Monitoring the spawning behaviour of chum salmon with an acceleration data logger

Abstract

Abstract – The continuous observation of salmon behaviour in some wild environments can be extremely difficult. We recorded spawning behaviour of female chum salmon (Oncorhynchus keta) in the field simultaneously using visual observation and fish‐borne data loggers with two‐axis accelerometer sensors. Using only acceleration records, behaviours were successfully classified into the eight well‐known components of spawning behaviour: swimming, nosing, exploratory digging, nest digging, probing, oviposition, covering and post‐spawning digging. To understand how the female chum salmon modulates spawning behaviours in relation to changes in environmental conditions, we compared the behaviours of salmon during normal flow of clear water to those of salmon during the heavy flow of turbid water after a storm. Salmon in the normal flow showed all eight behaviours, whereas salmon in the heavy flow showed only three behaviours: swimming, nosing and exploratory digging. The proportion of time spent on swimming was greater in the heavy flow than in the normal flow (mean of 98.47% vs. 92.84%). Moreover, the proportion of tail beating in swimming was greater in the heavy flow (77.86%) than in the normal flow (15.63%). Our results indicate that the behaviour of female chum salmon was strongly affected by the heavy flow of turbid water after a storm. The recording of accelerations is a promising method for clarifying the spawning behaviour of salmonids in the wild where continuous visual observation is too difficult.