Environmental Issues > Wildlands Main Page > All Wildlands Documents

Top of Report

Chapter 6


Increases in sediment and altered streamflows: Roads and logging degrade aquatic ecosystems by increasing levels of fine sediment deposited in streams and by altering natural streamflow patterns.

Adverse impacts on aquatic species: Increased fine sediment deposition in streams and altered streamflows and channel morphology result in increased adult and juvenile salmonid mortality, a decrease in aquatic amphibian and invertebrate abundance or diversity, and decreased habitat complexity.

Increases in sediment and altered streamflows: Roads and logging degrade aquatic ecosystems by increasing levels of fine sediment deposited in streams and by altering natural streamflow patterns.

Key Finding: Roads degraded stream habitat for aquatic species, including salmonids, by accelerating erosional processes and modifying natural drainage networks.

Source: Furniss, M. J., T. D. Roelofs and C. S. Yee. 1991. Road construction and maintenance. In Influences of forest and rangeland management on salmonid fishes and their habitats. American Fisheries Society Special Publication 19: 297-323.

The authors review research documenting the impacts of roads on stream habitat. Roads accelerate soil erosion rates due to surface erosion and mass soil movement such as slumps and earthflows, debris avalanches, debris flows, and debris torrents. High rates of stream sedimentation result from this increased erosion. Soil erosion rates (m3/hectare) were 30 to 300 times higher on forests with roads than undisturbed forest. Roads also altered streamflow rates and volumes, which along with increased sedimentation, resulted in altered stream channel geometry. Acting as new flowpaths for water, roads increased the channel network over watersheds, increasing the drainage density.

Research also demonstrated that roads degraded salmonid habitat by creating migration barriers like culverts and temporary dams caused by landslides. Erosion resulted in sedimentation of streams and declines in spawning habitat when too high a proportion of fine sediment was deposited. Macroinvertebrates, the primary food source of juvenile fish, also declined when large amounts of sediment were present.

Key Finding: Logging activities degraded stream habitat by changing the amount, quality, and timing of flowing water, increasing erosion rates, and reducing stream habitat diversity.

Source: Chamberlin, T. W., R. D. Harr and F. H. Everest. 1991. Timber harvesting, silviculture, and watershed processes. In Influences of forest and rangeland management on salmonid fishes and their habitats. American Fisheries Society Special Publication 19: 181-205.

The authors review the impact logging activities had on salmonid habitat through altering stream processes such as the amount, quality, and timing of flowing water, as well as changing the gravel substrate composition of the streambed, available fish cover, and food supplies. The studies reviewed show that logging altered streamflows by affecting snow accumulation rates in forests and snow melt rates. Because of vegetation removal, logging also changed evapotranspiration rates and soil water content, with resulting increases in annual runoff. Soil compaction changed infiltration rates and therefore runoff and erosion rates.

By removing streamside vegetation, logging changed stream temperatures, raising them in some cases, but lowering winter stream temperatures in more northern regions. Stream temperature was shown to affect the time required for salmonid eggs to develop and hatch. Stream channel structures were also altered after logging, with a corresponding loss of the habitat diversity required by fish populations. By accelerating erosion rates, logging increased sedimentation rates of streams. In the steep and high-rainfall forests of Oregon, Washington, British Columbia, and Alaska, for example, mass movements of soil were the dominant erosional process. Many of these mass movements originated on open areas after logging, with increases in frequency ranging from two to 31 times.

Key Finding: Soil erosion rates due to debris slides were many times higher on forests with roads, landings, and logging activity than on undisturbed forests.

Source: Amaranthus, M. P., R. M. Rice, N. R. Barr and R. R. Ziemer. 1985. Logging and forest roads related to increased debris slides in southwestern Oregon. Journal of Forestry 83: 229-233.*

The authors inventoried mass erosion events occurring over a 20-year period in the Siskiyou National Forest in the Klamath Mountains of southwestern Oregon. Aerial photos were analyzed from 24 forest sites and erosion attributed to roads, logging, or natural events. The volume of soil mass movements was estimated from the photographs, with partial field checking to confirm accuracy. Debris slides were found to be the primary type of mass erosion, accounting for about 80% of the volume of soil moved and 90% of mass erosion events inventoried. A total of almost 1.5 million yd3 of debris slide erosion occurred. Roads, occupying 2% of the area studied, were the sites for more than half the slides and 60% of the erosion volume. Clearcut areas, occupying 10% of the area studied, were the sites for 34% of the slide events and 18% of the slide volume. The authors also analyzed slides with respect to position on slopes, aspect, precipitation, and geology of study area.

* See also key finding in Chapter 5.

Key Finding: Roads were responsible for 61% of the soil volume displaced by erosion in northwestern California.

Source: McCashion, J. D. and R. M. Rice. 1983. Erosion on logging roads in northwestern California: How much is avoidable? Journal of Forestry 81: 23-26.*

The authors investigated erosion due to forest roads and logging in northwestern California. Their inventory covered 344 miles of roads in the Coast and Klamath Mountains. Roads were thinly rocked, graveled, or heavily rocked and regularly maintained logging roads. Slope, grade, aspect, cut-and-fill height, and soil volume displaced by erosion were recorded on each 1-mile road segment.

Mass erosion was the predominant form of erosion occurring in the study sites. Roads caused 152 of the 171 major erosional events inventoried (events that displaced more than 20 cubic yards of soil), and 61% of the soil volume displaced by erosion was due to these road-related events. The remainder was due to natural events and some logging-caused erosion. Road-related erosion increased with the slope traversed by the road. Seasonal roads had similar erosion rates to main-haul (and regularly maintained) roads.

In a separate study, erosion due to roads relative to logging areas was studied in 30,000 acres of commercial timberland in Six Rivers National Forest. The road network occupied less than 4% of the total logging area. Total erosion from the 30,000 acres was 137,800 cubic yards. Of this total, 40% came from the roads and 60% from the logged areas. The average erosion rate in the road rights-of-way (47 cubic yards per acre) was 17 times the average erosion rate in the logging areas (2.82 cubic yards per acre).

* See also key finding in Chapter 5.

Key Finding: Clearcutting increased the frequency of mass soil movements from hillsides.

Source: Gray, D. H. 1970. Effects of forest clear-cutting on the stability of natural slopes. Bulletin of the Association of Engineering Geologists 7: 45-66.*

A review of the scientific literature, including research from Alaska, Utah, California, Oregon, and Japan, demonstrated that clearcutting on slopes increased the frequency of mass soil movement events (landslides, earthflows, slips, etc.). The loss of forest cover was believed to affect slope stability in two principal ways:

1) Mechanical root support due to interconnected root systems was lost after logging. Research in Alaska, for example, indicated a time lag after clearcutting before landslide activity increased and a lack of landslide correlation with rainfall intensity. The authors believe this is due to the increased deterioration of root systems with time. Other studies similarly showed that with increasing age and maturity, the effectiveness of forest cover in preventing landslides increased.

2) A denuded slope was likely to reach critical soil saturation earlier than a forested slope (since no transpiration from trees can occur). Therefore, during a large storm, it was predicted that these soils would reach a critical failure condition earlier than a forested slope would.

