Search

As a group, birds are facing continued declines in conservation status. As such, ex situ management is becoming a more widely used and important tool in conservation programmes. However, adaptation to captivity and behavioural change during ex situ breeding programmes may reduce the success of these interventions. Changes in mating traits, such as vocal behaviour, may be particularly important. In this thesis, I examine the potential for vocal evolution during ex situ conservation to determine how it could impact conservation efforts, focusing on the Java sparrow (Lonchura oryzivora), an endangered estrildid finch. Firstly, I examine the existing literature to determine how vocal behaviour could impact on conservation actions and how potential problems could be mitigated (Chapter 2). Vocal learning and the formation of population dialects have particular significance in conservation programmes due to their effects on important behaviours, such as mate choice. Java sparrows learn a number of components of their song, including song complexity and note acoustic structure (Chapter 3). Birds also learned many temporal features of song, such as tempo, the inheritance of which has been less frequently studied (Chapter 4). Given the importance of vocal learning in song development, there is a high potential for vocal change and population divergence during ex situ breeding in this species. Such differences are important for mate preference; Java sparrow females preferred familiar, over unfamiliar, songs (Chapter 5). Preference for familiar songs may result in assortative mating between individuals of different population origin during conservation efforts, reducing the effectiveness of conservation programmes. As well as behavioural changes, other phenotypic changes may occur in captivity. Captive Java sparrows were both larger and had greater sexual size dimorphism than those of wild origin (Chapter 6). Morphological changes during captive breeding may reduce the performance of captive-bred individuals in wild environments, as well as affecting other, correlated traits, such as vocal behaviour. Vocal behaviour may also be affected by altered sound environments experienced by birds during ex situ breeding programmes. Human presence has a significant effect on the zoo soundscape; reduction in human presence during the COVID-19 lockdown was associated with a number of changes in the sound environment (Chapter 7). Overall, there is significant potential for vocal evolution during the captive breeding of Java sparrows, mediated by cultural processes resulting from social learning of vocalizations, environmental differences, and other, correlated changes. Changes in vocal behaviour in the Java sparrow could affect conservation programmes due to preferences for familiar vocalizations in this species by contributing to assortative mating. The findings from this species suggest that vocal behaviour can have an important role in the success of conservation efforts and requires further investigation in this and other species.

Historically, natural history museums have collected and preserved specimens to provide data on the occurrence and distribution of wildlife populations. Zoologists still track animals by recording footprints, collecting dung and spoor and observing, recording and quantifying behaviour from the ground. However, these traditional observational techniques allow only a few populations to be monitored at once at limited spatial scales and disturbance from the ground can disrupt observation of natural behaviour. We are now in a golden age of technological advances and are able to remotely monitor and track wildlife via a variety of electronic sensors. Significant questions remain about how best to methodologically apply these new technologies for the purposes of wildlife monitoring. In this thesis, I consider challenges of using Earth observation satellites and unmanned aerial vehicles (UAVs) to track wildlife and understand movement in relation to the expanding human footprint and anthropogenic risk. Specifically: (i) I collate and analyse spatially explicit data on the distribution of illegal hunting incidences via a systematic map. I show that hunting increases in proximity to roads, water bodies, and human settlement areas and there is a considerable lack of systematically collected quantitative data. (ii) I investigate acoustic disturbance to understand anthrophony from the species perspective. I create a mitigation method applied in the case of UAV noise using species weighted audiograms (iii) I test whether very high-resolution satellite imagery and machine learning can be used to automate the detection of African elephants in vast heterogeneous landscapes. This is achieved presenting a new method to monitor elephants (iv) I record the spatial relationship of African elephants in relation to the human footprint using GPS tracking data and satellite imagery. I show elephants readily adapt their foraging habits and itineraries, spatially and temporally in relation to human settlement. Accurate and up-to-date data is vital for effective wildlife conservation planning. Remote sensing technologies offer enhanced capabilities to understand the spatial relationship between wildlife and the increasing human footprint. This body of work contributes to the global wildlife conservation effort by devising methods that can enable more reliable data collection at larger spatial scales.

Tropical coral reefs are ecosystems of great beauty and value, but are degrading at unprecedented rates. New approaches are required to monitor and manage reefs worldwide, and to motivate societal changes that reduce the threats facing these vulnerable ecosystems. Healthy coral reefs are alive with sound, as a wide range of soniferous organisms create vibrant soundscapes that reflect the composition of the local community. Many fishes and invertebrates develop as larvae away from reefs before using sound to detect and settle to suitable reef habitat as juveniles, replenishing local populations. In this thesis, I present four field studies that develop and demonstrate the value of bioacoustic approaches to understanding and mediating responses of coral reef ecosystems to degradation. After an introduction to coral reef bioacoustics (Chapter One), I present data showing that soundscapes in the northern Great Barrier Reef have been significantly altered by degradation. Young fishes are less attracted to these post-degradation soundscapes than their pre-degradation equivalents (Chapter Two). Following this, I show that reef recovery can also be measured using acoustic recordings. Previously-damaged reefs restored by rubble stabilisation and coral gardening sound quantifiably similar to nearby healthy reefs and different from nearby degraded reefs (Chapter Three). I then present evidence that loudspeaker playback of healthy reef sound can enhance restoration efforts by increasing fish settlement and retention; fish community development is accelerated by acoustic enrichment of coral-rubble patch reefs compared with unmanipulated controls (Chapter Four). Furthermore, I show that the response of individual species to acoustic enrichment differs in relation to other restoration conditions (Chapter Five). Finally, I discuss the implications of these data for reef monitoring and management, and the use of bioacoustics to inspire public engagement with reef conservation (Chapter Six). In conclusion, bioacoustics offers novel methods to monitor and manage coral reefs, and can motivate people to adopt environmentally-friendly lifestyles. In a world where reefs are changing at unprecedented rates, much can be achieved by learning to listen.

Inexpensive mobile devices offer new capabilities for non-specialist use in the field for the purpose of conservation. This thesis explores the potential for such devices to be used by citizen scientists interacting with bioacoustic data such as birdsong. This thesis describes design research and field evaluation, in collaboration with conservationists and educators, and technological artefacts implemented as mobile applications for interactive educational gaming and creative composition. This thesis considers, from a participant-centric collaborative design approach, conservationists' demand for interactive artefacts to motivate engagement in citizen science through gameful and playful interactions. Drawing on theories of motivation, frequently applied to the study of Human-Computer Interaction (HCI), and on approaches to designing for motivational engagement, this thesis introduces a novel pair of frameworks for the analysis of technological artefacts and for assessing participant engagement with bioacoustic citizen science from both game interaction design and citizen science project participation perspectives. This thesis reviews current theories of playful and gameful interaction developed for collaborative learning, data analysis, and ground-truth development, describes a process for design and analysis of motivational mobile games and toys, and explores the affordances of various game elements and mechanics for engaging participation in bioacoustic citizen science. This thesis proposes research into progressions for scaffolding engagement with citizen science projects where participants interact with data collection and analysis artefacts. The research process includes the development of multiple designs, analyses of which explore the efficacy of game interactions to motivate engagement through interaction progressions, given proposed analysis frameworks. This thesis presents analysed results of experiments examining the usability of, and data-quality from, several prototypes and software artefacts, in both laboratory conditions and the field. This thesis culminates with an assessment of the efficacy of proposed design analysis frameworks, an analysis of designed artefacts, and a discussion of how these designs increase intrinsic and extrinsic motivation for participant engagement and affect resultant bioacoustic citizen science data quantity and quality.

This thesis represents a comprehensive examination of the phonotactic behaviour (i.e. attraction to sound) of the female Gryllus bimaculatus under laboratory conditions. Chapter 2 is the first study to analyze the effect of substrate texture on walking performance in crickets. Substrate texture is found to play an essential role in the phonotactic responses of G. bimaculatus. Smooth substrate texture has a detrimental effect due to slipping, whereas a rough texture results in optimal walking performance due to the friction with the walking legs. Chapter 3 represents the first detailed lifetime study analysing phonotaxis in crickets. My results demonstrate that the optimal age to test phonotaxis in G. bimaculatus females is from day 7 to 24 after the final moult. I also found that selectiveness was persistent with age. These findings contradict the female choosiness hypothesis. This study is also the first to describe the effect of senescence on phonotaxis in insects, as responsiveness decreases with age. Chapter 4 compares the phonotactic behaviour of female crickets from different laboratory-bred colonies. From six tested cricket lab colonies, I found three groups statistically different from each other. Females raised under laboratory conditions at the University of Cambridge and Anglia Ruskin University were most reponsive at a frequency of 4.5 kHz, whereas females bred in Tokushima University in Japan were tuned towards a higher frequency of 5 kHz. These results suggest a degree of artificial allopatric speciation. Comparisons with crickets bred under low-quality conditions in a local pet shop demonstrate a loss of responsiveness, indicating that breeding conditions have a direct effect on phonotactic responsivity. Chapter 5 is the first study to report the presence of phonotaxis in males of G. bimaculatus. Previously it was unknown if G. bimaculatus males were able to perform phonotaxis, given that they were only recognised as endurance signal producers. In the present study, only 20% of the studied males (N=70) performed a weak phonotactic response. This finding has potential ecological implications in terms of male cricket territory establishment, and male-male interactions in the wild, which are discussed. Chapter 6 explores the song pattern recognition of the female G. bimaculatus by changing the duration of either the first, second or third pulse of the chirps. A long first pulse decreased the phonotactic response whereas phonotaxis remained strong when the third pulse was long. Chirps with three pulses of increasing duration of 5, 20 and 50 ms elicited phonotaxis, but the chirps were not attractive when played in reverse order. The data are in agreement with a mechanism in which processing of a sound pulse has an effect on the processing of the subsequent pulse, as outlined in the flow of activity in a delay-line and coincidence-detector circuit.

How can we distinguish individual birds singing in a dawn chorus, and thereby improve measurements of biodiversity to aid conservation efforts? What are the key differences between the walking patterns of an unimpaired individual and a stroke survivor, and how can this inform physical rehabilitation? What is an optimal way to navigate a quadcopter through a complicated environment? This thesis aims to address these questions, and a few others, by linking techniques from optimization, signal processing, and geometry with problems in bioacoustics, biomechanics, and robotics. While the applications are highly interdisciplinary, the methodologies share some fundamental similarities -- all are grounded in physical principles and designed for robust application to real-world datasets. First, in Part 2, we address challenges encountered in acoustic monitoring of wildlife, and develop techniques for separating simultaneous sound sources and finding lower-dimensional structure in birdsong. In Part 3, we draw on geometric principles to study the characteristics of gait kinematics and the mechanisms behind visual perception. Lastly, in Part 4, we design and evaluate algorithms for trajectory optimization of constrained dynamical systems. Overall, we hope this work can be a contribution to both theoretical frontiers and applied sciences.

