The following is a position paper from the National Marine Fisheries
Service (NMFS) analyzing Oregon's Forest Practices Act, the laws governing private
industrial land logging. NMFS is a group of federal scientists in charge of protecting
endangered salmon species. Jim Brown, Oregon's state forester, told me that he thought
there was nothing wrong with Oregon's Forest Practices, and they have not contributed
at all to the decline of fish. He thinks the following NMFS conclusions are flawed.
Judge for yourself - who's right?
The ODF uses a process that is not spelled out in the Rules to identify both general areas and specific sites that may be high risk for future landslides. There are specific rules to minimize erosion and soil-surface disturbance associated with timber harvest, but these are associated only with roads constructed or in use since 1972. There are also rules that govern logging and road construction located directly on high-risk sites. However, there are no provisions to avoid logging or road construction on high-risk sites. This is a serious deficiency in the Rules, because landslides can add significant amounts of fine sediment to streams and can result in increased direct mortality to salmon through burial of redds and eggs. In extreme cases, landslides can also cause other significant effects to salmon such as blocking upstream spawning migration or severely damaging instream rearing habitat.
The Rules do not address the fact that soil shear strength decreases as tree roots decay following logging. This is a major factor controlling slope stability and it is largely affected by logging. Several syntheses of research (Swanston 1969, Burroughs and Thomas 1977, Krogstad 1995) found that lateral root reinforcement provides the only available reinforcement in deeper soils where most roots cannot reach to bedrock. These deep soils areas tend to be located in areas of convergent topography which feed directly into the stream channel network. Deep forest soils are widespread in coastal Oregon.
It may take several decades to recover soil shear strength after logging, depending on the local vegetation species and many other variables (Ziemer 1981). In general soil instability due to logging peaks at about nine years. Selective harvest is the only logging method that is likely to avoid substantially diminishing root reinforcement in conifer forests (Kragstad 1995). Both clear-cut and shelterwood harvest are likely to prevent sites from recovering 80% of the root reinforcement for at least 15 to 20 years, even assuming prompt reforestation (Burroughs and Thomas 1977, Ziemer and Swanston 1977).
The Rules for minimizing slope disturbance lack specificity, and the likely outcomes of Best Management Practices (BMPs) are difficult to evaluate. ODF has not yet demonstrated that the Rules can effectively prevent or minimize soil and debris delivery into streams.
The ODF claims that the overall result of implementing all the Rules "should be a relatively small increase over background in landslide impacts for a short time after tree removal and a somewhat larger increase over background in landslide impacts associated with road management," (ODF issues paper, December 1995).
Given the high rate of documented landslides after logging in the Klamath, Coast, and Cascade mountains of Oregon (l00%-1800% increase in landslides occurring in clear cuts as compared to those in unmanaged forests (Benda et al. 1991)), any increase in landslide impacts on streams and fish habitats is a serious concern. Furthermore, the rules do not address how the expected increases will affect channel morphologies and fish habitat.
The ODF claims that harvested areas with leave-tree areas had higher rates of slope failures than clearcuts without leave-tree areas. However, the ODF used the results from a study that only examined harvested areas on very steep concave slopes at the tops of small drainages, called headwalls, to defend this point. This study also appeared to support the contention that root strength is an insignificant cause of slope failure. In fact, Skaugest et al. (1993), found that slope failures were about 50% for harvested headwalls with leave trees (n=26), 38% for clear-cut headwalls (n=69), and 25% for forested headwalls (n=180). Thus, any harvest of forested headwalls substantially increases the number of slope failures on these inherently unstable sites.
Nonetheless, the ODF's implied conclusion is that slope failures in certain high-risk sites will not be diminished by leaving trees during clear cutting, therefore, it is just as logical, and considerably more expedient to clear cut an entire headwall as long as ground disturbance can be avoided. This begs the question of whether there are some high-risk sites that should even be logged at all. Since the only way that lateral root reinforcement can be maintained is by retaining most trees, then the only reasonable prescription is no clear cutting or, at most, only selective tree removal from certain high-risk sites. Simply minimizing ground disturbance will not sufficiently protect slope stability in many of the highly, or even moderately unstable sites.
As discussed above, even when there is virtually no soil disturbance due to timber harvest on a high risk site, merely removing the trees would still be expected to increase the number of landslides as the root strength decreases over a period of two to 10 years (Swanston and Swanson 1976, Ziemer 1981). The ODF apparently dismisses the documented relationships among root reinforcement, soil strength, and landslide initiation, while offering no specific data to justify timber harvest on unstable slopes. They also fail to consider slope gradient, soil depth, and degree of saturation when determining slope stability.
The Rules for harvesting high risk sites lack specificity, e.g., "minimize risk of mass soil movement while maintaining forest productivity," and cannot be evaluated in a measurable way. Lacking a measurable standard, how can these BMPs be thoroughly monitored?
Lacking a specific method to determine landslide run-out, how can the "risk of material entering waters of the state" be adequately evaluated? There are published methods for determining landslide run-out paths (e.g., Benda and Cundy 1990, for debris flows--summarized in Collins et al. 1994). This is an example of a BMP that needs to be developed.
ODF is to be commended for developing a process for identifying high-risk areas, and specific sites within those areas, that is based on commonly accepted determinants of slope stability. In order to ensure that the process includes all of the necessary components and is universally followed across the state, it would be helpful to formalize that methodology with peer review.
Although high-risk areas are fairly well described in the ODF issues paper of December, 1995, in terms of general landform characteristics, specific high risk sites cannot be accurately located without field inspection. For example, such a process would need to identify small-scale features indicative of slope movements, locate incised channels, and measure specific slope landforms and gradients. The inaccuracy entailed in simply using maps to determine slope gradients is discussed by Dragovich et al. (1993) when they state "topographic maps average out the important steep slope segments that control the location of slide initiation."
It is not clear when geo-technical specialists would be expected to assist foresters in locating high-risk sites. Areas of high risk can be screened by foresters using an objective process, but specific identification of those sites may require trained staff. The current approach reflects the one used in Washington before a method was developed, under Watershed Analysis, to systematically evaluate landslide hazards from forestry activities over an entire sub-basin.
By protecting only high-risk sites (no moderate risk sites are discussed), it is not clear that all the sites with a high potential for delivering sediment to streams are actually identified. Many factors influenced by forest operations, especially when combined with decreased soil strength during storms, effectively change moderately stable sites into highly unstable sites (Chatwin et al. 1994): "Landslides are rarely triggered during the actual logging operation. Rather, they occur on sites that are naturally moderately stable, but become unstable following tree root deterioration."
Storm flows channeled and delivered by road ditches, cut-slopes, and cross-drains are also common triggers that can saturate soils and change moderately unstable ones into highly unstable sites (Chatwin et al. 1994). (See the section on Road-Related Problems for a discussion of the concerns with older roads.) Proper road design, construction, and maintenance can greatly reduce, but not eliminate, the adverse effects of roads on slope stability. Furthermore it is not clear that all the active and inactive roads regulated by ODF (estimated at about 3 mi/mi2) are adequately maintained.
Any increase in landslide rates may potentially have a very serious impact on many fish habitats that are currently in a degraded condition due to decades of unrestrained road building and timber harvest.
It would be very helpful if ODF were to develop a process for estimating the sediment delivery of specific landslides into streams. There are tested models that can be used to determine the downstream extent of a particular landslide, and thus make some estimate of the potential damage to fish habitats (Benda and Cundy 1990). Use of such a model would enable ODF to develop a way to link the upslope processes to conditions within fish streams.
We understand that a landslide inventory will be updated for at least six areas because preliminary assessments suggest that the low-frequency floods of February 1996 have resulted in widespread road washouts, channel changes, and further degradation of anadromous fish habitat. An analysis of the causes and effects of each landslide would improve future management of slopes and channels and may dictate a review of the Rules.
The ODF's ongoing watershed analysis of mass erosion in the Kilchis River is particularly important in light of the anticipated Habitat Conservation Plan (HCP) being developed with ODF, focusing on the state forest lands. However, it would also be helpful to monitor a representative basin within each geologic type of the anticipated HCP. Such a study could accomplish the important task of establishing a quantitative relationship among forestry activities, slope stability determinants, sediment delivery, and channel conditions with respect to anadromous fish habitats. The results would be of great utility to the extent that it would then be possible to extrapolate and apply the data to areas with similar geology and thereby estimate channel responses to management.
