CRITICAL-EVALUATION OF THE TAXA IN DIDYMO SPHENIA (BACILLARIOPHYCEAE)

Didymosphenia populations originating from localities in Siberia/Russi a (Lake Baikal, River Angara), Pakistan (Himalaya, 3300 m above sea le vel), Ireland (Killarney) and from the Faroe Islands have been investi gated in LM and SEM. Other populations from Norway, Sweden, Finland, S cotland, Russia (Lake Onega, Lake Issyk), Kazakhstan (River Iii), and Svalbard we compared in LM only. We distinguish 5 taxa at the specific rank. 1.) Didymosphenia geminata (Lyngb.) M. Schmidt with three, prov isionally designated morphotypes: geminata sensu stricto, capitata and subcapitata. 2.) Didymosphenia siberica (Grun.) M. Schmidt (sensu str icto). 3.) Didymosphenia curvata (Skv. and Meyer) nov. stat. 4.) Didym osphenia clavaherculis (Ehr.) nov. comb. 5.) Didymosphenia pumila nov. spec. These taxa are differentiated by the structure of the areolae w hich is profoundly different in D. geminata and D. clavaherculis compa red to D. siberica, D. curvata, D. pumila. Nevertheless we do not inte nd to split the genus on this criterion, despite several recent cases of hasty splittings by other authors. Our observations are compared wi th older and more recent literature data, in particular with the numer ous infraspecific taxa introduced by Skvortzow and Meyer (1928) and Sk vortzow (1937). The taxa created by the latter authors cannot be disti nguished from each other. Moreover they do not represent real populati ons. From the critical biological view of modern taxonomy they should be seen as typologically described ''pseudo-taxa'' only. Of course aut hors always select and name single individuals as regular variants of the cell cycle. However, their studies on Lake Baikal did not reveal t he 5 real species that occur in 6 to 7 populations, when one takes the morphotypes (populations) into account. Despite these obvious shortco mings one of these taxa, D. geminata var. baicalensis fo. curvata is c hosen as the basionym for a new species: D. curvata. It is shown that Didymosphenia dentata Dorogostaisky sensu Skvortzow and Meyer should n ot be regarded as a distinct species or race. It is evidently based on an optical artefact. Under a certain focus the high profile of the tr ansapical costae of D. geminata resemble a regular row of spines at th e junction of the valve face and mantle. Two forms of spines occur in subapical positions. Continuous spines occur in more weakly silicified marginal crests in D. geminata and D. clavaherculis. In the other thr ee species however, completely isolated, more solid and longer spines occur on the edge of the mantle, without any traces of crests. Apparen tly they were overlooked (corroded?) by Dawson (1973b).