Two common and problematic leucochrysine species – Leucochrysa (Leucochrysa) varia (Schneider) and L. (L.) pretiosa (Banks) (Neuroptera, Chrysopidae): redescriptions and synonymies

Abstract We dedicate this article to the memory of Sergio de Freitas, FCAV-UNESP, Jaboticabal, São Paulo, Brazil (deceased, 2012). He was an active and enthusiastic Neuropterist and the cherished mentor and friend of Francisco Sosa. Leucochrysa McLachlan is the largest genus in the Chrysopidae, yet it has received relatively little taxonomic attention. We treat two problematic and common Leucochrysa species – Leucochrysa (Leucochrysa) varia (Schneider, 1851) and Leucochrysa (Leucochrysa) pretiosa (Banks, 1910). Both are highly variable in coloration and were described before the systematic importance of chrysopid genitalia was recognized. Recent studies show that these species occur within a large complex of cryptic species and that they have accumulated a number of taxonomic problems. We identify new synonymies for each of the species–for Leucochrysa (Leucochrysa) varia: Leucochrysa (Leucochrysa) ampla (Walker, 1853), Leucochrysa internata (Walker, 1853), and Leucochrysa (Leucochrysa) walkerina Navás, 1913; for Leucochrysa (Leucochrysa) pretiosa: Leucochrysa (Leucochrysa) erminea Banks, 1946. The synonymy of Leucochrysa delicata Navás, 1925 with Leucochrysa (Leucochrysa) pretiosa is stabilized by the designation of a neotype. The following species, which were previously synonymized with Leucochrysa (Leucochrysa) varia or Leucochrysa (Leucochrysa) pretiosa, are reinstated as valid: Leucochrysa (Leucochrysa) phaeocephala Navás, 1929, Leucochrysa (Leucochrysa) angrandi (Navás, 1911), and Leucochrysa (Leucochrysa) variata (Navás, 1913). To help stabilize Leucochrysa taxonomy, lectotypes are designated for Allochrysa pretiosa and Allochrysa variata. Finally, Leucochrysa vegana Navás, 1917 is considered a nomen dubium.

cochrysa), with ca 46 species, and Leucochrysa (Nodita), with ca 150 species (Brooks and Barnard 1990, Penny 1998, 2001, Freitas and Penny 2001, Freitas 2005, 2011a, 2011b. The genus has numerous taxonomic complications, many stemming from the poor status of its descriptive systematics. Most Leucochrysa species were described without reference to their internal anatomy (e.g., genitalia), and in many cases, these structures continue to remain unknown. As a result, identifications are difficult and a large number of cryptic species have gone unrecognized. In addition, polymorphisms and other forms of intraspecific variation are often interpreted as species differences; in some cases, males and females of the same species are described under different names. Moreover, systematic study of the group has been slow because types are sometimes difficult to locate or access; many are in poor condition. As a result, synonymies are numerous; species are difficult to identify with accuracy; a significant number of species remain undescribed; and the genus is a taxonomic enigma for systematists and ecologists who are interested in chrysopids.
Efforts to improve the descriptive systematics of Leucochrysa began in the late 1970s with work by Adams (1977Adams ( , 1979; his studies were the first to include the genitalia (male and female), and his findings were strongly rooted in careful examination of type specimens. Subsequently, in the last ca 10 years, interest in the descriptive systematics of Leucochrysa has increased. Regional treatments of Leucochrysa have been published; new species have been described; polymorphisms have been elucidated; and, larval morphology has been explored (Freitas and Penny 2001, Penny 2002, Tauber 2004, Freitas 2005, Mantoanelli et al. 2006, 2011, 2011a, 2011b. Unfortunately, in some of these studies, type specimens do not appear to have received appropriate attention. As a result, errors continue to creep into the literature on the taxonomy of the group. Recently we examined specimens of Leucochrysa (Leucochrysa) that resemble the relatively common species, Leucochrysa (L.) varia (Schneider). We were surprised to find that our specimens included numerous cryptic species. Among them was Leucochrysa (L.) pretiosa (Banks); most, if not all, of the other species appear to be undescribed. To help provide a sound morphological and systematic basis for future comparative study and description of the newly discovered species, we examine the taxonomy and morphology of the two previously named species. In doing so, we (a) redescribe and provide images of the two species, including the genitalia of both males and females, (b) identify new synonymies for each of the species, (c) help stabilize the nomenclature of the group by designating a neotype and two lectotypes, (d) reinstate to valid status three species that had previously been synonymized with one or the other of the species, and (e) designate one species as a nomen dubium. All of these taxonomic actions are verified with images of the appropriate types.

