Review of the continental Oriental species of Lilioceris Reitter (Coleoptera, Chrysomelidae, Criocerinae) closely related to Lilioceris impressa (F.)

Abstract Criocerine leaf beetles found in Nepal feeding on Dioscorea bulbifera (L.), an invasive weed of Asian origin, were identified as Lilioceris cheni Gressitt and Kimoto based on a synopsis of the Oriental Lilioceris species and review of the Lilioceris impressa species group. All the continental, Oriental species included in the group are diagnosed and illustrated, and a key for their identification is provided. Species status of Lilioceris thibetana Pic, 1916 is resurrected. The following new synonyms are proposed: Lilioceris coomani (Pic, 1928) = Lilioceris egena (Weise, 1922), and Lilioceris subcostata (Pic, 1921a), Lilioceris laticornis (Gressit, 1942), Lilioceris inflaticornis Gressit & Kimoto, 1961, and Lilioceris maai Gressit & Kimoto, 1961 = Lilioceris impressa (Fabricius, 1787). Lectotypes of the following species are designated: Lilioceris coomani Pic, 1928; Lilioceris impressa (Fabricius, 1787); Lilioceris laosensis (Pic, 1916); Lilioceris malabarica (Jacoby, 1904); Lilioceris ruficornis (Pic, 1921b); Lilioceris subcostata (Pic, 1921a); Lilioceris thibetana (Pic, 1916); and Lilioceris unicolor (Hope, 1831).


introduction
This study was initiated by the need to identify leaf beetles that were feeding on Dioscorea bulbifera L. (Dioscoreaceae) in Nepal. Dioscorea bulbifera (air potato) is a herbaceous, perennial twining vine that attains lengths of 20 m or more, rendering it capable of climbing over and smothering native vegetation (e.g., Schmitz et al. 1997;Langeland and Craddock Burks 1998;Gordon et al. 1999). The species was introduced to Florida from tropical Asia or Africa in 1905 (Morton 1976). By the 1980s, air potato vines were growing in thickets, waste areas, and hedges or fencerows in many parts of south and central Florida (Bell and Taylor 1982). By 1999, D. bulbifera was recognized as an invasive exotic capable of altering plant communities by displacing native species and altering community structure and ecological functions (FLEPPC 2003). Vegetative propagation occurs primarily through aerial bulbils (hence the name "air potato") that form in leaf axils during late summer, and may weigh up to 1 kg. These bulbils drop to the ground during the cooler months when the vines die back. Vines resprout during spring from subterranean tubers or from bulbils. Seed production is unknown in Florida and spread occurs mainly through anthropogenic dispersal of the bulbils (Schultz 1993). In order to find natural enemies of air potato, field explorations were conducted in Nepal and other Asian countries (Wheeler et al. 2007).
Leaf beetles that were found feeding on air potato in Nepal were not identifiable to species using existing keys (Gressit andKimoto 1961, Kimoto andGressit 1979), which are generally incomplete and often use unreliable characters. There are about 110 species of Lilioceris in the Oriental Region, about 80 of which occur on the continent. Synoptic study involving examination of the original descriptions and major leaf beetle collections (see list below) allowed us to identify a set of features that separated a relatively small group of species, including a species in question that is a potential air potato biocontrol agent, from the rest of the continental, Oriental Lilioceris. We refer to this group as the Lilioceris impressa species group and treat it in this paper, providing distinguishing characters of the group, a key to its species, and illustrations of characters, including the internal sac of the aedeagus.

Material and methods
Specimen observation and preparation follow the methods of Konstantinov (1998 The most reliable characters for discrimination of Lilioceris species are those of the internal sac of the aedeagus. In Criocerinae, these characters were used previously for separation of Lilioceris species from Iran (Berti and Rapilly 1976) and Oulema species related to O. melanopus (L.) (Berti 1989). Terminology for the sclerites of the internal sac is not overly complicated. Berti and Rapilly (1976) called the longer, dorsally situated structure in the everted internal sac, the "pièce terminale". We recognize three major sclerites in the internal sac of the aedeagus: dorsal [="pièce terminale" of Berti and Rapilly (1976)], median, and ventral (Fig. 23). All these three parts differ from species to species within the L. impressa group, but the most reliable and easily observed characters are those of the dorsal sclerite. Dissection and preparation of the sclerites of the internal sac of the aedeagus are relatively simple. We used slightly bent #1 entomological pins to extract the internal sac from the dorsal opening of the aedeagus that was soaked in hot 10% KOH solution for 15-20 minutes and washed in excess of water. Figures were generated using a Camera Lucida attached to a Zeiss Stemi SV 11 dissecting microscope. They were scanned and edited with Adobe Photoshop.

