16 July 1998 PROC. ENTOMOL. SOC. WASH. 100(3), 1998, pp. 467-483

SYSTENA CHEVROLAT (COLEOPTERA: CHRYSOMELIDAE: ALTICINAE): NOTES ON NOMENCLATURE, REDESCRIPTION OF THE GENUS, AND A PRELIMINARY DISCUSSION OF CHARACTERS AND PHYLOGENETIC RELATIONSHIPS

STEVEN W. LINGAFELTER, ALEXANDER S. KONSTANTINOV, AND JONG EUN LEE

(SWL, ASK) Systematic Entomology Laboratory, PSI, Agricultural Research Service, U.S. Department of Agriculture, c/o National Museum of Natural History, MRC-168, Washington, DC 20560 U.S.A. (email: [email protected]; akonstanti­ [email protected]); (JEL) Department of Biology, College of Natural Sciences, An­ dong National University, Andong, Kyungbuk 760-749, South Korea.

Abstract.-A summary of the taxonomic history for Systena Chevrolat (Coleoptera: Chrysomelidae: Alticinae) is presented. Detailed discussion and illustration of morpho­ logical characters are presented based on stereo, compound, and scanning electron mi­ croscopy of dissected exemplar specimens for eight species of Systena from North, Central, and South America. A discussion of putative relationships of Systena to other alticine genera is presented.

Key Words: Systena, Alticinae, Chrysomelidae, adult, morphology, character, nomencla­ ture, systematic, flea beetle, leaf beetle

To help provide a foundation for future TAXONOMIC HISTORY generic-level phylogenetic studies in Al­ Systena was proposed by Chevrolat ticinae, a nomenclatural discussion and (1836:390). Heikertinger and Csiki (1939) detailed morphological redescription with and Blackwelder (1946) incorrectly attrib­ preliminary phylogenetic discussion are ute authorship of Systena to Melsheimer provided for Systena. Although the genus (1847), who indicated Chevrolat as author Systena was proposed in 1836, the first of Systena. In Melsheimer (1847), one ad­ discussion of morphological characters of ditional species, Systena blanda, was de­ the genus was presented by Clark (1865). scribed, but no other species were listed. However, as is typical of many early Chevrolat (1848) stated that Systena was works, the characters he used were super­ proposed by "us" (meaning himself) and ficial and not extremely useful consider­ adopted by Dejean. In this work, Systena ing the paucity of knowledge of character s-littera (Linnaeus) and S. vittata (Fabri­ distribution among Alticinae on a world­ cius) were mentioned as additional mem­ wide level. When characters are viewed in bers of the genus. Clark (1865) in his ex­ this context, as we aspire to do here, their amination of South American Alticinae pro­ value as diagnostic and evolutionary in­ vided the first characterization of Systena, dicators is greatly enhanced. and probably for this reason, Gemminger 468 PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON

Table 1. Character state information for the eight Systena species used for the generic characterization. Character numbers refer to text. Symbol, '' +'' indicates that both states present or too difficult to interpret.

'illTaxa Characters-+ 5 10 15 20 23

Systena blanda (Melsheimer) 00101 00030 010+0 +ODDO 000 Systena carri Blake 00001 00020 01000 10??0 ??? Systena championi Jacoby 00001 01020 01000 00010 000 Systena elongata Fabricius 00000 00020 01000 00010 000 Systena frontalis (Fabricius) 00001 00030 01000 00110 000 Systena marginallis (Illiger) 00001 00020 01000 00000 000 Systena oberthuri Baly +0001 00030 01000 +0000 001 Systena s-littera (Linnaeus) 00001 00020 01000 +ODDO 000

and von Harold (1874), Hatch (1927), Hom erthuri Baly, and S. s-littera (Linnaeus). (1889), and Leng (1920) considered him to These taxa which occur in North, Central, be the author of the genus. However, the and South America are suggested to repre­ mere use of one or more available species­ sent the spectrum of morphological diver­ group names in combination or clearly in­ sity for Systena. Characters, many of which cluded under a new genus-group name (as are described for the first time in Chryso­ done by Chevrolat 1836), is sufficient for melidae, are discussed in detail below. A availability for names published before subset of the characters and states is coded 1931 (ICZN [1985] Article 12 b-5; sum­ in Table 1 to explicitly indicate variability marized in Barber and Bridwell 1940). Oth­ or constancy within Systena. The appendix er authors who also correctly attribute to provides a brief description of the charac­ Chevrolat authorship of Systena include ters and states used in Table 1. Chapuis (1875), Crotch (1873), Seeno and Characters of adult exemplars were ex­ Wilcox (1982), and Poole and Gentili amined using a Leica MZ-APO stereo­ (1996). Systena frontalis (Fabricius) was scope, Leitz Diaplan compound microscope designated as the type species by Monr6s fitted with transmitted light interference and Bechyne (1956). contrast apparati (ICT), and an AMRAY 1810 scanning electron microscope. To ease METHODS examination of sclerotized features, adult The discussion of the adult characters specimens were partially disarticulated and and terminology follows that of Konstanti­ placed in hot KOH for several seconds. Ste­ nov and Lopatin ( 1987), Konstantinov reo and compound microscopic illustrations (1994) and Konstantinov and Vandenberg were made using camera lucidas. Images (1996). Hind wing terminology follows Ku­ were scanned at 600 dpi using a Nikon kalova-Peck and Lawrence (1993, Fig. 36) Scantouch scanner and enhanced using since they provided the most explicit hy­ Adobe Photoshop 3.0 and labeled using potheses of homology based on examina­ Macromedia Freehand 5.5 on a Power Mac­ tion of all beetle families. Metafemoral intosh 8500/132 with 50 megabytes RAM. spring terminology follows Furth (1982). Many characters are described here for the REDESCRIPTION OF SYSTENA CHEVROLAT first time. Morphological characters are Body (Fig. 1): Moderately elongate, flat based on at least two examples of dissected, in lateral view. Body length/width ratio as disarticulated specimens of Systena blanda measured by length of elytra divided by (Melsheimer), S. carri Blake, S. championi greatest width, ranging from 1.42 (S. s-lit­ Jacoby, S. elongata Fabricius, S. frontalis tera)-1.91 (S. oberthuri). Elytron without (Fabricius), S. marginallis (Illiger), S. ob- apical declivity. VOLUME 100, NUMBER 3 469

