Our research examines the effects of forest management strategies by hcj


									        Effects of Contemporary Forest Harvest Practices on Aquatic Ecosystems:

                                    Trask Watershed Study


       Our goals in the Trask Watershed Study are to understand how aquatic systems,

particularly small streams, respond to forest harvest operations and to compare two management

practices to determine the effectiveness of BMPs. Our overall objectives are to determine:

      The effects of forest harvest on the physical, chemical and biological characteristics of

       small streams;

      The extent to which alterations in stream conditions caused by harvest along headwater

       channels infl Section II uence the physical, chemical and biological characteristics of

       downstream fish-bearing streams.

Co-science leads: Dr. Sherri Johnson, PNW Research, US Forest Service, Dr. Bob Bilby,
Weyerhaeuser Company. Science team: Dr. Jason Dunham, USGS FRESC; Dr. Arne Skaugset,
OSU Forest Engineering; Dr. Michael Adams, USGS FRESC; Dr. Joan Hagar, USGS FRESC;
Liz Dent, Oregon Dept of Forestry; Dr. Judy Li, OSU Fisheries and Wildlif; Dr. David Wooster,
OSU Ag Exp Station; Dr. Stephen Lancaster, OSU Geosciences; Maryanne Reiter, Weyerhaeuser
Company; Doug Bateman, OSU Forest Engineering; Linda Ashke nas, OSU Fisheries and
Wildlife; Bill Gerth, OSU Fisheries and Wildlife; Amy Simmons, OSU Forest Engineering.

                         Table of contents

Section I
Introduction ………………………………………………………………...

Objectives …………………………………………………………………..

Background …………………………………………………………………

Conceptual design ……………………………………………………….....

Researchers and collaborators ……………………………………………

Section II

Study site description ………………………………………………………

Hypotheses and methods ………………………………………………..…


Expected outcomes ……………………………………………………….

Literature cited ………………………………………………………….…

Section III
Contextual Analysis ………………………………………… separate document

Section I
      The Oregon Forest Practices Act (FPA) was the first in the United States to include

provisions for the protection of water quality and fish habitat. These rules, which define Best

Management Practices (BMPs) for commercial forest operations have been evolving since they

were adopted. While they have led to improvements in water quality and fish habitat, questions

still remain about the effectiveness of contemporary BMPs in protecting water quality and

aquatic biota during and after forest harvest.

       The Trask Watershed Study will examine the impacts of forest operations on small,

headwater streams and the extent to which these impacts are transferred downstream. Small

streams are tightly coupled with terrestrial ecosystems, making them potentially more responsive

to forest harvest than larger, downstream systems (Moore and Richardson 2003), and they have

not received the level of protection afforded to fish-bearing streams during forest operations.

These small, non-fish-bearing streams often comprise 80% or more of the stream length in a

drainage network (Benda et al. 2005). Impacts to small streams have the potential to have

downstream repercussions because they transport water, energy, organic and inorganic materials,

and influence the characteristics of downstream aquatic communities and habitats. However,

effects of forest management on headwater streams and the linkages between headwater and

downstream areas have not been quantified for many physical, chemical and biological

parameters (e.g., Reid 1998; Hartman 2004).

       The Trask Watershed Study will be conducted in the headwaters of the Trask River,

which flows into the Pacific Ocean at Tillamook Bay, in northwest Oregon (Figure 1).

Ownership of the upper basis is evenly split between the Oregon Department of Forestry (ODF)

and Weyerhaeuser Company, with a small area managed by Bureau of Land Management. ODF

and Weyerhaeuser Company are providing long term funding for this research, the results of

which will be shared in peer reviewed publications, reports, online datasets and workshops.

Additional funding is being sought through competitive grants. The science teams are

interdisciplinary, with primary involvement by researchers from Oregon State University, US

Forest Service Research, USGS Forest and Rangeland Ecosystem Science Center, and

Weyerhaeuser Company. Frequent collaborations occur with personnel from Oregon agencies

and local groups (see section “Researchers and Collaborators” for more details).

       Figure 1: a) Trask Watershed Study area; b) Watershed studies through the WRC

       The Trask Watershed Study is one of the studies collaborating through the Watershed

Research Cooperative (WRC) at Oregon State University. The three research projects in the

WRC (Figure 1b) are using comparable methods to evaluate the effectiveness of Oregon’s BMPs

in minimizing impacts to water quality, aquatic habitat and fish and insect communities from

forest operations. All three studies involve multidisciplinary teams of scientists from multiple

research organizations conducting research in cooperation with industrial, state, and federal

forest landowners. Each of the landowners are providing substantial financial support and have

adjusted their management plans to accommodate rigorous research designs. The approaches to

evaluating responses in the Trask study are coordinated with those in other WRC studies in an

attempt to provide a comprehensive assessment of the effects of forest harvest. With the coupling

of the empirical work and modeling, we hope to gain an understanding of the relationships

between fishes and forestry that will apply to the other studies and streams throughout the Pacific

Northwest. Cooperation and coordination among these projects increases understanding of

responses of stream ecosystems to forest operations, and ultimately provides a higher degree of

certainty regarding the extension of results to other locations in western Oregon than would be

the case if the studies were conducted independently. Although all three studies are monitoring

hydrologic and sediment dynamics as well as abundances of key food web communities, the

Trask Watershed Study scientists will be extending their research into study of mechanisms and

processes that influence potential responses.


       The goals of the Trask Watershed Study are to understand how aquatic systems,

particularly small streams, respond to forest harvest operations and to compare two management

the physical, chemical and biological characteristics of downstream fish-bearing practices to

determine the effectiveness of BMPs. Our objectives are to determine:

      The effects of forest harvest on the physical, chemical and biological characteristics of

       small streams;

      The extent to which alterations in stream conditions caused by harvest along headwater

       channels influence streams.


       Forest harvest can affect the water balance of a site by removal of overstory vegetation,

which influences the amount of water intercepted or evaporated, the rate of snow accumulation

and melt, and soil moisture storage via evapotranspiration ( ). Peak discharge, shown to have

potential increases after forest harvest, are linked with sediment transport dynamics. Without

vegetation canopy, precipitation has greater impact on soil, which has been shown to result in

disturbances and erosion ( ).

       Forest roads also have the potential to alter the water balance of a site by intercepting and

redistributing subsurface flow. In combination, these processes govern the rate and pathway by

which water is transmitted to stream channels. Thus, when processes such as infiltration,

percolation, evaporation, transpiration, and others are individually and cumulatively affected

following road construction, harvesting, or other forest practices, the way a watershed processes

incoming precipitation is also altered.

       Increases in water temperature of streams are a common impact of anthropogenic

activities. Removal of riparian canopy allows light and energy to reach the stream and increase

temperatures. At higher temperatures, fish metabolism increases and consumption increases, yet

effects of temperatures on biomass of insect prey is uncertain. Most fish growth models of

temperature effects assume that if fish are fed to satiation, they will grow larger at higher

temperatures, up to some optimal temperature. But if insects are smaller at higher water

temperatures, there may be reduced availability of calories to support fish growth.

