l2-ISIb. OANFORTH ANCHOR 

 20ft 1/4* CHAIN 



HANGER and HOOP DETAIL 



Figure 1. — Buoyed nets for sampling larval herring. 



A, positions of four nets during slack and flowing tides; 



B, arrangement of hanger and hoop (After Graham and 

 Venno, 1968). 



divided into the total number of larvae captured 

 and multiplied by 100 to obtain catches per 

 100 ml Thus, at each location (landward and 

 seaward) average catch rates from the two lines 

 of nets were obtained for two depths (shallow 

 and deep) during two tidal stages (flood and 

 ebb) , providing eight samples for each location 

 or 16 samples for a single overnight experiment. 

 The nets were fished during darkness because 

 the larvae avoided the nets during daylight. A 

 comparison of nighttime catches between two 

 lines of buoyed nets set in tandem within the 

 channel and a comparison between nighttime 

 catches of buoyed nets with those of a high speed 

 Boothbay Depressor Trawl (Graham, Cheno- 

 weth, and Davis, 1972) suggested that the larvae 

 did not avoid the net at night nor did they escape 

 from the nets during slack water. The average 

 catches above and below mid-depth determined 

 the concentration of larvae in the water layers 

 above and below the level of no-net-motion of 

 the tidal flows. Larvae in the upper layer would 

 experience a net seaward transport and those in 

 the lower layer would experience a net landward 

 transport. In addition to sampling larvae al- 

 ready in the channel, the two lines of nets in the 

 landward location would sample larvae trans- 

 ported into the channel by the ebb tide and the 



FISHERY BULLETIN: VOL. 70. NO. 2 



-i44°or 



Wiscasset ^/ 



LANDWARD SETS 



43°45' 



69°45 



69°36 



Figure 2. — Locations (circled) of buoyed 

 and anchored nets in the Sheepscot estu- 

 ary of Maine (after Graham and Davis, 

 1971). 



two lines in the seaward location would sample 

 larvae transported into the channel by the flood 

 tide. 



The arrangement of the nets and catches in 

 the estuarine channel constituted a factorial 

 sampling design. A factorial design is especially 

 adaptable to examining the effects of a number 

 of diff'erent factors which vary in a regular way 

 during an experiment. The source of regularity 

 in the experiments was the tidal currents trans- 

 porting the larvae. An example of the statistical 

 procedures is given in Table 1. Catch rates were 

 transformed into common logarithms to obtain 

 a normal distribution of the data for computa- 

 tions. In this paper the untransformed catch 



300 