Fig. 1. Systena blanda, male, dorsal view.

Head (Figs. 1, 7, 8): Antenna! calli pres­ tinct on anterior margin of antenna! calli but ent, well developed, each as large as anten­ otherwise shallow and not differentiated na! socket, longer than wide, extending to from surrounding area; not continuing be­ antenna! sockets. Supra-antennal sulcus dis- yond antenna! socket. Supraorbital sulcus 470 PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON shallow, not well-differentiated from sur­ constricted at apical two-thirds, becoming rounding area. Supracallinal sulcus varies cone-shaped apically. Second antennomere from well developed in S. frontalis to poor­ shorter than third, longer than fourth. Fifth ly developed in S. blanda. Anterofrontal antennomere longer than fourth and sixth. ridge with distinct, narrow crest, lower than Mesal margins of antennal sockets farther frontal ridge, contacting frontal ridge an­ apart than diameter of antennal socket (Fig. teapically (before anterior end). Orbital sul­ 7). cus absent (when present in other genera, Thorax: Pronotum (Fig. 1) without pos­ extends from dorsal end of the supraorbital terior (prebasal) transverse impression and sulcus to dorsal margin of eye); dorsal mar­ without postero-lateral longitudinal impres­ gin of eye not demarcated from adjacent sions. Anterolateral and posterolateral cor­ vertex. Frontal ridge well developed, ex­ ners of pronotum each bearing one long tending between antenna! calli to clypeus seta. Procoxal cavities closed (apex of in­ (declivous between antennal calli), not tercoxal prostemal process contacting hy­ higher than rim of antennal socket when pomeral projection). Intercoxal prostemal viewed from antero-ventral perspective process strongly expanded apically (at least along plane of labrum. Frontal-lateral sul­ 2.0 times as wide at apex as narrowest re­ cus poorly developed; region between fron­ gion between procoxae). Mesocoxa (Fig. tal ridge and antennal socket not well dif­ 32) broadly open laterally to mesepimeron ferentiated. Subantennal sulcus absent. Sub­ (mesostemum and metastemum not con­ genal suture well-developed. Tentorium in­ tacting one another on the outer lateral mar­ complete, with only posterior arms gin of the mesocoxa). Intercoxal mesoster­ extending dorsally less than one-fifth of dis­ nal process (Fig. 32) bilobed. Mesonotum tance from floor of the head to the vertex. (Fig. 34) strongly sclerotized and lacking Occipital opening nearly evenly oval, not stridulatory ridges. Mesoprescutum (scutel­ subquadrate. lum, Fig. 34) is generally triangular, but Mouthparts (Figs. 2-5): Labrum (Fig. 2) more truncate posteriorly in S. blanda and with submarginal row of fine sensilla on in­ S. s-littera. Metendosternite (Figs. 37, 38) ner (ventral) surface; four long sensilla on with apex of anterior arms defiexed and ta­ outer (dorsal) surface; two short setae on pering to a point; with long, narrow, dorsal mid-apical margin on inner (ventral) sur­ ridge from anterior midpoint to posterior face; and with erratic pattern of short, pitch­ midpoint (Fig. 37); with poorly developed er-shaped sensilla on inner (ventral) sur­ tendons of metafurcal-mesofurcal muscles; face. Prementum (Fig. 3) with four setae with poorly developed ventral process (Fig. (the posterior pair approximately the length 38) without visible contact to ventral pro­ of basal two labial palpomeres); base be­ jection; with poorly developed transverse tween labial palpi evenly rounded, not connection from anterior arm into middle mildly acute. Apex of penultimate labial region. Metanotum (Fig. 33) less than 1.5 palpomeres with two long setae, the longest times as wide as long. Metanotal ridge e of which extends at least to apex of the api­ and c2 fused, forming ridge c, thus appear­ cal palpomere. Apical maxillary palpomere ing thickened at region where both con­

(Fig. 5) about 1.5 times as long as penul­ verge with b 1 (terminology following Kon­ timate palpomere, with small, basal, quad­ stantinov, in press). Metanotal ridge d in­ rate digitiform sensilla patch. Mandible tersecting c at point one-third of e's length (Fig. 4) with five teeth evenly distributed before posterior end, distinctly posterior to along inner (ventral) and outer (dorsal) sur­ midpoint on c (in S. blanda, d intersects c face. Mandibular teeth unevenly dull-ser­ at its midpoint). Metanotal ridge b 1 inter­ rate. secting a below the median groove. Antenna (Fig. 6): Apical antennomere Wings and legs: Elytral punctation irreg- VOLUME 100, NUMBER 3 471

) / '\ / / ••• \ °'. !'~I.;\ ''1 ~I· i y ''{

)1 ,~· I I J) ' I I 2 c::,/

I I. r .,,,1 ,. \ \ \ ...

