' si Oty ‘ ‘ x : 4 ; ‘ , a te peesereen det the » rs miles tae, uterhem,™ ere Pe yarns alee a eee Reels eee eerie prea gseser tell sess aaneat peereer tae any vee poe pees ae aaase gel Pers ’ ate ee oe : Bette sneha me aya : cisce het ‘ : UNIVERSITY OF ILLINOIS LiGF®ARY AT URBANA-CHAMPAIGN BIOLOGY Jan 19 1994 ian a es Digitized by the Internet Archive In 2011 with funding from University of Illinois Urbana-Champaign http://www.archive.org/details/develoomentaleco53redb ILLINOIS BIOLOGICAL MONOGRAPHS . 27>) = Aa ny 4 EH ’ py . ge er \ F nD Nee teow v a4 Sue +a i eae hg The De of Man (Neuro] The person charging this material is re- sponsible for its return to the library from which it was withdrawn on or before the Latest Date stamped below.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. To renew call Telephone Center, 333-8400 UNIVERSITY OF ILLINOIS LIBRARY AT URBANA-CHAMPAIGN L161—O-1096 wpa UF iit MAY 1 2 1985 uve” ss IMgIS The Developmental Ecology of Mantispa uhleri Banks (Neuroptera: Mantispidae) KURT E. REDBORG and ELLIS G. MACLEOD ILLINOIS BIOLOGICAL MONOGRAPHS 53 UNIVERSITY OF ILLINOIS PRESS Urbana and Chicago ILLINOIS BIOLOGICAL MONOGRAPHS Volumes | through 24 contained four issues each. Beginning with number 25 (issued in 1957), each publication is numbered consecutively. Standing orders are accepted for forthcoming numbers. The ttles listed below are still in print. They may be purchased from the University of Illinois Press, 54 East Gregory Drive, Champaign, Illinois 61820. Out-of-print titles in the Illinois Biological Monographs are available from University Microfilms, Inc., 300 North Zeeb Road, Ann Arbor, Michigan 48106. KOCH, STEPHEN D. (1974): The Evagrostis-pectinacea-pilosa Complex in North and Central America (Gramineae: Eragrostoideae). 86 pp. 14 figs. 8 plates. No. 48. $7.95. KENDEIGH, S. CHARLES (1979): Invertebrate Populations of the Deciduous Forest: Fluctuations and Relations to Weather. 110 pp. Illus. Tables. No. 50. $12.50. LEVINE, NORMAN D., and VIRGINIA IVENS (1981): The Coccidian Parasites (Protozoa, Apicomplexa) of Carnivores. 205 pp. Illus. Glossary. Index. No. 51. $15.95. Board of Editors: M. R. Lee, Michael Lynch, Kenneth Robertson, David Young, Gilbert P. Waldbauer This monograph is a contribution from the Department of Entomology, University of Illinois, originally submitted in partial fulfillment for the first author’s degree of Doctor of Philosophy. First submitted to the Illinois Biological Monograph Board for publication June 1979. © 1985 by the Board of Trustees of the University of Illinois Manufactured in the United States of America Library of Congress Cataloging in Publications Data Redborg, Kurt E. (Kurt Eric), 1949- The developmental ecology of Mantispa uhleri Banks (Neuroptera: Mantispidae) (Illinois biological monographs; 53) Bibliography: p. Includes index. 1. Mantispa uhleri—Development. 2. Mantispa uhleri—Ecology. 3. Insects—Development. 4. Insects—Ecology. I. MacLeod, Ellis G. II. Title. III. Series. QL513.M3R43 1984 595.7747 84-32 ISBN 0-252-01085-X (alk. paper) Acknowledgments We thank Dr. Stanley Friedman and Dr. Joseph A. Beatty for extensive criticism of the original version of this paper, with special thanks to the latter for his invaluable help with the identification of spiders. We also wish to thank the Illinois Biological Monograph Editorial Committee and two anonymous reviewers for subsequent comments on the manuscript. Dr. Jerome S. Rovner provided timely advice regarding the successful rearing of spiders. Evaluation of the data and statistical analyses were completed with the help of Dr. Arthur Ghent and Dr. Richard B. Selander. To Alice Prickett we extend appreciation for her excellent line drawings of the immature stages of Mantispa uhleri. Some of the work reported here was conducted at the Univer- sity of Illinois Dixon Springs Agricultural Center. We express our thanks to the director, Dr. C. J. Kaiser, and staff of this institution. Finally, and most important, the first author wishes to thank his wife Annemarie for valuable discussions and moral support, and to express appreciation to Rachel Anderson of the University of Illinois Press for her painstaking preparation of the manu- script. Without their efforts this work could not have been published. a) ea ane i ‘ Vad Tee BF a Al ear ae er oT an ial oe Siiovmaara het! 1 4 ‘h aw v7? te z= ur ve! Lar, iw 7 a’ wee tos YT | =e y a (ot OOM A ire} | TT ORE ‘} Te. ae at} a a“ Lty, ih ey 4 / rs WEVORACT Vas ve 5 ' 4 °F . ’ F i" aif ’ 4 ‘ 4 ] Y a LAL } ! ’ ’ » a + ra id j ; tj | yi = i ade A y ri vee iy trl j i i MOAN a Le on T ’ | A ahi + 7 " i H moe inde ans 7 ' ( vist i a useytav el eR ight! ak oa iA ot Wee + novel lee ‘ rd .{ a 42 U9 ChPr REET ee wis iA js cc} Pac’ iv f Contents 1. Review of the Literature 2. Laboratory Culture of Mantispids Maintenance of Adults Larval Rearing-Chambers Spider Eggs Rearing and Mating Conditions Measurements EXPERIMENT 1: Adult Size Variation Results Oviposition Egg First Instar Larva Subsequent Development of the Larva Second and Third Instar Larvae Cocoon Pupa Adult Mating Behavior Discussion Eggs Adult Size Variation Possible Male Pheromone 3. Larval Strategies for Locating Spider-Egg Prey Methods and Results EXPERIMENT 2: Reaction of M. viridis and M. uhleri Larvae to Spiders EXPERIMENT 3: Behavior of M. viridis and M. uhleri Larvae toward Spider Egg Sacs EXPERIMENT 4: Developmental rates of M. viridis and M. uhleri 29 Discussion 30 4. Egg Sac Penetration 34 Methods and Results 34 EXPERIMENT 5: Discovery and Penetration of Egg Sacs by Unrestrained Larvae 35 EXPERIMENT 6: Direct Observation of Egg Sac Penetration by Larvae in a Confined Area 37 EXPERIMENT 7: Larval Feeding Responses to Real Sacs and Pseudosacs 39 EXPERIMENT 8: Larval Age and Egg Sac Integrity in Relation to Sac-penetrating Behavior 40 EXPERIMENT 9: Developmental Inhibition Related to Number of Larvae per Sac 40 EXPERIMENT 10: Egg Sac Age and Larval Development 42 Discussion 43 5. Boarding of Spiders 48 Methods and Results 48 EXPERIMENT 11: Spider Sex and Larval Boarding 48 EXPERIMENT 12: Spider Size and Larval Boarding 49 EXPERIMENT 13: Larval Behavior toward Previously Boarded Spiders 50 Observations of Spider Boarding 5] Discussion BS 6. Movements of First Instar Mantispid Larvae on Spiders 55 Methods and Results 5S EXPERIMENT 14: Larval Behavior on Nearly Mature Phidippus audax ao EXPERIMENT 15: Larval Behavior on Third and Fourth Instar Phidippus audax 57 EXPERIMENT 16: Larval Behavior on Lycosa rabida 57 Discussion 64 Positions First Adopted by Larvae after Boarding 64 Entry of Larvae into Book Lungs 65 9. Larval Movements Associated with the Adult Molt and with Egg-Sac Entry Transfer of Larvae from One Spider to Another . Species of Spiders Utilized Methods and Results Discussion . The Mantispid Seasonal Cycle Methods and Results Discussion Summary Addendum Appendix I. Mantispa uhleri Banks-Spider Associations in the Southern Illinois University at Carbondale Collection Appendix II. Spider Species Utilized by First Instar Mantispa uhleri Larvae Literature Cited Index Figures Fig. 1. Second instar mantispid larva in pseudosac Fig. 2. Third instar mantispid larva in pseudosac Fig. 3. Cocoon in pseudosac Fig. 4. A-D: First, second, third instar larvae Fig. 5. Egg clutch after hatching Fig. 6. Mature third instar larva Fig. 7. Mantispa uhlert Banks, adult male Fig. 8. Arena used in Experiment 5 Fig. 9. Arena for observing larvae in Experiment 6 Fig.10. Rearing cage for Lycosa rabida through fifth instar Fig.11. Rearing cage for L. rabida, sixth and later instars 101 1Y7 122 127 Fig. Fig. Fig. Fig. Fig. Fig. Fig. 12 13. 14. ey 16. Le 18. Fate of M. uhleri larvae on developing L. rabida First instar larva of M. uhleri on pedicel of L. rabida First instar larva of M. whleri in book lung of L. rabida Adult M. uhleri collections, 1974 Adult M. uhleri collections, 1975 Winter temperatures preceding summer collections A-C: Naturally occurring cocoons of M. uhleri in egg sacs of Philodromus vulgaris from shagbark hickory trees 62 65 67 83 84 85 88-89 1. Review of the Literature For more than two hundred years naturalists have been fascinated by members of the neuropteran family Mantispidae because of their remarkable resemblance to the Mantidae, brought about largely by the possession of an elongate prothorax and raptorial prothoracic legs. In fact, early systematists often confused the groups and described new species of mantispids as mantids. This phenomenon is now recognized as an excellent example of convergent evolution—these insects are not closely related phylo- genetically but have evolved similar adult structures due to similar selective pressures. The similarities are actually rather superficial, and close examination reveals many profound differ- ences between the groups. Most striking is the fact that preying mantids are hemimetabolous with nymphal stages that closely parallel the adult; mantispids are holometabolous with larval stages structurally and ecologically distinct from the adult. The family Mantispidae consists of two subfamilies: the Mantispinae and the Platymantispinae. The Platymantispinae will concern us little in this report. Suffice it to say that it is the more primitive subfamily and its members have a less mantid, more lacewing-like appearance. The larval stages of all man- tispids are predaceous. Larval platymantispines have been natu- rally associated with a number of different insects, and in the laboratory several species have been reared on a variety of sedentary arthropod foods (MacLeod and Redborg, 1982). In contrast, all known larval feeding associations for the Man- tispinae are with spiders. Our knowledge of the developmental ecology of the Man- tispinae begins with the patient work of Friderich Brauer who, in 1852, published an illustration of the egg and first instar larva of 2 Mantispa uhleri Banks Mantispa styriaca Poda, which had been secured from a rare, field-collected female. Three years later, while excavating in the soil, Brauer discovered a pupa of M. styriaca in its cocoon. In retrospect, it seems likely that this cocoon was actually within the egg sac of a burrowing lycosid that had been killed or frightened away by Brauer’s digging. It is a testament to his careful methods that his illustration (Brauer, 1855) of the reconstructed appear- ance of the cocoon in the soil shows it surrounded by what was, unbeknownst to him, the tattered remains of a spider egg sac. It was not until Rogenhofer’s (1862) serendipitous observation of the emergence of an adult of M. styriaca from the egg sac of a species of Lycosa that the pieces of this puzzle began to fall in place for Brauer. Even then, seven more years were to elapse before it became certain that spider eggs were the obligate larval food of this species (Brauer,1869). Brauer demonstrated that larvae of Mantispa styriaca can burrow directly through the wall of a spider egg sac. Seventy years later, when Hungerford (1939) discovered ten to fifteen first instar larvae on the pedicel of a field-collected spider, Arctosa littoralis (Hentz), additional information was finally obtained to show how spider eggs can be found by a first instar larval mantispid. Viets (1941) actually observed that first instar larvae of Mantispa interrupta Say, released into a container holding an adult female of an unidentified lycosid, boarded it and positioned themselves around the spinnerets. The egg sac produced by this spider yielded a pharate adult of M. interrupta and since larvae had ignored egg sacs that had been presented to them, Viets hypothesized that the larvae had entered the sac at the time of its production. Kaston (1938, 1940) also suggested the possibility of a mantispid larva’s first boarding a spider, in order to account for his rearing of a specimen of Mantispa fuscicornis Banks from an egg sac of Agelenopsis naevia (Walckenaer) (cited as Agelena naevia) which had been spun after the spider had been collected. In contrast to this route, McKeown and Mincham’s (1948) studies of Mantispa vittata Guerin are consistent with Brauer’s original observations of the direct penetration of egg sacs. They report that, when confined in a jar with a female spider, larvae made no attempts at boarding. In contrast, living mantispid larvae were discovered within two egg sacs that had been placed in Review of the Literature 3 the vicinity of searching first instar larvae, although the penetra- tion of the sac was not observed. The most recent word (as of 1979; but see Addendum) on the mantispid larval route to its food has come from Lucchese (1955, 1956), who concluded that larvae of Perlamantispa perla (Pallas) (cited as Mantispa perla, transferred to Perlamantispa by Hand- schin, 1960) were unable to penetrate the egg sacs of lycosids which he offered. The preferred larval targets were the immature spiderlings and adults of lycosids which the larvae boarded and remained upon (although Fig. 60 of the 1956 paper shows not only lycosids but also gnaphosid spiderlings carrying larvae) as the latter matured and from which they entered the egg sacs produced by mature females. Quite obviously, our understanding of the path by which larval mantispids locate their food has been fragmentary since Brauer’s time. We believe that one of the important results of our study is a clarification of this situation. In addition to the studies of Brauer, Viets, McKeown and Mincham, and Lucchese, two other investigations involved successful rearing of mantispines from the egg. Bisset and Moran (1967) reared an unidentified South African mantispid and described in detail its cocoon-spinning behavior. Davidson (1969) provided a brief description of the larva of Mantispa viridis Walker and reared three adults of this species that were fed a novel diet of crushed cabbage loopers, Trichoplusia ni (Hubner). Most of the remaining literature touching on the developmental ecology of these insects has been the result of casual observations, such as the unexpected emergence of a mantispid from the egg sac of a spider that had been collected for quite different purposes. Poujade (1898) and Main (1931) reported the rearings of Mantispa styriaca from the egg sacs of Drassodes hypocrita Simon and an unidentified gnaphosid [=drassid], respectively. Kishida (1929— original not seen; cited in Bristowe, 1932) described the associa- tion of Eumantispa harmandi Navas with a clubionid and a ctenizid. Kaston (1940) recorded the emergence of Mantispa interrupta from the egg sac of Gnaphosa muscorum (L. Koch). Also reporting on the emergence of adults from egg sacs, Milliron (1940) recorded Cupiennius sallez (Keyserling) and Parfin (1958) recorded A gelenopsis sp. (probably A. pennsylvanica C. L. Koch) 4 Mantispa uhleri Banks as spiders whose egg sacs are utilized by Mantispa viridis. Stein (1955) described the emergence of two green mantispids (un- doubtedly M. viridis) from spider egg sacs collected in New Jersey (Stein, personal communication). Although Stein identifies the sacs as that of lycosids, the illustration reveals that they are probably from a clubionid or gnaphosid. Another green man- tispid, M. viridula Erickson, was reported by Birabén (1960) from the egg sacs of Metepeira labyrinthea (Hentz). Mantispa decorata Erickson was reared from the egg sac of Lycosa poliostoma (Koch) by Capocasale (1971) and, finally, George and George (1975) reported the emergence of Climaciella brunnea (Say) from the egg sac of a lycosid, Tarantula sp. The fortunate collection of an adult mantispid, or the findings of a location in, or time of year during which adult mantispids were comparatively common have prompted several reports. Thus, Smith (1934) observed the adults and obtained eggs and first instar larvae of Climaciella brunnea (cited as Mantispa brunnea) as well as the eggs and first instar larvae of Mantispa sayi Banks. Smith’s paper also recorded the emergence of two pharate adults of M. interrupta from egg sacs of the salticid Philaeus militaris Hentz. The collection of numerous adults of Mantispa interrupta during one summer made possible Hunger- ford’s (1936) account of mating in this species. This same year Hoffman (1936) also described the eggs and first instar larval behavior of Climaciella brunnea (cited as Climaciella brunnea var. occidentalis Banks). Batra’s (1972) description of the behavior of this species suggested that the “host” might be hymenopterous, inconsistent with the rearing observation of George and George (1975). Kuroko’s 1961 report provides notes on the eggs and first instar larvae of two Japanese mantispids. Lastly, while his findings do not fall into any of the above categories, it is worth mentioning that Valerio (1971) studied the natural occurrence of larval Mantispa viridis and of the hyme- nopteran Baeus sp. in the egg sacs of Achaearanea tepidariorum (C. L. Koch). Review of the Literature 5 Previous studies on the Mantispidae have been hampered by their reliance on happenstance observation rather than upon data derived in significant amounts from controlled situations. It is the purpose of the present study to sharpen our understanding of these remarkable insects through an intensive examination of one particular species, Mantispa uhleri Banks. We hope that this endeavor will prove to be as important to future workers as Brauer’s investigations were to us. 2. Laboratory Culture of Mantispids The following materials and methods, which will be used through- out the monograph, have been successfully employed in rearing from the egg some thousand adults representing not only Mantispa uhleri but also M. fuscicornis Banks, M. interrupta Say, M. pul- chella (Banks), M. Say: Banks, and M. viridis Walker. Maintenance of adults Initial laboratory cultures used the offspring of two female Mantispa uhleri collected at Ferne Clyffe State Park, Johnson Co., Illinois, 25 August 1972 at ultraviolet light. The cultures have been periodically outcrossed to wild-caught stock from a number of Illinois locales. The mantispids were kept in small screw-top glass jars (5.9 cm diam x 6.4 cm high) with holes (0.2 cm diam) drilled in the Bakelite top for ventilation. The top and sides of each jar were lined with filter paper. Adults were given a house fly each day and had constant access to water from a small cotton pledget which was kept moist. Eggs were readily laid on the filter paper-lined sides and top. The egg clutches, still on the filter paper, were isolated in 2-dram shell vials as soon as detected, and were incubated in a glass chamber over a saturated water solution of KBr that maintains a relative humidity of 80% (Sheldon and MacLeod, 1971). Larval rearing-chambers Prior to egg hatching, artificial rearing chambers—‘“‘pseudo- sacs’’—were prepared by excavating cylindrical depressions in a hardened mixture of nine parts plaster of Paris to one part Laboratory Culture of Mantispids 7 powdered activated carbon, the optimal proportions for use in rearing several species of neuropterans (MacLeod and Spiegler, 1961; MacLeod, personal observation) and other small arthropods (Huber, 1958, and references therein). Water was added to the dry mixture until it became just fluid enough to pour in drops, as a series of small globules. A Petri dish served as a suitable mold and container, and, after hardening, the pseudosacs were drilled slightly deeper than wide, by means of a drill bit 4-inch in diameter. After complete drying, compressed air was used to blow the pseudosacs free of dust. For rearings, at least three layers of spider eggs were added to each pseudosac. A fine camel’s hair brush was used to place one first instar larva in each pseudosac, and the area around the hole was slightly moistened with distilled water. The pseudosacs were then closed by placing over each a 1-cm square of glass cut from a standard 1-mm thick microscope slide. This was gently pressed until a seal was made. These pseudosacs (Figs. 1, 2 and 3) area modification of the rearing chambers designed by McKeown and Mincham (1948) and later employed by Bisset and Moran (1967). After all moisture from the wetting procedure had dissipated, the pseudosacs were placed in the 80% relative humidity chamber. It is necessary to remove free water droplets from the vicinity of the eggs, because at 80% relative humidity these droplets will persist long enough to support the growth of mold which is lethal to mantispid larvae. Larvae were left undisturbed, except for being visually observed through the glass, until engorgement was well under way and there was no possibility of a larva’s escaping by slipping out between the glass and the plaster. The chambers were then opened so that more spider eggs could be added. After being spun, the cocoons (Fig. 3) were carefully removed from the pseudosac and transferred to individual 7-dram vials lined with filter paper. Vials were stoppered with cotton plugs wrapped with Kimwipes tissue, and returned to the humidity chamber. The tissue wrapping prevents the pre-tarsal claws of the newly eclosed adult from becoming entangled in the cotton fibers. During eclosion, the pharate adult bites its way out of the cocoon and climbs up the vial’s filter paper-lined side where it undergoes its final ecdysis. 8 Mantispa uhleri Banks Fig. 1. Second instar mantispid Fig. 2. Third instar mantispid larva in pseudosac. larva in pseudosac. Spider eggs Eggs from many species of spiders have proved successful as larval food for M. whleri, and it is likely that most spider eggs would be suitable. For instance, we have used the eggs of Argiope aurantia Lucas, which are found in late summer. Because these eggs are generally cemented together within the egg sac, they are suitable as larval food only after being separated from each other, care being taken to prevent any flow of yolk from punctured eggs. Because of their ready availability, how- ever, the eggs most used in our present study were those of a theridiid, Achaearanea tepidariorum, an agelenid, Agelenopsis sp., and a salticid, Phidippus audax (Hentz). Adult females of Achaearanea were collected in numbers during the summer and fall under bridges, under park benches, and around the windows of houses and garages. Females were placed in individual filter paper-lined 7-dram vials stoppered with cotton, and fed house flies, Musca domestica Linnaeus, every other day. Egg sacs for immediate use were removed as soon as produced and stored at 5°C to retard development. For long-term use, Achaearanea eggs can be killed by freezing and Laboratory Culture of Mantispids 5) Fig. 3. Cocoon in pseudosac; covering glass slide in place. stored at -20°C for long periods with no apparent effect on their subsequent suitability as mantispid food. During the winter months, when Achaearanea were not available, egg sacs of Agelenopsis sp. were collected from beneath the bark of trees. They were especially abundant beneath the bark of living Osage orange trees, Maclura pomifera (Rafin- esque) Schneider, which are commonly found in central I}linois as hedgerow plantings. Such eggs, collected in midwinter before development had been initiated, provided a very satisfactory larval food. Unlike Achaearanea eggs, however, Agelenopsis sp. eggs proved unacceptable as food after they had been killed by freezing. In the same winter environment where Agelenopsis eggs occurred, subadults of Phidippus audax were found in silken retreats. These were matured in the laboratory and eggs obtained from females about 2 weeks after mating. These served to span the period from early spring (when Agelenopsis eggs were no longer suitable because of their stage of development) to summer (when Achaearanea were again present). Phidippus audux eggs were also unsatisfactory after freezing. 10 Mantispa uhleri Banks Care was taken to give first instar mantispid larvae only eggs that showed no signs of development. We have observed that by the time the spiderlings’ appendages can be seen through the chorion, the mantispid larvae seem not to be feeding. In contrast, third instar larvae can feed somewhat on even newly hatched spiderlings, although small black marks occasionally appear on the mantispids, suggesting that the spiderlings are capable of damaging the larval integument. Since Davidson (1969) had some slight success using a non- spider diet of crushed cabbage loopers for Mantispa viridis, we attempted to substitute a more readily available larval food for M. uhlen. Eggs of the bagworm moth Thyridopterix ephemerae- formis (Hayworth) were tried, and, although third instar larvae did feed on them, the moth eggs had a toxic effect. Our experience suggests that an alternative food would probably be no easier to employ than spider eggs, since, with a little planning, an almost unlimited supply of these can be obtained. Rearing and mating conditions Egg incubation and larval-pupal rearing took place at a controlled photoperiod of L:D= 16:8, 25°C, and 80% relative humidity. Adult mating and oviposition occurred under these same conditions except that the relative humidity fluctuated between 20 and 50%. Mating behavior was observed under a GE 60 W 115-125 V ruby red light during the early scotophase of the daily cycle. Virgin males and females were used for observation of mating behavior. Pairings were made in plastic cages measur- ing 8.5 x 12.5 x 6.0 cm with 3.7-cm diameter screened holes cut in the top for ventilation. To reduce cannibalism, the adult mantis- pids used were well fed and at least 1 week old. Measurements Mantispid larvae used for measurements of size and duration of developmental stages were fed eggs of Achaearanea tep- idariorum. Third instar larvae were presented with an unlimited supply of these eggs so that some were still available when the larvae began spinning cocoons. Second and third instar larvae Laboratory Culture of Mantispids 1] were measured from material either preserved in Peterson’s KAAD fluid and stored in 95% ethyl alcohol or killed in hot water, fixed in Kahle’s fluid, and stored in 70% ethyl alcohol. Unfed first instar larvae were macerated in Nesbitt’s fluid (Nesbitt, 1945) and temporarily mounted in glycerine beneath a cover slip prior to measurement. Except for the lengths of engorged larvae, all measurements were made under a dissecting microscope with an ocular micrometer, calibrated with a stage micrometer. Because of their curved shape, engorged larvae were drawn on graph paper while being viewed through an ocular grid and were measured with a planimeter. The drawing of the unfed first instar larva (Fig. 4) was made from a cleared specimen prepared and mounted as described above but examined under a phase-contrast compound micro- scope. Drawings of second and third instar larvae were made from specimens macerated in Nesbitt’s fluid, stained with Chlorazol Black E (2% in 70% ethyl alcohol), and examined in glycerine. Egg counts were made by enlarging photographs of individ- ual mantispid egg clutches taken subsequent to hatching and after staining the filter paper background with India ink (Fig. 5). Colors noted in our descriptions were directly compared to a standard color atlas (Maerz and Paul, 1930) and the specific reference in parentheses following each of our subjective descrip- tions refers to the closest shade in this reference by plate number, column, and row. All measures of dispersion in this monograph are standard errors of the mean. Experiment 1: Adult size variation Since a considerable size variation in field-collected adult mantispids has been observed by numerous workers, an experi- ment was designed to examine the relationship between the amount of food ingested in the larval stages and adult size attained. Differing allotments of eggs of A. tepidariorum were made available to third instar larvae by placing the eggs in the pseudosacs of engorged, quiescent second instar larvae just prior to ecdysis. Allotments of 30, 40, or 50 eggs, or a quantity in excess 12 Mantispa uhleri Banks (oul ‘ AN Sean Fig. 4 A. Unfed first instar larva; dorsal view. . Third instar larva; head capsule, lateral view. B C. Second instar larva; abdominal tp, lateral view, segments VITI-X. D . Third instar larva; abdominal tip, lateral view, segments VII-X. Laboratory Culture of Mantispids 13 « Ow eje ¢ < ete? ey te * 7. 4 3 ee", gt * . ¢ —< ’ °e - ” oe * ~ eo * ~", ’ 348 or se ue . ° >“ e»* * Sao es Se * ae . 