Table Of Content

Memoirs of the American Entomological Society Number 48 The Gall Midges (Diptera: Cecidomyiidae) of Hickories (Juglandaceae: Carya) By Raymond Gagne J. SystematicEntomologyLaboratory,PSI AgriculturalResearchService,U.S.DepartmentofAgriculture c/oSmithsonianInstitutionMRC-168 P.O.Box37012 Washington,D.C.20013-7012 USA . (e-mail:[email protected]) — ABSTRACT Thismonographtreats the63 speciesofgallmidgesknowntofeedonNorthAmericanhickories {Carya:Juglan- daceae). The distributionand historyofhickories are outlined. Oncewidelydispersed across the Holarctic Region duringthe Eoceneepochand later,hickoriesarenowlimited inrange toeasternNorthAmericaand a smaller area insoutheasternAsia. Fifty-sixspeciesbelongtothesinglegenusCqryomyiathatisknownexclusivelyfromhickories. Eachisunivolrineand formsa characteristic,complexleafgallinwhichtheirlarvaeeventuallypupatethefollowingspring.Thesespeciesexhibitagreatnum- berofstrikingbiologicalandmorphologicaltraits.Sixotherspeciesinfourothergeneraareresponsibleforonecomplexandfive simplegallsontheleavesornuthusksofhickories,whileoneotherisaninquilineingallsformedbyCaryomyiaspp.Morphol- ogy,relationships,distribution,andbiologyofeachofthegallmidgetaxaaretreatedinturn.Keysfortheidentificationofhick- orygallsmadebygallmidgesand tolarvaeofcecidomyiidsonhickoriesareprovided. Nineteenofthe63 specieswereprevi- ouslynamed.Fifteenofthemareredescribedandtheremainingfourarerelegatedtojuniorsynonymy.Onespeciesisrenamed andanewgenusisdescribedforanother.Alreadydescribedspeciesare:CaryomyiaansericollumGagne,newnameforCaryomyia caryaecolaFelt;CaryomyiaantennataFelt;Caryomyiacaryae(OstenSacken) (=Caryomyiaarcuaria (Felt),newsynonym);Caryomyia caryaecola(OstenSacken);Caryomyiacynipsea(OstenSacken);Caryomyiaholotricha(OstenSacken) (=Caryomyiacaryae(Felt),new synonym);CaryomyiainanisFelt;Caryomyia-persicoid.es(OstenSacken);Caryomyiasanguinolenta(OstenSacken);Caryomyiatubicola (OstenSacken);Caryomyia thompsoni(Felt) (= CaryomyiaconsobrinaFelt,CaryomyiasimilisFelt,newsynonyms);Gliaspilotagluti- nosa(OstenSacken),newcombination;Harmandiolanucicola(OstenSacken),newcombination;andParallelodiplosiscaryae(Felt). Twonewgenera,CaryadiplosisandGliaspilota,aredescribed,asare49newspecies,asfollows:Caryadiplosisbiconvexa,Caryadiplo- sisvenicola, Caryomyiaaggregata,Caryomyiaalbipilosa, Caryomyiaarcuata, Caryomyiaasteris, Caryomyia biretta, Caryomyia caminata, Caryomyiacilidolium,Caryomyiaconoidea,Caryomyiacucurbitata,Caryomyiadeflexipili,Caryomyiaechinata,Caryomyiaeumaris,Cary- omyiaflaticrnstum,Caryomyiaglauciglobus,Caryomyiaglebosa,Caryomyiaguttata,Caryomyiahirtidolhim,Caryomyiahirtiglobus,Cary- omyia inclinata, Caryomyia inflata, Caryomyia lenta, Caryomyia levicrustum, Caryomyia leviglobus, Caryomyia marginata, Caryomyia melicrustum, Caryomyia ovalis, Caryomyiaprocumbens, Caryomyiapurpurea, Caryomyia recurvata, Caryomyiashmoo, Caryomyiasphe- rica,Caryomyia spiniglobus, Caryomyiaspinulosa, Caryomyia stellata, Caryomyiastriolacrustum, Caryomyiastriolata, Caryomyiasubu- lata,Caryomyiasupina,Caryomyiatuberculata,Caryomyiatuberidolium,Caryomyiatumida,Caryomyiaturbanella,Caryomyiaturbinata, Caryomyiaurnula,Caryomyiaviscidolium,Contariniabulliformis,andContariniacucumata. THEGALLMIDGES(DIPTERA:CECIDOMYIIDAE)OFHICKORIES JUGLANDACEAE:CARYA) ( INTRODUCTION while State Entomologist and afterwards he named 1080 species of gall midges from North Cecidomyiidae, the gall midges, are a large America and elsewhere. Thatnumber is phenom- family ofover 6,000 described species worldwide enal, even taking into account the standards of (Gagne 2004 & subsequently). They are of great the day, but the accomplishment left him little economic and ecological interest because at least time toreexamine thebodyofworkasitgrew. He two-thirds of their number feed during their lar- nonetheless left to posterity an overall well- valstageonfloweringplants,someofagricultural mounted and labelled collection of type material and horticultural importance. Many cecidomyi- as well as a major work on galls, Plant Galls and ids, by means of their larval salivary secretions, Gall Makers (Felt 1940). The book is somewhat cause their hosts to produce galls that surround dated,butstill usefuland uniqueinthatittreated the larvae and produce their food. Except for the all of the North American insect and mite gall moreeconomically importantpests, however,Ce- makers known to 1940, including what are still cidomyiidae remain one of the least well known the more noticeable and common gall makers of families of Diptera. A case in point is the present our fauna. Felt acknowledged Osten Sacken's monograph on the gall midges of hickory. The species and named nine more from hickory, all number of species treated in this study is aug- from the Albany, New York, vicinity except one mented from the 14 previously known from hick- from Massachusetts. Two species were based on ories to 63, 49 ofthemnew to science. adults netted on hickory (Felt 1907) and the re- The first notice of hickory gall midges was maining seven were reared from galls (Felt 1908, published in 1862 by Charles R. von Osten 1909, 1921), but his association of the reared Sacken. AEuropean amateur dipterist, he lived in species with particular galls wavered in consecu- the United States between 1856 and 1877, first as tive papers. Felt (1909) proposed the name Cary- Secretary of the Russian Legation in Washington, omyia for the hickory gall midges and later (Felt later as Consul General of Russia in New York, 1921) gave an overview of the included species. and finally as a retired person before returning Four of Felt's nine species are treated as junior permanently to Europe (Osten Sacken 1904). His synonyms here, and another, a junior homonym, 1862paperpresented the firstintroductory sketch is renamed. and synopsis of North American Cecidomyiidae. NoticesofsomeoftheseOstenSackenandFelt In that long article, still a delight to read for the species were published subsequently, including simple, straightforward prose, Osten Sacken de- summaries in Felt (1940), the posthumous Felt scribed and named many species of gall midges (1958) (as revised by Remington & Remington), he found in the vicinity of Washington, DC. and Gagne (1989), but only one further study, by Amonghisdiscoverieswasasuiteofeightspecies BertramW. Wells (1916), treated the cecidomyiid- from hickories. Seven of them form complex, de- caused hickory galls in a comprehensive manner. tachable leaf galls and belong to the genus Cary- Fascinatedbygallarchitecturewhileabotanystu- ornyia. The eighth species causes a sticky leaf spot dent and later assistant professor at Ohio State and is placed here in a new genus, Gliaspilota. Os- University, Wells published several papers on tenSackenreared theadultstagefromonlyoneof galls. In his 1916 paper, he made effective draw- the detachable galls; all the remaining species ings, whole and in section, of 28 kinds of ce- weredescribed chieflyfromtheirgalls,witha line cidomyiid hickory galls that he collected, mainly or two on the larvae of some. In 1870 Osten in southeastern Ohio. His two plates are repro- Sacken described a ninth species, placed here in duced here as Figures 1-33. Wells's drawings of Harmandiola, from swellings onhickorynuthusks gallsinlongitudinalsectiondemonstrated theim- and associated larvae that he collected in portance ofinternal gall structure, including wall Rhinebeck, NewYork. thickness, any intrusions from the exterior, and Additional species from hickory were named how each is attached to the leaf. He attempted to between 1907 and 1921 by Ephraim Porter Felt, assign Osten Sacken's and Felt's species names to distinguished State Entomologist of New York the galls but refrained from coining new names from 1898 to 1928. Felt was involved with many himself because he believed that task should be statewide insect problems, including mosquitoes done by those studying the insects (Wells 1916). and shade tree insects,butas a major side interest Wells subsequently moved to North Carolina RAYMOND GAGNE J. where he distinguished himself in the field of found.Thesegeneraareincludedinthelarvalkey plant ecology. Following his 1916 paper and the provided farther on but are not treated further in reviewbyFelt (1921), essentiallynothingnewhas thispaperexceptto listthemhere, as follows: been published on hickory gall midges for these Lestodiplosis. This is a large worldwide genus lasteighty-sevenyears. of over 160 described species that feed as larvae Table 2, pp. 17-18, lists the 63 species of Ce- on insects and mites (Gagne 2004). Larvae ofsev- cidomyiidae known to feed on American hicko- eral species were found during this study on the ries, allbutoneresponsibleforspecificgalls. They surface of hickory leaves or in galls caused by belong to six separate genera and show remark- Caryomyia,Phylloxera (Phylloxeridae),andvarious able morphological diversity and a variety of life groups of mites (Acarina). Larvae are shown in strategies. Thegenera are as follows: Figs. 48 and 53. These predaceous larvae and the Caryomyia. This genus comprises 56 species few adults reared from them during the present found exclusively on hickories. Each species study cannot now be identified with confidence forms a distinctive complex leaf gall that usually withouta separate revisionofLestodiplosis. dehisces between spring and fall (Figs. 36-164). Dasineura. This is the largest genus of gall Larvae remain in their galls over winter and pu- midges, a catchall category with more than 460 pate the followingspring. described species (Gagne 2004). On one occasion Caryadiplosis. This is a new genus that appears in early May, 2005, two full-grown larvae were to be the closest known relative of Caryomyia. It foundcloselyappressedtoanexpandingleafletof contains two new species that make simple galls, mockernut hickory. The larvae were not associ- oneacircular,biconvexleafblister(Figs. 167-168), ated withany apparentdamage. the other a vein or petiole swelling (Figs. 165-166). When fully fed, larvaeexitthe galls and drop to the soil where they overwinter and then THEHICKORIES pupatethe followingspring. Contarinia. To this large group in its broad Hickories belong to Carya, a genus of the wal- sense are added two new species known only nut family (Juglandaceae) that contains 12 North from their larvae. One species forms a complex American species including pecan. The present sac-like leaf gall (Figs. 172-173) in spring on sev- range of the genus encompasses the eastern half eral hickories, the other causes a bubble leafblis- of temperate North America, including southern ter (Figs. 170-171) onpecan. Ontario and Quebec, eastern United States, and Gliaspilota. Thisnew genus is described for Ce- isolated areas in the mountains of northeastern cidomyia glutinosa (Osten Sacken). Larvae of this Mexico, as well as parts of southeastern Asia species live freely on mucilaginous leafspots that (Grauke 2003, Manchester 1987, Stone 1990). The appear on the underside of new leaves in spring 12 North American hickories form two sections, (Figs. 175-177). Fully-fed larvae soon drop to the Apocarya with fourspecies and Encarya witheight soil to pupate and then emerge as