1 gal sigh Verge
+ Der yee cps ern) NY Ari aye at Alt Ghee?
eh AAs OF BERS: yh sy oa ‘x
vey
eet rites at Vee
My yytney ra
vats eerita
Cen as eM tity alge
ORM NSS Ft yyy ‘, CAUISIAN | SSL TY b y ‘ : , Des Op NB SDALt ie EAS rey adatnahe
Sear ayia Ng tp cade Wty 9 YEO Ae LS Oh HENNE OS ae
ynstens
RRO te eg Male eh, a iY
Pa RA vy i AN . é ‘ pease ‘ . i Xe - if = } Panna rite y , 1B vy x Wes Syivein'ss
: ds ARrcys DERN bs alate wala eee NS
ia raly ary ay ular fee ws ayy Ry
nt te AERP ANY
4
ae ee ety . Seen Pay ree i ‘ ars Saat ; RRA Soviets ys "f ra Aeneas me
Sah ROW AL Pcie
ARR RAD HCL
f pat Pe ara, Guat
eerea Iai x
Hy
eaten) Waa EO
iteike By
eT Gortok So St oe 4
eae cesses A
fet rgueete
avs
tag
yak
Posse Sa! Ky te
af hs
INST
a OT (a
& Wy ASH
(7 is YY INOSHS NOSIS
4911 LIBRARIES SMITHSONIAN INSTITUTION NOILALILSNI NVINOSHLINS S3IYVUSIT
S3le 4, INST “Ss =
é
Ne C Sal INST Sil INST
’
SMITHSONIAN
NVINOSHLIWS
SMITHSONIAN tp
% 4
NVINOSHLIWS
SMITHSONIAN
NVINOSHLIWS
TION NOILNLILSNI_ NVINOSHLINS S31YVYE!I1_ LIBRARIES SMITHSONIAN INSTITUTION
2 < z [e) OD i: & = if) > = s 7) = a a 4H = A a = fad me ng, dw e < a 4 a4 - si | <x o =" oc S fog a a = S) e S =" fs) = 3 z 4 2 = = = YGi7 LIBRARIES SMITHSONIAN INSTITUTION NOILALILSNI NVINOSHLINS S31Yu¥Vvud ma r ra - = - ei = = ce) = (o) = fe) = wo = oO — wo — Q - 5 2 5 2 5 > > FE > = = > p= = = 5. = 5 = . a z B Z 3 Ae D JTION NOILALILSNI NVINOSHLIWS SAlYVUEII_LIBRARIES | SMITHSONIAN INSTITUTION £ z = (2 ow = < @. = ~ a = z I Zz WN = z = a | z z SQ F § 5 VE 3 g° EW 2 : E f Ean: eaten a Z 3 ugi7_LIBRARIES SMITHSONIAN _INSTITUTION NOILOLILSNI_ NVINOSHLIWS SAtYVee te a oe ul A o a =. ~ tit 7 “4 = a = a a whe = < = : g S a Gry 5 S 3 = 5 a “Mf 35 = 5 e 3 z a 2 F z 3 2 ITION NOILNLILSNI NVINOSHLINS S3IYVYSIT LIBRARIES SMITHSONIAN_ INSTITUTION : Sle: : eae : = o = o = a = 5 2 5 2 5 2 5 = > \ CARE a = 3 3 a = SW a = ae = oD z a a 2 a z UaIT_ LIBRARIES SMITHSONIAN INSTITUTION NOILALILSNI_ NVINOSHLINS, S31 YW 817 _ oe = ; = = = E = GY. 5 z Ng 3 5 is. = [e} * SY [e) nae Zz MS = = == ‘ = » > > 7) 2 ea z
S3!IYVYSIT LIBRARIES
TION NOILNLILSNI_ NVINOSHLIWS SMITHSONIAN INSTITUTION
o = = = ui o a a a = = =f <x pa =A ae ow =r = (= Pee] = = = 3 S to = fo) z ¥ai7 LIBRAR TES SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLINS S3IYVuaIT = o S 3 ) Ee = = > ee e = -_ es = re a Z 2 e
JTION NOILNLILSNI_NVINOSHLINS S31YVYSIT LIBRARIES SMITHSONIAN
IOSHLINS S3IYVYaIT LIBRARIES SMITHSONIAN OSHLINS S3IYVYEIT_ LIBRARIES SMITHSONIAN OSHLINS S3ZIYVYUSIT LIBRARIES SMITHSONIAN
Z Z z x z “\ <7 = a SN {HSV AG <x ee 4 ; a : é ‘ » Si Sere AL EAL NE cy . OGYG TX oy! TF RX | Se [Se
a LE NS mf OD \e fie. MES 5 Pee ge oN ONG wy .% AF bie ee
a - ery 2 of fee TK Rdappys/ 5 “Mabey —
m = 5 Wasnt _ aL @ 2 e a m Yay zh Xe Wasi m .
” — wo = wn _
5 INSTITUTION NOILNLILSNI NVINOSHLINS S3INVYSIT LIBRARIES SMIT
g i ae ee ae
= 5 Wy 2 z 2 Uy > Nxt
re) as MENS re a 2 lp “aga YIN SG
Z E AS" Zz, = 2/7 = NY z
= ore <= = = ae
tae pos dWYaIT_LIBRARIES SMITHSONIAN INSTITUTION |, NOILMIILSNI_NVIN
= Wl z 2 2 Ww Ae.
is = it = g = a.
= a = a! fae _
= . = = = ao | <x = \
= FS = ce S es =e
3 = 3 =k 5 = 9
a : = z q = zs 2
S SMITHSONIAN INSTITUTION NOILMLILSNI NVINOSHLINS S3IYVUdIT LIBRARIES SMIT
a Lag = c z lee za
2 oD ° pa Ne o = °
- = 1 ~ JE EA
=a By 5 -) aS = -) 5.
ae: = = eas = = Fy
= ies = 2 = ra | 7,
” m 7) = n -_ ” é
= ; 7) = o eo 2 -
I NVINOSHLIWS Sa luvua Wu! BRARI ES SMITHSONIAN — NOLLOLILSNI_ NIN
=: 7)
§ 5 DY 2 @X F : 5
se 0 Oi Be NN fe) S =
E 2% = » 2 2 =
z 2 De hi a eae ee saiuvugin LIBRARIES SMIT
S SMITHSONIAN _INSTITUTION NOILOALILSNI NVINOSHLIWS
LIBRARIES NOILNLILSNI
LIBRARIES
LIBRARIES
x x IW NVINOSHLINS S3IYVYUGIT LIBRARIES SMITHSONIAN_INSTITUTION NOILNLILSNI NVI
SaINVYGIT LIBRARIES SMITHSONIAN
INSTITUTION NOILNLILSNI INSTITUTION NOILALILSNI
ire & z Le = = (e) o o = S ps) a =) > > = \ — _— 2) m m SMI
S SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLINS S3ZIYVYGIT LIBRARIES
N
SS es OS S3IYVYGIT LIBRARIES SMITHSONIAN INSTITUTION NOILNLILSNI_ NVI
NVINOSHLINS S31IYVUE!l
NVINOSHLIWS NVINOSHLIWS SMITHSONIAN
“a NVINOSHLIWS SMITHSONIAN
NVINOSHLIWS
>
S SMITHSONIAN
NOILNILILSN! NOILNLILSNI NOILNLILSNI LIBRARIES NOILNLILSNI
INSTITUTION NOILALILSNI NVINOSHLINS S3!lu¥vudiT LIBRARIES SMI
INSTITUTION
S3IYVYSIT LIBRARIES SMITHSONIA
INSTITUTION INSTITUTION saluvugi INSTITUTION Ne
NVINOSHLINS S3IYVYUEIT_LIBRARIES
of.
SMITHSONIAN INSTITUTION NOILNLILSNI_ NVI
eo z RS L195 es Nj SF ; | 9 oe =
N
; WSON, > Q
vy SYA S os SS
S; Cy és eA 3 $ o> ws
OSHLINS SAIYVYE!IT_ LIBRARIES
ISONIAN ISHLIWS
{SONIAN SYM Nast ISHLINS . Ss \\ 4HSONIAN e: Va Le cs
iS
a
— VOL. 87 JANUARY 1985 NO. | ! (ISSN 0013-8797)
PROCEEDINGS
i of the
ENTOMOLOGICAL SOCIETY ot WASHINGTON
PUBLISHED QUARTERLY
CONTENTS
BLANCHARD, A. and E. C. KNUDSON—New species of Phycitinae (Lepidoptera, Pyralidae)
i from Dexas:\ with descriptionof anew genus) (fio ee Gl altos ee 231 - DARLING, D. C. and D. R. SMITH—Description and life history of a new species of Nematus i (Hymenoptera: Tenthredinidae) on Robinia hispida (Fabaceae) in New York ........... 225
. DAVIS, D. R.—A re-examination of Enteucha cyanochlora Meyrick and its subsequent transfer
[ to the Nepticulidae (Lepidoptera: Nepticuloidea) ...................0 022 c eee eee eens 142 i DEITZ, L. L.—Placement of the genera Abelus Stal and Hemicentrus Melichar in the subfamily | G@entrotinae|(Homoptera: (Membracidae) ).) ish. U ayy ueibyi ee bidbaeieaiem ats & tee opie eb tema 161
i EVANS, E. W.—A key to nymphs of four species of the genus Podisus (Hemiptera: Pentatomidae)
} GUNOLHCASTEFTMUNOLEDVAMETACA') Abs. sty ie bie he Wh Raa a cle REGUS iRaeee ete Wed nLt ELA PEE Ones to Rout ie Wkly 94 Ht EVANS, W. G.—The intertidal zonation of Thalassotrechus barbarae (Horn) (Coleoptera: Ca- i GARNEAU AA a eat iON TCO R ORR ILOMEE EAE RCE LM AL EA g) Ste ta it has Oh 110 1h GAGNE, R. J.—Descriptions of new Nearctic Cecidomyiidae (Diptera) that live in xylem vessels i of fresh-cut wood, and a review of Ledomyia (S. str.) . 2... 0006. es 116 \ HARBACH, R. E., C. DAHL, and G. B. WHITE—Culex (Culex) pipiens Linnaeus (Diptera: } Culicidae): Concepts, type designations, and description ................ 0002.00. 000055 1
HOEBEKE, E. R. and W. T. JOHNSON—A European privet sawfly, Macrophya punctumalbum (L.): North American distribution, host plants, seasonal history and descriptions of the
i immature stages (Hymenoptera: Tenthredinidae) ....................00 000022 e eee 25 M! HOEBEKE, E. R. and A. G. WHEELER, JR.—Sitona lineatus (L.), the pea leaf weevil: first y records in eastern North America (Coleoptera: Curculionidae) ........................ 216 } KIMSEY, L. S. and D. R. SMITH—Two new species, larval descriptions, and life history notes i of some Panamanian sawflies (Hymenoptera: Argidae, Tenthredinidae) ................ 191 i LAVIGNE, R. J. and D. S. DENNIS—Ethology of three coexisting species of Efferia (Diptera: Asilidaeyant viexico t piso se SU oly HORE ee PIR DE SEPT RPE MRD Pardee hh biol lana ee i 146
(Continued on back cover)
THE
ENTOMOLOGICAL SOCIETY OF WASHINGTON
ORGANIZED MARCH 12, 1884
OFFICERS FOR 1985
DoNALD M, ANDERSON, President MICHAEL E. SCHAUFF, Program Chairman EDWARD M. Barrows, President-Elect GEOFFREY B. WuiTE, Membership Chairman THOMAS E. WALLENMAIER, Recording Secretary Vicror L. BLACKBURN, Custodian RICHARD G, Rossins, Corresponding Secretary MANYA B. STOETZEL, Delegate, Wash. Acad. Sci.
THOMAS J. HENRY, Treasurer HELEN SOLLERS-RIEDEL, Hospitality Chairman RAYMOND J. GAGNE, Editor
Publications Committee DaAvip R. SMITH THEODORE J. SPILMAN GEORGE C. STEYSKAL
Honorary President C. F. W. MUESEBECK
Honorary Members FREDERICK W. Poos ASHLEY B. GURNEY THEODORE L. BISSELL
All correspondence concerning Society business should be mailed to the appropriate officer at the following address: Entomological Society of Washington, c/o Department of Entomology, NHB 168, Smithsonian Insti- tution, Washington, D.C. 20560.
MEETINGS. — Regular meetings of the Society are held in the Natural History Building, Smithsonian Institution, on the first Thursday of each month from October to June, inclusive, at 8 P.M. Minutes of meetings are published regularly in the Proceedings.
MEMBERSHIP.— Members shall be persons who have demonstrated interest in the science of entomology. Annual dues for members are $15.00 (U.S. currency) of which $13.00 is for a subscription to the Proceedings of the Entomological Society of Washington for one year.
PROCEEDINGS. — Published quarterly beginning with January by the Society at Washington, D.C. Members in good standing receive the Proceedings of the Entomological Society of Washington. Nonmember subscriptions are $30.00 per year, domestic, and $35.00 per year, foreign (U.S. currency), payable in advance. Foreign delivery cannot be guaranteed. All remittances should be made payable to The Entomological Society of Washington.
The Society does not exchange its publications for those of other societies.
Please see p. 730 of the July 1984 issue for information regarding preparation of manuscripts.
STATEMENT OF OWNERSHIP Title of Publication: Proceedings of the Entomological Society of Washington. Frequency of Issue; Quarterly (January, April, July, October).
Location of Office of Publication, Business Office of Publisher and Owner: The Entomological Society of Wash- ington, c/o Department of Entomology, Smithsonian Institution, 10th and Constitution NW, Wash- ington, D.C. 20560.
Editor: Raymond J. Gagné, Systematic Entomology Laboratory, c/o U.S. National Museum NHB 168, Wash- ington, D.C. 20560.
Managing Editor and Known Bondholders or other Security Holders: none. SPB Be SORE ELE SE Re EE ee
This issue was mailed 31 January 1985 Second Class Postage Paid at Washington, D.C. and additional mailing office.
PRINTED BY ALLEN PRESS. INC., LAWRENCE, KANSAS 66044, USA
PROC. ENTOMOL. SOC. WASH. 87(1), 1985, pp. 1-24
CULEX (CULEX) PIPIENS LINNAEUS (DIPTERA: CULICIDAE): CONCEPTS, TYPE DESIGNATIONS, AND DESCRIPTION!
RALPH E. HARBACH, CHRISTINE DAHL, AND GRAHAM B. WHITE
(REH) Walter Reed Biosystematics Unit, NHB-165, National Museum of Nat- ural History, Washington, D.C. 20560, USA; (CD) Department of Zoology, Sec- tion of Entomology, University of Uppsala, Box 561, S-75122 Uppsala, Sweden: (GBW) Department of Entomology, London School of Hygiene and Tropical Medicine, Keppel Street (Gower Street), London WCIE 7HT, England.
Abstract.—The history of the concept of Culex pipiens Linnaeus is reviewed. An illustration of a syntype published by Réaumur is designated as the lectotype of pipiens. A lectotype is also designated for Culex bifurcatus Linnaeus, which is stabilized as a synonym of pipiens. A neotype for pipiens is designated in place of the non-extant lectotype-specimen. The adult, pupal, and larval stages of the neotype are described and illustrated. Sexual differences are described and illus- trated for the alloneotype. A description of diagnostic and variable characters is provided for each life stage of the species.
The present system of naming organisms originated with the work of Carolus Linnaeus, and the 10th edition of his Systema Naturae (1758) is the designated starting point of zoological nomenclature. In that volume, Linnaeus described six species of Culex, the first being Culex pipiens. The family Culicidae (Stephens, 1829) is based on the genus Culex for which pipiens is the type-species (Latreille, 1810). Of the other species described in 1758, only Culex bifurcatus remains in Culicidae, but has been treated as a synonym of pipiens (Martini, 1922; Edwards, 1932; Stone et al., 1959; Knight and Stone, 1977).
The Linnaean Collection is now the property of the Linnean Society of London. It includes two original mosquito specimens located in Diptera box 22 (transferred from old box 195). One of the specimens is a female bearing Linnaean labels (sense of Day and Fitton, 1978) inscribed “Culex.” and “‘l. pipiens” (Fig. 1A—D). This specimen represents a species of Aedes (Ochlerotatus), but it is badly damaged and unidentifiable to species. The other specimen is a male of the maculipennis complex of Anopheles (White, 1978), and it bears a Linnaean label inscribed ‘‘2 bifurcatus.” (Fig. 1E, F). These findings revealed the need to stabilize the Linnaean mosquito nomenclature relevant to the western Palearctic fauna (Dahl and White, 1977) and the taxonomy of the pipiens complex under investigation by the senior author as part of a revision of the Culex (Culex) of northern Africa and south- western Asia.
' The opinions or assertions contained herein are the private views of the authors and are not to be construed as official or as reflecting the views of the supporting agencies.
nN
PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON
Cdex.
8 0
L. ifuety,
E F
Fig. 1. A-—D, Linnaean labels (A) accompanying the original female specimen (B-D) of Culex pipiens Linnaeus in the Linnaean Collection. The specimen actually belongs to a species of Aedes (Ochlerotatus). E, F, Linnaean label (E) accompanying the original male specimen (F) of Culex bifur- catus Linnaeus in the Linnaean Collection. The specimen actually belongs to a species of the mac- ulipennis complex of Anopheles. B, dorsal aspect; C, F, left sides; D, might side.
There is no reason to doubt the authenticity of the mosquito specimens in the Linnaean Collection. Where Linnaean species are represented by a single specimen ‘hich bears a handwritten label with the specific name and the number of the species in the 10th edition of Systema Naturae, the specimen is likely to be a Linnaean original (Lindroth, 1957; Day and Fitton, 1978; Robinson and Nielsen,
VOLUME 87, NUMBER 1 3
1983). The specimen labelled pipiens also bears the generic name of Cu/ex because it represents the first species which Linnaeus described for the genus.
The taxonomic units conceived by Linnaeus were based on conspicuous mor- phological characteristics, not phylogenetic interpretations. His concepts were much broader than ours. Accordingly, the genus Cu/ex, as it was formalized in 1758, included species of Ceratopogonidae (pulicaris), Simuliidae (reptans,; equi- nus) and Empididae (sterocoreus) in addition to Culicidae (Dyar and Knab, 1909; Knight, 1972). Regarding the Culicidae, it is obvious that Linnaeus discerned only two general kinds of mosquitoes, the twittering mosquito (Culex pipiens) and the mosquito with a forked beak (Culex bifurcatus). These concepts were published in the first edition of Fauna Svecica (Linnaeus, 1746, species numbers 1116 and 1115, respectively) before they appeared in the 10th edition of Systema Naturae along with their assigned specific names.
Reference made by Linnaeus (1746) to “Culex ipse in sy/vis, praesitem Lap- poniae”’ is evidence that he included forest Aedes in his concept of pipiens. On the other hand, Linnaeus cited illustrations in the works of Swammerdam (1737: pl. 31, Figs. 4-8; pl. 32, Figs. 1-5), Réaumur (1738: pls. 43, 44) and Joblot (1754: pl. 13, Figs. A-E, H, I, L) which unequivocally depict a species of Cu/ex compatible with our modern concept of pipiens (see Réaumur’s plates which are reproduced here in Figs. 2 and 3). Moreover, observations made by Linnaeus concerning ““Hominibus & animalibus sono alarum & sanguinis suctu molestissimus”’ (1746) and “Jnsectum pipiens, pungens” (1758) imply that pipiens represented any culi- cine female known to him. Linnaeus (1758) recognized the ubiquity of these animals when he wrote “Habitat in Europae aquosis; copiosissima in Lapponia; etiam in America obvia.”’ This statement is the source for what are regarded as the type-localities of pipiens.
Virtually all workers of the late 1 8th and early 1 9th centuries (notably: Linnaeus, 1761, 1767; Sulzer, 1761; Geoffroy, 1762; Scopoli, 1763; Schaeffer, 1766; De Geer, 1776; Fabricius, 1781, 1787, 1794, 1805; Schrank, 1781, 1803; Herbst, 1787; de Villers, 1789; Gmelin, 1790; Rossi, 1790; Olivier, 1791; Meigen, 1804; Latreille, 1809) used the name of pipiens in the sense of Linnaeus (1758). It was not until Meigen (1818) separated the genera of Aedes and Anopheles from Culex that the present concept of pipiens began to develop. This concept was founded on the descriptions and illustrations in the works cited by Linnaeus, particularly the detailed account of the immature stages given by Réaumur (1738) (see Figs. 2, 3). The exact identity of pipiens remained rather obscure until Dyar and Knab (1909) figured the male phallosome based on a comparison of specimens from Denmark, France, Hungary, and the United States. The current concept was fully realized when Martini (1925) distinguished and described Culex torrentium, a species previously confused with pipiens in Europe.
Incidentally, De Geer (1776) substituted the name of Culex communis for Linnaeus’ concept of pipiens. This is evident from the synonyms De Geer lists under communis. Most important among these are Culex pipiens and the figures in Blankaart (1688), Swammerdam (1737), and Réaumur (1738) on which Lin- naeus partially based his concept of this species. De Geer’s two specimens of communis in the Naturhistoriska Riksmuseet, Stockholm, are actually species of Culiseta. This fact supports Linnaeus’ broad concept of pipiens as including other culicine species. De Geer exchanged correspondence and specimens with Linnaeus
4 PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON
Fig. 2. Plate 44 of Réaumur (1738) on which Linnaeus (1758) partially based his concept of Culex pipiens. Notice that the eggs illustrated upside down bear a corolla (“‘col’’) characteristic of Culex.
VOLUME 87, NUMBER 1 5
who identified much of De Geer’s material. Needless to say, Culex communis De Geer was treated as a synonym of Culex pipiens Linnaeus by most authors until Edwards (1921) revived the name for the species previously commonly referred to as Aedes nemorosus (Meigen). Aedes communis 1s now recognized as the nom- inotypical member of a sibling species complex (Ellis and Brust, 1973).
Culex pipiens is one of the most important mosquito species for reasons of its widespread abundance, biological characteristics, and taxonomic significance. Un- fortunately, it belongs to a complex of species which represents one of the major outstanding problems in mosquito taxonomy. Basic to the resolution of this prob- lem is the delimitation and fixation of the names of the taxa involved. The first step in this direction was taken by Belkin (1977) who established the priority of Culex quinquefasciatus Say over Culex fatigans Wiedemann and set the stage for Sirivanakarn and White (1978) to designate a neotype for quinquefasciatus. Re- cently, Harbach et al. (1984) designated a neotype to fix the identity of Culex molestus Forskal.
The need for a primary type-specimen for Culex pipiens is apparent. Obviously, acceptance of the venerable specimen in the Linnaean Collection as the “type” of pipiens (reasoning of Crosskey, 1974), or designation of this specimen as the lectotype (reasoning of Vane-Wright, 1975), would drastically upset the current concept of pipiens and catastrophically alter the accepted meaning of the genus Culex. Fortunately, the specimens portrayed in the published illustrations upon which Linnaeus founded his concept are valid syntypes of pipiens (Article 73b(i), International Code of Zoological Nomenclature, 1985). Accordingly, the larva illustrated by Réaumur (1738: pl. 43, Fig. 3) (see Fig. 3) is hereby designated lectotype of Culex pipiens Linnaeus. Unfortunately, Réaumur studied insects as natural history objects without preserving specimens for posterity, and the lec- totype-specimen is non-extant (this does not invalidate the lectotype designation, see Article 74c of the Code). Furthermore, Réaumutr’s illustration of the lectotype does not show the characters which distinguish pipiens from torrentium. For these reasons, a neotype is designated below. It should be noted that the existence of a paralectotype (the Linnaean specimen in London) does not preclude the desig- nation of a neotype (Article 75b(iii) of the Code).
The neotype designation satisfies all of the qualifying conditions of the Code except, perhaps, Article 75b(5) which requires that a neotype be collected as near as practicable to the original type-locality. As indicated above, the “original” type-locality of pipiens was very broad, comprising Europe, Lapland, and America (Linnaeus, 1758), but with the lectotype designation, the provenance of the type is restricted to France. However, the neotype selected below originated in Sweden. Thus, the neotype designation fixes a type-locality for pipiens which is not the same as that established by the lectotype. Nevertheless, we feel justified in selecting a neotype from elsewhere for the following reasons. First, emphasis was placed on obtaining specimens of the genus and species so well figured by Réaumur. Secondly, it seemed desirable to acquire material from a Swedish locality that could have been visited by Linnaeus. Finally, and most importantly, although material was available to us from France, none of it was individually reared and did not meet our high standards for type-material. A precedent for selecting a neotype from outside the original type-locality was established in the case of Culex
6 PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON
ps
43 fo 056.Mint 43 de Ltece keelncccte: Tom
Fig. 3. Plate 43 of Reaumur (1738) on which Linnaeus (1758) partially based his concept of Culex pipiens. “Fig. 3” illustrates the lectotype larva. Notice that the antenna and siphon of the larva are typical of Culex.
VOLUME 87, NUMBER 1 7
aegypti Linnaeus (Mattingly et al., 1962; International Commission for Zoological Nomenclature, 1964).
The case of bifurcatus can be dealt with in parallel with that of pipiens, although neotype designation is unnecessary. When Linnaeus described bifurcatus, he cited figures 1 and 2 of plate 40 in Réaumur (1738) (reproduced here in Fig. 4) which clearly illustrate a Cu/ex male, presumably belonging to the same species whose immature stages are illustrated on the plates which Linnaeus cites as representing pipiens. Unfortunately, the relation of the sexes was unclear to Réaumur, and he referred to the male as a separate species based on the character of the maxillary palpi. Furthermore, Réaumur regarded the palpi as a forked beak, and this, of course, influenced Linnaeus when he named bifurcatus. With this in mind, it appears that Linnaeus’ concept of bifurcatus included any mosquito with long palpi, 1.e., anophelines and male culicines. In any case, the figures cited by Linnaeus eventually led Martini (1922), the first reviser, to synonymize bifurcatus with pipiens. To stabilize this synonymy, the male illustrated by Réaumur (1738: pl. 40, Fig. 2) (see Fig. 4) is hereby designated lectotype of Culex bifurcatus Linnaeus. This action has the advantage of not destabilizing current usage of the name for whichever Palearctic species of Anopheles is represented by the Linnaean speci- men.
MATERIALS AND METHODS
Our original intention was to select a neotype for pipiens from specimens orig- inating in Linnaeus’ garden in Uppsala. To our disappointment, collections made there in the summer of 1983 contained only larvae of Culex torrentium. In August of the same year, two collections containing substantial numbers of larvae con- forming to the prevailing concept of pipiens were made near Veberéd, Scania, Sweden. These were expeditiously shipped to Washington, DC, where the larvae were individually reared for study. A neotype (specific information is provided with the designation below) and an alloneotype were chosen from one of these collections. Seventy-eight adults (9 males and 6 females with pupal exuviae; 34 males and 29 females with larval and pupal exuviae) and 14 fourth-instar larvae comprised the collection containing the neotype and alloneotype. Additionally, 34 adults (9 males with pupal exuviae; 14 males and 11 females with larval and pupal exuviae) and 10 fourth-instar larvae were obtained from the second col- lection which was made in an old boat on the shore of Lake Krankesj6 near the site where the neotype originated. The collections also yielded 392 specimens of Culex torrentium: some 67 males, 69 females, 136 pupal exuviae, 102 larval exuviae, and 18 fourth-instar larvae.
Detailed descriptions and illustrations of the adult, pupal, and larval stages of the neotype are provided. Characters which differ in the alloneotype are described, and some important adult sexual differences are illustrated. A description of diagnostic and variable characters for each life stage based on associated specimens is also included. Character measurements, setal counts, and setal branching counts were made on the neotype and alloneotype, and 10 specimens collected with them. The morphological terminology follows Harbach and Knight (1980), except that siphon indices were calculated using the basal width of the siphon rather than the width measured at midlength. Ratios of DV/D (Sundararaman, 1949) were de- termined using the method of Barr (1957).
8 PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON
=
SS
= RX ws SSS :
Fug.t0-
Fig. 4. Plate 40 of Réaumur (1738) showing figures | and 2 on which Linnaeus partially based his concept of Culex bifurcatus. “Fig. 2” illustrates the lectotype male.
VOLUME 87, NUMBER | 9
Culex (Culex) pipiens Linnaeus
Neotype (hereby designated): 6 (1-42) with associated larval and pupal exuviae and genitalia on slides, and with the following collection data: SWEDEN, Scania, Veberéd, Silvakra farm, near Lake Krankesj6, open water reservoir (vessel), 15 m ASL, 23.8.1983, Coll. C. Dahl. Deposited in the National Museum of Natural History, Washington, DC, Type No. 101370.
