow 1.96 1.97 127(14) neck 201 (16) 48.0(15) 50.9(14) yellow dull white w/black lines nigrolineatus Frectp OBSERVATIONS ON MAINLAND ANOLES 97 (15) (15) 1.48 (7) 1.78 (15) 229 ear to eye 182(16) 221(12) 46.3(9) 49.9(21) dark; blue dull white; gray streaks peraccae inner rim 1485 (2) beyond snout 211(2) 117.0(2) golden white anteriorly; greenish posteriorly princeps (2) 6.2 brown 9.0 CEG) eye to snout 400(1) 214(2) 73.0(1) 70.0(1) dark pale red; sp. (1) (1) black spot (megapholidotus). The area of the dew- lap was determined by tracing the outline on paper while the organ was held fully extended and then measuring it with a planimeter. Table 2 summarizes information about body temperature (first three columns) and reproduction (last two columns). It shows mean body temperature, range of body temperature, and adjacent air tem- perature for several species. The fourth column indicates observed occurrence, by month, of individuals indicating reproduc- tive activity: Females with enlarged folli- ces or oviducal eggs; and juveniles, or large young. Although most populations were not sampled throughout the year, breeding seasons are inferred (Table 2) from the observed occurrence of eggs and young, from the observed incubation peri- ods and growth-rates of several species (Fitch 1973a), and from observations that, in general, dry weather inhibits reproduc- tion of anoles and wet weather induces it. Table 3 shows preferred resting places for each of the 25 species, whether at ground level (leaf litter, rock surface) or above ground (height, and kind of perch such as tree trunk, rocks, foliage or stem). Perch-diameter is shown for those that used stems. Table 4 is also concerned with perches, but limited to the nine species for which most data were available, with in- traspecific comparisons of adult males, females and immatures. Table 5 summarizes interspecific co- occurrences. The alphabetical series at the top of the page includes 16 of this study and eight other commoner species that co-occur and interact with them. Al- though interactions are still poorly known in most instances, relative size, abundance, and extent of distribution are indicated, because these are factors that might influ- ence the outcomes of interactions. Ordi- narily, larger species would tend to domi- nate or displace smaller ones. Interactions between a common and/or widespread 98 Tue UNIVERSITY OF KANSAS SCIENCE BULLETIN species and a rare and/or geographically limited species would be relatively impor- tant and perhaps critical for the latter, but relatively inconsequential for the former. The displays were filmed with a Super-8 motion picture camera (Bolex 155 Macro- zoom) at a speed of 18 frames per second. A Bell and Howell Super-8 projector adapted for single-frame advancement was used in the analysis of the displays. Some of the displays filmed were induced in un- confined anoles by presenting them with a mirror or a tethered rival, others were filmed in Plexiglas enclosures when sev- eral individuals of the same species or different species were confined together. The aggressive displays of iguanid lizards are complex, stereotyped, and species-spe- cific (Carpenter, 196la, 1961b). These dis- plays may reflect both genetic relationships and environmental factors (Echelle e¢ al. 1971a). The displays have been filmed and analyzed in few kinds of mainland anoles, notably in Anolis nebulosus (Jenssen, 1971). It has been demonstrated by Jens- sen that interpopulational differences in displays may exist, and T. A. Jenssen (pers. comm.) has indicated that certain species show a variety of displays that differ in basic form. However, in such in- stances, one display type is much more frequent than the others and may be con- sidered the display. Various display modi- fiers may cause differences in detail, re- flecting the motivational state of the lizard. TABLE 2 Dara oN Bopy TEMPERATURE AND REPRODUCTION OF Anolis spp. Observed occurrences of ovarian follicles (OF) Air oviducal eggs (OE) Inferred Body temperatures temperature juveniles (J) and breeding SPECIES when active N_ whenactive _ large young (LY) season QUSOLE DiS ape eee tees wee OE: Feb. July-Sept. CUDTINUS BA sean ee eee” EN J. 30.7(34.3-26.2) 24 30.5 J: Sept.-Dec. July-Sept. QUIT ee ee eo Be etn OE: June-July July-Sept.? UG OULIN Sees, ee Pa ne oun. E 30.5 1 30.9 OF, OE: July July-Sept. J: Feb. MERADRONAOLUS __..---cneceocenenncneeene-no-- 17.5 OF, OE: June-July June-Sept. TIGLOLE PIA OLU SE 21-17 OF, OE: July July-Sept. BGLUICITCH GL ape ee eee OF, OE: Feb. year round? SCLOOCULGEE gaan ee 31.4(34.6-28.0) 13 30.0 OF, OE: July June-Oct. LOSI OTES yes het ee Pal, 23 29.1(32.8-26.2) 28 27.9 OF, OE: July July-Oct. Ye: Febss.Mar- CEULGLICT, Sige RO Oe cede OF, OE: Mar. year round? Qitenuaius: 21.5 18.0 OF: Mar., July, Aug., year round? J: Sept., Oct. BiScUlLC Chee eee 31.0(32.5-28.5) 36 32.6-27.3 OE: Dec., Jan., Feb., year round? Mar., May, Aug. GUEVETEGA) ee OE 27.3 1 ZO Ne Nove year round? LY: Feb. GG fUstanys a ee ee ee 28.4(31.9-26.0) 31 27.6 OF, OE: Feb. ? J, 1aY-" Reb: INSONISE eae ee eee ee OE. 200 Jin Mare ? GECUGIOTIGIIS ee 22.3(24.5-20.1) 2 2220) je Reb: ? DEIDTIOS USM ee ee 21.0(29.3-18.2) 34 20.4 year round ICT OMUNEGLUS I en ee eee 29.9(30.5-29.4) 5 30.5 OF, OE: Mar. Feb.-July? ER UCCUC ee Ee RP es! OF, OE, J, LY: year round? Feb., Mar. low Oct.-Dec. DYINCE PS des ees. Fin Sys ps Tas 27.4 1 Die. ? Ge ee eee en ne ae 26.0(26.4-25.6) 2 26.0 OF, OE: Feb. year round? J: Feb. 99 Frecp OBSERVATIONS ON MAINLAND ANOLES Pitee cpataala als 7 an 10°2 001 ois ERs di Rp ida oie si lle nies ae 0S0" cl eae aor Jee ite 00 ea en Sa Comte 5 ico as ie eae sdaoutd (vuvurq) +++ Bcckcessss aeeeSeeeg NN ceocaseeseee 06'I (enti, eee eae GE QUT Othe anise es ceca avIIDAIG (ore ) Ge neers 96 OO ee Saas cc ne 1 Cle er ae ac aces SNIDIUNOAT IU cooeerdS = mepereareni ae 071 O0T Sponsor ant [0 seeguoninonus oe ere aE bL'T Qe © RS EQ a are aa re as ee aca as SnSOULMIS Receeecee) | RROEREPAS er ee O0T SpRCESSE ORS arn SoS Ea ed aga aaa aS T1070 Z10" Se wesetesesces | ceeteeeceene CZ] Qi | SSSR eects 9 ecace i aaa sypiiojnbav Tgp eS eee seateals £8" [EA aD cca eI Le gg 5 egraptg anal 1zanB1ipos oScerseccec 4 0S°Z OOT eer ra | ac ati ata L080) soeeeecscs | Sescenccges me O0T sessseeeeees sesces Meee em TS. hy cimoxccca |= Etnaeetcee OZ b9/ sncenecose=s 9'§Z 6 SieeannaEnES 20005721770 poe Me Aer ee +++ 06 Ee O tae Wiarinps €9¢ gears ss ane hotuadin) corteeeeeees seen ee en eee LOT 9°99 Se eerie aa gee 0) a (ke, ea cece AI BUNISIG SOoereereee ERE eee eere SES 00°Z 676 SRFRERESERO Lal Co en enn 1272 120727, sorepecnossy Sicageety | Teeteeieey ge aes Um eerie in cle Noelle seosce er eg a STF) oe es eee OT [en G@eee ee 6°S he eae ee s4opav} dgi® eae Eee pee ate bI'I ORG = - Sees 9 CO sore: pass smreneae nase sLippnzoqgns SpecoeseesG |) eeeecoeoa | ree) |) eras as OOr I iamimamenmmmen 72) 2001970400, See 0 See ee ee ae ae O0l Qe ae, eShop Io oNoIUL Speereene aes Gz" CC seeseeeeee Gb [TD cm sngopyoydpsau fief Bpcreocceccs 9) Gees Peres rower ete ee eS ia eee pe oe oe rea fu Nonton Pe acne oeeteet ates voeeee ee ae Ce EE sa U ND, Wal See 4 eee LOT (HO rn er es SLT ORs aa ee Saree snutdns Stccawteg | Peto Aeeare Gb 0Z seteeeeeeeee 08 Co dn eae ICO/OSIuy Yoi9g unig aselloy oy 991] 2DIR] (uw) aoeyins BES y N Sa1OFdS (uw) Jaap jo yuniy, ISI fs yoy yeoT fs punois sv0qy [PA2]| punolry ‘dds S7]0uUpP AO SAHOUA ONILSAY © Gla Val 100 Tue UNIVERSITY OF KANSAS SCIENCE BULLETIN Z|RSRSSAsSaH ACCOUNTS OF SPECIES me Tue MExIcaNn SPECIES Be 2 |SQenKNgQrum Anolis anisolepis gee |cass aH tz This species was described by Smith, elas gee Burley and Fritts (1968) from 23 speci- Ze |SeSxssaga mens collected at San Cristébal de las = oD oot fa fe . B|o Casas and several nearby localities in west- ° . . . . lt cee ern Chiapas. In describing it the authors v e . . . 2 3 Rees oonSn stated that its specific distinctness from the closely related and similar Anolis haguei was demonstrated by sympatry; een 0 eS sete) pe both were found together between San e Zo 2 ee eee tn Cristébal de las Casas and Tenaja. The z ~ authors distinguished A. anisolepis from i ve ee its near relatives A. crassulus and A. “| |$8/8R ixS ig 's haguei by the following diagnostic ch ee Se haguet by the following diagnostic char- se acters. es S A. antsolepis: Dorsal crest prominent 8 B\oS ixo io iO (with scapular and sacral gaps); dorsal oD ° . scales enlarged, with largest, at mid-back, about half the size of ventrals, gradually eS aS becoming smaller laterally and grading mean height (meters) into granular scales of sides; dorsal stripe lacking; size relatively small (S-V up to TABLE 4 CoMPARISON OF REsTING PERCHES OF NINE Anolis spp. _ TZTInwowoco es \seeceesen oy £| |e § ciakn ta ||| A. crassulus: Dorsal crest lacking; dor- 3S 1 sal scales abruptly larger than small lateral ED lanme | scales, the largest about equal to ventrals, go eaRaooose ‘ith lateral dorsals | han medi gs is with lateral dorsals larger than median dorsals; dorsal stripe lacking; size rela- Z tively small. es LIN =INANNH AO - , fo |\N Ba SON RN oe A. haguet: Dorsal crest prominent Ev6 : ace) 5 (with scapular and sacral gaps); dorsal Se ian ee oh scales enlarged, with largest at mid-back Ors Oye . Jeane = = c E S/SSSSSESRSE about half the size of ventrals, gradually a) becoming smaller laterally and grading ol 5 . . = into granular scales of sides; dorsal stripe J § ; Belen e cease present; size relatively large (S-V up to 53 bo” mm). As do all other Mexican species here discussed, Anolis anisolepis and its near ‘| relatives belong to the CArysolepis species +} 5 4 ‘| series of Beta anoles. On 14 February 1971, we captured five eo A. anisolepis 45 km S of San Cristobal de “a's x = 8 os : S nalSseeassgs las Casas beside Highway 190. There had His = 2a Ss S's : : S(/SSSS 8s 58 § been frost on the preceding night (eleva- HIS SAPP SESS ; GAlP FSS SERS S tion approximately 2150 m), general ter- 101 FIELD OBsERVATIONS ON MAINLAND ANOLES ete een ee ene ee ee eee nee eee ‘ds Spas tes awages ogee aiats oe cane eae ee res sdaoutsd ee ers avoovnaad oe eee m mn e nec a nen nn enna wenn nnn e een nnneenn SnpaUYyoss it Paha ake ne eke Bale eee Nae ee SIAJUPAYNIVD UL ee errr snsouuas s1oqy? Spemmmaganey, anes -sigrcne Tene scene cena sypisojonban NVOIAWY HLOOS +44 1ZINBUpPOr NS a aac Susisut wee ee eww ewe ewe eee ete enn nett a ween enna meee ianavy Se me Pages Cetera NU 0 eco eeooeeocceperoe snuvisn{yop ee oe ea sana pear Cc 1MI{UIGADI — It A en nee ewww ween wee eee AITUNISIG Goauecorncen se apnacnnrecesenenraaasoc a= Worse SnJDNUIIID SN SS SS aaa I Gage snoyonbo NVORMANY IWYLNAD +-.— — oe ee i a ae ere 110)hD1 sb ob —— se al SLADINIOGNS eS a eee ee er a nrc a I Sees ge ee aes pyppnriiaivd Se ee a See snjopidajosiu = See eae a snjopyoy dviau ee ee ee ee ee unaopos —+ 4+ —+— —-+— rai ag teak css ah i ae og Se iar pe ee unp == snutidng sidjostup NVOIXSJ 8 = S q § 8 g x S rR Sy 8 Be 8 8 3 ~ > x. > oe XN 3 = = = 5 S: 8 3 8 S 8 8 8 S. ze 5 S = & s 8 Re =: - 3 § s 5S 8 Ss S ES 5 3S o: = Ss > bad Ss a *(—) w sdepiaao Ajaraur io (++) a8uer sassedurodua ‘(—) Juepunqe sso] Jo (+) JUepuNqe 210 ‘(—) Jo]][eUIs IO (+) Jase] ist satdads payetosse JayIYM Moys sjoquiAs vary] ay V9UaIINII0-09 YyIea 10g *(1addn) saidads payeidosse pue (aJo]) s2joue pur[UTeUT UMOUY IPI] 10 sue dUIOg "ddS S7J0Up AO SAONAAANIIO-OFD) OIAIOAASUALN] C W1dV.L Tue UNIversity oF KANsAs SCIENCE BULLETIN 102 +—+ 4++—-— 4+-— cine RR ee Pas AG. sree sted, ede c Loe (2 ial alba Se RT Oe ee ‘ds —+— ++— ——— Sale eee We) Somes eee irs (> Pires ae SS ee eee ae sda2uid ——+ ——+ ———— ee ee eR a a Sa RG a epee espe ns erect ae gee ee av220.41ad aes mek mene aes SNJDIUNOASIU é ++ + +- + + + + f+ ae We ee SUMUIAYNIDUL Gospel ee a a i ra ri era SNSOUUAIT ——-+ +—-+ —-+— Se Re i ee i er ie SIMOJYI I SL ee sypisoypnban NVOhnaNY HLA0g Soeaar cteng oo Goa ge”~ pee Chee IZINSIAPOd iat pi le Sees eR ore arene are SIUBISUL See 2 Rane eS hie a aa a 1anovy hig Sree teoeParesrer econ Ssnupisn{ op Se Sats teat fa ee oe) Wem Rt be Reececeee one cae n arma ore WI JUIGADI ae a a ee ge ae ae a AITUNISIG a ee eee Snyonuayyy Ne ea snoyonbp NVOIMAWY TWULNdad oS es ee ae 110404 pee een ee A anes ae SLAIDINIOGNS a a ee cee von Se ireane esac eee nae SnIOpId 210.19 1UL Sessa anne econ naan sinjopyoy dvs2u Sear Te Ba nSPh enone me ue apne snecn cao maopos be Rina ie She ea tuunp setitras 2 1y oc en cha snutidng il aera ee ae sid a]Ostup NVOIXa J “ds sdaoutad avoonsad SnSOUMIS S140] 49 SUOA{IULY] SMIpI MAP qUI SHIUNY tansvy snasdno SUMJUPANNIVUL sypisoppnbav sidathjod ee (panunuo?)—c ATAVL FIELD OBSERVATIONS ON MAINLAND ANOLES 103 BISCUTIGER Se Ee oh eee RODRIGUE ZI Inte ieee Fic. 3. Diagrams of Display-movements of Anoles. Explanation as in Figure 1. Upper: Anolis biscutiger, San Isidro del General, San José Province, Costa Rica. Lower: Anolis rodriguezi, Puerto Barrios, Izabal Province, Guatemala (typical display, and on right atypical short display). 104 Tue UNIVERSITY OF KANSAS SCIENCE BULLETIN rain was flat, open and xeric, but there was a roadside spring flowing into a marsh with vegetation including Scirpus, Typha and Equisetum. There was a dense thicket of Crataegus, with Cornus, Sambucus, Smilax, Ribes, Salix and other woody plants bordering the marsh along the road bank. The anoles were found, all within a 30 m radius, in this brush patch. One was on a horizontal stick .5 m over the spring, two were in dense grass at the edge of the marsh and two were in leaf litter beneath the bushes. They were relatively slow and clumsy and were easily caught. The three adults each weighed between 3 and 4 grams and their lengths (S-V) were: 53 (4), 53 (¢), and 56 (2). All thus equalled or exceeded the maximum size indicated for Anolis anisolepis by Smith, Burley and Fritts (1968). The only adult female among the five lacked oviducal eggs, but had somewhat enlarged ovarian follicles (diameter 5.6 left, 2.6 right). Absence of fully devel- oped eggs and of young less than half- grown suggested a cessation of breeding for several months and from analogy with other species it may be inferred that repro- duction of A. anisolepis is limited to the wettest part of the year (July-September). The typical full display-activity pattern began with 1 (20.99%) to 4 (125%) pre- liminary sets of bobs—most commonly with 2 (41.6%) or 3 (25%) performed with the dewlap retracted. Duration of bobs of these preliminary sets was as fol- lows: soet I 92 ==. 07> sec (ns = 9) Set 26) == 003 seen(i—nG) 0c LM 5/0) == 042 sec (n = 5). A bobbing series fol- lowed (mean duration, 153 = .121 sec, n = 23) during which the dewlap was pulsed in and out one to five times (usu- ally three or four times). Duration of the entire display was 5.2 = 39 sec (n = 16). Six of the 22 filmed sequences included no preliminary series of bobs and the figures for total time do not include these. The number of bobs per unit were noted for eight additional filmed sequences, al- though these were not analyzed frame by frame. In Figure 2 a display of A. ant- solepis is compared with those of A. cras- sulus and A. haguet. Anolis cuprinus The original description (Smith, 1964) was based on a single adult male from near Zanatepec, Oaxaca, stated to be a member of the “cupreus group” closely related to Anolis cupreus of the west coast of Central America, but larger (53 mm S-V), with a distinct frontal concavity, having interparietal smaller than ear open- ing, having a “bright, blood-red dewlap,” not bicolored as in A. cupreus and differ- ing in a few details of lepidosis. Lynch and Smith (1966) reported a series of 26 from the same area as the type, and agree- ing with it in most characters. Henderson and Fitch (1975) compared the structural niche of Anolis cuprinus with that of the smaller, syntopic A. sericeus. Collections were made on 10 and 22 February 1972, at Rancho Las Vigas, 8 km E Zanatepec, Oaxaca, Mexico. This locality is a relatively humid wooded can- yon between 270 and 340 m elevation. Al- though the type locality was stated to be Zanatepec in the original description, in- tensive search by us near the town failed to reveal Anolis cuprinus there and the locality is in open, arid terrain of the coastal plain at an elevation of 240 m. In the original description Smith stated that Fic. 4. Diagrams of Display-movements of Anoles. Explanation as in Figure 1. The diagrams for Anolis taylori and A. dunni shown here indicate relative amplitudes of dewlap and head movements (unlike other diagrams in Figs. 1-3 and A. gadovii in Fig. 4). Upper: Anolis gadovii, Tierra Colorada, Guerrero Province, Mexico. Middle: Anolis taylori, Puerto Marqués, Guerrero Province, Mexico. Lower: Anolis dunni, Palo Blanco, Guerrero Province, Mexico. FIELD OBSERVATIONS ON MAINLAND ANOLES GADOVII Stes an Gee all TAYLOR! weer (ol. DUNNI 105 106 Tue UNIVERSITY OF KANSAS SCIENCE BULLETIN the holotype was from 5,000 feet (about 1575 m) elevation, which suggests that the locality must have been several kilom- eters removed from Zanatepec, in the Sierra Madre to the north. Our specimens agree well with the description of the holo- type in most respects, but in Smith’s pho- tograph the holotype has much more prominent transverse dark bands on the body and more dark pigment on the ven- tral surface. The population found by us was lim- ited to a few hectares at the head of a small canyon. Doubtless there are other localized populations in the same general region, but adjacent canyons were drier, with much evidence of fire, and none of the lizards could be found in them. Many A. cuprinus were found on the ground in leaf litter. Others were found on vertical tree trunks and stems, usually suspended head downward within 2 m of the ground. A few were on the vertical rock-faces of cliffs or outcrops. All A. cuprinus were found within a few meters of the ravine bottom where there were large boulders and a trickle of water even at the height of the dry season. Even here there were signs of fire; trees, including large ones, had been partly burned. Some of the anoles found were on charred wood. Burning evidently had oc- curred within the previous year. Perhaps the lizards were able to escape incinera- tion only in unusually protected situations such as were provided by the ravine bot- tom, but anoles that wandered onto the adajacent exposed slopes probably would have been killed off whenever burning occurred. Most of the 37 A. cuprinus collected in February 1972, were adult, as were others seen that escaped, but a few were sub- adults. Evidently there had been no repro- duction for many months—since the rainy season. The females had neither enlarged follicles nor oviducal eggs at this time, at the height of the dry season. Lynch and Smith (1966) noted that their series of 26 collected between 6 September and 10 December consisted of adults and juve- niles. Probably egg-laying begins soon after the onset of the rainy season (July?) and continues into September or even Oc- tober, but ends in time for all but the latest eggs to develop into adults by Feb- ruary. If so, the annual cycle is like that of the more southerly Anolis cupreus, but with the breeding season even more con- centrated by the constraints of the long and severe dry season. Several males that were confined to- gether displayed, and three sequences of different individuals were analyzed (Fig. 1). The displaying lizard tilted its head far back to allow ample space for exten- sion of the relatively large dewlap, which was sometimes tipped or wobbled to make it more conspicuous to a nearby adversary. There was a series of 7-10 bobs (3.9-7.6 sec), a pause (.7-1.2 sec) and then three head drops with intervening pauses. Ex- tension of the dewlap began with the first bob and lasted 5.3-9.6 sec, with retraction in the pause before the first head drop or immediately after it, lasting 2.7-3.2 sec. The dewlap was spread again between or after the head drops, for 1.0-1.9 sec, but not maximally. Following the second re- traction by 2.1-3.4 sec was a third exten- sion, .8-1.4 sec, after the last head drop. The bobbing sequence and head drops lasted 10.4-13.8 sec. Anolis dunni The original specimen, from Agua de Obispo “between Rincén and Cajones,” Guerrero, was first reported by Smith (1933) as Anolis gadovit. Later the same author (Smith 1936) after having exam- ined gadovii material, named the present species and distinguished it from A. ga- dovit by its widely separated, narrow and ill defined frontal ridges, vertically com- pressed nares, supraorbital semicircles broadly in contact, separated from inter- FIELp OBSERVATIONS ON MAINLAND ANOLES 107 MACULIVENTRIS ATTENUATUS maaan loa to ast pis th aa i a ah ety GT el a ae ta ee Fic. 5. Diagrams of Display-movements of Anoles. Explanation as in Figure 1. Upper: Anolis macu- liventris, Tinalandia, Imbabura Province, Ecuador. Lower: Anolis attenuatus, Monteverde, Puntarenas Province, Costa Rica. 108 Tue UNIVERSITY OF KANsAs SCIENCE BULLETIN parietal by a single series of scales, and reddish dewlap. Davis (1954) mentioned additional characters. Davis and Dixon (1961) reported A. dunni from the same part of southern Guerrero, at elevations between 2,800 and 3,300 feet, in tropical deciduous forest and pine-oak forest ad- jacent to it at higher elevations. Duell- man (1961) reported it 300 km farther west in Michoacan. On 24 July 1971 an adult male was found in the lichen-covered crotch of a tree 2.13 m above ground in a humid can- yon near a stream 7.55 km S Palo Blanco, Guerrero. A gravid female was found on the same day about 5 km down the same ravine on the ground in herbaceous vege- tation at the edge of the water. No others could be found in a week of field work in the general area. Davis and Dixon (1961) found fully developed eggs in sev- eral females taken in June and July. Evi- dently the breeding season begins after the onset of summer rains. In life the male was olive brown with a lateral white stripe on each side begin- ning behind and below the eye on the supralabials and ending above the thigh insertion. There were four pairs of lateral white spots on the body, each about the size of the ear opening, the first, on the shoulder, twice as high as long, and the others successively more rounded. The ventral surface was dusky, with small and obscure white dots. Dewlap coloration consisted of four or five dark red bands on a lighter red background; the bands ran parallel with the free edge of the dewlap. The single male that we collected in July 1971 was kept alive for several months, and he displayed vigorously when- ever other anoles were placed with him. The display (Fig. 4) was almost identical to that of Anolis taylori, having the same three dewlap movements, preceded by sev- eral shallow head jerks. In three of the four displays analyzed there was a_ pre- liminary extension of the dewlap preced- ing the final three dewlap movements by 3.9 to 4.6 secs. Anolis gadovu This species was named by Boulenger (1905) from a specimen collected by H. Gadow at Tierra Colorada, 300 m in the mountains of southern Guerrero. It was not mentioned again in the literature until Mosauer (1936) rediscovered it and pub- lished a brief account of the habitat and saxicolous habits. Davis (1954) in describ- ing Anolis omiltemanus, A. subocularis and A. microlepidotus from Guerrero, compared each with A. gadovit and in- cluded the latter in his key to Guerreran anoles. He characterized A. gadovi as large (up to 80 mm S-V), with long hind limb (reaching nearly to eye), having upper parts gray with dark markings, dewlap reddish, ventrals smooth, and with four gulars between the anterior chin shields. Our observations on A. gadovit were made at Tierra Colorada, especially on a rocky hillside 1 km N Palo Gordo (an outlying village southeast of Tierra Colo- rada) in July 1971. The lizards were on steep rocky hillsides where there were loose boulders 1.5 to 4 m in diameter, sometimes in several layers with luxurious vegetation (including gnarled trees) grow- ing among the rocks and vines screening some of the rock-faces. The anoles were extremely abundant. All were adults, usu- ally associated in pairs, with intervals of usually 5 to 8 m between pairs. Most often the anoles were on vertical rock surfaces. They were shy and elusive, hiding be- neath concealing vines and relying on their cryptic pattern, or moving downward out of sight and out of reach beneath the rocks. “An estimated 25 per cent of those seen used tree trunks or stems ranging from 2 to 50 cm in diameter” (Fitch and Henderson, 1976). All the females seen appeared to be gravid. Ten that were dissected had FIELD OBSERVATIONS ON MAINLAND ANOLES 109 both oviducal eggs and enlarged follicles. Seven had an oviducal egg (10.5 to 14.6 mm) in each oviduct, and one enlarged follicle. The remaining three each had an oviducal egg on one side only and had a single follicle on the side opposite to the egg. We revisited Tierra Colorada and Palo Gordo in the dry season, in February 1972. Trees were bare and leafless then and herbaceous vegetation was withered and dry. In prolonged search only three ju- veniles were seen. Evidently activity is much reduced, the anoles tending to stay in relatively cool and humid situations deep beneath the rocks. The breeding sea- son of A. gadovii is believed to resemble that of A. taylori (July through October? ) but to be even more restricted by the more xeric conditions where it occurs. Five displays of two individuals were analyzed. The displays were characterized by one or two dewlap extensions each as- sociated with a short series of ascending head bobs (Fig. 4). Anolis megapholidotus This species was described by Smith (1933) from four males and two females, all from Agua de Obispo about 45 km S Chilpancingo, Guerrero, Mexico. The species was diagnosed as having ventral and dorsal scales strongly keeled, scales of supraorbital semicircles enlarged, strongly keeled and in contact or separated by a single row of scales, occipital about equal in size to ear opening, dorsal scales larger than ventrals; dewlap red, extending to mid-belly in male, small in female. Davis (1954) included Anolis mega- pholidotus in his key to Guerreran anoles and compared it with other species. Davis and Dixon (1961) found A. megapholido- tus at additional localities near Agua de Obispo, in tropical deciduous forest and pine-oak forest, from 850 to 1350 m. Females they collected in June were all gravid. Eight adult females we collected at Agua de Obispo 10-12 July, 1971, also were all gravid. Four had one oviducal egg, each in the left and right oviducts and four others each had an egg in only one oviduct. Also, each female had either one (in 3) or two (in 5) enlarged ovarian follicles. Only adults were seen, suggest- ing that there had been no reproduction for several months during the dry season. Meanwhile, young of the previous year had grown to adult size. Displays were easily elicited when males were confined together, but also were readily interrupted by movements of other lizards, and no fully complete se- quences were filmed. Displaying males spread their small, brightly colored dew- laps to the fullest extent and, with head elevated, strained upward in several slow, irregular, bobbing movements. The dew- lap movements were the most conspicuous part of the display. The dewlap was grad- ually and fully extended during the first 1.2 seconds, then collapsed, and finally gave three double pulses, each shorter and more rapid than the previous one, in the final three seconds. The entire display lasted approximately five seconds. Anolis microlepidotus This species was described by Davis (1954) from three specimens collected in the mountains of southern Guerrero near Chilpancingo, and was characterized as small and slender, with short legs, small ear opening, keeled ventrals, small keeled dorsals, grayish coloration, and ruby red dewlap. Davis and Dixon (1961) reported it from additional localities in the same general area. They found it in pine-oak and tropical deciduous forest types, in both xeric and swampy habitats. We found Anolis microlepidotus at sev- eral localities in southern Guerrero and Oaxaca, at medium altitudes in rocky open woodland of scrub oak (Quercus sp.). The lizards were always associated with leaf litter and were at or near ground level. 110 Tue UNIVERSITY OF KANSAS SCIENCE BULLETIN An adult female found 4 km S Almo- longa, Guerrero, June 1971, was perched on a small flat boulder in a clearing; another was on a horizontal stick 75 mm above the ground. In the mountains of northern Oaxaca, 3.2 km E Ixtepeji at about 1830 m, an adult male was found in a ravine bottom and ran into the cavity at the base of a large oak tree before cap- ture. Two immatures were captured along the same ravine the following morning. One was in a small brush pile of oak cuttings and one flushed in a thicket ran up onto the stem of an oak bush. On 8 February in the mountains south of Oaxaca 8 km N Portillo San Andrés, 2,220 m, a juvenile female flushed from leaf litter climbed 75 mm on an oak stem 13 mm in diameter. The sky was overcast and the lizard was active at air tempera- ture of about 17° C. In life, these lizards were dull-brown dorsally, reddish brown head contrasting with body, and dewlap bright purplish red. Their habitats were among the most xeric in which we have found anoles of any kind. One of the females collected near Al- molonga in early July contained in each oviduct an egg nearly ready to be laid (11.0 x 5.6 mm left, and 11.2 x 65 mm right) and also contained large ovarian follicles (6.0 mm left, and 2.7 mm right). The other had an egg only in the right oviduct (10.1 x 6.2), but had large ovarian follicles on each side (7.2 left, 4.5 right). Both females show concentrated egg pro- duction after the onset of summer rains. The presence of juveniles in February seems to show that egg-laying may con- tinue into November, and perhaps incu- bation and early growth are much slower at the altitudes where A. microlepidotus occurs than they are in species of the neighboring lowlands and foothills. When an adult male and two imma- tures were confined together they dis- played several times. One sequence, filmed from the adult male, may be complete. As in other species of Guerreran anoles the display was relatively simple and was notable for the dewlap movements. The displaying lizard did not greatly elevate its head as do some other kinds, and there was conspicuous antero-posterior rocking movement accompanying the display bobs. The display consisted of three slow and high bobs with intervening deep bows and with a pause preceding the last bob. Each bob was accompanied by an exten- sion of the dewlap, about halfway on the first one, fully on the second and about three-fourths on the third. This main part of the display lasted for 9 sec. In the following 7 sec after the bobs there were two small dewlap pulses, the second barely perceptible. Anolis parvicirculata This species has been reported only in the original description (Alvarez del Toro and Smith, 1956) from two specimens col- lected at Suspiro (between Ocozocuautla and Quechula) in west-central Chiapas, Mexico. The authors characterized the species as a member of the “heteropholido- tus group,” thus associating it with Anolis heteropholidotus, a slender, long-tailed, high-montane species of El Salvador. Ano lis parvicirculata has smooth ventrals, and has small scales in the supraorbital semi- circles. An adult female, captured near the type locality (3 km NW Suspiro) 15 February, 1972, was in life olive drab dorsally with a narrow pale tan, middorsal line; ventral surface pale grayish brown with coarse faint speckling laterally; legs were of the same color as body with no discernible markings, but the fingers and toes had faint dark bands; the dewlap was small and uncolored. This is an unusually slender species, most readily recognizable by the extremely large eye with dark iris. The female’s length of 50 mm (S-V) equalled the length FIELD OBSERVATIONS ON MAINLAND ANOLES 11] of the paratype and exceeded that of the male holotype (46 mm). As in most other rain-forest species the male probably is little, if at all, larger than the female and may average smaller. The female contained a full-sized ovi- ducal egg and a large ovarian follicle, although females of several other species collected in the same month in Chiapas and Oaxaca in more xeric habitats were all nonreproductive. Anolis subocularis In the original description, Davis (1954) listed 100 specimens from Tierra Colorada and nearby localities in southern Guerrero. He stated that some of the specimens had previously been reported as Anolis nebuloides, but they were different from that species in having dorsal scales noticeably smaller than ventrals. Other diagnostic characters mentioned were: Row of small intercalated scales between suboculars and supralabials; ventral scales keeled; long hind leg, reaching forward beyond eye; dewlap orange or reddish (p. 4, but also stated to be ruby red, p. 6); males considerably larger than females. Davis stated that Anolis subocularis was the commonest anole in Guerrero below 1500 m elevation. An adult male we collected at Puerto Marqués had the following appearance in life: Dorsal color yellowish olive, with narrow, jagged-edged lateral brown line from above axilla to above thigh on each side and a second lower line, from behind axilla to thigh; several whitish spots with dark borders on each side; two longi- tudinal ventrolateral rows of black dots; ventral surface dull white; five pairs of dorsolateral light spots with dark edges; broad, faint, dark bars on limbs, more distinct on toes; top of head spotted with black, largest spots in supraocular region; tail faintly banded; dewlap red, extending posteriorly to level of elbows. We found Anolis subocularis at many localities in southeastern Guerrero and southwestern Oaxaca, sometimes in great abundance, in a variety of habitats, as in- dicated by the following field notes in July 1971, and February 1972. 25 February, 12 km ESE Rio Grande, Oaxaca, in oceanside coconut grove. Many anoles were found mostly among dried palm fronds on ground. 26 February, .£8 km N Guachupin, Oaxaca. East-facing hillside with open type of scrub forest (trees leafless then in dry season) interspersed with gigantic boulders, some in groups. Anoles were found on bare rock-faces and on tree trunks, within 1 m of ground. 26 February, 9.0 km E Jamiltepec, Oaxaca. One juvenile found in thicket beside swift, rocky stream. 27 February, 7.2 km W La Estancia, Oaxaca. Low, rolling grassland, recently burned off. Along a gully, which then, in the dry season, had a mere trickle of wa- ter, there was a band of green vegetation, with trees, brush and some herbaceous growth. Anoles were abundant in ground litter and on stems and tree trunks near the ground. 27 February, Ometepec, Guerrero. One found at dusk beside swift, rocky, moun- tain stream. 28 February and 21 and 22 July, 12 km W Marquelia, Guerrero. Anoles were found in moderate numbers on coconut palms in an open grove. Most often they were between 1 and 2 m on the trunk, clinging head downward and were con- spicuous from a distance. They were somewhat wary and upon approach of a person would ascend the trunk far out of reach. 29 February, Puerto Marqués, Guer- rero. Several were seen on the steep moun- tainside (strewn with gigantic boulders, interspersed with gnarled trees and thick- ets). Here they were associated with the much more numerous Anolis taylori and the relatively scarce A. nebulosus. A. tay- 112 Tue UNIVERSITY OF KANSAS SCIENCE BULLETIN lori was most often on rock surfaces, whereas A. subocularis was here seen only on tree trunks. 21 July, 7.2 km ESE San Marcos, Guer- rero. Three adults were taken and others seen on the trunk of an isolated large thorny tree at the edge of a cultivated field. 21 July, 14.3 km N Cruz Grande, Guer- rero. Eight were captured and others seen on tree trunks in a riparian grove. All those collected or seen in late Feb- ruary and late July were adults. The February specimens were sexually inac- tive, with neither oviducal eggs nor en- larged follicles and with fat bodies minute. All seven July females were sexually active with a total of 9 oviducal eggs and 13 enlarged follicles. All but one had an enlarged follicle in each ovary; three each had 2 oviducal eggs, three each had 1, and one had none. Evidently, breeding had begun within the previous month, after the onset of the summer rains. Breeding must be limited to late summer and early autumn, probably ending in October or even earlier, because young had all at- tained adult size by late February. Sexual difference in size is much greater in Anolis subocularis than in most other mainland anoles. Mean female-to- male length ratio was 76.1°/,. Mean female- to-male weight ratio was 49.194 in Feb- ruary, but had increased a 83.59% in July when all females were gravid. Six displays of four individuals were analyzed. Initial and terminal portions of the display were often given separately. The initial portion, less stereotyped than the terminal portion, consisted of three or four bobs, a pause (2.5-5 sec) and another bobbing sequence, sometimes followed by a third. The terminal portion (Fig. 1) consisted of 11-17 bobs. Between the first two major bobs of the terminal series there were 3-6 short, rapid bobs. Each major bob consisted of a rapid upward movement from which the head was low- ered about one-third of the downward distance and paused .2-.6 sec before it low- ered the remaining distance. The dewlap was spread and retracted three times dur- ing the initial portion of the display and four times during the following pause, and was maximally extended during the long series of terminal bobs. It was re- tracted after the last bob but subsequently might be pulsed two or more times. Anolis taylori This species was described by Smith and Spieler (1945) from eight specimens collected in the vicinity of Acapulco, Guer- rero. It was stated to be a “rock anole” closely related to A. gadovii and A. dunni, separable from the former by having only 2 (rather than 3 or 4) gulars in contact with mental, and separable from the latter by having smaller ventrals and having 7 or 8 (rather than 6) scales between nasals. Smith and Taylor (1950) included A. tay- lori in their checklist of reptiles of Mexico, but the species seemingly has not been in- vestigated further in the last 25 years. Fitch and Henderson (1976) described the structural niche and field behavior of a population of A. taylori at Puerto Marqués 7 km SE Acapulco. These lizards are ex- tremely abundant in their chosen habitat of enormous boulders on steep, wooded mountainsides. In late July (early in the rainy season) only adults and adolescents could be found. All of the 13 adult females dis- sected were gravid. Five had an oviducal egg on each side and also an enlarged fol- licle in each ovary; 2 had an egg on each side, but had an enlarged follicle in only one ovary; 5 had an egg in one oviduct and a large follicle in each ovary; and one had only one egg and (on the opposite side) an enlarged follicle. Reproduction seemed near a peak at that time, with eggs being produced in rapid succession. In late February and early March (late in the dry season) the population sample con- Fie_p OpsERVATIONS ON MAINLAND ANOLES 113 sisted largely of immatures, mostly more than half-grown, and adult females were nonreproductive. The incidence of reproductive females and young of various sizes indicated sus- pension of reproduction in the dry season, perhaps from November through June, and resumption of breeding in July, with hatchlings beginning to appear around August and continuing to appear into Oc- tober or November. A relatively short and concentrated breeding season, as compared with those in other anole species, is indi- cated. Fitch and Henderson (1976) recorded a mean of approximately 29° (26.0 to 32.8) for 30 body temperatures of A. taylort. Generally air temperatures were near these levels when the lizards were most active. Activity tended to reach a peak about mid- morning and was much reduced at mid- day with some tendency to increase again in late afternoon. The display of A. taylori was distinc- tive in having the dewlap fully extended or in motion for the entire sequence and constituting the dominant component, whereas head movement was minimal, generally consisting of 3-5 preliminary twitches. The dewlap was first slowly and fully extended, then gradually retracted (1.90 = .092 sec, n = 18). The extension and retraction covered approximately equal intervals, with a slight intervening pause (.64 = .092 sec, n = 18) while the dewlap was held maximally extended. There was another slight pause (.98 = .056, n = 18) between the retraction and the pulse that followed, during which the dewlap was spread to no more than half its full extent. A third extension, usually to no more than three-fourths of full extent, ended the SeniccmGloon=—all9snsec, m — 18): Nhe pulse between the two major extensions ranged from a completely separate move- ment to a momentary pause in the second major extension. The active, pulsing move- ments rendered the dewlap-display highly conspicuous. The final retraction of the dewlap was often much more prolonged than the retraction occurring after the first dewlap display. In contrast to the first dewlap extension, the final extension im- mediately graded into retraction with no pause after maximal extension (Fig. 4). Some variation in the display was noted. The dewlap was occasionally ex- tended once, before (5 displays) or after (3 displays) the sequence described. Usu- ally there was a pause of two or more seconds between these extra dewlap exten- sions and the main sequence. Tue CENTRAL AMERICAN SPECIES Anolis aquaticus This species was described by Taylor (1956), who had one specimen from Pal- mar and two from Golfito, both in Pun- tarenas Province, Costa Rica. Diagnostic characters listed in the original description include: Heavy, dark, transverse bars wider than interspaces on body, limbs, and tail; dewlap, large, orange; tail com- pressed with a middorsal row of enlarged serrate scales; ventrals keeled and much larger than dorsals; scales of middorsal pair of rows on body markedly enlarged; suboculars separated from supralabials by two or three rows of small scales; three or four scale rows between supraorbital semi- circles. In our study many A. aquaticus were observed and others were collected and preserved at Finca Las Cruces near San Vito, Puntarenas Province, Costa Rica. All of the lizards were found along swift, flow- ing, rocky, mountain streams. Usually, they perched on rocks; sometimes they were on logs or sticks. They were wary and when approached they dove into the water, swam for distances up to several meters, and emerged in well concealed places, such as cavities beneath log jams or overhanging rocks, or in accumulations of drift. 114 Tue UNIVERSITY OF KANSAS SCIENCE BULLETIN Of 9 adult females collected at Finca Las Cruces, 14 March, 1974, two had neither oviducal eggs nor enlarged folli- cles, but the remaining seven were all re- productive. Four each had an oviducal egg on one side and an enlarged follicle on the opposite side. One had an enlarged follicle on each side (4.5 and 3.4mm), but lacked oviducal eggs. One had an oviducal egg (left), but no follicles. One had an enlarged follicle (5.9 mm, left), but no oviducal eggs. In other species of anoles, it is most typical for a female to have two oviducal eggs simultaneously—one larger than the other and destined for earlier laying— with ovarian follicles also developing, ready to replace each egg soon after it is laid. The smaller number of such repro- ductive units in the females of A. aquati- cus suggests that it has a lower reproduc- tive potential, although rate of develop- ment and extent of seasonal change are not known. In both February 1973 and March 1974, young of various sizes were seen (though in small numbers), suggest- ing year-round reproduction, as might be expected in the consistently warm and humid climate where A. aquaticus occurs. Anolis attenuatus Taylor (1956) named this anole (as a subspecies of Anolis wood1) from three specimens collected at Isla Bonita, altitude 1680 m, on the southeastern slope of Vol- can Pods, Heredia Province, Costa Rica. Diagnostic characters that he listed in the original description include: Large size (S-V 78-83 mm); tail 24 times body length; ventrals keeled and slightly larger than dorsals; approximately 140. scales about body; postanals enlarged; extended hind leg reaching forward beyond eye; extended foreleg reaching back to groin; body dull-red with black markings; dew- lap large, dark olive, with magenta scales. Anolis woodi was described by Dunn (1940) from El Volcan, Chiriqui, Panama, from a single female specimen, but Taylor (1956) reported it from Cafias Gordas, Puntarenas Province, Costa Rica, and other specimens have been reported from several localities in the southern half of Costa Rica. A. woodi and A. attenuatus are both large, slender, long-legged and long-tailed anoles with large dewlaps. However, they differ in body size, relative sizes of the sexes, color of body and of dewlap, and relative length of tail and limbs (Table 6). In view of these numer- ous and trenchant differences and the fact that their known ranges are well separated, so that there seems to be no opportunity for geographic integradation, they are best considered distinct species. We found Anolis attenuatus at Monte- TABLES CoMPARISON oF Anolis attenuatus To A. woodt A. attenuatus A. woodt Geographic range Size S-V mean for ¢ @ 84.5 in 24 mean for 2 @ 80.6 in 18 2 to 6 size ratio SEI E Dorsal color Dewlap color Dark olive or black Approximate number of scales around mid-body 140 Cordillera de Tilaran and C. Central Dull red with black markings Cordillera de Talamanca 80.8 in 4 69.9 in 10 86.6% Olive with rusty spots Bluish white at base, bordered with amber yellow; pink orange on outer edge 125 FIELD OBSERVATIONS ON MAINLAND ANOLES MS verde, Puntarenas Province, Costa Rica, 70 km W and 11 km N of the type local- ity; from March 1973 to March 1974, 49 records were accumulated, many of them by Dr. Richard K. La Val. Most of the lizards were found on tree trunks; a few were on the ground. These lizards were slow-moving and depended largely on cryptic behavior and coloration to avoid detection. Once seen, they were easily captured. On many oc- casions those that were captured and re- leased, or merely observed in the field, were seen again after intervals of hours in almost the same location and position. In their montane habitat, mean air tem- perature was about 18° C at the times of capture. The seasonal distribution of fe- males that appeared gravid (March, July, August, September, and October) and of immatures suggested year-round reproduc- tion. Taylor (1956) described the behavior of two of the original specimens from Isla Bonita, before they were collected, as fol- lows. - on the top of a tree fern . . fighting. The male with dewlap extended would attempt to bite the female. Then the female would retreat to another frond and shortly would return to attack the male . . . who at all times had the throat fan extended. ...” In our study, presenta- tion of mirrors and of transferred indi- viduals and placing individuals together in confinement did not elicit display, ex- cept when a juvenile was placed with a juvenile A. insignis and responded to its challenge. The displays were notable for the small amount of movement and the long pauses, with the large, dull-colored dewlap held fully extended (Fig. 5). The single display analyzed lasted 20 sec, with initial, middle and terminal phases sepa- rated by pauses of about 3.3 sec. The first phase consisted of a slow elevation of the head and spreading of the dewlap, then a slight dip and pause (1.4 sec) and a high bob interrupted by a half-second pause near the top, a very rapid dip closely fol- lowed by another high bob and dip, The middle portion following the first main pause, began with a small dip followed by a small bob, short pause, second dip and medium bob. The terminal portion consisted of a small dip and then a rapid small bob followed by a medium bob with a pause (.7 sec) at the top, a small dip, pause, and final retraction of dewlap. Anolis biscutiger Taylor (1956) named Anolis biscutiger from 11 specimens collected in Puntarenas Province, Costa Rica (Golfito 1—the holo- type, Palmar 8, 24 km WSW San Isidro del General 2). Among the diagnostic characters mentioned were: Small size (length S-V less than 40 mm); tail twice snout-vent length; extended hind limb reaching between eye and nostril; two pairs of much enlarged postanals; supra- orbital semicircles separated by one scale row; seven loreal series. Taylor did not specifically discuss the relationships of A. biscutiger and A. limifrons, but his de- scriptions of the two differ only in minor details and obviously they are closely re- lated. Savage (1973) in his list of the amphibians and reptiles of Costa Rica did not include A. discutiger, but considered it a synonym of A. limifrons (pers. comm.). Williams and Smith (1966) recorded it from Julieta and this locality remains the northwesternmost record. Accumulated evidence has led to a reconsideration of the relationship of A. biscutiger to A. limifrons. The differences between them, though small, are numer- ous and fairly constant. The known range of A. biscutiger is in the Pacific lowlands of Costa Rica from Julieta to Golfito, effectively separated from A. limi- frons in the Caribbean lowlands by the mountain-mass of the Cordillera de Tala- manca. The type locality of Anolis limt- frons is Cucuyas de Veraguas, Panama. Fresh material from there is needed for 116 Tue UNIVERSITY OF KANSAS SCIENCE BULLETIN comparison with other populations, in- cluding those of A. biscutiger and Costa Rican A. limifrons. However, it is known that Panamanian A. limifrons differ in size, proportions, dewlap-color and scala- tion from both these Costa Rican popula- tions. Some characters of A. biscutiger which reflect its relationships are the following: Female-to-male length ratio 105.6%, males 37.3 = 332 mm, 43-33 in 42, females 39.4 + 416 mm, 44-36 in 33; tail to snout-vent ratio 1.96%, 1.77 to 2.18 in 43; postanals much enlarged in males; adpressed hind limb reaches between eye and nostril; ad- pressed forelimb reaches beyond snout. A total of 33 body temperatures were recorded for A. biscutiger at Quepos. All but two were in the range 30-33 and nearly half were in the interval between 31 and 32. In samples collected at Quepos in December, January, February, March, May and August, most females were gravid and juveniles were not prominent in any of these samples. Hence, it seems that in the humid climate where A. biscu- tiger occurs, its breeding is extended throughout the year, or much of it, but probably with changing levels of inci- dence. The display begins with a series of 3-18 bobs (duration, 25 =: A7 sec, n = 11) as the dewlap is slowly extended. A series of 25 inverse head bobs (Echelle et al., 1971a) follows (duration 1.57 + .183 sec), during which the dewlap is gradually retracted. The pause between the two parts of the display is short (.12 + .051 sec). The dewlap is pulsed submaximally during each down movement of the in- verse head bobs, which are of small ampli- tude. The display is relatively short, aver- BIS nee ps) Se) Flop ee Anolis carpentert This species was described by Echelle, et al. (1971b) from 10 specimens collected at the Rio Reventazén, Turrialba, Cartago Province, Costa Rica. Distinctive features indicated were: Small adult size (35 to 45 mm S-V); grayish green color; buffy eye ring; female-to-male ratio 104.9 per cent; relatively short tail (1.61 times body length); hind limb extending to ear or slightly beyond; forelimb extending to midway between eye and nostril; and orange dewlap. Myers (1971) described the species (as Anolis procellaris) from a single male (illustrated with a photograph from life) from Veraguas Province, Pan- ama. A fairly extensive range in the Ca- ribbean lowlands of Central America is indicated. There is one specimen in the K.U. Natural History Museum from 10.5 km N and 9 km E Matagalpa, Nicaragua. One was seen, but not captured, at Bev- erly, Limén Province, Costa Rica, and numerous individuals have been captured or seen at Finca La Selva, Heredia Prov- ince, Costa Rica. At the latter locality, females examined and released in February, April, August, October and November all appeared to be carrying eggs and probably there is some reproduction throughout the year. In the display, an initial crouch was followed by a series of two to four ascend- ing push-up bobs and these were followed by a series of four to six slower and more uniform bobs, with a distinct pause after each. In three of seven displays, there was a preliminary series of two to four low, short bobs before the ascending push- ups, but in the remaining four displays this preliminary series was absent. The dewlap was fully extended during the entire display, which lasted about 15 sec- onds. In two displays, the tip of the conspicuous, red tongue protruded from the mouth. A more detailed description of the display was included in the original description of the species (Echelle et al., 19/iby Anolis dollfusianus This small Guatemalan species was FreLp OBSERVATIONS ON MAINLAND ANOLES 117 named by Bocourt (1873) from San Au- gustin, Volcan Atitlan, altitude 1200 m. Stuart (1955) made a detailed comparison of A. dollfusianus with A. cupreus, imply- ing that the two might be closely related and might even intergrade, as their ranges seem to be complementary. He stated that A. dollfusianus occurs commonly in the upper part of the tropical zone between altitudes of 600 and 1500 m and is abun- dant in coffee groves. He erroneously stated that the adult male of A. dollfusi- anus is only 25 mm in snout-vent length. Other diagnostic characters which he men- tioned were the yellow dewlap, keeled ventrals, rugose plates of occipital region, and anterior head shields with upturned edges and central keel, making them ap- pear tricarinate. Field observations indicate that A. doll- fustanus occurs in high population densi- ties in favorable habitats. However, it has remained little-known because of the small geographic range, extending from the vi- cinity of Escuintla in Guatemala west to the vicinity of Tapachula, Chiapas, Mex- ico, a distance of approximately 270 km, in a band of perhaps half that width, at low and medium altitudes on the south slope of the Sierra Madre del Sur. Our field observations on A. dollfusi- anus were made near Retalhuleu, Guate- mala, 16-24 February 1971, and Tapachula, Chiapas, Mexico 17 February 1972. At the Guatemalan locality four of the six adult females each had a single oviducal egg and an enlarged follicle in the opposite ovary. A fifth which lacked the oviducal egg had an enlarged follicle (3.4 mm). The sixth female had a follicle that was only slightly enlarged. Approximately half of the 64 anoles in this sample were immature, but only five were less than half-grown. In contrast, the series of 53 from south- ern Chiapas had only six immatures (from half-grown to adolescent size), and only one of 13 adult females had an oviducal egg, while one other individual had an enlarged follicle. Seemingly, the timing of reproduction is influenced by locality and weather. At low altitude in southern Chiapas, there was little reproduction in late fall and winter 1971-72, but at higher altitude in southwestern Guatemala in a somewhat cooler and moister climate in late fall and winter 1970-71, reproduction was maintained at a considerably higher level. Fifty-five displays were analyzed, 41 from 14 males from near Tapachula, Chia- pas, Mexico, and 14 from 9 males from Retalhuleu, Guatemala. The two localities are 7) km apart and average differences were discernible in the displays. The most common display was a series of 3-10 rapid low-amplitude bobs that averaged 2.32 + 1.77 sec for the Guatemalan lizards and 293) 2= -2270seq for the Mexicans lizards and were accompanied by a single dewlap movement of 1.66 = .094 sec (Mexican) on 3:29==)230 sec) (Guatemalan):) Inethe Mexican lizards, dewlap extension began 1.07 + .061 sec before the end of the last bob and 1.19 = .140 sec after the begin- ning of the first bob. For the Guatemalan lizards, corresponding figures were 2.28 = 198 sec and 66 + .111 sec. Although dewlap extension was definitely more pro- longed in the Guatemalan lizards, the displays in the two populations were es- sentially similar (Fig. 2). Anolis haguet In the original description of Anolis haguet, Stuart (1942) distinguished it from the closely related Anolis crassulus by its much smaller dorsal scales. Later the same author (Stuart 1948, 1955) relegated haguet to subspecific status, indicating that it replaced Anolis crassulus crassulus in cloud forests of Alta Verapaz, Guatemala at altitudes above 1300 m. Smith and Ker- ster (1955) reported a single specimen, al- legedly from “Region Soconusco” in the Pacific lowlands of southern Chiapas, 118 Tue University oF KANsAs SCIENCE BULLETIN Mexico, and suggested that haguer was a species distinct from crassulus. Smith, Burley and Fritts (1968) described Anolis anisolepis from San Cristobal de las Casas in the mountains of central Chiapas, as a near relative of both A. crassulus and A. haguei, and recognized an ‘anisolepis group’ of species characterized by small size and having scattered small groups of enlarged scales irregularly distributed over the sides where most of the scales are minute granules. Members of the ani- solepis group also have strongly keeled ventral scales, enlarged dorsals, and knobby, rugose dorsal head scales. The diagnostic characters that distinguish 4. haguei from A. crassulus and A. anisolepis are listed under the account of the latter species. On 16 February 1971, four of these anoles were captured and many others seen in a roadside hedge of agaves 2.4 km west of Quetzaltenango, Guatemala. The lizards were active and shy, responding to approach of a person by running back toward the stalk at the base of the leaf. The long, rigid leaves with recurved spines along their margins provided effective pro- tection. The dorsal pattern of grayish brown with a series of paired dark blotches rendered these lizards remarkably like Sceloporus in appearance and the impres- sion was heightened by their behavior and by the xeric aspect of the habitat. Two females had neither oviducal eggs nor en- larged follicles, but they may not have been fully mature. The display usually consisted of 4 or sometimes 5 series of rapid bobs inter- rupted by short pauses. The final or main series of bobs had more than the pre- liminary series. The first consisted of 2 quick bobs with no extension of the dew- lap. The second was longer (average 1.24 + 089 sec, n = 11) with 3 to 6 bobs and a pronounced dip of the head toward the end of the series. A dewlap flash of .42 = 045 sec, n = 11 accompanied this series. Compared with the display of 4. crassulus from Panajachel, Guatemala (Fig. 2), the display of A. haguet was more variable, with an apparent repetition of the second preliminary bobbing series of A. crassulus. Homology with the second series of A. crassulus was predicated on the form of the bobs and the fact that they were ac- companied by a dewlap movement. The third preliminary series was the shortest (.24 += .052 sec, n = 11) and usually con- sisted of a single quick bob occurring within a second before the final bobbing series. The latter involved 4-10 bobs and another dewlap flash (.61 + .065 sec, n = 11). Twice in displays of high intensity, two deep bobs (adding 2.6 and 4.1 seconds to the duration of the display) preceded the main display as described above. Dur- ing these bobs, the dewlap was extended, the mouth gaped, and the engorged red tongue was conspicuous. Anolis insignis This species was described by Cope (1871) from a specimen purportedly from San José, Costa Rica (collected by a Dr. Van Patten). Diagnostic characters men- tioned were: Giant size (total length up to 440 mm, snout-vent 145); ventrals smooth; tail base compressed; extended hind limb short of ear; ear opening half size of eye; body light brown, with four transverse double bands of greenish blue, a large ocellate spot, greenish blue with brown center anterior to axilla. Taylor (1956) did not find the species in the course of his extensive collecting in Costa Rica, but he cited published records including La Palma between Guapiles and San José, Cariblanco, Ballena, and El Valle, Panama. Peters and Donoso-Barros (1970) stated the range to be “Panama to Costa Rica in mountainous areas.” The single specimen obtained by us, at Monteverde, Puntarenas Province, Costa Rica, constitutes an extension of the known range northwestward into the FIELD OBSERVATIONS ON MAINLAND ANOLES 119 Cordillera de Tilaran. It was a hatchling and was on a horizontal branch 2.55 m above ground. Local people reported find- ing a giant anole, almost certainly 2. insignis, on the ground in a road, near the same location. When approached it dis- played but did not attempt to escape. Probably the rarity of A. insignis in col- lections results partly from highly arboreal habits and preference for large trees in dense primary forest. In displaying, the juvenile male held his head aligned with the body or only moderately elevated, not tipped far back as in most other species, and the dewlap was not extended to its maximum. The display consisted of a long series of bobs, with either two or three pauses of from 1 to more than 2 sec at low points between successive bobs. In each of two series, duration was 20 sec and there were 13 bobs (Fig. 1). Anolis rodriguezt This species was named by Bocourt (1873) from Panzos, Guatemala. Stuart (1948) discussed its similarity to Anolis limifrons of southern Central America and relegated A. rodriguezi to subspecific status under the former species. The range of A. rodriguezi extends from Central America into the Yucatan Peninsula, and various authors have commented on its distribution, habitat and lepidosis in that area. Whether anoles of the /imifrons species- complex occur uninterruptedly through the Caribbean lowlands from Costa Rica to Guatemala remains to be demonstrated, but there are extensive gaps in the known range and geographic intergradation be- tween rodriguezi and limifrons has not been established. Seemingly, there are well- defined habitat differences. A. limifrons occurs in primary lowland rain forest and in edge situations. In Yucatan, Duellman (1963) found A. rodriguezi in great abun- dance in xeric areas of scrub, but found it to be rare in rain forest. Etheridge (1960), in his study of skele- tal morphology of the anoles, found A. rodriguezi to agree with species of the Chrysolepis series, whereas A. limifrons and its near relatives were allocated in the Fuscoauratus series. In view of this dis- crepancy and other differences between the two (Table 7) it seems inadvisable to assign rodriguezi to the species A. limi- frons without actual proof of intergrada- tion. Differences in habitat, body size, proportions, sexual dimorphism, dewlap color and aggressive display provide fur- ther evidence against conspecificity of the two forms. Two distinct types of displays were ob- served in A. rodriguezi. The more com- mon display began with 2-5 preliminary bobs of 2.8 + .39 sec, n = 13. The dewlap was then fully extended during an as- cending series of bobs (2-17, most com- monly, 5, sim D/6,== Ie7isecene—— 15)iqtol- lowed by 2-3 descending bobs in 1.44 = 119 sec, n = 12. Thereupon, the dewlap was partly retracted during a “plateau” bob (1.78 = .169 sec, n = 11), followed by 1-16 bobs with the head held alternately high and then low for approximately equal intervals, producing a rectangularly shaped bob. Mean duration of this series of bobs was 6.2 == [2lisecen — 9 During these bobs the dewlap was pulsed—outward as the head lowered and inward as it rose. The rectangularly shaped bobs closely re- semble the “inverse head bobs” of A. &1- scutiger and those of A. limifrons de- scribed by Echelle e¢ al. (1971a). The less common type of display was observed only twice, performed by dif- ferent individuals. These displays lasted 5.3 and 6.3 sec, and consisted of two rec- tangularly shaped bobs, a plateau bob, and two final bobs. The dewlap was extended for the entire display in one instance and was pulsed in the other (Fig. 3). 120 Tue UNIversiry oF KANSAS SCIENCE BULLETIN Tue SoutH AMERICAN SPECIES Anolis aequatorialis This large species was described by Werner (1894) with no more specific lo- cality than “Ecuador.” Peters and Donoso- Barros (1970), described the geographic range as “middle altitudes of western slopes in Ecuador,” and they listed various diagnostic characters, including keeled ventrals, long hind leg extending beyond snout with tibia exceeding head length, smooth head scales and keeled supraocu- lars. Williams (1974) mentioned that noth- ing was known of the habits of 4. aequa- tortalis, but that, with A. mirus and his newly described A. parilis, it was a mem- ber of the eulaemus species subgroup of Alpha anoles. In this group, the toe pads are “Norops-like”—reduced as in strictly terrestrial kinds and not adhesive, so that the lizards are dependent on their claws for such climbing as they may do. Five Anolis aequatortalis were captured at Tandapi, Pichincha Province, Ecuador on 18, 19 and 23 February, 1975. All were between 2 and 3 m above ground, on outer twigs or foliage of small trees or herba- ceous plants. They were not wary or elu- sive, but seemed to depend on conceal- ment for escape, and were well hidden by screening vegetation. They were found in the same sorts of situations as the smaller and much commoner Anolis gem- mosus. Partitioning of resources was not evident. A live, adult male was described as fol- lows: Olive brown dorsally, with seven pairs of dark, dorsal chocolate marks each in the form of an elongate oval, meeting its counterpart middorsally in a chevron- like pattern; second series of oblique dark markings low on sides, partly merged with those of dorsolateral series; midventral area coppery; chin barred and speckled with black; supralabial region from loreals to ear dull white; limbs sharply barred; toes faintly barred; posterior half of tail black; top of head pale green spotted with black; large black patch with white specks on each side of neck; dewlap large, lacking bright colors but with strikingly checkered pattern of light and dark brown. A ju- venile of 43 mm (S-V) was much like the adults in color and pattern, but the area between eye and ear, dull white in adults, was malachite green, and the chin was pale blue. Anolts chloris This was one of several species de- scribed by Boulenger (1898) from Pa- ramba, Imbabura Province, in northwest- ern Ecuador. Peters and Donoso-Barros (1970) stated the range to be Pacific low- lands of Ecuador and Colombia, and Darién, Panama. They listed diagnostic characters, including keeled ventrals, ex- tended hind limb reaching a point between ear and eye, supraorbital semicircles sepa- rated by two rows of scales, five rows of loreals, and six supralabials anterior to ABE 7 Comparison oF Anolis rodriguezi to A. limifrons A. rodriguezt A. limifrons Geographic range Morphological relationships Beta anoles Size S-V 36 43.4 in 15 TS) 43.7 in 23 2 to € size ratio Dewlap color 100.9% Orange yellow Yucatan Peninsula and northeastern Guatemala Chrysolepis series of Central America in Caribbean lowlands Fuscoauratus series of Beta anoles 37.5 38.6 102.9% Dull white with faint yellowish central spot FIELD OBSERVATIONS ON MAINLAND ANOLES 121 center of eye. A. chloris is an Alpha anole of the Latifrons species series. In late February and early March, 1975, three of these lizards were captured at Tinalandia, Rio Pupusa, Pichincha Prov- ince, Ecuador, and several others were seen at the Rio Palenque field station. All were adults. Dorsal color was emerald green. The dewlap was medium-sized, pale greenish with a pale blue outer edge. The eyes were blue and the tongue was orange. Those seen at the Rio Palenque field station were all on balsa trunks, and only one was captured. Upon approach of a person the lizards would run several meters up the vertical trunks and escape. The sexes were not noticeably different in sIZe. Anolis gemmosus This species described by O’Shaughn- essy (1875, type locality unknown) was figured and further described by Boulen- ger (1885). It was included in the list of valid Neotropical anole species by Peters and Donoso-Barros (1970), who stated the range to be “lowlands of Pacific Ecuador” and summarized its scale char- acters and bodily proportions in their “ma- trix” for 116 anole species. Williams (1970) stated that its affinities were with other Ecuadorian species, the high-montane Anolis andianus, and A. fasciatus of the Pacific lowlands. All these are Alpha anoles of the Latifrons species series. Despite Peters and Donoso-Barros’ (op. cit.) statement, it is doubtful whether A. gemmosus occurs in the Pacific lowlands, as it seems to be a species of middle alti- tudes. It has remained poorly known. All published accounts are based upon preserved material, with erroneous state- ments concerning color and pattern. In 1967, John D. Lynch and Robert W. Henderson, while carrying on field work on frogs at Tandapi, Pichincha Province, Ecuador, found A. gemmosus to be abun- dant there and they collected a large series. Directed to Tandapi on Lynch’s and Hen- derson’s advice, we likewise found A. gemmosus to be fairly abundant on 18, 19, 23, 25 and 27 February and 15 March, 1975: Much variation, individual, sexual and ontogenetic, was evident among the speci- mens collected, and those seen in the field. In life the general ground-color was deep emerald green. As in many other kinds of anoles, females often had a pale, longi- tudinal, middorsal band, but its occurrence and development was variable. The band was pale tan to rust color, usually set off by black margins, but these margins varied from broad to narrow. Among 21 adult females, nine had a broad dorsal band, two had a medium band, three had a nar- row band, three had a broken or discon- tinuous band, two had remnants of a band and two were uniform-green. As in other species, the banded female pattern is al- ready present in hatchlings. Though lack- ing the dorsal band, males were almost as variable as females in their dorsal pat- terns. Some were almost uniform-green while others were heavily spotted. Small, round or oval, buffy yellow spots were arranged in lateral, slightly oblique series, with as many as eight spots on each side in each series. Irregularly arranged black dots were also present on some, concen- trated mainly in the ventrolateral area. Limbs were faintly barred, with black stip- pling on undersides of limbs and tail. Some white facial markings were present beneath the eye and extending onto the cheek. In most populations of mainland anoles, male dewlaps are notably homogeneous in color and markings, but 4. gemmosus is an exception. The highly-variable dew- laps did not seem to comprise well-defined classes but tended to form a continuum between extremes. At one extreme were dewlaps with little contrast, dull yellowish green on the basal area, shading to dull greenish yellow on the outer part. The 122 Tue UNIVERSITY oF KANSAS SCIENCE BULLETIN more contrasting and colorful type of dew- lap was similar in having a dull greenish yellow outer part, but the basal part was bluish green with six narrow sharply de- fined white stripes diverging from a cen- ter on the anterior basal portion. The stripes had bright blue edges proximally at their origins, but distally the blue changed to green and the stripes them- selves became suffused with the yellow background and finally blended into it and blended with their brighter colored edges. In our sample of 70, only 13 were im- mature. Three were judged to be in their first two months (33, 28 and 27 mm) and three others, of 39, 38 and 37 mm were almost surely younger than six months old and more than two months old. The re- maining six included four of adolescent size (53 and 52 mm, all females) and two that were more than half-grown (48 mm). A low incidence of reproduction through the fall and early winter months is indi- cated. However, at the time of sampling, in the rainy season, all females were repro- ductive. Seventeen each had two oviducal eggs; 11 each had an egg in only one ovi- duct and an enlarged follicle in the ovary of the opposite side; three each had only one oviducal egg, but no enlarged follicle; and three each had one enlarged follicle but no oviducal eggs. Some reproduction throughout the year is indicated, but with relatively low levels in the fall and early winter. Anolis maculiventris This small species was named by Bou- lenger (1898) from Paramba, northwest- ern Ecuador, and was stated to have a red dewlap, snout-vent length 45 mm, brown- ish and metallic purple dorsal color and smooth ventrals. Peters and Donoso- Barros (1970) listed additional characters: Frontal crests present; extended hind limb reaching between ear and eye; 3 or 4 series of scales separating supraorbital semicircles; 8 rows of loreals; 8 supra- oculars anterior to center of eye. A. macu- liventris is a Beta anole of the Fuscoaura- tus species series. A single adult male was found 22 Feb- ruary 1975 at Tinalandia, Rio Pupusa, Pichincha Province, Ecuador, low on a tree trunk. In life it was dull grayish brown dorsally, with a pale rusty-red dew- lap (having scales paler, yellowish), and with iris golden yellow. When confined with other male anoles, A. chloris and A. peraccae, this one displayed vigorously, although the others were unresponsive. The display was complex, consisting of a long series of bobs (16 in 17 seconds) highly variable in amplitude and duration, with occasional short pauses in both the elevated and depressed positions. The first bob was the highest and shortest (2.5 sec) and the dewlap was slowly extended to a maximum corresponding to the peak of the bob. It was then retracted about half- way and held so for 2.5 sec, further re- tracted and held for 4 sec, and finally was slowly spread (not quite to the maxi- mum) and retracted in the last 6 sec (Fig: 5). Anolis nigrolineatus This species was named by Williams (1965) on the basis of two specimens, from Machala and Guayaquil, El Oro Province, Ecuador. Williams (1974) men- tioned two additional specimens from Playa de Montalvo, Los Rios Province, Ecuador, and listed the following charac- ters: Scales across snout 8 or 9; scales between supraorbital semicircles 1 or 2, loreal rows 4 to 7; scales between semi- circles and interparietal 2 or 3; 7 to ll supralabials anterior to center of eye; en- larged postanals in male; one middorsal scale row enlarged. Williams called at- tention to the swollen rostrum of Anolis nigrolineatus, projecting beyond the tip of the chin. He interpreted this development as indicating relationship to other anoles FreLp OBSERVATIONS ON MAINLAND ANOLES 123 having bizarre nasal appendages—Anolis proboscis, A. phyllorhinus and A. laevis. These, with A. caquetae, A. dissimilts, A. nigrolineatus and A. punctatus were con- sidered to be near relatives comprising the Punctatus Group of Alpha anoles in the Latifrons species series. On 10 and 11 March 1975, many indi- viduals of this recently described species were observed in groves at the highway junction 5 km E Machala, El Oro Prov- ince, Ecuador. Males averaged larger than females (S-V 50.9 = .488, range 55-47, and 48.0 = 390, range 51-45). In life, the male was greenish with oblique rows of dark spots and fainter white spots on the sides and there was rusty suffusion on the sides and back of the head becoming fainter middorsally on the body. The undersur- face was whitish with faint dark spots, with limbs and toes barred; the bars were formed from dark specks in a reticulate pattern; eyelids were buffy tan and iris was yellow. The chin was speckled with dull green. The tail was black for the middle third of its length. The dewlap was relatively small and of unusual shape, more than twice as long as deep when fully extended, dull white with a jet black horizontal line along its base on each side. The female was similar, except in lacking the dewlap, in having a pale tan middorsal stripe with a double dark line along each edge, and in having faint dark spots on the sides. All the lizards seen were on cacao trees. They were shy and active, and at least half of those seen escaped. Most often they were first seen at heights between 1 and 3 m on the main trunks, but when approached they would dodge to the farther side of the stem and climb out of reach. All that were seen were adults. Ab- sence of young of any size indicated that there had been cessation of breeding activ- ity for several months, for an interval that must have included fall and early winter. However, 14 females dissected were all reproductive. Nine of them had an egg in each oviduct and the remaining five each had an egg in one oviduct and an enlarged follicle in the ovary of the opposite side. Thus, reproductive activity was near a peak at the height of the rainy season. However, southern Ecuador has a long and severe dry season and probably egg- laying does not occur during that interval. Anolis peraccae This species was described from Chim- bo, Ecuador, by Boulenger (1898). Peters (1959) synonymized Anolis irregularis (Werner, 1901, type locality “Ecuador”), indicating that nearly all characters listed by the describer were identical, and that the details of pattern, arrangement of head scales and numbers of digital lamellae in which the two taxons differed were sub- ject to much individual variation. Char- acters listed by Peters and Donoso-Barros (1970) for A. peraccae include: Smooth ventrals; frontal crests present; supraorbi- tal semicircles contacting supraoculars, separated medially by one or two series of scales and separated from interparietal by 3 or 4 series of scales; 5 or 6 rows of loreals; 6 or 7 supralabials anterior to cen- ter of eye. A. peraccae is an Alpha anole of the Latifrons species series. This small species was found to be rela- tively common and widely distributed in the lowlands of northwestern Ecuador. A total of 33 were captured in February and March 1975, at Santo Domingo de los Colorados, Rio Palenque field station, and vicinity of Tinalandia, Rio Pupusa. At least twice that number escaped. Com- pared with other small species, this one is extremely active and wary. Several were found on trunks of large trees in rain forest, a few others were on trunks of balsa trees and several were on roadside