Introduction

The Pantanal, a World Heritage Site and Biosphere Reserve, is the largest Neotropical seasonal freshwater wetland on Earth (ca. 160,000 km2). This extensive wetland is located in Brazil (states of Mato Grosso and Mato Grosso do Sul), Bolivia and Paraguay (Alho et al. 1988). There are at least two well-defined ecohydrology cycles in the Pantanal: dry and rainy. During the dry season, the fields are predominantly covered with native grassland; the surface water becomes scarce, being restricted to the perennial rivers (with defined beds) and large ponds (Girard et al. 2010; Alho and Sabino 2011). Conversely, during the rainy season the rainwater soaks into the soil and marshes, resulting in the overflow of ponds and rivers (Girard et al. 2010; Da Paz et al. 2014).

The Pantanal has a vegetation system arranged in a mosaic. There is a great diversity of forage species, which make up the main food source for large wild herbivores and also for cattle and horses (Seidl et al. 2001; Bergier 2013). The vast expanse of the grassland plains, allied with a favorable climate, promotes the extensive practice of cattle ranching in the Pantanal (Seidl et al. 2001). Over the last two centuries livestock production has been the dominant economic land use activity of the Pantanal (Seidl et al. 2001; Eaton et al. 2011), where approximately 80 % of the lands are used as native and introduced pastures (Eaton et al. 2011).

As a whole, Brazil has an estimated 176 million ha of native and introduced pastures (IBGE 2012). Native pasture contributes approximately 74 million ha, while introduced pastures cover approximately 102 million ha (IBGE 2012). The substitution of native grassland with introduced pastures has been implemented in many regions of Brazil (including the Pantanal) (Seidl et al. 2001; Alho et al. 2011) aiming to improve the technology of livestock production and achieving higher productivity of cattle (Figueiredo et al. 2012). African grasses of the genus Urochloa (Poaceae) occupy about 80 % of the cultivated pasture areas in the country (Figueiredo et al. 2012). This grass presented excellent adaptation to the Brazilian climate and due to its good nutritional values became widely cultivated for livestock (Rao et al. 1996).

In open areas of native or introduced pasturelands, dung beetles (Coleoptera: Scarabaeidae: Scarabaeinae) bury and utilize the manure deposited by animals for nesting and feeding their offspring, resulting in the cycling of soil nutrients (Slade et al. 2007; Yamada et al. 2007), improved soil fertility (Bang et al. 2005) and aeration (Mittal 1993), reduced fly parasites (Braga et al. 2012) and facilitated growth and development of plants (Nichols et al. 2008). The ecological services and benefit values (Beynon et al. 2015) performed by dung beetles is a practical and natural method of dung removal on pastures (for a review see Nichols et al. 2008), reducing the use of financial resources on the cattle health and on the conservation of pastures (e.g. Lousey and Vaughan 2006).

Many ecological studies have used dung beetles to assess the impacts of land use change on biodiversity (e.g. McGeoch et al. 2002; Nichols et al. 2007; Braga et al. 2013; Korasaki et al. 2013), mainly because dung beetles undergo changes in abundance, species richness and/or composition when subjected to environment changes (Halffter and Favilla 1993; Louzada et al. 2010; Barragán et al. 2011) of anthropogenic or natural origin. For example, the conversion of natural to introduced habitats (Halffter et al. 2007; Navarrete and Halffter 2008; Shahabuddin et al. 2010; Almeida et al. 2011), forest fragmentation and the isolation of forest remnants (Klein 1989; Halffter et al. 1992), and climate change (Larsen 2012). Additionally, they are sensitive, effective and low-cost bioindicators of environmental changes (e.g. Barlow et al. 2007).

In the Brazilian Pantanal, studies on dung beetle communities are scarce (only Louzada et al. 2007; Rodrigues et al. 2010; Tissiani et al. 2015) and the effect of the substitution of natural to introduced pasture on the dung beetle community was yet to be investigated. Therefore, we sampled dung beetles in native (e.g. Andropogon spp. and Axonopus spp.) and introduced pastures (Urochloa spp.) using pitfall traps baited with carrion, cattle dung, and human feces in the dry and rainy seasons to test the following hypotheses: (1) introduced pastures have lower abundance and species richness; (2) there are differences in dung beetle species composition between pasture systems in each season; and (3) the dominant functional guilds are different in the two landscapes.

