Novitates PUBLISHED by the AMERICAN MUSEUM of NATURAL HISTORY CENTRAL PARK WEST at 79TH STREET, NEW YORK, N.Y

AMERICAN MUSEUM Novitates PUBLISHED BY THE AMERICAN MUSEUM OF NATURAL HISTORY CENTRAL PARK WEST AT 79TH STREET, NEW YORK, N.Y. 10024 Number 301 1, 43 pp., 85 figures, 4 tables June 26, 1991

Cladistics and the Biogeography of Two Trans-Caribbean Hairstreak Butterfly Genera: Nesiostrymon and Terra (Lepidoptera, Lycaenidae)

KURT JOHNSON'

ABSTRACT Two sister genera composed of notably non- T. hispaniola Johnson and Matusik (southern His- vagile members distributed both in the Greater paniola); on the mainland, type species T. tera Antilles and mainland Neotropics are revised. (Hewitson) (central Mexico to Colombia); three Cladistic classification is employed from numer- species transferred from Thecla, T. calchinia ical analysis (PAUP, Swofford) ofthe ingroup and (Hewitson) (upper Amazon), T. cana (Hayward) three outgroups restricted to the mainland. Ne- (northwest Argentina), T. chilica (Schaus) (south- siostrymon includes five species: in the Antilles, east Brazil, Paraguay); and one new species, T. type species N. celida (Lucas) (with three interis- andevaga (Ecuador). Comparison of area clado- land subspecies-nominate [Cuba] N. c. aibonito grams of Nesiostrymon and Terra shows congru- (Comstock and Huntington) [Puerto Rico and ence of plesiotypic Antillean elements and apo- northern Hispaniola], N. c. baorucoensis, new sub- typic members occurring in Central America and species [southern Hispaniola]), and N. shoumatoffi South America, respectively. These are construed (Comstock and Huntington) (Jamaica); on the as more compatible with late-Cretaceous-Tertiary mainland, pan-Neotropical N. celona (Hewitson) vicariance of a proto-Greater Antilles from the (transferred from Thecla) and two new insular spe- Central and South American regions than a com- cies, N. milleri (cloud forests, Aragua, Venezuela) plex dispersal explanation requiring numerous and N. australivaga (scrub-steppe, Mendoza, Ar- concordant dispersals. gentina). Terra includes six species: in the Antilles,

' Research Associate, Department of Entomology, American Museum of Natural History.

INTRODUCTION Nesiostrymon Clench (1964) and Terra tory (FMNH); Hope Entomological Collec- Johnson and Matusik (1988) are sister genera tion, Oxford University (HEC); Instituto of notably nonvagile "hairstreak butterflies" Miquel Lillo (Tucuman, Argentina) (IML); (Theclinae, Eumaeini) occurring both in the Los Angeles County Museum (LACM); Mil- Antilles and the mainland Neotropics. Nu- waukee Public Museum (MPM); Museum merical cladistic analyses associated with the National d'Histoire Naturelle (Paris, France) description of Terra indicated that it and Ne- (MNHN); Museo Nacional de Historia Nat- siostrymon are the sister group of three eu- ural (Santo Domingo, Dominican Republic) maeine assemblages restricted to the main- (MNDR); Zoologisches Museum der Hum- land (Johnson and Matusik, 1988). Observing boldt Universitiit zu Berlin (ZMH). Speci- the local habitat restriction and widely dis- mens were also examined from three private junct distributions ofNesiostrymon and Ter- collections with extensive Antillean or aus- ra taxa, Johnson and Matusik (1988) sug- tral South American holdings: Robert C. Ei- gested that Antillean species of these genera sele Collection (Jujuy, Argentina) (REC); Da- (along with two other recently discovered vid Matusik Collection (Skokie, Illinois) south Hispaniolan endemics) most likely "re- (DM); and Albert Schwartz Collection (Mi- flect a tectonic relationship between ... ami, Florida) (ASC). regions . . . on the northern edge of the Ca- ribbean plate." SYSTEMATICS Recent biogeographical studies (Shields and Phylogenetic Analyses ofNesiostrymon and Dvorak, 1979; Buskirk, 1985; Johnson and Terra: Johnson and Matusik (1988) used nu- Descimon, 1989; Miller and Miller, 1989) merical cladistic analysis (PAUP; Swofford, recognized a component of Antillean Rho- 1985) to establish the monophyly of Nesios- palocera more suggestive of an ancient vi- trymon, Terra, and certain outgroup taxa of cariant origin than the prominent waif-dis- the lycaenid grade "Thecla" (sensu Bridges, persal origin attributed to the fauna by early 1988) (fig. 1; tables 1, 2). Focusing primarily workers (Comstock and Huntington, 1949; on the Caribbean region, Johnson and Ma- Scott, 1972; Riley, 1975; Brown, 1978). Mil- tusik (1988) distinguished Antillean mem- ler and Miller (1989), summarizing a prob- bers of Nesiostrymon and Terra and listed able vicariance and dispersal model for mainland relatives from the "Thecla" grade. Antillean butterfly origins, included Nesio- Subsequently, Johnson (1991) published a strymon and Terra as one of nine butterfly nomenclature for outgroup taxa (figs. 84, 85). groups most probably arising from late Cre- The cladistic study of Johnson and Ma- taceous-Eocene vicariance. tusik (1988) indicated that Terra and Ne- This study reports a cladistic taxonomy of Neotropical Nesiostrymon and siostrymon form a monophyletic group Terra and distinguished from outgroups by six synapo- presents the ecological and ev- biogeographic morphies of the wings, genitalia, and termi- idence supporting a vicariant origin for An- tillean members of two nal abdominal morphology (table 1: charac- these hairstreak but- ters 2, 4-6, 12, and Terra terfly genera. 14). taxa share a salient apomorphy: the prominent sclerotized bulb protruding ventrally between the MATERIALS AND METHODS cephalic and caudal elements of the ductus COLLECrIONS bursae (table 1, character 11). Nesiostrymon taxa share a specialized prominently pronged Specimens examined included samples tergite 8 in males (table 1, characters 1, 9). from the Allyn Museum of Entomology, High consistency values (fig. 1) for clado- Florida Museum of Natural History (AME); grams result from the various very distinctive American Museum of Natural History morphological characters which separate (AMNH); British Museum (Natural History) these taxa (e.g., the incised posterior cavity (BMNH); Carnegie Museum of Natural His- ofmale tergite 8, distribution ofmicrotrichia, tory (CMNH); Field Museum ofNatural His- etc.). "Outgroup" and "Lundberg outstate" 1991 JOHNSON: HAIRSTREAK BUTTERFLY GENERA 3

AB C D A B CD E F 5-' - 1 4 12 ' 2,4,5,6,12,1 4 1 , 9 T' caption, A caption, B caption, C (Tb. 1) 2,4,5,6,12,1l4 Fig. 1. Cladogram ofNesiostrymon (D), Terra (C), and relatives (A, B) (Johnson and Matusik, Fig. 2. Ingroup cladogram ofNesiostrymon (A, 1988). Four-taxon statement derived from parsi- B, C) and Terra (D, E, F). Six-taxon statement monious distribution ofunweighted characters and derived from parsimonious distribution of un- rooted using (1) outgroup[s] described in text (con- weighted characters and rooted using Lundberg sistency index = 0.923) and (2) Lundberg method outstates based on monophyly of Nesiostrymon (Swofford, 1985) based on presumed primitive and Terra, figure 1. Consistency index = 0.909. states listed at bottom oftable 2 (consistency index Apomorphies are specified by horizontal bars and = 0.889). Apomorphies for Nesiostrymon and Ter- represent characters enumerated in tables 3 and ra specified by horizontal bars enumerated from 4. Synapomorphies noted for Nesiostrymon/Terra 1 tables and 2. Outgroups and their apomorphies: are from figure 1, tables 1 and 2. Terminal groups A, B. Outgroup rooting preferred " Thecla uzza include A, N. celona/milleri/australivaga; B, N. complex" (fig. 84) as sister group ofNesiostrymon! celida; C, N. shoumatoffi; D, T. hispaniola; E, T. Terra based on characters 9, 10 with " Thecla cel- tera; F, T. andevaga/chilica/calchinia/cana. mus complex" (fig. 84) as the outgroup; Lundberg rooting preferred the opposite based on characters, 1, 3, 7, 13. C. "Thecla" outgroup (fig. 85), characters enumerated by Johnson and Matusik, 1988. demic T. hispaniola being the most primitive in character. Based on these results, species- level taxonomies for Nesiostrymon and Terra (Swofford, 1985) rootings (fig. 1) give iden- are enumerated below. tical results concerning the monophyly of Terra and Nesiostrymon. TAXONOMY OF NESIOSTRYMON and To determine a cladistic classification of TERRA species in Nesiostrymon and Terra, parsi- GENUS NESIOSTR YMON CLENCH monious distributions of characters were constructed for shared characters of species Figures 3-32, 57, 59-68 in each genus (fig. 2, tables 3, 4). Monophyly Nesiostrymon Clench, 1964: 248, 251. Brown and of the ingroup was assumed from Johnson Heineman, 1971: 4; 1972: 232. - Scott, 1986: and Matusik's (1988) results; species criteria 21. - Riley, 1975: 101. - Bridges, 1988: 1.74, were based on standard taxonomic proce- II.77. - Johnson and Matusik, 1988: 236. - dures involving consistent in Schwartz, 1989: 240. - Miller and Miller, 1989: differences 245. characters of the wings and genitalic and tergal morphology. Table 3 lists the apomorphic DIAGNOSIS: Compared to all Eumaeini: and plesiomorphic states ofcharacters delim- male tergite 8 modified to a "subcordate in- ited from the final rooted tree (fig. 2). This cised posterior cavity" (the "sipc" sensu Field, analysis indicates the following: Nesiostry- 1967a, 1967b; Johnson, 1988a, 1989a, 1989b; mon is a monophyletic group with two nested Johnson and Matusik, 1988) terminating in sets of taxa, the Jamaican endemic, N. shou- a dorsally produced lobe or prong, genital matoffi, being the most primitive in charac- valvae multiplanar with protruding, winglike ter; Terra is a monophyletic group with two lateral lobes; female genital ductus bursae ex- nested sets of taxa, southern Hispaniolan en- hibiting heavily sclerotized, opaque caudal 4 AMERICAN MUSEUM NOVITATES NO. 3011 and cephalic elements separated by a trans- eumaeines have an extended caudal lobe, with parent neck; both sexes' wing under surfaces only a disjunct, slightly sclerotized pad (the gray-white, crossed by a jagged or disjunct "extravalvular element" sensu Johnson and black medial band. Compared to Terra: Ter- Matusik, 1988); aedeagus elongate (length ra male tergite 8 unmodified and genital val- usually exceeding rest of genitalia by at least vae flat and uniplanar; female ductus bursae one-third) with caecum short (comprising with bulblike sclerotized element protruding one-fifth to one-seventh of aedeagus length, basally from transparent ductal neck; wing depending on the species) and terminus un- under surfaces with medial yellow, orange, or dulate; terminus ofaedeagus with two spine- reddish bands. like cornuti. Female Tergal Morphology and DESCRIPTION: Adult. Body fuscous, finely Genitalia (figs. 59, 62, 64, 67). Tergite 8 un- overlaid with gray to bluish hairlike scales specialized but papillae anales modified by and setae; eyes ringed with white; antennae heavy sclerotization to a steeply pointed, black, finely striped with white. Male. Upper nearly opaque terminus; ductus bursae di- Surface of Wings: apical and marginal areas vided into darkly sclerotized caudal and ce- of wings fuscous to black; rest of wings bril- phalic elements separated by a transparent liant iridescent blue. Large elliptic androco- neck, relative length and shape of respective nial element (hereafter, the "brand" sensu caudal and cephalic elements varying greatly Eliot, 1973) in distal area of discal cell var- between the species; ventral opening of duc- iously overlapping blue ground color. Hind- tus terminus with thin transverse lamellal lips; wing with long tail at terminus ofvein CuA2, juncture of ductus bursae to corpus bursae short tail at terminus of vein CuA 1. Under blunt, without distinctive structural elements Surface ofWings: ground color white to gray at the cervix bursae; signa broadly based with jagged to disjunct medial and/or sub- sclerotized plates with a raised, cephalically marginal bands formed by black dots or directed keel. slashes in the wing cells (most commonly a TYPE SPECIES: Thecla celida Lucas (1857) disjunct black medial band on each wing). by original designation. Clench (1964) des- Limbal area with prominent eyespot at mar- ignated N. celida as type, illustrating what he gin of cell CuA 1 (the "Thecla spot" sensu considered the Jamaican "subspecies" (shou- Clench, 196 1). Length of forewing: 8.0-12.5 matoffi Comstock and Huntington). Al- mm. Female. Upper Surface of Wings: sim- though the latter is clearly a distinct species, ilar to those ofmale but structural color dull- I consider T. celida the type, consistent with er, fuscous margins and apices wider, fore- Clench's taxonomy. wings without brands. Under Surface of DISTRIBUTION (fig. 81): Central Mexico Wings: Similar to those of males. Length of southward to central Argentina; in the An- Forewing: 9.0-13.0 mm. Male Tergal Mor- tilles (Cuba, Puerto Rico, Hispaniola, Ja- phology and Genitalia (figs. 57, 60, 61, 63, maica). 65, 66, 68). Sipc terminating in dorsal lobe DISCUSSION: Contrary to Clench's original or prong of varying length. Dorsoterminal view (1964), Nesiostrymon is not an endemic surface of genitalia with two small "brush Antillean genus (see Biogeography). Rather, organs" (sensu Eliot, 1973) abutting lateral it includes two Antillean species (one with edge of vinculum; length of brushes not ex- three distinctive allopatric populations) and ceeding terminus of labides. Genitalia with three species on the mainland (two of which lateral surface ofvinculum tapered smoothly are undescribed). Referring to the present re- between labides and saccus (not angled as in visionary study, Johnson and Matusik (1988) Terra); saccus greatly elongate, cephalically informally listed Thecla species possibly be- distending the vinculum; valvae, compared longing in Nesiostrymon and Bridges (1988) to other Eumaeini, very small, length ofcau- followed this list in his index of Thecla taxa. dal extensions not exceeding ventrocaudal Accordingly, and since Bridges (1988) did not length of vincular arc (fig. 57) and shaped use formal binomials in his index, I consider distinctly multiplanar with winglike lateral the present work as establishing the new com- lobes protruding at different lengths in the binations and synonymies cited in the tax- various species; valval terminus, where most onomy below. 1991 JOHNSON: HAIRSTREAK BUTTERFLY GENERA 5

