Field tests of a directional parametric acoustic alarm designed to alert manatees of approaching boats.

Abstract

The efficacy of the alarm was documented during controlled slow boat approaches under two experimental conditions: (1) approaches without an alarm and (2) same boat approaches with an alarm. Experiments conducted in a NASA security area provided controlled environmental conditions and minimized anthrogenic acoustical variables. An array of GPS instrumented buoys were deployed to acoustically and visually grid sites. HD aerial video synchronized with these buoys documented behavior and acoustic conditions at focal animal positions and throughout sites. Bathymetry and propagation measurements indicate shallow water constraints along with Lloyd’s mirror effect resulting in significant transmission losses at frequencies of 1 kHz. This could account for the lack of response observed during 94% of the boat approaches without the alarm. In contrast, 100%, of the alarm approaches elicited overt avoidance responses. The change in behavior during approach trials was significantly greater during alarm trials F=76.74, df=1, p<0.01). The distance at which manatees responded was significantly greater during alarm trials (F=143.42, df=1, p<0.01). Applying conservative critical ratio estimates for wideband noise, manatees could not detect no‐alarm approaches at distances of 9 m, while manatees responded to alarm approaches up to 35 m away at levels 18 dB above their estimated critical ratios. [Work funded by the U.S. Department of Defence Legacy Resource Management Program, (Navy), Florida Inland Navigation District, and Florida Fish and Wildlife Conservation Commission.]