Interactions Between Physical and Biological Processes in Riverine Landscapes VII: Ecohydraulics C
Presiding: J M Buffington, USDA Forest Service; C V Baxter, Colorado State University; A E Rosenberger, University of Idaho and USDA Forest Service
B52A-01 10:30h
Large wood dynamics and biophysical consequences for riparian forests: A comparison of an unconfined alluvial river in a temperate rainforest and a bedrock confined river in a semi-arid South African savanna.
Large wood shapes the geomorphology and ecology of rivers. We determined the origin, distribution, and fate of large wood in two rivers from contrasting environments. The Queets is an unstable temperate, rainforest river running from the Olympic Mountains (USA) through a glacial valley with colossal trees. In most years, the channel erodes a variety of forested landforms which forms jams that sculpt habitats. Many are displaced in a few years. Remaining jams initiate landform development and forest renewal. Thus, wood is stockpiled in the floodplain where it may become buried. Channel movements recapture most logs within 50 years. In contrast, the Sabie is a perennial river running through a confined bedrock channel in a fire-prone semi-arid South African savanna. Riparian trees are relatively small and many sink in water. A recent flood (February 2000) devastated the riparian forest, introducing wood to the channel. Jams formed on toppled trees, transported logs, and bedrock outcrops. Many trees survived and resprouted. Jams facilitated the establishment of woody plant seedlings and the intrusion of fire into riparian areas. Sunken wood formed unique depositional features. The Queets and Sabie rivers are strikingly different systems. However, large wood appears to promote the renewal and development of complex riparian forests in both rivers.
B52A-02 10:45h
Characterizing Dimensions and Abundance of Channel and Riparian Coarse Wood in Rocky Mountain Watersheds: Consequences of Spatial Variation
Coarse wood is widely recognized as a critical component of aquatic environments in forested ecosystems and fundamental to fish habitat formation, diversity, and stability, as well as sediment dynamics, hydrologic response, and channel complexity. Its significance has resulted in prescriptions for coarse wood abundance in streams, yet few studies have addressed the spatial variation in coarse wood counts and volumes, with many relying on samples taken from reaches of a few hundred meters. To examine these patterns and relations at the scale of individual pieces and in consecutive 50-m reaches, we continuously censused channel large wood and at 250-m intervals sampled riparian large wood along 1- to 5-km reaches in 13 western Montana basins with varying disturbance histories. Although most coarse wood in channels appeared to originate from riparian zones, frequency distributions of dimensions of channel coarse wood compared to those of riparian pieces indicated substantial in-stream processing. Despite apparent transport of coarse wood shorter than bankfull width in nearly all streams, frequency distributions of counts and volumes of channel coarse wood tended to be normal. Surprisingly, in most streams there was no correlation among counts or among volumes of wood in adjacent 50-m reaches, and although correlograms revealed large-scale coarse wood patches in several streams, patch size was unique to each stream. Finally, relations in piece counts (and volumes) between channels and riparian zones were erratic. Sample sizes necessary to estimate mean length and maximum diameter of individual pieces within 25% of the mean were modest (fewer than 30 pieces in most basins), whereas estimating piece volume required sampling over 150 pieces. To characterize coarse wood abundance with similar precision on average required sampling over 1.3 km of stream; to characterize volume required sample sizes exceeding 2.7 km. Consequently, we regard estimates from short stream reaches as unreliable predictors of total coarse wood abundance. Similarly, models predicting coarse wood abundance based on such estimates likely underestimate variation and may yield unrealistic results.
B52A-03 11:00h
Relating Hydrogeomorphic Attributes to Nutrient Uptake in Alluvial Streams of a Mountain Lake District
Stream form and hydrologic processes may indirectly drive nutrient uptake, however developing predictive relationships has been elusive. Problems in establishing such relationships may lie in the sets of streams analyzed, which often span diverse channel-sizes, geology, and regions, or are too geomorphically similar. We collected field data on stream geomorphology and hydrologic and nutrient transport processes using solute injections at 22 alluvial stream reaches in the Sawtooth Mountains, Idaho, USA. Many of these streams occur near lakes, which create contrasting fluvial form and functions that we hoped would produce a broad geomorphic dataset to compare to hyporheic and dead-zone transient storage and NO3 and PO4 spiraling metrics. Preliminary results suggest that storage zone residence time (Tsto) was best predicted by sediment D50, wood abundance (CWD), and discharge (r2=0.84, p<0.01) and relative hyporheic zone size was best related to D16 and channel width/depth ratio (W/D) (r2=0.48, p=0.05). Among site variation in NO3 uptake velocity (Vf) was best explained by CWD, W/D, and Tsto (r2=0.71, p=0.05) and PO4-Vf was best explained by bed mobility and Tsto (r2=0.96, p<0.01). If and by what mechanisms these models represent true geomorphic drivers of stream transient storage and nutrient cycling processes should be further considered and investigated.