Male (Neotype) (Fig. 5). A medium-sized mosquito closely resembling Culex torrentium Martini but without prealar scales and with different genitalia. Also closely resembling Culex quinquefasciatus Say and Culex restuans Theobald, but differing in the character of the scutal scaling, tergal banding, and genitalia. Head: Length of antennal flagellum 1.6 mm; flagellomeres 1-12 pale between whorls; flagellomeres 13 and 14 dark, combined length about 0.7 mm, approximately 0.45 of flagellum length; pedicel orange, black mesally. Proboscis mainly black- scaled, with ventral patch of white scales 0.3 to 0.7 from base; without ventral cluster of setae at false joint; length 2.2 mm, false joint 0.6 from base; labella dark. Length of maxillary palpus 2.9 mm, 1.3 of proboscis length, extending beyond tip of proboscis by length of palpomere 5; palpus mainly black-scaled, integument between palpomeres 2 and 3 pale; lateral surface of palpomere 3 with stripe of white scales 0.4 to 0.9 from base, stripe bordered ventrally by row of about 22 long black setae on distal 0.5 of palpomere, ventral surface devoid of scales and pale, with ventromesal row of about 18 small antrorsely-curved setae; palpomere 4 with nearly complete narrow ventral stripe of white scales; palpomere 5 with small patch of white scales ventrally at base; lateral surfaces of palpomeres 4 and 5 with long black setae that are longest at base of 4 and become gradually shorter toward apex of 5. Forked scales of vertex rather short, mainly dark, some pale medially; falcate scales narrow, pale yellow, slightly paler laterally; lateral spatulate scales yellowish white. Ocular setae black, antrorsely curved. Interocular space narrow, setae yellowish brown. Thorax (Fig. 5B): Pleural integument yel- lowish brown, darker anteriorly, faded posteriorly; scutal integument dark brown. Scutal scales fine, golden brown with slight reddish tint, somewhat finer on fossae and supraalar areas; integument and scales between supraalar and posterior dor- socentral setae not noticeably darker; pale yellow scales on outer margins of supraalar and prescutellar areas. Scutal setae nearly black (many prescutellar setae missing). Scutellum with narrow pale yellow falcate scales; 5 large setae on each lateral lobe (1 missing on left lobe), 8 on median lobe (6 missing). Antepronotum with 2 patches of narrow falcate scales, upper patch golden brown, scales of lower patch coarser and very pale yellow; setae mainly dark, pale ventrally. Postpro- notum with golden-brown falcate scales, paler and slightly coarser posteriorly; with 5 setae on posterodorsal margin, longer and paler posteriorly. Pleural setae golden, numbers on left side as follows: 19 upper proepisternal in more or less double row, 10 prealar, 5 upper mesokatepisternal, 10 lower mesokatepisternal with uppermost seta very prominent, 7 upper mesepimeral, and 1 prominent lower mesepimeral. Pleural scales spatulate, few below upper proepisternal setae yellowish, others white and in patches as follows: patches on upper corner and lower posterior border of mesokatepisternum, anterior patch on mesepimeron at nearly same level as upper mesokatepisternal patch, and small patch before upper mesepimeral setae; without postspiracular and prealar scales. Wing: Length 3.4
10 PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON
A af ae Culex (Culex) pipiens \ Ws IN : / ne un ’ \
an a Aye Sh Sates ay a |
K
~ J
10 SAN a H&E “ |G \) ai \, ~ ( LAN es
alloneotype
< 1-91 A
Cc D neotype L | i rete ] 1-42 \\ had aan \ \ vi \! pei )\ ) ( / ( HC 7 / ‘ 1-26 ( | f ; a) 01 1 1-73
+ \ i ama AKAN AUR EATEANY Vy NAN SENN NSSARSSS SSO alloneotype
220'= 4
neotype
S Likiiok
neotype alloneotype
VOLUME 87, NUMBER 1 11
mm; cell R; 2.9 of R,, 3; subcosta intersects costa before furcation of R,,,; cell M, 0.8 of cell R;; scales entirely dark. Dorsal scaling: broad squame scales on costa, subcosta, R, R,, and CuA; relatively narrow squame scales on R,.;, M,, M,, M;,4, mcu, and proximally on 1A; linear plume scales on R,, R>,3, Rs, R;, M, M,.5, and distally on 1A; remigium with 2 distinct rows of scales and 2 setae (3 on right wing) distally. Ventral scaling: squame scales on costa, subcosta, base of R,, R,, R>.;, bases of R, and R;, M,.,, and bases of M, and M,; plume scales on other veins and parts of veins except CuA before mcu and proximal 0.5 of 1A which are devoid of scales. Halter: Entirely pale. Legs (Fig. SE): Anterior surface of forecoxa mainly black-scaled, with small basal patch of yellowish scales, anterior surface also with many long, nearly black, ventrally-curved setae, apex with 4 setae On posterior margin, most proximal seta nearly perpendicular to surface, others project ventrally; midcoxa with midlateral longitudinal row of 4 long dark setae and 6 or 7 short pale setae, setae margined anteriorly by longitudinal patch of white spatulate scales, anterior surface with small patch of black scales and several ventrally-projecting setae at apex; posterolateral surface of hindcoxa with longitudinal row of 7 long golden setae becoming gradually shorter from base to apex, anterolateral surface with narrow longitudinal row of nearly colorless scales and 4 short ventrally-projecting setae at apex, mesal surface with 2 dark setae at apex. Ventral surfaces of trochanters with white spatulate scales; anteroventral surface of fore- and midtrochanters with some black spatulate scales. Apices of all femora with narrow dorsal border of yellowish scales; forefemur with anterior surface black-scaled, posterior surface white-scaled; midfemur like forefemur but black scales extending over dorsal surface toward apex; hindfemur mainly white- scaled, with complete anterodorsal stripe of black scales gradually widening to extend over entire anterior and ventral surfaces just before apex. Foretibia mainly black-scaled, with white scales posteroventrally; midtibia with black scales an- teriorly, white scales posteriorly; hindtibia mainly black-scaled, posteroventral surface with whitish scales, anterior and dorsal:surfaces with whitish scales at apex. Tarsi black-scaled, tarsomere | of fore- and midlegs with some pale scales on posterior surface. Pulvilli pale. Ungues black; anterior foreunguis larger than posterior foreunguis, posterior foreunguis about 0.75 length of anterior foreunguis, both stout, anterior foreunguis with small ventral tooth near midlength, posterior foreunguis with small tooth nearer base; anterior midunguis like that of foreleg, posterior midunguis shorter and more slender than posterior unguis of foreleg but with tiny ventral tooth near midlength; hindungues very small, simple. Abdomen (Fig. SG): Terga mainly black-scaled; tergum I golden setose, with posteromesal patch of black scales; posterior margins of terga II-VII with row of golden setae, median setae about length of basal band of next tergum, lateral setae longer, about 0.75 of tergum length; lateral scale-free areas of terga II-VII with long, laterally- directed, posteriorly-curved golden setae; tergum II with small basomedian spot
—
Fig” 5. Culex (Culex) pipiens Linnaeus. A, Scutum of alloneotype female. B, Thorax of neotype male (left side). C, Female cibarial armature. D, Female genitalia. E, Legs of neotype male (anterior aspect of left legs). F, Right wing of alloneotype female (dorsal). G, Abdomen of neotype male (dorsal). H, I, Abdomen of alloneotype female (H, dorsal; I, left side). Scales in mm.
12 PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON
of yellowish scales; terga II—VII with basal yellowish bands 0.35 of tergum length, bands of terga V-VII produced posteriorly along lateral scale-free areas, partic- ularly on terga VI and VII; anterior 0.5 of tergum VIII (ventral in position) with yellowish scales, posterior 0.5 with golden-brown setae and indistinct dusky scales, posterior margin without emargination in middle. Sterna II-VI mainly yellowish (scales same color as basal bands of terga), with some median black scales; sternum VIII (dorsal in position) mainly clothed with yellowish scales, with some subtle dark scales in middle; sterna II-VII golden setose, posterior margin of sternum VIII with row of golden setae. Genitalia (Fig. 6C-H): Ninth tergal lobes small, each with 12 setae in 2 irregular rows. Gonocoxite normal, ventrolateral setae strongly developed, these longer and stouter than lateral setae, mesal surface with 5 rows of small setae extending from base to level of subapical lobe; subapical lobe undivided, setae a—fin more or less straight row with gap between c and d, seta g immediately lateral to d-f, seta / lateral to g; a—c slightly curved, a stout and rodlike with blunt apex, ) and c each with stout base, tapering distally and with apex hooked and pointed; d—f shorter than a—c, hooked apically, d and e slender, flaterally flattened and appearing broad in lateral view; g foliform, longer than broad, slightly asymmetrical, apex not sharply produced; / slender, bent distally. Gonostylus stout, curved, concave dorsal surface with 2 small, slender setae on distal 0.3; gonostylar claw short, broadest apically, troughlike. Phallosome longer than broad with lateral plates and aedeagus of nearly equal length; lateral plate with definite dorsal, lateral, and ventral arms, dorsal arm broad, apex nearly truncate, diverging laterally from its mate of the opposite side, appearing slightly sinuous in lateral view; lateral arm broad in lateral view, its posterior margin more or less trilobed, the ventral lobe more prominent than the others and bent ventrolaterally, base of lateral arm with thumblike dorsal process, base of this process continuous mesally with dorsal aedeagal bridge; ventral arm narrow and sharply curved with apex directed laterally, DV/D —0.03. Aedeagus subcylin- drical, narrowed distally; ventral aedeagal bridge relatively wide, joining aedeagal sclerites just beyond midlength. Proctiger without distinctive features; paraproct with small, conical basal lateral arm, crown dark with numerous short, spinelike spicules. Cercal sclerite elongate, somewhat kidney-shaped, caudal margin irreg- ular; 6 and 5 cercal setae on left and right sides, respectively. Tergum X rectangular, adjoining paraproct below and behind basal lateral arm.
Alloneotype: 2 (1-91) with associated larval and pupal exuviae and same data as neotype.
Female (Alloneotype). Like neotype except as follows. Head: Length of antennal flagellum 2.1 mm, entirely dark, pedicel and flagellomere | with tiny pale scales on mesal surface. Ventral surface of proboscis with creamy-white scales 0.1 to 0.7 from base. Maxillary palpus entirely black-scaled; length 0.36 mm, about 0.16 of proboscis length. Forked scales of vertex more numerous, more pale scales medially. Cibarial armature (Fig. 5C, described and illustrated from specimens collected with the alloneotype): Cibarial crest concave, slightly produced in mid- dle: with about 28 short, blunt teeth. Cibarial dome nearly elliptical in dorsal outline, produced anteriorly in middle; surface largely granular, imbricated pos-
‘riorly. Thorax (Fig. SA): Scutal and pleural scales and setae same as neotype <cept scales and integument between supraalar and posterior dorsocentral setae noticeably darker, forming a pair of ovoid spots, and upper proepisternal scales
VOLUME 87, NUMBER | 13
more numerous, with small patch of indistinct dusky scales on mesal side of setae; postpronotum with 6 setae and few narrow spatulate scales interspersed with pale falcate scales posteriorly; numbers of pleural setae on left side differ as follows: 8 upper proepisternal, 8 prealar, 8 lower mesokatepisternal, and 9 upper mesepi- meral. Wing (Fig. 5F): Length 4.3 mm; cell R, 5.2 of R,,,; subcosta intersects costa beyond furcation of R,,3; cell M, 0.75 of cell R,; remigium of both wings with 2 setae. Legs: Like neotype except midcoxa with 5 large and 5 small setae in midlateral row; hindcoxa with 9 setae on posterolateral surface. Pulvilli distinct. Ungues small, simple, black; fore- and midungues stouter than hindungues. Ab- domen (Fig. 5H, I): Tergum II with basomedian spot of yellowish scales and lateral patches of white scales: terga III-VII with basal bands of yellowish scales and basolateral spots of white scales, spots becoming gradually larger on succeeding posterior terga and actually cover entire lateral surfaces of tergum VII, bands 0.25 of tergum length, slightly convex on terga III and IV and not quite reaching spots, straight on terga V—VII and contiguous with spots, bands of terga VI and VII slightly produced posteriorly along mesal side of spots; tergum VIII clothed with whitish scales. Sterna II—-VII like those of neotype; sternum VIII with whitish scales on lateral margins, broad median area without scales. Genitalia (Fig. 5D, described and illustrated from specimens collected with the alloneotype): Sternum VIII with rounded median posterior emargination. Tergum IX narrow, postero- lateral margin with 6-8 setae. Upper vaginal lip narrow, distinct; 8-11 insular setae in dense cluster immediately anterior to indistinct lower vaginal lip. Upper vaginal sclerite distinct, U-shaped. Postgenital lobe short, apex evenly rounded, with submedian row of 7 setae extending from dorsal to ventral surface over apex. Cercus short, laterally compressed, apex bluntly rounded, lateral and ventral surfaces setose; cercus/dorsal postgenital lobe index about 2.3.
Specimens collected with the neotype and alloneotype exhibited the following variation. Head: Proboscis length 2.16-2.26 mm, mean 2.20 mm. Maxillary palpus of female sometimes with some pale scales dorsally on palpomere 4, length 0.36-0.40 mm, mean 0.39 mm; length of palpus in male 2.80-3.00 mm, mean 2.9 mm, often extending beyond tip of proboscis by less than length of palpomere 5. Thorax: Ovoid spots of scutum usually moderately distinct in dorsal aspect, usually indistinct in lateral view; acrostichal and fossal scales often as dark as ovoid spots in females. Postpronotum with 5-9 setae, most often with 5. Pleural setae as follows: females with 8-13 and males with 18-25 upper proepisternal; both sexes with 8-12 prealar, 4-7 upper mesokatepisternal, 8-13 lower meso- katepisternal, and 5-12 upper mesepimeral. Prealar area occasionally with few inconspicuous pale scales immediately below setae (present on one side in 54 and 52 of 108 specimens examined). Wing: Length 4.3—4.5 mm in females, 3.4—3.7 mm in males; cell R,/R;,, 4.9-5.5 in females, 2.6-3.6 in males; cell M,/cell R, 0.72-0.82. Abdomen: Basal bands of terga always yellowish, usually 0.25-0.40 of tergum length in males, 0.15—0.35 in females, bands in males usually like those of neotype, bands in females variable as follows: basomedian spot of tergum II often nearly lost, tergum III frequently with convex basomedian spot, bands of terga VI and VII often very narrow, that of VII sometimes absent in middle. Tergum VIII of female with yellowish or whitish scales, scales usually paler than those of basal bands of proceeding terga; sometimes with some dark scales pos- teriorly in middle. Sterna almost always with dark scales medially, dark scales
14 PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON
sometimes on proximal 0.5 only, sometimes arrayed as speckles, often forming complete stripe; sternum VIII of males (dorsal in position) usually largely pale with dusky scales posteriorly in middle, sometimes entirely pale. Genitalia (of male): DV/D zero to —0.19, mean —0.09 (for 15 specimens).
Pupa (Neotype) (Fig. 6A, B). Character and positions of setae as figured. Ceph- alothorax: Lightly tanned, legs, scutum, metanotum, and metathoracic wings dark- er. Trumpet: Moderately tanned, subcylindrical, gradually widened distally, index 4.9; tracheoid area darker, extending 0.3 from base; pinna oblique, about length of tracheoid area. Abdomen: Lightly tanned, terga I-VI darker in middle; length 2.9 mm. Genital lobe: Lightly tanned; length 0.3 mm. Paddle: Lightly tanned, midrib and buttress darker; outer margin without distinct spicules; midrib distinct except at apex; length 0.9 mm, width 0.6 mm, index 1.5.
The alloneotype resembles the neotype except as follows: trumpet index 5.4, abdominal length 3.3 mm; genital lobe length 0.2 mm; paddle 1.0 x 0.7 mm with index 1.4.
Table | lists the range and modal number of branches for pupal setae observed in the neotype, alloneotype, and 10 associated specimens. Diagnostic and variable characters follow. Cephalothorax: Setae 1, 2-CT with 3-5 branches; 3, 4-CT with 2 or 3 branches, commonly 3; 5-CT usually with 5 branches (3-6); 6-CT usually with 3 branches (1-4); 7, 9, 11-CT double; 10-CT frequently with 6 branches (6- 13); 12-CT usually double (2-4). Trumpet: Index 4.8-6.9, mean 5.5. Abdomen: Seta 6-1, II single; 7-I, I] usually double; 1-II] usually with more than 20 branches (15-26); 1-III-V frequently with at least 6 branches, |-III usually with 8 branches (6-10), 1-IV usually with 6 or more branches (3-8), 1-V most often with 6 branches (4-6); 2-II, VII lateral to seta 1, 2-III-VI mesal to 1; 5-IV usually triple, sometimes with 4 or more branches; 5-V—VII almost always double; 6-III, IV most often with 3 branches, 6-V, VI most often with 4 branches. Paddle: Index 1.2-1.6, mean 1.4.
Larva (Neotype) (Fig. 7). Character and placement of setae as figured. Head: Wider than long, length 0.7 mm, width 1.0 mm; approximately anterior 0.5 lightly tanned, labiogula, lateralia posterior to seta 10-C, and dorsal apotome behind seta 5-C moderately tanned. Median labral plate narrow but distinct, anterior margin slightly emarginate between insertions of seta 1-C. Labiogula narrower anteriorly than posteriorly, length about same as posterior width; hypostomal suture complete, extended posterolaterally from posterior tentorial pit to near collar. Collar best developed along lateralia, heavily tanned. Mouthparts devel- oped for filter-feeding. Dorsomentum with 8 and 9 teeth on left and right sides of median tooth, respectively. Antenna: Length 0.5 mm, 0.7 length of head; moderately tanned, mesal surface with dark spot at base; seta 1-A 0.7 from base, part proximal to seta I-A with strongly-developed aciculae mainly on dorsal and lateral surfaces, distal part slender and smooth except for few short aciculae laterally near seta 1-A. Thorax: Integument hyaline, tubercles of all large setae
—
Fig. 6. Culex (Culex) pipiens Linnaeus, neotype male. A, B, Pupa (A, dorsolateral aspect of ceph- alothorax, left side; B, dorsal and ventral aspects of left side of metathorax and abdomen). C-H, Genitalia, aspects as indicated (C, gonocoxite; D, phallosome; E, F, lateral plate and aedeagal sclerite; G, icrgum IX; H, proctiger and terga X). Scales in mm.
VOLUME 87, NUMBER 1
Culex (Culex) pipiens
PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON
16
“A[UO SNJOIATY 5 (apo) adury q “Way? YUM pazoa]joo suatutoads Q] pur ‘adAjoouoyje pue adAjoou ay} UO apeU S]UNOD UO Paseg ¢
- = - = = = = = = = (Z) b-7 = = = ec> Wet. pen = Cara z - = = z z = e (9) €1-9 ~ =. = (O)6-o) Wipicae I I I (ITT c - = - (oz (b)S-E Ss ()S-ZsE (s) 9-€ a = (b) 9-€ = = = I (pr) 9-€ —— (€) S-7 (9) L-€ z (2) € ‘2 ( - - - ()9-z (p)S-Z_— (HSE (ED OZ (€) p-Z 1 (€) b-1 o— = = (a)icail (7) €-1 c (€) 9-7 (S) 8-b (S)) ise (bp) 9-b (S) 9-€ = = z z (p)s-z (ph) 9-E OTH (S) L-€ (b) 9-€ (9) 8b (€) €°7 =< = = (2) €°7 z z (9) L-b Z (2) € ‘7 (€) € ‘7 I a = I I I I I CO carl I (p) S-€ Cra - (roe (pr) SE (9) OH (9) BE (3) 01-9 (pz) 97-S1_ (9) LL-EE al) SE
a 7 I I I I I I if = _
XI ILA ILA IA A Al Il I I ALS} SIPPPd xposyiojeyday sjuawdag jeuIWOpgy
I
esnoeuury suaidid (xajn) XajnD Jo aeias jednd 10} sayouesg JO JaquinN
al ol a at Ol 6
Joquinny B1aS
“T 1921
VOLUME 87, NUMBER | 17
moderately tanned; setae 1-3-P and 9-12-P, M, T oncommon tubercles. Abdomen: Integument hyaline, tubercles of setae 7-I, 6-I-VI and 1, 3- VIII moderately tanned, tubercle of setae 2, 3-X heavily tanned. Segment VIII: Comb consisting of 40 and 42 scales on left and right sides, respectively; scales short, evenly fringed on sides and apex, arranged in 4 irregular rows. Siphon: Index 5.0; subcylindrical, broadest at base, slightly sigmoid in lateral view; moderately tanned, darker at base; acus attached, longer on posterior side of attachment. Pecten on basal 0.3 of siphon, that of left and right sides with 14 and 11 spines, respectively, spines increasing in size from base of siphon, larger spines with 3 long basal denticles. Seta 1-S in 4 pairs, la-S very near most distal pecten spine, lc-S distinctly out of line with others. Segment X: Saddle complete; moderately tanned, darker dorsally; posterodorsal area with minute spicules; length 0.3 mm. Ventral brush (seta 4-X) with 6 pairs of setae arising from grid, setae increasing in length pos- teriad, most posterior seta 7.5 length of most anterior. Anal papillae elongate, subacutely tapered; dorsal pair longer than ventral pair, dorsal pair twice length of saddle.
The alloneotype resembles the neotype except for minor differences in the number of branches of some setae, and principally as follows: length of head 0.8 mm, width the same; comb with 49 and 52 scales on left and right sides, respec- tively; siphon index 4.9, pecten with 17 and 16 spines on left and right sides, respectively.
Table 2 lists the range and modal number of branches for larval setae determined from the neotype, alloneotype, and 10 associated specimens. Diagnostic and vari- able characters include the following. Head: Seta 1-C slender, tapered distally, slightly bent mesad; 3-C distinct; 2-C absent; 4-C single, rather long; 5-C frequently with 4 branches (4-6); 6-C most often with 4 branches (3-5); 7-C resembles 5, 6-C, with 6-10 branches; 8, 10-C usually double (2, 3); 11, 12, 13-C double or triple, 13-C occasionally with 4 branches; dorsomentum most often with 10 teeth (8-11) on either side of median tooth. Thorax: Setae 1-3-P single, nearly of equal length; 4, 7, 8-P double rarely single; 11-P usually with 4 or 5 branches (3-7). Seta 1-M usually single, occasionally double, about 0.5 length of 3-M; 3-M single: 4-M usually double, occasionally single. Seta 1-T short, 0.5 or less length of 2-T, most often double (1-4); 2-T usually single or double, infrequently with 3 or 4 branches. Abdomen: Seta 3-I, VII usually single, sometimes double (3-I double on at least one side in 27 of 112 specimens examined; 3-VII double in only 16); seta 6-I-VI long, 6-I, II usually with 3 branches (2-4), 6-III-VI double; 7-I re- sembles 6-I, almost always double; 1-III-VI usually double, one branch longer than the other. Segment VIII: Comb with 37-57 scales, mean 44; seta 1-VIII most often with 5 branches; 3-VIII frequently with 7 branches; 5-VIII with 3 or 4 branches, more often with 4. Siphon: Always slightly S-shaped in lateral view; index 4.5—5.8, mean 5.0; pecten with 11-17 spines, most often with 14; usually with 4 pairs of seta 1-S (in 112 specimens examined, 3 with 3 on one side, 9 with 5 on one side and 2 with 5 on both sides). Segment X: Seta 1-X usually single (double on one side in 10, and both sides in 3, of 112 specimens examined): ventral brush (seta 4-X) almost always with 6 pairs of setae.
DISCUSSION
Culex pipiens and Cx. torrentium are the only species of Culex (Culex) known to occur in central and northern Europe. As mentioned earlier, Cx. torrentium
PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON
Culex (Culex) pipiens
neotype 1-42
\ Wr a hy Nv
C '
VOLUME 87, NUMBER | 19
was confused with pipiens until Martini (1925) recognized it as a separate species. Since these species frequently occur together in the same habitat, it is possible that the figures published by the authors cited by Linnaeus (1758) were based on specimens of either pipiens or torrentium or both. Nevertheless, the concept of pipiens which has been handed down to us is that explicitly defined by Dyar and Knab (1909). This is the species we collected, reared, and selected as the neotype, alloneotype, and associated specimens of Culex pipiens.
In the adult stage, Cx. pipiens is reliably differentiated from torrentium by the character of the male genitalia. In forrentium, the posterior margin of the lateral plate bears a spiculate lobe and the dorsal arms are pointed. Mattingly (1951) denoted that torrentium could be distinguished from pipiens by the presence of prealar scales, but this character is not totally reliable (Service, 1968; Jupp, 1979: Onyeka, 1982). Prealar scales are occasionally absent in torrentium and sometimes present in pipiens. We have noted that specimens of torrentium contained in our two collections bear an inconspicuous dorsal pale spot at the apex of the foretibia. The potential diagnostic value of this character needs to be investigated in other populations of torrentium.
The larvae of pipiens and torrentium are similar. Natvig (1948) could not differentiate them and considered all larvae examined from Scandinavia and Finland to be pipiens. Sicart (1954) reported that the character of seta 1-X might be useful for separating the larvae of these species. He observed that seta 1-X was single in pipiens and double in torrentium. Later, Callot (1957) and Doby and Rault (1960) relied on this character to separate pipiens and torrentium in France. But as Service (1968) discovered, this seta may be single or double in either species, and cannot be used to separate them with confidence. As a matter of course, we examined the larval chaetotaxy of torrentium for comparison with that of pipiens. In so doing, we found that forrentium could usually be separated from pipiens by the character of seta 1-T. The length of this seta was normally greater than one-half the length of seta 2-T in torrentium and less than one-half in pipiens. In cases where this seta was missing, obstructed from view, or its length was questionable, the following combination of characters identified rorrentium: setae 3-I, VII and 1-X usually double with 3-I sometimes triple. These setae are usually, but not always, single in pipiens. Never were all three setae single in any one specimen of torrentium or double in any one specimen of pipiens. Seta 3-I was never triple in pipiens.
The taxonomy of the pipiens complex is an enigma complicated by interpre- tational difficulties and controversy associated with a number of bewildering morphological, behavioral/physiological, and genetic issues. An understanding of these issues is intimately related to the resolution of the pipiens complex. On the other hand, the delimitation and fixation of the concepts and their names for the taxa involved is fundamental to all taxonomic, morphological, behavioral/phys- iological, and genetic studies. Neotypes have now been designated for the three
—
Fig. 7. Culex (Culex) pipiens Linnaeus, neotype male, larva. A, Head (dorsal and ventral aspects of left side). B, Thorax and abdominal segments I-VI (dorsal and ventral aspects of left side). C, Dorsomentum. D, Abdominal segments VII-X (left side). E, Pecten spine. F, Comb scale. Scales in mm.
PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON
0
5
‘(apoul) adury 4
“WdY) YUM paloay[oo suautoads Q] pur ‘adAjoauojye pure adAjoau ay) UO APE S]UNOD UO paseg -
a
I I I I I = = = (L1) O€-TI (Z) 71 (€) p-Z (pz) se-6 (e)e ‘ce (Cebp‘e (€)9-@ (ZII7-6 (2) b-I (pb) 9-€ (81) 97-SI = I I I GO saat (7) €-1 (Z) €°T I I I (Z) p-I (QIEST (ig LT MOS Orel (Z) pol (9) L-€ (7) b-1 (2) pI (p) L-€ I I I I I I I I I I (S)IL=C QO ee I I I I (Oita (9) L-S (S) L-p I (S)S5c (7) p-7 I I I I = (6) 91-9 (CS) 9-b (Cie I I (S)9-p (S)8-p (9)L-E (pb) S-T ()T°1 (L) 6-¢ (LS. I (é (OI) LI-8 (é c c Cc (SiGe (€) b-7 CU Gaal I I (€) p- (Z) p-Z (iss CUES) SSC eS (1) €-1 (p) L-7 I I I I Caicat (€) p-Z I (1) €-1 (S) Sse (9) OI-r (p) 9-7 (ren C I I I I I I (I) 771 (€) 9-1 I I I I I I I I I C1) fel (2) p-1 I (€) p-7 Cy eare (ge Meee Nera CG tS (¢) c-1 (2) p-1 Ciel I oad I I I I I I I —- = _ a1 1) 81-6 a x ILA IIA lA A Al Il II I at W d sluauaag jeulWwOpgy xesoy ]
(€) p-7 SI (Ue 1 al (€) p-Z tl (NEG cl (g) €°T II (IEC Ol (p) 9-€ 6 (t) € °T 8 (Z)01=9 a2 (p) S-€ iS) (p) 9-b iS (Tl v I 3 = c I I I 0 ) 4aq peoH -WwnN P19
a —————————————————_————————
»snoeuury suaidid (xajnD) XajN-D JO BAIR] JPISUI-YLINOJ dY1 JO aeIAS IO} SayouBIg JO AQGWNN “7 IQVL
VOLUME 87, NUMBER 1 21
principal nominal forms of the pipiens complex: Culex quinquefasciatus Say, Culex molestus Forskal, and Culex pipiens Linnaeus. With this, an important step has been taken toward resolving the taxonomy of this important group of species.