agaves. However, approximately 90 per cent were on banana trees. Most often the lizard was between one and three m above 124 Tue UNIvERsITY OF KANsAs SCIENCE BULLETIN ground on the banana stem, but at any movement in the vicinity it would run higher, having no difficulty in clinging to the smooth, vertical stems. Most ascended beyond reach and were lost in the crowns of the banana trees. Captures were made by using a straight pole, which was slowly extended until it touched the trunk above the anole. Then the pole was lowered along the trunk driving the lizard ahead until it had descended within reach, for capture by hand. Even those driven within reach usually escaped by running back up the stem, by jumping to another stem or leaf, or by finding shelter in a hole be- neath the roots. These anoles are dull and cryptically colored, with little difference between the sexes. The most conspicuous recognition marks were ventrolateral oblique dark streaks forming a fine grid-like pattern. The dewlap is relatively small in the male and absent in the female. Males were seen more often than females, probably because they perch in more exposed places. The 32 captured in late February and early March included 17 adult males, 10 adult fe- males, 3 half-grown and 2 small juveniles. This high ratio of adults indicated a low level of breeding in the months before the sampling. Seven of the adult females were dissected and all but one were reproduc- tive. Two each had oviducal eggs on both left and right sides; three each had one oviducal egg and an enlarged follicle on the opposite side; one had an oviducal egg but no follicle. This high level of reproduction, and scarcity of juveniles (only a little over 6 per cent of the sample) suggested recent increase from a low level of reproduction in fall and early winter. Anolis princeps Boulenger (1902) described this giant rain forest species from specimens from Rio Lita, Paramba, San Javier, and Sa- lidero, northwestern Ecuador. Peters and Donoso-Barros (1970) described the range as lowlands of northwestern Ecuador, and mentioned the following characters: Dor- sal color green with dark brown bars; ventrals not consistently keeled or smooth; extended hind limb reaching beyond snout; 4 or 5 series of scales separating supraorbital semicircles; 9 to 11 rows of loreals; 10 to 12 supralabials anterior to eye. Two adult males were captured at the Rio Palenque field station on 1 and 3 March 1975. In life they were olive green with two dark transverse bands on the neck and five more bands on the body, each consisting of a double series of five or six oval markings on each side (in both the anterior and posterior series). Some of the oval markings were like ocelli, hav- ing light centers. The ventrum was green- ish white. The area around the eye was divided into alternate paler and darker sectors. The large dewlap was white along its anterior edge, ivory or slightly greenish white on its posterior part. These large anoles were conspicuous on their low perches, and neither made any attempt to escape as the collector ap- proached and captured them by hand. One was .8 m above ground on a stem .05 m in diameter, the other was 1.2 m high on a trunk of .8 m diameter. Anolis sp. In February, 1975, at Tinalandia, Rio Pupusa, Pichincha Province, Ecuador, we found a species of anole that was obviously different from any described in published literature. Subsequently, at the Rio Pa- lenque field station, another specimen was found and the species was found to have been included in a key to the lizards of the Station, compiled by Kenneth A. Miyata of the Museum of Comparative Zoology, Harvard University. Mr. Miyata (pers. comm.) has informed us that the species is described in manuscript by E. E. Williams, and that it is tentatively asso- FIELD OBSERVATIONS ON MAINLAND ANOLES 125 ciated with Anolis lemurinus, a Beta anole of the Chrysolepis species series. Distinctive features of the species are the short, blunt head (with rostrum es- pecially shortened), long limbs and tail (Table 1), bold dark dorsal markings in- cluding a lyre-shaped occipital mark and a saddle-like sacral blotch and several in- tervening blotches on the olive ground color; dewlap that is relatively small and scarlet in the male, with a central black spot, about one-fourth as large and pink in the females, with the spot brown. At Tinalandia an adult male was in a thicket at the base of a steep bank, climb- ing 1.6 m above ground on the terminal twigs of a shrub. A juvenile of 49 mm S-V was flushed from screening vegetation on the vertical surface of a high rock wall and ran several meters along the wall be- fore capture. An adult female (73 mm S-V) cap- tured at the Rio Palenque field station in early March 1975 was larger than the male (70 mm). This female contained two large oviducal eggs. One seemingly complete display of the juvenile male from Tinalandia was filmed. It was relatively simple, beginning with a slow (4 sec) high bob which was followed by four that were lower and faster (2.0, ise L9eand 2.0 sec). The dewlap’ was fully spread near the peak of the first bob and was partly retracted with each down- ward movement. In view of the suggested relationship of this species to Anolis lemurinus, the display is of special interest for comparison with the latter. In five displays of three individuals of A. lemurinus bourgaei from Tikal, El Petén Province, Guatemala that were analyzed (A. F. Echelle ms) there were four slow bobs followed by a pause and then a final slow bob, with each ele- vation of the head accompanied by a dew- lap extension—a pattern somewhat similar to that of the present species. Displays of A. lemurinus lemurinus from the vicinity of Escuintla, Guatemala, did not follow this basic pattern and were less similar to A. 1. bourgaei and to the Ecuadorian species than these two were to each other. In each of four displays of the Escuintla anoles there were six bobs and one dewlap movement beginning at the end of the third bob or during the fourth bob, spreading maximally, then retracting at the end of the bobbing series (duration of display 3.4 = 52 sec, dewlap exposure 1.51 + .047 sec). Nine other dis- plays were much more complex, with long ascending series and descending series of rapid bobs, with the dewlap maximally extended most of the time and spread one to three times per display. Display posture of A. 1. lemurinus is distinctive. The head is pointed upward, and holding this posi- tion for long periods, the lizard occasion- ally flashes the dewlap in erratic fashion. The important differences between A. /. lemurinus and A. 1. bourgaei in display and posture, dewlap, and body size cast some doubt on their supposed conspeci- ficity. DISCUSSION The 25 kinds of anoles discussed above are almost equally divided among Mexi- can, Central American and Ecuadorian species, and the three geographic groups show some significantly different trends. Inasmuch as these three geographic areas are somewhat parallel in their range of climates, plant formations and agricultural crops, and each has many species of anoles, ecological equivalence might be expected among them. Each of the species included in this study and each of the better known common species in the same regions was compared with other species to determine the degree of ecological equivalence be- tween the three regions. For the common and dominant species of each region, especially, counterparts were sought that were similar in size, ap- pearance, habitat and habits. One example 126 Tue UNIVERSITY OF KANSAS SCIENCE BULLETIN of ecological equivalence is that of the semi-aquatic Anolis barker1, widely dis- tributed in Mexico, and its counterparts 4. lionotus, A. aguaticus and A. poecilopus, that are similar in size, habitat, and habits, in different parts of Central America. However, no other such clear-cut cases of ecological equivalence could be discerned. In Central America the three most abundant species are Anolis limifrons, A. humilis and A. cupreus. The former is a species of the humid, Caribbean lowlands in forest and forest-edge. It is found on the ground in leaf-litter and grass, but also is scansorial on shrubs, tree trunks and vines. A. humilis of the same area, but extending from the lowlands up into cloud forest, occurs on the forest-floor generally in deep shade, usually in leaf-litter but often climbs on buttressed roots and tree bases. A. cupreus of the relatively dry Pacific versant is somewhat more scan- sorial than A. humilis and somewhat less so than A. limifrons. It is most typically found in leaf-litter of evergreen gallery forests near streams. All three species thrive in some types of disturbed and al- tered habitats as well as in primary rain forest and in fact all have been found in near-maximum abundance in cacao groves. Also, they have all been found in banana and coffee groves. A. limifrons and A. cupreus, but not A. humilis, have been found in coconut groves on the tree trunks. None of the three seems to have close counterparts among any of the Ecuadorian species. In Central America, A. biscutiger and A. rodriguezi are near relatives of A. limifrons that seem to replace it ecologi- cally in some areas. A. dollfusianus is somewhat like A. cupreus and even more like A. limifrons in habits and behavior. In southern Mexico, both A. subocularts and A. cuprinus are much like the Central American A. cupreus, but both occur in more open and xeric situations. Anolis sertceus, widely distributed in Mexico and Central America, resembles A. cupreus somewhat, but it is smaller and more slender and where the two overlap, at least, A. sericeus is definitely more scan- sorial. None of the three commonest Cen- tral American species (A. limifrons, A. humilis, A. cupreus) has an approximate equivalent in Mexico. These three species, together with A. sertceus, A. lemurinus and A. tropidonotus occurring in both Central America and Mexico, and A. nebulosus and A. subocu- laris of southwestern Mexico, are all widely distributed and abundant, and sym- patric with many additional species. Throughout Mexico and Central America one or another of the seven species men- tioned tend to dominate habitats that are suitable for anoles and to exert selective pressure on other kinds that are less widely distributed and usually are less numerous in areas of co-occurrence. The Ecuadorian species studied oc- curred in low population densities (only A. gemmosus was moderately abundant), were mostly medium to large Alpha anoles with little sexual difference in size, with arboreal habits, long tails, and with dew- laps often having white (4), green (3), and less frequently blue (2), red (2), and yellow (1). Most of them were notably shy and elusive. In contrast the Central American spe- cies often occurred in high densities, were mostly Beta anoles of both Chrysolepis (5) and Fuscoauratus (2) series but with one Alpha. They ranged from extremely small to very large, were found at ground level or were scansorial low on shrubs and tree-trunks, had relatively short or medium-length tails, dewlaps usually yel- low or orange, and had but little sexual dimorphism, with the female larger in several. In the Mexican species, the trend was again somewhat different. Several kinds were found in high population densities, others were uncommon even in the limited areas where they occurred. All nine spe- FIELD OBSERVATIONS ON MAINLAND ANOLES 127 cies were Beta anoles of the Chrysolepis series, having short to medium tails, ground-living or rock-living habits (or if scansorial, occurring only on bushes and low on tree-trunks), small or medium size, dewlap often all red or partly red, sexual size-difference extreme in some, with males usually larger than females. Divergent ecological trends in the three geographic groups may have some phylo- genetic basis, with Alpha anoles tending to be larger and more arboreal and having longer tails. Also, climatic trends are closely correlated with the ecological trends observed. Most of the Mexican species have small geographic ranges or are even confined to a single locality within a few square kilometers. In those instances it seems that the populations are relicts, surviving precariously in a region that is almost too xeric to support anoles. They have managed to survive only where local conditions mollify the aridity of the climate, and have undergone some adap- tive specialization in the process. In arid climates, such as are inhabited by most of the Mexican species, successful reproduction can occur only in the rainy season. Limitation of breeding to one part of the year intensifies its stresses. Males, needing to establish territories, repel rivals and find and court females, are selected for size, aggressiveness, and effectiveness of display. Sexual dimorphism becomes more pronounced than in the species that live in the relatively uniform and benign en- vironment of a rain forest or cloud forest. LMERAGURE ClhlrED ALVAREZ DEL Toro, M. AND H. M. SmitTH. 1956. Notulae Herpetologicae Chiapasiae I. Herpe- tologica, 12:3-17. Bocourt, F. 1873. Recherches zoologiques. Etudes sur les reptiles. I Mission Scientifique au Mexique et dans |’Amerique Central. Livr. 3:28-1000. Notes erpetologique. Ann. Sci. INatjasers 5; vol: -17- BouLEeNcER, G. A. 1885. Catalogue of lizards in the British Museum (Natural History). Edition 2. London. Vol. 2. xiii + 97 pp. 1898. An Account of the reptiles and ba- trachians collected by Mr. W. F. H. Rosen- berg in western Ecuador. Proc. Zool. Soc. London, 107-126. 1902. Descriptions of new batrachians and reptiles from northwestern Ecuador. Ann. Mag. Nat. Hist., ser. 7, 9:51-57. 1905. Descriptions of new reptiles discov- ered in Mexico by Dr. H. Gadow, F.RS. Proc. Zool. Soc. London, 2:245-247. CARPENTER, C. C. 196la. Patterns of social behavior in Merriam’s canyon lizard (Sceloporus m. merriami—tiguanidae). Southwestern Nat., 6:138-148. ——. 196lb. Patterns of social behavior in the desert iguana, Dipsosaurus dorsalis. Copeia, 1961:396-405. Corr, E. D. 1871. Ninth contribution to the herpe- tology of tropical America. Proc. Acad. Nat. Sci. Philadelphia, 1871:200-224. Davis, W. B. 1954. Three new anoles from Mexico. Herpetologica, 10:1-6. Davis, W. B. anv J. R. Dixon. 1961. Reptiles (ex- clusive of snakes) of the Chilpancingo region, Mexico. Proc. Biol. Soc. Washington, 74: 37-56. Daupin, F. M. 1803. Histoire naturelle générale et particuli¢re des reptiles. Vol. 3, pp. 1-452; vol. 4, pp. 1-397. DutLt_MAN, W. E. 1961. The amphibians and rep- tiles of Michoacan, México. Univ. Kansas Publ. Nat. Hist., 15:1-148. DuELrMAN, W. E. 1963. Amphibians and reptiles of the rainforests of southern E] Petén, Guate- mala. Univ. Kansas Publ. Mus. Nat. Hist., 15:205-249. Dunn, E. R. 1940. New and noteworthy herpe- tological material from Panama. Proc. Acad. Nat. Sci. Philadelphia, 92:105-122. Ecuetie, A. A., A. F. ECHELLE anp H. S. Fitcu. 197la. A comparative analysis of aggressive display in nine species of Costa Rican Anolis. Herpetologica, 27:271-288. 1971b. A new anole from Costa Rica. Herpetologica, 27:354-362. EruerwwcE, R. E. 1960. The relationships of the anoles (Reptilia: Sauria: Iguanidae): An in- terpretation based on skeletal morphology. Ph.D. thesis, U. of Michigan, Ann Arbor, 236 pp. Fircu, H. S. 1972. Ecology of Anolis tropidolepis in Costa Rican cloud forest. Herpetologica 28:10-21. 1973a. A field study of Costa Rican lizards. Univ. Kansas Sci. Bull. 50:39-126. 1973b. Population structure and survivor- ship in some Costa Rican lizards. Occas. Pa- pers Univ. Kansas Museum Nat. Hist., 18: 1-41. 128 Tue UNIVERSITY OF KANSAS SCIENCE BULLETIN 1973c. Observations on the population ecol- ogy of the Central American iguanid lizard, Anolis cupreus. Carib. Jour. Sci. 13:215-230. 1975. Sympatry and interrelationships in Costa Rican anoles. Occas. Papers Mus. Nat. Hist., No. 40:1-60. Fitcn, H. S. anp R. W. Henpverson. 1976. The rock anoles of southern Mexico. Jour. Herpe- tology in press. Henperson, R. W. anv H. S. Fitcu. 1975. A com- parative study of the structural and climatic habitats of Anolis sericeus (Reptilia: Iguani- dae) and its syntopic congeners at four locali- ties in southern Mexico. Herpetologica, 39: 459-471. JensseN, T. A. 1971. Display analysis of Anolis nebulosus (Sauria, Iguanidae). Copeia, 1971: 1-34. Lyncu, J. D. anp H. M. SmitrH. 1966. New or un- usual amphibians and reptiles from Oaxaca, Mexico. IJ. Trans. Kansas Acad. Sci., 69: 58-7). Mosaver, W. 1936. The re-discovery of Anolis gadovii. Herpetologica 1:61-63. Myers, C. W. 1971. A new species of green anole (Reptilia, Sauria) from the north coast of Veraguas, Panama. Amer. Mus. Novitates 2470:1-14. O’SHAucHNEssy, A. W. E. 1875. List and revision of the species of Anolidae in the British Mu- seum collection, with description of new spe- cies. Ann. Mag. Nat. Hist., ser. 4, 15:270- 281. Perers, J. A. 1959. Nota miscelaneas sobre saurios del Ecuador. Ciencia y Naturaleza, 2:118-124. Peters, J. A. AND R. Donoso-Barros. 1970. Catalogue of the Neotropical Squamata: Part II. Lizards and Amphisbaenians. Smiths. Inst., U.S. Nat. Mus. Bull. 297, 293 pp. SavacE, J. A. 1973. A preliminary handlist of the herpetofauna of Costa Rica. Univ. Graphics, Los Angeles, 17 pp. SmitH, H. M. 1933. Notes on some Mexican lizards of the genus Anolis, with the description of a new species, A. megapholidotus. ‘Trans. Kansas Acad. Sci., 36:315-320. 1936. A new Anolis from Mexico. Copeia, 1936:19. 1964. A new Anolis from Oaxaca, Mexico. Herpetologica, 20:31-33. SmirH, H. M., F. W. Burtey anp T. H. Frirts. 1968. A new anisolepid Anolis (Reptilia: Lacertilia) from Mexico. Jour. Herpetology, Pela7alpil SmitrH, H. M. anp H. W. Kerstrer. 1955. New and noteworthy Mexican lizards of the genus Anolis. Herpetologica, 11:193-201. SmitrH, H. M. anp R. A. Sprerer. 1945. A new anole from Mexico. Copeia, 1945:165-168. SmitH, H. M. anp E. H. Tayror. 1950. An anno- tated checklist and key to the reptiles of Mex- ico exclusive of the snakes. U.S. Nat. Mus. Bull. 199:1-253. Stamps, J. A. anD G. W. Bartow. 1973. Variation and stereotypy in the displays of Anolis aeneus (Sauria: Iguanidae). Behaviour, 47:67-94. Stuart, L. C. 1942. Comments on several species of Anolis from Guatemala with descriptions of three new forms. Occas. Papers Univ. Michigan, 464:1-10. —. 1948. The amphibians and reptiles of Alta Verapaz, Guatemala. Misc. Pub. Mus. Zool., Univ. Michigan, 69:1-109. 1955. A brief review of the Guatemalan lizards of the genus Anolis. Misc. Pub. Mus. Zool. Univ. Michigan, no. 91, 31 pp. Taytor, E. H. 1956. A review of the lizards of Costa Rica. Univ. Kansas Sci. Bull., 38:1-322. WERNER, F. 1894. Uber einige Novitatem der Herpe- tologischen Sammlung des Wiener Zoolog. vergl. anatom. Instituts. Zoologischer Anzeiger, Jahrg. 17, no. 446:155-157. Witurams, E. E. 1965. South American Anolis (Sauria, Iguanidae): two new species of the punctatus group. Breviora, No. 233:1-15. 1970. South American anoles: Anolis apol- linaris Boulenger 1919, a relative of A. bipor- catus Wiegmann (Sauria, Iguanidae). Bre- viora, No. 358:1-11. 1974. South American Anolis: three new species related to Anolis nigrolineatus and A. dissimilis. Breviora, No. 422:1-15. Wittiams, K. L. anp H. M. Smirn. 1966. Note- worthy lizards of the genus Anolis from Costa Rica. Carib. J. Sci. 6:163-166. | i iii) SE 04553" \, 2, 1976 October 2 Re RS °°. oe, oe sel ae se a oe Fs sot Kod = oe oe 2 * Od iets ws we we % te 6 o8, s, Bs oe i : oe oe RS = te. “. eae S a % OO te ~ ROLF BORCHERT SCIENCE BULLETIN THE UNIVERSITY OF KANSAS COMPUTER-AIDED EVALUATION OF SHOOT-GROWTH PATTERNS IN TREES By Vol. 51, No. 4, pp. 129-140 as . MMe OE ORT T CCE OSEEEIIESSESEEISESSSEESSSSESSSSS SSS ESSENCES: Leteseipteselel alee AAC RRS A NRT EN TOTO E OE EEOC CELL EEL CLE AL SLEPT SIS 130 Tue UNIVERSITY OF KANsAs SCIENCE BULLETIN Quantitative data suitable as a base for modeling the rhythmic shoot-growth should comprise as many variables related to shoot-growth as possible and be ob- tained from a large number of trees. This report deals with an attempt to meet this requirement by measuring relatively few variables per tree in a large number of trees at regular intervals throughout the growing season. Consecutive processing of raw data by the computer involves the following steps: 1. Calculation of derived quantities from the measured variables. 2. Printing of measured and derived data in numerical form. 3. Calculation and printing of mean values for groups of trees growing under identical conditions or sub- ject to the same experimental treatment. 4. Presentation of data for each tree in graphic form. This computer-aided evaluation of growth-measurements facilitates the analy- sis of shoot-growth in hundreds of trees with a degree of precision nearly equal to that achieved by detailed growth-measure- ments of a few trees. ACKNOWLEDGEMENTS Support of this work by the University of Kansas General Research Fund, use of the facilities of the University of Kansas Computation Center and the reliable and competent help of Mrs. Louise Wilson and Miss Elizabeth Vacek in making growth- measurements and preparing the data for computer treatment are gratefully ac- knowledged. Dr. Robert Nunley, Depart- ment of Geography, kindly provided use of the MAPS Computer System for meas- urement of leaf-areas. MATERIALS AND METHODS CoLLECTION OF Data At the beginning of the growing sea- son, the initial state of each tree was de- termined, e.g., length and diameter of the stem, number of branches, weight, etc. Because of existing differences in morphol- ogy and growth-patterns, different genera of trees may require different evaluation procedures and the initial parameters used for the characterization of trees, the vari- ables to be measured at regular intervals and the derived quantities to be calculated may vary. Even where repeated measurements of shoot-growth in a population of trees are not possible, the amount and rhythmicity of shoot-growth is reflected in graphs of data from one measurement, each, for shoot, leaf, and internode length taken at the end of the growing season (compare Fig. fg; 36g). Data given and discussed below include measurements of the following variables made at regular intervals throughout the growing season in 1973 or 1974: a. Length of shoot developed in current season. b. Length of leaves exceeding 20 mm. c. In- ternode length between mature leaves. Leaves were considered mature when weekly increments in length were less than 2 mm. Also, the state of activity of the apical bud—active growth vs. rest—was observed on the date of each measurement. Data ProckssInc At the time of measurement, data were recorded by means of a tape recorder, later being transferred to data sheets containing 80 columns, each line corresponding to one computer card. Data were then punched on cards and evaluated by a pro- gram written in Fortran Y (which is available from the author ). Assuming that leaf growth is allometric, the leaf-area can be calculated from leaf- length by the formula for allometric growth: Apo inv Zee ho 2) ee ee ey JL, where A = leaf-area, L = leaf-length, and CoMPUTER-AIDED EVALUATION OF SHOOT-GROWTH PATTERNS IN [TREES 131 6 and & are constants (Bonner 1963). For example, 60 leaves of greenhouse-grown pin-oak were collected, their lengths in mm were determined with a ruler and their areas were determined with the MAPS (Multi-Dimensional-Analysis-Proc- essing-Spaces) Computer system, devel- oped at the University of Kansas by Dr. R. Nunley. Any other valid method for determination of leaf area can, of course, be used. The correlation coefficient be- tween In L and In A was found to be r = 0.9, the leaf area could be calculated by ln A = 1.93 ° ln L— 6.04 or A = 0.00238 - L1-°8, where L is in mm and A is in cm”. The following comparison between measured and calculated leaf areas illu- strates the quality of the calculations. Leaf Length Leaf Area,meas. Leaf Area, calc. mm cm? cm? 69 9 8.4 92 13.5 14.7 139 32.4 SIS) 145 36.4 33) 166 48.2 45.9 Whereas the formula for leaf-area in pin-oak has been used in the evaluation of data given below, leaf-areas for other spe- cies, as given in Table II, were calculated with the assumed formula 4 = L? - 0.003, because no empirical formulas were avail- able. Ourpurt oF MEASURED AND CALCULATED DaTA All variables not mentioned above as being measured are calculated from the measured data. The totality of measured and calculated data reflects the dynamics of shoot-growth much more clearly than measured data alone can do. Necessarily, the quality of the results depends strongly on the frequency and precision of meas- urements. Weekly measurements done with a ruler are sufficient to determine adequately the shoot-growth patterns and can be done with large samples of trees. Since the dynamics of shoot-growth manifests itself both in variations of many parameters in time and in variations along the axis of the shoot, data for growth must be shown as related to both. Inir1AL Data The record for each tree begins with the following information (Table I): a. Tree-species; 6. Code number for experi- mental treatment; c, Number of tree; d. Initial measurements of tree at beginning of growing season; e. Key for transforma- tion of calendar date into coded date. SHooT AND LEAF GrowTH Data REcoRDED AND PLoTTED FOR EAcH TREE Numeric data of shoot and leaf-growth against time are given in Table I, as printed by the computer. In an additional table (not given here), the output contains the following measured and calculated variables for each leaf: a. Maximum leaf length in mm. 6. Leaf area in cm* as cal- culated from length. c. Internode length in mm. d. Sum of all internodes below leaf to determine position of leaf along stem. e. Highest leaf growth rate achieved in mm/day. f. Date when leaf exceeded 40 mm length. g. Date when leaf reached maturity. 4. Days passed between reach- ing 40 mm and maximum length. 7. Days passed between consecutive leaves reaching 40 mm (plastochron index). A graphic representation of data for shoot and leaf-growth is shown in Fig. 1. For any given tree, the information in this graph equals that in conventional graphic presentations of shoot and_ leaf-growth (compare Greathouse et al. 1972, Critch- field 1960, 1970). Besides the tables and graphs described above, rhythmicity of each tree is numeri- cally characterized by recording, for each growth-flush, the beginning and ending dates of active shoot-growth, the duration of active growth and of the interflush rest- 132 periods and the number of leaves per flush. Other values for the evaluation of shoot- growth are stored during the computation of data and are printed in a final table for each tree, e.g., maximal rates of shoot and leaf-growth, mean leaf-area per shoot, total leaf-area per shoot, etc. MEAN VALUES FoR GRrouPs OF TREES For use in quantitatively comparing different groups of trees, the means of the following variables are tabulated at the end of each experimental group of trees: a. Total shoot length in mm. 6. Maximum shoot growth rate achieved in mm/day. c. Total leaf area per shoot in cm? (d,f). d. Area per leaf in cm?. e. Maximum leaf growth rate achieved in mm/day. f. Num- ber of leaves per shoot. g. Number of leaves per growth flush (fk). 4. Number of leaves in first flush. 7. Number of leaves in second flush (h,g). 7. Days of active shoot growth. k. Number of growth flushes. 7. Days per growth flush (j,k). m. Days in first growth flush. 2. Days in second growth flush (j,m.) 0. Duration of Tasce I. Quercus palustris. Young, single-shoot tree growing in greenhouse. Tue UNIVERSITY OF KANSAS SCIENCE BULLETIN rest periods in days (j,q). p. Duration of first interflush period in days. g. Duration of measurement period in days. r. Num- ber of trees in treatment. s. Vigor of tree as calculated from initial data on stem height and diameter. Several of these measured (a,f,j,k,q) and calculated (4,d,e,p) variables appear to be completely independent from each other. Others obviously depend on some of the independent variables (c,¢,/,0). They have been characterized above by giving in parenthesis the variables they depend upon. Other correlations should become apparent from future multivariate analyses of tree-growth data and should reveal as- pects of the mechanisms which control tree-growth. Data assembled from these tables have been used to prepare Table II. OBSERVATIONS EXAMPLEs OF RECORDED SHoot GrowTH PATTERNS 1. Quercus palustris—Pin-oak. Data given in Table I and Fig. 1 were Data of shoot and leaf growth as a function of time. For explanation of letters see legend Fig. 1. *** PIN OAK -- QUERCUS PALUSTRIS SPEGIES EZ TREATMENT NO. 1 PLANT NO. ] STEM LENGTH, UPPER AND LOWER STEM DIAMETERS IN MM 278. 3. as DATE/APR 1 5 10 15 20 25/MAY1 5 10 15 20 25/JUN1 5 10 15 20 25 JUL1 5 10 15 CODE Wo 8 ih 4) Ao) 4) 31 35 40 45 50 55 62G6 776) Sil86 92 96 101 106 chs UNIUS | ets NO DA SHOOT SHOOT IA NB L40 LMX LEAF LEAVES ie RATE CT AREA (1) “2) 3 40°5 96.0 7 8: -9), 10),.11' Wie aaa Kh asCe ae ata) f ] U.S8— “thal Opes 40 Ak Sig Chilo Wo ESic 2 0) S85 2 2 @ © 66. 83. 94.110.102. 3 / 85. WM I © W © 13% 93, 9911S 105% A Zee soe Ole ee Ae Or a4 735 Wo Oo Ws 0 5 SS Fo So) © 4 8 @ We © © Wo O5.G0o Blo as ko Ghee 6 &) @, @65 2 4 @ © Wee Ws 5 O85 OslOS Moi UIB ole 7 iG Cd. @ JY 4& @ OM Ws We O05 We OsWISoU Uo iitss WAsis lOz/c 8 > Go, O@ I © @ & O06 © @ @ Ws O Os 6 We S) Bie (qe 1 9 TOG5e Bole Gal 0.5 9) 0.4 (05, 0196. 4050: 0. 0. .0. 0: 0s 10> -0.025 us 2920 Salas il 7 88 058 0 © G&S © Be 5 @ WS Ob OF M5 WS Ws Oo BcilWoilss Gs Ese WZ ti) ft. (We i © © @ s& @s @5 06 5 Ws 5 © Oo WolWboloileee WUMs Iss ish 86 685 @ WT i @ 8 sees 5 @2 @5 6 Os @6 Ws We Wo Wo Wo Me Wo & 1 CS so 0. 4 la 100° 88 10) 4 CoMPUTER-AIDED EVALUATION OF SHOOT-GROWTH PATTERNS IN TREES 133 obtained from a 3-year-old, potted, single- shoot tree growing in the greenhouse from March to June 1974. A single look at the graphic presentation of shoot-growth shows that the tree passed through 3 con- secutive flushes of growth during the ob- servation-period, and that virtually every aspect of shoot-growth reflects rhythmic growth (Fig. 1). Curves for shoot-elonga- tion and shoot-growth rate show 3 short periods of rapid growth alternating with longer periods of rest (Fig. lc,e). During rapid shoot-elongation, relatively few leaves per flush begin growth in rapid sequence, almost simultaneously. They rapidly expand and reach maturity at about the same time (Fig. la,b,d). The dynamics of leaf-growth during 3 flushes is given in more detail in Fig. 1f, which shows that the major part of growth of each leaf is completed in less than 2 weeks. This rapid, almost simultaneous expansion of a few leaves per flush is characteristic of the so-called “determinate” pattern of shoot-growth, in which only leaves pre- formed in the bud expand during a growth-flush. Rhythmic growth is also shown by the periodic changes in inter- node-length (Fig. 1g), although the greater length of the first internode during each flush represents in reality the sum of 2 or 3 internodes, namely those between the first foliage leaf and the preceding bud scales, which do not enter into the record because of their reduced size. Although numeric data are less useful for analysis of growth-patterns of individ- ual trees, they are indispensable for quan- titative comparisons between groups of different trees. Data given in Table II were compiled from numeric output for individual trees and treatment-groups and clearly show the wide variability of growth-patterns for pin-oak and_ other species. 2. Tila platyphylla—Basswood. Although Fig. 2 clearly shows that the 3-year-old, potted basswood tree passed through two flushes during the 1973 ob- servation-period, there are significant dif- ferences between the growth-patterns of basswood and the oak tree represented in Fig. 1. In Tilia, periods of active growth are longer than rest-periods, and even dur- ing the rest-period caused by abortion of the shoot-apex and preceding resumption of growth by the uppermost lateral bud, shoot-elongation continues at a low rate (Fig. 2c, Date 70-92). During the first flush, leaves develop and mature over a longer period of time (Date 42-89), growth of individual leaves lasts approxi- mately 3 weeks and there are more leaves per flush than in oak. In contrast to this pattern for young basswood, flushing and leaf-growth of an adult tree were com- pleted within two weeks, showing a growth-pattern similar to that of a mature Oaleitee nies tt. 2h): 3. Populus deltoides—Cottonwood. The 2-year-old potted cottonwood tree shows typical “indeterminate” growth (Fig. 3): Shoot-growth continues for a long time; a large number of leaves ex- pand and mature consecutively. However, even though growth is never arrested, shoot-growth displays a marked rhythmic- ity which is reflected in shoot-growth rate, leaf-size and internode length, all of which pass through minima at about the same time (Date 87, leaves 31-33). Cor- responding fluctuations of growing condi- tions in the greenhouse were not observed. 4. Fraxinus americana—American ash. The 2-year-old potted ash grows slowly, but continuously, for a long period (Fig. 4). Shoot-elongation and number of leaves are low, compared to cottonwood (Fig. 3), but—because of the large leaves— leaf area increases rapidly. Data for leaf- growth reflect the decussate position of leaves. The development of consecutive pairs of leaves is similar. In ash, as in 134 Tue UNIversITY oF KANsAs SCIENCE BULLETIN Tasie II. Means of selected variables characterizing shoot growth in experimental groups of four tree species under different conditions. A—Potted sapling, one growing shoot, outside greenhouse B—Potted sapling, one growing shoot, inside greenhouse C—Lateral branch of approximately 15 m high tree (basswood, as or 5 m high tree (cottonwood) 4 4 E = Ee E = 2 nN g a = w 80 peepee wh oo g pela = aI = g cc Sg 3 g g 2 Be = =< SPECIES = = ai a 3 3 3 iS te Tilia platyphylla A 359 12 591 50 44 33 10 1.3 (basswood) B 343 11 816 73 38 27 9 1.4 Cc 145 7 374 56 1 15 9 1.0 Quercus palustris A 188 19 453 25 19 13 10 1S) (oak) B 294 24 938 40 33 14 13 2.4 C 127 11 493 45 19 17 11 ‘lait Fraxinus americana A 112 7 706 34 82 82 24 1.0 (ash) B 234 19 1262 67 70 70 21 1.0 Populus deltoides B 726 39 2458 63 81 81 44 1.0 (cottonwood) G 351 25 1328 54 79 79 24 1.0 Fic. 1. Quercus palustris, same tree as in Table I. Shoot and leaf growth data as a function of time (a-e) and as a function of leaf number (f,g). Lines have been drawn by hand. Explanation of letters: a—number of leaves having exceeded 40 mm length since last measurement date b—number of leaves having reached maturity since last measurement date c—length of current season’s shoot in mm d—total leaf area of shoot as calculated from leaf length in cm? e—shoot growth rate in mm/day f—leaf growth. Total height of each bar gives length of leaf at maturity in mm. Numbers forming bars indicate date (j) at which leaf has reached certain size. For example, leaf # 5 (h) had 60 mm length at date 5(k) = day 31 (j), 105 mm at date 6 = day 39, and 115 mm at date 7 = day 46. Letters A, B, C indicate dates 10, 11, 12. g—internode length below respective leaf in mm. Scale is 2X that of leaf length (1 space = 1 mm) h—leaf number beginning with oldest leaf i—activity of apical meristem. O0—active growth; 1,2—arrested growth j—coded measurement date (days after April 1) k—current number of measurement dates m—total number of mature leaves f’—leaf growth of mature tree in spring 1973 135 HOOT-GROWTH PATTERNS IN TREES S CoMPUTER-AIDED EVALUATION Of povpbrrryrbrryrryEeeECCEEEEEEEEEE CE EEE EEE EEEE EEE CZ C2eceecececeececcce ce LLL LLL LL prbbrpyyPVDDDEEEEEEEEEECECEEEEEEEEECEEE*S2CC eee eee ee cece eee ee eee eceeceeeeC CLL LLU L LALA LLL pobvrbbbbbyryrrrrYPPHPee EEE ELE CECE EE EE EE CE EE EE EE EEE EEE EEEEEEEEE EEE EEE CC Ccccceccccececccccceeeee ccc CCC CCCLLLLLLLLLLLLLLLL pobobbbbrpbrbrrbbrrPrPyyece CECE EE EEEE CE EEEE CE CECE CECE EEEEEEEE CC Ce cececeeeeeeeeccceecececcc ccc eCCCCCCCLLLLLLLLLLLLLLLLLLL brppVVLPPPEEEEEEEEEEEEECE CECEEEEEEEEECCCCee ccc eee eee ceeeecceeeeceec eee eC CCCLLLLLL LLL LLL LLLLL eee ——— el HLONI1 sV3a1 WW ObL cxxWd OL vl vl El ZL I LL I LLELLLLLLLLLELULLLLLLLL =p i LLLELELELULLLLELLLELLLELLLLLL = € LLULELLLLELELLLLLLLLLL HY 2 a LUULELELLLLLULEULELL _ HLONTT dV37 WW ObL zt O0L 08 09 Ov 02 0 4va1 S3AV31 4O HLMOUD ¥* 2°ON HdVU9 4% Zx*Wd OL tl rat OL 8 9 b 2 ) ($) wauv 4v37 WH 008 004 009 005 00b 00€ 002 00L 0 (X) HLON37 LOOKS OY able) a * 0000 0 9€L 2 * 000000 0 S2l LL : 00 IIIIII 0 9LL OL Se LOOOMMELTIDIP TONOUUG ie OOO SIT OMlOlee : 00 1 0S6 Z 000000000 1 028 9 * _ QOOIIIIIII 008 §G * ©000 IIII 099 » —_ * QOOIIIIIII O€S € SS Cen, Meola cr x ° IIIIIIIII Olb t * Sey L 0 ~~ (*)3LVY HLMOYD LOOHS -- ($)V¥3¥Y 4VI7 -- (X)HLONIT LOOHS -- (0)XWW1 -- (1)Ob1 -- LOWI AVG ON «x L "ON HdVuS ¥% ANSAS SCIENCE BULLETIN r THe UNIVERSITY OF 138 VVV666666666666666666666666666666666666666666666666666666 A VVVVVVVVVVVVVVVVVVVVVVVVV666666666666666666666666666666666666666666666666666666666666 L2 VVVVVVWVVVVVVVVVVVV'66666666666666666666666666666666666666666666666666666666666622/L2/LLLLLLLLLL LLL LLL LLL LLL LLL LLL LLL 02 VVVVVVVV66666666666666666666666666666666666666666666666666666666666666622L2LLLLLLLLLLLLLLLL LLL LLL LLL LL 6L LLLLLLLLLLLLL LLL LLL LLL LL kkk kk kkk Lk eke £££ £L£L£L2L219999999999999999999999999999999999999999999999999955555555555555 BL LELLLLLLLLLLL LLL LLL LLL LLL kkk kk Lh cL k££££L£2£LLL2L199999999999999999999999999999999999999999999999995S55555555555555555¢ HA) L99999SSSSSSSSSSSS5SS55S55SSSSS5SSSSSSSSSSSSSbb bbb bby bbb bby bbb bbb yb bboy bby bbb bby bby bry bbybbbY oI LLLLL99999955S55SSSS55SSSS5S5S5SSSS9SSSSSSSSSSSpbbbb bbb bby bbbb byob bbbbb bbb bbb bby by bDbbbbbbbyPY silt SSSSSSSSSSSSSSbbbb bbb bbb bbb ybbbb bb bbb bboy bbb bbb bbb bbb bbb bbb bbb bbb bb bbb Pb bVVbyvEEceeEECEEECEEEEEEEEEE val SSSSSSSSSSpbbbbbb bby bbi bbb bbbybb yb bybb bbb bbbb byob bbb bbbbb bbb bbb bbbbbb bby bb bby bby bby bY bbYVEEEEEEEECEEEEEEEEEEEEE Sil byppypvorvborvbVobprvVOvVL DV DHVEEEEE ECE EEEE EE EEE ELE CEE EEE EEE ECE CE CEE EEE CECE EE ECE EEE CECE EE EEE EEE CEC EC CCCcc ccc cc ccccccccce Al pobppbvrbrvbobvDbVVbL VOD UDVDDPVECEEEECEEEEEEE ECE EEE EE EEE EEE EEE EEE EEE EEE EEE EEE EEE EEE E EEE EEE EEE LEE EEE E CCC ccc cecc ccc cc cee LL CCCCCEEEECCeeceeccececeCCCCCCCCLLLALLULLLLLLLLLLL ELLE LL LLL LLL LULU LLL ELLE Liebe bbe eee tke eee OL CEECEEEE CEE CECE ccceccceececceccee coe COCCCCCOCCOCLLLLLLLLLL LLL LLL LLL LLLLLLULELL ELL Lb bebe eke b ke eke eee eee GAAS LET ETETE UTA CHAS RTETUIGTEUCTEURT RU VET TE CUETRTETRTVES ETI EY VTS} RYT EYB] GREY EYE) YE) 7) I) BY Ey) BTA EUV) VAT EVI STITT EYE) RT OU VV AT INE ETT) EUV} QT AVY BT} SOCLLLLLULELLLELLLLELLUL CL ULE ELLE LELLLCLULELLLLLLL EE LUCUCLL UCL LOLCLUL EEE L LCCC LL LLL Le Le kee ee CLELLLLELLLLLLLELELLELELLELELELELLLLELLLLLUL LLL LL LELEL LLL LLLLULELULULE LLL EAN ICM IETRT OE VTETEVRVATETINTOVRICVEVRIEVRIGTEVAVETANUTAVRVEYRIIVRCRTEVAVLTOVEUC) Gy IEICTAY EVV TGULIAVEVRTCUICILTGIGVAY VG TDIGNEVVG)RTEV AD BYETV GV AVES EI GUMTERETENK) AVR TR| 4) COCCLLLLLELLLLELLLLLLLLULEULELELULELULLLLELLLLULLLLLLLL LELLELECL ELE ECE LE ELLE ELEC ELE ELLE LLL EL LULL LEER LLL LEL ULL LELELELLLE ge ETT VATED AUT EYE] ATL hose seen aog HLONS1 4V31 WW OVL O2L O0L 08 09 Ov 02 0 4Vv3i1 S3AV31 4O HLMOUD x¥ 2 “ON HdVHD +x Zx*W0 OL vl ZL OL 8 9 17 2 0 ($) wauy Jv31 WN 008 002 009 005 00% 00€ 002 OOL 0 (X) HLON31 ue == * x ; L 9LL OL * X : 00 II 096 6 * X é 00 006 8 —— | x Il 08 £ —— ue X * 000 i to te 2) ener ae ete) ie) 1 0 02 4 =e ae 4) —— P00 Tt ORES: Y ee na XW " 00000 gill OPS € _O- $a X "000 I] 08b 2 * ‘ ee ee : Wt Thi (0) 27 al $x Of 0 * -- (x)3LVY HLMOYD LOOHS -- ($)V3uW 4V31 -- (X)HLONST LOOHS --(0)XWW1 -- (I)Ob1 -- LOWI AWO ON xx% L “ON HdVY9D xx ‘ydeis oy) ut Joy popraoid yySus] WNUTXeU 94) paps29xo Apysys T/L] pue pI-[] seavay ‘astmoyty “ydesrs oy) ur paptaoid asue1 ay) pape2dx9 kore Jeo] ‘/ a}ep WW ‘[ ‘Sly ur se sronay ‘asnoyuso1s ul Sutmois 90 poyod ‘sunoX ‘puvriauiy snuixvsy “b “Ol CoMPUTER-AIDED EVALUATION OF SHOOT-GROWTH PATTERNS IN TREES 139 basswood and cottonwood, the first set of leaves (1-10), which were enclosed in the resting bud, develop quickly at the be- ginning of the flush. Thereafter, leaves develop at longer intervals and expand more slowly. Adult ash (not shown here) displays one short annual flush of shoot- growth similar to those of oak and bass- wood (Fig. 1f’ and 2f’). CoMPARISON BETWEEN DIFFERENT Groups OF TREES As growth-patterns of individual trees growing under identical conditions are variable and the time-course of flushing is not synchronous within a group, an “ay- erage shoot-growth pattern” for all trees within one sampled group cannot be calcu- lated. However, if the means of several representative variables are calculated and tabulated, then characteristic, quantitative differences between groups are obtained, useful for statistically treating the data. VARIATION WITHIN SPECIES Potted trees of three species (basswood, oak, ash) were grown both outside and inside the greenhouse, the latter growing in environmental conditions characterized by higher average temperature and higher relative humidity. For all three species, overall growth was higher inside the greenhouse, with leaf-size being the single variable showing the greatest increase (Table H, Treatments A and B). A comparison between seasonal growth of lateral shoots of large trees and green- house-grown single-shoot saplings shows significant quantitative differences as well as differences in shoot-growth pattern. Whereas almost all adult trees passed through only one growth flush, young oak and basswood trees usually flushed re- peatedly (Table I, Treatments B and C). If normal intraspecific variability of shoot-growth patterns is as marked as shown in these few examples, even more dramatic differences should be obtained by experimental manipulation of trees. The analysis of such variation should lead to a better understanding of the mechanisms controlling shoot-growth. Variations in shoot-growth of oak as reported in the literature and observed in my experiments have been previously described, interpreted and related to predictions resulting from the model for rhythmic shoot-growth in frees (Borchert, 1973, 91975): VARIATION BETWEEN SPECIES Interspecific differences in shoot-growth patterns, as shown above for single trees (Fig. 1-4), are equally obvious in the com- parison of quantitative data given in Table II. DISCUSSION The analysis of shoot-growth data for broad-leaved trees described above offers the following advantages and possibilities: 1. Shoot-growth data for relatively large numbers of trees can be collected and prepared for computer analysis by personnel with little training in tree- morphology and punching of data cards. 2. Data are automatically recorded in numeric and graphic form and can be grouped and rearranged in any desired format. 3. Information contained in the meas- ured raw data is increased as follows: (3.1.) Meaningful subsets of data which are present in the raw data, but not recog- nized as such, can be extracted and com- piled from the raw data, e.g., data on flush- growth and on the time-course of growth in single leaves (Table I and Fig. 1-3). (3.2.) Derived variables, which often mean more than the original ones, can be calculated. For instance, the dynamics of shoot-growth is more clearly reflected in changes of shoot-growth rate over time than in increasing shoot-length. Likewise, leaf-area as calculated from leaf-length is important in determining some princi- pal physiological processes controlling 140 Tue UNIVERSITY OF KANSAS SCIENCE BULLETIN shoot-growth, especially transpiration and photosynthesis. 4. The automatic representation of data in graphic form is an essential prerequisite for evaluating and comparing — shoot- growth patterns in many separate trees by visual scanning. Despite progress in com- puter technology, the human eye and brain are still the most efficient instru- ments to evaluate complex patterns. The experienced observer will at a single glance recognize certain growth-patterns in the graphed data; he will detect similari- ties and discrepancies between trees of different groups and see correlations be- tween or trends in certain variables. This information cannot be obtained as effi- ciently by observing living plants or by studying numeric growth-measurements. Once detected by visual scanning, such correlations can be analyzed in greater detail in the numeric data and “indicator” variables can be singled out which show differences in growth-patterns between various groups of trees, whose means can be compared statistically, e.g., by multi- variate analysis. 5. When physiological measurements have been performed, their results can be treated together with morphologic obser- vations and the combined data used to construct and improve the model for shoot-growth. Predictions from the model may then be compared with measured data, i.e., there will be continuous inter- action between observation, description, and interpretation of shoot-growth data. LITERATURE, Cites Bonner, J. T. 1963. Morphogenesis. Atheneum, New York. BorcHert, R. 1973. Simulation of rhythmic tree growth under constant conditions. Physiol. Plant. 29: 173-180. BorcHert, R. 1975. Endogenous shoot growth rhythms and indeterminate shoot growth in oak. Physiol. Plant. 35: 152-157. CrircHFieLD, W. B. 1960. Leaf dimorphism in Populus trichocarpa. Amer. J. Bot. 47:699- VANE CriTcHFIELD, W. B. 1970. Shoot growth and het- erophylly in Gingko biloba. Bot. Gaz. 131: 150-162. CrircHFiELD, W. B. 1971. Shoot growth and _ het- erophylly in Acer. J. Arnold Arbor. 52: 240- 266. GreATHousE, D. C., W. M. LAEtscH, AND B. O. PHIN- Ney. 1971. The shoot-growth rhythm of a tropical tree, Theobroma cacao. Amer. J. Bot. 58: 281-286. Harré, F., anp R. Martin. 1968. Etude de la croissance rhythmique chez l’Hévéa (Hevea brasiliensis). Adansonia 8: 475-803. Koztowsk!, T. T. 1971. Growth and development of trees. Vol. I. Academic Press, New York. Tusss, F. R. 1973. Research fields in the interaction of rootstocks and scions in woody perennials. Horticul. Abstr. 43: 247-253, 325-335. | Ht {i || | 1 WATT HL , 1977 J 22 @ s oo are retatetebeeteeceebeeteeneececeeats April 14 2, LAS eetetetntes oe Sse tate THE UNIVERSITY OF KANSAS SCIENCE BULLETIN 2, 505 0 o.9.2 ©. Sess 2.2 8 @. sss BOREIDAE (MECOPTERA) By NORMAN D. PENNY A SYSTEMATIC STUDY OF THE FAMILY 2. o8 ste oe, oat mS : > sot “I (Ste ate oe ae = 2S . + eee > ; = = SOR RRR I RR RRR a SS ecietececeronenee RR NN NC RNR RRS oR setesedoneZoneceneneatatetatatctatetatecatatstatetatetatetstenssesccatecetenteeasesoatstav oreo” oeeere metetetateetetee detec eet teceeteetebene eset et eecee teeta anenatnanceneeane atest cease seen aaeseatec eneceseteeteteteteteteteteketetetetetesetececeseteeeeeseeceseseceesssterasas re: roceseecesestetatstatstetotetscoeetoecetetatetsterateatatet stone statetetetetenetatenenstenetsteneestensnerecerereutrermcenes erecestceestatetatatetatatetatetetete ANNOUNCEMENT The University of Kansas Science Bulletin (continuation of the Kansas Uns- versity Quarterly) is an outlet for scholarly scientific investigations carried out at the University of Kansas or by University faculty and students. Since its incep- tion, volumes of the Bulletin have been variously issued as single bound volumes, as two or three multi-paper parts or as series of individual papers. Issuance is at irregular intervals, with each volume prior to volume 50 approximately 1000 pages in length. Effective with volume 50, page size has been enlarged, reducing the length of each volume to about 750 pages. The supply of all volumes of the Kansas University Quarterly is now ex- hausted. However, most volumes of the University of Kansas Science Bulletin are still available and are offered, in exchange for similar publications, to learned societies, colleges and universities and other institutions, or may be purchased at $20.00 per volume. Where some of these volumes were issued in parts, individual parts are priced at the rate of 2 cents per page. Current policy, initiated with volume 46, is to issue individual papers as published. Such separata may be purchased individually at the rate of 3 cents per page, with a minimum charge of $1.00 per separate. Subscriptions for forthcoming volumes may be entered at the rate of $20.00 per volume. All communications regarding exchanges, sales and subscriptions should be addressed to the ExcHaNce LisrariaNn, UNIVERSITY oF Kansas Lipraries, LAWRENCE, Kansas 66045. Reprints of individual papers for personal use by investigators are available gratis for most recent and many older issues of the Bulletin. Such requests should be directed to the author. The International Standard Serial Number of this publication is US ISSN 0022-8850. Editor Eugene C. Bovee Ed:storial Board William L. Bloom Philip W. Hedrick Rudolf Jander Harvey Lillywhite Charles D. Michener Norman A. Slade Henry D. Stone George W. Byers, Chairman THE UNIVERSITY OF KANSAS SCIENCE BULLETIN Vol. 51, No. 5, pp. 141-217 April 14, 1977 A SYSTEMATIC STUDY OF THE FAMILY BOREIDAE (MECOPTERA) by NorMawn D. Penny’ TABLE OF CONTENTS ASTRO? Lo ee ee ee 2 ee eee a eee Ey ee pee ee Sens 141 “PTET BRE STEIOINY, et Ee Er Se Ae pe On Tee Oe 141 PUEEIODSEOESERESERVATION OF SPECIMENS. «2:.<-<::-2<<-0. Ficures 31-33, male, tip of abdomen, lateral view. Fig. 31, Boreus coloradensis Byers. Fig. 32, Boreus reductus Carpenter. Fig. 33, Boreus vlasovi Martynova. All 78x. 165 166 Tue UNIversITy oF KANsAs ScIENCE BULLETIN Ficures 34-39, tip of abdomen, dorso-caudal view. Fig. 34, Hesperoboreus brevicaudus (Byers). Fig. 35, Hesperoboreus notoperates (Cooper). Fig. 36, Boreus borealis Banks. Fig. 37, Boreus brumalis Fitch. Fig. 38, Boreus californicus Packard. Fig. 39, Boreus coloradensis Byers. All 66%. was constructed by weighting each char- acter state in the unweighted phylogeny according to the number of species in- volved (Ashlock, unpublished method). This would give more value to those char- acter states which help to unify groups of species. The final horizontal distance be- tween species gives a measure of how closely related the species are. Character states on the weighted phylogenies are listed by branching sequence in Table 2, p 169: The third phylogeny (Figure 68, p. 181) illustrates a weighted phylogeny which A Systematic Srupy oF THE FAMiLy Borerpar (MecoprTErRA) 167 Ficures 40-45, tip of male abdomen, dorso-caudal view. Fig. 40, Boreus elegans Carpenter. Fig. 41, Boreus hyemalis (L.). Fig. 42, Boreus intermedius Carp. Fig. 43, Boreus lokayi Klp. Fig. 44, Boreus mivoriundus Fitch. Fig. 45, Boreus nix Carpenter. All 66X. 168 Tue UNIvERSITY OF KANsAs ScIENCE BULLETIN uses not only the number of species in- volved but also the number of times a character state appears (Brothers, 1975). By assignment of only half value to a character state which has evolved twice, these character states lose much of their value. This is done because we are trying to determine the most probable line of evo- lution, and character states which appear to have evolved twice are more doubtfully correct and should be given less weight. By reducing their value such lines of evo- lution which use these character states appear to have diverged less. Figure 68 agrees with my own intuition as to the evolution of this group. In all three phylogenies, the two genera, Boreus and Hesperoboreus, appear quite distinct, as do most of the groups and sub- groups mentioned earlier. However, it will be noted that in the two weighted phylogenetic diagrams the v/asovi sub- group and hyemalis group do not appear to be very distinct, while the nivoriundus group is very distinct. This is due in part to the material available. All nivoriundus group species are represented, while kra- tochvili is lacking from the /Ayemalis group, and the only one vlasovi subgroup species is represented. Only one synapo- morphous character (Hennig, 1965) sepa- rates the nivortundus subgroup from the rest of Boreus, but because of the primitive position of this subgroup in the phylogeny and the large number of species involved, the two subgroups of the nivoriundus groups appear quite distinct from each other. Boreids evolved in two directions. One group, given the new generic name Hes- peroboreus, and containing H. brevicaudus and notoperates, lives in areas of rather warm climate along the west coast of the United States. These species possess two ocelli; large bristles are present along the anterior and posterior margins of the pro- notum; the eighth tergum and sternum are not fused in males; females possess a very short ovipositor with cerci incom- pletely fused; and no ninth tergal hood development or outer forewing spines are found in males. All species of Boreus have outer fore- wing spines in the males, and a long ovi- positor and fused cerci among females. These species can in turn be divided into one Nearctic group, two Holarctic groups, and one Palearctic group. The first group (nivoriundus group), containing B. beybienkot, brumalis, ele- gans, nivoriundus, nix, and pilosus, con- sists of both North American and Asiatic species. ‘These species possess pronotal bristles, and in males have unfused eighth and ninth abdominal terga and _sterna, and small tergal hoods. This group can be further divided into the nivoriundus subgroup, having males with entire ninth sterna and gradually tapering forewings, and the dramalis subgroup, having males with deeply notched ninth sterna and wings abruptly narrowed at mid-length. The second group (reductus group), containing B. reductus, vlasovt, and prob- ably most Asiatic boreids, is found both in western North America and Asia. These species agree in possessing no pronotal bristles, unfused male eighth tergum and sternum, and a small tergal hood in males. This group also can be subdivided into two subgroups. The reductus subgroup has unfused ninth tergum and sternum in males, while the v/asovz subgroup usually has fused male ninth tergum and sternum. Most Asiatic species belong in this second subgroup. B. orientalis may be an exception among Asiatic species. Martynova (1954) has il- lustrated part of the eighth abdominal seg- ment of this species, showing it as fused. This would be the only known Asiatic species with fusion in both eighth and ninth segments. Until more is known about this species, I feel that it should be considered as a part of the vlasovi sub- group of the reductus group. Other Asi- aaa aaa aaa A SS A Systematic Stupy oF THE FAmity Boremae (Mecoprera) 169 atic species which Martynova illustrated with fused ninth segment are chadzhi- girejl, vlasovt, and semenovi. Cooper (1973) stated that the ninth tergum and sternum of vlasovi were unfused, but this varies with the individual. Since navasi is very similar to chadzhigirepi, it prob- ably belongs in this subgroup. Although little is known about sjoestedti, perhaps it too belong to the vlasovi subgroup. To know with certainty, it will be necessary to collect males of these species. The third group (californicus group) includes B. borealis, californicus, colora- densts, and intermedius. These species all possess no pronotal bristles, show fusion of eighth terga and sterna among males but no fusion of male ninth terga and sterna, TABLE, 2 PHYLOGENETIC CHARACTER STATE SEQUENCE. Between points 1 and 2 abdominal segment-female 34. Development of eighth sternal notch-female Between point 2 and H. brevicaudus 14. Gain of bristles at caudal margin of wing pad-female Between point 2 and H. notoperates 29. Development of notched ninth sternum-male 35. Expansion of eighth sternal notch-female Between points 1 and 3 19. Development of small tergal hood-male 30. Lengthening of tenth abdominal segment- female 32. Complete fusion of cerci-female Between points 3 and 4 7. Gain of rostral bristles 11. Broadening of forewings at mid-length-male Between point 4 and B. elegans 8. Gain of 25 rostral bristles Between point 4 and B. nivoriundus none Between points 3 and 5 20. Tergal hood expanded to middle of denticu- lar areas-male Between points 5 and 6 29. Development of notched ninth sternum-male Between point 6 and B. brumalis none Between points 6 and 7 9. Gain of long pilosity Between point 7 and B. nix none Between point 7 and B. pilosus Loss of some denticles Between points 5 and 8 10. Loss of thoracic bristles Between points 8 and 9 1. Loss of median ocellus 1. Loss of median ocellus 2. Gain of rugulose occiput 4. Loss of antennal segments 5. Movement of antennal bases to below eyes 6. Widening of hypostomal bridge 12. Loss of outer forewing spines-male 31. Development of posterior process on tenth 4. Loss of antennal segments 22. Development of projections along lip of hood Between point 9 and B. reductus 13. Forewings reduced until they no _ longer cover hindwings-female 23. Development of medial projections along lip of hood-male Between point 9 and B. vlasovi 28. Fusion of ninth tergum and sternum-male 33. Gain of abruptly narrowed cerci at mid- length-female Between points 8 and 10 18. Fusion of eighth tergum and sternum-male 21. Tergal hood expanded to lateral margin of denticular areas-male 24. Gain of median septum to tergal hood-male Between points 10 and 11 none Between point 11 and B. borealis 27. Development of numerous setae covering interior of tergal hood-male Between point 11 and B. californicus 25. Development of narrow median septum Between 11 and B. coloradensts none Between point 11 and B. intermedius 26. Gain of short median septum with ventral fork-male Between points 10 and 12 15. Gain of transverse ridge on second abdominal tergum-male 16. Gain of transverse ridge on third abdominal tergum-male Between point 12 and B. hyemalis 3. Loss of occipital reticulations Between points 12 and 13 none Between point 13 and B. westwoodt none Between point 13 and B. lokayi 17. Transverse ridge of third segment reduced to a tubercle-male 26. Gain of short median septum with ventral fork-male 170 Tue Universiry oF KANsAs ScrENCE BULLETIN Ficures 46-49, tip of abdomen, dorso-caudal view. Fig. 46, Boreus pilosus Carp. Fig. 47, Boreus reductus Carpenter. Fig. 48, Boreus vlasovi Martynova. Fig. 49, Boreus westwoodi Hagen. All 66X. and have a large male tergal hood with median septum. All species of this group are Nearctic. In the Palearctic region a fourth group of species occurs, the hyemalis group, con- sisting of B. hyemalis, kratochvili, lokayi, and westwoodi. ‘These are similar to the californicus group but have transverse ridges on the second and third abdominal terga. The picture that emerges from the morphology and distribution of these spe- cies is the following: ancestral boreids de- veloped in the Nearctic region from a panorpodid-like ancestor, with flightless females, prothoracic bristles, and unfused eighth and ninth abdominal segments in the males. As is the Hesperoboreus spe- cies, females had incompletely fused cerci, and no elaboration had occurred on the male ninth abdominal tergum. This an- cestral stock then split. H. notoperates and brevicaudus represent the remnant of an ancestral species which further split into A Systematic Stupy oF THE FAMiLy Borerar (MeEcopTERA) 171 two populations as glaciers retreated after the ice ages. One population (notoperates) remained in southern California and sur- vived the warming trend by becoming adapted to warmer temperatures and mov- ing to the higher elevations of Mount San Jacinto (and perhaps associated ranges). The other population (brevicaudus) re- mained at a lower elevation and moved or remained northward in western Oregon and Washington. In the other ancestral stock, i.e., exclu- sive of Hesperoboreus, the female cerci fused entirely and the tenth abdominal seg- ment became more elongate. This stage of development is exemplified by the nz- voriundus group. This group is tranconti- nental in distribution in North America and is also found in the Ala-Too Range, north of the Himalayas. Although no specimens are now known from the Rocky Mountains of Colorado to southern IIli- nois, it seems likely that at one time en- vironmental conditions permitted an east- west movement of Boreus across northern North America. Perhaps in the cooler re- gions of northern Canada this distribution is still contiguous, although there are no specimens to verify this. The presence of this group in Asia indicates a possible early movement into the Palearctic region. The nivoriundus group is here subdi- vided into the nivoriundus subgroup (ele- gans and nivoriundus) and brumalis sub- group (drumalis, nix, and pilosus). B. beybienkot is probably in this group, since the original description mentions protho- racic bristles. However, I have not seen specimens of this species and its position within the nivortundus group is still not known. The next step in evolution of the Borei- dae was loss of thoracic bristles, as seen in the reductus group. The only member of this group in the Nearctic region is re- ductus, distributed in the arid mountain regions of western North America. There are some character states in this group that cause problems for phylogeneti- cists. Species of the reductus group have no median ocellus, in which character they resemble Hesperoboreus. The vlasovi sub- group has the shortest ovipositor found in any group or subgroup of Boreus; and the number of antennal segments is reduced to 19 or 20, as in some specimens of Hesperoboreus. However, these charac- ter states probably were independently achieved through reduction, because males of the reductus group have a well-devel- oped ninth tergal hood, a relatively elab- orate structure which seems unlikely to have evolved twice. There are other indi- cations, as well, of the relatively advanced state of evolution of the reductus group, such as the absence of thoracic bristles in both sexes and reduced number of hind- wing spines in males. Fusion of the eighth abdominal tergum and sternum in males of a population of an ancient reductus group species later oc- curred. This condition created the ances- tral species of the californicus group, which later disepersed in or into the Nearctic region. B. borealis became isolated on St. Paul Island, probably at a time when land connections existed, or possibly when per- manent ice covered the Bering Sea. The closely related species californicus and col- oradensis became widely distributed in the relatively arid mountain regions of west- ern North America from California to central Colorado and from central Arizona to northwestern British Columbia. B. in- termedius is a more northern counterpart of these western species in Alaska. In Europe, the Ayemalis group arose from species in which there evolved an enlargement of the male second and third abdominal terga into flat ridges. These ancestors dispersed into the Carpathian Mountains and western Europe. Martynova (1954) stated that B. chad- zhigireyt had tergal ridges but that they were almost absent. The figures accom- panying her description showed no fusion 172 Tue UNrversiry oF Kansas ScrENCE BULLETIN 51 50 : Q 53 54 | 57 ( 58 Ficures 50-52, male, ninth sternum, ventral view. Fig. 50, Boreus coloradensis Byers. Fig. 51, Boreus reductus Carp. Fig. 52, Boreus brumalis Fitch. Figs. 53-58, male, dististyle. Figs. 53-54, Boreus colora- densis Byers. Figs. 55,56, Boreus elegans Carp. Figs. 57,58, Hesperoboreus brevicaudus (Byers). Figs. 53,55, and 57, caudal view. Figs. 54,56, and 58, lateral view. All 88x. A Systematic Stupy oF THE FAmity BorerpaEe (MEcopTeERA) 173 60 Ficures 59-62, female ovipositors, lateral view. Fig. 59, Boreus coloradensis Byers, 50. Fig. 60, Boreus vlasout Martynova, 57X. Fig. 61, Hesperoboreus brevicaudus (Byers), 57X. Fig. 62, Hesperoboreus notoperates (Cooper), 57X. of the eighth tergum and sternum, but fusion in the ninth segment. These two character states indicate that this species is most likely a member of the v/asovi sub- group of the reductus group, and has no tergal ridges (cf. Cooper, 1972:276- hy: Both the nivoriundus and reductus groups are Holarctic, and little evidence is available to indicate the direction of their movement between continents. However, there may well have been multiple move- ments between the Palearctic and Nearctic regions. ECOLOGY A. PLant Assocrations—Boreidae have been taken from many mosses. Boreus brumalis larvae from eastern North Amer- 174 Tue Universiry oF KANsAs SCIENCE BULLETIN ica has been collected from Dicranella heteromalla and Atrichum angustatum (Webb et al., 1975), and Byers (1954) men- tioned that J. W. Leonard had found adults of this species in Sphagnum. Hes- peroboreus notoperates has been recorded from Grimmia apocarpa, G. laevigata, G. montana, Rhacomitrium sudeticum, Or- thotrichum rupestre, Tortula princeps, T. ruralis, Homalothectum aeneum, H. ne- vadense, H. pinnatifidum, and Campo- thecium amesiae in southern California (Cooper, 1974). Boreus reductus larvae and adults have been collected on club 64 mosses (Selaginella sp.) in southeastern Washington. It appears that the species of moss is of less importance than the moss texture. Mosses which grow in low, com- pact cushions and whose rhizoids are tightly matted more frequently contain adult and larval Boreidae than do mosses that grow upright or are more loosely matted. It has been my experience that upright or loosely matted mosses more frequently contain predaceous carabid beetle larvae. Perhaps, larval boreids sur- vive predation better in the more tightly compacted mosses. Ficures 63-65, female ovipositors, dorsal view. Fig. 63, Hesperoboreus brevicaudus (Byers). Fig. 64, Boreus coloradensis Byers. Fig. 65, Boreus vlasovi Martynova. All 66. A Systematic Stupy or THE Famity Boremar (MeEcoprerA) 175 ' e B H : N e e | ? i Cc Dp E le UY 4 16 © Qu 7 Boose! Kee : Mest fe eee eae ? ‘ Via ob on Uo ORC A UKs as an & 06 na ey et 1 i DF I ' ' ms V1 Ug ite ee ree. \l ' wi vi ter in) 1 ja } ' ' 1 ' ' ' -_e--- —-——<_—-— - lcm Ficure 66. Three dimensional unweighted phylogeny. Movement between plates in any direction of 1 cm indicates one character state change. (Direction of change in position between plates has no biological significance.) A is H. notoperates, B is H. brevicaudus, C is B. elegans, D is B. nivoriundus, E 1s B. brumalis, F is B. nix, G is B. pilosus, H is B. reductus, 1 is B. vlasovi, J is B. coloradensis, K is B. borealis, L is B. californicus, M is B. intermedius, N is B. lokayi, O is B. westwoodi, and P is B. hyemalis. B, Feeprnc—Withycombe (1922) men- Boreus is moss. Other authors (Brauer, tioned keeping one female of Boreus hye- 1855, and Steiner, 1937) have speculated malis alive for 37 days on juices of crushed that Boreus feeds on collembolans, al- flies, although in a later paper (Withy- though no evidence has been presented to combe, 1926) he states that the diet of support this. The European B. hyemalis 176 Tue Universiry of KAnsAs ScrENCE BULLETIN TABLES ScALED SIMILARITY MATRIX. 2 8 2 3 S = = fo fe s & § tS: yl See s = Ss SS Ss & <3 = 4-8 © § = 5.2. Tot ge ee Se RR Se Me Re H. brevicaudus Ses a6) 1S) 7 1S 18 8 225 7 7G eee | H. notoperates 08 2. I2 iS a8 ae} gi) ie) ii 23 8 16 15> emeelseeee | B. borealis AAS oN eee Sime el ( 4 2 GS 9 SP isle lel 4 B. brumalis 43 40 22 8 7 6 10 8 10 5 1 2 yi Jae NO B. californicus PAO ile 058 22 10) 4 Z 6 9 9 pofeee TE Sil 4 B. coloradensis AG bo 0s I ES ese 1 5 8 8 7 “20k @hOReSar B. elegans AD Sl 227 M6 27 24 ... 12. 10, 2 0, 7 8) Stim B. hyemalis #9 Sil ali 27) slit oes. 4 AN Ai OS Sees Z & B. intermedius £9) Sil 05 AA» LD 085 7 ike + 939 8. a A a B. lokayi (50) 463 SS 7 IG, eile se SN ei see ae SS 2 : B. nivoriundus AGm 49) 24, 40 24 229203 30) 224 30 6 7 LOO eit = B. nix BS Wh OS I IS) I a cl). Alle 1 8 Salil 7 B. pilosus AB A 22 HOD es 22pm 22 © 27 22 OS 03 ee OF 0 B. reductus £6 £0) BO al 2) 27 0 2 ese 7 2 4 13 B. vlasovi 13 i) 330) I) ) 27 BS) Se SID Se ] 13 B. westwoodi 540557 27 lily 208 327 205° Ne 7:05) 30) 30) 27 Soe Ratio of Differences/Total of Characters Used has been found to feed as larvae on Dicra- nella heteromalla, Byrum atropurpureum, Mnium hornum (Withycombe, 1926) and Pylaisia sp. (Boldyrev, 1914). Striibing kept larvae of B. hyemalis alive to pupa- tion on Polytrichum pilifferum. 1 have observed larvae of B. brumalis feeding on the rhizoids of Dicranella heteromalla in the field. In the laboratory, adults I ob- served placed the long rostrum among the moss leaves, as if to feed, although the tip of the rostrum could not be seen. Stomach contents of adults reveal only small globules of a liquid substance, possi- bly because of masticating activity of the proventriculus (see Richards, 1965). That adult boreids do indeed feed on mosses and occasionally other substances has been recorded by numerous other observers. C. Lire Cycre—Both Brauer (1855) and Steiner (1937) found large larvae of Boreus hyemalis in terraria within a few months of initial collection of adults. However, in both cases no mention was made of initially checking the moss for larvae. Withycombe (1922) claimed to have reared the same species to adulthood in one year, although he gave few details of his technique. Doubts were raised when Syms (1933) found larvae of hyemalis of two distinct sizes. He postulated a two- year life cycle. Striibing (1950) was able to rear this species and confirmed a two- year life cycle. Cooper (in litt.) reports early and late instar larvae occurring to- gether for B. brumalis, B. nivoriundus, B. elegans, H. brevicaudus and H. notoper- ates. This two-year life span seems reason- able in light of other boreal adaptations. Other cold-adapted insects have increased the length of the life cycle. In the high arctic environment most Lepidoptera spend two or more years in the larval stage (Downs, 1964), and even northerly or A Systematic Stupy oF THE FAmity Borerpar (MeEcoprera) 177 TABLE 4 List or PresENcE (++) vs. ABSENCE (—) oF A CHARACTER STATE. g ARES ates Hisbe eS eS SE Sapese a eS a ee a a Seme SP e a Se S ) Ge Ree eense UES Big e 3 a AS) WES, PSS tas RS Te cs Rs ate Sea erie ar gal 6g Seas ese Ss SM me 9S om my Pa... SS eee ae ae ee ee ee eee Dee eH’ te RN eg EAN ee oe ee ee Se 5 eee = oS SP ade Ae ORR tse SE Se eae eo ese ee at jh) sosststeeee - + —- =| =| =| =| F—-| FF FF F- | | $F $F = See Pep Eile eiad = Geet ft Se ep EN Gee. See si 5)S plot Lees = = Vin atlas ae if kor as Weenre en Pat ed WE a RT i, oe” CRE im ga es Phe ) _ ae ae ne EN Se pe eres © ait i, eee a ot U0) essaseoreate + + — + —- = + = = ~—~ + + 4+ - - ~°- I... Fe a re Se ae ae i) a a ee ee eee eS eae eS 13 _..... ee Ls eh ee ee. pes es way ON a ee ee eR 15 n=. => eT a ee - -—- — — + 16 la elena ice ak ed i sey SP eae 6 LU peers A a 18... = sey dhe Ss gre dp ESB SS SR) Se i Ge 19 = OSS SE PSR este Si eee Se AR are See apy aE DOME ee a... ya) RR Ee ee ee 21 on - ee ee ee Se Se 22 se ee gh Sty SR ee Se eat eee ae ab ee DS ea eee Sa SS SS ee SS SSS i SS i eS TAY eh eee - -—- —- —- = Sos iS SS iS ie eS SS > eer ee ee ee ee PASS are ree i ee set Lee en —- -—- =—- -—- = = = Mo 26S ies eS SO SS 7 a ng he Pisa 2 STL ee uit TE. 29 ences ae eo a a a ee ee S (Saat. 2 ee a. i i - — ae Papeete (ft, 2S eel ao ee tole xc a gee eg ek ne fie pm pet riage oy ee BF ce = eS ISSR SS ee Sep) eR Se ae ae re a 35 se eg tN ie Se ea oe ee ne or Boreas Rie areal ieked aioe eS OE, Ee 26 ees 1 ee ee ye Ca, (i Ae Ee ee montane races of one species may have a prolonged life cycle (Downes, 1965). Thus, finding this adaptation within another group of holometabolous insects is not sur- prising. However, Kaufmann (1971) has shown that an arctic carabid beetle, Ptero- stichus brevicornis, can have a variable length of life cycle. A two-year life cycle needs to be confirmed for other popula- tions and species of Boreidae. The pupal and adult stages of the life cycle usually occur during the cool or cold months of the year, yet pupae of a few species have been found as early as mid- August. Pupae of B. brumalis have been collected only from September 30 until October 18. Pupae collected on October 13 were maintained in the laboratory until the adults emerged from November 24 to December 4. Adults of this species have been collected from November 19 until April 24. Two adult specimens in the 178 Tue University oF KANsAs ScriENCE BULLETIN ‘TABLED List of CHARACTER STATES USED IN TABLE 4. 1. Presence of median ocellus. 2. Occiput very rugulose. 3. Occiput with fine reticulations. 4. 19 or 20 antennal segments. 5. Antennal base below eyes. 6. Wide hypostomal bridge. 7. Spines on caudal margin of rostrum. 8. More than 24 rostral spines. 9. Body covered with long pilosity. 0. Thoracic bristles present. 1 male. 12. Complete loss of outer forewing spines-male. 13. Forewings covering hindwings-female. 14. Bristles on caudal margin of wing pad- female. 15. Transverse ridge on second abdominal tergum- male. 16. Transverse ridge on third abdominal tergum- male. 17. Third tergal ridge reduced to a_tubercle- male. 18. Eighth abdominal tergum and sternum fused- male. 19. Tergal hood on ninth abdominal segment- male. 20. Lateral margin of hood reaching only medial margin of denticular areas-male. . Forewings abruptly narrowed at mid-length- 21. Lateral margin of hood reaching only middle of denticular areas-male. 22. Lateral margin of hood reaching lateral mar- gins of denticular areas-male. 23. Projections along lip of tergal hood-male. 24. Medial projections along lip of tergal hood- male. 25. Tergal hood with median septum-male. 26. Median septum narrow-male. 27. Median septum short and forked ventrally- male, 28. Interior of tergal hood obscured by numerous setae-male. 29. More than 25 denticles on either side of ninth tergum-male. 30. Ninth tergum and sternum fused-male. 31. Presence of deeply notched ninth sternum- male. 32. Tenth segment as short dorsally as ninth- female. 33. Posterior process on tenth abdominal seg- ment-female. 34. Complete fusion of cerci-female. 35. Cerci tapering evenly-female. 36. Eighth sternal notch present-female. 37. Eighth sternal notch large-female. Ohio State University Collection are la- belled August 15, 1935, but surely these are atypical. Adults of other species have also been collected generally from Novem- ber until April; however, adults of B. coloradensis have been collected as late as May 27 in the mountains around Logan, Utah, and those of B. borealis have been collected in the Pribilof Islands only dur- ing May, July, and August. Emergence appears to be delayed until moderate weather conditions exist for spe- cies living in montane and high latitudi- nal environments. The populations of B. coloradensis at 2,134 to 3,657 m. in Boulder County, Colorado, have only been col- lected in April, whereas this same species is collected from November until May at 1,676 m. around Logan, Utah. Boreus nix is found from November to February in Montana and British Columbia, but has only been collected in April in Alaska. B. intermedius also has only been collected in April in Alaska. Thus, emergence is to be expected later in northern and montane regions. In the laboratory, I maintained adults of B. coloradensis and nivoriundus for 48 and 51 days, respectively. This is possibly shorter than natural adult longevity, yet the rigors of the laboratory environment may have been offset by protection from predation or severe cold. D. Martine anp Ovirostrion—The mat- ing behavior of Boreidae has been dis- cussed by Cooper (1940, 1974) and various other authors. Briefly, the mating behavior is as follows: In Boreus, the male upon encountering a female, attempts to clasp her with his dististyles; if he is successful, the female becomes quiescent and the male uses his long, thin hook-like wings to pull the female upon his back. The female tucks her rostrum and antennae under her thorax and between her legs A Systematic Stupy oF THE Famity Borewar (MeEcopTerA) 179 while the male’s wings clasp her forelegs; the hypovalves of the female’s eighth ster- num are inserted into the male endoan- drium, where they are held in place by the basal lobes of the male dististyles. The male carries the female upon his back for an extended period of time; Marechal (1939) used the figure of 55 hours, al- though the duration is usually much shorter than that. During this time the female moves very little. Later, the female places small, white oval eggs among the moss rhizoids and the bases of the leafy stems. In Hesperoboreus, the male approaches and leaps at the female, grasping whatever part of her anatomy he can with his disti- styles. The female actively tries to escape, dragging the male with her. When she rests, he tries to grasp her with his wing hooks. If successful, he clutches her be- tween her head and mesonotum, holding her body parallel to his own. When the female becomes submissive, the male moves his dististyles to the ovipositor, and the wing grip is released. The valves of the female’s eighth sternum are pried down and apart. The male holds the valves down by hooking the basal notch of the valves with the basal lobe of his dististyles. The valves are forced into the subepandrial pockets and held there. The female then rocks backwards 90°, and rests with rostrum and antennae folded be- tween the forelegs. She remains perpen- dicular to the male in a “female _per- pendicular pose” that contrasts with the copulatory position of Boreus which Lamb (1922) has termed “female vertical pose,” but in which the final copulatory position of the female is above and paral- lel to the male. E. Moisture—The water required for boreid survival can probably be obtained from the food. However, the relative hu- midity requirements of the microhabitat may be a major limiting factor. The hu- midity must be high enough to support a good growth of moss, and the luxuriance of moss banks where most specimens of Boreus have been collected indicates high humidity most of the year. On the other hand, boreid larvae have sometimes been found in mosses that appeared desiccated and friable (Cooper, 1974). There has even been one collection of adults of B. coloradensis in sagebrush desert at Prom- ontory, Utah. H. notoperates survives in Grimmia, a moss of dry, rocky habitats. Therefore, no one statement can cover the moisture requirements of all species of Boreidae. F, Etevation—Elevation greatly influ- ences such climatic conditions as moisture and temperature; however, its effects on the distribution of Boridae have not been determined. B. brumalis is found in east- ern North America from sea level to 4,000 feet (1,220 m). H. brevicaudus has been collected from the bottom of the Colum- bia River gorge at 100 feet (31 m.) and on Spencer Butte near Eugene, Oregon, from 1,200 to 1,900 feet (366 to 580 m.). On the other hand, B. coloradensis is normally found in mountainous regions from 4,700 feet (1,457 m.) around Logan, Utah, up to 12,000 feet (3,660 m.) near Boulder, Colo- rado. G. Temperature—The effect of tem- perature, like that of moisture, can only be guessed at because of the insects’ be- havior in selection of microhabitat. How- ever, various species’ distributions do ap- pear to be governed to some extent by dif- ferences in average temperature. The dis- tributions of H. brevicaudus, B. elegans, and possibly H. notoperates appear to be influenced by the mild, modifying effects of the Pacific Ocean. None of these spe- cies has been collected more than 90 miles from the Coast. The mildness of tempera- tures in this region can be seen in monthly means at Mount San Jacinto, California, Portland, Oregon, Seattle, Washington, 180 sijewady -g thexO] -g IpOOM}SaM -g snipawialul -g sisuspelojo3 “g == sn3!ujojijea “g sijeaiog “g IAOSPIA “Q snjanpai “g Ssijewnig “g Xlu “g snsojid “g snpuntioalu -g suebaja -g Sa}eladojou “y SNPNedivgsqg “H Tue UNIversiry oF KANsAs ScIENCE BULLETIN © © © ©) ©) ) S S) ©) =) | cm Ficure 67. Phylogeny weighted by number of species involved. (Number of character state changes number of species involved equals horizontal distance.) 