PlanetCatfish.com

MatsP wrote:But Hypancistrus from Rio Xingu are definitely closely related to Rio Orinoco ones, as they can hybridise with fertile off-spring. If these fish were not relatively close related in an evolutionary sense.

I'm not saying there aren't possibly some "odd ones" in the group, but I'm not convinced that the argument that "they come from far apart, so can't be the same genus".

Ok, two queries then.

Firstly, I am note sure that's Shane's argument. Let's call them the Hypancistrus zebra group and the Hypancistrus contradens group. Have we seen those group freely hybridise?

Secondly, where have we seen lower Amazon tributary and Orinoco drainage hybrids?

For both of the above, apologies if I've missed a post or some other bit of info on this - just don't recall it.

Jools

Matt30 wrote:That's top of my list absolutely stunning
I bet they will be priced at Hypancistrus Zebar
prices though.

Not in the same league, these are going for about £425 retail. How many among us then would pay for a group of 6? How many importers would handle a box? I think that answers a few questions about which market they end up at.

Jools

The "hybrids" thread at the L-number days in 2009 had L129 x L28 which is Rio do Para x Rio Orinoco drainage...

See the last dozen or so pictures here:
http://www.aquazone.gr/forums/index.php?showtopic=33730

I've added H. debilittera (L129) type location to the above map.

--
Mats

I must say this is a great thread! Very informative!

@Mats

Yes, that answers the "in-group" but not the point, that I think also Shane was making re H zebra x H contravene groups?

Jools

So, what you are saying is that H. zebra wouldn't cross with one of the Orinoco spotted species?

But we have a picture of L66 x L136 in the Cat-eLog. And I'm pretty certain I've seen pictrues of L66 x H. zebra (by a Norwegian breeder?) - so if L66 can breed with both L136 (which is a Rio Negro spotted species) and H. zebra, I don't see why H. zebra should be much different.

--
Mats

But Hypancistrus from Rio Xingu are definitely closely related to Rio Orinoco ones, as they can hybridise with fertile off-spring. If these fish were not relatively close related in an evolutionary sense.

Mats, the map is cool but looks to include far more than the described spp in the genus. "L Numbers" and how they may or may not spawn between each other in artificial conditions (captivity) is not really a scientific argument that these fishes are closely related. There is every possibility that some, maybe many, of these L Numbers will not turn out to represent valid spp. and that some will be placed in genera other than Hypancistrus.

Interspecific hybrids will almost always have infertile offspring, so if the H. zebra complex (Xingu) and H. contradens complex (Orinoco) are in fact capable of hybridizing it would be of little concern to hobbyists as their fry could not propagate (they would be "mules"). If a cross between the two groups does produce fertile offspring, it actually would support the argument that the two populations are not congeners. So if your information above is correct that crosses between the two populations have produced fertile offspring, this is the nail in the coffin of the argument that they belong to the same genus.

-Shane

When populations of known Hypancistrus are added to the map (e.g. Lower Negro, Rio Demini), I fail to see the disjunction between these distributions. Subject to sampling biases, it appears to me these fishes have effectively dispersed, colonised and speciated in most available habitats (deep, rocky, clear/blackwater rivers) of the Amazon/Orinoco system. The reason they are not found in the western Amazon, is likely due to lack of suitable habitat, as rivers in this area tend to be more sediment rich.

However, Shane's assertion about taxonomic status of these undescribed forms holds true, but the monophyly of the genus can only be answered by looking at the anatomy or DNA.

All this is supposition though, really.

Interestingly, Armbruster (2004) does point towards a polyphyletic Hypancistrus, but more taxa and data are definitely required.

Hi Shane,

the "problem" with Hypancistrus is, that it appears to be possible to produce virtually any cross between those forms. And (to my knowledge) none of them turned out to be infertile (which, admittedly, has not been checked thoroughly in most cases).
The best documented cases are from Rolo, who gave the talk at the L-Numbers Days.
He "demonstrated" that L 28 and L 129 can interbreed, as do L 262 and L 270 (and a lot of others, such as L 46 and L 66).
However, he also showed that the offspring of the L 28xL 129 cross can interbreed with the L 262xL 270 hybrids. So, if I got your argument correctly, this cross should not exist?

