| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Faculty of Forestry, Univ. of British Columbia, Vancouver, BC V6T 1Z4, Canada
b British Columbia Ministry of Forests and Range, Research Branch, Vernon, BC V1B 2C7, Canada
c Dep. of Geography, Simon Fraser Univ., Burnaby, BC V5A 1S6, Canada
* Corresponding author (krzic{at}interchange.ubc.ca).
 |
ABSTRACT |
|---|
 TOPNOTES ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Abbreviations: BWBS, Boreal White and Black Spruce biogeoclimatic zone • CDF, Coastal Douglas-fir biogeoclimatic zone • CWH, Coastal Western Hemlock biogeoclimatic zone • f, porosity • fMBD, porosity at MBD • ICH, Interior Cedar–Hemlock biogeoclimatic zone • IDF, Interior Douglas-fir biogeoclimatic zone • LTSP, Long-Term Soil Productivity Study • MBD, maximum bulk density • PCA, principal component analysis • SBS, Sub-Boreal Spruce biogeoclimatic zone • gravimetric water content at which MBD was achieved
|
INTRODUCTION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
To maintain sustainable soil productivity, it is necessary to assess the ability of a soil to support plant growth after machines have traveled over it. Soil scientists studying sustainability have traditionally measured bulk density as an indicator of compaction, and this measurement is also made because bulk density is a key soil property for determining site nutrient contents. Despite this, limiting values for bulk density have not been defined for the wide range of soil conditions typical in forests, primarily because such limiting values are different for soils with varying texture, organic matter content, and other properties. Establishment of limiting values would be beneficial for soil scientists and land managers.
One approach to better evaluate the state of soil compaction among soil types involves expressing the actual bulk density as a percentage of some reference compaction state (Lipiec et al., 1991; Topp et al., 1997; Lipiec and Hatano, 2003). The idea of comparing soil physical conditions at field sites to a reference state was also proposed by Joosse and McBride (2003), who proposed comparisons based on the void ratio to evaluate the soil quality of agricultural sites. Such comparisons would allow conditions from a wide range of soil types to be evaluated using a single threshold limit, much as the critical limits of soil mechanical resistance and air-filled porosity appear to be relatively independent of soil type (Hakansson and Lipiec, 2000; Zou et al., 2001). Therefore, use of a reference state could potentially enhance interpretations in soil compaction studies.
Various parameters for a reference compaction state have been proposed (Carter, 1990; da Silva et al., 1994; Hakansson and Lipiec, 2000), but the maximum bulk density (MBD) determined by the standard Proctor compaction test (ASTM, 2000) is rigorously defined, is readily determined with standard test equipment, and has been used in several studies (Carter, 1990; Smith et al., 1997; Aragon et al., 2000). The potential advantages of using MBD as a reference compaction state can only be realized if the soil samples used to determine it reliably represent site conditions, and this can create challenges in forest soils. Unlike agricultural soils, where soil type is often relatively consistent within a particular field, the properties of forest soils are known to vary widely across short distances (Courtin et al., 1983) on many forested sites in response to more variable topography and the absence of tillage to mix and homogenize surface layers. Such variation would require a large number of samples to be taken to determine MBD, and some alternative method to predict MBD would be beneficial.
The standard Proctor method (ASTM, 2000) evolved from studies by civil engineers (Proctor, 1933) on the compaction of soils for dam and road foundations. Two parameters are obtained from this method: MBD and the critical water content at which MBD is achieved for a given amount of energy (WMBD). The compactive force applied in the Proctor test as it is used in engineering studies has evolved over the years to make it more applicable to changing needs. Despite this, no information is currently available to determine whether different levels of applied force would improve interpretations of compaction effects on growth (Hakansson and Lipiec, 2000). Therefore the standard Proctor test is commonly used in productivity studies (Carter, 1990).
The variation in MBD as determined by the standard Proctor test for a range of soils has been attributed to changes in soil organic matter, particle size distribution, Fe and Al oxides, or plastic and liquid limits. For example, quantity as well as quality of organic matter has been determined to have effects on MBD (Soane, 1990; Aragon et al., 2000), and both organic C (Donkin, 1991; Smith et al., 1997; Krzic et al., 2004) and readily oxidizable organic matter (Ball et al., 2000) have been used to predict MBD. Cementing agents, such as Fe, Al, or Mn oxides (in acidic soils) and carbonates (in calcareous soils) enhance aggregate stability, contributing to high soil shear strength (Yee and Harr, 1977). Dorel et al. (2000) reported that Caribbean Andosols and Nitisols (FAO, 1998) or Andisols and Alfisols (according to Soil Survey Staff, 2006) were more resistant to compaction because of the presence of stable microaggregates containing halloysite and Fe oxide. Larson et al. (1980) found that among 36 agricultural soils from around the world, soils with predominantly kaolinite or Fe oxide in the clay fraction had lower MBD than soils with predominantly 2:1 type clays.