181 tHE Famity Boremar (MEcopTERA) A SyYsTEMATIC STUDY OF [aouristp [ejuoztioy syenba paajoaut sotdads jo Joquinu X (ad1M} Sulsim990 asuvyd YyoIvo 104 %—sosuvyd a}ejs Jo}IvVIvYI Jo JoquinN) | *savodde ayeys JojovIeyD ve sow] Jo Joquinu puUe pdAaoAul satoads Jo raquinu Aq parystam Auasojx a ec Xo ow suare “oc =} @ SHE aac =_ 2 4 2 ca ie tO x. ¢€ = =) CA See 2 “ M = eo 5 6 SS sa}e1ad0}0u “H snpnesjAaig *H 182 Tue Universiry oF Kansas ScIENCE BULLETIN Degree of Difference s s . « > > ao w on (=) an Oo 05° s s . ie) no = an oO ao G0" . So So OL H. brevicaudus H. notoperates B. elegans B. nivoriundus B. pilosus B. nix B, brumalis B. reductus B. vlasovi B. borealis B. calffornicus B, coloradensis B. intermedius B. westwoodi B. lokayi B. hyemalis Ficure 69. Phenogram. and Vancouver, British Columbia—all lo- calities close to places where these three species of Boreidae have been found. The low mean monthly temperatures for these four localities range from 0° to 8°, and the high mean monthly temperatures range from 18° to 28° C. The high temperature recorded for these cities in 1971 was 46° and the low temperature was — 13° C. This contrasts sharply with temperature tolerances of species from the western plateau region and the East Coast. B. brumalis has been collected from Duluth, Minnesota, to Washington, D. C., with low mean monthly temperatures varying from — 13° to 3°, and high mean monthly temperatures ranging from 19° to 26° C. High and low temperatures for these two localities in 1971 were 34° and —32° C. Helena, Montana, is within the range of three western species, and it, too, has se- vere changes in temperature with high A Systematic Stupy oF THE Famity BorermAr (MeEcoprTerRA) 183 and low monthly mean temperatures of 20° and —7° C. The change in species composition with climatic zone seems to be particularly abrupt in Oregon and Washington, where B. californicus is found from the Cascade Range eastward, and H. brevicaudus is found from the western foothills of the Cascades west- ward. Microhabitats offer climatic diversity in these extreme environments. Bliss (1969) noted that “In these severe environments, where day-time temperatures at 0.7 m sel- dom exceed 13° to 18° C. and more fre- quently are 5° to 8° C,, it is of consider- able interest to note that soil surface temperatures in excess of 38° C. have been recorded by the writer in arctic Alaska as well as in the alpine tundras of Wyoming and Mt. Washington. Nighttime soil- surface temperatures for the same dates were 4° to 7° C.” Wilson (1957) found that air temperatures near a small clump of arctic Saxifraga oppositifolia were 0.5° Cera, Cat! cm above the clump, and 6° C. inside two flower buds. Wulff (cited in Porsild, 1951) recorded tempera- ture readings of 35° C. among dead leaves of a saxifrage and 10° C. in a clump of mosses in northern Greenland when air temperature was — 12° C. Thus, through behavioral preferences boreids can select microenvironmental temperatures far dif- ferent from local air temperatures. One collection of B. brumalis larvae was made on a day when air temperatures were about 24°, but the soil temperature where they were collected was only 10° C. The activity of adult boreids is clearly influenced by temperature. At 0° C.,, adults of B. coloradensis are active on the snow surface; however, this temperature appears to be close to the minimum for most boreid activity. Below this tempera- ture, they can be seen lying on the snow but will not move, even when picked up. Overnight temperatures of —37° C. in British Columbia killed specimens of B. elegans and H., brevicaudus | was trans- porting in plastic cartons for study. The maximum temperatures of adult activity are not known, but adults of B. brumalis survived a temperature of 20° for a few hours time. Herter (1943) recorded maxi- mal survival temperatures of 32-32° C. for B. hyemalis and 34-37° C. for B. west- wood. H. Tre or Activity anp Licur Re- actioN—During a collecting trip in north- ern Utah, I noticed that adults of B. colo- radensis were far more abundant on the snow surface at dusk than during the day. One active individual was even found by flashlight more than an hour after dark. This observation led to speculation that Boreidae are mainly crepuscular, or possi- bly nocturnal. Fraser (1943) made the same observation for B. hyemalis. While observing H. brevicaudus in Washington, I noted that the insects on snow did not move, even though the air temperature was 5° C, and the area was in bright sun- light. However, about three hours later (2:00 P.M.), at another collecting site 2° warmer, individuals of H. brevicaudus were active. In the laboratory, adults of B. brumalis occasionally move at night when temperatures are 7° C.; but most often remain immobile, clinging to moss stems. The two isolated observations of numerous boreids at dusk may have been due to chance, and more observations should be made. Marechal (1939) mentioned that B. hyemalis was attracted to electric lights, and Lestage (1940) speculated that this reaction explained why Boreuws was more frequently collected on bright, sunny days. Lestage’s suggestion is contrary to my own observations, although I recognize it ap- plies to a species I have not observed alive. All North American species I have col- lected are as abundant on cloudy days as on sunlit days. Temperature appears to be 184 Tue Universiry oF KANsAs ScrENCE BULLETIN a more important factor than light in their activity. TAXONOMY Key To THE GENERA OF BoREIDAE 1. Males with no outer forewing spines; fe- males with tenth abdominal tergum as short as ninth, bearing caudal, spinifer- ous extensions; female eighth abdominal sternum notched basally; cerci not fused apically yea. ’ April 1934, W. I. Lloyd, 3 males. 6 females (Carpenter, 1936, as gracilis). Gird’s Creek is about 5 miles straight east of Hamilton, Montana. The creek is on the dry eastern side of the Bitterroot River Valley, and is relatively inaccessible. However, this is the only locality where this species has been consistently collected. B. nix has been collected from 2 Nov. to 11 Feb. in Montana and southern British Columbia, while Alaskan specimens were collected from mid- to late-April. The climate and evaluation of the Alaskan lo- calities were discussed under B. interme- dius. The dry, hilly area east of Hamilton, Montana, seems to be considerably differ- ent from the glacial lowland of Alaska, but differences in climate may be modified by different emergence dates. When Carpenter (1936) discussed B. gracilis, he mentioned that this species differed from nix only in having male wings which were not abruptly narrowed at mid-length. However, in observing the series he used in describing the male, I noted that these specimens do have such abruptly narrowed wings. As there are no other differences, I consider gracilis as a junior synonym of nix. This species is a member of the dru- malis subgroup of the nivoriundus group, having bristles on the pronotum and a notched male ninth sternum. The long pilosity of the body separates this species from all others except pilosus. Males of pilosus have only about 19 denticles in each tergal area, while males of nix have about 33: Boreus pilosus Carpenter Boreus pilosus Carpenter, 1935, Psyche 42:114-115, Fig. 10. Holotype male and allotype female (no. 22360) in Museum of Comparative Zoology, Harvard University, Cambridge, Massachusetts. Present description based on 7 males, 9 females. Heap: Dark brown shading to light brown on tip of rostrum. Occiput finely reticu- late, with a few whitish setae. Median ocellus present. Antenna brown, with 18 to 20 flagellomeres. Anterior and posterior surfaces of rostrum with long, whitish setae, numerous near antennal bases, sparse over rest of rostrum. Ratio of length of maxillo-labial complex to rostrum = .80. Tuorax: Dark brown. Pronotum with indistinct transverse ridge at mid-length. Anterior and posterior margins of prono- tum with 2 to 6 yellowish-brown bristles. Meso- and metanotum occasionally with 2 bristles each. Small setae covering thorax. Lecs: Yellowish-brown with numerous dark brown and long, white setae. Apical femoral spine present, dark brown. Tibial spurs light brown. Mate wine: Yellowish-brown, abruptly narrowed at mid-length, with 19 to 22 inner and 16 to 19 outer forewing spines, 9 to 15 hindwing spines. FEMALE FOREWING: Yellowish-brown, oval; covering hindwing. Mate appoMEN: Dark brown except yel- lowish-brown on ninth segment, basistyles, and dististyles. Sclerites with many long, white setae. No tergal ridges on second and third segments. No fusion of eighth tergum and sternum or ninth tergum and 200 Tue University oF Kansas ScrENCE BULLETIN sternum. Ninth tergum with hood extend- ing laterally to mid-width of denticular areas; no median septum in hood. Den- ticles somewhat scattered, only 19 on each side. Dististyle with 17 denticles arranged from below basal lobe to base of dististylar claw. Ninth sternum deeply notched api- cally, extending to bases of dististyles. FEMALE ABDOMEN: Dark brown, except for rusty brown ovipositor and cerci. Ratio of length of ovipositor to rostrum = 1.09. Cerci evenly tapering to apex. Bopy wteNctH: Male, 2.6 to 2.9 mm., pinned, 3.0 to 4.0 mm., in alcohol; female, 4.1 to 4.6 mm., pinned, 5.2 mm., in alcohol. INTRASPECIFIC VARIATION: There is consid- erable variation in number of antennal segments and male wing spines. Some- times prothoracic bristles are rubbed off, giving this species the apearance of a com- bination of characters unlike any group. Holotype, allotype, and 11 paratypes from BRITISH COLUMBIA: Kaslo, Dec. 1932. Other collections (Fig. 75) are from ALBERTA: Jasper National Park, 2-4 Oct. 1964, 6200 feet, 1 male, 1 female; MONTANA: Missouri Co., 26 Feb. 1957, H.R. Dodge, 2 males; Missouri Co., Grant Creek, 11 April 1972, R. A. Haick, 1 male; Missoula Co., Elk Creek, 4000 feet, 22 Jan. 1972, R. A. Haick, 1 female. George Ball (personal communication) states that the Alberta specimens were collected in “an alpine meadow on a snow- covered, windblown, _ northeast-facing slope of Mount Edith Cavell, at 6200 feet.” These specimens were collected in associa- tion with B. californicus, as have been all other known specimens of B. pilosus. B. pilosus has been collected from 4000 to 6200 feet (1220 to 1891 m.) between 4 Oct. and 11 April. This species is part of the brumalis subgroup of the nivoriundus group, hav- ing thoracic bristles and a deeply notched male sternum. It can be separated from the eastern brumalis by the long pilosity, and from the western nix by its lighter coloration and by having fewer ninth ter- gal denticles in males. Boreus reductus Carpenter Boreus reductus Carpenter, 1933, Can. Entomol. 65:94-95, Fig. 1A. Holotype male and allotype female in Canada Department of Agriculture col- lection, Ottawa. Present description based on 25 males, 24 females. Heap: Dark brown. Occiput finely reticu- late, with numerous setal pits. Median ocellus absent. Antennae dark brown; 17 to 18 flagellomeres. Rostrum with many long setae near antennal bases, few setae over remainder of rostrum. Ratio of length of maxillo-labial complex to rostrum = 90. Tuorax: Dark brown. Pronotum with in- distinct transverse ridge at mid-length; no bristles present. No bristles on meso- and metanotum. Lecs: Dark brown with numerous light- colored and dark setae. No apical femoral spine. Tibial spurs light brown. Mate wines: Dark brown, abruptly nar- rowed at mid-length, with 19 inner and 16 to 19 outer forewing spines, three hind- wing spines. Femare wincs: Dark brown; reduced to small vestiges; forewing not overlapping hindwing; with short, whitish setae. Mate aspoMEN: Dark brown throughout. No tergal ridges on second and third seg- ments. No fusion of eighth tergum and sternum or ninth tergum and sternum. Ninth tergal hood extending laterally to mid-width of denticular areas. Hood pro- jecting dorso-caudally as a medial point (Fig. 47). Denticular areas large, with about 20 denticles on each side, forming a continuous area when medial cleft is not opened by internal pressure. Dististyle with about 13 denticles arranged from basal lobe to dististylar claw. Cleft be- A Systematic Stupy oF THE FAmity BoremaAr (MEcopTERA) 201 tween blunt-tipped basal lobe and rest of dististyle. Ninth sternum narrowed sharp- ly at mid-length, smoothly rounded at apex, not reaching bases of dististyles. FEMALE ABDOMEN: Dark brown. Ratio of length of ovipositor to rostrum = 1.10. Cerci dark brown, evenly tapering to apex. Internal sclerotized plate of eighth seg- ment shaped like oval disc with two ven- tral funnels. Bopy Lenctu: Male, 1.9 to 2.4 mm., pinned, 3.3 to 3.9 mm., in alcohol; female, 2.8 to 3.1 mm., pinned, 4.8 mm., in alcohol. INTRASPECIFIC VARIATION: Only variation noted is in number of antennal segments and male wing spines. Holotype and allotype from BRITISH COLUMBIA: Kaslo, 29 Dec. 1906, J. W. Cockle. In addition, there is one other female with the same data, but no type label, in the Museum of Comparative Zoology. Other specimens (Fig. 76) are from BRITISH COLUMBIA: Lillooet, Jam. 1917, A. W. A. Phair, 1 male, 2 fe- males; Ashnola River Valley, 23 Nov. 1963, P. R. Grant, 1 female; MONTANA: Ravalli Co., Hamilton area, from 27 Nov. tompeejan- Wy L.. Jellison; 15) .males;. 7 females; Missoula Co., from 6 Feb. to 11 March, John Chapman, D. S. Potter, and R. A. Haick, 6 males, 7 females; Mineral Go; St; Regis, 28 Dec. 1957, H. R. Dodge, 1 male, 6 females; Flathead Co., near Strick, 11 March 1961, W. L. Jellison, 1 male; WASHINGTON: Spokane Co., Spokane, 28: Dec» 1942, HH. W. Prescott, many specimens; Whitman Co., 8 mi. W. of Colton, 18 Feb. 1972, W. J. Turner and J. Brunner, 2 males, 8 larvae; IDAHO: Oneida Co., 5 mi. SW. of Juniper, 22 Nov. 1969, George Knowlton, 1 female; NE- VADA: White Pine Co., T18N, R54E, 1 Jan. 1971, Chambers and Nelson, 2 fe- males. This species has been taken in the arid Great Basin and mountainous regions of western North America, frequently in as- sociation with B. californicus. Specimens from Whitman Co., Washington were col- lected by Berlese funnel from Selaginella sp., a club moss. B. reductus is unusual in many ways, and should not be confused with any other North American species of Boreus. Fe- males of this species alone have forewings which do not overlap the hindwings. This is the only species for which an internal sclerotized plate is known in females. Males have no pronotal bristles and un- fused eighth tergum and sternum—a com- bination unique among North American species. The B. reductus group is the only group of Boreus with no median ocellus. The ninth tergal hood is distinctive in having a medial point. The nearest relatives of this species are found in the Old World, from the Crimea to the Pacific Ocean. Whether B. reductus is an early relic of this group in the Nearc- tic Region, or has more recently moved back into this region, cannot be deter- mined at this time. Key to Mates or Paearctic Boreus 1. Bristles present on margins of pronotum MANDI of Sealife Pal EP NOs beybienkoi No bristles present on pronotum .......... 2 2. Transverse ridge on second abdominal [0S CoA UNO al Pate athe ieee tery 3 No transverse ridge on second abdominal fen oun es ee ee eee 6 3. Ridge on second abdominal tergum di- widedsrmedialllype sa: meee kratochvili Ridge not divided «medially ae 4 4... Occiput-smoothy eee hyemalis Occiputnupuloses ne eee 5 5. Transverse ridge on third abdominal ter- (oj 10.0 Se ke ie ies Wine in westwoodi Tubercle on third abdominal tergum ........ Pais 6) Jt elt ned mtr rte od lokayi 6. Tergal hood present on ninth abdominal SEDIMEI Esse nan th EERE es tes Stee 7 No tergal hood on ninth abdominal seg- TINE C1 tgyetse, Cee eee chadzhigireji 7. Symmetrical points on either side of mid- lineofimarsiniot hood =: 22:2. vlasov1 202 Tue Unriversiry oF Kansas SciENCE BULLETIN Medial point on margin of ninth tergal aiicl igs verre that Le ee rind ee a 8 8. Medial point of tergal hood acute -........... ASE ae eed omer eee Semen eT semenovi Medial point of tergal hood smoothly rounded sapicallive sess. orientalts No males are known for sjoestedti and navast. Key to FEMALES OF PaLearctic Boreus 1. Bristles present on margins of pronotum ies a Ct ee i RS ee eke beybienkoi No bristles present on pronotum .......... Zz 2. Forewings not covering hindwings ...... 3 Forewings covering hindwings ...........- 4 3. Forewings similar in shape to hindwings ie VR = 0 eae tte navast Forewings longer and broader than hind- WEI OSueee ee eee Oe Ee chadzhigirepi 4. Cerci evenly tapering to a point ~........ 5 Cerci narrowing abruptly at mid-length LSet vlasovt 5. Known distribution—Europe ............... 6 Known distribution—eastern Asia ..... 8 G: Occiput smooth 3c hyemalts Occiput finely, reticulated — 7 7. Known distribution—mainly northern | SAUUK0) 073 eas eee ena A eee westwood Known distribution—the Balkan Penin- UU ee tan ha lel aay glee eee lokay1 8. Known distribution—Kamtchatka Penin- SHIEy ee Oe eee ERR oy sjoestedti Known _ distribution—Asian continent proper 23. semenovi and orientalis No females are known for kratochvilt. SPECIES DESCRIPTIONS Boreus beybienkoi Varbinsky Boreus bey-bienkoi Tarbinsky, 1962, Sbornik Entomologicheskikh Rabot. Aka- demiia Nauk Kirgizsk 1:134-135. Holotype, allotype, and all paratypes in Tarbinsky’s collection. The following is a translation by Mr. Christopher Starr of Tarbinsky’s original description: “Males smaller than females. Body short. Length of body 2.75 mm. Color of body black, short with metallic bronze. Pleura of thorax (episterna and epimera) and coxae of middle and hind legs dull. Entire body covered by short white setae. An especially heavy pubescence visible on abdomen, front coxae and base of rostrum. Abdomen wide. Legs long, forelegs short- er than middle legs, middle legs shorter than hind legs, hind legs more than twice as long as body. To transverse ridges. Antennae 21-segmented. Basal segment of antennae wide, second segment thick and round-oblong, apical segments almost twice as short as segments placed closer to base. Two pairs of wings. Forewings not bent hook-like, held straight along body. Setae along edges of forewings somewhat bent, and at ends directed to- ward wingtip. Forewings comparatively narrow at base, 0.25 mm; length of wing, including terminal spine 1.45 mm; termi- nal spine 0.2 mm. Length of wing almost six times greater than width at base. Fore- wings long (relative to length of body), with terminal spines, reaching to sixth abdominal segment. Along front and hind edge wings are covered by long spines, the longest of them shorter than terminal spine, placed at tip of wing. In middle part of wing all spines comparatively equal in length; shorter ones placed nearer to base of wing, but longer ones near tip of wing. Spines placed along front edge of wing began somewhat nearer to base of wing than spines along hind edge of wing. All spines evenly spaced. On dorsal sur- face of wings close to base up to 3 long bristles; entire wings covered by sparse but clearly visible hairs. Bases of wings brown, remaining parts black. Second pair of wings even more modified and in this species very thin, long and toward tip strongly curved in pincer-like form. Width of wing at base 0.15 mm.; length 1.05 mm. When extended, small spines visible from below in middle part of hindwings, but on inner edge in curved part of wing a fringe of very small setae. Hindwings brown. Pronotum short; along front edge of it 4 bristles placed in pairs, also some bristles on hind legs. A Systematic Stupy oF THE FAmity Boremar (MEcoprTeRA) 203 “No tergal ridges on dorsal side of sec- ond and third terga. Ninth tergum short; on top a comparatively deep and broad cavity. Caudal apex of tergal hood round- ed off, blunt, and densely covered with short, thick, blunt setae on each side. Above base of cavity, covering it (when examined from above), rises up a small, triangular plate which fuses at its base with middle of tergum. Ninth sternum (subgenital plate) with detached apical part and obtuse engraving along apical edge. Lateral edges of sternum with visi- ble cavities. In its total form ninth sternum similar to that of B. chadzhigirey Plig., differing from it by wider apical part and cavity along front edge. Length of ninth sternum along mid-line 0.65 mm. Gono- coxites massive, with wide tips, their bases with rather long end; dististyles with wide base, their width being from claw on inner side to lower edge of base. Apical part of dististyles terminates in massive and com- paratively short claw. Cavity between base of apical claw and claw on inner side of dististyles is uniformly deep; edge of cavity covered with short, easily visible denticles. Claw on inner side of dististyles blunt with shovel-shaped broadened upper part. “Females larger than males. Length of body, including ovipositor, 3.16 to 3.6 mm. Eighth tergum very large, one side making blunt angle with rounded tip. Ninth seg- ment comparatively short and wide. Tenth segment short and thick; when examined from above greatest width in middle, visi- bly narrowed toward tip and base. Twelfth segment wide at base with detached, bluntly cut off tip. Valves of ovipositor (eighth segment) with wide base. Distal ends of valves separate near small angle, but at base of them small cavity visible. From below, distal end of valves covered with strong, short spines, tips of which are directed toward base of valves. “Examined material: northern slope of Kirghiz Ala-Tau range (Fig. 78), low hills-foothills (the massif Paspeldik); 9 Dec. 1957, 2 males; 5 Jan. 1958, 16 males, 12 females; 4 Feb. 1960, 45 males, 40 fe- males. Author’s collections.” This species can be separated from all other Palearctic boreids by the presence of spines on the pronotum. For this reason B. beybienkot apparently is a member of the mvoriundus group. ‘The description indicates that the male ninth sternum (hy- pandrium) and uncurved wings are mark- edly different from those of other species, and may also be diagnostic characters. However, further description and illustra- tion of these structures is needed for clarity. Boreus chadzhigirejt Pliginsky Boreus chadzht-gireyt Pliginsky, 1914, Revue Russe d’Entomologie 14:366-367, figs. 9-13. The two female types from Sevastopol and Simferol, and 1 male, 2 females from Eupatoria, all in Pliginsky’s private col- lection have probably been lost (Marty- nova, 1954). Four males and 12 females from Eupatoria are in the collection of the Zoological Museum, Academy of Science, Leningrad. Following is Esben-Petersen’s (1921) translation of the original Russian descrip- tion, followed by my translation of the German summary: “Female shining; dark greenish metal- lic coloured or partly purplish shining. Rostrum yellowish-brown, darker at the tip. Palpi blackish. Eyes elliptical, yellow- ish. Head between the eyes evenly punc- tuate. Antennae 19jointed, blackish; 1st joint large and robust; 2nd globular, half as broad as Ist, and twice as long as broad; 3rd joint small, cylindrical, half as broad as 2nd, and twice as long as broad; 4th joint longer than 3rd; 5th-10th long, cylin- drical. 11th-18th cylindrical, somewhat longer than broad; 19th joint elliptical; all the joints clothed with very short blackish and whitish hairs. Prothorax as long as broad, narrowed in front, and finely rugosed transversely. Mesothorax as 204 Tue Universiry oF Kansas SciENCE BULLETIN broad as hind margin of prothorax, with parallel lateral margins and somewhat in- cised posterior margin. Metathorax broad, with a semi-lunar incision posteriorly and with a transversely placed groove in its centre. Wings yellowish; forewings ob- long and punctuate; hindwings have the shape of small narrow plates. Abdomen evenly clothed with short whitish hairs, transversely rugose. Ovipositor almost straight. Legs slender, yellowish, whitish haired; tarsi somewhat darker. “Length: 3 mm.; breadth: 0.75 mm. “Male. The male only differs from that of B. navasi in possessing 19-jointed anten- nae. “The female of this species is sepa- rated from the other species of the genus by a completely straight ovipositor and through the appearance of the forewings, which are strongly elongate and compara- tively large. The male is distinguished from B. navasi only by its 19-segmented antennae. “The area around Sevastopol and Sim- feropol, 2 females in my collection; Eupa- toria, 5 males and 14 females in the col- lection of the Zoological Museum of the Academy of Science and in my collection.” In addition to this description, further characters can be mentioned (Martynova, 1954). In the male, the eighth tergum and sternum are not fused, the ninth tergum and sternum are fused, and the ninth ster- num does not taper evenly to the apex, but narrows rather abruptly at mid-length, as in reductus, ending in a square-tipped apex. The female tenth tergum is rela- tively short, but no ovipositor-to-rostrum ratio is available. There are no caudal extensions of the female tenth segment below the cerci, and cerci narrow rather sharply, as in v/asov1. Spines are present only on the apical one-third of the hypo- valves. Pliginsky made no mention of abdom- inal tergal ridges, and Martynova stated that they were almost absent. Because this species appears to be a part of the reductus group, having fused ninth tergum and sternum, I am inclined to think that it has no tergal ridges. From Martynova’s illustration it ap- pears that males of B. chadzhigirejt have no tergal hood. However, the absence of pronotal bristles, fusion of ninth tergum and sternum, and shortening of female forewings are traits which I feel are de- rived, and evolved after development of the tergal hood. Therefore, Martynova’s illustration is probably incomplete; I sus- pect that this species does have a tergal hood that extends to the middle of the denticular areas. The three localities of Sevastopol (44° 36’ N, 33°32’ E), Simferopol (44° 57’ N, 34°06’ E), and Eupatoria (Yeupatoriya) (45°12’ N, 33°22’ E) are all’ located near the southern and western coast of the Crimean Peninsula of the U.S.S.R. in an outlying range of the Caucasus Mountains (Fig. 78). Martynova (loc. cit.) recorded the dates of collection of the type material from Eupatoria as 10 July 1905 and 18 Nov. 1905—3 Jan. 1906. The July record seems unusual, and perhaps was misread or mis- printed, although there have also been reports of collections of B. brumalis, hye- malis, and nivoriundus from mid-summer. Females of B. chadzhigireji are unusual in not having oval forewings covering the hindwings like most boreids, nor fore- wings reduced to small vestiges like re- ductus and navast, but rather intermediate forewings that are small enough not to cover hindwings, but large enough to form an elongate wing pad. Males may be identified by the sharply narrowing ninth sternum. Boreus hyemalis (Linnaeus) Panorpa hyemalis Linnaeus, 1767, Sys- tema Naturae 1:915. 12th edition. Gryllus proboscideus Panzer, 1796, Fau- nae insectorum Germanicae initia 27:18. A Systematic Stupy oF THE FAMiLy BoremAr (MEcopTERA) 205 Bittacus hiemalis (Linnaeus) Latreille, 1805, Histoire Naturelle générale et par- ticuliére des Crustacés et des Insectes 13:20. Boreus hyemalis (Linnaeus) Latreille, 1816, Nouveau Dictionnaire d’Histoire Naturelle 4:152-153. Boreus hiemalis (Linnaeus) Latreille, 1817, Le Regne Animal 3:433-434. Ateleptera hiemalis (Linnaeus) | Dal- man. Analecta Entomologica p. 34. (1823) Boreus gigas Brauer, 1876, Festschrift der K. K. zool—bot. Gesellschaft in Wien, pp. 263-300. Nomen nudum. Whereabouts of type specimens un- known. Present description based on 8 males, 8 females in alcohol. Heap: Dark brown shading to yellowish- brown on rostrum. Occiput smooth with numerous small setae. Median ocellus pres- ent, light brown. Antenna light brown basally, dark brown apically, with 20 to 21 flagellomeres. Many light setae on rostrum close to antennal bases, few setae over rest of rostrum. Ratio of length of maxillo- labial complex to rostrum = .89. TuHorax: Dark brown. Indistinct trans- verse ridge on pronotum at mid-length; no pronotal bristles. Lecs: Yellowish-brown with dark brown setae. Apical femoral spine present. Tibial spurs light brown. Mate wines: Light brown, abruptly nar- rowed at mid-length, with 22 inner and 8 to 10 outer forewing spines, no hindwing spines. FEMALE FOREWING: Yellowish-brown, oval; covering hindwing. Mate aspoMEN: Dark brown except yel- lowish-brown on ninth abdominal seg- ment, basistyles, and dististyles. Trans- verse ridge on second tergum tilted an- teriorly. Lateral margin of ridge with dorsal lobes. Tergal ridge of third seg- ment reduced to rugulose area or smaller ridge with dorso-lateral lobes. Eighth ter- gum and sternum fused. Ninth tergum and sternum not fused. Large hood of ninth tergum extending laterally to lateral edges of denticular areas. Median septum of hood long, narrow. About 20 denticles on each side of septum very sharply tipped, more like spines. About 17 spines along anterior margin of dististyle from below basal lobe to base of dististylar claw; basal lobe sharp-tipped with cleft between lobe and rest of dististyle. Ninth sternum smoothly rounded at apex, reaching bases of dististyles. FEMALE ABDOMEN: Dark brown except yellowish-brown on last 4 visible segments. Ratio of length of ovipositor to rostrum = 1.14. Cerci yellowish-brown, evenly tapering to apex. Bopy LENGTH: Male, 2.7 to 3.5 mm., in alcohol; female, 3.8 to 4.5 mm., in alcohol. INTERSPECIFIC VARIATION: I have seen some German specimens with light brown col- oration. The most important variation is in the shape of the ridges of the second and third abdominal terga, which some- times have no lobes. The shape of the ridges has been used to distinguish species of the hyemalis group. B. hyemalis has been collected from many localities (Fig. 81) in: Austria, Bel- gium, Czechoslovakia, Denmark, France, Germany, Great Britain, Italy, the Nether- lands, Norway, Poland, Rumania, Sweden, and Switzerland. Specimens have been collected from 7 Sept. to 13 April in Norway. The adult emergence period appears to be shorter than that farther south. One Dutch speci- men was supposedly collected on 22 June. This species has been collected at 1400 m. in Norway to 2000 m. in the Alps of southern France and northern Italy to 2500 m. in the French Pyrenees, but also has been collected near sea level of the Nether- lands and Belgium. B. hyemalis is the typical member of the Ayemalis group, having abdominal ter- 206 Tue University of Kansas ScreNcE BULLETIN gal ridges in the male. Of the other species in this group, B. kratochvili has a forked ridge on the second abdominal tergum, westwoodi has a ridge on the third ab- dominal tergum with no lateral lobes, and lokayi has a tubercle in place of the tergal ridge on the third abdominal tergum. Be- cause of the variation noted in the ridge of the third abdominal tergum of B. /ye- malis, this character may be unreliable in taxonomy. Meinander (1962) and others have noted that Ayemalis has a smooth occiput with scattered setal pits, while westwoodi and lokayi have a rugulose occiput. In the admittedly inadequate amount of material I studied, the third abdominal tergum was variable and dif- ferences in smoothness of the occiput hard to detect. It is my impression that with further study of a larger number of speci- mens, B. kratochvili, lokayi, and west- woodi will be synonymized. Boreus kratochvili Mayer Boreus kratochvili Mayer, 1938, Ento- mologicke listy 1:132, Figs. 3, 6, 9, 12, 15, 13, 24: Location of type specimens unknown. The following is my translation of the summary of Mayer’s (1938) original de- scription: “Large like Boreus hyemalis (male 25-355 mm.). “Body pitch black with metallic blue reflections. Rostrum, antennae, palpi, sec- ond and third thoracic segments, legs, wings, and genital appendages brown (like Boreus hyemalis L.). “Head: Surface very weakly wrinkled (high magnification! ), without large pits. Antennae: male with 25 segments; the 2 basal segments and the whole distal half dark. The first antennal segment is stout, somewhat smaller on the distal end. The second segment is broader in the middle than at the base and at the distal end; maximal breadth to length 1:15. Third and fourth antennal segments very short, the fifth through twelfth elongate to the distal end. From the thirteenth segment on, they are shortened again. “The second pair of mandibles brown, like Boreus hyemalis. Lacinia has on the base only nine, stout, dark brown bristles that sit on wider rings. On the distal end of the lacinia is a brush of long, whitish hairs. “The first thoracic segment is very dark, almost black, clearly cross-furrowed. The underside of the forewing has two rows of stouter, short, nearly black bristles; the outer row consists of 11 (5-+-1+1+3 +1) bristles. On the end the first pair of wings terminates in a long, black spine. “The second abdominal tergum bears before the hind margin a stout, somewhat foreward-bent ridge, that terminates on both sides in sharp points, but in the mid- dle is cut strongly in V-shape. For this reason the male of this species is conspicu- ously distinguished from the other males of this genus. On the third tergum is a much smaller ridge, that resembles the form of a single specimen of Boreus hye- malis; on the upper margin it is again pointed and weakly squared on the side. “The subgenital plate brown, like Bo- reus hyemalis; base to height 1:43:1, the end (from only 2 males) appears to be somewhat rounded. “The genital appendages are stout, two- branched; the upper branch is bent-hook- shaped and the second shortened; both are dark brown. “The tongue-shaped extension of the ninth episclerite brown, strongly chiti- nized, filled with long bristles, ending in a rather high, colorless, compact cone. Base to height 1.8:1. “Female unknown. “This species has until now only been collected from three localities in the vicin- ity of Brno together with Boreus hyemalis Wig Little is known about this species. Lest- age (1941) illustrated the second abdomi- 207 A Sysrematic Stupy oF THE FAamity Borerpar (MEcoprTerRA) “upatsaols "g pure ‘snipaultajul “g ‘stjvaiog snasog JO UOQNINsIp {98 ‘Si ‘teousuas *g puke syvjuatio snasog JO uoNNsIp ‘6/ “SIq ‘wausiyzpvy2 ‘gq pue toyustqhkag snasog JO WONNINsIp ‘g/ “SI “Wospja *g pue IsvavU snasog JO UOKNQIsIp ‘f/f “BLY “O8-LL sTANoIy og 62 ‘ypaysaols - !Aouawas “Gg @ Snipawiayur ~ sijejuano-g ow Sijeasoq * 82 anbiyzpeys-g ow Se i aa Ae s te i a 1 * Inosejn-g @ lo vaighaq ho) A poe ; Re cm ae ‘a ; ek ay, pea 3 im 208 Tue University oF Kansas SciENCE BULLETIN nal tergum of the male as having a bi- partite ridge, which would indicate that kratochvili is a member of the Ayemalis group. This character state also distin- guishes it from all other members of the group. This species has been collected only in association with hAyemalis, and therefore its validity comes into question. Although the ridge of the third abdominal tergum of Ayemalis varies considerably, the ridge of the second abdominal tergum appears to be fairly constant. No females are known, but this is not surprising because females of the hyemalis group are difficult to distinguish morpho- logically. The only known specimens of krato- chvili (3 males) were collected near Brno, Czechoslovakia (49°12 N, 16°40’ E) (Fig. 82). Boreus lokayi Klapalek Boreus lokayi Klapalek, 1901, Rozpravy Ceska Akademie cisare Frantiska Josefa pro Védy, slovesnost, a Umeni, Prague 10(21) :1-19. Location of type specimens unknown. Present description based on 2 males, 1 female, pinned. Heap: Occiput dark brown, finely reticu- lated. Median ocellus present. Rostrum light brown basally, shading to yellowish- brown distally. Antenna light brown ba- sally, dark brown distally, with 19 flagel- lomeres. Numerous setae below antennal bases, few setae on anterior and posterior surfaces of rostrum. Ratio of length of maxillo-labial complex to rostrum = .90. TuHorax: Light brown. Pronotum with indistinct transverse ridge at mid-length and no bristles along anterior and posterior margins. Meso- and metanota without bristles. Lecs: Yellowish-brown, with numerous setae and apical femoral spine. Tibial spurs light brown. Mate wines: Yellowish-brown, abruptly narrowed at mid-length, with about 25 inner and 8 to 10 outer forewing spines; 18 very small hindwing spines. FEMALE FOREWING: Yellowish-brown, oval; covering hindwing. Mate ABpoMEN: Shiny dark brown except light brown on basistyles and yellowish- brown on dististyles. Ridge of second ab- dominal tergum large, with acute dorso- lateral lobes, occasionally with two such lobes. Ridge of third abdominal tergum smaller than second, with a medial peak. Peak varies from indistinct to quite promi- ment tubercle. Eighth tergum and ster- num fused. Ninth tergum and sternum unfused. Ninth tergum with large hood extending laterally to lateral margins of denticular areas. Medial septum narrow, short. About 35 acutely-tipped denticles on either side of mid-line. About 23 den- ticles arranged from near base of each dististyle to near base of dististyle to near base of dististylar claw. Ninth sternum broadly rounded apically, reaching bases of dististyles. FEMALE ABDOMEN: Light brown. Ratio of length of ovipositor to rostrum = .66. Cerci light brown, evenly tapering to apex. Bopy LENcTH: Male, about 3.2 to 3.9 mm., pinned; female, 4.1 mm., pinned. INTRASPECIFIC VARIATION: Even on oppo- site sides of the same ridge of the second abdominal tergum, the shapes differ from one to two acute lobes. The third tergum is peaked medially, but may or may not have a transverse ridge. All three present specimens from Buc- secs, Transylvania (now Romania) (no date label). All known specimens of B. lokayi are from Bucsecs (Fig. 82) (see Cooper, 1972:280). This species is very similar, if not iden- tical, to B. hyemalis. However, I have not seen enough variation in the second and third tergal ridges of B. hyemalis to war- rant synonymy at this time. 209 A Systematic Stupy oF THE FAmity Borewaer (MEcopTeErRA) cg pue 1tayoIojp. 1480] Qe “1poomjsam *g JO UoNiNstp ‘Eg “Shy ‘Wyyo) “g (Y snaiog JO UoNNIMstIp ‘7g ‘shy “SIDULIAY snadog JO WoNNnQLNSIp 1s =o ysiewphy og ‘18 “SIT "€8-1g saunoLy 210 Tue Unrversiry oF KANsAs ScrENCE BULLETIN Boreus navasi Pliginsky Boreus navasi Pliginsky, 1914, Revue Russe d’Entomologie 14:364-366, Figs. 1-8. Boreus aktijari Pliginsky, 1914, Revue Russe d’Entomologie 14:367, Fig. 14. New synonymy. Syntypes (3 males, 1 female) of navasi and holotype female of akzijarz in Pligin- sky’s private collection. According to Martynova (1954) these type specimens have not been preserved. They are the only known specimens. I repeat here the translation of the original Russian description of navasi ap- pearing in Esben-Petersen (1921), followed by my translation of the German summary. “Dark-green, almost black, metallic shining. Rostrum yellowish, darker to- wards apex. First joint of palpi brownish yellow, the remainder blackish. Eyes el- liptical, yellowish. Antennae 20-segmented, blackish and clothed with whitish hairs; basal joint robust, as long as broad; sec- ond joint somewhat longer than broad; third joint oblong, half as broad as second; 4th-8th cylindrical, twice as long as broad; 9th-11th somewhat stouter and shorter than 8th; 11th-19th somewhat longer than broad; 20th joint elliptical. Prothorax nar- rowed in front, blackish and coarsely stri- ated transversely; in the male sex the hind part of prothorax is brownish yellow. Mesothorax broad, trapezoidal, punctuated and not so coarsely striated transversely as the prothorax; in the centre a transversely placed groove. Metathorax broader and shorter than the mesothorax, with a trans- versely placed groove in its centre and with a deep semi-lunar incision posteriorly. Ab- domen black, greenish metallic shining and finely rugosed transversely; in the male sex the abdomen is clothed with short white hairs; in the female sex it is naked. Legs tiny, yellowish; tarsi black- ish. “Male. Wings yellowish, reaching the o J o third abdominal segment. Forewing curved, darker at apex, shining, punctate, and finely rugose and haired dorsally; coarsely haired ventrally. Hindwings nar- rower and shorter than forewings; densely haired ventrally. Hindwings narrower and shorter than forewings; densely haired ventrally. Tip of the wings with a claw- like prolongation. No elevations on sec- ond and third abdominal segments. Sub- genital plate brownish yellow, very thin, punctuated downwardly and finely haired. The plate narrowly incised laterally. “Length: 2.5 mm.; breadth: 1 mm. “Female. Forewings very small; hind- wings have the shape of very small and narrow plates. All the wings yellowish coloured. Ovipositor blackish, curved. “Length including the ovipositor: 3 mm.; breadth: 0.75 mm. “The female of this species is differen- tiated from all presently known species by the tiny forewings and through the ap- pearance of the ovipositor which is bent under at a nearly straight angle. “The area around Sevastopol (Fig. 77), 20 Dec. 1913, 3 males and 1 female in my collection.” Mountains near Sevastopol (44°36’ N, 33°32’ E) ‘are a part of the w@agessus Range in the Crimea of southern U.S.S.R. Pliginsky separated aktijart from navast on the basis of aktijari’s 19-segmented an- tennae, rounded female forewings, lack of fine abdominal striations, and angle of depression of the cerci. B. reductus has antennae which vary from 19 to 20 seg- ments, and Cooper (1974) has found an- tennal segments to vary from 18 to 20 in H. notoperates, pointing out the ques- tionableness of making distinctions on this basis. Fine abdominal striations and de- pression of the cerci, which Pliginsky iden- tifies as the ovipositor, are not useful taxonomic characters due to variation in both. As can be seen in Pliginsky’s Fig- ures 6 and 14, the shape of the female forewings is almost identical. In addition, both nominal species have been collected A Systematic Stupy oF THE FAmity Boremar (MeEcoprTeEra) AIL only in the vicinity of Sevastopol. Because there are no type specimens, and because of lack of useful distinguishing character states, any future collections could be placed in either species with equal facility. Therefore, I feel that aktijari is a junior synonym of navast. The original description is very poor, but if other specimens are collected, the females should be immediately recognized by the low number of antennal segments and very reduced wings, as in reductus of North America. Boreus orientalis Martynova Boreus orientalis Martynova, 1954, Trudy Zoologicheskii institut, Akademia nauk S.S.S.R., Moscow 15:65, Figs. 6, 11, IDS 20 24.27: Types in the collection of the Zoologi- cal Museum, Academy of Science, Lenin- grad. The following is a translation by Chris- topher Starr of Martynova’s original de- scription: “Male. Front wing a little shorter than in B. semenovi Plig., but with base of same width; length of wing, including terminal spine, 1.4 mm.; along front edge of wing, spines begin a little closer to base and end a little closer to tip than along hind edge of wing; spines uniformly spaced. No ter- gal ridges on dorsal side of second and third terga. Ninth tergum almost trape- zoidal; at base almost same width as at tip; tip truncated, as in B. semenovi Plig., divided into two blades, forming some- what shorter slit; base of slit placed in depression, over which is small hood, as in B. semenovi Plig., but of somewhat dif- ferent form; base of hood without cavity, projection on distal side of it very gently sloping and cavities not as deep; ninth sternum (subgenital plate) very short, al- most one-third the length in B. semenou1 Plig.,. with very deep and wide cavity along top edge; gonocoxite with shorter and blunter base than in B. semenov1 Plig., also with wedge-shaped, but shorter distal end and with less massive claw on inner side of dististyles. “Female. Eighth tergum in form re- sembles eighth tergum of B. semenovi Plig., but outer corner of lateral line of it placed lower, therefore not in form of isos- celes triangle; front edge of eighth ster- num at base forms large projection, then sharply narrows and remains uniform width to the end; valves of ovipositor sep- arate gradually, forming cavity; ninth seg- ment of almost same form as in B. semen- ovi Plig.; tenth segment not as long, with deep cavity on dorsal side of distal end; this cavity in form similar to that of B. westwoodt Hag., but longer than latter; valves of ovipositor and first segment ter- minate at same level; twelfth segment in B. semenovi Plig. and B. orientalis sp. n. alike, but in B. orientalis lower side with two shallow cavities. “Examined material: Sovetskaya Har- bor, Maritime Provinces (Fig. 79), 29 VIIL-1 IX 1908, 1 male, 1 female (types), on outdoor woodpile (ZIN collection). “Both east-Siberian species closely re- lated, though also quite sharply different, one from the other.” There is no twelfth segment in female Boreus; Martynova was probably referring to the cerci. The tergal hood has a medial point, which is not as acute as in B. semenovt. The male ninth sternum appears to be extremely short, perhaps being broken off. Martynova’s Figure 6 shows the male B. orientalis with fused eighth tergum and sternum, and ninth tergum and ster- num. Because fusion of the ninth tergum and sternum and presence of a medial point on the tergal hood are only known in the reductus group, the eighth tergum and sternum are likely unfused. If B. orientalis has unfused eighth ter- gum and sternum, it is in the reductus group, and can be separated from other eastern Asiatic species by the less acute Di, Tue UNrversity oF KANsAs ScrENCE BULLETIN angle of the medial point of the hood. At this time the female cannot be definitely separated from B. semenovi and sjoestedtt. Sovetskaya Harbor (48°58’ N, 140°18’ E) is in the Tatarskiy Proliv between the Sea of Japan and the Sea of Okhotsk. Boreus semenovi Pliginsky Boreus semenovi Pliginsky, 1930, Revue Russe d’Entomologie 24: 230-231. Holotype male in collection of Acad- emy of Sciences Museum, Leningrad. I present here my translation of Lest- age’s (1941) French translation of Pli- ginsky’s original Russian description of B. semenovt. “Male. Black with slight purplish re- flections. Head metallic green, smooth above, save some points between the eyes. Frons smooth. Eyes large, elliptical. Rost- rum and palps yellow-brown. Antennae brown-black, with 24 segments. “Wings I yellow, sclerotized, punctu- ated, wrinkled, with little black spines on the borders, enlarged basally, narrowed at the apex, which terminates in a point, tri- angular seen from above, gradually nar- rowed, without lateral sinuousity, covered with a yellow pubescence directed back- wards. “Wings II forming yellow, narrow lances. “Abdomen laterally reddish; terga smooth and without ridges; sides near the edges with dark hairs; sterna covered with whitish hairs. “Ninth sternum very convex, conical, tapered caudally, its extremity slightly notched. Apex of gonopods acuminate and black. Femur I relatively thin. All legs clear brown. “Length: 4 mm. “Distribution. Russia, Aian-Nelkansk Expedition, 62nd kilometer.” Illustrations from Martynova (1954) give additional information about this spe- cies. Male with 10 inner and 14 outer fore- wing spines. Male eighth tergum and sternum unfused, ninth tergum and ster- num fused. Ninth tergum with hood ex- tending laterally to middles of denticular areas. Margin of tergal hood produced into a caudally directed, medial projection. Female eighth sternum with spines on apical one-third of hypovalves. Tenth seg- ment long, but ratio of length of oviposi- tor to rostrum not available. Tenth seg- ment may have short caudal extensions below cerci. Cerci tapering evenly to apex. The holotype was collected in U.S.S.R.: Yakut S.S.R., Jakoti, on 20. IV. 1903. The specimen was collected by the Aian-Nel- kansk Expedition on the slopes of the Oijski Mountain Range between the Lena and Aldan Rivers. Neither the lo- cality nor the mountain range can be found on available maps. However, the collection was probably made somewhere along the track that leads from Ayan (56°27’ N, 138°10’ E) to Nel’kan (57°40’ N, 136°13’ E) to Yakutsk (62°00’ N, 129°40’ E). One other collection of 7 males, 1 fe- male was made in 1926 at the confluence of the Tumpsu and Aldan Rivers (Fig. 79), Yakut S.S.R. The Tumpsu River can- not be located on available maps either. This species appears to belong to the reductus group, and males can be sepa- rated from other Asiatic members of this group by the medial projection of the hood, as in the North American reductus. Females cannot be distinguished at this time. Boreus sjoestedti Navas Boreus sjéstedti Navas, 1925, Arkiv for Zoologi 18B(2) :3-4, Fig. 2. Holotype female in Naturhistoriska Riksmuseet, Stockholm, Sweden. Present description based on original description and unpublished notes on the type specimen by George W. Byers. Heap: Black shading to brown on ros- trum. Occiput with fine surface sculpturing and setal pits. Antenna brown, with 18 A Systematic Stupy oF THE FAmity Boreraer (MEcoprTeErA) 213 flagellomeres. Rostrum with many whitish setae at antennal bases, few setae else- where on rostrum. Ratio of length of max- illo-labial complex to rostrum = .81. Tuorax: Black. Pronotum with indis- tinct transverse ridge at mid-length, an- terior and posterior margins without bris- tles, finely striated. Lecs: Brownish. FEMALE FOREWING: Brownish, oval; cov- ering hindwing. FEMALE ABDOMEN: Shiny black with short setae along caudal margins of terga and over all of sterna. Ratio of length of ovi- positor to rostrum = 1.14. Bopy LENGTH: Female, 4.9 mm., pinned. Holotype female (and only known specimen) from Achomten Bay (Bukhta Akhomten: 52°26’ N, 158°30’ E3, Kam- tchatka Peninsula, U.S.S.R. This specimen was collected on 12 June 1925 by R. Malaise. Although this summer collection date seems late, specimens of borealis from the Bering Sea area have been collected at a similar date, and there are a few scattered collections of other species at this time of year. Achomten Bay is a small, uninhabited inlet in Kamtchatka (Fig. 80), 42 miles south of Petropavlovsk, with 7000-foot- high Sopka Mutnaya volcano of the Kam- chatka-Kuril Range rising steeply to the west. This female specimen is very similar to other eastern Asiatic and Alaskan speci- mens of other species, allowing no easy differentiation. Navas gave no useful char- acters for separating this from other spe- cies. However, since this is the only species described within a radius of 820 miles, it may well be a valid species and more male specimens from the Kamtchatka Peninsula may give us additional characters for this species. Martynova (1954) did not include this species in her monograph of Boreidae of the U.S.S.R. Boreus vlasovi Martynova Boreus vlasovi Martynova, 1954, Trudy Zoologicheskii institut, Akademiia nauk Siso.R:, Moscow 15-61-62; Figs: 4 97713; 18, 22, 26. Whereabouts of type specimens un- known. Present description based on 7 males, 2 females in alcohol. Heap: Dark brown. Occiput rugulose with many white setae. Median ocellus absent. Antenna dark brown, with 16 to 17 flagellomeres. Many long setae near antennal bases of rostrum; few setae over rest of rostrum. Ratio of length of maxillo- labial complex to rostrum = 1.00. Tuorax: Dark brown. No _ transverse ridge at mid-length or marginal bristles on pronotum. Lecs: Dark brown with light-colored and dark brown setae. Apical femoral spine absent. Tibial spurs dark brown. Mare wines: Light brown, abruptly nar- rowed at mid-length, with 15 inner and 11 to 12 outer forewing spines, 3 to 4 hind- wing spines. FeMaLe ForEwING: Light brown, oval, short, reaching only middle of metanotum, barely covering hindwing vestiges, with many whitish setae. Mate aBpoMEN: Dark brown throughout. No tergal ridges on second and third seg- ments. No fusion of eighth tergum and sternum, but ninth tergum and sternum fused in many cases. Ninth tergum with hood extending laterally to middles of denticular areas. Margin of hood produced caudally on either side of mid-line. Small median septum of hood present. About 20 denticles on each side. Dististyle with about 15 denticles arranged from below basal lobe to base of dististylar claw; basal lobe blunt-tipped, with cleft between it and rest of dististyle. Ninth sternum smoothly rounded apically, reaching bases of dististyles. 214 Tue UNrversiry oF KaAnsAs ScrENCE BULLETIN FEMALE ABDOMEN: Dark brown. Ratio of length of ovipositor to rostrum — _ .93. Cerci dark brown, not evenly tapering to point, but sharply pointed at apex; fused. Bopy LENGTH: Male, 2.3 to 4.1 mm., in alcohol; female, 3.6 to 4.0 mm., in alcohol. INTRASPECIFIC VARIATION: One male was noted to have two spines at apex of ninth sternum; normally this species has no spines on ninth sternum. The sharp pro- jections of the caudal margin of the fe- male eighth tergum, illustrated in Marty- nova’s (1954) Figure 18, were not noted in specimens examined. Fusion of ninth ter- gum and sternum is complete in only about half the available specimens. (cf. Cooper, 1973). Collections of this species are from (Fig. 77) Turkmen S.S.R., Ashkabad (37° aN 3-25) aE.) 15. Jans, to 20.heb: 1932; and 1935, 148 males and females; Tadzhik S.S.R., 30 km N of Stalinabad (Dushan- be); 1100-2 (68°33" NN, 68°48" E), 10 to 12 March 1943, 3 males, 3 females; and Turkmen S.S.R., Askhabad, Dec. 1955, 7 males, 2 females. There is 1 male in the Deutsches Entomologisches Institut, Ber- lin, which I did not examine. This species is a member of the reduc- tus group having no thoracic spines and unfused eighth tergum and sternum. Al- though other Asiatic species are poorly known, this species probably can be dif- ferentiated by the two caudal projections of the lip of the hood in males, and the sharply narrowed cerci in females. Boreus westwoodi Hagen Boreus westwoodi Hagen, 1866, Ento- mol. Monthly Mag. 3:132. Boreus tarnanu Navas, 1911, Revue Russe d’Entomologie 11:277-278, Fig. 1. Boreus boldyrevi Navas, 1911, Revue Russe d’Entomologie 11:278, Fig. 2. Location of all type specimens un- known. In the original description, Hagen gave no information on type locality, although he listed the distribution as Germany, Fin- land, and England. As currently delim- ited, B. westwood: is a northern and alpine species, not found in England, as Mac- Lachlan (1869) pointed out. However, since the types cannot be located, it is im- possible to say whether the current B. westwoodi is the same one that Hagen en- visioned, or even if it is a valid species. Navas (1911) mentioned no repository of types of boldyrevi and tarnaii and they, too, cannot be located. However, the type localities are listed as “Russie, environs de Moscou” and “Pologne de la Russie: Novaja-Alexandria,” respectively. Marty- nova (1954) considered both boldyrevi and tarnanit as synonyms of westwoodt, which is commonly collected in the vicin- ity of Moscow. Present description based on 18 males, 14 females, pinned. Heap: Occiput black with fine reticula- tions. Rostrum yellowish-brown; fine setae on caudal surface, no stout setae. Median ocellus present between antennal lobes. Antenna light brown basally, becoming dark brown apically; with 21 to 23 fla- gellomeres. Ratio of length of maxillo- labial complex to rostrum = 85. Trorax: Dark brown. Indistinct trans- verse ridge at mid-length of pronotum; no pronotal bristles. Lecs: Yellowish. Apical femoral spine present, dark brown. Tibial spurs light brown. Mate wines: Yellowish-brown. Forewing abruptly narrowed at mid-length, with 26 to 29 inner and 6 to 9 outer forewing spines. Hindwing with ventral fringe of setae, no spines. FEMALE FOREWING: Yellowish, oval, cov- ering hindwing. Mate aspoMen: Dark brown, except yel- lowish-brown dististyles. Transverse ridge of second tergum with large dorso-lateral A Systematic Stupy oF THE Famiry Boremar (MecoprTerA) 5) lobes. Ridge of third abdominal tergum prominent, without dorso-lateral lobes. Eighth tergum and sternum fused. Ninth tergum and sternum not fused. Ninth ter- gum with large hood extending laterally to lateral margins of denticular areas. Me- dial septum short, thin. Only about 16 sharp-tipped denticles on each side. Ninth sternum smoothly rounded apically, reach- ing bases of dististyles. Denticles on disti- style arranged from below basal lobe to base of dististylar claw. Cleft between basal lobe and rest of dististyle. FEMALE ABDOMEN: Segments 1 to 8 dark brown, segments 9-10 and cerci yellowish- brown Ratio of length of ovipositor to rostrum = 1.41. Cerci evenly tapered to apex. Bopverencra: Male 3.0 “to 35° mmi, pinned; female, 4.0 to 4.5 mm., pinned. INTRASPECIFIC VARIATION: There is some variation in number of antennal segments and male wing spines. Lestage (1941) in- indicated variation in the male ridge of the third abdominal tergum. Svensson (1972) listed the distribution of westwood: as the following (Fig. 83): Bulgaria, Czechoslovakia, East Balticum, Finland, Germany, Italy, Norway, Poland, Sweden, Switzerland, and U.S.S.R. This species has been frequently col- lected from November to April, with a few scattered records from June, Septem- ber, and October. B. westwoodi is a member of the /ye- malis group, having tergal ridges. It can be separated from /yemalis by the fine reticulations on the occiput, from kratoch- vili by the unipartite second tergal ridge with dorso-lateral lobes, and from lokayz by the ridged third tergum (no tubercle). ACKNOWLEDGMENTS I would like to thank Drs. William L. Jellison and George Knowlton for their time and effort in collecting many boreid specimens. Dr. Lewis J. Stannard and Donald W. Webb are to be thanked for encouraging my early interest in this group and Drs. C. D. Michener and P. D. Ashlock for their helpful criticism. The National Science Foundation has been fi- nancially helpful through grant no. GB 30837 (G. W. Byers, principal investigator) and through travel funds administered by the K. U. Committee on Systematics and Evolutionary Biology (NSF grant no. GB 8785). Special gratitude is reserved for Dr. George W. Byers, who has made this project possible through his years of ex- perience with Mecoptera, his accumulated material, his concern and patience. LITERATURE CITED Banxs, N. 1923. Mecoptera. In McAtee, Insects, Arachnids, and Chilopods of the Pribilof Islands, Alaska. U.S.D.A., Bureau of Biologi- cal Survey, North American Fauna 46:158. Buss, L. C. 1962. Adaptations of arctic and alpine plants to environmental conditions. Arctic 153117-144. BotpyreEy, B. T. 1914. Notice sur le Boreus boldy- revt Navas (Neuroptera, Panorpidae). Revue Russe d’Entomologie 1:203-210. Braver, F. 1855. Beitrage zur Kenntniss des in- neren Baues und der Verwandlung der Neu- ropteren. Verhandlungen zoologisch-botan- ischen Gesellschaft in Wien 5:701-726. . 1857. Beitrage zur Kenntniss der Verwand- lung der Neuropteren. Verhandlungen zoo- logisch-botanischen Gesellschaft in Wien 7: 69-70. ——. 1863. Beitrage zur Kenntniss der Panor- piden-Larven. Verhandlungen zoologisch- botanischen Gesellschaft in Wien 13:307-324. 1876. Die Neuropteren Europas und _ins- besondere Osterreichs mit Riicksicht auf ihre geographische Verbreitung. Festschrift der zoologisch-botanischen Gesellschaft in Wien pp. 263-300. 1885. Systemisch-zoologische Studien. Sit- zungsberichte der Akademie der Wissenschaf- ten in Wien 41:237-413. Broruers, D. 1975. A revision of the aculeate Hy- menoptera with special reference to the Mu- tillidae. Kansas Sci. Bull. 50:483-648. Burtt, E. D. anp B. P. Uvarov. 1944. Changes in wing pigmentation during the adult life of Acrididae (Orthoptera). Proc. Roy. Entomol. Soc. London (a) 32:30-34. Byers, G. W. 1954. Notes on North American Mecoptera. Ann. Entomol. Soc. Amer. 47: 484-510. . 1955. A new species of Boreus (Mecoptera: Boreidae) from Colorado. Occ. Pap. Mus. of Zool., Univ. of Michigan 562:1-4. 216 Tue Universiry oF Kansas ScreENCE BULLETIN 1961. An unusual new species of Boreus (Mecoptera: Boreidae) from Oregon. J. Kan- sas Entomol. Soc. 34:73-78. CarPENTER, F. M. 1931. Revision of the nearctic Mecoptera. Bull. Mus. Comp. Zool., Harvard Univ. 72:205-277. . 1933. A new Boreus from British Columbia (Mecoptera). Can. Entomol. 65:94-95. 1935. New nearctic Mecoptera with notes on other species. Psyche 42:105-121. 1936. Descriptions and records of nearctic Mecoptera. Psyche 43:56-64. ———. 1939. Records and notes of nearctic Mecoptera and Raphidiodea. Bull. Brooklyn Entomol. Soc. 34:162-166. Corr, A. C., Jr. 1938. Insect collecting in the Great Smoky Mountains National Park, Ten- nessee. J. Tenn. Acad. Sci. 13:274-276. Coover, K. W. 1940. The genital anatomy and mating behavior of Boreus brumalis Fitch (Mecoptera). Amer. Midland Natur. 23:354- 367. —. 1951. Compound sex chromosomes with anaphasis precocity in the male mecopteran, Boreus brumalis Fitch. J. Morphol. 89:37-58. 1972. A southern California Boreus, B. notoperates n. sp. I. Comparative morphology and systematics (Mecoptera: Boreidae). Psyche 79 :269-283. —. 1973. Patterns of abdominal fusions in male Boreus (Mecoptera). Psyche 80:270. —. 1974. Sexual biology, chromosomes, de- velopment, life histories and parasites of Boreus, especially of B. notoperates. A south- ern California Boreus. II. (Mecoptera: Bore- idae). Psyche 81:84-120. Datman, J. W. 1823. Analecta Entomologica. Holmiae, Lindholm. 104 p. Downes, J. A. 1964. Arctic insects and their en- vironment. Can. Entomol. 96:279-307. 1965. Adaptations of insects in the Arctic. Ann. Rev. Entomol. 10:257-274. EsBEN-PETERSEN, P. 1921. Mecoptera. Collections Zoologiques du Baron Edm. de Selys Long- champs 5(2):1-172. Fircu, A. 1847. Winter insects of eastern New York. Amer. J. Agr. Sci. 5:274-284. Fraser, F. C. 1943. Ecological and biological notes on Boreus hyemalis (L.) (Mecoptera, Bore- idae). J. Soc. Brit. Entomol. 2:125-129. Goopwin, T. W. 1952. The biochemistry of locust pigmentation. Biol. Rev. Cambridge Phil. Soc. 27:439-460. Gostin, R. M. 1950. Additional records of Boreus brumalis (Mecoptera: Boreidae) from Ten- nessee. J. Tennessee Acad. Sci. 25:309. Hacen, H. A. 1866. Synopsis of the genus Boreus. Entomol. Monthly Mag. 3:132. Heinricu, B. AND G. A. BaArTHoLoMEw. 1972. Temperature control in flying moths. Sci. Amer. 226:71-77. Hepsurn, H. R. 1969. The skeleto-muscular sys- tem of Mecoptera: the head. Univ. Kansas Sci. Bull. 48:721-765. —. 1970. The skeleto-muscular system of Mecoptera: the thorax. Univ. Kansas Sci. Bull. 48:801-844. Herter, K. 1943. Die Beziehungen zwischen der Okologie und der Thermotaxis der Tiere. Biologia Generalis 17(1-2). Hing, J. S. 1901. A review of the Panorpidae of America north of Mexico. Bull. Sci. Lab. Denison Univ. 11:241-264. Hinton, H. E. 1958. The phylogeny of the panor- poid orders. Ann. Rev. Entomol. 3:181-206. KaurMann, T. 1971. Hibernation in the arctic beetle, Pterostichus brevicornis, in Alaska. J. Kansas Entomol. Soc. 44:81-92. Kapaexk, F. 1903. Ueber neue und wenig bekannte Arten der Palaarktischen Neuropteroiden. Bul- letin international (Sciences, mathematiques, et naturelles) Ceska akademie cisare Frantiska Josefa pro védy, slovesnost a umeni v Praze Talleiler Lams, C. G. 1922. The geometry of insect pairing. Proc. Roy. Entomol. Soc. London (B) 94: 1-11. LatreiLte, P. A. 1805. Histoire Naturelle générale et particuliére des Crustacés et des Insectes 13:17-21. —. 1816. Nouveau dictionnaire d'Histoire Naturelle 4:152-153. . 1817. Le Régne Animal 3:433-434. Lattin, J. D. 1956. The first California record of Boreus californicus fuscus Carpenter. Pan- Pacific Entomol. 32:81-82. Lestace, J. A. 1940. Pour Vhistoire des Boreus (Stegopteres-Mecopteres). I. Annales de Societe Royale zoologique de Belgique 71:5- Die 1941. Pour l’histoire des Boreus (Stégop- téres-Mécoptéres). III. Systematique. Annales de Société Royale zoologique de Belgique 72:105-125. Linné, C. 1967. Systema Naturae. 12th edition (facsimile). W. Junk, Gravenhage. 824 p. Lioyp, L. C. 1934. Two species of Mecoptera from Alaska. Pan-Pacific Entomol. 10:119- 120. MacLacuian, R. 1869. Note on Boreus hyemalis and B. westwoodii. Trans. Roy. Entomol. Soc. London S.4. 2:399-401. Marecuar, P. 1939. Contribution a la connaisance de Boreus hyemalis L. Bulletin et Annales de la Société Entomologique de Belgique 79:111-116. Martynova, O. M. 1954. Mecoptera fauna of U.S.S.R. Part I, Boreidae. Trudy zoologiche- skil institut Akademiia nauk S.S.S.R., Mos- cow 15:54-66. (In Russian) Mayer, K. 1938. Rod Boreus (Panorpata - Boreidae) Ceskoslovensku. Entomologické listy 2:129- 140, 1 plate. MEINANDER, M. 1962. The Neuroptera and Mecop- tera of eastern Fennoscandia. Fauna Fennica 13:1-87. Micxoteit, G. 1971. Das exoskelet von Notiothauma reedi MacLachlan, ein Beitrag zur Mor- A SysTEMATIC STUDY OF THE phologie und Phylogenie der Mecoptera (Insecta). Zeitschrift fiir morphologie und okologie der tiere 69:318-362. Navas, L. 1911. Deux Boreus nouveaux d’Europe (Neuroptera). Revue Russe d’Entomologie 11:277-278. (In French with Latin descrip- tions.) —. 1925. Entomologische Ergebnisse der schwedischen Kamtchatka-expedition. 1920- 1922. Part 6. Neuroptera et Mecoptera. Arkiv for Zoologi 18B(2):1-3. (In German with Latin descriptions.) Packarp, A. S. 1870. New or rare American Neuroptera, Thysanura, and Myriapoda. Proc. Boston Soc. Nat. Hist. 13:405-410. Panzer, G. W. F. 1796. Faunae Insectorum Ger- manicae. Felssecker, Nurnberg. Pererson, A. 1951. Larvae of Insects. Part II. Coleoptera, Diptera, Neuroptera, Siphonap- tera, Mecoptera, Trichoptera. Edwards Bros., Columbus, Ohio. v + 416 p. PricinsKy, V. G. 1914. New Boreus species from the Crimea (Neuroptera, Panorpidae). Revue Russe d’Entomologie 14:363-367. (In Russian with German summary.) 1930. A new species of the genus Boreus Latr. (Neuroptera, Panorpidae). Revue Russe d’Entomologie 24:230-231. (In Russian.) Porsitp, A. E. 1951. Plant life in the Arctic. Canad. Geograph. J. 42:120-145. Ricuarps, A. G. 1965. The proventriculus of adult Mecoptera and Siphonaptera. Ent. News 76: 253-256. SacE, B. L. 1973. Alaska and its wildlife. The Viking Press, New York. 128 p. SaLt, R. W. 1961. Principles of insect cold-hardi- ness. Ann. Rey. Entomol. 6:55-74. SNEATH, P. H. A. AND R. R. Soka. 1973. Numerical Taxonomy. W. H. Freeman and Company, San Francisco. 573 p. SrannarRD, L. J. 1957. The first records of Boreus Famity Borerpar (MEeEcoprerA) 27, (Boreidae, Mecoptera) in Illinois. Trans. Illinois State Acad. Sci. 50:279-280. STEINER, P. 1936. Beitrag zur Fortflanzungsbiologie und Morphologie des Genitalapparates von Boreus hiemalis L. Zeitschrift fur Morphologie und Okologie der Tiere 32:276-288. STEPHENS, J. F. 1829. A Systematic Catalogue of British Insects. Baldwin, London. 416 p. Srrupinc, H. 1950. Beitrage zur biologie von Boreus hyemalis L. TZoologische Beitrage (N.F.) 1:51-110. Svensson, S. A. 1972. Boreus Latreille, 1825 (Mecoptera). A synopsis of described species. Entomologica scandanavica 3:26-32. Syms, E. E. 1934. Notes on British Mecoptera. Trans. South London Entomol. Nat. Hist. Soc. 1933/1934:84-88. Tarpinsky, S. P. 1962. Finding scorpion-flies, Boreus sp. n. (Mecoptera, Boreidae) in the foothills of the Kirghiz Ala-Too. Sbornik Entomologicheskikh Rabot. Akademiia Nauk Kirgizsk 1:131-136. (In Russian.) Tweepiz, M. 1972. Northern and mountain insects. Animals 14:460-463. Wess, D. W., N. D. PENNy AND J. C. Martin. 1975. The Mecoptera, or scorpionflies, of Illinois. Bull. Illinois Nat. Hist. Surv. 31:251-316. WituraMs, F. X. 1916. The pupa of Boreus brumalis Fitch. Psyche 23:36-39. Witson, W. 1957. Observations on the temperatures of arctic plants and their environment. J. Ecol. 45:499-531. WirHycomsE, C. L. 1922. On the life-history of Boreus hyemalis L. Trans. Roy. Entomol. Soc. London 54:312-318. —. 1926. Additional remarks upon Boreus hyemalis L. Entomol. Monthly Mag. 62:81- 83. WutrF, T. 1902. Botanische Beobachtungen aus Spitzbergen. I. Uber die Transpiration der arktischen Gewachse. Lund. pp. 3-32. 4q z ‘2 7 La - 7 i 7 ™ py ad ere - = 2 7 ot PPE lea > a 7 pe | Spel anita ~e. = - < ~ : 5 va ‘ ; ae 2 ‘ : .; ae : . = =e ~ 2 ¢ < ine am . . Pa OR RR ee A DS Sos SA PERC AL SA ASDA ae eee oe eee eaten THE UNIVERSITY OF KANSAS SCIENCE BULLETIN © COMPARATIVE ANATOMY OF CAECILIAN ANTERIOR VERTEBRAE anacaeaesetetete te ehaaesesetetetetecanesnensetesectotanenanassteistatoee bea ROR SO RIE ee RI RO RN By EDWARD H. TAYLOR SSRIS SIRT A PPR oo. Oo Se 8 8 SRR a gS Perens cececearenaveraetaretatetatetaters etaera" iSeaseteeeteesteesseeeseceseeess Vol. 51, No. 6, pp. 219-231 June 24, 1977 ANNOUNCEMENT The University of Kansas Science Bulletin (continuation of the Kansas Uns- versity Quarterly) is an outlet for scholarly scientific investigations carried out at the University of Kansas or by University faculty and students. Since its incep- tion, volumes of the Bulletin have been variously issued as single bound volumes, as two or three multi-paper parts or as series of individual papers. Issuance is at irregular intervals, with each volume prior to volume 50 approximately 1000 pages in length. Effective with volume 50, page size has been enlarged, reducing the length of each volume to about 750 pages. The supply of all volumes of the Kansas University Quarterly is now ex- hausted. However, most volumes of the University of Kansas Sctence Bulletin are still available and are offered, in exchange for similar publications, to learned societies, colleges and universities and other institutions, or may be purchased at $20.00 per volume. Where some of these volumes were issued in parts, individual parts are priced at the rate of 2 cents per page. Current policy, initiated with volume 46, is to issue individual papers as published. Such separata may be purchased individually at the rate of 3 cents per page, with a minimum charge of $1.00 per separate. Subscriptions for forthcoming volumes may be entered at the rate of $20.00 per volume. All communications regarding exchanges, sales and subscriptions should be addressed to the ExcHaNnce LispraRIAN, UNIVERSITY or Kansas Lisraries, LAwRENCE, Kansas 66045. Reprints of individual papers for personal use by investigators are available gtatis for most recent and many older issues of the Bulletin. Such requests should be directed to the author. The International Standard Serial Number of this publication is US ISSN 0022-8850. Editor Eugene C. Bovee Ed:torial Board William L. Bloom Philip W. Hedrick Rudolf Jander Harvey Lillywhite Charles D. Michener Norman A. Slade Henry D. Stone George W. Byers, Chairman THE UNIVERSITY OF KANSAS SCIENCE BULLETIN Vol. 51, No. 6, pp. 219-231 June 24, 1977 Comparative Anatomy of Caecilian Anterior Vertebrae Epwarp H. Taytor Professor Emeritus University of Kansas Comparative Anatomy of Caecilian Anterior Vertebrae By Epwarp H. Taytor Professor Emeritus University of Kansas ABSTRACT The limbless caecelian amphibians have vertebral adaptations for their burrowing habit and mode of life that suggest an evolutionary change of certain thoracic vertebrae to assist motile, cervical functions, despite retention of rudimentary ribs on them. The changes appear related to distinct stresses placed on the vertebrae by movements necessary to burrowing. Also, the changes are distinct enough to be taxonomically useful for potential distinction of genera and/or species in the several families of caecelians. INTRODUCTION All caecilians differ from other am- phibians in lacking an appendicular skele- ton. Thus, without this, we lack the usual criteria for classifying the vertebrae. It is generally agreed that there is but a single cervical vertebra, the atlas, in amphibians. Yet, in caecilians the anterior four or five vertebrae function together in critical pivoting movements of the cranium. These retain free ribs, but although loss or fusion are commonly accepted as prerequisite of cervical vertebrae, that criterion is not in- fallible, as for example in lizards and crocodilians wherein all vertebrae anterior to the first bearing a rib articulated with the sternum are accepted as cervicals de- spite the fact that one or more posterior cervicals bear free ribs. A more realistic criterion for cervicals in limbless tetrapods lacking a sternum is a functional one, wherein the most anterior trunk vertebrae that serve importantly in pivotal move- ments of the cranium are accepted as cervicals. Therefore, in this study all of the anterior four or five trunk vertebrae are regarded as cervicals. In the evolution of animals, changes that occur in body characteristics are the result of a cause; otherwise we would have NO miracles! Usually, regions having the greatest stresses (as for example the pelvic region in frogs) tend to change most ex- tensively, to varying degree. In burrowing forms lacking limbs, the head-and-neck region evidently is just such a region of stress. For example, should an animal such as a skink find it difficult to compete with other animals for accustomed food, it either becomes extinct or acquires an ability to augment the food supply from another source. A source often sought by such types of animals, is within a few inches of the accustomed habitat, but involves burrowing in the earth, where insects, in- sect larvae, worms, etc. are usually avail- able. Animals adapting for exploitation of such food resources find limbs and digits actually a hindrance, particularly if the adaptation leads to the use of burrows for permanent habitat. Thus, in a number of cases, we find the digits are discarded one by one and finally the entire limb may be lost. Animals having run the full gamut of such burrowing adaptation, with com- plete loss of limbs, of necessity will be forced to use the head or neck for burrow- ing if it continues its course of adaptation CoMPARATIVE ANATOMY OF CAECILIAN ANTERIOR VERTEBRAE 221 to that way of life. Perhaps at this point, most snake phylogenies parted company with other invaders of the burrowing adap- tive zone, failing to evolve the subse- quently essential perfections that more successful burrowers, such as the Scoleco- phidia, caecilians and amphisbaenids, have devised. The present study is an attempt to determine whether support for this hy- pothesis may be evident in modifications, if any, that may have taken place in the anterior vertebrae of caecilians. Unfortunately, a limited number of species has been available for examina- tion. Thus, the number of samples I have used is far too small to justify generaliza- tions about consistency of structure in orders or families or perhaps even in genera. However, one may reasonably presume that the differences pointed out are obvious from the illustrations and would be evident in other examples of comparable size and age. I am not alone in considering generic and specific differences to exist in the con- tours of vertebrae. Recently, a single, small, broken vertebra found in Brasil served as the type for a new species and genus of fossil caecilians and led the de- scribers to postulate a closer relationship with an African genus than with other American genera. Such photographs as I have were pre- pared several years ago; but I was unable to obtain as much material for making skeletons as I desired and I withheld pub- lication until now, hoping to acquire a more adequate representation of species. ACKNOWLEDGMENTS The following abbreviations are used: AMNH—Anmerican Museum of Natural History, New York, New York; ANSP— Academy of Natural Sciences, Philadel- phia, Pennsylvania; DSBM—Division of Systematic Biology, Stanford University, California; EHT-HMS—Edward Taylor- Hobart M. Smith, Lawrence, Kansas; MCZ—Museum of Comparative Zoology, Harvard University; UIM—University of Illinois Museum of Natural History, Ur- bana, Illinois. DESCRIPTIONS OF CERVICAL VERTEBRAE Famity ScoLECOMORPHIDAE Scolecomorphus kirku Boulenger (Fig. 1). Upper portion of the first (atlas) vertebra very much longer than the ventral portion. No dorsal ridges evident on the vertebrae; ridges on ventral surfaces strongly marked, widening posteriorly, closely held by the anterior processes which are curved slightly inwardly. FamiLy TYPHLONECTIDAE Typhlonectes natans (Fischer in Peters) (Fig. 2). Total length of atlas much greater than median ventral length. Dorsal Fic. 1. Scolecomorphus kirkii Boulenger. MCZ 27106, Cholo Mountains, Malawi, Africa, “elev. 3600 ft.” (see Table 4, U. Kan. Sci. Bull., vol. 48, p. 602). A, Dorsal view of 5 cervical vertebrae. B, Ventral view of same. 222 Tue UNIVERSITY OF KANSAS SCIENCE BULLETIN Fic. 2. Typhlonectes natans (Fischer in Peters). MCZ 24524, Rio Magdalena, Cucuta, Colombia, South America. A, Dorsal view of 5 cervical vertebrae. B, Ventral view of same. ridges not or scarcely indicated, but the third and fourth vertebrae each with weak, paramedian dorsal grooves, the fifth ver- tebra with two stronger parallel dorsal grooves. On the ventral surface the ventral ridge is somewhat curving or saddle- shaped. The fifth has a deep groove on each side of the ventral ridge. Ribs still re- main on the second and fifth vertebrae. The anterior processes of the ventral part of each vertebra are moderately close to the preceding vertebra, and tend to curve in slightly. Famity IcHTHYOPHIIDAE Ichthyophis kohtaoensis Taylor (Fig. SAG Bs Pis.4b,.C). Inckie, 3, leneth of dorsal part of atlas 27 mm; length of ventral portion of axis 15 mm. Dorsal ridges very strong on second to fifth verte- brae; anterior processes of third vertebra grasping the ventral ridge of the preceding vertebra, those on the following two ver- tebrae curving inward around the pos- teriorly widened part of the ventral ridge of preceding vertebrae. In Fig. 4, in ventral view, the processes directed forward on vertebrae four and five widely separated. A ventral ridge barely indicated on the fourth vertebra. In dorsal view, only a vague suggestion of a dorsal ridge. Four have ribs attached. Ichthyophis beddomei Peters (Fig. 3C, D). Length of dorsal part of atlas much longer than ventral portion. The paired ventral processes extending forward from the third to the fifth vertebrae are nearly parallel, rather than curving inward. No dorsal ridges; strong ventral ridges, their posterior terminal portions widened. Two ribs attached. Ichthyophis muindanaoensis Taylor (Fig. 4A). A dorsal view of four vertebrae, the posterior three with a very low dorsal ridge. Caudacaecilia larutensis Taylor (Fig. 5A, B). Dorsal length of axis only a little longer than ventral length. A vague sug- gestion of dorsal ridges on vertebrae; well- defined ventral ridges widening posteri- orly. Processes on rib attachment some- what elongated anteriorly. Caudacaecilia nigroflava Taylor (Fig. 5C, D). Dorsal length of axis only two or three mm greater than ventral length. directed processes Anteriorly ventral parallel. FaMILy CAECILIIDAE Oscaecilia bassleri (Dunn) (Fig. 6A, B). On five anterior vertebrae, the dorsal ridge is only vaguely indicated. Ventral ridge on vertebrae 2 and 4 relatively very narrow throughout most of its length, widening suddenly posteriorly. A pair of small projections from the forwardly di- rected processes seemingly serve as braces. They are directed inwardly and somewhat backward. Oscaecilia ochrocephala (Cope) (Fig. 6C, D). Five anterior vertebrae with well- defined dorsal ridges. In ventral view, the CoMPARATIVE ANATOMY OF CAECILIAN ANTERIOR VERTEBRAE 223 A B Fic. 3. A-B, Ichthyophis kohtaoensis Taylor. Cc D EHT-HMS 3935, 10 miles N Chiang Dao, northern Thailand. A, Dorsal view of 5 cervical vertebrae. B, Ventral view of same. C-D, Ichthyophis beddomei Peters. EHT-HMS 3186, Kotegehar, India. C, Dorsal view of 4 cervical vertebrae. D, Ventral view of same. A B C Fic. 4. A, Ichthyophis mindanaoensis ‘Taylor. DSBM 20926, 11 km SE Buena Suerte, on side of Dapitan Peak, Mindanao, “‘elev. 3700 ft.’ Dorsal view of cervical vertebrae. B, Ichthyophis kohtaoensis Taylor. EHT-HMS 1838, Dansai Province, Thailand. Four cervical vertebrae, ventral view. C, Ichthyophis kohtaoensis Taylor. DSBM 25496, Kerala Forest, India. Body vertebrae, from near middle of body, dorsal view. processes enclosing the preceding vertebra much thickened, slightly curving in. The ventral ridge sharply defined and strongly widened posteriorly, somewhat pointed terminally. The differences between the dorsal and ventral views of these two presumed species of the same genus is re- markably great. Caecilia albiventris Daudin (Fig. 7A, B). Five anterior vertebrae, all except atlas with a well-defined narrow dorsal ridge. A well-defined ventral ridge on all but axis, all widening posteriorly. The vertebrae of albiventris are relatively nar- rower than most vertebrae of Caecilia. The forwardly-directed ventral processes are nearly parallel and lack small, inwardly- directed processes. Caecilia degenerata Dunn (Fig. 7C, D). Four anterior vertebrae, the second and third each with a vague, dorsal ridge 224 Tue UNIVERSITY OF KANSAS SCIENCE BULLETIN A B ¢ D Fic. 5. A-B, Caudacaecilia larutensis (Taylor). EHT-HMS 3359, Maxwell Hill, Larut Hills, Perak, Malaya (topotype, within % mile of type-locality). A, Dorsal view of 4 cervical vertebrae. B, Ventral view of same. C-D, Caudacaecilia nigroflava (Taylor). EHT-HMS 8375, Bukit Lagong Forest Reserve, Selangor, Malaya. C, Dorsal view of 5 cervical vertebrae. D, Ventral view of same. CoMPARATIVE ANATOMY OF CAECILIAN ANTERIOR VERTEBRAE jp: A B ( D Fic. 6. A-B, Oscaecilia bassleri (Dunn). EHT-HMS 4675, “Ecuador.” A, Five cervical vertebrae, dorsal view. B, Ventral view of same. C-D, Oscaecilia ochrocephala (Cope). UIM 41092, Gatun, Canal Zone, Panama. C, Five cervical vertebrae, dorsal view. D, Ventral view of same. 226 Tue UNIVERSITY OF KANsAs ScIENCE BULLETIN A B GC D Fic. 7. A-B, Caecilia albiventris Daudin. AMNH 49960, “Bogota” Colombia, S.A. A, Cervical vertebrae, dorsal view. B, Ventral view of same. C-D, Caecilia degenerata Dunn. AMNH 23354, “Colom- bia, S.A.”’ C, Cervical vertebrae, dorsal view. D, Ventral view of same. extending to posterior border of the verte- bra. In ventral view, with very narrow, clearly-defined ridges widening slightly posteriorly. The forward processes of third and fourth each with a small branch point- ing inward and somewhat backward. Caecilia occidentalis Taylor (Fig. 8A, B). Four anterior vertebrae relatively broad, with slight median ridges on last three. In ventral view, the last three with low ridges; forwardly-projecting processes widely separated, nearly parallel. Caecilia disossea Taylor (Fig. 8C). Ventral view of the first five vertebrae with last four showing a narrow, sharply-defined ventral ridge that curves down and widens a little posteriorly. The ventral processes directed forward, rounded at tips, not parallel, lacking any internal directed branches. Caecilia orientalis Taylor (Fig. 8D). A vague, dorsal ridge indicated on posterior three of the four anterior vertebrae. In ventral view, with three, sharply-defined ventral ridges tending to narrow mesially and widen slightly at posterior end. Proc- CoMPARATIVE ANATOMY OF CAECILIAN ANTERIOR VERTEBRAE DY Fic. 8. A-B, Caecilia occidentalis Taylor. ANSP 25568, Popayan, Cauca, Colombia, S.A. A, Cervical vertebrae, dorsal view. B, Ventral view of same. C, Caecilia disossea Taylor. EHT-HMS 1808, Alto Curaray, Napo Pastaza, Ecuador, S.A. Four cervical vertebrae, ventral view. D-E, Caecilia ortentalis Taylor. EHT-HMS 4677, “Ecuador,” §.A. D, Four cervical vertebrae, dorsal view. E, Ventral view of same. 228 Tue UNIVERSITY OF KANSAS SCIENCE BULLETIN A B C D Fic. 9. A-B, Gymnopis multiplicata Peters. EHT-HMS 4702, Rancho San Bosco, Tilaran, Guanacaste Province, Costa Rica. A, Five cervical vertebrae, dorsal view. B, Ventral view of same. C-D, Dermophis mexicanus Duméril and Bibron. UIM 66889, Chiapas, Mexico. C, Four cervical vertebrae, dorsal view. D, Ventral view of same. ComMPARATIVE ANATOMY OF CAECILIAN ANTERIOR VERTEBRAE 229 A B C Fic. 10. A-B, Siphonops paulensis Boettger. D AMNH 23433, “Brasil.” A, Four cervical vertebrae, dorsal view. B, Ventral view of same. C-F, Siphonops annulatus (Mikan). UIM 56668, Limon Cocha, Ecuador, S.A. C, Four cervical vertebrae, dorsal view. vertebrae, dorsal view. F, Lateral view of same. esses relatively slender, nearly parallel and rather widely separated. Gymnopis multiplicata Peters (Fig. 9A, B). Five anterior vertebrae, in dorsal view with practically no traces of ridges. In ventral view, with a relatively broad ventral ridge, a widening posteriorly, the forward-directed processes heavy, some- what blunted at their tips, not exactly par- allel. The third has a pair of small branches from the forward projecting processes. Dermophis mexicanus Duméril and Bibron (Fig. 9C, D). Four anterior verte- brae showing dimly a median ridge on last three. The lower lateral parts of the ver- tebrae show a more or less distinct eleva- tion. In ventral view, the last three with sharply defined ventral ridges widening but slightly posteriorly. Forwardly directed processes curving inwardly somewhat. Siphonops paulensis Boettger (Fig. 10A, B). Four anterior vertebrae, with distinct ridges the greater part of dorsal length. Sharply defined ventral ridges widening very slightly posteriorly; proc- D, Ventral view of the same. E, Four cervical esses at terminal forward part, parallel. Siphonops annulatus (Mikan) (Fig. 10C, D). Four anterior dorsal vertebrae with a very narrow dorsal ridge scarcely discernible. In ventral view, strong but rather narrow median ridge, widening posteriorly somewhat, the processes di- rected forward and curving inwardly as if grasping. In lateral view, foramina for blood vessels and nerves evident. Geotrypetes seraphini seraphint (A. Duméril) (Fig. 11A, B). Atlas smaller than usual. The third and fourth anterior dorsal vertebrae are elongated more than in most forms. In ventral view, anterior border of second vertebra circular rather than angular. Third and fourth vertebrae elongate, the processes nearly parallel, the ventral ridge prominent for most of length. Median ventral length of axis very short. Schistometopum gregoriu (Boulenger) (Fig. 11C, D). Four anterior vertebrae, the last three with a low dorsal ridge. In ventral view, the lower part of the atlas longer than in G. seraphint. The frontal 230 Tue UNIversiIry oF KANsAsS SCIENCE BULLETIN - lilt A B Fic. 11. A-B, Geotrypetes s. seraphini (A. Duméril). MCZ 3424, Metet, Cameroons, Africa. A, Four cervical vertebrae, dorsal view. B, Ventral view same. C-D, Schistometopum gregoru (Boulenger), MCZ 20146, Lake Peccatoni, Kenya, Africa. C, Four cervical vertebrae, dorsal view. D, Ventral view of same. border of the second ventral is strictly angular. The forward-directed processes are much thickened, nearly parallel and not terminally rounded; well-developed ventral ridges. Hypogeophis r. (Fig. 12A, B). Four anterior vertebrae, without trace of dorsal ridges. In ventral rostratus (Cuvier) view, each with a median ridge, at least on last half of ventral surface, widening posteriorly. The second is angular an- teriorly, the posterior part of ridge closely bordered by nearly parallel forward pro- jections; following vertebrae with processes longer and merging, extending outward somewhat. Uraeotyphlus oxyurus Duméril and Bibron (Fig. 12C, D). Four anterior ver- tebrae, the posterior ones with a vague dorsal ridge. In ventral view, with a thin, slightly developed ventral ridge, widening posteriorly, and curving down. slightly Forward-directed processes widely separated, nearly parallel anteriorly. posteriorly, Grandtisonia alternans Stejneger (Fig. 13). Five anterior vertebrae, not or only vaguely suggesting dorsal ridges on four posterior ones. Most ribs retained. In ventral view, each with a strongly-defined ventral ridge becoming widened at the posterior terminal, almost as if formed by two branches. Forward processes tending to curve inwardly. Gegeneophis ramaswamu Taylor (Fig. 14A, B). Four anterior vertebrae, the third and fourth showing vague dorsal ridges. In ventral view, posterior vertebrae with moderately well-defined ventral ridges, the CoMPARATIVE ANATOMY OF CAECILIAN ANTERIOR VERTEBRAE 231 A B C D Fic. 12. A-B, Hypogeophis r. rostratus (Cuvier). MCZ 48935, St. Anne Island, Mahe Coast, Seychelles Islands. A, Cervical vertebrae, dorsal view. B, Ventral view of same. C-D, Uraeotyphlus oxyurus Duméril and Bibron. MCZ 9484, Taliparabambia, Travancore, India. C, Four cervical vertebrae, dorsal view. D, Ventral view of same. Fic. 14. Gegeneophis ramaswamu Taylor. EHT- HMS no no., no locality. A, Four cervical vertebrae, dorsal view. B, Ventral view of same. A 5 terminal part of forward processes nearly Fic. 13. Grandisonia alternans Stejneger. EHT- HMS 4647 (formerly MCZ 15638), Mahe, Seychelles Islands. A, Five cervical vertebrae, dorsal view. B, Ventral view of same. parallel. < (/bmA ee? 2 Greer! —“sie~ /. arrange ) | | Pie “ i CMa by i mesetatesatenacacetenesscececeseaessscsessocnsraetesrarcrareretaretetateteratetecetstecatetonetatocstoneamiosseatonenetnnonstanenetatenataterstetera® asnoenngncencnennnegraHenRseenctsotnecnngoOsnn eseatatatatatetonatetcoonatotaceteteneteneneneioceaitena a eats anaret ee THE UNIVERSITY OF KANSAS SCIENCE BULLETIN Marine siolugical “Laporaiory TRARY JUL 27 1977 Woods Hole, Mass. RAPHICAL VARIATION IN NESTING BIOLOGY AND SOCIAL ORGANIZATION OF HALICTUS LIGATUS = ote, od = ES 7: Bo so 2.4 = eeatetete: _A OOO) yorereren 0.020. reese woeeceseetetatatatetstetetete ot otetetetetotats By ore seseceeeeecitearareratatctareveratePerate'e CHARLES D. MICHENER and FRED D. BENNETT rere SS nN RF RR OR We%e%e CK *, = wre ate te we SORE Ro "ee << see Se oe os "ewe ae, ed poet ee S ie oe By = ieee ate! By * se stot RR Vol. 51, No. 7, pp. 233-260 July 5, 1977 eee seeteteeeteees, SOX ANNOUNCEMENT The University of Kansas Science Bulletin (continuation of the Kansas Unt- versity Quarterly) is an outlet for scholarly scientific investigations carried out at the University of Kansas or by University faculty and students. Since its incep- tion, volumes of the Bulletin have been variously issued as single bound volumes, as two or three multi-paper parts or as series of individual papers. Issuance is at irregular intervals, with each volume prior to volume 50 approximately 1000 pages in length. Effective with volume 50, page size has been enlarged, reducing the length of each volume to about 750 pages. The supply of all volumes of the Kansas University Quarterly is now ex- hausted. However, most volumes of the University of Kansas Science Bulletin are still available and are offered, in exchange for similar publications, to learned societies, colleges and universities and other institutions, or may be purchased at $20.00 per volume. Where some of these volumes were issued in parts, individual parts are priced at the rate of 2 cents per page. Current policy, initiated with volume 46, is to issue individual papers as published. Such separata may be purchased individually at the rate of 3 cents per page, with a minimum charge of $1.00 per separate. Subscriptions for forthcoming volumes may be entered at the rate of $20.00 per volume. All communications regarding exchanges, sales and subscriptions should be addressed to the ExcHancr Lisrarian, UNIVERSITY or Kansas LiprariEs, LAWRENCE, Kansas 66045. Reprints of individual papers for personal use by investigators are available gratis for most recent and many older issues of the Bulletin. Such requests should be directed to the author. The International Standard Serial Number of this publication is US ISSN 0022-8850. Editor Eugene C. Bovee Editorial Board William L. Bloom Philip W. Hedrick Rudolf Jander Harvey Lillywhite Charles D. Michener Norman A. Slade Henry D. Stone George W. Byers, Chairman THE UNIVERSITY OF KANSAS SCIENCE BULLETIN Vol. 51, No. 7, pp. 233-260 July 5, 1977 Geographical Variation in Nesting Biology and Social Organization of Halictus ligatus Cuarves D. MICHENER Departments of Entomology and of Systematics & Ecology, and Snow Entomological Museum, University of Kansas, Lawrence, Kansas 66045, U.S.A., and Frep D. BENNETT Commonwealth Institute of Biological Control, Curepe, Trinidad, West Indies TABLE OF CONTENTS I. TEMPERATE POPULATIONS INestesitesi,: mee seen unt eens Chee SE eS Seasonal activity TE Tl, Seneorl ID Ete) eee Nest structure and development —........... Mihemsocial liter cycles an.ctee--e eee ee ee Caste differences Overwintering Warm temperate populations ____.................. II. TropicaL PopuLaTIons Methods IFES SCE. A a I Fi ee ae ee Seasonal activity Nest structure and development __............. Colony size [iain fines stag ese a ese eee ee ree Guards Foragers Table 2. Percentages of Insemination and Ovarian Classes in Bees from Curepe, Sti TC aC eee ek eee ieee Table 3. Relations among Attributes of Females from Curepe, Trinidad ............. 256 III. Discussion AND CONCLUSIONS ...........-..---.------ 257 Table 4. Percentages of Bees (excluding nest foundresses) from New York and Trinidad having ovaries of various size Classes? X(A=D))) a: Sth orn ote CF ee SE 1k 259 IVZCEeTDERATURE, CIGED) 23a s see eee 259 V. Ficures 1. Distribution of Halictus ligatus 235 2-4. Nests from Lafayette, Indiana -............. 239 5-6. Nests from Toronto, Ontario .............. 239 7-8. Nests from Lawrence, Kansas ............ 239 9. Pollen masses from cells from Indiana .... 241 10. Histogram, frequencies of head-width of females, eastern third of Kansas _......... 244 IE Nese Wake! Placids Flonday 247 12. Nests, Mitla, Oaxaca, Mexico .................- 247 13. Nest entrance, Curepe, Trinidad ............ 250 42, Cell Curepes hrinidad ee ee 2 [DElveeNestss Gurepes aranidadees 2 22 18. Histogram, frequencies of head-width of remalesssCurepe.wliri mica pss. eee oe 255 234 Tue UNIVERSITY oF KANsAs ScrENCE BULLETIN ABSTRACT The halictid bee Halictus ligatus ranges from southern Canada to Colombia and Trinidad. In cool, temperate regions it develops small, summer, eusocial colonies similar to those of many halictids and consisting of a queen and several workers which are usually unmated and which lay few if any eggs. These colonies collapse in late summer or autumn, when the old queen and workers die; only new young queens survive the winter. In the tropics and prob- ably also in Florida and elsewhere in the warm temperate regions, while such eusocial colonies may occur as an ontogenetic stage, the colonies become larger and presumably consist of daughters or more remote descendants of the foundress. They are active through all or most of the year and all adult females appear capable of surviving unfavorable seasons. While some of the bees are old foragers, unmated, with slender ovaries, i.e., workers, over half are mated, egg-laying foragers. Most of the eggs are laid by such foragers and in most colonies in these warmer regions there is no evidence of nonforaging reproductives, i.e., queens. The social structure in the tropics does not correspond to any of the named types of social organizations. There is a continuum in size with no bimodality from the smallest to the largest females. In temperate regions large individuals are produced primarily in late summer and are the young gynes which will overwinter. In the tropics, large individuals are produced at all seasons when there is reproduction. At least some large individuals are produced in large cells with elongate food masses. Other geographically-varying, biological features are as follows: From Florida southward, males seem to be produced in any months when females are produced. In cool temperate regions, perhaps with the exception of the northernmost population studied (Ontario), males are pro- duced principally in late summer; early summer progeny are nearly all females. Northern popu- lations usually make shallow nests which are often aggregated. From Kansas, southward, nests are commonly deep and rarely aggregated. Nest guarding is common in cool, temperate re- gions, rare in the tropics. INTRODUCTION This paper is an account of the nesting The subgenus Halictus is largely re- and social biology of a sweat bee, Halictus stricted to temperate regions and has its ligatus (Family Halictidae). In most parts center of abundance and diversity in Eu- of its range this species is common and is rasia. H. ligatus is the only species of presumably a major pollinator, especially the subgenus occurring in the American of many Compositae. It is important, tropics and it does not range south of therefore, that its biology be known, to northern South America. For these rea- serve as a basis for its possible future management. The objectives of this study are to make biological information avail- able for this species, to describe a novel type of social organization found in its tropical populations, and to elucidate the geographical variation in social behavior found in this species. Halictus ligatus is one of the most widespread bees, ranging from southern Canada to northern South America (Fig. 1). At least a little information on nests in burrows in the ground; it falls in the subgenus Halictus, but is a distinctive form not closely related to any other species. sons it seems likely that this species is a recent invader of tropical climates and that attributes of populations in temperate North America are primitive, relative to those of tropical populations. This view is supported by the fact that the social organization in cool temperate areas is similar to that of various other well known halictid bees such as Lasioglossum imt1- tatum and L. zephyrum. In contrast, the social organization of H. ligatus in the tropics is in some features unique among known insects and thus differs consider- ably from its relatives, including northern populations of the same species. NestiNc Brotocy AND SocrtAL OrcGANIZATION oF Halictus ligatus i) W VI Fic. 1. Distribution of Halictus ligatus with dots indicating the localities where field observations have been made. Areas within the general range where the species does not occur, such as high mountains and deserts, are not indicated. and seasonal cycles is available from di- verse localities, indicated by dots on the map. It is a primitively-social species, usually living in small colonies of females Halictus ligatus is probably much more abundant, now, in the formerly wooded parts of its range than in primeval times and may have extended its range sub- stantially. Thus, no early collector is known to have obtained it in Trinidad or in South or Central America. Yet in the present century it is common in those areas. The life cycle and social organization of the species in the temperate region is reviewed in some detail in part I, below. Our data from tropical America are pre- sented as part II. Part I is the work of the first author and is largely based on the unpublished and published works of others acknowledged therein. Far more than for most papers, we are indebted to the helpfulness and cooperation of these individuals. Part II is a joint work of both authors, based largely on observa- tions made in Trinidad and materials col- lected there. We are indebted, however, to Drs. M. J. West-Eberhard and W. G. Eberhard for greatly facilitating CDM’s brief work in Cali, Colombia. This paper was made possible by National Science Foundation grant num- ber GB 38502 to the University of Kansas. We wish to acknowledge dissections and 236 Tue Universiry oF KAnsAs ScIENCE BULLETIN measurements of bees by Kenneth W. Richards, James C. Trager, and Robert E. Gorton, as well as aid in statistical work by the last mentioned. We are especially indebted to Drs. George C. Eickwort and Marcia Litte for reading the manuscript, suggesting improvements, and generously providing information from their own observations. I. TEMPERATE POPULATIONS Except for the last subsection, this sec- tion is based on populations in the cool, temperate part of eastern North America, i.e., latitude 39° and northward and east of the 96th meridian. Information on populations of cool, temperate regions is from several sources. The first im- portant account of the species concerns populations in the vicinity of Lafayette, Indiana; it is by Chandler (1955), who has given us permission to include ma- terial from his unpublished work. It con- tains many important conclusions about the species, but few data are presented. Chandler apparently made few dissections of bees and no examinations of spermathe- cae. Parts of Chandler’s results were pub- lished by Roberts (1973), along with some information from Oregon. Additional material on the Indiana population is in- cluded in unpublished work by the late Dr. Rodney Kirkton (1963, 1968). A series of publications by Knerer and co-authors contains information on H. ligatus populations near Toronto, On- tario, Canada, and Kirkton (1968) also provides certain information on the same populations. A comprehensive treatment of the Toronto population has not been provided. Kirkton (1963, 1968) assembled infor- mation about caste differences and geo- graphical variation in H. ligatus. His information about nests is not only from Indiana and Ontario, but also from Oaxa- ca, Mexico. Litte (in press) made a study near Ithaca, New York, which is very im- portant in clarifying observations made at the other sites. She is the first author to provide appropriate statistical data to doc- ument her conclusions. Near Lawrence, Kansas, one of us (CDM) has made observations on nests of H. ligatus, some of them together with H. V. Daly; A. Wille, and Ex AgiGress: No major study was made, however, for the nests are sparse. A few of the results appeared in Sakagami and Michener, 1962. From the sources listed above, the following account of nesting behavior of H. ligatus in cool, temperate regions has been assembled; authors’ names used without dates refer to the sources cited above. Thereafter an account is given of our limited knowledge of the species in warm, temperate regions. Nest sites: The nests consist of bur- rows in gently sloping or flat soil, either bare or with short vegetation, exposing considerable soil. In Indiana, Chandler suggests, nests disappeared from an aban- doned garden when weeds became too dense. Nest sites are ordinarily exposed to the sun, but a few of the nests studied in Kansas were partially shaded by trees. Soils are varied, from soft, garden loams or mixed sand and clay to firm silt or hard-packed bare clay paths. There is a strong tendency for nests to be aggregated. Thus, Chandler reports an aggregation of nearly one hundred nests in Indiana, Litte studied three dense ag- gregations in New York, and Roberts mentions dense aggregations in Oregon. Kirkton believed that aggregations in In- diana and Ontario, usually on dry hilltops, resulted from limited suitable nesting habitats. Litte, however, believes that the aggregations result from bee behavior, not mere limitation of suitable soils. She notes (77 /itt.) that two aggregations con- tained 295 nests, an average of 15 nests Nestinc Biotocy AND SoctAL OrGANIZATION oF Halictus ligatus 237 Tasie 1. Seasonal Data for Halictus ligatus in the Cool Temperate Region, Eastern North America. Location (North Latitude) and Source Flight Season N. Dakota (Stevens, 1951; June 18-late Sept. (47°) Kirkton, 1968) Ontario (Knerer and mid May-mid Oct. (43°40’) Plateaux-Quénu, 1967a; Kirkton, 1968) New York —_(Litte, in press) May 8 (or 19)+- (42°30) Sept. 15> (or Oct. 14) Indiana (Chandler, 1955) May 3-Oct. 10 (40°30’) Indiana (Kirkton, 1968) late Apr.-mid Oct. (40°30’) Kansas (Kirkton, 1968) early Apr.-late Sept. (39>) First Last Young Old Workers Males Gynes Queens Aug. 10-? (peak mid Sept.) late June- late June-early Oct. mid Aug. mid July* early Sept. (peak early Sept.) June 21 (or July 28 (or 31)-? Aug.5 Aug. 31 July 11)-? June 26- July 31-Oct. 12 Aug. 10 Aug. 22 Aug. 29 peak mid Aug. peak mid Aug. * One reported in a nest on August 13 (Knerer and Plateaux-Quénu, 1966). +G. Eickwort (in litt.) reports flight in the warm April of 1976 as early as April 19, but nesting activity did not begin until June. per m*, the mean distance between nearest neighbors being only 14 cm. Solitary nests are also recorded from Indiana, but the large aggregations are striking when com- pared to sites in Kansas and more southern localities, where nests are usually solitary or in loose groups of only two to ten. Seasonal activity: Table 1 shows the timing of some major landmarks in the seasonal activity of H. ligatus in the north- ern half of its eastern North American range. The flight season begins with the emergence from the ground of the over- wintered gynes and ends with the dis- appearance into the ground of the new young gynes that will hibernate during the following winter. The column for workers gives the flight season for workers of matrifilial colonies, i.e., daughters of the overwintered gynes which are by now queens. Since workers are short-lived, no one worker lives for the whole period indicated. The first young gynes are the first daughters that will pass the following winter. The last old queens are the last of the overwintered gynes that have lived through the summer as queens. Kirkton’s data are based on museum specimens from squares 270 miles on a side, while data presented by Chandler, Litte, and by Knerer and Plateaux-Quénu are presum- ably from specific nesting sites. Nest structure and development: Nests in temperate areas have been described by Chandler, Kirkton, Litte, and Roberts. Sakagami and Michener (1962) also gave data on nests. In spring, overwintered gynes establish nests by making new burrows, appropri- ating burrows of the same or other species, or remaining in and refurbishing the overwintering burrows, ie., the nests of the previous summer. A circular tumulus usually develops; it is much larger at the entrances of new burrows than at those of the other nest types. Such tumuli range up to 5 cm in diameter and over 1 cm high. Similar tumuli are found at nest entrances during the summer, whenever the nests are being extended. Tumuli at entrances of certain nests found in Kansas were highly asymmetrical, fan-shaped, be- cause the upper two or three centimeters of the burrow sloped strongly and tumulus 238 Tue UNIversITY oF KANsAs ScIENCE BULLETIN material was ejected up the sloping burrow rather than symmetrically around a more or less vertical burrow entrance. The bur- row entrance is round; Chandler says it is often slightly smaller in diameter than the rest of the burrow, but contrary to that of most Halictinae, often not constricted. In Kansas, however, measurements show entrance diameters of 3 mm, burrow diam- eters of 5 mm. The soil particles of the tumulus are usually loose and easily blow or wash away, but walls of packed soil, probably consolidated with the aid of a secretion, sometimes extend from the upper part of the burrow up through the tu- mulus. In such cases, dispersal of the loose material of a tumulus by wind may leave a fragile turret at the nest entrance. The entrance is plugged with soil at night and in inclement weather; indeed, Litte reports that nests inhabited by a single gyne are open for only about 30 minutes per day, while those with multiple foundresses re- main open for about six hours. Below the soil surface the burrows commonly slope slightly, or often consid- erably in the lower parts where cells are excavated. Thus, cells are commonly at- tached to portions of burrows sloping as much as 45°. Litte, however, found most burrows that she examined to be vertical. Chandler says that the consolidated lining of the burrow often extends downward, so that only. the lower extremities lack it. Chandler reports horizontal, lateral bur- rows 2.5 to 10 cm below the soil surface, each ending in a roughly excavated en- largement (forage cell), in early spring nests. Other authors do not report such features in burrows of this or other halic- tids and their construction by this bee re- quires verification. These laterals and cavi- ties are filled with earth before brood cell construction. The spring nest is therefore an unbranched burrow at the time of cell construction and provisioning; Litte re- ports the average nest depth at this time to be 15 cm and Kirkton says that brood cells are between 7.5 and 12.5 cm deep. The first completed cell found by Chand- ler, in different nests in Indiana, occurred from as early as May 15 to as late as June 9. After a lone overwintered gyne has made and provisioned her brood cells, she closes the nest entrance with earth which fills the upper 3 cm or more of the burrow. She remains in the closed nest until emer- gence of her first adult workers. In Indi- ana, Chandler reported that nest closure occurred from June 3 to 14, reopening being on June 26 to 28, or even later. Litte found, however, that nests occupied by two or more overwintered bees do not have a distinct, inactive period between spring and summer phases, such colonies remaining more or less continuously active. With emergence of daughters, exten- sion of the burrow is begun and this com- monly involves branching of the burrow. New cells are mostly along the branches. The maximum number of branches is six. As shown in Figures 2-6, there is often no deep, “main” burrow; below about the level of the bottom of the spring nest or somewhat deeper, branching occurs with- out a recognizable main axis. Some deep nests in Kansas are still unbranched in August, however (Figs. 7,8). Perhaps lack of branches characterizes nests in dry soils where great depth is necessary to attain needed humidity. In Indiana, the cells in summer were less than 25 cm deep, an unsuitable layer of soil being at that depth (Chandler, in Roberts). Such shallow nests, however, appear normal in Indiana, New York, and Ontario. Kirkton says summer nests are 15-30 cm deep, uncommonly in dry summers 45 cm deep, while Litte reports average summer depth as only 18 cm, the cells obviously being even shallower. In a moist region such as the northeastern United States, the cells are probably con- sistently shallow. As in other halictids, they are deeper when the soil is dry than when it is wet, for they must be con- NestiNc Biotocy AND SoctIAL OrcANIZATION oF Halictus ligatus 239 !Ocm Fics. 2-4. Nests from Lafayette, Indiana, June, July 27, and August 13, 1963, respectively; contained three, seven, and eight adult females, respectively (after Kirkton). For explanation of symbols see Figures 15-17. Fics. 5-6. Nests from Toronto, Ontario, August 11 and 27, respectively; contained ten and fourteen adult females, respectively (after Kirkton). For explanation of symbols see Figures 15-17. Fics. 7-8. Nests from Lawrence, Kansas, August 29 and 22, 1953, respectively. Each contained four adult females. For explanation of symbols see Figures 15-17. structed where the soil is moist enough to be workable. Therefore, as summer passes, particularly in dry years, the cells are constructed progressively more deeply. In Kansas, nests are extended to depths of at least 70 cm in the dry summer months (Sakagami and Michener, 1962). On the basis of few excavated nests, spring cells seem to be at the same depths as in Ontario and Indiana, but in July and August, cells are constructed at depths of 40-65 cm. Brood cells are of the usual shape for Halictini—horizontal to sloping slightly downward, scattered along the burrows, much as in such Lasioglossum as L. (Dialictus ) imitatum or versatum (Miche- ner and Wille, 1961; Michener, 1966). Ex- cept for the short necks, they are lined with wax-like material; also, as in most other Halictini. Worker-producing cells are 9-11 mm long, 5-7 mm in diameter, with the neck joining them to the burrow 1-3 mm lone 240 Tue UNIversITY oF Kansas SciENCE BULLETIN (Chandler). Chandler reports that in the spring a queen constructs two cells and lines each, then provisions one, then the other. Then she constructs and _provi- sions each subsequent cell before be- ginning with the next. It seems unlikely that this is a consistent pattern. As soon as a cell is completed and provisioned and an egg laid in it, it is closed with earth which is so smoothed at the surface of the burrow that the loca- tion of the cell is not visible from the burrow. There is no evidence that the cells are opened again until emergence of the adult, although inspection of the young as in Lasioglossum zephyrum and other species of Dralictus (Michener, 1974, p. 67) may well occur. Soon after a bee emerges from its cell, the cell is filled with earth. Re-use of old cells has not been recorded and pre- sumably does not occur although new cells may be made in the same vicinity. Chand- ler notes that cells in which mold develops are somehow recognized in spite of the closed entrances and are packed with soil, a process which pushes moldy material into a mass at the far end of the cell and apparently inhibits fungal proliferation. (Similar behavior is known in Nomza; see Batra and Bohart, 1969.) A lone gyne makes up to six (Chand- ler) or seven (Litte) cells before closing her nest and awaiting emergence of the first workers. Chandler noted that during one spring in Indiana most gynes made only three cells each, a few up to five; another year when spring weather was better, most nests contained four to six cells before being closed. Litte noted even greater differences between years in New York, mean production being 1.9 vs. 4.0 daughters in two different spring seasons. She found nests with two founders some- times producing more workers, up to 14. After emergence of the first daughters, cell construction and provisioning go on continuously through the summer. In mid- September, one of Chandler’s nests in In- diana contained 40 cells, including the old, earth-filled ones, and a few other old cells may have been missed in opening the nest. This was the maximum size observed in Chandler’s Indiana study. On August 27, a nest illustrated by Kirkton from Indiana contained 50 cells, 12 of them earth filled. These authors give no numerical data but their illustrations suggest that nests often have 15 to 30 cells (including earth filled ones) in August and early September. Litte gives the maximum cell number found among nests in New York as 32 with means of only 10.1 and 6.8 in two different years; these figures do not include old earth-filled cells and thus exclude spring production of daughters. Although cells in the gyne nests in the spring are constructed in progressive se- quence, the shallowest first and subse- quent ones progressively deeper, locations of summer cells are often irregular or even retrogressive, so that young of dif- ferent ages are intermixed. Cells from which males emerge are similar to those that produce workers, possibly slightly larger in average size; Chandler gives their diameters as 6 to 7 mm, lengths 10 to 115 mm. Queen-pro- ducing cells, however, are markedly larger according to Chandler, having diameters of 5 to9 mm (X = 6,5) and lengths 12 to 17 mm (X = 14). Reproductives of both sexes are produced from cells irregularly interspersed among worker cells, according to Chandler. Provisions, as in most other Halictinae, form a smooth, firm, flattened ball on the top of which, in a weak longitudinal de- pression, a curved egg is laid with its long axis parallel to the long axis of the cell and only its two ends touching the food- mass. Pollen-balls in worker and male- producing cells are reported by Chandler to be about 3 mm wide, 4 mm long and 2 mm deep, but his photographs, in agreement with our Kansas data, show Nestinc Brotocy AND SocrAL OrcAnizAtTion oF Halictus ligatus 241 lengths and widths more nearly equal, the ball being nearly circular as seen from above (see Fig. 9). The pollen-masses in Fic. 9. Pollen-masses from cells from Indiana, traced from photographs by Chandler (1955). Mass from queen producing cell at left, from worker cell at right. queen-producing cells, however, are not only larger but also elongate, 3.5 mm wide and 5 to 9 mm long (Fig. 9), according to Chandler’s text and photographs. No other author reports such a modified cell size and pollen mass shape for queen production in H. ligatus, but the parallel with the Damitas population of Lasio- glossum (Dialictus ) umbripenne 1s obvious (Wille and Orozco, 1970). Two pollen masses from a nest in Florida (see below) were large and elongate, however, while another from a different nest was round (G. C. and K. R. Eickwort, unpublished). The social life cycle: For a short time after nests are first opened in the spring, Chandler, in Indiana, noted that the one to several gynes in a nest come and go, feed- ing at flowers. It was at this season that he reported construction of the “foraging cells” to which the gynes retreated when disturbed and at night. He reports that brood cell construction does not occur until the foundresses disperse, leaving only one in the parental nest as well as in each new nest. Litte’s data from New York appear to support such dispersal, for she found more bees in the early spring nests than later, when brood was being reared, but in different years she found 25 and 32 percent of the nests during brood rear- ing with two or more bees. Knerer and Plateaux-Quénu (1966) report that in On- tario 60 percent of the spring nests contain more than one (2-7) female. These found- resses, according to the same authors, are not always overwintering groups of. sis- ters, but may contain individuals from other nests. The largest female of a group becomes the queen, the others have less ovarian development and become, in ef- fect, workers or auxiliaries. Litte reached similar conclusions, although in three of eleven two-foundress New York nests, the two females had similar ovarian develop- ment. Kansas data on these matters are meager. Of six spring nests found in dif- ferent places and different years, only one contained two females in mid-May when cells were being provisioned in it. At least in one nest studied by Knerer and Plateaux-Quénu (1966), a marked auxiliary survived and before dying for- aged in early July along with the new workers, thus making an easy transition from a spring, semisocial colony of over- wintered mated gynes to the summer, matrifilial, or eusocial colony with a queen and unmated workers. Litte also found in four nests that auxiliaries retained their be- havioral roles for at least several days after daughters had emerged and started to forage and two nests opened in August each contained two large, worn, insemi- nated, foundress-ttype bees, presumably overwintered gynes; one in each nest had smaller ovaries and was presumably an auxiliary. In Indiana (Chandler) and New York (Litte) the first brood of offspring, appear- ing in late June, consists of workers, there being no males until late July. In Ontario, however, according to Knerer and Pla- teaux-Quénu (1967a), some males are pro- duced in late June, along with the first workers. Under laboratory conditions, the same authors (1967b) found that the mo- nogynous colonies produce _ principally males, the polygynous colonies, workers. 242 Tue UNIvErsITy oF KANsAs ScIENCE BULLETIN According to Chandler, male produc- tion increases as the season advances, 50 to 75 percent of the cells containing male pupae in the period June 30 to August 14, after which the percentage of male pupae decreases. Adult males leave the nests permanently soon after emergence from the cells. As many as 12 males were found developing in a single nest. In Indiana (Chandler), New York (Litte), and Ontario (Knerer and At- wood, 1967; Knerer and Plateaux-Quénu, 1967a) worker production continues through much of the summer, the last, newly-constructed worker cells being found about August 1 and the last, active, adult workers on August 29 in Indiana. The length of life of adult workers is so short that colonies never become large. Both Chandler and Litte found that the maximum number of adult workers in a nest at any one time was nine or ten; Litte reports average colony sizes (adult females, including queens) in August of two different years as 4.3 and 3.4 (N = 25 and 28, respectively). Chandler says that 18 to 20 workers in total may be pro- duced by a colony, but presumably less than half of them are usually living adults at any one time. Litte found lower pro- ductivity, a maximum for both sexes of 20, with means of only 8.3 and 4.8 in different years. The greatest number of developing workers found in cells of one nest at one time by Chandler was seven. Overwintered queens may survive through much of the summer (see Table 1, last column). Knerer and Plateaux- Quénu (1966) report excavating a func- tional, overwintered queen from her nest as late as August 13, although in Ontario most apparently die in July (Knerer and Plateaux-Quénu, 1967a). Litte also indi- cates considerable mortality of queens during the summer in New York, for 14.3 and 43.8 percent of the nests in July and August, respectively, lacked over- wintered queens. Chandler (1955 and in Roberts, 1973) emphasizes (apparently on the basis of few observations) that if a queen dies, her full function is not taken over by one of the workers, presumably because they are unmated. Instead, they construct some cells, produce male off- spring, and the colony dies. This corre- sponds to the late production of males in Indiana; they were first seen on July 31. Possibly in Ontario, where males develop as early as late June, young females in an orphaned colony may mate, producing re- placement queens, and the colony may thus continue more or less normal de- velopment. In view of the numbers of colonies reported by Litte to lack over- wintered queens, it seems clear that in New York also, replacement queens can be produced. Litte cites two nests in which young were being produced in the ab- sence of overwintered queens. She says, however, that even when the overwin- tered queens are absent, very few daugh- ters attain fully queenlike ovarian develop- ment. Young gynes first appeared as adults in mid-August, both in Indiana and Ontario. The large brood cells were first constructed in Indiana about July 20 while the last were constructed about August 27. This seems to be slightly later than male cell construction (July 10 to August 22). In New York, Litte found a probable young gyne in a nest on August 5. The number of young gynes produced by colonies in Indiana ranged from 2 to 15, with up to 8 developing in cells of one nest at a time. The prevalence of gyne and male brood in late summer, failure to produce work- er cells in August, and failure of young gynes to contribute to the work of the colony, result in a diminution of nesting (i.e., cell construction and provisioning) activities during August. After the end of that month, the nests are occupied only by young gynes. Caste differences: There are striking differences in average size between castes, Nestinc Biotocy ANp SocrtAL OrcANIzATION oF Halictus ligatus 243 as shown by the following measurements (mm) of head widths and forewing lengths: Chandler (Indiana) Workers: head widths—2.04-2.68 (X=2.34, N=51) wing length—4.33-5.82 (X=5.14, N=50) Gynes: head widths—2.42-3.19 (X=2.68, N=48) Bie length -4.93°6.79" (X==5.88., N=48) Litte (New York) Workers: head widths—2.1-2.7 (X=2.49, SE=0.13, N=76) wing length—5.0-5.9 (X=5.50, N=76) Gynes: head widths—2.4-3.4 (X=2.91, SE=0.19, N=84) wing length—5.3-7.4 (X=6.30; N=84) Chandler does not indicate the seasons when his samples was taken; Litte’s were taken in June and July (daughters), May and June (gynes). The implication of Chandler’s work is that there is a rec- ognizable switch from worker production to production of reproductives and that most females can be recognized as to caste by size. Heads of queens seem dispro- portionately large. In Chandler’s data one can see that minimal head width of gynes is larger than mean head width of work- ers, but minimal wing length of gynes is less than mean wing length of workers. There was probably better differentiation of castes by head width than by wing length. Examining colonies separately, Chand- ler and Litte both say that the largest worker is consistently smaller than its own queen. Chandler gives data from eight colonies to demonstrate this. (In one case the wing length of a worker exceeds that of the queen, but other measurements were smaller than those of the queen.) Knerer and Atwood (1966), in Ontario, found differences in mean size between castes similar to those described by Chand- ler and Litte, on the basis of measurements of 162 queens and 137 workers. They give no statistical treatment nor data on seasons of capture, but state that workers become progressively larger during the summer and that there is no abrupt change in size of food masses as produc- tion changes from workers to young gynes. Litte implies the same, although she is careful not to jump to the conclu- sions that the words “workers” and “gynes” imply. She does describe the in- creasing size of daughters as the summer advances. Probably, there is a continuum from small females (workers) produced early in summer to large females (gynes) produced late. Kirkton, by means of smoothed curves rather than data, showed bimodality in head width measurements of females throughout the range of the species. He gives no statistical evidence for bimodality, nor do other authors. To gain further insight into seasonal size variation, head widths of 206 females collected on flowers in the eastern third of Kansas were measured. There is no evidence of bimodality in the whole set of data or for any month. Bees taken in Spring (May) and Autumn (September to November) were lumped as presumed gynes, while each summer month was examined separately, with the following results (mm): Spring, Autumn (gynes), 1.61-2.94 (X=2.58, SE=.023, N=68) June, 2.23-3.06 (X=2.58, SE=.026, N=39) July, 1.99-2.92 (X=2.53, SE=.017, N=72) August; 2.12-2°67 ((X==2:41; SE=.027,, N=27) Analysis of variance showed significant heterogeneity among these four means (P << .001). A student-Neuman-Keuls test showed the August mean to be signifi- cantly smaller than the others (P << 01), but no significant differences appear among the other three means. Figure 10 is a histogram showing the frequencies of the various size classes. The relatively large size of the July bees and small size of August bees is a surprise, since one would expect July bees to be mostly workers and August bees to be partly 244 Tue Universiry oF KaAnsAs ScIENCE BULLETIN 30 Frequency S ro) Head width (mm) Fic. 10. Histogram showing frequencies of head width classes of females from the eastern third of Kansas. Black represents bees taken in August. young gynes. Apparently the foraging bees in August, however, are mostly workers, perhaps having matured in July. Samples taken from nests, like those of the authors mentioned above, would probably have been influenced by gynes in August and smaller bees in July. In- deed, there may well be a mixture of castes in our August sample, for the mean head width of the gynes (2.58 mm) is only 7 percent greater than the mean for August bees, compared to 14.5 and 16.9 percent difference in head widths be- tween gynes and daughter workers in Chandler’s and Litte’s nest populations. The Kansas data suggest a longer sea- son of activity than in the northeastern localities, with gynes not common away from their nests until September. Perhaps in an area with a long summer season, such as Kansas, the daughter bees on flowers in July, largely reared in June, are larger than the bees produced in mid- summer. This is suggestive of the sea- sonal size variation found in Lastoglos- sum xzephyrum (Batra, 1966; Kamm, 1974) and is in contrast to areas with short summers where the first daughters produced are perhaps smallest in mean size. Gynes live as adults for about a year— August or September of one year to July or August of the next. Workers probably survive as adults for only about three weeks, according to estimates by Chand- ler; Litte observed marked workers for up to 36 days with a mean of 21 (N = 46). (Males are estimated by Chandler to survive for two to two and one half months.) Gynes hibernate, but workers do not even live until the advent of cold weather. When nesting alone, gynes commonly go through a quiescent period after provisioning their spring nests; workers are not known to have quiescent periods. After a gyne has reared her first brood of workers, she leaves the nest only occasionally and never brings back pollen for cell provisioning. Chandler re- ports that even if all workers are removed, such a queen does not revert to cell con- struction and provisioning. In the absence of the queen, workers often do not fill vacated cells with earth, construct but little, and spend long periods in the nest between provisioning flights. In the matrifilial, summer nests in In- diana, according to Chandler, the queen with her large head is the usual guard. In a much more detailed study of divi- sion of labor, Litte in New York showed that in seven out of eleven nests the prin- cipal guard for a colony was the minimal forager, therefore perhaps the queen. This result seems to corroborate Chand- ler’s general statement. In other cases, however, the principal guard was found to be an uninseminated worker. Guards bite at potential enemies, but if sufficiently stimulated, turn and close the burrow with the dorsum of the abdomen. Chandler says that the early-spring, polygynic nests are guarded as in sum- mer, but that during the monogynous period of spring cell-construction and provisioning, the nest is often not guarded. Nestinc Biotocy AND SocrtAL OrcAnizATIon oF Halictus ligatus 245 Knerer and Plateaux-Quénu (1966) and Knerer and Atwood (1967), however, say that spring nests are never guarded, even polygynous ones, a surprising observation, since Litte found spring multifoundress nests to be regularly guarded. Chandler also notes that in autumn, after the work- ers are dead, the gynes in a nest some- times guard the entrance. Litte found that in H. ligatus, as in Lasioglossum zephyrum (Brothers and Michener, 1974), there is a marked ten- dency for workers to specialize, some being principal foragers, some being prin- cipal guards, and some being less active or doing little. Guards in spring nests averaged larger and had more ovarian development than did principal foragers. Of course, a prime difference between female castes is in ovarian development. Considering this factor, Litte classified females as follows: A, 5-6 ovarioles en- larged; B, 2-4 ovarioles enlarged; C, one considerably or several slightly enlarged ovarioles; D, all ovarioles slender. The symbols + and o are used to indicate presence or absence of sperm cells in the spermatheca. All young, adult females are class D; if ovarian enlargement occurs, they join other classes. In spring and summer ovo- position periods, egg layers are in classes epor b--) In eleven spting nests each with two foundresses, Litte found that the bees ranged from A+ to D+; the foundresses (one from each nest) with most enlarged ovaries had an average Ovanianescore or, 26 (A = 3, B = 2, D = 1) while for the other foundresses (workers) the score was 1.7. In three of the nests there was little size difference between ovaries of the two foundresses. In July and August the ovaries of workers often show some enlargement. Thus, 70 daughters (workers) taken in July and 81 taken in August are dis- tributed as follows in terms of percentages of ovarian classes and fertilization: ACL, Ao. B58, Bo! .C--, Go. Dt =Do julye 4 4.3. 24) 14, OL 286 258 7600 Aucust’ 0. 1317 86 37 0" 185° 8.6 568 In New York about 13.7 percent of the daughters (months lumped) have much enlarged ovaries (classes A and B) while a larger number (23.2°%), but none of them mated, have limited ovarian en- largement (class C). Gynes mate readily, but workers are reported by Chandler to be not attractive to males. This presumably is not entirely so, for Litte’s data show 5.6 percent of July daughter females and 17.2 percent of August daughter females to be mated. Overwintering: Mated gynes hibernate in the deeper parts of the nest burrows, presumably in nests in which they ma- tured. Hibernating gynes were found by Chandler at depths of 16-24 cm. The gynes close the burrows behind them with plugs of soil a centimeter or so in length, so that the overwintering cell is a burrow termination or section somewhat longer than the bee. The upper part of the bur- row is left open and unattended, but during autumn and winter it is closed by blowing and washing soil. Warm temperate populations: Few data are available concerning Halictus ligatus in the warm temperate areas. Kirkton gives conclusions on the season of activity in all temperate areas, but as they are based on museum specimens, the information may relate to activity of collectors as much as to activity of bees. For the southern half of Florida his maps suggest that the first females appear in late January and that the last females are in the field in late December. It is quite likely that there is no period when flight activity is absent, at least in some years. Mitchell (1960) indicates that there is activity every month and that males were taken on December 18. At the Archbold Biological Station, Lake Placid, Highlands County, Florida, however, Dr. M. Litte (personal communication) found no in- 246 Tue Universiry oF KANsAs ScIENCE BULLETIN dividuals of either sex on flowers and found no nests from January to late March, 1975. This was a warm but very dry winter. Nests found in warm temperate areas, like those in Kansas, have mostly been isolated. G. Eickwort (7n itt.) suggests that humid conditions of soil may be im- portant in determining nest distribution or survival, for the nests he found in Florida at the end of the dry season were either where they received drainage from a shed roof or along the banks of a drainage canal. Also, in southern New Mexico, 51 km south of Animas, near a water hole where the soil was moist, he found a large, dense, nest-aggregation similar to those in the northern part of the range. Data on two nests from the Archbold Biological Station have kindly been made available by Drs. George C. and Kath- leen R. Eickwort. The nests, opened on April 24, 1972, were in sandy soil, nearly vertical, 86 and 170 cm deep, and did not differ in form from other deep nests of the species, such as those from Kansas (Fig. 11). The bees in this region, like those in the tropics, are large and the burrows were 6-9 mm in diameter (most- ly 65-7). One had a tumulus 7.5 cm in diameter. The two nests contained 16 and 23 cells and 4 and 7 adult females, respectively. One cell and an earth-filled old cell were found at a depth of 21 cm, but otherwise the cells were located from 48 to 160 cm in depth. They contained young of all stages, including male pupae. The Eickworts’ observations verify Chandler’s report of some elongate pollen masses, no doubt producing large bees (gynes?). They measured one round pollen mass (5 X 5 mm X 3 mm high) and two elongate ones (9 X 5 mm, 4 mm high). In each nest the largest bee was mated and had much enlarged ovaries (class A) and in the deeper nest this individual was also much worn. One or two other in- dividuals in each nest also were mated and with enlarged ovaries (our class A or B). The unmated individuals had ovaries of classes B and C, except for one in class D, and were not or were little worn. The large number of mated bees with enlarged ovaries is suggestive of the trop- ical populations. This similarity is strong- ly supported by 17 females taken on flow- ers by Dr. G. C. Eickwort at Lake Okee- chobee and 23 miles southwest of Clew- iston, Florida, April 5 and 11, 1974. All these foragers, which were dissected by Dr. Eickwort, were mated and eight had at least somewhat enlarged ovaries, prob- ably falling in our classes B and C. Of course, these bees could have all been queens, but this was not suggested by their size nor by the populations of the two nests found at the Archbold Bio- logical Station, which is in the same gen- eral area. The only two bees taken in a nest in red volcanic soil 12 miles southwest of Apache, Arizona, August 16, 1972, and dissected by Dr. G. C. Eickwort, were mated, with enlarged ovaries. This nest was only 35 cm deep and contained but one active cell in addition to old, earth- filled cells and a dead female. From these meagre data, it seems prob- able that during the long flight seasons in warm temperate areas, colonies like those of the tropics develop. Il. TROPICAL POPULATIGNS Information on tropical populations has been obtained in several areas. The bulk of the data were obtained at Curepe, Trinidad, West Indies (10°40’N) by Ben- nett and Michener. Limited amounts of information, however, were gathered at Cali, Colombia (3°30’N) and 5 km north of Chame, Panama (8°30’N) by Miche- ner, at Turrialba, Costa Rica (9°40’N) by Nestinc Brotocy AND SociAL OrcGANIZATION oF Halictus ligatus 247 rt 10 cm Fic. 11. Nest from Archbold Biological Station, Lake Placid, Florida, April 24, 1974; contained seven adult females (from field notes by G. C. and K. R. Eickwort). c Fic. 12. Nests from Mitla, Oaxaca, Mexico, July 5, 1965; contained eleven adult females (after Kirkton). For explanation of symbols, see Figures 15-17. 248 Tue University oF Kansas ScrENCE BULLETIN Drs. George C. and Kathleen R. Eick- wort (unpublished) and 9 km east of Tehuantepec (16°20’N) and at Mitla (16°50’N), both in Oaxaca, Mexico, by Kirkton (1968). The more southern of these sites approximate the southernmost known localities for the species, both in the west (Colombia) and in the east (Trinidad) (Fig. 1). Methods: To roughly monitor sea- sonal abundance on flowers and to obtain material for studies of any _ seasonal changes in bee size, ovarian development, wear, frequency of mating, and the like, “special collections” were made at intervals through three years in the general vicin- ity of the nesting site at Curepe, Trinidad. They were made by walking in mid- morning along roadside patches of Brdens pilosa, which fortunately blooms through- out the year, capturing as many Halictus as practical in a period of about an hour. For some lots bees that had collected pollen loads on the scopae were kept separate from those that did not have large scopal loads. Except in January, 1972, the special collections were made by FDB. The nests at Curepe were watched casually over a period of years by FDB. Some of these nests, as well as those at Cali, Colombia, were excavated by CDM, using in general the techniques described by Michener et al. (1955). Burrows, how- ever, were marked by blowing powder from a plastic squeeze bottle into them instead of by pouring in plaster of Paris. Before opening it, each Trinidad nest was watched for an hour or more and all returning bees captured. Departing bees were allowed to go, but captured on their return. Thus, all individuals active in the field at the time were caught and kept segregated from bees remaining in the nest, which we call “nest bees.” Of course, some of these nest bees might well leave the nest at another time. Bees re- turning with pollen loads were also sep- arated from the rest for study. The ex- pression “bees associated with nests” is used for the nest bees plus foragers taken as they returned to nests, 62 in all. For statistical purposes “foragers” are the bees taken in special collections plus those that were returning to nests, 383 in all. Bees from the nests as well as from special collections were preserved in Diet- rich’s (Kahle’s) fixative and were meas- ured and dissected later, using the tech- niques described by Michener et al. (1955). Head-width was used as the measure of size, although forewing- length was also taken. Mandibular wear was coded one (unworn) to five (worn down to subapical mandibular tooth). Alar wear was recorded as the number of nicks in the apical margin of each forewing. An index of wear for each bee was obtained by adding its mandibular code number to the average number of nicks on the two forewings; a freshly emerged bee has an index of wear of 1. Bees were regarded as mated if the sper- matheca contained sperm cells; otherwise they were considered unmated. The ovaries of each bee were sketched and maximum ovarial width and length of the longest oocyte were also recorded. Some of the drawings were sent to Dr. Marcia Litte who classified them accord- ing to the system summarized in Part I above. Not all ovarioles were visible in the drawings, but she was confident of the categories. We are thus able to use, in addition to the measurements, lettered categories equivalent to hers. Nest sites: The site in Trinidad, which was active for at least six years, was near houses on a hill just north of Curepe. The nests were in soil consisting of decom- posed shale near the surface that could be cut with a knife, the soil being harder and with solid rocky regions at depths of 70 to 90 cm. The site was almost level, covered with coarse vegetation that was usually clipped short. The nests survived Nestinc Biotocy AND SocrtAL OrGANIZATION OF Halictus ligatus 249 a period of two months when the uncut grasses and weeds grew to a height of about 30 cm, but disappeared ultimately, possibly because vegetation was allowed to grow for a longer period. Six nests were found in an area of about one square meter. There were a few scattered nests nearby and there must have been many in the general area, for the bees were com- mon on flowers of Bidens pilosa and Me- lanthera nivea, both weedy Compositae. Pollen loads from each plant (distinguish- able by color) which were carried into the nests indicated that Bidens was more often visited than Melanthera. Nests were excavated in late December, 1971, and Jan- uary, 1972; otherwise the population was kept track of by monthly samples from the Bidens flowers. Three isolated nests near Turrialba, Costa Rica, were also in level ground with grasses and herbs. Near Cali, Colombia, and Chame, Pan- ama, nests were found in the very hard, bare soil of foot paths. The soil surface was flat or gently sloping. Most of the nests were isolated although three were about 20 cm apart. Kirkton’s three nests from Oaxaca were all isolated. At no tropical site have we encoun- tered dense nest-aggregations like those in Indiana, New York, and Ontario. Dr. Al- varo Wille of the Universidad de Costa Rica, however, has told us of a site near San Mateo, Costa Rica, inhabited by a moderately dense aggregation. Seasonal activity: This section is based entirely on data from Trinidad, although seasonal activity at other tropical locations is presumably similar. Nests excavated in January in Colombia appeared to be in the same condition as those excavated at the same time of year in Trinidad and, in fact, nests excavated in July in Costa Rica and Oaxaca were also similar. To understand the bee’s seasonal ac- tivity, it is necessary to have information on the seasonal cycle itself. Weather data are from the University of the West In- dies, St. Augustine, about 3 km from the nesting site. Temperatures are warm and rather uniform throughout the year. Monthly mean minima during the years 1972-1975 ranged from 19° to 23° C (67° to 74° F), with a tendency for lower mean minima during the Northern Hempishere winter and spring months. Monthly mean maxima ranged from 29° to 33° C (84° to 91° F), again with a tendency for lower means in the Northern Hemisphere winter and spring months. Rainfall occurs in every month of the year with a total annual rainfall of 1200 to 1475 mm (48 to 59 inches). The dry season, with 3 to 75 mm (0.12 to 3 inches) of rainfall per month, usually extends from January through May, in some years beginning in December or extending through June. Sometimes (e.g. 1972) January and May are wet so that the dry season is only February through April. The wet season, with a monthly rainfall of about 100 to 300 mm (4 to 12 inches), occupies the rest of the year. Data on seasonal activity in Trinidad are based largely on the special collections made from flowers at Curepe, but, in part, on dates of capture of specimens in the collection of the Commonwealth In- stitute of Biological Control in Curepe. These collections were somewhat irregu- larly distributed, but, over three years, one or more special collection was made or at- tempted in each month except November. Halictus ligatus has been taken on flowers at Curepe in every month except May and November. In November, al- though no special collections were at- tempted, there is good reason (abundance in late October and early December) to believe that activity continues uninter- ruptedly. In most months, a_ special collection, such as is described above, re- sulted in capture of 25 to 38 female Ha- lictus. Three different collections in April, however, produced 5, 1, and 4 bees each. In May, none were taken and in 250 Tue Universiry oF KAnsAs ScrENCE BULLETIN June only two. The collection of the Commonwealth Institute of Biological Control contains 13 taken in April, none in May and two in June. From these data, it seems that there is a period at the end of the dry season and perhaps continuing into the begin- ning of the wet when the species is scarce on flowers. There is no obviously com- peting flower that might account for its disappearance from Bidens at that time. The special collections in April to June were all made or attempted in 1974, a year in which the dry season continued through June, the rainfall that month being only 2.33 inches. In July, there is the same mixture of fresh and worn bees as in any other month. Thus, if adults of mixed ages be- come relatively inactive in April to June, they probably emerge and resume work after the rainy season begins. It is pos- sible that after a long dry season the soil is too dry to make cells and the bees be- come inactive until it is workable again. Males have been taken at Curepe in January, March, July, August, September and December and probably are active whenever females are to be found. Large females, suggestive of north temperate gynes, but even larger (work- ers, also, are larger in the tropics than northern workers), occur in the tropics. As explained later, however, such indi- viduals are not always gynes. In Panama and Oaxaca three such individuals were found, apparently starting new nests, in June and July. By itself, this is not evi- dence that nest establishment is highly seasonal, but it might occur predominantly at the beginning of the wet season when activity resumes. Colonies in large nests can also be found in July (Costa Rica, Oaxaca), suggesting that they do not break down in the inactive season as do colonies in the cool temperate autumn. The months of capture on flowers of the largest 10 percent of the females taken in Trinidad are January, March, April, August, Sep- tember, October, December. Thus, large females can probably be found on flowers whenever other females are there. The largest female taken at Curepe, markedly larger than any other, was on flowers in October. Fresh and unworn, it could not have started a nest in June or July. Nest structure and development: Three apparently new nests, shallow burrows each occupied by one large female with no cells or with a cell or two roughed out, have been found by CDM in Panama in early July and by Kirkton in Oaxaca in June. If new nests are started primarily by large, unworn females, as seems pos- sible in view of the situation in a tem- perate climate, it may be significant that essentially unworn, large females (head width over 3.5 mm) have been taken in Trinidad in January, March, July, Sep- tember, and October, four of them in the latter month. New nests were not found in Trinidad, despite a search in December and January and general surveillance throughout various years. The nests studied in Trinidad and Co- lombia in December and January are the bases for the following account, with par- enthetical comments for other regions. The nests were much larger than those in the northeastern parts of the range. Tu- muli of fresh soil up to 9 cm in diameter (Fig. 13) were found, showing extensive Fic. 13. Nest entrance at Curepe, Trinidad, showing large, fresh tumulus. The numbers represent inches (@5"cm) = inch): Nestinc Biotocy AND SoctAL OreAnizAtion oF Halictus ligatus 25 digging in some nests in December and January. Old tumular material, packed and washed by rain and sometimes with grass growing in it, but recognizable by the color of the subsoil, was also evident around many of the nests. Measurements of entrances and bur- rows are slightly larger than for those in Kansas, entrances being 3.5 to 5 mm in diameter, burrows 5.5 to 6 mm. Nothing suggestive of a turret was seen at any time, nor were nest-entrances ever seen closed by soil, except sometimes by fresh tumulus material being pushed out. There was no evidence of regular nocturnal closure, as is reported in the north. Although part-time guarding occurred early in the morning, and the sun was often on the nest sites, flights from the nests did not begin until about 9:30 a.m. Rain or threatened rain usually stopped activity in the early after- noon. While the nests were much deeper than in the north, the number of branches was not greater (Figs. 12, 15-17), the mean for six nests being 4.4 and the maximum being 6. The distribution of earth-filled, aban- doned cells showed that there had been other branches at depths from 15 to 60 cm, but they had evidently been filled and abandoned at earlier stages in nest-devel- opment. Presumably, the number of branches remains more or less constant, old ones being filled as new ones are made. The deepest nest (Trinidad) was 97 cm deep, the shallowest (Colombia), 79 cm (Costa Rica, 55 cm). The burrow walls are often lined 0.5 to 1.0 mm thick with soil from deep in the nest (as shown by color), bearing the marks made by tamp- ing with the apex of the abdomen. Chandler’s report of a consolidated soil lining is thus verified (as also in Arizona and Florida nests excavated by the Eick- worts). The tendency of burrows to slant, often in a single direction, is prominent in Trinidad (and Costa Rica) (Figs. 15- 17). The bottoms of some nests in Trini- dad were offset 60 cm or more horizontally from the entrance. In Colombia and Oa- xaca, the burrows were nearly vertical. Many earth-filled, old cells, mostly with feces, indicating probable survival and emergence of the occupants, showed that the colonies had been active for consider- able periods, months and possibly years, prior to our digging of them. In Trinidad the shallowest earth-filled cells found were 18 cm deep; in Colombia, 15 cm. This may be the depth at which nest foundresses make cells, but, if so, the nests were per- haps established by multiple foundresses, as there were many old cells at these depths. The uppermost, occupied cells found were 27 cm deep, both in Trinidad and Colombia (9 cm in Costa Rica); the deepest were 95 cm (19 in Costa Rica). They were often intermixed with earth- filled cells and young of different ages were often intermixed. Nonetheless, in general, eggs and younger brood were deeper in the nests than pupae. In six Trinidad nests with a total of 161 oc- cupied cells, 82 percent of the 61 cells less than 70 cm in depth contained pupae and only 7 percent contained eggs or feeding larvae. For the 100 cells over 70 cm deep, the figures are 51 percent for pupae and 37 percent for eggs and feeding larvae. The tendency of the tropical nests to be very deep, despite humid climate, is interesting, since in Europe and North America (Michener and Wille, 1961) halictid nests are deeper in dry than in humid seasons. Perhaps soil temperature is also important in influencing nest-depth. It is also true that the heat of the tropical sun tends to dry the soil more in a given time than the same period of sunshine in the north. In both Colombia and Trini- dad, the surface soil seemed rather dry much of the time in January, despite fre- quent rains, All cells (Fig. 14) found in the tropics correspond to Chandler’s male-and-work- ey) Tue Universiry oF Kansas SctENcE BULLETIN Fic. 14. Cell from Curepe, Trinidad. Black represents larval feces. Fics. 15-17. Nests from Curepe, Trinidad, December, 1971, and January, 1972; contained 14, 16, and 12 adult females, respectively. The symbols used in these and other illustrations of nests are as follows: Dotted cell = earth-filled, abandoned cell; m = moldy; P = pollen in open cell, being provisioned; € = empty, adult recently emerged; E = egg; SL = small larva; ML = medium-sized larva; PP = large larva, usually prepupa; ¢ 2 = pupae of the sex indicated. Nestinc Biotocy AND SoctAL OrcAnizATion oF Halictus ligatus 253 er-producing cells. The southern bees are somewhat larger, hence measurements are in general a little longer, as follows: Cell length (including the neck) 12 to 20 mm (Chandler’s measurements apparently did not include the neck); diameter 6 to 7 mm; diameter of neck 35 mm. The cell- closure was usually about 1 mm thick and most cells were 14 to 15 mm long including the neck or closure. Sometimes there was a small irregular hole in the closure. This might result if some of the loose soil fell out, but did not seem to be a result of the investigator’s work. The number of occupied cells per nest from six nests in Trinidad ranged from 18 to 44 (X = 34) (11 in Costa Rica). Earth-filled cells were at least as numer- ous and probably more so; counts were difficult to make and were not seriously attempted. This mean is enormously greater than means for Litte’s nests in New York, but the maxima are in the same range. Food-masses were shaped like those in Chandler’s worker-and-male-producing cells, but were slightly larger, 4.5 X 4.5 xX 2.8 mm, the last being the vertical measurement. Colony size: Colonies in the tropics are markedly larger than those in the cool temperate regions and those exca- vated were not obviously eusocial, but seemed more nearly communal (terminol- ogy of Michener, 1974). Eleven tropical colonies have been partially or fully examined. The smallest which was ex- cavated was from Costa Rica, with only four females (one from Trinidad had only one foraging bee and at least one other bee in the nest, but was not exca- vated). The largest colony fully studied was from Trinidad; it contained 16 fe- males. Another, from Colombia, con- tained 11 foraging bees and an estimated 21 in total. In six fully-excavated, tropical nests, three opened in June and July in Costa Rica and Oaxaca, three in Decem- ber and January in Trinidad, the mean number of females was 11.3. Immature stages: The nests studied in Trinidad in December and January, Colombia in January, Costa Rica in July, and Oaxaca in July contained immatures of all stages. The climatic regimes in these areas are reasonably similar and pro- duction of young of both sexes is presum- ably continuous with the probable excep- tion of the brief period when adults dis- appear from flowers (in Trinidad). Guards: In Trinidad and Colombia guards were sometimes seen at the nest entrances. They are inconspicuous com- pared to those in northern, summer colonies and in many cases an apparent guard may be only a bee about to leave the nest. Usually, they retreat when dis- turbed, but one turned and blocked the nest entrance with her abdomen. Four guards were collected and subsequently measured and dissected. From this small sample no suggestion of consistent size or internal differences between guards and foragers was detected. The guards were of moderate size, little to considerably worn, mated or unmated, with varying amounts of ovarian enlargement. Bees associated with nests: Data are available on 62 bees associated with nests in Trinidad. Except for one, possible nest- foundress and one or two possible, potential foundresses (large young bees), this sample is a reasonable tropical counterpart for Litte’s sample of daughter bees from nests in New York. Data are presented in Table 2. The following features are worthy of comment: (a) Frequency of mated bees, amounting to 58.1 percent of all bees. (b) Existence of some worn, unmated bees. (c) Absence of bees with ovaries of class A (fully enlarged). (d) Abundance of bees with moderately enlarged ovaries (class B), amounting to 51.6 percent of the individuals, mostly mated. (e) Lesser abundance of individuals in ovarian class C, amounting to 38.7 percent, mostly un- 254 Tue Universiry oF Kansas ScrENCE BULLETIN mated. (f) Scarcity of individuals with slender ovaries (class D), amounting to only 9.7 percent, mostly unworn and all unmated. The contrast with Litte’s data is striking, as noted in the Discussion be- low. The mean size of bees associated with nests did not differ significantly from that of foragers. For the few Trinidad nests which were completely excavated with all in- habitants captured, 47.5 percent of the 40 females were taken at the nest-entrances, returning with pollen. Pollen-collecting in- dividuals have approximately the same mean head-width and variance thereof as do all females from Trinidad. Of the pollen-collectors, 63.8 percent were ferti- lized. They ranged from unworn to the most worn of all the individuals taken at the nests; nearly half had indices of wear less than 5. Nearly all showed some ovar- ian enlargement (classes B or C), al- though 3.2 percent of the pollen-collec- tors have ovaries placed in class D. Among pollen-collectors, 23.7 percent had an oocyte 2.0 mm long or longer, often ap- parently nearly ready to lay; such bees were all fertilized. Not a single nest-bee, including the most queenlike individual, contained an oocyte approaching 2 mm in length. Thus, it seems that the foragers are the principal reproductives, as in soli- tary bees. It is clear, however, that not all the bees mate and lay eggs. Of the bees asso- ciated with nests, about one third had an index of wear of 5 or more and were therefore presumably the older individuals; of these 33.3 percent were not fertilized and had ovaries of classes C or D. Thus, about one sixth of the bees are apparently non- reproductive workers. Such individuals are not always small bees, but at least among foragers there was a significant, positive relationship among ovarian-width, mating, and head-width, suggesting that larger bees are more likely to become reproduc- tive than smaller ones. (For details see material below on Foragers.) Of the nest-bees (taken in the nests), over half were unmated, mostly with in- dices of wear of 1 or 2. Although such bees are presumably young adults that have foraged little if at all, some showed some ovarian enlargement (class C). Nest-bees showed a size-range and mean size essentially the same as those taken (nearly all carrying pollen) as they re- turned to their nests and as foragers in general. Foragers: To elucidate the attributes of foraging bees, all the females taken in Trinidad in the special collections from flowers were measured and _ dissected. Nearly all these bees were collecting pollen. The few that had little or no pollen on the scopa were in no other way different from those that were collecting pollen. Hence, all are termed foragers and were analyzed together with the foragers caught returning to nests with pollen, making a total of 383 females. Head-widths of the specimens from Curepe, Trinidad, in the collection of the Commonwealth Institute of Biological Control were also measured, thus providing size data on 147 additional females. The most noteworthy attribute of the foragers is the high percentage that have enlarged ovaries and are mated. Data on these features are shown in Table 2, and support, with some modifications, the com- ments on bees associated with nests, as follows: (a) Mated foragers, 55.6 percent of total, only slightly more numerous among worn than among little worn and presumably younger bees. (b) Existence of worn, unmated bees. (c) Small num- bers of bees with ovaries in class A. (d) Abundance of bees with moderately en- larged (class B) ovaries, totaling 49.9 per- cent, the great majority mated. (e) Abun- dance of bees with less developed ovaries (class C), amounting to 45.4 percent, most- ly unmated. (f) Scarcity of individuals Nestinc Biotocy ANp SoctAL OrcAnizaAtion oF Halictus ligatus 255 Tasce 2. Percentages of insemination (+, 0) and ovarian classes (A-D) in bees from Curepe, Trinidad. Index of Wear A-+ Ao B+ Bo C+ Co D+ Do N Associated with nests -......... 