January 5, 2005

Leslie Markham

California Department of Forestry and Fire Protection

135 Ridgeway Ave.

Santa Rosa, CA 95401

Re: THP 1-04-030 SON Hansen-Whistler

and associated conversion TCP 04-530

Dear Ms. Markham:

I would like to submit these revised comments on the Hansen/Whistler THP and Timberland Conversion THP 1-04-030 SON. The Friends of the Gualala River asked me to comment on this project this a few days prior to December 20, 2004 deadline. The deadline was subsequently changed to January 7, 2005. Therefore, in this letter, I have clarified some issues and reorganized my comments for improved readability. I have also added additional references.

The Friends of the Gualala River have asked me to comment on the proposed Negative Declaration for the Hansen-Whistler Timberland Conversion and associated. I was the Hydrologist for the Mendocino County Water Agency (MCWA) from May 1989 to November 1994. The Mendocino County Board of Supervisors appointed me as their representative on pre-harvest inspections for THPs with the potential to impact public water systems. As the MCWA Hydrologist, I also reviewed Proof-of-Water pump tests for the Town of Mendocino. I also routinely reviewed CEQA documents for projects before the Planning Commission. I have a Masters in Physical Science specializing in Hydrology from Chico State University. Since 1994 I have been a consulting Hydrologist. I have also taught Hydrology at California State University, Monterey Bay.

Timberland Conversion Project

The proposed Hansen-Whistler Timberland conversion is located in Sonoma County, approximately 1.7 miles northeast of Annapolis. The proposal is to replace 16 acres of coniferous forest with 14.8 acres of vineyards and 1.2 acres building site and reservoir. The project is on the northwest end of Brushy Ridge. About 14 acres of the proposed conversion drains towards Little Creek, a tributary of Buckeye Creek. The remaining 2 acres of the conversion drain directly towards Buckeye Creek. Buckeye Creek is a Class I stream and is a tributary of the South Fork of the Gualala River. Lower Little Creek supports steelhead and is therefore a Class I stream. Steelhead were listed as threatened, in the Northern California ESU, on June 7, 2000. The Gualala River is listed as impaired by sediment and temperature. As a result, the THP associated with the Hansen-Whistler Timberland Conversion must follow the Forest Practice Rule 916.9 Protection and Restoration in Watersheds with Threatened or Impaired Values.

Based upon my review of the following documents;

· the Mitigated Negative Declaration;

· the THP including the Erosion Control Plan

· Revisions by the RPF dated August 24, 2004;

· the Hydrology Report by O’Conner dated July 14, 2004

and other documents in the file, I find that the proposed Hansen/Whistler THP may cause significant adverse impacts and may result in violations to subsections of Forest Practice Rule 916.9. I also find that the Mitigated Negative Declaration for the Hansen-Whistler Timberland conversion is inappropriate because it is based on an incomplete analysis of the storm-water runoff process which ignores the importance of subsurface storm flows in routing peak flows to the channel network. The potential for erosion from the headcutting of Class III channels due to increased subsurface storm flows is also ignored in the THP, Hydrology Report and the Mitigated Negative Declaration. In addition, the cumulative effects analysis is very weak and does not support the conclusion of no cumulative impacts. Therefore, the CDF should deny the Hansen\Whistler THP 1-04-030 SON.

Hydrologic Affect of Timberland Conversion Clearcut

The Hansen/Whistler THP and associated Timberland conversion proposes clearcutting approximately 14 acres in the Little Creek watershed and about 2 acres in the Buckeye Creek watershed. A clearcut done in association with a Timberland Conversion is hydrologically different from a clearcut done in accordance with the standard restocking rules of the Forest Practice Act. The permanent loss of the forest canopy in a Timberland Conversion is central to the hydrologic difference between a Timberland Conversion clearcut and a restocked clearcut. An assessment of the hydrologic impact of a Timberland Conversion clearcut must consider the permanent loss of the forest canopy.

The hydrologic impacts of a clearcut are related to the following factors;

· Reduction in evapotranspiration,

· Loss of canopy interception,

· Increases in compaction.

The relative importance of these factors may be different when considering the project’s affect on stream flow during each of three different periods of the year. The three periods of hydrologic interest are early season storms, later season storms and during the dry season. The distinction between early season storms and late season storms can be made on the basis of the moisture content of the soil column. Early season storms can be considered those storms that occur when the soil column is relatively dry so that a portion of the rainfall goes to satisfying the soil moisture deficit. Later season storms are ones that occur after the initial soil moisture deficit has been replenished by the earlier storms. The dry season can be considered to from extend from late about April or early May until about mid-November, on average.

In a Timberland Conversion, the forest is not replanted. So the surface-area of the forest canopy is never matched by the subsequent leaf area of subsequent uses of the land, which in this case is a vineyard. During most of the rainy season, grape vines are leafless and so the vines do not provide much surface area to intercept rainfall. Thus the amount of rain intercepted and subsequently evaporated back to the atmosphere is permanently reduced when a forest in converted to a vineyard. The much smaller surface area of the leaves of a fully developed vineyard can store only a fraction of the rainfall as a +50-year second growth forest. The rainfall stored in the vegetative cover tends to evaporate, especially on a flat ridgetop that is completely exposed to the wind and sun, such as this project. So, converting a forest to a vineyard will allow a higher percentage of the rainfall to reach the ground. Dunne and Leopold (1978, p87-88) summarize research showing that the canopy of a coniferous forest intercepts about 22% of rainfall. Most of this increase in rainfall reaching the ground must eventually reach the stream channel network.