Conceptual Design

       Stream ecosystems are complex and there are multiple linkages among system

components and processes (Figure 2). Forest removal and associated management can directly

impact hydrology, sediment, nutrients                                 Forest harvest

and incoming energy; these effects have
                                              Stream flow             Sediment           Detritus   Incoming light
                                                                                        and wood     and radiation
the potential to cascade through stream
                                                 Stream chemistry
ecosystems influencing individuals,                                                         Stream

populations, communities, standing
                                                 Macroinvertebrates                Primary
                                                                                  production        Dissolved
stocks, and connectivity at multiple

points (Northcote and Hartman 2004).                     Amphibians

Uncovering the underlying processes           Figure 2: A simplified representation of stream ecosystem
                                               elements and linkages that may be influenced by forest
responsible for observed changes in

system condition is challenging.

       In the Trask Study, we will study the effects of forest operations through a combination

of field observations and experiments in headwaters streams and in downstream fish-bearing

reaches. We are interested in not only documenting responses to forest harvest but also

elucidating processes and mechanisms responsible for the observed patterns. We are measuring

key individual, community and ecosystem processes and rates and developing conceptual models

to identify linkages among ecosystem elements and processes, as well as between the headwater

treatment and downstream effect sites. Experiments will be designed to quantify the key

linkages that can not be determined through field observations.

       Many long-term studies have found wide fluctuations in abiotic and biotic parameters

even in undisturbed systems; this variation can make detection of the effects of disturbance

difficult or impossible. Although we intend to quantify hydrologic and sediment fluxes and

densities of vertebrates and invertebrates, we expect that by focusing on understanding and

quantifying processes, such as water flow paths, rates of growth and feeding, shifts in timing of

life history stages dispersal and drift, we will better understand responses to forest operations.

We also recognize there is inherent landscape variability in both geology and landuse history in

the Trask, and therefore, magnitude of responses of fishes, amphibians, macroinvertebrates,

primary producers, aquatic habitat, water quality and water quantity to forest operations may

vary. We will incorporate this variability by study of processes and will implement models with

these data to evaluate effects of forest operations on abiotic and biotic responses.

Researchers and collaborators

       Trask researchers are an interdisciplinary, multi-organizational group of scientists with

expertise in key areas. Dr. Sherri Johnson, PNW Research, US Forest Service and Dr. Bob Bilby,

Weyerhaeuser Company serve as Co-Science leads. The Science team includes Dr. Jason

Dunham, USGS FRESC; Dr. Arne Skaugset, OSU Forest Engineering; Dr. Michael Adams,

USGS FRESC; Dr. Joan Hagar, USGS FRESC; ; Dr. Judy Li, OSU Fisheries and Wildlif; Dr.

David Wooster, OSU Ag Exp Station; Dr. Stephen Lancaster, OSU Geosciences; Liz Dent,

Oregon Dept of Forestry; Maryanne Reiter, Weyerhaeuser Company; Doug Bateman, OSU

Forest Engineering; Linda Ashkenas, OSU Fisheries and Wildlife; Bill Gerth, OSU Fisheries

and Wildlife; Amy Simmons, OSU Forest Engineering.

       Research planning, sampling logistics and results are discussed at frequent Team

meetings and field trips. Watershed scientists outside of the Trask Study are frequently

consulted for input on study plans or to coordinate field visits to comparable projects. The Trask

Science group also communicates regularly outside of the traditional scientific community;

frequent interactions occur with community groups, including Tillamook County and NGOs

(such as Watershed councils, EcoTrust, Oregon Trout and Tillamook Estuary Partnership); forest

industry representatives (Oregon Forest Industries Council, National Council for Air and Stream

Improvement), participating land owners (Weyerhaeuser Company, Oregon Dept. of Forestry,

and Bureau of Land Management), industry representatives (Plum Creek Timber Company,

Roseburg Timber Company), state agencies (Oregon Dept. of Environmental Quality, Oregon

Dept of Fisheries and Wildlife) and federal agencies (USFS, USGS, BLM). Collaboration and

integration occurs among studies within the WRC in formal and informal venues; frequent

interactions and sharing of results occur in many venues and as a result of overlap of individuals

among projects and during local, regional and national scientific meetings and field tours.

       A major responsibility for Trask researchers is providing data and metadata to networked

databases so that others can use data in years to come. The Trask team has set up shared data

directories in the OSU Forestry Computing Network. Web sites of the WRC and ODF are used

to post study plans, cleaned data and metadata, participant activities and summaries of findings.

Section II

Study Site description

       The experimental design for the Trask River Watershed Study incorporates nested and

paired catchments, where on-site responses are evaluated in headwaters and downstream effects

and linkages with upstream harvest are studied in downstream fish bearing reaches (Figure 1a).

Headwater catchments, of 25 to 50 ha each, within the Trask study area have been grouped into

clusters (Figure 1a). Within each cluster, multiple catchments will be treated and onsite and

downstream effects studied. One catchment in each cluster will not be harvested so as to serve as

a within-cluster paired reference. The type of harvest treatments will be uniform within a cluster

to enable us to examine the downstream response to the multiple harvest treatments of the same

type. Catchments in two clusters (UM and BS) occur on Weyerhaeuser ownership and will be

clearcut harvest with no riparian tree retention along streams; these treatments represent current

harvest practices in commercial forestry. In PH cluster, on ODF ownership, riparian tree

retention will occur during clearcut harvest. In RK cluster, also ODF, no harvest will occur and

this cluster will serve as an undisturbed reference for evaluation of downstream-effects. Harvest

treatments in all basins will be implemented in 2011-2012 (Table y).

       Table 1. Area of catchments and clusters.

                   Cluster           Catchment         Area (ha)
                  Upper Main           UM 1              44.5
                                       UM 2              37.6
                                       UM 3              33.4
                                     Whole cluster      278.8
                 Bob and Sherri         BS 1             26.6
                                        BS 2             39.1
                                        BS 3             37.8
                                        BS 4             21.1
                                     Whole cluster      302.2
                     Pothole            PH 1             67.2
                                        PH 2             45.1
                                        PH 3             48.5
                                        PH 4             26.4
                                     Whole cluster      324.6
                  Rock Creek            RK 1             44.8
                                        RK 2             32.4
                                        RK 3             35.3
                                        RK 4            38.71
                                     Whole cluster      664.2

       Elevations range from 275 m at the downstream terminus of the study area to almost

1,100 m at the top of Trask Mountain. The area experiences wet winters, relatively dry summers

and mild temperatures throughout the year. Heavy precipitation results from moist air masses

moving off the Pacific Ocean. Mean annual precipitation ranges from 175 cm to 500 cm with the

majority of precipitation falling in the winter months of November, December, and January

(OWEB 1995).

