COS 69-6
The importance of fluvial geomorphology and hydrology in driving the distribution of a native, river-associated species, Carex nudata

Wednesday, August 12, 2015: 9:50 AM
338, Baltimore Convention Center
Matthew Goslin, Geography, University of Oregon, Eugene, OR

Carex nudata (torrent sedge) occurs in rivers throughout northern California and Oregon, and where prominent appears to function as an ecosystem engineer capable of altering river morphology. Previous studies have also found that C. nudata may indirectly facilitate the presence of other species, enhancing diversity. While the species appears to play a key role in river ecosystems, little is known about the drivers of the species distribution. The objective of this research is to determine the environment factors driving this species’ distribution and to test the hypothesis that distribution is driven by hydrological variables (stream power) and fluvial geomorphological variables (sediment size). This objective was pursued through a two-step process.  In the first step, I used herbaria data and a species distribution model designed for presence-only data (Maxent) to construct a range-wide distribution model.  Given the inherit problems in using presence-only data, the primary purpose of the range-wide distribution model was to serve as the basis for a stratified (by predicted probability of C. nudata) random sampling design for field surveys.  In the second step, I conducted basin-wide surveys in two representative basins, the John Day and Santiam basins.  At 30 study sites, I quantified abundance of C. nudata and measured associated environmental variables and river channel metrics.  I used boosted tree regression models with field-measured variables as well as digitally derived variables (bankfull discharge) to model which environmental factors best predict the abundance of C. nudata within river basins.


The Maxent draft model demonstrated that hydrological variables (mean annual discharge, velocity, stream power) dwarfed climate variables in importance in explaining C. nudata.  Data from the basin-wide surveys further elaborated on this model.  Critical variables in C. nudata distribution included stream power, sediment size and canopy cover.  With its strong, dense root systems, C. nudata appears to be a disturbance-adapted species able to survive in stream reaches with medium to higher stream power.  Stream power appears to sets upper and lower limits on the distribution.  In addition to stream power, C. nudata is associated with larger sediment sizes of the channel bed.  The sedge typically occurs in gravel bed and bedrock channels but not in channels with finer sediment sizes.  Finally, C. nudata appears to be tolerant of partial shade but not complete shade.  This suggests an interaction between valley constraint and surrounding vegetation type that limit C. nudata distribution. Constrained valleys coupled with dense forest types limit distribution.