8th Annual Review Abstracts

Interactions of Salmon, Bear and Riparian Vegetation in River Corridors of the Pacific Northwest

Interactions between Pacific salmon (Oncorhynchus spp.) and bear (Ursus spp.) may play an important role in influencing the stature and dynamics of riparian vegetation. It has been known for some time that anadromous salmon returning to their natal streams make significant contributions of marine-derived carbon and nutrients to the aquatic ecosystem. These inputs enhance productivity at various trophic levels within aquatic food webs, but they may also fertilize riparian vegetation. Preliminary data suggest that marine-derived nutrients accumulate in riparian vegetation, and that this enrichment is enhanced where the influence of bear is strongest. Bears may act as an important vector for moving nutrients from the stream to the riparian zone through the killing and consumption of salmon and subsequent deposition of carcasses and feces within the riparian forest. Inputs of marine-derived nutrients may enhance growth rates of riparian trees and influence the production of large woody debris, which may in turn affect structural habitat and productivity within the stream corridor.

The objectives of this project are to determine the extent to which riparian vegetation acquires marine-derived nitrogen from spawning salmon; to assess the impacts of marine-derived nutrients on riparian growth and community composition; and to assess the relative importance of bear activity (e.g., carcass and feces distribution) and hyporheic flow as pathways for transfer of nitrogen (N) from stream to riparian systems. The principal methods used in the evaluation of these objectives are the measurement of stable isotopes (15N/14N) and experimental manipulation of salmon carcasses and bear feces. The significance of this research lies in the discovery of how synergistic interactions between salmon and bear can act to influence riparian forests, and how materials from those forests act to influence the strength of salmon populations.

James Mark Helfield
Center for Streamside Studies
College of Forest Resources
Box 352100
University of Washington
Seattle, WA 98195-2100
(206) 685-8658
helfield@u.washington.edu

Juvenile Salmonids (Oncorhynchus spp.) in Off-Channel Habitats of the Queets River, Olympic Peninsula

The decline of Pacific salmon populations this century is often attributed to loss of freshwater habitat in the Northwest. In rivers subject to channel simplification and manipulation of natural flow regimes, the isolation of the channel from its floodplain reduces availability of and access to a diverse array of aquatic habitats lying off the main channel. Evidence and observation suggest that such habitats can serve as important summer rearing areas for juvenile salmonids (Oncorhynchus spp.). This study seeks to investigate off-channel habitat use by juvenile salmonids during the summer low-flow period in the unchannelized Queets River on the west coast of the Olympic Peninsula. Ultimate objectives include development of a classification scheme for off-channel habitats based on measurable habitat variables, and subsequent correlation of habitat type with juvenile salmonid standing crop, growth and survivorship. A number of these habitats were surveyed during the 1997 low-flow period using a combination of electrofishing and habitat characterization methods. Preliminary data indicate patterns in species composition, size, and abundance of juvenile salmonids among the different habitat types. These patterns can be related to differences in key habitat variables. Focused research on characteristics of habitats used by juvenile salmonids is critical to stream restoration efforts and mitigation or prevention of habitat loss. Treva Coe
School of Fisheries
University of Washington
Box 357980
Seattle, WA 98195-7980
(206) 616-9421
tcoe@u.washington.edu

In-Stream Factors Controlling Juvenile Chinook Salmon Migration: Preliminary Data and Analysis Approach

We are investigating the role of in-stream factors (e.g., light, temperature, and flow) on juvenile chinook migratory behavior and survival in the Snake and Columbia River systems. Research on environmental controls of migratory behavior in other species or in other areas provides evidence that these factors may be important. In the Spring of 1997, the National Marine Fisheries Service (NMFS) conducted a radio-tagging experiment on hatchery chinook salmon in the Grande Ronde River. During the period of out-migration, data were also collected on in-stream conditions at nine monitoring stations.

Our data and related data collected from NMFS and U.S. Geological Surveys will be combined to understand the influence of environmental variables on fish behavior. Analysis will be completed in three stages. First, the relationship between flow and in-stream conditions will be assessed over the sequence of monitoring sites. Second, we will test whether in-stream conditions have significant effects on migration patterns in the Grande Ronde River. Third, we will evaluate the degree to which observed relationships can be applied to other free-flowing and managed rivers and to other species.

Ashley Steel
Quantitative Ecology and Resource Management
University of Washington
Box 351720
Seattle, WA 98195-1720
(206) 616-9440
ashley@cqs.washington.edu

Influences of Streambed Topography and Subsurface (Hyporheic) Flow on the Chum Salmon Egg Pocket Environment

Experiments conducted in laboratories and artificial stream channels have demonstrated that dissolved oxygen (DO) decreases with increasing fine sediments, and may reach levels low enough to harm Pacific salmon embryos incubating in subsurface gravels. Field studies conducted in natural streams, however, have identified much unexplained variability in DO, survivorship of salmon embryos, interactions between fine sediments and DO, and success of embryos exposed to fine sediments. Much of this variability may be explained by oxygen-rich subsurface (hyporheic) flow patterns within streams, largely determined by surface bed topography at the habitat unit (pool/riffle) scale. This hypothesis was tested on an approximately 100-m reach of Kennedy Creek during 1996-1997. Piezometers installed in the streambed at various depths and locations were used to measure hydraulic head, permeability, and to sample subsurface water. Permeability, DO, chemical conductivity, and temperature were also measured in natural chum salmon egg pockets and artificial egg baskets installed throughout the study reach. Subsurface DO varies spatially within the study reach, and this spatial variation persists over time. DO within egg pockets/baskets also varies spatially, and is similar to that in the nearby hyporheic zone, except in very shallow pockets/baskets (<10cm from the surface). Hyporheic DO is highest in pool tail-outs and decreases with distance down riffles and glides. High DO observed in very shallow egg pockets/baskets is probably a result of surface water downwelling from directly above. Permeability is highly variable in the subsurface environment, ranging over three orders of magnitude, from very low to very high. Permeability in egg pockets/baskets is mostly high. Fine sediment (<0.85mm) measured in the upper 20cm of bed substrate is unrelated to DO and permeability in the hyporheic zone and in egg pockets/baskets.