So basically, while only little documentend information is publicly available, there are strong indications that any form of the Hypancistrus complex can interbreed with any other, and additionally, there is no hint that these offspring might be infertile.

Cheers, Sandor

Ability to interbreed or produce fertile offspring has nothing to do with taxonomic rank.

I would imagine that the two are in general, significantly inversely correlated, however.

Yes, the link to HuskyJim's photos clearly show off-spring of hybrids [at least if we believe the photos].

I understand that hybridization capability is not a scientific sign of close relation. And also that L-numbers are not always valid species. I also admitted that I didn't put markers for every "species". But to completely ignore fish are at least on superficial inspection appear to be in the Hypancistrus genus simply because they haven't been described and aren't from the known distribution of currently known Hypancistrus is probably equally wrong.

I agree that sometimes the genus given to an L-number is incorrect - "blue panaque" turns out to be rather than a Panaque species. But most of the time, I'd say it's not that far off the mark.

--
Mats

When populations of known Hypancistrus are added to the map (e.g. Lower Negro, Rio Demini), I fail to see the disjunction between these distributions. Subject to sampling biases, it appears to me these fishes have effectively dispersed, colonised and speciated in most available habitats (deep, rocky, clear/blackwater rivers) of the Amazon/Orinoco system.

Even taking into account some sampling bias (which I do not think there is a lot of given the commercial collection of these fish) the described spp, with the possible exception of H. inspector, all have very limited natural ranges and have not effectively dispersed. You can for example collect Panaque maccus in any one of 100s of Orinoco tributaries, but the Orinoco Hypancistrus only occupy a very small, restricted area. If they existed outside those areas commercial collectors would have found those populations. I understand that it is the same in the Xingu. I wish I could do a map like Mats did, but consider this:

Orinoco complex
H. contradens Rio Ventuari, 94 km E San Fernando de Atabapo
H. debilittera 38 kilometers north of San Fernando de Atabapo
H. furunculus 60 kilometers east of San Juan de Atabapo
H. lunaorum Río Guapuchi, first major riffle complex, Río Ventuari drainage

Negro
Hypancistrus inspector Rio Negro (Barcelos)

Xingo
H. zebra Xingu, near Altamira

The Xingu population of Hypancistrus is separated from the others by 1500 miles as the crow flies and effectively 3000 plus miles by the way rivers flow. The two populations are also separated by a massive natural barrier (the main Amazon river channel). I would call this a disjunction.

I agree that sometimes the genus given to an L-number is incorrect - "blue panaque" turns out to be Baryancistrus beggini rather than a Panaque species. But most of the time, I'd say it's not that far off the mark.

Ah, like when it was L 46 Hypostomus "zebra" in the first L Number books, or when it was L 46 Peckoltia sp? I am looking at Glasser's (1998) book right now and see: LDA 25 Hypostomus (actually Parotocinclus), L 124 Ancistrus, also listed in another L Number book as Hypostomus (actually Peckoltia), L 168 Lasiancistrus (Dekeyseria), L 2 Peckoltia (Panaque), I could go on and on. Look back at some of the L Number books from the mid to late 90s and you'll see what I am saying, they were off at least as often than not and I suspect they still are.

All this is supposition though, really.

That may be one of the few facts in this thread.

-Shane

Hi Shane and racoll,
So you guys do not believe that the capability of two or more different species to produce fertile hybrids has no bearing on a very close taxonomic relationship? I thought this kind of ability was evidence of a close relationship and recent divergence into various species within a genus.
I have followed this sort of discussion most closely with regard to the 3 species of Symphysodon, the Discus fishes. One major difference with Symphysodon from Hypancistrus is that they are all only found in different parts of the Amazon River system. Discus are probably quite a bit younger than even Hypancistrus, if we agree that the described species found in either the Amazon or Orinoco Basins are truly in the same genus, due to geological barriers that now exist between these two river basins. No natural Discus populations have ever been found in the Orinoco Basin.