* See also key finding in Chapter 5.

Key Finding: During storm events in southwestern Washington, average sediment levels in runoff from forest roads ranged from 500 mg/l to 20,000 mg/l.

Key Finding: Roads were direct sources of sediment delivery to streams, with approximately 34% of road drainage points entering stream channels.

Source: Bilby, R. E., K. Sullivan and S. H. Duncan. 1989. The generation and fate of road-surface sediment in forested watersheds in southwestern Washington. Forest Science 35: 453-468.

The authors studied the erosion of sediment from two kinds of forest gravel roads in southwestern Washington: heavily used, valley-bottom haul roads and midslope secondary haul roads. Sampling sites were located at the downslope of each cross-drain and at ditches draining from cut slopes. Traffic use of each road was also monitored.

The sediment produced from each road segment was related to traffic rate as well as to type of road surfacing material. The majority of the sediment produced (80%) was material finer than 0.004 mm. Steeper roads produced a higher proportion of coarser material (primarily sand). Average sediment concentrations from the secondary road sites were 2,000 mg/l, with a maximum of 19,500 mg/l. Hourly concentrations from the mainline road ranged from 500-700 mg/l, occasionally exceeding 20,000 mg/l.

Delivery of this sediment to streams was investigated by carrying out an inventory of road drainage sites in three watersheds. Two thousand drainage points, along 730 km of road, were identified. Of these, 34% directly entered streams rather than draining into the forest floor.

Key Finding: Very fine sediment washed from a forest road surface directly into a stream during rainfall events.

Source: Bilby, R. E. 1985. Contributions of road surface sediment to a western Washington stream. Forest Science 31: 827-838.

The size of sediment washing from a gravel-surfaced road and its fate after entering a small stream were examined in Johnson Creek, Washington. The study stream was a fourth-order stream, tributary to the Deschutes River. Road sediment entered the creek at a large culvert under the road. Areas upstream of the bridge (upstream of the road sediment entry point) and downstream of the bridge (downstream of sediment entry) were sampled. Freeze-core samples of sediment were taken from five salmonid spawning gravel areas above and below the bridge, three times during the year. Automatic pump samplers were used to measure suspended sediment and turbidity.

During the study period, the road had an average traffic rate of 290 axles daily, primarily logging trucks. During dry weather, there was little difference in stream turbidity upstream and downstream of the culvert. After rainfall events, sediment input from the road frequently increased the levels of suspended sediment downstream of the culvert compared to upstream levels. Maximum turbidity reached downstream was almost three times the maximum recorded upstream. Road runoff contributed a total of 20.4 metric tons of suspended sediment during the one-year study period. Sediment concentrations reached a peak during peak ditchflow and rapidly dropped off after that. Sediment was primarily very fine particles (more than 80% less than 0.004 mm in size) and was attributed to erosion from the road surface rather than roadside ditches or banks.

Key Finding: Forest road erosion was a source of fine sediment in stormflow runoff, even after mitigation measures.

Source: Swift Jr., L. W. 1984. Soil losses from roadbeds and cut and fill slopes in the Southern Appalachian Mountains. Southern Journal of Applied Forestry 8: 209-216.

The contribution of forest roads to soil erosion was investigated on a newly constructed timber sale access road in the southern Appalachian Mountains. The roadbed surface, cut slopes, and fill slopes were tested separately for soil loss. Dry weight was obtained of heavy particles deposited in collection troughs below cut and fill slopes and in stream sections ahead of each stream-gauging station. To measure lighter-weight particles, stormflow water was filtered and sediment concentrations determined by weight. Two sites were studied, one with a roadbed grade of 7% and the other with a grade of 5%.

The usual practices after road construction of grass seeding on cut and fill slopes and surfacing the roadbed with gravel were delayed for the purposes of this study. The greatest percentage of soil loss occurred during the first winter after road construction, with 42% of the total soil loss from roadbeds (tons/acre) occurring during this period, as well as 58% of the loss from fill slopes and 82% of the loss from cut slopes. Cut slopes had the highest soil erosion in the winter, due to dry ravel and frost heaving. Fill slopes had the highest erosion in early spring. Both cut and fill slopes generally experienced soil erosion of all particle sizes, while more than half the erosion from the roadbed surface was composed of finer particles. Soil erosion rates were higher on the roadbed of the steeper 7% grade site than on the 5% grade site.

After seeding and grading the road surfaces with gravel, soil loss rates were greatly reduced, especially from the grass-covered cut and fill slopes. However, some erosion from the roadbed continued. Despite the reduced erosion rates after these mitigation measures, soil loss from the entire roadway was calculated to be about 20 times the normal rate for undisturbed forest.

Key Finding: Gravel forest roads generated up to 440 tons of sediment/km/year from surface erosion.

Source: Reid, L. M. and T. Dunne. 1984. Sediment production from forest road surfaces. Water Resources Research 20: 1753-1761.

A one-year field study was conducted to determine how much sediment was generated from forest road surfaces and from ditches and cutbanks. Ten road segments were investigated in the Olympic Mountains of Washington State. Of these, eight were gravel roads and two were paved roads. Traffic use was categorized as heavy (more than four logging trucks per day), moderate (one to four trucks), light, and abandoned. During rainstorms, water discharge was measured at the mouth of each culvert and from natural lips on abandoned roads. Rainfall intensities were recorded at each sampling location.

Three factors - traffic intensity, road gradient, and road segment length - were investigated. Sediment loss was found to be related to traffic intensity and was highest on heavy-use gravel roads compared to unused roads or paved roads. Sediment yield from cutbanks and ditches alongside paved roads was less than 1% of that from gravel roads. Heavily used roads were calculated to produce 440 tons of sediment/km/yr over the period of study, compared to lightly used roads with 3.8 tons/km/yr and paved roads with 2 tons/km/yr.

Key Finding: The volume of fine sediment present in streams increased in direct proportion to logging in the watershed and stream crossings by roads.

Source: Eaglin, G. S. and W. A. Hubert. 1993. Effects of logging and roads on substrate and trout in streams of the Medicine Bow National Forest, Wyoming. North American Journal of Fisheries Management 13: 844-846.

The effects of logging and associated road construction on streams and on trout populations were studied in the Medicine Bow National Forest, Wyoming. Twenty-eight stream reaches (200 m each) were examined, with sampling conducted along transects at 4-m intervals. Trout standing stocks were estimated using a backpack electroshocker. The percentage area logged and the density of roads in areas upstream of the drainage were calculated. The density of road culverts was recorded as an index of the extent to which roads crossed watercourses within the drainage.

The amount of fine sediment in a stream reach increased, and the embeddedness of fine sediment (its coverage of large particles) in the substrate increased as the proportion of logged area increased and as the extent to which roads crossed watercourses increased. Trout standing stocks also decreased as the density of road culverts increased.

Key Finding: Logging and forest road construction led to an increase in landslides and surface erosion, disrupting the riparian vegetation along first- and second-order tributaries of a river in Oregon.

