SOIL COMPACTION ON A MECHANIZED TIMBER HARVEST OPERATION IN EASTERN OREGON By Richard R. Zaborske A PAPER submitted to Department of Forest Engineering Oregon State University Corvallis, Oregon 97331 in partial fulfillment of the requirements for the degree of Master of Forestry AN ABSTRACT OF THE PAPER OF Richard R. Zaborske for the degree of Master of Forestry in Forest Enqineerinq presented on May 26, 1989. Title: Soil Compaction on a Mechanized Harvest Operation in Eastern Oreqon Abstract approved: Henry A. Froehlich The use of mechanized harvest equipment has been increasing as an economical method to harvest small diameter timber. While the use of this equipment is increasing, little is known about associated soil compaction. In particular, little information exists on soil compaction caused by feller--bunchers. This study measured soil compaction caused by a mechanized harvest operation using 2 swing-boom, tracked feller-bunchers and 2 rubber-tired grapple skidders. The study took place on the LaGrande Ranger District of the Wallowa-Whitman National Forest, located in eastern Oregon on volcanic ash soils. USDA Forest Service definitions of detrimental compaction (20% or greater increase in soil density over pre-activity levels) were used as a guideline to determine if detrimental compaction had occurred. Results show that 54% of the total area was impacted by either the feller-bunchers, skidders or both. Feller- bunchers impacted 19% of the total area and caused a statistically significant increase in soil bulk density of 8.8% when compared to before logging densities. Main skid trails covered 12% of the total area and had a statistically significant increase in soil density of 36.3%, when compared to before logging densities and was considered detrimentally compacted. trails. These main skid trails were also feller-buncher Thus, 12% of the total area was impacted by both feller-bunchers and skidders while 7% of the total area was impacted only by feller-bunchers. Twenty three percent of the total area received 1 to 4 skidder passes and was not in identifiable skid trails. Even these non-skid trail areas showed a statistically significant increase in soil density of 9.6% when compared to before logging densities. An additional 12% of the total area received 5 to 8 skidder passes and was also not in identifiable skid trails. This area showed a statistically significant increase in soil density of 19.8% and was considered detrimentally compacted. Regression analysis showed that slash significantly reduced compaction caused by feller-bunchers and skidders. APPROVED: - Dr. Henry A. Froehlich - Major Professor Forest Engineering Dr. William A. Atkinson - Department Head Forest Engineering Paper presented May, 1989 DEDICATION To my wife, Vera Ellen For all the love, support and encouragement which made this paper possible. ACKNOWLEDGMENTS I would first like to thank the USDA Forest Service for providing me the opportunity to participate in the two-year advanced Logging Systems Program here at Oregon State University. Without this opportunity, this project would not have been possible. From the LaGrande Ranger District, I would especially like to thank Ranger Gail Kimbell for allowing me to do this study on her District. I would also like to thank John Wilkens, Operations Forester in Timber Administration and the Sale Administrators John Szymoniak and Bruce Rynearson for helping me locate the study area, providing me with the supplies I needed and offering encouragement during the data collection process; Dan Harkenrider and Art Kreger, Soil Scientists, who provided me with soils information; and Bob Dipold, Fleet Manager, who always had a vehicle available for me. From the Timber Sale Purchaser, Crisstad Enterprises Inc., I would like to thank Jack Barrymore, subcontractor. I would especially like to thank the 2 feller-buncher operators, Greg Jensen and Homer Gorbett and the skidder operator, Jerry Gorbett, for their cooperation, information and advice. I would particularly like to thank the members of my graduate committee, Professors Henry A. Froehlich, Loren D. Kellogg and Department Head Dr. William A. Atkinson. Their guidance, help and suggestions were crucial to the success of this project. Finally, I would like to thank Don Studier for his help in the data collection. Also, thanks to Don and my fellow graduate students in this program for all of the encouragement, ideas and support during the past 2 years. TABLE OF CONTENTS Paqe 1.0 2.0 3.0 INTRODUCTION 1 . 1.1 OBJECTIVES 2 1.2 SCOPE 3 LITERATURE REVIEW 4 2.1 METHODS OF MEASURING SOIL COMPACTION 4 2.2 AMOUNT OF AREA IMPACTED 7 2.3 INCREASES IN SOIL DENSITY 8 2.4 EFFECTS OF MULTIPLE PASSES 2.5 EFFECTS ON PRODUCTIVITY AND RATE OF RECOVERY FIELD STUDY DESIGN 11 . 14 19 3.1 STUDY AREA SELECTION 19 3.2 STUDY AREA DESCRIPTION 19 3.3 EQUIPMENT USED IN LOGGING 21 3.4 EQUIPMENT USED IN DATA COLLECTION 22 3.5 HARVEST PATTERN 24 3.6 3.5.1 FELLING PATTERN 24 3.5.2 SKIDDING PATTERN 25 DATA COLLECTION 27 3.6.1 MACHINE PASS GROUPINGS 29 3.6.2 SAMPLE POINT PREPARATION 32 3.6.3 BEFORE LOGGING 34 3.6.4 AFTER FELLING AND BUNCHING 35 3.6.5 AFTER SKIDDING 37 3.6.6 AMOUNT OF AREA IMPACTED 37 TABLE OF CONTENTS (continued) P aq e 3.7 4.0 LABORATORY ANALYSIS 3.7.1 SOIL MOISTURE DETERMINATION 3.7.2 CALIBRATION OF PROBE 39 . . . 39 39 RESULTS 41 4.1 NUMBER OF SAMPLE POINTS TAKEN 41 4.2 BULK DENSITY MEASUREMENTS 42 4.2.1 4.2.2 4.2.3 COMPARISONS WITH BEFORE LOGGING MEASUREMENTS 43 CHANGES IN BULK DENSITY WITH INCREASING DEPTH 46 CHANGES IN BULK DENSITY WITH INCREASING NO. OF PASSES 48 4.3 SLASH DEPTH 49 4.4 AREA IMPACTED 51 4.5 REGRESSION ANALYSIS 55 4.6 SOIL MOISTURE ANALYSIS 59 5.0 DISCUSSION OF RESULTS 6.0 MANAGEMENT IMPLICATIONS AND RECOMMENDATIONS 7.0 FURTHER RESEARCH 68 8.0 REFERENCES 71 61 . . . 66 APPENDICES APPENDIX A - PERTINENT TIMBER SALE CONTRACT CLAUSES APPENDIX B - CALCULATION OF NUMBER OF SAMPLE POINTS FOR BEFORE LOGGING AND CONFIDENCE INTERVALS FOR THE SAMPLE OBTAINED LIST OF FIGURES Page Figure 2.1: Relationship Between Increases in Bulk Density and the Number of Machine Trips 15 . Figure 3.1: Map of Unit 1A Showing Access Road and Landing Location 20 Figure 3.2: Theoretical Ground Pressure Distribution Under a Loaded Feller-Buncher 23 Figure 3.3: Schematic Diagram of Felling Pattern 26 . Figure 3.4: Schematic Diagram of Skidding Pattern . . 28 Figure 3.5: Schematic Diagram Showing Relationship Between Machine Pass Categories 31 Figure 3.6: Schematic Diagram Showing Relationship Between Skid Trail Types 33 Figure 3.7: Map of Unit Showing Areas Not Sampled by Machine Type 36 Figure 4.1: Graph of Average Soil Density by Condition Figure 4.2: Graph of Percent of Area Impacted and Percent Increase in Soil Density for the 0 - 12 Inch Soil Layer . 44 56 LIST OF TABLES Paqe Table 4.1: Number of Measurements Used for Results . . . 42 Table 4.2: Average Soil Densities for Each Grouping Sampled 43 Table 4.3: Soil Density Comparisons With Before Logging Values 46 Table 4.4: changes in Soil Density With Increasing Depth 47 Table 4.5: Changes in Bulk Density With Increasing Number of Passes 49 Table 4.6: Slash Depth 51 Table 4.7: Area Impacted by Harvesting Activities . . 52 Table 4.8: Relationship Between Area Impacted and Density Measurements 55 Table 4.9: Regression Analysis Results 58 Table 4.10: Soil Moisture Results 60 1.0 INTRODUCTION Mechanization of felling and skidding offers considerable economic advantages over conventional logging methods used to harvest small diameter timber. Mechanized harvesting has the potential to not only increase production but also to reduce labor costs resulting in lower total logging costs. The use of mechanized harvest equipment has been increasing. Silversides (1984) traced the use of mechanized forestry equipment from World War II to 1984. He noted that early development of mechanized equipment focused on chainsaws and ground skidding equipment while later development included mechanical fellers and multi-function machines. He reported that widespread use of chainsaws began in earnest following the war and, while crawler tractors had been used for skidding, it wasn't until 1956 that the first stripped-down-four wheel drive truck was tested as a skidding tractor. The first production model of a mechanical feller was introduced in 1965. From these beginnings, Schuh and Kellogg (1988) reported, from a 1985 western US industry survey, that more than 140 timber companies and logging contractors were using one or more pieces of mechanized logging equipment ranging from fellerbunchers to delimbers and chippers to various types of skidders. While research has shown that soil compaction reduces 2 future timber yields (Froehlich, 1979; Froehlich, Miles and Robbins, 1986; Laing and Howes, 1983), there is currently little information as to the effect that this new mechanized harvest equipment has on forest soils. Most current research has focused on soil compaction caused by the skidding phase of the operation, and the majority of these studies have taken place in the southeastern United States where soils and climatic conditions are considerably different from the Pacific Northwest. Few studies have focused on soil compaction caused by feller-bunchers and even fewer have followed a mechanized logging operation from start to finish. It is the purpose of this study to evaluate soil compaction caused by the felling and skidding phases of a mechanized harvest operation in eastern Oregon. 1.1 OBJECTIVES The objectives of this research project are: For a mechanized harvest operation using tracked, swing boom feller-bunchers and rubber-tired, grapple skidders; Estimate the amount of compaction and percent of area impacted by feller-bunchers. Estimate the amount of compaction and percent of area impacted by skidders. Compare the results of this study with respect to USDA Forest Service, Region 6 guidelines for acceptable levels of compaction. 3 1.2 SCOPE This field study reports on soil compaction resulting from logging a beetle damaged stand of lodgepole pine in eastern Oregon with feller-bunchers and rubber-tired, grapple skidders. It is a case study limited to one site, using logger selected pieces of equipment. The most important timber sale contract specifications which apply to this study are summarized below. These complete timber sale clauses are presented in Appendix A. CONTRACT CLAUSE CT6.O - OPERATIONS "Trees shall be skidded or yarded to the landing fulllength with one end suspended. Skid trails shall be at least 60 feet apart, center to center, except when converging. Single tractor passes off of skid trails are allowed as long as rutting or detrimental compaction (refer CT6.4# (Option 3)) is avoided." CONTRACT CLAUSE CT6.4* (Option 3) - CONDUCT OF LOGGING If designated skid trails are detrimentally compacted, they shall be ripped through the compacted zone but not to exceed a depth of 20 inches. Ripping shall be done with winged subsoilers during the normal operating season. Detrimental compaction is defined as a 20% increase in bulk density from undisturbed levels." 4 2.0 LITERATURE REVIEW Review of pertinent literature will focus on 5 major categories. They are Methods of measuring soil compaction. Amount of area impacted. Increases in soil density. Effects of multiple passes. Effects on productivity and rate of recovery. Inherent in topics 3 and 4 above is the significant effect soil moisture has in the amount of compaction caused by harvest operations. 