Reports
Within the Hayfork Ranger District, watershed assessments were completed on nine (9) sub-watersheds (Plate 1). As of March 1993, final reports were accepted by the TRFWRP, along with requests for funding on only four basins (Rattlesnake Creek, Middle South Fork, Upper South Fork and Lower Hayfork Creek). An additional two sub-basins (Butter Creek and Tule Creek) have had draft reports prepared that are currently being reviewed by the USFWS, Weaverville. Finally, the District has begun preparing reports for the three remaining basins (Salt Creek, Plummer Creek and Corral Creek).
Within the Yolla Bolla District of Shasta-Trinity National Forest, reports followed a different format. Instead of summarizing inventory results basin by basin, as on the Hayfork District, areas inventoried and program accomplishments are summarized for the whole District in an annual progress report. The report is divided into several sections which describe the District topography, climate, vegetation, geology and soils, the collective accomplishments during the year, a list of field data collected during the year, and finally a list of prioritized recommendations by sub-watershed or road where needed erosion control activities should be performed (USFS, 1990h; 1991d; 1992c).
Six Rivers National Forest also provides a yearly progress report to the TRFWRP that summarizes work accomplishments during the fiscal year. The reports include data from stream crossing and landslide inventories, as well as restoration activities such as erosion control along roads, revegetation projects, landslide stabilization projects and fish population and habitat surveys both within the South Fork watershed and on their lands elsewhere in the Trinity River basin (SRNF annual reports to TRFWRP, 1990, 1991, 1992).
Scope of inventories
Excluding watershed assessment work conducted in Grouse Creek where sediment budget studies were performed by SRNF, all other basins throughout the South Fork were inventoried only for future sediment sources. On STNF managed lands, the data sets indicate approximately 90+% of all sites or roads recommended for treatment consisted of either stream crossing excavation, maintenance or upgrading, gully erosion control activities, seeding and mulching, rock armoring road surfaces and ditches, or attempts to address stream channel bed and bank erosion. Less than 10% of all mapped and identified potential sediment sources were mass movement features. Erosion associated with mass movement features in the South Fork watershed are generally considered to be the dominant process of sediment production (CDWR, 1979, 1982b; Irizarry, et. al., 1985; Raines and Kelsey, 1991). The very low number of mass movement features identified during STNF inventories suggests only the most obvious features were mapped.
Proposed rehabilitation actions
Excluding fire rehabilitation measures, proposed erosion control and rehabilitation actions to benefit fisheries resources on the STNF can be divided into four types: 1) obliteration of roads, mainly dead-end spurs; 2) closure of roads; 3) individual upgrading or reconstruction projects located at specific road-related erosion problem sites; and 4) in-stream projects.
1. Road obliteration consists of removing stream crossings by excavating the fill and the culvert, installing rolling dips and frequent cross-road drains or large waterbars to disperse road runoff, ripping the road prism to decompact the surface, improve infiltration and improve revegetation potential; and/or outslope (re-contour) sections of road which pose risks of landsliding.
2. Road closure primarily involves placing large barriers or gates at the start of the road (frequently spur roads) to prevent all vehicular use, and performing any necessary maintenance or upgrading activities to winterize the existing stream crossings and the road, including cut and fill slopes. Most frequent tasks include the installation of additional or larger culverts, installing overside or berm drains, downpipes and waterbars, protecting culvert inlets with drop inlets or trash racks, installing rolling dips at stream crossings to prevent stream diversions, providing energy dissipation at culvert outlets, and in some situations, reshaping the road prism to control surface drainage patterns. The miles of road proposed for road closure will effectively receive little to no maintenance for an unspecified time period.
3. Individual re-construction projects include work along roads which are to be retained as part of the main transportation network. Tasks most frequently recommended consist of work to repair or upgrade culverts, stabilize existing gullies, address road related surface drainage problems or, in a few instances, involve excavating unstable fill from the edges of landings and along roads.
4. Stream channel rehabilitation projects are intended to stabilize stored sediment already in the stream channels, and to reduce rates and frequency of erosion along streambanks. The most frequent measures include planting of trees and other vegetation, construction of in-stream structures such as rock and log check dams and headcut stabilization structures, installing channel armor or clearing particular sections of stream, and installing stream bank revetment.
Erosion prevention estimates
Due to the limited scope of our project, we were only able to conduct minimal field verification of potential sediment source areas on inventoried lands. Thus, this analysis is based on our extensive experience in conducting wildland erosion control assessments in nearby basins, reviews of documents and reports, and discussions with numerous agency staff and the public.
1. The most compelling question related to the South Fork watershed inventories is whether the prioritization of project work locations within all federally inventoried lands are addressing the most immediate erosion control and prevention needs. We believe the prioritization of needed corrective work at culverted stream crossings is valid and will result in measurable net benefits to fisheries resources in the South Fork. However, the lack of complete inventories of potential mass movement features for tributaries to the South Fork Trinity River (excluding the Hayfork sub-basin) is a short-coming of the inventories. Specifically, basins underlain by South Fork Mountain Schist, Galice Formation and the western half of the Rattlesnake Creek Terrane bedrocks need detailed mass-movement assessments.
