We measured soil density and tree growth after wet-season, ground-based yarding on fine-textured soils at three clear-cut sites. Four treatment conditions were sampled on or near four skid trails (replicates): nontilled and tilled primary skid trails, and adjacent slash-treated areas; the fourth treatment was secondary skid trails at two locations and a logged-only control at the third location. The 16 treatment plots were split into 4 subplots, each randomly assigned to a species–stock type and planted with 30 seedlings. Tree data through year 8 after planting were analyzed as a randomized block, split-plot design. Compared with nontrail areas, bulk density in the 0- to 8-cm depth on primary skid trails after logging averaged 41–52% greater. Eight years later, bulk density in the 0- to 30-cm depth of primary skid trails still exceeded that outside trails by about 20%, yet tree survival was similar except for western hemlock (Tsugaheterophylla (Raf.) Sarg.) having poorer survival on nontilled trails at one location. Average tree height and volume of Douglas-fir (Pseudotsugamenziesii (Mirb.) Franco var. menziesii), Sitka spruce (Piceasitchensis (Bong.) Carr.), and hemlock (except height at one location) did not differ among treatments. Observed differences among treatment means were small. Power analyses indicated that 20% or larger differences in tree height or volume were detectable with 30–95% power.

A study of impacts of forest harvest on volcanic ash soils was conducted on three National Forests in the Blue Mountains of Eastern Oregon and Washington. Surface soil conditions were assessed in 11 forest harvest units; sampling employed a randomly positioned grid of randomly oriented line transects. In addition, air permeameter readings were tested in seven units for correlation with bulk density values obtained from core samples. Most of the detrimental soil conditions detected were attributable to compaction. Displacement, although readily apparent during harvest or slash piling with ground‐based equipment, was nearly undetected by the methods used on these units, which ranged from 14 to 23 yr old. Detrimental compaction averaged 28 or 19% of the harvest unit area when defined as either a 15 or 20% increase in bulk density, respectively. Area percentages included skid trails and landings but not the transportation system. Statistical tests showed that either 5 or 3 of the 11 units exceeded 20% area in detrimental condition, depending on degree of bulk density increase used to define detrimental compaction. Several other units were borderline in exceeding 20% of the area. Given slow recovery from compaction, there is potential for additional area to be adversely impacted by future entries. Air permeameter readings were poorly correlated with bulk density assessments, even though six of seven correlations were statistically significant. Less than 12% of the total variability was accounted for by regression analyses. The air permeameter does not appear to be accurate enough to measure compaction in volcanic ash soils.

AuthorsMonitoring soil compaction is time consuming. A desire for speed and lower costs, however, must be balanced with the appropriate precision and accuracy required of the monitoring task. We compared three core samplers and a cone penetrometer for measuring soil compaction after clearcut harvest on a stone-free and a stony soil. Precision (i.e., consistency) of each tool at depths of 0-10, 10-20, and 20-30 cm was determined from two adjacent samples at 21 or more sampling points in each harvested location. Because one bulk density (D b ) sampler provided a continuous sample of each decimeter depth, it was designated as the standard; thereby, the relative accuracy and bias of the two shorter core samplers could be calculated. Both shorter samplers overestimated D b as determined by the standard. At least 15 penetrometer samples could be taken and processed in the time required for three D b samples to the same 30-cm depth. Precision of measurements taken by the cone penetrometer, however, was clearly less than that with any of the D b samplers. Based on time requirements and precision of each tool, we examined the efficiency of double sampling (using a combination of penetrometer and core sampler) for estimating D b . Results from the stone-free soil indicated an advantage in both precision and efficiency in applying double-sampling theory to estimate D b rather than sampling exclusively by the more time-consuming core samplers.Keywords: Bulk density, measurement precision, relative accuracy, cone penetrometer, soil strength. SummarySoil bulk density (D b ) is the ratio of mass of dry solids to bulk volume of soil. Bulk density, expressed as Mg•m -3 or g•cc -1 , is commonly used to measure soil compaction. An increase in D b indicates that movement of air and water within the soil has been reduced, and that the soil may be less favorable for plant growth or be more likely to erode. Soil compaction also can be indicated by a penetrometer (PEN), a device forced into the soil to measure its resistance to vertical penetration. Soil strength, as measured in kPa by a PEN, is related crudely to soil resistance as encountered by roots. Because PEN readings can be strongly dependent on soil moisture content, field sampling for comparative purposes should be done when soils are near field capacity.Our general objective was to compare precision and efficiency of four tools conventionally used to monitor soil compaction. We compared three bulk density samplers and a cone penetrometer for measuring soil compaction over a range of sampling conditions in two young forests. Both stands had been planted after clearcutting and yarding by crawler tractors (D-8) in winter on wet silt loam soils. Both study locations were sampled during the spring rainy season; therefore, soils were probably at or near field capacity. Sampling points for each tool were assigned by subdividing each potential sampling plot, a 25-x 100-cm rectangle, into four equal-sized squares and randomly allocating a tool to each square. Two subsamples (replica...

