The effects of redd excavation by female quinnat salmon (Oncorhynchus tshawytscha walbaum) on the bentic invertebrate communities of two stable springfed spawning streams in the headwaters of the Rakaia River, South Island, New Zealand, were studied during January to September 1982. The gross effects of salmon spawning on the streams appear to be: the loss of instream macrophytes, algae and mosses, a loss of fines and detritus, and a modification of the pool-riffle character. Redd excavation may be associated with a reduction in benthic standing crop of up to 83.670, with only a minor effect on the benthic community structure. Benthic recolonisation rates were similar for both spawning streams, with almost complete recovery after 95 days.

The behavioral interaction between underyearling chinook salmon Oncorhynchus tshawytscha and brown trout Salmo trutta, both exotics in an unnatural sympatry in New Zealand, was tested during spring and summer in a simulated stream containing natural food. Both species were highly territorial and actively defended preferred drift-feeding sites and resting areas in the pool. Species dominance differed with season and was influenced by prior residence. In spring, chinook salmon were larger (because of their earlier emergence) and socially dominant, but they did not dominate when brown trout had prior residence. In summer, interspecific differences in size no longer existed, and brown trout were always socially dominant, although dominance required more time to establish when chinook salmon were prior residents. Most aggressive attacks of both species were initiated and won by the dominant fish in both spring and summer. Yearling brown trout in the pool in summer did not influence the interaction between underyearlings. Tactics in maintaining territory differed between species, with brown trout using a more energy-conserving strategy (frequent resting in preferred cover areas) than did chinook salmon (primarily free-swimming). Brown trout could adversely affect juvenile chinook salmon populations; therefore, this species should not be added to streams with chinook salmon.

The emergence timing, quality and diet of chinook salmon, Oncorhynchus ishawyrschn, fry were studied as they emerged from ten natural redds. The impact of selected environmental factors on emergence was studied also. Samples of fry were obtained daily using redd traps.The mean period of total emergence for redds with more than 450 fry emerging was 44.8 days (range 25-57 days), but for 80% of the fry in each redd to emerge the mean time elapsed was 10.8 days (range 6 21 days). No fry sampled had a visible yolk sac. The effect oftemperature and light on emergence timing was weak, but stream flow had a strong influence. Fry length was constant for each redd throughout emergence, but fry weight tended to decline; consequently condition factor declined also. There was evidence that fry fed before emergence, but they did not grow. The major prey item. Deleatidium spp. (47.4% of the contents of all stomachs sampled), was the most abundant invertebrate in stream benthos and drift. Empty stomachs were prevalent at the start of emergence, with full stomachs becoming more common as emergence progressed.The time distribution of numbers of fry emerging was usually unimodal, but occasionally bimodal. Over one third of fry emerged during the best 2 days of sampling.

A drift sampler for use in small streams is described. The mouthpiece is the rectangular end of a plastic adaptor for rainwater guttering, and a 1 m long, 45 µm mesh net is attached to the circular posterior. Three such adaptors, plus nets, are placed side by side 70 mm apart and attached to aluminium flat. The whole structure is attached to metal stakes driven into the stream bed. The catch of the sampler is compared with a standard drift sampler, and the replicates are analysed for variance.

No abstract