According to Paglia et al. (2012), the 21 species of small mammals registered in the SAES, represent about 21% of the fauna occurrence of small mammals in the Cerrado biome. This species richness is comparable to other studies in the same biome, such as Mares et al. (1986), Bonvicino et al. (1996), Lacher and Garlic (2001), Cáceres et al. (2011), Carmignotto and Aires (2011), Rocha et al. (2011), Bonvicino et al. (2012), Santos-Filho et al. (2012), Mendonça et al. (2018), Carmignotto (2019) and Carmignotto et al. (2022).
The species richness recorded in each sample site of the cerrado ss was lower than the species richness in the forest sites, more accentuated in the dry deciduous forest. This result is explained by the higher complexity of the forest habitats’ phytophysiognomy. Therefore, the more vertically structured and complex habitats can offer greater resource availability, thus harboring higher species diversity (August, 1983; Tews et al., 2004; Santos-Filho et al. 2012). Several works have found this positive correlation between the small mammals’ richness and the vegetation's structural complexity (August 1983; Santos-Filho et al. 2012; Mendonça et al. 2018; Carmignotto 2019; Carmignotto et al. 2022).
The deciduous dry forest also presented higher abundance than the other habitats. It is worth noting that, in addition to the vertical complexity of the vegetation, it has rocky outcrops, the preferred microhabitat of the T. apereoides species, the most abundant in this study (Garlic 1982; Mares et al. 1986; Nowak 1991; Lacher and Garlic 2001).
The fires’ regularity and the habitats’ complexity were important predictors structuring the species composition among the habitats sampled by the SAES. Of the 21 species recorded in this work, 17 were shared between at least two phytophysiognomies, and only 4 species are restricted with only one individual collected in each phytophysiognomy, Cerradomys scotti in the Cerrado ss; Guerlinguetus aestuans in the Babassu forest and Gracilinanus agilis and Dasyprocta azarae in the gallery forest. Species dissimilarity between forested and savannah environments in the Cerrado is well described in studies by Mares et al. (1986), Bonvicino et al. (1996), Garlic (2005), Carmignotto and Aires (2011), Santos-Filho et al. (2012); Carmignotto et al. (2014) Mendonça et al. (2018), Carmignotto (2019) and Ribeiro et al. (2020). However, our results show that most of the species are common in forest and savannah environments, which may reflect the homogenization of habitats in the absence of fires.
Fire effect:
By comparing the species richness observed at SAES in 1999/2000 with the species richness observed in 2016/2017, we concluded that there was no change in total richness. However, there was an increase in the richness of each phytophysiognomy, as well as the number of species shared among the phytophysiognomies, indicating a homogenization of environments in the face of the transformation from savannah to forested areas, due to the fire absence.
Our results show the replacement of species typical of savannah environments by forest species, of the six species sampled in 1999/2000 that were not captured in 2016/2017, three have a preference for savannah environments, Carterodon sulcidens = (Santos-Filho et al. 2012, Euryzygomatomys spinosus), Kunsia tomentosus and Calomys sp. And of the seven species that were recorded only in 2016/2017, six are generalist species or are mainly associated with forest environments, as is the case of Neacomys spinosus and Nectomys rattus (Santos-Filho et al. 2012; Carmignotto et al. 2014; Carmignotto 2019). Only the Cerradomys scotti species had a single specimen collected in the Cerrado ss. This species is typical of savannah habitats in the Cerrado, but less frequently, and can be found in veredas, riparian forests, and cerradão (Langguth and Bonvicino 2002; Furtado et al. 2021).
There was no record of the species Neacomys spinosus in 1999/2000. In 2016/2017, 50 individuals were collected, one of which was recorded in the Cerrado ss. However, for Bonvicino et al. (1996) and Patton et al. (2000) individuals of this species are frequently captured in primary forest and secondary vegetation, in upland and floodplain habitats.
Considering the species recorded in 1999/2000 only in forest phytophysiognomies and, in 2016/2017, also recorded in the Cerrado ss, the following stand out: Hylaeamys megacephalus, Marmosa (Marmosa) murina, Marmosa (Micoureus) constantiae and Metachirus nudicaudatus.
Only one individual of the species Hylaeamys megacephalus was collected in 1999/2000 in the gallery forest, in 2016/2017 95 individuals were collected, eight individuals in the Cerrado ss. This species inhabits a variety of biomes but is always restricted to forest habitats (Mares et al. 1986; Ochoa et al. 1993).
In 1999/2000, Marmosa (Marmosa) murina was recorded in gallery forest and Babassu forest. The species Marmosa (Micoureus) constantiae and Metachirus nudicaudatus were recorded in the Babassu forest. However, in 2016/2017 the three species were collected in the Cerrado ss. It is worth considering that they are forest habitats’ species (Mendes-Oliveira et al. 2010; Cáceres et al. 2012; Voss et al. 2019).
Thus, our results indicate that the fire absence is restructuring the small mammal assemblages in the SAES, since, in the fires absence, there is a continuous increase in the trees density, replacing grasses and shrubs with larger woody plants (Stevens et al. 2016; Maracahipes-Santos et al. 2018). The fires absence can reduce or extinguish typical habitats of the Cerrado savannah and its specific resources for fauna, such as grasses. Hence, the response of small mammals to these changes shows different preferences for habitat use (Emmons and Feer 1997; Santos-Filho et al. 2012; Ribeiro et al. 2020), occurring a substitution of species considered to be from the Cerrado savannah and the addition of species that use more forested habitats. Typically, Cerrado species such as Carterodon sulcidens = (Santos-Filho et al. 2012, Euryzygomatomys spinosus), Kunsia tomentosus and Calomys sp. disappeared from these areas with the change from savannah to forested areas in the absence of fire.
However, these results still need to be better investigated, because in studies with fire and small mammals’ assemblages in savannahs in Australia, Legge et al. (2019) observed that different threats can interact synergistically, affecting the small mammal assemblages. Therefore, the management of unique threats to biodiversity can be compromised if interactions between different threats are not considered (Legge et al. 2019).
In this way, we suggest further studies, where the introduction of controlled fire can be used in the SAES, as the vegetation because of fires forms three groups. One of initial succession, reaching the peak in two years, where they are adapted to open environments and to deal with fires; a second group reaches its peak between two and nine years, and the third group of late succession reaches its peak after ten years (Vieira 1999; Mendes-Oliveira et al. 2012; Vieira and Briani 2013; Holmes and Robinson 2016; Lindenmayer et al. 2016). However, the diversity and abundance of small mammals reach maximum values in the first stages of succession, and fire suppression can lead to these populations’ extinction through habitat modification and impaired demographic processes (Briani et al. 2004; Hutto 2008; Templeton et al. 2011).
Given the above, it is concluded that the fire extinction has negatively affected the maintenance of small mammal assemblages in the SAES, considering that fire is a determining factor for the dynamics of the Cerrado, which has a historical evolution modeled by fire processes (Fidelis 2020).