A vineyard uses significantly less water than a forest and so soil moisture should be higher at the end of a growing season after the conversion compared to the forest prior to clearcutting. In addition, vineyards are often irrigated during the summer, which replaces soil moisture with water from a reservoir or with ground water if the irrigation water is supplied by a well.

After a few years, a restocked clearcut’s water use begins to approximate that of the pre-existing forest. Thus, soil moisture conditions at the end of the dry season begin to approximate the pre-cut soil moisture conditions. Keppeler (1998) noted that increased summer flows on the South Fork of Caspar Creek lasted only 7 years. The soil moisture regime at the end of the dry season in a Timberland Conversion clearcut will not revert to the pre-cut condition.

Normal soil processes will tend to decrease the compaction on skid trails over time in a restocked clear cut. But, in a Timberland Conversion clearcut used to grow grapes, a permanent system of roads will be left between planting blocks. In addition, equipment is often driven down the lanes between rows of vines in a vineyard at various times during the year. I have not found studies that address the compaction from equipment use in vineyards. But, the gentle slopes and the use of cover crops will probably tend to promote infiltration between rows of vines. However, the permanent roads between vineyard blocks will increase the area subject to overland flow. It is imperative that the vineyard roads be properly designed to quickly drain and prevent the road runoff from concentrating.

The following table summarizes the factors that effect streamflow following a clearcut during three periods of the year.

Factor Increasing Streamflow	Dry Season Flow	Early Season Storms	Late Season Storms
Reduced Evapotranspiration	Yes	Yes
Reduced Canopy Interception		Yes	Yes

Increased Peak Flows to Little Creek

O’Conner’s July 2004 Hydrologic Assessment analysis of peak flows is incomplete. Page 4 of the Hydrologic Assessment states:

The conversion of forest vegetation to vineyard will reduce the interception and evaporation of rainfall by forest canopy. Experimental data indicate that forest canopy intercepts and evaporates approximately 20% of storm precipitation in temperate coniferous forests (Dunne and Leopold 1978, pp. 87-88). Removal of the forest canopy therefore is expected to increase the quantity of precipitation reaching the ground surface, potentially causing increases in

Infiltration of water to the soil and percolation to the groundwater aquifers

Summer streamflow

Storm runoff

The O’Conner report cites studies that suggest that logging results in an increase in peak flows but concludes that the gentle slope, distance to the nearest Class III watercourse and surface conditions found on the Hansen/Whistler conversion site make it unlikely that there will be an increase in peak flows from the conversion. Page 8 of the O’Conner report states that:

Peak flows in stream channels draining this project area might be expected to increase, however, several factors suggest that this is unlikely to occur at this project site. Gentle topography (mean slope within the conversion area is <3%), and the distance between the project area property boundary and the nearest stream channel (Class III), suggest that it is unlikely that potential peak flow increases associated with storm runoff will be effectively routed to the stream channel. Surface storage and roughness with respect to runoff characteristics between the project site and the Class III channel is relatively high, with irregular topography, substantial surface storage capacity and vegetative cover. There is no evidence of continuous channelized flow draining to the south across or within sight of the project boundary where most of the potential peak flow increase would be routed. Considered collectively, these facts indicate that stream peak flow increases observed at Caspar Creek are unlikely to occur as a result of this project. Consequently, there is little potential for increased channel and bank erosion. Nevertheless, potential peak flow increases are calculated for the project for comparison with other projects in the area and to establish the maximum potential flow increase.

O’Conner’s analysis, quoted above, focuses exclusively on infiltration-excess (Hortonian) overland flow and totally ignores subsurface storm flow routing to the channel network. Subsurface storm flow is probably the dominate pathway in routing runoff from the hillslope to the stream channel network in the forested lands (Church and Eaton 2001) of the Gualala River watershed. Saturated overland flow is probably a locally significant storm water pathway, for example, the bottom of hillslopes may experience saturated overland flow as subsurface flow from upslope raises the water table to the surface. Areas that tend to remain wet through the summer, such as the wet areas reported along the project property boundary, are another location where saturation overland flow will tend to be an important runoff pathway. In contrast, Hortonian overland flow is probably the least significant pathway (in terms of area involved) for storm water to move from the forested hillslopes to the channel network in the Gualala River watershed during all but the most extreme storm events. Of course, Hortonian overland flow is the dominate pathway for impermeable surfaces such as roads or compacted skid trails and bedrock. The O’Conner report focuses on the least likely pathway for storm water to be routed from the project site to the nearby Class III streams during a 2-year storm.

Since the project area is on relatively flat ground, it is reasonable to ask if subsurface storm flow would occur on the site. The proposed Hansen/Whistler timberland conversion project is entirely on Goldridge soils. The description of the soil sample used to type the Goldridge soil series as attached. The basic description of the soil is quoted below.

The Goldridge series consists of deep and very deep, moderately well drained soils formed in material weathered from weakly consolidated sandstone. Goldridge soils are on rolling uplands with slopes of 2 to 50 percent. The mean annual precipitation is 45 inches and the mean annual temperature is 56 degrees F.

The Goldridge soil on the project site lies on slopes from 2% to 9% and average slope is said to be about 3%. The slope tends to increase with distance from the crest of the ridge (watershed divide). The Goldridge soil on the project site was derived from the underlying Ohlson Ranch Formation. The O’Conner Hydrology report (2004. p 9) notes that;

The Goldridge soil typically has a subsurface soil stratum with higher clay content that can impede infiltration

This observation is consistent with the description of the soil sample used to type the Goldridge soil series. The type sample clay content is shown below.