adults the fol- species and one variety (Table 1, p. 4). Additional lowingyear. Carya spp. occur in southeast Asia, three to six Harmandiola. Until recently this group com- species depending on theauthority, andbelong to prised several gall-formers on poplars and chest- a third section, Sinocarya (Stone 1997). No galls nuts. Itnowincludesoneotherspecies,Harmandi- from Asian species have been reported and no ola nucicola (Osten Sacken), from hickory. Larvae sign ofgalls occurs on leaflets oftheir representa- formswellingsonnuthusks (Figs. 178-180). Fully tivesintheherbariumoftheNational Museumof grownlarvaedrop tothe soiltooverwinter. Natural History, Smithsonian Institution, Wash- Parallelodiplosis. A number of species in this ington, DC, so the present study treats only the genus live as parasitoids in galls formed by other gall midges ofNorthAmericanhickories. gall midges and cynipids. One, Parallelodiplosis Hickoriesarenow usedchieflyforfirewood or caryae (Felt), lives in galls of various species of charcoal-making, but their tough, resilient wood Caryomyia (Fig. 48). This species may have more was oncemore commonlyused forsuchthings as thanone generationperyear. spokes for wagon wheels and tool handles In addition to the taxa listed above, undeter- (Grauke 2003). Pecan, originally a Mississippi mined species of two additional genera were basin species, is now widely planted and culti- THEGALLMIDGES(DIPTERA:CECIDOMYIIDAE)OFHICKORIES JUGLANDACEAE:CARYA) ( Table 1. Scientific and commonnames ofNorthAmericanCarya species (from Stone 1997). The correct name for the section Eucarya is Carya, but Eucarya is used in this paper to avoid confusion with the genericname. SectionApocarya: C.aquatica (F. Michaux) Nuttall waterhickory C.cordiformis (Wangenh.)K. Koch bitternut C. illinoiensis (Wangenh.)K. Koch pecan C.palmeriManning Mexicanbitternut SectionCaryaorEucarya: C.floridanaSargent scrubhickory C.glabra (Miller)Sweet pignut C.laciniosa (F.Michaux)Loudon shellbark C. myristiciformis (F Michaux) Nuttall nutmeghickory C.ovata (Mill.)K. Koch shagbark C.ovatavar.mexicana (Helmsley)W.E.Manning Mexicanhickory C.pallida (Ashe)Engler&Graebner sandhickory C. texanaBuckley blackhickory C. tomentosa (Poiret) Nuttall mockernut vated throughout southeastern United States and Europe about 45 million years bp (Manchester abroadforitsnutmeat.Shagbarkandwhitehicko- 1987, Stone 1990). Post-Eocenehickoriesextended ries, especially north of the range of pecan, still across the Northern Hemisphere until after the have some value to people for their nutmeat. All mid-Miocene (ca. 15 million years bp), when the are important to wildlife, particularly tree squir- climate cooled and the individual elements of a rels (Manchester 1987). once uniformArcto-Tertiary forestbegan to sepa- The name "Carya" is Greek for walnut. The rate and undergo extinctions (Wolfe & Leopold common name "hickory" is derived from the 1967).FossilCaryafruithavebeenfoundinAlaska American aboriginal word "pawcahiccora" for a (Lower Miocene), Washington state (Miocene) mash made from the nutmeat. The word was and Colorado (Lower Oligocene), as well as in shortenedtohickoryforthenameofthetreebythe Eurasia (Middle Eocene to Pliocene) from central early Englishcolonists. "Pecan" isanotherAmeri- and eastern Europe, western Siberia, northern canaboriginalwordmeaning"hard-shellednut." China, and Japan (Wolfe & Leopold 1967, Man- Only one other genus ofJuglandaceae,Juglans chester 1987). The present narrow range of the or walnut, occurs in North America. It is widely genus is due to extinctions during the uppermost distributed throughout the temperate Holarctic Tertiary and Pleistocene (Manchester 1987). Region. The only gall midges recorded from During the Pleistocene glacial maxima, hicko- Juglansacrossitsentirerangearetwoundescribed ries and other deciduous trees were much more easternNorthAmericanspecies thatliveas larvae severely restricted in range than presently. Decid- in marginal rolls of leaflets, one a Dasineura on uous angiosperm trees survived insmall, isolated Juglans nigra, or black walnut, the other a Con- populations in the lower Mississippi valley and tatinia on Juglans cinerea, or butternut (Gagne on the southern coastal plain among the more 1989). widespread conifers, as determined by pollen de- Hickories are an ancient genus that was once posits in lakebeds (Davis 1983, Delcourtand Del- much more widely dispersed. Carya-like pollen is court 1984). When the glaciers of the most recent known from the early Paleocene, about60 million iceagebegantorecedeabout12,000yearsago,the years before the present (bp), but the earliest rec- various tree species that survived gradually ognizable Carya fossil fruit is from mid-Eocene spread north on a species basis and not as com- RAYMOND GAGNE J. munities, their individual paths dependent upon & Pratt (1986), and Stone (1997). Keys do not all their particular refugia and individual strategies agree in the number of species. After gaining (Davis 1983). Moving from southwest to north- someexperienceIfollowedElias (1980),Grauke& east, Carya expanded more slowly than some Pratt(1986),andStone (1997) inwhichCaryaovalis other genera, in partbecause of their heavy fruit; is placed under C. glabra and C. septentrionalis un- even so, hickories arrived early (10,000 bp) in the der C. ovata. Hickories are highly prone to hy- upper midwest at sites close to their present bridization (Grauke 1988, 2003) and one notices northernboundary,butwereslowtocrosstheAp- considerable variation in this genus, but only palachian Mountains, arriving in New England rarely did I encounter a tree not readily attributa- onlyabout5,000 yearsbp (Davis 1983). ble to a species. As ithappens, mostcecidomyiids Populations of hickories in Mexico (Map 1) inthisstudydonotappeartodiscernamonghick- may not have been greatly affected by glaciation. ories, particularly those feeding on the Eucarya Hickories there are part of an assemblage of de- section. — ciduous trees generically similarto thatofeastern Collections. Most galls and their cecidomyi- North America. This representation of northern ids used for this studywere collectedby me since trees in Mexico appears to have been in place 1993. Observations and collections were largely since the mid-Miocene when the cooling conti- made at many sites within 20 miles of my resi- nental climate allowed the deciduous forest to dence in Silver Spring, Maryland and at one site reach so far south (Graham 1999). Evidence for near Hedgesville, West Virginia. A major local this is that pollen and spore deposits in Mexico area sampled was the Beltsville Agricultural Re- show a late Cenozoic history of the assemblage's search Center in Beltsville, Maryland where trees presence there. In addition, it is unlikely that the remained undisturbed during this time period soil between central Texas and the relict popula- and where I was free to collect as an employee of tions in Mexico would have supported a rich de- the U.S. Department ofAgriculture. I made three ciduous forest more recently than the mid- separate circuit trips across midwestern and Miocene (Graham 1999). These points will be of southern United States to collect galls, two sepa- interest farther along when considering the rela- rate trips to Florida, several to New England and tive absence of galls on a Mexican population of New York, including once an extension into On- Carya ovatavar. mexicana. tario, and one trip to Tamaulipas, Mexico with a North American hickories vary greatly in dis- stop onthewaytosoutheasternTexas.Additional tribution.Afewspecies, C. cordiformis,C.glabra, C. collections were made for me by collaborators in ovata, and C. tomentosa, range widely and occupy Connecticut, Georgia, and Florida. In addition, I most of the combined range of the genus in studied types and other specimens from the Felt Canada and theUnited States indicated inMap 1. Collection, whichis theproperty ofthe NewYork The remaining species, except pecan, Carya illi- State Museum in Albany but currently on loan to noiensis, which has been extensively planted for the Systematic Entomology Laboratory in Wash- commerce,arerestricted tosmallpartsofthecom- ington,DC,andthefewmiscellaneousspecimens, bined range (Elias 1980). including the existing Osten Sacken types, de- posited overthe years in the NationalMuseum of NaturalHistory, Washington, DC. MATERIALSANDMETHODS Oneneedstocoveralotofgroundtofindhick- ories with galls, especially the less common ones. — Hickory names and identifications. In dis- The best collection sites were forest edges or cussions thatrefer toboth Carya and Caryomyia or open-grown trees where low-hanging branches Caryadiplosis, the abbreviation "C." is used to madecollectingconvenient. Samplingwhile trav- mean Caryomyia or Caryadiplosis; in those discus- eling is always haphazard, dependent on a place sions I use the common names of the hosts to to park a car or obtaining permission from prop- avoid confusion. Both scientific and common ertyowners. Oldchurchyards, cemeteries, and in- names of hickories are listed in Table 1. Hickory terstate highway rest stops, particularly the large identifications were made with the use ofvarious midwestern ones, are productive sites. Paths keys and treatments, including those of Brown & through woods are not as good when time is Brown(1972),Elias(1980),Gleason(1963),Grauke pressing because finding hickories can be uncer- — THEGALLMIDGES(DIPTERA:CECIDOMYIIDAE)OFHICKORIES JUGLANDACEAE:CARYA) ( tain and branches in forests are often too high to peat moss. If larvae normally exit the galls when reach. Some good collecting sites incentral Mary- full grown, they will crawl into the peat moss. In land were lost to development over the course of this way it is possible to obtain all stages of a thiswork: onesiteisnowapublic iceskatingrink speciesand learnsomethingoftheirbiology. Even with large adjacent parking lot, and another a so, I was unable to rear adults or obtain pupae of shoppingcenter. all of the species treated here. Many larvae die Collectingconsistedofharvesting gall-bearing from parasitism or other causes before they are leaflets and placing them in a plastic bag with a fully developed, so one needs a large number of localityandhostlabel. Igatheredgallsatasiteun- galls to successfully obtain gall midge pupae and tilIhadasufficientnumberoftheavailablekinds. adults. Galls of the less common species were Because these galls are small, close examination never found in sufficiently large numbers to suc- was done subsequently, usually in the evening of cessfullyrearadults. — the same day with the aid of a dissecting micro- Adult rearing. I have tried various methods scope. Representative gallswere cutopenand the to rear adults but that outlined in Gagne (1989, larvae removed and saved in alcohol. Samples of 1994) still works satisfactorily, with one change. I whole and cross-sectioned galls were glued to a place galls, or larvae when these escape the galls card and pinned together with a data label for topupateinthesoil, inpotsfilledwithdampened permanent reference. Remaining galls that were peat moss and place them in an outdoor cold excess but not numerous enough