The Battle over Anole Classification Ignores the War

Should Anolis be split into several genera, and why is this is the wrong question? The battle over anole classification is not about splitting Anolis into several genera; it is about changing the content of a well-understood taxon, by pointing the name Anolis to a different branch or node of the tree. The war, then, is about the failure to connect taxonomy to phylogeny in an evolutionarily meaningful way, which is that taxon names should be associated with evolutionary lineages (clades) and not with ranks. If one accepts this, then it is rarely necessary to change the association of a name with its taxon, as proposed by Nicholson et al. (2012) in the case of Anolis.

Below I respond to several misconceptions about taxonomy, some with reference to anoles. I am not claiming that Nicholson et al. (2012) espouse these explicitly, but they are germane to anole taxonomy.

Misconception 1. Taxonomic stability is ignorance. Put another way, stability in taxonomy is not necessary, or even desirable. In contrast, I argue that stability should be a basic characteristic of taxonomy.

Taxonomies become unstable when the association between a name and its taxon changes, i.e., when the name points to a different taxon. However, stability does not mean that taxonomies do not change at all. Stable taxonomies can change, that is, improve, by adding more information about hierarchy. That is, as new nodes are discovered, names are progressively applied to those nodes. The existing associations between names and taxa need not change.

Misconception 2. Taxonomies are primarily for systematists. Unfortunately, some systematists view taxonomy as a personal sandbox. Rather, taxonomies are reference systems that are fundamentally important to the community of non-systematists. If not conservative, taxonomies are confusing for those who need stable reference lists. Witness the controversy about Bufo, Rana, etc.

Misconception 3. Some people don’t like change.  Two types of change are at issue: (a) change in taxonomy, and (b) change in the practice of taxonomy. We who prefer a conservative taxonomy that maintains name-taxon stability are considered old-fashioned. Those who prefer taxonomy that breaks name-taxon stability, as has been proposed for anoles, are often considered progressive (see Misconception 1), under the assumption that any change is progress.

Ironically, a stable, “conservative” taxonomy requires a radical change in mindset about how taxonomy is done. Simply put, one maintains the association between name and clade, and applies new names when as needed to newly uncovered taxa. This approach reflects a growing understanding of the relationship between taxonomy and phylogeny. de Queiroz (1988) called attention more than 20 years ago to the failure of taxonomists to integrate taxonomy into the Darwinian Revolution.

A focus on ranks—arguing that eight genera of anoles are preferable to one—is inherently non-evolutionary. Thus, those who prefer to split a ranked taxon into several of equal rank are the resistors of change.

Misconception 4. Changes in stability between name and taxon are inevitable, especially in cases of paraphyly. De-stabilizing changes are not inevitable, and only result if one places primacy on rank-based taxonomy rather than taxonomy based on ancestor-descendant (evolutionary) relationships. N. B., I am not advocating that ranks should not be used, only that the emphasis on ranks is the cause of the controversy.

Elimination of paraphyly was the battle-cry of early cladists, but in reality the arguments about paraphyly were a distraction from the real issue. The Reptilia-Aves controversy was fundamentally about ranks, not paraphyly. Should Reptilia and Aves both be ranked as classes? If yes, then Reptilia is paraphyletic, because paraphyly follows from the use of ranks. The solution to the controversy was acknowledgement that Aves is nested within Reptilia, giving primacy of phylogeny over ranks. As Neil Shubin articulated, we all have an Inner Fish.

Paraphyly has consistently been a motivation for dismantling Anolis beginning with Guyer and Savage (1986). However, that the genus Norops (for example) is nested within the genus Anolis does not require splitting Anolis into several genera. One simple solution is to treat Norops as a subgenus within Anolis. The name of the species sagrei can then be written as Anolis (Norops) sagrei. The elegance of this is that Norops and Anolis, as nested taxon names, continue to refer to their traditional clades.

A second, more general solution is to use multiple levels of unranked clade names as done by Castañeda and de Queiroz (2013). They recognized as formal unranked taxa the clade Dactyloa; and within Dactyloa, clade Megaloa for the latifrons series, and clade Phenacosaurus for the heterodermus series. Because these are expicitly used as unranked names, they are not regulated by the International Code of Zoological Nomenclature (the Code).

As Cannatella and de Queiroz (1989:68) responded to Guyer and Savage (1986): “A phylogenetic taxonomy could have been effected by reorganizing sections, subsections, and series within Anolis, without generic level re-arrangements.”

