Proposal (1051) to South American Classification Committee

 

 

Merge Ixobrychus into Botaurus

 

 

Note from Remsen: With some minor modifications for SACC, this proposal is relayed from NACC, which passed it; NACC now treats all Ixobrychus in Botaurus.

 

Background: The SACC list includes three species of Ixobrychus: Least Bittern I. exilis and Stripe-backed Bittern I. involucris. Molecular studies have consistently found I. exilis to be more closely related to Botaurus than to most other species of Ixobrychus. In the DNA-DNA hybridization tree of Sheldon (1987a), for example, it was sister to American Bittern B. lentiginosus rather than to the other two species of Ixobrychus (I. minutus and I. cinnamomeus) included in the study. Moreover, in the DNA barcode paper of Päckert et al. (2014), who sampled seven species of Ixobrychus and three of Botaurus, I. exilis was sister to the three species of Botaurus (with 1.0 p.p.) rather than to any of the other six species of Ixobrychus:

 

 

 

Fig. 1. Mitochondrial barcode tree from Päckert et al. (2014) showing the sister relationship between Ixobrychus exilis and the three species of Botaurus.

 

New information: In the UCE tree (Fig. 2 below) from Hruska et al. (2023), I. exilis was strongly supported (100% bootstrap) as sister to the two species of Botaurus sampled for UCEs (B. lentiginosus and B. poiciloptilus); this clade was sister to another species traditionally placed in Ixobrychus (I. involucris), also with 100% bootstrap support. Thus, the two New World species of Ixobrychus were strongly supported as more closely related to Botaurus than to the Old World species of Ixobrychus.

 

 

Fig. 2. Phylogenetic tree based on sequences of UCEs (Hruska et al. 2023). All nodes received 100% bootstrap support except as noted.

 

In their Bayesian mtDNA tree (Fig. 3 below), these species were reasonably well supported (0.99 p.p. for both species) as more closely related to Botaurus lentiginosus than to other species of Ixobrychus, although this section of their ML mtDNA tree was unresolved.

 

 

Fig. 3. Phylogenetic tree based on Bayesian analysis of sequences of mtDNA (Hruska et al. 2023). All nodes received 100% bootstrap support except as noted.

 

Based on these results, Hruska et al. (2023) recommended that both I. exilis and I. involucris be transferred to Botaurus. An alternative to this would be to place I. exilis into a separate monotypic genus, which would presumably also necessitate placing I. involucris in a monotypic genus, but names may not be available for these genera.

 

Another alternative, given that some of the small bitterns of Ixobrychus are more closely related to Botaurus, thus blurring the morphological lines between the traditional limits of these genera, would be to merge all species of these two genera into a single genus. This would avoid placing exilis in a separate genus from the extremely similar minutus and sinensis, all previously thought to constitute a superspecies. There is also a fair amount of plumage and size variation among species of Ixobrychus, such that I. flavicollis, for example, overlaps some Botaurus in size. Although the nodes uniting Ixobrychus and Botaurus in the UCE and mtDNA trees above are deeper than those uniting other single genera, Hruska et al. (2023) noted that rates of sequence evolution in bitterns (both Ixobrychus and Botaurus) were faster than those in other herons, supporting the previous finding of Sheldon (1987b) and indicating that branch lengths in the Botaurinae are likely inflated. Although it seems that an expanded Botaurus (Botaurus Stephens, 1819, has priority over Ixobrychus Billberg, 1828) would be heterogeneous both phylogenetically and phenotypically, the branch lengths would be at least somewhat offset by the faster rates of evolution and an expanded Botaurus would be little more phenotypically heterogeneous than a Botaurus expanded to simply include exilis (and involucris).

 

An additional consideration involved in the option of simply transferring exilis to Botaurus, is that minutus, the type species of Ixobrychus, was not sampled in the UCE study of Hruska et al. (2023), introducing some uncertainty as to whether minutus is part of the clade of species not being transferred to Botaurus (i.e., whether minutus is truly not closely related to exilis/Botaurus). The data regarding the relationship of minutus to exilis-Botaurus are suggestive but not definitive. In the DNA-DNA hybridization tree of Sheldon (1987a), I. minutus and I. cinnamomeus were not part of the I. exilis-Botaurus lentiginosus clade, although the relationships of minutus and cinnamomeus were unresolved. In the mtDNA barcode tree (Päckert et al. 2014) the 11 samples of I. minutus formed a clade (with 1.0 pp) sister to the seven samples of I. sinensis, which was sampled by Hruska et al. (2023) and was deeply embedded in the clade not being transferred. Päckert et al. (2014) also found exilis to be sister to the three species of Botaurus sampled, again with 1.0 pp. Thus, there’s reasonably strong support for minutus not being closely related to exilis-Botaurus in the mitochondrial data, and some support in the nuclear data. In my view, this would be a more serious issue if we were proposing to make substantial changes to Ixobrychus, such as splitting it into two or more genera, rather than simply transferring one species to Botaurus. However, the majority of species currently in Ixobrychus, including type species minutus, would simply be retained in Ixobrychus, undergoing no change of status under this option.

