Proposal (388) to South American Classification Committee
Split Diomedea exulans into four species
Effect on SACC: This would divide an existing species, Diomedea exulans (Wandering Albatross), into four species, at least three of which have been recorded in South American waters.
Background: The Wandering Albatross (Diomedea exulans) as currently defined by SACC consists of four taxa:
(1) nominate exulans (Linnaeus, 1758) of circumpolar distribution across the Southern Ocean (e.g. South Georgia, Kerguelen, Heard, Macquarie),
(2) dabbenena (Mathews, 1929) from the temperate-zone island group of Tristan da Cunha and Gough,
(3) antipodensis (Robertson & Warham 1992) from the sub-polar Antipodes and Campbell Islands off New Zealand,
(4) gibsoni (Robertson & Warham 1992) from the sub-polar Auckland Islands off New Zealand.
The four taxa are characterized by a trend of neoteny and decreasing size towards the north, with an all-white adult plumage in the circumpolar exulans giving way to a complex pattern of brown immature plumage retention in the temperate and New Zealand taxa. The morphological distinctness between the two New Zealand taxa gibsoni and antipodensis is in question, and many adult individuals cannot be unequivocally identified, although adult gibsoni is generally paler (=less neotenous) and somewhat approaches exulans (Tickell 2000; Brooke 2004; Onley and Scofield 2007).
In 1983, a fifth taxon was described from the temperate Indian Ocean island of Amsterdam (Roux et al. 1983). This new taxon – breeding at equal latitudes as dabbenena – displays the most extreme form of neoteny with an all-brown adult plumage. Therefore, Roux et al. (1983) decided to award it species status: D. amsterdamensis. However, this treatment is not universally accepted, and some authors have proposed the inclusion of D. amsterdamensis in D. exulans (e.g. Penhallurick and Wink 2004).
New Information: In a controversial book chapter, Robertson and Nunn (1998) put forth a revised albatross taxonomy in which they elevated all nominal subspecies to species rank. Consequently, they recognized five Wandering Albatross species. Their proposal was only partly based on published phylogenetic evidence and was therefore rejected by some subsequent authors (e.g. Penhallurick and Wink 2004; Christidis and Boles 2008). However, despite widespread criticism in scientific circles, the new Robertson and Nunn classification has been widely adopted by authors of popular bird books (Tickell 2000; Brooke 2004; Onley and Scofield 2007) and by the conservation community (e.g. Birdlife International).
Penhallurick and Wink (2004) employed GenBank sequences of albatrosses and other seabirds to construct phylogenetic trees and to compute cyt-b divergences. They found divergences between the five Wandering Albatross taxa to be well below their 3.2% “cut-off” for species status and concluded that all five taxa be lumped into a single species. Penhallurick and Wink’s (2004) work came under criticism for analytical flaws and the rigid use of cyt-b divergences as a species threshold (Rheindt and Austin 2005).
Burg and Croxall (2004) examined sequences of the mitochondrial control region in four Wandering Albatross taxa (exulans, antipodensis, gibsoni, dabbenena). They found that exulans, dabbenena, and a clade comprising the two New Zealand taxa (gibsoni and antipodensis) were separated from one another by 4.5 – 5.2% divergence. The New Zealand taxa gibsoni and antipodensis were little differentiated between each other and shared several haplotypes, although there was one near-fixed difference. Additionally, Burg and Croxall (2004) analyzed nine microsatellite loci in most samples. They were unable to detect differentiation between any of their exulans populations, even between individuals from different oceans, while there were significant levels of population differentiation between antipodensis and gibsoni. Based on their genetic data, Burg and Croxall (2004) proposed a three-species treatment under which the taxon gibsoni is subsumed under D. antipodensis, while D. dabbenena and a monotypic D. exulans are also viewed as distinct species. Please note that Burg and Croxall (2004) did not address P. (e.) amsterdamensis.
Burg and Croxall’s (2004) results agree with phylogenetic studies on other Antarctic sub-polar seabirds that have demonstrated strong latitudinal differentiation in the absence of longitudinal differentiation. For instance, Jouventin et al. (2006) detected pronounced phylogenetic differences in rockhopper penguins and elevated the northern taxon from the temperate zone to species level (Eudyptes moseleyi), even though geographic distances between some northern and southern populations are much smaller than towards their conspecifics. This finding suggests that global sea currents of different temperature play an important role in seabird speciation.
Analysis and Recommendation: The biological species status of the five Wandering taxa must be assessed in the light of two types of considerations:
(1) The relevance of sequence divergence: A vast and growing body of literature shows that many avian biological sister species are generally differentiated by at least c. 2-3% divergence in a number of widely used mitochondrial coding genes (ND2, cyt-b, COI), although this figure can be substantially lower on account of rapid speciation (Johnson and Cicero 2004) or can be skewed by genetic introgression following hybridization (e.g. Funk and Omland 2003). Albatrosses in general and the five Wandering taxa in particular are characterized by low divergences (<1% cyt-b; Penhallurick and Wink 2004). These minute values led Penhallurick and Wink (2004) to unite the five taxa into a single species. Although Burg and Croxall (2004) found c. 5% divergence between Wandering taxa in the mitochondrial control region, this locus is known to evolve much faster than the widely used coding genes (ND2, COI, cyt-b) and, therefore, cannot be compared.
The use of mtDNA divergences as a species indicator intersects the barcoding debate and the molecular clock debate, and remains controversial. The latest studies seem to show that a homogeneous rate of c. 2.1% mtDNA divergence per million years may apply across a wide range of bird families for comparisons below the genus level (e.g. Weir and Schluter 2008). This would indicate that Wandering taxa are very young. However, important differences between rates of mtDNA evolution among bird families cannot be ruled out (Pereira and Baker 2006) and have even been detected within the Procellariiformes (Nunn and Stanley 1998).
There are other aquatic bird clades characterized by tiny interspecific mtDNA divergences (e.g. Anas ducks – Johnson and Sorenson 1999; Larus gulls – Liebers et al. 2004) possibly related to frequent hybridization and mtDNA introgression in these groups. Therefore, past introgression events may have artificially reduced divergences between Wandering taxa and may explain why some researchers perceive these divergences to be too low for the generally high level of morphological differentiation.
On the other hand, Friesen et al. (2007) have shown that speciation in pelagic seabirds may proceed at a much faster time scale than in most other tetrapods. Their demonstration that a speciation event between two biological species of storm-petrel dates back only 110,000 – 180,000 years opens up the possibility that the Wandering taxa may also have attained biological species status in a shorter timeframe than generally expected.
In summary, tiny mtDNA divergences among Wandering taxa may be an artifact of introgression. Even if not, the potentially young age of Wandering taxa does not rule out biological species status, because even more rapid speciation has been documented in other seabirds. Therefore, other data sources must be consulted to obtain evidence for or against the biological species status of Wandering taxa.