Kerry Bauman
School of Fisheries
University of Washington
Box 357980
Seattle, WA 98195-7980
(206) 296-1967
kerry.bauman@metrokc.gov

Influence of Riparian Vegetation
on Surface/Groundwater Interactions

Many of the current challenges in managing and restoring Northwest rivers revolve around riparian land uses. Continuing controversies over the size and management of riparian management zones (RMZ's) emphasize the need for a better understanding of how human actions affect streamside processes. Despite the importance of riparian processes to the maintenance of healthy streams, little research has actually addressed fundamental ecosystem questions about the efficacy of alternative RMZ sizes or management policies.

Surface/groundwater interaction zones are increasingly recognized as a key regulatory landscape feature in stream ecosystems. Processes within these ecotones control the nature and amount of materials delivered from groundwater aquifers to surface channels, thereby regulating stream responses to catchment disturbances or land use alterations. Processing of materials in transport within interaction zones also influences streamwater chemistry giving the stream a second chance to retain and process catchment inputs. Research in this area has focused on the role of subsurface processes such as hydrologic transport, residence time and biogeochemical transformations in determining the pattern and nature of subsurface chemical patterns. Little research has addressed the linkage between riparian surface communities and subsurface processes in alluvial floodplain reaches. If those linkages are strong, understanding them must be a key factor in predicting effects of riparian management on streams and lateral habitats. I will describe how riparian forest structure influences the diversity of subsurface patterns and processes in naturally vegetated floodplain rivers. I will then list hypothetical effects of common riparian land uses such as silviculture, agriculture and urbanization on surface/groundwater ecotones.

Rick Edwards
Center for Streamside Studies
College of Forest Resources
University of Washington
Box 352100
Seattle, WA 98195-2100
(206) 543-3507
ricke@pisces.fish.washington.edu

Influence of Forest Cutting on Nutrients, Temperature and Turbidity in Small Streams in the Hoh River Valley

Stream chemistry and temperature were monitored monthly in West Twin Creek Watershed (58 ha) in the Hoh River Valley, Olympic National Park, and in two adjacent harvested DNR watersheds - Rock Creek (235 ha) and Tower Creek (263 ha), from June 1996 to December 1997. In addition, turbidity was monitored monthly from June 1997 to December 1997. Harvesting was conducted in 1975 and 1981 in Rock Creek and 1987 in Tower Creek. A greater area was harvested in Tower Creek than Rock Creek. The original vegetation in all three watersheds was old-growth forest dominated by western hemlock and Pacific silver fir. Much of forest in the Rock and Tower Creek watersheds probably resulted from either a fire or extensive blow down in the early 1800s. The forest is older in West Twin Creek with the oldest trees being >650 years. Elevation ranges are 170 to 850 m for Rock Creek, 230 to 900m for Tower Creek and 180 to 850 m in West Twin Creek.

There were few differences in stream chemistry among the three watersheds. Stream pH fluctuated more in West Twin Creek (4.5 to 7.5) than Rock and Tower Creeks (6.8 to 8.1). Electrical conductivity was slightly higher in West Twin Creek. However, peaks of nitrate were higher in Rock and Tower Creeks (about 60 ueq/L) than West Twin Creek (32 ueq/L), but in general nitrate and ammonium concentrations were low in all streams. Average stream temperatures in West Twin Creek had much less variability (range 5.9 -11.1 C) than Rock and Tower Creeks (3.5 - 14.5 C). In general, temperatures in Rock and Tower Creeks were very similar. Daily maximum stream temperatures had a similar pattern to the average stream temperatures with greater seasonal fluctuations in Rock and Tower Creeks than West Twin Creek. Maximum temperatures in the summer were higher in Rock and Tower Creeks (15.4 C) than West Twin Creek (12.1 C) and lower in winter (3.7 and 6.0 C, respectively). Minimum temperatures also showed the same patterns with the lowest minimum temperature in Rock Creek (2.5 C) compared to 5.6 C in West Twin Creek. Turbidity was low in all three streams, although the streams in the harvested watersheds did have higher turbidity than West Twin Creek. Rock Creek had a maximum value of 21 NTU in January of 1998. Highest turbidity occurred in winter and lowest turbidity occurred during low summer flow in all streams. The main influence of harvesting was on stream temperature and turbidity, not on chemistry.

Robert L. Edmonds
College of Forest Resources
University of Washington
Box 352100
Seattle, WA 98195-2100
(206) 685-0953
bobe@u.washington.edu

Effects of Acid Mine Effluent
and Waste Rock Leachate on Watershed Ecology

The Alder mine is an inactive and abandoned gold, silver, copper, and zinc mine in Okanogan County, WA. It produces acidic metal-rich effluent that affects the quality of water in a mountain headwater stream. The results of a pre-liminary study conducted in August of 1997 indicate that the concentrations of heavy metals in the stream were above water quality standards and that the density and diversity of benthic macroinvertebrates were less below the mine than above. The evolution of carbon dioxide was elevated in soil and decomposing logs where groundwater was contaminated by acidic mine effluent. The pH of the mine effluent was 4.5 where it flowed into the ground, increased to 5.5 as it flowed towards the creek and reached normal levels (6.5) when it eventually mixed with the water in the creek. The Alder Creek watershed, therefore, is a good site for experiments to test the hypothesis that chemical criteria do not measure the secondary effects of chemical pollution such as bioconcentration and bioaccumulation.

