cross breeding

[MatsP]

It's a bit like "All cats have a tail." "Dogs have a tail, therefore dogs are cats".

You can clearly say that because two (proposed) different species CAN NOT produce fertile offspring, they are NOT the same species. However, if they are able to produce fertile offspring does not in itself make them the same species. Surely, no one is suggesting that

Hypancistrus zebra

and

Hypancistrus sp(L260)

are the same species?

[And yes, I know, given that, using modern methods, we probably would arrive at the conclusion that not all dogs are the same species...]

--
Mats

[Shane]

i have been reading through everyones post and this actually seems interesting but if 2 different species had the same DNA to make a crossbred child, wouldnt that mean at 1 point these fish MAY have been the same species???? because they have the same DNA

This is a good question. The 1970s era science most of us learned from (which was probably from school textbooks that had been around since the 1950s) pretty much taught us all that, by definition, a species was valid and separate from another if the two organisms could not produce fertile offspring. Animal and plant biologists have strayed from this definition, in varying distances among various life science specialties, over the last 50 plus years.

Ironically, the only group of living beings on the planet we still apply this definition to is ourselves (aren't we special). Be your ancestors short, gracile San bushman from South Africa superbly adapted for life in the Kalahari, one of the central African tribes whose red blood cells have evolved clear down at the molecular level to make you genetically immune to malaria, or two meter tall Nordics whose skin and hair have lost almost all pigmentation allowing their bodies to utilize the very limited available sunlight in the far north to process vitamin D to keep them alive... you my friend are Homo sapiens because all matings between all humans on the planet produce viable offspring. If we applied this definition to fishes, for example, there would only be a Baensch Atlas I and we would all save some money. We are able to use a different definition for the rest of life because making an argument that Panaque maccus is "better" than P. cochliodon is just plain silly, while making the argument that race X is "better" than race Y has pretty much been the history of organized human violence.

One recent study I read pointed out that upon completing DNA testing of 14 species of Lake Victorian c i c h l i d from 9 different genera (different genera!) there was less genetic difference between them than between a typical population of humans. That pretty much means that a very small town in Iowa has more genetic diversity among its population of H. sapiens than several hundred c i c h l d s described as distinct species and belonging to over a dozen genera.

The important thing to remember, from a hobbyist's point of view, is that there is currently no universally agreed on definition of a "species" and that the definitions out there range far and wide. In fact several scientists are now arguing that it is time to just throw out the entire concept of "species" (won't that make Jool's and Mats' managing of the Cat-eLog less time consuming!) as it is based on faulty underpinnings. The species concept came about at a time when scientists believed, as the religious texts stated, that every organism was present in its final form as made by the Creator. All we had to do was find them all (as Noah did) and catalog them... that is why we are able to choose a single specimen from a population, called a holotype, to represent an entire species. Linnaeus wrote Systema Naturae in 1735 and gave us Taxonomy 135 years before Darwin wrote On the Origin of Spp and gave us evolution. Then we realized that evolution was in fact an ongoing process. However, we were still ok as we adopted the idea that evolutionary change takes place over very, very long periods of time.

Now we know this is not true at all. 12,000 years ago Lake Victoria was dry. Those 500 species of c i c h l i d s did not come about after millions of years of evolution, but rather after a few thousand. To bring the point home to H. sapiens; a recent study of black Americans showed that only 8% still retain the genetic immunity to malaria. Once the evolutionary pressure for immunity to malaria was removed the population dropped the adaptation not in 1,000s of years... but in a dozen or so generations.

So, the argument goes, it makes no sense to attach a static "species" name to what we know are constantly evolving gene pools. In fact, a species described in 1809 may not even be valid today if it has changed sufficiently in the last 200 years.

-Shane

[Janne]

Shane,
Excellent explanation

The new generations of scientist will have lots of work ;)

Janne

[racoll]

An eloquent and interesting synopsis Shane, although I must contest a few points.

f we applied this definition to fishes, for example, there would only be a Baensch Atlas I

Variation in humans is generally strongly clinal along environmental gradients. The gradation of phenotypes is consistent with the recent population expansion. Insufficient time has elapsed, and too much gene flow is taking place for differentiation and assortative mating to occur. Taxonomists will look for discontinuities and unique characters, rather than recognise these clines.

