The grant workplan and budget anticipated that water quality monitoring would be a component of the project. However, there was no water body or stream in the project area. Photomonitoring was done, and is described below.

Photomonitoring

Each parcel had before and after photos taken; ambient condition photos before treatment and after treatment were also taken. Photos were also taken during the treatment process.

Parcel photos

Pre- and post-treatment photos were taken of each parcel. A range pole marked in one foot increments was used on each site location (one foot increments indicating a sense of scale). Each parcel was assigned a number; numbers were posted on the range pole. Surveyor flags were used to identify camera position and range pole position. The same camera and setting was used pre- and post-treatment. See the Final Grant Report (PDF*, 5.2 MB) for the photos.

Water Quality Impacts

The project field coordinator met with Mike Brenner, Natural Resource Conservation Service (USDA-NRCS) staff person for Placer County to assess impacts to water quality from this project. Mr. Brenner offered two perspectives, the official agency position on mastication, with background material, and a perspective derived from the Universal Soil Loss Equation (USLE).

The NRCS position regarding impacts from thinning using mastication is that impacts are negligible. As a basis for this position, Brenner referenced Effects of Management on Water Quality in North American Forests by Brown and Binkley (USDA, USFS, General Technical Report RM-248, pps. 13-14). The following table summarizes management effects in the first column as noted in the technical report; the second column notes the applicability to the Colfax mastication project.

Potential management effect on water quality from Brown and Binkley	Applicability to Colfax mastication project
Detachment of soil particles by impact of raindrops	Minimal impact, as canopy was reduced only to 60%, and all materials were masticated and left in place as mulch.
Mass movement on steep slopes (e.g., debris slides)	In one of the heaviest winters in memory, only one area showed any signs of slide, and that was a very large oak located on the edge of the project which was uprooted during snow load.
Stream channel bank erosion	No channel bank erosion was noted in the stream located below the project area.
Removal of vegetative cover	Thinned trees were masticated and vegetation was mulched in place, increasing vegetation cover.
Machinery compaction of soil	Masticators are designed to put less than 10 pounds per square inch of pressure on the earth from their tracks. Further, they work from a series of small locations, where they vibrate in one place as they work, thus minimally compacting a series of small footprints, rather than larger areas or swaths.
Harvesting leading to greater soil moisture due to reduced interception and transpiration	Harvest was not part of this project. Thinning generally increases rates of growth, resulting in greater transpiration.
Road building	An old logging road was cleared with a dozer as preparation for masticator access, though the shape of the land was not altered. A culvert was installed. Thus, some minor impacts occurred as a result of roadbuilding.
Decay of tree roots after harvest leading to slope failure	No major trees were harvested. Trees masticated were all under 10 inches in diameter, with the vast majority in the 2-4 inch diameter. Root rotting from over-crowded small trees is not problematic.

NRCS concludes generally that mastication does not negatively impact erosion; it was Mike Brenner's professional judgment that this was the case for the Colfax Hillcrest Project.

Universal Soil Loss Equation (USLE)

The NRCS staff also noted the USLE as an indicator of potential soil loss from the management practices. The USLE can give relative guidelines for the amount of erosion that was being avoided by reducing the risk of catastrophic fire. Essentially, the exercise was to determine the relative factor of the mastication management soil loss as opposed to the erosion that might occur due to catastrophic fire. The reference used was Biotechnical Slope Protection and Erosion Control by Gray and Leiser (Van Nostrand Reinhold Co, New York, pps. 16-22).

The soil loss equation is: X=RKSLCP where:

X = the computed soil loss in tons (dry weight) per acre from a given storm period
R = the rainfall erosion index for the given storm period
K = the soil erodibility factor
L = the slope length factor
S = the slope gradient factor
C = cropping management (vegetation) factor
P = erosion control practice factor

It was noted that for a mastication project, the only variable that is changing in the USLE is C = vegetation management. C factors for woodlands are listed below:

The existing tree canopy of nearly 100% cover was reduced by the project to 60–70% cover. However, the reduction in cover was masticated and serves as "mulch", equivalent to "forest litter" in the chart above. NRCS considers mastication a minor trade-off of canopy reduction and forest litter increase, with no net impact to erosion, and effectively no change through mastication to the C-factor. The range of C-factor from the table would be from .001–.004.