TABLE 1 Characters of Nesiostrymon, Terra, and Relatives TABLE 1-(Continued) Used for Figure 1 Cladogram Construction (apomorphic (A) and plesiomorphic (P) states for 10. Female, condition adjacent transparent neck: (A) characters of table 2 matrix; descriptions and ter- transparent neck with prominent ventrally directed ovate minology as noted belowa) sclerotized element (the "bulb" [fig. 69] sensu Johnson and Matusik, 1988) (figs. 69, 70, 71, 75-77); (P) see entry Tergal Morphology 9, A (figs. 59, 62, 64, 67). 1. Male, condition of tergite 8: (A) tergite 8 heavily 11. Female, condition of bulb (fig. 69): (A) bulb lo- sclerotized throughout, forming incised posterior cavity cated about equidistant from caudal and cephalic ductal ofsubcordate shape (the "sipc" sensu Field, 1967a, 1967b; termini (fig. 69, 70, 71, 75-77); (P) bulb occurring at Johnson, 1988a, 1989a, 1989b) (figs. 57c, d, 60, 61, 63, cephalic base of ductus bursae, abutting distal end of 65, 66, 68); (P) tergite 8 normal (sensu Johnson, 1989a) corpus bursae (see Johnson and Matusik, 1988: 241) (fig. (figs. 58c, d, 69, 72, 73, 74, 78-80). 84b). Genitalia 12. Female, configuration of papillae anales: (A) pa- 2. Male, configuration of saccus: (A) length of saccus pillae anales with heavy, caudally inclined, lateral scler- equal to or exceeding 2.5 times "radius of vincular arc" otizations forming a pointed terminus (figs. 59, 62, 64, (fig. 57a, b); (P) ventrum ofvinculum generally entire or 67, 69, 70, 75, 77); (P) papillae anales lobate, with evenly with saccus development negligible (saccus length equal distributed sclerotized elements (taxa of figs. 84; 85a-c to or less than radius of vincular arc) (figs. 84a, b, d-g; [papillae anales not figured]). 85a-c, e). Wing Pattern 3. Male, configuration of vinculum: (A) "caudoven- 13. Both sexes, pattern of under surface (figs. 3-56): tral diameter ofvincular arc" (fig. 57a) equal to (or within (A) under surface ofhindwing with pattern elements oc- 0.2 standard deviation from saccus length (fig. 85c); (P) curng (a) medially at least in cells SC + RI, Rs, Ml, caudoventral diameter of vincular arc equal to at least M2, M3, CuAl, CuA2, and 2A, and (b) postbasally at twice saccus length (fig. 84a, d-g; 85a-c, e). SC + RI; (P) pattern more elaborate than above includ- 4. Male, configuration of vinculum: (A) radius of ing (a) distally expansive, continuous or lunular median vincular arc less than length of saccus (figs. 57, 58); (P) bands, (b) extensive markings in the postbasal area, and radius of vincular arc equal to or greater than length of (c) pronounced limbal markings in addition to the usual saccus (figs. 84a, b, d-g; 85a-c, e). Thecla spot and accordant ground color suffusions (see 5. Male, valvae "caudal extension" (figs. 57a-58a): Johnson and Matusik, 1988: 241). (A) length ofvalves not exceeding caudoventral diameter 14. Both sexes, pattern of under surface: (A) under of vincular arc (figs. 57-58, 61, 63, 65, 66, 68, 72, 73, surface ofhindwing with pattern elements crossing inside 74, 78-80); (P) valvae caudal extensions tapered to var- discal cell, and occurring either (a) postbasally in at least ious lengths exceeding caudoventral diameter ofvincular one of the following cells: cell SC + RI, cell CuA2, cell arc (figs. 84a, e-g; 85a-c). 2A; or (b) basally in the discal cell; (P) see entry 13, A. 6. Male, condition of valval apparatus at caudoven- External Secondary Sexual Characteristics tral extreme ofvincular arc (figs. 57-58): (A) area caudad 15. Male, dorsal forewing androconial structures: (A) of valvae apex and basal to arch of falces with a mem- forewing dorsum of male with ellipsoidal androconial brane-enfolded, ellipsoidal sclerotized pad (the "extra- brand (sensu Eliot, 1973) encompassing distal end of valvular element" sensu Johnson and Matusik, 1988, discal cell and basal portion of cells M3, M2, Ml, and figs. 57a-58a); (P) see entry 5, P (figs. 84a-g; 85a-c). the radial veins (figs. 3, 7, 11, 13, 17, 21, 25, 27, 29, 33, 7. Male, aedaegus: (A) length (measured along lateral 37, 41, 45, 49, 53); (P) androconial scales lacking or, in surface, including length ofcaecum displaced outside the a few outgroups cases, concentrated in a small andro- plane ofaedeagal shaft) equal to or exceeding 2.75 times conial mark limited to the area basad the crossvein of the caudoventral diameter of vincular arc (figs. 57aa- the forewing discal cell (see Johnson and Matusik, 1988: 58aa); (P) length (measured as above) variously less than 240). 2.75 arc times the caudoventral diameter of vincular a (measured as above) (taxa of fig. 84a, d-g; 85a-c [ae- For consistency, descriptions of apomorphic states deagus not figured]). follow Johnson and Matusik (1988); terminology follows 8. Female, configuration ofductus bursae: (A) ductus Johnson and Matusik (1988) and Johnson (1989a, 1989b, bursae divided into respective caudal and cephalic ele- 1989c; 1991) with first usage of any term marked with ments connected by a "transparent neck" (figs. 59, 62, quotation marks. 64, 67, 69, 70, 71, 75-77); (P) ductus bursae of contiguously opaque sclerotization (figs. 84b-h; 85a-c, e). 9. Female, condition of transparent neck: (A) transparent neck smooth (uninterrupted by sculpturing or ad- Studies of Antillean Lepidoptera have tra- ditional components) over entire circumference (the ditionally used subspecies categories for dis- "simple" condition characterized by Johnson and Ma- tinctive interisland populations. Some of tusik, 1988) (figs. 59, 62, 64, 67); (P) see entry 8, P (figs. 84b-h; 85a-c, e). these subspecies complexes exhibit general geographic patterns (Comstock and Hunting- 6 AMERICAN MUSEUM NOVITATES NO. 3011

TABLE 2 Character State Matrix for Nesiostrymon, Terra, and Relatives (columns 1-15, characters enumerated in table 1; rows as enumerated belowab) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 A 1 0 1 1 1 1 1 1 1 0 0 1 1 1 1 B 0 0 1 1 1 1 1 1 0 1 1 1 1 1 1 C 0 1 1 0 0 0 1 1 0 1 0 0 1 0 1 D 0 1 1 0 0 0 1 0 0 0 0 0 1 0 1 E 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 L 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 a Taxa A-E. (A) Nesiostrymon, (B) Terra, (C) uzza complex (enumerated in Systematics and fig. 84), (D) celmus complex (enumerated in Systematics and fig. 84), (E) outgroup (enumerated in Systematics and fig. 85). b L. Characters used for Lundberg rooting of parsimonious network. ton, 1943; Riley, 1975; Scott, 1972; Johnson (lobes nearly absent in N. shoumatofji, greatly and Descimon, 1989; Miller and Miller, produced in mainland taxa). Female genitalia 1989). Subspecies usage in Nesiostrymon is with caudal and cephalic ductal elements of compatible with cladistic results since Antil- about equal size (cephalic element diminu- lean Nesiostrymon exhibits two nested sets of tive in N. shoumatoffi and mainland taxa). taxa. Consistent with an equal rank criterion, Hindwing under surface with nearly com- each nested set corresponds to a species, one plete, black medial band and with yellow with three allopatric island populations hav- Thecla spot (N. shoumatoffi with mottled, ing three subspecies. On the mainland, there discontinuous band and black Thecla spot; appears to be no utility in using infraspecific mainland taxa with one or two continuous categories (Johnson, 1989c). Below, species brown bands and yellow to orange Thecla of Nesiostrymon are organized according to spot). the regional nested sets apparent in the in- DESCRIPTION: Male. Upper Surface of group cladistic analysis (fig. 2, table 2). Wings: ground color brilliant iridescent azure blue from base to subapex, forewing with elliptic black androconial brand contiguous ANTILLEAN TAXA with black coloration ofwing apex, hindwing Nesiostrymon celida (Hewitson) with thin black marginal band. Under Sur- Figures 3-20, 57, 59-61 face of Wings: ground color chalky white; Thecla celida Lucas, 1857: 610. - Kirby, 1871: forewing with black postmedial line, costa to 393. - Hewitson, 1869: 125, pl. 49, figs. 246, cell CuA 1; hindwing with medial band of 247. - Draudt, 1919: 803, pl. 159. - Kaye, closely adjacent black slashes vividly con- 1931: 535, pl.39, fig. 10 (referring only to shou- trasting the ground color, scattered black matoffi Comstock and Huntington, see next en- slashes in the postbasal area and light crenate try). - Wolcott, 1936: 402. - Comstock and black slashes along the outer margin; limbal Huntington, 1943: 76; 1958-1964 [1959]: 181. area with Thecla spot yellow to orange-yellow Strymon celida: Bates, 1935: 194. and a central black dot. Length of Forewing: Nesiostrymon celida: Clench, 1964: 248, 251. - 11.5-13.0 mm. Female. Upper Surface of Lewis, 1974: 67. - Brown and Heineman, 1972: Wings: similar to those ofmale but with fus- 232. - Scott, 1986: 21. - Riley, 1975: 101. - Bridges, 1988: I.74, 11.77. - Johnson and Ma- cous marginal areas more extensive and with- tusik, 1988: 236. - Schwartz, 1989: 240. out forewing androconial brand. Under Sur- face of Wings: similar to those of males. DIAGNosIs: Male tergite 8 with elongate Length of Forewing: 11.5-13.5 mm. Male dorsoterminal prong (diminutive lobe in N. Tergal Morphology and Genitalia (figs. 57, shoumatofli, shorter terminal prongs in main- 60, 61). Sipc typical of genus but with dor- land taxa N. celona, N. milleri, and N. aus- soterminal spine elongate (length exceeding tralivaga); male genitalia with lateral lobes three-fourths width of the tergite dorsum). protruding about one-half the valvae length Brush organs typical ofgenus. Genitalia with 1991 JOHNSON: HAIRSTREAK BUTTERFLY GENERA 7

TABLE 3 Characters Nesiostrymon and Terra Species TABLE 3-(Continued) and Species Groups Used for Figure 2 Cladogram Construction sae with prominent, ventrally directed ovate sclerotized (apomorphic (A) and plesiomorphic (P) states for element (the "bulb," figs. 69, 70, 71, 75, 76, 77); (P) characters of table 4 matrix; description and ter- same area without bulb. minology as in table 1) 10. Female, condition ofjuncture of caudal and cephalic ductal elements: (A) sclerotized elements robust Tergal Morphology and divided into caudal and cephalic elements by a trans- 1. Male, condition of tergite 8: (A) tergite 8 with parent central neck and adjacent sclerotized bulb (fig. specialized subcordate incised posterior cavity ("sipc" 69); (P) sclerotized elements robust but contiguous and sensu Field, 1967a, 1967b; Johnson, 1988a) with dor- entire (fig. 84). soterminal margin caudoventrally produced (figs. 57c, 11. Female, condition of sclerotized bulb: (A) scler- d, 60, 61, 63, 65, 66, 68); (P) tergite 8 unspecialized otized bulb robust, widest ventral diameter equaling ap- (without sipc) (figs. 58c, d, 69, 72, 73, 74, 78-80). proximately half the length of cephalic ductal element 2. Male, condition of sipc terminus: (A) dorsoter- and bulb occurring adjacent caudal end ofcephalic half minal margin of subcordate incised posterior cavity ter- of the ductus bursae (fig. 69); (P) sclerotized bulb robust minating in a caudoventrally directed lobe (not an elon- as above [note: as opposed to character 12 (A), maximal gate prong) (fig. 63); (P) dorsoterminal margin oftergite ventral bulb diameter less than one-fifth length of ce- 8 entire. phalic ductal element in apotypic Terra (fig. 2, taxa 5, 3. Male, condition of sipc terminus: (A) dorsoter- 6)] and occurring adjacent the cephalic end of the ce- minal margin of subcordate incised posterior cavity ter- phalic half of the ductus bursae (fig. 84). minating in an elongate, caudoventrally directed, prong 12. Female, condition of caudal and cephalic ductal (fig. 57c, d, 60, 65, 66, 68); (P) dorsoterminal margin of elements: (A) cephalic ductal element elongate and thin, tergite 8 entire. length equaling or exceeding five times maximal ventral diameter of bulb; bulb located in caudal half of entire Genitalia ductus length (figs. 70, 71, 75-77); cephalic ductal ele- 4. Male, configuration of valvae: (A) lateral margin ment robust, length approximately two times maximal of each valve with prominent terminally directed lobe ventral diameter of bulb; bulb located in cephalic half (fig. 57a, b); (P) lateral margin of each valve generally of entire ductus length (fig. 69). entire (figs. 58, 84). Wing Pattern 5. Male, condition ofvalve lateral margin: (A) lateral marginal lobe of each valve constricted to a point with 13. Both sexes, pattern of under surface: (A) under length oflobe equal to or exceeding that ofcentral valval surface of hindwing with postmedial pattern elements lobe (figs. 65, 66, 68); (P) lateral marginal lobe not con- colored golden yellow and occurring disjunctly in each stricted, elongate or in any other way sculptured so as cell from SC + R1 at least cell CuA2 and ofabout even valval lobe expression (figs. 38, 40); (P) under surface of hindwing to equal or exceed the length of the central with postmedial pattern element in cell RS obsolescent (fig. 57a, b). to lacking, causing postmedial band to appear basally 6. Male, condition ofvalvae central lobes: (A) central directed from vein M2 to the costa (figs. 34, 36). lobes tapered contiguously from valval base and paired in caudoventral one-third to one-fourth of valve length 14. Both sexes, pattern of under surface: (A) under (figs. 63, 72); (P) central lobes paired from caudoventral surface offorewing and hindwing with pattern elements one-half to entire valval length (figs. 84, 85a-c) [note: variously congealed into thickened lines or bands and considering primitive condition, central lobe = caudal colored yellow-orange, yellow-brown, brown, or red (figs. extension (fig. 57a)]. 42, 44, 46, 48, 50, 52, 54, 56); (P) see entry 13(A). 7. Female, configuration of ductus bursae: (A) respective opaquely sclerotized caudal and cephalic elements of ductus bursae connected by a smooth con- vinculum smooth and lacking ventral spurs. stricted and transparent neck (figs. 59, 62, 64, 67) (e.g., without "bulb"); (P) ductus bursae contiguous with Valvae with lateral winglike lobes widely el- opaque sclerotization (fig. 84). liptical but short (distal expanse only about 8. Female, condition ofcephalic ductal element: (A) one-half the length of the valval termini); contiguous opaquely sclerotized length ofcephalic ductal lateral margins of lobes variously produced element comprising not less than 0.4 that of contiguous in allopatric subspecies. Saccus robust, length opaquely sclerotized expanse of caudal ductal element about equal to caudoventral expanse of vin- (fig. 59); (P) ductus bursae with opaquely sclerotized "antrum" (figs. 84b, 85d sensu Klots, 1970) contrasting di- cular arc (fig. 57). Aedeagus typical of genus minutive cephalic element ofless than one-fourth ductus but with caecum more prominent than con- bursae length (figs. 84b, 85a, b, d). geners (comprising up to one-fifth aedeagal 9. Female, condition of juncture of caudal and ce- length) and terminus less recurvate. Female phalic ductal elements: (A) transparently sclerotized Tergal Morphology and Genitalia (fig. 59). juncture of caudal and cephalic elements of ductus bur- Tergite 8 typical ofgenus. Papillae anales with 8 AMERICAN MUSEUM NOVITATES NO. 3011

TABLE 4 Character State Matrix for Nesiostrymon and Terra Ingroup (columns 1-14, characters enumerated in table 3; rows as enumerated belowa b) 1 2 3 4 5 6 7 8 9 10 1 1 12 13 14 A 1 0 1 1 1 0 1 0 0 0 0 0 0 0 B 1 0 1 1 0 0 1 1 0 0 0 0 0 0 C 1 1 0 0 0 1 1 0 0 0 0 0 0 0 D 0 0 0 0 0 1 0 0 1 1 1 0 0 0 E 0 0 0 0 0 0 0 0 1 0 0 1 1 0 F 0 0 0 0 0 0 0 0 1 0 0 1 0 1 L 0 0 0 0 0 0 0 0 0 0 0 0 0 0 a Taxa A-F. (A) N. celona/milleri/australivaga, (B) N. celida, (C) N. shoumatofji, (D) T. hispaniola, (E) T. tera, (F) T. andevaga/chilica/calchinia/cana. b L, Characters used for Lundberg rooting of parsimonious network based on monophyly of Nesiostrymon and Terra in figure 1, tables 1 and 2. terminus more lobate than congeners. Gen- Nesiostrymon celida celida (Lucas) italia with lengths ofrespective opaque ductal Figures 3-6, 57, 60 elements generally similar, cephalic ductal el- CHARACTERIZATION: Black apices of fore- ement length comprising not less than two- wings broad, completely encompassing fifths of the respective caudal element; ter- the minal ductal opening narrow, edged with thin male forewing brands (fig. 3); male genitalia lamellal lips. (fig. 57, 60a-d) with lateral valval lobes pro- TYPE: Deposition uncertain. Comstock and truding at near right angles to the central val- Huntington (1958-1964 [1955]) and Bridges, val lobes with little, if any, marginal produc- 1988 speculated BMNH but this is uncon- tion. firmed (P. R. Ackery, BMNH, personal com- TYPES: See generic treatment. mun.) and I have not found it there. The DISTRIBUTION (fig. 8 1): Restricted to Cuba; name, however, is unambiguous. Type Lo- mainly known from old material with data cality: Cuba. suggesting mesic forest habitats typical ofthe DISTRIBUTION (fig. 81): Cuba, Hispaniola, species elsewhere. and Puerto Rico confined to dense mesic for- MATERIAL EXAMINED: N. c. celida: CUBA. ests and their immediate margins (see Bio- Guantanamo (16, 1 ) (AMNH); Holquin (26) geography). (AMNH); "Cuba" (66, 2Q) (BMNH); "Cuba" leg. Poey (26) (CMNH); Santiago (18) (BMNH); Bayate (18) (BMNH); Guantana- Allopatric Island Endemics of mo (1I ) (BMNH); no data [with Cuban spec- N. celida in the Antilles imens] (2Y) (BMNH); "Cuba" (26, 2!) (MNHN); Cotorro, 1884 (18) (MNHN); As noted under the generic entry, interis- "Cuaba" [sic] (one male) (MNHN); Guan- land subspecies have been traditionally tanamo (16, 1Q) (MNHN). employed by lepidopterists and these populations often have biogeographically informa- Nesiostrymon celida aibonito tive characters. Hitherto, only the Cuban and (Comstock and Huntington) Puerto Rican populations of N. celida have Figures 7-12, 60, 61 been characterized and names have not been available for the Hispaniolan elements. In- Thecla celida aibonito Comstock and Huntington, terisland endemics are reviewed below in an 1943: 76. abridged descriptive format. A new subspe- CHARACTERIZATION: Forewing apex dark- cies is described from the southern Hispa- *ened only at apex, leaving male forewing niolan paleoisland. Subspecies of N. celida brand surrounded by iridescent blue (figs. 7, differ negligibly in the female genitalia but 1 1). Lateral lobe ofmale genitalia with slight- have distinctive characters of the wings and ly produced sclerotized ridges (figs. 60c, 6 le, in the tergal and valval structures of males. f). 1991 JOHNSON: HAIRSTREAK BUTTERFLY GENERA 9

Jh 3 4 5

)

I 7 ..,I t. t ,. I 7 ,. ...t

A., 4t -A #,o I .-I -1 6 7 8 I

I. AL

9 10 II Figs. 3-1 1. Adults ofNesiostrymon. N. c. celida: 3, 4. Upper surface, under surface, male, Guantan- amo, Cuba (AMNH); 5, 6. Upper surface, under surface, female, same data (AMNH). N. c. aibonito: 7, 8. Upper surface, under surface, holotype male; 9, 10. Upper surface, under surface, allotype female; 11. Upper surface, male, Polo, Barahona Prov., Dominican Republic (AMNH).