2. It appears many estimates of sediment prevention and total unit treatment costs for erosion prevention projects recommended for implementation are unrealistic. Shasta-Trinity National Forest reports discuss the volume of erosion which is likely to occur at a work site or along a section of road if it were left untreated. This is an important amount to quantify when performing a watershed assessment. But, in order to effectively prioritize and treat the most urgent and critical future sediment source sites first; the volume of sediment likely to be delivered to stream channels must also be determined. All the reports discuss erosion prevention and the volume of erosion prevented or stabilized, not the volume saved or prevented from entering stream channels. As a result, the volume estimates do not appear to reflect accurate net benefits to fisheries brought about by the dollars spent to perform the work.
Most watershed assessment reports project between 10,000 to over 170,000 yd3 of erosion prevention if the listed measures are implemented. Within these totals, several sites have estimates of erosion prevention (yd3 of sediment saved) which are unrealistic, and appear inflated. For example, in the Middle South Fork basin, two proposed stream crossing reconstruction projects are estimated to prevent over 18,000 yd3 of erosion at a cost of nearly $59,000. Erosion of this magnitude is almost impossible to imagine occurring at or associated with the failure or diversion of two stream crossings.
Another example is the estimated volume of erosion prevented at several proposed road obliteration and road closure locations. As explained to us, estimates of erosion from roads are based on summing projected soil loss from stream crossings, and from roads and landings over some time frame, say 20 years. To determine road and landing erosion, identified mass movement volumes were added to the estimated volume of surface erosion using a modified Universal Soil Loss Equation. Surface area of the road to be rehabilitated is determined, and an estimate is made as to how much erosion (lowering) of the road surface will occur if the road is not obliterated or closed. Using USFS reports, back calculations suggest some road surfaces would have to be lower by over 12 inches if no work is performed to achieve the stated erosion estimates. This is not likely even with intense levels of use.
More importantly, whatever volume of road surface erosion is expected to occur, only a percentage of the eroded material would be expected to actually be delivered to a stream channel. The type of road construction (ditched or outsloped), whether the road is surfaced or unsurfaced, the number and frequency of stream crossings or ditch relief culverts, the usage level of the road (heavy, light or abandoned) and other variables will dictate the erosion rate and influence what percent of the eroded material is actually contributed to nearby streams. These road characteristics must be assessed for each road segment in order to determine the extent of erosion, and what percent of the eroded material will be stored on the hillslopes or delivered to streams.
Reid (1981) in calculating sediment yield from various classes of logging roads, found that abandoned and light-use roads produce 10 to 70 times less sediment than moderate-use roads. Heavy-use roads produced 12 times more sediment than moderate-use roads. Studies elsewhere in mountainous terrain indicate delivery ratios of sediment eroded from roads, cutbanks, fillslopes and bare soil areas to be between 30% and 70% (Megahan, 1984; Reid, 1981; USSCS, 1986; Burroughs and King, 1989; Raines and Kelsey, 1991; Weaver, et. al., in press). The lower values should be applied to roads which are outsloped with no inboard ditch, or where ditch relief culverts discharge storm flow into protected areas well away from stream channels.
Applying a uniform rate of erosion to roads without considering the variables affecting sediment movement and delivery to streams will result in erroneous estimates of sediment yield. If most roads slated for closure on STNF are light-use or already abandoned, then this data suggest estimated sediment prevention values are substantially high.
3. Road closure as proposed by STNF is primarily intended to maintain road and resource investments since continued road maintenance funds are not as readily available as in years past (John Veevaert, personal communication). A secondary goal is to prevent road and resource degradation by intentionally excluding uncontrolled use. Based on our map analysis, many segments of road which are proposed for closure or obliteration are located on hillslopes which pose low risk of sediment delivery to stream channels. Thus, sediment prevention volumes (which should be defined as sediment saved or prevented from being delivered to streams) appear inflated, and the treatments appear to benefit silvicultural (ie. reforestation) needs far more than fisheries needs.
"Closing a road" without excavating all the stream crossings and excavating potentially unstable fill materials does little to "erosion-proof" a road. Stream crossing culverts should not be left in place (even with rolling dips added to prevent stream diversions) if the crossings are not going to receive regular maintenance. Without regular maintenance, most culverted stream crossings will eventually fail and deliver sediment directly to stream channels and impact downstream fish habitat. Road closure, as defined by the USFS, may be a misnomer that could lead to reduced resource protection and increased sediment yields. The excavation of stream crossings and unstable fill material must be a component of any road closure project in order to effectively reduce the potential for sediment yield to stream channels. As this report goes to press, the USFS is currently re-evaluating it's road closure procedures and standards (Gary Brimhall, personal communication).