In order to estimate the widest range of impact from catastrophic fire, the C-factor from bare soil was used from the following table, companion to the table above:

The C-factor for worst case catastrophic fire from this table would be first line "no appreciable canopy" where fire had burned all vegetation down to scorched tree boles, and zero percent ground cover from column 4. Thus, the worst scenario C-factor for catastrophic fire is .45. Since the C-factor is the only variable that changes in the USLE, the worst case erosion would change .45/.004=112 to .45/.001-450. The change in erosion, according to USLE, could be a factor of 112 to 450 times due to worst case catastrophic fire, an example of which is shown in the picture of the 1960 Volcano Fire. Mike Brenner pointed out that this was absolutely maximum theoretical change based on the USLE, which is a tool developed for and used most accurately to predict soil losses on Midwestern farm conditions.

This photo (click for larger view) is an example of worst scenario soil loss conditions due to catastrophic fire and severe weather in the first year after the fire.

Erosion Potential using USFS WEPP FUME model

The Project Manager and Registered Professional Forester also met with USFS TNF soil scientist and watershed specialist Carol Kennedy for the purpose of applying the Water Erosion Prediction Project Fuel Management Tool (WEPP FuME) to estimate avoided soil loss, which is a comparison of mastication and worst scenario catastrophic fire. The following material from USFS introduces the model:

Watershed Analysis for Fuel Management Operations, by William J. Elliot
U.S.D.A. Forest Service, Rocky Mountain Research Station, Moscow, ID. February 2005.

Draft chapter for a General Technical Report on the Environmental Consequences Toolkit for Applied Wildland Fire Research in Support of Project Level Hazardous Fuels Planning.

This chapter discusses the main components for completing a watershed analysis to supportfuel management activities. The main tool is the Water Erosion Prediction Project FuelManagement Tool (WEPP FuME). Other tools will be discussed that provide more detailed analysis.

Introduction

One of the main products of many forests is surface water. The main pollutant in most forest streams is sediment. Upland management disturbances including fuel management activities and forest roads can cause erosion, leading to increased stream sedimentation and reduced water quality. Forest managers need to evaluate the impact of most forest activities on stream sedimentation, including fuel management.

A special computer interface has been developed to assist with analyzing soil erosion rates associated with fuel management activities. This interface estimates background erosion rates, and predicts erosion associated with mechanical thinning, prescribed fire, and the road network. The interface uses the Water Erosion Prediction Project (WEPP) model to predict sediment yields from hillslopes and road segments to the stream network. The WEPP model is a physically-based soil erosion model developed over the past 15 years to predict soil erosion and sediment yields for agriculture, rangeland, and forest conditions (Laflen et al., 1997). The simple interface has a large database of climates, vegetation files, and forest soil properties to support this and other interfaces, including Disturbed WEPP for forests, and WEPP: Road for road segment analyses (Elliot, 2004). The soil databases for roads and disturbed forested hillslopes are based on rainfall simulation and natural rainfall studies carried out over the past 20 years (Elliot and Hall, 1997).

For this application, the WEPP hillslope interface is used to model a single strip of hillslope (Figure 1). It is assumed that the sediment generated from this hillslope from a number of disturbances will be routed through the watershed. In the year of the disturbance, there is likely to be considerable deposition of sediment from the disturbed hillslope in the stream network. This sediment is gradually routed through the watershed in subsequent wet years. If the years are dry, there is unlikely to be any sediment routed. As the disturbed hillslope recovers, erosion from that hillslope will gradually decline. This application assumes that road erosion occurs every year, with the magnitude dependent only the level of traffic and the weather during the year.

Description of the Tool

The WEPP FuME interface carries out erosion prediction runs for seven forest conditions:

Undisturbed mature forest

Wildfire

Prescribed fire

Thinning

No traffic roads

Low traffic roads

High traffic roads

The climate, soil texture, topography, road density, wildfire return interval, prescribed fire cycle and thinning cycle are specified by the user.

