Project Topics
Get Complete Project Material File(s) Now! »
Mammalian responses
The analysed papers included responses of 99 species of the 150 non-volant terrestrial mammals present in Cerrado (Table 2). I found 37 papers covering the topic of non-volant mammal responses to land-use changes in the Cerrado (Table 3). Most of these studies were conducted at community level, but 13 were at species level, e.g., Maned Wolf (Chrysocyon brachyurus) (20 papers), Giant Anteater (Myrmecophaga tridactyla) (17 papers), Cougar (Puma concolor) (14 papers), Jaguar (Panthera onca) (8 papers), White-lipped Peccary (Tayassu pecari), and Bush Dog (Speothos venaticus). There were significantly more papers focussing = on carnivorous/insectivorous species (t= -3.1452, df = 21, p-value = 0.004886).
Herbivore community organization
The multivariate analysis found dimensions 1 and 2 explained 42.44% of the total variance (Figure 3b). The third axis, alone, explained 17.79%, but when plotted with the first axis we found no significant changes to community structuration. The first dimension was positively correlated with mean body mass and negatively correlated with forest use. The second dimension was strongly correlated with use of open and forested areas. This axis is also largely due to the generalist (negative correlation)/specialist antagonism. Finally, the use of open or forested area contributed to both axes being positively correlation with dimension 1 and negatively correlate with dimension 2. Dimension 1 allow the clear differentiation of two families (Suidae and Tapiridae) from other taxa due to their high body mass. The family Cervidae was also part of the large-bodied Cerrado mammal group, but seperated out due to their preferential use of open areas (Figure 3a). The other 16 families were more discriminated by their diet than by use of natural habitat. Some families (Caviidae, Cricetidae, Dasypodidae, Echimyidae), overlapped substantially in diet. In terms of percentage of generalist or specialized species among herbivorous mammals, the first group represented 55% of the total community, while specialists represented 31%. However, when plotted species, were recorded a slight overlap between generalist and specialist species (Figure 3c), although, generalist species tended to have higher body masses, with the Cervid Blastocerus dichotomus being the only specialist species group to have a high bodymass. Specialist species also tend to use open areas. We found a clear segregation of species by use of natural habitat (Figure 3d).
Carnivore community organization
The two first axes of the multivariate analysis (dimensions 1 and 2) explain 68.21% of the total variance (Figure 4a). Compared to the herbivore community, that of the carnivores seems more structured. The first dimension is positively correlated with the use of open areas and, to a lesser extent, by body mass (Figure 4b). The second dimension is positively correlated with species from open areas and negatively correlated to species with a preference for open/forest areasThe diet vectors are antagonistic for carnivores and insectivores participate either for Dimension 1 and 2 separately. Were see that dimension two separated off five of the seven carnivorous families (Canidae, Dasypodidae, Mephitidae, Mustelidae and Procyonidae). The use of natural habitat is, therefore, a strongly structuring variable. The family Felidae, whose carnivores have a larger body size, used mainly forest areas. When plotting species, we saw a clear structuring between insectivorous and carnivorous species. Insectivores are mainly structured by the dimension 2 and use open areas most extensively. Most carnivorous species use forested areas and have the greatest body masses.
Study system and species data
The data presented in this paper came from a subset of a savanna monitoring networks in Bodoquena Plateau region (20°25’29.28” to 21°44’19.72” S and 56°52’24.46”56° to 17’23.36” W), CAPES-COFECUB Project. The plateau is located in Mato Grosso do Sul State, southern Brazil (Figure 5). The region is characterized by a mountain chain (altitude 450-800m), a tropical climate (20 to 22°C) with some variation along the plateau from tropical climate in the northern part to tropical humid in the southern part and annual rainfall (1300-1700 mm). The landscape of the region is composed of two main vegetation formation s: the Cerrado (Brazilian savanna) and the Atlantic Forest domains, mainly in the region inside ‘Parque Nacional da Serra da Bodoquena (PNSB)’. This area represents an important conservation unit because it is an ecological corridor for biodiversity between the plateau and the BAP plain.
The species data set used in this paper was extracted from camera trap images sampled from the 193 survey site Bodoquena Plateau monitoring network. At each site, a camera trap was deployed for 25 consecutive days, between June 2016 to December 2017. The camera trap network was deployed across the landscape gradient (Figure 6b). Each camera trap was positioned following an standardized protocol to collect data on multiple mammal species: at all 18 sites were installed 193 randomized camera traps; so, for each site, we deployed a set of 10 to 15 cameras trap randomized inside of each site, with the exact number reflecting the percent of forest cover at each site, (i.e., the greater the percentage of forest cover, the more cameras deployed). Camera traps were positioned 40 cm above the ground on the nearest tree to a computer-generated random point, at an angle of approximately 10° relative to the ground. Cameras were set to take a series of three shots when triggered. Camera trap delays to detect passing animal were set at an interval of 3 secs. At the end of the sampling period, SD cards were removed from the cameras and images were analysed by visual observations by the researcher using Wild.ID software (https://github.com/ConservationInternational/Wild.ID). Four herbivore mammal species representing a range of ecological strategies within the regional herbivorous mammal community were used in this paper (see Table 4).
Table of contents :
PRESENTATION AND THESIS STRUCTURE
References
CHAPTER I GENERAL INTRODUCTION
Effect of land-use change on terrestrial mammals of the Cerrado: large knowledge deficits in a
biodiversity hotspot
Methods
Study region
Ecological characteristics of species and their responses to land-use change .
Results
Mammalian responses
Discussion
References
CHAPTER II Spatial extent of landscape effects on herbivorous mammals in one region of the Cerrado and Atlantic Forest hotspots
Abstract
Resumo
Introduction
Methods
Study system and species data
Landscape data
Data Analysis
Results
Sensitivity of change –detection on landscape metric response across buffer sizes
Effect of changing buffer size on landscape metrics
Landscape metric effects on selected species
Discussion
References
CHAPTER III Idiosyncrasies of herbivore mammal responses to land-use changes in a region increasingly affected by agribusiness in the Cerrado hotspot
Abstract
Resumo
Introduction
Methods
Results
Discussion
Implications for conservation
References
