Island birds (1 Viewer)

[Peter Kovalik]

Valente, L., Phillimore, A.B., Melo, M. et al. A simple dynamic model explains the diversity of island birds worldwide. Nature (2020). https://doi.org/10.1038/s41586-020-2022-5

Abstract:

Colonization, speciation and extinction are dynamic processes that influence global patterns of species richness. Island biogeography theory predicts that the contribution of these processes to the accumulation of species diversity depends on the area and isolation of the island. Notably, there has been no robust global test of this prediction for islands where speciation cannot be ignored, because neither the appropriate data nor the analytical tools have been available. Here we address both deficiencies to reveal, for island birds, the empirical shape of the general relationships that determine how colonization, extinction and speciation rates co-vary with the area and isolation of islands. We compiled a global molecular phylogenetic dataset of birds on islands, based on the terrestrial avifaunas of 41 oceanic archipelagos worldwide (including 596 avian taxa), and applied a new analysis method to estimate the sensitivity of island-specific rates of colonization, speciation and extinction to island features (area and isolation). Our model predicts—with high explanatory power—several global relationships. We found a decline in colonization with isolation, a decline in extinction with area and an increase in speciation with area and isolation. Combining the theoretical foundations of island biogeography with the temporal information contained in molecular phylogenies proves a powerful approach to reveal the fundamental relationships that govern variation in biodiversity across the planet.

 

[TomDerutter]

I was trying to find data for Humblotia flavirostris - which I found was sampled for this study
Apparently it's embedded within Muscicapa, sister to Muscicapa adusta

 

[Jim LeNomenclatoriste]

Apparently it's embedded within Muscicapa, sister to Muscicapa adusta

Where did you see that? There are no trees in the paper

 

[jts1882]

Try the Supplementary materials. There is a file Humblotia_maxcred.tre under Maximum clade credibility trees. It's a Nexus file, which I can't review.

Phylogenetic trees used for global analysis of island birds

Maximum clade credibility trees and posterior distribution from the BEAST analyses of island birds. 80 different phylogenetic analyses. For the 11 previously published trees, please contact the authors.

data.mendeley.com

Found a way of previewing. It is indeed sister to M. adjusta and together they are sister a clade containing M. aquatica, M. cassini, M. gambagae, and M. striata, and they are all sister to M. sethsmithi.

 

[TomDerutter]

You can open the trees on IcyTree for example

 

[l_raty]

Try the Supplementary materials. There is a file Humblotia_maxcred.tre under Maximum clade credibility trees. It's a Nexus file, which I can't review.

Here is this tree, viewed in FigTree. Branch label = posterior probability.

 

[Jim LeNomenclatoriste]

In the extended Date table 3, it's written "Chasiempis / Elepaio". Don't tell me that Elepaio is an undescribed new genus ? 😮

 

[Mysticete]

In the extended Date table 3, it's written "Chasiempis / Elepaio". Don't tell me that Elepaio is an undescribed new genus ? 😮

Presumably a typo incorporating the common name by mistake

 

[andrew147]

Valente, L., Phillimore, A.B., Melo, M. et al. A simple dynamic model explains the diversity of island birds worldwide. Nature (2020). https://doi.org/10.1038/s41586-020-2022-5

I’ve been going through the cladograms in the supplementary information of this paper – and there are definitely some interesting points for discussion. I’m not qualified to critique their methodology but some novel findings in other recent papers seem to be supported here, amongst (e.g.) Estrildidae, Ploceidae, Monarchidae & Dicruridae. As the diagrams are well-hidden, and there are a lot of them, I thought I’d share some of the highlights. So, in no particular order...

Muscicapidae: the position of 'Humblotia' flavirostris has already been discussed. It is embedded in Muscicapa, sibling to M. adusta. The overall tree is very close to other studies (e.g. Voelker et al. 2016) adding credence to the result.

Nectariniidae: Anabathmis is polyphyletic with 'A', hartlaubi sibling to Dreptes thomensis in a different area of the family. Other genera are polyphyletic, as expected, but not enough spp are included to be able to do much about it.

Cisticolidae: the Prinia tree was the most interesting to me, as it is an heterogeneous group and few spp have been included in previous studies. Here, ‘Prinia’ buchanani is outside Prinia, closest to Orthotomus sutorius (the only Orthotomus sp included); the two do look very similar in photographs. (In my personal list, I’ve separated it in a monospecific Franklinia. I’ve also divided the remaining members of the traditional Prinia into 4 genera - Herpystera, Drymoica, Suya and Prinia - as this allows Urorhipis & Heliolais to be retained).

