BY n. J. TILLYARD. 209 



ation and habits than are some other aquatic insect larvae, such as those of May- 

 flies. It seems clear that their numbers have been much decreased since the trout 

 were freed in these lakes and rivei-s, although I am unable to estimate this reduc- 

 tion as clearly as in the cases of Stone-fiies and May-flies. 



TJie Caddis-flies. 



Observations in other parts of the world, as well as in other parts of New 

 Zealand, show that Caddis-fly larvae foi-m one of the most important articles of 

 diet for the trout. Most of these larvae construct cases for themselves out of 

 weeds, sticks, sand or small pebbles; and one would imagine that such habits as 

 these would serve as eflicient protection for them. But this is not the case. The 

 trout know well the habits of the Caddis larvae. They watch carefully for any 

 suspicious movement amongst the weeds, sticks, etc., that strew the bottoms of the 

 lakes and streams, and they pounce upon the larvae and swallow them whole in 

 their cases. The substance of the case is usually indigestible; but the larva itself 

 is a succulent, fat morsel, and an excellent food for the fish. Those most sought 

 after are the elongated, more or less cylindrical eases of the Leptoeerid-ae and 

 Sericostomatidae, the former usually made from weeds, pieces of leaves or sticks, 

 the latter from grains of sand or a thin transparent substance secreted by the 

 iarva itself. In the Leptoceridae, the genera Notanatolica, Triplectides and 

 Oecetis are abundant throughout New Zealand; in the Sericostomatidae the same 

 is true of Olinga and Piicnocentria. Throughout the Hot Springs Region the 

 trout have most seriously diminished the number of these and other Caddis-flies. 

 Only two species now remain at all common, viz., Oecetis unicolor, whose larva 

 is still common, feeding in the green NitellaAj^As in the lakes, and Ilydropsyche 

 colonica, whose larvae form fixed houses of small pebbles attached to rocks. This 

 latter species still exists in great numbers in such places as the Okere Rapids, 

 where the rush of water is too swift for the trout to search for it. Its comparative 

 absence in other parts is strong- evidence of the reduction of the Caddis-fauna, 

 due to the trout. 



The most striking instance of the almost complete loss of the original rich 

 Caddis-fauna is afforded by the condition of the Te Wairoa stream, flowing into 

 Lake Tarawera. Except in the rough water below the Falls, where no trout 

 exist, it is almost impossible to oI)tain any caddises in the stream. But a 

 short distance off there is a nuich smaller stream, rising from a hill 

 near the lake. This stream has been dammed off by boards, and the water 

 drawn off close to its exit into the lake by a force-pump. No trout pass up 

 this small stream. On examining it I found that it was swarming with caddis- 

 larvae under every stone and stick, and upon the gravelly bed and sides of the 

 stream these little creatures were most abundant. Yet a day's search in the Te 

 Wairoa stream yielded far less than I was able to ]iick out in the course of ten 

 minutes in this tiny stream a mile away from it. 



It would not be overestimating the depredation caused by the trout amongst ' 

 the Caddis-fauna of the Hot Springs Region, if the loss were put at 90 % of the 

 original fauna. 



TirE St.\tk of the Insect Food Supply. 



My survey of the insects of the Hot Spring-s Region quickly convinced me 

 that the balance of nature has been completely upset by the introduction of the 

 trout. By comparison with the state of affairs in most parts of the South Island, 