2.1 METHODS OF MEASURING SOIL COMPACTION Alexander (1985) defined soil compaction as a reduction of soil volume due to a decrease in pore space. Compaction can be measured directly by changes in soil bulk density, or indirectly by changes in soil strength or changes in macropore space. Alexander (1985) and the American Society of Agricultural Engineers (1958) report that changes in soil strength can be measured by a cone penetrometer. Changes in macropore space can be measured by changes in infiltration rates. Both of the above references have identified 4 methods for measuring changes in bulk density. The first 3 methods involve sampling a volume of soil while the fourth method allows determination of bulk density in place. A brief 5 explanation of each method follows. Core Method - This method involves driving a cylindrical core of known volume into the soil and digging the core out to obtain the sample which is then taken to a lab where it is oven dried and weighed. Bulk density is the dry weight of the core divided by it's volume. This method is fairly simple and can give accurate results. Care must be taken in collecting and transporting samples and cores are difficult to collect in gravelly soils or in soils with large amounts of roots. Clod Method - This procedure involves collecting clods of soil in the field, coating them with paraffin or Saran to prevent moisture loss and transporting them to a laboratory for analysis. In the lab, weight, volume and moisture content of the clod is determined and bulk density calculated. This method works well in soils which are too hard or stony for collecting core samples. However, it can be quite time-consuming and tends to overestimate bulk density by underestimating the amount of pore space, since clods generally correspond (at least partly) to soil ped faces and thus the void space between clods is not measured. Irreqular Hole Method - This method involves excavating a shallow hole, obtaining dry weight of soil removed and using a material such as sand or water to determine volume of soil removed. This method works well on near-level surfaces, although Flint and Childs (1984) have 6 developed a method using styrofoam beads which works well on steep slopes. This method is time consuming, with difficulty in measuring the volume of a hole in the side of a soil pit. The equipment used in this method can be cumbersome in the field. 4) Transmission of Gamma Radiation - This is the method of measuring soil bulk density used in this study. Vomocil (1954) describes this method as "a means of making rapid measurements of bulk density of soil . does not require the removal of a large sample." It . .that involves use of gamma rays, since the amount of gamma radiation absorbed by the soil is proportional to it's mass. Measurement is carried out by placing a radiation source (usually Cesium 137) in the soil a fixed distance from a radiation detector and counter. The source is usually a probe inserted in the soil at a known depth with the Cesium source at the tip. The detector is either located in the body of the device on the soil surface, or in a different probe inserted in the soil at the same depth as the source at a fixed distance. The detector counts the number of pulses it receives and, by suitable calibration, soil density can be determined (Van Bavel 1958). In this study, the number of pulses recorded will be referred to as the number of counts. This method is quick and can give precise results. is important to determine soil moisture content and the It 7 probe must be calibrated for the soil types being studied. Since the gauge measures amounts of radiation absorbed by all material between the radiation source and detector, it is important that the sample spot be properly prepared by removing all litter and debris which could affect measurements. Also, reported bulk densities include gravel and stones found in the soil. Thus, results from very stony or gravelly soils may not be accurate in terms of reporting fine earth bulk densities. 2.2 AMOUNT OF AREA IMPACTED Studies which have investigated effects of logging on soils have frequently reported the percent of harvested area disturbed. Soil disturbance includes mixing of litter with soil in place and removing litter and/or duff to expose bare mineral soil. Froehlich and McNabb (1983) note that Pacific Northwest forest soils are generally low in bulk density and strength and are highly porous. Because of these factors, many Northwest soils are easily disturbed and compacted during timber harvest activities. In northeast Washington, Laing and Howes (1983) reported that 41.8% of an area logged with feller-bunchers and rubber-tired skidders had suffered soil damage. bunchers covered 20 - 25% of the area logged. skidders raised the impacted area to 41.8%. has found similar results: Feller- Clam bunk Other research 8 Dyrness (1965) - 26.8% of a tractor logged stand in the - western Cascades of Oregon was compacted. Aulerich, Johnson and Froehlich (1974) - 16 - 27% of - an area tractor-logged near Corvallis, Oregon was in skid trails. - Steinbrenner and Gessel (1955) - 26% of an area tractor logged in southwestern Washington was in skid trails. - Hatchell, Ralston and Foil (1970) - 32% of an area tractor-logged on the Atlantic Coastal Plain was in skid trails. - Campbell, Willis and May (1973) - 23% of an area in Georgia logged with rubber-tired skidders had been disturbed. Preplanning skid trails may be a way to reduce the amount of area impacted by logging. In the Oregon Coast Range, Sidle and Drlica (1981) found that 13.6% of the total area yarded with an FMC skidder was in skid trails, location of which had been preplanned. Of this area in skid trails, 25% showed less than a 10% increase in soil density, 38% showed a 10 - 27% increase in soil density and 37% of the impacted area showed an increase in soil density greater than 27%, when compared to before logging densities. 2.3 INCREASES IN SOIL DENSITY Much research has been done which investigated the 9 relationship between logging and soil compaction. Adams and Froehlich (1981) point out a number of factors which contribute to how much a soil is compacted. These factors include the amount and type of pressure and vibration applied, depth and nature of surface litter, and soil texture, structure, and moisture content. Since bulk density increases with compaction, changes in bulk density are often used to describe soil compaction. In Arkansas, Moehring and Rawls (1970) found that wet weather logging significantly decreased soil macropore space and increased soil density in the 0 - 2 inch layer, compared to adjacent, undisturbed sites (alpha = 0.05). They also reported that 5 years after logging, traffic on 3 or 4 sides of trees significantly reduced tree growth (alpha = 0.05). Logging traffic on 1 or 2 sides of trees did not influence tree growth. They found that while dry weather logging did scarify the soil surface, it did not affect tree growth, soil density or macropore space (alpha = 0.05). In North Carolina on a loblolly pine plantation, Gent, Ballard and Hassan (1983) found a 23% increase in soil density at a depth of 12 inches in primary skid trails, when compared to pre-harvest densities. Rubber-tired feller- bunchers and skidders were used to whole-tree and treelength yard the unit. Other studies have found similar results: - Stokes and Sirois (1982) - logging corridors 10 showed a 15% increase in soil density at the 2 inch depth and an 11% increase in soil density at the 4 inch depth. These increases were significantly higher (alpha = 0.05) than densities found in undisturbed soils. The study was done in southeast Louisiana immediately after a unit had been felled, limbed, bucked and piled by a tracked machine, but prior to skidding. - Steinbrenner and Gessel (1955) - an average increase in soil density of 34.9% in skid trails. Densities in areas adjacent to logged areas ranged from 0.479 to 1.117 gm/cc. Densities in skid roads ranged from 0.763 to 1.394 gm/cc. - Dyrness (1965) - 48% increase in soil density. Before logging densities were 0.657 gm/cc while after logging densities were 0.975 gm/cc - Campbell, Willis and May (1973) - average increase in soil density of 16% in disturbed areas versus undisturbed areas. - Hatchell, Ralston and Foil (1970) - soil densities increased from 0.75 gm/cc in undisturbed soils to 0.92 gm/cc in secondary trails, and to 1.08 gm/cc in primary trails, a 22% and 44% increase, respectively. - Aulerich, Johnson and Froehlich (1974) - reported the following increases in soil density on tractor thinning of young-growth Douglas-fir: 11 BULK DENSITY INCREASE (%) 0-6 in. Major skid trails Secondary skid trails Lightly used trails 2.4 21 16 13 6-12 in. 17 12 6 EFFECTS OF MULTIPLE PASSES Research has shown that only a few passes over the same area are needed to increase soil densities to near their maximum. On a volcanic soil, Lenhard (1986) reported that soil density reached a maximum after 4 trips with an unloaded skidder traveling at slow speeds. While soil density remained statistically constant with up to 32 trips, pore size distribution changed with increasing number of passes between 4 and 32. In a vehicular compaction test, Hatchell, Ralston and Foil (1970) reported that an average of 2.5 trips resulted in soil densities within 10% of the maximum attained. - Other studies have found similar results: Froehlich (1978) - reported the following results with increasing number of passes on 3 sites where a loaded, low ground-pressure, torsion-suspension skidder was used: Mt. Hood Site - At the 2 inch depth, soil density increased in the first few passes and changed little after 6 passes. At the 6 inch depth, density increased with increasing number of trips up to 6, at which point it reached its maximum and remained virtually the same, even after 20 trips. At the 10 inch depth, soil density changed 12 little regardless of the number of passes. Soil moisture content was 43% and surface litter ranged from 2 to 8 inches in depth. Soil densities reported were: SOIL DENSITY (GM/CC) Depths j (inches) SOILS Undisturbed Skidder Trail 1-3 trips 6-10 trips 15-20 trips 50-60 trips 2 0.65 0.87 0.88 1.03 0. 89 1 . 06 1 .01 1.12 1.24 0.96 0.96 0.92 0.95 0.90 1.01 Umpqua Site - Soil density changed little at any depth, This lack of change was attributed to even after 20 trips. high initial densities, thick litter layer (5 to 9 inches) and lower soil moisture (29%). Soil densities reported were: SOIL DENSITY (GM/CC) Depths of (inches) SOILS Undisturbed Skidder Trail 1-3 trips 6-10 trips 15-20 trips 90-100 trips 2. 1.06 1 . 09 1.08 1.09 1.33 LQ 1.16 1.06 1.06 1.08 1.04 1.17 1 .04 1 .08 1.18 1.13 1.16 1.34 1 . 29 1 .33 1 . 