The MBD was significantly correlated with clay, fine silt, coarse silt, medium sand, and fine sand, and the clay + silt fraction had the strongest (inverse) correlation with MBD (R = 0.79) on 26 South African forest soils (Smith et al., 1997), while Nhantumbo and Cambule (2006) also showed a relationship between clay content and MBD. Peakedness (kurtosis) and symmetry (skewness) of the particle size distribution curve have also been suggested as parameters in predicting MBD (Webster and Oliver, 1990). Well-graded soils, as indicated by a low coefficient of kurtosis, tend to have a higher MBD. A linear relationship between MBD and kurtosis (R2 = 0.82) was reported by Moolman (1981), while Smith et al. (1997) showed that MBD decreased as the degree of kurtosis increased, but the relationship was not strong (R = –0.48).
Particle density of mineral soils dominated by Fe oxides and heavy minerals can range from 3.0 to 5.0 Mg m–3 (Padmanabhan and Mermut, 1995; Ruhlmann et al., 2006), while organic soil may have a particle density as low as 0.84 Mg m–3 (Redding and Devito, 2006). Since soil bulk density is influenced by particle density, consideration of particle density should be made when predicting MBD. This is particularly important when the soils examined have a wide variation in particle density, as they may for groups of soils developed on diverse geologic materials in mountainous terrain.
Due to the complex interrelationships among soil properties, attempts have also been made to combine several soil properties when predicting MBD. For example, variation in MBD was predicted well by the liquid limit, organic C, and sand (R2 = 0.98) in a study performed by Howard et al. (1981) on 14 forest and rangeland soils in California. Similarly, Ball et al. (2000) showed that MBD, WMBD, and total porosity at MBD were predicted (R2 = 0.49, 0.55, and 0.43, respectively) by a combination of liquid limit and readily oxidizable organic matter for a range of cultivated soils in Great Britain.
Based on these findings, and as part of a larger study performed throughout British Columbia to determine the effects of compaction on forest soil productivity and tree growth, we evaluated the potential for predicting MBD based on properties that can be determined on samples normally collected during field evaluations of bulk density on forested sites. Our objectives were to: (i) evaluate the relationships between MBD and WMBD as determined by the standard Proctor test and other soil properties for a wide range of British Columbia forest soils; (ii) identify the soil properties most important for predicting MBD; and (iii) describe a proposed method for using MBD as a reference bulk density in forest soil compaction studies.
|
MATERIALS AND METHODS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
The sites sampled for this study included 12 long-term soil productivity study (LTSP) installations, three long-term landing rehabilitation trials, seven provincial park sites, five oil-exploration-disturbed sites, four road rehabilitation sites, and two stumping-disturbed sites (Fig. 1 ). The LTSP sites in British Columbia are part of the North American LTSP network that includes the USDA Forest Service, Canadian Forest Service, British Columbia Ministry of Forests and Range, and various universities and industry groups (Powers, 2006).
|
35 kg each) were collected at 0- to 0.1-, 0.1- to 0.2-, or 0- to 0.2-m depth after removal of the forest floor (if present). The number of samples collected at each site varied between two and 12.