2 ea ¢ s” Fig. 5. Egg clutch after hatching. of what the larva would consume were utilized. For this last group, enough eggs were placed in the pseudosac so that an excess remained at the time of cocoon spinning. The following criteria served as indices of adult size: the width of the head capsule measured through the eyes at their widest point in anterior view; the length of the forewing from the base of the subcosta to the wing tip; and the length of the pronotum, as seen in lateral view, from the posterior point of articulation with the cervical sclerites to the point of articulation with the mesothorax. These were compared with similar measurements taken from the largest and smallest wild-caught males and females available to us. Head capsule measurements were com- pared by means of a two-way ANOVA via multiple regression analysis using food groups and sex of the mantispid as main effects, assuming no interaction. RESULTS Oviposition Large individual clutches of eggs were deposited as a series of curved rows produced by the slow back-and-forth movement of 14 Mantispa uhleri Banks the abdomen of the ovipositing female mantispid (Fig. 5). Each egg was attached to the filter-paper jar lining by a short stalk; stalk length varied from clutch to clutch. Ten laboratory-reared females, mated only once, averaged an egg clutch every 3-5 days and produced 12.6+1.7 fertile egg clutches throughout their lifetimes. The mean number of eggs per clutch from four wild- caught females from three different localities ranged from 614 to 2,976 (Table 1). Thus a large female is capable of laying more than 35,000 eggs during her lifetime. Table 1. Egg production of wild-caught Mantispa uhleri females Egegs/Clutch Female (Mean + SE) % Hatch ] 614415 (N = 3)2 98.84+ 0.45 2 1,802+ 101 (N = 4) 97.88+ 0.60 3 2,3074 220 (N = 5) 98.37+ 0.60 4 2,976 149 (N = 4) 98.4522 0:37 *N = number of clutches. Egg Ellipsoid; prior to any embryonic development varying greatly in color from one clutch to another, from tannish yellow (9H2) through pale tan (1OB3) to ight pink (9A3); eyes become visible at 5.6+0.2 days (N = 10 clutches) and the usual neuropteran pattern of transverse banding is visible through the chorion at 7.1 + 0.2 days (N = 10 clutches; each clutch is from a different female). Egg measurements are given in Table 2. First instar larva (Fig. 4A) Campodeiform, with a series of transverse reddish brown (6K11) segmented bands on abdomen, each band consisting of a posterior bilateral pair of hook-shaped pigmented areas with the point of the hook directed cephalad and the shank of the hook perpendicular to the dorsal vessel which is visible between the medial ends of the pair; each thoracic segment with a pair of dark brown (7E10) laterodorsal sclerites; background color of the unfed first instar larva light peach (9B5). Average length 0.884 mm. Duration of first stage 7.4 days (Tables 2 and 3). Laboratory Culture of Mantispids 15 Table 2. Measurements of developmental stages of Mantispa uhleri Developmental stage Measurements in mm (Mean + SE) Egg Length, micropyle excluded 0.362 + 0.003(N=15) Width’ 0.182 + 0.002(N=15) First instar larva Length of unfed larva 0.884 + 0.008(N=7) Length of engorged mature larva? 1.621 + 0.007(N=7) Width of head capsule* 0.126 + 0.002(N=7) Mature second instar larva Length” 2.945 + 0.097(N=6) Width of head capsule* 0.243 + 0.004(N=7) Mature third instar larva Length” 10.6 +0.4 (N=7) Width of head capsule* 0.468 + 0.012(N=7) @ At widest portion. > Measured from tip of mouthparts to posterior margin of tenth tergum. Table 3. Duration of developmental stages of Mantispa uhleri Developmental stage Duration in days (Mean + SE) First instar larva 7.44 0.3(N=11) Second instar larva 2.3+0.1(N=11) Third instar larva prespinning 2.5+ 0'2(N=11) Third instar larva postspinning 5.7£0.1(N=11) Pupa 9.9 + 0.2(N=11) Adult“ 114.0+ 7.0(N=8) a a . Maximum longevity of mated laboratory-reared females. Subsequent development of the larva Prior to ecdysis, the engorged first instar larva passes through an immobile quiescent period during which the cuticle becomes shiny and transparent and the eyes of the pharate second instar larva can be discerned within. The onset of ecdysis is signaled when a dorsal rent in the old cuticle appears at the rear margin of the head capsule and then proceeds to extend caudad for about one-half the length of the 16 Mantispa uhleri Banks , larva. At no time does the old head-capsule split. The method by which the second instar larva frees itself from the old cuticle does not differ appreciably from that described for Chrysopa oculata Say by Smith (1922). The newly ecdysed larva is immobile for a short time before it resumes feeding. Ecdysis of the third instar larva follows an identical sequence. Second (Fig. 1) and third instar (Figs. 2 and 6) larvae Physogastric, the legs disproportionately shorter and less heavily sclerotized than those of first instar larvae and lacking any important locomotor function. At a magnification of 30X, only slight morphological differ- ences appear to exist between these instars, the most pronounced involving an increase in the number of setae in the third instar and the modification of the tenth abdominal segment into a spinneret in this form (Figs. 4C and 4D). The color of the second and third instar larvae is determined by the color of the egg contents which they ingest. Larvae fed Achaearanea eggs are a pale tan (10B3) with an overlay of white mottlings produced by isolated portions of the larval fat body. Cocoon (Fig. 3) Cocoon spinning is basically the same as that described by Bisset and Moran (1967), with the spinneret tracing the same “figure of eight’”” movements in the final phase of spinning aftera “haphazard” foundation has been erected. The internal portion of the cocoon is composed of a series of silken panels, each made up of a number of “figure of eight” figures. Each panel is connected to the one previously made bya strand of silk so thatifa cocoon is cut open and one panel withdrawn, a series of panels then follows like scarves from a magician’s hat. Pupa Exarate, with the prothorax not greatly enlongate; two pairs of dorsal abdominal hooks on each of the third and fourth abdomi- nal segments, similar to those described for the hemerobiid Wesmealius quadrifasciatus (Reuter) by Killington (1936) and for “MOIA [RIL] ‘BAIL, LeIsUT piTy) sine, “9 ‘Sq 18 Mantispa uhleri Banks the mantispid studied by Bisset and Moran (1967); on each segment, the points of the anterior pair of hooks project cephalad and the points of the posterior pair project caudad, the dorsal vessel running between the members of each pair. After ecdysis to the adult, a dark, liquid meconium is emitted from the anus. This is in contrast to the hard meconial pellet produced at eclosion by other Neuroptera except the Conioptery- gidae (Withycombe, 1925) and the berothid genus Lomamyia (MacLeod, unpublished). Adult (Fig. 7) Pronotum dark grey (8A9) with a darker mid-dorsal line; femur and tibia of prothoracic leg mostly shiny black (48L7) but, variably, a paler patch beside the femoral teeth laterally and at the proximal end of tibia; coxa pale, with a dark line running the length of its anterior surface; meso- and metathoracic legs pale tan (10B3); abdomen yellow (9J2), marked with black (48L7), a broad black line down the venter, a black line laterally intersecting the spiracles although area directly around spiracles yellow, a black Fig. 7. Mantispa uhleri Banks, adult male. Laboratory Culture of Mantispids 19 line down the midline of the dorsum expanded laterally at the posterior margin of each segment. Sexual dimorphism as follows: Female pterothoracic epimeral plates, episternal plates, mera, trochantins, and coxae black (48L7) with only the area directly around sulci pale yellow (9E2); corresponding parts in the male pale yellow with only occasional splotches of black; black abdominal lines of female somewhat broader, especially the ventral line, so that female abdomen carries proportionately more black than the male. Mating behavior The observations summarized here were made from detailed notes taken during six pairings with four successful matings. Subsequent observations of many additional matings have pro- duced no substantial deviations. Two matings were followed closely enough for accurate timing of some of the behavior components (Table 4). The mating ritual usually begins with the male and female on the top of the cage in an upside down position. Less commonly they are positioned on the side of the cage, or, rarely, on the floor of the cage. Courtship starts with the male and female facing each other. Beginning with either sex, the male and female reciprocally slowly extend forward the coxa and femur of one foreleg, followed by a movement which returns the leg to its original position. This is followed by an identical sequence of movements of the other foreleg. We term these alternating bouts of extension and flexion “sparring.” As noted, sparring is usually reciprocal in that the movements of the right foreleg of one sex are followed by identical movements of the right foreleg Table 4. Timings of epigamic components in Mantispa uhleri Engagement of genitalia completed Abortive Minutes after Minutes after Minutes Pair sparring o and ? initiation in number cycles placed together of sparring copulation ] 0 520 0.75 25825 2 9.63 0.72 35.90 20 Mantispa uhleri Banks of the other, then followed by a similar exchange between the sexes involving the left foreleg, and so on. Sparring will some- times be stopped and restarted resulting in individual extensions and flexions by either partner, but once both sexes begin doing it at the same time it is always reciprocal. While sparring, the male may repeatedly flick his abdomen up and down. At this time a sweetish odor can be detected. Prior to or between bouts of sparring, male and female may stand closely facing each other, motionless, except for rapid antennal movements by both. Antennal contact does not occur. After sparring, the male turns away from the female and assumes a characteristic position in front of her in which his abdominal segments are extended, his wings held vertically over the body with their dorsal surfaces together, while both forelegs are simultaneously slowly extended and flexed. We refer to this position as the “‘wings-up stance.’ During this time that the male is in front of the female he is usually turned away from her, although his body may be perpendicular to hers with his head to the right or left of her body. At no time is he facing the female, nor does he move while in the wings-up stance. During this time the female continues to spar and may move toward the male. The male then backs up and, if necessary, rotates his body so as to achieve an orientation parallel to that of the female. He then’ curves the tip of his abdomen toward the female’s genitalia and touches her. If adequate contact is made, the male continues to rotate until his head is pointed in the opposite direction to that of the female. As is usual in the Neuroptera, the tips of the two abdomens have their ventral surfaces in contact so that, when this final orientation is attained, the male’s abdomen is twisted 180 degrees. While the male is attempting to make contact, his wings remain with their dorsal surfaces together. After copulation begins, his wings are lowered from their fully raised position, but, like the female’s, they are held high enough to prevent interference with copulation. Upon separation, the male and female turn toward each other and the courtship is often ended with a quick jab of the forelegs, usually from the female. The male can often be seen to apply his mouthparts to his genitalia shortly after mating. A large opales- cent spermatophore protrudes from the tip of the female’s Laboratory Culture of Mantispids 21 abdomen after a successful mating. This is no longer visible after 24 hours. Courtship can be terminated in several ways. The male and female may begin sparring, but one or both may walk away. The female may abandon the male after he assumes the wings-up stance or may back away from him when he attempts copulation. Sometimes several cycles of sparring and of assuming the wings- up stance occur before a successful engagement of the genitalia is achieved. DISCUSSION Eggs The number of eggs laid per clutch is quite variable among female mantispids (Table 1), although the totals vary much less among clutches from the same female. From subjective observa- tions of the egg clutches from many wild-caught females, it appears likely that the number of eggs per clutch and the ovipositing female’s body size are directly proportional. Al- though no measurements were taken, we recall that Female | (Table 1) was exceptionally small and by far the smallest of the four females listed. Size of the female may account for the variable stalk lengths observed, since larger females may lift their abdomens higher from the egg-laying surface before releasing the egg. Adult size variation Under culture conditions, third instar larvae begin cocoon- spinning when they have consumed all of the eggs in the pseudosac or when they are unable to locate any of the remain- ing eggs. After spinning begins, the larvae ignore any additional eggs provided, even though these would have been consumed if given prior to the initiation of spinning. A direct relationship exists between the amount of food ingested by the third instar larva and the body size of the resulting adult (Table 5). This factor in all the estimates of adult size is obvious; therefore, a detailed analysis was performed only for head capsule width. 22 Mantispa uhleri Banks Larvae fed an unlimited number of spider eggs were nearly twice as large as those fed only 50 eggs and were even larger than any of the others. Because of the disproportionate weight that this class would contribute to the multiple regression analysis, it was excluded from consideration. Among the first three groups, S€Xaqj Was not significant (Fy), = 1.428, 0.5 > P > 0.25); but food groups,q; were very highly significant (F, ,,; = 47.204, P > 0.001, MSE = .0008). This ability of mantispids to reach maturity with an exception- ally wide range of food supplies is in direct contrast with the situation in Chrysopidae and Hemerobiidae (MacLeod, unpub- lished), in which death occurs to a third instar larva unless it receives a nearly full supply of food. Only a single third instar M. whleri larva successfully matured on a diet of 30 Achaearanea eggs, suggesting that the ingestion of some minimal amount is probably necessary for successful cocoon spinning, pupation, and adult eclosion and that 30 Table 5. Adult size variation in reared and wild-caught Mantispa uhleri (Exper. 1) Adults Adult measurements (mm) (Mean + SE) Eggs measured Head capsule Wing Pronotum provided Number Sex width length length 30 ] 2 1.200 (ee) 2.0 40 2 & 1.3002 0025 7.7200 9-222 2 ? 1.312+,.0.037 8020.5 222508 50 3 & 1408+ 0.008 8.140:.2 2329 7 ? 1423+ 0010 89201 24255 Excess 3 & 2.442 + 0.117. 14.32,0.5 4. eee 2 2 2.475+ 0.049 162+02 4040.0 Representative Smallest * 1.550 9.0 2.55 wild-caught Smallest 2 1.550 9.9 24 adults* Largest o& 2.475 lo 4.] Largest 2 2.375 14.5 Pe “Approximately 100 specimens of each sex were examined. Laboratory Culture of Mantispids 23 Achaearanea eggs are close to providing this minimum. Above this minimum, our data suggest that (regardless of the availability of food) the probability of reaching the adult stage is quite high, although the size of the adult produced is closely related to the amount of food consumed. The ability of a third instar larva to spin after a minimal amount of food has been eaten is assuredly adaptive, for once a larva enters a spider egg sac its food supply is fixed because it has no way of reaching another egg sac. The food content within an egg sac (dependent on egg size and number) is probably variable within any spider species and is most certainly variable among species. Since there 1s good evidence that under natural conditions Mantispa uhleri enters the egg sacs of spiders of a number of species belonging to several families, it is probable that the food supply available for the larvae will be quite variable. The size variation of these mantispids contrasts with the relatively uniform size that is found in most insects, so we must look for a satisfactory explanation of the presumed advantage of producing variable-sized adults, particularly since extremes in adult size could easily present problems in mating. Aside from the physical handicaps associated with the copulation of different- sized mantispid individuals, small representatives of either sex might risk becoming a meal instead of a mate when approaching a larger individual. Thus, in consideration of this problem alone, it would seem that selection should favor size equality of the sexes, as it seems to in most insects, so that adults of smaller, but of nearly uniform, size might be expected from egg sacs regardless of the size of these sacs. Every female must divide the energy available for reproduction between that utilized in positioning eggs in the environment and the energy actually contained in the eggs themselves. Time may well be the limiting factor for a female insect that spends a great deal of energy in locating sites, such as host animals or plants, for oviposition. For such a female there is no advantage in being able to produce more eggs than she has time to lay. But there should be an advantage in increased egg-laying capacity for a female that invests primarily in the eggs themselves. Although we are ignorant of exactly where female M. uhleri lay their eggs, it seems likely that they expend little energy in 24 Mantispa uhleri Banks positioning eggs. This conclusion is derived from the very large size of the egg clutches of M. uhleri, which indicates a low probability of larval success in locating spider eggs and that the larvae are pretty much on their own in locating food. Thus, the number of offspring a female can produce is not likely to be limited by time, but rather by the number of eggs she can physically synthesize, which is probably correlated with her size. Large size, whenever this is allowed by a surfeit of available spider eggs, should be valuable in species such as M. uhleri where the low probability of larval survival must be compensated for by the production of large numbers of eggs. Of course, such arguments do not explain the advantage of large size to the male. This feature in the male may simply be carried along by selection in the female. Or perhaps, faced with a population of variable-sized females, large size in the male is valuable because an increase in size decreases the likelihood of being eaten by a larger female during courtship. It would seem advantageous for a male to mate with as large a femaleas possible, since this would increase his reproductive potential, and this may be more easily accomplished if the male is as large as his larval food supply permits. Possible male pheromone Eltringham (1932) postulated the existence of a male phero- mone as the product of an eversible glandular sac between the fourth and fifth abdominal segments of M. styriaca. An analogous situation occurs in M. uhleri since, while there is no eversible sac, there are porous areas with associated glandular epithelia in the flexible cuticle between the third and fourth, fourth and fifth, and fifth and sixth abdominal tergites of the male (MacLeod, unpub- lished). These areas are likely sources for the sweetish odor which can be detected during courtship. The observed flicking of the male abdomen and extension of his abdominal segments may serve to broadcast this pheromone. What seems to be this same odor has been apparent to us when examining freshly thawed specimens that had been preserved by freezing. Closer examina- tion of these specimens under a dissecting microscope has revealed small liquid droplets exuding from these porous areas of the abdomen. 3. Larval Strategies for Locating Spider-Egg Prey The means by which a searching first instar mantispid larva locates a spider egg has remained an intriguing question for more than a century after Brauer (1869) demonstrated that larvae of M. styriaca can burrow directly through the wall of a spider egg sac. The reports of Hungerford (1939), Kaston (1938, 1940), Viets (1941) and Lucchese (1955, 1956) suggest that the larvae of their respective species board spiders; McKeown and Mincham’s (1948) study, like Brauer’s, confirms the findings of direct penetration of egg sacs by mantispid larvae. The seeming contradictions in these observations are magnified by the implication that each investigator is presenting a fragment of evidence of some general tactic common to all larval mantis- pids. We contend that they can better be identified as two rather distinct strategies: the direct penetration of spider egg sacs that have been spun and left in the environment, and the mantispids’ boarding of spiders prior to egg sac production and entrance into the sac at the time of spinning. We shall also argue that these two maneuvers may not necessarily be mutually exclusive in a particular species of mantispid, different mantispid species utilizing either tactic, or both, to varying extents. Previous studies have been hampered by reliance on random observation rather than upon data derived in significant quanti- ties from controlled situations. Negative findings are particularly difficult to evaluate, since, for example, if a larva of a particular species fails to penetrate an egg sac it is possible that the type and condition of the sac or its presentation may account for the failure of the larve to penetrate. We have circumvented this problem and others through a series of experimental studies in which such variables as larval age and 25 26 Mantispa uhleri Banks species of spider producing the egg sac or presented for boarding were controlled. Further, we have employed as the experimental subjects larvae of Mantispa viridis and M. uhleri. Because of their rather different behaviors, detailed below, each species acts as a partial control for the results obtained from the other. Our investigations of these strategies are dealt with in Experi- ments 2, 3, and 4. In Experiment 2, larvae of both species were exposed simultaneously to the same spider to see how their behavior differed toward it. In Experiment 3, the behavior of the two species toward egg sacs was investigated in a situation in which the surface area to be searched was an experimental variable. The activities within the egg sac and the developmental rates of the larvae of M. uhleri and M. viridis were compared in Experiment 4, and the important differences found between these are discussed relative to the distinctions in their methods of finding spider eggs. Our results suggest that M. viridis is an obligate egg sac penetrator only, while M. uhleri is capable of this same penetrat- ing behavior as well as the tactic of boarding spiders prior to egg production, entering the sacs subsequently as these are spun. These two strategies can be related to the findings of previous workers on other mantispid species. METHODS AND RESULTS First instar larvae of both M. uhleri and M. viridis were obtained from adults reared in the laboratory using the techniques already described (pp. 6-11). The culture of M. viridis was descended from females collected at Little Grand Canyon, Jackson Co., Illinois. Under low magnification, larvae of the two mantispid species were easily distinguished by the pattern of abdominal pigmenta- tion, which in M. uhleri is composed of a series of transverse bands and in M. viridis consists of a bilateral pair of dorsal longitudinal stripes extending the length of the abdomen. All experiments were conducted with larvae that had hatched from the egg on the day on which the experiment was begun. Larval Strategies for Locating Spider-Egg Prey 27 Experiment 2: Reaction of M. viridis and M. uhleri larvae to spiders Spiders of several species (Table 6) representing five different families were collected. With the exception of the specimen of Admestina tibialis (C. L. Koch), collected at Carbondale, Illinois, all were taken in the area of Urbana, Illinois. Each spider was placed in a separate screw-top jar measuring 8.9 cm high and 7.2 cm in diameter with an internal surface of 233 cm?. Holes had been drilled in the Bakelite jar tops for ventilation, and a circular piece of filter paper placed between the top and lip to prevent the escape of larvae. At the beginning of each exposure, 20 first instar larvae of Mantispa viridis and 20 of M. whleri were released on the filter paper-lined jar top with a spider situated in the bottom of the jar. The jars were then left undisturbed for 24 hours, after Table 6. Boarding frequency of Mantispa uwhleri and Mantispa viridis on various spider species (Exper. 2) Spider species Number and sex Larvae boarding spider used M. uhleri__M. viridis Dysderidae Ariadna bicolor 5 ? 26/1004 1/100 Agelenidae Agelenopsis sp. 1 immature 5/20 0/20 Lycosidae Schizocosa sp. 3 immature 17/60 0/60 Pardosa milvina 2 oy 6/40 0/40 Thomisidae Philodromus vulgaris 3 subadulto’ 40/60 0/60 Salticidae Phidippus audax 3 subadult ? 40/60 0/60 Metacyrba undata 3 (2, o& immature) 20/60 0/60 Admestina tibialis ] ? 1/20 0/20 Total larvae on spiders/total larvae presented to spiders 155/420 1/420 * 26 larvae boarded out of a total of 100 larvae presented to the spiders. 28 Mantispa uhleri Banks which the spiders were quickly placed in a Syracuse dish containing 70% ethyl alcohol. Spiders were then scored as to the number of larvae found either still attached or in the alcohol. Of the 420 larvae of each species exposed to 21 spiders, 155 M. uhleri and 1 M. viridis were found associated with a spider (Table 6). Most of the M. whleri were still on the spider, with only a few loose in the alcohol of the dish. These larvae most frequently took up a position on the pedicel of the spider, although larvae were also found between the spinnerets and around the bases of the legs. The single larvae of M. viridis was found loose in the alcohol of the dish. Experiment 3: Behavior of M. viridis and M. uhleri larvae toward spider egg sacs The egg sacs used were those of an agelenid, Agelenopsis sp., collected during the winter months in and around Urbana, Illinois, from beneath the loose bark of Osage orange trees. For laboratory rearings, we have collected overwintering A gelenopsis egg sacs from these same sites for the past five years and have never found any naturally occurring mantispid larvae within them; it is thus virtually certain that the larvae found within them at the conclusion of our investigations were our experimental larvae. The egg sacs are lenticular in shape and usually constructed flat against the inner surface of the bark and covered with a cone of loose silk and debris. Sacs were carefully separated from this cone before use in the experiment. Two sets of three screw-top jars of increasing total surface were used: 135 cm? (6.4 cm high, 5.9 cm diam); 233 cm? (jar used in the preceding experiment); and 573 cm? (17.1 cm high, 9.5 cm diam). One Agelenopsis sac was placed in each of the six jars. At the beginning of each replication five newly hatched larvae of Mantispa viridis were placed in each of the three jars of one set, on the filter paper-lined Bakelite tops that sealed the jar; five M. uhleri larvae were placed in each of the three jars of the other set. Jars were left undisturbed for 24 hours after which they were opened and the egg sacs scored for larvae in and on the sac. Ten replications were run. In each case, all larvae were accounted for to ensure that none had escaped from the jar. The independence of Larval Strategies for Locating Spider-Egg Prey 29 Table 7. Mantispa uhleri and Mantispa viridis penetration of A gelenopsis sp. egg sacs in controlled surface areas (Exper. 3) Number of M. uhleri Number of M. viridis larvae (N=50) larvae (N=50) Surface area Associated Not associated Associated Not associated of chambers with sac with sac with sac with sac 135 cm? 192° 3] 46 4 233 cm? 92.4 4] ggh4 12 573 cm? 4ae 46 4ghe 6 * Frequencies of larvae associated with sac and not associated with sac differ significantly among three chambers (X 2= 13.90, P < 0.05, df = 2). > Frequencies of larvae associated with sac and not associated with sac not significantly different among three chambers (X ?= 5.54, 0.1 > P > 0.05, df = 2). © Frequencies of larvae associated with sac and not associated with sac differ significantly between species in small chamber (X ? adj = 29.71, P < 0.005, df = 1). 4 Frequencies of larvae associated with sac and not associated with sac differ significantly between species in medium chamber ( X 2? adj = 31.47, P < 0.005, df = 1). © Frequencies of larvae associated with sac and not associated with sac differ significantly between species in large chamber (X ? adj = 60.94, P < 0.005, df = 1). the effect of surface area of the jar to be searched and mantispid species on the location of larvae was tested using the chi-square distribution. Significantly more larvae of M. viridis than of M. uhleri were associated with the egg sac for all jar sizes (Table 7). The frequency of M. uwhleri larval association decreased with increas- ing jar size with the difference among the three sizes being significant. The corresponding frequencies for M. viridis were not significantly different. Experiment 4: Developmental Rates of M. viridis and M. uhleri Larvae of both species were reared with use of Achaearanea tepidariorum eggs in pseudosacs covered by glass slides. Larvae were left undisturbed except for visual observations through the glass until the third instar, at which time enough spider eggs were added so that an excess remained at the time of cocoon spinning, ensuring that the duration of this stage reflected the larva’s ontogenetic program rather than the amount of food available. Differences in the duration of corresponding stages in the two 30 Mantispa uhleri Banks Table 8. Duration of developmental stages of Mantispa uhleri and Mantispa viridis t (Exper. 