Materials and methods

Study sites

This study was conducted in Aquidauana, Mato Grosso do Sul, Brazil (19°54′36″S, 55°47′54″W), at Aquidauana’s Pantanal. The climate of the region, according the Köppen classification is Aw, i.e. tropical hot-wet, with a rainy summer and a dry winter (Alvares et al. 2014). The average temperature is 26 °C (12–40 °C), with maximum peaks recorded between September and October. Monthly rainfall averages 30 mm during the dry period between the months of May and September, and it increases to an average of 160 mm during the rainy period between October and April.

The study site possesses extensive areas of natural pasture (resistant to the seasonal flooding that occurs in this biome) that are used as grazing for cattle (Santos et al. 2011). The low fertility soil and limitations on quantity and quality of native forage, results in pastures with a low support capacity of around 4 ha per animal unit (Santos et al. 2011). Consequently, farms in the Pantanal are often very extensive (>10,000 ha), and it has been reported that large pastures are important for dung beetle conservation (Buse et al. 2015). In the 1970s African grasses from the genus Urochloa were introduced as a strategy to increase the support capacity, and therefore, the stock productivity of pastures in the Pantanal. These grass species were adapted to low soil fertility and resistant to flooding, increasing the support capacity to 1.5 animal unit ha−1 (Seidl et al. 2001).

We sampled dung beetles in four areas of native pastures (e.g. Andropogon spp. and Axonopus spp.) and four of introduced pasture (Urochloa spp.). Areas of the same pastures type (e.g. native pastures) were separated by approximately 0.5 km, while areas of different pastures (e.g. native vs. introduced) were separated by approximately 1 km.

Experimental protocol

In each of the eight pasture areas, we installed four linear transects (250 m) located approximately 50 m from the edge habitat and separated from the others by 100 m for the allocation of the traps. The distance (100 m) between transects was used to ensure independence of the samples (Larsen and Forsyth 2005). Therefore, transects were considered as replicates (n = 16 transects) for each habitat in each sampling season. In each transect we demarcated six points separated by 50 m. At each point we installed a set of three traps separated from each other by 3 m in a triangular shape; the traps of each set were baited with approximately 40 g of carcass (decaying beef), cattle dung, and human feces. We sampled dung beetles with different baits in order to assure an accurate sampling and also due to their different functional and trophic specializations.

Dung beetle sampling

We sampled dung beetles in July 2011 (dry season) and January 2012 (middle of the rainy season) with traps remaining active for 48 h in each of the eight pasture areas. Dung beetles were sampled using pitfall traps (15 cm diameter; 9 cm depth), installed at ground level. The pitfall traps were covered with plastic lids (15 cm diameter) supported with three wooden sticks (25 cm), so as to reduce desiccation of the baits and avoid possible damage in case of rain. Within each trap 250 ml of a saline solution + neutral detergent (1.5 %) were added. The baits were placed in plastic containers (50 mL) at the center of each trap using a wire as bait holder. Overall, we installed 576 traps per sampling season, with a total sampling effort of 1152 traps during the study. Sampling in both pastures occurred simultaneously.

The dung beetles were sent to the Universidade Federal de Mato Grosso (UFMT; Cuiabá, Mato Grosso, Brazil) where they were identified by Dr. Fernando Zagury Vaz-de-Mello. Voucher specimens were deposited at the Entomology Section of the Zoological Collection of the UFMT and in the Entomology Laboratory at the Universidade Estadual de Mato Grosso do Sul (UEMS; Aquidauana, Mato Grosso do Sul, Brazil).

To determine the size estimates for each species a sample of 20 individuals were measured (from the clypeus to the pygidium) with a digital caliper. If this number of individuals were not available due to the number of individuals collected in the traps, the size was estimated for all the individuals available.

Data analysis

We used individual-based rarefaction analysis to compare patters of species and sample effort in two pasture types. Comparisons between habitats were carried out visually with a 95 % confidence interval (CI) of the rarefaction curves implemented with the EstimateS software (Cowell 2005).

We used generalized linear mixed modelling (GLMMs) to verify differences in dung beetle abundance and species richness between native and introduced pastures. We used pasture types (fixed factor) and seasons (random factor). For this, we used the type of pasture as explanatory variables and the abundance and species richness as response variables. We used “Poisson” errors for the abundance and richness. All GLMMs were subjected to residual analysis for fitting of the distribution of errors (Crawley 2002). We conducted these analyses with the software R (R Development Core Team 2015). We plotted species rank-abundance distributions to visually compare patterns of species dominance in the two pasture types.