One recent study I read pointed out that upon completing DNA testing of 14 species of Lake Victorian c i c h l i d from 9 different genera there was less genetic difference between them than between a typical population of humans

I don't know which study, but this variation was almost certainly at a very limited selection of loci. Only a genome wide survey could ascertain whether how truly different they are. Different forces act on different parts of the genome.

The important thing to remember, from a hobbyist's point of view, is that there is currently no universally agreed on definition of a "species" and that the definitions out there range far and wide.

So, the argument goes, it makes no sense to attach a static "species" name to what we know are constantly evolving gene pools. In fact, a species described in 1809 may not even be valid today if it has changed sufficiently in the last 200 years.

While it is certainly not always black and white, most taxonomists and evolutionary biologists are happy with the phylogenetic species concept, and it is the only concept implied or stated in the overwhelming majority of species descriptions, even before the advent of cladistic theory.

Marked discontinuities can be observed between most species, and populations of interbreeding or recently interbreeding organisms with diagnosable synapomorphic characters can, by and large, be called species. This is not theory, but observation and pattern from nature. Empirical data has shown the jump in branching rate at the boundary between speciation and gene coalescence. Most species conform to this model, but groups undergoing recent expansions such as discus, humans, Victoria cichlids or salmonids are going to cause problems when we need to recognise independent evolutionary units.

I do not believe species are some magical entity, but I do believe the concept of species is a workable and defensible one.

[apistomaster]

I am much more agreement with racoll on this one especially since he was so kind enough to include some of the fish I care a lot about; the Symphysodon species and Salmonids which happen to share many of the same fuzzy lines that the Hypancistrus spp present and cause endless confusion.

Hey racoll,
I am getting a dozen F1 Nhamunda Blue Discus from a beautiful pair of fish very soon.
I do plan to keep the line pure but I should also get a couple of spare females I can breed with one of my favorite male wild Alenquer and then cross one female with a Stendker Brilliant Turquoise. I think the latter will produce a reasonable facsimile of a wild royal blue that my typical customers will like. The pure wild lines are appreciated more by only a few Discus purists like myself. It takes at least a year for a wild type Discus to show it's full potential colors and most folks want their Discus colorful from a very small size and lack the patience for the wild types to finally show their true colors. One other friend's wild blue pair has been spawning but he has been in a slump. I think I have begun to get him thinking about raising a brood thus adding to the diversity of F1 wild Discus I plan to keep going. There has been a recent trend towards a growing interest in keeping wild Discus but having tank raised wild types will make breeding them easier and they will be free of the parasites common to wilds.
Next step: Getting a line of well marked F1 Red Spotted Green Discus established.
These Discus projects will take a few years to establish well.

[Janne]

I do not believe species are some magical entity, but I do believe the concept of species is a workable and defensible one.

It has a time limit, when enough time of evulotion have past it don't fit ?... it's either a new species or one old species and one new species or maybe even more species. When 2 species interbreed in nature and these offspring etablish a new population somewhere near or far away... is that a new species or is all 3 the same species? There are not much study when it comes to Loricariidae but for corals that also is an animal are much more recently studies made... and they interbreed every year in some places. Are humans more intelligent than mother nature or is't just our need to keep statistics on everything ? ;) I'm asking these questions of curiosity and because it's something I don't have so much knowledge about.

Janne

[Mike_Noren]

Ironically, the only group of living beings on the planet we still apply this definition to is ourselves (aren't we special).

Sorry, that is not true. First of all, the humans today are not considered species on the grounds that they are not diagnosable; there is no single character or combination of characters which would make it possible to distinguish species among humans. Every single character grades between races to the point it is even impossible to say how many races there are - 4 or 400.
The exception is fossil humans. Every single fossil of hominids tend to get described as a new species, so that there have been something like 15 described species of human alive during the last 2-3 million years. And in that context one should know that there exist no proof of any speciation event after Homo erectus (who grades imperceptibly into H. sapiens, ie H. erectus is a name given not to a separate species, but to a portion in time of the gene pool of humans, a "chronospecies"). Human biology as a whole is a cesspool of confusion and bullshit. The extremely confused debate around neanderthals and 'flores hobbits' depress and annoy me unspeakably, because no one pretends to know that paleontological species concepts are very different from biological ones, and everyone pretends that you can tell if two groups of animals were different species or not based on that one had a slightly larger bump on a bone in their foot, or were 0.1% different from modern humans in a mitochondrial gene.

One recent study I read pointed out that upon completing DNA testing of 14 species of Lake Victorian c i c h l i d from 9 different genera (different genera!) there was less genetic difference between them than between a typical population of humans.