Soil Analysis
Maximum Bulk Density
The MBD and WMBD were determined using the standard engineering Proctor test (Proctor, 1933; ASTM, 2000). Soil samples were air dried until friable and then passed through a 19-mm sieve followed by sieving through a 4.75-mm sieve and further air drying. To carry out the test, an initial estimate was made for each sample of the water content at which MBD would be achieved. Because WMBD is typically slightly less than the plastic limit, the initial estimate involved determining the water content of a sample that had been moistened to the point where, after squeezing in the hand, it would remain in a lump when hand pressure was released, but would break cleanly into two pieces when "bent" (ASTM, 2000). Water was then added to a 2.3-kg subsample until it reached the estimated water content, and then four more subsamples were prepared, two with soil water content (W) below, and two with W above this value. The five subsamples were then left in sealed plastic bags to equilibrate overnight. During the test, soil was compacted in a standard mold (9.43 x 10–4 m3) using a 2.5-kg rammer falling freely from a height of 0.3 m. The soil was added to the mold in three layers and 25 blows of the rammer were applied to each layer. Total compactive effort applied to the sample was approximately 600 kN m m–3 (or 595 kJ m–3). The compacted sample was used to determine bulk density and corresponding water content. Soil water content was determined gravimetrically (w/w) by drying samples at 105°C for 16 h. Dry bulk densities vs. W values were plotted on a graph and the points were fitted with a best-fit curve (third-order polynomial). From the resulting compaction curve, MBD was determined from either (i) the peak of the curve, or (ii) the highest sample value when the peak of the curve lay below that level. Approximately 0.5 kg from each sample was sieved through a 2-mm sieve to determine the percentage of fine fraction, which was used to correct MBD. The volume of mineral coarse fragments was determined from dry mass and assumed to have a particle density of 2.65 Mg m–3. Fine-fraction MBD was calculated as the mass of dry, coarse-fragment-free mineral soil per volume of moist soil, where volume was also calculated on a coarse-fragment-free basis. All MBD values are reported on a fine-fraction basis.
Particle Density
Particle density was determined by the gas displacement method (Flint and Flint, 2002), which was modified so that the expansion chamber (instead of the sample chamber) was pressurized to 239 kPa with room air. The expansion chamber was then opened to the sample chamber. Volumes of the two chambers, hose, and transducer were not measured directly; instead, a model describing the volume–pressure relationship was derived based on the changing volume of the sample chamber with known-volume plastic disks. Soil samples <2 mm, oven dried at 60°C for 48 h, were used for the test. The samples used for determination of particle density were not treated to remove organic matter, so the reported particle density for each sample reflects an average particle density for the entire fine fraction, including organic and mineral components.
Soil Organic Matter
Soil total C was determined by the dry combustion method (Nelson and Sommers, 1996) using a LECO analyzer (LECO Corp., St. Joseph, MI) on a sample that had passed through a 2-mm sieve. The estimate of the "readily oxidizable organic matter" was obtained as the weight difference before and after treatment of the sample with H2O2 and expressed as the gravimetric fraction of the original mass of soil. Hydrogen peroxide treatment involved heating a 50-g sample with 75 mL of H2O2 (30%) and 300 mL of water to 80°C and adding small increments of H2O2 until no further reaction was observed.
Soil Oxides
Soil oxides of Al, Fe, Mn, and Si were extracted by 0.2 mol L–1 acid ammonium oxalate solution. This method extracts active Al, Fe, Mn, and Si oxides, including a fraction of oxides bound by organic matter (Loeppert and Inskeep, 1996). The extracted ions were measured by inductively coupled plasma spectrometer.
Particle Size Distribution
Soil particle size distribution was determined by the hydrometer method (Gee and Or, 2002). Samples were pretreated with H2O2 (30%) and heat, while samples from the IDF zone that may have contained carbonates were also treated with a NaOAc buffer. Particle size distribution was described using the Canadian System of Soil Classification (Sheldrick and Wang, 1993) in terms of the percentage of clay (<0.002 mm), fine silt (0.002–0.005 mm), medium silt (0.005–0.02 mm), coarse silt (0.02–0.05 mm), very fine sand (0.05–0.10 mm), fine sand (0.10–0.25 mm), medium sand (0.25–0.50 mm), coarse sand (0.50–1.00 mm), and very coarse sand (1.00–2.00 mm).
Peakedness (kurtosis) and symmetry (skewness) of the particle size distribution curve were calculated using the equations proposed by Webster and Oliver (1990):
 | [1] |
 | [2] |
(Xi – µ)2 is the second moment of the distribution about the mean of the observation, M3 = (1/n)(Xi – µ)3 is the third moment of the distribution about the mean of the observation, M4 = (1/n)(Xi – µ)4 is the fourth moment of the distribution about the mean of the observation, n is the number of observations, Xi is the ith observation, and µ is the mean of the observations.
