Summary
A grid of underwater click recorders has been run in the vaquita refuge for about three months each year since 2011, and this paper reports the 2016 results. Acoustic detections fell by 49% between 2015 and 2016 (95% credible interval, 82% decline to 8% increase), and by more than 90% overall between 2011 and 2016. Assuming detections track population size, about 30 vaquitas (95% credible interval 8 to 96) remained in November 2016.
Why it matters
It gives the 2016 figure of about 30 vaquitas, and it showed that acoustic activity was still falling after the 2015 gillnet suspension began.
Key findings
- Acoustic detection rate fell 49% from 2015 to 2016 (posterior mean; 95% credible interval 82% decline to 8% increase); probability of a decline was 0.96.
- Average change from 2011 to 2016 was 0.61 per year (95% credible interval 0.48-0.74), a 39% decline per year (95% credible interval 26% to 52%), about 90% in total.
- Projected population on 2 November 2016: approximately 30 (posterior mean 33, median 27, 95% credible interval 8 to 96).
- The projection starts from the autumn 2015 estimate of Taylor et al. (2016), a lognormal with mean 66 and SD 33, multiplied by the 2015-2016 change in acoustic activity.
- Forty-six loggers (C-PODs) were deployed June to August each year 2011-2016 in the refuge; 47 more were added in mid July 2016.
Caveats
The 2015-2016 change is too imprecise to say whether it differs from earlier years, and its interval includes no change. Turning acoustic activity into population size assumes vaquita click behaviour and detection have not changed, and that the trend outside the refuge (about 20% of the population in 2015, 12 of an estimated 59) matches the trend inside it.