       Fish –bearing stream channels within the study area range from relatively wide, low

gradient reaches to highly constrained, relatively steep channels. Headwater streams also exhibit

a wide range of channel gradients and confinement due to variable geology. The bedrock

geology of the study area is a mix of igneous and sedimentary formations dating back 40 to 60

million years (Wells et al. 1994). Lower-gradient, unconfined channels tend to be more

prevalent on the sedimentary formations. Large wood is scarce in many streams throughout the

watershed due to past fires and subsequent salvage logging.

       Much of the study area was impacted by the Tillamook Burns, which consisted of

multiple fires between 1933 and 1951. In 1933, the first fire extended over 240,000 acres near

Hebo. The second Tillamook fire occurred in 1939 and burned approximately 190,000 acres near

Saddle Mountain. The 3rd fire occurred in 1945 and burned 180,000 acres between the Wilson

River and Salmonberry Creek fires. The final burn in the Elkhorn and North Fork Trask area

occurred in 1951 across 33,000 acres. All except the first fire included the Trask study area.

Salvage logging of the both burned and unburned areas occurred in the 1950s within the research

basin and areas were replanted. The overstory vegetation is predominantly second-growth

Douglas-fir (Pseudotsuga menziesii), while red alder (Alnus rubra) predominates along many

stream channels. Since 2000, some thinning has occurred in PH and a limited amount of clearcut

harvest has occurred on portions of Weyerhaeuser property.


     Year                                      Trask Activities

   2005-2006      Pre-harvest abiotic and biotic studies begin; survey of stream network for
                  fish distribution; installation of downstream gauge sites and TTS stations.
      2007        Installation of gauges and TTS on headwater basins. Fish weirs installed
                  on downstream sites. Abiotic and biotic studies at headwaters and
                  downstream fish reaches
   2008-2011      Pre-harvest abiotic and biotic studies continue
   2011-2012      Harvest treatments in 9 headwater basins
   2012-2017      Post-harvest abiotic and biotic studies, final reports and publications

Hypotheses and methods

       Abiotic factors


   H1 headwaters: Water yield and the frequency of small to moderate peak flows will increase

   following harvest.

   H1 downstream: Changes in flow will not be detectible.

(methods for discharge are described below with those for sediment and turbidity)

Sediment and turbidity

   H2 headwaters: Export of suspended inorganic sediment and turbidity will increase in

   headwater streams following harvest. The increase in transport will be associated with both

   an increase in runoff and an increase in mobility of sediment within harvested unit.

   H2a chronic sediments from roads

   i. Existing roads: High traffic use combined with moderate-high intensity precipitation will

   result in local increases in turbidity when roads are hydrologically connected to the stream.

   Otherwise, there will be no detectable effect of roads constructed and maintained to current

   FPA standards on turbidity under average winter storm conditions and road use.

   ii. New roads: Under the same conditions as above, new roads will result in greater increases

   in turbidity when roads are hydrologically connected to the stream.

   iii. All roads: There will be no detectable effect of newly constructed or existing roads built

   and maintained to current FPA standards on turbidity under average winter storm conditions

   and road use.

   H2b episodic sediment from roads: Road-related landslides and washouts/diversions will

   occur with mean annual flood flows with an increasing occurrence associated with increasing

   road risk. These landslides and washouts/diversions will result in local channel scour and


   H2 downstream: As distances from road-related sediment sources increase, our ability to

   detect a chronic road-effect will decrease. We will detect downstream cumulative impacts on

   channel scour and deposition, turbidity, and suspended sediment if there are multiple

   episodic road-failures.

Methods for discharge, sediment, and turbidity: To quantify effects of harvest on water

quantity and quality, stream gauging and threshold turbidity sampling (TTS stations will be

installed on small, perennial, non-fish-bearing streams and in downstream study reaches. In each

of the clusters that will have forest operations, gauges will be installed in two headwater streams

and at a downstream site. One of these headwater gauges will be at the mouth of a harvested

basin, and the second will be at the mouth of a similarly-sized, unharvested basin. Within the

reference Rock Creek (RK) basin, one gauge will be established on an untreated headwater

basin. These headwater installations will use pre-calibrated fiberglass flumes to quantify

discharge. Downstream gauging stations, using pressure transducers for measuring stage in open

channels, have been installed at a downstream site in each cluster, plus one on the East Fork

South Fork Trask above the Rock Creek confluence. Rating curves for stage and discharge are

being developed at these five downstream gauges.

       Instrumentation at gauge sites includes data loggers, pressure transducers, air and water

temperature sensors, turbidity sensors, specific conductance probes, and automated pump

sampler and measurements are recorded every 10 minutes. Storm event water samples are

collected with the automated pump sampler during storms at pre-determined turbidity levels, and

then transported to OSU for analysis of suspended sediment. Statistical relationships are

developed between in-stream turbidity and concentrations of suspended sediment and used to

extrapolate real-time, in-stream turbidity values to sediment fluxes. Pre-harvest data for flow,

turbidity and sediment fluxes will be compared with post-harvest data to evaluate effects of

forest practices.

       Road Surveys for all open and closed roads (30-35 years old) will be conducted prior to

treatment, after road construction, and immediately following treatment, and periodically there

after for the life of the study. Road surveys will quantify hydrologic connection, functional

rating for road surface, risk of failure and a sample of recreation trails. Traffic use will be

measured. With the road survey results, road segments most likely to affect water quality will be

monitored. This may include the use of cap rod and flumes in ditches to estimate turbidity and

suspended sediment delivery to streams from roads.

Nutrient concentrations and fluxes

    H3 headwaters: Nutrient concentrations, including nitrogen, phosphorus, carbon and various

    cations, in headwater streams will increase during and after harvest. Increases will be greatest

    in the first autumn flows following harvest and where riparian buffers are removed.

    H3 downstream: Changes in nutrient concentrations will decrease between upstream and

    downstream reaches and will not be detectible downstream.

Methods: Nutrient analyses will be conducted on water samples from selected storm flows and

during baseflows.

Stream temperature, shade and heat budgets

   H4 headwaters: Solar radiation reaching headwater streams will increase following harvest.

   This in turn will result in increased maximum and minimum spring and summer stream

   temperatures. However, if logging slash covers streams after harvest, increases in

   temperature may be delayed (until the slash decays) or not occur at all if streamside

   vegetation shades the channel before the slash decays. Sites where riparian buffers remain

   intact after harvest will not show increased temperatures.

   H4 downstream: Increases in headwater maximum stream temperatures will decline with

   distance downstream from harvested areas. Major processes responsible for decline will be

   advection and conduction and dilution from cooler groundwater or hyporheic flows.

