What’s all that racket! Soundscapes, phenology, and biodiversity in estuaries

frequency ranges. Pearson’s correlation (r) between monthly averages of Shannon-Wiener diversity index and monthly averages SPLs in (E) high (7000–40,000 Hz), (F) low (50–1200 Hz), and (G) broadband (1–40,000 Hz) frequency ranges. Linear regression between monthly averages of temperature as the independent variable and monthly averages of (D) species rich- ness and (H) Shannon-Wiener diversity index as dependent variables between 2016 and 2018 at all stations combined. For correlations N = 86 and for regression N = 90. (TIF) S16 Fig. Correlation and regression analysis of species abundance, sound pressure levels, and temperature. Pearson’s correlation between monthly averages of species richness and monthly averages of sound pressure levels (SPLs) in (A) high (7000–40,000 Hz), (B) low (50– 1200 Hz), and (C) broadband (1–40,000 Hz) frequency ranges. Linear regression between monthly averages of temperature as the independent variable and monthly averages of (D) spe- cies abundance between 2016 and 2018 at all stations combined. For correlations N = 86 and for regression N = 90. (TIF) S1 Table. Significance of specific variables on sound pressure levels. Results of the Boruta, a wrapper algorithm based on random forest, that tested the significance of specific variables on high (7000–40000 Hz), low (50–1200 Hz), and broadband (1–40000 Hz) frequency sound pressure levels (SPLs). Decision was confirmed important at p < 0.01; N = 130,679. (XLSX) S2 Table. Significance of specific variables on sound pressure levels with files that con-

tained physical sounds and anthropogenic noise removed. Results of the of the Boruta, a wrapper algorithm based on random forest, that tested the significance of specific variables on high (7000–40000 Hz), low (50–1200 Hz), and broadband (1–40000 Hz) frequency sound pressure levels (SPLs). Decision was confirmed important at p < 0.01; N = 130,679. (XLSX) S3 Table. Results of first positive Posterior Probability of change (PP). Year, station, mean spring water temperature, date, and day of year of first PP of change � 0.5 with corresponding day anomaly, and value of PP for high (7000–40000 Hz) frequency sound pressure levels (SPLs). (XLSX) S4 Table. Results of first positive Posterior Probability of change (PP). Year, station, mean spring water temperature, date, and day of year of first PP of change � 0.5 with corresponding day anomaly, and value of PP for low (50–1200 Hz) frequency sound pressure levels (SPLs). (XLSX) S5 Table. List of species that were caught and quantified during seining conducted one or two times per month in close proximity to passive acoustic stations in the May River, SC. (XLSX) S6 Table. Locations of National Oceanic and Atmospheric Administration (NOAA) weather stations located close to the May River, SC. Stations were used to obtain rainfall data for each day. Data were obtained from: https://www.ncdc.noaa.gov/cdo-web/search. (XLSX) S1 Data. (XLSX) PLOS ONE Estuarine soundscapes and biodiversity PLOS ONE | https://doi.org/10.1371/journal.pone.0236874 September 3, 2020 15 / 18