Mauritius January - July 2013, Round Island

Round Island, 25 km North of Mauritius functions as a refugium for critically endangered species, mostly reptiles and plants. Many plants that can be found here have vanished from other places, like the Hurricane Palm, a species which was down to just one last individual, but is now on the return thanks to MWF (individuals can be seen on Iles aux Aigrettes). The reptile community is just as unique for the Keel-scaled Boa, Durrel´s Night Gecko, Guenther´s Day Gecko, Ornate Day Gecko and Telfair´s Skink were restricted to this small island (recent translocations have enlarged their distribution). MWF is still working hard on the restauration of the island and the protection of its inhabitants.

Guenther´s Day Gecko (Phelsuma guentheri)

We were allowed to help with the restauration for 10 days, which was for us a highlight of our stay in Mauritius. This work enabled us to get up close with the breeding seabirds on this island. Round Island for instance hosts the largest colony in the Indian Ocean of Red-tailed Tropicbirds (2.000 - 2.500 pairs) and Wedge-tailed Shearwaters (40.000 - 80.000 pairs). An estimated 750 to 1.500 pairs of White-tailed Tropicbirds also breed on the island. Since 40ties breeding petrels have been sighted. These have always been identified as Trindade Petrels (Pterodroma arminjoniana), but today it is a bit more complicated than that. In in 1987 the first Bulwer´s Petrel was found, in 1994 the first Kermadec Petrel (P. neglecta) and in 1995 the first Herald Petrel (P. heraldica). Currently seven Pterodroma species have been sighted on the island and three of those breed (Black-winged Petrel (P. nigripennis) made several attempts but all of them failed). The three species that breed on Round Island are Trindade, Kermadec and Herald Petrel. These three species are possibly hybridising, which makes the identification of individual birds on Round Island difficult to say the least. In the next paragraphs I will try to give some clues to the identification so that it might increase the insight of this difficult species complex.

Red-tailed Tropicbird (Pheaton rubricauda)

White-tailed Tropicbird (Phaeton lepturus)

Trindade Petrels
The vast majority of the Pterodromas on Round Island are Trindade Petrels (120 to 170 breeding pair). The majority of these birds belong the dark morph. The dark morph on Round Island differs a bit from 
the dark morph on Trindade Island, for the birds in the Indian Ocean almost never show a pale throat. The intermediate and light morph are much less numerous. I even did not see any light morphs at all. The genetics of the Round Island form of Trindade Petrel also differs a bit from the other population in the genetics, but it is hard to give them a (sub)specific status due to the hybridisation that causes the species boundaries to fade...

Intermediate morph Trindade Petrel (Pterodroma arminjoniana)

Dark morph Trindade Petrel (Pterodroma arminjoniana)

Kermadec Petrels
An estimated 10 to 15 pairs of Kermadec Petrel breed on Round Island, but the total number of birds varies from month to month. Also this species is mostly represented here by dark en intermediate morphs. The white-headed morp of this species has even only been seen twice with certainty. Hybridisation with this species and Trindade has been proven. Because of high variety within the species and hybridisation it is hard to identify species just by plumage: head, upperwing and underwing patterns and colouration (although I got the feeling that Kermadec Petrels have a warmer colouration in general). It is therefore, in my opinion, only possible to indentify extreme plumages with certainty, unless the colouration of the primairy shafts can be observed. Kermadec Petrels have white primairy shafts and Herald and Trindade Petrels have dark primairy shafts. The primairy shafts of hybrids are apparently half dark half light. Besides the plumage charactaristics, sound is a very usefull way to identify Kermadec Petrel.

Dark morph Kermadec Petrel (Pterodroma neglecta)

Intermediate morph Kermadec Petrel (Pterodroma neglecta)

Herald Petrels

About the same amount of individual live on Round Island for the Herald Petrel as for the Kermadec Petrel (10 to 15 breeding pairs), in my opinion however, this species is overlooked even more often. The most obvious variant (and according to Carl Jones the only variant occuring here) is the lightest variant. Most of individuals are smaller with a shorter bill (but the there is quite a lot of overlap in the morphologic characteristics of all the three species). Herald petrels also have subtile vermiculations on the chest and light lores. I have also seen birds that have half of the characteristics of a Herald Petrel and half those of a Trindade petrel. Are these individuals hybrids or just deviation within the species?