   Downstream study reaches may show increases in minimum summer temperatures after


Methods: Gauged streams will have instantaneous measures of water and air temperature every

10 minutes year round. In-stream data loggers will be used to monitor half hour temperature in

ungauged streams from June through September each year. Before being deployed , thermistors

will undergo accuracy checks. Changes to the vegetative cover near streams can have strong

influences on microclimatic conditions. This in turn can impact energy exchange between the

stream and the atmosphere and can alter stream food web structure and dynamics. During biotic

sampling, the amount of overhead cover above the stream reach will be quantified using a

densiometer ( ). Measurements of overhead cover will be made at center of each transect.

Annually, potential incoming radiation and shade will be quantified at each site using

hemispherical photography. A proposal has been submitted to purchase four portable

microclimatic stations, which will allow us to evaluate microclimate at multiple sites and to

study longitudinal effects and processes. This will allow us to quantify heat budgets for streams

(Johnson 2004) before and after harvest, as well as examine influences of upstream harvest on

downstream microclimates and heat budgets.

Instream habitat

   H5 headwaters: Channel form may be moderately influenced by changes in inputs, transport

   and deposition of sediment following harvest. Pools will become shallower with increased

   fines in the stream bed. However, changes would not be expected to be dramatic, because

   major effects likely occurred following earlier fires, harvest and post-fire salvage. Channels

   may not have fully recovered from these past activities.

   H5 downstream: Pools will become shallower with increased fine sediment deposited on the

   stream bed.

Methods: Widths, depths and characterization of particle sizes will be measured at 5 points

within each of 6 instream transects during each sampling for macroinvertebrates and algae

standing stocks. Monumented cross-sections will be established in headwater and downstream

study reaches to examine changes over time in cross-sectional areas, substrate composition, and

embeddedness. In downstream fish reaches, gradient, geomorphic habitat unit types (i.e. pools

riffles etc), wetted area and volume, and residual volume of pools will also be quantified during

summer fish sampling.

       Biotic responses

Primary production

   H6 headwaters: Increased light and nutrients following harvest will lead to increased rates

   of primary production, chlorophyll a and algal biomass. This will be most apparent in early

   summer, when leaf out occurs on undisturbed streams and would have resulted in shading of

   the headwaters.

   H6 downstream: Effects of harvest on primary production will dissipate with distance

   downstream from harvest. At downstream study sites, no increases in standing stocks or

   primary production will occur.

Methods: Standing stocks of benthic algae and chlorophyll a (a major indicator of

photosynthetic activity) will be measured at headwater and downstream sites. Collections are

made in early spring, prior to leaf-out, and mid-summer. Sample reaches are same as those for

invertebrate sampling (see below), and are approximately 30 meters in length in headwater

streams and 100 meters at downstream sites. For each sample (four replicates in headwater

streams, six in downstream sites), a minimum of three subsamples is composited. Sample

collection consists of scrubbing known areas of hard substrates (such as gravels, cobbles or small

boulders) or collecting small shallow cores in fine sediments. Samples are placed in the dark on

ice and returned to the laboratory within 48 hours of collection.

       In the laboratory, each sample is split into two aliquots. One portion is filtered onto pre-

combusted and pre-weighed Whatman GFF filters, dried at 60oC, weighed to determine dry

mass, combusted at 500oC and reweighed to determine AFDM. The second portion is filtered

through a pre-combusted GFF filter and analyzed for chlorophyll a and phaeophytin content by

the hot ethanol extraction method (Sartory and Grobbelaar 1984). Samples are analyzed within

one month of field collection to prevent degradation of the photopigments.

       Luminescence-based optical dissolved oxygen (DO) sensors will be used to evaluate

spatial and temporal variability in DO and to quantify ecosystem production and respiration.

These measures of whole stream metabolism provide a more robust estimate of overall benthic

productivity than chamber techniques. Whole-stream metabolism will be measured using the

two-station method ( ), when possible, and reaeration quantified using conservative tracer

releases (Hall et al., Oxford book ).

Detritus and instream wood

   H7 headwaters: Where there is complete removal of overstory vegetation near streams,

   inputs and exports of detritus will decrease immediately following harvest, unless slash

   remains on site, in which case there may be short term increases. Amounts of wood in many

   streams are currently low, but in streams where wood is present, we expect a decrease in

   wood abundance over time after harvest without buffers. Changes are not expected to occur

   on small streams where buffers are retained.

   H7 downstream: Detritus is expected to decrease in downstream reaches following harvest

   due to decrease in downstream transport of litter from headwater streams with removal of

   riparian vegetation.

Methods: During low flows, instream and near-stream wood will be quantified. String transects

(Wallace and Benke 1984, Swanson et al. 1984) will be conducted at each of the channel

measurement transect sites, extending 1 m into riparian areas, and through each of the major

accumulations within study reaches. Measurements of wood diameters, made using calipers, can

be converted to surface area.

       Detritus will also be quantified during summer sampling. A 6-cm diameter corer will be

inserted several cm into substrate and benthic organic material within core collected. Then

subsurface material will be stirred and suspended and a subsample of fine benthic organic

material collected. The volume of supernatant within the corer will be recorded. Organic matter

samples will be dried (60oC) and ashed (500oC) to quantify AFDM/m2.


   H8 headwaters: After harvest, abundances of benthic and drifting macroinvertebrates will

   increase. The community composition of macroinvertebrates that feed by grazing is expected

   to increase. Other taxa may shift their diet from detritus to increased consumption of algae,

   resulting in a shift after harvest in their stable isotope ratios (lower δ13C after harvest). The

   sizes of representative taxa of functional feeding groups will shift due to increased stream


   H8 downstream: Increases of drifting macroinvertebrates will be greatest closest to the

   harvested, headwater basins. We expect this effect to dissipate with distance downstream.

   Thus, we do not expect a change in macroinvertebrate drift or benthic community structure

   following harvest. Benthic macroinvertebrates will likely not increase at downstream effects

   sites unless there is an increase in primary production at those sites.

Methods: Responses of macroinvertebrate communities in headwaters and downstream sites will

be evaluated through the collections made from the substrate (i.e., benthic samples) and from the

water column (i.e., drift samples) (Barbour et al. 1999). Macroinvertebrate communities will be

sampled from headwater watersheds in April and June of each year. At each headwater site, six

benthic samples will be collected and composited before being counted. Drift samples will be

collected at each site for 24 hrs. Water depth and velocity will be measured at the point of

placement of drift nets to calculate the volume of water sampled. At the four downstream sites,

macroinvertebrate communities will be sampled in a similar fashion. Additional sampling for

macroinvertebrates will occur at the downstream sites when fish diets are being sampled in

August. August sampling will not be conducted in the headwater sites due to low stream flows.

       Samples will be processed in the laboratory at Oregon State University. A 500 count

subsample will be taken from the surber samples and the entire drift samples will be processed.

