albatross taxonomy and more

Subject: albatross taxonomy and more
From: David James <>
Date: Mon, 28 Feb 2000 14:53:33 +1000
Tony Russell has opened up that big can of worms again, albatross taxonomy.
What may seem like a simple question has no simple answer. I think the
issue requires considered answers, and, at the risk of boring some of you,
I'm going to rave on some philosophical issues covering science, taxonomic
principles, and conservation. So be warned and tune out now if this is not
for you. This is long

The taxonomy of albatrosses was stabilised, in part, in a unique fashion in
1965 when 15 seabird gurus from around the world published a taxonomy of
albatrosses which recognised 13 species in two genera (Alexander et al
1965). This was the taxonomy used universally until a couple of years ago,
but is now dated. The only major change for many years was the description
of the form of "great" albatross breeding on Amsterdam Is, Indian Ocean, as
a new species, Diomedea amsterdamensis, by French seabird biologists (Roux
et al. 1983).

Robertson & Warham (1993), recognising that geographical variation in
Wandering Albatross was very much overlooked, described two new subspecies
of Wandering Albatross from New Zealand's Subantarctic islands, namely
Diomedea exulans antipodensis from the Antipodes Islands and D. e. gibsoni
from the Auckland Islands. At the albatross conference held in Hobart in
September 1995, Robertson & Nunn presented a radical revision of
albatrosses recognising four genera (they resurrected two old genus names,
Thalassarche and Phoebastria, both originally proposed by Reichenbach in
1852) and elevating all existing terminal taxa (see SPECIES CONCEPTS below)
to species level. Thus, they recognised 24 species instead of 14. Their
presentation was published in the conference proceedings three years later
(Robertson & Nunn 1998).

Neither the original presentation, nor the published paper contained any
data in support of the proposed taxonomy. This is a very important point,
NO DATA. However, Nunn et al (1996) published a revision of Albatrosses at
the generic level based on molecular data (cytochrome B analysis) which
provides an argument in support of the recognition of four genera.

Given the lack of data in Robertson & Nunn (1998), the main argument in
support of their proposal was that the species concept previously used for
Albatross taxonomy is inappropriate. Thus, they proposed a change from
using the "biological species concept" (BSC) to using the "phylogenetic
species concept" (PSC).  There is a lot of international scientific debate
about species concepts at present and I am no expert, but I will wade in to
the shallows here.

The BSC is the concept behind most bird taxonomy in the latter 20th
century. It is used for example in the Birds Australia Checklist
(Christidis & Boles 1994) and its predecessors, and Peters and Clements
world checklists, and Schodde & Mason's (1999) new <Directory of Australian
Birds>. In the BSC the finest level of evolutionary divergence is the
<subspecies>. Species can be either polytypic (having two or more
subspecies) or monotypic (having no subspecies). Both subspecies and
monotypic species are "terminal taxa", ie the finest level of divergence.
("Taxon" or the plural "taxa" refers to any unit in taxonomic
classification such as genus, species or subspecies). In contrast the PSC
does not recognise subspecies at all but treats the finest level of
divergence (terminal taxa) as species. Thus a monotypic BSC species is a
PSC species, a BSC subspecies is also a PSC species and a polytypic BSC
species is several PSC species. And so the PSC recognises 24 species of
albatross whereas the BSC recognises 14. They both recognise 24 terminal
taxa at present. The PSC would also recognise perhaps 25,000 species of
bird world wide (no one knows for sure) versus 9-10,000 in the BSC.

The reason for these differences is the way the two concepts define
"species". The BSC uses (in theory) hybridisation or lack there of to
decide if two taxa are one species or not. Unfortunately this does not work
at all for allopatric forms (ie forms that have separate ranges such as
Eastern and Western Yellow Robins) and so in practice taxonomists have
estimated whether or not forms are conspecific (the same species) based on
how morphologically similar they are. The PSC gets around this nicely by
defining species as the finest unit of evolutionary diversity that can be
diagnosed. However, it gets in to trouble with sympatric forms (forms whose
ranges overlap) that hybridise extensively. Hybrids are neither one species
nor the other, they are just ignored. Clines (where there is continuous
gradation from one form to another as in Fairy Gerygone) go in the too-hard
basket. Another problem of the PSC is that it masks relationships between
terminal taxa. When all terminal taxa are treated as separate species there
is no information about which terminal taxa are most closely related to
which and which are not closely related to any others.