Source: Ryan, S. E. and G. E. Grant. 1991. Downstream effects of timber harvesting on channel morphology in Elk River Basin, Oregon. Journal of Environmental Quality 20: 60-72.

Aerial photos were used to investigate the downstream impacts of logging in the Elk River Basin of Siskiyou National Forest in southwestern Oregon. The river and its tributaries were reported to be good habitat for salmonid species such as chinook salmon, steelhead trout, cutthroat trout, and coho salmon.

Commercial logging in the basin began in the mid-1950s. Aerial photos from 1956, 1964, 1969, and 1979 were used to reconstruct the disturbance history of Elk River's tributaries. Openings in riparian canopies due to a disturbance were identified on first- through fifth-order channels and attributed when possible to landslides, debris flows, large floods, or excessive sedimentation from surface erosion.

In first- and second-order tributaries, the number of sections with open riparian canopies (open reaches) increased continuously through the study period. The largest increase in open reach length occurred between 1964 and 1969 after a major 1964 storm. Of the new open reaches, 74% were initiated by landslides; the majority of these were associated with clearcuts or roads. Twenty percent of the open reaches created were attributed to surface erosion; all of these were associated with clearcuts or roads.

Open reaches also occurred on higher order channels (fourth- and fifth-order) but did not show significant changes in number or size over the study period. Upslope forestry activity could not be linked, therefore, to openings downstream in these higher-order tributaries. The authors attribute this to the lack of debris flows in this system, low harvest levels before the major 1964 storm, and slope constraints.

The distribution of gravel bars in the main stem, Elk River (sixth-order), was measured to obtain a rough estimate of increases in sedimentation due to changes upstream. Overall, there was a 77% increase in the number of gravel bars over the 26-km segment of river analyzed. The authors noted that this was a qualitative indication of increased sedimentation, though their study did not demonstrate causation.

The authors note that their methods provide useful coarse-scale information on the erosional impacts of logging and roads, but only where sediment volume or management activity is extensive enough to have removed the riparian canopy.

Key Finding: Roads intercepted subsurface flow on mountainous slopes in the Idaho Batholith, converting it to surface flow.

Key Finding: Subsurface flow converted to surface flow by intercepting roads would be likely to trigger soil erosion and soil mass movement.

Source: Megahan, W. F. 1972. Subsurface flow interception by a logging road in mountains of Central Idaho. pp. 350-356 in Watersheds in Transition. Proceedings of a symposium on "Watersheds in Transition." Fort Collins, Colorado, June 19-22, 1972. AWRA. Urbana, Illinois.

The author's study site was located in the Idaho Batholith, where water in undisturbed forest rarely flows overland after a heavy rainstorm or snowmelt, but instead is primarily subsurface flow. The author measured the volume of subsurface flow intercepted by a road in two undisturbed micro-watersheds. The forest was composed primarily of ponderosa pine, Douglas-fir, and Engelmann spruce. Slopes ranged from 35% to more than 70%.

Water was collected at road cut banks where bedrock was exposed. Subsurface flow emerged on the face of the bedrock and ran down to a collection trough. Subsurface flow was therefore converted to surface flow. The author calculated that along a given length of road, the amount of subsurface flow intercepted by the road was 7.3 times greater than surface runoff from the road alone after a precipitation event.

The author discusses the impacts of converting subsurface flow to surface flow. Total watershed runoff volume probably increases. Surface flow commonly causes significant surface erosion and excess soil water can result in mass erosion. In addition, the author mentions the potential broader ecological impacts of rerouting subsurface flow from downslope habitat, such as the alteration of vegetation species composition and growth rates.

* See also key finding in Chapter 5.

Key Finding: The peak rate of subsurface flow increased by an average of 27% after clearcutting, and due to its interception by a road cut and conversion to surface flow, was believed likely to lead to increased levels of erosion from the road and the slopes below the road.

Source: Megahan, W. F. 1983. Hydrologic effects of clearcutting and wildfire on steep granitic slopes in Idaho. Water Resources Research 19: 811-819.

A paired watershed study was conducted on headwater watersheds in the Idaho Batholith to evaluate the effects of clearcut logging on the watersheds' hydrology. Data on inflow, storage, and outflow from three years before logging and three years after logging were analyzed. Within one year after logging a wildfire burned through both the logged and unlogged watersheds, so its effects were also included.

A logging road ran along the lower boundary of both watersheds. Subsurface flow, intercepted by the road cut, was collected in a trough at the bottom of the cut slope. Data on annual snow accumulation were also collected. In contrast to the unlogged watershed, the clearcut watershed had a highly significant increase in snow water content the year after logging, as well as a significant increase over the next two years after the wildfire. There was no detectable increase in snow water content on the unlogged watershed even after the fire. Subsurface outflow, as measured at the road cut, increased after clearcutting. Both the volume and peak rate of subsurface flow increased, the former by 96%, the latter by an average of 27%. Because this increased subsurface outflow was intercepted by the road cut, the author considered it likely that erosion due to surface flow along and below the road would increase.

Key Finding: Roads and clearcut logging increased peak stream discharges and advanced the timing of peak discharges in multiple paired watershed studies, most likely because of subsurface flow being converted to surface flow at road cuts.

Key Finding: Even after many years, roads and clearcut logging, both together and separately, resulted in significant increases in stream peak discharges.

Source: Jones, J. A. and G. E. Grant. 1996. Peak flow responses to clear-cutting and roads in small and large basins, western Cascades, Oregon. Water Resources Research 32: 959-974.

The authors studied paired watersheds in the western Cascades and examined road building, logging, and peak discharge records to compare streamflow peaks pre- and post-treatment. Records for two pairs of small basins extended over 34 years, and records for three adjacent large basin pairs extended over 50 to 55 years.

One of the small watersheds was 100% clearcut without road construction. After clearcutting, a significant number of storms resulted in higher peak discharges and volumes, and began earlier. A higher-than-expected number of runoff events had greater peaks and volumes. Sixteen to 22 years after clear-cutting, average peak discharges were still significantly higher (almost 40%) than pre-logging levels.

The second small basin provided four years of data on the impact of roads alone, before logging began. Roads occupied 6% of the watershed. After road construction, a higher-than-expected number of storm events had higher peak discharges and began earlier. After clearcutting 25% of the watershed, average peak discharge increased by 50% in the first five years, and storm discharges began an average of six hours earlier than pre-treatment. After 25 years, average peak discharges were still significantly (more than 25%) higher than pre-management levels.

Similarly, in the three large basin pairs, peak discharge increased as cumulative area logged increased. Begin times were not reported.

The authors note that the most likely mechanism for the increase in peak flow due to just roads was road cuts converting subsurface flow to surface flow, which was then routed directly to stream channels. Logging, they conclude, had an impact on streamflow due to changes in evapotranspiration and snow accumulation and melt rates.

Key Finding: Roads formed new surface flow paths to natural channels and incised new gullies, so increasing the routing efficiency of water; thereby probably explaining some higher stream peak flows.