5 —~ 66.6 a ee Se 28.6 4.8 21 [ROna eho weee ee < 530. x 100. Fic. 7. Cells composing the sporangial walls seen Fic. 8. Section through a sporangium showing its in face view (tangential) showing their sinuous out- attachment to the abaxial surface by a short stalk lines. X 145. (arrow). X 180. “ft ‘ % pat + oa '? . 7 ~ ‘4 \ 7 rom. ~~ # _ ae Fic. 9. Portion of distal sporangial wall showing the Fic. 10. Single bean-shaped spore showing the highly elongate, thick-walled cells strongly suggestive of a distinctive pattern of exine sculpturing. Note that rudimentary annulus. 750. some ridges bifurcate. & 890. 288 Tue Unrversiry oF Kansas ScrENCE BULLETIN rangium. The spores are either monolete or alete and possess a distinctively sculp- tured exine. This exine-sculpturing is highly-consistent, being composed — of broad, occasionally bifurcated and anasto- mosed ridges or muri, each typically 2.0- 35 wm wide, running parallel to one another along the primary axis of the spore. Ridges are typically 1.5-2.0 »m high and from 4.0-6.5 »m apart. The exine of the spore is about 2.5-3.0 »m in thickness. DISCUSSION There is little doubt that the fertile- fern pinnules described here are assignabie to the leptosporangiate fern-family, Schiz- aeaceae. This family has solitary sporangia not grouped into sori. Also, the sporangia are borne abaxially in two regular rows on either side of the midrib (in Schizaea and Anemia) and are often protected by an incurling of the pinnule margin. The spo- rangia are typically large and may be shortly-stalked or sessile. The first-formed tracheids are finely scalariform. The ring of thick-walled cells presumed to be an inconspicuous annulus is distinctly schiz- aeaceous and the highly-distinctive spore ornamentation closely resembles some spe- cies of the extant and fossil genus Schizaea. When considered together, the pinnule morphology, the sporangial form and ar- rangement, the nature of the annulus and the distinctive sculpturing of the spore- exine unquestionably place this fertile fern in the family Schizaeaceae. The only genus of schizaeaceous fern previously described from the Carbonifer- ous is Senftenbergia (Radforth, 1938, 1939), and is the only other member of the family Schizaeaceae of approximately equivalent age to that in the Middle Pennsylvanian coal-ball specimens de- scribed in this report. Boureau (1970) lists 5 species of Senftenbergia, several of which have morphological features similar to those described for our coal-ball speci- mens. The sporangia of Senftenbergia ophiodermatica are borne abaxially on the pinnule in two rows, one on either side of the midrib, arranged with their long axes parallel; each sporangium is __ shortly stalked. The pinnule margins, with their associated sporangia, are incurled toward the midvein, often giving the appearance of a crowded mass of sporangia. In both S. ophiodermatica and S. sturi the annuli are indistinct, the former species having been originally described by Kidston (1924) as exannulate. The sporangium of S. sturi is similar in form to that of our coal-ball specimens in being elongate, spindle- shaped, and shortly-stalked at the basal end. The spores of such species as S. pen- naeformis, however, are basically spherical and have, on the irregular system of anas- tomosing ridges, stout peg-like projec- tions. Also, all species of Senftenbergia produce spores with an obvious, triradiate commissure. These spore features are un- like those described for our coal-ball speci- mens, eliminating them from inclusion in the genus Senftenbergia. Had the sporangia and highly-distine- tive spores described in this report been associated with Upper Mesozoic or young- er strata, a prompt and confident assign- ment could have been made to the genus Schizaea. Selling (1944) has reviewed sev- eral fossil, schizaeaceous ferns and without qualification referred two to Schizaea (S. miocenica and S. skottsbergu). Cookson (1957) described S. digitatoides from Plio- cene-aged coal in New Guinea, and Lu- kose (1964) later added S. kashmuriensts from the Upper Mesozoic of India. The spores of S. confusa, described by Selling (1947) from specimens growing in Mada- gascar, are virtually identical with those described here from Middle Pennsylva- nian-aged material. Cookson (1957) noted that such distinctively-ridged exospore sculpturing, considered along with a bi- FERTILE SCHIZAEACEOUS FERN FROM MippLE PENNsYLVANIAN Iowa CoA BALLS 289 lateral, monolete form, “provides a reliable basis” for assigning ferns with such spores to the genus Schizaea. Lukose (1964) is even more convinced of this, stating that “From the available knowledge of the morphology of the recent and fossil Schiz- aeaceae spores, bilateral bean-shaped, stri- ated, monolete spores are met with only in the genus Schizaea.” We note, too, that Arnold (1947) considers Schizaea and Senftenbergia closely related. We believe that the botanical afhinity of the fertile-fern specimens here described clearly lies with the Schizaeaceae. No other group of ferns, fossil or extant, re- sembles those specimens so closely as does the family Schizaeaceae. The available material strongly supports classification of this fern to the Schizaeaceae. This paper is believed to represent the first report of the remains of a fertile schizaeaceous fern from the North American coal-balls, estab- lishing the presence and describing the morphology of distinctive, large, ridged, monolete spores from Middle Pennsyl- vanian-aged deposits. Pi Eee LUREY CilLED Anprews, H. N. 1943. Contributions to our knowl- edge of American Carboniferous floras. VI. Certain Filicinean fructifications. Ann. Mo. Bot. Garden. 30:429-443. Boureau, Epovarp, ed. 1970. Traité de Paleobota- nique. Tome IV (Filicophyta) :257-284. Mas- son et C'® Editeurs, Paris. CHANDLER, M. E. J. 1955. The Schizaeaceae of the south of England in early Tertiary times. Bull. British Museum (Natural History). 2(7): 291-314. Cookson, I. C. 1957. On some Australian Tertiary spores and pollen grains that extend the geo- logical and geographical distribution of the living genera. Proc. Royal Soc. Victoria. 69: 41-53. Harris, T. M. 1945. Notes on the Jurassic flora of Yorkshire. 19. Klukia exilis (Phillips) Raci- borski. Ann. Mag. Nat. Hist. ser. 11, 12:257- 265. . 1961. The Yorkshire Jurassic flora. Part I. Thallophytes and Pteridophytes. British Mu- seum (Natural History) :128. Joy, K. W., A. J. WixLs, anp W. S. Lacey. 1956. A rapid cellulose peel technique in paleobotany. Ann. Bot. (new ser.) 20(80) :635-637. Kinston, R. 1924. Fossil plants of the Carboniferous rocks of Great Britain. Great Britain Geol. Surv. Mem. 2(5):379-522. LuxosE, N. G. 1964. New species of Schizaea spore from the Upper Mesozoic of Kashmir, India. Nature. 204(4958) :566-567. Raprortu, N. W. 1938. An analysis and compari- son of the structural features of Dactylotheca plumosa Artis sp. and Senftenbergia ophio- dermatica Goeppert sp. Trans. Roy. Soc. Edin. 59(14) :385-396. ——. 1939. Further contributions to our knowl- edge of the fossil Schizaeaceae; genus Senften- bergia Trans. Roy. Soc. Edin. 59(27) :745-761. SELLING, O. 1944. Studies in the recent and fossil species of Schizaea, with particular reference to their spore characters. Meddel. fr. Gote- borgs. Bot. Tradgard. 16:1-112. il ML vil SE 02183 Pe ee Satis ty ; ¥ AMANe $B ee a A, ye RR RR OR Ry renee eocetetecetecececestarara*anata*atena*etetata'ehs"a%e*s"e" = soe 'e See eee See Px ote’ ss we ie ott ee ote! tee ote. = oS fee lee get Reo Se a oS OS eet oe e net oe. mR = oS ef of ‘eve tet fone ee PE od °° eet ae ee 525 ae eee ate me ate oo re o Se ete oe tone set se = SS ss ‘eve set Ye ‘e's Nee et tege eee Se ‘e Sa we "em 88, et eee be . On < %, om oy bod 5] fate I ¥ se ss Se ee we RSs RR seeees eeseatetecatetotetocate esereleeececeteets * ae ey *. Ss BS ist “owe RX ue. RI se me essence eceeneeneseseenen Sa ate teeta ERA aeaeceaeteeeeateeeeceeteeteeteettatteeteet sieSeseteceseteceseteSesetecesesecosearceascrceaectaratetatatetacateronatecatetasatetecetetateteratetetatetetetecensterenaterenenet THE UNIVERSITY OF KANSAS > SCIENCE BULLETIN cata grr : f hae : ; ——E——E—— : ' jl 1 hij a ay ft - \ y yt VV UUUS riOi’, ty FP sactencenn erences st Ss names mate me THE PARASITIC GROUPS OF HALICTIDAE (HYMENOPTERA, APOIDEA) By Charles D. Michener Vol. 51, No. 10, pp. 291-339 January 27, 1978 ANNOUNCEMENT The University of Kansas Science Bulletin (continuation of the Kansas Uni- versity Quarterly) is an outlet for scholarly scientific investigations carried out at the University of Kansas or by University faculty and students. Since its incep- tion, volumes of the Bulletin have been variously issued as single bound volumes, as two or three multi-paper parts or as series of individual papers. Issuance is at irregular intervals, with each volume prior to volume 50 approximately 1000 pages in length. Effective with volume 50, page size has been enlarged, reducing the length of each volume to about 750 pages. The supply of all volumes of the Kansas University Quarterly is now ex- hausted. However, most volumes of the University of Kansas Science Bulletin are stil! available and are offered, in exchange for similar publications, to learned societies, colleges and universities and other institutions, or may be purchased at $20.00 per volume. Where some of these volumes were issued in parts, individual parts are priced at the rate of 2 cents per page. Current policy, initiated with volume 46, is to issue individual papers as published. Such separata may be purchased individually at the rate of 3 cents per page, with a minimum charge of $1.00 per separate. Subscriptions for forthcoming volumes may be entered at the rate of $20.00 per volume. All communications regarding exchanges, sales and subscriptions should be addressed to the ExcHANcE LipraRIAN, UNIVERSITY oF Kansas Lipraries, LAWRENCE, Kansas 66045. Reprints of individual papers for personal use by investigators are available gratis for most recent and many older issues of the Bulletin. Such requests should be directed to the author. The International Standard Serial Number of this publication is US ISSN 0022-8850. Editor Eugene C. Bovee Editorial Board William L. Bloom Philip W. Hedrick Rudolf Jander Harvey Lillywhite Charles D. Michener Norman A. Slade Henry D. Stone George W. Byers, Chairman THE UNIVERSITY OF KANSAS SCIENCE BULLETIN Vol. 51, No. 10, pp. 291-339 January 27, 1978 The Parasitic Groups of Halictidae’ (Hymenoptera, Apoidea) Cuar_es D. MICHENER Departments of Entomology and of Systematics and Ecology, and Snow Entomological Museum, University of Kansas TABLE OF CONTENTS PENNE SS TERA Cs aa oe on ey es ARE nr ese ra TE 2 2 OT oe sete et ve aes A 292 TISSUE UST ROIN < aceasta ee ee 292 (SGONVERGENTEREAT URES nee eee ee ee ee ee NE eS St RS POP aN Ee 292 REDUCTION AND Loss OF STRUCTURES ...... Se ae ee ee ee Cee 8 eR A eB ee ee ee ee 293 EGAN GENCENT-UCAND BSPEGCIALIZATION ROBO TRU CHU RIES o.<:oee ene ere oe as ee eee noe oe ea oe Nees Soe cee ce Sete zeeee 302 SIMILARITY OF PARASITIC FEMALES TO MALES ................... SAS reget re 8 poate eh sen NOU eel RN ae Ste 303 REMAUKVER DIVERSUGY, (OF PPARASITIC "AND INONPARASITIG FIARICTINAE: 2 ..22c22-c-c-202-c2.-2accssees ee cc beset ee eee nese eee 303 TRLSGIF LSB NTO INI SE Te AS, ee aN es se eS ll SRE re a ec Pi eee ae Eee OE 304 Seen CATS ADRS ACT VEN UREN fe ay a Ee See ra pS oe AEE Se monn a en ee ed 308 (Gams, PATTI he os oo a tee aa Na at a ee OD te ar ae ee ee 308 \GenTESWPLICRL NT CICS E ene eee en Ne OE MNS es, 2 AS, oe LEO he Ue y POE NU SS 310 FMS TAA OA TAT OID LEE si he RRS ee REM OE RE ON oe eae EN me Re 311 REC TUES II RCTI177 OS OUIs (Lame ret Oe re A er epee arch Oe we Rv ce 313 PEERS P HE CODEN GROW Ditete ees ene irene eee war: teeta ENE oA TT) ed dence et eee a Nees 318 (Canis VACA IN (Gan, iss) Sede) ee Minne ele a ee) ene ee See ee ene ee 315 Genus @Witcrosphecodes. . se ee ee Ee Sie eee ON rr Ate ee nl et ee 317 (Gamns VATU Bis ko Senet a lay et RE pete to ee OR ROI ee en OPPS 319 SnD CEILISI 4 POLE 1.5112 amen ee are mr EE OE SID OS a en I toe sae ee Swe ee 322 STD CIISM NCS CEU P CLE. d Gimeerene mea nee eae. eed at VOPR a RN ee ee ei 2 eee 322 Canis GWAC wespscepre ttn ee EN et, a ere Od Red Sire i 6 Sak ea) Se ee ee Ba PRE Oe 322 Sub enusieS P2CcOd Comte ame ete Smet a a te. oe be er ese Re Ne RAL eae une Leeann Ie Se 325 SUDSERUSEeATLOSPRCCOUCSMNeRS UD one eee ten Ne ies ee Poe Se et ee ee ed 327 APPENDIX ....... See ees ae ene ee Ne ree a et oe Sees Uhl) vee a 2 327 PUL OLNTATIGOST O97) LESCIVT 1 C1 Gee ee IC ney Ie Ea cep eee Se B27. EALGLUTATICOSL OM URCIOR LTA eee iee Te ae ie eo BSR ee ees fe eee a 328 IN OT LACT DCIS! sapere ae Ce Ee nS oR Rh rely w, SON IPE CAO OEE a a er es | 328 Genus EIOCle piss en ee are te kL Degen 2 ES es 328 IKSEVETO SPECIES IOP Se CIMOCIED TTS ieee 2 he see Ie ete ooo .ds coat ecaeeeestkches a oe 328 Piloclepus, tomentosa aspy 2 es ee pe ee en ee I 328 IPE OClEplise POLY IONE SAN asp epee eee et eek ee ne eee 330 PilocleptisNercRWOTL Teasp-y ie Ee. ero eA Bass oe 331 Eupetersia (Eupetersia) guillarmodi n, sp. .........----..------------------ pith ee ele. a ee 332 Eupetersia (Neoeupetersia) flava n. sp. Ee ee ee Orne 334 Sphecodes (Austrosphecodes) convergens Nn. Sp. .----.-------------------- Sihete ot oS PoE EAN et RAs oe 335 IAGKNOWICED GEMIE NGM Race 2 henner fee an oe ae Bel ete oe eS a hen Ne See ee 337 Ie siereru timers! Ghia aay ess eae aaa SE A ER I Rene ee ee ee ae ee ee te ee ee ee es 337 » Contribution number 1611 from the Department of Entomology, The University of Kansas, Lawrence, Kan- sas 66045, USA. 292 Tue UNIVERSITY OF KANsAsS SCIENCE BULLETIN ABSTRACT Eight parasitic genera of halictid bees are recognized, derived from five different nonpara- sitic ancestral groups. The parasites vary from those, like Paralictus, which are similar to their nonparasitic relatives and are presumably recent derivatives of them, to others, like Sphecodes, which appear to be ancient parasitic groups derived from nonparasitic ancestral genera that quite likely no longer exist. Convergent morphological features among parasites include re- duction or loss of structures for pollen collecting and carrying and of structures used in nest construction. A review of host relationships is provided, as is speculation on the relationships between type of parasitization and structure within the genus Sphecodes. In the taxonomic treatment, a new genus (Pz#ilocleptis) and three new species (P. tomen- tosa, polybioides, and eickworti) are described, as are Sphecodes subgenus Austrosphecodes and the species S. (4.) convergens and Eupetersia guillarmodi and flava. INTRODUCTION Among most families of bees there are genera or species that do not make or pro- vision their own nests, but put their eggs in, or actually live in, the nests of other species of bees. Larvae of such social para- sites or cleptoparasites do not feed on the hosts, but eat the provisions supplied by the host. In some cases, adult hosts are killed by the female parasite, but in most cases only the eggs or young larvae in cells are killed, either by the adult female para- site or by the young larvae of the parasite. In the large bee family Halictidae, para- sitic forms have arisen at least five times from different nest-making ancestors. The subjects of this paper are the parallel evo- lution that has occurred among. these forms, their host relationships, and their taxonomic groupings. Although some of them have not yet been reared from nests of their hosts, para- sitic halictids can be recognized by the absence of pollen-gathering and _pollen- carrying structures in the females. There are parasitic species of other families of bees that have not lost pollen-handling equipment (Bombus, Braunsapis; see Michener, 1974), but no such forms are known in Halictidae, although individuals of nest-making species sometimes usurp nests of the same or other species (Knerer and Plateaux-Quénu, 1967). The parasitic halictid genera and their antecedents are listed in Table 1. Charac- ters supporting the judgments as to which are the ancestral groups are indicated in the taxonomic treatment. Convergent fea- tures characteristic of the parasitic groups are listed below and in Table 2. CONVERGENT FEATURES Parasitic halictids are characterized by numerous similarities. These are among the principal characteristics by which they differ from the nonparasitic or nest-mak- ing genera from which they arose. Many of these features involve reduction or loss of pollen-gathering, pollen-carrying, and nest-making structures, while others ap- pear to be defensive and offensive equip- ment for encounters with nest owners. In either case, the structures are those of fe- males, and the rest of this section pertains only to female characters. Some of these characters, such as the reduced hairiness, relatively coarse punctation, and sometimes even spine-like setae on the hind tibiae, occur in males of some species also. Their functional significance is probably limited to females, however; presumably the de- Tue Parasitic Groups or Harictipar (HyMENopTERA, AporDEA) 293 velopmental mechanisms happen to op- erate in some males. In this section each feature which seems to be an adaptation of parasites to their way of life and which appears to have arisen independently in different parasitic groups, is lettered a to s. Each of these features is scored on a scale 0 to 1, 0 repre- senting the structure as found in _nest- making ancestral halictids, 1 representing the extreme modification found among parasites. The intermediate conditions are subjectively assigned scores such as 0.2 or 0.6 to give an idea of the morphological degree of difference from the extremes, 0 and 1. Scores for the parasitic genera are shown in Table 2 and many of them are indicated in parentheses in the following account. REDUCTION AND Loss oF STRUCTURES OF Nest-Makinec Hatictips: (a) The body and legs of nest-making Halictidae are usually hairy, most of the longer hairs and many short ones being plumose (Fig. 1). This extensive vestiture presumably serves to capture loose pollen, which can later be brushed together and transferred to the scopa by modified grooming movements (Jander, 1976). Comparison of the right sides of Figures 1 and 2 shows the difference in vestiture of a typical host and a highly modified parasite. Paralictus (0) has the hairiness of its presumed ancestor; all other para- sites (1) are markedly less hairy than their presumed ancestors. In Pulocleptis and some species of Austrosphecodes a dense covering of short plumose hairs occurs on certain parts of the body, but this is very different from the longer, more or less erect hairs of nest-making halictids. (b) The front basitarsus of most pollen collecting halictids is somewhat flattened with a sharp ridge or carina along the pos- terior (i.e., outer) margin of the segment, the distal half or more of this ridge giv- ing rise to a row of closely placed setae forming a comb, the anterior basitarsal comb (the anterior basitarsal brush of Eickwort, 1969b). The entire segment ex- cept for the strigilar area is covered with a brush of hairs which are usually longer than those of the comb, so that the latter may be inconspicuous. The hairs of the brush are usually rather dense but are sparse in small forms like Homalictus and Lasioglossum (Dialictus). The brush and the comb contained within it appear to be important in removing pollen from anthers Taste 1. Parasitic Halictidae and the Groups from which they Presumably Arose. PARASITES 1. Paralictus 2. Echthralictus 3. Parathrincostoma 4. Temnosoma Augochlorini The Sphecodes Group: of Genera 5. Ptilocleptis 6. Microsphecodes 7. Eupetersia 7a. Eupetersia s. str. 7b. Nesoeupetersia 8. Sphecodes 8a. Sphecodes s. str. 8b. Austrosphecodes PROBABLE ANCESTORS Lasioglossum (Dialictus) Homalictus s. str. Thrinchostoma Sphecodes (Austrosphecodes) Sphecodes (Austrosphecodes) Sphecodes s. str. DIsTRIBUTION North America Samoa Madagascar Neotropical Halictus-Lasioglossum Group Neotropical Neotropical Africa Madagascar to India Holarctic, Oriental, African; also penetrating Australian and Neotropical regions Neotropical 294 Tue University oF Kansas ScrENCE BULLETIN Fic. 1. Sphecodes monilicornis (Kirby), female, hairs omitted on left. Fic. 2. Lasioglossum malachurum (Kir- by), female, hairs omitted on left. Scale line represents 1.0 mm. as well as various parts of the body. The posterior ridge and the comb arising from it vary from essentially unmodified (0) to lacking (1) among parasitic groups. When they are lacking, the basitarsus is more nearly round in cross-section, less flattened than in nest making forms. Although de- velopment of the comb varies in the Augo- chlorini and it seems to be absent in some unrelated, nonparasitic genera, its loss es- pecially in the Halictini seems clearly re- lated to parasitism. (c) The base of the mid femur and the apex of the mid tibia of pollen collecting halictids each bears a comb or brushlike row of hairs on the under surface, that on the tibia arising from an oblique ridge. They were termed the mesofemoral brush and mesotibial comb by Eickwort (1969b). As pointed out by Jander (1976), these structures are opposable and serve to re- move pollen from the fore legs. In para- sites the tibial brush is retained but the femoral brush varies from little modified (0) to absent (1). Usually the tibial brush is more brushlike, less comblike, than in many nest making bees. (d) In pollen-carrying forms, there are Tue Parasitic Groups oF Harictipar (HyMENopTERA, APOIDEA) 295 long and usually branched hairs on the underside of the hind trochanter (Fig. 4). On the femur in the Halictini, there are some long branched hairs at the base on the underside, but otherwise the underside is nearly bare. Long branched hairs from the anterior (= outer) surface (Figs. 4, 5) and a longitudinal row of similar hairs from near the middle of the posterior (= inner) surface (Fig. 6) curl down and meet below the femur to form a femoral corbic- ula, the cavity of which is somewhat en- larged by the straight or usually concave, rather than convex, under surface of the femur. The pollen carrying corbicula is less specialized in the Augochlorini. Long branched hairs arise from a more extensive basal area on the under side of the femur; on the posterior surface, while the hairs are directed downward, there is no row of unusually long curled hairs. The corbicula is shorter than in Halictini because of the larger basal hairy area and is open, since very long curled hairs arise only from the anterior side of the femur. Earlier illustra- tions and accounts of halictid scopas are those of Eickwort (1969b) and Eickwort and Fischer (1963). In all parasitic forms the scopa is reduced (Figs. 3, 7-14). In Paralictus the trochanteral scopa is un- modified while the long branched femoral hairs are either somewhat reduced in num- ber and density but curled to form a cor- bicula which often contains some pollen grains (0.2), or the scopa is further modi- fied by loss of the long hairs on the an- terior surface of the femur (0.5). In all other parasitic genera the trochanteral hairs are short and the femoral hairs are short, relatively sparse, and not or little branched. The underside of the femur is relatively bare, however, and the hairs are directed about as in the pollen carrying forms. In some (0.9) the underside of the femur is about as in the pollen collecting ancestral groups (concave in Echthralictus, straight in Temnosoma) while in others (1) the curvature of the under surface is slightly to markedly convex. (ce) The posterior tibia in nest making forms is covered with long plumose pollen carrying hairs except for the relatively bare under surface. Near the anterior (outer) margin of the bare area, i.e., along the lower margin of the outer surface of the tibia, the hairs are especially long (ex- cept in Homalictus) and coarsely branched and while the main axes are directed downward and distally, the apical parts are curled upward and distally. The tibial scopa is modified in all the parasitic groups (Figs. 7-14). In Paralictus (0.3) the long hairs near the lower margin of the outer surface are not much larger than other tibial hairs, are only gently curved, and lack the characteristic coarse plumosity of these hairs in pollen carrying forms. In Echthralictus (0.5) the same is true but the hairs are even shorter, those of the lower margin of the outer surface being quite short. In other parasitic genera, the ventral relatively bare area is not or weakly evident, the hairs have few barbs or are simple, and those along the lower margin of the outer surface are smaller than those elsewhere on the tibia. Forms having a relatively uniform coverage of simple hairs are scored (1) while those with more plumose hairs and a sparsely hairy under surface are scored as (0.8) or (0.9). (f) The hind basitarsus is flattened, usually as long as the remaining tarsal seg- ments, the hairs of its lower margin un- usually long, coarse, straight, simple or with branches only basally. Probably these features function to comb pollen off of the metasoma. In parasitic genera they vary from almost unmodified (0.2), to less flattened and lacking most or all of the coarse, straight hairs (0.7), to those in which the basitarsus is also markedly shorter than the remaining tarsal seg- ments (1). Because of variability this fea- ture is rather weak, but seems worth more bh LS NWAWSs LORE = SSS: ‘ WANS NS Fic. 3. Outer (= anterior) surface of hind leg, Sphecodes monilicornis (Kirby), female. Fic. 4. Same, Lasto- glossum malachurum (Kirby). Fics. 5, 6. Anterior and posterior views of hind femur of Lasioglossum mala- churum (Kirby), female. Most hairs are indicated only by their sockets. Scale line represents 0.5 mm. Tue Parasitic Groups or Harictipar (HyMENopTERA, APoIDEA) 297; ( \ | f x CAM RES, Make PEACE Fics. 7-14. Outer views of posterior femora and tibiae of females. Fic. 7. Paralictus asteris Mitchell. Fic. 8. Echthralictus extraordinarius (Kohl). Fic. 9. Parathrincostoma seyrigt Bliithgen. Fic. 10. Temnosoma smarag- dinum Smith. Fic. 11. Ptilocleptis tomentosa Michener. Fic. 12. Eupetersia coerulea Bliithgen. Fic. 13. Sphe- codes (Austrosphecodes) chilensis Spinola. Fic. 14. Sphecodes (Sphecodes) confertus Say. Scale lines represent 1.0 mm. 298 Tue UNIverRsITY oF Kansas ScIENCE BULLETIN Tas.e 2. Scores for Various Characters of Females of Parasitic Halictid Genera 1 to 8.* 0 = same condition as in nest making ancestral group; 1 = extreme of modification among para- sites. 1 2 3 - 5 6 7 8 a. Vestiture 0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 b. Tarsus I 0 0 1.0 1.0 1.0 1.0 1.0 1.0 c. Mid-femoral brush 0.5 0 1.0 1.0 1.0 1.0 1.0 1.0 d. Femoral scopa 0.2-0.5 0.9 1.0 0.9 1.0 1.0 1.0 1.0 e. Tibial scopa 0.3 0.5 1.0 1.0 1.0 1.0 1.0 0.9 f. Hind basitarsus 0.2 0.7 0.7 0.7 0.7 0.7 0.7 0.7-1.0 g. Ventral scopa 0.2 0.6 1.0 0.8 0.9 0.9 1.0 0.9-1.0 h. Basitibial plate 0.1-0.3 0 0.9 0.7 1.0 1.0 0.7 0.7-1.0 i. Prepygidial region 0.5 0.6 1.0 1.0 1.0 1.0 1.0 1.0 j. Sixth tergum 0.3 0.3 0.4 0.6-1.0 0.9 0.9 0.9 0.9 k. Labrum 0.5-0.7 0.2 0.2 0.8 1.0 0.9 0.9 0.9 1. Tibial spur 0.2 0.2 1.0 1.0 1.0 1.0 1.0 1.0 m. Penicillus 0.5-0.8 0.8 0.5 1.0 1.0 1.0 1.0 1.0 n. Sculpturing 0 0 0.8 1.0 0 0 0-0.6 0.4-0.8 p. Mandible 0.5-1.0 1.0 0 0.5 0.5 1.0 1.0 0 -1.0 q. Head 0.5-1.0 1.0 0 0 0 1.0 1.0 0 -1.0 r. Legs 0 0 0.1 0 0.5 0.5 0.5 0 -1.0 s. Spines 0 0.2 0 0 0.1 0.1 0.1-0.3 0.1-1.0 Mean 0.25-0.38 0.44 0.64 0.72-0.74 0.76 0.83 0.82-0.87 0.69-0.97 *1 = Paralictus, 2 = Echthralictus, 3 = Parathrincostoma, 4 = Temnosoma, 5 = Ptilocleptis, 6 = Micro- sphecodes, 7 = Eupetersia, 8 = Sphecodes. consideration than mere inclusion under general hairiness (feature a, above). (g) Nest making Halictinae vary in the extent to which there is a ventral meta- somal scopa for carrying pollen. Usually there is a preapical zone of rather long plumed hairs, often simple but curled at the tips, across each metasomal sternum, these hairs being longer and more erect and therefore more important in pollen carrying on basal than on apical sterna. In Homalictus, however, the ventral scopa becomes the main pollen carrying struc- ture. The bands of hairs are reduced, usually to a single row on each sternum, but these hairs are enomous, coarsely plumed, directed posteriorly and at their tips often upward toward the sterna to form a corbicula behind each row. More- over, similar giant hairs arise on the lateral, ventral parts of the terga and are directed postero-mesally. In all parasitic forms the sternal scopa is reduced. In Paralictus (0.2) the hairs are only somewhat shorter and less plumose than in the presumed ancestral type. In Echthralictus (0.6), a Homalictus derivative, the large sternal hairs are much reduced in numbers, curv- ature, and in plumosity, some being sim- ple, but are still evident as a few, unusually large hairs; large lateral tergal hairs are absent. The other parasitic genera have only simple hairs of moderate length on the areas concerned, and are scored 0.8 to 1 depending on the density of the hairs, the latter indicating the lower density. (h) The basitibial plate is defined, at least along its posterior margin, and com- monly around the whole margin by an elevated carina in nest making species (Fig. 15). It functions to support the bee working in the nest as it pushes its legs out against the burrow walls (Batra, 1964). In parasitic genera (Figs. 16-22) it may be essentially normal (0), less well defined with the marginal carina lower (0.1-0.3), feebly defined and often only posteriorly (0.7), barely detectable (0.9) or entirely absent (1). (i) The fifth metasomal tergum of nest- Tue Parasitic Groups oF Harictipar (HyMeENoprerA, APoImDEA) 299 Fics. 15-22. Outer views of bases of posterior tibiae of females, including basitibial plates. The scale line repre- sents 0.25 mm. Fic. 15. Lasioglossum malachurum (Kirby). Fic. 16. Echthralictus extraordinarius (Kohl). Fic. 17. Ptilocleptis tomentosa Michener. Fic. 18. Temnosoma smaragdinum Smith. Fic. 19. Eupetersia coerulea Bluthgen. Fic. 20. Sphecodes chilensis Spinola. Fic. 21. Sphecodes confertus Say. Fic. 22. Sphecodes monilt- cornis (Kirby). Scale line represents 0.25 mm. making species has a dense prepygidial fimbria of hairs extending across the seg- ment, apically. This hair band is broad- ened middorsally where the hairy region extends toward the base of the segment. In the middle of this broad area is a sharp- ly defined specialized longitudinal zone lacking ordinary hairs but covered with minute dense hairs. The surface in this zone is very minutely punctate. There is a deep cleft in the tergal margin, dividing this zone longitudinally in the Augochlo- rini but not in the Halictini. In all para- sitic forms this elaborate structure is modi- fied. In no case is there a sharply defined specialized longitudinal zone. In Paralic- tus there is an ill-defined longitudinal area or a small, median, apical region of minute punctation and fine hairs. Although su- perficially the tergal surface and vestiture are uniform from side to side, there is a broad triangular, median wedge lacking long hairs but with scattered hairs of moderate length. This zone includes the small median apical region and is roughly an equilateral triangle; posterolaterally it is continuous with the apical impuctate tergal margin (0.5). In Echthralictus, the specialized zone is also broadened poste- riorly to form a wide, hairless, triangular wedge, only minutely roughened and shiny, cutting into the hairy part of the 300 Tue UnIversiry oF KansAs ScIENCE BULLETIN tergum (0.6). Posterolaterally this zone is continuous with the bare posterior mar- ginal band of the tergum. The posterior margin of the zone is longer than the other two margins which are about at right angles to one another. The zone is similar to that of Paralictus but it is larger and much better defined. Both genera have hairs along the extreme tergal margin. In other parasitic genera the tergal surface and margin show no specialized region, dense to sparse hairs being distributed across it (1). When the hairs form a dense fimbria it is often less dense medially; such a sparsely hairy area may or may not be a rudiment of the specialized zone found in nest-making forms. In all cases it lacks the short fine hairs and dense minute punctation of this zone in nest making forms. (j) The sixth metasomal tergum of pollen-collecting halictids is provided with two heavily sclerotized plates (Fig. 23) which perhaps contribute to the effective- ness of the tamping movements character- istic of halictid nest building (Batra, 1964). The gradulus is strongly arched to the rear and forms a carina delimiting poste- riorly and laterally the more anterior, ele- vated plate, the suprapygidial plate of Eickwort (1969b). An inclined zone, broad anteriorly and tapering posteriorly, slopes downward as it extends backward from the suprapygidial plate and merges 26 4 ' \W a, Fics. 23-25. Median parts of sixth metasomal terga of females. Fic. 23. Lasioglossum malachurum (Kirby). Fic. 24. Paralictus asteris Mitchell. Fic. 25. Ptilocleptis polybioides Michener. Fics. 26-28. Labra of females. Fic. 26. Lasioglossum malachurum (Kirby). Fic. 27. Sphecodes monilicornis (Kirby). Fic. 28. Ptilocleptes tomentosa (Michener). Tue Parasitic Groups of Harictmar (HyMeNnopTerA, ApompeA) 301 with the median basal part of the pygidial plate, which is a flat process projecting beyond the rest of the tergal margin. On either side of the pygidial plate is a hairy region, the hairs constituting the pygidial fimbria, which is broken by the pygidial plate and the inclined zone anterior to it. All parasitic forms have these plates less well developed (Figs. 24, 25). In Paralic- tus (0.3), the arch of the gradulus is less prominent although distinct, the supra- pygidial plate being a relatively small ele- vated area, delimited by a carina which is angulate rather than rounded posteriorly. The inclined zone is absent, and the py- gidial fimbria is sparse, but continuous, across the tergum where the inclined zone would be if it were present. Echthralictus (also 0.3) is similar, but the gradulus is very weak, not carinate, and the supra- pygidial plate therefore is inconspicuous; however, the pygidial fimbria is interrupt- ed medially. Parathrincostoma (0.4) agrees with the comment on Echthralictus except that the whole gradulus is almost invisible and some Temnosoma (0.6) likewise agree except that the gradulus is absent lateral to the suprapygidial plate. In the Sphe- codes group (0.9) the gradulus is only weakly arcuate posteriorly in the middle, there being no hint of a suprapygidial plate, and the fimbria or hairy zone is con- tinuous or narrowly interrupted medially. In some Temnosoma (1) the gradulus and of course the suprapygidial plate are absent and the fimbria is continuous. (k) The labrum of nest making forms (Fig. 26) consists of a transverse, more or less rectangular, thick sclerite or body, sometimes called the basal area of the labrum, from the distal margin of which projects an apical process which is flat- tened, depressed below the surface of the body, relatively thin, pointed or rounded, and provided with a strong dorsal keel and a row of coarse bristles along the mar- gins. The keel is reduced in parasitic forms, and the process is usually less pointed (Figs. 27, 28). In the genera scored 0.2 in Table 2, the labrum is modified principally by the reduction of the keel on the process to a carina. In Paralictus (05- 0.7) the line between the body of the lab- rum and the process is not sharp, the sur- faces sometimes being at or nearly at the same level, and the process is broad and lacks the keel completely, and is bluntly pointed, rounded, or subtruncate at the apex. In one species there are scattered bristles on the surface of the process. In Temnosoma (0.8) the labrum is similar, but the process is relatively short, trun- cated or emarginate, and its bristles scat- tered. In the Sphecodes group, the body of the labrum is reduced to a transverse basal ridge, the process lacks a keel and is broadly rounded, truncate, or bilobed, and there are often bristles other than the marginal ones or the row of bristles is mesal to the margin (Figs. 27, 28). Most members of this group are scored 0.9, but Puilocleptis, because of the greatly short- ened process, is scored 1.0. (1) The hind tibial spurs each has two margins between which is a concave sur- face. Both margins of the outer spur and the outer margin of the inner spur are minutely serrate. The inner margin of the inner spur, in nest making species, varies from more coarsely serrate to wavey, even almost straight, to produced into one or a few long processes. In genera scored 0.2 the inner margin of the inner hind tibial spur is produced into a few large projec- tions, as in the forms ancestral to these parasites, but the projections are smaller (as noted for Paralictus by Eickwort and Fischer, 1963). In all the rest (1); that margin is minutely serrate like the other margins. (m) In nest-making forms, the apex of the hind basitarsus, on the outer sur- face, is prolonged as a flat process bearing a dense brush of hairs, the penicillus, 302 Tue University oF KANnsAs SciENCE BULLETIN shaped like a paint brush and used for spreading liquid on the cell lining (Batra, 1964). A small but slender penicillus is present in Parathrincostoma (0.5); in genera scored 0.8, a small projection re- mains but the brush is absent, while in those scored 1 both projection and brush are absent. ENLARGEMENT AND SPECIALIZATION OF SrructurES AMONG PARASITES: (n) Nest-making Halictinae vary greatly in the coarseness of the punctation and propodeal areolation. Moreover, there is no single measure of such coarseness for a specimen or species. Yet in general it seems clear that parasitic forms are more coarsely sculptured, and the cuticle firmer, than in their nest-making relatives. Tem- nosoma is noteworthy for the coarse pitting of the first two metasomal terga which end in thickened margins instead of thin mar- gins as in other forms, for the coarse trans- verse ridges on the constricted bases of the second and third terga, and for the coarse- ly punctate tegulae. Presumably such fea- tures provide defense against the stings and jaws of host bees. The numbers in Table 2 for this feature are more subjec- tive than for others, but nonetheless give a general idea of sculpturing. (0) The proccipital carina and the an- terolateral angles of the pronotum are highly variable in nest-making forms, as also in the parasites. In general, however, they appear to be more prominent in para- sites, probably providing defense for the neck. Acute produced prothoracic angles occur in Echthralictus and Parathrincos- toma. The latter also has an unusually prominent preoccipital carina. In Temno- soma the head is unusually produced back- ward forming a sharp ridge behind the vertex, protecting the neck region. Per- haps this explains the lack of enlargement of the pronotal angles in this genus. These features are not scored for Table 2. (p) Mandibles in nest-making Halic- tinae have a preapical inner tooth. They are unmodified in forms scored 0 in Table 2. In those scored 0.5, the mandibles are of more or less normal length but lack the inner tooth, or in some Paralictus, are elongate with a small inner tooth; 1 indi- cates mandibles that are very long, pointed, and simple. (q) Most nest-making Halictinae have heads about as wide as the thorax, the clypeus less than three times as wide as high, and the eyes usually converging be- low. Various parasitic groups (0) retain these characteristics. Others (1) have heads usually wider than the thorax, the clypeus over three times as wide as high, and the eyes not converging below. This places the mandibles farther apart, presumably making them more effective in grasping hosts, and is generally associated with long mandibles. (r) In nest-making species the legs are relatively slender, the hind femur being straight or concave on the under surface (see d) (Figs 5, 6). Similarly shaped femora are found in some parasitic forms (0). In Parathrincostoma the legs are ex- traordinarily slender but the under surface of the hind femur is slightly convex (0.1). In others, the under surface of a moder- ately slender femur is slightly convex (0.5) or the femur is more thickened (1), with the under surface slightly convex and the upper surface especially near the base strongly so. Thickened legs presumably house stronger muscles for pushing re- sistant host bees along the burrows (see Figs. 7-14). (s) Nest-making Halictinae lack spine- like tibial setae and the outer surface of the apex of the hind tibia often has a mere tubercle representing the apical tibial spine. Some parasitic groups (0) are similar, others (0.1) differ only in having a some- Tue Parasitic Groups oF Haricriar (HyMENopTERA, APOIDEA) 303 what more recognizable tibial spine, but broader than long. Echthralictus (0.2) has the numerous setae of the outer margin of the hind tibia thickened, tapering, mostly simple, and thus spinelike. Other forms (0.3) have spinelike setae intermixed among hairs on the outer margin of the hind tibia, although the tibial spine is broader than long. In many Sphecodes the spinelike setae become more promi- nent, shorter, stouter, often almost peglike and in some species similar setae occur on the middle tibia. Along with such strength- ening of the spinelike setae, the hind tibial spine is much better developed, becoming two or three times as long as broad. Such forms are scored as 1 (Fig. 3). Probably the spinelike setae and the apical tibial spines provide for a firm grip on the bur- row walls when a parasite pushes a host bee. The means in Table 2 give an idea of the amount of morphological evolution that a parasitic group has undergone since diverging from its nest making ancestors. Obviously Paralictus is less different from Lasioglossum (Dialictus) than other para- sitic genera are from their nest making ancestral groups. Echthralictus, also, is moderately close to Homalictus from which it was derived. Every genus has scores ranging from 0 to 1, showing that acquisition of the characteristics of para- sites has not followed the same sequence in all. The genera from Temnosoma to Sphecodes are all highly modified as para- sites, Temnosoma being more specialized than the least modified Sphecodes, ac- cording to the table. This is probably true, but it should be noted that the range of means for a genus is based on the scores in the column above. There may not be any species having such extreme means, however. For example, in Paralictus, there is no species that combines all the low scores, nor is there a species that combines all the high scores. Thus there is no Para- lictus species with a mean score as low as 25, nor is there one with a mean score as high as .38. SIMILARITY. OF PARASITIC FEMALES TO MALES Males of halictids leave the nests early and play no roles in nest making, defense, or parasitization. It is therefore not sur- prising that in most features relating to pollen gathering and transport and nest construction, the parasites approach males. Specifically, this applies to characters to m of the preceding sections and able 2. In character p also, parasites often re- semble halictid males and it may be that loss of the mandibular tooth should be looked at as loss of a structure important in nest-making, rather than as acquisition of simple, sicklelike mandibles for fighting. Indeed, the Sphecodes species famous for fighting and killing hosts with its mandi- bles [S. monilicornis (Kirby) | is one that retains the inner mandibular tooth. In gross appearance, however, most halictid males, whether belonging to para- sitic genera, or not, are quite different from females, having longer, more slender meta- somas and longer antennae than females. Some nest-making forms have males that look like females [e.g., Lasioglossum lus- trans (Cockerell)|, and some parasites have sexes that look alike, more or less intermediate in shape between typical ha- lictid males and females (e.g., Eupetersia). RELATIVE DIVERSITY OF PARASITIC AND NONPARASITIC HALICTINAE The nonparasitic Halictinae constitute an enormous group of species of impres- sive, morphological homogeneity. The similarity in morphology is suggested by the series of papers by Vachal and Bliith- gen. Vachal, who was a specialist in halic- tines early in this century, placed all non- 304 Tue University oF Kansas ScrENCE BULLETIN parasitic species of the subfamily in a single genus, Halictus. Blithgen, a later specialist who was a “splitter” at the ge- neric level in other groups with which he worked, placed nonparasitic groups as iso- lated as Homalictus, Lasioglossum, Pachy- halictus, and Halictus in a single genus. Only intensive studies and long searches for group characters have made possible the present classification of nonparasitic halictines, dividing the subfamily into two tribes, Augochlorini and Halictini, and many genera (Eickwort, 1969a, b). By contrast, parasitic halictines have almost consistently been accorded generic rank since their species were first studied, and it is perfectly obvious that they are morphologically markedly different from the nonparasitic forms and from one another. No numerical phenetic study has been made to demonstrate the differentia- tion of the parasitic genera quantitatively, but such a study hardly seems necessary for that purpose alone. It seems clear that on all continents, in spite of a wide diversity of habitats, social organizations, nesting substrates, and floral relationships, the nonparasitic forms re- mained basically similar, presumably as a result of selection favoring their various characteristic structures. The peculiarities of the labrum, scopa, pygidial plates and pygidial and prepygidial fimbriae, for ex- ample, vary only within narrow limits. Among parasites, however, these struc- tures and others listed above, mostly in- volved in nest construction and_ pollen gathering and carrying, appear to have escaped from the forces that elsewhere limit their variation. The result is the structural diversity found among parasitic genera. HOST RELATIONSHIPS Most parasitic bees have their own close relatives as hosts. This rule appears to apply to the majority of parasitic Halic- tinae, although some species of Sphecodes are exceptions. Table 3 gives known or probable hosts for parasitic halictids. All hosts are halictids except as otherwise indi- cated. Of course, the most indisputable host data are those in which the parasite has been reared from host cells, or its larvae, pupae, or teneral adults found there. Such records are annotated (R) in the table. Finding a female parasite in the burrow of another bee is not entirely convincing, un- less repeated, since parasitic bees may enter various burrows at night, in bad weather, or in searching for host nests. Such rec- ords, as well as reports of the parasites fly- ing about nesting sites of the hosts, are annotated (N) in Table 3. Reports of hosts based on finding the parasite in the vicinity of or on the same flowers as the supposed host, the latter being the only local bee of the appropriate size and sea- sonality, are mere surmises, but probably are commonly correct, and are annotated (S). Less definite associations have been ignored, except for genera like Echthralic- tus and Parathrincostoma for which no host data exist, but whose hosts are prob- ably correctly surmised. In the table, each parasite name begins a separate paragraph. Specific names of hosts are omitted to save space when dif- ferent host species of the same genus or subgenus are reported. The publications cited after host names provide such infor- mation. Examination of the table suggests no high level of host specificity by most spe- cies of parasites. Some appear to attack any halictines of the right size that are available to them, while different-sized individuals of a parasitic species are re- ported by several authors to be associated with host species of different sizes. Other parasites are doubtless more host specific. Some are primarily or exclusively parasites Tue Parasitic Groups of Harictiar (HyMeNopTerRA, APoIDEA) 305 of halictids, others of andrenids. Certain species of Sphecodes are believed to para- sitize both halictids and andrenids, how- ever, and S. albilabris has been reared from an andrenid and probably from a colletid, but appears to be primarily a parasite of large halictids. Larvae of parasitic Halictinae are very similar to those of other Halictinae, having no specialized structures and_ behavior, such as characterize the parasitic antho- phorid and megachilid larvae that attack host eggs or larvae. In halictid parasites whose behavior has been examined, the adult female parasite opens a host cell, destroys the egg, replaces it with her own egg, and recloses the cell. Evidence for this statement is largely that in repeated exca- vations of nests parasitized by Paralictus (Michener, unpublished), Microsphecodes (Eickwort and Eickwort, 1972), and Sphe- codes (various authors), many cells con- taining immature stages and presumably eggs of parasites must have been seen, although often such cells were not distin- guished from unparasitized cells. G. Eick- wort (personal communication) was able to distinguish the shorter, straighter eggs of Sphecodes autumnalis and sp.? from those of hosts (Perdita and Evylaeus). The Sphecodes eggs were in the typical halic- tine egg position on the provisions and were not easily distinguishable from those of the hosts in the field, although measure- ments served to permit recognition in the laboratory. All cells, both those examined by Eickwort and by other authors, con- tained only one egg or young, indicating destruction of the host-egg by the adult parasite. All were closed, except, no doubt, in those species of Lasioglossum (Evy- laeus) whose cells are left open part of the time. In no case was a partly eaten pollen ball found with an egg, indicating that the parasites do not destroy host larvae and replace them with eggs, but only attack fully provisioned cells in which the host larva has not started to feed. Hosts quickly oviposit in and close any fully provisioned cell. Hence, in nearly all cases, destruction of a host egg or possibly freshly hatched larva must precede laying of the egg of a parasite. Parasites enter the nests of both solitary and eusocial hosts. After ovipositing in appropriate cells, they close the cells and either leave, or remain inside the nest to deposit more eggs on subsequent days, depending at least in part on the number of cells available for parasitization (Knerer and Atwood, 1967; Ordway, 1964). Para- lictus, Microsphecodes, and some species of Sphecodes accomplish parasitization with- out killing adult hosts (Ordway, 1964; Eickwort and Eickwort, 1972; Michener, unpublished) while other Sphecodes spe- cies regularly do so. Legewie (1925) re- corded the results of 76 attacks of Sphe- codes monilicornis on colonies of Lasio- glossum malachurum. The Sphecodes suc- ceeded in 75 cases and 283 Lasioglossum were killed. S. prmpinellae, however, ap- pears to chase away the hosts, Augochlo- rella, spp., without killing them (Ordway, 1964). Species such as Sphecodes monilicornis and pimpinellae attack social hosts and the attack appears to result in destruction of the colony. In the case of solitary hosts, the Sphe- codes may lay in appropriate cells and depart before the host returns, thus being cuckoo-like, as is Nomada. Eickwort and Eickwort (1972) give an account of Microsphecodes kathleenae, fe- males of which enter nests of Lastoglos- sum (Dialictus) umbripenne, a eusocial host, either by passing the guard like a Lastoglossum or by locating unguarded nests. Female Microsphecodes were found in nests along with workers and queens of the host and are believed to locate the host cells, open them, eat the eggs, deposit their own, and close the cells. More than 306 Tue UNIVERSITY oF KansAs ScrENCE BULLETIN Taste 3. Hosts of Parasitic Halictid Bees. (R) = reared or immature parasite taken from host cell, (N) = found in or about nests, (S) host surmised on the basis of seasonal and local occurrence. Paralictus asteris Mitchell—Lasioglossum (Dialictus) imitatum (Smith), (R) Michener and Wille, 1961, parasite misidentified as P. cephalotes (Dalla Torre); (N) Michener, unpublished. Paralictus cephalotes (Dalla Torre)—Lasioglossum (Dialictus) zephyrum (Smith), (N) Robertson, 1901, 1926. Paralictus simplex Robertson—Lasioglossum (Dialictus) versatum (Robertson), (N) as unidentified Paralictus, Michener, 1966; (N) Michener, unpublished. Echthralictus spp-——Homalictus spp. Homalictus is the only possible host in the islands where Echthralictus occurs. Parathrincostoma sp—Thrinchostoma sp. Probable host as judged by size and relationship. Temnosoma spp—Augochlorini such as Augochloropsis or Augochlora. Probable hosts as judged by size, abundance, similar coloration, and relationship. Microsphecodes kathleenae (Eickwort)—Lasioglossum (Dialictus) umbripenne (Ellis), (R) Eickwort and Eickwort, 1972. Microsphecodes russeiclypeatus (Sakagami and Moure)—Lasioglossum (Dialictus) seabrai (Moure), (N) Sakagami and Moure, 1962. Microsphecodes sp—Lasioglossum (Evylaeus) sp., (R) Michener, unpublished. Microsphecodes sp—Habralictus sp., (N) Michener, unpublished. Sphecodes albilabris (Kirby) (= fuscipennis Germar)—Halictus spp., (N) Blagoveshchenskaya, 1955; (N) Bliithgen, 1919; (N) Fahringer, 1922; (N) Friese, 1926; (?) Stoeckhert, 1954. Colletes cunicularius (Linnaeus), COLLETIDAE, (N) Alfken, 1912; (S) Blithgen, 1930; (R?) Malyshev, 1927; (N) Moschler, 1938. Meliturga clavicornis Latreille. ANDRENIDAE, (R) Rozen, 1965. Andrena ovina Klug, ANDRENIDAE, (N) Friese, 1926. Major hosts of this species are apparently Halictus quadricinctus (Fabricius) and sexcinctus (Fabricius) but the only actual rearings from cells are from Colletes and Meliturga. Sphecodes alternatus Smith—Lasioglossum (Evylaeus) nigripes (Lepeletier) (N) Knerer, 1968; (N) Knerer and Plateaux-Quénu, 1970. Sphecodes arvensis Patton (det. Cockerell)—Halictus rubicundus (Christ), (N) Hicks, 1934. Sphecodes atlantis Mitchell—Lasioglossum (Dialictus) pilosus (Smith), (R) G. Eickwort, personal com- munication. Sphecodes autumnalis Mitchell—Perdita octomaculata (Say), ANDRENIDAE, (R) G. Eickwort, personal communication. Sphecodes barbatus Bliithgen—Lasioglossum (Lasioglossum) caspicum (Morawitz), (S) A. W. Ebmer, personal communication. Sphecodes brachycephalus Mitchell—Calliopsis andreniformis Smith, ANDRENIDAE, (S) Mitchell, 1956. Sphecodes chilensis Spinola—Corynura lepida Alfken, (R?, N) Claude-Joseph, 1926; (R? N) Janvier, 1933. Sphecodes crassus Thomson (= variegatus Hagens)—Lasioglossum (Evylaeus) spp., (N) Alfken, 1912; (N) Bliithgen, 1919, 1923a; (R) Bliithgen, 1934; (N) Rapp, 1945; (R) Stoeckhert, 1933. Stoeckhert (1933) depreciates reports that Halictus maculatus Smith is a host and Valkeika (1962) reports but gives no evidence that Lasioglossum (Lasioglossum) sexnotatulum Schenck is a host. Sphecodes cristatus Hagens—Lasioglossum (Evylaeus) nigripes (Lepeletier), (S) Blithgen, 1934. Sphecodes davisii Robertson—Agapostemon virescens (Fabricius), (R) G. Eickwort, personal communi- cation. Sphecodes divisus Kirby—Lasioglossum (Evylaeus) spp., L. (L.) spp. and Halictus spp., (N) Alfken, 1913a; (N) Bliithgen, 1916, 1919, 1934; (N) Chambers, 1949; (S) Morice, 1901; (N) Rapp, 1945; (N) Scholz, 1912. There are also scattered records of this species entering Andrena nests, ANDRENIDAE, (N) Alf- ken, 1913a; (N) Morice, 1901. Sphecodes fasciatus Hagens—Lasioglossum (Dialictus) spp., L. (Evylaeus) spp., and Halictus (Seladonia) tumulorum (Linnaeus), (N, S) Alfken, 1912, 1913b; (S) Bliithgen, 1934; (S) Morice, 1901; (N) Perkins, 1887. Sphecodes ferruginatus Hagens (= rufescens Hagens)—Lasioglossum (Evylaeus) spp., (S) Bluthgen, 1934; (N) Stoeckhert, 1933. Sphecodes near fragariae Cockerell—Perdita nuda Cockerell, ANDRENIDAE, (R) Torchio, 1975. Sphecodes friesei Herbst—Leioproctus (Lonchopria) zonalis (Reed), COLLETIDAE, (R?) Janvier, 1933. Sphecodes gibbus (Linnaeus)—Halictus spp., (N) Aflken, 1913a, 1939; (N) Bluthgen, 1923a, 1934; (R) Breitenback, 1878; (N) Fahringer, 1922; (S) Friese, 1926; (N) Méschler, 1938; (R) Nielsen, 1903; (S) Per- kins, 1887; (S) Sanders, 1948; (N) Stoeckhert, 1933. Lasioglossum (Evylaeus) malachurum (Kirby), (N) Blith- gen, 1934; (N) Stoeckhert, 1933. Lasioglossum (Lasioglossum) leucozonium, (N) Perkins, 1887. Records of Tue Parasitic Groups or Haricripar (HyMENopTERA, APOIDEA) 307 this species in or near nests of unrelated bees involve Dasypoda hirtipes (Fabricius), MELITTIDAE (Friese, 1920, 1923, 1926), Andrena spp., ANDRENIDAE (Morice, 1901; Minkiewicz, 1935, but see Stoeckhert, 1954), and Eucera longicornis (Linnaeus), ANTHOPHORIDAE (Friese, 1920). Of these only the Dasypoda record appears to be based on repeated observations. Sphecodes granulosus Sichel and rugulosus Sichel—Caenohalictus rostraticeps (Friese), (R?, N) Claude- Joseph, 1926; (R?, N) Janvier, 1933. Corynura (Callochlora) chloris (Spinola), (R?) Janvier, 1933. Sphecodes hyalinatus Hagens—Lasioglossum (Evylaeus) spp., (N) Blithgen, 1934; (N) Rapp, 1945; (N) Stoeckhert, 1933. Sphecodes levis Lovell and Cockerell (?)—Lasioglossum (Evylaeus) quebecense (Crawford), (R) G. Eickwort, personal communication. Sphecodes longuloides Blithgen—Lasioglossum (Evylaeus) aglyphum (Pérez), (N) Bliithgen, 1934. Sphecodes longulus Hagens—Lasioglossum (Evylaeus and Dialictus) spp., (N) Alfken, 1912; see also Blithgen, 1934 and Valkeila, 1962. Sphecodes majalis Pérez—Lasioglossum (Lasioglossum) pallens (Brullé), (S) Bliithgen, 1934; (S) Stoeck- hert, 1954. Sphecodes marginatus Hagens (= nigritulus and atratus Hagens)—Lasioglossum (Evylaeus) spp., (N) Alfken, 1913a, 1939; (N) Blithgen, 1934; (S) Moschler, 1938. Sphecodes miniatus Hagens (= dimidiatus Hagens)—Lasioglossum (Evylaeus) spp., (N) Alfken, 1939; (N) Bluthgen, 1919; (S) Perkins, 1887; (N) Stoeckhert, 1933. Blithgen (1916) reported Halictus (Seladonia) tumulorum as a probable host. Sphecodes minor Robertson—Lasioglossum (Evylaeus) cinctipes (Provancher), (R) Knerer and Atwood, 1967. (R, Atwood, 1933, probably also concerns S. minor.) Sphecodes monilicornis Kirby (=subquadratus Smith, quadratus Meyer)—Lasioglossum (Evylaeus) mala- churum (Kirby), (N) Blithgen, 1934; (N) Ferton, 1898, 1923; (N) Grandi, 1961; (N) Knerer, 1968, 1973; (R) Legewie, 1925; (N) Marechal, 1894; (N) Rapp, 1945; (N) Stoeckhert, 1933, 1954. L. (Evylaeus) spp., (N) Alfken, 1913a, b; (N) Bltithgen, 1934; (N) Fahringer, 1922; (N) Grandi, 1961. Halictus rubicundus (Christ), (N) Alfken, 1913a, b; (S) Perkins, 1887. L. (Lasioglossum) spp., (N) Alfken, 1913a, b; (N) Frisby, 1914. Many of the references marked (N) report the Sphecodes killing nest guards or burrowing into nest entrances. Sphecodes niger Sichel—Lasioglossum (Evylaeus) spp., (N) Blithgen, 1916, 1934; (N) Rapp, 1945. Sphecodes pellucidus Smith—Andrena spp., ANDRENIDAE, (N) Alfken, 1913a; (N) Bliithgen, 1934; (N) Chambers, 1949; (N) Fahringer, 1922; (N) Friese, 1920; (N) Moschler, 1938; (S) Perkins, 1919; (N) Stoeckhert, 1933. Dasypoda hirtipes (Fabricius), MELITTIDAE, (N) Friese, 1923. Lasioglossum (Evylaeus) nigripes (Lepeletier), (N) Grandi, 1961. Lasioglossum (L.) spp., (S) Morice, 1901; (?) Valkelia, 1962. Sphecodes pimpinellae Robertson—Augochlorella spp., (R) Ordway, 1964. Sphecodes puncticeps Thomson—Lasioglossum (Evylaeus) spp., (N) Alfken, 1912, 1913a, but see 1939; (S) Bliithgen, 1919, 1934. Sphecodes reticulatus Thomson—Andrena spp., ANDRENIDAE, (N) Alfken, 1913a, b, 1939; (N) Bliith- gen, 1919, 1934; (N) Legewie, 1925; (S) Perkins, 1919; (S) Richards, 1944. Dasypoda hirtipes (Fabricius), MELITTIDAE, (N) Friese, 1920, 1923. Sphecodes rubicundus Hagens—Andrena spp., ANDRENIDAE, (N) Alfken, 1912, 1913a; (N) Blithgen, 1934; (N) Chambers, 1949; (N) Méschler, 1938; (S) Perkins, 1919; (N) Saunders, 1898; (R) Saden, 1895; (N) Stoeckhert, 1919, 1933; (N) Torka, 1913. Sphecodes ruficrus (Erickson) (= hispanicus Wesmael)—Lasioglossum malachurum (Kirby), (N) Fer- ton, 1898. Sphecodes rufiventris (Panzer) (= subovalis Schenck)—Halictus maculatus Smith, (N) Blithgen, 1934; (N) Stoeckhert, 1933. Lasioglossum (Evylaeus) spp., L. (Lasioglossum) spp. and Halictus (Seladonia) sp., (S) Blithgen, 1923. Sphecodes schencki Hagens—Lasioglossum (Lasioglossum) discum (Smith) [= morbillosum (Kriech- baumer)], (R) Grozdani¢, 1971. Sphecodes spinulosus Hagens—Lasioglossum xanthopus (Kirby), (?) Alfken, 1912; (S) Blithgen, 1916, 1934; (N) Perkins, 1889. Halictus rubicundus (Christ), (S) Yarrow, 1943. Records for Andrena are probably incorrect (Stoeckhert, 1933). Sphecodes subovalis Schenck—Halictus maculatus Smith, (N) Rapp, 1945. Sphecodes sp—Calliopsis spp., ANDRENIDAE, (N) Ainslie, 1937; (R) Michener, 1953 (misidentified as Neopasites); (N) Mitchell, 1960; (N) Rau and Rau, 1916; (N) Shinn, 1967. Sphecodes sp.—Perdita spp., ANDRENIDAE, (N) Michener, unpublished; (N) Mitchell, 1960. 308 Tue UNrversity oF Kansas ScIENCE BULLETIN one Microsphecodes was often found in a single nest and the parasites appear to re- main in the host-nest for up to a day or so. The Chilean Sphecodes reported upon by Claude-Joseph (1926) and Janvier (1933), all of them probably in the subgenus Austrosphecodes, probably behave similar- ly. The several North American species of Paralictus are parasites in Lasioglossum (Dialictus) nests and adults of host and parasite are regularly found together in the same nest. Probably they, too, have habits similar to those of M. kathleenae. There are probably morphological cor- relates in females of Sphecodes related to the manner of parasitization. Thus, those that actively dig and push their way into nests, against the defense of hosts, have heavy legs, strong hind-tibial spines, spine- like or peglike tibial setae, and a partially recognizable basitibial plate, all features which presumably make the parasite more effective in pushing its hosts. S. monilt- cornis has these features very well devel- oped (Figs. 3, 22) and also has a heavy, quadrate head, which provides space for the strong mandibular musculature; the females of this species puncture or crush the heads of the host with their long mandibles. On the contrary, the subgenus Austrosphecodes and the genus Micro- sphecodes, whose females apparently live more or less peacefully in nests of the host, have slender legs without the projections and basitibial plates to improve traction. Unlike many Sphecodes, which com- monly visit flowers, adults of Microsphe- codes perhaps feed only in cells of their hosts. Most of the known specimens have been collected only recently, by persons studying nests of halictine bees. No speci- mens are known to have been taken on flowers although M. russeiclypeatus was obtained in a place where much floral col- lecting has been done by me and by those associated with Padre J. S. Moure. Pub- lished observations of M. kathleenae in Costa Rica indicate that it was not found on flowers (Eickwort and Eickwort, 1972). I found the unnamed species listed in Ta- ble 3 as parasites of Habralictus and Lasto- glossum (Evylaeus) to be common flying about nesting sites of their hosts in Co- lombia, as well as in the nests, yet not one was seen on a flower. Another Colombian species was taken about a bank inhabited by Lasioglossum (Dialictus) where exten- sive sweeping of the flowers visited by the Dialictus failed to reveal even one Micro- sphecodes. These observations suggest that the principal, if not the only, feeding place of Microsphecodes adults may be in the nest of the host, where they probably feed on food masses stored in cells. Such be- havior is suggestive of that of the parasitic genera of allodapine bees which likewise do not visit flowers but feed in the nests of the host. SYSTEMATIC TREATMENT For genera 1 to 4 of Table 1, the com- mentary in this section indicates the place- ment relative to their nonparasitic ances- tors. For genera 5 to 8, the Sphecodes group, however, more detailed, compara- tive descriptions are given, because these genera are closely related and require sub- stantiation. Moreover, they probably had a common parasitic Sphecodes or Sphe- codes-like ancestor and thus constitute a natural group, unlike genera 1 to 4, each of which arose from a different nest- making ancestor (Table 1). Genus PARALICTUS Rosertson (Bigs. 7,24, 29) Paralictus Robertson, 1901, Canadian Ent., 33:229. Type species: Halictus cephalicus Robertson, 1892 (not Morawitz, 1873) = Halictus cephalotes Dalla Torre, 1896, by original designation. This is a North American group of several species, resembling and presumably Tue Parasitic Groups or Haricrmar (Hymenoptera, AporeA) 309 derived from the subgenus Dvralictus of Lasioglossum. It resembles Dialictus in small size, presence of some greenish color- ation on the head and thorax, weakened second and third transverse cubital veins (second sometimes absent) and second re- current vein, presence of a few (two to four) coarse teeth on the inner margin of the inner, hind tibial spur (but these teeth smaller than is usual in Dialictus), male genital structure including the large, re- trorse, ventral lobe on the base of the gono- stylus, as well as various other features marked as zero or near zero in column 1, Table 2. Paralictus differs from Dialictus in those features having high numbers in column 1, Table 2. Included among these are reductions in various pollen-handling and nest-making structures, such as char- actersec to! 7 and mm. (For details, see the explanations for the characters listed in Table 2). Other outstanding features of Paralictus are the following: (kf) Apical process of labrum of female round- ed at apex, without keel. (p) Mandible of female large, sometimes with subapical tooth, although mandible narrower and more pointed than in most Dialictus, some- times acutely pointed, without distinct, subapical tooth. (This feature is ap- proached in some Dyalictus, as is the next.) (q) Head of female often quadrate, inner margins of eyes sometimes parallel or di- verging below so that face is wide below, genal areas commonly greatly broadened, much wider than eye. The males are not distinguishable from those of Dialictus by characters that appear to be of generic importance (but see Mitchell, 1960). As emphasized to me by Dr. George C. Eickwort, who has studied Paralictus in some detail and who provided me with an identified set of specimens for study, there is considerable diversity among the species, some being more like Dialictus than others (see variation indicated in column 1, Ta- ble 2). P. michiganensis Mitchell and sim- plex Robertson are perhaps the most Dia- lictus-like, having a relatively unmodified head (especially in simplex), bidentate mandibles (inner tooth smaller and _nar- rower than in Dialictus), the body of the labrum distinctly elevated above the apical process, and the carina margining the basitibial plate strong. On the other hand, in some features the same species are quite different from Diéalictus. Thus the hind femur lacks long, scopal hairs on the anterior surface, the penicillus is much re- duced, and the labral process in simplex is subtruncate and bears scattered bristles other than those near the margin. At the other extreme are species such as P. cepha- lotes (Dalla Torre) and asteris Mitchell which have large heads, the inner orbits nearly parallel or diverging below, the mandibles enormous and untoothed, the labral process weakly differentiated from the body of the labrum, and the carina margining the basitibial plate weak. The long, curled, scopal hairs on the anterior surface of the hind femur in asteris and the well formed penicillus in cephalotes, however, are Dyalictus-like features of these species. (The largest penicillus, but still reduced compared to Dialictus, oc- curs in an undescribed species from Law- rence, Kansas.) The question arises as to whether Para- lictus is a monophyletic group or consists of parasites derived independently from different species of Dialictus. Perhaps this question can be answered if Dialictus be- comes well enough known that the group or groups from which Paralictus arose can be determined. At present it is impossible to reach a decision. Obviously, if Paralic- tus is polyphyletic it must either be divided into two or more genera (or subgenera of Lasioglossum) or synonymized with Dia- lictus. For the present, I maintain Paralic- tus as a genus because it differs from Dia- 310 THe UNIverRsITY oF KANSAS ScIENCE BULLETIN lictus by many more, and more striking, characters than those that separate the non- parasitic groups such as Dialictus, Evy- laeus, Lasioglossum s. str., and even Halic- tus from one another. Genus ECHTHRALICTUS PERKINS AND CHEESMAN (Figs. 8, 16, 30) Echthralictus Perkins and Cheesman, 1928, Hymenoptera—Apoidea, Sphecoidea and Vespoidea, in Insects of Samoa, British Museum (Natural History), p. 14. Type species: Halictus extraordinarius Kohl, 1908, by original designation. This genus contains at least two species restricted to Samoa, where it presumably arose from Homalictus (Homalictus) spe- cies. As indicated by Michener (1965) it resembles Homalictus proper in the fol- lowing features: small size; presence of some greenish or bluish coloration on the head and thorax; presence of a frontal carina; weakened third transverse cubital and second recurrent veins; presence of a 29 few, coarse teeth on the inner margin of the inner, hind tibial spur; presence of a few, very long hairs representing the dis- tinctive sternal scopa of Homalictus; rather elongate male genitalia with short gono- styli and without a basal, retrorse, gonosty- lar lobe, as well as various other features marked as zero or near zero in column 2 of Table 2. Echthralictus differs from Homalictus in those features having high numbers in column 2, Table 2. Among these are re- ductions in pollen-carrying — structures (characters d to g) and in some probable nest-making structures (characters 7, 7, m), as explained in the paragraphs about char- acters listed in Table 2. The femoral scopa is not recognizable, except for the row of long hairs on the posterior surface which is reduced to simple hairs, the longest about two-thirds as long as the maximum fe- moral diameter. The tibia has a distinct, relatively bare, under surface, but the hairs margining it are not longer than those else- where on the tibia and have only short PRETO ES Fics. 29, 30. Facial views of females of Paralictus asteris Mitchell and Echthralictus extraordinarius (Kohl). Tue Parasitic Groups or Haricripar (HyMENopTERA, ApoIDEA) 311 branches. Except for some long hairs aris- ing near the tibial spurs, the longest tibial hairs are the thick, mostly-simple bristles along the outer margin (character s). Other outstanding generic characters of Echthralictus are: (k) Apical process of labrum of female rather narrowly rounded (rather than pointed) at apex; the keel reduced to a strong carina. (0) Dorso- lateral angle of pronotum produced to acute spine, rounded at apex. (p) Mandi- ble of female acute, without subapical tooth. (q) Head of female quadrate, inner margins of eyes subparallel, so that face is wide below, genal area much wider than eye. Another distinctive feature, perhaps re- lated to the loss of the tergal and great reduction of the sternal scopa, is the round- ed lateral metasomal margin. In Homalic- tus, the metasoma is compressed so that its lateral margin is a strong bend, or crease, in the terga where their ventral surfaces join the dorsal surfaces. The inner margin of the inner, hind-tibial spur of the male is coarsely pectinate, suggesting a female. In features such as the wide head and the coarse teeth on the inner hind-spur of the male, Echthralictus suggests Homalic- tus ctenander Michener from Australia, but in the labral process, female mandible, scopa, etc., that species does not show evi- dence of parasitic behavior (Muichener, 1965). While the male of Echthralictus is easily distinguished from that of most Homalictus, species like H. ctenander eliminate useful, constant, generic charac- ters for males. Genus PARATHRINCOSTOMA BLUTHGEN (Figs. 9, 31-38) Parathrincostoma Blithgen, 1933, Mitt. Zool. Mus. Berlin, 18:389. Type species: Para- thrincostoma seyrigi Blithgen, 1933, by original designation and monotypy. This genus, known only from Mada- gascar, was first described from the male. Although Bliithgen could not determine whether the pollen-carrying scopa of the female was present or absent, he suggested that the genus was probably a parasite de- rived from Thrinchostoma. Benoist (1962) described the female and noted the absence of the “brosse collectrice aux tibias III,” but made no point of this finding, nor did he note the lack of a femoral scopa. Pre- sumably he was confused by the fact that one of his two, supposed species was a Nomia with pollen-collecting apparatus (see Appendix). Thus the parasitic nature of the genus is here established for the first time, based on the lack of the pollen- carrying apparatus. Parathrincostoma agrees with Thrin- chostoma subgenus Eothrincostoma in the characters listed below: Both sexes: Nonmetallic black, large and slender, 11-14 mm long; metasoma elongate, widest at third segment. Clypeus produced downward and strongly protu- berant forward, a line tangent to the lower ends of eyes crossing clypeus near middle or at lower third; malar space conspicuous, but shorter than basal mandibular width. Paraocular area extending as a _ strong, right-angular lobe into clypeus. Mouth- parts long and slender, glossa linear and much exceeding the short galea and palpi. Pronotum with horizontal, dorsal surface rather broad, margined anteriorly by high carina or lamella, not overhung by scu- tum. Dorsal surface of propodeum much longer than metanotum. Apical wing veins strong; first transverse cubital arising well away from margin of stigma, vein r being about three times as long as wide, first recurrent vein approximately inter- stitial (or in one specimen of P. seyrigi, second transverse cubital absent so that there are only two submarginal cells); no area of dense hair along second transverse cubital vein, this vein not angular or thick- 312 Tue UNiversiry oF Kansas ScrENCE BULLETIN Fics. 31-34. Parathrincostoma seyrigi Bluthgen. Fics. 31, 32, male and female. Fics. 33, 34, faces, male and female. ened; anterior margin of third submarginal cell about half as long as posterior margin; marginal cell minutely truncate and ap- pendiculate at apex. Apical margins of terga 1-4 of females and 1-6 of males broad- ly impunctate and slightly depressed. Male: Labrum with strong, median, apical process similar to that of female, this process margined with bristles, with- out keel, but with feeble, longitudinal, median ridge. Basitibial plate an elongate, hairless, shining area, not defined by a carina. Parathrincostoma differs from Thrin- chostoma in the features represented by high scores in column 3, Table 2, and dis- cussed in the paragraphs associated with that table as well as below: (a) Body and legs largely without short hairs except behind pronotal lobe, sparsely across dor- sum of pronotal collar, and in males on large areas with short, dense, brown hairs on terga 3-5. Depressed apical tergal mar- gins without laterally directed hairs in female or in male with such hairs poorly represented. (/) Basitibial plate of female feebly elevated, shining, not clearly de- fined. (7) Fifth metasomal tergum of fe- male hairy, with hairs nearly to margin, without any evidence of median special- ized area, surface not hidden by hairs, apical margin weakly produced medially. (7) Sixth tergum with pygidial and supra- pygidial plates united, the gradulus weak or absent. (Rk) Labral keel of female re- duced to a strong carina, high and almost a keel apically. Other distinctive features of Parathrin- costoma are as follows: Dorsolateral, pro- notal angle obtuse or produced to acute spine, connected mesally with carina or lamella across anterior margin of collar. Propodeal triangle broader than in Thrin- chostoma, posterior margin curving onto posterior pronotal surface. Male: Face and legs black; flagellum of moderate length, all segments longer than broad or second Tue Parasitic Groups oF Harictipar (HyMENopterA, APomDEA) 313 as broad as long; hind tibia slender, black, unmodified, the spurs in normal positions; first two hind tarsal segments not fused, but articulation much broader than more distal articulations; seventh tergum with- out pygidial plate, but with broadly round- ed, apical margin as seen from above, this margin being a sharp separation of dorsal and ventral surfaces of tergum. Sterna 4 and 5 unmodified. It is interesting that the unmodified fourth and fifth sterna and hind tibia of the male and the nonfused first and second hind tarsal segments of the male appear to be ancestral to those of any Thrincho- stoma. Relative to most Thrinchostoma, several other characters, especially those of the wings, are ancestral, being more like ordinary halictids. It is in these latter characters that Parathrincostoma resembles the subgenus Eothrincostoma. Eothrinco- stoma is restricted to Africa, so far as known not being found in Madagascar or Asia. Unless the ancestral wing, leg, and sternal characters are reversions somehow related to parasitic behavior, the implica- tions are that Parathrincostoma is an an- cient parasite that arose from a common ancestor with Eothrincostoma either in Africa or at a time when Eothrincostoma was in Madagascar. Eothrincostoma and Thrinchostoma have the common derived male features of modified sterna, fused first and second hind tarsal segments, and enlarged and modified hind tibiae. Genus TEMNOSOMA Smitru (Figs. 10, 18) Temnosoma Smith, 1853, Catalogue of hy- menopterous insects in the collection of the British Museum, 1:38. Type species: Temnosoma metallicum Smith, 1853 (monobasic). Micraugochlora Schrottky, 1909, Rev. Mus. La Plata, 16:138. Type species: Mzer- augochlora sphaerocephala Schrottky, 1909 (monobasic). Temnosomula Ogloblin, 1953, Bol. Soc. Ent. Argentina, 2:2. Type species: Temno- soma (Temnosomula) platensis Ogloblin, 1953, == Temnosoma_ sphaerocephala (Schrottky, 1909), by original designation and monotypy. This genus, which ranges from south- ern Arizona to Argentina, consists of bril- liant metallic-green species with coarse punctation (suggestive of chrysidids), the only known parasites in the tribe Augo- chlorini. (All other forms treated in this paper are in the tribe Halictini). Temno- soma has been described and illustrated by Eickwort (1969b) and its features related to the presumed parasitic behavior are summarized in Table 2 and the accom- panying text and figures. More detailed treatment is not necessary here. In some respects there is more variation in the genus than Eickwort recognized. Thus the labral process of the female is sometimes emarginate instead of truncate and the body of the labrum may have a single median elevation instead of a pair of them. The gradulus of the sixth tergum of the female may be absent, so that there is no suprapygidial plate and the hairy zone or pygidial fimbria extends across the tergum in front of the pygidial plate unin- terruptedly. On the other hand, as illus- trated by Eickwort, there may be a weakly defined, small, suprapygidial plate and be- hind it, a break in the hairy zone. THE SPHECODES GROUP The content of this large and wide- spread group is indicated in Table 1. It contains halictids that are highly modified as parasites. Probably because of its an- tiquity, its most closely related nonpara- sitic relatives are not readily identifiable. The distal veins of the forewing are strong, a character shared with Halictus and differentiating the parasites from Lasioglossum, Homalictus, and their rela- 314 Tue UNIversiry oF KANnsAs ScrENCE BULLETIN Fics. 35-38. Parathrincostoma seyrigi Blithgen, male. Fics. 35, 36. Dorsal-ventral and lateral views of genitalia. Fics. 37, 38. Seventh and eighth metasomal sterna, dorsal at left. Fics. 39-43. Ptilocleptis tomentosa Michener, male. Fic. 39. Seventh tergum. Fics. 40, 41. Dorsal-ventral and lateral views of genitalia. Fics. 42, 43. Seventh and eighth metasomal sterna, dorsal at left. Fics. 44-46. Ptilocleptis eickworti Michener, male. Fic. 44. Seventh tergum. Fics. 45, 46. Dorsal-ventral and lateral views of genitalia. Scale lines represent 0.5 mm. Tue Parasitic Groups of Harictipar (HyMeENopTERA, APOIDEA) 315 tives. Unlike Halictus, however, the distal margins of the metasomal terga are usu- ally rather broadly hairless, never with hair bands. Also, unlike Hadlictus, the male gonostylus often has a small basal ventral lobe, probably homologous to the retrorse lobe that is widespread in the Lasioglossum-Homalictus group. The rather large gonostyli, however, are sug- gestive of those of some Halictus. Partly because of the great variability among species of Sphecodes, the genera of the Sphecodes group are not easily defined. The more diagnostic characters are itali- cized in the following descriptions. PTILOCLEPTIS New Genus (Figs: 11, 17, 25, 28, 39-54) Type species: Ptilocleptis tomentosa new species. The three species of this genus, all of them described in the Appendix, resemble certain polybiine wasps because of the elongate body, dark costal margins of the forewings, and in some cases the yellowish integumental markings, in others the pat- tern of yellow brown pubescence. Both sexes: Punctation moderate, dense on head and thorax, not involving coarse pitting as is usual in Sphecodes. Head Iit- tle wider than long, eyes strongly converg- ing below, clypeus only about twice as broad as long, not biconvex. Eyes with scattered hairs, extremely short to about one-fourth as long as diameter of scape. Hypostomal carina a very high, thin la- mella, gradually reduced toward posterior end, abruptly reduced near angle so that transverse part is low. Hairs of flagellum mostly short, but dorsal side with some hairs one fifth as long as diameter of fla- gellum, basal segments with some as much as one third that length. Preoccipital ca- rina present, complete. Pronotum with horizontal surface of collar very-short me- dially, a strong carina across anterior mar- gin of collar between lateral angles; \ateral angle right-angular, extended downward as strong vertical carinate ridge which merges with a laterally directed lamella that extends toward coxal base; a carina from lateral angle of pronotum onto pos- terior lobe, continuing at least two thirds of way across lobe. Anterior extremity of scutum narrowly vertical, then abruptly curving onto dorsal surface, vertical area that lacks punctures present, but small, or absent in P. polybioides Michener. Scu- tellum rather flat medially or biconvex. Propodeum with dorsal area bearing strong longitudinal carinae, the area about as long as or longer than metanotum, distinctly shorter than scutellum; posterior and lat- teral surfaces of propodeum with short plumose hairs. First metasomal tergum as long as or longer than wide. Wings with rather long hairs over en- tire surface; stigma of moderate size; mar- ginal cell with apex narrowly rounded almost on wing margin, free part beyond submarginal cells much longer than part subtended by marginal cells; marginal cell unusually broad because vein beyond sub- marginal cells is gently curved almost tts entire length; submarginal cells extending well beyond apex of stigma; second and third submarginal cells each receiving a recurrent vein, or second transverse cubital vein absent, so that there are only two submarginal cells, the second receiving both recurrent veins. Female: Mandible without subapical tooth, relatively short. Labrum with apical process a little over twice as wide as long to four times as wide as long. Legs slender, hind femur well over four times as long as wide; basitibial plate tot- ally absent; “bia rather densely covered with relatively short, slender, essentially simple hairs, some near upper margin with a few short barbs, some shorter plumed hairs interspersed, no spinelike setae; hind tibial spine absent. 