 

For this issue, votes are needed as follows:

 

1. Revise the taxonomy of the bitterns, YES or NO

 

If YES on #1, then vote for either 2A or 2B (unless you favor a different option, which should be specified):

 

2A. Transfer I. exilis to Botaurus, retaining I. sinensis and I. minutus in Ixobrychus

2B. Subsume Ixobrychus into Botaurus, transferring all three NACC species

 

(Note from Remsen: see SACC voting structure at end of proposal)

 

Recommendation: Numerous molecular studies indicate that the taxonomy of the bitterns requires revision, so I strongly recommend a YES vote on Part 1. Regarding Part 2, this is a tough call, with good arguments for and against the proposed treatments. I slightly favor transferring only I. exilis to Botaurus, at least for now, because (1) it is the least disruptive option; (2) it limits the genus transfers to a species endemic to the Americas, rather than also involving two Old World accidentals (as would happen if all NACC species were transferred to Botaurus); and (3) this option seems more consistent with the depth of nodes in the phylogenetic trees, even adjusting for a more rapid evolutionary rate among the bitterns. I view point 2 above as a major obstacle to subsuming Ixobrychus into Botaurus at this time. Should a global body such as WGAC endorse the merger of Ixobrychus into Botaurus, then point 2 above becomes moot and the balance is likely tipped towards transferring all species of Ixobrychus to Botaurus. In that case we could re-vote, but for now I recommend the transfer of exilis only.

 

 

Terry Chesser, May 2025

 

 

Note from Remsen on SACC voting structure: Note that NACC ended up expanding Botaurus to include all of Ixobrychus.  So, let’s just simplify this.  A YES vote on the SACC proposal means you favor merging Ixobrychus into Botaurus.  A NO vote means either (1) you favor transferring the two New World Ixobrychus into Botaurus and thus restricting Ixobrychus to Old World taxa (or some other solution) or (2) waiting for another, independent data set given the surprising results in the genetic studies so far, and one that would also include sampling of the type species of Ixobrychus (minutus) in the UCE study.

 

 

 

Voting Chart: https://www.museum.lsu.edu/~Remsen/SACCPropChart1044+.htm

 

Comments from Remsen: “YES.  Note that the potential for confusion when reading some of this is high here because the original paper did not even mention the obvious alternative hypothesis, i.e. subsuming Ixobrychus into Botaurus; this was added by NACC as a possibility (as Option 2B above) even though there is no mention of this obvious alternative in the paper on which the proposal is based.

 

“Also, note that there is a technicality on which one could dismiss the results entirely, namely, the type species of Ixobrychus is minutus (Ardea minuta Linnaeus 1766).  But minutus was not sampled by Hruska et al.  The only mention of minutus in their text is its initial mention as a member of traditional Ixobrychus and its inclusion in Ixobrychus sensu lato in the classification table.  Step 1 of any valid taxonomic revision is analysis of the type taxa.  Its absence in the analysis of the type species for the genus could be sufficient grounds for voting NO on the proposal, with a statement along the lines of …  How can we revise the limits of Ixobrychus without its type species being included in the analysis?  A homologous problem in all this is that the type species of Botaurus is stellaris (Ardea stellaris Linnaeus, 1840), so how can we revise the limits of that genus without sampling its type species also?

 

“Huang et al. (2016) sampled minutus but not exilis.  Päckert et al. sampled minutus, sinensis, and exilis, but not dubius.  So, the only data relevant to the relationship between exilis and the type species of Ixobrychus is Päckert et al.’s mtDNA barcode tree, which indeed found exilis as sister to Botaurus, not minutus.  But that same tree conflicts with Hruska et al. in placing cinnamomea and eurhythmus as sister to everything else, including minutus and flavicollis, the latter sometimes placed in the monotypic genus Zupetor.  However, I suppose it is safe to proceed based on the reasonably safe assumption that Ixobrychus minutus is not sister to all other bitterns, as we did at NACC.

