UC San Diego

The past two centuries have seen an increased exploitation of marine habitats by humans, so a growing appreciation of the role ambient noise plays in cetacean studies has resulted. To achieve a broad acoustical view of understudied areas (namely Mexican waters), this dissertation tackles three overarching principles: (1) parameterizing current baseline ambient acoustic environments for subsequent comparisons, (2) determining whether the sounds that animals introduce into their environments can provide employable information for population estimation purposes, and (3) cataloguing and characterizing the sounds whales use to communicate so they can later be compared across time, geography, and ambient noise levels. Only when we have a handle on the relative contributions of anthropogenic, physical / geological, and biological sounds in a whales habitat can we begin to understand how each source may elicit and / or change their calling behaviors.

Chapter 2 uses an 8-year acoustic dataset from Laguna San Ignacio (LSI), México, to investigate a complex acoustic environment, and to quantify the extent that human-generated noise contributes to this environment relative to natural sound sources. This lagoon has been deemed to be a critical habitat for breeding whales and calving mothers, so vessel traffic is minimized and regulated. I found that humans contribute some noise to the lagoon, but crepuscular snapping shrimp and dusk-centric croaker fish are more intense and pervasive. Therefore, my research validates current management policies for the lagoon and provides a baseline account of a stable acoustic environment that can be compared to future years, should tourist traffic increase.

Chapter 3 develops a model to estimate humpback whale density using ambient noise arising from the songs of many individual animals. What initially began as an analytical model for wind-generated ambient noise, using a collection of randomly distributed sources near the surface, was adapted for randomly-distributed calling whales. The model was tested using data collected in the Los Cabos region of México, which is a breeding ground and part of the migration route for the North Pacific substock. A Generalized Linear Model was used to link humpback-generated ambient noise intensity to concurrent visual surveys. I found that the analytical model provided good predictions of how the intensity of humpback-generated noise varies as a function of acoustic frequency, singer population size, and singer spatial density. In particular, the ambient noise model accurately predicted that singing humpback whales maintain the same separation distance from each other, regardless of singer population size.

Chapter 4 assembles a catalogue of social calls used by humpback whales in the same Los Cabos region and compares them to known social calls from Alaska, Hawaii, and Australia. By using acoustic tags to collect twenty-one samples from three different social group types over two years, I was able to determine the geographical uniqueness, call rate, repertoire diversity, and repertoire entropy of the whales in that breeding ground. I then determined the variability of these behaviors between differently composed social groups. This work provides a starting point for subsequent studies of social calling behavior of Mexican whales, such as temporal changes in social call repertoire, behavioral context of calls between different social groups, and geographical comparisons between the acoustic composition of other humpback whale habitats.

By tackling all three of these acoustic principles, this dissertation aims to set the stage for understanding ambient noise as part of cetacean habitats by laying out how to measure it, how to use it to our advantage, and how to characterize and compare calls within it. With the new novel tools to quantify sound sources in ambient acoustic environments and to monitor population levels that this dissertation provides, my future research can delve into exploring any geographic and temporal commonalities and disparities.