ACKNOWLEDGMENTS
Sincerest gratitude is expressed to the following individuals for critically re- viewing and/or commenting on the manuscript: Curtis W. Sabrosky and F. Chris- tian Thompson, Systematic Entomology Laboratory, IIBIHI, USDA, Washington, D.C.; M. W. Service, Department of Medical Entomology, Liverpool School of Tropical Medicine, Liverpool; Richard V. Melville, Secretary, International Com- mission on Zoological Nomenclature; R. I. Vane-Wright, British Museum (Nat- ural History), London; and Bruce A. Harrison, E. L. Peyton, Ronald A. Ward, and Thomas J. Zavortink, Department of Entomology, Walter Reed Army In- stitute of Research, Washington, D.C. Special recognition is due Curtis W. Sa- brosky and F. Christian Thompson who took time from busy schedules to offer advice and discuss taxonomic procedures and nomenclature with the senior au- thor.
We are grateful to the Linnean Society of London, especially M. G. Fitton, Curator of Insects, for permission to examine and photograph the Linnaean mos- quito specimens. We are also grateful to P. I. Persson and B. Gustafsson, Natur- historiska Riksmuseet, Stockholm, for permitting the examination of the De Geer Collection. Thanks are due Ellen Welles, Special Collections, as well as Ruth Schallert and Sylvia Churgin, Natural History Library, Smithsonian Institution, Washington, D.C., for assistance with rare books and permission to have the plates in Réaumur photographed by the Medical Audio-Visual Services, Walter Reed Army Institute of Research. Appreciation is also expressed to Taina Litwak who produced Figs. 5-7, and Olimpia Areizaga who prepared the drafts of the manuscript. This study was supported in part by the Swedish National Science Research Council (grant no. 3318-110 to C. Dahl).
LITERATURE CITED
Barr, A.R. 1957. The distribution of Culex p. pipiens and C. p. quinquefasciatus in North America. Am. J. Trop. Med. Hyg. 6: 153-165.
Belkin, J. N. 1977. Quinquefasciatus or fatigans for the tropical (southern) house mosquito (Diptera: Culicidae). Proc. Entomol. Soc. Wash. 79: 45-52.
Blankaart, S. 1688. Schou-Burg der Rupsen, Wormen, Ma’den, en Vliegende Dierkens daar uit voorkomende. Door eigen ondervindinge by een gebragt. Amsterdam. vi + 232 pp., 17 pls.
Callot, J. 1957. Sur Culex torrentium Martini. Ann. Parasitol. Hum. Comp. 32: 438-442.
Crosskey, R. W. 1974. The British Tachinidae of Walker and Stephens (Diptera). Bull. Br. Mus. (Nat. Hist.) Entomol. 30: 267-308.
Dahl, C. and G. B. White. 1977. Culicidae, pp. 390-395. Jn Illies, J., ed., Limnofauna Europaea. Gustav Fischer Verlag, Stuttgart.
Day, M. C. and M. G. Fitton. 1978. Re-curation of the Linnaean Hymenoptera (Insecta), with a reassessment of the taxonomic importance of the collection. Biol. J. Linn. Soc. 10: 181-198.
De Geer, C. 1776. Mémoires pour servir a histoire des insectes. Vol. 6. Stockholm. 523 pp., 30 pls.
de Villers, C. J. 1789. Linnaei entomologia, faunae suecicae descriptionibus aucta; D. D. Scopoli, Geoffroy, De Geer, Fabricii, Schrank, &c. speciebus vel in systemate non enumeratis, vel nuperrime detectis, vel speciebus Galliae australis locupletata, generum specierumque rariorum iconibus ornata. Lugduni. ii + 656 pp., 4 pls.
Doby, J. M. and B. Rault. 1960. Complément a l'étude des caractéres morphologiques différentiels
22 PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON
de Culex pipiens Linné, 1758 et de Culex torrentium Martini, 1924 (Diptera Culicidae). Cah. Nat. 16: 113-121.
Dyar, H. G. and F. Knab. 1909. On the identity of Culex pipiens Linnaeus. Proc. Entomol. Soc. Wash. 11: 30-39.
Edwards, F. W. 1921. A revision of the mosquitoes of the Palaearctic Region. Bull. Entomol. Res.
12: 263-351.
1932. Genera Insectorum. Diptera. Fam. Culicidae. Fascicle 194. Belgium. 258 pp.
Ellis, R. A. and R. A. Brust. 1973. Sibling species delimitation in the Aedes communis (Degeer) aggregate (Diptera: Culicidae). Can. J. Zool. 51: 915-959.
Fabricius, J. C. 1781. Species insectorum exhibentes eorum differentias specificas, synonyma, auc- torum, loca natalia, metamorphosin adjectis observationibus, descriptionibus. Vol. 2. Hamburgi et Kilonu. 517 pp.
———. 1787. Mantissa insectorum sistens species nuper detectas adjectis characteribus genericis, differentiis specificis, emendationibus, observationibus. Vol. 2. Hafniae. 382 pp.
—. 1794. Entomologia systematica emendata et aucta, secundum classes, ordines, genera, species adjectis synonymis, locis, observationibus, descriptionibus. Vol. 4. Hafniae. vi + 472 pp.
———. 1805. Systema Antliatorum secundum ordines, genera, species adjectis synonymis, locis, observationibus, descriptionibus. Brunsvigae. xiv + 372 pp. + 31 unnumbered pages.
Geoffroy, E. L. 1762. Histoire abrégée des insectes qui se trouvent aux environs de Paris; Dans laquelle ces animaux sont rangés suivant un ordre méthodique. Vol. 2. Paris. 690 pp., 22 pls.
Gmelin, J. F. 1790. Caroli a Linné, Systema naturae per regna tria naturae, secundum classes, ordines, genera, species, cum characteribus, differentiis, synonymis, locis. Ed. 13, Vol. 1, Pt. 5. Lipsiae, pp. 2225-3020.
Harbach, R. E. and K. L. Knight. 1980. Taxonomists’ glossary of mosquito anatomy. Plexus Pub- lishing, Inc., Marlton, New Jersey. xi + 415 pp.
Harbach, R. E., B. A. Harrison, and A. M. Gad. 1984. Culex (Culex) molestus Forskal (Diptera: Culicidae): neotype designation, description, variation, and taxonomic status. Proc. Entomol. Soc. Wash. 86: 521-542.
Herbst, J. F. W. 1787. Kurze Einleitung zur Kenntniss der Insekten fiir Ungeiibte und Anfanger. Vol. 3. Berlin und Stralsund. 200 pp., 48 pls.
International Code of Zoological Nomenclature. 1985. Third edition. International Trust for Zoolog- ical Nomenclature, London. xx + 338 pp. [Publication of the Code was originally scheduled for 31 December 1984. In the preparation of this manuscript, the authors were kindly permitted by a member of the Editorial Committee to quote from the final draft of the Code in anticipation of its publication in 1984. The provisions of the Code are effective as of 1 January 1985 (Bull. Zool. Nomencl. 1984, 41: 196).]
International Commission for Zoological Nomenclature. 1964. Opinion 711. Culex aegypti Linnaeus, 1762 (Insecta, Diptera): validated and interpreted under the plenary powers. Bull. Zool. No- mencl. 21: 246-248.
Joblot, L. 1754. Observations d’histoire naturelle, faites avec le microscope, sur un grand nombre d’insectes, & sur les animalcules qui se trouvent dans les liqueurs préparées, & dans celles qui ne le sont pas, &c. avec la description & les usages des differens microscopes, &c. Vol. 1, Pt. 2. Paris. 124 pp. + 15 pls.
Jupp, P.G. 1979. Culex (Culex) pipiens and Culex (Cx.) torrentium (Diptera: Culicidae) in England: Notes on their taxonomy and biology. Mosq. Syst. 11: 121-126.
Knight, K. L. 1972. History of mosquito systematics Part I. Eighteenth century. Mosq. Syst. 4: 10-15.
Knight, K. L. and A. Stone. 1977. A catalog of the mosquitoes of the world (Diptera: Culicidae). 2nd edition. Thomas Say Found. 6: 1-611.
Latreille, P. A. 1809. Genera crustaceorum et insectorum secundum ordinem naturalem in familias disposita, iconibus exemplisque plurimis explicata. Vol. 4. Parisiis et Argentorati. 399 pp., 4 pls.
—. 1810. Considérations générales sur l’ordre naturel des animaux composant les classes des Crustacés, des Arachnides et des Insectes avec un tableau méthodique de leurs genres disposés en familles. Paris. 444 pp.
{indroth, C. H. 1957. The Linnaean species of carabid beetles. Linn. Soc. J. Zool. 43: 325-341.
Linnaeus, C. 1746. Fauna Svecica sistens Animalia Sveciae Regni: Qvadrupedia, Aves, Amphibia, Pisces, Insecta, Vermes, distributa per classes & ordines, genera, & species. Cum differentiis
tN w
VOLUME 87, NUMBER 1
specierum, synonymis autorum, nominibus incolarum, locis habitationum, descriptionibus
insectorum. Stockholmiae. xxiv + 411 pp., 2 pls.
1758. Systema naturae per regna tria naturae, secundum classes, ordines, genera, species, cum characteribus, differentiis, synonymis, locis. Ed. 10, Vol. 1. Holmiae. 824 pp.
—. 1761. Fauna Svecica sistens Animalia Sveciae Regni: Mammalia, Aves, Amphibia, Pisces, Insecta, Vermes. Distributa per classes & ordines, genera & species, cum differentiis specierum, synonymis autorum, nominibus incolarum, locis natalium, descriptionibus insectorum. Editio Altero. Stockholmiae. xlv + 578 pp., 2 pls.
—. 1767. Systema naturae per regna tria naturae, secundum classes, ordines, genera, species, cum characteribus, differentiis, synonymis, locis. Ed. 12 (ref.), Vol. 1, Pt. 2. Holmiae. pp. 533- 1327.
Martini, E. 1922. Zur Nomenclatur der fiir Mitteleuropa wichtigsten Stechmiicken. Entomol. Mitt. 11: 106-126.
—. 1925. Zwei bemerkenswerte Culiciden von einem eigenartigen Biotop. Int. Rev. Hydrobiol. Hydrogr. 12: 333-337.
Mattingly, P. F. 1951. Culex (Culex) torrentium Martini, a mosquito new to Britain. Nature (Lond.) 168: 172.
Mattingly, P. F., A. Stone, and K. L. Knight. 1962. Culex aegypti Linnaeus, 1762 (Insecta, Diptera); proposed validation and interpretation under the plenary powers of the species so named. Bull. Zool. Nomencl. 19: 208-219, 1 pl.
Meigen, J. W. 1804. Klassification und Beschreibung der europaischen zweifliiglichen Insekten (Dip- tera Linn.). Vol. 1, Pt. 1. Braunschweig. xxviii + 152 pp., 8 pls.
———. 1818. Systematische Beschreibung der bekannten europaischen zweifliigligen Insekten. Vol. 1. Aachen. xxxvi + 333 pp., 11 pls.
Natvig, L. R. 1948. Contributions to the knowledge of the Danish and Fennoscandian mosquitoes— Culicini. Nor. Entomol. Tidsskr. (Suppl. I). xxi + 567 pp., 12 pls., 1 map.
Olivier, G. M. 1791. Encyclopedie methodique. Histoire naturelle. Insects. Vol. 6. Paris. 704 pp.
Onyeka, J. O. A. 1982. The taxonomic value of pre-alar scales in the identification of Culex (Culex) pipiens L. and Culex (Cx.) torrentium Martini (Diptera: Culicidae). Mosq. Syst. 14: 41-52.
Réaumur, R. A. F. de. 1738. Mémoires pour servir a l’histoire des insectes. Vol. 4. Paris. xxxvi + 636 pp., 44 pls.
Robinson, G. S. and E. S. Nielsen. 1983. The microlepidoptera described by Linnaeus and Clerck. Syst. Entomol. 8: 191-242.
Rossi, P. 1790. Fauna Etrusca sistens Insecta quae in provinciis Florentina et Pisana praesertim collegit. Pt. 1. Liburni. xxii + 348 pp. + 6 unnumbered pages, 10 pls.
Schaeffer, J.C. 1766. Elementa Entomologica. Einleitung in die Insektenkenntniss; 135 ausgemahlte Kupfertafeln. Regensburg. 1x + 135 pp., 135 pls.
Schrank, F. von P. 1781. Enumeratio insectorum Austriae indigenorum. Augustae Vindelicorum.
xxii + 548 pp. + 4 unnumbered pages, 4 pls.
1803. Fauna Boica. Vol. 3, Pt. 1. Landshut. 272 pp.
Scopoli, J. A. 1763. Entomologia Carniolica exhibens insecta Carnioliae indigena et distributa in ordines, genera, species, varietates methodo Linnaeana. Vindobonae. xxxvi + 420 pp. Service, M. W. 1968. The taxonomy and biology of two sympatric sibling species of Culex, C. pipiens
and C. torrentium (Diptera, Culicidae). J. Zool. (Lond.) 156: 313-323.
Sicart, M. 1954. Présence de Culex torrentium dans les Pyrénées et comparaisons avec Culex pipiens du méme gite. Bull. Soc. Hist. Nat. Toulouse 89: 228-230.
Sirivanakarn, S. and G. B. White. 1978. Neotype designation of Culex quinquefasciatus Say (Diptera: Culicidae). Proc. Entomol. Soc. Wash. 80: 360-372.
Stephens, J. F. 1829. A systematic catalogue of British insects: being an attempt to arrange all the hitherto discovered indigenous insects in accordance with their natural affinities. Part I. Balwin and Cradock, London. 388 pp.
Stone, A., K. L. Knight, and H. Starcke. 1959. A synoptic catalog of the mosquitoes of the world (Diptera, Culicidae). Thomas Say Found. 6: 1-358.
Sulzer, J. H. 1761. Die Kenntzeichen der Insekten nach Anleitung des Konigl. Schwed. Ritters u. Leibarztes Carl Linnaeus durch 24 Kupferfeln erlautert und mit derselben natiirlichen Ges- chichte erlautert; mit einer Vorrede des Herrn Johannes Geesners. Ziirich. xxviii + 67 pp.
Sundararaman, S. 1949. Biometrical studies on intergradation in the genitalia of certain populations of Culex pipiens and Culex quinquefasciatus in the United States. Am. J. Hyg. 50: 307-314.
24 PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON
Swammerdam, J. J. 1737. Bybel der natuur, door Jan Swammerdam Amsteldammer. Of histoire der Insecten etc. Joannis Swammerdammii Amstelaed. Biblia naturae; sive Historia Insectorum, in classes certas redacta, nec non exemplis, et anatomico variorum animaliculorum examine, aeneisque tabulis illustrata. Insectis numerosis rariorum naturae observationibus. Omnia Lingua Batava, Auctori vernacula, conscripta. Accedit Praefatio, in quo vitam Auctoris descripsit Hermannus Boerhaave Med. Prof. Latinam versionem adscripsit Hieron. Vol. 1. Leydae. 550 pp. + 56 unnumbered pages.
Vane-Wright, R. I. 1975. The butterflies named by J. F. Gmelin (Lepidoptera: Rhopalocera). Bull. Br. Mus. (Nat. Hist.) Entomol. 32: 19-64, 6 pls.
White, G. B. 1978. Systematic reappraisal of the Anopheles maculipennis complex. Mosq. Syst. 10: 13-44.
NOTICE TO CONTRIBUTORS
Page charges for publication in the Proceedings have remained unchanged for several years although production costs have continued to rise. The approximate cost of production is now $55 per page. By action of the Executive Committee, page charges have been increased effective with the publication of this issue, Vol. 87, No. 1, but will not affect those manuscripts received by the Editor before this date. Cost to members for regular papers will now be assessed at $35 per page; for immediate publication and for pages exceeding fifteen, $55 per page.
PROC. ENTOMOL. SOC. WASH. 87(1), 1985, pp. 25-33
A EUROPEAN PRIVET SAWFLY, MACROPH YA PUNCTUMALBUM (L.): NORTH AMERICAN DISTRIBUTION, HOST PLANTS, SEASONAL HISTORY AND DESCRIPTIONS OF THE IMMATURE STAGES (HYMENOPTERA: TENTHREDINIDAE)
E. RICHARD HOEBEKE AND WARREN T. JOHNSON
Department of Entomology, Cornell University, Ithaca, New York 14853.
Abstract.—Macrophya punctumalbum (L.), a widespread European sawfly known in North America previously from Ontario, Quebec and British Columbia, is recorded from New York; this is a new United States record. Known host plants are members of the Oleaceae, with privet (Ligustrum spp.) appearing to be the primary food plant. Seasonal history and habits were studied at Ithaca, New York, during 1982-83. Overwintering occurs in the last larval (or prepupal) stage. Adults begin to appear in mid- to late May. Eggs, laid under the upper epidermis of leaves, begin to hatch by early to mid-June, larvae mature by early to mid-July, and drop to the ground by late August to construct earthen cocoons. Adult feeding produces irregular “‘rasping”’ marks and rectangular holes on the upper leaf sur- faces, and larvae chew circular holes in the leaves. A diagnosis of the adult is given, and the egg and last instar larva are described and illustrated.
Macrophya punctumalbum (L.), a European privet sawfly, occurs widely in all of Europe (including the British Isles) to the Caucasus (Benson, 1952; Schwenke, 1982). It is the only species of the genus Macrophya Dahlbom known from the Nearctic and Palearctic Regions. Presumably introduced into Canada from Europe some time in the early 1900s or before, it is now well established in eastern Canada (numerous records given by Gibson, 1980). Earliest records are from Toronto, Ontario, in 1932, and Vancouver, British Columbia, in 1934 (Gibson, 1980).
Our attention first focused on this introduced sawfly as early as the spring of 1979, when one of us (WTJ) noticed unfamiliar feeding damage to an ornamental planting of California privet (Ligustrum ovalifolium Hassk.) on the Cornell Uni- versity campus (Ithaca, New York). At that time, no adult or larval insect could be associated with the damage. It was not until late May 1983, when we collected adult females of M. punctumalbum, that we were able to implicate this sawfly as the pest species involved.
In this paper we review the North American distribution of this introduced species and give new United States records based on our own collecting in New York. We also summarize our observations on the biology, habits, and seasonal history at Ithaca, New York; provide recognition features for the adults; and describe and illustrate the egg and last instar larva.
NorTH AMERICAN DISTRIBUTION
In addition to published records from Ontario, Quebec and British Columbia (Gibson, 1980: 133), the following new records for the United States can be given.
26 PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON
New York: Tompkins Co., Ithaca (Cornell University campus, Red Barn), 24 May 1982, early June 1982, and additional collections in May and early June 1983. Niagara Co., Niagara Falls, 12 June 1983 (this record is based solely on adult feeding damage to California privet (L. ovalifolium) in a hedge of a private residence). A thorough examination of undetermined sawflies in the Cornell Uni- versity Insect Collection produced a single female specimen collected at Ithaca, NY, in late May 1975.
Host PLANTS AND DAMAGE
Known host plants of M. punctumalbum are members of the olive family, the Oleaceae. Fraxinus and Ligustrum appear to be the primary hosts across its native European range (Enslin, 1913; Korolkov, 1913; Schwenke, 1982). Fraxinus ex- celsior L. and Ligustrum vulgare L. are recorded as hosts in the British Isles (Cameron, 1882; Benson, 1952). Pomerantzev (1930) observed the species causing “considerable damage to ash [F. excelsior] in European Russia.” Korolkov (1913) noted that larvae of this sawfly severely damaged ash trees in the Alexander Garden (Moscow, Russia), “the leaves being skeletonized.”
In North America (eastern Canada), Gibson (1980) recorded “Fraxinus penns.” [= F. pennsylvanica Marsh.] and “*F. americana?” as hosts, citing ecological data from labels affixed to specimens examined during the course of his studies of North American Macrophya. Since 1940, M. punctumalbum has been found in- festing privet hedges in the Toronto area (Ontario) (Anon., 1949; Brown, 1942; Foott, 1973). In addition, it has caused “serious injury to privet hedges” in the St. Catharines area (Niagara Peninsula, Ontario) (Anon., 1959), and in Montreal, Quebec (Anon., 1965).
Gibson (1980) noted that M. punctumalbum has been reared from Syringa (lilac) in Canada; Syringa villosa Vahl. was cited as an associated host plant at Etobicoke Twp. (Toronto metropolitan area). At our study site, at least two species of Syringa (S. vulgaris L. and chinensis Willd.) occur within 15 meters of the privet hedge where our observations were made; foliage of only the common lilac (S. vulgaris) showed some feeding damage by M/. punctumalbum.
On California privet, adult feeding damage consisted of irregular “rasping” marks on the upper epidermis of leaf surfaces (Fig. 1). Adult sawflies also produce shiny, black fecal material which is deposited as irregular tarlike spots (up to 2 mm in diameter) on the upper leaf surfaces and new shoots (Fig. 1). These fecal spots wash or drop off the leaves by late summer.
Early instar larvae excavate circular holes (I—3 mm in diameter) in the interior of the leaves (Fig. 2). Larval feeding occurs primarily, if not exclusively, on the lower surfaces, except in the final feeding stage. The young larvae move into a curled (head to tail) posture to feed. Mature, full-grown larvae (Fig. 3), which reach a maximum length of about 18 mm, are capable of consuming the foliage, leaving only the midrib and lateral veins intact.
SEASONAL HISTORY AND HABITS
The biology and habits of 4. punctuma/bum in Europe given by Pomerantzev (1930), Korolkov (1913), and Mrkva (1965) are similar to what we report here, based on our observations at Ithaca, New York, during 1982-83.
Adults (predominantly females; males are generally scarce) first appeared on
VOLUME 87, NUMBER 1 27
Figs.1-2. Privet foliage damaged by European privet sawfly, Macrophya punctumalbum. 1, Privet
foliage damaged by adult sawfly feeding, characterized by the irregular “rasping” marks and rectangular
holes, and the tarlike spots. 2, Feeding damage by adults and larvae; the small round holes in the
interior of the leaves and the irregular leaf margins are caused by larvae.
privet foliage and began feeding by mid- to late May of both seasons, producing characteristic “rasping’’ marks and rectangular holes on the leaf surfaces. By June 6 (in 1982), females had oviposited under the upper epidermis of leaves, with eggs placed singly or in chains of 2—7. The oviposition site is marked by a small “blister” (per single egg), usually at the apex or periphery of the leaf (Fig. 5A, B). The embedded eggs are also visible from the under surface of the leaf. Also on June 6, early instar larvae were present on the foliage and beginning to feed. Females apparently oviposit over an extended period as first instar larvae can be found in the field until late June. Adult females become scarce by mid- to late June, and disappear shortly thereafter. By early July, nearly mature larvae were found on the foliage, together with various intermediate stage larvae. By early August, fewer larvae can be found; mature larvae have dropped from the foliage and presumably go into the soil to construct earthen cocoons. However, some mature and intermediate stage larvae have been observed on the foliage as late as August 24 (in 1982). The last stage larva (or prepupa) probably overwinters. A single generation is produced annually on privet in New York. We collected or
28 PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON
Figs. 3-4. Life stages of the European privet sawfly, Macrophya punctumalbum. 3, Mature (feeding stage) larvae; note feeding damage and fecal pellets (length of mature larvae ranges from 14.4-18 mm). 4, Adult female on privet leaf.
VOLUME 87, NUMBER 1 As)
observed very few adult males; the species is mainly parthenogenetic, with males poorly represented in most populations (Benson, 1952; Novak, 1976).
ECONOMIC IMPORTANCE
Macrophya punctumalbum is expected to be of economic significance in North America. If adults and larvae occur in large numbers, its primary food plant (privet) can be defoliated. From our field and laboratory observations, feeding and oviposition occurs primarily on foliage of the tender, terminal shoots, and often on foliage of the shaded portions of the plantings. Recommended control measures, especially when the host plant is used in an ornamental hedge, may be simplified by hedge trimming in late spring before the larvae can mature.
LABORATORY STUDIES
In 1982 we attempted to rear VM. punctumalbum under caged conditions in a room with natural lighting and no temperature control. Temperatures during these studies ranged from 26°C (in the summer) to 11°C (in the winter). Six adult females were caged on May 19, 1982, with two potted, nursery-grown plants of Ligustrum ovalifolium. The cage (30.5 x 30.5 x 91.5 cm) was covered on 3 sides and the top with a fine mesh ninon fabric; the front, a sliding door, was made of clear Plexiglas, approx. 3 mm thick.
The adult sawflies immediately began feeding. As foliage was depleted, twigs from uninfested L. ovalifolium plants were stuck into the pots as a food supple- ment. By May 27 the potted plants were nearly defoliated and were removed from the cage. The supplemental twigs were placed in a container of water-saturated sand and all remaining foliage from the potted plants was removed and attached to the supplemental food plants. Two days later 2 new potted plants, Ligustrum vicaryl Rehd., also nursery grown, were placed in the cage. Within two days all adult sawflies were dead, presumably killed by a residual insecticide on the foliage of these new plants. Eggs had been laid in leaves of new plants and also were apparent in the supplemental foliage, as well as foliage taken from the original potted plants. On June 2 the progeny were found—14 days after the adults were first introduced into the cage. The larvae developed slowly. By mid-July the second pair of potted plants had been defoliated by larval feeding. From this time all food was from supplemental sources. By August 24, there were 35 larvae of various stages of development on the foliage. By September 10, larvae were abundant but there was little movement or feeding. By December, only 3 live, but quiescent, larvae were observed on the foliage. Observations were not made again until March 1983; no live larvae were found at that time. The soil and duff were removed from the cage and filtered through 3 sieves (screen sizes 8, 10 and 18). Twenty- two dead, shriveled larvae were found and no pupae. Thus, the rearing conditions were not suitable for pupation and the colony was lost.
DESCRIPTION
Adult recognition features.—Females of M. punctumalbum are unlikely to be confused with those of any other North American species of Macrophya. The bright, rufous hind femora, contrasting with the predominantly black body, dis- tinguishes this attractive species (Fig. 4). All other North American Macrophya
30 PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON
Fig. 5.
illustrating the “‘blisterlike” oviposition wound. B, scanning electron photomicrograph of the same (30x).
Oviposition site of the adult female of Macrophya punctumalbum. A, schematic drawing
have a black or black and white hindfemur. The body and appendages of the female are black with the following areas white or white-yellow: small spots along ihe occipital carina of the head; broad area along the posterior margin of the pronotum; large spot over most of the scutellum, and the scutellar appendage;
VOLUME 87, NUMBER 1 31
Fig. 6. Mature (feeding stage) larva of Macrophya punctumalbum. A, Head and thorax. B, Third abdominal segment. C, Eighth, ninth and tenth abdominal segments. D, Maxilla, dorsal. E, Right (r) and left (1) mandibles, ventral, F, Epipharynx. G, Integument of abdominal annulet, scanning electron photomicrograph (700 ~ ).
anteroapical surfaces of the femora, and anterior surfaces of the tibiae and tarsi of the fore and middle legs; basolateral spot on the hindcoxa and elongate subapical spot on the hindtibia; and lateral tergal spots and ninth tergite dorsally of the abdomen. The male, which is usually smaller than the female, is almost entirely black. There are small white spots on the dorsoapical margin of the pronotum adjacent to the scutellum, and the fore and middle legs are similar in coloration to those of the female.
Description of egg, Fig. SA, B.—Length 1.36-1.44 mm (x = 1.40 mm; n = 3); maximum width 0.68-0.88 mm (X = 0.76 mm; n = 3). Elongate-ovoid, somewhat depressed along the long axis. Chorion minutely sculptured at extremely high magnification (>2000 x), otherwise nearly smooth at lower ranges of magnifi- cation (250-500 x ).
Description of larva, Fig. 6A-G.—Known larvae of Nearctic Macrophya are variously patterned with spots or bands on the head and body (Gibson, 1980). Mature larvae of M. punctumalbum are entirely lime green, except for the yel- lowish head capsule and a black eye spot. Lorenz and Kraus (1957) gave a brief description of the larva of M. punctumalbum and included it in a key to larvae of European Macrophya. Gibson also provided a short diagnosis of the larva.
In late instar (feeding stage), head capsule pale yellow-brown with black eye spot; thoracic legs pale green; body lime green above, somewhat paler below. Integument as in Fig. 6G. Length 14.5-18.0 mm (X = 16.6 mm; n = 11).
