As the population and interest in recreational activities increase in the Portland/Vancouver area, public concern for forest management practices has also grown. Historically, clear-cutting practices have led to dense road networks, degraded aesthetics, damaged aquatic ecosystems, and low biodiversity. More sound ecological alternatives to clear-cutting are being researched, experimented with, and implemented throughout the state of Washington. As forest practices change to alleviate environmental concerns, a detailed harvest plan must be outlined combining engineering and ecological design principles.

During the Spring of 1998, the senior class of Forest Engineering undergraduates of the University of Washington, under the direction of Dr. Peter Schiess, developed a harvest plan for 13,000 acres of the Northwest portion of the Washougal Landscape (T03N R05E and T03N R06E, WM) east of Vancouver, WA. After months of preparation and three weeks of fieldwork, a final report and presentation were submitted to interested officials of the Department of Natural Resources in June of 1998, A Management Plan for the Washougal Landscape. The report you are reading today, is a condensed version of the harvest plan proposed to the Southwest Region. The original report involved a large amount of engineering design that will not be discussed here, such as road design, setting design, and watershed analysis. This report focuses on the silvicultural alternatives of harvesting practices in the area. It is explicitly written for decision-makers of DNR forest management activities in the Washougal area that do not have the desire to wade through hundreds of pages of technical content.

The main objectives for the Southwest Region harvest planners, are to maximize economic benefit to Skamania County, and to follow guidelines outlined in the Department of Natural Resources’ 1997 Habitat Conservation Plan (HCP). Due to the intense fire history, severe wind shear, and rocky soil, the trees in the area demonstrate severely stunted growth rates. Based on a habitat survey in 1997, the area does not meet minimum habitat conditions (as defined in the HCP) for dispersal habitat of the Northern Spotted Owl. The HCP requires that at least 50 percent of the area that could potentially support habitat, needs to meet the given requirements for dispersal habitat. Currently, just under 50 percent of the dispersal areas meet the habitat requirements. The stunted growth of the trees also causes concern for the merchantable volumes that can be harvested. When conducting our analysis, the following constraints were considered:

1. Maintain at least 50 percent of the dispersal areas as suitable dispersal habitat for Northern Spotted Owls ("Final Habitat Conservation Plan" IV.12) at any given time.

• Relative Density Index > 50 (approximate to 70 percent canopy closure)
• Quadratic Mean Diameter (QMD) of 100 largest trees > 11 inches
• Height of 40 largest trees from above diameter class > 85 feet
• At least three large snags

1. Harvesting must produce economic benefit to the Trust.

In order to meet the 50 percent requirement of dispersal habitat, management activities in the near future must create both habitat and revenue. Dispersal stands not currently designated as habitat should then be coaxed into growing faster, or left alone to reach habitat conditions on their own. Different types of thinning have proven to stimulate growth in remaining stands. Therefore, it was first hypothesized that the Washougal area may respond well enough to thinning to create habitat conditions in the near future and create revenue. With this in mind, three main harvesting alternatives for the Washougal Planning Area were considered in the analysis:

• Non-Commercial Thinning
• Commercial Thinning
• Regeneration Harvest (Clear-cutting)

The following sections include the conclusions, recommendations, and discussion of our analysis. The Conclusions Section describes and interprets the results of our analysis of the three silvicultural alternatives. The Recommendations Section lists suggestions that we have for the implementation for harvesting strategies in this area. The Discussion Section goes into the details of analysis. It describes the three alternatives, our design constraints, the current stand conditions, our growth modeling results, a summary of the cost comparisons of the alternatives, and our harvest scheduling results.

From our analysis of three different harvesting alternatives, non-commercial thinning, commercial thinning, and regeneration harvests, we found that regeneration harvests are the most ecological and economical in the Washougal Planning Area. Our main objectives were to comply with the HCP by maintaining 50 percent dispersal habitat for the Northern Spotted Owl, and maximize economic benefit to the beneficiaries of state forests. Our growth models have shown that no thinning option in the Washougal will create dispersal habitat conditions in a stand within the next 25 years. They have also shown that 85 percent of the area will reach habitat conditions on its own within the next five years. Since the thinning options do not show benefit in the creation of dispersal habitat, and are much more costly than the traditional method of regeneration harvests, they were not considered realistic in the Washougal Planning Area. We believe that it is reasonable to sustain a harvest of four million board feet a year from the area in regeneration harvests, and still maintain the HCP requirements for dispersal habitat.

