Taxonomic Merge 72486
(Borrador)
|
|
combinado en |
|
Comentarios
The fact that the statement says: ""Two subspecific populations of G. tigrina may occur"
Genetic data are often not conclusive and often misleading. We have had botanists having to resurrect several genera after sinking them based on DNA work " not support such differentiation " and then when more DNA work was done, it was discovered that the classical delimitation was very well supported.
I am totally against this merger, until the data are conclusive, and the authors of the paper state ""Two subspecific populations of G. tigrina do not occur, and substantial DNA data show that the populations are totally panmictic"
No one is using G. t. methi -- by location, all these should be methi and not tigrina: https://www.inaturalist.org/observations?nelat=-28.513406594363396&nelng=34.63732716209307&place_id=any&swlat=-31.13413265689118&swlng=26.761698988741784&taxon_id=41595
If no one is using it, then by definition iNat's taxonomy is mis-specified: G. t. tigrina is already sensu lato and includes G. t. methi. The taxon change would simply reflect current usage.
Incorrect usage just not justify dropping subspecies. It suggests that the IDs should be reviewed and fixed.
@jwidness: on what grounds do you make the statement "by location, all these should be methi and not tigrina:" ?
According to https://www.ewt.org.za/wp-content/uploads/2019/02/17.-Cape-Genet-Genetta-tigrina_LC.pdf
Two subspecific populations of G. tigrina may occur;
G. t. tigrina, ranging from southern Western Cape Province to south KwaZulu-Natal and
G. t. methi from the Pondoland area, southern KwaZulu-Natal (Roberts 1951),
although morphometric and genetic evidence did not support such differentiation (Gaubert et al. 2005a). Historic distributions of G. tigrina elsewhere mentioned in Lack
(1977), Wemmer (1977) and Skinner and Smithers (1990) now reclassify to other species, with the Kenyan population for example reclassifying to G. maculata.
There is only a single instance of potentially methi currently on iNat based on distribution identified as subsp. trigrina, https://www.inaturalist.org/observations/33951974 , so that statement must surely be wrong?:
I'm very confused. There are a whole handful of observations from right near https://www.inaturalist.org/observations/33951974. If 33951974 is methi, why would the others not be?
Additionally, you yourself have been calling those individuals G. tigrina tigrina.
We can't simultaneously support both G. t. tigrina sensu stricto and sensu lato. If people have previously been using the sensu lato definition, the best way to correct this is through a taxon split rather than just adding more IDs.
The comment from @bushboy on https://www.inaturalist.org/observations/11136687 ("Only Genet at this location and south is G.t.t.") suggests to me that the iNat community has been using G. t. tigrina sensu lato, not sensu stricto.
You will need to ask him. But south goes a long way before we get to Pondoland ....
What characteristics differentiate these taxa? Are there at least any phenotypic differences? Is there any isolation between these two supposed populations?
Isolation: a forest-preferring species: so potentially lots of isolated populations, especially inland versus coast.
The southern African iNaturalist community does not have access to the morphological differences between the subspecies. Due to the taxonomic mess within the tigrina group over the last 30 years, none of the field guides mention the diagnostic or characteristic features of the two subspecies - they all still deal with the issues of distinguishing rubriginosus (which is not listed as a synonym or species on iNat), trigrina and maculata, which were all subspecies, or confused, and not matching the taxonomic concepts to the north.
The only data commonly available is the geographical differentiation.
My tuppence worth:
I'd never heard of methi, until Tony decided to rechristen the Umdoni Forest stand-up fellow.
"Pondoland" is pretty fuzzy. The Pondoland Centre of (plant) Endemism is however pretty well defined - Msikaba Sandstone topped with sandy, leached soils supporting impressive endemic vegetation from Port St Johns north, petering out at about Oribi Gorge, n/w of Port Shepstone. Base vegetation is vigorous grassland with a few forest patches in protected gorges.
This is well south of Pennington's Umdoni Forest, and a different habitat.
Sadly, there are no observations on iNat of any Genet at all in Pondoland - a huge gap from Pennington all the way down to Kenton-on-Sea.