32 PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON
Clypeus usually with 2 setae on each side (individiuals sometimes with 3 setae on one side and 2 on the other; see remarks below). Labrum with 3 setae on each side, apical margin emarginate at middle; epipharynx with 8 clavate spines located in arcuate row on each anterolateral half (Fig. 6F). Each mandible with | seta on outer lateral surface; left mandible with 2 ventral teeth, 3 lateral teeth, and 1 molar tooth (Fig. 6E); right mandible with 3 lateral teeth and 3 molar teeth (Fig. 6E). Maxillary palpus four-segmented; second segment of palpus with | seta on outer surface at apex; palpifer with 3 setae; stipes with | seta; galea large, digitlike; lacinia with 8-9 clavate spines (Fig. 6D). Labial palpus three-segmented; 3 setae on each side of prementum at apex; second labial palpal segment with | seta on inner surface at apex.
Thorax without spines, tubercles or glandubae. Thoracic legs normal; femur longer than tibia; setae present on all surfaces of each segment. Prothoracic spi- racles not winged.
Abdominal segments | through 8 each with 7 dorsal annulets (typical segment shown in Fig. 6B); without spines, tubercles or glandubae. Eighth, ninth and tenth segments as in Fig. 6C; also without spines, tubercles or glandubae. Fine setae numerous on suranal and subanal areas of tenth segment. All abdominal spiracles not winged.
Remarks.—The subgenus Pseudomacrophya, originally proposed by Enslin (1913) for the single species /. punctumalbum, differs from the nominate sub- genus by the degree of convergence of the inner orbits of the eyes of the adults. Lorenz and Kraus (1957) stated that the larva of M4. punctumalbum has 3 setae on the clypeus (on each side). They suggested, on the basis of this character and the one seta on each mandible, that Psewdomacrophya might be considered a valid genus. Of 22 mature larvae examined in our studies (by ERH), 18 had 2 setae on each side of the clypeus; however, 4 larvae had 3 setae on one side and 2 on the other. Also, in the New York populations, larvae lacked tubercles or glandubae on annulets 2 and 4 of the typical abdominal segment (Abd. 3) that Lorenz and Kraus (1957) reported for European material.
ACKNOWLEDGMENTS
We thank Howard H. Lyon (Cornell University) for providing us with the photographs of Figs. 1-4, David R. Smith (Systematic Entomology Laboratory, IIBII, USDA, Washington, D.C.) for confirming the identification of the sawfly and for reviewing the manuscript, Deborah Martens for the line drawing of Fig. 5A, and A. G. Wheeler, Jr. (Penn. Dept. Agric., Harrisburg, PA) for a critical review of an earlier draft of the manuscript.
LITERATURE CITED
Anonymous. 1949. Insects affecting greenhouse and ornamental plants. Can. Agric. Insect Pest Rev. 27: 153, 177-178.
—. 1959. Insects on ornamentals. Can. Agric. Insect Pest Rev. 37: 246-247.
1965. Pests of ornamentals. Can. Agric. Insect Pest Rev. 43: 118.
Benson, R. B. 1952. Hymenoptera (Symphyta). Family Tenthredinidae, pp. 51-137. Jn Royal En- tomological Society of London, Handbooks for the identification of British insects, v. 6, pt. 2(b). London.
Brown, A. W. A. 1942. Summary report from the forest insect survey, 1942. Can. Agric. Insect Pest Rev. 20: 265-276.
VOLUME 87, NUMBER | 33
Cameron, P. 1882. A monograph of the British phytophagous Hymenoptera. Vol. I. 340 pp. + 21 pls. London.
Enslin, E. 1913. Die Tenthredinoidea Mitteleuropas. Deutsch. Ent. Z., Beiheft, 1913, pp. 99-202.
Foott, W. H. 1973. Garden pests in and around Toronto. Can. Agric. Insect Pest Rev. 51: 35-37.
Gibson, G. A. P. 1980. A revision of the genus Macrophya Dahlbom (Hymenoptera: Symphyta, Tenthredinidae) of North America. Mem. Entomol. Soc. Can., No. 114: 116 pp.
Korolkov, D. M. 1913. [Insects injurious to gardens: materials for the study of the injurious insects of the Government of Moscow during the year 1912]. Moscow: Zemstvo of the Government of Moscow; 1912-1913: pp. 1-25.
Lorenz, H. and M. Kraus. 1957. Die Larvalsystematik der Blattwespen (Tenthredinoidea und Me- galodontoidea). Berlin, 339 pp.
Mrkva,R. 1965. [Contribution on the morphology, bionomics and parasites of Tomostethus nigritis.]. Prace Vyzkumn. Cst. Lesn. CSSR, n. 30, pp. 33-64. [In Czech. with Russian and German summaries. ]
Novak, V. 1976. Atlas of insects harmful to forest trees, vol. 1, pp. 7-123 (incl. 115 pls.). Elsevier Scientific Publishing Company, Amsterdam, The Netherlands.
Pomerantzev, D. V. 1930. On the biology of the sawflies of the ash-tree. Bull. N. Cauc. Pl. Prot. Sta., Rostov, Don. 6-7: 27-32. [In Russian with German summary.]
Schwenke, W. 1982. Die Forstschadlinge Europas. Band 4, Hautfliigler und Zweifliigler. Paul Parey, Hamburg and Berlin, 392 pp.
PROC. ENTOMOL. SOC. WASH. 87(1), 1985, pp. 34-43
A NEW GENUS AND TWO NEW SPECIES OF ANTILLOCORINI FROM THE NEOTROPICS, WITH NOTES ON RELATED TAXA (HEMIPTERA: LYGAEIDAE)!
JAMES A. SLATER
Section of Systematic and Evolutionary Biology, University of Connecticut, Storrs, Connecticut 06268.
Abstract.—Schuhocoris gracilis is described as a new genus and species of An- tillocorini from Peru, Brazil and Panama. Its phylogenetic position is discussed. The dorsal aspect of the adult and anatomical details of the abdomen, paramere, and sperm reservoir are illustrated. Paradema antennata is described as a new species from Brazil. There is a discussion of variability in Paradema oculata Slater. New records include Bathydema cubana from Puerto Rico and Caeneusia obrienorum from Brazil.
The tribe Antillocorini in the Neotropics has been the subject of two recent papers (Slater, Sweet, and Baranowski, 1977; Slater, 1980). Recently I have had the opportunity to study additional material including a striking new genus with a somewhat ozophorine habitus and an undescribed species of Paradema. Also included is additional information on several previously described species. All measurements are in millimeters.
Schuhocoris NEW GENUS
Dorsal surface chiefly pruinose. Head and abdomen subshining, remainder of body dull. Head granulose not conspicuously punctate. Pronotum, scutellum and hemelytra coarsely punctate; clavus with three distinct rows of punctures. Head non-declivent, vertex strongly convex. Pronotum with lateral margins deeply sinuate, much broader across rounded humeri than across anterior lobe; margins bluntly calloused or ‘“‘subcarinate’’; transverse impression complete; calli little differentiated from remainder of pronotal surface. Posterior margin of pronotum straight before base of scutellum, produced laterad of scutellum posteriorly as rounded lobes. Scutellum with a prominent Y-shaped elevation. Apical corial margin shallowly concave near inner end. Hemelytral membrane hyaline. Gular trough elongate, nearly reaching base of head, terminating in a tapering acute point. Metathoracic scent gland auricle strongly curved posteriorly so that distal end projects caudally. Evaporative area occupying only inner ’2 of metapleuron, convexly rounded on outer margin. Abdomen with well developed scent gland scars present between terga 3-4, 4-5, and 5-6. Inner laterotergites present on segments 4, 5, and 6 (Fig. 2). All spiracles ventral, those of sterna 3, 4, and 5
This work was supported in part by a grant from the National Science Foundation.
VOLUME 87, NUMBER 1 35.
located on sternum below sternal shelf (Fig. 6). Suture between sterna 3 and 4 reaching marginal shelf, that between sterna 4 and 5 obsolete dorsally. Tricho- bothria of sternum 5 located as follows: anterior trichobothrium placed near 3-— 4 suture; two posterior trichobothria located close together dorso-ventrally and well forward of spiracle of abdominal sternum 5 (Fig. 6). Antennae very elongate and slender, terete; first segment slightly “bent,” exceeding apex of tylus by more than ' its length. Legs elongate, slender. Fore femora not incrassate, mutic. Ovipositor short, not dividing 6th sternum. Posterior margin of 6th abdominal sternum with a small truncate median posteriorly produced projection. Sperm reservoir of ““generalized”’ type, not reduced; ejaculatory tube not greatly enlarged (Figs. 4, 5).
Type species: Schuhocoris gracilis new species.
It is a pleasure to dedicate this interesting genus to Randall T. Schuh (American Museum Natural History) for his important contributions to the systematics of the Hemiptera.
The systematic position of Schuhocoris within the Antillocorini and its per- plexing “mix” of characters is an indication of how much remains to be understood of the phylogeny of these interesting insects. Schuhocoris will not run to any genus in my (Slater, 1980) key to Western Hemisphere genera. The apical corial margin is only slightly concave which might cause it to be assigned to the first half of couplet | where it would run to Paradema at couplet 2. It is not at all related to this genus which has rounded lateral pronotal margins, an only slightly posteriorly curved metathoracic scent gland auricle, spiracle four located on the sternal shelf, etc. More importantly, as discussed below, Schuhocoris has a conventional sperm reservoir and thus lacks the enormously enlarged ejaculatory reservoir of Para- dema (see Slater, 1980), If one follows the second half of couplet 1, Schuhocoris will key to Antillocoris at couplet 10 (except for the position of spiracle four). It differs from Antillocoris in many ways: general habitus (body elongate and slender in Schuhocoris, short and stout in Antillocoris); spiracle of abdominal sternum 4 located below the sternal shelf in Schuhocoris, on the shelf in Antillocoris. Antil- locoris lacks a well developed scent gland opening between terga 3-4, in Schu- hocoris it is strongly developed; in Antillocoris the posterior trichobothria are located below spiracle 5, in Schuhocoris well anterior to it. In Schuhocoris the metapleuron is pruinose, in Antillocoris it is shining.
The systematic position of Schuhocoris within the Antillocorini is very inter- esting but also perplexing. My (Slater, 1980) cladogram of Western Hemisphere antillocorine genera is inadequate to accommodate this genus. In that cladogram the development of the incomplete suture between sterna 4 and 5 is hypothesized as having been evolved twice. This does not present a problem here as in one case (Paradema) the suture loss is in a clade with taxa having highly apomorphic sperm reservoirs and ejaculatory ducts. In Schuhocoris the reservoir and duct clearly show the plesiomorphic condition (Figs. 4, 5). Schuhocoris thus clearly belongs to the alternate clade for while it does not have elongate body hairs and the apical corial margin is only shallowly concave, the lateral pronotal margins are conspicuously calloused, the spiracle on abdominal tergum 4 lies below the sternal shelf and the abdominal scent gland openings between terga 3—4 are well developed.
My 1980 cladogram has three factual mistakes. (1.) There is no synapomorphy
36
Fig. 1.
PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON
Schuhocoris gracilis, dorsal view.
VOLUME 87, NUMBER 1 oH)
Figs. 2-6. Schuhocoris gracilis. 2, Abdomen, dorsal view. 3, Paramere, 4, Sperm reservoir, lateral view. 5, Sperm reservoir, dorsal view. 6, Abdomen, lateral view.
in the cladogram linking Paurocoris, Botocudo and Cligenes. This was a lapsus and the synapomorphy should be “‘pronotal margins carinate.’” However, there are a number of species at present placed in Botocudo where the lateral pronotal margins are actually calloused rather than sharply carinate. (2.) Some species now placed in Botocudo have the posterior trichobothria of sternum five placed dorso- ventrally rather than being on the same level (linear). It is obvious that a study of species now considered to belong to Botocudo must at least have reached the
38 PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON
stage where the basic “‘groups” can be established before the phylogeny of the Antillocorini can be much better understood. (3.) The statement that the deeply grooved prosternum is a synapomorphy that will distinguish Cligenes from Bo- tocudo is true for the type species Cligenes distinctus but is not true of Cligenes subcavicola Scudder, Darlington, and Hill.
Schuhocoris shows derived conditions in the placement of the abdominal spi- racles, strongly posteriorly curved scent gland auricle, calloused pronotal margins and contrastingly shining head and pruinose body. The well-developed 3-4 ab- dominal scent gland auricle precludes it at present from the clade with Antillocoris and its allies. However, that clade is based on a loss character. I suggest that until the complexity of Botocudo is better understood that Schuhocoris should be ten- tatively considered as the plesiomorphic sister group of the Antillocoris et al. clade, which has retained the original 3 abdominal scent gland condition and has an only slightly concave apical corial margin.
Schuhocoris gracilis NEW SPECIES Figs. 1-6
General coloration nearly uniformly bright reddish brown including antennal segments 1, 2, and 3, becoming slightly darker on distal 2 of corium. Fourth antennal segment completely white. Legs uniformly pale yellow. Body clothed above with scattered, short, but rather conspicuous, erect or suberect hairs.
Head acuminate anteriorly. Eyes large, occupying most of lateral head surface but set away from antero-lateral pronotal margin. Ocelli conspicuous, elevated, set much closer to compound eye than to one another. Head length 0.60, width 0.60, interocular space 0.30. Pronotum with well defined punctate anterior ‘‘col- lar’ area; posterior lobe more elevated than anterior but broadly depressed in middle. Pronotal length 0.64, width across anterior lobe 0.66, width across humeri 1.10. Scutellar length 0.54, width 0.56. Hemelytra with lateral corial margins nearly straight, very shallowly concave at level of apex of scutellum; radial vein strongly elevated. Length of claval commissure 0.26. Midline distance apex clavus- apex corium 0.60. Midline distance apex corium-apex membrane 0.76. Abdomen with areas about bases of trichobothria pale, pruinose and conspicuously differ- entiated from subshining abdominal surface. Labium extending between meso- coxae, first segment remote from base of head. Labial segment lengths I 0.40, I] 0.38, III 0.30, IV 0.26. Antennal segment length I 0.64, Il 0.72, II 0.60, IV 0.64. Total body length 3.48.
Bulb of sperm reservoir elongate-elliptical. Reservoir wings sloping moderately caudally, elongate tapering (Fig. 5). Basal area of ejaculatory duct with a mght angled curvature with a short thick basally directed projection at angle (Fig. 4). Paramere broad, without an inner projection, but produced along outer margin (Fig. 3).
Holotype: é PERU: Junin: San Ramon de Pangoa, 40 km SE Satipo, 750 meters, (soil litter layer in high secondary forest), 25.III.1972 (R. T. & J. C. Schuh). In American Museum Natural History. Paratypes: | ¢, 1 2, same data as holotype. 5 6, 7 2, same 23.III.1972. 1 4, same 30.1.1974. PANAMA: 1 9, Bocas d. T. Corriente Grande, 100 meters, 9°17'30’N, 82°32'11”W, 4.1V.1980 (Henk Wolda). BRAZIL: 1 4, Sao Paulo, Jacupiranga XII.1963 (F. Plaumann). | 4, Rio de Janeiro, Silva Jardin, VIUI.1974 (F. M. Oliveira). In American Museum of Natural His-
VOLUME 87, NUMBER 1 39
tory, Instituto de Biologia UNAM Mexico D.F., P. D. Ashlock and J. A. Slater collections.
Three specimens show antennal oligomery. In one specimen (from Brazil) the second and third segments of the right antenna appear to be fused to form an elongate segment, with segments one and “‘four” normal. Two Peruvian specimens have the left antenna identically modified. In these specimens the left antenna although three segmented is nearly as long as the mght antenna, but the second segment is longer than the normal second. The third (terminal segment) is mostly white but has a dark brown basal portion which suggests that there has been fusion of the normal third and fourth segments.
This species probably is part of the large fauna of Lygaeidae that feed on the fallen seeds of Ficus. R. T. Schuh took an adult and nymphs representing two instars at the type locality on 30.1.1974 below a large fig tree in litter that contained an abundance of fig seeds.
DESCRIPTION OF NYMPHS
Third Instar Nymph (type locality).— Head, legs and first three antennal seg- ments dull yellow, fourth antennal segment white. Head suffused with brown posteriorly behind eyes. Thorax above and below nearly uniformly brown. Ab- domen mottled with red. First abdominal tergum with a narrow transverse well differentiated strap-like sclerite across central area. Dorsal abdominal scent gland openings present and of equal width between terga 3-4, 4-5, and 5-6. None of these openings surrounded by extensive darkened sclerotized plates. Sterna also lacking median sclerotized plates. Sternum 7 with a conspicuous median spine at posterior margin (as in adults).
Head convex across vertex; epicranial stem well developed. Tylus extending only to proximal 2 of first antennal segment. Head length 0.54, width 0.42, interocular space 0.28. Pronotum subquadrate. Pronotal length 0.28, width 0.50. Mesothoracic wing pads very slightly produced over antero-lateral margins of metanotum. Wing pad length 0.20. Abdomen much broader than head or thorax, maximum width (across segment 5) 0.88. Labium slightly exceeding metacoxae, first segment remote from base of head. Labial segment lengths I 0.30, II 0.38, III 0.24, IV 0.24. Antennal segment lengths I 0.44, II 0.44, III 0.40, IV 0.48. Total body length 2.30.
Fifth Instar Nymph (from type locality).—General coloration and structure as above but with head, thorax and mesothoracic wing pads nearly uniformly dull yellow. Reddish coloration of abdomen obscure. Pronotum more strongly tapering from humeral angles to anterior margin with lateral margins acute and slightly concave. Mesothoracic wing pads almost attaining suture between abdominal terga three and four. Labium extending posteriorly only between mesocoxae. Head length 0.68, width 0.60, interocular space 0.40. Pronotal length 0.46, width 0.80. Mesothoracic wing pad length 1.00. Abdominal length 1.56. Labial segment lengths I 0.38, If 0.50, III 0.26, IV 0.26. Antennal segment lengths I 0.70, II 0.70, III 0.58, IV 0.66. Total body length 3.24.
Paradema antennata NEW SPECIES
Head and anterior pronotal lobe black, former shining, latter covered with gray pruinosity except for two large quadrate calli patches. Pronotal pruinosity con-
40 PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON
Table 1. Comparison of interocular distance and antennal segment lengths in females of Paradema oculata Slater and Paradema antennata new species.
Interocular
Distance Length of Antennal Segments Il Wl IV N Mean Range Mean Range Mean Range Mean Range antennata 3. 53 (.50-.54) 82 (.78-.84) 53 (.50-.56) 85 (.82-.88) oculata Rio Calceone 8 .48 (.46-.48) 70 (.66-.74) 47 (.46-.50) 58 (.56-.58) Serra Lombard 7 .49 (.48-.52) 70 (.68-.76) 48 (.46-.54) 57 (.54-.66) Serra do Navio 5 .50 (.50) .77 (.70-.80) 52. (.48-.54) 60 (.60)
tinued as a posteriorly tapering ‘V’ on meson of anterior 73 of posterior pronotal lobe; a broad dark brown area present on either side of this pruinosity and on humeral angles; remainder of posterior pronotal lobe yellow. Scutellum dark reddish-brown, elevated areas paler. Hemelytra variegated, dark red brown and yellow to almost white. Pale hemelytral markings as follows: an elongate streak on basal 2 of clavus just within claval suture and entire distal end of clavus; a stripe on basal 2 of corium adjacent to claval suture and a second interrupted stripe running along outer margin of corial furrow to level of end of claval com- missure; most of lateral explanate corial flange (but with dark patches at level of middle of scutellum, at level of end of claval commissure and at apex of corium); the pale area between dark apex and median dark area extending inward as a broad comma-shaped macula. Membrane with variegated patches of dark and light coloration. Antennae almost uniformly dark red brown. Femora darkened subdistally, with pale distal ends; fore femora darker than middle and hind femora. Remainder of legs pale yellow except for third tarsal segment which is dark red brown. Body surface as in ocu/ata with numerous elongate upstanding hairs pres- ent.
Body shape and proportions as in oculata. Labium extending between meta- coxae. Fore femora with a prominent ventral spine. Head length 0.66, width 0.90, interocular space 0.54. Pronotal length 1.02, width 1.46. Scutellar length 0.90, width 0.78. Midline distance apex clavus-apex corium 1.00. Midline distance apex corium-apex membrane 0.60. Labial segment lengths I 0.70, II 0.74, III 0.54, IV 0.36. Antennal segment lengths I 0.40, II 0.84, III 0.56, IV 0.88. Total body length 4.60.
Holotype: 2 BRAZIL: Marituba 24.X.1961 (J. & B. Bechyne). In American Museum Natural History. Paratypes: | 2 same data as holotype. 1 2? same 1.VI.1961. In P. D. Ashlock and J. A. Slater collections.
P. antennata is certainly closely related to P. oculata Slater, agreeing with the latter in having a fore femoral spine, similar protruding eyes, and in pronotal shape, color and overall habitus. P. antennata may eventaully prove to be a subspecific geographic isolate of ocul/ata. However, the extremely elongate fourth antennal segment of antennata falls so far out of the range of oculata that a distinct population is evident and I am according it specific status at this time (Tables | ind 2). The length of the fourth antennal segment of oculata as given by Slater (1980) has the numbers reversed. The length is 0.62 not 0.26 mm.
VOLUME 87, NUMBER 1 41
Table 2. Ratios of interocular distance to antennal segment III and of antennal segment III to antennal segment IV in females of Paradema antennata and Paradema oculata.
Ratio Antennal Segment IV Ratio Antennal Segment IV
Interocular Distance Antennal Segment III
N Mean Range Mean Range antennata 3 1.61 (1.59-1.64) 1.60 (1.57-1.64)
oculata
Rio Calceone 8 2 (1.17-1.26) 1.22 (1.16-1.26) Serra Lombard 7 1.16 (1.08-1.27) 1.20 (1.13-1.26) Serra do Navio 5 1.20 (1.20) Lats (1.07-1.25)
All three females of antennata have the lateral corial margins evenly meeting the membrane which, if its remains constant when a larger series is available, will separate antennata from those females of ocu/ata with an angulate corial-mem- brane junction (see below).
Paradema oculata Slater Paradema oculata Slater, 1980: 214-215.
This species was originally described from the holotype male from Colombia and a single paratype female from Guyana. It was readily distinguishable by being the only member of the genus to have a ventral spine on the fore femur. It was placed in a group with /ongisetosa Slater and bathydemoides Slater, characterized by the lateral corial margins of the females joining the membrane at a distinct angle.
P. D. Ashlock has recently made available for study a long series representing several Brazilian localities. To my surprise this material clearly indicates that the condition of the corial membrane junction is variable in females of this species. This suggests caution in the use of couplet 2 of my (Slater, 1980) key to species as the condition could prove to be variable in other species as well.
There is evidently geographic variation in the expression of this feature as may be seen in Table III. There do not seem to be any other significant differences between these populations, and one must conclude that we are dealing with a single species.
Bathydema cubana Slater & Baranowski Bathydema cubana Slater and Baranowski, 1977: 349, 351.
Slater, Sweet, and Baranowski (1977) recognized eight species in the genus Bathydema and subdivided the genus into three groups. Of these the ‘‘ohscura- group” has subsequently been elevated to generic status under the name Antil/- lodema by Slater (1980).
The “darlingtoni-group”’ comprised three species, one each on the Greater Antillean Islands of Jamaica (darlingtoni), Cuba (cubana) and Hispaniola (his- paniola).
Through the kindness of Drs. C. W. and L. B. O’Brien, I have now been able to examine the first specimens of Bathydema from Puerto Rico, a macropterous female from ““Guilarte For. Res. Hwy. 131 & 158, 23.VI.1979 (G. B. Marshall)”
42 PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON
Table 3. Comparison of corial margin conditions in Paradema oculata slater.
Number with = Number with
Angulate Non-angulate Conal Conal N Margins Margins Locality Males Females Females Females Rio Calceone lg. do Tigre Brazil 6/8.8 1961 (J. & B. Bechyne) 2 8 8 Serra Lombard, Limao, Brazil (J. & B. Bechyne) 2.8.1961 1 1 15.8.1961 | 1 | 20.8.1961 1 1 21.8.1961 1 1 27.8.1961 1 2 2 31.8.1961 1 1 19.8.1961 1 1.9.1966 2 Serra do Navio, Brazil 11.7.1961 (J. & B. Bechyne) 7 6 6
Regina, Brazil 14.8.1961 (J. & B. Bechyne) 1 1
Ozémar, Brazil (Diniz) 1
and a macropterous male and two females from “Carib. N.F. el ToroNegroD Hwy. 143, K16H4, 21.VII.1979 (L. B. O’Brien)” (in J. A. Slater collection). They will key to cubana in Slater, Sweet, and Baranowski (1977) although the head ratios differ slightly from those given for cubana. Measurements of the Guilarte For. Res. specimen are: head length 0.44, width 0.60, interocular space 0.34; ratio width head/interocular space 1.76, ratio length head/interocular space 1.29. All of the specimens resemble Cuban material in being of a predominently dark color. Bathydema cubana has a much shorter head than does hispaniola. While these specimens may represent a distinct species they are so similar to cubana that specific recognition does not seem warranted.
It is important to note that the ‘‘darlingtoni-group” (the plesiomorphic one) is now known from all of the four main islands of the Greater Antilles.
Caeneusia obrienorum Slater Caeneusia obrienorum Slater, 1980: 211-212.
This species was originally described from near Tingo Maria, Peru. I have recently examined a single typical female from Serra Lombard, Brazil (2. VIII.1961 J. & B. Bechyne).
ACKNOWLEDGMENTS
I am indebted to R. T. Schuh (American Museum of Natural History) for the collection and loan of the greater part of the type series of Schuhocoris and for the use of his field notes; to P. D. Ashlock (University of Kansas), Harry Brailovsky (University of Mexico), Dodge Engleman (Coco Solo, Panama) and C. W. and L. B. O’Brien (Florida State University) for the loan and/or gift of important material; and to Mary Jane Spring and Elizabeth Slater (University of Connecticut) for
VOLUME 87, NUMBER 1 43
preparation of the illustrations and aid in preparation of the manuscript, respec- tively.
LITERATURE CITED
Slater, J. A. 1980. Systematic relationships of the Antillocorini of the Western Hemisphere (He- miptera:Lygaeidae). Syst. Entomol. 5:2: 199-226.
Slater, J. A.. M. H. Sweet, and R. M. Baranowski. 1977. The systematics and biology of the genus Bathydema Uhler (Hemiptera:Lygaeidae). Ann. Entomol. Soc. Am. 70:3: 343-358.
PROC. ENTOMOL. SOC. WASH. 87(1), 1985, pp. 44-48
EUCARAZZIA ELEGANS (FERRARI), AN APHID NEW TO THE WESTERN HEMISPHERE, WITH ARCHIVAL DATA (HOMOPTERA: APHIDIDAE)
MANYA B. STOETZEL
Systematic Entomology Laboratory, IIBIII, Agricultural Research Service, USDA, RM 6, Bldg 004, BARC-West, USDA, Beltsville, Maryland 20705.
Abstract.—The collection and identification of alates of Eucarazzia elegans (Ferrari) from five counties in southern California constitute a new record for North America and the Western Hemisphere. In the literature, the aphid has been reported from the Mediterranean area, the Canary Islands, India on various hosts in the plant family Labiatae, the mints. While only viviparous females have been collected and while E. elegans has been collected on a wide variety of hosts in California, there is no indication that the true host or hosts are not a plant or plants in the Labiatae. Information is given on those hosts and locations reported in California and in the literature, the collection data of specimens in the U.S. National Collection of Insects, the history and synonymy of this genus and species, and characteristics useful in the identification of live and slide-mounted females.
This paper is provided to call attention to the first collection of Eucarazzia elegans (Ferrari, 1872) in North America and the Western Hemisphere. Infor- mation is given on those hosts and locations reported in California and in the literature, the collection data of specimens in the U.S. National Collection of Insects, the history and synonymy of this genus and species, and characteristics useful in the identification of live and slide-mounted females.
