Agricultural intensification is one of the main drivers of biodiversity loss in worldwide (Flynn et al., 2009; Tuck et al., 2014). Understanding how different farming systems influence biodiversity has become essential for reconciling food production with conservation objectives (Tuck et al., 2014). In this context, organic farming has been proposed as a more biodiversity-friendly alternative to conventional agriculture, as it avoids synthetic fertilizers and pesticides, promotes crop rotation, and generally involves lower land-use intensity, thereby potentially mitigating negative impacts on wildlife while maintaining agricultural productivity (Katayama et al., 2019; Tuck et al., 2014).

The Pampa biome, one of the most threatened ecosystems in Brazil, is dominated by natural grasslands but also includes forests, savannas, palm groves, rocky outcrops, dunes, wetlands, and water bodies (Andrade et al., 2023). In southern Rio Grande do Sul, the Pampa is regionally and continentally important for family farming, legally defined in Brazil as agricultural systems based on family labor and diversified land use (Cicconeto and Verdum, 2012). In this landscape, biodiversity conservation depends not only on the type of agricultural management but also on the structure and composition of the surrounding landscape matrix (Iop et al., 2020).

Amphibians are among the vertebrate groups most threatened vertebrates globally (Harfoot et al., 2021). Their complex life cycles, low dispersal capacity, and dependence on both aquatic and terrestrial habitats make anurans particularly suitable ecological models for assessing the effects of agricultural intensification (Collins and Fahrig, 2017). Most anuran species rely on waterbodies for breeding and larval development, while the surrounding terrestrial environment is essential for foraging, shelter, and dispersal (Moreira et al., 2020). Consequently, anuran communities are structured by processes operating at multiple spatial scales, including local characteristics of breeding sites, such as pond area, depth, vegetation structure, and the presence of predators, as well as landscape attributes that influence habitat availability, connectivity, and movement among ponds (Queiroz et al., 2015; Moreira et al., 2020).

In agricultural contexts, conventional farming practices have been consistently associated with reduced anuran richness and abundance, whereas organic systems often mitigate these effects (Maltchik et al., 2017; Moreira et al., 2014). However, responses to agricultural management are not uniform and may strongly depend on landscape context (Agostini et al., 2021). The amount of native vegetation, habitat heterogeneity, and the spatial arrangement of aquatic and terrestrial habitats can mediate the effectiveness of farming practices in conserving amphibian diversity (Collins and Fahrig, 2017). In grassland-dominated farming such as the Pampa, where breeding habitats are often artificial and embedded within managed matrices, disentangling the relative roles of local habitat conditions and surrounding landscape structure is particularly important (Moreira et al., 2020; Oro et al., 2024).

Although many studies evaluating the impacts of agricultural intensification on amphibians have focused on taxonomic diversity, other dimensions of biodiversity, such as functional diversity, remain comparatively understudied in agricultural landscapes (Pereyra et al., 2018). Functional diversity, defined as the range and distribution of functional traits within a community (Violle et al., 2007), provides insights into community assembly processes and reflects the capacity of communities to respond to environmental change (de Bello et al., 2021). Traits related to body size, reproductive mode, habitat use, and dispersal ability are particularly relevant for amphibians in agroecosystems, as they influence species’ sensitivity to habitat alteration (Pereyra et al., 2018). Moreover, community composition analyses can reveal whether different management systems promote distinct assemblages or lead to biotic homogenization across the landscape (Oro et al., 2024; Preuss et al., 2024).

In this study, we classified farming using pesticides, and high land-use intensity as conventional, whereas organic farming relied on natural alternatives for pest and soil management, lower chemical inputs, and reduced land-use intensity (Carvalho et al., 2025). Importantly, organic management in the study area does not inherently imply greater native vegetation cover around ponds, allowing us to evaluate management effects independently from landscape composition. Both farming systems are embedded within heterogeneous agricultural landscapes characterized by a mosaic of grasslands, forest patches, crops, and artificial ponds, which serve as key breeding habitats for amphibians (Iop et al., 2020).