(2) The question of taxon allopatry: Akin to Proposal 166 on Shy Albatrosses, it is relevant to examine whether the Wandering taxa are allopatric in essence. One sub-polar taxon (exulans) stretches around the globe and closely approaches the range of the other three taxa at its breeding colonies in Macquarie Island (620 km to nearest gibsoni, 700 km to nearest antipodensis) and South Georgia (2500 km to nearest dabbenena). In contrast, the Atlantic breeding colonies of exulans (South Georgia) are 5100 km removed from the nearest colony in the Indian Ocean (Prince Edward Island), and the minimum distance between its Indian Ocean breeding grounds at Kerguelen and its New Zealand colony at Macquarie is 5800 km. Although Macquarie was not sampled by Burg and Croxall (2004), their microsatellite data show a panmictic population structure of exulans between South Georgia and Crozet across 6000 km. The uniform plumage reflects this lack of genetic differentiation across the range of exulans and contrasts with the distinct neotenous adult plumages of the other taxa.
The lack of genetic data on Macquarie notwithstanding, the small exulans colony on this island only numbers c. 10 birds and is subject to current immigration and emigration to/from the Indian Ocean (de la Mare and Kerry 1994). It can, therefore, not be very different from Indian Ocean colonies in terms of population genetics. In fact, exulans is a widespread visitor to New Zealand and Australian waters in numbers that exceed the size of the Macquarie colony. The fact that exulans has colonized an island within the range of gibsoni and antipodensis, but fails to interbreed with them, suggests that they may have attained prezygotic isolation mechanisms.
Recommendations: Based on the above considerations, I suggest the following three options for treatment:
(1) One species: Lump all five taxa into D. exulans. This is the status quo, although I am not sure about SACC’s current stance on the extralimital taxon amsterdamensis.
(2) Four species: Recognize all taxa as distinct species, except for gibsoni, which is retained in D. antipodensis. This is Burg and Croxall’s (2004) proposal, though – here again – they did not comment on amsterdamensis.
(3) Five species: Recognize all taxa as distinct species. This is Robertson and Nunn’s (1998) treatment.
I advise against Option 3, because gibsoni and antipodensis are poorly differentiated morphologically and genetically (Burg and Croxall 2004). The significant microsatellite structure between both taxa is consistent with subspecies treatment.
As far as Options 1 and 2 are concerned, I do not feel that there is overwhelming evidence for either treatment. However, if I were forced to make a recommendation, I would advocate Option 2, because distributional data indicate that the New Zealand taxa don’t interbreed with exulans even though they could. By yardstick analogy, the temperate-zone dabbenena (which may have a different life-history owing to its warm-current environment) would be at the species level because its control-region differentiation towards the other taxa is even more pronounced than that of the New Zealand clade (Burg and Croxall 2004). Data on the extralimital amsterdamensis are lacking, but on account of its high level of morphological differentiation (=extreme neoteny) it may be best to go with the describers’ recommendation for species status (Roux et al. 1983) until and unless other data have been presented.
Friesen VL, Smith AL, Gómez-Díaz E, Bolton M, Furness RW, González-Solís J, Monteiro LR (2007) Sympatric speciation by allochrony in a seabird. Proc. Natl. Acad. Sci. U S A 104, 18589–18594.
Jouventin P, Cuthbert RJ, Ottvall R (2006) Genetic isolation and divergence in sexual traits: evidence for the northern rockhopper penguin Eudyptes moseleyi being a sibling species. Mol. Ecol. 15 (11), 3413-3423.
Pereira SL, Baker AJ (2006) A mitogenomics timescale for birds detects variable phylogenetic rates of molecular evolution and refutes the standard molecular clock. Mol Biol Evol 23, 1731-1740.
Roux JP, Jouventin P, Mougin JL, Stahl JC, Weimerskirch H (1983) Un nouvel albatros Diomedea amsterdamensis n. sp. decouvert sur L'ile Amsterdam (37°50'S, 77°35'E). Oiseau et la Revue Francaise d'Ornithologie 53 (1), 1-11.
Tickell WLN (2000) Albatrosses. Yale University Press, New Haven, Connecticut.
Weir JT, Schluter, D (2008) Calibrating the avian molecular clock. Mol. Ecol. 17, 2321–2328.
Frank Rheindt, January 2009
Remsen addendum: Thus a YES vote would indicate favoring a four species treatment, and no would favor single species (status quo). If you favor a 5-way split, then vote YES on this one, indicate that you would favor 5, and if there is enough support, I’ll create a new proposal to go from 4 to 5 species.
Comments from Robbins: “YES, given the current state of knowledge, Frank’s suggestion of recognizing four species seems the best course of action.”
Comments from Zimmer: “YES. I find the Burg & Croxall (2004) proposal to be the most compelling. I can’t support treating gibsoni/antipodensis as separate species, and based on current evidence, I’m willing to give amsterdamensis the benefit of the doubt.“
Comments from Jaramillo: “YES – This is a subject I have been following for a while, and certainly the treatment of polytypic Wandering, and a separate Amsterdam didn’t jive with me. Certainly if Amsterdam was a species, the other group that shows considerable neoteny in plumage, particularly that of females (the NZ group, gibsoni and antipodensis) seemed to warrant species status too, for example. The issue of latitude and water temperature is indeed of considerable importance in the argument to separate these taxa, and fortunately it is finally being given the consideration it deserves. We think it logical to separate closely related yet different (voice, display, morphology, genetics) species east of the Andes and west of the Andes, well the same is true of temperate vs. Subantarctic seabirds, water temperature differences may be a barrier as great as a mountain range or broad river to a landbird. I have not seen any work that analyzes the displays of these different forms. It would be nice to have that data to work with as well, but alas it is not out there in a manner that we can use to assess behavioral barriers to reproduction. But as the proposal states, the potential for mixing is there, because many of these forms overlap in the broad sense that the at sea range may include breeding islands of members of other taxa in the complex. Finally, it may not sit well with some people that some of these taxa are extremely similar to each other, or even possibly unidentifiable other than at the breeding island (at least some individuals of the population); however. this is common in Procellariiformes. Divergence in plumage coloration is not great in seabirds, and sometimes within-species plumage variation is greater than between-species plumage variation (observed in various polymorphic Pterodroma, or even Leach’s Storm-Petrel). Four species seems like a reasonable course of events. Although there are some clear differences between gibsoni and antipodensis, there are also greater similarities between these two than with the rest.
“Finally, three of the four have occurred within our waters:
exulans broadly in the southern oceans; antipodensis off Chile (photos and satellite tracked individuals); dabbenena in Argentina – Uruguay and I assume Brazil.”
Comments from Pacheco: "YES. Diante do exposto, eu creio que o melhor arranjo seja aquele que contempla a adoção de 4 espécies, tal qual recomendado por Burg e Croxall (2004).”
Comments from John Penhallurick:
Frank Rheindt discusses the taxonomy of five taxa, four of which (nominate exulans, dabbenena, antipodensis and gibsoni) have been traditionally treated as subspecies of Diomedea exulans; and one (amsterdamensis) which was originally described as a separate species (Roux et al. 1983), but which a number of recent publications (Bourne, 1989; Vuilleumier et al. 1992; Dickinson, 2003; Penhallurick and Wink, 2004; Christidis and Boles 2008) have treated as a subspecies of D. exulans. Rheindt appears to recommend recognising four species: exulans, dabbenena, antipodensis (including gibsoni) and amsterdamensis.