Horizon	Depth Range	Clay Content
A	0"-24"	10-20%
B	24"-73"	25-35%
C	73"-80"	15-25%

The higher clay content in the B horizon indicates that the B horizon has a lower hydraulic conductivity than the overlying A Horizon. The O’Conner Hydrology Report shows that the B horizon (subsurface) has a lower infiltration rate (0.2 to 0.63 in/hr) than the surface A horizon (0.63-2.0 in/hour), according to the Sonoma County Soil Survey. The lower infiltration capacity of the B horizon compared to the A horizon is the typical situation in forest soils (Whipkey and Kirkby, 1978, p127).

The reduced infiltration capacity of the subsurface B horizon, due to the higher clay content, provides the very conditions needed to promote subsurface storm flow along the interface of the A and B horizons. There will also be another zone of subsurface flow along the interface between the C horizon and the underlying sandstone of the Ohlson Ranch Formation. In addition, groundwater that percolates through the Ohlson Ranch Formation will encounter the nearly impermeable Franciscan Formation and the groundwater will move downslope along the interface between the Ohlson Ranch Formation and the Franciscan Formation. So, subsurface storm flow will be routed along both the A-B horizon interface and the interface between the C horizon and the Ohlson Ranch Formation. In addition, ground water will move downslope along the interface between the Ohlson Ranch formation and the Franciscan Formation.

The removal of the trees will result in higher antecedent soil moisture levels, increase subsurface storm flow and a potential increase in percolation to deep groundwater. The increased water moving along the subsurface storm flow pathways will eventually surface into the drainage network and potentially erode the head of the channel. Jaeger (2004) offers the following summary of the dynamics of channel head formation.

The channel head represents the start of the drainage network, and its location is influenced by the underlying bedrock, soil characteristics, climate regime, and land use (Montgomery and Dietrich 1988, 1989; Prosser 1996; Wemple et al. 1996). Past workers have proposed that the processes driving channel-initiation and channel head locations can be mathematically described through exceedence of an erosion threshold (Dietrich et al. 1992, 1993; Montgomery and Dietrich 1994). Such an erosion threshold is specific to the particular mechanism controlling channel-initiation (e.g. overland flow, shallow landsliding, and seepage erosion) and is expressed in terms of the contributing drainage area (Acr) and local ground surface slope (q). For example, the required contributing area required for channel-initiation by overland flow is given by

Acr = C/(tanq)^a (1)

where C is constant as a function of rainfall intensity and site-specific physical field characteristics (Montgomery and Foufoula-Georgiou 1993). These threshold models predict systematic source area-slope relationships as presented in Figure 1.

Figure 1. Source area-Slope Relationship. A Schematic taken from Montgomery and Dietrich (1994) where landscape is divided according to dominant channel initiation processes.

Figure 1 (of this letter, presented after the references) shows the proposed Hansen/Whistler timberland conversion superimposed on the 1971 Preliminary Geologic Map of Western Sonoma County and Northwesternmost Marin County, California. The Annapolis portion of the map was drawn on the 7.5’ Annapolis quadrangle from 1943 with 25’ contour lines. The superimposed project location is approximate because differences in the base topographic maps make it difficult to overlay the geologic map on the topographic map showing the project location. The geologic map shows that, in the project vicinity, Brushy Ridge is underlain by the Ohlson Ranch Formation which in turn sits on top of the Franciscan Formation. The Franciscan Formation is much less permeable than the Ohlson Ranch Formation and ground water typically only flows along cracks in the essentially impermeable rock. Therefore, it is likely that a saturated zone develops above the Ohlson Ranch/Franciscan boundary. A small portion of the water in the perched saturated zone may leak into fractures in the underlying Franciscan but most of it will tend to move downslope above the contact and increase the depth of saturation near the head of the Class III channels. The increased depth of saturated soil will increase the storm discharge into the channel network. In addition to the increased flow along the Ohlson Ranch/Franciscan boundary, there will be increased flow along the A-B horizon interface and increased flow along the interface between the C horizon and the underlying Ohlson Ranch Formation. The increased flow along each of these three subsurface pathways will increase the depth of saturation at the contact between the Ohlson Ranch Formation and the Franciscan.

In Appendix 2 of the North Coast Watershed Assessment Project (NCWAP) Report for the Gualala River Fuller and Custis (2002, p.27) discuss the characteristics of the Ohlson Ranch Formation.

The youngest consolidated geologic formation in this subdivision is the Ohlson Ranch Formation. The relatively young marine sediments of this formation are poorly consolidated sands, silts and gravels that tend to slump or flow when saturated on slopes such as those near the contact with the underlying Franciscan formation.

So, the increased depth of saturation of the Ohlson Ranch Formation at the contact with the Franciscan Formation, after the timber harvest, is expected to cause slumping or earth-flows which will result in the headward erosion of the Class III watercourses downslope of the project.

Figure 2 shows the approximate location of the Ohlson Ranch/Franciscan boundary on a 3-D view of the 7.5 minute topographic map. Figure 2 also shows the likely general direction of the subsurface storm flow and some example hollows (swales) where the increased subsurface storm flow may have the potential to initiate headcutting of the channel-head. Headcutting of the channel head would cause sediment to enter the channel network and result in a violation of Forest Practice Rules 916.9(a) (1)(2)(7) which are quoted below.