On April 6, 1984, I identified as Eucarazzia elegans (Ferrari) two alate aphids collected 3-30-84 on bell peppers (Capsicum annuum) [Solanaceae] in Mecca, Riverside County, California, and one alate aphid collected March 16, 1984 on Fuchsia sp. [Onagraceae] in Fallbrook, San Diego County, California. Kono and Reeves (1984) published the first report of this genus and species in North America and the Western Hemisphere. Alates of E. elegans have now been collected in California in the additional counties of Imperial, Orange, Santa Barbara, and San Bernardino. Since the first two submissions, alates have been collected from the following plants: red root pigweed (4maranthus retroflexus) [Amaranthaceae]; rose periwinkle (Catharanthus roseus), oleander (Nerium oleander) [Apocyna- ceae]; taro (Colocasia esculenta) [Araceae]; chrysanthemum (Chrysanthemum sp.), prickly lettuce (Lactuca serriola), sow thistle (Sonchus sp.), sunflower (Helianthus sp.), Mare’s-tail [reference slip lists Erigeron canadensis, not Hippuris vulgaris] [Compositae]; watermelon (Citru/lus lanatus) [Cucurbitaceae]; sweet corn (Zea mays) [Gramineae]; mint (Mentha sp.), catnip (Nepeta cataria), basil (Ocimum sp.), thyme (Thymus sp.), rosemary (Rosmarinus sp.), sage (Salvia sp.) [Labiatae]; unspecified bean plant, snail vine (Vigna caracalla) [Leguminosae]; mallow (Mal- va sp.) [Malvaceae]; eucalyptus (Eucalyptus sp.) [Myrtaceae]; privet (Ligustrum sp.) [Oleaceae]; fan palm (Livistona sp.) [Palmae]; Pittosporum tobira ‘Variegata’
VOLUME 87, NUMBER 1 45
[Pittosporaceae]; rose (Rosa sp.), strawberry (Fragaria sp.) [Rosaceae]; Sierra cur- rant (Ribes nevadense) [Saxifragaceae]; petunia (Petunia sp.) [Solanaceae]; lantana (Lantana sp.) [Verbenaceae]. Four alates were collected in water-pan traps, 7 alates were collected on yellow sticky boards, and | alate was “found in an office.”
Immatures (lI st-instar nymphs) have been collected only from Riverside Coun- ty, California and only from the following hosts: basil (Ocimum sp.) [Labiatae], Thermal, April 11, 1984, 10 immatures, 3 alates; mint (Mentha sp.) [Labiatae], Riverside, April 14, 1984, 1 immature, 2 alates; mallow (Malva sp.) [Malvaceae], Thermal, April 18, 1984, 2 immatures, 2 alates; Sierra currant (Ribes nevadense) [Saxifragaceae], Idyllwild, May 4, 1984, 1 immature, 5 alates. While FE. elegans has been collected on a wide variety of hosts in California, I believe that these hosts represent accidental alightings and that the true host or hosts will probably prove to be plants in the Labiatae. Alates have been collected many different times Over a wide area; only 1-5 alates have ever been collected at any one time on any one plant. Despite the fact that developing populations have not been found in California, it is assumed that this aphid will not be important econom- ically. Rearing experiments are now being conducted in California by the U.S. Department of Agriculture to establish whether E. elegans can reproduce on non- Labiatae hosts. Its proficiency as a virus vector is also to be tested. Stary et al. (1971) reported that Praon volucre Haliday is a parasite of E. elegans on Salvia sp. in France.
In the literature, only viviparous females of E. elegans have been described or reported and only on plants in Labiatae. Hille Ris Lambers (1953) stated that E. elegans had been collected on various Labiatae (Coleus sp., Lavandula sp, Mentha sp., Sa/via sp.) around the Mediterranean (Italy, Spain, Morocco, Asia Minor, Egypt). Specific collections have been reported as follows: CANARY ISLANDS on Mentha sp. (Gomez-Menor, 1963; Tambs-Lyche, 1971); FRANCE on Salvia sp. (Stary et al., 1971); INDIA on an unknown plant (Ghosh, 1974); ISRAEL on Salvia sp. (Bodenheimer and Swirski, 1957); ITALY (del Guercio, 1921); KENYA in suction trap (Eastop, 1957); SICILY (Barbagallo and Stroyan, 1978, 1982) on Clinopodium vulgare, Mentha pulegium, M. x sativa, Nepeta cataria, Salvia of- ficinalis, and S. verbenaca (Barbagallo and Stroyan, 1982); PORTUGAL on La- vandula latifolia and Salvia verbenaca (Iharco, 1979) and in Moericke trays (Lourenco and IIharco, 1982); SOUTH AFRICA in yellow tray (Miller and Scholl, 1958); SPAIN on Lavandula stoechas (Mier Durante and Nieto Nafria, 1979) and on Lavandula pedunculata (Mier Durante and Nieto Nafria, 1983); and TURKEY on Mentha piperita (Tuatay, 1972). The collection in the British Museum (Natural History), London, contains, in addition to several of the above, material from IRAQ (mint), IRAN (Salvia splendens), and ZIMBABWE (yellow trap) (V. F. Eastop, pers. comm.).
The National Collection of Insects contains the following slide-mounted ma- terial: (1) originally identified as Rhopalosiphoninus salviae Hall on Lavandula stoechas {Labiatae]—Rabat, Morocco, February 1936; Mamora, Morocco, July 31, 1936 and January 20, 1939; (2) intercepted in quarantine at Boston and originally identified as Rhopalosiphoninus chicotei Gomez-Menor—with Mentha sp. [Labiatae] from Portugal, November 19, 1963; with Mentha sp. from Azores, October 14, 1968; on unknown plant from Portugal, November 15, 1975; (3) identified as Eucarazzia elegans (Ferrari)—on Mentha sp. from Cyprus, December
46 PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON
2
Figs. 1-2. Eucarazzia elegans. 1, Alate female, in life, illustrating the waxy bands on the head, thorax, and abdomen and the black patch on the dorsum of the abdomen. 2, Slide-mounted alate female illustrating the dark spots on the wing margin at the end of the veins, the black patch on the dorsum of the abdomen, the short cauda, and the distinctively swollen cornicles.
17, 1967; (4) intercepted in quarantine at Boston and identified as Eucarazzia sp.—on Mentha sp. from Azores, November 14, 1969.
In 1872 Ferrari described e/egans and placed it in the genus Rhopalosiphum. In 1921 del Guercio described as new the species picta from aphids collected in
VOLUME 87, NUMBER | 47
1919 from leaves of Nepitella sp. in Sicily. Del Guercio stated that picta had generic characteristics similar to those of Rhopalosiphum; however, because picta lacked antennal tubercles and a well-defined cauda, del Guercio described the new genus Eucarazzia and included Rhopalosiphum calthae Koch (1854) and R. na- jadum Koch (1854) with his picta. Eastop and Hille Ris Lambers (1976) placed calthae Koch in Rhopalosiphoninus, treated najadum Koch as a synonym of Rhopalosiphum numphaeae (L.), and listed Eucarazzia picta, Anuraphis (Clavi- siphon) elegans del Guercio (1930), Rhopalosiphoninus chicotei Gomez-Menor (1950), and Rhopalosiphoninus salviae Hall (1926) as synonyms of Eucarazzia elegans (Ferrari). A full discussion of the history and synonymy of this genus and species is given in Hille Ris Lambers (1953). Besides EF. elegans, only E. caucasicus (Aizenberg, 1956), described from alates collected on leaves of Pterocarya frax- inifolia in Abkhazia, ASSR, USSR, is now in the genus Eucarazzia (Eastop and Hille Ris Lambers, 1976).
In life an alate female of EF. elegans (Fig. 1) is striking in appearance with a silvery, waxy covering on its head and thorax and in bands on those abdominal segments not bearing a large, black patch. Each vein of the forewing ends in a dark spot on the wing margin, and the anal vein has a fuscous border for its entire length. This wing coloration is distinctive and it, along with the silvery and black coloration, readily identifies an alate in the field. An apterous female also is covered with silvery wax, but it does not have a large, black, dorsal patch on its abdomen.
A slide-mounted alate female of E. elegans (Fig. 2) is readily identified by the ornamented wings, the black, dorsal abdominal patch, antennal segment II with 16-27 large, tuberculous sensoria, and antennal segment IV with 3-9 sensoria. An apterous female does not have a black, dorsal abdominal patch; but it does have 2-20 tuberculous sensoria irregularly distributed on the distal portion of antennal segment III and 2—9 on antennal segment IV. Both alate and apterous females have distinctive cornicles that are strongly swollen on the distal half and that have 3-4 rows of reticulations apically. The cornicles are dark except for the proximal 4th in the alata and are pale except for the swollen area which is brownish in the aptera. Both forms have a pale cauda that is hardly longer than wide, is acute, and has 5-7 setae. A full discussion of the characteristics of the alate and apterous females of E. elegans is given in del Guercio (1921) and Hille Ris Lambers (1953).
At first glance the distinctively swollen cornicles of E. elegans suggest that this species is in the genus Rhopalosiphoninus, but closer examination reveals that the two genera share almost no other morphological characters. In his description of E. caucasicus, Aizenberg (1956) noted Hille Ris Lambers’ (1953) treatment of E. elegans and concluded that **... there are differences in the marking, in the configuration of the wing, etc.’’ According to Aizenberg, his caucasicus has flocky waxen powder only on the abdomen and, in addition to the triangular spots on the margin at the end of all of the veins, a light-brown band running the length of the cubitus in the forewing.
ACKNOWLEDGMENTS
I thank Tokuwo Kono, California Department of Food and Agriculture, Sac- ramento, for his willingness to provide collection data, many samples of the aphids
48 PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON
themselves which have been deposited in the U.S. National Collection of Insects, and a review of this article. For permitting me to use the photograph in Fig. 1, I thank George N. Oldfield, Boyden Fruit & Vegetable Entomological Laboratory, ARS, USDA, Riverside, CA, and Max Badgley, University of California, Riv- erside. For their critical review of this manuscript I thank the following members of the Systematic Entomology Laboratory, ARS, USDA: E. Eric Grissell and Douglas R. Ferguson, Research Entomologists, and Louise M. Russell, Resident Cooperating Scientist.
LITERATURE CITED
Aizenberg, E. E. 1956. New data on the systematics of aphids (Aphidoidea, Homoptera). Trudy Vses. Entomol. Obshchest. 45: 155.
Barbagallo, S. and H. L. G. Stroyan. 1978. Rilievi preliminari sulla composizione dell’afidofauna Siciliana. Proc. XI Cong. Naz. Ital. Entomol. (Portici-Sorrento) p. 249.
—. 1982. Osservazioni biologiche, ecologiche e tassinomiche sull’afidofauna della Sicilia. Frus- tula Entomol. 3: 139.
Bodenheimer, F. S. and E. Swirski. 1957. The Aphidoidea of the Middle East. Weizmann Science Press of Israel, Jerusalem, p. 277.
Eastop, V. F. 1957. The periodicity of aphid flight in East Africa. Bull. Entomol. Res. 48: 306.
Eastop, V. F. and D. Hille Ris Lambers. 1976. Survey of the World’s Aphids. Dr. W. Junk b.v., Publishers, The Hague, p. 193.
Ferrari, P, M. 1872. Species Aphididarum Hucusque in Liguria lectos. Ann. Mus. Civ. Stor. Nat. Genova 3: 217.
Ghosh, A. K. 1974. A list of aphids (Homoptera: Aphididae) from India and adjacent countries. J. Bombay Nat. Hist. Soc. 71: 210.
Gomez-Menor, J. 1950. Algunas especies nuevas de afidos (Homoptera, Aphidae) Rev. Espan.
Entomol. 1: 110-113.
1963. ‘“‘Aphidoidea” de las Islas Canarias. II. Anu. Estud. Atlanticos 9: 519-605. Guercio, G. del. 1921. Specie nuove e nuovi generi per l’afidofauna Italica. Redia 14: 129-136. 1930. Osservazioni intorno al Gen. Anuraphis del Guercio. Redia 19: 194.
Hall, W. J. 1926. Notes on the Aphididae of Egypt. Min. Agric. Egypt, Tech. Sci. Serv. Bull. 68: 42-44.
Hille Ris Lambers, D. 1953. Contributions to a monograph of the Aphididae of Europe. V. Tem- minckia 9: 31-35.
Ilharco, F. A. 1979. 1. Aditamento ao catalogo dos afideos de Portugal continental. Agron. Lusit. 39(4): 265.
Koch, C. L. 1854. Die Pflanzenlause Aphiden, getreu nach dem Leben abgebildet und beschrieben. Durk von Fr. Campe & Sohn, Niirnberg. Heft II, p. 45-49.
Kono, T. and E. Reeves. 1984. Eucarrazia elegans (Ferrari), an aphid new to North America (Homoptera: Aphididae). Calif. Plant Pest Disease Rpt. 3: 53-54.
Lourengo, A. and F. A. Ilharco. 1982. Andalise das capturas de afideos efectuadas por armadilhas de Moericke num campo de favas en Oeiras - Outubro 1978/Maio 1979. Agron. Lusit. 41(3- 4): Chart 1.
Mier Durante, M. P. and J. M. Nieto Nafria. 1979. Nuevos datos afidologicos para la provincia de Salamanca (Hom. Aphidoidea). Bol. Asoc. Esp. Entomol. 3: 159.
—. 1983. Aportaciones a la afidofauna de Galicia, If (Hom. Aphidoidea). Bol. Asoc. Entomol. 6: 329.
Miiller, F. P. and S. E. Schéll. 1958. Some notes on the aphid fauna of South Africa. J. Entomol. Soc. South. Africa 21: 398.
Stary, P., G. Remaudiére, and F. Leclant. 1971. Les Aphidiidae (Hym.) de France et leurs hdétes (Hom. Aphididae). Entomophaga 5: 35.
Tambs-Lyche, H. 1971. Aphids from the Canary Islands. Entomol. Scand. 2: 128.
Tuatay, N. 1972. Aphidoidea: Aphididae. /n Muzesi’s Bocek Katalogu (1961-1971). Turkiye Cum- huryeti Tarim Bakanligi Zurai Mucadele ve Zirai Karantina Genel Mudurlugu Yayinlari, Mes- leki Kitaplar Serisi, Ankara, p. 20.
PROC. ENTOMOL. SOC. WASH. 87(1), 1985, pp. 49-79
STUDIES OF CECIDOMYIID MIDGES (DIPTERA: CECIDOMYIIDAE) AS COCOA POLLINATORS (THEOBROMA CACAO L.) IN CENTRAL AMERICA
ALLEN M. YOUNG
Invertebrate Zoology Section, Milwaukee Public Museum, Milwaukee, Wis- consin 53233.
Abstract.—The role of cecidomyiids (Diptera: Cecidomyiidae) as floral visitors and pollinating agents of “cocoa,” Theobroma cacao L. (Sterculiaceae), was ex- amined in Costa Rica and, to a lesser degree, in Belize. Emphasis was placed on: (1) describing the mechanism of pollination, (2) determining the most abundant pollinating species, (3) estimates of adult population densities of midges in cocoa flowers, (4) confirmation of pollinating ability with the use of experimental field cages, (5) describing diurnal activity patterns of cecidomytids in cocoa farms and in relation to similar patterns of anthesis in cocoa flowers, (6) description of communal roosting in spider webs by some species in cocoa farms, (7) observations on how midges interact with specific floral parts in cocoa, (8) estimates of seasonal changes in abundance of midges in cocoa flowers and in relation to annual flow- ering patterns of cocoa, and (9) determination of breeding microhabitats of cec- idomyiids in cocoa farm habitats. By far the most abundant species in cocoa flowers at two localities in Costa Rica (Finca La Tigra and Finca Experimental La Lola) were Clinodiplosis sp. 1 and Mycodiplosis sp. 1, both new to science. Adults possess functional elongate mouthparts adapted for sucking liquids such as floral nectar, unusually long abdominal and leg hairs that trap cocoa pollen grains, very large compound eyes, and dusk-night-dawn activity patterns that match well the diurnal cycle of flower opening and pollination receptivity in cocoa. Female cecidomyliids are far more abundant as cocoa flower visitors than males, and both crawl inside enclosed floral parts associated with reproductive behavior in cocoa. The mechanism of pollination is similar to that described for cerato- pogonids, but one major difference is the exploitation of the vivid yellow petal ligules as feeding sites by cecidomyiids, while these structures are largely ignored by ceratopogonids. Cecidomyiids feed at stomate-type nectaries lining petal ligules and the inner surfaces of petal hoods. Although as many as ten genera and 14 species of cecidomyiids occur in cocoa farms in Central America, only a few of these behave as consistent floral visitors to cocoa and related species of Theo- broma. Cocoa flower-visiting forms are free-living with larval stages thriving in fungus-infected rotting organic debris, most of which is arboreal and includes: animal-generated holes in cocoa pods, rotting diseased cocoa pods and wilted cherelles, and epiphytic mosses. The observations of cecidomyiid densities being somewhat elevated in flowers during the drier months of the year may be due to either concentration of populations in shaded areas of cocoa or an actual increase in abundance at these times. Some cecidomyiids breed in the dry exocarp of rotting pods in trees during the dry season. Communal roosts in spider webs are
50 PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON
re-used over several successive days as daytime resting sites for midges, probably as a predator (primarily anoles, mantids, and jumping spiders)-escape adaptation. The ‘“‘pollinator reward syndrome” in cocoa and other Theobroma species is discussed from the standpoint of cecidomyiids, ceratopogonids, and wild bees as natural pollinating agents.
Several studies in different cocoa-growing regions (Theobroma cacao L.) of the world have concluded that tiny biting midges (Diptera: Ceratopogonidae) are the primary effective pollinators, although not exclusively so (e.g. Jones, 1912; Wel- lensiek, 1932; Cope, 1939; Billes, 1941; Posnette, 1942, 1944; Soetardi, 1950; Gnanaratnam, 1954; Van Der Knapp, 1955; Saunders, 1959; Walker, 1959; Des- sart, 1961; Glendenning, 1962; Gorrez, 1962; Hernandez, 1965; Sampayan, 1966; Vello and Magalhaes, 1971; Edwards, 1973; De La Cruz and Soria, 1973; Kauf- mann, 1973a, b, 1974, 1975a, b, c; Amponsah, 1972; Winder, 1972; 1977a, b, 1978a, b; Winder and Silva, 1975; Wirth and Waugh, 1976; Bystrak and Wirth, 1978; Young, 1982, 1983). Relative to the many genera and species of cerato- pogonids suspected or proven to be effective pollinators of cocoa, several workers have noticed high abundances of similarly-sized cecidomyiid midges (Diptera: Cecidomyiidae) in cocoa farms, both in the New and Old World tropics (e.g. Privat, 1979; Winder, 1977c; Kaufmann, 1973b; Soria et al., 1980, 1981; Lucas, 1981), although the role of these midges in cocoa pollination is thought to be minimal (Winder, 1977b, 1978a). Yet Kaufmann (1937b), using experimental cages, demonstrated effective but low frequency pollinating activity of one species of cecidomyiid midge during the dry season in Africa, and subsequent studies using isotopes demonstrate that these midges transport cocoa pollen grains on their bodies in high frequency in some African farms (Decazy et al., 1980). Herein I present data indicating that several genera and species of Cecidomyiidae are effective cocoa pollinators in Central America, most notably in Costa Rica, and also summarize new ecological and behavioral information as related to the role of these midges as pollinators, and to their adaptation for breeding in cocoa farm habitats. I tentatively conclude that both ceratopogonids and cecidomylids con- stitute one constellation of cocoa pollinators, but that other major pollinators (bees) have played equally important roles in the evolution of cocoa as a tropical rain forest tree species. Some studies have emphasized the role of small Diptera as effective, opportunistic pollinators of various plant species in shaded, moist forest habitats (e.g. Mesler et al., 1980; Levesque and Burger, 1982), environments that tend to accumulate high species density of Diptera (e.g. Vanhara, 1981), including cocoa farms (Bigger, 1981). Although it is generally maintained that Diptera are occasional pollinators (Percival, 1965; Kevan and Baker, 1983), sev- eral studies indicate more consistent associations with some plant species (Bau- mann, 1978; Vogel, 1978; Mesler et al., 1980; Gilbert, 1981; Steiner, 1983; Phil- brick, 1983; Young, 1984a). With respect to Theobroma species (Sterculiaceae) overall, it may well be that Diptera are opportunistic “minor” pollinators that thrive best in moist, shaded habitats in the tropics, while “‘major”’ pollinators, ones coevolved with the floral structure and its associated physiology, are pri- marily wild bees (A. M. Young, in preparation) associated with Theobroma pop-
VOLUME 87, NUMBER 1 51
ulations in more open, exposed habitats. Small flying insects such as tropical Cecidomyliidae are most active in shaded moist habitats (e.g. Mamaev, 1975; Fisher and Teetes, 1982; Barnes, 1930; Summers, 1975; Petralia et al., 1979; Shazli and Mostafa, 1980; Mogal et al., 1979; Kaufmann, 1973b) where they often exhibit marked diurnal cycles of activity (e.g. Shazli and Mostafa, 1980; Summers, 1975; Brewer, 1981; Brown and McGavin, 1982) and immature stages of free- living forms associated with fungi, molds, and rotting organic substrates such as leaf litter (e.g. Winder, 1977a; Gagné, 1977; Parnell, 1969). Because cecidomyiids and other allied dipterans often exhibit marked cycles in adult activity and breed- ing, both in the temperate zones and the tropics (e.g. Fontanilla-Barroga, 1962; Charlwood et al., 1982; Chiang, 1968; Kay, 1983), determination of potential roles in pollination of specific plant species such as cocoa has to be related to the known diurnal cycles of flower opening and anthesis (Wellensiek (1932) for cocoa). Because even decaying organic substrates can be limiting factors in the breeding structure of dipteran populations (e.g. Carpenter, 1983; Binns, 1980; Fallis and Snow, 1983; Siefert, 1980; Siefert and Barrera, 1981; Kaufmann, 1973b; Privat, 1979; Winder, 1977a), it is also necessary to determine whether or not cecido- mylids as cocoa pollinators breed within the cocoa farm environment, and this was done in the present studies. Although very few published reports indicate a possible role of cecidomyiids as pollinators (e.g. Soderstrom and Calder6én, 1971), Winder (1977b) found cecidomyiids bearing cocoa pollen grains and also entering the petal hoods of cocoa flowers where they came into contact with viable pollen. Winder (1977b) suggests that the high abundance of these midges in cocoa farms indicates a possible role in cocoa pollination. Others, however, suggest that these midges are of negligible importance in cocoa pollination (e.g. Entwistle, 1972).
LOCALITIES AND STUDY METHODS
The following studies were conducted between 1978 and 1983 primarily at two localities on the Atlantic watershed of Costa Rica: (1) mechanism of pollination of cocoa by cecidomyiid midges and determination of most effective pollinating species; (2) breeding habits and life cycles of some species pollinating cocoa; (3) estimates of adult population densities in cocoa flowers at various times of the year; (4) experimental tests with cages to determine pollinating abilities of ceci- domyliids; (5) estimation of overall cecidomyiid faunas associated with cocoa farms; (6) communal roosting habits of adult cecidomyiids in cocoa farms and relation to pollinating activity; (7) diurnal activity patterns of adult cecidomyiids in relation to cocoa tree flowering patterns and availability of possible pollination “rewards” in flowers. The greater portion of these studies were conducted at “‘Finca Experimental La Lola, near Siquirres (10°06'N, 83°30’W), Limon Province (see also Young (1983) for further desceiption of this site), and fewer studies at “Finca La Tigra,” near La Virgen (10°23’N, 84°07'W), Heredia Province (see Young (1982) for additional information on this locality). In addition to these Costa Rican studies, I made preliminary surveys of adult cecidomyiids on cocoa flowers, and the larval microhabitats of some species, at the Hummingbird Hershey Cocoa Farm (approx. 17°08’N, 88°38'’W), near Belmopan, Belize. Visits were made to all three localities during both wetter (rainy) and drier (dry) times (seasons) for one to three years. Because rainfall pattern has a marked influence on both the cycles of flowering in cocoa (e.g. Wellensiek, 1932; Alvim, 1977; Hutcheon, 1981;
52 PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON
asof LA TIGRA (TIRIMBINA)
LA LOLA
RAINFALL IN MM
JFMAMJJASONDJFMAMJJASONDJFMAMJJSJASONDJFMAMJJASONDJFMAMJJASONDJSJFMAMJJA 1978 1979 1980 1981 1982 1983
SUCCESSIVE MONTHS AND YEARS
Fig. 1. Monthly patterns of rainfall, over several years and including the study periods, for “‘Finca La Tigra” and ‘Finca Experimental La Lola” in the Atlantic (Caribbean) lowlands of Costa Rica.
Myers, 1930; Allen, 1981; Young, 1982, 1983, 1984b) as well as on the population dynamics and associated densities of pollinating insects (e.g. Winder and Silva, 1975; de la Cruz and Soria, 1973; Leston, 1969; Soria and Abreu, 1976; Soria et al., 1981; Kaufmann, 1973b; Young, 1982, 1983, 1984a), monthly patterns of rainfall for ‘“‘La Lola” and “‘La Tigra” are summarized here (Fig. 1).
The rainfall data (Fig. 1) encompass the periods of midge study at both localities, and it is apparent in the data from both localities that the driest times of the year are generally the months January through March, although the duration of reduced rainfall periods varies greatly from year to year. For the purposes of this paper, I assume that both localities have a short and irregular “dry season” (“‘veranillo”’) generally falling between January and March each year. Other studies (Young, 1982, 1983, 1984b) have shown that the dry season is a period of drastically reduced flowering in cocoa at these localities.
The study areas at La Lola include heavily-shaded cocoa, with the predominant shade tree being Erythrina sp. (Leguminosae), although other areas at the same locality have little shade. The La Lola studies included surveys of cecidomyiids in both the shaded (Area A) and unshaded (Area B) UF-29 cocoa studied in Young (1983) as well as a well-shaded clonal garden area with several clonal varieties (UF-677, UF-613, Pound-7, etc.). The La Tigra study area consisted of many
n ww
VOLUME 87, NUMBER 1
varieties of cocoa planted beneath a thinned-out natural canopy of advanced secondary forest. At La Tigra there is a high incidence of rotting cherelles on cocoa trees while at La Lola these are removed quickly as they appear. In the La Tigra study area, many trees have sizable accumulations of epiphytic mosses and other epiphytes on the lower branches.
During both wetter and drier periods between 1978 and 1983, observations were made at both localities on the behavior of adult cecidomyiid midges at freshly-opened cocoa flowers at various hours of the day and night. Daylight transects through cocoa trees were taken by walking slowly and examining open flowers on each tree to determine the presence of cecidomyiids on or inside flowers. When a midge was encountered, it was scored as either “resting’”’ motionless on an external portion of the flower, or as moving inside the flower. In the latter case, notes were kept on where the midge was located in the flower. Time of day was also recorded. These surveys generated a series of voucher specimens for those cecidomyiids actually found in contact with cocoa flowers. Other voucher series of cecidomyiids were obtained by collecting adult midges from communal roosts in spider webs in cocoa trees, and occasionally by rearing midges from larval samples.
Daylight observations at open cocoa flowers generally extended at intermittent intervals from 0600 to 1800 hours in both sunny and rainy weather, and nighttime samples from 1900 to 2300 and from 0400 to 0530 hours. Nighttime observations were aided by using red cellophane over a flashlight and occasionally by using a Night Vision infra-red viewing scope outfitted with a 50mm macrolens. Whenever midges were found moving inside flowers, detailed observations were made on this behavior. Additionally, behavior suggesting that midges were feeding on specific floral parts was followed up as much as possible. For example, early into the studies, I discovered that cecidomyiids (eventually determined to be Myco- diplosis sp. 1 and Clinodiplosis sp. 1) had their mouthparts embedded into the pliable, spongy tissue of the blade-like ligules extending from the petal hoods of cocoa flowers (see Cuatrecasas, 1964 for a detailed discussion of gross floral structure in cocoa), and when encountered in the field, observations were extended until the feeding stopped. Through brief observations at various times, I was able to determine how cecidomyliids interact with cocoa flowers and how this behavior relates to pollinating activity.
Opportunistically, immature stages of cecidomyiids breeding in cocoa farms were obtained by surveying rotting cherelles, arboreal and ground-cover leaf litter beneath cocoa trees, and by collecting repeated samples of epiphytic moss. The emphasis here was on obtaining a qualitative picture of breeding substrates rather than a quantitative sampling of population densities of immature stages. Collected breeding substrate samples were usually confined to tightly-shut large clear plastic bags to allow larvae to complete development and adults to eclose. Voucher series of larvae and adults were kept for determination studies. For all cecidomyiid voucher samples obtained, the insects were preserved in the field in vials of 70% ethanol and later shipped to the Insect Identification and Beneficial Insects In- troduction Institute, ARS, USDA at Beltsville, Maryland for identification.
To estimate the densities of adult cecidomyiids in open cocoa flowers, I made repeated censuses of 70 mature trees, consisting of a mix of clonal varieties UF- 677, UF-613, and Pound-7, in one of the clonal gardens at La Lola; the same set
54 PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON
of trees was censused at various times of the year and data were also kept on the abundance of open flowers at each census for each tree. These censuses were usually taken at the same time of day each time (generally between 0900-1100 hours) and in both cloudy (overcast) and sunny weather. I attempted always to examine all open flowers within easy eye-viewing at each tree.
