What is Rhabditidae?

• Rhabditidae is a family of "free-living" nematodes in the order Rhabditida.  Worms in this family are usually specialized to feed on bacteria (decaying organic matter) and are often associated with animal hosts or vectors.  The life-cycle can be as short as 3.5 days, but these worms can also develop through an alternative larval stage (the dauerlarva, probably preadaptive to the infective stage of parasites) which is specialized to disperse or to resist harsh conditions.  Many are easy to culture on bacteriological plates.  We maintain a collection of strains and a database for these and other strains (WSRN, see Resources).
   
• Several scientifically, agriculturally, and medically important nematodes are derived from, are closely related to, or are themselves Rhabditidae.  For example, vertebrate parasites of order Strongylida have phylogenetic origins in Rhabditidae (Fitch & Thomas 1997; Blaxter et al. 1998; below).  C. elegans is a member of Rhabditidae and a major model for fundamental biological research, drug design and functional genomics (see C. elegans WWW server).
   
• The taxonomy of Rhabditidae is in flux.  Classically, different systematists have had very different ideas about the composition of species groups in this taxon or about how these groups are related to each other.  Two major influences on rhabditid systematics are Drs. Istvan Andrássy (1983, 1984) and Walter Sudhaus (1976 and later).  Sudhaus and Fitch (1999) have provided an English translation of part of the 1976 monograph by Sudhaus.

  [Return to top of page.]

Data we use to infer Rhabditidae phylogeny

• We have been using primarily DNA sequences from 18S ribosomal RNA genes (rDNA) for inferring species phylogenies (see also P. De Ley's list of nematode 18S sequences).
   
• Although rDNA is a family of repeated genes, the 18S sequences in the different repeated units are very homogeneous.  That is, all the 18S sequences share evolutionary changes, a phenomenon known as concerted evolution (see figure on concerted evolution).
• Using dye-primer sequencing chemistry, we have demonstrated for all of our 18S sequences that intraspecific polymorphisms, though detectable, are very rare.  We have detected rare substitution polymorphisms (see figure) and even rarer insertion/deletion polymorphisms (see figure).
• Our high-throughput Genetics Analysis Facility (GAF), funded by NSF and NYU, allows highly accurate and extensive sequencing reads (see Facilities).
• One of our alignments of these 18S sequences uses predicted secondary structures of the RNA products (based on models provided by R. De Wachter, Univ. Antwerp):
    [PDF files with secondary structures of 18S rRNA from representative taxa]
• A portion of this alignment is available for downloading (see also the Resources page):
    [Alignment with secondary structures in MS Word98 format]
   
• We have also begun to use DNA sequences from other genes; e.g., the genes for RNA polymerase II and let-60 Ras.  We are also exploring other genes for their phylogenetic informativeness.  These data will be incorporated into our phylogenetic analyses in the next few years.
   
• We also use morphological characters for phylogenetic inference, but using new data on character homologies and states obtained from electron microscopic and developmental studies.
   
• In analyses at the single cell level, we have been able to trace the developmental origins of all cells that make up the ray and phasmid sensilla of the male tails of several representatives of Rhabditidae (see Evolution and Morphogenesis pages).  In all Rhabditidae species we have characterized so far, the patterns of these unique origns for each sensillum are identical, providing a means to identify which sensilla are homologous between species (Fitch & Emmons 1995; Fitch 1997).
• We have used these homologies (and other features identified at the cellular level) to compile a matrix of male tail morphological (and cellular) characters.  This matrix has been published (Fitch 1997) and is available from the journal site or our Resources page.
• In collaboration with Dr. Sudhaus, these and additional morphological characters will be incorporated into our phylogenetic analyses in the next few years.

  [Return to top of page.]

Phylogenetic results so far

• Our preliminary analyses with 18S rDNA alone (>50 Rhabditidae species) have resulted in 3 phylogenies with different degrees of resolution and confidence (evaluated by maximum likelihood using appropriately tested models):
   
• The first phylogenetic tree (cladogram) only depicts those relationships which are the most strongly supported, and for which we have nearly 100% confidence.
• The second tree (also a cladogram) depicts additionally resolved relationships for which we have a somewhat lower confidence.  This is the tree we have used for tracing evolutionary changes (see pages on Rhabditidae evolution).
• The third tree is the maximum-likelihood tree.  Several of the relationships in this completely dichotomously branching tree (marked by arrows in Tree 3) are not well-supported.  But the advantage of this tree is that it shows branch lengths.
• Because these are still preliminary results, please do not use them without permission.
   
• Our previous analyses with 18S rDNA and male tail morphological characters (10 Rhabditidae species) showed that these two kinds of data yield phylogenetic results that are not significantly incongruent (Fitch 1997).  When combined, the two data partitions are complementary and together yield a more highly resolved phylogeny.

  [Return to top of page.]

Significance of these results

• Adding particular taxa increases resolution for species groups in Rhabditidae, even without adding data from more genes.  This increased resolution results from the "breaking" of long branches separating particular taxa.
   
• But there are more relationships that need to be resolved when more taxa are added; some of these relationships are not well resolved because there are not enough evolutionary change occurring in 18S rDNA to provide phylogenetic information about these particular relationships.  This means we need data from additional characters (DNA, morphology) to resolve these relationships.
   
• Several previous phylogenetic hypotheses about Rhabditidae are completely overturned by our new data:
   
• In the classical view, species groups in Rhabditidae that do not have a "glottoid apparatus" (Parasitorhabditis, Protorhabditis) are primitive and anciently diverged.  However, our data strongly suggest that these taxa are derived, and that the glottoid apparatus must have been lost at least twice independently (see Evolution pages).
• Rhabditoides was classically thought to be closely related to species in a group that Sudhaus called "Eurhabditis".  However, our data strongly suggest that Rhabditoides contains some of the most anciently diverging species lineages in Rhabditidae and could be polyphyletic.
• Cruznema is not closely related to Mesorhabditis-like species, as classically thought, but is much more closely related to "Eurhabditis" species.
• "Peloderinae", a subfamily erected by Andrássy containing many species groups with "Peloderan" male tails, is polyphyletic; several independent changes between peloderan and leptoderan tails have occurred in Rhabditidae (see pages on Evolution).
• Species classically placed in a different family or subfamily (genus Diploscapter) are derived from within the Protorhabditis genus.
• Two completely different orders are derived from species groups within Rhabditidae:  Strongylida (comprising some vertebrate parasites) is derived from the "Eurhabditis" group (as previously predicted by Fitch & Thomas 1997 and Blaxter et al. 1998) and Diplogasterida is closely related to some of the Rhabditoides species (as predicted by nobody).
• As predicted by Sudhaus (1993), the insect-parasitic family Heterorhabditis is derived from within Rhabditidae ("Eurhabditis").
   
• Because at least two parasitic groups are derived from within Rhabditidae, these phylogenetic analyses are required for reconstructing the origin of specialized adaptations associated with parasitism in these species.  Our Rhabditidae phylogeny also provides a means of predicting taxa most closely related to these parasites for comparative studies that could lead to the development of taxon-specific control agents (e.g., anti-parasite drugs).
   
• Rhabditidae includes a major model system, Caenorhabditis elegans, which provides a wealth of detailed information about biological mechanisms.  Our phylogeny for Rhabditidae provides the comparative context needed for interpreting the relevance of information derived from the C. elegans system for other nematodes and eventually other metazoan systems, such as humans.

  [Return to top of page.]