The soils around the Hansen/Whistler Timberland conversion are shown in Figure 3 along with the location of the boundary between the Ohlson Ranch Formation and the underlying Franciscan Formation. A profile line is shown in Figure 3, which may represent a potential flow path from the watershed divide to the Class III watercourse downslope of the wet area on the property line. The elevations along the profile line are shown on Figure 4. Along the profile line, the boundary between the Goldridge and Josephine soils appear to be located at the boundary between the Ohlson Ranch Formation and the underlying Franciscan Formation. Both of these interfaces appear to be located at the point where Brushy Loop Road crosses the swale. This overlap of the geologic and soil interfaces may be artificial. The soil maps for the Sonoma County Soil Survey are known to not align correctly with 7.5-minute topographic maps. And as noted previously, the 1971 Preliminary Geologic Map also does not line up perfectly with the 7.5-minute topographic map. So, it is possible that one or both of the geologic and/or soil interfaces occurs at the sharp break in slope at the 720’ contour line shown in Figure 4. This potential error in the location of the geologic contact and the boundary of the soils demonstrates the need to gather field data in the swales upslope of the geologic contact to more accurately predict the potential for headcutting of the Class III watercourses.

Dr. Robert Curry (personal communication) reports that he has observed the following affect of a timberland conversion for a vineyard in Napa County:

After the conversion, old barely-detectable "fossil" swales saturated and developed incipient and then real runoff with overland flow in what would have been seen as a "zero order" healed in-filled swale. We didn't know it was there until after it eroded.

Dr. Curry’s observations verify that headward erosion of the channel network does occur after forest clearing for vineyards.

In addition, the increased subsurface storm flow may increase the saturated area around the head of the channel leading to increased saturated overland flow which would rapidly enter the channel network. The increase in saturation near the channel head may also trigger debris-flows in the steeper channel heads such as those that drain towards Buckeye Creek. Debris-flows are triggered when the drainage out of a mass of colluvium can not keep pace with the subsurface inflow to the colluvial mass. Typically, the drainage imbalance is triggered by intense rainfall but the higher soil moisture induced by the increased subsurface stormflow to the channel head may decrease the rainfall intensity needed to trigger a debris-flow.

The increased subsurface storm flow would enter the channel network potentially increasing the storm peaks over pre-harvest levels which may in turn induce erosion of the channel bed and banks downstream of the channel head. This affect is predicted by the O’Conner report (p5), if the increase in runoff can reach the channel. O’Conner assumes that the runoff can not reach the channel because his analysis only considers Hortonian overland flow and ignores the more likely subsurface storm flow pathway.

The O’Conner report finds that the largest increase in storm runoff will occur in vineyard block 3 (along the western boundary, the wet area shown in Figure 2 is in block 3). O’Conner predicts a 33% increase or 0.86 cfs in peak storm flow from block 3. Block 5 is adjacent to and south of block 3 is predicted to have a 21% increase in peak flow or an additional 0.51 cfs. In total, the south draining (drains towards Little Creek) conversion areas are predicted to have a total increase peak flow increase of 14% or 2.06 cfs. A significant portion of that flow may be concentrated along the profile line shown in Figures 3 and 4.

The Caspar Creek study, cited by O’Conner, presents evidence from a watershed near Fort Bragg, CA that demonstrates that peak flows (storm runoff) are increased by clearcutting. Clearcutting is a required step in the Timberland Conversion process. Therefore, it is likely that the Class III watercourses draining towards Little Creek will experience a significant increase in peak flows and erosion. The Caspar Creek studies found a mean peak flow increase of 27% for storms with a 2-year recurrence interval on clearcut watersheds ranging from 25 to 67 acres (O’Conner 2004).

The increases in peak flows predicted by the O’Conner Hydrology Report are in line with the Caspar Creek studies. The flaw in the O’Conner report is that it assumes that the predicted increase in peak flows can not reach the channel network by overland flow but ignores the subsurface storm flow pathways. The O’Conner report (p 11) estimates that an additional 9 acre-feet per year might be delivered to the soil each year. The decrease in soil permeability at the A/B horizon interface and at the interface between the C horizon and the underlying Ohlson Ranch formation indicate that subsurface storm flows are very likely to occur. The a portion of the subsurface storm flow can be expected to move through soil pipes and macro-pores and so can be expected to reach the channel network quickly (Church and Eaton, 2001). Some of the subsurface storm flow is in the form of saturated flow along the interface with layers of lower permeability. Subsurface storm flow can also occur in the unsaturated zone (Montgomery and Dietrich, 2002).

Increased Sediment Discharge to Little Creek

The O’Conner report (p5) summarizes some of the findings of the Caspar Creek studies regarding increases in sediment load after logging as follows:

In summary, watershed experiments at Caspar Creek indicate substantial increases in annual water yield, summer minimum flows, and storm runoff following clearcut harvest in the North Fork Caspar Creek. In addition, suspended sediment yield for small watersheds (about 25 to 70 acres) increased substantially. Increased annual water yield is due largely to increased storm runoff which results from decreased canopy interception of rainfall and increased soil moisture; increased summer flows are significant, but represent a smaller portion of the increased annual yield. Increased summer minimum flows result primarily from reduced growing-season evapotranspiration and higher soil moisture. The increasing trend in these parameters and approximate magnitude of change is likely to be similar conversion of forest to vineyard at the project site near Annapolis. (emphasis added)

In addition, O’Conner reports that:

Lewis (1998) found that suspended sediment yield measured from the small watersheds increased on the order of 200% (a three fold increase after harvest). Although the source of this increase in suspended sediment was not determined, it was suggested that a substantial portion was caused by accelerated channel or bank erosion associated with observed increases in stream flow.

Therefore, the expected increase in the magnitude of peak flows in the Class III watercourses near the Hansen-Whistler Timberland Conversion are also expected to have a significant and measurable increase in suspended sediment load.

