PlanetCatfish.com

Regarding the identity of the common bristlenose catfish (). Following on from this thread, I had the idea of seeing how many specimens were in the DNA barcoding database, to see if there were any matches to the common bristlenose (which I have sequences for).

There weren't that many, but there were several wild caught specimens of , from both the lower and upper Parana basin. Assuming these were correctly identified then it's unlikely that the common bristlenose is this species, as they were very different (nearly 10%, which is a lot for this gene). Unfortunately, there are currently no close matches in the database to the common bristlenose.

Further reading, see this article.

Héhé!
Who had doubts about this? The identity of the common Ancistrus as A. cf. cirrhosus is just another urban legend! But try to say the contrary, and nobody will trust you, even when you are one of the leading specialist of the group... or when you provide evidences as you just did it.

Great idea, and great job!
Cheers
Raph.

racoll wrote: There weren't that many, but there were several wild caught specimens of

Ancistrus cirrhosus

, from both the lower and upper Parana basin. Assuming these were correctly identified then it's unlikely that the common bristlenose is this species

There's the rub right there. FishBOLD is brimming with crap; there's no telling what you've compared to.

In addition most sequences in Fish-BOL have never been submitted to GenBank and are hence not available for scientific study.

Can't say I'm surprised, though it's nice to see it with evidence.

This isn't really that surprising, but also doesn't account for the possibility of the common being a hybrid. At 10%, assuming the correct ID, it is highly doubtful they are the same species. However, I doubt ever finding a real heritage for these. They've been around the hobby for a very long time, and I'm fairly sure have been hybridized multiple times either accidentally or intentionally. The odds of L144 having been the same species is definitely questionable, but the genes were introduced to produce the blue eye gold strain. I doubt this being the only incidence.

I'm suggesting a name change from common Ancistrus to flowerhorn Ancistrus

Larry

pleco_breeder wrote:This isn't really that surprising, but also doesn't account for the possibility of the common being a hybrid.

It doesn't account for it being a hybrid, but if the maternal species was in the database, the common bristlenose would cluster with it. It clearly isn't in the database, though.

Hi,

What is writen above is not exactly correct. Fish-BOL is not sensu stricto a database, it is a barcoding project. Sequence in the BOLD database are also deposited in GenBank, since you have to submit them directly to the GenBank database (using Bankit for barecodes, see at NCBI). This is specified by BOLD, and of course sequences are available, via both GenBank and BOLD. What would be the interest to create such a huge database if the data were not accessible? Second, when you submit a barecode, you also have to provide a voucher refering to the specimen. Normally, sequences belonging to this specimen are correctly identified, but in case of doubts, you can turn back to the specimen.
I checked the database to see who made the submissions, and sequences for Ancistrus should be correct.

So what was the question: is the common Ancistrus an Ancistrus cf. cirrhosus? and the reply was : No. Of course it can be a hybrid, but this does not make it an A. cf. cirrhosus.

As I specified, even when you provide evidences, people doubt. Of course, they read it in an article... yes, but an aquarist article.

Good luck
Raph

Mike Noren wrote: There's the rub right there. FishBOLD is brimming with crap; there's no telling what you've compared to.

Hi Mike. I understand your scepticism, and I also am sceptical of the overall usefulness of the database in relation to bad IDs (in fact I am about to submit a manuscript on the subject).

However, this doesn't mean you should "throw the baby out with the bath water", i.e. just because some fish are misidentified, doesn't mean they all are. You just have to judge each entry on its own merits--who did the ID, is there supporting information, is the locality accurate and correct?

Any hybrid conclusion is a serious leap of logic. Ancistrus spp have shown that they are highly, highly unlikely to hybridize. I have checked 40 years of aquarium literature (including all past publications by the UK and US catfish associations) and the internet and have found exactly one documented case of hybridization between two clearly distinct Ancistrus spp.

Before we went with I was leaning towards the common Ancistrus possibly being one of the northern spp from a location where we no longer receive aquarium imports from. There is also of course the very real possibility that it is an undescribed sp. After Chaetostoma, Ancistrus is probably (as a genus) the least well documented genus of its family.

Regardless, we will certainly need a new name for an old friend.

-Shane

Raph wrote:What is writen above is not exactly correct. Fish-BOL is not sensu stricto a database, it is a barcoding project. Sequence in the BOLD database are also deposited in GenBank

Try finding the A. cirrhosus sequence in GenBank.

Fish-BOL sits on the sequences, some times for years, before submitting them to GenBank. Worse, they have at least in the past submitted placeholder sequences ("unknown species, unknown sequence"). And to top it off a high percentage of sequences are misidentified/contamination.

What would be the interest to create such a huge database if the data were not accessible?