I also know that the relationships between different fishes do not follow any sort of rigid classification based on whether or not different populations can interbreed and produce viable offspring. There are many examples of what are considered to be the same species of African Killifish from different locations having problems producing fertile offspring. And many species can produce infertile hybrids. Killies are probably more primitive and are quick to begin to differentiate when populations become isolated from the main range of a species.

apistomaster wrote:Hi Shane and racoll,
So you guys do not believe that the capability of two or more different species to produce fertile hybrids has no bearing on a very close taxonomic relationship?

Since it is so hard to prove/disprove in a lab setup, it breeding information is relatively rare in scientific literature in general - most fish used in science aren't alive anywehre even remotely long enough to breed - generally they are killed within a few minutes of comming out of the water.

Disregarding the "most fish are dead" problem, how long would you keep Hypancistrus sp A with Hypancistrus sp B to be certain that they do/don't breed? And if they haven't bread after four years, is it because they can't, or because you haven't provided the right combination of conditions. May not be too difficult with Hypancistrus, but I'm sure just like everyone here, you know of people who (or you yourself) have kept and tried to breed some fish for dozens of years [assuming they live that long] and not succeeded in breeding them. Many aquarium fish are NOT being bred even in commercial breeding, because they are just not easy to breed.

What about plants that only flower one night every 100th year and another, closely related that flowers every 107th year? You'd be waiting many hundreds of years to check it out...

So breeding together is a pretty poor way to deduce what is/isn't a species, genus, etc. And it certainly isn't easy with dead fish ;)

I thought this kind of ability was evidence of a close relationship and recent divergence into various species within a genus.
I have followed this sort of discussion most closely with regard to the 3 species of Symphysodon, the Discus fishes. One major difference with Symphysodon from Hypancistrus is that they are all only found in different parts of the Amazon River system. Discus are probably quite a bit younger than even Hypancistrus, if we agree that the described species found in either the Amazon or Orinoco Basins are truly in the same genus, due to geological barriers that now exist between these two river basins. No natural Discus populations have ever been found in the Orinoco Basin.

Clearly the (lack of) ability to breed together is one indication of relatioship. But scientists prefer to use other methods to determine that. One way is to use DNA, another is to look at similarities and differences in morphology.

Of course, inability to breed may not be genetic - it could be physiological, visual or hormonal as well. If a guppy male's (or a Tatia) gonopodium isn't fitting in the females hole, it's not going to be any babies. If the male doesn't "look right" for the female (or other way around), we probably won't get any breeding [guppies clearly don't bother with this!]. If the male's or female's feromones aren't attractive to the opposite part, again, we're not going to get to the stage of getting any breeding. So it may not show us much of an indication on how closely related they are from that - it only takes one small change in a tiny portion of the DNA to alter feromones, pattern or physiology in one place.

I also know that the relationships between different fishes do not follow any sort of rigid classification based on whether or not different populations can interbreed and produce viable offspring. There are many examples of what are considered to be the same species of African Killifish from different locations having problems producing fertile offspring. And many species can produce infertile hybrids. Killies are probably more primitive and are quick to begin to differentiate when populations become isolated from the main range of a species.

So clearly, scientist don't use DNA to tell these apart [or can't tell from the DNA that they are too far apart - it may not be that easy to tell!]

--
Mats

There is also the argument that fishes that may encounter each other in the wild "learn" not to interbreed. Our very limited range (hey, another fact!) Hypancistrus don't encounter each other in the wild (except perhaps when washed up in the lower Xingu) and so haven't had to "fix that problem".