Dan Peplow
College of Forest Resources
University of Washington
Box 352100
Seattle, WA 98195-2100
(206) 524-4652
dpeplow@u.washington.edu

The Function of Riparian Buffers in Urban Watersheds

The riparian buffer serves as a transition from the terrestrial ecosystem to the stream ecosystem. Under natural conditions, an intact riparian buffer serves numerous functions that maintain the integrity of the water quality, hydrology and biology of the stream ecosystem. Numerous studies have shown that riparian buffers can protect the stream n watersheds which are developed for either forestry or agricultural practices. However, in urban watersheds, which have many unique characteristics, the ability of the buffer to protect the stream may be reduced. In general urban streams have more intense land use, increased impervious area, altered hydrology, sediment loads, and sources of pollutants. This talk will focus on the functions a riparian buffer should provide for a stream system, and the functions expected to be provided in an urban watershed.

Many of the examples in this discussion are based on a case study of Rock and Richardson creeks near Portland, OR. Both of the watersheds are currently developed with small farms and housing developments, however because both of these watersheds lie within the urban growth boundary, they will be completely developed over the next 50 years. As part of the planning and development process, Portland METRO is evaluating the current condition of each watershed, with the goal to develop the area with the least impact on the stream system.

The ability of the riparian buffer to protect the stream will be evaluated in terms of which functions the buffer can provide and which functions would be bypassed in an urban watershed. Based on the evaluation of each of the riparian buffer functions, recommendations related to both the effectiveness of a buffer and non-buffer practices will be discussed as they relate to the protection of the stream ecosystem in an urban watershed.

Jenna Leavitt
Civil Engineering
University of Washington
Box 352700
Seattle, WA 98195-2700
(206) 543-6272
jleavitt@u.washington.edu

Seston and the Streamside Forest - The Effects of Logging on the Food Quality of Suspended Particles

Suspended particulate matter (seston) plays an important role in the trophic ecology of flowing waters. In most headwater forested ecosystems, allochthonous inputs are the primary source of seston. However, silvicultural activities in and along the riparian corridor may change the type and abundance of available allochthonous material and may alter the existing level of autochthonous production. The hypothesis involved here is that changes in the riparian canopy significantly alter various attributes of the seston, ultimately compromising the food quality of the trophic base of the stream in question. Baseflow percent organic matter, chlorophyll a concentration, and C:N ratio are the major parameters which will be used to analyze seston food quality.

Scott Elliott
College of Forest Resources
University of Washington
Box 352100
Seattle, WA 98195-2100
(206) 685-8658
elliott@u.washington.edu

Human Influences on Hydrology in the Dungeness River Basin

The Dungeness River flows northward out of the forested mountains and foothills to cross the coastal plain of the Northern Olympic Peninsula on glacial till and outwash. Since the 1895 initiation of irrigation in response to the surprisingly low rainfall patterns in the shadow of the Olympic Mountains, land use decisions have had a profound impact on both the river itself and on the distribution, quantity, and quality of underground and small stream water in the larger Dungeness Valley.

Land use and land cover changes are discussed, the impact of changing land use practices on instream flows are presented, and the implications of these changes for subsurface water quantity and quality are discussed. A model for predicting impacts on subsurface water, developed in 1983 and in use since, is presented and alternative explanations are discussed.

Penny Eckert
College of Forest Resources
University of Washington
Box 352100
Seattle, WA 98195-2100
(206) 543-0102
penelope@u.washington.edu

Post-Disturbance Recovery of Large Woody Debris Abundance and Pool- Formation in Forest Channels

Recovery rate of the pool-forming function of large woody debris (LWD) in forest channels is dependent on both channel characteristics and riparian forest development. Larger stream channels require larger LWD to initiate pool formation, and pool formation is more sensitive to LWD abundance in moderate-slope (0.01-0.04) channels than in low-slope (<0.01) channels. Number and size of trees available for LWD recruitment in riparian forests is a function of species and age. Red alder (Alnus rubra) tends to grow faster than Douglas fir (Pseudotsuga menziesii), western hemlock (Tsuga heterophylla), or red cedar (Thuja plicata) during the first 20 to 40 years. However, red alder typically dies out between age 70 and 90, and conifer species attain greater heights and diameters than red alder beyond age 50. Pool formation by LWD initially recovers more quickly in smaller streams because red alder LWD is of sufficient size to initiate pool formation. Also, for an equivalent change in LWD abundance, moderate-slope channels experience a greater increase in pool abundance than do low-slope channels. Recruitment of LWD that is large enough to form pools in larger channels is delayed when stands are alder-dominated or are densely stocked conifer. Therefore, active management of riparian forests may decrease the time to recruitment of LWD large enough to initiate pool formation.

Tim Beechie
College of Forest Resources
University of Washington
Box 352100
Seattle, WA 98195-2100
(206) 543-5506
Tjbeechie@aol.com

The Influence of Sediment Supply and Large Woody Debris on Pool Characteristics and Habitat Diversity

The influence that large woody debris (LWD) and sediment supply has on pool characteristics and habitat diversity was examined in streams within the Skagit and Stillaguamish Watersheds. The study focused on streams with channel slopes of 0.5% to 3.1%, and bankfull widths between 7.4 m and 53.7 m. Linear regression and analysis of variance were used to assess relationships between sediment supply and LWD and the following habitat characteristics: residual pool depths, pool spacing, percent pool, and habitat diversity. A Brillouin Diversity Index, based on habitat unit area, was used to describe habitat diversity. Sediment supply estimates from mass wasting inventories were used to describe sediment supply conditions within the stream reaches studied.