4) Mean days + SE t-test (two-tailed) Developmental M.uhleri M. viridis comparison between Stage (N = 11) (N = 14) species First instar AE 03 44 Os t= 10.10*** Second instar Pearce Dal Los O.1 t = 2.32* Third instar 2p.a 10.2 2.0 + 0.0 t = 3.28** Prepupa (cocoon spinning to pupation) as) | 6 ie sh t = 3.45** Pupa 99 = 02 9.0 = 02 t = 3.63** Total Zee OFF yaaa es | 7 = \Fed eggs of Achaearanea tepidariorum at 25° C, L:D = 16:8, relative humidity = 80%. * =0.05 >P> 0.01. ** =0.01 > P > 0.001. *** = P < 0.001. species were compared using a t-test. The durations of the developmental stages for M. viridis were significantly shorter than those for M. uhleri (Table 8). DISCUSSION Our first insight into the possibility that different species of mantispids might have basically different ways of reaching their spider-egg prey came while rearing mantispids in the laboratory and after placing a hatching egg clutch of M. uhleri and one of M. viridis in a large jar with a female jumping spider, Phidippus audax. After 24 hours the spider had spun a flimsy retreat around herself, and, with the naked eye, we could see hundreds of larvae milling around and over her body. The larvae of these species are easily distinguished under the magnification of a dissecting microscope and an examination revealed that the only larvae that had actually boarded the spider were Mantispa uhleri, while the larvae milling around her and oriented in close proximity to the silken retreat were all M. viridis. It is likely that M. viridis larvae will only seldom be found on spiders (Table 6). Because each of the spiders used in this experiment represented a portion of the surface area available to Larval Strategies for Locating Spider-Egg Prey 31 mantispid larvae of both species, one might expect to count some larvae as having boarded that were simply walking across the surface of the spider at the instant that it was removed from the Jar, and this is probably the explanation for the one M. viridis associated with Arzadna bicolor (Hentz). Although it is possible that Mantispa viridis is a spider boarder specific for families or species we did not test, this seems unlikely, in view of the apparently nonspecific nature of its egg sac-penetrating behavior. Achaearanea and Agelenopsis sacs were the first presented to Mantispa viridis larvae, and both were readily entered. Later, egg sacs from the jumping spider, Metacyrba undata (De Geer), were made available to searching Mantispa viridis larvae, and these, likewise, were penetrated. Such impartial penetrating and feed- ing behavior is not consistent with a narrow boarding range. Larvae of M. uhleri have boarded every spider presented to them in the laboratory, including hunting spiders of the genera Herpyllus, Lycosa, Phidippus, and Salticus, as well as the web spinners Achaearanea, Araneus, and Argiope, provided that these latter were prevented from suspending themselves in a web. However, the natural prey range is probably somewhat narrower. The rather high incidence of spider boarding indicated by these data suggests that the location of spiders by Mantispa uhleri may not be random. Although both species exhibit sac-locating and sac-penetrating behavior, M. viridis is more successful at this than M. uhleri (Table 7). An increase in the surface area on which the sac was located did not seem to affect M. viridis, since there were no significant differences in the numbers of larvae penetrating sacs in the three different-sized jars. However, the differences were significant in M. uhlerz. The decreased efficiency in locating egg sacs in successively larger jars suggests that the location of egg sacs by this species is more nearly random. These data are consistent with the hypothesis that M. uhleri encounters sacs through random searching, while larvae of M. viridis are able to orient to egg sacs from a distance. Furthermore, the data are also consistent with our added suggestion that some element of M. uhleri’s search for spiders may not be random, which leads to the prediction that M. uhleri differs importantly from M. viridis in actively orienting its search behavior toward 32 Mantispa uhleri Banks spiders and directly penetrating sacs only if these are accidentally encountered along the way. Before this idea can be accepted, however, certain complicating factors must be eliminated. For example, larvae of M. uhler: may show strong orientation toward egg sacs and penetrating behavior after some specific amount of time has passed without encountering a spider. The area searched by the larvae of the two species in a given amount of time might also have its own significance. Quite obviously, a closer examina- tion of such possibilities is needed. The rates of development of these two species are very likely re- lated to the differences in the routes by which these mantispids reach spider eggs. Thus, the more rapid development of M. viridis, especially in the first instar, should be adaptive for an egg sac penetrator that will probably encounter eggs in varying stages of development and occasionally have to contend with hatching spiderlings. But larvae of M. uhler, entering the sac during its construction, from a position on the spider, could afford to feed in a comparatively leisurely fashion. Other behavioral aspects parallel the developmental rate. Third instar M. uhleri are extremely sluggish and can be damaged and killed by hatching spiderlings. Disturbing these larvae by touching them produces, at most, slight and extremely slow reactive movements. In contrast, third instar larvae of M. viridis are quite active and whip violently away when touched, a possible protective mechanism against hatching spiderlings. On the basis of the preceding considerations we should like to suggest that, despite the fragmentary nature of some of the observations, all of the other mantispid species studied by previous workers also show one of the two basic strategies demonstrated in these experiments (Table 9). Although Brauer (1869) did not expose larvae of M. styriaca to spiders, and McKeown and Mincham’s (1948) observations of M. vittata and spiders were not extensive, we would not expect to see spider- boarding activity in either of these species because of their documented overwintering as unfed first instar larvae. Spiders present for boarding in the spring would also have been available the preceding fall and would seem to offer a more protected overwintering site than the one actually used, which in M. vittata is an aggregation of larvae ina mass beneath loose bark. Larvae of Larval Strategies for Locating Spider-Egg Prey 33 Table 9. Method of egg sac entry by various mantispids, as extrapolated from the literature Mantispid species Reference Spider Egg sac boarding penetration Mantispa interrupta Say Viets, 1940 =F ? Mantispa styriaca Poda Brauer, 1869 - + Mantispa uhleri Banks Present report a + Mantispa viridis Walker Present report - cf Mantispa vittata Guer. McKeown and Mincham, 1948 - + Perlamantispa perla (Pallas) Lucchese, 1956 ST ? Perlamantispa perla (Lucchese, 1956) and Mantispa uhleri do, in fact, overwinter on spiders. In the case of M. whler this overwintering behavior provides not only protection and insula- tion, but food as well, since the larvae actually feed on spider blood during this time (Redborg and MacLeod, 1983b). Thus, for a spider-boarding mantispid, there would be little advantage in waiting until spring to search for spiders. Viets’s observations are not adequate to rule out egg sac penetration for M. interrupta. Only one example of negative evidence is given, and the conditions under which the sacs were presented are not described in detail. Lucchese’s findings that egg sac penetration by Perlamantispa perla does not occur are more convincing. Still, he is not specific about controls for his observations. It is not known, for example, whether those larvae that ignored egg sacs were simultaneously exposed to spiders of the same species that they did, in fact, board. In addition, Lucchese used large numbers of larvae in his investigations, a situation that might reduce the likelihood of sac penetration in favor of spider boarding in a species that does both. It does seem likely to us, however, that mantispids will be found which specialize in spider-boarding behavior to the exclusion of any direct penetration of egg sacs. Perlamantispa perla may well be one of these. The clarification of these two rather different methods of locating spider eggs should help to illuminate such other aspects of mantispid biology as the possible choice of oviposition sites by females and the sensory cues used by searching first instar larvae. 4. Egg Sac Penetration In our experiments the mantispid larvae were confined for long periods (ca. 24 hrs) with spiders or egg sacs, and were not allowed to leave the area; this of course is contrary to natural conditions. It could thus be argued that either egg sac penetration or spider boarding is a laboratory artifact brought about by the repeated, prolonged exposure of larvae to egg sacs or spiders. The boarding of spiders can be confirmed in nature by examining wild-caught spiders for the presence of mantispid larvae. It is difficult, however, to secure data showing that in nature larvae of M. uhleri directly penetrate egg sacs. Collected in the field, an egg sac containing a larva tells us nothing of the larva’s method of entrance. Direct penetration under laboratory conditions might be only a small facultative component of a much more elaborate sequence of behaviors following the obli- gate boarding of a female spider. A larva may ideally reach the eggs as they are being deposited. Direct egg sac penetration might be employed if and only if the larva reaches the vicinity of the eggs after the sac has been partially constructed. In a series of six experiments we have examined some of the details of egg sac-penetrating behavior. The problem of whether direct egg sac penetration is normal larval activity or an artifact was addressed directly in Experiments 5, 6, and 7. Experiment 8 is concerned with some of the factors that affect egg sac penetration. Finally, Experiments 9 and 10 focus on problems encountered by larvae after egg sac penetration. METHODS AND RESULTS Egg sacs used in the following experiments were either those of Achaearanea tepidariorum, obtained as soon as spun from laboratory maintained, field-collected adults, or those of Agele- 34 Egg Sac Penetration 35 nopsis sp. collected from beneath the loose bark of Osage orange trees and prepared as detailed for Experiment 3 (p. 28). We used adult crab spiders, Philodromus vulgaris Hentz, taken from the same habitats as the Agelenopsis sacs. All three spider species were collected in the vicinity of Urbana, Illinois. All experiments were carried out at a temperature of 25°C with 80% relative humidity and a photoperiod of L:D= 16:8. Frequencies were analyzed using the chi-square (X 2) distribution or the G statistic. Means were compared by a t-test. Experiment 5: Discovery and penetration of egg sacs by unrestrained larvae An experimental arena was constructed with an inverted 11.5- cm diameter Petri dish top, as illustrated in Figure 8. Phil- odromus spiders or Agelenopsis egg sacs, or both, were placed at positions P!] through P6 of Figure 8. Spiders were restrained by folding a small piece of cheesecloth in half, placing the spider between the two pieces thus formed, and closing the three open cheesecloth =< — = enclosure my for spider == Petri dish top eee used for arena peee DOttom i | 3 4 Fig. 8. Arena used in Experiment 5. 36 Mantispa uhleri Banks sides with Plasticine. Crab spiders were specifically chosen because of their inability to escape from such a restraint. Eggs or larvae of Mantispa uhleri were positioned in the center of the arena at the beginning of each experimental run. The arena was covered by the inverted bottom of the Petri dish, supported by three pieces of Plasticine; the wide gap thus created between the arena floor (Petri dish top) and arena covering (Petri dish bottom) allowed larvae to leave the arena freely around its circumference. Since the underside of the arena bottom was not completely flat, it was placed on a large sheet of cotton wool to ensure even contact between the bottom and the surface on which the arena was placed so that larvae would not become trapped on the surface of the arena. The experiment was run four times, with different combinations of spiders and egg sacs as follows: Run 1. Spiders were placed at positions P1, P3, and P5 and egg sacs were placed at positions P2, P4, and P6. Egg sacs were placed in cheesecloth enclosures identical to those used for the spiders. Just prior to hatching, one entire M. uhleri egg clutch on filter paper was placed in the center of the arena. Run 2. Egg sacs were placed at all six positions with only those at Pl, P3, and P5 being placed in cheesecloth enclosures. Just prior to hatching, an entire mantispid egg clutch on filter paper was again placed in the center of the arena. Run 3. Spider egg sacs were placed at all six positions. No cheesecloth enclosures were used. A group of 3-day-old man- tispid larvae hatched from a single egg clutch was allowed to disperse from the center of the arena. These larvae had been stored in a cotton-stoppered 2-dram shell vial since hatching. The vial was opened and placed vertically in the center of the arena at the beginning of the run. Run 4. Egg sacs were placed at all six positions without cheesecloth enclosures. One hundred newly hatched first instar larvae from the same clutch were released into the arena, 20 ata time, over a 3-day period. The clutch was kept in a stoppered shell vial and larvae were individually transferred with a small camel’s hair brush to a small piece of filter paper that was placed in the center of the arena. Each increment of 20 larvae was released only when all larvae from the previous release had dispersed and were no longer visible. Egg Sac Penetration 37 All runs were scored for the number of larvae found on spiders or in egg sacs. Under CO anesthesia, spiders and egg sacs were examined through a dissecting microscope. Empty egg shells from the clutches used in the first three runs were counted to determine the number of larvae available in the arena. Results are summarized in Table 10. Table 10. Pentration of Agelenopsis sp. egg sacs by unrestrained first instar Mantispa uhleri (Exper. 5) Available Larvae on Larvae in Run larvae spiders egg sacs 1 1,398 127 0? 2 1,087 not used 3b.d 3 842 not used sve 4 100 not used 4cd * Frequencies of spider boarding versus sac penetration significantly different ( Xs a 71,864, P< 0.001, df = 1). > Frequencies of sac penetration between Runs 2 and 3 not significantly different (Gigi 2.722, 0.1 > P > 0.05, df = 1). “Frequencies of sac penetration between Runs 3 and 4 not significantly different (Gg; = 3.195, 0.1 > P > 0.05, df = 1). 4 Frequencies of sac penetration between Runs 2 and 4 significantly different (Ga = 8.330, 0.005 > P > 0.001, df = 1). Experiment 6: Direct observation of egg sac penetration by larvae in a confined area A 27-cm? arena was constructed by filling the inside of a5.9-cm diameter Bakelite jar lid with Plasticine, to weight it down, and placing it right side up intoa Petri dish of larger diameter (Fig. 9). Water was poured into the Petri dish until it just reached the top of the Bakelite lid, forming a circular surface surrounded by water that effectively isolated a first instar larva on an island. Achaeara- nea egg sacs were utilized in this experiment, one sac being placed in the center of the arena at the beginning of each run. One first instar Mantispa uhlert was placed at the circumference of the arena with a fine camel’s hair brush. The larvae used in this experiment varied in age from | to 4 days. The different lots were not siblings. Movements of the larvae were observed and times were recorded for the following events to occur: (1) egg sac hit (a hit was recorded when a larva disappeared from sight beneath the sac viewed dorsally); (2) egg sac mounting; (3) initiation of egg sac penetration (initiation was recorded when walking ceased and 38 Mantispa uhleri Banks water reservoir Bakelite | jar lid \ Petri dish i Ud. bok Lad ba bok be i ba A Sa yan 4 “+ ba > (eee tei i Fig. 9. Arena used for observing larvae in Experiment 6. side-to-side movements of the head capsule against the silk were observed); (4) time interval required to penetrate an egg sac and completely disappear from view. The behavior of 12 larvae was recorded (Table 11), their movements observed through a dissect- Table 11. Behavior of twelve first instar larvae of Mantispa uhlert on egg sacs of Achaearanea (Exper. 6) Number of Minutes encounters Mounting Time until Time from Larva with sac of sac mounting mounting to penetration ] ] = 14 18 Zz ] a 14 2 3 ] tp ll 17 4 ] a5 2 6 5 ] ly 5 ae 6 ] - - - 7 2 a 7 20 8 2 + 6 10 9 2 a 6 12 10 3 = 32 9 1] 4 =f 12 52 12 4 = 16 38 Egg Sac Penetration 39 ing microscope at 10X magnification. Observations were termi- nated after a period of 45 minutes if the larva had not yet mounted the egg sac. Only | in 12 larvae failed to mount and subsequently penetrate the egg sac within the allotted period. Experiment 7: Larval feeding responses to real sacs and pseudosacs Pairs of Achaearanea egg sacs of the same age were collected and stored at 10°C until used. One sac of each pair was randomly selected, opened, and its contents placed in a standard pseudosac. The other sac of the pair was placed unopened in a 2-dram shell vial containing in the bottom % inch of a hardened mixture of plaster of Paris and activated carbon, the same material used for the pseudosac. Two first instar larvae on their day of hatching from the same egg clutch were used in each run. With a fine camel’s hair brush one larva was placed on the eggs in the pseudosac which was then closed by wetting the area around it with distilled water and gently pressing a small piece of standard 1-mm thick microscope slide over the opening until a seal was made. The second larva was placed on the Achaearanea sac in the vial and a piece of cotton wrapped in a Kimwipes was then pushed into the vial until it just touched the egg sac. The pairs were scored as to the final developmental stage reached by each larva, and, if an adult was produced, the length of time taken until eclosion was noted (Table 12). No significant differences were observed between the two groups. All mortality beyond the first larval stadium occurred in egg Table 12. Behavior of first instar Mantispa uhleri larvae toward Achaearanea spider eggs in real egg sacs and pseudosacs (Exper. 7) Total Number Number Days to larvae Number dying of adult exposed not during adults emergence Eggs in to eggs feeding development produced (Mean + SE) Pseudosacs 44 pa ig 312 24.2 + G9" Real sacs 44 63 92 292, -24.4+ 0.20" * Frequency distributions between pseudosacs and real sacs not significantly different (x?= 2.267, 0.75 > P > 0.50, df = 2). b Means not significantly different (t = 0.723, 0.5 > P > 0.2, 2-tailed). 40 Mantispa uhleri Banks sacs from which spiderlings hatched before larval development was complete. In most instances black dots and blotches could be observed on the larval cuticle. Experiment 8: Larval age and egg sac integrity in relation to sac-penetrating behavior The same type of medium-sized jars described in Experiment 3 were used. Each jar contained one Agelenopsis egg sac, situated on the jar bottom. At the beginning of each replication, five first instar Mantispa uhleri larvae were released on the filter paper- lined top of each jar. After 24 hours each egg sac was scored for the number of larvae inside. The experiment was run three times, each with 10 replications. In Run 1, freshly hatched larvae and completely closed egg sacs were used. In Run 2, newly hatched larvae from the same clutch as Run | were used, but the Agelenopsis sacs were slightly torn so that an open path to the eggs existed. Run 3 was identical to Run 1, using 2-day-old larvae from the same clutch. Results (Table 13) indicate that neither larval age nor egg sac integrity 1s a significant factor in the mantispid’s egg sac- penetrating behavior. Table 13. Effect of larval age and egg sac integrity on penetration frequency of Agelenopsis egg sacs by first instar Mantispa uhleri (Exper. 8) Number of larvae Location of 0-1-day-old 2-3-day-old larvae Whole sac Torn sac Whole sac Torn sac In egg sac4 4 6 8 _ Not in egg sac 46 44 42 - “Frequencies of larvae in egg sacs not significantly different among three groups ( X? = 1.515, 0.5 >P > 0.1, df = 2). Experiment 9: Developmental inhibition related to number of larvae per sac For each replication three Achaearanea egg sacs of the same age were opened and their contents emptied into each of three Egg Sac Penetration 4] standard pseudosacs. Freshly hatched Mantispa uhleri larvae from the same clutch were placed in the three pseudosacs as follows: One larva in sac #1 (Group 1); two larvae in sac #2 (Group 2); and three larvae in sac #3 (Group 3). Larvae were inserted and the pseudosacs closed as described in Experiment 7. No eggs were added. Under our visual monitoring each pseudo- sac was left undisturbed until a larva reached the spinning stage or until hatching spiderlings were evident within. The time elapsed and the stage reached by each larva were recorded (Table 14). Fourteen replications were made. The larvae dying during the first stadium in all three groups showed no signs of feeding. The larva dying during the second stadium in Group 2 and four of the dead third instar larvae and the dead pupa in Group 3 died in association with hatching spiderlings. These dead larvae also showed black discolorations on their cuticles. The dead second instar larva and the remaining two unsuccessful third instar larvae in Group 3 had been killed and eaten by another larva. The frequency of larval survival past the first stadium decreased through the three groups with the differences among them significant ( X2 = 8.810, .025 >P > .01, df = 2). The number of days to cocoon spinning increased significantly from Group | to Group 2 (t = 3.518, 0.001 > P > 0.0005, 1-tailed) and from Group 2 to Group 3 (t = 2.005, 0.05 > P > 0.025, 1-tailed). In no instance did more than one larva reach maturity in any given pseudosac. Table 14. Survival of competing larvae in an egg sac (Exper. 9) Number of Larvae Number Number of mantispids larvae sur- Days to per of Final stage reached by larva viving past cocoon pseudo- pseudo- Larval instar Pupa Adult first spinning sac sacs ] 2 3 instar/total (Mean + SE) ] 14 2 0 0 is 12/14 10.0 + 0.21 (N = 12) 2 14 15 1 0 0 is 13/28 15 2-38 (N = 12) 3 1425 od gebiex, ul 9 17/42 12.74 0.49 (N = 11) 42 Mantispa uhleri Banks Experiment 10: Egg sac age and larval development Ten Achaearanea egg sacs from laboratory-maintained females were collected on each of four consecutive days and stored at 25°C. The experiment was begun the last day of collection, with the 40 egg sacs in four age groups as follows: Group | (0-24 hours old); Group 2 (24-48 hours old); Group 3 (48-72 hours old); and, Group 4 (72-96 hours old). Each sac was placed in a 2- dram shell vial at the bottom of which was % inch of a hardened mixture of plaster of Paris and activated carbon moistened to raise the humidity. All larvae used were siblings from a single egg clutch. One freshly hatched Mantispa uhleri larva was placed on each sac, and a Kimwipes-wrapped cotton plug was inserted into the end of each vial until it just touched the egg sac. The sacs were left undisturbed until the emergence of an adult mantispid or spiderlings. If spiderlings emerged, the sac was examined to determine the stage reached by the larva. The number of larvae successfully reaching the adult stage decreased as the age of the egg sac increased. Larvae that did not develop beyond the first stadium failed either to penetrate their sacs or to begin feeding. Larvae dying in the second or third stadia had entered sacs that produced hatching spiderlings before the larvae had reached the cocoon-spinning stage. As noted for Experiments 7 and 9, the cuticles of most of these dead second and third instars exhibited black marks. Results are summarized in Table 15. Table 15. Effect of age of Achaearanea egg sacs on the development of Mantispa uhleri larvae? (Exper. 10) Number of mantispids Group/Age Stage reached by larvae after penetration of eggs Number Failing to Larval instar Pupa Adult (days) of sacs penetrate ] yd 3 1/0-1 10 0 0 2 0 f 2/1-2 10 2 0 0 3) 0 5 3/2-3 10 0 0 3 6 0 ] 4/3-4 10 ] 2 6 1 0 0 “Frequencies of larvae reaching the adult stage for Groups | and 2 combined vs. Groups 3 and 4 combined were significantly different (x? 4, = 11.396, P< 0.005, df = 1). Groups were combined to avoid cells with expected frequencies less than 5. Egg Sac Penetration 43 DISCUSSION Results of several experiments are consistent with the conclusion that direct egg sac penetration is a naturally occurring phenome- non. In Runs 2-4 of Experiment 5 a total of 15 larvae penetrated the Agelenopsis sacs. These larvae were not confined and had the opportunity of leaving the arena as an alternative to mounting and penetrating an egg sac. Where larvae were confined (Experi- ment 6), 11 of 12 penetrated an egg sac within 64 minutes, and 5 of these larvae entered immediately after their first encounter with the sac. These encounters were thus neither prolonged or repeated and gave no indication of being “artificial.” If direct penetration is not a part of Mantispa uhleri’s natural behavior, then the barrier presented by intact egg sacs (Experiment 7) might be expected to produce a decrease in survival or an increase in development time, compared to larvae placed directly with spider eggs. Such differences could possibly be explained in terms of the time wasted in a prolonged search for spiders and the final, inefficient penetration of the sac. The fact that no significant differences were found suggests that larvae penetrated quickly and efficiently. Our results do not bear out the notion that egg sacs are directly entered only after repeated encounters. Like Pandora’s Box, however, Experiment 5 suggests much more than was first anticipated. The information obtained in Run 1, in which both spiders and egg sacs were available and no larvae penetrated or were even found close to the egg sacs, was totally unexpected. In addition to the 72 larvae found on the spiders, a total of 154 dead larvae were found directly beneath the spiders. Presumably the spiders could accommodate only so many larvae, and since larvae feed on spider blood after boarding (Redborg and MacLeod, 1983b), these dead individuals may have starved while attempting to board or been killed by their successful competitors. As indicated, this apparent orientation to the spiders was highly significant. This result might be accounted for either by factors increasing the likelihood of spider boarding or by factors decreasing the probability of egg sac penetration. Factors increasing spider boarding might involve the ability of larvae to locate spiders from a distance, whereas egg sacs are encountered only by random search, or—if both spiders and sacs 44 Mantispa uhleri Banks are located at random—a greater frequency of abandonment and continued searching after encountering an egg sac than after locating a spider. But the number of larvae penetrating in Runs 2, 3, and 4 where spiders were not present was surprisingly low, since direct observation (Table 11) indicates that any larva encountering an egg sac has a high probability of entering. This result suggests the existence of a factor that decreases the probability of egg sac penetration under the conditions of Experiment 5. A hypothesis consistent with all experiments is the existence of an inhibitory mechanism that prevents larvae from penetrating egg sacs located close to their hatching egg clutch, but which does not affect spider boarding. Such a mechanism could be adaptive since an egg sac located near hatching eggs of M. whleri is likely to be encountered by many larvae. As detailed below, there is good evidence that only one larva will reach maturity in any egg sac. Thus, if two larvae penetrate an egg sac simultaneously, only one larva is likely to survive to cocoon spinning. Behavior discouraging a larva from penetrating a sac under such conditions should be favored by selection. This would not seem to be the case for the boarding of spiders which, unlike stationary egg sacs (excepting those of Lycosidae and Pisauridae which are carried by the egg-laying female), may move briefly through the vicinity of hatching larvae; a mobile spider temporarily located near a hatching mantispid egg clutch would not be as likely to pick up many larvae, compared to an egg sac. That an egg sac will be encountered by many larvae is assured. Thus, an appropriate reaction for a larva might be to ignore, or at least be wary of, egg sacs near its hatching site, but to board a spider, in any event. There would seem to be two possible ways in which an inhibitory mechanism acting to prevent the penetration of nearby egg sacs might work. First, larvae might be unresponsive to egg sacs for a given amount of time, allowing them sufficient opportunity to disperse some distance from the egg clutch. The time might be measured by some sort of energy clock related to distance traveled. If this were the case, larval age would be expected to have considerable effect on penetration activity. In Run 3 (Table 10) where larvae were 3 days old, the number penetrating was not significatly greater than in Run 2 where freshly hatched larvae were used. This lack of any correlation Egg Sac Penetration 45 with larval age is also corroborated by the results