Source: Wemple, B. C., J. A. Jones and G. E. Grant. 1996. Channel network extension by logging roads in two basins, western Cascades, Oregon. Water Resources Bulletin 32: 1195-1207.

Two fifth-order basins (Lookout Creek and Blue River), in the western Cascades of Oregon, were studied to determine the mechanism by which logging roads may alter stream peak flows by changing water routing efficiency. The road density in each basin was 1.9 km/km2, and roads occupied 3% of each basin's area.

A sample of 62 km of the road network was surveyed. A total of thirty-one 2-km transects was selected, and the transects were subdivided into segments at each culvert. Study sites were distributed between valley, midslope, and ridgetop sites and among roads ranging in construction period from the 1950s to the 1990s. A subsample was also studied immediately after storm events.

Road culverts delivered water to natural stream channels at stream crossings, into new gullies incised below culvert outlets, or onto hillslopes, where water reinfiltrated the soil. The first two mechanisms of surface flow linked the roads directly to the stream channel network. More than 57% of the total road length surveyed was calculated to be connected to the stream network by these two flowpaths. Of the 436 culverts examined, 33% crossed streams and 23% were ditch-relief culverts with gullies incised below. Thirty-four percent of the road length drained to stream channels and 24% drained to gullies. Of the gully-forming culverts studied immediately after storm events, approximately half directed surface runoff to a nearby channel or saturated area. The authors estimated that these new flowpaths due to roads resulted in an increased drainage density of 36% and 39% in the two basins, although they noted that these figures would probably vary by season and by the degree to which gullies were connected to streams.

The authors describe an earlier study in the same basins (Jones and Grant 1996), in which stream peak flows increased after road construction and logging. They hypothesize that the present study, documenting the integration of roads with stream networks and the increased drainage density in the two basins, was a possible mechanism for increased water routing efficiency and therefore increased peak flows.

Key Finding: Almost 30 years after clearcut logging occurred, average and peak stream flows in the watershed studied were still higher than pre-logging flows.

Source: Troendle, C. A. and R. M. King. 1985. The effect of timber harvest on the Fool Creek Watershed, 30 years later. Water Resources Research 21: 1915-1922.

The Fool Creek watershed study was part of a paired watershed experiment in the Rocky Mountains started in the 1940s. More than 10 years of data on pre-logging streamflow were available. The study evaluated the effects of clearcut logging on streamflow and snowpack accumulation for a 28-year period after the first trees were cut.

Streamflows from April to September increased after logging, with the average annual increase being 40%, or 8.2 cm, higher than that expected based on pre-logging data and the control watershed. Twenty-eight years after logging, the increase observed was still high (10.2 cm). Peak mean daily discharge also increased by an average of 23%, or about 55 l3/s, after 28 years, and peak mean daily discharge occurred about 7.5 days earlier in the year. These increases were attributed to snow melting sooner in exposed, logged areas and less water being used for soil saturation. Peak water equivalent (total snowmelt) of the entire watershed increased by 9%, a significant change. The authors review other studies from sites near their watershed, showing that even partial cutting of trees resulted in an increase in the overall deposition of snow and less loss of moisture through interception by trees. The authors conclude that one third of the increase in streamflow may be due to increased peak water equivalent and note that vegetation regrowth over the past 28 years has had little effect.

Key Finding: Natural streamflow rates during periods of high flow were significantly altered in two watersheds after logging road construction.

Key Finding: Subsurface flow intercepted by logging roads was converted to surface flow and was the most likely cause for increases in streamflows during snowmelt runoff and heavy summer storms.

Source: King, J. G. and L. C. Tennyson. 1984. Alteration of streamflow characteristics following road construction in north central Idaho. Water Resources Research 20: 1159-1163.

The effects of logging roads on streamflow were monitored on six headwater watersheds in Nez Perce National Forest in north central Idaho, as part of the Horse Creek project. Watersheds ranged from 28 to 148 ha in size. The study was conducted after logging roads were constructed, but before the logging occurred. Daily water discharge data were collected at the mouths of the watersheds and compared with pre-road data for the same watersheds and with a control watershed that had no roads.

Two of the six watersheds showed significant changes. In one watershed, with 3.9% of its area disturbed by roads, there was an increase in the 25% exceedance flows (streamflow during snowmelt runoff and summer storms). The authors attributed this increase in streamflow to interception of subsurface flow by the roads and conversion to surface flow. A majority of the road length had cut slopes greater than 6 m in height, and being located midslope, the road had the potential to intercept upslope flow from 67% of the watershed's area. The authors observed high interception of subsurface flow along one of the cut slopes, with long periods of flow in the ditches.

Another watershed, with 4.3% of its area in roads, showed a significant decrease in the 5% exceedance flow, which represents the period of highest flow (usually snowmelt runoff). The other four watersheds showed no significant change in the any of the seven streamflow variables.

Key Finding: Stream peak flows increased significantly in a watershed with 12% of its area in roads, before any logging occurred.

Key Finding: Stream peak flows increased as the percentage of watershed area clearcut increased.

Source: Harr, R. D., W. C. Harper and J. T. Krygier. 1975. Changes in storm hydrographs after road building and clear-cutting in the Oregon Coast Range. Water Resources Research 11: 436-444.

Six small watersheds were studied in the Alsea River basin of the Oregon Coast Range. Changes in stormflow after road building, logging, and slash burning were compared to pre-logging data and to a control watershed. Storm events were monitored over three years of rainy seasons, with fall data separated out from winter data. Only one year of data was available for changes due to roads alone, before logging commenced.

After road construction, but prior to logging, peakflow volumes increased significantly in only one watershed. Average peakflows increased (by 5 ft3/s/mi2) in this watershed, which had the greatest extent of roads (12% of its area). Peakflow changes in the other watersheds were smaller and inconsistent. Data for streamflow were limited, however, since there was only one year of information.

After clearcutting, three of the five logged sites showed significant increases in peak flow, with the largest increases being in the watersheds with the greatest area clearcut. For example, fall peak flow increased (by 27 ft3/s/mi2) in the watershed that was 90% clearcut.

Key Finding: Forest roads extended the natural channel network, initiated new channels, and increased the susceptibility of steep slopes to landsliding.

Key Finding: Road cuts intercepted subsurface flow and diverted it to roadside ditches.

Source: Montgomery, D. R. 1994. Road surface drainage, channel initiation, and slope instability. Water Resources Research 30: 1925-1932.

Field surveys were conducted at three sites in the western United States to investigate road drainage and associated landsliding and channel network extension. The study sites were located in 1) the southern Sierra Nevada; 2) on Mettman Ridge in the Oregon Coast Range; and 3) on Huelsdonk Ridge on the Olympic Peninsula. Drainage area and slope were determined to be the key criteria contributing to slope instability (so leading to landslides) and initiation of new water channels. The author mapped all discharge points from the roads and estimated the contributing drainage area. In each area, average ground slopes were also measured.

In the southern Sierra Nevada site, road drainage resulted in the road surface acting as an extension of the natural channel network. Road cuts had diverted both surface and subsurface flow into ditches. Four hollows had lost natural drainage waters due to diversion by the roads. Three different hollows received extra drainage from the road system. The overall drainage density of the area studied (1.2 km2) had increased by a factor of 1.6.