4

Figs. 2-6. Adult anatomy of Systena blanda. 2, Labrum, dorsal (outer) view. 3, Labium, posterolateral view. 4, Right mandible, ventral (inner) view. 5, Right maxilla, dorsal (outer) view. 6, Antenna. Abbreviations: ds = digitiform sensillum; pm = prementum; ps = pitcher sensilla; sms = submarginal sensilla. 472 PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON

dm

di

Figs. 7-10. Adult anatomy of Systena blanda. 7, Head, anterolateral view. 8, Head, lateral view. 9, Left metafemur, posterior view. IO, Metafemoral spring, posterior view, dorsal side oriented to left. Abbreviations: ac = antenna! callus; afr = anterofrontal ridge; de = declivous margin; dl = dorsal lobe; dls = dorsolateral apical spines; dm = dorsal margin; dms = dorsomesal apical spines; ea = extended arm; fis = frontolateral sulcus; fr = frontal ridge; rf = recurved flange; sas = supraantennal sulcus; sgs = subgenal suture; sos = supraorbital sulcus; vi = ventral lobe. ular, confused arrangement (Figs. 1, 35). rounded (Fig. 1, 35). Elytral epipleuron Ventrally elytra with two small, separate subvertical, continuous nearly to apex. Hind binding patches not extending beyond basal wing CuA2 not attached to CuA in most half of elytron (Fig. 35). Elytral apices species. In others such as S. blanda, S. car- VOLUME 100, NUMBER 3 473 ri, S. oberthuri, and S. s-littera, a dark pig­ study on developmental segmentation needs mentation appears to connect the CuA2 to to be done. [Cox (1996) showed variability the CuA, this indicated by a "+" in Table in numbers of tergites on pupae within Al­ 1. Hind wing AA unbranched and connect­ ticinae but this variability was not dis­ ed to CuA3 +4 at about half the distance from cussed.] Tergite VII (Fig. 39) base of fe­ the origin of CuA. Hind wing RP-MP2 not male with dense, rnicrotrichia of two types: reaching R4 • Hind wing venation not de­ (a) compound (with 3-6 parallel spicules of veloped beyond basal 213 of wing. Dorsolat­ differing lengths) and (b) spiniform (regu­ eral apex of metatibiae with row of 22-30 lar, ordered distribution, with short spicule small spines (Figs. 1, 9). Dorsomesal apex originating from apex of triangular cuticular of metatibiae with row of 15-25 long, thick plate. Tergite VII apex of female with nu­ spines (Fig. 9). Dorsal surface of metatibiae merous long, tapering setae, evenly distrib­ convex at basal %. Apical 14 nearly flat with uted. Apex with about 10-15 longer, api­ small convexity near middle. Metafemur cally-curved setae (Fig. 39). Sternite VII (Fig. 9) moderately broad (l.8-2.2 times (fifth visible sternite) (Fig. 40) of female longer than wide). First metatarsus moder­ with about 12-16 long, apically-curved se­ ately long, approximately V:3 length of me­ tae along apical margin. Otherwise with tatibia (Fig. 9). Tarsal claws appendiculate. regularly distributed normal setae. Sper­ Third metatarsomere bilobed. Metafemoral matheca (Fig. 43) with short, abruptly spring (Figs. 11-31) with variably devel­ curved pump (apical region). Spermathecal oped recurved flange; elongated arm por­ duct (basal, sclerotized region, fig. 43) com­ tion of dorsal lobe ranging from 0.25 to plex in shape, twisted. Spermathecal gland 0.35 length of metafemoral spring; dorsal base extending nearly one-half length of margin of dorsal lobe variable in curvature; sclerotized portion of spermathecal duct dorsal margin of the ventral lobe varies in (basal region). Spermathecal gland surface the angle to the dorsal-ventral axis of the micro-convoluted and irregular; nearly metafemoral spring; tightness of the coiling length of spermatheca. Vaginal palpus (Fig. of the dorsal and ventral lobes variable. 42) with 6-8 fleshy setae. Vaginal palpi Abdomen and genitalia: Apex of sternite slightly divergent at base, nearly contiguous VII of males with median truncate exten­ at apex. Ventral, anterior sclerotization of sion bordered by two apico-marginal im­ vaginal palpi slender, longer than posterior pressions. Aedeagus from lateral view (Fig. sclerotization, nearly as long as posterior 45) with only slight concavity along ventral part of palpi behind point of their connec­ profile; slight convexity along dorsal pro­ tion. Tignum apex (Fig. 41) (sternite VIII) file. Aedeagus apex arrow-shaped with fine unmodified, with apical margin of short se­ striae at extreme apex (Fig. 46). Aedeagus tae. Tergite VIII (Fig. 41) of female with apex dorsally with broad, flat, raised area darkly pigmented lateral margins and an extending anteriorly to narrow, raised ridge apical margin with row of short setae. In A. (Fig. 44). Central part of median lobe dor­ frontalis, this tergite is moderately and sally with broad ridge, narrowing anterior­ evenly sclerotized and spoon shaped. ly, and margined on either side by long nar­ row ridge (Fig. 44). Aedeagus from ventral DIAGNOSTIC CHARACTERS view with basal opening slightly constricted The following characters are apparently posteriorly; posterior opening narrower unique (autapomorphies) for Systena: sper­ than at anterior end (Fig. 46). mathecal duct irregulary shaped with sev­ Female abdomen with seven visible ter­ eral widened and curved regions (Fig. 43); gites and seven pairs of spiracles. The last basal, sclerotized portion of spermathecal tergite, tergite VII, may be homologous gland long and roughly parallel sided (Fig. with tergite VIII of other beetles, but more 43); and metanotal ridged intersecting c at 474 PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON

Figs. 11-31. Right metafemoral springs of Systena spp. Dorsal lobes are oriented to the top except in the proximal end views in which the dorsal lobe is generally oriented to the top right. White bar represents 100 micrometers. 11, S. blanda, dorsal (posterior) view. 12, S. blanda, proximal end view. 13, S. blanda, ventral (anterior) view. 14, S. clzampioni, dorsal (posterior) view. 15, S. clzampioni, proximal end view. 16, S. clzampioni, ventral (anterior) view. 17, S. elongata, dorsal (posterior) view. 18, S. elongata, proximal end view. 19, S. elongata, ventral (anterior) view. 20, S. frontalis, dorsal (posterior) view. 21, S. frontalis, proximal end view. 22, S. frontalis, ventral (anterior) view. 23, S. marginallis, dorsal (posterior) view. 24, S. marginallis, proximal end view. 25, S. marginallis, ventral (anterior) VOLUME 100, NUMBER 3 475

point one-third of e's length before posterior fourth. Our observations reveal a great sim­ end, distinctly posterior to midpoint on e. ilarity in the female genitalia (including the Characters which occur in Systena and spermathecae) of the two genera. The se­ other hypothetically closely related genera tation of the last abdominal tergite in fe­ (potential synapomorphies for Systenini, males of Systena and Platiprosopus are the see discussion below) include: parallel sid­ same except there are many more com­ ed body (Fig. 1); elytron without apical de­ pound microtrichia (Fig. 39a) distributed clivity; frontal and anterofrontal ridges along the basal margin of the tergite in Pla­ (Figs. 7, 8) well developed but not high; tiprosopus. anterofrontal ridge lower than frontal ridge; Heikertinger and Csiki (1939) arranged a and labrum with four long sensilla on outer worldwide catalog based on their ideas (dorsal) surface (Fig. 2); apex of tergite VII about relationships among flea beetle gen­ of female with about 10-15 longer, apical­ era. They placed Systena between Prasona ly-curved setae and compound microtri­ Baly and Agasicles Jacoby. These taxa were chiae (Fig. 39). followed by Tanygaster Blatchley, Heiker­ tingeria Csiki, and Aeanthonyeha Jacoby. PRELIMINARY DISCUSSION OF RELATIONSIIlPS In 1962 Scherer synonymized Prasona with OF SYSTENA Systena, although he did not explicitly in­ We provide here a summary of most of dicate, "New Synonymy." Scherer men­ the genera that have been implicitly or ex­ tioned that Jacoby previously had reserva­ plicitly suggested to share a close relation­ tions about the validity of Prasona (no ref­ ship to Systena. We have examined exem­ erence indicated, however), but Jacoby did plars of these taxa and in cases where these not formally synonymize them. do not represent the type species of the ge­ Our examination reveals that Prasona nus, we consider them representative of the (based on P. viridis Baly) shares several pu­ genus. tative synapomorphies with Systena: sper­ Clark (1865) in the first characterization matheca with short, abruptly curved pump of the genus believed that Systena is closely (Fig. 43), spermathecal duct wide, and me­ related to Oxygona Chevrolat (now Plati­ tanotal ridge d intersecting e at point one­ prosopus Chevrolat). The diagnostic char­ third of e's length before posterior end, dis­ acters he provided to distinguish the two tinctly posterior to midpoint on e (Fig. 33). genera include the shape of the body and They also share the bilobed intercoxal me­ pronotum. In Systena, the body is more par­ sostemal process (Fig. 32), and dorsally allel-sided and narrower than in Platipro­ flattened apical fourth of the metatibia. Dif­ sopus, and the pronotum is more quadrate ferences among these taxa include the and rectilinear with the posterior comers shape of the spermathecal receptacle, pos­ sharp and not rounded. Our examination of terior part of the tignum, abdominal tergites Platiprosopus aeutangula (Chevrolat) has VIII and IX, median lobe of aedeagus, and shown other differences and similarities. the width of the frontal ridge. From other We observed that Platiprosopus has the el­ studies it is known that these latter char­ ytron with an apical declivity, labrum with acters typically vary at the species level. at least 8 long setae, and fifth antenna! seg­ Therefore, we support the synonymy of ment longer than sixth but shorter than Prasona with Systena. Our investigation of

view. 26, S. oberthuri, dorsal (posterior) view. 27, S. oberthuri, proximal end view. 28, S. oberthuri, ventral (anterior) view. 29, S. s-littera, dorsal (posterior) view. 30, S. s-littera, proximal end view. 31, S. s-littera, ventral (anterior) view. 476 PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON

37

·.: 34 ( 'i i '··" .