Plastic and Liquid Limits
The plastic limit was determined as the gravimetric water content at which a soil sample could be rolled by hand into a thread of 3-mm diameter without breaking (McBride, 2002). The liquid limit was determined using the one-point Casagrande method (McBride, 2002). The soil water content at which 20 to 30 blows are required to close a groove along a distance of 13 mm was determined gravimetrically after drying at 105°C for 16 h. The liquid limit was calculated using the following equation:
 | [3] |
Statistical Analysis
Distribution of the data was summarized by principal component analysis (PCA) using the SAS PRINCOMP procedure (SAS Institute, 1990). Before the PCA, "missing values" in the data matrix (i.e., 34 samples without plastic limit and three without liquid limit) were filled by the SAS PRINQUAL procedure with the minimum generalized variance method (SAS Institute, 1990). Simple regression analyses between dependent (i.e., MBD, WMBD) and independent variables (i.e., soil properties) were run. Samples without plastic or liquid limits were not included in the regression. In addition, multiple regression analysis was used to select soil properties that were highly correlated with dependent variables. A stepwise method was selected in multiple regression analysis because certain soil properties (e.g., plastic and liquid limit, total C, and oxidizable organic matter) may have multiple impacts on soil MBD. Significance levels of the 
2 score for variable entry and stay were set at 0.25 and 0.10, respectively.
|
RESULTS AND DISCUSSION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
|
Although compaction effects on forest soil productivity are more directly related to changes in porosity, f (volume of pores/total soil volume), than bulk density, it is bulk density (and MBD) that is most commonly used to describe the forest soil condition (Smith et al., 1997). For groups of soils where particle density varies only slightly, bulk density and MBD can accurately reflect changes in f as they affect plant growth because the change in MBD is inversely related to the volume of pores. For groups of soils with a range of particle densities, however, changes in bulk density and MBD will reflect both differences in the volume of solids and particle densities, and the compaction state as it affects growth might be better described by the change in f rather than by the bulk density. In Fig. 2 , we plotted the Proctor test curve of the soil with the highest particle density (2.97 Mg m–3), then replaced its particle density with the lowest one (2.33 Mg m–3) and plotted the five points again. The derived MBD of the hypothetical soil was 28% lower than that of the soil with the highest particle density (Fig. 2a), even though f remained the same (Fig. 2b).
|
Soil Organic Matter
The MBD was negatively and WMBD positively related to total C and oxidizable organic matter (Table 3). The MBD is commonly considered to be linearly related to organic matter (Soane, 1990; Zhang et al., 1997; Krzic et al., 2004). In this study, however, for both total C and oxidizable organic matter, exponential models gave a better description of the relationship (R2 = 0.70 and 0.64, respectively; data not shown) than linear models (R2 = 0.65 and 0.61, respectively; data not shown). Linear models described well the relationships between WMBD and total C and oxidizable organic matter (R2 = 0.65 and 0.54, respectively; data not shown).
Greacen and Sands (1980) reported that increased organic matter in sandy soils under radiata pine (Pinus radiata D. Don) forests in South Australia was associated with reduced compaction (measured as bulk density) under a given load. In a study performed by Ball et al. (1989) on Gleysol and Cambisol (FAO, 1998) or Inceptisol (Soil Survey Staff, 2006) soils from Scotland, a reduction in MBD of 0.18 g cm–3 per increase of 1% organic C was observed. In our study, total C had a similar effect on MBD.
Oxidizable organic matter is used as an indicator of the soil organic matter quality, since H2O2 oxidizes the colloidal, humified organic matter but not the fibrous residues (Day, 1965). The relationship between MBD and oxidizable organic matter has been reported in several studies. For example, Soane (1990) indicated that highly humified material increased soil aggregate stability and soils with high oxidizable organic matter tended to be less compacted. Ball et al. (2000) have shown that oxidizable organic matter explained 63% of variation in MBD of British agricultural soils.
In our study, MBD and WMBD were better correlated to total C than oxidizable organic matter (Table 3). It has been reported that oxidation of the soil organic matter by H2O2 is restricted in the presence of Fe, which tends to stabilize soil organic matter during oxidative degradation (Oades and Townsend, 1963). The weaker relationship of MBD and WMBD with oxidizable organic matter than with total C could be attributed to the presence of Fe oxides, indicating the importance of including several soil properties in a model to increase compactability prediction. On the other hand, Soane et al. (1972) reported lower correlations between MBD and WMBD and organic C (R = –0.72 and –0.61, respectively) compared with oxidizable organic matter (R = –0.81 and –0.74, respectively) for 58 British agricultural topsoils. They used sodium dichromate mixtures to test organic C, which did not completely oxidize the organic compounds. In addition, the reaction between soil Fe and Mn oxides with dichromate may have further lowered the organic C (Nelson and Sommers, 1996).