Although sperm whales are a cosmopolitan species, male and female sperm whales are sexually dimorphic, and the sexes have differences in behavior and habitat preference that result in differences in their distribution and seasonality. Understanding the complex spatiotemporal distribution patterns and demographic composition can be difficult with traditional, logistically challenging shipboard methods given the vast distances and depths these animals travel. Since sperm whales produce highly distinctive echolocation clicks while foraging and navigating, they’re excellent candidates for passive acoustic monitoring (PAM), an alternative method to eavesdrop on these deep-diving animals. Here we show the utility of PAM as a robust tool for advancing our understanding of sperm whale ecology. This study incorporates acoustic data from over 40 recording sites across the northern hemisphere, yielding valuable insights into demographics, acoustic density estimation, and the identification of high-use areas and habitat associations. In remote regions, PAM has enabled us to identify areas where sperm whales are adapting to changing environmental conditions by expanding their potential range in response to climate change, exemplified by observations in the eastern Canadian Arctic. Furthermore, our findings challenge conventional assumptions about male and female preferred habitats, as evidenced by the presence of females in high-latitude regions like the Gulf of Alaska and Bering Sea/Aleutian Islands. Our long-term PAM efforts have significantly expanded our knowledge of demographic specific presence, spatiotemporal distribution, acoustic density, and habitat associations of sperm whales. This underscores the importance of tailored conservation and management strategies that account for demographic variations for effective stewardship of this endangered species.

Blue whales, Balaenoptera musculus, are the largest animals on Earth and yet there is much unknown about their life history. Blue whales produce low frequency sounds including short (1-4 s) down-sweeping D calls, commonly associated with foraging. This study used long term passive acoustic monitoring to investigate a new sequence of D calls observed in data collected from 2007 to 2020 in Southern California and localized sequences on July 22nd 2019. On that day, 28 sequences occurred in a two hour and twenty-minute period. The D calls during this time had an average 7.3 s +/- 1.7 s inter-call interval within a sequence and 7 calls +/- 1 call per sequence. In the data from 2007 to 2020, sequences were present in all the years with a peak in sequences per day in 2019 with the maximum number of sequences per year also occurring in 2019. To understand how sequences were produced spatially, the D calls on July 22nd 2019 were localized. This resulted in 194 locations for all D calls and an average location for each sequence. Average locations were scattered within a 250 m by 350 m area, indicating that an individual whale may be producing the sequences. Using the location for each call, source level was computed using received level and transmission loss. The average root mean square source level calculated over 30 to 80 Hz was 162.3 +/- 2.2 dB re 1µPa at 1 m. By combining both observations of calling behavior and localizations, we can gather insight into the temporal and spatial behavior of blue whales when emitting D calls.

Abstract: Bioacoustic analyses have been used for a variety of purposes including classifying vocalizations for biodiversity monitoring and understanding mechanisms of cognitive processes. A wide range of statistical methods, including various automated methods, have been used to successfully classify vocalizations based on species, sex, geography, and individual. Here, I used several classification techniques, namely Discriminant Function Analyses (DFAs), Support Vector Machines (SVMs), and Artificial Neural Networks (ANNs), for sex-based classification of zebra finch (Taeniopygia guttata) distance calls using acoustic features measured from spectrograms. We found that all three methods (DFAs, SVM, and ANNs) correctly classified the calls to respective sex-based categories with high accuracy between 92% and 96%. Frequency modulation of ascending frequency, total duration, and end frequency of the distance call were the most predictive features underlying this classification in all our models. My results presented here agree with previous results and identified highly convergent acoustic features involved in this classification. My comparison of classification methods gives researcher much needed information to select an optimal classification method. Refining and modifying experimental procedures play a vital role in improving methodology while also reducing animal distress. Standardized experimental procedures are often modified for obtaining optimal responses. Many automated operant conditioning procedures have been developed and modified over time. Here, I asked if an increase in feed time duration affects discrimination in an operant go/no-go task. Specifically, we used zebra finches’ sexually dimorphic distance calls as acoustic stimuli to test whether there were any significant differences in performance on an operant discrimination task requiring zebra finches iii to classify calls according to the sex of the producer when a key experimental parameter, feed time duration, was increased from 1 second to 2 seconds. We found no significant differences in learning speed (trials to criterion) between birds that were given 1 sec or 2 sec of food access following a correct go response. Our results indicate doubling food access duration did not impact the speed of acquisition of distance call discrimination in zebra finches. These findings suggest that we can provide twice as much time for zebra finches to access food, potentially improving animal welfare, whilst still being able to compare new results with historical results.

Abstract: The culture of ecology is shifting towards collaborative, integrative approaches that use ‘big data’ to solve big problems. Passive acoustic monitoring (PAM) has the potential to play a role in this new paradigm because it uses in-situ autonomous recording units (ARUs) to collect a permanent archive of audio recordings. PAM research groups across the world are collecting vast amounts of acoustic data that could be integrated to understand ecological phenomena at a global scale; however, there are several hurdles that must be overcome. First, accurate algorithms that automatically scan acoustic recordings (hereafter, “recognizers”) are required to efficiently determine the species detected within the recorded soundscapes. Second, an understanding of the context of these recognizer-processed datasets is essential for data integration and can improve how recognizer data is used in ecology. Unlike survey data collected by human observers, recognizer data is typically treated as a binomial dataset with minimal context beyond date and time of observations. In this thesis, I demonstrate the importance of spatial and behavioural context of recognizer data using the common nighthawk (Chordeiles minor) as a model species. The common nighthawk is a crepuscular and highly mobile bird species that consumes aerial insects and is declining in most parts of its breeding range across North America. First, I demonstrate a fundamental principle that is crucial to using and integrating recognizer data. I show that the classification probability of species detections reported by a recognizer is related to detection distance, and that the classification threshold applied thus defines the survey area. Understanding this spatial context is necessary for estimating density from recognizer data and for integrating multiple datasets. Next, I show that the behavioural context of recognizer data can provide important insight into ecological analyses. I use VHF telemetry to show that the wing-boom display of the common nighthawk is a territorial signal, which I then use to study behaviour-specific habitat use of this species in the boreal forest. I use the wing-boom to separate territorial from home range behaviour and show that the scale that most strongly predicts habitat use corresponds to the movement range of that behaviour. I then use the wing-boom to separate territorial from extraterritorial behaviour to confirm that the common nighthawk is a ‘disturbance specialist’ species in the boreal forest, but only for nesting territories. Finally, I combine these spatial and behavioural contexts to demonstrate a novel method for density estimation that can be applied to single ARUs at broad spatial scales. The goal of this approach is to improve regional, national, or range-wide population estimates, especially for regions that are poorly covered by human surveys or for species that have large home ranges. Together, this density modelling approach and the spatial principles my thesis presents will facilitate future integration of PAM datasets collected with varying methodologies as well as with other data types. The behavioural context component of my thesis encourages PAM users to ensure they put their recognizer data into appropriate ecological context, particularly for wide-ranging species. Collectively, the ecological inferences in my thesis provide a major advance in understanding common nighthawk ecology in the boreal forest and the tools developed will help future research and conservation of this enigmatic species.

Abstract: Habitat alteration is the leading cause of bird population declines globally. Therefore, understanding the processes influencing habitat selection are important for the identification and protection of important areas for birds. Long-distance migrant birds are particularly vulnerable and are experiencing disproportionate declines relative to other avian groups. Preventing further declines requires a more thorough understanding of how quality habitat is assessed by birds and new measures for determining the potential productivity of a territory. Differences in territory settlement date (arrival) have been shown to reflect habitat quality on small spatial scales, thus differential arrival timing of migrant birds may help us address fundamental questions of habitat selection across landscapes. Despite widespread acceptance of arrival time as an indicator of habitat quality, a large-scale multi-species demonstration of this phenomenon is absent due the logistical difficulty of collecting the appropriate data. I investigated the potential value of habitat assessment through measured differences in relative migrant arrival date estimated from provincial scale long term bioacoustic monitoring. I first assess the descriptive ability of two different operational arrival definitions used previously to justify the arrival date of three boreal migrants (first detection and first instance of three-day consecutive detection) and an additional novel definition (the detection-gap definition). The descriptive ability of each was assessed through the strength of an assumed positive latitudinal relationship. I found that all definitions performed similarly, producing the anticipated latitudinal relationship in 8 of 9 species-arrival definition combinations. Therefore, I adopted the first detection definition to describe arrival in this system as it provides similar descriptive ability with reduced sampling effort. I then address large-scale settlement patterns of three migrants: Ovenbirds (Seiurus aurocapilla), Tennessee Warblers (Leiothlypis peregrina), and Yellow-rumped Warblers (Setophaga coronata) across the boreal regions of Alberta. I also examined the local settlement patterns of Ovenbird by comparing both relative arrival timing and measured density changes between nearby territories. Using predicted density as a habitat quality proxy, I found increasingly early arrivals in higher quality sites both provincially and locally for species where habitat specificity has been reported. I also found that sites where higher densities of Ovenbird are predicted to occur locally are filled to higher densities before sites of lower density. Finally, I found that additional life history characteristics including habitat associations and migration distance may be measurable through arrival timing comparison. In this study, I present the first landscape-scale examination of simultaneous multi-species migrant arrival time using bioacoustics. By correlating migrant arrival time and density, I demonstrate that density is a reasonable measure of habitat quality. Combined, density and arrival data provide a relatively low-cost way of assessing habitat quality and should be used to inform land management decisions in the boreal forest region.