Primary pool depth was significantly related to basin area and sediment supply (r2 = 0.65, P < 0.001). Large woody debris in combination with the width to depth ratio described the greatest proportion of the variation observed in pool spacing (r2 = 0.69, P < 0.001). As LWD loading increased and the width to depth ratio increased, pool spacing decreased. LWD described a significant amount of that variation (r2 = 0.37, P < 0.001). Large woody debris was also associated with 58% of the pools observed. Pool area was primarily influenced by LWD that was in contact with the summer flow channel, and channel slope (r2 = 0.722, P < 0.001). Habitat diversity was best explained by LWD and substrate size characteristics (r2 = 0.47, P < 0.001).

The results of this study suggest sediment supply estimates based on mass wasting inventories are useful for assessing impacts to channel and habitat characteristics. Results also indicate habitat diversity can be described using a diversity index, and that it appears to be sensitive to morphological conditions. However, both require further study and refinement. Under most conditions, the greatest influence sediment supply has on pool characteristics is the effect it has on pool depth. Large woody debris was negatively related to pool spacing and positively related to pool area and greater habitat diversity.

Kurt Nelson
College of Forest Resources
University of Washington
Box 352100
Seattle, WA 98195-2100
(206) 543-5506
knels@u.washington.edu

Residence Time of Large Woody Debris
in the Queets River, Washington

Instream large woody debris (LWD) provides a number of useful functions in freshwater ecosystems, including sediment and nutrient retention, salmonid habitat enhancement, and stable colonization sites for incipient floodplain vegetation. In this study, increment cores from instream LWD were crossdated against cores from living riparian conifers to estimate the year each LWD piece was recruited to the channel. Debris pieces that were decayed or otherwise incompetent to provide cores were dated using standard 14C techniques. Size and species composition were compared to that of the riparian forests from which the LWD originated. The depletion rate of LWD from the channel followed an exponential decay curve in which most LWD pieces were not more than 30 years old, although several pieces have remained in the channel for 300 years or more. Hardwood species were better represented in riparian forests than as instream LWD, and conifers were better represented as LWD than in riparian forests, indicating that hardwoods are depleted from the channel faster than conifers. An assumed long-term equilibrium between LWD recruitment and depletion suggests that harvesting of large conifers from riparian forests could have deleterious impacts to streams within three decades.

Tim Hyatt
College of Forest Resources
University of Washington
Box 352100
Seattle, WA 98195-2100
(206) 685-8658
thyatt@u.washington.edu

The Basin Wide Delineation of Instream Woody Debris Placement Sites Using a Geographic Information System

The use of digital elevation models (DEM) was explored to determine their usefulness in identifying sites for instream placement of wood. Initial model development was carried out in two watersheds on the Oregon coast. The GIS/DEM method of site selection was successful in reducing the length of streams needing further field verification by 63-80%. Fifty-nine sites were sampled to test the accuracy of the selection method. The selection method overestimated the length of potential sites in 3 out of 29 (10%) sites sampled for overselection and underselected the length of potential sites in 1 out of 30 (3.3%) of the sites sampled to test underselection. There was good agreement between the length of potential wood placement sites identified by the digital elevation data and those identified with previous channel survey information. Further development of the model has been carried out to increase the accuracy of the channel gradient information. The improved model is currently being tested in the Willamette River basin in Oregon.

Barry Thom
Oregon Dept. of Fish and Wildlife
Corvallis, OR
(541) 757-4263 ext. 256
thomb@fsl.orst.edu

Automating Watershed Analysis on GIS:
What Can DEMs Tell Us?

Watershed Analysis is generally a pencil-and-paper process, but the advent of Geographic Information Systems (GIS) allows a more computational approach to watershed analysis. The pencil-and-paper approach has focussed on delineating areas of highest concern for environmental impact, which then have become subject to restrictive or prohibitive regulation. Modern desktop computing allows the application of quantitative process models to digital maps of soil, topography, vegetation, etc., to allow the direct computation of the environmental impacts of any range of management options.

The most informative of these digital maps are Digital Elevation Models (DEMs), the most available of which are the USGS quads that describe topography with a regularly spaced grid of elevations. Common GIS software packages include a range of tools for computation of topographic and hydrologic variables (which can be used to identify watershed processes and hazard areas). These include; steepness and convergence (mass wasting), channel gradient and confinement (channel response), elevation and aspect (rain-on-snow), flow accumulation (LWD size), and topographic shading (stream temperature).

Finn Krogstad
Forest Engineering
College of Forest Resources
University of Washington
Box 352100
Seattle, WA 98195-2100
(206) 685-2091
FKrogstad@aol.com

Reproductive Success of Steelhead in Forks Creek, Washington

Steelhead (Oncorhynchus mykiss) are an important ecological, recreational, and economic resource in the Pacific Northwest. Currently many steelhead populations are threatened due to human activities, including introduction of hatchery stocks to supplement wild populations. Steelhead have complex life histories and are typically assumed to have evolved to environmental conditions of their natal streams. Conversely, introduced fish are thought to be less well adapted to conditions in a new stream. Recently, hatchery steelhead have been introduced to Forks Creek, a tributary of the Willapa River in southwestern Washington, where a wild population of steelhead already exists. The genetic impacts of hatchery fish on wild populations are currently unclear. One of the primary ways hatchery populations differ from wild populations is in the reproductive environment: wild fish choose and compete for mates while humans choose the mating scheme and which fish will be spawned in a hatchery. This creates a different selective regime for hatchery spawners than for fish spawning under natural conditions. As a first step in understanding this hatchery/wild reproductive dichotomy, an analysis of reproductive success of hatchery spawned fish is being conducted using DNA microsatellite markers. Steelhead spawned in a hatchery were sampled. The progeny of these parents were then sampled as pre-smolts. DNA samples are being used to match parents to offspring. This is a nested study within a larger reproductive success study of hatchery and wild steelhead in Forks Creek.