07 Malheur Site - Density of surface soils increased in the first few trips with little change thereafter. There was a lesser increase in density at depths of 4 and 6 inches. Densities did not change at the 10 inch depth. Soil moisture was 13% and the litter layer ranged from 0.2 13 to 2.5 inches deep. Reported soil densities were: SOIL DENSITY (GM/CC) Depths of (inches) SOILS Undisturbed 1.09 1.05 1.03 1.06 Skidder Trail 1-3 trips 6-10 trips 15-20 trips 1.21 1.24 1.25 1.12 1.12 1.17 1.11 1.11 1.13 1.01 1.01 1.08 - 2 Burger, et al. (1985) - using an unloaded rubber- tired skidder and crawler tractor on a clay soil found that neither machine had an effect on soil density at the 6 to 8 inch depth. Density of the top 2 inches sharply increased with the first 3 passes with smaller increases for up to 9 passes. The level of soil moisture significantly affected the degree to which the top 2 inches of soil were compacted (alpha = 0.05). At 18% soil moisture content, overall density increased by 0.10 gm/cc while at 21% moisture the increase was 0.18 gm/cc. Despite the fact that the crawler tractor had a mean ground contact pressure 3.7 times that of the rubber-tired skidder, the effect of the 2 machines on soil density was the same. - Froehlich et al. (1980) - developed a regression model which only contained terms for number of trips and cone index. This model explained 54% of the variation in change in bulk density. They found that about 70% of the increase in bulk density had occurred by the fifth trip. density curve produced from their prediction equation is A 14 given in Figure 2.1. Their results are based upon data collected on the Tahoe National Forest on 4 soil types, ranging in texture from loam to gravelly loamy sand, and 3 different skidders (rubber-tired, crawler tractor and a torsion-bar suspension machine). 2.5 EFFECTS ON PRODUCTIVITY AND RATE OF RECOVERY Compaction alters soil properties such as strength, density pore space, and size. These soil properties control root penetration and aeration, which in turn affect the ability of plants to uptake moisture and nutrients, and ultimately its growth rate. Thus, altering these soil properties can reduce future growth and forest productivity. Growth loss studies have used weighted averages of trees found throughout the width of skid trails in determining growth reductions. While it may be possible that growth losses could be minimized by careful placement of planted seedlings (planting all seedlings near the edges of skid trails and none up the middle or not planting at all in skid trails and planting more seedlings in undisturbed areas), it cannot be said with certainty what effect, if any, this would have on minimizing growth losses due to soil compaction. There are many other factors besides soil compaction which may affect growth reductions due to logging activities. Some of these factors may be puddling of the soil surface resulting in reduced infiltration rates and 2 4 6 8 15 Number of Vehicle Trips 10 20 Figure 2.1 RelationshIp between Increases in bulkdensity and the number of machine trips (Froehlich et al. 1980) 5 30 16 removal or displacement of the nutrient rich surface layer. Froehlich, (1979) reported that 16 years after logging on the Ochoco National Forest in Oregon, soil densities in skid trails at 3 and 6 inch depths were 18% higher than undisturbed soils. Densities at 9 and 12 inch depths were 9% higher than undisturbed soils. Ponderosa pine (Pinus ponderosa) growth rates were reduced by 6% on moderately impacted trees and 12% on heavily impacted trees over this same 16 year period. Cascades, On a tractor logged site in the Oregon Wert and Thomas (1981) found that heavy compaction existed over 25% of an area logged 32 years earlier. Only the top 6 inches showed some recovery from compaction. New stand volume was reduced by 11.8%. Other studies have shown similar results: - Froehlich, Miles and Robbins (1986) - studied ponderosa pine and lodgepole pine (Pinus contorta) sites logged 23 and 14 years earlier, respectively, in central Washington. They reported that on the ponderosa pine sites soil densities in skid trails were 15% higher when compared to undisturbed soils. Over the past 5 years, trees in skid trails had a 5% reduction in height growth, 8% reduction in diameter, and a 20% reduction in volume growth, when compared to trees in undisturbed areas. Skid trails in lodgepole pine sites had densities 28% higher than undisturbed soils. They found, however, that increases in soil density were not significantly related to growth of 17 this species. - Hatchell, Ralston and Foil (1970) - using regression, estimated it would take 18 years for soils under log decks to return to soil densities found in undisturbed soils. - Laing and Howes (1983) - found an average increase in soil density of 36% when compared to pre-activity levels. They predicted a 35% reduction in volume on these areas over the next rotation. - Froehlich, Miles and Robbins (1985) - reported that on a volcanic soil in Central Idaho, soil densities in skid trails at depths of 2, 6 and 12 inches had not returned to densities found in undisturbed soils 23 years after logging. The differences in soil densities between skid trails and undisturbed soils were significant (alpha = 0.01). - Helms and Hipkin (1986) - intensively studied a 1.2 acre site within a 16 year old pine plantation in California. Bulk density increased 30% in a skid trail compared to soils of lowest bulk density measured. They also stated that in this skid trail, 15 years after planting establishment, a 29% reduction in mean tree volume, combined with 37% mortality, resulted in a reduction of 55% in volume per unit area. In areas adjacent to the skid trail, bulk density was 18% higher and had a reduction of 13% in volume per unit area. 18 Youngberg (1959) - studied growth rates on planted - Douglas-fir (Pseudotsuqa menziesii (Mirb] Franco) seedlings in Lane Co., Oregon. he reported the following changes in bulk density: Bulk Density (gm/cc) Condition Plantation 1 Depth (in.) Plantation Cutover 0 - 6 6 - 12 0.87 0.98 0.88 0.88 Skid Roads 0 - 6 6 - 12 1.58 1.73 1.52 1.59 Road Berms 0 - 6 6 - 12 1.01 1.05 0.89 0.88 He also reported the following average height growth rates for 2 year old seedlings. The differences in height growth between cutover and both skid road and road berms was significant (alpha = 0.01). Condition Cutover Skid Roads Road Berms - Heiqht Growth (inches) Plantation 1 Plantation 2 6.38 4.73 5.71 7.01 3.16 4.98 Cochran and Brock (1985) - studied 2 small clearcuts on the Deschutes National Forest. Eight years after logging, 37% of the combined total area had increases in soil density of 15% or more when compared to soils outside of the clearcut boundaries. Using regression, they concluded that compaction reduced ponderosa pine height growth for the 5-year period after planting. 19 FIELD STUDY DESIGN 3.0 All measurements were made during the period July, 1988 to August, 1988, on cutting unit 1A of the Hoosier Timber Sale, LaGrande Ranger District, Wallowa-Whitman National Forest, located in eastern Oregon. 3.1 Study Area Selection Several cutting units of this timber sale were to be harvested using feller-bunchers and grapple skidders during the summer of 1988. This particular unit was chosen because it was easily accessible and logged when enough time was available to do the study. Field data were collected during the months of July and August, 1988. 3.2 Study Area Description The unit was 26 acres in size. Midpoint elevation was 5700 feet above sea level. Average slope was 12% with a few areas having slopes of 25%. Figure 3.1 shows the outline of the unit, access roads and landing location. Unpublished soil survey data classifies this unit as a Grand-fir/Huckleberry/Douglas-fir plant association. A US Forest Service timber cruise estimated an average net volune of 41 cunits per acre. Almost three-quarters of this volume was dead, both standing and down, lodgepole pine killed within the past 10 years by bark beetle attacks. diameters were generally less than 10 inches. The Tree 0 Feet 400 Approximate Scale Access Road Landing Unit Boundary Figure 3.1 Map of Unit 1A Showing Access Road and Landing Location 800 - - KEY 21 silvicultural prescription called for removing all live and dead lodgepole pine. All healthy Douglas-fir trees were to be left, and areas of existing regeneration impacted as little as possible by logging information). (USDA Forest Service Data on average tree diameter, piece size and trees per acre were not available. Soil inventories for the unit showed soils derived from volcanic ash overlaying a clayey subsoil. Total depth to bedrock was 20 to 40 inches. Thickness of the ash layer ranges from 14 to 20 inches. Texture of the ash layer is silt loam with generally less than 5% cobbles. of the surface was covered with stones. Less than 1% The soil is well drained, has moderate permeability rates throughout it's depth and has moderately high erosion potential (Unpublished soil survey information). Soil moisture contents found during the course of this study ranged from 22% to 27% throughout the surface 12 inches of soil. 3.3 Equipment used in locqinq Two feller bunchers were used to fell the stand; a Mitsubishi MS-180 and a Hitachi UHL 7-7. Both machines are tracked, swing boom, converted backhoes (shearing head and additional guarding added) with hydraulic shears capable of shearing a tree up to approximately 20 inches in diameter at ground level. Both machines had 30 foot booms and weighed approximately 53,000 pounds, unloaded. Tracks were 22 25 inches wide and 12 feet long (ground contact) with 1 inch grousers. Static, unloaded ground pressure on level ground is approximately 7 pounds per square inch, assuming uniform pressure distribution. Figure 3.2 shows the theoretical ground pressure distribution under the feller-buncher. Two rubber-tired, grapple skidders were used to skid the unit. One was a John Deere 548D Turbo and the other a John Deere 540. Detailed tire information such as footprint size, inflation pressure and tire type was not collected so ground contact pressure could not be calculated. The 548D had a gross weight of approximately 25,000 pounds, tire width of 29 inches, and tire diameter of five and one-half feet. The 540 had a gross weight of approximately 20,000 pounds, tire width of 18 inches, and tire diameter of five feet. All timber was chipped on site at the landing using a chain flail delimber/debarker prototype developed by Gibson Chip Co., and a MORBARK model 23 Total Chipharvester. 3.4 Equipment used in Data Collection Bulk density was selected as the measure of soil compaction because it is a direct method and easily measured in the field. A single probe nuclear densiometer was used. The single probe meter gives the "average" soil density for the soil layers between the probe tip and the soil surface. This meter does not give point measurements or soil density Figure 3.2 Theoretical Ground Pressure Distribution Under a Loaded Fefler-buncher (Lysne and Burditt, 1983). Longer Imes indicate greater pressure. 24 at a particular depth. Originally, a double probed meter was to be used so densities at specific depths could be measured and results compared with other studies. However, just before the study began, the double probed meter malfunctioned and no substitute could be found. Rather than delay the entire project, the decision was made to proceed with the single probe meter. Bulk density measurements were taken at depths of 4, and 12 inches. 8 Four inches was considered to be representative of the surface layer. The 8 and 12 inch measurements were taken in an attempt to determine the pattern of change in density in the surface 12 inches of soil. In addition to the single probe meter, a soil corer was used to take soil samples for gravimetric soil moisture determination. This is necessary in order to convert all density measurements to dry weight basis for comparison. 