Abstract
Tidal inlets such as the FORCE demonstration site are dynamic regions that provide important habitat for harbour porpoise (Phocoena phocoena). Harbour porpoise use echolocation to hunt and communicate (Kastelein et al. 2002), and they are known to be susceptible to noise disturbance (Tougaard et al. 2009). Few studies to date have focused on exposure to continuous low frequency noise such as that emitted by tidal turbines. The tidal dynamics inform the presence of porpoises in these areas in complex ways. Hence, long-term and ongoing monitoring of this variability has been an important component of understanding the impacts of installing tidal turbines at this site. FORCE contracted SMRU Consulting (Canada) to complete equipment calibration and click detection data analysis relating to the deployment of passive acoustic monitors (C-PODs) in support of its marine mammal environmental effects monitoring program (EEMP). C-PODs can detect echolocating cetacean species including dolphins, but not whales. Monitoring began on 5 May 2011 and is ongoing. This report encompasses monitoring up to May 1, 2018, and includes the 160 day Cape Sharp Tidal Venture (CSTV) turbine trial at Berth D of FORCE test site. The installation of the CSTV tidal turbine occurred on 7 November 2016. The cable connection to the turbine was disconnected on April 21, 2017 (during which time the turbine had the ability to free-spin) and the turbine Was removed June 15, 2017.
This report firstly summarizes the dynamic temporal patterns in porpoise presence in Minas Passage from 2011-2018 related to key environmental covariates, notably annual, seasonal, tidal and day versus night variability. It is important to note that temporal coverage was intermittent over this seven year period, with just two winter-early spring periods of baseline that overlapped with the CSTV turbine trial. We provide a statistical analysis of the distribution and activity of harbour porpoise around the FORCE demonstration area in response to the installation and operation of the turbine using data from 2 mid-range C-PODs (within 230 m), and three located at least 1 km away.
From May 2011 through to May 2018, there have been 1210 monitoring days and 4651 C-POD days, spread across eight locations within and immediately adjacent to the FORCE site. Overall, harbour porpoises have been detected on 98.9% of days at a median of seven detection positive minutes per day and maximum of 44 minutes. When the turbine was operating, porpoises were detected on 82% of days. No dolphins were detected during any of the C-POD deployments at any of the eight C-POD locations. C-PODs a reported to have an effective porpoise click detection range 188 m (Nuuttila et al. (2018), with decreased performance reported when ambient noise exceeds 105 dB re. 1 μPa (rms) and false alarm rates less than 2% (Clausen et al. 2018). Comparison of porpoise detection rates using C-PODs and post-deployment analysis of broadband recordings using alternate click detection software such as PAMGuard (Porskamp et al. 2015, Sarnocinska et al. 2016, Jacobsen et al. 2017) are subject to the effects of variable hydrophone sensitivities, issues with determining false classifications and variability in deployment methods. Nevertheless, these studies do highlight porpoise detection probabilities using buoy-mounted C-PODs should be considered minimum estimates.
A statistical model using data from the five C-PODS that overlapped in time with the CSTV turbine trial was built to test the effects of the turbine while accounting for the natural dynamics of the FORCE test site. The model results confirmed an effect on porpoise presence at the closest two mid-field monitoring stations. Additionally, porpoise presence varied significantly by time of year (i.e., peak period May/June and lower secondary peak October/November), by current speed and tidal height (i.e., preference for 0-2.5 m/s ebb tides), by time of day (i.e., higher activity at night) and across the lunar cycle (i.e., affected by the position in the spring-neap tidal cycle). CPODs monitor underwater noise each minute, however, in each minute, the units will only ‘listen’ until a maximum buffer is reached (called the ClickMax buffer). If this buffer is reached then the remainder of the minute monitoring is lost (termed Percent Time Lost). C-POD performance restrictions (Percent Time Lost) also varied due to noise effects, notably due to non-biological clicks associated with sediment transfer during periods of relatively high current velocity. Percent Time Lost had little effect on data quality between an ebb current speed of <2.4 m/s (95% of 10-minute periods) and a flood current speed of <2 m/s (71% of 10-minute periods). At ebb current speeds up to 2.9 m/s (99% of 10-minute periods) and flood current speeds up to 3.5 m/s (95.5% of 10-minute periods), Percent Time Lost does not exceed 65%. Despite the use of statistical methods to take Percent Time Lost into account, C-POD monitoring performance above these current speeds is clearly less reliable, noting that these speeds only occur a very small fraction of the tidal cycle.
During the CSTV turbine trial, porpoises were detected at all five monitoring locations on 128 of the 130 days, and 106 of 129 days at D1, and 76 of 130 days at East1. Declines in daily porpoise presence with the operating turbine from similar time periods of baseline monitoring equated to ~35% for East1 and ~7% for D1. Consequently, there was no evidence of porpoise exclusion of the mid-range (210 – 1710 m) study area post-installation, noting that changes in the overall distribution of porpoise within the vicinity of the turbine is considered of higher importance.
A statistical model (that accounted for the environmental variables and Percent Time Lost) tested for changes in the distribution and activity of harbour porpoise in relation to the installation and operation of the turbine. The overall effect of turbine operations on porpoise detection rates were found to be significant (P<0.01). East1, a site 210 m north of the turbine at 41 m depth, and D1, a site 230 m south of the turbine at 33 m depth both showed statistically fewer porpoise detections post installation of the turbine. Both these sites had overall lower activity levels both with and without the turbine, whereas the sites > 1 km west and south of the turbine had overall higher activity levels and showed no decrease in porpoise detections with the turbine. The obvious and immediate drop in detections observed during the previous analysis of the first 73 days post-installation at East1 as well as West1 and West2 are believed to represent disturbance from vessel activity and initial turbine installation, while subsequent dips observed after this period considered to reflect continued winter and also lunarscale fluctuations related to lower detection performance of C-PODs during all spring tides (higher % lost time). These observations coupled with high levels of inter-annual and site variability and the short (166) post-installation period, result in the overall conclusion that further C-POD data collection is required before robust inferences can be drawn, but preliminary statistical results suggest a mid-range turbine effect within 230 m. Continued C-POD monitoring will allow for a better understanding of the variables that inform the environmental dynamics of the FORCE demonstration area.