Language: English with bilingual summary in English and French
The genus Xiphinema is redefined and a list of valid species and species inquirendae is presented. An overview of the geographical distribution of all species described to date is given, except for those belonging to the X. americanum lineage. Distribution patterns are discussed and illustrated for the most widespread species. The anterior end, the female reproductive system and the adult tail shape of virtually all described species are illustrated, as well as the tail ontogeny of those taxa for which such data were at least partly available. Morphological characters traditionally used in species descriptions are critically evaluated, revealing that many of them are only of limited usefulness in phylogenetic reconstruction due to considerable intraspecific variability and multiple parallel evolution within the genus. Phylogenetically informative characters are identified and their reliability was tested in the light of other information derived from geographical distribution, mode of reproduction and ontogeny. Examination of the developmental stages in over 100 species demonstrated strong unidirectional pattems in tail ontogeny. The major transformation modes during tail development in the genus Xiphinema were identified and partitioned in a number of discrete steps, each characterized by a particular morphological tail type. Simple coding of tail shapes as multi-state characters in each developmental stage proved inadequate and potentially deceptive since tail morphologies in ontogenetic transformation series are biologically dependent. Comparison of the observed tail type in a particular stage with that of the preceding stage demonstrated that identical tail types could result from non-homologous developmental processes. More robust character scoring was achieved by shifting the focus from the pattern to the process and by taking into account the timing of expression of a particular tail type during ontogeny. Trial analysis based exclusively on tail characters and including only those species for which tail data were known in at least three juvenile stages revealed a phylogenetic succession of different tail types and allowed delimination of a number of (mostly paraphyletic) groups. These groups were subjected to further analysis by including all taxa with similar adult tail shape and by employing all phylogenetically informative characters. Composition, morphological details and possible monophyly constraints enforced during the search were discussed. Relationships inferred from the resulting optimal or consensus trees were interpreted in a zoogeographic context.
The long-tailed (or filiform) species were identified as a polyphyletic ancestral stock comprising a basal group of species occupying ancestral positions, and a monophyletic clade standing in apposition to a round-tailed species group. Cladistic analysis also clearly illustrated the polyphyletic status of the “round-tailed species” with (1) certain species having originated from an ancestral stock with long filiform tails, (2) other species being derived (twice) from ancestors with digitate convex-conoid tails, and (3) some species having convergently evolved within the X. americanum lineage.
Species displaying ventrally arcuate to elongate conoid tails form a paraphyletic group showing a gradual transition between both tail types. The evolution of sexual dimorphism within this group is characterized by two major events: (1) abrupt shortening of the long ventrally arcuate tail type in the male, and (2) subsequent gradual shortening of the female tail, eventually leading to secondary loss of sexual dimorphism.
Cladistic analysis suggested that species with conoid to dorsally convex tails have evolved from ventrally arcuate- to elongate conoid-tailed ancestors. In addition, there is a morphological continuum from non-digitate, over subdigitate to digitate tails. One of the monophyletic clades within the paraphyletic group of species with non-digitate conoid tails consists of all representatives currently accommodated in the X. pachydermum- and X. americanum-groups. Parsimony analysis of the subdigitate-tailed species revealed a basal dichotomy leading to a monophyletic clade and a paraphyletic branch demonstrating the transition to the digitate-tailed species. Evolution within the digitate-tailed species is marked by a gradual displacement of the terminal peg from a ventral to a central (on the terminus) position and its subsequent reduction to a mere bulge, eventually leading to a pegless tail. Species wit convex conoid to broadly rounded tails, with or without central peg or bulge, represent the terminal lineage in the phylogenetic tree of the genus. The present phylogenetic analysis confirmed that the subgeneric division of the genus as proposed in the 1970s is purely artificial and bears no relation to reality.