Rheindt writes relatively approvingly of Robertson and Nunn (1998), who announced that all terminal albatross taxa (monotypic species and what had previously been considered subspecies) should be considered species in terms of the Phylogenetic Species Concept (PSC), to which they indicated their allegiance. He notes that the Robertson and Nunn classification “has been widely adopted by authors of popular bird books (Tickell 2000; Brooke, 2004 Onley and Scofield 2005) and by the conservation community (Birdlife International).” It should be remembered that Robertson and Nunn’s treatment was published as an unrefereed book chapter, not in a refereed journal. Furthermore, Robertson and Nunn adduced no evidence at all in support of their proposal: it was simply a matter of fiat in terms of the PSC.
Rheindt also writes approvingly of Burg and Croxall (2004), and notes that his final preference, that is, recognising four species, agrees with Burg and Croxall’s conclusion. It must be remembered that both papers by Burg and Croxall (2001, 2004) wrote not in terms of the multidimensional Biological Species Concept (BSC Mayr 1996). In terms of what species concept they were using, Burg and Croxall cite Moritz (1994a and 1994b), who described the differences between management units (MU) and evolutionary significant units (ESU): ESUs are two groups that show reciprocal monophyly of mtDNA haplotypes and significant differences in allele frequencies at nuclear loci. MUs on the other hand show significant differences in allele frequencies without regard to the phylogeny of the markers. They also cite Avise & Wollenberg (1997), who endorsed the Phylogenetic Species Concept (PSC), which emphasizes the criteria of phylogenetic relationships, and not reproductive relationships. Thus it appears that they are using either the ESU model, which stresses conservation values, or the PSC, which treats all subspecies as good species. Burg & Croxall’s discussion is irrelevant to any discussion of albatross taxonomy in terms of the multidimensional BSC.
It should be remembered that many recently discussions of albatross taxonomy have widely relied on conservation values. This is understandable: there are severe threats to many albatross taxa from practices like long-line fishing. It is extremely regrettable that much conservation legislation is written in terms of species, not subspecies, the latter of which, being geographical representatives of species, are often at risk. As Schodde and Mason (1999) have stated: “ ‘subspecies’, as genetically distinct regional populations, are the building blocks of evolution and the ‘real’ units of biodiversity, and…many more of them are and endangered than ‘biological species’.” Schodde and Mason proposed a new term “ultrataxon” to refer to all terminal taxa, the second reason for doing so being that it averts a re-circumscribing of Australian bird fauna under alternative definitions of species, which would lead to ambiguity and confusion in classification. However, in the body of their book, Schodde and Mason continued to classify birds in terms of species and subspecies in the fashion of the multidimensional BSC.
Rheindt, correctly, recognises the role of neoteny in the differentiation of these taxa, although it is unclear exactly what he means when he cites dabbenena as displaying “the most extreme form of neoteny”. The plumage of gibsoni, antipodensis and amsterdamensis recalls stages 1-3 of nominate exulans. Such changes probably reflect epigenetic change, and specifically gene silencing, in that the one or more of the genes that are responsible for the change from juvenile or immature plumage are switched off. Neoteny is also apparently responsible for the differences between the two taxa in Diomedea epomophora, in that sanfordi retains black upperwings. The juvenile stages of both taxa appear to be almost identical, involving some black speckling on the back, slightly more in sanfordi, and largely black upperwings; the nominate shows more white, although the juvenile sanfordi also shows “fine white fringing to greater median and some lesser secondary coverts” (Marchant & Higgins 1990: 282). The fact that the cytochrome-b Tamura-Nei distance between these two taxa is 0.0000% suggests that they diverged very recently. This suggests that the epigenetic factors involved in such plumage differences can evolve very quickly indeed, and morphological differences due to such factors should not, of themselves, be claimed as evidence of species status.
In their critique of Penhallurick and Wink (2004), Rheindt and Austin (2005) cite specifically Abbott and Double (2003 a, b) as well as Burg and Croxall (2001,2004) as studies that have uncovered new evidence for the species status of at least some of these forms. I have explained above why I do not consider Burg and Croxall’s papers as relevant to the question of species concepts under the multidimensional BSC. Similarly I do not believe that anything in Abbott and Double’s two papers is relevant to the question we address here: should a number of taxa traditionally treated as subspecies within a single species be treated as comprising two or more species. Of critical importance in this discussion are the species concepts utilised in these studies.
Abbott and Double (2003a) initially stated that they were adopting the species nomenclature suggested by Robertson & Nunn (1998). Since Robertson and Nunn (1998) explicitly said that they were working within the Phylogenetic Species Concept, and since Robertson and Nunn’s paper was not published in a refereed journal, this is not a promising beginning for any discussion in accordance with the Multidimensional BSC. In their abstract, Abbott and Double (2003a) stated that their analysis confirmed the separation of the shy/white-capped pair and the Salvin’s/Chatham pair but did not provide species-level resolution (Emphasis added).
An admitted limitation of Penhallurick & Wink (2004) was that it relied on a single gene: cytochrome-b, which was criticised by Rheindt and Austin (2005) and also in Rheindt’s Proposal 388. And it is obviously desirable to confirm these findings with studies of other, particularly nuclear, genes. However, it should be remembered that many studies have confirmed the utility of both cytochrome-b and Bayesian inference. For example, May-Collado & Agnarsson (2006), in a study of cetacean phylogeny, stated in their abstract:
Until more genes are available for a high number of taxa, can we rely on readily available single gene mitochondrial data? Here, we estimate the phylogeny of 66 cetacean taxa and 24 outgroups based on Cyt-b sequences. We judge the reliability of our phylogeny based on the recovery of several deep-level benchmark clades. A Bayesian phylogenetic analysis recovered all benchmark clades and for the first time supported Odontoceti monophyly based exclusively on analysis of a single mitochondrial gene. The results recover the monophyly of all but one family-level taxa within Cetacea, and most recently proposed super- and subfamilies. In contrast, parsimony never recovered all benchmark clades and was sensitive to a priori weighting decisions. These results provide the most detailed phylogeny of Cetacea to date and highlight the utility of Bayesian methodology in general, and of Cyt-b in cetacean phylogenetics. They furthermore suggest that dense taxon sampling, like dense character sampling, can overcome problems in phylogenetic reconstruction.
What Penhallurick & Wink (2004) said in relation to Diomedea exulans and Diomedea epomophora was as follows:
We provide a distance matrix for albatrosses as Table 2. In the text that follows, nucleotide distances are given, with amino acid distances following in brackets. Considering the distances in the Table 2, those between the species that were split by Robertson and Nunn (1998) are much smaller than those between previously recognised "good" species of albatross. For example, within the D. exulans complex, the distance between Robertson and Nunn's D. chionoptera [= nominate exulans] and D. antipodensis is 0.52 % (0.00 %); in the case of their D. exulans [= dabbenena], 0.87 % (0.00 %); and in the case of gibsoni, 0.52 % (0.00 %). D. gibsoni shows a percentage difference of 0.00 % (0.00 %) from D. antipodensis and 0.70 % (0.00 %) from dabbenena. Compare these distances, all of less than 1.0 %, with the distances ranging from 3.2 % to 3.6 % between both D. e. epomophora and D. e. sanfordi from all of the taxa in the exulans complex. We conclude that gibsoni, antipodensis and dabbenena are better recognised as subspecies of D. exulans than as good species in their own right. We note that in the case of antipodensis and gibsoni, both were described as a subspecies of D. exulans in their original description by Robertson and Warham (1992: 74 and 76).