Figs. 32-38. Adult anatomy of Systena blanda. 32, Meso- and metastemum, left ventrolateral view. 33, Metanotum (abbreviations from Konstantinov, in press). 34, Mesonotum and scutellum. 35, Right elytron, ventral view. 36, Hind wing (terminology from Kukalova-Peck and Lawrence 1993). 37, Metendostemite, ventral view. 38, Metendostemite, dorsal view. Abbreviations: aaa = anterior arm apex; bp = binding patch; dr = dorsal ridge; mes = mesepimeron; mg = median groove; msp = mesostemal process; mss = mesostemum; mts = metastemum. VOLUME 100, NUMBER 3 477

40

Figs. 39-41. Adult anatomy of Systena blanda. 39, Tergite VII, female (a) compound microtrichia, (b) spiniform microtrichia. 40, Sternite VII, female. 41, Female genitalia, oblique lateral view. Abbreviations: gu = digestive tract; st8 = sternite VIII; te8 = tergite VIII; ti = tignum; va = vaginal palpi.

Disonychodes exclamationis (Boheman) sclerotized plate with a very complicated shows agreement in most characters with shape; vaginal palpi are absent; the tignum Systena. is robust and strongly sclerotized; the sper­ Agasicles (based on A. connexa Bohe­ mathecal receptacle is wider than long; the man) has a well developed recurved flange basal part of the duct is horizontal; the basal of the metafemoral spring. The female gen­ part of the pump is long and extremely italia are extremely unusual: tergite VII weakly sclerotized; and the horizontal part lacks the setal and microtrichial characters of the pump is short and sclerotized as found in Systena; tergites VIII and IX of strongly as the receptacle. On the metano­ females are transformed into a strongly tum, ridge c is extremely close to the me- 478 PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON dian groove. The metendosternite has an somere bilobed; short, abruptly curved extremely narrow stalk. The procoxal cav­ pump of spermatheca; curved setae on ster­ ities are open behind. The dorsal side of the nite VII of female; and well-developed su­ labrum has at least 6 setiferous pores. The pracallinal sulcus. Differences include the anterofrontal ridge is extremely low, es­ following: anterofrontal ridge with medial pecially at the middle and has patches of depression; antenna! calli well-developed; setae laterally. The only character that is pronotum with antebasal transverse and shared by Agasicles and Systena is the elon­ longitudinal impressions; strong elytral cos­ gate, parallel sided body and elytron with­ tae; metatibiae with long, well-defined out an apical declivity. We therefore believe ridge dorsally (at least 1h length of tibia); Agasicles is only distantly related to Sys­ spermathecal duct coiled; and vaginal palpi tena. robust, with long, strong sclerotization an­ Our examination of specimens of Hei­ teriorly. Some of these differences are kertingeria was limited since only males of shared with Altica Geoffroy, and the sper­ an undetermined species were available for matheca in particular suggests that Chla­ study. These specimens have an elytral api­ mophora is more closely related to Altica. cal declivity and a slightly convex dorsal Seeno and Wilcox (1982) recognized the surface of the metatibiae with an apically tribe Systenini containing, besides the type fiat longitudinal ridge. genus, Prasona, Egleraltica Bechyne and Like specimens of Agasicles, Acanthon­ Bechyne and Pseudogona Jacoby. The ycha females (A. jacobyi Bechyne) exam­ group of related genera separated from the ined lacked the usually developed tergites tribe contains Dysphenges Hom, Cyrsylus VIII and IX and apparently also the vaginal Jacoby, Trijiniocola Bechyne and Bechyne, palpi although there are 2 transverse scler­ Andiroba Bechyne and Bechyne and Agas­ otized structures below tergite VIL The tig­ icles. No specimens of Egleraltica, Dys­ num in Acanthonycha is robust and more phenges, Andiroba, Pseudogona, and Tri­ strongly sclerotized than in Systena. In finocola were available for examination, so Acanthonycha, the procoxal cavities are Seeno and Wilcox's idea on their relation­ open behind, the intercoxal mesostemal ship to Systena cannot be addressed. process is narrow and not bilobed, the se­ The last references to show a grouping of tation of tergite VII of females varies great­ taxa with Systena are Furth (1985, 1989). It ly from Systena, and the metendosternite is unclear whether or not he intends the has an extremely slender stalk. morphogroups to reflect potential phyloge­ Chapuis (1875) included Systena, Pra­ nenetic affinity, although he has discussed sona, and Chlamophora Chevrolat ( = Cla­ patterns of congruence between metafemor­ mophora Jacoby) in the taxon called Cre­ al spring morphogroups to generic prox­ pidoderites (Crepiderini) based on the ob­ imity in catalogues (Furth 1989: 499-503). longo-oval or elongate shape of the body Furth (1985) included Systena in morpho­ and the presence of the transverse impres­ group 3, containing Dysphenges Horn, sion on the pronotum. Hom ( 1889) also rec­ Phyllotreta Chevrolat, Pseudolampsis ognized the close relationship of Chlamo­ Hom, Pachyonychus Chevrolat and Lu­ phora with Systena, but believed it was a praea Jacoby. Our study of Systena meta­ mistake to include them in the Crepidod­ femoral springs (Figs. 10-31) reveals ex­ erites and rather proposed a suprageneric treme variation in the length of the dorsal taxon called Systenae. Specimens of Chla­ lobe's extended arm, the curvature of the mophora (C. meridionalis) share similari­ dorsal margin of the dorsal lobe, the angle ties with Systena in the following character of the dorsal margin of the ventral lobe to states: four long sensilla on the labrum; ab­ the dorso-ventral axis, the development of sence of supraorbital sulci; third metatar- the recurve flange, and in the tightness of VOLUME 100, NUMBER 3 479