Forest roads studied in Oregon and Washington were both ridgetop roads. Roads had initiated new channels. Road-associated landsliding was highest on the steepest slopes and on slopes having the greatest drainage area. Drainage density due to new water flowpaths increased by a factor of 1.23 at the Oregon study site; no figure was reported for the Washington site. Road discharge points were studied immediately after rainfall only at Oregon site. At other sites, the author estimate that mapping accuracy of drainage areas was within +/- 30%.

Adverse impacts on aquatic species: Increased fine sediment deposition in streams and altered streamflows and channel morphology result in increased adult and juvenile salmonid mortality, a decrease in aquatic amphibian and invertebrate abundance or diversity, and decreased habitat complexity.

Key Finding: Salmonid survival rates decreased after logging and road construction as fine sediment levels in streams increased and as important habitat characteristics, including the number of pools and winter cover, decreased.

Source: Hicks, B. J., J. D. Hall, P. A. Bisson and J. R. Sedell. 1991. Responses of salmonids to habitat changes. In Influences of forest and rangeland management on salmonid fishes and their habitats. American Fisheries Society Special Publication 19: 483-518.

The authors review research on the effects of logging on salmonids and their stream habitat. Studies from Oregon, Idaho, British Columbia, and Alaska, for instance, showed that salmonid abundance and fry survival decreased as fine sediment levels increased after logging. Fine sediment in deposits or suspension also reduced the availability of food in streams by reducing invertebrate abundance and primary production. Suspended sediment increases were shown to affect salmonids in various ways, including avoidance, cessation of feeding, and disrupted social behavior.

The increased frequency of landslides and other mass erosion events due to logging and roads changed channel morphology, reducing pool area and depths and resulting in stream reaches that were wider, shallower, and more prone to bank erosion. Studies in British Columbia, for instance, showed that pool habitat was reduced by an average of 79% in streams affected by debris torrents and suitable winter cover was reduced by an average of 75%. Coho salmon winter survival averaged 1.8% in stream reaches affected by debris torrents compared to survival rates of 24.5% in unaffected streams.

The authors discuss studies showing salmonid abundance initially increasing after clearcutting. They note that these increases were documented only over the short term and that over the longer term (after 10 to 15 years), other research had indicated that populations could eventually decline to levels lower than those in old-growth forest.

Key Finding: Coho and chum salmon fry survival declined after logging and associated increases in fine sediment deposited in spawning areas.

Source: Scrivener, J. C. and M. J. Brownlee. 1989. Effects of forest harvesting on spawning gravel and incubation survival of chum (Oncorhynchus keta) and coho salmon (O. kisutch) in Carnation Creek, British Columbia. Canadian Journal of Fisheries and Aquatic Sciences 46: 681-696.

The effect of logging practices on salmonid spawning gravel composition and fry survival was investigated in Carnation Creek on the west coast of Vancouver Island. The study included pre-logging, logging, and post-logging data. Three clearcut logging treatments were conducted at varying distances along the stream: 1) clearcutting with a strip of vegetation left along the streamside; 2) intense streamside treatment; and 3) careful streamside treatment with some shrubby vegetation left intact. Permanent survey sites were established at 3-m intervals in each study section. Streambed cores were obtained from each of these sites using a freeze-core technique, driving steel probes 30 cm into the bed. Peak flows and suspended sediment levels were also measured. Annual egg-to-fry survival rates for coho and chum salmon were estimated by calculating egg deposition rates from adult female numbers and sizes and by counting emerging salmon fry at traps at a downstream fence.

After logging, the percentage of fine sediment increased in the streambeds, although the patterns of deposition and proportion varied among treatments and timing of logging. These fines particles appeared to originate from erosion of streambanks after the loss of living roots from streamside vegetation and the loss of large organic debris post-logging. They were transported through the creek primarily as bedload rather than as suspended sediment. Streambank erosion increases were highest in the intense streamside treatment and lowest in the treatment with a buffer strip.

After logging and a subsequent large snowmelt event, coho salmon fry survival to emergence rates declined to 16.4%, compared to a prior survival rate of 29.1%. The decline was correlated to decreasing mean particle sizes in the lower layers of the streambed cores. Survival to emergence of chum salmon fry declined from a prior rate of 22.2% to 11.5% post-logging and was correlated to decreasing mean particle size in the whole streambed core and in the top layers of the core. Peak survival occurred during years when pea gravel and sand were washed out from the top layer. The authors attribute the difference in streambed layers affecting chum and coho survival to differences in natural egg deposition depths between the two species.

Key Finding: Survival rates of coho salmon and steelhead trout fry decreased as the proportion of fine sediment in spawning gravel increased.

Source: Phillips, R. W., R. L. Lantz, E. W. Claire and J. R. Moring. 1975. Some effects of gravel mixtures on emergence of coho salmon and steelhead trout fry. Transactions of the American Fisheries Society 3: 461-466.

Laboratory experiments were conducted at the Alsea Watershed Study field station to investigate the relationship between the proportion of fine sediment in spawning gravel and the survival of coho salmon (Oncorhynchus kisutch) and steelhead trout (Salmo gairdneri) fry. Six different gravel sizes were mixed in troughs to create spawning gravel similar in composition to natural coho salmon redds in Deer Creek, in the Oregon Coast Range. The proportion of fine sediment (sand 1-3 mm in diameter) was then increased by 10% increments to create eight gravel mixtures with 0-70% sand by volume. Coho salmon and steelhead fry were buried in the gravel, and their date of emergence, survival, and weight were recorded. Six replicates were tested.

As the proportion of fine sediment in the gravel mixtures increased, coho salmon fry emerged earlier and were smaller in size. Their survival rates decreased as fine sediment percentage increased, from 96% survival in the control gravel mixture to 8% survival in the mixtures containing 70% sand. Fine sediment proportions had no effect on the timing of steelhead fry emergence. However, their survival patterns were similar to those of coho salmon fry, with 99% survival for steelhead fry in the control mixture and 18% in the 70% sand mixture.

The authors note that sediment sizes smaller than 1 mm were not tested in their experiment and that total emergent fry survival could be even lower under conditions that included finer sediment. They also note that if fish were exposed to high sediment levels for a longer time period, from egg fertilization through development, mortality due to indirect effects such as low oxygen concentrations could be higher.

Key Finding: Brook trout populations declined significantly after stream sedimentation levels increased.

Key Finding: Populations of stream benthic invertebrates (the major food source of brook trout) declined significantly after stream sediment levels increased.

Key Finding: Higher fine sediment levels in a stream resulted in a loss of pool habitat, fish cover, changes in stream velocity, and higher summer water temperatures.

Source: Alexander, G. R. and E. A. Hansen. 1986. Sand bed load in a brook trout stream. North American Journal of Fisheries Management 6: 9-23.

The effects of sedimentation on populations of brook trout (Salvelinus fontinalis) and stream channel physical characteristics were investigated over a period of 15 years in Hunt Creek in the Lower Peninsula of Michigan. Trout populations were monitored for five years prior to sand deposition, for five years during which sand was introduced into the stream, and then five more years without adding sand.

