Rapids habitats in a fluvial landscape may affect rheophilic fish and invertebrate assemblages much the way islands in the ocean affect assemblages of terrestrial organisms like birds and lizards. Each rapid is a spatially discrete zone of solid substrate and elevated water velocities in which philopatry and specializations for station-holding and substrate feeding are valuable. One of the largest and most complex rapids habitats in the Amazon Basin is the Volta Grande (Portuguese for Big Bend, see map at left), a horseshoe-shaped bend in the lower course of the Xingu River that consists of a series of reticulated rapids and anastomose channels, some over 60 m deep. Since 2013, I have been a participant in the NSF-funded Biological Survey and Inventory of the rapids of the lower Xingu River (NSF DEB-1257813, the iXingu Project). I have been responsible for developing next-generation, restriction associated DNA sequencing libraries and bioinformatics pipelines to infer a high resolution picture of gene-flow within several species spanning these rapids. I am particularly interested in hybridization zones within the rapids (see photos at right) and understanding the relative influence of biotic (e.g., life history traits) vs. abiotic (e.g., habitat traits) factors on gene flow. Preliminary results suggest the presence of significant, unrecognized diversity within the rapids and we are now trying to understand why.
EXAMPLE 1: The iXingu Project: a collaborative project with Mark Sabaj Pérez, John Lundberg, Brice Noonan, Leandro Sousa, Lucia Rapp Py-Daniel and Kirk Winemiller
EXAMPLE 2: Lujan, N.K. and K. Conway. 2015. Life in the fast lane: a review of rheophily in freshwater fishes. pp. 107–136 in: R. Riesch, M. Plath, M. Tobler, eds. Extremophile Fishes. Springer, Dordrecht.
As part of a recent book chapter on rheophilic fishes, I developed a conceptual model describing longitudinal shifts in evolutionary vs. ecological processes controlling the composition of rheophilic fish assemblages in large tropical river basins. Much of my research in the coming years will be guided by predictions of this model.