Ok, so I went back to some of the original source material. The subspecies definition is here: https://journals.co.za/content/nfi_annalstm/21/1/AJA00411752_470. (It may also be worth noting that he describes 4 new subspecies in this document, the other 3 have all already been invalidated.) Roberts compared 1 male and 1 female of "methi" to 3 males, 1 female, 6 unsexed of tigrina. He gives no pelage differences, only that methi is larger. He gives no range for the subspecies and the male and female were caught at the same location: "in the bush at the mouth of the Umsigaba River, Pondoland".
In Roberts 1951, he provides a range description: "So far known definitely only from the type locality, the mouth of the Umsigaba River, but probably extending over a considerable area to the south and north."
In summary, there is very little evidence to support this subspecies being distinct, and there is certainly no confidence in using either pelage or location to give subspecies IDs.
We must have read different articles, because there are a few differences:
in colour and colour markings, including the pale tail rings and marks on the underside similar to G. t. tigrina of the Cape of Good;
but larger in general size and skull,
especially in the back teeth, which are larger than in the even larger G. rubiginosa ...
in both the type (male) and cotype (female) the inner cusp of p3 is well developed.
The male lacks, but the female has, a rusty shade in the dark spots on the upper parts of body.
But lots of subspecies are size and dentition forms, and dont show any morphological features. I dont think that it is so easy to dismiss subspecies. I would prefer to see a proper study.
A Pondo Genet at last!
See https://www.inaturalist.org/observations/39786848
More discussion here: https://www.inaturalist.org/comments/4230320
To add to the discussion:
In the Kingdon Field Guide to African Mammals (Academic Press, 1997), Johnathan Kingdon uses information that was ruling genet taxonomy prior to Philippe Gaubert's extensive work and revisions of the genus Genetta. He describes 8 species of genets, and under "Blotched genet Genetta tigrina", he lists the following subspecies, and indicates that a revision is needed:
G. t. maculata; G. t. poensis; G. t. pardina; G. t. mossambica; G. t. bini;, G. t. aequatorialis; G. t. stuhlmanni; G. t. schraderi; [+ at least 20 other races]
After Gaubert et al.'s work (but which still needs to be expanded, notably by means of genomic analyses), we now have 14-15 species (Genetta felina having not yet been unequivocally accepted; no doubt about Genetta genetta felina as (at least) subspecies though) and among the new species, the first three subspecies listed above have been upgraded to species level:
Genetta maculata (Central African large-spotted genet or rusty-spotted genet); Genetta poensis (King genet) and Genetta pardina (West African large-spotted genet or pardine genet).
G. mossambica or G. t. mossambica has been shown to be a hybrid of G. angolensis and G. maculata. Other subspecies have not been elevated to species level, and several of those are synonyms (among them or synonym with currently accepted species) or invalid.
Interestingly, Kingdon forgot to list G. t. tigrina, which was described in 1878 already. Since the later described G. t. methi does not exist, it means that G. tigrina is monotypic and using "Cape genet G. tigrina" is the correct way to go based on current knowledge. Here is what Gaubert et al. (2005) wrote:
"The specimens attributed to G. t. methi (from southern KwaZulu-Natal: Oribi Gorge, Underberg and Donnybrook) grouped unequivocally with south-western populations of G. tigrina, thus rejecting the hypothesis that methi may constitute a hybrid population between G. maculata and G. tigrina (see Crawford-Cabral & Pachecho, 1992)."]
The work by Roberts (1948-1951) is based on a handful of individuals, and even lists colour differences between sexes based on the examination of 1 male and 1 female!
"The male lacks, but the female has, a rusty shade in the dark spots on the upper parts of body".
Unfortunately the above does not make much sense, and the presence of a rusty shade or not varies from individual to individual. I have seen rusty-spotted genets G. maculata that have very rusty spots, and others that have no rusty tinge at all. In addition, I personally do not know of any intersexual variation in coat colour pattern in any of the currently described genet species. If Roberts work were submitted to a journal today, it would even not be sent to reviewers. With the huge inter-individual variation observed in the Durban and Pondoland area (this being only partly linked to hybridisation between G. tigrina and G. maculata), I am sure that by collecting randomly 100 male-female pairs of genets and comparing them to 5-6 G. tigrina "tigrina" individuals from Western Cape, one could likely describe several new (but of course invalid) subspecies.