Materials and methods
We searched for specimens in a large number of entomological collections. Below are the institutions (with acronyms) where we found and used specimens. red-wine coloration throughout; (6) pedicel generally cream-colored, with inner margin shaded brown; (7) flagellum cream-colored to yellow, covered with pale bristles, basal three to five flagellomeres generally with inner margin streaked with black; (8) frons and clypeus white to cream-colored or variably red suffused; (9) gena tinged with pink, red or reddish brown; (10) labrum yellowish, generally not marked; (11) maxillary and labial palpi yellowish; (12) cervix with small, red, lateral mark; (13) pronotum greenish to yellowish, unmarked; (14) mesonotum, metanotum variably pigmented with red, reddish brown or black (pigmentation with wide range of intraspecific and interspecific variation); (15) abdominal tergites 5-6 with dark brown oval spot, bordered with red and yellow.
Leucochrysa (L.) varia was reported from British Guiana, Suriname, Central America, and Mexico; however we have not confirmed any of these records. Banks (1945: 168) questioned the Navás (1928a: 131) record for Central America (Guatemala). Based on our study here, we also question this record. We confirmed that the Banks (1944: 31) record for British Guiana was in error; a specimen in the AMNH with the data he reported is probably an undescribed species; its abdomen is missing. The specimen(s) that he identified as L. (L.) varia from Suriname (Banks 1944: 31) were not found in the MCZ (P. Perkins, personal communication), nor in the AMNH. This record probably applies to L. (L.) pretiosa or another, undescribed species. In summary, we have not seen specimens of L. (L.) varia from northern South America, the Caribbean region, Central America, or Mexico.
The type is in fairly good condition. Its primary label is hand-written, apparently in Schneider's hand; it reads: "varia nov. sp. / Schneider / Brasilia". The exact collection site is unknown. In addition to Fig. 1 here, images of the external and male features are in the MCZ Type Database (website: http://insects.oeb.harvard.edu/MCZ/ index.htm).
Walker originally referenced two specimens under the name Chrysopa internata ("a" from Brazil and "b" without locality data); he also stated that C. internata had two varieties. Because one of the specimens must represent Walker's lettered variety ("var. β"), that specimen must be excluded from the type series of C. internata under Art. 72.4.1 (see Oswald 2007). Kimmins (1940: 444) identified the specimen without locality data (specimen "b") as the primary C. internata type (a lectotype). Because the actual type series of C. internata consists of only a single specimen, his lectotype designation was unnecessary; we recognize Kimmins' action as identification of the holotype.
The excluded "var. β" specimen (specimen "a" from Brazil), was subsequently designated the type of Leucochrysa walkerina Navás (see below).  The C. internata holotype is in fairly good condition; the head is separated from the body and the dissected abdomen is in glycerin, in a vial attached to the pinned specimen. The labels are hand-written and printed (Fig. 2B). The specimen carries no locality data; its collection site is unknown. All of its features, including the male terminalia, correspond to those of the L. (L.) varia type (see Figs 2-3).   Walker originally referenced two specimens ("a" and "b") under the name Chrysopa ampla; he also stated that C. ampla had two varieties. Because one of the specimens must represent Walker's lettered variety ("var. β"), that specimen must be excluded from the type series of C. ampla under Art. 72.4.1 (see Oswald 2007). Thus, the actual type series of C. ampla consists of only a single specimen, the one that was listed by Walker as "var. α" and that carries no locality data. Therefore, this specimen constitutes the holotype by monotypy. This designation is consistent, nomenclaturally, with the action by Kimmins (1940: 444), who considered the specimen as the name-bearing type when he designated it as the "lectotype".
The excluded "var. β" specimen was reported from "Georgia". Tauber (2004Tauber ( : 1132 identified it as Leucochrysa (L.) insularis Walker, and it bears her label with that name.
The C. ampla holotype is in fairly good condition; the dissected abdomen is in glycerin, in a vial attached to the pinned specimen. The labels are hand-written (not by Navás) and printed (Fig. 4A). The specimen carries no locality data; its collection site is unknown. All of the features (external and genitalic) of this type correspond to those of female L. (L.) varia (see Figs 4-5).