Somewhat surprising is the distance evidence relating to D. amsterdamensis, which has sometimes been treated as a good species since its description by Roux et al. (1983), although Bourne (Table 4 1989: 112) treated it as a subspecies of D. exulans. The fact that it is only 0.52 % (0.00 %) distant from antipodensis, gibsoni and exulans, and only 0.87 % (0.00 %) removed from dabbenena strongly suggests that it belongs among the subspecies of exulans.
In our unpublished reply to Rheindt and Arndt (2005), which was, I believe very unfairly, denied publication, we made the following point:
The Tamura-Nei distance between nominate epomophora and sanfordi is 0.0000%, and the same distance is found between D. e. gibsoni and D. e. antipodensis. This figure suggests that the divergence between these taxa was very recent. Whereas Burg and Croxall (2004), in terms of ‘ESUs’, suggested splitting exulans from antipodensis/ gibsoni and both from dabennena, the TN distances of 0.902% between nominate exulans and dabbenena; of 0.539% between exulans and antipodensis; of 0.539% between exulans and gibsoni; of 0.540% between exulans and amsterdamensis; are well below the TN distance between exulans and epomophora of 3.797%. These data suggest that in terms of the Multidimensional BSC, we have only two species: D. exulans and D. epomophora, although it would seem appropriate to class dabbenena as a semi-species.
On the question of taxon allopatry, Rheindt concludes “The fact that exulans has colonized an island within the range of gibsoni and antipodensis, but fails to interbreed with them, suggests they may have attained prezygotic isolation mechanisms.”(Emphasis added). To claim that because a small number of birds (ca. 10) on Macquarie Island have failed to interbreed with birds on islands 620 km north in the case of gibsoni and 700 km north in the case of antipodensis proves or even strongly suggests that they could never interbreed again stretches the imagination.
Rheindt concludes “tiny MtDNA divergences among Wandering taxa may be an artifact of introgression…”(Emphasis added). Introgression refers to the movement of a gene from one species into the gene pool of another by backcrossing an interspecific hybrid with one of its parents (Dowling and Secor 1997). It is a long-term process; it may take many hybrid generations before the backcrossing occurs. Given the extreme philopatry of all Diomedea albatross taxa, the idea of such lengthy processes of hybridisation, of which there is no evidence whatsoever, seems implausible. I conclude that mention of two possibilities, without any evidence for the reality of either, does not amount to any kind of proof.
In Proposal 388, Rheindt refers to criticism of Penhallurick and Wink (2004) for “rigid use of cyt-b divergence as a species threshold (Rheindt and Austin 2005).” Yet Rheindt himself cites as authoritative a study by Burg and Croxall which appeals to “c.5% divergence between Wandering taxa in the mitochondrial control region”, although he concedes “this locus is known to evolve much faster than the widely used coding genes (ND2, COI, Cyt-b)”….” There is an obvious contradiction here.
Rheindt and Austin (2005) identified several problems under this heading. ‘Firstly, saturation and multiple substitutions are a serious problem as one goes deeper in the phylogeny.’ I agree. Because of its fast rate of mutation, cytochrome-b would be unsuitable for investigating relationships at the order level. But most of the taxonomic issues Penhallurick and Wink were addressing in their 2004 paper concerned levels of divergence of 5% or less. And it is highly unlikely that multiple substitutions, which require two or more changes at a single site, are going to be a problem at that level. The use of Tamura-Nei weighted distances also reduces potential saturation problems.
A second problem identified by Rheindt and Austin (2005) is “the difference in rates of molecular evolution among bird lineages’’. They made the mistake of assuming that Penhallurick and Wink were applying a universal rate of evolution: specifically, that we ‘tend to reject separate species status for any taxon pair with a divergence of <2%.’ What we were actually trying to apply was the principle enunciated by Helbig et al. (2002) in discussing the types of evidence that might be applied to determine the species status of allopatric taxa. They referred to: DNA sequences, and the sum of the character differences corresponding to or exceeding the level of divergence seen in related species that coexist in sympatry. Throughout, when dealing with possible species, we used the distance between well-established species in the same group as a measuring stick. And it was this judgement that guided our decisions on status.
I believe that is both unfair and incorrect of Rheindt and Arndt (2005) to say that Penhallurick and Wink (2004) propose to ‘once again revise’ the albatross taxonomy. In relation to albatrosses, we supported the analysis contained in Mayr & Cottrell (1979) and in Marchant & Higgins (1990, vol. 1. Part A: 264-354, except that that source treated the taxon we call dabbenena as the nominate of D. exulans, and referred to the taxon we call exulans as D. e. chionoptera); and also in del Hoyo, Elliott and Sargatal (1992); Dickinson (2003) and more recently Christidis and Boles (2008). Diomedea exulans gibsoni and Diomedea exulans antipodensis were both described as a subspecies by Robertson & Warham (1992). Since Robertson & Nunn (1998) claimed to be relying on genetic distances, our main concern was to point out that a more reasonable view of distances supported the traditional analysis.
Finally, I would like to point to a glaring lack in relation to proposals submitted to the SACC: specifically, the lack of any criteria for making judgments about whether the difference between taxa should be at the species or subspecies level. In the absence of such criteria, one does not know how to evaluate proposals such as 388. Nowhere in his proposal does Rheindt explicitly state in terms of what species concepts he is making his proposal. It does not appear to be the strictest form of the PSC, which views all terminal taxa as species, since he proposes to treat antipodensis and gibsoni as conspecific. Perhaps he, like Burg and Croxall, is working in terms of the ESU or MU concepts of Moritz (1994a and 1994b). But if the multidimensional BSC is at least implicitly the standard by which such proposals must be judged, I submit that Proposal 388 fails the test.
Abbott, C. L. and Double, M. C. (2003a). Phylogeography of shy and white-capped albatrosses inferred from mitochondrial DNA sequences: implications for population history and taxonomy. Molecular Ecology 12: 2747-2758.
Abbott, C. L. and Double, M. C. (2003b). Genetic structure, conservation genetics and evidence of speciation by range expansion in shy and white-capped albatrosses. Molecular Ecology 12: 2953-2962.
Avise J. C. and Wollenberg, K. (1997). Phylogenetics and the origin of species. Proceedings of the National Academy of Sciences of the USA 94: 7748-7755.
Bourne, W. P. (1989) The evolution, classification and nomenclature of the great albatrosses. Le Gerfaut 79: 105-116.
Burg, T. M. and Croxall, J. P. (2001). Global relationships amongst black-browed and grey-headed albatrosses: analysis of population structure using mitochondrial DNA and microsatellites. Molecular Ecology 10: 2647-2660.
Burg, T. M. and Croxall, J. P. (2004). Global population structure and taxonomy of the wandering albatross species complex. Molecular Ecology 13: 2345-2355.
Christidis, L. and Boles, W. E. (2008). Systematics and Taxonomy of Australian Birds. CSIRO Publishing, Collingwood, Victoria.