Improbable, but depends on what genes were compared. For most genes both all humans and all victoria cichlids are identical or very nearly so. We're still talking animals with distinct ecologies and with highly distinctive pigmentation due to female choice - ie, they don't, or rarely, hybridize in nature, and any person would say they look different. Genetic distance means little to nothing.

Genera, families etc higher ranks have no intrinsic evolutionary meaning; they're simply "boxes" we use to partition the diversity of species to make it manageable. They're practical but ultimately mean nothing.

In fact several scientists are now arguing that it is time to just throw out the entire concept of "species" (won't that make Jool's and Mats' managing of the Cat-eLog less time consuming!) as it is based on faulty underpinnings.

Imperfect underpinnings, perhaps... I think you're talking about typology, the concept that all members of a species were imperfect representations of the "essence" of their species. We now know that reality is messy and that it is hard to determine the exact time of speciation, where the lineages branch, but abandoning species both flies in the face of evidence (it's obvious and self-evident that an oak tree is a different class of organism than a mallard duck) and would make studying evolution very difficult indeed.

So, the argument goes, it makes no sense to attach a static "species" name to what we know are constantly evolving gene pools. In fact, a species described in 1809 may not even be valid today if it has changed sufficiently in the last 200 years.

I don't understand this point. The species is valid into infinity or until it branches into two new species, whichever comes first.

[Janne]

Sorry, that is not true. First of all, the humans today are not considered species on the grounds that they are not diagnosable; there is no single character or combination of characters which would make it possible to distinguish species among humans. Every single character grades between races to the point it is even impossible to say how many races there are - 4 or 400.

Does this means that there is one species of humans and 4-400 subspecies of humans? Whats the different between races and subspecies?

The species is valid into infinity or until it branches into two new species, whichever comes first.

Yes, at this moment and in the future when it no longer exist it was a species in the past, but the evulotion don't ever stop; so at which point in the evulotion can we (humans) decide this is the same species even that they look different?

Janne

[Suckermouth]

I do not believe species are some magical entity, but I do believe the concept of species is a workable and defensible one.

It has a time limit, when enough time of evulotion have past it don't fit ?... it's either a new species or one old species and one new species or maybe even more species. When 2 species interbreed in nature and these offspring etablish a new population somewhere near or far away... is that a new species or is all 3 the same species? There are not much study when it comes to Loricariidae but for corals that also is an animal are much more recently studies made... and they interbreed every year in some places. Are humans more intelligent than mother nature or is't just our need to keep statistics on everything ? ;) I'm asking these questions of curiosity and because it's something I don't have so much knowledge about.

Janne

I don't know of any situations where when two species interbreed, and their offspring found a new population far away. What is more common is that the two species have relatively extensive ranges and they have a zone right between where they hybridize; AFAIK, this is the general thing that happens with coral reef fishes that hybridize. This is not a small problem, and the problem is highlighted by so-called "ring species." In terms of your question, it could be either all one species, or two species. Hybridization never results in speciation, so it will never be three species.

Sorry, that is not true. First of all, the humans today are not considered species on the grounds that they are not diagnosable; there is no single character or combination of characters which would make it possible to distinguish species among humans. Every single character grades between races to the point it is even impossible to say how many races there are - 4 or 400.

Does this means that there is one species of humans and 4-400 subspecies of humans? Whats the different between races and subspecies?

Janne

You may be missing his point. His (and racoll's) point is that it is difficult to draw any lines between groups of humans and that it all blends together, and that variation is all in the form of gradients. If you choose an arbitrary definition you might get a few groups of humans, and if you choose another you might have to split humans excessively to get monophyletic groupings. There's no point in breaking up the humans into subspecies because each one would be minutely different from the next. A similar problem exists in splitting some large genera into meaningful genera.

The difference between "race" and "subspecies" is probably irrelevant since we don't use both on any one species, but if we did, the difference between races and subspecies would likely be the same as the difference between a common name and a scientific name for species. Sometimes they correspond exactly, and sometimes they don't.

Here's a piece from Wikipedia on race:

Some argue that the taxonomic concept of race, although valid in regards to other species, does not (currently) apply to humans. Many scientists have pointed out that traditional definitions of race are imprecise, arbitrary, have many exceptions, have many gradations, and that the numbers of races delineated vary according to the culture making the racial distinctions. Thus, those rejecting the notion of race typically do so on the grounds that such definitions and the categorizations which follow from them are contradicted by the results of genetic research.

