SSSAJ Journal of Natural Resources and Life Sciences Education
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via ISI Web of Science (6)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Mecke, M.
Right arrow Articles by Ilvesniemi, H.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Mecke, M.
Right arrow Articles by Ilvesniemi, H.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Mecke, M.
Right arrow Articles by Ilvesniemi, H.
Soil Science Society of America Journal 64:485-492 (2000)
© 2000 Soil Science Society of America

DIVISION S-1-SOIL PHYSICS

Prediction of Near-Saturated Hydraulic Conductivity in Three Podzolic Boreal Forest Soils

M. Mecke, C.J. Westman and H. Ilvesniemi

Dep. of Forest Ecology, Univ. of Helsinki, P.O. Box 24, FIN-00014, Helsinki, Finland

marja.mecke{at}helsinki.fi


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 Materials and methods
 Results and discussion
 Conclusion
 REFERENCES
 
Steady-state infiltration fluxes into the soil were measured with a tension infiltrometer at supply potentials of -0.35, -0.70, and -1.10 kPa, and the near-saturated hydraulic conductivities (K) were calculated using an exponential model. Measurements were conducted in four mineral soil horizons at three forest sites, representing contrasting textures. The analysis was concentrated on K at -0.35 kPa [K(-0.35)] since this potential corresponds to the 1-mm pore diam., which is often considered to be the limit between macropores and mesopores. The average K(-0.35) of the site varied in the parent soils of the three sites from 0.46 to 40.98 cm h-1, while in the two uppermost horizons the variability was smaller: 0.30 to 0.69 cm h-1. Three multiple linear regression models of log[K(-0.35)] were constructed by stepwise regression analysis. The retained water content at the seven potentials; textural fractions; dry bulk density; and Al, Fe, and C contents were suggested as predictor variables. In addition, simple functions of these variables were suggested. In Model 1, all horizons were included ; in Model 2, all horizons except the upper illuvial horizon were included ; and in Model 3, only the lowest horizon was included . Adding predictor variables increased r2 in all models. The water content at -100 kPa, which depends on pore-size distribution and C content (which produce a strong retarding effect on water flow), were the most important predictors for K(-0.35). Similarly, by gradually excluding horizons where pedological and biological processes had changed the structure and pore-size distribution, r2 increased from 0.86 (Model 1) to 0.99 (Model 3).

Abbreviations: {alpha}, the constant in the exponential K({psi}) function • Db, dry bulk density • K, hydraulic conductivity • Ks, saturated hydraulic conductivity • Kse, extrapolated saturated hydraulic conductivity • K({psi}), hydraulic conductivity at water potential {psi} • {theta}s, total porosity • {theta}({psi}), water content at water potential {psi} • {psi}, water potential • TI, tension infiltrometer


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
Materials and methods
 Results and discussion
 Conclusion
 REFERENCES
 
FEW STUDIES ARE AVAILABLE on the hydraulic properties of forest soils, especially for podzolic soils of the boreal climate region (Espeby, 1990). In these young soils, formed by land–ice activity, the texture of the single-grained parent material is often coarse (Tamminen and Starr, 1994). Assorted layers with soil material showing considerable local variation in particle size may occur, especially in glaciofluvial deposits. In the podzolization process, organic matter accumulates in the topsoil as litter and root exudates, and dissolved organic C migrates downward with the percolating water. Depending on the fertility of the site, the rooting zone may become enriched by substantial amounts of organic C (Liski and Westman, 1997). The precipitation of dissolved Al and Fe in various compounds in an illuvial B horizon is also typical for the podzolization process (Petersen, 1976; Browne, 1995). In the upper illuvial layer, a secondary structure develops because of the accumulation of amorphous material (Al, Fe, C) on pores and within them. In addition, freezing and thawing (Westman and Jauhiainen, 1994), as well as the movement of tree roots when the stems are bent by the wind (Hintikka, 1972), decrease the density of the surface layers and together with the root channels, increase the number of large pores. These processes result in the pore system being separated into matrixpores and macropores (Luxmoore, 1981; Moore et al., 1986). Simple textural properties can seldom give sufficient explanation for models of soil hydraulic properties because of the complexity of soil pore-size distribution created by the soil-forming processes.

