Knowing how many individuals there are in a population is a fundamental problem in the management and conservation of freshwater and marine fish. We compare abundance estimates (census size, Nc) in seven brook trout Salvelinus fontinalis populations using standard mark–recapture (MR) and the close‐kin mark–recapture (CKMR) method. Our purpose is to validate CKMR as a method for estimating population size. Close‐kin mark–recapture is based on the principle that an individual's genotype can be considered a “recapture” of the genotypes of each of its parents. Assuming offspring and parents are sampled independently, the number of parent–offspring pairs (POPs) genetically identified in these samples can be used to estimate abundance. We genotyped (33 microsatellites) and aged c. 2,400 brook trout individuals collected over 5 consecutive years (2014–2018). We provide an alternative interpretation of CKMR in terms of the Lincoln–Petersen estimator in which the parents are considered as tagging the offspring rather than the offspring “recapturing” the parents. Despite various sources of uncertainty, we find close agreement between standard MR abundance estimates obtained through double‐pass electrofishing and CKMR estimates, which require information on age‐specific fecundity, and population‐ and age‐specific survival rates. Population sizes (trueN^) are estimated to range between 300 and 6,000 adult individuals. Our study constitutes the first in situ validation of CKMR and establishes it as a useful method for estimating population size in aquatic systems where assumptions of random sampling and thorough mixing of individuals can be met.

Mortality and predation of tagged fishes presents a serious challenge to interpreting results of acoustic telemetry studies. There is a need for standardized methods to identify predated individuals and reduce the impacts of “predation bias” on results and conclusions. Here, we use emerging approaches in machine learning and acoustic tag technology to classify out-migrating Atlantic salmon (Salmo salar) smolts into different fate categories. We compared three methods of fate classification; predation tag pH sensors and detection data, unsupervised k-means clustering, and supervised random forest combined with tag pH sensor data. Random forest models increased predation estimates by 9-32% compared to pH sensor data, while clustering reduced estimates by 3.5-30%. The greatest changes in estimates were seen in years with large class imbalance or low model accuracy. Both supervised and unsupervised approaches were able to classify smolt fate, however, in-sample model accuracy improved when using tag sensor data to train models, emphasizing the value of incorporating such sensors when studying predator-prone fish. Sensor data may not be sufficient to identify predation in isolation due to Type I and Type II error in predation sensor triggering. Combining sensor data with machine learning approaches should be standard practice to more accurately classify fate of tagged fish.

Mortality and predation of tagged shes presents a serious challenge to interpreting results of acoustic telemetry studies. There is a need for standardized methods to identify predated individuals and reduce the impacts of "predation bias" on results and conclusions. Here, we use emerging approaches in machine learning and acoustic tag technology to classify out-migrating Atlantic salmon (Salmo salar) smolts into different fate categories. We compared three methods of fate classi cation; predation tag pH sensors and detection data, unsupervised k-means clustering, and supervised random forest combined with tag pH sensor data. Random forest models increased predation estimates by 9-32% compared to pH sensor data, while clustering reduced estimates by 3.5-30%. The greatest changes in estimates were seen in years with large class imbalance or low model accuracy. Both supervised and unsupervised approaches were able to classify smolt fate, however, in-sample model accuracy improved when using tag sensor data to train models, emphasizing the value of incorporating such sensors when studying predator-prone sh. Sensor data may not be su cient to identify predation in isolation due to Type I and Type II error in predation sensor triggering. Combining sensor data with machine learning approaches should be standard practice to more accurately classify fate of tagged sh. Highlighted Student PaperThis paper contributes signi cantly to the eld of ecology by introducing a standardized work ow for analyzing telemetry data which is greatly needed to reduce biases in study results.