Translocations of kiwi (Apteryx spp.) are one of the most common and growing types of conservation translocations in New Zealand. However, their outcomes remain mostly unpublished, which does not allow for sharing of lessons learnt from past developments. We reviewed 102 kiwi translocations from the 19th century until 2018 and identified factors affecting their outcome. North Island brown kiwi (A. mantelli) was the most translocated species, but the highest impact of translocations on the improvement of conservation status was for the rarest taxa: little spotted kiwi (A. owenii), rowi (A. rowi), and Haast tokoeka (A. australis ‘Haast’). Translocations are typically used for creating secure populations and, more recently, for ecosystem restoration and meta-population management. We developed a set of criteria to evaluate the outcome of introductions and reintroductions based on demographic parameters alongside current recommendations on the genetic make-up of translocated populations. Based on these criteria, only a few translocated populations can be considered successful in the medium–long term: 15+ years following the release of a genetically diverse population (40+ unrelated individuals). Most historical translocations failed or require further genetic and habitat management. However, the majority of kiwi translocations have occurred over the last two decades and, while several populations have successfully established, for most of them, it is too soon to assess their medium-long term outcome. An analysis of factors affecting translocation outcomes revealed that, despite ongoing predator control, populations at small, unfenced sites on the mainland suffer from dispersal and predation, which has negative demographic and genetic consequences. Releases to larger mainland sites and predator-free areas have increased survival times, indicating higher chances for a positive translocation outcome. Moreover, translocated wild-caught and captive-sourced birds survived longer than birds from the Operation Nest Egg (ONE) programme, particularly at sites that were not predator-free. We highlight the need for genetic considerations in the planning and adaptive management of proposed and existing translocated populations. Specifically, we suggest that differences in kiwi survival, based on the type of released birds and release site’s area size and predator status, should be considered during translocation planning. The kiwi translocation review identified significant inconsistencies and often insufficiency of post-translocation monitoring. We demonstrate the utility of post-translocation monitoring methods in a recent translocation case study: a reintroduction of roroa–great spotted kiwi (A. maxima) in the Nina Valley, Lake Summer Forest Park. In 2015, eight wild-caught adults were translocated from the Hawdon Valley, Arthur’s Pass National Park, following the release of ten ONE subadults between 2011–13. We tracked the translocated kiwi by radio telemetry between 2015–17 to monitor post-release survival, dispersal, and ranging behaviour. Dispersal was highly variable among the released wild birds. The straight-line distance from the release site to the last recorded location ranged between 0.5–10.3 km. Based on the dynamic Brownian bridge movement model, seven of the wild birds survived, remained in the Nina Valley, and covered up to 1700 ha (95% utilisation distribution). Releasing the wild birds had no measurable impact on the ranging behaviour of previously released subadults. Additionally, we used occupancy modelling to analyse passive acoustic monitoring data (PAM) from the Nina and Hawdon valleys to monitor changes in distribution and growth of the translocated population and the impacts of the translocation for the source population. We analysed data from two survey years 2012–13 and 2017–18, being before-and-after the 2015 translocation. Occupancy estimates increased significantly at both study areas, despite the translocation of approximately 20% of known territorial adults (four pairs) from the Hawdon to the Nina. Moreover, at least three out of four territories, where adult birds were removed, were re-occupied by new pairs within 2.5 years. Site occupancy increased in the Nina from 0.20 (SE 0.10) to 0.72 (0.10), and in the Hawdon from 0.63 (0.10) to 0.95 (0.04). Detectability varied significantly between study areas and was influenced by the length of survey night, breeding/non-breeding season, and wind speed. The differences between the naïve and estimated occupancy values underscore the benefits of occupancy modelling for measuring response to conservation management. This study demonstrates the utility of PAM in monitoring translocation outcomes: tracking changes in occupancy and local distribution and assessing impacts on the source population following the birds’ removal for translocation.

Acoustic imaging systems such as biomedical devices and sonar play an important role in situations where optical systems cannot be used due to limited visibility (e.g. in living tissue or in turbid water in the ocean). Despite the wide variety of these systems, a major challenge for state-of-the-art devices is how to increase the image resolution. There are two ways to address this challenge. First, the resolution could be enhanced by increasing the number of transducer elements on the phased array of sensors. However, most imaging systems have a centralized architecture in which synchronizing the transducers complicates the hardware and increases costs significantly. This limits the number of transducer elements that can be recorded simultaneously. Second, the resolution could be increased by judiciously processing the measured acoustic field. Comparisons between artificial and biological systems (e.g. human-made sonar versus biosonar) reveal that the former could be improved significantly but how to do so is still an open question. In this work, we propose novel strategies to improve acoustic imaging systems using both metamaterial-based and bio-inspired approaches. First we propose the concept of out-of-band phase conjugation to enable imaging/communication systems with excellent performance inside complex media. In this method, a decentralized metasurface composed of arrays of non-synchronized sensing and source transducers responds locally to the received field and generates higher harmonic waves that propagate back to the original source. We demonstrate the concept in electromagnetics and acoustics where second harmonic waves were generated from the original wave. Due to the high frequency nature of the outgoing waves, excellent self-focusing properties with enhanced spatial resolutions can be achieved in dynamically varying inhomogeneous media. We further investigate this novel benefit using a metasurface that down-converts the wavelength by five orders of magnitude using sound-to-light transduction with an acousto-optical metasurface (AOM). In this work, we study the physics that drive this AOM imaging, and demonstrate that this method achieves promising image quality. Unlike most conventional acoustic imaging devices, our proposed AOM operates in a totally decentralized manner and thus the complexity scales linearly with the number of transducer elements in contrast with conventional synchronized systems where complexity scales polynomially. This advantage will enable devices with more sensors at a low level of complexity, resulting in images with a significantly improved spatial resolution. Systems that exploit the non-synchronized down-conversion method described above rely on large sensor arrays. However, there are applications where an image may need to be formed using measurements from a limited number of sensors. To gain insight into how to effectively use information from a limited number of sensors, we examined a biological system, the bottlenose dolphin. Acoustic data are used by these animals for echolocation, localization, and communication in the marine environment. In this work, we studied the echolocating bottlenose dolphins in an experimental setting. A physics-based model of the acoustic environment, experimental data collected from the animal, and a machine learning-based model were used to investigate how dolphins discriminate targets using echolocation. These models were validated using both experiments and numerical simulations in a case study with three scenarios. The proposed framework provides insight into key acoustic features from the echoes that may be used for target identification and indicates that adaptive positioning is used by the animals to modulate the acoustic information needed for target discrimination.

Abstract: Accurate assessment of how animals distribute themselves across the landscape is an essential component of ecological research. Ecologists often conduct surveys to subsample a representative portion of an area of interest and extrapolate their findings to a larger region. Acoustic surveys are frequently used to study a variety of organisms including birds, which rely on acoustic communication and vocalize regularly. The development of autonomous recording units (ARUs) has greatly increased the popularity of acoustic surveys, due to the ability to record surveys and leave recording equipment unattended for extended periods of time. However, variation in the detectability of wildlife is difficult to quantify, and can be strongly influenced by the focal species, the surrounding environment, and method of recording and processing data. Furthermore, count data from ARUs are typically considered to be indices of abundance, and obtaining density requires significantly more effort and data. My objectives were to investigate and quantify variables that influence detectability, determine how variation in detectability influences estimates of relative abundance, describe methods to address and correct for variation in detectability, and develop methods to efficiently estimate animal density using ARUs. I found that detectability is greatly influenced by the surrounding environment and that detectability greatly differs between open and closed vegetation types. Ignoring variation in detectability can lead to estimates of relative abundance in open vegetation types that are more than double that of closed vegetation types, due to increases in the area sampled by ARUs. Furthermore, detectability is influenced by the type of recorder used, and while sometimes equivalent to the detectability of human observers, standardization of detectability is required to properly integrate data from multiple sources. Finally, I presented methods to predict correction factors for different species and vegetation types based on the generalized sound characteristics of their vocalizations or songs. These correction factors should offset differences in detectability between different environmental conditions without the need for calibration studies. I also present a new method for estimating animal density from ARUs by estimating the distance of a vocalizing individual from the sound level measured on an audio recording. I show that these estimates are more accurate and objective than distance estimation in human observers and can provide estimates of animal density through conventional distance sampling. Integrating and standardizing data from ARUs and other sources is important when looking at large-scale ecological patterns. My research contributes to the growing number of studies that investigate variation in detectability of acoustic signals in different environments and the implication on wildlife surveys. I also provide innovative methods which should increase the versatility, quality and accuracy of data obtained from audio recordings and allow ecologists to integrate data from multiple sources to answer questions at different ecological scales.

Abstract: The Yellow Rail (Coturnicops noveboracensis) is a small, secretive, wetland bird, which is apparently rare throughout most of its range. Almost nothing is known about its abundance and density in the wetlands of the western boreal forest. Emerging technologies have enabled us to effectively survey for Yellow Rail in remote wetlands by using ground-based remote sensors (autonomous recording units; ARUs) to conduct passive acoustic monitoring. This technique was employed to survey Yellow Rail populations across two large study areas: one in the taiga plains of the Northwest Territories, and the other in the boreal plains of Alberta, Canada. For the Edéhzhíe Indigenous Protected Area (NWT), a predictive map of Yellow Rail density was developed based on data obtained from a systematic avian survey conducted in 2016, using 205 ARUs. Counts of Yellow Rail were converted to density estimates using habitat specific effective detection radii obtained via call-playback experiments. Generalized linear models and covariates from a detailed landcover classification effort were used to develop the spatial model. Yellow Rail appeared to breed at relatively high densities (0.07 males/ha compared to average densities of 0.04-0.05 males/ha) in Edéhzhíe and they were strongly associated with marsh wetlands. The Mills Lake wetland complex was identified as an important breeding area for Yellow Rail in the Northwest Territories based on a population estimate of ca. 560 breeding pairs. For the Alberta Oilsands Region, a predictive map of Yellow Rail breeding abundance was developed using acoustic data compiled from the first five years (2013-2017) of an ongoing bioacoustic monitoring program. Recent developments in open-access satellite data, cloud computing (Google Earth Engine), and data science were leveraged to secure large-scale, high-resolution (10 m) landcover data. Multiple satellite remote sensors were used to derive fifteen predictor variables: Sentinel-1 synthetic aperture radar, Sentinel-2 optical imagery, and Advanced Land Observation Satellite digital elevation maps. Gradient boosted regression was used to develop the spatial model. Six remote sensing predictors (DPOL, ΔVH, REIP, ARI, VH, and SWI), were identified as having strong predictive capacity. Several predictors had complex non-linear responses and multiple important interactions were identified. Approximately 1.5% of available wetland habitat in the region was predicted to be highly suitable for Yellow Rail.