Greg Mackey
School of Fisheries
University of Washington
Box 357980
Seattle, WA 98195-7980
mackeyg@pisces.fish.washington.edu

Use of Microsatellites, Behavior, and Life History Traits to Assess Reproductive Success of Pink Salmon

Reproductive success of Pacific salmon, Oncorhynchus spp. is influenced by the complex interactions of many factors. In the past, the different aspects of reproductive success have been looked at individually with reproductive success defined through indirect means such as courtship and/or spawning observations. The objective of this study is to assess the importance of fecundity, egg size, spawning site characteristics, dominance, density, superimposition, body size, secondary sexual characteristics, and stream longevity to the reproductive success of pink salmon. Reproductive success is to be defined by number of offspring returning to spawn with parentage determined through microsatellites. A path model will be designed that will define statistically the interaction between these factors and their influence on reproductive success. The results of this ongoing four year study will probably differ significantly from past finding due to previous attention being placed on individual importance rather than interactive importance of these factors.

Bobette Dickerson
School of Fisheries
University of Washington
Box 357980
Seattle, WA 98195-7980
bobette@fish.washington.edu

Summer Movement Patterns of Juvenile Salmonids
in Big Beef Creek, Washington

The objective of this study is to examine the movement patterns of stream-dwelling salmonids in a coastal Washington stream. The specific goals are to (1) determine the home range size and movement patterns of individual salmonids, and (2) estimate the growth and survival rates of fish with differing levels of mobility.

During the summer of 1997, fish within a 330 m study section of Big Beef Creek, Kitsap County, were given individual marks using the Photonic tagging technique. Weirs with two-way fish traps were constructed at both ends of the study section to account for migration. Fish movement was determined by direct observation of marked individuals during snorkeling surveys, by collection of fish at the weirs as they emigrate, and by seining. Multiple observations of individual fish have provided a pattern of movement rather than only a total displacement measurement. Individual growth was determined by comparing initial measurements of weight and length with subsequent measurements. The growth and survival of marked individuals will be compared to their movement patterns to determine if there may be a correlation between movement, and growth and survival.

Tom Kahler
School of Fisheries
University of Washington
Box 357980
Seattle, WA 98195-7980
(206) 616-9419
tkahler@fish.washington.edu

Fish Communities with Bedrock and Alluvial Dominated Streambeds in the Willapa Basin

Many Pacific Northwest salmonid populations are in decline. This study examines a possible contributor to this decline that has recieved relatively little attention. This investigation explores how stream substrate affects fish communities and fish movements. Land use practices, log jam removal, debris torrents and other events have changed the proportion of alluvial and bedrock dominated streambeds in the Pacific Northwest. Many streams have become increasingly dominated by bedrock substrate which is often less productive than gravel-rich streams. We are examining the effects of substrate on all vertebrate stream dwelling species found in the streams. Relatively few studies in the Pacific Northwest have examined the interactions between salmonids and other species, although most streams contain one or more non-salmonid species. This work will is being conducted on small (2nd to 4th order) streams in the Willapa basin. The Willapa basin was selected for this investigation because it provided opportunities for collaboration with forest products company researchers and other scientists at the University of Washington and Oregon State University on the geology and botany of the specific stream reaches. Mark and recapture work, using individual tagging techniques, and surveys of the biological and physical characteristics of the stream reaches are being used to study this influence on the productivity and diversity of fish in these streams.

Michael Erickson
School of Fisheries
University of Washington
Box 357980
Seattle, WA 98195-7980
(206) 616-9419
(360) 943-8895
mikee@fish.washington.edu

Experimental Analysis of Salmon Hatchery Supplementation:
A Behavioral Approach

Every year Washington State salmon hatcheries release hundreds of millions of juvenile salmon into streams to supplement natural salmon production. However, evidence that hatcheries have contributed to the decline of natural populations has left many skeptical of this technology. In addition to the concerns that hatchery populations introgress with wild populations reducing genetic diversity, it has been hypothesized that hatchery juvenile salmon compete with naturally rearing fish after they are released in streams.

Field and aquarium experiments were designed to compare the performance of hatchery-reared coho juveniles with naturally reared coho juveniles over the summer. Performance was measured as survival, movement, growth and competitive ability. The hypothesis that naturally-reared coho out-perform hatchery-reared coho was rejected. Hatchery-reared coho and naturally-reared coho survived in equal proportion in the field experiment, they moved in similar patterns, and evidence suggested that hatchery reared coho grew at faster rates. Hatchery-reared fish dominated naturally-reared fish when size matched in aquarium experiments and superior competitive ability might have accounted for their faster growth rates in the field. Naturally-reared coho had the competitive advantage of prior territorial residence, however aquarium experiments indicated that a prior residence advantage is overcome by a size advantage and/or high levels of aggressive behavior induced by winning experience. If managers decide to supplement natural populations of coho with hatchery-reared parr they should not assume that the hatchery-reared fish are inferior. A simulation model suggested that there will be a greater trade-off in recruitment of naturally-derived fish as the number of hatchery fish planted is increased and/or as the size advantage of hatchery fish is increased and/or as carrying capacity is approached.