 

“Here’s the problem with maintaining Ixobrychus for the Old World genera.  The four traditional Ixobrychus species exilis, minutus, dubius, and sinensis are so similar phenotypically that they have always been considered part of a species complex.  Mayr & Short (1970) officially designated them as a superspecies (parapatric replacements of monophyletic lineage).  This was followed subsequently by basically all authorities to one degree or another.  This is not to say that Mayr & Short didn’t make mistakes in their superspecies designations, but they usually got it right (because recently diverged species typically remain morphologically similar and parapatric or allopatric), and in this case no one that I know of familiar with the group would fault them for having made this designation based on their parapatric replacement pattern and similar plumage and morphology, despite Hruska et al.’s subsequent falsification of it being a monophyletic group (but without having sampled minutus).

 

To illustrate this point, b­­elow is a photo of F. Jutglar’s excellent illustration in HBW Vol. 1 (with taxon names added by me).

 

 

“Note the extreme similarity between our I. exilis and the Australian taxon dubius (treated as a subspecies of minutus or as its own species depending on classification).  They are so similar that I suspect without doing some advance studying, all of us would misidentify dubius in the field as our familiar exilis (and on a just a quick view would likely pass off male minutus as exilis also).  Note that Hruska et al.’s three sentence discussion of the issue includes no mention of this extreme phenotypic similarity, yet they want to put these remarkably similar taxa in different genera.

 

“I can think of no analogous case in birds. Of course, DNA has revealed some big surprises that have violated our world view of the value of plumage as a phylogenetic character.  But this is different.  Exilis, minutus, dubius, and to some degree sinensis are among the most ornamented of herons, with a suite of shared plumage pattern characters not found elsewhere in the group or even the family.  If they were a cluster of pattern-less dull birds, then finding that they weren’t each other’s closest relatives might just have been hidden by an absence of characters – no big deal.  But this is the opposite, at least within heron phenotypes.  Add to this their parapatric replacement pattern and perhaps even vocal similarities (see Jon’s comments), and placement in separate genera, when a reasonable phylogenetic alternative that maintains them as congeners is available, will likely draw eyerolls, head shakes, and snickers from many ornithologists.

 

“I can think of a recent case in which we expanded one genus to include another despite morphological discontinuity: Catoptrophorus into Tringa.  The Willet really differs phenotypically from typical Tringa, which are fairly homogeneous, but when faced with the genetic results, we and NACC expanded Tringa to include morphologically disparate Catoptrophorus.

 

“The proposal made the point that broad Botaurus would be too heterogeneous phenotypically.  Here’s my counter-argument.  Structurally, one could make a case that Botaurus look superficially like giant juvenal-plumaged Ixobrychus.  Check out Jutglar’s plate above.  One could also make a case that involucris is intermediate between Ixobrychus and Botaurus in plumage (and thus no surprise that it is on the branch with Botaurus sensu stricto, as the recent genetic data suggest).  From the illustrations, female cinnamomeus looks like a miniature Botaurus, as does female eurhythmus … in my opinion.  The oddball flavicollis closes the size gap to some extent between big Botaurus and small Ixobrychus; body weights in HBW (1992, v. 1) show overlap between flavicollis and the low end of our American Bittern.  My point is that once you look at the diversity of bitterns, beyond just our familiar Least and American, a single genus becomes slightly more palatable, at least in my opinion, certainly much more so than placing phenotypically similar allotaxa like exilis and minutus in different genera.  Placing those two in separate genera would make those two genera morphologically undiagnosable.

 

“As noted above, Hruska et al. did not sample B. stellaris.  If exilis and minutus are not closely related, then it is unsafe to assume that stellaris has to be closely related to the two Botaurus sampled.  Subjectively, the exilis and minutus look more similar to one another to me than do lentiginosus or poiciloptilus and stellaris.

 

“I would change my mind (as long as minutus and stellaris are also included in the analysis) if a time-calibrated tree indicated that the split between Hruska et al.’s broad Botaurus and constricted Ixobrychus was ancient, e.g. early Miocene as in many groups labeled as genera in birds.  As is, I’m ok with the admittedly painful solution of a broad Botaurus.”