A major criticism of the BSC is that it does not serve conservation very
well. Conservation needs to occur at the finest level of biodiversity and
BSC species are not necessarily units of biodiversity. On the other hand
PSC species are, or nearly are, units of biodiversity. Around the world,
people and governments talk about "threatened SPECIES". Outside of
Australia, most legislation protects only to the level of species. Thus,
under the BSC, subspecies (which are units of biodiversity) are not
protected adequately. The public and politicians do not understand this, so
the PSC sidesteps the problem and redefines species as units of biodiversity.

The PSC is perhaps well suited to albatrosses because the various forms are
allopatric. However, as Trevor Worthy (a NZ taxonomist) has pointed out,
albatrosses are a great example of a group showing all levels of divergence
from family to terminal taxon. Perhaps a ten-tiered system would not do
justice to the family of albatrosses (the PSC is two- tiered [genus &
species], the BSC is three-tiered). Looking at the big picture, adopting
the PSC for all groups of birds world wide would create a little turmoil
and uncertainty to say the least.


To oversimplify, there are generally two levels of taxonomic revision.
Primary revision involves new research and presents new arguments and novel
reviews, generally of a small group of birds at a time. Then, secondary
reviews, or checklists, assess published arguments (in primary reviews) on
their merits and accept or reject them. (In practice, publications like
Schodde & Mason 1999 and Sibley et al's works incorporate both into one).

Christidis & Boles (1994) was a landmark taxonomic revision of exceedingly
high scientific merit that presented no new data and presented no novel
taxonomic arrangements or conclusions. It merely reviewed the taxonomic
literature since 1975, assessed each argument on its merits and followed
the arguments to their logical conclusions (there are a few mistakes of
course). It used the BSC.

So, if the Christidis & Boles process was applied to albatrosses the
following would happen.
1). The species concept used by Robertson & Nunn would be rejected for the
sake of consistency with the rest of the families in the checklist.
2).  The arguments of Robertson & Nunn (1998) would be assessed on their
merits and rejected as unsubstantiated.

3). The generic revision of Nunn et al (1996) would be assessed on the
merits of the molecular data presented. Not being a gell-jockey it seems
rather good to me, but Les Christidis tells me that it is not very
convincing data in his opinion.

Schodde & Mason also reviewed the literature and used the BSC, but also
presented original data in their recent checklist. They do not have access
to enough material to do a new and comprehensive revision of albatross
taxonomy world wide, so they would be in the same situation as Christidis &

However, neither Christidis & Boles nor Schodde & Mason have a monopoly on
taxonomy nor do they have the final say. Not only is it an international
arena, but also anyone is free to follow whatever taxonomy they wish.
Christidis & Boles enjoys its status as the "official" Australian list only
by consensus, but that consensus is being eroded by a string of checklists
produced by CSIRO (Schodde & Mason 1997; 1999; Stanger et al. 1998).

Robertson & Nunn has been accepted as an "interim taxonomy" by some
management and research organisations around the world, but mostly those in
Australia. Environment Australia and the Australian Bird and Bat Banding
Scheme The Antarctic Division and SOSSA immediately accepted the taxonomy
before it was even published. The taxonomy has been adopted by Stephen
Garnett as the basis for the Draft Action plan for Australian birds and is
set to be enshrined in Australia's endangered species legislation. I find
this latter situation very alarming.

Peter Milburn wrote on Birding Aus
>>My personal view is that since much research is in progress we have to be
satisfied with the INTERIM taxonomy as has been adopted.  It has to be
emphasized that changes are likely so one may only be authorative about the
fact that revisions were and probably still are necessary>>.

I can not agree at all. As a scientist I was trained to question. When I
question this subject I find no basis. I am not satisfied at all with a
taxonomy based on NO DATA. I have talked to quite a few taxonomists and
<marine ornithologists> around the world who feel likewise, so I must also
disagree entirely with Peter's statement that the interim taxonomy is being
followed <in most of the world by marine ornithologists>. Only disciples
and the ignorant would follow a taxonomy unsupported by data.


At 10:09 24/02/00 +1000, Tony Russell wrote:

>>Is there anyone out there who can authoritatively advise as to the
present status of various albatross species / subspecies? ie, which are now
full species and which are still subspecies?

Tony, the only person who can give you an answer is yourself, because it is
a matter of opinion a matter of definition, not a matter of fact.
Fortunately, the taxonomy that you or anyone else uses does not change the
birds. Firstly , you have to decide on a definition of species, but even
then, species boundaries are still in dispute.