The expected increase in peak flow magnitude in the watercourses near the Conversion results in apparent violations of the three Forest Practice Rules listed below (emphasis added):

916.9(a)(1) Comply with the terms of a Total Maximum Daily Load (TMDL) that has been adopted to address factors that may be affected by timber operations if a TMDL has been adopted, or not result in any measurable sediment load increase to a watercourse system or lake.

916.9(a)(2) Not result in any measurable decrease in the stability of a watercourse channel or of a watercourse or lake bank.

916.9(a)(7) Result in no substantial increases in peak flows or large flood frequency.

The expected increases in peak flows will violate 916.9(a)(7). The expected increase in peak flows from the timberland conversion will measurably decrease the stability of the bed and banks of upper Little Creek by initiating headcutting of the channel head, in violation of Rule 916.9(a)(2). The anticipated erosion of the bed and banks of the Class III channels and possibly the Class I and Class II watercourses downstream in Little Creek will result in a measurable increase in sediment load in violation of Rule 916.9(a)(1).

The expected erosion of the channel-head and the bed and banks of the Class III watercourses downslope of the Hansen-Whistler Timberland Conversion, as they adjust to the expected increased magnitude of peaks flow delivered by subsurface storm flow, may result in a direct potential adverse impact to the steelhead habitat in Little Creek downstream of the Conversion. The steelhead habitat in Little Creek may also be incrementally impacted by the sediment generated from the Hansen-Whistler Timberland Conversion and other harvests or conversions in the Little Creek watershed.

Since steelhead, a federally listed species, are known to inhabit the lower reaches of Little Creek, a complete analysis of the environmental impacts, including cumulative impacts, of the Hansen/Whistler Timberland Conversion require analyzing the impact of increased peak flows and the resulting increase in sediment and other harvests or conversions in the Little Creek watershed. The Forest Practice Rules quoted above require that the Timberland Conversion not generate and measurable sediment or measurable increases in peak flows and must not destabilize and channel bed or bank.

The potential adverse environmental impacts from increased subsurface storm flow and the resulting increases in peak flow and erosion of the Class III channel heads on the steelhead in lower Little Creek invalidates the conclusion that the timberland conversion permit for the Hansen-Whistler conversion project can be issued under the proposed Mitigated Negative Declaration. CDF should either deny the application for the Hansen-Whistler Timberland Conversion or should require an EIR.

Incomplete Project Description

The THP (page 280) claims that there are no Class I, Class II or Class III watercourses within the plan area. This is not the case. The THP states (page 280) that:

No Class I, Class II or Class III watercourses occur within the plan area. An off-site Class III watercourse is afforded a minimum 25 foot riparian buffer no harvest zone to protect the filter strip vegetation, vegetative cover and habitat.

This statement shows that the Class III watercourse was given a standard WLPZ and treated as part of the plan. The THP goes on to say that:

A wet area occurs on the property has been avoided and excluded from the project. It shall be provided an ELZ with a no harvest prescription to protect the filter strip vegetation, vegetative cover and habitat.

Again, this statement shows that the wet area is a part of the plan and that the plan provides specific protections for the wet area and the surrounding vegetation. The wet area and the Class III watercourse are part of the plan but have been excluded from the cut and provided a certain level of protection.

Page 213 of the THP indicates that;

There is a wet area along the property line to the east, for which a 150+ foot setback will be maintained. This area has been set aside as part of a 200 foot wildlife corridor. A small seasonal wet area exists on the property line to the south and this area has been set aside along with a few surrounding redwood clusters.

The seasonal wet area described as being on the south property line is probably the one that has been provided a ELZ and was discussed above. However, while neither of these wet areas will be subjected to harvest activities under the proposed timberland conversion THP, they have not been adequately described. These two wet areas may play a crucial role in the subsurface hydrology of the project site. Since they have not been adequately described, from a hydrologic perspective, it is not possible to discern their importance in understanding the subsurface flow system

The conversion plan does not clearly layout crucial features of the vineyard conversion such as;

· how many vines per acre will be planted,

· how often the vines will be watered initially or after they have become established,

· how much water will be applied to each vine per irrigation cycle,

· how many irrigation cycles are expected per year

· how much irrigation water will be required during a drought.

Until this information is disclosed, it is not possible to accurately assess the impact of the proposed project on the local water resources. It is also important to answer these questions since the economic viability of the proposed vineyard conversion depends on having sufficient water to establish and maintain the vines.

The watershed area feeding the proposed reservoir also needs to be clearly defined. Calculations need to be shown concerning the amount of rain required to fill the reservoir. Calculations must also be shown regarding how much water the reservoir can expect to receive during a 20-year 1-hour storm and how the reservoir spillway will handle the excess flow. Plans must also be prepared for how to handle the reservoir discharge during a 20-year 1-hour rainstorm to prevent the formation of a new channel.

Inadequacy of the Existing Well

Page 37 of the THP is a report by Luciani Pump Company, Inc evaluating the existing well on the Hansen Whistler property, dated 9/14/01. The pump evaluation shows that the well produced 15.8 GPM and caused 30 feet of drawdown after 4 hours and 10 minutes of pumping. The comments on the pump report are quoted below.

Comments: The well is plumbed in copper and PVC. It looks like the homeowner did the work. Electrical is o.k. But has Romex run in PVC conduit.

An estimate for a larger pump is difficult to give at this time. We need to know the total vines, irrigation block sizes, etc. before a pump can be sized. My best guess at this point figuring a 30 GPM pump, pipe, pressure tanks, plumbing, and electrical would be approximately $3,500-$4,000. You should be very careful to not oversize the pump at this point, because we do not know if the groundwater will change over the years, or what affect this well will have on surrounding wells, if any.