In order for harvesting practices in the Washougal Planning Area to meet the 1997 HCP requirements and still provide economic benefit to the Trust, the following recommendations are suggested:

1. If new analysis results have not proven otherwise, harvesting practices in the Washougal Planning Area should be restricted to regeneration harvests.
2. Road construction, and harvest scheduling should be done in a similar order as SNAP suggests ("Management Plan", Chapter 14) in order to maintain dispersal habitat, adjacency regulations and minimize costs.
3. If new analysis results have not proven otherwise, four million board feet a year may be harvested from the Washougal Planning Area in the next 25 years in the form of regeneration harvests.
4. After regeneration harvests, new stands must be planted under intensive management conditions in order to create more dispersal habitat 40 to 50 years later. These newly planted stands may respond well to thinning, fertilizing, and other techniques to promote growth.
5. When better forest inventory data becomes available (from FRIS), reevaluate our analysis of future growth and volume predictions.

Our analysis included evaluating the feasibility of three harvesting alternatives in the Washougal Planning area. We had to analyze the current stand conditions and model how well they would grow under the three treatments: non-commercial thinning, commercial thinning, and regeneration harvests. Under each scenario, we then had to reevaluate our dispersal habitat requirements and make sure at least 50 percent of the area met the minimum criteria (listed in the following section). We also had to run a cost analysis of the scenarios, and develop a harvest schedule that would produce adequate revenue.

In the following discussion, we are going to talk about the design criteria to our harvest plan, the current stand conditions, growth model results of the three treatments, a brief cost comparison of the three alternatives, scheduling and optimization analysis, and the limitations to our models. The design criteria are the requirements that the harvesting option must meet in order to be a feasible harvesting option for the DNR. After we talk about those requirements in detail, we will discuss the current status of our stands and how they rate in terms of those criteria. Next, we will talk about how we projected the stands into the future, and simulated the three treatments. We will talk about the results of the growth modeling and how the future stands match up to the design criteria. The fourth section will summarize costs of the three alternatives. The fifth section will discuss our scheduling and optimization analysis using a program called SNAP, and how the dispersal habitat changes with our proposed harvest schedule over a 25-year period. The last section discusses possible sources of error with our models.

As mentioned briefly in the introduction, two main criteria have to be met in order for a sound harvest plan to be created: 50 percent of the area must be maintained as dispersal owl habitat, and the harvesting option must provide economic benefit to the stakeholders of these state lands.

The Washougal Planning Area has been designated by the DNR as a potential dispersal area for the Northern Spotted Owl. As a result, harvest planners must maintain at least 50 percent of the watershed as dispersal owl habitat that meets several requirements. The criteria for each acre of dispersal habitat are as follows:

• Relative Density Index > 50 (approximate to 70 percent canopy closure)
• Quadratic Mean Diameter (QMD) of 100 largest trees > 11 inches
• Height of 40 largest trees from above diameter class > 85 feet
• At least three large snags

After the HCP for owl habitat is satisfied, the DNR must consider the economic revenues that can be generated for the public stakeholders. It is vital to many state institutions, that the DNR create and maintain revenue from state forested lands. Therefore, the harvesting alternative selected must also show an effective way to produce revenue.

The Washougal Planning Area currently exhibits fairly uniform stand conditions due to the Yacolt fires and passive management history. As stated previously, the area is extremely unproductive due to severe wind shear, and damaged soils. As a result, stands demonstrate unusually low heights for their diameter classes. Average trees from the area gain little height after they reach a DBH of 9.5 inches. The average tree does not grow taller than 90 feet. Since one of the criteria of owl dispersal habitat is 40 trees over 85 feet tall per acre, this is most likely going to be the "limiting" factor to whether or not our areas can be considered habitat.

As a result of the 1997 Habitat Conservation Plan, an owl habitat survey of the area was conducted by the DNR to establish current habitat areas. The habitat survey designated the landscape into three categories: Dispersal Habitat, Dispersal Non-Habitat, and Non-Dispersal Non-Habitat. The Dispersal Habitat areas were designated as meeting the four requirements for dispersal habitat outlined in the HCP. Dispersal Non-Habitat areas were designated as not meeting the criteria presently, but could potentially meet the criteria in the future. Non-Dispersal Non-Habitat areas are areas that could never support habitat, such as rock outcrops and non-ownership lands. The results of the survey are summarized in Figure 1.

As you can see from Figure 1, about 50 percent of the area is now considered suitable dispersal habitat. Twelve percent of the area can never be habitat, so our main focus in prescribing silvicultural practices will depend greatly on the 38 percent of potential habitat left in the Washougal. Plenty of large snags are present in the Washougal Area residual from the Yacolt fires. Therefore, in our growth modeling, we shall consider the other three criteria, especially watching for dominant tree heights over 85 feet.