To Fish-BOL or to the rest of the world? Their interests do not necessarily coincide.

racoll wrote:

Mike Noren wrote:However, this doesn't mean you should "throw the baby out with the bath water", i.e. just because some fish are misidentified, doesn't mean they all are. You just have to judge each entry on its own merits--who did the ID, is there supporting information, is the locality accurate and correct?

If you as the end user have to evaluate the metadata in order to find out whether to trust the results, then BOLD has failed. It's supposed to be a quick-and-easy way for non-experts to accurately identify specimens and produce, but what it actually provides is an unpublished clustering algorithm to group unpublished sequences of questionable quality.

Throw the baby out with the bathwater? I'm not sure there IS any baby in the bathwater. If I want to barcode something I do a BLAST search on GenBank, it is not clear to me that BOLD provides any additional value.

Hi Mark,

You are right, the only sequences of Ancistrus cirrhosus deposited in GenBank are mine, and they are not COI. The only public sequence is the one of A. brevipinnis (also present in the BOLD database as it should be the case for all other sequences). This is rather strange since it is explicitely written here :http://131.104.97.143/wp-content/upload ... icies1.pdf at iBOL that data release should be done within few weeks, and made available to the community.

And if you make reference to the huge release of sequences of Siluriformes sp. in GB last year by BOLD, I fully agree that it caused serious troubles with Blast searches (GB is not a trash).

Cheers
Raph

Oups sorry Mike of course!

I'm aware of one other incidence of hybridization by a hobbyist several years ago. They had been selling TR fish to a shop I was financially involved with. After almost two years of selling these to the shop, they decided to sell their breeders to the shop. The female appeared to be a common. The male was a very dark grey/green color with no spotting or markings in the fins. He was also substantially more narrow when viewed from above than the common and the nose not as rounded. I suspect similar spawnings have been occurring regularly during the captive maintenance of whatever is being maintained.

Likewise, the TFH write-up which introduced the first pics of albino Ancistrus I'm aware of was a fish imported from Trinidad per the article. I know this is unlikely because of the locality not being a regular export site. The catelog shows only two species of Ancistrus from Trinidad. While it is possible this is the same species, or albino randomly appeared in the common species around the same time, I would be more inclined to believe an outcross was made to establish the trait. That is not documented as far as I know, but would not be the first time it has happened.

Again, as is stated several times on forums around the world, the only imported specimen of the true L144 was from Paraguay. Although it isn't documented that I'm aware of, the common belief is this fish was crossed back to the common strain to produce the blue eye gold strain now commonly sold as L144 and bred frequently with other strains of common.

I'm not saying it isn't possible these were within a species. However, if the above mentioned possibilities did take place, it would be highly unlikely to be spawns within the same species. The geographic distance, as well as the saltwater if the albino was from fish actually from Trinidad, would definitely qualify as a geographic bottleneck to separate the species.

I can accept that there is some speculation involved in my theory. However, on the flip-side, it has to be reasonably considered that the fish we keep as commons in the hobby have been far removed from their original species either unintentionally by unknowing hobbyists or commercial endeavors.


Larry

Raph wrote:This is rather strange since it is explicitely written here :http://131.104.97.143/wp-content/upload ... icies1.pdf at iBOL that data release should be done within few weeks, and made available to the community.

Yeah, I know.

And if you make reference to the huge release of sequences of Siluriformes sp. in GB last year by BOLD, I fully agree that it caused serious troubles with Blast searches (GB is not a trash).

I didn't know about the siluriforms. It was in 2009 - 2010, if memory serves I was working with skates and eels at the time. I complained to GenBank that "no name" sequences were worse than no sequences, as they diluted searches. GB said it was out of their hands, Fish-BOL had promised to add metadata within days and then stalled. I tried to keep tabs and when I moved on, almost a year after the sequences went public in GenBank, they still had no metadata.

Likewise, the TFH write-up which introduced the first pics of albino Ancistrus I'm aware of was a fish imported from Trinidad per the article. I know this is unlikely because of the locality not being a regular export site.

This is what I was talking about when I said, "I was leaning towards the common Ancistrus possibly being one of the northern spp from a location where we no longer receive aquarium imports from."

I would be more inclined to believe an outcross was made to establish the trait.

Outcrossing will not establish a trait, only dilute it. Traits are usually established through heavy inbreeding, not outbreeding (hence all the chronic health problems from inbreeding associated with so many modern dog breeds). If this albino from Trinidad were the ancestor, it would have been bred to a member of the same sp again, then its own offspring, then, their offspring, etc until the recessive "stacked" enough to become fixed. This would be the same process for all the other fixed deformities we see in the common Ancistrus. The fact that so many traits have been fixed tells us we are working with a single gene pool of a single sp. that is well understood by those farming this sp.
-Shane

Shane wrote:The fact that so many traits have been fixed tells us we are working with a single gene pool of a single sp. that is well understood by those farming this sp.

