Figure 1.— Chart of Grays Harbor, WA, showing the four strata and random trawl sites selected from grid lines. 



trips per month coincided with low tides when 

 navigation and location of stations within the nar- 

 row channels characteristic of the North Bay and 

 Inner Harbor would be easiest. Stations situated in 

 the shallow, unmarked channels were sampled at 

 low water, while stations in more navigable areas 

 were sampled during the following periods of high 

 water. All hauls were confined to slack tide periods 

 so that strong tidal currents would not interfere 

 with the operation of the beam trawl. We sampled 

 the offshore study area once a month from April 

 through September. 



Field Techniques 



Within the estuary, the plumb staff beam trawl 

 was operated from a 6 m outboard boat, while off- 

 shore, either a 17 m or 20 m vessel was chartered. 

 Mechanical or hydraulic winches allowed the re- 

 trieval of the beam trawl while underway, reducing 

 escapement of captured organisms from the net. 



At all stations, the beam trawl was hauled for a 

 distance dependent on the amount of epibenthic 

 material expected. Hauls of <10 minutes duration 



(200-300 m) were necessary within the estuary 

 because the small mesh cod end liner would clog with 

 mud and organic debris, making the net difficult to 

 retrieve. Although the same problem existed in the 

 open coast survey area, hydraulic lifting gear en- 

 abled hauls of up to 20-30 min duration (800-1,000 

 m) at most stations. We attempted to tow at 3 km/h 

 with a scope of about 5:1; at depths <5 m, the scope 

 was increased to 10:1. 



Within the estuary, marker buoys were deployed 

 where we estimated the trawl started and stopped 

 fishing. The distance fished was determined by 

 measuring between buoys with an optical range- 

 finder. Along the open coast, Loran C readings were 

 recorded at points we estimated the trawl first con- 

 tacted and left bottom and were later converted to 

 distance fished. 



We sorted most of the catch by species (some 

 fishes to family only), and information was recorded 

 for total weight and number per haul. Random sub- 

 sampling of the catch (never <20% by weight) was 

 performed when necessary to speed processing. 

 Length frequencies and individual lengths and 

 weights were recorded for selected species. 



824 