The study area was divided into two 1-mile sections, with the upper section of the stream serving as a control throughout the study. For five years, sand was introduced daily into the treated section of the stream, increasing total sediment concentrations from approximately 20 ppm to 80 ppm to replicate concentrations reported for trout streams with severe streambank erosion. Cross sections were established at 100-ft intervals to document changes in stream channel characteristics. Brook trout were collected from spring through fall every year, as were samples of benthic invertebrates (their primary food source).

The volume of sand deposited on the streambed gradually increased over the study period. A significant decrease occurred in brook trout populations in the treated section of the stream, a decrease particularly evident four years after the initial introduction of sand. Total trout numbers dropped by 51%, a statistically significant change. Trout of all sizes and ages declined in number in the sand-treated section compared to the control section of the stream. There was no change in growth rates.

After sand introduction, populations of benthic invertebrates also dropped to less than half their pre-treatment populations. The insect orders of Ephemeroptera, Diptera, Coleoptera, Trichoptera, and Plecoptera showed the most significant declines. Fish stomach analyses revealed that the majority of these taxa were important food sources for brook trout.

Stream physical characteristics also changed with increased levels of sedimentation. The stream became wider and shallower, pools disappeared, and the stream bottom lost all fish cover after becoming uniformly covered by sand. Water temperatures in the summer increased. Deeper stream depths near the banks disappeared.

Key Finding: Salmonids avoided water with suspended sediment in Alaskan streams and lakes.

Source: Lloyd, D. S., J. P. Koenings and J. D. LaPerriere. 1987. Effects of turbidity in fresh waters of Alaska. North American Journal of Fisheries Management 7: 18-33.

The authors review research on the effects of turbidity (due to suspended particles) in Alaskan streams and lakes. They note that there are natural sources of particles from glacial meltwater, as well as unnatural sources from placer mining, logging, and road construction. Primary production (algae and vascular plant growth), the foundation of the aquatic food chain, was found to be positively related to the clarity of streams. The authors review studies indicating that various salmonid species, including Arctic grayling, coho salmon, chinook salmon, and rainbow trout, avoided turbid waters. This avoidance was attributed to their reduced ability to find food, as well as interference with visual cues during migration.

Key Finding: Delivery of fine sediments to streams and deposition on spawning and rearing substrate decreased after a moratorium on logging, but increased again after logging resumed.

Source: Platts, W. S., R. J. Torquemada, M. L. McHenry and C. K. Graham. 1989. Changes in salmon spawning and rearing habitat from increased delivery of fine sediment to the South Fork Salmon River, Idaho. Transactions of the American Fisheries Society 118: 274-283.

The effects of fine sediment delivery to rivers from logging and road construction were studied in habitat for chinook salmon (Oncorhynchus tschawytscha) and steelhead (O. mykiss, formerly Salmo gairdneri). Spawning and rearing areas were studied after a logging moratorium was declared in the watershed of the South Fork Salmon River, which drains part of the Idaho Batholith. The authors reported results of earlier studies indicating high levels of fine sediment present in the river due to accelerated erosion from logging, road construction, large storm events, and road washouts.

Ten transects were established at each of five chinook salmon spawning areas, and substrate characteristics were measured for 20 years. After logging ceased, there was a significant decline in the percentage of fine sediment (material <4.75 mm in diameter) on the surface of 84% of the spawning area locations. Overall sediment declines over the 20 years varied at each of the five spawning areas, but ranged from a decrease by 16.7% at one area to a decrease by 76.5% at another. The percentage of gravel and rubble correspondingly increased. Within two years of resuming logging, however, surface fine sediments increased at all five spawning areas, with overall increases of 22.2% to 83.8%.

In salmon rearing areas, transects were established at 15-m intervals at 47 sample stations. Data were collected from these areas for six years. The percentage of fines on the surface of rearing areas decreased by 73.5% over the study period. Overall, rearing areas had lower levels of fine sediment deposition from logging than spawning areas did.

Key Finding: Fine sediment deposition on cobble substrates decreased the availability of interstitial spaces (used as winter refuges), and winter densities of juvenile chinook salmon decreased correspondingly.

Source: Hillman, T. W, J. S. Griffith and W. S. Platts. 1987. Summer and winter habitat selection by juvenile chinook salmon in a highly sedimented Idaho stream. Transactions of the American Fisheries Society 116: 185-195.

The authors investigated the effect of fine sediment on juvenile chinook salmon (Oncorhynchus tschawytscha). One particular criterion they looked at was the impact of fine sediment deposition on winter survival. Two sites on Red River, Idaho, were modified in September 1985 by placing cobble in randomly located 1-m2 plots center stream and under the banks. The cobble was expected to provide more refuge areas for salmon during the winter. Fish populations were estimated by three-pass electrofishing and by snorkeling several times through the winter.

Salmon winter rearing densities increased eightfold in glide areas (slow, shallow areas) after cobble was added, compared to densities the previous year. Densities in treated areas were nine times higher than densities in untreated areas during the same period. A significantly higher density of young chinook salmon (five times higher) used interstitial spaces in the altered areas than in the unaltered areas. By March 1986, however, when the cobble had become heavily embedded with fine sediment, juvenile salmon densities had decreased by more than 90% and were similar to densities pre-alteration.

Key Finding: Chum salmon eggs were susceptible to mortality from increased streambed scour associated with logging and/or roads.

Source: Montgomery, D. R., J. M. Buffington, N. P. Peterson, D. Schuett-Hames and T. P. Quinn. 1996. Stream-bed scour, egg burial depths, and the influence of salmonid spawning on bed surface mobility and embryo survival. Canadian Journal of Fisheries and Aquatic Sciences 53: 1061-1070.

Spawning by chum salmon (Oncorhynchus keta) in streambed gravel was studied in a stream near Juneau, Alaska. Egg burial in summer was followed by a winter incubation period, during which the stream experienced the greatest discharge volumes due to high rainfall. The depth to which egg pockets were buried in the gravel was measured in 40 salmon redds (nests). To measure the depths to which the streambed was scoured during winter peak flows, 104 scour monitoring chains were distributed throughout the study section of the stream.

Depths of bed scour and egg pockets varied, with scour depths ranging from 0-60 cm (mean of 13.4 cm) and egg pocket depths ranging from 9.8-48.9 cm (mean of 22.6 cm). The authors found that for the majority of egg pockets, burial depth was just enough to protect them against natural scour during the peak winter discharge. The authors indicate that this protection was most likely a result of finely tuned adaptation by the salmon to natural rates of sediment transport and scour depths.

They note research by other authors reporting that increases in scour depths were related to increases in stream discharge and velocity and increases in fine sediment transport. The authors therefore conclude that increases in scour due to increased sedimentation from logging or roads could significantly increase the mortality of buried salmon eggs.

Key Finding: Salmonid embryo survival rates decreased as the proportion of fine particles in stream spawning substrate increased and dissolved oxygen levels decreased.

