We developed a gillnetting protocol by sampling 17 nonwadeable rivers across northern Ontario over the course of four summers from 2005 to 2008. The rivers represented a range of habitats; however, all had a coolwater to warmwater fish community characterized by walleyes Sander vitreus, northern pike Esox lucius, white suckers Catostomus commersonii, yellow perch Perca flavescens, and smallmouth bass Micropterus dolomieu. During the study, three net designs were used. Each net had one to four configurations that varied in height, length, monofilament diameter, and sequence of the mesh sizes. We found that most fish (87%) were captured in the lower half of 1.8‐m‐high gill nets and that there was no difference in catch between gill nets with serial panels (i.e., sequentially increasing mesh sizes) and those with randomly organized panels. Net that were set perpendicular to shore or angled to shore caught approximately twice as many fish as nets that were set parallel to shore. Contrary to expectations, there were no differences in net durability and catch between thick‐ and thin‐diameter monofilament nets. Based on the results of these experiments, we propose an improved net design that we call the large‐mesh riverine index net for use in rivers; it is more versatile, reduces net drag, and fishes more effectively than the traditional gill nets that are used in lakes. The large‐mesh riverine index net has eight panels (3.1 m long × 0.9 m high) and totals 24.8 m in length. Panel stretch measures (and the random order used) for this design are 127, 76, 178, 25, 102, 203, 51, and 152 mm. The nets are set perpendicular to shore in rivers from July 1 to October 1 and should be fished in areas where water velocity is less than 0.1 m/s.
Received January 12, 2010; accepted November 2, 2010