‘ ‘ ' ‘ ’ Aft, Fa if r , ' hp o4 ens J bpns Hp ebb PLEAS > ’ ’ i-8 tay re, Uh ee i i The person charging this material is re- sponsible for its return to the library from which it was withdrawn on or before the Latest Date stamped below. Theft, mutilation, and underlining of books are reasons for disciplinary action and may result in dismissal from the University. UNIVERSITY OF ILLINOIS LIBRARY AT URBANA-CHAMPAIGN BUILDING use L161— O-1096 ° “- aa a a + —_— E : — i mm A, oe te | ry > ¢ 7 , ; » , = » aare NM 52 4 , L' + 2 a dat . Li Hl 1p TIE oY, et week eee Digitized by the Internet Archive in 2011 with funding from University of Illinois Urbana-Champaign http://www.archive.org/details/obligatoryfacult25bald Obligatory and Facultative Insects in Rose Hips Ther R ecognition and Bronomics W. V. BALDUF ILLINOIS BIOLOGICAL MONOGRAPHS: Number 96 RENE ASTRO EER DA AY SA OER CARD THE UNIVERSITY OF ILLINOIS PRESS URBANA, 1959 ILLINOIS BIOLOGICAL MONOGRAPHS is the general title for a cee of ‘mono: ’ graphs in botany, entomology, zoology, and allied fields. Volumes 1 through 24 con- tained four issues each and were available through subscription. Beginning with number 25 (issued in 1957), each publication will be numbered consecutively. No subscriptions will be available, but standing orders will be accepted for forthcoming numbers. Prices __ of previous issues still in print are listed below, and these may be purchased from the University of Illinois Press, Urbana, Illinois. Requests for exchange arrangements should be addressed to the Exchange Department, University Library, Urbana, Illinois. __ BAKER, FRANK COLLINS (1922): The Mol- luscan Fauna of the Big Vermilion River, Illinois, with Special Reference to Its Mod- ification as the Result of Pollution by Sew- age and Manufacturing Wastes. 15 pls. Vol. 7, No. 2. $1.25. BEAVER, PAUL CHESTER (1937): Experi- mental Studies on Echinostoma revolutum (Froelich), a Fluke from Birds and Mam- mals. 8 pls. 20 charts. Vol. 15, No. 1. $1.00. BENNETT, HARRY JACKSON (1936): The Life History of Cotylophoron cotylopho- rum, a Trematode from Ruminants. 9 pls. Vol. 14, No. 4. $1.50. CAHN, ALVIN ROBERT (1937): The Turtles of Illinois. 31 pls. 20 eas 15 figs. Vol. 16, Nos. 1-2. $3.00. CHEN, HSIN KUO (1935): Development of the Pectoral Limb of Necturus maculosus. 11 pls. Vol. 14, No. 1. $1.00. COOPER, ARTHUR REUBEN (1918): North American Pseudophyllidean Cestodes from Fishes. 13 pls. Vol. 4, No. 4. $2.00. CREGAN, SISTER MARY BERTHA (1941): Generic Relationships of the Dolichopo- didae (Diptera) Based on a Study of the Mouth Parts. 30 pls. Vol. 18, No. 1. $1.00. DOAK, CLIFTON CHILDRESS (1935): Evolu- tion of Foliar Types, Dwarf Shoots, and Cone Scales of Pinus with Remarks Con- cerning Similar Structures in Related Forms. 32 figs. Vol. 18, No. 3. $1.50. EDDY, SAMUEL (1934): A Study of Fresh- Water Plankton Communities. 9 figs. Vol. 12, No. 4. $1.00. ESSEX, HIRAM ELI (1928): The Structure and Development of Corallobothrium, with Descriptions of Two New Fish Tape- worms. 5 pls. Vol. 11, No. 3. $1.00. GAMBILL, WILLIAM G., Leguminosae of Illinois. va 22, No. 4. FAUST, ERNEST CARROLL (1918): Life His- tory Studies on Montana Trematodes. 9 — pls. 1 fig. Vol. 4, No. 1. $2.00. FISHER, HARVEY I., and GOODMAN, DONALD c. (1955): The Myology of the Whooping tt Crane, Grus americana. 40 Heh Vol. i P No. 2. $2.50. $3.00. GOODNIGHT, CLARENCE JAMES (1940): The Branchiobdellidae (Oligochaeta) of North American Crayfishes. 8 pls. Vol. 17, aa 3, $1.00. GUTBERLET, JOHN EARL (1915): Vol. 2, No. 2. $.50. HAYES, WILLIAM PATRICK (1930): "ieee 4 phology, Taxonomy and Biology of Larval Scarabaeoidea. 15 pls. Vol. 12, No. ae 5s) $1.25. HEISS, ELIZABETH M. (1988): A Classifica- Me tion of the Larvae and Puparia of the _ Syrphidae of Illinois, Exclusive of Aquatic — Forms. 17 pls. Vol. 16, No. 4. $1.50. HIGGINS, GEORGE MARSH (1921): The iy Nasal Organ in Amphibia. 10 pls. ‘Vol. 6, No. 1, $1.00. HIGLEY, RUTH (1918): Morphology Soe Biology of Some Turbellaria from the © Mississippi Basin. 3 pls. Vol. 4, No, 3. $1.25. HOFF, C. CLAYTON (1942): The Ostrangds Rr of Illinois—Their Biology and Taxonomy, 9 pls. Vol. 19, Nos. 1-2. $2.50. HOFFMEISTER, DONALD F. (1951): A rexel nomic and Evolutionary Study of the — Pifion Mouse, Peromyscus truei. 5 pls. bs, figs. Vol. 21, No. 4. Cloth only, $3.50. | (1958): “The ‘ 7 4 = i. ir “ve " be pte ac tg SS —~ a. On the Osteology of Some of the Loricati. 5 bs , > a ae of oe ee ee, aT. = = a ee ae vy iy OBLIGATORY AND FACULTATIVE INSECTS IN RosE Hips Obligatory and Facultative Insects in Rose Hips Their Recognit won and Bronomics W. V. BALDUF ILLINOIS BIOLOGICAL MONOGRAPHS: Number 26 THE UNIVERSITY OF ILLINOIS PRESS URBANA, 1959 Board of Editors: WILLIAM R. HORSFALL, H. ORIN HALVORSON, FRANCIS J. KRUIDENIER, WILSON N. STEWART, AND AUBREY B. TAYLOR. FOR LIBRARY OF CONGRESS CATALOG ENTRY, SEE PAGE 194, THIS MONOGRAPH IS A CONTRIBUTION FROM THE DEPARTMENT OF ENTOMOLOGY, UNIVERSITY OF ILLINOIS. DISTRIBUTED: MARCH 16, 1959. © 1959 BY THE BOARD OF TRUSTEES OF THE UNIVERSITY OF ILLINOIS. MANU- FACTURED IN THE UNITED STATES OF AMERICA. CONTENTS MMMEIEMIOUAGOIN ttbitice Sis Ps - 70 The Long-Tailed Parasite. Hymenoptera..................+-: (2 Scambus hispae (Harris), Ichneumonidae, Hymenoptera... ... 72 Euderus cushmani (Crawford), Eulophidae, Hymenoptera.... 74 AN UNIDENTIFIED ECTOPARASITE OF Cydia packardi................ 75 BUENOAUOES: OF CUOIG DOCKING. 6 ori. aoc 2 she oushantrrn os $e vs trees wma es 75 Enoclerus? sp. and Phyllobaenus sp., Cleridae, Coleoptera.... 75 Carposina scirhosella 4. s., CARPOSINIDAE, LEPIDOPTERA...........- ne THE SCAVENGEROUS FLY, Lonchaea polita sAY, LONCHAEIDAE, DIPTERA.. 78 THE ROSE cCuRCULIO, Rhynchites bicolor (¥FABR.), CURCULIONIDAE, PER OPAEE NA WAT St Aes 23s eRe eo Ia ae vas ad gard oe 78 Rhynchites bicolor bicolor (Fabr.) in the East.............. 83 Biynchites bicolor insthe Wests. (me 52 tare ee 2 eects 96 NATURAL ENEMIES OF Rhynchites bicolor (FABR.)............+.++- 105 Luchatema baldufi Walkley, Ichneumonidae, Hymenoptera.... 105 Bracon mellitor Say, Braconidae, Hymenoptera.............. 114 Vv Vi CONTENTS Bracon variabilis (Prov.), Braconidae, Hymenoptera.......... ils Zatropis rosaecolis Burks, Pteromalidae, Hymenoptera........ 120 Eurytoma sp., Eurytomidae, Hymenoptera.................. 121 Chaetogaster sp., Oligochaeta, Chaetopoda, Annelida........ 122 SCAVENGEROUS MITE: .5.. «4 0s (hts ua ase oes dws > oe) 3 ee 123 Caloglyphus spinitarsus (Herm.), Acaridae, Acarina, Arachnida 123 THE ROSE SEED CHALCIDS, Megastigmus spP., TORYMIDAE, HYMENOP- PER 6 Ab sed ek HLS Sota ate sew PR eee I Oe 123 Megastigmus aculeatus (Swederus)-....:.2.2..5: . eee 124 Megastigmus nigrovariegatus Ashmead, The American Rose Seéd Chaleid: . 5 6.015058 eR ee A ae 130 PARASITES OF Megastigmus nigrovariegatus................ 0000s ee 145 Eupelmus, Kupelmidae, Hymenoptera ..2.....>.-22 5) Seer 145 ADDITIONAL PARASITIC HYMENOPTERA FROM HIP JARS...........-.- 145 SUMMARYS. 9.04 ede Ge ere Sroka 0 te eee eo EP we 146 ROSE. HOSTS OF! HIPMINSECIS 6. kei. 2) ee 150 BIBLIOGRAPEING «is hon Se aye Ao wee set eke Shale 4 tte eae A rr 15] PEATE ere Ih 4 Sesh sce Grd argpailcana, Gist We RED WO Oe et er 163 INTRODUCTION For some years, my daily walks between home and office took me past a fine stand of Rosa rugosa growing northwest of the “Old Ag. Building” on the University campus at Urbana, Illinois. Here, in the mid-twenties I discovered a red-and-black beetle boring into rose buds with its long beak. During the ensuing years I permitted my interest to extend only to the accumulation of the earliest dates of its annual appearance, pre- suming that so common a species had certainly already been studied adequately. Then in mid-June of 1943, I found several honey-yellow chal- cids, new to me, engaged in inserting their long terebras into the young hips of this rose. My guess was that they were parasitizing some other insect in the hip—possibly the larva of the snout beetle. Seeking to gain data about the geographical range of this chalcid, I picked some hips of R. acicularis at “Eaglesnest,” in northern Minnesota, on September 17, 1943. Upon opening the package at Urbana on October 7, I was sur- prised to find seven stramineous dipterous puparia that had obviously developed from larvae formerly resident in the hips. While removing many such larvae from hips at Eaglesnest in August of 1944, I made the acquaintance of a small caterpillar that shared the fleshy coat of the hips with the dipterous larva. Thus, by August, 1944, I had learned that rose hips in Minnesota and Illinois are sometimes inhabited by four phytophagous species of insects. These proved to be the rose snout beetle, Rhynchites bicolor (Fabr.), Curculionidae; the American rose hip fly, Rhagoletis basiola (O.S.), Trypetidae; the cherry fruitworm, Cydia (Grapholitha) packardi (Zell.), Olethreutidae, and the American rose seed chalcid, Megastig- mus nigrovariegatus Ashm., Torymidae. In later years, the European rose seed chalcid, M. aculeatus (Swed.), was reared from hips obtained from Illinois and states to the east. But these phytophagous residents of rose hips later proved to consti- tute only the smaller portion of the inhabitants more or less characteristic of this microhabitat. From 1944 to 1953, an impressive series of para- sitic Hymenoptera, several new to man, were obtained from the above phytophags. The facts in their lives were learned insofar as permitted by the limits of my time as a teacher and the amounts of host and parasite material available. Since the hips and insects processed in that decade originated princi- pally in Minnesota, Wisconsin, and Illinois, with fewer samples from other northern states, it is probable that still other phytophags and para- 1 2, INSECTS IN ROSE HIPS sites utilize rose hips as food or shelter, particularly in more southerly and westerly areas of the United States. Not only may further investi- gation add more species to the list, but it will certainly serve to further clarify the modes of life and roles of certain species found only in small numbers to date. Perusal of the literature has revealed several additional “hip insects” native to Europe and Russia. The principal objectives were to learn what species of insects depend more or less on rose hips for food and/or shelter, their bionomic rela- tions to the hip or to each other, and the life histories of each species. These goals were reached to varying degrees, depending chiefly on the quantities of living material available for investigation. No instruments were employed to determine the influence of temperatures and relative humidities on the crops of hips or the development and populations of the insect species. Exhaustive technical descriptions have not been in- cluded for most adult and immature stages of the insects found in or upon the hips. Such studies of the immature stages may properly consti- tute the subject matter of another treatise. The characteristics described are believed adequate to enable interested persons to distinguish the several stages of any one species from those of others. Samples of hips from 32 species or varieties of roses were processed in the course of this work. While the bulk of the materials originated in Minnesota, Wisconsin and Illinois, from Rosa acicularis, R. blanda, and R. carolina, 1 collected other lots of hips from Ohio and Indiana, and received shipments from cooperating persons in Washington, Idaho, California, Maine, New York, Virginia, North Carolina, Mississippi, Texas, Nebraska, Kansas, Iowa, and Alaska, as well as in Saskatchewan, Manitoba, and Ontario. Approximately 14,000 adult specimens of rose hip insects were reared, and additional thousands of larvae of the phytoph- agous species have been dissected to obtain the endoparasitic larvae, of which numerous specimens are preserved in alcohol. Altogether, about 45 species of animals—insects excepting an acarid mite and an annelid worm—were found to be associated with rose hips. They are divisible into three groups according to the nature of their bionomic relations to the hips; namely, incidental, facultative and ob- ligatory. The incidental forms, being paurometabolous and physically not adapted for life within the hip, occasionally visit the hips to feed, or to attach or insert their eggs, but also utilize other similar hosts to perform these functions. About 15 species fall into this category, and are not treated in the present monograph. The facultative and obligate groups comprise only holometabolous in- sects, and their bionomics constitutes most of the subject matter of this treatise. Their more or less legless, or adaptive, larvae reside, feed, and develop out of sight within some part of the rose hip. Present know]- INTRODUCTION 3 edge of their host relations shows that certain of these Holometabola confine themselves to rose hips, developing in no other plants, or parts of plants. Such obligatory species are Rhagoletis, Rhynchites and Mega- stigmus; Opius and Halticoptera, endoparasites of Rhagoletis; and Luch- atema, an internal parasite of Rhynchites. By contrast, it is well known that some other Holometabola also pass their larval life in plants other than Rosa. Such facultative members of the rose hip complex are the cherry fruitworm, Cydia packardi, which also inhabits apple, cherry and blueberry, and the parasites Glypta, Scambus and Euderus on the cherry fruitworm, and Bracon, whose host list in- cludes, beside Rhynchites, many coleopterous and lepidopterous phyto- phags of similar habit in plants other than rose. Besides suggesting the nature of the subject matter contained herein, the main title is intended to imply my conviction that all bionomic ma- terial regarding the species is properly included whether it relates to rose hips directly or to alternate host plants and host insects. The bionomic data pertinent to the species of the rose hip community were obtained by four simple methods. First, direct observations in the field yielded records on seasonal occurrences of the adult insects, and information on the activities of mating, oviposition, feeding and move- ment. Second, dissection of samples of immature hips at intervals of 10 days was a means of learning the typical niches of the several phy- tophagous insect larvae and their ectoparasites, their developmental pattern in the course of the season, and variations in populations with reference to year, locality and species of rose or insect host. In turn, dissections of the phytophags yielded similar information as to their endoparasites. Third, in September and October quantities of more or less ripe hips, gathered in various northern states, Alaska, and Canada, were stored, mostly at Urbana, Illinois. Cloth covers placed on the glass “hip jars” prevented both excessive drying of the hips and escape of the emerged larvae and adults of the ectoparasites. The larvae were removed, usually on the day of their exit from the hips, and either preserved for subsequent description and identification, dissected in order to trace the later developmental aspects of the endoparasites, or transferred to rearing cages designed to yield the pupal and adult forms. Fourth, four-inch flower pots, filled with soil and surmounted with old- fashioned lantern globes closed above with fine-meshed cloth, constituted cages suitable for rearing the rose hip fly and the rose curculio and sev- eral species of endoparasites. The curculionid larvae readily enter the soil, where they winter; puparia of the Rhagoletis winter fairly well when placed between two discs of cloth on the surface of the soil. The pots, minus the chimneys, were set out-of-doors, countersunk in the ground, 4 INSECTS IN ROSE HIPS under natural conditions from November to May, then taken into an open shelter and topped with the chimneys to await the emergence of the adult beetles, flies or hymenopterous parasites. While some cherry fruitworms survived the winter among hips in jars, better results could probably have been secured had pithy plant stems been provided for the larvae to enter for preparation of their hibernac- ula. Both species of the rose seed chalcids winter as matured larvae in the achenes, and the adults may be reared readily in the spring from hips stored in glass jars. Mould growing profusely on the decomposing moist hips appeared not to affect adversely the development of the chal- cids, whereas extreme hardening through desiccation seems to have blocked the emergence of some adults. The ectoparasites of Cydia and Rhynchites pupate in the hips. Accord- ingly, the adults of these emerged in the hip jars. Since some individuals or species emerge in the spring, the hips need to be kept under observa- tion during the winter. Valuable host-parasite data were obtained by isolating in vials the ectoparasitized hosts in the hips until the parasitic larva or pupa became adult. Acknowledgments In a comprehensive project such as this, the investigator is dependent, to a degree, on other persons for funds, performance of routine duties, acquisitions of live material from outlying areas and for the technical services of specialists in the identification of plants and insects. I am pleased to acknowledge all such aids with appreciative thanks, even though space does not permit the naming of all individuals concerned. The research board of the Graduate College of the University of IIli- nois provided funds for preparation of drawings, bibliographical work, care of rearing cages, for pinning, labeling, and arranging many adult specimens reared, and typing a part of the manuscript. The drawings were made by Donald Van Horn, then a graduate student in zoology. Quantities of rose hips were gathered at the places and by the persons named herewith, and mailed to me for processing. Alaska: S. E. Lienk, C. O. Esselbaugh, Katherine Sommerman. Saskatoon, Saskatchewan: Mrs. Arni Arnason. Manitoba: R. D. Bird. Ottawa, Ontario: T. N. Free- man. California: Albany and Berkeley, R. F. Smith and J. W. McSwain; Cedarville, Sam Rea. Washington: Spokane, Pullman, and Wawawai; and Moscow, Idaho, C. O. Esselbaugh. Utah: Mapleton, C. E. Hopla; Mt. Timpanogos, Utah County, C. L. Hayward. Kansas: Manhattan, R. H. Painter. Nebraska: Crete, W. F. Rapp. Oklahoma: Norman, H. P. Brown. Missouri: Ironton, David Lauck. Minnesota: Cook County, Mr. and Mrs. Edgar Love, Katherine Love; Ely, Mr. and Mrs. Jalmer Lake. North Carolina: Ash, Guilford and Transylvania Counties, and Raleigh, W. B. INTRODUCTION 5 Fox. Ohio: East Liberty, A. J. Sharp; West Alexandria, R. T. Bell. New York: Geneva, S. E. Lienk. Maine: Orono, F. H. Lathrop. Virginia: East Falls Church, L. H. Weld. Illinois: Rochelle, S. E. Lienk, John Porter; Seward and Lanark, J. Porter; Alton, David Lauck; West F rankfort, Wayne Carlisle; San Jose, D. M. Tuttle; Rising, C. E. Wester; Philo, Mr. and Mrs. F. H. White. I received 12 lots of wild rose hips through the cooperation of Wilbur A. Rush, Chief, Division of Lands and Waters, Iowa State Conservation Commission. These were secured by twelve members of his field staff in their respective areas of supervision. Also, four samples came to me by courtesy of the Colorado Forestry and Horticulture Association, Denver. The hips involved in the course of this investigation represent approxi- mately 30 species or varieties of roses. Taxonomic interpretation differs somewhat among students of the genus, which botanists generally recog- nize as a difficult one to determine. For their painstaking work in identi- fying most of my rose materials, I am indebted to Dr. John W. Moore, then botanist at the University of Minnesota, and Dr. G. Neville Jones, systematic botanist at the University of Illinois. Some materials were named by the late Dr. L. R. Tehon, Illinois State Natural History Survey; Dr. E. L. Stover, Eastern Illinois State College, Charleston; the late Messrs. James S. Hutchinson and S. W. Hall, University of Illinois. The following botanists identified the roses growing in the areas of their resi- dence: Dr. B. F. Harrison, Brigham Young University, Provo, Utah; Dr. R. C. Russell, University of Saskatchewan, Saskatoon; Prof. Fay High- land, University of Maine, Orono; Prof. W. B. Fox, North Carolina State College, Raleigh; Dr. C. F. Gates, Kansas State Agricultural College, Manhattan. The hips received from Fairbanks, Alaska, had been de- termined by Father Le Page, according to Dr. S. E. Lienk, who sent them to me. I am indebted also to Dr. G. Neville Jones, University of Illinois, and Dr. Reino O. Alava, rosarian of the Missouri Botanical Gardens, St. Louis, for checking the scientific names of all plants mentioned herein, with reference to the accepted status of specific and subspecific nomials, specific synonyms and names of authorities insofar as this was possible within a genus so involved as Rosa. Dr. Leland Shanor, mycologist, then at the University of Illinois, identified a fungus parasitic on larvae of the rose curculio. The insects and other invertebrates observed in the course of this study were identified by the following staff specialists who are, or were, at the United States National Museum, Washington, D.C.: Diptera, Alan Stone and C. W. Sabrosky; Chalcidoidea, A. B. Gahan and B. D. Burks; Braconidae, C. F. W. Muesebeck; Ichneumonidae, H. K. Townes and 6 INSECTS IN ROSE HIPS Luella M. Walkley; Lepidoptera, H. W. Capps, J. F. Gates Clarke and the late Carl Heinrich; larvae of Cleridae, O. L. Cartwright, and Acaridae by E. W. Baker. L. J. Stannard, Illinois State Natural History Survey, determined some lots of mites, and L. J. Thomas, University of Illinois, identified the an- nelid worm, Chaetogaster sp., which I have from Rhynchites. I am indebted to The Ohio State University for the loan of the doc- toral thesis of N. D. Blackburn (1944), and to the author for permission to use some of the original unpublished material from the thesis: also to the University of Wisconsin and Dr. D. A. Dever (1953) for like favors. Special thanks are merited by my wife, Willma, for her interest, pa- tience, and assistance. She willingly permitted me to convert the back porch of our home into an insectary and laboratory in order to bring my night work of several years conveniently to hand. She also mounted thousands of specimens on pins. THE RosrE Hip F ies, TRYPETIDAE, DIPTERA As interpreted today, at least three species of trypetid flies are known to develop as larvae in the fleshy hypanthium of rose hips. They are Carpomyia Schineri (Loew), Rhagoletis alternata (Fallén) and Rhago- letis basiola (Osten Sacken). Carpomyia Schineri (Loew) Loew (1862) described this species under the name Oedaspis Schineri. It is cited in the literature also with the generic names Trypeta and Orel- lia. According to Hendel (1927) and Seguy (1934), it occurs throughout central and eastern Europe. The species has been of minor importance as a pest, and there is relatively little information at hand about its mode of life. Rodendorf (1938) includes the species in a key to a number of Trypetidae that are most often found in fruits at quarantine stations in Russia. It has three relatives that have attracted some notice owing to their injury to fruits—C. incompleta Beck (Rodriguez, 1926) and C. vesuviana A. Costa (Khare, 1923) for their attack on berries of Zizyphus, and C. caucasica Zaitz (Prinz, 1919) as a pest in melons. These I regard as significant here because their habits are perhaps indicative of those of C. Schineri. Eppelsheim (1871) apparently was first to discover the host plant of C. Schineri. He reared adults from larvae found in the hips of Rosa spin- Osissima at Griinstadt near Speyer and Heidelberg in the Rheinland— Pfalz, Germany. One adult, reared in 1872, was sent to Jos. Mik (1898) THE ROSE HIP FLIES, TRYPETIDAE, DIPTERA ii and a fine series to Kaltenbach (1874). Perris (1876) reported finding a “very large” dipterous larva in the overripe hips of Rosa canina. He gathered a quantity of hips and reared many such larvae from them, later obtaining adults, which he identified as C. Schineri with the help of M. Gobert and the literature of Schiner. These specimens were ob- tained around Mont-de-Marsan, north of the river Adour in the prov- ince of Gascony, in southwestern France. Von Roder (1891) added an- other species, Rosa gallica damascena, to the host list, based on adult insects that he received from Dr. G. Dieck in Zoschen, near Merseburg, who reared them from the fruits. Eppelsheim observed the mature larvae to leave the hips in August. Those reported by Perris emerged after mid-September. These observers agreed that the larva descended to, or into, the soil where they pupari- ated in the fall. The adult flies emerged from the puparia in the follow- ing June (Eppelsheim), and beginning with the last days of June, to July 10, 1875 (Perris). While oviposition seems not to have been ob- served, it may safely be assumed that the fly inserted the eggs into the hypanthium. The larvae ate out winding galleries in the flesh of the hip and did not disturb the stony seeds within it (Eppelsheim). No para- sites were recorded in the literature cited. Rhagoletis alternata (Fallén) Rhag. alternata, known to European entomologists since Fallén de- scribed it in 1820, and Rhag. basiola are very similar in their physical characteristics and apparently also in their modes of life. As a conse- quence, some doubt remains concerning their taxonomic status. For the present purpose, I am adopting the view of Alan Stone (1951) that it seems better for now to consider them as distinct species. In Germany, Rhag. alternata is called die Hagebuttenfliege; in Sweden, Nyponflugan. Distribution and hosts. Rhag. alternata is widespread in Europe, and has been reported from the hips of many species of Rosa. Following are the records known to me: Rosa canina, middle and northern Europe (Hendel, 1927); Surrey County, England (Niblett, 1936), Scotland (J. Hardy, 1873); R. villosa, south of the Alps (Hendel, 1927), and Scot- land (J. Hardy, 1873); R. rugosa, Kiev, Russia (Belski, 1924); R. hugo- nis, England (Collin, 1947); R. spinosissima, Griinstadt, Germany (Ep- pelsheim, 1871); Rosa spp. not determined, Austria (Schiner, 1864), Koblenz, Germany (Klein-Krautheim, 1937), England (Pearce, 1928), Ireland (Wilson, 1931), and Sweden (Anon., 1941). Wilson (1935) obtained this fly from the following 27 forms of Rosa in the gardens of the Royal Horticultural Society at Wisley, England: Rosa alpina var. pyrenaica; blanda; calocarpa; canina, vars.; Chaberti; cinnamomea; feneliensis; glauca; gymnocarpa; lucida; macrantha; mollis- § INSECTS IN ROSE HIPS sima; moschata, var. alba; Moyesii; multibracteata; nutkana; oxyodon; parviflora; pomifera; eglanteria; rubrifolia; rugosa; Selongiana; sericea; swenginzowii; virginica; woodsii. Without knowledge of the food plant, Rhag. alternata has been re- corded also from the Piedmont, Italy (Rondani, 1870); France (Gobert, 1887); Boitsfort, Belgium (Guilliaume, 1904); Kiel, Germany (Boie, 1847), Swinemiinde and Berlin (Zacher, 1919), Berlin-Buch. (U.S. National Museum); and Maribo, Denmark (U.S. National Museum). Fallén (1820), who named this fly, Rhag. alternata, gave its “habitat” as Esperod Scaniae, Sweden. The United States National Museum also has specimens from hips sent from Sweden and intercepted at Cleve- land and New York in 1947 and 1948. In addition to the above records of Rhag. alternata from Rosa spp., the literature contains three reports of its occurrence in other plants. However, the latter three are now known to be in error. These cases are reviewed briefly. Von Roser (1834, 1840) stated he had Rhag. alternata from the fruits of Lonicera xylosteum L. Subsequently Mik (1898), Loew (1862) and Klein-Krautheim (1937) showed that von Roser actually had two species of Trypetidae from these fruits—Rhag. cerasi L. and Acidia speciosa Lw. —and that he had mistaken A. speciosa for Rhag. alternata. Fletcher (1883) reported Rhag. alternata as having been obtained by J. B. Hodgkinson from mines in the leaves of Impatiens nolitangere L. Fletcher (1883a) later corrected the error. Hodgkinson had inadvert- ently thrown mined leaves of Impatiens and rose hips together in his greenhouse. Unaware that the hips contained dipterous larvae, he sup- posed the flies to have originated in the leaf mines. A third error in host records is obviously due to confusion on the part of the abstracter of the article by Zacher (1916) in the Review of Ap- plied Entomology, Series A, 8(1920):288. Here it is stated that Zacher found the trypetid “Zonosema alternata (Fall.)” infesting barley. Scru- tiny of the original article showed that the fly attacking barley was a chloropid, Lasiosina cinctipes Meig., and that Zacher reported Rhag. alternata from rose hips. 3 Food of the larva. That the larvae inhabit rose hips, and, when mature, emerge from them, has been long known in Europe. This has sometimes led to the incorrect assumption that the larvae inhabit the seed cavity and feed on the “seeds,” i.e., achenes, thereby injuring them. The occur- rence of puparia among rose seeds, reported by Wilson (1931), was probably due to the artificial conditions in which the inhabited hips were stored, and would not be found in nature. However, the belief that the larvae consume or injure the achenes has persisted until recent years. In the abstract of an article by Reichert (1935), the Review of THE ROSE HIP FLIES, TRYPETIDAE, DIPTERA 9 Applied Entomology, A, 23:506, represents him as stating that the larva injures the seeds in the hip. Klein-Krautheim (1937), in reviewing some published statements on this point, noted that Mik (1898) assumed that the seeds were injured, whereas such attack was disclaimed by Lucet (1898). In his own studies, Klein-Krautheim found that the larvae do not feed on the seeds. If the food habits of Rhag. alternata are like those of its very close relative, Rhag. basiola, the larvae limit both their feed- ing and residence to the hypanthium. Damage to rose hips. As the feeding larvae mine the hypanthium, the normally green or red hip becomes discolored with brown to black strips or blotches, and as more of the tissue is consumed by the larger larvae, the hypanthium collapses and wrinkles. Zacher (1919) cited destruction of the entire crops of hips at Swinemiinde in 1915 and previous years, and at the Biologisches Reichsanstalt in Berlin in 1918. In the latter case, all that remained on the rose plants in August were the hip petioles and tattered scanty remnants of the fruit flesh. Remnants of the larger hips at the tops of the plants, including the achenes, had fallen to the ground. Presence of the larvae in hips still hanging on the plants was indicated by the dull red patches over the feeding burrows. Vitamin C content of rose hips. In recent decades, and particularly during World War II, the hypanthium of the rose hip was found to be an important source of ascorbic acid or vitamin C, valuable in human nutrition. Consequently, Rhag. alternata came to be regarded as a pest in some Eurasian countries. An anonymous article, entitled “Nyponflu- gan” (1941), reported interest in vitamins of rose hips in Sweden, and briefly described the life cycle of this enemy fly. Belski (1924) compiled a list of “pests” of medicinal plants occurring in the vicinity of Kiev, Russia, and included Rhag. alternata as one of the enemies of rose hips, which he regarded as medicinal because of their vitamin C content. Through personal correspondence with a resident of Canada, I learned that the Canadian government was urging its citizens to utilize rose hips as an available source of ascorbic acid. Subsequently, Melville and Pyke (1947) investigated the relation of specific variability and the environ- ment on the content of vitamin C in English rose hips. Wilson (1935) had found that this fly develops in 27 species and varieties of roses in England. Hence, the destructive larva has gained economic significance in that country. The extensive damage reported by Zacher (1919) indi- cated that this fly sometimes had a similarly detrimental effect on the yield of rose hips as this source of the vitamin in Germany. Life cycle and habits. Bouché’s report in 1834 apparently derived en- tirely from his own observations. In Germany, during the summer he found the larvae eating out tunnels in the “envelope” or hypanthium of 10 INSECTS IN ROSE HIPS Rosa villosa, thereby destroying its “fruits.” Late in August, the larvae descended into the ground, where they pupariated, and in June of the next year, the adults emerged from the ground. Thus, Bouché con- cluded that Rhag. alternata can, and does, complete only one cycle or generation in a year. Some of these facts had also been reported by Eppelsheim (1871), J. Hardy (1873), and Kaltenbach (1874), and further observations were made by Zacher (1919), Wilson (1935), Niblett (1936), and Klein- Krautheim (1937). Adult flies remained alive 5 to 16 days when Klein- Krautheim supplied them with moistened sugar. He observed that copu- lation in captivity sometimes continued uninterrupted for at least 2 hours. Search of the literature has revealed no statement regarding ovi- position, but it may be assumed that Rhag. alternata inserts the eggs singly into the hypanthium, as its close relative, Rhag. basiola, is known to do. The early instars apparently have not been described. Collections by various investigators from mid-August to October have shown that the larvae of Rhag. alternata mature and emerge from the hips from August (Bouché, 1834; Eppelsheim, 1871) to November (J. Hardy, 1873), breaking through the skin of the hips to escape. J. Hardy (1873) was obviously in error in stating that the hips show the openings by which the maggots gain access; and also where he added that the larva appears not to confine itself to a single fruit, and readily enters a fresh one when presented to it. He possibly had reference to the entrance of matured emerged larvae into hips stored in captivity, where the normal soil medium for pupariation was lacking. Pupariation has been observed by writers quoted above to take place from a few hours to a few days after the larvae issue from the hips. J. Hardy (1873) commented that the “puparia were mostly cemented to the object, on which they lay, by a moisture exuded from the vent.” Fraenkel and Brooks (1953) describe the same phenomenon in other Diptera, but have shown that the cement issues from the mouth. In describing Rhag. alternata, writers agree that the larvae normally pupariate on or in the soil, and that the puparia constitute the over- wintering stage. But whether the insect is a prepupa or a pupa in this season seems not to have been ascertained. The adults reportedly issue the following year, chiefly in June. Additional data on habits and development are given by Klein-Kraut- heim (1937) and Niblett (1936). At Giessen, Germany on October 7, 1933, Klein-Krautheim (1937) received a quantity of the hips of Rosa canina from Koblenz. Larvae emerged on and after October 8, but some had issued previously. Puparia kept in a cool room yielded males and fe- males, April 4 to June 9, 1934. At Surrey, England, Niblett (1936) ob- tained a lot of the hips of R. canina on October 13, 1934. Larvae began THE ROSE HIP FLIES, TRYPETIDAE, DIPTERA 11 emerging the next day, and adults issued from the resulting puparia June 5-21, 1935. A second lot of puparia observed by Niblett (1936) originated in the hips of Rosa hugonis. They yielded flies during the first week of the next June. Klein-Krautheim (1937) noted that when their cage stood in sun- shine the flies ran and flew about in a lively manner. When walking, they made rowing movements with the wings such as are characteristic of the Trypetidae and related flies. Klein-Krautheim presented a hypo- thetical chronology of the life stages, based on his observations on Rhag. alternata and data from the literature on this and other species of Rhagoletis. This hypothesis holds that Rhag. alternata passes its larval life in the hips of roses, emerges to pupariate in the fall, and winters as a puparium, thus completing one life cycle in a year. No reports have been found in the literature concerning parasites on any stage of Rhag. alternata. Rhagoletis basiola (Osten Sacken ) This is the fruit fly found in the hips of roses in North America. Ac- cording to Stone (1951) it is known in the literature under several generic and specific names. Distribution. My records from the literature (Osten Sacken, 1877; Doane, 1899; Gibson, 1915; Robertson, 1923; Crane, 1932; V. T. Phillips, 1946), the U.S. National Museum, and my own rearings show that Rhag. basiola occurs from near the 35th line of latitude in the United States to near the Arctic Circle in Alaska, and from the Atlantic to the Pacific oceans. Its distribution appears, therefore, to be coextensive with that of the genus Rosa as represented by its wild-growing species. It is doubtlessly true that the acquisitions of additional records from the col- lections of other scientific institutions would serve to intensify but not materially extend the limits of its known range. All the southern proy- inces of Canada are represented, but I have no records for 13 states that lie wholly north of latitude 35°. More specifically, the extreme northern limit of its occurrence is repre- sented by flies reared from hips sent to me by S. E. Lienk and C. O. Esselbaugh from 12 localities in northwestern Canada and Alaska along the Alcan, Richardson, and Steese highways. Miller House, located about 66 miles south of the Arctic Circle, is the presently known northernmost point of its occurrence. On the other hand, San Luis Obispo, California, Las Vegas, New Mexico, and North Carolina are the southernmost points for which I have records. A challenging distributional problem is posed by the lack of reports for southern United States. If Rhag. basiola actu- ally wanes southward, what may be the factors that cause it to do so? Hosts. Records from the literature cited above, from the U.S. National 14 INSECTS IN ROSE HIPS Museum, and, for the most part, from my own rearings give a total of 19 species of Rosa now known to be hosts of Rhag. basiola. They are listed here with the geographical localities where the fly has been ob- tained from them: Rosa acicularis, northwest Canada and Alaska; R. acicularis, blanda and Macounii, northeastern counties of Minnesota, north of Lake Superior; R. blanda, carolina and arkansana, Wisconsin; R. Macounii, woodsii, nutkana, alcea and pyrifera, British Columbia, Saskatchewan and Manitoba; R. Macounii, Nebraska; R. woodsii, Utah; R. blanda, Ottawa, Ontario; R. spaldingii, nutkana, and ultramontana, northwestern United States; R. rugosa, Rhode Island and New York; R. nitida and virginiana, Orono, Maine; R. eglanteria, North Carolina; R. canina, palustris, gallica, setifera and carolina, Ohio, Indiana and Illinois. I made no systematic search in nature for other plants Rhag. basiola may inhabit. Had the fly any other hosts except Rosa, some of these would almost certainly have been discovered and reported in the course of earlier investigations on insect pests of rosaceous fruits. Adult (Fig. 1). Males, 3.8 to 5.7 mm., usually 5.2 to 5.5 mm.; females, 4.2 to 6.2 mm., mostly 5.0 to 6.0 mm. long. Dorsum of body largely dull orange, rest of body and the legs varying in shades of yellow. Postscutel- lum with a pair of black marks sometimes extending to metanotum. Body generally smooth, shiny; hairs and bristles moderately dense. Wings (Fig. 3) with a somewhat variable pattern of brownish black areas and bands, but “apical dark spot . . . usually joined to the subapical transverse band or narrowly separated” (Stone, 1951, Fig. 3). In the female, ex- posed apical abdominal segment trapezoidal, more rigid than preced- ing segment, and, in life, darker, brownish. Ovipositor consisting of three tapering terminal segments telescoped into abdomen when not in use, the distal one hard, pointed for penetration of rose hips. In male, genital segment subspherical, bearing a short, soft, pale, median capitate process and a pair of flat, tapering, firm, diverging orangish brown claspers usu- ally held directed forward in contact with venter of abdomen. Egg (Fig. 2). As it occurs naturally in the hypanthium of the rose hip, the egg consists of two structurally distinct parts, the egg proper—com- posed of chorion, embryonic membrane and yolk or embryo—and a cap borne on the caudal end of the egg. The egg as a unit, including the cap, varies from 1.0 to 1.25 mm. in length, but usually is 1.2 mm.; at its widest point it is about one-third as broad as it is long, and consider- ably broader than deep. In form, the egg proper is elongate-oval length- wise, oval also in its transverse diameter, moderately rounded at both ends but more sharply so at the apex, subdepressed above and bellied or moderately convex beneath. When the egg is newly laid, its smooth, THE ROSE HIP FLIES, TRYPETIDAE, DIPTERA 13 semicoriaceous chorion is milky to pale yellowish white, but becomes medium brown with time. As the embryo advances toward maturity, the chorion becomes readily separable from the underlying thin white mem- brane, hence may remain as a lining of the egg pit when older eggs are dissected out of the hypanthium. This fact may lead to the false impres- sion that the outer skin or layer is not a part of the egg, but a product of the injured hypanthium. | The caudal cap, or “brush,” reminiscent of the cupule around the base of acorns, is fairly rigid, brown, with numerous small pointed bristle- like processes and a larger, central knob or tassel-like process on its sub- flattened outer surface; its ental side is concave, and cuplike, embrac- ing the round basal end of the chorion, but extending farther down over the lateral, than over the broader top and bottom, sides. The cap is perhaps an ornamental, not an essential functional structure, and probably develops and becomes a part of the egg when the latter is still contained in the ovarian tube. Larva. The following key embodies the principal structural features that serve to distinguish the three larval instars of Rhagoletis basiola as they occur in rose hips from Illinois to northeastern Minnesota. 1. Larvae usually 2.0 mm. or less in length; mouth hooks with a prom- inent medioventral tooth but with a dorsobasal lobe (Fig. 11); hypo- stomium slender, unarmed (Fig. lla); no prothoracic respiratory “fans,” but two obscure U-shaped to oval caudal respiratory slits on each pos- mera Chie ear eb, 20) 2. Rs OP eh he BI First instar. 1—Larvae usually 2.0 mm. or more in length; mouth hooks with a narrow or obscure preapical tooth (Figs. 12, 13); hypostomium shorter, with prominent stout ventral processes; food channels, i.e., lattice-like stomal sieves (Fig. 10), subhypostomia and parastomia present (Figs. 12, 13); also prothoracic respiratory “fans,” and three spiracular slits Gimeden posterior stigmatic platé (Figs. 5,°6,'7, 9)... 2.0. Sete ee oe 2. Larvae mostly 2.0 to 4.0 mm. long; mouth hooks with a slender, narrow preapical tooth (Fig. 10); slits of posterior stigmatic plates oval, about twice longer than their maximum width (Fig. 6) ...Second instar. 2—Larvae mostly 4.0 to 7.0 or 8.0 mm. long, found either in or emerged from hips; mouth hooks with a broad, short, rounded preapical tooth (Fig. 13); slits of posterior stigmatic plate elongate, slender and eetoro times longer than wide (Fig. 7)... 20.00.0400 Third instar. Puparium. Newly formed puparia are dark ocherous, becoming pale stramineous in a few days as the cuticle dries. Thoroughly wetting the puparium restores the dark color. Form oval, somewhat depressed; length varies from 3.0 to 6.45 mm., the maximum width from 1.0 to 2.85 mm. The length and diameter in a large sample of puparia varied 14 INSECTS IN ROSE HIPS ratiowise, from 1.83 to 1.0 in the shortest small individual, to 3.59 times longer than wide in the slenderest large one (Balduf, 1958). Expansion of the wings. On the newly emerged flies, the inflated ptil- inum is sometimes temporarily visible. The wings are then mere small sacks that reach caudad only to the base of the abdomen. They are soft and crumpled, and their lumina still contain more or less bodily fluid that presumably represents blood and constitutes the agent which effects expansion of the wings. The process of broadening and thinning of the wing sacks is first visible at the base, whence it spreads distad. In the intermediate stage of expansion, the wings are long and narrow, their front and hind edges parallel, and they lie well apart from one another and horizontal to the top of the body. In their final, fully developed state, the apices are broader and carried elevated at an acute angle to the length of the body, also in a transversely oblique position—the anal area adjacent to the abdomen, the costa at a higher level. During the process of wing expansion, the flies cling to a vertical sur- face, head upward. The rate of expansion was observed to be slower on cool rainy days, as compared with bright warm spring days. Ordi- narily, the fly engages in its characteristic movements within a few hours of its emergence from the puparium. Like other picture-winged flies, Rhag. basiola may stride grandiosely about while pirouetting the ele- vated colorful wings. Time of emergence. In the course of attending the rearing cages in the earlier years of this investigation, I casually observed that the adults of Rhag. basiola seemed to emerge from their puparia only in the hours between dawn and noon of the same day. Scarcely prepared to believe that they issue only in one restricted period, day by day, I instituted a plan of checks in which the first inspection of the rearing cages was made at about 6:00 a.m., and, in many cases, at or before dawn, and thereafter as many times in the day as other duties permitted. The num- ber of flies found in each cage was recorded at each inspection. The crop of flies in each cage was removed between 4:00 p.m. and midnight of each day. Extending this program of checks through the seasons of emergence of the years 1946, 1947 and 1948, it was established that Rhag. basiola indeed has a period of emergence each day that begins approximately at sunrise and terminates, with few exceptions, by 10:00 A.M., standard time, of the same day. This emergence pattern holds true for Eaglesnest, Minnesota as well as Urbana, Illinois; hence it is not merely a phenomenon induced by some peculiarities of local environ- ment. Data on these observations are presented in Table 1, and are based on 12 complete daily records taken in June 1946, 1947 and 1948, when THE ROSE HIP FLIES, TRYPETIDAE, DIPTERA 15 emergence from the overwintered puparia was at its numerical peak. The percental values show that emergence occurs largely between sun- rise and noon, and that the time of inception and completion of the daywise emergence is influenced by temperatures, being delayed when the average readings Fahrenheit for the 10 hours preceding 6:00 a.m. varied between 51 and 59°, and occurred earlier and earlier in the morning as these average temperatures increased to 70 to 74°, in shade. Excepting one day of the 12, the morning hours averaged warmer than the predawn hours. TaBLeE 1. PERiopIC EmeRGENCE OF Rhagoletis basiola (Selected days in June, 1946, 1947, 1948) Percentage of flies emerged Average temperatures in 10 Prior 8:00 10:00 hours Number to to to After preceding of flies 8:00 10:00 12:00 12:00 6:00 A.M. issued A.M. A.M. noon noon SLs ee 38 0 24 32 44 60-64 43 0 86 14 0 65-69 141 28 69 3 0 0 70-74 63 94 6 0 This periodicity apparently may be explained by the probability that the process of pupal development to the emergence-ready adult state is alternately retarded by the relatively cool temperatures of night and accelerated by the warmer daytime temperatures during May and June. Let us choose several successive days out of the season in order to illustrate the nature of the alternation. Each daily yield represents a fraction of the total sample. On the first day, we find a number of pupae that have matured to the state of emergence, and these issue as flies during the daylight hours of the ensuing morning. In the total sample of pupae there is also a number of individuals so immature that they cannot quite develop to the state of emergence as adults in the remain- ing warm afternoon hours of the day. The further development of this number is retarded by the falling temperatures that usually characterize the ensuing night. Since night temperatures mostly reach the lowest point just before dawn, accelerated progress of this fraction toward the state of emergence must await the warmer sunlight hours of the new day. Thus accumulates a crop of subadult individuals from noon to dawn, due to the submaturity of the pupae and the retarding influence of cool night temperatures. This crop will emerge on the morning of the 16 INSECTS IN ROSE HIPS second day. The individuals that constitute the daily fractions or crops of flies, have attained the emergence-ready state to various degrees, as a result of which their emergence is not simultaneous, but spread over the several hours of the forenoon. To demonstrate a theory that irregular temperature is one of the fac- tors causing this restriction of emergence to the daylight hours of the morning, three lots of puparia were subjected to a constant temperature of 80° F. and a water-saturated atmosphere, beginning June 6, 1947. The sample of puparia reared from Rosa acicularis, growing at the mouth of the Cascade river on Lake Superior, yielded 56 flies between June 8 and 24; those from R. arkansana, Madison, Wisconsin, yielded 55 flies, June 12 to 23, and a third lot from R. carolina, Kickapoo Park, Illinois, 84 flies, June 12 to 28. The constant temperature disrupted the natural pattern of emergence somewhat, in that flies began, on some days, to issue before 6:00 a.M., and others appeared after 3:00 p.m. However, the majority of them still emerged between the hours of 6:00 and 10:00 a.m. Had the puparia been held under the above constant conditions for a much longer period, e.g., from mid-April, preceding the attainment of the emergence-ready state, the appearance of the flies might have been continuous, rather than periodic. Dunning (1956) describes a diurnal rhythm in the emergence of Pegomyia betae (Curtis), Anthomyidae, Diptera, from the puparium. In its general aspects, at least, this rhythm resembles the periodic emergence of Rhag. basiola described above. From experiments designed to identify the initiators of this rhythm in P. betae, Dunning suggests the phenomenon may be “induced by the effect of temperature or light fluctuation at some time before the late pupal stage, or that it might be inborn in the species, and that it is further regulated by temperature variation at the time of emergence. Further work will be necessary be- fore the suggested explanation can be shown to be valid or otherwise.” Behavior. The adults are inordinately secretive under natural condi- tions, hence are seen but infrequently in association with their obligatory host plants. Their presence is usually first indicated by the dark speckles made on the rose hips by the ovipositor. They remain out of sight on a clear day, and are then resting, or engaged in limited walking or flight, in the shaded recesses of the rose bushes or adjacent vegetative cover. The flies obviously dislike direct sunlight, and because of their prefer- ence for the obscurity of deep shade, an observer might easily assume that the flies are rare and that only a few individuals produce all the eggs found in local situations. In my observations of this species, they came out to plain view in the cool shade of evening and during the daytime only when the sky was THE ROSE HIP FLIES, TRYPETIDAE, DIPTERA LL ¢ heavily overcast. Even when the males, which were visible more often than the females, were abroad as if alert to find mates, they mostly con- fined themselves to the undersides of leaves and also the hips. Possibly because males are less sensitive to light, or are impelled by sex urge, they appeared somewhat less secretive than females. The females seem to be less aware of, or responsive to, light when preoccupied with ovi- position, yet this function is discontinued when the sun breaks through the clouds for a few seconds. Feeding. The flies which had been caged for observation of mating and oviposition fed readily on water solutions of sugar or honey with or without supplements of baker’s yeast. They also drank the water droplets which condensed on the glass chimneys. Preening. The captive flies also spent much time in brushing them- selves with their legs. In doing so, the front legs briskly stroked the sometimes flickering proboscis and the head, including the eyes, whereas the hind legs preened the abdomen, the wings, the ovipisitor after it was withdrawn from the hip following oviposition, and were also used to rub each other. Copulation. Mating pairs are seen but rarely in nature. Within the confines of the cages, it was observed that the advances of the usually aggressive male are sometimes resented, as evidenced by the vigorous sidewise movements of the female. Once the union is effected, the pair becomes comparatively imperturbable although flight is possible for short distances during copulation. The accepted male stands quietly astride his mate, his fore tarsi resting near the base of the female wings, while the other four feet contact the wings, abdomen, or the support on which the pair stands. In the copulatory posture, the apex of the female ab- domen appeared linked to the venter of the male terminal segment. While so united, the male sometimes elevates his abdomen, and thereby draws the invaginated terminal female segments, that form the ovi- positor, out to their full length. This seems to mean that the very apex of the ovipositor is attached to the male. The pair of pointed chitinous claspers are then seen to have their flat side applied against the antero- lateral faces of the everted ovipositor. In a few instances, pairs were dis- covered in copulo on the first day of adult life where no food was pro- vided. Seven other pairs were in their second to sixth days when first seen united, but these were not certainly the first matings. The females of the seven pairs lived 21 to 33 days when fed as stated above, the males from 17 to 28 days, during the period, July 16 to Au- gust 21. Copulation by the seven pairs was seen to be repeated from four to 10 times in as many days. The unions generally occurred on successive days, but in some cases, at intervals of one to several days. The interrup- 18 : INSECTS IN ROSE HIPS tions appeared to be associated with temperatures in the sixties. The mat- ing activities of all the pairs was most intense during the first half of the recorded periods of adult life, and particularly so in the first two to three days. In a specially noteworthy instance, a pair continued joined night and day for 33 consecutive hours. It seems certain in this case that the then low temperatures prolonged the union. Copulation is probably not initiated in such temperatures, but seems to persist so long as cool weather prevails following its inception. No mating was observed in the last 3 to 7 days of the life of six pairs. In the case of the longest-lived pairs, whose male died at age 28 days and the female at 33, copulation was not seen in the final 17 days. Oviposition and egg. Data pertinent to the process of placing eggs into rose hips, and the development of the embryo, were obtained by the following means: (1) oviposition was observed both in the labora- tory and the field; (2) approximately 30,000 hips were examined in the field in order to determine the percentages that bore ovipositional punctures, and dates of same; (3) samples of punctured hips, totaling about 3,000, were dissected microscopically to learn the manner of placement of the eggs and the gross development of the embryo. The females that actually laid eggs under observation were about 12 days old. In general, in this procedure the fly settles upon a hip, and bends the apical part of the abdomen downward, sometimes obliquely backward, sometimes at an angle of about 90° with the base. Next the entire body, pivoting on the legs, rocks alternately and vigorously for- ward and backward for some seconds or longer. By this means, the awl- like terminal section of the ovipositor effects an oval puncture in the skin and a pit in the underlying flesh of the hypanthium. When the rocking ceases, the distal two segments of the drill have entered the hypanthium, while the end of the broader, blunt third segment rests against the skin of the hip. Then the abdomen is lengthened and shortened alternately in a pul- sating manner, and exerts backward thrusts for several minutes. In some cases, the proboscis flickered rapidly during the pulsating thrusts. In this phase, the egg probably passes from an oviduct, through the ovi- positor, and into the egg pit. The fly then remains still for a minute or more before she withdraws the ovipositor from the pit. With captive flies, the work of drilling an egg pit and placing an egg into it was ob- served to require 12 to 24 minutes. The duration is probably determined by the hardness of the hypanthium more than any other factor. The egg pit necessarily conforms closely to the length and form of the egg, being approximately 0.3 mm. in diameter and 1.0-+ mm. in depth. Initially, the broadly oval aperture is not visible without careful search and magnification; but in less than a week, the marginal skin tissue of THE ROSE HIP FLIES, TRYPETIDAE, DIPTERA 19 the hypanthium, disrupted by the ovipositor, turns dark brown to black. Then the number of punctures can readily be counted without aid of a microscope. However, brown specks of a corky texture that probably represent a kind of plant disease, may sometimes be mistaken for egg punctures, until scrutinized with a microscope. The outer or caudal end of the inserted egg sometimes protrudes slightly above the adjacent surface of the hypanthium so that the orna- mental brushlike caudal cap remains visible externally; at the other ex- treme, the egg is often distinctly countersunk, as much as one-fifth its own length. The hypanthium sometimes swells convexly over and around the egg, a condition that appears associated with the hard texture of green, growing hips. Rarely a fly makes two pits through a single ex- ternal puncture, the pits then forming an inverted V, but only one of the pits has contained an egg. Numerous dissections of hips (40 per cent of all recorded for R. blanda and 19 per cent for R. acicularis) leave no doubt that many pits are incomplete, and therefore contain no egg. The pits vary in depth, from mere shallow breaks in the skin to full depth, so far as one can judge. Old incomplete pits are filled with a hard conical brown plug formed of desiccated sap that oozes from the hy- panthial tissue. Many careful examinations seem to indicate that the hardness of the hypanthium is the principal cause of incomplete pits, since they are found largely on hips that have grown to full size yet remain green and so flinty hard that small pieces may be chipped off with a scalpel. However, in some cases, absence of eggs from apparently complete pits may signify interruption of the ovipositing fly by natural causes, e.g., leaves of the plant blown against her by the wind. It is also reasonable that the fly may become so tired from boring into hard hy- panthium that she desists before an egg can be inserted. The positions taken by the fly when she implants the eggs may be learned from observing her in action, and also from the angles at which the eggs lie in the dissected hypanthium. The longitudinal axis of the fly more or less closely parallels the long axis of the hip, with the fly’s head directed toward the floral end as she clings to the lower, shaded and comparatively green ventral side of the hip. Rough diagrams made during dissection of the rose hips, to indicate the position of the eggs in the hypanthium, show that most eggs were set at various slants or angles, rather than vertically. The angles ranged from vertical to sub- parallel with the surface at the point of insertion, i.e., 15 to 90°, with the majority lying at intermediate angles of 35 to 85°. Both the sketches and direct observations showed that all but a few eggs are placed into the basal half of the hip and into its green under- side. Of 810 punctures recorded from R. blanda, 92 per cent occurred on the petiolar or basal half, 7 per cent on the transverse or equatorial 20 INSECTS IN ROSE HIPS line, and 1 per cent on the apical half. For R. acicularis, the number on the free or distal half was somewhat larger. Only rarely were eggs in- serted into the lateral faces or the sun-reddened upper sides. This pre- dominance of eggs or incomplete pits on the basal and ventral areas seems to result from two factors—one environmental, the other