Dickinson, E. C. (2003) The Complete Howard and Moore Checklist of the Birds of the World. Rev and enlarged 3rd edn. Princeton University Press, Princeton, New Jersey.
Dowling, T. E. and Secor. C. L. (1997). The role of hybridization and introgression in the diversification of animals. Annual Review of Ecology and Systematics. 28: 593-619.
Helbig, A. J., Knox, A. G., Parkin, D. T., Sangster, G. and Collinson, M. (2002) Guidelines for assigning species rank. Ibis 144: 518-525.
del Hoyo, J., Elliott, A. and Sargatal, J. eds. (1992) Handbook of the Birds of the World, Vol. 1. Lynx Edicions, Barcelona.
Marchant, S. and Higgins, P. J. (1990) Handbook of Australian, New Zealand and Antarctic Birds. Vol. 1. Oxford U. P., Melbourne.
May-Collado, L. M. and Agnarsson, I. (2006) Cytochrome b and Bayesian inference of whale phylogeny Molecular Phylogenetics and Evolution 38, 344-354.
Mayr, E. (1996). What is a species and what is not? Philosophy of Science 63, 262-77.
Mayr, E. and Cottrell, G. W. (1979) 'Checklist of Birds of the World'. Vol. 1, 2nd edn. Museum of Comparative Zoology, Cambridge, Mass.
Moritz, C. (1994a). Defining ‘evolutionarily significant units’ for conservation. Trends in Ecology and Evolution 9: 373-374.
Moritz, C. (1994b). Applications of mitochondrial DNA analysis in conservation: a critical review. Molecular Ecology 3: 401–411.
Penhallurick, J. M and Wink, M. (2004) Analysis of the taxonomy and nomenclature of the Procellariiformes based on complete nucleotide sequences of the mitochondrial cytochrome-b gene. Emu 104: 125-47.
Rheindt, F. E. & Austin, J. J. (2005) Major analytical and conceptual shortcomings in a recent revision of the Procellariiformes – a reply to Penhallurick and Wink (2004). Emu 105: 181-186.
Robertson, C. J. R. and Nunn, G. B. (1998) Towards a new taxonomy for albatrosses In Albatross biology and conservation (Eds. G. Robertson and R. Gales) pp. 13-19. Beatty and Sons, Chipping Norton.
Robertson, C. J. R. and Warham, J. (1992) Nomenclature of the New Zealand Wandering Albatrosses. Bulletin of the British Ornithologists' Club. 112 (2): 74 - 81.
Roux, J.-P., Jouventin, P., Mougin, J.-L., Stahl, J.-C., and Weimerskirch, H. (1983) Un nouvel albatros Diomedea amsterdamensis n. sp. découvert sur l'Ile Amsterdam (37o50'S,77o35'E). Oiseau 53: 1-11.
Schodde, R. and Mason, I. J. (1999) The directory of Australian birds: passerines. CSIRO Publishing, Collingwood, Vic.
Vuilleumier, F., Le Croix, M. and Mayr, E. (1992) New species of birds described from 1981-1990. Bulletin of the British Ornithologists’ Club, Centenary Supplement, 112A: 267-309.
Rheindt’s response to the comments by Penhallurick:
Penhallurick’s comments are rather antagonistic and suggest my proposal seeks to establish one taxonomic treatment of Wandering Albatross at the expense of the other. However, my main conclusion states that – as far as the “1-species” or “4-species” solutions are concerned – “…I do not feel there is overwhelming evidence for either treatment…” So essentially, I concur with Penhallurick to the point that his favored “1-species solution” may well turn out to be the best treatment at some point in the future when more evidence is available. However, with the current evidence at hand, I believe that the 4-species solution would make more sense within the framework of the Biological Species Concept, for reasons expounded in the proposal.
In his interpretation of my proposal, Penhallurick has misunderstood a number of important points and missed the significance of others. He dwells on particulars of a scientific debate carried out in the journal Emu in 2004/05 (Penhallurick and Wink 2004; Rheindt and Austin 2005) that – in my opinion – has only marginal relevance to the current proposal, resulting in the mix-up of a number of unrelated issues. For instance, Burg and Croxall’s (2004) incidental finding of 5% control region divergence is not the reason why my proposal cites this study as supporting BSC species status of Wandering taxa. Rather, the phylogenetic structure these authors found among Wandering taxa is used in conjunction with ecological observations (range sympatry) to advance this case.
I do wish to offer a clarification on four central points brought forth by Penhallurick.
(1) Penhallurick asks SACC members to disregard some of the most pertinent research carried out on Wandering Albatross systematics because the authors phrased their results under a different species concept. I believe we should not throw out these detailed studies on Wandering Albatross phylogenetic structure only because their authors opt to describe their findings in terms of “ESUs” rather than biological species. Despite Penhallurick’s concern, SACC members have the ability of re-interpreting those results under their own species concept.
(2) Similarly, Penhallurick is confused about which species concept underlies the present proposal. In line with SACC’s adoption of the multi-dimensional BSC, my proposal followed the other 387 SACC proposals in utilizing this species concept. This is obvious by mention of such terms as “essential allopatry”, which would be immaterial to a debate within the confines of most other species concepts widely used in ornithology.
(3) Penhallurick states that I write “approvingly” of Robertson and Nunn (1998) and Burg and Croxall (2004). Apart from being irrelevant, this is not true. In fact, if anything, I wrote disapprovingly of Robertson and Nunn’s (1998) “controversial book chapter”. Penhallurick then goes into a lengthy discussion on how the conservation community has welcomed or pushed the treatment of Wandering taxa as separate species. Here again, I fail to see the relevance to the current SACC discussion. Just as conservation concerns should not motivate a split in albatross taxa, the illegitimacy of this interference should not be used as an argument to motivate a lump.
(4) Most significantly, I take issue with Penhallurick’s portrayal of genetic introgression in albatrosses as an “…implausible…” and “…lengthy process of hybridization of which there is no evidence whatsoever…” Introgression rather than conspecificity as a cause of near-zero mtDNA divergences among Wandering taxa is not just a theoretical exercise, but also a plausible scenario. The following three points are meant to illustrate that genetic introgression is not a marginal process, but an all-encompassing phenomenon:
a. MtDNA introgression is pervasive in the biological world and greatly diminishes the utility of low mtDNA divergences as a true yardstick of taxon divergence. (Note that high divergences are less gravely affected). Judging by the inflationary number of studies reporting on introgression, it seems to occur in most animal species in one form or the other. In birds alone, patterns of genetic introgression have been detected across the whole taxonomic spectrum, from passerines to Galloanseres. A small number of example studies include the following (full citation not provided): Gill 1997; Brumfield et al. 2001; Helbig et al. 2001, 2005; Rohwer et al. 2001; Bensch et al. 2002; Sætre et al. 2003; Lovette 2004; Kulikova et al. 2004; Mank et al. 2004; Shapiro et al. 2004; Grant et al. 2004; Dabrowski et al. 2005; Borge et al. 2005; Secondi et al. 2006; Kvist and Rytkönen 2006; Vallender et al. 2007; Peters et al. 2007; Martínez-Cruz and Godoy 2007; Rheindt et al. 2009; Carling and Brumfield 2008, 2009; Gay et al. 2008.
