Tomkins and Milne (1991) presented information suggesting that the
Galapagos and Hawaiian populations of Dark-rumped Petrel might be distinct
at the species (rather than subspecies) level. Galapagos birds are 16%
larger, but Hawaiian birds lay eggs that are 18% larger. (Presumably
Hawaian females make louder squeals during egg-laying.) Black markings on
the white forehead are variable (sometimes lacking) in Galapagos birds but
are lacking in Hawaiian birds. Hawaiian birds have a short, well-defined
nesting period, but Galapagos birds have an extended, almost-always nesting
period (but there are several island colonies). And, sonagrams of the two
populations are said to be very different. However, Tomkins and Milne
(1991: 12) show sonagrams, and Simons (1985:236) show sonograms, but I find
them impossible to compare. Thre is no evidence (but there is speculation)
that they occur together in non-breeding times. This 1991 paper was
apparently convincing enough that Sibley and Monroe split the species in
1993. [Why Burt did not do that for the Check-list is another continuing
mystery. He kept them as groups.] They used the names Hawaiian Petrel and
Galapagos Petrel, maintaining Dark-rumped for the combination.

Browne et al. (1997) summarized the above and did allozyme electrophoresis
on blood from members of the two populations (no voucher specimens). They
found one fixed allelic difference between them, out of 13 loci. They say
this "supports" their recent elevation to species status by S&M. This paper
also summarizes morphological differences and mentions vocal differences.

One of the coauthors of Browne et al. (1997) was also a coauthor of the BNA
account on Pterodroma phaeopygia (Simons and Hodges 1998). The Condor
article was submitted in Dec. 1996, accepted April 1997, and in the August
issue. The BNA account was submitted in Jan. 1997, published in 1998. Thus
work on the two papers was simultaneous. However, the BNA account makes NO
mention of the genetic work, and the possible specific status of the two
forms was downplayed. That is very interesting.

Pratt and Pratt (2001) recognize sandwichensis at the species level.

Recommendation: I think we should split these populations into species.

References
Browne, R. A. et al. 1997. Condor 99:812-815.
Pratt and Pratt. 200l. SAB on Hawaii
Sibley and Monroe 1993. Supplement
Simons, T. R. 1985. Condor 87:229-245. Re Hawaiian birds
Simons and Hodges. 1998. BNA 345
Tompkins, R. J., and B. J. Milne. 1991. Notornis 38: 1-35.

Addendum from Remsen: No one on AOU CLC, including Dick, was impressed one way or another on the evidence either way on this one, but in the end we decided unenthusiastically to go with the split because there seemed to be less evidence for considering them the same species than for considering them separate (and recognizing that the original lump of the two was probably based on "data-free" opinion.) We fearfully assumed that the described differences in voices are OK, and we were impressed with the differences in reproductive biology between them.

Ridgway described sandwichensis as distinct species. Of interest is that Jouanin & Mougin (1979) ("Peters") considered sandwichensis as a "doubtfully distinct" species, and this was repeated by Carboneras (1992) ("HBW"); this influenced Dickinson (2003) to recognize them only as subspecies (and that's the genesis of SACC's current classification).

[I haven't checked to see if more recent data have been published -- please chime in if you know of some.]

References:

CARBONERAS, C. 1992b. Family Procellariidae (petrels and shearwaters). Pp. 216-257 in "Handbook of the Birds of the World, Vol. 1" (J. del Hoyo et al., eds.). Lynx Edicions, Barcelona.

JOUANIN, C., AND J.-L. MOUGIN. 1979. Order Procellariiformes. Pp. 48-121 in "Check-list of birds of the World, Vol. 1, Second Edition" (Mayr, E. & G. W. Cottrell, eds.). Museum of Comparative Zoology, Cambridge, Massachusetts.

============================================

Comments from Robbins: "[YES]. I did not appreciate how weak the data are for this split, but given that it is suggestive and everyone has been splitting these I guess we should follow. Thus, I vote "yes" for recognizing them as two species."

Comments from Stiles: "This is one I'd love to abstain on, but if passing the buck isn't to be allowed, I will most reluctantly vote YES, largely because the AOU checklist committee did so (and as "hijito", it might not be good to flout "Papa's" decisions, however off-base we might consider them - and I do think this was not one of their wiser decisions). I am decidedly skeptical about using genetic analysis to decide species-level questions, especially if sampling was not dense, and the breeding-season argument cuts both ways - there are a number ot cases of seabird populations on different islands differing in breeding seasonality - Sooty Tern comes to mind). A more thorough statistical analysis of both vocalizations and measurements would also help."

Comments from Zimmer: "YES, although the evidence is weak."

Comments from Jaramillo: "YES ­ I think that Procellariformes will give Scytalopus a run for their money in the end. This is a not a group that tends towards divergence in plumage pattern, and my guess is that many taxa currently classified as subspecies are probably good and very well differentiated species. Admittedly this is a gut feeling more than anything, but as soon as more vocal data are available, and when a careful molecular analysis is performed on the group we may end up being surprised at the results. I will stop there and just say that I am comfortable with this split."

Comments from Nores: "YES. Acepto reconocer a Pterodroma sandwichensis como especie. Aunque cabe reconocer que la evidencia es liviana y el trabajo genético que sería lo más fuerte, tiene la contradicción (o falta de aceptación del coautor) indicada por Richard."