Potential impacts on zooplankton of seismic surveys

Environmental Impact Assessment

Richardson, Anthony ; Matear, Richard ; Lenton, Andrew

CSIRO

DOI  :  10.4225/08/59724f38211cd
RP-ID  :  EP175084

学科分类：地球科学(综合)

澳大利亚|英语

来源: CSIRO Research Publications Repository

【 摘 要 】

Context: While the impact of seismic surveys on marine mammals has been well studied, there has been little work on zooplankton. The study by McCauley et al. (2017) published on the 23 June 2017 is the first large-scale field experiment on the impact of seismic activity on zooplankton. Their study overturns the conventional idea of limited and localised impact on zooplankton. They found that air gun exposure significantly decreased zooplankton abundance, and increased the mortality rate from a natural level of 19% per day to 45% per day (on the day of exposure). These impacts were observed out to the maximum assessed range of 1.2 km. McCauley et al. (2017) state that there is an urgent need to conduct further study to mitigate, model and understand potential impacts on plankton.Methods: Here we simulate the large-scale impact of a seismic survey on zooplankton, assuming the mortality rate associated with air gun exposure reported by McCauley et al. (2017). Our approach models a hypothetical survey on the edge of the Northwest Shelf during summer. The survey area was 80 km by 36 km in water 300-800 m deep and the simulation was conducted over a 35-day period,. To simulate the movement of zooplankton by currents, we used a hydrodynamic model that seeded 0.5 million particles into CSIRO’s Ocean Forecast Australia Model. Zooplankton particles could be hit multiple times by the air gun if they were carried by currents into the future survey path. Each particle represents a zooplankton population exhibiting logistic population growth. The greatest limitation in this approach was accurate knowledge of the natural growth and mortality rates of zooplankton. We thus tested the sensitivity of the model to different recovery (growth-mortality) rates, and also the sensitivity of our results to ocean circulation by undertaking simulations with and without water motion. We report the relative zooplankton biomass in our simulations – the ratio of zooplankton biomass following a seismic survey relative to biomass in the absence of a survey, from 0 (all dead) to 1 (no impact). We report results on four regions relevant to management and of varying size.Results: Simulations that included ocean circulation showed that the impact of the seismic survey on zooplankton biomass was greatest in the Survey Region (0.78, i.e., 22% of the zooplankton biomass was removed) and declines moving to the Survey Region + 15 km (0.86), and the Survey Region + 150 km regions (0.98, see Table for values); there was no discernible effect on the entire Northwest Shelf Bioregion. The time to recovery (to 95% of the original level) for the Survey Region and Survey Region + 15 km recovery was 39 days (38-42 days) after the start of the survey and 3 days (2-6 days) after the end of the survey.Simulations with no ocean circulation showed a much greater impact of the seismic survey on relative zooplankton biomass: 0.65 for the Survey Region; 0.78 for the Survey Region + 15 km; 0.97 (0.97-0.97) for the Survey Region + 150 km; and no discernible effect on the entire Northwest Shelf Bioregion. The time to recovery for the Survey Region from the start of the survey was 64 days (49-100 days) and from the end of the survey was 26 days. Discussion: Applying the mortality rate from McCauley et al. (2017), we found considerable impact within the seismic survey area and within 15 km of it. However, these impacts are not discernible at the largest scale of the Northwest Shelf Bioregion and are barely discernible within 150 km of the survey area. Zooplankton populations recovered quickly after seismic exposure due to their fast growth rates, and the dispersal and mixing of zooplankton from both inside and outside of the impacted region. Finally, we make suggestions about how future studies could be designed and optimized using tools developed in the current study – to test the findings of McCauley et al. (2017).