This test is for informational purposes only and the conclusions are of the well in its present condition. This test may not show seasonal fluctuations and cannot predict either the future quantity or quality of water that the well will produce.

This report shows that the pump was tested at 15.8 gpm. Robert Plum of Luciani Pump used an estimate of 30 GPM for the purposes of estimating the cost of upgrading the well pump. The 30 GPM estimate is not based on the actual performance of a pump in the well. Mr. Plum states that placing a larger pump in the well could result in the lowering of the groundwater surface or affect neighboring wells.

The THP and the Mitigated Negative Declaration both assume, without proof, that the well can produce 35 GPM or more. If the well can not consistently supply the assumed amount it is possible that the vineyard will fail. CDF must require a 72-hour pump test performed by a qualified hydrogeologist, before approving the Mitigated Negative Declaration. The 72-hour pump test should be conducted in late August and be preceded by a 72-hour monitoring period to determine if the water table surface is static or declining. Until a 72-hour pump test is performed, the well should not be regarded as adequate. In fact, the project proponent appears to have doubts about the ability of the well to supply irrigation water for the project since some of the land will be used to create a reservoir.

The THP (p 207) and the Mitigated Negative Declaration (p 3-23) both assume that pumping the existing well for irrigation water will only impact neighboring wells only if the neighboring wells are located close to the pumping well. It is true that the cone-of-depression from a pumping well in an unconfined aquifer has a small radius of influence. However, a well that is pumping from an aquifer of limited extend, such as the Ohlson Ranch Formation under the northwest end of Brushy Ridge, can lower the groundwater table and thereby adversely impact neighboring wells that are much further away that the pumping radius-of-influence would suggest. A properly conducted 72-hour pumping test should be able to determine if extended pumping to irrigate the proposed vineyard might lower the ground water table.

The THP, the Erosion Control Plan, the Hydrology Assessment and the Mitigated Negative Declaration fail to present any detailed information to support estimates of the amount of water required to irrigate the vineyard. Such calculations are essential to assess the environmental impact of the project.

The Mitigated Negative Declaration (p3-23) states, without any supporting calculations that the project will require about 4 acre-feet annually to establish the vines and 2 acre-feet per year after the vines are established. Using the planting density of 2,450 vines per acre given on page 10 of the Timberland Conversion Permit and Application, and a watering rate of 5-gallons per vine per watering cycle, along with the assumption that a total of 12 watering cycles would be required, I estimate that the irrigation requirements for the proposed 14.8 acres of vineyard is 6.68 acre-feet. This is 167% of the unsupported estimate given in the plan. This is a crucial calculation and the details need to be revealed before the environmental impact can be determined. Also, no information is given about the amount of water that would be required during a drought such as the 6-year dry period from 1987-1992.

During the drought of 1987-1992, the Smith family rainfall records (Table 1), for a site near Annapolis, showed annual total ranging from 34.6” to 47”, which is well below their 53.6” average for the 18 years of record. The Independent Coast Observer rainfall records for Annapolis (Table 2) show that the annual totals between 1987 and 1992 ranged from 33.7” to 51.7”. The average for 28 years of record recorded by the ICO was 59.2”. Both sets of data show that about 34” of rainfall was the minimum that occurred in the recent drought of 1987-1992. Figures 5 and 6 show the statistics and result of extending the Smith family’s rainfall record using the Fort Ross precipitation record. The exceedence probability of 34” of rainfall occurring, based on the extended data set (Table 3)1, is about 95%, that is only 5% of years would have a lower rainfall. Therefore, the water resources facilities of the proposed vineyard should be designed to meet the irrigation needs of a year with only 34 inches of rainfall.

Much more information is required to determine if the existing well can supply the proposed vineyard, especially during the establishment of the vines. FPA Rule 1105.2, quoted below, requires that adequate quality and quantity of water be available to ensure the economic viability of the project.

1105.2 Director's Determination

The Director shall determine the applicant's bona fide intention to convert in light of the present and predicted economic ability of the applicant to carry out the proposed conversion; the environmental feasibility of the conversion, including, but not limited to, suitability of soils, slope, aspect, quality and quantity of water, and micro-climate; adequacy and feasibility of possible measures for mitigation of signification adverse environmental impacts; and other foreseeable factors necessary for successful conversion to the proposed land use.

Dry Season Flow

The Caspar Creek watershed studies suggest that the removal of trees increases summer streamflow because less moisture is removed from soil moisture storage by the replacement vegetation than was used by the trees. In general this is true but in itself is insufficient to estimate the expected impacts on the Hansen-Whistler Timberland conversion.

O’Conner Environmental (July 14, 2004) prepared an, Assessment of Potential Hydrologic Effects, Hansen/Whistler Timber Harvest Plan and Conversion. The O’Conner report argues that the North Fork of Caspar Creek was clearcut and therefore would allow reasonable extrapolation of the results of the Caspar Creek study to the Hansen/Whistler conversion. Portions of the North Fork of Caspar Creek were clearcut and so would, at first glance, appear to be directly applicable to a timberland conversion to vineyard. However, the North Fork of Caspar Creek has significant areas of north exposure (higher soil moisture content than flat or south facing slopes) and there are significant areas with slopes greater than those found on the Hansen-Whistler conversion. The effect of slope, aspect and relative soil water content were not examined in the Caspar Creek studies. However, the Caspar Creek studies, particularly those of the South Fork selective logging offer some insights concerning the variables that were not directly investigated in the Caspar Creek studies.