I am convinced that the future (once full genome analyses from a large sample of genets have been carried out) will bring some fascinating twists to the genet taxomy story, but for now I am of the opinion (i.e my opinion only) that parsimony and caution should be adopted.
This has probably been one of the most meaningful discussions on subspecific taxa I have ever seen on iNaturalist. I appreciate everyone's participation and thoughtful input. Based on the current understanding of Genetta sp. taxonomy, the most accurate representation of G. tigrina is monotypic.
Unless there are any additional comments, the swap should be committed.
It is not valid to evaluate old publications by todays standards. Had the standards been in place, they would probably have been met. But techniques and computing and analytical opportunities have advanced considerably since then.
What has gone backwards is our field work and appreciation of differences between species in the field. Choosing between a modern and older field worker, I know which one I would prefer on my team any day.
My point above was that a new subspecies cannot be described based on a sample of 2, with one male and one female. One can publish such results in a short note and highlight the differences with previously known samples/specimens, and encourage larger sampling to clarify the matter, but not claim this is a subspecies. It is clear that the standards where not in place at the time, and this is no one's fault. Roberts can only be praised for his monumental work and efforts to advance knowledge. What is inappropriate is to take observations/results from the past and continue to use them in modern times when the current state of knowledge clearly suggests that the scientific grounding of such observations/results is shaky.
I couldn't agree more with you Tony. I am a field biologist and this is where I learn about my studies species, and come up with new questions and hypotheses. There are less and less young people who go in the field and "open their eyes", unfortunately. In my line of work, my job is about encouraging local students to go back to basics. I only do lab work when there is no alternative (often determined by lack of funding, time or manpower), and I am not a big fan of advanced statistics. The results are often difficult to interpret and boring to read. When you publish results describing e.g. the types of resting sites used by a small carnivore or a rodent, you know these observations will be valid forever and there is no dispute about it. [The sad thing is that journals are reluctant to accept descriptive work, and you can only publish this in journals with low impact factor. This also means less money for research. But I stick to what I like.] Genetic results, in contrast, may be overturned in a matter of 6-24 months, but of course this is more "sexy" and trendy.
As my old professor said: taxonomy is easy when you only have one or two specimens.
The bottom line is that 99% of our species were described when there were only one or two specimens. For three hundred years collectors toured southern Africa and sent back specimens to Europe where they were described on the basis of a single specimen.
Roberts had two specimens - he was ahead of his time!
And now, using todays standards you want a significant sample size? You are going to have to throw out millions of species by that restriction.
Yes, we need all these species to be checked and validated. What we dont need are fashions where subspecies are denied, or all elevated to species based on a quick and dirty DNA analysis of a few megabytes of sequence of a few genes. We need proper studies.
All I am saying is dont throw out the baby with the bathwater. These old guys knew their stuff. Keep their concepts as the straw dog for further studies. Dont trash them and then in 50-100 years let us discover that they were right all along (even with their limited resources).
If Roberts collected two specimens from Pondoland, it is because he noticed that the animals were different. The two specimens are all that he managed to sample from this different population: how else could he report his results other than to compare it to the other specimens to hand. he did and they were different. Quo vadis: we have a new species/subspecies.
((Unless you have evidence that Roberts had no idea that the Pondoland animals were different and cheated by selecting these two specimens - after he had done all the measurements on all the specimens - on the basis of these measurements, and then created the subspecies.))
And now 80 years later, we sit and criticize him because he did not collect 100 specimens from Pondoland, and increase the number of specimens from the rest of the country by 100 000, and therefore trash his results as "unsupported", "unstatistical" and "speculative". Come on: be fair to the poor guy, and the data. I maintain that we should recognize his subspecies until WE get enough data to disprove (or not) it.