Justin Rhodes
School of Fisheries
University of Washington
Box 357980
Seattle, WA 98195-7980
jrhodes@fish.washington.edu

Stream Corridor Restoration: Principles, Practices and Processes -

A U.S. Government Interagency Reference Currently Scheduled for Release in Summer 1998

It is widely acknowledged, both nationally and internationally, that that many of the values of stream corridors are being and have been lost through uninformed or misguided actions on waterways and watersheds that are nourished by them. With recognition of this and increasing evidence of stream corridors as critical ecosystems, sixteen agencies of the U.S. Government began cooperating in 1994 in an unprecedented effort to develop a document of restoration technology to serve as common reference for federal work in stream corridors. These agencies have undertaken a four-year effort to develop this reference to aid in developing stream corridor restoration projects. It is a reference only-not a policy document of the cooperating agencies.

The document is divided into three principal parts and an Appendix:

Part I provides background on the ecological structure and functions of the stream corridor. It presents the many natural processes that shape the corridor and lead it to a state of dynamic equilibrium. The discussion concludes with an overview of the causes and symptoms of stream corridor degradation.

Part II focuses on identifying and explaining a general restoration plan development process. It includes a discussion of the fundamental steps involved in planning and implementing stream corridor restoration: getting organized, identifying problems and opportunities; developing goals and objectives; selecting and designing restoration alternatives, and implementation, monitoring and evaluation.

Part III examines the issues associated with the how to of implementing a restoration plan. This section uses the information developed in the previous sections to consider how knowledge of the riparian corridor and its surrounding system and the results of a tailored planning process can be applied in a restoration initiative. This section includes information on both passive and active restoration options.

The Appendix is a collection of fact sheets describing various restoration techniques organized around the categories of: instream practices, bank treatment, backwater management, channel reconstruction, stream corridor measures, discharge manipulation, and watershed management practices. The development of these fact sheets was a major contribution of the NRCS Watershed Science Institute.

Carolyn Adams
Watershed Science Institute of the USDA Natural Resources Conservation Service
26 Johnson Hall
University of Washington
Box 351310
Seattle, WA 98195-1310
(206) 616-5724
houston@forsterite.geology.washington.edu

Seasonal and Successional Variation in
Soil Nitrogen Flux in a Floodplain Forest

In the Pacific Northwest, nitrogen (N) fixation associated with red alder (Alnus rubra) leads to rapid N accumulation in recently deposited riparian soils. In order to gauge the potential for N exchanges between riparian soils and the aquatic ecosystem, seasonal and successional variation in retention and loss of soil N was studied on the Queets River floodplain. Soil and groundwater nutrient concentrations were sampled monthly in five patches of differing soil and vegetation successional ages.

Seasonal variations in N concentrations in groundwater and soil solution were observed at all sites, with a sharp peak in N occurring in the fall. Although the old growth sites had the highest total soil N, groundwater and soil solution N concentrations (nitrate and dissolved organic N) were much higher in in the alder dominated sites, with highest concentrations in the oldest alder sites. Total soil N increased rapidly during the first 40-70 years of soil development. Accretion of N and C was strongly related to soil particle size, which in turn reflected fluvial processes. Some stable organic N is being imported associated with fine inorganic sediments during floods. Retention of indigenously produced N also appears to be related to the availability of adsorption sites on fine sediment particles.

Scott Bechtold
College of Forest Resources
University of Washington
Box 352100
Seattle, WA 98195-2100
(206) 616-9421
sbech@u.washington.edu

Streamflow and Temperature Issues in Headwater Streams

Headwater streams are receiving increasing attention from policy makers and regulators in Washington. In large watersheds, up to 80-90% of the river system may consist of headwater streams. Their number far outweighs the small size. Anything that affects these streams, either from a management or ecological perspective, can have far reaching effects.

Many riparian functions are affected by alterations to the areas surrounding headwater streams. However, much less is known about headwater streams than larger streams. Unanswered questions include, where does the water come from in headwater streams, how does heat get transferred to headwaters streams, can harvest increase groundwater temperature, do RMZ's meet management objectives on larger streams, and what is the reliability of current stream temperature models when used on headwater streams? An on-going study on the Olympic Peninsula attempts to address some of these issues. Mid-summer stream source locations, types, discharges and temperatures were monitored. Information is being collected for comparison between pre- and post- harvest conditions and for paired harvested and unharvested streams.

Jenelle Black
College of Forest Resources
University of Washington
Box 352100
Seattle, WA 98195-2100
blackjs@u.washington.edu

Measurement of Forest Harvest Effects
on Rain-on-Snow Processes

Despite considerable evidence that clearcut forest harvesting can increase water delivery to the soil during rain-on-snow and spring-melt events, little is know about the effect of alternative forest harvest prescriptions (such as green tree retention) on such events. To better quantify the hydrologic response of forest harvesting over a wide variety of harvest prescriptions, measurements of snow accumulation, melt, and micro-meteorological are being taken in existing mature forested units and shelterwood units in the Umpqua National Forest. Data are collected in the transient snow zone of the western Cascades of southwestern Oregon as part of the Demonstration of Ecosystem Management Options (DEMO) project. Pre-harvest data have been collected from 1994 to 1997. In the summer of 1998, all units (excluding one control) will be harvested at different retention levels and additional years of data will be collected. Micrometeorologic measurements include incoming and reflected shortwave radiation, incoming longwave radiation, wind speed, relative humidity, air temperature, and precipitation. Snow accumulation and melt at the unit scale are measured weekly on a twenty point snow course. Point observations of snowpack outflow consists of two 2.6 m2 non-weighing snow lysimeters in each unit.