Macroinvertebrates will be identified to the lowest feasible taxonomic category, genus for most

insect taxa and order or family for non-insects. Surber samples allow for an estimate of benthic

density and density in drift samples will be estimated as the number of individuals per liter of

water sampled. The presence and dominance of sensitive and tolerant taxa will be determined;

sensitive and tolerant taxa will be identified from published work on macroinvertebrate

responses to human disturbances (Barbour et al. 1999). Relative importance of food resources

will be determined by assigning taxa to functional feeding groups (Merritt and Cummins 1996)

and through stable isotope analysis.

       Several taxa from each order will be selected as indicator taxa. Changes in size, condition

and rate of growth of the indicator taxa will be measured. Individuals will be collected from the

samples taken for the community composition work and from additional monthly samples (see

below). For these additional samples the indicator taxa will be sorted in the field, preserved, and

taken to the laboratory for analysis. Developmental stage of individuals will be determined from

head capsule widths and wing pad development. Size and biomass will be measured and

compared to published length-weight relationships. Estimations of the time and size at

emergence will be calculated based on the size and time at which late instars (based on head

capsule width and wing pad development) are found. Examining individual condition as well as

community and population metrics of macroinvertebrates may improve our ability to detect

responses to forest harvest.

       Additional funding is also being sought to study responses of crustaceans, signal crayfish,

Pacifasticus leniusculus) to forest harvest. At headwaters and downstream sites, crayfish will be

captured using baited traps and individuals injected with a coded wire tag. Marked individuals

will be sexed, weighed, and carapace and cheliped length measured. Reproductive status of

females will be noted. Monthly trapping (April-June/July at headwater sites and April-

September at downstream sites) will be conducted. Mark-recapture efficiencies will be used to

estimate local crayfish population size. Individual growth rates will be determined for each

sampling interval. Changes in reproductive status will also be noted. Diet of crayfish will be

determined from a subsample of non-marked individuals using stable isotope analysis.

Experiment on temperature influences on macroinvertebrate growth and emergence

       We are conducting laboratory experiments at the Oregon Hatchery Research Center to

examine influences of temperature on growth, biomass and timing of emergence for indicator

species of macroinvertebrates. According to existing but limited research on instream

macroinvertebrate responses to temperature, increases in temperature lead to faster growth

through each instar (Vannote and Sweeney 1980, Harper and Peckarsky 2006). This faster

growth has been suggested to translate into smaller sizes and reduced biomass as well as earlier

emergence. For insect taxa with multiple generations per year, temperature changes could result

in more generations per year of smaller individuals.

       Because fish diets are greatly influenced by the biomass of insects that they consume, and

we know little about the factors influencing biomass of their prey, we will experimentally

examine the influences of stream temperature on macroinvertebrate size and biomass. We will

also examine the timing of emergence as a function of temperature, which would influence the

life stage of the invertebrates and their availability to fish.

        We will examine the influences of temperature on aquatic larval growth rates, sizes, wing

development and timing of emergence as adults for representative mayflies, stoneflies and

caddisflies. Experimental organisms will be chosen in part for their ubiquity and appropriate

developmental stage in Oregon coastal range headwaters. Larval macroinvertebrates will be

collected from nearby headwater streams and reared in covered baskets in flow-through troughs

or stacked flow-through incubations trays in an inside laboratory at the Oregon Hatchery

Research Center (OHRC). These insect taxa are easily grown in laboratory settings, especially

with flow-through systems.

        Responses of macroinvertebrates to stream water of three temperature regimes will be

studied. One pair of troughs and incubation stacks will temperatures typical of undisturbed

headwater streams. A second pair of troughs and stacks will have elevated temperatures of 2-

3oC, typical of temperatures in headwater streams where riparian vegetation has been disturbed.

Whereas most growth and temperature experiments in the literature have been conducted at non-

fluctuating temperatures (Quinn et al. 1994), both these treatments will be allowed to fluctuate

with diel temperature regimes. To isolate the effect of temperature alone, and eliminating

possible nutrient, turbidity or sediment differences between Fall and Carnes Creek, a third pair of

troughs and stacks will have warmed water heated to approximately 18oC; temperatures in this

pair of troughs will not fluctuate from day to night. These experiments will begin July 2007 and

if successful, a second set of will begin in spring 2008 and continue into early summer to

examine taxa responses to increased temperature in spring and early summer.


    H9 headwaters: Abundance and condition of tailed frogs will change after clearcut

   harvesting due to increased temperature and shifts in food resources to algae. We expect

   lower δ13C after harvest. Timing of metamorphosis will shift to earlier in season due to

   interactions of food quality, availability, and temperature

   H9 downstream:

Methods: Tailed frogs (Ascaphus truei) are abundant and salamanders are present. Amphibians

can be particularly sensitive to changes in water temperature and fine sediment (Pilliod et al.

2003), both of which may be affected by forest harvest. We will explore approaches for

measuring individual movement, growth, survival, and condition. A clear understanding of how

to measure individual and population-level responses will be critical to the success of

understanding responses of amphibians to forest harvest and related changes in habitat



H10 downstream: Length at age, condition, and survival of salmonids and sculpins will increase

at focal sites, due to possible changes in temperature and food availability. However, fish

condition may decline if increased water temperature and sedimentation are sufficient to offset

any benefits due to increased length of growing season or trophic productivity. δ13C values of

fishes will decrease due to a shift to a more algal-based foundation of the food web.

To evaluate how forest operations are influencing biota, it is important to understand how habitat

and organisms are distributed in both space and time and concurrently how organisms interact

across gradients of habitat, relative abundance and season. We will monitor individual fish to

understand how physiological and behavioral responses are linked to observed changes at the

population and community levels (e.g., Railsback et al. 2003). This research will also incorporate

processes responsible for any observed changes in fish populations. As a result, we expect to be

able to understand cause-and-effect linkages between fish and spatial and temporal changes in

habitat, and also to identify responses that are most sensitive to habitat changes linked to forest

management. Over the longer-term, we intend to use this life history information in life cycle

models for key species (e.g., coho; Lawson et al. 2002) and to understand responses to changes in

habitat caused by forest harvest within the context of other factors fish experience during their

lifetimes (e.g., Holtby 1988).

       Our proposed approach to studying fish responses to forestry in TWS will adapt novel

approaches to examining habitat requirements and forestry impacts on headwater fishes. The

focus will be on coastal cutthroat trout, since this species dominates in headwater streams.

Detailed site-specific studies of individual fish condition, growth, mortality, immigration, and

emigration will be coupled with monitoring of physical habitat conditions in the stream (e.g.,

channel structure, water quality) and biotic conditions (e.g., primary and secondary productivity)

to understand how fish and habitat interact at this scale.

This work strives to look at how changes in individuals actually scale up to population processes.

Typically studies of the effects of forestry on fish look only at a single response of individuals

(e.g., growth or survival) or aggregated responses of populations (e.g., abundance or size

structure) without making an explicit, quantitative linkage between individual responses and

population-level consequences. The challenge is how to link these individual responses in a

meaningful and quantitative way to understand how they function to influence the dynamics of

populations (Figure c). With the development of powerful individual-based simulation models of

fish populations, it is now possible to explore the linkages much more clearly. While it is

impossible to collect information on every possible relevant response in the field, simulation

models can help to explore likely scenarios.