We used a non-metric multidimensional scaling (NMDS) to express graphically the changes in dung beetle community composition among the pasture types in each sampling season. To verify these differences we used an analysis of similarity (ANOSIM) (Clarke and Warwick 2001). All data were performed over a Bray–Curtis dissimilarity matrix with standardized and square-root transformed data (Anderson and Willis 2003). These analyses were performed using the software Primer v.6 with PERMANOVA+ (Clarke and Gorley 2006).

We use the Indicator Value analysis (IndVal; Dufrene and Legendre 1997) to identify the species that were significant and reliable indicators of each pasture system. We used 5000 randomizations to determine the statistical significance of the observed indicator value (Monte Carlo test; p < 0.05). Species with significant IndVal results above 70 % were regarded as indicator species for a given land use. Species with intermediate IndVal, i.e. 45–70 %, were considered detector species (Verdú et al. 2011).

To compare the functional guilds (“foraging strategies”), we classified the species sampled as dwellers, rollers, or tunnelers (as proposed by Hanski and Cambefort 1991). We also classified the sampled individuals as small when they were <10 mm in mean length or large when ≥10 mm (Arellano et al. 2005). The individuals were then allocated in their respective functional guilds and classified as: small or large dweller, roller, and tunneler beetles.

Results

Abundance and species richness

We collected 7086 individuals belonging to 32 species of 16 genera and six tribes of dung beetles (Table 1). In the native pastures we captured 30 species in the rainy season and 19 species in the dry season, while in the introduced pastures we collected 25 dung beetle species in the rainy season and 14 species in the dry season (Table 1). Of the 32 species sampled, 23 were found in both pastures, whereas five were found exclusively in native pastures and only two in introduced pastures (Table 1).

Table 1 Biodiversity of dung beetles sampled in native and introduced pastures in the dry and rainy seasons in the Pantanal, Mato Grosso do Sul, Brazil
Full size table

In the native pastures, four of 30 species were collected in low numbers with just a few individuals (n < 10). From the 3924 individuals sampled, the dominant species were: O. aff. hirculus (19.9 %), D. nisus (15.3 %), Canthidium aff. barbacenicum (14.6 %), Canthidium aff. pinotoides (7.3 %), and Canthon edentulus Harold (4.5 %) (Fig. 1). In contrast, in the introduced pastures 11 of the 25 species were represented by just a few individuals. From the 3162 individuals sampled in this pasture type, the dominant species were: Dichotomius nisus (Olivier) (34.1 %), Trichillum externepunctatun Preudhomme de Borre (32.6 %), Canthon mutabilis Lucas (7.6 %), Onthophagus aff. hirculus (5.5 %) and Dichotomius bos (Blanchard) (5.4 %) (Fig. 1).

Fig. 1
figure 1

Rank-abundance distribution of dung beetle species in native (circle) and introduced pastures (triangle)

Full size image

The observed species richness [Sobs (Mao Tau)] in native pasture was higher than that observed in introduced pasture (95 % CI) (Fig. 2). The dung beetles’ abundance was similar among the native and introduced pastures (\(\upchi^{2}_{(1,60)}\) = 1.12, p = 0.28) (Fig. 3). Species richness was greater in the native pasture (\(\upchi_{(1,60)}^{2}\) = 10.38, p < 0.001) (Fig. 3).

Fig. 2
figure 2

Individual-based species accumulation curves for dung beetles sampled in native and introduced pastures. The dotted lines are 95 % CI, illustrating that there was significant difference among pasture types

Full size image
Fig. 3
figure 3

Mean abundance and species richness of dung beetles sampled in native and introduced pastures in the Pantanal. Different letters above the bars indicate statistically significant differences among pasture types (p < 0.01). Error bars represent SE

Full size image

Of the 32 species analyzed, eight (25 % of the total) were considered indicator or detectors species by classification used by Verdú et al. (2011). Two species were considered indicators and four detector species of native pastures, and two were considered detector species of introduced pastures (Fig. 4).

Fig. 4
figure 4

Indicator values (IV) and abundances of indicator species (IV > 75 %) and detector species (IV: 45–75 %). Bars indicate IV. Dot with error bars indicates mean abundances ±SE

Full size image

Species composition

The ordering of sample points produced a pattern consistent with the species composition in both seasons, with a clear difference in species composition between each pasture system forming a distinct cluster on the NMDS plot in the rainy (ANOSIM; R = 0.86, p = 0.001) and dry seasons (ANOSIM; R = 0.56, p = 0.001) (Fig. 5).

