In Which I Talk About “Genera” – Again

This post is about taxonomy and “The Genus Question.” If you do not want to read, I suggest not going below the fold. If you wish to continue, you will be rewarded with a pretty picture.

The Paleobiology Database, is a taxonomic database started by and largely run by the Australian Research Council, and contains values such as changes in taxonomy, citing authors, locality information, geologic settings, and so forth, a broad range of data that is useful to the paleobiologist. Anne Lagomarcino of the University of Cincinnati’s Dept. of Geology joined with Arnold Miller combed this database for Cenomanian through Maastrichtian marine taxa.

They localized these taxa for informativeness and for region, choosing to focus on three: the Gulf Coast, the North American Cretaceous Seaway, and the European Epicontinental Sea, and used maps of these regions for each of four segments of time (Cenomanian, a combined Turonian-Coniancian-Santonian bin, Campanian, and Maastrichtian); while the Gulf Coast represented open-ocean, the other two represented epicontinental seas, and thus are influenced by pelagic versus near-shore or shallow-water taxa. Using a variety of metrics including the power law function (S = cAz), the authors computed their taxa against area and regressed concentrations for each area with each other.

Some outliers exist: the Turo-Coni-Santonian taxon sampling for area is particularly low on the Gulf Coast, while for the Maastrictian this diversity is low in the NA Cretaceous Seaway (likely due to the near absence of the seaway post-Campanian); the Tur-Coni-Santonian samples for all regions display relatively high slopes, indicative of high diversity in small and broad areas, but not moderate areas, while these slopes suggest rather that during other periods slopes are generally consistent.

Some of this is over my head, so some should pardon me for not covering all the details. The paper is free (as from PLoS One, it is open-access), and I encourage the interested to check it out for themselves. My concern is something that readers familiar with my blog will anticipate. In their abstract, Lagomarcino and Miller write:

Recent studies indicate that spatial distributions of taxa and the kinetics of taxon origination and extinction may have differed in these two settings. Against this backdrop, we analyze regional Genus-Area Relationships (GARs) of Late Cretaceous marine invertebrates in epicontinental sea and open-ocean settings using data from the Paleobiology Database.

This is enough to generally pique my interest, as I have been looking forward to finding whether there are useful metrics for evaluating the quality, if any, of a genus. The authors continue in their introduction:

While most present-day studies of the relationship between area and diversity have been conducted at the species level, one could just as easily look at these relationships at higher taxonomic levels, and, because of concerns about data quality, all analyses in this paper were conducted at the genus level, which has been the level of choice for a broad suite of investigations of ancient marine diversity conducted over the past two decades.

So the challenge is on: Is there a difference between genus-level versus species-level diversity over time in these three regions? What might this question resolve?

First, it should be noted that even given narrow qualifications of the term “genus,” “genera” or “genera-equivalent” nomenclature generally encapsulate more than one species in extant organisms, it is considered a sort of metataxon, qualifying a form of diversity by which scientists can measure “richness,” “likelihood of speciation,” etc. Yet “genera” do not exist, in at least the same way that species do. What then is meant when scientists use “genera” in the stead of species? For the most part, they assume that “genera” represent a form of actual genetic diversity, which can be quanitified (despite no one having done so, in at least a phenetic manner; and the same, generally, cannot be said of species, due to the plethora of species definitions, for none are forthcoming for “genera” that themselves do not depend on species). Hence, “genus” more resembles a metataxon, and not an actual taxon, and the same is true for all other supra-specific Linnaean ranks.

When Lagomarcino & Miller note that their option of taxon choice was the “genus,” it is because it is easier to compare this to previous studies; however, this ignores the fact that, unlike species, “genera” can shift by containing fewer or more species, and thus species remain constant while moving in and out of particular containing clades. The question similarly becomes moot if, say, we were to simply collapse all “genera” to their type species, and thus expand all fossil species to have a “genus” and species (we’d just call this a binomen: a species name preceded by a praenomen). This solves some tricky issues, but of course it brings up new ones, especially given that current ecological and conservation studies depend on the broader categories, rather than the specific, in order to define what is largely an irrelevant distinction: It is easier to conserve a taxon if it seems rarer, by further refining the taxonomy to restrict the possible population, or alternatively to subsume diversity into large categories and define “genera” with lone species to encapsulate how rare the breadth of lineage being considered really is.

Let me just say, that while I applaud conservation and preservation of modern diversity as far as man’s interference goes, playing political gamesmanship by alternately expanding or narrowing the nomenclature of these animals does nothing for scientific analysis by itself and should not be considered when assessing the genetic diversity of living or extinct organisms. Taxic diversity doesn’t necessarily equate to genetic diversity, nor the inverse.