systematics of L. impressa species group
Species of the Lilioceris impressa group share the following characters: 1) glabrous scutellum almost completely lacking setae, with only a few setae occasionally present near the base (Fig. 7); 2) antennomeres 5-10 distinctly flattened, quadrate or even transverse (antennomere 5 often slightly elongate) and covered with dense, short, appressed setae (in clear contrast with more basal antennomeres bearing longer, more erect and much sparser setae, not obscuring the view of glabrous cuticle) (Fig. 6); and 3) structure of the sclerites of the internal sac of the aedeagus. The dorsal and ventral sclerites are roughly plank-like in shape, the dorsal one being longer and more complex shaped, especially anteriorly where it bears two relatively long lateral processes that vary in shape.
To a lesser extent color (traditionally often used in distinguishing and keying out Lilioceris species) is helpful to recognize representatives of the species group. Typically, they have a black head, thorax, abdomen and legs, and brownish-yellow to reddish elytra. Black legs and pale unspotted elytra are consistent throughout all species group members, while the head and thorax may be partially or entirely brown, reddish-brown or reddish; these paler colors are variably present around the abdominal apex.
The three characters mentioned above, in combination, define the group. As to the similar species, there are, on one hand, several similar looking black and red (yel-low) species with a glabrous scutellum and similar structure of male genitalia, which have distinctly, often substantially elongate (never close to quadrate or transverse) antennomeres 5-10. On the other hand, a group of mostly entirely reddish species with the antennal characters identical to those of the impressa-group possess consistently a completely setose scutellum and differently shaped sclerites of the aedeagal internal sac (in particular, the dorsal sclerite is narrow, elongate and thread-like, at least apically). A few species with consistently spotted elytra (e.g., L. bakewelli Baly, L. ruficollis Baly) possess all of the characters in agreement with the impressa-group as it is outlined above.
Below, we are introducing several new synonymies and rearranging several previously proposed synonyms. There are two primary reasons for these changes in taxonomy of the Oriental Lilioceris. First, in the course of this study we introduced for the first time the use of male genitalia characters into the diagnostics of the species, in particular focusing on the sclerites of the internal sac of the aedeagus. In general, male genitalia morphology in the Oriental species appeared to be quite conservative. However, minor differences in the shape of the aedeagus and characters of the aedeagal sclerites of the internal sac are stable and consistent across vast species ranges and among multiple individuals within populations. This consistent set of aedeagal characters provides reliable diagnostics among species that are otherwise monotonous with respect to color and punctation. As a result, some misidentifications were corrected and species identities clarified. Second, the accepted concept of L. impressa (Gressit andKimoto 1961, Kimoto andGressit 1979), considered to be the most widespread and variable species, was found to be erroneous after the study of Fabricius' authentic material. Contrary to the opinion of Gressit and Kimoto (1961) and illustrations in their monographs, the true L. impressa has an isolated oblique setose band on the posterior part of the outer metasternal disc (see Fig. 13c and Fig. 14c in Gressit andKimoto 1961 andGressit 1979, respectively). Consequently, their concept of L. impressa corresponds primarily to L. laosensis, with some L. egena specimens identified as L. impressa occasionally.