Fig. 5
figure 5

Non-metric multidimensional scaling results (NMDS), constructed from Bray–Curtis matrices, for dung beetle communities in native and introduced pastures in the rainy and dry sampling seasons

Full size image

Functional guilds

In native pastures we sampled small (<10 mm) and large (≥10 mm) dung beetles of the three functional guilds (dweller, roller and tunneler), although the proportion of large dwellers was very low. In introduced pastures the guild of the large dweller beetles was absent (Fig. 6). Small tunneler beetles were the most dominant guild on native pastures, whereas large tunneler and small dwellers beetles were the most dominant on the introduced pastures (Fig. 6).

Fig. 6
figure 6

Proportional change in functional dung beetle guild sampled in native and introduced pastures. Numbers inside the figure represent the species numbers in each functional group

Full size image

Discussion

Patterns of abundance and species richness

This study presented for the first time the effect of conversion from native pasture to introduced pasture on dung beetle communities in the largest freshwater wetland on Earth. Our results reveal a reduction in dung beetle species richness due to the land use change. The biodiversity loss of dung beetles, beyond being a conservation matter, may also negatively affect ecological services, such as soil revolving, nutrient cycling and pest control (see review of Nichols et al. 2008). Therefore, studies on the response of animal groups that provide important services to the ecosystem, due to anthropic disturbance, are important to supply baselines for conservation policies (Louzada et al. 2010; Correa et al. 2013a; Korasaki et al. 2013), that could help in the overall protection of ecosystems.

Our study of dung beetle communities in native and introduced pastures in the Brazilian Pantanal confirmed the occurrence of common species of open areas such as introduced pastures in the Cerrado-Pantanal ecotone (Aidar et al. 2000; Puker et al. 2014, Correa et al. 2013b, 2016), introduced pastures (Louzada and Carvalho e Silva 2009; Almeida et al. 2011; Abot et al. 2012) and native grasslands in the Cerrado (Brazilian savanna) (Almeida et al. 2011). Additionally, we report for the first time the occurrence of Coprophanaeus bonariensis Gory (Phanaeini) for the state of Mato Grosso do Sul (for an annotated checklist see Vaz-de-Mello et al. 2016). It also contributes to the understanding of local diversity of these beetles.

Although dung beetle abundance did not significantly differ between pasture types, species richness did. In this study two species (D. nisus and T. externepunctatum) contributed more than 60 % of the total of individuals sampled on introduced pasture. In general, introduced pastures are considered simple and homogeneous habitats with low diversity of food resources (Louzada and Carvalho e Silva 2009; Almeida et al. 2011). This habitat simplification and homogenization of food resources may modify the distribution patterns and abundance of species (Quintero and Halffter 2009). Thus, only dung beetle species that can adapt to the microclimatic conditions of these introduced pastures, with high temperature and intense solar exposure (e.g. Navarrete and Halffter 2008), and use cattle dung as a food and nesting resource (Louzada and Carvalho e Silva 2009; Correa et al. 2016) can maintain and/or increase their populations. The dominance of a few dung beetle species with a high abundance rate observed in this study reflects a pattern very well documented in the introduced pastures of Brazil (e.g. Aidar et al. 2000; Koller et al. 1999, 2007; Louzada and Carvalho e Silva 2009; Almeida et al. 2011; Abot et al. 2012; Correa et al. 2013b, 2016).

The substitution of native pastures for introduced pastures has been implemented in many regions of Brazil (Seidl et al. 2001; Alho et al. 2011), with the aim of improving the technology of livestock production and achieving higher productivity of cattle (Figueiredo et al. 2012). Information on the effects of substituting native pastures for introduced pastures on the assembly of dung beetles in the Brazilian Pantanal does not yet exist, and given the increasing substitution of native pastures for introduced pastures in this ecosystem, our results demonstrate that native grasslands are important for conservation of dung beetle biodiversity, as already documented for native pastures in the Cerrado (Brazilian savanna) (Almeida et al. 2011). Of the 32 species sampled, 23 were found in both pastures, with five occurring exclusively in native pastures. These findings suggest that the conservation of native pastures may be a source of biodiversity for dung beetles that could potentially colonize introduced pastures in the Pantanal, and thus provide ecological services and benefit values (Beynon et al. 2015) in this agroecosystem.

