Application of machine learning to identify predators of stocked fish inLake Ontario: using acoustic telemetry predation tags to inform management

Abstract

Understanding predator–prey interactions and food web dynamics is important for ecosystem‐based management in aquatic environments, as they experience increasing rates of human‐induced changes, such as the addition and removal of fishes. To quantify the post‐stocking survival and predation of a prey fish in Lake Ontario, 48 bloater Coregonus hoyi were tagged with acoustic telemetry predation tags and were tracked on an array of 105 acoustic receivers from November 2018 to June 2019. Putative predators of tagged bloater were identified by comparing movement patterns of six species of salmonids (i.e., predators) in Lake Ontario with the post‐predated movements of bloater (i.e., prey) using a random forests algorithm, a type of supervised machine learning. A total of 25 bloater (53% of all detected) were consumed by predators on average (± S.D.) 3.1 ± 2.1 days after release. Post‐predation detections of predators occurred for an average (± S.D.) of 78.9 ± 76.9 days, providing sufficient detection data to classify movement patterns. Tagged lake trout Salvelinus namaycush provided the most reliable classification from behavioural predictor variables (89% success rate) and was identified as the main consumer of bloater (consumed 50%). Movement networks between predicted and tagged lake trout were significantly correlated over a 6 month period, supporting the classification of lake trout as a common bloater predator. This study demonstrated the ability of supervised learning techniques to provide greater insight into the fate of stocked fishes and predator–prey dynamics, and this technique is widely applicable to inform future stocking and other management efforts.

Acoustic telemetry predation sensors reveal the tribulations of Atlantic salmon (Salmo salar) smolts migrating through lakes

“…A paper by Klinard et.al. (2021) revealed a 17% false positive rate, utilising the larger V9 tags, and this could impact our study as well. However, based on the movement (lack of movement) of predated smolts and the findings by Halfyard et al (2017) showing that false positives first were observed after 47 days for the V5 tag, we argue this is not likely to be a substantial source of error in this study.…”

Implications of habitat use and movement of stocked juvenile dusky flathead (Platycephalus fuscus) on stock enhancement release strategies

“…A VPS requires close spacing of numerous receivers and can only normally cover a small proportion of a study area, particularly in the presence of macrophytes (Espinoza et al, 2011; Steel et al, 2014), whereas Lake Macquarie is the largest coastal lagoon in Australia, with extensive seagrass beds along its eastern shore. Fish would only need to move a few hundred metres to occupy adjacent habitat to avoid detection within the VPS array, so the reduction in detections we observed over time may have been a result of predation on released fish, which is difficult to differentiate from dispersal (Sherwood et al, 2004), although “predation tags” that can detect predation on tagged fish (e.g., Klinard et al, 2021) may help to distinguish mortality from dispersal.…”