We aimed to compare taxonomic, functional, and beta diversity, as well as community composition of anurans between organic and conventional agroecosystems in the Brazilian Pampa. Additionally, we evaluated how local habitat characteristics of ponds and landscape composition at different spatial scales influence diversity patterns and community assembly. We hypothesized that organic agroecosystems would support higher taxonomic and functional diversity and distinct community composition compared to conventional systems, although functional composition may remain similar if assemblages are functionally redundant. We further expected that higher native vegetation cover in the surrounding landscape would positively influence diversity metrics.

We recorded 20 native anuran species and one exotic Aquarana catesbeiana in the study area, with sampling coverage above 90% (Fig. S1). Organic farming hosted 20 species, compared to 17 in conventional agroecosystems. Boana faber, Leptodactylus mystacinus, Limnomedusa macroglossa, and Rhinella achavali occurred only in organic systems, while Ololygon berthae was exclusive to conventional ones (Table S1).

No spatial autocorrelation was detected in model residuals (Tables 2–4). The results of the principal coordinates analysis (PCoA) on the measured variables characterizing the ponds indicate that the first axis accounts for 65.76% of the total variation, while the second axis explains an additional 20.62% of the total environmental variation (Fig. S2). Ponds with positive values on the first axis (hereafter referred to as PCoA1) are associated with higher vegetation cover on their surface, whereas ponds with negative PCoA1 values tend to have a greater presence of fish (Fig. S2). In contrast, the second axis (hereafter referred to as PCoA2) shows that positive values are linked to ponds with larger surface area, while negative PCoA2 values correspond to deeper ponds (Fig. S2).

Species richness was higher in organic ponds than in conventional ponds, whereas functional diversity did not differ between agricultural management types on this scale (Table 2; Fig. 1A–B). However, neither metric was related to forest or silviculture cover within 200 m (Table 2).

Fig. 1.

Influence of agricultural management type on A: species richness and B: functional diversity (SES.FD), in ponds sampled within organic and conventional farming in Southern Brazil.

The anuran species composition differed significantly between ponds from organic and conventional agroecosystems (PERMANOVA, F(1,25) = 2.245, p = 0.017; PERMDISP = F(1,25) = 2.63, p = 0.131). Taxonomic beta diversity was 0.58, driven by turnover (0.38) and nestedness (0.20) (Fig. S2). In contrast, functional composition did not differ significantly (PERMANOVA, F(1,25) = 1.15, p = 0.31; PERMDISP = F(1,25) = 0.16, p = 0.683), with functional beta diversity of 0.70 dominated by nestedness (0.62) (Fig. S3).

The first two PCoA axes explained 66.79% of trait variation among species (Fig. 2A). Axis 1 represented a gradient from terrestrial species with benthic larvae and mixed activity periods to arboreal or fossorial species with nektonic or suspension-feeding larvae. Neither PCoA1 nor PCoA2 scores were significantly associated with management type, pond characteristics, or landscape variables (Table 3).

Fig. 2.

The community weighted mean (CWM) traits and their relationship with the characteristic’s ponds (PCoA1 and PCoA2). Only significant trends in CWM are presented here. In section A: PCoA of species traits, B-C: Dots in blue are ponds in organic farming, and red are ponds in conventional farming.

The second PCoA axis was linked to spawning sites and adult body size (SVL), ranging from surface-spawning species to those using foam nests or arboreal eggs (Fig. 2A). While management type had no effect, pond characteristics influenced traits (Table 3): surface-spawning species were less common with greater vegetation cover, which favored foam-nesting or arboreal-egg species (Fig. 2B). Larger ponds supported more surface-spawning species (Fig. 2C). No landscape-scale associations were detected (Table 3).

At the farming level, patterns mirrored those at the pond scale: organic farming had higher species richness and functional diversity than conventional ones (Table 4; Fig. 3A–D). The interaction between management and landscape features was not significant for richness (Table 4; Fig. 3B–C), but functional diversity increased with forest cover in organic agroecosystems (Table 4; Fig. 3E–F).

Fig. 3.

Relationship between the species richness and A: farming management; B-C: interaction between farming management and landscape features (scale 500 m, forest and silviculture, respectively); relationships between functional diversity (SES.FD) and farming management; E-F: interaction between farming management and landscape features (scale 500 m, forest and silviculture, respectively).

