364 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I25 



and Gigantodax aquamarensis show definite preference for stones and 

 rocks. The other species favor leaves, stems, and twigs as substrata. 



Of all the species listed, only aquamarensis is lOO percent specific 

 insofar as the substratum to which the larvae and pupae attach (stones 

 and rocks). This is probably due to the nature of its breeding places 

 which do not support leafy plants owing to the very low pH. Pupae 

 of Gigantodax wrighti and .S. tricornis, although usually found at- 

 tached to stones, rocks, leaves, stems, and twigs, were also sometimes 

 found loose in the mud at the bottom of the streams. They were col- 

 lected by passing water over the mud which was held in a coarse 

 strainer. 



The three principal anthropophilic species, S. ochraceum, metalli- 

 cum, and callidum, all preferred leaves for attachment of their larvae 

 and pupae, although they were frequently collected from stones and 

 rocks. 



Since rapids and sudden drops in level are so prevalent in almost 

 all the smaller streams of the Pacific slope, it may be that several of 

 the larvae and pupae included under normal streams should have been 

 listed as having been collected in cascades. However, for the purpose 

 of this study, a cascade was defined as a small waterfall having at 

 least a 5-foot drop. 



No species was found exclusively in waterfalls. Of all the species 

 listed in table 30, only 5". carolinae, larvispinosum, microbranchium, 

 and ethelae had been taken frequently in this habitat. At the time he 

 described 6". carolinae, De Leon (1944) considered the heavily sclero- 

 tized, pigmented plates associated with the X-shaped sclerite of the 

 larvae as an adaptation for life in the waterfalls where the larvae 

 would be subjected to the force of the torrents. However, the structure 

 of these plates, as well as observations of the larvae in their natural 

 habitats, do not corroborate his beliefs. Since the fixation of the larvae 

 to the substratum is by means of the anal disc (posterior sucker), and 

 since the sclerotized plates are situated at a marked angle to the disc, 

 the plates would not be able to enter the substratum. Also, since al- 

 most 99 percent of the larvae of this species attach to rocks, the plates 

 would, at best, be of use to only the small number attaching to vege- 

 tation. It would be expected that many more than 41 percent of the 

 larvae and pupae of this species would be found in waterfalls, if this 

 adaptation was evolved to serve the function De Leon assigned to it. 

 Only 25 percent of the larvae and pupae of S. larvispinosum, which 

 has a still more elaborate arrangement of sclerotized plates around the 

 anal disc, were found in waterfalls. 



