Date of Award
Master of Science in Ocean Engineering
James H. Miller
Networks of high frequency (69kHz) acoustic fish tracking tags and receiver arrays are commonly used by marine biologists and fisheries management programs to determine the spatial and temporal distribution of a marine species. Despite the widespread use of these networks for monitoring fish movement ecology, there is often limited information to inform suitable receiver array spacing for optimal detection probability, or confidence estimates in tag detections. This is further confounded by the effects that seasonal and environmental changes can have on acoustic propagation. To evaluate these effects, the effective range of high frequency acoustic transmitters for medium and large fish (Vemco V16®) was measured in active passive acoustic receiver locations throughout Narragansett Bay under varying environmental and seasonal conditions. The physical factors occurring in the recorded field data were validated using acoustic modeling techniques, showing that receiver array spacing for any location can be determined through simulation if the environmental characteristics of the acoustics soundscape are known, thereby mitigating the need for rigorous and expensive field testing to determine tag detection ranges for this system. Receiver detection data was used in a logistic regression generalized linear model to estimate probability of detection with range and to inform a detection a threshold level at 50% detectability. The detection threshold was set where modeled results reached a level of 8 dB above recorded noise levels and modeled results using Bellhop yielded detection ranges between 688.7 - 878.8 meters for the Narragansett Bay range testing sites. Several hypothetical modeling scenarios which examined the effect of changing individual acoustic environmental characteristics other than noise, revealed that decreased detection ranges were associated with seafloor attenuation characteristics, especially in downward refracting environments (summer thermoclines). Changes in detection range due to varying receiver depth were found to be minimal in shallow water locations (20-60 meters), however they were more substantial in deeper water locations with significant thermoclines (150+ meters). Conversely, model scenarios with tags and receivers located below thermoclines exhibited increased detection ranges across all seasons. In summer months, seafloors comprised of silty sediments had detection ranges far lower than scenarios with a sandy seafloor. When considering receiver array construction, these results reveal that caution should be taken when placing receiver arrays near the sea surface, and that bottom sediment type should also be accounted for, as environments with silty sediments will experience greater loss than those with sandy seafloors. Additionally, receiver array spacing should be adjusted during summer months to account for additional transmission loss due to downward refracting rays and increased bottom losses.
Bonney, Jacob, "CHARACTERIZING THE SEASONAL AND ENVIRONMENTAL EFFECTS ON HIGH FREQUENCY ACOUSTIC FISH TAG DETECTION IN SHALLOW WATER SETTINGS" (2021). Open Access Master's Theses. Paper 2084.