Proposal (829) to South American Classification Committee

Merge Oceanodroma into Hydrobates

Note from Remsen (May 2019): This proposal is a spinoff of Shawn Billerman’s proposal to NACC, which was passed unanimously. The proposal is self-explanatory. The taxa that would be affected in the SACC classification are as follows:

Oceanodroma microsoma Least Storm-Petrel (NB)

Oceanodroma tethys Wedge-rumped Storm-Petrel

Oceanodroma castro Band-rumped Storm-Petrel

Oceanodroma leucorhoa Leach's Storm-Petrel (NB)

Oceanodroma markhami Markham's Storm-Petrel

Oceanodroma hornbyi Ringed Storm-Petrel

Oceanodroma melania Black Storm-Petrel (NB)

_________________________________________________________________________-

Merge the storm-petrel genus Oceanodroma into Hydrobates

Background and New Information:

The northern storm-petrels, Hydrobatidae, are currently placed into two genera, Hydrobates and Oceanodroma. The genus Hydrobates includes only a single species, the European Storm Petrel (Hydrobates pelagicus), whereas all other species of northern storm-petrel are placed in the genus Oceanodroma. Although there are still relatively few studies that look at the phylogenetic relationships of the storm-petrels, recent work has shown that Oceanodroma is paraphyletic with respect to Hydrobates, with the European Storm-petrel embedded within the larger Oceanodroma (Kennedy and Page 2002, Penhallurick and Wink 2004, Robertson et al. 2011, Wallace et al. 2017). Most studies have found that the European Storm-Petrel is sister to Fork-tailed Storm-Petrel (O. furcata) (Robertson et al. 2011, Wallace et al. 2017; Fig. 1). As a result of the paraphyly of Oceanodroma, most taxonomic authorities (e.g., Dickinson and Remsen 2013) have merged the two genera, with Hydrobates Boie, 1822, having priority over Oceanodroma Reichenbach, 1853.

Figure 1. Bayesian phylogeny (based on sequence data from cytochrome-b and 5 nuclear introns), where ‘*’ indicates posterior probabilities of 1.0 and all posterior probabilities above 0.8 are given. Note that Hydrobates pelagicus is sister to Oceanodroma furcata, which is in turn sister to a clade of New World Oceanodroma. This larger clade is in turn sister to the rest of the Oceanodroma. Adapted from Wallace et al. 2017.

Although the European Storm-Petrel is often found to be sister to Fork-tailed Storm-Petrel, most other relationships within the family are not well resolved, making it difficult to speculate on any well-supported clades within the family (Robertson et al. 2011, Wallace et al. 2017).

Recommendation:

Based on the findings of several recent molecular phylogenies (Penhallurick and Wink 2004, Robertson et al. 2011, Wallace et al. 2017), I recommend merging the genus Oceanodroma with Hydrobates, given that Oceanodroma is paraphyletic with respect to Hydrobates and that Hydrobates has priority. At this time, I propose no change in the linear sequence of the family given the lack of resolution of many relationships. This would result in the following changes to the AOS Checklist:

European Storm-Petrel (Hydrobates pelagicus)

Fork-tailed Storm-Petrel (Hydrobates furcata)

Ringed Storm-Petrel (Hydrobates hornbyi)

Swinhoe’s Storm-Petrel (Hydrobates monorhis)

Leach’s Storm-Petrel (Hydrobates leucorhoa)

Townsend’s Storm-Petrel (Hydrobates socorroensis)

Ainley’s Storm-Petrel (Hydrobates cheimomnestes)

Ashy Storm-Petrel (Hydrobates homochroa)

Band-rumped Storm-Petrel (Hydrobates castro)

Wedge-rumped Storm-Petrel (Hydrobates tethys)

Black Storm-Petrel (Hydrobates melania)

Guadalupe Storm-Petrel (Hydrobates macrodactyla)

Markham’s Storm-Petrel (Hydrobates markhami)

Tristram’s Storm-Petrel (Hydrobates tristrami)

Least Storm-Petrel (Hydrobates microsoma)

References

Kennedy, M., and R.D.M. Page. 2002. Seabird supertrees: combining partial estimates of Procellariiform phylogeny. The Auk, 119: 88-108.

Penhallurick, J., and M. Wink. 2004. Analysis of the taxonomy and nomenclature of the Procellariiformes based on complete nucleotide sequences of the mitochondrial cytochrome b gene. Emu, 104:125-147.