316 Tue Universiry oF Kansas SciENCE BULLETIN Fics. 47-54. Ptilocleptis. Fics. 47-49. P. tomentosa Michener, holotype female. Fic. 50. Face of allotype male of P. tomentosa. Fics. 51, 52. P. polybioides Michener, holotype female. Fic. 53. Wing of P. tomentosa, allotype male. Fic. 54. Propodeum and base of metasoma of P. tomentosa, holotype female. THE Parasitic Groups or Harictipaz (HyMENopTERA, APOIDEA) 317 Fifth tergum with apical fringe of plu- mose hairs longer than those of preceding terga. Pygidial plate a thin slightly up- turned apical process in front of which a bare zone extends anteriorly toward mid- dle of tergum in P. tomentosa. Male: Antenna of moderate length as in female, not thickened, first flagellar seg- ment broader than long or as long as broad, others longer than broad or second about as long as broad. Labrum with api- cal process two to four times as broad as long. Second hind tarsal segment shorter than to longer than third, base broader than base of third. Gonocoxite not striate, without depres- sion; gonostylus without basal lobe. This genus is known from Mexico (un- known locality) to southern Brazil. It differs from Sphecodes in the features italicized above, especially those of the labrum, clypeus, and hind tibia. Moreover, it does not look like a Sphecodes, not only because of its form and relatively fine punc- tation, but because of lack of red colora- tion and the extensive coverage of the body with short plumed hair, as indicated in the descriptions of the species. In its elongate body, slender legs and relatively fine punctation this genus sug- gests Eupetersia. It is much less similar to Sphecodes than is Eupetersia, however. Presumably it is an independent derivative of Sphecodes, as indicated by the charac- ters emphasized above as well as by the relatively narrow head and short mandi- bles, features suggestive of S. convergens which may represent the group from which Ptilocleptis arose. The generic name is based on pzilon, feather or plume, and elptis, thief, and is feminine in gender. Genus MICROSPHECODES Eick wort AND STAGE (Figs. 55-57) Microsphecodes Eickwort and Stage, 1972, Jour. Kansas Ent. Soc., 45:501. Type spe- cies: Sphecodes kathleenae Eickwort, 1972, by original designation. Both sexes: Punctation rather weak, not involving such coarse pitting as is usual in Sphecodes. Clypeus three to four times as broad as long, not biconvex. Eyes with only very short, scattered hairs. Hairs on upper surface of antennal flagellum often one third as long as diameter of flagellum or more, such long hairs present on all flagellar segments. Preoccipital ca- rina present or absent. Pronotum with horizontal surface or collar poorly defined, rounded onto declivous anterior surface, narrow medially, lateral angle rounded or obtusely angulate with no marked vertical ridge extending downward from it and no carina or lamella extending toward coxal base; a weak ridge extending to pronotal lobe, but not continued as a carina across lobe. Anterior extremity of scutum gently convex except adjacent to pronotum, no large strongly convex or vertical region, but with small transverse more or less verti- cal area without punctures. Scutellum bi- convex or not, the midlateral areas (con- vexites when present) largely impunctate. Propodeum with dorsal area _ slightly shorter than or as long as scutellum, semi- lunar, with a few, often irregular carinae. Posterior and lateral surfaces of propo- deum with numerous, short, plumose, white hairs in addition to scattered longer hairs. Wings with rather long hairs over en- tire surface; stigma large; marginal cell pointed at or near wing margin; free part of marginal cell more than twice as long as part subtended by submarginal cells, which do not extend beyond apex of stig- 318 Tue University oF Kansas ScreENCE BULLETIN Fics. 55-57. Microsphecodes kathleenae (Eickwort and Stage). Fics. 55, 56. Faces of female paratype and allo- type male. Fic. 57. Wing of allotype male. Fics. 58, 59. Eupetersia flava Michener, face and dorsum of propodeum and base of metasoma. ma; second and third submarginal cells each receiving a recurrent vein. First tergum slightly longer than broad, more elongate in male; metasoma moder- ately elongate, more slender in male than in female. Almost no constriction between first and second terga as seen in lateral view. Female: Mandible without subapical tooth. Labrum as usual in Sphecodes, with broad rounded apical process about twice as broad as long. Legs slender, hind femur over three times as long as wide, upper surface near base scarcely convex; basitibial plate en- tirely absent; hairs on outer side of hind tibia simple or nearly so, upper margin of hind tibia without spinelike setae or pegs; hind tibial spine absent or nearly so. Fifth tergum with apical margin bare, broadly impunctate, like preceding terga. Pygidial plate narrow, rounded apically, parallel-sided, marginal carinae extending only a short distance onto tergal disc. Male: Antenna not much longer than in female, flagellum not thickened, first two flagellar segments both a little broader than long. Labrum as in female. Second hind tarsal segment slightly longer than third, narrowed toward base and articulated like third. Gonocoxite not striate, without de- pressed area. Gonostylus broadly rounded, its basal ventral setose lobe rather small and directed apically. This is a neotropical genus of minute species, 3.25 to 6 mm long, often with more or less extensive yellowish areas on the head and thorax. Species are known from Costa Rica to southern Brazil. The characters italicized in the above descrip- tion separate Microsphecodes from most Tue Parasitic Groups oF Haricripar (HyMENopTerA, AporpeA) 319 Sphecodes. Some of these characters, such as the simple mandibles, occur also in vari- ous groups of Sphecodes, but others seem very distinctive and have led me to elevate Microsphecodes to the genus-level. Among such characters are the gently convex anterior part of the scutum, the enlarged stigma and short submarginal cells relative to the marginal cell, and the lack of an apical fringe on the fifth tergum of the female. Thus Microsphecodes is fully as distinctive as Eupetersia. The described species of the genus are listed and discussed by Eickwort and Stage (1972). Various small species of Sphecodes re- semble Microsphecodes in appearance and some species of Austrosphecodes from South America may represent a group of Sphecodes related to Microsphecodes. In other areas, similar species obviously result from convergence. For example, Sphe- codes (Sphecodes) antennariae Robertson has the size, coloration, shining thorax with small punctures, lack of areolation, and simple female mandibles of Micro- sphecodes. However, the distinct origin of S. antennariae is shown by most of the above italicized characteristics and by the presence of a few spinelike setae on the outer margin of the hind tibia of the fe- male, the partially recognizable basitibial plate of the female, the thickened articu- lation between the first and second hind tarsal segments of the male, etc. Genus EUPETERSIA ButrucEen (Figs. 12, 19, 58-66) Both sexes: Punctation moderate to fine, varying from dense to sparse, not in- volving such coarse pitting as is usual in Sphecodes. Head much wider than long, clypeus three or more times as wide as long, not or feebly biconvex. Eyes hairless or with scattered very-short hairs. Hairs of antennal flagellum all short or some as Fics. 60, 61. Eupetersta guillarmodi Michener, dor- sal-ventral and lateral views of male genitalia. Scale lines represent 0.5 mm. much as one-fifth as long as diameter of flagellum. Preoccipital carina present, at least at sides where lower extremities ap- proach or join posterior extremities of hypostomal carinae. Pronotum with horizontal surface of collar reduced almost to the vanishing point medially (broader in geillarmod?), but forming lateral angle which is usually prominent (weak and rounded in coerulea Blithgen and ruficrus Blithgen and rela- tives) and below which vertical ridge ex- tends downward, a carina usually evident from lateral angle onto posterior lobe. An- terior extremity of scutum strongly con- vex, the vertical anterior surface sometimes lacking punctures and therefore with an area sharply different from rest of scutum. 320 Tue Universiry oF Kansas ScreENcCE BULLETIN Scutellum variable. Propodeum with dor- sal area coarsely-rugose, aerolate, markedly longer than scutellum, usually slightly con- cave, broad laterally so that it is semilunar and broadly rounded posteriorly (not tri- angular); posterior and lateral surfaces of propodeum with numerous short plumose pale hairs in addition to scattered longer hairs. Wings with rather long hairs over entire surfaces; stigma of moderate size; marginal cell narrowly truncate to pointed on or near wing margin; free part of mar- ginal cell beyond submarginal cells much longer than part subtended by submarginal cells, which usually extend but little be- yond apex of stigma. First tergum usually about as long as broad; metasoma in both sexes moderately elongate, usually more slender than in most female Sphecodes and not as slender and parallel sided as in most male Sphecodes. (In E. guillarmodi, metasoma relatively robust, same shape in male and female.) In lateral view, a constriction between first and second terga (except in E. gz«l- larmodi). Posterior margins of terga 2-4 broadly depressed, hairless, impunctate. Female: Mandible without subapical tooth. Labrum as usual in Sphecodes with broad, rounded apical process about twice as broad as long. Legs slender; hind femur over three times as long as wide, upper surface near base scarcely convex; basitibial plate en- tirely absent or slightly elevated; long hairs on outer side of hind tibia simple or nearly so; upper margin of hind tibia with or without spinelike setae; hind tibial spine broad, blunt, not longer than broad. Fifth tergum with apical margin fringed, unlike preceding terga, and some- times with a conspicuous hairless zone in front of fringe. Pygdial plate usually nar- row, the sides parallel, but broad in E. guillarmodi and flava. Male: Antenna short as in female, flagellum not thickened, first two flagellar segments both distinctly broader than long. Labrum with apical process short so that it is three to several times as broad as long. Second hind tarsal segment about as long as third, base broader than base of third. Gonocoxite not striate, with dorsolat- eral depressed area with lamella or flange on either side and often with one or more longitudinal ridges across depressed area. Gonostylus long, with basal setose lobe broad and directed mesobasad. This genus ranges widely through sub- Saharan Africa, south as far as Cape Prov- ince; it also occurs in Madagascar, the Sey- chelles, and has recently been found in southern India. Baker (1974) has cata- logued the described species. [He missed records of E. sakalava Blithgen by Benoist, 1962, as well as a species, E. constricta (Benoist), new combination, described in the same paper as a Sphecodes and here transferred to Eupetersia on the basis of the description.| Keys to the species were given by Bliithgen (1928, 1935). Compared to Sphecodes, the body form is rather elongate, about the same in the two sexes, with the antennae of the male short as in females. In species such as E. bequaerti (Meyer), the pitting of the meso- scutum, mesepisterna, and to a lesser de- gree the rest of the thorax is quite coarse, suggestive of Sphecodes. Also, the first metasomal tergum is not as elongate as in most species (length to breadth about as 9:10) and the part of the marginal cell subtended by submarginals extends well beyond the stigma. In these respects as well as coloration E. bequaerti resembles Sphecodes. Unfortunately, I have not seen males of this species, but it is a reasonable assumption that such Sphecodes-like forms were ancestral to the species typical of Eupetersia. The existence in Africa of Sphecodes Tue Parasitic Groups oF HaricrmAr (HyMENopTeERA, APOIDEA) 321 Fics. 62-65. Eupetersia guillarmodi Michener. Fics. 62, 63. Faces of holotype female and allotype male. Fics. 64, 65. Dorsal views of median parts of body, holotype female and allotype male. species (unidentified, Karen, Nairobi, Kenya, Univ. of Kansas) with simple mandibles in the female, a strong preoc- cipital carina, and a narrow pygidial plate, nearly bridges the gap between Sphecodes and Eupetersia and supports the suggestion of Benoist (1962, p. 124) that Eupetersia is not generically distinct from Sphecodes. I have concluded, however, that Eupeter- sia, although an offshoot of Sphecodes, is different enough to receive generic status. The distinctive combination of characters is italicized in the above description, but it must be noted that few of these char- acters, by themselves, are diagnostic. Most Eupetersia are distinguished from Sphecodes by the fine thoracic punctation, either rather dense or sparse, as well as the body form. The coloration is commonly unlike that typical of Sphecodes, ranging from entirely black or dark metallic blue | E. coerulea Blithgen and allies, placed in Calleupetersia by Cockerell (1938) only because of the blue color| to specimens with the thorax red and the rest of the body black (some specimens of E. para- doxa Bluthgen) to the entirely reddish yel- low E. flava described below. Slender legs are diagnostic for females. The slender, parallel-sided pygidial plate occurs in some otherwise ordinary Sphe- codes, such as S. crassus Thompson, gibbus (Linnaeus), Aeracle: Robertson, etc. The antennae of known males do not resemble those of any Sphecodes known to me. The flagellum is relatively short, not thickened as in Sphecodes, the middle segments slightly broader than long to longer than broad. The first two segments are both much broader than long. In Sphecodes, even species with relatively short male antennae, such as S. cressoni Robertson, atlantis Mitchell, and costaricensis Friese, the flagellum is longer and thicker than in the female and although the first segment is much broader than long, the second is nearly as long as broad or longer than 322 Tue University oF Kansas ScreENcE BULLETIN broad. Unfortunately males of the most Sphecodes-like Eupetersia, such as E. be- quaerti, are not known to me. For the same reason the genitalic characters listed above cannot with certainty be considered diagnostic of all species. SuscENUs EUPETERSIA BLUTHGEN S. STR. Eupetersia Blithgen, 1928, Deutsche Ent. Zeitschr., 1928:p. 49. Type species: Eu- petersia neavei Blithgen, 1928, by original designation. Calleupetersia Cockerell, 1938, Rev. Zool. Bot. Africaines, 30:329. Type species: Halic- tus lasureus Friese, 1910, by original desig- nation. Scutum more or less densely punctate. Scutellum with surface bigibbous, the two convexities not or sparsely punctate, punc- tures denser around them and on midline separating them. Recurrent veins both entering third submarginal cell or the first interstitial. Depressed area of male gono- coxite preapical. This subgenus is limited to Africa and Madagascar. Calleupetersia is distinctive only in its metallic blue integumental col- oration and does not warrant subgeneric status, SuscENus NESOEUPETERSIA BLUTHGEN Nesoeupeterisa Blithgen, 1935, Deutsche Ent. Zeitschr, 1935:182. Type species: Sphe- codes scotti Cockerell, 1912, by original designation. Scutal punctures scattered. Scutellum gently convex, not bigibbous, with scat- tered punctation similar to that of scutum. Second and third submarginal cells each receiving a recurrent vein, or the first vein interstitial. Depressed area of male gono- coxite extending much of length of gono- coxite, without carinae across it (examined only in P. sakalava Blithgen). This subgenus is known from Mada- gascar, the Seychelles, and south India, with a possibly incorrectly placed species from Zaire (see Baker, 1974). Genus SPHECODES LatrEILLE (Figs. 13, 14, 20-22, 27, 67-72) Both sexes: Punctation of head and thorax usually coarse, but punctures, espe- cially of mesoscutum, sometimes widely separated by shining ground. Head much wider than long (or scarcely so in S. con- vergens Michener), clypeus about two to usually more than three times as wide as long, often biconvex due to longitudinal median depression which is often reduced to a closely punctate zone or absent. Eyes usually nearly hairless, but with variably long hair in S. diroi Friese and in an un- identified Argentine species, and with scat- tered short hairs in various other species. Hairs of antennal flagellum all or nearly- all very short. Preoccipital carina usually absent in the holarctic region (occasionally present as in S. scabricollis Wesmael), but distinct in some groups from South Amer- ica, south Asia to Australia, and Africa. Pronotum with horizontal surface of collar almost absent medially, but forming lateral angles which are usually prominent (weak and rounded in brachycephalus Mitchell and some small Neotropical spe- cies) and below which a vertical ridge extends downward (ridge weak in forms with weak lateral angles and in S. conver- gens Michener, persimilis Lovell and Cock- erell, etc.), vertical ridge approaching or merging with a more laterally directed ridge that extends toward coxal base, or laterally directed ridge weak or absent in many species; a carina from lateral angle across posterior lobe of pronotum. Anterior extremity of scutum strongly convex, ver- tical anterior surface nearly always with area that lacks punctures (or has them smaller and sparser than in adjacent areas Tue Parasitic Groups oF Harictipar (HyMENopTERA, APOIDEA) 323 “ a 72 | Fic. 66. Eupetersia guillarmodi Michener, wings of paratype female. Fics. 67, 68. Sphecodes mansku (Ray- ment), face and forewing of holotype female. Fic. 69. Sphecodes convergens Michener, face of holotype female. Fic. 70. Sphecodes mansku, dorsum of median part of body of holotype female. Fics. 71, 72. Sphecodes conver- gens. in S. convergens Michener) and is there- fore sharply different from rest of scutum. Scutellum gently convex or sometimes weakly biconvex due to feeble longitudi- nal median depression. Propodeum with dorsal area coarsely rugose, usually strongly areolate, usually about as long as scutellum, shorter than scutellum in a few species [e.g., S. spinulosus Hagens, albila- bris (Kirby), and rufichelis Strand], area broadly rounded posteriorly; posterior and lateral surfaces of propodeum usually with- out or with few short plumose hairs in addition to longer hairs, but such hairs 324 Tue Universiry oF Kansas ScrENCE BULLETIN sometimes rather conspicuous (e.g., in S. heraclei Robertson, prosphorus Lovell and Cockerell, pulsatillae Cockerell, pecosensts Cockerell, rubicundus Hagens, and in the subgenus Austrosphecodes). Wings with hairs short and dense api- cally, somewhat longer and less dense toward bases [hairs rather long and dense throughout in S. manskii (Rayment) |; stigma of moderate size; marginal cell pointed to narrowly truncate at apex; free part of marginal cell beyond submarginal cells as long as to twice as long as part subtended by submarginal cells, which extend well beyond apex of stigma. Sec- ond and third submarginal cells each re- ceiving a recurrent vein, or as occasional variants first recurrent and second trans- verse cubital interstitial, or in some species first transverse cubital absent so that there are only two submarginal cells. First tergum usually broader than long, in various Neotropical Austrosphecodes species as wide as long or even longer than wide. In lateral view base of second tergum sometimes depressed so that there is a weak constriction between first and second terga. Posterior margins of terga 2-4 broadly depressed, hairless, commonly impunctate. Female: Mandible with or without subapical tooth. Labrum with broad, flat apical process which is rounded, truncate, or bilobed and usually about twice as broad as long, but varies to nearly as long as broad [e.g., in S. monilicornis (Kirby) |. Legs commonly robust, hind femur little over twice as long as wide (S. mont- licornis Kirby) to three or more times as long as wide, nearly always strongly con- vex on upper surface near base except in some Austrosphecodes; basitibial plate ab- sent to demarked along posterior margin and at apex; long hairs on outer side of hind tibia plumose, barbed, or (rarely) mostly simple (e.g., in S. monilicornis (Kirby), profugus Cockerell); upper mar- gin of hind tibia with or without spine- like setae; hind tibial spine variable. Fifth tergum, unlike preceding terga, with apical margin fringed except in S. brachycephalus Mitchell, often with dense prepygidial fimbria in front of apical fringe, fringe sometimes interrupted me- dially. Pygidial plate typically broader than in Expetersia, but slender in various species [e.g., S. gibbus (Linnaeus), pulsa- tillae Cockerell]. Male: Antennae long to as short as in female, flagellum often thickened, first flagellar segment broader than long, sec- ond variable, but usually longer than first, the first and second never both very short as in Eupetersia. Labrum with apical process shorter than in female, usually several times as wide as long, but rarely (e.g., S. chilensis Spinola) little over twice as broad as long. Second hind tarsal segment shorter than to longer than third, base broader than or equal to base of third. Gonocoxite usually striate, sometimes with dorsal basal depression, but without dorsolateral margined depression as in Eupetersia. Gonostylus variable, common- ly with basal setose lobe. This genus consists of hundreds of spe- cies and is found on all continents, al- though it is nearly absent in Australia, being represented there by only two spe- cies found in the northern part of the continent. Major papers on the taxonomy of Old World species of the genus are by Blithgen (1923b, 1924, 1927, 1928); com- prehensive taxonomic treatment of New World species is limited to those of eastern North America by Mitchell (1960). Meyer (1919) listed and gave copies of descrip- tions of the species of all geographical regions. The genus is quite variable and doubt- less, with adequate study, could be divided into various subgenera. For the present, however, only two major groups are ac- Tue Parasitic Groups oF HatictipaE (HyMENopTeERA, APoIDEA) 325 corded subgeneric rank, the Neotropical Austrosphecodes and the rest of the genus, Sphecodes proper. The mandible of the female usually has a subapical inner tooth. Sometimes it is reduced to a small tooth close to the main axis of the mandible (e.g., in S. brachy- cephalus Mitchell, costaricensis Friese sensu Michener, 1954, and pycnanthemi Robertson). In various species the sub- apical tooth is absent so that the mandible is simple and sharply pointed. Such man- dibles characterize Austrosphecodes, as well as species such as S. antennariae Rob- ertson, confertus Say and stygius Robert- son. The related genera Eupetersia, Ptlo- cleptis, and Microsphecodes also have sim- ple mandibles. It seems likely that this feature has arisen independently in differ- ent species groups. Male antennae are usually quite elon- gate, with most flagellar segments, includ- ing the second, much longer than broad, and with the first, by contrast, much short- er and broader than long. In various spe- cies, however |[e.g., most Austrosphecodes and antennariae Robertson, atlantis Mitch- ell, costaricensis Friese, cressonit Robert- son, tllinoensis (Robertson) |, the first and second segments are not very different in length, each a little broader than long. The second antennal type is most common in small species. All intergrades between the two types exist (e.g., in heracler Rob- ertson, confertus Say, and chilensts Spi- nola). The legs of females are usually robust, the upper surface of the hind femur strong- ly convex basally, near the trochanter. To a variable extent this is not so in Austro- sphecodes and in a few other species (S. brachycephalus Mitchell). The hind tibia of the female typically has some spinelike or even peglike setae, shorter than the nearby hairs, along the outer margin. At the apex, on the outer surface of the tibia, there is a strong tibial spine. These fea- tures are reduced in some, but not all, small species. In some small species like S. brachycephalus Mitchell, the spinelike setae are few in number and pale, hence dificult to see. The tibial spine is some- times only a rounded prominence |e.g., in S. illinoensis (Robertson) |. In Austro- sphecodes the spinelike setae are absent or are as long as nearby hairs and sometimes with barbs or branches. The same is true of the very large S. rufichelis Strand. The spinelike setae are usually absent in males, but are present in some species |e.g., S. ruficrus (Erichson), spinulosus Hagens]. The preoccipital carina is usually ab- sent, the posterior surface of the head being rounded peripheral to its central concavity, but such a carina is present and very strong in S. mansku (Rayment), distinct in S. chilensis Spinola and some other Austro- sphecodes as well as in profugus Cockerell, scabricollis \Wesmael, and an unidentified African species (see discussion under Eupetersia). The male gonocoxites are typically longitudinally striate over most of the up- per and outer surfaces. Such striae are absent or weak and limited to certain areas in Austrosphecodes. SuscENUs SPHECODES LATREILLE S. STR. Sphecodes Latreille, 1804, Nouvelle diction- naire dhistoire naturelle | Deterville |, Paris, Tableaux méthodiques, 24:182. Type species: Sphex gibba Linnaeus, 1758, monobasic. Dichroa Illiger, 1806, Mag. Insektenk., 5:39. Type species: Sphex gibba Linnaeus, 1758, designation of Sandhouse, 1943, Proce: SsNat.. Muss 92:545; Sabulicola Verhoeff, 1890, Ent. Nachr., 16: 328. Type species: Sabulicola cirsu Ver- hoeff, 1890, — Andrena albilabris Kirby, 1802, monobasic. Drepanium Robertson, 1903, Ent. News, 14: 103. Type species: Sphecodes falcifer 326 Tue UNIVERSITY oF KaANnsAs ScrENCE BULLETIN Patton, 1880, — S. confertus Say, 1837, monobasic. Proteraner Robertson, 1903, Ent. News, 14: 103. Type species: Sphecodes ranuncult Robertson, 1897, monobasic. Sphecodium Robertson, 1903, Ent. News, 14: 103. Type species: Sphecodium cressonii Robertson, 1903, by original designation. Machaeris Robertson, 1903, Ent. News, 14: 104. Type species: Sphecodes stygius Robertson, 1893, by original designation. Dialonia Robertson, 1903, Ent. News, 14: 104. Type species: Sphecodes antennariae Robertson, 1891, monobasic and original designation. Callosphecodes Friese, 1909, Ann. Mus. Nat. Hungarici, 7:182. Type species: Sphe- codes (Callosphecodes) ralunensis Friese, 1909, monobasic. Mandible of female usually with sub- apical tooth. Labral process of male much shorter than in female, three or more times as wide as long. Preoccipital carina usually absent. Legs almost always robust, hind femur two to three times as long as broad, strongly convex on upper surface near base. Hind tibia of female almost always with spinelike setae on outer margin; basitibtal plate of female defined along posterior margin and sometimes also at apex. Sec- ond hind tarsal segment of male shorter than or equal to, rarely longer than, third, its base broader than base of third, so that articulation to first is broader than articu- lation of third to second. Posterior surface of propodeum usually without short plu- mose hairs among long erect hairs. First metasomal tergum considerably broader than long. Male gonocoxites striate on most of outer and dorsal surfaces. This subgenus is not known in South America, but is found on all other conti- nents. As indicated by the discussion and descriptions above, it is quite diverse and may well be subdivided. The status of the name Callosphecodes remains in doubt. It was described by Friese from a large, robust specimen with a metallic blue-black metasoma from New Britain in the Bismarck Archipelago. Un- fortunately, Dr. Jeno Papp of the Hun- garian Natural History Museum reports that the type (and only specimen) cannot be found. The blue color suggests some species of Eupetersia but it is not likely that Callosphecodes is a senior synonym of that name. The locality is far to the east of the known range of Eupetersia, and metallic species of that genus are known only in Africa. Moreover, the robust form (Friese says 3 mm wide, 9-95 mm long) does not suggest Eupetersia. It seems likely that Sphecodes (Callo- sphecodes) ralunensis Friese is related to S. mansku (Rayment) from northern Aus- tralia, as suggested by Michener (1965). Rayment described this form as having a metallic purplish green abdomen, although the type entirely lacks such coloration at present, the metasoma being shining black. S. mansku is a very unusual Sphecodes, and if it is similar to S. ralunensis, the subgeneric name Callosphecodes may be used to unite the two species. Interesting features of S. mansku (Figs. 67, 68) in- clude the unusually large subapical man- dibular tooth of the female, the very strong preoccipital carina, the tooth at the pos- terior end of the hypostomal carina, and the unusually well defined, shining, slight- ly elevated basitibial plate of the female. The hind femur of the female is more slender (3 times as long as wide) and less convex on the dorsal surface near the base than usual for Sphecodes s. str., especially for a large species with conspicuous spine- like setae on the outer margin of the hind tibia and a large, long tibial spine. The fifth tergum has a less dense prepygidial fimbria than most large species, the apical fringe being absent in the middle part of the margin, and a smooth shining hairless area being in front of this fringeless area. A unique feature of S. manski among Sphecodes that I have seen is the vestiture Tue Parasitic Groups oF HaricrmaAr (HyMENoprTerRA, APOIDEA) 327 of the wings, the hairs being long and dense as in Eupetersia. Unfortunately the species is known from only a single female specimen. AUSTROSPHECODES New Suscenus Type species: Sphecodes chilensis Spinola, 1851. This subgenus includes the rather nu- merous South American species of the genus. I have studied unidentified speci- mens of many species, and identified speci- mens of S. mutillaeformis Schrottky, speci- mens of S. chilensis compared with Spi- nola’s type by H. Toro, and of course S. convergens described below. Mandible of female simple. Labral process of male two to three or more times as broad as long. Preoccipital carina often present. Legs relatively slender, hind fe- mur not strongly convex on upper surface near base. Tegula usually larger than in Sphecodes s. str. Hind tibia of female without spinelike setae on outer margin or if present they are as long as the nearby hairs; basitibial plate of female entirely absent. Second hind tarsal segment of male as long as or longer than third, nar- rowed at base like third so that its articu- lation to first is similar to that of third to second. Posterior surface of propodeum with short plumose hairs among long erect hairs. First metasomal tergum vari- able but often as long as broad or longer than broad. Male gonocoxites without striae or striae fine, inconspicuous, and occurring only in limited areas. In various features Austrosphecodes re- sembles Eupetersia. Both have simple mandibles and rather slender legs; some species of Austrosphecodes have the slender body and even a somewhat female-like body form in the male, suggestive of Eupetersia; some also have a preoccipital carina. Austrosphecodes also has short pale plumose hairs on the vertical surfaces of the propodeum, as does Eupetersia. Aus- trosphecodes, however, looks much more like Sphecodes than does Eupetersia, resem- bling Sphecodes in wing vestiture and venation, in basal flagellar segments of the male, and in lacking the margined de- pressed area of the male gonocoxite found in Eupetersia. Presumably the similarities of Austro- sphecodes and Eupetersia are convergences since the latter probably arose from Afri- can Sphecodes and not from South Ameri- can forms. The subgeneric name, meaning south- ern Sphecodes, refers to the South Ameri- can range of the group. APPENDIX Descriptions of new species needed for inclusion in the preceding account, to- gether with other taxonomic conclusions at the specific level, are presented here. Moreover, a revision of the new genus Ptilocleptis is included. Parathrincostoma seyrigi Blithgen Parathrincostoma seyrigi Blithgen, 1933, Mitt. Zool. Mus. Berlin, 18:390, male. New localities are as follows: Mada- gascar: District de Fanovana, Region orientale de forets humides, September- October, 1938 (C. Lamberton) ; east Mada- gascar, forest 600-1200 m, December-Feb- ruary, 1930-31 (C. Lamberton); Manombo s.l., Tulear Province, March 31, 1968 (K.M.G. and P.D.). The Lamberton ma- terial is in the American Museum of Natural History, New York; the other specimen is in the British Museum (Nat- ural History). The female (one specimen from “east Madagascar”) has a strongly produced clypeus, almost as in the male; the antero- lateral angles of the pronotum are pro- duced to acute, almost spinelike angles; the punctation is relatively sparse, punc- 328 Tue UNIversity oF Kansas SCIENCE BULLETIN tures of the frons below the ocelli being separated by much more than puncture widths; and metasomal segments 4-6 are brown, the rest of the body brownish black and the legs more noticeably reddish. Parathrincostoma elongatum Benoist Parathrincostoma elongatum Benoist, 1962, Verhandl. Naturf. Ges. Basel, 73:131. This species, known only from the type female, has a less produced clypeus than P. seyrigi; the anterolateral angles of the pronotum are obtuse; the punctation is denser, punctures of the frons below the ocelli being separated by less than a punc- ture width, and metasomal segments 5 and 6 are brown, the rest of the body and legs being black. Nomia ambrensts (Benoist) new combination Parathrincostoma ambrense Benoist, 1962, Verhandl. Naturf. Ges. Basel, 73:130. As indicated above, the remaining spe- cific name proposed in Parathrincostoma is based on a specimen of Nomuia in the broad sense. It is a slender black bee, with wings dark except for the clear bases, and thus is superficially extraordinarily similar to Parathrincostoma. However the lack of a preepisternal groove, the short prono- tum without a transverse carina or lamella, the pale femoral scopa, the elongate third submarginal cell, and the short face clearly show the incorrect placement. Genus P'TILOCLEPTIS This genus, described above, is known from only seven specimens taken in the American tropics from Mexico to southern Brazil. Three species are easily recogniza- ble, separated by the following key. Key To THE Species o—r PTILOCLEPTIS 1. Terga 2-6 with extensive areas of dense, yellow brown, plumose hair; scutellum of male with an impunc- tate area on each side, but not strong- ly bisibbous 4.222... 5 =e tomentosa Terga 2-6 with hairs mostly simple, not obscuring surface; scutellum of male strongly bigibbous, an elevation on either side of midline being im- punctate at summit 1. eee 2 Scutum with many plumose hairs although surface not entirely ob- scured; scutellum and metanotum yellow (female) 2522 polybioides Scutal hairs mostly simple and sur- face not at all obscured; scutellum and metanotum black (male) .. etchworti bo Ptilocleptis tomentosa new species (Figs. 11, 17, 28, 39-43, 47-50, 53, 54) This middle American species is mark- edly larger than the other members of the genus, from which it also differs by having conspicuously hairy eyes and areas of dense, plumose, brown hair on the metasoma. Female: Length 10 mm; forewing length 8 mm (7 mm in paratype). Black; the following parts red-brown: labrum, mandible, lower two-thirds of clypeus (en- tire clypeus in paratype), scape, pedicel, underside of flagellum (almost whole of two basal segments, red-brown coloration more restricted in area and darker in color toward apex), posterior pronotal lobe, teg- ula, axillary sclerites, tibiae (mid and hind with infuscated postmedian band in para- type), tarsi, apices of femora, apical half of sixth metasomal segment (the last blackish in paratype). Metasomal segments 2-5 with apical translucent brown bands, narrow on terga 2-3, progressively broader on 45. Wings light yellowish-brown, costal area of forewing from stigma to apex of wing (including entire marginal cell) dark brown; veins and stigma brown. Pubescence yellow-brown to yellow- testaceous, on sixth metasomal segment dusky except laterally, that on venter of thorax and in paratype also on side of Tue Parasitic Groups of Haricrmar (HyMeENoprTera, APoIDEA) 329 thorax and posterior surface of propodeum whitish. Hairs of head rather short, yel- low-testaceous, plumose ones hiding most of paraocular area, frons, vertex, and genal area. Thorax with hairs rather short and simple, some long hairs on scutellum and metanotum, surface easily visible except for the following areas which have dense coverings of short, plumose, yellow-testa- ceous hairs obscuring the surface: dorsum of pronotal collar, margin of pronotal lobe and corresponding mesepisternal depres- sion, pre-episternal groove (not entirely covered), line along mesepisternal-meta- plural suture, rest of metapleuron (not entirely covered), line along scuto-scutellar suture, posterior margin of scutellum, disc of metanotum, posterior and posterolateral surfaces of propodeum. Metasoma with scattered, moderately long, erect, simple hairs, very sparse on disc of first tergum, longer and slanting posteriorly on apical terga dorsally, on all terga laterally, and on all sterna; second to fifth terga also with short hairs which are plumose, dense and form broad, yellow-testaceous basal band across tergum 2, basal half of ter- gum 3 similar, otherwise plumose_ hairs less dense and light brown; short, simple hairs on mid-dorsal convex part of tergum 2 and across posterior parts of terga 2-4; hairs on both terga and sterna extending almost to posterior margins, on terga 2-4 short hairs, more erect and laterally di- rected than the rest, forming weak, nar- row, subapical, yellow-testaceous bands. Fifth tergum with hairs somewhat longer than on preceding terga, some arising al- most to apex of tergum, but no recogniz- able subapical band and no dense prepy- gidial fimbria. Eye with rather abundant hairs about one-fourth as long as diameter of scape. Head narrower than thorax, somewhat wider than long (100:88), upper and lower interorbital distances as 59:45, shorter than eye length (65). Clypeus convex, longer than clypeoantennal distance (27:20). An- tennocular :interantennal :antennocellar :in- terocellar:ocellocular distances as 15:14.5: 25:13.5:15. Labrum with broad, transverse, basal, elevated area covered with short hairs, apical process as long as basal area, about four times as wide as long, rounded at apex, margined with narrow band of slender hairs. Supraclypeal area gently elevated up to frontal tubercle, above which it is gently declivous to frons. Scape, pedi- cel, and upper surface of flagellum near base with rather conspicuous hairs, under surface of flagellum with only very short hairs; flagellum with first two segments broader than long, third longer than broad, middle segments slightly longer than broad. Dorsal area of propodeum slightly shorter than metanotum. Hamuli 8-10. First metasomal segment about as long as broad. Head closely punctate, ground between punctures minutely roughened, anterior part of clypeus most coarsely punctate, rest of clypeus and lower supraclypeal area more coarsely so than frons and vertex, which are rather finely punctate except for area of coarser, sparser punctures above antennal base; genal area finely and rather sparsely punctate, lower genal area near foramen magnum striate. Mesoscutum closely punctate, punctures similar in size to largest clypeal punctures, anterior part rugose, punctate. Scutellum much more finely punctate than scutum, with some shining ground between punctures; meta- notum closely and rather finely punctate; mesepisternum above scrobe somewhat more finely punctate than scutum, else- where coarsely reticulopunctate, strigose below; pre-episternal groove a series of pits separated by carinae; mesepisternal- metapleural suture marked by similar series of smaller pits; upper convexity of metapleuron with about four horizontal striae, rest of metapleuron with fine punc- tures and less conspicuous transverse striae, 330 Tue University oF Kansas ScreENcE BULLETIN especially above. Basal area of propodeum with irregular, longitudinal carinae, seven or eight on each side, connected by a ca- rina posteriorly and separating pits which middorsally are twice as long as wide, or more; posterior surface and posterior part of lateral surface of propodeum irregularly coarsely areolate. First metasomal tergum shining with only scattered minute punc- tures. Remaining terga with numerous small punctures, mostly separated by a puncture width or less, extending onto the slightly depressed brownish margins of terga 2-5, but punctures on these margins smaller and = sparser than elsewhere; ground between punctures shining, but minutely lineolate, especially on more pos- terior terga; sterna shining, coarsely and sparsely punctate, apical margins of more anterior sterna broadly impunctate, sur- face between punctures minutely lineolate, progressively more strongly so on more posterior sterna. Male: Similar to description of female, differing as follows: Forewing length 7 mm. Black, the following parts red- brown: scape, first flagellar segment, an- terior tibia (darker brown along outer sur- face). Other parts described as red-brown in female are dark brown. Metasomal ter- ga 1-2 without translucent margins, 3 with a rather narrow translucent brown margin, 4.6 with such margins broad. Hairs of head as well as sides of thorax largely whitish, sparser than in female, par- tially obscuring surface only on paraocular area, scutoscutellar line and center of meta- notum, fully hiding surface under dense plumose yellowish hair only on dorsum of pronotal collar. Basal band of plumose, yellowish hairs across tergum 2 supple- mented by weaker preapical band, strong- est dorsolaterally; similar, yellowish, plu- mose hairs occupying most of dorsum, an- terior to translucent margins, of terga 3-6. Hair of eyes very short. Structure, including facial proportions, about as in female. Second flagellar seg- ment as long as broad, succeeding seg- ments all longer than broad. Scutellum as in female. Genal area somewhat more sparsely, but no more finely punctate than vertex. Upper part of metapleura with only two or three horizontal striae, rest of meta- pleuron areolate. First metasomal tergum shining with sparse large punctures and sparse smaller punctures intermixed. Holotype female: 2 miles west of Pal- mares, Alajuela Province, Costa Rica, Jan- uary 23, 1965 (D. H. Janzen), in the Snow Entomological Museum, University of Kansas. Allotype male: “Mex.,” in the Academy of Natural Sciences of Phila- delphia. Paratype female: Hacienda Cap- olinas, 5 km N.W. of Quezaltepeque, El Salvador, 450 m altitude, Dec. 26, 1964 (M. E. Irwin), in collection of the Uni- versity of California at Riverside. Ptilocleptis polybioides new species (Figs. 25, 51, 52) This Peruvian species differs from the other members of the genus in the yellow pronotum, scutellum, metanotum, and much of the first metasomal tergum, a pat- tern which enhances its wasplike aspect. Female: Length 7 mm; _ forewing length 6 mm. Black, the following parts yellow (reddish-yellow in paratype): la- brum, mandible (apex red-brown), clypeus, scape, pedicel, first two flagellar segments (reddish-yellow) (second dark brown in paratype), prothorax (lamella across front of collar dark), tegula (translucent), axil- lary sclerites, scutellum (reddish), meta- notum, legs, basal two-thirds of first meta- somal tergum, first sternum, and base of second sternum. Lower anterior part of mesepisternum with diffuse, yellow brown area (absent in paratype, which has sides of thorax dark brown). Metasomal terga and sterna with narrow, apical, dark, Tue Parasitic Groups oF Haricripar (HyMENopTeERA, APoIDEA) 331 brownish bands; fifth and sixth segments more extensively brownish. Wings light yellowish brown, costal area of forewing from about level of vein cu-v to apex of marginal cell (including marginal cell ex- cept for narrow strip along its posterior border, or entire marginal cell in paratype) dark brown; stigma and veins proximal to it including basal vein dark brown, other veins yellow brown. Pubescence yellowish-white (that of head and thorax more golden in paratype), that of posterior half of metasoma more brownish-white and of sixth segment dusky. Hair pattern of head and thorax as described for P. tomentosa, but scutum with abundant, plumose hair partly obscur- ing surface, plumose hairs (white) other- wise perhaps less dense and obscuring sur- face less than in P. tomentosa, conspicu- ously less abundant on scutellum. Hair pattern of metasoma as described for P. tomentosa, but short hairs are longer, and none are plumose. Eye with scattered, very short hairs. Head about as wide as thorax, some- what wider than long (75:63), upper and lower interorbital distances as 45:34.5, shorter than eye length (48.5). Clypeus convex, longer than clypeoantennal dis- tance (18:12). Antennocular :interantennal: antennocellar :interocellar:ocellocular — dis- tances as 11:12:225:105:12. Labrum with broad, transverse, basal, elevated area cov- ered with short hairs; apical process about twice as wide as long. Supraclypeal area and antennal pubescence as described for P. tomentosa; flagellum with first segment broader than long, second as long as broad or appearing longer than broad in para- type, the remainder all distinctly longer than broad. Scutellum biconvex, summits of convexities impunctate. Dorsal area of propodeum longer than metanotum. Ham- uli 8. First metasomal tergum about as long as broad. Punctation similar to that of P. tomen- tosa, differing from description of that species as follows: ground between punc- tures of clypeus and supraclypeal area smooth and shining, these areas similarly punctate; striae of lower genal area weak. Anterior part of mesoscutum punctate, but scarcely rugose; mesepisternum above scrobe as coarsely punctate as adjacent areas. Basal area of propodeum with more- regular longitudinal carinae, pits between median ones about three times as long as wide. Impunctate tergal margins broader. Sterna less coarsely punctate, less lineolate. Holotype female: Tingo Maria, Peru, 620 m altitude, October 5-12, 1964 (H. C. Porter), in the Museum of Comparative Zoology, Harvard University. Paratype fe- male: Monzon Valley, Tingo Maria, Peru, November 2, 1954 (E. I. Schlinger, E. S. Ross), in the California Academy of Sci- ences. The specific name is based on the wasp genus Polybia, because of the bee’s resem- blance in general form and coloration to such polybiine wasps. Unlike other known specimens of the genus, the paratype of P. polybioides has only two submarginal cells through loss of the second transverse cubital vein. Puilocleptis eickworti new species (Figs. 44-46) This species from southern Brazil is the smallest of the genus and also the darkest, having neither the rather extensive yellow markings of P. polybioides, nor the large areas of yellowish plumose hairs of P. tomentosa. Male: Length 6 mm; forewing length 5 mm. Black, the following parts red- brown: scape, pedicel, first flagellar seg- ment, tibiae, tarsi, and parts of femora; the following dark brown, sometimes blackish: labrum, mandible, clypeus, pro- notal lobes and diffuse area on side of pronotum, tegula, axillary sclerites, BBY) Tue University oF KAnsAs SciENCE BULLETIN coxae, trochanters, much of femora, and much of first tergum. Extreme base of first tergum yellow, this color continued mid- dorsally to behind middle of tergum. Wings light brownish, marginal cell of forewing dark brown, dark color extend- ing less intensely beyond apex of marginal cell as well as basal to stigma; veins and stigma brown. Pubescence yellowish white, yellow on legs, brownish on metasoma, longer hairs of posterior segments dusky. Plumose hairs (white) less dense than in other spe- cies, not covering surface, most conspicu- ous on face and pronotal collar, present also on sides of thorax and sides and pos- terior surface of propodeum and surface clearly visible in these areas; plumose hairs largely absent on scutum so that surface is clearly visible. Hairs of metasoma and eyes as in P. polybioides. Head about as wide as thorax, wider than long (64:57), upper and lower inter- orbital distances as 40:28, shorter than eye length (42). Clypeus convex, longer than clypeoantennal distance (15:11). Anten- occular :interantennal :antennocellar :intero- cellar:ocellocular distances as 8:11:21:10: 10.5. Labrum with broad, transverse, basal, elevated area with short hairs; apical proc- ess over twice as broad as long. Supra- clypeal area and antennal pubescence as described for P. tomentosa; flagellar seg- ments as described for P. polybioides. Scu- tellum bigibbous, with a strong elevation, impunctate at summit, on each side of midline. Dorsal area of propodeum longer than metanotum. Hamuli 7. First meta- somal tergum distinctly longer than broad. Punctation as described for P. poly- bioides. Holotype male: Nova Teutonia, Santa Catarina, Brazil, March 24, 1966 (Fritz Plaumann) in the Museum of Compara- tive Zoology, Harvard University. Para- type male, same locality and collector (27° 11’ S, 52° 23’ W), no date, im the tsame collection. This species is named for Dr. George C. Eickwort, who assembled most of the known specimens of Pilocleptis and kind- ly permitted me to study them. Eupetersia (Eupetersia) guillarmodt new species (Figs. 60-66) This species differs from other species of the genus by the black body and red scape and legs; other species have the body partly red, yellow, or blue. The species is larger than other species except E. neavei Blithgen, from which it differs in the black metasoma, the broad pygidial plate of the female, the largely black pubes- cence, and other characters. Female: Length 10 mm (9.5 mm in one paratype); forewing length 8 mm. Black, the following parts red brown: la- brum, mandible (except blackish apex which grades into the red basally), scape, pedicel, first two flagellar segments, legs including apices of coxae (but greater part of coxae black). Posterior lobe of prono- tum, tegula, and axillary sclerites of wings slightly darker red-brown; third and fol- lowing flagellar segments dark-brown. Wings fuliginous, veins and stigma dark brown. Longer hairs black, those of labrum, mandible and lower clypeal margin red- dish black, apical fringe of fifth tergum reddish. Anterior surface of front tibia and posterior surfaces of mid and hind tibiae with hair pale-red. Hairs of under- sides of tarsi red. Short, white hair present among longer black hairs on much of head and thorax; most of hair of frons and metapleuron pale; short, pale hairs con- spicuous among longer black hairs on pro- notal collar, around hind wing base and on lateral and posterior faces of propo- deum. Dense fringe on posterior pronotal Tue Parasitic Groups oF Harictar (HyMENoprera, AporpeA) 333 lobe and dense hairs in corresponding de- pression of mesepisternum white. Head as wide as thorax, much wider than long, upper and lower interorbital distances subequal (71:70.5), greater than eye length (60.5). Clypeus scarcely bicon- vex, higher than clypeoantennal distance (19.5:15). Antennocular:interantennal :an- tennocellar : interocellar : ocellocular dis- tances as 22:12.5:27:16:20. Labrum with sharp, transverse, basal ridge with only a few short hairs on distal surface; process beyond ridge about twice as broad as long, rounded, margined with strong hairs, two or three irregularly placed preapical hairs. Supraclypeal area strongly elevated to fron- tal tubercle, then abruptly declivous up- ward between antennal bases. Flagellum with first two segments much_ broader than long, third longer than broad, mid- dle segments slightly broader than long. Preoccipital carina continuous, but weak dorsally. Pronotal collar broader medially than in other species, anterolateral angles sharp although somewhat obtuse, a strong, high carina extending across pronotal lobe. Scutum with longitudinal median depres- sion extending from anterior almost to posterior margin. Scutellum distinctly biconvex. Legs rather slender; basitibial plate with posterior margin indicated by distinct smooth line which curves ante- riorly at apex and indicates end of plate; hairs along upper margin of hind tibia with branches along convex surfaces, spinelike setae absent, but there are some simple bristles especially near base and apex; hairs of outer surface of hind tibia appearing simple in lateral view, but with a few short branches on convex sides visi- ble from above. Apex of marginal cell narrowly subtruncate with appendage. Hamuli seven. Pygidial plate rather broad, but lateral margins converging somewhat anteriorly so that plate is nearly round. Head dull, coarsely and closely punc- tate, transversely rugose-punctate on ver- tex between ocelli and preoccipital carina, punctures finer and arranged to suggest weak striae on genal area; hypostomal area shining, with widely scattered, coarse punc- tures, surface between them minutely shallowly punctulate. Scutum and meta- notum dull, closely punctate, more coarse- ly so than head, transversely rugose an- teriorly on scutum. Scutellum more finely, closely punctate except for the two large, shining, convexities with scattered punc- tures. Mesepisternum coarsely rugose- punctate, becoming more coarsely reticu- late below and ventrally; preepisternal groove consisting of a series of pits sepa- rated by carinae, a similar series of pits along anterior margin of metapleuron; upper convexity of metapleuron with about four horizontal striae, rest of metapleuron rugose-punctate, more finely so below. Dorsal surface of propodeum shining, lon- gitudinal carinae irregular and connected by numerous transverse carinae, so that surface is areolate, at margins areolae of similar size extending down on lateral and posterior surfaces of propodeum, except that anterior, lateral surface is finely punc- tate. First metasomal tergum polished, at first glance impunctate, but with scattered small punctures laterally and even less con- spicuously extending across tergum mid- dorsally. Terga 25 with punctures densest laterally, quite fine on 2, progressively coarser on 3-5; posterior marginal zones of terga 25 broadly shining, impunctate, hair- less; marginal zones as well as surface be- tween punctures nearly smooth on 2, pro- gressively more distinctly lineolate on 3-5; marginal zones slightly depressed, most noticeably so sublaterally because of gently elevated, less closely punctate convexities of punctate areas in front of marginal zones; marginal zone of tergum 2 occupies one third exposed length of tergum, the zones being progressively broader on terga 35, occupying most of exposed length of 5. Anterior sterna with only scattered 334 Tue UNIversiry oF Kansas ScIENCE BULLETIN coarse punctures and large median im- punctate areas; sterna progressively more closely and more finely punctate poste- riorly. Male: Length 10 mm (to 8 mm in one paratype); forewing length 8 mm (to 7 mm in one paratype). Coloration as in female, but posterior lobe of pronotum black; tegular and axillary sclerites brown- ish black. Pubescence colored as in female, but hairs of sixth and seventh terga black or dusky. Upper interocular distance somewhat longer than lower (64:58), subequal to eye length (60). Clypeal length to clypeoan- tennal distance as 20:15. Antennocular :in- terantennal :antennocellar :interocellar :ocel- locular distances as 19:13:26:16.5:185. La- bral process shorter than in female, about three times as broad as long. Third as well as middle flagellar segments about as broad as long. Longitudinal scutal depression weaker than in female and not noticeable behind middle of scutum. Hairs of hind tibia with fewer and shorter branches, those of outer surface simple. Apex of pygidial plate broadly rounded with mar- gin slightly raised. Gonocoxite with de- pressed area about as broad as long with several longitudinal carinae across it; gono- stylus over half length of gonocoxite, broad, rounded, with broad basal ventral lobe. Terga much as described for female, but 5 should read 6 throughout; second tergum with punctures sparse dorsally and surface between them not or scarcely lineo- late. Holotype female: Mamathes, Lesotho (= Basutoland), March 8, 1953 (C. Jacot- Guillarmod). Allotype male, same data but January 2, 1960. Two female and five male paratypes, all from same locality, females taken January 12, 1947 (L. Bevis) and February 18, 1945 (A. Jacot-Guillar- mod), the males on November 19 and 22, 1945 (L. Bevis), January 12, 1947 (L. Bev- is), February, 1940 (C. Jacot-Guillarmod), and February 11, 1945 (C. Jacot-Guillar- mod). The holotype and allotype are in the British Museum (Natural History), para- types in the Albany Museum in Grahams- town and the Snow Entomological Mu- seum, University of Kansas. This species is named for one of its collectors, Charles Jacot-Guillarmod, Di- rector of The Albany Museum, Grahams- town, South Africa. Eupetersia (Nesoeupetersia) flava new species (Figs. 58, 59) This species differs from other members of the genus by the uniformly reddish- yellow color, so that the insect is suggestive of various nocturnal Hymenoptera. E. scottt (Cockerell) and madagasca Blith- gen are species of Nesoeupetersia with partly red head and thorax, but no others have the wholly reddish yellow body of E. flava. Female: Length 6 mm; _ forewing length 5.5 mm. Reddish yellow, apices of mandibles red, upper surface of flagellum brown; wings nearly clear, veins and stig- ma brown. Hairs yellowish white, longer hairs of lower part of face dusky yellow, hairs of tarsi yellow, longer hairs of posterior sterna and sides of posterior terga dusky in certain lights. Head slightly wider than thorax, much wider than long, upper and lower inter- orbital distances as 35:34, about equal to eye length (36). Clypeus uniformly con- vex, upper margin concave medially, height of clypeus greater than clypeoanten- nal distance (8:6). Antennocular:inter- antennal :antennocellar :interocellar :ocelloc- ular distances as 11:5:23:7.5:11. Labrum with sharp transverse basal ridge with a Tue Parasitic Groups of HaricripAr (HyMENopTERA, APOIDEA) 335 few hairs on distal surface; labral process over twice as broad as long, margined with strong hairs and without premarginal hairs. Supraclypeal area weakly elevated to frontal tubercle, then gently descending above, between antennal bases. Flagellum with first three segments all much broader than long, middle segments broader than long. Preoccipital carina continuous, but weak dorsally. Anterolateral angles of pronotum nearly right angular, a weak carina extending across pronotal lobe. Scu- tum without longitudinal depression. Legs slender, basitibial plate slightly elevated and short; hairs along upper margin of hind tibia with branches along convex sur- faces, these intermixed with some coarse, simple, curved, pointed bristles about half as long as longest hairs or longer; hairs of outer surface of tibia as described for E. guillarmodi. Apex of marginal cell point- ed almost on wing margin. Hamuli six. Pygidial plate rather broad, sides diverging anteriorly. Body shining, with scattered minute punctures, lower half of face and entire genal area minutely roughened between punctures, upper part of frons smooth be- tween punctures separated by several punc- ture widths, vertex almost impunctate; thorax with punctures coarser than those of head, punctures of scutum separated by one to two puncture widths, scutellum with similar sized punctures separated by three of four puncture widths, space be- tween punctures minutely roughened, more conspicuously so on scutum than scutellum; sides and venter of thorax smooth, with scattered small punctures, about two horizontal striae on upper con- vex part of metapleuron. Dorsal surface of propodeum with longitudinal, slightly irregular carinae, terminating at posterior margin; posterior and lateral surfaces of propodeum finely roughened and punc- tate, not areolate except for a few, large, incomplete areolae just above attachment of metasoma, carinae delimiting them ex- tending only about half way up from meta- somal attachment to summit of posterior face of propodeum. Metasomal terga en- tirely shining and nearly impunctate, mi- nutely lineolate except on first tergum, weakly so on second; dorsal hairs almost absent except on fifth and sixth terga; basal parts of terga and apical marginal zones not differentiated except sublaterally on terga 1-4, which have distinct sublateral swellings in front of marginal zones. Sterna 1-2 nearly impunctate, 3-6 with coarse punctures posteriorly and laterally, interspaces conspicuously lineolate. Holotype female: East Madagascar, Forest 600-1200 meters altitude, December to February, 1930-31 (C. H. Lamberton, collector), in the American Museum of Natural History, New York. The specific name refers to the uni- formly yellow color of the body. Sphecodes ( Austrosphecodes) convergens new species (Figs. 69, 71, 72) This species differs from other Sphe- codes known to me by the relatively nar- row face, with eyes strongly converging below. In this and other respects it re- sembles the genus Pzilocelptis and suggests the origin of that group from Austrosphe- codes. The abundant, snow white hair is unusual, and the areas of dense, white hair at the sides of the terga probably distinguish it from other species such as S. cordillerensis Jorgensen. Female: Length 7 mm; _ forewing length 5.5 mm. Black, the following parts light red: labrum, mandible (apex red- brown), antenna (upper surface red- brown, basal half of scape brown in some paratypes), pronotal lobe, legs (coxae and in some paratypes trochanters infuscated, front coxa essentially black; hind tarsal segments 1-3 infuscated in some para- 336 Tue UNIVERSITY OF KANSAS SCIENCE BULLETIN types), metasomal terga 1-3 and basal half of 4, sterna (fifth and sixth brownish). Clypeus dark brown. Tegula and basal wing sclerites translucent testaceous. Wings transparent, veins and stigma dark brown. Hairs snowy-white, those of apex of abdomen yellowish; shorter hairs largely densely plumose, obscuring paraocular area and extreme sides of clypeus, slightly less dense on frons and supraclypeal area, most of clypeus with only scattered sparse sim- ple hairs; genal area hidden by dense hair. Thoracic surface largely hidden by plu- mose hairs in the following areas: posterior margin of pronotum and outer surface of pronotal lobe, anterolateral, lateral and pos- terior margins of mesoscutum, lateral part of scutellum, metanotum, lateral and ven- tral surface of thorax, posterior and lateral surfaces of propodeum. Hair of legs rather abundant, longer hairs of hind tibia longer than tibial diameter, mostly branched, none thickened or spinelike; hairs of under surfaces of tarsi pale-yellowish. Lateral parts of terga 15 with diffuse areas of white hairs, mostly not quite dense enough to hide surface, these areas extending me- sally toward midline on middle parts of terga 2 and 3 as areas of laterally directed hairs, conspicuous from some angles. Api- cal fringe of fifth tergum broadly inter- rupted medially so that medially this ter- gum has broad, hairless margin like pre- ceding terga; prepygidial fimbria rather weak, nowhere fully obscuring surface of fifth tergum. Eye with scattered short hairs. Head as wide as thorax, broader than long (about 80:70); eyes strongly converg- ing below, upper and lower interorbital distances as 48:38, both less than eye length (50). Clypeus slightly biconvex, only slightly more than twice as wide as long (37:16), longer than subanten- nal distance (16:14). Antennocular:in- terantennal :antennocellar :interocellar :ocel- locular distances as 14:65:19S:145=iR Mandible simple. Labrum with trans- verse basal elevated area about four times as wide as long, summit of this area feebly emarginate medially, distal sur- face of this area with short hairs; process beyond elevated area rounded, about twice as broad as long, hairs of marginal row slender, not bristles. Supraclypeal area gently convex, frontal carina not steeply declivous above frontal tubercle. Flagellum with first segment almost twice as broad as midventral length, segments 2-9 sub- equal in length, slightly longer than broad (3-5 most distinctly so) to about as broad as long. Preoccipital carina absent dor- sally, distinct laterally and extending to posterior end of hypostomal carina where the two form a posteriorly directed acute angle. Anterolateral angle of pronotum with dorsal, transverse, horizontal carina, angle about right-angular seen from above, no strong ridge, but only a rounded sur- face extending downward from angle, but ventrolaterally directed ridge extending to- ward coxal base distinct. Scutum with vertical, anterior surface convex, with punctures and therefore less sharply differ- entiated from rest of scutum than usual. Scutellum feebly biconvex. Legs slender, hind femur width to length as 12.5:47. Hind tibia without basitibial plate, apical spine, or spinelike setae; hairs of outer surface of hind tibia with branches along convex sides. Apex of marginal cell nar- rowly truncate with short appendage. Hamuli five. First metasomal tergum nearly as broad as long. Pygidial plate rather broad, margins diverging anteriorly. Head closely and rather finely punctate, dull because of punctation except clypeus, which is shining with coarser round punc- tures separated by about a puncture width, and hypostomal area which has only very scattered punctures, but whose surface is minutely roughened, dull. Scutum with punctures nearly as large as those of cly- Tue Parasitic Groups oF Harictipar (HyMENopTERA, AporDEA) 557, peus, separated by about one-third punc- ture width or almost contiguous anteriorly; scutellum more finely punctate; metano- tum finely reticulate; sides of thorax reticu- late punctate, finely so above, weakly and finely so on metapleuron and anterior part of propodeum; upper part of metapleuron with about three horizontal striae. Dorsal surface of propodeum rather finely reticu- late, nearly smooth around the arcuate posterior margin; posterior and posterior lateral surfaces of propodeum more coarse- ly reticulate-areolate. First metasomal ter- gum shining, with small punctures mostly separated by three or four puncture widths, except laterally in some areas where they are coarser and closer and posteromiddor- sally, where they are finer and sparser. Second tergum similarly punctured, de- pressed apical margin not differently punc- tured, coarsest punctures widely separated and on elevation anterior to depressed margin. Remaining terga with punctures closer, surface between them somewhat less shining, marginal areas impunctate posteriorly but basal parts with punctures. Sterna with surface rather dull, finely- lineolate. Holotype female and three female para- types (one headless): Zapala, Prov. Neu- quén, Argentina (Juan Foerster). One paratype is dated March, the others have no dates. One female paratype: Torn- quist, 500 m altitude, Sierra de la Ventana, Prov. Buenos Aires, Argentina, December, 1954 (F. H. Walz). The holotype is in the Snow Entomological Museum, University of Kansas. Paratypes are in that collection, the American Museum of Natural History, and the Museo Argentino de Ciencias Naturales “Bernardino Rivadavia.” The specific name refers to the inner ocular margins which are strongly conver- gent below, unlike other Sphecodes. ACKNOWLEDGEMENTS I am indebted to Dr. Gisela K. Fashing, Williamsburg, Virginia, for aid in assem- bling the data on host relationships of the parasitic halictids. For loan of important type material I wish to thank E. Konigsmann, Berlin; J. C. Cardale, Canberra; and C. Baroni- Urbani, Basel. For the loan of other ma- terial I am indebted to J. G. Rozen, Ameri- can Museum of Natural History, New York; C. Jacot-Guillarmod, Albany Mu- seum, Grahamstown, South Africa; George C. Eickwort, Cornell University, Ithaca, New York; and to the authorities of the British Museum (Natural History). I am indebted to Dr. Denis J. Brothers, University of Natal, Pietermaritzburg, South Africa, for the drawings of whole bees and whole legs, and to Ms. Mary Makepeace for certain other drawings. Especially, I acknowledge suggestions and ideas from Dr. George C. Eickwort that significantly improved the manuscript. This paper was made possible by grant DEB 73-06815 A03 from the National Sci- ence Foundation. EITERATURE,-CLVED Arnstiz, C. N. 1937. Notes on the biology of two panurgine bees. Canadian Ent., 69:97-100. ALFKEN, J. D. 1912. Die Bienenfauna von West- preussen. Bericht. Westpreuss. Bot.-Zool. Ver. Danzig, 34:1-96, pls. 1-2. . 1913a. Die Bienenfauna Bremen. Abh. Nat. Ver. Bremen (1912), 22:1-220. —. 1913b. Die Bienenfauna von Ostpreussen. Schriften physik-okonom. Gesellsch. Konigs- berg i. Pr. (1912), 53:113-182. 1939. Die Bienenfauna von Bremen. Mitt. Ent. Ver. Bremen, 1939:6-30. Atwoop, C. E. 1933. Studies on the Apoidea of western Nova Scotia with special reference to the visitors of apple bloom. Canadian Jour. Res., 9:443-457. Baker, D. B. 1974. Eupetersia, a genus of parasitic halictine bees (Hym., Apoidea) new to the oriental region. Ent. Monthly Mag., 110:59- 63, pl. II. 338 Tue University oF Kansas ScrENCE BULLETIN Batra, S. W. T. 1964. Behavior of the social bee, Lasioglossum zephyrum, within the nest. In- sectes Sociaux, 11:159-185. Benoist, R. 1962. Apides (Hyménoptéres) recueil- lis 4 Madagascar par le Dr. Fred Keiser en 1957-1958. Verhandl. Naturf. Ges. Basel, 73: 107-148. BLAGovEscHENSKAYA, N. N. 1955. The nest building of the moss-footed bees Dasypoda plumipes Pz. in the Ulyanovsk district. Ent. Oboz., 34: 157-163 (in Russian). BLirucen, P. 1916. Ein Beitrag zur Bienenfauna Nortwestthringens. Mitt. Ent. Gesellsch. Halle, 1916(10):10-40. 