to attempt to frame for the winter. Many galls in that situation rear were kept dry in labelled plastic bags for fu- undergo deterioration from the dampness and ture comparison and extraction of larvae as fungi, allowing entry by various arthropods, needed. Larval samples were placed in alcohol mainlymitesandsciarids. Thisisprobablynodif- andlabelled,someeventuallytobeslidemounted ferent than what happens in nature, but it means orusedforSEMphotos. Attheproper season, de- that one has to have a largenumber ofgalls to in- pending on the species, galls with full-grown lar- sure success. At the end ofFebruarywhenIbring vae or, for genera other than Caryomyia, larvae the pots indoors, I place each pot in a separate that had escaped the galls, were kept through clear plastic bag. The bag is large enough to winter to rear adults. Some reared adults were spread well away from the pot at the bottom so pinned but most were kept in 70% ethyl alcohol any emerging specimens are noticeable when the foreventual slidemounting. bag is held against the light. The top ofthe bag is Sweeping or otherwise trapping adults in closedbybringing togethertwoendsofthe open- flight is not practical. Associating both sexes of a ing in a simple knot. Condensation does not col- species is usually not possible for most plant- lectonthe inside ofthebagifthepotisnotoverly feeding gall midges unless they are reared to- wetin the firstplace and thebag is kept outofdi- gether from similar galls. Any sweep of hickory rectsunlight. Ifmoisturedoescondense on thein- branches can yield many species of gall midges, side of the bag, a wetbag canbe changed repeat- some just passing by and having nothing to do edly to a dry one until the pot loses the excess with hickory. While gall midges caught in flight moisture. This seems to be a more efficient may with practice be identifiable to generic level method than the one I formerly used, when I when viewed dried or in alcohol, they must be placed the pots individually in cardboard shoe- painstakingly mounted on microscope slides for boxes, each with a vial fit tightly into an opening further critical viewing, and even then most at the end. With that method I occasionally lost flight-caught phytophagous cecidomyiids cannot somespecimens thatcould squeeze outthrougha usually be determined beyond genus with confi- cover that did not fit tightly or escape when I dence. Slide mounting is a time-consuming activ- opened thecoverfora peek. ity better spent on reared samples. The best Specimens examined and their preparation. methodforobtainingadultsofphytophagousgall Slide mounting of specimens follows the tech- midges is to collect a sufficiency of galls before niqueinGagne (1989, 1994)usingCanadabalsam. they begin dropping from the trees, usually in Most of the larvae and adults studied during this September and October for hickory galls, and study can be traced to particular galls that are ei- keep them over the winter. Galls with full-grown ther on pinned cards or in alcohol. Holotypes of larvae are placed in pots filled with dampened newspeciesandremainingspecimensusedinthis — RAYMOND GAGNE J. study, except those from the Felt Collection, are include: For. = Forest; Mt. = Mountain;Natl = Na- deposited in the insect collection of the National tional; MCZ = Museum of Comparative Zoology, Museum of Natural History in Washington, DC Cambridge, Massachusetts; Ref. = Refuge; St. - (USNM). State; USDA = U.S. Department of Agricuture; — Descriptions. In the species descriptions that USNM = National Museum of Natural History, follow, the gall is described first because it is the SmithsonianInstitution,Washington, D—C. mostobviousmanifestationofa species. Eachgall Species concept within Caryomyia. I consid- descriptionlistselementsofthematuregallinthis ered a distinct gall the phenotypic extension of a order: frequency (common, occasional, or rare), discrete species. In only one case might the gall distributiononhickories (Apocarya or Eucarya sec- distinction seem minor, viz. between C. leviglobus tion or both); arrangement of galls on a leaflet and C. spiniglobus. The first (Fig. 62) has a smooth (single or clustered), position on a leaflet (upper exterior, the other (Fig. 64) iscoveredwithstiffse- orlowerleafsurface,betweenoronveins);height tae. Becausethetwokindswereneverfoundinter- mm or length in and external shape; external tex- spersed, I consider the gall makers distinct, even ture; connection to the leaflet; wall structure in though I cannot separate them morphologically. longitudinal section; and shape and texture ofthe Two other species arenoteworthyherebecause of larval chamber. A concluding sentence or two their variability. Caryomyia thompsonihas two dis- usuallyfollows to differentiatethe gallfromother tinct variations in gall surface, one kind more or gallswithwhichitmightbeconfused. less sparsely haired and the gall surface visible Somespeciesarenamedand describedhereon (Fig. 113), the other with thick hair obscuring the thebasisofonlythegallandlarvabecauseduring gallsurface(Fig. 114).Thesecanoftenbefoundin- the period of this study I did not succeed in rear- terspersedonaleaflet(Fig. 112), and themorphol- ing the adult. In a comprehensive study the ab- ogy ofCaryomyia specimens fromeachis indistin- sence of adults of some species should not pre- guishable. Except for the type specimens under clude describing species from larvae alone. the species name, I made no further distinctionin Larvaehave good distinguishing characters, their the list between the two kinds and considered biology andhosts are known, and theircharacter- them to be made by the same species. The other istic galls canbe found again, possibly to success- special case is C. caryae, in which the galls inter- fullyrear adults. grade. The extreme forms (cf. Figs. 55-56 with Adult abdominal tergal setation is described Figs. 57-58) each seen alone certainly appear dif- from the third tergum posteriad because the pos- ferent and I might have treated them as separate teriorsetalrowsofthefirstandsecondtergitesare species were it not for commonly found inter- always discontinuous mesally. Larval length of- grades. No differences in any stage of the gall fers a rough size estimate. In any instar a larva midgefromthetwoextremesofgallscouldbede- grows considerably, so the larger number of the tected. range isthe moremeaningful one. Lists of collection and specimen localities. GENUS CARYOMYIAFELT Collection localities of galls of each cecidomyiid speciesinthis revisionarelisted alphabeticallyby Caryomyia Felt 1909: 292. Type species, Cecidomyia tubi- state or province and county or town. Voucher colaOstenSacken,byoriginaldesignation.Felt1911: specimensforthegallrecordsaredepositedinthe 56 (review); Felt 1921: 94 (review); Gagne 2004: 100 USNM, either pinned or in alcohol. Listed also, (catalog). with citations, are galls from publications when — accompanied by identifiable figures. A separate Diagnosis. This genus may be differentiated list of larvae, pupae, and adults of each species from other genera ofCecidomyiidiby the follow- studied gives the collection and rearing (if any) ingcombinationofcharacters:Adultwithoutdor- data in alphabetical order by state or province, salprotuberanceonocciput;wing(Fig.245)broad- and town or other site, followed by collection est onbasal half, C unbroken atjuncture with R 5; date, collector if not me, and the associated life legsconsiderablyforeshortened (Fig. 183), shorter stage.Allthesespecimens,includingtypesofnew thanwinglength; abdominal sternites lacking an- species, but excepting the Felt Collection speci- terior pair of trichoid sensilla; gonostylus short, mens, are deposited in the USNM. Abbeviations broad, with wide tooth; protrusible ovipositor ta- — THEGALLMIDGES(DIPTERA:CECIDOMYIIDAE)OFHICKORIES JUGLANDACEAE:CARYA) ( pered abruptlybeyond eighth segment, short, the gate, stiff, pigmented dermal structure along its ninth tergite dorsally with narrow, pigmented, length. Sternites rectangular, 2-3 times as wide as stiff, dermal structure, and cerci partially fused long, without trichoid sensilla; second through dorsally at base (Figs. 184-185). Larva with only sixthsterniteswithhorizontalroworrows ofpos- 1-4 lateral papillae; dorsal thoracic papillae re- teriorsetaeandvariablehorizontalrowsofmostly ducedto4;terminalpapillaeallsimilar,flatorcon- shorter setae near midlength; seventh and eighth vex,asetose(Fig.271). sterniteswithmorenumeroussetaeandbothrows Description. Adult: Head: Eyes connate; adjacent.Genitalia(Figs. 186,247-257):cercishort, facetsmostlyhexagonoid and closelyadjacentex- broad, with setae along margin; hypoproct paral- cept slightly farther apart and circular at mid- lel-sided to broadly widened posteriorly, with height of head. Occiput without dorsal protuber- scattered short setae along posterior and lateral ance. Frons setose. Labrumwithorwithout setae. edges, covered with setulae except, in most Labella separate or coalesced, with or without se- species, dorsum bare frombase to posterior third tae.Palpus1-to4-segmented.Antennawith 12-20 or fourth; aedeagus taperingfrombase to apex or flagellomeres, when 12, often with an additional broad, cylindrical, rounded or notched apically; bud atapexoftwelfthcountedasa thirteenth,but gonocoxite cylindrical, short, broad; gonostylus this one without circumfila; male flagellomeres short, broad, ovate, with dense setae and entirely (Figs. 235, 237, 239) generally foreshortened, bin- setulose. Female abdomen (Figs. 183-185): First odal with foreshortened internodes and necks, or through eighth tergites entire, rectangular, about cylindrical (Fig. 240) and lacking internodes and 6 times as wide as long; vestiture as in male. Sec- necks;eitherwith3circumfila, 1 onbasalnodeand ond through seventh sternites rectangular, about 2 ondistalnode, or the3 evenly distributed along 2.5 times as wide as long, vestiture as in male. flagellomere length, loops of the circumfila sub- Ovipositorslightlyprotrusible, softtorigid, grad- equal in length, short, not attaining bases ofnext ually tapering beyond eighth segment; ninth ter- distal circumfilum, rarely a loop connected be- gitedorsallywithnarrow,pigmented,stiffdermal tweentwocircumfila,orwith9-10interconnected sclerite, laterally and ventrally setose; cerci ovoid horizontal circumfila; female flagellomeres (Figs. or triangular, partially fused anterodorsally, cov- 236, 238, 241) eachprogressively shorter, cylindri- ered with setae but without distal peglike setae; cal with shortnecks, ringed with 2 appressed cir- hypoproctshort,broad,with2posteriorsetae. cumfila connected by two longitudinal bands or Pupa (Figs. 187-214, 267-268): Integument un- several anastomozing bands. Thorax: Scutum pigmented. Head: Vertex convex, on each side with 4 longitudinal rows ofsetae. Scutellum with withanterolateralpairofpapillae,oneofeachpair, lateral group of setae on each side. Pleura with occasionally both, with conspicuous seta. Anten- vestiture only on setose mesepimeron. Wing (Fig. nalbasesforminganteriorlydirectedhornsofvar- 245) broadest on basal half, unmarked, width iouslengthandshape.Anteriormostanglesofface 0.4-0.5 wing length; C unbroken atjuncture with