In order to predict stand development in the future, we used a program called LMS (Landscape Management System) developed at the University of Washington, Silviculture lab. It is a windows interface between growth modeler FVS (Forest Vegetation Simulator, USFS), UVIEW (Landscape Visualization System, Forest System Engineering Lab), SVS (Stand Visualization System), and LOGSORT (Log Sorting and Bucking). More information can be found in the "Management Plan" Chapter 10 or in McCarter 96: 17-23. The FVS growth rate was modeled with the West Cascades Gifford Pinchot Variant. Using this technology and current stand information, we were able to "grow" stands and predict how well they would respond in the future under different harvesting conditions. We were also able to set up the program to evaluate whether or not the three remaining habitat criteria had been met: relative density, QMD, and height.

When predicting future stand conditions in a growth model, it is very important to have good current stand information. Initial stand data for the growth modeling was derived from the biased 1997 habitat survey plot cards, 24 plots taken during the field portion of the project, and a 10 meter Digital Elevation Model (DEM). Other information was provided in other DNR GIS coverages provided, but was quickly found to be inaccurate, outdated, or irrelevant ("Management Plan", Chapter 6). Additional information from the Forest Resource Inventory System (FRIS), to be conducted this year, will prove to be invaluable to the future analysis of the area.

The idea behind the non-commercial thinning alternative is to create dispersal habitat from dispersal non-habitat quickly by removing competitors for growing space that might have naturally died out if left. This would hopefully enable the remaining stand to accelerate height growth and meet dispersal habitat conditions within the near future. This type of thinning operation is not intended to be profitable, hence the term "non-commercial", but rather, it is intended to create more dispersal habitat so that other more economical harvesting practices can occur elsewhere.

Using LMS, 10 dispersal non-habitat stands were "thinned" in the year 1998, and grown to the year 2023. The thinning was modeled as the removal of 30 percent of the smallest trees by diameter class. The stands were grown in five-year increments, and none of them reached habitat conditions in the 25 years analyzed. The results of two stands are shown in Table 1.

As you can see from the results above, the stands grown had no trouble meeting the diameter requirements (QMD) of at least 11 inches, but failed to produce either a high enough relative density (greater than 50), or 40 trees that were above 85 feet tall. Stand number 60385 did not meet either criterion of relative density or 40 trees above 85 feet, and stand number 60399 failed to meet the relative density requirement of at least 50. As stated earlier, the fourth criterion, of at least 3 large snags per acre was assumed to be met because of the numerous amounts of snags located in this region.

Another silvicultural alternative, would be to thin stands to meet both objectives of promoting habitat and creating revenue. The non-commercial thin took out only 30 percent of the smallest trees, where a commercial thin may take 40 percent or more of merchantable timber. Since we are thinning for profit, larger stands are needed to make it a profitable adventure. In our particular area, this is probably unreasonable due to the small volumes expected. Never the less, this alternative was analyzed with ten dispersal habitat stands thinned at 40 percent from below. The results of two stands are located below in Table 2.

Even though the stands initially met habitat conditions before they were thinned, the relative density of the stand never recovered sufficiently after 25 years to reach 50 in either stand. This suggests to me that the stands in the area are too mature to respond to commercial thinning. Because this was a thin from below, the largest trees were left and continued to maintain the other two criteria for dispersal habitat, diameter (QMD) above 11 inches and 40 trees above 85 feet. Both stands in Table 2 excelled far and beyond in those two categories. As in the Non-Commercial Thinning option, at least 3 large snags per acre were assumed to exist because of the numerous amounts of snags in the area.

The traditional method of timber harvesting is producing a series of clear-cuts. Public pressure has forced the timber industry to look into other alternatives, but few have proven to be as economically profitable. Clear-cuts have seemly severe initial effects on the landscape, but in our case, it appears that they have the same impact as thinning as far as removing acreage from dispersal habitat conditions. An individual stand that has been clear-cut will obviously not meet habitat requirements for a good 40 or 50 years into the future. The only requirement that will be met in this period will be snags that are left for wildlife trees. Our growth model results show that 85 percent of the area will meet habitat requirements in the next five years without any treatment at all. It may then afford the DNR to clear-cut and start over with planting a new stand under intense management, which in the long run, may be more beneficial towards the development of habitat. Regeneration harvests are definitely a more economical operation.

The three treatments in our analysis have all failed to meet all four habitat criteria in individual stands over a 25-year period. The results are summarized in Table 3.

