What can goby otolith morphology tell us?
The taxonomic information inscribed in otoliths has been widely ignored in ichthyological research, especially in descriptions of new fish species. One reason for this is that otolith descriptions are per se qualitative, and only a few studies have presented quantitative data that can support assignments of otoliths to individual species or permit differentiation between higher taxonomic levels. On the other hand, in palaeontology, otoliths have been employed for the identification and taxonomic placement of fossil fish species for over 100 years. However, palaeontological otolith data is generally regarded with suspicion by ichthyologists. This is unfortunate because, in the Cenozoic, the fossil otolith record is much richer than that based on skeletons. Thus fossil otoliths are a unique source of information to advance our understanding of the origin, biogeographical history and diversification of the Teleostei. This case study deals with otoliths of the Oxudercidae, which, together with the Gobiidae, encompasses the 5-branchiostegal-rayed gobiiforms. The objective was to determine whether the five lineages of the Oxudercidae, and individual species of the European Pomatoschistus lineage, could be distinguished based on the quantification of otolith variations. The data set comprises otoliths from a total of 84 specimens belonging to 20 recent species, which represent all five lineages of the Oxudercidae (Mugilogobius, Acanthogobius,Pomatoschistus, Stenogobius, Periophthalmus), and five fossil otoliths of †Pomatoschistus sp. (sensu Brzobohatý, 1994). Ten measurements were taken on each otolith and 23 otolith variables were computed and used for univariate and multivariate analyses. The results indicate that otolith morphometry (i) is capable of identifying the Pomatoschistus andPeriophthalmus lineages among the Oxudercidae, but is of limited use in the separation of the other three lineages; (ii) can reliably distinguish the sand gobies (a distinct clade within the Pomatoschistus lineage) from other members of the Pomatoschistus lineage; and (iii) supports a previous assignment of fossil otoliths to †Pomatoschistus sp. as an ancient genus of the sand gobies. With its middle Miocene age (15 m.y. ago), †Pomatoschistussp. represents the oldest record of a sand goby species to date. We discuss possible relationships between distinct otolith morphologies, biogeographic distribution and lifestyles for Pomatoschistus and the Periophthalmuslineage, and also for the sand gobies. We conclude that otolith morphology, combined with morphometry, can be considered as an autapomorphy for an individual species from the studied groups. In addition, it also appears to contain a phylogenetic signal, but more work is needed to evaluate this fully.