Fringillidae: amongst the Hawaiian finches, the Hemignathus/Loxops group is not monophyletic, with Akiapolaau and Anianiau both closer to Pseudonestor and these three in turn closer to Loxioides/Telespiza.

In the crossbill (Loxia) tree, there are three groups of red crossbills: a north Eurasian group, including scotica & pytyopsittacus, which is sibling to a North American group + an Asian group. Amongst the white-winged spp, L. megaplaga is closest to bifasciata, with leucoptera sibling to the pair.

Sittidae: the nuthatches are organised differently to other studies (e.g. Päckert et al, 2020; Pasquet et al, 2014) showing a primary division into a micro and a macro clade.

Sturnidae: 'Lamprotornis' shelleyi/hildebrandti have previously been recovered as distant to the rest of genus (e.g. Lovette et al, 2007), but here they completely outside Lamprotornis, sibling to Hylopsar + Notopholia.

Bombycilloidea: relationships amongst are represented as: Phainopepla (Phainoptila + Ptiliogonys) + Dulus (Hylocitrea (Bombycilla)). In the same tree, Mohoidae is shown as sibling to Irenidae.

 

[awiner]

Bombycilloidea: relationships amongst are represented as: Phainopepla (Phainoptila + Ptiliogonys) + Dulus (Hylocitrea (Bombycilla)). In the same tree, Mohoidae is shown as sibling to Irenidae.

Is that result placing Irenidae in Bombycilloidea or placing Mohoidae outside of it? And is this a novel result?

 

[andrew147]

Is that result placing Irenidae in Bombycilloidea or placing Mohoidae outside of it? And is this a novel result?

Mohoidae + Irenidae is completely outside of Bombycilloidea. I don't recall seeing this alignment before.

 

[Jim LeNomenclatoriste]

I’ve been going through the cladograms in the supplementary information of this paper – and there are definitely some interesting points for discussion. I’m not qualified to critique their methodology but some novel findings in other recent papers seem to be supported here, amongst (e.g.) Estrildidae, Ploceidae, Monarchidae & Dicruridae. As the diagrams are well-hidden, and there are a lot of them, I thought I’d share some of the highlights. So, in no particular order...

Muscicapidae: the position of 'Humblotia' flavirostris has already been discussed. It is embedded in Muscicapa, sibling to M. adusta. The overall tree is very close to other studies (e.g. Voelker et al. 2016) adding credence to the result.

Nectariniidae: Anabathmis is polyphyletic with 'A', hartlaubi sibling to Dreptes thomensis in a different area of the family. Other genera are polyphyletic, as expected, but not enough spp are included to be able to do much about it.

Cisticolidae: the Prinia tree was the most interesting to me, as it is an heterogeneous group and few spp have been included in previous studies. Here, ‘Prinia’ buchanani is outside Prinia, closest to Orthotomus sutorius (the only Orthotomus sp included); the two do look very similar in photographs. (In my personal list, I’ve separated it in a monospecific Franklinia. I’ve also divided the remaining members of the traditional Prinia into 4 genera - Herpystera, Drymoica, Suya and Prinia - as this allows Urorhipis & Heliolais to be retained).

Fringillidae: amongst the Hawaiian finches, the Hemignathus/Loxops group is not monophyletic, with Akiapolaau and Anianiau both closer to Pseudonestor and these three in turn closer to Loxioides/Telespiza.

In the crossbill (Loxia) tree, there are three groups of red crossbills: a north Eurasian group, including scotica & pytyopsittacus, which is sibling to a North American group + an Asian group. Amongst the white-winged spp, L. megaplaga is closest to bifasciata, with leucoptera sibling to the pair.

Sittidae: the nuthatches are organised differently to other studies (e.g. Päckert et al, 2020; Pasquet et al, 2014) showing a primary division into a micro and a macro clade.

Sturnidae: 'Lamprotornis' shelleyi/hildebrandti have previously been recovered as distant to the rest of genus (e.g. Lovette et al, 2007), but here they completely outside Lamprotornis, sibling to Hylopsar + Notopholia.

Bombycilloidea: relationships amongst are represented as: Phainopepla (Phainoptila + Ptiliogonys) + Dulus (Hylocitrea (Bombycilla)). In the same tree, Mohoidae is shown as sibling to Irenidae.