Now, the other thing that bothers me about Hypancistrus is the fact they are deeper water small fishes. It's my gut feeling (that's the end of the facts" that there are as many of them out there as Corydoras (like 150 species) although I fear I will not live to see it. And, as with that genus, they're due for a split (e.g. what Shane is suggesting). Now, in the 1960's if you'd suggested that about Corys, a lot of people would have called you an idiot, but here we are with 140 odd described species, 150+ c numbers and 40 odd CW numbers.

ANYWAY, nice to have another pretty Hypancistrus. If we follow H. zebra, it should be under £100 in about 8 years time.

Jools

apistomaster wrote:Hi Shane and racoll,
So you guys do not believe that the capability of two or more different species to produce fertile hybrids has no bearing on a very close taxonomic relationship?

On the scale of different genera vs. different species vs. different populations, it really isn't very effective. Some animals of different genera can interbreed, while some animals of the same species, as you mention, can't interbreed. This is partially due to some biological effects as mentioned by Mats, but it's also partially due to the arbitrary "distances" that we ascribe to genera.

So, basically, ability to hybridise has no bearing on generic placement?

Jools

When molecular biology was just barely being conceived, Jorgen Scheel wrote his classic book, Rivulins of the Old World, published in 1968. Scheel cross bred nearly 100 species of African Killiefish. He ended up with hybrids with intermediate chromosomal counts between the different species used and showed hundreds photos and diagrams of the chromosomes, the parent fish and the hybrids.
I am suddenly very distressed to learn I do not know where my original copy of the book is. I know I have had it since 1968.

I do not expect anyone would attempt anything like this work with fish like Hypancistrus. Like has happened to these species of old world Killies, the Hypancistrus are likely be found to belong to different and often newly named genera. Scheel anticipated this and instead of using the contemporary scientific names he decided to use 3 letter codes to identify species. For example, where in 1968 Aphyosemion gardneri was referred to as GAR. Now it is called Fundulopanchax gardneri which actually turned out to be a complex of several species and subspecies per current names in good standing. His coding identification was not all that different than the contrived L-number system. It was just a way to use a descriptive term for the sake of providing a common reference. But the point is, Scheel did a comprehensive range of experimental hybrids so anyone who has experimented with Hypancistrus hybrids whether on purpose or by accident is not doing something unique in the history of fish taxonomy in examining whether different looking species can hybridize and whether or not these hybrids are fertile or even viable. Many killiefish hybrids could not be raised to adults and Scheel inferred that these were probably too distantly related to produce a fish which could even survive to reach sexual maturity.

Dr Joanne Norton and others did similar work with fishes in the genus Xiphophorus, the Platies and Sword Tails and many of the hybrids are now popular aquarium fishes. Others pursued the same line of hybrid research with various species of the genus Poecilia, Guppies have been crossed with Mollies.
Not much come of the latter except an interesting veil tail molly X Guppy hybrid which never was viable enough to become an established aquarium strain.
Many popular aquarium strains of Mollies are the results of early hybridization experiments.
Dr Eduard Schmidt-Focke experimented with hybridization of of the 3 species of Symphysodon and laid the ground work that led to the development of domestic hybrid fancy Discus around which an entire branch of the aquarium hobby is based and an entire industry of commercial farming of expensive Discus.

So there is an extensive amount of research using hybrids to establish relationships between different species before molecular biology could contribute much to our understanding. I think the influence of these peoples' work is being over looked in this discussion as if hybrids were irrelevant.
Clearly they have had considerable influence on both the science and hobby of fish keeping.
No one would approach sorting out the phylogeny of the Hypancistrus this way now that we have more sophisticated methods available but it is scientifically interesting whenever we learn which species can interbreed. It does tell us something even if it is of little immediate practical value.
Pure research often has little or no immediate relevance. That is what makes it research. Eventually little bits of information ultimately contributes to a better understanding of a larger question. In examples cited above there were many fish which became commercially important to the aquarium hobby. It would be ignoring history if we dismiss the likely possibility that some Hypancistrus-like fish hybrids do not ultimately become a popular aquarium fish.

