2.1 Study Area

We carried out the study in the Aquidauana municipality, Mato Grosso do Sul state, Brazil (19°54′36″ S, 55°47′54″ W; 130 m a.s.l.), which is located in the southern part of the Brazilian Pantanal subregion of Rio Negro. The regional climate is tropical hot-wet (Aw, according to the Köppen classification), with rainy summers and dry winters (Alvares et al. 2014). The annual average temperature is 26°C (ranging from 12°C to 40°C), with the highest average temperature occurring between September and October, and the annual precipitation ranges from 1200 to 1300 mm (Cristaldo et al. 2017).

Livestock production has been the main economic activity in Brazilian Pantanal for the last 70 years, where approximately 80% of the land is used for grazing in native and introduced pastures (Eaton et al. 2011). The studied area has been historically modified by livestock farming activities, in which the 1970s African grasses from the genus Urochloa were introduced as a strategy to increase the support capacity. More recently, since the 2020s, soybean agriculture is expanding in the landscape (Song et al. 2021).

2.2 Dung Beetle Sampling

We sampled dung beetles during the rainy season (January 2023), the region's most appropriate period to sample dung beetles (Correa et al. 2018). Beetles were collected in 15 sampling sites, each one consisting of native grasslands (n = 5 sites), exotic pastures (n = 5), and soybean croplands (n = 5 sites). Native grasslands (e.g., Axonopus spp.) are currently utilised for livestock meat production without intensive management (no use of fertilisers, herbicides), ranging in size from 20 to 80 ha. Exotic pasture areas are large open areas, ranging in size from 100 to 400 ha, cultivated with African grasses (Urochloa ssp.), for livestock meat production without intensive management (no use of fertilisation, and no application of herbicides), with stocking rates between 0.8 and 1.3 livestock per hectare and often use of veterinary drugs (e.g., ivermectin, usually 0.2 mg/kg an annual dose). Soybean cropland areas range in size from 5 to 40 ha, cultivated with intensive management (use of fertilisation, application of herbicides and insecticides). The soybean crops were in the third consecutive year of cultivation, and beetles were sampled when soybean began pod maturation (R7) (see Machado et al. 2023). Sites of the same system (e.g., exotic pasture sites) were separated by approximately 0.5 km to ensure the independence of the samples (da Silva and Hernández 2015), while sites of different habitats (e.g., native grasslands vs. soybean cropland sites) were separated by approximately 2 km.

At each sampling site, we placed a 300 m linear transect, 100 m away from its edge and delimited four sampling points along the transect (100 m apart from each other). At each sampling point, we set up two traps, 2 m apart, one baited with about 20 g of carrion (decaying beef) and the other with fresh human faeces. We used different bait types to accurately represent the local dung beetle functional and trophic groups (Correa, Puker, Ferreira, et al. 2016). In total, we had a sampling effort of 120 traps (2 traps * 4 sampling points * 15 sampling sites) and 40 per habitat type (native grassland, exotic pasture, soybean croplands).

Each trap consisted of a plastic container (15 cm diameter, 9 cm depth) installed at ground level, covered with plastic lids (15 cm diameter) supported with three wooden sticks (25 cm) to reduce desiccation of the bait and to avoid rainwater accumulation. Within each trap, 250 mL of a solution (salt + neutral detergent; 1.5%) was added. The baits were placed in plastic containers (50 mL) at the centre of each trap using a wire as a bait holder. The traps remained active for 48 h at each site, after which the specimens were removed and packed in plastic bags containing 70% alcohol for further sorting and taxonomic identification.

Dung beetles were separated to the genus level (Vaz-de-Mello et al. 2011) and identified to species level (Mota et al. 2023), of which the species identification was posteriorly confirmed by one of us (Fernando Z. Vaz-de-Mello). Voucher specimens are deposited in the Coleção Entomológica de Mato Grosso Eurides Furtado (CEMT) at the Universidade Federal de Mato Grosso—UFMT (Cuiabá, Mato Grosso, Brazil).

2.3 Data Analysis

To assess the survey's completeness for each habitat type, we estimated the sample's completeness using the sample coverage analysis with the ‘iNext’ package (see Chao et al. 2014; Hsieh et al. 2016). This measure of sample completeness reveals the proportion of the number of individuals in an assemblage belonging to the taxonomic groups (i.e., species) represented in the sample. We estimated the sample coverages using an individual-based approach.