TYPES: Holotype male, AMNH labeled terial listed by Schwartz (1989, sexes not not- "Thecla celida aibonito Comstock and Hun- ed) (included in fig. 81): 14 km SW Jamao al tington, holotype male," "holotype," "Ai- Norte, 534 m, Espaillat Prov. (one); 1 km S bonito, P.R., July 14-17, '14." Type Locality: Constanza, La Vega Prov. (one); 16 km NE Aibonito, Puerto Rico. La Romana, 61 m, La Altagracia Prov. (one). DISTRIBUTION (fig. 8 1): Known from mesic PUERTO RICO. See Types entry above; Ai- forest in Puerto Rico and on Hispaniola in bonito, Puerto Rico, July 14-17, 1914 (para- regions on, or directly adjacent to, the north- type male) (AMNH). ern palaeoisland. MATERIAL EXAMINED: In addition to types listed above-DOMINICAN REPUBLIC. Nesiostrymon celida baorucoensis, Broadleaf Cloud Forest, 19 km N Cabral, new subspecies Barahona Province, June 29, 1985, leg. K. Figures 13-16, 59, 60 Johnson and D. Matusik (28) (AMNH); Bam- CHARACTERIZATION: Male forewing brand boo Hole Canyon, nr. Jarabacoa, La Vega and black apex conjoined like the nominate Prov., June 23, 1985, leg. K. Johnson; "Do- but with lateral lobes of male valvae greatly mingo" (16, 29) (BMNH). Unexamined ma- produced along their distal margins. 10 AMERICAN MUSEUM NOVITATES NO. 3011

I ,z;.. i-I ,kj . . ' )= 1; '7

12 -%

/I 13 14I It

-A .4 '%I i tI.I e .1;

c- 1- I t-, *,, ,*.J, 15 16 \. 17

* 18 19 20 Figs. 12-20. Adults of Nesiostrymon. N. c. aibonito: 12. Under surface, male of figure 11. N. c. baorucoensis: 13, 14. Upper surface, under surface, holotype male; 15, 16. Upper surface, under surface, allotype female. N. shoumatoffi: 17, 18. Upper surface, under surface, holotype male; 19, 20. Upper surface, under surface, allotype female. DESCRIPTION: Male. Upper Surface of Length of Forewing: 10.0-12.5 mm. Male Wings: ground color brilliant iridescent azure Tergal Morphology and Genitalia (fig. 60b, blue; forewing with brand and black apical 61g). Typical of species (fig. 61) but with color conjoined; hindwing with long tail, ter- greatly produced lateral valvae lobes. Female minus vein CuA2, short tail, terminus vein Tergal Morphology and Genitalia (fig. 59). CuA 1. Under Surface ofWings: ground color Differing negligibly from the nominate. dingy bluish gray; forewing with black post- TYPEs: Holotype male, allotype female, medial line, costa to cell CuAl; hindwing with "Upper Abejas" (sensu Johnson and Matu- medial band ofnearly conjoined black slash- sik, 1988), July 5-17, 1986, deposited es, a discal and two to three black postbasal AMNH. Paratypes. AMNH: "Las Abejas" slashes and a light blackish crenate line along (see Remarks) July 1, 1984 (16); July 5-17, the outer margin; limbal area with small, or- 1986 (186, 14Q); CMNH: same data but July angish Thecla spot. Length ofForewing: 9.5- 13-18, 1987 (36, 3Q); AME: same data but 12.5 mm. Female. Upper Surface of Wings: June 30-July 3, 1988 (10 specimens); DMC: similar to male but with fuscous marginal same data but July 1-5, 1984 (36, 2Q); July areas more extensive and without forewing 5-17, 1986 (56, 42); June 29-July 3, 1988 (46, brands. Under Surface ofWings: as on males. 4Q). 1991 JOHNSON: HAIRSTREAK BUTTERFLY GENERA I1I

DISTRIBUTION (fig. 81): Known only from talia cephalically recurvate and with caudal mesic broadleaf deciduous forest in deep ra- lamellae distally pronged. Wing under sur- vine areas of the Sierra de Baoruco uplands faces with all markings disjunct and lineal; (1 150-1200 m) in the southwestern Domin- Thecla spot black. ican Republic. Extremely habitat restricted, DESCRIPTION: Male. Upper Surface of occurring only in areas of filtered sunlight in Wings: ground color iridescent sky blue, fore- dense wet woods; to date collected in num- wing with small, elliptic androconial brand bers almost exclusively while perching on contiguous with black wing apex; hindwing Gyrotaenia Grisebach (Urticaceae) (see Bio- with thin black marginal band, long tail, ter- geography). minus vein CuA2, short tail, terminus vein REMARxs: This name is needed for the CuAl. Under Surface ofWings: ground color southern Hispaniolan subunit of the Antil- bright white; forewing with thin, black post- lean distribution of Nesiostrymon. Johnson medial line of disjunct dashes, costa to cell and Matusik (1988) first commented on the CuAl; hindwing with medial band of thin field behavior of N. celida based on popula- black dashes, lightly marked discal and post- tions located in the southwestern Dominican basal dashes and disjunct crenate markings Republic (see Biogeography). Along with the along the outer margin; limbal area with small structural characters differentiating N. celida black Thecla spot. Length ofForewing: 10.0- subspecies, medial under surface stripes are 11.5 mm. Female. Upper Surface of Wings: generally continuous in N. c. baorucoensis and marked as male but without forewing brand. N. c. celida and disjunct in N. c. aibonito. It Under Surface ofWings: marked as male but should also be noted that variation in fore- with yellow occasionally surrounding Thecla wing length in N. c. baorucoensis is probably spot. Length of Forewing: 10.0-11.5 mm. attributable simply to the large sample size. Male Tergal Morphology and Genitalia (fig. ETYMOLOGY: Refers to the Sierra de Bao- 63). Sipc with dorsal terminus short, limited ruco of Hispaniola. to a terminally protruding lobe. Brush organs MATERIAL ExAMINED: In addition to types typical ofgenus. Genitalia with vincular mar- listed above-DOMINICAN REPUBLIC. gin entire, lacking ventral spurs. Valvae with Unexamined material listed by Schwartz lateral, winglike lobes occurring only as (1989) (included in fig. 81): 4-7 km NE El rounded keels distad valval termini; central Aguacate, 519-824 m, Independencia Prov. valval lobes produced, tapering a concave (2); 1 km SE Banano, 143 m, Pedernales Prov. terminus with short distal prongs. Extraval- (2); Las Abejas, 11 km NW Aceitillar, 1220 vular element more elongate than in conge- m, Pedernales Prov. (1); 1 km N Aceitillar, ners. Saccus elongate, length exceeding cau- 1281 m, Pedernales Prov. (1). doventral expanse of vincular arc. Aedeagus elongate, nearly three times caudoventral expanse ofvincular arc, and with shaft curvate; Nesiostrymon shoumatojfi caecum comprising one-seventh or less ofae- (Comstock and Huntington), deagal length. Flat valval ventrum more sim- revised status ilar to Terra than to congeners (see Discus- Figures 17-20, 62, 63 sion below). Female Tergal Morphology and Theca celida shoumatoffi Comstock and Hun- Genitalia (fig. 62). Tergite 8 typical ofgenus. tington, 1943: 75; 1958-1964 [1963]: 119. - Papillae anales with thick, marginally in- Kaye, 1931: 535, pl. 39, fig. 10. - Schwartz, cised, lateral sclerotizations and pointed ter- 1989: 240. minus. Genitalia with caudal element ofduc- Nesiostrymon celida shoumatoffi: Clench, 1964: tus bursae greatly elongate (exceeding length 248, 251. - Brown and Heinemann, 1971: 4; ofcephalic element by four to five times) and 1972: 232. - Scott, 1986: 21. - Riley, 1975: with cephalic element curvate; terminal la- 101. - Bridges, 1988: I.319, 11.77.- Johnson mellae flared widely and distally pronounced and Matusik, 1988: 236. into two terminally directed prongs. Overall DiAGNOSIS: Male tergite 8 terminating in genital shape and terminal elements more small protruding dorsal lobe; genital valvae reminiscent of mainland N. celona than An- with little lateral production. Female geni- tillean N. celida (see Discussion below). 12 AMERICAN MUSEUM NOVITATES NO. 3011

TYPEs: Holotype male, allotype female, (MNHN); Wilson Run, Trelawny Parish, June AMNH, respectively labeled "Thecla celida 29, 1951, leg. H. Heineman (19) (AMNH); shoumatoffi Comstock and Huntington, ho- Polly Ground, track to Bat Cave, 6 February lotype male," "holotype," "Christiana, Ja- 1957, leg. H. Heineman (19) (AMNH). Listed maica, 26 vii 1933," "collectors A. Avinoff by Brown and Heineman (1972) but not lo- and N. Shoumatoff;" same, but "July 26 1933 cated in this study: St. Thomas, Corn Puss [sic]." Type Locality: Christiana, Jamaica. Gap, July 12, 1949, leg. "I.J.," Cuna Cuna, DISTRIBUTION (fig. 81): Primarily Jamaican July 15, 1933 (sighted), leg. Avinoff and uplands (Blue Mountains and Cockpit coun- Shoumatoff; St. Catherine, Bog Walk at Flat try) in deciduous forest and forest openings Bridge, June 7, 1959, leg. Bengry. (see Discussion). DISCussIoN: Hitherto, based on simple wing MAINLAN TAXA pattern comparison, this species had been Nesiostrymon celona (Hewitson), considered a subspecies ofN. celida (Clench, new combination 1964; Riley, 1975; Schwartz, 1989). How- Figures 21-26, 64, 65 ever, when wing pattern and morphology are considered in context with all congeners, it is clear that N. shoumatofij is a distinct spe- Thecla celona Hewitson, 1863-1878 [1874]: 179. cies. The male tergite 8 and genital valves - Comstock and Huntington, 1958-1964 exhibit a primitive condition. Flat ventral [1959]: 181. - Bridges, 1988: I.75, 11.77. - surfaces on the valvae resemble sister genus Johnson and Matusik, 1988: 236. Terra; overall shapes of the female ductus DIAGNosIs: Upper surface ground color in bursae and male aedeagus are similar to those males dark azure blue with lustrous black in mainland congener N. celona. As a result, ovate brand dominating distal end of fore- species status for N. shoumatoffi appears un- wing discal cell; female dull blue to brownish. equivocal, irrespective of the systematic Under surface of both sexes with wide, con- method used to cluster taxa of the group. tinuous, lunular orange-brown medial bands Brown and Heineman (1972) noted the extending across forewing and hindwing. Male "rarity" ofN. shoumatoffi. Historical records genitalia with lateral lobes elongate (length primarily indicate upland deciduous forest equaling that of central valval lobes) and ta- habitats and surrounding margins; some old pering to pointed termini. Compared to in- specimens have data suggesting more low- sular mainland congeners N. milleri and N. land locales and may not be reliable. As with australivaga, N. milleri with under surface N. celida, small historical samples likely re- bands discontinuous, a unique postbasal band sult from collectors not locating the species' and male genitalia with elliptic valval lobes microhabitat. Remarks in Brown and Heine- flanked by lobate vincular spurs; N. austra- man (1972) recorded the species from "bush- livaga with under surface bands straight, thin, es." The latter may be Gyrotaenia, noted basally directed and suffused, male genitalia heretofore as the perching substrate for N. with diminutive lateral lobes and female gen- celida. italia robust as in N. celida. MATERIAL ExAMIED: In addition to types DESCRIPTION: Male. Upper Surface of listed above-JAMAICA. Newcastle (26) Wings: ground color dark iridescent azure (BMNH); "Jamaica" (18) (BMNH); "Jamai- bordered by black margins, apices and sub- ca," 1914, leg. Avinoff (16) (CMNH); Flam- apices. Forewing with lustrous black andro- stead, 1883, 3900 ft (AME); Allsides, July conial brand outstanding over azure ground 26, 1933, leg. "A & S" (16) (CMNH); Low color; hindwing with long tail at vein CuA2 River at Trelawny, July 30, 1933 (36, para- terminus, short tail at vein CuAl terminus. types) (CMNH); Christiana, July 26, 1933, Under Surface of Wings: ground color dis- leg. Avinoff and Shoumatoff (46, paratypes) tally light gray, basally gray-brown, divided (CMNH); Coleyville, Manchester, July 29, on both wings by a lunular (sometimes jag- 1933 (56, paratypes) (CMNH); Moneague, St. ged) orange-brown medial band bordered Ann's, July 21, 1933 (68) (CMNH); Mt. Dia- distally with black; submargins on both wings blo, July 10, 1936 (96); "Jamaica" (1Q) with white-edged brown dashes occurring in 1991 JOHNSON: HAIRSTREAK BUTTERFLY GENERA 13

-v.

, I 21 22 23

row

4 *.1

26, 24 26 25 vt,v.

27 28 29 -

".... .A .. .I s ; .N , , 30 31 32 Figs. 21-32. Adults ofNesiostrymon. N. celona: 21, 22. Upper surface, under surface, topotype male; 23, 24. Upper surface, under surface, topotype female; 25, 26. Upper surface, under surface, male, Jalapa, Mexico (AMNH). N. milleri: 27, 28. Upper surface, under surface, holotype male. N. australivaga: 29, 30. Upper surface, under surface, holotype male; 31, 32. Upper surface, under surface, allotype female. each cell; limbal area with small orange to brands. Under Surface of Wings: similar to brownish Thecla spot. Length of Forewing: that of males but with bands more pro- 8.0-11.0 mm. Female. Upper Surface of nounced and often more jagged. Length of Wings: ground color dull iridescent blue to Forewing: 9.0-11.0 mm. Male Tergal Mor- brownish; fuscous marginal areas more ex- phology and Genitalia (fig. 65). Sipc with dor- tensive than in males and forewings lacking soterminal prong ofmoderate length (slightly 14 AMERICAN MUSEUM NOVITATES NO. 3011

ti 33 34 35 r t?; .t -I I I pN .6 36 37 38

'P.;. I I

C-1 'i t IV O'

I z.' 39 40 41 Figs. 33-41. Adults of Terra. T. hispaniola: 33, 34. Upper surface, under surface, allotype female; 35, 36. Upper surface, under surface, male, "Middle Abejas," Pedernales Prov., Dominican Republic, July 30, 1990, leg. D. Matusik (AMNH). T. tera: 37, 38. Upper surface, under surface, topotype male; 39, 40. Upper surface, under surface, topotype female. T. andevaga: 41. Upper surface, holotype male. exceeding one-halfthe tergite dorsum width). duced, forming a pair of terminally directed Brush organs typical ofgenus. Genitalia with prongs as in N. shoumatoffi and N. australi- vincular margins smooth, lacking ventral vaga. spurs. Valvae with lateral lobes steeply point- TYPES: Holotype male, BMNH, labeled ed and lengths ranging from four-fifths to "Espiritu Santo, Hewitson Coll. 79-69. equal that of the central valval lobes. Saccus Thecla celona. 1.," "male, Thecla Type ce- robust, shorter than most congeners, with lona," "B.M. Type No. Rh. 981." Type Lo- length usually less than that of vincular arc. cality: Espiritu Santo, Brazil. Aedeagus elongate as typical of genus and DISTRIBUTION (fig. 81): From east central recurvate in the terminal one-third, caecum Mexico southward, excluding the Amazon comprising only one-seventh to one-eighth basin, to southeastern Brazil and central Peru. aedeagal length. Female Tergal Morphology DISCUSSION: There is remarkably little and Genitalia (fig. 64). Tergite 8 typical of variation in the male and female morphology genus. Papillae anales with pointed terminus ofspecimens from central Mexico southward typical ofmainland congeners. Genitalia with to southeastern Brazil. There appears to be ductus bursae more fluted than in congeners, little doubt that all these populations repre- transparent neck restricted and located rela- sent the same species (figs. 25, 26). Though tively close to a swollen cephalic terminus. this character homogeneity is in marked con- Lateral edges ofductus bursae terminus pro- trast to similarly distributed populations of 1991 JOHNSON: HAIRSTREAK BUTTERFLY GENERA 15