As these studies progressed, with the assistance of Raymond J. Gagné of the Systematic Entomology Laboratory, IIBIII, in Washington, D.C., it was possible to narrow down the number of genera and species of these midges most consis- tently found in cocoa flowers, and this allowed me to conduct some preliminary field tests with small experimental cages to determine the potential for these midges to pollinate cocoa. Such tests can only give a possible indication of pollinating activity, however, and the results may have little or nothing to do with whether or not these same species of midges actually pollinate cocoa under natural con- ditions. The bias and pitfalls of this sort associated with cage tests involving small flying insects as potential pollinators has been pointed out by Winder (1978a) in the case of cocoa. In the present field studies, small sleeves of fine-mesh silk bolting cloth (Tetkco Co., Elmsford, New York) were stretched over wire frames embedded into sections of branches on cocoa trees bearing lots of floral buds and open flowers. Details of this methodology are outlined in Young (1983). These tests were conducted at La Lola using the self-compatible UF-29 variety of cocoa, a condition that permitted confined insects such as cecidomyiids to fertilize con- fined flowers successfully without the need for cross-pollination with other vari- eties. A total of ten cage tests were conducted, five each at two different times during two successive rainy seasons (1981 and 1982); of the five cages used each time, three were “‘controls’” having no midges introduced into them and the remaining two, experimentals. Experimental cages generally received anywhere from five to 30 cecidomyiids over a 1-3 day period, and the numbers of open flowers scored each day in all cages. Prior to initiation of these tests, all open flowers were removed from all cages and subsequently opening flowers were shield- ed by the bolting cloth from any possible external pollinating agents other than the confined cecidomyiids. Midges were usually introduced at different times of the day; sources of cecidomyiids included collections from open flowers and to a lesser extent from communal roosts. It was not always possible to determine the species of cecidomyiid used, although emphasis was placed on testing Mycodiplosis sp. | and Clinodiplosis sp. |.
At two different dates at La Lola, I tagged the positions of several communal roost sites for adult cecidomyiids in a heavily-shaded area (the Area A of Young, 1983) of UF-29 cocoa trees to determine (1) the day-to-day changes in roosting habits and degree to which the same site is re-used over short periods of time (i.e. a succession of days), and (2) the diurnal patterns of roost formation and dissolution. I was also interested in determining which species of cecidomyiids participated in roosting behavior at La Lola, and which of these species were also visiting cocoa flowers and possibly exhibiting pollinating behavior. Several pre- vious studies in the Neotropical Region have described the unusual phenomenon of cecidomyiids roosting communally in the silken webs of various species of spiders (Fig. 2), including both natural forest and cocoa farm habitats (e.g., Lah- mann and Zuniga, 1981; Kaufmann, 1973b). As sometimes individual roosts on a given date contain more than one species of cecidomyiid (Dr. W. E. Eberhard,
VOLUME 87, NUMBER | 55
Fig. 2. Communal roosting aggregate of unidentified Cecidomyiidae in the Hummingbird Hershey Farm in Belize.
per. comm.), it was necessary to collect samples of midges from roosts at various dates to estimate the degree to which roosts in this particular case were mono- specific in composition. Tagged roost sites were checked almost hourly and the number of midges in them counted and recorded. Particular emphasis was given to observing roost formation and dissolution as related to dawn and dusk. Ob- serving roosts in this manner provided the opportunity to relate diurnal activity patterns of cecidomyiids to the diurnal cycle of flower opening and anthesis in cocoa. Only at the end of observation periods were samples taken of adults in roosts for determination purposes.
Although to be summarized in detail elsewhere (A. M. Young, in preparation),
56 PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON
an experiment of setting out sticky squares of flypaper around open cocoa flowers at both La Lola and La Tigra and checking them for trapped flying insects at various times of the day and night, provided some additional information on diurnal activity patterns of cecidomyiids in the immediate vicinity of cocoa flow- ers. Relevant portions of these data are mentioned in this paper. The experiments involved placing several hundred such small squares (anchored on No. 3 insect pins) each time within a few mm of open cocoa flowers on 1-5 trees at each locality. Squares were then scored as to whether they were “occupied” by insects, what kinds, and time of the day.
Cecidomyiids surveyed at Hummingbird Hershey in Belize included the col- lections of larvae from rotting cherelles and rearing adults from these. Information was obtained on the locations of cecidomyiid larvae in these substrates. Also, collections of animal-damaged rotting pods hanging in trees were made to deter- mine the presence of cecidomyiid larvae (at La Lola and Hummingbird Hershey) and to rear these to adulthood.
A small sample of adult cecidomyiids was collected from cocoa flowers and preserved dry to examine with the Scanning Electron Microscope for the presence of cocoa pollen (Taylor, 1965); it is easy to match grains from flowers with those collected from insects. Laboratory analysis by R. J. Gagné also included an ex- amination of the mouthparts of those species most consistently found in cocoa flowers. This study was done to determine if these particular species had functional mouthparts and if so, to describe these and possibly associate with the observed behavior of the midges in the flowers. While some cecidomyiids possess vestigial mouthparts as adults (e.g. Ehler, 1982), long-lived free-living forms have func- tional mouthparts (R. J. Gagné, pers. comm.).
RESULTS
Pollination Mechanism and Behavior.—Cecidomyiid midges, mostly Myco- diplosis sp. 1, Clinodiplosis sp. 1, Coquillettomyia sp., and Aphodiplosis trian- gularis, have been found between 0600 and 0700 hours, and again between 1500 and 2100 hours, exhibiting two forms of behavior associated with moving inside enclosed parts of cocoa flowers: (1) individual midges crawl along a staminode, parallel to, and adjacent to, the pistil, and oriented towards the basal ovary area of the flower, and (2) individual midges alight on a sepal, petal ligule, or exteri- or surface of a petal hood, and proceed to crawl inside the petal, often moving from one petal hood to the next and so on. For example, on 9 December 1982 at 0700 hours at La Lola, one female of Clinodiplosis sp. 1 landed at the tip of a staminode and crawled along it and rubbed against the pistil; examination of the pistil area after collecting the midge revealed a slight smear of several pollen grains. Sometimes a female remains motionless on petal ligules for several minutes to an hour beore moving. On 21 November 1981 at La Lola, one specimen (Mycodiplosis sp. 1) was discovered with mouthparts embedded into the distal surface tissue of a ligule at 1100 hours and was still there at 1500 hours. A male of the same species was seen crawling along a staminode at 0900 hours and with mouthparts clearly scraping against the staminode tissue, as if feeding. Another female Clinodiplosis sp. 1 alighted on the tip of a staminode at 0600 hours on 19 November 1981 at La Lola, stayed there for one minute, and then quickly crawled down the staminode to the ovary area of the flower; this cecidomyiid remained
VOLUME 87, NUMBER 1 57
Table 1. Adult cecidomyiid midge behavior in open flowers of Theobroma cacao as related to effective pollination.
Descnption of Behavior Species Sex Date Time Locality Related to Pollination
Clinodiplosis female 9-XII-82 0700 La Lola lands on tip of staminode and sp. | quickly crawls towards ovary,
brushing against pistil
female 9-XII-82 1630 La Lola crawling in narrow space between staminodes and pistil
female 11-XII-82 0730 La Lola lands on floral petiole, crawls over to staminode and crawls along it and enters into inside of petal hood, exits few seconds later and enters a second petal hood
female 12-ITI-82 2000 La Lola captured while inside petal hood Mycodiplosis female _11-XII-82 0730 La Lola crawling along ligule towards petal sp. | hood opening female 13-XI-81 1600 La Lola crawling around inside petal hood female 13-XI-81 1730 La Lola perching on ligule and suddenly
“flips” around and enters petal hood and moves around violently inside it and causes petal hood to momentarily change shape
female 12-III-82 1700 La Lola captured inside petal hood
female 21-XI-81 0600 La Lola crawling into petal hood
female 21-XI-81 0630 La Lola lands on outside of petal hood and crawls inside it
female 9-XII-82 1630 La Lola captured inside petal hood
female 9-XII-82 1630 La Lola captured inside petal hood
female 9-XII-82 1645 La Lola captured in narrow space between
staminodes and pistil
on the ovary tissue for about four hours before flying off. While adult cecidomyiids are readily found resting on petal ligules, or upon dense mats of fungal growth (hyphae) covering squirrel or woodpecker-generated holes in cocoa pods (Fig. 3), of particular interest are those observations of these midges actually moving within confined spaces of the flower (Table 1). Midges hovering about cocoa flowers are sometimes captured as prey by the smaller nymphal instars of the mantid (Or- thoptera) Chaeteessa filata (Burmeister) (Dictyoptera: Mantidae) which is abun- dant on branches of cocoa trees at La Lola (Fig. 3). One rainy season census (August-1983) of 70 adjacent trees at La Lola revealed that close to half (32 trees) had one individual of this mantid at 0900 hours in sunny weather, and that 20 of these individuals were nymphs and the remainder, adults (A. M. Young, unpubl. data). Various species of salticid and other spiders also capture these cecidomyiids routinely, and Anolis lizards, which are very abundant in Costa Rican cocoa farms, also feed upon them (A. M. Young, unpubl. obs., 1978-1983). Small dipterans sometimes comprise sizable portions of anole diets in the American tropics (e.g. Floyd and Jenssen, 1983).
58 PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON
Fig. 3. Various aspects of cecidomyiid biology in Central American cocoa plantations. Clockwise, beginning in upper left photo: adult cecidomyiid perched on petal hood of cocoa flower during the daytime; squirrel-generated hole in cocoa pod, and filled with fungi and rotting pod tissues, a micro- habitat where cecidomyiid larvae are found; the mantid Chaeteessa filata on branch of cocoa tree— in earlier nymphal stages, a predator on adult cecidomyiids; adult cecidomyiid perched in hyphae of fungi on hole in cocoa pod.
While more midges are seen resting on staminodes, petal hoods, and petal ligules of cocoa flowers than seen moving in them at both La Lola and La Tigra (Tables 2 and 3), I suspect that such a paucity of direct observations of midges pollinating flowers, or at least visiting flowers in such a manner so as to possibly pollinate
VOLUME 87, NUMBER 1 59
them, is due to (1) my inability to find and record accurately all such instances at a given time of day, and (2) the very likely fact that only relatively few species of the total cecidomyiid faunas of cocoa farms are actually visiting cocoa flowers for the purpose of feeding or gaining some other reward. In the shaded understory conditions of cocoa farms cecidomyiids in general alight on light-colored objects. While the species most abundantly seen on or inside cocoa flowers fall within the body length range of 2-4 mm, the larger Mycodiplosis sp. 2, with variegated wings, is commonly found alighting in groups on the bottoms of large white-plastic cups used to simulate bromeliads (see Young, 1983). Thus it is necessary to distinguish between cecidomyiids that visit cocoa flowers for some purpose other than a resting place, namely for feeding. My observations suggest that forms such as Clinodiplosis sp. 1, Mycodiplosis sp. 1, Coquillettomyia sp. and A. triangularis, are the most likely pollinators of cocoa in Costa Rica. Of these, most of the observations on actively moving cecidomyiids inside cocoa flowers are of Cli- nodiplosis sp. 1 and Mycodiplosis sp. 1. Interestingly, females are far more abun- dant on or inside cocoa flowers than males for those genera and species found in greatest frequency associated with flowers (Tables 2 and 3). Examination of fresh- ly-open flowers on other species of Theobroma trees growing adjacent to cocoa at La Lola and La Tigra also suggest that cecidomyiids are attracted to these flowers (Table 4), even though there is considerable difference in size, color, and three-dimensional design of the flowers among these species (see Cuatrecasas, 1964). Young et al. (1984) and M. Strand and A. M. Young (in preparation) also found considerable differences in the types of fragrance compounds among these species, although the general positioning of floral nectaries and other possible reward-generating structures is similar among these species. For the purposes of this paper, I tentatively conclude that Clinodiplosis sp. 1 and Mycodiplosis sp. 1 are the most frequent cecidomyliids visiting cocoa flowers at La Lola and La Tigra, and are the best candidates for being effective pollinating agents. Our totally new observations of adult cecidomyiids being attracted to the flowers of Guazuma (Sterculiaceae) at La Tigra add indirect evidence for the existence of a generalized pattern of associaiton of these cecidomyiids with sterculiaceous flowers in the American tropics (Table 4).
A good review of the mechanism of pollination in cocoa by ceratopogonid midges is given by Bystrak and Wirth (1978) and my observations with cecido- myiids suggest a very similar pattern of association with flowers. Basically I propose that cecidomylids such as Clinodiplosis sp. 1 and Mycodiplosis sp. 1 pick up small numbers of viable cocoa pollen grains on long caudal hairs on the abdomen and elsewhere when midges crawl around inside petal hoods in the early morning and late afternoon. Pollen grains thus removed from the petal hoods are then smeared upon the pistil when pollen-bearing midges crawl along the narrow space between staminodes and pistil. Kaufmann (1973b) and Winder (1977b) both report smear-type pollination by cecidomyiids in Africa and Brazil, respec- tively.
Both Clinodiplosis sp. 1 and Mycodiplosis sp. 1 have well-developed compound eyes. While the mouthparts of both species possess no specialized adaptations for flower-feeding, both have elongate functional mouthparts adapted for sucking up fluids such as nectar in flowers (R. J. Gagné, pers. comm.). Interestingly members of both genera typically possess long hairs on the legs and abdomen (R. J. Gagné,
60 PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON
Table 2. The abundance of adult Cecidomyiidae on/in freshly-opened Theobroma cacao flowers* for five widely separated census dates? for 1000 flowers examined each date‘ at Finca Experimental La Lola, near Siquirres, Limon Province, Costa Rica.
No. Midges
Resting on No. Crawling Communal Now Occupied Flowers? Inside Flowers* Roosts!
Species Flowers 22 58 28 22 38 T 22 33 Ty Clinodiplosis sp. | 17/1000 13 0) 13 3) 0 5 (0) 0 0) Clinodiplosis sp. 2® 0/1000 = - = - 8 ey uls) Clinodiplosis sp. 3 0/1000 — — - = 2 l 3 Mycodiplosis sp. 1" 63/1000 100 16 116 1 0 1 25 - 27 “52 Mycodiplosis sp. 2 0/1000 - = 10 0 10 Mycodiplosis sp. 3 0/1000 = — _ = 10 0 10 Coquillettomyia sp.’ 0/1000 _ = _ 30 9 39 Lestodiplosis sp. 2 5/1000 2 5 q/ = — 0 0 0 Aphodiplosis triangularis 19/1000 30.18 48 _ - 0 0 0) Winnertzia sp. 4/1000 2 4 6 - 0 0 0 Ledomyia sp. 1/1000 ] (0) 1 — = 0 0 0 Bremia sp. 2/1000 1 l 2 _ = 0 0 0 Trisopsis sp. 1/1000 1 0 1 — = 0 0 0 Chrybaneura sp. 0/1000 = _ = — 0 5 5 Undet. Cecidomyiidae 6/1000 3 10 13 — — 0 0 0
Total det. species: 14
Total det. species, individuals, and sex ratio on/in flowers, resp.: 9, 194, 150:44. Total no. occupied flowers: 112/1000
Usual range of midge number/flower: 1-3
“Most censuses of flowers occurred between 0630-0900 hours and freshly-opened flowers were recognized by the horizontal orientation of the sepals.
+ 18-23 Nov. 1981, 10-14 March 1982, 17-23 July 1982, 8-11 Dec. 1982, 12-15 March 1983.
© A total “pool” of approx. 1000 flowers was taken each date from usually 20-70 trees depending upon the time of the year and availability of flowers; tree varieties used were UF-677, UF-613, and Pound-7.
4 Defined as one or more midges suspended motionless usually from staminodes, ligules, sepals or petal hoods of open flowers.
© Defined as a midge visibly seen crawling either between staminodes and pistil, or inside petal hood of freshly-open flower.
‘ Ageregates of several midges on spider webbing in trees; data combine one or more roosts.
® Tentatively named as new species denotata (R. J. Gagné, pers. comm.).
» Tentatively named as new species /igu/ata (R. J. Gagné, pers. comm.).
‘Tentatively named as new species obliqua (R. J. Gagné, pers. comm.).
pers. comm.) and a preliminary SEM survey of wild-caught Mycodiplosis sp. | (N = 6 observed) revealed that one cecidomyiid had a cocoa pollen grain adhering to one caudal hair on the abdomen. While members of both genera occur world- wide, the particular species associated with cocoa flowers may have tropical dis- tributions, and these species will soon be described by Gagné (in press). Individual cecidomyiids approach an open cocoa flower with considerable “bobbing and weaving” flight paths around a flower. Seemingly healthy flowers are often passed by completely by cecidomyiids flying along branches in cocoa trees. The flies do not appear to “swarm” about open flowers in large numbers. But Cecidomyiidae sometimes swarm along branches in cocoa trees (i.e. 12 Feb- ruary 1981 at 1700 hours at La Lola) and with both sexes represented. Winnertzia sp. was found swarming (approx. 20 midges seen) at 1500 hours on 18 November
VOLUME 87, NUMBER 1 61
Table 3. The abundance of adult Cecidomyiidae on/in freshly-opened Theobroma cacao flowers* for six widely separated census dates? for 500-1000 flowers examined each date‘ at Finca La Tigra, near La Virgen, Heredia Province, Costa Rica.
No. Midges Resting No. Crawling on Flowers? Inside Flowers* Communal Roosts' No. Occupied ee
Species Flowers 2 od T 22 38 Ti 29 36 T Clinodiplosis sp. | 14/1000 23 0 23 —_ — = — Mycodiplosis sp. 1® 18/1000 25 3 28 1 0) 1 = = Mycodiplosis sp. 2 3/1000 0) 3 3 a = _ - Mycodiplosis sp. 4 1/500 0) 1 1 _ — — - Coquillettomyia sp." 2/1000 2 0) 0 = aa — - Lestodiplosis sp. | 2/1000 2 0 2 — _ — = Aphodiplosis triangularis 3/500 3 0 3 = - - _ Winnertzia sp. 1/500 1 0 1 _ — — — Ledomyia sp. 1/1000 1 0 I = _ _ ~ Cantarina sp. 2/1000 4 1 5 — — - - Feltiella sp. 4/1000 3 3 6 = ~ = Undet. Cecidomyiidae 7/1000 tg ? ) - = - -
Total det. species: 11
Total det. species, individuals, and sex ratio on/in flowers, resp.: 11, 75, 64:11. Total no. occupied flowers: 68/500—1000
Usual range of midge number/flower: 1-2
° Most censuses of midges in flowers occurred between 0630-0900 hours following an initial period of determining diurnal patterns of peak abundance, and freshly-opened flowers were readily recognized by the marked horizontal orientation of the sepals (which are more vertically positioned in older flowers).
» 11-24 March 1979, 1-4 Nov. 1980, 4-18 Feb. 1981, 12-17 Nov. 1981, 8-13 and 27-31 July 1982, and 2-6 March 1983.
© A total “pool” of usually 500-1000 flowers was taken as a sample each date from usually 20-50 trees in the same area (many mixed hybrid varieties of undetermined origin) depending upon the time of year and availability of fresh flowers; because cacao flowers drop off the trees within 24—48 hours after opening under most conditions, the occupancy data given in the table are expressed in terms of 500 or 1000 flowers, i.e., a single census.
* Defined as one or more midges suspended motionless usually from staminodes, ligules, sepals or petal hoods of open flowers.
© Defined as a midge seen actively crawling either between staminodes and pistil or inside petal hood (including entering and exiting) of freshly-opened flower.
‘In the La Tigra study as compared to the La Lola study (Table 2), little attention was given to roosts.
84 See species determinations given for same footnotes in Table 2.
1981 (La Lola) along a branch in a cocoa tree. Micromya sp. was found swarming about 5 cm above some rotting slices or discs of banana tree trunks (Young, 1983) (La Tigra, 26 July 1981).
Freshly-opened cocoa flowers between 0530 and 0700 hours are heavily visited by cecidomyiids, and these exhibit the potential pollinating behavior described above. While some of this activity may also occur in the pre-daylight hours of the dawn period, the rather compressed daylight period of cecidomyiid activity at dawn probably precedes the formation of communal roosts (see data below). While the basis for attraction to freshly-opened flowers awaits study, when three of the five petal hoods are surgically removed from flowers, cecidomyiids are found within a few minutes resting on various exposed floral parts: on 19 No- vember 1981 at 0800, five were found perched on exposed anthers, petal ligule
62 PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON
on one of the remaining petal hoods, and the tip of a staminode. Although the flower was deliberately damaged, this act did not lower attractiveness of the flower to potentially pollinating insects.
The results of a preliminary experimental study involving the confining of Mycodiplosis sp. 1 to experimental cages on flower-laden branches of cocoa trees at La Lola provide additional evidence that cecidomyiids can be cocoa pollinators, and confirms what others have concluded for other genera and species of these midges (Kaufmann, 1973b; Winder, 1977b). On 30 July 1981 a series of 14 such cages (2 experimentals and 12 controls) were set up (see Young (1983) for view of these cages) in the Area B study site of Young (1983) using UF-29 cocoa trees. One of the two experimental trees had 95 floral buds within the caged section of the branch and the second one had 23 buds at the time the test was initiated. The control trees had a range of about 30 to 100 floral buds at this time (these are tagged trees nos. 62, 63, 66, 76, 78, 79, 80 of Young, 1984b). The two experimental cages combined received a total of 33 cecidomyiids between 1-3 August 1981. By 10 August, one cherelle was well in evidence in one of the two experimental cages and no other cages had pods. The experiment was repeated on 12 March 1982 at La Lola, but this time in the Area A study site of UF-29 cocoa described in Young (1983). This time, however, I used only one control and one experimental cage, both on the same tree (tagged tree No. 15 of Young, 1984a). The confined area of the branch covered by the control cage had 64 floral buds while the experimental cage had 51 buds. A total of 39 individuals (mixed sexes) of what I judged to be Clinodiplosis sp. | were introduced into the experimental cage between 12-14 March. By 29 March three new cherelles were evident in the experimental cage and none in the control cage. In both tests, the cages were left in place on the trees until the final census (of pods) was taken.
Abundance Patterns of Cecidomyiids in Cocoa.— Typical adult abundance pat- terns in open cocoa flowers are readily seen from some data collected at La Lola (Area A of UF-29 cocoa in Young 1983) during March 1982: from a total of 70 flowering trees examined almost exhaustively, and representing a total pool of about 2000 open flowers, a total of 29 adult cecidomyuids were found perched on flowers, with a range of 1-2 per “occupied” flower and with five observations of three cecidomyiids on a single flower (and all observations made 1100 to 1300 hours in hot, sunny weather). During the March 1983 dry season at Hummingbird Hershey, a total of 39 cecidomyiids were counted from a total of 28 freshly- opened flowers at 1100-1200 hours on one date; total flower abundance on these trees just exceeded 5800 at this time (5880 open flowers). But unlike La Lola, these trees also had a total of 1586 pods, mostly cherelles, from natural pollination, whereas at La Lola pod numbers were less than 200 and the result largely of hand- pollination. The Hummingbird Hershey census area was well-shaded due pri- marily to self-shading cocoa trees, and an adjacent section of relatively little such shade yielded a total of 9 cecidomyiids found from a total sample of 3725 open flowers on an additional 70 trees. This area had a total of 245 cherelles. Generally, adult cecidomyiids are of very low density in open cocoa flowers, as also seen by some data from La Tigra: on 3 November 1980 at 1100 hours (hot, sunny day), a total of 20 flowering cocoa trees (mixed varieties) yielded a total of 640 open flowers (X + SD = 31.00 + 18.39 flowers per tree for N = 20) and a total of 16 cecidomyiids (¥ + SD = 0.85 + 0.72 per tree) from open flowers. Total abundance
VOLUME 87, NUMBER 1 63
of cecidomyiids, i.e. all captured flying about flowers and landing on branches and leaves at this time was N = 69 or x + SD = 3.45 + 1.14, a sample that undoubtedly included many non-pollinating species. An additional census at La Lola in March 1982 yielded a total of 54 cecidomyiids (28 female & 26 male) or ¥ + SD = 1.42 + 0.77 per tree and with a range of 1—4 per flower and 25 flowers with one cecidomyiid each, 6 with 2 each, 3 with 3 each, and one flower with 4 cecidomyiids each. Another census in the clonal garden area at La Lola in August 1983 revealed that nine of 70 trees had one or more cecidomyiids on or inside open cocoa flowers, and a total of 13, with 1-2 per “occupied” flower (5 female & 8 male); most of these appeared to be Mycodiplosis sp. 1. At this time, flowering was very low, with a range of about 0-5 open flowers per tree (N = 70 trees) and with slightly more than 50% of the trees with no flowers at all. In all censuses of cecidomyiids on flowers, virtually every specimen found was perched on one of the following floral parts: petal hood (abaxial surface), petal ligule, sepal, or stam- inode. All observations of specimens actually crawling inside flowers were limited to 0600-0800 hours (most observations) and 1600-1800 hours for daylight sur- veys taken. Other studies (Young, 1983; Young et al., 1984) indicate that cocoa flowers generally begin opening in the late afternoon and are fully open by 0400- 0500 hours at both La Lola and La Tigra (see also the data of Wellensiek, 1932). Such a pattern is consistent with the early morning daylight activity period of cecidomyiids observed in the present study (Tables | and 4). Yet it is not un- common to find large numbers of these midges perching motionless on open flowers at other times of the day. Casual observation indicated that midges were at least three times as abundant in this well-shaded area (Area A of Young, 1983) as contrasted with an unshaded area (Area B) at this time (dry season). Subsequent determinations indicated that most of the midges in all of these censuses were mixes of Mycodiplosis sp. 1, Clinodiplosis sp. 1, and Coquillettomyia sp. Differ- ential partitioning of adult midges between shaded and sunny cocoa habitats at La Lola during the drier months was also demonstrated by an examination of aggregates of Clinodiplosis sp. 2 on the undersides of large plastic cups suspended from cocoa trees in both areas (see Young 1983 for details): on 13 March 1983 at 1400 hours, a total of 6 out of 20 such cups in Area A (shade) had a total of 51 midges with 5-16 per cup, while only one of 20 cups in Area B (sunny) had one midge. A 12 March 1982 census (i.e. one year earlier) of the same areas yielded a total of 56 midges from 7 “occupied” cups in the shaded area and none in the sunny area. During the drier months at La Tigra (such as March) typically many cecidomyiids are found on open cocoa and Guazuma flowers between 0600-0800 hours, but that after this time there is marked decline and virtual absence of specimens at flowers until much later in the day (near dusk). Such diurnal decline patterns are less pronounced during the rainy season for cocoa flowers. Additional data on cecidomyiid abundance in cocoa flowers are seen in Table 2. A total of 12 genera and 16 species were collected in the La Lola studies, of which nine genera and species were found visiting cocoa flowers. A total of 116 individuals of Mycodiplosis sp. 1 (99 female & 17 male) on flowers (Table 2) represents about 60% of the total fauna of 193 determined midges) on flowers, followed by A. triangularis with about 20% of the fauna on flowers (N = 53 or 28 female & 25 male), and then by Clinodiplosis sp. 1 (all female) at about 6% of the total sample taken (Table 2). At La Tigra the most abundant species found
64 PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON
Table 4. Records of adult Cecidomyiidae in freshly-opened or open flowers of Theobroma species other than cacao at Finca Experimental La Lola and Finca La Tigra in Costa Rica.*
Time of Midge Species Date Day Sex Tree Species Other Comments La Lola Clinodiplosis Feb. 1981 1600 male T. simiarum on staminode sp. | Mar. 1983 0800 male T. speciosum on ligule Mar. 1983 0700 male 7. mammosum _ on ligule Prodiplosis sp. Feb. 1981 0700 1 female, T. simiarum on staminodes 3 males Clinodiplosis Mar. 1983 0600 = 2 females, T. mammosum — on staminode sp. 4 1 male Mycodiplosis Mar. 1983 1600 female T. silvestre on staminode sp. | Bremia sp. Mar. 1983 1700 female T. speciosum inside petal hood Mar. 1983 1700 male T. simiarum inside petal hood Asynapta sp. Mar. 1983 HA ? T. mammosum — on staminode La Tigra Clinodiplosis Mar. 1983 1000 - T. mammosum communal roost” of 18 sp. | (mixed sexes) Mar. 1983 0700 3 females Guazuma sp. landing on open flowers Bremia sp. Mar. 1983 1000 _ T. bicolor communal roost of ap- prox. 400 (mixed sexes)*
® Examination of flowers made opportunistically on various days and not following a particular pattern.
> This roost was located on spider webbing between leaf petiole and branch and about 10 cm from nearest open flowers.
© Also Sterculiaceae.