Robertson, B.C., B.M. Stephenson, and S.J. Goldstien. 2011. When rediscovery is not enough: taxonomic uncertainty hinders conservation of a critically endangered bird. Molecular Phylogenetics and Evolution, 61: 949-952 .

Wallace, W.J., J.A. Morris-Pocock, J. González-Solís, P. Quillfeldt, and V.L. Friesen. 2017. A phylogenetic test of sympatric speciation in the Hydrobatinae (Aves: Procellariiformes). Molecular Phylogenetics and Evolution, 107:39-47.

Submitted by: Shawn M. Billerman

Date of Proposal: 3 December 2018

Comments from Jaramillo: NO. This is what I sent NACC on this proposal:

“Dear committee members, I wanted to reach out and make sure you were aware of a potential issue in the data brought forward in the current proposal: Merge the storm-petrel genus Oceanodroma into Hydrobates. The issue has to do with the placement of markhami (Markham’s Storm-Petrel) as sister to, and essentially equivalent to Black Storm-Petrel (melania) in Wallace et al. 2017 (Molecular Phylogenetics and Evolution, 107:39-47). In Table 2 you will see that the divergence between these two species is very small, with “0” being included in the range of divergence possibilities. The problem is that in reality, biologically, and in every way other than both being storm-petrels and all dark, the two are very different. This is an incredibly surprising result!! I am quite sure that the reason for this is that the Peruvian specimen of markhami that is used (noted as soft tissue in Table 1) is actually a non-breeding melania. Oceanodroma melania is very common in Peruvian waters and recently field observers have noted that they can be more common than markhami there. This is recent information that is available due to good digital photos, historically it was assumed that markhami was the “default” large dark storm petrel in those waters. I noted this potential error to Vicki Friesen, she replied with this note:

‘As I’m sure you know, Markham’s storm-petrels are (or were) hard to sample, hence the use of a museum specimen. This specimen came from the Academy of Natural Sciences in Philadelphia (specimen number 11751). We didn’t actually see it, but trusted the expertise of the biologists and curators who collected and preserved it. We would definitely be interested in re-visiting the systematics of this species (and any others represented only toe pads in our study) if new samples could be collected. Thank you for contacting us. – Vicki’

“As it stands it is likely that no genetic information exists for markhami, and its relationships are therefore still unclear. I am willing to bet that it will be in the group that it most closely resembles in flight style, coloration etc, which is with Clade B, the one that includes leucorrhoa. If future proposals consider re-arranging the sequence of the storm petrels, and I think there is ample reason to do so with available data, do take into consideration that markhami is almost certainly not closely related to melania.

“Finally, I ask you to consider that merging the entire family into one genus creates a rather uninformative genus! Some of these nodes are thought to be over 11 mya, and there are some clear and well-defined clades within the family. I would suggest that it would be better to separate them out into 4-5 genera to better represent and segment the diversity within the family. For example, the Halocyptena genus (melania, tethys, microsoma) shares a distribution in the Eastern Pacific, they also have particularly long tarsi for northern storm petrels, as well as very dark chocolate colored plumage lacking the gray tones (when fresh) of other all dark storm-petrels. Note that white in the rump is not an informative character for storm petrels. The pale ulnar bar on these species is restricted to the greater coverts, while on typical Oceanodroma the ulnar bar extends to the bend of the wing and on to the distal median and lesser coverts. The very long legs of melania, was distinctive enough that previously it was classified in its own genus (Loomelania), I think that Halocyptena predates it for the group name, however. I would predict that once other details such as voice can be added to the dataset, the cohesiveness of this genus will become even more clear cut. Similarly, the Band-rumped storm-petrels would make a good genus, one that has a wide distribution in the northern hemisphere and probably contains more species than currently understood. The main clade would remain as Oceanodroma, and I would provisionally include markhami in there. Finally, the more contentious issue would be if one lumps but furcata and pelagicus in one genus (Hydrobates) or a different genus is chosen for furcata. The two are a clade but they are not particularly closely related. In any case, this multi-genus organization would make for a much more informative way to subdivide the family. Creating a one genus family, for a group which is both highly widespread in distribution and holds relatively old lineages does not make much sense.

Comments from Stiles: “NO, as it stands. If branch lengths are any indication of lineage ages, there are two genus-level groupings here: the group from melania through furcata (perhaps excluding markhami as stated by Alvaro), and the “other Oceanodroma” (jabejabe through homochroa). The former group would be Hydrobates. Because the type species of Oceanodroma is furcata, another name would be needed for the latter; Cymochorea Coues 1864, type species leucorhoa, seems the most likely name.”

Comments from Robbins: “NO, based on comments by both Alvaro and Gary.”

