Species Tree Estimation of Heroine Cichlids

[cichla]

Species Tree Estimation and the Historical Biogeography of Heroine Cichlids

Hulsey, Keck & Hollingsworth

Molecular Phylogenetics and Evolution in press

Abstract
Heroine cichlids are major components of the fish faunas in both Central America and the Caribbean. To examine the evolutionary patterns of how cichlids colonized both of these regions, we reconstructed the phylogenetic relationships among 23 cichlid lineages. We used three phylogenetically novel nuclear markers (Dystropin b, Myomesin1, and Wnt7b) in combination with sequence data from 7 other gene regions (Nd2, Rag1, Enc1, Sreb2, Ptr, Plagl2, and Zic1) to elucidate the species tree of these cichlids. The species examined represent major heroine lineages in South America, Central America, and the Greater Antilles. The individual gene trees of these groups were topologically quite discordant. Therefore, we combined the genetic partitions and inferred the species tree using both concatenation and a coalescent-based Bayesian method. The two resulting phylogenetic topologies were largely concordant but differed in two fundamental ways. First, more nodes in the concatenated tree were supported with substantial or 100% Bayesian posterior support than in the coalescent-based tree. Second, there was a minor, but biogeographically critical, topological difference between the concatenated and coalescent-based trees. Nevertheless, both analyses recovered topologies consistent with the Greater Antillean heroines being phylogenetically nested within the largely Central American heroine radiation. This study suggests that reconstructions of cichlid phylogeny and historical biogeography should account for the vagaries of individual gene histories.

[Bojan Dolenc]

See also from the same authors:
Temporal diversification of Central American cichlids
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2944184/

[Bojan Dolenc]

Graphical abstract

Research highlights

► Ten gene species tree of Central American and Caribbean cichlids. ► Individual gene trees topologically very discordant. ► The concatenated and coalescent-based tree differed in support and topologically. ► Greater Antillean cichlids phylogenetically nested within Central American heroines.

Keywords: Anomaly zone; Cichlidae; Cuba; Haiti; Middle America; Coalescent
http://www.sciencedirect.com/science?_o ... archtype=a

[Bas Pels]

I can be very wrong indeed - but if I'm not, this might help

The current greater Antilles islands - such as Cuba and Hispanolia are said to have formed a landbridge between northa and south america before the current Central America showed up

It might be safe to assume the Nandopsis species on these islands evolved from a predecessor living on this landbridge - and perhaps this predecessor also had offspring on parts of the landbridge which later moved to the west, and became part of central America

Put differently - not the fishes, but the islands moved. This might result in less unlikely patterns. But as said - I'm not certain I remember this ancient geography correctly

[Darrell Ullisch]

Actually, DNA work was done (15-20?) years ago by Glen Collier and others, using reptiles, amphibians, and cyprinodont fishes, to show that the most of the fauna on Cuba and Hispaniola are relict lines that were stranded when the old land bridge broke up. I sent a copy of the paper to Willem a few years ago after seeing his talk at OCA, PM your email to me and I will send it to you. The larger islands were a part of of a land bridge, yes, but I'm pretty sure they moved east, and a new land mass moved in from the west to form the current CA. This one was colonized again from the south (so far as killifish and reptiles are concerned), I forget the exact timeline, but it was millions of years in the process. There was never any contact between the current isthmus and the Greater Antilles except where it came into contact with the two continental land masses millions of years later.

That paper also determined that Trinidad and most of the lesser Antilles were colonized by SA lines during the last Ice Age, when ocean levels were very low. At that time they made up a peninsula, which was easily invaded from the mainland. There are several species on Trinidad that also exist on the mainland of SA, so the connection was fairly recent by geological time.

As usual, the cichlid researchers are a decade or more behind the killifish science. (Sorry, couldn't help myself.)