A new species of the suckermouth armored catfish genus Peckoltia is described from the lower Ventuari River, a tributary of the upper Orinoco River in Amazonas State, Venezuela. Specimens of this species were formerly included in the wide-ranging Amazonian species P. vittata, but a recent molecular phylogeny found Orinoco individuals to be distantly related to Amazon Basin individuals spanning the range of P. vittata syntypes. Detailed morphological examination confirmed distinctiveness of Orinoco specimens, and found them to be diagnosable from true P. vittata by having generally greater than 25 teeth (vs. less), spots on the nape (vs. nape lacking spots), the upper lip with two to three black bar-shaped markings in a line like a moustache (vs. lips generally with a hyaline wash), and by the snout having a medial black line disconnected from the moustache markings (vs. medial snout stripe connected to a bar just above the lip). Peckoltia wernekei displays remarkable genetic similarity to its sister species, P. lujani, but differs morphologically by having dentary tooth rows meet at an angle less than 90° (vs. >90°), by having large faint blotches on the abdomen (vs. abdomen with no blotches), by a smaller internares width (21.2–26.6% vs. 28.5–46.5% of interorbital width), and a larger dorsal spine (148.1–178.6% vs. 80.1–134.5% of abdominal length).

Ancistrini , Hypostominae , Molecular Phylogeny, Morphology, Peckoltia , Systematics, Taxonomy

Peckoltia Miranda Ribeiro, 1912 is a genus of suckermouth armored catfishes (Loricariidae) with 18 currently described species (Armbruster et al. 2015). Armbruster (2004; 2008) had restricted Peckoltia to those species of Ancistrini that had dentary tooth rows meeting at an angle of 90° or less and that lacked the synapomorphies of similar genera like Hypancistrus and Panaqolus (the latter then the Panaque dentex group); however, this arbitrary definition was not supported by the molecular analysis of Lujan et al. (2015), so Armbruster et al. (2015) recognized an expanded Peckoltia that currently lacks a morphological diagnosis, but is strongly monophyletic based on molecular evidence (Lujan et al. 2015).

Armbruster (2008) had recognized a wide-ranging Peckoltia vittata (Steindachner, 1881) that included specimens from the Maranhão to the Madeira and upper Orinoco, but suggested that this putative range of P. vittata included multiple species. The syntype series of Peckoltia vittata itself contains specimens spanning more than 600 km of the main channel of the Amazon River, from the Xingu River to the Madeira River, making it difficult to know where one might reliably find the true P. vittata. Specimens from the Xingu, Madeira and Orinoco that were morphologically consistent with Peckoltia vittata were found to be polyphyletic in Lujan et al. (2015) and Lujan et al. (in review, Fig. 1). In addition, P. greedoi was recognized as distinct from P. vittata by Armbruster et al. (2015; specimens of P. greedoi were originally included among the specimens of P. vittata examined by Armbruster 2008).

Figure 1. 

Phylogenetic relationships within the Peckoltia Clade (sensu Lujan et al. 2015), from Lujan et al. (in review). Results based on analysis of a 4293 base pair alignment consisting of two mitochondrial (16S, Cyt b) and three nuclear loci (RAG1, RAG2, MyH6). Node support values given in order as Bayesian posterior probability, maximum likelihood bootstrap and maximum parsimony bootstrap.

Lujan et al. (2015) found the putative specimens of Peckoltia vittata from the Orinoco to be strongly supported as sister to another Orinoco species, P. lujani Armbruster, Werneke and Tan, and part of a clade with another undescribed species known as Peckoltia sp. n. Meta L147. In this study, we re-examine specimens identified as P. vittata from the upper Orinoco and describe them as a new species clearly distinguishable from Amazon Basin P. vittata. In addition, we provide a key to all of the current species of Peckoltia.

Methods follow Armbruster (2003) with the addition of counts of mid-dorsal and mid-ventral plates (the number of plates in these series from the head to caudal fin and excluding the last, triangular plate, which is beyond the hypural). Institutional abbreviations are as in Sabaj Pérez (2014). Names of skeletal characteristics are as in Schaefer (1987) and Geerinckx et al. (2007) and of plate rows as in Schaefer (1997). Full morphometric dataset is presented in Suppl. material 1, locality information for species described below is presented in Suppl. material 2.

The disparity between the morphological and molecular phylogenies of Armbruster (2004; 2008) and Lujan et al. (2015) and Lujan et al. (in review) are likely due to homoplasy and convergence in the morphological dataset, and nowhere is this more obvious than in the upper Orinoco clade containing Peckoltia lujani, P. wernekei and Peckoltia sp. n. Meta L147 (Fig. 1). Peckoltia wernekei was included in PeckoltiasensuArmbruster (2008) because of its stocky body and dentaries forming an angle of less than 90° (Fig. 6). The molecular phylogeny found P. wernekei to be sister to P. lujani, a species with dentary tooth rows forming almost a straight line and a more elongate body (Fig. 5a). Despite the differences in jaw angle, which likely has ecological and functional repercussions (Lujan and Armbruster 2012), P. wernekei and P. lujani differ from one another by only one of approximately 600 base pairs (bp; <0.2%) sequenced from the mitochondrial 16s gene and three of approximately 1150 bp (<0.3%) sequenced from the mitochondrial cytochrome b gene, whereas P. wernekei has 9.6% cyt b sequence divergence from P. vittata in the Xingu and 7.5% divergence from P. vittata in the Madeira (Lujan et al. in review).