·.

sg 42

.

.

44 46 0 Figs. 42-46. Adult anatomy of Systena blanda. 42, Vaginal palpi. 43, Spermatheca and gland. 44, Aedeagus, ventral sculpturing. 45, Aedeagus, lateral view. 46, Aedeagus, dorsal view. Abbreviations: pu = pump; re = receptacle; sd = spermathecal duct; sg = spermathecal gland; ssd = sclerotized spermathecal duct. 480 PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON the coiling of the dorsal and ventral lobes. larly revealing of differences in the coiling Systena approximates morphogroup 2 or 3, of the dorsal and ventral lobes and the ex­ but we cannot unambiguously assign it be­ tent of the recurve flange. Compound mi­ cause the differences between them are un­ croscopy is hampered because of the diffi­ clear. We extract their definitions here: culty to position the metafemoral spring in Morphogroup 2.-Furth (1980:267): "is a standard position for comparison with distinctive but is apparently close to [mor­ other taxa. A slight rotation on one axis can phogroup 1]"; Furth (1982:20): "very sim­ alter the appearance of structures or hide ilar to morphogroup 1 with one major dif­ others. Compound microscopy does not al­ ference; the presence of the recurve flange ways reveal the complicated folding patters as an extension of the ventral lobe. Some­ accurately and does not always reveal the times the recurve flange is highly devel­ recurve flange when it is in fact present. oped . . . yet in others . . . the recurve flange The characteristics of the metafemoral is only slightly developed."; Furth (1989: spring defining the morphogroups should 508): "simple modification of morpho­ be reassessed with discrete states, perhaps group 1, possessing the addition of a well­ in a matrix format so they can aid in iden­ developed recurve flange attached to the tifying generic groups and be used in phy­ ventral lobe." logenetic investigations. Only with explicit Morphogroup 3.-Furth (1980:267): character states can genera be assigned un­ "also distinctive from [morphogroup 1 and ambiguously (since morphogroups were morphogroup 5] but seems to have some proposed, [Furth 1980], about 10 genera similarities to both"; Furth (1982:22): "rel­ have been reassigned [Furth 1982, 1989]). atively short extended arm of dorsal lobe Comments on potential relationships of which is significantly depressed apically the remaining taxa in morphogroup 3 (Furth and out of line from horizontal axis of dor­ 1985) are listed here based on comparison sal lobe. Also the ventral lobe is extended with many of the character states found in into recurve flange, though not well devel­ the examined Systena species. Pseudolamp­ oped"; Furth (1989:508): "has a shorter sis (P. guttata LeConte) and Pachyonychus dorsal lobe (but noticeably longer than ven­ share many features (apically swollen last tral lobe), apically depressed, with a small metatarsomere; short, straight, dorsally flat recurve flange." metatibia with preapically inserted metatar­ It is clear that both morphogroups are sus; flat and moderately wide frons) with characterized by a recurve flange but it is Monoplatini, a well supported lineage in unclear how the flange differs, if at all. Ap­ Alticinae, and are only distantly related to parently morphogroup 3 differs from 2 in Systena. Phyllotreta species (including Tan­ having a "relatively short extended arm" y gaster, synonymized by Smith, 1979) although it is unclear if it is shorter relative share with Systena a moderately flat body to morphogroup 1 or 2 or both, and it is shape and elytra without an apical declivity. impossible to know at what point the ex­ Phyllotreta differs by its undeveloped or tended arm becomes long enough to merit extremely poorly developed supracallinal, placement into another morphogroup. The midfrontal and suprafrontal sulci, much apical depression of the dorsal lobe, char­ longer dorsolateral row of small spines on acteristic of morphogroup 3, is unclear and the apex of the metatibiae, arcuate intercox­ is not strikingly different in any of the il­ al mesosternal process, open procoxal cav­ lustrations of Furth (1985, 1988, 1989). ities, metendosternite well developed, with Scanning electron microscopy is the best fully developed ventral process contacting procedure to accurately determine the char­ the ventral projection, and fully developed acteristics of the metafemoral spring. The tendons of the metafurcal-mesofurcal mus­ views from the proximal end are particu- cles. These differences and the fact that VOLUME 100, NUMBER 3 481