I'm not so sure that this is true.

In domestic canaries there have been fixed many many many more traits and these are hybrids.

Fixing a genetic trait is as easy in hybrids as in a single species. If the hybrids are fertile of course.

In birds it is/was common practice to create hybrids in order to bring genetic traits from one species into an other. I can't see why this should be impossible within Ancistrus. Deliberately or not.

Shane wrote:Any hybrid conclusion is a serious leap of logic.

Ah, this old chestnut!

Ancistrus spp have shown that they are highly, highly unlikely to hybridize.

But how do you know that? Why, are they so unlikely to hybridise when so many other loricariid genera are not?

I have checked 40 years of aquarium literature (including all past publications by the UK and US catfish associations) and the internet and have found exactly one documented case of hybridization between two clearly distinct Ancistrus spp.

Ah, but surely we should be looking for hybridisation between two clearly indistinct Ancistrus spp?

Remember your that you attempted to take control of the US market with? Where did you get them from?

I would propose that the vast majority of these fish that are bred are bred by fish breeders who buy them at auction or from other keepers as trades. I don't see wild caught bred anything like as often. Any they are very inconsiderately paired because, as this discussion proves, it's hard to tell fish apart that live thousands of miles apart.

In speaking with Czech breeders, they tell me, although I have no proof, that male captive or strain bred Ancistrus will spawn with pretty much anything wild caught.

The other question that's always bothered me is if this is a naturally occurring species, why do we never see it imported. We only see things similar to it.

Jools

Shane wrote:Outcrossing will not establish a trait, only dilute it. Traits are usually established through heavy inbreeding, not outbreeding (hence all the chronic health problems from inbreeding associated with so many modern dog breeds). If this albino from Trinidad were the ancestor, it would have been bred to a member of the same sp again, then its own offspring, then, their offspring, etc until the recessive "stacked" enough to become fixed. This would be the same process for all the other fixed deformities we see in the common Ancistrus. The fact that so many traits have been fixed tells us we are working with a single gene pool of a single sp. that is well understood by those farming this sp.
-Shane

Once the albino is established, because it is a recessive trait it only takes one individual displaying the trait to start the outcross. The progeny of this F-1 albino and a common would be heterozygous for albino. A sibling cross would then be very likely to produce 25% albino.

This is assuming they fry resemble the common strain. If not, those most similar to commons can be separated into pairs or small colonies to breed for albino to show again. Since all the fry from the initial cross would be heterozygous, it shouldn't take very long to get a fish displaying the albino trait with the same general features of the common strain.

This outcross of the original albino species just passed to this species with basic mendellian genetics.

Larry

pleco_breeder wrote: This outcross of the original albino species just passed to this species with basic mendellian genetics.

This is true. I'd however point out that COI is a mitochondrial gene and hence maternally inherited: if the albino male was a different species the hybridization would leave no trace in COI, because the offspring get their mitochondria from their mother.

Also: a 10% difference in COI between two morphologically almost identical (and presumably closely related) species is surprisingly high. It's not unheard of (e.g. european and american perch) but for comparison 10% is the distance between and , and twice the distance between and .

What is COI? I understand it's a mitochondrial gene, but apparently now being used heavily to establish species profiles for comparison. It's been quite a while since I took a genetics class, so these may sound a bit dated.

Larry

COI is Cytochrome c Oxidase Subunit I, it's the most commonly used gene for barcoding (species identification).

I am not sure what you are basing this statement on. I have managed to find 14 documented cases of "hybridization" in the family. Only ONE of these is a documented case of hybridization between two described spp (Pseudacanthicus leopardus X Pseudacanthicus spinosus). Of the remaining 13 cases, eight (61%) are between Hypancistrus, one Pseudacanthicus, one Leporacanthicus, and one Ancistrus. There are about 95 genera in loricariidae! So out of over 1,000 described spp, 500 plus L and LDA Numbers, and several hundred more known but undescribed spp we have possible hybridization on a baker's dozen of occasions between three of the family's 95 genera. Take out Hypancistrus (which appears to be a huge exception) and you are better off buying a lottery ticket than betting on two loricariids to hybridize.

Err, so you are happy with the assumption that out of the genera of loricariids that have not been spawned in the aquarium that 100% of them will not hybridise? If not, then your logic is a wee bit flawed. On documented cases, you are missing Pterygoplichthys and (see LH numbers). I have also seen Peckoltia, , and hybrids. I am saying that, out of the commonly spawned hypostomine loricariid genera, most are known to hybridise. Your leap of logic is to say that the ones that don't or haven't been spawned don't hybridise. I would also suggest that the creators of hybrids don't exactly rush (or are typically inclined) to document them.

