The global demand for food, fiber, and biofuels has intensified land use and accelerated the loss and fragmentation of native habitats (Tilman et al., 2011). In human-modified regions, these processes reorganize spatial patterns and erode ecological integrity, driving biodiversity decline across multiple taxa (Newbold et al., 2020). As landscapes simplify, native vegetation becomes restricted to marginal areas (Mantero et al., 2020), and ecological communities shift toward generalist and disturbance-tolerant species (Gámez-Virués et al., 2015), reducing functional diversity and disrupting ecosystem services (Le Provost et al., 2023).

Landscape heterogeneity, encompassing both land-cover diversity and spatial configuration, has emerged as a key factor modulating the effects of habitat loss (Priyadarshana et al., 2024). Heterogeneous landscapes tend to support higher biodiversity by providing complementary resources and reducing isolation (Fletcher et al., 2024), whereas homogenized landscapes promote biotic and functional homogenization (Gossner et al., 2023), with cascading ecological impacts (Wang et al., 2021). Understanding how habitat amount and landscape heterogeneity jointly shape biodiversity is therefore essential for managing multifunctional landscapes.

Acoustic indices offer a promising approach by summarizing the distribution, variability, and complexity of sounds, providing a cost-effective way to infer biodiversity patterns across large spatial and temporal scales (Bradfer-Lawrence et al., 2023, 2024). Despite ongoing methodological refinement, these indices show strong associations with species richness and activity across taxa (Dröge et al., 2024) and respond sensitively to habitat loss and land-use intensification (Scarpelli et al., 2021; Burivalova et al., 2022). Because acoustically active taxa exhibit strong behavioral responses to environmental variation, soundscapes can serve as indicators of ecological integrity in human-modified systems (Burivalova et al., 2022).

Despite growing interest in soundscape monitoring, few studies evaluate the combined influence of habitat amount and landscape heterogeneity on acoustic patterns. Research often treats these drivers separately, limiting our understanding of whether heterogeneity can buffer soundscape degradation in landscapes with reduced habitat. Moreover, although habitat amount is a key biodiversity driver in fragmented landscapes (Watling et al., 2020), its interaction with landscape composition (Fahrig et al., 2011) remains poorly explored for acoustically active taxa, particularly in tropical regions undergoing rapid land-use transitions. The Atlantic Forest of Brazil provides an ideal context for addressing these knowledge gaps, given its extensive habitat loss, strong land-use gradients, and marked variation in landscape configuration (Vancine et al., 2024).

Here, we investigate how native vegetation amount and landscape heterogeneity jointly influence soundscapes in the Alto Paranapanema watershed, southeastern Brazil. We focus on acoustically active taxa (vertebrates and insects), particularly birds and anurans, given their distinct circadian activity peaks and sensitivity to habitat change (De Araújo et al., 2024). We sampled 72 landscapes spanning a full gradient of habitat amount and heterogeneity, calculated acoustic indices and evaluated five alternative hypotheses: (i) if soundscapes are primarily driven by habitat amount, acoustic indices should increase with the proportion of native vegetation; (ii) landscape heterogeneity may enhance soundscape indices by increasing the diversity and availability of habitats; (iii) both factors may act independently, producing additive effects; (iv) landscape heterogeneity may modify the influence of habitat amount on soundscape indices through an interaction between the two factors; and (v) these relationships differ between diel periods due specific biological groups activity and different acoustic patterns. This framework allows us to evaluate how landscape composition shapes soundscape patterns and their implications for landscape management in human-modified tropical landscapes.

Our results reveal contrasting responses of acoustic diversity (ADI) and acoustic complexity (ACI) to landscape composition. ADI decreased with increasing habitat amount, whereas ACI showed an opposite pattern. Importantly, the positive effect of habitat amount on ACI was strongly modulated by landscape heterogeneity. At lower habitat proportions (<40%), heterogeneous landscapes maintain relatively high ACI, while homogeneous landscapes exhibited marked reductions. Only when habitat cover exceeded ∼40−50% did homogeneous landscapes converge toward higher ACI levels. This indicates that heterogeneity buffers soundscape degradation, sustaining complex acoustic communities even where native vegetation is limited, whereas homogeneous landscapes rely more on high habitat availability to maintain similar acoustic complexity.

The negative association between ADI and habitat amount contrasts with general expectations from habitat–biodiversity relationships (Watling et al., 2020) and may reflect the influence of dominant or continuous acoustic signals, indicating that ADI does not necessarily track species diversity directly. Although previous studies have reported positive relationships between ADI and species richness in less intensive landscapes (Drodge et al., 2024), we found no relationship between heterogeneity and ADI at the landscape scale. A plausible explanation is that altered landscapes are dominated by generalist or edge species (Gámez-Virués et al., 2015; Hendershot et al., 2020) and insects (Burilova et al., 2022), producing broad and continuous acoustic signatures (Gasc et al., 2018). Additionally, higher levels of anthropophony (e.g., machinery and agricultural activities) and geophony (e.g., wind) may increase spectral variability and inflate ADI values beyond those generated by biophony alone. This suggests that low-habitat landscapes may exhibit higher acoustic saturation despite reduced biodiversity, potentially hindering heterogeneity effects. In contrast, more forested landscapes harbor acoustically complex communities, particularly birds and amphibians (Gasc et al., 2015; Alcocer et al., 2022), supporting the positive relationship between habitat amount and ACI.