In the context of reproductive resilience, the dynamics of phenotypically plastic spawning behaviors are shaped by the environmental and anthropogenic stressors that a species is exposed to. Spawning locations, timing, and periodicity are three characteristics that are progressively being recognized as important aspects of productivity and resilience in fish species. The chapters of this dissertation examine those factors as they are filtered through the ecological contexts of the environment, life history traits and fisheries exploitation. The following research covers spatio-temporal scales from a species and site-specific study of spawning behavior to a multi-species, regional study of spawning behavior and the interactions with fisheries. Chapters two and three illustrate how reproductive resilience is realized in species that spawn frequently across spatially diverse locations. These reproductive traits have evolved to maintain reproductive success and population stability in response to the unpredictable and highly variable environments the species inhabit. Chapter four investigates the use of passive acoustic techniques to locate and monitor spawning in species that produce sound during the spawning process. That study seeks to resolve the spatial resolution of passive acoustic sampling and monitoring in shallow estuarine environments. Chapter five provides further support for including spawning behaviors in the study of productivity and resilience in fish populations by showing that species that aggregate to spawn, along with large changes in density, for short and predictable time periods, are the most vulnerable to targeted fishing and are more likely to have been historically overfished. This work builds on the concept and framework of reproductive resilience in fish and provides both theoretical support and empirical evidence for further consideration of the spatial and temporal components of spawning in stock assessments, climate variability assessments and the conservation of fish populations

Abstract: Remote camera traps are often applied to large mammal conservation and management programs because they are cost-effective, allow for repeat surveys, and can be deployed for long time periods. Additionally, statistical advancements in calculating population metrics, such as density, from camera trap data has increased the popularity of camera usage in mammal studies. However, drawbacks to camera traps include their limited spatial coverage and tendency for animals to notice the devices. In this study, we compared autonomous recording units (ARUs) to cameras in their detectability of gray wolves (Canis lupus) through a paired study design in northeastern Alberta. The use of ARUs to survey for large, low-density predators, like wolves, is just now emerging as a viable passive monitoring method, but to our knowledge, a comparison of ARU and camera detectability for wolves has never been done. We also tested the random encounter and staying time model (REST), a new means of estimating the density of an unmarked population, using human volunteers and simulated camera surveys. We found ARUs to be comparable in their detectability of wolves to cameras, despite only operating a fraction of the time that cameras were active. We also found the REST method to produce unbiased estimates of density, regardless of changes in human abundance, movement rates, home range sizes, or simulated camera effort. These advances in surveying technology and statistical methods provide innovative avenues of large mammal monitoring that have the potential to be applied to a broad spectrum of conservation and management studies, provided assumptions for these methods are rigorously tested and met.

Soniferous reef animals produce sound through intentional behaviors like communication and as byproducts of activities like feeding and defending. Sounds from individuals collectively create habitat-specific and temporally variable biological soundscapes. As sounds propagate away from their sources, they may influence larval fishes that use acoustic cues during settlement. Therefore, animal behavior creates spatiotemporally-variable soundscapes, which in turn influence the behavior of fishes during a critical life history transition. These different scales of marine bioacoustics were investigated. The soundscapes across a range of reef quality were recorded for six weeks in an anthropogenically-degraded region of Caribbean Panama. The same contributors to the soundscapes were observed across sites, with taxa-specific variation in spatiotemporal patterns of acoustic behavior. The observed rate of these potential cues was predicted by an individual-based model to improve larval settlement, suggesting resiliency of some components of these acoustic cuescapes. The toadfish Ampicthyses cryptocentrus was a dominant contributor to the soundscapes and showed site-specific temporal variation. Analyses of recordings of ten focal males and males in neighboring burrows demonstrated that males were responsive to multiple neighbors, not just the perceived loudest competitor, and a null model supported that grunts overlapping neighbors’ calls were used as acoustic competition. These results suggest the contribution of toadfish to the soundscape is dependent on both abundance and spatial distribution. Calibrated propagation modelling was used to predict the acoustic fields created by reef-based sounds. It was found that spatial heterogeneity in cue strength increases with frequency and changes with source location, with implications for cue detection. Detection of reef-based cues is enhanced when the otoliths receive energy from the bladder’s response to pressure fluctuations. Changes in bladder size and otolith-bladder distances are likely to create ontogenetic changes in pressure sensitivity, which was tested with a combination of micro-computed tomography and finite-element modelling using larval Sciaenops ocellatus. Otolith-bladder distances increased with fish size, reducing predicted pressure sensitivity. Bladder volume also increased with growth, compensating for increasing distances by partially recovering predicted pressure sensitivity. Combined, these results highlight the system’s complexity: social behavior, spatiotemporal variation, and ontogenetic patterns dynamically influence how soundscapes influence larval behavior

Killer whales produce pulsed calls, which are used for communication. Calls are highly stereotyped and repertoires are unique to individual pods. Discrimination amongst these calls and comparison of call repertoires between pods can help determine population structure in killer whales and can be used to track pod movements. Calls were detected in underwater acoustic recordings in August and September 2017 in the Arctic waters of Eclipse Sound, in Nunavut, Canada. We present a repertoire of killer whale calls recorded. Eleven stereotypic call types, three biphonic and eight monophonic, were identified using manual call organization and manual whistle contour extraction. A higher diversity of calls was detected in the hydrophone located in the known narwhal aggregation site in Milne Inlet, than at the second hydrophone deployed at the mouth of Eclipse Sound which is the proposed entrance and exit point for the killer whales. The potential for increased killer whale presence and magnitude of predation on narwhals is a source of concern for management of the population and by Inuit subsistence hunters who rely on narwhals for food and economic benefit. Describing the acoustic repertoire of killer whales seasonally present in the Canadian Arctic may help understand their behavior and seasonal movements. The results presented may provide a basis for future acoustic comparisons across the North Atlantic and aid in characterizing killer whale ecotypes making seasonal incursions into Arctic waters.

Abstract: Anthropogenic disturbance is known to have negative population consequences and alter animal behaviour, and a growing body of research on the effects of anthropogenic noise is finding similar negative impacts on wildlife. Noise in natural environments can mask important acoustic signals used for animal communication. Owls use vocal communication to attract mates and defend territories, and rely on acoustic cues to locate their prey. Industrial noise has been shown to negatively affect owl hunting success and reduce foraging efficiency by affecting their ability to detect prey, but whether this results in reduced habitat suitability for owls in areas near industrial noise sources is largely unknown. It is also not known if the abundance of small mammals, the primary prey of owls, is affected by industrial noise. Communication between small mammals is unlikely to be directly affected by noise because they primarily communicate using olfactory rather than vocal signals, but small mammals could be indirectly affected through altered predation dynamics. I sought to determine if owls avoid the areas surrounding chronic industrial noise sources, if prey availability is affected by chronic industrial noise, and the relative importance of noise compared to other types of disturbance resulting from industrial development on owl habitat use. I used autonomous recording units and automated recognizers to survey for owls and scan recordings for owl vocalizations, to assess the impacts of disturbance from industrial development on owls in northeastern Alberta, a region that has seen increased development in the oil and gas industries in recent years. I found that barred owls (Strix varia), great horned owls (Bubo virginianus), and boreal owls (Aegolius funereus) were equally likely to occupy noisy sites compared to sites with no noise, indicating that site-level occupancy (representing a home range scale) was unaffected by the presence of noise sources on the landscape. I found no difference in abundance or activity of red-backed voles (Myodes gapperi) and deer mice (Peromyscus maniculatus), indicating that these important prey species are not strongly affected by noise. Finally, I found each owl species responded differently to the different disturbance types. Barred owls were less likely to be present in areas with greater proportions of human footprint and roads, whereas great horned owls were more tolerant to disturbance and were more likely to be present in areas with greater proportions of soft linear features (e.g. seismic lines, pipelines), though they did avoid areas with large industrial facilities. Boreal owl presence was more strongly affected by forest composition than disturbance; they prefer more coniferous forests, and it was less clear if they were sensitive or relatively tolerant to disturbance. Assessing the relative impacts of multiple types of disturbance and how the accumulation of disturbances can impact wildlife is important in understanding species declines and can help focus conservation efforts. My thesis contributes to research on the impacts of anthropogenic disturbance and suggests the effect of noise on barred owls, great horned owls, and boreal owls is minimal at the spatial scales I looked at. Similarly, the effect of noise on abundance and activity of red-backed voles and deer mice was weak, suggesting similar prey availability for owls. For the owl species I studied, forest composition and other types of disturbance on the landscape are more likely to influence their habitat use.

Abstract: For male breeding songbirds, song rate varies throughout the breeding season and tends to be correlated with breeding-cycle stages. Although these patterns have been well documented, to our knowledge, this relationship has not been used to predict a bird’s breeding status through acoustic monitoring. The first objective of this study was to determine if variation in song rate can be used to predict the breeding status of the Olive-sided Flycatcher (Contopus cooperi; OSFL), a Species at Risk in Canada. In 2016, song rates from 27 male OSFLs in Alberta and the Northwest Territories were collected from human observers (n = 454 5-min counts), and breeding status (i.e. single, paired, and feeding young) was monitored throughout the breeding season. I evaluated the predictive ability of three modeling approaches (i.e. regression, hierarchical, and machine learning) using model sensitivity and specificity. The hierarchical model was the best at predicting all three breeding statuses, with 69%, 50% and 87% sensitivities and 80%, 82% and 78% specificities for predicting single, paired, and feeding young, respectively. This resulted in a mean sensitivity of 69%, compared with 54% and 50% from the regression and machine learning models, respectively. A second objective was to use the hierarchical modelling framework to predict breeding status from song rates collected by Autonomous Recording Units (ARUs) processed using automatic recognition software. For 24 of these OSFLs, I collected 4,302 5-min song counts and used daily song rate to compare the relationship of rates and breeding status as determined by ARUs versus human-observers. We then tested four hierarchical models accounting for imperfect detection. Song rates derived from ARU data followed a similar pattern to that of human-observer song rates, where single males had the higher rates, paired males had lower rates, and those feeding young had lowest rates, but the absolute values for rates were much lower with ARUs. All ARU data predictive models performed poorly at predicting single (sensitivity range 0 – 7%) and well at predicting paired (sensitivity range 77 – 84%). The ARU models had mixed success at predicting feeding young (sensitivity range of 25 – 68%) but adjusting for imperfect detection did not improve model sensitivity to predict any breeding statuses. Low predictive ability was likely due to the low detectability of ARUs (e.g. bird movement out of detection range of ARU) and the automatic recognition software we used. Considering the high predictive ability of models using human-observer data and that the challenges currently associated with our acoustic processing methods can be addressed, I recommend that the breeding status of forest birds should be monitored using acoustic data. I provided a hierarchical modelling framework than can be applied to other species and improved to account for bird movement or number of conspecifics. This novel approach could provide a cost-effective tool to infer much needed demographic information over large spatial extents, and inform species status assessments, recovery strategies, and management plans for many species of conservation interest.