Given the nested phylogenetic position of Peckoltia wernekei among two other species with relatively straight tooth rows (P. lujani and P. n.sp. Meta L147; Fig. 1), its highly-angled dentary tooth row angle is likely derived and an example of convergence upon the condition observed in many other Peckoltia. Given the limited sequence divergence between P. wernekei and P. lujani, this jaw evolution has either occurred very quickly or there has been very recent mitochondrial introgression. Mitochondrial introgression was observed between the sympatric Panaqolus koko and Peckoltia otali (Fisch Muller et al., 2012) where the two species had similar sequences for the mitochondrial COI gene suggesting close relationships, but sequences in the nuclear F-Reticulon4 gene showed no such close relationship. In this case, P. wernekei and P. lujani are not known to be sympatric and nuclear genes are also very similar (≤0.3% divergence in RAG1, RAG2 and MYH6; Lujan et al., in review), so we believe that the genetic similarity between these species is indicative of close evolutionary relationships and not hybridization. If such a pattern can be seen among closely related species, then many elements of jaw morphology are likely to be convergent across the phylogeny. Indeed, highly-angled jaws were seen in several clades even within the morphological phylogeny (Armbruster 2004; 2008). We are now testing this hypothesis with a phylogenetically explicit examination of jaw morphological diversity across the Hypostominae.

Peckoltia wernekei is illustrative of an increasingly clear biogeographical pattern in which the fish fauna of the Orinoco River upstream of its confluence with the Ventuari undergoes replacement or turnover between this confluence and the large Autures Rapids downstream. We hypothesize that this is due to an environmental filter similar to that documented by Winemiller et al. (2008) in the nearby Casiquiare Canal. The limnology of the Casiquiare Canal displays a gradient from relatively neutral or slightly acidic pH with few tannins at its origin in the upper Orinoco, to highly acidic pH with high tannin load at its confluence with the Negro River. Likewise, the physicochemical parameters of the clearwater upper Orinoco River shift a short distance downstream from the Ventuari River in response to its confluence with one major blackwater tributary (the Atabapo River) and one major whitewater river (the Guaviare/Inirida River) along its left-bank.

Many loricariid species are known only from the Ventuari River and nearby reaches of the Orinoco main channel. These include: Baryancistrus demantoides, B. beggini, Hemiancistrus subviridis, Hypancistrus contradens, H. furunculus, H. lunaorum, Leporacanthicus cf. galaxias, L. triactis, Hypancistrus vandragti, Peckoltia lineola, Pseudolithoxus tigris, and Pseudancistrus pectegenitor (Armbruster 2008, Armbruster et al. 2007, Lujan and Armbruster 2011, Lujan and Birindelli 2011, Lujan et al. 2007, 2009, Werneke et al. 2005). At least two other species have a disjunct distribution inclusive of the Ventuari River and the Caura River but are absent from intervening reaches of the Orinoco River main channel (Limatulichthys nasarcus, and Pseudolithoxus anthrax; Lujan and Birindelli 2011, Londoño-Burbano et al. 2014), and in four of these instances there is strong morphological or molecular evidence that sister species are allopatrically distributed upstream vs. downstream of the confluence of the Orinoco and Atabapo/Guaviare/Inirida rivers. In addition to the Peckoltia lujani/P. wernekei pair, there is Hypancistrus debilittera and H. furunculus, Pseudolithoxus kelsorum and P. tigris, and Hemiancistrus sp. n. L128 (Dignall 2014) and Hemiancistrus subviridis.

This project was supported by NSF grants DEB-0107751, DEB-0315963, and DEB-1023403 to JWA and OISE-1064578 (International Research Fellowship) to NKL, with additional support to NKL from DEB-1257813 (the iXingu Project), National Geographic Committee for Research and Exploration Grant #8721-09, the Coypu Foundation of New Orleans, and the Canada Department of Fisheries and Oceans. We are indebted to numerous people for help when visiting museums and for help in collecting specimens including: Mark Sabaj Pérez, John Lundberg, Marry Anne Rogers, Barry Chernoff, Phillip Willink, Mark Westneat, Richard Vari, Susan Jewett, Jeffrey Williams, Karsten Hartel, Lawrence Page, Robert Robins, Lesley de Souza, Donald Taphorn, Roberto Reis, Luiz Malabarba, Pablo Lehman, Paulo Buckup, Michael Retzer, Kevin Cummings, Christine Mayer, Oscar León Mata, Augusto Luna, Brooks Burr, Jeffrey Stewart, Matt Thomas, Mario de Pinna, Oswaldo Oyakawa, John Friel, Hernan Ortega, Tiago Carvalho, Vanessa Meza-Vargas, and Erling Holm. This paper is contribution No. 724 of the Auburn University Museum of Natural History.