Leucochrysa vegana. Type(s) by original designation, probably missing. Navás (1917: 278), in his original description, did not indicate the depository or sex of the type; he reported that it was from the "Coll. Br.
[Brother] Apolinar Maria [María]" (not examined). It was not found in the Navás collection (Monserrat 1985: 240), the BMNH, the MNHN, or the MCZ. And despite searches by colleagues (see acknowledgements) in Colombia where Br. Apolinar María lived, it was not found. Apparently, the Br. Apolinar María collection was housed in the Museo de la Universidad de la Salle and was destroyed during a political upheaval in 1948.
Navás mentioned the similarity between L. vegana and L. varia ("variae Schn."), but he did not point out why he considered that L. vegana was different. Most of the features that he described for L. vegana are found on L. (L.) varia specimens, including the markings at the base of the L. vegana forewing that he illustrated. Thus, although it is unlikely that Banks actually saw the L. vegana type, his synonymy (based on Navás' description) made sense at the time.
The type locality of L. vegana (La Vega, which is in the Cordillera Oriental of Colombia), is considerably north of the currently known northern limit of L. (L.) varia [the western regions of the Amazon drainage basin of Ecuador and Peru (e.g., the Yasuní Reserve in the Province of Napo, Ecuador and the Tambopata district of Peru)], but well within the ranges of other, undescribed Leucochrysa (L.) varia-like species. At this time, we suspect that L. vegana is not synonymous with L. (L.) varia. However, the region remains very poorly collected, so it is possible that our suspicion is in error. Thus, while we await the collection of L. (L.) varia from La Vega or nearby, to confirm Banks' synonymy, we consider the species name to be a nomen dubium. Leucochrysa phaeocephala. Type(s), by original designation, Hamburg, probably destroyed during W.W. II, sex unknown (not examined).
Banks ' (1944: 30) synonymy was made on the basis of the description; it does not appear that he saw the type. Navás (1929b: 21) reported that the specimen was collected in Dutch Guiana in 1908 (now, Suriname). Because we have found no records of L. (L.) varia from the northern regions of South America, we reverse Bank's synonymy; we will deal with the species in a later publication.
The walkerina type is slightly teneral, but in good condition; the cleared abdomen is in glycerine, in a vial attached to the specimen. The pin carries four labels below the specimen (including a locality label, "Brazil") ( Fig. 6A); it also has a "Paralectotype" label above the specimen (Fig. 6B). The external characteristics (Fig. 6) are consistent with those of L. (L.) varia, and the elongate, coiled, spermathecal duct is that of L. (L.) varia (Fig. 7).

Diagnosis.
As the name varia implies, adults of L. (L.) varia exhibit a wide range of color variation; there are black and red morphs, with and without coloration on the mesoscutellum (See Fig. 8 here, Fig. 7 in Mantoanelli et al. 2006). In addition, preserved specimens of varia-like species tend to loose their natural coloration very quickly; old specimens are especially discolored. Thus, it is important to emphasize that accurate identification of the varia-like species can only be made by careful examination of the male or female genital characters.  darkening and suffusion around these veins is generally uniform, whereas in L. (L.) pretiosa, the distal Psm-Psc crossvein is usually darker and more prominent than the other Psm-Psc crossveins or the outer gradates.
In L. (L.) varia males, the sclerotized mediuncal plate is elongate; its rods are narrow and parallel; its membranous connection to the gonarcal bridge does not extend laterally beyond the gonocornua and is soft [not broad, leathery and stiff, as in L. (L.) pretiosa]. The L. (L.) varia female is distinguished by a very long, strongly coiled spermathecal duct and a spermatheca that is scoop-shaped distally and has a convoluted, tubular, basal section leading to a broad, fluted bursal duct.
