Can someone (Silurus?) please tell us a little more about the proposed split of Pimelodidae into Pimelodidae, Pseudopimelodidae and Heptapteridae?
Why is the split considered necessary? If motivated by monophyly, what was not monophyletic about the "old" Pimelodidae?
What are the primary distinguishing characteristics of each? Can any one of these three families be descibed as more plesiomorphic than the others?
Is there some sense in which Pseudopimelodidae is related to the Loricarioids? If so, what?
Dinyar
Splitting Pimelodidae
-
Dinyar
- Posts: 1286
- Joined: 31 Dec 2002, 00:34
- My articles: 3
- My images: 227
- My catfish: 10
- My cats species list: 3 (i:10, k:0)
- Spotted: 94
- Location 1: New York, NY, USA
- Interests: Mochokidae, Claroteidae, Bagridae, Malepteruridae, Chacidae, Heteropneustidae, Clariidae, Sisoridae, Loricariiadae
-
Silurus
- Posts: 12419
- Joined: 31 Dec 2002, 11:35
- I've donated: $12.00!
- My articles: 55
- My images: 893
- My catfish: 1
- My cats species list: 90 (i:1, k:0)
- Spotted: 424
- Location 1: Singapore
- Location 2: Moderator Emeritus
The split arose primarily of Mario de Pinna's Ph.D. dissertation (which has yet to be formally published, BTW) in which he looked at the phylogenetic relationships of all catfishes. And yes, the split arose because of the non-monophyly of the old Pimelodidae.
In his strict consensus of 360 equally most parsimonious trees, <i>Pseudopimelodus</i> was found to be the sister group of the sisoroid (akysids, amblycipitids,aspredinids, erethistids, and sisorids) + loricarioid (amphiliids, astroblepids, callichthyids, loricariids, nematogenyids, scoloplacids and trichomycterids) clade. Heptapterids were found to be the sister group of the <i>Olyra</i>+<i>Phreatobius</i> clade. What was left (Pimelodidae sensu stricto) was found to be the sister group of the bagrids. All this in different parts of the tree.
The three groups are not characterized by any single character, but rather by a combination of characters that are all osteological (and too technical to list here).
The problem with a study of that scope is always the effects of inadequate sampling on the tree topology (e.g. he sampled bagrids very poorly and got a completely different topology than what Mo obtained; I would favor Mo's cladogram as it is supported by the weight of evidence). The results are summarized in de Pinna (1998) and I can send a scan of that if anyone is interested. Rui Diogo's upcoming book claims to have a "better" catfish phylogeny than de Pinna's. We shall see.
You get an apparently different picture with the molecules. Mike Hardman's work indicated that the three groups are indeed one monophyletic unit (the Pimelodidae of the old). Confusing, eh?
In his strict consensus of 360 equally most parsimonious trees, <i>Pseudopimelodus</i> was found to be the sister group of the sisoroid (akysids, amblycipitids,aspredinids, erethistids, and sisorids) + loricarioid (amphiliids, astroblepids, callichthyids, loricariids, nematogenyids, scoloplacids and trichomycterids) clade. Heptapterids were found to be the sister group of the <i>Olyra</i>+<i>Phreatobius</i> clade. What was left (Pimelodidae sensu stricto) was found to be the sister group of the bagrids. All this in different parts of the tree.
The three groups are not characterized by any single character, but rather by a combination of characters that are all osteological (and too technical to list here).
The problem with a study of that scope is always the effects of inadequate sampling on the tree topology (e.g. he sampled bagrids very poorly and got a completely different topology than what Mo obtained; I would favor Mo's cladogram as it is supported by the weight of evidence). The results are summarized in de Pinna (1998) and I can send a scan of that if anyone is interested. Rui Diogo's upcoming book claims to have a "better" catfish phylogeny than de Pinna's. We shall see.
You get an apparently different picture with the molecules. Mike Hardman's work indicated that the three groups are indeed one monophyletic unit (the Pimelodidae of the old). Confusing, eh?
