W-IZ-4
Diversity Patterns and Associations of Brasilian Lotic Fish Species

Wednesday, September 11, 2013: 9:00 AM
Izard (Statehouse Convention Center)
Robert M. Hughes , Amnis Opes Institute, Corvallis, OR
Ruanny Casarim , Departamento de Biologia, Universidade Federal de Lavras, Lavras, Brazil
Miriam Castro , Departamento de Biologia, Universidade Federal de Lavras, Lavras, Brazil
Nara Junqueira , Departamento de Biologia, Universidade Federal de Lavras, Lavras, Brazil
Cecilia Leal , Departamento de Biologia, Universidade Federal de Lavras, Lavras, Brazil
Rafael Leitão , Laboratório de Sistemática e Ecologia de Peixes, Instituto Nacional de Pesquisas da Amazônia, Manaus, Brazil
Diego Macedo , Departamento de Biologia Geral, Universidade Federal de Minas Gerais, Belo Horizonte, Brazil
Barbara Sanches , Programa de Póst-Graduação em Zoologia de Vertebrados, Pontifícia Universidade Católica de Minas Gerais, Belo Horizonte, Brazil
Bianca de Freitas Terra , Universidade Federal Rural do Rio de Janeiro, Seropédica, Brazil
Francisco Gerson Araújo , Universidade Federal Rural do Rio de Janeiro, Seropédica, Brazil
Marcos Callisto , Departamento de Biologia Geral, Universidade Federal de Minas Gerais, Belo Horizonte, Brazil
Paulo Pompeu , Departamento de Biologia, Universidade Federal de Lavras, Lavras, Brazil
Gilmar Santos , Programa de Póst-Graduação em Zoologia de Vertebrados, Pontifícia Universidade Católica de Minas Gerais, Belo Horizonte, Brazil
Jansen Zuanon , Laboratório de Sistemática e Ecologia de Peixes, Instituto Nacional de Pesquisas da Amazônia, Manaus, Brazil
Biological diversity, frequently measured by species richness, is a common social goal and management objective that is measured at a local or site scale and extrapolated to landscape, basin, or riverscape scales.  However, increasingly the insufficiencies of such extrapolations are being recognized.  In this study, we sampled 24 to 80 randomly or stratified-randomly selected stream or reservoir sites in 7 river basins through use of multiple gears, depending on site characteristics.  After identifying each individual collected to species, we calculated alpha (mean site), gamma (basin), and beta (gamma-alpha, or turnover) species richness.  We also collected landscape and site environmental data for all sites.  Gamma (and alpha) species richness in the basins ranged from 20(4) in a highland savanna ecoregion to 111(22) in an eastern lowland Amazon basin.  Beta species richness was 2 to 8 times, but typically 4 times, alpha richness.  Estimated gamma richness and sampling effort curves indicated that more than 40 sites and site sampling distances greater than 40 times mean stream widths were needed to reach sampling effort asymptotes.  Natural landscape, land use/cover, site, and combined models explained 20%, 11%, 32%, and 47% of the variability in savanna fish species, respectively.