Source: Chapman, D. W. 1988. Critical review of variables used to define effects of fines in redds of large salmonids. Transactions of the American Fisheries Society 117: 1-21.

The author reviews laboratory and field studies on salmonid embryo survival. The majority of studies showed that salmonid embryo survival rates decreased as the percentage of fine sediments in stream substrate increased. With increasing fine sediment levels, dissolved oxygen levels decreased, as did gravel permeability and pore space. Dissolved oxygen levels were found to be critical to the survival of embryos and their later development. Size of emergents was also found generally to decrease as fine sediment levels increased.

Despite the variability among studies in quantitative results, they consistently showed the adverse impacts of fine sediments on salmonid survival.

Key Finding: Adult and juvenile salmonids exposed to suspended fine sediment in streams had an increasingly negative response as concentrations and duration of exposure increased.

Source: Newcombe, C. P. and J. O. T. Jensen. 1996. Channel suspended sediment and fisheries: a synthesis for quantitative assessment of risk and impact. North American Journal of Fisheries Management 16: 693-727.

The authors review 80 published studies on the response of fish to suspended sediment in streams. Data from these studies were used to develop models quantifying the response of fish to varying sediment concentrations and varying durations of exposure. This response was defined as "severity of ill effect," which included effects such as reduced growth rates, reduced fish density, reduced fish population size, and habitat damage. The data were also used to provide estimates of the onset of sublethal and lethal effects in fish.

Data were grouped into six subcategories based on species, age, and sediment size. Adult and juvenile salmonids exposed to particle sizes of 0.5-250 (m showed an increasingly negative response as sediment dose increased, and sublethal and lethal effects occurred at high doses. The equations derived for the model were tested against newer data and validated.

Key Finding: Juvenile coho salmon avoided water with high turbidity levels.

Source: Bisson, P. A. and R. E. Bilby. 1982. Avoidance of suspended sediment by juvenile coho salmon. North American Journal of Fisheries Management 4: 371-374.

An aquarium experiment was designed to investigate the effect on juvenile coho salmon of suspended sediment from logging roads. Young coho salmon were collected from a tributary of the Deschutes River, Washington, and fine sediment was collected from a catchment basin next to a heavily used, unpaved road. One group of fish was acclimatized to clear water, the other to water with a low turbidity, to replicate natural winter turbidities in this tributary. After being held for three weeks, 10 individuals from each group were placed in an aquarium with suspended sediment added to one side only. Fish could move freely and their preference for a particular side of the chamber was then monitored.

Both groups of fish avoided turbid waters once a threshold of tolerance had been reached. This threshold was lower for the salmon acclimated to clear water than to salmon acclimated to slightly turbid water. The exception was fright behavior, when, due to a lack of any other cover, fish grouped in the turbid portion of the chamber.

Key Finding: Basins with more than 25% of their area logged had lower stream habitat diversity, as measured by the number of pools and pieces of wood, than basins with less than 25% of their area logged.

Key Finding: The diversity of juvenile anadromous salmonid populations was lower in basins with more than 25% of the area logged than in basins with less than 25% logged.

Source: Reeves, G. H., F. H. Everest and J. R. Sedell. 1993. Diversity of juvenile anadromous salmonid assemblages in coastal Oregon basins with different levels of timber harvest. Transactions of the American Fisheries Society 122: 309-317.

The diversity of juvenile anadromous salmon populations was examined in relation to the extent of logging in coastal Oregon stream basins. Fourteen stream basins were sampled. Basins with 25% or less of the watershed area logged were classified as having low logging levels. Basins with greater than 25% of their area logged were classified as having high logging levels.

As an index of stream habitat complexity, pools and number of pieces of wood per 100 m of stream length were counted. This data was collected for three paired streams, selected to minimize variation in other stream characteristics. Juvenile anadromous salmon were sampled in all streams over five years by divers. Species identified were chinook salmon (Oncorhynchus tshawytscha), coho salmon (O. kisutch), steelhead (O. mykiss), and cutthroat trout (O. clarki). Counts were used to estimate relative population numbers.

In the paired comparisons, streams in low-harvest basins had higher habitat diversity than streams in high-harvest basins. Number of pieces of wood per 100 m was significantly higher (2-12 times higher) in low-harvest basins and number of pools was significantly higher in two out of three comparisons, ranging from 10-47% higher.

Salmon species dominance was used as the primary attribute of community diversity. Juvenile anadromous salmonid assemblages were significantly more diverse in low-harvest basins, with single species dominating the fish assemblage more often in high-harvest basins than in low-harvest basins. There were no differences in total mean densities of fish between different logging levels.

Key Finding: The density of all three stream amphibian species studied was lower in streams affected by sediment due to road construction than in control streams. Two of three species had significantly lower numbers in all five stream microhabitats.

Source: Welsh, H. and L. M. Ollivier. 1998. Stream amphibians as indicators of ecosystem stress: a case study from California's redwoods. Ecological Applications 8: 1118-1132.

The authors analyzed the impact of highway construction and resulting erosion on the abundance of stream amphibians in California old-growth redwood forest. A major storm during road construction resulted in large volumes of sediment from mass wasting and surface erosion entering stream channels. Five streams affected by sediment were compared with five control streams in the same basin. The three most abundant native amphibians were sampled - larval Pacific giant salamanders (Dicamptodon tenebrosus), larval tailed frogs (Ascaphus truei), and larval and adult southern torrent salamanders (Rhycotriton variegatus). Salamander densities were surveyed in transects placed throughout more than 3 km each of affected stream habitat and control stream habitat. Different habitat types were sampled, including pools, glides/runs, riffles, step runs, and step pools.

A total of 267 transects, 0.6 m wide, was sampled, with 540 individual amphibians captured. The density of Pacific giant salamanders and southern torrent salamanders was significantly lower in the sedimented than in the control streams. The density of tailed frogs was lower in their preferred riffle and step run habitat in sedimented streams as opposed to control streams, although results were not statistically significant.

Key Finding: A higher proportion of fine sediment occurred in streams flowing through forest stands with logging than streams flowing through unlogged forest stands.

Key Finding: Abundance, density, and biomass of all aquatic amphibian species studied were lower in streams flowing through logged forest than unlogged forest streams.

Source: Corn, P. S. and R. B. Bury. 1989. Logging in western Oregon: responses of headwater habitats and stream amphibians. Forest Ecology and Management 29: 39-57.

The authors compared the occurrence and abundance of amphibians in streams flowing through unlogged forest versus streams flowing through forests with prior logging in Oregon's Coast Range. Stream segments 10 m long were examined in 43 streams, which included first-, second-, and third-order streams. Stream segments were chosen partly based on their accessibility and partly on how typical they appeared of the rest of stream. Twenty-three streams were in unlogged forest stands ranging in age from 60 to 400 years old. Twenty streams were in logged stands ranging in age from 14 to 40 years old, with canopy cover having been reestablished over the streams. No study sites had road crossings upstream of them. Streams were surveyed from June to early August for two years. Stream physical characteristics were recorded and thorough searches conducted for all amphibians.

