A Contribution to the Taxonomy of Some British Species of Sappaphis Matsumura 1918 (Homoptera, Aphidoidea].*

The taxonomy of the British species of Sappaphis Matsumura using Crataegus as primary host has been investigated by means of free choice transference tests to secondary hosts and biometric studies on both wild material and experimental cultures derived from transference tests. The results confirm the existence of seven different taxa causing galls on Crataegus leaves in spring. These belong to two species-groups, one consisting only of S. ranunculi (Kltb.) and the other of four species (S. lauberti (Börner), angelicae (Koch), petroselini (Bömer) and crataegi (Kltb.)), one subspecies (S. crataegi ssp. aethusae (Borner)) and one biological race or infrasubspecies of S. crataegi (kunzei Börner). The existence of other subspecies of S. crataegi proposed by Börner (anthrisci and aegopodii) has not been confirmed in extensive field collections, and as no morphological diagnoses have been given these are regarded as subspecies dubiae in view of the not infrequent records of anomalous host choice in this group of Sappaphis. All members of the group have been found to be about equally abundant in the Cambridge area, with the exception of angelicae whose distribution is limited by its secondary host Angelicae sylvestris but relative abundance of the different species fluctuates from year to year. Examination of data for nymph colour in galls on Crataegus and degree of wax powdering is held to discredit the use of these characters for taxonomic purposes in Sappaphis. Examination of hair lengths show that these can vary within limits which do not invalidate them as taxonomic characters except in S. ranunculi (Kltb.); an. estimate of limits of variability for abdominal hair length maxima within the second (crataegi) species-group points to a limiting value of about 24 for the Coefficient of Variability within homogeneous samples. It appears that certain phenomena arising in the analysis of variance data for certain characters in wild and experimental material, together with observations on the characteristics of field collections from Crataegus and secondary hosts, can most simply be accounted for by the existence of a proportion of hybridity among the population on Crataegus in spring. A preliminary experiment shows that cross-mating is possible between Sappaphis crataegi (Kltb.) and its biological race kunzei Bonier. Although transference tests from Crataegus to secondary hosts were used as a valuable complement to morphological studies it has been deduced from theoretical considerations that such tests have no more intrinsic reliability than traditional morphological methods except where secondary host differences bring about effective isolation through the selective disadvantage attaching to hybrids and they are thus no more effective than morphological methods for the detection of incipient divergences in an evolving complex.