3.5 3.5.1 Harvest Pattern Fellinq Pattern Both operators worked in a similar fashion. They would basically stay in 1 spot, reaching up to 20 feet into the stand in front of them and on both sides of the machine, collecting material to make a bunch. Occasionally, they would need to move forwards or backwards a few feet to 25 reposition themselves or to reach around obstacles such as leave trees. The more dense the residual timber, the more repositioning they would have to do. When finished in a spot, they would move forward 10 - 20 feet and repeat the Again, the more dense the residual process described above. timber, the shorter the distance between spots. They would cut trails in the unit approximately 40 feet apart, laying the trees in bunches next to their trail with the butts pointing towards the landing. along a unit boundary. The first trail would be When a convenient stopping place such as a dog leg in the unit was reached, the operator would back out on the trail just cut, move into the unit approximately 40 feet and begin cutting a new trail. Due to the heights of the trees, the previously cut trail would be covered with the boles or tops of trees now being cut. bunch contained approximately 15 trees. Each Since the material was being chipped, all sound material which could be bunched was taken. Thus, considerable time was spent bunching down material, which was jackstrawed. Because of this, the operators estimated it took them approximately 40% longer to cut and bunch this unit compared to a unit of similar size having predominately standing timber. Figure 3.3 is a schematic diagram of the tree bunch pattern. 3.5.2 Skiddinq Pattern Both skidder operators worked in a similar fashion. Figure 3.3 Schematic Diagram of Felling Pattern * Direction To Landing Tree Bunch Trail Cut Order in Which Feller-buncher Trail Unit Boundary KEY 27 They would operate on every second or third feller buncher trail, depending upon the number of tree bunches. Where the stand volume was high, every other feller-buncher trail was used. In more open areas, every third trail was used for skidding. taken. Generally, one bunch at a time would be When skidding a portion of the unit, one trail would be their "main" trail over which all bunches in that portion of the unit would pass over. They would take all bunches along the "main" trail and then back off this trail over one or two other feller-buncher trails to get those bunches. As much as possible, they would attempt to follow the same path each time they backed off the "main" trail. Thus, the skidder operators would be traveling over every second or third feller buncher trail a number of times, making it a skid trail. Occasionally two small bunches would be combined into one large load. To combine bunches, the operator would pull the first bunch along the "main" trail and then back up and pick up the second bunch. Bunches would then be combined along the "main" trail and taken to the landing. Figure 3.4 is a schematic diagram of the skidding pattern. 3.6 Data Collection The intent of this study is to assess the impacts caused by a mechanized harvest operation. For this reason, exact number of passes for each machine at each sample point E - D Direction To Landing Loaded Skidder Path Skid Trail Trail Feiler-buncher Tree Bunch KEY Figure 3.4 Schematic Diagram of Skidding Pattern: Skidders back up to tree bunch and pull it out over same path NL' /0 Nfr 29 were not recorded. Rather, groupings of number of passes for each machine (which reflected normal operating procedure) were made and the area in each of these groupings was sampled. The groupings were based upon observations made during the harvesting of unit 1A and the unit harvested prior to unit 1A. Both units were harvested by the same operators with the same equipment. Work began in unit 1A as soon as the previous unit was logged. For this study, one pass is defined as the passing of the machine, either loaded or unloaded, over a specific point. 3.6.1 Machine Pass Groupinqs Feller-buncher, 1 - 5 passes. This was for cases where the machine basically progressed forward in cutting it's strip. Occasionally it would have to back up to reposition itself or to move around clumps of regeneration, leave trees or other obstacles. Feller-buncher, 6 - 10 passes. This occurred in areas of dense timber with a large amount of down material. The operator would have to reposition the machine many times in order to make bunches. Feller-bunchers were not observed to have more than 10 passes over any area. Skidder, 1 - 4 passes. This occurred in areas where the skidder backed off the trail across one feller-buncher trail to pick up tree bunches. Very little repositioning 30 by the operator was required to position the tree bunch in the grapples. No combining of bunches was done. Skidder, 5 - 8 passes. This occurred when the operator backed off and across two feller-buncher trails, when bunches needed to be combined, or when the operator had to do considerable maneuvering in order to correctly position the bunch in the grapples. In the field, there was no way to visually distinguish between the areas having 1 - 4 and 5 - 8 skidder passes. The soil surface appeared the same in regards to slash depth, tire tracks and impact on existing vegetation. Skidder, 50+ passes. This was for major skid trails where the skidder passed over at least 50 times. Figure 3.5 shows the relationship between the machine pass groupings. There were also short secondary skid trails and portions of main skid trails where the number of skidder passes was between 8 and 50. The length of these short, secondary trails was generally less than 50 feet. The number of these secondary trails varied according to the shape of the unit; the more irregular the unit shape, the more secondary trails there would be. The length of trail at the ends of the main skid trails having 8 - 50 skidder passes varied from approximately 100 feet in dense timber to over 200 feet in sparse timber. Both of these areas were not sampled, as an accurate count on the number of machine Direction To Landing Tree Bunch Location Skidder - 1 to 4 Passses Skidder - 5 to 8 Passes 1 -Sand6-lOPasses Feller-buncher Trail Skid Trail - 50+ Passes Figure 3.5 Schematic Diagram Showing Relationship Between Machine Pass Categories h KEY 32 passes could not be made. Figure 3.6 shows a representation of the layout of skid trails in relation to each other. Soil density measurements were taken before logging, after the feller-buncher, and after skidding, in order to access the changes in bulk density caused by logging. Stratified sampling was used, meaning that p'ost activity measurements would only be taken where the machine had passed. For the feller-buncher, 1 - 4, and 5 - 8 skidder passes, sample points would be taken in the machine track or tire track, as it was felt that this would be the area of maximum impact. For the main skid trails (50+ skidder passes), sample points would be located across the width of skid trail, as it would not be possible to determine the exact path of tire travel in these areas. Slash depth measurements would also be taken at each post-activity measurement point to determine if slash had any effect on reducing soil compaction caused by logging equipment. Total slash plus litter depth would be measured at each point by laying a shovel across the slash and measuring the perpendicular distance from the shovel handle to bare mineral soil. 3.6.2 Sample Point Preparation For all sample point locations (before and after logging) a square area approximately 1.5 feet by 1.5 feet was cleared to bare mineral soil. A spot slightly larger Figure 36 Schematk Diagram Showing Relationship Between Skid Trail Types: Shows Skid Trails Sampled and Not Sampled. Feller-buncher Trails Not Shown Sampled 50+ SkIdder Passes Main Skid Trails Not Sampled 8 -50 SkIdder Passes Secondary Skid Trails Sampled 1 -8 SkIdder Passes Ends of Skid Trails Unit Boundary KEY 34 than the probe base was carefully smoothed and a probe access hole driven to a depth of 12 inches. A minimum of three, 15 second counts were taken at each depth (4, 8, and 12 inches). At positions where the readings varied by more than 5%, one or two additional readings were taken. The maximum number of possible readings for any depth was 5. The mean of the counts was determined for each depth at each point. After all measurements were taken, soil samples were obtained by using the soil corer. the 0 to 6 and 6 to 12 inch depths. Samples were taken from The samples were sealed in plastic bags and marked for later soil moisture content determinat ion. 3.6.3 Before Loqqinc Prior to logging, an "average" bulk density for the unit was obtained. This was done to have a "base-line" or before treatment bulk density for comparison purposes. The points were selected using a grid placed over a map of the area, with 30 sample locations selected using a random number table. These 30 points were then located in the field using a hand compass and pacing distances. This number of sample points equates to an approximate sampling intensity yielding 90% confidence that the average bulk density obtained is within density for the stand. .1 gm/cc of the true mean bulk Ninety-five percent confidence 35 intervals for the before logging mean bulk densities measured are: 0 - 4 inch layer, .672 to .726 gm/cc; 0 - 8 inch layer, .669 to .711 gin/cc; 0 - 12 inch layer, .626 to .676 gm/cc (Appendix B). After Fellinq and Bunchinq 3.6.4 The two feller-buncher machines were operating at opposite ends of the unit. One machine was selected and the number of passes it made along the trail was recorded. When a trail had been cut out, and before it was covered with trees from the next trail, sample points were randomly located in the track marks and the required data were obtained. After the trail was sampled, the second feller- buncher was observed and sampled using the same process. Data collection for the feller-bunchers ended when approximately 90% of the unit was felled and bunched. This was necessary because skidding in the unit had already begun. Figure 3.7 is a map showing the area where feller- buncher data were not collected. The location of each bunch of trees was marked by spraying paint on a nearby stump, rock or leave tree. This was done so the number of passes for major skid trails could be estimated by counting the number of paint marks which led into the trail. 0 Feet 400 ® No Feller-buncher Data No Skidder Data - Access Road O Landing Unft Boundary Figure 3.7 Map of Unit Showing Areas Not Sampled By Machine Type 8D0 Approximate Scala KEY 37 After Skiddinq 3.6.5 Due to the fact that when skidding began, data were still being collected for the feller-bunchers, skidding data were only taken on the northern one-half of the unit. Figure 3.7 shows the area where skidder data were not collected. skidder. Most sample points were taken behind the 648D Density measurements were made once the number of passes was determined over a given point in a tire track. For major skid trails, the number of bunches passing over the trail was estimated by counting paint marks Sand determining the location where at least 50 passes had occurred. Sample points were then randomly located from that point up to the landing. These sample points were located along the entire width of the trail and not just in the tracks. 