THE CLASSIFICATION OF THE EPHEMEROPTERA

E. Riek

Abstract:Although there has been general agreement at the family or subfamily level, there has been little agreement on the higher classification of the Ephemeroptera. EDMUNDS (1962) ably reviewed the earlier schemes (NEEDHAM, TRAVER and Hsu 1935; BURKS 1954; EDMUNDS and TRAVER 1954; DEMOULIN 1958), and presented a concise outline of a new classification. EDMUNDS, ALLEN and PETERS (1963) made slight changes to this classification and dealt with it in more detail although the interrelationships between the superfamilies are not adequately discussed. ILLIES (1968) accepted this classification. Earlier classifications have tended to emphasise the structure, especially wing venation and tarsal segmentation, of the adults, whereas I consider that the nymphs offer a sounder basis for an outline of the higher classification of the order, although supporting evidence of the distinctiveness of some nymphal types is supplied by the adults. I agree with EDMUNDS and ALLEN (1966) that a knowledge of the immature stages is essential to the task of reconstructing the probable phylogeny of the order. LANDA (1959) has shown that the internal anatomy of the nymphs is also significant in this respect. The proposed classification (Fig. 1) differs from that of EDMUNDS, ALLEN and PETERS (1963) mainly in separation of the Heptageniidae from the Baetoidea and elevation to superfamily status equivalent to the Leptophlebioidea. However, there are apparently greater differences between the two schemes when one considers the presumed interrelationships between the superfamilies. Although the Prosopistomatidae and Baetiscidae are referred to the same superfamily at present, the differences, both in nymphs and adults are such that a more-detailed study may indicate that they are not as closely related as at present assumed. The Prosopistomatidae, in which the venation is very reduced, are the only Recent mayflies in which there is no evidence of a triadic branching of MA. The Baetiscidae, on the other hand, have the most generalized venation of all Recent species; the venation differs from that of the Lower Permian Misthodotidae mainly in the basal alignment of the veins. There is also a slight reduction in the anal field. The position of the Caenoidea is also somewhat doubtful but, as pointed out by EDMUNDS, ALLEN and PETERS (1963), the evidence seems to indicate that the superfamily is most closely related to the Baetiscidae. There is a distinct similarity between the nymphs, if one makes allowance for the extreme development of the carapace and the correlated changes that have taken place in the Baetiscidae. However, there, is a difference in the caudal filaments. They

Biology