With the guidance of USFS soil scientist Carol Kennedy, three areas were selected on the Hillcrest project for WEPP model runs. Those sites are indicated on the following map:

Related links:

Applied Wildland Fire Research in Support of Project Level Hazardous Fuels Planning
Environmental Consequences Toolkit (WePP FuMe can be found here)

WEPP FuME Model Results

First Site

Climate	COLFAX	CA
Soil texture	clay loam
Hillslope length	1000	ft
Hillslope gradient	0 30 40	%
Buffer length	250	ft
Wildfire cycle	20	y
Prescribed fire cycle	5	y
Thinning cycle	10	y
Road density	4	mi mi^-2

Output summary based on 50 years of possible weather

Line	Source of sediment	Sediment delivery in year of disturbance (ton mi^-2)	Return period of disturbance (y)	"Average" annual hillslope sedimentation (ton mi^-2 y^-1)
1	Undisturbed forest		1	2809.6
2	Wildfire	82880	20	4144.0
3	Prescribed fire	12160	5	2432.0
4	Thinning	3244.8	10	324.5
5	Low access roads	77.3 to 150.9	1	77.3 to 150.9
6	High access roads	132.4 to 150.9	1	132.4 to 150.9

The first site indicates an "average annual hillslope sedimentation (ton/sq. mi./yr)" as 2809.6 tons. (The Hillcrest site has a relatively high rate of average annual erosion because the top of the hill is essentially a densely roaded suburban environment; water sheets off the residential impervious surfaces and gains speed until it hits the forest at mid-slope.) In column three, the "sediment delivery in year of disturbance" is shown to be 82880 tons. The worst case catastrophic scenario from this model is the wildfire sediment delivery in year of disturbance divided by the average annual hillslope sedimentation, or 82880 tons/2809 tons = 29.5. The difference, then, of average condition to worst case castrophic fire scenario is approximately thirty fold the rate of sediment. Note that the model takes into account many variables that are not used in the USLE, resulting in a more moderate prediction of erosion. (For more complete documentation of the analysis, see the example full analysis report.)

Second Site

The site for the second model run was similar in character to the first site. The model run on site 2 had the following results:

Climate	COLFAX	CA
Soil texture	clay loam
Hillslope length	750	ft
Hillslope gradient	0 35 45	%
Buffer length	100	ft
Wildfire cycle	20	y
Prescribed fire cycle	5	y
Thinning cycle	10	y
Road density	4	mi mi^-2

Output summary based on 50 years of possible weather

Line	Source of sediment	Sediment delivery in year of disturbance (ton mi^-2)	Return period of disturbance (y)	"Average" annual hillslope sedimentation (ton mi^-2 y^-1)
1	Undisturbed forest		1	2515.2
2	Wildfire	76582.4	20	3829.1
3	Prescribed fire	13331.2	5	2666.2
4	Thinning	3296	10	329.6
5	Low access roads	51.7 to 171.0	1	51.7 to 171.0
6	High access roads	148.3 to 171.0	1	148.3 to 171.0

Using the same approach as above, sediment delivery in year of disturbance exceeds average annual hillslope sedimentation by a factor of 30.5, approximately the same as at site one.

Third Site

The third model run site is quite distinct from the other two. The slope is more gentle, it does not have a residential housing/impervious surface component and is entirely forest condition. The model results for site three are as follows:

Climate	COLFAX	CA
Soil texture	clay loam
Hillslope length	250	ft
Hillslope gradient	0 30 0	%
Buffer length	100	ft
Wildfire cycle	20	y
Prescribed fire cycle	5	y
Thinning cycle	10	y
Road density	1	mi mi^-2

Output summary based on 50 years of possible weather

Line	Source of sediment	Sediment delivery in year of disturbance (ton mi^-2)	Return period of disturbance (y)	"Average" annual hillslope sedimentation (ton mi^-2 y^-1)
1	Undisturbed forest		1	748.8
2	Wildfire	22156.8	20	1107.8
3	Prescribed fire	2668.8	5	533.8
4	Thinning	908.8	10	90.9
5	Low access roads	0.0 to 1.5	1	0.0 to 1.5
6	High access roads	0.0 to 1.5	1	0.0 to 1.5

In the third example, even though the site conditions were different with regard to slope, road density, and residential component, the sediment delivery in year of disturbance exceeds average annual hillslope sedimentation by a factor of 29.6, approximately the same as at site one. wildfire sediment delivery in year of disturbance divided by the average annual hillslope sedimentation, or 22156 tons/748.8 tons = 29.6.

Conclusion

The conclusion from the theoretical application of USLE and the three model runs is that fuel reduction which prevents catastrophic fire can avoid the significant erosion caused by catastrophic fire.

* Get Adobe® Reader