Despite the native and introduced pastures being relatively very similar (Almeida et al. 2011), indicator species (Canthidium aff. barbacenicum and Onthophagus aff. hirculus) were only found in native pasture. Species considered indicators are highly specific to a certain environment (McGeoch et al. 2002), therefore they are more susceptible to changes in habitat. These two species can be important as study tools and monitoring of native pastures. Detector species found in both pasture types possess a moderate specificity, with different degrees of preference among various ecological states (McGeoch et al. 2002). Their requirements are less rigorous regarding the structure of the environment, when compared to indicator species, making adaptation to new habitat conditions more rapidly, since they can move further into adjacent habitats (Verdú et al. 2011).

Species composition

The dung beetle species composition differed between native and introduced pastures in each sampling season. Generally, in dung beetle communities the number of individuals within populations reduces or species are lost when they are unable to adapt to new environmental conditions (Navarrete and Halffter 2008; Shahabuddin et al. 2010; Almeida et al. 2011). Moreover, despite pastures with introduced grasses being relatively similar to native pastures (Almeida et al. 2011), this change in the vegetation structure may influence the spread of odor from the food source (Correa et al. 2016), favoring species with high dispersal ability (e.g. Hanski and Cambefort 1991). Changes in vegetation structure directly influence the composition of the local fauna of dung beetles (Louzada et al. 2010; Almeida et al. 2011). This can be due to changing environmental factors that directly affect the biology of the species, such as luminosity, temperature and humidity, which provide support for reproduction, nesting, foraging and development of dung beetles (e.g. Halffter and Edmonds 1982; Hanski and Cambefort 1991).

A higher density of cattle on introduced pastures may also play a fundamental role on the construction of local structural heterogeneity on the dung beetle assemblages, altering the vegetal succession by exerting a control in the development of foraging plants (Olff and Ritchie 1998; Adler et al. 2001). The changes in structural heterogeneity caused by intense grazing imply an alteration of habitat diversity, bringing consequences such as a more homogenous environment and transformations in local biodiversity (Wallis-de-Vries et al. 2007). In contrast, structural heterogeneity increases the quantity of ecological niches to species adapted to high or low vegetation (Debano 2006), a characteristic of native pastures in the Pantanal (Santos et al. 2011). Additionally, an increase in cattle density may reduce vegetal coverage, while also compacting the soil surface (Vzzoto et al. 2000), having a negative effect, especially on dung beetle species that build their nests below ground (Bang et al. 2005).

Functional guilds

Small tunneler beetles were dominant in native pastures. We believe that these beetles are dominant in native pastures because their size may permit a greater number of individuals and species to share the same resource (Correa et al. 2013b). In contrast, the large tunneler and small dweller beetles were dominant in the introduced pastures. We believe that the predominance of large tunneler beetles in introduced pastures may in the first instance be associated with higher availability of food recourses. Moreover their large size permits individuals of this guild to use a greater quantity of dung pad (e.g. Louzada and Carvalho e Silva 2009; Correa et al. 2013b, 2016), making it less available to other beetle guilds such as dwellers and rollers. The domain of small and large tunneler beetles in native and introduced pastures, respectively, may also be related to adaptation of these species to open habitats, such as pasturelands.

In summary, we assert that subtle changes in vegetation structure in open areas, beyond altering the composition and the richness of dung beetle species as reported by Almeida et al. (2011) in the Cerrado, also affects the functional guilds of dung beetles.

Conservation of dung beetles in the native pastures

In recent decades (<40 years), cattle production has been considered the main source of land use change in the Pantanal (Seidl et al. 2001; Alho et al. 2011; Silva et al. 2011). Approximately 45 % of the Pantanal has been affected by human activities, including alteration and loss of natural habitats, and if the current rate of degradation continues (2.3 % per year), natural habitats of the Pantanal will be destroyed in approximately 50 years (Silva et al. 2011). The greater dung beetle species richness in native pastures highlights the importance of these ecosystems for the conservation of dung beetle biodiversity. The Pantanal is one of the least known ecosystems in terms of biodiversity of Brazil, and it is known that invertebrates play an important functional role in its conservation (Lewinsohn et al. 2005). Thus, it is necessary to develop a management plan for livestock production in native and introduced pastures in the Pantanal that may encourage the use of native pastures, including agricultural techniques such as rotational management for more efficient use of native pastures (Eaton et al. 2011) and consequently the conservation of dung beetle biodiversity and their ecological services.