Mysticetes (baleen whales) often make long, annual migrations from high latitude summer feeding areas to low latitude wintering areas. Eastern North Pacific gray whales (Eschrichtius robustus) migrate within a few kilometers from shore for most of their route from summer feeding areas in the Bering, Chukchi, and Beaufort Seas to wintering areas in the lagoons along the south-western coast of the Baja California Peninsula in Mexico. This dissertation combines passive acoustic recordings, infrared camera video, and visual sightings to investigate gray whale behavior and how it changes across different timescales. I use a four-element hydrophone array in central California to present the first published full-season acoustic monitoring and tracking of migrating gray whales. I describe the characteristics of calls produced by migrating gray whales and analyze how these characteristics change due to propagation. I show that gray whale behavior changes on diel and seasonal timescales. Notably, gray whales increase their vocalizations at night but their mean swimming behavior does not change, contrary to previous assumptions used in population size estimates. Over seasonal timescales, vocalizing gray whale swimming behavior aligns with previous observations. I explore how passive acoustic and infrared camera monitoring can help quantify whales by calculating cue rates or call and blow rates for migrating gray whales. Acoustic calling rates indicate that the gray whale population size is greater than estimated using visual sightings alone and that calling rate increases over the southbound migration. Infrared camera blow rates are less affected by whale behavior and are useful for daytime and nighttime monitoring, but are limited by visibility and distance. To understand gray whale behavior over seven migration seasons, I use visual daily counts at two sites and single-hydrophone call detections which indicate that migratory behavior seems to be driven more by intrinsic than the tested environmental factors. I find that the proportion of the population using a coastal route through the Southern California Bight, especially past Los Angeles, increased over these years. Understanding the behavior of migrating gray whales will help improve abundance estimates and determine how these whales may be impacted by nearshore anthropogenic activities and climate change.

Passive acoustic monitoring is a transformative tool for increasing the scale of ecological monitoring. The biggest challenges remain in analyzing large volumes of recordings to produce ecological information useful for decision making. Ecoacoustics, a holistic, quantitative analysis of soundscapes and their emergent properties is an appealing way to distill information from many recordings, but new tools require rigorous studies to measure their efficacy tracking ecological patterns. This thesis contains three studies testing the efficacy of ecoacoustic approaches to monitoring species and communities. The first study uses simulations to test if a variety of indices of soundscape complexity, intensity and differences can measure realistic bird community changes in simulated dawn choruses. In addition, I explore combining indices to increase predictive power, and how the addition of anthropogenic noise and changes in bird behavior impact the usefulness of acoustic indices. The second study examines the acoustic monitoring of Cassin’s Auklets (Ptychoramphus aleuticus), a chorusing colonial seabird that is notoriously challenging to monitor. I test how a traditional automated call detection compares to a novel approach based on energy spectrums of the entire soundscape to index auklet activity. I discuss how for some species, this approach may be more effective in indexing activity, and in this case is a reliable predictor or relative burrow density, and could be applicable for other nocturnal, vociferous seabirds. In the third study I move from single seabird species to seabird communities. I apply acoustic indices of soundscape differences, complexity and intensity to measure seabird restoration outcomes in the Western Aleutian Islands. I find that acoustic richness, and the spectral differences from a pristine reference island are predicted by years of seabird recovery and other factors that promote seabird recovery. Acoustic indices provided similar information to traditional and more laborious approaches. In summation, ecoacoustic approaches are a valuable part of the acoustic monitoring toolbox. Across all three studies, acoustic indices reflected meaningful ecological patterns, at a fraction of the time and effort needed for other approaches. This can greatly improve the scale and speed at which monitoring programs can inform management decisions and iteratively improve conservation outcomes.

This project started early in the summer of 2016 when it became evident there was a need for an effective and efficient signal analysis toolkit for the Littoral Acoustic Demonstration Center Gulf Ecological Monitoring and Modeling (LADC-GEMM) Research Consortium. LADC-GEMM collected underwater acoustic data in the northern Gulf of Mexico during the summer of 2015 using Environmental Acoustic Recording Systems (EARS) buoys. Much of the visualization of data was handled through short scripts and executed through terminal commands, each time requiring the data to be loaded into memory and parameters to be fed through arguments. The vision was to develop a graphical user interface (GUI) that would increase the productivity of manual signal analysis. It has been expanded to make several calculations autonomously for cataloging and meta data storage of whale clicks. Over the last year and a half, a working prototype has been developed with MathWorks matrix laboratory (MATLAB), an integrated development environment (IDE). The prototype is now very modular and can accept new tools relatively quickly when development is completed. The program has been named Banshee, as the mythical creatures are known to “wail”. This paper outlines the functionality of the GUI, explains the benefits of frequency analysis, the physical models that facilitate these analytics, and the mathematics performed to achieve these models.

Abstract: Habitat loss and fragmentation are thought to be leading causes behind local to global amphibian declines. Recent expansion of the energy sector in the western boreal forest of northeastern Alberta raises concerns for amphibians that occupy this landscape. Methods for monitoring anuran amphibians typically involve call surveys to detect vocalizing males during the breeding seasons. I used passive acoustic monitoring with autonomous recording units (ARUs) to maximize detections of uncommon and common species. ARUs were deployed in the Lower Athabasca Planning Region (LAPR) of northeastern Alberta during the months of April to August (2012 – 2016). ARUs collected acoustic recordings of the soundscape that were processed through human listening of raw recordings and through use of an automated recognizer to identify the unique vocalizations of the Canadian toad (Anaxyrus hemiophrys, Cope 1886) and boreal chorus frog (Pseudacris maculata, Agassiz 1850). I generated a predictive generalized linear model using logistic regression to identify breeding habitat used by the Canadian toad in northeastern Alberta. I did this using a combination of remote sensing from ARUs and landscape characteristics collected from satellite and LIDAR technology. I found that the presence of this species is associated with fen wetland and upland edge habitat, along with the occurrence of coarser soils that allow toads to dig below the frost line during winter months. A positive relationship between presence and the proximity of roads was seen in the model, indicating that suitable habitat for toads exists near roads. This model had good predictive capacity and accuracy, and would be an effective tool for locating additional breeding locations for this species of concern. I also used ARUs to understand if boreal chorus frogs show a negative response to roads and associated noise pollution, a major cause of habitat degradation. I used call intensity codes as a measure of the relative abundance of chorus frogs and evaluated if road metrics, calculated with spatial data layers, influenced relative abundance. Boreal chorus frog abundance was shown to be positively influenced by road proximity, suggesting that the species breeds in roadside ditches. I further examined acoustic interference by passing vehicles, by measuring the full chorus amplitude (dB) for boreal chorus frogs before, during and after a vehicle passage. No significant difference in chorus amplitude could be detected, indicating that this species is not affected by traffic noise. It seems this anuran is not affected by roads and human activity on roads in this system. In conclusion, I provide an argument for the use of passive acoustic monitoring for surveying temperate anuran amphibians. ARU detections greatly increased the sample size of sites and dates surveyed, as well as expanding the ecological questions that I could address. Future monitoring and conservation plans for anurans in the LAPR are encouraged to use the findings of this research and to continue to use and develop methods for acoustic sampling.

Abstract: Limited information exists on the recovery of different ecosystem components following reclamation of oil and gas wellsites in the boreal forest of Alberta. Songbird response to wellsite reclamation efforts in the boreal forest was previously unexamined, despite the abundance of wellsites, frequent use of songbirds to assess ecosystem state, and importance of the boreal forest as breeding habitat for songbirds. Determining local scale impacts of small disturbances characteristic of energy sector on songbirds in the boreal forest, and how these impacts change with regeneration requires spatially accurate data on use or avoidance of these features. Conventional methods for surveying songbirds in the boreal vary in their ability to provide these spatially accurate data. Many bird surveys now utilize bioacoustic approaches. Standard approaches to collect biacoustic data do not overcome challenges associated with conventional methods for accurate estimation of bird singing locations. However, certain bioacoustic approaches, including the use of an acoustic location system have potential for collecting data with the spatial accuracy to determine where songbirds sing in relation to small boreal disturbances. The purpose of this thesis was to use an acoustic location system to determine how Ovenbirds (Seiurus aurocapilla), and songbird communities as a whole respond to vegetation regeneration on reclaimed wellsites in the deciduous boreal forests of Alberta, as a measure of ecological recovery. Songbird community similarity between reclaimed wellsites and the adjacent forest increased with vegetation regeneration on the wellsite. Understanding this relationship required data on relative use of the wellsite and adjacent forest by the songbird community provided by the acoustic location system, and could not be detected from presence/absence data on songbird assemblages detected only within the wellsite footprint, and from a standard biacoustic survey. I used Ovenbirds as a focal species, as their response to recovery of disturbances in the boreal forest has been studied extensively. Individual Ovenbirds were identified by their songs, and tracked using the acoustic location system. Ovenbirds would sing from reclaimed wellsites and edges more frequently with increasing canopy cover on the wellsite, and less frequently with presence of conspecifics. Current wellsite reclamation practices result in vegetation recovery which facilitates use of wellsites by songbird communities in upland deciduous boreal forests. This thesis demonstrates that an acoustic location system can be used to provide precise spatial locations of multiple individual songbirds concurrently, and can be used as an effective alternative to conventional bird survey methods.