A single, channelized landslide can disturb many kilometers of stream. Degrees of disturbance range from displacement of LWD and resorting of spawning gravels to removal of many mature riparian trees and down-cutting streambeds by several meters, or even cutting an entirely new channel. Channels with extensive sediment deposits often lack surface flow during summer. Unstable road segments, whether old or new, constitute a high risk if they have the potential to initiate channelized landslides.
Geologists differentiate several types of landslides and related peak flows, e.g., deep-seated failures, shallow-rapid landslides that may either stop at a channel or become debris flows when channelized, and dam-break floods. The last mentioned are the most destructive and occur when either floating organic debris or landslide debris plugs a channel during a storm.
Dam-break floods occur when a temporary pond forms above a channel constriction and then releases with terrific force. Resulting peak flows can be many times greater than the maximum predicted for even a 100-year flood. Dam-break floods extend the adverse impacts of slope failures further downstream (and upslope) into low gradient anadromous fish habitats than do debris flows (Johnson 1991). Streams that have had dam-break floods are considered by researchers to be susceptible to recurring dam-break floods if sufficient organic material, e.g., from logging slash or landslide debris, has been re-deposited (Coho and Burges 1994).
Channel morphologies in many anadromous fish habitats are now influenced by large amounts of sediments that have been delivered by a variety of sources over the past century (Swanson et al. 1988, Sullivan et al. 1987). Researchers have found that a slug of sediment, once introduced into a low to moderate gradient stream, will move downstream at rates of about 200 to 1000 m/yr (Madej 1982, Kelsy 1980). The long residence times of coarse sediments within many fish streams suggests that the relative risks of watershed-level channel morphology alteration will continue to be high to moderate for many more decades.
There are a number of Rules intended to protect small, non-fish-bearing streams. Because these streams are so prevalent on the landscape, especially in the wet mountains of western Oregon, fully adequate protection for streambanks, shade, and large wood recruitment would greatly reduce the area dedicated to industrial forestry.
These small, mostly non-fish-bearing streams typically constitute more than 80% of the total stream network in a fifth-order size basin (Chamberlin et al. 1991).
There is a well-documented method of surveying small streams to establish fish distribution in Oregon. It is entitled Surveying Forest Streams (ODF and ODFW, 1995). According to the ODFW, doing this task on every stream statewide is an enormous task that has no identified schedule for completion. Any HCP that addresses anadromous fish habitat would need to include a schedule for doing this work throughout the plan area.
The reason for doing this is that some recent surveys of the upstream limits of anadromous fish have found coho rearing in small- to medium- sized streams at gradients near 20% (Jill Silver, Hoh Indian Tribe, personal communication, 1995). Previously, areas such as this were thought to be outside the coho's range.
A study by Montgomery (1994) in western Oregon investigated the integration of road drainage structures with the existing stream network. Increases in stream networks occurred in areas where wide spacing of cross drains (shallow ditches used to channel and redirect surface water from roads) allowed too much water to collect, causing these ditches to erode headward up to the road drainage structures. In addition to increasing the effective stream channel network, this resulted in substantial erosion and delivery of fine sediment to streams. Further, a network of roads with densities common for forest operations (e.g., three to five mi/mi2) can be expected to increase the overall stream network of small watersheds by 12% to 35%, according to several Federal watershed analyses (Upper and Middle Lewis River Watershed Analyses, Gifford Pinchot National Forest, 1995).
The Rules do not make clear what extent of disturbance is to be tolerated during cable yarding across small Type N streams. The rules simply discuss "minimizing" disturbances in the stream channel and retaining stream-side vegetation. Also, merely stating that introduction of sediments to any stream will be minimized provides no assurance that a given action will actually avoid damaging aquatic habitats.
In order to fully meet the Rules' intent to maintain the morphology and function of these small streams, it is necessary to fully protect streambanks and ensure long-term recruitment of large woody debris (LWD) from riparian areas. Since there is no requirement to retain riparian trees along small Type N streams, the Rules are not likely to provide either streambank protection (reinforcement of banks by tree root systems) or sufficient long-term recruitment of LWD to store fine sediment and prevent it from routing directly to downstream fish-bearing streams. Mature standing conifers in riparian areas also appear to moderate the effects of channelized landslides. For Type N streams, the list of riparian functions requiring protection therefore must include sediment storage, streambank stability, and reducing the effects of channelized landslides.
Megahan (1982) surveyed 1,715 in-channel obstructions in small, steep streams on the Idaho batholith, and found that wood obstructions trapped 49% of the stored sediment. Additionally, "15 times more sediment was stored behind obstructions than was delivered to the mouth of the drainage as average sediment yield."
Streambank stability is maintained by a live root mass that is half as wide as the tree-crown diameter (Burroughs and Thomas 1977). Another way to state this is that the minimum riparian buffer width necessary to maintain root strength is about 25-30% of the height of a site-potential tree (FEMAT 1993). Based on a range of site potential tree heights in western Oregon, these riparian widths range from 20-60 feet.
Regarding channelized landslides, Coho and Burgess (1994) discuss the buttressing effect of mature riparian conifers, i.e., large trees standing within the flow path can cause rapid deposition and substantially limit landslide run-out. The width of streamside buffers necessary to retain this function was not described.
The lack of a long-term ability to recruit large wood in small non-fish-bearing streams places the important sediment storage function of these headwater channels at risk. The timing, rate, and amounts of sediment delivered to fish habitats are greatly influenced by LWD in small streams providing upstream sediment storage capacity (Swanson and Fredriksen 1988, Bisson et al. 1992). If sufficient instream structures providing sediment storage are not maintained in headwater streams over the long-term, then in creased amounts of fine and coarse sediments are expected to be transported to anadromous fish streams, further damaging habitats that have already been severely degraded.
The Rules do not directly address potential changes in hydrology that may result from forestry operations. According to an issues paper prepared by ODF technical staff in December of 1995, the Rules intended to minimize the extent of surface soil disturbance will indirectly maintain surface water hydrology. Also, the Rules that require prompt reforestation and minimize disturbance from slash burns are expected to prevent hydrologic changes. Watercourses and wetlands are protected by a 20-foot no-touch zone and hydrologic connectivity is maintained between streams and wetlands.
ODF states that there is no need for explicit Rules that assess and address potential changes in hydrology as well as associated changes in fish habitat. Evidence to the contrary is dismissed as being either conflicting or insignificant.
The ODF acknowledges that older logging roads are potential sources of continued detrimental channel change. But unless a particular road is (or becomes) active under current forest operations, there is no rule requiring that erosion or drainage problems be corrected. This is a serious deficiency in the Oregon Forest Practice Rules, since landslides and adverse channel changes often are triggered by roads that were constructed some time ago under less restrictive standards. (See the discussion of this issue under the section on Road-Related Problems.)
Because there are no Rules that limit the extent or severity of harvest operations within a given watershed, changes in hydrology resulting from forestry activities are not adequately addressed. The rules simply attempt to minimize soil surface disturbances; this cannot begin to address the many ways that hydrologic changes are triggered by roads and logging.
In order to build links between forest operations and their effects on anadromous fish habitat, it is necessary to focus on aspects of channel morphology and dynamics that are sensitive indicators of perturbation and to consider the affected channel's specific type and position in the channel network. The ODF needs to establish a process for assessing channel morphologies and a way to link these studies with riparian and upslope management. A recently published approach for classifying stream reaches based on response to changes in the hydrologic and sediment regimes may be a useful model upon which to build these processes (Montgomery and Buffington 1993).
Following is a brief summary of flow changes in various hydrologic regimes.
(1) Extensive clear-cutting results in short term increases in low flows.
(2) In areas dominated by fog-drip, clear-cutting can cause decreases in low flows.
(3) There may be a decrease in low flows as clear-cut areas become revegetated.
(4) In areas susceptible to rain-on-snow storms, there is an increase in peak flows for many years (decades) after clear-cutting.
(5) In areas dominated by either snow or rain alone, peak flow increases after clear-cutting are usually slight.
(6) Roads and compacted soils act synergistically with clear-cuts to increase peak flows.