Misconception 5. Subgenera are not used much in herpetology. Even if this were true, it is not a reason to reject the use of subgenera. Regardless, the data don’t support this claim; the use of subgenera is rising. They are a very useful tool, but have constraints imposed by the Code (these can be easily fixed).

Misconception 6. The most recent classification must be used as the standard. To recognize this fallacy one need only read the first Principle of the Code, which embraces taxonomic freedom. A common question from the community-at-large is, Which classification is the “correct” one? The answer is of course that there is no “correct” classification, and taxonomists who claim this do a disservice to the general community.

It is, in fact, time for a new classification of anoles, but one that truly integrates evolutionary principles with taxonomy, reflecting progress and not just change.

Acknowledgements. This essay is strongly influenced by Kevin de Queiroz, who articulated many of these ideas >25 years ago. David Wake engaged in helpful discussions.

References

Cannatella, D. C., and K. de Queiroz. 1989. Phylogenetic systematics of the anoles: is a new taxonomy warranted? Syst. Zool. 38:57-69.

de Queiroz, K. 1988. Systematics and the Darwinian revolution. Phil. Sci. 55:238-259.

del Rosario Castañeda, M., and K. de Queiroz. 2013. Phylogeny of the Dactyloa clade of Anolis lizards: New insights from combining morphological and molecular data. Bull. Mus. Comp. Zool. 160:345-398.

Guyer, C., and J. M. Savage. 1986. Cladistic relationships among anoles (Sauria: Iguanidae). Syst. Zool. 35:509-531.

Nicholson, K. E., Crother, B. I., Guyer, C., and J. M. Savage. 2012. It is time for a new classification of anoles (Squamata: Dactyloidae). Zootaxa 3477:1–108.

11 thoughts on “The Battle over Anole Classification Ignores the War

  1. I believe this post also contains several misconceptions, the most severe of which is:

    Misconception 1. A taxon remains the same even as its content changes. As an example, a Reptilia that includes Aves and a Reptilia that does not are different mappings of the name “Reptilia” onto taxa. Keeping the name the same does not ensure taxonomic stability–changes in content of taxa are also a source of taxonomic instability! Further, Reptilia is paraphyletic if Aves is excluded from it regardless of ranks. Paraphyly is not a result of ranks, but of the mapping of the content of a taxon onto a phylogenetic tree.

    This problem is also apparent in the following:

    Taxonomies become unstable when the association between a name and its taxon changes, i.e., when the name points to a different taxon. However, stability does not mean that taxonomies do not change at all. Stable taxonomies can change, that is, improve, by adding more information about hierarchy. That is, as new nodes are discovered, names are progressively applied to those nodes. The existing associations between names and taxa need not change.

    What happens when our understanding of phylogenetic relationships changes? Suppose, as an extremely unlikely hypothetical scenario, that we discovered Aspidoscelis to be nested within the clade of Anolis. Well, we have two options: either the names persist unchanged (and Anolis is paraphyletic) or we change the association between names and taxa (e.g., by changing the content of Anolis so that it includes Aspidoscelis). So far as I can tell, it is Dr. Cannatella’s position that the latter does not constitute any change in the association between taxa and names–that “Anolis” means precisely the same thing regardless of what species are included within it. I believe this is incoherent. If whiptails were to become anoles, that would be a major taxonomic change regardless of whether or not any binomials were changed as a result. Interestingly, changing the binomials would not result in taxonomic instability as defined here by Dr. Cannatella, because synonymy preserves “the association between a name and its taxon”. Aspidoscelis exsanguis, as a synonym of Anolis exsanguis, would still denote the same species.

    Unless we endorse paraphyletic taxa, taxonomic change cannot be limited only to the naming of new taxa. We cannot simply move taxa among named clades and pretend that no change has occurred! Some degree of taxonomic instability is necessary. Other sources of taxonomic instability, however, are optional and should be avoided (e.g., dividing a monophyletic genus into more monophyletic genera).

    1. Hi Patrick.

      I think this issue becomes resolved if names are assigned to nodes rather than nested taxonomic ranks. For instance, if the name Reptilia is defined as “the clade descended from the node representing the MRCA of crocodylians, squamates, turtles, and tuatara”; and, furthermore, Aves is defined as “the clade descended from the MRCA of all birds”; then both names can stably survive the discovery that Aves is nested within Reptilia, with their original meaning (if not ranks) intact.