The authors do not consider the species-level when analyzing their datasets, nor then do they consider relative naming conventions have on genetic diversity being approximated. Instead, they (and most other analyses of fossil organisms before them, as cited in the paper) use “genus” as a proxy for this concept, and assume it is effective in conveying the broader concept of diversity. I, respectfully, disagree. The question, of relative species and “genus” diversity, never comes up, and thus the question asked of the authors is instead mapping relative open-ocean and epicontinental sea diversity by generic defintions of their taxonomic proxies. I cannot help but think that this paper, well written and relevant in many ways to assessing diversity, fudges the quality and relevance of the work by not using the database at hand and use species instead.

Lagomarcino, A. J. & Miller, A. I. 2012. The relationship between genus richness and geographic area in Late Cretaceous marine biotas: Epicontinental sea versus open-ocean-facing settings. PLoS One 7(8):e40472.

As promised, a pretty picture:

Mama should be worried her Rinchenia-let is eating her feathers....

This entry was posted in Paleobiology, Paleontology, Taxonomy and tagged . Bookmark the permalink.

8 Responses to In Which I Talk About “Genera” – Again

  1. Fabrizio says:

    Any species having ITS OWN genus would be much more practical, i think. Are there any contrindications to it

    • Establishing a praenomen for all species would be fantastically huge, an undertaking that would take decades. The slippery slope argument is also relevant: How would you know what to start naming? Would domestic dogs each get their own praenomen, or each variety of dog? I like the idea that we should start this project on fossils only, and perhaps just craniates, or even just chordates. This would reduce the taxonomic load. Currently, more new vertebrate taxa are being coined as “genus-species couplets,” which means slowly, this concept of “genus” is largely becoming redundant with species — for fossils.

  2. I think it’s time to drop the whole “only species are real” cliche. Only clades are real, and species AND genera (and families, orders, phyla, etc.) can be clades, and increasingly are conceived of that way, since no one likes paraphyly anymore. Note many species concepts are therefore not real. Not just the LITU concept we use with Mesozoic dinosaurs, but also the capable of interbreeding concept, since two populations that are separated and will lead to two distinct taxa in the future could still interbreed for a while. While it’s true that it’s subjective which level you place any genus at, the same is true for species. Is there just Microraptor zhaoianus, or are there M. zhaoianus, M. gui, M. hanqingi and Cryptovolans pauli? Or for a modern example, are the different giraffe populations separate species or not?

    You could choose an objective definition for any taxonomic level (species, genus, order) with living members by using interbreeding ability, percent of genetic difference or (more problematically) divergence time, but these will not work for most fossils, so any attempted measure of diversity like Lagomarcino and Miller is going to be subjective.

    • (I forgot to note which pose this was in reply to, so I am moving this back over)

      I am drawing at least one objective line in the sand to prevent myself fault: that there are explicit species definitions to fall back on; there are no “genus” definitions which are not themselves based on species that are at any point qualified. To avoid the issues that come from rank-based taxa, and ranks themselves, I am eschewing ranks entirely. This leaves the concept of the unique biological entity, in which for extant taxa the biological species concept seems apt; while for fossil “species” the morphological species concept (with quantifications) seems apt. These are ideas, though. There has been a larger movement to qualify species, and virtually none for “genera,” a reflection of a concensus that there is some form of biological aspect to “species,” even when they represent lineage segments. This becomes my “line in the sand.”

  3. Your reply (left on the next post) was-

    “I am drawing at least one objective line in the sand to prevent myself fault: that there are explicit species definitions to fall back on; there are no “genus” definitions which are not themselves based on species that are at any point qualified. To avoid the issues that come from rank-based taxa, and ranks themselves, I am eschewing ranks entirely. This leaves the concept of the unique biological entity, in which for extant taxa the biological species concept seems apt; while for fossil “species” the morphological species concept (with quantifications) seems apt. These are ideas, though. There has been a larger movement to qualify species, and virtually none for “genera,” a reflection of a concensus that there is some form of biological aspect to “species,” even when they represent lineage segments. This becomes my “line in the sand.””

    It’s simply not true that no one has proposed genus definitions which are not themselves based on a species definition. For an obvious example, Sibley and Alquist’s (1990) tome of bird phylogeny proposed any two individuals with a Delta T(50)H value of 2.2 or less were in the same genus, with that value referring to the temperature at which half the hybridizable DNA is still hybridized. They had equivalent definitions for other Linnaean ranks.

    I just told you how the biological species concept is NOT always a unique biological entity- “two populations that are separated and will lead to two distinct taxa in the future could still interbreed for a while”. You say it can still have a “form of biological aspect”, but what does that even mean?

    If your line in the sand is that more people have tried to define the species than have tried to define the genus, that’s a pretty meaningless line. Just because more people have tried to make one concept objective does not mean you’re right to treat that one as objective and another as subjective.