3.6.6 Amount of Area Impacted Total area in machine trails was desired, including the area beneath the machine between tracks or tires. This could be measured for skidders by recording the area in tire tracks and area showing evidence of trees being dragged along the surface. For feller-bunchers, the area beneath the machine between tracks was not touched and thus showed no signs of disturbance. Rather than try to "estimate" this area beneath the feller-bunchers, only area in feller- 38 buncher tracks was recorded. Upon completion of skidding, transects were run to determine the amount of area impacted by each machine. total of 15 transects were taken. feet in length. A Each transect was 100 The starting point and compass direction for each transect was randomly located throughout the unit. For each transect, a 100 foot tape was stretched and the number of feet in each of the five categories below was recorded. No disturbance - No visible signs of disturbance. Feller-buncher track - Imprints of feller-buncher tracks were visible. Skid track - Imprints of skidder tires were visible. Tree bunch sweep - No evidence of track or tire imprints but evidence of trees being dragged over the ground. Evidence looked for was pieces of slash on the ground, broken seedlings or shrubs or the appearance of the ground being "swept over" by tree tops. Major skid trail - Obvious major skid trail, as though a machine had cleared the trail down to bare soil. 39 3.7 Laboratory Analysis 3.7.1 Soil Moisture Deterniination Each soil saniple was weighed and then dried in an oven at 105 degrees Celsius for 24 hours. The samples were weighed again and the percent of moisture on a dry weight basis was calculated using the following formula: MOISTURE (WET SOIL WEIGHT - DRY SOIL WEIGHT) CONTENT (%) = (DRY SOIL WEIGHT) X 100 Calibration of Probe 3.7.2 The probe was calibrated for this soil type by using soil brought back from the unit harvested. This was done by packing a known weight of soil into a calibration box of known volume. Bulk density was calculated by dividing the weight of soil by it's volume. Bulk densities of .599, .9, 1.0, and 1.2 gm/cc were produced. .8, These compared to the range of bulk densities encountered in the field during this study. The number of counts in 15 seconds for 4, 8, and 12 inch depths were recorded in the same manner as the field counts. Regression analysis was then used to develop an equation for estimating bulk density based upon the average probe count for each depth. R-squared for these 3 equations was .98 or greater. The equations derived were used to estimate wet bulk density. Wet densities were converted to a dry weight basis 40 for the moisture content for each point and depth using the following formula: WET BULK DENSITY DRY BULK DENSITY = (MOISTURE CONTENT + 1.0) (as a decimal) 41 4.0 RESULTS Data were originally separated by model of fellerbuncher and skidder. Since so few points were sampled for the John Deere 640 skidder (7 out of the 60 total taken for the skidding operation), results for both skidders were combined. Results for both models of feller-bunchers were also combined. These machines had identical dimensions, weights, and operating pattern. A comparison of average bulk densities for both feller-bunchers showed no statistical difference for all 3 soil layers (alpha = 0.01). Results for both feller-bunchers, 1 - 5 passes (62 sample points) and 6 - 10 passes (12 sample points), were also combined. There was no statistical difference between the average bulk densities for all 3 soil layers (alpha = 0.01). 4.1 Number of Sample Points Taken During the course of this study, a total of 165 points were sampled. With 3 measurements taken at each point (4, 8 and 12 inch depths), a total of 495 bulk density measurements were taken. Of this total, 21 were discarded from the data set for various reasons such as organic matter in the soil sample resulting in high soil moisture content, lost soil samples so that soil moisture content could not be determined, extremely rocky soils, and atypical sample point location which did not represent the conditions intended to be studied. Results presented here are based upon the 42 remaining 474 measurements (Table 4.1). Table 4.1 - Number of Measurements used for Results No. Discarded Original no. No. Used for Taken Results Layer (in.) Layer (in.) Layer (in.) 0-4 0-8 0-12 0-4 0-8 0-12 Cateqory 0-4 0-8 0-12 Before 27 30 30 30 3 2 2 28 28 Logging After FellerBuncher 75 75 75 1 1 1 74 74 74 After 26 Skidder 1-4 passes 26 26 1 0 0 25 26 26 14 After Skidder 5-8 passes 14 14 3 2 2 11 12 12 After 20 Skidder 50+ passes 20 20 1 1 1 19 19 19 4.2 Bulk Density Measurements Results presented are average dry bulk densities for each of the 5 categories (before logging, after the fellerbuncher, after 1 - 4 skidder passes, after 5 - 8 skidder passes and after 50+ skidder passes) for all 3 soil layers studied (0 - 4 inch layer, 0 - 8 inch layer and 0 - 12 inch layer). Pooled t tests are used to determine if differences in soil density are significant. It needs to be kept in mind that the bulk density measurements reported are average densities for the 3 soil layers and not point measurements taken at specific depths. Table 4.2 presents bulk density 43 measurements for each of the 5 categories sampled. Table 4.2 AVERAGE SOIL DENSITIES FOR EACH GROUPING SAMPLED SOIL LAYER NO. OF SAMPLES (INCHES) AVERAGE BULK DENSITY (GM/CC) BEFORE LOGGING 0-4 0-8 0-12 27 28 28 0.699 (0.065) 0.690 (0.055) 0.651 (0.063) 74 74 74 0.717 (0.064) 0.750 (0.057) 0.708 (0.062) AFTER FELLER--BUNCHER 0-4 0-8 0-12 AFTER SKIDDER, 1 - 4 PASSES 0-4 0-8 0-12 25 26 26 0.694 (0.074) 0.735 (0.075) 0.712 (0.078) AFTER SKIDDER, 5 - 8 PASSES 0-4 0-8 0-12 12 12 0.796 (0.064) 0.799 (0.057) 0.780 (0.081) 19 19 19 0.874 (0.058) 0.854 (0.039) 0.887 (0.108) 11 AFTER SKIDDER, 50+ PASSES 0-4 0-8 0-12 (standard deviation) 4.2.1 Comparisons with Before Loqqinq Density Measurements Table 4.3 compares bulk densities for the different conditions sampled with before logging measurements for all 3 soil layers. Figure 4.1 displays the average soil layer CD 0 o a) C 0.6 0.65 0.7 0.75 0.8 0.85 0.9 0.95 Soil Layer Depth (inches) 0-8 0-12 Figure 4.1 Graph of Average Soil Density by Condition 0-4 H H H I Skidder 50+ Passes Skidder 5-6 Passes Skidder 1 -4 Passes Feller-Buncher Before Logging KEY 45 density for each of the 5 categories for all 3 soil layers. The feller-buncher caused a significant increase in soil density for the 0 - 8 and 0 - 12 inch layers (alpha = The fact that the 0 - 4 inch layer did not show a 0.01). significant increase in density may be explained by the churning action of the grousers on the feller-buncher This churning may be mixing the top 1 to 2 inches tracks. of soil, thus breaking up the compaction caused by the machine. Churning will greatly affect the density of the top 4 inch layer of soil because 25 to 50% of this layer will be affected by the grousers. On the other hand, churning will have very little effect on the density of the 0 - 8 and 0 - 12 inch layers because the top 1 - 2 inches make up a smaller proportion of these soil layers. One to 4 skidder passes also caused significant increases in soil density for the 0 - 8 and 0 - 12 inch layers (alpha = 0.01). The fact that the 0 - 4 inch layer did not show a significant increase in density may be explained by a churning action similar to that described above for feller-bunchers. Here, instead of grousers, the tire tread and sweeping action of trees being dragged along the surface may be mixing the top few inches of soil, thus breaking up the effects of compaction caused by the machine. As previously discussed, this mixing of the top few inches will greatly affect density of the top 4 inches of soil and have little effect on the 0 - 8 and 0 - 12 inch layers. 46 Five to 8 and 50+ skidder passes caused a significant increase in soil density in all 3 layers (alpha = 0.01). Apparently the surface is compacted enough that the mixing due to tire tread and dragging tree bunches is no longer occurring. I Table 4.3 SOIL DENSITY COMPARISONS WITH BEFORE LOGGING VALUES SOIL LAYER (INCHES) NET CHANGE FROM BEFORE LOGGING PERCENT CHANGE FROM BEFORE LOGGING SIGNIF. (GM/CC) (¼) LEVEL +2.56 +8.73 +8.81 0.11 >0.01 >0.01 -0.69 +6.60 +9.34 0.40 >0.01 >0.01 +13.95 +15.84 +19.83 >0.01 >0.01 >0.01 +25.00 +23.81 +36.30 >0.01 >0.01 >0.01 AFTER FELLER-BUNCHER +0.018 +0.060 +0.057 0-4 0-8 0-12 AFTER SKIDDER, 0-4 0-8 0-12 1 - 4 PASSES -0.005 +0.045 +0.061 AFTER SKIDDER, 5 - 8 PASSES 0-4 0-8 0-12 +0.097 +0.109 +0.129 AFTER SKIDDING, 50+ PASSES 0-4 0-8 0-12 4.2.2 +0.175 +0.164 +0.236 Chanqes in Bulk Density With Increasinq Depth Table 4.4 compares soil densities for each of the 3 soil layers within each of the 5 categories to see how soil 47 density changed with depth. Before logging, the soil had a fairly uniform bulk density throughout. There is no statistical difference in soil density between the 0 -4 and 0 - 8 inch layers (alpha 0.1). While the decrease in density between the 0 - 8 and TABLE 4.4 CHANGES IN SOIL DENSITY WITH INCREASING DEPTH SOIL LAYER (INCHES) BULK DENSITY CHANGE (GM/CC) PERCENT CHANGE (%) SIGNIF. LEVEL BEFORE LOGGING 0-4 0-8 -0. 0091 0 - 12 _0.0392 - 1.291 - 5.652 0.29 >0.01 + 4.601 - 5.602 >0.01 >0.01 + 5.901 - 3.132 0.10 0.21 + 0.381 - 2.382 0.46 0.25 + 2.291 + 3.862 0.27 0.16 AFTER FELLER-BUNCHER 0-4 0-8 +0. 0331 0 - 12 _0.0422 AFTER SKIDDER, 0-4 0-8 0 - 12 1 - 4 PASSES +0.0411 _0.0232 AFTER SKIDDER, 5 - 8 PASSES 0-4 0-8 +0. 0031 0 - 12 _0.0192 AFTER SKIDDER, 50+ PASSES 0-4 0-8 -0. 0201 0 - 12 +0.0332 NOTES 1 Change between 0 - 4 and 0 - 8 inch layers 2 Change between 0 - 8 and 0 - 12 inch layers 48 o - 12 inch layers is statistically significant (alpha = 0.01), it is only a difference of 0.04 gm/cc. For all practical purposes, densities in all 3 of these layers can be considered the same. The increase in soil density between the 0 - 4 and 0 8 inch layers after the feller-buncher is significant (alpha = 0.01) The decrease in soil density between the 0 - 8 and 0 - 12 layers is also significant (alpha = 0.01). 4, After 1 - 5 - 8 and 50+ skidder passes, soil densities are constant throughout their depths. There is no significant change in soil density between the 0 - 4, 0 - 8 and 0 - 12 inch soil layers within each of these 3 machine pass categories (alpha = 0.1). 4.2.3 Chanqes in Bulk Density With Increasinq No. of Passes Table 4.5 compares bulk density changes for each machine pass grouping to see if bulk density increased with increasing number of passes. The amount of compaction caused by the feller-buncher and 1 - 4 skidder passes is the same. For all 3 soil layers, there is not a significant difference in bulk densities between "after the feller-buncher" and after 1 - 4 skidder passes (alpha = 0.05). There is a significant increase in soil density between 1 - 4 and 5 - 8 skidder passes (alpha = 0.01) for all 3 soil layers. There is also a significant increase in soil density between 5 - 8 and 50+ 49 skidder passes (alpha = 0.01) for all 3 soil layers. This indicates that bulk density does increase with increasing number of passes, up to 50. TABLE 4.5 SOIL LAYER (INCHES) CHANGES IN BULK DENSITY WITH INCREASING NUMBER OF PASSES BULK DENSITY CHANGE (GM/CC) PERCENT CHANGE (%) SIGNIF. LEVEL - 4 SKIDDER PASSES VRS. AFTER FELLER-BUNCHER1 1 0-4 0-8 0 - 12 -0.023 - 3.17 - 1.96 + 0.48 -0. 015 +0. 004 5 - 8 SKIDDER PASSES VRS. 0-4 0-8 0 - 12 +0. 102 +0. 064 +0. 068 1 0.07 0.13 0.42 - 4 SKIDDER PASSES2 +14. 74 + 8.67 + 9.60 >0.01 >0.01 >0.01 50+ SKIDDER PASSES VRS. 5 - 8 SKIDDER PASSES3 0-4 0-8 0 - 12 +0. 078 +0.055 +0.107 + 9.70 + 6.89 +13. 