1919. Die Biennenfauna Pommerns. Stet- tiner Ent. Zeitung, 80:65-131. 1923a. Zur Biologie der Bienengattung Sphecodes Latr. Zeitschr. wiss. Insektenbiol., 18:19-23. 1923b. Beitrage zur Systematik der Bienen- gattung Sphecodes Latr. Deutsche Ent. Zeit- schr., 1923:441-514. 1924. Beitrage zur Systematik der Bienen- gattung Sphecodes Latr. Deutsche Ent. Zeit- schr., 1924:457-516. ——. 1927. Beitrage zur Systematik der Bienen- gattung Sphecodes Latr. III. Zool. Jahrb., Abt. f. Syst., Okol. u Geogr. d. Tiere, 53:23-112. 1928. 2. Beitrag zur Kenntnis der athio- pischen Halictinae. Deutsche Ent. Zeitschr., 1928:49-72. 1934. Die Wirte der paladarktischen Sphe- codes-Arten. Zeitschr. Wiss. Insektenbiol., 27: 33-42, 61-66. 1935. 4. Beitrag zur Kenntnis der athio- pischen Halictinae. Deutsche Ent. Zeitschr., 1935:177-190. BREITENBACH, W. 1878. Uber Halictus quadricinctus Fabr. und Sphecodes gibbus L. Stettiner Ent. Zeitung., 39:241-243. Cuampers, V. H. 1949. The Hymenoptera Aculeata of Bedfordshire. Trans. Soc. British Ent., 9: 197-252. CLAupE-JosEPH, F. 1926. Recherches biologiques sur les Hyménoptéres du Chili. Ann. Sci. Nat., Zool., (10)9:113-268 (translation: 1960, Publ. no. 1, Centro Estudios Ent., Fac. Filos. y Educ., Univ. de Chile, pp. 1-60, pls. 1-18). CockERELL, T. D. A. 1938. Bees of the genus Sphe- codes from the Belgian Congo. Rev. Zool. Bot. Africaines, 30:327-329. E1ckwort, G. C. 1969a. Tribal positions of Western Hemisphere green sweat bees, with comments on nest architecture. Ann. Ent. Soc. Amer., 62 :652-660. 1969b. A comparative morphological study and generic revision of the augochlorine bees. Univ. Kansas Sci. Bull., 58:325-524. anp K, R. Erckwort. 1972. Aspects of the biology of Costa Rican halictine bees, III. Sphecodes kathleenae, a social cleptoparasite of Dialictus umbripennis. Jour. Kansas Ent. Soc., 45:529-541. AND R. Fiscuer. 1963. A morphological study of the tibial spurs of bees of the sub- family Halictinae (Hymenoptera: Halictidae) of eastern North America. Ann. Ent. Soc. Amer., 56:348-354. anp G. I. Stace. 1972. A new subgenus of neotropical Sphecodes cleptoparasitic upon Dia- lictus. Jour. Kansas Ent. Soc., 45:500-515. Faurincer, J. 1922. Hymenopterologische Ergeb- nisse einer wissenschaftlichen Studienreise nach der Tiirkei und Kleinasien (mit Aus- schluss des Amanusgebirges). Archiv. Natur- gesch., (A) 1922 (9) :149-222. Ferton, C. 1898. Sur les moeurs des Sphecodes Latr. et des Halictus Latr. Bull. Soc. Ent. France, 1898:75-77. 1923. La vie des abeilles et des guépes, oeuvres choisies, groupées et annotées par Etienne Rabaud et Francois Picard. Paris: Chiron, xv + 376 pp., 72 figs. Friese, H. 1920. Ueber die Lebensweise der Grab- wespenbiene, Sphecodes Latr. Zeitschr. wiss. Insektenbiol., 12:175-179. -———. 1923. Die europaischen Bienen. Berlin und Leipzig: Gruyter & Co., vii + 456 pp., pls. 1-32. 1926. Hymenoptera, Teil 1, Die Bienen, Wespen Grab- und Goldwespen, in C. Schré- der, Die Insekten Mitteleuropas, vi -+ 192 pp. Frissy, G. E. 1914. Solitary bees. Proc. Holmes- dale Nat. Hist. Soc. [London], 1910-1913: 62-79. Granpi, G. 1961. Studi di un entomologo sugli imenotteri superiori. Boll. Istituto Ent. Univ. Bologna, 35:xv + 660. Grozpanié, S. 1971. Biologische Untersuchungen an den Bienen. Deutsche. Ent. Zeitschr., (NF) 18:217-226. Hicks, C. H. 1934. Some reared insect parasites and their hosts. Univ. Colorado Studies, 21:265- Das JanvER, R. 1976. Grooming and pollen manipulation in bees (Apoidea). The nature and evolution of movements involving the foreleg. Physiol. Ent., 1:179-194. Janvier, H. 1933. Etude biologique de quelques Hyménoptéres du Chili. Ann. Sci. Nat., Zool. (10)16:209-355 (translation, omitting section on parasitic halictines, 1967, Publ. no. 9, Centro Estudios Ent. Fac. Filos. y Educ., Univ. de Chile, pp. 1-58, pls. [1-9]. Knerer, G. 1968. Zur Bienenfauna Niederdster- reichs: Die Unterfamilie Halictinae. Zool. Anz., 181:82-117. 1973. Periodizitat und Strategie der Schma- rotzer einer sozialen Schmalbiene, Evylaeus malachurus (K.). Zool. Anz., 190:41-63. anp C. E. Atwoop. 1967. Parasitization of social halictine bees in southern Ontario. Proc. Ent. Soc. Ontario, 97(1966) :103-110. anp C, PLaTEaux-QuéNnu. 1967. Usurpation de nids etrangers et parasitisme facultatif chez Tue Parasitic Groups oF HarictipAr (HyMENopTERA, APOIDEA) 339 Halictus scabiosae (Rossi). Insectes Sociaux, 14:47-50. 1970. The life cycle and social level of Evylaeus nigripes, a mediterranean _halictine bee. Canadian Ent., 102:185-196. Lecewirz, H. 1925. Zum Problem des tierischen Parasitismus. I Tiel: Die Lebensweise der Schmarotzerbiene Sphecodes monilicornis K. (= subquadratus Sm.). Zeitschr. Morph. Okol. Tiere, 4:430-464. MatysHEv, S. I. 1927. Lebensgeschichte des Colletes cunicularius L. Zeitschr. Morph. Okol. Tiere, 9 :390-409. MarecuHar, P. 1894. [Note on parasitism by Sphe- codes.| Bull. Soc. Ent. France, 1894(9) :cxv. Meyer, R. 1919. Apidae—Sphecodinae. Arch. Nat- urgeschichte, 1919, Abt. A, pp. 79-242. MicuEeNner, C. D. 1953. Comparative morphological and systematic studies of bee larvae with a key to the families of hymenopterous larvae. Univ. Kansas Sci. Bull., 35:987-1102. 1954. Bees of Panama. Bull. Amer. Mus. Nat. Hist., 104:1-176. 1965. A classification of the bees of the Australian and South Pacific regions. Bull. Amer. Mus. Nat. Hist., 130:1-362, pls. 1-15. 1966. The bionomics of a primitively social bee, Lasioglossum versatum. Jour. Kansas Ent. Soc., 39:193-217. 1974. The social behavior of the bees. Cambridge: Harvard Univ. Press, xii + 404 Pp: AND A. Witte. 1961. The bionomics of a primitively social bee, Laszoglossum incon- spicuum. Univ. Kansas Sci. Bull., 42:1123- 1202. Minxiewicz, R. 1935. Myrmosa brunnipes Lepel. et autres Hyménoptéres Aculéates méridionaux ou rares, trouvés en Pologne centrale. Frag- menta Faun. Mus. Zool. Polonici, 2:191-227, pl. ix. MircHeLt, T. B. 1956. New species of Sphecodes from the eastern United States. Jour. Elisha Mitchell Sci. Soc., 73:206-222. 1960. Bees of the eastern United States, I. North Carolina Agric. Expt. Sta. Tech. Bull. [Raleigh], 141:1-538. Morice, F. D. 1901. Observations on Sphecodes. Ent. Monthly Mag., 37:53-58. Moscuier, A. 1938. Ein Beitrag zur Bienenfauna in Ostpreussen, insbesondere der kurischen Nehrung. Schriften physik.-skonom. Gesellsch. Konigsberg i Pr., 70:243-288. NieEvsen, J. C. 1903. Om Bislaegten Sphecodes, Latr. Ent. Mettl., 1903:22-30. Orpway, E. 1964. Sphecodes pimpinellae and other enemies of Augochlorella. Jour. Kansas Ent. Soc., 37:139-152. Perkins, R. C. L. 1887. Notes on some habits of Sphecodes, Latr., and Nomada, Fab. Ent. Monthly Mag., 23:271-274. 1889. Is Sphecodes parasitic? Ent. Monthly Mag., 25:206-209. ——. 1919. The British species of Andrena and Nomada. Trans. Ent. Soc. London, 1919:218- 319, pls. xi-xv. Rau, P. anp N. Rau. 1916. Notes on the behavior of certain solitary bees. Jour. Anim. Behavior., 6:367-370. Rapp, O. 1945. Die Bienen Thiringens unter be- sonderer Berticksichtigung der faunistisch- oekologischen Geographie. Museum fir Nat- urkunde, Erfurt, 150 pp. Ricuarps, O. W. 1944. Observations on aculeate Hymenoptera. Proc. R. Ent. Soc. London, (A) 19:133-136. Rogsertson, C. 1901. Some new or little-known bees. Canadian Ent., 33:229-231. 1926. Phenology of inquiline and nest mak- ing bees. Psyche, 33:116-120. Rozen, J. G., Jr. 1965. The biology and immature stages of Melitturga clavicornis (Latreille) and of Sphecodes albilabris (Kirby) and the recog- nition of the Oxaeidae at the family level. Amer. Mus. Novitates, 2224:1-18. SaKAGAMI, S. F. anv J. S. Moure. 1962. Sphecodes russeiclypeatus, n. sp., obtido de um ninho de Dialictus (Chlorictus) seabrat (Moure, 1956). Bol. Univ. Parana, Zool., no. 2:1-6. Scuoiz, J. R. 1912. Nestbau des Halictus quadri- cinctus F. Jahresheft Ver. schlesische Insektenk. Breslau, 1912(5) :18-19. SHINN, A. F. 1967. A revision of the bee genus Calhiopsis and the biology and ecology of C. andreniformis. Univ. Kansas Sci. Bull., 46: 753-936. SLADEN, F. W. L. 1895. Sphecodes rubicundus and other bees near Dover. Ent. Monthly Mag., 1895 :256-258. STorcKHERT, F. K. 1933. Die Bienen Frankens. Beiheft Deutsch. Ent. Zeitschr., 1932:vii ++ 294 pp. 1954. Fauna Apoideorum Germaniae. Abh. Bayerischen Akad. Wiss., Math.-naturwiss. Klasse, (n.f.)65:1-87. STOCKHERT, E. 1919. Beitrage zur Kenntnis der Hy- menopterenfauna Frankens. Mitt. Munchner Ent. Gesellsch., 9:4-12, 17-32, 37-49. Torcuio, P. F. 1975. The biology of Perdita nuda and descriptions of its immature forms and those of its Sphecodes parasite. Jour. Kansas Ent. Soc., 48:257-279. Torka, V. 1913. Die Bienen der Provinz Posen. Zeitschr. Naturwiss. Abt., Deutsche Gesellsch. Kunst Wiss. Posen, Abt. 20:97-181. VaLkeEILa, E. 1962. Ohjeita vasta-alkajille myrkky- pistidisten kerailyssdé. Ylipainos Lounais- Hameen Luonto, 13:73-85. Yarrow, I. H. H. 1943. Collecting bees and wasps. Amateur Ent., 7:55-81, pl. 1. AY S07 eA) pean walt Salsing et i) bse eo 5 iy AY a) pe \ y } , . 5 Mig tin et > f ad A) tye an | a a iF . . ; ; : Si i) 3 : Tuy en mi 7? m On) ash 7 7 e 5 2 ma 6 a F uty ve id b , - U # Se ee My > , LC Ws ~ ~ : . 2 ' Many ’ - 9 ’ - ' ; { iii 5 WHS Se 02184 — \N om 1978 e ° 3 January 27 aceettaratecasetreteteranetens etetetetee SY ee"ere” terete orene. Oe) GENUS NEOPANORPA By Helen Chwei-Sia Chau and George W. Byers sere" 07 ee seatetatetocets Re ‘ ELLE EE EEE SSEEE SSS REECE EPEC a THE UNIVERSITY OF KANSAS SCIENCE BULLETIN THE MECOPTERA OF INDONESIA Vol. 51, No. 11, pp. 341-405 mS SS RR SE RR RR OS I OO SASS Lee Ee eee ett ae atatasetatnttahetat tate atatetattseeetattesntetatteneasattanets eoreteeceaeteniassentatstetasetetocnsetocs: a a af seeresecetecetetecececeteneetecesetetesetececeees oe SISSIES ESSE EESTI SSIES SEE SEES SSSS SSS SSE SASS SSPE: ce ANNOUNCEMENT The University of Kansas Science Bulletin (continuation of the Kansas Uns- versity Quarterly) is an outlet for scholarly scientific investigations carried out at the University of Kansas or by University faculty and students. Since its incep- tion, volumes of the Bulletin have been variously issued as single bound volumes, as two or three multi-paper parts or as series of individual papers. Issuance is at irregular intervals, with each volume prior to volume 50 approximately 1000 pages in length. Effective with volume 50, page size has been enlarged, reducing the length of each volume to about 750 pages. The supply of all volumes of the Kansas University Quarterly is now ex- hausted. However, most volumes of the University of Kansas Science Bulletin are stil] available and are offered, in exchange for similar publications, to learned societies, colleges and universities and other institutions, or may be purchased at $20.00 per volume. Where some of these volumes were issued in parts, individual parts are priced at the rate of 2 cents per page. Current policy, initiated with volume 46, is to issue individual papers as published. Such separata may be purchased individually at the rate of 3 cents per page, with a minimum charge of $1.00 per separate. Subscriptions for forthcoming volumes may be entered at the rate of $20.00 per volume. All communications regarding exchanges, sales and subscriptions should be addressed to the ExcHANcE LiprariAN, UNIVERSITY or Kansas Lipraries, LAwRENCE, Kansas 66045. Reprints of individual papers for personal use by investigators are available gratis for most recent and many older issues of the Bulletin. Such requests should be directed to the author. The International Standard Serial Number of this publication is US ISSN 0022-8850. Editor Eugene C. Bovee Editorial Board William L. Bloom Philip W. Hedrick Rudolf Jander Harvey Lillywhite Charles D. Michener Norman A. Slade Henry D. Stone George W. Byers, Chairman hd oO THE UNIVERSITY OF KANSAS SCIENCE BULLETIN Vol. 51, No. 11, pp. 341-405 January 27, 1978 The Mecoptera of Indonesia: Genus Neopanorpa HELEN CHwer-S1a CHAu AND GeEorcE W. Byers TABLE OF CONTENTS foe TESOL UPR IT ee ak RN is Eo 2 ee ne 343 Op PSL PD TETL TESTBED. al ee YS ey eee ae en Ee ae Ene ee ee ee 345 IEE CRE TSP RTT See ERT AE Gh, eee Se see ee a ee ae ee seat Aces cuelect seco cee 345 UACSTAB SVT ES iE © ata ele ee ee a a a ee oe 346 TopoGRAPHY, CLIMATE AND VEGETATION ...........2------------0-------- RTE a Rs ee One ed We 349 jos 1 STR OSCTOUE 1d 7s A a ees ON WNT Bee ME A Pee (Aen Saree ee eee 354 OTS TSS hae aot Sea I a ee en 356 PET? De UNIE TSS) (OTRIIN Cay 77171 7377 MRD mwa oe NG este EE Pe ee en 357 CTE Ti LESBRIA UBTSS (GTS YE) 212) (10) 7 01 as ne Sena Se =e RE ne 358 [DS STSTE STDIN: EeB AGES ST 9 Se ee ee 359 NEOPANORPA MUELLERI GROUP ........-...2------2---220-00--- Se ENE Se Be Me Ae Ee 359 EE eee AON CESS) AN Sn sen a a 359 INS SOLEIL DS) 0S) PLES eS oe Re ae 366 IN, YORE RELIES AMEN NLS) 2 EOS ses ee 369 105 EARS DIL TI Toe UN IP CS DE) ST ee oo ees CO et 372 Min E INGE AMANCEIST TAPICEIEA GROUP 52.5 on hee es Lalas Bac Ad Siete zsses 376 DNB tists PEC Gy gINOW ESCOICS) Se5 ore. carn A es eS A oda eate ae 376 NT EUDICT: GUNES IY SOLS eco 8 ace ee me Sr OME ae ne 378 iRGE AMOR ME ERAC TRAM GROMD se ees ler fe ee ee tk eevee 380 INF EP REPETITIONS) ot ee Pn Ee as ena So Or 380 IIS GET IET od CEG IOVS SY Se a ned ea ee re 385 INice SARAIVA WINEW {SPCCIES) a... 2 3 ss eta Be ee ee ed aan es Dee 388 ele CIDANORPAGROR NEE RSIS" GROUP ese oe hae a 2 a Sel Basescu ee cccpeneedance cen ooeeeae secede 39] INT. (QGIAEIE PGE MEN SS lace a ne nae 391 IN TOLER Dig Sa a ee ee ee eee 392 INES ARE, VB SSCS a eee cee errr 395 DR eae Fe CNET SVE OES eee ee eve er tec a ee eee ceed ta ee cceanee ses 396 Jaa SON EREIT NASER ES EY 03 CICS Ne Ce a CEE 396 CoMPARISON BETWEEN NEOPANORPA AND LEPTOPANORPA_ ...........------ tins Had) elie bio Reet Res Pa 398 MEMBERS SOD SUCHR I UNIO PED Ae pate cd ne on ee oan Sach dese se chen sea ean ocicatecdeeteencpes 401 oorsep em okt ere Maas RE ae he ee ne gE oS ae Oe RO 2 Pe eee eer a 402 gers veseusre cea (CHT Sane eg a ee ee 404 Tue Mecoptera oF INponesiA: GENus Neopanorpa 343 The Mecoptera of Indonesia: Genus Neopanorpa' ABSTRACT The Mecoptera (scorpion flies) of Indonesia are limited to species of the genera Lepto- panorpa and Neopanorpa of the family Panorpidae. Our detailed examination of all available museum specimens of Indonesian Neopanorpa indicates they form four groups: A Neopanorpa mueller: group, including N. muelleri, N. hyalinata, N. fuscicauda n. sp., and N. umbonata n. sp.; a Neopanorpa angustiapicula group, including N. angustiapicula n. sp., and N. diloba n. sp.; a Neopanorpa fractura group, including N. fractura n. sp., N. crinita n. sp. and N. sumatrana n. sp.; a Neopanorpa borneensis group, including N. borneensis, N. flavicauda and N. spicata. One other species, Neopanorpa lieftincki n. sp., cannot be placed with any of the groups. Established species are redescribed and new species described. The genera Neopanorpa and Leptopanorpa are compared and distinguished. INTRODUCTION Indonesian insects of the order Mecop- tera, commonly known as scorpion-flies, have been studied by van der Weele (1909) and Lieftinck (1936), and inci- dentally by a few others. Mecoptera of some other areas of southeastern Asia have also been examined in detail: Japan (Mi- yake, 1913; Issiki, 1933), China (Cheng, 1957), and Indo-China (Byers, 1965). The Mecoptera are represented in Indo- nesia only by the family Panorpidae and by only two of its three genera, Lepto- panorpa and Neopanorpa. Weele’s work concerned only three forms of Neopanorpa (which we think are all of a single spe- cies) and four species now assigned to Leptopanorpa. Lieftinck later dealt in de- tail with Leptopanorpa only. The primary objectives of this study are to describe the Indonesian species of Neopanorpa and to attempt to show how they are related to one another and to the mainland species of the genus. Another purpose is to eval- uate the characters used to differentiate Neopanorpa and Leptopanorpa to deter- mine whether these nominal genera are distinct. * Contribution no. 1641 from the Department of Entomology, The University of Kansas, Lawrence, Kansas 66045, U. S. A. Supported by NSF Research Grants GB-7045X and GB-30837. Weele (1909) first proposed Neopa- norpa as a subgenus of Panorpa. West- wood (1842) described its type species, Panorpa angustipennis, from either Java or Tenasserim (peninsular Burma), not explaining why he was uncertain of the locality. The female holotype is la- beled, “Java.” In revising Panorpa (1846), Westwood reversed the order of possible type localities as “Tenasserim, India, vel Java.” Others, however, have regarded angustipennis as a Javanese species (Weele, 1909; Enderlein, 1912; Esben-Petersen, 1913; Roepke, 1916), but the Javanese species they described was actually Neo- panorpa muellert, as Banks (193la) sup- posed. Byers (1965) judged that angusti- pennis is an Indo-Chinese (mainland) spe- cies, rather than Indonesian. He showed that the thoracic markings and the genital plates of the holotype of angustipennis are characteristic of mainland species, and the wing markings, while very similar to those of muelleri, are also like those of some other Indo-Chinese species. He also found that one of two female Malayan specimens matched the type, and judged that those and associated males were con- specific with the holotype. Weele (1909) separated the Indonesian 344 Tue UNIveRsITY oF KaAnsAs ScrENCE BULLETIN Mecoptera into two genera, Panorpa Lin- naeus and Leptopanorpa MacLachlan. Be- cause he assigned to the genus Leptopanor- pa those males with length of the abdomen about twice that of the wings, he placed in Panorpa four species currently considered as Leptopanorpa. Weele differentiated subgenus Neopanorpa from typical Pa- norpa because the notal organ of males of the Indonesian species is conspicuous, whereas that structure is but slightly de- veloped in the European species of Pa- norpa with which he was familiar. He also separated from Panorpa (Neopanor- pa) muelleri a new, smaller subspecies, ungaranensis, that has the pterostigmal band divided into two spots in the hind wings. Karny (1923) said that those characters had no taxonomic significance and he could not tell typical muellert and m. ungaranensis apart, for some smaller individuals do not have the pterostigmal band of the hind wing broken into two spots and some larger ones do. Enderlein (1910) erected the genus Campodotecnum for Panorpa angustipen- nis Westwood and the genus Himantu- rella for Leptopanorpa tubifera Enderlein and L. nematogaster MacLachlan. In 1912, he retained angustipennis (actually muel- leri) in Campodotecnum and described two new species, Campodotecnum lemnis- catum and C. cingulatum. He raised Neo- panorpa to generic status, but designated as its type species Panorpa nematogaster MacLachlan (now in Leptopanorpa), not knowing that the type species, angust- pennis, had been fixed by Weele. These errors were noted by Esben-Petersen (1913) and Roepke (1916). Esben-Petersen (1913) described a new species, Ayalinata, from Java and trans- ferred Panorpa jacobsoni Weele, P. javan- ica Westwood and P. pi Weele to the genus Leptopanorpa. He also synony- mized Campodotecnum with Neopanorpa, and Himanturella with Leptopanorpa. Navas. (1913) accepted Campodotec- num as the generic name for his new spe- cies falcatum from Java and added to that genus the species effusum Navas, from Sikkim. In 1915 Esben-Petersen transferred Pa- norpa nematogaster MacLachlan and Neo- panorpa linguata Navas to the genus Lep- topanorpa, both as synonyms of Lepto- panorpa charpentieri Burmeister. Later, in 1921, he reviewed the Javanese species of Neopanorpa, including as valid species angustipennis Westwood, hyalinata Esben- Petersen, muellert Weele, lemniscata En- derlein and cingulata Enderlein. He sy- nonymized Campodotecnum falcatum Na- vas with Ayalinata and muellert ungara- nensis with muellert. Variation, especially of color, in the latter pair was earlier dis- cussed by Roepke (1916), who presumed that they were the same species. Banks (1931b) described Neopanorpa flavicauda, the first mecopteran recorded from Borneo, and suggested (193la) for the first time, that Weele’s angustipennis is only a form of muelleri. Lieftinck (1936) accepted the new stat- us of some Leptopanorpa species trans- ferred from Panorpa by Esben-Petersen (1913) and described six new species as L. erythrura, L. filicauda, L. inconspicua, L. petersent, L. robusta, and L. sarangana, and one new subspecies, L. pz decorata, adding some morphological and biological observations. He judged that Leptopanor- pa charpentiert, regarded by Esben-Peter- sen (1915) as synonymous with L. nemato- gaster MacLachlan, was from Bengal, not from Java. Burmeister, 1839, had said that he would describe a new species from Ben- gal, but in his description of charpentiert, supposedly that new species, he gave the locality as “Ostindien.” Since charpentiert was unknown to Lieftinck, he thought it was distinct from nematogaster. He also suggested that Campodotecnum lemni- scatum Enderlein was a Neopanorpa and Tue Mecoprera oF INpoNnEsIA: GeNus Neopanorpa 345 that C. cingulatum Enderlein might be identical with his L. erythrura. The for- mer is a synonym of Leptopanorpa pi Weele and the latter, the senior synonym of L. erythrura (Byers, 1967). Lieftinck made the following statement about synonymy of Leptopanorpa and Neopanorpa: “Enderlein, in the Zool. Anzeiger, 35, 1910, p. 192, has erected the generic name Himanturella, with tubtfera Enderlein as genotype, but since tubifera End. 1910 is the same species as longicauda Weele 1909, which is a true Leptopanorpa, Enderlein’s unfortunate selection is invalidated. More- over, Enderlein created a second invalid name in making Panorpa javanica West- wood the type of his new genus Campo- dotecnum, a species belonging undoubt- edly to Leptopanorpa. In 1912, Enderlein adds further to the confusion in placing his Himanturella as a synonym of Neo- panorpa, while nematogaster 1s erroneous- ly fixed as the type of Hinanturella End. Thus, as has been clearly ascertained by Esben-Petersen in his ‘Synonymic List of the order Mecoptera’ (Entom. Meddelel- ser, 10, 1915), Neopanorpa Weele 1909 (type: P. angustipennis Westwood) is of earlier date than Campodotecnum Ender- lein 1910, and should be used as the name of this genus, while Himanturella becomes an absolute synonym of Leptopanorpa MacLachan 1875, of which ritsemae is the type.” Byers (1966) described two additional species of Neopanorpa from Borneo and redescribed N. flavicauda Banks. The only key for Indonesian Neo- panorpa (Weele, 1909) concerned primar- ily those species now assigned to Lepto- panorpa. ACKNOWLEDGEMENTS We are deeply indebted to Dr. M. A. Lieftinck, of Rhenen, Netherlands, retired, who helpfully arranged the loan of speci- mens and gave us information on collect- ing localities and other label data. Special thanks go to Dr. D. C. Geijskes who sent us all the specimens from the Rijksmu- seum van Natuurlijke Historie, Leiden, Netherlands, and permitted us to study and dissect the lectotype of Neopanorpa muelleri, and to Dr. John E. H. Martin who sent us specimens from the collection of the Canada Department of Agriculture. We thank, too, the many persons who col- lected the specimens on which this study is based, particularly K. W. Dammerman, F. C. Drescher, E. Jacobson, H. H. Karny, M. A. Lieftinck, H. Lucht, M. C. Piepers, L. J. Toxopeus and M. E. Walsh. METHODS AND MATERIALS We examined 395 dried, pinned speci- mens, nearly all from the Riyksmuseum van Natuurlijke Historie, Leiden. Most type specimens of older species were ex- amined and drawn (by camera lucida) by Byers in 1964. The lectotypes of mueller and wngaranensis were reexamined in 1972 and that of muelleri dissected. Terminology for wing venation is that of Comstock and Needham; for wing markings, that of Esben-Petersen (1921); for male and female genitalia, by Byers (1965). The genital bulb was removed from the male (sometimes also the 7th and 8th abdominal segments), and the 8th and following abdominal segments from the female. These were boiled in water for one or two minutes to soften them and to prevent breakage during dissection. For the male, the ninth sternum (including hypovalves) and ninth tergum were re- moved together from the genital bulb. For detailed examination of genitalia the ae- deagus was excised by cutting along the inner bases of the basistyles. Dissected parts were preserved in glycerin in micro- vials, these being attached to the individ- ual specimen’s pin. For the female, the 346 Tue UNIversiry oF KAnsAs ScIENCE BULLETIN terminal segments were boiled, the sub- genital plate of the 8th sternum cut off, and the exposed genital plate was removed. These were preserved on an insect mount- ing point in a drop of polyvinyl alcohol, which is water soluble, so that the dis- sected parts can be washed off when re- examination is desired. Line drawings were made with the aid of a camera lucida. For wing photographs, specimens were relaxed for two days be- fore removal of the wings. Measurements of antennae, wings and body length were made with dividers and a millimeter scale. Body length was meas- ured as the distance from the front of the head to the end of the dististyles for the male, from front of head to tips of cerci for the female. Wing length is the straight- line distance from point of the wing’s at- tachment to its apex. This is adequate, since precise measurements are not signifi- cant in taxonomy of Mecoptera and the method gives a general impression of the insects’ sizes. MORPHOLOGY GENERAL DEscrIPTION: Adults of Neopanorpa (Fig. 1) are me- dium-sized insects. The body length is about 14 to 19 mm for males and 11 to 16 mm for females, depending on the species. Wing length is about 12 to 14 mm, variable among species. Head: Elongation of the clypeus, subgenae and certain mouthparts produces the characteristic beak-like rostrum. The compound eyes are large and black; three large, transparent, brown ocelli are close together on a raised triangle. The term “dorsum of head” used here descriptively refers to the top of the head, from the antennal sockets to the post-occipital mar- gin. Chewing mouthparts are at the end of the rostrum. The small labrum is not separated by a distinct suture from the greatly elongated clypeus. The mandibles cross each other on preserved specimens. The stipites of the maxillae are greatly elongated and the submentum is elongated and membranous. Tips of maxillary and labial palps are usually darkened. The antennae are long (about 12 mm) and slender, each composed of a thick- ened scape, a nearly spherical pedicel and a flagellum, slightly tapered from the base to the apex, consisting of 40 to 46 cylindri- cal segments or flagellomeres (number varies with species and individuals). Thorax: The wide pronotum bears bristles or hairs along its anterior margin. Below the pronotum are an undivided pleuron and the front coxae. At each side, between the propleuron and the posterior margin of the head, is a large cervical scle- rite. The mesonotum is clearly divided into scutum, scutellum and postnotum, the first two separated by the scutoscutellar suture, and the scutellum from the post- notum by a membranous, intersegmental conjunctiva. The pleural suture is dis- tinct, dividing the pleuron into epister- num and epimeron. The metathorax is about the same size and form as the meso- thorax. The mesothoracic spiracle is in a membraneous area between the pronotum and mesepisternum, and the metathoracic spiracle is between the mesepimeron and metepisternum. Legs: The coxae and mera are well developed, especially on the mesothorax and metathorax. The femora and tibiae are long and each is of rather uniform diameter throughout. At the apex of each tibia are two large, tibial spurs. There are five tarsomeres, basitarsus much longer than the others. The two pretarsal claws bear five pectinations each. Wings: There are two pairs of long, narrow wings. Wing shape and venation are useful in generic diagnosis. For Leptopanorpa, the wing is narrower than for Neopanorpa, especially the wing Tue Mecoprera oF Inpones1a: Genus Neopanorpa 347 Fic. 1. Neopanorpa muelleri (Weele), male, left lateral aspect; wings elevated to show venation; right legs and most of antennal flagella omitted. Abbreviations: A—anal veins; AB—apical band; bs—basistyle, Cu—cubitus; ds—dististyle; hvy—hypovalve (abdominal sternum 9); M—media; pnp-——posterior notal process; pp3t—pos- terior process of abdominal tergum 3; PTB—pterostigmal band; R—radius; Sc—subcosta; 9t gum 9, abdominal ter- 348 Tue University oF KANsAs SCIENCE BULLETIN base. Neopanorpa differs from Panorpa in the length of vein 1A, which reaches the wing margin before the level of the origin of the radial sector in the former, well beyond it in the latter (Esben-Petersen, 1921). The wing markings (Fig. 1) are smoky dark brown to faint brown, de- pending on the species. They include: An apical band (AB) extending from anterior to posterior margin at the apex of the wing; a pterostigmal band (PTB), which usually extends from anterior to posterior margin and may or may not fork at mid-length into a proximal (basal) branch and a distal (apical) branch, but which may be reduced to a pterostigma only (hyalinata, angustiapicula); a mar- ginal spot extending generally from C to near R4+5; a basal band from C to 1A or Cuz, sometimes broken into two spots; and a basal spot near the base of Cui and Cuz. The wing markings help in primary identification, varying from the most com- plete pattern (presence of all the mark- ings, as in typical muelleri) to the most re- duced, obscure one (diloba). They may vary much within one species (muvellert) ; or the same pattern may occur in different species (umbonata, muelleri and suma- trana). The hind wings are usually slight- ly smaller than the fore wings and have less extensive markings. Abdomen of the male: There are nine, readily visible abdominal segments. The first tergum is divided into an an- terior sclerite joined to the metathorax and a separate posterior sclerite. The median posterior process of tergum 3 (upper part of the notal organ) projects backward from the usually shallowly emarginate ter- gum and rests on a small elevation on an- terior tergum 4. Weele (1909) used this to differentiate Neopanorpa as a subgenus from typical Panorpa. It varies from spe- cies to species, sometimes enough to be used in species diagnosis. The tergum and sternum are fused and the pleura obliter- ated in segments 6-8. Paired spiracles are present on the first eight segments. The enlarged ninth segment (genital bulb) is connected to the eighth by a short pedicel. The ninth tergum (epian- drium, or preépiproct; Fig. 1, 9t) bears two subapical processes that extend ven- trally around the tenth segment. In Neo- panorpa, the apex of tergum 9 has a shape that distinguishes this genus from Lepto- panorpa, but it does not vary sufficiently within the genus to be useful in species recognition. The ninth sternum (hypan- drium) is divided apically into two hypo- valves, the length, shape and curvature of which are useful as specific taxonomic characters. Between the ninth tergum and hypo- valves are the stout, paired basistyles (gon- ocoxites, or coxopodites; Fig. 1, bs). At the apex of each is a chelate dististyle (stylus, or harpagone; Fig. 1, ds), the usually slender, apical half of which is curved mesad and tapers to a sharp point. Projecting from the inner margin of the dististyle near its attachment is the basal lobe (basituberculus of ‘Tjeder, 1970), which varies in size and shape according to the species. For most species, such as muellert or those of the fractura group, the basal lobes are knob-like, but for a few (the angustiapicula group) they are flat- tened (Figs. 83, 87). Two to four long, stout, curved, black spines project mesad from the dorsal surface of the basal lobe in the muellert group (Fig. 26), but are absent in other Indonesian species. Smaller, curved, black spines (4 to 25 in number) may also occur in a group on the antero- dorsal surface of the basal lobe. The structure of the aedeagus assists specific diagnosis but has not been used in earlier taxonomic treatments of Indonesian Neopanorpa. Most of the aedeagus is only weakly sclerotized, but certain parts are firm and darkly pigmented. The aedeagus (Figs. 17, 27) consists of a pair of blunt Tue MecoprTera oF INDoNEsIA: GENus Neopanorpa 349 ventral valves (penis) slightly divergent at their apices, a usually somewhat smaller pair of dorsal valves, and associated ap- pendages. It is braced between the basi- styles by the lateral processes. These proc- esses are simple for nearly all Indonesian species, but in the angustiapicula group they are complexly divided into two lamel- lae enclosing a deep pocket. Ventral para- meres are not present on Indonesian Neo- panorpa except in the angustiapicula group, where they are large, project ven- trally from the genital bulb and are partly visible, laterally (Fig. 71). The dorsal parameres (penunci of Tjeder, 1970) are ordinarily small, vertically oriented blades of varying shape, above the dorsal valves of the aedeagus. They are, however, com- plicated for fuscicauda and the fractura group, extending ventrally alongside the ventral valves of the former and bearing a small, flattened, dorsal appendage, pro- jecting dorsad, on the latter. Segments 10 and 11 comprise the proc- tiger. They are small and ordinarily tele- scoped beneath tergum 9. The single- segmented cerci arise between segments 10 and 11. Abdomen of the female: The ab- domen of the female of Neopanorpa is very similar to that of Leptopanorpa. Ten clearly recognizable segments taper gradu- ally to the rear, and paired spiracles are present on the first seven. No lateroter- gites are present on the seventh and eighth segments as on some Panorpa. Part of segment 8 is modified as the subgenital plate, and the tergum and sternum are fused on segment 10. The two-segmented cerci arise from the so-called cercifer (cer- cal bases, in Ferris, 1939) of the tenth segment. The eighth and ninth segments include external female genitalia. The subgenital plate (subgenitale) of sternum 8 is broad, usually notched and whitish apically (ex- cept in angustiapicula), and bears long setae on each posterior lobe and along the apical, outer margin of some species (Fig. 28). Above the subgenital plate are the so-called genital plates (internal skeleton, in Miyake, 1913; medigynium, in Tjeder, 1970). These provide the best specific characters for the female. Each (Fig. 29) is composed of an axial portion (gonoclavi of Tjeder, 1970) above and behind which are two blade-like arms (distal plate or lamnae), twisted somewhat at their bases. The axial portion is usually elongate in Indonesian species, but is short and oval for a few (fractura group, fuscicauda, an- gustiapicula group and borneensis group). The taxonomic characters used by most authors who have dealt with Indonesian Mecoptera, such as differences of size, of wing markings and body color, and over- all shape of the genital bulb in males, are not sufhcient for species distinction. For example, Weele (1909) separated a new subspecies, Neopanorpa muelleri ungara- nensis, from his new species muelleri by its smaller size and the division of the ptero- stigmal wing band into two spots in the hind wing, but he ignored the identical structure of the genitalia of the two forms. Some authors later regarded certain indi- viduals of muelleri as Javanese specimens of N. angustipennis Westwood because of their general appearance, especially wing pattern; but they ignored the absence of thoracic markings on muelleri and its dis- tinctive genitalia. Wing pattern and body color have taxonomic value only when as- sociated with genitalic and other char- acters. Accurate distinction of species re- quires comparison of the combination of all available characters. TOPOGRAPHY, CLIMATE AND VEGETATION Indonesia now includes Sumatra, Java, Sulawesi (Celebes), Kalimantan (Indone- sian Borneo), Nusa Tenggara (the Lesser 350 Tue University of Kansas ScrENCE BULLETIN Sunda Islands), Maluku (the Moluccas, including Ceram and Halmahera), Flores, Sumbawa, part of Timor, West Irian (ap- proximately the western half of New Guinea) and some 3000 smaller islands. In the past, the name Indonesia was ap- plied to various combinations of these and adjacent territories. Most of these islands belong to young, Tertiary mountain sys- tems that lie completely in the tropics and within the Indo-Australian monsoon re- gion. Indonesia therefore has a climate with high temperatures, high humidity and abundant rains. The mean annual temperature at sea level is about 27° C (80° F) and the mean humidity 80%. The montane climate is more temperate. There is a decrease of temperature between 5.5 and 6° C (10-11°F) for each rise of 1000 m, e.g., the average temperature is 22.1° C (71.8° F) at Bandung, 730 m above sea level, and only 15.9° C (60.8° F) at Tosari, 1734 m in altitude (Bemmelen, 1949). Since Mecoptera are known only from Sumatra, Java, Borneo and possibly Hal- mahera, the following geographical com- ments pertain only to these islands. Sumatra (Fig. 5), located southwest of the Malay Peninsula, has an area of 435,- 000 sq. km. It is 1650 km long by 100-200 km wide in the northern part and about 350 km wide in the southern part. It lies between latitude 5°45’ N and 6° S and longitude 95°20’ E and 106° E. The Bari- san Mountains extend along the entire southwestern coast and separate the north- eastern and southwestern coastal lowlands. The slope toward the Indian Ocean is generally steep. The eastern and southeast- ern parts of the island are jungle lowlands traversed by five major rivers. The south- western coastal range (Barisan Mountains) contains many peaks from 2000 m to 4000 m high, often of volcanic origin, some still active. With the equator crossing Sumatra near its center, the climate is equatorial. The average, annual, lowland temperature is about 27° C (80° F); rainfall is heavy. The slopes, exposed to a continual south- western wind, receive more than 300 cm of rainfall, annually, that is distributed fairly evenly over the island (Robequain, 1954). The warm, moist air favors growth of dense forests (myrtaceous and ficaceous trees, oaks, guttapercha, camphor, teak, pine, etc.) that cover about 90% of the island. In the forests, the Sumatran Neopa- norpas flourish. Hesse (1937) believed en- demic animal species of Sumatra were primarily found south of the mountains; but most of the six endemic Sumatran species of Neopanorpa are distributed on both the southwestern and _ northeastern slopes of the moutains. Hesse (1937) also thought faunal differences between the two coastal lowlands of Sumatra exceed the differences between the northeastern slopes and the Malay Peninsula; but, again, this is not the case with the Su- matran species of Neopanorpa. The spe- cies of Sumatra have not been found in nearby Malaya. Java (Figs. 2-4), located southeast of Sumatra and south of Borneo, between the Java Sea and the Indian Ocean, has an area of 127,000 sq. km. It is 975 km long by 206 km wide at its widest part and lies between 5°52’ and 8°47’ S and 105°13’ and 114°37’ E. Three physiographic proy- inces can be distinguished: 1. West Java (west of the longitude of Tjeribon) (Fig. 2). 2. Central Java (between Tjeribon and Semarang) (Fig. 3). 3. East Java (east of Semarang and Jogjakarta) (Fig. 4). The west-to-east mountain chain in- cludes 112 major volcanoes, 35 still active. The northern coast is lowland, of volcanic deposits or marine sand and clay, but the southern coast has limestone ridges. In East Java, the mountains are generally near the southern coast. These include Java’s highest volcanic peak, Mahameru Tue MecopTera oF INpoNnEsIA: GENus Neopanorpa 25) Fic. 2. Map oF WESTERN JAva. Circles indicate localities where Mecoptera have been collected. Numbers, in sequence from left to right, correspond to locality numbers appearing in the alphabetical gazetteer. Black spots indicate major cities; D—Djakarta. Contour lines at 100 m, 1000 m, and 2000 m, to indicate coastal lowlands, intermediate elevations, and high mountains. Fic. 3. Map oF CENTRAL Java. See Fig. 2 for explanation. J—Jogjakarta, S—Semarang, T—Tyjeribon. 352 Tue University oF KansAs SciENCE BULLETIN (3711 m; 12,060 ft), and the volcanic Tengger, Ijang and Ijen highlands, which form a wide peninsula extending eastward toward Bali. In Central Java, the moun- tains are more central and the main water- shed is toward the south. Here, the vol- cano Slamet reaches 3472 m. The Preanger Mountains in southeastern West Java are lower (the major peak less than 3000 m) and include extensive upland basins. Fur- ther westward the summits are lower, but rise again in the northwest to nearly 2000 m. There are four major rivers, draining generally northward. The coastal temperature of Java aver- ages 26-27° C (78-80° F), while temperate conditions occur above about 700 m. In general, humidity is high (759-907). The Asiatic (northwest) monsoon, from No- vember to April, brings much rain, espe- cially in the north. From May to October, the southeast monsoon season brings some rain to the southern coast but in general is drier than the northwest monsoon. Western Java (except northwest Java) and the entire southern coast have high annual rainfall (more than 200 cm), quantity de- pending largely on the altitude and situa- tion of the mountains. Bogor (Buiten- zorg), 290 m above sea level, has more than 406 cm of rain per year; Bandung, at 730 m, 195 cm. Maximal rain falls in the Djampangs, on Mt. Tangkuban Prahu, Mt. Malabar, various mountains of East Priangan, and Mt. Tjerimai (Tjaréme) in Tjeribon. In Central Java, annual rainfall is 669 cm on the upper northern slopes of the mountains, but only 215 cm at Sema- rang on the northern coast. The rainfall is not so evenly distributed as in West Java. There is an obvious dry season from June to September. Semarang receives only 13°% of its total rainfall during the four-month dry season. In Central Java, the maxima of rainfall are found on Mt. Slamat and the range east of it, including Mt. Ungaran, the Merbabu and Merapi volcanoes. East Java is drier than Central Java. The foehn-like, southeastern, moon- soon winds from the Australian deserts are generally warm and relatively dry. They bring less rainfall to the eastern end of the island than to the central part, eastern volcanic peaks becoming islands of heavier rainfall. The northeastern coastal low- lands get only 150-200 cm of rain, annually, the months of July to October being prac- tically dry (Bemmelen, 1949). In East Java, regions of maximal rainfall are on Mt. Lawu, the Tengger Mountains, the southern slopes of Mt. Raung and Idjen Plateau. Rainfall decreases from western Java to eastern Java, and because the southern mountain slopes force the monsoon winds to rise, there is also a decrease from south to north. This pattern of rainfall coincides with the distribution of Javanese Neo- panorpa (Figs. 2-4), just as it does with that of Leptopanorpa (Lieftinck, 1936). The genus Neopanorpa requires high at- mospheric humidity. The distribution and mode of life of Indonesian scorpion flies are closely related both to the total amount of rain and the seasonal rainfall pattern. Java’s many volcanoes and its heavy equatorial rainfall contribute to the fertil- ity of the soil and the abundant plant and animal life. The moist mountain slopes have an extensive evergreen forest with the tall timber tree, rasamala, common in the west, but absent in the east where mountain Casuarina is plentiful. Teak is more frequent and the evergreen rain for- est less robust in Central Java than in West Java. Many species of scorpion flies are apparently restricted to the luxuriant forest in the mountains of western Java. Mayr (1944) called the fauna of Java im- poverished, compared to that of Sumatra (or Borneo). However, collections of scorpion flies suggest the opposite. Mayr thought the volcanic lava and ashes during the Pleistocene may have exterminated Tue Mecoprera oF INponEsIA: GENus Neopanorpa 353 Fic. 4. Map oF Eastern JAva. See Fig. 2 for explanation. many local species and that Java is less humid and poorer in habitats than Suma- tra, more peripheral and less accessible to colonizing species. We suggest that hu- man depletion of forests in Java (20-30% of the surface covered with forests, as con- trasted with 90°% in Sumatra), destroying much of the shaded habitat of scorpion flies, must have had a profound effect upon the present fauna. Borneo, the second largest island of the Malay Archipelago (736,000 sq. km), is roughly triangular, with two small penin- sulas at its eastern side, and lies between latitude 7°3’ N and 4°20’ S and longitude 108°50’ E and 119°20’ E. Its northern and central parts are hilly and mountainous, but mostly under 1500 m. The chief ranges extend northeast to southwest, with west- ern, southwestern and eastern branches. The highest point is Mt. Kinabalu, 4485 m (13,456 ft), in Sabah, the northernmost part of the island. Alluvial plains between ranges are often swampy near the ocean because of tidal flows. Most of Borneo has evergreen rain forest of rich and varied vegetation. Coastal temperatures usually range from 28-34° C (82-93° F). The rainfall is often violent; mean, annual rainfall is about 310 cm at Sandakan. The few records of Mecoptera from Borneo are from highlands around Mt. Kinabalu. Since Panorpidae usually frequent moist, northern, continental, temperate for- ests, their penetration into Indonesian trop- ics requires explanation. Perhaps panorpids migrated there from the Asiatic mainland when Indonesia was much cooler than now. As the climate warmed, the panor- pids moved to suitably cool and forested habitats on mountain slopes, where they survive today. Some of the species appear to us to have narrow vertical limits. N. angustia- picula is found only above 2000 m; N. lieftincki at approximately 1900 m; N. 354 Tue University oF Kansas ScrENCE BULLETIN fractura at 1400 m; N. diloba and N. fusct- cauda around 1000 m; N. sumatrana be- tween 920 and 1200 m; N. crinita between 500 and 1000 m. Those species having nar- row vertical limits are also often limited to one area only: N. angustiapicula, N. diloba and N. fuscicauda have been found only in East Java; N. lieftincki and the fractura group in Sumatra only. Some other species have a much broader vertical distribution: N. muelleri can be found from low country (at 100 m, 500 m and 1000 m) to above 2000 m altitude; N. um- bonata ranges from the coast to 920 m; and N. hyalinata is found from 300 to 1850 m. Other species tend to have wider horizontal or overlapping distributions: N. muelleri occurs throughout the islands of Java and Sumatra; N. hyalinata occurs both in West Java and East Java but as yet is unknown in Central Java. N. muelleri is the only species known from both Java and Sumatra (and possibly Halmahera). N. muelleri might have developed before these two islands became separated. The doubtful record of N. muelleri from Halmahera is discussed elsewhere in this paper (see Zoogeography). ZOOGEOGRAPHY Neopanorpa occurs only in southeastern Asia, where 83 species (including 8 new species from Indonesia) are known. Thirty are in nine provinces of southern and southeastern China, including Sikang, Szechwan, Yunnan, Kweichow, Anhwel, Kiangsu, Kiangsi, Fukien, Kwangtung (Cheng, 1957); 17 are found in southern India, southern Nepal, Sikkim, Assam and Burma (Rust and Byers, 1976); 14, in Indo-China (Byers, 1965); 9, in Taiwan and 13 (including the 8 new species) in Indonesia. There appears to be no overlap of specific ranges from one of these re- gions to another. This may be due to restricted ranges, but more probably to in- sufficient collecting. The 13 Indonesian species of Neopa- norpa are restricted to Java, Sumatra, and Borneo (and perhaps Halmahera); one occurs in Java and Sumatra (possibly also Halmahera), 4 others on Java, 5 others only on Sumatra and 3 others only on Borneo. There is as yet no record of Mecoptera from Celebes, Lesser Sunda Is- lands (Lombok, Sumbawa, Flores, Timor, etc.) or New Guinea. Lack of records of Neopanorpa and other Mecoptera from these islands may be due to inadequate collection or to physical barriers, or both. In early Miocene or late Oligocene, the earth’s crust buckled downward along the western Sumatran islands (Simeulue, Nias, Kepulauan, Batu, Siberut, Sipura, etc.), Timor, Kai, Ceram and Halmahera (Earle, 1845, in Mayr, 1944). In late Mio- cene, a second folding formed parts of Sumatra and Java. At first most of this fold was under water, some islands not emerging until mid-Pleistocene. The Pleis- tocene ice-age lowered the sea level to about 100 m lower than now, establishing several land connections between present islands (map, pp. 16, 17, Townes and Chiu, 1970). Sumatra, Java, Balreand Borneo and the Malay Peninsula formed “Sundaland,” an extension of the Asiatic mainland. Lombok, however, remained separated by the 312 m depth of Lombok Strait (Mayr, 1944). We presume that Mecoptera from mainland Asia migrated across these land connections to Sumatra, Java, Borneo and the Philippines during the cool Pleistocene. The Indonesian Mecoptera (all in the family Panorpidae) have no taxonomic familial connections with Mecoptra of Aus- tralia (families Bittacidae, Choristidae, Nannochoristidae, Apteropanorpidae and Meropeidae). Instead, the Mecoptera of Australia show connection with those of South America, e.g., the genus Nanno- chorista, occurring also in Chile and Ar- gentina (Riek, 1954). Many elements in Tue Mecoprera oF INponEsia: GENus Neopanorpa 355 Fic. 5. Map oF SumatrRA. See Fig. 2 for explanation of symbols. Contour lines at 200 m, 1000 m, and 3000 m, to indicate coastal lowlands, intermediate elevations, and high mountains. other groups of insects (Edmunds, 1957, and Hardy, 1951, in Gressitt, 1958), in vertebrates and plants are common to Australia, southern South America and, in some cases, Africa (Gressitt, 1958). This tends to support the existence of an orig- inal southern supercontinent (“Gond- wanaland”) which may have existed some 150 million years ago, of which an Aus- tralian fragment broke off and moved northward, joining Asia. Geological changes (sea level, orogenic forces, coral formation, etc.), later broke the connection between Asia and Australia, re-established it and broke it again (Robequain, 1954). Animals and plants originating in Asia developed, in Australia, as unique life forms, due to the long period of isolation. As pointed out by Lieftinck (1936), all the Indonesian Mecoptera are endemic; even the species of Sumatra are not known from nearby Malaya. Of the two Indo- nesian genera, Neopanorpa and Lepto- panorpa, only the former occurs on main- land Asia. Leptopanorpa, as we have lim- ited the genus, is known only from Java, but possibly occurs on Sumatra as well. We have examined the syntypes (1 male now without abdomen and 1 female) of L. charpentieri (Burmeister), described 356 Tue UNiversiry oF KANsAs ScIENCE BULLETIN from “Ostindien.” They bear no locality labels. Lieftinck, in 1936, assumed the spe- cies was from Bengal as indicated in the Introduction above. Esben-Petersen (1921) correctly synonymized charpentiert and nematogaster; Lieftinck did not agree. We believe that nematogaster is a synonym of charpentiert, and we conclude that L. char- pentieri is from Indonesia, not from India. The other two supposed Himalayan Lep- topanorpas, L. furcata and L. effusa, are large Neopanorpa species (Byers, 1971). Endemicity of Indonesian Mecoptera is probably due to evolution in the present ranges following migration of ancestral species from mainland Asia. These mi- grants were isolated by water barriers resulting from the melting of the glaciers during and at the end of the Pleistocene. Warming lowland temperatures prob- ably further isolated these Mecoptera, driv- ing them up into their somewhat isolated ranges on mountains. Mecoptera are not strong fliers; their low vagility intensifies spatial isolation, inbreeding and species formation, giving rise to endemic species or even endemic genera in the particular environments of the isolated tropical is- lands. We now find no species common to Malaya and Sumatra, and only one, N. muelleri, occurring both on Java and Su- matra. Two specimens of N. muelleri in the Canadian National Collection are labeled as from Halmahera. How did these Java- Sumatran species cross approximately 1450 km (900 mi) from Java to Halmahera? Perhaps the specimens are mislabeled. Townes and Chiu (1970: 234) doubted the supposed occurrence of the ichneumonid wasp, Xanthopimpla f. fastigiata, on Hal- mahera, the specimens thereof in question having been collected by the same collector whose label appears on the supposed Hal- mahera N. muelleri. Wallace’s Line may hold its validity for such weakly flying in- sects as Mecoptera, if the labels are correct, for both flying and flightless Indo-Malayan elements have been found even in Timor (Rensch, 1936, in Mayr, 1944); and the straits separating the Indonesian islands have not been absolutely insuperable ob- stacles to the migration of animal and plant species (Robequain, 1954). Mayr (1944) states that each of the straits in the Lesser Sunda Islands is to some extent a zoogeographic barrier, with the Lombok Strait more effective than any of the others, because of its depth and its persistence through periods of low sea level during the Pleistocene. TAXONOMY We divide the Indonesian Mecoptera into four groups, as follows, primarily on the basis of genitalial structure of the males and to a lesser degree on the basis of genitalial characters of the female and wing pattern. One species (lieftincki, n. sp.) must be placed alone. Neopanorpa muelleri group: muellert van der Weele hyalinata Esben-Petersen fuscicauda, new species umbonata, new species Neopanorpa angustiapicula group: angustiapicula, new species diloba, new species Neopanorpa fractura group: fractura, new species crinita, new species sumatrana, new species Neopanorpa borneensts group: borneensis Byers flavicauda Banks spicata Byers Ungrouped species: lieftinck1, new species Tue Mecoprera oF INponEsIA: GEeNus Neopanorpa 357 KEY TO MALES OF NEOPANORPA IN INDONESIA This key is based mainly on readily visi- ble characters, not requiring dissection for visibility. Identifications may be verified by reference to structural details of genital bulb, etc., in the species descriptions. 1. Hypovalves of sternum 9 each bear- ing a lightly sclerotized projection from dorsal surface near mid-length seca ima lateral aspect) 222)... Hypovalves of sternum 9 without Spent prOyectionsyes..- 20. Fs 2. Hypovalves expanded basally and at- tenuate toward apex! 2-2 )--2--t-.+ Hypovalves not expanded basally or not attenuate toward apex ................ 3. Basal lobe of dististyle conspicuously divided into a proximal and a slight- ly larger distal protuberance (fig. 15 4))) ge Nel lie Raa Ree ee flavicauda Basal lobe of dististyle divided into a ventral portion and a more thick- ened dorsal portion (fig. 127) ........ niece te i ee borneensis 4. Dorsal surface of hypovalves bearing fomprhaics (fig) 104) 04 2... crinita Dorsal surface of hypovalves with- out long hairs 5. Apical band and pterostigmal band entire; dorsal sclerotized projections of hypovalves small, forming about a 25° angle with hypovalves (fig. 110) | Sa eee sumatrana Apical band almost absent, or re- duced to spots; pterostigmal band much constricted near mid-length; dorsal sclerotized projections of hy- povalves expanded, forming about a 35° angle with hypovalves (fig. 3.5) | Ae I wae wae de a fractura 6. Hypovalves acuminate in apical half or abruptly narrowed in apical one- iii | es lee AES oe Sele hte ae Hypovalves gradually narrowed, TOUnGednateapex sate tes Ue i. 7. Posterior process of abdominal ter- 2) ne gum 3 extending across about two- thirds length of tergum 4; hypo- valves overlapped mesally near mid- length (fig. 67); pterostigmal band much reduced, with only conspicu- ous stigma remaining (fig. 66); gen- eral body color reddish brown; oc- CULLING sila |AVas ee angustiapicula Posterior process of abdominal ter- gum 3 long, extending to mid-length of segment 6 (fig. 138); hypovalves separated throughout their length (fig. 139); pterostigmal band com- plete; general body color dark brown to brownish black; occurring in BORICO pra eet es ee ee .... Spleata . Wing markings almost obliterated, very faint smoky brown; basal lobe of dististyle without stout, black dor- sal spines fos. 4 20. eee ee eS 9 Wing markings conspicuous, dark smoky brown, even if only ptero- stigma remains; basal lobe of disti- style bearing stout, black dorsal Spimess (dn oe26)) ee Sh ne ae 10 Outer edges of hypovalves greatly infolded, the folded margins sub- parallel in ventral aspect (fig. 142); basal lobe of dististyle concave mes- sally, not flattened (fig. 145) .... lteftincki Outer edges of hypovalves slightly infolded, the folded margins not subparallel in ventral aspect (fig. 81); basal lobe of dististyle flattened, divided into ventral and dorsal parts (fig 3335 OU) xe. 5 ee ee Bone diloba . Wings almost clear, hyaline, with only conspicuous pterostigma dark- ened: (igs: 30-32) 22 hyalinata Wing markings more complete, usu- ally including apical and pterostig- malt bands, -* 236 2 ea ie 11 Ventral and dorsal portions of basal lobe of dististyle distinctly separated, ventral portion flattened, rounded, directed ventrocaudad, not concealed by hypovalves (fig. 56); dorsal sur- 358 Tue UNiversiry oF KANsAs ScIENCE BULLETIN face of dorsal portion of lobe bearing three sizes of spines in three proxi- mate groups (fig. 62) ............ umbonata Ventral and dorsal portions of basal lobe of dististyle either not clearly separated or not far apart, both usu- ally concealed by tips of hypovalves, directed mesad; dorsal surface of lobe bearing two sizes of spines in two separated groups 2.2.22 12 Genital bulb uniformly dark brown to blackish brown; basal lobe of dis- tistyle about as wide as long (fig. 7g) ects laa La aoa hae ee an enc fuscicauda Genital bulb mostly yellowish brown to brown (brown throughout in a few individuals); basal lobe of disti- style much less wide than long (figs. Doe Open een eee ee muelleri KEY 10° FEMALES OF NEOPANORPA IN INDONESIA Specific identification of females usu- ally requires examination of the subgenital plate and the genital plate, the latter visible by simple dissection described earlier. The female of diloba is unknown; that of flavi- cauda has not been examined in detail. ie Wing markings much reduced or nearly obliterated; markings may be only faint brown, or apical band may be reduced to a few spots, pterostigmal band either entire or interrupted near mid-lIength; or wings may be almost clear and hy- aline, with only pterostigma remain- Up eaten Paes | 20) in eae el Be B Wing markings dark smoky brown, usually including entire apical band and pterostigmal band (includes flavicauda and probably diloba) . Pterostigmal band reduced to ptero- stigma only, or with only a small triangular extension behind stigma. 3 Pterostigmal band faintly indicated, or entire or interrupted near mid- len ethicae. RO SAPS Lee eee ee 4 3. General body color reddish brown; apical band usually present but not entire; axial portion of genital plate ovoid (figs. 66, 77) ........ angustiapicula General body color dark brown to blackish brown; apical band absent, wings almost clear-hyaline except for conspicuously darkened stigma; axial portion of genital plate forked anteriorly (figs. 30, 43) ............ hyalinata . Wing markings faint brown, much reduced and almost obliterated; axial portion of genital plate forked an- teriorly, without lateral lobes (figs. PT 150) ee lieftincki Wing markings dark smoky brown, distinct; apical band reduced to spots, pterostigmal band either en- tire or interrupted near mid-length; axial portion of genital plate ovoid with rounded lateral lobes -............... 5 . Pterostigmal band entire, with con- spicuous proximal branch, distal branch obliterated or represented by a spot; subgenital plate narrowed smoothly to apex; lateral lobes of genital plate conspicuous, darkly sclerotized (figs. 101, 110) -......... crinita Pterostigmal band interrupted or greatly constricted near mid-length, proximal branch usually reduced to a spot; subgenital plate narrowed abruptly near mid-length; lateral lobes of genital plate inconspicuous, visible only by transmitted light (figs:,89,99)) 2:0 fractura . Axial portion of genital plate forked anteriorly 226 coecec hae 7 Axial portion of genital plate not forked “anteriorly (2.2.0 eee 8 . Tips of anterior branches (apo- demes) of axial portion of genital plate thick, bent sharply laterad (fig. OA) Ee Eee ee ee um bonata Tips of anterior branches of axial portion of genital plate slightly thick- ened, not bent (fig. 29) ............ muellert Tue MecopTera oF INponEsIA: GeNus Neopanorpa 359 8. Axial portion of genital plate short, conical; bases of arms of plate ex- panded to form a darkly sclerotized, subquadrate plate cleft medially above apex of axial portion (fig. 5) at ea a fuscicauda Axial portion of genital plate short but ovoid; bases of arms of plate not forming a subquadrate plate —.......... 9 9. Genital plate without rounded lat- eralmobes (ie. 140) Prot e eee spicata Genital plate with rounded lateral MODES Gir. Ace mene ese ae Cee a 10 10. Subgenital plate narrowed from near mid-length to apex, with pale median streak or zone (fig. 130) .... ctl a borneensis Subgenital plate narrowed in _ pos- terior one-third, without pale median zone but with pale area just before apical notch (fig: 120) 2....._-..- sumatrana DESCRIPTIONS OF SPECIES NEoPANORPA MUELLERI Group Neopanorpa muelleri Weele, 1909 Panorpa miilleri Weele, 1909. Panorpa miillert ungaranensis Weele, 1909. Panorpa angustipennis; Weele, 1909 (mis- identification). Van der Weele based his description of muelleri on one male and five females and that of muelleri ungaranensis on one male and four females. The male specimen of each of these is hereby designated as re- spective lectotype. Head: Dorsum glossy _ blackish brown to black; frons below antennal sock- ets dark brown to blackish brown; rostrum reddish brown to dark brown anteriorly, yellowish brown to brown laterally or laterobasally; mouthparts reddish brown to dark brown, black on tips of palps in some specimens. Antennal scape yellowish brown (lectotype) to brown, pedicel brown (lectotype) to dark brown; flagellum dull dark brown basally, remainder dull black- ish brown to black; 41 flagellomeres. Thorax: Pronotum black, with 25 bristles at each side along anterior margin. Mesonotum and metanotum black, with small dark brown areas near wing bases in most specimens; paler on scutellum in some specimens. Pleura and coxae dingy yellowish brown, dark brown near wing bases in a few specimens, to dull dark brown throughout; femora yellowish brown to dark brown; tibiae and tarsi dark brown. Wings (Figs. 6-14): Almost color- less, slightly iridescent, with smoky dark brown bands and spots. Apical band usu- ally complete, with variously shaped proxi- mal prominence in both fore and hind wings, or reduced to narrow darkened zone along apical margin and one or two other slender spots, usually along cross- veins (six specimens). Pterostigmal band usually complete, narrowly joining apical band along costal margin, not forked in most specimens. Marginal spot conspicu- ous to small or absent. Basal band entire, of variable width, or reduced to one or two spots, or absent. Basal spot present or ab- sent. Abdomen of male: Terga 1-6 black- ish brown to nearly black; sterna 25 yel- lowish brown to dull dark brown; segment 6 paler ventrally than dorsally. Segments 7-8 yellowish brown, reddish brown or dull dark brown. Posterior process of tergum 3 narrowly triangular, extending about half-way across tergum 4. Hypovalves of sternum 9 darkened or not (dark brown at least basally, where they are separated by a nearly circular space, below which is a semicircular, pale membranous area), widened at mid-length, often overlapping, paler and covered with long hairs apically (Fig. 15). Ninth tergum yellowish brown to dark brown, paler and shallowly emar- ginate apically. Cerci pale at base, dark brown at apex. Basistyles mostly dark 360 Tue University oF KansAs ScriENcCE BULLETIN Fics. 6-8. Neopanorpa muelleri (Weele), wings, showing variation in degree of pigmentation. 6, specimen from Central Java; 7, from West Java; 8, from East Java. See also Fics. 9-14. Tue MecorTera oF INDONESIA: GENus Neopanorpa 361 Fics. 9-11. Neopanorpa muelleri (Weele), wings, showing variation in degree of pigmentation. 9, specimen from Sumatra; 10, 11, specimens from West Java. See also Fics. 6-8, 12-14. 362 Tue University oF Kansas ScrENcE BULLETIN brown, lighter brown where concealed by hypovalves. Dististyles darkened basally, paler apically, or of uniform color through- out, outer margin slightly concave near base, abruptly curved near apex. Ventro- mesal concavity of basal lobes of dististyles medium (lectotype) to large, with pendant mesal and lower margins (Figs. 22, 24); 2-4 large, apically curved, black spines pro- jecting from dorsal surface of basal lobes (Fig. 26); cluster of 8-25 apically curved smaller spines projecting from anterodor- sal surface of basal lobes; inner pendant margins of basal lobes bearing hairs (Fig. 25). Aedeagus lightly sclerotized, reddish brown; ventral valves blunt, slightly di- vergent at apex, projecting a little beyond dorsal valves; lateral processes broad; dor- sal parameres compressed, blade-like, pro- jecting dorsad; ventral parameres absent (Figs. 17, 27). Abdomen of female: Terga 16 dark brown to black, corresponding sterna sor- did yellowish brown to dark brown. Seg- ments 7-10 dark brown. Cerci black. Sub- genital plate of sternum 8 (Fig. 28) yellow- ish brown to dark brown, notched and whitened apically, bearing setae on each side of notch and along outer margin, with or without laterobasal translucent, membranous areas. Axial portion of geni- tal plate of variable length, anterior apo- demes straight, only slightly divergent; arms of distal plate long, twisted basally (Fig. 29). Body length: Male, 12.3-15.6 mm; fe- male, 9.0-145 mm. Length of fore wing: Male, 12-13.6 mm; female, 7.9-14 mm. Holotype: Male, collected at Bahia, Java (no date recorded), by S. Miller. Additional specimens examined: West Java: 1 8, Mt. Malang, Tengeh, Djampangs, 200-800 m, Jan. 1940, Walsh; 2 ¢ 4, the same data as above, except Dec. 1937, Walsh and 1931, Betrem; 13 6 6, 14 @@, Omg. Sukanegara, Djampangs, 700-1000 m, 23 and 28 Dec. 1931, Lieftinck; 26 6, the same data as above, except 700 m, 19 April 1935, Toxo- peus; 1¢, 19, Bibidjilan, Banten (district) south coast, low country, 1937, Walsh; 1¢6, 229, Mt Bunder, Salak, 6 April 1931, Lieftinck; 2 ¢ 4, Depok, Karang, 26 Dec. 1920 and 17 Feb. 1924; 2 29, Preanger, 1500 m, Oct. 1937, Jacobson; 1é, 19, Mt. Gedeh, 1200-1700 m, Dec. 1912 and 12 Dec. 1929, Jacobson; 17 ¢ 6,18 2 2, Mt. Gedeh, Tapos, 700-800 m, 1 Nov., Apr. 1933 and Aug. 1936, Kalshoven; 1 @, same data as above, except 16 Oct. 1932, Lieftinck; 3 ¢ 6, 6 2 9, Mt. Gedeh, Tapos, 1200 m, without date, Kalshoven; 4 @ 6, 4 99, Mt. Guntur, Kamodjang, 1400 m, May 1935 and Sept. 1938, Overbeck; 3 646, 5 @9, Mt. Karang, Bantam, 800 m, 27 May 1931, Lieftinck; 1 @, the same data as above, except 1000 m, Oct. 1930, Paine; 8 ¢ ¢6, Kawa Kamodjan, near Garut, Mt. Guntur, 1450 m, May 1935, Overbeck; 1 Q, Mt. Madiun, Java, 1200 m, 10 Jan. 1930, Leeuven; 1 @, Mt. Manglajang, Preanger, 1400 m, 11 May 1931, Toxopeus; 2 ¢ 46, 1 2, Mt. Megamendung, 800-1000 m, 27 March and 25 Sept. 1932, Lieftinck; 1 ¢&, Pangerango, Oct. 1908, Jacobson; 1 @, Pa- teungteum, 1300 m, 24 Noy. 1935, Toxopeus; 1 6,2 2 Q, Patjet, Mt. Bésér, 1300 m, 30 Sept. 1934 and 10 March 1940, Lieftinck; 1 6, 1 Q, Pelabuanratu (Wynkoops Bay), southern West Java, low country, Nov. 1935; 1 @, Pantjar Mts., Mt. Gedeh, 1500 m, 8 July 1931, Tutbing; 1 6, Pre- anger, Piepers; 1 Q, Radjamandala, 335 m, June 1935, Jacobson; 2 ¢ ¢@, Selabintana, near Sukabumi, Mt. Gedeh, 1000 m, Dec. 1938, Walsh; 5 ¢@ 6, 15 @ 2, Sukabumi; 1 ¢,3 2 9, Mt. Tangkubanprahu (Tangkuban Prahu), Preanger, 1333-1700 m, 4 Jan. 1924 and Jan. 1936, Drescher; 4 6 6,5 2 Q, the same data as above, except 1400 m, 3 Nov. 1940, Olthof; 2 ¢ 4,3 92 @, Tjibodas, Mt. Gedeh, 1400 m, 22 May 1935, Vecht; 3 @ 2, the same data as above, except 1550-1700 m, 27 July 1930 and 28 March 1932, Lieftinck; 1 92, Tjiomas, Warung Loa, Salak, 9 Nov. 1932, Lieftinck; 1 ¢, 2 929, Tjipeundeng, Leuwibiang, 900 m, 22 July 1934, Lieftinck; 5 ¢ ¢,1 92, Tjisarua, Mt. Gedeh-complex, 1050: m, 6 July 1930; Lieftincks Oo "GeGeuiees Tjisarua, Mt. Panggerango, Mt. Gedeh, 1000-1400 m, 27-30 Aug. 1931, 2 Jan. 1936, 30 May 1937, and 16 April 1950, Lieftinck; 3 6 6, Mt. Tjisura, Tengah, Djampangs, 500 m, Sept. 1932, Walsh; 2 992, Ungaran Gunung, 800-1200 m, Oct. 1909, Jacobson. Central and North Java: 2 ¢ 6,3 @ 9, Kaliu- rang, Djokjakarta, C. J., low country, Noy. 1936, Overbeck; 2 @ @, Mt. Muria, Macrodjambangan, N. J., 1000 m, 14 and 17 Dec. 1935, Walsh; 3 6 ¢, 10 22, Mt. Raung, Bajukidael, C. J., 500-800 m, 8 May, Sept. and Dec. 1932, Lucht; 4 6 6, 4 2 9, Mt. Telamojo, res. Kedu, C. J., 1200-1400 m, 29 Oct. 1939, Lieftinck; 2 6 ¢, 4 9 2, Tjolo, Muria Mts., C. N. J., 800 m, 20 and 24 Oct. 1939, Lief- tinck. East Java: 6 6 6,9 @ 9, Blawan, Idjen Plateau, 950 m, 12 Dec. 1934, 15 March and 26 Dec. 1935, 5 May and 13 Dec. 1936, Lucht; 1 6, 1 9, Mt. Kendeng, 2000 m, 13 Nov. 1933, Lucht; 2 ¢ 6, Tue MecoprTera oF INpoNnEsIA: GeENus Neopanorpa 363 ail Sa aa i so eee ae ee ee i Ss ™ = Pi . a 14 Fics. 12-14. Neopanorpa muelleri (Weele), wings, showing variation in degree of pigmentation. 12, specimen from West Java. 13, specimen from Central Java; note broken pterostigmal band in hind wing, a character used by Weele in differentiating subspecies uwngaranensis from typical muellert. 14, fore wing only of specimen showing broken pterostigmal band. See also Fics. 6-11. 1 2, Ongop-ongop, Idjen Mts., 1600-1850 m, May 1924, Dammerman and 27 Dec. 1935, Toxopeus. Sumatra: 1 @, Air Tarbis, Dec. 1913, Jacobson; 4 2 9, Brastagi, northeastern Sumatra, 950-1400 m, 12 Nov. 1950, Lieftinck; 1 ¢, 4 9 9, Giesting, Mt. Tanggamus, southwestern Lampongs, southern Su- matra, 800 m, 27 Dec. 1934, Lieftinck; 1 6,2 2 @, Serapai Kur., July 1915, Jacobson; 2 ¢ @, Suban Ajam, July 1916, Jacobson; 4 66, 3 29, Tan- djunggadang (west coast), 1000-1200 m, Oct. 1925 and Feb. 1926, Jacobson. Neopanorpa muelleri is the commonest Indonesian Neopanorpa, with the broadest range of habitats and of vertical or hori- zontal distribution. It ranges from low country to above 2000 m, throughout the islands of Java and Sumatra, and possibly on Halmahera. This relatively great range probably has helped produce some of the complicated variation within this species. The wing markings vary (Figs. 6-14) from clear except for the pterostigmal band and reduced apical band to a complete pat- tern including apical band, pterostigmal 364 Tue UNIVERSITY OF KANsAs SCIENCE BULLETIN Fics. 15-24. Neopanorpa muellert (Weele), structural details of male. 15, genital bulb of lectotype, ventral aspect. 16, same, right lateral aspect. 17, aedeagus, lectotype, ventral aspect. 18, abdominal terga 3 and 4, showing notal organ, dorsal aspect, reconstructed from lectotype. 19, same, left lateral aspect. 20, ninth abdominal tergum, lectotype, dorsal aspect. 21, same, right lateral aspect. 22, right dististyle, specimen from Sumatra, dorsal aspect, showing only three large spines and many smaller spines on elongate basal lobe. 23, right dististyle, specimen from Sumatra, dorsal aspect, showing only two large spines and smaller group of small spines on shorter basal lobe. 24, left dististyle, male lectotype of subspecies wngaranensis (Weele), ven- tral aspect, showing prolonged anterior (ventral) portion of basal lobe. Tue Mecoprera oF INpoNEsIA: GEeNus Neopanorpa 365 0.5mm SO ee es eee | G.w.B, Fics. 25-29. Neopanorpa muelleri (Weele), deatils of male and female genitalia. 25, base of right dististyle, male lectotype, ventral aspect. 26, base of left dististyle and dorsal parameres, male lectotype, dorsal aspect. 27, aedeagus, male from Ongop-ongop, East Java, right lateral aspect. 28, subgenital plate of female, ventral aspect. 29, genital plate of female, ventral aspect. band, marginal spot, basal band and basal spot. The apical band varies from a re- duced, incomplete type to large with con- spicuous proximal prominence or enclos- ing a small posterior pale spot. The ptero- stigmal band may be either complete or incomplete, broad or narrow, forked pos- teriorly or not. The distal branch may be complete or isolated as a spot. The basal band may be entire, reduced to one or two spots, or absent. Marginal and basal spots are either present or absent. Color varia- tion in N. muelleri is not affected by alti- tude (z.e., the dark form both in body and wings is found both in the mountains and in the lowlands). Hypovalves and disti- 366 Tue UNIversiry of KANsAs ScIENCE BULLETIN styles may be darkened or not. The basal lobe of the dististyle has a medium-sized concavity (or cup) on all Javanese forms of muelleri, but bears a larger cup in all Sumatran specimens. The number of smaller, curved spines on the anterior dor- sum of the basal lobe varies from as few as 8 on most Javanese forms to 25 on most Sumatran forms (Fig. 22). Weele (1909) and some other authors thought certain of these variants were dif- ferent species, Neopanorpa muellert and N. angustipennis. We found no speci- mens from Indonesia having the same structure and coloration as the holotype of angustipennis, supporting the view (Byers, 1965) that the type of angustipennis is from Tenasserim (peninsular Burma), not Java. Weele (1909) also described ungara- nensis as a subspecies of muellert, based on the smaller size of the former and its more reduced wing pattern, but Karny (1923) dissented. As Karny did, we find that variation in body size and color and wing markings occurs in a distribution that can- not sensibly be geographically subdivided, and we judge the nominal subspecies wn- garanensis to be a variant of muelleri. Once, we thought these might be a group of very similar species (and initially sorted these specimens into four “species”), but we prefer the more conservative view that these variant forms represent instead one widespread, variable species. Neopanorpa hyalinata Esben-Petersen, 1913 Neopanorpa hyalinata Esben-Petersen, 1913: 227-228. Campodotecnum falcatum Navas, 1914. Esben-Petersen based his description of hyalinata on the male holotype, female allotype, and one male, one female para- types. Head: Dorsum glossy black; frons below antennal sockets dark brown; ros- trum reddish brown anteriorly, yellowish brown laterally and brown apically; mouthparts brown with tips of palps dark brown. Antennal scape brown, pedicel dark brown, basal flagellomeres dull brown, others dull dark brown, flagellum with 44 segments (extreme apex broken in holotype). Thorax: Pronotum brownish black to black (holotype), bearing 3-4 bristles at each side along anterior margin. Mesono- tum and metanotum black. Pleura and coxae pale yellowish brown to sordid brown (holotype). Femora sordid yellow- ish brown (holotype) to dull dark brown; tibiae brown; tarsi dark brown. Wings: (Figs. 30-32): Almost color- less, slightly iridescent, with spots dark smoky brown. Apical band greatly re- duced, from almost absent to slightly tinged with light brown on distal margin of wing. Pterostigmal band usually ab- sent, pterostigma conspicuous, long and very dark. Behind inner end of pterostig- ma a small, dark brown spot connected to pterostigma in some specimens. Marginal spot, basal band and basal spot absent. Abdomen of male: Terga 15 brownish black, corresponding sterna yel- lowish brown to dark brown. Segment 6 black dorsally, blackish brown ventrally. Segments 7-9 brownish black. Posterior process of tergum 3 narrowly triangular (Fig. 36). Hypovalves of sternum 9 wid- ened and overlapping near mid-length, bearing a few long hairs, darkest near base, paler in apical half, narrowed to pointed apex. Tergum 9 pale and emarginate api- cally. Dististyles slender, dark brown ba- sally, reddish brown distally, outer margins slightly concave near mid-length, abruptly curved near tips. Basal lobes of dististyles concave ventrally, bearing long pale hairs on distal margin and short pale hairs on anterior margin; each with 3-4 long, api- cally curved, stout, black spines on meso- dorsal surface (4 in male syntype), about Tue Mecoprera oF INponesiA: Genus Neopanorpa 367 Fics. 30-32. Neopanorpa hyalinata Esben-Petersen, wings, showing variation in wing pattern. 30, typical pattern; note long, dark pterostigma; male, from East Java. 31, male, from East Java. 32, female, from East Java; this pattern also occurs in males. 6 smaller, apically curved spines on antero- dorsal surface (Fig. 40). Ventral valves of aedeagus broad, extending a little be- yond dorsal valves (Figs. 33, 42); lateral processes broad; ventral parameres absent, possibly represented by sclerotized bands in membrane below lateral processes (Figs. 33, 42); dorsal parameres conspicuous, blade-like, apices directed dorsomesad (Fig. 42). Abdomen of female: Terga 1-6 black, terga 7-10 blackish brown; sterna 368 Tue UNIversITy oF KANsAs ScIENCE BULLETIN Cw.B. 44 Fics. 40-44. Neopanorpa hyalinata Esben-Petersen, details of male and female genitalia. 40, base of right dististyle and dorsal parameres, dorsal aspect; male from Malang, Java. 41, same as 40 except ventral aspect without parameres. 42, aedeagus, right lateral aspect. 43, genital plate, ventral aspect; female from East Java. 44, same as 43, right lateral aspect. male, Nongkodjadjar, Java, Jan. 1911, Ja- cobson, in Rijksmuseum van Natuurlijke Historie, Leiden; 1.6, 1” 2. (eype. speci- mens of Campodotecnum falcatum Na- vas), Java, 1891, Fruhstorfer, in Naturhis- torisches Museum, Wien. Additional speci- mens: 3! 