    • Very well, there is an explicit attempt to justify “genus.” I will take that as an example which has also never been supported in other analyses. You will admit that “you can put lipstick on a pig, but it’s still a pig” will apply in this case to Sibley and Ahlquist, and I will even agree that it will apply to “species.” Problematically, there is a qualifiable point at which allopatric populations cease capability of breeding, and this is, so far, one of the strongest and most grounded criteria for explicating two groups into a reasonable case of taxonomy. Hence, the biological species definition. It matters not, then, how this taxonomy was recognized in the past, or should be in the future — the taxonomy should reflect some attempt at finding scientific grounds. The alternative is either let current practices continue (whatever the heck we want, let’s have semantic debates, yada yada etc.) or go further and just name EVERYTHING; or in the opposite direct, and dump all taxonomy PERIOD (organisms are known only by their individual specimen numbers/DNA barcodes). The necessity of quantifying organisms as we recognize by our vulgar nomenclature through to formal nomenclature (crossing language barriers), and the need to use shorthand in active communication (no reference sheets, scanners, etc., but words, in the best way recognizable) means we must accommodate the science in this process. That is, the nomenclature has to be as Science-y as possible. This means we are not permitted to simply pick and choose at leisure which idea for species or higher groups, but must apply tests for reasonable recognition, meaningfulness, and replicability. Obviously, for most of these, this is not necessarily possible.

      Phylogenetic nomenclature solves these problems in large part by turning everything to algebra, but we do this by treating everything as a clade. Where the fineness of our cladistic splitting becomes unusable for nomenclature for things like striped chipmunks is that we do not have to go down the fine-tune path to understand how possible-interbreeding populations chose not to breed (behavioral separation), cannot breed (geographic, morphologic), but merely through recognition of the fact that they don’t breed. It is only by historic fact that a species is subdivided than a multitude of species exists when it comes to things like squirrels on either side of the Grand Canyon, or refugia-protected animals separated by deforestation or whatever.

      My “line in the sand” is making an objective stand (I never said otherwise) while at the same time making the LEAST objective decision regarding taxonomy in general. Recognize a single quality (morphological, genetic or behavioral) which segregates two things, apply tests, and when the two are not compatible, sort them into names that best fit them; if they should be equivalent to another taxon but do not fit inside it, name it.

      • So you agree even species are subjectively denoted, but I would argue your choice is not the least subjective possible. The biological species definition is terrible for fossils and not followed for the majority of living species either. How many species have been tested to see if they can produce viable offspring with other related species? Less than 0.001% I’m sure. You also say that the fact two populations don’t interbreed is what matters, not why they don’t, but that leaves us with ridiculous examples like introduced populations being separate species.

        The slippery slope argument is never a good one, and your example is no exception. We can choose to have some rules for defining “genus” (must be monophyletic, must not be inside another genus, must not include -idae, -inae or -ini clades, etc.) and yet still recognize that within that framework it’s completely subjective at what level we place each genus. It’s not science. Genera aren’t equivalent to each other. That’s okay. Notice how few others care. No one else is still harping on the horror of someone being able to propose brancai is in a new genus while leaving Psittacosaurus to include ten species.

        • I would pick the phenetic morphological species concept, because it is the easiest way to make fossils and extant taxa compatible, but this isn’t really viable in my view: As I noted in a previous response, there is more than one thing that prevent species from interbreeding:
          genetically incompatible (sex is possible, may occur under circumstances, but foetuses cannot develop or come to term, or progeny is sterile),
          morphologically incompatible (organisms cannot, actually, breed, so that either the egg cannot be fertilized by the sperm, or the male cannot actually procreate with said female — such as “lock in key” genitalia),
          behaviorally/chemically incompatible (animals do not recognize others as compatible, regardless of genetic capability), and
          geographically incompatible (the organisms are not in the same area, thus do not encounter one another).

          So, I choose to differentiate species concepts. We already use genetic concepts to of late help define our extant species (through phylogenetic analysis), bolsterted by morphology, behavior, geography, and behavior, where available. This is the BEST way. But it cannot be used for fossils: We have incomplete data, so we can use only ONE of these (morphology) when the quality of our data is so incomplete. If we had a high-level of data resolution, we can use geography and potentially diet, but these cannot be used for EVERY thing we find, so we do not make it mandatory. I mentioned on this blog a long way back methodology I can use to tell me the likelihood of species referrability, and of these morphology is the most informative, sorted by material, and distributed by geography and stratigraphy. This seems to be a way to help inform me the quality of a morphological species. If it is distinct, therefore, I name it (or recognize it) … as a binomen. As a species.

          I can only suggest finer gradients of taxonomic utility to be useless, and formulative of a worse slippery slope: the sliding degrees of “subspecies,” “races,” and so forth: Shall we taxonomically differentiate populations of humans because there IS genetic variance? You plop your taxonomic nomenclature boundary at a point that causes the least problems, and this (for me) is that you will largely ignore extant taxa. (As I said.)

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s