74 >0.01 >0.01 >0.01 NOTES 1 1-4 skidder pass density minus feller-buncher density 2 5-8 skidder pass density minus 1-4 skidder pass density 50+ skidder pass density minus 5-8 skidder pass density 4.3 Slash Depth Slash depth was recorded after the machine had passed over the sample point. There is the possibility that slash may have been deposited or removed while the machine passed over the point and thus the slash measured was not the actual depth at the time the machine went over the point. Since the feller buncher did not drag trees along the ground 50 but stacked them in bunches, the chance of slash being removed or deposited was slight. During the process of skidding, trees were dragged along the ground and the chance of slash being swept away by tree tops or deposited due to breakage was greater. However, since the exact sample point location could not be determined until after the skidder had passed, it was felt that the measurement procedure followed would be the most consistent one. Table 4.6 shows the average slash depth and range of slash depths encountered at each sample point location. There was very little duff present, generally less than one inch. The majority of material on the ground was down trees and broken limbs, especially considering the large amount of dead material found in this unit. The most noticeable slash depth is the 11 inches after 50+ skidder passes. This sample point was located just beyond a half-buried log where a depression had formed and filled in with slash. Since it could not be determined if the depression had formed during skidding or not, it was assumed it had been there prior to logging and the sample point was retained. Bulk density measurements for this point were not abnormally high or low. In addition, there does appear to be a trend of decreasing slash depth and more area swept clear of slash as the number of passes increased. There is a significant decrease in slash depth between 1 - 4 and 50+ skidder passes (alpha = 0.01). There is no significant difference in slash depth 51 between 1 - 4 and 5 - 8 skidder passes or between 5 - 8 and 50+ skidder passes (alpha = 0.1). Differences in slash depth between "after the feller-buncher" and all 3 skidder pass groups in Table 4.6 are not significant (alpha = 0.1). TABLE 4.6 SLASH DEPTH CATEGORY AFTER FELLER BUNCHER AVERAGE SLASH DEPTH (IN.) DEEPEST SLASH DEPTH (IN.) NO. OF POINTS NO SLASH PERCENT POINTS NO SLASH 2.4 (2.2) 13 11 15 AFTER SKIDDER 1 - 4 PASSES 4.4 (3.0) 10 4 15 7 2 17 11 14 74 (N = 74) (N = 26) AFTER SKIDDER 5 - 8 PASSES 2.9 (1.8) (N = 12) AFTER SKIDDER 50+ PASSES 1.6 (3.2) (N = 19) (standard deviation) 4.4 Area Impacted Table 4.7 displays the results of the 15 transects run to determine amount of area impacted by each machine. The percent of area impacted was calculated by dividing the number of feet recorded in each category by the total distance measured (1500 feet). in the transects. There was much variability All of the 5 impact categories were not found on every transect. This may explain the rather large 52 TABLE 4.7 AREA IMPACTED BY HARVESTING ACTIVITIES TOTAL DISTANCE RECORDED (FT.) CATEGORY PERCENT OF TOTAL DISTANCE NO DISTURBANCE 682 45.5 (18.8) FELLER-BUNCHER TRACK 100 6.7 (8.8) SKID TRACK 137 9.1 (7.0) TREE BUNCH SWEEP 395 26.3 (15.2) MAIN SKID TRAIL 186 12.4 (14.8) (standard deviation) standard deviations. For purposes of discussion and management implications, a determination of how these 5 categories of area impacted correspond to the 5 groupings of bulk density measurements taken was made as follows: The feller-buncher track area corresponds to the area impacted by feller-bunchers. The percent of area in feller- buncher track seems low considering that the feller-buncher trails were approximately 40 feet apart. However, due to the nature of the skidding operation, all main skid trails were first feller-buncher trails. Using this reasoning, 19.1% of the area would have been impacted by the fellerbunchers. In addition, only area with evidence of actual feller-buncher track imprint was recorded. The area underneath the machine, between the tracks, would also be considered part of the feller-buncher trail but would show no sign of disturbance. In order to accurately measure area 53 impacted by feller-bunchers, transects would have had to been run after felling and before skidding. This was not possible in this study because, for a time, skidding and felling were going on at the same time. Also, after approximately 60% of the unit had been felled, and before skidding began, there was a brief rainstorm (amount of precipitation received was unknown) which may have washed out some feller-buncher tracks. For these reasons, it is likely that the area measured as being impacted by the feller-buncher in this study may be lower than the actual amount. There was no visual way to determine how many skidder passes the skid track and tree bunch sweep categories correspond to. Some of this area had 1 - 4 skidder passes and the rest would have had 5 - 8 passes. During the skidding phase, it was observed that approximately twice as much area had 1 - 4 skidder passes than did 5 - 8 skidder passes. Thus, 65% of the total skid track and tree bunch sweep areas had 1 - 4 skidder passes and 35% of this area had 5 - 8 skidder passes. The majority of the area with 1 - 4 and 5 - 8 skidder passes were not impacted by feller-bunchers. These areas were mainly found between feller-buncher trails as the skidder backed off trails to pick up bunches. It is recognized that when the skidder backed across a fellerbuncher trail to pick up bunches, there would be small areas 54 impacted by both feller-buncher and skidder, where the trails intersected (Fig. 3.5). Also, at the ends of feller-buncher trails there would be areas impacted by both machines. This area would be the length of trail in front of the last few tree bunches in a trail. Both of these areas were included in the areas sampled for the 1 - 4 and 5 - 8 skidder pass categories. Not enough sample points were located in these areas impacted by both the feller-buncher and skidder to analyze as separate populations. The main skid trails were impacted by both fellerbunchers and skidders. This area is composed of main skid trails having 50+ passes, which was sampled, and portions of skid trails having less than 50 passes but having enough passes to be easily distinguished as a skid trail (predominately bare ground). Observations made of skidding indicates that this happened after 10 - 15 passes. At the time the transects were run, it was not possible to determine if the skid trail encountered had 50+ or less than 50 passes. However, since a 19.8% increase in density was found in the 0 - 12 inch layer after 5 - 8 skidder passes (Table 4.2), it is likely that the portions of the skid trails with less than 50 passes had at least this much of an increase in bulk density, and probably more. These areas would have had more than 8 skidder passes and the data indicate that soil density increases with greater number of passes above 8. 55 Table 4.8 shows the relationship between area impact categories and bulk density measurement groupings. These values will be used for discussion and management implication purposes. TABLE 4.8 RELATIONSHIP BETWEEN AREA IMPACTED AND DENSITY MEASUREMENTS BULK DENSITY GROUPING AREA IMPACTED CATEGORY PERCENT OF AREA IMPACTED Before Logging No disturbance 45.5 (18.8) After fellerbuncher Fel ler-buncher 6.7 (8.8) After skidding 1 - 4 passes Skid track and Tree bunch sweep 23.0 (11.8) After skidding 5 - 8 passes Skid track and Tree bunch sweep 12.4 (6.4) After skidding 50+ passes Main skid trail 12.4 (14.8) track (standard deviation) Figure 4.2 is a summary chart showing the percent of the unit in each of the 5 impact categories and the average percent increase in soil density (compared to before logging) for the 0 - 12 inch soil layer. 4.5 Reqression Analysis Regression analysis was used to determine how much of the variability in dry bulk density was explained by the variables measured (soil moisture content, slash depth and category of machine passes). Since groupings of passes and 12.4% of Area 50+ Skidder passes and feller-buncher Distibutlon of Area Impacted 5-8 SkIdder passes 12.4% of Area V 36.3% Increase 19.8% Increase 9.3% Increase 8.8% Increase No Increase Bulk Density Increase (%) 0 - 12 Inch Soil Layer Figure 4.2 Percent of Area Impacted and Percent Increase In Soil Density for the 0 - 12 Inch Layer for Unit 1A, After Logging 23% of Area 1-4 Skidder pses Feller-buncher only 6.7% of Area No impact - 45.5% of Area 57 not individual numbers were recorded, 0, 1 indicator variables for each grouping of passes (after feller-buncher, 1 - 4 skidder passes, 5 - 8 skidder passes and 50+ skidder passes) were used. It was found that for all 3 soil layers sampled, soil moisture content did not significantly contribute to the regression (alpha = 0.1). Only 1% of the variability in dry bulk density measurements was explained by this variable. This lack of effect of soil moisture is probably due to the fact that soil moisture was fairly constant throughout the course of this study and within the top 12 inches of soil. Had there been a wide variation in moisture content over time or within the top 12 inches of soil, it would be expected that moisture content would be a significant factor in explaining variations in bulk density. Also, this stand was logged when soils were dry (generally less than 25% moisture content). Had logging been done when the soils were wetter, it is possible that increases in soil density would have been higher and soil moisture content may have then been a significant factor in the regression. By far, machine pass category explained the greatest amount of variability. For the 0 - 4, 0 - 8 and 0 - 12 inch layers, machine pass category explained 42%, 38% and 45%, respectively, of the variability in dry bulk density. 4.9 displays the regression coefficients for the most Table 58 TABLE 4.9 REGRESSION ANALYSIS RESULTS VARIABLE* COEFFICIENT STANDARD ERROR SIGNIF. LEVEL 0 - 4 INCH SOIL LAYER Constant Slash Depth(inches) FB Skidl Skid2 Skid3 0.702 -0.010 0.039 0.034 0.115 0.188 0.012 0.002 0.014 0.019 0.021 0.018 >0.01 >0.01 >0.01 0.07 >0.01 >0.01 0.689 -0.010 0.085 0.091 0.139 0.181 0.010 0.001 0.013 0.017 0.019 0.016 >0.01 >0.01 >0.01 >0.01 >0.01 >0.01 0.651 -0.015 0.093 0.125 0.172 0.260 0.012 0.002 0.016 0.021 0.024 0.020 >0.01 >0.01 >0.01 >0.01 >0.01 >0.01 R - Squared = .50 0 - 8 INCH SOIL LAYER Constant Slash Depth(inches) FB Skidl Skid2 Skid3 R - Squared = .47 0 - 12 INCH SOIL LAYER Constant Slash Depth(inches) FB Skidl Skid2 Skid3 R - Squared = .56 * NOTES Slash Depth has no value for before logging if for feller-buncher, 0 otherwise Skidl = 1 if for 1 - 4 skidder passes, 0 otherwise Skid2 = 1 if for 5 - 8 skidder passes, 0 otherwise Skid3 = 1 if for 50+ skidder passes, 0 otherwise FB = 1 meaningful regression, the regression using slash depth and machine pass category. density in gm/cc. The regression predicts dry bulk This is the dependent variable. The independent variables are slash depth (in inches) and machine pass grouping (either 0 or 1). Thus, the 0 - 12 59 inch regression equation is: Dry bulk density (gm/cc) = + + + 0.651 - (0.015 * slash depth) (0.093 * FB) + (0.125 * Skidl) (0.172 * Skid2) (0.260 * Skid3) Slash depth also significantly contributed to this regression (alpha = 0.01) for all 3 soil layers. A negative relationship between dry bulk density and slash depth was found. As slash depth increased, predicted dry bulk density decreased within the same category of machine passes. For the 0 - 4 inch layer, slash depth accounted for 8% of the variability found in dry bulk density measurements. For the 0 - 8 and 0 - 12 inch layers, the amount of explained variability was 9% and 11%, respectively. 