>>In the absence of trinomials I find these very difficult to sort out and
they seem to be changing at a rate that the literature can't keep up with.
Are the Robertson and Nunn proposals being adopted or have they fallen by
the wayside?

Depends who you talk to, again a matter of opinion. If data are presented
maybe it will substantiate the taxonomy. but meantime? (BTW, there are no
trinomials in a PSC classification.)

>The birds in question are as follows:
>Diomedea exulans               - Wandering Albatross
A problem. The oldest (senior) name for wandering Albatross in the broad
sense. The type specimen (which links the name to the actual form the name
belongs to) was collected somewhere in the world and the specimen has long
since been lost. It was not known what island it was hatched on and not
described well enough to help. Mathews subsequently restricted the type
locality to Cape of Good Hope, which still doesn't help. Under the BSC D.
exulans remains a species (Wandering Albatross), but under the PSC it may
have to be declared indeterminate and sink into oblivion. Alternatively it
could be considered to represent the Tristam population, in which case D.
dabennena would be come a junior synonym. There can not be both. The
outcome depends on the quirks of the code of zoological nomenclature.
Robertson & Nunn did not seem to deal with this problem.

The "exulans" complex consists of
>Diomedea antipodensis          - Antipodean Alb.
>Diomedea gibsoni               - Gibson's Alb.
>Diomedea dabennena- Tristam Alb. (or D. exulans?).
>Diomedea amsterdamensis        - Amsterdam Alb.

These four forms are closely related, the first 3 formerly treated as
subspecies of Wandering Albatross, the fourth treated as a separate species
since described in 1983. They are all small, low latitude breeders. They
are neotonous to varying degrees, meaning that they become adult and breed
with traces of (ancestoral) juvenile characters such as brown plumage and
dark cutting edges to the bill.

These forms are particularly poorly defined and diagnosed, and Sandy Bartle
(curator of birds at the Te Papa National Museum of NZ) insists that
specimens of gibsoni and antipodensis can not be reliably identified on the
museum bench. If this is so, then under the PSC they should not be treated
as separate species. Under the BSC all four would be treated as subspecies
of one species, but would they be known as D. dabennena or D. exulans?  Or
even D. chionoptera?

>Diomedea chionoptera? Type locality Kerguelan Is.
Formerly included in wandering albatross, this form, the <Snowy Albatross>
that breeds in higher latitude islands (closer to the pole), is much bigger
and is more white when adult. An argument could be mounted for treating
this as a separate BSC species to the above group based on morphology and
ecology, but none has presented.

>Diomedea epomophora            - Southern Royal Alb.
>Diomedea sanfordi              - Northern Royal Alb.
Formerly subspecies of one species. An argument could be mounted for
treating them as separate BSC species, but none has been presented.

>Thalassarche impavida  - Campbell Alb.
Formerly the Campbell is. ssp of Black-browed. Apparently a fair bit of
hybridisation and gene exchange allegedly occurs between populations so
this one might struggle to be recognised as more than a ssp under the BSC.

>Thalassarche chlororhynchos    - Atlantic Yellow-nosed Alb.
>Thalassarche bassi (= T. carteri)              - Indian Yellow-nosed Alb.
Formerly conspecific, split by Robertson & Nunn. The name carteri
Rothschild 1903 is senior to bassi Mathews 1912.

>Thalassarche cauta             - Shy Alb.
T. steadi               White-capped Alb.
>Thalassarche salvini           - Salvin's Alb.
>Thalassarche eremita           - Chatham Alb.

Four forms once treated as subspecies of T. cauta; cauta is the "shy
albatross' that breeds in Tasmania; steadi is a very similar form that
breeds in NZ subantarctic and was described as a separate subspecies by
Falla (1933). Sandy Bartle maintains that this form cannot be reliably
identified on the museum bench and no convincing information on the
separation of this form has ever been published. It is certainly not a good
BSC species and possibly not a good PSC species or BSC subspecies. However,
it would seem appropriate to recognise NZ and Australian forms as separate
units of biodiversity - a tough issue.

Arguments could be mounted for treating salvini and eremita as full BSC
species on morphological and ecological grounds.

>>I think it's time for a definitive statement on these, something like
Schodde has done with passerines ( even though he gives me some grief with
his lumpings and splittings).

Sorry, but a definitive statement is just a dream. And there are far more
important issues at stake than life lists. Albatross taxonomy and
conservation are intimately entwined, for better or worse.

