On August 26, 2996, the Eighth Street wildfire ignited in the Boise Foothills. By the time it was contained, it had burned over 15,000 acres. The overall fire intensities were 17% high, 64% moderate, and 19% low. The fire reduced effective ground cover from a prefire 30-50% e.g.c. to less than 10% over the high intensity burned areas. With these intensities, all vegetation (predominantly Douglas fir) was killed. At moderate intensities some upland shrubs had remnant leaves and small twigs, while riparian plants remained green or with dead leaves.
The watershed is dominated by fluvial slopes that are moderately dissected with a dendritic pattern. The soils are of two parent materials: Idaho Batholith and a lacustrine deposit of Glens Ferry member of Payette Lake Formation. The soils of granitic origin have dry soil moisture, and have textures ranging from loams to coarse gravelly loams, depending on slope position. Steepness also varies with slope position, ranging from 15-35% in the lower watershed to 40-75% in the upper watershed. Most of the area has moderately deep to very deep soils. After the fire hydrophobic soils developed on 32% of the forest land. Using the USLE (Universal Soil Loss Equation), double-checked by a WEPP run (Water Erosion Prediction Project), erosion was predicted to be around 32 tons per acre per year during the first two seasons after the fire.
The average annual precipitation ranges from 14-30 inches per year; increasing with elevation. Although most of this falls from November through May, rain-on-snow has historically caused only minor flooding, while summer/early fall thunderstorms have a history of flooding.
Like the monitoring team, the BAER team was interagency. In order to satisfy everyone's management goals, a detailed analysis of alternatives with mitigations and effects on various resources was done. A key consideration was protection of life and property in the city of Boise. Once the rehab plan was implemented, a monitoring plan was developed. The 1998 report assessed results after two growing seasons, and outlined monitoring plans for the third season based on what was learned to date. The following is a synopsis of the second season's results.
Three workgroups were formed to implement the plan: hydrology, soil stability, and vegetation. The plan called for assessment of first, fire effects on infiltration capacity, runoff and interrill erosion on the watershed, secondly, effectiveness of erosion control treatments, and third, the success of seeding prescriptions and distribution techniques on the vegetation community.
The hydrology workgroup assessed fire effects rather than treatment effectiveness, so none of the sample sites had burned area emergency (BAER) treatments.The group's objective was to quantify differences in infiltration capacity, runoff, and interrill erosion between burned and unburned areas in order to validate the need for BAER treatment to offset runoff effects predicted during summer thunderstorms. A rainfall simulator was used to apply 'raindrops' at a known rate of intensity (0.4 inches/hour).
The fire had its greatest impact on intensely burned south-facing slopes where infiltration was reduced from 2.1 to 1.3 inches per hour. North-facing slopes also showed a significant fire effect, though it was only the reduction in infiltration rate from that of south slopes. One important trend for burned and unburned sites alike was that runoff consistently began between 2 and 4 minutes after rainfall started on these sandy, easily erodible soils. The effect of the fire on soil erosion was even greater than its effect on runoff; nearly 40 times greater on south-facing burned slopes than on same-aspect unburned slopes. One year following the fire, what was thought to be a 100 year thunderstorm event delivered 0.4 inches in the first ten minutes. Upon closer investigation, the team learned that such events, involving intense localized storm cells, are not uncommon, occurring once every five years or so somewhere in the Boise Foothills.
Treatments included various upslope and channel treatments. Each treatment type had a slightly unique rehab objective and also separate questions, or monitoring objectives. In summary, these were:
Checkdams were designed to maintain channel stability - both strawbale checkdams and gravel bag checkdams were constructed in ephemeral draws, except in the oversteepened headwater reaches. Each design had five sites where photo points and cross-sections were established. Cross-sections both above and below the dams were monitored to see if the dams maintained structural integrity and whether they caused channel erosion. Both dam styles worked as designed, although the gravel bags were mostly disintegrated by the second season. The integrity of the channel at all sites was maintained without downcutting or bank erosion, therefore the dams were judged to be valuable rehabilitation tools.