Key to the species of Lilioceris impressa-group
1 Outer parts of metasternal disc mostly free of setae; isolated latero-posterior setose patches or lateral extensions of anterior setose margins present in some species and widely scattered single setae may present on unworn individuals ..  Comments. The holotype of L. cheni is housed in the Bishop museum collection. This collection contains 5 more specimens of this species all identified as L. cheni by Kimoto in 1967 and1977, none of them are marked as paratype. Biology. Pale white, oblong eggs of L. cheni are deposited in loosely aggregated clusters on leaves of its host plant, Dioscorea bulbifera (air potato: Dioscoreaceae). Females deposit, on average, more than 1200 eggs during their lifetime. The eggs become yellowish as the embryo develops and dark reddish eye spots appear mid-way through the incubation period. The entire incubation period requires about 4 days. The larvae are yellowish at first, becoming grayish in later instars, with black legs, head capsule, and prothoracic shield. They are often covered with a slimy substance to which fecal material adheres. Larvae feed gregariously and skeletonize the leaves from the underside. Young leaves are preferred but they also consume older, tougher leaves and are able to feed on the aerial bulbils. Complete development of the four larval instars requires about 8 days, with each instar lasting about 2 days each. When fully grown, larvae drop from the host plant to the soil which they quickly enter. They then produce a whitish oral exudate that hardens into a foam-like cocoon. Pupation often occurs gregariously, with several pupae clumped together within a matrix of this material. Adults emerge in about 16 days, begin mating in about 10 days following emergence, and initiate oviposition about 5 days later. The adults live 3 months or more and can survive a month without food.
In the Katmandu Valley of Nepal, the host plant drops its leaves during the cool, dry winter forcing the adult beetles to over-winter beneath debris on the ground. The adults emerge during mid-May to early June. Oviposition begins during late May and continues till mid-June. Females lay about 90 eggs/day during a 13-day period of ovipositional activity. Overwintered adult beetles live until mid-July, for about 76 days after emergence.  (Weise, 1922) http://species-id.net/wiki/Lilioceris_egena Figs 2,9,15,25 Crioceris egena Weise, 1922:41 (Type locality: Fukien. Type depository, unknown). Crioceris coomani Pic, 1928:88 (Type locality: Vietnam, Tonkin. Lectotype designated here, MNHN). New synonym Diagnosis. Occipital area without longitudinal furrow. Pronotal disc without or with few small punctures, larger punctures present only along pronotal mid-line at least in anterior half. Lateral sides of pronotum around constriction with large punctures. At least half of metepisternal disc covered with setae. Outer metasternal disc mostly free of setae. Anterior setose fringes not expanded in antero-lateral corners. Apical elytral punctures strong. Apical elytral intervals distinctly raised. Internal sac of aedeagus with posterior part of dorsal sclerite in dorsal view more or less triangular, directed forward. Posterior part of dorsal sclerite in lateral view more or less triangular, directed forward. Comments. We were not able to find type material for this species. The type specimens were not found in Berlin (HUB) nor in Hamburg (ZMUH) (Peters, personal communication, August-September 2009). There is a strong possibility that the types were destroyed in Hamburg during World War II. However, two specimens identified as L. egena by J. Weise himself (including a male) were available for the study, and we based our concept of the species on these specimens. However, we are reluctant to designate the neotype at this moment, since an exhaustive search in other potential depositories has not been undertaken during the preparation of this manuscript. The male holotype of L. coomani perfectly corresponds to males of L. egena as we understand it now. This species is collected feeding on Lilioceris impressa (Fabricius, 1787) http://species-id.net/wiki/ Lilioceris_impressa  Figs 3, 7, 10, 16, 26 Crioceris impressa Fabricius, 1787:88 (Type locality: "Siam". Lectotype designated here, BMNH). Crioceris subcostata Pic, 1921a:2 (Type locality: China, "Shin-guey-fu". Lectotype designated here, MNHN). New synonym Crioceris ruficornis Pic, 1921b: 136 (Type locality: China, "Pe Yen Tsing (Yunnan)".

Lilioceris egena
Lectotype designated here, MNHN). Gressitt and Kimoto 1961:59 (synonymy). Crioceris laticornis Gressit, 1942:300 (Type locality: Hainan, Kwantung. Type depository: unknown). New synonym Lilioceris inflaticornis Gressit and Kimoto, 1961:50 (Type locality: "SE China [Kwantung, Fukien]. Type depository: CAS). New synonym Lilioceris maai Gressit and Kimoto, 1961:53  Comments. All five species put here into synonymy with L. impressa correspond to it well. There is slight variability in the size of the outer metasternal setose patch, and color varies from almost completely reddish-brown to typical black and yellow, but pronotal and elytral punctuation and sculpture as well as aedeagus characters are consistent between specimens. We did not observe any meaningful variability in the shape of the expanded distal antennomeres, which was used by Gressit and Kimoto in describing L. inflaticornis and L. laticornis. However, even these authors had some doubts about the use of these characters, even in the reference to the type series (Gressit and Kimoto 1961).
We did not thoroughly search for L. impressa-group specimens originating from outside of the area of interest, in particular from the islands and archipelagoes of the Oriental Region. However, all the island specimens in the USNM identified as L. impressa do not belong to this species and all lack the outer metasternal setose patch. So, the prevailing concept of L. impressa, being widely distributed across almost the entire Oriental Region, seems to be at least questionable. All the USNM L. impressagroup specimens originating from Sri Lanka, Andaman Islands, Greater and Lesser Sunda Islands, and Philippines were not identified as any of the species dealt with in this study and might indeed represent undescribed species closely related to other L. impressa-group species.

Diagnosis.
Occipital area with a shallow furrow, at least with a deep small fovea at midpoint. Pronotum posteriorly more or less convex, no distinct impressions present. Pronotal disc without or with few small punctures, larger punctures present only along pronotal mid-line, at least in anterior half. Lateral sides of pronotum around constriction impunctate. Outer metasternal disc mostly free of setae. Anterior setose fringes of metasternum not expanded in antero-lateral corners. Setae on metepisternae occupy only narrow line along inner margin. Apical elytral punctures weakened. Apical elytral intervals not raised. Internal sac of aedeagus with posterior part of dorsal sclerite in dorsal view wide and short, slightly triangular, directed forward. Posterior part of dorsal sclerite in lateral view widely triangular with dorsal side longer than ventral, directed forward, but at the same time posteroventrally.