Invasive silver (Hypophthalmichthys molitrix) and bighead (H. nobilis) carp (collectively bigheaded carp) dominate the Mississippi River Drainage, outcompete native species, and continue to expand northward. Silver carp demonstrate an unusual jumping behavior to boat traffic. Understanding this response is important from a behavioral standpoint and critical to developing effective control methods. Experiments in outdoor concrete ponds (10 x 5 x 2 m), investigated the impact of pure tones (500 – 2000 Hz) and a recording from an outboard motor (0.06 – 10 kHz) on bigheaded carp swimming. Speakers broadcasting the sound stimuli were placed at either end of the concrete pond and bigheaded carp behavior was assessed using overhead cameras. Silver carp were consistently directed away (mean: 11.8 ± 1.3 consecutive responses) from the broadband sound (outboard motor recording) and the bighead carp were more responsive, exhibiting a median of 20.0 consecutive responses (1st Q: 12.0, 3rd Q: 23.0). However, both species were less reactive and habituated quickly to pure tones (after 1-2 trials). Field observations of silver carp jumping frequency, orientation, and direction in response to fast moving (16 – 40 km/hr) boats suggest that jumping is non-random, as the fish primarily move away from (79.8%) and behind the boat (> 90%), avoiding the area directly astern (< 4 m) in the most turbulent part of the wake. Finally, broadband sound (> 150 dB) stimulated wild silver carp to jump in a small tributary of the Illinois River. The fish demonstrated a different pattern than those reacting to the fast moving boats, with carp jumping all around the slow moving boat. This research suggests that sound can be used to alter the behavior of bigheaded carp with implications for deterrent barriers or other uses (e.g., herding fish to increase harvest).

Anthropogenic sound in the ocean, such as sonar, has been related to mass strandings of marine life, in particular the Cuvier's beaked whale (Ziphius cavirostris). Since opportunities to study behavioral responses are limited, furthering our knowledge of whales through simulation has become a well adopted process. To simulate sound reception, the Ziphius ear complex was implemented as a three dimensional finite element model with two mechanisms that enable hearing; pressure loading from surrounding soft tissues and bone conduction from skull vibration. The numerical formulation was summarized to demonstrate an assumed-strain technique using nodally integrated continuum elements (NICE) for a forced harmonic vibration problem. This work focuses on obtaining hearing responses using both mechanisms in conjunction with each other, while also looking into the effects of each mechanism individually. Mesh refinement with surface and volume smoothing reveals variations of simulated auditory response for the two loading mechanisms considered. Effects of the amplitudes and phase shifts, used in loading, along with material damping were reviewed for cases of only one mechanism applied at a time, then with both mechanisms active together. A synthetic audiogram was generated from the stapes velocity transfer function (SVTF) based on relative displacement of the hearing structure. Results found here provide insight on the auditory sensitivity of the Cuvier's beaked whale and its ability to capture sound in the known range of mid-frequency sonar. Since this marine mammal has mass strandings heavily linked to sonar activity, this work allows for an understanding of the whale's hearing over a given range of frequencies.

Throughout the global ocean, there is an abundant and diverse assemblage of fishes aggregated at mesopelagic (200-1000 m) depths. These fishes are critical to pelagic food webs and carbon transport. In the southern California Current Ecosystem, with naturally hypoxic mesopelagic waters, mesopelagic fishes may be vulnerable to predicted ocean deoxygenation. Additionally, water property discontinuities at oceanic fronts can disproportionately affect abundance, compositions, and reproduction of marine animals. In this dissertation, I investigate responses of mesopelagic fishes to ocean deoxygenation and fronts. First, I correlated acoustically-detected Deep-Scattering Layers (DSL) of mesopelagic fishes with midwater oxygen, irradiance, and temperature, and found that the lower DSL boundary correlates most with oxygen. The upper boundary correlates with both oxygen and irradiance. Assuming current deoxygenation rates, I predicted both lower and upper boundaries will shoal. Next, I measured activities of the aerobic enzymes Citrate Synthase and Malate Dehydrogenase, and the anaerobic enzyme Lactate Dehydrogenase to test for changes in metabolic activities of mesopelagic fish in response to dissolved oxygen. There was an apparent suppression of activity at low oxygen concentrations for all species combined. There was no increased reliance on anaerobic activity at depressed oxygen concentrations. Although there is evidence that some mesopelagic fishes may have a rare alternate anaerobic pathway catalyzed in part by Alcohol Dehydrogenase, I did not detect any Alcohol Dehydrogenase activity for 16 species studied.Finally, I compared responses of mesopelagic fish assemblages at three frontal systems. I found no abundance changes across fronts, except for larvae which were elevated at the most stable system. Non-vertically migratory assemblages were uniform across frontal gradients, while migratory and larval fish assemblages were typically altered across fronts. Changes in population growth potential were detected across the two more stable frontal systems for migrators and larvae, though not for non-migrators. These results suggest that deoxygenation may cause habitat compression andmetabolic suppression, while changes to frontal frequency could impact the structure andpopulation growth of mesopelagic fish assemblages. Ongoing monitoring of thesepopulations using existing and novel technologies will allow further understanding ofmesopelagic fish responses to these and other environmental changes.

Abstract: Black-capped chickadees (Poecile atricapillus) are a common North American songbird that produce numerous types of vocalizations with various functions. The vocal repertoire of black-capped chickadees have been the focus of numerous vocal production and perception studies. Black-capped chickadees make an excellent model for studying acoustic communication because their vocal repertoire has been so well-studied. In addition to producing a learned song, parts of their chick-a-dee call are also learned. In many species of songbird, the species’ song is a long, complex vocalization, while the species’ calls are short and acoustically simple. In contrast to this, the fee-bee song of black-capped chickadees is a short and relatively simple two-note tonal vocalization, while their chick-a-dee call is a long and relatively complex vocalization. Although the acoustic structure of fee-bee songs is relatively simple, the function of male songs is similar to the function of other songbird songs: mate attraction and territory defense. In addition, male songs contain information regarding the individual producing the signal, including cues about the male’s dominance rank. In Chapter 2, I conducted a bioacoustic analysis on male songs recorded in different geographic locations. I found that the acoustic features that indicate a male’s dominance rank vary with geographic location, in addition, I found other subtle features within the song that vary with geographic location. Next, I used two operant conditioning techniques (i.e., choice preference task and a go/no-go discrimination task) to examine the perception of dominance cues in male fee-bee songs (Chapter 3). The results suggest that preference and discrimination performance varies depending on the location-of-origin of the singer, the sex of the signal receiver, and the category of songs that is rewarded during the task. I also examined the perception of geography-based acoustic cues in male fee-bee songs using a go/no-go operant discrimination task (Chapter 4). Results from this study provide evidence that male songs contain geography-related cues that are perceived by chickadees. In addition to examining acoustic cues in male songs, I also examined the production of songs by male and female chickadees. In many temperate songbirds, including black-capped chickadees, research has focused on male-produced song. However, in many temperate songbirds it is now recognized that both males and females produce song. To examine the production of songs by female black-capped chickadees, both males and females were recorded and a bioacoustic analysis was conducted on their songs (Study 4), revealing that while male and female songs have overall structural similarity (i.e., two notes), at least one acoustic feature (fee glissando) varies between the sexes. An operant conditioning task revealed that male and female songs belong to separate perceptual categories, but the biological salience of the songs affects the discrimination performance of the birds. In addition, this study revealed that acoustic features within the song’s first note (fee note) likely contain information regarding the singer’s sex (Study 5). Taken together, these studies reveal that a relatively simple vocalization, the fee-bee song, contains multiple types of information and birds can use this information when discriminating among songs; however, the biological relevance of the acoustic signal influences the discrimination performance of the birds.

Abstract: The chick-a-dee call of the black-capped chickadee (Poecile atricapillus) is composed of four main note types (A, B, C, and D) that occur in a fixed order. Sex differences have been identified in a number of black-capped chickadee vocalizations (including tseet calls and fee-bee songs) and in the chick-a-dee calls of other chickadee species (specifically, Carolina chickadees [P. carolinensis]). In the current study, I investigated twelve acoustic features in black-capped chickadee chick-a-dee calls including frequency, duration, and amplitude measurements. Using permuted discriminant function analyses, these features were examined to determine which feature, or combination of features, could be used to identify the sex of the caller. Only one note type (A notes) allowed for the discrimination of male and female calls at levels approaching significance. In particular, the start frequency of A notes provided the best discrimination. This finding is consistent with previous research on Carolina chickadee chick-a-dee calls that found that the starting frequency differed between male- and female-produced A notes (Freeberg et al. 2003). Future research will investigate the behavioural response of black-capped chickadees as they discriminate male and female chick-a-dee calls as well as acoustically manipulated calls. The results of this and future projects will add to our knowledge of the proximate mechanisms underlying vocal communication of black-capped chickadees in particular and, more generally, will add to our knowledge of vocal communication in animals that use learned vocalizations, including humans.

Bioacoustics is a very challenging field due to the necessity of putting fragile, powerhungry pieces of instrumentation out in isolated, hostile environments for long periods of time. Making these instruments as durable as possible, while also considering weight, power consumption, size, and affordability is a constant struggle. In this dissertation, multiple engineering challenges associated with these environments are addressed.

Ocean-bottom seismic networks are often deployed in remote regions to study the structure of the oceanic crust and mantle, and are able to detect both seismic signals and ocean acoustic signals. This thesis examines records from a large ocean-bottom seismic network to decipher the various acoustic signals that were recorded over nearly one year, including baleen whale calls and volcanic landslides. Specifically, ten months of broadband seismic data, recorded on twentynine ocean-bottom seismographs located in and adjacent to the Lau Basin, were utilized to identify baleen whale species and to determine species’ spatial, temporal, and diel patterns. Probable whale sounds that could not be matched to published spectrograms, as well as nonbiologic sounds that are likely of volcanogenic origin, were also recorded. An automatic detection algorithm utilizing empirical orthogonal functions and multinomial logistical regression was developed to detect select sound types from the large data set.

Videotape and audio recordings of a patch reef in Ant Atoll, Pohnpei, Micronesia were analyzed for the presence of bioacoustic activity produced by species of the teleost family Pomacentridae. The acoustic parameters of fish calls were measured and then identified as belonging to either Amphiprion melanopus or Stegastes nigricans. Visual and statistical methods of analysis were combined to compare the calls of the two species and to test for measurable differences in the acoustic parameters of their calls. Soundscape data showed similar patterns to previously observed tropical reef environments. Fish produced single or multi-pulsed calls in the frequency range of 0-1000 Hz that were often accompanied by observable behaviors such as a chases and displays of dominance. Results indicated that the species on this atoll produce similar calls that cannot be distinguished using acoustic data alone. Visual confirmation was necessary for definitive identification of sound producers and in many cases was not possible beyond the family level. Implications for passive acoustic monitoring are that acoustic data collection alone may be able to capture fish abundance but may not be sufficient to fully capture species richness in regions where populations of closely related fish have similar signals.

