Assessing the effects of agricultural intensification on natural habitats and biodiversity in Southern Amazonia

produce species diversity maps considering these three groups of terrestrial vertebrates in Bra- zil [5]. These groups were further subdivided into the categories small-ranged species, threat- ened species, and endemic species. Small-ranged species were defined as those species that have a range smaller than the median for that taxon (2,250,813 km2 for birds, 1,230,901 km2 for mammals, 66,979 km2 for amphibians) in Brazil. For example, a bird species is considered to be small-ranged by occurring naturally in a range of less than 2,250,813 km2, which resem- bles the median distribution range for that taxon in Brazil. Threatened species were defined as vulnerable, endangered, or critically endangered according to the IUCN Red List [70]. Consid- ered threatened species are globally threatened as the use of taxanomies was inconsitent in the national RedList [5]. Endemic species were defined as having at least 90% of their range within Brazil and no part of their range extending more than 50 km beyond the Brazilian border. Overall, 1703 bird species, 637 mammal species and 875 amphibian species were considered in this study. We decided against the inclusion of the category total species richness in our assess- ment. Total species richness as an indicator for biodiversity can be misleading as it is mainly driven by wide-ranged species [5, 71] which might even be benefit from degraded habitats [72] while especially endemic and small-ranged species are dependent on the intactness of their respective ecosystems [73].

We calculated the BII accroding to [38] in order to assess the impact of LULCC on overall vertebrate diversity in the time from 2010 to 2030. The BII is defined as the population of a species group i under land-use activity k in ecosystem j, relative to a reference population on the same ecosystem type according to eq 1. BII ¼ SiSjSkRijAjkIijk SiSjSkRijAjk ð1Þ Eq 1: Biodiversity Intactness Index [38]. Iijk, the “population impact”, is the population of a certain species group i under land-use activity k in ecosystem j. Ajk is the area of land-use k in ecosystem j, Rij the number of species of taxon i in ecosystem j. Since the calculation is done on grid-cell level, each cell is associated with a specific land use type. The number of species is the sum of bird species, mammal species, and amphibian spe- cies assigned to one cell respectively. In order to formulate the population impact, a combina- tion of impact values from [74, 75] and [39] was employed. These values indicate the reduction of mean species diversity in respect to a certain type of land use. The values employed are shown in Table 2. A BII value of 1 indicates a species abundance on the pre-colo- nial level. An index of 0,5 indicates that the species abundance is reduced by half in reference to the pre-colonial level. A decreasing BII value is an expression for further reduction of biodiversity intactness due to LULCC affecting regions characterized by the occurrence of different species of different taxa. An increasing BII value expresses a recovery of biodiversity intactness mainly due to the

displacement of anthropogenic land-use out of these regions or by replacement of certain land-use types by “less harmful” land-use types (e.g. cropland to fallow land) within these regions. 3. Results & discussion The first section summarizes the main LULCC characteristics for each scenario. Thereafter, we assess the resultant effect on vertebrate species diversity, (1) by relating natural habitat area conversion and vertebrate species diversity as well as distribution ranges and (2) by calculating the BII. PLOS ONE Agricultural intensification effects on natural habitats and biodiversity PLOS ONE | https://doi.org/10.1371/journal.pone.0225914 November 25, 2020 7 / 21