4 This roost located on webbing stretching across two leaves and about 15 cm from nearest open flowers.
in cocoa flowers (and also in Guazuma sp. flowers) was Clinodiplosis sp. 1 (23 of 44 or about 50%). Eight genera and 13 species were collected from flowers at La Tigra. At La Lola all five Clinodiplosis sp. 1 seen moving inside floral parts of cocoa were females as well as the two Mycodiplosis sp. | (Table 4). Several genera that were not found in cocoa flowers (Table 3) at all were found forming communal roosts in cocoa habitats. Caution is exercised in interpreting these patterns as the data are small. Some species, such as 4. triangularis, while abundant in flowers, were never found in roosts (Table 3), but Mycodiplosis sp. 1, while common in flowers, also constituted about 69% of all cecidomyiids found in communal roosts (Table 3). In all studies, however, relative to abundance of open flowers, ceci- domyiids are relatively rare in cocoa flowers in both rainy and dry seasons in Central America (Table 5). For the three most abundant species in the La Lola samples, namely, Mycodiplosis sp. 1, A. triangularis, and Clinodiplosis sp. | (in this order), a total of 122 individuals were collected on two different ““dry” season dates and 66 on three wet season dates, even though the same number of trees was covered on all dates and samples were always taken at the same times of the day. For both seasons, all census dates, these three species constitute 188/194 or
VOLUME 87, NUMBER | 65
Table 5. Estimates of population density of adult Cecidomyiidae on freshly-opened flowers* of Theobroma cacao in the La Lola (Costa Rica) and Hummingbird Hershey (Belize) cocoa farms in Central America.
No Percent-
Total No Total No. Midge age (%)
Census Open Adult Occupied Flower Dates “Season”? Flowers* Range "am be) DL Midges x + SD Flowers Occupancy
La Lola 21-VII-82 mid-rainy 650 0-53 Ose = 1120 7 0.11 + 0.36 6 0.92% 8-XII-82 late rainy 1577. 0-75 22-52) +-19:55 19 0.27 = 1.25 7 0.44% 14-I11-83 dry 1712 0-300 24.45 + 41.45 61 0.77 = 215 31 1.82% 5-VIII-83 mid-rainy 3278 1-178 38.24 + 47.04 30 1.50! 2: 7 14 0.42% Hummingbird Hershey 22-11-83 dry $880 10-108 73.50 + 48.64 39) 10:535-=21209) 28 0.47%
* For each locality, all of the data on both flowers and midges were taken always from the same set of mature and healthy cocoa trees. A total of N = 70 trees were sampled each time at each locality (in the case of La Lola for the repeated censuses). At La Lola, the tree sample was almost evenly divided among the clonal varieties of UF-677, Pound-7, and UF-613 (these rows being contingent), while at Hummingbird Hershey the 70 trees were of mixed undetermined varieties. Only freshly- opened flowers were examined for midges and only midges resting on or inside definite floral parts were counted.
> “Season” is considered here as a relative condition of “wetter” and “drier” periods as defined by available rainfall data (Fig. 1). As such, this condition changes considerably from year to year at each locality.
© For 21-VII-82 census at La Lola, 153/650 or 23.5% of open flowers with horizontal sepals, 1.e., flowers freshly-opened that morning (as opposed to being 1-day old flowers); similar level for 8-XII- 82; for 14-III-83 the 31 flowers occupied by cecidomyiids all but 4 with horizontal sepals and 36.1% of all flowers at this time with this condition of freshness. In Belize sample about 50% of all flowers with horizontal sepals.
about 97% of all cecidomyiids collected (Table 2). For La Tigra, only rainy season data (2 dates) are available.
The experimental studies using small flypaper squares to trap flying insects within the vicinity of open cocoa flowers and away from them (Table 6) revealed very low abundance of cecidomyiids in both day and night samples at La Lola during the rainy season. While the sample size is small, there is a slight trend towards finding more specimens in “experimental” sticky squares than in the “‘control’’ squares away from flowers (Table 6). Prior to these results, a total of 48 sticky squares distributed almost randomly among 15 trees (all Pound-7) at La Lola and within one area yielded two cecidomyiids (both Mycodiplosis sp. 1; females) captured in the traps between 1700 and 0700 hours (dusk-to-night-to- dawn) on 16 July 1982, and the same test repeated on the following evening produced one cecidomyiid of undetermined species designation. From 27 to 31 July 1982 similar tests conducted at La Tigra provided similar results: in one patch of four cocoa trees a total of 127 squares were placed around 60 open flowers at 1700 hours (27 July) and checked at 0730 the following day. A total of 18 squares had trapped flying insects (21) including one female Clinodiplosis sp. 1. A second test conducted with the same trees and same number of squares and flowers yielded no cecidomyliids for the period 0800 to 1700 hours (28 July) even though 49 other flying insects were trapped during this daytime census. Tentatively these findings taken together suggest that cecidomyiids are not abundantly active
66 PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON
Table 6. The occurrence of Cecidomyiidae versus other flying insects in small squares of sticky flypaper adjacent open cocoa flowers (experimentals) and on bare spots on branches (controls) in an area of typically-shaded Pound-7 variety cocoa trees in a clonal garden at Finca Experimental La Lola, Costa Rica.
No No Control No. Squares No Expen- No Squares with Cecido- Date mental Flowers (0 Trapped myuds Intervals Time Range Tree No. Squares Used Flowers) _ Insects Trapped Species
20 July 0630-1730 ] 46 25 (0) 14/46 ] Mycodiplosis sp. | (dawn—-day-— 2 22 20 0) 5/22 2 Mycodiplosis sp. | dusk) 3 0 0 30 3/30 0 (all females)
Total trapped insects: 44 (all squares)
20-21 July 1730-0630 1 46 25 0 15/46 ] Clinodiplosis sp. 1 (dusk-night— 22 20 0 8/22 l Mycodiplosis sp. | dawn) (0) 0 30 10/30 (both females)
Total trapped insects: 40 (all squares)
21-22 July 1730-0630 1 35 16 0 6/35 0 Clinodiplosis sp. 1 (dusk—night— 2 10 6 0 5/10 2 (both females) dawn) 3 0 0 30 2/30 (0)
4(new) 14 8 0 3/14 0
Total trapped insects: 15 (all squares)
at cocoa flowers at any time of the day or night, assuming that the trapping method used 1s effective in capturing these insects, but that cecidomyiids are very likely active both at dawn and dusk at flowers.
Communal Roosts of Cecidomyiids in Cocoa.— When a total of nine different communal cecidomyiid roosts were followed in the shaded UF-29 cocoa at La Lola (Area A of Young, 1983) in November 1981, all but three roosts lasted at least five successive days in the same locations (usually in branch crotches or cherelle petiole-branch crotches) even though actual size (numbers of cecido- mylids) in each roost varies greatly from day to day. Roosts both increase and decrease in size over successive days. In one roost studied, a pattern of increase over three days was from 6 to 18 to 19, while the decline pattern of a second roost for the same period was 90 to 5 to 10. Roosting midges occupy small spider webs anchored to cocoa tree branches in shaded spots, and most roosts are within 0.1 to 3.0 m of the ground. Occasionally living spiders also occupy the webs used by roosting cecidomyiids; these include Theridiidae (possibly Achaearenea sp.) and Leucauge sp. (Araneidae). While most roosts at La Lola appear to consist of one primary species, one mixed-species roost of about 100 cecidomyiids collected at La Lola in March 1982 included 30 Mycodiplosis sp. 1 (15 female & 15 male), one Mycodiplosis sp. 2 (1 male), 10 Coquillettomyia sp. (8 female & 2 male), and Feltiella sp. (1 male). On a day to day basis, very large roosts, defined here as consisting of 100 or more midges at any one time, frequently break up into much smaller aggregates at the original roost site. As long as a web remains undamaged (by rain, wind, abrasions, etc.), midges may occupy the same web on several days. Cecidomyiids are most prevalent on roosts during daylight hours and vacate roosts at night; in clear sunny weather these frequently leave roosts by 1745 hours, as do those perching motionless on open cocoa flowers. On 20 November 1981 I
VOLUME 87, NUMBER | 67
witnessed the return of several cecidomyiids into a tagged roost site (web) at 0630 hours, and most roosts are formed between 0600-0700 hours each morning. Unless disturbed, the flies do not normally leave roosts in the daytime. The majority of specimens found in roosts appear to have much darker cuticles than many found on freshly-opened flowers in daylight hours, a condition suggesting that freshly- eclosed adult midges, particularly females (see Table 2), seek out cocoa flowers before entering into roost formations. On rainy, overcast days, dispersal from roosts may occur earlier in the afternoon, between 1700 and 1730 hours. Periods of heavy daytime rains do not disturb midges from roosts. During March 1982 in the shaded UF-29 cocoa habitat at La Lola, a total of 25 roosts were discovered among 50 cocoa trees, representing a total of 723 cecidomyiids (mostly M/yco- diplosis sp. 1) with average roost size being ¥ +SD = 28.92 + 31.21 for a total of 18 trees having one or more roosts; the range in numbers of specimens per roost was 1-125 for a census at 1100 hours in hot, sunny weather. In the clonal garden area at La Lola in December 1982 (late rainy season) only three roosts were found on a total of 70 cocoa trees, with 20-40 cecidomyiids per roost; specimens were also scarce on open flowers at this time. Lowest roost densities in cocoa trees occur in the rainy season at La Lola. At Hummingbird Hershey during August 1981 (rainy season) a total of five roosts were found within a 10 by 10-meter area within Field Block 18, with range in midges being 25-150 at 0900 hours and all midges being Coquillettomyia sp. Mortality of roosting midges is probably low, but on 20 July 1982 at La Lola one roost of 25 specimens was totally encased in hyphal growths of fungi when discovered at 0900; the roost was found following an evening period of heavy cold rains. When the webbing of a roost site is destroyed, sometimes new webbing appears in the same spot a day or so later, and new webs are sometimes recolonized by midges. Earwigs scattering along branches in cocoa trees at La Lola (A. M. Young, in preparation) occasionally fall into webs with roosting cecidomyiids, and these quickly disperse in response to such disturbances. By following diurnal patterns of roost formation and dis- solution, it was determined that roosting in cocoa cecidomyliids is a daytime behavior, and that they are active principally at dusk, night, and dawn (Table 7). Roosting cecidomyiids are not disturbed by actively hunting insect predators such as Anolis lizards and salticid spiders. The flies are most vulnerable to these predators when hovering above branches or landing on branches, leaves, and flowers. The exploitation of spider webs in cocoa trees may function to protect cecidomyiids from such predators at times of the day when these insects are largely inactive.
Cecidomyiid Midge Life Cycle Substrates.— Cocoa trees with branches covered with epiphytic mosses generally have many adult cecidomyiids flying about them during the early daylight morning hours. Most cocoa-associated cecidomyiids most likely breed close to the ground, colonizing various kinds of rotting organic materials for completion of the life cycle. Small areas within cocoa farms that consistently have high numbers of cherelles are also sites of sustained high adult population densities of cecidomyiids and ceratopogonids (A. M. Young, field obs. 1978-1983). At Hummingbird Hershey, a total of six large cocoa pods in various stages of rotting, and with large woodpecker-generated holes in them, yielded a total of 158 cecidomyiid larvae, with a range of 6-60, and for a total sample of 20 damaged pods (all but 4 150-180 mm long) examined (of which six had midge
68 PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON
Table 7. Comparative abundance of cecidomyiid midges (mostly /ycodiplosis sp. 1) in communal roosts* between day and night in one section of cocoa plantation” in Finca Experimental La Lola, near Siquirres, Limon Province, Costa Rica, during two different rainy season census periods.
0700 to 1700 hrs (“Day”) 1800 to 0600 hrs (‘Night’) No Max. No. Midges Max. No. Midges Roosts Range in Seen in Roosts Range in Seen in Roosts. Day Present Roost Size Day Roost Size Day
18-22 November 1981 (“late rainy”’)
1 5 2-6 26 0 - 2 9 2-110 131 0-10 10 3 9 2-18 47 0 - 4 9 1-22 73 0-16 16 5 9 0-34 34 0 =
Total midge “‘observations”’:* 341 X + SD (N = S days): 68.20 + 54.85
16-22 July 1982 (“mid-rainy”’)
l 5 10-45 109 0 - 2 i 1-20 67 0 _ 3 14 2-150 S17 0 - 4 15 2-60 217 (0) _ 5 15 1-88 25) (0) 135 6 17 7-51 322 0 = 7 17 2-53 293 (0) —-
Total midge “observations”’:* 1776 X + SD (N = 7 days): 253.71 + 148.70
* Roosts were discovered and positions tagged with small, yellow plastic tags (numbered); most roosts in spider webbing draped between branches or pods and branches in cocoa trees.
» Well-shaded “habitat” of UF-29 cocoa—the “Area A” of Young (1983).
© Defined as the total number of midges obtained by summing the “maximum number seen” figures in the table.
larvae), the average abundance was ¥ + SD = 7.85 + 17.52 midges per pod (23- 24 March 1983 census). Within a given pod having larvae, larvae are distributed in a patchy fashion. Larvae occur in both moist and dry exocarp as well as on moist rotting seeds inside pods. Shaded areas of cocoa have many rotting cherelles and tend to have more roosts of cecidomyiids than exposed areas. Pockets of several larvae frequently occur inside pods. At Hummingbird Hershey in March 1983, I witnessed several instances of female cecidomyiids perched on rotting pods in cocoa trees, and with the terminal tips of swollen abdomens (perhaps filled with mature eggs) probing the pod surface as if determining an egg-placement site. Larvae clustered in pockets in pods may reflect a clustered egg-laying habit in cecidomyiids. A hanging pod at La Lola in March 1982 and having a large squirrel-generated hole in it was blanketed with the hyphae of various fungi (Fusarium, Phytophthora, and Thielaviopsis sp.) and several adult cecidomyiids (Clinodiplosis sp. | and several undetermined ones as well) having swollen ab- domens were perched in the fungal growth (Fig. 3). When the pod was collected and opened, many larvae of Clinodiplosis sp. 1 were found inside, within the rotting tissues exposed by the wound in the pod wall. Prior to collecting, adults on the fungus remained there for several hours, and adults were also present the
VOLUME 87, NUMBER 1 69
next day as well. Clinodiplosis larvae make audible clicking sounds and jump several centimeters when disturbed. Various genera and species of cecidomyiids dwelling in cocoa farms in Central America undergo their life cycles in a range of rotting organic debris (Table 8). Samples of epiphytic moss from cocoa trees brought into the laboratory (Milwaukee) exhibit a gradual day-to-day emergence of Mycodiplosis sp. 1 adults: a sample collected in November 1981 at La Tigra generated 3 adults on 28 December 1981, and one each on 7, 8, 11 and 12 January 1982, and with all emergences occurring 0700-1000 hours under laboratory con- ditions. Both Mycodiplosis and Clinodiplosis species are known to be fungus- feeders in the larval stages and none are predatory or gall-makers (R. J. Gagné, pers. comm.). Cocoa farm habitats appear to be suitable for breeding in cecido- myiids, some of which may be cocoa pollinators (Table 8).
DISCUSSION
Other studies have indicated that some Cecidomyidae are pollinating agents of cocoa in both the New and Old World humid tropics (Kaufmann, 1973b; Winder, 1977b, 1978a), although the general conclusion has been that these midges are only incidental or fortuitous cocoa pollinators (e.g. Entwistle, 1972; Winder, 1978a). The present findings indicate that a few species of cecidomylids, such as Clinodiplosis sp. 1 and Mycodiplosis sp. 1, are natural pollinators of cocoa in Central America, and several species also visit the flowers of other Theobroma species. The observed mechanism of pollination of cocoa by cecidomyiids in Costa Rica did not appear to include the ““wing vibration” syndrome of several specimens in one flower as observed by Kaufmann (1973b) in Ghana. Rather, pollination appeared to be largely limited to the movements of individual ceci- domyiids through enclosed reproductive floral parts in the present study.
In general the mode of cocoa pollination by cecidomyiids in Costa Rica matches well the typical pollination system involving ceratopogonid midges (e.g., see the review in Entwistle (1972) and in Bystrak and Wirth (1978) with these notable differences: (1) cecidomiids tend to carry fewer pollen grains than ceratopogonids (as also noted by Winder, 1977b), (2) cecidomyiids are mostly active in cocoa flowers at dawn, dusk, and at night while ceratopogonids are generally most active in the later daylight hours of the morning (e.g. Fontanilla-Barroga, 1962; Soria et al., 1980), and (3) cecidomyiids exhibit apparent feeding behavior at the fleshy petal ligules in areas filled with stomate-type nectaries (see Young et al., 1984) whereas this floral part is largely ignored by ceratopogonids; such feeding in cecidomyiids generally precedes the movement of midges into the enclosed re- productive parts of the cocoa flower, and might therefore be a prerequisite to actual pollination by cecidomyiids. Decazy et al. (1980) found many cocoa pollen grains adhering to the cuticle of cecidomyiids in South Cameroon, and determined that these midges often fly up to 35 meters and can therefore serve as cross- pollinating agents. Winder (1977b) also noted pollen grains in the cuticle of cec- idomyiids in Brazil and Privat (1979) describes the movement of cecidomyiids in the petal hoods of cocoa flowers in Costa Rica. The question of whether or not this interaction is fortuitous awaits further study, given the general conclusion that dipterans are generalist floral visitors (e.g. Percival, 1965; Faegri and Pijl, 1971: Kevan and Baker, 1983). Yet some field studies indicate a more specialized role of some dipterans as pollinators (e.g. Baumann, 1978; Vogel, 1978; Mesler et
70 PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON
Table 8. Rearing records for Cecidomyiidae in some Central American cocoa farms.
Species
Mycodiplosis sp. |
Mycodiplosis sp. 2
Clinodiplosis sp. |
Clinodiplosis sp. 3
Coquillettomyia sp.
Micromyia sp.
Ledomyia sp.
Lestodiplosis sp. Porricondaylia sp.
Camptomyia sp.
Resseliella sp.
Undet. Cecidomyiidi
Locality
La Tigra (Costa Rica)
La Lola (Costa Rica)
La Tigra (Costa Rica)
La Tigra (Costa Rica)
La Lola (Costa Rica)
Hummingbird Hershey (Belize)
La Lola (Costa Rica)
La Lola (Costa Rica)
La Lola (Costa Rica)
La Tigra (Costa Rica)
La Tigra (Costa Rica)
La Tigra (Costa Rica)
La Tigra (Costa Rica)
Hummingbird Hershey (Belize)
Hummingbird Hershey (Belize)
La Lola (Costa Rica)
Rearing Conditions
10 from larvae on moss on trunk of tree in forest adjacent to cocoa; July-August, No- vember 1981
1 from larva in experiment rotting disc (slice) of banana tree trunk (Young, 1983), March 1983
4 from larvae in ground-cover rotting cocoa leaf litter; July 1982
9 from larvae in arboreal plastic cup (artificial bromeliad) filled with rotting cocoa leaves (Young, 1982) and from moss on trunk of tree in forest adjacent to cocoa; Dec. 1978, Nov. 1981 and March 1983
30 larvae in squirrel-generated wound in ripe cocoa pod in tree (with pod hole covered with hyphae of various fungi); March 1982
8 larvae just beneath dry exocarp layer of hanging cocoa pod destroyed by black pod rot; March 1983
from larva in arboreal plastic cup (artificial bromeliad) filled with rotting cocoa leaves (Young, 1983); February 1981
from larva in arboreal plastic cup (artificial bromeliad) filled with rotting cocoa leaves (Young 1983); August 1981
from larvae in experimental rotting discs (slices) of banana tree trunk (Young, 1983); July 1981
from larva in arboreal plastic cup (artificial bromeliad) filled with rotting cocoa leaves; December 1978
from larva in ground-cover rotting cocoa leaf litter and 2 from larvae in arboreal plastic cup (artificial bromeliad) filled with rotting cocoa leaves; Dec. 1978 and March 1979
from larva in ground-cover rotting cocoa leaf litter; March 1979
2 from larvae in ground-cover rotting cocoa leaf-litter,; December 1978
tN
—
12 larvae in both wetter and drier portions of exocarp galleries in 3 arboreal rotting cocoa pods; March 1983
10 larvae inside woodpecker-generated gouge or hole in arboreal cocoa pod in early stages of rotting cycle; March 1983
5 larvae in squirrel-generated wound in ripe cocoa pod in tree (with pod hole covered with hyphae of various fungi); March 1982
VOLUME 87, NUMBER 1 71
al., 1980). The presence of functional elongate mouthparts and exceptionally long hairs on the legs and abdomen of Clinodiplosis and Mycodiplosis, and the observed apparent feeding by these midges on petal ligules and other floral parts involved in the reproductive biology of cocoa, point to a specialized association with flow- ers. Both Kaufmann (1973b) and Winder (1977b) observed cecidomyiids inside petal hoods of cocoa flowers in Ghana and Brazil, respectively, and related this behavior to pollination. Other genera of these midges formed the basis of their studies. Patterns of feeding on nectar and pollen by some Diptera may vary within a group of closely related species and may be expressed in terms of divergence in mouthpart structure (e.g. Gilbert, 1981). Although there exist some extensive records of life cycle associations in Neotropical Cecidomylidae (Gagné, 1968, 1977), a pollination association with cocoa has not been emphasized in great detail (with the exceptions of Kaufmann, 1973b and Winder, 1977b). The ob- served heavily skewed abundance of female midges in cocoa flowers suggests a possible feeding association somehow linked to the reproductive physiology of these midges. But Kaufmann (1973b) found a 1:1 sex ratio in Parallelodiplosis triangularis (Cecidomyiidae) in cocoa flowers in Ghana, and the females were sexually mature. Pollen-feeding by female Heliconius butterflies is an adaptation for acquiring some amino acids essential for egg development (Gilbert, 1972). Female cecidomyiids in cocoa farms may seek out flowers for pollen or nectar, and then proceed to breed in nearby microhabitats such as epiphytic mosses and rotting cocoa pods still on the trees. In some studies of mosquitoes, there are good correlations in females between breeding sites and feeding sites from the previous evening (e.g. Kay, 1983). While Baker et al. (1953) noted associations of ceci- domyiids with predaceous larval stages feeding on mealybugs associated with cocoa and other Theobroma in Brazil, free-living non-predaceous forms such as species found as adults in cocoa flowers may have larval stages associated with these trees in a different way: larvae are found in the rotting tissues of either animal-damaged or diseased pods still on the trees. Toxomyia, a genus evolu- tionarily close to Mycodiplosis, has larval stages associated with rust spores on leaves in the West Indies (Parnell, 1969). Unlike the association of other ceci- domyiids with the flowers of grasses in the American tropics (Soderstrom and Calderon, 1971), larval stages of cocoa-pollinating cecidomyiids have not been found in the flowers. Kaufmann (1937b) in her Ghana studies has reported larvae of cocoa-associated cecidomyiids in flowers.
There is an abundant data base indicating that small-bodied midges such as cecidomyiids are largely limited in activity to moist forest habitats or to times of the day and evening when ambient conditions are moist and cool (e.g. Fisher and Teetes, 1982: Barnes, 1930; Summers, 1975; Shazli and Mostafa, 1980; Brewer, 1981; Vanhara, 1981; Brown and McGavin, 1982; Charlwood et al., 1982; Lums- den, 1952; Willmer, 1982). My data indicate a strongly crepuscular and nocturnal activity pattern of cocoa-pollinating cecidomyiids, as determined from surveys of midges in cocoa flowers, and by observing roosting habits in some species. Kaufmann (1973b) noted similar diurnal activity patterns for cocoa-associated cecidomyiids in Ghanian cocoa farms, and at least one tropical study indicates that Toxomyia (a genus close to Mycodiplosis) adults emerge in the late afternoon and early evening from their pupae (Parnell, 1969). Lucas (1981) reported ceci- domyiids laden with cocoa pollen grains being collected at or before 0700 hours
122 PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON
in cocoa farms in the Ivory Coast, and Decazy et al. (1980), using radio-isotopes as midge markers, determined two peaks of cecidomyiid activity in cocoa plan- tations in South Cameroon as 0730-0930 and 1630-1830 hours. In general cocoa flowers begin to open in the late afternoon, and gradually open all night and are fully open by 0500 hours (e.g. Wellensiek, 1932; Sampayan, 1966; Gorrez, 1962; Soetardi, 1950). Although Sampayan (1966) concludes that anther dehiscence in cocoa flowers peaks between 0600-1000 hours and stigma receptivity to pollen between 1000-1300 hours in the Philippines, similar data from Costa Rica (La Lola) indicates pollen-release throughout most of the day and night, and greatest fertilization of flowers in the early and mid-morning hours (Young et al., 1984). Given these data and the observed activity patterns of cecidomyiids in the Costa Rican studies, I propose, tentatively, that these midges pick up viable cocoa pollen grains during their dusk and evening activity periods, and deposit pollen on stigmas and pistils primarily in the dawn hours (e.g. 0530-0700 hours). Thus cocoa flowers behaviorally function as males during most of the day and evening following anthesis, and as females in the dawn hours, even though the flowers are morphologically and physiologically hemaphroditic (e.g. Walker, 1959). Differ- ential rates of sexual maturation in male and female floral parts in relation to pollinator behavior have been noted in other plant species exhibiting synchronized flowering (e.g. Philbrick, 1983; Essig, 1971; Gibbs et al., 1977).
The habit of some cocoa flower cecidomyiids forming roosting aggregates in spider webs on cocoa trees may assist in the concentration of midges near floral rewards. Although such roosts in the present study were found to contain about equal numbers of male and female midges, Kaufmann (1973b) noted twice as many females as males in Ghanian cecidomyiids in cocoa. As in the present study, Kaufmann (1973b) also noted dissolution of roosts in the evening and the for- mation of roosts in the daylight hours. Although not yet determined, roosting behavior in cecidomyiids may function primarily to conceal midges from pred- ators that forage on the trunks and branches of cocoa trees (see also Lahmann and Zuniga, 1981) as they apparently have little or no function in mating behavior (Sivinski and Stowe, 1980). As in the present study, Kaufmann (1973b) too noted that midges often roost in spider-occupied webs, suggesting little threat of pre- dation from these animals. By far, the greatest densities of midge predators in cocoa trees are seen in Anolis lizards and the mantid Chaeteessa filata. Dipterans are known to be a major component of the diet in Anolis (e.g. Floyd and Jenssen, 1983), and the high densities of these lizards in cocoa habitats like La Lola (see Andrews, 1979) suggests that considerable selection pressure is operating for ef- fective predator defense behavior in small-bodied flying insects in the same hab- itats. Aggregative behavior has been long known to be one form of protection against predators in invertebrate animals (Allee and Rosenthal, 1949 and several included references). The observation that the same sites (webs) tend to be re- used over several successive days suggests a social organization associated with roosting cecidomyiid species in cocoa habitats. Roosting in these midges, with daily recruitment of new individuals in some instances, may represent a form of “local enhancement” to advertise optimal resources within the cocoa habitat (e.g. Ward and Zahari, 1973). Because reproductive effort (i.e. numbers of eggs de- posited in the environment per unit area) in forest-dwelling dipterans can be relatively high (e.g. Fallis and Snow, 1983), predators may selectively forage on
VOLUME 87, NUMBER | 73
numerically-abundant species such as cecidomyiids. The high turnover in organic rotting materials in cocoa farms very likely promotes the maintenance of ceci- domyiid breeding populations, unlike closed laboratory systems in which enforced migrations can sometimes occur in response to contamination arising from high population densities, as shown for some fungus-dwelling Sciaridae (Binns, 1980). Preferential egg-laying behavior in cecidomyiids may allow midges to exploit a range of different microhabitats, both terrestrial and arboreal, as a means of resource partitioning as shown for other dipterans in the tropics (e.g. Siefert, 1980: Siefert and Barrera, 1981; Gagné, 1968, 1977; Carpenter, 1983; Winder, 1978b and many included references). A major component of the evolutionary associ- ation of Cecidomyiidae with cocoa trees may include the exploitation of arboreal breeding microhabitats routinely found in 7. cacao populations in the American tropics, namely, rotting holes in cocoa pods made by animals such as squirrels and woodpeckers, and epiphytic mosses.