inherent in the behavioral characteristics of the fly. As to the first, the sepals of the calyces largely lie recurved over the floral or distal part of the growing hip, thus affording a rough and effective support for the fore legs of the fly during oviposition, and, at the same time, they prevent the in- sertion of the ovipositor into this same distal area of the hip. Second, the pronounced negative phototropism of the fly appears to explain the strong preference for the underside of the hip as the place for oviposition. The predilection for the basal half favors the larvae since the hypanthium is thicker, bulkier here, affording them more food and living space. However, this same preference is a handicap because the area ripens and mellows later than the sunned upper surface. This circumstance tends to retard or thwart the feeding and burrowing of the first instar until it reaches the more penetrable, riper parts of the hypanthium. Rhag. basiola never inserts more than one egg into a single pit, al- though two or more deposited eggs per hip are not unusual. In such cases, differences in the age of the embryos show the several eggs are not laid in close succession by a single fly. In no instance was an egg found in the central or achene cavity of the hip. Limitation of the egg to the hy- panthium may be explained by the observations that the majority of the eggs are placed in the thick basal part of the hip wall, and most of them lie more or less obliquely in it. Normally, the capless cephalic end of the egg is first to enter the pit, with the result that the larva, upon hatching, enters directly into the flesh of the hypanthium which consti- tutes its food. Six of the hundreds of eggs found in situ appear to have been inserted caudal end first, for the dark brown mouth hooks of the advanced embryos were visible in the outer end of the pit. Thus, the brown, brushlike cap, which normally embraces the cauda of the egg, lay over the cephalic end. Where such inverted embryos lie consider- ably countersunk in the hypanthium, the hatching larva may possibly bore into the hip directly from the upper end of the egg pit. Oviposition and rose hosts. The number of pits and eggs differed no- ticeably with reference to the hips of R. acicularis and R. blanda investi- gated at Eaglesnest from 1945 to 1950. Of the 6,652 hips of acicularis examined 907 bore 1,094 pits, or an average of 1.21 per hip, whereas 1,239 of 19,305 hips of blanda, including R. Macounii, had 2,084 punc- tures, or an average of 1.69 per hip. However, the percentage of eggs in the pits was higher in acicularis. Eggs had been laid into 81 per cent THE ROSE HIP FLIES, TRYPETIDAE, DIPTERA DA | of the pits in acicularis, whereas only 60 per cent of the pits in blanda contained eggs. The hips of acicularis not only develop and ripen earlier in the over-all ovipositional period, but are generally more penetrable to the ovipositor and the resulting larvae, as compared with the later- ripening, harder hips of blanda. Although the hips of blanda mellow later in the season and the fly then achieves a greater success than earlier, the over-all advantages in acicularis are more favorable to success. The differences in texture and ripening time of the hips of these rose species go far toward explaining the larger population of Rhag. basiola in acicularis. Embryo. As the embryo develops, the chorion changes from the pale yellowish white of the newly laid egg to pale brown, a change in color possibly effected by the stain of fluid released from the cells of the hypanthium ruptured by the ovipositor. Recently laid eggs remain free within the pit walls and are readily removed intact by chipping bits of the contiguous tissue away with a fine scalpel. The chorion later be- comes fused to the wall of the pit, probably by the cell sap that dries around it. As the advanced phase of the embryo is reached, the chorion separates from the inner membrane, and the more or less mature embryo can be worked out with probes, leaving the chorion as a lining in the egg pit. These conditions facilitate gross identification of successive phases of the developing embryo. Three quickly identifiable phases of embryonic growth are recog- nized: early, intermediate, and advanced. The first is indicated by the pale freshness and plump form of the chorion, the granular texture and stringy consistency of the grayish white yolk, that seems to signify an incipient and grossly amorphous embryo. The intermediate phase is characterized by its definite larva-like form, still straight tubular ali- mentary canal and the absence of bodily segments and mouth hooks. The advanced phase is recognizable at a glance by its over-all resem- blance to the first instar of the larva, ie., by the well-developed seg- mentation, the reddish to dark mouth hooks and the convoluted ali- mentary canal clearly visible in the caudal half of the abdomen. Casual observations indicate that the first two phases are of short duration, but the third persists through more than half of the embryonic period. Deficient eggs. A small percentage of the eggs found in rose hips had chorions of normal color, size and form, but were subnormal in that they contained only a fluid having the color and consistency of cow’s milk. I have considered two theories to explain this condition. First, it may result from failure of fertilizaton. But this cannot be the cause, since the normal, grayish white, granular yolk forms in the ovarian tubules independently of the sperm. Second, it may signify a deficiency in 292 INSECTS IN ROSE HIPS nutritive materials such as may characterize old, reproductively ex- hausted flies. Also observed repeatedly in egg pits, in the absence of larval burrows, were dry, collapsed chorions of dull color. These appeared to be empty, or, at least, they contained no macroscopically recognizable solids. This condition suggests that these are deficient eggs from which the milky fluid evaporated in the course of the summer. This belief is strengthened by the fact that collapsed empty chorions were found from late July into September, i-e., in the later part of the reproductive season, when adequate time for desiccation had elapsed. It therefore appears prob- able that the milky state of the egg antedates the dry collapsed condition of the chorion. Larval habits. It may be taken as a fact that the embryo, when de- veloped to the state of hatching, uses its mouth hooks as a mechanism for escape from the enveloping chorion. All but a few of the embryos lie headed toward the inner end of the egg pit; hence, as hatching larvae, they issue directly into the substance of the hypanthium. The rate of progress in burrowing, and also of growth, of the first instar is definitely influenced by the condition of the hypanthium. Many rose hips are hard when immature and green, and the degree of hard- ness in this stage of their development varies from species to species. For example, those of Rosa blanda are then flinty in texture, as compared with the inherently less firm hips of R. acicularis. One of the impressive facts learned in dissection of the hips of blanda was that the first instars of Rhag. basiola suffer a high rate of mortality. Many of their burrows were incipient to short, with thickly encrusted rough brown walls. The dead, blackened shrunken larva was then found at the fresher end of the burrow. This frequently observed effect obviously resulted from the inability of the larva to ingest any of the hypanthial tissue, or enough of it to survive. Later in the season, August to September, the hypan- thium of R. blanda mellowed and ripened, and only a few newly dead first instars were then found. When the hypanthium is sufficiently soft, the larva rasps a simple sub- cylindrical burrow, which gradually but slightly enlarges in diameter as the insect doubles in thickness in the course of growth from hatching to the first molt. As the larva advances through the hypanthium, the walls of the older sectors of the burrow become encrusted, particularly when the hypanthium is unripe and juicy. The crusty layer probably results from solidification of juices that ooze from the cells disrupted by the larval mouth hooks, forming a rough brown surface and reducing the diameter of the burrow. But encrustation is less in extent, or lacking, later in the summer when the hypanthium softens and ripens, particu- larly in R. acicularis. THE ROSE HIP FLIES, TRYPETIDAE, DIPTERA 23 The burrows of the first instars occur at various depths and areas of the hypanthium. In hard unripe hips, they may pass just under the skin of the hypanthium, and, being encrusted and dark, their straight or winding paths may be traced externally. Other burrows extend along the floor of the hypanthium, adjacent to the achene cavity, but do not penetrate it. The majority pass through the fleshy middle level. While 90 per cent and more of the eggs of Rhag. basiola are inserted into the shaded green underside of the basal half of the hip, and most burrows are therefore initiated in this area, the larvae do not limit themselves to it. Many of the first and second instars were found to inhabit the thick basal part of the hypanthium, and the petiole was circled by some of them. Others proceeded toward or to the floral half, where they not infrequently ringed the necklike constriction at the base of the calyx. Again, some larvae invaded the mid or equatorial area of the hip, includ- ing both the green underside and the reddening or ripe sunned surface, while others bored lengthwise in a straight or sinuous line from base to apex. Desipte these variations in location, depth and direction of the bur- rows, I hold that the burrowing is not wholly fortuitous. Assuming the behavior of the larva to be made up of a succession of reflex responses, it seems probable that the textural, chemical and nutritional properties of the hypanthium, the orientation of the hip on the plant and the posi- tion or posture of the eggs in the hypanthium, determine the pattern of the burrows and that the boring behavior may be fundamentally uniform. The maturing larvae, and especially the third instars, due to their greater bulk and feeding capacity, and also the concurrence of soften- ing ripe hypanthium, are not limited to the cylindrical type of burrow. Instead, by increasing the width of the burrows and by creating branches, cavities of greater or lesser size are produced that may be described as blotch mines. The larvae lie in the margins of these mines, in contact with the living tissue and removed from their excrement (which resembles apple butter in appearance and consistency) in the area previously occupied. The mine may involve a quarter or more of the entire hypanthium, depending on the size of the hip and the larva; or several mines, representing a number of larvae in a hip, may involve almost all of the hypanthium. Rarely was any one hip found to harbor simultaneously more than a single third instar but hips, particularly of R. acicularis, were commonly inhabited also by two or three earlier instars. In the latter cases, the slender tunnels of the small individuals may be disrupted, and the larvae themselves incommoded, by competi- tion with the large larva for food and space. The hypanthium overlying the undermined sectors turns dark, and 94 INSECTS IN ROSE HIPS when the ripe mushy tissue is rasped off cleanly, the skin proves to be whitish and transparent, or window-like. The hips so infested also lose their plump convex form as the skin collapses into the mines. The mature larvae of the fly emerge head first by making local, ir- regular breaks in the skin at various parts of the hypanthium above their blotches, from petiole to distal half. The flaps of the skin involved in the exit hole may resume their natural position after the larva has wormed out. The discovery of the holes as evidence of emergence therefore may require close inspection of the inhabited hip. Larval population. The number of Rhag. basiola larvae present in any one time, place or rose host is most conveniently ascertained by collect- ing the mature larvae as they normally issue from hips in late summer and fall. The data pertaining to larval populations reported below were secured by storing more or less ripened hips in fruit jars, and removing the emerged larvae or puparia, usually at daily intervals. In the years 1944 to 1950, approximately 70,000 hips were under observation, and 9,752 mature larvae or puparia of this species were obtained. Assuming one individual per hip, 14 per cent of the hips contained a larva. The error due to the occasional presence of two, or rarely three, larvae per hip is probably more than counterbalanced by another error due, in some samples, to the emergence of the earliest fraction of the larvae while the hips still hung on the bushes in nature. Sixty-five separate samples compose the over-all lot of hips observed. The component lots vary from 20 to 5,835 hips representing 10 species of roses: R. acicularis, blanda, Macounii, arkansana, palustris, canina, virginiana, eglanteria, and woodsii. The tentatively tabulated data derived from the above rearings showed first a striking range in number of larvae per unit number of hips. The most impressive aspect of this percentile variation is the somewhat steady decline in the ratio of larvae to hips from the northern to the southern limits of the range. A series of examples representing the successively lower latitudes shows this feature clearly. At Ely and the nearby vicinity of Eaglesnest Lakes in northeastern Minnesota, 54 per cent of the 5,500 hips of R. acicularis processed from 1944 to 1950 yielded mature larvae of Rhag. basiola. Similarly, 40 per cent of 5,104 hips of R. blanda from the same years and part of Minnesota, produced larvae. Farther south, in extreme northwestern Wisconsin, 2,276 hips of R. blanda proved to be inhabited at the average rate of only 7 per cent over a three-year period. At Chetek and Rice Lake, Wisconsin, 17,751 hips from R. blanda and R. carolina contained larvae at the rate of 9 per cent. And finally, from northern to central Illinois, so far as Ridgefarm below Danville, and Taylorville, 16,340 hips of R. carolina yielded 682 larvae, or a percent- THE ROSE HIP FLIES, TRYPETIDAE, DIPTERA 25 age of only 4. Some large samples taken in the Illinois area were entirely without the species. Below latitude 35°, Rhag. basiola has not, to my knowledge, been reported. In addition to this variation in the north-south direction, secondary differences occur locally, that may be correlated with varying suit- ability of the several rose hosts, conditions of drought that shrink the hips and render them intolerable to the developing larvae, severity of the winter seasons, and certainly, as will be shown below, the numerical fluctuations of several endoparasites of the larvae and puparia. There- fore, larval populations of Rhag. basiola are locally, as well as geo- graphically, subject to much quantitative variation. Pupariation. In September and October, at Urbana, the mature larva of Rhag. basiola undergoes the change to the puparial form within the day it emerges from the rose hip. In its natural habitat, its several hours of prepuparial mobility doubtlessly are spent in establishing itself in the duff or soil under the rose plant. When contained within a glass vessel, this larva crawls about energetically for an hour or more, then gradually reduces its rate of travel until the inevitable process of puparia- tion completely immobilizes it. Larvae that emerge in the cool later part of fall required as much as 10 hours to transform and, at near-freezing point, a few failed to pupariate. During the change from larva to pu- parium, the body shortens, thickens and becomes oval, and the soft white cuticle becomes leathery, then hard and ocherous yellow. If kept in dry atmosphere for several days, the puparium loses a considerable amount of weight and its cuticle turns brittle and pale stramineous, but if then placed on water-soaked cotton, it regains its lost weight, leathery flexibility and ocherous color (Balduf, 1957). Commonly the puparia became so firmly attached to the surfaces of the glass containers in which they had transformed that their shells sometimes broke, with fatal consequences, when they were pried loose. At first I supposed that the adhesion resulted from the sticky hip juices through which the larvae had crawled. But according to Fraenkel and Brookes (1953), who observed similar attachment in other higher Dip- tera, the almost fully contracted larvae suddenly pour forth a clear liquid from their then enlarged salivary glands. This secretion spreads over the venter of the body, hardens quickly and serves to cement it to the support. Possibly this is the mechanism of attachment also in Rhag. basiola. Prepupa. For three to five days of normal fall temperatures at Urbana, the newly formed puparium contains a body which, in this form or stage of dipterous insect, has sometimes been regarded as a fourth larval instar, and in which it still resembles a larva both in external and internal 96 INSECTS IN ROSE HIPS organization. Within a few days it becomes transformed into a pupa, with marked structural and physiological changes. This process sug- gests that the hypothetical fourth instar is not larval. As a substage of Diptera, it is homologous with that occurring just before the pupal stage of other holometabolous insects, where it is recognized as a prepupa. In Rhag. basiola, this stage begins with pupariation and terminates with the exuviation of the prepupal cuticle and the externalization of the leg cases and wing pads and the division of the body into three distinct regions. A casual examination of the viscera of a prepupa one to three days old reveals its more obvious structural features. Most prominent is the extensive bulky fat body. As in the larval instars, this consists of a thin network composed of numerous minute clusters of “fat” cells, each of the clusters consisting in turn of many still smaller cells and linked to other clusters by short threads. The network conforms to the shape of the body cavity and the visceral organs for which it forms a lining. The web disintegrates into its numerous separate component clusters when the insect attains the early state or phase of the pupa, i.e., about the third to fifth day following the inception of pupariation. The alimentary canal retains its larval form but is now empty, transparent and therefore obscure amid the massive fat clusters that envelop it and all other organs in the body cavity. Only the fusiform sector of the mid intestine, that lies just anterior to the Malpighian tubes, remains readily visible due to its content of an elongate slender spindle-shaped reddish brown solid mass of feces-like substance. Another system of organs easily dis- tinguished amid the voluminous fat body is the pair of yellow, or some- times greenish yellow, Malpighian tubes. One tube arises from each lateral side of the alimentary canal at the juncture of the mid and hind intestines. Each consists of a short basal stem, which bears two long, filiform, flexible chainlike arms of equal length. On close inspection, the pair of whitish, wiener-shaped salivary glands, supported by their short slender ducts, can be seen. But the status of the muscles, heart, nerves and tracheae cannot be ascertained by casual examination, and presum- ably will be generated from the imaginal disks during the later phase of pupation. Gas in prepupae. The abdomen of the prepupa of Rhag. basiola was regularly found to be inflated with a quantity of clear gas. It may be seen in the body removed intact from the puparium, or, in some cases, through the puparial wall. On account of their inflated state, the pre- pupae are so reduced in weight vs. bulk that they float lightly on water. Reference to my numerous notes on this phenomenon shows that the gas is present during September and October, and specifically in puparia containing prepupae aged one to four days but mostly about three days, THE ROSE HIP FLIES, TRYPETIDAE, DIPTERA 27 but not in the subsequent early phase of the true pupa. Small, weak pre- pupae lacked the gas, whereas prepupae parasitized by first instars of the chalcid Halticoptera rosae Burks contained moderate amounts of it. Since the gas represents inspired air and occupies the body cavity of the prepupa, it seems likely that it incidentally provides the supply of oxygen for the parasitic larva. It also appears likely that the pressure exerted against the inner wall of the prepupal body cavity may aid the eversion of the pupal appendages, which marks the termination of the prepupal stage. The gas was found to vary in amount from: little to so much that it occupied most of the lumen of the abdomen. The amounts may possibly be correlated with parasitism and ordinarily the age of the prepupa, the older individuals possibly having consumed larger quanti- ties of it in respiration. Snodgrass (1924, p. 25) states concerning Rhag. pomonella Walsh that “this gas is evidently the respiratory medium of the pupa,” and that it gains entrance through the anterior tracheal tubes. Pupation. At Urbana, the prepupa transforms to the true pupa in three to five days after pupariation. This takes place primarily in September and October. In this study, the newly formed pupa is designated the “early phase pupa,” but in more formal terms it has been called, by Snodgrass (1924) and other investigators quoted by him, the “crypto- cephalic substage” of the pupa. Its salivary glands, Malpighian tubes, and the alimentary tract with its small red-brown mass of waste, remain identical with those of the prepupa. The prepupal skin, having been molted at the termination of the prepupal stage, is retained unbroken around the pupa and forms a lining to the puparial wall. The three dis- tinct body regions are evident, as are also the evaginated wing pads and leg sacks. These appendages appear as empty chitinous shells except that their lumina contain a few, free-floating particles of fat body. Particularly the leg sacks become obviously longer after the inception of pupation, and their extension backward to the apex of the abdomen is an easily identified indication that the early phase of the pupa has been attained. Insofar as gross observation can determine, all normal individuals of this species uniformly attain precisely this structural or developmental early pupal state during August to late fall, and, moreover, appear to undergo no further changes, either externally or internally, until the following spring. Pupal diapause. This early phase pupa exists in a state of suspended development or diapause of eight or nine months over the fall, winter and spring seasons. The small early fraction of the mature larval popula- tion attains this condition in late August, the rest in September and Octo- ber and even as late as early December, as in the case of larvae obtained 98 INSECTS IN ROSE HIPS at Kickapoo Park in 1946. Were it unfavorably low atmospheric tempera- tures that induce the diapause, the larvae maturing in the warm weather of late August or early September would have advanced farther toward, if not to, the adult stage than those emerging from the hips from late September to late October, or thereafter, when cooler to cold weather prevails. Instead of wintering as prepupae, pupae in different phases of development, or as adults, the species uniformly, and regardless of the season or temperature, attains one developmental stage only, that of the early phase pupa, before winter sets in. This invariable confinement of the winter stage within these narrow structural limits implies a causa- tive factor of an equally constant character which is independent of environmental temperature. Heart beats of pupa. From the dorsal aspect, the beating heart of three early phase pupae was observed. The observations were possible when pupae were extracted unbroken from the puparia, or as breaks in the dorsal puparial walls afforded a view of the pupae within. The pressure of the air bubble in the pupae caused the meshes of the adipose web to spread, and the pulsations of the underlying heart alternately enlarged and diminished the aperture, so that the number of beats per minute was readily counted. The rate of the beat in two pupae was 85 and 100 per minute. In the first there was a slight pause about every three or four beats which indicated an occasional prolonged dilation, as if pulsations were about to stop, but the pause was followed by a quick contraction. In the third pupa the beats occurred at the rate of 90 in one sample minute, and increased to 125 after approximately 15 minutes of exposure to warmth from the microscope lamp. When this pupa had cooled about 15 minutes, the pulsations had decreased to 120 per minute. Spring development of pupa. When the puparia, kept on soil in flower pot cages during the winter outside, are brought in, in early May, they contain pupae which have remained in the early phase they had at- tained in the previous fall. The disintegrated adipose material still fills the body cavity, and the salivary, digestive and excretory organs retain their winter condition. Pupal development is resumed between the third week of May and early June. A perceptible change takes place in the texture of the fat particles. These are now further disintegrated, the clusters being resolved into their much smaller component granules, giving the fat substance a smooth, homogenized, creamy appearance. Gradually this mass of material is consumed, being apparently utilized for the formation of the organic systems of the adult insect from the imaginal discs. For example, the tarsal cuticles which, in the early pupal phase, evidenced segmentation only as shallow, broadly rounded con- THE ROSE HIP FLIES, TRYPETIDAE, DIPTERA 29 strictions, now contain the developing segments of the adult tarsus within. At first vague and devoid of vestiture, but gradually more distinct and spiny, the tarsal segments serve as a readily observable measure of the progress of the pupa toward the mature adult state. Simultaneous with the development of the tarsi, the well-formed head, antennae and mouthparts, the reddening eyes and their concise facets, the advent of the dark bristles against the background of yellow and orange body regions, and the appearance of veins in the immature wings were fur- ther indications that the fly would soon employ its ptilinum to effect escape from the puparium. However, not a few puparia yield no flies, or yield them belatedly. Parasitism by the larvae of Opius and Halticoptera permits the host to pupariate but destroys it before the pupal stage is attained, or com- pleted, respectively. Undernourishment due to competition with other larvae of their own species, or those of Cydia packardi in the same hip, or the chance that they may inhabit dry, diseased or overripe hips, may result in larvae so weak and puny that they can barely pupariate, but develop no farther. Prolonged pupal diapause. Of special interest were nine normal pu- paria which failed to yield adults in due season. When these were dis- covered, they had persisted three to six months beyond June to July, the months in which adults normally emerge. They had originated from hips of four different species of roses in Minnesota, Utah and Nebraska. However, they had two features in common: all were among the largest and most vigorous of their lots, and the eight which I dissected con- tained large healthy pupae, still in the early developmental wintering phase. So far as treatment is concerned, the nine puparia suffered various degrees of desiccation due to my neglect for three months or more. I wetted the ninth puparium with water, suspecting it to be in a state like that of the eight. This puparium then yielded a large normal fly, but a full year after the usual June date. It pupariated in the fall of 1947, therefore should normally have become an adult in June of 1948, but, presumably because it had suffered desiccation, issued as an adult in June, 1949. Perhaps the vigor of these pupae, or the character of the puparial walls, enabled them to tolerate the partial desiccation, whereas indi- viduals of less vitality might have perished. Presumably the eight would have produced flies in the manner of the ninth, had they been spared and wetted. These instances of delay exemplify a form of development which differs from that characteristic of all early phase pupae of Rhag. basiola only in that it persisted a year longer. Such cases warrant further study with reference to the causative factors. Life cycle. All the pertinent data obtained show that Rhag. basiola 30 INSECTS IN ROSE HIPS completes one entire cycle, or generation, in a year in all the states and provinces from which materials were obtained. These data were derived from cage rearings of hundreds of adults, from emergence records for more than 5,000 mature larvae from hips, dissection of more than 700 hips from the field to determine the season of embryonic and larval de- velopment, and from examination of puparia of all ages to learn the winter state and subsequent progress of pupation. Acquired primarily in the years 1945 to 1950, the records pertain chiefly to northeastern Minnesota, northwestern Wisconsin and east central Illinois. Adult life begins with the emergence of the flies from the overwintered puparia, which was continuous at Urbana from May 17 to July 12, an over-all period of 56 days. Year by year, the inclusive periods of emergence were 1946: May 30, July 3, or 35 days; 1947: May 17, July 10, or 54 days; 1948: May 28, July 6, or 40 days; 1950: May 30, June 27, or 29 days. When fed on an aqueous solution of sugar or yeast, females lived 21 to 33 days, males 17 to 28 days. Allowing a preovipositional period of 12 days, egg-laying may begin at Urbana before June 1; however, egg pits were first found on R. palus- tris on July 4, 1949. The number of punctured hips increased rapidly to 90-+- per cent by August 16. Possibly the latest occurrence of oviposit- ing flies in the field (at Eaglesnest) was September 3 to 4, as indicated by the discovery of newly laid eggs in hips. The embryonic period involves at least the months of July and August. Five-year records for Eaglesnest show that only embryos, i.e., no larvae, occurred on July 10 to 12 in hips of R. acicularis. Some embryos remained in the field samples of hips from August 23 to September 1, indicating that oviposition may continue well into September. The dura- tion of embryonic development may be estimated by dissecting eggs of known age and by observing first dates of oviposition and newly hatched first instars. At 11 days, the embryo had already reached the advanced phase, as indicated by its maggot-like form, convoluted alimentary tract and fully developed mouth hooks. Embryos 17, 23 and 24 days old had made no macroscopically observable advance over that seen in the 11-day embryo. The first egg was laid on R. palustris, Urbana, on July 4, 1949, and the earliest first instar was discovered August 11, or 38 days later. These evidences indicate that embryonic life may be as long as five weeks or more, and that the early and intermediate phases are completed probably in much less than 11 days. The unusual duration of the ad- vanced state suggests that a limited diapause is involved. That this phase is comparatively long is shown also by the fact that by far the most em- bryos found through several summers at Eaglesnest were in this state. Larval stages, as well as other stages, are influenced by the prevailing THE ROSE HIP FLIES, TRYPETIDAE, DIPTERA 31 temperatures. Some first instars had hatched on August 11 at Urbana, but these were probably not the first, considering that small first instars were found July 17 to 19 in a five-year study in relatively colder northeastern Minnesota. Some first instars persisted at Eaglesnest to September 1, and the comparatively large numbers showed that they may extend far into this month. Second instars were found from August 18 to Septem- ber 1, while third-stage larvae also occurred during all that time. Clearly, second instars continue far into September, and rearing records prove conclusively that the third instars mature and emerge from the hips at Eaglesnest from August 22 to October 15. At Chetek, Wisconsin, the mature larvae emerge from near September first to October 24. In east central Illinois, emergence begins about mid-September, commonly con- tinues into early November, and in the instance of one sample from Kickapoo Park, larvae issued to December 13, 1945. Thus, the emergence of mature larvae from hips appears to occur later from north to south, or from the colder latitude in northeastern Minnesota to the warmer in central Illinois. The time required for the development of individual larvae may be estimated or approximated from the dates of the earliest first instar and the dates the first mature larvae appear from the hip. Such data, ob- tained from Rosa acicularis and R. blanda at Eaglesnest in five years, indicate that the individual larval development required 30 to 35 days. Since pupariation occurs within a few hours after the mature larvae leave the hips, and the early phase of the pupal state is attained in 3 to 5 days in average temperatures, the time period of this early pupa be- gins only half a week after the dates of larval maturity, i.e., from mid- August to early December, but usually late October, and persists through the winter into mid-May or June of the ensuing year. In May to June, pupal development resumes and continues into July. The emergence of the adults from the puparia—between May 17 and July 12—marks the termination of the pupal stage which may extend from mid-August of one year to July of the second year. The adult flies, after a preoviposi- tional period of about 12 days, initiate the next annual cycle when the females introduce their eggs into new rose hips. Under average conditions, the yearly development proceeds approxi- mately as follows in the over-all area involved. The larvae in the hips mature and emerge, and pupariate within a few hours, from late August to late October. In three to five days the larva undergoes transformation through the prepupa to the early or cryptocephalic phase of the pupal stage which is then followed by a diapause lasting through fall, winter, and spring to late April or early May. Pupal development resumes in May, and adults issue from the puparia during a period beginning late in May and ending early in July. The flies oviposit into the hips from June oo INSECTS IN ROSE HIPS to early September, and the larva completes its three stadia from late July or early August to October. HyMenopterous Parasites oF Rhagoletis basiola I have found four species of parasitic Hymenoptera that attack the immature stages of Rhagoletis basiola. Three of these have been iden- tified as new; two species of Opius, Braconidae, and one of the chalcid genus, Halticoptera, Pteromalidae. The identity of the fourth remains in doubt, owing to its scarcity and my fruitless efforts to rear the adult. Some evidences, to be stated later, indicate that it may be a chalcid of the family Eupelmidae. All these parasites are solitary and internal, and, with the possible exception of the supposed eupelmid, have one generation in a year in coordination with the single annual cycle of the host. Opius rosicola Mues. and Opius balduft. Mues., Braconidae, Hymenoptera Hosts and distribution. These new species were named and described by Muesebeck (1949). Rhag. basiola is their only host known to date. The specific characteristics of hips from six species of Rosa (R. acicu- laris, blanda, carolina, arkansana, virginiana and setigera, and some varieties of the foregoing) appear not to affect observably the utilization of Rhag. basiola as a host. In general, it may be said that the parasites have been obtained from all species of Rosa for which adequate samples of hips, well-inhabited by Rhag. basiola, have been available for investi- gation. One notable exception was Rosa Macounii from Crete, Nebraska, whose hips yielded 146 flies but no parasites. Opius rosicola was reared from puparia of Rhag. basiola that origi- nated at Saskatoon, Saskatchewan; Eaglesnest, Ely, and along the Minne- sota-U.S. Highway 61, in northeastern Minnesota; and at Solon Springs, Sarona, Chetek and Madison, Wisconsin. Opius baldufi emerged from puparia of Rhag. basiola deriving from three points in Alaska (milepost 326 at Richardson highway, Fairbanks; milepost 297.3 Richardson highway at Big Delta, and 144 miles north of the Canadian border on the Alcan highway); from Eaglesnest, Ely, and the intersection of the Minnesota-U.S. highway 61 with Pike Lake Road, and at Bally Creek, near Grand Marais, Minnesota; and from Chetek, Wisconsin; Orono, Maine; Urbana, Illinois, and Kickapoo State Park, near Urbana. The above records indicate that both species are widely distributed in North America, but apparently limited to areas northward from the HYMENOPTEROUS PARASITES OF Rhagoletis basiola a8 latitude of central Illinois, which appears also to be the range of Rhag. basiola. Recognition of Opius spp. While the adults of these species of Opius are structurally very distinct, distinguishing bodily characteristics of the larvae were not readily observable and the matter was not pursued. No specific differences in modes of life have been observed, and for this reason the bionomics of the two species are treated simultaneously. The following distinctive characteristics that mark the adults of O. rosicola and O. baldufi are chosen from the original descriptions by Muesebeck (1949). Most distinctive are the wing characteristics: com- pare Figures 14 and 15. Some of the data on the number of antennal segments are based on specimens in my collection. Opius rosicola (Fig. 14). Head ferruginous; mesoscutum with a long, slitlike median impression on posterior half; first tergite of abdomen rugulose, with two dorsal keels converging to about the apical third, where they are obsolescent; stigma short, broad; antennae of females with 35 to 42 segments, males 31 to 42 and for the species as a whole, usually 36 to 38. Opius baldufi (Fig. 15; wings). Head black, only the lower parts fer- ruginous; mesoscutum with a median dimple-like impression before hind margin; first tergite of abdomen rugulose, with two nearly parallel, widely separated keels on basal half; stigma long and narrow; antennae of females with 40 to 47 segments, males 40 to 47 and for the species as a whole, usually 44 to 47. Both sexes of O. baldufi and rosicola have small inequalities in the number of segments composing the right and left antennae. Sex ratio. In the years 1945 to 1951, at least 300 adults of O. rosicola and 100 O. baldufi were reared. The series of rosicola comprised 136 males and 153 females, and of baldufi, 42 and 48 respectively. These small samples indicate the probability of a 50-50 ratio of males and fe- males in both species. Adult and egg. In the field, the brown adults (Fig. 14) were infre- quently seen resting on the leaves of roses or plants adjacent, as they flew somewhat briskly from place to place, usually for short distances. The eggs (Fig. 16) of these Opius species, obtained by dissecting larval Rhag. basiola, are about 0.5 mm. long, with a broadly rounded, large cephalic end and a much narrower tapering caudal end. The chorion appears smooth, whitish and translucent. The large quadrate head of the advanced embryo is visible externally. The activity of oviposition was observed only once in the field. When discovered, the female had already inserted the terebra deeply into a hip of Rosa acicularis, and lay quiet with the venter of her body fully 34 INSECTS IN ROSE HIPS in contact with the upper surface of the hip. That an egg had been de- posited was shown by subsequent dissection of the hip, revealing a first instar larva of Rhag. basiola, which, in turn, contained a newly hatched larva of Opius that had not yet ingested food. The eggs found invariably remained free and unattached in the body cavity of the host. While eggs were not infrequently found in the host larvae, few of them contained living embryos; their scarcity indicated a brief embryonic life, probably one or two days. The majority of eggs present in the hosts contained no identifiable embryos but only dull grayish white homogeneous granules, apparently yolk. The frequency of such eggs leads me to conclude they were dead, probably unfertilized. The chorions of some such dead eggs had collapsed to various extents and had even become twisted or curled, and usually more or less brown. For convenience, I refer below to the latter as “brown bodies.” These occurred in living second and third instars of the host, but mostly in puparia. Thus deformation and discoloration of the dead egg developed over a period of weeks in some instances, since the female attacks first instars as well as seconds and thirds. The best evidence as to the host stages attacked by the ovipositing Opius are my records of early phase first instars found in hosts dissected. The retention of all or much of the gelatinous envelope around the body, and the simple thysanuroid form of the parasite are reliable indicators of such early phase larvae. Twenty-five such larvae were found in host larvae: 5 in first instars, 14 in second instars, 6 in third instars. The pu- paria of Rhag. basiola never contained parasites of this early phase. First larval instar. The number of instars has not been ascertained but gross external structural characteristics indicate there are at least three. As in other opiine Ichneumonidae, the first instars of the present species differ sharply in form from the subsequent ones (Clausen, 1940, p. 48). The newly hatched first instar of O. rosicola and O. baldufi in its usu- ally moderately curved resting form, is 0.6 to 0.7 mm. long, and en- larges to 1.5 mm. during its slow development to the molting state. The prothorax at first is barely as broad as the head, while the abdomen tapers behind, effecting an over-all thysanuroid form. Gradually the thorax and abdomen expand in all directions to the shape of a broadly round sack (Fig. 17), while the head retains essentially its original size and form. In this advanced phase of the first instar, the postcephalic region of the body remains relatively depressed but is twice as broad and thick as the head. The head is quadrate, pale brown, and sclerotized, with a pair of median denticles on the anterior edge of the labrum; antennae are short, and appear soft and flexible; the mandibles are brown, prominent, and robust, falcate in form, attached under the protruding labrum, and may HYMENOPTEROUS PARASITES OF Rhagoletis basiola 35 be opened very widely; two obscure fleshy mamma-like extensible ap- pendages arise from the sternum of the prothorax, one near each outer edge; the apex of each appendage bears a pair of obscure divergent spines. Being concolorous with the body and contractile, the “mammae” are usually not visible, but become briefly so when extended as the larva rolls over sidewise, or writhes energetically in a drop of water. The postcephalic part of the body consists of 13 segments, the small terminal segment being obscured by invagination in some postures. A broad-based minute spur is present on the mid-dorsal line of each of segments 6 to 9. The sternum of at least the first nine postcephalic segments bears a pair of minute bristles. The abdomen terminates as a pair of soft, transparent, extensible pointed processes, which appear joined to the thirteenth seg- ment. Some investigators (Cameron, 1941; see also Clausen, 1940, p. 48) assert that the concave side of certain first instar Opius is the dorsal rather than the ventral aspect. I did not inquire into this point in the instance of Opius spp. from Rhag. basiola in rose hips. Initially, the first instars are wholly enveloped in a thick layer of a gelatinous, grayish white substance composed of minute granules or spherules. Similar envelopes have been reported for Opius humilis Silv. and Opius fletcheri Silv. by Pemberton and Willard (1918) and O. ilicis Nixon by Cameron (1941). Cameron describes the substance as “a loose mass of large serosal cells with large nuclei,” thereby indicating it originates in the embryo of the parasite. As a consequence of enlargement of the parasite by feeding, and pre- sumably also from contact with structures in the host, the envelopes of first instar O. rosicola and O. baldufi break here and there, and hang in tatters, and are, in a few days, entirely lost, leaving the body bare. Similar envelopes do not recur on subsequent instars. Boese (1936) described what may have been similar envelopes around hymenopterous parasites in lepidopterous hosts. For example, he re- ported that the first instars of Apanteles glomeratus ( Linn.) in the larva of Pieris brassicae (Linn.) become encysted by phagocytic haemo- cytes which undergo mitosis, forming a capsule of cells around the parasite. He observed further that the larger parasitic larvae are not subject to such attack. The changes in form and size observable as the first instar of Opius develops from its newly hatched to its latent molting state results largely and directly from the ingestion of much food material from the host’s body cavity. This dilates the alimentary canal which, in turn, presses outward against the body wall to enlarge the entire larva, excepting its head. The food taken by the first instar consists always of a pale yellow fluid, which is probably host lymph, and golden yellow globules of many 36 INSECTS IN ROSE HIPS sizes suspended in the fluid. The globules are probably free fat originat- ing in the fat body of the host. Larval diapause. After feeding for an undetermined time, but perhaps a week, the first instar attains its maximum size and sacklike form. Then a curious phenomenon is observed. Whereas most immature insects which have reached the advanced phase proceed quite promptly to molt and continue their growth, the first instars of Opius baldufi and O. rosi- cola persist in their latent molting condition for periods varying prob- ably from a few days to a few weeks. The duration depends on the age or developmental state which the host has reached when the female Opius introduces her egg. The parasite may persist in the latent molting state for several weeks when the egg is placed in a first instar of the host, or only one or two weeks or a few days when the host attacked is a second or third instar. The persistence of the Opius in this advanced, molt-ready or latent phase of the first instar, although of variable duration, constitutes a true diapause. How the diapause is initiated remains unknown, but presum- ably by some physiological factor—possibly a juvenile hormone, pro- duced in the parasite itself. This inhibiting agent occurs in all the first instars of Opius spp. in Rhag. basiola and invariably suspends their further development, usually for a matter of a few weeks. The nature of this diapause differs from most delays known among insects in that its termination appears to be effected by some factor operative in the host. This belief was indicated by the results of dissec- tions of several large samples of larval and puparial hosts. So long as the host remained in its larval stage in the hip, the parasite remained a first instar, although invariably attaining its latent phase. However, pu- parial hosts only three to five days old contained Opius that had molted and advanced beyond the first instar. The coincidence of the molt of Opius and the prepupation of the host in the puparium suggests strongly that a hormone, produced by the host, activates the first ecdysis of the parasite as well as its own trans- formation to the pupal stage. More specifically, this molt of Opius prob- ably takes place within a few hours after the full-grown larval host emerges from the rose hip and pupariates. In other words, the first in- star and its diapause seem to terminate approximately at the inception of the prepupal phase of the host. The suggested dependence of the parasite on the host for the stimulus to molt implies that the hormone-producing glands of Opius spp. have deteriorated functionally in a very long, well-adjusted relation with Rhag. basiola. Advanced larvae. The exuviation of the dilated sacklike first instar HYMENOPTEROUS PARASITES OF Rhagoletis basiola 37 reveals a larva of very different structural characteristics (Fig. 18). The sclerotized head, the prominent falcate mandibles, the prosternal “mam- mae” and caudal processes of the first instar are lacking. The head re- mains somewhat quadrate, but is now like the rest of the white body in color and texture, and its anterior rounded face has a distinct vertical notch at the middle. There seem to be 12 postcephalic segments. These remain recognizably distinct until the pressure from much ingested food later obscures or obliterates their boundaries. The body is apparently bare, smooth and semitransparent excepting the area rendered opaque by the food and retained fecal matter. When removed from the puparial host, the larva lies on its side and, as then seen, is prominently and roundly humped above and concave beneath—a form reminiscent of the larva of the Colorado potato beetle. As seen from above, the total body has a mere sacklike shape, with the head narrower. Larvae of this form range from about 2.0 to 3.5 mm. in length. It is likely that larvae of the above general type represent two instars. The final instar is distinguishable by its size—mostly 3.0 to 4.0 mm. long—its plump oval form, the relatively large speckles of white adipose substance under the cuticle, and its vestiture of numerous minute, uni- formly short, close-set cuticular spinulae over the dorsum and perhaps other parts of the body. The spinulae arise from cuticular pits—the latter joined by minute grooves. Three pinkish red facial marks have been seen on some full-grown larvae, and presumably occur on all. These may represent structures of the pupa foreshadowed within the larval cuticle. First, it should be emphasized that the entire development of the Opius larvae, subsequent to the first stadium, takes place within the puparium of the host. It is initiated with the molt of the first instar, and terminates with the eventual complete consumption of the host fat body. The exuviae of the first instar may usually, with but little search, be found near the young second instars. If newly shed, the discarded cuticle com- monly remained entire, with a longitudinal break in the thoracic terga. Sometimes head and postcephalic parts were found separated. This larval form is readily obtained by prying one end off the puparium and pressing it with needles, causing the larva, which is then still much smaller than the host lumen, to slither out into a bit of water in the dissecting dish. The parasite then contains varying amounts of yellowish fluid and globules of fat that dilate the blind alimentary tract to various extents. This form is retained, but with increase in size, until the final instar is reached. The growth of the parasite to the final instar within the puparium is unusually rapid, when compared with the long duration of the first stadium. Dissections of numerous puparia varying from barely hardened to several days of age showed that the parasites had reached their maxi- 38 INSECTS IN ROSE HIPS mum size and the final instar in about two to three days after their first exuviation. In this short time, the body enlarges so much that it fills the puparium snugly and makes the removal of whole specimens difh- cult. The extreme dilatation of the body to its full size results from in- gestion of the disintegrated fat body of the host. The body cavity of puparial Rhag. basiola normally is filled with such particles of fat body. These are readily sucked into the tract of the Opius larva, with appar- ently little immediate change in their form. The process appears to be essentially mechanical, and may be likened to the siphoning of a fluid from one barrel to another. The increase in size of the body through the two to three days may perhaps therefore be described as enlargement or dilatation, rather than growth. At any rate, the removal of the volumi- nous fat body is so complete that the host is now dead, with only the chitinous organs remaining depressed within the puparium. This dilated state of the full-grown parasite is not tantamount to maturity. The ingested food gradually becomes reduced in volume, and as it presumably undergoes chemical modifications in digestion, it changes also from a yellow to a dusky brown contracted soupy mass, or meconium. Now the shrunken mature larva, filling the host lumen loosely, and enveloped in the host’s tough unbroken larval cuticle, is easily extracted in a whole state from the puparium. In this matured state, the parasitic larva exists during the ensuing fall and winter. When dissecting puparia to obtain parasites, I sometimes recorded the direction in which the mature Opius lay headed within. No one posi- tion proved to be taken, but most of those observed lay headed toward the cephalic end of the puparium. It was noted also that the walls of parasitized puparia are usually thin, brittle and dull yellow as com- pared with the thicker, harder stramineous walls of parasite-free puparia, or such as contain first instars. A second diapause. It was stated above that the first instar of Opius persists in its latent molting state for periods of varying length, and that the termination of that diapause appears to depend on stimulation from hormones in the prepupal host. A second diapause begins with the maturity of the last instar of Opius in late summer to fall. It persists as such a full-grown final instar in the host puparium until spring of the next year, when it terminates in pupation after a duration of up to nine months. Inasmuch as the host has then been long since consumed, such hormonal stimulation as may be presumed to break the diapause in the next spring must originate within the parasite itself, incident to its larval-pupal transformation. Precocious and delayed development. Four examples were discov- ered in which the opiine parasite completed the development from egg HYMENOPTEROUS PARASITES OF Rhagoletis basiola 39 to adult within the summer and fall of one calendar year. All originated in hips of Rosa blanda and R. carolina taken at Chetek, Wisconsin. The hosts pupariated in September and October, at Urbana, and the para- sites proceeded to pupate without the usual winter diapause in the last instar. Two of the four were found in 1948, two in 1951. The two hosts of 1951 pupariated on September 24. When dissected on September 30, each contained a female pupa of Opius, whose eyes had turned brown, indicating that development to the adult stage had begun. The four puparia had been stored in shell vials in artificially warmed rooms with- out adequate moisture. If these conditions induced the precocious pupa- tion, the diapause of the full-grown larval Opius would seem to be un- stable in character. On the other hand, I discovered a few instances of delayed emer- gence of adult Opius from the puparia of Rhag. basiola. The adult state was attained a year later than usual. Presumably most of the additional year was spent in the mature larval state. Particulars surrounding these instances are lacking. Pupa. The full-grown larva of Opius spp. spins a scanty meshwork of silk on the inner wall of the host puparium. Transformation to the pupal form occurs in the warmer parts of the spring season. In the newly formed pupa, the abdomen is bright white, while the head, thorax, basal parts of the legs, the wings and the antennae are pale yellowish white. First structures to show changes from yellowish white are the compound eyes, ocelli and mandibles, which first turn pale red, and later maroon. The rest of the body passes through shades of yellow to the final brown, the abdomen being last to attain its mature color. The antennae extend caudad straight along the inner edges of the folded legs and the wing pads, and, being longer than the body, curl around the end of the abdomen. The short, sturdy ovipositor bends vertically upward. Life cycle. Like their host, Rhag. basiola, Opius rosicola and O. baldufi complete one cycle in a calendar year, their several life stages necessarily lagging chronologically behind the corresponding stages of the host. Puparia which originated in Saskatoon, northeastern Minnesota and northwestern Wisconsin yielded adult Opius rosicola at Urbana between June 22 and October 4, in the over-all period 1945 to 1951. Puparia from Alaska, northeastern Minnesota, Chetek, Wisconsin, Orono, Maine, and eastern Illinois, yielded adult Opius baldufi at Urbana between May 21 and September 21, in the same years. Accordingly, the emergence of adult O. rosicola extended through 96 days; of O. baldufi, 123 days. Actually, the spread of emergence in any one year and locality was con- siderably shorter that those numbers indicate; the extremes recorded for the several localities in one year were 11 to 79 days. It is worthy of 40 INSECTS IN ROSE HIPS note that the three adult males from Alaska issued between May 21 and 28. This antedated the earliest emergence from any other source by about two weeks, hence may signify an unusual hereditary tolerance of low temperatures. In a few cases, emergence of adults was delayed to the summer of the second year after parasitization. In the course of observations in the field at Eaglesnest, I discovered 19 adult Opius on Rosa acicularis and R. blanda (incl. Macounii) be- tween July 31 and September 5, 1945 and 1946. It is probable that they are abroad both before and after these dates. The rare discovery of live embryos in host larvae dissected suggests that the embryonic stage is very short. Other evidence indicates that the embryonic development is completed in less than 24 hours. Allowing some days for preoviposition and the retarding effects of northern latitude, the above data indicate that adult Opius are abroad parasitizing the larvae of Rhag. basiola at Eaglesnest chiefly from July into September. Further evidence on the time of oviposition is afforded by the dates of occurrence of unfed first instars of Opius less than 1.0 mm. long and still enclosed in their complete jelly-like envelopes. Larvae of this description were found at Eaglesnest between August 5, 1946, and September 11, 1949. Since the embryonic period may be less than one day in length, the dates of oviposition and of hatching are prac- tically identical. As stated above, the first instars of these Opius species may be pres- ent in any of the three instars of Rhag. basiola, but never in the puparia. Since the host larvae occur from July to October, the first instars of Opius are potentially present in these months. At Eaglesnest, I took the first instars during August and September, but, since investigations there had to be terminated before mid-September, the latest dates of their occurrence were not learned. The instars beyond the first are confined to the host puparia, and com- plete their development within two to three days after pupariation, destroying the host before it reaches the early phase pupal stage. Pupa- riation of the host is known to occur from mid-August to late fall in the over-all area investigated. Accordingly, the Opius larvae become full- grown from August to October. These full-grown parasite larvae also form the wintering stage, hence may be present from August to late spring and mid-summer of the ensuing year. I did not learn the duration of the pupal stage of individuals. The records relating to the emergence of the adults indicate that, excepting a few precocious individuals, pupation in the generation as a whole may occur from May to October, but ordinarily through late June to mid- August. Accordingly, the adult Opius appears from June forward, when the host has again attained the larval stage essential to parasitization. HYMENOPTEROUS PARASITES OF Rhagoletis basiola 41 Halticoptera rosae Burks, Pteromalidae, Hymenoptera Types and sex ratio. This new species ranks with Opius rosicola and O. baldufi in importance as a parasite of Rhag. basiola. It was described by Burks (1955) from 58 females and 48 males. Six of the female para- types and three of the male paratypes originated at Newport, Rhode Island in 1943 and 1944. The rest of the types were reared by me from hosts taken at various places in Wisconsin and northeastern Minnesota in 1945-1951. In those years, I reared 1,089 adults of H. rosae from places and rose hips named below. The sample consisted of 551 females and 538 males. Hosts and distribution. My series of adults, and numerous larvae dis- sected from the hosts, originated entirely in stages of the rose hip fly, Rhag. basiola. According to Burks (1955), the U.S. National Museum has specimens reared at Newport, Rhode Island, from Rhag. basiola, also others from Rhag. alternata. Subsequent to the