The Keppeler and Ziemer (1990) discuss the factors associated with variations in the streamflow response. They note on page 1674 that:

High antecedent moisture conditions preceding and during the hydrologic year were related to an increase in the South Fork flow relative to the North Fork.

This reflects the idea that when the soil is at saturation, the actual evapotranspiration is close to the potential evapotranspiration (PET). As soil moisture declines the trees have to reduce their actual evapotranspiration to levels well below that of the PET. So, the Caspar Creek study supports the idea that the magnitude of any increase in summer streamflow resulting from logging depends on the antecedent soil moisture conditions. That is, there will be less or no increase in summer streamflows following dry winters. This is further supported by the following quote from the 1998 article by E.T. Keppeler discussing the Caspar Creek study. The abbreviation SFC refers to the South Fork of Caspar Creek

On SFC, the minimum discharge (instantaneous daily flow) increased an average of 38 percent or 0.25 L/s/sq-km between 1972 and 1978. The maximum increase was 0.42 L/s/sq-km in 1973, the final year of timber harvesting on this watershed. No increases were detected in 1977, the driest year of record. Summer discharge minimum returned to prelogging levels beginning in 1979. (Emphasis added).

The effect of removing trees on summer streamflow is also not uniform in space. The Caspar Creek study can be seen to support the idea that removing trees from drier sites will have less impact on summer streamflow than removing trees from wetter sites especially those with nearly saturated soil.

Chang (2003, p. 195-199) provides a discussion of the factors that affect the water yield after timber harvest. The factors discussed by Chang are:

· Forest-cutting intensity

· Species

· Precipitation

· Soil topographic conditions

Chang’s discussion of soil topographic conditions on water yield is quoted below.

Hydrological responses to forest harvests vary among watersheds due to the type and depth of soils along with the steepness and orientation of the watershed. Soils of deep and fine textures have a much greater water-holding capacity than soils of shallow and coarse textures, consequently a greater potential for yield increase. Rowe and Reimann (1961) stated that water yield could not be appreciably increased in soils with depth less than about 1 m.

Slope aspect affects solar radiation, precipitation and wind speed, consequently soil and air temperatures, snow accumulation, snowmelt, ET, and vegetation type and growth. Forest transpiration is generally greater in northern than southern slopes because of denser vegetative cover and deeper soils (Bethlahmy, 1973). Actual observations showed that clearcutting on south-facing slopes caused only about a one-third increase of that measured on north-facing slopes in Idaho (Cline et al., 1977) and at Coweeta, North Carolina (Douglas, 1983). (Emphasis added)

Cermak and Kucera (1987) note that;

Transpiration rate decreases under conditions of drought. The decrease is proportional to soil water potential, but modified within certain limits by the physiological properties of plants in the given stands.

The Hansen-Whistler Timberland conversion is on fairly flat ground or on gentle south facing slopes. The proposed conversion is on Goldridge soil is a well-drained fine sandy-loam. Therefore, both the aspect and the soil type suggest that the conversion will remove trees from relatively dry sites. In contrast, a significant portion of the North Fork of Caspar Creek faces north and so would be expected to have higher soil moisture content. In addition, the clearcuts in the North Fork of Caspar Creek probably removed trees that were closer to the stream channel network and so grew on sites with elevated soil moisture than the trees that will be removed as part of the Hansen-Whistler conversion. So, the actual change in summer streamflow may be significantly less downstream of the Hansen-Whistler conversion than the changes seen in the Caspar Creek study.

Studies presented in Dunne and Leopold (pages 253-274) show that areas with saturated or nearly saturated soil moisture conditions tend to be concentrated near the bottom of hillslopes above stream channels. The saturated area tends to contract as the time after precipitation increases but the saturated area continues to be centered on the stream channel network. These studies support the idea that the location of tree removal, with respect to topographic position, plays a significant role in determining the response of the summer low flow to logging.

Furthermore, properly assessing the impacts of the hydrologic changes associated with the timberland conversion should not be based on “average conditions” but on extremes. During dry years there may be no increase in summer streamflow associated with the removal of trees during the timberland conversion.

Inadequate Mitigations

The Mitigated Negative Declaration does not have any proposed mitigations to deal with the increased storm flow runoff and erosion of the channel-heads in nearby Class III watercourses. The engineering plans for the reservoir and its outlet works are not given in the plan. Therefore, it does not appear that adequate provisions have been made to route storm water from the reservoir spillway to an appropriate channel. Page 3-24 of the Mitigated Negative Declaration says that, in case of dam failure, the escaping water would flow over 150 feet of soils prone to erosion. However, no indication of how normal storm water would be route from the reservoir spillway to the channel network is given.

Conclusions

The initial study relies on reports that use superficial and incomplete analysis to determine that there would be no significant adverse impacts to the environment resulting from the Hansen-Whistler Timberland conversion and associated THP. A careful analysis shows that a variety of significant hydrologic impacts may arise from the Hansen-Whistler Timberland conversion. These impacts include but are not limited to:

· Insufficient irrigation water for establishment of the vineyard particularly during dry years may lead to development of additional water sources with unknown environmental impacts or even abandonment of the vineyard.

· Pumping the groundwater within the Ohlson Formation may lower the ground water table and impact neighbors

· Subsurface storm flow is expected to significantly increase the peak flows in the class III watercourses that drain to Little Creek.

· Subsurface storm flow and increased percolation to the groundwater table are expected to cause headcutting of the Class II channels and the increased peak flows are expected to erode the bed and banks of the Class III watercourses downstream of the channel head. The resulting additional sediment load may have an adverse impact on steelhead and their habitat in lower Little Creek.

