Habitat - Riparian
Date Published: August 2008
Number of Pages: 53
Author(s): Aimee P. McIntyre, Marc P. Hayes, and Timothy Quinn
A report submitted to the Landscape and Wildlife Advisory Group, Amphibian Research Consortium, and the Cooperative Monitoring, Evaluation, and Research Committee.
Between June 2004 and August 2006, we conducted an assessment to determine whether enough non-fish-bearing basins with the appropriate characteristics were available to populate study blocks in the proposed Type N Experimental Buffer Treatment Study. As originally proposed, this research study was to comprise a total of 20 sites, with five blocks of four different riparian buffer treatments (including reference sites) distributed across three physiographic regions. Site-selection criteria (limits on geographic range, elevation, gradient, lithology, basin size, stream order, and stream network geometry) were developed and basins meeting those criteria were identified using Geographic Information Systems (GIS). Cooperation was established with landowners, and ownership criteria (e.g., stand age and harvest timing) were determined. We field-verified the criteria identified by GIS and specified by landowners, the presence of target amphibian species, and fish endpoints (Type F/N boundary). A database was developed to manage this information and assist in the selection of suitable basins, resulting in the final selection of basins and grouping of those basins into blocks.
We identified a pool of 35,957 non-fish-bearing basins in western Washington using the Washington Department of Natural Resources (WADNR) GIS hydrolayer recently updated to include a last-fish point based on a GIS-logistic regression model. Application of site-selection criteria using GIS reduced this pool to 6,125 qualifying basins from the target regions: the Olympic Peninsula, the Willapa Hills, and the South Cascades. Ownership criteria further reduced this pool to 496 qualifying basins. Field verification of GIS and ownership criteria then reduced the pool to 131 basins. Of these, amphibian surveys were conducted at 115 basins to verify presence of Forests and Fish (FFR)-designated stream-breeding amphibian species (coastal tailed frog [Ascaphus truei], and three species of torrent salamanders [Rhyacotriton]). We used tailed frog detection as an indicator of the presence of stream-associated FFR taxa because presence of the coastal tailed frog is highly correlated with presence of the other target amphibian species. Coastal tailed frog was found in 58 basins of the 115 surveyed basins, roughly half of the basins in which amphibian surveys were conducted. After field-verifying the location of fish endpoints at these sites, we determined that 26 of the basins from the remaining pool were either smaller or larger than the basin size defined in our site-selection criteria, leaving 32 basins remaining.
Fourteen of the remaining 32 basins were unavailable for inclusion after final consideration of landowner and logistic requirements. The remaining 18 sites were available for inclusion in the study, enough to populate four of the five proposed study blocks. Results of a power analysis support three blocks as being the minimum sufficient for implementing the amphibian portion of the study. Reduction in the number of study blocks from five to four will slightly reduce the power to detect a difference among blocks for the amphibian demography portion of the study; however this change is insignificant in context of the overall study design.
Hydrological monitoring was restricted to two study blocks because monitoring weirs could not be placed in streams that lacked a suitable site. Reduction in the number of blocks will reduce the sample size and the power to detect a change, but hydrology data is continuously recorded which will increase measurement accuracy and resolution; and its analysis will not depend on the repeated measures ANOVA model proposed for other response variables. The subsequent reduction of within-site variability will increase the power to detect a significant difference between treatments if a difference exists, at least partially offsetting the loss of replication.
Only eight basins (two blocks) had the stream length (from fish endpoint to downstream tributary junction) needed to conduct the fish-sampling portion of the study. As a result, the fish portion of the study cannot be included in the repeated measures ANOVA analysis. However, the assessment of the response of fish, as a series of case studies, will still provide insight into fish response under differing treatment conditions.
McIntyre, A. P., M.P. Hayes, and T. Quinn. 2009. Type N Feasibility Study. A report submitted to the Landscape and Wildlife Advisory Group, Amphibian Research Consortium, and the Cooperative Monitoring, Evaluation, and Research Committee. 48pp.