Measuring and interpreting the temporal variability in the soundscape at four places in Sequoia National Park

The soundscape was recorded in four selected places in Sequoia National Park CA, to quantify and assess the diurnal and seasonal character of the park's soundscape. The recording sites were selected to represent a combination of elevation and vegetation diversity. Hour-long sound recordings were made by four individuals at each place during fall, spring, summer and winter at dawn, midday, dusk, and midnight with identical recording instrumentation. The recordings of the soundscape were made in an old growth forest (Crescent Meadow), in a foothill oak savanna (Sycamore Spring), in an upland savanna chaparral (Shepherd Saddle) and in a foothill riparian location adjacent to the Kiawah River (Buckeye Flat). Sound recordings were analyzed using a normalized Power Spectral Density (PSD) algorithm and partitioned into 1 kHz intervals based on 12 subsamples from each of the 64 h-long sound recordings. Biological signals (biophony) were based on the highest PSD value within the range of 2-8 kHz. A multilevel analysis (MLA) was used to examine temporal patterns of biophony at four locations in Sequoia National Park. Unsupervised Landsat Thematic Mapper Satellite Imagery identified 25 vegetation regimes in Sequoia National Park. Satellite signatures of the habitat where recordings were made were extracted from the imagery to scale to the region.