Various studies have shown that cocoa-pollinating Ceratopogonidae exhibit wide fluctuations in population structure correlating well with seasonal patterns of rainfall (e.g. Soria and Abreu, 1976). While the drier months of the year at Costa Rican cocoa-growing localities such as La Tigra and La Lola tend to be times of reduced levels of flowering in cocoa trees (e.g. Young, 1982, 1983, 1984b) the preliminary data on abundance of adult cecidomyiids given in the present paper suggest an increase in abundance during the drier months. These data were not gathered in such a way so as to distinguish between greater abundance being an artifact of sampling, i.e. in which areas of cocoa sampled in the dry season tended to have greater accumulations of midges, or an actual indication of greater absolute abundance in populations at these times. But Kaufmann (1973b) deter- mined that cecidomyiids were more abundant in cocoa flowers during the dry season in Ghana. If the observed high densities of these midges during the drier periods is a real biological effect and not an artifact of small sample size, and if it represents a concentration of these insects into moist pockets of cocoa, there is expected to be an increased dispersion of the populations during the rainy months. Flowering tends to be greater in the wetter months at the Costa Rican localities studied (Young, 1984b) and there will be a greater floral reward for pollinators over larger areas of cocoa habitats. Because the egg-to-adult developmental time in some free-living tropical cecidomyiids is only about ten days (e.g. Parnell, 1969), many overlapping generations can occur each year. During the drier months, these midges may exploit dry substrates as breeding sites as shown by the discovery of larvae in relatively dry exocarp of cocoa pods in the dry season at Hummingbird Hershey. Particularly during the drier months, small-bodied Diptera may accu- mulate in the more shaded areas of cocoa farms in response to thermal and moisture stress, and shaded areas of cocoa habitats tend to have a greater diversity of Diptera than sunny habitats (Bigger, 1981; Gibbs and Leston, 1970; Young, 1983). Seasonally-related changes in the spatial distribution and abundance of moist or wet microhabitats used by Cecidomyiidae and other Diptera in the tropics result in changes in the availability of adult insects in breeding populations (e.g. McLachlan and Cantrell, 1980). Levels of natural pollination in cocoa are deter- mined in part by the ways in which seasonal changes in rainfall influence popu- lation cycles in natural pollinators (De la Cruz and Soria, 1973). Times of the year in which flowering is highest and synchronous within a population of a
74 PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON
tropical tree species are also times of greatest abundance of pollinators at flowers, since more pollinators are attracted to more intense floral displays (e.g. Augs- purger, 1981).
Although it was determined, by light microscopy, that cocoa flowers possess nectary-like structures on specific floral parts (Stejskal, 1969), recent studies using scanning electron microscopy reveal stomate-type nectaries on petal ligules and on the inner surfaces of petal hoods in these flowers (Young et al., 1984). These nectaries have diameters suitable to accommodate the sucking mouthparts of cecidomyiids such as Clinodiplosis sp. 1, Mycodiplosis sp. 1, Coquillettomyia sp., and Aphodiplosis triangularis. The surface of the floral ovary is blanketed with peg-like glandular structures (Stejskal, 1969; Young et al., 1984) of unknown function, but perhaps an additional component of the “pollinator reward syn- drome” of Theobroma flowers (Young et al., 1984). Cecidomyiids and cerato- pogonids, as well as wild bees (Young et al., 1984), may visit cocoa flowers and those of other Theobroma for both pollen and nectar as floral rewards. A fragrance- producing ring of glandular structures located between the sepals and petals of Theobroma flowers, and greatly reduced in faintly-scented cocoa flowers (Young et al., 1984), may constitute the mechanism by which floral visitors are attracted to flowers. Once inside petal hoods, cecidomyiids often follow the pigmented “nectar guides” leading towards the basal area of the flower, and they are perhaps also following a fragrance being produced by the basal glandular ring, albeit faint in cocoa. Similar behavior, resulting in pollination, has been noted for certain species of fungus gnats associated with 4sarum (Vogel, 1978).
Elsewhere (Young et al., 1984) the hypothesis is advanced that the basic floral design in Theobroma may have originally evolved as a bee-pollination system as most species of the genus have colorful flowers with strong pleasant fragrances, two conditions reduced in 7. cacao. It is suggested that 7. cacao which thrives well in cool, moist shaded forest understories, even in its natural state (Allen, 1981), has evolutionarily replaced bees with Diptera as frequent natural pollinators (Young et al., 1984). While some wild bees, such as 7rigona jaty (Hymenoptera: Apoidea: Meliponinae) may not directly pollinate cocoa flowers (Soria, 1975; Young, 1981), such heavy-bodied insects may sufficiently jostle pollen loose from anther sacs to facilitate eventual contact with small midges crawling through petal hoods. Pollen grains frequently lodge on the inner surface of petal hoods, adjacent to the anthers they enclose (Young et al., 1984). In the closely related genus Herrania, pollen grains adhering to the side of petal hoods are picked up by phorid flies, the natural pollinators (Young, 1984c). Still other kinds of small-bodied wild bees may directly pollinate cocoa flowers (e.g. Kaufmann, 1975c). In Costa Rica, wild bees tend to forage primarily on cocoa flowers in the sunlit upper portions of cocoa trees, while midges (cecidomyiids and ceratopogonids) are more com- monly encountered on shaded, lower branches (A. M. Young, unpubl. data). In some plant species, both dipterans and bees jointly contribute to the overall reproductive success in populations (e.g. Motten et al., 1981). When bee-pollinated plant species occur in cool, moist environments, generalized or specialist flies tend to replace bees as routine pollinating agents (e.g. Levesque and Burger, 1982; Anderson and Beare, 1983). In the American tropics, some tree species such as Hevea possess specialized associations between flowers and their midge pollinators (Steiner, 1983), and it is sometimes necessary to view a plant species as having
VOLUME 87, NUMBER | dl
nn
both “major” and “minor” sets of pollinators, often belonging to very different taxa (Lewis and Zenger, 1983). Wild bees and midges (Ceratopogonidae and Cecidomyiidae) may comprise such a constellation of pollinator sets in Theobroma and other Sterculiaceae.
ACKNOWLEDGMENTS
This research was funded by grants from The American Cocoa Research In- stitute of the United States of America. I thank J. Robert Hunter and Gustavo A. Enriquez for logistical assistance and sharing of information during the Costa Rican studies, and Gordon Patterson for similar assistance in Belize. This research would not have been possible without the assistance of Raymond J. Gagné, Sys- tematic Entomology Laboratory, IIBIII, USDA, in Washington, D.C., for making determinations of cecidomyiids and for many helpful discussions regarding these insects as pollinators. I also thank Dr. Gagné for providing me with the excellent and interesting Vogel reference. Figure | was prepared by Tammy McCarthy. I thank David W. Inouye (Department of Zoology, University of Maryland) and Michael R. Mesler (Department of Biology, Humboldt State University) for read- ing and commenting upon an earlier draft of this paper.
LITERATURE CITED
Allee, W. C. and G. M. Rosenthal, Jr. 1949. Group survival value for Philodina roseola, a rotifer. Ecology 30: 395-397.
Allen, J. B. 1981. Collecting wild cocoa at its centre of diversity. Proc. 8th Int. Cocoa Res. Conf., Cartegena, Colombia, 655-662.
Alvim, P.deT. 1977. Cacao, pp. 279-313. Jn P. de T. Alvimand T. T. Kozlowski, eds., Ecophysiology of tropical crops. New York: Academic Press.
Amponsah, J. D. 1975. Studies in natural pollination of cocoa in Ghana. Acta Horticult. 49: 223- 229.
Anderson, R. C. and M. H. Beare. 1983. Breeding system and pollination ecology of Trientalis borealis (Primulaceae). Am. J. Bot. 70: 408-415.
Andrews, R. M. 1979. Evolution of life histories: a comparison of Anolis lizards from matched island and mainland habitats. Breviora, Museum of Comp. Zool., Harvard, No. 454.
Augspurger, C. K. 1981. Reproductive synchrony ofa tropical shrub: experimental studies on effects of pollinators and seed predators on Hybanthus prunifolius (Violaceae). Ecology 62: 775-788.
Baker, R. E. D., F. W. Cope, P. C. Holliday, B. G. Bartley, and D. J. Taylor. 1953. The Anglo- Colombian cacao collecting expedition. Report on Cacao Research (Governmental), St. Au- gustine, Trinidad, B.W.I. 133-154.
Barnes, H. F. 1930. On some factors governing the emergence of gall midges (Cecidomyiidae: Diptera). Proc. Zool. Soc. London I: 381-393.
Baumann, E. 1978. Rennfliegen (Diptera: Phoridae) als Bliitenbesucher. Kritische Sicht. der Literatur. Flora 167: 301-314.
Bigger, M. 1981. Observations on the insect fauna of shaded and unshaded Amelonado cocoa. Bull. Entomol. Res. 71: 107-119.
Billes, D. J. 1941. Pollination of Theobroma cacao L. in Trinidad, B.W.I. Trop. Agric. 18: 151- 156.
Binns, E.S. 1980. Field and laboratory observations on the substrates of the mushroom fungus gnat Lycoriella auripila (Diptera: Sciaridae). Ann. Appl. Biol. 96: 143-152.
Brewer, J. M. 1981. The influence of light on circadian emergence patterns of Monarthropalpus buxi (Diptera, Cecidomyiidae). Acta Entomol. Bohemoslov. 78: 152-161.
Brown, V. K. and G. C. McGavin. 1982. The biology of some mine and gall forming Diptera on bracken, Pteridium aquilinum (L.) Kuehn. J. Nat. Hist. 16: 511-518.
Bystrak, P. G. and W. W. Wirth. 1978. The North American species of Forcipomyia, subgenus Euprojoannisia (Diptera: Ceratopogonidae). U.S. Dept. Agric. Tech. Bull. No. 1591.
76 PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON
Carpenter, S. R. 1983. Resource limitation of larval treehole mosquitoes subsisting on Beech detritus. Ecology 64: 219-223.
Charlwood, J. D., J. Lopes, and P. C. Whalley. 1982. Light intensity measurement and the biting behaviour of some sylvatic mosquitoes of the Amazon Basin (Diptera: Culicidae). Acta Amazon. 12: 61-64.
Chiang, H.C. 1968. Ecology of insect swarms. V. Movement of individual midges Anarete pritchardi, within a swarm. Ann. Entomol. Soc. Am. 61: 584-587.
Cope, F. W. 1939. Agents of pollination in cacao. Cacao Res. 1939: 13-19.
Cuatrecasas, J. 1964. Cacav and its allies—a taxonomic revision of the genus Theobroma. Contrib. U.S. Nat. Mus. 35: 379-614.
Decazy, B., F. Massaux, C. Tchiendju, and C. Misse. 1980. Use of radioisotopes (32P) in the study of pollinating agents of the cacao tree in the ecological conditions of South Cameroon. Proc. 8th Int. Cocoa Res. Conf., Douala, Cameroon.
De La Cruz, J. and S. de J. Soria. 1973. Estudio de fluctuaciones de polinizacion del cacao por las mosquitas Forcipomyia spp. (Diptera, Ceratopogonidae), en Palmira, Valle, Colombia. Acta Agronom. 23: 1-17.
Dessart, P. 1961. Contribution a l’étude des Ceratopogonidae (Diptera). Bull. Agric. Congo Belge 52: 525-540.
Edwards, D. F. 1973. Pollination studies on Upper Amazon cocoa clones in Ghana in relation to the production of hybrid seed. J. Hort. Sci. 48: 247-259.
Ehler, L. E. 1982. Ecology of Rhopalomyia californica Felt (Diptera: Cecidomyiidae) and its parasites in an urban environment. Hilgardia 50: 1-32.
Entwistle, P. F. 1972. Pests of cocoa. London: Longmans.
Essig, F. B. 1971. Observations on pollination in Bactris. Principes 15: 20-24.
Faegri, K. and L. van der Pijl. 1971. Principles of pollination ecology. New York: Pergamon Press.
Fallis, S. P. and K. R. Snow. 1983. Distribution of the eggs of Aedes punctor (Diptera: Culicidae). Ecol. Entomol. 8: 139-144.
Fisher, R. W. and G. L. Teetes. 1982. Effect of moisture on sorghum midge (Diptera: Cecidomyiidae) emergence. Env. Entomol. 11: 946-948.
Floyd, H. B. and T. A. Jenssen. 1983. Food habits of the Jamaican lizard Anolis opalinus: resource partitioning and seasonal effects examined. Copeia 1983: 319-331.
Fontanilla-Barroga, S. 1962. A progress report on the study of insects associated with pollination of Theobroma cacao with special emphasis on midges. Philippine J. Agric. 27: 147-159.
Gagné, R. J. 1968. A catalogue of the Diptera of the Americas south of the United States. 23. Family Cecidomyiidae. Dept. Zool. Secret. Agric. Sao Paulo, Brazil 23: 1-62.
—. 1977. The Cecidomyiidae (Diptera) associated with Chromolaena odorata (L.) K.&R. (Com- positae) in the Neotropical Region. Brenesia 12/13: 113-131.
—. 1984. Five new species of neotropical Cecidomyiidae (Diptera) associated with cacao flowers, killing the buds of Clusiaceae, or preying on mites. Brenesia 22: 123-138.
Gibbs, D. G. and D. Leston. 1970. Insect phenology in a forest cocoa-farm locality in West Africa. J. Appl. Ecol. 7: 519-548.
Gibbs, P. E., J. Semir, and N. Diniz da Cruz. 1977. Floral biology of Ta/auma ovata St. Hil. (Magnoliaceae). Cienc. Cult. 29: 1436-1441.
Gilbert, L. E. 1972. Pollen feeding and reproductive biology of Heliconius butterflies. Proc. Nat. Acad. Sci., U.S.A. 69: 1403-1407.
Gilbert, F. S. 1981. Foraging ecology of hoverflies: morphology of the mouthparts in relation to feeding on nectar and pollen in some common urban species. Ecol. Entomol. 6: 245-262.
Glendenning, D. R. 1962. Natural pollination of cacao. Nature 193: 1305.
Gnanaratnam, J. K. 1954. Pollination mechanism of the cacao flower. Trop. Agric. 31: 98-104.
Gorrez, J. 1962. The flower biology, morphology, and pollinating and crossing habits of cacao. Philippine Agric. 46: 288-302.
Hernandez, J. 1965. Insect pollination of cacao (Theobroma cacao L.) in Costa Rica. Doct. Dissert., Univ. of Wisconsin, Madison.
Hutcheon, W. V. 1981. Physiological studies on cocoa (Theobroma cacao L.) in Ghana. Doct. Dissert., Univ. of Aberdeen, Scotland.
Jones, G. A. 1912. The structure and pollination of the cacao flower. West Indian Bull. 12: 347- 350.
VOLUME 87, NUMBER 1 77
Kaufmann, T. 1973a. Behavioural ecology of a cocoa pollinator Forcipomyia inornatipennis (Diptera: Ceratopogonidae) in Ghana. J. Kansas Entomol. Soc. 47: 541-548. 1973b. Preliminary observations on cecidomyiid midge and its role as a cocoa pollinator in
Ghana. Ghana J. Agric. Sci. 6: 1093-1098.
1974. Pollinators of Theobroma cacao in Ghana. Proc. 4th Int. Cocoa Res. Conf., West
Africa.
—. 1975a. Cocoa pollination by males of Forcipomyia squamipennis (Diptera: Ceratopogonidae)
in Ghana. Trop. Agric. 52: 71-74.
. 1975b. Biology and behaviour of cocoa pollinating Ceratopogonidae in Ghana. Env. Entomol.
4: 347-351.
1975c. An efficient, new cocoa pollinator, Lasioglossum sp. (Hymenoptera: Halictidae) in
Ghana, West Africa. Turnalba 25: 90-91.
Kay, B. H. 1983. Collection of resting adult mosquitoes at Kowanyaince, Northern Queensland and Charlesville, South-Western Queensland. J. Aust. Entomol. Soc. 22: 19-24.
Kevan, P. and H. G. Baker. 1983. Insects as flower visitors and pollinators. Ann. Rev. Entomol. 28: 407-453.
Lahmann, E. J. and C. Ma. Zuniga. 1981. Use of spider threads as resting places by tropical insects. J. Arachnol. 9: 339-341.
Leston, D. 1969. Phenology of cocoa and some associated insects in Ghana. Proc. 3rd Int. Cocoa Res. Conf., Accra 197-204.
Levesque, C. M. and J. F. Burger. 1982. Insects (Diptera, Hymenoptera) associated with Minuartia groenlandica (Caryophyllaceae) on Mount Washington, New Hampshire, U.S.A., and their possible role as pollinators. Arct. Alp. Res. 14: 117-124.
Lewis, W. H. and V. E. Zenger. 1983. Breeding systems and fecundity in the American Ginseng, Panax quinquefolium (Araliaceae). Am. J. Bot. 70: 466-468.
Lucas, P. 1981. Etude des conditions de pollinisation du cacaoyer au Togo. Café Cacao Thé 25: 113-120.
Lumsden, W. H.R. 1952. The crepuscular biting activity of insects in the forest canopy at Bwamba, Uganda. A study in relation to the Sylvan epidemiology of yellow fever. Bull. Entomol. Res.
42: 721-760.
Mamaevy, B. M. 1975. Evolution of gall forming insects—gall midges. Leeds: The British Library, Maney Ltd.
McLachlan, A. J. and M. A. Cantrell. 1980. Survival strategies in tropical rain pools. Oecologia 47: 344-351.
Mesler, M. R., J. D. Ackerman, and K. L. Lu. 1980. The effectiveness of fungus gnats as pollinators. Am. J. Bot. 67: 564-567.
Mogal, B. H., A. R. Mali, S. G. Rajput, and K. L. Pawar. 1979. Emergence of sorghum midge (Contarinia sorghicola Coq.) from diapausing larvae. J. Maharashtra Agric. Univ. 4: 323.
Motten, A. F., D. R. Campbell, D. E. Alexander, and H. L. Miller. 1981. Pollination effectiveness of specialist and generalist visitors to a North Carolina population of Claytonia virginica. Ecology 62: 1278-1287.
Myers, J. G. 1930. Notes on wild cacao in Surinam and British Guiana. Kew. Bull. 1930: 1-10.
Parnell, J. R. 1969. The biology and morphology of all stages of Toxomyia fungicola Felt (Diptera: Cecidomyiidae) in Jamaica. Proc. R. Entomol. Soc. London (A) 44: 113-122.
Percival, M. 1965. Floral biology. London: Pergamon Press.
Petralia, R. S., A. L. Wuensche, G. L. Teetes, and A. A. Sorenson. 1979. External morphology of the mouthparts of larvae of sorghum midge, Contarinia sorghicola. Ann. Entomol. Soc. Am. 72: 850-855.
Philbrick, C. T. 1983. Contributions to the reproductive biology of Panax trifolium L. (Araliaceae). Rhodora 85: 97-113.
Posnette, A. F. 1942. Natural pollination of cocoa, Theobroma leiocarpa Bern. on the Gold Coast.
Il. Trop. Agric. 19: 188-191.
1944. Pollination of cacao in Trinidad. Trop. Agric. 21: 115-118.
Privat, F. 1979. Les Bromeliacees, lieu de développement de quelques insectes pollinisateurs des fleurs de cacao. Brenesia 16: 197-212.
Sampayan, T.S. 1966. Flower biology, fruiting habit and compatibility relationship in cacao. Phil- ippine J. Plant Indust. 31: 193-201.
78 PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON
Saunders, L. G. 1959. Methods for studying Forcipomyia midges, with special reference to cacao- pollinating species (Diptera, Ceratopogonidae). Can. J. Zool. 37: 33-51.
Shazli, A. and T. M. Mostafa. 1980. Frequency of Thomasiniana oleisuga Targ. (Dipt., Cecido- myiidae) and its parasites and predators in Amman, Jordan. Z. Angew. Entomol. 89: 269-277,
Siefert, R. P. 1980. Mosquito fauna of Heliconia aurea. J. Anim. Ecol. 49: 687-697.
Siefert, R. P. and R. Barrera. 1981. Cohort studies on mosquito (Diptera: Culicidae) larvae living in water-filled floral bracts of Heliconia aurea (Zingiberales: Musaceae). Ecol. Entomol. 6: 191- 197.
Sivinski, J. and M. Stowe. 1980. A kleptoparasitic cecidomyiid and other flies associated with spiders. Psyche 87: 337-348.
Soderstrom, T. R. and C. E. Calderon. 1971. Insect pollination in tropical rain forest grasses. Biotropica 3: 1-16.
Soetardi, R. G. 1950. De Betekenis van Insecten Bij Bestuiving van Theobroma cacao L. Arch. Kofhecult. Indonesie 17: 1-31.
Soria, S. de J. 1975. O papel das abelhas sem ferrao (Meliponinae) na polinizacao do cacaueiro na América tropical. Rev. Theobr. (Brasil) 5: 12-20.
Soria, S. de J. and J. M. Abreu. 1976. Dinamica populacional de Forcipomyia spp. (Diptera, Cer- atopogonidae) na Bahia, Brasil. 1. Fluctuacao estacional dos polinizadores do cacaueiro rela- cionada com chuva e balanco de agua (Thornthwaite). Rev. Theobr. (Brasil) 6: 47-54.
Soria, S. de J., W. W. Wirth, and R. K. Chapman. 1980. Insect pollination of cacao in Costa Rica. 1. Preliminary list of the ceratopogonid midges collected from flowers. Rev. Theobr. (Brasil) 10: 61-68.
Soria, S. de J., R. K. Chapman, J. K. Knoke. 1981. Cacao pollination in Costa Rica. 2. Breeding sites of ceratopogonid (Diptera, Nematocera) midges. Rev. Theobr. (Brasil) 11: 119-123.
Steiner, K. E. 1983. Pollination of Mabea occidentalis (Euphorbiaceae) in Panama. Syst. Bot. 8: 105-117.
Stejskal, M. 1969. Nectar y aroma de las flores del cacao. Oriente Agropeuc. 1: 75-92.
Summers, C. G. 1975. Daily adult emergence in the sorghum midge, Contarinia sorghicola. Env. Entomol. 4: 495-498.
Taylor, A. S. 1965. Studies in the pollen morphology of Theobroma and Herrania. Cacao 10: 1-9.
Van Der Knapp, W. P. 1955. Observations on the pollination of cacao flowers. Rep. 14th Int. Hort. Congr. 2: 1287-1293.
Vanhara, J. 1981. Lowland forest Diptera (Brachycera, Cyclorrhapha). Acta Sci. Nat. Acad. Sci. Bohemoslov. Brno. 15: 1-32.
Vello, F. and W. S. Magalhaes. 1971. Studies on the role of the cacarema ant (Azteca chartifex spiriti Forel) on the pollination of cacao in Bahia. Rev. Theobr. (Brasil) 1: 29-42.
Vogel, S. 1978. Pilzmiickenblumen als Pilzmimeten. Flora 167: 329-366; 367-398.
Walker, C. 1959. Notes on the botany of cocoa. Agric. J. Fiji 29: 56-61.
Ward, P. and A. Zahari. 1973. The importance of certain assemblages of birds as “information centres” for food-finding. Ibis 1973: 517-537.
Wellensiek, S. J. 1932. Flower-biological observations with cocoa. (translat. from Dutch). Arch. Koffeekult 6: 87-101.
Willmer, P. G. 1982. Hygrothermal determinants of insect activity patterns: the Diptera of water- lily leaves. Ecol. Entomol. 7: 221-231.
Winder, J. A. 1972. Cacao pollination: Microdiptera of cacao plantations and some of their breeding places. Bull. Entomol. Res. 61: 651-655.
—. 1977a. Some organic substrates which serve as insect breeding sites in Bahian cocoa plan- tations. Rev. Brasil. Biol. 37: 351-356.
—. 1977b. Field observations on Ceratopogonidae and other Diptera: Nematocera associated with cocoa flowers in Brazil. Bull. Entomol. Res. 67: 57-63.
——. 1978a. Cocoa flower Diptera: their identity, pollinating activity and breeding sites. PANS 24: 5-18.
—. 1978b. The role of non-dipteran insects in the pollination of cocoa in Brazil. Bull. Entomol. Res. 68: 559-574.
Winder, J. A. and P. Silva. 1975. Current research on insect pollination of cacao in Bahia. Proc. 4th Int. Cocoa Res. Conf., Trinidad & Tobago, 553-565.
Wirth, W. W. and W. T. Waugh. 1976. Five new Neotropical Dasyhelea midges (Diptera: Cerato- pogonidae) associated with culture of cocoa. Stud. Entomol. 19: 223-236.
VOLUME 87, NUMBER | 79
Young, A. M. 1981. The ineffectiveness of the stingless bee, 7rigona jaty (Hymenoptera: Apidae: Meliponinae) as a pollinator of cocoa (Theobroma cacao L.). J. Appl. Ecol. 18: 149-155.
1982. Effects of shade cover and availability of midge breeding sites on pollinating midge populations and fruit set in two cocoa farms. J. Appl. Ecol. 19: 47-63. —. 1983. Seasonal differences in abundance and distribution of cocoa-pollinating midges in relation to flowering and fruit set between shaded and sunny habitats of the La Lola Cocoa Farm in Costa Rica. J. Appl. Ecol. 20: 801-831.
1984a. Mechanism of pollination by Phondae (Diptera) in some Herrania species (Stercu- liaceae) in Costa Rica. Proc. Entomol. Soc. Wash. 86: 503-518.
1984b. Flowering and fruit-setting patterns of cocoa trees (Theobroma cacao L.) (Sterculi- aceae) at three localities in Costa Rica. Turrialba 34: 129-142. Young, A. M., M. Schaller, and M. Strand. 1984. Floral nectaries and trichomes in relation to
pollination in some species of Theobroma and Herrania (Sterculiaceae). Am. J. Bot. 71: 466-
480.
PROC. ENTOMOL. SOC. WASH. 87(1), 1985, pp. 80-84
MAYACNEPHIA FORTUNENSIS (DIPTERA: SIMULIIDAE), A NEW BLACK FLY SPECIES FROM PANAMA
JOHN L. PETERSEN
Gorgas Memorial Laboratory, Panama, Republic of Panama. Correspondence: Box 935, APO Miami, Florida 34002.
Abstract. —The larva, pupa, male and female of Mayacnephia fortunensis, new species, are described and illustrated. This report extends the distribution of the genus from the western United States, Mexico and Guatemala, southward to include Panama.
Field studies related to a black fly pest problem at the Fortuna Hydroelectric Project, Chiriqui Province, Panama, revealed 10 simuliid species (Petersen et al., 1983). Nine of these belong to the genus Simu/ium; the tenth is described here as a new species of the genus Mayacnephia Wygodzinsky and Coscaron (1973). Nine species of Mayacnephia have been described, ranging from western U.S.A. south to the highlands of Mexico and Guatemala. The species described below represents the first member of this genus reported from Panama. Terminology follows Peterson (1981).
Mayacnephia fortunensis Petersen, NEW SPECIES Figs. 1-14
Female.— Wing length 3.2 mm.
Head: Dichoptic; frontal angle 50-55° (Fig. 5). Antenna about 0.7 mm long; scape and pedicel light brown; flagellum brown with nine flagellomeres. Palpus dark brown. Sensory vesicle larger than in male, elongate with neck distad of vesicle midline. Frons and clypeus brown, with white pruinosity, irregularly clothed with yellow and black hairs. Postocciput brown, densely clothed with yellow hairs. Distal cornua of cibarium slender, sharp-pointed, heavily sclerotized at ends, median distal space hyaline and smooth.
Thorax: Scutum dark brown, densely covered with yellow hairs; prescutellar region with a few long fine black hairs. Postpronotum and proepisternum with yellow hairs. Scutellum light brown, with long black hairs and short yellow hairs. Postnotum glabrous, with gray pruinosity. Pleural sclerites glabrous, with gray pruinosity. Wings, 3.2 mm long and |.5—1.6 mm wide; Sc sparsely pilose ventrally; R, pilose dorsally with spines along distal third; R, completely pilose; CuA, and A, arcuate; basal cell present; A, ending before wing margin (Fig. |). Legs brown except for tarsi which are light brown; calcipala well-developed, pedisulcus absent. Claws with large, blunt, subquadrate, basal tooth (Fig. 3).
Terminalia: Spermatheca somewhat kidney shaped, sclerotized, with at most
VOLUME 87, NUMBER 1 81
JR E
2
4 1 mm
Figs. 1-7. Mayacnephia fortunensis. Figs. 1-5, Female. 1, Wing. 2, Genital fork. 3, Claw. 4, Spermatheca. 5, Portion of head showing frons and clypeus. Figs. 6-7, Male. 6, Ventral plate, median sclerite and paramere, ventral view. 7, Gonostylus, ventral view.
a slight membranous area at junction with spermathecal duct, without internal spicules; spermathecal duct not sclerotized (Fig. 4). Genital fork without lateral, anteriorly directed apodemes (Fig. 2).
Male.—General body color dark brown. Length: body, 2.1 mm (dry pinned specimens), 2.3 mm (alcohol preserved specimens); wing, 2.5 mm.
Head: Holoptic; frons and clypeus brown with silvery pruinosity; clypeus with pale yellowish to brownish hairs. Antenna with scape, pedicel, and base of Ist flagellomere dull yellow, rest of flagellomeres light brown. Palpus dark brown. Sensory vesicle globular and smaller than in female.
Thorax: Scutum dark brown, with short yellow hairs distributed evenly over entire surface, with long black hairs near scutellum, pleural sclerites glabrous, with grayish pruinosity. Scutellum light brown, with long black hairs and short yellow hairs, ventral portion with faint silvery pruinosity. Halter light brown. First abdominal tergite dark brown, with very long yellow hairs appearing brown-
82 PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON
13 1mm
: 9 1mm
KE kOe 5 mm Imm
Figs. 8-14. Mayacnephia fortunensis. Figs. 8-11, Pupa. 8, Respiratory filaments. 9, Variation in respiratory filament number |. 10,