For measuring hydraulic conductivity (K) at high potentials, tension infiltrometer (TI) techniques have been applied; the most commonly used TI techniques are based on measuring sorptivity and steady-state flow (White and Sully, 1987), steady-state flow from discs of two different radii (Smettem and Clothier, 1989), or steady-state flow from a single disc at several supply potentials (Ankeny et al., 1991; Reynolds and Elrick, 1991). Hussen and Warrick (1993) and Cook and Broeren (1994) compared these methods and found that they usually gave compatible results.

The objective of this study was to determine near-saturated K of different horizons in three podzolic forest soils of varying parent material and texture by applying the TI technique with the Ankeny et al. (1991) procedure. Multiple linear regression was performed to determine whether K can be predicted by texture, content of organic C and pedogenic Fe and Al, dry bulk density (Db), and the soil water content at given supply potentials.


   Materials and methods
TOP
 ABSTRACT
 INTRODUCTION
 Materials and methods
Results and discussion
 Conclusion
 REFERENCES
 
Site Description and Measurement Locations
Three sites with Haplic podzols (FAO-Unesco, 1990) were selected. The parent material was very coarse (Site 1), medium-coarse (Site 2), and fine (Site 3) sand. Sites 1 and 2 were sorted glaciofluvial sand deposits (groundwater table between 4 and 10 m) and Site 3 was a glacial till with some boulders and a layer of stones in the upper 20 cm of the mineral soil. At Site 3 the thickness of the soil layer above the bedrock varied between 45 and 160 cm. The measurement depths were selected according to the morphological layers, the average depths of which were rather similar at all sites: eluvial horizon (0–5 cm), upper illuvial horizon (5–12 cm), lower illuvial horizon (12–50 cm), and parent soil (>50 cm).

Seven measurement points at 3-m intervals were made at Sites 1 and 2. At Site 3, eight measurement points were selected to cover the variation in soil depth above the bedrock with an average interval of 7 m. At each point, TI measurements were conducted at 0-, 7-, 23-, and 50-cm depths in the mineral soil. At one point at Site 3, TI measurements were not feasible at the deepest level since the soil depth was <50 cm. Three measurements were rejected since the membrane of the TI disc had obviously leaked during the measurement and a fourth was discarded because the infiltration rate was too low to be registered. Our material thus consisted of 83 TI measurements. After TI measurements were taken, a volumetric core sample was taken precisely below the measurement surface with a steel cylinder (5.7-cm diam., 5.8-cm height, n = 83) to determine the Db and soil water–retention characteristics.

Measurement and Calculation of Hydraulic Conductivity
The steady-state infiltration fluxes into the soil were measured with a TI (SW-080, 8.75-cm disc diam., Soil Measurement Systems, Phoenix, AZ) at three supply potentials: -0.35, -0.70, and -1.10 kPa (Ankeny et al., 1991). An infiltration ring (SMS SW-092) was used to preserve the soil surface structure (i.e., to prevent soil compaction). The contact material was fine sand (0.2–0.06 mm), which is reported to have an air-entry potential value between -3.5 and -7.0 kPa and is sufficiently permeable to not be a hydrologically limiting layer (Watson and Luxmoore, 1986). Less than 1.5-mm thickness of contact sand (Reynolds and Zebchuk, 1996) was applied slightly moist to prevent its falling into larger pores and forming wicks (Messing and Jarvis, 1993). The measurements were postponed if the soil was so wet that smearing of fine material during preparation of the surface would change the structure.

Calibration of the TI supply potentials was checked occasionally during actual measurements (Jarvis and Messing, 1995). A descending supply potential sequence was used. According to Reynolds and Elrick (1991), this may introduce hysteresis effects, since progressive drainage will occur close to the disk while wetting continues at and near the infiltration front. No differences, however, between ascending and descending measuring sequences were observed while testing.