ACI increased with habitat amount, consistent with evidence that more intact forests sustain complex vocal communities with greater behavioral diversity and acoustic variability (Gasc et al., 2015). The moderating role of heterogeneity suggests that landscape configuration influences not only species occurrence but also the structure of acoustic communities (Barbaro et al., 2022). Heterogeneous landscapes can enhance resource complementarity (Priyadarshana et al., 2024), reduce isolation (Fletcher et al., 2024), and maintain microhabitat diversity, allowing acoustically complex taxa to persist even when habitat cover is low. This aligns with broader ecological evidence showing that land-use homogenization drives biotic simplification (Wang et al., 2021; Gossner et al., 2023) and impairs ecosystem functioning (Newbold et al., 2020), whereas heterogeneous landscapes increase resilience (Estrada-Carmona et al., 2022).

Homogeneous landscapes required substantially more habitat to maintain ACI values comparable to heterogeneous landscapes with <40% native vegetation. This pattern echoes threshold ranges identified in the Atlantic Forest for multiple taxa, typically between 30% and 50% habitat cover (Martensen et al., 2012; Melo et al., 2018; Anunciação et al., 2021). The convergence ∼40–50%, where homogeneous landscapes recover higher ACI, is particularly noteworthy given the broad recommendation of maintaining at least 40% habitat to prevent rapid biodiversity loss (Arroyo-Rodríguez et al., 2020). However, as landscapes become increasingly simplified under intensification, larger amounts of native habitat may be required to sustain biodiversity. Given that many tropical landscapes retain <50% native vegetation, maintaining habitat alone may be insufficient if heterogeneity declines. Our results indicate that heterogeneity can partially buffer these effects by maintaining diverse habitat conditions.

The buffering effect of landscape heterogeneity is especially relevant in regions where agricultural intensification and land consolidation reduce complexity (Kennedy et al., 2019; Morteo-Montiel et al., 2021). In such settings, even landscapes with moderate habitat amounts may experience rapid declines in ecological integrity if heterogeneity is lost (Kennedy et al., 2019). Conversely, heterogeneous mosaics with mixed land-uses and varied vegetation structure can sustain acoustically complex communities even when habitat is scarce. This aligns with growing evidence that heterogeneity supports multifunctional landscapes capable of reconciling biodiversity conservation and production (Perfecto and Vandermeer, 2010; Tscharntke et al., 2021; Barbaro et al., 2022).

From a management perspective, these findings highlight the need for context-dependent conservation strategies. In landscapes with low habitat cover, maintaining or enhancing heterogeneity is critical to buffer biodiversity loss. In contrast, in homogeneous and intensively managed landscapes, increasing habitat amount becomes necessary to restore soundscape complexity. In landscapes undergoing intensification, conservation should prioritize retaining or promoting heterogeneity, whereas in already simplified landscapes enhance habitat amount.

Positive correlations observed between Day and Night indices, particularly for ACI, suggest that soundscape responses are broadly consistent across taxa active at different times. Despite the distinct ecological and behavioral traits of diurnal birds and nocturnal anurans, both respond similarly to habitat amount and landscape heterogeneity. These results support the use of acoustic indices as indicators of ecological patterns, although they integrate biophony, geophony and anthropophony and should be interpreted cautiously. As highlighted in recent studies (e.g., Alcocer et al., 2020; Bradfer-Lawrence et al., 2025a, 2025b; Sugai and Costa-Pereira, 2025), the ecological interpretation of acoustic metrics can benefit from complementary data sources, such as field-based biodiversity surveys or species-level acoustic identification approaches. Within these limitations, ecoacoustics represents a scalable tool for biodiversity monitoring in rapidly changing tropical landscapes.

Overall, maintaining or enhancing landscape heterogeneity can mitigate losses in acoustic complexity in landscapes with reduced habitat amounts, whereas homogeneous landscapes require substantially higher levels of native vegetation to sustain comparable soundscape structure. These findings bridge landscape ecology and ecoacoustics and provide practical guidance for conservation and land management in the Atlantic Forest and other tropical regions under intensification. In regenerating landscapes, soundscape monitoring may also offer a scalable approach to track ecological recovery and the return of acoustically active taxa over time.

Our findings show that habitat amount and landscape heterogeneity jointly shape soundscape patterns. While ACI increases with habitat cover, heterogeneous landscapes maintain relatively high acoustic complexity even at low habitat amounts (<40%), indicating that heterogeneity can buffer reduced habitat availability, whereas high habitat amounts sustain complexity in homogeneous landscapes. In contrast, ADI showed a different pattern, likely reflecting the influence of dominant or continuous acoustic signals in altered landscapes and highlighting the need for cautious interpretation of acoustic indices. These results indicate that under ongoing land-use intensification in tropical landscapes, maintaining the same proportion of native vegetation may still lead to biodiversity decline if landscape heterogeneity is lost. Effective conservation strategies must therefore integrate habitat amounts with actions that preserve or enhance landscape heterogeneity, to sustain resilient tropical soundscapes.

Júlio Cesar Lima de Araújo: Methodology, Investigation, Formal analysis, Writing – Original Draft, Writing – Review & Editing. Vinicius de Avelar São Pedro: Methodology, Writing – Review & Editing. Alexandre Uezu: Methodology, Writing – Review & Editing. Alexandre Camargo Martensen: Conceptualization, Methodology, Writing – Review & Editing, Funding acquisition, Formal analysis.

During the preparation of this work, the author(s) used ChatGPT tool in order to assist with English translation, spelling and grammar. After using this tool/service, the author(s) reviewed and edited the content as needed and take(s) full responsibility for the content of the published article.

This work was supported by São Paulo Research Foundation (FAPESP) [grant number 2018/20501-8]; and Coordination of Improvement of Higher Education Personnel - Brazil (CAPES) [grant number001]

The datasets generated during and/or analyzed during the current study are available on https://doi.org/10.5281/zenodo.15556586.