betweenantennalbases and flagellaoftenconvex. R5; R5curved apically,joining C posterior towing Face anterior to midlength smooth or indented apex;Rsincomplete,closertoarculusthantoapex withanextruded posterior lip oneachside. Frons ofRj. Legsgenerallyforeshortened, femora,tibiae and clypeus convex or conspicuously expanded, and tarsi of approximately equal length or pro- base of frons with pair of setose papillae. Labella gressively shorter from femora to tarsi. Tarsal triangular, convex. Thorax: Prothoracic spiracles claws (Fig. 246) untoothed, curved beyond mid- elongate, pointed. Thoracic papillae short-setose, length; empodia as long as claws; pulvilli minus- situatedoneachsideasfollows: 1onprothoraxan- cule, much shorter than claws. Male abdomen: terior and medial to spiracles, 3 along anterior First through seventh tergites entire, rectangular, edge ofnotum, 1 anteriortowingbase, andoneat about 5 times as wide as long; with setae in 1-4 posterior end of notum. Notum with row of tiny rows along posterior marginor onlyposterolater- spiculesoneach side ofmedialsuture. Wingpads ally and with anterior pair of trichoid sensilla but extending just beyond second abdominal seg- no othervestitureexceptformicrosetulae entirely ment,legsextendingapproximatelytoendofthird covering sclerites; eighth tergite may be more abdominal segment. Abdomen: Spiracles sessile. weakly sclerotized than preceding tergites and Firstabdominaltergumofallspeciescoveredwith have fewer setae; tenth tergumwithnarrow, elon- short spicules and additionally with horizontal RAYMOND GAGNE J. row of 6 short-setose papillae barely longer than circumfila of the male flagellomeres (Gagne 1989, spicules;intersegmentaltergalareassmooth,bare; 1994). Below thatlevelthegenusfitsintothelarge second through eighth tergites as for first or with tribe Cecidomyiini that contains most of the Hol- severalhorizontalrowsoflarge,closelyset,spine- arctic plant-feeding Cecidomyiidi, mostly genera like spicules on anterior third of segment. Pleura that lack a dorsal occipital protuberance, have smooth anteriorly and posteriorly but elsewhere usually simple tarsal claws that are curved be- coveredwithsmallspicules,apairofsetosepapil- yond midlength, and larvae with a particular lae present similar to those on terga. Sterna wrin- makeup of the terminal papillae (Gagne 1989). kled, rough,withminutepointedverrucae anteri- While Cecidomyiini typically have four pairs of orlyandlaterally,papillaenotdetected. terminal papillae, three of them short-setose and Larva: Third instar (Figs. 215-231, 269-271): the other corniform, enlarged, and recurved dor- White, cylindrical to ovoid, anteriorandposterior sally (as in Fig. 334), all eight terminal papillae of ends rounded or tapered. Head directed anteri- Caryomyia are without setae and similar in ap- orly or ventrally Antenna about twice as long as pearance (Fig. 271). The relationship of the genus wide. Spatula (Figs. 274-327) present, variously to other Cecidomyiini is nonetheless clear due to shaped. Papillaewithorwithoutsetae, theirbases the fortunate discovery during this study of the prominent or not. Lateral papillae 1—4. Ventral new genus Caryadiplosis that is described below. papilla on each side of spatula closely adjacent to While the two genera are evidently related due to lateralpapillae. Dorsalpapillaeofthorax4,offirst shared, derived adultcharacters, Caryadiplosis has through seventh abdominal segments 4-6. First retained the larval form commonly found in the through eighth abdominal segments always with tribe. In the key to genera of Cecidomyiidae in verrucae laterally, dorsum and venter smooth or Gagne (1981), Caryomyia runs to couplet 135 with lined and/or variously covered with rounded to Contarinia, but that placement does not argue for pointedverrucaeandventrallywithroundedver- anyaffinitybetweenthosetwo genera. rucae or pointed spicules. Terminal segment of Felt (1909) first described Caryomyia to include same length to more than twice as long as eighth eightspecies andeventually added anotherseven abdominal segment, with 8 similar papillae with- (Felt 1921). Three were included only for conve- outsetae,4oneachside,slightlytoconspicuously nience because they occurred on hickories. Felt convex.Analregionwith0-2papillae. Second in- (1909, 1921) wrote that Caryomyia bore some re- star (272-273): White, ovoid. Head directed ante- semblance to Planetella (as Hormomyia), but the riorlyorventrally.Antennaabouttwiceaslongas only grounds he offered was that the two genera wide. Spatula present in all species except C. albi- shared "somewhat the appearance," a suggestion pilosa and C. striolata, with apically pointed tooth that does notbear scrutiny. Felt (1921) laterwrote and elliptical to quadrate shaft. When spatula thatCaryomyia mightbearelativeofPolystepha (as present, collar segmentoneachsidewithpatchof Cincticornia), a genus in the tribe Asphondyliini, enlarged spicules. Papillae with or without setae, because themale antennal circumfila ofboth gen- their bases prominently raised or not. Lateral era are short and Carya and Quercus, the host of papillae 1-4. Dorsal papillae of thorax 4, of first Polystepha, are somewhat related. Because of this, through seventh abdominal segments 4-6. First Caryomyia was listed in Felt (1958) with the tribe through eighth abdominal segments variously Asphondyliini, and that placement was contin- covered dorsally with rounded and or pointed ued in Foote (1965). Male antennal flagellomeres verrucae. First instar (Fig. 35): Translucent to and their circumfila are foreshortened in some white, fusiform at first, becoming flattened and species ofCaryomyia that Felt knew (Fig. 237), but discoid. Spiracles present on first thoracic and had he seen a male of the Caryomyia cilidolium eighth abdominal segments only. Antenna about group with its reticulated flagellomeres (Fig. 240) threetimesas longaswide. hewouldhavehadbetterbutstill superficial sup- Egg(Fig. 34): Elongate-ovoid, red. port. The antennae of the C. cilidolium species group are generally similar to those found in sev- eral species of Polystepha spp. (cf. Fig. 30a, p. 154 Affinities of Felt 1916). We now know that the two genera Caryomyia belongs to the supertribe Ce- are in separate tribes because Polystepha shares a cidomyiidi because of the separate many-looped number of apomorphies with the Asphondyliini 10 THEGALLMIDGES(DIPTERA:CECIDOMYIIDAE)OFHICKORIES JUGLANDACEAE:GARYA) ( (Mohn 1961, Gagne 1994), the most importantbe- gonostylus setulose only on the basal half and ing the enlarged female seventh sternite. That the longitudinally ridged beyond, is the one most antennaeoftwo groups ofspecies in two separate commonly found in the subfamily Cecidomyi- tribes are so similar is only one more example of inae,includingMacrodiplosisand Thecodiplosis,but adaptive homoplasy that can be found in Ce- notParadiplosis. Acompletely setulose gonostylus cidomyiidae. Other examples of parallel modifi- is definitely the less common state in Cecidomyi- cations are listed in the following discussion. ini. It will be noticed that Paradiplosis shares at These serve as a good argument for comprehen- least three characters with Caryomyia and Ca- sive revisions, in which context these happen- ryadiplosis: a sclerite on the distal half of the stances canbe appreciatedbetter. ovipositor, theunbrokencostalvein, andthecom- A relationship between Caryomyia and Cary- pletely setulose gonostylus. This genus is dis- adiplosis is supported by three synapomorphies cussed furtherunder Caryadiplosis. and one other character they have in common. Caryomyia differs from Caryadiplosis in many The first synapomorphy is the presence of a nar- ways, presumablyall apomorphic, as follows: row, elongate, dorsal sclerite on the female ninth segment, the posterior, more rigid half of the ovi- 1.Galls are complex, the result of a complete re- positor(Figs. 258-264,331). Bothgenerahavepro- structuringofthehosttissue,insteadofsimple trusible, yet relatively short ovipositors on which swellings. the structure is presumably used as a stiffener. I 2.Larvae have only one to four lateral thoracic know of a corresponding structure on only two papillae on each side of the midline (Fig. 274) other genera of Cecidomyiini, Macrodiplosis and instead of six lateral thoracic papillae in two Paradiplosis. Macrodiplosis, a large genus of leaf tripletsoneachside (Figs. 335). Suchreduction gallmakers onoaks (Gagne 1989), has a relatively is not uncommon in Cecidomyiinae that form short ovipositor also, but the stiffener extends galls, especially complex galls, but the full only along the posterior two-thirds of the seg- complement is the rule in non-phytophagous ment. InParadiplosis, a genusoffourspecies onfir Cecidomyiinae. and spruce, the ovipositor is extremely long (Os- 3.Larvae have only four instead of six thoracic good and Gagne 1978). dorsal papillae. Except for Caryomyia, all other The second synapomorphy is the unbroken Cecidomyiidi known to me have six dorsal costal vein at its junction with R5. An unbroken thoracicpapillae. Iknow ofasimilarreduction costal wing vein is rare in Cecidomyiini and un- intwogenera,Asteromyia (Gagne 1968) and Se- known in other tribes of the supertribe Ce- mudobia (Roskam 1977), both members of the cidomyiidi. The trait occurs in two cecidomyiine supertribe Lasiopteridibutinseparate tribes. genera that infest Pinaceae, Paradiplosis and The- 4.The larval terminalpapillae are all similar and codiplosis, neither of which seems especially asetose (Fig. 271), instead of dissimilar and ei- closely related to one anotheror to Caryomyia and ther setose or corniform (Fig. 334). This reduc- Caryadiplosis, so I suppose the closed costa tion is uniqueinCecidomyiidi. evolved independently in this tribe at least three 5.Pupation occurs in the gall rather than in the times. Macrodiplosis, mentioned in the previous soil. paragraph,has abrokencosta. 6.Adult tarsi of Caryomyia are foreshortened, no The third synapomorphy is the apparent loss longer than the tibiae and 0.33 or shorter than ofapairofpeglikesensilla usuallyfound nearthe the entire leg, while in Caryadiplosis and most apex of the female cerci. Almost all Cecidomyiidi other Cecidomyiini the tarsi are appreciably have these two (sometimes more) distinctive sen- longer than the tibiae and 0.40 or more than silla that have broad sockets and wide, blunt- the entireleg. tipped setae (as in Fig. 362). Paradiplosis and The- 7.Adultabdominaltergiteshaveneitheralateral codiplosisbothhave the pairofsensilla. group ofsetae (setaethatliejustanteriortobut Another character Caryomyia and Caryadiplosis separate from the posterior row of setae) nor haveincommonisacompletelysetulosegonosty- scales (Fig. 258). In Caryadiplosis and most lus,butIcannotsaywithconfidencewhetherthis other Cecidomyiini, these setae and scales are is primitive or derived. The alternative state, the usuallypresent (seenin Fig. 330).

ebook img

The gall midges (Diptera: Cecidomyiidae) of hickories (Juglandaceae: Carya) PDF

2008·78.4 MB·English
by  
#additional_collections #biostor
Save to my drive
Quick download
Download
Upgrade Premium
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.