Organic matter affects the compaction process in at least two ways: (i) it increases soil resistance to compaction by enhancing the contact between soil particles (Soane 1990); and (ii) its low particle density (Redding and Devito, 2006) compared with soil mineral particles reduces the overall particle density and therefore bulk density, especially when organic matter content is high. For our soils, total C accounted for 30.2% of the variation in particle density. The coefficient of determination for the relationship between fMBD and total C, where the density effect of total C was removed, was 5% lower than that for MBD with total C (Table 3). Therefore, we conclude that the organic matter had the strongest effect in improving soil resistance to the compactive force for our group of soils.
Soil Oxides
The MBD was negatively and WMBD positively related to Al and Fe oxides (Table 3). Linear relationships for Al oxide with MBD and WMBD were both stronger than those for Fe oxide (Table 3), while adding an exponential component further improved the relationships with Al oxide (R2 = 0.53 and 0.44; data not shown). In these relatively young soils of British Columbia that have developed since the most recent glaciation, oxides of Fe and Al are the main cementing agents that enhance aggregate stability (McKeague and Sprout, 1975).
Addition of Fe oxides along with organic C, liquid limit, and sand into the prediction model used in a study by Howard et al. (1981) improved the predictability of MBD by 1% (R2 = 0.99). Contrary to our findings, Fe oxide was positively related to MBD in that study. Howard et al. (1981) used the citrate–bicarbonate–dithionite extraction method, which removed the total Fe oxide. The positive relationship observed in their study appeared to reflect the effect of Fe on soil mass rather than on soil strength, as soil particle density increases with Fe content. The negative relationship between MBD and active Fe oxide (extracted by ammonium oxalate) in our study reflected the enhanced soil strength due to the presence of soil oxides. Consequently, testing of active oxides along with total oxides could provide important information to determine the mechanisms by which these materials affect compaction.
Particle Size Distribution
Our results show that particle size distribution was not the major factor related to variations in MBD or WMBD across the entire range of soils we studied, even though correlation coefficients were significant between MBD and medium silt, and between WMBD and clay, fine silt, medium silt, fine sand, medium sand, and coarse sand (Table 3). In previous studies, increasing clay content has either been associated with lower MBD (Smith et al., 1997; Nhantumbo and Cambule, 2006) or had little effect on MBD (Ball et al., 2000; Aragon et al., 2000).
Kurtosis and skewness were previously considered to be useful parameters in predicting soil MBD (Moolman, 1981). A low coefficient of kurtosis indicates a well-graded particle size distribution and is expected to lead to a higher MBD, while a low absolute value of the skewness coefficient indicates high symmetry of the distribution curve and higher MBD. In our study, kurtosis ranged from –1.67 to 3.46 and skewness varied from –0.26 to 2.26, but we found no relationship between MBD or WMBD and kurtosis or skewness (Table 4 ). Smith et al. (1997) suggested that the relationship between MBD and kurtosis would be confounded by the significant relationship between MBD and organic matter, which may also have occurred in our study.
|
Plastic and liquid limits integrate several soil properties such as particle size distribution, organic matter content, and clay mineralogy. Our findings are similar to those of Soane et al. (1972), who tested 13 properties of 58 Scottish topsoils and found that MBD and WMBD were highly related to the plastic and liquid limits (R = –0.80 and –0.68 for MBD, 0.74 and 0.69 for WMBD). Howard et al. (1981) found that the MBD of California forest and rangeland soils was significantly correlated to the liquid limit (R = –0.96) but they did not report the plastic limit. Ball et al. (2000) reported that the liquid limit accounted for 43 and 48% of the variation in MBD and WMBD, respectively, of British soils, while the relationship between MBD or WMBD and the plastic limit was lower (R2 = 0.29 and 0.36, respectively). Relative to our data, the correlation coefficient between MBD and liquid limit was higher in a study by Howard et al. (1981) and lower in Ball et al. (2000). The former study tested 14 Californian soils predominated by loam texture, while the latter evaluated 146 British agricultural soil samples with a wide variation in texture. Differences in the soil textures might have accounted for the difference in correlation coefficients among the above studies. In the study by Ball et al. (2000), lower correlation coefficients may have been observed because some nonplastic soils had missing values filled by a statistical tool before running the correlation analysis.