of Experiment 8 in which there was no apparent difference in egg sac-penetra- tion frequency between 0-1 and 2-3 day-old larvae. A second possibility is the inhibition of sac-penetrating activity by the presence of other larvae. As larvae disperse, the number of larvae per unit area would decrease until at some point penetrat- ing behavior is activated. This mechanism would have the advantage of being self-adjusting for egg clutches of different sizes that are likely to occur because of M. uwhleri’s extreme adult size variation. By chance, the number of larvae hatching from the egg clutches decreased in Runs 1-3 and, by design, Run 4 had no more than 20 larvae in the arena at any particular time. Thus, the larval density at any given time was lower in each succeeding run. The number of larvae penetrating sacs increased from Runs 1-4, and the differences in the frequency of penetration between Runs 2 and 4 were significant. This partial result is consistent with the hypothesis just outlined. Very likely, additional possibilities accounting for these results could be proposed, and the existence of some such mechanism is worth further investigation. One factor that could affect egg sac penetration is the texture and composition of the egg sac. The difficulty experienced by a larva of M. uhleri in penetrating an egg sac may conceivably vary with the species of spider producing the sac. Achaearanea egg sacs are penetrated in less than one hour. In this case, the larva crawls over the sac until it finds a suitable point of entry. The head, appressed to the surface of the sac, is then moved from side to side, appearing to abrade the surface. The larva enlarges the opening thus made, moves into the sac, and disappears beneath the silk. Although microscopy does not reveal any obvious adaptations in Mantispa uhleri, it is possible that the mouthparts may either cut the silk or release a matrix-dissolving enzyme. Closer study of the exact mechanism of penetration is required. The low rate of penetration of Agelenopsis sacs (Experiment 5) cannot be explained solely by difficulty of entrance, since sacs torn open to facilitate entry (Experiment 8) had no influence on the number of larvae entering sacs. Once a sac is entered, a larva may encounter several problems, a major one being penetration by additional larvae. Our findings (Table 14) strongly suggest that no more than one Mantispa uhleri will reach maturity in any egg sac. Interestingly, it also 46 Mantispa uhleri Banks appears that only one begins development, the other unsuccessful larvae dying (or being killed) without appreciable feeding. Indeed, the lack of feeding by the unsuccessful larvae suggests that the larvae actively search out one another and interact until just one remains alive. Only then does feeding begin. Of course, the possibility exists that the larvae simply refrain from feeding until starvation eliminates all but one, but it seems unlikely that a larva would passively starve with available food present. Although it might seem strange that larvae do not begin feeding immediately in an attempt to “‘outeat’’ competitors, there may be forces operating against this. Noteworthy is the fact that the larva that begins to feed at once will also be the first to reach the more vulnerable, quiescent intermolt period, during which assassination by a smaller, more agile larva might be compara- tively easy. In all three instances in which two larvae began development, one was eventually killed by the other and its contents ingested. Further data derived from Experiment 9 (Table 14) support the hypothesis that larvae compete with each other in the egg sac and mutually inhibit development. Here the number of larvae beginning development decreased as the number of larvae per sac increased. The significant increase in time to cocoon spinning among the three groups 1s also consistent with the conclusion that as the number of larvae per sac increases, so does the time until elimination of all but one larva. A second factor at work is the age of the egg sac. If the sac is too old, a larva may not have enough time to complete feeding and spin a cocoon before the spider eggs hatch (this would not happen to larvae entering freshly spun sacs after having first boarded the spider). This problem is documented in Table 15. Development in Achaearanea is quite rapid, with spiderlings hatching from the egg sac 11-12 days after construction at 25°C in the laboratory. As the age of the egg sac prior to its entry bya larva increases, the number of adult mantispids emerging decreases and the average final stage reached by unsuccessful larvae decreases. Thus, only a single adult was obtained from 20 egg sacs 2-4 days old, while 20 egg sacs 0-2 days old produced 12 adult mantispids. This difference was highly significant: seven of ten larvae developed to the third instar in Group 3, while only one of ten did so in Group 4. Egg Sac Penetration 47 These results indicate that unless a Mantispa uhleri larva locates an Achaearanea egg sac within 48 hours of its construc- tion, the probability of the larva’s developing to the adult is extremely low. Depending on the species of spider, the larval maneuver of direct egg sac penetration may be a rather risky affair. Achaearanea, for example, has such a short developmental period that the 11-12 days needed for the hatching of the spiderlings is only slightly longer than the time needed for development to cocoon spinning by larvae of Mantispa uhleri. But such rapid spider egg development may be the exception, and species such as Lycosa rabida Walckenaer and Phidippus audax, which need approximately 30 days in the laboratory to produce emerging spiderlings, would allow Mantispa uhleri a longer time for successful penetration. It is interesting that larvae in these experiments penetrated and began feeding in sacs in which they were destined not to survive. It would be adaptive for mantispid larvae to be able to test the developmental state of eggs within an egg sac and to abandon sacs that would provide insufficient time for successful cocoon spinning. Achaearanea’s rapid egg development might interfere with such an ability and it would be worth investigating a possible existence of such a mechanism in relation to spiders with a longer egg development time. 5. Boarding of Spiders Our understanding of spider boarding has been limited to the supposition that it affords the boarding mantispid larva an opportunity to enter an egg sac while it is being produced. Larval behavior toward spiders prior to and after boarding has heretofore been a mystery. In this chapter we examine factors affecting spider boarding in a series of three experiments in which first instar larvae of M. uhleri were exposed to spiders under various conditions and scored as to whether boarding occurred. The actual boarding process is described from observations of restrained spiders. METHODS AND RESULTS The following experiments were carried out at a temperature of 25°C and a photoperiod of L:D = 16:8. Frequencies of spider boarding in the various experiments were analyzed using the chi-square distribution. Experiment 11: Spider sex and larval boarding Mature males and females of the salticid Metacyrba undata were collected from overwintering retreats, usually beneath the loose bark of trees, in several Illinois localities. A series of three screw-top jars with increasing internal surface areas identical to those described in Experiment 3 were used. As in Experiment 3, small holes were drilled in the Bakelite jar tops for ventilation, and a piece of circular filter paper placed between the top and jar lip effectively confined larvae within the container. Five first instar larvae of Mantispa uhleri were placed on the filter paper 48 Boarding of Spiders 49 Table 16. Behavior of first instar Mantispa uhleri larvae boarding Metacyrba undata adults (Exper. 11) Surface area Number of Larvae Boarding/ Combined data of experimental Available larvae from ¢ and o& chambers ? Spiders o& Spiders spiders 135 cm? 1/25 13/40 20) Bon. 233 cm? 15/25 21/40 36/65 >< 573 cm? 6/25 8/40 14/65 “4 Totals 28a" 42/120 * 70/195 “Boarding frequencies (number of larvae boarding spiders versus number not boarding) of male and female spiders not significantly different (X? ,q; = 0.034, 0.9 > P > 0.5, df = 1). Data for three jar sizes were lumped for each sex. Homogeneity (X2 = 0.109, df = 2), not significant, indicating that lumping was justified. >Boarding frequencies significantly different (X” adj = 7.058, 0.01 > P > 0.005, df = 1). “Boarding frequencies significantly different (X” aaj = 14.332, P < 0.005, df = 1). 4 Boarding frequencies not significantly different (X* aa) = 0.996, 0.5 > P > 0.1, df = 1). of the tops of each of the three jars at the beginning of each replication and one spider was placed in each jar. All spiders within a replication were the same sex. In eight replications male spiders were used and, in five, females were used. All fifteen larvae of any replication were siblings that had hatched the day of the experiment. After 24 hours, spiders were removed from the jars, dropped in a small dish of 70% ethyl alcohol, and scored for the number of larvae that had boarded. A total of 42 out of 120 larvae boarded males of Metacyrba undata, while 28 of 75 larvae boarded females (Table 16). Larvae boarded either sex with equal frequency. Experiment 12: Spider size and larval boarding Forty-two immature salticids, Phidippus audax, were reared to the second (20), third (10), and fourth (12) stadia from eggs obtained from laboratory-maintained females. Spiders of each instar were placed individually in 2-dram vials with one first instar Mantispa uhleri. Two groups of second instar Phidippus audax were used: those that were newly emerged from the egg sac and active (10) and those that had fed for several days on Drosophila melanogaster Meigen and were partially engorged 50 Mantispa uhleri Banks (10). Vials were stoppered with cotton plugs wrapped with Kimwipes and placed in a humidifier at 80% relative humidity. They were examined daily until the mantispid boarded the spiderling, was eaten, or had died. Larvae eaten by spiderlings were easily identified by their shriveled and mutilated appearance. Larvae dying from other causes were dehydrated, but not mutilated. Results are presented in Table 17. No larvae successfully boarded unfed second instar Phidippus. As spider size increased, so did the frequency of larvae successfully boarding. Table 17. Immature Phidippus audax spiders boarded by first instar Mantispa uhleri larvae (Exper. 12) Developmental Fate of mantispid larvae stage of spider Eaten Boarding Dead Total Second instar 10 0 0 10 Unfed Second instar 5 5 0 10 Engorged Third instar 5 5A 0 10 Newly molted Fourth instar ] 10° l 12 Newly molted *Boarding frequencies (number of mantispid larvae boarding spiders vs. number not boarding) for unfed second instar and newly molted third instar spiderlings combined (to avoid contingency table cells less than 5) and compared to boarding frequencies for newly molted fourth instars. Frequencies significantly different (Xx? ada 8.061, P< 0.005, df = 1). Experiment 13: Larval behavior toward previously boarded spiders Naturally infested spiders of the species Phidippus audax and Metacyrba undata, each carrying one Mantispa uhleri first instar larva, were collected at the University of Illinois Dixon Springs Agricultural Center in southern Illinois. Each infested spider and an uninfested control spider of the same species, state of maturity, and sex ( in the case of adults) were individually confined in 2-dram shell vials with one laboratory-hatched M. uhleri larva as described in Experiment 12. After 24 hours, Boarding of Spiders 5] spiders were anesthetized under CO, and scored as to whether boarding by the laboratory-reared larva had occurred. Since larvae on spiders slowly feed on spider blood (Redborg and MacLeod, 1983b), wild-caught larvae were easily distinguished from laboratory-reared larvae by their darkened midgut. Twelve pairs of spiders were tested during the period of several days necessary to collect them. Wild-caught larvae were either reared to the adult for identification or were identified by means of an unpublished key devised by the second author. The frequencies of larvae boarding naturally infested spiders and uninfested controls were the same (Table 18). None of the wild-caught larvae moved from their original positions after the spiders were boarded by a laboratory-reared larva. Table 18. Mantispa uhleri larvae boarding spiders already carrying a larva (Exper. 13) Number of Spiders Tested Boarded Not boarded 5 Infested immature male Phidippus audax 4 5 Uninfested controls 4 2 Infested immature female Phidippus audax 2 0 2 Uninfested controls 2 0 2 Infested immature Metacyrba undata 0 2 2 Unifested controls 0 Z 3 Infested adult male Metacyrba undata 3 0 3 Uninfested controls 3 0 Total infested spiders 9 3 Total controls 9 3 Observations of spider boarding A 27-cm? arena was constructed by filling the inside of a 5.9- cm diameter Bakelite jar lid with Plasticine to weight it down and placing it right side up in a Petri dish of larger diameter. Water could be poured into the Petri dish until it just reached the top of the Bakelite lid, forming a circular arena surrounded by water that could effectively confine a first instar larva. A 52 Mantispa uhleri Banks mature female of Salticus scenicus (Clerck) was restrained in the center of the arena by applying a small drop of Elmer’s Glue-all over the tip of each leg while the spider was immobilized by C0, anesthesia. After the glue had dried the spider was allowed to revive and the reservoir surrounding the arena was filled with water. A single first instar larva Mantispa uhleri was placed at the edge of the arena and its actions were noted through a dissecting microscope. Four spiders were observed, each being exposed, sequentially, to several different larvae. The movements of a total of 20 larvae were observed. During the initial stages of our observations on the boarding behavior of larvae it proved impossible to restrain adult females of such larger spider species as Phidippus audax, as the spiders pulled free, and other means of restraint did not produce a natural orientation of the spider. Salticus scenicus was then chosen because of its small size. Although this species is not known to be utilized by Mantispa uhleri in nature, we do not believe that the observations arising from this experiment are likely to be misleading since the range of spider species utilized by M. uhleri is so broad. Adult females were used rather than immature spiderlings or males, since, if larvae ever discriminate between immature and adult, or between mature male and female, we expect that it is the latter that should be preferred. Direct observations of spider boarding revealed no discernible pattern to the larva’s movements before encountering the spider. The larvae spent most of their time running about, circling the water margin and periodically crossing the arena. After passing beneath the spider’s legs, however, each larva suddenly appeared “alert.’’ In a typical case the larva circled the spider, spending most of its time in the area around the abdomen and near the pedicel. Twitching of the spider’s abdomen and flexion of its legs were observed in response to larval contact. Boarding occurred when the larva lifted its legs toward the spider’s abdomen, remaining attached to the substrate by its caudal pygopod, and was whisked aboard the spider as the latter brushed against the larva. In a series of short, jerky movements, the larva made its way to the pedicel where it wrapped itself around it like a belt. In all instances, the larva boarded within 30 minutes of its first encounter with the spider. Boarding of Spiders 53 DISCUSSION Mantispid larvae showed no clear-cut preference (Table 16) for male or female spiders in any of the three experimental jars. The larvae boarded both sexes, although, if given a choice, a prefer- ence might have been indicated. On the assumed equivalency of larval behavior toward male and female spiders, male and female data were combined and the numbers of larvae that boarded spiders in each of the three test jars were analyzed. A significantly greater number of larvae boarded spiders in the medium-sized jar. This unexpected finding suggests complexities in the larval searching-behavior that we do not yet understand. One possible explanation would involve the antagonistic effects of random larval searching (in which the number of boardings would decrease with increasing surface area) and larvae interfering with each other’s boarding attempts (in which interference might decrease with increasing jar size, resulting in an increase in boarding). First instar Mantispa uhleri are within the prey size range for second, third, and fourth instar Phidippus audax (Table 17). The number of larvae boarding spiders, increasing in correlation with spider instar and size, presumably reflects a decreasing desirability or visibility of the mantispid larva as a food item. We assume that there is a certain spider size above which first instar Mantispa uhleri are never consumed as prey. We observed that larvae are eaten by large spiders only after having been—by chance—picked off the spider’s body by the grooming move- ments of the legs; these movements often were elicited in response to an apparent irritation caused by the larva. This seems to occur more commonly with long-legged spiders, such as lycosids, than with shorter-legged forms, such as salticids. Our results also suggest the possibility that the nutritional state of the spider may affect the larva’s success in boarding. Thus, while no unfed second instar Phidippus audax was boarded by a larva, half of the engorged second instar salticids were boarded. On the basis of our previous evidence that only one larva will reach maturity in a given egg sac, it seemed conceivable that the presence of a resident Mantispa uhleri larva on a potential host spider might prevent a second larva from boarding, either by an 54 Mantispa uhleri Banks inhibition of the boarding behavior of the second larva or by an aggressive action by the resident larva after boarding. Experi- mental results (Table 18) indicate that no such mechanisms are in operation. Consistent with this conclusion was the observa- tion that no resident wild-caught larva changed its position after a subsequent boarding by a laboratory-reared larva. This would not have been expected if the first larva were aware of the second larva’s boarding and had tried to prevent it. Although we know that a single larva on a spider will enter an egg sac at the time of oviposition, we have not seen reactions of multiple larvae on a spider at the time of egg production. Such multiple boardings do occur in nature. This phenomenon warrants further investiga- tion. 6. Movements of First Instar Mantispid Larvae on Spiders We now consider the behavior of a Mantispa uhleri larva after it has boarded a spider. Of particular interest are such questions as whether the larva can negotiate a spider molt; what is the preferred location on the spider’s body adopted by a resident larva; whether, in fact, larvae on spiders are ultimately able to enter egg sacs from this position during oviposition by the spider; what happens to larvae that have boarded male spiders. We have investigated these questions in a series of three laboratory experiments (14, 15, 16) with the salticid spider Phidippus audax and the lycosid Lycosa rabida, both of which are boarded by larvae of Mantispa uhleri in nature. In these experiments single first instar larvae were allowed to board immature spiders which were then reared—to the adult stage where possible. METHODS AND RESULTS All procedures were carried out at a photoperiod of L:D = 16:8 and a temperature of 25°C. The three experiments detailed below have certain features in common as they were carried out sequentially, with refinements added to Experiments 15 and 16 as we learned of the complexity of the phenomenon under study. The various activities in all experiments were analyzed in two- by-two contingency tables using the chi-square distribution or the Fisher Exact test. Experiment 14: Larval behavior on nearly mature Phidippus audax Large but immature Phidippus audax were collected in the vicinity of Urbana, Illinois, and at the University of Illinois Dixon.Springs Agricultural Center (DSAC) in southern Illinois 55 56 Mantispa uhleri Banks during the late fall and winter. Each spider was placed in a 2- dram shell vial with one newly hatched first instar larva. Vials were stoppered with a cotton plug wrapped with Kimwipes and placed in an 80% relative humidity chamber. The spiders were examined at 24-hour intervals until larvae were no longer observed crawling in the vials, at which time each spider was examined under CO, anesthesia in order to note and record larval positions. Spiders were then returned to their respective shell vials and each was fed one house fly, Musca domestica Linnaeus, daily until ecdysis occurred. After ecdysis, spiders were again examined under CO, and each larva’s presence and position were recorded. Spiders and mantispids recover quickly from CO, anesthesia and we have observed no apparent behavioral abberations in either organism. Approximately 2 weeks after reaching maturity spiders were paired in small plastic cages measuring 8.5 X 12.5 X 6.0 cm. Pairings involved not only spiders carrying a larva, but also other spiders that had not been exposed to mantispids. Three types of pairing were made: (1) females carrying a larva and males without a larva; (2) pairings with each sex carrying a larva; (3) females without a larva and males with a larva. Courtship and copulation were carefully observed under a dissecting microscope at 10X when this could be done without disturbing the spiders. After mating or—in some instances—cannibalism, both spiders were examined under CO, anesthesia to ascertain any changes in larval position. Some spiders proved to be subadults and molted only once; other spiders molted twice before reaching maturity. Larval movements on the latter are summarized in Table 19. Movements Table 19. Movement of Mantispa uhleri larvae on immature Phidip- pus audax spiders molting twice to maturity (Exper. 14) Number of Spiders Movement of mantispid larva 2 oe Body —* pedicel — pedicel 5 2 Body — book lung — _ book lung ] 7 Body — book lung — pedicel Z 0 Body — book lung — gone 8 0 Total 16 9 a . . Arrow indicates a molt. Movements of First Instar Mantispid Larvae on Spiders 57 of larvae on P. audax molting only once are found in Table 20 which additionally condenses and compares all data from Experi- ments 14 and 16. A total of 21 egg sacs were obtained from the spiders matured and mated in Experiment 14. The larval move- ments associated with the production of these sacs are summar- ized in Table 21. Experiment 15: Larval behavior on third and fourth instar Phidippus audax P. audax were laboratory-reared to third and fourth instars from eggs obtained from laboratory-matured and mated females. After hatching, spiderlings were kept in 7-dram shell vials covered with a piece of nylon screening and fed Drosophila melanogaster daily until they reached the desired stage. Boarding of each spider by a single first instar larva of Mantispa uhleri was induced in a 2-dram shell vial as in Experiment 14. Larvae were easily visible on these small spiders without CO, anesthesia. After boarding by a larva, spiderlings were returned to their original 7-dram vials and fed Drosophila melanogaster daily. After each ecdysis, spiderlings were examined under CO, anesthesia and any changes in larval position noted. Most spiders were reared successfully through only two addi- tional molts after larvae had boarded. Significantly more larvae (Table 22) entered the book lungs of the more mature spiderlings. Experiment 16: Larval Behavior on Lycosa rabida Several adult females of L. rabida carrying egg sacs or spider- lings were collected in early September at DSAC. These were kept in individual plastic cages (8.5 X 12.5 X 6.0 cm) until the second instar spiderlings, which remain on their mother’s abdomen through this instar, began leaving of their own volition. These spiderlings were isolated singly in the plastic cages just described, modified as follows (Figs. 10 and 11): A 1-cm diameter hole was drilled in the upper right corner in each of the two cage ends and closed with a cork. This permitted food to be added with the plastic top in place. A 1.7-cm diameter hole was drilled in the lower right corner of one side and through this a 2-dram shell vial, filled with water and stoppered with cotton, was ‘(payie1-OM) say, exy IY ISLA RIA Z00'0 < d < 400'0) Stepids ayeutay pue s[euUt Us29MI0q JWU9II}ITP Apuroyiusis Suteaddestp pue 1four ise] 1ayye staprds uo aearey jo Aouanbar, ‘(paytei-om) ‘¢000'0 = 159.L yoexy Jaypsty 1A g) Siapids ayeutay pue seu UsaMj0q JUII9IIIP Anueoyiusis 3uteaddestp pue jour se] 191jye stapids uo aevarey jo Aouanbaly , ‘(patei-OM) “1610 = 1S9L yexY Jaypsty eIA g) Siaprds ayeutay pue s[eUl UsIMIJoq JUIIIF;TP AfuRotytuSis JOU [aIpad uo pue sdun] Yoo ul pauontsod sear] jo Aouanbay p ‘(pattei1-OM) “{ = ISA, WeXy JaySty PIA q) SJaptds ayewiay pue seul UIaMI9q JUIIIIjIP Afjueotytusis JOU [a21ped uo pauontsod pure ssun| yoo Sutiajua aearey jo Aduanbaly | (1 = yp ‘1000 < d < 9000 ‘ecrs= !PeX) saaptds ayeuray Jo sdnoi8 0m) uaamiaq IarayjIp APURITTUSIS [221pad uo pauontsod pure ssun{ yoo SuliajuUa aeAre] JO Auanbaly q (1 = JP ‘1000 >d ‘OF'0Z = fpe 2X) stapids a[euloy pue afeul UsaMI9q IUIIIjjIP Afuesy1usts [a1pad uo pauontsod pure ssun, yoo Sutiajua aearey Jo AQuanbaly p Ne ee ee ee a8 aél I 6 6 Il po pld 0 0 al 266 5 20 Gc 20ne0e tow Uce pl apoc= cf 0 sO JOSE 2 (Q{ ‘todxq) vpiqns “7 18 18 Os 58 4 Bal Boe 5 ; 0 i alll. (p[ ‘todxq) Atimew 30 y4 0 0 G [i e = : 0 G L 2 0} 99IM} SUT}[OW xXDpnD ‘q : : e- 2 e7 e DS wt : bo qge6l qebh 868 (pl wodxy) Aime - - - - - - - - - I ef ice 0] 90U0 SUN[OW xvpnv ‘J SS ee ee ee SS ae ee guoyg) = iapidg peaq 2u0y) [901 sun] [aor Sun, Ajsno 92u0%d Jeor Sun, xag satvads rapids uO -Pad YOOY -PId ACO -tAaid -Ped ACCP sutAp /pe9d [OUI Se] yOu J[OU Ise] O} J[OUI ISITJ jaye Sun, yooq \Se] 19Ije UOTITSOd roud ysn{ uontsog jaye UONISOg WO} [BATAING ee —————————————— deAIe] JO JaquinN eee eee (g] pue p] ‘tadxq) ssopids vpiqvs vsorkT pur xvpnv sndqipiyd Uo asealey 14a;YyN vgsiyUvpy JO IW IAC = *0Z FATGBL Movements of First Instar Mantispid Larvae on Spiders 59 Table 21. Mantispa uhleri larva penetration of Phidippus audax egg sacs (Exper. 14) Position of larva Number of spiders * Number of larvae after prior to oviposition spinning egg sacs egg sac production by spider In sac Dead Missing On pedicel 17 10 2 5 Elsewhere ] 0 ] 0 In book lung 3 2 ] 0 *One mantispid larva per spider. Table 22. Movement of Mantispa uhleri larvae associated with the first molt on immature Phidippus audax spiderlings (Exper. 15) Number of spiderlings Group I: Group II: Larval movement Spiderlings molting Spiderlings molting associated with molt from 3rd to 4th instar from 4th to 5th instar Body — * pedicel 17> 136 Body — book lung 8b 43b Body — gone ] 10 Total 26 66 * Arrow indicates a molt. > Numbers of larvae traveling to the pedicel and to book lungs significantly different in the two groups (X? aq; = 13.005, P < 0.005, df = 1). inserted to provide spiderlings with constant access to free water. As water evaporated, the cotton pledget was drawn into the vial and continued to provide a wet surface. A 2.7-cm diameter hole on the opposite side of the cage, screened with nylon mesh, provided ventilation. In the bottom of each cage was a small 4- dram vial containing Carolina Biological Supply Instant Droso- phila Medium®. A culture of Drosophila was thus established in each cage to provide a constant source of food for the second instar spiderlings. Third instar spiderlings received the same food as the preceding stage. Drosophila medium was removed from the cage at the beginning of the spiders’ fourth stadium and first instar soybean loopers, Pseudoplusia includens (Walker), and first instar crickets Acheta domestica Linnaeus, were sub- stituted as food. Fifth and sixth instar spiderlings were fed in the same way but with progressively larger loopers and crickets, 60 Mantispa uhleri Banks while sixth instar spiderlings were also given house flies, Musca domestica, if accepted. The horizontal water vial was removed when spiders reached the sixth instar and were too large to crawl inside it; the opening thus created was stoppered with a cork. Water was provided for the larger spiders by means of a moistened cotton pledget on a small plastic disc. Beginning with the seventh instar, only house flies were offered as food, and sugar cubes were added to each cage to prolong the life of uneaten flies. This last arrangement remained unchanged until the experiment was complete. As far as possible, spiderlings were reared to the adult stage and were transferred to a clean cage after each ecdysis. First instar Mantispa uhleri were allowed to board the spider- lings when these had reached the fifth or sixth instar. This was accomplished as detailed in Experiment 14 (p. 55). When boarding had been accomplished, the position of the larva was recorded and the spider returned to its plastic rearing cage. After each subsequent ecdysis, spiders were examined under CO, anesthesia and any change in larval position recorded. A group of control spiders, never exposed to a mantispid and treated identically, including examination under CO, anesthesia, were simultaneously reared. Approximately 2 weeks after maturing, spiders were paired in a shallow enameled pan (40 X 75 cm) and observations of the ensuing mating or, in some instances, cannibalism, were made. Each male or female spider carrying a larva was paired with a laboratory-reared spider of the opposite sex from the control group. Both spiders were examined under CO, for any larval movement after mating. In instances of cannibalism, the remain- ing spider was examined after it had consumed its partner. Some males carrying larvae were mated more than once, and several were paired with previously mated females to induce cannibalism by the female. A total of 56 spiders were successfully matured out of 61 originally boarded by larvae. Larval movements noted on these spiders after each ecdysis are summarized in Table 23, while Figure 12 indicates the ultimate fate of these larvae. In the control group, 40 of 44 spiders were successfully matured. Movements of First Instar Mantispid Larvae on Spiders 61 ‘vial with water and cotton pledget § screened hole Z \ for ventilation — water-soaked cotton on plastic disc sugar cubes for Musca Fig. 11. Rearing cage for Lycosa rabida for sixth and subsequent instars. 