Understanding the threshold of species loss by combining local knowledge with in-field data on mammal and bird species in the Cerrado Hotspot

Abstract

Human-induced land use changes are the most important threat affecting the survival of several species and reducing the provision of ecosystem services for local communities. Recent studies have shown that once the vegetation cover of the landscapes reaches certain levels, the number of species decline faster, which is called threshold of species loss. It is urgent to undertake such studies in human-modified landscapes in order to provide relevant information for decision making and conservation in private and public lands. Furthermore, it is highly relevant to include non-forested ecosystems in the scope, such as the Cerrado Hotspot, since those types of ecosystems have been often neglected for conservation in Brazil. This study approaches the threshold of species loss in humanmodified landscapes in three different perspectives. First, it makes a review of empirical studies word-wide that use threshold of species loss approach with birds, and finds 31 papers published from 1994 to 2018, with 24 studies conducted at temperate latitudes and seven in tropical regions, remarking the increasing tendency of the studies and their potential application to conservation and restoration strategies of landscapes for bird conservation. Then, it performs an empirical research at the Serra da Bodoquena Plateau with 18 medium and large mammal and six bird species using data collected with camera traps. It focuses on the 9 mammals and 2 birds that negatively responded to native vegetation loss at 500 m buffer, resulting in an average threshold of 45.97% (S.E. 12.68) of native vegetation cover. When interpolating this value in modeled anthropogenic use conversion maps for 2030 and 2050 to project how the occupancy probability will change over time, the predicted annual loss found was 22.6 km2 above the average threshold value, indicating that almost half of current area with values above the threshold will be below them by 2050. In addition, this thesis explores a more social component by investigating the local inhabitant’s perception about mammal richness at property level. In order to understand differences in perceptions according to main economic activity, I performed interviews to 37 local inhabitants dedicated to agriculture production, cattle ranching and tourism in a Cerrado area of Mato Grosso do Sul state. Although I find no significant difference in the total richness perceived according to economic activity, there is a significant difference in the richness perception of open areas and forested species within touristic properties (t 0.0194, n=6), which suggest that this category of land owners, relying on economic activities directly related to biodiversity, have a better knowledge about their biodiversity and may be willing to protect larger tracts of forests within their properties.