Detection efficiency and habitat use to inform inventory and monitoring efforts: juvenile coho salmon in the Knik River basin, Alaska

Abstract

AbstractImperfect detection associated with sampling gear presents challenges for wildlife inventory and monitoring efforts. We examined occupancy dynamics and habitat use of juvenile coho salmon, Oncorhynchus kisutch, in shallow lake environments over a summer and early fall season in the Knik River area of south central Alaska using models which control for and estimate sampling gear detection efficiency. In addition, we present statements for the probability that observed absences at a survey site or from a survey area (a collection of sites) are true absences given some amount of sampling effort and analysts' beliefs about site occupancy and sampling gear detection efficiency which can be used to guide inventory and monitoring efforts for juvenile salmon or other wildlife and plant species. Occupancy modelling results demonstrate that minnow traps were effective at sampling juvenile coho in shallow lake environments, with a mean probability of detection across the study period of 0.68 (i.e., probability of detecting the presence of juvenile coho given that they are present at a trap site; SE = 0.03). Juvenile coho salmon migrated into shallow water lakes in late summer and early fall, presumably to seek out overwinter habitat. N‐mixture modelling examination of habitat use demonstrated that once in shallow lake environments, juvenile coho were widely distributed across a range of microhabitats, with some evidence for preference for shallower depths and warmer water temperatures.