Terra (see below), there are distinctive main- tos, leg. E. May (18) (BMNH); Tarapota (18) land congeners of Nesiostrymon occurring (BMNH). sympatric with N. celona on the Guyana Shield (N. milleri, below) and in montane Nesiostrymon new species central Argentina (N. australivaga, below). milleri, Contrasting Antillean congeners, but par- Figures 27, 28, 66 alleling the historical records of Terra spe- DIAGNosIS: Upper surface iridescent blue cies, females of mainland Nesiostrymon taxa darkest of congeners, with male forewing are poorly known. However, their characters brands located more distally and obscured by figure heavily in the evaluation of relation- the iridescent ground basad thin black apical ships (figs. 1-2, tables 1-4). Because ofjagged borders. Under surface with an iridescent blue under surface bands, black and white photos sheen over gray ground color; hindwing with of some N. celona females may resemble ei- postbasal line paralleling the medial band and ther sex of certain species of Caerofethra ground color between these distinctly darker Johnson, particularly C. iambe (Godman and gray than rest of wing. Male genitalia with Salvin) (fig. 84). However, the under surfaces lateral lobes widely elliptical, length about ofCaerofethra taxa have tawny ground colors three-fourths that ofcentral valval lobe; vin- and bright orange wing bands. culum with unique ventrolateral spurs. MATERIAL ExAMINED: In addition to type DESCRIPTION: Male. Upper Surface of listed above -BOLIVIA. Rio Surutu, April Wings: ground color deep iridescent navy 1915 (18), October 1913 (16) (both CMNH); blue; forewing with blackened androconial Rio Yapacani, 600 m, Sept. 1914 (16) brand located at extreme distal end of discal (CMNH); Rio Yapacani, E. Bolivia (16) area; hindwing with long tail at vein CuA2 (CMNH). BRAZIL. Gavea, Parana, 1 Feb- terminus, short tail at vein CuAl terminus. ruary 1930 (16) (MPM); Amazonas, Tonan- Under Surface of Wings: ground color dull tins, July-September 1880, leg. de Mathan gray with iridescent blue sheen; forewing with (18) (BMNH); Santarem, December 20, 1922, discontinuous black postmedial line, costa to leg. C. S. Larsen (MNHN). COLOMBIA. Cali, cell CuA 1. Hindwing with discontinuous Cauca, April 3, 1909, leg. Paines and Brink- black medial band paralleled by a prominent ley (16) (BMNH); Novelle Granada, leg. de black discal line extending from cell RS to Methan, September 15, 1990 (12) (BMNH); CuA I and suffused basally with dark brown. Cauca, Cali, April 3, 1909, Paines and Brink- Submargin with white dashes in each cell; ley (16) (BMNH); Puerto Colombia, July 11- limbal area with black Thecla spot slightly 12, 1920, dry scrub area (18) (CMNH). EC- bordered with orange. Length of Forewing: UADOR. Baiios, 1800 m, Tungurahua, July 10.0 mm (holotype). Female. Unknown. Male 1936 (18) (BMNH); Puyo, Napo, 1000 m De- Tergal Morphology and Genitalia (fig. 66). cember 14, 1938 (18) (BMNH); Balzapamba Sipc typical ofgenus but with dorsoterminal (16) (BMNH). MEXICO. Catemaco, Vera- prong robust, moderate in length (length two cruz, May 1961, leg. Escalante (16) (AME); times maximal width) and blunt at terminus. Yucatan, Piste, July 15, 1959, leg. E. C. Wel- Brush organs typical ofgenus. Genitalia with ling (26) (CMNH); Yucatan, Quintana Roo, ventrolateral margin ofvinculum marked by leg. E. C. Welling, November 17, 1955 (16) prominent spurs overlapping ventroterminal (CMNH); Colima (16, 19) (CMNH); Colima surface of valvae. Valvae with lateral lobes (16) (HEC); S. Mexico, leg. Townsend (18) winglike, with elliptical termini, lengths ex- (CMNH); Quintana Roo, Yucatan, Novem- ceeding three-fourths that of the central val- ber 17, 1957, leg. Welling (18) (CMNH); Pre- val lobes. Saccus robust, length about equal sidio, April 1940, leg. Hoffman (6) (AMNH); to that ofvincular arc. Aedeagus robust, length Catemaco, May 1961 (6) Escalante (AMNH); exceeding rest ofgenitalia by only about one- Piste, Yucatan, August 15, 1960 (26) fifth, caecum comprising about one-fifth the (AMNH); "locality 5092," Moeck (16) length and undulate in terminal one-half. (MPM); Chiapas, leg. Hoffman (18) (AMNH). TYpPEs: Holotype, male, Aragua, Porto- PANAMA. Summit, Canal Zone, May 3, chuelo Pass, Parque Nacional Henri Wittier, 1964, leg. Shull (16) (CMNH). PERU. Iqui- (Rancho Grande) Cloud Forest, July 24, 1981, 16 AMERICAN MUSEUM NOVITATES NO. 3011

Station VE1O, Venezuela, leg. Lee D. Miller, minutive and lateral aspect robust. Saccus deposited AME. robust, more elongate than in N. celona (ex- DISTRIBUTION (fig. 81): Known only from ceeding length of vincular arc). Aedeagus the type locality. elongate as typical of genus and terminally DISCUSSION: Apparent insular occurrence curvate, caecum comprising about one-sixth of N. milleri reflects a distribution pattern aedeagal length. Brush organs typical of ge- seen in numerous recent revisionary studies nus. Female Genitalia (fig. 67). Ductal ele- of Neotropical hairstreak genera-a species ments tapering as in N. celona but with length oflimited distribution on the Guyana Shield of caudal element slightly more than twice sympatric there with a congener ofmuch wid- that of cephalic element (caudal/cephalic ra- er geographic range (Johnson 1988a, 1989a, tio 0.55); cephalic element with slight dorsal 1989b, 1989e). inclination. Ductal terminus with short, ter- ETYMOLOGY: Patronym for Lee D. Miller, minally directed, lateral prongs as in N. ce- who collected the type. lona and N. shoumatoffi. Papillae anales typical of mainland congeners. TYPES: Holotype male, allotype female, Nesiostrymon australivaga, new species "Mendoza, Argentina" (see Discussion be- Figures 29-32, 67, 68 low), ex. C. S. Larsen Collection, deposited DiAGNosIs: Males with upper surface fus- MNHN. Paratype. MNHN: same data as pri- cous wing borders extremely wide, obscuring mary types (one female). the forewing brands; females brown. Under DISTRIBUTION (fig. 81): Known only from surface medial bands thin, straight, and ba- "Mendoza" in Argentina (see Discussion be- sally directed. Male genitalia with central and low). lateral valval lobes diminutive; female gen- DISCUSSION: Recent studies of temperate italia with tapered configuration like N. ce- zone South American Theclinae indicate a lona but relative size of ductal elements re- distinctive endemic fauna (Johnson, 1988a, sembling that of N. celida (see below). 1989a, 1989b, 1989e; Johnson et al., 1988, DESCRIPTION: Male. Upper Surface of 1990). Representatives of this fauna derive Wings: ground color darkly suffused irides- both from recent collections and ill-studied cent blue bordered by wide blackened apices samples forwarded by early workers to var- and subapices obscuring the ovate forewing ious European museums. Notable among brands. Hindwing with long tail at vein CuA2 these are Argentine, Patagonian, and Chilean terminus, short tail at vein CuAl terminus. collections by C. S. Larsen and R. Martin at Under Surface of Wings: ground color gray- the MNHN and Argentine collections by K. brown; forewing with thin, straight, basally Hayward and E. Giacomelli at the BMNH. directed brown medial band; hindwing with The types of N. australivaga derive from the similar, very lineal band; limbal area with C. S. Larsen material. As detailed by Johnson prominent yellow-orange Thecla spot marked (1989e, in press), MNHN curator Percy Lathy centrally with black dot. Length ofForewing: described several species from early C. S. Lar- 9.0 mm (holotype). Female. Upper Surface sen material. However, many other austral ofWings: ground color darkly suffused brown specimens were never incorporated into the with wide, blackened apices and subapices MNHN collection. Johnson et al. (1990) have and no forewing brand. Under Surface of noted that "Mendoza" data on C. S. Larsen Wings: similar to that of males. Length of specimens may not correspond to the current Forewing: 10.0 mm (allotype); 10.0 mm Mendoza Province of Argentina. Specific lo- (paratype). Male Tergal Morphology and calities listed on Larsen labels including the Genitalia (fig. 68). Sipc typical of genus but notation "Mendoza" include sites north to with length of dorsoterminal prong slightly modem La Rioja Province and south to Rio exceeding that of N. celona and with base of Negro Province. prong greatly produced and then tapered ETYMOLOGY: The suffix vaga ("roamer") is steeply to a sharp terminus; lateral edges of added to the geographic term "austral," re- prong with prominent microtrichia. Genita- ferring to the southerly occurrence ofthe spe- lia with valval central and lateral lobes di- cies. 1991 JOHNSON: HAIRSTREAK BUTTERFLY GENERA 17

1,, ~4.

I 4 i 42 43 f 44 J

4., '4.'

45 46 47

A 6,. le 4, I - -i

48 49 50 Figs. 42-50. Adults of Terra. T. andevaga: 42. Under surface, holotype male; 43, 44. Upper surface, under surface, allotype female. T. cana: 45, 46. Upper surface, male, "Tucuman," leg. Steinbach, (BMNH); under surface, male, "El Paraiso," Tucumin Province, Argentina, December 31, 1987, leg. R. Eisele (AMNH); 47, 48. Upper surface, under surface, female, Villa Nougues, Tucuman Province, January 21, 1931, leg. K. J. Hayward (IML). T. calchinia: 49, 50. Upper surface, under surface, male, Ega, Brazil (BMNH) [black and white reproduction of color slide].

GENUS TERRA JOHNSON AND MATUSIK yellow, orange, brown or red, not black as in Figures 33-56, 58, 69-80 Nesiostrymon. DEscRIPrION: Adult. Body blackish, cov- ered with gray to bluish hairlike microtrichia; Terra Johnson and Matusik, 1988: 235. - eyes ringed with white; antennae black, finely Schwartz, 1989: 241. striped with white. Male. Upper Surface of DIAGNOsIs: Compared to all Eumaeini, fe- Wings: ground color dull iridescent blue to male genitalia with an outstanding, ventrally violet with fuscous borders ofvarious width; directed, sclerotized bulb (figs. 69-80) at- forewings with elliptic, gray to black, andro- tached to the transparent neck of the ductus conial brands variously conjoined to the fus- bursae. Compared to Nesiostrymon, male ter- cous distal ground; hindwing with long tail gite 8 without sipc and with genital valvae at vein CuA2 terminus, short tail, vein CuA 1 uniplanar and ventroterminally flat. Hind- terminus. Under Surface of Wings: ground wing under surfaces with medial bands of colors ranging from white to dull gray-brown 18 AMERICAN MUSEUM NOVITATES NO. 3011

contrasting fuscous-edged bands of yellow, species. Corpus bursae with two small, orange, brown or red- forewing with disjunct pronglike signa. or continuous band extending from costa to TYPE SPECIES: Theca tera Hewitson (1863- inner margin, hindwing with medial band 1878 [1878]) by original designation. comprised of variously pronounced lunular DISTRIBUTION: Primarily mainland with and/or lineal elements usually angled in a five generally allopatric species replacing each W-shape toward the anal margin. Hindwing other in ranges from central Mexico south limbal area with bright yellow, orange, or red- through the Andes and Amazon headwaters dish Thecla-spot bordered by various suffu- to montane northern Argentina and south- sive marks. Length of Forewing: 10.0-14.0 eastern Brazil; one species endemic to south- mm. Female. Upper Surface of Wings: sim- ern Hispaniola. ilar to that ofmale but fuscous margins wider DIscussIoN: Terra species are grouped be- and forewings lacking brands. Under Surface low according to the two nested sets of taxa of Wings: similar to that of males but often apparent in the cladistic analysis-Antillean with wing bands more pronounced. Length and mainland. Of the five members com- of Forewing: 10.0-14.0 mm. Male Tergal prising the mainland set, three allopatric taxa Morphology and Genitalia (figs. 58, 72-74, (T. tera, T. andevaga, T. cana) exhibit lu- 78-80). Tergite 8 unspecialized (lacking sub- nulate yellow to orange under surface bands cordate incised posterior cavity), terminal and are referred to as the "yellow-banded margin entire and edged with thin, elongate species." Wing patterns in these species differ microtrichia. Dorsoterminal surface of gen- mostly in the shape and extent of bands and italia with two small brush organs abutting occurrence of adjacent yellow or orange pat- lateral edge of vinculum; length of brushes tern elements in the discal area. Of the yel- not exceeding terminus of labides. Genitalia low-banded members, T. tera is the most with lateral surface ofvinculum more angled widely distributed, reminiscent of N. celona than in Nesiostrymon, saccus thin and elon- in Nesiostrymon. Other yellow-banded mem- gately parabolic. Valvae diminutive, length bers are regionally isolated. Wing patterns of not exceeding caudoventral diameter of vin- remaining mainland congeners (T. chilica, T. cular arc (fig. 57), valval ventrum uniplanar calchinia) exhibit bold brown or red bands and mostly flat. Valvae termini occurring as with numerous additional pattern elements two, centrally located caudal extensions (fig. in the adjacent basal or distal areas. These 58) each terminating with elongate, proxi- latter taxa are grouped below as the "red- or mally protruding, microtrichia; distad these, brown-banded species." a lightly sclerotized extravalvular element. It is important to note for diagnostic pur- Bilobed areas of valvae (fig. 58) with bilat- poses that color and pattern differences in erally symmetrical patches of nearly trans- Terra taxa, though striking in actual speci- parent sclerotin, ventral valval shapes and mens, become less evident in black and white lengths of caudal extensions varying greatly photographs. Lightly colored under surfaces between species. Aedeagus length usually ex- (white in some Terra species) require a fast ceeding rest ofgenitalia by one-fourth to one- shutter speed favoring pattern detail over third, shaft generally straight, caecum com- ground contrast. As a result, photographs of prising one-fourth to one-fifth aedeagal length. some under surface patterns (the photos with Terminus of aedeagus with two, spinelike, darker backgrounds, figs. 33-56) appear more cornuti. Female Tergal Morphology and uniform than the actual specimens and vivid Genitalia (figs. 60-71, 75-77). Tergite 8 un- contrasts between brown and blue upper sur- specialized but papillae anales heavily scler- face colors and white wing fringes are not otized and tapering steeply to a point. Gen- always apparent (figs. 53, 55). In addition, italia with sclerotized cephalic and caudal pattern (but not color) similarities occur be- ductal elements joined by a constricted, tween species of Terra and some outgroup transparent neck; directed ventrally from taxa. Confusing instances are pointed out in transparent neck, a prominent sclerotized Discussion sections of appropriate species bulb (the "bulb," figs. 69-71, 75-77). Rela- entries. tive lengths and breadths of ductal elements Cladistic analysis indicates little structural and adjacent bulb differ greatly between the homogeneity among the wing pattern groups 1991 JOHNSON: HAIRSTREAK BUTTERFLY GENERA 19

ai I 'r, ,V, '? I: ,rX. lb '.- 0- ,7 51K; ~- 52 53

I 10 V7 t ( r.4. 1 ¾. .\ 1 V., L }' j 54 55 56 Figs. 51-56. Adults of Terra. T. calchinia: 51, 52. Upper surface, under surface, female, "Amazon" (HEC). T. chilica: 53, 54. Upper surface, under surface, male, Rolandia, Brazil (AMNH); 55, 56. Upper surface, under surface, female, same data (AMNH). comprising Terra. For instance, red-banded DIAGNOSIS: Superficially similar to yellow- T. calchinia has a square-shaped valvae like banded mainland congeners reviewed below. that in yellow-banded T. tera but brown- From these, differing by wings' duller upper banded T. chilica exhibits parabolic valvae surface lavender color and under surface ob- like that in yellow-banded species T. cana solescent bands in which lack ofan outstand- and T. hispaniola. Remarkably, one yellow- ing medial mark in cell RS causes the band banded species (T. hispaniola) stands out to appear basally directed from vein M2 to structurally in the cladistic analysis. Such het- costa. Female genitalia robust, caudal ductal erogeneity results in my treating Terra as six element uniquely larger than cephalic ele- species. Johnson and Matusik (1988) includ- ment with ventral bulb thickly conjoined to ed a sixth taxon (Thecla hyccara Hewitson) latter element. Male genitalia more typical of in a list ofprobable Terra species. The female genus but with diminutive, elliptical valvae of T. hyccara was unknown at the time; sub- terminating with a slightly recurvate caudal sequent inclusion of male and female char- extension. acters in the cladistic analysis (fig. 85) indi- DESCRIPTION: Male. Upper Surface of cated T. hyccara belonged to another, mostly Wings: ground color dull violet blue, margins undescribed generic level outgroup of Ne- and apices suffused black; forewing with out- siostrymon and Terra (fig. 85). standing elliptical brand; hindwing with long tail at vein CuA2 terminus, short tail at vein CuAl terminus. Under Surface of Wings: ANTILLEAN TAXA ground color, gray-white, forewing with gray- suffused postmedial line, costa to cell CuA1; Terra hispaniola Johnson and Matusik hindwing with medial band ofdiscontinuous Figures 33-36, 69, 72 yellow or yellow-gray patches, darkly suffused at edges, with no outstanding medial Terra hispaniola Johnson and Matusik, 1988: 241. mark in cell RS causing band to appear ba- - Schwartz, 1989: 241. sally directed from vein M2 to costa; limbal 20 AMERICAN MUSEUM NOVITATES NO. 3011 area with light yellow Thecla spot centered upland and xeric than N. celida. T. hispaniola with a black dot. Length of Forewing: 11.0- is more vagile than N. celida. At Las Abejas, 13.0 mm. Female. Upper surface of Wings: in contrast to the perching habits ofN. celida similar to that of male but fuscous marginal (see its entry and Biogeography), T. hispanio- areas wider and forewings without brands. la is a limited patroller, ranging from areas Under Surface of Wings: similar to that of of broken canopy at 1190-1200 m altitude males. Length of Forewing: 12.5 mm (allo- to the margin with pine forest at 1250 m type). Male TergalMorphology and Genitalia altitude. To date, T. hispaniola is known from (fig. 72). Tergite 8, brush organs, and genital nine males and one female. Such disparity of vinculum, saccus, and aedeagus typical ofge- males and females parallels that seen in series nus. Valvae diminutive, elliptical in bilobed of T. cana, the only other Terra species re- area, tapered terminally to slightly recurvate cently collected in numbers. These circum- caudal extensions. Female Tergal Morphol- stances may indicate females of Terra species ogy and Genitalia (fig. 69). Tergite 8 and pa- are more habitat specific than patrolling pillae anales typical of genus. Genitalia with males. Herein, along with the allotype female length ofcephalic ductal element nearly twice of T. hispaniola (figs. 33, 34), I illustrate a that of caudal element and with both ele- recently caught male with reduced under sur- ments very robust (cephalic element length face bands (figs. 35, 36). only 6-7 times width compared to 15-21 MATERLAL ExAMIED: In addition to types times length incongeners). Bulb located rath- listed above-DOMINICAN REPUBLIC. er flush with terminus of cephalic ductal el- Same locality as primary types, July 5, 1984, ement, not on semidetached stalk typical of leg. D. Matusik (16) (AMNH/HS 108 trans- congeners. ferred to DMC) (DMC), July 30, 1990 (16) TYPES: Holotype male, allotype female, (AMNH, figs. 35, 36), (1 ) (AME),16 (DMC); CMNH. Type Locality: "Middle Abejas" "Upper Abejas" (sensu Johnson and Matu- [sensu Johnson and Matusik, 1988], Las Abe- sik, 1988), patrolling path, July 13, 1987, leg. jas, Pedernales Province, Dominican Repub- K. Johnson, on lifetime loan to Albert lic, July 6, 1986, leg. K. Johnson (genitalic Schwartz (18) (AMNH/HS 109 transferred to dissection "AMNH/HS #106, 107," trans- AME) (AME); same location as primary types, ferred to CMNH). nectaring on small blue flowers in sunlight, DISTRIBUTION (fig. 82): Known only from July 15, 1987, leg. K. Johnson (18) (AMNH/ the Las Abejas rain forest in the Sierra de HS 110) (AMNH); "Upper Abejas," June 29, Baoruco of the Dominican Republic (south- 1988, leg. D. Matusik (18) (DMC), June 30, ern paleoisland). 1988, leg. K. Johnson (16) (AME). DIscussIoN: The original description did not mention the basal displacement of the under surface hindwing spot in cell M1. This MAmILAND TAxA character became apparent in studying this species in a revisionary context and exam- As previously noted, structurally hetero- ining additional specimens. It is also appar- geneous mainland taxa are divisible into two ent from recent collections (figs. 33, 34) that groups based on color of the under surface yellow banding can be very reduced on some bands. Since yellow-banded species are most specimens. Because these specimens resem- similar to T. hispaniola, they are treated first. ble Nesiostrymon celida, we suspect T. hispaniola may have been collected by early workers and misidentified as "Thecla" celi- Yellow-banded Species da. As a result, curators should carefully ex- Terra tera (Hewitson) amine series of N. celida for possible speci- Figures 37-40, 70-73 mens of T. hispaniola. Thecla tera Hewitson, 1863-1878 [1878, vol. 1]: Johnson and Matusik (1988) commented 211; [1878, vol. 2]: pl. 84, figs. 714, 715. - first on the field ecology of T. hispaniola. Its Godman and Salvin, 1879-1901 [1887, vol. 2]: occurrence is limited to mesic broadleaf de- 86. - Hoffman, 1940: 717.- Comstock and ciduous forest, where it inhabits areas more Huntington, 1958-1964 [1964]: 124. - Clench, 1991 JOHNSON: HAIRSTREAK BUTTERFLY GENERA 21 k 58