Female. Height of S6 ca 0.75 times length, S7 height ca 0.60 times length. Callus cerci round, diameter 0.17-18 mm, with 30-35 trichobothria. T8 roughly quadrate (lateral view) with rounded corners, similar in depth to T6. T9+ectoproct elongate, slanting anteriorly; ventral margin slightly convex, extending slightly below level of gonapophyses laterales. Dorsal margin of S7 with slight taper basally, becoming more pronounced distally; terminus unmodified, with terminal (posteroventral) setae slightly more dense than in other areas. Gonapophysis lateralis rounded to slightly acute dorsally, rounded distally, ventrally, ca 0.53-0.60 height of T9+ectoproct; inner membranous surface not expandable, with ca two vertical rows of short setae. Colleterial complex consisting of membranous gland connected to colleterial reservoir via broad duct, and elongate ribbon-like accessory gland; colleterial gland elongate, delicate, transparent; colleterial reservoir smaller, delicate, transparent, tapering to narrow, granulose, spiny duct; accessory gland narrow, elongate, forked distally, with spiny surface; accessory gland and colleterial duct connected to lightly sclerotized, widened, flattened platform extending from below transverse sclerite; transverse sclerite curved, lightly sclerotized, slender throughout, with long teeth (setae?). Spermatheca with initial (posterior) section scoop-shaped, broad, thick, tapering slightly at base (ca 0.25 mm wide along distal margin x ca 0.20 mm height from tip of distal margin to base of scoop), with elongate, broad, smooth, convoluted tube extending down one side, looping in U-shaped turn, then twisting through several loops before joining bursal duct [tube ca 1.1 mm in total length, ca 0.05 mm in width throughout]; spermathecal invagination not specifically identified. Spermathecal duct extremely long, well sclerotized throughout, densely, tightly coiled, arising from side of scoop-shaped section of spermatheca; coiled length ca 3-5 mm, including membranous, brushy, less coiled distal section, uncoiled length much greater. Bursal duct extending from tip of tubular spermathecal velum, basal section membranous, broad, curved, fluted; surface with striated folds, lateral margins of major folds heavily granulose. Bursa copulatrix small, saccular, extended over spermatheca, slightly into section of S7; ventral surface with small striated folds; dorsal surface smooth; pair of clear, elongate, tubular bursal glands attached dorsally to base of bursa via clear, pipe-fitting-like bases; bursal glands very long, unbranched; surface lightly granulose. Subgenitale with smooth (unfolded), rounded surface, broad, rectangular, robust, bilobed projection extending distally at ca 90° angle to subgenitale surface, lobes large, with minute setae on surface, region between lobes extending distally as smaller, acute lobe; base of bilobed projection with dense transverse folding, with sclerotized, knob-like mesal lobe projecting from pair of scalloped, sclerotized arms. Larva. All instars described (Mantoanelli et al. 2006). Biology. Developmental and survival rates of immature stages under five constant temperatures and larval trash-carrying behavior studied by Mantoanelli et al. (2006) and Mantoanelli and Albuquerque (2007).
Variation. The coloration of the head, mesonotum and metanotum is the most obvious variation expressed by L. (L.) varia [see above]. However, color is not the only feature that varies among L. (L.) varia specimens. For example, the expansion of abdominal segments 4-9 and the degree of sclerotization of the integument varies greatly among both male and female specimens (see Figs 10, 12). Differences in sclerotization are particularly noticeable in the ventral apodemes of the male T9+ectoproct (Fig. 10).
We suspect that these features (abdominal expansion and integumental sclerotization) are at least partially a function of age and developmental or reproductive maturation.
Leucochrysa (Leucochrysa) variata (Navás). Valid status reinstated. See below. Leucochrysa delicata Navás [1925Navás [ ] (1925 fig. 18) original description: "Costa Rica: Reventazón, 15 de Marzo de 1923. Janson et Sons. Col. m.". Banks (1945 synonymy with L. pretiosa; Adams (1979: 97) species list, note regarding absence of type; Brooks and Barnard (1990;276) listing as a synonym of L. (L.) pretiosa; Oswald (2007) catalog listing of L. delicata and L. (L.) delicata as synonyms of L. (L.) pretiosa. Neotype designation below. Leucochrysa erminea Banks [1946Banks [ ] (1945 (Banks 1944: 31) are all well within the confirmed range of L. (L.) pretiosa. We have seen specimens from most of the areas he reported; however, we have not examined his specimens. The northern-most specimen of L. (L.) pretiosa that we have seen is from Chiapas, Mexico. Valencia Luna et al. (2006: 48) also reported the species from Morelos, Mexico; however, we have not seen specimens to confirm the report. The southern-most record is based on a single specimen, with somewhat obscure data that we interpreted as referring to the Mennonite Colony of Sommerfeld, ca 210 km east of Asunción, Paraguay. This locality is far south of the second southern-most record for the species (north-central Ecuador).
Here, to stabilize the nomenclature of this taxonomically difficult group of lacewings, we recognize the specimen with an abdomen (female) as the Lectotype [present designation].