they share no presumed synapomorphies of LITERATURE CITED Systenini, strongly suggests they belong in Barber, H. S. and J. C. Bridwell. 1940. Dejean Cata­ different alticine lineages. Lupraea (L. lon­ logue names (Coleoptera). Bulletin of the Brook­ gicornis Jacoby) shares no provisional syn­ lyn Entomological Society 35(1): 1-12. apomorphies of Systenini. Several character Blackwelder, R. E. 1946. Checklist of the Coleopter­ states place it in different groups of genera: ous insects of Mexico, Central America, the West Indies, and South America, Bulletin of the United the orbit is extremely narrow, the frontocly­ States National Museum 185, Part IV: 551-764. peal suture is undeveloped, the labrum has Chapuis, E 1875. Famille des Phytophages, Vol. XI. more than 15 setae on the dorsal surface, In Lacordaire, T. and E Chapuis, Histoire Natu­ the antero and posterolateral callosities of relle des Insectes. Genera des Coleopteres Librair­ pronotum are undeveloped, the lateral mar­ ie Encyclopedique de Roret, Paris. 420 pp. Chevrolat, L. A. 1836. pp. 361-443. In Dejean, P. E gin of the pronotum is broadly explanate, M. A., Catalogue des coleopteres de la collection the prostemal intercoxal process is narrow de M. le comte Dejean, livr. 5, Paris [second edi­ and short, and the procoxal cavities are tion]. open. Lupraea also exhibit sexual dimor­ Chevrolat, L. A. 1848. In D'Orbigny, Dictionnaire Universe! d'Histoire Naturelle, vol. 12, 816 pp. phism in the size of the eyes. Cyrsylus Ja­ Paris. coby specimens (based on C. recticollis Ja­ Clark, H. 1865. An examination of the Halticidae of coby) also do not possess any of the pro­ South America. Journal of Entomology 2 (13): visional synapomorphies of Systenini. The 375-412. remaining genera in Furth's (1985) morpho­ Cox, M. L. 1996. The pupae of Chrysomeloidea, pp. 119-265. In Jolivet, P.H. A. and M. L. Cox, eds., group 3, Glenidion Clark and Phydanis Chrysomelidae Biology, vol. 1: The Classifica­ Hom, were unavailable for examination, so tion, Phylogeny and Genetics. SPB Academic comments on their relationship to Systena Publishing, Amsterdam. 444 pp. cannot be made here. Crotch, G. R. 1873. Check List of the Coleoptera of Our study suggests that many of these America, North of Mexico. Naturalists Agency, Salem, Mass, 136 pp. genera which have previously been aligned Furth, D. G. 1980. Inter-generic differences in the me­ with Systena, either in a catalog listing or tafemoral apodeme of flea beetles (Chrysomeli­ morphological grouping share few of the dae: Alticinae). Systematic Entomology 5: 263- uncommon characters (provisional synapo­ 271. morphies) of the tribe Systenini, and there­ Furth, D. G. 1982. The metafemoral spring of flea bee­ tles. Spixiana 7: 11-27. fore probably belong to different alticine Furth, D. G. 1985. Relationships of Palearctic and Ne­ lineages. Platiprosopus seems to be the arctic genera of Alticinae. Entomography 3: 375- only genus that shares with Systena some 392. of these characters. Furth, D. G. 1988. The jumping apparatus of flea bee­ tles (Alticinae)-The metafemoral spring, pp. 285-297. In Jolivet, P., E. Petitpierre, and T. ACKNOWLEDGMENTS Hsiao, eds., Biology of Chrysomelidae. Kluwer Academic Pub!., Dordrecht. 615 pp. We thank David Furth, Eric Grissell, Furth, D. G. 1989. Metafemoral spring studies of some Darlene Judd, Eric Smith, and Charlie neotropical genera of Alticinae. Entomography 6: Staines for reviewing this manuscript and 497-510. providing valuable suggestions. Discussion Gemminger, M. and Harold, E. von 1874. Catalogus of logic, nomenclature, and wing veins with coleopterorum hucusque descriptorum synonymi­ cus et systematicus, Vol. 11: 3233-3478. Norman Woodley, R Christian Thompson, Hatch, M. H. 1927. A systematic index to the keys for and Kunio Suzuki, respectively, was very the determination of the Nearctic Coleoptera. helpful. Jong Eun Lee received a grant from Journal of the New York Entomological Society Andong National University, South Korea 35: 279-306. Heikertinger, E and Csiki, E. 1939. Coleopterorum Ca­ which enabled study at the National Mu­ talogus, pars 166, Chrysomelidae: Halticinae, 1, seum of Natural History, Smithsonian In­ pp. 1-336. [Vol. 25]. stitution, Washington, D.C. Horn, H. 1889. A synopsis of the Halticini of boreal 482 PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON

America. Transactions of the American Entomo­ 1. Body proportions ratio measured as logical Society 16: 163-320. length of elytron divided by greatest ICZN. 1985. International Code of Zoological Nomen­ clature, Third Edition, adopted by the XX General width of elytra. (0) 1.40-1.80; (1) Assembly of the International Union of Biological >1.80 Sciences. International Trust for Zoological No­ 2. Elytron apex. (0) without apical decliv­ menclature, London, xx + 338pp. ity; (1) with apical declivity Konstantinov, A. 1994. Comparative morphology and 3. Antenna! calli. (0) as large as antennal some evolutionary trends in flea beetles (Altici­ nae). pp. 383-391. In Jolivet, P. H., M. L. Cox, socket, longer than wide; (1) not as and E. Petitpierre, eds., Novel aspects of the bi­ large as antennal socket, wider than ology of Chrysomelidae. Kluwer Academic Pub­ long lishers. The Netherlands. 4. Anterofrontal ridge. (0) lower than Konstantinov, A. S. In press. Genus Aphthona Chev­ rolat of the Palearctic region (Coleoptera: Chrys­ frontal ridge; (1) higher than frontal omelidae: Alticinae). Memoirs on Entomology, ridge International. 5. Supracallinal sulcus. (0) well devel­ Konstantinov, A. S. and I. K. Lopatin. 1987. Compar­ oped; (1) poorly developed (indistinct) ative-morphological study of the metendosternite 6. Orbital sulcus. (0) absent; (1) present in leaf-beetles of the subfamily Alticinae (Cole­ optera: Chrysomelidae). Entomological Review 7. Frontal ridge. (0) well developed; (1) 66(4): 154-163. poorly developed Konstantinov, A. S. and N. Vandenberg 1996. Hand­ 8. Dorsal surface of labrum. (0) with 4 se­ book of Palearctic Flea Beetles (Coleoptera: tae; (1) with more or less setae Chrysomelidae: Alticinae). Contributions on En­ tomology, International 1(3): 237-439. 9. Fifth antennomere length relative to Kukalova-Peck, J. and J. F. Lawrence. 1993. Evolution fourth and sixth. (0) longer than 4&6; of the hind wing in Coleoptera. The Canadian En­ (1) shorter than 4&6; (2) 4=5>6; (3) tomologist 125: 181-258. 4>5=6 Leng, C. W. 1920. Catalogue of the Coleoptera of 10. Procoxal cavities. (0) closed; (1) open America, north of Mexico, Mount Vernon, N.Y. 470 pp. 11. Prosternal intercoxal process. (0) Melsheimer, F. E. 1847. Descriptions of new species strongly expanded at apex; (1) narrow­ of Coleoptera of the United States. Proceedings of ly expanded at apex the Academy of Natural Sciences at Philadelphia 12. Mesoprescutum. (0) truncate posteri­ 3: 158-181. Monros, F. and J. Bechyne. 1956. iiber einige verkann­ orly; (1) triangular in shape te Chrysomeliden-namen. Entomologischen Ar­ 13. Metendosternite anterior arms. (0) de­ beiten aus dem Museum G. Frey, Tutzing 7(3): flexed anteapically; (1) not deflexed at 1118-1137. ends Poole, R. W. and P. Gentili, Eds. 1996. Nomina Insecta 14. Metanotal ridge d. (0) intersecting c at Nearctica. Entomological Information Services, Rockville, Maryland, 827 pp. point one-third of e's length before pos­ Scherer 1962. Bestimmungsschliissel der neotropisch­ terior end, distinctly posterior to mid­ en Alticinen-Genera (Coleoptera: Chrysomelidae: point on c (1) not as in 0, d intersecting Alticinae). Entomologische Arbeiten aus dem Mu­ c at point much anterior (at least ante­ seum G. Frey 13: 497-599. Seeno, T. N. and Wilcox, J. A. 1982. Leaf beetle gen­ rior to midpoint of d). era (Coleoptera: Chrysomelidae). Entomography 15. Elytral punctation. (0) irregular, con­ 1: 1-221. fused; (1) arranged more or less in stri­ Smith, E. H. 1979. Genus Tanygaster Blatchley, a new ae synonym of Phyllotreta Chevrolat (Coleoptera: 16. CuA • (0) not attached to CuA; (1) Chrysomelidae: Alticinae). Coleopterists Bulletin 2 33(3): 359-362. reaching CuA 17. Dorsolateral apex of metatibia. (0) with APPENDIX 22-30 small spines; (1) less than 20 Characters and states for Systena species spines present compared in Table 1. More thorough de­ 18. Coiling of dorsal lobe. (0) coiled nearly scriptions given in text. completely along dorso-ventral axis; VOLUME 100, NUMBER 3 483

(1) coiled only at extreme distal end pound and spiniform microtrichiae; (1) (near attachment to tibia). without both types 19. Extended arm of metafemoral spring. (0) 22. Spermathecal duct. (0) irregulary between 0.35 and 0.25 length of spring; shaped with several widened and (1) less than 0.25 length of spring. curved regions; (1) not as in 0 20. Aedeagus apex. (0) arrow shaped with 23. Spermathecal gland. (0) basal, sclero­ fine striae at extremity; ( 1) not arrow tized portion long and roughly parallel shaped and without striae. sided; ( 1) not as in 0 21. Tergite VII of female. (0) with com-