A historical overview of the classification of the Neotropical tribe Zammarini ( Hemiptera , Cicadidae ) with a key to genera

An overview is given of the past century’s confusions concerning some key characters within this Neotropical cicada tribe. Th e limits of the genera Zammara and Orellana have been redefi ned and a comprehensive checklist is included for the tribe. An illustrated key to the genera of the tribe Zammarini (Hemiptera: Cicadidae) is provided; this is the fi rst key to both males and females of this tribe. Odopoea perspicua Distant, 1905 is transferred from Zammarini to the African tribe Platypleurini to become Canualna perspicua (Distant, 1905), comb. n.


Introduction
Th e Neotropical tribe Zammarini (Hemiptera: Cicadidae), belongs to the subfamily Cicadinae, the latter characterized by the presence of tymbal covers.Zammarini are easily recognizable by their strongly produced pronotal fl anges.Males of this tribe have a more or less bulbous tymbal cover, which is sometimes useful as a specifi c character.
Over the past century, the tribe Zammarini has been studied and discussed by several hemipterists, probably because these cicadas are relatively large, colorful, and possess unique, strongly produced pronotal fl anges.
Th is paper is the fi rst one in a series of papers dealing with the taxonomy of the tribe Zammarini, with the ultimate goal of publishing a monograph on the tribe based on morphological, molecular, ecologic, biogeographic and acoustic data.

History of the tribe
Th e tribe Zammarini was described by Distant (1905) under the "division" name Zammararia.Distant (1905) defi ned his Zammararia to consist of Zammara Amyot & Audinet Serville, 1843 (7 spp.), Odopoea Stål, 1861 (13 spp.), two new genera, Orellana and Miranha, which are based on respectively Z. columbia Distant, 1881 andZ. imbellis Walker, 1858, and two new species Orellana brevis and Odopoea perspicua.Th ese 24 species attributed to 4 genera were catalogued a year later in Distant 's catalogue (1906a).In 1915, Van Duzee fi rst used the tribal name Zammarini.Distant (1912) described the genus Uhleroides as a new genus of Zammarini; Uhleroides is restricted to Cuba and Hispaniola.Without explanation, Kato (1932) transferred Uhleroides to the tribe Th ophini.Moulds (2001) recently transferred Uhleroides back into Zammarini, based on the following characters: "the presence of transverse grooves towards the distal ends of the postclypeal ridges, an antennal plate that reaches almost to the eyes, fore wings which carry blotch-like infuscations at distal ends of apical veins 1-7 and at base of apical cells 2, 3 and 5, a narrow basal cell to the fore wing with veins M and CuA meeting the cell close together and a hind wing anal lobe that is narrow."Haupt (1918) and Delétang (1919) both described new genera within Zammarini (Adusella and Edholmbergia, respectively); however, both generic names were synonymized to Odopoea by Torres (1945).Distant (1920) described Juanaria, a monospecifi c Zammarini genus endemic to Cuba.Davis (1928Davis ( , 1934) ) added two more genera to the tribe, Borencona (a monospecifi c genus endemic to Puerto Rico) and Chinaria (restricted to Mexico and the Dominican Republic).
Th e most recent generic addition to Zammarini was made by Boulard and Sueur (1996), who described the monospecifi c genus Zammaralna (a group endemic to Venezuela) as the "taxon frère de Zammara" [sister taxon to Zammara].

