Laserfiche WebLink
Surface Observation <br />Surface observation is difficult in swift, turbulent, or turbid water. <br />Many potential subjects may be overlooked or mis-identified. Therefore, <br />observation techniques are best applied in clear water where the Fronde <br />number (f = V/ gd, where V is the velocity, g is the acceleration due to <br />gravity, and d is the depth) is <0.5. <br />Electrofishing <br />The limitations of this method were discussed earlier. They may be <br />overcome by using a dead man switch to spot-sample discrete areas, <br />rather than running the field continuously. <br />Drift Gill Nets <br />Drift gill nets can only be safely used where the channel is free from <br />snags. Fish sampled by gill netting should give a distinct reaction <br />(e.g. shovelnose sturgeon (Scaphirhynchus platorynchus) break to the <br />surface) when first netted, so that their positions .may be documented. <br />SCUBA or Snorkel Observation <br />Fish may be alarmed by the presence of a diver, depending. on whether he <br />approaches them from upstream or downstream. Fish reactions may vary, <br />so it may be wise to experiment somewhat with approach patterns. <br />Baited Hoop Nets (and Trot Lines) <br />These are inherently biased since the fish are attracted to the sampling <br />location from somewhere else in the stream. However, for some species, <br />such as channel catfish Ictalurus unctatus they are the most effic- <br />ient sampling devices. Sampling bias may be reduced somewhat by elim- <br />inating subjects drawn from outside the sample reach. This may be <br />accomplished by setting a battery of "control" hoop nets at the most <br />downstream limit of the sample reach, or by placing an "electric weir" <br />across the lower end of the reach. These controls will intercept most <br />of the in-migrants to the section. Hoop nets within the sample reach <br />should be situated so that they are sampling a discrete habitat area. <br />Sampling bias may be reduced by a statistically random selection of <br />sampling areas, and by using a combination of more than one sampling or <br />observational technique. It is unlikely that bias can be eliminated <br />completely, but it should be reduced wherever possible. <br />28 <br />