Kelloggella, Jordan & Seale, 1905

Tornabene, Luke, Deis, Brian & Erdmann, Mark V., 2018, Evaluating the phylogenetic position of the goby genus Kelloggella (Teleostei: Gobiidae), with notes on osteology of the genus and description of a new species from Niue in the South Central Pacific Ocean, Zoological Journal of the Linnean Society 183, pp. 143-162 : 158-159

COMPARISONS WITHIN KELLOGGELLA

Kelloggella avaiki can easily be distinguished from all other species by the presence of bright blue spots on the body and fins, with a dark grey, dark blue or black background. Kelloggella cardinalis is uniformly green to light greenish-grey, with red-orange fins ( Fig. 18; also see photos in Allen, Steen & Orchard, 2007 and Allen & Erdmann, 2012). All other species of Kelloggella possess distinct vertical bars (or in females of some species, vertically paired spots or blotches) down the side of the body ( Fig. 19).

Five species of Kelloggella have been described that possess distinct vertical bars in males, including K. oligolepis (type locality Hawaii), K. quindecimfasciata (type locality Ryukyu Islands, Japan), K. centralis (type locality Rarotonga, Cook Islands), K. tricuspidata (type locality Hiva Oa, Marquesas) and K. disalvoi (type locality Easter Island) ( Fig. 19). Female coloration is either similar to the males or consists of paired spots or vertically oriented blotches instead of vertical bars. There has been confusion regarding which of these species are valid, with K. centralis being synonymized with K. quindecimfasciata ( Sawada, 1977) and K. quindecimfasciata being considered as synonym of K. oligolepis ( Hoese, 1975; the latter two are both currently considered valid; Eschmeyer, Fricke & van der Laan, 2017). Among these five nominal species with vertical bands, K. tricuspidata ( Fig. 19I) is perhaps the most distinct and is distinguished from the others in having fewer dark vertical bars (7 vs. 9–11), which are distinctly narrower (white interspaces equal to or wider than dark bars vs. being notably narrower than dark bars), and in possessing six or seven inner rows of tricuspid teeth in each jaw (vs. two to four inner rows). The other four nominal species have been distinguished from each other by authors by a combination of characters including modal differences in the number of bars on the body of males (i.e. 9 vs. 10 vs. 11); the presence or absence of sexual dichromatism; differences in the thickness of vertical bars on body; dentition; modal differences in number of branched caudal rays (15 vs. 16); modal differences in the number of pectoral rays (12 vs. 13 vs. 14); modal differences in the number of anal rays (I, 7 vs. I, 8) and modal differences in second dorsal rays (I, 10 vs. I, 11 vs. I, 12). In examining both K. oligolepis and K. disalvoi from their respective type localities, there is considerable variation in the number and width of the bands on the body, both between individuals of a population, and also before and after preservation within a single individual (e.g. Fig. 19A–F). In addition, sexual dichromatism has been variously reported for K. oligolepis . Hoese (1975) noted that K. oligolepis lacked pronounced sexual dichromatism. At the time, he was basing this on his observations of individuals from both Hawaii and Easter Island (= K. disalvoi ). Greenfield & Randall (2004) and Randall (2009) noted that K. oligolepis in Hawaii were sexually dichromatic ( Fig. 19K, L), and Randall noted the K. disalvoi were not. On the contrary, recent collections of K. disalvoi indeed showed variation in colour pattern with sex, and to a lesser extent, with size ( Fig. 19A–G). Given the overlap among some of the aforementioned characters (e.g. number of vertical bars), intraspecific variation in others (e.g. width of vertical bars and female colour patterns), the difficulty in accurately observing dentition from specimens that are not cleared and stained and the fact that some characters have been incorrectly reported in past studies (e.g. branched caudal-ray counts; see Winterbottom & Emery, 1986 for details), we are hesitant to take a strong stand regarding which species are valid. Therefore, we unfortunately refrain from including a key for the genus at this time. Additional geographic sampling is needed to sort out the validity of this species group, ideally in concert with an integrative taxonomic approach that includes data from DNA sequences, morphology and live/fresh coloration of both sexes. This approach has recently proved useful, if not essential, in clarifying the taxonomy of other diminutive, morphologically similar groups of gobies from the ultra-diverse tropical Indo-Pacific region (e.g. Winterbottom et al., 2014; Greenfield & Tornabene, 2015; Tornabene et al., 2015, 2016a).

Comparative cleared and stained material examined: USNM 440505, Kelloggella oligolepis ; USNM 253375, Eviota sp. ; USNM 225183, Eviota sp. ; USNM 221756, Eviota sp. ; USNM 230083, Eviota sp. ; USNM 71405, Eviota abax ; USNM 216580, Eviota distigma ; USNM 223063, Eviota sp. ; USNM 287162, Acentrogobius madraspatensis ;USNM30626, Lythrypnusdalli;FMNH 121346, Lythrypnus dalli ; USNM 216191, Gobiopsis macrostomus ; USNM 147966, Coryogalops ocheticus ; USNM 306266, Gobiodon citrinus ; USNM 287159, Oplopomus oplopomus ; USNM 244143, Amblygobius decussatus ; USNM 236664, Parioglossus raoi ; USNM 287163, Bathygobius cf. soporator; USNM 287161, Istigobius ornatus ; uncataloged, Nes longus ; USNM 435307, Bryaninops sp. ; USNM 435308, Cryptocentrus strigilliceps ; USNM 435309, Gobitrichinotus radiocularis ; USNM 439365, Kraemeria samoensis ; USNM 143153, Kraemeria bryani . UW 157173, Eviota sp. nov.; UW 157174 Pleurosicya mossambica ; UW 19036, Ptereleotris microlepis ; UW 11991, Paragobiodon echinocephalus ; FMNH 107439, Paratrimma nigramenta .