New Hypancistrus

[Silurus]

Araújo, F, M Ferreira, I Monteiro & W Wosiacki, 2024. A new species of Hypancistrus Isbrücker & Nijssen 1991 (Loricariidae, Siluriformes) from the rapids of the middle Rio Tocantins. Journal of Fish Biology doi: 10.1111/jfb.15971

Abstract

The Hypancistrus genus is recognized in the Río Orinoco basin and Rio Xingu in the Guiana and Brazilian Shields, respectively. Some of its species are important in ornamental fishing. Despite this significance, many other undescribed species are still awaiting to be named. Here we describe a new species of Hypancistrus found on bedrock in the Rio Tocantins, representing an extension of the distribution of the genus. Also, a multigene phylogeny is presented to evaluate the taxonomic position of this species concerning congeners. The new species differs from all congeners by (1) hypertrophied odontodes on cheeks reaching beyond the cleithrum, (2) a supraoccipital crest conspicuously elevated, (3) a supraorbital crest slightly convex, (4) oblique bars on the anterior part of the body, (5) a dark E-shaped mark on the snout, (6) three oblique dark bars on the anterior part of the body and horizontal vermicular bars from the pectoral girdle to the posterior insertion of the dorsal fin, (7) a thin light gray bar on the posterior of the head extending across the branchial opening, (8) a tan background color, (9) a developed suspensorium with a diminished appendix in the metapterygoid, and (10) a dentary plate robust significantly fused with the angulo-articular bone. The molecular phylogenetic results show the new species forming a group with Hypancistrus zebra (Brazilian Shield—Rio Xingu) as a clade, a sister group of a monophyletic group consisting of all congeners from the Río Orinoco.

[bekateen]

Hypancistrus parkateje

[Jools]

It looks a _lot_ like a Peckoltia - that's really interesting. I can't think of another vertically striped species?

Cheers,

Jools

[HaakonH]

I discussed this fish with Leandro Sousa when he and hos crew were on a survey in the Rio Tocantins. We did speculate at first if it was indeed a Hypancistrus, but we concluded it was

Peckoltia sp(L080)

. The dentition doesn't really fit with either genus. I'm unsure if there were more than one species collected, or if the authors are aware of the existence of Peckoltia sp. L80. Leandro is sceptical to this description.

- Haakon

[bekateen]

There's no way this new species is the same as the fish we show here on PCF as

Peckoltia sp(L080)

. Compare the mature male L080 shown in our CLOG page to the holotype of

Hypancistrus parkateje

(which appears to be a mature male) from the paper - They have totally different sexual dimorphisms.

Weirdly, the paper reports finding no sexual dimorphism, in spite of the paper using 21 specimens (the holotype and 20 paratypes), choosing a big male as the holotype, and choosing to list the giant cheek odontodes as a species-distinguishing trait (rather than as a sex dimorphism); moreover, the paper says this about the species:

"Fin spines and rays supporting odontodes; conspicuously stretched on pectoral-fin spines."

That made me wonder if all the specimens were males. I'm now looking at the supplemental data, specifically the Excel sheet they provide with morphometrics. My first observation of that file is that, for some inexplicable reason, they did not include the raw data on SL (except for holotype). Put that aside and move on to sexual dimorphism: Thinking of how male

Hypancistrus

typically have wider, shorter heads than females, I compared the raw data on relative cleithral width (as an estimate of head width) to relative head length, then compared the ratio of these (CW/HL) against head length. I obtained this outcome:

What I see here is that even though the holotype has one of the longest bodies collected (~65mm SL, vs SL of paratypes being 44-70mm), it also has the shortest relative head length (31% of SL; see "holotype" on graph). Then when you compare all the paratypes, most of their relative head lengths and cleithral widths cluster nicely (with about the same relative cleithral width as the holotype), except one individual (dot in red circles) that has a very long and very narrow head. The other interesting detail in this data is how stable the CW/HL ratio is for the 19 clustered paratypes, in spite of how variable their head lengths are (see green horizontal line in second graph). Could it be that these are all subadults/juveniles which haven't started to express sexual dimorphism? (meaning the other two are one male and one female).