To compare the dung beetle diversity among habitat types, Hill numbers were used. Diversity was estimated considering ⁰D (species richness, which is independent on each species abundance), ¹D (exponential of Shannon, which considers the relative abundance of each species, thus representing the number of abundant species in the assemblage), and ²D (inverse of Simpson, which gives a higher weight to species abundance than q1, thus representing the number of dominant species) (Chao et al. 2014; Hill 1973; Jost 2006). Diversity numbers were calculated in iNEXT software v. 1.0 (Hsieh et al. 2016).

We used Generalised Linear Models (GLMs) to analyse the effects of habitat types (native grasslands, exotic pastures, and soybean croplands) on dung beetle species richness (⁰D), ¹D, ²D, and number of individuals. The assemblage attributes were the response variables, and habitat types were the explanatory variables. All GLMs were submitted to residual analysis to evaluate error distribution adequacy (Crawley 2013). Poisson errors were used for dung beetle species richness (⁰D), gaussian for ¹D and ²D, and negative binomial errors for number of individuals. We undertook contrast analysis to test pairwise differences (Crawley 2013). Models with negative binomial errors were conducted with the package ‘MASS’ (Venables and Ripley 2002) and were analysed in R software (R Core Team 2024).

To verify differences in assemblage structure among habitat types, we used Permutational Multivariate Analysis of Variance (PERMANOVA) (Anderson 2001). To test the heterogeneity of multivariate dispersions of samplings (i.e., sampling sites) among the different habitat types, we used Permutational Multivariate Analysis of Dispersion (PERMDISP) (Anderson 2001). The graphical exploration of the differences in assemblage structure of dung beetles among habitat types was performed by using Non-Metric Multidimensional Scaling (NMDS) (Anderson and Willis 2003). The NMDS ordinations, PERMANOVA, and PERMDISP were performed based on the Bray–Curtis dissimilarity matrix, which is sensitive to species abundances. To reduce bias due to the discrepancies of species abundances, we transformed data (square-root). PERMANOVA, PERMDISP, and NMDS were implemented in the Primer with PERMANOVA+ software version 6.0 (Clarke and Gorley 2006).

We used indicator value method following Dufrene and Legendre (1997), to identify dung beetle species that were significant and reliable indicators of each habitat type. We used 5000 randomizations to determine the statistical significance of the observed indicator value (Monte Carlo test; p < 0.05). This analysis was performed with the ‘indicspecies’ package in R software (Cáceres et al. 2022; R Core Team. 2024).

3.1 Diversity and Abundance

Species richness per sampling site ranged between 7 and 19 species in native grasslands, between 13 and 19 species in exotic pastures, and between 12 and 17 species in soybean crops (Table S1). The average species richness (⁰D) did not differ among habitat types ($$$ {\chi}_{2,12}^2 $$$ = 1.89, p = 0.38—Figure 2A). The number of abundant (¹D) (F_2,12 = 6.96, p = 0.01—Figure 2B), and dominant (²D) (F_2,12 = 5.32, p = 0.02—Figure 2C) dung beetle species was higher in native grasslands and exotic pastures when compared to soybean croplands. Regarding dung beetle abundance, native grasslands had between 19 and 152 individuals recorded in each sampling site, exotic pastures ranged from 97 to 793 individuals, while soybean croplands recorded between 103 and 230 individuals (Table S1). The average abundance was higher in the exotic pastures than in native grasslands and soybean croplands ($$$ {\chi}_{2,12}^2 $$$ = 13.24, p < 0.01—Figure 2D), which did not differ in their abundances.

3.2 Species Composition and Indicator Species

Eighteen species were recorded in all three habitat types (Figure 3). The native grasslands had two species recorded exclusively in this habitat and shared five species with exotic pastures and five species with soybean croplands (Figure 3). Three species were recorded exclusively in exotic pastures (Figure 3). Only two species (Canthon aff. dentatus and Gromphas inermis Harold, 1869) were exclusively collected in soybean croplands (Figure 3).

NMDS ordination showed distinct groups, corresponding to the three habitat types, where all of them were significantly different from each other (Pseudo-F = 6.39; p < 0.01—Figure 4; Table 1). Habitat types showed differences in the multivariate dispersion of points (Permdisp-F = 16.86; p < 0.01—Figure 4; Table 1), where native grasslands had a higher dispersion value (34.95 ± 3.35 SE) than the exotic pastures (20.26 ± 2.02 SE) and soybean croplands (16.74 ± 1.12 SE). Finally, of the 37 species analysed, one species (Canthidium sp. 5) was considered an indicator of native grasslands, four indicatorsof exotic pastures, and one species (D. bos) was an indicator of soybean croplands (Figure 1; Table S2).

