Investigating audible and ultrasonic noise in modern animal facilities.

The environmental housing conditions of laboratory animals are important for both welfare and reliable, reproducible data. Guidelines currently exist for factors such as lighting cycles, temperature, humidity, and noise, however, for the latter the current guidelines may overlook important details. In the case of the most common laboratory species, the mouse, the range of frequencies they can hear is far higher than that of humans. The current guidelines briefly mention that ultrasonic (>20 kHz) frequencies can adversely affect mice, and that the acoustic environment should be checked, though no recommendations are provided relating to acceptable levels of ultrasonic noise. To investigate the ultrasonic environment in a large mouse breeding facility (the Mary Lyon Centre at MRC Harwell), we compared two systems, the Hottinger Bruel and Kjaer PULSE sound analyser, and an Avisoft Bioacoustics system. Potential noise sources were selected; we used the PULSE system to undertake real-time Fourier analysis of noise up to 100 kHz, and the Avisoft system to record noise up to 125 kHz for later analysis. The microphones from both systems were positioned consistently at the same distance from the source and environmental conditions were identical. In order to investigate our result further, a third system, the AudioMoth (Open Acoustic Devices), was also used for recording. We used DeepSqueak software for most of the recording analysis and, in some cases, we also undertook further spectral analysis using RX8 (iZotope, USA). We found that both systems can detect a range of ultrasonic noise sources, and here discuss the benefits and limitations of each approach. We conclude that measuring the acoustic environment of animal facilities, including ultrasonic frequencies that may adversely affect the animals housed, will contribute to minimising disruption to animal welfare and perturbations in scientific research.