publication of his arti- cle, Doctor Burks informed me by letter that the record from Rhag. alternata is an error and should read “Rhag. basiola” instead. Stone (1951) regards these flies as distinct species. All the Halticoptera rosae reared to date therefore appear to have originated in Rhag. basiola. The species of Rosa and the localities from which Halticoptera rosae is now known to me are as follows: Rosa acicularis vars. Fairbanks and Big Delta, Alaska; Eaglesnest and Ely, Minnesota; also at several points along the north shore of Lake Superior, Minnesota. Rosa blanda vars. Eaglesnest and Bally Creek, near Grand Marais, and Tower, Minnesota; also Chetek, Sarona and Gordon, Wisconsin, and Ottawa, Ontario. Rosa arkansana. Chetek and Madison, Wisconsin. Rosa carolina. Chetek, Wisconsin. Rosa lunelli and R. Macounii. Saskatoon, Saskatchewan. Rosa woodsii. Beulah, Manitoba, and Provo, Utah. Rosa virginiana. Orono, Maine; Newport, Rhode Island (Burks, 1955). Rosa rugosa. Newport, Rhode Island (Burks, 1955). The specimens from Newport were secured by Mayer, Kisliuk, and Mallia. These data show that Halticoptera occurs at least from Alaska on the north to Madison, Wisconsin, and from Provo, Utah, eastward to Ottawa, Ontario, and Orono, Maine. Although numbers of Rhag. basiola were reared from hips of several species of roses, including R. virginiana, pal- ustris, canina, eglanteria, gallica, carolina and setigera, in Ohio, Indiana, and eastern Illinois, this chalcid was not obtained. According to these evi- dences, H. rosae ranges southward to approximately the Wisconsin- 492 INSECTS IN ROSE HIPS Illinois state line, and appears to be more limited in distribution than its native host, Rhag. basiola. The adult. The characters given herewith are selected from the de- tailed description prepared by Burks (1955), and will serve to distin- guish Halticoptera rosae from other chalcids obtained from rose hips. Female (Fig. 19). Length 3.5 to 4.0 mm. Head black with bright green or faintly brassy metallic reflections; antennae mostly very dark brown, scape dark, metallic green; thorax above bright metallic blue-green; pleura and sterna black with faint brassy reflections; propodeum bright metallic green; abdominal petiole and gaster black with faint metallic green reflections. Male. Length 3.0 to 3.5 mm. Head and thorax bright metallic green; antennal scape yellow at base, shading to tan at apex, pedicel dark at base, distally golden tan, funicle uniformly golden tan, club slightly darkened; enlarged terminal segments of maxillary palpus slightly longer than wide, apex of terminal segment nipple-like and hirsute. Egg (Fig. 22). Length about 0.2 mm., apetiolate, pale whitish, chorion smooth; narrowly rounded at the caudal end, with gradual widening to the broadly rounded cephalic end. Larva (Figs. 23-27). The newly hatched first instars (Fig. 23) be- fore feeding are from 0.2 to 0.25 mm. long. In their mature state (Fig. 25), they attain a length of 0.6 to 0.9 mm., and a corresponding increase in diameter. The body of this instar is approximately fusiform, broadest in the thoracic region, tapering abruptly toward the head but gradually toward the cauda. The head is somewhat longer than wide, suboval in dorsal view, pale brown and lightly sclerotized, with the oral aperture subapical and ventral. The thorax and abdomen appear to comprise 13 segments, but sometimes there seemed to be 12, perhaps due to the re- traction of the small terminal member. The first two thoracic segments are bare, but the metathorax and each of the abdominal segments bears a transverse series of several spines on the dorsocaudal margin. The small terminal segment is distinct in that it bears spines of diverse lengths on its ventral side. Of the internal systems, only the alimentary tract is macroscopically visible, due to the opacity of its food contents. It appears to be a simple undifferentiated sack, narrow in the head and bulging little to much in the abdomen, depending on the amount of the contained food or waste. It remains closed behind, and extends caudodorsad to terminate blind in what appears to be the tenth postcephalic segment. I did not determine the number or characteristics of the instars that develop after the first molt. But the advanced larvae differ sharply from the first instar in form of body. Newly molted second instars (Fig. 26) HYMENOPTEROUS PARASITES OF Rhagoletis basiola 43 are about 1.2 mm. long, 0.6 mm. wide, with smooth, shiny cuticles and few or no small speckles of whitish adipose visible through the cuticle. This larva soon becomes sacklike, dilating increasingly as more host substance is ingested, and, as a consequence, the segmental lines of thorax and abdomen become obliterated. What I take to be last instars of Halticoptera rosae (Fig. 27) are 1.8 to 3.2 mm. long, robust, with 13 postcephalic segments; the head is short and retracted into the thorax, and the abdomen terminates bluntly and bends slightly ventrad at the apex. These larvae are wholly white, with a pearly luster, bare, smooth, with prominent transverse rounded dorsomedian ridges and bunlike pleural swellings on the thoracic and abdominal segments. Many small white speckles of adipose appear under the cuticle. The size and number of the speckles, the presence of ridges and swellings, and the absence of spinulae on the pearly cuticle distinguish the full-grown larvae of Halticoptera from those of Opius as these are found in the puparia of Rhag. basiola. Emergence and movements. The young adults escape by gnawing a subcircular hole through the side near the apex of the puparium as do also the adults of Opius spp. In the field, they were observed at rest on rose plants, and also moving deliberately over leaves and buds. Their locomotion is a combination of jumping and flying for short distances. Mating and oviposition. Apparent attempts at mating by one-day-old individuals that issued from puparia in capitivity were observed in sev- eral instances. Whether coitus occurred, it was impossible to ascertain. However, the observed impulse suggests that copulation may occur very early in adulthood and that the activity is of very short duration. In June, 1945, I discovered a chalcid, later identified as H. rosae, ap- parently engaged in ovipositing into a hip of Rosa acicularis in the field at Eaglesnest, Minnesota. Dissection showed the chalcid was working at a point where an egg of Rhag. basiola had been inserted, and also that the egg contained a minute living larva. By exposing fly eggs, laid in captivity, to reared females of H. rosae 11 days old, I established that the minute larva was indeed that of this chalcid. It demonstrated too that the green chalcids which emerged from puparia of Rhag. basiola institute the parasitic aspect of their life by inserting their eggs into those of the rose hip fly implanted into the hypanthia of the rose hips; also that the resulting larvae hatch within the embryonic host. These preliminary findings were confirmed hundreds of times in the years 1945 to 1951 by the discovery of such parasitic larvae in host embryos and larvae dissected from hips. The female of H. rosae discovered on a hip in June, 1945, rapidly 44 INSECTS IN ROSE HIPS tapped the locus of the fly egg with the apices of the antennae. Then several minutes were spent in stepping alternately forward and back- ward over the egg site, meanwhile palping the place with the antennae and applying the abdomen to it as if attempting to insert the ovipositor. When the host was examined four days later, the parasite had already hatched within it. In subsequent observations, it was disclosed that the eggs are placed free within the body of the host embryo, i.e., not attached to any interior structure. I have found the eggs of H. rosae in the early, the intermediate and the advanced phases of the host embryo. But since the “egg” stage of Rhag. basiola appears to last as much as 40 days, and the embryo is in its advanced phase probably eight-tenths of this period, the female H. rosae chances to place the large majority of her eggs in the advanced form of the host. The egg is inserted shallowly into the cauda of the host embryo, as is shown by the consistent occurrence of the young first instar in the posterior quarter or fifth of the first instar of Rhag. basiola. The above lone instance indicates the so-called “egg stage” is less than four days, and probably one or two days, if not mere hours, in duration. That it is brief is shown also by the discovery of very few embryos as compared with many newly hatched larvae in the host em- bryos scrutinized. “Brown bodies” (Figs. 20, 21). In dissecting the eggs, three instars and the pupae of Rhag. basiola I found, not infrequently, small inert bodies which, after considerable puzzlement I identified as dead eggs of H. rosae (Figs. 20, 21). The best-preserved specimens retained the form of normal eggs and contained recognizable amounts of gray granu- lar yolk, yet no clear evidence of embryos. Older eggs had collapsed, with the terminal thirds medium brown and the intermediate third pale brown, while still older specimens had collapsed completely, turned dark brown and sometimes became more or less curled or twisted, cork- screw-like. The increase in their numbers as the summer advanced, and failure to find embryos in these bodies, seem to mean the eggs were not fertilized and that the sperm supply for the late-season or old females may have been exhausted. This condition might develop in species, like H. rosae, that exhibit a distinct protandry, with the consequent absence of males when the late fraction of females emerges from the host puparia. These “brown bodies” are comparable, as to their origin, to those de- scribed above for Opius. They were more numerous in Halticoptera. Larval life. The parasite hatches within the embryo, not in the first instar, of the host. This is what may be expected since embryonation of HYMENOPTEROUS PARASITES OF Rhagoletis basiola 45 the parasite is completed in a few hours or days, whereas that of the host seems to require about 40 days. Among numerous first instars of H. rosae obtained by dissection from the eggs of Rhag. basiola, only one was taken in the process of hatch- ing. The infrequency of this occurrence in itself indicates that the “egg” stage is short and the activity of hatching of brief duration. In the one case discovered, the posterior three-fourths of the parasite still lay enveloped in the translucent chorion. By means of bodily pulsations and feeble writhing, the larva extricated itself from the chorion. One parasite per host embryo is the predominating condition, but two are not infrequently present. In some cases of duality, one parasite was found dead, but I have seen no evidence of cannibalism; moreover, the probability that the parasite feeds little, if any, in the host embryo, indicates that death is not due to competition for food supply. Further, not uncommonly a single host embryo contained a living parasite and also one or two brown bodies, or two brown bodies alone. In the extreme instance of multiple parasitism noted (discovered on August 7, 1945, at Eaglesnest) an advanced host embryo simultaneously contained two brown bodies and two living, newly hatched first instars. Eventually, all supernumerary parasites die. I have no evidence to show that a host can support more than one parasite to maturity. First instar. When the parasitized host embryo hatches to become a free-moving, boring larva in the hypanthium of the rose hip, it contains in its body cavity the small, early phase instar of H. rosae. Soon the supply of pale white yolk which it carries over from its embryonic state seems to be depleted, for the young instar of a few days of age contains substance it has ingested from the first instar host. Its alimentary tract contains a pale yellowish fluid, probably blood plasma, in which are suspended globules of rich yellow material that vary much in size and probably represent droplets of fat from the host’s fat body. The rate of ingestion of host substance is very slow or small. The parasite increases in size, but slowly, multiplying its bulk as a first instar by four to five times while the host grows from a young first instar about 1.0 mm. long to a mature larva about 8.0 or 9.0 mm. long. The amount of material ingested is such that the growth of the host is not visibly retarded in any of the instars. The Rhagoletis soon attain a plump appearance despite the presence of the first instar H. rosae, and I have no conclusive evidence that mortality from normal parasitism occurs in host larvae of any instar. Many young first instars of the hip fly die when they inhabit the very hard hypanthium of some rose hips, but usually these are parasite-free. As the weeks of August, September and October pass, and the host 46 INSECTS IN ROSE HIPS larva develops from the first to the third instar and transforms to the early pupal phase in the puparium, the larva of H. rosae persists as a first instar. However, during this span of weeks, this first instar slowly, gradually increases in size, and acquires a sharply different shape, yet basically identical form of body (Fig. 25). It has now added much bulk in the postcephalic region, while the head retains essentially its original form. The dorsal and caudal spines, as well as the general fusiform outline, attest to the fact that this is still the first instar. Its enlargement appears to result basically from a gradual inflation of the digestive tract to its full capacity; the dilated tract in turn swells the body wall and seems to occupy all the space of the body cavity. Why is the development of the first instar of H. rosae so slow as to require a period of time that corresponds in length to almost the total larval life of the host? By way of answer, I propose a theory based on the nutrition and mouth parts of the parasite. The three host instars inhabited by the first parasite larva contain a normal fat body of web- like form. It is possible that, owing to the structure of its mouth parts, the parasite cannot acquire the particles of the fat body as bound in the webbed state, and thus must utilize blood plasma and such host fat as floats free in the blood. This implies that the substance of the fat body is more effective than plasma and free fat as a promoter of growth, but is not available to the parasite owing to its type of mouth parts. Diapause. The first instar of H. rosae involves two life phases. The first is characterized by the function of feeding-growing, and occurs entirely in the host larva. This phase terminates with the full-fed state of the first instar, an event of late summer or fall. The second, or dia- pausing phase begins with the termination of the feeding-growing phase, and persists after the host larva pupariates in late summer to fall, and while the host winters in its early pupal phase. At approximately the time when the host resumes its pupation in May or June, the diapausing parasite finally molts to the second instar. The supposed precise con- currence of the molt and the resumption of pupation of the host sug- gests that the pupational hormone may trigger the molt of Halticoptera rosae. These phases of the first instar are emphasized to stress that the first phase is not a diapause, even though, due to the very slow and prolonged growth, it may superficially appear to be such. Phagocytes and the first instar. Whereas the young first instars of Opius are invariably enveloped in a gelatine-like substance, only a few Halticoptera were found so enwrapped, and these were not young in- stars. In seven years, I discovered, by dissection of hosts, only four such cases. Whereas the gelatinous substance enveloping Opius was never HYMENOPTEROUS PARASITES OF Rhagoletis basiola 47 fatal, that enveloping Halticoptera destroyed the parasite. All the four were found in September, and had developed to the advanced or dia- pausing state and occurred within living puparial hosts one to seven days of age, ie., either prepupae or early phase pupae. In three of the four cases, the parasites were dead, brown, somewhat shrunken, and completely enveloped with the gelatine-like material. The body of the fourth was smooth and turgid, the head and the anterior part of the thorax remained free and lifelike, and pulsations indicative of life per- sisted in the thoracic part of the body cavity. But a large part of the enveloped abdomen had turned black and was presumed dead. Since the phagocytes, presumed to be active in these instances, oc- curred in the host puparia, it is possible they may have been identified in some way with the process of histolysis which transpires incident to pupation of the host. But the fact that only a few of the parasites found in the puparia were affected in this manner indicates that the above victims of phagocytosis, or the phagocytes themselves, were un- usual in some way. “Milky” first instars. Only in the fall of 1951, I discovered a number of advanced first instars of Halticoptera at Eaglesnest, Minnesota, and Chetek, Wisconsin, which appeared normal except they were subopaque and diluted milky white. The affected parasites measured 0.6 to 0.9 mm. long, and varied from slenderish to plump in form, and inhabited pu- paria 2 to 25 days of age that contained prepupae and early phase pupal hosts. In the same year, many first instars from the same geographic sources were normal in color. Advanced larvae. As already stated, the first instar of Halticoptera is terminated with the molt that takes place in the spring season at about the time the host resumes its postwinter pupation. By now the adi- pose web of the pupal Rhagoletis is normally disintegrated, and appears creamy in texture, and the second and later instars of the parasite ob- viously possess a suctorial type of feeding organ which permits quick ingestion of relatively larger amounts of food than the first instar is able to take. This homogenized adipose may also be highly nutritious. At any rate, the parasite grows apace and attains its full size in probably only a few days, ie., in much less time than was required for the growth phase of the first instar alone. In the course of this rapid mechanical transfer of host substance into the parasite, the host is depleted and killed. The rapid rate at which the bulk of adipose is conveyed into the parasite suggests that it undergoes little chemical change at once, and that therefore the enlargement of the parasite does not constitute true growth. This physical dilatation is followed later by physiological ma- turity, for the parasite, first greatly distended with food, shrinks per- 48 INSECTS IN ROSE HIPS ceptibly in time as the ingested adipose changes from yellow to greenish, dull brown or purplish. These changes indicate that the nutrients are gradually absorbed from it, eventually leaving the indigestibles in the form of a dark contracted semisolid mass in the peritrophic membrane. Contraction waves. Upon opening a puparium of Rhagoletis in May, I discovered a second instar of H. rosae 1.55 mm. long, the actions of whose digestive tract were readily observable. In nine minutes and forty seconds, the walls of the tract completed 31 contraction waves. These ran the full length of the tract and thoroughly churned the adipose newly ingested from the host. The individual waves required from 10 to 40, but mostly 14 to 25, seconds to run their full course. Twenty-five of them proceeded from head to cauda; only six from cauda to head. Predomi- nance of the cephalocaudal waves seems to expedite the removal of food from the oral area into the midgut. Similar waves were also seen in an early first instar in the embryo of the host. The food in the digestive canal was probably derived from the parasite egg, and flowed alternately backward and forward as the larva shortened and lengthened itself by means of contractile movements in its postcephalic segments. Pupa. The meconium is discharged incident to the prepupal molt that discloses the newly formed white pupa. It appears that the full-grown larva does not spin a cocoon of silk, so that the pupa lies bare in the empty puparial shell of the host. Life cycle. H. rosae, like Opius baldufi, O. rosicola, and their host Rhagoletis basiola, completes one cycle or generation in a year. At Urbana 1,034 adults of H. rosae, reared from host puparia, emerged between June 2 and July 31 in the over-all period of 1946 to 1951. All originated in areas north of Ilinois. The retarding influence of northern climate on dates of emergence is reflected in the case of two lots of puparia which originated and win- tered at Eaglesnest, Minnesota. These yielded adult H. rosae from July 7 to August 26, 1946, or four to five weeks later than their counterparts wintered at Urbana. Adults were seen in the field at Eaglesnest between August 6 and 31, 1945, and eggs were found in the host embryos as early as June 30, 1946, and as late as September 6, 1945. At Eaglesnest, Minne- sota, Halticoptera rosae places her eggs into those of Rhag. basiola be- tween late June and late August. Embryonic development appears to require less than four days, and the larva persists, apparently making little or no growth, in the host embryo for varying lengths of time, since this host stage may have a life as long as 40 days. The first instar of H. rosae is a slow-growing parasite, developing from the newly hatched state to its much dilated mature phase while the host larva grows through HYMENOPTEROUS PARASITES OF Rhagoletis basiola 49 all its three instars; this, I have calculated, requires about 30 to 35 days. The larva of Rhag. basiola becomes full grown, emerges from the hip, and pupariates beginning about mid-August. After three to five days in the prepupal stage, the host enters its early pupal phase, a true pupa, and invariably enters into diapause. In the absence of parasites, the process of pupation is resumed and completed in the subsequent months of May to June. While the host winters as an early phase pupa, the parasite persists as a mature first instar. The next advance in the development of the parasite occurs in May to June. It then molts to the second instar; this molt appears to be concurrent with the resumption of pupation of the host, and, moreover, may be dependent on it in the sense that the host’s pupational hormone may also trigger the molt of the parasite. The development of H. rosae through all the instars following the first is completed rapidly, certainly in less than a week. These instars feed upon the adipose of the host pupa, which substance is now in disinte- grated, easily available form; this may be a major reason why these in- stars grow relatively rapidly. Thus, the host is killed as a pupa in late spring and early summer, and the matured parasitic larva pupates in the puparium. Accordingly, the total larval stage of the H. rosae utilizes three suc- cessive major metamorphic stages of the host—embryo, larva and pupa —in its development. Chronology of Rhag. basiola and its parasites. This statement is based on the dates of emergence of the adult Rhag. basiola and the adults of Opius and Halticoptera from puparia in the years 1946, 1947, and 1948. They were reared in approximately natural temperatures at Urbana, Illinois. All but a few of the host puparia originated in northeastern Minnesota. From the dates of earliest and latest emergence in the three years, it follows that the adults of Rhag. basiola issued between May 28 and July 10; those of H. rosae, between June 10 and August 10; and of Opius rosicola and/or O. baldufi between June 30 and August 6 and for- ward. While a considerable amount of intergeneric overlapping is indi- cated, this is not so significant as it appears to be, for the late dates commonly involve only very small fractions of the totals reared, or occasional stragglers. The large majority of each group emerged within narrower time limits than those indicated, and when viewed in terms of these majorities, the periods of emergence of the host, H. rosae and Opius spp. prove to be fairly distinct. The order of seasonal appearance of the adults is then as follows: the host, Rhag. basiola, H. rosae and Opius spp. H. rosae can begin its parasitization earlier than Opius, since it utilizes the host embryo, that 50 INSECTS IN ROSE HIPS may last forty days individually, while Opius oviposits in one of the subsequent larval instars of the host. The fact that Opius spp. diapause and winter as full-grown larvae, whereas H. rosae does so as an advanced first instar, may at first lead one to anticipate that Opius has a head start, developmentally speaking, and therefore might be expected to become adult decidedly before H. rosae. That this is, however, not the actual sequence of events provokes the theory that the chalcid early resumes its larval development in the spring due to stimulation from the pupa- tional hormones of its living host, whereas Opius, inhabiting the dead empty host puparium, is dependent on its own hormonal stimulation alone, and thus comes out second to H. rosae in its advance to adult- hood. Eupelmidae spp., Hymenoptera In the years 1946 to 1951, I obtained 13 small larvae that performed as solitary endoparasites in the three instars and puparia of Rhag. basiola. The similarity of the first instars to Clausen’s illustration (figure B, p. 198:1940) suggest they may be a kind of eupelmid chalcid. In the same years, I had a series of adults of a Eupelmus sp. from quantities of hips picked at Eaglesnest, Minnesota, and in Wisconsin, and Illinois. Whether these larvae and adults represent one and the same species was not established. Larvae from Rhag. basiola. The first instar (Fig. 47) of this supposed eupelmid differs noticeably from the first instars of Halticoptera rosae and Opius spp. Length 0.5 to 0.8 mm. Head pale brown, wider than long, broadly rounded in front, sides subparallel, about as broad behind as base of thorax; a transverse row of at least two pale spines on sternum of each of first six, or more, postcephalic segments; similar spines also on - terga of some segments; abdomen narrowing behind and terminating in a pair of identical, flexible, depressed pointed processes which curve ventrad and diverge strongly at their apices. Figure 48 illustrates what I take to be the second instar. Lengths ob- served were 1.2 to 2.0 mm. The head is concolorous with rest of body, short, bilobed in front; postcephalic region much dilated, broader and longer than that of first instar; abdomen sacklike, soft, with a caudal pair of short, pointed processes, one at each outer posterior angle. One such larva appeared to have 12 postcephalic segments, these being rather obscurely demarcated due to inflation of body with food. These parasitic larvae were found in Rhag. basiola that originated in northeastern Minnesota. Six living first instars occurred in host larvae of the first and second instars and in mature emerged larvae and in puparia one day old. The second instars inhabited puparia five days old. All the larvae were discovered, by dissection of hosts, between August THE ROSE HIP MOTHS, Cydia spp., OLETHREUTIDAE, LEPIDOPTERA 51 21 and October 3. Because the first instars appeared in first instars of the host, the female presumably oviposits either in them or in the eggs. Moreover, the persistence of first instars in one-day puparia, and the presence of seconds in five-day-old puparia suggests that the parasite may delay its molt to the second instar until the host enters its prepupal period. In other words, this is possibly another instance wherein the parasitic first molt depends on stimulation of a sort from the host, as was observed in Opius and Halticoptera in Rhag. basiola above. Eupelmus from hip jars. The 18 adult individuals obtained alive from hips originating at Eaglesnest, Minnesota, Chetek, Wisconsin, and two localities in Illinois, appeared in hip jars between August 16 and Sep- tember 30 of four years, excepting one (Taylorville, Illinois) which emerged on May 27. The two Illinois specimens came from localities and hips that rarely or never yielded the rose hip fly; this fact casts doubt on the supposition that Eupelmus sp. is the adult of the Eupelmidae? sp. larvae found in this fly. Tue Rose Hie Morus, Cydia spp., OLETHREUTIDAE, LEPIDOPTERA In the earlier literature, these small hip-inhabiting caterpillars are re- ferred to generally under the names Grapholitha, Tortricidae, but in later publications as Cydia, Olethreutidae. Accordingly, they are near relatives of the codling moth, Cydia pomonella (L.) and the oriental fruit moth, C. molesta (Busck), and resemble these species in mode of life, and especially in that they inhabit and ingest the pulpy parts of rosaceous fruits. Among the phytophags that inhabit rose hips, they are bionomically comparable to the Trypetidae, Rhag. basiola, Rhag. alter- nata and Carpomyia Schineri, in that they limit their activities as larvae to the flesh of the hip, thereupon emerging to pupate externally. Two species of Cydia are known to sustain such relations to rose hips, namely the American form, C. packardi (Zell.), now officially called the cherry fruitworm, and the European form, C. roseticolana (Zell.). Cydia roseticolana (Zell. ) The early synonymy is given in Zeller (1871), Eppelsheim (1871), and Staudinger and Rebel (1901). Zeller (1849) described C. rosetico- lana from a female obtained from Tuscany. The literature listed by Zeller (1871) records this species from Breslau, Baden, at Halle an der Saale, Nassau, at Glogau, Oesel, Schlesia, Meseritz and Stettin, Germany. Eppelsheim (1871) had it from Griinstadt, Germany, and Barrett (1872- 73) recorded it as Endopisa roseticolana from Britain. A more thorough investigation will probably show it occurs widely in northern Europe. UNAERSITY OF BLINDS LIBRARY 2. INSECTS IN ROSE HIPS Little appears to have been published pertaining to the life history. The principal observations were made within the quarter-century after Zeller (1849) named it in the genus Grapholitha. Zeller (1871) and Eppelsheim (1871) made known such basic facts as we have today. Kaltenbach (1874) quoted the observations of Eppelsheim in his book on insect enemies of plants. C. roseticolana is the only insect from rose hips treated by Liistner (1931) in an extension circular dealing with the most important enemies of rose in Germany. At Griinstadt, Eppelsheim found the larvae as early as the middle of August in the fruits of Rosa canina, a scattered but omnipresent plant. The larvae eat their way into the hip just below the crown at the base of the calyx. The point of entrance is easily recognized as a small black spot. While they are small, the mines are indicated as dull pale brown areas in contrast to the undamaged parts of the hips. As they grow, they gradually destroy the hip flesh; hence the surface wrinkles, sags and becomes dark-spotted. In August to September, 1870, Eppelsheim se- cured 400 to 500 inhabited hips, from which the mature larvae emerged later that year. According to earlier reports by Koch, v. Heinemann and Rossler, the mature larvae were supposed to enter the ground for pu- pation; but, seeing one boring into the end of a dry rose stem in the field, Eppelsheim gathered dry stems of Eryngium campestre L. to further test his field observation in the laboratory, and soon was gratified to see that numerous larvae entered these stems. They prepared burrows 1.5 times the length of their bodies and just wide enough to permit them to turn end for end. The entrance was walled off from the outside with a silken web. At Meseritz and Stettin, Zeller (1871) also found the larvae in the hips of R. canina, but not in “rose galls,” as Scholz claimed (Zeller, 1849). About the middle of September, he reared many larvae from red rose fruits. Black patches on the fruits indicated inhabitation by larvae. These remain in the hip until full-grown. Much excrement is voided into the burrows, which is at first light red, later, when dry, red-brown. He also observed that they do not eat the achenes of the hip; ie., they confine their activity to the hypanthium. Upon emerging from hips in a rearing jar, some larvae spun themselves in with silk, but most of them dried up, which indicated that the natural cocooning site was lacking. The mature larvae are 8 to 9 mm. long and pinkish. Eppelsheim re- ported seeing dark spots on the dorsum of the eighth body segment. I presume these were the incipient testes of the adult male moth. The larvae observed by Eppelsheim wintered as such, and pupated in late winter and early spring, in an artificially warmed room. The chrysalises, excepting the last abdominal segment, emerged from the THE ROSE HIP MOTHS, Cydia spp., OLETHREUTIDAE, LEPIDOPTERA 53 -cocoons shortly before the moths issued, from April forward. Zeller (1871) reported the same movements of the pupae, and also the emer- gence of one adult on May 13. He adds emergence records for May 22, and also June 6, and presumed the natural flight period begins in the last third of May, and had observed it to extend throughout June. The above data indicate that Cydia roseticolana has one generation in a year where observed, with the mature larvae as the wintering stage. I have seen no records of parasites attacking this species. Cydia packardi (Zell.), The Cherry Fruitworm Synonymy. Owing to its damage to the fruits of several cultivated plants, C. packardi, named and described by Zeller (1876), has been comparatively well investigated and a considerable body of facts made known concerning its mode of life and host relations. Heinrich (1926) summarizes the principal synonymy of the species. Previous to 1950, the bibliographies largely employed the name Grapholitha packardi Zell., but Cydia packardi ( Zell.) in the present decade. Sanderson (1900, 1903) called it “the apple bud-borer.” However, the larva appears to prefer fruits to buds or twigs, and currently is rated as a pest in culti- vated cherries. While wild cherry is clearly one of its principal native hosts, the hips of wild roses rank prominently as native food plants. Distribution. Between 1945 and 1951, some thousands of individuals of C. packardi were obtained by dissection of immature larvae from rose hips, and by rearing mature caterpillars in hip jars and some adults from cocoons. Most lots of the source hips were secured personally, others through collaborators. By these means, C. packardi proved to occur in rose hips in the following areas: northeastern Minnesota, from the north to the south limits of Wisconsin, numerous localities of northern and east central Illinois, several widely placed points in Indiana and Ohio, in Guilford County, North Carolina, Geneva, New York, Nebraska, Mis- souri, Oklahoma, Kansas, Idaho, Washington, and Saskatoon, Saskatche- wan. The literature affords the following distribution records, few of which have reference to the species in rose hips: Texas, Mississippi, Missouri, Arkansas, Illinois, Michigan, Maryland, West Virginia, Virginia, New Jersey, Delaware, Massachusetts and New Hampshire (Heinrich, 1926); New York (Forbes, 1923); British Columbia, Canada (Downes, 1929, 1929a); Washington (Breakey and Webster, 1939, Newcomer, 1950); Iowa (Richardson, 1943); Colorado (Hoerner and List, 1952); Wis- consin (Dever, 1953); and Michigan (Vergeer, 1954). Cydia packardi has a wide distribution in North America, the extreme points of its recorded occurrence being British Columbia, Saskatchewan, 54 INSECTS IN ROSE HIPS Washington, Texas, Virginia and New Hampshire. Most of the records refer it to the region north of the latitude of the Ohio river. Rose hosts. In northeastern Minnesota, the larvae of C. packardi were found to inhabit the hips of Rosa acicularis vars. and R. blanda, includ- ing Macounii; in Wisconsin, the rose hosts were R. blanda vars., R. arkan- sana, and R. carolina vars.; in Illinois, usually R. carolina, and sometimes arkansana, palustris, virginiana, eglanteria, canina, setigera, but not hu- gonis; Indiana, R. carolina and palustris; Ohio, R. palustris, rubifolia and R. spp.; Nebraska, R. Macounii; Missouri, R. carolina; Kansas, R. suffulta; Idaho and Washington, R. spaldingii; Saskatchewan, Macounii and lunelli. C. packardi has the same general range and rose hosts as Rhag. basiola. Adult. The moth is dark grayish to brownish, with a wing expanse of about 10 mm. In describing “Grapholitha” packardi Zell., Heinrich (1926) states that “the adult is smaller than that of G. molesta and has a faint but distinguishable dark median fascia on the forewing. Its most striking character is a strong patch of blackish sex scaling upon the upper surface of the hind wing and a similar patch on the under surface of the forewing of the male. This character, so far as I know, is shared by no other North American species of Grapholitha or Laspey- rezia. Forbes (1924) reported the adult from New York, under the name “Laspeyresia pyricolana Murtfeldt.” Egg and oviposition. The egg is 0.5 mm. in diameter, circular in out- line, depressed, slightly convex, opalescent to yellow, the contents turn- ing grayish white as they develop (Dever). I have seen a few empty pale grayish white chorions on rose hips. Their scarcity suggests that they may be ingested by the newly hatched larvae, or fall from the hip after hatching takes place. The moth obviously prefers natural depres- sions for oviposition. On rose hips, the eggs were usually placed in the rounded, necklike constriction at the apex or floral end of the hip. In the case of cherries cultivated in British Columbia, the eggs are laid on the fruit, usually in the suture near the apex, but some at the basal or stem end (Downes, 1929). In Colorado, Hoerner and List (1952) found them deposited often in the suture or any rough spot on cherries, and sometimes around the stem, or on the stem close to the fruit. In Wisconsin, Dever (1953) usually discovered them “next to the suture at the base of the petiole or at the calyx and next to the pistil scar” of cherries. Observing this species in Michigan, Vergeer (1954) reported the eggs “often cemented in the groove between the sepal and the cen- tral disk of small blue berries.” In contrast with the above observations that the eggs are laid on fruits, Garman (1918) reported they are placed on the upper and under side of the leaves of apple, in which case the resulting larvae bore into twigs. THE ROSE HIP MOTHS, Cydia spp., OLETHREUTIDAE, LEPIDOPTERA 59 Larva. The descriptions available in the literature presumably pertain to the mature larva. Murtfeldt (1891) described it in general terms, and Dever (1953) provides an illustrated technical account, including the setal pattern. In his article of 1954, Dever included setal maps and also the following verbal description: “Mature larva—general color whitish- pink, ventral surface not so pink as dorsal; head, from mottled yellowish- brown to dark brown, shiny; mouth parts, pale gray, almost white; an- tennae, white at base, slightly darker towards tip; thoracic shield, shiny, pale yellow to light brown, divided in the middle by a longitudinal paler line; thoracic legs, white; prolegs white, crochets uniordinal, uniserial, and in a complete circle; crochets on anal prolegs uniordinal, uniserial, and in a transverse band; setae on grayish-white pinaculae; anal shield, mottled grayish-black; anal comb, dark brown with four to six prongs of irregular length.” By dissecting buds and hips of Rosa, I obtained evidently newly hatched larvae hardly more than 1.0 mm. long, and pale in color. Ac- cording to Vergeer (1954) larvae of the first stage are 2.0 mm. long, white, with a dark brown head; the head, thoracic shield and the anal plate become smoky brown as the larva grows, and at maturity it changes to bright reddish pink above while remaining decidedly lighter be- neath. The full-grown larva bears an anal plate of three separate parts, and attains sizes varying from 5.0 to 9.0 mm. in length. From rose hips, I reared some female larvae 10.0 mm. long, and also individuals only 5.0 mm. long. The smaller sizes seem to result from undernourishment in unusually small hips, in deteriorating hips in the field or jars, and in hips shared by other larvae of their kind, or Rhag. basiola. I have no direct evidence that endoparasitism alone produces larvae of the smaller sizes. The pinkish coloration indicates impending larval maturity. Among the larger larvae, the males may readily be distinguished by the paired clusters of reddish to purplish, wine-colored gonads situated in the dorsal area of the fifth abdominal, or eighth postcephalic, segment. Each cluster comprises four filbert-shaped bodies appearing attached to to each other by their edges. In one instance, among many gonads ob- served, one cluster of the pair consisted of five bodies, indicating that the testicular tubes of the adult may be subject to slight numerical varia- tion. The number of instars comprising the larva appears not to have been determined. Molts may be anticipated soon when the diameter of the thorax strongly exceeds the width of the head; on the other hand, a recent molt is indicated when the head and thorax are about equal in width, and the body color remains more or less pale. Food and niches of the larva. Whatever their food plants, all the instars of C. packardi exist as borers, seeming to enter the plant directly 56 INSECTS IN ROSE HIPS from the egg, and live invisibly within buds, shoots or fruits of the host, and appear not to feed first externally on leaves. Heinrich (1926) listed apple, rose and Crataegus as the food plants, and rated packardi as “a species of some importance as an enemy of apple... . It usually attacks the growing tips, seldom the fruits. It has also been reared from rose tips and there are a couple of doubtful records from peach.” However, it appears now to be more frequent and important eco- nomically in the following fruits: Cherry, particularly cultivated sour varieties, in British Columbia (Downes, 1929, 19292); Washington (Breakey and Webster, 1939; Baker, 1930; Newcomer, 1950); Colorado (Hoerner and List, 1952); Wisconsin (Dever, 1953, 1954). Blueberry (Vaccinium ) in New Jersey (Lipman, 1936; Tomlinson, 1951); Michigan (Vergeer, 1954). Apple, in Missouri (Murtfeldt, 1891); Delaware (San- derson, 1900, 1903); Maryland (Garman, 1918); Iowa (Richardson, 1943). I have found it to be a major species in the hips of roses in Minne- sota, Wisconsin, [linois, Indiana, and Ohio. The larva is unique among the major phytophags in rose hips in that it also utilizes rosaceous fruits other than roses, and also blueberry. Larva in apple (Apple bud-borer). According to Murtfeldt (1891), whose record appears to have been the first for apple, the larva “bores the shoots of the second growth of apple in August and September, occasionally on recently planted trees, inflicting serious damage. The larva spins scarcely any web, but bores downward through the terminal bud, entering the stem from half an inch to an inch, sometimes blacken- ing all the growing points of a young tree.” The mode of attack described by Sanderson (1900) is similar. On hatching, the larvae mine through the outer terminal leaves and bore down into the twig often for a distance of one or two inches. Then the twig dies, and new shoots grow from the first lateral bud below it. This new shoot is attacked, and so on, one new growth after another, until the twig has been badly stunted and given a zigzag, knotty form. Tender “water sprouts” on trunks of old trees are generally infested. Garman (1918) adds that the larva infests “apple fruit and twigs; peach, twigs,” but the growing shoots of apple are its favorite food. On the basis of observations made at Bevington, Iowa, Richardson notes that the larvae seem to bore only in the young growth of twigs on young trees. Larva in blueberries. In New Jersey, the larva of C. packardi com- monly infests the early, immature fruits of cultivated blueberries (Lip- man, Tomlinson), but it varies in importance from field to field and year to year. The larvae feed in several berries in the cluster before THE ROSE HIP MOTHS, Cydia spp., OLETHREUTIDAE, LEPIDOPTERA 57 maturing, so that only a few worms may destroy many berries. Vergeer informs us that they spin the berries together when they move from one to another. He notes further that they occasionally make holes to the outside of the fruit in addition to that by which they enter, but keep only one open, covering the others with silk spun transversely over the opening. The newly hatched larva eats into the fruit through a small hole difficult to detect. The location of these “pinprick” holes indicates that “the great majority of larvae enter the berries on the sides nearer the calyx than the stem end. Some enter the calyx next to the sepals.” Larva in cherries. As described by Hoerner and List (1952), the en- trance holes made by the newly hatched larvae into the fruit of cherry are “very minute and usually have a small amount of frass present around the opening. The larvae tunnel through fruit around the stone.” As Downes (1929, 1929a) expressed it, the larva bores into the fruit and mines between the flesh and pit of the cherry. He found the native host in British Columbia to be wild cherry. Larva in rose. Forbes (1924) reported that the larvae of the “spring brood” bore into the buds of roses and blast them. This appears to be the first observation of the larva in rose buds. Heinrich (1926) names rose “tips” as one of the larval food plants. Since I found numerous larvae in the buds and hips of various wild roses, I conclude that rose is evidently another of the native American host plants. In east central Illinois, the yearly development of C. packardi in- volves both the buds and hips. No less than two generations are pro- duced here, the first or spring cycle predominantly in the buds, the last largely in the hips. Larva in rose buds. The behavior of the larvae in relation to rose buds was observed many times, particularly at Urbana. Upon hatching from embryos that developed on the neck of the incipient hip, the small larvae entered the lumina of the buds in one of two places—at the neck or at the side of the bud itself. In a random sample, 70 larvae had entered by way of the neck, and seven bored in through the sides, usually in the basal two-thirds of the bud. When entering at the sides, they some- times penetrated only the corolla exposed between the sepals, and less frequently perforated both sepal and the underlying petals. The initial entrance is oval and about 1.0 mm. in diameter but is enlarged for sub- sequent exit of the mature larva. Buds inhabited for a day or more by the larvae are readily and re- liably identified by the small mass of pale brown to black fecal frass that accumulates externally at the entrance. While some of the feces is expelled through the aperture to the outside, the bulk of it produced during larval development is deposited in the lumen of the bud. 58 INSECTS IN ROSE HIPS The rose bud constitutes a nearly ideal situation for the development of the larva. It affords darkness which this evidently negatively photo- tropic larva requires, and a diversity of food material in the form of floral structures. This bud lumen also is the place where the larva molts. Especially the discarded head capsules may be found mingled with the silk-bound vegetable and fecal frass, those of the first instar sometimes lying at the entrance to the neck of the hip. Usually the larva is found lying curled on the floor of the bud within the circular space between the mass of short pistillary tubes and the bases of the filaments and petals. In this position, the mandibles can readily reach the bases of the stamens and corolla, which parts consti- tute much of the food. The pistils, a clump of short, threadlike structures, seem not to suffer extensive, if any, attack. The pollen from the anthers is ingested in large amounts, for the digestive tract is often packed with yellow particles. Pollen may become available to the larva when the anthers fall to the bottom of the bud lumen after the filaments are sev- ered. In this connection, it is of interest that the larva appears sometimes, inadvertently, to perform the role of pollinator. In several instances, I discovered that the hip contained normal numbers of achenes despite the fact that larvae had severed the corolla from the receptacle and pre- vented its unfolding so, as a consequence, the anthers were not available to the usual external pollinizing agents. The larva may chance to effect pollination when it chews off the filaments of the stamens whose anthers then fall upon the pistils. Rose bud damage varies both in rate and kind; up to 65 per cent were found infested at Urbana. Small larvae that migrate to other buds while still immature cause little damage to the buds first occupied. In other cases, when the larva severs only some of the petals, the flower opens but has an asymmetrical shape. Often destruction is complete, as when all the petals are cut at their bases from the receptacle. The coro!la then retains its normal conical shape and remains insecurely attached to the incipient hip, but turns gradually from the normal living red to dull red to dry brown, shrinking as it dries, and eventually falls off. Such dry conical buds are easily lifted intact from the hip. Migration of larvae from bud to bud, or bud to hip, is probably not infrequent, even though one rarely discovers the larva in the act. Indirect evidence of such movements is often seen. I have found numbers of buds with small entrance holes but meager injury typical of this species, which can mean only that the inhabitant was small and had migrated elsewhere. Again, a bud may be inhabited by an advanced larva, which had fed but little and had deposited only little excrement, a circumstance indi- cating recent advent of the larva and its earlier growth in another bud. THE ROSE HIP MOTHS, Cydia spp., OLETHREUTIDAE, LEPIDOPTERA 59 In one such case, the larva was 8.0 mm. long and pink, submature, and little excrement occurred with it. Larva in hips. In late June and early July, when budding and flowering are at an end, the larva invades the rose hip, which has now attained various sizes up to full-grown, yet is hard and unripe. I believe that some of the earliest inhabitants of the hips may be migrant, partly grown larvae that initiated their development in buds. But the prevalence of small larvae in the area of the neck and shoulder of hips shows that the major- ity of the larvae found in hips during July originated directly from eggs laid there. Commonly small larvae are found in incipient burrows at the equator of the hip, perhaps having entered between two hips hanging in contact. A small number occurs in the thick basal or petiolar end where they probably enter via the concavity around the stem. A few larvae gain entrance into the hypanthium through egg pits prepared by Rhag. basiola. When inhabiting hips, the larvae of C. packardi are almost always confined to the hypanthium. Although usually residing so deep as not to be visible externally, some larvae bore along just beneath the skin so that the pattern of the burrow can be traced from outside. Rarely the larger larvae had eaten through the floor of the hypanthium, and some fecal matter lay on the exposed achenes. However, the achenes were not eaten. Such exceptional cases were discovered, mostly in the hips of Rosa carolina, and seem to occur in hypanthia so shallow that pressure of the ceiling of the burrow on the dorsum of the larva may have stimu- lated it to chew into the floor in order to gain space requisite to its life and avoid exposure to light. The passages made through the flesh of the hip by the smaller larvae are narrow, nearly cylindrical tunnels. Their diameter increases gradu- ally as each instar develops, but enlarges abruptly as the borer molts between instars. At some point in the more advanced phase of the larval stage the burrow is broadened without corresponding elongation. This the larva does by feeding at the lateral edges of the passage, thereby producing a broad, somewhat rounded blotch or mine. Superficially these cavities are recognizable by the dark color of the overlying hip skin, and closely resemble the dark blotches made by the advanced larvae of Rhag. basiola, but are readily distinguishable from the latter by their content of pellet-like excrement. That of the maggot is a homogeneous, semifluid brown paste. Moreover, the cavities of packardi are usually provided with one to three small apertures in the skin above the run- ways, the small first one at least being usually plugged with fecal pellets. Disposal of the excrement, by expelling it through the holes, is appar- ently practiced to some extent also by submature larvae, as is shown by the large heaps of fresh fecal pellets frequently seen in hip jars. How- 60 INSECTS IN ROSE HIPS ever, much waste is voided directly into the burrow, for the older sectors of mines still inhabited, and deserted ones as well, are packed with ex- crement. The inhabited sectors of mines are devoid of feces and remain clean. When first eliminated, the pellets are rich reddish brown, but later turn dark brown to black, and then also are commonly overgrown with the mycelia of a fungus and apparently bound with silk. The pattern or design of the system of burrows made by the larvae in the hips seems to assume various forms. However, this variety may be more apparent than real, so one asks whether there may be a single basic design. Perhaps, were the observer to note carefully the position of the hip on the rose bush, and then consider the design from that view- point, he might discover that the direction of the cavities with reference to the position of the hip bears a constant relation. Several factors, such as the direction of the sunlight, the pull of gravity, and the texture and odor, bitterness or sweetness of the hypanthium may be found to com- bine to produce the pattern. It is perhaps such complexity of interacting stimuli on insects that accounts for the inconclusiveness of our knowledge in the area of insect behavior. Oral excretions. Mature emerged larvae sometimes emit sizable amounts of heavy, reddish brown fluid through the mouth when dis- turbed. It appears that the larvae concerned had fed on decomposing hypanthium in their final days in the hip. When irritated with a needle, the larvae produced small quantities of the fluid at each of several suc- cessive stimulations. The principal interest in these cases lies in the observation that the amount of dark substance present in the fore part of the digestive tract lessened visibly at each stimulation, proving that the substance originated in this tract, and not in the salivary glands. The contents of the glands are white and of small volume in the full-grown larvae of C. packardi. Hibernation. Investigators are in general agreement that, whatever the host plant is, the cherry fruitworm spends the winter as a mature larva. Upon leaving the fruit or branch in which it develops, the larva constructs a protective hibernaculum, whose composition and location are subject to some variation, as is indicated by the following notes from the original literary sources. Sanderson (1900) observed the larvae to winter in burrows in the “terminals” of apple. Their presence is often indicated by a leaf petiole remaining on the tip of a twig, the larval cell being present at the base of the petiole. Occasionally a case is formed on the trunk or branch of a tree. Hibernacula were composed of silk covered with bits of bark and dirt. Downes (1929) reported that the mature larvae leave cherries to THE ROSE HIP MOTHS, Cydia spp., OLETHREUTIDAE, LEPIDOPTERA 61] search for a winter place and bore into dry twigs or bark to a depth of one to 1.5 inches. Stubs of pruned branches are preferred. Breakey and Webster (1939) stated the species hibernates in the stubs of twigs of sour cherry. Since they found no hibernating pupae, they deduced that the larva alone passes the winter. Also in Colorado, where Hoerner and List (1952) observed the species, the full-grown larvae find winter shelter in the pruned twigs of cherry, under the bark, in the stems of weeds or grasses or in the ground. In the pruned stubs they tunnel the pith to a depth of 0.5 to 1.25 inches. They then line the excavated cavities with silk, and plug the opening where they remain through thegwinter. Dever (1953), in his work in Wisconsin, found that the mature larvae pass the winter, constructing their winter quarters by spinning a nest under a piece of loose bark, in a roughened stub of a broken branch or twig, or they may bore into the stubs of pruned branches. “It appears that any crevice, bark curl, or similar de- pression which will serve for protection of the overwintering larvae are utilized.” The hibernacula have a light gray silken cover. This is easily removed, and then reveals two additional layers, which are very black, quite tough and brittle. Describing the wintering of the cherry fruitworm with reference to blueberry, Vergeer (1954) states the larvae wander about on the bushes and commonly enter the pith of pruned twigs, tunneling as much as two inches, but usually less. “The opening in the end is closed smoothly and tightly with silk so that it normally escapes observation. They also encase themselves in cracks of the bark lower down on the bushes or even in plain sight on the outside of twigs.” Some mature larvae from rose wintered successfully in jars, where they spun their cocoons among hips or under the cloth covers on the jars. The observed high rate of winter mortality could, in view of the above re- ports, probably have been avoided by providing plant stems for the larvae to enter. Pupation. After wintering in the hibernacula described above, the mature larvae of C. packardi transform through the pupal to the adult form. The particulars of the process of transformation have not been made known. The dates and duration of pupation are given in the fol- lowing section. Life cycle. According to published statements, the number of genera- tions that develop in a year varies from one to three, or four. The re- ports from British Columbia (Downes), western Washington (Breakey and Webster), Door County, Wisconsin (Dever), and Grand Junction, Michigan (Vergeer ), and my observations in northeastern Minnesota and northwestern Wisconsin, show that the species undergoes its life cycle 62 INSECTS IN ROSE HIPS once per year in these areas. According to Hoerner and List, there is ordinarily one generation in a year in cherries in northern Colorado, but they found indications that a partial second occurred in 1946. Forbes (1924) reported three “broods” for New York, these appearing in May, June, and August, the “spring brood boring into buds of rose and blasting them; the later broods normally feed on the leaves. Similar habits on apple and peach” (Forbes). Sanderson (1900) gives data which seem to him to indicate four “broods” in a year in Delaware. The time data which I obtained from dissections of immature larvae from hips and rearing mature larvae and moths, strongly indicate there are two genera- tions yearly at Urbana, Illinois, and vicinity. One generation per year. In British Columbia (Downes), pupation occurred during the second week of May, when the transformation of a series of individuals required an average of 42 days. The adults emerged from late May to mid-June. In 1929, the height of emergence was reached on June 23 to 24. In Door County, Wisconsin, pupation took place “in the spring,’ and adults were trapped from July 6 to 26, 1950; June 12 to July 1, 1951; and June 9 to 18, 1952 (Dever). The adults therefore begin to appear there two to four weeks after the cherry petals fall, and continue for 14 to 21 days. Pupae were not uncommonly seen projecting from the hibernacula just before and after the emergence of the adults. At Grand Junction, Michigan, the overwintered larvae pupated early the next May in the hibernacula, and the adults emerged over a six- week period that began during the last week of May, but did so mostly during the first three weeks (Vergeer). In Colorado (Hoerner and List), the wintered larvae pupate the following May, the average duration of the pupal stage there being 29 days. As determined by the use of bait traps, the first moths appeared near the first of June. In 1947 and 1949, the peak of emergence was on June 15, and in 1948 it was reached on June 7. Moths were captured as early as May 17 and as late as July 7 in northern Colorado. In British Columbia, eggs were laid as early as June 15 in 1927, and the first ones on June 24, 1929. Incubation required 10 to 11 days, and the resulting larvae reached maturity in about 30 days (Downes). In Colorado, eggs were readily found during the second half of June in heavily infested cherry orchards. Under laboratory conditions, a few eggs hatched in seven days. In an average season, practically all the larvae have left the fruits by the time the crop