The unaccounted for volume of accelerated, man-caused erosion and sediment delivery associated with USFS proposed road closure methods, coupled with the very low amount of calculated surface erosion from roads which would actually be delivered to stream channels, indicates STNF sediment prevention estimates appear unrealistic. As a result, estimates of the net benefits to fisheries resources and cost-effectiveness (dollars spent per cubic yard saved) to perform the various projects are likely to be significantly over-valued.
Treatment cost estimates
1. Dollar estimates to perform erosion control activities on STNF managed lands are based on using existing USFS personnel and heavy equipment (through the use of USFS force accounts). If work was contracted to local private construction firms or contractors, it is estimated implementation costs could triple the estimates included in the assessment reports (John Veevaert, personal communication). Based on the USFS reports, estimated costs for road obliteration average $7900 per mile of road treated, which seems reasonable for most situations. However, the estimated road closure costs, which average $2350 per mile of road appear unrealistically low. Likewise, estimates to perform erosion control work at specific sites along roads very greatly and are difficult to assess. Listed individual site treatment costs range from a low of $30 (which seems reasonable) to a high of $34,500 per individual site where reconstruction or repair is proposed.
2. Cost-effectiveness values (dollars spent per cubic yard of sediment prevented from entering stream channels) to perform the various work tasks vary widely. At several sites the values seem very low, while at others the cost-effectiveness values are extremely high. Estimates for road obliteration costs on STNF lands ranged from $1.21 to $1.49/yd3, for road closure they ranged from $1.17 to $1.61/yd3, and for individual work sites ranged from $0.14 to $12.43/yd3 of erosion prevention. Virtually all the road obliteration and road closure estimates for major earth moving activities are well below actual implementation costs which are documented from other erosion control programs such as those conducted on steep, wildland forested basins at Redwood National Park (Terry Spreiter, personal communication; Weaver and Sonnevil, 1984), the Mount Baker-Snoqualmie National Forest in Washington (Harr and Nichols, 1993; Roger Nichols, personal communication), the Hoopa Indian Reservation (Pacific Watershed Associates, 1990a, 1990b), on Pacific Lumber Company lands (Pacific Watershed Associates, 1992a, 1992b), and Six Rivers National Forest (Caroline Cook, personal communication).
The low unit-cost for erosion prevention work is either related to field staff unintentionally inflating sediment production volumes, to the use of erosion volumes instead of sediment delivery volumes, to the possibility that all necessary work sites along a road segment may not have been identified, therefore estimated costs don't reflect the true extent of work required to "erosion-proof" the road, and/or to the relatively low hourly wages and rental rates paid by the USFS when in-house personnel and equipment are used for erosion control activities.
3. Finally, cost-effectiveness does not appear to have been a selection criteria in determining which road-related and in-stream rehabilitation project sites have been prioritized by the USFS for treatment, and which are excluded from treatment. For example, in the Rattlesnake Creek Watershed Inventory Report, several proposed stream channel work sites have estimated cost-effectiveness values of between $22 and $37/yd3 to implement the proposed treatment. In the Upper South Fork area, 26 proposed in-stream structures (out of a total of 56 structures to be built) will cost $40.62/yd3 of treated soil or sediment. In the Tule Creek basin, many proposed work sites involving in-stream or slope stability structures are estimated to cost from $15 to $300/yd3 of erosion prevented. Similar examples exist for proposed in-stream and streambank erosion control projects on the Yolla Bolla District.
Fortunately, for any projects to be implemented using TRRP funds, the proposed projects must meet specific criteria spelled out by the TRRP (Bill Brock and Tom Stokely, personal communication). This provides a check on the proposed projects, but does not identify whether or not "cost-effectiveness" calculations or estimates are realistic.
We believe once watershed inventories have been completed, all sites should be evaluated as to whether or not the potential projects can be implemented in a cost-effective manner. Available funds to conduct watershed and fisheries improvements are not unlimited. Projects such as those described above should probably not be implemented because of their excessively high unit costs. Many other much more cost-effective sites are present throughout the South Fork where we speculate considerably larger benefits can be obtained. As a general rule, an upper limit of $7/yd3 spent (in some special circumstances a maximum of $10/yd3) should be the cutoff in deciding whether a particular site is or is not treated. Exceptions should be closely scrutinized.
This discussion points to the need for unbiased evaluations in order to conduct and insure watershed restoration programs are addressing treatable sediment sources in a cost-effective manner. All potential sources of man-caused accelerated sediment production and yield (ie. both fluvial and mass movement sources) must first be identified and evaluated as to risk of damaging fisheries resources and water quality. Once inventories are completed, prioritization of the most critical sediment sources can proceed while adhering to rigorous cost-effectiveness standards. Chapter 11 continued