(7) Water yields increase after about 20% of the forest cover in a basin has been removed.
(8) In snow-dominated systems, vegetation removal advances the timing of spring melt and peak flows.
(9) All of the above changes are usually most evident in relatively small basins.
Rain-dominated Hydrology
Peak flows can be increased via (1) soil compaction which reduces both infiltration rates and infiltration capacity leading to increased magnitude of overland flow (Gardner and Chong 1990, Purser and Cundy 1992); (2) reduced evapo-transpiration leading to higher soil moisture and water tables (Ruprecht and Schofield 1989); (3) increased drainage density, especially through road building (Montgomery 1994); and (4) interception of subsurface flow at road cuts with subsequent conversion to surface flow (Megahan 1972).
The changes in peak flow in rain-dominated watersheds tend to be smaller and more variable than in systems with both snow and rain (MacDonald and Ritland 1989). However, it must be kept in mind that flows in rain-dominated systems are much more variable than those in snowmelt-dominated areas (Fountain and Tangborn 1985).
Rain-on-snow Hydrology
Logging operations have the same effects on soil, vegetation, and topography listed above, but they also create additional effects that can increase peak flows. First, timber harvest decreases snow interception and increases the accumulation of snowpack (Berris and Harr 1987). Second, openings increase melt rates, especially via increases in convective energy transfer to the snow-packs (Berris and Harr 1987). Several studies indicate that increases in peak flow resulting from clear cuts and roads are both large and long-lasting (Harr 1986; Coffin and Harr 1992). Notably, there is a great deal of private timberland in Oregon and Washington in the transient snow zone (roughly 1600-3000 feet elevation), where changes in hydrology from roads and harvest are most pronounced. The most damaging flood events in the Northwest typically occur during rain-on-snow events; these events are also associated with the greatest mass failure magnitudes, possibly due to high levels of soil saturation and attendant high pore pressures (e.g., Iverson and Major 1986).
Snow-melt Hydrology
Logging-related activities increase peak flows via the effects discussed above but also by greatly increasing snow accumulation (Megahan 1984, Reid 1993) and melt rates (Megahan 1984) via increased solar and convective transfers (Megahan 1984, Harr 1986, and MacDonald and Ritland 1989). Increased snowpack depth increases melt duration and, therefore, amplifies the magnitude of saturated areas by elevating water tables (Megahan 1984). This leads to increased peak flow (Rhodes 1985). When larger areas are saturated for longer periods of time, overland flows increase, as well as annual water yields (Rhodes 1985, Ruprecht and Schofield 1989), and peak flows (Moore et al. 1986). Watershed-level studies have consistently found that logging increases peak flows in a statistically detectable manner in snow-dominated regions (Harr 1986, MacDonald and Ritland 1989, King and Tennyson 1984, King 1989, Cheng 1989, Coffin and Harr 1990, Chamberlin et al. 1991) though there are some exceptions (generally due to poor study design) . Researchers have noted that increased snowmelt due to logging and roads can increase sediment transport significantly (King 1989, MacDonald and Ritland 1989). Heede (1991) found that ephemeral channels expanded significantly in the high-elevation forested region of Arizona after logging and inferred that increased peak discharge was the cause.
Annual Water Yield and Low Flows
Deforestation increases annual water yields by decreasing evapo-transpiration; this is done by shunting water from storage into surface runoff (Ruprecht and Schofield 1989), and by decreased infiltration capacity from soil compaction. Generally, increases in low flows also occur when evapo-transpiration is reduced, especially during the summer where this factor makes up a large portion of the hydrologic budget. Many studies indicate that low flows do not appear to increase until a substantial proportion of a watershed is deforested (Bosch and Hewlett 1982).
However, the increase in low flows may be short-lived when forests are regenerated because the initial stages of second growth have higher evapo-transpiration than late-seral forests. Some investigations indicate that as clearcuts become revegetated, low flows decrease to a point below pre-logging levels (Chamberlin et al. 1991). Losses in base flow caused by road cuts likely last as long as the road exists. Systems dominated by fog drip may undergo some flow reduction until the stands are reestablished (Harr 1982).
Recommendations for Potential Hydrologic Changes
- The ODF needs to recognize the preponderance of studies demonstrating that logging and roadbuilding cause changes in hydrologic function. Further, the ODF needs to develop watershed-scale methodologies for assessing the likely effects of proposed forest operations.
- The ODF should delineate the areas in Oregon where the various hydrologic regimes prevail (e.g., fog-drip, rain dominated, rain-on-snow zone, and snow melt).
- Establish a process for assessing channel morphologies, and linking them to riparian and upslope management considerations.
- Manage road design, construction, and maintenance to minimize both the interception of sub-surface water and the altered routing of surface waters. Road altered flows commonly trigger slope failures as well as cause adverse changes in anadromous fish habitat.
Issue 4.
Cumulative Effects
There is no well-defined process to address cumulative effects of forestry activities in the State of Oregon's Revised Forest Practice Rules. The ODF's position is that since each Best Management Practice (BMP) will minimize adverse "immediate" effects associated with a specific activity, the overall risk from adverse cumulative effects of all the BMPs associated with a particular project is likely acceptable.
"Immediate effects" do not include effects that occur later in time (after triggering events such as floods, and fires), and also do not include indirect and/or off-site effects of the actions, e.g., blanketing of downstream redds with sediment from activities further upstream in a watershed. The contributions to overall cumulative effects of past and reasonably foreseeable future actions are also not addressed.
The Rules (ORS Section 527.770-15-2), however, do require that "a study of harvest rates and cumulative effects related to forest practices on forest land in Oregon" shall be delivered in a report to the Sixty-eighth Legislative Assembly. In addition, the Rules (ORS Section 527.710-8) include a provision allowing the Board of Forestry (Board) to, "based upon the analysis required in section 15(2) [above]... and as the results become available, and [if] the Board determines that additional rules are necessary... the board shall adopt forest practice rules that reduce to the degree practicable the adverse impacts of cumulative effects on air and water quality, soil productivity, fish and wildlife resources and watersheds."
Until that time, the Rules simply require monitoring of selected BMPs, which is intended to point ODF toward changing those BMPs that need improvement. This approach is not adequate to assess cumulative effects on aquatic resources such as salmon. Cumulative effects must include the effects of multiple activities in time and space, and should be evaluated on a watershed-by-watershed basis. Appropriate watershed-specific practices could then be identified and applied to adequately minimize cumulative effects.
It is important to note that the provisions in the Rules to develop a cumulative effects assessment process rely considerably on the discretion of the State Board of Forestry. The Rules do not address the appropriate scale at which such an assessment should be conducted (the watershed). Furthermore, the actual development of such a process is apparently pending conclusion of several studies. Only one study report has been provided to NMFS, and it consists of an extensive literature review (Beschta et al. 1995); it appears that this is not the study specifically required in the Rules. Based on a review of that report, however, it appears that there is a process under consideration; however, currently it has not been formalized or subjected to peer review.
Although there are provisions under the rules for watershed specific practices for watersheds that either are water quality limited or have listed (Threatened or Endangered) species, there are no examples of watersheds that have had special rules established. There is apparently no other process to either analyze or explicitly address watershed scale cumulative effects in the rules.
The rules require monitoring of selected BMPs, which is intended to point ODF toward changing those BMPs that need improvement. This approach is simply inadequate to assess cumulative effects to aquatic resources such as salmon. Cumulative effects must include the effects of multiple activities in time and space, and should be evaluated on a watershed-by-watershed basis. Appropriate watershed-specific practices could then be identified and applied to adequately minimize cumulative effects.
Commissioned Report and Suggested Approach
ODF recently commissioned a lengthy report that is summarized in a 39 page executive summary (Beschta et al. 1995). In that summary, a reasonable process to analyze cumulative effects of forest practices on aquatic biota or water quality is apparently dismissed as much too complicated. The dynamic nature of forest ecosystems, along both spatial and temporal scales, combined with the stochastic nature of natural processes, uncertain knowledge of current conditions relative to historic (or reference) conditions, and complex interactions of effects and competing resource needs are all considered overwhelming, and would result in such a complicated and demanding process that ODF simply could not commit to such a vast undertaking.