Today many scientists study human genotypic and phenotypic variation using concepts such as "population" and "clinal gradation". Large parts of the academic community take the position that, while racial categories may be marked by sets of common phenotypic or genotypic traits, the popular idea of "race" is a social construct without base in scientific fact.

[Janne]

This is not a small problem, and the problem is highlighted by so-called "ring species." In terms of your question, it could be either all one species, or two species. Hybridization never results in speciation, so it will never be three species.

Ok, when this happen what makes them to 2 species, if they succeed to etablish a new stable population they will be 2 species, and between that they are temporarily the same and 1 species?

You may be missing his point. His (and racoll's) point is that it is difficult to draw any lines between groups of humans and that it all blends together, and that variation is all in the form of gradients.

So this problem is the same for all other animals too, there are small variation within the same species but not enough to separate them as their own species, colouration, habitat adaptation etc. doesn't matter. Does this also meen that all existing subspecies today will be synonyms and become one species in the future?

Janne

[apistomaster]

I don't see a once size fits all solution to species differentiation.
There are many diverging populations of a single or superspecies that don't quite warrant their own species but need differentiation because they have begun to diverge too much for a satisfactory inclusive species delineation without further qualifiers such as subspecies or local races.
And so many aquarium fish appear to be in a transition where making a clear distinction isn't possible.

[racoll]

Whats the different between races and subspecies?

Subspecies do confuse matters, unnecessarily I think

I refer back to my post a few pages back.

Subspecies were created to describe intraspecific variation under the biological species concept, where species were proposed to be entities that are unable to produce viable offspring.

Now, I don't like the biological species concept one bit, as it recognises diversity far too late in the process of divergence, ignoring evolutionary independent lineages which are clearly different, but may well hybridise. The subspecies concept was created to recognise this problem for those entities within a biological species that are morphologically different, but would be able to breed together and produce offspring given the opportunity.

In reality I would guess that only a handful of creatures have ever been described on the basis of breeding experiments, and these were certainly carried out in captivity, which does not reflect natural conditions. This to me makes the biological species concept irrelevant, with essentially untestable hypotheses.

So, what has happened , especially in birds and insects, is the species rank used when "large" differences are clear, while the subspecies rank being ascribed to variation that is only "slightly" different. These relative, and totally subjective differences are basically used as a proxy for whether something is capable of producing fertile offspring or not. This I think is poor science.

There has been a few papers criticising the recognition of many avian subspecies, pointing to lack of reciprocal monophyly as evidence of misleading classification; i.e. that subspecies should be monophyletic, and indicating that workers have been describing local populations with slightly different morphologies as different subspecies.

My first choice would be to recognise all independently evolving, monophyletic groups as separate species, no matter how small the difference is, and providing data are presented and defended in concordance with our current knowledge of the systems involved. There is no need for subspecies therefore, as the species rank becomes the smallest unit of formal biodiversity. I am glad that most if not all current ichthyologists agree with this idea, and therefore do not use the subspecific taxon.

Hybridization never results in speciation

It can do actually. See here.

[apistomaster]

There are some fish like Oncorhynchus clarki, the North American Cutthroat trout that has about 14 recognized extant subspecies; some recently met with extinction. They cover most of Western North America, share much of the same range as Oncoryhynchus mykiss, Rainbow Trout.
No one argues these are two species but they freely interbreed and produce intermediate phenotypes which are fully fertile.
The Rainbow is further divided into subspecies.
These classifications are very subjective in some ways and objective in others. Wild fisheries managers have to agree on which fish are what fish by some convention. Their criteria may but does not necessarily include molecular genetics. They and fishermen must obey regulations specific to certain phenotypes of these fish. It comes down to a need for whatever the reasons, to be able to label the different fish with different names. Social science crosses practical paths with objective science where neither school of thought is happy with the definitions of the other but are legislated to do so. See http://en.wikipedia.org/wiki/Cutthroat_trout

[Janne]

My first choice would be to recognise all independently evolving, monophyletic groups as separate species, no matter how small the difference is, and providing data are presented and defended in concordance with our current knowledge of the systems involved. There is no need for subspecies therefore, as the species rank becomes the smallest unit of formal biodiversity. I am glad that most if not all current ichthyologists agree with this idea, and therefore do not use the subspecific taxon.

