Exploring the soundscape of small freshwater lakes

So many toadfish, so little timea)

The oyster toadfish, Opsanus tau, has been a valuable biomedical model for a wide diversity of studies. However, its vocalization ability arguably has attracted the most attention, with numerous studies focusing on its ecology, behavior, and neurophysiology in regard to its sound production and reception. This paper reviews 30 years of research in my laboratory using this model to understand how aquatic animals detect, integrate, and respond to external environment cues. The dual vestibular and auditory role of the utricle is examined, and its ability to integrate multimodal input is discussed. Several suggestions for future research are provided, including in situ auditory recording, interjecting natural relevant ambient soundscapes into laboratory sound studies, adding transparency to the field of acoustic deterrents, and calls for fish bioacoustics teaching modules to be incorporated in K-12 curricula to excite and diversify the next generation of scientists.

Evidence for rapid divergence of sensory systems between Texas populations of the Mexican tetra (Astyanax mexicanus)

Population divergence is often quantified using phenotypic variation. However, because sensory abilities are more difficult to discern, we have little information on the plasticity and rate of sensory change between different environments. The Mexican tetra (Astyanax mexicanus) is a fish distributed throughout Southern Texas and Northern Mexico and has evolved troglomorphic phenotypes, such as vestigial eyes and reduced pigmentation, when surface ancestors invaded caves in the past several hundred thousand years. In the early 1900s, surface A. mexicanus were introduced to the karstic Edwards-Trinity Aquifer in Texas. Subsequent cave colonization of subterranean environments resulted in fish with phenotypic and behavioral divergence from their surface counterparts, allowing examination of how new environments lead to sensory changes. We hypothesized that recently introduced cave populations would be more sensitive to light and sound when compared to their surface counterparts. We quantified divergence using auditory evoked potentials (AEPs) and particle acceleration levels (PALs) to measure differences in sound sensitivity, and electroretinography (ERGs) to measure light sensitivity. We also compared these results to measurements taken from native populations and lab-born individuals of the introduced populations. Honey Creek Cave fish were significantly more sensitive than proximate Honey Creek surface fish to sound pressure levels between 0.6 and 0.8 kHz and particle acceleration levels between 0.4 and 0.8 kHz. Pairwise differences were found between San Antonio Zoo surface and the facultative subterranean San Pedro Springs and Blue Hole populations, which exhibited more sensitivity to particle acceleration levels between 0.5 and 0.7 kHz. Electroretinography results indicate no significant differences between populations, although Honey Creek Cave fish may be trending toward reduced visual sensitivity. Auditory thresholds between wild-caught and lab-raised populations of recently invaded fish show significant differences in sensitivity, suggesting that these traits are plastic. Collectively, while these results may point to the rapid divergence of A. mexicanus in cave habitats, it also highlights the responsive plasticity of A. mexicanus auditory system to disparate environments.

Visual Active Learning for Labeling: A Case for Soundscape Ecology Data

Labeling of samples is a recurrent and time-consuming task in data analysis and machine learning and yet generally overlooked in terms of visual analytics approaches to improve the process. As the number of tailored applications of learning models increases, it is crucial that more effective approaches to labeling are developed. In this paper, we report the development of a methodology and a framework to support labeling, with an application case as background. The methodology performs visual active learning and label propagation with 2D embeddings as layouts to achieve faster and interactive labeling of samples. The framework is realized through SoundscapeX, a tool to support labeling in soundscape ecology data. We have applied the framework to a set of audio recordings collected for a Long Term Ecological Research Project in the Cantareira-Mantiqueira Corridor (LTER CCM), localized in the transition between northeastern São Paulo state and southern Minas Gerais state in Brazil. We employed a pre-label data set of groups of animals to test the efficacy of the approach. The results showed the best accuracy at 94.58% in the prediction of labeling for birds and insects; and 91.09% for the prediction of the sound event as frogs and insects.

Temperate freshwater soundscapes: A cacophony of undescribed biological sounds now threatened by anthropogenic noise

The soundscape composition of temperate freshwater habitats is poorly understood. Our goal was to document the occurrence of biological and anthropogenic sounds in freshwater habitats over a large (46,000 km²) area along the geographic corridors of five major river systems in North America (Connecticut, Kennebec, Merrimack, Presumpscot, and Saco). The underwater soundscape was sampled in 19 lakes, 17 ponds, 20 rivers and 20 streams, brooks and creeks that were grouped into broad categories (brook/creek, pond/lake, and river). Over 7,000 sounds were measured from 2,750 minutes of recording in 173 locations over a five-week period in the spring of 2008. Sounds were classified into major anthropophony (airplane, boat, traffic, train and other noise) and biophony (fish air movement, also known as air passage, other fish, insect-like, bird, and other biological) categories. The three most significant findings in this study are: 1) freshwater habitats in the New England region of North America contain a diverse array of unidentified biological sounds; 2) fish air movement sounds constitute a previously unrecognized important component of the freshwater soundscape, occurring at more locations (39%) and in equal abundance than other fish sounds; and 3) anthropogenic noises dominate the soundscape accounting for 92% of the soundscape by relative percent time. The high potential for negative impacts of the anthropophony on freshwater soundscapes is suggested by the spectral and temporal overlap of the anthropophony with the biophony, the higher received sound levels of the anthropophony relative to the biophony, and observations of a significant decline in the occurrence, number, percent time, and diversity of the biophony among locations with higher ambient received levels. Our poor understanding of the biophony of freshwater ecosystems, together with an apparent high temporal exposure to anthropogenic noise across all habitats, suggest a critical need for studies aimed at identification of biophonic sound sources and assessment of potential threats from anthropogenic noises.