Therefore, a simple approach is described in the executive summary that "is not a quantitative methodology but rather... provides a framework for identifying state, regional, and basinwide cumulative effects and a landscape level inventory." The intention is to clearly state goals, objectives, and assessment criteria, conduct the assessment, and finally specify "forest practices that are designed to fit the goals and objectives for the area." The hypothetical example presents very general prescriptions that give only vague direction to watershed managers: e.g., "Minimize roading and sedimentation that destroy fish habitat . . . Removal of dead or down material will not be permitted from existing intermittent channels . . . Due to slope gradients and erosion hazard, uphill cable yarding would be advisable."
The above approach has not been compared to other models of cumulative effects assessment and management, and appears to be generally inconsistent with both the CEQ definition of cumulative effects, and regional salmon conservation needs.
Best Management Practices
The method or approach to managing cumulative effects on Oregon's lands is to use BMPs which (by definition) are applied on a site-specific basis. Cumulative effects of forest practices may include changes in sediment, temperature, and hydrologic regimes, resulting in direct, indirect or eventual loss of key habitat components (e.g., clean gravel interstices, large woody debris (LWD), low temperature holding pools, and protected off-channel rearing areas) necessary for spawning and rearing of anadromous salmonids. These changes often are not expressed "immediately" at the project site, but instead may occur subsequent to triggering events (fire, floods, storms) or are manifested off-site (downstream) of where the effects are initiated.
Pitfalls of Depending on BMPs
The prevention of potentially adverse impacts at the project site is indeed necessary, but not sufficient, to avoid adverse cumulative effects under the CEQ definition of cumulative effects (CEQ 1971). As Reid (1993) states: "The BMP approach is based on the premise that if on-site effects of a project are held to an acceptable level, then the project is acceptable, regardless of activities going on around it. Interactions between projects are beyond the scope of BMP analysis, and operational controls are applied only to individual projects."
In summary, however useful site specific BMPs are in minimizing effects of individual actions, they still do not address the cumulative effects of multiple actions occurring in the watershed which, though individually "minimized" through application of the site-specific BMPs, may still be significant, in their totality, and have undesirable consequences for beneficial uses such as salmon populations and salmon habitat.
It should be stated that the entire approach of the Revised Oregon Forest Practices Rules to cumulative effects relies upon an untested assumption that minimizing site specific actions by application of site specific BMPs somehow avoids adverse cumulative effects to important beneficial uses of streams and watersheds, such as salmon. In many cases this is an unreasonable assumption. Because it is sometimes unreasonable, and has yet to be tested, there are significant risks in its application across large portions of the landscape.
The argument that applying a BMP while conducting a specific forest practice minimizes site specific effects and thus also minimizes cumulative effects is logically flawed. Every BMP is an action and has an effect (ODF maintains that some actions have "neutral," "subtractive," or "decrementally synergistic" effects; however, to our knowledge these are unproven in the scientific literature) thus, generally, the more the BMPs are applied, the greater the cumulative effect. Only by minimizing the number of actions, i.e., the number of individual applications of BMPs, would cumulative effects be minimized. This is precisely why a cumulative effects assessment is needed -- to establish the watershed-specific limits and excesses of BMP application.
Even if present practices (i.e., revised forest practice rules) meet some "standard" of minimizing future contributions to cumulative effects, the legacy of past practices, and of the implementation of older rules, or "no rules" (i.e., actions prior to the adoption of forest practice rules) still exists. The best example is so called "legacy roads," many of which are still present on the landscape, unmaintained, prone to sudden failure, and currently supplying chronic, non-point source sediment pollution to forested streams and downstream habitats.
Examples of Cumulative Effects
Although individual management activities by themselves may not cause significant harm to salmonid habitats, incrementally and collectively they may degrade habitat and cause long-term declines in fish abundance (Bisson et al. 1992). Changes in sediment dynamics, streamflow, and water temperature are not just local problems restricted to a particular reach of a stream, but problems that can have adverse cumulative effects throughout the entire downstream basin (Sedell and Swanson 1984; Grant 1988). For example, increased erosion in headwaters, combined with reduced sediment storage capacity in small streams, from loss of stable instream LWD, can overwhelm larger streams with sediment (Bisson et al. 1992). Likewise, increased water temperature in headwater streams may not harm salmonids there but can make water too warm downstream (Bjornn and Reiser 1991).
Cumulative effects on sediment and hydrology worsen as the area affected by timber harvest increases (Rhodes and McCullough 1994). The amount of sediment delivered to streams and fine sediment in pools increases with increasing timber harvest and road construction (Chen 1992; Lisle and Hilton 1992). Water yield increases in proportion to the areas devegetated (Haar 1983), and peak flows increase in proportion to roads and soil compaction (Haar et al. 1979). Pool depth and frequency, LWD, and channel complexity decrease with increased logging (Bisson et al. 1992; Reeves et al. 1993; Murphy 1995).
Habitat disturbances that are anthropogenic in origin combine with natural disturbance events to create cumulative effects. Habitat disturbances can be cumulative in the sense that different factors acting sequentially or concurrently can limit population size or growth during different phases of the freshwater and estuarine rearing cycles of anadromous salmonids (Elliott 1985). Habitat disturbances may also be cumulative in the sense that more than one important physical habitat factor or component (i.e., those components necessary for salmon survival and reproduction) may be altered at the same time, or over a period of time (Bisson et al. 1992; Reeves et al. 1993; Ralph et al. 1994; Murphy 1995).
Cumulative Loss of Habitat Complexity
The most pervasive cumulative effect of past forest practices on habitats for anadromous salmonids has been an overall reduction in habitat complexity (Bisson et al. 1992), from loss of multiple habitat components. Habitat complexity has declined principally because of reduced size and frequency of pools due to filling with sediment and loss of LWD (Reeves et al. 1993; Ralph et al. 1994). However, there has also been a significant loss of off-channel rearing habitats (e.g., side channels, riverine ponds, backwater sloughs) important for juvenile salmon production, particularly coho salmon (Peterson 1982). Cumulative habitat simplification has caused a widespread reduction in salmonid diversity throughout the state of Oregon, and throughout the region.
Cumulative effects of individual, dispersed timber-harvest and road-related activities commonly increase system-wide risks of habitat damage, recovery times, and susceptibility to large runoff events and related disturbances (Peterson et al. 1992). Depending on the physiographic characteristics of a watershed, such events can adversely affect not only water quality but also riparian and aquatic habitats (Wissmar et al. 1994). Common hillslope, floodplain, riparian, and stream responses include mass wasting, streambank erosion, and changes in channel geomorphology (Wissmar et al. 1994). When adverse effects occur, they are usually in violation of Section 319 of the Federal Clean Water Act (Wissmar et al. 1994).
Cumulative effects of land and water uses over the past century have greatly altered the health of river basins in eastern Washington and Oregon (Wissmar et al. 1994). Environmental effects resulting from timber harvest, fire management, livestock grazing, mining, and irrigation and other factors over long periods of time have become significant collectively. Cumulative effects induced by upland forest practices include changes in hydrology, temporary and long-term sediment production, transport and storage, and off-site or downstream effects (Swanson 1986).
Of specific concern is the pathway for creating significant cumulative effects through reduction of future supplies of LWD to fish bearing streams. Large woody debris is a critical component of salmonid habitat, and is the major pool forming and sediment storing element in many streams in Oregon. Beschta et al. (1995) cite a study done by the ODF (OFIC 1993) that indicated that logging practices at that time reduced streamside conifers by 61 percent in western Oregon. The remaining conifers represented, on average, only 50 percent of the potential large conifer debris that would have come from the undisturbed stand. In eastern Oregon, streamside conifers were reduced by 33 percent after harvest using current logging practices. Beschta et al. (1995) conclude that "[t]hough instream loadings of large woody debris present prior to harvest were not affected by current harvesting practices, future supplies of large wood are significantly reduced by current practices in western Oregon. This reduction in large wood recruitment has the potential to continue a state of diminished fish habitat quality and quantity."