Our results demonstrate that anuran diversity patterns in Brazilian Pampa farming are shaped by the combined effects of agricultural management, local habitat characteristics, and landscape context. At the pond scale, species richness was consistently higher in organically managed farming, indicated that organic farming can mitigate biodiversity loss in agricultural landscapes (Tuck et al., 2014; Katayama et al., 2019). These benefits are likely associated with reduced pesticide and lower land-use intensity, which contribute to more favorable conditions for amphibians (Maltchik et al., 2017; Katayama et al., 2019).

In contrast, functional diversity did not differ between management types at the pond scale, indicating that although organic systems support a greater number of species, the range of functional traits represented locally remains similar between organic and conventional ponds. This pattern suggests some degree of functional similarity among anuran species occupying agricultural ponds, whereby species turnover does not necessarily translate into changes in functional diversity (Pereyra et al., 2018). Similar decoupling between taxonomic and functional diversity has been reported in amphibian assemblages exposed to agricultural intensification (Pereyra et al., 2018).

Despite this functional similarity at the local scale, taxonomic composition differed significantly between ponds in organic and conventional farming. Species replacement rather than nested species loss dominated taxonomic beta diversity, indicating that management practices filter species identities without eliminating functional strategies. Conventional farming was likely dominated by disturbance-tolerant and generalist species, while organically managed systems supported a broader set of species (Flynn et al., 2009; Agostini et al., 2021). This taxonomic turnover without corresponding functional divergence highlights the importance of considering multiple dimensions of biodiversity when assessing the ecological effects of agricultural practices (Pereyra et al., 2018.

Analyses of community-weighted mean traits further revealed that functional trait composition was more strongly associated with breeding site characteristics than with management type. Pond vegetation cover and pond size influenced the distribution of reproductive strategies and body size, favoring species that use foam nests or arboreal egg deposition in structurally complex habitats. For instance, species such as Phyllomedusa iheringii and foam-nesting Leptodactylidae were favored by structurally complex ponds, whereas open, simplified ponds tended to support more generalist taxa (Moreira et al., 2020). Vegetation plays a critical role during amphibian reproduction by providing oviposition sites, shelter, and foraging habitat for larvae, while also reducing exposure to agrochemicals (Agostini et al., 2021). The loss of aquatic vegetation, often associated with herbicide use in agricultural systems, can therefore negatively affect amphibian reproductive success and survival (Agostini et al., 2020).

At broader spatial scales, landscape context emerged as a key driver of functional diversity. While landscape composition within 200 m of ponds was not associated with diversity metrics, functional diversity at the farming level increased with forest cover in organic systems. This scale-dependent response suggests that local breeding habitats alone are insufficient to maintain functional diversity and that access to surrounding natural habitats is critical for sustaining a wider range of ecological strategies (Ribeiro et al., 2017; Moreira et al., 2020). Forest patches likely function as refuges and sources for recolonization, enhancing functional complementarity in organically managed landscapes (Ribeiro et al., 2017).

The contrasting effects of forest cover and silviculture further emphasize the importance of landscape quality. Monoculture plantations, such as eucalyptus plantations common in this region, can reduce habitat suitability and connectivity for amphibians, particularly when combined with intensive management practices (Saccol et al., 2017). In contrast, landscapes that integrate organic farming with native vegetation may better support functionally diverse assemblages.

Overall, our findings indicate that organic farming in the Brazilian Pampa support higher anuran species richness and, at broader spatial scales, greater functional diversity when embedded within landscapes containing native forest cover. These results underscore the importance of integrating sustainable agricultural practices with landscape-level conservation strategies to mitigate biodiversity loss and maintain ecosystem functioning in grassland agroecosystems (Agostini et al., 2021). Policies that promote organic farming alongside the conservation and restoration of natural habitats may therefore play a crucial role in reconciling agricultural production with biodiversity conservation.

CJCP thanks CAPESfor a master's scholarship. This research was funded by FAPERGS (proc.21/2551-0001950-6), andCNPq (proc.406414/2021-1). LD research has been funded by a CNPq Productivity Fellowship (grant 306951/2022-3).