N.e ce--- ~~~~~~~~~ce--

Ce- - Ni bo- -

b~~~~~~~~~~~~~~~~~~

d Figs. 57, 58. Generalized male genital and tergal structures of Nesiostrymon (57) and Terra (58). a. ventral view, genitalia with aedeagus removed and placed along side (aa); b. lateral view, genitalia; c. lateral view, terminal tergite and microtrichia; d. dorsal view, terminal tergite and microtrichia. ce, caudal extension of valvae (= cl, central lobe, in Nesiostrymon); ee, extravalvular element; bo, brush organs; N 1, caudoventral diameter ofvincular arc; N2, radius ofvincular arc; stipples, areas ofrelatively transparent sclerotization.

1964:251. - Ross, 1976:198. - Bridges, 1988: DIAGNOSIS: Differing from other yellow- 1.344, II.77. banded mainland species (T. cana, T. an- Terra tera: Johnson and Matusik, 1988: 235 (as devaga) by (i) upper surface iridescence dull type species). blue with fuscous borders comparatively nar- 22 AMERICAN MUSEUM NOVITATES NO. 3011

61 (cont.)

Figs. 59-61. Morphology of Nesiostrymon celida. 59. Female. Genitalia. a. ventral; b. lateral; c. terminal tergite, papillae anales, and microtrichia. 60, 61. Male. Terminal tergite in N. celida subspecies. (60). a. nominate, lateral view (left), dorsal view (right); b. N. c. baocucoensis, dorsal; c. N. c. aibonito, dorsal. Valvae in N. celida subspecies. (61). a. nominate, ventral (both lobes, with midline dashed) (left), lateral (right); b. N. c. aibonito, Puerto Rico, ventral (left of midline only) (left), lateral (right); c. N. c. aibonito, Hispaniola, same views; d. N. c. baorucoensis, same views. tn, transparent neck; stipples, areas of relatively transparent sclerotization. 1991 JOHNSON: HAIRSTREAK BUTTERFLY GENERA 23 row (others with wide apical, subapical, or element diminutive lamellae; cephalic ele- submarginal fuscous borders), (ii) under sur- ment flared widely and with slight dorsal in- face medial band generally continuous from clination. Bulb robust and elliptical, length cell Ml to anal margin with adjacent discal comprising about one-fifth length of caudal streak obsolescent or absent. Male genitalia ductal element. with valval base robust and square-shaped, TYPES: Holotype male, ZMH, labeled "Or- caudal extensions (comprising up to one-third igin," "Thecla tera Hew.," "Coll. Stgr.," valval length) prominent compared to con- "Zool. Mus. Berlin" (Johnson and Matusik, geners; female genitalia with ductal elements 1988: 242 was an inadvertent listing error). most robust of mainland taxa, length of ce- Type Locality: original description lists Chi- phalic element exceeding that of caudal ele- riqui, Panama. ment by only about one-half (former greatly DISTRIBUTION (fig. 82): Known from Ve- elongate in T. andevaga and the red- and racruz State, Mexico, southward to the Cauca brown-banded species). Valley of Colombia. DESCRIPTION: Male. Upper Surface of DISCUSSION: In an index, Bridges (1988) Wings: ground color dull iridescent blue with included tera under Nesiostrymon, based thin black marginal and apical borders. Fore- probably on Clench's (1964) remarks. Con- wing with postmedial elliptic gray brand sur- sistent with my taxonomic discussion under rounded by iridescent blue. Hindwing with the generic entry, and because Bridges did long tail at vein CuA2 terminus, short tail at not use formal binomials, I do not include vein CuAl terminus. Under Surface ofWings: his index citation in the synonymy. Ross ground color chalky white; forewing with thin (1976) noted the habitat of T. tera occurring yellow-orange postmedial line, narrowly bor- in the Sierra de Tuxla of Yucatan, Mexico, dered with black and sometimes with adja- as "Lower Montane Rain Forest" character- cent gray discal slash; hindwing with gener- ized by a Terminalia-Dalbergia association ally continuous medial band of elongate at 50-3000 m. This habitat was noted as drier yellow-orange patches bordered distally with than upland "Montane Forest or Cloud For- black. Discal area sometimes with light gray est." This xeric affinity of Terra is consistent or obsolescent slash; outer marginal line with with our observations ofNesiostrymon celida dull crenate markings; limbal area with and Terra hispaniola in the Antilles (see Bio- prominent yellow Thecla spot centered with geography) but does not appear true of T. black dot. Length ofForewing: 11.0-13.0 mm. andevaga (noted herein from "submacroth- Female. Upper Surface of Wings: similar to ermic rain forest") and T. cana (label data that of male but with fuscous margins wider indicating "mesic to hydric forest"). Thus, and forewing without brands. Under Surface some habitat segregation may be active in of Wings: similar to that of male. Length of certain Terra taxa (see T. andevaga Discus- Forewing: 11.0-13.0 mm. Male Tergal Mor- sion). phology and Genitalia (fig. 57, 73). Tergite 8, MATERIAL EXAMINED: In addition to type brush organs, genital vinculum, and saccus listed above-COLOMBIA. Rio Sucio (1Q) typical of the genus. Valvae ventrum robust (BMNH); Q. Tasajera, 1300 m, February 19, with square base and prominent caudal ex- 1949 (18) (AMNH). COSTA RICA. Iraziu, 6- tensions comprising up to one-third entire 7000 ft (26, IQ) (BMNH); Costa Rica (16) valval length. Aedeagus shaft bowed, caecum (BMNH). MEXICO. Catemaco, Vera Cruz, comprising about one-fourth aedeagal length. September 1957, leg. Escalante (16) (AME); Female TergalMorphology and Genitalia (fig. Chiapas, no other data (36), September 1955 70). Tergite 8 and papillae anales typical of (16), September 1949 (16), August 1942 (16) genus. Genitalia with ductal elements robust (all AME); Jalapa, 4600 ft, November 1912, compared to mainland congeners: caudal el- leg. A. Hall (2Q) (AME); Presidio, Vera Cruz, ement length about one-half that of cephalic December 1935, leg. Escalante (16), May element; cephalic element thin (diameter 1945, leg. Escalante (1Q), May 1948, leg. Es- about one-half that of caudal element) but calante (16), April 1942, leg. Escalante (16) not greatly elongate as in T. andevaga, T. (all AME); San Lui Potosi, May 1973, leg. H. calchinia, and T. chilica. Lamellae of caudal A. Freeman (16) (AME), June 2, 1973, leg. 24 AMERICAN MUSEUM NOVITATES NO. 3011

Freeman (1) (both AME); Santa Rosa Comi- wing discal cell. Under Surface of Wings: as tan, May 1960, leg. Escalante (18) (AME). on males. Length ofForewing: 13.0 mm-14.0 GUATEMALA. S. Geronimo, leg. Cham- mm. Male Tergal Morphology and Genitalia pion (12) (BMNH). NICARAGUA. Jinotaga, (fig. 74). Tergite 8, brush organs and genital November 1905, wet season, 3400 ft (18) vinculum, saccus, and aedeagus typical ofge- (BMNH). PANAMA. Chiriqui, 2500-4000 nus. Valvae ventrally diminutive with square ft, leg. Champion (19) (BMNH). GUYANA. base and prominent, thickly tapered, caudal Mt. Roraima, leg. H. Whitely (18) (BMNH). extensions comprising about two-fifths over- VENEZUELA. Miranda, D.F., Rio Chacai- all valval length. Aedeagus elongate as typical to, 950 m, February 15, 1936, leg. Lichy (18) ofgenus, shaft bowed, caecum comprising no (AME); Metotan, September 2, 1937 (18) more than one-fifth aedeagal length. Female (AME). Tergal Morphology and Genitalia (fig. 71). Tergite 8 typical ofgenus. Papillae anales less evenly tapered toward terminus, and more declined than Terra andevaga, new species terminally congeners. Genita- Figures 41-44, 71, 74 lia with elongate, thin, cephalic ductal element similar to that of T. calchinia and T. DIAGNOSIS: Distinctive in large size and chilica but much longer, nearly twice length broad wing expanse (FW to 14.0 mm) with of caudal ductal element. Bulb elliptical and broad upper surface black borders and ad- robust, length equaling about one-fourth that jacent forewing brands widely separated by of caudal ductal element. iridescent blue ground color. On under sur- TYPES: Holotype male, allotype female, face, medial yellow patches more pro- Santo Domingo de los Colorados, Pichincha, nounced than in T. tera or T. cana and greatly Ecuador, "submacrothermic rain forest," disjunct. Male genitalia with base of valvae December 24, 1969, leg. H. Descimon, de- very small, square, and terminating with posited AMNH. Paratypes. AMNH: same thickly tapered caudal extensions comprising data as primary types (one male); BMNH: about three-eights ofthe valval length; female Cashaurcu [Casahuaycu?], E. Ecuador, 4000 genitalia with cephalic ductal element greatly ft, leg. Palmer (one female); MNHN: "Equa- elongate, nearly twice length of caudal ele- teur" (16). ment. DISTRIBUTION (fig. 82): Known from up- DESCRIPTION: Male. Upper Surface of land submacrothermic rain forest in Ecua- Wings: ground color dull light iridescent blue dor. with wide, black, subapical and submarginal DIscussIoN: Though poorly represented in forewing borders; pronounced elliptic post- collections, this species has been collected in medial forewing brands bordered distally by series as recently as 1969. In contrast to re- iridescent blue for at least 1.5 mm basad of gionally sympatric N. celona, T. andevaga ap- fuscous wing borders. Hindwing with long pears distinct from T. tera ofthe Colombian tail at vein CuA2 terminus, short tail at vein Cordillera. Along with the structural char- CuAl terminus. Under Surface of Wings: acters, the widely disjunct forewing andro- ground color chalky white, forewing with conial brand and under surface bands appear prominent bipartate yellow-gray postmedial to be stable characters, occurring on an ap- band and discal slash; hindwing with pro- parent dwarf specimen (10.0 mm) as well as nounced yellow to yellow-orange patches on the large specimens characterizing the forming disjunct medial band and prominent types. Johnson (1989e) noted range disjunc- discal slash. Outer margin bordered with pro- tion dividing thecline species of the genus nounced yellow-gray crescents; limbal area Orcya into respective distributions in the Co- with large yellow-orange Thecla spot cen- lombian Cordillera and nuclear Ecuadorian tered with a prominent black spot. Length of highlands. Certain insular species were re- Forewing: 13.0-14.0 mm. Female. Upper stricted to upland cloud forest while wide- Surface ofWings: similar to that ofmales but ranging xerophiles were conspecific through- with fuscous marginal areas much more ex- out the northern Neotropical region. From tensive, blackening all areas distad of fore- the known habitat data, it appears T. andeva- 1991 JOHNSON: HAIRSTREAK BUTTERFLY GENERA 25 62 ..po b

Figs. 62-65. Morphology of Nesiostrymon shoumatoffi and N. celona. N. shoumatoffi: 62. Female. Genitalia (format as in 59). 63. Male. Terminal tergite. a. lateral; b. dorsal; valvae; c. ventral; d. lateral. N. celona: 64. Female (format as in 62); 65. Male (format as in 63). Stipples, areas ofrelatively transparent sclerotization. ga may be restricted to upland Ecuadorian listed above - ECUADOR. Balzapamba, cloud forest while N. celona ranges widely Prov. de Bolivar, 1893-1894, leg. de Mathan throughout the region. (16 [forewing 10.0 mm, construed as a dwarf, MATERLAL EXAMINED: In addition to types see Discussion above] (BMNH)). 26 AMERICAN MUSEUM NOVITATES NO. 3011 66

Figs. 66-68. Morphology of Nesiostrymon milleri and N. australivaga. N. milleri: 66. Male (format as in 63, except e. lateral view, unique vincular spur). N. australivaga: 67. Female (format as in 59). 68. Male (format as in 63). Stipples, areas of relatively transparent sclerotization.