6 Gs 4-2 Sy wbengeer Vics, Nongkodjadjar, East Java, 1200 m, 2 May 1938, Walsh; 2 6 6, 1 2, Djyunggo, Ma- fangs Java, 22 Dee: 193i Betrem; di, 1 2, Ongop-ongop, Idjen, Java, 1850 m, May 1924, Dammerman; 2 2 2, Sumber- brantas, Mt. Ardjuno, East Java, 1500 m, Jan. 1936, Overbeck; 2 2 2, Bodjongka- long, Djampangs, West Java, 300 m, June 1938, Walsh. We place this species in the muelleri group because of the large spines on the basal lobes of the dististyles and other sig- nificant details of the male and female Tue MecopTera oF INponeEsIA: GeNus Neopanorpa 369 genitalia. It is easily recognized by the hyaline, slightly marked wings, for which the species is named, by its conspicuous pterostigma, and by the somewhat more blackish body color than that seen in other species. We find that the small spot below the inner end of the pterostigma is also present in some females (absent accord- ing to Esben-Petersen, 1913). Neopanorpa fuscicauda, new species Description based on 2 males, 5 females, pinned. Head: Dorsum glossy black; frons below antennal sockets dark brown; ros- trum brown anteriorly, yellowish brown laterally; mouthparts brown. Antennal scape and pedicel brown, flagellum dull dark brown with 41 segments (both an- tennae missing from holotype). Thorax: Pronotum black, bearing 5 bristles at each side along anterior mar- gin. Mesonotum and metanotum black. Pleura and coxae yellowish brown (holo- type) to reddish brown or blackish brown (2 females). Femora yellowish brown; tibiae and tarsi brown (holotype) to dark brown. Wings (Fig. 45): Almost colorless, slightly iridescent, with markings dark smoky brown. Apical band incomplete posteriorly, with small proximal promi- nence in one female. Pterostigmal band complete but not forked posteriorly, nar- rowly joined to apical band along anterior margin, narrow in males, broader in fe- males. Marginal spot, basal band and basal spot usually absent (marginal spot present and basal band represented by a spot between end of Cur and Cuz in one female). Abdomen of male: Terga 15 brownish black; sterna 25 sordid yellow- ish brown. Segment 6 brownish black. Segments 7-8 brown basally, dark brown distally. Segment 9 blackish brown (holo- type) to dark brown. Posterior process of tergum 3 narrowly triangular (as in muel- leri), extending about half-way across ter- gum 4. Hypovalves of sternum 9 blackish brown (holotype) to dark brown at base, remainder grayish brown, with inner mar- gins separated or slightly overlapping near mid-length, outer margins infolded dor- sally, the folded edges subparallel in ven- tral aspect (Fig. 48). Tergum 9 slightly emarginate apically (Fig. 49). Basistyles blackish brown (holotype) to dark brown; dististyles blackish brown (holotype) to dark brown basally, brown apically. Basal Fic. 45. Neopanorpa fuscicauda, new species, wings of female paratype. 370 Tue University oF KANsAs ScrENCE BULLETIN Fics. 33-39. Neopanorpa hyalinata Esben-Petersen, structural details. 33, genital bulb, ventral aspect, ninth sternum removed to show aedeagus; male from Malang, Java. 34, same as 33, left lateral aspect, ninth sternum in place. 35, ninth abdominal sternum and hypovalves, ventral aspect, shaded to indicate color pattern; male from Malang, Java. 36, reconstruction of abdominal terga 3 and 4, showing notal organ, dorsal aspect; male holotype. 37, same as 36, left lateral aspect. 38, ninth abdominal tergum, dorsal aspect; male from Malang. 39, subgenital plate, ventral aspect; female from East Java. 15 light brown (allotype) to sordid brown, sterna 6-7 brown to dark brown. Sub- genital plate of sternum 8 dark brown, slightly notched and whitish apically (Fig. 39), bearing long setae at each side of notch and on lateral margins, membra- nous basally. Axial portion of genital plate elongate, arms broad and twisted basally (Fig. 43). Body length: Male, about 13-15 mm; female, about 10-12 mm. Length of fore wing: Male, 12.8-13.5 mm (holotype 13.2 mm); female, 12.1-12.9 mm. Holotype: Male, and allotype, fe- Tue MecopTerA oF INDoNEsIA: GEeNus Neopanorpa S71 Fics. 46-53. Neopanorpa fuscicauda, new species, details of male and female paratypes. 46, genital bulb, ven- tral aspect, with sternum 9 removed to show aedeagus; ds—dististyle. 47, base of left dististyle and dorsal parameres, dorsal aspect. 48, ninth abdominal sternum and hypovalves (hv), ventral aspect, partly shaded to indicate color pattern. 49, ninth abdominal tergum, dorsal aspect. 50, same as 49, left lateral aspect. Sle genital plate of female, ventral aspect. 52, same as 51, right lateral aspect. 53, subgenital plate, ventral aspect, partly shaded to indicate color pattern. Scale a—Fics. 46, 48-50; scale b—Fics. Ave Sil-53: 372 Tue UNIVERSITY OF KANsAs SCIENCE BULLETIN lobes of dististyles subrectangular, with long hairs on posterior surface, 2-3 stout, black, apically curved spines projecting from mesodorsal surface, and about 8 smaller, apically curved spines on antero- dorsal surface; some pale hairs arising on anterior margin (Fig. 47). Ventral valves of aedeagus long and slender, projecting beyond dorsal valves in dorsal aspect (Fig. 47); lateral processes deeply concave; ven- tral parameres absent; dorsal parameres well developed, broadly triangular, mesally concave blades, projecting somewhat dor- sad and connecting ventrally to sides of ventral valves (Fig. 46). (Note that entire aedeagus as shown in Fig. 46 may be atypi- cally displaced upward, or caudad). Abdomen of female: ‘Terga 1-6 black; sterna 25 yellowish brown to dark brown (allotype), sternum 6 dark brown. Terga 7-9 brownish black, corresponding sterna dark brown. Cerci dark brown. Subgenital plate of sternum 8 short and broad, mostly dark brown, notched and with large whitish area apically, bearing several setae at each side of notch, other- wise sparsely hairy; basal portion slender, pale, membranous. Axial portion of geni- tal plate very short, conical; arms slender, sclerotized on outer basal margins, trans- lucent on inner margins (Fig. 51); base of each arm expanded to form a subquad- rate plate, the two plates separated by a median cleft above apex of axial portion and each darkly sclerotized adjacent to axial portion (Fig. 51). Body Length: Male, about 12-15 mm (holotype 15 mm); female, about 9-15 mm (allotype 11 mm). Length of fore wing: Male, 12.213 mm (holotype 13 mm); female, 12.3-13.3 mm (allotype 13 mm). Holotype: Male, Tengger Mts., Nongkodjadjar, East Java, 1200 m, 5 May 1938) Walsh: “Allotype 1? os ande4 9 paratypes, same data as holotype; 1 2°, same data as holotype, except Jan. 1911, Jacobson. Holotype, allotype and 3 2 paratypes in the Rijksmuseum van Na- tuurlijke Historie, Leiden; 1 ¢3)2 4 paratypes in the Snow Entomological Mu- seum, The University of Kansas. We place this species in the muelleri group because of its wing pattern and the spiniferous basal lobes of the male disti- styles. It can be easily recognized by its dark brown to blackish brown genital bulb, from which the species is named (also found in some muellert but not as uniformly as fuscicauda) and the charac- teristic dorsal parameres of the aedeagus (extending ventrally alongside the ventral valves). Females can be identified by the unique structure of the genital plate (base of each arm expanded to form a subquad- rate plate). Neopanorpa umbonata, new species Description based on 15 males, 21 fe- males, pinned. Head: Dorsum glossy brownish black (holotype) to black; frons below antennal sockets dark brown; rostrum brown with yellowish brown median and lateral stripes; mouthparts brown except tips of palps dark brown. Antennal scape yellowish brown, pedicel brown, flagellum mostly dull dark brown, brown at base, with 42 flagellomeres. Thorax: Pronotum brownish black, with 3-4 bristles at each side along anterior margin. Mesonotum and metanotum brownish black. Pleura mostly yellowish brown (holotype) to reddish brown, dark brown on episternum near wing base. Coxae yellowish brown (holotype) to red- dish brown. Femora yellowish brown; tibiae and tarsi dark brown. Wings (Figs. 54, 55): Nearly color- less, slightly iridescent, with bands and spots smoky dark brown. Apical band interrupted posteriorly or including vari- ously shaped clear areas. Pterostigmal band with complete proximal branch, dis- Tue MeEcopTera oF INponEsIA: Genus Neopanorpa 373 Fics. 54-55. Neopanorpa umbonata, new species, wings of male paratypes from two localities in southern Sumatra, showing variation in pigmentation. 54, male from Tandjong Sakti. 55, male from Wai Lima, Lampongs. tal branch reduced to spot on posterior wing margin. Marginal spot small (holo- type) to conspicuous. Basal band slender, usually interrupted in cell M (holotype), complete in some paratypes. Basal spot absent. Abdomen of male: ‘erga 1-3 brownish black, terga 45 dark brown; sterna 25 usually yellowish brown. Seg- ment 6 dark brown. Segments 7-9 brown- ish yellow. Posterior process of tergum 3 (Fig. 61) narrow with subparallel sides, extending about half-way across tergum 4. Hypovalves of sternum 9 dark brown at base to brown at tip, inner margins over- lapped near mid-length, tips usually con- cealed by basal lobes of dististyles. ‘Ter- gum 9 (Fig. 58) narrowed just before truncate apex, elongate, nearly concealing cerci. Basistyles slender, glabrous and darkly sclerotized mesally. Dististyles dark brown, their outer margins generally straight in basal two-thirds, then smoothly curved toward tips (Fig. 56). Basal lobes of dististyles divided into thin, expanded posteroventral and anterodorsal portions together forming a large cup-like recess (Fig. 56); posteroventral portion directed ventrocaudad, usually covering tips of hy- povalves, covered with long hairs; more 374 Tue UNIVERSITY OF KANsAs ScIENCE BULLETIN Fics. 56-65. Neopanorpa umbonata, new species, structural details of holotype and male and female paratypes. 56, genital bulb of male, ventral aspect, ninth sternum removed to show aedeagus. 57, ninth abdominal sternum and hypovalves, ventral aspect, partly shaded. 58, ninth abdominal tergum, dorsal aspect, most hairs omitted. 59, genital bulb, left lateral aspect; holotype. 60, abdominal segments 3 and 4, left lateral aspect, showing notal organ; holotype. 61, same as 60, dorsal aspect. 62, basal lobe of left dististyle, dorsal aspect. 63, sub- genital plate of female, ventral aspect. 64, genital plate of female, ventral aspect. 65, same as 64, right lateral aspect. Scale a—Fics. 56-61, 63; scale b—Fics. 62, 64-65. Wings (Fig. 66): Tinged with yel- type. Pterostigmal band incomplete, re- low and slightly iridescent, markings duced to a small spot (holotype), or only smoky brown. Apical band incomplete, stigma pigmented (two paratypes), tri- bearing variously shaped proximal promi- angular and including stigma in females. nence in some paratypes, reduced almost to Marginal spot absent in all males, present a light, small spot in one (teneral) para- in females. Basal band absent. Tue Mecoprera oF INpoNnEsIA: GENus Neopanorpa 375 Fic. 66. Neopanorpa angustiapicula, new species, right wings of female paratype. Abdomen of male: Terga 1-3 black- ish brown, terga 45 dark brown (holo- type) or dark brown in the middle, grad- ing into light to yellowish brown at mar- gins (paratypes); sterna 25 yellowish brown to light brown (holotype). Seg- ment 6 brown dorsally, reddish brown ventrally; dark brown in one paratype. Segments 7-9 reddish brown, 7-8 much longer than in any other Javanese Neo- panorpa (Fig. 71). Posterior process of tergum 3 elongate, stout, fringed with hairs at base, extending across about two- thirds length of tergum 4 (Figs. 69, 70). Elevated process on tergum 4 bifurcate. Smooth, shiny area on tergum 4 beneath posterior process of 3. Male genitalia (based mainly on dissected paratype): Hypovalves of sternum 9 apically and me- sally darkened, long, expanded and over- lapping at mid-length, narrowed and bear- ing sparse long hairs on apical one-third (Fig. 67). Ninth tergum (Fig. 72) broad, reddish brown (holotype) to yellowish brown, paler and almost squarely trun- cate apically. Cerci pale basally, dark brown apically. Genital bulb short and stout. Dististyles short, their outer mar- gins generally straight or slightly concave but abruptly curved at tips (Fig. 67). Basal lobes of dististyles large, subrec- tangular in posterior mesal aspect, flat- tened, darkened, bearing many long yel- lowish hairs (Figs. 68, 73). Mesal margins of dististyles above basal lobes also some- what pointed, bearing a long setae dorsally (Fig. 73). Aedeagus complex, with con- spicuously projecting ventral parameres (Figs. 71, 74-76). Ventral valves small, projecting caudad only about as far as dorsal valves; dorsal valves concealed be- tween dorsal parameres. Ventral para- meres thick, darkly sclerotized, strongly curved ventrad and caudad, abruptly nar- rowed to mucronate tip (Figs. 74, 76). Dorsal parameres thickened, _ bluntly rounded, each continued dorsocephalad as a thin, slightly hooked blade (Figs. 74, 75). Lateral processes complexly divided into two parts separated by a deep, sclero- tized pocket (Fig. 74). Abdomen of female: Terga 1-6 dark brown (allotype) to blackish brown, corresponding sterna dull brown to light brown (allotype). Segments 7-8 brown basally, reddish brown apically (allotype) or entirely reddish brown. Segments 9-10 blackish brown (allotype) to reddish brown basally, dark brown apically. Cerci black. Subgenital plate yellowish, broadly 376 Tue UNIVERSITY OF KANsAs SCIENCE BULLETIN dorsal portion of lobe with 2-3 long, stout, black spines projecting from its postero- dorsal surface, about 12 smaller black spines on mesodorsal surface and about 9 small apically curved spines on anterodor- sal surface (Fig. 62). Ventral valves of aedeagus divergent basally, extending slightly beyond dorsal valves and lateral processes; ventral parameres absent; dor- sal parameres conspicuous, tongue-shaped blades, close together basally but with apices divergent (Fig. 56). Abdomen of female: Terga 1-6 brownish black, corresponding sterna dull yellowish brown (allotype) to dull brown. Terga 7-10 dark brown; sternum 7 brown; cerci black. Subgenital plate of sternum 8 (Fig. 63) brown, whitish and notched apically, bearing long setae at each side of notch, weakly sclerotized anterolaterally. Axial portion of genital plate, greatly en- larged anteriorly (Fig. 64), bearing con- spicuously protruding, ventral processes (Fig. 65). Arms broad, slightly twisted near base. Body length: Male, about 13-15 mm (holotype 14 mm); female, about 9-11 mm (allotype 11 mm). Length of fore wing: Male, 11.9-14 mm (holotype 11.9 mm); female, 12.2-13.3 mm (allotype 12.6 mm). Holotype: Male, Giesting, Mt. Tanggamus, southwestern Lampongs, southern Sumatra, 500-800 m, 24 and 27 Dec. 1934, Lieftinck. Allotype, 10 ¢ 4, 1D) 292) vsame datas as holotype: 1-6; -3 2 2, same data as holotype, except 500- 1000 m, 29 June and Sept. 1933-34, Toxo- peuss i> é,, 1 2, Lampongs», War Lima, southern Sumatra, 11 Dec. 1921, Karny; I, 63, Muara, Sako,..Sumatea, Oct. 1915, Jacobson; 1 2, Fort de Kock (Bukittingi), Sumatra, 920 m, 1924, Jacobson; 1 ¢, Tan- duong Sakti, Benkulen, southern Suma- tra, 650 m, June 1935, Walsh. Holotype, allotype and most paratypes in the Rijks- museum van Naturrlijke Historie, Leiden; 2 6,1 2 paratypes in the Snow Entomo- logical Museum, The University of Kansas. We place this species in the muelleri group because of its wing pattern, the large spines on the dorsal surface of the basal lobes of the dististyles and the structure of the aedeagus in males. It can be easily recognized by the prominent, rounded components of the basal lobe of the disti- style, from which the species is named. Especially, the entirely exposed, ventro- caudally directed part is conspicuous and characteristic. For the female, the peculiar structure of the genital plate (protruding processes on anteroventral axial portion) is diagnostic. NEOPANORPA ANGUSTIAPICULA Group Neopanorpa angustiapicula, new species Description based on 4 males, 7 females, pinned. Head: Dorsum of head _ glossy brownish black except brown bordering eyes; frons below antennal sockets dark brown to brown (two paratypes); rostrum in males reddish brown anteriorly, bright yellowish brown laterally, dark brown in most females; mouthparts dark brown. Antennal scape light brown, brown or reddish brown (holotype), pedicel dark brown, basal flagellomeres dull dark brown, others dull blackish brown; flagel- lum with 46 segments (paratypes; both antennae broken in holotype). Thorax: Pronotum blackish brown to black (holotype), bearing many short, dark hairs but usually no bristles along anterior margin (three short bristles at each side in some females). Mesonotal and metanotal scuta blackish brown to black (holotype) with elongate, diffuse, yellowish brown spots at each side, above bases of wings. Mesonotal and metanotal scutella dark brown. Pleura and coxae yel- lowish brown to reddish brown (holo- type). Femora yellowish brown; tibiae brown; tarsi dark brown. Tue Mecoprera oF INponEsIA: GENus Neopanorpa 377 \ \ Fi G.w.B. Fics. 67-71. Neopanorpa angustiapicula, new species, structural details. 67, genital bulb, ventral aspect; male holotype. 68, right dististyle, posteromesal (dorsomesal) aspect, to show shape of basal lobe; paratype. 69, abdominal terga 3 and 4, dorsal aspect, showing notal organ; holotype. 70, same as 69, left lateral aspect. 71, terminal abdominal segments, left lateral aspect, holotype. 72, ninth abdominal tergum, dorsal aspect; paratype; most hairs omitted. 378 Tue UNIversiTy oF KansAs ScrENCE BULLETIN oval, notched apically between two round- ed lobes bearing long setae (Fig. 79), lightly sclerotized and translucent latero- basally. Genital plate small (compared to that of other species), axial portion ovoid, transparent basally (Fig. 77); arms short, not twisted mesal edges translucent, lateral edges darkened. Body length: Male, 16.7-18.7 mm (holotype 18.7 mm); female, 10.9-16.6 mm (allotype 16.6 mm). Length of fore wing: Male, 12.9-13.1 mm (holotype 13 mm); female, 12.2-13.0 mm (allotype 13 mm). Holotype: Male, Sumberbrantas, Mt. Ardjuno-Mt. Andjasmoro, East Java, about 1900 m, Jan. 1936, M. E. Walsh. Al- lotype, female, and 2 2 paratypes, same data as for holotype; 3 4 4,3 2 9, same data as for holotype except altitude about 2000-2300 m. This species is known from East Java only. Holotype, allotype and most paratypes in the Rijksmuseum van Natuurlijke Historie, Leiden; 1 3,1 2 paratypes in the Snow Entomological Mu- seum, The University of Kansas. Neopanorpa angustiapicula appears closely related to N. diloba, forming with it the angustiapicula group on the basis of the aedeagal structure. Both species have well developed ventral parameres (absent in most other Indonesian species, except spicata and borneensis) ; both have shorter, thicker dorsal parameres than other spe- cies; the lateral processes are large and complicated; and both species have large, flattened basal lobes on the dististyles and lack the long, thick, black spines present in the muelleri group. Neopanorpa angustiapicula can be eas- ily distinguished from diloba and other species by its reduced wing pattern, gen- erally reddish brown body color, and for males by the elongated process of abdom- inal tergum 3, the notched process of ter- gum 4, the long segments 7-8 (more than 15 times longer than in other species), the robust genital bulb and the narrowed tips of the hypovalves, from which the species is named. The female differs from other species in the ovoid axis of the geni- tal plate and in having no apical pale or hairless areas on the subgenital plate. Neopanorpa diloba, new species Description based on one callow male, pinned. Head: Dorsum glossy _ blackish brown; frons below antennal sockets dark brown; rostrum yellowish brown, change in color between frons and rostrum abrupt; mouthparts brown, tips of palps dark brown. Antennal scape brown, pedicel dark brown, basal flagellomere dull brown, others dull dark brown (both antennae broken). Thorax: Pronotum brownish black, bearing 45 bristles on each side along an- terior margin. Mesonotum and metono- tum brownish black, with paler areas near wing bases. Pleura and coxae dull brown, covered with short, whitish hair. Femora yellowish brown; tibiae dull brown; basi- tarsi dull brown, other tarsomeres dark brown. Wings (Fig. 80): Almost colorless, slightly iridescent, with faint smoky brown markings. Apical band small and incom- plete, with detached spots in cell Ist Rs and cell 2nd Rs. Unforked pterostigmal band slender, faint and incomplete. Mar- ginal spot absent. Basal band reduced to faint, large spot in cell R and smaller one in cell M. Abdomen of male: Terga 1-4 dark brown; tergum 5 dark brown basally, brown apically. Sterna 25 brown. Seg- ment 6 dark brown; segment 7 brown basally, pale yellowish brown apically; segment 8 pale yellowish brown; segment 9 yellowish brown except hypovalves dark brown. Posterior process of tergum 3 stout, extending over about two-thirds length of tergum 4. Hypovalves of ster- num 9 darkened, very large, hairless on Tue MecopTera oF INDONESIA: GeNus Neopanorpa 379 Fics. 73-79. Neopanorpa angustiapicula, new species, details of male and female genitalia. 73, left dististyle, ventral aspect; male paratype. 74, aedeagus, left lateral aspect; male paratype; vp—ventral paramere. 75, ae- deagus, posteroventral aspect, left half with ventral paramere removed; paratype. 76, left ventral paramere and edge of lateral process, ventral aspect; paratype. 77, genital plate of female, ventral aspect; paratype. 78, same as 77, left lateral aspect. 79, subgenital plate of female, ventral aspect; paratype; hairs partially omitted. 380 Tue UNIversIty oF KANsAs ScrENCE BULLETIN mesal margins (Fig. 81). Ninth tergum (Fig. 84) broad, short, yellowish brown, with slightly rounded apical margin; ven- tral processes directed anteriorly (Fig. 85). Cerci pale basally, dark brown apically. Genital bulb broad and short; basistyles yellowish brown, dististyles yellowish brown, short and stout, their outer margins generally straight, curved at tips. Basal lobes of dististyles large, darkened, bear- ing many long hairs, flattened, each sub- divided into two lobes, ventral one smaller, pointed, dorsal one larger, bluntly rounded (Figs. 83, 87). Above basal lobe on mesal surface of each dististyle a rounded ridge (Fig. 83) appearing as a blunt, angular projection in ventral aspect (Fig. 87). Aedeagus complicated, lightly sclerotized, yellowish brown, with more densely sclero- tized ventral parameres and lateral proc- esses. Ventral valves small, set unusually far dorsally on aedeagus (Figs. 87, 88); dorsal valves minute, concealed between dorsal parameres in lateral aspect and by ventral valves in ventral aspect. Ventral parameres somewhat spatulate, curved ventrad and caudad from ventral surface of aedeagus (Figs. 87, 88); dorsal para- meres short, thick, bluntly rounded at apices (Figs. 86, 88). Lateral processes di- vided, with thick inner lamellae below and before ventral valves (latter subquadrate in outline in Fig. 88) and outer lamellae each bearing a knob-like process, with a deep pocket at each side, a little below ventral valves, separating lamellae (Figs. 87, 88). Body length: Male, about 14.0 mm. Length of fore wing: 115 mm. Holotype: Male, southeastern Djam- bangan, Tengger Mts., East Java, 1800 m, July 1939, N. van Bemmel-Lieneman. We consider species to be closely related to angustiapicula on the basis of the struc- ture of genitalia, as discussed under an- gustiapicula. Both species are restricted to East Java. Neopanorpa diloba can be easily recognized by its faint, almost obliterated wing pattern, its long, wide hypovalves, and the characteristic flattened, divided basal lobes of the dististyles, from which the species is named. NEOPANORPA FRACTURA Group Neopanorpa fractura, new species Description based on 15 males, 18 fe- males, pinned. Head: Dorsum glossy black; frons below antennal sockets dark brown; ros- trum dark brown anteriorly, yellowish brown laterally; mouthparts brown except tips of palps dark brown. Antennal scape yellowish brown, pedicel brown, flagellum dull dark brown on basal segment, other- wise dull blackish brown, with 41 flagel- iomeres. Thorax: Pronotum black, bearing about 4-5 bristles on each side at anterior margin. Mesonotum and _ metanotum blackish brown to black. Pleura and coxae yellowish brown to dingy dark brown (reddish brown in holotype). Femora and tibiae dingy yellowish brown; basitarsi dull brown, other tarsomeres dark brown. Ss - ~ a ee Fic. 80. Neopanorpa diloba, new species, right fore wing of male holotype. Tue MecoprTera oF INDoNEsIA: GENus Neopanorpa 381 Fics. 81-88. Neopancrpa diloba, new species, structural details of male holotype. 81, genital bulb, ventral aspect. 82, terminal abdominal segments, left lateral aspect. 83, right dististyle, posteromesal (dorsomesal) aspect, to show shape of basal lobe. 84, abdominal tergum 9, dorsal aspect. 85, same as 84, left lateral aspect. 86, basal lobe of left dististyle and aedeagus, dorsal aspect. 87, left dististyle, apex of basistyle, and aedeagus, ventral (posterior) aspect. 88, aedeagus, left lateral aspect. Scale a—Fics. 81-85; scale b—Fics. 86-88. 382 Tue UNIVERSITY OF KANsAs ScIENCE BULLETIN Fics. 89-90. Neopanorpa fractura, new species, right wings of two male paratypes from north central Sumatra, showing variation in pigmentation. Wings (Figs. 89, 90): Clear, slightly iridescent, with markings smoky dark brown. In fore wing, apical band frag- mented, reduced to brown tinge in distal portion of outermost row of radial cells and transverse spot, variable in extent, at level of distal fork of Re. Pterostigmal band variable from complete and forked posteriorly (1 female), to entire but strong- ly constricted at mid-length and lacking distal branch (holotype and 10 paratypes), to interrupted at or near Mi with proximal branch represented by one or two spots near wing margin (in most paratypes). Basal band usually represented by one or two spots, entire in holotype and 6 para- types. Marginal spot conspicuous, basal spot absent. Hind wing: incomplete ptero- stigmal band present, weak apical band and marginal spot present in holotype but Tue Mecoprera oF INponEstA: GENus Neopanorpa 383 96 = { wit f Uy 14 = 7 ESS Gw.B, 95 Fics. 91-97. Neopanorpa fractura, new species, structural details of male. 91, genital bulb, ventral aspect; holotype. 92, genital bulb, ventral aspect, with ninth sternum removed to show aedeagus; paratype. 93, hypo- valve of ninth sternum, left lateral aspect. 94, ninth abdominal tergum, left lateral aspect. 95, same as 94, dorsal aspect. 96, base of left dististyle and aedeagus, dorsal aspect. 97, supposed rudimentary ventral para- meres, ventral aspect. Scale a—Fics. 91-95; scale b—Fics. 96-97. absent in most paratypes; basal spot pres- ent in some specimens. Abdomen of male: TYerga 15 usu- ally black; sterna 25 yellowish brown. Seg- ment 6 blackish brown. Segments 7-9 yellowish brown. Posterior process of ter- gum 3 a narrow triangle, widened basally, extending about half-way across tergum 4. Hypovalves glossy dark brown at base to dull brown at tip, bent mesad near mid- length, with inner margins broadly over- lapping, outer margins strongly infolded dorsally, nearly touching inner margins; a stout, dorsal, lightly sclerotized projection near mid-length in lateral aspect (Fig. 93). Tergum 9 (Figs. 94, 95) broadly rounded to nearly truncate apically; ventral process with incurved edge. Basistyles yellowish brown; dististyles dark brown, their outer margins nearly straight except abruptly 384 Tue Universiry oF KANsAs ScrENCE BULLETIN ie) 0.5mm [= ee Fics. 98-100. Neopanorpa fractura, new species, details of genitalia of female allotype. 98, subgenital plate, ventral aspect; hairs partially omitted. 99, genital plate, ventral aspect. 100, genital plate, right lateral aspect. curved at tips. Inner basal lobes of disti- styles directed anteriorly (downward), bearing long hairs anteroventrally, cleft mesally so that each basal lobe seems di- vided into two parts (Fig. 96), with a long seta projecting from near cleft, a few (4-8) smaller setae arising from mesodorsal sur- face of each basal lobe, and tuft of pale hairs beneath basal lobe (Figs. 92, 96). Ventral valves of aedeagus extend a little beyond dorsal valves; lateral processes sim- ple, concave ventromesally; ventral para- meres absent or possibly represented by moderately sclerotized bands on ventral surface of aedeagus, bearing small, mesally directed points (Figs. 92, 97). Dorsal para- meres conspicuous, thin, translucent, broad and outwardly curved at base, mucronate at apex, each with a small flattened dorsal appendage projecting dorsad with out- wardly deflected tip (Fig. 96). Abdomen of female: Terga 1-6 black; sterna 25 pale yellowish brown, sternum 6 dark brown. Segments 7-10 brown. Cerci black. Subgenital plate (Fig. 98) yellowish brown, pale at apex and base, with indistinct, dark-bordered pale spot at margin near base, narrowed gradually from near mid-length, notched apically, bearing many long setae along lateral margins and each side of notch. Axial portion of genital plate ovoid; arms broad, twisted basally (Fig. 99). Body length: Male, about 11-13 mm (holotype 13 mm); female, about 8-11 mm (allotype 9 mm). Length of fore wing: Male, 12-13 mm (holotype 13.2 mm); fe- male, 12.0-13.2 mm (allotype 12 mm). Holotype: Male, Brastagi, northern central Sumatra, 1400 m, Nov. 1950, M. A. Lieftinck. Allotype and 14 6,17 2 para- types, same data as for holotype. Holotype, allotype and most paratypes in the Rijks- museum van Natuurlijke Historie, Leiden; 2 6,2 2 paratypes in the Snow Entomo- logical Museum, The University of Kan- sas. Tue Mecoptera oF INponEs1A: GEeNus Neopanorpa 385 This species with swmatrana and crinita form the fractura group because of the structure of the hypovalves (lightly sclero- tized projection on dorsal surface), cleft basal lobes of the dististyles, and structure of the aedeagus of males and genital plates of females. It can easily be recognized by its incomplete apical band, strongly con- stricted or interrupted pterostigmal band, and the sclerotized projections of the hypo- valves. The species is named from the divided appearance of the hypovalves (Fig. 93). It differs from crinita in not having long dorsal hairs on the hypo- valves, in the shape of the hypovalves (Fig. 93) and the more rounded apex of the 9th tergum. It can be distinguished from sumatrana by the more complete wing markings and less prominent dorsal projections on the hypovalves in that spe- cies. Females can be distinguished from those of sumatrana by the wing pattern. Further differences are found in the sub- genital plate and genital plate, which in sumatrana has distinct lateral lobes (com- pare Figs. 99, 124). Neopanorpa crinita, new species Description based on 6 males, 1 female, pinned. Head: Dorsum shining black; frons below antennal sockets dark brown; ros- trum dark brown anteriorly, yellowish brown laterally; mouthparts brown except tips of palps dark brown. Antennal scape yellowish brown, pedicel brown, flagellum dull dark brown at base, grading into dull blackish brown, with 40-41 flagellomeres (right flagellum broken in holotype). Thorax: Pronotum black, bearing 35 bristles on each side along anterior margin. Mesonotum and metanotum blackish brown to black. Pleura dark brown to blackish brown near wing bases, yellowish brown to reddish brown near coxae, or discolored to completely dark brown (holotype and 1 male paratype) to black (allotype). Coxae yellowish brown (holotype) to dull brown, blackish brown in allotype. Femora yellowish brown (hol- otype and most paratypes) to dull dark brown (allotype); tibiae and tarsi brown. Wings (Fig. 101): Almost colorless, Fic. 101. Neopanorpa crinita, new species, right wings of male paratype. 386 Tue UNtversiry oF KANsAs ScIENCE BULLETIN slightly iridescent, with dark smoky brown bands and spots. Apical band very weak, almost obliterated (holotype and 1 male paratype), or fragmented, reduced to brown tinge in distal portion of outermost row of radial cells and transverse spot, variable in extent, at level of distal fork of Re and terminating before vein Rs. Pterostigmal band variable, entire but lacking distal branch in holotype and 3 paratypes; distal branch reduced to a dif- fuse spot in 2 paratypes. Marginal spot present. Basal band usually entire, broken into two spots in 1 paratype. Basal spot absent or only faintly developed. Abdomen of male: Terga 15 black- ish brown to black (holotype); sterna 25 yellowish brown to sordid dark brown (holotype). Segment 6 blackish brown. Segments 7-9 brown, dark brown apically on segment 7. Posterior process of tergum 3 narrowly triangular, extending about half-way across tergum 4. Hypovalves dark brown at base to brown at tip, notched laterally and appearing bent near mid- length, with slight process on mesal mar- gin near base (Fig. 103); inner margins overlapped mesally, outer margins strong- ly infolded dorsally, bearing long, crinkled hairs (except possibly broken off in 1 para- type), and a stout, dorsal, moderately sclerotized projection near mid-length in lateral aspect (Fig. 104). Tergum 9 slight- ly emarginate (Fig. 105) to nearly truncate apically, its ventral processes subrectangu- lar. Basistyles yellowish brown to brown mesally; dististyles dark brown. Basal lobes of dististyles (Figs. 102, 108) di- rected anteriorly, bearing long hairs on ventral surface and tuft of shorter hairs anteroventrally, shallowly cleft mesally so that each basal lobe seems divided into two parts; a long seta projecting from near cleft, about 4 smaller spines arising from mesodorsal surface of each basal lobe. Ventral valves of aedeagus extending ven- trad, somewhat hooked (Fig. 107); lateral processes simple, concave; ventral para- meres absent; dorsal parameres thin, sub- quadrate blades, closely appressed in dor- sal aspect (Figs. 107, 108). Abdomen of female: Terga 16 black; sterna 2-6 sordid brown. Terga 7-9 blackish brown, corresponding sterna dark brown. Cerci blackish brown. Subgenital plate yellowish brown, pale and notched apically, with 5-6 setae on lobe at each side of notch, pale apical area otherwise hair- less; low folds from anterobasal corners converging medially, area between folds pale, without hairs (Fig. 109). Axial por- tion of genital plate broadly subconical with darkly sclerotized, rounded lateral lobe at each side above base of distal arm; arms moderately sclerotized, twisted basal- ly (Fig. 110). Body length: Male, about 12-14 mm (holotype 14 mm); female (allotype) about 9 mm. Length of fore wing: Male, 12-14 mm (holotype 13 mm); female (al- lotype), 12.3 mm. Holotype: Male, Suban Ajam, Su- matra, July 1916, Jacobson. Allotype, same data as for holotype. One 4 paratype, Giesting, Mt. Tanggamus, southwestern Lampongs, southern Sumatra, 800 m, 27 Dec. 1934, Lieftinck; 1 ¢, Wai Tebu, Mt. Tanggamus, southern Sumatra, 500 m, 24 June 1934, Toxopeus; 1 ¢, Mt. Tangga- mus, 1000 m, 12 July 1934, Toxopeus; 1 ¢, Air Njuru, Dampu, Sumatra, 1400 m, Aug. 1916, Jacobson; 1 ¢, Muara Sako, Sumatra, Oct. 1915, Jacobson. Holotype, allotype and 3 @ paratypes in the Rijks- museum van Natuurlijke Historie, Leiden; 2 6 paratypes in the Snow Entomological Museum, The University of Kansas. N. crinita is very similar to N. fractura, but the two have quite different ranges. N. crinita may be distinguished from frac- tura and sumatrana by the long, crinkled hairs on the hypovalves, from which the species is named, and by the more abruptly bent shape of the hypovalves (cf. Figs. 91, Tue MecoprTerA oF INDONESIA: GENUS Neopanorpa 103 Imm 0.5mm 105 G w.B, 106 Fics. 102-108. Neopanorpa crinita, new species, structural details of male paratype. 102, genital bulb, ven- tral aspect, with ninth sternum removed to show aedeagus. ventral aspect, only partly shaded to show color pattern. 103, ninth abdominal sternum and_ hypovalves, lateral aspect; most hairs omitted. 104, ninth abdominal tergum and sternum, left 105, ninth abdominal tergum, dorsal aspect; most hairs omitted. 106, abdominal terga 3 and 4, dorsal aspect, showing notal organ. 107, aedeagus, right lateral aspect. 108, base of left dististype and dorsal parameres, dorsal aspect. Scale a—Fics. 102-106; scale b—Fics. 107-108. 387 388 Tue University oF KANsAs ScrENCE BULLETIN Fics. 109-111. Neopanorpa crinita, new species, details of genitalia of female allotype. 109, subgenital plate, ventral aspect; hairs partially omitted. 110, genital plate, ventral aspect. 111, genital plate, right lateral aspect. 103) and less conspicuous sclerotized pro- jection on the dorsal surface of each hypo- valve (Figs. 93, 104). It can be further differentiated from fractura by its longer ventral valves, the thin, subrectangular dor- sal parameres (more outwardly rounded in fractura) and the more complete ptero- stigmal band, especially the proximal branch. The thin, appressed dorsal para- meres and the simpler wing pattern sepa- rate this species from samatrana. In the female, the pale mesal zone and bordering folds of the subgenital plate and the dark- ened lateral lobes of the axial portion of the genital plate confirm the identification suggested by the wing patterns and to- gether differentiate crinita from fractura and sumatrana. Neopanorpa sumatrana, new species Description based on 1 male, 4 females, pinned. Head: Dorsum glossy black; frons below antennal sockets blackish brown; rostrum grayish brown basally, reddish brown apically, except yellowish brown along sides; mouthparts brown, dark brown at apex of palps. Antennal scape brown, pedicel dark brown, flagellum dull dark brown, with 41 segments (left flagel- lum broken on holotype). Thorax: Pronotum black, bearing 4 bristles on each side along anterior mar- gin. Mesonotum and metanotum black. Pleura mostly dull blackish brown, brown near coxae. Coxae and femora dull brown; tibiae and tarsi dull dark brown. Wings (Fig. 112): Membrane col- orless and slightly iridescent, bands and spots dark smoky brown. Apical band in- cluding small posterior pale spot in holo- type and most paratypes. Pterostigmal band complete, forked posteriorly, joining apical band narrowly along costal margin, distal branch smaller than proximal branch. Marginal spot conspicuous. Basal band complete. Basal spot absent in holo- type and most paratypes. Abdomen of male: Terga 15 black- ish brown, corresponding sterna yellowish brown. Segment 6 blackish brown, seg- ments 7-9 reddish brown except hypovalves Tue Mecoprera oF [nponestA: GENus Neopanorpa 389 112 Fic. 112. Neopanorpa sumatrana, new species, right wings of female paratype. of sternum 9 darker. Posterior process of tergum 3 triangular, extending about half- way across tergum 4 (Fig. 116). Hypo- valves of sternum 9 dark brown at base to brown at tip, inner margins overlapped mesally, outer margins strongly infolded dorsally; near mid-length on dorsal sur- face, a stout, sclerotized projection (Fig. 114). Tergum 9 (Fig. 118) brown, trun- cate apically. Basistyles short, slightly longer than dististyles; dististyles long, their outer margins slightly concave before mid-length (Fig. 113). Basal lobes of disti- styles (Fig. 123) directed downward (an- teromesad), bearing long ventral hairs, cleft (Fig. 119) so as to form two parts; a long seta projecting from near cleft; about 7-10 smaller apically curved spines projecting from mesodorsal surface of each basal lobe (Figs. 119, 121); tuft of pale hairs in notch between basal lobe and base of dististyle. Ventral valves of aedeagus extend well beyond small dorsal valves; lateral processes simply concave, ventral parameres absent or possibly represented by small, moderately sclerotized points on anteroventral surface of aedeagus; dorsal parameres large, outwardly convex blades, appressed basally but with dorsal margins divergent (Fig. 121), more ventral mar- gins curved outwardly around dorsal valves (Fig. 122). Abdomen of female: Terga 1-6 black; sterna 25 yellowish brown, sternum 6 dark brown. Terga 7-9 blackish brown, corresponding sterna brown. Segment 10 and cerci black. Subgenital plate yellow- ish brown with whitish, notched apex, bearing many long setae along lateral mar- gins and at each side of notch (Fig. 120). Axial portion of genital plate ovoid, with conspicuous, rounded lobe at each side; arms of distal portion broad, twisted ba- sally, with outer margins pale (Fig. 124). Body length: Male, about 10 mm (holotype); female, about 8-10 mm (allo- type 9mm). Length of fore wing: Male, 12.7 mm (holotype); female, 12.1-13.5 mm (allotype 13.5 mm). Holotype: Male, Fort de Kock (Bukittinggi), Sumatra, 920 m, 1924, Ja- cobson. Allotype, Tandjunggadang, near west coast of Sumatra, 1200 m, Feb. 1926, Jacobson. Paratypes: 1 2, same data as for holotype; 2 2 2, same data as for allotype, except 1000 m. Holotype, allo- type and 2 2 paratypes in the Rijksmu- seum van Natuurlijke Historie, Leiden; 1 2 paratype in the Snow Entomological Museum, The University of Kansas. 390 Tue UNIVERSITY OF KANsAs SCIENCE BULLETIN 116 118 f) 1mm Gw.B 117 115 Fics. 113-120. Neopanorpa sumatrana, new species, structural details of male holotype and female paratype. 113, genital bulb, ventral aspect, ninth sternum removed to show aedeagus. 114, ninth abdominal tergum and sternum, left lateral aspect. 115, ninth sternum and hypovalves, ventral aspect. 116, abdominal terga 3 and 4, dorsal aspect, showing notal organ. 117, same as 116, left lateral aspect. 118, ninth tergum, dorsal aspect. 119, right dististyle, posteroventral (slightly posteromesal) aspect, to show shape of basal lobe. 120, subgenital plate of female, ventral aspect. We place this species in the fractura group because of the structure of the hypo- valves, basal lobes of the dististyles, and aedeagus in the male and of the genital plate in the female. It may be distin- guished from fractura by its complete, extensive wing markings, both in male and female, less protruding dorsal projec- tion of the hypovalves, and the very char- acteristic, dorsal parameres, of which the dorsal part is much larger in sumatrana than that in fractura. The female is dif- ferentiated from that of fractura by the subgenital plate (narrowed more abruptly in sumatrana) and by the conspicuous lat- eral lobes of the genital plate. Tue MecoptTerA oF [NpoNEsIA: GENUs Neopanorpa 39] 121 125 GwW.B. Fics. 121-125. Neopanorpa sumatrana, new species, details of genitalia of male holotype and female paratype. 121, basal lobe of left dististyle and dorsal parameres, dorsal aspect. 122, aedeagus, right lateral aspect. 123, basal lobe of left dististyle, ventral aspect. 124, genital plate of female, ventral aspect. 125, same as 124, right lateral aspect. NEoPANORPA BORNEENSIS Group Neopanorpa borneensis Byers, 1966 Description mostly based upon original description (Byers, 1966), based on 2 males and 6 females, pinned. Head: Dorsum glossy black; frons below antennal sockets dark brown; genae light yellowish brown; rostrum dark brown anteriorly, yellowish brown later- ally, especially near base, except for thin line of brown along edge of eye; margin of labrum black. Antennal scape dark yellowish brown, pedicel brown; flagellum dark brown basally, grading into black, with 41 flagellomeres. Thorax: Pronotum black, with 3 bristles at each side on anterior margin. Mesonotum and metanotum almost whol- ly black, each with small, poorly defined, yellowish brown areas on _ posterolateral shoulders of scutum near wing bases. Pleura and coxae dark yellowish brown. Femora yellowish brown; tibiae darker yellowish brown; tarsi brown, darkest apically. Wings (Fig. 126): Almost colorless, marked with bands and spots of dark 392 Tue UNIVERSITY OF KANsAs SCIENCE BULLETIN E> XE E €£ ” = ° ° ° 130 fof fe) Vv 132 Fics. 126-132. Neopanorpa borneensis Byers, structural details of male holotype and female allotype. 126, left fore wing of male. 127, genital bulb, ventral aspect. notal organ. 129, genital bulb, left lateral aspect. 128, abdominal terga 3 and 4, dorsal aspect, showing 130, subgenital plate of female, ventral aspect. 131, genital plate, ventral aspect. 132, same as 131, right lateral aspect. Scale a—Fic. 126; scale b—Fics. 127-130; scale c—Fies. 131-132. nean species has been discussed by Byers (1966). Neopanorpa flavicauda Banks, 1931 Description modified from Banks (1931b) and Byers (1966). Rostrum pale laterally. Antennal scape pale; left and half of right flagella lacking in lectotype. Legs yellowish brown. Wings (Fig. 133): Clear, bands and spots almost black. Apical band complete. Pterostigmal band complete, broadly joined to apical band along costal margin, forked posteriorly, with distal branch as Tue MecopTera oF INpoNnEsIA: Genus Neopanorpa 393 Fics. 133-136. Neopanorpa flavicauda Banks, structural details of male paralectotype. 133, right fore wing. 134, genital bulb, ventral (posterior) aspect. 135, genital bulb, right lateral aspect. 136, abdominal terga 3 and 4, dorsal aspect, showing notal organ. Scale a—Fic. 133; scale b—Fics. 134-136. broad as proximal branch; distal branch joined to apical band at margin in some specimens (not in lectotype). Marginal spot conspicuous. Basal band entire, slight- ly spurred proximad between veins Cui and Cuz. A small basal spot near mid- width of wing before basal band in most specimens. Abdomen of male: Segments 7-9 pale yellowish brown. Segment 6 cylindri- cal and plainly longer than 5th; segments 7-8 shorter than 6th, 7th longer than 8th. Posterior process of tergum 3 linguiform in some males (Fig. 136), longer and more slender in lectotype than in male illu- strated here. Dististyles thick, their outer margins markedly concave in basal half, abruptly curved at tips. Basal lobes of dististyles conspicuously divided into prox- imal and distal protuberances, distal one slightly the larger. Hypovalves of sternum 9 expanded basally, attenuate apically (Fig. 134), with a rounded basal lobe on mesal margin and a blunt dorsolateral projection near mid-length (Fig. 135). Aedeagus not examined. Females have not been examined. Lectotype: Male, Kenokok, Mt. Kinabalu, Sabah (British North Borneo), 1100 m, 26 April 1929, in the British Mu- seum (Nat. Hist.), London. Paralecto- types: 4 ¢ 3, 4 2 9%, same data as for 394 Tue UNIVERSITY OF KANsAs ScIENCE BULLETIN smoky brown. Apical band entire, some- what faded posteriorly in both males, with included pale spot posteriorly in two fe- males. Pterostigmal band entire, forked posteriorly, connected along costal margin to apical band. Cross-veins in area be- tween apical and pterostigmal bands pale. Marginal spot extending from costa to M1: +2 or farther. Basal band entire, strong- ly constricted near mid-length in male, not in female. Small basal spot present over cross-veins Ist m-cu and Ist cu. Abdomen of male: Terga 14 black, tergum 5 dark brown; corresponding sterna sordid yellowish brown mottled with brown. Segment 6 dark brown with yellowish brown apex; segments 7-8 yel- lowish brown; segment 9 dark yellowish brown, except tips of dististyles brown. Posterior process of tergum 3. slender, parallel-sided in apical half, extending about half-way across tergum 4 (Fig. 128). Prominence on tergum 4 rounded and smooth posteriorly, flattened and _ hairy anteriorly. Hypovalves dark yellow-brown, darkened apically, widened near mid- length, acuminate in apical half and bear- ing a few long hairs near apex, each with a stout, strongly sclerotized dorsolateral projection (Figs. 127, 129). Tergum 9 evenly tapered toward apex, with truncate apical margin. Dististyles stout, outer mar- gins slightly concave near base, darkly sclerotized beyond mid-length. Basal lobes of dististyles comprising an elongate, hairy, ventral (posterior) projection and a more dorsal, thickened, less hairy portion sclero- tized on lower, mesal margin (Fig. 127). Ventral valves of aedeagus sclerotized, rounded in lateral aspect, not projecting between hypovalves; bases of ventral para- meres in form of narrow, acutely tipped blades projecting backward (ventrad) (shown as shaded V-shaped structure be- tween bases of hypovalves, in Fig. 127). Abdomen of female: All terga brownish black, sterna mottled, sordid brown; cerci black. Subgenital plate of sternum 8 broadly and shallowly notched at apex, with long setae on each apical lobe and on lateral margins of plate; pig- mentation of plate in definite bilateral pattern with pale median streak (Fig. 130). Axial portion of genital plate broadly oval, flanked by wide, rounded, dorsoven- trally thickened lateral lobes (Figs. 131, 132); arms of plate broadly spatulate, slightly twisted basally. Body length: Male, about 12-13 mm (holotype 13 mm); female, about 10-14 mm (allotype 14 mm). Length of fore wing: Male, 13.6-14.4 mm (holotype 13.6 mm); female, 13.8-14.8 mm (allotype 14.6 mm). Holotype: Male, Tenompok, 48 km E of Jesselton, Sabah (British North Bor- neo), 1460 m, 26-31 Jan. 1959, T. C. Maa. Allotype, female, Keningau, Sabah (Brit- ish North Borneo), 12-17 Jan. 1959, T. C. Maa. Paratypes: 1 ¢, 5.9 99 Bunee Tukan, Sabah (British North Borneo), 18 Feb. 1959, T. C. Maa. Holotype, allo- type and 4 paratypes in collection of the Bernice P. Bishop Museum, Honolulu, Hawaii; 1 ¢,1 2 paratypes in Snow En- tomological Museum, The University of Kansas. A male lacking abdominal seg- ments beyond the 6th, from Kiam Base, north of Mt. Kinabalu, 4 Feb. 1959, was not made a paratype. The three known Bornean species, flavicauda, spicata and borneensis, are easily differentiated from the Javanese and Sumatran species by the longer ab- dominal segments 7-8, the shapes of the accuminate-tipped hypovalves and aedea- gus of the male, characters which also show the close relationship among them. The axial portion of the genital plate of Bornean species is usually rounded, as in species of the fractura group, and the lat- eral lobes of genital plate are much larger and more conspicuous, if present. The supposed relationship among these Bor- Tue Mecoprera oF INponEsIA: GENus Neopanorpa 395 140 Fics. 137-140. Neopanorpa spicata Byers, structural details of male holotype and female allotype. 137, left fore wing, male. 138, abdomen of male, left lateral aspect, showing elongate posterior process of tergum 3 (notal organ). 139, genital bulb of male, ventral aspect. 140, genital plate of female, ventral aspect. Scale a—Fics. 137-138; scale b—Fic. 139; scale c—Fic. 140. lectotype, except 22, 23 and 26 April 1929, in the British Museum; 1 ¢, in the Uni- versity Zoological Museum, Copenhagen; Perce 2 2? and I 6 “not labelled as a type but with the same data as the lecto- type, in the Museum of Comparative Zoology, Harvard University. For the probable relationships of this species, see under N. borneensis. Neopanorpa spicata Byers, 1966 Description mostly repeated from orig- inal description (Byers, 1966), based on 11 males and 4 females. Head: Dorsum glossy black; frons below antennal sockets dark brown; ros- trum brown anteriorly, yellowish brown laterally; mouthparts yellowish brown. Antennal scape yellowish brown, pedicel brown; flagellum (both incomplete in hol- otype) black, with 40 to 42 flagellomeres. Thorax: Pronotum black, with 4 or 5 bristles at each side on anterior margin. Mesonotum and metanotum mostly dark brownish black, each with poorly defined, pale yellowish brown areas on posterolat- eral shoulders of scutum near wing bases. Pleura and coxae sordid yellowish brown. Femora, tibiae and basitarsi sordid yellow- ish brown, apical tarsomeres brown. Wings (Fig. 137): Mostly clear, slightly iridescent, with bands and _ spots 396 Tue UNIversiry oF KANsAs ScIENCE BULLETIN smoky brown. Apical band in fore wings weakly indicated along margin in holo- type, allotype and 7 paratypes, broader but deeply indented or including variously shaped clear areas in other paratypes. Pterostigmal band complete, constricted and almost broken near mid-length, forked posteriorly in some specimens. Marginal spot small, basal band short, not darkly colored (holotype) to complete but slender. A small basal spot in cell 1st Cui in one paratype. Abdomen of male: Terga 1-3 dark brown, terga 45 dark yellowish brown; sterna 25 dark yellowish brown. Segment 6 yellowish brown, segments 7-8 pale yel- lowish brown, segment 9 dark yellowish brown. Posterior process of tergum 3 long, stout, curved upward, densely set with black hairs especially on underside, extending to or slightly past mid-length of segment 6 (Fig. 138). Hypovalves of sternum 9 not darkened, separated throughout their length with small mesal lobe near base, widened near mid-length and tapered to acuminate tips (Fig. 139). Tergum 9 slightly narrowed toward trun- cate apex, with subapical ventral process on each side bent around segment 10. Dististyles long, slender, outer margins nearly straight except curved at tips. Basal lobes of dististyles strongly sclerotized me- sally, shallowly excavated below. Ventral valves of aedeagus sclerotized, projecting posteriorly and ventrally (backward) be- tween hypovalves; bases of ventral para- meres in form of narrow, blunt-tipped blades projecting backward. Abdomen of female: Terga 1-4 and about basal half of 5 brownish black; cor- responding sterna pale yellowish brown; segments 6-10 dark yellowish brown; cerci black. Subgenital plate of sternum 8 api- cally notched with 2 thick, finger-like lobes above gonopore on dorsal surface. Axis of genital plate ovoid; genital plate with- out conspicuous lateral lobes (Fig. 140); arms of plate spatulate, somewhat twisted near base. Body length: Male, about 11-13 mm (holotype 12 mm); female, about 9-11 mm (allotype 10 mm). Length of fore wing: Male, 12.2-13.2 mm (holotype 13.2 mm); female, 13.1-13.6 mm (allotype 13.6 mm). Holotype: Male, Tenompok, 48 km E of Jesselton, Sabah (British North Bor- neo), 1460 m, 17-21 Oct. 1958, T. C. Maa. Allotype, female, same data as for holo- type. Paratopotypes: 2 2 2, 17-21 Oct. 1958, L. W. Quate; 2 ¢ ¢, 20 Oct. 1958, L. W. Quate; 3 ¢ ¢,; 19,2631 Jane T. C. Maa. One additional ¢ paratype, Singkor, Sabah (British North Borneo), 19 Jan. 1959, Maa. The holotype is pinned together with a small ichneumonid wasp on which it was feeding at the time of capture. Holotype, allotype and 10 para- types in the Bernice P. Bishop Museum, Honolulu, Hawaii; 2 ¢ ¢ and 1 2 para- types in the Snow Entomological Museum, The University of Kansas. The probable relationships are men- tioned under N. borneensis. The elongate notal organ (posterior process of abdom- inal tergum 3), the longer abdominal seg- ments 7-8 and the shapes of the hypo- valves and aedeagus of the male allow spicata to be easily differentiated from other Bornean species of Neopanorpa and from the Javanese and Sumatran species. Uncroupep SPECIES Neopanorpa lieftincki, new species Description based on 3 males, 1 female, pinned. Head: WDorsum glossy brownish black; frons below antennal sockets dark brown; rostrum dark brown anteriorly, yellowish brown laterally; mouthparts dark brown, tips of palps blackish brown. Antennal scape brown, pedicel dark brown, flagellum dull brown at base, grad- ing into dull dark brown, with 42 (holo- type) to 43 flagellomeres. Tue MecopTera oF INpoNEsIA: GENus Neopanorpa 397 Fic. 141. Neopanorpa lieftincki, new species, right wings of female paratype. Thorax: Pronotum brownish black, bearing 4 bristles on each side along an- terior margin. Mesonotal and metanotal scuta mostly black, brown along postero- lateral margins and dark brown near wing bases; scutella dark brown. Pleura and coxae dull yellowish brown, brown (two paratypes) or dark brown (female). Fe- mora yellowish brown; tibiae and tarsi dark brown. Wings (Fig. 141): Almost colorless and slightly iridescent, with faint smoky brown markings. Apical band reduced to small, faint spots, or almost absent. Ptero- stigmal band indistinct, slender, constricted or broken at mid-length, with distal branch present only as separate spot at posterior margin of wing. Marginal spot absent. Basal band represented by two faint spots. Hind wing clear except for faint, poorly defined pterostigma. Abdomen of male: Terga 15 black- ish brown to dark brown (one paratype); sterna 2-5 yellowish brown. Segment 6 blackish brown dorsally, dark brown ven- trally. Segments 7-9 brown (7 reddish brown in one paratype), except hypovalves of sternum 9 dull yellowish brown and dististyles dark brown apically. Posterior process of tergum 3 short, triangular, ex- tending about one-third across tergum 4 (process deformed in holotype, bearing a projection laterally on left, Fig. 147). Hy- povalves of sternum 9 dull yellowish brown, expanded in apical 2/3, mesal mar- gins broadly overlapped, outer margins strongly infolded dorsally, folded edges subparallel in ventral aspect (Fig. 142). Ninth tergum yellowish brown, slightly constricted near apex and bearing long hairs at both sides of very slight emargina- tion (Fig. 146). Cerci pale basally, dark brown apically. Basistyles brown, with two large, membranous pouches (probably pheromone sacs) anteroventrally between them (Figs. 142, 145); dististyles brown at base, dark brown apically, outer mar- gins slightly concave in basal half (Fig. 142). Basal lobes of dististyles semicircu- lar in ventral aspect, with crescentic ven- tromesal concavity; a few large pale hairs on ventral and distal surfaces (Fig. 145); a long seta projecting from dorsal surface of lobe and about 4 apically curved, small- er spines projecting from anterodorsal sur- face (Fig. 144). Dorsal, basal margins of 398 Tue UNIversiry oF KANsAs ScIENCE BULLETIN basal lobes clear, thin subacute blades. Aedeagus slightly sclerotized, reddish brown; ventral valves large, narrowing ventrally to broad, appressed blades, slight- ly divergent at apex, projecting well be- yond much smaller, blackened dorsal valves (Fig. 148); lateral processes simple, concave posteromesally; dorsal parameres subrectangular with uneven dorsal margin (Fig. 148); ventral parameres absent. Abdomen of female: Terga 1-6 black; sterna 25 sordid dark yellowish brown, sternum 6 dull brown. Segment 7 dark brown. Terga 8-10 brown, corre- sponding sterna light brown. Cerci dark brown. Subgenital plate mostly yellowish brown, brown laterally, with — slightly notched apex and a whitish, hairless streak (about one-third length of plate) anterior to notch; bearing long setae along lateral margins and at each side of notch. Axial portion of genital plate ovoid, with stout, divergent anterior apodemes (Fig. 150); arms of plate broad, subtranslucent toward apex, twisted and yellowish brown basally. Body length: Male, 14.3-19.1 mm (holotype 18.3 mm); female (allotype), 16.7 mm. Length of fore wing: Male, 13.214 mm (holotype 13.7 mm); female (allotype), 14 mm. Holotype: Male, crater of Mt. Tang- gamus, southwestern Lampongs, southern Sumatra (see map), 1900 m, 1 Jan. 1935, M. A. Lieftinck. Allotype, female and 2 male paratypes, same data as for holotype. Holotype, allotype and 1 ¢ paratype in the Rijksmuseum van Natuurlijke His- torie, Leiden; 1 4 paratype in the Snow Entomological Museum, The University of Kansas. Lieftinck (1936) says this species was immediately recognized as undescribed be- cause of its vivid grass-green color during life. The four specimens he found were in the ancient crater of Mt. Tanggamus, about 33 m below the summit (2100 m). The crater holds a well-protected marsh with a luxuriant growth of Rhododendron. This species cannot be placed in any of the groups recognized above. It is close to the muelleri group on the basis of struc- ture of aedeagus and the basal lobes of dististyles (except that it lacks the black dorsal spines). In the obscure, faint wing pattern, it looks identical with diloba. It differs from diloba by the shorter hypo- valves and the way they are infolded. The genital bulb is also a little longer and more slender than those of diloba and muellert, especially the dististyles. This species is named in honor of its discoverer, Dr. M. A. Lieftinck, in recognition of his great contributions to the knowledge of Indonesian Mecoptera. COMPARISON BETWEEN NEOPANORPA AND LEPTOPANORPA The genus Leptopanorpa, as mentioned earlier, occurs only in Indonesia, possibly in Java, only (there is a record of L. char- pentiert from Sumatra, but it is doubtful), and contains 12 species. The records of Leptopanorpa from elsewhere than Java (effusa, Sikkim; furcata, Nepal; and ja- vanica, in Hainan, Burma, etc.), pertain probably to Neopanorpa instead of to Lep- topanorpa (Byers, 1965, 1971; Rust and Byers, 1976). Because of this confusion, Neopanorpa is compared with Leptopa- norpa below. Neopanorpa and Leptopanorpa differ obviously in the body length of males. The body of males of Neopanorpa is about as long as the length of the wings or some- what longer, but in Leptopanorpa it is usually much longer than the length of the wings (except cingulata and javanica). The different abdominal length is due mainly to the shape of abdominal segments 6-8. In Neopanorpa, these are robust and usually not elongate. In contrast, in Lepto- panorpa they are ordinarily long and slen- Tue Mecoprera oF INpoNnEsIA: Genus Neopanorpa 399 G.wW.B. ‘ 147 ° Fics. 142-148. Neopanorpa lieftincki, new species, structural details of male. 142, genital bulb, ventral aspect; holotype. 143, ninth abdominal tergum and sternum, right lateral aspect; paratype. 144, base of right dististyle and dorsal parameres, dorsal aspect; paratype. 145, base of left dististyle and aedeagus, ventral aspect; para- type. 146, ninth abdominal tergum, dorsal aspect, most hairs omitted; paratype. 147, abdominal terga 3 and 4, dorsal aspect showing atypical (deformed) notal organ; holotype. 148, aedeagus, left lateral aspect; paratype. Scale a—Fics. 142-143, 146-147; scale b—Fies. 144-145, 148. der (except in cinglata, inconspicua, javan- ica and petersen1). Wings of Neopanorpa are usually broad, whereas those of Lepto- panorpa are slender and much narrowed basally, except in javanica (similar to Neo- panorpa). The shape of the genital bulb also dif- ferentiates these two genera. In Neo- panorpa, it is broad, generally oval in ven- tral aspect, and without a stalk at the base of the 9th segment. In Leptopanorpa, it is slender, generally elliptical, and has 400 Tue UNIversity oF Kansas SciENCE BULLETIN — ow Ss Ga a ~~ N —— Se 150 Fics. 149-151. Neopanorpa lieftincki, new species, details of genitalia of female paratype. 149, subgenital plate, ventral aspect, hairs partially omitted. 150, genital plate, ventral aspect, only partly shaded. 151, genital plate, right lateral aspect. either a fairly long, basal petiole (7itsemae, etc.) or a shorter one (cingulata, etc.). In the males of Neopanorpa, the hy- povalves are usually broad, overlapping mesally or distally. Those of the Indian species N. hirsuta and N. chillcotti are more slender, as in some species of Lepto- panorpa. Wowever, in some species of Leptopanorpa, the hypovalves are broader than those of most Neopanorpa, yet never overlapping distally. The ninth abdomi- nal tergum most easily differentiates these two genera. In most male Neopanorpa, tergum 9 has a broadly rounded distal margin, but in some species it is shallow- ly emarginate (e.g. muelleri); and in most it has subapical, lateroventral proc- esses curving around segment 10 toward the interior of the genital bulb. In Lepto- panorpa the median distal margin of ter- gum 9 either protrudes caudad slightly (as in charpentiert), or forms a distinct process (ritsemae, etc.). Terga of either genus have the lateroventral processes directed toward the interior of the bulb, as in most species of Neopanorpa. The aedeagus of Neopanorpa is usually located more or less equidistantly from sternum 9 and tergum 9, whereas that of Leptopanorpa is more ventral in position, so that the dorsal parameres are approxi- mately halfway between the ninth tergum and sternum. The ventral parameres of the aedeagus of Neopanorpa may be either fairly long on mainland Chinese species and a few Indonesian species (diloba and angustiapicula), short on most Indo-Chi- nese species, or absent, as from most Indo- nesian species, whereas those of Lepto- panorpa are always large and complicated. The dorsal parameres of Neopanorpa are small, shorter than the ventral valves and Tue Mecoprera oF INponEsIA: GENus Neopanorpa 401 almost invisible from dorsal view of the intact genital bulb, whereas those of Lep- topanorpa are large, longer than the ven- tral valves and conspicuous in dorsal aspect. The ventral valves of Neopanorpa are the most conspicuous structures of the aedeagus in most species (except diloba and angustiapicula), whereas those of Leptopanorpa are much reduced and hard- ly visible. The lateral processes of Neo- panorpa are broad and are evenly divided in diloba and angustiapicula, whereas those of Leptopanorpa are small. One important difference is the shape of the basal lobes of the dististyles. They are rarely bilobed in Neopanorpa (except diloba) and most are concave anteroven- trally (Indonesian and Indo-Chinese spe- cies), a few concave anteriorly (Indian species); but they are usually bilobed hori- zontally (in ventral aspect) in Leptopa- norpa, except in petersent, cingulata, filt- cauda and inconspicua, and are often con- cave dorsally. The female of Neopanorpa is much like that of Leptopanorpa in general appear- ance, but the genitalia of the two genera are much different. The subgenital plate in Neopanorpa is pale, apically, on most Indonesian species and some Indo-Chinese species. That of Leptopanorpa has a generally uniform color. The genital plates of Neopanorpa are much more flattened (except umbo- nata) than those of Leptopanorpa (except of ritsemae, filicauda). In Neopanorpa, the axial portion of the genital plate is sometimes forked anteriorly, whereas in Leptopanorpa it is either bluntly rounded or with broad basal plates, or protrudes ventrally, or bears a pair of small basal processes. The arms of the genital plate of Neopanorpa are usually broader than those of Leptopanorpa (except robusta, cin gulata and sarangana, in which they are as broad as those of Neopanorpa). This difference 0.5mm Fics. 152-153. Leptopanorpa charpentiert, female syn- type. 152, genital plate, ventral aspect. 153, genital plate, left lateral aspect. is not easily noticed, unless specimens to compare are available. As discussed above, under the heading of Zoogeography, we have examined the syntypes (1 6, 1 2) of Leptopanorpa charpentieri (Burmeister) and have com- pared the genital structures of the female with those of all Indonesian species. The genital plate (Figs. 152, 153) is clearly that of a Leptopanorpa and, in our opinion, the same as that described for L. nematogaster (MacLachlan) by Lieftinck (1936: 302, plate 14). Key To GENERA OF PANORPIDAE 1. Vein 1A in front wing joining wing margin far beyond level of origin of radial sector; two cross-veins between 1A and 2A; in hind wing, 1A united with Cuy for only a short distance NC ahee Wikre DASE. a eee eee Panorpa Vein 1A short, extending little more than half distance from wing base to origin of radial sector in front wing; one cross-vein between 1A and 2A; in hind wing, 1A fused with Cuy nearly its entire length except for short distal portion resembling diagonal cross- [si 0 VAs MNO DED ote Bo 2 RC 2 402 Tue UNIversiry oF KAnsAs ScrENCE BULLETIN 2. Abdomen of male long and slender, segment 7 twice or more (usually 2.5- 3.5 times) length of segment 5; seg- ment 3, excluding appendage, as long as or longer than its width; segment 9 tapering to slender, often elongate petiole at base; abdominal tergum 9 with small to conspicuous median caudal projection; axial portion of genital plate of female dorsally ex- panded, appearing thick when viewed PLO tne Siders oh! ner (ea Ra Ce Leptopanorpa Abdomen of male shorter, extending only slightly beyond wing tips, with segment 7 not twice length of seg- ment 5 and usually 1-1.5 times as long; segment 3, excluding append- age, wider than long; segment 9 nar- rowing abruptly to short, inconspicu- ous petiole at base (except N. effusa and N. furcata of Himalayas); ter- gum 9 without median caudal projec- tion; axial portion of genital plate of female without dorsal expansion ........ Diets Tae 9 YG SEE ES ee Neopanorpa GAZE LTEER Localities in the following gazetteer are arranged alphabetically. The number before each listed locality corresponds to a numbered locality on accompanying maps (Figs. 25). On the maps, localities are numbered from left (west) to right (east), except the maps of Java precede that of Sumatra. Data pertinent to each locality appear in the following sequence: 1. Name of locality as recorded on or translated from pin label. 2. Variant spellings of name, or synonymous names, in parentheses. 3. General region (W. J.—West Java; E. J—East Java; C. J.—Central Java) in which locality is situated. 4. Geographic coordinates, if avail- able, or reference to another locality that is well known. 5. Elevation where collection was made, if available. 55. uo) 50. i: 44. te 41. 6. Brief description of habitat, if available. LocaLirtieEs . ‘Airbangis, Sumatra; 0°12" N7 99-23 not below 1200 m; in the Padang High- land, not far from Fort de Kock. Andalas, Tandj, Sumatra; 0°07’ S, 100° 41’ E; no elevation recorded; a village. Bajukidul (Baju-Kidul), Mt. Raung, E.. J.; 8°12” S, 114°09 #3 500-300 sm, estate at foot of Mt. Raung, west of Ban- juwangi; on the eastern coast of Java. Bandarbaru (Bandar-Baru), northeastern Sumatra; 3°16’ N, 98°33’ E; 950 m; a village. Baung, Banten, E.-J.; 8°02 Sp 113-22" E; at some altitude above Malang; a for- est reserve. . Mt. Bengbreng, E. Djampang, W. J.; 7°13! S,-107°20) E; 400m. _ Mt. Bésér,; W. J.; 7°11’ S, 109°3% E; 1300 m. Bibidjilan, Distr. Banten, south coast, W. J.; 200 m. Blawan, Idjen Plateau, E. J.; 7°59” S, 114°09’ E; 950 m. Bodjongkalong, Djampangs, W. J.; 7° 00’ S, 107°42’ E; 300 m. . Brastagi, northeastern Sumatra; 1400 m; a mountain resort at foot of Mt. Sibajak, some hours by car south of capital, Medan. . Bukittinggi—see Fort de Kock. Mt. Bunder, Salak, W. J.; about 6°42’ S, 106°44’ E; 800 m. . Mt. Dempu, Air Njuruk, Sumatra; © 07’ S, 104°55’ E; no elevation recorded. . Depok, W. J.; 6°24’ S, 106°50’ E; 50-100 m; a small nature reserve between Dja- karta and Bogor. Djambangan, Tengger Mts., E. J.; 7°58’ S, 113°37’ E; 1800 m. Djampangtengah, Mt. Malang, W. J.; 7° 03’ S, 106°48’ E; no elevation recorded; a village. Djunggo, Malang, E. J.; 7°48’ S, 112°31’ E; no elevation recorded. Fort de Kock (Bukittingi), Sumatra; 0° 19’ S, 100°22’ E; 920 m; a city. 20. 60. 36. aS) 47. Dee 53. ihe 37. 35. Tue Mecoprtera oF [NponEsIA: GENus Neopanorpa 403 Mt. Gedeh, Tapos, W. J.; 6°47’ S, 106° 59’ E; 700-800 m and 1200-1700 m. Giesting, Mt. Tanggamus, southwestern Lampongs, southern Sumatra; 5°25’ S, 104°42’ E; 450 m; an estate on slope of Mt. Tanggamus. . Mt. Guntur, Kamojang, W. J.; 7°08’ S, 107°50’ E; 1400 m. Jesselton (Kota Kinabalu), Sabah; 5°59’ N, 116°04’ E. Kaliurang (Kali Urang), Djokjakarta (Djocja), C. J.; 7°36’ S, 110°25’ E; low country; a town near Djokjakarta. Kamodjang, Preanger, Mt. Guntur, W. J.; 7°10’ S, 107°48’ E; 1400-1450 m; collection was made in crater, near Garut. Mt. Karang, Bantam, W. J.; 6°03’ S, 106°09’ E; 800-1000 m. Kendeng Mts., Idjen Plateau, E. J.; 8°00’ S, 114°15’ E; 2000 m (1200-1300 m, acc. to Lieftinck). Mt. Kinabalu (Kini Balu, Kina Balu, Kina Batu), Sabah (British North Bor- neo); 6°05’ N, 116°30’ E; collection at 3300 ft (1015 m). Lembang, Preanger, W. J.; 6°49’ S, 107° 36’ E; 900 m; a summer resort above Bandung. Lubuksikaping (Lubuk-Sikaping), west- ern Sumatra; 0°08’ N, 100°10’ E; 450 m; a town. PeMie Madiun. . Ji2-7 7377S, 111°31’ E; 1200 m. Mt. Malang, Djambangan, W. J.; 6°28’ S, 106°43’ E; 800 m. Mt. Manglajang, Preanger, W. J.; 6°52’ S, 107°44’ E; 1400 m. Marcodjambangan, Mt. Muria, northern Java; about 6°36’ S, 110°53’ E; 1000 m; one of the elevations of the Mt. Muria Complex. . Mt. Megamendung, W. J.; 6°38’ S, 106° 55’ E; 600 m; south-east of Bogor, in the hills near Mt. Gedeh. Muarasako (Muara Sako), Sumatra; 2° 07’ S, 101°13’ E; no elevation recorded. . Nongkodjadjar (Nongkodjadar), Teng- Bere Mts. Ea est / 254 9, 1249" E: 1200 m; a summer resort. Oengaran—see Mt. Ungaran. 48. De. 58. 1B Se) . Mt. Tangkubanprahu . Tyibodas, Mt. Gedeh, W. J.; Ongop-ongop, Idjen Plateau, E. J.; about 8°05’ S, 114°15’ E; 1850 m; a village. Mt. Pangerango, W. J.; 6°46’ S, 106°57’ E; no elevation recorded. 4 Pantjan Mts. Mt. Gedeh; W. Je36° 35'S; 106°54’ E; 1500 m. . Pateungteung, W. J.; 7°09’ S, 107°46’ E; 1300 m; a town. . Patjet, Mt. Bésér, W. J.; 7°34’ S, 109°50’ E; 1300 m. Preanger, W. J.; 7°12’ S, 106°47’ E; 1500 m; an upland area in W. J. ) Radjamandalay WJes 6 5001S; 107-20" i; 335 mi;a village: Selabintana, Mt. Gedeh, W. J.; 6°52’ S, 106°57’ E; 1000 m; near Sukabumi. Y Semarane.aCs jisG25 605) 1102259 9 ino elevation recorded. Serapai, Mt. Kurintji (Korinchi), Su- matra; about 1°42’ S, 101°16’ E; 1000- 1400 m. Subanajam (Suban Ajam), Sumatra; about 4°15’ S, 103°00’ E; 800-1000 m; in Bengkulen. . Sukabumi, W. J.; about 6°13’ S, 106°46’ E; no elevation recorded. . Sukanegara, Djampangs, W. J.; 7°06’ S, 107°07’ E; 700-1000 m; in a damp forest. Sumberbrantas (Sumber Brantas), Mt. Ardjuno-Mt. Andjasmoro, E. J.; about 7°45’ S, 112°30’ E; 1500-2400 m. Tandjunggading, west coast of Sumatra; 3°18’ N, 99°20’ E; 1000 m; a village. Mt. Tanggamus (crater), southwestern Lampongs, southern Sumatra; 5°25’ S, 104°42’ E; 1900 m; Neopanorpa lief- tincki n. sp. was taken here, in the forest below summit of Mt. Tanggamus. (Tangkuban Prahu), Preanger, W. J.; 6°44’ S, 107°36’ E; 1400 m. + Melomojo} Resse Keduyi C5.) 547" 22048; 110°24’ E; 1200 m. Tenompok, 48 km E of Jesselton, Sabah; approximately 6°00’ N, 116°30’ E; south foot of Mt. Kinabalu; elevation not re- corded but probably near 2000 m. 65S. 107°01’ E; 1550-1700 m. 404 Tue University oF KANsAs ScrENCE BULLETIN 2. Tjibunar Estate, Mt. Gedeh-Pangrango, W. J.; 6°50’ S, 106°04’ E; 1000 m; on west slope of Mt. Gedeh-Pangrango. 38. Tyilo, Muria Mts., C. J.; 6°40’ S, 110°54’ E; 800 m. 8. Tjiomas, Warungloa (Warung Loa), Mt. Salak, W. J.; 6°36’ S, 106°45’ E; 300-400 m; SE of Tjipeundeuj, Warungloa is a village at the foot of Mt. Salak, near Bogor. 5. Tjipeundeuj, Leuwiliang, W. J.; about 6°40’ S, 106°29’ E; 800-900 m; a village W of Bogor, in the hills near Mt. Gedeh. 17. Tjisarua, Mt. Pangerango-Gedeh, W. J.; 6°40’ S, 106°59’ E; 1000-1300 m. 10. Mt. Tjisuru, W. J.; about 7°24’ S, 106° 48’ E; 500-600 m; in the middle of the Djampangs District. 33: Mt., Ungaran, C..J= 7°12’ S, 110°20' E; 800-1200 m; where E. Jacobson collected many years. 61. Wai Lima, Lampongs, southern Suma- tra; no elevation recorded; a coastal vil- lage in the Lampong District, extreme southern Sumatra. 60. Wai Tebu, Mt. Tanggamus, southern Su- matra; 300-500 m; a river near Giesting at foot of Mt. Tanggamus. 6. Wynkoops Bay, Pelabuhanratu, south- western Java; 7°03’ S, 106°27’ E; no ele- vation recorded; in low country. LITERATURE CITED Banxs, N. 193la. Some neuropteroid insects from the Malay Peninsula. J. Federated Malay States Museums 16:377-409. —. 193lb. Neuropteroid insects from North Borneo, particularly from Mt. Kinabalu. J. Federated Malay States Museums 16:411-429. BEMMELEN, R. W. van. 1949. The geology of Indo- nesia. Vol. 1A. Hague Government Printing Office. 732 p. BurMEIsTeR, H. 1839. Family Panorpina. Handbuch der Entomologie, vol. 2:951-959. Byers, G. W. 1965. The Mecoptera of Indo-China. Pacific Insects 7:705-748. —. 1966. Mecoptera from Borneo and Tioman Island. Pacific Insects 8:885-892. . 1967. Synonymy in the Panorpidae (Mecop- tera). J. Kansas Entomol. Soc. 40:571-576. —. 1971. A new Neopanorpa from Nepal. J. Kansas Entomol. Soc. 44:534-539. CHENG, F. Y. 1957. Revision of the Chinese Mecop- tera. Bull. Mus. Compar. Zool., Harvard Coll. 116:1-118, 23 plates. ENpERLEIN, G. 1910. Uber die Phylogenie und Klassifikation der Mecopteren unter Bertick- sichtigung der fossilen Formen. Zoologischer Anzeiger 35:385-399. 1912. Zur Kenntnis der Mecopteren Javas. Notes, Leyden Museum 34:235-239. EsBEN-PETERSEN, P. 1913. Mecoptera and Planipen- nia collected in Java by Edward Jacobson. Notes, Leyden Museum 35:225-236. —. 1915. Synonymic list of the order Mecop- tera. Entomologiske Meddelelser 10:216-242. —. 1921. Mecoptera. Collections Zoologiques du Baron Edm. de Selys Longchamps. Fasc. Ds lf2ip: Bruxelles: Ferris, G. F. anv B. E. Rees. 1939. The morphology of Panorpa nuptialis Gerstaecker (Mecoptera: Panorpidae). | Microentomology 4:79-108, plates 36-51. GressitTT, J. L. 1958. Zoogeography of insects. Ann. Rev. Entomol. 3:207-230. Hesse, R., W. C. ALLEE AND K. P. Scumipr. 1951. Ecological animal geography. Second edition. John Wiley and Sons, Inc. New York. 674 p. Karny, H. H. 1923. Panorpidae from Java. Treubia 3:381-383. Liertinck, M. A. 1936. Studies in Oriental Mecop- tera. I. The genus Leptopanorpa in Malaysia. Treubia 15:271-320, plates 4-14. MacLacuian, R. 1869. Synopsis of the species of Panorpa occurring in Europe and the adjoin- ing countries; with a description of a singular new species from Java. Trans. Entomol. Soc. London, ser. 4, 2:59-70, plate 4. Mayr, E. 1944. Wallace’s line in the light of recent zoogeographic studies. Quart. Rev. Biol. 19: 1-14. Navas, L. 1914 (1913). Neuroptera asiatica. Ento- mologicheskoe Obozrenie (Revue Russe d’En- tomologie), ser. 2, 13:424-430. Rrex, E. F. 1954. The Australian Mecoptera. Aus- tralian J. Zool. 2:143-168. Roseouain, C. 1954. Malaya, Indonesia, Borneo and the Philippines. Second edition. Translated by E. D. Laborde. Longmans, Green and Co., London. 457 p. RoepKE, W. 1916. Panorpa’s auf Java (Panorpa angustipennis Westw.). Tijdschrift voor En- tomologie 59:170-174. Rust, M. K. anp G. W. Byers. 1976. The Mecoptera of India and adjacent regions. Univ. Kansas Sci. Bull. 51:19-90, 159 figs. TyepEerR, Bo. 1970. Mecoptera. In, Tuxen, S. ie Taxonomist’s Glossary of Genitalia of Insects, Second edition, p. 99-104. Munksgaard, Co- penhagen. Townes, H. anv S.-C. Curu. 1970. Indo-Australian Xanthopimpla (Hymenoptera). Mem. Amer. Entomol. Inst., No. 14. 303 p. Tue MecopTera oF INpoNnEsIA: GEeNus Neopanorpa 405 vAN BEMMELEN (see BEMMELEN). Westwoop, J. O. 1842 (1841). A monograph of the Van DER WeEELE, H. W. 1909. Mecoptera and Plani- genus Panorpa, together with description of pennia of Insulinde. Notes, Leyden Museum species of various allied genera. Annals and 31:1-100, plates 1-5. Magazine of Natural History 8:298-300. Wattace, A. R. 1869. The Malay Archipelago. Re- Note: This is a summary of a monograph print, 1962, Dover Publications, Inc. New under a similar title published in 1846 in York. 497 p. Trans. Entomol. Soc. London 4:184-196. . be see a batho ‘i ser, Ky ee 7 | 7 e ih te ee oa ae P a ae - at 5 7 ‘ 0 Pele 7 Ce | i - | ; h . onl A ny rare if. Part 2 - ne | M F $x,’ = — ‘ ne na 7] ~ | *. h LU | ty, - }) { a At ay ~~ oS as 3 \\ FWRI ON tie eS ee ee eo Ap See re cit mites og he tp 8 es Aneta rb \ Sr bo Dek BEN, Bi ivind ‘ig iar Pee Apt Hehe Seve cen 4 s F a |