b. As pointed out in the original proposal, mtDNA introgression can affect all/most members of entire radiations, such as Anas ducks (Johnson and Sorenson 1999, Peters et al. 2007) and Larus gulls (Liebers et al. 2004). Plumage differences between various Larus gulls are arguably smaller than between an adult male Snowy Albatross (D. exulans) and his neotenous cousin from the north (e.g. D. amsterdamensis). While introgression is widely accepted as the cause for near-zero mtDNA divergences in Larus, why should it not be plausible in Diomedea? If we were to subscribe to Penhallurick’s “divergence-only” approach, we would also have to be prepared to challenge the biological species status of Greater and Lesser Black-backed Gulls based on their near-zero mtDNA divergence (Liebers et al. 2004).
c. Penhallurick paints genetic introgression as a long-term process requiring many generations, and thus finds it to be an “implausible” phenomenon in albatrosses. Penhallurick errs on two accounts. Firstly, genetic introgression can be extremely fast. For an avian example, take Mank et al.’s (2004; Conservation Genetics 5) research showing how Mottled Duck and Mallard microsatellites went from distinct to almost identical within only 58 years. Secondly, Penhallurick’s statement belies the vast time scales of the evolutionary process. Why – in the last tens of thousands of years – could there not have been, say, a climate-induced 500-year period (or repeated episodes of such periods) during which exulans expanded into the ranges of the other Wandering taxa and occasionally hybridized with them, leading to introgression from one taxon into the other and a “re-setting” of mtDNA divergence to zero… There is no reason to presume introgression in Wandering Albatrosses must have happened fast and now.
Based on these considerations, I re-iterate my view that near-zero mtDNA divergences in Wandering Albatross taxa may not be too informative to this taxonomic debate, and that other lines of enquiry (e.g. their essentially sympatric occurrence around New Zealand) may provide more information on their BSC status.
Comments from Stiles: “YES. I think that the 4-species treatment is the best way to express what we know about these birds, and I think that Rheindt has satisfactorily answered the objections of Penhallurick.”
Additional comments from Penhallurick: “Some final comments on Rheindt’s Proposal. If we consider Rheindt’s proposal, and ignore the appeals to irrelevant papers, the core of his proposal comes down to two statements of “possibilities”. He states that introgression “may” explain the low DNA distances. He offers no actual evidence that this has occurred. And he needs to explain not just how introgression occurred between two Wandering Albatross taxa but between all of them. This beggars the imagination. Secondly, we are invited to assume that because a small number of exulans are present on Macquarie Island, this represents breeding isolation between that taxon and taxa on Subantarctic Islands some 700 km away. Think about this! It would be laughable in any serious discussion. Again, Rheindt presents no evidence that this most unlikely convergence ever occurred. I thought science was about evidence. And Rheindt has presented none!”
Additional comments from Rheindt: “Please see Liebers et al. (2004; Proc. R. Soc. Lond. B) for a well-documented example of a species swarm of approximately a dozen BSC species of Larus gull that all share near-identical mtDNA, presumably as a result of introgression. Similar patterns (but with less comprehensive documentation) have been detected in Anas duck mtDNA - see previous contribution. It is questionable why such a pattern should ‘...beggar imagination...’ in albatrosses."
Comments from John Croxall (12 Jan. 2010): “I was recently informed of the existence of this proposal and although I would not normally get involved in such issues, I felt that the SACC might find a few comments useful to help it reach a decision. (For the record: (a) I have undertaken research on albatrosses and petrels for some 30 years to my retirement in 2006; (b) I am not a taxonomist; such work on albatrosses that I have co-published were by-products of studies on population genetics and parentage largely conducted by Theresa Burg; (c) I have never met, or refereed papers by, Rheindt or Penhallurick!).
Additional published material
“There are at least 4 supplementary sources of information that are relevant to your deliberations:
“1. The very recent paper (despite its June 2009 date, the issue only appeared in November) by Chambers et al. (2009) Phylogenetic analysis of the 24 named albatross taxa based on full mitochondrial cytochrome b DNA sequences. Notornis 56: 82-94. To me, on a quick perusal (and despite its abundant typographic errors), this provides a good and pragmatic overview of many of the issues, not least those involved in considering the approach and conclusions propounded by Penhallurick and Wink (2004). While the paper does not consider relevant biological (especially ecological) information (see below), I believe the essence of its conclusions in respect to the taxa relating to your proposal are entirely realistic and would support the proposal for splitting. I attach a copy of the paper.
“2. The paper by Techow et al. (2010) Speciation and phylogeography of giant petrels Macronectes. Molecular Phylogenetics and Evolution 54: 472-487, as this has considerable bearing on the issue of cytochrome b sequence divergence (see below). I attach a copy.
“3. The publications of the Taxonomic Working Group of the Agreement on the Conservation of Albatrosses & Petrels (ACAP), which has reviewed several of the Diomedea issues relevant to your proposal, specifically Diomedea antipodensis and gibsoni at its 2006 meeting and D. exulans and D. amsterdamensis at the 2008 meeting. They have not yet reviewed D. exulans/D. dabbenena. However, given the principles and precedents involved and that Burg and Croxall (2004) found there was more genetic distinction between dabbenena and exulans (sensu stricto) than between exulans and antipodensis/gibsoni, it would seem unlikely that the ACAP Taxonomic Working Group would not support this split (it is the existing ACAP position). I attach the documents (AC2, AC3, AC4) from to the relevant meetings of this Working Group; these are available publicly through the ACAP website. I am not a member of this Taxonomic Working Group but its discussions are relevant to BirdLife International (see below), for which I am the ACAP representative to its Advisory Committee and Meeting of Parties.
“4. In respect of Diomedea dabbenena, the publication by Cuthbert et al. (2003) Separating the Tristan albatross and the wandering albatross using morphometric measurements. Waterbirds 26: 338-344, is also relevant. Tristan albatross, although showing reasonable genetic divergence, is not easy to distinguish in the field from D. exulans, without considerable experience of age and sex-related plumage patterns of the latter. In essence, their ranges overlap at sea only in the wintering grounds, particularly in the Brazilian and Benguela Current systems where Tristan albatross may co-occur with D. exulans from South Georgia towards the northern limits of the latter’s wintering range. There are numerous recent observations and uplinks from tracking studies confirming the regular occurrence of Tristan albatross in the SACC area.
“5. SACC also raises the issue of D. amsterdamensis, material of which was unavailable to Burg and Croxall (2004); some relevant data were provided in Milot et al. (2007) Proc. Roy. Soc. Lond. Series B 274: 1779-1787 and I understand there is a paper forthcoming which summarises all data relevant to its distinctiveness at the species level. However this species is, to date, unrecorded for the SACC area (indeed it has quite a restricted (at all seasons) main distribution at sea in the Indian Ocean).