Does this mean that for example: Peckoltia sabaji consist of several species instead of one? The population in Rio Tapajos is different in the colouration compared to Rio Xingu's populaion and these 2 populations is different in the colouration and fin pattern to Peckoltia sabaji in Guyana and the one in Colombia, all of these populations are separated and isolated populations, no gene flood between them... today? That they didn't have access to the other "species" except the holotype thats why all become the same species, or maybe it's we hobbyist's that put them under one species but they don't belong there and are considered as undescribed species?
Many Apistogramma species is living separated and isolated even that they are considered to be the same species, they differ in pattern or colour for example A. bitaeniata we have a yellow colour form living isolated and separated from the blue colour form, they should also be considered as 2 different species?

I don't state anything, I ask these questions to learn and to understand the logic.

Janne

[Carp37]

Hybridization never results in speciation

It can do actually. See here.

There are also a number of fish groups that are tetraploid- e.g. the catostomid "suckers" of North America- I think these are believed to have originated from hybridisation events, but maybe someone with more current knowledge might dispute that.

[apistomaster]

I have never heard about natural tetraploidy in Catostomus Suckerfish species.
We consider them a "rough fish" and most caught suckers are left on the river bank for the raccoons and skunks.
We have millions of them up to 30 inches long.

[Carp37]

I have never heard about natural tetraploidy in Catostomus Suckerfish species.

I'm sure they have a much higher than usual chromosome count- maybe it's only proposed that their high chromosome count came about via tetraploidy, possibly by hybridisation. The books/papers I read this in would be 30-40 years old now.

[Carp37]

http://www.wdc-jp.biz/pdf_store/isj/pub ... /34415.pdf

The above pdf confirms that they're tetraploid- thought I was going crackers there!

[apistomaster]

I looked using Google and found this:
http://www.wdc-jp.biz/pdf_store/isj/pub ... /34415.pdf
Catostomus commersoni is a common local species.
It does indeed have an unusually high chromosomal count; 2n=96-100.
I have been amazed but I have caught a few large specimens on weighted fly imitations of a large stone fly nymph.
Their chromosomes are much more numerous than the fish I have read more about their karyotypes; Old World Killiefish. 2n=~40 is closer to theirs.

[apistomaster]

Hi carp,
That was a tie on who got to Google first.

[Suckermouth]

Hybridization never results in speciation

It can do actually. See here.

Insert foot in mouth. But thanks for the article, now I know.

My first choice would be to recognise all independently evolving, monophyletic groups as separate species, no matter how small the difference is, and providing data are presented and defended in concordance with our current knowledge of the systems involved. There is no need for subspecies therefore, as the species rank becomes the smallest unit of formal biodiversity. I am glad that most if not all current ichthyologists agree with this idea, and therefore do not use the subspecific taxon.

Does this mean that for example: Peckoltia sabaji consist of several species instead of one? The population in Rio Tapajos is different in the colouration compared to Rio Xingu's populaion and these 2 populations is different in the colouration and fin pattern to Peckoltia sabaji in Guyana and the one in Colombia, all of these populations are separated and isolated populations, no gene flood between them... today? That they didn't have access to the other "species" except the holotype thats why all become the same species, or maybe it's we hobbyist's that put them under one species but they don't belong there and are considered as undescribed species?
Many Apistogramma species is living separated and isolated even that they are considered to be the same species, they differ in pattern or colour for example A. bitaeniata we have a yellow colour form living isolated and separated from the blue colour form, they should also be considered as 2 different species?

I don't state anything, I ask these questions to learn and to understand the logic.

Janne

P. sabaji and its lookalikes haven't had much research done, so similar species probably aren't named due to the research not having been done/published yet, not that they are the same species. As far as I know, few fish have a distribution that includes rivers as distant as the Essequibo and the Xingu.

I will mention that isolated populations have not always warranted naming separate species. You have to also show that they are also different. Isolated populations can be pretty similar if the isolation has happened in recent time, and especially if the separate populations are actually really large.

[MatsP]

Milton, I just asked J. Armbruster, and his opinion is that P. sabaji are actually considered the same species both in the Guyana Shield and Brazilian Shield. He (or "we" as the e-mail reply said) couldn't find any difference, but the fish were found in big rivers, so are likley capable of moving around quite a bit.

I agree that fish with such a wide distribution is unusual.

--
Mats

[Suckermouth]

Milton, I just asked J. Armbruster, and his opinion is that P. sabaji are actually considered the same species both in the Guyana Shield and Brazilian Shield. He (or "we" as the e-mail reply said) couldn't find any difference, but the fish were found in big rivers, so are likley capable of moving around quite a bit.