This also makes me wonder what the SL values were for these paratypes? Was this one individual particularly large or small? Were the 19 other paratypes that clustered together about the same SL as each other, or similar to the other two outlying individuals? Might the authors have had only one female in the whole series (or since SLs are not given for paratypes, maybe this individual is a juvenile) and that may be why they didn't define sexual dimorphism.

The other possibility is that as the paper implies, even females have massive cheek odontodes. But if that's true, what about head shape and pectoral odontodes on those paratypes?

Cheers,
Eric

[bekateen]

Looking back at the original DATZ photo (HERE), I'm wondering if the first

L080

was a Peckoltia or if it was

Hypancistrus parkateje

. The solitary photo in that article shows a rather long maxillary barbel, so I'm guessing the true L080 is a Peckoltia, not a Hypancistrus.

Cheers, Eric

[bekateen]

The dentition doesn't really fit with either genus.

Maybe a new genus is in order?

[HaakonH]

The dentition doesn't really fit with either genus.

Maybe a new genus is in order?

Maybe? Peckoltia has become a bit of a trashbin and needs splitting eventually, but this particular fish is still kind of shrouded in mystery. As you so brilliantly pointed out in your analysis of the material in the description, it seems to me we are missing some details here. I would certainly like to take a closer look at more specimens and compare them in more detail to Peckoltia sp. L080. It's certainly not unusual that you find a Hypancistrus and a Peckoltia looking quite similar, sharing the same habitat. That's another thing Leandro and I discussed while he was investigating the material he collected. However, I don't think Leandro collected any fish looking quite like the holotype in question here, so it's quite possible that he in fact only collected the actual Peckoltia sp. L80. I don't know.

Just a thought on how you find two specimens being different from the remaining cluster of paratypes; I don't suppose they may actually have included both Peckoltia sp. L80 and this Hypancistrus in their material without noticing the difference? Surely that can't be?

-Haakon

[bekateen]

Just a thought on how you find two specimens being different from the remaining cluster of paratypes; I don't suppose they may actually have included both Peckoltia sp. L80 and this Hypancistrus in their material without noticing the difference? Surely that can't be?

-Haakon

They provided another supplemental document with DNA data. It appears they tested only four specimens for DNA for one gene and two specimens for another. According to the paper, DNA testing was done on tissue samples saved at the museum, and the specimen codes provided don't match the codes for the holotype or paratypes... which in my mind indicates that these were obtained from fish that are not in the type series.

Knowing all that about the DNA testing, it would not exclude the possibility that at least one L080 accidentally was included in their morphometrics of paratypes. That said, with the dentition being so weird, if they looked at teeth in every specimen (and I belive they did based on the supplemental data chart), I would hope they confirmed all had the same teeth. But no, it appears that two don't have any tooth counts provided and three others have only partial data. This might be explained by just some specimens having damaged teeth, but it does open the door for error.

Cheers, Eric

[Jools]

Good discussion.

Before I read it, or the paper (thanks Eric), I went to sexual dimorphism. Those around when

L080

was introduced will remember it being imported "hairy" even at quite small size. Indeed, Ingo's original photos of it here show that very well. It was one of the first L-numbers bred in captivity (C. Seidel, 1994) and had a big hand in the "you need a cave" revolution from our German friends that launched the l-number fishkeeping tribe. It was very well known that males developed long back body odontodes.

Now I've read the paper, the holotype is clearly a male and clearly not "hairy". The live photograph also so. Colouration aside, the live photo most reminded me of

L260

.

Another thought, if they are cave dwellers, it is quite possible that mostly males (even of both spp.) formed the sample because they would be much easier to collect. Males stay in caves and can be teased out with a chopstick, females scatter into the surrounding area. This makes more sense too when we consider the authors write, "moderate water flow in a bedrock background".

