Page 7 - zt03963p054

FIGURE 3. Distribution of COI pairwise K2P distance values, performed by ABGD, showing the "barcode gap" between the
putative maximum co-specific divergence (0.01) and the minimum congeneric divergence (0.05).

Discussion

Being aware of the controversial taxonomic status of Kyphosus species and the historical problems in delimiting
the different species and identifying them, we carefully examined the specimen collected off Sicily. Its
morphometric and meristic data matched those of K. vaigiensis reported in Knudsen & Clements (2013a). The
mitochondrial DNA sequences matched those of K. vaigiensis as well: its COI and 16S sequences were placed
within a monospecific clade together with haplotypes from Indo-Pacific K. vaigiensis and Atlantic K. incisor, the
latter consequently having been synonymized with K. vaigiensis by Knudsen & Clements (2013a).

Its full stomach indicates that the Sicilian specimen was in good condition and agrees with its known
herbivorous habits (Clements & Choat, 1997).

Overall, the data presented here resolve the taxonomy of the Sicilian specimen, and highlights some problems
with Sakai & Nakabo’s taxonomy (2014).

The specimen collected in the Ionian Sea (K. saltatrix, in Kiparissis et al., 2012) “matches K. sectatrix in both
morphological and molecular… characters” (Knudsen & Clements, 2013a: 60). These authors consider the
specimens recorded by Tortonese (1954), Merella et al. (1998), Hemida et al. (2004) and Francour & Mouine
(2008) as K. sectatrix, yet caution: “Whether other previous reports of K. sectatrix from the Mediterranean are
valid is difficult to say” (Knudsen & Clements, 2013a: 60). The NJ trees clearly support the genetic identity of this
Ionian specimen as K. sectatrix, and indicate a pattern consistent with the present-day existence in the
Mediterranean basin of two Kyphosus species, namely K. sectatrix and K. vaigiensis. Such conclusion does not
solve the discrepancies between the two mentioned revisions dealing with the "sectatrix-case". In fact, the sectatrix
name is treated by Knudsen & Clements (2013a) as a widespread Atlantic–Mediterranean–Indo-Pacific species,
while Sakai & Nakabo (2014) split it into two endemic Atlanto–Mediterranean species, and the only specimens
analyzed here comes from Atlanto–Mediterranean region.

However, the vaigiensis-incisor clade supports the existence of a wide ranging species, over the Atlantic–
Mediterranean–Indo-Pacific region, which matches the revision by Knudsen & Clements (2013a). The ABGD
analysis was applied to the COI dataset to prove the efficiency of the DNA barcoding in delimiting the
Mediterranean species. According to the ABGD analysis the vaigiensis-incisor clade seems to be well delimited.
Nonetheless, it is increasingly recognised that accurate species delineation requires a multilocus approach taking
into account two or more independent markers from the same individuals, which necessarily involves sequencing
nuclear genes (see Knudsen & Clements, 2013a).

The presence of Kyphosus species in the Mediterranean Sea is not a recent event. An older specimen, collected
in the Gulf of Palermo, Sicily, and described as Pimelepterus boscii sicula by Doderlein (1883), was regarded as
junior synonym of K. bigibbus (Lacepède, 1801) by Knudsen and Clements (2013a), but considered as incertae
sedis by Sakai and Nakabo (2014).

KYPHOSUS VAIGIENSIS RECORDS FROM THE MEDITERRANEAN SEA Zootaxa 3963 (1) © 2015 Magnolia Press · 51