Quantifying Fish Habitat in Streams: Transect Spacing, Sample Size, and a Proposed Framework

Procedures for evaluating fish habitat in streams have focused largely on specific methods used to sample individual habitat variables, and studies ofsampling design are uncommon. We used data from 86 sites on 58 Wisconsin streams to determine the optimal number and spacing of habitat transects needed to characterize the means of commonly measured habitat variables. The optimal number of transects varied with stream width; approximately 13 transects, spaced every three mean stream widths (MSW), were required on narrow streams (<5 m wide), and approximately 20 transects, spaced every 2 MSWs, were needed on wider streams (5–35 m wide). Spacing transects in terms of MSW yields equal sample sizes regardless of stream width, is easier to apply than random spacing, and is logistically more reasonable than spacing transects at regular intervals without regard for stream width. Estimates based on transects spaced 2 MSWs apart within a stream reach of 35 MSWs were within 5% of the true values 95% of the time, Statistical differences that could be detected by means of 20 transects were reasonable and probably biologically meaningful.