Various hillslope treatments were done with the objective of reducing overland flow and soil erosion. These included tillage, straw wattles and hand trenches. The tillage, in particular, was intended to increase infiltration by breaking up the hydrophobic layer. The wattles and contour felling were all intended to increase infiltration by retarding surface runoff. This in turn would reduce erosion. The treatments were monitored to measure soil movement, using 3-F erosion bridges and photo points, at five treatment sites and one control site. Rilling as a result of the August '97 rainstorm caused much of the significant soil loss, and the '98 readings indicated acceleration of rill erosion, with the control and the wattle treatment having the least change in '98. The tilled site showed the greatest and most consistent soil loss, leading to the conclusion that tillage benefits are minimal for BAER. Hand trenches were only effective the first year. Straw wattles were effective.
Mechanical trenches were excavated to stop overland flow in a 100 year event, and in the chance of soil movement, to trap sediment. Every 50 feet, a 'baffle' is built into the trench, to minimize drainage area in the event of failure. Trenches were strictly an upslope treatment; that is, the trenches did not extend into swales, as they were not designed for fluvial processes. The monitoring objectives were detection of failure in trapping runoff, to see if the trenches stabilized soil stability and if they maintained adequate storage capacity the first two years. Thirty percent of the trenches were visually inspected for failures. Thirty permanent cross-sections were established in the trenched treatment area to monitor ability to store sediment and loss of trench capacity. Win X SPRO software was used to plot changes and calculate volume stored. The first year decrease in storage volume was 1.3 cubic feet per cross-section, and about half again as much was lost the second year. This amounted to a 14% reduction of total trench storage capacity by year two.
The mechanical trenches have prevented watershed damage from runoff, even given the intense storm after the first growing season. They are 'self-healing' since they fill in and revegetate over time. The team felt that they were therefore a valid emergency treatment where downstream values are high,
Sediment basins were installed to collect and store increased water. The six sediment basins were photo-documented to see if any design changes were needed for future applications. The basins were deemed a valid BAER treatment although they were found to be in need of more erosion control fabric to stabilize the spillways.
Various seeding and planting prescriptions were done to increase soil cover, encourage recovery of the native plant community, and control the spread of invasive weeds. The monitoring objectives were to identify when ground cover reaches 90% of prefire conditions, and to determine the success of various prescriptions on establishment of seeded species, native plant recovery, and controlling the spread of invasive plants.
In a nutshell, the documented results were an increase in basal ground cover due to an increase in the litter from annual grasses. By 1998 all but 1 of 15 sites had exceeded the cover goal. The seeded grasses were most successful on the drilled sites. Native plants at higher elevations showed little variance in stem density between ripped and unripped treatment areas, but the lower elevation ripped sites had higher densities. Cheatgrass increased in many sites at nearly exponential rates, however the amount was the same or only slightly exceeded the densities observed on control sites. The ridge areas were responsive to the drilling treatments in controlling the spread of invasive weeds and reducing fire potential. It appears that natural communities on the north slopes have been able to successfully outcompete weed species.
The noxious weed study objectives included evaluating the effectiveness of herbicide treatments. This involved six plots of 0.32 acres in size - 3 drilled and 3 undrilled sites. One herbicide treatment was determined to be more effective than the others. Low level photography was planned to be used to expand the study in 1999.
For details of the monitoring methods or results by the vegetative or other workgroups, contact the authors. Each year a report is produced which includes updated plans for the next monitoring season as well as results to date.
While the Eighth Street monitoring team has gained much knowledge on what works for BAER in the Boise Foothills, caution should be applied when transferring techniques to other places. The natural processes acting on a particular landscape, as well as the values at risk, need to be recognized. Negative side effects from a given treatment, such as the potential for trenches to adversely affect slope hydrology in landslide-prone ground, should be carefully considered.-