Need for Watershed Scale Assessment
The amount of cumulative effects to anadromous fish populations and their supporting habitats cannot be successfully evaluated, controlled, or mitigated at a site-specific level. The effects of individual actions, such as dispersed, separate harvest units and road building, should be considered in the context of all other previous and ongoing activities in the watershed (Murphy 1995). This is not to say that specific BMPs applied at a particular time and place do not have an important role, merely that they need to be tailored to a watershed or basin scale context to be effective. As the National Academy of Sciences, the so called "Supreme Court of Science," states: "There is an increasing need to understand cumulative effects not only on a site-specific basis, but also across entire watersheds. Only through a broad geographic perspective can the unique qualities of each watershed and their spatial and temporal effects of aquatic habitats be effectively understood." (NRC 1995)
As an example, non-point source pollution by sediment, i.e., erosion from unstable hillslopes and roads and subsequent transport and deposition of sediment in streams, should be analyzed at a watershed or basin scale. Sediment produced on hillslopes, or from channel migration and other streambank erosional processes, moves though a stream network in a downstream direction. Evaluation at a watershed scale is particularly necessary because effects may be generated in one place within a watershed and felt in another. A key component of effective cumulative effects management is identification of sediment sources that are beyond what the watershed's stream network can manage, i.e., that are in excess of expected sediment input rates from natural events, or that exceed the channel's storage capacity. This excess sediment can have tremendous on-site and off-site (downstream) impacts, including damage to tributary and mainstem spawning habitats and rearing areas, blockage of culverts and road failures, filling in of reservoirs, siltation of bays and estuaries, increased dredging and spoil disposal costs, and reduced water quality for downstream agricultural, recreational, and municipal uses.
Need for Watershed-Specific Prescriptions
To continue to use sediment as an example, the cumulative effects assessment for a particular watershed should result in prescription or application of watershed specific practices that reduce or minimize sediment production in the watershed. This will avoid both on-site and off-site cumulative effects. An assortment of watershed specific practices are possible. However, they generally should include the following types of actions:
-- identifying high risk roads and road segments within a watershed (those prone to immediate failure causing massive inputs of sediment) and putting these roads to bed
-- identifying other roads and road segments in a watershed that supply chronic sediment inputs to streams, in excess of natural erosion rates, and implementing appropriate road maintenance, on a regular basis, to control those inputs
-- identifying and replacing improperly installed culverts and road crossings, whose failure would deliver large amounts of fine sediment to the stream
-- identifying and relocating valley bottom roads that deliver fine sediment at greater rates due to proximity of these roads to the stream channel
-- protecting unstable and potentially unstable hillslopes and prohibiting timber harvest or road building on the most unstable of these sites (i.e., those with slopes greater than 65% (33 degrees) and one or more of the following: soils less than 10 feet in depth and lying on uniform slopes, soils with low cohesiveness, and/or landforms with known potential for mass wasting; see section of these comments on "mass wasting").
Several Methodologies Available
Although the ODF argues that cumulative effects of forestry activities are too complex to manage effectively, there are a quite a number of methodologies already in use that are field tested, peer reviewed, and reasonably effective in capturing the major pathways and mechanisms of cumulative effects to aquatic systems.
Reid (1993, page 27-36) provides a comparison of eight different cumulative effects methodologies that could be used, at least six of which may be useful in evaluating cumulative sediment effects from logging and roading. Reid (1993) also recommends using a coarse screening procedure (similar to the. method used by the California Department of Forestry) to identify the mechanisms for cumulative effects within watersheds, followed by more specific analysis of identified mechanisms, as necessary, using appropriate modular assessment tools. Many of the needed tools already exist as "modules" of both the Federal and Washington State Watershed Analysis guides. The National Marine Fisheries Service has recently developed a "matrix" of factors and indicators (NMFS 1995) that would provide a rapid screening procedure to identify pathways for effects where more in depth analysis may be needed.
In addition to methods to evaluate the effects of logging and roading on sediment input dynamics, similar cumulative effects analysis procedures and watershed specific practices are needed to address (1) small stream protection (i.e., for temperature control/shading, LWD recruitment, energy inputs to streams, in-channel sediment storage, and other pathways for on-site and off-site cumulative effects), and (2) potential changes in hydrology (see other sections of this analysis).
Any method to assess cumulative effects should address downstream effects, for example, the violation of water quality standards downstream of where the effects may originate -- i.e., upstream forested watersheds. A cumulative effects methodology should incorporate this concern and address downstream changes not only to water quality parameters, such as temperature, pH, and turbidity, but also to habitat related parameters important to anadromous fish, such as pools, percentage of fine sediment in gravels (or cobble embeddedness), volume of fine sediment in pools (Lisle and Hilton 1992), and amount or presence of off-channel rearing areas in mainstem habitats of salmon producing rivers (Peterson 1982). Ideally, a process should also consider the contribution of upstream forest practices to social and economic costs, such as flood relocations and reparations, dredging costs to remedy sedimentation and infilling of bays and estuaries, and water treatment costs to ensure healthy recreational and consumptive uses of water.
The revised Oregon Forest Practice Rules contain no mechanism whereby cumulative effects on a watershed scale can be determined and the appropriate "watershed specific practices," which are a provision of the revised rules, can be applied. The State of Idaho has developed a cumulative effects methodology, and the State of Washington's Timber, Fish, and Wildlife program addresses cumulative effects through completion of watershed analysis. We are not suggesting that ODF adopt either of these state-specific approaches, however we do recommend that ODF develop an effective method or process for cumulative effects assessment.
Recommendations
The ODF needs to develop a reasonable process for cumulative effects assessment at the watershed scale that is both consistent with the CEQ definition of cumulative effects and regional salmon conservation objectives. The assessment process should be applied to evaluate cumulative effects both on State Forest lands and on private lands regulated by the State's Forest Practice Rules. NMFS suggests that ODF consider using or adapting the NMFS "matrix" of factors and indicators (NMFS 1995) as a rapid and readily adaptable screening process for determining watershed scale (direct, indirect and) cumulative effects. The advantage of this method over other available screening procedures is that it allows the watershed's current environmental baseline to be established, and allows adjustment of target values of key parameters (ecological pathways and indicators affecting salmon and their essential habitats) for unique watersheds or ecoregions. Appropriate watershed specific practices and prescriptions can then be developed to adequately minimize watershed scale cumulative effects. Note: additional Rules may need to be developed as recommended in other sections of this analysis, or as identified during watershed-specific analyses. Alternatively, NMFS suggests that ODF utilize currently available, peer reviewed and field tested watershed analysis and cumulative effects methodologies.
Issue 5.
Inadequate Long-Term Wood Recruitment into Streams
When ODF developed the latest Forest Practice Rules (which were enacted in September 1994), several scientists from academia and industry assisted in the process of modeling woody debris inputs and tree growth in the riparian areas. The rationale for the approach the ODF used is well documented in The Oregon Forest Practices Act Water Protection Rules, Scientific and Policy Considerations, December 1994 (Lorenson et al. 1994).
The overall strategy is good, but the widths of the riparian management areas (RMAs) are too narrow and the tree densities to be retained after logging are too low to provide optimal riparian function. Only fish-bearing streams are managed to provide large wood.
Fish-bearing streams are classed according to channel size, with 100-foot RMAs on both sides of the streams larger than about 50 feet wide, 70-foot RMAs on medium-sized streams, and 50-foot RMAs on small streams. The inner 20 feet are "no-touch" zones where all trees are retained. Trees within the outer (managed) area of each RMA are managed under the Rules to supply both commercial harvest and large-sized conifers that could someday contribute to instream structure. The ODF assumes that in managed forests, the inner 20 feet are dominated by hardwoods.
Mature stands provide most riparian functions and inputs in greater quality and quantity than do young stands. In streams bordered by young stands, shortages of large, persistent woody debris are particularly noticeable. Conifer stands with large trees are the best suppliers of this large, persistent woody debris (Murphy 1995).
Historically, the forest landscape contained riparian stands of all ages and species compositions, ranging from early successional to old-growth. Wildfire, windstorms, floods, insects, disease, and beaver activity were the agents of periodic disturbance. Nevertheless, across the forest landscape and at any given time, a large portion of all riparian areas supported stands of mature forests. In contrast, the riparian areas on private lands in Oregon now primarily support younger age classes -- very little is left of the mature age classes (Lorenson et al. 1994).