62 Mantispa uhleri Banks 6/ Larvae on Spiders 5 Spiders Dying | During Development 32_on Males a Fon Females 27 Book Lung /1 BookLung 2 Pedicel 9 Gone 2 Dead 2 Pedice/ 2 Gone Dying before Sacs Spun / Spider 2 Spiders 2 Dying 3 Sacs Spun— 3 Sacs Spun— Accidentally Producing before Larvae Failing Larvae Entering killed No Sacs Sacs Spun To Enter | 3 Adult Moantispa uhleri Fig. 12. Fate of Mantispa uhleri larvae on developing Lycosa rabida. Table 23. Movement of Mantispa uhleri larvae on Lycosa rabida spiders reared from fifth or sixth instar to maturity Number of spiders Mantispid larva movements = 9 through several spider molts 4 i Xe) ~I Pedicel— book lung +> book lung — book lung Pedicel— book lung--> book lung — gone Pedicel— book lung—book lung— gone — gone Pedicel— pedicel— gone— gone— gone Pedicel— book lung— book lung— pedicel Pedicel— pedicel book lung— pedicel Pedicel— book lung— pedicel— book lung Pedicel— book lung— pedicel— gone Pedicel— book lung— dead Pedicel— dead— > dead Totals lel et et et COs SE Oo °° 9° 6 =} = N SS iS) nN ihe) ee 4 Arrow signifies a spider molt. Spiders were generally mature at the ninth or tenth instar, so that larvae experienced three to five spider molts. Terminal spider instar was always an adult. > These arrows equal one to four spider molts. All other arrows equal only one molt. Movements of First Instar Mantispid Larvae on Spiders 63 Tables 20, 24, and 25 summarize data from all or some combination of the three experiments. The sites occupied by larvae after first boarding spiders in all three experiments are specified in Table 24. Table 20 catalogues larval movements during spider development in Experiments 14 and 16. A total of 32 male Phidippus audax and 34 male Lycosa rabida carrying larvae were paired with females. Larval movements during the Table 24. Initial locations of Mantispa uhleri larvae after boarding Phidippus audax and Lycosa rabida spiders Number of larvae boarding spiders Location of 3rd and 4th 5th and greater 5th and 6th M. uhleri larva instar instar instar on spider P. audax P. audax L. rabida Pedicel 110 84 54 Under edge of carapace 8 13 5 Between or around coxae 7 4 pd Between sternum and leg bases 2 10 0 Spinnerets 0 12 0 Legs 0 8 0 Totals 127 131 61 Table 25. Transfer of Mantispa uhleri larvae from male to female spiders during mating and cannibalism Number of mantispid larvae At spider mating During cannibalism of o& spider by @ Spider species Transfer _No transfer Transfer No transfer Phidippus 0 20 in book lung 5 from book lung 1 in book lung audax 4 on pedicel 1 from pedicel 1 on pedicel Total 24 6 2 Lycosa rabida 0 20in book lung 2 from book lung 12 in book lung 64 Mantispa uhleri Banks mating and cannibalism associated with these pairings are summarized in Table 25. DISCUSSION As noted previously, the three experiments reported here have certain features in common and their separate data have a strong interrelationship. Accordingly, we shall discuss our results under four general headings, rather than deal with each experiment separately. Positions First Adopted by Larvae after Boarding It is clear from all of the experiments (Tables 19, 20, 22, and 23) that mantisipid larvae are capable of negotiating a spider molt. Since the phenomenon has never been reported before, we were surprised to find that in many instances it is associated with entrance into one of the spider’s book lungs. This is, however, never the first location after boarding. The pedicel is the usual site (Table 23), and the larva positions itself anywhere around the exposed circumference of this structure (Fig. 13). Further- more, on particularly large spiders, a larva may be found in one of the two “‘pits’” formed laterodorsally by the telescoping of the pedicel into the base of the abdomen. In such cases the larva may be completely hidden from view until the abdomen is pulled back to reveal these pockets. We have observed spiders pick larvae off their bodies during grooming movements of their legs, after which they eat the larvae. In contrast, the pedicel as the resting site for a larva has the advantage of being relatively inaccessible to the spider. Such a location also provides a larva with easy access to the next instar at ecdysis because the spider’s old exoskeleton splits laterally along the pedicel to expose the new cuticle only a short distance away. Finally, the pedicel is membranous, as are all other areas that a larva might immediately occupy after boarding. Since they often feed on the spider’s blood prior to the production of an egg sac (Redborg and MacLeod, 1983b), larvae must find tissue thin enough for their mouthparts to penetrate. Movements of First Instar Mantispid Larvae on Spiders 65 f dorsal cephalo- & thorax Fig. 13. First instar larva of Mantispa uhleri on pedicel of Lycosa rabida. Arrow indicates position of larva. Entry of Larvae into Book Lungs Movement of larvae into the book lungs of spiders was first discovered when we studied larvae on Phidippus audax spiders that were undergoing their final molt to the adult (Experiment 14). The findings (Table 20) revealed a striking difference in larval movements during molting, with respect to the two spider sexes; mantispid larvae entered the book lung of significantly more males than females. We were first inclined to explain this seemingly preferential behavior on the basis of the close proxi- mity of the male’s genital opening to the book lungs and the possibility that in such a location a larva might subsequently migrate to the palp at the time of sperm charging and be transferred to the female’s abdomen during copulation. This expectation was not borne out in our analysis of a more general survey of larval movements on immature spiders during two or more molts (Tables 19 and 23). After boarding immature females of P. audax (which needed two molts to reach maturity), larvae entered the book lungs at the first molt in significantly 66 Mantispa uhleri Banks larger numbers (Table 20) than did those larvae that had boarded subadult females and experienced only one molt. This demon- strated that larvae did enter the book lungs of female spiders prior to sexual maturity and suggested that larvae made no distinction between males and immature females, responding differently only to adult females. To study the book-lung-entering behavior further, we had intended to rear the third and fourth instar P. audax of Experi- ment 15 to the adult stage and to follow the movements of the larvae. Unfortunately, nutritional difficulties prevented most of the spiders from passing through more than one or two molts, and entrance into the book lungs could not be studied further until these difficulties were overcome (Experiment 16). In this experiment, virtually all mantispid larvae on both sexes of immature Lycosa rabida spiders entered a book lung at the first spider molt (Table 23); frequencies of book lung entry were not significantly different for larvae on pre-adult male and female spiders (Table 20). Notwithstanding our failure to rear them to the adult stage, the spiders of Experiment 15 did yield some information con- cerning the relationship of the size of the spider to the time that book-lung-entering behavior is first observed. At the first molt experienced by the larvae, significantly more of them entered the book lungs of spiders molting from the fourth to the fifth instar (Table 22—Group II), than of spiders molting from the third to the fourth instar (Group I). But several of the Group I larvae that were ultimately counted as having gone to the pedicel were seen attempting to enter a book lung immediately following the spider’s ecdysis. In both groups of spiders many of the larvae that did enter a book lung had legs or abdomens protruding from the lung slit, suggesting that the book lungs of some of these spiders were simply too small to accommodate a larva. The increased number of book lung entries observed during the ecdysis from the fourth to fifth instar is likely due to the increase in the size of the book lung after this molt. In all three of our experiments, larvae entered a book lung only in association with a spider molt. The reasons for this are unknown, but conceivably attempts by a larva to enter a book lung during an intermolt period may alert the spider, giving it an opportunity to dislodge its would-be parasite. However, entry Movements of First Instar Mantispid Larvae on Spiders 67 is probably comparatively easy immediately following ecdysis while the spider’s movements are restricted during the tanning of its new cuticle. It is also possible that the sclerotized margins of the book lung cannot be pried apart as successfully as they can immediately after ecdysis. Upon first entering the book lung of L. rabida (Fig. 14), larvae entered the right and left book lungs with equal frequency (26 left, 31 right, X2 = 0.439, 0.9 >P > 0.5, df =1). Larvae generally remained through subsequent molts in the book lung originally entered. Thus, it is likely that a larva located in a book lung can remain there during a spider’s molt, since, if it did leave, there is only a 50% probability that it would re-enter the same book lung. Of 117 instances in which larvae remained in book lungs, 106 occupied the book lung in which they had been located just prior to ecdysis. This result is significantly different (X? 4) = 75.521, P< 0.001, df=1) from the frequency distribution expected if larvae were leaving their respective book lungs and re-entering either the right or left book lungs at random. The few instances where larvae switched book lungs were usually coxa of fourth leg “s : . . Se es \ ‘ i Fig. 14. First instar larva of Mantispa uhleri in book lung of Lycosa rabida. Arrow indicates position of larva which is barely visible. 68 Mantispa uhleri Banks correlated with observations of damage (swelling, discoloration, congealed blood) to the book lung previously occupied. These larvae may have been forced to leave the resident book lung during ecdysis because of this damage, or they may have deliberately moved to the more habitable book lung. Compared to a location on the pedicel of the host spider, a book lung would seem to offer several advantages to a larva awaiting the spider’s sexual maturity. Since gas exchange occurs across the book lung membrane, its cuticle should be thinner and thus more easily penetrated by the jaws of a feeding larva. Also, once inside the book lung, the larva runs no risk of being dislodged or damaged by the spider’s potentially abrasive groom- ing movements. Finally, the elevated humidity probably associated with the book lung may benefit the larva. Larval Movements Associated with the Adult Molt and with Egg-Sac Entry Our data have suggested that, without regard to the spider’s sex, a larva enters the book lung of an immature spider at the earliest molt which produces a sufficiently large book lung. Thus, just prior to their final ecdysis, 50 of the 52 surviving L. rabida of Experiment 16 contained larvae in their book lungs, with no significant difference in this regard between male and female spiders (Table 20). Most of these larvae had successfully negotiated two, or more, spider molts in the book lung and there had been only a very low incidence (2 out of 81) of either death or disapperance. We were, therefore, puzzled to observe a much higher rate of disappearance of larvae associated with the final molt of female spiders. At this time only 12 of the 21 larvae located in a book lung immediately prior to ecdysis were to be found on the spider after the molt (10 of these remained in a book lung, while 2 successfully moved out onto the pedicel—see Table 23). One of the 9 larvae that had previously resided in the book lung died in the process of this molt, but the remaining 8 could no longer be found on their spiders. In contrast, all 29 larvae that had been located on subadult male spiders were found alive on the spider after the molt, 27 of these remaining in a book lung. The Movements of First Instar Mantispid Larvae on Spiders 69 number of book lung larvae that succeeded in remaining on their spiders at the final molt varied significantly between male and female spiders (Table 20). This differential behavior with respect to the two spider sexes at the final molt is not restricted to L. rabida. A comparison of the movements of larvae on males and females of Phidippus audax, which molted twice to maturity in Experiment 14 (Table 20, lines 3 and 4), showed notable differences also. In this case all 7 larvae aboard males survived, compared to only 3 out of 11 on females. Some light is shed on the fate of the larvae missing from the adult female spiders by the coincidence that one of the spiders carrying a larva (Experiment 16) had been confined in a vial while molting. Within an hour after ecdysis we observed the larva crawling around the bottom of the vial, suggesting that for some reason this larva, and presumably the others, had not been able to remain on their spider through the final ecdysis. It seems unlikely that simply leaving a female spider at this time is part of a larva’s normal developmental behavior since, under natural conditions, the larva—once having left the spider— would have little chance of reboarding because the spider will shortly move away. Additional factors are undoubtedly at work in this process. The possibility that adult spiders actively try to rid themselves of larvae, as well as the possibility that larvae may inefficiently shift from the book lungs to some other location more favorable for entry into an egg sac are considerations to be explored. In view of the damage to a book lung caused by a resident mantispid larva, as well as the lowered fitness of female spiders whose egg sacs are successfully entered by these same larvae, it would not be surprising to find that spiders have evolved mechanisms for ridding themselves of mantispid larvae. It is possible, therefore, that a larva attempting to remain in a book lung that has optimal access to an egg sac 1s effectively foiled by some structure or behavior of female spiders. For several reasons, this does not seem likely to us. That it is not inherently difficult for a larva to negotiate a pre-adult spider molt in this location is demonstrated by our observation (e.g., in Experiment 16) that during 79 of 81 spider molts from one immature instar to another, larvae in book lungs successfully 70 Mantispa uhleri Banks remained in situ. It is only in the final molt of female spiders that such a purging mechanism seems to operate, if indeed it exists at all: yet such riddance should also be of advantage to a spider at earlier stages when larvae are feeding as ectoparasites and causing damage to the book lung’s delicate structure. Such a mechanism should also be of advantage to male spiders, yet the loss of larvae is restricted to females. Further, since similar results were obtained with two species of spiders belonging to different superfamilies, it seems unlikely that both P. audax and Lycosa rabida would have independently evolved a similar ant'- mantispid mechanism. Finally, the notion that larvae seek to remain in the book lungs of female spiders because this is the most advantageous location from which to enter egg sacs is inconsistent with our findings. With Phidippus audax spiders that were undergoing a single molt to sexual maturity, boarding larvae had the opportunity to enter female book lungs but failed to do so in 19 of 23 cases (Experiment 14 and Table 20, lines 1 and 2). The second possibility seems more reasonable, although it does not account for all of our results. We suggest that larvae leave the book lungs in order to station themselves on an exposed surface (probably the spider’s pedicel) so that they can more easily enter egg sacs; we suggest that a gain in the facility with which larvae might enter egg sacs from the spider’s pedicel compared to a position in a book lung more than offsets the ineptness with which the book lungs are abandoned. There are four maneuvers that a mantispid larva may perform during the spider’s ecdysis: the larva may (1) remain on the pedicel (or sometimes move from another membranous area to the pedicel); (2) move from the pedicel (or elsewhere) into a book lung; (3) remain in a book lung; (4) move from a book lung to the pedicel. From what we have seen of these four activities, we postulate that leaving a book lung is an intrinsically difficult maneuver. During a spider’s ecdysis, the most serious obstacle to the larva is probably the barrier imposed by the old cuticle, since a larva located on a remote area of the exuvia may, after a molt, take too long to get back onto the spider. Since the tanning spider may remain only briefly in contact with its exuvia, a dislocated larva might possibly lose all contact with the spider. In the first two Movements of First Instar Mantispid Larvae on Spiders 71 maneuvers noted above, this danger is minimal, since to reach the cuticle of the next instar a larva need only move into the nearby ecdysial tear that runs laterally down the pedicel. The larva can then remain on the pedicel or enter a book lung while the outer margins of the latter are still soft and untanned. Our data show that such larvae, positioned on the pedicel, survive a spider’s ecdysis with a high rate of success. In the third situation, a book lung larva need only push against the outward pull of the thin cuticle withdrawn from the book lung during ecdysis in order to tear through it and remain in the same book lung of the next instar. Again, our data indicate that larvae are also adept in this maneuver. A larva that attempts to leave a book lung during a molt, however, must remain in the book lung long enough to tear through the old cuticle as it is withdrawn from the lung slit, for if the larva leaves too soon it risks being stranded on the exuvia some distance from the spider’s new cuticle and may fail to regain the spider. This may have been the fate of the disappear- ing larvae of Experiments 14 and 16. But a book lung larva that has successfully pierced the cuticle that is being pulled out of the book lung slit must now leave quickly lest it be trapped in the book lung. Just as in entering, leaving a book lung can possibly be done only while the soft cuticle renders the spider immobile. If the cuticle becomes sclerotized before the larva can leave, it may stay in the book lung rather than risk being eaten by the spider aroused by the larva’s attempt to escape. That the spider pedicel is the objective of larvae exiting the book lungs is suggested by the movements of larvae on P. audax (Experiment 14) and by the high success rate of larvae in entering egg sacs from this position. Of the 23 larvae on females of P. audax that had reached maturity in one molt—so that the larvae did not have to contend with the postulated difficulties of a book lung location—19 of these positioned themselves on the pedicel rather than entering a book lung (Table 20, line 2). Further, as summarized in Table 21, 10 of 17 larvae stationed on the pedicel of this species successfully entered egg sacs, along with 2 of 3 larvae located in a book lung. The principal difficulty with this explanation is in under- standing why mantispid larvae that have entered the book lungs 72 Mantispa uhleri Banks of subadult females leave this position if their rate of success in remaining on the spider is so low. This is particularly puzzling since the data just cited, as well as those of Figure 14 pertaining to Lycosa rabida, show that larvae that have remained in book lungs are also successful in entering egg sacs. Our sample size is not large enough to compare, quantitatively, the success rate of egg sac entry from these two locations. Possibly, additional studies with a larger number of spiders would prove valid the posited advantages of the pedicel over the book lung, and would demonstrate an improved ability to reach sacs from this site which offsets the high rate of dislodgment from the spider that larvae experience when they exit book lungs. Studies of larval movements during the spider’s final instars, using the procedures developed in the experiments described here and with additional species of spiders, are needed in order to observe more closely the details of such behavior. Transfer of Larvae from One Spider to Another No Mantispa uhleri larva attempted to transfer from male to female during our observations of 44 spider matings of Phidippus audax and Lycosa rabida (Table 25). In the pairings of Phidippus audax, some of the females were carrying a larva; transfer behavior of the larva on the male may thus have been inhibited. But since for all Lycosa rabida matings the females were free of larvae, inhibited transfer behavior seems unlikely. Several L. rabida males were mated more than once to test the possibility that the first mating may convey some necessary preliminary information to the larva, but all of these results were also negative. Since our laboratory conditions were conducive to the spider’s epigamic behavior and successful copulation, it seems a reason- able inference that they would also be suitable for larval transfer, if this should occur. We thus conclude that, although one might have expected selective pressure for such behavior, Mantispa uhleri larvae seldom, if at all, transfer from male to female spiders during mating. Although this conclusion might seem counterintuitive, it emphasizes that the theoretical advantage of some heritable trait, deduced by a priori considerations, does not Movements of First Instar Mantispid Larvae on Spiders 13 ensure that natural selection will necessarily have produced such a trait. Although the act of spider egg-laying may be similar in most species, enabling mantispid larvae to enter almost any egg sac, epigamic behavior, because of its role as an isolating mechanism, may vary in detail with different kinds of spiders. Therefore, the utilization of a number of spider species by M. uhleri may be incompatible with the evolution of behavior that allows an efficient larva transfer from male to female spider at mating. We argue this since the ability of a larva to synchronize its movements to the details of a particular spider’s mating behavior, permitting a transfer from male to female, would be of no advantage to the offspring of that larva unless the same species of spider were often boarded. Consideration of the very large number of spider species naturally utilized by M. uhleri makes this latter circumstance seem doubtful. Nevertheless, larvae boarding male spiders are not necessarily destined to die without reaching maturity. When certain of the Phidippus audax males (Table 25) eventually died, larvae were observed to leave the corpse and, presumably, could resume search activity. In nature such larvae may have an enhanced survival and search time because they had fed on spider blood. Also of potential importance to larvae located on male spiders is a larva’s ability to transfer to a new spider during cannibalism. Several larvae in our experiments thus moved from a male to a female spider (Table 25). This action has the advantage of transferring a larva from any spider to any other regardless of spider sex, state of maturity, or even species. Just such an opportunity was observed by one of us (K. E. R.) under field conditions with the collection of two adult females of Meta- phidippus galathea (Walckenaer), one being consumed by the other. The prey spider had a larva of Mantispa uhleri on its pedicel. Although this encounter was permanently disturbed when the spiders were brought into the laboratory and examined under CO, anesthesia, the potential advantage for a transfer of the larva to the predator spider existed. More knowledge of how frequently spiders prey on other spiders under natural conditions is needed for a more accurate evaluation of this behavior as it pertains to M. uhleri. 7. Species of Spiders Utilized Most references briefly note the larval association of mantispines with such general statements as “‘[they] are parasitic in the egg sacs of ground spiders” (Borror and Delong, 1971), in a persistent echo of the fact that the first published account of an adult emergence recorded that it had developed in the egg sac of a lycosid. Except for occasional notes documenting the rearing of a particular mantispid species from the egg sac of a particular spider, virtually nothing has been published about mantispid- spider associations, including the important details of how wide a range of spider species a given mantispid may utilize. The fact that mantispid species often occur sympatrically suggests they may in some way be apportioning available prey spiders among themselves. As a result, different mantispid species might use spiders of restricted taxonomic groups or might concentrate on spiders found in certain specific habitats. One obvious way to evaluate these possibilities is to collect the data that link adult mantispids with the egg sacs of field- collected female spiders. A second, quite effective method can be used with those mantispid species whose first instar larvae board spiders. Here, with the assumption that the presence of the larva on the spider indicates an eventual egg sac entry, idenufication of spider and larva establishes the crucial association. We have used this latter approach in surveying the range of spider species utilized by M. uhleri: in our search to find associated larvae, we studied a large, identified collection of spiders, as well as collections we ourselves made in areas where M. uhleri was commonly found. We advance the hypothesis that M. uhleri depends on a number of species of hunting spiders, but rarely if ever boards, or 74 Species of Spiders Utilized 75 eats the eggs of, web-building spiders. Within the broad designa- tion “‘hunting spider,’’ however, this mantispid appeared not to be restricted to particular taxa. In pursuit of such evidence, we also collected field data on the frequency of larval infestation of the hunting species Metacyrba undata and three nonhunting spiders, Arzadna bicolor, and two species of Agelenopsis, occur- ring in the same habitat as Metacyrba. METHODS AND RESULTS To obtain a relatively unbiased view of the variety of spiders used by larvae of Mantispa uhleri, we studied the Ilinois spider collection of Southern Illino University at Carbondale (SIUC). Assembled by Dr. Joseph A. Beatty and his students, the collection is the record of an intensive sampling of the spider fauna of southern I]linois where M. uhleri is common, and is intended to be as nearly complete a representation of the species in this area as possible. Where species presumed to occur in the region were lacking, special efforts were made to obtain specimens. The magnitude of this effort is apparent in the fact that more than one hundred species have been added to the state list (Beatty and Nelson, 1979). Most important, no spiders were collected on the supposition that they might bear mantispid larvae. Under a dissecting microscope, and in a Syracuse dish contain- ing alcohol, the entire surface of each spider was carefully scrutinized, the abdomen being retracted to reveal the pedicel. Both book lungs were gently opened with forceps. When larvae were found, they were removed and mounted on slides. The second author’s (E. G. M.’s) unpublished key to first instar larvae was used in identification. Any egg sacs attended by spiders and collected with them were also opened and examined for mantis- pid larvae. Additional records were gathered by the first author (K. E. R.) from spiders collected at the University of Illinois Dixon Springs Agricultural Center (DSAC) in Pope County and a few other Illinois locales. Over the summer many spiders were collected during the day by sweeping vegetation and examining foliage; at night, lycosids, and occasionally other spider families, were 76 Mantispa uhleri Banks collected with a head lamp. During the winter months, spiders were gathered from beneath the loose bark of trees and from the leaf litter. Living spiders were examined under CO, anesthesia. Larvae were identified either by using the key just mentioned or by rearing them to the adult stage. Spider specimens totalling 5,761 were examined from the SIUC collection, and the species and number of individuals were listed in a sequence of families recommended by J. A. Beatty. Sixteen larvae of M. uhleri were found, all of which were associated with hunting spiders of the Lycosoidea and Clubio- noidea (see Appendix I [p. 101] and Table 26). Appendix II [p. 117] summarizes fundamental data for all species of spiders which we have found to be utilized by M. uhleri, and includes larvae from the SIUC collection (extracted from Table 26) as well as records of 110 additional larvae which we obtained from spiders collected at DSAC and other locations. The arrangement is alphabetical by family, genus, and species. For each species is listed information on all larvae removed from spiders of that species including the sex and maturity of the spider, location of larva on spider, the site where collected, and the date when acquired. No locale is specified if the spider was part of our supplemental collecting at DSAC. If more than one larva was found on a single spider this fact is noted. The total larval associations involved 31 species of spiders distributed in 22 genera (Table 27), and represents all but two of the families of the superfamilies of hunting spiders noted above. During the winters of 1974-75 and 1975-76 a survey of the occurrence of larval M. uhleri on the spiders Metacyrba undata and Ariadna bicolor was conducted. These spiders were collected from silk retreats beneath the bark of shagbark hickory, Carya ovata K. Koch, from ground level to a height of approximately 10 feet. The woods surrounding DSAC were entered from several arbitrarily chosen points and all shagbark hickories encountered were examined for loose bark. All spiders that could be located on each suitable tree were collected and brought into the laboratory where, under CO, anesthesia, they were examined for larvae. In 1976 a similar survey was made of the number of larvae on two agelenids, Agelenopsis kastoni Chamberlin and Ivie and A. Species of Spiders Utilized Fig | Table 26. Summary of Mantispa uhleri Banks-spider associations from the examination of the SIUC Collection Spider genera Spider species Total spiders Spider suborder, Larvae Larvae Larvae superfamily, family Examined present Examined present Examined present MYGALOMORPHAE ATYPOIDEA Antrodiaetidae Z ) 2 0 68 0 CTENIZOIDEA Ctenizidae ] 0 ] 0 4 0 ARANEOMORPHAE DICTYNOIDEA Amaurobiidae ? 