      In your case of Aspidoscelis nested within Anolis, this could be resolved by ‘giving phylogeny primacy over rank’ and thus recognizing Aspidoscelis as a named subclade of Anolis. Aspidoscelis retains its original meaning as the clade descended from the MRCA of all whiptails, whereas Anolis retains its original meaning as the clade descended from the MRCA of all anoles, etc. Whether we assign the rank ‘genus’ to Anolis or Aspidoscelis is, of course, arbitrary. There are circumstances in which nomenclatural incongruencies cannot be resolved in this way – for instance, when phylogenetic information tells us that two names (of the same or different rank) are associated with the same node. If they are of different rank, this is not a huge issue – we just have monotypic genera or monogeneric families (etc.). If they are the same rank, then we have a genuine taxonomic incongruity to address.

      – Liam

      1. Hello Liam,

        I do not agree that an Anolis that includes Aspidoscelis “retains its original meaning”. “The clade descended from the MRCA of all anoles” is not a taxon, it is a rule for determing what taxon would be denoted by a name given a particular tree. Without a reference tree, the meaning of the name is ambiguous. As the reference tree changes, the taxon indicated by the name changes. The rule may stay the same while the name refers to different taxa (within whatever bounds are set by the rule) in reference to different trees.

        IMO, this is a way of forcing taxonomic instability to follow rules. On the plus side, it would eliminate the ability for researchers to create arbitrary taxonomic instability. On the minus side, it mandates taxonomic instability whenever the placement of the species included in the rule associated with a name changes. Whether the balance comes out in favor of this way of defining taxa or not depends on how good our phylogenetic inferences are when the rules defining the application of names to taxa are created. At the extremes, if our phylogenetic inferences are very good, the taxon denoted by a name will change little if at all; however, if our phylogenetic inferences are dramatically wrong, the taxon denoted by a name may vary wildly without any ability for taxonomists to exercise judgment and reduce the it.

        Personally, I think the best approach is simply to keep the current nomenclatural system and for the taxonomic community to unite in rejecting needless changes. As a community, we can reject needless changes under the ICZN / ICNafp as is; no new system is needed, we just ought to do it. However, we ought to allow some discretion. When our understanding of phylogeny changes dramatically, we ought to allow (and encourage!) taxonomists to modify names in the way that results in the least disruption, rather than embracing a rule-bound system that may result in massive instability that we know is not in the best interests of communication. Imagine, for instance, that Oxybelis were defined a couple decades ago so that Oxybelis argenteus were one of the defining species. What, exactly, would the content of Oxybelis be now? Would that name have anything remotely approximating its original meaning? Moving a small number of species out of Oxybelis seems quite an improvement over expanding Oxybelis beyond all recognition!

  2. This exchange with Patrick brings to light a misconception that I failed to point out; call it Misconception 0: The name of a taxon is defined by its content. In a phylogenetic (evolutionary) system, the name of a clade is not defined by the species it contains, but rather by the node to which it points (is associated with).

    Note the distinction between a taxon and its name. A taxon is just a group of organisms; it need not have a name. Names refer to taxa; under a phylogenetic system, names refer to clades, since it’s commonly accepted that taxa should be clades.

    Patrick wrote:

    As an example, a Reptilia that includes Aves and a Reptilia that does not are different mappings of the name “Reptilia” onto taxa. Keeping the name the same does not ensure taxonomic stability–changes in content of taxa are also a source of taxonomic instability! Further, Reptilia is paraphyletic if Aves is excluded from it regardless of ranks. Paraphyly is not a result of ranks, but of the mapping of the content of a taxon onto a phylogenetic tree.

    You’re correct when you say that “…a Reptilia that includes Aves and a Reptilia that does not are different mappings of the name ‘Reptilia’ onto taxa.” But the issue is not the mapping, but rather the association of a name with a node.

    Mapping, as Patrick seems to use it, is not the association of a name with a node, but simply the assignment of taxa to Reptilia, i.e., specifying the content of Reptilia.

    Regarding Patrick’s example of Aspidoscelis:
    Let’s consider the biological implication of Aspidoscelis being nested within Anolis. It would mean that Aspidoscelis is a highly derived anole. Conversely, it would also mean that the “teiid” morph is convergent in Anolis and in Teioidea.

    Aspidoscelis, whether highly derived or not, is a descendant of the last common ancestor of the species included in Anolis, so Aspidoscelis is part of (the clade) Anolis, just as Aves is part of (the clade) Reptilia.

    Patrick wrote that in this situation there are two options:

    Option 1: “either the names persist unchanged (and Anolis is paraphyletic)”

    Referring again to Misconception 0, it is not the name that persists unchanged. It is the association of the name with its clade–the “name-taxon stability” that I referred to. Anolis is paraphyletic only if one insists that Anolis must be a genus and Aspidoscelis must be a genus. As I remarked in my post, paraphyly would not result if there were no ranks.