Tree densities within the managed portions of the RMAs are to be managed as fixed basal area targets (i.e. targets are measured as total cross-sectional ft2/acre). Part of the logic behind the establishment of basal area targets was that these targets would encourage retention and growth of larger-sized trees, adjacent to the inner "no-touch" zones. In reality, basal area targets can be met by lots of small diameter trees, a few very large diameter trees, or some combination of both. There is therefore no assurance that a significant number of larger diameter trees will be available to provide LWD recruitment from the outer managed RMA zone.
Tree species used to meet basal area targets are intended to include conifers primarily. Larger coniferous species do not deteriorate as fast, and provide significantly greater benefits in terms of both habitat function and channel stability and integrity, than smaller hardwood species. However, hardwood species may be included in the basal area targets under certain conditions, and in fact it is found that the inner 40 feet of RMAs is presently dominated by hardwoods in many streams in Oregon (This is documented by ODF themselves). Also the rotation of these stands is assumed to be 50 years. Typically 60 to 90 year rotations are required for large conifer growth and recruitment. There is therefore no assurance that management of the outer portion of RMAs can presently supply, or will likely supply in the future, adequate quantities of larger coniferous tree pieces, particularly mature or defective conifer boles that could readily fall into streams. Large conifer boles with intact root wads are the most stable and long-lasting form of instream LWD. These stable "key pieces" of woody debris remain in place for a significantly longer period of time, often even during extreme flood events, and are more effective at trapping other smaller woody debris pieces, and sediment. [Add cites, e.g. L Bryant]
Specifically, the goal of the Rules is to meet a desired future condition of mature streamside trees, dominated by conifers between 80 and 200 years old. We agree with this goal, but we do not agree that the prescribed basal area targets, and the assumed rotation length (50 years), will achieve that goal within the desired time period, which is stated in the Rules to be the next 25 years. This is largely because the of the currently low baseline conditions of riparian forest stands, which contain mostly hardwoods and young conifers. This current species composition will not allow 100 percent of the desired future complement of mature (80-200 year old) conifer trees. The rotation length (50 years) is also not long enough to allow riparian conifers to grow to maturity (80-200 years) throughout multiple rotations.
Amounts of Future Large Wood
The Rules generally do not provide for sufficient LWD recruitment in any but the largest fish-bearing streams. According to an analysis of the figures for riparian basal area targets for each stream size, the Rules would eventually provide a maximum of no more than 92% of the potential sources of LWD along large fish streams, 83% along medium-sized fish streams, and only 56% along small fish streams. In addition, virtually none of the necessary large wood would be retained along non-fish-bearing streams. After 25 years of growth (mid-rotation age for many clearcuts), only 73% to 83% of the potential sources of large wood would remain next to large, fish-bearing streams in five forest types in western Oregon. The medium-sized fish streams would have 66% to 75% of the potential sources of large wood, and small fish-bearing streams would only have 24% to 30%.
We have discussed this analysis with ODF policy staff, who informally concur that the Rules may meet only 30-80% of the necessary large wood, but the political climate at the time the Rules were developed in 1994 dictated this level of riparian tree retention.
Optional Placing LWD for Stream Restoration
It is not known how much LWD exists in anadromous fish habitats in Western Oregon, but it is generally considered to be deficient throughout Oregon on non-Federal lands. Restoring the proper levels of LWD in streams largely depends on natural inputs from windthrow or other mortality. Natural inputs of LWD are minimal for young riparian stands until hardwoods are 40 to 65 years old and conifers are more than 80 years old (Grette 1985, and Heimann 1988).
Under the Rules, landowners have the option of placing some large wood in fish-bearing streams (under the ODF and ODFW guidelines for proper instream placement) and counting that as part of the riparian tree retention. We like this approach to encouraging instream restoration where suitable. However, even under this option, a specified minimum basal area must be maintained.
In practice, some small landowners often retain all trees within an entire RMA because (1) tree sizes do not allow harvest until a basal area minimum is attained, and (2) even when tree sizes and numbers approach maturity and would thereby enable some harvest, it is far simpler to measure the full width of the RMA and not measure basal areas. Of course, industrial timber growers are expected to manage for maximum economic gain.
According to a study of the sizes of LWD observed to be functioning in various stream sizes, large fish-bearing streams (about 50 feet wide) need trees at least 20 inches in diameter (Bilby and Ward 1989). Tree lengths greater than the channel width tend to be stable during transport flows. Medium-sized fish-bearing streams (about 25 feet wide) need trees at least 14 inches in diameter. Small streams (about 10 feet wide) need trees at least 10 inches in diameter.
Conifers are generally preferred to hardwoods for instream function because they have greater strength and last much longer in water. Fallen alders tend to decompose entirely after five to ten years in a stream, while many conifer species remain solid for decades. A few species of cedar and redwood will last for centuries.
Recommendations for Inadequate Long-Term Wood Recruitment
- Manage for RMAs as wide as a site-potential tree height (about 120 to 170 feet) to ensure the potential supply of future large wood.
- Basal area targets need to be increased to provide 100% of the necessary wood recruitment at mid-rotation.
- A wider no-touch zone would better maintain streambank protection and shade. This should be 25-30% of the site-potential tree height.
- Trees growing within the inner no-touch zone should not be counted toward the basal area target.
- An approach to placing large wood tin streams hat takes into account the specific channel type would be much more likely to actually improve fish habitats.
- Riparian trees need to be provided along all non-fish-bearing streams. This will ensure a long-term supply of structural elements that will store sediments and maintain riparian functions in these headwater channels.
Issue 6.
Road-Related Problems
Over the last century, forest practices have left many older roads and railroad grades, i.e., "legacy roads." Only roads that have been used since 1971 (when the Forest Practice Act was first developed) are addressed by the Rules. According to the ODF, there is no process for any state agency to inspect or address the potential slope failures associated with these legacy roads. Monitoring done in 1988 found these older roads were major sources of landslides.
There is very little information available on the density or sediment delivery potential of the legacy roads. One rough estimate of their density is one mi/mi2, compared to an estimated three mi/mi2 for newer roads that are regulated by the ODF (K. Mills, ODF geologist, pers. comm. 1996).
The ODF admits that older roads, which were constructed under different standards, "have in some cases created a legacy of potential instability. Many landslides over the last few years occurred as the result of construction practices of many decades ago. Over-steepened fill and decomposing debris in fills can fail years after construction. Maintenance activities can reduce, but not eliminate, the potential for landslides on these older roads," (ODF issues paper, December 1995).
The latter statement assumes that maintenance may be conducted on some older roads, but these roads are entirely ignored unless needed for ongoing forest operations. Water that saturates unstable fills or is diverted by older roads onto sensitive slopes during storms is a leading cause of slope failures (Chatwin et al. 1994).
Regarding the risks of channel morphology alteration, the ODF stated in their December 1995 issues paper that "in western Oregon the risk may be moderate for watersheds with many old and abandoned roads, and/or old railroad grades." There is actually a high risk that older roads or railroad grades will trigger slope failures that will deliver large amounts of sediments into anadromous fish streams.
In light of the recent floods of February, 1996, the ODF will be conducting a landslide inventory of six areas in western Oregon. It is expected that affected channels will also be examined in order to determine what changes resulted from sediment delivery and debris flows. The range of effects has been well documented for anadromous fish streams, but the ODF must still establish a clear link between landslides and the changes in fish habitats on lands regulated by the Forest Practice Act and Rules.
Recommendations for Road-Related Problems
- There needs to be a process for identifying and correcting potential erosion from older roads and railroad grades.
- Newer roads that the ODF regulates need to be adequately maintained to avoid potential erosion problems and sediment delivery to anadromous fish habitats.
- It is necessary to monitor for compliance all activities conducted under the Rules and report on their effectiveness.
- Establish a clear link between landslides and changes in fish habitats.
References:
Benda, L. 1995. Stochastic geomorphology in mountain basins, Ph.D. thesis, University of Washington, Seattle, WA
Benda, L. and Cundy, T. 1990. Predicting deposition of debris flows in mountain channels, Canadian Geotechnical Journal 27:409-417.
Benda, L. Collins, B., Schiess, P., and Martin, D. 1991. Methods for testing the effectiveness of Washington Forest Practices Rules and Regulations with regard to sediment production and transport to streams: report prepared for the Washington Department of Natural Resources, Olympia, Washington, by Pentec Environmental, Inc., DNR-CMER Report TFW-WQ8-91-008, 124 p.