0 3 0 12 0 Dictynidae 2 0 9 0 218 0 Oecobiidae ] 0 ] 0 2 0 Uloboridae ] 0 ] 0 10 0 DYSDEROIDEA Dysderidae 2 0 2 0 7 0 SCYTODOIDEA Scytodidae 2 0 2 0 48 0 ARANEOIDEA Pholcidae 2 0 2 0 22 0 Theridiidae 17 0 36 0 810 0 Linyphiidae 26 0 40 0 515 0 Araneidae 23 0 44 0. 1,141 0 Symphytognath- idae Z 0 2 0 2 0 Mimetidae 2 0 3 0 18 0 LYCOSOIDEA Agelenidae 4 0 1] 0 248 0 Hahniidae 2 0 3 0 21 0 Pisauridae 2 ] 8 ] 76 ] Lycosidae 7 0 24 0 614 0 Oxyopidae ] 0 3 0 289 0 CLUBIONOIDEA Gnaphosidae 10 0 21 0 108 0 Clubionidae 7 2 24 2 350 2 Anyphaenidae 3 ] 7 l 118 ] Thomisidae 1] 3 a2 3 406 3 Salticidae 21 4 38 5 654 9 78 Mantispa uhleri Banks Table 27. Taxonomic summary of all spider species utilized by Mantispa uhler Spider genera examined Spider species examined Spider Superfamily, Larvae Larvae Family Total present Total present LYCOSOIDEA Agelenidae 4 ] 1] ] Hahniidae 2 0 3 0 Pisauridae 2 2 8 2 Lycosidae 7 2 me 4 Oxyopidae ] 0 3 0 CLUBIONOIDEA Gnaphosidae 10 ] 19 ] Clubionidae di Z 24 2 Anyphaenidae 3 ] 7 2 Thomisidae iy! 5 ae 6 Salticidae 21 8 38 13 Totals 68 22 169 31 emertoni Chamberlin and Ivie. Although all of the adult speci- mens collected belonged to these two species, many specimens were immature and could not be identified with certainty. Therefore, all specimens are hereafter referred to as Agelenopsis spp. These collections were made in April, after the weather had moderated sufficiently to allow the spiders to move from their overwintering sites and construct funnel webs in the leaf litter and around shrubs and fallen branches. The areas searched were generally the same as where Metacyrba and Ariadna had been collected during the preceding two winters. All funnel webs found were examined and spiders occupying them were coaxed into collecting vials and preserved in 70% ethyl alcohol. These spiders were examined for larvae as described previously for specimens in alcohol. Incidence of larvae on these spiders is recorded in Table 28; analysis reveals almost no larvae from Ariadna bicolor and Agelenopsis spp., while 8.5% of Metacyrba undata were infested with mantispid larvae. Species of Spiders Utilized 79 Table 28. Number of first instar larvae of Mantispa uhleri Banks found on overwintering spiders collected at Dixon Springs Agricultural Center Total Specimens Specimens specimens without with Spider collected examined larvae larvae Metacyrba undata 484 443 4] Ariadna bicolor 148 148 Agelenopsis spp. 216 215 ] DISCUSSION Very early in the course of collecting the spider specimens it became apparent that, in nature, larvae of Mantispa uhleri were boarding a wide variety of hunting spiders. During this period web-building spiders were also collected, but as no mantispid larvae were found on them, collecting unavoidably concentrated on those groups of spiders that harbored larvae. Thus, we realized that to single out effectively the major groups of spiders used by M. uhleri it would be necessary to obtain an impartial sampling of spiders. The SIUC collection fulfilled this require- ment and permitted us to test the hypothesis that, in nature, searching first instar larvae board primarily, or solely, spiders belonging to the non-web-spinning, hunting groups. We found 16 larvae of M. whlerz, 13 on spiders, 3 within egg sacs (out of a total of 35 sacs) attended by spiders (Table 26 and Appendix I), and, as noted, these were associated solely with the hunting groups (Table 26, superfamilies Lycosoidea and Clubio- noidea). A formal statistical analysis of these data was inappro- priate since, although gathered at random with respect to the presence of mantispid larvae, the collecting effort was not neces- sarily standardized with respect to the species sampled or to the time of year. The complete absence of larvae associated with the 2,987 specimens of nonhunting species (Table 26, Families Antrodiaetidae through Mimetidae) is impressive and, we strong- ly feel, consistent with our hypothesis. The extensive additional data on hunting species collected at DSAC (App. II) and the absence of larvae on nonhunting groups collected from that locale support our generalization. That members of nonhunting groups of spiders may occasion- ally be utilized by M. uhleri is suggested by some of our 80 Mantispa uhleri Banks laboratory observations. For instance, females of even a web builder like Achaearanea tepidariorum are readily boarded if confined in a vial with a larva and, by means of a cotton plug, restricted to a small space so that they cannot suspend themselves within a web. Likewise, although none of the 148 Ariadna bicolor of the DSAC field sample carried a larva, even though collected from the same microhabitat that yielded abundant larvae-bearing Metacyrba undata, in the laboratory this spider species is boarded with ease when restrained from spinning a retreat. The single larva taken from our substantial series of Agelenopsis spp. compared to the large number taken from Metacyrba undata also suggests that Mantispa uhleri’s failure to utilize nonhunting spiders is related more to the infrequency of direct contact than to avoidance of them. The data of Appendix II also show that, at least insofar as sexually mature spiders are concerned, both males and females are boarded in nature (48 male spiders bearing larvae, 36 females). This corroborates an inference from our earlier experi- mental findings (Table 16) that mantispid larvae do board male spiders under natural conditions. Many of these larvae were found in the book lungs of both male and female spiders. This is again consistent with our laboratory findings (Table 22)— particularly our contention that entrance into the book lungs is not associated solely with male spiders. In our field sample, however, for a number of the larvae-bearing species, the adults are probably too small for larvae to enter the book lungs. Our findings, related to the preferential residence in the book lungs by mantispid larvae during the development of immature and adult male spiders, are thus appropriate only with the larger spiders such as those we used in our experiments. Mantispid larvae are generally thought to be rare. Indeed, even after observing numerous larval boardings in the laboratory, we pessimistically felt that field confirmation of this behavior might be like looking for the proverbial needle. Fortunately, we found that larvae were quite common, at least at Dixon Springs (Table 28). One of twelve overwintering Metacyrba undata had been boarded by a larva, which suggests that mantispids prey on the eggs of some spider species much more commonly than has been previously supposed. 8. The Mantispid Seasonal Cycle Our investigation of this predator-prey relationship has demon- strated what 1s probably a complex seasonal cycle and points to the fact that Mantispa uhleri relies on a large number of spider species to govern—by means of their own biological time- clocks—the timing of mantispid seasonal development. A general model for M. uhleri’s seasonal cycle can be developed and related to results of two years of seasonal distribution records of larvae and adults derived from collections made in southern Illinois. Supplementary information collected in other Illinois locales has also been incorporated. METHODS AND RESULTS Adult M. uhleri were collected in two light traps, which were similar to one illustrated by Metcalf and Luckmann (1975, p. 329) and were set up about 10 yards from the edge of extensive woodlands in the Shawnee National Forest at DSAC in southern Illinois. The traps were situated about 50 yards apart and each trap consisted of four 15-watt ultraviolet fluorescent tubes placed above a wide-mouthed metal funnel leading into a screened cage approximately 2 feet on each side. A wooden roof above the lights prevented rain from entering the traps. Metal flanges on both sides of the lights aided in the capture of flying insects. The distance from the lights to the ground was approximately 10 feet. In theory, insects attracted by ultraviolet radiation hit the lights or flanges and fell through the funnel into the box below. Frequently, dry ice, insulated and covered by crumpled news- papers, was placed in the bottom of a waste basket-sized con- 81 82 Mantispa uhleri Banks tainer immediately below the funnel, so that insects passing through the funnel were anesthetized by the sublimating carbon dioxide. This reduced mutilation of mantispids by crawling beetles and fluttering moths and left most of them in an apparently undamaged condition when released the following morning. With the exceptions noted later and in Figures 15 and 16, the traps were operated every night from mid May 1974 to mid April 1976. Traps were checked in the early morning and the number and sex of any M. uhleri were recorded. Except for adults that were kept for rearing purposes, undamaged mantispids were marked and released next to the traps the morning after their capture, in the hope that recapture would give some indication of their longevity under natural conditions. Small spots of colored fingernail polish were placed on the undersurfaces of the wings in a characteristic pattern so that individual insects could be identified if re-collected. Temperature data were recorded for the periods of collection by means of a hygrothermograph located near the traps. Figures 15 and 16 record the number of M. uhleri collected during 1974 and 1975, respectively. High and low temperatures for each collecting night are indicated in both figures. The highest temperature usually occurred at sunset and the lowest at sunrise. Approximately equal numbers of male and female mantispids were collected; there was no indication of a male- biased sex ratio such as that observed by New and Haddow (1973) from their light-trap collections of mantispids at Entebbe, Uganda, in 1961 and 1962. In order to evaluate the severity of the winters preceding our collections, the average weekly tempera- tures were plotted for December-April of 1973-74 and 1974-75 (Fig. 17). During the summer of 1974, eggs were obtained on 6 July from the earliest trap-collected female M. uhleri that produced fertile eggs in captivity. Larvae from these eggs were reared in an unheated screened insectary at DSAC so that the generation time under field conditions could be estimated and an indication obtained as to when to expect the earliest possible emergence of adults of a second summer generation. The first adults from these eggs eclosed on 12 August (Fig. 15), representing a field- ‘g[eWay SUNISOCIAO IS11j] JO sddo wio1] AreqISUT IOOPINO UL patel SI|NPe ISI © ‘sden syp ur sinpe jo goueievadde jsitj ay) Woy skep L¢ jo oun UONeIIUas B SUTUINSSE ‘UONRIIUIS PUOIAS B JO IUIBIIW SIT[IL9 @ ‘sden jo uoneiado 1sitj ‘palda[ [OI 319M sugutwads 2497YN “PY etOur 10 9UO YITYM Suinp syystu Ajtusts WIeISOISIY ay} JO aseq IYI Ie SMOLIY “IUTT P Aq pavuuo)d BULsq SOUIIXI IY) qYystu SUNII{[OI YIRI IO}J soinjesgduiad MO] pue YSTY sy} UdATS o1e WeISOISTY IY) BAOGY “S1oyJO UY) LopIM Apuanbasuod aie sieq oy) JO UNOS :papnUT JOU 91k UNI IJOU d19M sden ay) Pry uo siyStN ‘uo e1edo den jenpe jo siysiu Q] sutinp paiva{[oo sprdsnueus jo Joquinu oy) sjuasoidal eq YY “PLS ‘SUONII{[OI layyn vdsyuvyp iopy “St “3td 420 Bay cane ch i © © 2 ee Hill a , = sl : s¢ a | | | ! ydas —$ = Ajo o °o fo) N syinpy jo Jequnn (0) 0 — SSaee a — wo f°) Temperature °F fo} t os | | , i | " | | | j ‘palda[[OI alam t4azyn ‘W jo suauideds a10Ul 10 9UO YSTYM SuLInp siystu AjIUsSIs UeIBOISTY IYI JO aseq ay) 1B SMOLIY ‘guy & Aq palwauUOd Suraq sowanxs sy) ‘IYStU SuUNII{[OI Yea 1OJ saimiesoadwi9a} MO, pue YsrYy 94) UDATS ov WIRISOISTY IY) VAOGY “s1oyIO UeY) JapIM ATJUINbasuOoD oie Seq VY) JO 9ULOS ‘papnpuI JOU a1e UNI 10U ataM Ssdey ay) YOTYM UO SIYSIN ‘UoNeIIdo den [enioer jo siystu Q[ SuLMp pada] [09 sprdsnueul jo taquinu oy) sjuasoidar 1eq YOeY “G/G] ‘SUOTIII[[OI Luazyn vdsijuvp I[NpYy “9T “Sy 190 c Jludy ss Ane = s]inpYy jo JequNN . Temperature °F i | | : in ‘potiad Aep-1 auo jo jutodprut sy) sjuasaidai yutod yey “uooU ZI 1 SUIUUTSaq S[PAIOUT INOY-g 1e PpoplOdo1 ‘sgumeraduiay A[Iep INO} SUISeIIAR Aq poye[NI[B atom spotted Avp-/, 10} soinietod wo) Uva “SUOT 1 tady | youPw | 484 S/-y16| e——* vl-€161 %----° Ja] [09 JaWUUNs Surtpadeid somnjesaduis} JUIN “LT ‘Sq | uof l 22q | AON l aunyosadway do 86 Mantispa uhleri Banks laboratory generation time of 37 days. Also marked in this figure is an estimation of the earliest emergence date of a second summer generation, assuming a 37-day generation time and that eggs were laid in the field on 20 June, the first day an adult was collected. A total of 88 adults of M. uwhlert were marked and released in 1974 and 34 in 1975. Three were recaptured in 1974: one male was recaptured twice—on 10 September and 12 September— having originally been collected 7-9 September; a second male caught 7-12 September was recaptured 13 September. No recap- tures occurred in 1975. Several adult females of M. whlerz captured toward the end of the 1975 season (September and October) were kept in the insectary under natural conditions to observe any indications of an adult diapause: none was noted. Females produced eggs until they eventually succumbed to the cold weather. Successive egg clutches developed and hatched until, they too, were halted by the cold. No fertile eggs that failed to hatch in the fall survived the winter. The last eggs to hatch were laid on | October and hatched 27 October; the confined, unfed first instar larvae displayed no signs of arrested development. They exhibited typical searching behavior as previously observed in the labora- tory at long photoperiods. In approximately one week they all died. On the basis of our trapping data, we attempted to find M. uhlert in naturally occurring egg sacs of spiders that had overwintered as subadults or adults. Our survey was conducted 10-27 June 1979 during four collecting trips to wooded areas at Lake of the Woods, Champaign Co., Illinois, and along the banks of the Vermilion River, Vermilion Co., Illinois. We looked for egg sacs under the loose bark of shagbark hickory. The mantispids collected were brought into the laboratory and kept at 25°C and a photoperiod of L:D = 16:8. The dates of their adult emergence were recorded (Table 29). A few egg sacs of Phidippus audax and Metacyrba undata were found, but by far the majority of the egg sacs discovered belonged to the crab spider, Philodromus vulgaris. All seven mantispid associations (Table 29)—five of them verified Mantispa uhleri—were with egg sacs of Philodromus vulgaris. In six of the The Mantispid Seasonal Cycle 87 Table 29. Presence of Mantispa uhleri in Philodromus vulgaris egg sacs collected in 1979 Date of Number of egg Number and Date of adult collection sacs developmental state M. uhleri in 1979 examined of mantispids* found emergence 10 June 5 1 third instar 23 June 17 June 10 1 prepupa in cocoon — killed during collection 19 June not recorded 1 pupa 29 June 27 June not recorded 2 pupaeb 2 July 1 pharate adult collected dead® 1 empty cocoon4 4 Fach association from a separate egg sac. > Pupae in separate egg sacs. One sac attended by female spider which had spun a second sac. “Specimen recently dead: found next to the egg sac from which it had exited. Adult recently emerged: the attending female spider had spun a second egg sac. associations, the egg sacs were guarded by the female spiders that had constructed them. The one unattended egg sac containing a mantispid could still be categorized as recently made because the sac contained a few live spiderlings that had not been eaten earlier as eggs by the mantispid. We were amazed at the number of ragged and tattered mantis- pid cocoons that could still be detected in the remains of spider egg sacs from years past. Thirty to forty such cocoons were found. We wonder how many naturalists have observed but not recognized these relics. Figure 18 illustrates the typical appearance of cocoons of Mantispa uhleri as we found them in Philodromus vulgaris egg sacs. Seasonal data concerning the occurrence of first instar larvae of M. uhlert on wild-caught spiders in southern Illinois (Appendix II) are summarized in Table 30. A total of 123 larvae were recovered from 115 spiders, some of the boarded spiders bearing two, or even three, larvae. Our procedure was not standardized in making these spider collections; therefore, no conclusions as to the relative abundance of these larvae can be drawn. It is important, however, to note their occurrence in nearly all months of the year. ‘s1eah ysed WO1J—UO00IOD pa19Ne) plo ‘YS IY) OT, \(UOOIOD [eaAaI 07 paUadoO Jes) apew A[IUIIII Ja] VY 01 UODDDOTD “YW ‘uU00I0D B JO UOTIISOd ay) SaIedIpUL MOIIY ‘S9a1) AIOYITY YIeGSeys Jo yIeG sy) YeIUaq PI}II[ [OI $2403]NA. SNWOLPO]LY J JO SIeS B39 JY) Ul WazyN Vds1UYPF JO SUOOIOD BuILINIIO ATTeINIeEN “gy “Sy ‘y[npe ayereyd ay) Jo ajoy 1x9 ‘UO0I0D [BIAII 0} pouadO gg] JO IVS “yg SUIMOYS UOOIOD paiedeA A[JUIIII SUTUTRJUOD IVS *|F 90 Mantispa uhleri Banks Table 30. Monthly collections of first instar Mantispa uhleri on spiders in southern Illinois Number of larvae Number of larvae January 16 July February 20 August March 16 September 31 April 9 October May 6 November 3 June 2 December DISCUSSION Mantispa uhleri has, along with some internal parasitoids, the unusual attribute of a larval diapause that is apparently induced and terminated by another, host, organism. Thus, after boarding a spider, a larva ceases development and, in the case that the spider is a pre-adult, waits for the spider to attain sexual maturity and produce an egg sac. The larva then enters the sac and resumes development. The cues for diapause induction and termination are presumably linked to the behavior of the spider (i.e., the act of boarding by the larva, and egg sac construction by the spider, respectively) but, since larvae ingest spider blood, vital chemical cues may also pertain. On larvae reared in pseudosacs at short photophases and/or cool temperatures, we have seen no developmental effects that might indicate the mechanism of any diapause control by these factors of the physical environment. Thus, M. uhleri apparently depends on the spider it boards to monitor the crucial environmental cues necessary for the seasonal timing of its reproductive cycle. The numerous spider species that are boarded by M. whleri spin their egg sacs at varying times of the year. Even if only a fraction of these spiders are successfully used by M. uhleri in producing its own adults, one would expect these adults to emerge at several different times during a season. Thus, if, on the day after a mantispid larva has boarded it, a spider spins an egg sac, the larva will begin development within a day. But if the larva has boarded a young spider that is overwintering before maturation, development of the mantispid adult will be resultantly delayed until some characteristic time the following year. The Mantispid Seasonal Cycle 1 All of our field observations suggest that a spider-inhabiting first instar larva is the only overwintering stage of M. whleri. In both years of our study, adult mantispids did not appear in the traps until late June, the initial dates in these two years differing by only eight days. During May and early June of both years many days were warm enough to lure mantispids to the traps, and if M. uwhleri had overwintered as a fully fed larva, pupa, or adult, one might expect that adults would be seen before late June. But larvae and pupae of M. uhlerz collected from the egg sacs of Philodromus vulgaris in June 1979 had most certainly not overwintered in that condition, since the egg sacs had been spun only that spring. P. vulgaris overwinters as a subadult and matures during April in New England (Kaston, 1948). The specimens of Mantispa uhleri that we collected (Table 29) had without doubt overwintered as first instar larvae on the spiders. Also in support of our conclusion is the negative evidence characterized by our repeated failures to collect any other over- wintering stage, our observations in the insectary of fall-collected adults and their offspring, and our inability to bring about laboratory-induced diapause in any other stage. The first adult mantispids of the year probably emerge from late spring or early summer egg sacs of spiders that had overwintered as adults or subadults. For these spiders, feeding activity would probably begin in March or early April, and egg sacs would be produced sometime thereafter. Assuming a devel- opmental time of 28 days (Table 3), our first adult mantispids (Figs. 15 and 16) probably emerged from egg sacs produced sometime in late May or early June. It is noteworthy that the mantispids that we know entered egg sacs under just such conditions (Table 29), albeit slightly farther north, emerged as adults during the same seasonal period as the earliest adults trapped in southern Illinois. Larvae will also have overwintered on less mature spiders of other species that will be maturing and spinning their egg sacs later in the year. Thus, it might be expected that emergence of the overwintered generation of M. uhleri would occur through- out the summer. We suggest that this emergence curve might be skewed toward the earlier months of the warm season to the extent that the incidence of larvae on species of early egg-sac- 92 Mantispa uhleri Banks spinning spiders may be greater than on those of species spinning later since, assuming comparable population sizes of these different groups, the spiders spinning early would have been available for a longer period during the previous year, thereby increasing the likelihood of their being boarded by a larva. At some time during the summer the eggs produced by the mantispid adults that have matured from the overwintered generation of larvae should begin to appear. Some of the hatching larvae from these mantispids would board spiders destined to spin egg sacs that same year. As a result, a second generation of adult M. uhlerz should emerge, possibly over- lapping the still-continuing emergence of the overwintered generation. The complexity of this cycle may be illustrated with Phidippus audax and Lycosa rabida, from each of which we have taken larvae of Mantispa uhleri in the field. Our observations in Illinois agree with those of Kaston (1948), who concluded that both of these spiders have one generation per year in New England. Phidippus audax overwinters as a subadult in a silken retreat, matures in the spring and soon afterward spins its egg sacs. The offspring then develop slowly through the summer. Lycosa rabida overwinters as a small spiderling—probably in the leaf litter—and sexual maturity and sac production take place in mid to late summer. We expect that an overwintered larva on Phidippus audax would enter an egg sac early in the year and would emerge as an adult, would mate and produce larvae while overwintered larvae on Lycosa rabida were still on the spider. These early summer, first instar larvae—the offspring of those overwintering on Phidippus audax—might then board small immature P. audax spiders that would have hatched a few weeks before, overwinter on subadults of these spiders and enter an egg sac the following spring, approximately one year after hatching from the egg. Some of the same group of early summer larvae might board nearly mature Lycosa rabida spiders and enter egg sacs within a few weeks, so that the resultant adults would be emerging during the same summer. This simplified analysis considers only two spiders: the true situation is certainly even more complicated The Mantispid Seasonal Cycle 93 because of the great number of spider species utilized by Mantispa uhleri. Peck and Whitcomb (1978), in a study of the phenology of cursorial spiders in the south central United States, demon- strated ‘‘a shifting numerical dominance from species to species through the seasons as the adults of one species displaced those of another closely related one.” This is precisely what we suggest is occurring with the spiders utilized by M. whlerz. In fact, two of the spiders from the Peck and Whitcomb study, Schizocosa ocreata Hentz and Lycosa rabida, are included in the host range of Mantispa uhleri (Appendix II). Schizocosa ocreata adults appear in midsummer (June-August), between the early (April- May) maturity of Phidippus audax and late (July-September) maturity of Lycosa rabida. The number of adult M. whlert emerging at any particular time of year, then, should be directly proportional to the number of larvae on various spider species spinning egg sacs approximately one month earlier. For spiders spinning in early summer, these totals derive entirely from overwintered larvae and are probably proportional to the length of time the spiders were available for boarding the previous year and the increasing numbers of mantispid larvae present through the previous summer. For later spinning spiders, producing egg sacs after new summer larvae have appeared in the field, the number of larvae on spiders at the time of egg sac production should increase as the season advances and be composed of an ever-decreasing proportion of overwintered larvae and an increasing proportion of larvae that have hatched during that summer. A given lineage may conceivably have one, two, or even three generations a year, depending on the host spider species. Three generations would seem to be a maximum, based on our calculated 37-day generation time. Three generations should be relatively uncommon, however, since each larva in a given lineage would have to locate spider eggs immediately (by direct sac penetration or boarding a spider about to spin) twice in succession. An average of two generations per year seems more likely. This analysis predicts a complex emergence curve, compound- ed of two separate curves representing overwintered and summer generations. The two components of the overall curve do not 94 Mantispa uhleri Banks imply two separate generations of M. uhleri each year, but rather the combined effect on the overall emergence curve by two groups of larvae with distinctly different histories. The component of adults of the additional summer generations is probably skewed toward the later months of the year because, with the passage of time, there is the likelihood of an increasing frequency of larvae as a result of the growing population of searching larvae and the greater time with which the spiders are exposed to these new summer larvae. In reality the resultant emergence curve could vary considerably depending on the proportionate contributions from the different species of spiders. Peaks could result from large numbers of certain spider species spinning their sacs at particular times. If our prediction of multiple, overlapping generations is correct, we would expect to find first instar larvae on spiders throughout the summer. The data in Table 30 are consistent with this expectation. Although the methods used in collecting these larvae-bearing spiders were not standardized, it is our subjective impression that at no time were such spiders particular- ly difficult to secure. Data obtained through light-trapping should furnish a relative index of the actual population at the times of collection. Thus, if M. uhleri is long-lived in its natural habitat, we would expect to find the capture frequency of adults of the population produced by our predicted emergence patterns to be gradually increasing through much of the season. It is of interest, then, that our data for 1974 (Fig. 15) show a decrease in adult collections during midsummer, suggesting that this curve is more closely approximating emergence. Although our procedures were not originally designed to collect data directly bearing on the following discussion, we shall briefly touch on several circumstances that might cause data drawn from a standing population to approximate the form of an emergence curve. It is possible, for example, that adults of M. uhleri are, in fact, relatively short-lived, so that samples actually are being drawn from such an emerging population. Although we have kept adults alive in the laboratory for several months, it is difficult to equate this with natural conditions where heavy The Mantispid Seasonal Cycle 95 predation or sensitivity to changing weather conditions might produce a much shorter mean life span. A further possibility is that individuals of M. uhleri are attracted to our traps only during a brief period of their life span, such as while searching for mates or during a time of migration or intensive feeding immediately after eclosion. Under these circumstances, too, our collecting data would approximate those derived from an emergence curve. One can visualize additional, special circumstances relating to the stress of capture or the weather conditions. It is obvious that our study is only a preliminary examination of a situation in which many addi- tional data bearing on the basic features of the ecology of the adults of M. whleri are needed. The collecting data for 1975 (Fig. 16) seem completely different from those of 1974. We believe it likely, however, that the overwintering and additional summer generation curves are independent in some degree. Thus, factors that might affect the overwintering population and lower the number of mantispids collected during the first part of the year might not substantially alter the emergence curve of additional summer generations if the offspring larvae from the overwintered survivors encountered summer spiders often enough. This may have happened in 1975, since the number of mantispids collected from late August to early September was virtually the same in both 1974 and 1975. We have no idea what factors might have lowered the overwin- tering population of 1975 or, perhaps, enhanced the winter survivals of 1974. There seemed to be no obvious temperature differences between the winters of 1973-74 and 1974-75 (Fig. 17). Only additional studies can demonstrate whether such year-to- year fluctuations are the rule for M. uhleri and from what they derive. 