Scale-dependent landscape effects on soundscape: A case study in Guangzhou, China

As a key acoustic indicator of ecosystem health and human well-being, acoustic indices have attracted growing interdisciplinary attention from the fields of ecology, urban planning and conservation biology. They are increasingly being used for rapid biodiversity assessment, habitat quality monitoring, and the management of urban ecosystems. However, the scale dependency of acoustic indices in relation to different landscape characteristics remains poorly understood, particularly in urban environments, where quantitative research on the effects of multiple spatial scales is still limited. Using two to four months of audio data collected at 32 passive acoustic stations in 11 urban parks in Guangzhou, we examined the explanatory power of three acoustic indices at radii of 25 m, 50 m, 100 m and 200 m, and quantified the effects of landscape characteristics at each scale. All indices exhibited significant scale dependency, with the 50 m scale identified as optimal for the acoustic diversity index (ADI), the acoustic entropy index ( H ), and the normalized difference soundscape index (NDSI) (marginal R 2 = 0.48, 0.84, and 0.61, respectively). At this critical scale, forestland and shrub-grassland areas were significant positive drivers of acoustic diversity. In contrast, construction land area significantly suppressed the H and NDSI indices at broader scales (100 m and 200 m). The 25 m buffer consistently demonstrated the lowest explanatory power for ADI and H . Our findings highlight the existence of distinct scale thresholds for urban green space soundscape. We recommend integrating vegetation configuration, building density, and landscape connectivity into urban park planning across multiple spatial scales to directly enhance acoustic diversity and residents’ well-being.