LAUSR

More than half of the diminutive North American catfish species known as madtoms (Noturus spp.) are considered imperiled due to range reductions from habitat fragmentation and degradation. In response, government agencies regularly conduct sampling that targets these cryptic fishes. Unfortunately, the effectiveness and biases associated with common sampling gears has rarely been quantified for non‐game fish, including madtoms. To improve sampling protocols for diverse madtom species, we: (a) quantified the detection probabilities and effort necessary to detect five species of madtom with relative certainty using complimentary sampling gears, and (b) described biases associated with size‐selectivity of the same gears when used to sample madtoms. Five madtom species that are native to Mississippi streams and have two distinct coloration patterns were selected: uniform pattern, Brown Madtom (Noturus phaeus) and Freckled Madtom (Noturus nocturnus); patterned, Piebald Madtom (Noturus gladiator), Least Madtom (Noturus hildebrandi), and Brindled Madtom (Noturus miurus). These species were captured using four standardized sampling gears (i.e., backpack electrofisher, seine, dipnet, and Gee‐style minnow trap) from 13 isolated stream reaches in Mississippi during baseflow conditions between August and November 2016. Each gear was randomly assigned to up to five transects at each site resulting in a maximum of 20 unique samples per stream reach. We estimated the detection probability of each species with each gear using presence–absence capture data. Cumulative detection probabilities were calculated for each species and gear combination to describe the number of samples necessary to achieve a detection probability of 0.95. Additionally, we pooled all capture data across sampling sites to compare madtom size distributions among gears. Electrofishing was generally the most efficient gear for detecting madtoms, but slightly overestimated fish size distributions. By seining, managers can expect to capture all species evaluated with a modest amount of effort; however, are likely to catch relatively small individuals. Seining was the most effective gear for only Piebald Madtom, a species petitioned for listing under the Endangered Species Act. To be relatively certain that Piebald Madtom are absent from a stream reach would require sampling ten 100 m transects by seining. In contrast, the same certainty would require more than twenty 100 m electrofishing transects. Minnow traps and dipnetting generally performed poorly and failed to detect some species at occupied sites. Biologists can now consider using each of these complimentary gears assessed when evaluating species status and population structure for diverse madtom species, while explicitly acknowledging the biases associated with each gear.