Sound production is a widespread phenomenon in fishes; however, the importance of acoustic signals and their potential to influence reproduction has not been determined. This dissertation examines fish acoustic courtship signals to investigate whether sound has a role in reproductive success. The pre-spawning sounds of several fishes were recorded and analyzed. The male advertisement call of two species of Belizean toadfish, Sanopus astrifer and Batrachoides gilberti, were found to significantly differ. These data, coupled with data in the literature suggest an influence of habitat characteristics on the calling behavior of toadfishes. Additionally, acoustic playback experiments were employed to investigate the role of male courtship sounds in the Malawi cichlid species, Tramitichromis intermedius. Playback results indicated that male sounds may initiate egg-laying behavior in females, but may not be behaviorally relevant to conspecific males. A discussion of confounding factors in aquarium playback experiments is presented. Technical aspects of fish sound recording, playback, and analysis were also examined to provide information for future fish bioacoustics studies. It was determined that digital cameras are a useful method of recording fish sounds to describe metric characteristics; however, temporal parameters are more accurately captured by hydrophones, which are optimal for use in scientific description of fish sounds. Underwater speakers commonly used in fish playback experiments were tested for fidelity when producing a low-frequency pulsed fish sound. The Electro-Voice UW30 speaker was found to perform the best playback at low sound pressure levels (120 dB re 1 μPa) and at greater distances than the UW30. Many fish sounds have been described in the literature; however, there is no standardization of sample size used in species descriptions. A method is presented that can be used to estimate the level of inclusiveness of sound variability in sound descriptions, and to approximate sufficient sample sizes of recordings. The courtship calls of Dascyllus albisella and Batrachoides gilberti were examined to illustrate this method and to provide a benchmark for future sound descriptions.

Improving vertebrate pest control operations relies on increasing pest animal encounters and interactions with control devices, primarily toxic baits and traps. Encounters and interactions are encouraged using a variety of baits and lures that stimulate an animal’s visual, olfactory or auditory senses, which attract and orientate the target species towards the control device. Likewise, attractants can be used to improve interaction rates with monitoring and detection devices providing more accurate estimates of the distribution and changes in the relative abundance of invasive pest species. In New Zealand, the Australian brushtail possum (Trichosurus vulpecula) is a major invasive pest of conservation and agricultural estates and is the subject of on-going management to keep population densities below environmental damage and disease transfer thresholds. Possum control operations have, traditionally, used visual and, to a lesser extent, olfactory lures for attracting possums to control devices and much research has been undertaken to make improvements to traditional lures. However, the potential for an auditory stimulus to attract possums remains largely unexamined and underutilised for possum control in New Zealand. Research presented in this thesis expands on previous studies with captive animals and examines the development and field testing of an audio lure for possum control. Preliminary field trials showed that possums found audio-lured devices sooner than silent devices, and that a greater proportion of audio-lured devices were located by possums over time. In addition, possums were recorded investigating audio-lured sites at higher rates compared to silent sites, suggesting that possums were more likely to interact with a control device if it had an audio lure. In areas where possum populations were high (>5/ha), audio lures were no better than traditional lures at attracting possums, with all sites becoming saturated by possum activity, suggesting audio lures may be most effective in areas with lower possum densities. Monitoring trials in lower density possum populations undertaken to compare the sensitivity of two detection devices to possum presence revealed that Chew-track-cards (CTCs) were more sensitive at detecting possums than WaxTags®. The palatable olfactory and gustatory peanut-butter stimulus contained within CTCs was believed to encourage possums to bite the detection device and likely contributed to their higher level of sensitivity. The addition of an audio lure increased the sensitivity of both detection devices further and made both monitoring devices equally as sensitive to possum presence. One trial also showed that possums were unlikely to make consecutive visits to detection devices for more than three nights indicating that detection devices should be checked for possum activity and follow-up possum control undertaken to remove detected possums every three nights during possum surveillance monitoring. Audio-lured bait stations spaced on 300 m grids at Mt Misery were as effective at controlling possum populations as the current best practice placement of bait stations on 150 m grids. More than 90% of possums captured within the 300 m grid had interacted with audio-lured bait stations. The large proportion of the possum population found to have interacted with audio-lured devices on 300 m grid suggests that alternative audio-lured control devices, such as multi-kill, self-setting possum traps, could be deployed at similar densities to achieve high levels of possum control. Simulated possum control operations using a possum control model developed from the outcomes of the Mt Misery field trial showed that using audio-lured bait stations on a 300 m grid also provided large infrastructure and labour cost savings. A simulation using the model to control an 81 ha effective area showed that infrastructure costs using standard best practice possum control strategies with bait stations cost 123% more than using audio lured bait stations on 300 m grids. In addition the fieldworker component of the labour costs were 1.5 times greater using standard best practice and required field staff to deliver over 100 kg more non-toxic bait than would be used on the 300 m grid. Overall, it appears that audio lures do have the potential to attract possums to control and detection devices faster, more efficiently and more conveniently than traditional visual and olfactory lures and will provide a useful tool for possum monitoring and control in New Zealand.

Interaction of birds with wind turbines has become an important public policy issue. Acoustic monitoring of birds in the vicinity of wind turbines can address this important public policy issue. The identification of nocturnal bird flight calls is also important for various applications such as ornithological studies and acoustic monitoring to prevent the negative effects of wind farms, human made structures and devices on birds. Wind turbines may have negative impact on bird population. Therefore, the development of an acoustic monitoring system is critical for the study of bird behavior. This work can be employed by wildlife biologist for developing mitigation techniques for both on-shore/off-shore wind farm applications and to address bird strike issues at airports.Acoustic monitoring involves preprocessing, feature extraction and classification. A novel Spectrogram-based Image Frequency Statistics (SIFS) feature extraction algorithm has been developed and was compared against traditional feature extraction techniques such as Discrete Wavelet Transform (DWT) and Mel Frequency Cepstral Coefficients (MFCC). Unlike traditional MFCC, signals were first cleaned with wavelet-denoising during the preprocessing stage. Additionally, a mixed MFCC-SIFS (MMS) technique was also developed. Features extracted from proposed algorithms were then combined with various classification algorithms such as k-NN, Multilayer Perceptron (MLP) and Hidden Markov Models (HMM) and Evolutionary Neural Network (ENN). SIFS and MMS algorithms, combined with ENN and MLP, provided the most accurate results. Proposed algorithms were tested with real data collected during the spring migration, around Lake Erie in Ohio, of five nocturnally migrating bird species native to Northwest Ohio. Also, sparrows, warblers and thrushes passing over the University of Toledo, the Ottawa National Wildlife Refuge and Ohio State University’s Stone Lab between April 20 – May 29 were calculated. Quantification of class level migratory bird has been presented. The developed bird flight call recognition system is suitable for deployment for both on-shore & off-shore wind turbine locations. This system would be suitable for 24/7 remote sensing.

Animal communication systems, which rely upon complex cognitive behavior, specific social contexts, and environments that permit effective transmission, are vulnerable to disruption by anthropogenic disturbance. Forests in urban landscapes are known to differ from rural forests in terms of vegetation, avian communities, and anthropogenic noise. Although these urban-associated differences can elicit demographic consequences, little is known about sub-lethal behavioral effects. Recent studies have implicated anthropogenic noise as a cause of changing bird song in urban areas; however, few have considered alternative explanations, nor the evolutionary and ecological consequences of altered songs. I investigated song variation in an urban landscape by asking the following questions: 1) How do the structural and behavioral components of bird song change across a rural-urban landscape gradient? 2) Which aspects of urbanization best predict changes in song properties? and 3) Does urbanization alter relationships among song, indicators of fitness and male quality? I investigated these questions by recording vocal behavior and monitoring the breeding activity of individually-marked male Northern Cardinals (Cardinalis cardinalis) at nine sites distributed across riparian forests within a rural to urban landscape gradient in central Ohio. Forests within urban versus rural landscapes differed ecologically so that urban forests had greater densities of conspecifics, denser understory vegetation with greater numbers of large trees, louder ambient noise (primarily from traffic), and smaller birds than more rural forests. As expected, cardinal song also changed with urbanization, with songs becoming longer, faster and with higher minimum, maximum and peak frequencies (Hz) as urbanization increased. Ambient noise at the territory level explained shifts in minimum frequency, whereas changes in conspecific densities best explained temporal variation in song structure (e.g., length and syllable rate). Peak and maximum frequencies were not well explained by habitat or morphological models and instead may be driven by factors not considered, such as cultural dialects. Relationships between song and indicators of fitness and quality were not consistent across landscapes for some, but not all, song characteristics. In rural landscapes, the largest birds sang songs that were short and slow; however, song of urban males was not associated with body size. The lack of an association between song and body size in urban birds may be due to increased interspecific aggression from the aggregation of high numbers of territorial males. Thus urbanization may weaken the usefulness of temporal aspects of song as a reliable indicator of size. Relationships between song and parental care were consistent across landscapes, where birds that sang short, slow songs provisioned less and had nestlings in poorer condition. Males that used higher minimum frequencies also had nestlings in poor condition, a possible effect of habitat quality in noisy territories. Male song was not related to reproductive output or territory quality in either landscape. To my knowledge, this is the first study to simultaneously evaluate the evidence for multiple potential mechanisms of urban-associated changes in song, as well as investigate consequences for signal reliability and fitness. These results suggest that a variety of factors, not anthropogenic noise alone, influence song in urban birds and these changes can affect signal reliability in novel environments.

