Effect on SACC: This moves the positions of several genera in our current linear sequence.
Background: Our current linear sequence is the product largely of historical momentum and interpretations of comparative morphology. Fortunately, two recent gene-based phylogenies corroborate or are consistent with much of the traditional sequence. This proposal addresses the minor conflicts with these new studies.
New data: Tavares et al. (2006) sequenced 8 genes, mitochondrial and nuclear, for a total of almost 6400 bp. Their taxon sampling included 25 of 30 genera of Neotropical parrots. The well-supported nodes that affect our sequence are
• Deroptyus and Pionites are sisters (100% Bayesian support, 100% ML bootstrap support)
• Myiopsitta and Brotogeris are sisters (100% Bayesian support, 100% ML bootstrap support)
• Nannopsittaca and Bolborhynchus (lineola) are sisters (100% Bayesian support, 100% ML bootstrap support)
Wright et al. (2008) also sampled sequenced 3900+ bp of mitochondrial and nuclear genes, including a different set of genes from Tavares et al., and included 23 genera of Neotropical parrots plus 46 from elsewhere. Therefore, the results of this study represent a largely independent assessment. Their results were impressively concordant with those of Tavares et al., namely:
• Deroptyus and Pionites are sisters (99% ML bootstrap support)
• Myiopsitta and Brotogeris are sisters (100% ML bootstrap support)
They also found that:
• Nannopsittaca and Touit are sisters (100% ML bootstrap support)
However, they did not sample Bolborhynchus, and Tavares et al. did not have Touit, so this is not a conflict. Nannopsittaca and Touit are adjacent in our linear sequence.
And also found that:
• Leptosittaca is sister to Diopsittaca + Guarouba (84% ML bootstrap support), and is not close to Aratinga, with which it is frequently associated
Analysis and Recommendation:
To make our linear sequence correspond to the best available hypotheses on the relationships within the family requires only minor changes, namely:
1. move Deroptyus from its second-to-last position to immediately following Pionites
2. move Myiopsitta downward to follow Brotogeris
3. move Bolborhynchus (plus presumed close relatives Psilopsiagon) downward to precede Nannopsittaca + Touit
4. move Leptosittaca upward to follow Guarouba
The resulting sequence is consistent with the phylogenetic hypotheses above, and other than the placement Deroptyus, the changes are minor. I recommend a YES on this one – it doesn’t get much better than two concordant independent data sets.
TAVARES, E. S., A. J. BAKER, S. L. PEREIRA, AND C. Y. MIYAKI. 2006. Phylogenetic relationships and historical biogeography of Neotropical parrots (Psittaciformes: Psittacidae: Arini) inferred from mitochondrial and nuclear DNA Sequences. Systematic Biology 55: 454470.
WRIGHT, T. F., E. R. SCHIRTZINGER, T. MATSUMOTO, J. R. EBERHARD, G. R. GRAVES, J. J. SANCHEZ, S. CAPELLI, H. MULLER, J. SCHARPEGGE, G. K. CHAMBERS, AND R. C. FLESICHER. 2008. A multilocus molecular phylogeny of the parrots (Psittaciformes): support for a Gondwanan origin during the Cretaceous. Molecular Biology & Evolution 25: 2141-2156.
Van Remsen, November 2008
Comments solicited from Tim Wright: “I support the proposed changes to the linear sequence of Neotropical parrot genera in the SACC. As detailed in the proposal, the changes would bring the linear sequence in line with the recent independent molecular phylogenies by Tavares et al and Wright et al. that show similar topologies.”
Comments from Stiles: “YES. Again, massive multigene data from two independent studies agree, so I can see no reason not to adopt these changes.”
Comments from Nores: “YES, aunque pienso que Forpus debería ser incluido entre Pionopsitta y Touit. Aunque esta relación no tiene fuerte soporte, tampoco la tiene la que agrupa Myiopsitta con Brotogeris, Nannopsittaca y Touit.”
Comments from Zimmer: “YES. Two concordant, independent data sets make the case that this is the best arrangement.”
Comments from Jaramillo: “YES – I see no conflicts and no surprises here really. It is good to have large datasets that are concordant as these are.”
Comments from Pacheco: "YES. Bem suportado por dois trabalhos independentes.”