Science should not be about trashing knowledge because it is outdated: it should be about taking that knowledge and vigorously testing it. There is good precedent: Medicine started off assuming all herbal remedies were valid and then testing them and finding those that worked (best). Even today, the starting point for new drugs is to check what the herbalists and quacks are using, and starting from there: they do not start off my saying it is all rubbish and must be ignored, and then testing all 200 000 species from scratch.
We should do the same.
I reiterate. Roberts tested his hypothesis and created a new subspecies. Keep the subspecies until we can test if it is true/untrue with proper data.
I agree with several of the points above, especially considering the twists and turns in taxonomy (subspecies, then species, then back to subspecies) that have been witnessed as the number/type of molecular markers included in studies increases. However, the issue at hand here is at another level:
Someone (let's forget about names and times, as it is really not of the essence here) described a new subspecies, and subsequent studies, using the same metrics, additional metrics and a larger sample size, as well genetic analyses (which was not the case in the previous study) have shown that there is no difference between two taxa (here tigrina and methi). Therefore either we believe the first study and continue to consider both subspecies, or we believe the subsequent study and drop the subspecies (in this particular case also because the subspecies "tigrina" was used to differentiate from individuals with the "maculata" phenotype, not "methi" ones). If we go ahead with the first option, then we must also restore the hundreds of thousands of taxa which have been invalidated by subsequent research (and I am not only talking about synonyms). This is possible but will this be helpful?
Second, the same study which invalidated the existence of methi has been used (possibly unknowingly) to elevate the subspecies Genetta genetta felina to the rank of species, Genetta felina. The authors of the study recommend to use the species level until further studies confirm this. There is clearly a difference (even the phenotypic ones are obvious) between felina and other taxa, but it is not yet clear what is the overall genetic difference (could be said in terms of % of nucleotides/genes that differ), and the question is whether this will ultimately deserve species level.
I have therefore difficulty to understand how the same study could be used to 1) keep a subspecies for which no substantial evidence has been found, and 2) elevate a subspecies to species level when it is not yet 100% sure whether it is a species. In fact, both decisions go against the scientific precautionary principle. Even when people determine records on this site, they will maybe start with features that correspond to the order, then go to the family level, genus, species and lastly subspecies if clear-cut criteria exist. If elements/evidence are missing at one level, then that's where the determination should stop. Similarly, if more advanced analyses tell us that a subspecies does not exist, the precautionary principle should encourage us to stick to the species level. If further studies were to find that several subspecies exist within Genetta tigrina (even if only at the genomic level), then only can we start using subspecies, although even here it would be impossible to allocate those subspecies to records if no clear phenotypic criteria exist (distribution could be used if the distribution range is clearly established in the paper and there are e.g. allopatric populations).
The above has nothing to do with a personal opinion, nor is linked to methods used per se. This is just based on logic and the precautionary principle. We all wish we knew more, but it's not the case.
Lastly, it is true that a lot of species have been described based on a single specimen. This was linked to logistic issues and the objective of bringing back samples of as many species as possible. In several cases though, 2-5 samples were collected and only one used as the holotype. Whether there was a single specimen or several of them, however, does not mean at all that we have to throw out million of species. Even if the Plains zebra has been described with a single specimen, the fact is that there are thousands of individuals with the same phenotype which can be observed on a daily basis to confirm that this is indeed a very diagnosable taxon (and which differs e.g. from Mountain Zebra or Grevy's Zebra). In the case of the proposed subspecies methi, the diagnosed features were based on 2 methi individuals vs 5 tigrina individuals, and when the tigrina sample was increased, it was noticed that the methi metrics were within the range of the values observed for tigrina. Therefore, no subspecies can be considered, and logically the diagnosable characters (if they exist) would at least not be those described in the paper.
To be clear, Genetta felina is considered a valid species on iNaturalist (https://www.inaturalist.org/taxa/925659) because the taxon is recognized as valid by iNaturalist's external authority on mammals, the ASM Mammal Diversity Database (MDD) (https://mammaldiversity.org/species-account.php?genus=genetta&species=felina).
Yes, indeed. It was upgraded from subspecies at some stage over the past two years, likely based on the 2005 work I mentioned. The work also indirectly indicated that Genetta tigrina is monotypic.
@jwidness @tonyrebelo thoughts?