Terra cana (Hayward), new combination Thecla tera [not tera Hewitson, 1863-1878 (1878)]: Figures 45-48, 75, 78 Hayward, 1973: 158 (misidentification ofT. cana females, see Discussion below). Thee/a cana Hayward, 1949: 571, fig. 7. - Coin- stock and Huntington, 1958-1964 [1959]: 174. DiAGNOSIS: Medial elements of hindwing - Hayward, 1973: 155. - Bridges, 1988: 1.68, under surface band generally disjunct (often IL.105. - Johnson and Matusik, 1988: 235. appearing as arc of spots) and paralleled by 1991 JOHNSON: HAIRSTREAK BUTTERFLY GENERA 27 a prominent, orange-bordered discal slash or Locality: Villa Nougues, Tucuman Province, line from cell RS to cell CuAl; forewing up- Argentina. per surface with wide fuscous apices, ob- DISTRIBuTIoN (fig. 82): Known only from scuring identification of sexes by means of mesic woodland localities in Tucuman and the androconial brand (see Discussion be- Jujuy provinces of northwestern Argentina. low). Male genitalia with wide-rimmed, el- DIscussIoN: Hayward (1973) reported both liptical valvae; female with laterally com- Thecla tera and T. cana from Tucuman and pressed ductal elements of about equal size, did not closely associate the species. In his terminating with widely flared lamellae. 1949 description of the latter species he did DESCRIPTION: Male. Upper Surface of not mention T. tera and T. cana as similar. Wings: ground color pale iridescent blue to This seemed odd until Hayward's IML ma- lavender, margins and apices clouded fus- terial was examined: a series identified as cous; forewing with brand generally contig- males and females of T. cana was, in fact, all uous with subapical fuscous; hindwing with males; a series identified as males and females long tail at vein CuA2 terminus, short tail at of "T. tera" was, in fact, all females of T. vein CuAl terminus. Under Surface ofWings: cana. These misidentifications are attribut- ground color chalky white; forewing with able to the wide fuscous forewing apices of faint, discontinuous, yellow postmedial band; T. cana noted in the above Diagnosis. Even hindwing with faint, discontinuous, yellow noting this distinction, I have confused the medial band paralleling a postbasal line from sexes ofthis species without dissection. Hay- cell RS to cell CuA 1. Outer margin lined with ward's small series was mostly in very poor faint yellow-gray crescents; limbal area with condition. An exception is the female illus- faint yellow Thecla spot. Length ofForewing: trated in figures 43-44. Recently, R. Eisele 11.0-12.5 mm. Female. Upper Surface of and B. MacPherson have collected large Wings: similar to that of male but with fus- numbers of T. cana in mesic to hydric forest cous marginal borders wider and forewings areas of Jujuy and Tucuman provinces (see without brands. Under Surface of Wings: locality descriptions 5A in Johnson et al., similar to that of males. Length of Forewing: 1988; 153A, 160B,C in Johnson et al., 1990). 11.0-12.5 mm. Male TergalMorphology and The disparity of sexes in recent samples of Genitalia (fig. 78). Tergite 8, brush organs T. cana parallels our experience with T. his- and genital vinculum, saccus, and aedeagus paniola in the Dominican Republic. As here- typical ofgenus. Valvae ventrally elliptic with tofore noted, the wetter habitat indicated for thick lateral rims tapered to blunt caudal ex- T. cana parallels that reported for T. andev- tensions (length oflatter slightly less than nar- aga in Ecuador but contrasts that generally rowest breadth ofindented valval base). Base attributed to T. tera in Mexico and Antillean of valvae indented laterally to form two, T. hispaniola. shorter basal lobes. Aedeagus bowed, caecum MATERIAL EXAMIND: In addition to types comprising about one-fourth aedeagal length. listed above-ARGENTINA. "Tucuman," Female TergalMorphology and Genitalia (fig. leg. Steinbach (1, 19) (BMNH); "Tucuman," 75). Tergite 8 typical ofgenus. Papillae anales 1100 m, January-February 1905, leg. Stein- with lateral surface heavily sclerotized, form- bach (12) (BMNH); "Tucuman," May 31, ing a raised midline. Genitalia with caudal 1922 (19) (BMNH); Villa Nougues, Tucuman ductal element laterally compressed, lamellae Province, January 21, 1931, leg. K. J. Hay- with widely flared lateral flaps; cephalic duc- ward (16, 32), January 1929 (19) (IML); Que- tal element about same length as caudal el- brada de Lules, Tucuman Province, January ement, but half as wide and not dorsally in- 1929, leg. R. A. (18) (IML); Tucuman Prov- clined. Bulb rather ovate, length equaling only ince, Dept. Tafi Viejo, Las Pitas, 6 km N "El about one-ninth that of caudal ductal ele- Portezuelo," Rte. 340, 1150 m, Rio Raco ment. watershed, W slope Cumbre de Taficillo, Jan- TYPES: Holotype male, labeled "Holotipo, uary 19, 1987, leg. R. Eisele(1 , 12) (AMNH); Thecla cana Hayw., holotipo male, K. J. Hay- Tucuman Province, Dept. Yerba Buena, Yer- ward Det., preparacion genitalia male 4007, ba Buena ("El Paraiso") to Anta Muerta, Villa Nougues, 1 1931" deposited IML. Type Cumbres de Javier, Rte. 338, km posts 15 to 28 AMERICAN MUSEUM NOVITATES NO. 3011

18, 700-900 m, December 31, 1987, leg. R. orange suffusion along the anal lobe. Length Eisele (AMNH) (226, 1Q); Jujuy Province, of Forewing: 11.5-12.5 mm. Female. Upper Cucho, south slopes Cerro Labrada, 1200- Surface ofWings: ground color brown, slight- 1500 m, January 11-16, 1987, leg. B. ly flushed with blue iridescence. Under Sur- MacPherson (36, 2Q). face ofWings: similar to that ofmale but with bands wider and hued red-brown. Length of Forewing: 11.5-12.5 mm. Male Tergal Mor- Red- and Brown-banded Species phology and Genitalia (fig. 79). Tergite 8, Terra calchinia (Hewitson), brush organs and genital vinculum, saccus, new combination and aedeagus typical of genus. Valvae with Figures 49-52, 76, 79 wide distal rims and, depending on degree of slope along cephalic margin, squarish to hex- Thecla calchinia Hewitson, 1868: 21. - Kirby, agonal in shape with gradually tapered, lobate 1871: 393. - Druce, 1907: 600. - Comstock caudal extensions. Aedeagus with shaft quite and Huntington, 1958-1964 [1959]: 172. - straight, caecum comprising about one-fourth Bridges, 1988: 1.66, IL 105.- Johnson and Ma- aedeagal length. Female Tergal Morphology tusik, 1988: 235. and Genitalia (fig. 76). Tergite 8 typical of DIAGNoSIS: Medial under surface bands genus. Papillae anales distinctively lobate, prominently red-orange over white ground with rather square termini differing from the color, a pattern similar only to that of T. steeply pointed structures ofcongeners. Gen- chilica which has prominent yellow-brown italia similar to those of T. chilica and T. forewing band but a dappled hindwing pat- andevaga with relatively elongate cephalic tern and distally angled wings. On T. calchi- ductal element. Of these, however, T. cal- nia the red-orange hindwing band is distinct- chinia is most robust with latter element ex- ly lunular adjacent vein RS and there is ceeding length of caudal element by about brilliant orange along the anal margin be- 1.3. Bulb relatively small, length comprising tween the band and anal lobe. Male genitalia about one-eighth that of caudal ductal ele- have square to hexagonal bases (much like ment. those of T. tera) with short lobate caudal ex- TYPES: Holotype, male, BMNH, labeled tensions; female genitalia exhibit an elongate "Thecla calchinia Hew. Type. Godman-Sal- cephalic ductal element as in T. chilica and vin Coll. 1912023., Ega, U. Amazons, H. W. T. andevaga but, compared to these, T. cal- Bates, B.M. Type. Rh 969." Type Locality: chinia is much more robust (see below). Ega, Upper Amazons, Brazil. DESCRIPTION: Male. Upper Surface of DISTRIBUTION (fig. 82): Known only from Wings: ground color dull iridescent blue on a few localities in the upper Amazon Basin basal two-thirds offorewing and across entire of Amazonas State, Brazil. hindwing except submargins, distal borders DISCUSSION: This species is poorly repre- of both wings fuscous. Forewing with large sented in collections but very distinctive. It elliptical brand slightly darker than fuscous has received little report in the literature and distal ground color. Hindwing with long tail label data suggest all ofthe specimens are old. at vein CuA2 terminus, short tail at vein CuAl Consequently, and because hairstreak iden- terminus. Under Surface of Wings: ground tifications are uncertain at many museums, color bright gray-white; forewing with prom- it is possible that additional specimens are to inent tripartite band, deep red-orange cen- be found in old, unsorted materials. In black trally bordered on both sides by black and and white photos, the female under surface white; submargins with blackish-brown of T. calchinia somewhat resembles that of slashes in each cell; hindwing with tripartite Sipaea hycarra (Hewitson). However, the lat- medial band disjunct, consisting of a large ter species has brilliant blue on the hindwing red-orange costal spot, a jagged continuous upper surface and, on the under surface, band from vein RS to a pronounced "W"- brown bands over tawny ground color, not shaped angle near the anal margin, and en- red bands over gray-white ground color as in larged orbicular spot at the discal cell ter- T. calchinia. minus. Limbal area with brightly contrasted MATERiAL EXAMINED: In addition to type white ground color, bold red Thecla spot, and listed above-BRAZIL. Ega, Upper Ama- 1991 JOHNSON: HAIRSTREAK BUTTERFLY GENERA 29

69 70 71

72 75 74\&IJL Figs. 69-74. Morphology of Terra species. T. hispaniola: 69. Female. Genitalia, lateral (right), ventral caudad of bulb (b) (left, above). Terminal tergite, papillae anales, and microtrichia, lateral (below). 72. Male. Terminal tergite and microtrichia, dorsal (above); valvae, ventral (below). T. tera: 70, 72. (same views as for T. hispaniola). T. andevaga: 71, 74. (same views as for T. hispaniola). b, bulb; stipples, areas of relatively transparent sclerotization. zon, leg. Bates (16, 19) (BMNH); Tefe, Soli- Bridges, 1988: I.78,1I.105. -JohnsonandMa- m6es, November 1919, leg. Parrish (16) tusik, 1988: 235. (CMNH); Amazon, Bates, [ 1 8]35 (19) (HEC); DiAGNosIs: Wing apices distally angled, Sao Paulo de Olivenca (18) (MNHN). forewing brand in males consequently appearing very large. Ground color of forewing Terra chilica (Schaus), new combination (particularly in females) brown-hued, con- Figures 53-56, 77, 80 trasting bright blue hindwing marked distally Thee/a chi/ica Schaus, 1902: 422. - Comstock with vivid white fringe and tails. Under sur- and Huntington, 1958-1964 [1959]: 184. - face with wide yellow-brown medial band on 30 AMERICAN MUSEUM NOVITATES NO. 3011

75 76 (

78 J7Y9) tuu Figs. 75-80. Morphology of Terra species. Same views as figures 69-74. T. cana: 75,78. T. calchinia: 76, 79. T. chilica: 77, 80. Stipples, areas of relatively transparent sclerotization. forewing, hindwing of dappled appearance ductal element elongate (though not extreme from enlarged costal spot and variously dis- as in T. andevaga) with a short, robust caudal junct discal and submarginal lines comple- element. menting the medial band. Male genitalia with DESCRIPTION: Male. Upper Surface of valvae more prominent than in congeners Wings: forewing ground color iridescent gray- (valve length only slightly less than caudo- blue, hued brown, with prominent gray brand ventral diameter of vincular arc) and base closely adjacent fuscous apices and margins; widely elliptical with extremely short caudal hindwing ground color brighter blue with tails extensions; female genitalia with cephalic at vein CuA2 and vein CuAl termini elon- 1991 JOHNSON: HAIRSTREAK BUTTERFLY GENERA 31 gate and, along with adjacent fringe, colored Some wing pattern variation is apparent in vivid white. Under Surface ofWings: ground the samples: in southeastern Brazil, wing un- color chalky white to gray; forewing with light der surfaces are more dappled and the hind- gray submarginal line and prominent yellow- wing band quite lunulate; westward in Parana brown medial band; hindwing with lunular State, Brazil, and in Paraguay, under surface band ofyellow to yellow-brown patches, cos- ground colors are more concolorous and tal element greatly enlarged and discal and bands nearly linear (looking somewhat like submarginal areas more prominently marked those of Nesiostrymon celona). with suffused elements than most congeners; MATERIAL EXAMINED: BRAZIL. Castro, limbal area with yellow-brown Thecla spot. Paranfa, 2900 ft, Jones (26, 22) (BMNH); Length ofForewing: 10.0-12.0 mm. Female. Iguassu, Parana, December 30, 1921 (19), Upper Surface ofWings: ground color offore- December 22, 1921 (16, 12) (BMNH); Pal- wing dull brown; hindwing markedly more mar, Parana, November 2, 1937 (19) (MPM); blue with vivid white fringe. Under Surface Rolandia, Parana, December 1956 (29) of Wings: similar to that of males but some- (MPM); Rolandia, leg. Maller (19) (AMNH); times with broader forewing band. Length of Nova Teutonia, 1944 (1 ) (CMNH); Petro- Forewing: 10.0-12.0 mm. Male Tergal Mor- polis, 1939-1941 (19) (CMNH); Santa Bar- phology and Genitalia (fig. 80a). Tergite 8, bara, October 12, 1926 (19) (MNHN). PAR- brush organs and genital vinculum, saccus, AGUAY. Villarrica, December 1923, F. and aedeagus typical of genus. Valval more Schade (12) (BMNH); Villarrica (18) (MPM), prominent than in congeners, widely ellipti- October 10, 1926 (16, 19) (MPM); Paraguay cal with terminus tapering abruptly to two (46) (MNHN). very diminutive caudal extensions (length comprising about one-sixth of total valvae BIOGEOGRAPHY length). Aedeagus elongate as typical of genus, shaft bowed, caecum comprising one- HISTORICAL REVIEW fourth to one-fifth aedeagal length. Female Clench (1964) erected Nesiostrymon as a Tergal Morphology and Genitalia (fig. 77b). monotypic genus for certain hairstreak but- Tergite 8, in specimens examined, appearing terflies he considered endemic to the Greater more caudoventrally lobate than congeners. Antilles. He divided his type species, N. cel- Papillae anales steeply pointed as typical of ida, into three interisland "subspecies": genus. Genitalia with cephalic element of nominate (Cuba), c. shoumatoffi (Jamaica), ductus thin and elongate, length twice that of and c. aibonito (Puerto Rico). At the time, comparatively robust caudal element. Cau- these butterflies had seldom been collected dal element diameter three times that of ce- and little was known of their habitats or ac- phalic element and ventroterminal opening tual distributions. No specimens had been thin, extending over terminal one-half of el- reported from Hispaniola. Clench speculated ement and surrounded centrally by collarlike that the hairstreak Thecla tera Hewitson, dis- lips. Bulb rather rectangular, overlapping base tributed from central Mexico to northern ofcaudal ductal element and joined to trans- South America, might be a mainland "rela- parent ductal neck by a short stalk. tive" ofNesiostrymon. Earlier, Draudt (1919) TYPE: Holotype male, #5956, National had grouped a number of taxa superficially Museum of Natural History (Smithsonian similar to N. celida in his "Thecla celida Institution). Type Locality: Castro, Parana, Group": T. celida, t. tera, T. celona Hewit- Brazil. son, and T. chilica Schaus. Riley (1975) and DISTRIBUTION (fig. 82): Known from south- Schwartz (1989) reported small numbers of eastern Brazil (Rio de Janeiro and Parana Nesiostrymon from Hispaniola. Johnson and states) west through southern Paraguay. Matusik (1988) differentiated Nesiostrymon DIscuSSION: Like T. calchinia, this species from Terra and described the sole Antillean has received little attention in the literature member of Terra. In their work, Johnson and though every collection examined, with sig- Matusik also made preliminary comments on nificant holdings from southwestern Brazil the restricted local distributions and field and Paraguay, contained a few specimens. ecology of N. celida and T. hispaniola in the 32 AMERICAN MUSEUM NOVITATES NO. 3011

N1%

N..

N" 0C S___N___ N%.I.JEzN

I s' I ..i:! Fig. 81. Inset, map of Greater Antilles indicating geographic distribution of Nesiostrymon; below: map of Neotropical realm between 26°N and 40°S indicating geographic distribution of Nesiostrymon. Taxa Key: O N. celida celida; O c. aibonito; 0 c. baorucoensis; 0 N. shoumatoffi; 0 N. celona; A N. milleri; V N. australivaga.

Antilles. Although relatively few specimens est of the Dominican Republic (Schwartz, ofN. celida had been previously reported from 1989). Miller and Miller (1989) listed a num- the Antilles (Schwartz 1989), Johnson, Ma- ber of Antillean butterflies whose secretive tusik, and members of a 1987 CMNH His- habits and extreme microhabitat restriction paniolan expedition, secured large numbers make them improbable dispersers. As they after locating the microhabitat ofthe species. note, documenting the local ecology of such Contrary to Clench's (1964) view, cladistic species is germane to discussing their cladis- analysis indicates Nesiostrymon is not an en- tic and distributional data. demic Antillean genus. Rather, Nesiostrymon and a sister lineage Terra both occur in the Antilles and on the mainland. T. hispaniola LOCAL ECOLOGY OF was discovered in habitat immediately ad- NESIOSTRYMON AND TERRA jacent to Nesiostrymon and, to date, has not On Hispaniola members of Nesiostrymon been collected outside of the Las Abejas for- and Terra are sympatric in remote upland 1991 JOHNSON: HAIRSTREAK BUTTERFLY GENERA 33 mesic forests (Johnson and Matusik, 1988; Terra hispaniola is still poorly known (nine Schwartz, 1989) of the southern paleoisland males and one female at Las Abejas since (Schwartz, 1989). Nesiostrymon has also been 1985). Mainland species of Terra are also not recorded from four localities on the northern common in collections. Published ecological paleoisland (Schwartz, 1989). In contrast to notes mention Terra tera as "rare" in Ter- a few historical collection records, N. celida minalia L. (Combretaceae)-Dalbergia L. (Fa- can be readily collected once its microhabitat baceae) associations of"Lower Montane Rain is discovered. At "Las Abejas" (Sierra de Forest" (50 to 2500-3000 ft) in the Sierra de Baoruco, Dominican Republic; see Johnson Tuxtla of Yucatan, Mexico, an association and Matusik, 1988) nearly a hundred speci- noted as more xeric than adjacent upland mens have been collected since 1981. The "Montane Forest or Cloud Forest" (2500- samples are mostly from early morning (9- 3000 to 5300 ft) (Ross, 1977). Xeric affinity 1000 hr) in isolated enclaves ofupland mesic of Terra, and mesic affinity of N. celida, is deciduous forest (1100-1200 m) where N. consistent with our observations ofcongeners celida occurs in damp ravine-bottom stands in the Antilles. Las Abejas specimens of T. of Gyrotaenia [probably] myriocarpa (Urti- hispaniola have all been collected in areas of caceae) (i.d. C. Whitefoord, Botany Depart- slightly broken canopy on narrow paths 50 ment, BMNH). The microhabitat also sup- m to 0.5 km upland from the mesic habitats ports other seldom-collected, locally restricted of N. celida. At Las Abejas, T. hispaniola Hispaniolan endemics Greta diaphana (Dru- appears more vagile than N. celida, an indi- ry) (Ithomiidae), Anetia jaegeri Menetries vidual seen in pine forest bordering the Las (Danaidae), Battus zetides Munroe (Papili- Abejas mesic forest at 1600 m altitude in onidae), and recently described Leptotes 1987. As Schwartz (1989: 498-500) noted, idealus Johnson and Matusik (Lycaenidae). T. hispaniola is one ofseveral species ofbut- Local distribution of N. celida is very lim- terflies currently known only from Las Abe- ited. Even in a supportive Gyrotaenia stand, jas. He speculated that if T. hispaniola and the butterflies are not easily observed. If fo- other Las Abejas endemics occur more wide- liage is disturbed, a few individuals may be ly on Hispaniola, they are probably restricted seen flying short distances of5-10 m between to the upland Sierra de Baoruco of the Do- shaded branches but, aside from this activity, minican Republic and perhaps once occurred adults remains mostly sedentary. Midday sun in the adjacent Massif de la Selle of Haiti brings occasional individuals to nectar on before its deforestation. Of five endemic Ly- flowers of adjacent clearings and paths. Be- caenidae described from Las Abejas (John- cause such individuals are more easily ob- son and Matusik, 1988), only Strymon an- served, I suspect most historical Antillean drewi has been collected more widely in the collections (including the small number re- Sierra de Baoruco (Schwartz, 1989). Schwartz ported by Schwartz, 1989) have occurred in attributed this to the latter species being an this manner. Our early experience with the upland pine forest xerophile while mesic de- species was through such occasional sight- ciduous forest biome characterizing the other ings. We never collected N. celida in numbers species is rapidly being destroyed (Schwartz, until we located small stands of Gyrotaenia 1989; Johnson and Matusik, 1988, In press; and sampled them in the early morning be- Johnson, 1989d). fore many other butterflies were flying. Gy- rotaenia is the primary perching substrate for N. celida at Las Abejas but since diverse veg- CLADISTICS AND ANTILLEAN BIOGEOGRAPHY etation is dense amid and around growths of With the advent of modern plate tectonic Gyrotaenia, we have been unable to deter- theory as a reference point, historical litera- mine whether it is the larval foodplant. At ture concerning biogeography of Antillean Las Abejas, N. celida does not venture far butterflies can be generally grouped in four from Gyrotaenia for nectar. The most com- categories: (1) pretectonic dispersalist expla- mon nectar source, Palicourea barbinervia nations (Comstock and Huntington, 1949; (Rubiaceae) (i.d. E. de J. Marcano, MNDR) Michener, 1943; Munroe, 1950; Clench, grows within a few feet of the Gyrotaenia 1964; Miller, 1966, 1968); (2) posttectonic, stands we have sampled. primarily dispersalist, explanations (Scott, 34 AMERICAN MUSEUM NOVITATES NO. 3011