           Fish Abundance

 (Birth + Immigration) – (Death + Emigration)

                        Σ (Individual Fitness)

                          Growth, Survival

Figure C. A simplified depiction of how fish abundance (top) can be decomposed into four fundamental
processes, which in turn are determined by individual fish that are presumably acting in ways to minimize
their fitness. Fitness in turn is a function of growth and survival of individual fish. The arrows in this
diagram indicate how abundance decomposes into basic responses of individuals, such as growth and
survival. Alternatively, this can be thought of as a pathway from individual growth and survival to
processes that determine fish abundance, distribution, and other population-level responses. Bold text
indicates responses we can measure in the field, in addition to abundance.

Short-term fish study plans and long-term goals

During pre-harvest periods, our objectives are to:

      Develop a conceptual foundation for developing preliminary hypotheses about the likely

       effects of contemporary forest practices on coastal cutthroat trout in headwater streams.

      Design and implement a field study to evaluate the potential responses of trout to

       variability in environmental conditions in locations that are most likely to be impacted by

       planned forest treatments.

      Collect data to address key areas of uncertainty regarding fish-habitat requirements in

       headwater streams.

      Continue development of a long-term framework and approach to addressing questions

       about fish and forestry in the Trask that can also be generalized to a broader array of

       locations and scenarios. This includes additional proposal development and active

       pursuit of funding external to those allocated to the TWS.

       Our depiction of pathways of influences (Figure D) leading to fish responses recognizes

both individual and population-level factors. For individual fish, we recognize three factors of

primary importance: food availability, predation risk, and metabolism. Much of our logic is

based on that offered by inSTREAM and associated publications


  Supply,            Fine sediment       Upstream
  routing,               supply         hydrological                                               Solar radiation
 retention                               processes               Number and                         Subsurface
                                                                   sizes of                             flow
 Instream                Channel               Low flow
   Wood                  infilling             discharge
                                                                 Intra- + Inter

   Riparian               Instream                 Pool depth
                                                  Hiding cover                     of predators    Metabolism
      Habitat               Substrate
  for terrestrials           (fines)
                                                                     cover avail
 & aquatic adults

     inputs                                                                         Predation       Individual
                                                                                       risk         Movement
                                                       Food Avail                                     growth
   Riparian                                                                                           survival
                                Water                   Aquatic                                      condition
                             temperature               Food Avail                     Food
                           (Solar radiation)                                        availability

       Number and                       Intra- + Inter                                             abundance
         sizes of                          -specifc
                                         competition                                               production

Figure D. Influence diagram depicting major relationships between environmental variables

and the responses of individuals and populations of fish. Solid lines indicate positive influences

and dotted lines indicate negative influences. If influences are nonlinear (e.g., temperature and

fish metabolism), both lines are shown. Of necessity, this diagram emphasizes major influences

on fish while greatly simplifying relationships for other ecosystem components.

             Food availability and metabolism influence growth and condition of fish, and possibly

survival via starvation. If fish are sufficiently starved, they may also engage in risky behavior to

obtain food, and thus become more vulnerable to predators. Predation risk is the third primary

factor and one that is poorly understood for headwater fishes. Each of these primary factors is

influenced by a series of physical and biological processes that trace back to the outer edges of

the influence diagram. Beyond these outer edges are the effects of forestry on upstream forest,

hydrologic, and geomorphic processes, such as wood recruitment, sediment routing, energy

balance (temperature), and nutrient flux. Also of possible importance is the influence of

downstream and marine conditions on the abundance of coho salmon and steelhead trout, both

major potential competitors of cutthroat trout.

Methods: Field study in 2007

To address the variety of individual and population-level responses shown in Figures C and D,

we are establishing “focal sites” which will enclose 200-250m sections of stream with two-way

(upstream/downstream) traps at the top and bottom. Focal sites in each cluster are the closest

reaches downstream of headwater catchments that contain fish. Fish growth and behavior will be

monitored during low flows, which a time of year when survival of larger (age 1+) cutthroat

trout has been shown to be lowest (Aaron Berger, personal communication) in other watershed

studies (Hinkle Creek). Furthermore, this is a time of year when weirs can be effectively

maintained, and when access is most feasible. The anticipated schedule of events will include

the following:

  2007                                            Activity
  May-June                                        Prepare for field work, begin testing of
                                                  weir and trap designs
  June-July                                       Begin installation of weirs and traps
  July (last week)                                Close weirs and conduct initial mark-
                                                  recapture population estimates, tagging and
                                                  measurement of fish
  August-15 September                             Weir and trap maintenance. Includes
                                                  tagging of fish moving out and into sites
  August 20-24                                    Mark-recapture population estimate at all
                                                  sites, measurements on captured

  September 17-21                               Mark recapture population estimate at all
                                                sites, measurements on captured
                                                individuals and weir removal
  September 30                                  Summer 2007 field work completed
  >30 September                                 Data analysis and report preparation

Fish measurements. Fish within sites will be initially captured and individually tagged with PIT

tags. Mark-recapture population estimates will be made during the beginning, middle, and end

of the study period. Measurements on all captured fish will include species, size, condition, and

initial location within the site. Individual rates of growth will be estimated for recaptured fish, as

well as changes in individual condition, emigration, immigration, and survival. In addition to

data collected at focal sites, the Oregon Department of Fish and Wildlife has two screw traps

located downstream of the TWS. These traps will capture out-migrating salmon and trout,

providing additional information on individuals at a later time.

       Overall our goal in 2007 is to refine our methods for a focal site approach and

successfully collect information on fish and habitat characteristics during a time of year when we

expect fish to be highly vulnerable to possible changes in habitat related to possible influences of

forestry. In future years, we will consider complementing this season of observation with work

in other times of the year, as resources allow. Similarly, we will consider adding other focal site

locations as resources allow. In 2007, we hope to learn the following:

      Assessment of multiple biological responses within each of the four focal sites, including

       individual growth rates, condition, size distributions, immigration/emigration, abundance,

       and survival rates.

      An analysis of variability in these responses within and among sites.

      Comparison of other biotic and abiotic conditions among the four focal sites.

Addressing key uncertainties in fish-habitat requirements

        There are many unknowns for fish-habitat requirements, and we have focused on

predation as a key uncertainty in headwater streams. Much work has focused on fish growth,

movement, population abundance and size structure. Less has focused on specific mechanisms

influencing survival. It is well-established that survival may vary seasonally for many species,

but why survival is lower or greater is not well understood. In other words, is survival tied to

starvation, predators, disease, or other factors? We suspect predators play a major and unstudied


        Predators such as birds and mammals have strong impacts on salmon and trout in larger

lakes and rivers, but studies in small streams such as those studied in the three paired watersheds

are lacking. One reason is that observing these predators directly is a daunting logistical

challenge. Predators may only visit an area once in a year or even less frequently and have an

important influence on fish populations. Thus, it is necessary to devise indirect approaches for

understanding how fish may interact with predators in small streams, and how the ability of fish

to avoid predators is tied to local habitat conditions. This is a key linkage between the response

of fishes to forestry and natural processes that influence availability of habitat that fish may need

to avoid predators (hiding cover; Figure D).