Predicting Maximum Bulk Density by a Set of Soil Properties
Principal component analysis (a multivariate analysis tool to examine relationships among several quantitative variables in a data set) showed that the first three components accounted for 67% of the variation in the data set (Table 5
). The first component mainly explained MBD, fMBD, and WMBD and soil properties like liquid and plastic limit, total C, and oxidizable organic matter. The second and third components mainly explained soil texture and Al oxides. This indicated that soil organic matter and liquid and plastic limits had the greatest impact on MBD, fMBD, and WMBD, while particle density, Al and Fe oxides, and some of the particle size classes were of secondary importance.
|
|
|
For the majority of our soils, oxidizable organic matter was preferred to total C when predicting MBD, illustrating the importance of quality-related (i.e., oxidizable organic matter) rather than quantity-related (i.e., total C) soil organic matter in compaction studies. In a study by Howard et al. (1981), organic C was the most important variable in predicting MBD, but they did not determine the active oxides and subgroups of sand and silt size fractions, hence the importance of oxidizable organic matter was not known. Oxidizable organic matter was the second most important variable in predicting MBD in a study by Ball et al. (2000).
In our study, both organic matter (either total C or oxidizable organic matter) and oxides were important for the prediction of MBD (Table 6). In all groups, organic matter (i.e., total C) showed a strong relationship in decreasing MBD (R2 = 0.59–0.72), which is similar to the results of Smith et al. (1997), who found a strong negative relationship between MBD and total C (R2 = 0.88), and also to Aragon et al. (2000), who showed a high dependence of MBD on organic C. Including particle size distribution further improved the prediction. Even though clay came second in predicting MBD for the "overall" group, particle size components usually ranked third or lower in the prediction. Unlike Smith et al. (1997), who found a strong relationship between MBD and clay + silt (R2 = 0.63), there was no relationship between MBD and clay + silt in our nonplastic and moderately plastic groups. Only in the highly plastic group did clay + silt show a high correlation with MBD (Fig. 4 ), but the positive effect we observed was opposite to that reported by Smith et al. (1997). Smith et al. (1997) found that in the lower range of clay + silt (0–400 g kg–1), the effectiveness of clay + silt and total C appeared to offset each other in the compaction test; only in the higher range of clay + silt (400–1000 g kg–1) did clay + silt enhance the effect of total C in reducing MBD. Hence, the importance of these two properties cannot be compared directly in their study. As total C was positively correlated to clay + silt (R2 = 0.33) in their study, the strong effect of clay on MBD may be just a covarying result of organic matter on MBD. For the highly plastic group in our study, there was also a close correlation between clay and total C, which was not apparent in the other groupings (Fig. 5 ), but the relationship we observed showed total C content declining with increasing clay content, opposite to what is commonly expected for a range of soils. Our highly plastic soils group appeared to contain a mix of two subgroups of soils: (i) those with very high clay content but low or intermediate organic matter; and (ii) those with very high organic matter content but low clay content. This may clarify why multiple regressions explained the smallest amount of variation for the highly plastic group.
|
|
Proposed Method for Using Maximum Bulk Density as a Reference in Forest Soil Compaction Studies
To use MBD as a reference value in soil compaction studies, a method for obtaining the best estimate of MBD across a variable site is required. The standard approach to determining MBD using the Proctor test relies on collecting a 10-L sample from the site and carrying out the laboratory test. On typical forestry sites, such a method may be impractical because site variability makes it difficult to identify the "typical" condition that will best represent conditions throughout the site. A better approach would be to collect samples from each variant of the soil conditions, but this too may become unwieldy because of the large number of samples required. Generally, bulk density sampling requires high numbers of samples to account for the natural variation on forestry field sites (Courtin et al., 1983; Page-Dumroese et al., 1999). The method we propose takes advantage of the strong relationships we have observed between MBD and soil properties that are relatively easy to measure. The method involves four steps.
|
CONCLUSIONS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Currently, only bulk density is used widely as a parameter to assess the compaction state of a soil. We have described a method to predict MBD from readily measured soil properties that could enable more effective means of providing reference values for compaction studies. This would be particularly beneficial where these attributes exhibit high point-to-point variation, such as in British Columbia's forest soils. Prediction would involve first determining the plasticity for a soil sample, then using the appropriate equation to determine MBD.
|
ACKNOWLEDGMENTS |
|---|
|
NOTES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Received for publication February 21, 2007.
|
REFERENCES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  H. Blanco-Canqui, L. R. Stone, A J. Schlegel, D. J. Lyon, M. F. Vigil, M. M. Mikha, P. W. Stahlman, and C. W. Rice No-till Induced Increase in Organic Carbon Reduces Maximum Bulk Density of Soils Soil Sci. Soc. Am. J., September 11, 2009; 73(6): 1871 - 1879. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Crop Science | |||
| Journal of Natural Resources and Life Sciences Education |
Vadose Zone Journal | ||||
| Journal of Plant Registrations | Journal of Environmental Quality |
The Plant Genome | |||