Results to take with a grain of salt ?

 

[andrew147]

Results to take with a grain of salt ?

Most of the results seem well aligned with other studies. The Psittaciformes tree was bizarre but otherwise the novelties seemed feasible (unlike e.g. recent hummingbird/Charadriiformes studies).

Do you think the methodology is flawed?

 

[Jim LeNomenclatoriste]

Do you think the methodology is flawed?

I don't know, but I would have liked these figures to be included in the article rather than in separate files, especially files in .tre

 

[andrew147]

I don't know, but I would have liked these figures to be included in the article rather than in separate files, especially files in .tre

Agreed, but the subject of this paper was island biogeography, rather than phylogenetics - so I understand why the cladograms were appended.

 

[l_raty]

Most of the sequences they used were taken from GenBank: the results are well aligned with other studies largely because the data were the same.

The Mohoidae in this particular tree are represented by short (< 325 bp) cytochrome b sequences only (taken from the Fleischer et al 2008 dataset; the longest sequences in the original study were of RAG-1, to which pieces of cytb, 12s, ATP6 and 8, and bFib5 and 7 were added). Additionally, some of the sequences might not be perfect, as they are oldish and were (unavoidably) obtained from old museum specimens.

 

[Jim LeNomenclatoriste]

Additionally, some of the sequences might not be perfect, as they are oldish and were (unavoidably) obtained from old museum specimens.

Example?

 

[l_raty]

Example?

Example of what ?
Mohoidae are extinct, thus no fresh material is available for any of them. The methods for sequencing old DNA have improved quite a lot in recent years, but in somewhat older sequences, at least, it's not rare to have sequencing errors. I'm saying nothing more than this.

 

[Jim LeNomenclatoriste]

Example of what ?

Example of imperfect sequences, but I thought you were talking about all the families and not only the Mohoidae x)

 

[Jim LeNomenclatoriste]

I’ve been going through the cladograms in the supplementary information of this paper – and there are definitely some interesting points for discussion. I’m not qualified to critique their methodology but some novel findings in other recent papers seem to be supported here, amongst (e.g.) Estrildidae, Ploceidae, Monarchidae & Dicruridae. As the diagrams are well-hidden, and there are a lot of them, I thought I’d share some of the highlights. So, in no particular order...

Muscicapidae: the position of 'Humblotia' flavirostris has already been discussed. It is embedded in Muscicapa, sibling to M. adusta. The overall tree is very close to other studies (e.g. Voelker et al. 2016) adding credence to the result.

Nectariniidae: Anabathmis is polyphyletic with 'A', hartlaubi sibling to Dreptes thomensis in a different area of the family. Other genera are polyphyletic, as expected, but not enough spp are included to be able to do much about it.

Cisticolidae: the Prinia tree was the most interesting to me, as it is an heterogeneous group and few spp have been included in previous studies. Here, ‘Prinia’ buchanani is outside Prinia, closest to Orthotomus sutorius (the only Orthotomus sp included); the two do look very similar in photographs. (In my personal list, I’ve separated it in a monospecific Franklinia. I’ve also divided the remaining members of the traditional Prinia into 4 genera - Herpystera, Drymoica, Suya and Prinia - as this allows Urorhipis & Heliolais to be retained).

Fringillidae: amongst the Hawaiian finches, the Hemignathus/Loxops group is not monophyletic, with Akiapolaau and Anianiau both closer to Pseudonestor and these three in turn closer to Loxioides/Telespiza.

In the crossbill (Loxia) tree, there are three groups of red crossbills: a north Eurasian group, including scotica & pytyopsittacus, which is sibling to a North American group + an Asian group. Amongst the white-winged spp, L. megaplaga is closest to bifasciata, with leucoptera sibling to the pair.

Sittidae: the nuthatches are organised differently to other studies (e.g. Päckert et al, 2020; Pasquet et al, 2014) showing a primary division into a micro and a macro clade.

Sturnidae: 'Lamprotornis' shelleyi/hildebrandti have previously been recovered as distant to the rest of genus (e.g. Lovette et al, 2007), but here they completely outside Lamprotornis, sibling to Hylopsar + Notopholia.

Bombycilloidea: relationships amongst are represented as: Phainopepla (Phainoptila + Ptiliogonys) + Dulus (Hylocitrea (Bombycilla)). In the same tree, Mohoidae is shown as sibling to Irenidae.

Where are you in your bird list?