Berris, S.N. and Harr, R.D. 1987. Comparative snow accumulations and melt during rainfall in forested and clear-cut plots in the western Cascades of Oregon. Water Resour. Res., 23: 135-142.
Beschta, R.L., J.R. Boyle, C.C. Chambers, W.P. Gibson, S.V. Gregory, J. Grizzel, J.C. Hagar, J.L. Li, W.C. McComb, T.W. Parzybok, M.L. Reiter, G.H. Talyor, J.E. Warila, 1995. Cumulative effects of forest practices in Oregon: literature and synthesis. Prepared for Oregon Department of Forestry, 2600 State Street, Salem, Oregon, 97310.
Bilby, R.E., and J.W. Ward. 1989. Changes in characteristics and function of woody debris with increasing size of streams in western Washington. Trans. Am. Fish. Soc. 118:368-378.
Bisson, P.A., T.P. Quinn, G.H. Reeves, and S.V. Gregory. 1992. Best management practices, cumulative effects, and long-term trends in fish abundance in Pacific Northwest river Systems. Pages 189-232, in R.J. Naiman (editor) Watershed management: balancing sustainability and environmental change. Springer Verlag, New York.
Bosch, J.M., and J.D. Hewlett. 1982. A review of catchment experiments to determine the effect of vegetation changes on water yield and evapotranspiration. Journal of Hydrology 55:3-23.
Bjornn, T.C., and D.W. Reiser. 1991. Habitat requirements of salmonids in streams. Pages 83-138, in W.R. Meehan (editor) Influences of forest and rangeland management on salmonid fishes and their habitats. Special Publication 19. American Fisheries Society, Bethesda, Maryland.
Burns, D.C. 1991. Cumulative effects of small modifications to habitat. Fisheries 16:12-17.
Burroughs, E.R., and Thomas, B.R., 1977. Declining root strength in Douglas-fir after felling as a factor in slope stability. Research Paper INT-190. USDA Forest Service, Intermountain Forest and Range Experiment Station, Ogden, UT
CEQ 1971. CEQ Guidelines, 40 CFR 1508.7, issued 23 April 1971.
Chamberlin, T.W., Harr, R.D., and Everest, F.H. 1991. Timber Harvesting, Silviculture, and Watershed Processes. Influences of Forest and Rangeland Management on Salmonid Fishes and Their Habitats, Am. Fish. Soc. Special Pubi. 19: 181-205.
Chatwin, S.C., Howes, D.E., Schwab, J.W., and Swanston, D.N. 1994. A guide for management of landslide-prone terrain in the Pacific Northwest, 2nd Edition: British Columbia Ministry of Forests, Land Management Handbook Number 18, 220 p.
Chen, G.K. 1992. Use of basin survey data in habitat modeling and cumulative watershed effects analysis. FHR Currents, Fish Habitat Relationships Technical Bulletin No. 8. USDA Forest Service, Pacific Southwest region, Eureka, California.
Cheng, J.D. 1989. Streamflow changes after Clear-cut logging of a pine beetle-infested watershed in southern British Columbia, Canada. Water Resour. Res. 25: 449-456.
Coho, C. and Burges, S.J. 1994. Dam-break floods in low order mountain channels of the Pacific Northwest, Washington State Department of Natural Resources TFW report, TFW-5H9-93-00l, Olympia, WA
Collins, B., T. Beechie, L. Benda, P. Kennard, C. Veidhuisen, V. Anderson, and D. Berg. 1994. Watershed assessment and salmonid habitat restoration strategy for Deer Creek, North Cascades of Washington, report to the Stiliaguamish Indian Tribe and the Washington State Department of Ecology by the 10,000 Years Institute, Seattle, WA
Dietrich, W.E., and T. Dunne. 1978. Sediment budget for a small catchment in mountainous terrain. Z. Geomorphol. Suppl. 29:191-206.
Dragovich, J.D., Brunengo, M.J., and Gerstel, W.J. 1993. Landslide inventory and analysis of the Tilton River-Mineral Creek area, Lewis County, Washington; Part 1: terrain and geologic factors: Washington Geology, v. 21, n. 3, p. 9-18
Elliott, J.M. 1985. Population dynamics of migratory trout, Salmo trutta, in a Lake District stream, 1966-83, and their implications for fisheries management. J. of Fish Biol. 27 (Supplement A) :35-43.
Fountain, A.G. and Tangborn, W.V. 1985. The effect of glaciers on streamflow variations. Water Resour. Res., 21: 579-586.
Gardner, B.D. and Chong, S.K. 1990. Hydrologic responses of compacted forest soils. J. Hydrol., 112: 327-334.
Grant, G. 1988. The RAPID technique: a new method for evaluating downstream effects of forest practices on riparian zones. General Technical Report PNW-GTR-338. USDA Forest Service, Pacific Northwest Research Station, Portland, Oregon.
Grette, G.B. 1985. The role of large organic debris in juvenile salmonid rearing habitat in small streams. M.S. thesis, Univ. WA, Seattle, WA. 105 p.
Harr, R.D. 1980. Streamflow after patch logging in small drainages within the Bull Run Municipal Watershed, Oregon. USDA, Forest Service Research Paper PNW-268. Portland, OR.
Harr, R.D. 1982. Fog-drip in the Bull Run Municipal Watershed, Oregon. Water Resour. Bull., 18: 785-789.
Harr, R.D. 1986. Effects of clearcutting on rain-on-snow runoff in western Oregon: A new look at old studies. Water Resour. Res., 22: 1095-1100.
Harr, R.D. 1989. Cumulative effects of timber harvest on stream flows. Paper presented at Soc. A., Foresters 1989 Annual Convention. Spokane, WA. 24p.
Haar, R.D., R.L. Fredriksen, and J.S. Rothacher. 1979. Changes in stream flow following timber harvest in southwestern Oregon. Research Paper PNW-249. USDA Forest Service, Pacific Northwest Forest and Range Experiment Station, Portland, Oregon.
Harr, R. D. and Coffin, B.A. 1990. Effects of forest cover on snowmelt during rainfall. New Perspectives for Watershed Management Symposium Seattle. Wash., p. 62, Center for Streamside Studies, Univ. of Wash., Seattle, WA.
Heede, B.H. 1991. Response of a stream in disequilibrium to timber harvest. Env. Manage., 15: 251-255.
Heimann, D.C. 1988. Recruitment trends and physical characteristics of coarse woody debris in Oregon Coast Range streams. M.S. thesis, Oregon State University, Corvallis, OR
Hicks, B.J., Beschta, R.L., Harr, R.D. 1991b. Long-term changes in streamflow following logging in western Oregon and associated fisheries implications. Water Resources Bulletin 27:217-226.
Iverson, R.M. and J.J. Major. 1986. Groundwater seepage vectors and the potential for failure and debris flow mobilization. Water Resources Research 19:1543-1548.
Johnson, A.C. 1991. Effects of landslide-dam-break floods on channel morphology. Washington State Department of Natural Resources TFW report, TFW-5H17-91-00l, Olympia, WA.
Kelsey, H.M. 1980. A sediment budget and an analysis of geomorphic process in the Van Duzen River basin, north coastal California, 1941-1975. Geol. Soc. Am. Bulletin 91:1119-1216.
King, J.G. 1989. Streamflow Responses to Road Building and Harvesting: A Comparison With the Equivalent Clearcut Area Procedure. USFS Res. Paper INT-401, Ogden, UT.
King, J.G. and Tennyson, L.C. 1984. Alteration of streamflow following road construction in north central Idaho. Water Resour. Res., 20: 1159-1163.
Krogstad, F. 1995. A physiology and ecology based model of lateral root reinforcement of unstable hillslopes, MS thesis, University of Washington, Seattle, WA, 44 pages.
Lisle, T.E., and S. Hilton. 1992. The volume of fine sediment in pools: an index of sediment supply in gravel-bed streams. Water Resources Bulletin 28:371-383.
Lorenson, T., Andrus, C.W., and Runyon J. 1994. The Oregon Forest Practices Act water protection rules, scientific and policy considerations. Oregon Dept. Forestry, Salem, OR.