Preview The gall midges (Diptera: Cecidomyiidae) of hickories (Juglandaceae: Carya)

Memoirs of the American Entomological Society Number 48 The Gall Midges (Diptera: Cecidomyiidae) of Hickories (Juglandaceae: Carya) By Raymond Gagne J. SystematicEntomologyLaboratory,PSI AgriculturalResearchService,U.S.DepartmentofAgriculture c/oSmithsonianInstitutionMRC-168 P.O.Box37012 Washington,D.C.20013-7012 USA . (e-mail:[email protected]) — ABSTRACT Thismonographtreats the63 speciesofgallmidgesknowntofeedonNorthAmericanhickories {Carya:Juglan- daceae). The distributionand historyofhickories are outlined. Oncewidelydispersed across the Holarctic Region duringthe Eoceneepochand later,hickoriesarenowlimited inrange toeasternNorthAmericaand a smaller area insoutheasternAsia. Fifty-sixspeciesbelongtothesinglegenusCqryomyiathatisknownexclusivelyfromhickories. Eachisunivolrineand formsa characteristic,complexleafgallinwhichtheirlarvaeeventuallypupatethefollowingspring.Thesespeciesexhibitagreatnum- berofstrikingbiologicalandmorphologicaltraits.Sixotherspeciesinfourothergeneraareresponsibleforonecomplexandfive simplegallsontheleavesornuthusksofhickories,whileoneotherisaninquilineingallsformedbyCaryomyiaspp.Morphol- ogy,relationships,distribution,andbiologyofeachofthegallmidgetaxaaretreatedinturn.Keysfortheidentificationofhick- orygallsmadebygallmidgesand tolarvaeofcecidomyiidsonhickoriesareprovided. Nineteenofthe63 specieswereprevi- ouslynamed.Fifteenofthemareredescribedandtheremainingfourarerelegatedtojuniorsynonymy.Onespeciesisrenamed andanewgenusisdescribedforanother.Alreadydescribedspeciesare:CaryomyiaansericollumGagne,newnameforCaryomyia caryaecolaFelt;CaryomyiaantennataFelt;Caryomyiacaryae(OstenSacken) (=Caryomyiaarcuaria (Felt),newsynonym);Caryomyia caryaecola(OstenSacken);Caryomyiacynipsea(OstenSacken);Caryomyiaholotricha(OstenSacken) (=Caryomyiacaryae(Felt),new synonym);CaryomyiainanisFelt;Caryomyia-persicoid.es(OstenSacken);Caryomyiasanguinolenta(OstenSacken);Caryomyiatubicola (OstenSacken);Caryomyia thompsoni(Felt) (= CaryomyiaconsobrinaFelt,CaryomyiasimilisFelt,newsynonyms);Gliaspilotagluti- nosa(OstenSacken),newcombination;Harmandiolanucicola(OstenSacken),newcombination;andParallelodiplosiscaryae(Felt). Twonewgenera,CaryadiplosisandGliaspilota,aredescribed,asare49newspecies,asfollows:Caryadiplosisbiconvexa,Caryadiplo- sisvenicola, Caryomyiaaggregata,Caryomyiaalbipilosa, Caryomyiaarcuata, Caryomyiaasteris, Caryomyia biretta, Caryomyia caminata, Caryomyiacilidolium,Caryomyiaconoidea,Caryomyiacucurbitata,Caryomyiadeflexipili,Caryomyiaechinata,Caryomyiaeumaris,Cary- omyiaflaticrnstum,Caryomyiaglauciglobus,Caryomyiaglebosa,Caryomyiaguttata,Caryomyiahirtidolhim,Caryomyiahirtiglobus,Cary- omyia inclinata, Caryomyia inflata, Caryomyia lenta, Caryomyia levicrustum, Caryomyia leviglobus, Caryomyia marginata, Caryomyia melicrustum, Caryomyia ovalis, Caryomyiaprocumbens, Caryomyiapurpurea, Caryomyia recurvata, Caryomyiashmoo, Caryomyiasphe- rica,Caryomyia spiniglobus, Caryomyiaspinulosa, Caryomyia stellata, Caryomyiastriolacrustum, Caryomyiastriolata, Caryomyiasubu- lata,Caryomyiasupina,Caryomyiatuberculata,Caryomyiatuberidolium,Caryomyiatumida,Caryomyiaturbanella,Caryomyiaturbinata, Caryomyiaurnula,Caryomyiaviscidolium,Contariniabulliformis,andContariniacucumata. THEGALLMIDGES(DIPTERA:CECIDOMYIIDAE)OFHICKORIES JUGLANDACEAE:CARYA) ( INTRODUCTION while State Entomologist and afterwards he named 1080 species of gall midges from North Cecidomyiidae, the gall midges, are a large America and elsewhere. Thatnumber is phenom- family ofover 6,000 described species worldwide enal, even taking into account the standards of (Gagne 2004 & subsequently). They are of great the day, but the accomplishment left him little economic and ecological interest because at least time toreexamine thebodyofworkasitgrew. He two-thirds of their number feed during their lar- nonetheless left to posterity an overall well- valstageonfloweringplants,someofagricultural mounted and labelled collection of type material and horticultural importance. Many cecidomyi- as well as a major work on galls, Plant Galls and ids, by means of their larval salivary secretions, Gall Makers (Felt 1940). The book is somewhat cause their hosts to produce galls that surround dated,butstill usefuland uniqueinthatittreated the larvae and produce their food. Except for the all of the North American insect and mite gall moreeconomically importantpests, however,Ce- makers known to 1940, including what are still cidomyiidae remain one of the least well known the more noticeable and common gall makers of families of Diptera. A case in point is the present our fauna. Felt acknowledged Osten Sacken's monograph on the gall midges of hickory. The species and named nine more from hickory, all number of species treated in this study is aug- from the Albany, New York, vicinity except one mented from the 14 previously known from hick- from Massachusetts. Two species were based on ories to 63, 49 ofthemnew to science. adults netted on hickory (Felt 1907) and the re- The first notice of hickory gall midges was maining seven were reared from galls (Felt 1908, published in 1862 by Charles R. von Osten 1909, 1921), but his association of the reared Sacken. AEuropean amateur dipterist, he lived in species with particular galls wavered in consecu- the United States between 1856 and 1877, first as tive papers. Felt (1909) proposed the name Cary- Secretary of the Russian Legation in Washington, omyia for the hickory gall midges and later (Felt later as Consul General of Russia in New York, 1921) gave an overview of the included species. and finally as a retired person before returning Four of Felt's nine species are treated as junior permanently to Europe (Osten Sacken 1904). His synonyms here, and another, a junior homonym, 1862paperpresented the firstintroductory sketch is renamed. and synopsis of North American Cecidomyiidae. NoticesofsomeoftheseOstenSackenandFelt In that long article, still a delight to read for the species were published subsequently, including simple, straightforward prose, Osten Sacken de- summaries in Felt (1940), the posthumous Felt scribed and named many species of gall midges (1958) (as revised by Remington & Remington), he found in the vicinity of Washington, DC. and Gagne (1989), but only one further study, by Amonghisdiscoverieswasasuiteofeightspecies BertramW. Wells (1916), treated the cecidomyiid- from hickories. Seven of them form complex, de- caused hickory galls in a comprehensive manner. tachable leaf galls and belong to the genus Cary- Fascinatedbygallarchitecturewhileabotanystu- ornyia. The eighth species causes a sticky leaf spot dent and later assistant professor at Ohio State and is placed here in a new genus, Gliaspilota. Os- University, Wells published several papers on tenSackenreared theadultstagefromonlyoneof galls. In his 1916 paper, he made effective draw- the detachable galls; all the remaining species ings, whole and in section, of 28 kinds of ce- weredescribed chieflyfromtheirgalls,witha line cidomyiid hickory galls that he collected, mainly or two on the larvae of some. In 1870 Osten in southeastern Ohio. His two plates are repro- Sacken described a ninth species, placed here in duced here as Figures 1-33. Wells's drawings of Harmandiola, from swellings onhickorynuthusks gallsinlongitudinalsectiondemonstrated theim- and associated larvae that he collected in portance ofinternal gall structure, including wall Rhinebeck, NewYork. thickness, any intrusions from the exterior, and Additional species from hickory were named how each is attached to the leaf. He attempted to between 1907 and 1921 by Ephraim Porter Felt, assign Osten Sacken's and Felt's species names to distinguished State Entomologist of New York the galls but refrained from coining new names from 1898 to 1928. Felt was involved with many himself because he believed that task should be statewide insect problems, including mosquitoes done by those studying the insects (Wells 1916). and shade tree insects,butas a major side interest Wells subsequently moved to North Carolina RAYMOND GAGNE J. where he distinguished himself in the field of found.Thesegeneraareincludedinthelarvalkey plant ecology. Following his 1916 paper and the provided farther on but are not treated further in reviewbyFelt (1921), essentiallynothingnewhas thispaperexceptto listthemhere, as follows: been published on hickory gall midges for these Lestodiplosis. This is a large worldwide genus lasteighty-sevenyears. of over 160 described species that feed as larvae Table 2, pp. 17-18, lists the 63 species of Ce- on insects and mites (Gagne 2004). Larvae ofsev- cidomyiidae known to feed on American hicko- eral species were found during this study on the ries, allbutoneresponsibleforspecificgalls. They surface of hickory leaves or in galls caused by belong to six separate genera and show remark- Caryomyia,Phylloxera (Phylloxeridae),andvarious able morphological diversity and a variety of life groups of mites (Acarina). Larvae are shown in strategies. Thegenera are as follows: Figs. 48 and 53. These predaceous larvae and the Caryomyia. This genus comprises 56 species few adults reared from them during the present found exclusively on hickories. Each species study cannot now be identified with confidence forms a distinctive complex leaf gall that usually withouta separate revisionofLestodiplosis. dehisces between spring and fall (Figs. 36-164). Dasineura. This is the largest genus of gall Larvae remain in their galls over winter and pu- midges, a catchall category with more than 460 pate the followingspring. described species (Gagne 2004). On one occasion Caryadiplosis. This is a new genus that appears in early May, 2005, two full-grown larvae were to be the closest known relative of Caryomyia. It foundcloselyappressedtoanexpandingleafletof contains two new species that make simple galls, mockernut hickory. The larvae were not associ- oneacircular,biconvexleafblister(Figs. 167-168), ated withany apparentdamage. the other a vein or petiole swelling (Figs. 165-166). When fully fed, larvaeexitthe galls and drop to the soil where they overwinter and then THEHICKORIES pupatethe followingspring. Contarinia. To this large group in its broad Hickories belong to Carya, a genus of the wal- sense are added two new species known only nut family (Juglandaceae) that contains 12 North from their larvae. One species forms a complex American species including pecan. The present sac-like leaf gall (Figs. 172-173) in spring on sev- range of the genus encompasses the eastern half eral hickories, the other causes a bubble leafblis- of temperate North America, including southern ter (Figs. 170-171) onpecan. Ontario and Quebec, eastern United States, and Gliaspilota. Thisnew genus is described for Ce- isolated areas in the mountains of northeastern cidomyia glutinosa (Osten Sacken). Larvae of this Mexico, as well as parts of southeastern Asia species live freely on mucilaginous leafspots that (Grauke 2003, Manchester 1987, Stone 1990). The appear on the underside of new leaves in spring 12 North American hickories form two sections, (Figs. 175-177). Fully-fed larvae soon drop to the Apocarya with fourspecies and Encarya witheight soil to pupate and