The accurate prediction of which Douglas fir [Pseudotsuga menziesii (Mirb.) Franco] stands may or may not respond to N fertilization should improve the economics of the practice. Our objective was to determine the efficacy of various stand (site index, age, and relative density), climatic (total precipitation, average daily solar radiation), site (elevation, soil depth, and available water‐holding capacity), and soil‐test variables (anaerobically mineralized N, total N, organic matter, and C/N ratio) to predict relative and absolute response to a single application of 224 kg N ha−1 as urea. The core equation with stand variables accounted for 70% of residual variation in average annual volume growth. Predicting response of fertilized stands proved much less precise. The best core equation explained 37% of the residual variation for average percentage response in volume growth and explained less variation in absolute response in both volume and basal area. Of the site, climatic, and soil‐test variables, C/N ratio in the surface soil was the only one that significantly increased precision of the core equations. The best combined equation explained 46% of the variation in percent volume response. The anaerobic N mineralization test failed to make a significant contribution to the core equation and had a lower correlation with response than did the C/N ratio. Stand variables remain the most reliable predictors of fertilizer response in this region; any improvement from including soil data for N or organic matter is not justified because of their additional cost.

In a nitrogen-deficient plantation in southwest Washington, we (1) compared effects of 224 kg Nlha as ammonium nitrate, urea, and biuret on volume growth of dominant and codominant Douglas-fir (Pseudotsuga menziesii var. menziesii (Mirb.) Franco); (2) determined how 8-year response of these trees to fertilization was related to their distance from a strip of the plantation interplanted with nitrogen-fixing red alder (Alnus rubra Bong.); and (3) observed effects of biuret on understory vegetation. On both sides of the strip centerline, we grouped subject trees into 30 plots of 4 trees each, based on slope position and distance from alder. We randomly assigned three fertilizers and a control within each plot. We analyzed separately data from east and west of the mixed stand centerline. Initial volume differed greatly among the 120 trees on each side, so we used covariance analysis to adjust observed treatment means. Adjusted mean volume growth was increased (p ~0 . 1 0 ) by 22 to 28 percent on the east side and by I 1 to 14 percent on the west side, with no significant difference in response to the three fertilizers. Only biuret stimulated growth within the mixed stand. Biuret had no visible toxic effect on competing vegetation during 8 years after application.Keywords: Pseudotsuga menziesii, Douglas-fir, Alnus rubra, red alder, nitrogen fertilization, biuret, ammonium nitrate, urea, volume growth. SummaryUrea is the most commonly used source of nitrogen (N) for fertilizing Douglas-fir (Pseudotsuga menziesii var. menziesi~) forests in the Pacific Northwest. Although ammonium nitrate is an alternative source, about 20 percent more gain in cubic volume growth would be necessary to offset its greater treatment cost (material plus spreading). A third potential source is biuret from which nitrogen is released more slowly than from urea and, especially, from ammonium nitrate. Although biuret is toxic to most agricultural plants, some coniferous tree species, including Douglas-fir, tolerate biuret and respond positively in seeding size and weight. We report the first comparison of biuret versus other conventional sources of nitrogen under field conditions. The scope of our comparison was restricted to an N-deficient plantation in southwest Washington. At the time of fertilization, this 57-year-old plantation still retained most of the trees originally planted at a 2.4-by 2.4-m spacing after the 1928 growing season. Douglas-fir numbers were fewer, but volume of individual dominant trees averaged about 2.6 times greater, in a 27-m-wide strip through the plantation that had been interplanted at year 4 with N-fixing red alder (Alnus rubra Bong.). We surveyed a 15-by 15-m grid east and west of the mixed stand and selected the largest undamaged Douglas-fir closest to each grid point as our subject trees. The 240 subject trees were either 15, 30, 45, or 60 m from the outer row of red alder or within the mixed stand (0 m). At each of these distances, we randomly assigned one of four treatments (urea, ammonium nitrate, biuret, or...