Long-term acoustics monitoring of zooplankton dynamics in Saanich inlet (British Columbia, Canada)
Long-term acoustics monitoring of zooplankton dynamics in Saanich inlet (British Columbia, Canada)
Abstract:
Ocean Networks Canada (ONC, http://www.oceannetworks.ca/) operates underwater cabled observatory networks to deliver real-time data for monitoring the marine environments and ecosystems in the NE, subarctic Pacific, and Arctic. The longest continuous time series on the ONC network is located in Saanich Inlet, an inverse estuary at the southeastern end of Vancouver Island (British Columbia, Canada). Here, a ‘zooplankton acoustic profiler’ (inverted echo-sounder) has been mounted on the Saanich Inlet instrument platform since 2006. The single-frequency Zooplankton Acoustic Profiler (ZAP, 200kHz) and more recently a multi-frequency Acoustic Zooplankton Fish Profiler (AZFP, 38, 125 and 200 kHz) have been deployed during 2006-2016 and 2016-present, respectively. Measurements of environmental and oceanographic indices (e.g. temperature, salinity, oxygen and tide height) have been collected via multiple observatory system sensors concurrently. These thirteen-year (2006-2019) high-resolution time-series allow for the first characterization of zooplankton dynamical features and exploration of driving forces/processes on multiple time scales using Empirical Dynamic Modelling. The fluctuations in acoustically measured migration zooplankton biomass (mostly contributed by Euphausia pacifica) and variations in measured non-biological indices display significant nonlinearity (θ > 0) on weekly, biweekly, monthly and quarterly scales. Temperature, salinity and oxygen are identified as significant causal factors of nonlinear dynamics in the migratory zooplankton biomass time series on quarterly and monthly scales in this system, while tide height significantly drives the nonlinear dynamics on biweekly and weekly scales. In addition, we demonstrate forecasting on migration zooplankton biomass using single variable (e.g. bioacoustics) and using multiple variables.