Comments on the proposal to SACC
“1. The statement (New Information para 1) that “despite widespread criticism … popular bird books” is incorrect. First, while Penhallurick and Wink (2004) – on rather tendentious grounds – and Christidis & Boles (2008) – on grounds which may equally reflect lack of any real experience of the taxa involved (and both representing at most personal/ regional perspectives on this complex)- were critical of the adoption of the Diomedea “splits”, the vast majority of the scientific community has accepted them, as evidenced by references in the reports of the ACAP Taxonomic Working Group. Second, this acceptance extends to recent monographic treatments, e.g. Tickell (2000); Brooke (2004); Onley and Scofield (2007). The authors of these works would be somewhat dismayed to be regarded as non-scientists! Tickell was a pioneer researcher on albatrosses as long ago as 1958; Brooke has published for over 20 years on avian ecology, especially with respect to seabirds, including describing new procellariiform taxa; Scofield is a leading New Zealand museum-based researcher with wide field experience of albatrosses and petrels. Thus all recent global reviews have chosen to maintain most of the splits in Diomedea taxa that are the subject of your proposal.
“2. The BirdLife International position is as follows. As the authority for the IUCN Red List, BirdLife maintains a Taxonomic Working Group (BTWG). For regions with authoritative checklist/classification committees, these are very closely followed by BirdLife. However, for taxonomic groups of more global distribution, BirdLife prefers to use the latest authoritative global review. For Procellariiformes, the default authority is Brooke (2004), with updates as determined by BTWG in its annual review process. For albatrosses, however, BirdLife currently follows the advice of the ACAP Taxonomic Working Group, regarding this as the most appropriate taxonomic authority for the species covered by this Agreement.
“1. It is increasingly recognised that, compared with most avian taxa, Diomedeidae, probably Procellariidae and possibly Procellariiformes, show unusually low levels of sequence divergence even between well-established species.
“2. This topic is reviewed and discussed for albatrosses by Milot et al. (2007). The difficulty, of course, is how to compare distances between taxa that are mainly (indeed entirely for all problematic cases) allopatric, given the limited guidance available from congeneric sympatric taxa. For me, there are two main guidelines here. First, Black-browed and Campbell Albatross, which now breed sympatrically at Campbell Island, are separated by 0.8%. Second, the two giant petrels are separated by about 0.4%. The latter is quite an important result because we know from extensive (multi-year of banded individuals) field studies of these cryptic sibling species that hybridisation is almost non-existent (the few cases are invariably Southern males with Northern females), especially noteworthy considering that the loose breeding aggregations of both species often overlap spatially, that the observable discriminating morphological characteristics are essentially only in bill colour and that their annual cycles are only staggered temporally by some 5-6 weeks. Everyone who has worked in the field with these two species where they breed sympatrically recognises the correctness of their discrimination by Bourne and Warham (1966) as distinct species. It is remarkable that Penhallurick and Wink (2004) chose to ignore the wealth of biological and ecological data on these taxa – literally a dozen or more publications whereby species distinctiveness had not only never been questioned but invariably clearly substantiated- in favour of applying some arbitrary genetic distance standard.
“Even recognising that Macronectes is in a different family of the Order, it nevertheless has many biological and ecological similarities to albatrosses (diurnal, surface breeders with vocal and plumage characteristics highly conserved), and emphasises that procellariiform taxa which may look almost identical to us and which have very low cytochrome b divergence, may still behave as perfectly good biological species.
“This is not the place to rehearse the arguments for the taxonomic status of the Diomedea and Thalassarche taxa, perhaps beyond recognising that there are probably three difficult cases, for most of which more data would be useful:
(a) Diomedea antipodensis/gibsoni, discussed in some detail by Chambers et al (2009), where the ACAP TWG decision (in AC2 paper) was to retain as subspecies, the identical cytochrome b sequences and limited microsatellite differences finding more favour than the plumage differences at most ages and the very distinct foraging areas and migrations outside the breeding season (the former taxon occurring regularly (annually on migration) in the SACC area; the latter mainly restricted to the Tasman Sea when not breeding)).
(b) D. epomophora/sanfordi (both of which occur in the SACC region), whose distinctive plumages render them diagnosable at almost all ages and stages yet cytochrome b differences are very low (0.08%) and hybridisation does occur (albeit at the not entirely “natural” site of Taiaroa Head). The ACAP TWG addressed this species pair in its AC4 paper and maintained full species status for both. [For the record, based on my experience with Thalassarche melanophrys and T. chrysostoma, hybridisation events can occasionally take place between species which are as different as these two, if there are mixed colonies with substantially disproportionate numbers of the two species involved. In Diomedea, the relatively high frequency of extra-pair copulations, including forced copulations, may also lead to hybrid pairs at sites where individuals of one species are sometimes attracted to breeding sites populated by another species].
(c) Thalassarche cauta and T. steadi (the latter occurring in the SACC region) which currently lack 100% reliable morphometric or plumage diagnostics for all ages and stages (but bill colour allows discrimination in almost all cases and breeding seasons differ by about 8 weeks) and have cytochrome b distances of about 0.2%. However, these two taxa have been the subject of a dedicated phylogeographic PhD thesis and the ACAP Taxonomic Working Group (in its AC3 paper) reached a unanimous decision based on the published papers there from; consequently BirdLife has followed this decision.
“I do not wish to extend these comments to address issues raised by Penhallurick. However, I would simply observe that, whatever the merits of the higher order taxonomic and nomenclatural modifications suggested in the Penhallurick and Wink (2004) paper, over 100 albatross and petrel scientists, meeting at the Fourth International Albatross and Petrel Conference in Uruguay in 2004, were unanimous in agreeing that most, if not all, of the conclusions relating to the species status of albatrosses, giant petrels, and prions were in direct disagreement with the experience of those scientists who had worked most extensively with these species in the field. It is evident that the Penhallurick and Wink (2004) conclusions disregarded the substantial body of biological and ecological information available on the distinctiveness of most of the taxa in the above groups, where modern data on genetics, vocalisations, ecology, core foraging areas etc. are demonstrating the distinctiveness of these mainly allopatric taxa in ways consistent with their recognition at species level under all of the currently accepted concepts.”
Additional comments from Robbins: “Given Rheindt and Croxall’s comments, I see no reason to change my original vote of supporting recognition of four species.”
Additional comments from Stiles: “Here, I merely reiterate my YES vote, especially in view of Croxall’s reply to Penhallurick’s latest salvo. The fact that a group of world experts on the albatrosses has collectively supported species status for these forms is not to be lightly dismissed. As a definite non-expert I am strongly inclined to accept their verdict.”
Comments from Schulenberg: “YES. As several have noted, this case is a tough nut to crack, as the biological species concept is poorly suited to treat essentially allopatric taxa, and other taxonomies are reasonable. All that said, similar as all these taxa are, I am impressed by the differences among them, both in terms of genetic structure between populations, and in terms of trends towards neoteny in plumages (correlated with latitude). Therefore I think that recognizing four species (including the extralimital amsterdamensis) is the best approach. Also, while fully aware of the distance between Macquarie Island (breeding site for exulans) and Campbell and Auckland islands (breeding sites for antipodensis/gibsoni), I'm more impressed by the relative proximity of these sites. Low genetic distances notwithstanding, the morphological and ecological differences point to me towards recognizing these taxa as species.