Analysis of pre-harvest data suggests that our experimental design is adequate to assess differences in seasonal snow accumulation and event outflow among harvest units. However, the data collected are insufficient for rigorous testing of models of snow accumulation and ablation. Weekly snow courses yield little information on snow interception processes that control snow accumulation under the canopy. Furthermore, a large number of small lysimeters can only characterize each harvest unit in an average sense; they say nothing about processes at the scale of an individual tree's canopy. Since canopy throughfall is highly variable over small spatial scales, any estimate of snowmelt (outflow minus throughfall) measured by small lysimeters may not be representative of the canopy as a whole. To overcome these limitations, we began a pilot study during the winter of 1996/97 to extend the on-going field program. To more adequately investigate processes affecting snow accumulation and ablation under a mature forest canopy, two 25 m2 weighing lysimeters were constructed. Each lysimeter was installed around an individual mature (approximately 40 meter tall) Douglas-Fir tree. The platform is weighed every minute by four 5-ton load cells. Outflow from the lysimeter is routed to tipping bucket volumetric flow gages for measurement.

The canopy collectors generate high quality data regarding snow accumulation and ablation under the forest canopy. However, they cannot, by themselves, provide any information on the processes controlling snow interception and mass or meltwater release to the ground snowpack. To more fully investigate snow interception, one slightly smaller (12.5 m 2 ) weighing lysimeter was placed in a nearby shelterwood. Since this "open lysimeter" is not influenced by canopy interception, snow interception and release by the canopy can be inferred by comparing the changes in snow water equivalent of the canopy lysimeters relative to the open lysimeter. Data from all three collectors can be used to quantify differences in snow pack water equivalent for time scales ranging from individual events to the entire snow season.

To measure snow interception directly, two lysimeters capable of direct weight measurements of individual trees were installed in a nearby forest plantation (~20 years old). Each of these "tree-weighing lysimeters" contains a 6 m2 weighing lysimeter constructed around the stump of an existing plantation tree. The stump is left to protrude though the ground pack lysimeter. Tree weight is measured by a device containing a 1000 lb load cell which is anchored to the existing stump. By employing a design which weighs the ground snowpack separately from the snow intercepted on the tree, direct measurements on the fate of intercepted snow can be made. For example, snow falling from the tree will register with a decrease in tree weight and an increase in ground snowpack weight. Drip of meltwater off the tree will register first as a decrease in tree weight with a simultaneous increase in snowpack weight and subsequent measurement of lysimeter outflow.

Susan Bolton
College of Forest Resources
University of Washington
Box 352100
Seattle, WA 98195-2100
(206) 685-7951
sbolton@u.washington.edu

Urban Stream Rehabilitation in the Pacific Northwest:
Physical, Biological, and Social Considerations

Last spring, the Center for Urban Water Resources Management was awarded a three-year grant from the joint National Science Foundation/U. S. Environmental Protection Agency interdisciplinary program, "Waters and Watersheds." Faculty and graduate students in several departments across campus are involved: Derek Booth, Stephen Burges, Chris Konrad, and Marit Larson (Civil Engineering); Sally Schauman, Sandra Salisbury, Karen Billica, and Cory Parker (Landscape Architecture); and James Karr and Sarah Morley (Department of Fisheries). Our overall goal in this project is to document the consequences of urban development on the physical and biological condition of urban streams and to use that knowledge to demonstrate specific rehabilitation strategies likely to restore valued properties of those systems.

Since the beginning of the project in April 1997, we have accomplished several tasks:

1. Site selection for examining process and elemental changes and cultural context of stream degradation. We have selected eighteen sites that span a gradient from low to high urbanization. Since highly urban streams exhibit cumulative changes in most every physical and biological condition, our sites emphasize the lower end of this spectrum, where degradation can be seen incrementally, in an effort to define critical causal mechanisms that initiate degradation. Physical, biological, and social conditions have been characterized for all sites, though there are areas of special emphasis for evaluating processes of stream degradation.

2. Evaluation of visual preferences for rehabilitation design and human behaviors exhibited toward urban riparian areas and rehabilitation designs. One masters' thesis has already been completed on the preliminary evaluation of visual preferences for typical restoration techniques. In addition, an initial scoping study of observed behaviors towards urban riparian systems has been made to provide us with an initial set of terms that describe human interactions with urban streams.

3. Assessment of rehabilitation project costs and outcomes. Although this topic is a primary focus only in subsequent project years, we are beginning to compile the available data on biological and physical characteristics of stream channels, before and after construction of stream-rehabilitation projects, in order to start quantifying benefits and to identify where additional targeted monitoring in Year 2 of this project will be most beneficial. In subsequent years of the project, we look to apply an improved understanding of degradation processes to the very tangible problems of minimizing urban consequences on natural stream systems, long-term, with only limited available resources.

Derek Booth
Center for Urban Water Resources Management
University of Washington
Box 352700
Seattle, WA 98195-2700
(206) 543-7923
dbooth@u.washington.edu

Bioavailability of Dissolved Organic Carbon in the Hyporheic Zone of the Queets River, Washington

Heterotrophic bacteria living in hyporheic (saturated subsurface) sediments of rivers require allochthonous sources of organic matter for growth and reproduction. In floodplain rivers of the Pacific Northwest, channel inputs via downwelling and lateral inputs from riparian soils are the primary sources of organic matter for microbial production. The lability of this organic matter is determined by the quality of the source and the degree of processing that occurs as the material travels along flowpaths. This research is a preliminary study investigating the potential of carbon limitation on epilithic productivity in the hyporheic zone of floodplain rivers. The goals of this study are: 1) describe the chemistry of dissolved organic matter (DOM) in the hyporheic zone, 2) quantify changes in the bioavailability of DOM along a flowpath, and 3) evaluate the bioavailability of DOM from riparian soils for bacterial growth. Hyporheic water from several locations on the floodplain and leachates of riparian soils of varying ages were analyzed for dissolved organic carbon and carbohydrate concentration. Initial results indicate differences in the amount of carbohydrate entering from channel and riparian sources.