is harvested in the latter half of July (Hoerner and List). Dever found that the larvae in cherries complete their development in about three weeks, and that some in- dividuals attack several fruits before they become full-grown. His work in the years 1950 to 1952 showed the larvae sometimes mature and leave the fruit before cherry harvest in late July to early August, but in other THE ROSE HIP MOTHS, Cydia spp., OLETHREUTIDAE, LEPIDOPTERA 63 years, some still remained in the ripe fruits. Near Fruitport, Michigan, Vergeer examined 207 blueberries which bore evident damage by insects on July 8, 1953, and found only one such berry that still contained a cherry fruitworm. He states in effect that the larvae become full-grown toward the end of June and the first part of July, and do not develop beyond this condition until the following spring; i.e., they undergo a prolonged diapause. Although buds and flowers of Rosa acicularis and blanda were still present in July in northeastern Minnesota, my notes contain no mention of larvae found in buds. Dissection of unripe hips produced the first larvae on July 21; these were small, 2.0 to 3.0 mm. long. Examination of sample hips at intervals during the months of August and early Septem- ber in 1945 to 1951 revealed larvae varying greatly from first to last in- stars, and showed also that some had already matured and emerged from the hips by mid-August. Samples of ripening hips from Eaglesnest, kept in jars, yielded mature larvae up to October 15 at Urbana. These records indicate that C. packardi in rose hips begins its larval life in the first half of July, and that full-grown larvae appear from mid-August to mid- October to enter their winter diapause. All the lots of hips which originated in northwestern Wisconsin were picked and placed in jars at Urbana during the first third of September, 1945 to 1951. Some full-grown larvae issued within a day after the hips were picked, showing that emergence, particularly at Chetek where the majority of hips was obtained, began in August. Emergence continued into October from four of the five annual samples, and, in the extreme instance, to October 28, 1946. Larvae emerged as late as November 2 and 5, 1947, from hips taken at Madison and Black River Falls, Wiscon- sin, respectively. Two generations per year. In east central Hlinois, Cydia packardi com- pletes its life cycle twice in a year, the first largely in rose buds, the second in rose hips. This fact is clearly indicated by the dates of emer- gence of the adult moth and the times of occurrence of the early and late, or extreme, instars. Adults, which developed from mature larvae in captivity, emerged at two distinct seasons of the year, namely May 3 to 28, and June 29 to July 29. My data show that the maturity and emergence of larvae from rose buds and hips occurred about June 5 to mid-July, and again be- tween August 22 and December 8. I do not pretend to assign the larvae of July 7 to August 16 as to generation; they may have been late mem- bers of the first generation or early comers of the second, for there is doubtlessly some overlapping in the larval stage of the two cycles. Early instars appeared in two distinct time periods. The first to third instars predominated numerically from June 5 to 20 and again from July 64 INSECTS IN ROSE HIPS 7 to August 16. The low incidence of larvae in hips on June 27 to July 6 indicates this was the preovipositional and ovipositional interval between generations. The seasons of the embryonic and pupal stages may be approximated by inference from the dates of larval and adult stages. Pupation occurs in late April forward, and during parts of June and July; oviposition largely in May and July. Parasites oF Cydia packardi The following parasitic Hymenoptera have been reported in the litera- ture. In Delaware, the hibernating larvae of C. packardi were para- sitized, probably to the extent of 50 per cent, by the braconid, Bracon mellitor Say (Sanderson, 1900). List (1932) reared the chalcid, Eury- toma tylodermatis Ashm. from the larva in Colorado, and Putnam (1935) reported the chalcid, Psychophagus omnivorus Walk., a braconid, Asco- gaster carpocapsae Vier., and an ichneumonid, Glypta rufiscutellaris Cress. from this host in Ontario. About one per cent of the cherry fruit- worms collected by Hoerner and List (1952) in 1948 were parasitized, mostly by Ascogaster quadridentatus Wesm., a few by a species of “Bassus.” Driggers (1932) found that chalcids of the genera Eurytoma and Eupelmus were apparently hyperparasites of G. rufiscutellaris in the cocoons of Cydia molesta (Busck) in New Jersey. I have discovered several species of hymenopterous parasites and two genera of clerid beetles attacking the larva of C. packardi in rose buds and hips. My observations, combined with information from the litera- ture, are given below for these species. Glypta rufiscutellaris Cress., Ichneumonidae, Hymenoptera This native North American parasite was investigated extensively be- tween 1925 and 1945 as a result of the discovery that it constituted some- what of a check on the newly introduced oriental fruit moth, Cydia mo- lesta. The principal contributions pertaining to its immature stages and bionomics are by A. W. Crawford (1933) and Montgomery (1935). Glypta rufiscutellaris (Figs. 28, 29-32) was named and first described by E. T. Cresson in 1870 from specimens of both sexes which originated in Connecticut, New Jersey and Illinois. The original description states that the rufous scutellum and the black mouth readily distinguish it from its relatives. Distribution and hosts. G. rufiscutellaris occurs from the Atlantic Ocean west to Alberta, Canada, and to Utah, in the Transitional and Upper Austral faunal zones (Muesebeck, et al., 1951). I have found it present in all the main areas in which the insects of rose hips were in- PARASITES OF Cydia packardi 65 vestigated, as follows: Eaglesnest, Minnesota, north shore of Lake Su- perior between Little Marais and Grand Marais, Minnesota; Solon Springs, Gordon, Rice Lake, Chetek, Endeavor, and Madison, Wisconsin; at several rose stations in Urbana, at Mayview and Philo, and at Velma (near Taylorville), all in Illinois. The members of the braconid subfamily Banchinae, which includes the genus Glypta, are internal parasites of caterpillars (Muesebeck, et al., 1951). G. rufiscutellaris appears to select small lepidopterous hosts which live enclosed in such parts of plants as fruits, slender stems, husks on nuts, and leaf rolls, where penetration of the protective plant wall by the ovipositor is necessary to reach the host within. A native host frequently reported for rufiscutellaris is the ragweed borer, Epiblema strenuana Wlk. in the stems of Ambrosia artemisiifolia L. The adult has been reared from this host at St. Davids in the Niagara fruitbelt of Ontario (Putnam, 1935; van Steenburgh, 1934; Boyce, 1946), in Connecticut (Friend, 1942), western New York (Daniel, 1930), New Jersey (Driggers, 1930), Delaware (Rice, 1935), Virginia (Bobb, 1939), Ohio (Stearns and Neiswander, 1930; Blackburn, 1944) and Indiana (Montgomery, 1933). Bobb reared it also from Epiblema otiosana Clem. in Bidens bipinnata L. and Xanthium canadense Mill. in Virginia. Investi- gators have found the ragweed borer to be a valuable alternate host of rufiscutellaris as a factor in the control of the oriental fruit moth. For this reason, several eastern entomologists have advocated that the rag- weed be permitted to grow in and near peach orchards. Other native hosts of G. rufiscutellaris recorded to date are: Laspey- resia caryana Fitch in the husks of Carya ovata ( Mill.) K. Koch (Putnam, 1935); the lesser apple worm, Grapholitha prunivora Walsh (Putnam, 1935); Cydia (Graph.) packardi in the fruit of cultivated red cherry (Putnam, 1935); Argyrotoxa semipurpurana Kearft. on Quercus rubra L. and Q. coccinea Muench. (Putnam, 1935). All these records by: Put- nam pertain to St. Davids, Ontario. Putnam (1938) adds other native hosts of G. rufiscutellaris in Ontario as follows: Epiblema scudderiana (Clem. ), a gall-former on stems of goldenrod; Epiblema obfuscana Dyar, boring in stems of goldenrod; Melissopus latiferreanus Whshm., in acorns of red oak, and a leaftier on Crataegus L., tentatively identified as Epin- otia laracana Kearft. Garman (1930) reared this Glypta from the straw- berry leaf roller, Ancylis comptana Froel., in Connecticut. From Poly- chrosis carduana Busck, Blackburn (1944) obtained small numbers of it that bored in the tips of thistle in Ottawa County, Ohio, and he also obtained it from the goldenrod gall moth, Gnorimoschema gallaesolida- ginis (Riley). In California, Mackie (1944) found that the potato tuber moth, Gnor. operculella ( Zell.) was useful as a host for rearing this para- site. In Washington, Oregon, and western California, Dohanian (1942) 66 INSECTS IN ROSE HIPS discovered it attacking the filbert worm, Cydia (Mellisopus) latifer- reana (Wlsm.) in hazelnuts (Corylus). Additional hosts of G. rufiscutellaris are named in the Synoptic Cata- log (Muesebeck, et al., 1951), whose original literary sources I have not unearthed. These are the codling moth, Carpocapsa pomonella (L.), and Proteotera aesculana Riley. I have several times obtained Glypta rufiscutellaris from Illinois, Wisconsin and Minnesota, as a parasite of Cydia packardi in the buds and hips of roses. Also many rearing records of G. rufiscutellaris relate to the oriental fruit moth, Cydia molesta, as a host, and usually it is rated as second only to the imported braconid, Macrocentrus ancylivorus Rohwer as a natural check on this pest. In this relationship, it is known both from the Niagara peninsula and the southwestern area of Ontario, where peach is grown (Boyce, 1946). It is regarded as significant also in Mary- land (McConnell, 1932), New Jersey (Driggers, 1940), Delaware (Haden, 1935), Pennsylvania (Anon., 1934), western New York (A. W. Crawford, 1933), Virginia (Bobb, 1939), Ohio (Stearns and Neiswan- der, 1930), Michigan (Merritt, 1933, 1933a) and Missouri (Wingo, 1940). Butler (1933) reared a few individuals from C. molesta in Rome County, Tennessee. Six colonies of this parasite were liberated by agents of the Bureau of Entomology, U.S. Department of Agriculture, in peach orchards in Georgia and the Carolinas (Marlatt, 1931). It has also been exported from the United States to foreign countries, including Japan (Ishii, 1940), Goulburn Valley, Victoria, Australia (Anon., 1938), Uru- guay (Lepage and Fadigas), and Italy (Grandi). Adult. Females 5.0 to 6.6 mm. long, males 4.6 to 5.8 mm.; form slender, the males more so; body predominantly dark brown or black; mesonotum and thoracic pleura more or less reddish brown to rufous, the notum some- times almost black; antennae and sheaths of ovipositor brown, the sheaths 3.5 to 4.4 mm. long; legs mostly light brown, fore coxae white, fore and middle tibiae paler than femora, the hind tibiae white except the apical thirds black; fore tarsi uniformly light brown; hind tarsi black with white bands, mid tarsi same, but paler. Abdomen beneath whitish, membranous and concave; ovipositor sheath about as long as the abdo- men (Fig. 28). Fed a combination of sugar and water, the adults can live for a month or more in cages. The males are sexually active upon emerging from the cocoons, and mating takes place at once under favorable conditions. A. W. Crawford (1933) described the activity of mating. Parthenogenetic reproduction has been reported by three observers. Montgomery (1935) found larvae of Glypta in a few oriental fruitworms which had been exposed only to virgins. Putnam (1938) reported an unmated female which parasitized newly hatched fruitworms, in which 12 males then PARASITES OF Cydia packardi 67 developed. Also Crawford learned that uninseminated females will ovi- posit and the embryos develop normally. In all cases, the progeny was male. These, when mated with virgins, produced offspring of both sexes. Oviposition. Each ovary of Glypta consists of 22 to 28 ovarioles of the meroistic type (A. W. Crawford, 1933). The grown eggs remain in the lower chambers of the tubes until deposited. As many as 86 fully de- veloped eggs occurred in a single female. Oviposition begins within 2 to 6 days after emergence from the cocoon. The maximum deposition was secured at 70 to 80° F. and when the larvae of the peach moth were made available in the morning or late afternoon. The female was incited to oviposit only when frass or silken webbing of the host larva was present at the entrance to the burrow. These materials, rather than the larva itself, appear to trigger the egg-laying response. The host was attacked only when concealed within its burrow. The first indication that oviposition is about to occur is the rapid move- ment of the antennae as the female starts feeling over the twig in an effort to locate the point of entrance of the host larva. Upon finding this hole, the female prods into it with the ovipositor. The instant the terebra comes into contact with the host larva, she immediately ceases all movements of the body for 15 to 25 seconds. She then withdraws the terebra and usually crawls down the twig, wiping the head and antennae with the fore legs. However, sometimes a second egg may be laid in the same larva before the female leaves the twig, or, in other cases, she may re- turn after a brief absence for a second oviposition. From a number of timed depositions, the average duration of the process was found to be 19 seconds. Eggs are laid singly, and placed free in the body cavity close to, but never in, the digestive tract, and never in the fat body or muscles. The newly laid egg measured 0.300 to 0.308 mm. long, and 0.090 to 0.1125 mm. wide at the broadest, or cephalic, end. It is slightly curved, and the caudal part more broadly rounded. The thick chorion gives the egg a somewhat solid appearance (Fig. 32). Embryogenesis requires about 41 hours at 75 to 80° F. (Crawford). Reaction of host to attack. The response of the oriental fruitworm sub- jected to attack by female Glypta depends, according to Crawford (1933), largely on the depth of the larval burrow in the peach twig. When the larva inhabits a burrow so short that it cannot retreat, it emerges and crawls rapidly away, unless the parasite holds it with her antennae until she places an egg into it. If the burrow is longer than the ovipositor, the larva retreats hastily into it as a means of escaping attack. But if the burrow is intermediate in depth so that neither of the above responses is possible, the larva is liable to a number of thrusts with the terebra. It appears that the host is not paralyzed by the attack, 68 INSECTS IN ROSE HIPS for Crawford observed the larva is strong and natural in its movements thereafter. Larval life. The oriental fruitworm may be parasitized in any of the instars, from the first to the last. However, if one-day-old larvae are attacked by the female Glypta, very few parasites reach maturity. Again, if eggs are inserted into late third instars, the parasite fails also, for the reason that the host completes its larval development long before the parasite becomes full-grown, and the large majority of the hosts pu- pated. Almost 100 per cent survival of the parasites was obtained with host larvae of the late first and early second instars (Crawford). Also Montgomery (1935) secured best results in the form of adult Glypta with young larvae of Cydia molesta. Larvae approximately 24 hours old when exposed to ovipositing Glypta showed parasitization of 80 per cent, whereas larvae two days old when so exposed, showed 53 per cent, and larvae four days old, only 16 per cent. I have seen only the first two instars of G. rufiscutellaris in larvae of Cydia packardi (Figs. 30, 31). Both Crawford and Montgomery describe and figure four instars. These are similar in being cylindrical in form. The first instar is distinct in having a pale brown, moderately sclerotized head capsule. When newly hatched, the parasite is 0.52 mm. long and 0.11 mm. wide; the full-fed fourth instar varies from 5.0 to 6.98 mm. in length by 1.5 to 1.98 mm. in width. All the instars bear a short caudal process, which becomes shorter and shorter from the first to the fourth instar. Exclusive of this process, all the instars have 13 postcephalic seg- ments. The first three instars lack spiracles, the fourth has them visibly developed (Crawford, Montgomery ). Throughout the course of growth of the instars, the larva of Glypta rufiscutellaris appeared to be free in the body cavity of the oriental fruit- worm (Montgomery, 1935). The host’s viscera became crowded against the body wall as the parasites grew. The fourth instar then consumed the tissues of the host, and the full-grown parasite filled the host skin to distension. Finally the head of the mature parasite larva appeared through the host’s cuticle. Montgomery states the head emerged through the dorsa of the last segments of the host larva. Crawford observed that the opening was made either in the anterior or posterior end, depending on the direction the emergence-ready larva had last headed in the host. Both these investigators saw that the old host skin is worked off by vigorous wriggling or contortions of the parasite. To expedite this process, the parasite may even bend the head around to the tip of the abdomen. As Crawford also pointed out, the escape of the full-grown parasite larva from the host skin is therefore a process of shedding rather than emergence. PARASITES OF Cydia packardi 69 Cocoon, prepupa, and pupa. Having discarded the host skin, the full- grown larva of Glypta encloses itself by spinning its cocoon of white silk. The cocoon “has the appearance of a transparent silken membrane when completed and has an average measurement of 10 mm. by 35 mm.” (A. W. Crawford). In three to four days after shedding the host skin, the parasite has transformed to a pupa. I presume this statement applies only to the summer generation, since the other winters as mature larvae. Half of the three to four days is spent as a prepupa. Crawford’s figure indicates that the prepupa resembles the larva in general form, but has the dark pupal eyes visible through the cuticle. At first white, with red- dish eyes, the pupa grows darker as it develops, assuming the color of the adult as it does so. The female pupae average 7.9 mm. long by 2.02 mm. wide. The males are slightly smaller. Superparasitism. Montgomery (1935) found an egg and a first instar of G. rufiscutellaris in one larva of Cydia molesta. | discovered two instances of duplication in C. packardi. In one instance, two eggs were found in a caterpillar of this moth; in the other, two living immature larvae occurred together in one host. Life cycles. The number of cycles or generations completed by Glypta rufiscutellaris varies from one and a partial second to three in a year, according to reports published to date. Stearns and Neiswander (1930) believed it to be “two-brooded” in northern Ohio in 1927 to 1929. There is evidence of two generations at St. Davids, Ontario (Putnam, 1938). In 1934, the adults of the first generation had already emerged there before May 14, when the first collection of the host, Epiblema strenuana Wk. in ragweed, was made. In 1935, adults appeared from the same host from May 8 to 24, and in 1936, from May 4 to 14. Adults of the second generation issued through August 8 to 18, 1932, and August | to Septem- ber 10 in 1934. A. W. Crawford (1933) reported one and a partial second generation from E. strenuana in western New York. Adults of the spring “brood” emerged between June 10 and July 22, 1931; those of the partial second, between August 12 and 28, 1931. But when attacking Cydia molesta, which completes three cycles in a year in New York, Glypta undergoes its life cycle three times yearly, one for each genera- tion of the host (Crawford). Timewise, these were found to be distrib- uted as follows: full-grown larvae of the first generation are in their cocoons about the latter part of June and the first part of July; the full- grown larvae of the second are in their cocoons in latter July and during August; those of the third cycle occur there about the middle of Sep- tember. Crawford found that the emergence of the adults of the three generations began as follows: first generation, July 2, 1931, and July 7, 1932; second, August 8, 1931, and 1932; third, September 22, 1931, and 70 INSECTS IN ROSE HIPS September 30, 1932. Males began to appear before the females in both these years. My rearings of adult Glypta from Cydia packardi indicate two cycles per year in eastern Illinois. Of the 43 adults obtained from hip jars, some emerged between June 19 and July 17, whereas others appeared in August and September. A quantity of hips from San Jose, Illinois, yielded 29 adults from late September to October 27. That Glypta rufiscutellaris winters as a mature larva is indicated by all reports. Driggers (1930) stated that the winter is passed in New Jersey by the full-grown larva in its own cocoon which lies within that of the host. Crawford (1933) wrote that the parasite exists as a full- grown larva within its cocoon until the middle of the winter, then develops to a prepupa, with eyes, and continues as such until the first part of April. Also the report of Putnam, for Ontario, indicates the species winters as a full-grown larva in the host’s cocoon in the host plant. The fact that McConnell (1928) in Maryland, and other investigators in eastern United States, reared adult Glypta from overwintered oriental fruitworms, indicates that the parasite winters in some immature stage. At temperatures ranging between 75 and 80° F., embryogenesis aver- aged 35 hours. The four instars required 5 to 6 days, 3 to 4 days, 4 to 5 days and one to 2 days, respectively. The prepupal form is retained about 2 days, and is terminated by the molt that discloses the pupa. The average survival of adult males of three generations was 19.6, 16.5 and 20.7 days; of females, 28.4, 28.2 and 31.9 days, respectively (Crawford, 1933). When kept at high humidities, the average lifetime of male Glypta was 30.69 days; of females, 49.43 days (Rice, 1935). Ascogaster sp.?, Braconidae, Hymenoptera That my series of 51 first instars dissected from the larvae of Cydia packardi in rose hips represents a species of Ascogaster is indicated by several records, as follows. I obtained one adult of an undetermined species of this genus from a jar of hips inhabited by this caterpillar. Put- nam (1935) reared A. “carpocapsae Vier.” from the cherry fruitworm in Ontario; Hoerner and List (1952) had A. quadridentatus Wesm. from this host in cherry in Colorado. The most convincing evidence that my 51 first instars probably are a species of Ascogaster lies in the very close structural similarity between them (Fig. 46) and Cameron’s (1938) drawing of the first instar of A. quadridentatus from Cydia nigricana Steph. Also my observations on the bionomic features of these instars agree with the accounts given by Cameron (1938) and by Rosenberg (1934), who investigated the parasites of Cydia pomonella L. in France. The first instar is easily distinguished from all other parasitic larvae PARASITES OF Cydia packardi 71 found to date in insect hosts developing in rose hips. When newly dis- sected from the host, the parasites varied from 0.6 to 1.9 mm. in length, and were about one-fifth as wide; sides subparallel; body decidedly de- pressed, flat above, form varying from straight to sinuous as viewed from above, smooth with no obvious vestiture. Head pale brown, sclerotized, subquadrate as seen from above, tend- ing to bulge at the sides behind; mouth parts obscure, protruding slightly from the middle of the anterior margin of the head. Postcephalic parts white, the several segments uniform in size and shape, each protruding prominently and roundly at the sides. Abdomen terminates in a short but distinct, broadly rounded vesicle narrower than preceding segments (Fig. 46). The thorax, abdomen and caudal vesicle constitute 13 seg- ments. When left in water only a few minutes, the body swells so that the segments, and particularly the caudal ones, tend to lose their form and identity, and even the head assumes an unnatural distended form. Usually the newly exposed larvae lie straight in the water in the dissect- ing dish. When absorbing water, they may assume a sinuous form. My series of 51 first instars was obtained in the course of dissecting larvae of Cydia packardi from the hips of Rosa acicularis, blanda, caro- lina, palustris and a Rosa sp. in the years 1946 to 1951. The parasite occurred at Eaglesnest, Minnesota; Chetek, Wisconsin; Jamestown, Indiana; Taylorville and Mahomet, Illinois, and West Alexandria, Ohio. Nine individuals of the series were discovered between August 7 and 19, at Eaglesnest, in the larvae of C. packardi one-half to two-thirds grown and therefore still feeding in the hips. These parasites were 0.6 and 0.7 mm. long. The remaining 42 were contained in mature host larvae which had issued from the hips. These parasites had grown to 1.75 to 1.9 mm., but still had not developed beyond the first instar. These were dissected out of the hosts between August 29 and November 15. That they remained as first instars so late as mid-November strongly indicates the parasite winters as the first instar, in the normally overwintering mature larva of C. packardi. Accordingly the first instar, after a pro- longed slow growth to its advanced, mature state, seems to enter into diapause which may terminate in late spring or early summer of the en- suing year. Rosenberg (1934) and Cameron (1938) found that the mature first instar of A. quadridentatus in Cydia pomonella and C. nigricana go into diapause that persists through the winter in the mature host larva. They found also that the females insert their eggs into the embryos of the host moths. Many species of practically all microlepidoptera of habits similar to C. packardi are parasitized by A. quadridentatus, according to the above investigators. 72 INSECTS IN ROSE HIPS The Long-Tailed Parasite. Hymenoptera Specific identification of this endoparasite of the larval Cydia packardi has not been made because only the first instar is known to me. The common descriptive designation employed above is, of course, pro- visional. This first instar (Fig. 50) is structurally unique among parasites ob- tained from hosts in rose hips in being extremely slender, with a taper- ing caudal process that is almost as long as the head, thorax and abdomen combined. Head longer than thick, its venter provided with a blunt process; in some views it appears also to have on its venter a pair of minute, obscure, forked appendages which are directed caudoventrad; postcephalic part whitish, clear, subcylindrical, tapering moderately to- ward the caudal process; this latter single appendage is clear, broad at base, and narrows gradually to a fine point behind, normally moderately arched and directed almost straight caudad. The postcephalic area, ex- clusive of the caudal appendage, appears to comprise 12 segments, each of which is divided into an anterior and a posterior half by a superficial transverse constriction. When newly removed from the host to water, the parasite reacts by rolling sidewise quickly and often, and bending the cephalic end stiffly ventrocaudad, whereby the body takes a J-shaped posture; or the cephalic end and the caudal process may be bent ventrad U-shaped, and again straightened alternately, while the “tail” flails about slowly. Only six individuals of this form were discovered in the course of dis- secting cherry fruitworms for parasites from 1945 to 1951. These were taken as follows: Jamestown, Indiana, four. One measured 1.3 mm. long, from host one-third grown, October 2, 1945; two measured 1.8 and 2.0 mm., from mature, emerged hosts, October 19, 1947, and one which was 2.0 mm. long, from a mature, emerged host, October 31, 1947. All originated in hips of Rosa palustris. Philo, Illinois; one parasite 2.0 mm. long, from a host about one-fourth grown, July 8, 1948, in hip of Rosa carolina. Fox Ridge State Park, Charleston, Illinois; one parasite, 2.1 mm. long, from a host four-fifths grown, July 31, 1949, in hip of Rosa virginiana. Because only first instars were discovered in the over-all period of July 8 to October 31, the species may prove to complete one generation in a year and undergo a prolonged diapause, including the winter season, as a first instar. Scambus hispae (Harris), Ichneumonidae, Hymenoptera The adults in my collection are slender, dark brown to black, the PARASITES OF Cydia packardi 73 males measuring about 4.7 mm. long, the females 5.0, with sheaths of the ovipositor 3.0 mm. long; coxae and femora brownish yellow; front legs to hind, the tibiae are white, to white, ringed with black. This ichneumon has been treated in the literature under several scien- tific names. See the Synoptic Catalog by Muesebeck, et al., (1951, p. 187), for synonyms. This source states also that the species is trans- continental in the Canadian, Transitional and Upper Austral faunal zones of North America, and that “the species of Scambus are parasites of small Lepidoptera in buds, fruits, leaf rolls, leaf mines, galls, or similar retreats.” S. hispae has been reared from thirty species of small Lepi- doptera, not including Cydia packardi, and appears to be the most common member of the genus. It is a solitary ectoparasite. My belief that S. hispae parasitizes Cydia packardi in rose hips is based on two kinds of evidence. First, I have obtained a series of adults from jars of rose hips in which the cherry fruitworm was the only lepi- dopteron present. Also, the published reports show that hispae parasitizes host caterpillars having habits similar to those of packardi. Second, in June, 1951, I picked a quantity of buds of Rosa carolina at Urbana that bore the unmistakable injuries made by larvae of packardi. On July 3, 1951, an adult male hispae appeared in the cage. Also fourteen adults were reared at Urbana from quantities of rose hips stored in fruit jars, as follows: Eaglesnest, Minnesota, one o', September 24, 1950; two o's, Septem- ber 12 and 13, 1951, and one ?, September 22, 1951; one ?, May 25, 1948, and one 6, May 28, 1948. All these originated in hips of Rosa acicularis. Solon Springs, Wisconsin, one (sex unknown), October 7, 1946; three os, October 6, 11, and 24, 1947. From Rosa blanda. Rice Lake, Wisconsin, one 6, October 8, 1947. From hips of Rosa blanda or carolina. Chetek, Wisconsin, two o's, October 5, 1949, September 22, 1951. From hips of R. blanda or carolina. Urbana, Illinois, one ¢, July 3, 1951. From bud of Rosa carolina. Only fragments of bionomic data appear in the literature, so far as I have seen it. Cushman (1927) observed it as a solitary ectoparasite on the larvae of the pine tip moth, Rhyacionia frustrana (Comst.). It attacks full-grown or large larvae, and its cocoon consists of a silken lining to the burrow of the host. He added that it is “a common parasite of internally feeding microlepidopterous larvae, with many known hosts.” Doner (1936) found it (Pimpla indagatrix Cr.) parasitizing the mature larvae of Coleophora pruniella Cl. in Wisconsin. Under the name Epiurus in- dagator (Cr.), Porter and Garman (1923) reported S. hispae as one of the parasites reared in Connecticut from the European “apple and thorn 74 INSECTS IN ROSE HIPS skeletonizer,” Anthophila (Hemerophila) pariana (Clerck). It oviposits on the full-grown larva, and the adult emerges from the host cocoon. Brunson and Allen (1948) reported it as a “cocoon parasite” of Cydia molesta in New Jersey and other eastern states, and McConnell (1928) obtained adults from the overwintered larvae of molesta in Maryland. As observed by Coppel (1946) in British Columbia, it performed as a primary parasite of the spruce budworm, Choristoneura (Archips) fumi- ferana (Cl.), and wintered as a secondary in the cocoons of the ichneu- monid Phytodietus fumiferanae Rohwer. According to Schaffner and Griswold (1934), it (Epiurus nigrifrons Vier.) is a solitary parasite in New Hampshire and New Jersey where it hibernates in the host pupa and has at least two generations in the year on Halisidota tesselaris S. and A. and Hemerocampa leucostigma S. and A. On two occasions I reared Scambus hispae from hosts obtained in the vicinity of Urbana, Illinois. In both cases, the species was identi- fied by R. A. Cushman, U.S. National Museum, as “Epiurus indigator Cress.” In the first instance, adults emerged from the central cell of the gall made by the cynipid, Disholcaspis mamma (Cress.) and appeared to have parasitized the larva of this insect, but the host may have been some lepidopteron that entered the vacated gall to cocoon (Balduf, 1926). In the second instance, it attacked the common bagworm, Thy- ridopteryx ephemeraeformis (Haw.). Bags dissected in the early part of the period of March 31 to April 18, 1936, contained overwintered mature parasitic larvae in the chrysalises of both male and female hosts. Most bags contained one parasite, but a few had two larvae or cocoons in the crysalids. The series of 16 adults reared that spring comprised 15 females and one male (Balduf, 1937). Euderus cushmani (Crawford), Eulophidae, Hymenoptera This chalcid shares the role of solitary ectoparasite on the larva of Cydia packardi with Scambus hispae. The recorded hosts, Polychrosis viteana (Clem.), Anarsia lineatella Zell., Argyresthia thuiella (Pack.) and Coleophora pruniella Clem., as given in the Synoptic Catalog (Muesebeck, et al., 1951), are, like those of S. hispae, larvae of small Lepidoptera which live concealed in parts of their food plants. Accord- ing to the same source, E. cushmani has previously been reported from Pennsylvania, Wisconsin and California. I have it from Wisconsin and Illinois. J. C. Crawford (1915) described E. cushmani from specimens whose host was Polychrosis viteana. My adults from Cydia packardi are dark, the males approximately 1.0 mm. long, the females about 2.0 mm. long. Kiefer and Jones (1933) reared E. cushmani (Crawf.) (Secodella cush- mani) from the immature larvae of Anarsia lineatella, on which it ap- PREDATORS OF Cydia packardi 75 parently was an ectoparasite. Adults appeared on July 21, 1932, and March 6 to 8, 1933, in California. I have obtained 10 adults, all of which emerged from hips stored in jars, as follows: Chetek, Wisconsin, two 2s, September 13, 1951; three 9s, September 10 and 18, 1951, and one 6’, September 18, 1951. From hips of Rosa carolina or blanda. Madison, Wisconsin, one ?, September 2, 1946. From hips of Rosa arkansana. Urbana, Illinois, one 9, May 10, 1949. From hips of Rosa palustris. Also two o's, August 16 and September 14, 1951, from hips of R. carolina. The male secured at Urbana on August 16, 1951, developed from an advanced black pupa found in a mine of Cydia packardi in a rose hip on August 5. The host’s burrow originated at the neck of the hip and ex- tended to the equator, where the host had last fed and was overcome by the parasite. The pupa, adjacent to the remains of the host, was not encased in a cocoon. The small length and breadth of the burrow indi- cated that the host was very small when it was attacked by the female Euderus. AN UNIDENTIFIED EcroparasitE oF Cydia packardi At Eaglesnest, on August 25, 1947, I discovered an egg 1.9 mm. long (Fig. 51) adjacent to a limp submature larva of C. packardi in its mine in a hip of Rosa acicularis. The parasite—probably a large ichneumonid sp.—had killed the caterpillar with a paralyzing thrust of the terebra. Prepators oF Cydia packardi Enoclerus P sp. and Phyllobaenus sp., Cleridae, Coleoptera Besides attack by the parasites discussed above, Cydia packardi is at- tacked occasionally by predators. From time to time since 1945, I have come upon coleopterous larvae either in the burrows or mines of the caterpillar in the hypanthia of rose hips, or free among hips stored in rearing jars. Mr. O. L. Cartwright of the U.S. National Museum ex- amined some of my specimens and determined them as members of the genera Phyllobaenus and “probably Enoclerus,” of the family Cleri- dae. I have compared the rest of my specimens with those determined by Cartwright and find all are identical with those he returned. Reference was made also to the work of B6ving and Champlain (1920) on the larvae of this interesting family. The following descriptions pertain to what appear to be mature individuals, and will serve to distinguish the two generic types found in hips of roses. Enoclerus? sp. Length, 9.0 mm.; body slender, its width approximately uniform throughout; numerous white bristles tend to be bunched on the 76 INSECTS IN ROSE HIPS tergal and pleural aspects of the postcephalic segments, shorter ones over the venter; body whitish, except cerci or caudal processes, pronotum and all of head brown to reddish brown and sclerotized; cerci sturdy, straight, moderately divergent, their apices blunt except dorsocaudal angle of each member pointed. Phyllobaenus sp. Length, 8.0 mm.; spindle-shaped; numerous white bristles distributed somewhat generally over the body; body mostly whitish or suffused with purple, except head which is medium brown, sclerotized, also pronotum brown back to the heavy, broadly V-shaped purple figure; a pair of small dark lunate marks on meso- and metanota, these nota also pale purplish; abdominal tergites 3, 4, 5 and 7 dark purple, and suranal plate dark brown; prosternum with a T-shaped pur- ple design, venter of body otherwise light; coxae and femora ectally brown. No cerci of any sort present. Three individuals of Enoclerus? sp. were obtained, as follows: Mayview, Illinois, August 1, 1948. In an extensive mine of C. packardi in a hip of Rosa carolina, I found a head of the caterpillar, with shrunken postcephalic parts attached. The size of the head indicated the victim was nearly full-grown when attacked, from behind. In the same mine occurred a larva of Enoclerus? sp., about 6.0 mm. long, or two-thirds grown. The facts and circumstances here make it practically certain that Enoclerus? preyed on the larva of C. packardi. Mayview, Illinois, August 9, 1948. A mine of C. packardi in a hip of Rosa carolina extended from neck to base of petiole, or a distance of 12 mm. The sizable dead caterpillar had been greatly reduced in volume, and the mine was occupied by a larva of Enoclerus? sp. 5.5 mm. long, and estimated to be two-thirds grown. Velma, Illinois, October 14, 1945. The clerid larva, 10.0 mm. long, was moving over the bottom of a jar containing hips of Rosa carolina. When crawling, it progressed smoothly and without elevating any parts of the body. An anal proleg supported the end of the abdomen. Five individuals of Phyllobaenus sp. were taken. All originated in extensive, mature stands of wild Rosa carolina growing in an area of shifting sand near San Jose, east of Havana, Illinois. They came to my notice between August 1 and October 25, 1951. The larva obtained on August 1 occupied a large mine of C. packardi that opened at the peti- olar end of the hip. Since I failed to see the victimized caterpillar, I conclude it had been killed and largely consumed by the clerid larva, which was 5.5 mm. long. Its distended form indicated it had recently fed, or was about to molt. The other four larvae of this genus had apparently become full-grown since they had emerged from the hips and were taken moving free among the hips in the jars. They measured 7.5 to 10.0 mm. in length. PREDATORS OF Cydia packardi iid These clerids probably preyed only on the caterpillars of Cydia pack- ardi, and not on the maggots of Rhag. basiola. Evidences to this end are twofold. First, the rose hip fly was not reared from hips taken at San Jose, and was infrequent at Mahomet and Velma. Second, the habits of the cherry fruitworm favor the entrance of the predators into its burrow or mine, while those of the hip fly do not. The larvae of packardi perforate the hypanthium when they enter from the eggs, and subsequently prepare one to four holes to the outside where they eject pellets of excreta. Especially the latter type of holes made by the larger instars are likely to afford entrance for the slender clerid larvae. This may explain why the victims are usually large. The further fact that much of the excrement is ejected to the outside, or, when left in the mine, takes the form of semihard yet penetrable pellets, indicates the predators may pervade the mines to locate the caterpillars, and to attack them at the defenseless caudal end. By contrast, the larvae of Rhag. basiola enter the hypanthium through the inner end of the egg pit, which becomes plugged externally with dried hip substance, make no holes through the skin of the hip until they are full-grown and ready to emerge, and more or less fill the mine behind them with wet pasty obstructive excrement. Thus, the maggot inhabits a space that appears effectively sealed off against access of the clerid predators, whereas the caterpillar of Cydia packardi inadvertently provides an entrance and thereby makes itself vulnerable to predatory attack. Carposina scirhosella H. S., CARPOsINIDAE, LEPIDOPTERA In a footnote to an article that describes a new Megastigmus from Pseudotsuga, Wachtl (1893) gives the names of several species of in- sects that inhabit the hips of roses. Among these is a “Gelechiid moth, (Carposina scirhosella Tr.).” He gives no references to literature, and my search of the Zoological Record, 1893-1951, and series A of the Review of Applied Entomology, yielded none. The genus and species were named by Herrick-Schaffer in 1855 (Horn and Schenkling, 1928). It is recorded from southeastern Europe and Asia Minor (Meyrick, 1913). An idea of its probable habits may be gained from published studies of its generic relatives. In New Zealand, the larva of the raspberry-bud moth, C. adreptella Walk., bores into leaf buds, fruits or the succulent shoots, and may feed on leaves, or migrate from bud to bud to obtain sufficient food (Jeffreys, 1939). The larvae of C. sasakii Mats. bore into the fruits of apple, pear and others, where they complete their growth. Pupation is in the soil. It has been observed in Japan and Korea (Mura- matsu, 1927). 78 INSECTS IN ROSE HIPS THE SCAVENGEROUS FLy, Lonchaea polita Say, LONCHAEIDAE, DIPTERA This is a sturdy bluish black fly about 3.5 mm. long, with long wings that extend far beyond the end of the abdomen. The larva is believed to feed on the excrement and/or plant refuse caused by the feeding of the larvae of Rhagoletis basiola and Cydia packardi in the hypanthium of rose hips, and in similar hosts. Some puparia were found in the abandoned mines of these phytophags, and adults emerged from old hips kept in fruit jars. Although not numerous, this is the principal scav- engerous animal known to me from hips. I obtained a series of 36 individuals from jars of rose hips, and one from nature. The hips originated in Minnesota, Wisconsin, and Illinois. All the 36 flies were reared at Urbana, Illinois. Four of these are now in the collection of the United States National Museum. The rest bear data as follows: Eaglesnest, Minnesota, three from Rosa blanda, July 31 and August 9, 1946. Gordon, Wisconsin, 10 from R. blanda glandulosa, June 11 to 20, 1946. Urbana, Illinois, one from R. palustris, June 17, 1947; one from R. vir- giniana, July 12, 1947. Pana, Illinois, two from R. carolina, June 11 to 12, 1946. Philo, Illinois, 16 from R. carolina, June 9 to 14, 1946. Also four empty puparia found in old hips in jars, June 14, 1946. This shows that the larvae pupariate in the hip, and do not emerge to transform on the ground, as Rhagoletis does. Pechuman (1937) reared numbers of L. polita from elm logs, usually in association with bark beetles, and stated it is probably a facultative parasite of scolytids, although usually acting as a scavenger. On the other hand, Kaston (1939) believed it to be predatory on the elm bark beetle, Hylurgopinus rufipes (Eich.), in whose egg galleries the fly pupates. Claassen (1921) gives an account of this fly in his report on insects associated with Typha. Dr. C. W. Sabrosky, who identified my specimens, commented in his report to me that polita is “undoubtedly a scavenger, following damage by other insects.” THE Rose Curcurio, Rhynchites bicolor (Fasr.), CuRCULIONIDAE, COLEOPTERA Common names. This red-and-black beetle, sitting on rose buds and boring into their sides with its long proboscis, is a familiar and unwel- come insect (Fig. 33). It is the best known of the several species of THE ROSE CuRCULIO, Rhynchites bicolor (¥ABR. ) 79 insects that develop in rose hips. Popularly it goes by the name “rose bug,” and in the literature is referred to variously as the “rose snout beetle,” “red rose beetle,” “rose weevil,” “black-snouted rose beetle,” and “rose bud curculio.” However, “rose curculio” is the official com- mon name adopted by American entomologists. Taxonomy. Fabricius (1792-94) described this North American species under the generic names Attelabus and Curculio. Herbst (1796) es- tablished the genus Rhynchites and assigned bicolor to it. Sharp (1889) subdivided Rhynchites into further genera, one of which was Merhyn- chites, which included bicolor. Pierce (1913) adopted Sharp’s new genera and his placement of bicolor in it. Thus the name of this beetle became Merhynchites bicolor (Fabr.). Moreover, Pierce (1913) recognized six varieties of bicolor in North America, four of them new to him, which were based largely on differences in color. Following are the six vari- eties, with their then known geographical distributions as cited by Pierce. M. bicolor bicolor (Fabr.): Washington, D.C., New Hampshire, Maine, Massachusetts, Manitoba, New Jersey, Wisconsin, Iowa, Utah, New Mexico. M. b. wickhami Cockerell (Cockerell, 1912): Vancouver, British Columbia; Washington, Oregon, California, Idaho, Utah, Wyoming, Montana, Colorado, Arizona, New Mexico, North Dakota. M. b. cockerelli Pierce: California, Idaho, Utah, North Dakota. M. b. ventralis Pierce: South Dakota. M. b. piceus Pierce: Oregon. M. b. viridilustrana Pierce: Arizona. Buchanan (1939) reexamined the genera of Sharp, and, for reasons given by him, suggested that Rhynchites be temporarily retained in its broad sense and that the name Rhynchites bicolor (Fabr.) be continued for the rose curculio. As regards the varieties recognized by Pierce, Green (1920) believed that some of the forms are capable of specific definition by means of constant structural and sculptural differences. I am inclined to agree with this view for the reason that Rhyn. b. bicolor differs definitely in some of its bionomic features from Rhyn. b. wickhami, which Hoerner (1936) investigated in Colorado. Distribution. Chittenden (1901) enumerated the records from speci- mens in the collection of the U.S. National Museum, and Pierce later (1913) did the same. Since these reports are generally available today, I do not repeat their data here, but give only the records known to me from later publications, and those added through my own studies. Massachusetts: Woods Hole (Cockerell, 1912), Worcester (Gates, 1909). 80 INSECTS IN ROSE HIPS Connecticut: New Haven (Britton, 1928). New York: (Crosby and Leonard, 1922; Griswold, 1937); New York City (Blatchley and Leng, 1916). New Jersey: (Smith, J. B., 1899); New Brunswick (Dickerson, 1910). Ohio: Oak Harbor, West Lafayette, Homeworth (Balduf, original). Michigan: Flint (Pettit, 1927). Indiana: Idaville, Jamestown (Balduf, original). Illinois: Many localities in northeastern quarter of state (Balduf, orig- inal). Wisconsin: From Solon Springs south to Madison, along U.S. highways 53 and 12 (Balduf, original). Minnesota: (Lugger, 1899); St. Paul (Marcovitch, 1916). Oklahoma: Norman (Balduf, original ). Kansas: Manhattan (Balduf, original). Nebraska: Crete (Balduf, original). North Dakota: Crosby (Truax, 1946). Manitoba: (Robertson, 1923). Saskatchewan: (Twinn, 1933); Saskatoon (Balduf, original). Alberta: (Twinn, 1933). British Columbia: (Ruhmann, 1933). Montana: “Far into the mountains” (Cooley, 1925). Idaho: Moscow (Balduf, original). Oregon: (Zeller and Schuh, 1944); Russellville (Lovett, 1915). Colorado: (Cassidy, 1888; Cockerell, 1912); many localities around Fort Collins, up to 10,185 feet altitude at Leadville (Hoerner, 1936). Utah: Mapleton in Utah Co.; Mt. Timpanagos, 6,800 feet, near Provo (Balduf, original ). California: Towle, Placer Co. (Branigan, 1915). New Mexico: (Cockerell, 1894, 1912, Riley and Howard, 1891); Santa Fe (Cockerell, 1894). I have not taken Rhyn. bicolor itself in northeastern Minnesota, but two of its hymenopterous parasites obtained here indicate it occurs in small numbers. The parasites are Zatropis rosaecolis and Luchatema baldufi, reared from rose hips from the Cascade river near Lake Superior and from Ely, respectively. LeConte and Horn (1876) reported this curculio as “an abundant species from the Atlantic to the Pacific coast.” Blatchley and Leng (1916) added that it “ranges from the Atlantic to the Pacific, but principally northward in the Transition Life Zone.” The thirty-fifth line of latitude marks the approximate southern limit of its occurrence, whence it seems to extend northward somewhat beyond the fiftieth line in the Canadian provinces. Were all the records from insect collections assembled here, they would probably show that R. bicolor occurs in all the states and THE ROSE cuRCULIO, Rhynchites bicolor (¥FABR. ) 81 provinces within these limits. These coincide approximately with the known boundaries of the rose hip fly, the cherry fruitworm and the native rose seed chalcid. Host plants of Rhynchites bicolor bicolor. As defined here, a “host” is the plant from which the larva derives its food and shelter while de- veloping through its several instars. As applied to the eastern or typical form of the rose curculio, “host” excludes plants on which only the adult may be found to feed. Accordingly, only species of the genus Rosa con- stitute the host plants of this variety as observed to date in the eastern states. I have obtained what are probably the larvae of Rhyn. bicolor bicolor from the hips of rose species as follows: Rosa blanda, from Solon Springs to Endeavor, Wisconsin; R. arkansana, Madison, Wisconsin; R. a. suffulta, Huntley, Illinois; R. suffulta, Manhattan, Kansas; R. carolina, Chetek to Endeavor, Wisconsin; R. carolina, many localities in the northeastern quarter of Illinois, and Idaville, Indiana; R. palustris, Homeworth, Ohio and Urbana, Illinois; R. virginiana, Fox Ridge State Park, Charleston, Illinois; R. Macounii, Crete, Nebraska; R. rugosa, Urbana, Illinois. Larvae presumed to be this species inhabited hips of a Rosa species that were sent to me from Norman, Oklahoma. The following host or food records for Rhyn. b. bicolor are taken from the literature: Rosa rugosa, New Brunswick, New Jersey (Dickerson, 1910); Worcester, Massachusetts (Gates, 1909); St. Paul, Minnesota (Marcovitch, 1916); and New York (Blatchley and Leng, 1916; and Griswold, 1937). Also Lugger (1899) reported roses as the host of this curculio in Minnesota. Food plants and injury. Information available to date indicates that the adult of Rhyn. b. bicolor feeds only on plants of the genus Rosa. A single dissenting record is the statement by Griswold (1937) that the beetle also gnaws holes in flower buds of peonies (Peonia). When attacking Rosa, the curculios may utilize floral parts, hips and also vegetative parts of the plant. However, they appear to prefer the new, unopened flower buds. These form a conical structure of convenient size and shape to be embraced by the legs, and the tender corolla is readily penetrated by the long curved proboscis. Sometimes the snout pierces an overlying sepal as well. The filaments and anthers of the stamens within the lumina of the buds are thus accessible to the beak, and constitute no little part of the beetle’s food. While the injury done to flower buds is less conspicuous than that to the flower, it is actually more serious and deep-seated, for buds may open imperfectly, or they may fail to open, remaining in place to die and dry up intact. The injury done to the petals is most conspicuous and may come about 82 INSECTS IN ROSE HIPS in one of two ways. First, and chiefly, a straight row or series of several holes of uniform size and shape appears on a petal, or on several adja- cent petals. These originate when the proboscis bores through several petals or layers of the intact bud, and become apparent when the bud opens into a flower. Secondly, the holes thus formed may be enlarged after the petals open. But such enlargements seem to occur infrequently on Rosa carolina for the reason that the waxy surfaces of open petals afford but insecure footing to the beetle desiring to feed, and also the mandibles cannot readily seize the smooth surface and remove bits to ingest. But when the beetle stands on the comparatively rough-surfaced overhanging sepal or leaf, or on the pistillary center of the flower, its footing is secure, and from such a position of advantage, it has been seen to reach over to adjacent punctured petals and feed. The hole is enlarged by gnawing away the margin. Also it may then bore into, or clip the edges of, previously uninjured petals, or penetrate into the in- cipient hip under the pistils. Thus the open flowers may, in extreme cases, be reduced to tatters. The tender, newly developed terminal shoots of branches, the leaf buds and succulent new leaflets appear to be secondary choices when flowers and buds are available at the same time. I have observed three circumstances in which vegetative parts are eaten, first when some beetles emerge from the ground before the flower buds develop; second where these buds will not occur as a result of severe pruning of the plants late in the spring; and third, when the crop of floral buds and flowers is inadequate for large populations of curculios. The injuries to hips of Rosa carolina observed on the university campus at Urbana may be regarded as of two kinds. First are the pits or tunnels bored directly through the hypanthium by the female for deposition of her eggs. These are usually distinguishable by the covering of black secretion over the external opening. Second, and similar to the first, are the pits made by adults of either sex incident to feeding only; these con- tain no eggs, and lack the external covering of black material. The feed- ing punctures almost invariably occur on the petiolar or basal half of the hip. This area is utilized for the same reasons that it is employed for oviposition by Rhag. basiola, as stated above. Not only the hy- panthium but also the achenes may be ingested insofar as the proboscis is long enough to penetrate to the inner cavity of the hip. The extent of the damage inflicted on cultivated roses by Rhyn. b. bicolor is shown by the following instances. According to Gates (1909), the beetles ruined practically every bloom of Rosa rugosa at Worcester, Massachusetts, and their feeding extended even to clusters of tender, unfolding leaves. Dickerson (1910) reported that the beetles fed first in floral buds and flowers of R. rugosa at New Brunswick, New Jersey, THE ROSE CURCULIO, Rhynchites bicolor (FARR. ) 83 then attacked the hips so extensively that almost all bore one or more punctures. Also pertinent here are two of the more serious instances of injury to cultivated R. carolina at Urbana, Illinois. On June 22, 1949, a random sample of 573 hips checked in the field showed 40 per cent had been punctured by feeding or ovipositing adults. Since the period of | adult life had not terminated on that date, the level of injury for the season eventually became higher. At another station, in a sample of 426 hips examined July 4, 1949, the over-all injury reached 37 per cent. At Urbana, I found that the extent of damage varied not only from year to year but rose species to rose species and also from one part of a garden or hedge to another. Several explanations are suggested by these observations, as follows: unequal parasitism of the larvae in the several rose plantings; cultivation of the soil while the diapausing larvae and pupae inhabited it; the texture or penetrability of the soil to the emerg- ing young adults and the descending mature larvae; isolation of plant- ings by intervening obstruction such as large buildings, trees and shrub- bery; immunity of some rose species to attack, and the tendency of the beetle to concentrate on parts of a planting where an abundance of blooms occurs. For example, R. carolina bears few flowers in dense shade. On July 4, only two per cent of the hips in the shaded part of a hedge bore punc- tures, while two parts growing in sunlight had been punctured to the extent of 47 and 50 per cent, respectively, on that date. By contrast, a fine long hedge of Rosa hugonis which consistently bore heavy crops of sizable hips year after year completely escaped attack by all phytophags associated with rose hips, probably because it flowered before the reproductive adults emerged at neighboring plantings of other rose species. The bionomic features of Rhyn. bicolor as recorded by observers in the western mountains and in the eastern area of the United States are so divergent that separate treatments of western and eastern forms seem most expedient. The known facts for the eastern form, which, according to Pierce (1913), appears to be the typical subspecies, Rhyn. b. bicolor (Fabr.), will be given first, followed by an account of the subspecies known in the west. Several of the forms delineated by Pierce occur in western North America. However, of these, only Rhyn. bicolor wickhami Cockerell has been identified subspecifically by the investigators. We are indebted to Hoerner (1936) for his recent comprehensive study of this form in Colorado. Rhynchites bicolor bicolor (Fabr.) in the East Adult (Fig. 33). According to Pierce (1913, p. 371), “the typical §4 INSECTS IN ROSE HIPS specimens have elytra, prothorax, and base of head to the eyes red, the remainder of the body black. These specimens usually have the elytral strial punctures quite plainly larger than the interstitial punctures. The front between the eyes is moderately punctate, and rugose near base of beak.” Emergence. In my personal observations at Urbana the young pale reddish adults, after shedding their pupal cuticles, came up out of their pupal cells in the soil of the rearing pots in May to June, digging out from depths varying from one to four, but mostly two to three, inches. The length of time required for emergence was not determined, but is certainly affected by the physical properties of the soil. In 1947 the peak of emergence per day was reached in approximately 20 days after the beginning of issue; in 1949, this figure was 16 days. Locomotion. Besides its six legs for crawling over the plant, the beetle flies readily from bush to bush, particularly when the local supply of rose buds is diminishing, and, as many entomologists have observed, it promptly folds its legs under itself to fall, usually to the ground, when disturbed by the abrupt approach of people. It is presumed that the plant is reached again by crawling. These activities, and especially flight, are performed more freely at higher atmospheric temperatures. The red color or odor of the carolina rose appears to stimulate a positive response in the flying beetle: an individual, during a flight of three feet, passed over a red rose about two feet below, and made an abrupt U-turn back to the flower, landing precisely upon it. When the waxy petal provided no foothold, the beetle clung to an adjacent leaf, from which it could reach the flower with the proboscis. Proximity would have indicated feeding on the leaf, but, due to some sensory reaction, the petal was pre- ferred. Additional performances of similar kind were observed. Copulation. This activity seems to begin when the beetles have been out of the ground about 10 to 15 days. Since emergence at Urbana be- gins about mid-May, copulation commences near June, and, where larger numbers of beetles were concentrated locally, continued past the middle of July. The male climbs to the dorsum of the female, where the union is effected. During the period, the somewhat larger female may crawl about and engage in feeding. Not infrequently one may see a male approach another adult, but instead of entering into the mating position, turns aside at once. Such behavior probably means that the individual approached is another male, or, that males cannot identify the sex of their species until they make physical contact. The proboscis sometimes comes into use in the activity of mating. On one occasion an aggressor, presumably a male, assaulted a mating pair by inserting his snout between the two, and after some failures, suc- ceeded in separating them. THE ROSE CURCULIO, Rhynchites bicolor (¥FABR. ) 85 Oviposition. Dickerson (1910) noted that Rhyn. bicolor rarely fed on the tender tips of Rosa rugosa branches at New Brunswick, New