304 



FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE 



best fyke-netting sites. Here the river varies in 

 width from about 400 feet to 1,000 feet, with 

 depths to more than 15 feet and surface water 

 velocities to at least 8 feet per second. The 

 bottom consists mostly of gravel less than 4 

 inches in diameter. The riverbanks ai-e steep 

 because of the erosion that occurs nearly every 

 year during high water. The river downstream 

 from the Kaskanak Flats is affected by tides, 

 causing variable water velocities unfavorable 

 for fyke netting. 



The index site (site B) used for smolt enu- 

 meration was approximately 4 miles down- 

 stream from the outlet of Iliamna Lake. Two 

 other sites designated A and C were 1/2 n^ile 

 and 2 miles downstream from the outlet (fig. 

 4). At the index site the river forms two chan- 

 nels, 570 feet wide and 225 feet wide. The fyke 

 net was set on a submerged, slightly sloping 



Scole in fee) 



^—^ — I — t-H 



1000 2000 



Koskanok 

 Flots - 



Figure 4. — Fyke-net sites A, B (index), and C, Kvichak 

 River. (Map by D. W. Linn.) 



gravel bar near the centei- of the main channel. 

 This gravel bar extends at the same depth for 

 at least 100 yards above and below the index 



site. 



Fyke-Netting Procedure 



The fyke net was set before 2200 hours each 

 night and tended from a 20-foot skiff by two 

 or more men. To check the net, the cod end was 

 first raised out of the water by one man. In this 

 position the throat of the net was closed to the 

 passage of fish. The cod end with the catch was 

 detached by removing a single locking pin and 

 an empty cod end immediately put in place and 

 lowered into the water to continue fishing. 



The second man pulled the release cord or 

 zipper on the cod end and spilled the catch into 

 a large weighing basket that was immersed in a 

 tub of water. This weighing basket was then 

 removed from the tub and hung on a spring 

 balance of 40-pound capacity, suspended from 

 a weighing stand in the skiff. The weighing 

 basket was allowed to drain about 10 seconds 

 before the weight of the fish and the time of 

 catch were recorded. During peak migration 

 the fish weighed about 20 pounds at a net check. 

 Immediately after being weighed, the fish were 

 returned to the river. Elapsed time for a net 

 check was less than 30 seconds. The number of 

 fish in a 1-pound sample was usually counted 

 four times an hour and the number was used 

 for conversion of total weight of fish to total 

 number. 



To avoid excessive mortality during periods 

 of heavy migration, the net was checked as the 

 fish accumulated. In an extreme instance in 

 1958 it was necessary to check the index fyke 

 net nearly twice a minute. When the migra- 

 tion was very light, the net was checked every 

 hour. 



Fishing Season 



The experimental fishing season during this 

 study began in the spring before any smolt 

 migration takes place and continued until only 

 a few fish were caught each day. The smolt 

 migration started after ice breakup (fig. 5) and 

 following a rapid rise in water temperatures in 

 Iliamna Lake. 