Sandra Clinton
College of Forest Resources
University of Washington
Box 352100
Seattle, WA 98195-2100
(206) 616-9421
sclinton@u.washington.edu

Watershed Scale Relationships Between Land Use
and Water Quality

Human land use alterations are the major source of disturbance in most catchments. Predicting long-term consequences of population growth and basin development requires techniques that can integrate complex combinations of individual small scale process impacts over entire basins or ecoregions. Process-based hydrologic/water quality models yield insights into the effect of specific processes alterations, but application of such models over large time and space scales is expensive and problematic. Empirical non-point source models that relate basin-scale landscape trends to water quality offer promise, but all require assumptions about the scale and grain of those relationships. The most fundamental problem is determining the valid upstream area within which aggregated landscape properties exert a measurable influence over the variable of interest at a specific downstream location. Past basin-scale studies suggest that landscape-water quality relationships change systematically with position along the drainage. We are developing a simple landscape-water quality modeling approach that can be used to empirically test which portions of the upstream landscape best explain measured water quality trends. As well as offering a more systematic approach to model development, this approach may be used to explore the potential effectiveness of alternative stream buffer sizes for stream restoration and management.

Rick Edwards
College of Forest Resources
University of Washington
Box 352100
Seattle, WA 98195-2100
(206) 543-3507
ricke@u.washington.edu

Effects of River Regulation on Downstream Biota

There are over 75,000 dams on rivers in the United States, and almost every major river in North America is regulated in some manner. Despite these facts, little is known about how the synergy of modified flow and temperature ultimately influences the ecology of river systems in the Pacific Northwest. In free-running Western Washington rivers, channel morphologies are largely shaped by peak flow events. Sediment aggradation and scour shape a continuum of instream habitat which provide a template for the biotic features of the river. In regulated waters, habitat heterogeneity is often reduced, because the transport capacity of the river is compromised by diminished peak flows. Furthermore, the regulation of rivers often alters the temperature regime downstream, thereby affecting the growth rates and life histories of the biota downstream. As a result, the abundance and composition of food available to fish as drift may also be modified by regulation. I intend to determine how the manipulation of discharge and thermal regimes on the South Fork of the Tolt River is manifested on the populations of resident rainbow trout, Oncorhynchus mykiss.

The Tolt River in Western Washington offers a unique opportunity to quantify the effects of river regulation in the Pacific Northwest. While the North Fork of the Tolt is free-running, the South Fork has been dammed to form a reservoir for Seattle Public Utilities. Two study sites have been identified on each fork to provide within and between fork comparisons of (i) the density and composition of invertebrate drift, and (ii) the growth rates, condition analysis, population size estimates, and stomach contents of resident salmonids.

It is unrealistic to think that a river can be dammed without adversely influencing the continuity of the watershed, but it may be possible to minimize those effects once they are identified. This study has direct implications in the management and protection of Western Washington's wild resident rainbow trout and presmolt steelhead. As the number of dammed rivers in the United States increases, a growing percentage of the stream habitat available for salmonids exhibits altered water temperature and discharge regimes. It will be in the best interest of fishery managers to be able to anticipate the effects of river regulation on the abundance and composition of invertebrate drift, and salmonid growth rate.

Jamie Glasgow
School of Fisheries
University of Washington
Box 357980
Seattle, WA 98195-7980
(206) 543-7443
glasgow@fish.washington.edu

Environmental Limitations to Vegetated Stormwater Biofilters

To combat nonpoint source water pollution federal and local governmental agencies throughout the country have sanctioned construction of low cost, free water surface (FWS) stormwater filtration facilities typically built conjoined or adjacent to retention/detention ponds. Biofiltration swales (also called bioswales or biofilters) are open channels possessing a dense cover of grasses and/or wetland plants through which runoff is directed during storm events. Aboveground plant parts (stems, leaves and stolons) physically filter particulates and their associated pollutants as runoff passes slowly through the channel. Herbaceous cover is considered to be well correlated with treatment efficiency (Daniels and Gilliam, 1996; Hawkins et al., 1995; Horner, 1988).

Environmental conditions and construction history were examined for eight biofiltration swales in King County, WA to determine relative importance of and threshold values for the various factors influencing vegetation establishment and growth. Vegetation and plant litter biomass from soil surface to 10cm height (that which is most important to sediment filtration) was well correlated to insolation (r = 0.5, R2 = 0.25), design storm flow velocity (r = -0.53, R2 = 0.31) and mean summer base flow depth (r = -0.53, R2 = 0.28). For the one-third of the plots which possessed mean summer soil moisture potentials less than -15 MPa, vegetation biomass was strongly correlated with soil depth (r = 0.87, R2 = 0.76). Three swales were retrofitted with new soil and hydroseeded in September, 1996. Because limiting factors proved overwhelming for two swales, only one swale supported more abundant vegetation after one year.

A nested two factorial greenhouse experiment tested the response of four turfgrass species (each with three different cultivars) commonly seeded in bioswales to three inundation regimes plus a control. Festuca arundinaceae (Tall Fescue) accrued significantly more biomass than Agrostis alba (Redtop), Poa pratensis (Kentucky Bluegrass) and Alopecurus geniculatus (Meadow Foxtail) over four weeks in three of the four treatments. The treatment in which seeds were inundated 12 of 14 days for two consecutive periods produced equally minimal germination amongst all species. To improve future biofiltration swale design and performance, more consideration to providing seeded grasses with environmental conditions conducive to germination and establishment is recommended. However, in the absence of such consideration, and given the difficulty of proper construction and doubtful functionality, I recommend against the use of bioswales as a best management practice in favor of alternative facilities such as constructed wetlands.

Greg Mazer
College of Forest Resources
University of Washington
Box 354115
Seattle, WA 98195-4115
(206) 685-7494
mazer@u.washington.edu