According to our modeling, an individual stand cut in 1998 by any of the three treatments we are considering will not meet dispersal habitat requirements by 2023. The major factor is that the relative density index does not recover to at least 50 in any of the scenarios. In consequence, in order to meet the minimum requirement of 50 percent dispersal habitat in our area, we must carefully schedule our harvest units so that 50 percent of dispersal habitat is left alone. Potential habitat will have to become habitat of its own accord, and the harvested areas will have to become habitat in 40 or 50-year cycles.

Detailed equipment analysis, yarding costs, and revenues based on the timber market were analyzed in our report ("Harvest Plan" Chapter 6,7,14). To summarize our findings, the following table ranks the three treatments by cost, the highest cost is number 1.

In our analysis, we found that Non-Commercial Thinning is about three times more expensive then Regeneration Harvests, and Commercial Thinning was about twice that of the cost of Regeneration Harvests. Based on a purely economic point of view, regeneration harvests are the best option.

The three silvicultural alternatives are not necessarily stand-alone options. It is very possible that a combination of the alternatives would be a better plan. There are also economic decisions to be made on when and where harvesting should take place to minimize costs and maximize profits. To answer some of these questions, we ran an analysis with a program called SNAP, Scheduling and Network Analysis Program ("Harvest Plan" Chapter 14; Bettinger 565-577). SNAP was given a proposed road network, setting polygons, timber volumes, growth rates, and detailed cost information. It then decided, based on a rule of minimizing costs, what roads should be constructed, and which units should be harvested using the given silvicultural options. SNAP was especially useful for scheduling harvest units while maintaining 50 percent habitat (modeled at 55 percent) and also maintaining current FPA adjacency regulations. This analysis was run in the same five-year periods as those of LMS. SNAP also grows trees using the West Cascades Gifford Pinchot Variant, but is more simplified than the LMS program. LMS is a more accurate growth model because it can account for more variation within stands. Given a goal of harvesting four million board feet a year, SNAP was successful in scheduling harvest units while maintaining the 50 percent (55 percent) minimum habitat over 25 years. A graph showing the amount of habitat over the five periods are shown in Figure 2.

The figure shows that habitat is currently at about 50 percent and it peaks to 85 percent in the next five years (one period). The percentage of area of habitat then decreases gradually as more and more of the landscape is harvested, but 55 percent habitat is still maintained. The four million board feet a year goal was estimated by looking at both a 25-year harvesting period and a 100-year harvesting period. According to SNAP, these volumes are sustainable. In the 5 periods, SNAP chose only to thin two units, all the others were regeneration harvests. According to the SNAP program, it was more profitable to clear-cut as long as 55 percent of the area was maintained as habitat.

There are three main causes for concern with our analysis; the validity of our initial stand data, the projected growth rates by LMS and SNAP, and the SNAP method of thinning. The initial stand data was derived from a biased DNR habitat survey, in which plots were intentionally located in areas that were likely to be habitat, 25 plots run during the field portion of the analysis, and slope aspect derived from a DEM. As more stand data becomes available from the FRIS forest inventory, the accuracy of our analysis, if reran, would benefit enormously.

The growth modeling that took place overestimated the growth of the stands. Even though a growth rate specific to the region was used in the model (West Cascades Gifford Pinchot Variant), the model could not predict severe wind shear and localized characteristics that may have been unique to our planning area. As a result, trees continued to put on height growth through late seral stages, that may have not been realistic. This could cause us to overestimate the amount of dispersal habitat in our planning area.

The final limitation worth mentioning, is that SNAP does not have the ability to "thin from below" it took out 40 percent of the trees at random, and did not take 40 percent of the smallest diameter class. This would cause SNAP to overestimate the value of timber being harvested from thinning units, which implies that the two units SNAP chose to thin in our planning period, may not be economically realistic.

Bettinger, P., K. N. Johnson, and J. Sessions. "Forest planning in an Oregon case study:

Defining the problem and attempting to meet goals with a spatial-analysis technique."

Environmental Management 20 (Jul/Aug 1996): 565-577.

McCarter, James, et al. "Landscape Management through Integration of Existing Tools

and Emerging Technologies." Journal of Forestry 96 (June 1996): 17-23.

Washington State Department of Natural Resources. The Final Habitat Conservation

Plan. Commissioner of Public Lands. Washington: September 1997.

Washington State Department of Natural Resources, Southwest Region. A Management

Plan for the Washougal Landscape. University of Washington Forest Engineering

Undergraduates. Washington: June 1998.

Report Outline Chapter 14