Levu Kirkaldy, 1906

Zelazny, B. & Webb, M. D., 2011, 3071, Zootaxa 3071, pp. 1-307 : 36-37

Genus Levu Kirkaldy

Levu Kirkaldy, 1906: 434 . Type species: Levu vitiensis Kirkaldy, 1906 , by monotypy.

Diagnosis. In the genus Levu Kirkaldy the forewing venation is similar to that of the genera Rhotana Walker , Rhotanella Fennah , and Dichotropis Kirkaldy. The genus can be recognized by the combination of the following characters:

1. Smaller and with more narrow forewings compared to the species of Rhotana , Rhotanella and Dichotropis .

2. Facial carinae always meeting in front of the eyes.

3. Forewings powdered with 5 narrow costal cells (the powder may have been lost, however, in museum specimens). In live specimens (3 species observed) the wing surfaces form a common plane.

4. Forewing with M leaving Sc+R usually near the base of the basal median cell.

5. Elaborate colour patterns on the forewings (like in Rhotana and Rhotanella ) missing.

6. Male pygofer never with prominent lateral projections.

The largest species of Levu resemble the smallest species of Rhotana . However, the combination of the above other characters rarely leave doubt about the assignment of a species.

Gender. We consider the gender of the genus name ' Levu ' to be masculine. Kirkaldy did not explain the origin of the name or stated its gender. The genus and the type species ( Levu vitiensis ) have been evidently named after the Fijian island of Viti Levu , the type locality. In Fijian the word 'levu' means 'big' ( Viti Levu = big land) and therefore, the name is not of Latin or Greek origin. This means the gender of the genus name should be decided according to articles 30.2.3 and 30.2.4 of the Code. However, both articles appear to contradict each other in this particular case. Article 30.2.3 ("...the name takes the gender indicated...[by the]...adjectival species-group names ...") means the gender could be masculine or feminine, but not neuter because the neuter form would be "vitiense". In contrast to this, the name should be neuter according to article 30.2.4 since it ends in "-u". The wording of the articles under section 30.2 of the Code indicates that the articles should be followed sequentially to decide on the gender. E.g. 30.2.3 ("If no gender was specified...") refers to cases where 30.2.2 (the gender is specified by the author) does not apply and 30.2.4 ("If no gender was specified or indicated ...") refers to cases where neither article 30.2.2 nor 30.2.3 applies. We therefore, consider article 30.2.3 to have priority over 30.2.4 and we exclude the neuter option. This is also in the interest of taxonomic stability, because adopting the neuter option would mean that we have to change the type species to 'vitiense'. We are then left with the masculine and feminine options, where we reject the latter because it would apply only to names ending in "-a" as stated in 30.2.4. It is not clear why in the past Levu has been treated as feminine in most cases. On the other hand, Muir ( Levu rufulus ) and Zelazny ( Levu elongatus ) used masculine names. In no case was a neuter name used.

Distribution. Unlike the other genera of Rhotanini , the genus is commonly found in the Pacific islands. The distribution of the genus is shown in Figure 121 and in Table 4.

Grouping and keys. While some species of Levu are obviously closely related, grouping the species appears not to be possible for the large majority without the help of a molecular phylogenetic analysis. We describe only one species group composed of 7 species. After describing the remaining species we provide keys to the species of the Pacific islands and New Guinea.