    In fact, in the new tree the association of the names Anolis and Aspidoscelis with their clades would stable, and more importantly, this stability would point out an amazing instance of evolution.

    Patrick wrote further:

    Option 2: “or we change the association between names and taxa (e.g., by changing the content of Anolis>/em> that it includes Aspidoscelis). As far as I can tell, it is Dr. Cannatella’s position that the latter does not constitute any change in the association between taxa and names–that “Anolis” means precisely the same thing regardless of what species are included within it.”

    This reflects the same issue: we have not changed the association between a name and its taxon. The name Anolis still points to the same node, and the name Aspidoscelis still points to the same node. A clade is not defined by reference to its content, but by reference to its ancestor.

    So yes, given that Anolis points to the same node, Anolis does mean precisely the same thing, “regardless of what taxa are included in it.”

    If Aspidoscelis were found to be nested within Anolis and it became accepted that Aspidoscelis is an anole, it would be almost as important a change in thinking as accepting that birds are dinosaurs and would give Jonathan reason to write another book!

    As an aside, breaking up of Cnemidophorus into Aspidoscelis, etc. by Reeder et al. (2002) was as unnecessary as breaking up Anolis. In their phylogeny, Cnemidophorus was simply paraphyletic with respect to Ameiva, Kentropyx, etc. Thus, Ameiva are simply highly modified “cnemis”.

    Last, when I wrote that “…stability should be a basic characteristic of taxonomy” and “The existing associations between names and taxa need not change.” I didn’t mean that the association between a name and its clade will never change. Rather, if we re-orient our thinking toward a truly phylogenetic framework, one that goes beyond “taxa must be monophyletic,” we’ll realize that taxonomic instability is neither inevitable nor a sign of progress.

    Dave

  3. Hello David,

    In replying to this, I think I can stop at:

    “Misconception 0: The name of a taxon is defined by its content.”

    That is the crux of the biscuit. Either one thinks of taxa in terms of what actual organisms are included within them, or one does not. Adopting “content-free” taxa does simplify some things. However, a set of names that is not aimed first and foremost at providing a means of referring to organisms is not taxonomy. To the extent that we divorce names from organisms, we fail to provide a usable and meaningful taxonomy. Without content, a “clade” is nothing and a “taxon” likewise. We cannot separate the two without losing sight of the purpose of taxonomy–names that refer to organisms. The organisms referred to–the content–are not optional. They are the point of the whole exercise.

    A node is, of itself, a useless hypothetical entity. It becomes real and useful only to the extent that it can be used to mark a particular set of living things. Taxonomy starts and ends with actual organisms, not trees.

    1. Nor, for that matter, hypothetical common ancestors. There must have been a common ancestor to all extant Anolis, certainly. However, if you’re going to hang taxonomy on that ancestor you need to be able, at the very least, to point at it! Show me that ancestor, determine its characteristics, establish firmly its place in the phylogeny, and then we can have something to work from.

      A taxonomy based on hypothetical entities entirely misses the point. We don’t need names for hypothetical organisms, we need names for actual, observed critters.

  4. A practical question:

    I would like to name the clade of species I am working with in a paper.

    Following the debate so far I would not mind using the name “Audantia, a subgenus within Anolis”.
    Is there any consensus about this can be implemented in a citation? Do the “subgenera” have to be formally published as such first?
    Or would I have to write “within the genus Anolis, a clade that has recently been renamed as genus Audantia (Nicholson et al., 2014)”.
    Alternatively, would “Anolis, clade Audantia” make sense?

    just wondering.

    1. [I’m going to omit italics here because adding the HTML to display it is cumbersome.]

      When a new genus is described (created), its corresponding subgenus is created implicitly, and has the same author and date. Even if no one has ever used Audantia as a subgenus, it can be used at that rank with no special nomenclatural action.

      So you can refer to Anolis (Audantia) as in “This paper is a taxonomic revision of Anolis (Audantia).” with no other qualification.

      Alternatively, you could also just refer to “clade Audantia” if there’s no particular reason to use the subgenus rank.

      Note that it is (too) common in the literature to see Anolis (Audantia) used when the intent is to indicate that Audantia is a junior synonym of Anolis. This is incorrect usage, according to the Code. Even Anolis (=Audantia) can be ambiguous.

      Dave

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Optionally add an image (JPEG only)