Although wildlife conservation actions have increased globally in both number and complexity, the dearth of scalable, cost-effective monitoring hampers effective intervention to understand conservation efficacy. Automated sensors and computer aided analysis have expanded the tools available for conservation monitoring. For threatened and elusive seabirds, acoustic monitoring presents a cost-effective, scalable alternative to traditional monitoring methods that are limited by remote or inaccessible locations requiring skilled field personnel. A key assumption in monitoring population trends acoustically is that measures of acoustic activity are correlated with the relative nesting abundance of seabirds at colony sites. Here we tested the efficacy of acoustic activity as an index of seabird nesting abundance at colonies. Sensors recorded ambient noise at Forster's Tern (Sterna forsteri) breeding colonies in San Francisco Bay for two breeding seasons. We used an automated method (spectrogram cross-correlation) to detect and count tern advertisement calls from recordings. We calculated mean seasonal call rate and compared it to the mean active nest count at colonies. Our results show that acoustic activity was a reliable index of colony size among breeding sites and an accurate index of change in colony size between years (adj. R2=0.94, n=5, p

Abstract: Songbirds (order Passeriformes, suborder Oscines) have captured the attention of scientists and non-scientists alike with their vocal signals. The black-capped chickadee (genus Poecile) uses its namesake call, chick-a-dee, to convey a variety of information. In Chapter 2 and 3, I examine the relationship between season and diurnal cycle and the production of three vocalizations of black-capped chickadees. In the natural habitat chick-a-dee call production was highest in autumn and winter generally at the meridian. Fee-bee song production increased once in the winter and once in the spring, and occurred almost exclusively at dawn. Gargle production did not differ significantly by season but most occurred during the meridian (Chapter 2). In the laboratory, the patterns of production were in general agreement with the patterns in the natural habitat (Chapter 3). In Chapter 4, I examined what role the phylogenetic relatedness of a heterospecific individual had on neural activity, measured via an immediate early gene, in the auditory brain areas. Using natural calls I found that there was no difference in the amount of neural response from closely individuals but there was less response from a distantly related individual. To further examine this I used calls that were more similar in their bioacoustic structures. With these calls I found no difference in the amount of neural activity regardless of phylogenetic proximity. In Chapter 5, I used mobbing calls to explore whether ‘degree of threat’ is encoded in the auditory processing brain areas. Degree of threat was indicated by the amount of neural activity with high threat mobbing calls and high threat predator calls generating the most activity followed by low threat mobbing calls and low threat predator calls. Thus the ‘degree of threat’ was related to the amount of neural activity and within a threat level there was difference in the amount of activity regardless of the source of the threat. Finally, hand-reared birds had greater neural activity in response to mobbing calls which they had experience with than predator calls which they had no experience with. This result suggests that threat is a learned response and that the neural response is affected.

Many scientists have expressed a need for a new scalable approach to monitoring biodiversity that takes advantage of recent advances in technology, which could allow us to quickly identify and respond to ecological changes. Monitoring biodiversity through the use of remotely recorded sound is one method that has been proposed to fulfill these requirements. The objective of this study is to develop an index (the Modified Acoustic Entropy Index (H’)), using remotely recorded sound, to examine variations in acoustic entropy in agricultural soundscapes, and to test the ability of this index to indicate relative biodiversity levels in these systems. To do this, the H’ index examines the level of niche utilization of the acoustic soundscape, which is hypothesized to relate to the number of species and individuals present in a location. The resulting H’ values can then be used to compare soundscape patterns and acoustic diversity levels over time and between locations.Six locations with divergent Agricultural Health Index (AHI) levels, were selected to test the H’ index. Analyses were conducted to address five key questions: (1) Is there a significant relationship between the acoustic entropy (H’) of sites and their agroecosystem health index values, and/or the AHI’s component variables (Plant landscape diversity (BD), soil quality (SOIL), topography (ELV), social organization (SOC), farm economics (CAUV), and land economics (MKT)); (2) Are H’ levels comparable to those found using traditional measures of biodiversity (plant transect, insect pitfall, and avian call number); (3) How does the sampling radius of H’ compare to the sampling radius of traditional methods; (4) What is the length of sampling needed to ensure significant power of analysis; and (5) can the H’ index be used to identify and analyze anthropogenic disturbances.The results of this study indicate that (1) during certain time periods the H’ index is significantly related to the Agricultural Health Index, plant heterogeneity, and soil quality; (2) In contrast, the standard methods were not related to the biodiversity measure, Plant heterogeneity. Furthermore, none of the traditional methods tested were correlated with each other, calling into question their usefulness as a surrogate for total diversity; (3) Additionally, the sampling radius of the H’ index (> 1,000 m) was found to be greater than that of the traditional methods; (4) For six sites, a sampling period of greater than seven weeks is recommended. Because the index is calculated over a longer period of time, however, it can be used to discern diel and seasonal patterns of animal activity, which cannot be detected as easily using more traditional methods; (5) While the H’ index could be used to identify anthropogenic acoustic disturbances, no difference in rebound time between sites was detected.Therefore, although the H’ index was not correlated with any of the traditional methods, the consistency of the patterns seen and the agreement of the H’ with the AHI tentatively suggest that the H’ index may be used to detect significant differences in diversity as or more consistently than traditional methods of biodiversity assessment

Abstract: To understand the communicative functions of any vocalization it is important to first classify, describe, and measure the elements of that vocalization. Mexican (Poecile sclateri) and boreal (P. hudsonica) chickadees both produce a name-sake chick-a-dee call. Here, the note types present in samples of Mexican and boreal chick-a-dee calls are identified and described. Frequency and temporal measures of each note type are analyzed and show that frequency measures may be useful for note-type and individual discrimination. Call syntax is also analyzed and shows that both Mexican and boreal chickadees produce the notes within their chick-a-dee calls in a fixed order with the potential for any note type to be repeated or omitted within the sequence. This work provides a foundation for future studies aimed at understanding the communicative significance of this call within these species, as well as for comparative work on the chick-a-dee call among all chickadee species.

Purpose: The goal of this project was to categorize the fundamental frequency and durational patterns of chinchilla (laniger) vocal productions relative to typical interaction situations. Methods: This project focused on 4 distinct call types: Exploratory, Contact, Bark, and Alarm from three sources: doctoral dissertation supplements, pet owner posts, and newly collected samples from a single chinchilla. Praat was used to extract the fundamental frequency (F0) contour from the recordings. Results: Primary characteristics of the Exploratory utterances were: token Fo contours had a rapid decrease in frequency (136 ST/s), (2) token durations and token periods overlapped to a large extent across animals, 77% of all tokens contained a final Fo up-sweep tag, and utterances contained an average of 9 tokens.The Contact utterances contained 2-4 Exploratory-like tokens preceding a few transitional tokens, segueing into a sequence of the Contact tokens, Fo contours were complex, variable, and low pitched (300-800 Hz), there is a typical brief low Fo dip of less than an octave from the preceding and following Fo, and the utterances contained an average of approximately 7.Bark utterances were characterized by a brief tonal segment followed by a distinctive noise interval, an abrupt intensity onset followed by a more gradual intensity offset, an increase in token period duration across the utterance, an intensity decrease across the tokens, with a variant of the Bark category including an inspiratory tone preceding or following the token, and the utterances contained an average of 6 tokens. The Alarm utterance token was a high intensity call that included a large, very rapid frequency jump, and maintained a high intensity throughout.

The objective of this study was to provide an overview of the vocal behavior of the American crow, Corvus brachyrhynchos, and to thereby address questions about the evolutionary significance of crow behavior. I recorded the calls of 71 birds of known sex and age in a family context. Sorting calls by their acoustic characteristics and behavioral contexts, I identified and hypothesized functions for 7 adult and 2 juvenile call types, and in several cases found preferential use of a call type by birds of a particular sex or breeding status. My findings enrich our understanding of crow social behavior. I found that helpers and breeders played different roles in foraging and in protecting family territories from other crows and from predators. My findings may also be useful for human management of crow populations, particularly dispersal attempts using playbacks of crows' own vocalizations.

Fighting can be extremely costly in terms of energy use and potential for injury or death. Therefore, the ability to accurately assess a threat is advantageous, allowing an animal to retreat from conflict if its chances of winning are low. Game-theoretical models of agonistic behavior suggest that animals should base decisions about conflict escalation on an assessment of their opponent’s fighting ability or resource-holding potential (RHP). Individuals can assess an opponent’s fighting ability relative to their own using a number of physical and behavioral cues. Some aggressive displays, including some vocalizations, are believed to function as such cues, or advertisement displays, of an animal’s RHP. Studies of advertisement displays have typically focused on contests between individuals. However, just as correct assessment of a rival is advantageous in conflict-escalation decisions between individuals, it might be expected that correct assessment of group size would be an important criterion in determining potential costs in contests between groups. Social animals would benefit from accurate perception of group size when assessing the fighting ability or RHP of a competing group in the same way that individual competitors benefit from assessment of their single competitor’s size and strength. A number of studies have been published in recent years investigating group-level assessment in agonistic contests. These studies have focused on territorial vocal signals of various group-living species to determine whether competing groups glean information about group size from these acoustic signals and adjust their behavior in a manner consistent with the predictions of game-theoretical models of agonistic behavior. The goal of this thesis research was to determine whether group-size information is available in the group yip-howl of the coyote. The approach to this investigation was two-fold: a) conduct a field experiment with radio-collared coyotes and measure their movement and vocal responses to playbacks of group yip-howl recordings from different-sized groups, and b) analyze the acoustic features of the coyote group yip-howl. The aim of the field playback experiment was to measure the responses of coyotes to the agonistic vocalizations of a potential competing group. Acoustic features of the coyote group yip-howl were analyzed to better understand the potential cues contained in the signal that might allow for the perception of group size.

To examine the relationship between social organization and vocal behavior, I studied wild family groups of eastern American Crows, Corvus brachyrhynchos brachyrhynchos in Ann Arbor, Michigan. Some individuals tagged as nestlings acted as helpers at the nest in subsequent years. Helpers did not seem to gain inclusive fitness benefits, at least in the short term. Instead, ecological variables unrelated to group size were more important in nest success. Helpers could have been gaining parenting experience. Crows were partially migratory and group-territorial year round, though some individuals moved in and out of non-territorial summer flocks and roosting winter flocks. Males often remained with their parents longer than females, who may have been more likely to migrate and disperse. Caws recorded from wild and captive crows had a call-like structure, particularly in being harsh and graded. The repetitive, rhythmic, and complex organization of these caws into series, however, resembled vocalizations used by other songbirds in mating competition contexts. Some caw series types appeared to serve as general alarm signals, while unstructured cawing was clearly associated with predator mobbing. However, some caw series types are strongly associated with song-like contexts and one caw series type, doubled short caws, tracked seasonal changes in territorial bout frequency. Crows responded more strongly to playback of doubled cadence than to regular cadence, indicating a role for rhythm in repertoire size determination and supporting a hypothesis that doubled short caws are more territorial, or song-like, in function. Crows also responded slightly differently to caw series of similar cadence but different caw durations. Breeding pairs of crows responded more strongly to playback of strangers than did larger family groups. This result is consistent with a helper-reduced territoriality hypothesis: the presence of helpers lowers territorial response levels because they can usually passively deter intrusions or because the group's odds of success in a fight are great even if intruders engage them. Territoriality in a changing social milieu may provide selection pressure for song-like ritualization and elaboration of otherwise call-like signals.