“The question of English names is not, to my knowledge, before us. But if SACC accepts the split, then I foresee issues with English names. The name "Wandering Albatross" has long standing, and maybe there will be resistance to changing it. But the name "Wandering Albatross" would prove useful for (the great many) individuals of the complex (sensu lato) that might not be identifiable in the field. In turn, a different English name would be called for to use for Diomedea exulans (sensu stricto). Onley and Scofield (2007) use "Snowy Albatross" for exulans. I don't think that this has been followed by other authors, but ... they may be on to something.”
Comments from Remsen: “NO. My understanding of these albatrosses is that they will breed in subadult plumages with fully adult-plumaged individuals. If that is incorrect, please let me know. If this is true, then the neotenic plumage differences among island populations seem unlikely to act as any barrier to gene flow, consistent with the low degrees of genetic divergence. The correlation with latitude suggests an ecological response among the populations rather than any intrinsic biological difference that indicates that they are lineages that have reached levels of divergence associated with species rank in other birds. As for differences among island populations in timing of breeding, migration routes, and wintering areas, such differences occur among island populations of many seabirds that no one would consider to consist of multiple species, and are therefore not really persuasive on their own. To a non-seabird biologist such as myself, all I see is discrete geographic variation in plumage characters correlated with an ecological variable. Vocal differences are alluded to, but I would like to see published data on this. Seabird biologists working with shearwaters have been tuned in to vocal differences and their importance for some time, but I am not aware of a parallel emphasis in albatrosses – if this only reveals my ignorance, someone please correct me.
“Having discussed species limits with a number of seabird biologists, they have openly lobbied me for species rank for populations because that would increase their status in terms of conservation. Thus, their taxonomy has an agenda, driven by politics, and this arouses my suspicions. I do not want to accuse any of the 100 albatross and petrel scientists (mentioned by Croxall) of such non-scientific, subjective reasoning, but I am concerned that such reasoning may enter their decision of endorsement of narrower species limits, if only subconsciously. I’d also like to know how many of them have training in systematics rather than just ecology. Such comments will undoubtedly insult this active and extensive community of researchers, but I only mean to provoke production of the sort of data that will fortify these taxonomic decisions … particularly in this case data on actual isolating mechanisms (voice and display?). Finally, I want to emphasize that my vote on this has nothing to do with degree of genetic (mtDNA) differentiation.”
Comments from Cadena: “[NO]. I have remained on the sideline regarding this proposal because I know close to nothing about the biology of these birds and wanted to have comments from most committee members and others, hoping they would illustrate me. I am actually uncertain as to how to vote, so hopefully the following comments will stimulate some additional discussion that may finally lead us to reach a decision soon. What I see here is that much of the Rheindt vs. Penhallurick discussion boils down to a rather simple issue: Rheindt is only saying that there are various reasons why low (or no) mtDNA differentiation need not imply conspecificity and that because of those reasons one ought to look at other sources of data to decide how to rank all these forms. I have to say I agree with Rheindt on this 100%. Then, what else do we have? I am not certain about all the details, but as Van puts it, the strongest evidence for splitting these forms are differences in plumage and, based on his first remark, plumage variation is unlikely to be a barrier to gene flow (right?). To me, differences in timing of breeding are persuasive arguments for species status for sympatric populations, but not for populations existing in allopatry that may be tracking changes in e.g. resource availability and lack fixed genetic differences in genes controlling breeding behavior, such that they would mate if they were to come into contact (e.g. Ecuadorian populations of Zonotrichia capensis separated by 25 km differ drastically in breeding seasons as a result of climatic differences between sites, but nobody seems to have claimed these should be considered different species). Alvaro mentioned differences in water temperature between sites as a possible isolating factor, but this to me seems like a difference in the environment, not a difference in the organisms, so using it as an indication of reproductive isolation (following the BSC) appears problematic as essentially all biological species show some ecological discontinuities in their ranges and comprise populations occupying ecologically different environments (take again the Zonotrichia case, with populations occurring from sea level to very high-elevation areas). Is there anything else? If not, I think my vote would lean towards NO. I realize this might challenge the opinion of albatross experts mentioned in the proposal, but if these experts were not weighting the evidence for reproductive isolation following the criteria we try to follow here, then I would be fine with reaching a different conclusion.”
Comments from Nores: NO. Because it depended on my vote whether the Wandering Albatross
were considered 1 or 4 species, I have devoted considerable time to analyze the
fundamentals given by Rheindt, Penhallurick, Croxall, and the other members of
the Committee, supporting one position or another, in order to make a decision.
Although conservation of the species cannot be a reason for analyzing taxonomic
issues, I had initially thought of voting YES because seemed there were good
reasons for either treatment, and an affirmative vote would favor conservation
of the taxa. However, I now
consider that the reasons for keeping all the taxa within a single species are
more important than those for splitting them. I want to make it clear that this
decision does not mean to contradict a number of experts in marine birds
(Tickell, Brooke, Onley, Scofield, Croxall or any of the 100 albatross and
petrel scientists mentioned by the latter) because I believe that this is
ultimately a taxonomic matter rather than one of the biology of the species.
There seem to me to be two fundamental reasons for keeping all the populations
in a single species:
1) Taxa allopatry: the breeding areas of the five taxa clearly show that all of them are allopatric. D. e. exulans occurs in the southern part of the species range (Southern Georgias, Crozet, Kerguelen, Prince Edwards and Macquarie islands), whereas the all other taxa are to the north: Tristan da Cunha and Gough I. (dabennena), Amsterdam I. (amsterdamensis), Auckland I. (gibsoni) and Antipodes I. (antipodensis) (see Figure), which mated the typical distribution of subspecies.
The fact that exulans has been breeding in Macquarie Island, 620 km from the nearest gibsoni and 700 km from the nearest antipodensis, is not a proof of sympatry as suggested by Rheindt. This author has also mentioned “The fact that exulans has colonized an island within the range of gibsoni and antipodensis, but fails to interbreed with them, suggests they may have attained prezygotic isolation mechanisms”. I agree with Penhallurick that to claim that a small number of birds (ca. 10) on Macquarie Island having failed to interbreed with birds on islands 620 km north, in the case of gibsoni, and 700 km north in the case of antipodensis suggests that they could never interbreed is seems implausible.
2) Low Molecular divergence: the five Wandering Albatross taxa are characterized by very low divergences. The Tamura-Nei distance between D. e. gibsoni and D. e. antipodensis is 0.0000%. The TN distances of 0.902% between nominate exulans and dabennena, of 0.539% between exulans and antipodensis, of 0.539% between exulans and gibsoni and of 0.540% between exulans and amsterdamensis are well below the TN distance between exulans and epomophora of 3.797% These data suggest that in terms of the Multidimensional BSC, we have only two species: D. exulans and D. epomophora (Comments from Penhallurick). Although Burg and Croxall (2004) found c. 5% divergence between Wandering taxa in the mitochondrial control region, this locus is known to evolve much faster than the widely used coding genes (ND2, COI, cyt-b) and, therefore, cannot be compared (Comments from Rheindt). Although Rheindt concluded that “tiny MtDNA divergences among Wandering taxa may be an artifact of introgression (and he may be correct), he offered no actual evidence that this has occurred. He only offered examples in other birds (e.g. Anas ducks – Johnson and Sorenson 1999; Larus gulls – Liebers et al. 2004). Moreover, he needs to explain not just how introgression occurred between two Wandering Albatross taxa but between all of them.