Light morph Herald Petrel (Pterodroma heraldica)

Intermediate Morph Herald Petrel (Pterodroma heraldica) (picture taken by Anna van der Kaaden)

Phoenix Petrels

Besides these three breeding species and their hybrides there are some rumors about a fourth species that might be breeding here: the Phoenix Petrel. I think the description of this species sounds a lot like an intermediate variant of Trindade Petrel. According to Only and Scofield (Albatrosses, Petrels and Shearwaters of the World) are white undertail coverts and dark underwings good features for a Phoenix Petrel. However, I have seen quite a lot of Trindade Petrels (light underwings) with white undertail coverts. Are these deviating Trindade Petrels? Besides that I have seen a small number of birds that meet the criteria for a Phoenix Petrel, but with a silver haze on the underwings. The photos below show the best candidates for Phoenix Petrel and in Carl Jones´s opinion they actually are.
Update: I´ve send the pictures below to several experts like Paul Scofield, Robert Flood and Daniel Dankwerts. The bird on the upper picture is most likely a Herald Petrel. It shows the pale undertail coverts, bicoloured feet and a pale leading edge, that are good fore Phoenix Petrel, but the pale crescent formed by the pale bases of the primaries does not fit and the dark breast appears to messy and ragged. Trindade Petrel can be excluded by wing shape, for this bird has rather slender wings and does not fit the broad-winged jizz of Trindade Petrel. The bird on the lower picture is most likely a dark or intermediate morph Trindade Petrel, for even though it had white undertail coverts, the white on the underwings is way to extensive for a Phoenix Petrel

Possible Phoenix Petrel that turned out to be most likely a dark morph Herald Petrel (Pterodroma heraldica)

Possible Phoenix Petrel that turned out to be a dark morph Trindade Petrel (Pterodroma arminjoniana)

Luckily birds can also be identified by sound and at least Kermadec Petrel is an easy one to pick out,  because it produces a donky like call. Vocalisations of Trindade and Herals Petrel sound a lot alike. Both produce a chittering noise, but Trindade Petrel ends with a series of low gutteral sounds. Herald Petrel does not produce these low noises. It´s chittering call fades against the end. Sonograms can show this difference really well. Unfortunately I have the feeling that Trindade Petrels tend to leave the low noises out of their call and to stop abruptly after chittering. I think that Phoenix Petrels sounds a lot like a Herald Petrel of a prematurely stoping Trindade Petrel. This all does not make identification much easier. 

The three species (excluding Phoenix Petrel) are closely related and speciation most likely occured not so long ago as the consequence of them breeding in different colonies. This not only makes hybridisation possible but also causes the progeny to be fertile. This causes introgression of at least Kermadec Petrel genes in the Trindade Petrel population on Round Island. (Hybridisation with Herald Petrel has not yet been shown, but it is very plausible. I also am not aware of any interbreeding between Herald and Kermadec Petrels, but this seems less plausible due to the low numbers of breeding birds in both species.) The introgression of genes causes the species boundary to fade more and more with in the end a big hybrid swarm with barely identifiable individual birds. While I was on Round Island, I could see the interspecies interaction myself, for I observed display flights between Trindade Petrels and Herald Petrels and Trindade Petrels and Kermadec Petrels. However I did not see any  Kermadec x Kermadec or Herald x Herald display flights. Currently a lot of DNA research is being conducted on these petrels, but the DNA is not being linked to plumage charactaristics, which makes the research fairly useless for identification issues. So there is still a long way to go...

Kermadec Petrel (Pterodroma neglecta) intermediate morph in display flight with Trindade Petrel (Pterodroma arminjoniana) intermediate morph

Possible Kermadec X Trindade Petrel Hybrid (Pterodroma neglecta X Pterodroma arminjoniana) (picture taken by Anna van der Kaaden)

Access to Round Island itself is not permetted for the "normal birder", but boat trips around the island can be organised and this gives birders a reasonable chance to see Trindade Petrels. Seeing the other species will be a bit more difficult and probably mostly depend on luck. I saw several Trindade Petrels flying between Flat Island and Guenners Coin. Other birds of interest were Brown and Lesser Noddies, Sooty Tern and Subantartic Skua (a vagrant to Mauritius)

For more information about this matter read “Lost Land of the Dodo” by A. Cheke and P.J. Hume, “The breeding biology of the Round Island Petrel and factors determining breeding succes” by R.V. Tatayah, “Occurence of two surface breeding species of Pterodroma on Round Island, Indian Ocean” by M. Brook, M.J. Imber and G. Rowe and “Range expansion and hybridization in Round Island Petrels (Pterodroma ssp.): evidence from microsatellite genotypes” by R.M. Brown, R.A. Nichols, C.G. Faulkes et al.

Round Island (picture taken by Anna van der Kaaden)