"Cauca" probably refers to the Valle del Cauca, which is in southwestern Colombia. We could not locate an "Inmba" (spelling in original description), "Inmbo" (spelling on lectotype label and by Banks 1944: 31), or "Jumba" (spelling by Navás 1913: 158); we suspect that these names are all misspellings of Yumbo, a town near Cali, Valle del Cauca Department.
From his discussion, it appears that Banks (1945: 167) did not see the type of A. angrandi; even so, he suggested that the species was synonymous with L. pretiosa and, with one exception (Adams 1979: 97), it has been treated as synonymous. Our comparison of the lectotypes of the two species (both females) indicates that they are distinct . Indeed, the A. angrandi lectotype lacks some characteristics of the varia-like species that we have examined. For example, it appears to have red lateral stripes on the prothorax (Figs 16C-E), no suffusion on the forewing crossveins, and the base of the forewing lacks dark markings (Fig. 16I). In addition, the fifth and sixth tergites of the cleared abdomen (Fig. 17) lack the large spots that are typical of varia-like species. This characteristic sometimes is not visible in cleared varia-like specimens, especially teneral ones, and thus its absence from the A. angrandi type is not definitive for excluding the species from the group. This species will be re-described elsewhere.
In the original description, Navás stated that he retained the type in his personal collection; it is not there now (Monserrat 1985: 240), nor was it found at the MNHN or the BMNH. Presumably, it is lost (see Adams 1979: 97). Given the collection site of the type (Costa Rica), it is possible that it was either L. (L.) angrandi or L. (L.) pretiosa. Because L. (L.) pretiosa is the more common and therefore most likely of the two, we have no basis for altering Banks' synonymy; thus it pertains. To help stabilize the taxonomy of this taxonomically difficult group of Leucochrysa, we designate a female specimen in the William F. Barr Entomological Collection (UID), University of Idaho, as the Neotype. This specimen is from Turrialba, a locality in the Reventazón River basin of Costa Rica; its labels read: (1) "COSTA RICA. Cart. / Turrialba, CATIE / 26-29 Jun 1986 / Nadeer Youssef", (2) "NEOTYPE" / Leucochrysa delicata / Navás, 1925   Leucochrysa erminea Banks. Holotype by original designation, MCZ (sex unknown, abdomen missing, examined). No other type material was found in the MCZ (P. D. Perkins, personal communication). Banks (1945: 169) did not compare his type of L. erminea with the L. pretiosa specimens that he had described many years earlier. He was probably unaware of the variation in body color expressed by this species. We make the synonymy on the basis of external features, notably the wings, which are almost identical on the A. pretiosa Redescription. Head (Fig. 22): 1.7-1.9 mm wide (including eyes). Frons, clypeus white, red suffused, or entirely red; gena pink to dark red; maxillary and labial palpi yellowish. Vertex with central area raised, yellowish to green, with prominent, dark red to dark brown, V-shaped mark along anterior margin; lateral margins with narrow to wide stripes; post-ocular area with red spot. Antenna: scape tinged with red-wine coloration, especially mesally; pedicel yellowish green, inner margin with dark mark; flagellum cream-colored, with amber bristles; inner margin of basal ca three flagellomeres tinged with black; dorsal antennal fossae marked with red laterally. Thorax (Fig. 22): Cervix with red mark laterally. Pronotum 1.1-1.5 mm long, 1.2-1.4 mm wide, yellowish green, unmarked. Mesonotum, metanotum variable, entirely dark brown to yellow with brown marks of various sizes. Posterior margin of mesoscutum with prominent dark brown, transverse stripe; metascutum with small or large brown marks; mesoscutellum, metascutellum yellow to light green, without marks.
Larvae. Unknown. Biology. Unknown. Variation. Leucochrysa (L.) pretiosa expresses considerable variation in head, mesothoracic and metathoracic coloration, as well as wing size, shape and degree of suffusion on various veins. The range of variation in several of these traits is shown on Figs 9 and 22. As in L. (L.) varia, the expansion of abdominal segments 4-9 and the degree of sclerotization of the apodemes and sclerites varies considerably among specimens (see Figs 23,25). This variation may be related in part to age and/or maturation; however, both male and female specimens with large, expanded (apparently mature) abdomens often had very soft and delicate integuments that tore easily during dissection. Thus, a delicate integument may not necessarily be associated with a teneral status.