7p- <> . .

H~~ ~ ~ 'Pa.a~ ~~~~~~~~~~~~~~~~~~~~

Fig. 82. Map of Neotropical realm between 26°N and 40°S indicating geographic distribution of Terra. Taxa Key: 0) T. hispaniola; * T. tera; 0 T. andevaga; A T. calchinia; E T. chilica; v T. cana. 1972, 1986; Brown and Heineman, 1972; view and presented a combined vicariance/ Riley, 1975; Brown, 1978); (3) posttectonic dispersal model for the evolution of the An- vicariance-oriented reassessments without tillean butterfly fauna. They present no cla- specific cladistic analyses (Shields and Dvo- distic data in their study but, based on a phy- rak, 1979; Buskirk, 1985; Miller and Miller, logenetic view of contemporary taxonomy 1989); and (4) specific cladistic analyses rel- and distributional data, discuss the plausi- evant to establishing any general vicariant bility of dispersal or vicariance explanations pattern among the butterflies (Ackery and concerning Antillean butterflies and make Vane Wright, 1984; Shuey, 1986; Friedlan- numerous testable predictions. der, 1986 [1987]; Johnson and Matusik, Dispersal has undoubtedly affected the 1988). composition ofthe Antillean butterfly fauna. Since the general acceptance of plate tec- Historically, nearly all authors have charac- tonic theory, concensus has developed among terized the Antillean butterfly fauna as "de- lepidopterists that both dispersal and some pauperate," composed of taxa with generally form of vicariance have contributed to pres- incoherent taxonomic relations to adjacent ent-day Antillean butterfly distributions. mainland groups. Indeed, ecological and dis- Miller and Miller (1989) summarized this tributional traits ofmuch ofthe fauna do not 1991 JOHNSON: HAIRSTREAK BUTTERFLY GENERA 35 contradict an assessment of long-term ran- since the advent of plate tectonic theory- dom invasion coupled with subsequent cli- Antillean endemics with interisland distri- matic, eustatic, and ecological change influ- butions possibly concordant with hypothe- encing differential extinction (Fox, 1963; sized tectonic sequences (Shields and Dvo- Scott, 1972, 1986; Brown and Heineman, rak, 1979; Shuey, 1986; Buskirk, 1985; 1972; Riley, 1975; Brown, 1978; Buskirk, Johnson and Matusik, 1988; Miller and Mil- 1985). Miller and Miller (1989) estimated ler, 1989). Cladistic analyses of these groups roughly 60% of Antillean butterflies fit this are necessary to determine ifa general pattern profile. A smaller percentage of this number of vicariance is apparent among some Antil- (5%, Scott, 1972) are well-documented long- lean butterflies. distance dispersers. Biogeographers generally conclude that most Antillean butterfly taxa, like birds and mammals, are related to Mid- dle American groups with less numerous, dis- VICARIANcE BIOGEOGRAPHY AND cordant elements, showing affinities to North NESIOSTRYMON AND TERRA and South America (Scott, 1972, 1986; Cruz, Vicariance biogeography emphasizes areas 1974; Riley, 1975; Bond, 1978, 1979; Brown, of endemism as potential remnants of for- 1978; Olson, 1978; Buskirk, 1985; Miller and merly continuous ancestral distributions and Miller, 1989). seeks to determine whether there are gener- Nonvagile elements have also always been alized, nonrandom, patterns in the affinities recognized in the fauna and biogeographers ofendemic taxa correlating with information have noted the importance ofthese sedentary concerning geological history. Methods and elements in assessing the origin of the Antil- applications ofvicariance biogeography have lean fauna (Fox, 1963; Brown, 1978; Scott, been articulated in works by Platnick and 1972; Riley, 1975; Buskirk, 1985; Johnson Nelson (1978), Nelson and Platnick (1981), and Matusik, 1988; Miller and Miller, 1989). Wiley (1981), Rosen (1975), Patterson (1981) When the stable-Earth assumption dominat- and Savage (1982). ed biogeography, early Caribbean workers Rosen (1975), using a tectonic scenario pri- explained sedentary Antillean endemics by marily derived from Malfait and Dinkelman land bridges (Fox, 1963) or an "ancient in- (1972), proposed a seminal model for vicar- fusion" of farflung geographic elements by iance in the Caribbean region. Although this dispersal (Brown, 1978). Indeed, some au- view was disputed (Pregill, 1981, and others), thors invoked the small numbers of such subsequent geological literature (Pindell and nonvagile butterflies as a rationale for assum- Dewey, 1982; Burke, 1988; Donnelly, 1988) ing discordant dispersal as the origin ofmost generally supports the plausibility of a late island faunas (Comstock and Huntington, Mesozoic-Tertiary tectonic origin for the 1949; Scott, 1972, 1986; Robbins and Small, Greater Antilles. Discussing the Neotropical 1981). In the Antilles, least vagile butterfly herpetofauna, Savage (1982) summarized vi- groups, like the ithomiids, polyommatines, cariance views of Antillean origins and enu- primitive danaids, and certain endemic sa- merated predictions relevant to determining tyrids emphasized by Miller and Miller (1989) a general pattern. Miller and Miller (1989) show limited occurrences, high endemism, refined these predictions with regard to An- and little taxonomic commonality among the tillean Lepidoptera and, based on a tectonic islands. Included are members of Leptotes scenario slightly modified from Pindell and (Lycaenidae) and Archimestra (Nymphali- Dewey (1982), proposed a vicariance model dae) linked by most authors to ancient Af- for the origin of several groups of Antillean rican antecedents, groups ofspecies (like Ne- butterflies. According to Miller and Miller, siostrymon and Terra ) known from restricted at least 5 clusters of butterfly genera suggest ecological niches, and certain endemic genera a late Mesozoic-Cretaceous vicariance pat- (like Calisto, Satyridae) exhibiting remark- tern including Africa and the New World, 9 able intraisland endemism. Miller and Miller genera fit a late Cretaceous-Eocene vicariance (1989) provided detailed lists ofthese groups. of a "proto-Greater Antilles" from Central Most important to discussion in the pres- and South America, and some 20 intrage- ent study is a group of elements recognized neric groups reflect post-Eocene breakup and 36 AMERICAN MUSEUM NOVITATES NO. 3011 accretion among various tectonic elements of maican L. perkinsae (Kaye) (Johnson and the Caribbean plate. Matusik, in press). A second hairstreak, Elec- Cladistic analysis is basic to testing hy- trostrymon minikyanos Johnson and Matu- potheses of vicariance. Similar to the func- sik, may have a unique relation to E. pan tion of synapomorphies in character clado- (Drury) ofJamaica. However, because ofthe grams, vicariance biogeography construes problem of polyphyly in current usages of congruence of unique areas of occurrence in Electrostrymon, affinities in the genus are un- area cladograms as evidence ofa relationship. certain (Johnson and Matusik, 1988; John- Generalized patterns of vicariance can only son, 1991). be established by studying many groups of The view that T. hispaniola may be a organisms. However, study of particular northern Hispaniolan element stems from its groups provides the elements for these anal- local sympatry with N. celida, a species oc- yses. Congruences among area cladograms, if curring on northern Hispaniola, Cuba, and evidence of a general pattern, could result Puerto Rico. Two factors, however, suggest from either concordant dispersal or vicari- that this sympatry results from dispersion of ance of an ancestral biota. Regarding Antil- N. celida on Hispaniola. Firstly, another up- lean butterfly origins, congruence in area land Sierra de Baoruco endemic (Strymon cladograms is particularly relevant since bio- andrewi Johnson and Matusik) occurs on the geographers have debated whether the Great- north "rim" of the southern paleoisland and er Antilles have long remained in situ or its sister taxon, S. toussainti (Comstock and whether they originated by eastward tectonic Huntington), is pan-Hispaniolan. Secondly, drift of ancient elements from the Central the characters of Las Abejas N. celida bao- American region. Ifthe former, Antillean bi- rucoensis are generally autapomorphic, not ota must reflect over-water dispersal, a ran- primitive. dom phenomenon not open to objective sci- Significantly, area cladograms of Nesios- entific testing. If the latter, however, trymon and Terra expressed as four taxon distributions of Antillean biota may be vi- statements suggest late Cretaceous-Tertiary cariant remnants of such movement and re- vicariance irrespective ofthe above interpre- flect a recognized tectonic pattern. tations concerning T. hispaniola (fig. 83). Us- Two views of a vicariant origin are com- ing either interpretation, two features of Ne- patible with the area cladograms of Nesios- siostrymon / Terra area cladograms are trymon and terra. The relative simplicity of consistent: (1) plesiotypy of the Jamaican/ these views appears compelling compared to South Hispaniolan elements in both genera a dispersal explanation requiring several co- and (2) evidence of three taxon congruence. incidental and concordant over-water dis- Compatible with the Miller and Miller persals. Both views ofvicariant origin require model, if T. hispaniola is considered a north- consideration since it is debatable whether T. ern Hispaniolan element, proto-Antillean el- hispaniola represents a northern Hispaniolan ements [PGA] and Jamaican elements [J] in or southern Hispaniolan distributional ele- Nesiostrymon and Terra vicariated in con- ment. cert. This view requires only that the proto- Historically, T. hispaniola may be a south- Antillean element ofNesiostrymon maintain ern Hispaniolan element. Of five lycaenid a strong ancestral range affinity to South species described by Johnson and Matusik American components [SA] of the genus, an (1988) and known to date only from the Si- observation compatible with Miller and Mil- erra de Baoruco uplands (Schwartz, 1989), T. ler's view that the older the vicariance (e.g., hispaniola and two [perhaps three] other spe- Late Mesozoic-Cretaceous, p. 237-239) the cies have immediate Jamaican affinities. The stronger the affinity ofproto-Greater Antilles hairstreak Heterosmaitia abeja Johnson and to South America. The three-taxon congru- Matusik is the sole known sister ofJamaican ence in this scenario is J/PGA + SA, Central H. bourkei (Kaye). The polyommatine Lep- America [CA]; PGA/SA + CA for Nesiostry- totes idealus Johnson and Matusik shares its mon and PGS/SA + CA for Terra. This sce- salient single hindwing eyespot and various nario is also compatible with Miller and Mil- structural characters only with upland Ja- ler's model when the proto-Greater Antilles 1991 JOHNSON: HAIRSTREAK BUTTERFLY GENERA 37

PALEOCENE EOCENE

OLIGOCENE E. MIOCENE

PGA or PLIOCENE CA,SA PGA J SH CA SA - 14 12

Fig. 83. Paleocene to Pliocene tectonic scenario published in Miller and Miller's (1989) proposal of a vicariance model for Antillean Lepidoptera origins; below, right, area cladogram ofNesiostrymon and Terra species based on figure 2. MX, Mexico; CA, Central America; PGA, proto-Greater Antilles; SA, South America; J, Jamaica; SH, Southern Hispaniola; W&C CU, western and central Cuba; H, His- paniola; ECU & NH, eastern Cuba and northern Hispaniola; PR, Puerto Rico; PR & CH, Puerto Rico and central Hispaniola; LA, Lesser Antilles. element in Nesiostrymon subsequently vicar- Paleocene to Pliocene vicariance model (fig. iates as Cuba/Hispaniola + Puerto Rico. 83) indicates the three-taxon statements of If T. hispaniola is a southern Hispaniolan Nesiostrymon and Terra are congruent in this element [SH], further reference to Miller and case ifthe Jamaica/southern Hispaniola dis- Miller's models is necessary. Their five-step tributions in both genera are viewed as them- 38 AMERICAN MUSEUM NOVITATES NO. 3011

bl ~ ~ ..

f X 1h'

Fig. 84. Morphology of "Thecla" outgroups used in figure 1 cladogram construction (grouped according to Johnson and Matusik, 1988 [subsequent generic assignments by Johnson, 1991, in brackets]). All types BMNH and all views ventral, unless otherwise indicated. "T. uzza complex": a. T. uzza Hewitson [Uzzia], type male (left, male genitalia; right, male valvae, lateral view) showing lateral lobes considered precursor to Nesiostrymon; b. Uzzia cotera Johnson, holotype female, MNHN (genitalia, lateral view) showing condition primitive to female genital "bulb" of Terra. "T. celmus complex": group 1 [Celmia]; c. Thecla celmus Cramer (male left, female right), Presidio, Mexico, AMNH; group 2 [Caerofethra]; d. T. seudiga Hewitson (type male left, type female right, stipples here and in e, g, and h, illustrating membranous character of ductus bursae caudad of corpus bursae); e. T. emandatus Druce (type male left, type female right); f. T. carnica Hewitson, type male; g. T. hesychia Godman and Salvin (type male left, topotypical female right, Rio Sucio, Costa Rica, BMNH); h. T. iambe Godman and Salvin, type female. an, antrum; b, bulb; bl, bilobed configuration; ce, caudal extension. 1991 JOHNSON: HAIRSTREAK BUTTERFLY GENERA 39