Jer- sey, and he made careful search of the punctures to determine if eggs were laid in the branches. He found no eggs in them. When pursuing the larval life of Cydia packardi, at Urbana I examined hundreds of the flower buds of Rosa carolina which also bore punctures made by the rose curculio. In no case did I find eggs of the beetle in these buds, and, moreover, I did not discover the beetle engaged in oviposition in buds. Samples of hips which bore beetle punctures, picked at intervals of three days to two weeks through June to August of 1947 to 1951 were also dissected. I found good numbers of eggs, and larvae of various sizes, in the bottoms of the egg pits in the hypanthium or in the achene cavities of the hips. As a further check, I again inspected quantities of R. carolina buds at Urbana from May 21 to June 4, 1955, when both flower buds and the rose curculio occurred. A good percentage of the buds bore punctures made by the proboscis of the beetle, but none such contained eggs. At the same time, a high proportion of new hips were found to contain eggs and first instars, showing that oviposition was in progress when the buds were being inspected. From these positive and negative results, I conclude that the eggs of Rhyn. b. bicolor in Mlinois, and probably also in states to the east, are laid only in the hips. However, further explorations in the northern midwestern states are desirable in view of the observations by Marco- vitch (1916) in Minnesota. He reported differences in susceptibility of rose varieties to oviposition by Rhyn. bicolor, having “found that in the form, Rosa rugosa, the common Japanese rose, the eggs are laid within the hips, and not in the flowers or elsewhere. However, with Rosa rugosa alba, a white variety, the larvae were found within the flowers on July 22, to be already about half-grown. It appears, then, that in this variety, the eggs are laid in the flowers, and later [the larvae] may crawl into the hips.” Oviposition in the hip of R. rugosa is described by Dickerson (1910). When excavating the egg pit with her beak, the beetle spreads her legs as if to brace herself. The boring continues until the beak is inserted to the eyes, and the antennae are pressed close together and upward against the top of the head. The final operation “appeared to be the smoothing down of the sides of the puncture and enlarging the bottom; the whole operation taking somewhat over a quarter of an hour. ... Having com- pleted the puncture, the insect turned about, rested its anal extremity in the cavity, and remained there for about half a minute while it ovi- posited. Then turning about again, it spent slightly more than half a minute in apparently pushing the egg down into the bottom of the 86 INSECTS IN ROSE HIPS cavity and covering the opening. . . . Whether the covering of the punc- ture is a secretion of the plant caused by the injury, or in part is some secretion of the weevil is not quite evident. At any rate, it forms a dis- tinctive covering, light in color at first and gradually darkening.” Several times I watched the beetle oviposit, and found the procedure is essen- tially as described by Dickerson. Marcovitch had reported previously (1916) that the “egg-laying punc- ture” can be distinguished from the feeding punctures in being blackish and covered over, with no noticeable depression. I have repeatedly con- firmed that such a black covering is constructed over the mouth of the egg pit, and that it is lacking where eggs have not been inserted. I have only circumstantial evidence to support my belief that the black cover- ing substance originates in the beetle, not from the hip. First, when the cover is being formed, the beetle stands with the end of the proboscis introduced into the mouth of the puncture, then manipulates the mouth parts rapidly and produces a heavy gray-brownish fluid. This may be regurgitated food, a salivary secretion or both. Were it plant sap oozing from the wound, a black covering would form also over feeding punc- tures. Not infrequently, solidified plant sap or protoplasm from the punc- tures plugs the pit below the outer surface. Such plugs do not turn black. Further, when one carefully dissects off a piece of the black cover from egg pits, he finds it is uniformly thin and slightly transparent. These characteristics suggest it was laid down and molded by a mecha- nism capable of controlled manipulation, such as mouth parts, and not a product poured out by chance from a wound in the hip. The cover lies over the center of the somewhat star-shaped scar within which the pit is located. The scar is formed as the mandibles scoop out bits of hypan- thial skin preparatory to boring of the egg pit. The shiny, coal black covers not only are almost certain indicators of oviposition, but also of the extent of embryonic development. With en- largement of the hips, the covers become split into pieces whose number and extent of separation roughly indicate the time elapsed since oviposi- tion and thus also the approximate age of the embryos. However, if the hip is already full-grown when the egg is laid, the cover remains intact and has no such indicative value. The deposition site of the eggs of Rhyn. b. bicolor varies within the hips of Rosa carolina investigated for the most part at Urbana. Some occur within the hypanthium when it is of such thickness that the snout cannot reach the achene cavity. However, most eggs are found in that cavity. They are set in a hole gouged in a single achene, or in a receptacle chewed out between two contiguous achenes, or lie free upon the pe- ripheral fruits. Rarely, an egg lies amid the pollen tubes in the apical half of the achene cavity; in such instances it may have been inserted through THE ROSE CuRCULIO, Rhynchites bicolor (¥FABR. ) 87 the natural aperture at the floral end of the hip. Almost without excep- tion, the deposition sites occur within the basal petiolar area of the hip. Variations in the size, “set,” age, or species of the hip determine the depth to which the beetle can penetrate. In the course of the fruiting season, R. carolina produces several “sets” or clusters, whose component hips generally are successively smaller as the season advances. When most ovipositions coincide with maturing of the first cluster, a larger number of eggs should logically be found within the relatively thicker hypan- thium than when oviposition coincides chiefly with the maturing of the smaller later hips whose hypanthium is relatively thinner and thus more penetrable by the proboscis. Egg (Figs. 34, 35). The egg is subrotund, its ends very broadly rounded, with one end perceptibly wider than the other. In size, it averages 0.95 mm. in length, 0.8 mm. in height at the middle, and 0.75 mm. in maximum diameter. Some eggs found in the hips are covered with a thin layer of whitish frostlike material, which presumably orig- inates in the reproductive tract, and peels off readily when the egg is removed from its nidus. The bared egg, when still young, appears shiny, smooth, hyaline and pearly whitish, later becoming yellowish as the embryo within approaches maturity. The chorion is tough and leath- ery, and the egg so turgid and resilient as to be held with difficulty with needles, and sometimes pops audibly when punctured. The advanced embryo takes on the form of the larva which then be- comes visible through the chorion, and the reddish mandibles may be seen to open and close as when chewing food. It therefore seems certain that these mouth parts later serve to rupture the chorion in the process of hatching. Larva (Fig. 36). The larva of Rhynchites bicolor bicolor is the only phytophagous form of insect that lives free in the achene cavity of the rose hip. Structurally the larva is legless and white, excepting the exposed part of the head and mandibles, which are brown; the body is moder- ately curved, the anterior end thicker. The head is almost entirely retracted into the thorax. The postcephalic segments have prominent transverse folds that bear a few short bristles. The number of instars comprising the larval stage has not been determined but, whatever it proves to be, all have the general features stated above. In their normally curved posture, these robust grubs range from about 1.0 mm. to 3.0 to 5.0 mm. in length in the course of growth from hatching to maturity. The variation in size of the full-grown larvae emerging from the hips is to some extent associated with latitude, or, more specifically, perhaps with differences in climate. For example, both the larvae and adults reared from Rosa carolina and blanda taken at Chetek, Wisconsin, are foto} INSECTS IN ROSE HIPS visibly smaller than those from R. carolina grown at Urbana, Illinois. The cuticle of the larvae is pearly white, but the body of the first instars appears reddish after they have fed on achenes. As they approach maturity, the larvae accumulate an increasingly larger amount of whitish adipose that lends the body a dull chalky appearance. Small numbers of larvae were discovered within the egg pit in the hypanthium, from which location some migrated to the achene cavity. The larvae from eggs deposited in an achene commonly made the achene their nidi and fed on the contained seeds. These afforded food enough for the first instar, for the exuviae were sometimes found in the achenes. Sometimes also second instars occupied the achenes in the same manner, at least for a part of the stadium. Not infrequently, the aperture chewed in the occupied achene by the ovipositing female was so obscured with frass that the larvae were not readily discovered. The fact that these small larvae leave the shells of the achenes intact, except to enlarge the exit holes, is evidence that they ingest chiefly the nutrient dicotyledons within. Also when eggs are laid in the concavity shared by two adjacent achenes, the small larvae eat the seeds. Small larvae sometimes feed for a short time on the inner face of the hypanthium. These individuals seem to originate from embryos located in the inner extremity of the egg pit. From these several initial egg sites, the growing larvae gradually eat their way deeper into the cavity, creating ever increasing amounts of a sawdust-like pale brown frass. This certainly consists in part of hard, dry bits of ripe rose seeds and in part of excretal pellets, but largely of particles of achene shells broken up in the process of feeding. Conse- quently, the large larvae are often found deeply embedded in a mass of frass and whole achenes deep in the center of the basal half of the achene cavity. The achenes present in any one hip vary considerably in number; hence some hips may conceivably contain food matter inadequate to enable a larva to realize its growth potential. In such cases, the larva might complete its growth by feeding on the inner face of the hypan- thium, thereby perhaps causing the dark discolored areas sometimes visible externally on the hip. Mortality. Some fatalities occur among both the embryos and the larvae of the rose curculio. Occasional embryos may die from two possi- ble causes: first, from being encased in the solidified protoplasm that has oozed from the ruptured tissues along the wall of the egg pit; second, the female beetle may break some of her eggs as she pushes them deeper into the pit with the hard, pointed proboscis. Some dark, dead embryos and collapsed chorions have been found that suggest these causes. With the earlier instars, I have repeatedly suspected cannibalism where two or more black-covered punctures occurred in a hip but only THE ROSE CuRCULIO, Rhynchites bicolor (¥FAxR. ) 89 one larva was found, although I have obtained no good direct evidence to support this theory. However, ectoparasitism by chalcid larvae of Eurytoma spp. and Zatropis rosaecolis unquestionably accounts for some failures to find curculio larvae where black covers indicated they should be present. More persistent searches occasionally disclosed the shrunken dead dark bodies of small Rhynchites obscured amid frass in the hip cavity along with full-grown larvae or pupae of the parasites or their remains. Although the endoparasitic larva of the ichneumonid, Luchatema baldufi, sometimes inhabits large percentages of the host larvae, including the smaller instars, it does not cause mortality while the host resides in the hip. Emergence. The full-grown larva of Rhynchites works its way through the frass in the hip cavity to the inner face of the hypanthium, and makes a hole, usually through the basal half of the hypanthium, to escape. En route, the larva ingests at least some of the usually red hypanthial ma- terial, as indicated by reddish substance in the food canal of some newly emerged larvae. Emergence is sometimes heralded by the appearance of an external discolored patch above the spot in the hypanthium where the larva will issue. How long the emerged larvae remain above ground in nature was not observed. However, numerous individuals removed from hip jars and placed on soil in flower pots were seen to lose no time in burrowing into it. The larger larvae that had been out of hips in jars as long as 24 hours remained vigorous enough to bore down promptly when soil was provided. Influence of temperature. While removing emerged larvae from hip jars in the falls of 1945 to 1947, I gradually became aware of an un- expected relation between emergence and atmospheric temperatures. Instead of increase in warm days and decrease on cold days, it was found that a much greater number of larvae emerged from hips at lower than at higher temperatures. To check on this interrelation of emergence and temperature, I secured large quantities, usually one to several thousands per year, of hips from wild Rosa blanda and R. carolina at Chetek, Wisconsin, early in September of 1948 to 1951. Brought to Ur- bana, Illinois, the hips were stored each fall in jars kept under approxi- mately natural temperatures, where the mature larvae emerged in Sep- tember and October. Day-by-day records were kept on the numbers emerging during most of these yearly periods. The daily crop of larvae was removed from the jars at the close of the day. The temperatures recorded below pertain to an approximate 24-hour period preceding the daily removal of the emerged larvae. (See Table 2, p. 90.) To illustrate this relationship of temperature and emergence, data 90 INSECTS IN ROSE HIPS TABLE 2. RELATION OF TEMPERATURE TO EMERGENCE OF LarvaL Rhynchites bicolor bicolor (September 14 to October 6, 1948 and 1950) Number Per cent of 2,626 Extremes of temperature of larvae larvae issued at Low High emerged various temperatures 34-40° F. 55-69° F. 676 25:15 41-47° 61-75° 809 30.81 48-54° 64-83° 829 Sead 55-61° 64-90° 233 8.87 62-69° 74-94° 79 3.00 2,626 100 per cent (Table 2) were chosen for the period, September 14 to October 6, for both the years 1948 and 1950, when daily records were taken consistently. Kept in temperatures approaching those of natural conditions (an open shelter enclosed with wire screens), the lots of hips yielded a total of 2,626 mature larvae. In the table, the emergence records are grouped into five categories graduated chiefly in accordance with the range of the low extreme temperatures. It will be readily understood why these low ranges are not quantitatively equivalent to the spread shown by the high extremes, since day and night as well as daily temperatures in nature are subject to inconstant or differential variation. Briefly interpreted, the percentile data in Table 2 indicate that the large majority, or 88 per cent of the 2,626 larvae, emerged from the hips at relatively low temperatures of 34 to 54° F., and only 12 per cent at the higher levels of 55 to 69° F. It appears also that temperatures around freezing, or 34 to 40°, are not quite so effective as activators as 41 to 54°, yet are greatly more so than temperatures of 55 to 69° F. The daily records, involving 44 days in the 2 years, demonstrate the pattern of alternation or fluctuation between peaks and troughs in the phenomenon of emergence more patently than the data of Table 2. From these it seems evident that the major activity of the larvae in the hips is of two comparatively distinct types. One, they engage in their prime function of feeding-growing at the higher temperatures, and as a conse- quence, a fraction of them becomes full-fed, mature, and conditioned for emergence. The number of emergence-ready larvae accumulated during this phase depends on the duration of the higher temperatures that favor development. Two, the matured larvae so accumulated in the hips engage in the alternate activity of emergence when the atmosphere of night, or fall days, cools decidedly. However, there is rarely observ- THE ROSE CURCULIO, Rhynchites bicolor (FARR. ) 91 able a complete line of demarcation between the above two types of activity, ie., neither the feeding-growing nor the emergence function is completely inhibited, except possibly in late fall. Very few, if any, larvae fail to attain the emergence-ready condition on account of pro- hibitory low temperatures. Two instances are cited here from the daily records that reflect the alternating character of the larval activity in an impressive way. First, on the four days, September 18 to 21, 1948, when the 24-hour periodic temperatures ranged high, from 60 to 94°, only three larvae emerged: in the next four days of comparatively cool weather, September 22 to 25, temperatures varied between 44 and 77 degrees, and 335 larvae issued from the same hips. Second, in the three days, September 20 to 22, 1950, at extreme temperatures of 58 to 85°, 143 larvae issued: September 23 to 25, when temperatures fell to 37-72°, 766 larvae emerged from the same hips. It may then be deduced that a species that exhibits two such opposite characteristics can have evolved, and is able to perform successfully, only under climatic conditions characterized by alternation of periods of lower and higher temperatures. Accordingly the basic reason for this associa- tion of two antipodal behavioristic characteristics lies in the fact that the immature larvae and the mature larvae are physiologically two distinct entities despite their conspecificity. The first is geared to feed and grow in higher temperatures, and is congenial to a limited dark space within a hip, a situation favorable to growth. The mature larva no longer feeds, and seeks the large open spaces by emerging, to adopt an entirely differ- ent mode of life thereafter. In fact, this “mature larva” is to be regarded as a prepupa, with physiological characteristics distinct from those of the immature larva. Weights of mature larvae. More than 2,000 mature larvae available in the fall of 1950 and 1951 were utilized in a study of their weight char- acteristics. One thousand eight hundred twelve larvae emerged from about 1,800 hips gathered from Rosa blanda and R. carolina at Chetek, Wisconsin, on September 5, 1950, and 392 were reared from 2,262 hips obtained from the same sources on September 6, 1951. The daily yields were weighed collectively within one to 24 hours after the larvae emerged. The total mass of 1,812 larvae reared in 1950 amounted to 26.307 grams, and the average weight per individual larva was 0.0145 gram. The cor- responding data for the 392 larvae of 1951 were 5.528 grams and 0.0141 gram. Thus, the average weights of individuals for the two years were very nearly identical. Of greater interest, however, is the steady, progres- sive reduction in these average weights during the 49-day season of 1950, September 8 to October 26. The average individual weight of larvae 92 INSECTS IN ROSE HIPS that emerged in the first week was 0.0159; in the final week, it had de- creased to 0.0119 gram. The weighed larvae of 1951 issued in eleven weeks. The average larva that emerged in the first week weighed 0.0156, whereas that of the tenth week weighed 0.0118 gram. In this year, the weekly decline was less regular than in 1950, and the weekly samples consisted of only 5 to 95 individuals, as compared with 15 to 901 in 1950. How may this seasonal reduction in weight of individuals be accounted for? The larvae feed almost entirely on seeds. These are ripe by Septem- ber 1, and remain in good preservation in storage. Thus, the condition of their food is probably not an important factor in the over-all reduction in the average weight of the larvae. Instead, it is likely that the progres- sive lowering of average temperatures as the fall season advances, favors less and less of feeding with resulting smaller larval bodies, while at the same time, this cooler average weather stimulates the emergence of the larvae. Their smaller size and lower weights indicate that many may emerge before they attain the maximum of which they are capable under warmer conditions. Gas in the digestive tract. My notes contain references to 23 instances in which varying amounts of a clear gas were discovered in the larvae of the rose curculio. Still other cases were not recorded. Where the gas was specifically located, it occurred in the crop, which is a large bulbous sac in this species. When the inflated crop was punctured, a series of small bubbles escaped into the water in the dissecting dish. Larvae of all ages, from first instars that had taken some food, to mature individuals recently issued from the hips, contained gas. In all cases, the larvae affected contained some food stuff in the crop. Larval life in the ground. If one may make a deduction from their leglessness and their ineptness in locomotion on smooth surfaces, the lar- vae in natural environment probably fall rather than crawl to earth. Many of my newly emerged larvae were placed on black Illinois soil in four-inch flower pots. Here they began promptly to bore in, head first. Upon examining the “cages” weeks to months later, the larvae were found to have burrowed to depths of one to four inches, the large major- ity halting at depths of one and a half to two inches. In nature, the depth of penetration is obviously determined by the hardness of the soil, which may be so firm that exploratory crawling is required to find a penetrable place; or descent into the soil may actually be impossible. Larvae that were later removed from soil in the pots had constructed oval cells that had a smooth inner surface, but not a well-defined periph- ery. The cells were formed of particles of soil cemented together with a substance that doubtlessly issues from the mouth, and is probably THE ROSE CuRCULIO, Rhynchites bicolor (¥ABR. ) 93 glandular in origin. It permeates the surrounding earth to a thickness of several millimeters, being more compactly fused on the inner face and more loosely adherent toward the periphery of the cell, so that the external layer remains rough and crumbly. The over-all cell wall is about 3 mm. thick. Under the somewhat artificial conditions of the soil pot, overstocked sometimes with 100 to 200 larvae, the cells lie so close together that the soil which surrounds them may become fused into a dense, firm, im- pervious, imporous gummy mass. Such a condition of the soil seems to induce a state of oxygen deficiency in the larvae, for when first removed, although appearing normal in color and form, they have been observed to remain inactive, as if dead, but resume activity on exposure to the atmosphere for an hour or so. Larval diapause. Concealed within their cells, the larvae lie in the ground for eight to nine months, from about September to May. This was established by unearthing larvae from soil pots during November to June. In that time, they undergo no change in external appearance or internal condition so far as gross physical observations can disclose, and appear to lose little weight, as indicated by weighing a small sample unearthed after four weeks. The digestive tract is empty and collapsed from the beginning, excepting sometimes a small amount of red hypan- thial matter ingested as the larvae chew their way out of the rose hip. The body cavity is full of a grayish white soft adipose substance. In its disintegrated form in the pupa, it constitutes numerous small spherical particles. In submature and younger larvae, these are united to form large slab-shaped bodies, all of which are bound together to constitute one mass. All the other viscera are obscured or enveloped by this body, only the digestive tract retaining bulk sufficient to be readily identified. This inert condition, or diapause, is exhibited by all mature larvae emerging from mid-August to November. Hence, the diapause is of a physiological type initiated during the larval life in the hip. This is the prevailing form of suspended development in the life of Rhyn. bicolor. Prolonged diapause. The prolonged type of diapause first came to my notice when I scrutinized the soil in certain pots on July 14, 1947, and discovered nine mature larvae of Rhyn. b. bicolor still present a full month after emergence of adult curculios is usually completed. Dissec- tion of the nine showed that seven individuals contained the parasitic larvae of the ichneumonid, Luchatema baldufit, which apparently in- hibited the pupation of the hosts. However, the other two lacked para- sites and remained normal in color, form and size. The nine larvae emerged September 14 and October 27, 1956 from a quantity of hips gathered at Chetek, Wisconsin, that same fall. They had 94 INSECTS IN ROSE HIPS wintered out of doors in soil pots, and were returned to the open-air shelter in early May for observation on development. The two retarded, parasite-free mature larvae raised the question whether such individuals die sooner or later, pupate and become adult out of season, or may persist through a second winter and transform to adult in the usual spring season but one year late. In the hope of obtaining the answer to these questions, I employed the larvae from another lot of hips picked at Chetek in September of 1948. The larvae were placed in ten soil pots, and set out-of-doors for the winter. Adult Rhyn. b. bicolor developed in these pots, and duly emerged in June of 1949. Inspection of the soil in two of the pots in the fall of 1949 revealed a few healthy larvae, several months after trans- formation was normally completed. These tardy larvae were replaced in a soil pot and kept under observation along with the other eight un- disturbed soil pots of that set. The nine pots were set out into natural weather for the second winter, 1949-50, then returned to an open-air shelter for the following spring, when the soil was wetted thoroughly with water as needed. Fifteen normal adults emerged from the nine pots in the period of June 6 to 9, 1950, demonstrating that, under some conditions, the mature larvae do persist for an entire additional year and then change to pupae and adults in the usual period of transforma- tion in June. Despite the prolongation of the larval life, the species con- forms to the same seasonal distribution of the developmental stages as it does in the usual annual life cycle. Thus, the adult makes its appear- ance at the essential budding-fruiting time of the host plant. Prepupa and pupa. Along in May, both the normal and prolonged diapauses of the overwintered larvae are terminated. The exsertion of the head from its previously retracted position in the thorax is the first external indication that the prepupal body is formed within. The ap- proach of the final larval molt is also betokened by the loosening and crumpling of the cuticle and the change of the body from the curled to a straight form. Internally the slabs of fat body disintegrate into a homo- geneous mass of creamy white particles. When the larval cuticle is shed after prepupation, all the external structures of the white pupal body are seen to have formed. The pupal cuticle bears a sparse set of bristles which appear to be arranged in constant positions (Figs. 37, 38). They occur as 3 pairs on the dorsum of the snout; 3 transverse rows on the pronotum, comprising a row of 13 on the anterior margin, a row of 4 on the disc, and a row of 13 on the hind margin; a total of 8 bristles on the disc of the mesonotum, ar- ranged in two transverse rows of 4 each; bristles on the disc of the metanotum are the same as those on the mesonotum. The dorsal abdominal segments have bristles on the rounded trans- THE ROSE CuRCULIO, Rhynchites bicolor (¥FABR. ) 95 verse ridges as follows: first tergite, 12 bristles, with six on either side of the midline; second, 10, with groups of 2 and 3 on each side of meson; third, 10; fourth, 14, the median two small, lateral ones much larger; fifth, 12, in two series of 6 each; sixth, 10; seventh, one bristle on each caudolateral corner; eighth, 10, in an irregular transverse row; ninth, or terminal tergite, with an arcuate row anterior to the pair of strong, brown, caudal tapering spurs which are subparallel and directed dorso- caudad. There is also a pair of minute bristles on each of the humeral bulges of abdominal segments IV, V, and VI. The “knee” of each leg bears two apical bristles directed laterad. Venters of thorax and abdo- men are bare. Life cycle. Data from field and laboratory studies show conclusively that Rhynchites b. bicolor completes its life cycle once in about 12 months, excepting the minority of the population that may require two years due to a prolonged diapause in the mature larval state. Moreover, one cycle per year appears to characterize other varieties of the species wherever its development has been studied. (1) Cycle at Urbana, Ilinois. The statement below is based on data, and a few estimates, derived from several seasons of work with particular reference to the life of the curculio in Rosa carolina. The diapausing mature larvae in their hibernacula in the ground be- gin pupation about May 5, and complete the transformation by June 18. Emergence of adults from the ground extends from May 20 to June 18; they continue abroad to August 7, the last date adults were seen. The over-all adult season is therefore about 80 days. The occurrence of newly laid eggs in hips shows that oviposition extends from June 10 to August 7, or about 59 days. Adding 10 days for embryogenesis gives an over-all embryonic or “egg” period of 69 days. The earliest first instars were found in hips on June 20, and the latest emergence-ready mature larvae occurred about October 15. Since the first larvae issued from hips on August 15, and such were still present in soil pots as late as June 8 of the next spring, the full-fed last instars have an over-all diapause plus wintering stage of approximately 298 days, or 10 months. When the 117 days of over-all life in the hip are added to the diapausing period, the rose curculio proves to be present, in one or another of its larval instars, a total of 415 days. Being more than a year, this means the earliest first instars may have hatched in June before the latest wintering larva has pupated in the same June. Pupae were shown to occur in soil pots between May 5 and June 18, or for 44 days. Estimating the pupal period as 15 days, adults may appear as early as May 20. (2) Cycle at Chetek, Wisconsin. Although the life cycle of the rose curculio at Chetek probably differs from that described above for Ur- bana only in the initial and concluding dates of the several stages, a few 96 INSECTS IN ROSE HIPS significant data should be given here. The 1,517 larvae obtained from hips of Rosa blanda and carolina early in September, 1950, at Chetek, were distributed chronologically in 13 soil pots, as they emerged at Urbana between September 13 and 25. At intervals during the period October 29, 1950, and July 19, 1951, the insect contents of sample pots were removed and identified to note their developmental status. The essential results follow. Samples scrutinized in October and November contained only mature larvae in their hibernacula. In ever decreasing numbers, larvae persisted to July 13, 1951, some of these being parasitized by the ichneumonid, Luchatema baldufit, which destroys the mature larval host. Pupae. oc- curred in the sample pots from May 4 to June 7, and adults were found only in the pots examined on June 7 and 15. The last five of the 13 cages, explored between June 21 and July 19, had yielded their adults, but contained numbers of the parasite, which varied from a few first instars in the host to many pupae in cocoons, and fewer adults not yet emerged from the soil. (3) Cycle in Atlantic states. Additional data bearing on the life cycle of Rhyn. b. bicolor in the eastern United States are contained in the re- ports of Gates (1909) and Dickerson (1910). According to Gates, the adults were first noticed at Worcester, Massachusetts, late in May; they became more common in June and still remained numerous on July 10, when the observations were discontinued. At New Brunswick, New Jersey, Dickerson discovered that the adults appeared in May, and con- tinued active through June and July. Oviposition persisted into late June, when feeding and copulation also were noticed. Hatching began there in the first part of July, and most eggs had hatched by mid-July, and no unhatched ones were found on July 22. Some larvae had matured by August, and many had become full-grown soon after mid-August. Very few immature individuals remained in early September, and most larvae had left the “seed capsules” at that time. Marcovitch (1916) investigated an undetermined form of Rhyn. bi- color in Minnesota, probably in the vicinity of St. Paul. He first discov- ered the adults puncturing the buds of Rosa rugosa on June 10, and found eggs deposited within the hips on June 28. These hatched in 6 to 10 days. The larvae reached maturity about the middle of September, when they entered the soil and pupated the following spring. Rhynchites bicolor in the West Five of the six varieties of Rhyn. bicolor Fabr. recognized by Pierce (1913) appear to be more or less western in their distribution. These are Rhyn. b. wickhami Cockerell, Rhyn. b. cockerelli Pierce, Rhyn. b. ven- tralis Pierce, Rhyn. b. piceus Pierce, and Rhyn. b. viridilustrana Pierce. THE ROSE CURCULIO, Rhynchites bicolor (¥FABR. ) 97 The types of another variety, Rhyn. b. erythrosoma, described by Cock- erell and Harris (1924), were taken at Boulder, Colorado. Excepting Hoerner (1936), who investigated the life of Rhyn. b. wickhami in Colo- rado, the writers have not specified the varieties with which they were concerned. The bionomic differences in Rhyn. bicolor as reported by Cooley (1903 ) in Montana, Lovett (1915) in Oregon, Robertson (1923) in Manitoba, and by Hoerner (1936) in Colorado may prove to coincide with some of the varieties described by Pierce. Following is an abridgment of Hoerner’s (1936) description of the metamorphic stages of Rhyn. b. wickhami. Adult. Thorax and elytra bright red; venter of body and femora, beak, tibiae, tarsi and antennae black. The original description (Cockerell, 1912) gives characteristics of sculpture, form and vestiture. Beaks of twelve males averaged 1.82 mm. in length; of fifteen females, 1.95 mm. Egg. Pearly white, elliptical, finely pitted or reticulated; average size 1.24 x 1.1 mm. Larva. Newly hatched, white, legless, robust, tough, about 442 x %; inch when coiled; full-grown, about 44 x 4% inch, stramineous, subcylin- drical, broadest through postcephalic segments three to five, tapering bluntly at ends; body strongly wrinkled; head retractile, its anterior third brown; pronotal shield pale, obscure; thoracic legs represented by enlarged setose tubercles; thoracic and abdominal segments sparsely and finely hairy; spiracles on all segments except 2, 3 and 12. Pupa. White at first, about 14 inch long by 1% inch across mesothorax; a pair of curved appendages on caudal end of abdomen. Hosts of western Rhyn. bicolor. Published reports of Cooley (1903), Lovett (1915), Robertson (1923), and Hoerner (1936) indicate that, in their areas, Rosa is the only plant group in which the larva of the rose curculio develops. Cooley (Montana) was never able to find larvae except in rose hips, and he concluded “this is doubtless the normal place for their development.” While Lovett (Oregon) found eggs in the flower buds of blackberry, few hatched, and none of the larvae fed in the dead buds, indicating that blackberry is not a host plant. On the other hand, Lovett emphasized that the larvae feed in the “seed pods,” ie., hips of rose. In Manitoba, Robertson found the larvae feeding on seeds in the hips of some cultivated roses and in native species Rosa acicularis, blanda, Macounii and suffulta (pratincola). The larvae of Rhyn. bicolor wickhami, the variety which Hoerner investigated in Colorado, develop in the mummified flower buds of roses. In my own investigations, I have reared larvae of Rhyn. bicolor from hips sent to me from three places in the west. These presumably repre- 98 INSECTS IN ROSE HIPS sent varieties other than the typical form, Rhyn. b. bicolor. The hosts and places were: Rosa acicularis, lunelli, alcea, pyrifera and Macounii from Saskatoon, Saskatchewan; R. woodsii, Mapleton, and Mt. Timpanagos, near Provo, Utah; and R. ultramontana and spaldingi, Moscow, Idaho. Excepting the discovery of eggs in the flower buds of blackberry, by Lovett (1915), all the above records of host plants indicate that the larvae of western varieties of Rhyn. bicolor feed and develop in the hips, or in the mummied flower buds of roses. The exceptional case of black- berry may be questioned until it can be established whether or not the larvae can develop in the flower buds of this berry plant. In the light of these evidences, we are probably to interpret other reports of feeding on blackberries, raspberries and loganberries as having reference to the adult curculio, and, if so, these berry species are to be regarded as food plants, not as hosts. Correction. Reference is made here to the short essay by Craw (1894) under the title “Attelabian Snout Beetle, Rhynchites bicolor Fabr.” This is obviously a confusion of facts about two species of snout beetles. The description of the adult seems to refer to an Attelabus sp., which lays eggs in small nests made by the female on leaves of trees, including oak. The statement that the beetle frequently punctures the ripe fruits of blackberries and raspberries, causing them to decay, presumably refers to Rhynchites bicolor. But the assertion that “this beetle is very destruc- tive to grapevines” in portions of Europe does not apply to Rhyn. bicolor. Chittenden (1901) quoted Craw on the injury to blackberries and rasp- berries as having been done by Rhyn. bicolor. Adult: food and damage. The western forms of Rhyn. bicolor include a wider variety of rosaceous plants in their dietary than do their eastern relatives. While roses appear to be the principal food, and thereby suffer the greater injury, related plants of the genus Rubus, also are sometimes utilized and injured. The curculio is known to attack petioles, leaf buds, flower buds, flowers and their expanded petals and fruits of both Rubus and Rosa. However, some of the reports do not specify the parts eaten or the nature and extent of the resulting damage. When associated with roses, the adult western Rhyn. bicolor appears to prefer the buds, boring holes into them, with the consequence that the petals do not open. Robertson (1923) counted as many as 20 punc- tures in a single rose bud. Hoerner (1936) adds that whereas the beetle mostly penetrates only the petals, it sometimes bores through both the petals and sepals of the bud, or even its base. Some punctured buds open, he relates, exposing the riddled petals. Another frequent injury results from cutting the flower stem, or petiole, just below the calyx, so that the buds wilt and die, and drop to the ground. One or more of the above THE ROSE CURCULIO, Rhynchites bicolor (¥FAB8R. ) 99 types of feeding attack are reported by Lovett (1915) for Oregon, Rob- ertson (1923) for Manitoba, Hoerner (1936) for Colorado, Riley and Howard (1891) and Cockerell (1894) for New Mexico, and Chittenden (1901) for Montana, New Mexico and California. According to Lovett (1915), leaf clusters are taken as food in Oregon. Hoerner (1936) adds, regarding the use of vegetative parts of the rose, that young shoots may be eaten off, or the stem punctured so that the tips will curl and dry, when flower buds are not plentiful. The newly emerged young beetles usually began feeding on young growing tips of the rose shoots, eating the tender leaves or stems, or the rose buds. Then as the flower buds on the early varieties became scarce the beetles fed on later varieties. Little notice is given in the published reports to feeding and injury to rose hips; perhaps for the reasons that some of the western varieties of the rose curculio may not oviposit in hips, hence do not puncture them, and that the achenes have had little commercial value because of infre- quent use in propagation of roses. Lovett (1915) recorded that the curculio punctured rose hips in Oregon. The injuries to roses by the adult curculio may have serious conse- quences for the grower of roses. Comments from the literature are cited to this effect. In Montana, the beetle was the cause of numerous com- plaints from rose growers in 1923 and 1924 (Cooley, 1903, 1925). At Santa Fe, New Mexico, it was “very troublesome” about 1894 (Cocker- ell), and in Manitoba it did considerable damage to roses about 1923 (Robertson). Hoerner (1936) states that the damage by Rhyn. b. wick- hami was so severe in some sections of Colorado for a number of years that it was almost impossible to grow roses out-of-doors. That intensity of natural light affects the amount of injury by the beetle was observed by Robertson (1923). The wild roses, R. blanda and R. acicularis, growing under trees were infested by the curculio to the extent of 29.6 per cent, whereas when growing exposed to more sunlight, 60 per cent were infested. When associated with species of Rubus, western varieties of Rhyn. bicolor feed in the same manner and on the same parts attacked when feeding on Rosa; namely, leaf buds, flower buds and stems, or the berries. Mentioned in this connection are raspberry and blackberry, and their hybrid loganberry, and thimbleberry, with the first two cited most frequently and widely. Specific instances are: injury to blackberry in Colorado (Cassidy, 1888); serious injury to blackberry, raspberry and thimbleberry sometimes done in Oregon (Zeller and Schuh, 1944); and ripe fruits of blackberry and raspberry punctured, causing them to decay in California (Craw, 1894). Lovett (1915) describes the injury to black- berry more fully, for Russellville, Oregon. One to several beetles oc- 100 INSECTS IN ROSE HIPS curred on almost all clusters of flower buds, and affected one-half to two- thirds of them, causing them to turn brown and die as a result of numer- ous “feeding and egg punctures.” Also the stems for a short distance below the bud were dead, and the buds drooped over. In Manitoba, injury to blackberry and raspberry was negligible (Robertson, 1923). Hoerner (1936) did not find wickhami feeding on raspberry in nature, but the beetle ate the buds when no rose buds were available in cages. Mating, fighting, and locomotion. Mating occurs at any time of the day, and each male and female may mate several times with different individuals (Hoerner, 1936). This observer, on several occasions, also saw males fighting, raising their beaks above their opponents and bring- ing them down with considerable force. When disturbed, the adults draw their legs to the body, depress the beak, and thus usually fall off the plant, where they lie motionless for a time. Cooley (1903) observed that the beetles are shy and soon disappear under a leaf or stem when a person approaches. Lovett (1915) described them as very sluggish, dropping readily when their activity is interrupted. Oviposition and larval life. The accounts of these activities of Rhyn. bicolor in the west differ particularly with reference to the parts of the host plant wherein the egg is placed and the larva feeds and develops. In Montana, Cooley (1903) found the eggs placed in various parts of the rose plant. Most of them were either in the unexpanded petals of the flower bud, or in the young fruit. He also discovered an egg in the tender extremity of a new rose cane, and one in a cynipid gall. Pre- sumably the gall grew on the rose plant. In all cases the eggs lay in the holes made with the beak, and well below the surface. A hip that contains an egg “shows a blackened scar on the side which marks the spot where the parent beetle bored in to deposit the egg.” He observed also that the beetle bored holes into the stems at right angles to the axis, causing the buds to wilt and dry, and hang from the stems. I quote Cooley's next statement in full below because it emphasizes what appears to be a dis- tinct difference in the larval niche of this insect in Montana as compared with that of the larva in some other areas of the northwest. “We have not been able to see any particular significance in the bor- ing of holes into the stems, though when we began the studies, it was thought possible that the buds were caused to wilt and dry for the pur- pose of preparing a suitable food for the young. Though very many such buds have been broken open and examined, we have never found a larva feeding in one.” Later in the article Cooley added the following about the larva. “We have never been able to find larvae except in the rose hip or fruit, and this is doubtless the normal place for their devel- THE ROSE CuRCULIO, Rhynchites bicolor (FAR. ) 101 opment ... They feed upon the seeds which fill the greater part of the cavity of the fruit. The fleshy coating of the fruit is not eaten so far as we have observed. Examination of a fruit containing a nearly full- grown larva shows a part or all of the seeds excavated to mere shells and the body of the larva buried in a mass of waste and excrement.” Lovett (1915) seems to have made no search for the eggs and larvae in punctured rose buds, but did discover that “the larvae feed in the seed pods,” “devouring the seeds and leaving little but the shell.” According to the report by Robertson (1923) for Manitoba, the eggs of Rhyn. bicolor are placed in punctures in buds of roses, generally one to each bud. The resulting larvae “live for a short time within the flow- ers and then go to the hips, where they are found in late August feeding on the seeds.” Hoerner (1936) has given us the most complete account of the process of oviposition and larval life. The curculio bores “egg punctures” in the flower buds of roses, and deposits the eggs into the punctures. She like- wise punctures the stem that supports the bud, causing both bud and stem to wilt and die. The eggs were found deposited at various depths from just beneath the surface to about 2.0 mm. below the surface of the buds. The eggs sometimes lay under the outside petal; again under four or five petals. A few had been placed inside single petals. With wild roses, or in buds with few petals, the eggs may be introduced directly among the sta- mens. Rarely, the eggs occur in the base of the bud; all the rest were in folded petals, but here no definite position seemed to be preferred. Rarely more than one egg is found to have been deposited in a bud un- der field conditions. When the egg has been placed into the puncture, the opening is in- variably plugged with bits of petal chewed off the edges of the punc- ture, as Hoerner has observed, and can then seldom be detected except as a slight scar, or sometimes as a small protuberance on the bud, or swelling on a petal. As observed by Hoerner (1936), the actual number of eggs laid varied from 16 to 143 per female; the average number deposited by 64 individuals was 37. The largest number laid by one female in 24 hours was 8. The larvae of Rhyn. b. wickhami hatch and develop to maturity within the dead rose buds as these either hang on the bush or lie on the ground. Under natural conditions, the small larvae are in most cases found among the stamens, and, Hoerner states, pollen probably is their main food when they are first hatched. “Later, when the larvae become two- thirds or almost full-grown, they are practically always found in the base of the dried bud.” 102 INSECTS IN ROSE HIPS The discovery, by Lovett (1915), of eggs in the buds of blackberry suggests that Rubus may be a second host of a western variety of the rose curculio. However, his observation that “the young grubs did not feed at all” in the buds, leaves the question of hostship open to doubt. More thorough investigation is indicated. Summary: oviposition and larval life. Western forms of Rhyn. bicolor usually place their eggs either into the flower buds of rose, or into the rose hip, and larvae variously inhabit the buds only, the hips only, or both buds and hips. For example, Robertson states that the eggs are placed in the bud, where the larvae reside for a time, then move to the hips. Cooley reported finding eggs in the “unexpanded petals,” or rose buds, but larvae only in the hips. The careful studies by Hoerner re- vealed both eggs and larvae are confined to rose buds. Moreover, both eggs and larvae of the eastern Rhyn. bicolor have been found only in the hips of roses. Thus we have a graded series of variations with reference to the spe- cific sites of the eggs and the larvae. This condition suggests more in- vestigation to determine whether the observed variations, and possibly others still remaining unknown, may be correlated with forms of the rose curculio. The bionomic facts already at hand may justify recog- nition of two or more species in the taxonomic entity now treated as Rhynchites bicolor. Such a division was indicated by Green (1920) on the basis of certain structural differences noticed among western varieties. Behavior of the mature larva. The earlier observers, Cooley (1903) and Lovett (1915), learned that the full-grown larvae leave the rose hip. Cooley’s discovery of openings in hips that no longer contained larvae indicated to him that the larvae “eat out of” the hips and go to the ground. But his supposition that they pass the winter as pupae has since been shown to be in error. Also, Lovett concluded that they leave the hips when he found in late August and September that the inner por- tions of the “seed pods” had been devoured and the invaders had dis- appeared. Robertson (1923) stated positively that the mature larvae bore their way out of the hips, fall to the ground and bury themselves about 1.5 to 4.0 inches below the surface of the soil, where they hiber- nate. Hoerner (1936) learned that the mature larvae “leave the buds and enter the ground, where they pass the winter in small earthen cells about 6 mm. in diameter and about 14 to 2 inches below the surface. Larvae were dug up from the soil beneath rose bushes in October, No- vember, February, March, April and early May.” By this means, he proved that the matured larvae constitute the wintering stage of Rhyn. b. wickhami. THE ROSE CuRCULIO, Rhynchites bicolor (¥FABR. ) 103 Pupation. Cooley (1903) and Lovett (1915) were unsuccessful in finding the pupae because they looked for them in the fall season. Robert- son (1923) stated that pupation begins in the middle of May in Mani- toba. Hoerner found pupae in the earthen cells made by the larvae in the ground. “They habitually lie on their backs, and can move around quite readily in their cells by means of abdominal movements, aided by the caudal spines.” By about the second day of pupation, the com- pound eyes show as brown spots on the otherwise white pupa, and in five or six days more the eyes, mandibles, tarsi and caudal spines be- come light brown. These, and other parts, gradually darken until the mature adult state is reached. Emergence of the adult. To determine the age of the adult beetle before it works its way from the pupal cell to the surface of the ground, Hoerner (1936) secured pupae that were about ready to transform into adults. He placed them in the bottom of a small glass vial, covered with a pupal cell and an inch of dirt. Thus he learned that “it takes the adult four or five days to become fully colored and hardened, and a day or two to work its way to the surface.” The beetle appeared unable to emerge through dry soil, but worked its way out in a day through moistened soil, and began feeding soon thereafter. Life cycle. The data supplied by investigators of the western forms of Rhyn. bicolor agree either in indicating, or in stating expressly, that this species has one generation in a year. However, Hoerner (1936) dug a few mature larvae from the ground in April which did not pupate as expected but persisted all the following summer and winter as larvae. Thus, some individuals of the western forms, like their eastern relative, may enter upon a larval diapause of unusual duration. In Montana, Cooley (1903) observed that the beetles appeared on rose bushes early in June, and continued there until the latter part of August. He estimated that the egg stage lasts about 7 to 10 days, and speculated that the emerged larvae leave the hips and go into the ground, trans- form to pupae, and pass the winter as such. But he probably erred in regard to the stage of wintering. Near Russellville, Oregon, Lovett (1915) found the adults abroad and ovipositing in flower buds of blackberry on May 27, 1912. He made several unsuccessful attempts to work out the life history, and therefore gives the following “approximate life history.” The adults are present and active from late spring until late summer and early fall. Eggs were found from May until late July. “The egg stage is about twenty days.” Larvae from very small to full-grown sizes were found in rose hips in late August and September, while larvae matured and emerged from others in the same period of time. Although Lovett failed to find pupae 104 INSECTS IN ROSE HIPS at this time, he predicted that the adults emerge in spring from over- wintering pupae. On the basis of his incomplete knowledge, he believed the species to be “single-brooded” in Oregon. In Manitoba, Robertson (1923) noted that the adults begin to appear in the early part of June, and are most abundant at the middle of this month. Hatching of the embryos begins in the second week of July. The larvae, which first feed within the flowers, are found in the hips in late August. From about the fourth of September, the larvae bore their way out of the hips and enter the ground, where they hibernate. In the middle of the next May, pupation begins, and after two weeks the first adults emerge to appear on the rose bushes. These facts indicate one generation per year in Manitoba. Hoerner (1936) showed that Rhyn. b. wickhami has one generation in a year in Colorado. His findings cover a number of bionomic features of the rose curculio for which no data were previously available. The adults emerge from the ground during the latter part of May and the first part of June. The average length of life of the emerged beetles was found to be about 7 weeks; 109 individuals lived an average period of 41.78 days. Males lived longer than females, or an average of 43.06 and 40.72 days, respectively. Two females lived the unusual period of 97 days. Embryonic development required about 10 days, but the period varied considerably with environmental temperature. In the latter part of May, the eggs hatched after about 11 or 12 days; in the warmer weather of July, they did so in 8 or 9 days. The extreme periods of incubation were 8 and 15 days, the average 10.47 days. However, it appears that some factor in addition to atmospheric temperature helps to determine the duration of incubation, for Hoerner found that eggs deposited on the same day varied as much as four days in the period required for embry- onic development. The mouthparts, especially the mandibles, of the embryo can be seen moving inside the chorion about 36 hours before hatching. By exposing rose buds to laying females for a day and keeping these buds until the larvae emerged from them, Hoerner obtained values of 93 and 142 days for the developmental period of the larva. Inhabited buds collected in the field in early summer yielded mature larvae from July 16 to October 6. These facts were taken to indicate that the grow- ing phase of the larval period may be as short as 45 or 50 days. Hoerner (1936) showed that in Colorado also the mature larvae pass the winter, since he dug them from the soil beneath rose bushes, between October and early May. Hence, the over-all life of the full-grown larvae may extend from July of one year to May of the next. Pupation required about 9 days in April and early May in Colorado, NATURAL ENEMIES OF Rhynchites bicolor (ABR. ) 105 and the resulting adults emerged from the ground about two weeks later, in late May and the early part of June. NaturaL Enemies oF Rhynchites bicolor (Fasr. ) To date I have only one reference to natural enemies of the rose cur- culio from the literature. This pertains to an ichneumonid determined by R. A. Cushman as a species of Temelucha. Hoerner (1936) found it parasitized three per cent of the larvae of Rhynchites bicolor wickhami in Colorado. In the course of my investigation of Rhyn. bicolor bicolor, five species of hymenopterous parasites were reared. The most numerous of these was the ichneumonid, Luchatema baldufi, a solitary endoparasite con- fined to the larval host. The lesser four were solitary ectoparasites on the larvae. These comprised two species of the family Braconidae, Bracon mellitor and B. variabilis, and two Chalcidoidea, Zatropis rosaecolis, and Eurytoma sp. I have also obtained specimens of an oligochaete annelid, determined by Dr. L. J. Thomas, University of Illinois, as a species of Chaetogaster. These were associated with immature stages of the rose curculio un- earthed from soil pots at Urbana. Dr. Leland Shanor, then mycologist, University of Illinois, cultured the fungus Beauveria globulifera (Speg.) Pic from mature larvae of the curculio which I submitted to him for study. Mites of the species Caloglyphus spinitarsus (Herm.), Acaridae, were commonly present in soil pots that housed larvae and pupae of the rose curculio or contained the meconia of L. baldufi, but performed in the role of scavengers. Luchatema baldufi Walkley, Ichneumonidae, Hymenoptera In the years 1946 to 1951, inclusive, I accumulated a series of 151 adults—62 males and 89 females—of this parasite. Most of them were reared from the larvae of Rhynchites bicolor. A smaller number of them were captured on rose bushes in Urbana. In addition to these, the col- lection of the U.S. National Museum contained a number of specimens accumulated in previous years. Luella M. Walkley (1956), specialist in the Ichneumonidae at the National Museum, found that the above materials represent “a genus and species of Tersilochini new to the de- scribed Nearctic fauna.” Distribution. The specimens in the national collection originated in Massachusetts, New York, Iowa and South Dakota. Those obtained in Massachusetts were taken from flowers of Physocarpus, Rubus, and Hydrangea grown in the Arnold Arboretum, Boston. This observation 106 INSECTS IN ROSE HIPS indicates the adults feed on pollen and nectar. To date, the larva of the rose curculio, Rhyn. bicolor is the only known host. Here it is a solitary endoparasite. I have reared the species from the rose curculio contained in rose hips available to me in Urbana from Saskatoon, Saskatchewan; Mapleton and Mt. Timpanagos, Utah county, Utah; Ely, Minnesota; Solon Springs, Chetek, Rice Lake, Black River Falls, Endeavor and Madison, Wisconsin; Crete, Nebraska; Rochelle, San Jose, Bath, Velma, Rising, Mahomet, Urbana, Mayview, and Philo, Illinois, and Jamestown, Indiana. The above records indicate a wide distribution of Luchatema balduft over the northern United States and southern Canada from at least the Rocky Mountains to the Atlantic Ocean. Adult (Fig. 39). The distinctive characteristics given here are from Miss Walkley’s original description of the species. Only enough char- acters are selected to distinguish the adult from other Ichneumonidae reared from rose hips. Length of females, 4.3 to 7.0 mm.; males, 4.2 to 7.0 mm. Body black, except clypeus, mandibles, palpi, legs, abdomen and ovipositor, which are reddish brown, with the antennae, legs, mandibles, clypeus and mouth parts more testaceous. Head transverse, broader than long and broader than thorax. Thorax at least seven-eighths as high as long, polished, pubescent, and finely but densely punctate; hind basitarsus nearly as long as remaining tarsal segments combined. Abdomen pol- ished, lacking pubescence, with long slender petiole about two-thirds length of thorax; ovipositor about as long as body, slender, upcurved, sheaths blackish with reddish tinge, slender, sparingly hairy and ridged. Through the day and until sunset, I have seen adults on wing, zig- zageing at a fairly fast speed among the tops of the rose hedges on the campus at Urbana. Now and then they alighted to crawl over leaves and hips as if exploring for