Juan Mazar Barnett WROTE

>> A cry for help follows. Many of you will be aware of the recent volume
on Albatrosses produced by Robertson and Gales?.As a taxonomic review, it
has primarily failed absolutely  by not including a complete list of taxa
with author and date for each. Rather strangely (and annoyingly I must say)
there is a list of taxa and date (just date) of description, which without
reference to a list of synonyms is pretty much useless.

Good point

>> Also, there is the annoying mention (in full) of the taxon Diomedea
chionoptera in one chapter, and it switches to Diomedea dabbenena in the
other, which nonetheless are different taxa, the former being the "darker
higher-latitude" and the latter the "lighter lower-latitude" (and smaller)

Lower latitudes are closer to the equator. chionoptera is larger, lighter,
and closer to the pole.

>> D/T.(cauta) STEADI is!?, who described it and where.

Covered by Ross Silcock
"Thlassarche cauta steadi Falla 1933 Rec. Auck. Inst. Mus. 1:179-Foveaux

>> Also, does anyone know if the Thalassarche (bulleri/platei) sp nov. with
type loc in Cavancha, Chile has been (re)described yet or if will be at all?

Robertson & Nunn claim that platei Reichneow 1898 is synonomous with
bulleri Rothschild 1893 leaving <Northern Bullers albatross> undescribed
They cited Murphy (1936) and Robertson (pers obs.). Murphy did not say
this, he simply treated the bullers albatross as monotypic. Presumably the
type of platei still exists in the Berlin Museum and has been examined by
Chris Robertson, but he has failed to support his point with any data. If
it does not exist then it would have to be considered indeterminate.

> And does anyone know if the actual data/results/information used to
create this new taxonomy and phylogeny proposed by Robertson and Nunn has
been actually published yet? Any strong views in favour or against it?

Ha! What do you want, science or something. Don't be ridiculous!

To Conclude, I think that to accept a taxonomy before the research has been
done to collect the data to support the findings, is unwise. The albatross
situation is a fine example of terrible science pushed hard by political
interference. While we are waiting for the long promised ground breaking
data the existing trinomial classification allows data on separate
populations to be collected just as well as does the <interim taxonomy>. I
can't help but be highly sceptical of the motives of those who are
promoting Robertson & Nunn's hollow proposal.


Alexander, W.B.R.  et al. 1965. The Families and Genera of petrels and
their names. Ibis 401-405.
Christidis & Boles 1994. The taxonomy and species of Birds of Australia and
Its Territories. RAOU Monograph 2, RAOU, Melbourne.
Murphy, R.C. 1936. Oceanic Birds of South America. AMNH, New York
Nunn, G.B.  et al. 1996. Evolutionary Relationships among extant
Albatrosses? Auk 113: 784-801.
Robertson C.J.R. & Nunn, G.B. 1998. Towards a new taxonomy for Albatrosses.
Chapter 2, pp.13-19 in Robertson, G. & Gales, R. (eds). Albatross Biology &
Conservation. Surrey Beatty & sons, Sydney.
Robertson, CJR.& Warham, J. 1992. Nomemclature of the New Zealand Wandering
Albatrosses Diomedea exulans. Bull Brit Ornith Club. 112: 74-81.
Roux, J.-P. et al 1983. Un nouvel albatross Diomedea amsterdamensis n. sp.
decouvert sur l'Ilse amsterdam (37 50'S, 77 35"E). Oiseau 53: 1-11.
Schodde R. & Mason, I. 1997. Zoological Catalogue of Australia 37.2 Aves
(Columbidae to Coraciidae). CSIRO, Canberra.
Schodde, I. & Mason, I .1999 The Directory of Australian Birds, Passerines.
CSIRO, Canberra.
Stanger, M., Clayton, M., Schodde, R., Wombey, J. & Mason, I. 1998. CSIRO
List of Australian Vertebrates: A Reference with Conservation Status.
CSIRO, Canberra.

David James
PO BOX 5225
Townsville Mail Centre,
Qld 4810, Australia

To unsubscribe from this list, please send a message to

Include ONLY "unsubscribe birding-aus"
in the message body (without the quotes)

<Prev in Thread] Current Thread [Next in Thread>
  • albatross taxonomy and more, David James <=

The University of NSW School of Computer and Engineering takes no responsibility for the contents of this archive. It is purely a compilation of material sent by many people to the birding-aus mailing list. It has not been checked for accuracy nor its content verified in any way. If you wish to get material removed from the archive or have other queries about the archive e-mail Andrew Taylor at this address: andrewt@cse.unsw.EDU.AU