MacDonald, A. and Ritland, K.W. 1989. Sediment Dynamics in Type 4 and 5 Waters: A Review and Synthesis. TFW-012-89-002, Wash. Dept. of Natural Resour., Olympia, WA.
Madej, M. A. 1982. Sediment transport and channel changes in an aggrading stream in the Puget Lowland, WA: in Sediment budgets and routing in forested drainage basins, USDA Forest Service, General Technical Report PNW-141, p.97-108.
Megahan, W.F. 1972. Subsurface flow interception by a logging road in mountains of central Idaho. Proceedings: National Symposium on Watersheds in Transition, pp. 321-329, Am. Water Resour. Assoc., Bethesda, MD.
Megahan, W.F. 1982. Channel sediment storage behind obstructions in forested drainage basins draining the granitic bedrock of the Idaho batholith: in Sediment budgets and routing in forested drainage basins, USDA Forest Service, General Technical Report PNW-141, p. 114-121.
Megahan, W.F. 1983. Hydrologic effects of clearcutting and wildlife on steep granitic slopes in Idaho. Water Resources Research 19:811-819.
Megahan, W.F. 1984. Snowmelt and logging influence on piezometric levels in steep forested watersheds in Idaho. Soil Reinforcement and Moisture Effects on Slope Stability, pp. 1-8, Transportation Res. Record 965, Transportation Research Bd., Nat. Rec. Council, Washington. D.C.
Moore, I.D., Mackay, S.M., Wallbrink, P.J., Burch, G.J., and O'Loughlin, E.M. 1986. Hydrologic characteristics of a small forested catchment in southeastern New South Wales. Pre logging condition. J. Hydrol., 87: 307-335.
Montgomery, D. 1994. Road surface drainage, channel initiation, and slope instability, Water Resources Research, vol. 30, no. 6, pp. 1925-1932.
Montgomery, D.R., and Buffington, J.M. 1993. Channel classification, prediction of channel response, and assessment of channel condition. TFW-SHlO-93-002., Wash. Dept. of Natural Resources, Olympia, WA.
Murphy, M.L. 1995. Forestry impacts on freshwater habitat of anadromous salmonids in the Pacific Northwest and Alaska - -requirements for protection and restoration. U.S. Department of Commerce/NOAA Decision Analysis Series No. 7. Alaska Fisheries Science Center, Auke Bay Laboratory, Auke Bay Alaska. 156 p.
National Resource Council (NRC). 1995. Upstream -- Salmon and Society in the Pacific Northwest. Prepublication Copy, National Academy Press, 2101 Constitution Avenue N.W., Washington D.C. 388 p.
Oregon Forest Industries Council (OFIC) 1993. 1992 Oregon stream monitoring project. 37 p.
Peterson, N.P. 1992. Immigration of juvenile coho salmon (Oncorhynchus kisutch) into riverine ponds. Canadian Journal of Fisheries and Aquatic Sciences 39:1308-1310.
Peterson, N.P., A. Hendry, and T.P. Quinn. 1992. Assessment of cumulative effects on salmonid habitat: some suggested parameters and target conditions. TFW-F3-92-001. Center for Streamside Studies, University of Washington, Seattle, Washington. 75p.
Purser, M.D. and Cundy, T.W. 1992. Changes in soil physical properties due to cable yarding and their hydrologic implications. West. J. Appl. For., 7: 36-39.
Ralph, S.C., G.C. Poole, L.L. Conquest, and R.J. Naiman. 1994. Stream channel morphology and woody debris in logged and unlogged basins of western Washington. Canadian Journal of Fisheries and Aquatic Sciences 51:37-51.
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.
Reid, L.M. 1993. Research and Cumulative Watershed Effects. Gen. Tech. Rep. PSW-GTR-141. Albany, California: Pacific Southwest Research Station, Forest Service, U.S. Department of Agriculture; 118 p.
Reid, L.M., and T. Dunne. 1984. Sediment production from forest road surfaces. Water Resources Research 20:1753-1761.
Rhodes, J.J. 1985. A reconnaissance of hydrologic nitrate transport in an undisturbed watershed near Lake Tahoe. Unpublished M.S. thesis, Univ. of Nevada, Reno, NV.
Rhodes, J.J., and D.A. McCullough. 1994. A Coarse Screening Process for Potential Application in ESA Consultations. Prepared for the National Marine Fisheries Service. Columbia River Inter-Tribal Fish Commission, Portland, Oregon
Ruprecht, J.K. and N.J. Schofield. 1989. Analysis of streamflow generation following deforestation in southwest western Australia. J. Hydrol., 105: 1-17.
Sedell,. J.R., and F.J. Swanson. 1984. Ecological characteristics of streams in old-growth forests of the Pacific Northwest. Pages 9-16, in W.R. Meehan, T.R. Merrell, Jr., and T.A. Hanley (editors) Proceedings, fish and wildlife relationships in old-growth forests symposium. American Institute of Fishery Research Biologists, Asheville, North Carolina.
Sidle, R.C. 1989. Overview of cumulative effects concepts and issues. Water Resources Session, Soc. Am. Foresters National Convention, Spokane, WA.
Sidle, R. 1992. A theoretical model of the effects of timber harvesting on slope stability, Water Resources Research, vol. 28., no. 7, pp. 1897-1910.
Sidle, R., A. Pearce, and C. O'Loughlin. 1985. Hillslope stability and land use, Water Resources Monograph 11, American Geophysical Union, Washington, D.C.
Skaugest, A., Froehlich, H., and Lautz, K. 1993. Forest management on landslide prone sites: the effectiveness of headwall leave areas; Oregon State University, Coastal Oregon Productivity Enhancement (COPE) Report, v. 6, n. 1, p. 3-6.
Sullivan, K., T. E. Lisle, C. A. Dollof, G. E. Grant, and L. M. Reid. 1987. Stream channels: the link between forests and fishes: in Streamside management: forestry and fishery interactions., E. 0. Salo, and T. W. Cundy, eds., Univ. WA, Seattle, WA.
Swanson, F.J. 1986. Comments from a panel discussion at the AGU symposium on cumulative effects. Pages 66-69, in .Cumulative Effects - the UFOs of hydrology. Tech. Bull. 490. National Council of Air and Stream Improvement, Inc. New York.
Swanson, F. J., R. J. Janda, and T. Dunne. 1982. Summary: sediment budget and routing studies: in Sediment budgets and routing in forested drainage basins, USDA Forest Service, General Technical Report PNW-141, p. 157-165.
Swanson, F. J., and R. L. Fredriksen. 1982. Sediment routing and budgets: implications for judging impacts of forestry practices: in Sediment budgets and routing in forested drainage basins, USDA Forest Service, General Technical Report PNW-141, p. 129-137.
Swanston, D. N. 1969. Mass wasting in coastal Alaska. Res. Paper PNW-83. Portland, OR: Pac. NW Forest & Range Expt. Stn., Forest Service, USDA. 15 p.
Swanston, D. N., and F. J. Swanson. 1976. Timber harvesting, mass erosion, and steepland forest geomorphology in the Pacific Northwest: in Geomorphology and engineering. p. 199-221. D.R.Coates, ed., Dowden, Hutchinson, and Ross, Inc., Stroudsburg, PA.
Washington Forest Practices Board. 1995. Board Manual: standard methodology for conducting watershed analysis under chapter 222-22 WAC, Version 3.0. Washington Department of Water Resources, Olympia, Washington.
Wissmar, R.C., J.E. Smith, B.A. McIntosh, H.W. Li, G.H. Reeves, and J.R. Sedell. 1994. Ecological health of river basins in forested regions of eastern Washington and Oregon. Gen. Tech.
33 Rep. PNW-GTR-326. U.S. Department of Agriculture, Forest Service, Pacific Northwest Research Station, Portland, Oregon. 65 p.
Ziemer, R. R. 1981. Roots and the stability of forested slopes: in Erosion and sediment transport in Pacific Rim Steeplands. Christchurch, New Zealand. Intl. Assoc. Hydrol. Sci. Publ. 132:343-361.
Ziemer, R. R., and D. N. Swanson. 1977. Root strength changes after logging in southeast Alaska. USDA Forest Service Research Note PNW-306.