Proposal (167) to South American Classification Committee

Recognise the subfamilies Hydrobatinae and Oceanitinae within the family Hydrobatidae

Background

In the SACC Baseline document, as viewed on March 1, 2005, the following statement is made:

Many classifications (e.g., Carboneras 1992c) divide the family into two subfamilies, Hydrobatinae and Oceanitinae, but this classification follows the AOU (1998) in not recognizing these until corroborated by multiple independent data sets that mark major, deep branches within the family. Recent genetic data, however, supports recognition of these two subfamilies (Penhallurick & Wink 2004).

Proposal

Below we reproduce Figure 2 from Penhallurick and Wink (2004). This shows a Neighbourhood Joining tree showing a Bootstrap Cladogram (1000 replicates) using Jukes-Cantor as a distance algorithm.

Bootstrap cladogram (1000 replicates) reconstructed with the Neighbourhood Joining (NJ) method using Jukes Cantor as a distance algorithm.

The Storm-Petrel family, Hydrobatidae, has two basal nodes. One node, containing taxa from Oceanites oceanica Wilson's Storm-Petrel to Fregetta tropica Black-bellied Storm-Petrel, has 99% Bootstrap support. The other, containing taxa from Hydrobates pelagicus European Storm-Petrel to Oceanodroma castro Band-rumped Storm-Petrel, has 100% Bootstrap support. This provides strong support for the division of the Hydrobatidae into two subfamilies: Oceanitinae, comprising taxa from Oceanites oceanica to Fregetta tropica; and Hydrobatinae, with taxa from Hydrobates pelagicus to Oceanodroma castro. (see tree in Penhallurick & Wink -- ask Remsen for pdf if needed).

	78	79	80	81	82	83	84	85	86	87	88
[78] Oceanites oceanicus		0.0541	0.0321	0.0321	0.0348	0.0709	0.0737	0.0781	0.0681	0.0794	0.0681
[79] Garrodia nereis	0.1076		0.0458	0.0430	0.0403	0.0880	0.0880	0.0781	0.0681	0.0794	0.0681
[80] Pelagodroma marina	0.0936	0.0962		0.0348	0.0403	0.0794	0.0822	0.0907	0.0765	0.0880	0.0822
[81] Fregetta grallaria	0.0927	0.0989	0.0884		0.0159	0.0652	0.0709	0.0781	0.0681	0.0794	0.0681
[82] Fregetta tropica	0.0892	0.1015	0.1076	0.0744		0.0709	0.0765	0.0844	0.0681	0.0765	0.0851
[83] Hydrobates pelagicus	0.1277	0.1409	0.1330	0.1269	0.1312		0.0159	0.0844	0.0681	0.0765	0.0851
[84] Halocyptena microsoma	0.1409	0.1505	0.1435	0.1286	0.1382	0.1006		0.0443	0.0159	0.0403	0.0294
[85] Thalobata castro	0.1538	0.1615	0.1635	0.1462	0.1490	0.1212	0.1163		0.0413	0.0657	0.0565
[86] Hydrobates furcatus	0.1330	0.1435	0.1356	0.1321	0.1339	0.0822	0.0971	0.1173		0.0267	0.0348
[87] Cymochorea leucorhoa	0.1382	0.1531	0.1479	0.1260	0.1356	0.1067	0.1085	0.1240	0.0989		0.0485
[88] Halocyptena melania	0.1295	0.1505	0.1417	0.1295	0.1347	0.1085	0.0796	0.1346	0.1076	0.1129

The table above reproduces uncorrected p-distances, based on complete sequences of the mitochondrial cytochrome-b gene, below the diagonal, and amino-acid distances, based on the 300 bases within cytochrome-b involved in the production of amino acids, above the diagonal.

In terms of the uncorrected p-distances, the within-group mean distance for taxa within the Oceanitinae is 9.501% (S.D. 0.94). The within-group mean distance for taxa within the Hydrobatinae is 10.631% (SD 1.55). In contrast, the between-group mean distance is 14.086% (SD 0.98). These data confirm strongly that we are dealing with two distinct subfamilies.

Recommendation: That the subfamilies Oceanitinae and Hydrobatinae be recognized with the family Hydrobatidae.

John Penhallurick, March 2005

Comments from Nores: "SI, los datos moleculares de Penhallurick son muy convincentes, aunque no veo porque hay que tratar esto ahora. Aunque en los comentarios sobre la familia Hydrobatidae se habla de estas subfamilias, en la check-list no aparecen subfamilias. De todos modos, pienso que es valorable la propuesta de Penhallurick."

Comments from Stiles: "YES. In this case the genetic data are appropriate to the question and the study seems solid. My only query is regarding our somewhat nebulous criterion of "several independent studies" for such cases. Presumably the original division into two subfamilies was based upon a series of morphological and perhaps other characters - would this constitute "independent evidence"? Or should we interpret the criterion of multiple independent studies to pertain specifically to genetic studies? The implication here is that no single study, however careful or convincing, would suffice to change the 'status quo'."

Comments from Pacheco: "YES. Considero a proposição positiva, na medida em que confirma informação prévia sobre as relações intra-familiares. Os resultados obtidos por Penhallurick & Wink parecem demonstrar a existência desses dois subgrupos."

Comments from Jaramillo: "YES - I think that the higher level patterns of relationship noted in Penhallurick and Wink (2004) are solid. Although, they are pretty much the same patterns noted by Nunn and Stanley (G. B. Nunn, S. E. Stanley 1998. Body size effects and rates of cytochrome b evolution in tube-nosed seabirds. Mol. Biol. Evol. 15: 1360-1371.), but this is not surprising as the two sets of work are based at least partially on the same data. So to address Gary's question, there are separate recent works that come to the same conclusion, but they are not entirely independent. Going back to older work and the initial suggestion that the Storm Petrels fit into two main groupings will almost certainly be based on morphological data so that is surely independent data. In general, the southern group (Oceanitinae) have much longer legs than the northern group."