prospective hosts. On some short visits to the hedge, seven to nine individuals were seen in June. Mating and feeding were not observed. Oviposition. Although always alert to discover the females engaged in oviposition, I witnessed this procedure only once. The sun had just set on July 4, 1949, at Urbana. This individual clung with her legs to a rose hip that bore an egg puncture of Rhynchites bicolor. The three pairs of tarsi were set close together, not spread widely, and the body stood out almost vertically from the hip over the egg puncture. The terebra, with- drawn from its sheath, was held parallel with the venter of the body and directed cephalad so it extended straight into the egg pit. Then the abdomen vibrated vigorously sidewise while bringing pressure on the terebra as it bored into the pit. About one minute was spent in NATURAL ENEMIES OF Rhynchites bicolor (¥FABR. ) 107 drilling into the puncture, after which she remained quiet for a few seconds, while the egg presumably passed through the terebra into the host. The terebra was then withdrawn and replaced in the sheath. In less than a minute, this adult repeated the above ovipositional process on a second egg pit on an adjacent hip 1.5 inches from the first. The two operations combined required 4.5 minutes. The two hips concerned here were dissected 1.5 hours later. Each contained a first instar of Rhynchites bicolor, and each larva inclosed a whitish, smooth, apetiolate egg about 0.35 mm. long with no trace of embryo. The form is shown in Fig. 44. In each larva, the egg had been inserted into the body cavity. These, and other eggs of this parasite, discovered while I dissected other host larvae, lay unattached among the viscera of the host, from which they were readily disentangled, and sometimes popped out when the host’s body wall was opened. To check on the possibility that the egg may be inserted into the host egg, I dis- sected also eggs of the curculio during the season of parasitization. All results were negative. If Luchatema baldufi paralyzes the host incident to oviposition, the inactivating effect is of short duration and slight intensity. The above larvae were normally active 1.5 hours after parasitization. Host instars attacked. Discovery of parasite eggs in many small larvae of the curculio dissected has caused me to believe that mostly first and some second instars are attacked by the ovipositing female. It is very likely that selection of these instars is determined by their location at or near the inner end of the egg pit of the curculio. In other words, these prospective hosts lie in a direct line with the egg pit where the terebra is certain to strike them when it is inserted into the hip. More advanced larvae are likely to escape parasitization for the reason that they move from the egg site as they feed and grow, thereby reaching a location that is out of line with the introduced terebra. Reproductive capacity. Indications of the capacity of L. baldufi were obtained from dissection of a gravid female taken in nature on June 20. Each ovary consisted of 18 ovarian tubes. Each tube contained nine eggs sufficiently developed to permit identification. Thus, this individual was potentially capable of inserting 324 eggs into host larvae. Moreover, some eggs may have been deposited before the female was captured, and others might have developed had she been permitted to live. Since the species is proven to be essentially solitary, two eggs per host being in- frequently found, the potentialities for reduction of the host population by a single female may exceed 400. Each ovary has a cup-shaped egg calyx at its base and is joined to the median oviduct by a slender lateral oviduct almost as long as an 108 INSECTS IN ROSE HIPS — ovarian tube. The fact that the ova did not swell while they lay in cold water for a half hour indicates an impervious chorion. Larval instars. On July 4, 1949, I found a larva of the parasite with its caudal process still enveloped in the broad end of the chorion. The para- site then was about 0.35 mm. long. How many instars compose the larval stage of this Luchatema was not determined. But, like many other endoparasitic Hymenoptera, the present species has a very distinct first instar, both structurally and bio- nomically. The fact that this instar occupies about nine-tenths of the annual lifetime of its species, accounts for the relatively large amount of data presented below concerning it, as compared with that for subse- quent instars. I first became aware of the unusually prolonged first stadium in 1943, while investigating the rose seed chalcid, Megastigmus nigrovariegatus at Urbana. Incidental to that study, I discovered what were obviously first instars of a Hymenoptera species in larvae of the rose curculio in hips of Rosa rugosa. Dissections were continued all that summer in the hope that the advanced forms of the parasite would be found. But only first instars were present. It was not until the spring that the adult, re- quired for specific identification, was reared. That this is the normal course of development in this species has been confirmed again and again in subsequent years. When newly hatched in the body cavity of the host, this slender, apodous caudate parasite is about 0.35 mm. long and one-eleventh as wide (Fig. 40). From July to October, it grows to a maximum length of approximately 2.0 mm., with a comparable diametrical enlargement (Fig. 41). In the early developmental phase of the instar, the bases of the head and caudal process are as broad as the thorax and abdomen; but as the instar grows, these anterior and caudal parts distend but little, whereas the intermediate region enlarges from 3.0 to 3.5 times, yet remains almost cylindrical in shape. Thus, in its advanced or mature phase, this turgid instar resembles a rolling pin in form, with the narrow head and the caudal process representing the handles. Excepting these changes in sizes and form, the first instar possesses essentially the following structural characteristics throughout its life: head pale brown, sclerous, about 1.5 times longer than thick, with minute subterminal ventral mouth parts; postcephalic parts white; thorax and abdomen (exclusive of caudal process) moderately depressed, subcylin- drical, and about uniform in diameter, composed of 13 segments which vary little in size, and each constricted transversely at the middle so it appears double; particularly distinctive are the two pairs of udder-like, fleshy pseudopods, one pair on the ultimate abdominal segment, the other on the penultimate segment; the pseudopods appear to be at- NATURAL ENEMIES OF Rhynchites bicolor (FABR. ) 109 tached to the outer ventral edges of the segments, and are capable of elongation and shortening; abdomen terminates in a fleshy but subrigid, tapering, downcurved process; on the dorsal side, at the base of the caudal appendage, is a marginal notch, beneath which the blind hind intestine terminates. In addition to the small to larger spheres of yellow food substance, the mid intestine infrequently contains bubbles of gas. In one such case, a first instar 1.5 mm. long contained three largish bubbles amid the golden food in the posterior half of the tract. The slender form of the first instar seems well adapted to move freely among the semifluid viscera of the host. But if one may judge from ob- servations on larvae in water, their movements are largely nonlocomotor. The extent of these activities varies decidedly with the degree of bodily dilatation; in its younger slender phase, the instar is flexible and changes posture easily, but as it becomes more and more distended during growth, it becomes less agile, more turgid and stiff, and eventually tends to be lethargic and motionless until disturbed. In the later, less active phase, the specific gravity decreases to the extent that the larva, newly removed from the host, floats lightly on water. This condition facili- tates finding the parasite among the viscera of the dissected host. From the normal linear posture, the young flexible instar commonly assumes two other positions. In one instance, it turns the cephalic and caudal ends sharply and quickly dorsad, thereby taking the form of a capital letter U, but with the caudal process raised higher than the cephalic end. It also goes easily into the shape of an attenuated letter S, and may simultaneously roll over and over sidewise in a brisk manner. It is especially while the larva performs the lateral roll that the ab- dominal pseudopods are elongated and projected to be best seen, al- though only for an instant. Growth of first instar. The larva of Luchatema baldufi remains a first instar from early summer of the year it hatches from the egg until late spring to mid-July of the next year, or as long as a year to 13 months. During the period indicated above, it undergoes two functionally dis- tinct phases—feeding-growing and diapausing. First instars totaling 461 removed from host larvae of various ages showed growth from 0.35 mm. at hatching in June and July, to 1.3 to 2.2 mm. in length during mid- August to November. The first mature hosts emerged from hips about mid-August, the last so late as early December, in Urbana. The para- sites of the larger sizes, and removed from hosts in the later months, had completed their growth, and assumed the form of a rolling pin. This range in size does not signify varying degrees of immaturity, for larvae of like variations occurred in the hosts during the ensuing months of winter, spring and early summer, i.e., to the time of their first molt. 110 INSECTS IN ROSE HIPS Persistence of mature first instars through three seasons indicates they have entered diapause of an intense character that is probably induced by some inherent physiological factor. The first instar appears to persist in the diapause so long as the fat body of the host retains the form of whitish slabs normal to all the in- stars. The normal disintegration of the fat body into the puréed state appears to coincide with the advent of prepupation. The breakdown of the fat body and the first molt of Luchatema may therefore prove, with more intensive study, to be precisely concurrent events and actuated by one and the same causative agent—possibly the pupational hormone of the host. The unusually prolonged feeding-growing phase of this parasite has prompted me to suggest what may be the cause. Since the mouth parts are obscure and inoffensive, they, along with the nature of the food ingested, may explain the slow growth. During its entire feeding phase, the instar seems to ingest the yellow blood plasma of the host, and small globules of free fat. Small amounts of the plasma flow from all host instars when their body walls are punctured. The voluminous white semisolid slabs, which fill the mature host larva snugly, appear at no time to be disrupted by the parasite, and the nature of the substance present in the digestive tract indicates it ingests no fat body. If so, the fluids the parasite swallows may be so small in amount and so lacking in nutritive value that growth can be made only at a slow rate. Advanced larva. The development of the parasite from the time of its first molt to its full-grown state is certainly relatively brief, for it requires only a few days as compared with the 8 to 12 months devoted to the first stadium. This surprising acceleration in rate of growth may perhaps be due to two further developments in the host-parasite relation. These are the more simplified but efficient suctorial mouthparts of the advanced parasite, and the puréed state of the adipose of the host. These modifica- tions appear to combine to accomplish a single result, namely the quick voluminous transfer of the host’s inner substance into the gut of the parasite. In this way, the parasite quickly attains its full size while the host is killed as it is robbed of its large bulk of creamy, homogenized fat body. The number and distinctive characteristics of the advanced instars of Luchatema baldufi have not been determined. As the larva progresses from the second instar to the last, it increases in length from about 2.0 to 5.5 mm. (Figs. 42, 43). The body is then elongate-oval, its plumpness or form varying with the amount of food previously ingested; more or less depressed or sagged, both ends blunt, the prominent caudal process of the first instar now greatly reduced to lacking; head white like thorax and abdomen; body appears 13-segmented until dilatation from feeding NATURAL ENEMIES OF Rhynchites bicolor (FARR. ) 111 obliterates the intersegmental folds; cuticle of some larger larvae bare, shiny, finely wrinkled, while that of others appears finely granulated; these differences in sculpture may indicate two separate instars. On a larva 2.0 x 5.5 mm., I observed a small tubercle, probably spiracular, on the anterior half of each pleuron of segments 4 to 11. Small white sub- cuticular speckles of adipose visible externally on living larvae also prob- ably indicate the last instar. The two pairs of pseudopods present on the abdomen of the first instar have not been seen on the advanced larvae. Cocoon, pupa and emergence of adult. The cuticle of the dead host is filled snugly by the full-grown parasitic larva and is then papery and fragile so that it is readily sloughed off as a crumpled mass. This lies in the posterior end of the host’s cell in the ground in late spring. Now the parasite proceeds to spin its cocoon. These events mark the beginning of the prepupal stage, during which the parasite retains its external larval form while the external aspect of the pupa develops within. The cocoon (Fig. 45) generally fills the host’s cell completely and there- fore reflects approximately the form and size of this cell. In the period of May 30 to July 19, 1951, I measured 136 new cocoons, whose hosts originated at Chetek, Wisconsin. Their average length was 5.24 mm., their average maximum diameter 3.19 mm. The two smallest cocoons measured 4.4 x 2.2 and 4.2 x 2.7 mm.; the largest, 6.08 x 3.80 and 6.25 x 4.00 mm. Fifty-eight cocoons that contained female pupae averaged 5.49 x 3.33, while 34 spun by male larvae averaged 4.85 x 2.94 mm. The small- est female cocoon was 4.5 x 2.4, the smallest male 4.4 x 2.2; while the largest were 6.25 x 4.00 and 5.65 x 3.65 mm., respectively. The measure- ments of the 136 specimens also showed a considerable variation in the ratio of length to widest diameter, i.e., differences in form or shape. The cocoons consist of two distinct silken layers of different textures, an outer one of dull white loose meshwork, and an underlying one with a smooth inner surface and composed of thin sheets resembling onion skin in texture. Into the outer layer may be woven a transverse median band of whiter silk of variable width, and sometimes also a longitudinal ridge of silk along each lateral surface. Frequently I found “lifesaver”-like rings of golden color in the wall of the cocoon, especially in the middle of the upper side. At first a cause of perplexity as to their nature and origin, they proved to be droplets of water evaporated from the prepupa and the meconium within. Some were suspended from the inner surface of the cocoon, others enmeshed in the external webwork. They varied in size, and adjacent ones were, by pres- sure of a needle on the cocoon, caused to merge to form larger drops. They were present in largest numbers when the occupant or material were soft and moist, but few or lacking when the advanced pupa had developed a more impervious cuticle. 1B INSECTS IN ROSE HIPS The termination of the prepupal phase is heralded by the advent of reddish lunate compound eyes visible through the cuticle on the pupal head within the prepupal pronotum; also by the decided thickening of the abdominal area to an elongate-oval form, and by the elimination of the meconium, this discharge taking place shortly before the ecdysis of the prepupal cuticle. When eliminated, this accumulated waste has the form of a half disc (Fig. 45a), its anterior face flat and the posterior side convex to conform to the posterior inner wall of the cocoon where it lies. It takes various positions, from almost vertical to different degrees of obliquity. It is a dark brown, thick soupy substance when newly eliminated, and the end of the pupal abdomen may become inserted into it. The discharge of this waste leaves the prepupa, and the new pupa, entirely white, excepting the eyes. Shortly after defecation, the shedding of the prepupal cuticle reveals the pupa. Its head and thorax are at first pure white, excepting the brownish red compound eyes, whereas the abdomen is white tinged with yellow. Particles of adipose are visible through the semitransparent, flexible membranous venter of the abdomen. The sheaths of the ovi- positor curve dorsad around the apex of the abdomen and extend for- ward along the dorsum of the entire body, including the head. The cocoons were found at various depths; empty cocoons found four inches below the surface indicate that the young adults can emerge from this depth of soil. When cocoons containing fully developed young adults were unearthed and opened, the insects sometimes proved able to use their legs for running, and some also took to wing, although awk- wardly, as soon as they were released. Extent of parasitism. The records involved here were obtained in the years 1946 to 1951, by dissecting host larvae of all instars from several species of roses growing in localities in Saskatchewan, Utah, Wisconsin, Illinois and Indiana. The random sample concerned was processed only during the summer and fall in which the host larvae developed; the parasites were therefore all first instars. A total of 790 individual first in- stars were discovered in 1888 larvae of Rhynchites bicolor. Because mul- tiple parasitism in this case was found to be so infrequent as to be almost negligible, the over-all sample was parasitized to the extent of about 42 per cent. However, the degree of parasitism varied decidedly as to place and years, as the following instances show: Chetek, Wisconsin, 1946-51, 55 per cent; Madison, Wisconsin, 1946-47, 36; Urbana, Illinois, 1947-51, 42; Mayview and Philo, Illinois, 1947-51, 15; and San Jose, Illinois, 1951, less than one per cent of 417 hosts dissected. Superparasitism. As defined by Imms (1931), superparasitism is acci- NATURAL ENEMIES OF Rhynchites bicolor (¥FABR. ) 113 dental, not obligatory, in the instance of Luchatema baldufi. Of 790 larvae of Rhyn. bicolor parasitized by first instars, very few contained more than one individual. In five cases discovered, the host larvae had matured and remained alive. In four instances, one of the two parasites remained alive. The four living individuals were first instars, whereas three dead ones were first instars, and the fourth was a second instar, with the exuviae still adhering to the body. In the fifth instance, both parasites were first instars. One living larva measured 1.8 mm., whereas the associated dead one was 0.8 mm. long, and, moreover, enveloped in a large mass of transparent gelatinous substance, possibly phagocytes. Another dead parasite had become greatly inflated, probably by gas of decomposition. None of the dead parasites bore evident marks of injury or loss of form that might indicate cannibalism. Life cycle. The records on which this account is based were secured by dissecting first instars from host larvae, and by rearing the advanced instars, pupae and adults from hosts placed in pots of soil. The hips con- cerned here grew at Chetek, Wisconsin, and Urbana, [linois, and were processed at Urbana. The approximate period of emergence of young adults from the hiber- nacula of the host larvae in soil is indicated by records for 1949 and 1951. Thirty-two adults issued between June 28 and July 19, 1951, and 53 between June 24 and August 13, 1949. Accordingly the females are abroad parasitizing the smaller instars of the rose curculio in hips from June to late August. That embryogenesis requires at most a few days, is indicated by the infrequent discovery of eggs in small hosts parasitized by very young first instars of Luchatema. Small, recently hatched first instars were discovered by dissection of host larvae as early as the first week of July. Growing very slowly, these first instars attained their mature state from late summer to October. The mature first instars are then in physiological diapause, which con- tinues through fall, winter and spring while the full-fed, emerged host larva lies in its hibernaculum in the ground. Exuviation of the first instars appears to be concurrent with the prepupation of the host, in the period of approximately early May to early June. Thus, the over-all period of the first instar of Luchatema extends from late June of the first year to mid-July of the second involved in the cycle. The advanced instars complete their development in probably less than a week, and simultaneously, by feeding, invariably destroy the host before it can pupate. Advanced larvae were found in soil pots be- tween May 15 and July 20, 1951. The parasite was in its pupal form from May 29 to July 19. Adults then emerged between June 24 and August 13—the period when small first and second instars of Rhynchites are available for parasitization. 114 INSECTS IN ROSE HIPS Therefore, like its host, Luchatema completes one cycle per year, its successive life stages lagging, timewise, behind those of the host. Bracon mellitor Say, Braconidae, Hymenoptera Synonymy and distribution. Of four species of solitary ectoparasites found attacking the larvae of Rhynchites bicolor in rose hips, Bracon mellitor Say, named and described by Thomas Say (1836), was the most numerous. Subsequently it was placed under the name Microbracon (Muesebeck, 1925), and later relocated in Bracon (Muesebeck, et al., 1951). In his revision of 1925, Muesebeck showed that the names xantho- stigma Cress., vernoniae Ashm., anthonomi Ashm., and pembertoni Brid- well are synonyms of mellitor Say. The literary sources of these names are given in the Synoptic Catalog, p. 166 (Muesebeck, et al., 1951). According to this catalog, the species occurs in the United States, Mexico and Hawaii. My series of 43 females and 11 males appeared in jars containing rose hips from Chetek and Madison, Wisconsin; Ironton, Missouri; Crete, Nebraska; Fox Ridge Park and Kickapoo Park, Illinois, and San Jose, Philo and Mayview, Illinois. Hosts. Species of Bracon prefer hosts from which the ovipositing adults are “separated by an inert covering” (Salt, 1931). In its relation to Rhyn- chites bicolor, B. mellitor conforms to that pattern of behavior. Ento- mological publications show that mellitor has also been reared from numerous other hosts of like habits in the Coleoptera and Lepidoptera. The names of the hosts, with the literary sources, so far as known to me, follow herewith. COLEOPTERA CURCULIONIDAE Anthonomus grandis Boh. (Carpenter, 1944; Fenton and Dunnam, 1929; Folsom, 1936; Hixon, 1935; Hunter and Hinds, 1904; Marlatt, 1933; Miller and Crisfield, 1930; Muesebeck, 1925; Muesebeck, Krom- bein, Townes, 1951; Pierce, 1908; Smith, 1936; Strong, 1934, 1935). Anthonomus albopilosus Dietz (Pierce, 1908; Muesebeck, 1925; Muesebeck, Krombein, Townes, 1951). Anthonomus eugenii Cano (Muesebeck, 1925; Muesebeck, Krombein, Townes, 1951). Anthonomus fulvus Lec. (Muesebeck, 1925; Muesebeck, Krombein, Townes, 1951; Bigger, 1931, 1933; Satterthwait, 1946, 1948). Anthonomus signatus Say (Chittenden, 1893; Muesebeck, Krombein, Townes, 1951). Anthonomus squamosus Lec. (Pierce, 1908; Muesebeck, 1925; Muese- beck, Krombein, Townes, 1951). Desmoris scapalis Lec. (Pierce, 1908; Muesebeck, 1925; Muesebeck, Krombein, Townes, 1951). NATURAL ENEMIES OF Rhynchites bicolor (FAR. ) 115 Desmoris constrictus (Say), (Satterthwait, 1946; Muesebeck, Krom- bein, Townes, 1951). Hypera nigrirostris (Fabr.), (Underhill, 1924; Muesebeck, Krombein, Townes, 1951). Coeliodes inequalis Say (Brooks, 1918). Cylas formicarius elegantulus (Summers), (Cockerham, 1944). Chalcodermus aeneus Boh. (Bissell, 1940). Tachypterellus quadrigibbus Say (Hammer, 1936). Conotrachelus nenuphar Harris (Quaintance and Jenne, 1912). LEPIDOPTERA AEGERIIDAE Aegeria scitula Harris (Underhill, 1935). GELECHIIDAE Anarsia lineatella Zell. (Basinger, 1935). Frumenta nundinella (Zell.), (Montgomery, 1933; Muesebeck, Krom- bein, Townes, 1951). Pectinophora gossypiella Saund. (Bridwell, 1920; Willard, 1927; Ohlendorf, 1926; Muesebeck, 1925; Muesebeck, Krombein, Townes, 1951). OLETHREUTIDAE Crocidosema lantanae Busck (Bridwell, 1920). Cryptophlebia illepida Butl. (Bridwell, 1920). Polychrosis viteana Clem. (Johnson and Hammer, 1912; Muesebeck, 1925; Muesebeck, Krombein, Townes, 1951). Cydia molesta (Busck), (Smith, 1928; van Steenburgh, 1934). Rhyacionia frustrana (Comst.), (Cushman, 1927). PYRALIDAE Homeosoma _ electellum WHulst (Satterthwait and Swain, 1946; Muesebeck, Krombein, Townes, 1951). Myelois ceratoniae Zell. ( Bridwell, 1920). Pyrausta nubilalis Hubn. (Baird, 1929). Pyrausta penitalis Grote (Chittenden, 1918). COSMOPTERYGIDAE Mompha stellata Busck (Dickerson and Weiss, 1920; Muesebeck, Krombein, Townes, 1951). NOCTUIDAE Heliothis obsoleta Fabr. (Bibby, 1942). Stibadium spumosum Grote (Satterthwait, 1948; Muesebeck, Krom- bein, Townes, 1951). 116 INSECTS IN ROSE HIPS Direct evidence that B. mellitor attacks the grubs of Rhyn. bicolor came in the discovery of a live adult male in a hip of Rosa carolina from San Jose, Illinois, in association with much shrunken dead remnants of a larva of the curculio amid feces and pulverized achenes. In a number of other hips from the same locality, I found similar dead remains of the curculio and a few tough cocoons, such as the larvae of mellitor are said to spin. Extent of parasitism. The incidence of Bracon mellitor on the rose curculio was generally low, and certainly under one per cent, except at Chetek and San Jose. Where the cotton boll weevil, Anthonomus grandis, was the host, investigators agree that this was the most numerous para- site present, but, on the other hand, they concur in the view it does not effect a desirable degree of control. Parasitism of the boll weevil by all species found by investigators caused 8.22 per cent mortality at Madison Parish, Louisiana, in 1932, due principally to B. mellitor (G. L. Smith, 1936), and Strong (1936) stated that about 80 per cent of all the para- sites reared from the boll weevil in the cotton-growing area were B. mel- litor. In Georgia, the average parasitism by mellitor, in 1930, increased from 10 to 18 per cent in bolls in June, to 35 per cent in squares on the plants in August and September (Miller and Crisfield). In 1904, 24.4 per cent of 5,548 squares collected in Texas produced adult weevils, but only 1.3 per cent of the population was parasitized, 90 per cent by Bracon mellitor (Hunter and Hinds). Miller and Crisfield (1930) also reported an observation, in which Pierce, Fenton and Dunham, and Folsom con- cur, that parasitism by mellitor in bolls and squares on the plant is higher than in fallen ones. Bracon mellitor occurred in 65 per cent of the heads of sunflower (Helianthus) examined by Satterthwait (1946) at Lovington, Illinois, where it parasitized Anthonomus (“Desmoris”) fulvus. Bigger (1933) reported that 80 per cent of the parasites reared from the same host insect and plant in 1928 to 1932 were B. mellitor. However, the highest rate attained was only 6.2 per cent, in 1929. When attacking the oriental fruit moth, Cydia molesta, B. mellitor constituted only 0.4 per cent of the parasites reared in Michigan in 1930 to 1932 (Merritt, 1933), and was obtained from that host only in small numbers in Ontario, in 1926 and 1927 (C. W. Smith, 1928). Willard (1927) considered it the most valuable parasite against the pink bollworm (Pectinophora gossypiella) in Hawaii, yet only 2.33 per cent of this host were parasitized by it in 1918 and 1919. Adult life. The adult is fairly robust, with dusky filiform antennae, shiny reddish brown head and thorax, orangish yellow abdomen, promi- nent pterostigma, venation typical of the family. The female bears sturdy NATURAL ENEMIES OF Rhynchites bicolor (¥FABR. ) IN sheaths around the terebra, which is about two-thirds as long as the body. According to Willard (1927), length of the body varies much, depending on the amount of food available to the larva from the host. From small pink bollworms, adults not over 1.5 mm. long have been reared, but the usual length of those developing on this host is 4.0 mm. Similar variations in size have also been reported by other observers; e.g., W. D. Pierce (1908) obtained “a small melanistic form” of B. mellitor, which had previously gone under the name Bracon dorsator Say. Willard (1927) and Folsom (1936) showed that the length of adult life varies with sex, food and temperature. “Unfed” males averaged three days of life, and females five days (Folsom). Females fed on water and honey survived 24 to 102 days (Willard), and averaged about two weeks during August (Folsom). After the insects emerge from the cocoon, mating often takes place at once, and may be repeated ( Wil- lard). It is, as a rule, performed in 15 to 30 seconds. The procedure is described by Willard. However, insemination is not essential to repro- duction. Both Willard and Folsom showed that the progeny of unmated females was invariably male. When Folsom mated these males with females, normal eggs and larvae resulted. Oviposition. Deposition of eggs usually begins on the second day after eclosion (Folsom), and often not for 5 to 7 days (Willard). The female chooses the place for oviposition by the tactile sense. A cotton square that contains an advanced weevil larva yields to the touch of the terebra because it is hollow. A hard, resisting spot makes the terebra bend, whereas a soft one is penetrable, and an egg is placed here. Since soft spots indicate larger weevil larvae, the parasite chances to prefer large hosts (Folsom). The attacking female always stings and paralyzes the pink bollworm before she deposits an egg. As the terebra momentarily pierces the cuti- cle, it injects enough poison to cause paralyzation. After some time, the host becomes capable of only slight movements. An egg is deposited on it as soon as it grows quiet (Willard). According to Fenton and Dun- nam (1929), the larvae of the boll weevil are always completely and permanently paralyzed, but not killed at once. The pulse beat slowed in one to 3 days after paralysis. Folsom found the pulse of a normal larva increased with activity from 42 to 60 pulsations per minute, whereas it varied from 24 to 30 per minute after paralysis. Eggs are commonly attached anywhere on the body, but sometimes to the wall or lint of the host’s cell. The egg is pearly white, smooth, attenuated behind, broadly rounded at the head, curved, and 0.866 to 1.256 mm. long. Deposition of eggs required an average of 15.4 minutes during August and September (Folsom). Three confined females laid 116, 121 and 213 eggs, respectively (Willard). Both these investigators 118 INSECTS IN ROSE HIPS discovered that the females often went through the motions of oviposi- tion, including actual paralysis, without delivering an egg. One female observed by Willard deposited 63 eggs in 4 consecutive days, and later paralyzed 9 larvae on which no eggs were laid. Some special circumstances have been observed to affect parasitiza- tion. The female will not attack host larvae exposed outside the cotton bolls. Squares hanging normally on the plant are parasitized at a higher rate than others lying on the ground in the field. By placing pupae of the boll weevil in cells designed for oviposition, Folsom showed the fe- males can distinguish between pupae and larvae, avoiding the pupae. Carpenter (1944) performed experiments involving B. mellitor and the boll weevil, which showed that the number of hosts available to single females affects their productivity. When a female was caged with 20 larvae of Anthonomus grandis, the average yield per female was 7.5 adult mellitor, but when 5, 10, 15, 30 and 60 females were placed with 20 larvae, the production per female decreased to 2.85, 1.92, 1.47, 0.74 and 0.41 adult mellitor, respectively. Larva. The mature embryo hatches by biting away the anterior end of the chorion with its pointed mandibles. The newly hatched larva is 1.1 to 1.4 mm. long, translucent white, with a distinct head and 13 post- cephalic segments. Willard (1927, pp. 8-11) describes and figures four instars, which are similar and hymenopteriform. The mandibles end in a sharp tip that effectively punctures the body wall of the host. The young larva travels freely over the smooth body of the cotton boll worm with a crawling-looping movement. It begins feeding by lacerating the cuticle with its pointed jaws, and so sucks fluids from the host. The new first instars may frequently move to new locations on the same host. This travel incidentally brings the parasite into contact with other larvae of its kind that may be present. The resulting combat reduces the num- ber to one, and thereby assures an ample food supply for the survivor. However, two larvae may mature on a large bollworm (Willard). Only the skin and the head of the boll weevil larva remain when the parasite is full-fed. Differences in size of the host larva are reflected in the “enormous” variation in size of the resulting adult B. mellitor (Fol- som ). Cocoon and pupa. The cocoon is spun inside the square on the wall of the host’s cell. Folsom fully described the process of cocooning. He and Willard give diverse values for the duration of the prepupal stage; these seem to result from their different concepts of this phase of life. The prepupa discharges the accumulated digestive waste within the cocoon. Exuviation of the prepupa requires a few minutes, and the newly exposed pupa is glistening white, except for the reddish brown NATURAL ENEMIES OF Rhynchites bicolor (¥FABR. ) 119 eyes. The duration of the pupal stage varies from 3.5 to 5.5 days in temperatures that ranged from 71 to 86° F., in Hawaii. The adult emerges by biting a hole in the end of the cocoon ( Willard). Life cycle. The incubation period averaged 24 hours in August and September (Fenton and Dunnam, 1929), a little less than 24 (Folsom, 1936) and 18 to 28 hours in Hawaii (Willard, 1927). At Tallulah, Louisi- ana, the feeding-growing period of the larva required only three days, as a rule, but four were sometimes needed in September and October (Folsom). At Florence, South Carolina, development from hatching to cocooning in July and August involved three days (Fenton and Dun- nam). Willard determined the duration of the four instars to be 23 to 24, 20 to 23, 28 to 30, and 19 to 115 hours, respectively, in Hawaii. The temperatures varied between 68 and 80° F. The following differing values given by investigators are attributed to the varied numbers of life phases encompassed by their accounts. According to Willard, the egg, larval, prepupal and pupal periods lasted 28, 187, 28 and 132 hours, respectively, or a total of about 16 days. Fenton and Dunnam found that 10.19 days were required for development from oviposition to eclosion, and Folsom reported the normal period from hatching to eclosion, in August and September, as 12 days. The first three of the 12 days were spent as a feeding larva, the next 6 “as a larva” in the cocoon, and 3 days as a pupa. According to Folsom, oviposition ended in nature about mid-Novem- ber, 1934, when frosts killed the weevil larvae in the squares. Some of the cocoons contained living larvae of Bracon mellitor in the fall, which indicated the parasite can probably winter as such. My data for B. melli- tor from rose hips suggest the parasite winters as an adult, and completes its cycle once per year when Rhyn. bicolor, a univoltine host, is attacked in Wisconsin and Illinois. Bracon variabilis (Proy.), Braconidae, Hymenoptera The taxonomy of this species is treated by Muesebeck (1925; 1951, Synoptic Catalog). The literature contains very little information on its bionomics. Distribution and hosts. The above references give the distribution of B. variabilis as Quebec to Virginia, and west to Missouri and Texas. Other reports have it from Ontario (Garlick), Michigan (Gentner), Kansas (Parker and Lamerson) and New York (Leonard). I have it from Urbana, nearby Mayview and Kickapoo State Park, Illinois. Larvae of Lepidoptera and Coleoptera that develop hidden within parts of plants serve as hosts of B. variabilis. From the above literature and other sources cited here, I have the following host records: grape berry- moth, Polychrosis viteana (Clem. ), Olethreutidae, Pennsylvania; pine tip 120 INSECTS IN ROSE HIPS moth, Rhyacionia frustrana var. bushnelli Busck, Olethreutidae; Blasto- dacna sp., Cosmopterygidae; plum curculio, Conotrachelus nenuphar Herbst, Curculionidae, Virginia; apple curculio, Tachypterellus quadri- gibbus Say, Virginia and T. q. magnus List, Curculionidae; and an un- identified larva in seed pods of Oenothera, Virginia. The above host records lead me to suppose that, in rose hips, B. vari- abilis parasitized the larvae of Rhynchites bicolor and/or Cydia packardi. Both these phytophags occurred in the hips of Rosa carolina in the localities where my series of eight adults originated. Six of them ap- peared alive in hip jars between May 15 and June 11, and two on Sep- tember 10 and 21, in the years 1946, 1947, 1949 and 1951. Zatropis rosaecolis Burks, Pteromalidae, Hymenoptera This parasite on the larva of Rhynchites bicolor was described re- cently as a new species by B. D. Burks (1955a) from a series of 3 males and 33 females which I obtained from rose hips in Illinois, northwestern Wisconsin and northeastern Minnesota. Adult. The adult may be distinguished from other chalcids from rose hips by the following characters taken from the original description by Burks (1955a). Female—Length 2.0-4.0 mm. Head and body black with faint bluish- green or blue metallic coloration visible from oblique angles; antennal scape yellow or tan, shading to dark brown at apex, flagellum dark brown or black; tegulae and wing veins light brown; coxae black, femora very dark brown with apices yellow, each tibia yellow at base and apex, shaded with brown in the middle. Face clothed with numerous short hairs. . . . Thoracic dorsum with a few scattered, golden-yellow hairs. “Male—Length 1.5-2.0 mm. Color of head and thorax without metallic sheen.” Dr. Burks comments that Z. rosaecolis is the only North Ameri- can species of its genus which has the face clothed with simple hairs and has two rows of bristles behind the marginal vein on the underside of the fore wing. Distribution and rose hosts. My series of 79 adults, including the above 36 types, consists of 64 females and 15 males. They originated in hips of roses in localities as follows: Rosa carolina, Ironton, Missouri; R. rugosa, Urbana, Illinois; R. carolina at Urbana, Mayview, Philo, Velma, and San Jose, Illinois; R. Macounii, Crete, Nebraska; R. arkansana, Madison, Wisconsin; R. blanda and R. carolina, Chetek, Wisconsin; R. blanda, Solon Springs, Wisconsin; and Rosa acicularis bourgeauiana, near U.S. highway 61, in northeastern Minnesota. Rhynchites bicolor as host of Z. rosaecolis. While dissecting hips of Rosa carolina obtained at Urbana, Mayview, Philo and San Jose, Illinois, NATURAL ENEMIES OF Rhynchites bicolor (¥FABR. ) Lut and Ironton, Missouri, I discovered five larvae attacking small grubs of Rhyn. bicolor in the achene cavities. The inclusive dates were June 16 and August 7. Six small pupae lying adjacent to dead, shrunken, more or less dark larvae of Rhyn. bicolor were disclosed between June 9 and August 27. Eleven adults developed from these parasitic larvae and pupae, in the over-all period from June 15 to September 11, through the years 1947, 1948, 1949, 1951 and 1952. In addition, I discovered 5 adults still enclosed within the achene cavities along with shrunken remains of the host larvae, between July 22 and August 9. These, and the above reared series of 11 adults, consti- tute positive evidence that Zatropis rosaecolis parasitizes the larvae of Rhyn. bicolor in the achene cavity of rose hips. It is ectoparasitic and solitary. Also 61 living adults were found alive in jars of hips grown in areas where Rhyn. bicolor is known to occur. The majority of these issued during July to September of the years the larvae developed on the host, but smaller numbers wintered in some undetermined stage in the hips, the adults emerging during May and June of the ensuing year. Since no host stages inhabit the hips during winter, these spring emergers cannot represent a generation distinct from the adults issued the pre- vious year. The Zatropis larva (Fig. 49) spins no cocoon. The mature black pupae lie bare in the achene cavity. My rearing records show that pu- pation required 7 to 10 days in natural summer temperatures in the laboratory. That the female Zatropis inserts the terebra into the egg pit pre- pared by the female Rhyn. bicolor is indicated by the observations that the hosts are small instars and that these normally remain near the inner end of the pit to feed. Their position here probably renders them susceptible to parasitization. Eurytoma sp., Eurytomidae, Hymenoptera Most species of this chalcid genus are black, with the thorax um- bilicately punctate and pronotum transverse, strap-shaped; abdomen of female compressed and highly polished. Being unusually similar, the species are distinguishable with difficulty, which fact probably explains why many remain unidentified in the literature. It is possible that two or more species are represented in my series of 22 adults from rose hips. The females vary from 2.1 to 3.5 mm. in length, the males 1.5 to 2.3 mm. Two individuals of the series originated from hips of Rosa carolina taken at Chetek, Wisconsin, and 15 from hips of Rosa spp. sent to me from five widely separated localities of Iowa in April, 1953. The Iowan specimens emerged from hips between May 7 12 and 25, 1953; one from Chetek on May 18, and the rest from Wiscon- sin and from Illinois appeared in hip jars from July to October. This diversity of emergence dates may signify two generations per year in Illinois and Iowa. I reared a male Eurytoma from a newly formed pupa found lying on the dead, dusky shrunken remains of an advanced stage larva of Rhyn- chites bicolor in a hip of Rosa carolina at Urbana, Mlinois, on August 31, 1947. The adult issued from the hip, in which the pupa lay, on Sep- tember 9, 1947. The pupal development required about 10 days. In the summer of 1937, I reared another adult from a larva found attacking a grub of the rose curculio in a hip of Rosa rugosa at Urbana. These instances form certain proof that my Eurytoma sp. at least sometimes parasitizes the larva of Rhyn. bicolor. Like B. mellitor, it performs as a solitary ectoparasite. bo INSECTS IN ROSE HIPS Chaetogaster sp., Oligochaeta, Chaetopoda, Annelida From September 9 to 24, 1950, I set up a series of 13 cages to house approximately 1,500 mature larvae of Rhynchites bicolor that emerged from hips of Rosa blanda and R. carolina taken at Chetek, Wisconsin, on September 5. The cages consisted of four-inch flower pots almost filled with black soil from my house yard at Urbana. The larvae were the wintering stage of this curculio and were kept out-of-doors in pots countersunk in the ground, from November to May. Evidence of parasitic life. During May to July, 1951, sample pots were scrutinized at about weekly intervals as a means of determining the developmental status of the beetle and Luchatema baldufit, its endo- parasite. In the course of examining such a sample on June 17, 1951, I discovered several small, slender, active whitish worms, later deter- mined as an oligochaete annelid of the genus Chaetogaster by Dr. L. J. Thomas, University of Illinois. These were associated with a dead, but still fresh, whitish new pupa of the rose curculio in the upper 1.5 inches of the soil. No less than 40 individuals of Chaetogaster swarmed from the pupa when it was immersed in well water. They varied from less than 1.0 mm. to 6.0 mm. in length, excepting one 12.0 mm. long. Also on June 15, 1951, a smaller number of such worms was present in an- other of the 13 pots. These seemed to have scattered into the surround- ing soils subsequent to the destruction of the host. No worms were found in the other 11 pots. Presuming this Chaetogaster to be a normal inhabitant of the ground, I concluded that it probably was present in the soil used to fill the pots, or crawled in during the fall or spring when the pots sat in the soil. It seems very unlikely that the parent worm of the aggregation of 40 origi- nated in some rose hip at Chetek. Because the worms found on June 17 SCAVENGEROUS MITE; ROSE SEED CHALCIDS 123 inhabited an apparently recently killed pupa, they may be regarded as true parasites. SCAVENGEROUS MITE Caloglyphus spinitarsus (Herm.), Acaridae, Acarina, Arachnida In late June and early July of 1947 to 1951, numbers of this small whitish plump-bodied mite were discovered in soil pots of the type described above under Chaetogaster. The pots of soil had been the wintering site of the larvae of Rhynchites bicolor, some of which con- tained the larvae of the endoparasite, Luchatema balduft. In the months indicated above, the eggs, nymphs, and adults of Caloglyphus spinitarsus were found in numbers associated with dead larval bodies of the cur- culio, and also dead bodies and meconia voided in the cocoons of the parasite. The frequent occurrence of many mites in association with these dead organic substances leaves no doubt that the species is a scavenger, not a parasite. I have not found references to it in the Review of Applied Entomology, series A. Several congeneric species have been reported there as associated with decomposing plant matter, mushroom culture and corn processing plants, all under circumstances which indi- cate scavengerism. THE RosrE SEED Cuatcips, Megastigmus spp., TORYMIDAE, HYMENOPTERA Most known species of Chalcidoidea perform as entomophagous para- sites. Prominent among the phytophagous minority are many members of the genus Megastigmus, which have been shown to develop in the seeds of various plants. Two species of the genus are known to develop in rose hips, namely M. aculeatus (Swederus) of Europe (Fig. 53), and M. nigrovariegatus Ashmead of North America (Fig. 52). Their life histories are similar. Each completes one life cycle in a year—the em- bryos, larvae and pupae residing in the achenes. The mature larvae are the wintering stage and have a prolonged diapause. Pupation and emer- gence of the adult occur in the following spring. The rose seed chalcids are at most 3.0 mm. long, with variable amounts of black and yellow on the body, and particularly on the dorsum. As the generic name signifies, they have a comparatively large stigma on the anterior margin of the fore wings. The females of both species bear long terebras enclosed in bristly sheaths. Specific differences. These species, although closely related, are for 124 INSECTS IN ROSE HIPS the most part readily distinguished by the following characteristics I have taken from Milliron’s (1946) descriptions. Megastigmus nigrovariegatus Megastigmus aculeatus (Fig. 53) (Fig. 52) A cloudy infuscation surrounding No cloud surrounding stigma stigma No black coloration on outer two- Outer two-thirds of axilla black thirds of axilla, except on very dark females Ovipositor sheath almost constant in Ovipositor sheath distinctly longer length; distinctly shorter than en- than the total body, except in va- tire body riety nigroflavus Hoffm. Megastigmus aculeatus (Swederus ) Synonymy. Swederus named and characterized this European rose seed chalcid in 1795, under the generic name Pteromalus. In the subse- quent 80 years, it was renamed as follows: Megastigmus transversus Walker (1833); Torymus (Megastigmus) collaris Boheman (1833); Torymus punctum Foerster (1841); Megastigmus vexillum Ratzeburg (1848); Megastigmus flavus Foerster (1859); Megastigmus cynorrhodi Perris (1876). Milliron (1949, p. 289) is my source of this information. Although obviously not as variable in size and amounts of yellow and black on the body as M. nigrovariegatus, the European species, M. aculeatus nevertheless displays some variation, particularly as to color. This fact is indicated by Hoffmeyer’s (1929) recognition of a variety nigroflavus. I share the experience of Milliron in having discovered, among reared specimens, individuals which diverge from the color pat- tern of both nigroflavus and the typical form aculeatus. Food habit. A controversy that engaged the attention of several en- tomologists in Europe and North America during the nineteenth cen- tury was whether Megastigmus is “phytophagic” or parasitic. Milliron (1949, p. 355 and forward) gives an excellent account of this historical episode as it pertains to the genus as a whole. Following is an abbrevi- ated statement concerning M. aculeatus, in particular. Two entomological facts about the food habits of Chalcidoidea de- layed the acceptance of the phytophagy of aculeatus. First, some species of chalcids had early been proved to be parasitic on other insects, and it was therefore generally supposed that all but a few members of the superfamily are parasitic. Second, the larvae of the few exceptional species were found or believed possibly to be of “mixed” or “gemischter” food habit. The early instars of the “mixed” species consumed the small insect host, whereas the advanced instars developed on plant tissue left by the phytophagous host. W. J. Phillips (1927) and Puzanova-Maly- sheva (1936) describe instances of such mixed habit. THE ROSE SEED CHALCIDS, Megastigmus spp. — - 125 For about 75 years after Swederus named it in 1795, M. aculeatus was presumed to be a parasite on some phytophag in rose hips. Because of its larval habitation in the hypanthium and its common occurrence, the European rose hip fly, Rhagoletis alternata (Fallén, 1820) was be- lieved to be the host. This view was held, or copied, by Ratzeburg, 1848; Rondani, 1872; Thomson, 1875; Cameron, 1879; Dalla Torre, 1898, and others. Ratzeburg (1848, p. 182) stated that M. aculeatus (vexillum) was reared from this fly by P. F. Bouché, but gives no literary reference for the statement. Since Bouché (1833, 1834) does not mention such a host relation, it is likely that the report was received through personal com- munication, for these men were contemporaries and acquainted. Reinhard (1857) seems to have made the first observations that cast doubt on the parasitic habit of aculeatus. Knowing that the larva of the hip fly leaves the hip in the fall season, he was perplexed when he reared large numbers of adult aculeatus from hips collected in winter. Mayr (1874) repeated Reinhard’s statement on the dubious parasitic relation. But especially the studies by Perris (1876) in France, and Wachtl (1884) in Austria, produced evidences that helped to undermine the belief in the parasitic habit. Perris reared many aculeatus from hips of Rosa sempervirens in the spring season, and, noting holes in the seeds, questioned whether the Megastigmus might have inhabited the seeds. Moreover, the hips contained no other insects which might serve as its host. Concluding that aculeatus (“cynorrhodi’) is more than probably “spermophagous, Perris intended to continue his investigations, but published nothing further on the question. Wachtl (1884) did more than anyone of his period to reveal the true food habit of aculeatus (“collaris”). He reared adults from rose hips and traced their emergence holes from the periphery of the hips to the achenes. No seed at all remained in these, hence the kernel can have been destroyed only by the larvae of “collaris.” Later Wachtl found larvae filling the inner spaces of the whole achenes, and no other insects present that may have served as host of the “collaris.” But Howard (1892) and Riley (1893) remained unconvinced and per- sisted in their opinion that M. aculeatus might be a species of “mixed” habits. They cited a number of cases they believed to be mixed. For example, Howard then considered “Eurytoma funebris,” now known as Bruchophagus gibbus—the clover seed chalcid—to be parasitic on Dasy- neura leguminicola (Lintner )—the clover seed midge. Wachtl (1893) responded to their objection with the declaration that we know only a small fraction of the life of chalcids, and insisted that M. aculeatus is phytophagous. He has been supported by Crosby (1909), Milliron 126 INSECTS IN ROSE HIPS (1949) and others. Moreover, it is the belief of observers today that aculeatus confines its life to the achenes of Rosa spp. Few claims to the contrary have been made. Milliron (1949, p. 402) mentions Evonymus sp. and Rhamnus sp. reported in the literature as food plants of M. aculeatus. He regards these as dubious records that stand in need of confirmation. Escherisch (1923) quotes Fuchs (1912) to the effect that M. aculeatus parasitized the snout beetle, Otiorrhynchus sensitivus Scop., whose larvae feed at the roots of evergreen trees. This is obviously an error. Distribution and rose hosts. Below I give such records of rose hosts and geographical occurrence as have come to my attention through the literature, together with my own acquisitions. Placed first are those per- taining to the typical form aculeatus aculeatus (Swed.); second, those concerned with the dark form that resembles, more or less, the variety nigroflavus of Hoffmeyer; and third, records having to do with series not identified as to variety. I. The typical or light form, M. a. aculeatus. In the course of his re- visional study of the genus Megastigmus, Milliron (1949, 291-92) studied typical material “reared from, or associated with the following species of Rosa,’ and obtained from places named here: Rosa medwedewii, Caucacus; Rosa alpina, France and Switzerland; R. rugosa, Germany and U.S.S.R., and the following from U.S.S.R.