I agree that fish with such a wide distribution is unusual.

--
Mats

Man, I'm just messing up all over the place, aren't I? Haha. I didn't know that he's looked at some of these other fishes, so I suppose I'll defer to his opinion. On the other hand, "couldn't find any difference" doesn't mean there isn't one, and he even hints in the paper about a separation between Orinoco and Essequibo populations in his description, although I suppose what he knows might've changed since then.

[Janne]

My first choice would be to recognise all independently evolving, monophyletic groups as separate species, no matter how small the difference is, and providing data are presented and defended in concordance with our current knowledge of the systems involved.

Milton, I just asked J. Armbruster, and his opinion is that P. sabaji are actually considered the same species both in the Guyana Shield and Brazilian Shield. He (or "we" as the e-mail reply said) couldn't find any difference, but the fish were found in big rivers, so are likley capable of moving around quite a bit.

But they are obvious different, everyone can see it when you compare them.
Not the best pictures and I can take new ones later, these are a few years old at the time some wanted them to be Ancistomus but I can see differencies between them. I have another pic somewhere showing a white colour form.. the fins are white with spots from Rio do Para, maybe back home in my pc.

P. sabaji - Rio Tapajos

P. sabaji - Rio Xingu

Both these populations is really large and separated.

Janne

[Jon]

There can be no well defined answer because the species concept is misconceived and outdated.

One recent study I read pointed out that upon completing DNA testing of 14 species of Lake Victorian c i c h l i d from 9 different genera (different genera!) there was less genetic difference between them than between a typical population of humans.

I read that paper. I'm pretty sure it was EST based, which means transcriptomes were compared, not DNA. Which is more or less meaningless if you want to infer sheer genetic relatedness.

edit: never mind, wrong paper.

Genetic distance means little to nothing.

I don't think they were really proclaiming anything with the finding. After all, it's an obvious conclusion. The lake, as we know it, is only a few tens of thousands of years old. We had long diversified from that time.

[Mike_Noren]

So this problem is the same for all other animals too, there are small variation within the same species but not enough to separate them as their own species, colouration, habitat adaptation etc. doesn't matter. Does this also meen that all existing subspecies today will be synonyms and become one species in the future?

I am actually of the opinion that nearly all described subspecies should be considered separate species. Can you diagnose the group? Fine, call them a species. You can't? Well, they're a population.
With fish subspecies tend to be distinct & diagnosable - in other words, species.

[apistomaster]

Hi Mike,
I agree completely with you.
In the instance of the North American Cutthroat trout there are or were at least 14 different subspecies recognized.
Some only grow to 6 inches and others as large as 40 inches. All share the red slash markings and most can hybridized producing fertile hybrids but they otherwise have a great deal of different color types and many were completely isolated since the end of the last ice age until different populations were introduced to each other by fisheries managers.

[Jon]

Despite the seemingly wide breadth in phenotypic variance, cutthroats are relatively genetically homogenous.

[apistomaster]

Hi Jon,
And this is different from Hypancistrus, how?

[racoll]

I am actually of the opinion that nearly all described subspecies should be considered separate species. Can you diagnose the group? Fine, call them a species. You can't? Well, they're a population.

I am also entirely of this opinion.

There can be no well defined answer because the species concept is misconceived and outdated.

Quite a bold statement. Care too elaborate? If you conceptualise species, not as a absolute truth, but as a scientific hypothesis, it becomes much easier to reconcile the seeming conflicting information.

A species to me represents a macropopulation with an independent evolutionary trajectory. To recognise a species one must demonstrate, using supporting data, limited gene flow and divergence. Recognising populations on the cusp of independence will always be difficult to justify, but as long as hypotheses are supported with data, this is not problem.

Does this mean that for example: Peckoltia sabaji consist of several species instead of one?

Regarding the P. sabaji, these seem to look different to me too, but the way the system works, is that a hypothesis has been proposed stating that on the basis of morphometrics and meristics of preserved specimens, these populations should be inclusive of one another.

"couldn't find any difference" doesn't mean there isn't one

Exactly, and this does not mean they cannot be split further in future, but this would have to be subject to finer scale analysis using live colour patterns and DNA. But until these data are available, I think Dr Armbruster et al. have taken a defensible approach, and chosen not to inflate the number of names, potentially creating confusion and work for other taxonomists to clean up.