then emerge as adults the fol- species and one variety (Table 1, p. 4). Additional lowingyear. Carya spp. occur in southeast Asia, three to six Harmandiola. Until recently this group com- species depending on theauthority, andbelong to prised several gall-formers on poplars and chest- a third section, Sinocarya (Stone 1997). No galls nuts. Itnowincludesoneotherspecies,Harmandi- from Asian species have been reported and no ola nucicola (Osten Sacken), from hickory. Larvae sign ofgalls occurs on leaflets oftheir representa- formswellingsonnuthusks (Figs. 178-180). Fully tivesintheherbariumoftheNational Museumof grownlarvaedrop tothe soiltooverwinter. Natural History, Smithsonian Institution, Wash- Parallelodiplosis. A number of species in this ington, DC, so the present study treats only the genus live as parasitoids in galls formed by other gall midges ofNorthAmericanhickories. gall midges and cynipids. One, Parallelodiplosis Hickoriesarenow usedchieflyforfirewood or caryae (Felt), lives in galls of various species of charcoal-making, but their tough, resilient wood Caryomyia (Fig. 48). This species may have more was oncemore commonlyused forsuchthings as thanone generationperyear. spokes for wagon wheels and tool handles In addition to the taxa listed above, undeter- (Grauke 2003). Pecan, originally a Mississippi mined species of two additional genera were basin species, is now widely planted and culti- THEGALLMIDGES(DIPTERA:CECIDOMYIIDAE)OFHICKORIES JUGLANDACEAE:CARYA) ( Table 1. Scientific and commonnames ofNorthAmericanCarya species (from Stone 1997). The correct name for the section Eucarya is Carya, but Eucarya is used in this paper to avoid confusion with the genericname. SectionApocarya: C.aquatica (F. Michaux) Nuttall waterhickory C.cordiformis (Wangenh.)K. Koch bitternut C. illinoiensis (Wangenh.)K. Koch pecan C.palmeriManning Mexicanbitternut SectionCaryaorEucarya: C.floridanaSargent scrubhickory C.glabra (Miller)Sweet pignut C.laciniosa (F.Michaux)Loudon shellbark C. myristiciformis (F Michaux) Nuttall nutmeghickory C.ovata (Mill.)K. Koch shagbark C.ovatavar.mexicana (Helmsley)W.E.Manning Mexicanhickory C.pallida (Ashe)Engler&Graebner sandhickory C. texanaBuckley blackhickory C. tomentosa (Poiret) Nuttall mockernut vated throughout southeastern United States and Europe about 45 million years bp (Manchester abroadforitsnutmeat.Shagbarkandwhitehicko- 1987, Stone 1990). Post-Eocenehickoriesextended ries, especially north of the range of pecan, still across the Northern Hemisphere until after the have some value to people for their nutmeat. All mid-Miocene (ca. 15 million years bp), when the are important to wildlife, particularly tree squir- climate cooled and the individual elements of a rels (Manchester 1987). once uniformArcto-Tertiary forestbegan to sepa- The name "Carya" is Greek for walnut. The rate and undergo extinctions (Wolfe & Leopold common name "hickory" is derived from the 1967).FossilCaryafruithavebeenfoundinAlaska American aboriginal word "pawcahiccora" for a (Lower Miocene), Washington state (Miocene) mash made from the nutmeat. The word was and Colorado (Lower Oligocene), as well as in shortenedtohickoryforthenameofthetreebythe Eurasia (Middle Eocene to Pliocene) from central early Englishcolonists. "Pecan" isanotherAmeri- and eastern Europe, western Siberia, northern canaboriginalwordmeaning"hard-shellednut." China, and Japan (Wolfe & Leopold 1967, Man- Only one other genus ofJuglandaceae,Juglans chester 1987). The present narrow range of the or walnut, occurs in North America. It is widely genus is due to extinctions during the uppermost distributed throughout the temperate Holarctic Tertiary and Pleistocene (Manchester 1987). Region. The only gall midges recorded from During the Pleistocene glacial maxima, hicko- Juglansacrossitsentirerangearetwoundescribed ries and other deciduous trees were much more easternNorthAmericanspecies thatliveas larvae severely restricted in range than presently. Decid- in marginal rolls of leaflets, one a Dasineura on uous angiosperm trees survived insmall, isolated Juglans nigra, or black walnut, the other a Con- populations in the lower Mississippi valley and tatinia on Juglans cinerea, or butternut (Gagne on the southern coastal plain among the more 1989). widespread conifers, as determined by pollen de- Hickories are an ancient genus that was once posits in lakebeds (Davis 1983, Delcourtand Del- much more widely dispersed. Carya-like pollen is court 1984). When the glaciers of the most recent known from the early Paleocene, about60 million iceagebegantorecedeabout12,000yearsago,the years before the present (bp), but the earliest rec- various tree species that survived gradually ognizable Carya fossil fruit is from mid-Eocene spread north on a species basis and not as com- RAYMOND GAGNE J. munities, their individual paths dependent upon & Pratt (1986), and Stone (1997). Keys do not all their particular refugia and individual strategies agree in the number of species. After gaining (Davis 1983). Moving from southwest to north- someexperienceIfollowedElias (1980),Grauke& east, Carya expanded more slowly than some Pratt(1986),andStone (1997) inwhichCaryaovalis other genera, in partbecause of their heavy fruit; is placed under C. glabra and C. septentrionalis un- even so, hickories arrived early (10,000 bp) in the der C. ovata. Hickories are highly prone to hy- upper midwest at sites close to their present bridization (Grauke 1988, 2003) and one notices northernboundary,butwereslowtocrosstheAp- considerable variation in this genus, but only palachian Mountains, arriving in New England rarely did I encounter a tree not readily attributa- onlyabout5,000 yearsbp (Davis 1983). ble to a species. As ithappens, mostcecidomyiids Populations of hickories in Mexico (Map 1) inthisstudydonotappeartodiscernamonghick- may not have been greatly affected by glaciation. ories, particularly those feeding on the Eucarya Hickories there are part of an assemblage of de- section. — ciduous trees generically similarto thatofeastern Collections. Most galls and their cecidomyi- North America. This representation of northern ids used for this studywere collectedby me since trees in Mexico appears to have been in place 1993. Observations and collections were largely since the mid-Miocene when the cooling conti- made at many sites within 20 miles of my resi- nental climate allowed the deciduous forest to dence in Silver Spring, Maryland and at one site reach so far south (Graham 1999). Evidence for near Hedgesville, West Virginia. A major local this is that pollen and spore deposits in Mexico area sampled was the Beltsville Agricultural Re- show a late Cenozoic history of the assemblage's search Center in Beltsville, Maryland where trees presence there. In addition, it is unlikely that the remained undisturbed during this time period soil between central Texas and the relict popula- and where I was free to collect as an employee of tions in Mexico would have supported a rich de- the U.S. Department ofAgriculture. I made three ciduous forest more recently than the mid- separate circuit trips across midwestern and Miocene (Graham 1999). These points will be of southern United States to collect galls, two sepa- interest farther along when considering the rela- rate trips to Florida, several to New England and tive absence of galls on a Mexican population of New York, including once an extension into On- Carya ovatavar. mexicana. tario, and one trip to Tamaulipas, Mexico with a North American hickories vary greatly in dis- stop onthewaytosoutheasternTexas.Additional tribution.Afewspecies, C. cordiformis,C.glabra, C. collections were made for me by collaborators in ovata, and C. tomentosa, range widely and occupy Connecticut, Georgia, and Florida. In addition, I most of the combined range of the genus in studied types and other specimens from the Felt Canada and theUnited States indicated inMap 1. Collection, whichis theproperty ofthe NewYork The remaining species, except pecan, Carya illi- State Museum in Albany but currently on loan to noiensis, which has been extensively planted for the Systematic Entomology Laboratory in Wash- commerce,arerestricted tosmallpartsofthecom- ington,DC,andthefewmiscellaneousspecimens, bined range (Elias 1980). including the existing Osten Sacken types, de- posited overthe years in the NationalMuseum of NaturalHistory, Washington, DC. MATERIALSANDMETHODS Oneneedstocoveralotofgroundtofindhick- ories with galls, especially the less common ones. — Hickory names and identifications. In dis- The best collection sites were forest edges or cussions thatrefer toboth Carya and Caryomyia or open-grown trees where low-hanging branches Caryadiplosis, the abbreviation "C." is used to madecollectingconvenient. Samplingwhile trav- mean Caryomyia or Caryadiplosis; in those discus- eling is always haphazard, dependent on a place sions I use the common names of the hosts to to park a car or obtaining permission from prop- avoid confusion. Both scientific and common ertyowners. Oldchurchyards, cemeteries, and in- names of hickories are listed in Table 1. Hickory terstate highway rest stops, particularly the large identifications were made with the use ofvarious midwestern ones, are productive sites. Paths keys and treatments, including those of Brown & through woods are not as good when time is Brown(1972),Elias(1980),Gleason(1963),Grauke pressing because finding hickories can be uncer- — THEGALLMIDGES(DIPTERA:CECIDOMYIIDAE)OFHICKORIES JUGLANDACEAE:CARYA) ( tain and branches in forests are often too high to peat moss. If larvae normally exit the galls when reach. Some good collecting sites incentral Mary- full grown, they will crawl into the peat moss. In land were lost to development over the course of this way it is possible to obtain all stages of a thiswork: onesiteisnowapublic iceskatingrink speciesand learnsomethingoftheirbiology. Even with large adjacent parking lot, and another a so, I was unable to rear adults or obtain pupae of shoppingcenter. all of the species treated here. Many larvae die Collectingconsistedofharvesting gall-bearing from parasitism or other causes before they are leaflets and placing them in a plastic bag with a fully developed, so one needs a large number of localityandhostlabel. Igatheredgallsatasiteun- galls to successfully obtain gall midge pupae and tilIhadasufficientnumberoftheavailablekinds. adults. Galls of the less common species were Because these galls are small, close examination never found in sufficiently large numbers to suc- was done subsequently, usually in the evening of cessfullyrearadults. — the same day with the aid of a dissecting micro- Adult rearing. I have tried various methods scope. Representative gallswere cutopenand the to rear adults but that outlined in Gagne (1989, larvae removed and saved in alcohol. Samples of 1994) still works satisfactorily, with one change. I whole and cross-sectioned galls were glued to a place galls, or larvae when these escape the galls card and pinned together with a data label for topupateinthesoil, inpotsfilledwithdampened permanent reference. Remaining galls that were peat moss and place them in an outdoor cold excess but not numerous enough to attempt to frame for the