4.1 Impacts of Land Use on Dung Beetle Assemblage

We found that the number of species (⁰D) was similar among the three habitat types. Our result is the opposite of previous studies in the Amazon that have demonstrated a negative impact of soybean croplands on dung beetle species richness (Machado et al. 2023). Since the region of our study was originally dominated by open ecosystems, such as native grasslands (Pott and Pott 2009), our results suggest that the impact of conversion to exotic pastures and soybean croplands on dung beetle species richness in the Brazilian Pantanal is less severe than other regions with closed canopy (e.g., Amazon rainforest). Nevertheless, Oliveira et al. (2021) and Carvalho et al. (2022) found no significant differences in the dung beetle species richness among the Cerrado, exotic pastures and soybean areas in the Brazilian Cerrado. In this sense, Pantanal and Cerrado ecosystems may harbour a diversity of dung beetle species that successfully use open areas (Correa, Braga, Louzada, et al. 2019; Correa et al. 2022; Macedo et al. 2020; Machado et al. 2023; Vaz-de-Mello et al. 2017). In contrast to species richness results, abundant species (¹D) and dominant species (²D) were higher in the native grasslands than in soybean croplands. This indicates that a higher number of species are represented equitably in terms of abundance in native grasslands than in soybean croplands, in which a few species dominate the local assemblage (Machado et al. 2023). This trend is related to the assemblage dynamics established in both conserved and disturbed environments, in which a few species dominate the local assemblage (e.g., Halffter and Arellano 2002; Korasaki et al. 2013). Indeed, in the soybean croplands, more than 50% of the species recorded were considered rare (e.g., singletons or doubletons), indicating that few species successfully occupy the niches available in soybean croplands in the Brazilian Pantanal. In this sense, our results may allow us to propose the hypothesis that most dung beetle species are using the soybean croplands as transitional areas (e.g., ecological corridors or step-stones; Almeida et al. 2011; Costa et al. 2017) to move between native grasslands and exotic pastures, rarely occurring in soybean croplands. Finally, we found higher dung beetle abundance in the exotic pastures than in native grasslands and soybean croplands. Although exotic pastures are simple and homogeneous systems (Arellano et al. 2023), these environments can provide benefits for species that show capacity to use cattle dung for food resources and/or nesting breeding, develop under high temperatures, intense sun exposure, and cattle-compacted soils (Maldaner et al. 2024). These factors determine their persistence in introduced pastures (Arellano et al. 2023), resulting in an increase in the populations of these species, surpassing, in some cases, the populations of the most preserved natural systems (Escobar et al. 2007; Correa, Braga, Puker, et al. 2019).

Regarding the dung beetle diversity, vegetation structure, availability of mammalian dung resources, and evolutionary history of a region are determinants for species establishment (Hanski 1991; Pêssoa et al. 2023; Raine and Slade 2019). Vegetation structure is a crucial factor in the organisation of dung beetle communities in tropical landscapes (Hanski and Cambefort 1991; Macedo et al. 2020), given that it alters environmental conditions that directly affect species biology, such as luminosity, temperature, and humidity (Hanski and Cambefort 1991). Thus, although natural grasslands, exotic pastures, and soybean croplands show open canopies, the herbaceous density and complexity (e.g., shrubs, native herbs, and plant biomass) (Macedo et al. 2020) may alter the local microclimate conditions (Edmondson et al. 2016; Ozkan and Gokbulak 2017), which negatively affect many dung beetle species in open ecosystems (Correa and da Silva 2022; Larsen 2012). Regarding food resources, most large mammal species are adversely affected by croplands (Canale et al. 2012; Parry et al. 2007), which implies a reduction of dung resource availability for dung beetles, and consequently poorer dung beetle assemblages (Nichols et al. 2009; Raine and Slade 2019). However, as soybean croplands in Pantanal are subject to no-tillage management, there is a relatively high amount of decaying matter on the soil surface, which many species of dung beetles feed on (Oliveira et al. 2021). Finally, in relation to the history of the region, as soybean croplands are the more recent land use in the Brazilian Pantanal (e.g., 5 years) (Song et al. 2021), the native dung beetles have had more time for adaptation and for colonising exotic pastures (e.g., 50 years), which may help to explain poorer dung beetle assemblages in soybean croplands (Pêssoa et al. 2023). However, dung beetle assemblages change throughout time (e.g., Correa et al. 2024; Escobar et al. 2008), and dung beetle species from conserved open ecosystems may invade recently established croplands (Machado et al. 2023). Therefore, it is important to keep monitoring the novel agriculture scenario that comes together with soybean cropland expansion in the Brazilian Pantanal.