—R. davurica, R. jundzilli, R. rugosa kamtschatica, and R. mollis. Milliron redescribed M. a. acu- leatus, from specimens reared from Rosa at Jamaica Plain, Massachusetts, and Ithaca, New York. He states it is known to occur also in France, Germany, Switzerland, at Moscow, Gagri and Tiflis, Russia, and in Iran, Peking, China, and French Somaliland, Africa. Nikol’skaya (1934) re- ported it reared from seeds of Rosa cinnamomea in the province of Ku- ban, North Caucasus, and Poltava, in the Ukraine, of the U.S.S.R. My series of light-colored specimens were reared at Urbana, Illinois, from hips acquired as follows: Rosa eglanteria, Ashe County, North Carolina; R. canina, Homeworth, Columbiana County, Ohio; Rosa, prob- ably palustris, East Liberty, Ohio; R. eglanteria, West Lafayette, Coshoc- ton County, Ohio, and Rosa, near canina, Geneva, New York. II. The dark form, M. a. nigroflavus. Milliron (1949) recorded it from New Jersey; Virginia; Yokohama, Japan; and Saratow, Russia. The types from which Hoffmeyer described the variety were reared at Copen- hagen, Denmark, from seeds of Rosa multiflora received from Japan. The “New Jersey” record pertains to the series reared by Weiss (1917) from seeds of Rosa multiflora shipped from Japan. My dark forms, which approximate the nigroflavus Hoffm., originated in Rosa eglanteria, Ur- bana and Kickapoo State Park, Illinois; Rosa sp., Mahomet, Illinois; and R. virginiana, Fox Ridge State Park, [linois. THE ROSE SEED CHALCIDS, Megastigmus spP. 127 III. Form of Megastigmus aculeatus not identified. (1) Germany. Reported as “Torymus punctum,” Foerster (1841), and “Megastigmus flavus,” Foerster (1859): as “Megastigmus vexillum,” Ratzeburg (1848); Heilbronn, from Rosa, sp., Crosby (1913) and Baut- zen, Sachsen, from Rosa sp., Reinhard (1857): as “Megastigmus sp.,” which, I suspect, is M. aculeatus, Doyer (1926). (2) France. Mont-de-Marsan, Gascony, as “Megastigmus cynorrhodi,” from seeds of Rosa sempervirens, Perris (1876); Oise, Elbeouf, M. acu- leatus, “Parasite de Rhodites spinosissima, Spilographa alternata,”’ Cou- lon (1931). (3) Denmark. Copenhagen, “Aus Bromme Plantage,” Hoftmeyer (1930). (4) Austria. Vienna, from Rosa spp., Wachtl (1884, 1893). (5) Sweden. “Habitat in Uplandia,” July, August, September, as “Pteromalus aculeatus,” Swederus (1795); as “Torymus collaris,” “Habi- tat in Smolandia, Westrogothia et Scania sat frequens,” Boheman (1833); “Hakan Ohlssons, Lundae,” from Rosa sp., Thomson (1875). (6) Finland. “N: Tvarminne (Vorentaus), Helsinge,” lives in seeds of Rosa, Hellén (1934); in seeds of Rosa, Kangas (1945). (7) England. As “Megastigmus transversus,’ “on grass in woods, near London,” Walker (1833); “parasite of Trypeta continua Meig., a dweller in the berries of the rose,” Cameron (1879); Morley (1910); “Europe, Britain,” Laidlow (1931). (8) Italy? As “M. transversus Wlk. (=Torymus punctum Forst),” “nelle larvae della Zonozema alternata Fall. cire continua di Megn.,” Rondani (1872). (9) Russia. Reared from seeds of Rosa, Rodzianko (1908); “Larva destroys the seeds of roses (Rosa). Known from middle Europe, Japan, China, North America (introduced) and Kurak district of Russia,” Rimsky-Korsakov (1932). (10) Japan. From Rosa sp., Yano and Koyama (1918); in rose seeds, Japan (Anonymous, 1926); Japan and Europe, “parasitic in seeds of rose,” Ishii (1932). (11) China. From rose seeds imported from Peking. Specimens in collection of U.S. National Museum, Crosby (1913). (12) New York. Ithaca, from Rosa, Crosby (1913); Leonard (1926). (13) Rhode Island. Kingston, from seeds of Rosa multiflora. Speci- mens in collection of U.S. National Museum. Dated June, 1941. (14) Maryland. Rockville. In house, March, 1944. Specimens in col- lection of U.S. National Museum. (15) Washington. Seattle. Specimens in collection of U.S. National Museum. Dated September 31, 1944. Spread of M. aculeatus. The fact that M. aculeatus was originally de- 128 INSECTS IN ROSE HIPS scribed from Sweden 160 years ago, in 1795, and its wide occurrence from western Asia to England, constitute strong indications that the species is indigenous to the Old World. Its nativity in Eurasia is sug- gested also by its comparatively recent discovery in Japan, China and North America. The egress of the species from the place of its origin was favored by two conditions. First, it appears to prefer as hosts such roses as eglanteria and canina of Europe, and virginiana of North America, which also are favorites of gardeners, hence are widely dis- tributed through commercial channels. These are the species from which all but a few of my specimens were obtained. Second, the mature larva inhabits the achenes in the hips, undergoing a diapause beginning in midsummer and terminating about April to May of the following year. Since the ripe hips commonly persist on the rose bush until the spring after their development, and the rose hosts, desired for their economic importance, are generally distributed by dealers during the period when the chalcid larva is in diapause, the spread of aculeatus is greatly aided by human agencies, both from continent to continent, and from one country to another. On the other hand, the chalcid appears not, or little, to disseminate locally to native roses in Illinois through its own powers of flight. The major specific means of international spread is through shipment of the “seeds” of Rosa multiflora from the orient to Europe and North America. The multiflora rose appears very well adapted to the success of M. aculeatus, as is indicated by the large numbers reared from its achenes in Europe (Hoffmeyer, 1929; Doyer, 1926), and in New Jersey (Weiss, 1917). Today, and for some years past, the multiflora is and has been advocated and employed in the United States as an ornamental in parks, as a fence to confine domestic animals, and as a shelter for game animals. Sex ratio. Whereas M. nigrovariegatus involves approximately equal numbers of males and females, M. aculeatus produces a strongly dis- proportionate percentage of females over males, hence is decidedly parthenogenetic. This fact holds for all reports of large numbers reared, as follows: “more than 800 females and 25 males” reared by Hoffmeyer (1929) in Denmark from R. multiflora grown in Japan; 1,147 adults reared at Urbana in 1946 to 1953 from canina, eglanteria, and virginiana that originated in New York, North Carolina, Ohio and Illinois totaled 34 males and 1,113 females, a ratio of 1 to 33; two males among “a very large number of females” in France (Perris, 1876); and Reinhard (1857) reared “large numbers” of females, but no males, in Germany. The num- bers given by Hoffmeyer (1929) indicate a proportion of the sexes very similar to those quoted above. THE ROSE SEED CHALCIDS, Megastigmus spp. 129 Life cycle. The dates of emergence of adult M. aculeatus from hips indicate that it completes one cycle in a year in all areas where it has been observed. With few exceptions, the adults are reported to emerge in May and June. Specimens dated decidedly earlier or later than these months were probably either forced in artificial warmth, reared in subnormal temperatures, or discovered as dead some weeks after they issued. Emergence dates that fall within May and June are as follows: May 2, Kingston, Rhode Island (U.S. National Museum); May 13 to June 24, my lot of 1,147 individuals; June, France (Perris) and Western Russia (Nikol’skaya); May and June, Austria (Wachtl); latter part of May, New Jersey (Weiss); June 26, Finland (Hellén), and May 9 and 25 (Milliron). Only Reinhard (1857) seems to have seen the females “boring,” or ovipositing into the hips of wild rose. The eggs are probably inserted into the immature achenes, where, it may be supposed, the larvae also develop. Oviposition begins a few days after the adults leave the old wintered hips, or approximately from mid-May to July. Weiss (1917) observed that the “larva appears to destroy the entire interior of the seed, leaving nothing but the hard outer covering.” On March 5, 1953, I opened a sample of achenes that had hung during the previous winter on a bush of Rosa virginiana in Fox Ridge Park, Illinois, and found full-grown larvae snugly filling the achenes. The emergence of adults from hips of the same source in May and June is evidence that the wintered mature larva pupates in the achenes, probably from late April to June. Reinhard (1857) stated it pupates in the hip, probably a deduction from his observation that the adult, and not another stage, issues from the hip. Wachtl (1884) observed circular holes for escape both in the achenes and in the hypanthium. Accordingly, the larva must develop in the “seeds,” not in the fleshy outer cover, nor as a parasite within the larva of Rhagoletis alternata, as had theretofore been sup- posed to be its habit. The larva grows up as a phytophag, consuming all the seed while this remains more or less soft, and becomes full-grown about midsummer when the seed hardens, undergoes a long diapause and winters as a mature larva in the now seedless achene, pupates in late April to June of the next year, and emerges as adult in May or June. Economic importance. Commercial damage has been reported where M. aculeatus develops in the seeds of the “multiflora rose,” Rosa multi- flora, which is still being advocated and used extensively as a fence to confine farm livestock, as cover for game animals and as a rural orna- mental. Weiss (1917) stated that rose growers in New Jersey complained, at different times during the preceding years, of the failure of the seeds 130 INSECTS IN ROSE HIPS of Japanese Rosa multiflora to germinate. Consumption of the seeds by the larvae of M. aculeatus was found to be the cause. Weiss reared aculeatus from nearly every shipment of seeds consigned from Japan to New Jersey in 1917. An anonymous article from Japan (1926) named M. aculeatus in rose seeds as one of the principal pests found in plants intended for export. As an indication of the destructive potential of aculeatus, it is also significant that Hoffmeyer (1929) reared more than 800 adults from the seeds of R. multiflora shipped from Japan to Den- mark. Doyer’s (1926) report that a Megastigmus sp., which may have been aculeatus, was reared from seeds of R. multiflora at the seed-testing station at Wageningen, Germany, indicates this is a species capable of large populations in favorable rose hosts. Megastigmus nigrovariegatus Ashmead, The American Rose Seed Chalcid Type localities and synonyms. Ashmead’s description (1890) of M. nigrovariegatus was based on a number of females that were collected at Greeley, Colorado, and Vancouver Island, British Columbia, by H. F. Wickham. The male was described by Crosby (1913) from a series pre- served in alcohol, without giving data. Milliron (1949) redescribed this sex from specimens reared from seeds of Rosa rugosa. There are no synonyms, except two which resulted when Crosby (1909) confused nigrovariegatus Ashm. with aculeatus Swed., and Laidlaw (1931) mis- spelled the specific name as “nigrovarietus.” These nomenclatural errors are pointed out by Milliron (1949, p. 295). Hosts and distribution. Both the rose hosts and geographical sources for M. nigrovariegatus are known, as indicated in the following list. Rosa acicularis vars., Fairbanks, Alaska (U.S. National Museum); seven locations in northwestern Canada to near Fairbanks, Alaska (Balduf, original); Eaglesnest, Ely, Tomahawk road, Temperance river at Lake Superior, in northeastern Minnesota (Balduf, original). Rosa blanda vars., Eaglesnest, Ely and Grand Marais in northeastern Minnesota; Solon Springs, Gordon and Sarona, Wisconsin; Ottawa, Canada; Atkinson, Illinois (Balduf, original). Rosa blanda and R. carolina vars., Chetek and Rice Lake, Wisconsin (Balduf, original). Rosa carolina vars., Fall Creek, Endeavor, and Black River Falls, Wis- consin; Ironton, Missouri; the following localities in Illinois—Apple River State Park, Freeport, Kirkland, Rochelle, Wichert, Watseka, San Jose, Mahomet, Mayview, Philo, Urbana, Kickapoo State Park, Ridgefarm, Alton, Velma, Taylorville, Sumner, West Frankfort (Balduf, original); Champaign county (Balduf, 1945). THE ROSE SEED CHALCIDS, Megastigmus spp. 131 Rosa rugosa, Madison, Wisconsin (Crosby, 1909, 1913), Dane-Colum- bia county line on U.S. 45, and Ashland, Wisconsin (Balduf, original); Minneapolis, Minnesota (Milliron, 1949); Urbana and Sidney, Illinois (Balduf, 1945), Onarga and Taylorville, Illinois (Balduf, original); Kingston, Rhode Island (U.S. National Museum); Orono, Maine (Balduf, original). Rosa palustris, Urbana, Illinois, and Homeworth, Ohio (Balduf, original); Urbana, Illinois (Balduf, 1945); Vienna, Virginia (U.S. National Museum); Virginia (Milliron, 1949). Rosa eglanteria, Champaign county, Illinois (Balduf, 1945); Kickapoo State Park, Illinois, West Lafayette, Ohio ( Balduf, original). Rosa spinosissima and R. xanthina, Urbana, Illinois (Balduf, 1945). Rosa_ virginiana, Urbana, Illinois (Balduf, 1945); Orono, Maine (Balduf, original); Vienna, Virginia (U.S. National Museum). Rosa Macounii, Eaglesnest, Minnesota; Saskatoon, Saskatchewan; Crete, Nebraska (Balduf, original). Rosa setigera, Urbana, Illinois (Balduf, original). Rosa arkansana, Poynette and Madison, Wisconsin (Balduf, original ). Rosa woodsii, Beulah, Manitoba, and Provo, Utah (Balduf, original). Rosa californica, Albany, California (Balduf, original). Rosa pyrifera and R. lunelli, Saskatoon, Saskatchewan (Balduf, original ). Rosa spaldingii, Pullman and Spokane, Washington (Balduf, original). Rosa ultramontana, Moscow Mts., Idaho (Balduf, original). Rosa suffulta, Saskatoon, Saskatchewan; La Harte, Illinois (Balduf, original ). Adults of M. nigrovariegatus have also been reared from hips of un- determined wild roses sent to me at Urbana by collaborators from West Alexandria, Ohio, 1947; East Falls Church, Virginia, 1946; and from samples of hips sent in 1953 from the following places in Iowa; state park at Keosauqua; “Strawberry Point”; Black Hawk State Park; Lake- view; Waubonsi State Park, Hamburg; Lake Aquabic State Park, Indian- ola; Oak Grove State Park, Hawarden; Chariton, and Beeds Lake State Park, Hampton. I also have this species from hips taken at Kenosha Pass, Colorado. Crosby (1913) had M. nigrovariegatus “from seeds of roses” obtained in New York, Illinois, New Hampshire, Massachusetts, Delaware and Utah. According to data on specimens in the collection of the U.S. Na- tional Museum, the chalcid occurs in Canada, Illinois, Washington, D.C., Maine, Idaho, Colorado, British Columbia, New York, Pennsylvania, Massachusetts, Virginia, Minneapolis and Alaska. Cockerell (1894, 1899) found it on a rose at Santa Fe, New Mexico, and Britton (1938) re- ported it for Connecticut. Milliron (1949) adds the following provinces 130 INSECTS IN ROSE HIPS of Japanese Rosa multiflora to germinate. Consumption of the seeds by the larvae of M. aculeatus was found to be the cause. Weiss reared aculeatus from nearly every shipment of seeds consigned from Japan to New Jersey in 1917. An anonymous article from Japan (1926) named M. aculeatus in rose seeds as one of the principal pests found in plants intended for export. As an indication of the destructive potential of aculeatus, it is also significant that Hoffmeyer (1929) reared more than 800 adults from the seeds of R. multiflora shipped from Japan to Den- mark. Doyer’s (1926) report that a Megastigmus sp., which may have been aculeatus, was reared from seeds of R. multiflora at the seed-testing station at Wageningen, Germany, indicates this is a species capable of large populations in favorable rose hosts. Megastigmus nigrovariegatus Ashmead, The American Rose Seed Chalcid Type localities and synonyms. Ashmead’s description (1890) of M. nigrovariegatus was based on a number of females that were collected at Greeley, Colorado, and Vancouver Island, British Columbia, by H. F. Wickham. The male was described by Crosby (1913) from a series pre- served in alcohol, without giving data. Milliron (1949) redescribed this sex from specimens reared from seeds of Rosa rugosa. There are no synonyms, except two which resulted when Crosby (1909) confused nigrovariegatus Ashm. with aculeatus Swed., and Laidlaw (1931) mis- spelled the specific name as “nigrovarietus.” These nomenclatural errors are pointed out by Milliron (1949, p. 295). Hosts and distribution. Both the rose hosts and geographical sources for M. nigrovariegatus are known, as indicated in the following list. Rosa acicularis vars., Fairbanks, Alaska (U.S. National Museum); seven locations in northwestern Canada to near Fairbanks, Alaska (Balduf, original); Eaglesnest, Ely, Tomahawk road, Temperance river at Lake Superior, in northeastern Minnesota (Balduf, original). Rosa blanda vars., Eaglesnest, Ely and Grand Marais in northeastern Minnesota; Solon Springs, Gordon and Sarona, Wisconsin; Ottawa, Canada; Atkinson, Illinois (Balduf, original). Rosa blanda and R. carolina vars., Chetek and Rice Lake, Wisconsin (Balduf, original). Rosa carolina vars., Fall Creek, Endeavor, and Black River Falls, Wis- consin; Ironton, Missouri; the following localities in Illinois—Apple River State Park, Freeport, Kirkland, Rochelle, Wichert, Watseka, San Jose, Mahomet, Mayview, Philo, Urbana, Kickapoo State Park, Ridgefarm, Alton, Velma, Taylorville, Sumner, West Frankfort (Balduf, original); Champaign county (Balduf, 1945). THE ROSE SEED CHALCIDS, Megastigmus spp. 131 Rosa rugosa, Madison, Wisconsin (Crosby, 1909, 1913), Dane-Colum- bia county line on U.S. 45, and Ashland, Wisconsin (Balduf, original); Minneapolis, Minnesota (Milliron, 1949); Urbana and Sidney, [linois (Balduf, 1945), Onarga and Taylorville, Illinois (Balduf, original); Kingston, Rhode Island (U.S. National Museum); Orono, Maine (Balduf, original ). Rosa_ palustris, Urbana, Illinois, and Homeworth, Ohio (Balduf, original); Urbana, Illinois (Balduf, 1945); Vienna, Virginia (U.S. National Museum); Virginia (Milliron, 1949). Rosa eglanteria, Champaign county, Illinois (Balduf, 1945); Kickapoo State Park, Illinois, West Lafayette, Ohio ( Balduf, original). Rosa spinosissima and R. xanthina, Urbana, Illinois (Balduf, 1945). Rosa virginiana, Urbana, Illinois (Balduf, 1945); Orono, Maine (Balduf, original); Vienna, Virginia (U.S. National Museum). Rosa Macounii, Eaglesnest, Minnesota; Saskatoon, Saskatchewan; Crete, Nebraska (Balduf, original). Rosa setigera, Urbana, Llinois ( Balduf, original). Rosa arkansana, Poynette and Madison, Wisconsin (Balduf, original). Rosa woodsii, Beulah, Manitoba, and Provo, Utah (Balduf, original). Rosa californica, Albany, California (Balduf, original). Rosa pyrifera and R. lunelli, Saskatoon, Saskatchewan (Balduf, original ). Rosa spaldingii, Pullman and Spokane, Washington (Balduf, original ). Rosa ultramontana, Moscow Mts., Idaho (Balduf, original). Rosa suffulta, Saskatoon, Saskatchewan; La Harte, Illinois (Balduf, original ). Adults of M. nigrovariegatus have also been reared from hips of un- determined wild roses sent to me at Urbana by collaborators from West Alexandria, Ohio, 1947; East Falls Church, Virginia, 1946; and from samples of hips sent in 1953 from the following places in Iowa; state park at Keosauqua; “Strawberry Point”; Black Hawk State Park; Lake- view; Waubonsi State Park, Hamburg; Lake Aquabic State Park, Indian- ola; Oak Grove State Park, Hawarden; Chariton, and Beeds Lake State Park, Hampton. I also have this species from hips taken at Kenosha Pass, Colorado. Crosby (1913) had M. nigrovariegatus “from seeds of roses” obtained in New York, Illinois, New Hampshire, Massachusetts, Delaware and Utah. According to data on specimens in the collection of the U.S. Na- tional Museum, the chalcid occurs in Canada, Illinois, Washington, D.C., Maine, Idaho, Colorado, British Columbia, New York, Pennsylvania, Massachusetts, Virginia, Minneapolis and Alaska. Cockerell (1894, 1899) found it on a rose at Santa Fe, New Mexico, and Britton (1938) re- ported it for Connecticut. Milliron (1949) adds the following provinces 132 INSECTS IN ROSE HIPS or states: Nova Scotia, Quebec, West Virginia, Indiana, Michigan, Kansas, North Dakota, and Oregon. Our first indication of the host plant of M. nigrovariegatus came from the observation of Cockerell (1894), who discovered females “on the fruits of roses in Mr. Boyle’s garden” in Santa Fe, New Mexico. Cocker- ell considered it “extremely probable that the species breeds in rose fruits.” Crosby (1909) observed females ovipositing on the rose hips during June and July, at Ithaca, New York, and his discovery of larvae and pupae in the seeds of rose established that rose seeds are their specific niche and food, a fact which is now generally accepted. In view of this knowledge, earlier claims that M. nigrovariegatus is an ento- mophagous parasite are untenable. Milliron (1949, p. 297) relates two such instances of error, both orig- inating from data on old specimens in the collection of the U.S. National Museum. In one of these cases, originally published by Crosby (1913), the alleged host was a moth, “Clisiocampa plumalis,” under observation at the Washington Agricultural Experiment Station in 1898. This record was subsequently copied by Maria Ruhl (1914). _ In the second instance (Riley, 1893), the host was said to be a gall- forming cynipid, “Callirhytis scitulus Bassett.” Two of the three speci- mens, to which Riley referred, were rediscovered and examined by Mill- iron (1949), who found they were “small atypical nigrovariegatus’ and commented that “their identity throws all doubt on the correctness of Riley’s note.” It is of interest also that Riley (1893) had cited this case in support of his contention that Megastigmus, as a group, is essen- tially parasitic, chiefly on gall-making Cynipidae. Summary of distribution. Of the northerly areas of North America, all excepting Vermont, South Dakota, Wyoming, and Alberta are included in the above list. Probably the chalcid also occurs in these excepted states and provinces. M. nigrovariegatus is, accordingly, widely distributed in the northern reaches of this continent from coast to coast. The northern- most recorded occurrence is in the vicinity of Fairbanks, Alaska, at 65° north latitude, and its southern limits lie near the 36th line of latitude, as indicated by the records from New Mexico, Arizona, and Virginia. Variation (Fig. 52). As stated above, this species is usually readily distinguishable from its much more color-stable relative, M. aculeatus. While studying more than 8,000 reared nigrovariegatus, I observed that the males vary from entirely yellow on small, to wholly black on large, specimens, with many intergrades in size and color pattern. The varia- tion in color is decidedly less extensive in the females (Balduf, 1957a). More specifically, the females of the present species have ovipositor and sheaths shorter than the body and both sexes have the stigmal knob THE ROSE SEED CHALCIDS, Megastigmus spp. 133 surrounded by a narrow, yet distinct fuscous ring. M. aculeatus lacks this ring around the knob, and the female bears sheaths and terebra longer than the body. Concurrence of M. nigrovariegatus and M. aculeatus. Only three of the many lots of hips that I had under observation from 1943 to 1953 yielded adults of both these rose seed chalcids. The essential data per- taining to these three cases are presented in Table 3. TABLE 3. CONCURRENCE OF THE ROSE SEED CHALCIDS Sources of hips Rosa spp. M. nigrovariegatus M. aculeatus 2 3 2 } West Lafayette, O. eglanteria 0 2 2 0 Kickapoo Park, III eglanteria 0 2 21 0 Fox Ridge State Park, Ill. virginiana 0 5 65 2, Emergence. Still in the achene in the hip, the newly formed adult discards the pupal theca, then employs the mandibles to cut a circular hole through the pericarp of the achene and the more or less dry hypan- thium directly above it. The exits are gnawed through the broadly con- vex surface of the achene that lies in contact with the concave inner face of the hypanthium. And because the inhabited achenes normally lie in the petiolar area of the hip cavity, the exits are confined to this portion of the hypanthium. However, the adults do not always succeed in emerging. Dissections of the relatively small achenes on a shaded stand of Rosa carolina at Urbana revealed some adults dead in the achenes and still encased in the pupal thecae, while others remained stuck in the partly prepared exits in the achenes. Such fatalities are believed to result from the inadequacy of the small seeds of this rose bush for the development of vigorous larvae. Even the adults that did emerge averaged small for the species (Balduf, 1945). While rearing adults from hips in the laboratory, Milliron (1949) ob- served that they usually remained relatively inactive for a day or more after they emerged from the pupal thecae, requiring some time for drying or hardening the wings, or to remove any adhering parts of the thecae before they became very active and engaged in flight. Sex ratio and protandry. The number of males in relation to the num- ber of females is approximately equal, according to samples reared from hips. Milliron (1949) reports having reared 5,785 adults from a single lot of the hips of Rosa rugosa at St. Paul, Minnesota. Of this number, 134 INSECTS IN ROSE HIPS 3,099, or 53.4 per cent were males, and 2,697 were females, or 46.6 per cent. These figures give a ratio of 1.0 female to 1.15 males. In the years 1943 to 1953, I reared approximately 9,935 adults. These emerged from hips of many species of Rosa grown at places so widely separated as Alaska, California, Minnesota, Maine (Orono), and central Illinois. The sample consisted of almost equal numbers of males and females; 5,146 males or 52 per cent, and 4,779 females, or 48 per cent. My observation (1945) that M. nigrovariegatus is definitely protan- drous has been confirmed by Milliron (1949). He reported that “in general, males make their appearance about three days in advance of females, and reach their maximum emergence a few days after the latter have begun to emerge. In more restricted numbers they continue to ap- pear along with females throughout the period of emergence.” Some of my data on this feature appeared in my article of 1945, and others were accumulated in subsequent years. On June 3 and the early afternoon of June 4, 1943, only males were present on bushes of Rosa rugosa at Urbana. The first females were discovered at 6:00 p.m. of June 4. Males still predominated on June 5, but on June 9, the ratio of females to males was 15 to 1, and on June 13, it had increased 35 to 1. From June 15 to 19, no males remained, whereas 82, 60, 62, 50 and 38 females were present, respectively, on the same five days. Two other sets of data indicate the nature of protandry in this rose seed chalcid. The first set concerns adults from hips of Rosa rugosa stored in large glass tubes through a period of eight successive days. Emergence per day, by sexes, was as follows: first day, 26 o's, 39s; second day, 23 o's, 7 2s; third, 20 o's, 10 2s; fourth, 13 o's, 26 ¢s; fifth, 6 os, 28 2s; sixth, 2 #s, 10 2s; seventh, 3 o's, 0 2, and eighth dave Ih is 0 ©. The second set of data pertains to 227 females and 135 males reared from hips of rugosa between May 22 and June 27. The first males emerged on May 22, the first females on May 25. The numerical peak of emergence of the males was reached in May 26 to 30, that of the females in May 29 to June 4. Both sexes continued to appear thereafter, but the daily issue of females gradually overtook that of the males as the period of emergence progressed, thus repeating the same pattern of relative emergence of sexes demonstrated in the preceding instances and as ex- pressed in the statement quoted above from Milliron. Similar patterns of protandry were discovered also in other hip insects that develop equal numbers of males and females and were reared in adequate quantity. Mating. In May and June of the years 1943 and 1944, I observed a strong tendency of adult males to gather on and about certain old hips of Rosa rugosa at Urbana, while, at the same time, they avoided many THE ROSE SEED CHALCIDS, Megastigmus spp. 135 other hips of similar appearance and age. Then on May 30, 1944, as I stood watching the movements of such an aggregation of males, I dis- covered the significance of this behavior. Presently the head of a female slowly emerged into view in the corky hypanthium of the hip. No sooner had she completed her emergence than the group of males rushed upon her, and one succeeded in mating. Somehow, they are obviously capable of sensing the presence of adults of their kind still invisible in the hip. It is desirable to learn whether they respond in this manner to emerging males as well as females, i.e., whether they can also distinguish between the sexes while these are still out of sight in the hip. While thus attracted to the old hip, the males (as many as fifteen in one instance) either rested quietly on it, or on branches and leaves nearby, or sometimes ran excitedly over it, or made short quick flights within a foot from the hip, but always returned again to it. Often some of them vigorously pursued others as if to drive them away. In doing so, they sometimes raised and lowered their wings as resting butterflies may do. Dissections of several hips, around which males had previously con- centrated, showed that they no longer contained adults when the interest of the males in the hips had ceased. I believe this is the normal place and time of mating in the field, but have seen a few instances of mating elsewhere than on old hips. Milliron (1949, p. 384) stated that mating occurs very commonly in captivity and “usually within a few hours after the females emerge from the seeds. However, in a limited confinement with a relatively large population of males aged several days, copulation may occur within the short period of 15 or 20 minutes after emergence of the female.” In Milliron’s description of mating activity, I recognize four some- what distinct steps. First, in approaching the female, the male momen- tarily examines her with his antennae, becomes much excited and im- mediately takes a mounted position. Second, during the courting phase, the antennae of the male are vibrated rapidly in close proximity to, or touching those of the female; characteristically he also momentarily elevates the vibrating wings to a vertical position, during which he usu- ally surges somewhat downward and forward, often thereby touching her antennae or head with his mandibles. The wings may be elevated in that way several times before copulation is attempted. Also, if the female is especially inclined to move actively about, the male may drop his wings forward at her sides and hold them there momentarily each time. Third, when ready to copulate, the female slowly elevates the abdomen and lowers the sterna, thereby exposing the genital opening. Instantly the male is in place, and may then copulate from one of two positions, either almost caudolateral—or a directly caudal one. Small males usually did so from the caudolateral position. In either case, the 136 INSECTS IN ROSE HIPS male quickly moves backward, bends the abdomen into a ventral posi- tion, and clings to the wings, the side or to the ovipositor of the mate. Copulation usually requires between 15 to 20 seconds, during which time the female remains in a stationary position with the abdomen elevated. Fourth, after copulating, the male immediately remounts directly from the position taken when in copulo, and proceeds to repeat his courting behavior, but in no instance did Milliron observe a second consecutive copulation. Eventually, the male is dislodged by the female, or he leaves of his own accord. Milliron reported also that large males are usually not successful in mating with small females, but that small males commonly mated with large females. His observation, that small light-colored females from achenes of Rosa palustris mated with small males from R. rugosa, is fundamental in demonstrating that forms of nigrovariegatus from differ- ent species of roses can be cross-mated and therefore are not host specific. As a rule, females mate more than once, according to Milliron (1949, p. 384). My observations pertain to the chalcid on Rosa virginiana and other host roses, and have inclined me to the view that repetitions are not likely to occur often because the males seem to restrict themselves to old hips so long as females remain in them, and the newly emerged, mated females soon leave the site of their origin in favor of new hips of the year. While males intermingle with the ovipositing females early in the adult period of the species, they soon become the numerically minor sex, and later are wholly absent while numbers of ovipositing fe- males remain. This probably signifies that the males become widely dis- persed, or, rather, that they live a relatively short time. Being a pro- tandrous species, the males do not live so far into the season as the fe- males, even if it be assumed that their lifetimes are as long as those of the females. Matings at the sites of oviposition appeared to be of rela- tively infrequent occurrence. Oviposition. The age at which females commence to deposit eggs into the new growing hips has not been satisfactorily determined. That newly emerged individuals do not contain eggs ready for deposition was dem- onstrated by Milliron, who examined the ovaries. My observations in 1944, indicated that the first eggs are laid when the females are about five days old. According to Milliron (1949, p. 385), females that are pre- pared to begin ovipositing, frequent more consistently the ends of the branches where hips are borne. At this time, many hips are “still green and contain soft seeds in which the cotyledons are small and en- veloped in a watery or jelly-like medium. Ordinarily rose hips such as these constitute the type in which early egg depositions are made, but as the season progresses, somewhat older fruits must be selected. Rose hips are entirely unsuitable for oviposition after they become red and THE ROSE SEED CHALCIDS, Megastigmus SPP. 137 contain seeds with hard coats and firm cotyledons.” I have observed (1945, p. 193) that “oviposition is in progress when the hypanthium arid pericarps are still soft and the seeds remain in a semiliquid pulpy state.” The period of time during which seeds exist in this favorable condition for oviposition is prolonged by the development of several “sets” of hips in chronological succession. The following description of procedure in oviposition represents an integration of observations from the accounts published by Milliron (1949) and myself (1945). Employing her antennae, the female may carefully examine the entire surface of the hip before a location on it is finally selected. This place is necessarily restricted to the basal portion of the hip where the developing achenes lie. There the female stands with legs widespread to hold her to the hip surface. Wavering movements of the body begin as soon as the location is selected. The abdomen is next elevated sharply and strongly arched, and the caudal end brought for- ward ventrad. Thereby the long sheath and the contained ovipositor, or terebra, are moved ventrocephalad from their usual dorsocaudal resting position until they stand perpendicular to the hip and their tips contact its surface. Next, the sheath is released, presumably when the terebra has superficially pierced the hypanthium, and flips backward, at first assuming a horizontal posture, then gradually a vertical position as the terebra penetrates deeply into the hip. The sheaths are more rigid than the terebra itself, and therefore are a means of bringing this drill into the required position and also appear to assist in the initiation of drill- ing by preventing it from bending as pressure is applied on it from the abdomen. During the short time required to insert the terebra, the antennae are held porrect, the wings in an inclined position, and the thorax, and especially the abdomen, waver slowly and conspicuously. The ovipositor does not always enter the hypanthium in a steady con- tinuous manner, but drills by a series of alternating thrusts and partial withdrawals. When the terebra is entirely inserted, the abdomen becomes more natural in shape; its venter then almost or actually contacts the surface of the hip. A brief interval of inactivity follows the insertion of the ovipositor during which the egg is presumed to pass from the oviduct into the seed. The terebra is then withdrawn in a quick steady manner and at once snaps back into the sheath; these structures then resume the original resting posture. The female then may turn about to examine the puncture, but as often directly seeks another location to repeat the oviposition. In several instances observed in 1943 and 1944 at Urbana, particularly on Rosa virginiana, females appeared unable to withdraw the terebra from the hip, and were found dead suspended by that appendage. The time required for each of several ovipositions was between 4 and 138 INSECTS IN ROSE HIPS 7 minutes, although some required as much as 15 minutes ( Milliron). In a number of depositions observed on Rosa rugosa at Urbana, the whole process consumed from 100 seconds to 10 minutes. Milliron found that unsuccessful attempts to insert eggs are characteristically of shorter duration. Milliron (1949, p. 386) related that individual females which have deposited several eggs in a hip may become thoroughly accustomed to that hip and take possession of it, behaving in a manner as if to guard against oviposition by others. They move about excitedly with the wings elevated or inclined, and make thorough examinations of the hip at vary- ing intervals, driving away any intruding females. Occasionally the in- truder refuses to depart and a furious battle ensues in which the invading female may be the victor and herself take possession. The time of day when oviposition is performed is evidently deter- mined by the prevailing temperatures. Concerning this point, Milliron (1949) stated that “on warm days, oviposition is seldom noted in the direct rays of the sun during hours with maximum temperature,” but it “is continued in the shade during this time. Females restrict much of their oviposition on such days to the early morning or late afternoon hours.” The same relation between oviposition and temperature was observed at Urbana (Balduf, 1945). On June 17, 1943, 60 females were visible in the morning when the rugosa rose was shaded by trees and buildings; only 43 were exposed in the warmth of the early afternoon; 63 females came into view when evening shades again cooled the rose bush. Likewise, the thrice daily counts were 50, 20, and 41 on June 18, and 18, 5, and 13, respectively on June 20. In the mornings and eve- nings, the females moved about in full view on the new hips or leaves, and a number engaged in oviposition, whereas in midday they resorted to the shelter of leaves, and laid eggs only in hips that were then shaded. On the above dates, midday temperatures rose to 90 degrees or above in the sun. Egg and embryo. I found eggs and small first instars lodged in the soft tests of the seed or in the end of the seed that lay nearest the hy- panthium. Milliron (1949, p. 374) observed that “eggs are deposited entirely within the jelly-like mass surrounding the seed embryo, and the end of the anterior pedicel is attached usually to the side of the innermost envelop of the endosperm.” The terebra appeared usually to bore only through the hypanthium of the hip and into the cortex of the achene to place the egg into the seed substance, and infrequently deeper than the outer layer of achenes into the secondary underlying tier. Obviously, the depth to which eggs are placed may be determined by the relative length of the terebra and the thickness of the hypanthium, if not by the texture of the achene and THE ROSE SEED CHALCIDS, Megastigmus spp. 139 seed coat. Approximately 95 per cent of the larvae of nigrovariegatus which I dissected from achenes of Rosa rugosa in 1943-44 (1945) oc- curred in the outermost layer of fruits, the rest in the next deeper layer. We are indebted to Milliron (1949) for much of our knowledge of the embryonic, larval, and pupal stages of M. nigrovariegatus. These stages are illustrated in his Figures 32 to 44. He has described the egg as comprising a long, narrow anterior pedi- cel, an elongated oval body, and a short, spurlike posterior pedicel. Ovarian eggs measured an average of 0.985 mm. They are grayish white, turgid, and the surface smooth, shining and without ornamentation. The average length of eggs measured after deposition was 1.477 mm. The posterior end of the deposited egg usually becomes slightly attenuated, and the entire egg is somewhat more elongated. The anterior pedicel finally collapses to a ribbon-like form except sometimes its distal part. The posterior pedicel, which is sometimes wanting, is either extended, recurved toward the body of the egg, or appressed to its surface. The ovaries are “meroistic-polytrophic.” Dissections of several gravid females revealed the presence of 10 to 25 fully developed eggs per individual. These numbers do not represent the reproductive capacity of single females. In all late embryos examined by Milliron, the cephalic end was directed toward the end bearing the long, narrow pedicel which later collapses into a ribbon form. Prior to hatching, the embryo becomes separated from the chorion and _sur- rounded by a colorless fluid. Shortly before eclosion occurs, a delicate longitudinal trachea can be seen in each lateral region of specimens placed in glycerine. By their outward movement, the conspicuous mandi- bles of the late embryo appear to rupture the chorion and effect eclo- sion. Larval life. According to the studies by Milliron (1949, p. 376), the developing larva of nigrovariegatus comprises five instars. He ascer- tained this “originally by size, shape and degree of sclerotization of the mandibles, and by the presence of setae on the body segments. In order to check the instars as discovered by using these characters,” he made a limited number of measurements of the widths of the head, and these likewise indicate that five instars do occur. The size of the head capsule appears to be very significant as an aid in recognizing the second, third and fourth instars, which usually offer much difficulty. The larger setae present on the fifth instar readily distinguish it from the fourth. The instars are described fully by Milliron (1949, pp. 377-79, and his Plate 5). Because it is an excellent statement of original observations on larval development in the seeds, I quote Milliron’s account in full: “It has been pointed out that more than one egg may be placed within a single 140 INSECTS IN ROSE HIPS seed. If such be the case and the resulting larvae do not come in direct contact with each other, they may continue to develop through approxi- mately the third instar. Increase in size of the seed embryo, as well as increase in size of the larvae themselves eventually force them into contact with each other and cannibalism results. Frequently two or three larvae, representing as many instars, are found in a single seed but in no instance does more than one complete its larval development. The oc- currence of more than one young larva per seed is especially common during the early part of the oviposition period. During this time, when the seed embryo is yet comparatively small, the greater part of the seed content is watery to jelly-like in consistency, and the young larvae may be located almost anywhere inside consuming this material without any apparent damage to the seed or without impairing its growth. As the seed embryo grows, the surrounding medium becomes grayish white and paste-like, and at or during this stage a larva is commonly forced to the distal end of the seed or it may come to lie to the side of one of the cotyledons. In either event, when the larva is ready to begin feeding on the cotyledons, there is every assurance that it will not begin at the base of the seed embryo, which continues to enlarge and provide ample food for the larva to complete its development. The outer covering of the endosperm mostly remains intact until the later larval stages are attained when it is then consumed along with the basal portion of the embryo.” I have observed that the larva consumed all of the seeds excepting small hard dry brownish particles which remain scattered throughout the seed cavity and over the full-grown body. Probably incidental to its writhing movements, the full-fed larva works this frass to the narrower anterior and posterior ends of the seed cavity where it becomes com- pacted into two concave caps, shaped to the ends of the body and the cavity. The distinction between full-fed or full-grown larvae and mature larvae warrants a note. The alimentary tract of full-grown individuals is distended with a mixture of dark reddish to purple food and fecal waste, that gives the larva a dark color and fills the seed cavity of the achene so snugly that removal of the larva without puncturing the body wall requires great care. Some weeks after the larva has become full- fed, it has attained maturity, having utilized, physiologically, a consider- able part of the colored stuff in the tract so that the body is now notice- ably paler and shrunken and fills the cavity loosely. By April, i.e., after hibernation and shortly before pupation, only a small cluster of black solidifying pellets of feces remain to darken the midabdominal area of the matured body. Transformation. The process of change from larva, through prepupa, THE ROSE SEED CHALCIDS, Megastigmus spp. 141 to pupa have been described by Milliron. The larvae which are about to become prepupae are sluggish, somewhat contracted, and expel black pellets of excreta. The abdominal region becomes somewhat arched and distinctly differentiated from the thorax as the external form of the pupa within is developing. Conspicuous imaginal buds on the lateral and lateroventral surfaces of the thorax mark the approximate locations ot the wing pads and legs of the future pupa. The pupal head can be seen within the prothorax. The completed prepupa has a form intermediate between larva and pupa. At the end of this stage, the larval cuticle becomes slightly flaccid. The initial break appears on the thorax, and seems to be caused by pressure applied there by the contractions of the pupal abdomen within. The cuticle is slowly forced backward, chiefly by the contrac- tions and expansions of the abdomen. When the prepupa was removed from the normal situation in the seeds, this process required one or two hours. The changes in color undergone by the pupa from its beginning were described by Milliron (1949, pp. 381-82). The body of a pupa that is about to molt and disclose the adult, is no longer turgid; the theca loses its gloss and smoothness, and becomes greasy in appearance. The pupal theca is broken across the pronotum or just behind the head; the break results from increase in activity of the body and legs, and contractions of the abdomen that appear to concentrate pressure on the cephalic end. By its expansions, contractions and bending, the abdomen plays a most significant role in effecting the molt, although once the theca passes back to the abdomen, the hind legs may also be employed in removing it. Once exposed, the wings expand as a consequence of fluid forced into them from the body cavity, and this pressure also appears to be the principal means of removing the theca from the wings. That part of the theca still intact on the head and antennae is discarded with the aid of the front legs. Life cycle. Crosby (1909) has given data relative to the chronological distribution of the main events in the life cycle of nigrovariegatus. These show that it completes one generation in a year at Ithaca, New York. He observed that the adults emerge through round holes ‘from the “seeds” in the hips, and may be found ovipositing on the rose hips during June and July. The larvae, according to Crosby, become full-grown in the fall, when they so nearly fill the seeds that it is very difficult to remove one without crushing it. These mature larvae pass the winter, then pupate the next spring. Balduf (1945) and Milliron (1949) have pub- lished additional data, which essentially confirm while supplementing the facts from Crosby (1909). Since 1945, I have gathered many addi- tional records relative to the dates of emergence of adults and host 142 INSECTS IN ROSE HIPS relations from hips obtained in Ilinois and other states. The data pre- sented from Milliron pertain to St. Paul, Minnesota. In the years 1947 to 1951, I reared approximately 4,200 adults at Ur- bana, from hips of several Rosa spp. obtained from Alaska, northeastern Minnesota, Wisconsin, Saskatchewan and east central Illinois. In the over-all period, adults issued between May 12 and July 31, or 81 days. The yearly periods were, of course, considerably shorter, i.e., 51, 47, 57, 52 and 36 days, respectively. In 1943 and 1944, the first adults emerged from R. rugosa on June 3 and May 25 (Balduf, 1945). Unpublished records from the files of the University of Minnesota are cited by Mill- iron (1949): in 1914, emergence began at St. Paul about June 4, largest numbers emerged on June 15 and 23, and high numbers appeared con- sistently between June 22 and 26, after which, numbers per day de- creased gradually to July 7; in 1939 and 1940, Milliron noted the largest population and maximum activity during the second and third weeks of June. Hips of R. rugosa taken by me at Ashland in northern Wisconsin and transported at once to Eaglesnest, Minnesota, yielded their seed chalcids between July 2 and 29, 1945. A quantity of hips of R. acicularis picked at Eaglesnest and also wintered there, yielded the chalcids be- tween late June and July 6, 1949. These cases involving Ashland and Eaglesnest reflect the retarding influence of the cooler northern spring climate. On the other hand, lots of hips native to Llinois yielded the seed chalcids 3 to 16 days later than did lots of hips native to Maine, Minnesota, Wisconsin, the Rocky Mountains and Alaska, but wintered at Urbana. This may mean that the chalcids acquire some tolerance for the more stringent northern climate in which the larvae matured, and that this tolerance was not wholly lost in the interval spent in the moderate climate of central Illinois. Milliron (1949, p. 389) made laboratory tests to determine the sur- vivability of this seed chalcid under varied conditions of temperature and humidity. In all cases cited here, the temperature was held at 15° C., the food consisted of soaked raisins and, later, solutions of brown sugar, and the samples involved 101 to 314 individuals, except one of 17 fe- males. When the relative humidity was held constant at 30 per cent, males lived an average of 18.83 days, and the sample of 17 females av- eraged 21.24 days. Under identical conditions, but with humidity incon- stant, the average survival of males was 36.45; of females, 49.62 days. In another test, the moisture was not regulated, and no food or water were supplied. Males averaged 5.06; females, 7.65 days. From those experimental data, males may be estimated to live three or four weeks in nature, and females a somewhat longer time. If so, the chalcids may be expected to occur in nature at Urbana from late May THE ROSE SEED CHALCIDS, Megastigmus spp. 143 to the first part of August. I have seen females as late as July 30 (1947) at Urbana. According to Milliron (1949, p. 375), the incubation period of M. nigrovariegatus is “relatively short and probably does not exceed four or five days.” At Urbana, the over-all embryonic period, or “egg stage,” extends from late May to early July. I saw females ovipositing as early as June 3, 1944, and June 4, 1943, and as late as July 30, 1947, which indicates that some embryos may be found early in August if suitable rose hips for reception of the egg are then present. However, females are usually few by late June, hence the peak of the embryonic period occurs through the month of June. The earliest possible oviposition could not have been much before mid-June, 1939, at St. Paul (Milliron). Almost identical results are cited for 1940. Because the season of 1941 was two weeks earlier, Mill- iron believed oviposition could not have begun before the first week of June. The peak of oviposition appeared to be during the third week of June, 1939, and the activity had ceased by July 10. In 1940, females were seen ovipositing until the middle of July, but in 1941 the peak was reached by mid-June; none were found ovipositing between July 3 and it At Eaglesnest, I saw females oviposit on wild roses from July 15 to 26, 1945, and on July 15, 1950. Since these were but casual observations, it is reasonable to assume that the species probably oviposits there both before and after the dates given. Because the period of embryogenesis is probably only four or five days long, the inception of larval life is almost simultaneous with that of ovi- position. Milliron (p. 376) learned from dissections of inhabited achenes that growth of the larva through all its instars is completed in a month or less. Full-grown larvae were commonly obtained by July 19, in 1939, and by almost the same date in 1940. He discovered numerous submature larvae in achenes on June 30, 1941. In Urbana, at intervals of time, I dissected samples of achenes which showed that the over-all develop- ment of the larvae from hatching to maturity, or near it, took place from mid-June to late July, 1943, and 1944. However, the larval period does not terminate when the larva reaches its full-fed, full-grown state through late June and the first two-thirds of July. With few exceptions, the species persists, without perceptible change, as a full-grown larva through the remainder of the summer, the fall and winter, and to about April 20 to early June of the next year. Since this diapause is initiated in the warmest season of the year, it apparently may not be attributed to low temperatures, but probably to some nutritional or chemical factor in the seed substance ingested by the later larval instars. 144 INSECTS IN ROSE HIPS The full-grown, mature larval state usually terminates with transforma- tion to the pupal form in the spring following the year of larval devel- opment. For exceptions, see “Precocious pupae,” below. At St. Paul, this transformation occurred “during the latter part of April or the early part of May,” according to Milliron (1949). He found also that the pre- pupal period averages about three days in length under laboratory conditions. The pupal stage of females varied from thirteen to sixteen days, and that of the males was about twelve days. “Under conditions of cooler outside temperature, this period is probably considerably longer. On the other hand, full-grown larvae retarded until July, when both outside and inside temperatures were approximately the same, passed through the pupal stage as follows: female, twelve days; male, eight days.” Achenes dissected at Urbana at intervals during April and May showed that pupation began during the last ten days of April, 1943, and 1944; also that the peak of pupation was reached between May 8 and 22, and pupal life terminated with emergence of the adult from mid-May to June 20 of these years. Precocious pupae. While dissecting achenes of Rosa rugosa at Ur- bana, I discovered five pupae of M. nigrovariegatus in decidedly unusual seasons of the year, one on November 28, 1942, and one each on July 16, 21, 28, and 29, 1944, respectively. Instead of delaying pupation until the following spring, as is usual for the species, these had transformed to pupae in the same year in which the larvae developed. One of the pupae of 1944 was still white when found, but the rest had advanced to the brown color phase, which suggested that they would soon have become adults. Both sexes were represented, and all were large and vigorous in appearance, reflecting the favorable volume of the seed contained in the achenes of this rose. These physical features of the seed may afford clues to the cause of the premature pupation observed here. An alternate possibility as to the cause of the precocious pupation is suggested by the work of Williams (1958) on the juvenile hormone in certain macromoths, and the observations by Milliron (1949) on can- nibalism in this rose seed chalcid. In his studies Williams implanted the corpora allata of the young larva of Bombyx mori Linn. into the fifth instar, and found the latter continued to grow and then changed into a giant pupa and adult. Milliron observed that cannibalism among the larvae of Megastigmus nigrovariegatus is common in the achenes of Rosa rugosa, and eventually only one individual survives. These facts from Milliron and Williams served to suggest that larger larvae of the seed chalcid, when consuming smaller ones in the achene incidentally obtain an increased supply of juvenile hormone that may influence the de- velopment of larger-than-normal larvae, followed by precocious pupa- PARASITES OF Megastigmus nigrovariegatus 145 tion, quite apart from the food supply, indicated above as the cause of precocity. Acceleration of the life cycle. Although the single annual cycle is normal for M. nigrovariegatus, its duration may be greatly shortened by subjecting achenes inhabited by the mature larvae to unnatural treat- ment. This fact was established experimentally by Milliron (1949, p. 383). In late summer, the environmental temperature of such achenes was gradually lowered to zero centigrade, or slightly below, and held for approximately a month, and then gradually returned to room tem- perature. A few adult males emerged from these achenes during the middle of October. The larval diapause was shortened by about six months. Parasites OF Megastigmus nigrovariegatus Eupelmus, Eupelmidae, Hymenoptera In the spring of 1944, I found a few dead, dry greenish chalcids in the achenes of R. rugosa at Urbana. Dr. B. D. Burks identified them as a species of Eupelmus. The circumstances of their occurrence indicated they were parasites of M. nigrovariegatus. In subsequent years, I obtained 25 living adults of an Eupelmus sp. from jars containing rose hips acquired as follows: Taylorville, Illinois, one male, May 27, 1947, from old hips of Rosa canina; San Jose, Hlinois, one male, August 16, 1951, from live hips of R. carolina picked July 21, 1951; Eaglesnest, Minnesota, two females, September 18 and 21, 1951, from new hips of R. acicularis; Chetek, Wisconsin, one male and twenty females, September 9 to 30, 1948, 1950, and 1951, from hips of R. blanda and carolina. It is now not known whether one or more species are represented in this series of 25 adults, nor has it been established what the hosts may be. It is possible this is the same species found dead in the achenes of rugosa, as stated above. Again, it may be the species whose first instar I discovered in the larva of Rhagoletis basiola (see p. 50). Evidence from the host relations of Eupelmus spp. recorded in the literature (Muesebeck, et al., 1951), would indicate that it may attack directly either Rhynchites bicolor or Cydia packardi, or some of the primary parasites of these phytophags in rose hips. ADDITIONAL Parasitic HYMENOPTERA FROM Hip JARS Ten species of parasitic Hymenoptera have been obtained in addition to those previously described in connection with their hosts. These orig- 146 INSECTS IN ROSE HIPS inated in jars containing hips of several species and localities. All are represented by one to several specimens. I have no certain information about their host relations. Below is given a list of the species merely for the sake of whatever interest they may prove to be in the future. Trimorus columbianus (Ashm.), Scelionidae, Proctotrupoidea. Haltichella sp., Eupelmidae, Chalcidoidea. Eupelmella vesicularis (Retz.), Eupelmidae, Chalcidoidea. Tetrastichus faustus Burks, Tetrastichidae, Chalcidoidea. Zatropis incertus (Ashm.), Pteromalidae, Chalcidoidea. Sympiesis ancylae Gir., Eulophidae, Chalcidoidea. Campoplex sp., Ichneumonidae, Ichneumonoidea. Pycnocryptus director (Thbg.), Ichneumonidae, Ichneumonoidea. Cymodusa distincta (Cr.), Ichneumonidae, Ichneumonoidea. Orthopelma mediator (Thbg.), Ichneumonidae, Ichneumonoidea. SUMMARY Thirty species of animals are now known to comprise a microcom- munity which centers in the rose hip. With the exception of the mite, Caloglyphus, and the annelid worm, Chaetogaster, all the members of the community are holometabolous insects. Their larvae develop either in the hypanthium or the achene cavity of the hip, hence are obligatory, while the embryo, pupa and adult sustain no single positional relation to the hip. The hexapodous adherents to the community con- stitute three distinct bionomic groups, or strata, when classified with reference to food preferences and the niches they occupy. First, are the phytophags which feed directly upon the living substance of the hip; second, the predatory and parasitic members, which attack some stage or stages of the phytophags, and third, species of scavengers that feed upon the waste materials that result from the first and second groups. The bionomic features of these several groups are summarized below in the sequence in which they are described in the foregoing pages. The phytophagous species are members of four major orders of insects, Diptera, Lepidoptera, Coleoptera and Hymenoptera. The Diptera are represented by three species of Trypetidae—Carpomyia Schineri and Rhagoletis alternata of Europe, and Rhag. basiola, widely distributed in northern North America. Rhag. basiola places the eggs singly in pits bored into the hypanthium, the larva feeds on the fleshy tissue here and, when mature, emerges and pupariates, presumably amid duff on the ground. Development then advances to the early phase of the pupa, in which stage the species winters. Pupation is resumed in May and SUMMARY 147 June of the next spring, and flies appear as the new crop of hips de- velops during the summer. The observations made in Europe on Rhag. alternata and Carpomyia Schineri strongly indicate that their relation to the hips and the sea- sonal distribution of the life stages are very similar to those of Rhag. basiola. Rhag. basiola is attacked by four species of solitary endoparasitic Hymenoptera—two Braconidae, Opius baldufi and O. rosicola; a ptero- malid chalcidoid, Halticoptera rosae, and a species that appears to be a eupelmid chalcidoid. The species of Opius place eggs into the instars of Rhag. basiola, the larva persisting as a first instar until the host larva matures, leaves the hip and pupariates. The parasite then molts and quickly completes its larval growth in the puparium, consuming the host before it can pupate, and passes the winter as a mature larva in the puparium. Pupation and emergence of the adult occur in May to July after the hibernating period. In contrast with Opius, Halticoptera rosae inserts the egg into the egg of Rhag. basiola in the hypanthium. Its larval life begins in the host embryo, and continues as a first instar while the host undergoes its three larval stages and winters as an early phase pupa. The first molt apparently occurs when the host resumes pupation the next spring. In a few days, the parasite completes its growth, destroying the host pupa. H. rosae pupates in the puparium of Rhag. basiola, and the emergence of the adult in June and July antedates that of the Opius species. Only the first instar of the eupelmid species was seen. It occurred both in the larval stages and the puparia of Rhag. basiola, and hence may undergo a prolonged diapause in the first instar, as H. rosae does. The moths, Carposina scirhosella, Carposinidae, and Cydia roseti- colana, Olethreutidae, are native to Europe, whereas Cydia packardi is American. The female packardi cements the eggs mostly to the “neck” of rose buds and rose hips. There is one generation yearly in northeastern Minnesota but two in east central Illinois, the first developing largely in buds, the second in hips. The caterpillar feeds within the rose bud and in the hypanthium of the hip, and, when mature, emerges to spin up in various places of concealment. The winter is passed as a mature larva, and pupation takes place in April and May. The observations made on Cydia roseticolana and Carposina scir- hosella in Europe indicate that their cyclic development and relation to the hip are very similar to those of Cydia packardi. The larvae of two species of Cleridae, Coleoptera, preyed on the larvae of C. packardi in the hypanthium. The mature predators crawled free in hip jars. They are members of the genera Phyllobaenus and Enoclerus(?). Small numbers of solitary ectoparasites—the ichneumonid, 148 INSECTS IN ROSE HIPS Scambus hispae and the eulophid, Euderus cushmani—attacked the larvae of packardi in the hips. Three solitary endoparasites of the larvae of Cydia packardi were dis- covered. Of these I know two only as first instars; one is probably a species of Ascogaster, Braconidae; the other I call the “long-tailed para- site.” The third is the ichneumonid, Glypta rufiscutellaris, most numerous of the three. Glypta attacks also the oriental fruit moth, Cydia molesta, and many other lepidopterous larvae living in concealment in plants. The number of cycles developed in a year is said to coincide with that of the host. The terebra is inserted into the openings to the host’s bur- row, and the egg placed in the small instars of the host. The mature parasite leaves the dead host and pupates in a cocoon. The rose curculio, Rhynchites bicolor bicolor, Curculionidae, Coleop- tera, of the eastern United States, prepares an egg pit in the hypanthium, and through it, usually places the egg on the superficial achenes in the hip. The larvae eat the seeds, appearing to shun the achene shells or pericarps. The mature larva drops to the ground and forms a cell which serves as a hibernaculum and, in the spring, for pupation. The adults issue from the ground in May and June when the new crop of hips is developing. The life stages of Rhyn. bicolor wickhami, investigated by Hoerner (1936) in Colorado, are distributed seasonally like those of bicolor bicolor. However, the female wickhami places the eggs in rose buds, and the larvae feed and develop in the mummified buds. Observers of Rhyn. bicolor vars. in areas of northwestern America have reported lo- cations of the eggs and larvae that differ, in some cases, from those de- scribed by Hoerner for Colorado and Balduf for Ilinois. Chaetogaster sp.