9. Summary Mantispa uhleri is a member of the neuropteran family Mantis- pidae and develops exclusively in the egg sacs of spiders. In our field and laboratory investigations of this species we have attempted to discover how the larvae of this species locate food, to identify the kinds of spiders utilized, and ultimately to provide a model for the mantispid’s seasonal cycle. Culture methods that we devised for successful laboratory maintenance of M. uhleri and several other mantispid species should be applicable to additional studies throughout the family. We have recorded measurements and descriptions of the developmental stages of M. uhleri, and have described its mating behavior. Every three to five days, mated adult females of M. whleri each lay a clutch of eggs containing from several hundred to several thousand eggs that are attached to the substrate by a short stalk. The number of eggs per clutch is proportional to the body size of the ovipositing female. Extreme variation in adult size is quite common in this species (and other mantispids, as well), due to the varying amount of egg material in the spider egg sacs which larvae enter. Larvae are ‘“‘locked in”’ to their food supply, having no way of locating additional eggs. A third instar larva of M. uhleri thus begins spinning a cocoon whenever its supply of spider eggs is exhausted. After hatching, first instar larvae of M. uhleri must find their own food supply of spider eggs. We have no evidence that females oviposit in any preferential areas that would benefit the larvae. Two distinctive tactics are used by mantispid larvae to locate spider eggs: the direct penetration of an egg sac already constructed in the environment, or the boarding of a female spider prior to sac production and entering of the egg sac at the time of spinning. Some mantispid species employ only one of 96 Summary 97 these tactics. Mantispa viridis, for instance, penetrates egg sacs only, and displays no interest in spiders. Mantispa uhlevi, however, facultatively uses both egg sac penetration and spider boarding, although its penetration technique is neither as rapid nor as predictable as that of M. viridis. When larvae were given a choice of penetrating an egg sac or leaving its vicinity, they only occasionally penetrated the sac. But when given a choice between a restrained spider or an egg sac, the larvae always boarded the spider. These results suggest that immediate penetration of an egg sac is of minor importance to M. uhleri and that the principal larval activity consists of boarding spiders. Although these experiments were not designed to test for the utilization of chemical or tactile cues in locating egg sacs or spiders, no evidence suggested that either are located by anything but random searching by the larvae. To enter an egg sac a larva appresses its head to the surface of the sac and abrades the silk by back-and-forth head movements. As it enlarges the opening, the larva pushes forward until it has entirely invaded the egg sac and is out of sight beneath the silk. Once it is inside the sac, several factors may interfere witha larva’s survival, including age of the sac and simultaneous invasion of the sac by other larvae. Larvae will board spiders of either sex and any state of maturity, although they apparently fall within the prey size-range of very small spiders, making successful boarding of these spiders difficult. Larval boarding seems not to be affected by the presence of another mantispid already on the spider. After climbing on a spider, larvae of M. whleri preferentially wrap themselves around the spider’s pedicel. They negotiate the molts of immature spiders with remarkable success, and at the first molt that produces a sufficiently large spider, they enter its book lungs, regardless of the spider’s sex. These larvae generally remain in the book lungs during the spider’s subsequent development. At the molt which produces a mature female spider, however, larvae attempt to leave the book lungs for the pedicel, probably in anticipation of the spider’s impending oviposition. This maneuver is often unsuc- cessful; many larvae are dislodged from the spider and cannot regain their position. We did not find that any larvae transfer from mature male spiders to females during mating; such 98 Mantispa uhleri Banks transfers, however, did sometimes occur while male spiders were being cannibalized by females. During their tenure on a spider, larvae feed on spider hemo- lymph by penetrating the thin cuticle around the pedicel or within a book lung, but they do not engorge. During egg sac production the larvae enter the egg sac and only then resume development by piercing and draining the contents of the spider’s eggs. There are three larval instars. The mature third instar spins a cocoon within the spider egg sac, using silk from its Malpighian tubules. The pharate adult bites its way out of the cocoon and egg sac and crawls some distance away before undergoing the final ecdysis to the adult. The number of spider species used by larvae of M. uhleri was investigated by the examination of a collection of over 5,000 spiders that had been independently assembled during a study of the spiders of southern Illinois. This sample, unbiased with respect to the likely presence of mantispid larvae on spiders, included sixteen first instar larvae of M. whlerz. All of them were found on species of the hunting groups Lycosoidea and Clubio- noidea. This, together with the absence of M. whleri larvae on the nearly 3,000 specimens of nonhunting spiders, suggests that first instar M. uhlert are often or always associated with hunting spiders but rarely, if ever, with web spinners. A large sample of spiders collected by the authors from a restricted locale in southern Illinois supports this conclusion. The two collections together provide association of over one hundred larvae of M. uhleri with 31 species of hunting spiders distributed among 21 genera and almost every family of hunters. Our field studies of M. whleri in southern Illinois showed that this species overwinters as first instar larvae on spiders; it seems likely that this is the only overwintering stage in this area. Beginning in the spring, the larvae enter the egg sacs of the spiders on which they overwintered, the time of entrance being controlled by the phenologies of the different species of spiders. Adults of these larvae begin emerging in late June or early July. In midsummer, probably before all the overwintered larvae have entered sacs, new larvae produced by early emerging adults appear in the field. The timing of the transformation of these new larvae to the adult stage depends also on the species of spiders that Summary 99 they board. Direct penetration of egg sacs at any time can be considered comparable to’ the boarding of a spider that spins an egg sac immediately. There is thus a continuing emergence of adult M. uhlert throughout the summer, a given lineage produc- ing one, two, or possibly, three generations every twelve months. The seasonal presence of adults in black-light traps in operation throughout the warm months of two consecutive years supports our theory of the emergence pattern of adult M. uhleri. Future studies will undoubtedly show that this insect plays a much greater role in the forest ecosystem than has been previously supposed. Addendum Several relevant papers dealing with the developmental and behavioral ecology of the Mantispidae have appeared during the three-year interim from the date of acceptance to the date of publication of this monograph. These papers are: Gilbert and Rayor (1983); Killebrew (1982); MacLeod and Redborg (1982); Opler (1981); Redborg (1982a, 1982b, 1983); Redborg and Mac- Leod (1983a, 1983b). RP Wop MPE Med) ty. etarttetta oviteritphaoes Ag MY seer ume ee tt ae ane 8 et gaiieed sro sided Se Pynantet Sghwertyrsery & zed) td Seesefa ine peeved: ae ip aes » dni vom etd cylin ARTISAN tay ce ig tar intl epi fekistsmperette vrteerauWtive ety Mega) ws dep sriagdiar: roe tox “+i i, Tee Vo debe doting: Sheree? t- yRoien wre ood adits codeine istry gee 1 + - ait esta, ane “Ley nce uend! oh agp ‘ ‘ ene a sey nell ke # it stu? Adi é , — 7) LF } j j } (iia: 2 ' : de s/ By oO” . ! - (he i 1 wi “ ti 2 ' , Al oP uj if (hd trai mPa eo on ine Appendix I. Mantispa uhleri Banks-Spider Associations in the Southern Illinois University at Carbondale Collection Number of Number of spiders Number of spiders per species mantispid Spider families, examined Im- larvae Months of genera, and species per family mature &” @ removed collection Antrodiaetidae 68 Antrodiaetus unicolor 26. TO«30 IV-VIIX-XI (Hentz) Atypoides hadros Coyle 2. 1 0 X, XI Ctenizidae 4 Ummidia sp. 4 0. 0 VIL VII,X Amaurobiidae 12 Amaurobius bennetti (Black- 0 1 0 IV wall) Titanoeca americana Emerton 2 0 O VI Titanoeca brunnea Emerton 8 ae) V-VIl Dictynidae 218 Dictyna sp. ] Oo” 0 IV Dictyna bellans Chamberlin 2 |) IV,VI-VIII Dictyna bicornis Emerton 0 Ds 2 if VI Dictyna cruciata Emerton 9. 25 o V-VII Dictyna foliacea (Hentz) LOL C22 V-VIL IX Dictyna formidolosa Gertsch 2 0 0 V,vVI and Ivie Dictyna hentzi Kaston ] ee 0) IV Dictyna sublata (Hentz) 30 33 =«~OO IV-VII,X Dictyna volucripes Keyserling J 40" 10 IV-VIII,X Lathys immaculata 0 6 0 II Chamberlin and Ivie Oecobiidae 2 Oecobius cellariorum ] yarg V,VIl (Duges) Uloboridae 10 Uloborus glomosus 2 Ol 42 V, VIL VIII (Walckenaer) 101 102 Mantispa uhleri Banks Appendix I (cont.) Spider families, genera, and species Dysderidae Ariadna bicolor (Hentz) Dysdera crocata C. L. Koch Scytodidae Loxosceles reclusa Gertsch and Mulaik Scytodes thoracica (Latreille) Pholcidae Pholcus phalangioides (Fuess- lin) Spermophora meridionalis (Hentz) Theridiidae Achaearanea porter: (Banks) Achaearanea globosa (Hentz) Achaearanea tepidariorum (C. L. Koch) Argyrodes cancellatus (Hentz) Argyrodes elevatus Taczanomski Argyrodes fictilium (Hentz) Argyrodes trigonum (Hentz) Crustulina altera Gertsch and Archer Ctenium frontatus (Banks) Dipoena nigra (Emerton) Enoplognatha sp. Euryopis funebris (Hentz) Euryopis quinquemaculata Banks Latrodectus mactans (Fabricius) Number of Number of spiders Number of spiders per species mantispid examined Im- larvae Months of per family mature o& @ removed collection 7 0 28 IV,V,VIII, XI 0 0 2 IV 48 12 i 23 IV-X 0 $7 il IV,VI,VII,X 22 5 ~ Jib lal) II,V, VII, VIII, XI Me wk 0 I1,X 810 5 a V,VI, VIII, X1 3 2 2 V, VII, VII 78 195 169 LIV-XI 2 ] V-VII 2 7 IV,V,IX 0 Opel xX 2 4 0 V,VI 3 Li aaa 1,111, V,TX,XI1 ] Sae0 I, VII, VII 2 6 0 V-VIll 0 0 1 X a 0. 2 IV,V,VIII ] 0 0 V 2 Lie =0 V-X Appendix I 103 Appendix I (cont.) Number of Number of spiders Number of spiders per species mantispid Spider families, examined Im- Months of genera, and species per family mature &% @ removed collection Theridiidae (cont.) Latrodectus variolus 2 2 te V Walckenaer Nesticus pallidus Emerton 0 0 VI Pholcomma hirsuta Emerton 2 2 IlI,V,XI Phoroncidia americana 0 0 V (Emerton) Spintharus flavidus Hentz 3 20 6 VII-IX Steatoda borealis (Hentz) 4A 248 IV,V,VULIX Steatoda triangulosa 10 34 10 LILIV,V,VIII- (Walckenaer) XII Stemmops ornatus (Bryant) 4 0 VILVIUl Theridion alabamense 2 0 IV-VI Gertsch and Archer Theridion albidum Banks l ae) VI-VIII Theridion antonii Keyserling 0 a) Xx Theridion berkeley: Emerton 0 et Vv Theridion differens Emerton 23. 25. 3 IV-X Theridion flavonotatum ] 2 10 V,VI Becker Theridion frondeum Hentz A’ Wy: t42 VIL VII Theridion glaucescens Becker 3 7 0 IV,VI,X,XI Theridion lyricum 5) lod V,VII-X Walckenaer Theridion murarium 4 2» J0 VI, VII, XII Emerton Theridion pictipes b <2 gah V, Vill Keyserling Theridula opulenta 3. 1%), a V,VII (Walckenaer) Thymoites pallida ] a IX (Emerton) Thymoites unimaculata o Bi we IV-VI (Emerton) 104 Appendix I (cont.) Spider families, genera, and species Linyphiidae Bathyphantes pallida (Banks) Centromerus cornupalpis (O.P.-Cambridge) Centromerus latidens (Emerton) Ceraticelus creolus Chamberlin Ceraticelus fissiceps (O.P.- Cambridge) Ceraticelus laetabilis (O.P.- Cambridge) Ceraticelus micropalpis (Emerton) Ceraticelus minutus (Emerton) Ceratinella brunnea Emerton Ceratinopsidis formosa (Banks) Ceratinopsis sp. Ceratinopsis laticeps Emerton Ceratinopsis purpurescens (Keyserling) Ceratinopsis tarsalis Emerton Cornicularia sp. Cornicularia brevicornis Emerton Eperigone sp. Eperigone banksi Ivie and Barrows Eperigone maculata (Banks) Eperigone tridentata (Emerton) Mantispa uhleri Banks Number of Number of spiders Number of spiders per species mantispid examined Im- larvae Months of per family mature o @Q removed collection bil 0 Ss jw V,XI 7 57,0 IV,X,XI 3 8 1 1, III, X,XI ] a 1 IV,VI 10 She IV-IX 0 b ie Ill 0 2.990 VIII,IX 0 S40 VI,X, XI 1 Org IV 2 a VIILIX,X 0 Teo 0 V 0 pore V 4: WO) As IV-VII,X ] Oy a vil ] a 0 IILIV,XI ] wg IX ] Wak a X ] ro X 4 ae) IV,V,VIII,XI 2 6 LIV,VI,X Appendix I Appendix I (cont.) Spider families, genera, and species Linyphiidae (cont.) Eridantes erigonoides (Emerton) Erigone autumnalis Emerton Gonatium rubens (Blackwall) Grammonota inornata Emerton Grammonota vittata Barrows Islandiana flaveola (Banks) Islandiana longisetosa (Emerton) Floricomus spp. Floricomus rostratus (Emerton) Florinda coccinea (Hentz) Frontinella pyramitela (Walckenaer) Linyphia spp. Linyphia maculata Emerton Linyphia marginata C.L. Koch Linyphia waldea Chamberlin and Ivie Meioneta spp. Meioneta angulata (Emerton) Meioneta micaria (Emerton) Meioneta unimaculata (Banks) Origanates rostratus (Banks) Paracornicularia bicapillata Crosby and Bishop Phanetta subterranea Emerton Pityohyphantes costatus (Hentz) 105 Number of Number of spiders Number of spiders per species mantispid examined Im- Months of per family mature o° collection 0 a | IV-VI 15 50 II, V-VII,XI 0 hooG XI 2 a 30 VI,VII,X ] 0 O X 2 1 a VI,XII 1 0 O Vil ] ay 10 V,X,XI l 0 0 VI 4) JAS IV,V,VII,X 14) 55: 21 IV-VII,X,XI l 1 0 VII, VIII l I 6 VI,VUl 40, 83:23 IV-X ] 0 O Vv |e Me se IV-X 0 1 0O IX ] ae Vill,x 0 bio IX 3 146690 1,111, V,1X-XI ] 2.0 XI ] 0 1 XII 0 Zavin's IV,IX,X 106 Mantispa uhleri Banks Appendix I (cont.) Number of Number of spiders Number of spiders per species mantispid Spider families, examined Im- larvae Months of genera, and species per family mature & removed collection Linyphudae (cont.) Tapinopa bilineata Banks 0 Ok Vill Tennesseellum formicum 2 L 0 V-VII (Emerton) Walckenaera vigilax (Black- ] 0 0 Vill wall) unidentified 0 12> 90 1,IV-VII,X,XI Araneidae 1,141 Acacesia hamata (Hentz) 6 toate VII, VUIIX Alpaida calix (Walckenaer) 0 1 +6 V Acanthepeira stellata (Marx) 9 “5586 IV-VIIX,X Araneus gigas (Leach) 0 : 10 VI Araneus bonsallae (McCook) ] 204 VI Araneus juniperi (Emerton) ] 0 O Vil Araneus marmoreus Clerck I 25: 1g IX,X Araneus niveus (Hentz) 0 a! IX Araneus pratensis (Emerton) 10 a5 IV-VI,VIII,X Araneus thaddeus (Hentz) 0 2 0 X,XI Araniella displicata (Hentz) 0 bad V Argiope aurantia Lucas 5 8 it VII-X, XII Argiope trifasciata (Forskal) 3.) ee IX,X Cyclosa conica (Pallas) ] 0 @ IV Cyclosa turbinata ] 2 0 V,VI,VUI (Walckenaer) Gea heptagon (Hentz) 2 Zoe IV,V,VII,X Hypsosinga funebris 0 hore VI (Keyserling) Hypsosinga rubens (Hentz) 3 14902 IV-VI,X Hypsosinga pygmaea 0 Gia VI, VIII,X (Sundevall) Leucauge venusta 10 36 6 V-IX,XI (Walckenaer) Mangora gibberosa (Hentz) 23"! 15: 2a: VI-IX Mangora maculata is: 35 2 VI-IX (Keyserling) Appendix I Appendix I (cont.) Spider families, genera, and species Araneidae (cont.) Mangora placida (Hentz) Mastophora phrynosoma Gertch Meta menardii (Latreille) Metepeira labyrinthea (Hentz) Micrathena gracilis (Walckenaer) Micrathena mitrata (Hentz) Micrathena sagittata (Walckenaer) Mimognatha fox: (McCook) Neoscona arabesca (Walckenaer) Neoscona domiciliorum (Hentz) Neoscona hentzii (Keyserling) Nuctenea cornuta (Clerck) Pachygnatha brevis Keyserling Singa keyserlingi McCook Tetragnatha elongata Walckenaer Tetragnatha laboriosa Hentz Tetragnatha pallescens F.O.P.-Cambridge Tetragnatha seneca Seeley Tetragnatha straminea Emerton Tetragnatha versicolor Walckenaer Verrucosa arenata (Walckenaer) Number of spiders Number of spiders per species __ mantispid examined Im- larvae per family mature o& 9@ removed Number of @ 44505 0 yb 8 by 23°33 G Syed G 82 3 1s on 20 1 oe a 3 27 00 & 36) 5 OF 30 S467) 70 Io 4b 3 0 i 4a Oo 20 25) 28hcie 23) S927 OF) 16, 0 0 daze ee! l rar a) AQh Ss 107 Months of collection IV-VIII Vill I,X,XI VILIX III, VI-X VII-X VIil-X VL VII,X IV,VI-XI IX,X VIil-XI IV-XI ? VX VI-XI IV-X VI,X VII, VIII,X V,VIII VI-VIII,X IV,VI-X 108 Appendix I (cont.) Spider families, genera, and species Araneidae (cont.) Wixia ectypa (Walckenaer) unidentified Symphytognathidae Maymena ambita (Barrows) Mysmena guttata (Banks) Mimetidae Ero sp. Ero furcata (Villers) Mimetus epeiroides Emerton Mimetus puritanus Chamberlin unidentified Agelendiae Agelenopsis emertoni Chamberlin and Ivie Agelenopsis kastoni Chamberlin and Ivie Agelenopsis naevia (Walckenaer) Agelenopsis pennsylvanica (C. L. Koch) Cicurina arcuata Keyserling Cicurina brevis (Emerton) Cicurina ludoviciana Simon Cicurina pallida Keyserling Coras lamellosus (Keyserling) Coras taugynus Chamberlin Tegenaria domestica (Clerck) Hahniidae Hahnia cinerea Emerton Mantispa uhleri Banks Number of Number of spiders Number of spiders per species mantispid examined Im- larvae Months of per family mature o 9 removed collection 3 P4ui VIlI-X O22 eG VII,X, XII 2 0 Pg Vv 0 rane Vv 18 0 }) 99 IV 2 Oe VX Oy Cat Vill l G70 Vil, Vill 6 Fy 6 VI-VIII 248 IG) 37,46 VI-XI ae IV-VI,X 2 MO he VI,VIII-X 9° lS VI,VII,X 10 Woe I,11,1V-VIII,X, XI 4 4.0 IV,V,IX-XI Z £40 IV,VI,X,XI CO 28 IV,X AS ee A III-VII,IX-XI 0 | a V 3 Zi ee IV, VI, VIll,X Zl Appendix I 109 Appendix I (cont.) Number of Number of spiders Number of spiders per species mantispid Spider families, examined Im- larvae Months of genera, and species per family mature o%” 9 removed collection Hahniidae (cont.) Hahnia flaviceps Emerton O° aie IIL,IV,X Neoantistea agilis (Keyserling) 8 30 III,V,VI,IX-XI Pisauridae 76 Dolomedes sp. 0 be V,VI Dolomedes albineus Hentz 0 1 a IV,VI Dolomedes scriptus Hentz ] Ly IV,VIII Dolomedes tenebrosus Hentz 1 4 0 1 IV-VIILIX Dolomedes triton 4 4 4 IV,V,VILIX (Walckenaer) Dolomedes vittatus 0 ee V,X Walckenaer Pisaurina sp. ea II Pisaurina brevipes (Emerton) Ii) Se II-V Pisaurina dubia (Hentz) 5 lid ILIV,V Pisaurina mira (Walckenaer) 4 3 IV-VILIX Lycosidae 614 Arctosa funerea (Hentz) 2 20 oe VLIX,X Geolycosa missouriensis 0 r ©o II (Banks) Lycosa sp. Oy Oxo IV Lycosa antelucana Montgomery 0 I IV Lycosa aspersa Hentz 0 2 a0 V Lycosa balttmoriana 0 jeer! VI (Keyserling) Lycosa carolinensis it a V,IX-XI Walckenaer Lycosa frondicola Emerton 0 Sis wel IV,V Lycosa georgicola ] 24180 Vill,x Walckenaer Lycosa gulosa Walckenaer a 6x0 II,IV,IX-XI Lycosa helluo Walckenaer a OR es V-VIII,X, XI Lycosa hentzi Banks 0 2 IV,IX Lycosa pulchra (Keyserling) 4 230 11,X,XI 110 Mantispa uhleri Banks Appendix I (cont.) Number of Number of spiders Number of spiders per species mantispid Spider families, examined Im- larvae Months of genera, and species per family mature & @ removed collection Lycosidae (cont.) Lycosa punctulata Hentz 6 liotd IV,V,IX,X Lycosa rabida Walckenaer 9 .\ (a2 VI-X Pardosa milvina (Hentz) 29 66 13 IV-XI Pirata spp. ] Oral VX Pirata alachua Gertsch and 31. 29 3 V-IX Wallace Pirata insularis Emerton VI 0 lis Pirata maculatus Emerton 3 3 0 VI,VIIIIX,X Pirata spiniger (Simon) ] OF ¥2 IX Schizocosa sp. 0 1 40 VI Schizocosa avida 5 930 IV-VII (Walckenaer) Schizocosa bilineata 23 3: V,VI (Emerton) Schizocosa ocreata (Hentz) 85 83:1 IV-X Schizocosa saltatrix (Hentz) 52 1 ise0 IV-VII Trochosa avara Keyserling 15: SES) 40 III-VI,X,XI unidentified 5 9° 10 IV-VIII,X, XI Oxyopidae 289 Oxyopes sp. 2 lt #0 Vil Oxyopes aglossus 5 a 10 V-VIII Chamberlin Oxyopes scalaris Hentz 0 O14 IX Oxyopes salticus Hentz 68 69 140 V-VIII,X Gnaphosidae 108 Cesonia bilineata (Hentz) 3 0 a0 V,VI Drassylus aprilinus (Banks) 6 G22 II, 11I,V,VI,XI Drassylus covensis Exline 2 2, 00 V,VI Drassylus creolus ] Lao V Chamberlin and Gertsch Drassylus depressus ] O60 V (Emerton) Drassylus fallens Chamberlin ] ee | VI Appendix I Appendix I (cont.) Number of Number of spiders spiders per species Spider families, examined Im- genera, and species Gnaphosidae (cont.) Drassylus frigidus (Banks) — i=) Drassylus virginianus 9 7 Chamberlin Gnaphosa fontinalis ] 2 Keyserling Gnaphosa sericata (L. Koch) i Haplodrassus bicornis 3 ] (Emerton) Haplodrassus signifer 2 3 (C. L. Koch) Herpyllus ecclesiasticus 8 8 Hentz Litopyllus rupicolens 0 ] Chamberlin Micaria laticeps Emerton Rachodrassus exlinae ] ] Platnick and Shaalab Sergiolus sp. 07 20 Sergiolus decoratus Kaston 0 ] Sergiolus famulus ] 0 Chamberlin Sergiolus variegatus QO 6 (Hentz) Zelotes duplex Chamberlin 4 2 Zelotes hentzi Barrows 5 ] unidentified 2 4 Clubionidae 350 Castianeira amoena 0 ] (C. L. Koch) Castianeira cingulata 4 Il (C. L. Koch) uN Castianeira crocata (Hentz) Castianeira descripta (Hentz) OF) He per family mature & 9 111 Months of collection II V,VI,IX V-VIil VLVII V,VI IV-VI III-VIll,xX Vill IV,V III, VI IX X IV-VI IlI-V,X V,VII VII III, V-1X V-VII IV 112 Mantispa uhleri Banks Appendix I (cont.) Number of Number of spiders Number of spiders per species mantispid Spider families, examined Im- larvae Months of genera, and species per family mature o& @ removed collection Clubionidae (cont.) Castianeira longipalpus 2) Bade V,VLIX-XI (Hentz) Castianeira trilineata (Hentz) ] 0 0 IV Castianeiva variata Gertsch ] 1S VIL, VIII Chiracanthium inclusum ] 0 °0 Vil (Hentz) Chiracanthium mildei 7 (oe | 1 IV,V, VILIX-XI L. Koch Clubiona abboti L. Koch Ce ee V-VIII,X Clubiona catawba Gertsch 4 a. IV-VI Clubiona excepta L. Koch he Ae fe V-XI Clubiona kastoni Gertsch 0 4 0 V,VI Clubiona maritima L. Koch 4 a VI, VIII Clubiona obesa Hentz 3 4 0 ] IV-VI,X Clubiona pygmaea Banks ] 0 O VI Phrurotimpus alarius Sr 28) 65 II-IX, XI, XII (Hentz) Phrurotimpus illudens Se 0 a0 V-VII Gertsch Scotinella sp. 0 a) V,IX,X Scotinella britcheri 2 Pr 6 IV,V,X (Petrunkevitch) Scotinella formica (Banks) ] 250 II, VIII Scotinella goodnight 18) 20ty 0 II-XI (Muma) Scotinella redempta (Gertsch) 31/0, 0 III,V,VII,X Strotarchus piscatorius 1 bya V,VI (Hentz) Trachelas deceptus (Banks) ] 2 IV,V unidentified 0 poo VII Anyphaenidae 118 Anyphaena sp. OQ ‘pQyexd ] X Keyserling Appendix I 113 Appendix I (cont.) Number of Number of spiders Number of spiders per species mantispid Spider families, examined Im- larvae Months of genera, and species per family mature o 9 removed collection Anyphaenidae (cont.) Anyphaena celer (Hentz) ] 1 0 X Anyphaena fraterna (Banks) 5 5 0 IV-VII,X Anyphaena maculata (Banks) ] ‘RON X Anyphaena pectorosa 6.6). ).0 VI-VII L. Koch Aysha sp. OO) 27 IV Aysha gracilis (Hentz) 6 er | IV,V,VII, VIII Wulfila alba (Hentz) ] 4 0 VI, VIII Wulfila saltabunda (Hentz) lg 2a V-VII Ctenidae 2 Anahita animosa 0 0. a IV (Walckenaer) Zora pumila (Hentz) OO" al III Thomisidae 406 Coriarachne floridana Banks 0 Log IV Coriarachne versicolor ] 1 igs IV,X Keyserling Ebo sp. 0 0)! a XII Ebo latithorax Keyserling ] Ol V,XI Misumenoides aleatorius 9 16 10 IV,VI-X (Hentz) Misumenops asperatus 43) A IV-VII,IX (Hentz) Misumenops celer (Hentz) ] ae ] IV-VILIX,X Misumenops oblongus 1 ed | V-VIII,XI (Keyserling) Oxyptila monroensis QQ! 4: ke I,IV-VI Keyserling Philodromus bimuricatus 1 2G IV,VI Dondale and Redner Philodromus infuscatus Oe a X,XI 114 Mantispa uhleri Banks Appendix I (cont.) Number of Number of spiders Number of spiders per species mantispid Spider families, examined Im- larvae Months of genera, and species per family mature o& 9? removed collection Thomisidae (cont.) Philodromus keyserlingi 5 TAD VIVII Marx Philodromus laticeps ] ie | XII Keyserling Philodromus marxii 5 > © IV-VII Keyserling Philodromus minutus Banks 3 2. 0 V,VI Philodromus montanus ] 0 0 IV Bryant Philodromus placidus Banks a VI Philodromus pratariae 15 ‘ae | VIILIX (Scheffer) Philodromus rufus 4 | IV-VI Walckenaer Philodromus vulgaris 5 Db) > 4i IL,IV-VII,X (Hentz) Synema sp. ] Pe Vivi Synema parvula (Hentz) 40 32 24 IV-X Tibellus duttoni (Hentz) ] ae | II,VI Tibellus oblongus 0 | II (Walckenaer) Thanatus rubicellus ] 0 0 X Mello-Leitao Tmarus angulatus 0 5 VLVII (Walckenaer) Xysticus auctificus 0 PEG V-VIL,XI Keyserling Xysticus discursans ] 0 of IV Keyserling Xysticus elegans Keyserling 4 0 8 IV,VI Xysticus fraternus Banks 16 ha | V,VI Xysticus funestus Keyserling sin 14: IV-VI,VIII- XII Xysticus ferox (Hentz) 9 ay V-VII,X Appendix I 115 Appendix I (cont.) Number of Number of spiders Number of spiders per species mantispid Spider families, examined Im- larvae Months of genera, and species per family mature o @ removed collection Thomisidae (cont.) Xysticus texanus Banks ] yg VII Xysticus triguttatus 0 eX) VI Keyserling Salticidae 654 Agassa cyanea (Hentz) 4 6 60 V-VII Ballus youngi Peckham 72 0 O IV Evarcha hoyi (Peckham) 0 ao 00 V,VIll Gertschia noxiosa (Hentz) 0 2 0 V Habrocestum pulex (Hentz) 7 ae V,VI,XI Habronattus sp. Oo woe I Vill Habronattus agilis (Banks) ] 0 O IV Habronattus decorus 0 1 40 VI (Blackwall) Hentzia mitrata (Hentz) 3 10 19 2 ~ IV-VII,X,XI Hentzia palmarum (Hentz) 5 » 0 V,VI Icius elegans (Hentz) ] 78 VILVII Icius harti1 Emerton ] 0 O VI Icius similis Banks 0 Deed VI Maevia inclemens o~ 1 0 1,V-VIl (Walckenaer) Marpissa formosa (Banks) 5 Av? VL VII Marpissa lineata ] 1 0O IV,V (C. L. Koch) Marpissa pike: (Peckham) 2 Vike a A8 IV-VI, VIII Metacyrba undata (De Geer) 12 ilo a] IV-VI,IX, XI Metaphidippus spp. 7 is 0 III,IV-V1,X Metaphidippus canadensis G2. 