If the authors had recorded collection method, it would have helped understanding that.

Does anyone have more up-to-date collection data on L080? Maybe @LeandroSousa? Also, how does one pronounce the new name? "Park at EDGY"?

L080 not being mentioned is interesting and given the lack of advanced fishkeepeing references, it's hard to say what consideration the authors had of this species. I am guessing that L080 and the new sp. would not live together with perhaps the Peckoltia preferring a less barren biotope.

So, it's another brilliant mystery we didn't know we had. Seems L080 is not H. parkateje and we all will be looking a bit closer at these fishes.

Cheers,

Jools

PS @Yann and @jjphoto.dk 's pictures currently in L080 may be the new species. Not very sure. Opinions welcome!

[bekateen]

I contacted the lead author. He contributed the photo we have and he explained why the team didn't discuss sexual dimorphism. I'll have to add that later, as I'm traveling now. He also sent me the complete data table (with SL values) for the paratypes, so later I'll analyze that and see what pops out.

Cheers, Eric

[bekateen]

If the authors had recorded collection method, it would have helped understanding that.

From the paper, the authors wrote this:

Morphometric data from type material were taken from recently collected specimens by the authors that are deposited in the fish collection of Museu Paraense Emilio Goeldi (MPEG).

I interpreted that the specimens were collected by the authors, since they also have photos of live specimens. But it depends on how you interpret the grammar: did they collect the specimens and perform morphometric measurements, or did they only perform the morphometric measurements using specimens already in the museum?
Cheers,
Eric

[bekateen]

Quick update, the outlier with the (relatively) long narrow head was one of the smallest fish captured. Graph to follow.

[bekateen]

Before getting into the graphs, here's what the lead author of the paper told me about why they did not list sexual dimorphism in the paper. Basically, he said that since the researchers didn't open the abdomens of the fish to inspect the gonads (for testes vs ovaries), they didn't want to infer the sexes of the fish based on external traits. He said he hopes they can get back to that later.

Okay, now onto the updated data sheet and revised graphs... I reimagined the graphs a little:

1. CW/HL vs SL (a new analysis since I didn't have SL data before)
2. CW/HL vs HL (this is the same analysis as before)
3. # Cheek odontodes vs SL (also new data not clearly available in publicly posted Excel file)
4. Relative lengths of pectoral and pelvic fin spines. (I wondered if there might be some dimorphism emerging with size)

Interpretations: As noted earlier, the holotype male, very "macho" with the widest head, was also one of the largest specimens. Also now clear, the individual with the narrowest head is one of the smallest individuals (but not the absolute smallest). Also, I see that the holotype male has the most odontodes of any fish examined, and the narrowest-headed fish has among the fewest odontodes (but not the fewest). There was one fish, a 45mm SL fish, for which only a (-) was recorded for cheek odontodes. So that fish does not appear on the third graph. Finally, although weak, there is a general correlation between SL and number of cheek odontodes, but really no correlation between the coracoid width/head length ratio and SL (if you exclude the two outliers, graph 1).

The graph of fin spine lengths revealed relatively little change as the fish grow and that there might be more homogeneity at larger sizes than there is at smaller sizes. My thoughts are these:

• First, the slightly higher variation at smaller SL values might just be due to greater chance for minor errors when values are smaller, so that means nothing to me.
• Second, the overall stability could mean that there is no sexual dimorphism in fin spine lengths in this species (I know that in some pleco species, either pectoral or pelvic fins might be longer in males than in females, but I had no preconceived notion for this Hypan).
• However, if there is any sexual dimorphism in fin spine lengths in this species, then we would expect the values to diverge as the males and females matured and we don't see any divergence here. So if there is measurable sexual dimorphism in fin spine lengths, that argues for the type series as being all the same sex, at least at the larger sizes.

Food for thought. Cheers, Eric

Cheers,
Eric