· Increased saturation of the Ohlson Ranch Formation at its contact with the underlying Franciscan Formation is expected to result in slumping of the material into Class III watercourses.

· The potential for significantly increased levels of suspended sediment in the class III watercourses in upper Little Creek. The resulting additional sediment load may have an adverse impact on steelhead in lower Little Creek. The additional sediment load may also have an adverse effect on the aquatic habitat of upper Little Creek.

· Lack of a properly designed channel to route flood water from the reservoir spillway to the channel network.

The Project Description in the Mitigated Declaration is incomplete and misleading.

Because of these deficiencies in the THP, the Project Description, and inadequacy of the proposed Mitigated Negative Declaration, CDF should deny the Hansen\Whistler THP 1-04-030 SON.

Sincerely,

Dennis Jackson

Hydrologist

Cc: Allen Robertson, Deputy Chief, California Department of Forestry and Fire Protection

References

Blake, M.C. Jr, Judith Terry Smith, Carl Wentworth and Robert H Wright, Preliminary Geologic Map of Western Sonoma County and Northwesternmost Marin County, California. USGS Open-File Report 71-0044, 1971.

Cermak, J. and J. Kucera, 1987. Transpiration of mature stands of spruce (Picea abies (L.) Karst.) as estimated by the tree-trunk heat balance method. In Forest Hydrology and Watershed Management – Proceedings of the Vancouver Symposium, IAHS, August 1987.

Chang, Mingteh, 2003. Forest Hydrology: An Introduction to Water and Forests, CRC Press, New York.

Church, Michael, Brett Eaton, 2001, Hydrological Effects of Forest Harvest in the Pacific Northwest Department of Geography, The University of British Columbia, Vancouver, British Columbia, V6T 1Z2 Riparian Decision Tool; Technical Report #3, June, 2001

Dunne, T. and L.B. Leopold, 1978. Water in Environmental Planning. W.H. Freeman and Company.

Fuller and Custis, Gualala River Watershed Assessment Report. North Coast Watershed Assessment Program Appendix 2, Report on the Geologic and Geomorphic Characteristics of the Gualala River Watershed, California, by, December 2002, California Geological Survey

Jaeger, Kristin Channel-Initiation and Surface Water Expression in Headwater Streams of Different Lithology. A Master’s thesis University of Washington, 2004

Keppeler, E.T. 1998. The summer flow and water yield response to timber harvest. PSW-GTR-168, Pacific Southwest Research Station.

Keppeler, E.T. and R.R. Ziemer, 1990. Logging Effects on Streamflow: Water Yield and Summer Low Flows at Caspar Creek in Northwestern California. Water Resources Research, Vol 26, No. 7, pages 1669-1679, July 1990.

Montgomery, David R., William E. Dietrich, 2002, Runoff generation in a steep, soil-mantled landscape Water Resources Research, VOL. 38, NO. 9, 1168, doi:10.1029/2001WR000822, 2002

O’Conner Environmental, Assessment of Potential Hydrologic Effects, Fairfax Timber Harvest Plan and Conversion Number 1-01-171 SON, Grasshopper Creek and Annapolis Watersheds, Sonoma County, March 15, 2002

Rantz, S.E. and T.H. Thompson, 1967. Surface-Water Hydrology of California Coastal Basins Between San Francisco Bay and Eel River. U.S. Geological Survey Water-Supply Paper 1851. Prepared in cooperation with the California Department of Water Resources.

Rantz, S.E. 1974. Mean Annual Precipitation in the San Francisco Bay Region, California, 1931-70. Miscellaneous Field Studies Map MF-613.

Reid, L.M., 1998. Cumulative Watershed Effects: Caspar Creek and Beyond. PSW-GTR-168, Pacific Southwest Research Station.

Soil Survey Staff, Natural Resources Conservation Service, United States Department of Agriculture. Official Soil Series Descriptions [Online WWW]. Available URL: http://soils.usda.gov/soils/technical/classification/osd/index.html

Whipkey, R. Z., and M. J. Kirkby. c1978. Flow Within the Soil. Pages xvi, 389 p. : in M. J. Kirkby, editor. Hillslope hydrology. Wiley,, Chichester ; New York :.

Ziemer, R.R. 1997. Caspar Creek Thornthwaite potential evaporation, water years 1990-1995. U.S. Forest Service, Pacific Southwest Research Station, Redwood Sciences Laboratory, Arcata, CA. See web site:

http://www.rsl.psw.fs.fed.us/projects/water/Thornthwaite.html,

Figure 1. The proposed Hansen/Whistler timberland conversion is shown superimposed on the 1971 Preliminary Geologic Map of Western Sonoma County and Northwesternmost Marin County, California. The Annapolis portion of the map was drawn on the 7.5’ Annapolis quadrangle from 1943 with 25’ contour lines. The project location is approximate because differences in the base topographic maps makes perfect alignment difficult.

Figure 2. The Hansen/Whistler conversion (project area is shown as hatched) is on a ridgetop separating Little Creek from Buckeye Creek. The top of the ridge is underlain by the Ohlson Ranch formation which lays on top of the Franciscan Formation. Removal of the trees on about 14 acres that drain towards Little Creek is expected to cause headcutting of Class III streams upslope of the Ohlson Ranch/Franciscan boundary approximated by the red dashed line in the figure. The short blue dashed lines in the figure show examples of hollows that might experience headcutting as a result of the Hansen/Whistler conversion. The wet area shown along the project boundary near the 840-foot contour line needs to have a more complete description in order to determine its role in the hillslope hydrology of the in an near the project area.