You avoid answering my question on the source of your Ancistrus.

Jools

Shane wrote: I have managed to find 14 documented cases of "hybridization" in the family.

Isn't it possible that this number is so low because most of the hybrids (if they exist) are bred by people who think they are breeding with fish from the same species?

Some people can't see the difference between a P. gibbiceps and a bristelnose, do you think that they would write an article when two different bristelnoses breed in their tank?

Shane wrote: So the facts we have are:
1) hybridization is very rare among Ancistrus,

Can you use what you want to prove as proof?

Shane wrote: 2) limited barcoding shows the common Ancistrus to be evolved from a single lineage

I'm not an DNA-specialist but I thougt maternal DNA was used. So there is no prove there was never a male from another species involved.
In the catelog it is clearly stated that the original L144 was a male.

Shane wrote: the line breeding of genetic mutations is most easily accomplished within a small gene pool from the same sp.

This is not true. The chance a mutation pops up is bigger when you inbreed (pure or hybrid, that doesn't matter) but fixing a mutation is always accomplished by line breeding and line breeding happens by definition within a small gene pool, hybrid or not.

The presence of some mutations is no proof for a pure species. Take a look at Xiphophorus.

It think Jools made a little mistake with his magic tricks.

Bijn wrote:It think Jools made a little mistake with his magic tricks.

Whoops, sorry Shane, my bad, I think I nuked some of your post! Luckily I quoted some of it.

Jools

I think we're getting a bit ahead of ourselves with this speculation about hybrids.

One step at a time. The first stage is to identify the maternal species with mtDNA, then once that is established, then investigate the possibility of hybridisation using DNA from the nucleus (biparental inheritance).

Unfortunately, having decent reference specimens is the limiting factor here.

Mike Noren wrote: a 10% difference in COI between two morphologically almost identical (and presumably closely related) species is surprisingly high.

I doubt that all brown Ancistrus share a common ancestor, but yes, it is a lot (actually 8.1% according to BOLD).

If anyone is interested, here's the common bristlenose sequence ...

>common_bristlenose
CTTTACCTAGTGTTTGGTGCCTGAGCCGGAATGGTTGGTACAGCCCTCAGTCTCTTAATTCG
AGCTGAGTTAAGCCAACCCGGTTCTCTATTAGGTGATGACCAGATTTATAATGTCATCGTTA
CCGCACATGCTTTCGTAATAATTTTCTTTATAGTCATGCCAATCATAATTGGGGGCTTTGGA
AATTGACTAGTTCCACTAATGATTGGGGCACCCGATATAGCCTTCCCACGAATAAATAACAT
GAGCTTCTGACTACTGCCCCCCTCATTCCTTCTTCTACTGGCCTCTTCAGGGGTTGAAGCGG
GAGCTGGGACAGGTTGAACTGTATACCCACCCCTCGCCGGAAACCTGGCCCACGCAGGAGCT
TCCGTTGACCTGACTATTTTTTCACTACACCTGGCTGGTGTTTCTTCAATTCTGGGGGCAAT
TAACTTCATTACCACAATCATTAACATAAAGCCCCCGGCTATTTCACAATACCAAACCCCCC
TATTTGTGTGAGCCGTACTTGTTACAGCGGTCCTACTCCTGCTTTCCTTGCCCGTTCTGGCC
GCCGGCATTACAATACTGCTCACAGATCGAAATCTAAACACCACATTCTTTGACCCTGCGGG
CGGTGGAGATCCTATCCTTTATCAACACTTA

racoll wrote:I think we're getting a bit ahead of ourselves with this speculation about hybrids.

One step at a time. The first stage is to identify the maternal species with mtDNA, then once that is established, then investigate the possibility of hybridisation using DNA from the nucleus (biparental inheritance).

Unfortunately, having decent reference specimens is the limiting factor here.

Would sending fin clippings work? I would also suggest that we should have a wide international sampling of "the common ?

Jools

Where was this particular common ancistrus bought?

Where was this particular common ancistrus bought?

LFS in New Zealand. The "L144" variety I collected had an identical haplotype to the common. Two albino bristlenose have also been sampled in South Africa, and are in BOLD. One was identical to mine, and the other has two mutations over 651 bp.

Here's the fish:

Jools wrote:Would sending fin clippings work? I would also suggest that we should have a wide international sampling of "the common Ancistrus?

I was actually thinking about reference specimens of wild fish, but having said that, if hybridisation was more widespread than just a one-off event among common Ancistrus, then we would see multiple lineages. Sampling the trade would at least indicate that possibility, assuming introgression was not just from the male direction.