Abstract Detail



Systematics

Duvall, M [1], Burke, Sean [2], Clark, Dylan C. [3].

Evolution of the PACMAD grasses; finding the path to warm-season grasslands.

            Grasses underwent a major evolutionary bifurcation 55 or more million years ago. One lineage radiated into cooler, forested, and/or wetter habitats.  The other diversified into over 5,000 species that live in warmer grasslands. These latter are the “PACMAD” grasses (the Panicoideae, Aristidoideae, Chloridoideae, Micrairoideae, Arundinoideae, and Danthonioideae). Exactly how the PACMAD grasses evolved to dominate warm open habitats is a question that can be tested by phylogenomic study. Evidence-based explanations are twofold; the “aristidoid sister” and the “panicoid sister” hypotheses, in which the named subfamily in each case is sister to remaining PACMAD taxa.
            Complete plastid genomes (plastomes) from 169 grasses drawn from all 12 subfamilies and including five new aristidoid plastomes were analyzed by maximum likelihood (ML) and Bayesian Inference (BI) methods. Analyses were repeated 10 times with the successive addition of more gapped positions introduced by sequence alignment until all such positions were included.
            When all gapped positions were removed, the panicoid sister hypothesis was retrieved with moderate support. As gapped positions were gradually added back into the input matrix, support values for the panicoid sister hypothesis initially fell until there was a transition to a weakly supported aristidoid sister hypothesis or a transitory and weakly supported ((P,A),CMAD) topology. Support values for these two alternative hypotheses then rose reaching a moderately high plateau until all gapped positions were included. Similar patterns were seen with both ML and BI methods.
            Extensive study of data partitions in a 250 plastome matrix compared the effects of coding, positively selected, and gapped sites. The deep phylogenomic topology of PACMAD grasses is relatively insensitive to all but the gapped positions introduced by alignment. The fact that the aristidoid sister hypothesis was retrieved in plastome phylogenomic analyses only when gapped positions found in 10% or more taxa were included suggests that this result may be artifactual. Gapped positions tend to cluster in noncoding regions where no alignment can be completely unambiguous.
            The transition of ancient grasses from tropical forests into open grasslands is inexplicable when the ancestor is already adapted to those grasslands as the aristidoid sister hypothesis suggests. Note that Aristidoideae are uniformly open habitat species of warm grasslands. By contrast, sublineages of basal Panicoideae show variation in shade tolerance suggestive of gradual multiple transitions to unshaded habitats eventually leading to a diversity of derived open habitat lineages of PACMAD taxa. 


1 - Northern Illinois University, Plant Molecular And Bioinformatics Center And Department Of Biology, 1425 W Lincoln Hwy, Dekalb, IL, 60115, United States
2 - Northern Illinois University, Biological Sciences, 1425 W. Lincoln Hwy, DeKalb, IL, 60115, United States
3 - Northern Illinois University, Department of Biological Sciences, 1425 W Lincoln Hwy, DeKalb, IL, 60115, United States

Keywords:
PACMAD clade
plastome
Phylogenomics
Aristidoideae
Panicoideae
DNA sequence alignment gaps.

Presentation Type: Oral Paper
Number:
Abstract ID:287
Candidate for Awards:None


Copyright © 2000-2018, Botanical Society of America. All rights reserved