FISHERY BULLETIN: VOL. 74, NO. 1 



all four streams. Chef and Waterloo creeks also 

 contain steelhead trout, S. gairdneri, and chum 

 salmon, O. keta. Chum salmon occasionally 

 spawn intertidally in Lymn Creek. 



METHODS AND MATERIALS 



Sampling the Populations 



In Lymn Creek, sculpins were collected inci- 

 dentally to salmonids from April to July 1968. A 

 sampling schedule for cottids was initiated in 

 August and terminated in December 1968. Chef, 

 Cabin, and Waterloo creeks were sampled during 

 September and October. 



Fish were collected in the estuaries by seine at 

 low tide. In the streams proper, collections were 

 made with a 440- V DC fish shocker (Smith-Roote 

 Laboratories, Mark V^). In both environments, 

 discrete sections of stream, usually 15- to 30-m 

 sections, were sampled and all fish captured were 

 removed. 



Specimens were preserved in 5% Formalin. In 

 the laboratory, total length was measured to the 

 nearest millimeter and body weight to the near- 

 est 10 mg. Otoliths were removed for age deter- 

 mination. 



No attempts were made to quantify the relative 

 or absolute efficiencies of the two sampling 

 methods. The habitat seined lent itself to efficient 

 seining, and it is considered that any increased 

 capture efficiency or size-related sampling bias 

 usually associated with electrical fishing devices 

 was, at least in part, cancelled by the increased 

 complexity of habitat typical of the stream proper 

 and the concentration of the two youngest age- 

 groups in the lower stream, including the es- 

 tuaries. Increased stream flow and turbid water 

 following the first significant rains in the late fall 

 probably reduced the efficiency of both collecting 

 methods to a considerable but unknown extent. 

 Therefore, growth and survivorship estimates 

 were based on data collected prior to the onset of 

 the rainy season. 



In the laboratory, breeding activity was fol- 

 lowed by keeping adults allopatrically in 150- 

 liter fiber glass tanks at ambient freshwater 

 temperature with flow-through conditions, a rub- 

 ble substrate, and normal photoperiod. Em- 



'Reference to trade names does not imply endorsement by the 

 National Marine Fisheries Service, NOAA. 



bryological development and larval responses to 

 salinity, illumination, current, and food were 

 investigated. Egg masses of known age and their 

 resulting larvae were kept in 3-liter glass jars 

 filled with aerated fresh water or seawater; and 

 mortality and feeding responses of larvae to mi- 

 crozooplankton were observed. The responses of 

 larvae of known age and salinity history to over- 

 head illumination and water currents were inves- 

 tigated in a Perspex test chamber. 



Drift nets were set at several stations in 

 Lymn Creek during the hatching period in the 

 spring to document the timing and extent of the 

 hatching period, upper limits of the spawning 

 ground and characteristics of the fry moving 

 seaward. 



Population Estimates 



Estimates of population size in Lymn, Cabin, 

 and Waterloo creeks were attempted in the fall 

 for both species of sculpin. Population estimates 

 for Chef Creek were precluded by the large size of 

 the stream, which prevented representative sam- 

 pling across the stream at most stations. In the 

 other three streams, catches from individual sta- 

 tions were assumed to be representative of that 

 stream section, and population was calculated as 

 follows: 



N= ICD 



where C = station catch (fish/meter of stream) 

 where each station is representative 

 of a larger stream section D 

 D - stream section (in meters). 



The estimated populations were distributed 

 among the various age-classes so as to reflect the 

 age-class composition of the station catches. Ad- 

 mittedly, these estimates are rather crudely de- 

 rived yet they yielded fairly consistent trends in 

 annual mortality, particularly for the Lymn 

 Creek populations (see Results, Annual Growth, 

 Mortality, and Length-Weight Relations). At- 

 tempts to apply mark and recapture techniques to 

 the problem of population estimation proved 

 fruitless due to extensive behavior changes 

 in marked fish following their release. These 

 changes (movement downstream or into the 

 streambed) seriously affected their vulnerability 

 to recapture and led to large scale overestimates 

 of actual population size. 



132 