4.6 Soil Moisture Analysis Table 4.10 shows the results of soil moisture determination done in the laboratory. The one rainstorm which occurred during the felling added a small amount of moisture to the surface layer. The increase in soil moisture between 1) after skidding, 2) before logging, and 3) after feller-buncher, for the 0 - 4 and 0 - 8 inch soil layers is significant (alpha = 0.05). Though this difference is statistically significant, it is less than a 5% increase and probably had little bearing on results. Differences in soil moisture for the 0 - 12 inch layer for 60 all 3 categories are not significant (alpha = 0.1). TABLE 4.10 SOIL MOISTURE RESULTS CATEGORY 0 - 4 SOIL LAYER (INCHES) 0 - 12 0 - 8 BEFORE LOGGING 22.6* (8.8) 23.1* (8.4) 23.7* (8.1) AFTER FELLERBUNCHER 22.6* (10.8) 22.8* (10.3) 22.9* (94) AFTER SKIDDING ALL PASSES 27.3** (7.2) 26.7** (6.9) 25.4* (6.9) (standard deviation) NOTES * No statistically significant difference at alpha = 0.10 ** Statistically significant difference at alpha = 0.05 61 5.0 DISCUSSION OF RESULTS Some comparisons of trends in results between this study and others can be done. Previous studies found that soil densities approached their maximum after 4 to 6 machine passes and changed little with a greater number of passes. For the feller-buncher, this appears to be the case. Considering how the feller-buncher operated, it is also possible that most of the compaction occurred in the first 1 or 2 passes. As previously explained, the machine basically sits in one place or moves very slightly. It rocks back and forth as it swings around picking up material to make bunchers. Much dynamic loading occurs under the tracks along with vibration caused by the motor and sudden movement. Thus it seems possible that 1 or 2 passes would be enough to cause most of the compaction that would occur. As there was no significant increase in soil density between 1 - 5 and 6 - 10 passes (section 4.0), the maximum increase in density may have occurred within the first 5 passes and changed very little with a greater number of passes. The number of passes required for the soil density to reach its maximum under the skidder paths cannot be determined in this study. However, it is not likely that it took 50 skidder passes to compact the soil to near its maximum. It is more likely that near maximum density was reached between 8 and 50 skidder passes and additional passes caused little additional change in soil density. 62 There was a significant increase in soil density between 1 - 4 and 5 - 8 skidder passes and between 5 - 8 and 50+ skidder passes. This compares well with Figure 2.1 developed from Froehlich et al. (1980). Burger, et al. (1985) and Lenhard (1986) did their studies with empty skidders driving in straight lines where both tires went over the exact same point. Dynamic loading was probably not changing much in their studies as opposed to this study where skidder speeds and loading changed with each pass. Dynamic loading was probably much higher in this study. Soil density increases with number of passes may also be influenced by slash. The regression analysis (section 4.6) found that slash significantly reduced compaction produced by the logging machinery. Slash may tend to reduce the amount of compaction that can occur by spreading the load beyond the surface area under the wheels of the skidder, with larger pieces of slash better able to distribute the load than smaller pieces. Thus, down logs and tree limbs are more effective than duff or leaf litter in reducing compaction, and more passes over slash would be required to cause the same changes in density than over bare soil, litter and duff layers. Hatchell, Ralston and Foil (1970) did not mention slash depth in their study. Burger, et al. (1985) reported an average litter layer of 1.3 inches, considerably less than the litter cover found in this study. Froehlich (1978) and Lenhard (1986) reported 63 duff and litter depths in the ranges of those found here but as discussed above, the larger pieces of slash found here would be more effective in spreading out the compactive force. Soil moisture content may also help explain the results found in this study. Generally, a drier soil requires more force to compact it to a particular density than does the same soil at a higher soil moisture content. Soil moisture contents in this study were approximately 25% (table 4.7). Hatchell, Ralston and Foil (1970) classified their soil moisture content as high for the Atlantic Coast Plain, but provided no figures. Considering that this study occurred in August, which is in the dry season in eastern Oregon, it is quite likely that the soil moistures found in this study were lower than those found by the Hatchell et al. study. (1970) Froehlich's (1978) Mt. Hood and Jmpqua sites also had soil moisture contents higher than the soil moisture contents found in this study. The literature review also found that effects of compaction decreased with increasing soil depth, with very little change in soil density at depths of 12 inches. Data in tables 4.2 and 4.4 support this trend. For the feller-buncher, soil density significantly increased from the 0 - 4 to the 0 - 8 inch layers and significantly decreased from the 0 - 8 to the 0 - 12 inch layers. This indicates that soil density reached it's 64 maximum in the vicinity of 8 inches and decreased with greater depths. It is likely that the density at a depth of 12 inches was changed very little. This seems possible when one considers that in order for the 0 -12 inch layer to have a lower average density than the 0 - 8 inch layer, the average density in the 8 - 12 inch zone would have to be lower than the 0 - 8 inch layer in order to reduce the average density of the 0 - 12 inch layer, since both measurements passed through the same 0 - 8 inches of soil. This same trend of decreasing compactive effects with increasing depth is also the case for 1 - 4 and 5 - 8 skidder passes. For both of these cases, the highest average densities were found in the 0 - 8 inch layer. the differences in densities between the inch layers were not significant, th While 0 - 8 and 0 - 12 fact that the 0 - 12 inch layer had a lower average density indicates that the densities in the 8 - 12 inch zone again had to be lower than the average density for the layer above, indicating that the maximum density would be found in the vicinity of 8 inches in depth. For 50+ skidder passes, this trend is not apparent. Density decreases between the 0 - 4 and 0 - 8 inch layers and then increases between the 0 - 8 and 0 - 12 inch layers. It is not known what processes are occurring to cause the unusual readings for the 0 - 8 inch layer. Perhaps there is a root zone in the 4 - 8 inch zone which acts as a "sponge." 65 When the skidder passes over, this zone absorbs very little compactive force but is merely compressed and transmits the force to the soil below. 66 6.0 MANAGEMENT IMPLICATIONS AND RECOMMENDATIONS A.s presented in the Introduction (Section 1.2) detrimental soil compaction for volcanic ash/pumice soils is defined as an increase in soil bulk density of 20% or more over undisturbed levels (USDA Forest Service, 1987). This definition applies to any soil layer at least 4 inches thick between the depths of 4 to 12 inches (personal conversation with Bob Meurisse, Regional Soil Scientist, USDA Forest Service, Region 6). From table 4.2 it can be seen that the feller-buncher and 1 - 4 skidder passes have not caused the soil density to cross this threshold. The increases in bulk density for these 2 cases are 8.8% and 9.3%, respectively for the 0 - 12 inch soil layer. Conversely, 50+ skidder passes did cause detrimental compaction. The increase in soil density was 36.3% for the 0 -12 inch layer. The situation is not as clear for 5 - 8 skidder passes. Here, the increase in bulk density was 19.8%, not quite 20% but very close. Since the increase in density for 5 - 8 skidder passes is so close to the unacceptable level, for purposes of discussion here it will be taken as having a 20% increase in density and thus detrimentally compacted. This seems likely when one considers that all it would take for the increase in density to be 20% instead of 19.8% would be for the average density of the 0 - 12 inch layer for 5 - 8 skidder passes to be 0.001 gm/cc higher than that reported. This is a very small change when considering all of the 67 variables involved. Thus, 24.8% of the total area suffered detrimental compaction. As previously stated, areas receiving 5 - 8 skidder passes were predominately located between feller-buncher trails where the skidder had backed across 2 feller-buncher trails (Fig. 3.5). This area shows little evidence of impact on the surface and is almost identical in appearance to the 1 - 4 skidder pass areas. Thus, it would be extremely difficult, if not impossible, to identify this 5 8 skidder pass area in the field. For this reason, it is recommended that the skidders only be allowed to back across 1 feller-buncher trail thus confining the area suffering detrimental compaction to the easily identifiable skid trails. area being in skid trails. This will result in more It is estimated that the area in skid trails would increase from approximately 12% to approximately 18%. One benefit would be that the total area detrimentally compacted would be reduced from almost 25% to approximately 18%. Another benefit would be that most of the area suffering unacceptable increases in bulk density would be easily identifiable in the field when mitigation measures, as required by the timber sale contract, would be done. 68 7.0 FURTHER RESEARCH For future studies, the following changes and modifications in study design and procedure are recommended to attempt to answer some of the questions created by this study. Use of a double probed density gauge. This would allow point measurements at specific depths to be taken resulting in easier interpretation of results and easier comparisons with other studies. The maximum depth the effects of compaction reached could be determined. Also, a compaction curve for the soil could be developed, which may be useful to the land manager or other researchers. Attempt to measure areas where the exact number of feller-buncher passes is known, particularly areas having 1, 2, 3, and 4 passes. This would help to answer the question on how many passes it takes to cause most of the compaction that will occur under the feller-buncher. Do impact transects similar to those done in this study after both the feller-buncher and skidder. This would allow more precise estimates on the amount of area impacted by the feller-buncher to be made. Thus, comparisons between swing boom and tree-to-tree feller-bunchers could be made in regards to amount of area impacted. In addition, even though the feller-buncher in this study only caused an increase in soil density of 8.8%, on a different soil or under different conditions the increase in density could 69 become unacceptable. Then, knowing the amount of area adversely impacted would become very important to the land manager for possible mitigation measures. Attempt to access the amount of area having between 8 and 50 skidder passes and what increases in soil density is occurring in this area. As previously noted, the amount of area in this category was small in this unit. In other units or with different operators, this area could be quite large, especially if there are a lot of short, 20 - 30 foot long feller-buncher trails. It is not known from this study what is going on in this area in terms of soil compaction. Also, from this study it could not be determined how many skidder passes it takes to compact the soil to the point where additional passes cause no significant increases in soil density. 