B52A-04 11:15h
Linking Watershed Land-Use, Stream Geomorphology, and Aquatic Biodiversity in a Hierarchical Classification Scheme
Twenty-five streams in northwestern Vermont spanning a range of geomorphic conditions were surveyed to determine the linkages between watershed land-use, stream geomorphology, and aquatic biodiversity as part of an overall watershed classification project. Watershed land-use was evaluated in a geographic information system using Landsat TM land cover data. Field-scale geomorphic assessments included rapid geomorphic assessments (RGA) and rapid habitat assessments (RHA) following Vermont Department of Environmental Conservation protocols. Quantitative field data was collected using detailed site surveys including cross-sections, longitudinal profiles, and pebble counts. Fish, bird and macroinvertebrate data were collected and diversity indices calculated to assess site biodiversity. As a first step, the ability of site-level geomorphic data to predict ecological condition was evaluated. The RGA and RHA were found to be reliable in explaining some of the variance seen in the biotic data. At the next level, the watershed-level characteristics were evaluated to assess their ability to predict geomorphic condition and biotic integrity. Finally, all factors were combined into a hierarchical model identifying the most significant factors at each level used to predict ecological condition.
http://www.cem.uvm.edu/~hession/epa
B52A-05 11:30h
Changes in Salmon Spawning Habitat Distributions Following Rapid and Gradual Channel Adjustments in the Cedar River, Washington
Anthropogenic controls on rivers such as dams, hardened banks, and land uses limit the interactions between main river channel and floodplain ecosystems and contribute to decreased habitat diversity. These system controls dampen the frequency and magnitude of natural disturbances that contibute to physical habitat structure and variability. Under natural and altered disturbance regimes river systems are expected to exhibit resiliency. However, in some cases, disturbances cause fluctuations in the trajectory of the mean system state that can have implications for river recovery in the short- and long-term by changing the spatial and temporal dimensions of available habitat relative to specific biological requirements. Historic and contemporary salmon spawning data are analyzed in the context of changing disturbance regimes in the Cedar River, Washington. Historic data are presented for active channel conditions and spawning fish distributions. Contemporary data are presented for an intensively studied reach that received a landslide that deposited approximately 50,000 m3 of sediment in the main channel, temporarily damming the river. Biologically, the spatio-temporal spawning distributions of Chinook (Oncorhynchus tshawytcha) and sockeye (O. nerka) salmon responded to modifications of physical habitat.
B52A-06 11:45h
Studying coupled hydrological and micro-biological processes by means of tracer injections and mathematical models
To throw light on coupled hydrological, chemical and microbiological processes in treatment wetlands, this study uses both radioactive water and reactive tracers. A tracer mixture consisting of tritiated water, P-32 in the form of PO4- and N-15 in the form of N2O was injected to the 2.6 hectare large Ekeby wetland, Sweden. From the breakthrough curves of tritium, the mean residence time of water in pond 1 can be estimated to be about 3 to 3.5 days. The total injected activity of phosphorus was 17.98 GBq and about 13.73 GBq was recovered at the outlet during the investigation period ending 10 days and 16 hours after the start of the injection. This implies that 24% of the phosphate solution was removed in the November - December period in which the experiment was performed. The total injected amount of N-15 was 42.1 grams and 29.6 grams was retained at the effluent. This means that 30% of the nitrogen was either retained in the wetland or removed due to denitrification. An analysis of regular monitoring data shows that the annual removal rate in the entire wetland (each flow line passes two ponds in series) is about 50% for total phosphorus and 25% for total nitrogen. Probably, the most important mechanism for this removal is adsorption onto particulate matter and deposition. Analyses of vegetation material indicate that a certain (minor) fraction was adsorbed to submersed and emerging macrophytes, like Elodera Canadensis, Thypa sp. (Cattail) and Glyceria sp. (Manna grass). A 2D mathematical model for both water flow and solute transport could explain the N-transport through the wetland. The model accounts for the rate-limited exchange with bed sediments and denitrification in the water and bed sediment. Independent batch tests indicate a particularly high microbiological activity in the bed sediments. The rate-limited exchange with the bed limits also the denitrification capacity of the wetland.