        In this study we will measure characteristics of hiding cover and quantify use and

selection of hiding cover by salmon and trout in the three paired watersheds. We will take

advantage of existing marked fish to study their patterns of habitat use in relation to unused

habitats to measure the strength of selection of habitats by individuals. By understanding what

hiding cover is, and obtaining a measure of the quality of hiding cover (i.e., selectivity), we will

have a much better sense of how fish use habitat to avoid predation. Work will occur during

seasonal low flows in summer and fall, where fish may be more confined and vulnerable to

predators (D. Bateman, personal communication). To determine if fish are using hiding cover,

we will approach the stream with mobile antennas submerged to detect locations of marked

individual fish. Fish that do not move after detection are assumed to be using hiding cover,

whereas fish that leave the vicinity are assumed to be using a different tactic (evasion) for

avoiding what they perceive to be predators (humans entering their habitat). When possible,

underwater observation will be used for verification. Measures of hiding cover will include size

of pools (e.g., area, depth) and availability of instream cover (e.g., wood, un-embedded substrate,

undercut banks, boulders, turbulence). Other factors (e.g., velocity, light) may also be important.

At the level of individuals, natural selection should drive individuals to select habitats that

provide greater opportunities for feeding and for surviving the threat of predators. The relative

benefits of different habitats should also be conditioned on characteristics of the individual (e.g.,

size) and external factors such as temperature and the density of con-specific and hetero-specific

competitors, as well as availability of different habitat types (e.g., if cover is abundant and

available, then selectivity should be lower).

In summary the objectives of this work, funded by FRL, as part of a Master of Science thesis

project for Heidi Vogel (Department of Fisheries and Wildlife, Oregon State University) will be


         Measure habitat use and availability for coastal cutthroat trout, steelhead trout, and coho

          salmon to model patterns of habitat selection within individual locations (stream reaches)

      Compare results among locations to understand conditions that may modify habitat

       selection (e.g., among paired watersheds, sites)

      Identify habitat characteristics within each location that are most important to different

       species in terms of the strength of selection observed by individual fish.

This Masters thesis will be completed in 2009.

Long-term framework

       We consider field work in the Trask River to represent one of several critical lines of

evidence and inference needed to understand cause and effect in terms of how fish relate to

habitat, and ultimately to the potential influences of forest practices. Our detailed site-specific

studies of individual fish condition, growth, mortality, immigration, and emigration and

corresponding measures of physical habitat conditions in the stream (e.g., channel structure,

water quality) and biotic conditions (e.g., primary and secondary productivity) can be linked with

statistical relationships to understand how fish and habitat are associated at this scale. This is an

important advance, but still does not fully demonstrate cause and effect. Such is not possible

with any observational field study.

Use of experiments to evaluate key uncertainties. Whereas integration of rigorous field studies

and simulations can provide unprecedented insights into population dynamics and habitat

relationships, these approaches cannot conclusively demonstrate cause and effect. Furthermore,

both lines of inquiry will undoubtedly reveal key uncertainties that need to be resolved. For

example, our uncertainty about what hiding cover is for fish in headwater streams has prompted

a graduate study.

       Experiments are a valuable complement to observational field studies and simulation

studies because they can effectively isolate the influences of one or a few factors on fish. While

experiments allow more precise manipulation, control, treatment, measurement, and more

powerful study designs, they may not necessarily have relevance to actual field conditions,

where conditions are complex and uncontrolled. In our case, however, we have access to a new

facility that offers an opportunity to conduct controlled, manipulative experiments in semi-

natural conditions that offer a useful complement to inferences from the field.

       For experimental work we hope to make use of new experimental stream channels at the

Oregon Hatchery Research Center, http://www.dfw.state.or.us/OHRC/. Patterns, processes, and

uncertainties identified through the first years of field work in the Trask, as well as initial model

simulations can help to identify the critical questions that are most important to address under

experimental conditions.

       A preliminary list of questions includes the following:

   1. Importance of predation as a source of mortality in headwater streams
         a. What makes fish take risks? It is believed that fish take more risks to acquire
             resources (e.g., food) when they are scarce (e.g., when fish are not acquiring
             enough food to meet their daily ration requirements). Does this occur in the field,
             or under semi-natural conditions?
         b. Can we directly study the potential influences of predators (e.g., mink, raccoons)
             in experimental stream channels to better understand the consequences of
             individual behaviors in terms of predation risk? Is predation risk conditioned on
             availability of cover, food, or space within streams?
   2. Movement
         a. What makes fish emigrate? Are fish that emigrate in lower condition? What are
             the sizes of fish that emigrate relative to those that do not? How does emigration
             influence population dynamics (e.g., growth and survival of non-emigrating fish)?
         b. How strong is local site fidelity? Do fish hold rigid stations or defend fixed
             territories, or do they exhibit some form of local movement or ranging behavior
             within streams? What factors influence the degree of site fidelity within streams?
   3. Competition
         a. Intraspecific – How strong are interactions among individuals within a cohort
             versus among cohorts?

           b. Interspecific – Do other species common in headwater streams (e.g., coho
              salmon) have a substantial influence on the population dynamics of coastal
              cutthroat trout? Are these influences greater than those resulting from changes in
              habitat that may result from forestry or natural changes?

       By spanning our inferences from a generalizable simulation model to tests of key

hypotheses in experimental stream channels, and analysis of field sites within the Trask

Watershed, we will be able to span the chain of ecological inferences needed to address the

complex questions at hand.

Modeling and linking of abiotic and stream food webs

In the Trask study, we intend to provide:

      An integrated and biologically realistic analysis of process that is not possible with

       classical statistical approaches focused on pattern detection (e.g., BACI or “Before-After-


      Models to simulate potential future scenarios (e.g., “futuring” the longer-term effects of

       forestry and climate change) that are impossible to examine with empirical data alone.

      Modeled scenarios to develop rigorous hypotheses and testable predictions about how

       fish and more broadly food webs respond to the potential impacts of forestry. These

       provide the baseline for evaluating effects of forestry rather than a short (and statistically

       noisy) “before” dataset.

      Perspectives on stream responses to forestry that can be generalized outside of the Trask

       Watershed and provide useful insights regardless of the outcome of the planned forest

       treatments. In other words, results that can be applied across a much broader scope than

       the time and location of the Trask study.