Tarsomeres and head width: contradictions and confusions in the literature
Th e presence of 2 or 3 tarsomeres, a character that should be easy to evaluate, has led to considerable confusion within Zammarini.Within Cicadidae, a reduction of the number of tarsomeres is an unusual condition; indeed, within Zammarini, Zammara is the only genus with 2 tarsomeres instead of 3. Amyot and Audinet Serville (1843) originally described Zammara as having 2 or 3 tarsomeres; the type species, Z. tympanum (Fabricius, 1803), has 2 tarsomeres.Th e other species of Zammara described in this work, Z. strepens, was stated (erroneously) as having: "les tarses distinctement de trois articles" [tarsi clearly three-jointed].
Distant (1881a) described Z. columbia as having 3 tarsomeres (in error).Later, Distant (1905) erected the genus Orellana, and assigned Z. columbia as the type species of this genus; oddly, he stated clearly in his generic description for Orellana: "tarsi twojointed," based on a type species that he previously described as having 3 tarsomeres.In this same publication, Distant described the species Orellana brevis as having "twojointed tarsi."One year later, Distant (1906a) transferred Z. nigriplaga Walker, 1858 to the genus Orellana based on having the "head (including eyes) only about two-thirds the breadth of base of mesonotum" in contrast to "about as wide" for Zammara.Jacobi (1907) suggested that both Distant (1881a), and Amyot and [Audinet] Serville (1843) made errors in interpreting tarsomere morphology for respectively, Orellana columbia and Zammara strepens.Jacobi implied that the errors might have been due to the "nicht ganz gutem lichte und geringeren optischen hilfmitteln" [suboptimal lighting and inferior optical tools] than his "Zeissschen Binokularlupe mit 35facher vergrösserung."[Zeiss dissecting microscope with 35× magnifi cation].He also transferred O. nigriplaga (Walker, 1858) back into Zammara.
Subsequently, Distant (1914) transferred Z. nigriplaga and Z. brevis, again, into Orellana, based on the breadth of the head (as discussed above) and because these species supposedly have 2 tarsomeres, since he described the genus Orellana again as having tarsi "two-jointed."Schmidt (1919) discussed Jacobi's and Distant's works and moved Z. nigriplaga, Walker 1858 and O. brevis Distant, 1905 (back) to Zammara because both species have 2 tarsomeres.He also moved Z. strepens Amyot & Audinet Serville, 1843 to Orellana and described the new species Orellana bigibba; additionally, Schmidt suggested that the genus description for Orellana should be changed to "tarsen dreigliedrich" [tarsi three-jointed].Goding (1925) based the separation of Zammara and Orellana on Distant's (1905) key to genera, namely "head (including eyes) only about two-thirds the breadth of base of mesonotum" for Orellana and "head (including eyes) about as wide as the breadth of base of mesonotum" for Zammara.Based on this separation, Goding transferred Z. nigriplaga and Z. brevis again into Orellana and described two more species in the genus: O. pulla and O. brunneipennis, not mentioning the number of tarsomeres for either species.Boulard (1975Boulard ( , 1976) ) treated O. brevis, O. bigibba and O. columbia as belonging to Zammara, though without specifi cally transferring them back into Zammara.It is not clear whether, in doing this, he intended not to recognize Orellana as a valid genus, since by transferring O. columbia (the type species of Orellana) back to Zammara, Orellana would become a junior (subjective) synonym of Zammara.Boulard and Sueur (1996) praised Distant for his insight in the importance of the number of tarsomeres as a generic character, and gave the example where he [Distant] used it to separate Miranha (trimére) and Orellana (dimére).So it can be assumed that Boulard (1975Boulard ( , 1976) ) did not intend to synonymize the genus Orellana with Zammara.Boulard and Sueur (1996) discussed the number of tarsomeres briefl y when they erected the new genus Zammaralna, defi ning it in the following way: "Habitus de Zammara, mais tarses á trois articles; nervures médiane et cubitale naissant du meme angle de la cellule basale et poursuivant, juxtaposes, plus ou moins longuement (pas de tronc commun)" [Similar to Zammara, but with 3 tarsomeres; median and cubital veins arising at the same angle from the basal cell and from there on juxtaposed, over a more or less long distance (no common base)].

Results
My research for the present paper proves that Jacobi (1907) was correct that the number of tarsomeres in the descriptions of Orellana columbia Distant, 1881 and Zammara strepens Amyot & Audinet Serville, 1843 are wrong and that in fact Orellana columbia has 3 tarsomeres, while Zammara strepens has 2 tarsomeres.Schmidt (1919) was also correct in that the genus description for Orellana should be changed to "tarsi threejointed" (i.e.having 3 tarsomeres).Furthermore both nigriplaga and brevis should be transferred to Zammara because they each possess 2 tarsomeres.Th e species bigibba has 3 tarsomeres and therefore should stay in Orellana.I have not seen the types of Orellana pulla, Goding 1925 or Orellana brunneipennis Goding, 1925, nor specimens that belong to either of these species.Because the type specimens have not been located yet, the generic placement of the latter two species is tentative.Th is is especially so because Goding (1925) separated Zammara and Orellana based on the width of the head, a characteristic not valid to separate these genera.Distant (1905) described Odopoea perspicua and listed the type locality as "Saõ Th omé".After seeing pictures of the type, it is clear that this species is misplaced in Zammarini as it clearly belongs to the African tribe Platypleurini.Th is species is very closely related to Canualna liberiana (Distant, 1912), a species occurring on the island of Saõ Th omé (M.Villet, personal communication).Distant erroneously thought that Saõ Th omé referred to a location in Brazil, but instead this specimen is almost certainly from the African Island Saõ Th omé, even more so because the collector was "Negreiros", a painter born on that island.For the above reasons the species is transferred to Canualna perspicua (Distant, 1905), new combination.

Checklist of genera and species currently included within the tribe Zammarini.
In the following check-list only the type locality for each species is given, further distribution records will be discussed in future papers.

Key to the genera of the tribe Zammarini
Below, the fi rst key to both males and females for all (described) species within the Tribe Zammarini is provided.Th ere are several genera (Plautilla Stål 1865, Procollina Metcalf 1963, Daza Distant 1905, Aragualna Champagnet, Boulard & Gaiani 2000, Onoralna Boulard 1996) that are at present placed outside of Zammarini, which may prove to be members of the tribe.Currently molecular data and additional morphological data are being collected for each to help determine their taxonomic position within the Cicadidae.

Figure 1 .
Figure 1.Accumulation of species / publication year for the tribe Zammarini, since Distant recognized the tribe in 1905.

Figure 2 .
Figure 2. CuA and M veins arising together (left) or separate (right) from the basal cell.