I am not promoting the hybridization of any plecos. I do think there are breeders somewhere in the world who will pursue them and create fishes we can't even begin to imagine.
I am against the distribution of hybrids as much as anyone. Just because most of us do not want them does not mean enough people will find them interesting to see where hybrids may take them in their commercial exploitation. Might be wise to keep track of and understand these developments which are sure to come.
Scheel never distributed any hybrid Killiefish he produced however beautiful a few types were. Pioneering fish geneticists' experiments ultimately did enhance the science and the hobby of keeping tropical fish.

Apistomaster wrote:So you guys do not believe that the capability of two or more different species to produce fertile hybrids has no bearing on a very close taxonomic relationship?

Larry, as you say, ability to interbreed and produce fertile offspring can and certainly has been used as a taxonomic character*, however, turning this around and extrapolating ability to interbreed from taxonomic rank will not work. Generally yes, but specifically no.

Shane wrote: That may be one of the few facts in this thread.

Haha, yes. But an interesting discussion regardless. What riles me slightly though, is when people read these ideas/opinions, and repeat them as fact.

* unfortunately it's only of limited use, because as well as for the reasons Mats explained, for allopatric species this ability occurs quite late in the speciation process - often much later than diagnostic morphological traits.

Rupert,
I think you know me well enough to know I understand and I appreciate the subtleties of these things.
I qualify most everything I write. I know enough to know how little I really know. I always learn new things or gain from others' perspectives in these discussions.
It is fun to speculate on how fish become species and how the species are related. Partly mental time machine trips and partly imagining the geological changes that allow fish to end up where they are by stream capture, stream isolation and the other mechanisms that drive distribution and evolution.

Apistomaster wrote:So you guys do not believe that the capability of two or more different species to produce fertile hybrids has no bearing on a very close taxonomic relationship?

I'm Scottish, and thus not great at subtle. The word you missed was useful, e.g. "no [useful] bearing".

Jools

Shane wrote:Ah, like when it was L 46 Hypostomus "zebra" in the first L Number books, or when it was L 46 Peckoltia sp? I am looking at Glasser's (1998) book right now and see: LDA 25 Hypostomus (actually Parotocinclus), L 124 Ancistrus, also listed in another L Number book as Hypostomus (actually Peckoltia), L 168 Lasiancistrus (Dekeyseria), L 2 Peckoltia (Panaque), I could go on and on. Look back at some of the L Number books from the mid to late 90s and you'll see what I am saying, they were off at least as often than not and I suspect they still are.

Sure, but this isn't the whole picture. That's to say, it's not l-number books, it's literature. The Burgess Atlas, going further back, shows all these failings too. Sands atlases are also so. And the scientific literature, if it were not so we would not need synonyms. It is the nature of this that changes will happen as more things are known.

IMHO, the point is accuracy versus time, if something if way off when it is published then it deserves a pounding, however, otherwise, not.

Jools

Well, Jools,
It is always going to be hard to know what information is useful since biology is the result of both chance and necessity.

racoll wrote:What riles me slightly though, is when people read these ideas/opinions, and repeat them as fact.

Indeed, certainty walks with ignorance. Just don't know how to stop it though. Or if it's worse just not to discuss because this happens. It's this single thing that may kill some forums.

Jools

update some pictures

Hi!!

to me they look like beautiful patterned Hypancistrus sp L340



They do not look a lot more different than thoses
Cheers
Yann

Yann wrote:to me they look like beautiful patterned Hypancistrus sp L340

Exactly! I think, it is not more than just a very pretty variety of L 340.

Hi,

compared to this L 340, Hypancistrus sp. "Platinum" is really "a lot more".
But perhaps, the "Platinum" variation leads to the origin of LDA 19, as well known as "Mega Clown".

Recently we had a shipment from Colombia and we picked out 1 exceptionally looking piece from the lot and a few others that look somewhat similar. The rest were just plain Jane. So wondering if it's the same "Corroncho Plata" that were discussed here....as the shipment did originate from Colombia.

It seems more like a L129 to me.