winter. Many galls in that situation rear were kept dry in labelled plastic bags for fu- undergo deterioration from the dampness and ture comparison and extraction of larvae as fungi, allowing entry by various arthropods, needed. Larval samples were placed in alcohol mainlymitesandsciarids. Thisisprobablynodif- andlabelled,someeventuallytobeslidemounted ferent than what happens in nature, but it means orusedforSEMphotos. Attheproper season, de- that one has to have a largenumber ofgalls to in- pending on the species, galls with full-grown lar- sure success. At the end ofFebruarywhenIbring vae or, for genera other than Caryomyia, larvae the pots indoors, I place each pot in a separate that had escaped the galls, were kept through clear plastic bag. The bag is large enough to winter to rear adults. Some reared adults were spread well away from the pot at the bottom so pinned but most were kept in 70% ethyl alcohol any emerging specimens are noticeable when the foreventual slidemounting. bag is held against the light. The top ofthe bag is Sweeping or otherwise trapping adults in closedbybringing togethertwoendsofthe open- flight is not practical. Associating both sexes of a ing in a simple knot. Condensation does not col- species is usually not possible for most plant- lectonthe inside ofthebagifthepotisnotoverly feeding gall midges unless they are reared to- wetin the firstplace and thebag is kept outofdi- gether from similar galls. Any sweep of hickory rectsunlight. Ifmoisturedoescondense on thein- branches can yield many species of gall midges, side of the bag, a wetbag canbe changed repeat- some just passing by and having nothing to do edly to a dry one until the pot loses the excess with hickory. While gall midges caught in flight moisture. This seems to be a more efficient may with practice be identifiable to generic level method than the one I formerly used, when I when viewed dried or in alcohol, they must be placed the pots individually in cardboard shoe- painstakingly mounted on microscope slides for boxes, each with a vial fit tightly into an opening further critical viewing, and even then most at the end. With that method I occasionally lost flight-caught phytophagous cecidomyiids cannot somespecimens thatcould squeeze outthrougha usually be determined beyond genus with confi- cover that did not fit tightly or escape when I dence. Slide mounting is a time-consuming activ- opened thecoverfora peek. ity better spent on reared samples. The best Specimens examined and their preparation. methodforobtainingadultsofphytophagousgall Slide mounting of specimens follows the tech- midges is to collect a sufficiency of galls before niqueinGagne (1989, 1994)usingCanadabalsam. they begin dropping from the trees, usually in Most of the larvae and adults studied during this September and October for hickory galls, and study can be traced to particular galls that are ei- keep them over the winter. Galls with full-grown ther on pinned cards or in alcohol. Holotypes of larvae are placed in pots filled with dampened newspeciesandremainingspecimensusedinthis — RAYMOND GAGNE J. study, except those from the Felt Collection, are include: For. = Forest; Mt. = Mountain;Natl = Na- deposited in the insect collection of the National tional; MCZ = Museum of Comparative Zoology, Museum of Natural History in Washington, DC Cambridge, Massachusetts; Ref. = Refuge; St. - (USNM). State; USDA = U.S. Department of Agricuture; — Descriptions. In the species descriptions that USNM = National Museum of Natural History, follow, the gall is described first because it is the SmithsonianInstitution,Washington, D—C. mostobviousmanifestationofa species. Eachgall Species concept within Caryomyia. I consid- descriptionlistselementsofthematuregallinthis ered a distinct gall the phenotypic extension of a order: frequency (common, occasional, or rare), discrete species. In only one case might the gall distributiononhickories (Apocarya or Eucarya sec- distinction seem minor, viz. between C. leviglobus tion or both); arrangement of galls on a leaflet and C. spiniglobus. The first (Fig. 62) has a smooth (single or clustered), position on a leaflet (upper exterior, the other (Fig. 64) iscoveredwithstiffse- orlowerleafsurface,betweenoronveins);height tae. Becausethetwokindswereneverfoundinter- mm or length in and external shape; external tex- spersed, I consider the gall makers distinct, even ture; connection to the leaflet; wall structure in though I cannot separate them morphologically. longitudinal section; and shape and texture ofthe Two other species arenoteworthyherebecause of larval chamber. A concluding sentence or two their variability. Caryomyia thompsonihas two dis- usuallyfollows to differentiatethe gallfromother tinct variations in gall surface, one kind more or gallswithwhichitmightbeconfused. less sparsely haired and the gall surface visible Somespeciesarenamedand describedhereon (Fig. 113), the other with thick hair obscuring the thebasisofonlythegallandlarvabecauseduring gallsurface(Fig. 114).Thesecanoftenbefoundin- the period of this study I did not succeed in rear- terspersedonaleaflet(Fig. 112), and themorphol- ing the adult. In a comprehensive study the ab- ogy ofCaryomyia specimens fromeachis indistin- sence of adults of some species should not pre- guishable. Except for the type specimens under clude describing species from larvae alone. the species name, I made no further distinctionin Larvaehave good distinguishing characters, their the list between the two kinds and considered biology andhosts are known, and theircharacter- them to be made by the same species. The other istic galls canbe found again, possibly to success- special case is C. caryae, in which the galls inter- fullyrear adults. grade. The extreme forms (cf. Figs. 55-56 with Adult abdominal tergal setation is described Figs. 57-58) each seen alone certainly appear dif- from the third tergum posteriad because the pos- ferent and I might have treated them as separate teriorsetalrowsofthefirstandsecondtergitesare species were it not for commonly found inter- always discontinuous mesally. Larval length of- grades. No differences in any stage of the gall fers a rough size estimate. In any instar a larva midgefromthetwoextremesofgallscouldbede- grows considerably, so the larger number of the tected. range isthe moremeaningful one. Lists of collection and specimen localities. GENUS CARYOMYIAFELT Collection localities of galls of each cecidomyiid speciesinthis revisionarelisted alphabeticallyby Caryomyia Felt 1909: 292. Type species, Cecidomyia tubi- state or province and county or town. Voucher colaOstenSacken,byoriginaldesignation.Felt1911: specimensforthegallrecordsaredepositedinthe 56 (review); Felt 1921: 94 (review); Gagne 2004: 100 USNM, either pinned or in alcohol. Listed also, (catalog). with citations, are galls from publications when — accompanied by identifiable figures. A separate Diagnosis. This genus may be differentiated list of larvae, pupae, and adults of each species from other genera ofCecidomyiidiby the follow- studied gives the collection and rearing (if any) ingcombinationofcharacters:Adultwithoutdor- data in alphabetical order by state or province, salprotuberanceonocciput;wing(Fig.245)broad- and town or other site, followed by collection est onbasal half, C unbroken atjuncture with R 5; date, collector if not me, and the associated life legsconsiderablyforeshortened (Fig. 183), shorter stage.Allthesespecimens,includingtypesofnew thanwinglength; abdominal sternites lacking an- species, but excepting the Felt Collection speci- terior pair of trichoid sensilla; gonostylus short, mens, are deposited in the USNM. Abbeviations broad, with wide tooth; protrusible ovipositor ta- — THEGALLMIDGES(DIPTERA:CECIDOMYIIDAE)OFHICKORIES JUGLANDACEAE:CARYA) ( pered abruptlybeyond eighth segment, short, the gate, stiff, pigmented dermal structure along its ninth tergite dorsally with narrow, pigmented, length. Sternites rectangular, 2-3 times as wide as stiff, dermal structure, and cerci partially fused long, without trichoid sensilla; second through dorsally at base (Figs. 184-185). Larva with only sixthsterniteswithhorizontalroworrows ofpos- 1-4 lateral papillae; dorsal thoracic papillae re- teriorsetaeandvariablehorizontalrowsofmostly ducedto4;terminalpapillaeallsimilar,flatorcon- shorter setae near midlength; seventh and eighth vex,asetose(Fig.271). sterniteswithmorenumeroussetaeandbothrows Description. Adult: Head: Eyes connate; adjacent.Genitalia(Figs. 186,247-257):cercishort, facetsmostlyhexagonoid and closelyadjacentex- broad, with setae along margin; hypoproct paral- cept slightly farther apart and circular at mid- lel-sided to broadly widened posteriorly, with height of head. Occiput without dorsal protuber- scattered short setae along posterior and lateral ance. Frons setose. Labrumwithorwithout setae. edges, covered with setulae except, in most Labella separate or coalesced, with or without se- species, dorsum bare frombase to posterior third tae.Palpus1-to4-segmented.Antennawith 12-20 or fourth; aedeagus taperingfrombase to apex or flagellomeres, when 12, often with an additional broad, cylindrical, rounded or notched apically; bud atapexoftwelfthcountedasa thirteenth,but gonocoxite cylindrical, short, broad; gonostylus this one without circumfila; male flagellomeres short, broad, ovate, with dense setae and entirely (Figs. 235, 237, 239) generally foreshortened, bin- setulose. Female abdomen (Figs. 183-185): First odal with foreshortened internodes and necks, or through eighth tergites entire, rectangular, about cylindrical (Fig. 240) and lacking internodes and 6 times as wide as long; vestiture as in male. Sec- necks;eitherwith3circumfila, 1 onbasalnodeand ond through seventh sternites rectangular, about 2 ondistalnode, or the3 evenly distributed along 2.5 times as wide as long, vestiture as in male. flagellomere length, loops of the circumfila sub- Ovipositorslightlyprotrusible, softtorigid, grad- equal in length, short, not attaining bases ofnext ually tapering beyond eighth segment; ninth ter- distal circumfilum, rarely a loop connected be- gitedorsallywithnarrow,pigmented,stiffdermal tweentwocircumfila,orwith9-10interconnected sclerite, laterally and ventrally setose; cerci ovoid horizontal circumfila; female flagellomeres (Figs. or triangular, partially fused anterodorsally, cov- 236, 238, 241) eachprogressively shorter, cylindri- ered with setae but without distal peglike setae; cal with shortnecks, ringed with 2 appressed cir- hypoproctshort,broad,with2posteriorsetae. cumfila connected by two longitudinal bands or Pupa (Figs. 187-214, 267-268): Integument un- several anastomozing bands. Thorax: Scutum pigmented. Head: Vertex convex, on each side with 4 longitudinal rows ofsetae. Scutellum with withanterolateralpairofpapillae,oneofeachpair, lateral group of setae on each side. Pleura with occasionally both, with conspicuous seta. Anten- vestiture only on setose mesepimeron. Wing (Fig. nalbasesforminganteriorlydirectedhornsofvar- 245) broadest on basal half, unmarked, width iouslengthandshape.Anteriormostanglesofface 0.4-0.5 wing length; C unbroken atjuncture with betweenantennalbases and flagellaoftenconvex. R5; R5curved apically,joining C posterior