4.2 Impact of Land Use on Species Composition

We found that native grasslands harbour dung beetle species composition markedly distinct from those that inhabit exotic pastures and soybean croplands. As the species composition is one of the parameters that most respond to environmental changes (e.g., Audino et al. 2014; Schmidt and Diehl 2008), this result indicates an imperative need for conserving native grasslands. Thus, we recorded a species composition segregation pattern among native vegetation and agricultural systems, evidencing the high degree of habitat specificity of dung beetles in the Brazilian Pantanal (Correa, Puker, Korasaki, et al. 2016; Correa, Puker, Ferreira, et al. 2016; Gonçalves et al. 2022), which had already been documented for dung beetle assemblages in other ecosystems (e.g., Cerrado—Oliveira et al. 2021; Amazon—Machado et al. 2023). In addition, we found that native grasslands had a higher dispersion value (β-diversity) than exotic pastures and soybean croplands. Thus, greater beta diversity between natural grassland sites is related to its environmental heterogeneity (Whittaker 1972) and indicates that resources are distributed in these environments in a heterogeneous manner, which can increase the number of available niches (Stein et al. 2014; Tews et al. 2004), enabling the coexistence of a larger number of dung beetle species (da Silva et al. 2018).

Our results demonstrate well-defined assemblages in each habitat type, highlighting the importance of each of them for conserving the diversity of dung beetles in the studied landscape (Correa, Braga, Puker, et al. 2019), which may help to maintain biodiversity and their associated ecological functions in human-modified landscapes (Costa et al. 2017). The distinct composition of dung beetle species may be related to biotic and/or abiotic factors present in each environment, which provide specific conditions for dung beetles (Hanski and Cambefort 1991; Larsen 2012). Indeed, 48% of species were shared among native grasslands and agricultural systems, demonstrating that the conservation of natural systems may be a primary source of biodiversity for dung beetles in the Brazilian Pantanal. In this context, the management of native grasslands and exotic pasture areas for livestock production, besides being an important economic alternative for sustainable livestock production (Eaton et al. 2011; Santos et al. 2017), may be an alternative to avoid the loss of dung beetle biodiversity (Correa, Puker, Korasaki, et al. 2016; Correa, Puker, Ferreira, et al. 2016).

4.3 Indicator Species

We found the presence of indicator species in the three habitat types, which suggests that dung beetle species respond differently to land use change in the Brazilian Pantanal (Correa, Puker, Korasaki, et al. 2016; Correa, Puker, Ferreira, et al. 2016; Gonçalves et al. 2022). In the case of indicator species of native grasslands (Canthidium sp. 5), it is more susceptible to changes in habitat (McGeoch et al. 2002). Thus, with the constant agricultural expansion in the Brazilian Pantanal, this species is the most susceptible to extinction. Among the four species considered indicators of exotic pastures, Genieridium bidens and Trichillum externepunctatum are common in open areas, widely distributed in South American pasturelands (Maldaner et al. 2024) and considered of high importance for Brazilian pastures (Tissiani et al. 2017). These species could benefit from open-habitat conditions that come from the expansion of exotic pastures in the Brazilian Pantanal and can be important to study for the management and monitoring of exotic pastures. Finally, Dichotomius bos, a large tunneler and nocturnal beetle with a wide distribution, was considered an indicator of soybean croplands. Machado et al. (2023) demonstrated that this species is consistently found in soybean monoculture areas with high abundance, which reinforces the idea that this species can colonise soybean croplands. However, it is important to highlight that the constant use of herbicides and insecticides in soybean croplands poses a potential risk to Dichotomius bos health (see Cavallaro et al. 2023; Villada-Bedoya et al. 2019). Therefore, we suggest that long-term studies that assess the population dynamic and the body condition of this species may be important to determine how the population and individuals are physiologically responding to soybean cropland conditions (e.g., herbicides and insecticides use) and the real effect of these agrochemicals on their health.