Results were analyzed for the four amphibian species reported to be the most common and the dominant vertebrates of small streams in the Oregon Coast Range: tailed frogs (Ascaphus truei), Pacific giant salamanders (Dicamptodon ensatus), Olympic salamanders (Rhyacotriton olympicus), and Dunn's salamanders (Plethodon dunni). All four species occurred more frequently and had higher density and biomass in the streams flowing through unlogged as opposed to logged forest stands. The only stream physical variable found to be significantly different between stand treatment was that streams in logged stands had more fine sediment.

Key Finding: Stream insects preferred fully exposed cobble on the streambed to cobble partly or fully embedded in fine sediment.

Source: Brusven, M. A. and K. V. Prather. 1974. Influence of stream sediments on distribution of macrobenthos. Journal of the Entomological Society of British Columbia 71: 25-32.

Aquatic insects in the orders of Ephemeroptera, Plecoptera, Trichoptera, and Diptera were studied in the laboratory and in natural streams to determine what streambed material they preferred. Five species were collected in the field (Pteronarcys californica, Arctopsyche grandis, Ephemerella grandis, Brachycentrus sp., and Atherix variegata) and introduced into artificial streams. Substrates of different particle sizes were placed in test quadrants of the stream.

All five species preferred fully exposed cobble to cobble half-embedded in fine sand. Likewise, in natural streams, four of the five species occurred in rocky streams, where they lived on the surface of coarse substrate materials or used the spaces between and below for retreat. One species, Atherix variegata, lived in a wider range of substrate habitat.

Key Finding: Adult aquatic insects were an important part of the insect community in forests adjacent to streams, and were believed to be an important part of the food web for forest animals such as birds and bats.

Source: Jackson, J. K. and V. H. Resh. 1989. Distribution and abundance of adult aquatic insects in the forest adjacent to a northern California stream. Environmental Entomology 18: 278-283.

The authors studied the distribution of adult aquatic insects in mixed evergreen forest adjacent to a stream in the northern California Coast Range. Sticky traps were set up in a 100-m section of streamside forest for 26 days, during the period of peak aquatic insect emergence from the stream. Traps were hung in oak trees that were 5, 40, and 150 m from the stream and at heights of 2, 5, and 8 m above the ground.

A total of 5,402 individuals, including insects in 27 aquatic insect taxa were collected. Near the stream, up to 40 m into the forest, adult aquatic insects represented more than 30% of the total arthropod numbers and 25% of total arthropod biomass. Further from the stream, at 150 m, aquatic insects represented 15% of total arthropod numbers and 11% of total arthropod biomass.

The authors note that adult aquatic insects are an important link between aquatic and terrestrial habitat. They were a significant part of the arthropod community even 150 m away from the stream and may be an important part of the land food web as demonstrated by research on songbird and bat diets.

Key Finding: The remaining intact watersheds in southeast Alaska are key to maintaining sustainable salmon stocks.

Source: Bryant, M. D. and F. H. Everest. 1998. Management and condition of watersheds in southeast Alaska: the persistence of anadromous salmon. Northwest Science 72: 249-267.

The authors review the history of forest management practices in the Tongass National Forest since the 1950s. They note that, although logging practices have greatly improved during the past 40 years, most logging in the Tongass occurred before riparian protection policies were in place. As a result, there is a legacy of watersheds with degraded salmonid habitat. Although there is a lack of systematic data available on salmonid populations pre- and post-logging in Alaska, the authors predicted that logged watersheds would be less resilient to environmental stresses than intact watersheds and that salmonid populations would therefore be more vulnerable to environmental disturbances such as decreased marine survival, drought, landslides, flooding, etc. They also note studies on the contribution of large trees to stream channels that have shown that stream habitat deterioration may not be apparent for decades after logging and that habitat quality is unlikely to be restored for more than 100 years after logging ceases. The authors therefore conclude that streams in unmanaged, intact watersheds of southeast Alaska are critical to maintain sustainable salmon stocks.

Key Finding: Trout standing stocks decreased as the density of road culverts (a measure of the extent to which roads crossed watercourses) increased.

Source: Eaglin, G. S. and W. A. Hubert. 1993. Effects of logging and roads on substrate and trout in streams of the Medicine Bow National Forest, Wyoming. North American Journal of Fisheries Management 13: 844-846.*

The effects of logging and associated road construction on streams and on trout populations were studied in the Medicine Bow National Forest, Wyoming. Twenty-eight stream reaches (200 m each) were examined, with sampling conducted along transects at 4-m intervals. Trout standing stocks were estimated using a backpack electroshocker. The percentage area logged and the density of roads in areas upstream of the drainage were calculated. The density of road culverts was recorded as an index of the extent to which roads crossed watercourses within the drainage. The amount of fine sediment in a stream reach increased and the embeddedness of fine sediment (its coverage of large particles) in the substrate increased as the proportion of logged area increased and as the extent to which roads crossed watercourses increased. Trout standing stocks also decreased as the density of road culverts increased.

* See also key finding above.

Key Finding: During the summer, adult chinook salmon preferred pool habitat and cool stream reaches over other kinds of stream habitat.

Source: Torgersen, C. E., D. M. Price, H. W. Li and B. A. McIntosh. 1999. Multiscale thermal refugia and stream habitat associations of chinook salmon in northeastern Oregon. Ecological Applications 9: 301-319.

The distribution and behavior of adult spring chinook salmon relative to stream temperature and physical habitat was investigated in the Middle Fork and North Fork of the John Day River basin, northeastern Oregon. Salmon were counted by divers during July and August in selected stream sections, their habitat use recorded, and their location recorded using a global positioning system. Water temperature patterns were measured by remote sensing and by handheld thermometers. Temperature maps were compared with salmon locations using a geographical information system (GIS).

The total number and average density of chinook salmon was greater in the North Fork, which was relatively undisturbed, than in the Middle Fork, which had relatively intense land use including logging, grazing, and agriculture. Chinook salmon were found to use pools more frequently than riffles and in a proportion greater than the pools' availability. During this study period, 98% of the salmon were found in pools in the Middle Fork, and 82% were found in pools in the North Fork. The distribution of salmon was also strongly associated with temperature. Overall, salmon indicated a preference for stream reaches that were cooler and had greater pool volume.

Get Updates and Alerts

See the latest issue >

Give the Gift That Will Make a Difference: Leader of the Pack

NRDC Gets Top Ratings from the Charity Watchdogs

Charity Navigator awards NRDC its 4-star top rating.
Worth magazine named NRDC one of America's 100 best charities.
NRDC meets the highest standards of the Wise Giving Alliance of the Better Business Bureau.

Donate now >

Related Stories

Q&A: Documentary Filmmaker Ken Burns on National Parks
Ken Burn spoke to OnEarth about his motivation for his new documentary series on America's national parks.
In the Canadian Boreal Forest, a Conservation Ethic at Work
After fighting successfully for years to keep destructive logging, hydropower and mining projects out of their traditional territory, the people of Poplar River are now working to secure permanent protection for their boreal forest homeland.
Shop Smart, Save Forests
Share | |
Find NRDC on