*an

Fig. 85. Morphology of"Thecla" outgroups used in figure 1 cladogram (grouped according to Johnson and Matusik, 1988 [subsequent generic assignments by Johnson, 1991, in brackets]). All types BMNH and all views ventral, unless otherwise indicated. "Outgroup complex" [Aubergina]: a. T. alda Hewitson (type male left, type female right, below dashed line, lateral view ofcervix bursae); b. T. paetus Godman and Salvin (type male left, type female right, below dash line, lateral view ofcervix bursae); c. T. hicetas Godman and Salvin, type male. Additional outgroup taxa of present study: group 1 [Sipaea]; d. T. hyccara Hewitson (type male left, female right, "Amazon," BMNH, showing conditions primitive to Nesiostrymon/Terra); group 2 (Dicya); e. T. dicaea Hewitson (type male left, showing condition primitive to NesiostrymonITerra, female right, Rio de Janeiro, Brazil, MPM). an, antrum. 40 AMERICAN MUSEUM NOVITATES NO. 3011 selves representing the old Central American species, a dispersal explanation for their An- component ofthe ancestral distribution. With tillean origin is cumbersome. On the con- this assumption, the area cladograms are trary, the same cladistic and distributional J/PGA + SA,CA; PGA/SA + CA for Ne- data do not contravene a view that Nesios- siostrymon and SH [= cusp of PGA, Paleo- trymon and Terra arose through late Creta- cene-Eocene]/SA + CA for Terra. ceous-Tertiary vicariance of elements on the The ancestral distribution for proto-Ne- Caribbean tectonic plate. siostrymon!Terra has been assumed here as proto-Greater Antilles, South America, and lower Central America, compatible with the ACKNOWLEDGMENTS Paleocene. It is important to point out that I am particularly grateful to J. E. Rawlins both Nesiostrymon and Terra have modern and the O'Neil Fund for Invertebrate Zool- southern Mexican, Central American, and ogy (CMNH) for support of fieldwork and northern South American components (N. published report concerning butterflies in the celona, T. tera) which are apotypic relative Dominican Republic. D. Matusik (FMNH) to Antillean congeners. In accord with the from 1981-1988, Rawlins and R. Davidson interpretation that N. shoumatoffi and T. his- (CMNH) in 1987, and K. Matusik in 1988 paniola represent remnants of the ancestral were field companions at Las Abejas. Sub- Central American distribution of these gen- sequently this work has been aided by a close era, it is inferred that contemporary distri- association with A. Schwartz (Miami Dade butions of N. celona and T. tera reflect dis- County Community College) L. D. and J. Y. persion across the Pliocene-post-Pliocene Miller (AME), and D. Spencer Smith (HEC), reconnection of Central America and South the latter three providing review comments America, as is typical of many other groups on drafts of this manuscript. These, and re- of organisms (Cruz, 1974; Bond, 1978; Ol- view comments at AMNH, were extremely son, 1978; Savage, 1982). If widespread ex- helpful. For material examined I am grateful tinction has not been a historical factor, it to the following institutions, curators, and appears that Terra is relatively younger than collectors: AME, L. D. and J. Y. Miller; Nesiostrymon, since the former lacks a large AMNH, F. H. Rindge; ASC, A. Schwartz; proto-Greater Antilles component and the BMNH, P. Ackery and R. Vane Wright; sole Antillean member is of limited distri- CMNH, J. E. Rawlins; DMC, FMNH, D. bution. Matusik; HEC, D. Spencer Smith; IML, Z. A dispersal model for Nesiostrymon/Terra A. deToledo; MNDR, K. Guerrero R.; is complex. The least complex view requires MNHN, G. Bernardi, J. Pierre; MPM, A. no fewer than two simultaneous and three Young, S. Borkin; REC, R. Eisele, B. sequential coincidental dispersals. Initially, MacPherson; ZMH, H. J. Hannemann and an element ofNesiostrymon (N. shoumatoffi) W. May. must disperse solely to Jamaica while another element (N. celida complex) independently reaches Cuba, Hispaniola, and Puerto Rico in precisely the sequence of the Oligocene- REFERENCES early Miocene Antillean breakup. Simulta- Ackery, P. R., and R. I. Vane Wright neously, an element of the sister genus Terra 1984. Milkweed butterflies, their cladistics and (T. hispaniola) must disperse solely to south- biology. Ithaca: Comstock Publ. Assoc., ern Hispaniola in a fashion allowing it to co- Cornell Univ. Press, 425 pp. exist with N. celida populations in isolated Avinoff, A., and N. Shoumatoff upland mesic forest biomes. Even ifone spec- 1946. A list of butterflies of Jamaica. Ann. Cam. Mus. 30: ulates that were more 263-295. mesic forests formerly Bates, M. cosmopolitan in the Antilles, such dispersals 1935. The Butterflies of Cuba. Bull. Mus. appear exceedingly unlikely. Comp. Zool. 78: 63-258. Considering the high consistency value in Bond, J. cladistic characters ofNesiostrymon and Ter- 1978. Derivations and continental affinities of ra and the geographic occurrence of member Antillean birds. In F. B. Gill (ed.), Zo- 1991 JOHNSON: HAIRSTREAK BUTTERFLY GENERA 41

ogeography in the Caribbean, pp. 119- ers (family Picidae). Caribbean J. Sci. 128. Philadelphia: Acad. Nat. Sci. [Spec. 14: 183-188. Publ. 13]. Donnelly, T. W. 1979. Derivations of Lesser Antillean birds. 1988. Geologic constraints on Caribbean bio- Proc. Acad. Nat. Sci. Philadelphia 131: geography. In J. K. Liebherr (ed.), Zoo- 89-103. geography ofthe Caribbean Insects, pp. Bridges, C. A. 15-37. Ithaca: Cornell Univ. Press. 1988. Catalogue of Lycaenidae & Riodinidae Draudt, M. (Lepidoptera: Rhopalocera). Urbana, 1919. Thecla. In A. Seitz, Macrolepidoptera Ill.: published by the author, 811 pp. of the World, pp. 795-811. Stuttgart: Brown, F. M. Verlag, vol. 5 ("Vol. II"), pp. 593-1139, 1978. The origins ofthe West Indian butterfly vol. 5 Plates ("Vol. II Plates"), pl. 1- fauna. In F. B. Gill (ed.), Zoogeography 194. in the Caribbean, pp. 5-30. Philadel- Druce, H. H. phia: Acad. Nat. Sci. [Spec. Publ. 13]. 1907. On Neotropical Lycaenidae with de- Brown, F. M., and B. Heineman scriptions of new species. Proc. Zool. 1971. A checklist of butterflies of Jamaica. Soc. London 1907(3): 566-632. Mid.-Cont. Lepid. Ser. 2: 1-9. Eliot, J. N. 1972. Jamaica and its butterflies. Hampton, 1973. The higher classification ofthe Lycaeni- Middlesex: E. W. Classey, 478 pp. dae (Lepidoptera): a tentative arrange- Burke, K. ment. Bull. Br. Mus. Nat. Hist. Ento- 1988. Tectonic evolution of the Caribbean. mol. 28: 373-505. Annu. Rev. Earth Planetary Sci. 16: 201- Field, W. D. 230. 1967a. Preliminary revision ofbutterflies ofthe Buskirk, R. E. genus Calycopis Scudder (Lycaenidae: 1985. Zoogeographic patterns and tectonic Theclinae). Proc. U.S. Natl. Mus. history ofJamaica and the northern Ca- 119(3552): 1-48. ribbean. J. Biogeography 12: 445-461. 1967b. Butterflies ofthe new genus Calystryma Clench, H. K. (Lycaeanidae; Strymonini). Proc. U.S. 1964. A synopsis ofthe West Indian Lycaeni- Natl. Mus. 123(3611): 1-31. dae with remarks on their zoogeogra- Fox, R. M. phy. J. Res. Lepid. 2: 247-270. 1963. Affinities and distribution of Antillean 1961. Lycaenidae. In P. R. Ehrlich and A. H. Ithomiidae. J. Res. Lepid. 2: 173-184. Ehrlich (eds.), How to know the butter- Friedlander, T. flies, pp. 176-220. Dubuque, Iowa: Wm. 1986 [1987]. Taxonomy, phylogeny and bio- C. Brown. geography ofAsterocampa Rober 1916 Comstock, W. P., and E. I. Huntington (Lepidoptera, Nymphalidae, Apaturi- 1943. Lycaenidae of the Antilles (Lepidop- nae). J. Res. Lepid. 25: 215-338. tera: Rhopalocera). Ann. New York Godman, F., and 0. Salvin Acad. Sci. 45: 119-130. 1879-1901. Biologia Centrali-Americana. Zo- 1958-1964. An annotated list of the Lycaeni- ology: Insecta, Lepidoptera. 3 vols. dae (Lepidoptera, Rhopalocera) of the London: Taylor & Francis, Vol. 2 (Pier- western hemisphere. J. New York En- idae, Papilionidae, Lycaenidae, Hes- tomol. Soc. 66 [1958]: 103-118; 67 peridae) [1887-1901], 782 pp., vol. 3, [1959]: 59-95, 163-212; 68 [1960]: 49- (pls.), [1879-1886], 1-47[48?]; [1887- 62, 105-122, 1976-186, 232-240; 60 1901], [48?] 49-112. [1961]: 54-58, 105-118, 157-176, 191- Hayward, K. J. 200; 70 [1962]: 39-49, 100-118, 177- 1949. Neuvas especies de Lycaenidae de la Ar- 179; 71 [1963]: 45-57, 115-119, 189- gentina (Lep. Rhop.). Acta Zool. Lil- 197, 262-264; 72 [1964]: 62-64, 120- loana 8: 567-581. 130. 1973. Catal6go de los Ropaloceros Argenti- 1949 (1950). Origins and relationships ofMex- nos. Opera Lilloana 23: 1-318. ican and Antillean Papilionoidea (Lep- Hewitson, W. C. idoptera). Ann. Inst. Biol. Mexico 20: 1863-78 [1869, 1874]. Illustrations of Diurnal 385-391. Lepidoptera, Lycaenidae. London: John Cruz, A. van Voorst, 1: 229 pp.; 2 (pls.): 92 pls. 1974. Distribution, probable evolution and 1868. Descriptions ofsome new species ofLy- fossil record of West Indian woodpeck- caenidae. London: John van Voorst, 36 42 AMERICAN MUSEUM NOVITATES NO. 3011

pp. [paginated as in specimen in Hew- to Rico (Nymphalidae; Charaxinae). itson, on butterflies, 1867-1877 (1972), Caribbean J. Sci. 25: 45-53. London: E. W. Classey, 210 pp.] Johnson, K., R. C. Eisele, and B. MacPherson 1869-77 [1870]. Equatorial Lepidoptera col- 1988. The "Hairstreak Butterflies" (Lycaeni- lected by Mr. Buckley, or Remarks on dae: Theclinae) of northwestern Argen- and descriptions of new species of but- tina. I. Introduction, Calycopis, Calys- terflies collected by Mr. Buckley in Ec- tryma, Tergissima and Femniterga. Bull. uador. London: John van Voorst, 96 pp. Allyn Mus. 123: 1-49. [1870, pt 4, pp. 49-80]. 1990. The "Hairstreak Butterflies" (Lycaeni- Hoffman, C. C. dae: Theclinae) of northwestern Argen- 1940. Catalago sistematico y zoogeographico tina. II. Strymon, sensu stricto. Bull. Al- de los Lepidopteros Mexicanos. Pri. lyn Mus. 130: 1-77. Parte Papilionoidea. Ann. Inst. Biol. Johnson, K., and D. Matusik Mexico 11: 639-739. 1988. Five new species and one new subspe- Johnson, K. cies of butterflies from the Sierra de 1988a. Tergissima and Femniterga, new sister Baoruco of Hispaniola. Ann. Carnegie genera of Calycopis Scudder and Calys- Mus. 57(10): 221-254. tryma Field from the south-central An- In press. Two new butterfly species from His- des. Insecta Mundi 2(1): 28-42. paniola with notes on two other species 1 988b. [Review]. Taxonomy, phylogeny and recently described from the island. Bull. biogeography of Asterocampa Rober Allyn Mus. 1916 (Lepidoptera, Nymphalidae, Apa- Kaye, W. J. turinae) by T. Friedlander. J. New York 1931. Additions and corrections to the au- Entomol. Soc. 96: 482-485. thor's "Butterflies of Jamaica (1926)." 1989a. A revisionary study of the Neotropical Trans. Entomol. Soc. London 79: 531- hairstreak butterfly genus Noreena and 537. its new sister genus Contrafacia. J. New Kirby, W. F. York Entomol. Soc. 97: 11-46. 1871. A synonymic catalogue of the diurnal 1989b. A revision ofthe Neotropical hairstreak Lepidoptera. London: John van Voorst, butterfly genera Tergissima and Fem- 692 pp. niterga (Lepidoptera, Lycaenidae). In- Klots, A. B. secta Mundi 3: 1-24. 1970. Lepidoptera. In S. L. Tuxen (ed.), Tax- 1989c. Revision of Chlorostrymon Clench and onomists' glossary ofgenitalia ofinsects, description oftwo new austral Neotrop- pp. 97-110. Copenhagen: Munksgaard. ical species (Lycaenidae). J. Res. Lepid. Lewis, H. L. 43: 120-146. 1974. Butterflies of the World. Chicago: Fol- 1 989d. Letter to Editor [concerning deforesta- lett, 312 pp. tion in Parque Nacional Sierra de Bao- ruco]. Lepid. News 43(1): 8-9. Lucas, P. H. 1 989e. The new hairstreak butterfly genus Or- 1857. Ordre des Lepidopteres. In G. de la Sa- cya, a revision of the Neotropical gra (ed.), Histoire physique, politique, " Thecla" orcynia assemblage (Lepidop- et naturelle del l'Ile de Cuba, pp. 474- tera: Lycaenidae). J. New York Ento- 750. Paris: Anbimaux articules, Ber- mol. Soc. 98: 50-87. trand, vol. 7. 1991. Neotropical hairstreak butterflies: gen- Malfait, B. T., and M. G. Dinkelman era of the Calycopis/Calystryma grade 1972. Circum-Caribbean tectonic and igneous of Eumaeini (Lepidoptera, Lycaenidae, activity in the evolution of the Carib- Theclinae) and their diagnostics. Rept. bean plate. Geol. Soc. Am. Bull. 83: 251- Mus. Nat. Hist. Univ. Wisconsin 21: 1- 272. 127. Michener, C. D. In press. Types of Neotropical Theclinae (Ly- 1943. A review ofthe genus Calisto. Am. Mus. caenidae) in the Museum National Novitates 1236: 1-6. d'Histoire Naturelle, Paris. J. Lepidop- Miller, L. D. terists' Soc. 1966. A review ofWest Indian "Choranthus." Johnson, K., and H. Descimon J. Res. Lepid. 4: 259-274. 1989. Proper generic and specific status ofAn- 1968. The higher classification, phylogeny and tillean "Prepona" butterflies with de- zoogeography of the Satyridae. Mem. scription ofa new subspecies from Puer- Am. Entomol. Soc. 24: 1-174. 1991 JOHNSON: HAIRSTREAK BUTTERFLY GENERA 43

Miller, L. D., and J. Y. Miller Ross, G. N. 1989. The biogeography of West Indian but- 1964. An annotated list ofbutterflies collected terflies (Lepidoptera: Papilionoidea. in British Honduras in 1961. J. Lepid. Hesperioidea): a vicariance model. In Soc. 18: 11-26. C. A. Woods (ed.), Biogeography of the 1975-77. An ecological study ofthe butterflies West Indies, pp. 229-262. London: Ox- of the Sierra de Tuxtla, in Veracruz, ford Univ. Press. Mexico. J. Res. Lepid. 14 [1975 (1976)]: Munroe, E. G. 103-124,169-188, 233-252; 15 [1976]: 1950. The systematics of Calisto. (Lepidop- 41-60,109-128,185-200,225-240; 16 tera, Satyrinae), with remarks on evo- [1977]: 87-130. lutionary and zoogeographic signifi- Savage, J. M. cance of the genus. J. New York 1982. The enigma of the Central American Entomol. Soc. 58: 211-240. herpetofauna: disperals or vicariance. Nelson, G., and N. I. Platnick Ann. Missouri Bot. Gard. 69: 464-547. 1981. Systematics and biogeography. New Schaus, W. York: Columbia Univ. Press, 567 pp. 1902. Descriptions of new American butter- Olson, S. L. flies. Proc. U.S. Natl. Mus. 24(1262): 1978. A paleontological perspective of West 383-460. Indian birds and mammals. In F. B. Gill Schwartz, A. (ed.), Zoogeography of the Caribbean, 1989. The Butterflies of Hispaniola. Gaines- pp. 99-117. Philadelphia: Acad. Nat. ville: Univ. Florida Press, 590 pp. Sci. [Spec. Publ. 13]. Scott, J. A. Patterson, C. 1972. Biogeography of Antillean butterflies. 1981. Methods of Paleobiogeography. In G. Biotropica 4: 32-45. Nelson and D. E. Rosen (eds.), Vicari- 1986. Distribution of Caribbean butterflies. ance biogeography, a critique, pp. 446- Papilio (n. ser.) 3: 1-26. 489. New York: Columbia Univ. Press. Shields, O., and S. K. Dvorak Pindell, J., and J. F. Dewey 1979. Butterfly distribution and continental 1982. Permo-triassic reconstruction of west- drift between the Americas, the Carib- ern pangea and the evolution ofthe Gulf bean and Africa. J. Nat. Hist. 13: 221- of Mexico/Caribbean region. Tectonics 250. 1: 179-211. Shuey, J. A. Platnick, N. I., and Nelson, G. 1986. The ecology and evolution of wetland 1978. A method ofanalysis for historical bio- butterflies with emphasis on the genus geography. Syst. Zool. 27: 1-16. Euphyes (Lepidoptera, Hespieriidae). Pregill, G. K. Ph.D. dissertation, Ohio State Univer- 1981. An appraisal of the vicariance hypoth- sity (Columbus), 145 pp. esis of Caribbean biogeography and its Swofford, D. application to West Indian terrestrial 1985. PAUP (Phylogenetic Analysis Using vertebrates. Syst. Zool. 30: 147-155. Parsimony) and Users Manual [a com- Riley, N. D. puter software package made available 1975. Field guide to the butterflies ofthe West by the Illinois Natural History Survey Indies. London: Collins, 244 pp. and the author]. Robbins, R. K., and G. B. Small Wiley, E. 0. 1981. Wind dispersal of Panamanian hair- 1981. Phylogenetics, the theory and practice streak butterflies (Lepidoptera: Lycae- ofphylogenetic systematics. New York: nidae) and its evolutionary significance. Wiley, 439 pp. Biotropica 13(4): 308-315. Wolcott, G. N. Rosen, D. E. 1936. "Insectae borinquenses," a revised 1975. A vicariance model of Caribbean bio- checklist of the insects of Puerto Rico. geography. Syst. Zool. 24: 431-464. J. Dep. Agric. Puerto Rico 20: 1-600. Recent issues of the Novitates may be purchased from the Museum. Lists of back issues of the Novitates, Bulletin, and Anthropological Papers published during the last five years are available free of charge. Address orders to: American Museum of Natural History Library, Department D, Central Park West at 79th St., New York, N.Y. 10024. i

THIS PUBLICATION IS PRINTED ON ACID-FREE PAPER.