       In this work we will base much of the field effort in the Trask Watershed. However, we

wish also to generate results that can be applied to a variety of forest land ownerships (e.g., BLM

and Forest Service, as well as ODF and private lands) and impacts (e.g., temperature, sediment,

etc.). By using a modeling framework, we will be able to look at a broad variety of potential

scenarios that are relevant to both possible scenarios in the TWS as well as land ownerships

outside of the TWS. By spanning our inferences from a generalizable simulation model to tests

of key hypotheses in experimental stream channels, and analysis of field sites within the Trask

Watershed, we will be able to span the chain of ecological inferences needed to address the

complex questions at hand.

Overview of observational studies, models, and experiments.

       By integrating multiple approaches to understanding relationships between fish and

habitat, we hope to overcome limitations of each approach, as well as take advantage of the

strengths of each approach. Models allow an integration of process that is not possible with field

observations or experiments, and allow us to examine a number of possible scenarios

representing hypotheses about how habitat will change and fish will respond. Fish responses

represent testable predictions that result from these different scenarios or hypotheses.

Observational studies allow measurement of responses of fish in the field under natural

conditions to changes in habitat in space or time. Statistical models of fish and habitat

relationships can help to identify patterns and potential mechanisms or processes. Such studies

are by nature retrospective and provide a useful baseline for comparison to results from modeling

different scenarios.

       Models and observational field studies cannot resolve all of the uncertainties we have

about fish and habitat relationships. For key uncertainties identified by these approaches, we can

conduct experiments designed specifically to design them. The results of experiments can feed

back on models and field studies so that we may re-design models or conduct new studies to

better understand both pattern and process.

Expected outcomes

       Results from the first generation of watershed studies had a profound impact on

development of contemporary forest practices (Hall et al. 2004). We expect findings of the

Trask Studies will similarly benefit state and private forest landowners and natural resource

managers by expanding the understanding of key linkages between forest practices and aquatic

habitat. This improved understanding will enable state and federal agencies to develop and refine

contemporary forest management strategies to protect and restore aquatic habitat while enabling

forest owners to profitably manage their lands. In addition, Trask research in conjunction with

other studies will provide fundamental information on the organization and function of stream

ecosystems in the Pacific Northwest. This improved knowledge will ultimately enable more

efficient and effective distribution of resources available to restore streams in western Oregon.

       The results of our research must be easily accessible by technical and non-technical

audiences to be maximally useful. Information will be disseminated to the scientific community

through journal publications and presentations at scientific meetings and to managers and

broader audiences through regional workshops, field tours, newsletters, and information posted

on the Trask and WRC websites. The Trask and WRC are working with the outreach program at

the OSU College of Forestry to ensure the communication strategies are as effective as possible.

       Literature Cited

Barbour, M.T., J. Gerritsen, B.D. Snyder, and J.B. Stribling. 1999. Rapid Bioassessment
      Protocols for Use in Streams and Wadeable Rivers: Periphyton, Benthic
      macroinvertebrates and Fish, 2nd Ed. EPA 841-B-99-002. U.S. Environmental Protection
      Agency, Office of Water, Washington, D.C.
Benda, L., M. A. Hassan, M. Church and C. L. May. 2005. Geomorphology of steepland
      headwaters: the transition from hillslopes to channels. Journal of the American Water
      Resources Association 41: 835-851.
Gregory, S.V., J.S. Schwartz, J.D. Hall, R.C. Wildman, and P.A. Bisson. In press. Long-term
      trends in habitat and salmonid populations in the Alsea Basin. History of Alsea,
Hartman, G.F. 2004. Effects of forest management on watershed processes. Pages 271-302 in
      T.G. Northcote and G.F. Hartman, editors. Fishes and forestry: worldwide watershed
      interactions and management. Blackwell Science, Oxford, UK.
Hall, J.D., C.J. Cederholm, M.L. Murphy, and K.V. Koski. 2004. Fish-forestry interactions in
        Oregon, Washington, and Alaska, USA. Pages 365-388 in T.G. Northcote and G.F.
        Hartman, editors. Fishes and forestry: worldwide watershed interactions and
        management. Blackwell Science, Oxford, UK.
Holtby, L. B. 1988. Effects of logging on stream temperatures in Carnation Creek, British
       Columbia, and associated impacts on coho salmon (Oncorhynchus kisutch). Canadian
       Journal of Fisheries and Aquatic Sciences 45: 502-515
Moore, R.D. and J.S. Richardson. 2003. Progress towards understanding the structure, function,
      and ecological significance of small stream channels and their riparian zones. Can. J.
      Forest Research 33: 1349-1351.
Northcote, T.G. and G.F. Hartman, editors. 2004. Fishes and forestry: worldwide watershed
      interactions and management. Blackwell Science, Oxford, UK.
Pilliod, DS, Bury RB, Hyde EJ, Pearl CA, Corn PS. 2003. Fire and amphibians in North
        America. Forest Ecology and Management 178:163-182.
Railsback, S.F., H.B. Stauffer, and B.C. Harvey. 2003. What can habitat preference models tell
       us? tests using a virtual trout population. Ecological Applications 13:1580–1594.
Reid, L.M. 1998. Cumulative Watershed Effects: Caspar Creek and Beyond. Chapter 13 In:
       Ziemer, Robert R., technical coordinator. Proceedings of the conference on coastal
       watersheds: the Caspar Creek Story; 6 May 1998; Ukiah, CA. Gen. Tech. Rep. PSW-
       GTR-168. Albany, CA: Pacific Southwest Research Station, Forest Service, U.S.
       Department of Agriculture; available on the internet:
Ringler, N.H., and Hall, J.D. 1975. Effects of logging on water temperature and dissolved
       oxygen in spawning beds. Trans. Am. Fish. Soc. 104: 111–121.

Rosenberger, A. and J. Dunham. 2005. Validation of abundance estimates from mark–recapture
      and removal techniques for rainbow trout captured by electrofishing in small streams.
      North American Journal of Fisheries Management 25: 1395-1410.
Sartory, D. P. and J. E. Grobbelaar. 1984. Extraction of chlorophyll a from freshwater
       phytoplankton for spectrophotometric analysis. Hydrobiologia 114:177-187)
Swanson et al 1984. Organic debris in small streams. GTR PNW166
Wallace and Benke 1984. Quantification of wood habitat in subtropical coastal plain streams,
      CJFAS 41:1643-1652

Harper, M.P. and B.L. Peckarsky. 2006. Emergence cues of a mayfly in a high-altitude stream
       ecosystem: Potential response to climate change. Ecological Applications 16: 612-621.
Quinn, J.M., Steele, G.L., Hickey, C.W. and M.L. Vickers. 1994. Upper thermal tolerances of
       twelve New Zealand stream invertebrate species. New Zealand Journal of Marine and
       Freshwater Research 28:391-397
Vannote, R.L. and B.W. Sweeney. 1980. Geographical analysis of thermal equilibria: a
       conceptual model for evaluating the effect of natural and modified thermal regimes on
       aquatic insect communities. The American Naturalist 115:667-695.


To top