towing Face anterior to midlength smooth or indented apex;Rsincomplete,closertoarculusthantoapex withanextruded posterior lip oneachside. Frons ofRj. Legsgenerallyforeshortened, femora,tibiae and clypeus convex or conspicuously expanded, and tarsi of approximately equal length or pro- base of frons with pair of setose papillae. Labella gressively shorter from femora to tarsi. Tarsal triangular, convex. Thorax: Prothoracic spiracles claws (Fig. 246) untoothed, curved beyond mid- elongate, pointed. Thoracic papillae short-setose, length; empodia as long as claws; pulvilli minus- situatedoneachsideasfollows: 1onprothoraxan- cule, much shorter than claws. Male abdomen: terior and medial to spiracles, 3 along anterior First through seventh tergites entire, rectangular, edge ofnotum, 1 anteriortowingbase, andoneat about 5 times as wide as long; with setae in 1-4 posterior end of notum. Notum with row of tiny rows along posterior marginor onlyposterolater- spiculesoneach side ofmedialsuture. Wingpads ally and with anterior pair of trichoid sensilla but extending just beyond second abdominal seg- no othervestitureexceptformicrosetulae entirely ment,legsextendingapproximatelytoendofthird covering sclerites; eighth tergite may be more abdominal segment. Abdomen: Spiracles sessile. weakly sclerotized than preceding tergites and Firstabdominaltergumofallspeciescoveredwith have fewer setae; tenth tergumwithnarrow, elon- short spicules and additionally with horizontal RAYMOND GAGNE J. row of 6 short-setose papillae barely longer than circumfila of the male flagellomeres (Gagne 1989, spicules;intersegmentaltergalareassmooth,bare; 1994). Below thatlevelthegenusfitsintothelarge second through eighth tergites as for first or with tribe Cecidomyiini that contains most of the Hol- severalhorizontalrowsoflarge,closelyset,spine- arctic plant-feeding Cecidomyiidi, mostly genera like spicules on anterior third of segment. Pleura that lack a dorsal occipital protuberance, have smooth anteriorly and posteriorly but elsewhere usually simple tarsal claws that are curved be- coveredwithsmallspicules,apairofsetosepapil- yond midlength, and larvae with a particular lae present similar to those on terga. Sterna wrin- makeup of the terminal papillae (Gagne 1989). kled, rough,withminutepointedverrucae anteri- While Cecidomyiini typically have four pairs of orlyandlaterally,papillaenotdetected. terminal papillae, three of them short-setose and Larva: Third instar (Figs. 215-231, 269-271): the other corniform, enlarged, and recurved dor- White, cylindrical to ovoid, anteriorandposterior sally (as in Fig. 334), all eight terminal papillae of ends rounded or tapered. Head directed anteri- Caryomyia are without setae and similar in ap- orly or ventrally Antenna about twice as long as pearance (Fig. 271). The relationship of the genus wide. Spatula (Figs. 274-327) present, variously to other Cecidomyiini is nonetheless clear due to shaped. Papillaewithorwithoutsetae, theirbases the fortunate discovery during this study of the prominent or not. Lateral papillae 1—4. Ventral new genus Caryadiplosis that is described below. papilla on each side of spatula closely adjacent to While the two genera are evidently related due to lateralpapillae. Dorsalpapillaeofthorax4,offirst shared, derived adultcharacters, Caryadiplosis has through seventh abdominal segments 4-6. First retained the larval form commonly found in the through eighth abdominal segments always with tribe. In the key to genera of Cecidomyiidae in verrucae laterally, dorsum and venter smooth or Gagne (1981), Caryomyia runs to couplet 135 with lined and/or variously covered with rounded to Contarinia, but that placement does not argue for pointedverrucaeandventrallywithroundedver- anyaffinitybetweenthosetwo genera. rucae or pointed spicules. Terminal segment of Felt (1909) first described Caryomyia to include same length to more than twice as long as eighth eightspecies andeventually added anotherseven abdominal segment, with 8 similar papillae with- (Felt 1921). Three were included only for conve- outsetae,4oneachside,slightlytoconspicuously nience because they occurred on hickories. Felt convex.Analregionwith0-2papillae. Second in- (1909, 1921) wrote that Caryomyia bore some re- star (272-273): White, ovoid. Head directed ante- semblance to Planetella (as Hormomyia), but the riorlyorventrally.Antennaabouttwiceaslongas only grounds he offered was that the two genera wide. Spatula present in all species except C. albi- shared "somewhat the appearance," a suggestion pilosa and C. striolata, with apically pointed tooth that does notbear scrutiny. Felt (1921) laterwrote and elliptical to quadrate shaft. When spatula thatCaryomyia mightbearelativeofPolystepha (as present, collar segmentoneachsidewithpatchof Cincticornia), a genus in the tribe Asphondyliini, enlarged spicules. Papillae with or without setae, because themale antennal circumfila ofboth gen- their bases prominently raised or not. Lateral era are short and Carya and Quercus, the host of papillae 1-4. Dorsal papillae of thorax 4, of first Polystepha, are somewhat related. Because of this, through seventh abdominal segments 4-6. First Caryomyia was listed in Felt (1958) with the tribe through eighth abdominal segments variously Asphondyliini, and that placement was contin- covered dorsally with rounded and or pointed ued in Foote (1965). Male antennal flagellomeres verrucae. First instar (Fig. 35): Translucent to and their circumfila are foreshortened in some white, fusiform at first, becoming flattened and species ofCaryomyia that Felt knew (Fig. 237), but discoid. Spiracles present on first thoracic and had he seen a male of the Caryomyia cilidolium eighth abdominal segments only. Antenna about group with its reticulated flagellomeres (Fig. 240) threetimesas longaswide. hewouldhavehadbetterbutstill superficial sup- Egg(Fig. 34): Elongate-ovoid, red. port. The antennae of the C. cilidolium species group are generally similar to those found in sev- eral species of Polystepha spp. (cf. Fig. 30a, p. 154 Affinities of Felt 1916). We now know that the two genera Caryomyia belongs to the supertribe Ce- are in separate tribes because Polystepha shares a cidomyiidi because of the separate many-looped number of apomorphies with the Asphondyliini 10 THEGALLMIDGES(DIPTERA:CECIDOMYIIDAE)OFHICKORIES JUGLANDACEAE:GARYA) ( (Mohn 1961, Gagne 1994), the most importantbe- gonostylus setulose only on the basal half and ing the enlarged female seventh sternite. That the longitudinally ridged beyond, is the one most antennaeoftwo groups ofspecies in two separate commonly found in the subfamily Cecidomyi- tribes are so similar is only one more example of inae,includingMacrodiplosisand Thecodiplosis,but adaptive homoplasy that can be found in Ce- notParadiplosis. Acompletely setulose gonostylus cidomyiidae. Other examples of parallel modifi- is definitely the less common state in Cecidomyi- cations are listed in the following discussion. ini. It will be noticed that Paradiplosis shares at These serve as a good argument for comprehen- least three characters with Caryomyia and Ca- sive revisions, in which context these happen- ryadiplosis: a sclerite on the distal half of the stances canbe appreciatedbetter. ovipositor, theunbrokencostalvein, andthecom- A relationship between Caryomyia and Cary- pletely setulose gonostylus. This genus is dis- adiplosis is supported by three synapomorphies cussed furtherunder Caryadiplosis. and one other character they have in common. Caryomyia differs from Caryadiplosis in many The first synapomorphy is the presence of a nar- ways, presumablyall apomorphic, as follows: row, elongate, dorsal sclerite on the female ninth segment, the posterior, more rigid half of the ovi- 1.Galls are complex, the result of a complete re- positor(Figs. 258-264,331). Bothgenerahavepro- structuringofthehosttissue,insteadofsimple trusible, yet relatively short ovipositors on which swellings. the structure is presumably used as a stiffener. I 2.Larvae have only one to four lateral thoracic know of a corresponding structure on only two papillae on each side of the midline (Fig. 274) other genera of Cecidomyiini, Macrodiplosis and instead of six lateral thoracic papillae in two Paradiplosis. Macrodiplosis, a large genus of leaf tripletsoneachside (Figs. 335). Suchreduction gallmakers onoaks (Gagne 1989), has a relatively is not uncommon in Cecidomyiinae that form short ovipositor also, but the stiffener extends galls, especially complex galls, but the full only along the posterior two-thirds of the seg- complement is the rule in non-phytophagous ment. InParadiplosis, a genusoffourspecies onfir Cecidomyiinae. and spruce, the ovipositor is extremely long (Os- 3.Larvae have only four instead of six thoracic good and Gagne 1978). dorsal papillae. Except for Caryomyia, all other The second synapomorphy is the unbroken Cecidomyiidi known to me have six dorsal costal vein at its junction with R5. An unbroken thoracicpapillae. Iknow ofasimilarreduction costal wing vein is rare in Cecidomyiini and un- intwogenera,Asteromyia (Gagne 1968) and Se- known in other tribes of the supertribe Ce- mudobia (Roskam 1977), both members of the cidomyiidi. The trait occurs in two cecidomyiine supertribe Lasiopteridibutinseparate tribes. genera that infest Pinaceae, Paradiplosis and The- 4.The larval terminalpapillae are all similar and codiplosis, neither of which seems especially asetose (Fig. 271), instead of dissimilar and ei- closely related to one anotheror to Caryomyia and ther setose or corniform (Fig. 334). This reduc- Caryadiplosis, so I suppose the closed costa tion is uniqueinCecidomyiidi. evolved independently in this tribe at least three 5.Pupation occurs in the gall rather than in the times. Macrodiplosis, mentioned in the previous soil. paragraph,has abrokencosta. 6.Adult tarsi of Caryomyia are foreshortened, no The third synapomorphy is the apparent loss longer than the tibiae and 0.33 or shorter than ofapairofpeglikesensilla usuallyfound nearthe the entire leg, while in Caryadiplosis and most apex of the female cerci. Almost all Cecidomyiidi other Cecidomyiini the tarsi are appreciably have these two (sometimes more) distinctive sen- longer than the tibiae and 0.40 or more than silla that have broad sockets and wide, blunt- the entireleg. tipped setae (as in Fig. 362). Paradiplosis and The- 7.Adultabdominaltergiteshaveneitheralateral codiplosisbothhave the pairofsensilla. group ofsetae (setaethatliejustanteriortobut Another character Caryomyia and Caryadiplosis separate from the posterior row of setae) nor haveincommonisacompletelysetulosegonosty- scales (Fig. 258). In Caryadiplosis and most lus,butIcannotsaywithconfidencewhetherthis other Cecidomyiini, these setae and scales are is primitive or derived. The alternative state, the usuallypresent (seenin Fig. 330).

See more

The list of books you might like

Upgrade Premium
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.
DMCA & Copyright: Dear all, most of the website is community built, users are uploading hundred of books everyday, which makes really hard for us to identify copyrighted material, please contact us if you want any material removed.
Copyright @ PDFdrive.to, 2022 | We love our users