—an oligochaete annelid worm—was found associ- ated with the pupae of Rhyn. bicolor in Illinois in a manner that sug- gests a parasitic relation. Ectoparasites of Rhyn. bicolor. Four species of solitary ectoparasitic Hymenoptera were reared in small numbers that attacked the larvae of bicolor in hips in Illinois. These are Bracon mellitor and B. variabilis, Braconidae, which are reported to paralyze the host before oviposition; and Zatropis rosaecolis, Pteromalidae, and Eurytoma sp., Eurytomidae. All pupate in the rose hips near the remnants of the host. The ichneumonid, Luchatema baldufi, is the most numerous parasite — of Rhyn. bicolor in the Midwest. It is solitary and endoparasitic in the larva. The terebra is inserted through the egg pit made by the female beetle, and the egg deposited in the small instars of the host. Only the first instar is present while the host feeds in the hip and winters in the soil. The parasite appears to complete its first molt when the host is about to prepupate in spring, and quickly completes its larval growth, SUMMARY 149 destroying the host before it can pupate. The parasite pupates in the host’s hibernaculum, and the adults emerge from the ground in June and July when the young host larvae again occur in the new hips. The rose seed chalcids—Torymidae, Megastigmus aculeatus, native to Europe, and M. nigrovariegatus, the American species, are very sim- ilar in their relations to rose hips. The long terebras penetrate the hypan- thium and the then still soft-walled achene, and place the egg in the soft pulpy seed. Here the larvae feed and develop during June and July; the mature larvae occupy the achenes through the fall and winter. Pupation is an event of May and June, and the adults chew out through the pericarp of the achene and the overlying hypanthium. The achenes in the new hips are attacked chiefly in late June and July. Adults of an Eupelmus sp. found dead in a few achenes of R. rugosa at Urbana were the only parasites obtained from Megastigmus. It ap- peared to be a very infrequent parasite on the larva of M. nigrovarie- gatus. Two species of scavengers were discovered, both in only a few in- stances. The puparia of Lonchaea polita, Lonchaeidae, occurred in va- cated burrows of Rhag. basiola and Cydia packardi in old hips in jars, and adults were reared from same. The acarine mite, Caloglyphus spinitarsus, developed on dead larvae of Rhyn. bicolor and remnants of the parasite, Luchatema baldufi, in the hibernacula of the beetle. Rhag. basiola and its endoparasites, Opius and Halticoptera, have strictly univoltine life cycles, as do also Rhyn. bicolor and its endo- parasite Luchatema, and the seed chalcids, Megastigmus spp. All also exhibit more or less prolonged diapauses—Rhag. basiola as an early phase pupa in the puparium, Opius spp. as mature larvae in the host puparia, and Halticoptera as a mature first instar in the wintering phase pupa of the host fly. Rhyn. bicolor diapauses as a mature larva, and its parasite, Luchatema, as a mature first instar in the wintering mature host larva. Both species of Megastigmus diapause as mature larvae from August to the next May in the achene shells in the hip. A small per- centage of M. nigrovariegatus pupated precociously. The above phytophags and their parasites, so far as now known, seem obligated to the rose hip and their rose-hip-inhabiting hosts, respectively. By contrast, Cydia packardi develops in the fruits of various plant groups besides Rosa, and its principal endoparasite, Glypta, likewise utilizes a number of hosts other than C. packardi. The number of generations developed in a year by both Cydia and Glypta is not uniform through- out their geographical range. Also Bracon mellitor and B. variabilis, ectoparasites on larvae of Rhyn. bicolor, attack other hosts of similar habit, and develop at varying rates in the over-all area of their occur- rence. Rost Hosts oF Hip Insects Rosa acicularis Lindl. Rosa acicularis bourgeauiana Crep. Rosa acicularis engelmanni Crep. Rosa alcea Greene Rosa alpina L. (=pendulina L.) Rosa arkansana Porter Rosa arkansana suffulta (Greene) Cock. Rosa blanda Ait. Rosa blanda glandulosa Schuette Rosa californica Ch. and Schl. Rosa calocarpa Willm. Rosa canina L. Rosa carolina L. Rosa carolina villosa L. Rosa Chaberti Desegl. Rosa cinnamomea L. Rosa dawurica Pall. Rosa eglanteria L. (=rubiginosa L.) Rosa gallica L. Rosa gallica damascena Voss. Rosa gymnocarpa Nutt. Rosa hugonis Hemsl. Rosa jundzilli Bess. Rosa lucida Ehrh. Rosa lunelli Greene Rosa Macounii Greene Rosa macrantha Desf. Rosa mollis Smith Rosa mollissima Willd. Rosa Moyesii Hemsl. and Wils. Rosa multibracteata Hemsl. and Wils. Rosa multiflora Thunb. Rosa nitida Willd. Rosa nutkana Presl. Rosa oxyodon Boiss. Rosa palustris Marsh Rosa parviflora Ehrh. Rosa pomifera Herrm. Rosa pyrifera Rydb. Rosa rubrifolia R. Br. Rosa rugosa Thunb. Rosa rugosa kamtschatica (Vent.) Regel Rosa sempervirens L. Rosa sericea Lindl. Rosa setigera Michx. Rosa spaldingii Crepin Rosa spinosissima L. Rosa suffulta Greene Rosa ultramontana (S. Wats.) Heller Rosa villosa L. Rosa virginiana Mill. Rosa virginica Roess. Rosa woodsii Lindl. Rosa xanthina Lindl. Rosa spp. Appendix Rosa alpina var. pyrenaica (?) Rosa canina vars. Rosa dumetorum (?) Rosa feneliensis (?) Rosa glauca (?) Rosa hugonis platyacantha (?) Rosa medwedewii (?) Rosa moschata alba (?) Rosa Selongiana (?) Rosa swenginzowii (?) BIBLIOGRAPHY Anon. (1926), Outline of the plant quarantine service of Japan. Yokohama: Imperial plant quarantine service, 18 pp. (In Rev. Appl. Ent., A, 15[1927]: 181.) Anon. (1934), Entomology, Pa. Dept. Agr. Bul. 17, pp. 26-30. Anon. (1938), Entomological problems. Rept. Council Sci. Industr. Res., Australia, 12(1937-38), pp. 22-27. Anon. (1941), Nyponflugan [The rose hip fly], Flygbl. Vaxtskyddsanst. No. 57; 4 pp., 2 figs. Stockholm. (In Rev. Appl. Ent., A, 31[1943]:322.) Ashmead, W. H. (1890), On the Hymenoptera of Colorado; descriptions of new species, notes and a list of the species of the state. Colo. Biol. Assoc. Bul. 1, pp. 3-47. Baird, A. B. (1929), The present status of corn borer parasites in Canada. Ann. Rept. Ent. Soc. Ont., 59, pp. 38-40. Baker, W. S. (1930), Insect pests of sour cherries and berries. Proc. 26th Ann. Meeting, Wash. St. Hort. Assoc., pp. 207-12. Balduf, W. V. (1926), On the bionomics of some Hymenoptera from a bur oak cynipid gall. Can. Ent., 58:135-43. . (1937), Bionomic notes on the common bagworm, Thyridopteryx ephemeraeformis Haw. and its insect enemies. Proc. Ent. Soc. Wash., 39: 169-84. . (1945), Bionomics of the rose seed chalcid, Megastigmus nigrovarie- gatus Ashm: (Hym., Callimomidae). Proc. Ent. Soc. Wash., 47:185-98. . (1957), The weight of puparia of Rhagoletis basiola (O.S.). Ent. News, 68:163-65. . (1957a), Variations in Megastigmus from rose hips (Torymidae, Hymenoptera). Ann. Ent. Soc. Amer., 50 (5) :478-83; plate. . (1958), The puparium of Rhagoletis basiola (O.S.): variation in length and form. Ohio Jour. Sci., 58 (1):7-14. Barrett, C. G. (1872-73), On the British Tortrices of the genera Dicrorampha and Endopisa, as restricted by Wilkinson. Ent. Mo. Mag., 9:25-31. Basinger, A. J. (1935), Notes on the parasites of the peach twig borer in southern California (Anarsia lineatella Zell.). Calif. Dept. Agr. Mo. Bul. 24:245-46. Belski, B. I. (1924), List of pests of medicinal plants from observations made in the environs of Kiev during 1919-1922. Trans. 4th All-Russian Entomo- Phytopath. Meeting, Moscow, Dec. 1922, pp. 158-70. Petrograd. (In Rev. Appl. Ent., A, 12[1924]:358. ) Bibby, F. F. (1942), Some parasites of Heliothis armigera (Hbn.) in Texas. Jour. Econ. Ent., 35:943-44. Bigger, J. H. (1931), Another parasite of the sunflower weevil, Desmoris fulvus Lec. Jour. Econ. Ent., 24:558. . (1933), Parasites of the sunflower weevil, Desmoris fulvus Lec., dur- ing 1931 and 1932. Jour. Econ. Ent., 26:652. Bissell, T. L. (1940), Curculionidae, Bruchidae, Lepidoptera and their para- sites, infesting the seed pods of cowpeas and various wild plants. Jour. Econ. Ent., 33:844-47. 151 152 INSECTS IN ROSE HIPS Blackburn, Norris D. (1944), Some ecological studies on the parasites of the oriental fruit moth (Grapholitha molesta [Busck.]) in Ohio. Doctoral dis- sertation. The Ohio State University, Columbus. Blatchley, W. S., and Leng, C. W. (1916). Rhyncophora or weevils of north eastern America. Indianapolis: The Nature Publishing Co. 682 pp. Bobb, M. L. (1939), Parasites of the oriental fruit moth in Virginia. Jour. Econ. Ent., 32:605-607. Boese, Gerhard (1936), Der Einfluss tierischer Parasiten auf den Organismus der Insekten. Zeitschr. f. Parasitenkunde, 8:243-84; figs., refs. Boving, A. G., and Champlain, A. B. (1920), Larvae of the North American beetles of the family Cleridae. Proc. U.S. Nat. Mus., 57:575-649; bibl., 13 pls. Boheman, C. H. (1833), Skandinaviska Pteromaliner. Svensk. Kongl. Vetens. Akad. Handl., 54:329-80. Boie, F. (1847), Zur Entwicklungsgeschichte mehrer Trypeta-Arten. Stett. Ent. Zeitung, 8:326-31. Bouché, P. F. (1833), Naturgeschichte der schadlichen und niitzlichen Garten- Insekten und die bewahrtesten Mittel zur Vertilgung der ersteren. Berlin. Pp. i-vi, 1-176. (1834), Naturgeschichte der Insekten besonders in Hinsicht ihrer ersten Zustande als Larven und Puppen. Erste Lief. Berlin. 216 pp. Boyce, H. R. (1946), Long term trends in parasitism of twig-infesting oriental fruit moth larvae. Ann. Rept. Ent. Soc. Ont., 77, 21-34. Breakey, E. P., and Webster, R. L. (1939), Life history and control of the cherry fruitworm in western Washington. Wash. Agr. Exp. Sta. Bul. 384 (or 49th Ann. Rept. for year ended June 30, 1939), p. 34. Bridwell, J. C. (1920), Miscellaneous notes on Hymenoptera. Proc. Hawaiian Ent. Soc. (1918-20), 4:109-65. Britton, W. E. (1928), Twenty-eighth report, state entomologist of Connecti- cut, Conn. Agr. Exp. Sta. Bul. 305, pp. 666-768. . (1938), Additions to the check-list of the insects of Connecticut. Conn. Geol. and Nat. Hist. Surv. Bul. 60, pp. 9-169. Brooks, F. E. (1918), The grape curculio. USDA Bul. 730, pp. 1-19. Brunson, M. H., and Allen, H. W. (1948), Oriental fruit moth coccon para- sites. Jour. Econ. Ent., 41:446-50. Buchanan, L. L. (1939), Changes in names in Carabidae and Rhynchophora (Coleoptera). Proc. Ent. Soc. Wash., 41:79-82. Burks, B. D. (1955), A new species of the genus Halticoptera (Hym., Ptero- malidae). Proc. Ent. Soc. Wash., 57:81-82. . (1955a), A redescription of the genus Zatropis, with descriptions of three new species (Hymenoptera, Pteromalidae). Proc. Ent. Soc. Wash., 57:31-87. Butler, H. G. (1933), Larval parasites of the oriental fruit moth in Roane County, Tenn. Jour. Econ. Ent., 26:982-87. Cameron, Peter (1879), On some new or little known British Hymenoptera. Trans. Ent. Soc. London, 27:107-19. Cameron, Ewen (1938), A study of the natural control of the pea moth, Cydia nigricana Steph. Bul. Ent. Res., 29:277-318; pls., refs. . (1941), The biology and post-embryonic development of Opius ilicis n. sp., a parasite of the holly leaf-miner (Phytomyza ilicis Curt.). Parasitol- ogy, 33:8-39; figs., refs. BIBLIOGRAPHY 153 Carpenter, J. M. (1944), The effect of population density on the productivity of the parasite Microbracon mellitor Say. Proc. and Trans. Texas Acad. Sci. (for 1943), 27:69-70. Cassidy, James (1888), Notes on insects and insecticides. Colo. Agr. Exp. Sta. Bul. 6, pp. 1-20. Chittenden, F. H. (1893), The strawberry weevil, Anthonomus signatus Say. Insect Life, 5:167-86. . (1901), Some insects injurious to the violet, rose and other ornamen- tal plants, USDA Div. Ent., n.s. Bul. 27, pp. 1-114. . (1918), The lotus borer, Jour. Econ. Ent., 11:453-57. Claassen, P. W. (1921), Typha insects: their ecological relationships. Cornell Univ. Agr. Exp. Sta. Memoir 47, pp. 457-531. Clausen, C. P. (1940), Entomophagous insects. New York: McGraw-Hill. 688 pp. Cockerell, T. D. A. (1894), Notes from New Mexico. Insect Life, 7:207-11. . (1899), A first list of the hymenopterous families Tenthredinidae, Siricidae, Chrysididae, Proctotrypidae, Cynipidae, Chalcididae, and Bra- conidae collected in New Mexico. Trans. Kan. Acad. Sci., 26:212-16. . (1912), Rhynchites bicolor. Ent. News, 23:82. , and Harris, R. C. (1924), A new form of Rhynchites (Coleop., Cur- culionidae). Ent. News, 35:144. Cockerham, K. L. (1944), Some parasites of the sweet potato weevil. Jour. Econ. Ent., 37:546-47. Collin, J. E. (1947) The British genera of Trypetidae (Diptera), with notes on a few species. Ent. Record and Jour. of Variation, 59[1]:(1)-(8). Cooley, R. A. (1903), Two insect pests. Mont. Agr. Exp. Sta. Bul. 46, pp. 107-18. . (1925), Montana insect pests for 1923 and 1924, being the 20th Report of the state entomologist of Montana. In Univ. Mont. Agr. Exp. Sta. Bul. 170. 30 pp. Coppel, H. (1946), The collection of spruce budworm parasites in British Columbia with notes on their overwintering habits. Ann. Rept. Ent. Soc. Ont., 77, pp. 38-40. Coulon, L. (1931), Catalogue systematique et biologique de la famille des Hymenoptera chalcidides du musée d’Elbeuf (Collection europienne). Bul. Soc. Sci. Nat. d’Elbeuf, 49:73-86. Crane, H. A. (1932), Spilographa setosa Doane. Calif. Agr. Exp. Sta. Mo. Bal 2p. 217. Craw, Alexander (1894), Entomology and quarantine. Fourth biennial re- port, Calif. State Bd. Hort., for 1893-1894, pp. 79-109. Crawford, A. W. (1933), Glypta rufiscutellaris Cr., an ichneumonid larval parasite of the oriental fruit moth. N. Y. Agr. Exp. Sta. Tech. Bul. 217. 29 pp.; figs., refs. Crawford, J. C. (1915), The genus Secodella in North America. Proc. Ent. Soc. Wash., 17:142-44. Cresson, E. T. (1870), Descriptions of new species belonging to the sub- family Pimplariae found in America north of Mexico. Trans. Amer. Ent. Soc., 3:143-72. Crosby, C. R. (1909), On certain seed-infesting chalcis-flies. Cornell [N. Y.] Agr. Exp. Sta. Bul. 265, pp. 377-79. . (1913), A revision of the North American species of Megastigmus Dalman. Ann. Ent. Soc. Amer., 6:155-70. 154 INSECTS IN ROSE HIPS , and Leonard, M. D. (1922), Insects that hunt the rose. Amer. Rose Annual (Amer. Rose Society. Harrisburg), pp. 89-100. Cushman, R. A. (1927), The parasites of the pine tip moth, Rhyacionia frustrana (Comst.). Jour. Agr. Res., 34:615-22. Dalla Torre, C. G. de (1898), Catalogus Hymenopterorum, Chalcididae et Proctotrupidae, Vol. 5. Leipzig. 598 pp. Daniel, D. M. (1930), Oriental fruit moth parasite work in New York. Jour. Econ. Ent., 23:215-17. Dever, D. A. (1953), An analysis and study of the cherry insect pest com- plex in Wisconsin. Doctoral dissertation, Univeristy of Wisconsin, Madison. . (1954), Identification of the larvae of the more important insect pests of sour cherry in Wisconsin. Wis. Acad. Sciences, Arts and Letters, 43: 83-88. Dickerson, E. L. (1910), Notes on Rhynchites bicolor Fabr. Jour. Econ. Ent., 3:316-17. Doane, R. W. (1899), Notes on Trypetidae with descriptions of new species. Jour. N. Y. Ent. Soc., 7:177-98; pls. Dohanian, S. M. (1942), Parasites of the filbert worm. Jour. Econ. Ent., 35: 836-41. Doner, M. H. (1936), Hymenopterous parasites of Coleophora pruniella Cl. and parasites recorded from other species of Coleophora. Ann. Ent. Soc. Amer., 29:224-44. Downes, W. (1929), The cherry fruit worm, Grapholitha packardi Zell. Proc. Ent. Soc. British Columbia, 26:39-43. . (1929a), The cherry fruit worm and its control in British Columbia. Can. Dept. Agr., Ent. Branch, Circ. 79, pp. 1-3. Doyer, L. C. (1926), Seed injury from fungi and insects. Internat. rev. of the science and practice of agr. (Internat. Inst. of Agr., Wageningen), n.s. iv, No. 1, pp. 150-54. Driggers, B. F. (1930), Oriental peach moth investigations. Rept., N. J. Agr. Exp. Sta., 1929-1930, pp. 148-59. . (1940), Oriental fruit moth parasitism as related to infestation. Jour. Econ. Ent., 33:353-57. Dunning, R. A. (1956), A diurnal rhythm in the emergence of Pegomyia betae (Curtis) from the puparium. Bul. Ent. Res. 47:645-53. Eppelsheim, F. (1871), Beitrage zur Entwicklungsgeschichte der Klein- schmetterlinge. Stett. Ent. Zeitung, 32:315-16. Escherich, K. (1923), Die Forstinsekten Mitteleuropas. Bd. 2. Berlin. 663 pp. Essig, E. O. (1926), Insects of western North America. New York: Mac- millan Co. 1035 pp., 766 figs. Fallén, Carl Fr. (1820), Diptera Sueciae. No. 15, Ortalides, pp. 1-34. Fenton, F. A., and Dunnam, E. W. (1929), Biology of the cotton boll weevil at Florence, $.C. USDA Tech. Bull. 112, pp. 1-75. Fletcher, J. E. (1883), Notes on Diptera. Ent. Mo. Mag., 20:163. . (1883a), Alleged breeding of Trypeta alternata Fall., from Impatiens. Ent. Mo. Mag., 20:275. Foerster, Arnold (1841), Beitrage zur Monographie der Pteromalinen Nees. Heft 1. Aachen. 46 pp. . (1859), Zweite Centurie neuer Hymenoptera. Verh. Naturhist. Verein preuss. Rheinl. und Westph., 16:87-124. Folsom, J. W. (1936), Observations on Microbracon mellitor (Say), in re- lation to the boll weevil. Jour. Econ. Ent., 29:111-16. BIBLIOGRAPHY 155 Forbes, W. T. M. (1923), Lepidoptera of New York and neighboring states. N. Y. [Cornell] Agr. Exp. Sta. Memoir 68. 729 pp., 439 figs. Fraenkel, G., and Brookes, V. J. (1953), The process by which the puparia of many species of flies become fixed to a substrate. Biol. Bul., 105(3): 442-49, Friend, S. B. (1942), Forty-first report, 1941. State entomologist of Connecti- cut. Conn. Agr. Exp. Sta. Bul. 461, pp. 463-548 [+3]. Fuchs, Gilbert (1912), Generationsfragen bei Riisselkafer. Naturw- Zeige f. Forst- u. Landw., 10:43-54. Garlick, W. G. (1934), Some observations on the grape berry moth. Ann. Rept. Ent. Soc. Ont., 65, pp. 108-12. Garman, Philip (1918), A comparison of several species of Lepidoptera in- festing peach and apple in Maryland, with additional notes on the oriental peach moth. Md. Agr. Exp. Sta. Bul. 223, pp. 103-26. . (1930), The oriental peach moth in Connecticut. Conn. Agr. Exp. Sta. Bul. 313, pp. 401-51; 93 refs. Gates, B. N. (1909), The rose curculio (Rhynchites bicolor Fabr.) in Massa- chusetts. Jour. Econ. Ent., 2:465-66. Gentner, L. G. (1925), Some important grape insects. Mich. Agr. Exp. Sta., Special Bul. 148. 19 pp. Gibson, Arthur (1915), The entomological record, 1915. Notes on captures. Ann. Rept. Ent. Soc. Ont., 46, pp. 194-232. Gillette, C. P., and List, G. M. (1921), The western rose snout beetle, Rhyn- chites wickhami Ckl. Colo. Ent. Cire. 34, p. 20. Gobert, Emile (1887), Catalogue des Diptéres de France. Caen. 88 pp. Grandi, G. (1937), Parasitti della Cydia molesta Busck riscontrati nell Emilia. Boll. Ist. Ent., Bologna, 9:250-52. Green, J. W. (1920), Notes on American Rhynchophora. Ent. News, 31:193- 201. Griswold, Grace (1937), Common insects of the flower garden. N. Y. [Cor- nell] Ag. Exp. Sta. Bul. 371. 59 pp. Guilliaume, A. (1904), Matériaux pour un catalogue des Trypetidae de la Belgique. Ann. Soc. Ent. Belgique, 48:426-31. Haden, W. R. (1935), Parasitism of the oriental fruit moth, with special ref- erence to the importance of certain alternate hosts. Del. Agr. Exp. Sta. Bul. 194, 42 pp. Haeussler, G. J. (1930), Parasites of the oriental peach moth, Laspeyrezia molesta Busck, in North America. Jour. Agr. Res., 41:365-77. Hammer, O. H. (1936), The biology of the apple curculio (Tachypterellus quadrigibbus Say). N. Y. [Geneva] Agr. Exp. Sta. Tech. Bul. 240. 50 pp. Hardy, D. Elmo (1949), Studies in Hawaiian fruit flies. Proc. Ent. Soc. Wash., 51:181-205; figs., refs. Hardy, James (1873), Memoirs on Scottish Diptera, No. V. Trypeta alternata Fallén (The Rose-hip Fly). The Scottish Naturalist, 11:21-23. Heinrich, Carl (1926), Revision of the North American moths of the sub- families Laspeyresiinae and Olethreutinae. U.S. Nat. Mus. Bul. 132. 216 pp.; figs., pls. Hellén, Wolter (1934), Verzeichnis der Callimomiden (Hym., Chale.) Finn- lands nach Bestimmungen von E. Hoffmeyer. Memoranda Soc. Faun.-Flor. Fenn., 9:184-91. Hendel, Friedrich (1927), Trypetidae, in Lindner, Erwin: Die Fliegen der Palaearktischen Region. Bd. 5, Pt. 2, Fam. 49, pp. 1-221, Lief. 16. 156 INSECTS IN ROSE HIPS Herbst, J. F. W. (1796), Natursystem aller bekannten in- und auslandischen Insecten. 11 vols. Berlin. Hixon, E. (1935), Boll weevil control results for 1933-1934. Rept. Okla. Agr. Exp, Stas pa 272: Hoerner, J. L. (1936), Western rose curculio, Rhynchites bicolor wickhami. Ckl. Colo. Agr. Exp. Sta. Bul. 482, pp. 1-19; figs., refs. , and List, G. M. (1952), Controlling cherry fruitworm in Colorado. Jour. Econ. Ent., 45:800-805; fig. refs. Hoffmeyer, E. B. (1929), Aus Samen geziichtete Callimomiden. Ent. Meddel., 16:324-26. . (1930), Beitrage zur Kenntnis der danischen Callimomiden, mit Be- stimmungstabellen der europaischen Arten (Hym., Chalc.). Ent. Meddel., 17:232-85. Horn, W., and Schenkling, Sigm. (1928), Index Litteraturae Entomologicae. Vols. I, II, Ill, pp. 1-1426. Dr. Walther Horn, publ. Howard, L. O. (1892), The biology of the hymenopterous insects of the family Chalcididae. Proc. U.S. Nat. Mus., 14:567-88. Hunter, W. D., and Hinds, W. E. (1904), The Mexican cotton boll weevil. USDA, Div. Ent., Bul. 45. 116 pp. Imms, A. D. (1931), Recent advances in entomology. Philadelphia: Blakiston. 374 pp. Ishii, T. (1932), In Yano, M., Nippon Konchu Zukan, Iconographia insectorum japanicorum. Tokyo. 2470 pp., 124 pls. Jeffreys, F. J. (1939), The raspberry-bud moth (Carposina adreptella Walk.). New Zealand Jour. Sci. Tech. (A) 21, 114a-125a. Wellington. (In Rev. Appl. Ent., A, 28[1940]:318-14. Johnson, F., and Hammer, A. G. (1912), The grape-berry moth. USDA, Bur. nts BuleeiGs pp ailegle Kaltenbach, J. H. (1874), Die Pflanzenfeinde aus der Klasse der Insekten. Stuttgart. i-viii, 1-848; woodcuts. Kangas, Esko (1945), Eine neue Megastigmus-Art (Hym. Callimomidae) als Samenschadling an Abies sibirica. Ann. Ent. Fenn., 11:177-84. Kaston, B. J. (1939), The native elm bark beetle, Hylurgopinus rufipes (Eich- hoff) in Connecticut. Conn. Agr. Exp. Sta. Bul. 420. 39 pp. Khare, J. L. (1923), Ber (Zizyphus jujuba) fruit and its fly pest. Agr. Res. Institute, Pusa, Bul. 143. Calcutta. 16 pp. (In Rev. Appl. Ent., A, 11[1923]: 439.) Kieffer, H. H., and Jones, L. S. (1933), Some parasites of Anarsia lineatella Zell. in California. Calif. Dept. Agr. Mo. Bul. 22, pp. 387-88. Klein-Krautheim, F. (1937), Uber die Hagebuttenfliege (Rhagoletis alternata Fall.). Zeitschr. f. Angew. Ent., 23:603-14. Laidlaw, W. B. R. (1931), Megastigmus in Scotland; with an addition to the Scottish list. The Scottish Forestry Journal, 45:177-93. Le Conte, J. L., and Horn, G. H. (1876), The Rhynchophora of America north . of Mexico. Proc. Amer. Philos. Soc., 15(96) :1-455. Leonard, M. D. (1926), A list of the insects of New York, with a list of the spiders and certain other allied groups. Cornell Univ. Agr. Exp. Sta. Memoir 101. 1121 pp. Lepage, H. S., and Fadigas, M. (1940), A mariposa oriental das frutas, Grapho- litha molesta (Busck). Biologico, 10(5):135-40. Sao Paulo. Lipman, J. G. (1936), Blueberry fruitworm. Report of the director. Ann. Rept. N. J. Agr. Exp. Sta. 57, pp. 40-41. BIBLIOGRAPHY lar Loew, H. (1862), Die europaischen Bohrfliegen (Trypetidae). Wien. Re- printed, Berlin, 1913. Lovett, A. L. (1915), The rose curculio, Rhynchites bicolor Fab., injures blackberry buds. Rept. Dept. Ent., Oregon Agr. Exp. Sta. (In Biennial Crop Pest and Hort. Rept., 1913-14, pp. 95-202). (This article is wrongly cited in Rev. Appl. Ent., A, 3[1915]:265. For correct reference, see Exp. Sta. Rec., 32[1915]:651.) Lucet, E. (1898), Les Insectes nuisibles aux Rosiers sauvages et cultivés en France. Rouen. Liistner, G. (1931), Die wichtigsten tierischen Shadlinge der Rosen und ihre Bekampfung. Flugblatt Biol. Reichsanst. f. Land- u. Forstw., Nos. 115-16, 8 pp. (In Rev. Appl. Ent., A, 19[1931]:661.) Lugger, Otto (1899), Beetles injurious to fruit-producing plants. Minn. Agr. Exp. Sta. Bul. 66, pp. 85-331. (Also as Fifth Ann. Rept., Entomologist Minn. Agr. Exp. Sta., pp. 1-248.) McConnell, H. S. (1928), Insect investigations. Univ. Md. Agr. Exp. Sta. Bul. 298, pp. 179-80. (1932), Oriental fruit moth parasites in Maryland. Jour. Econ. Ent., 25:367-70. Mackie, D. B. (1944), One year with the oriental fruit moth. Calif. Dept. Agr. Bul. 33, No. 1, pp. 4-17. Marcovitch, S. (1916), The red rose beetle. Circ. 36, Office of the state ento- mologist, St. Anthony Park, Minn. 4 pp., 6 figs. Marlatt, C. L. (1931), Report of the chief of the Bureau of Entomology. USDA. 87 pp. . (1933), Report of the chief of the Bureau of Entomology USDA. 47 pp. Mayr, Gustav (1874), Die europiischen Torymiden biologisch und system- atisch bearbeitet. Verh. d. Zool.-bot. Gesell. Wien, 24:53-142. Meigen, J. W. (1826), Systematische Beschreibung der bekannten europiis- chen zweifliigeligen Insekten. Trypeta, 5:312-52. Melville, Ronald, and Pyke, Magnus (1947), The effect of specific variability and the environment on the vitamin C content of British rose hips. Proc. Linn. Soc. London, 159:5-16; refs. Merritt, J. M. (1933), Native parasites of the oriental fruit moth. Mich. Agr. Exp. Sta. Quarterly Bul. 15, pp. 224-55. . (1933a), Oriental fruit moth parasites in Michigan. Jour. Econ. Ent., 26:788-92. Meyrick, E. (1913), Lepidopterorum Catalogus, Carposinidae. Pars: 13) pp. 1-8. Dr. W. Junk, editor. Grauenhage. Mik, Jos. (1898), Zur Biologie von Rhagoletis cerasi L. nebsts einigen Be- merkungen iiber die Larven und Puparien der Trypetiden und iiber die Fihler der Musciden-Larven. Wien. Ent. Zeitung, 17:279-92. Miller, J. H., and Crisfield, G. F. (1930), The presence of Bracon mellitor Say, a parasite of the cotton boll weevil in Georgia. Jour. Econ. Ent., 23:607-608. Milliron, H. E. (1949), Taxonomic and biological investigations in the genus Megastigmus, with particular reference to the taxonomy of the Nearctic species (Hym., Chalcidoidea, Callimomidae). Amer. Midland Naturalist, 41:257-420; refs. pls. Montgomery, B. E. (1933), Preliminary studies of insect parasites in Indiana. Can. Ent., 65:185-90. 158 INSECTS IN ROSE HIPS . (1935), Notes on the biology and the developmental stages of Glypta rufiscutellaris Cress. (Ichn., Hym.), a larval parasite of the oriental fruit moth. Jour. Econ. Ent., 28:371-76. Morley, Claude (1910), Catalogue of British Hymenoptera of the family Chalcididae. London: Brit. Mus. Nat. Hist. Muesebeck, C. F. W. (1925), A revision of the parasitic wasps of the genus Microbracon occurring in America north of Mexico. Proc. U.S. Nat. Mus., 67:1-85. . (1949), Two new species of Opius from rose hips (Hymenoptera, Braconidae). Can. Ent., 81:254-56. , Krombein, K. V., and Townes, H. K. (1951), Hymenoptera of America north of Mexico. Synoptic Catalog. USDA, Agricultural Monograph No. 2, Wash., D.C., 1420 pp. Muramatsu, S. (1927), Notes on the life histories and habits of four fruit moths in Chosen. I (Japanese). Jour. Agr. Exp. Sta., Chosen, 16:1-37, Suiger, Korea. (In Rev. Appl. Ent., A, 16[1928]:483.) Murtfeldt, Mary E. (1891), Entomological notes for the season of 1890. USDA, Div. Ent., Bul. 23, pp. 45-56. Newcomer, E. J. (1950), Orchard insects of the Pacific Northwest and their control. USDA Circ. 270, pp. 1-63. Niblett, M. (1936), British Trypetidae, additional notes. Ent. Record, 48: 13-14. Nikol’skaya, M. N. (1934), List of chalcid flies (Hym.) reared in U.S.S.R. Bul. Ent. Res., 25:129-43. Ohlendorf, W. (1926), Studies of the pink bollworm in Mexico. USDA Bul. 1374. 64 pp. Osten Sacken, C. R. (1877), Western Diptera. U.S. Geol. and Geogr. Survey Bul. 3, pt. 2, pp. 189-354. Parker, R. L., and Lamerson, P. G. (1934), Hymenopterous parasites of the western apple curculio in northeastern Kansas. Jour. Kan. Ent. Soc. 7(3): 90-95. Pearce, E. K. (1928), Typical flies. A photographic atlas of Diptera. Ser. III. London: Cambridge University Press. 64 pp., 162 figs. Pechuman, L. L. (1937), An annotated list of insects found in the bark and wood of Ulmus americana L. in New York State. Brooklyn Ent. Soc. Bul. 32, pp. 8-21. Pemberton, C. E., and Willard, H. F. (1918), A contribution to the biology of fruitfly parasites in Hawaii. Jour. Agr. Res. 15:419-65; figs., refs. Perris, Eduard (1876), Nouvelles promenades entomologiques. Ann. Soc. Ent. France, 45(ser. 5, no. vr) :171-244. Pettit, R. H. (1927), Report of the section of entomology. Ann. Rept. Mich. St. Bd. Agr. 66, 1926-27, pp. 339-56. Phillips, W. J. (1927), Eurytoma parva (Gir.) and its biology as a parasite of the wheat jointworm, Harmolita tritici (Fitch). Jour. Agr. Res., 34:743- 58. Phillips, V. T. (1946), The biology and identification of trypetid larvae. Amer. Ent. Soc. Memoirs 12. 161 pp.; refs., pls. Pierce, W. D. (1908), Studies of parasites of the cotton boll weevil, USDA, Bur. Ent., Bul. 73, pp. 1-68. . (1913), Miscellaneous contributions to the knowledge of the weevils of the families Attelabidae and Brachyrhinidae. Proc. U.S. Nat. Mus., 45: 370-72. | BIBLIOGRAPHY 159 Porter, B. A., and Garman, P. (1923), The apple and thorn skeletonizer. Conn. Agr. Exp. Sta. Bul. 246, pp. 247-64. Prinz, J. (1919), Biology of the melon fly, Carpomyia caucasica Zaitz. Scien- tific conclusions dept., Tiflis Bot. Gardens, No. 1, 7 pp. (In Rev. Appl. Ent., A, 7[1919]:347-48. Russian. Putnam, W. L. (1935), Notes on the hosts and parasites of some lepidopterous larvae. Can. Ent., 67:105-109. . (1938), Further notes on some alternative hosts of the oriental fruit moth parasite, Glypta rufiscutellaris Cress. Can. Ent., 70:89-90. Puzanova-Malysheva, E. W. (1936), Syntomaspis eurytomae n. sp., eine Chal- cide mit gemischter Ernahrung im Larvenstadium. Zeitschr. f. Angew. Ent. 26:631-42. Quaintance, A. L., and Jenne, E. L. (1912), The plum curculio. USDA, Bur. Ent., Bul. 103. 250 pp. Ratzeburg, J. T. C. (1848), Die Ichneumonen der Forstinsekten. Bd. 2. Berlin. 238 pp. Reichert, A. (1935), Rosenschadlinge. Kranke Pflanze, 12(6):98-99; 1 pl. Dresden. (In Rev. Appl. Ent., A, 23[1935]:506.) Reinhard, H. (1857), Beitrage zur Geschichte und Synonymie der Pteromali- den. Berliner Ent. Zeitschr., 1:70-80. Rice, P. L. (1935), Notes on the ragweed borer (Epiblema strenuana Wk.) and its parasites. Trans. Peninsula Hort. Soc., 49:89-94. Richardson, C. H. (1943), Record of the oriental fruit moth in Iowa, and comments on an apparently new pest in Iowa apple orchards. Trans. Iowa St. Hort. Soc., 78:135-37. Riley, C. V. (1893), Is Megastigmus phytophagic? Proc. Ent. Soc. Wash. 2: 359-64. and Howard, L. O. (1891), Rhynchites bicolor injuring cultivated roses. Insect Life, 4:137. Rimsky-Korsakov, M. N. (1932), in Stachelberg, A. A.: Verzeichnis der schad- lichen Insekten der Palaarktischen Region. Teil I. Schadlinge der Land- wirtschaft. The Lenin Acad. Agr. Sciences in U.S.S.R., Institute for Plant Protection, Ser. 1, Entomology, No. 5, pp. i-xx, 1-500. Robertson, H. A. (1923), The rose curculio in Manitoba, with notes on other insects affecting roses. Ann. Rept. Ent. Soc. Ont., 54, pp. 12-16. Rodendorf, B. B. (1938), Key to the larvae of fruit flies. Moscow, 30 pp. (In Rev. Appl. Ent., A. 27[1939]:439. ) Rodriguez, J. A. (1926), El “Gusano” de las azufaifas (Carpomyia incompleta Becker). [The Jujube Maggot]. Bol. Estacion Pat. veg., 1(4):140-43; 3 figs., 3 refs. Madrid. (In Rev. Appl. Ent., A, 15[1927]:224-25.) Rodzianko, W. N. (1908), Commentatio de Torymidis quarum larvae in sem- nibus pomacearum vitam agunt. Soc. Imp. Nat. Moscow Bul. 21, for year 1907 (n.s. No. 4), pp. 599-600. Rondani, Camillo (1870), Ortalidinae Italicae. Part 7, Fasc. 4. Bol. Soc. Ent. Ital., 2:25-26; (1871) 3:186-88. . (1872), Degli insetti parassitti e delle loro vittime enumerazione con note (Part 3). Bol. Soc. Ent. Ital., 4:45, 53. Rosenberg, H. T. (1934), The biology and distribution in France of the larval parasites of Cydia pomonella L. Bul. Ent. Res., 25:201-56; figs., refs. Riihl, Maria (1914), Liste neuerdings beschriebener oder gezogener Parasiten und ihre Wirt., No. III. Soc. Ent., 29:17-18. 160 INSECTS IN ROSE HIPS Ruhmann, M. H. (1933), Twenty-eighth report, provincial entomologist of British Columbia. Ann. Rept. Dept. Agr. for 1933, pp. y39-y40. Victoria. Salt, G. (1931), Parasites of the wheat-stem sawfly, Cephus pygmaeus Linn. in England. Bul. Ent. Res., 22:479-545. Sanderson, E. D. (1900), Apple bud-borer (Steganoptycha pyricolana Murt- feldt). Rept. of the entomologist. 12th Ann. Rept. Del. Agr. Exp. Sta. for year ending June 30, 1900. pp. 194-99; 11 photes. . (1903), The larva and pupa of the apple bud-borer (Steganoptycha pyricolana Murt.). Can. Ent., 35:158-61. Satterthwait, A. F. (1946), Sunflower seed weevils and their control. Jour. Econ. Ent., 39:787-92. . (1948), Important sunflower insects and their insect enemies. Jour. Econ. Ent., 41:725-31. and Swain, R. B. (1946), The sunflower moth and some of its natural enemies. Jour. Econ. Ent., 39:575-80. Say, Thomas (1830), Descriptions of North American dipterous insects. Jour. Acad. Nat. Sci., Phila., 6:149-78, 183-88. . (1836), Descriptions of new species of North American Hymenoptera and observations on some already described. Jour. Boston Soc. Nat. Hist., 1(3):210-305. Schaffner, J. V., and Griswold, C. L. (1934), Macrolepidoptera and _ their parasites reared from field collections in the northeastern part of the United States. USDA Misc. Publ. 188, 160 pp. Schiner, J. R. (1864), Fauna Austriaca, Die Fliegen (Diptera), nach der analytischen Methode bearbeitet. Zweiter Teil. Wien. 658 pp. Seguy, E. (1934), Faune de France. 28. Dipteres (Brachycera) (Muscidae Acalypterae et Scatophyidae). Paris: Lechevalier. 832 pp., 27 pls. 903 figs. Sharp, David (1889), Il. The Rhyncophorous Coleoptera of Japan. Trans. Ent. Soe. London, 37:41-73. Smith, C. W. (1928), Parasitism of the oriental peach moth in Ontario, with special reference to biclogical control experiments with Trichogramma minutum Riley. Ann. Rept. Ent. Soc. Ont., 59, pp. 72-80. Smith, John B. (1899), Insects of New Jersey, a supplement to the 27th Ann. Rept. St. Bd. Agr. (Catalogue of insects.) Trenton. 755 pp. Stearns, L. A., and Neiswander, R. B. (1930), Oriental fruit moth investiga- tions in Ohio. I. Ohio Agr. Exp. Sta. Bul. 457. 36 pp. Stone, Alan (1951), The Rhagoletis of roses. Proc. Ent. Soc. Wash., 53:45- 48; refs. Strong, L. A. (1934), Report of the chief of the Bureau of Entomology. USDA, 20) pp: . (1935), Report of the chief of the Bureau of Entomology and Plant Quarantine. USDA. 96 pp. Swederus, Nils S. (1795), Beskrifning pa et nytt genus Pteromalus ibland Insecterna, hoerande til Hymenoptera, uti Herr Arch. och Ridd. vy. Linnes Systema Naturae. Svenska Kongl. Vet. Acad. Handl., 16, pp. 201-205, 216- 22: Thompson, W. R. (1945), A catalogue of the parasites and predators of in- sect pests. Belleville, Ont.: The Imperial Parasite Service, 13 vols., 1945-53. Thomson, Carl G. (1875), Skandinaviens Hymenoptera bearbetade af C. G. Thomson, 4e Delen, pp. 1-259, Lund. Tomlinson, W. E. (1951), Control of insect larvae infesting immature blue berry fruit. Jour. Econ. Ent., 44:249-50. BIBLIOGRAPHY 161 Truax, A. L. (1946), More about “Rose Bugs.” North and South Dakota Hor- ticulture, 19( Mar.) :48. Twinn, C. R. (1933), A summary of insect conditions in Canada in 1933. Ann. Rept. Ent. Soc. Ont., 64, pp. 62-80. Underhill, G. W. (1924), The lesser clover weevil. Quart. Bul. Va. State Crop Pest Commission, 6:1-7. . (1935), The pecan tree borer in dogwood. Jour. Econ. Ent., 28:393- 96. Van Steenburgh, W. E. (1934), Parasites of the oriental fruit moth, Laspey- rezia molesta Busck, in Ontario. A summary, 1932-33-34. Ann. Rept. Ent. Soc, Ont., 65, pp. 68-72. Vergeer, Teunis (1954), The cherry fruitworm (Grapholitha packardi) as a blueberry pest in Michigan. Mich. Agr. Exp. Sta. Quart. Bul. 36, pp. 370-73. von Réder, V. (1891). Ueber Orellia Schineri Lw. Entomologische Nachrich- ten, 17(14) :209-10. von Roser, C. L. F. (1834), Verzeichniss der in Wiirtemberg vorkommenden zweifliiglichen Insecten. Correspondenzbl., Landwirtsch. Verein, Wiirtem- berg (Stuttgart), Heft 8, p. 19. . (1840), Verzeichniss der in Wiirtemberg vorkommenden zweifliigli- chen Insecten. Nachtrag, Correspondenzbl., Landwirtsch. Verein, Wirtem- berg, Heft 1, pp. 49-64. Wachtl, Fritz A. (1884), Beitrag zur Kenntniss der Lebensweise der Mega- stigmus collaris Boh. Wien. Ent. Zeitung, 3:38-39. . (1893), Ein neuer Megastigmus als Samenverwiister von Pseudotsuga Douglasii Carr. Wien. Ent. Zeitung, 12:24-28. Same also In Centralbl. f. d. gesam. Forstw., 19( 1893) :6-10. Walker, Francis (1833), Monographie Chalcidum. Ent. Mag., 1:12-29, 115-42. Walkley, Luella M. (1956), A new tersilochine parasite of the rose curculio (Hymenoptera, Ichneumonidae). Ent. News, 67:153-56. Weiss, H. B. (1917), Megastigmus aculeatus Swed., introduced into New Jersey from Japan (Hymen.). Jour. Econ. Ent., 10:448. Willard, H. F. (1927), Parasites of the pink bollworm in Hawaii. USDA Tech. Bul. 19, pp. 1-15. Williams, C. M. (1958), The juvenile hormone. Scientific American, Feb., pp. 67-74. Wilson, G. F. (1931), Insects associated with the seeds of garden plants. Jour. Royal Hort. Soc., 56(Pt. 1):31-47. London. (In Rev. Appl. Ent., A, 19[1931]:246.) . (1935), Additional hosts of Anomoea antica Wied. and Spilographa alternata Fall. Ent. Mo. Mag., 71:58-60. Wingo, C. W. (1940), The importance and use of parasites in the control of oriental fruit worm. Proc. Missouri Acad. Sci., 5:122-24. Yano, M., and Koyama, M. (1918), On wasps parasitizing the seeds of conif- erous trees. Report of Forest Experiments, Forestry Bureau, Tokyo. No. 17, pp. 35-58. (In Rev. Appl. Ent., A, 6[1918]:402-403. ) Zacher, F. (1919), Bericht iiber die Tatigkeit der biologischen Reichsanstalt fiir Land- und Forstwirtschaft in den Jahren 1916, 1917 und 1918. II. Wis- senschaftliche Untersuchungen. Art. 9. Beobachtungen iiber einige schad- liche und niitzliche Insekten. Heft 17, pp. 28-29, Berlin. (In Rev. Appl. Ent., A, 8[1920]:288.) Zeller, P. C. (1849), Verzeichnis der von Herrn Jos. Mann beobachteten 162 INSECTS IN ROSE HIPS toscanischen Microlepidoptera (Fortsetzung). Stett. Ent. Zeitung, 10:231- 56. (1871), Lepidopterologische Beobachtungen im Jahre 1870. Stett. Ent. Zeitung, 32:57-60. (1876), Beitrage zur Kenntniss der nordamerikanischen Nachtfalter, besonders der Microlepidopteren. Verh. Zool.-Bot. Gesell. Wien, 25:207-360. Zeller, S. M., and Schuh, J. (1944), Diseases and insect pests of cane fruits in Oregon. Ore. Agr. Exp. Sta. Bul. 418, pp. 50-51. Puate I. Rhagoletis basiola ]. Adult. 2. Egg. 8. Right wing. 164 165 PuaTE II. LarvaL Structures oF Rhagoletis basiola Posterior respiratory apparatus of first instar: a, external dorsocaudal view of spiracular plate; b, atrium; c, trachea. Same of second instar. Posterior spiracular plate of second instar. Posterior spiracular plate of third instar visible within advanced second instar. Mouth hooks of first instar, with connecting structure broken apart. Anterior respiratory “fan” of second instar. Mouth hook of second instar, lateral view, also showing food channels, i.e., “stomal sieves.” Cephalic part of oral armature of first instar, in profile: a, hypostomium. Same of second instar: a, hypostomium; b, parastomium; c, subhyposto- mium. Same of third instar: a, b, and c, as in Fig. 12. 166 167 14. 15. 16. thes 18. PLATE III. Opius rosicola ann O. baldufi Adult female of O. rosicola. Right wings of O. baldufi. Egg of Opius rosicola or baldufi. First instar of Opius rosicola or baldufi, dorsolateral view. Second instar of Opius rosicola or balduf. 168 169 Puiate IV. Halticoptera rosae 19. Adult. 20 and 21. “Brown bodies,” or dead eggs. Dig Beran. 170 171 Puate V. Halticoptera rosae 23. Early phase of first instar. 24. Intermediate phase of same. 25. Advanced phase of same. 26. Second instar. 27. Mature larva. Puate VI. Glypta rufiscutellaris 28. Adult female of Glypta rufiscutellaris in profile. 174 7) PLATE VII. Glypta rufiscutellaris. Adult male of Glypta rufiscutellaris. First instar of same, lateral view. Second instar of same, lateral view. Egg of Glypta rufiscutellaris. 176 N77 33. 34. 35. 36. 37. 38. Puate VIII. Rhynchites bicolor Adult of Rhynchites bicolor, lateral view. Egg of same, dorsal view. Egg of same, lateral view. Mature larva of same, lateral view. Pupa of same, ventral view. Pupa of same, dorsal view. Lio, ee i 5, 179 PLATE IX. Luchatema balduft 39. Adult female of Luchatema baldufi, in profile. 180 40. 4l. 42. 43. 44. 45. Puate X. Luchatema balduft Early phase of first instar, lateral view. Advanced phase of same. Second instar. Mature larva. Egg. Cocoon, showing position of larval excrement (a). 183 46. AT. 48. 49. 50. Dl. Puate XI. Parasites First instar of Ascogaster sp. from Cydia packardi. First instar of eupelmid? from Rhag. basiola, dorsolateral view. Supposed second instar of eupelmid? from Rhag. basiola, dorsal view. Full-grown larva of Zatropis rosaecolis, on Rhynchites bicolor. First instar of “long-tailed” parasite from Cydia packardi. Egg of a parasite, presumably on Cydia packardi. 184 185 PLATE XII. Megastigmus 52. Female of Megastigmus nigrovariegatus Ashmead. The American rose seed chalcid. 53. Female of Megastigmus aculeatus (Swed.). European rose seed chalcid. 186 187 INDEX Achene. See under Rose hips Aegeria scitula, 115 American rose seed chalcid, 130-45 Anarsia lineatella, 74, 115 Ancylis comptana, 65 Anthonomus albopilosus, 114 Anthonomus eugenii, 114 Anthonomus fulvus, 114 Anthonomus grandis, 114 Anthonomus signatus, 114 Anthonomus squamosus, 114 Anthophila (Hemerophila) pariana, 74 Apple and thorn skeletonizer, 73-74 Apple bud borer, 53 Argyrotoxa semipurpurana, 65 Ascogaster carpocapsae, 64, 70 Ascogaster quadridentatus, 64, 70, 71 Ascogaster sp.?: evidence of parasit- ism in Cydia packardi, 70; first in- star described, 70-71; diapause, probable life cycle, 71 Ascorbic acid in rose hips, 9 Attelabus, 79, 98 Bagworm, common, 74 “Bassus” sp., 64 Beauveria globulifera, fungus, 105 Bionomics of hip insects: summary— phytophags, parasites, predators, scavengers, 146-49 Bracon anthonomi, 114 Bracon mellitor: facultative in rose hips, 3; parasite of Cydia packardi, 64; synonymy, distribution, 114; published host records — Coleop- tera, Lepidoptera, 114-15; parasite of Rhynchites bicolor, 114; evi- dence of attack on Rhyn. bicolor, 116; extent of parasitism, 116; adult —described, factors affecting size, mating, sporadic parthenogenesis, 116-17; oviposition and _ paralysis, egg described, 117-18; circum- stances affecting parasitization, 118; 189 larva—hatching, locomotion, de- scription, 118; cocoon and pupa, 118-19; life cycle, duration of stages, 119 Bracon pembertoni, 114 Bracon spp., facultative in rose hips, 3 Bracon variabilis: distribution and hosts, 119-20; evidence of attack on Rhynchites or Cydia, 120 Bracon vernoniae, 114 Bracon xanthostigma, 114 Caloglyphus spinitarsus, 123 Carpocapsa pomonella, 66 Carpomyia Schineri; distribution, rose hosts, larval life, activities, 6-7 Carposina adreptella, 77 Carposina sasakii, 77 Carposina_scirhosella: known occur- rence, habits of generic relatives, ele Chaetogaster sp., Annelida; discovery of, evidence of parasitic life, source, Chalcids (rose seed), 123-45. See also under Megastigmus Chalcodermus aeneus, 115 Cherry fruitworm, 1, 53 Choristoneura (Archips) fumiferana, 74 Clerid larvae, predators, 75-77 Coeliodes inequalis, 115 Coleophora pruniella, 73 Conotrachelus nenuphar, 115 Crocidosema lantanae, 115 Cryptophlebia illepida, 115 Curculio, 78 Cydia molesta, 64, 66, 68, 69, 74, 115 Cydia nigricana, 70 Cydia packardi: discovery of, 1; facul- tative resident, 3; synonymy, 53; common names, 53; records of oc- currence, 53-54; known rose hosts, 54; adult and egg described, 54; 190 oviposition, 54; first and last instars described, 55; gonads in male lar- vae, 55; food and niches of larva, 55; habits and damage in apples, blueberries, cherries, rose buds and rose hips, 56-60; larva as potential pollinator in rose buds, 58; oral ex- cretions of larva, 60; hibernation, 60-61; pupation, 61; cycles per year described, 61-64; parasites of C. packardi, 64-75 (see also under Parasites); predators, 75-77 (see also under Predators) Cydia pomonella, 70 Cydia roseticolana: synonymy and lit- erature, 51; known occurrence, 51; rose hosts and larval life, 52; co- coon, pupal activity, 52-53; season of adult, 53; one cycle per year indicated, 53 Cylas formicarius elegantulus, 115 Desmoris constrictus, 115 Desmoris scapalis, 114 Diapause: in Rhagoletis basiola, 27- 29: Opius rosicola, O. baldufi, in first and final instars, 36, 38; Halti- coptera rosae, 46; Eupelmidae spp., 51; Ascogaster sp.?, 71; long-tailed parasite, 72; Rhynchites bicolor bi- color — ordinary and prolonged types of, 93-94; Rhyn. b. wickhami, 103; Luchatema baldufi, 109, 1138; Megastigmus aculeatus, 129; M. nigrovariegatus, 141, 143 Diapauses of parasite terminated by host hormone? In Opius spp., 36, 38; Halticoptera, 46; Eupelmidae spp., 51; Luchatema baldufi, 110 Disholcaspis mamma, 74 Ectoparasite (unidentified) of Cydia packardi, 75 Emergence from puparia, rhythmic in Rhagoletis basiola, Pegomyia betae, 14-16 Enoclerus? sp.: mature larva de- scribed, 75-76; records from rose hips in Illinois, 75 Epiblema obfuscana, 65 Epiblema otiosana, 65 INSECTS IN ROSE HIPS Epiblema scudderiana, 65 Epiblema strenuana, 65, 69 Epinotia laracana, 65 Epiurus indagator, 73, 74 Epiurus nigrifrons, 74 Euderus cushmani: facultative in rose hips, 3; nature of parasitism and hosts attacked, 74; adult described, 74; host records (from literature and rose hips), parasitism on Cydia packardi, 75 Eupelmidae spp.: parasites in Rhago- letis basiola, 50; first and second in- stars described, 50; host stages at- tacked, 50; adults from hip jars, 51 Eupelmus sp.: as hyperparasite of G. rufiscutellaris, 64; in achenes in- habited by Megastigmus nigrovarie- gatus, 145; uncertain host relations of others from hip jars, 145 European rose seed chalcid, 1, 123-30 Eurytoma sp.: as hyperparasite of G. rufiscutellaris, 64; characters of the genus, 121; sources of adults ob- tained, 121-22; adults reared from Rhynchites bicolor, 122 Eurytoma tylodermatis, 64 Facultative insects of rose hips, 2, 3 Frumenta nundinella, 115 Glypta rufiscutellaris: facultative in rose hips or buds, 3; distribution and hosts, 64-66; adult described, 66; sporadic parthenogenesis, 66; oviposition, 67; reaction of hosts to attack, 67-68; larval life—instars of host preferred, four instars de- scribed, behavior in host, emer- gence, 68; cocoon, prepupa, pupa, described, 69; superparasitism, 69; life cycles — coordinate with host generations, seasonal distribution of stages, effect of temperatures on, 69-70 Gnorimoschema gallaesolidaginis, 65 Gnorimoschema operculella, 65 Grapholitha prunivora, 65 Halisidota tesselaris, 74 Halticoptera rosae: obligatory in rose INDEX hips, 3; type specimens and sex ra- tio, 41; hosts and known occur- rence, 41-42; adult, egg, larva de- scribed, 42-43; emergence and movements of adult, 43; mating and oviposition, 43-44; egg stage, 44; “brown bodies” (dead eggs), 44; larva — solitary and multiple parasitism, 45; life of first instar— slow growth and diapause, attack by phagocytes, 45-47; milky white first instars, 47; advanced larvae— rapid growth and why, 47-48; con- traction waves observed, 48; pupa, 48; life cycle, 48-49; life cycles of Halticoptera, Opius, and Rhago- letis compared, 49-50 OW tion, 105-106; adult—description, activity, 106; oviposition, 106-107; host instars attacked, 107; repro- ductive potential, 107-108; first in- star — recognition, growth, food, movements, feeding and growing phases, diapause, cause of slow growth, 108-10; advanced larvae— rate of growth, instars involved, structural characteristics, 110-11; cocoons—size, form, structure, loca- tion, 111; prepupa and pupa, 112; parasitism—variations in rate of, 112; superparasitism, phagocytes, 112-18; life cycle, seasons of life stages, 113-14 Heliothis obsoleta, 115 Macrocentrus ancylivorus, 66 Hemerocampa leucostigma, 74 Megastigmus aculeatus: synonymy, Homeosoma electellum, 115 Hypanthium. See under Rose hips Hypera nigrirostris, 115 Key to instars of Rhagoletis basiola, 13 Laspeyresia caryana, 65 Lesser apple worm, 65 Life cycles: Rhagoletis alternata, 9- 11; Rhag. basiola, 29-32; Opius rosicola, O. baldufi, 39-40; Halti- 124; controversy over food habit— “phytophagic” or parasitic, 124-25; distribution and rose hosts—of light form (M. a. aculeatus), dark form (M. a. nigroflavus), 126; form not identified, 127; spread from Europe to Asia, North America, 127-28; sex ratio, parthenogenesis, 128; life cycle, seasons of life stages, 129; diapause, 129; economic import- ance, 129-30 coptera rosae, 48-49; Glypta rufi- Megastigmus nigrovariegatus: obliga- scutellaris, 69-70; Ascogaster sp.?, 71; long-tailed parasite, 72; Rhyn- chites bicolor bicolor, 95-96; Rhyn. b. wickhami and _ undetermined western varieties, 103-105; Lucha- tema baldufi, 113-14; Bracon melli- tor, 119; Megastigmus aculeatus, 129; M. nigrovariegatus, 141-44 Lonchaea polita: adult described, lo- cation of puparia, sources of series from rose hips, beliefs on food of larva, 78 Long-tailed parasite (in Cydia pack- ardi): first instar, cycle and dia- pause, 72 Luchatema baldufi: obligatory in rose hips, 3; occurrence in northeastern Minnesota, 80; inhibits pupation of host, 93; destroys mature host, 96; sources of type series, 105; distribu- tory in rose hips, 3; type localities and synonyms, 130; rose hosts and known occurrence of the chalcid, 130-32; summary of distribution, 132; variation in color and size of adult, 132; concurrence of nigro- variegatus and M. aculeatus, 133; emergence of adult from hip, 133; sex ratio, protandry, 133-34; mat- ing, 134-36; oviposition—age of fe- male, state of seeds, process, monopoly of hips, influence of tem- perature, 136-38; nidi of embryo, egg described, reproductive capa- city of female, 138-39; larval life —five instars recognized, develop- ment in seed, full-grown vs. mature larvae, 139-40; prepupation and pupation, 140-41; life cycle, seasons of life stages, 141-42: laboratory 192 tests on adult survivability, 142-43; diapause of mature larva, 143-44; precocious pupation, causes of, 144- 45; acceleration of cycle under arti- ficial conditions, 145 Megastigmus spp.: obligatory in rose hips, 3; described, compared, 123- 24 Melissopus latiferreanus, 65 Merhynchites bicolor: varieties and occurrence in North America, 79 (see also Rhynchites) Methods of study, 3 Microbracon, 114 Mompha stellata, 115 Myelois ceratoniae, 115 INSECTS IN Objectives of the study, 2 Obligatory insects of rose hips, 2, 3 Opius balduft, O. rosicola: obligatory in rose hips, 3; parasites of Rhago- letis basiola, 32; hosts and distribu- tion, 32-33; recognition of species, 33; sex ratio, 33; movements of adult, 33; egg, “brown bodies,” ovi- position, 33-34; first instar—de- scribed, food and growth, diapause, 34-36; advanced larvae, 36-38; sec- ond diapause, in full-grown larvae, 38; precocious and delayed larval development, 38-39; pupa _ de- scribed, 39; life cycle, seasonal dis- tribution of stages, 39-40 Opius spp., obligatory in rose hips, 3 Oriental fruit moth, 64, 66, 74 Parasites of Cydia packardi: Glypta rufiscutellaris, 64-70; Ascogaster sp.?, 70-71; long-tailed parasite, 72: Scambus hispae, 72-74; Euderus cushmani, 74-75; unidentified para- site, 75 Parasites of Megastigmus nigrovarie- gatus: Eupelmus sp., 145 Parasites of Rhagoletis basiola: Opius rosicola, O. baldufi, 32-40; Halti- coptera rosae, 41-50; Eupelmidae spp., 50-51 Parasites of Rhynchites bicolor: Te- melucha, 105; Luchatema_ baldufi, 105-14; Bracon mellitor, 114-19; B. ROSE HIPS variabilis, 119-20; Zatropis rosae- colis, 120-21; Eurytoma sp., 121-22; Chaetogaster sp., 122-23 Parasitic Hymenoptera: species from hip jars only, list of, 145-46 Pectinophora gossypiella, 115 Pegomyia betae, rhythmic emergence from puparium, 16 Phyllobaenus sp.: mature larva de- scribed, records from rose hips in Illinois, 76 Phytodietus fumiferanae, 74 Phytophags. See American rose seed chalcid, European rose seed chalcid, Rose curculio, Rose hip flies, Rose hip moths Pimpla indagatrix, 7 Pine tip moth, 73 Polychrosis carduana, 65 Polychrosis viteana, 74, 115 Predators of Cydia packardi: Enocle- rus? sp., Phyllobaenus sp., 75-77; evidences of attack on packardi, not on Rhagoletis basiola, 77 Psychophagus omnivorus, 64 Pyrausta nubilalis, 115 Pyrausta penitalis, 115 Rhagoletis alternata: common names, 7; distribution and hosts, 7, 8; food of larva, 8, 9; damage to rose hips, 9; life cycle and habits, 9-11 Rhagoletis basiola: discovery of, 1; obligatory in rose hips, 3; distribu- tion, 11; rose hosts, 11-12; adult, egg, instars, puparium described, 12-14; expansion of wings, 14; peri- odic emergence from puparia, 14- 15; adult behavior—reaction to light, feeding, preening, copulation, 16-18; oviposition, 18-21; phases of embryonic growth, 21; deficient eggs, 21-22; rose species in relation to larval burrowing, feeding and survival, 22-24; emergence of lar- vae, 24; populations, rose species, and latitude, 24-25; pupariation, 25; viscera of prepupa, 26; gas in pre- pupa, 26-27; pupation, 27; pupal diapause, 27-29; pupa, heart beats, spring development of, 28; life cy- INDEX cle, seasonal distribution of stages, 29-32; hymenopterous parasites of, 32-51 (see also under Parasites ) Rhagoletis spp., obligatory in rose hips, 3 Rhyacionia frustrana, 73, 115 Rhynchites bicolor: discovery of, 1; obligatory in rose buds or hips, 3; common names, 78; taxonomy and varieties, 79, 96-97; distribution, 79- 81, 96-98: eastern hosts, 81; west- ern hosts, 97-98; niches of larvae varied, 97-98; published error cor- rected, 98; rosaceous plants eaten, damaged by adult, 98-100; oviposi- tion and larval life, 100-102; Rubus possibly a host, 102; behavior of mature larva, 102; pupation, 103; life cycle, 103-105 Rhynchites bicolor bicolor (eastern variety): host plants, 81; food plants and damage, 81-83; adult described, 83-84; emergence from ground, locomotion, 84; copulation, 84; oviposition, 85-87; egg de- scribed, 87; larva—described, its niches in hips, food, movements, mortality, 87-89; larval emergence from hips, influence of temperatures on, 89-91; weights of mature larvae, 91-92: gas in digestive tract of larva, 92; larval life in ground, cell structure, 92; effects of pot over- stocking, 93; larval diapause—or- dinary and prolonged, 93-94; pre- pupa and pupa, described, 94-95; life cycle, seasonal distribution of life stages, 95-96; natural enemies, 105-23 (see also under Parasites ) Rhynchites bicolor cockerelli, 79, 96 Rhynchites bicolor erythrosoma, 97 Rhynchites bicolor piceus, 79, 96 Rhynchites bicolor ventralis, 79, 96 Rhynchites bicolor viridilustrana, 79, 96 Rhynchites bicolor wickhami (western variety): taxonomy, 79; distribu- tion, 96-97, 98; adult, egg, larva, pupa described, 97; hosts, 97-98; larvae in mummified flower buds, 97; adult—food, damage, 98-100; 193 mating, fighting and locomotion, 100; oviposition, larval life, 100- 102; behavior of mature larva, 102; pupation, 103; emergence of adult from ground, 103; life cycle, sea- sons and duration of stages, 103- 105; Temelucha, parasite in, 105 Rhythmic emergence of Rhagoletis and Pegomyia, 14-16 Rose curculio, 78-105 Rose hip flies, 6-32. See Carpomyia Schineri, Rhagoletis alternata, Rhagoletis basiola Rose hip moths, 51-64. See Cydia roseticolana, Cydia packardi, Car- posina scirhosella Rose hips, achene cavity of: utilized by Rhynchites bicolor, 87-88; by some western forms of, 100-101 Rose hips, achenes of: utilized by Megastigmus aculeatus, 126-27; M. nigrovariegatus, 139-40 Rose hips, hypanthium of: utilized by Carpomyia Schineri, 7; Rhagoletis alternata, 8-9; Rhag. basiola, 22- 24: Cydia roseticolana, 52; C. pack- ardi, 55, 59; Carposina scirhosella, rai Rose hosts of hip insects, listed, 150 Rose seed chalcids, 123-45 Rose snout beetle, 78-105 Scambus hispae: facultative in rose hips, 3; adult described, 72-73; syn- onyms, distribution, habits of hosts utilized, 73; evidence of attack on Cydia packardi, 73; sources of adults reared, 73; host records from literature, 73-74 Scavengerous fly (Lonchaea polita) in rose hips: description, records, 78 Scavengerous mite (Caloglyphus spi- nitarsus) in pupal cells of Rhyn- chites bicolor, 123 Secodella cushmani, 74 Spruce budworm, 74 Stibadium spumosum, 115 Strawberry leaf roller, 65 Tables, 15, 90, 133 Tachypterellus quadrigibbus, 115 194 INSECTS IN ROSE HIPS Temelucha sp., 105 Zatropis rosaecolis: occurrence, distri- Thyridopteryx ephemeraeformis, 74 bution, rose hosts, 80, 120; sources of type specimens, 120; adult de- scribed, 120; evidence of parasitism Vitamin C in rose hips, 9 on Rhynchites, 120-21; habits, 121 The Library of Congress has cataloged this publication as follows: BaLpuF, WALTER VALENTINE, 1889— Obligatory and facultative insects in rose hips: their recognition and bionomics. Urbana, University of Illinois Press, 1959. 194 p. illus, 26 cm. (Illinois biological monographs, no. 26). 1. Insects—Biology. 2. Rose hips. 3. Insects—U. S. 1. Title. QL509.B3 595.7 58-6993 Library of Congress ee | HOFFMEISTER, DONALD F, and GOODPAS- _ TURE, wooprow w. (1954): The Mam- mals of the Huachuca Mountains, South- astern Souci 27 figs. Vol. 24, No. 1. $3.00. _ HOPKINS, SEWELL HEPBURN (1934): The 4 Papillose Allocreadiidae—A Study of Their £ fi Morphology, Life Histories, and Relation- _ ships. A pls. 6 figs. Vol. 18, No. 2. $1.00. | HUMES, ARTHUR GROVER (1942): The _ Morphology, Taxonomy, and Bionomics of _ the Nemertean Genus Carcinonemertes. 4 _ pls. 1 map. Vol. 18, No. 4. $1.50. _ HUNTER, GEORGE WILLIAM, 111 (1930): Studies on the Caryophyllaeidae of North _ America. 7 pls. Vol. 11, No. 4. $1.25. __ JONES, GEORGE NEVILLE (1946): American _ Species of Amelanchier. 23 pls. 14 maps. — Vol. 20, No. 2. Cloth only, $2.00. KAMM, MINNIE WATSON (1922): Studies _ on Gregarines II—Synopsis of the Polycys- _ tid Gregarines of the World, Excluding _ Those from the Myriapoda, Orthoptera, _and Coleoptera. 4 pls. Vol. 7, No. 1. $1.00. _ KENDEIGH, S. CHARLES (1941): Territorial _ and Mating Behavior of the House Wren. $2 figs. Vol. 18, No. 3. $1.50. (1952): Parental Care and Its Evo- | lution in Birds. 35 figs. Vol. 22, Nos. 1-3. $4.00. _ KINDRED, JAMES E. (1919): The Skull of _ Amiurus. 8 pls. Vol. 5, No. 1. $1.25. KRAMER, sOL (1950): The Morphology and Phylogeny of Auchenorhynchous Homoptera (Insecta). 6 charts. 16 pls. Vol. 20, No. 4. $2.00. KUDO, RICHARD ROKSABRO (1934): Studies on Some Protozoan Parasites of Fishes of - Milinois. 8 pls. Vol. 18, No. 1. $.75. (1944): Morphology and Develop- - ment of Nosema notabilis Kudo; Parasitic _in Sphaerospora polymorpha Davis, a Para- _ site of Opsanus tau and O. beta. 12 pls. 7 _ figs. Vol. 20, No. 1. $1.25. ‘MACKIN, JOHN GILMAN (19386): Studies on _ the Morphology and Life History of Nem- atodes in the Genus Spironoura. 6 pls. 2 figs. Vol. 14, No. 3. $1.00. MANTER, HAROLD WINFRED (1926): Some North American Fish Trematodes. 6 pls. 2 charts. 1 fig. Vol. 10, No. 2. $1.50. + y+ MAY, HENRY GusTAV (1920): Contribu- tions to the Life Histories of Gordius ro- bustus Leidy and Paragordius varius (Leidy). 21 pls. Vol. 5, No. 2. $1.50. MEGLITSCH, PAUL A. (1940): Cytological Observations on Endamoeba blattae. 8 pls. Vol. 17, No. 4. $1.00. MILLER, EDWIN LYNN (1936): Studies on North American Cercariae. 8 pls. Vol. 14, No. 2. $1.50. MIZELLE, JOHN DARY (1938): Compara- tive Studies on Trematodes (Gyrodacty- loidea) from the Gills of North American Fresh-Water Fishes. 5 pls. Vol. 17, No. 1. $1.00. NICHOLS, RAY JANNEY (1937): Taxonomic Studies on the Mouth Parts of Larval Anura. 3 figs. 8 charts. Vol. 15, No. 4. $1.00. RAY, JAMES DAVIS, jR. (1956): The Genus Lysimachia in the New World. 20 pls. 11 maps. Vol. 24, Nos. 3-4. $2.50. ROBACK, SELWYN S. (1954): The Evolution - and Taxonomy of the Sarcophaginae (Dip- tera, Sarcophagidae). 9 charts. 34 pls. Vol. 28, Nos. 8-4. $4.00. ROSS, HERBERT H. (1931): Sawflies of the — Sub-Family Dolerinae of America North of Mexico. 6 pls. Vol. 12, No. 3. $1.25. (1987): A Generic Classification of the Nearctic Sawflies (Hymenoptera, Symphyta). 17 pls. Vol. 15, No. 2. $2.00. SCHOOF, HERBERT FREDERICK (1942): The Genus Conotrachelus Dejean (Coleoptera, Curculionidae) in the North Central United States. 9 pls. Vol. 19, No. 3. $1.50. SCHOPF, JAMES MORTON (1943): The Em- bryology of Larix. 6 pls. 86 figs. Vol. 19, No. 4. $1.50. SMITH, FRANK (1923): The Calciferous Glands of Lumbricidae and Diplocardia. 12 pls. Vol. 9, No. 1. $1.25. SOLHEIM, WILHELM GERHARD (1930): Morphological Studies of the Genus Cercospora. 4 pls. Vol. 12, No. 1. $1.00. SPOONER, CHARLES STOCKMAN (1938): The Phylogeny of the Hemiptera, Based on a Study of the Head Capsule. 24 pls. 2 figs. Vol. 16, No. 3. $1.00. SPRAGUE, vicToR (1941): Studies on Gregarina blattarum with Particular Ref- erence to the Chromosome Cycle. 6 pls. 2 figs. Vol. 18, No. 2. $.75. STANNARD, LEWIs J., JR. (1957): The Phy- logeny and Classification of the North American Genera of the Suborder Tubu- lifera (Thysanoptera). 14 pls. No. 25. $2.50. STEVENS, FRANK LINCOLN (1924): Para- sitic Fungi from British Guiana and Trini- dad. 19 pls. 1 map. Vol. 8, No. 3, $1.25. (1927): Fungi from Costa Rica and Panama. 18 pls. 1 map. Vol. 11, No. 2. $1.25. and RYAN, SISTER MARY HILAIRE (1939): The Microthyriaceae. Vol. 17, No. 2. $1.50. STICKNEY, FENNER SATTERTHWAITE (1928): The Head-Capsule of Coleoptera. 26 pls. Vol. 8, No. 1. $2.00. TEHON, LEO roy (1985): A Monographic Rearrangement of Lophodermium. 5 pls. Vol. 13, No. 4. $2.00. VAN CLEAVE, HARLEY J. (1953): Acantho- cephala of North American Mammals. 13 — pls. 10 figs. Vol. 23, Nos. 1-2. $4.00. VAN DEVENTER, WILLIAM CARL (1987): Studies on the Biology of the Crayfish © Cambarus propinquus Girard. 46 graphs. Vol. 15, No. 3, $1.00. VESTAL, ARTHUR G. (1949): Minimum Areas for Different Vegetations; Their De- — termination from Species-Area Curves. 6 — figs. Vol. 20, No. 3. $2.00. WATSON, MINNIE ELIZABETH (1916): | Studies on Gregarines I—Including De- scriptions of Twenty-One New Species — and a Synopsis of the Eugregarine Rec- ords from the Myriapoda, Coleoptera, and Orthoptera of the World. 15 pls. Vol. 2, No. 8. $2.00. YUNCKER, TRUMAN GEORGE (1921): Re- vision of the North American and West — Indian Species of Cuscuta. 18 pls. Vol. 6, — Nos. 2-3. $2.00. é | . 4 th¢ A: : Ad et ae thi a HAL , "ut Whi ay Gt be ebay 44 D ir A Pied OO I eh Halen Ma bt Peck. Me | ; ins ea i, ol x an 1] Hit NY | f i / i Vinay “ik i i } af i, ; i aM ater in| : iy ay ¢ al ; 1 an 7, 4 ; a i Liatys * K ALF } : } ih We Ie he : panty V } i iy, Ad} i } ape Pia we en | ie ny, iL Vv } Sey et TAI 7, Woe e we at Sey Sy ik as eee sae We SONS