CRYPTIC, GENETICALLY EXTREMELY DIVERGENT, POLYTYPIC, CONVERGENT, AND POLYMORPHIC TAXA IN MADAGASCAN TROPIDOPHORA (GASTROPODA, POMATIASIDAE)

Madagascar's magnificent and environmentally threatened endemic radiat ion of the land-snail genus Tropidophora Troschel has recently been cl assified into three subgenera, 95 species and 142 varieties, based on often subtle conchological variation among small samples; it seems bes t to ignore temporarily this confusing plethora of names until true bi ological species and their relationships are better understood. The au thor's field work in 1990 succeeded in obtaining live Tropidophora fro m 40 populations distributed throughout the island. Allozyme analysis (108 snails, 15 loci, 117 alleles) yielded a cladogram suggesting nine genetically coherent taxa (T. taxa A-I), each represented in the coll ections by 1-10 populations. Comparisons among shells (total 1634) and penes (total 31 from 20 populations representing eight of the nine ta xa) revealed: (1) two conchologically indistinguishable taxa (H and I) fixed for alternative alleles at 13 of 15 loci, with only a very subt le difference in penes, and with mosaic and overlapping geographical d istributions in the Northeast; (2) two extremely polytypic taxa (C in the Southwest, F in the Southeast) with parallel trends toward depress ed, broadly umbilicate, heavily sculptured shells with apertural lips widely reflected at the umbilicus at inland, more arid sites, resultin g in sympatric convergence; (5) one southeastern taxon (G) in which th e penis apparently doubles in length but the shell does not change 6.0 km to the northwest, but in which the shell shifts dramatically in sc ulpture, colour, and lip reflection 0.5 km to the westnorthwest; and ( 4) generally such extreme intra- and interpopulation variation in shel l and male-genital characters as to render many of them dubious at bes t for systematics. Thus the Madagascan Tropidophora present a fascinat ingly complex problem in evolutionary morphology/ecology, the solution of which will require even more extensive collecting, followed by mol ecular comparisons or detailed anatomical comparisons, or both. The to tal number of biological species is still likely to be quite large, de spite irrelevance of much of the current taxonomy, because many smalle r species remain to be discovered. (C) 1995 The Linnean Society of Lon don