TACHINIDAE OF ORIENTAL REGION 107 



larly it is very difficult to classify the genera included in the Goniinae into easily 

 distinguishable tribes or subtribes - yet some hierarchical classification between 

 the subfamiliar and the generic levels is essential for practical identification as 

 well as for the more erudite purpose of reflecting the supposed phyletic interrelation- 

 ships. 



Some of the tribes or subtribes that have been recognized in recent years, such 

 as the Acemyini and Siphonini (which some specialists consider would be better 

 placed in the Tachininae), are fairly discrete entities and appear to be phyletically 

 natural taxa in which the members have a generally similar external adult facies, 

 a similar reproductive habit, and attack similar host groups. Other sections of 

 the subfamily that have been variously recognized as tribes or subtribes (such as 

 the Sturmiini, Carceliini, Goniini, Eryciini, Trypherina, Masicerina, Erythrocerina, 

 etc., of various specialists) are little more than haphazard aggregates of genera 

 vaguely united by the common possession of a few attributes in the adult flies 

 of (probably) little or no phyletic significance. It is, however, one thing to recognize 

 the artificiality of the existing system of so-called tribes or subtribes and quite 

 another to find a workable system to replace it - the existing classification within 

 the Goniinae, unsatisfactory though it is even at the utilitarian level, is at least 

 an approximation to a practical way of assembling the multifarious genera into 

 manageable units and of differentiating such units (be they called either tribes or 

 subtribes for mere convenience). 



Herting (i960) has put forward, in an inchoate way, an arrangement of most of 

 the west Palaearctic genera of carceliine-sturmiine-goniine-eryciine tachinids in 

 which the genera (with a few exceptions) are aggregated into two major groups 

 depending upon whether the reproductive habit is that of ovolarviparity or microovi- 

 parity. Such a system, properly formalized, implies two suprageneric taxa with 

 redefined limits to which - if ranked at tribal level - the names Eryciini and Goniini 

 correctly apply under the rules of nomenclature. There seems little doubt that 

 Herting's (op. cit. and personal communication) scheme is much closer to the 

 'phylogenetic truth' than is the usual classification into several tribes or subtribes 

 (Sturmiini, Carceliini, Goniini, Eryciini) made solely on the basis of resemblance 

 in external adult characters, and that delimitation of suprageneric groupings on 

 the basis of reproductive biology would result in much more natural taxa from the 

 evolutionary viewpoint. 



The trouble with tribal categorization on the basis of reproductive habit is that 

 the resulting taxa are almost impossible to define and key out, since there is little or 

 no correlation between the range of variation shown by the external adult facies 

 and a particular reproductive method. In practical taxonomy, where identification 

 is vital (often of forms for which the reproductive habit is in any case unstudied), 

 it is therefore impossible to adopt the scheme of recognizing Eryciini and Goniini 

 on a redefined basis; instead it is necessary, at least as an interim measure, to retain 

 the old entities Sturmiini, Carceliini, Eryciini and Goniini, since in spite of the fact 

 that these groups are rather obviously polyphyletic and not always easy to dis- 

 tinguish they nevertheless retain practical value (especially when dealing with 

 little known faunas like that of the Oriental Region) . 