Comments from Remsen: “NO. In a reversal of my vote and comments in NACC, which were flawed because I did not consult the original paper, which has a figure not presented in the proposal:

“This shows that regardless of resolution within clades, genetic support is strong for recognition of four clades as genera, all of which are predicted to be evolving separately since the Miocene, i.e. older than most groups we label as genera. The importance of treating them as genera is to emphasize the morphological conservatism in this group relative to other avian lineages, and thus emphasize one of the most important features of storm-petrel evolution. One problem with this is that “markhami” is actually a melania sample, as indicated in Alvaro’s comments (yet another example of why double-checking ID of vouchers is important!). Thus, placement of markhami would be based on educated guesswork. Another problem is that a new genus would need to be named for extralimital monteiroi-jabejabe (“not our problem”); tentative inclusion in Clade C would be the solution NACC could follow. However, I prefer sorting out the markhami problem than obscuring 10 my-old diversity in one broad genus. For what it’s worth, Mathews named four additional genera in the group, including Loomelania, which was still used by AOU (1957). A new proposal recognizing multiple genera would need to sort out the nomenclature carefully.”

Comments from Claramunt: “YES. The molecular data is robust and I don’t see any way of subdividing this family into clear and logical genus-level taxa. It is also premature to base any subdivision on divergence times; the same paper presents much younger estimates as plausible alternatives. Subdividing these storm petrels into multiple genera will not only create a discrepancy with NACC but also, I think, will produce confusion and frustration among users of the classification. Maybe in the future we can consider a proposal for subdividing this family, but at this point, the only option is to lump Oceanodroma into Hydrobates. Otherwise, for how long will we be maintaining a paraphyletic Oceanodroma?”

Additional comments from Remsen: “YES, reversing my vote, based on Santiago’s comments, which actually reflect the comments of most NACC members. Until all the details are sorted out, including a thorough investigation of the backbone nomenclature, I prefer maintaining a single broad genus rather than trying to do an adequate job re-arranging generic limits within the context of a SACC proposal. Looking forward to a publication that does this properly.”

Comments from Schulenberg: “Hydrobates (the older name) apparently is embedded in Oceanodroma. One solution is to recognize a single genus (Hydrobates) for all (e.g., Dickinson and Remsen 2103, del Hoyo and Collar 2014, NACC, my own preference), or one can opt to partition these species into multiple genera (from two up to perhaps six genera).

“One suggestion I have is anyone who votes NO on the single genus solution ought to specify exactly what they see as the desirable number of genera to recognize.

“More importantly, SACC should recognize that accepting putting some species into Hydrobates while retaining others in Oceanodroma is not a option that is on the table. the type species of Oceanodroma is furcata, which apparently is sister to Hydrobates - so Oceanodroma disappears no matter what.

“At some point the nomenclatural experts will need to be consulted. my take - and I don't bill myself as a nomenclatural expert - is that the most likely options (based on Figure 3 in Wallace et al. 2017 - I haven't checked other relevant papers) - include:

“two genera:

Hydrobates: pelagicus, furcatus, tethys, melania, microsoma

Cymochorea Coues 1864, type leucorhoa: hornbyi, leucorrhoa, socorroensis, cheimomnestes, monorhis, homochroa, castro, monteiroi, jabejabe, matsudairae, tristrami

Hydrobates in turn could be split into two genera:

Hydrobates: pelagicus, furcatus

Halocyptena Coues 1864, type microsoma: tethys, melania, microsoma

or into three genera:

Hydrobates: pelagicus, furcatus

Halocyptena: tethys, microsoma

Loomelania Mathews 1934, type melania: melania

and Cymochorea also could be split into two genera:

Cymochorea: hornbyi, leucorrhoa, socorroensis, cheimomnestes, monorhis, homochroa, matsudairae, tristrami

Thalobata Matthews 1943, type species castro: castro, monteiroi, jabejabe

“or three genera

Cymochorea: hornbyi, leucorrhoa, socorroensis, cheimomnestes, homochroa, tristrami

Thalobata: castro, monteiroi, jabejabe

Pacificodroma Bianchi 1913, type species monorhis: monorhis, matsudairae

“or, again, there's nothing wrong with classifying all in Hydrobates and walking away.”

Additional comments from Robbins: “After reviewing this again following Tom's comments, I feel at a minimum two genera should be recognized because of the very deep split, i.e., Hydrobates and Cymochorea. Obviously, if one wants to atomize things, one could recognize 4 or even 6 (why go that far as it starts to become meaningless to communicate relationships).