0 V,X (Banks) Metaphidippus galathea 2 ae) IV-VII (Walckenaer) Metaphidippus protervus 94 22 40 3 IV-VILIX,X (Walckenaer) Paraphidippus aurantius 12+ 4 Gewrel 1 V-VII,xX (Lucas) 116 Appendix I (cont.) Number of spiders examined Im- per family mature o 9 Spider families, genera, and species Thomisidae (cont.) Paraphidippus marginatus (Walckenaer) Phidippus sp. Phidippus audax (Hentz) Phidippus clarus Keyserling Phidippus mystaceus (Hentz) Phidippus princeps (Peckham) Phidippus putnami (Peckham) Phlegra fasciata (Hahn) Salticus scenicus (Clerck) Sassacus papenhoei Peckham Sitticus cursor Barrows Sitttcus fasciger (Simon) Synemosyna formica Hentz Thiodina sp. Thiodina puerpera (Hentz) Thiodina sylvana (Hentz) Zygoballus bettint Peckham Zygoballus nervosus (Peckham) Zygoballus sexpunctatus (Hentz) Totals 5,761 20 11 — coor OOO N = — 96 Oi, 2 M2 BY ya Le 7 0 0 0 0 = ie Dg UL, 2 °C O78 0 0 4 0 15) 95 10 0 a, JC 5-0 5,761 16 Mantispa uhleri Banks Number of spiders Number of per species mantispid larvae Months of removed collection IV,V,VII,VII,x IX I-VI VI-VIII VIiTl,Ix IV-VI VIII III-VI L,1,1V,XI,XII IX,X IV,VI IV-VII IV-X IV-VII,X VII-IX IV-VI,X Appendix II. Spider Species Utilized by First Instar Man- tispa uhlert Larvaet Agelenidae Agelenopsis kastoni Chamberlin and Ivie female: left book lung; IV-15-76 Anyphaenidae Anyphaena sp. immature male: left side of pedicel; Union County*; X-7-67 Anyphaena fraterna (Banks) male: dorsal pedicel in left pit; IV-16-76 male: left side of pedicel; IV-17-76 Anyphaena pectorosa L. Koch female: pedicel; IX-4-74 Clubionidae Chiracanthium mildei L. Koch immature female: right book lung; Jackson County*; IV-21-68 Clubiona obesa Hentz male: dorsal pedicel; IV-16-76 female: in egg sac; Williamson County*; IV-18-71 Gnaphosidae Herpyllus ecclesiasticus Hentz male: right side of pedicel; II-8-76 Lycosidae Lycosa pulchra (Keyserling) immature: pedicel; VIII-6-74 immature: right side of pedicel; VIII-7-74 immature: ventral pedicel; VIII-7-74 immature female: book lung; VIII-12-74 male: left side of pedicel; I-8-75 {Each line entry beneath a species name gives data (sex of spider: position of larva; collection locale; date of collection) for an individual of that species from which at least one larva of M. uhleri was removed. Multiple larvae from a single spider are recorded by a number. An asterisk following the collection locale signifies that that spider is from the SIUC Collection; if no locale is indicated, the collection was made at DSAC (see also p. 76). 117 118 Mantispa uhleri Banks Lycosa punctulata Hentz immature female: right book lung; IX-4-74 male: right side of pedicel; [X-10-74 Lycosa rabida Walckenaer male: dorsal pedicel in left pit; VIII-7-74 Schizocosa ocreata (Hentz) female: left edge of carapace; VIII-6-74 female: right side of pedicel; VIII-7-74 Pisauridae Dolomedes tenebrosus Hentz male: left book lung; Pope County*; V-19-71 Pisaurina mira (Walckenaer) immature male: dorsal pedicel in left pit; IX-29-74 immature male: ventral pedicel; I-28-75 immature female: 2 larvae—right side of pedicel, right book lung; IX-10-74 Salticidae Evarcha hoyi (Peckham) female: right book lung; X-16-74 Hentzia mitrata (Hentz) immature male: dorsal pedicel; Jackson County*; X-13-66 immature female: dorsal pedicel; IV-18-76 female: ventral pedicel; Hamilton County*; VII-20-72 Maevia inclemens (Walckenaer) immature male: pedicel; IX-3-74 Metacyrba sp. immature female: dorsal pedicel; XI-13-74 Metacyrba undata (De Geer) immature: immature: immature: immature: immature: immature: immature: immature: immature: immature: immature: immature: left side of pedicel; 1-23-75 left side of pedicel; I-23-75 right side of pedicel; II-4-75 ventral pedicel; II-4-75 right side of pedicel; II-4-75 dorsal pedicel; II-14-75 ventral pedicel; II-14-75 ventral pedicel; II-14-75 left side of pedicel; II-14-75 right side of pedicel; II-8-76 ventral pedicel; II-8-76 ventral pedicel; [1-26-76 Appendix II Salticidae (cont.) Metacyrba undata (De Geer) (cont.) immature: ventral pedicel; III-4-76 immature: left side of pedicel; III-15-76 immature: right side of pedicel; III-15-76 immature: right side of pedicel; III-15-76 immature: left side of pedicel; III-16-76 immature: dorsal pedicel in left pit; III-16-76 immature: left side of pedicel; III-16-76 immature: right side of pedicel; III-25-76 immature: right side of pedicel; III-25-76 immature male: pedicel; Alexander County; I-15-74 immature male: ventral pedicel; II-4-75 male: male: male: male: male: male: male: male: male: male: male: male: male: male: male: male: male: pedicel; Alexander County; I-15-74 pedicel; Hardin County; I-16-74 pedicel; Hardin County; I-16-74 left side of pedicel; I-7-75 right side of pedicel; I-7-75 ventral pedicel; 1-23-75 ventral pedicel; I-31-75 pedicel; II-2-75 right side of pedicel; I-3-75 dorsal pedicel; II-14-75 right side of pedicel; II-14-75 left side of pedicel; II-14-75 ventral pedicel; II-14-75 dorsal pedicel; III-15-76 left side of pedicel; III-15-76 right side of pedicel; III-16-76 ventral pedicel; III-16-76 female: pedicel; Alexander County; I-15-74 female: pedicel; Alexander County; I-15-74 female: ventral pedicel; I-7-75 female: dorsal pedicel in right pit; II-14-75 female: dorsal pedicel in left pit; II-14-75 female: ventral pedicel; II-15-76 female: spinnerets; III-16-76 Metaphidippus galathea (Walckenaer) female: pedicel; V-12-75 Metaphidippus protervus (Walckenaer) male: ventral pedicel; Jackson County*; X-15-66 119 120 Mantispa uhleri Banks male: ventral pedicel; Pope County*; V-19-71 male: left side of pedicel; Pope County*; V-19-71 Paraphidippus aurantius (Lucas) male: ventral pedicel; Jackson County*; VI-18-71 Pellenes sp. female: ventral pedicel; [X-17-74 Phidippus audax (Hentz) immature: dorsal pedicel; [X-12-74 immature: between leg bases and carapace; IX-17-74 immature: right book lung; IX-27-74 immature: left side of pedicel; X-12-74 immature: 2 larvae—ventral pedicel, left book lung; XI-9-74 immature male: right stde of pedicel; [X-17-74 immature male: right side of pedicel; [X-17-74 immature male: left side of pedicel; [X-17-74 immature male: dorsal pedicel in left pit; X-10-74 immature male: right side of pedicel; X-10-74 immature male: dorsal pedicel in left pit; X-11-74 immature male: left side of pedicel; X-11-74 immature female: 3 larvae—ventral pedicel, dorsal pedicel in left pit, between leg bases and carapace; IX-7-74 immature female: 3 larvae—between sternum and leg bases, dorsal pedicel in right pit, left book lung; [X-19-74 immature female: ventral pedicel; [X-20-74 immature female: left book lung; [X-20-74 immature female: dorsal pedicel; X-6-74 immature female: dorsal pedicel in left pit; X-10-74 immature female: ventral pedicel; I-10-75 male: ventral pedicel; Pope County*; V-19-71 male: right book lung; VI-16-75 female: dorsal pedicel in right pit; Pope County*; V-19-71 Phidippus clarus Keyserling immature: left book lung; FX-19-74 immature: pedicel; [X-20-74 Phidippus princeps (Peckham) immature male: right book lung; IX-20-74 immature female: right book lung; [X-20-74 Phidippus putnamii (Peckham) male: on abdomen just above pedicel; Jackson County*; VIII-30-no year Appendix II 121 Phidippus whitmani Peckham immature female: left book lung; IV-2-76 immature female: dorsal pedicel; IV-4-76 Thomisidae Misumenoides aleatorius (Hentz) female: 2 larvae—left side of pedicel, left book lung; IX-3-74 Misumenops sp. immature: pedicel; [X-24-74 Misumenops celer (Hentz) female: in egg sac; Pope County*; V-19-71 Philodromus vulgaris (Hentz) female: in egg sac; Jackson County*; VI-5-71 female: dorsal pedicel; III-16-76 Tibellus duttoni (Hentz) immature female: on left abdomen just above pedicel; Jackson County*; no date Xysticus ferox (Hentz) female: right book lung; IV-15-76 Xysticus funestus Keyserling male: 2 larvae—ventral pedicel, between leg bases; IX-10-74 Literature Cited Batra, S. W. T. 1972. Notes on the behavior and ecology of the mantispid, Climaciella brunnea occidentalis. J. Kansas Entomol. Soc. 45: 334-40. Beatty, J. A., and J. M. Nelson. 1979. Additions to the checklist of Illinois spiders. Great Lakes Entomol. 12: 49-56. Biraben, M. 1960. Mantispa (Neuroptera) parasita en el cocdn de Mete- peira (Araneae). Neotropica 6: 61-64. Bisset, J. L., and V. C. Moran. 1967. The life history and cocoon spinning behavior of a South African mantispid (Neuroptera: Mantispidae). J. Entomol. Soc. Southern Africa 30: 82-95. Borror, D. J., and D. M. DeLong. 1971. An introduction to the study of insects. 3rd ed. Holt, Rinehart and Winston, New York. Brauer, F. 1852. Verwandlungsgeschichte der Mantispa pagana. Archiv ftir Naturgeschichte 18: 1-2. Brauer, F. 1855. Beitrage zur Kenntnifi der Verwandlung der Neuropte- ren. Verh. Zool.-Bot. Ges. Wien 5: 479-84. Brauer, F. 1869. Beschreibung der Verwandlungsgeschichte der Mantispa styriaca Poda and Betrachtungen tiber die sogenannte Hypermeta- morphose Fabre’s. Verh. Zool.-Bot. Ges. Wien 19: 831-40. Bristowe, W. S. 1932. Mantispa, a spider parasite. Entomol. Mon. Mag. 68: 222-24. Capocasale, R. 1971. Hallazgo de Mantispa decorata Erichson para- sitando la ooteca di una Lycosa poliostoma (Koch) (Neuroptera, Mantispidae; Araneae, Lycosidae). Rev. Brasileira Biol. 31: 367-70. Davidson, J. A. 1969. Rearing Mantispa viridis Walker in the laboratory (Neuroptera: Mantispidae). Entomol. News 80: 29-31. Eltringham, H. 1932. On an extrusible glandular structure in the abdomen of Mantispa styriaca Poda (Neuroptera). Trans. Entomol. Soc. London 80: 103-105. George, L. D., and N. L. George. 1975. Notes on the three hundred and fifty-eighth meeting: A new record of mantispid reared from spider. Pan-Pacific Entomol. 51: 90. 122 Literature Cited 123 Gilbert, C., and L. S. Rayor. 1983. First record of mantisfly (Neuroptera: Mantispidae) parasitizing a spitting spider (Scytodidae). J. Kansas Entomol. Soc. 56: 578-80. Handschin, E. 1960. Beitrage zu einer Revision der Mantispiden (Neuroptera). II. Teil. Mantispiden des ‘“‘Musee Royal du Congo Belge’ Tervuren. Rev. Zool. Bot. Africaines 62: 181-245. Hoffman, C. H. 1936. Notes on Climaciella brunnea var. occidentalis Banks (Mantispidae—Neuroptera). Bull. Brooklyn Entomol. Soc. 31: 202-203. Huber, I. 1958. Color as an index to the relative humidity of plaster of Paris culture jars. Proc. Entomol. Soc. Washington 60: 289-91. Hungerford, H. B. 1936. The Mantispidae of the Douglas Lake, Michigan, region with some biological observations. Entomol. News 47: 69-72, 85-88. Hungerford, H. B. 1939. A note on Mantispidae. Bull. Brooklyn Entomol. Soc. 34: 265. Kaston, B. J. 1938. Mantispidae parasitic on spider egg sacs. J. New York Entomol. Soc. 46: 147-53. Kaston, B. J. 1940. Another Mantispa reared. Bull. Brooklyn Entomol. Soc. 35: 21. Kaston, B. J. 1948. Spiders of Connecticut. State of Connecticut, Hartford. Killebrew, D. W. 1982. Mantispa in a Peucetia egg case. J. Arachnol. 10: 281-82. Killington, F. J. 1936. A monograph of the British neuroptera. Vol. I. The Ray Society, London. Kishida, K. 1929. On the oviposition of a clubionid spider, Chi- racanthium rubicundulum. Lansania |: 73-74 (In Japanese). Kuroko, H. 1961. On the eggs and first-instar larvae of two species of Mantispidae. Esakia 3: 25-32. Lucchese, E. 1955. Richerche sulla Mantispa perla Pallas (Neuroptera Planipennia—Fam. Mantispidae). Ann. Fac. Agrar. Univ. Stud. Perugia 11: 242-62. Lucchese, E. 1956. Richerche sulla Mantispa perla Pallas (Neuroptera Planipennia—Fam. Mantispidae). Ann. Fac. Agrar. Univ. Stud. Perugia 12: 83-213. McKeown, K. C., and V. H. Mincham. 1948. The biology of an Australian mantispid (Mantispa vittata Guerin). Australian Zool. 11: 207-24. MacLeod, E. G., and K. E. Redborg. 1982. Larval platymantispine mantispids (Neuroptera: Planipennia): Possibly a subfamily of generalist predators. Neuroptera Intern. 2: 37-41. 124 Mantispa uhleri Banks MacLeod, E. G., and P. Spiegler. 1961. Notes on the larval habitat and development peculiarities of Nallachius americanus (McLachean) (Neuroptera: Dilaridae). Proc. Entomol. Soc. Washington 63: 281-86. Maerz, A., and M. R. Paul. 1930. A dictionary of color, 1st ed. McGraw- Hill, New York. Main, H. 1931. A preliminary note on Mantispa. Proc. Entomol. Soc. London 6: 26. Metcalf, R. L., and W. Luckmann. 1975. Introduction to insect pest management. John Wiley and Sons, New York. Milliron, H. E. 1940. The emergence of a neotropical mantispid from a spider egg sac. Ann. Entomol. Soc. Am. 33: 357-60. Nesbitt, H. H. 1945. A revision of the family Acaridae (Tyroglyphicae), order Acari, based on comparative morphological studies. Part I. Historical, morphological, and general taxonomic studies. Canadian J. Research Section D Zoological Sciences 23: 139-188. New, T. R., and A. J. Haddow. 1973. Nocturnal flight activity of some African Mantispidae (Neuroptera). J. Entomol. Ser. A Gen. En- tomol. 47: 161-68. Opler, P. A. 1981. Polymorphic mimicry of polistine wasps by a neotropical neuropteran. Biotropica 13: 165-76. Parfin, S. 1958. Notes on the bionomics of the Mantispidae (Neuroptera: Planipennia). Entomol. News 19: 203-207. Peck, W. B., and W. H. Whitcomb. 1978. The phenology and popula- tions of ground surface, cursorial spiders in a forest and a pasture in the south central United States. Symp. Zool. Soc. London 42: 131-38. Poujade, G. A. 1898. Observation sur les moeurs de Mantispa styriaca Poda. Bull. Soc. Entomol. France 3: 347. Redborg, K. E. 1982a. Mantispidae parasitic on spider egg sacs: An update of a pioneering paper by B. J. Kaston. J. Arachnol. 10: 92-93. Redborg, K. E. 1982b. Interference by the mantispid Mantispa uhleri with the development of the spider Lycosa rabida. Ecol. Entomol. 7: 187-96. Redborg, K. E. 1983. A mantispid larva can preserve its spider egg prey: Evidence for an aggressive allomone. Oecologia 58: 230-31. Redborg, K. E., and E. G. MacLeod. 1983a. Climaciella brunnea (Say) (Neuroptera: Mantispidae): A mantispid that obligately boards spiders. J. Nat. Hist. 17: 63-73. Redborg, K. E., and E. G. MacLeod. 1983b. Maintenance feeding of first instar mantispid larvae (Neuroptera, Mantispidae) on spider (Arach- nida, Araneae) hemolymph. J. Arachnol. 11: 337-41. Literature Cited 125 Rogenhofer, A. 1862. Beitrag zur Kenntnis der Entwicklungsgeschichte von Mantispa styriaca Poda (pagana Fab.). Verh. Zool.-Bot. Ges. Wien. 12: 583-86. Sheldon, J. K., and E. G. MacLeod. 1971. Studies on the biology of the Chrysopidae II. The feeding behavior of the adult of Chrysopa carnea (Neuroptera). Psyche J. Entomol. 78: 107-21. Smith, R. C. 1922. The biology of the Chyrsopidae. Cornell Univ. Agr. Exp. Sta. Mem. 58: 1287-1372. Smith, R. C. 1934. Notes on the Neuroptera and Mecoptera of Kansas, with keys for the identification of species. J. Kansas Entomol. Soc. 7: 120-45. Stein, R. J. 1955. An insect masquerader. Nat. Hist. 11: 472-73. Valerio, C. E. 1971. Parasitismo en heuvos de arafia Achaearanea tepidariorum (Koch) (Aranea: Theridiidae) en Cost Rica. Rev. Biol. Trop. 18: 99-106. Viets, D. 1941. A biological note on the Mantispidae (Neuroptera) J. Kansas Entomol. Soc. 14: 70-71. Withycombe, C. L. 1925. Some aspects of the biology and morphology of the Neuroptera. With special reference to the immature stages and their possible phylogenetic significance. Trans. Entomol. Soc. London 72: 303-411. ey at ct sestihin/ h abi pense wrth VE NAN Seer ya mig ‘hl it Tis Sa oe Dihientae). Pye saith’ vet Why Heol aod! aon ana abe nie ey Nabe “fs Ye totyesiod iynibes) sAtT AT ae Mini, OE, hash gai See atin ee. WE ad hee ewe ae AEH wales) wy ait) goloid wit pod a ee yo 2O Linklinggepe Rate Caqientt EO SE? He dqusaent as begins ely wdeoloh ata D4, me , Vioghiatel -OpRS bo) ij ‘ ie Rea or +" ev so otk. b Werk eats 7 ‘a re Tule? Index (Page numbers in boldface indicate illustrations.) Achaearanea tepidariorum, 4, 8, 9, 29, 31, 54, 38, 39, 40, 42, 45, 46, 47, 80 Acheta domestica, 59 Admestina tibialis, 27 Adult: Mantispa uhleri Banks, 18, 19; collections, 1974, 83; 1975, 84; man- tispids: 1, 4, 6; coloration, 18, 19; mea- surements, 13; emergence times, 91, 93; preying mantids, | (see also Ma- turity; Spider) Agelena naevia, see Agelenopsis naevia Agelenopsis emertoni, 78 Agelenopsis kastoni, 76 Agelenopsis naevia, 2 Agelenopsis pennsylvanica, 3 Agelenopsis sp., 27, 28, 31, 35, 39, 40, 43, 45, 75, 78, 79, 80 Anesthesia, CO,: advantages during collecting, 82; mantispid, during col- lection, 82; for spider examination, iy 52,"56; 57, 60) 73, 76; for*spider immobilization, 37, 52; spiders after mating, larval movement on, 60 Araneidae, 106 Araneus, 31, 106 Arctosa littoralis, 2 Arena for laboratory experiments, 35, 38 Argiope, 31 Ariadna bicolor, 27, 31, 75, 76, 78, 79, 80 Associations: mantispid larvae with spiders, 77, 78 Baeus sp., 4 Bagworm moth. see Thyridopterix ephemeraeformis Behavior: after recovery from CO, anesthesia (spiders and mantispids), 56; book lung-entering, male/female differential, 66, 69, and spider size, 66; egg-sac penetration, 34, 39-40; epi- gamic, 72-73; of spiders, during board- ing by larvae, 34, 52; and larval dia- pause, 90; predatory, 73, searching, larval, 39, 43-44, 53, 73, 79, 83, 94, 96; transfer, of larvae during spider mat- ings, 72 Boarding: behavior of larvae, 2, 52, 53, 97 (see also Behavior); of spiders, in nature, 34, 80; multiple, 54. See also Spider boarding Book lung, spider: damage to, 68, 69; and egg sac access, 69; entry by larvae, 57, 64, 65, 66, 67, 68, 70; larvae too large to fit, 66; preference for male/ female, 66, 68, 80; larval exit from, 70; size increased after fifth instar, 66 Cabbage loopers. See Trichoplusia ni Cannibalism: 10, 46, 56, 60, 64, 73, 98; and comparative size, 24; induced, 60; larval transfer to new spider during, 73 Carya ovata, 76. See also Shagbark hickory Chrysopa oculata, 16 Climaciella brunnea, 4 Cocoon, M. uhleri: 2,7; remains of, in nature, recognition, 87; spinning, 3, 9, 16, 21, 23, 42, 46, 47, 87, 88, 89, 96 Collecting: locales, 6, 8, 9, 55-56, 57, 82, 83, 98; M. uhleri, 81-82; spiders, 75-76; techniques, 75-76 127 128 Collections: spider, 74, 75, 94; adult mantispids, 74, 75, 82, 83, 84, 85, 86, 87, 94, 95, 98 Color (see also Pigmentation): adult, 18, 19; eggs, 14; larva, 14, 16; standard reference, 11 Competition, larval, 41, 43, 45, 46, 54 Confinement, contrary to natural con- ditions, 34 Coniopterygidae, 18 Convergent evolution, | Copulation. See Mating Courtship, spider, 56 Cupiennius sallez, 3 Cuticle: larval, discoloration, 10, 40, 41, 42, 47; spider: penetration by larva, 68; shedding (ecdysis), 15-16, 39, 64, 65, 70; tanning, 67 Damage: to larvae by spiders, 10, 32, 39, 41, 42; to spiders by larvae, 68 Death. See Mortality Development. See entity involved Developmental rates: laboratory, 15, 29, 30, 32, 40, 41, 47; field, 86 Diapause, 86, 90, 91 Dimorphism, sexual, 19 Disappearance of larvae from spiders, 68 Drassodes hypocrita, 3 Drosophila, 49, 57; culture, as food source for spiderlings, 59 DSAC collection, 76, 79 Eating of mantispid larvae by spiders, 50, 53, 64, 71 Ecdysis (see also Molt): mantispid, 1, 11; to adult, 18; in laboratory condi- tions, 7, 11, 15, 56, 59; spider, 59, 64; while carrying mantispid larva, 56, 64, 70-71 Eclosion (hatching), 7, 39-40 Egg: clutch, after hatching, 13; in labo- ratory, 6, 21, 44, 86; count method, 11; -laying capacity, 14, 21; bagworm moth, as larval food, 10; sac, spider: age and larval development, 42, 46; Agelenopsis, for laboratory experi- ments, 28; mantispid emergence from, 3; near hatching site, 44; penetration Mantispa uhleri Banks (of): 2, 25, 31, 33, 34, 38, 39, 97; inhibi- tion of, 44, 45; in nature, 3, 43; lack of, 43, 44; and texture of sac, 45; produced in laboratory, 57; spinning, 94; spider, as food for mantispid larvae, 7, 8, 9, 25; mass in sac, 8; toxicity for larvae, 10 Emergence: curve, 95; mantispid, from egg sacs, 3; overwintered M. uhleri, 91; summer generation, 93, 94, 95, 99 Eumantispa harmandi, 3 Evolution, convergent, 1 Experiments, by number: see Contents Feeding (see also Food): ability, larval, 10; strategies of larvae, rationale, 46 Fifth instar larva, 61 First instar larva, 12, 14; ability to feed, 10; on L. rabida, 65, 67 Food: for larvae, allotment of, 12-13; intake and adult size, 21, 22; Man- tispinae: spider(s), eggs, 1, 2, 7, 8; minimum larval intake to reach adult stage, 22-23; Platymantispinae larvae: various insects, sedentary arthopods, 1; spider blood, 33, 43, 51, 64, 68, 73, 90; for spiders, under laboratory con- ditions, 57; supply, larval search for, 96 Freezing of eggs for storage, accepta- bility after, 9 Gas exchange across book lung mem- brane, 68 Generations per year, 92, 93, 94, 99 Glue for spider immobilization, 52 Gnaphosa muscorum, 3 Grooming, spider. See under Spider Habitat, natural, for spiders, 76, 94 Hatching. See Eclosion Herpyllus, 31 House fly. See Musca domestica Humidity: in book lung, 68; relative, for mantispid mating, 10; in rearing chambers (pseudosacs), 6, 7, 10, 35, 42; regulation of, 6 Hypothesis, of inhibitory mechanism preventing egg-sac penetration by larvae, 44 Index Immobilization of spiders with CO, anesthesia, 52 Inhibition, of larval feeding, 40, 41, 46 Inhibitory mechanism, hypothesis re egg sac penetration, 44-45 Instar: see by number—first, etc. Key to larvae, 51, 75 Laboratory experiments: arenas, 35, 38; rearing cages, 61. See also Contents Larva(e): 12, eaten by spiderlings, 53; identification, 76; mature third instar, 17; wild-caught, 51, 54 Larval: age, and egg-sac penetration, 44-45; for experiments, 38; develop- ment, conditions influencing, 42, 46, 90; diapause, factors involved, 90; dis- lodgement during ecdysis, 97; move- ment on spiders, 56, 57, 62, 63, 65; positions on molting spiders, 64, 65; stages, mantispids, predaceous, | Light trapping, 81, 82, 94, 99 Lomamyia, 18 Lycosa: 31; poliostoma, 4; rabida, 47, 55, 57, 58, 62, 63, 65, 66, 67, 68, 69, 70, 12,92, 93 Maclura pomifera. See Osage orange Mantidae, resemblance of Mantispidae to, | Mantids, hemimetabolous, | Mantispa: brunnea. see Climaciella brunnea; decorata, 4; fuscicornis, 2, 6; interrupta, 2, 3, 4, 6, 33; perla: see Perlamantispa perla; pulchella, 6; say, 4, 6; styriaca, 2, 3, 24, 25, 32, 33; uhleri, adult male, 18; development exclusively in spider egg sacs, 96; egg sac penetration, spider boarding, 26- 33; larva, principal activity, boarding of spiders, 62, 97; viridis, 3, 4, 6, 26- 33; viridula, 4; vittata, 2, 32, 33 Mantispid(s): adult, maintenance of, 6; holometabolous, 1; larva(e), “locked in” to food supply, 96; predaceous, 1; -spider associations, data-collecting methods, 74 ff. Mantispinae, 1 Marking and release of adults, 82, 86 129 Mating: mantispid, 10, 19-21; timing, 19; spider, 56; larval position associ- ated with, 56, 60 Maturity: larvae reaching, 41, 73; con- ditions for reaching, 44, 45, 46, 53-54; spider molts before reaching, 56, 65 Measurements: of larvae, 10-11; of adult mantispids, 13; of developmen- tal stages, 15 Measures of dispersion, 11 Meconium: liquid; pellet, 18 Metacyrba undata, 27, 31, 48, 49, 50, 51, 75, 76, 78, 79, 80, 83 Metaphidippus galathea, 73 Metepeira labyrinthea, 4 Mold, lethal to eggs, 7 Molt(ing), 15, 55, 56, 70; and book lung entry/exit, 64-65; spider, larval sur- vival of, 55; after boarding by mantis- pid larva, 57, 66, 97. See also Ecdysis Mortality, larval: and failure to feed, 42; and hatched spiderlings, 39, 42; in presence of spiders and egg sacs, 43, 53; in relation to instars, 41 Movement, larval, on molting spiders, 64-65 Musca domestica: as food for adult mantispids, 6; as food for spiders, 8, 56, 59; sugar cubes as food for, 59 Nail polish, colored; wing marking for identification, 82 Natural selection, possible, of heritable trait, 73 Nutition(al): minimum, for M. uhleri larvae, 22-23; state of spider, and effect on spider boarding by larvae, 53 Osage orange (Maclura pomifera), 9, Overwintering, 32, 79, 83,91, 92, 93, 95, 98 Oviposition: adult mantispids, 13, 23, 96; adult spiders, 97 Pardosa milvina, 27 Pedicel, spider: position of larva on, 2, 64, 65, 68, 70, 71, 73; M. whleri larval preference for, 97 Perlamantispa perla, 3, 33 130 Pheromone, male, 20, 24 Phidippus audax, 8, 9, 27, 30, 31, 47, 49-53, 55-57, 58, 60, 63, 65, 66, 69-73, 86, 92, 93 Philaeus militaris, 4 Philodromus vulgaris (crab spider), 27, 35, 86, 87, 91 Pigmentation (see also Color; Cuticle): discoloration of book lung, 68; dis- tinguishing M. uhleri, M. viridis, 26 Platymantispinae, | Pseudoplusia includens, 59 Pseudosacs, 6, 7, 8, 9, 29, 39, 41 Pupa, 2, 16 Purging mechanism, possible; in fe- male spiders, 76 Rearing: chambers for mantispid lar- vae, see Pseudosacs; spiders, 57; cages for, 61; conditions, for mantispid lar- vae, 10, 55 Relative humidity. See Humidity Restraint, experimental, of spiders, 35- 30) 45,002 Salticid, 4, 53 Salticus scenicus, 31, 52 Schizocosa: ocreata, 93; sp., 27 Sclerotization, book lung margins; bar to larval entry, 67, 71 Search, larval. See Behavior Second instar larva, 8, 12, 16 Sexual dimorphism, 19 Shagbark hickory, spider habitat, 76, 86 SIUC collection, 76, 79 Sixth instar, 61 Size variation, of adult mantispids, 11, 13, 21-24; advantage, 23 Spider(s): behavior (see also Behavior), epigamic, 72-73; blood, for mantispids, 33, 51, 64, 68, 90; boarding: 25, 30, 33, 43-44, 57; larva on pedicel, and ab- dominal “‘pits,’”’ 64; larval fate, on male spiders, 55; larval preferences, 49; in nature, 34, 55; and nutritional state, 53; of web-builder, in laboratory, 80; book lung, q.v.; cannibalism, q.v.; Mantispa uhlert Banks control group, 60; cursorial, seasonal displacement, 93; ecdysis (q.v.; see also Molt) while carrying mantispid larva, 56; egg(s) (q.v.): age of, and effect on larval development, 42; per clutch, proportional to body size of female, 96; development, suitability for food, 8, 9, 10; entry facility of, 70; in laboratory, for mantispid rearing, 8-10, 29, 57; food in laboratory, 57; sac, in nature, 2; grooming, 53, 64, 68; hunting species, 74-75, 79, 98; imma- ture (see also Spiderlings), larval boarding of, 55; long-legged, and eat- ing of larvae, 53; mating, 56, 72-73; molt (see also Ecdysis; Molt), and larval entry of book lung, 66; mature, molts to reach; larval movement, 57; movement restricted during cuticle tanning, 67; naturally infested with M. uhleri, 50; nonhunting species, 79, 98; sex, and larval preference, 65; size: correlation with mantispid larva, as food, 53, 55; and larval boarding, 50, 55, 66; tanning, q.v.; web-building/ spinning, 74-75, 78, 98 Spiderlings, 3, 10, 46; eating larvae, 50 Statistical tests, 13, 29, 36, 48, 55 Storage, egg sac, 8 Sympatric occurrence of mantispid species, 74 Tanning, spider cuticle, 67-68, 70, 71 Tarantula sp., 4 Temperature, 85 Texture of egg sac, 45 Third instar larva, 8, 12, 16, 17, 21 Thyridopterix ephemeraeformis, 10 Time element, for adult mantispid growth, 47; in developmental phases, 47 Toxicity of certain eggs for larvae, 10 Transfer of larvae from spider to spider, 63, 72; book lung to book lung, 67 Trichoplusia ni (cabbage looper), 3, 10 Wesmealius quadrifasciatus, 16 Wild-caught larvae, 51, 54 A Note on the Authors KURT E. REDBORG received his Ph.D. from the University of Illinois, at which he has subsequently been a Research Associate in the Department of Agronomy, a Visiting Specialist in the Department of Physiology and Biophysics, and now is Education Services Coordinator at the Thames Science Center in New London, Connecticut. Extensively interested in insects throughout his youth, he collected and became intrigued with the family Mantispidae while serving as a teaching assistant for a course in insect taxonomy. He is particularly interested in behavioral ecology and evolutionary theory, and plans to use the family Mantispidae as a tool in their study. This is his first book. It was awarded the 1981 Balduf Research Award by the Department of Entomology for excellence in entomological research and publication among graduate students. ELLIS G. MacLEOD received his Ph.D. from Harvard University and is currently Associate Professor of Entomology/Genetics and Develop- ment at the University of Illinois. He is interested in all aspects of the evolution of insects. These include the mechanisms of speciation, as well as an analysis of the broad-scale features of the adaptive radiation of the insect orders as deduced from the fossil record and from comparative studies of living species. His research has dealt with numerous facets of the biology of the Neuroptera, including their reproductive cycles and an account of the taxonomy and zoogeography of the extinct Neuroptera of the Baltic amber. veh . } Vary (NMR tis 2 i Te ee a cin ot =A | . \ Ans ee) a inept ™ ; “ah iii " cre tig % 9, Wh, fad iA iM ay RY A r By iiss Rb I May RR ey alm 2s re HN Rel, as Mei aL aA ‘ NY Ata! : ; ve hwvenvudl ein aval 0 ; etre tp = & Ve in) ie ie wi eaape Hwa yey Tait: pine pat | \ f teh i ots Hy tte a7 fe Wii Me «joer tel rc ew ae i OP Oe a ae ‘ i Joven 4 ; i Cio rhy.rh) : i it J } = Fri) i ee AC RL bene 16 amy hh gees wiAS TVPG o i * . , ond GAP © ye hs ied ‘ae veya ah ie Ch) | a BS t Atel om ab aie hae wel iva! ef mame! heen roa Argh cos 7 wey SR i bert a nye) 0 A aa ye Re re ty ae Camry Ty te we MPs), Wash PUT MOH a, Ts eee it 4 yy Th. ; Wire F 4 8 Th i PLM ob ee! ee Le Trees OF ‘ Piik Tin’ 4jfA Pome WL Tmt ee US er ed iA A {iiaptig TIA ‘ate fie) eee iy teks a aay By ao ky C18), eerste oy suey : : ha a OEE LT RE EO elie RP a Tei {Ue Saha pa 4) POM eh can tLe eed ae ye ar Ph yest Sh Lithia “ie Phu ey Lane, df venon 4 ee ah fe wea a { eer a Wis pee hye ty Ped’ Cron el ide ee ta (ng YY Me, iy oe tes i ay ee Wh" ae Mi gg by Pony i boas VP ee 197! sya y ravi a ie ‘equ wPr) AP, AER's 1) GD aM 7 cen Cee RE a site hes ds at Sie? wy iy hy ab Ai Tito ~ iM a i “4 . i ahd rn aiir ad’ Cee i mney 1 Oe 7 Poe vie bt Vertis ra Pan 4 ‘ f et a” 7 > i j na = Hf i ) v 4 Ps i y ; } ; * ‘ as ' ; a F ; Ht : ii i va >) } La til f ie 12, a) ve Oe ki BUNS Ural ; ye 4 a yo) ieee (oe ; ‘ A bi) x j \ ery 7 a Sia we yi ot) ee rine ay 4 UNIVERSITY OF ILLINOIS-URBANA 570.5ILL C001 ILLINOIS BIOLOGICAL MONOGRAPHS$URBANA it | 01 INN 6675 51-55 1981-86 ve ‘ . ve Ke Se Se Ue rarNn aise Peay EN ENE oie ‘ AY eA SHE in . vara he SePPHeeriht te Ce. . “ Pere peli Wy ‘ ve werent sc ee x “ . i veiiy nan Ly ht i ‘ ae ‘ 1 , ‘ ak mane ' : rene pcan ¥ we