4, Suggested groupings of skidder passes are 1 5 - 8, 9 - 16, 17 - 30, and 30+. - These groupings might be better able to answer these questions. Collect more data in the 5 - 8 skidder pass area than was collected in this study. This area appears to be the critical number of passes for the break between detrimental and non-detrimental compaction, using the definition given. This break is critical to land managers in deciding if treatment is needed in order to mitigate the detrimental effects of compaction. Collect more information on the amount of area impacted by each machine due to the wide variation found in this 70 study. This would allow more precise estimates of the amount of area impacted by each machine to be made. 71 8.0 REFERENCES Adams, Paul W. and Henry A. Froehlich, 1981. Forest Soils, PNW 217. Compaction of Alexander, Earl B., 1985. Soil Disturbance and Compaction in Wildland Management. USDA Forest Service, Pacific Southwest Region, Watershed Management Staff, Earth Resources Monograph 8. Am. Soc. Ag. Eng. - Soil Sci. Soc. Am. Soil Compaction Committee Report. 1958. Concepts, terms, definitions and methods of measurement for Soil Compaction. Ag. Eng., 39:173-176. Aulerich, D. Edward, Norman K. Johnson and Henry Froehlich, 1974. Tractor or Skyline: What's Best for Thinning Young-growth Douglas-fir? For. md Mag., 101(12):4245. Burger, J. A., V. Perumpral, R. E. Kreh, J. L. Torbert and S. Minaei, 1985. Impact of Tracked and Rubber-Tired Tractors on a Forest Soil. American Society of Agricultural Engineers, 28(2):369-373 Campbell, Robert G., James R. Willis and Jack T. May, 1973. Soil Disturbances by Logging with Rubber-tired Skidders. J. Soil Water Cons., 28:218-220. Cochran, P. H. and Terry Brock, 1985. Soil Compaction and Initial Height Growth of Planted Ponderosa Pine. USDA For. Serv., Res. Note PNW-434, 4 pp. DeVore, Jay and Roxy Peck, 1986. Statistics, The Exploration and Analysis of Data. West Publishing Company, 701 pages. Dyrness, C. T., 1965. Soil Surface Conditions Following Tractor and Highlead Logging in the Oregon Cascades. J. For., 63:272-275. Flint, Alan L. and Stuart Childs, 1984. Development and Calibration of an Irregular Hole Bulk Density Sampler. Soil Sci. Soc. Am. J., 48:374-378. Froehlich, Henry A., 1978. Soil Compaction from Low Ground-Pressure, Torsion Suspension Logging Vehicles on Three Forest Soils. Forest Research Laboratory, Oregon State University, Corvallis. Research Paper 36. 12p. 72 Froehlich, H. A., 1979. Soil Compaction From Logging Equipment: Effects on Growth of Young Ponderosa Pine. J. Soil Water Cons., 34:276-278. Froehlich, H. A., J. Azevedo, P. Cafferata, and D. Lysne, 1980. Predicting Soil Compaction of Forested Land. Final Project Report, Coop. Agreement No. 228. USDA Forest Serv., Equip. Div. Cent., Missoula, Mont. Froehlich, Henry A. and David H. McNabb, 1983. Minimizing Soil Compaction in Pacific Northwest Forests. A paper presented at the 6th North American Forest Soils Conference on Forest Soils and Treatment Impacts. Froehlich, H. A., D. W. R. Miles and R. W. Robbins, 1985. Soil Bulk Density Recovery on Compacted Skid Trails in Central Idaho. Soil Sci. Soc. Am. J., 49:1015-1017. Froehlich, H. A., D. W. R. Miles and R. W. Robbins, 1986. Growth of Young Pinus ponderosa and Pinus contorta on Compacted Soil in Central Washington. Forest Ecology and Management, 15(1986):285-294. Gent, J. A., Jr., R. Ballard and A. E. Hassan, 1983. The Impact of Harvesting and Site Preparation on the Physical Properties of Lower Coastal Plain Forest Soils. Soil Sci. Am. J., 47:173-177. Hatchell, G. F., C. W. Ralston and R. R. Foil, 1970. Disturbance in Logging. J. For., 68:772-775. Soil Helms, John A. and C. Hipkin, 1986. Effects of Soil Compaction on Tree Volume in a California Ponderosa Pine Plantation. Western Journal of Applied Forestry, Vol. 1, No. 4: 121-124. Laing, Larry E. and Steven W. Howes, 1983. Detrimental Soil Compaction Resulting From a Feller Buncher and Rubber Tired Skidder Timber Harvest Operation: A Case. Study. USDA Forest Service, Pacific Northwest Region, Range and Watershed Management, l3pp. Lenhard, R. J., 1986. Changes in Void Distribution and Volume During Compaction of a Forest Soil. Soil Sci. Am. J., 50:462-464. Lysne, D. H. and A. L. Burditt, 1983. Theoretical Ground Pressure Distribution of Log Skidders. Transactions of the ASAE, Vol. 26, No. 5, pp. 1327-1331. Moehring, D. M. and J. K. Rawls, 1970. Detrimental Effects of Wet Weather Logging. J. For., 68:166-167. 73 Schuh, Donald D. and Loren D. Kellogg, 1988. TimberHarvesting Mechanization in the Western United States: An Industry Survey. Western Journal of Applied Forestry, Vol. 3, No. 2: 33-36. Sidle, R. C. and D. M. Drlica, 1981. Soil Compaction From Logging with a Low Ground Pressure Skidder in the Oregon Coast Range. Soil Sci. Soc. Am. Proc., 45:12191224. Silvercides, C. R., 1984. Mechanized Forestry, World War II to the Present. Forestry Chronicle, 60(4):231-235 Steinbrenner, E. G. and S. P. Gessel, 1955. The Effect of Tractor Logging on Physical Properties of Some Forest Soils in S. W. Washington. Soil Sci. Soc. Am. Proc., 19: 372-376. Stokes, Bryce J. and Donald L. Sirois, 1982. Operational Characteristics of a Harvester in Intermediate Cuttings. A paper presented at the Second Biennial Southern Silvicultural Research Conference, Atlanta, Georgia, November 4-5, 1982. Van Bavel, C. H. M., 1958. Soil Densitometry By Gamma Transmission. USDA Forest Service, North Carolina Agricultural Experiment Station, Journal Paper No. 887, pages 50-58. Vomocil, James A., 1954. In Situ Measurement of Soil Bulk Density. Ag. Engr. 35:651-654. Wert, S. And B. R. Thomas, 1981. Effects of Skid Roads on Diameter, Height and Volume Growth of Douglas-fir. Soil Sci. Soc. Am. J., 45:629-632. Youngberg, C. T., 1959. The Influence of Soil Conditions Following Tractor Logging on the Growth of Planted Douglas-Fir Seedlings. Soil Sci. Am. J., 23:76-78. APPENDIX A USDA FOREST SERVICE TIMBER SALE CONTRACT CLAUSES PERTINENT TO THIS STUDY. A-i PART CT6.O - OPRATI0N5 CTS.42f - Soecial Yardino Objectives. (4/82) Special yarding metho shall oc ac:omplisneo oy tne te or yarding equipment listed herein. Methods other than those specified may be approved in writing If methods meet the objective to protect the residual stand and to detrientaj demae to the soj.. uct approva sna inc uce Increases appropriate. in urrent ontrac: ates wnere Sale Area Map shall be revised to record such approval. Soecial Yartf no Methods Pvment Unit 1. Skid or yard trees full-length to the landing 2. Skid trees with one end suspension 3. Skid tra±ls shall be at least 80 feet apart, center to cen:e; encept where converging. 4. Skid trail. shal2. be at lea.: 60 feet apart, center to center, encept where converging 5. Skid trails shall not be wider than 10 feet. Single tractor passes off of skid trail. are allowed as long as rutting or detrinentaj. ccnpac:ion (refer cT6.4# (OptLom 3) is avoided. l,2,3,4,3,6,7,8,LA,2A, 3A,4A,5A 3A,4A,5A Acceotable Tardjn 1. reve Eauioment GA,74 3A,4A,SA All. Payment unit Any tractor or line system capable of accomplishing the above methods AU -.1 P age A-2 a.- PART c76.D - OPERATIONS CT6.4# (Dotfon 3) - Conduct of Locolna. (10/85' Other methods or requfremen e agreed in writing between Purc:iaser and Forest Service. The following hlng measures are applicable for use in addition to those in 8T6.4: s.,- Recuirements Aoollcable to Clearcuttino Unt. (I) Notwithstanding 872.31 all live and dead trees, except Reserve Trec subject to CTZ.3#, which exceed 4 Inches in diameter breast high are designated and required to be cut. Trees reouired to be cut, that meet the minimum 0311 in AT2, sha}l be felled c:ncurrenty with the logging operation. Concurrent felling does not apply when Stage logging is authorized. Recuirementz Aoolicable to Partial Cuttinc Areas. The use of roll-out tractors will, not be permitted. Payment Units of Sale Area where tree length logging is prohibited are so designated on Sale Area Map. Tree length loaglng is defined as skidding, yarding or otherwise moving a tree in one piece not Including the stump or top (that part of the tree less than the d.i.b. specified in AT2). (ill) When necessary to meet the terms and cunditions of 3T6.42, the winch line shall be unspooled from the tractor winch anc manually pulled to the location of individual logs for skidding purposes. 'c) General Loacino Reauirements. (i) If Sale Area Nap designates any cuttIng unit for stage loggln'a, timber in such a unit shall be removed in two or more separate felling and yardina operations in accordance with a plan agreed to by Purchaser and Forest Service. Stage logging by species will not be permitted on Sale Area, unless Purchaser and Forest Service acree to conditions for such an operation, including but not limited to an increase In performance bond amount. (if) If designated on Sale Area Map, landings, Temporary Roads, and/or tractor skid trails or portions thereof which are detrimentally compacted as defined below shall be ripped through the compacted zone but not to exceed a depth of 20 inches. Ripping shall span the width of compacted area, and the distance between ripped furrows shall not exceed 36 inches. Ripping shall be done with winged Subsoilers during Normal Operating Season unless otherjise approved by the Forest Service. Detrimental compactton Is defined as a 15 percent increase in soil bulk density frme the undisturbed level for payment units 6A and 7A ano a like increase of 20 percent for payment units 1. 2, 7, 8, lA, 2A, 3A, 4A, and 5A Page / A -3 , 4, , b, APPENDIX B Calculation of Number of Sample Points for Before Loqqinq Conversations with the two Forest soil scientists indicated that the soils in this unit were fairly uniform in soil properties such as texture, rock content, parent material and bulk density. A reasonable estimate of average, before logging bulk density was .7 gm/cc. Assuming most before logging bulk densities would fall in the range of .55 to .95 gm/cc (standard deviation of .25 gm/cc), and desiring to have 90% confidence that the bulk density determined would be less than +.1 gm/cc of the true mean bulk density, the following formula was used to determine the number of before logging sample points needed (DeVore and Peck, 1986). ta ()2 n= where: n t s B = = = = B2 sample size (solved for) student's t value for 90% confidence estimated standard deviation specified error of estimation Substituting in values: n = (1.645)2 (.25)2 (.075)2 = (5.48)2 : 30 Thus, 30 sample points should be sufficient before logging. B-i Calculation of 95% Confidence Intervals for Measured Before Loqqin Bulk Densities To determine the actual confidence in the measured, before logging soil densities, 95% confidence intervals were calculated with this formula (DeVore and Peck, 1986): S CI = X + t * n1'2 where X CI t n S = = = = = observed mean bulk density for the soil layer confidence interval student's t value for 95Z confidence number of sample points taken standard deviation of the mean density The chart below summarizes the calculations for each soil layer. These confidence intervals are for the number of measurements used for the results. Soil Layer - (in.) X (gm/cc) o - 4 0.699 27 0.065 2.06 .013 0 - 8 0.690 28 0.055 2.05 .010 0 - 12 0.651 28 0.063 2.05 .012 S S (gm/cc) n t Computed Confidence Intervals 0 - 4 inch layer = .699 ± .027 = .672 to .726 gm/cc 0 - 8 inch layer = .690 .021 = .669 to .711 gm/cc 0 - 12 inch layer = .651 ± .025 = .626 to .676 gm/cc B-2 n1'2