What should be considered regarding Tom's suggestion of recognizing a single genus vs multiple genera is how is the timing of the various branches of this clade compared to other currently recognized seabird genera? That is we should attempt to be consistent in what we call a genus if time is considered a parameter in defining genera.”

Additional comments from Claramunt: “I don’t think that even a two-genera solution is warranted for two important reasons:

“1) We don’t know the basal split. The base of the tree is, essentially, a polytomy. There is no statistical support for the basal relationships that are resolved in some of the trees.

“2) We don’t know how old the genus is. Wallace et al. estimates range from 5 to 22 million years. In addition, it is obvious that the evolutionary clock is ticking slower in these petrels, so the use of the absolute age is not completely justified as a yardstick.”

Comments from Pacheco: “YES. Because NACC has adopted the single-genus treatment in agreement with Santiago’s comments, the reasons for arrangement in two or more genera become less obvious.”

Comments from Stotz: “YES. This might not be the best answer, but I think it is the only approach we can currently justify.”

Comments from Bonaccorso: “YES. At least for now. Even if the basal split lacks support. Naming four genera without solving sampling issues would generate more instability.”

Comments from Areta: “NO. I agree with Alvaro and with Van´s first comments. Having a single-genus family including multiple species (limits of which are in need of further study) with fairly deep divergence times makes no sense to me: it hides whatever is shared by species groups and does not facilitate meaningful communication, nor helps condense what is known of these birds (two important features of any classification). This proposal has no analysis or suggestion of other alternatives, and no attempt at all to put biological or morphological data in the game, as such, it is an unbalanced view whose outcome will most likely be the merging of everything under a single genus. Voting NO is uncomfortable because it leaves a paraphyletic Oceanodroma, but I see no reasonable and solid alternative proposed.

“I am in no position to propose in how many genera the family Hydrobatidae could be reasonably split because a) it is a daunting work that should be done with deep biological knowledge of the birds, b) it demands careful examination of nomenclatural facts, c) there are more taxa that need to be sampled, and d) it should be published as a review paper dealing with the pros and cons of different classification schemes. Recently, Howell & Zufelt (2019) used a three genera approach with Thalobata, Halocyptena and Hydrobates (I thank Quillén Vidoz for sharing images of this book).

“Regarding stability, lumping everything in no way serves stability better than other alternatives. It just sweeps complexity (and knowledge) under the carpet. To be sure, yes, a single Hydrobates genus including all species is consistent with phylogenetic information. But is this useful?

“I am seeing that several molecular phylogenetic works lead to proposals that favor the creation of large genera. However, these proposals seldom attempt to analyze uniting features of smaller units that could be reasonably separated at the genus level (a largely subjective decision, we know). In my view, classifications should integrate these data into a useful research and communication device, and my take is that just looking at trees without understanding the birds will not produce the best possible classification. Maybe the single-genus option is informative enough and other options do not add anything meaningful, but Alvaro´s comments, Howell & Zufelt ´s 2019 book and the level of genetic differentiation suggest that there are several features to be uncovered to subdivide Hydrobatidae into genera. Indeed, looking at the original descriptions of those old names would be an excellent starting point. But this should be done comparatively, integrating modern evidence, etc., and not by me in a vote on a proposal. I am sure that Steve Howell can contribute loads of ideas and information here. The tension between the recognition of multiple smallish genera and single large genera needs to be sorted out based on integrative evidence. I do not see such an integration here, and so I feel uncomfortable with the proposal and with the options debated so far.

“Howell, S.N.G. & Zufelt K (2019) Oceanic birds of the World. Princeton University Press, NJ.”

Comments from Zimmer: “YES. I’m in broad agreement with Alvaro and others that lumping everything into Hydrobates would mask some pretty deep divisions, and that there are some pretty significant biological/ecological/biogeographical (and, almost certainly vocal) distinctions between these storm-petrels that are worthy of recognition at the generic level. However, I think that any such novel classification needs its own separate proposal, and that proposal needs to be based upon some published analysis that integrates multiple data sets (i.e. not just the genetic data). Meanwhile, I think it is far preferable to recognize a single, overly broad genus that at least is monophyletic, rather than continue with a classification that also conceals considerable variation, while also being demonstrably paraphyletic.”

Additional comments from Stiles: “I change to YES to sinking Oceanodroma into Hydrobates … if only to preserve monophyly. Hopefully someone like Alvaro can get together with a genetics lab and resolve the basal polytomy and incorporate data on plumage, flight, and vocalizations to get a good robust phylogeny!”