The method of Ankeny et al. (1991) is based on the equation of Wooding (1968), which describes unconfined steady-state infiltration from a circular water source. By making a few assumptions, the equation of Wooding can be written as follows:

(1)
where Q is the water flux, r is the radius of a circular water source, {psi} is the water potential, K({psi}) is the hydraulic conductivity–water potential relationship, and {alpha} is the constant in the exponential K({psi}) function of Gardner (1958) (see Eq. [2]). The {alpha} constant can be considered a soil structure parameter in that {alpha} increases as more flow is driven by gravity, and decreases as more flow is driven by soil capillarity.

By assuming a simple exponential relationship between K and {psi} (Gardner, 1958), K({psi}) can be calculated for a known value of saturated hydraulic conductivity (Ks) as follows:

(2)

The TI measurements can be limited to a given pore system by the range of supply potentials used. By applying a small negative supply potential, Smettem (1987) eliminated the influence of macropores in his sorptivity measurements. Messing and Jarvis (1993) summarized the temporal variation of K in a clay soil using a model with separate {alpha} values for macropores and mesopores. If {alpha} does not vary considerably across the supply potential range used, a one-{alpha} model can be applied to calculations of K. This approach will yield a single equation for simpler modeling (Ankeny et al., 1991).

K({psi}) values were calculated with Eq. [1] and [2] using a one-{alpha} model as an approximation. The mean Ks and {alpha} values for each measurement were first calculated with the three supply potentials. Ks and {alpha} were then used to calculate the different K({psi}) values (Ankeny et al., 1991; SMS User's Manual: Tension Infiltrometer, Soil Measurement Systems, Tucson, AZ). Application of Eq. [2] using a constant {alpha} includes the approximation that the measured potential range corresponds to a single-pore system. Thus, Ks is here rather an extrapolated saturated K of this pore system, Kse. The pore system with larger diameters than 1 mm corresponding to -0.35 kPa (Luxmoore et al., 1981; Luxmoore, 1983) can be expected to have larger {alpha} and Kse values (e.g., Messing and Jarvis, 1993).

Laboratory Determination of Soil Physical and Chemical Properties
The water-retention curves were measured with a pressure plate extractor (no. 1600, Soil Moisture Equipment, Tucson, AZ) at potentials of -1.0, -3.2, - 6.3, -10, -100, and -1585 kPa. At -1585 kPa, a homogenized subsample with one-third the volume of the original sample was used. The Db (after oven-drying to 105°C) and water contents were determined gravimetrically. The total porosity ({theta}s) was assumed to be equivalent to the measured water content value at saturation.

Particle-size distribution in seven fractions and the organic-C content were determined from the soil fraction having a <2-mm diam. and corrected for the proportion of particles >2 mm. The particle-size distribution of the finer fractions was measured, using a sedimentation method according to Elonen (1971). Total C content was measured using a Leco CSN-1000 Analyzer (Leco, St. Joseph, MI). To reduce the analytical variability, a 50-mL subsample was powdered before analysis and replicate analyses were performed. The organic-matter content was calculated by multiplying the C content by a factor of two (Huntington et al., 1989). From the <0.6-mm soil fraction, acid ammonium oxalate-extractable Al and Fe were determined according to Wang (1981) and corrected for the proportion of larger particles.

The Models
To predict the near-saturated using simpler soil properties, stepwise multiple linear regression analysis was applied. The correlation analysis of possible predictor variables was calculated earlier in connection with water-retention models (Mecke and Ilvesniemi, 1999). In the stepwise analysis, the first predictor variable is chosen as the one with highest (absolute) pairwise correlation with the dependent variable. The remaining predictor variables are chosen by seeking a variable with the highest partial correlation with the dependent variable, adjusted for the variables already chosen. The tested predictor variables were Db and {theta}s; particle-size distribution in seven fractions; content of gravel; content of organic C, Fe, and Al; as well as {theta}({psi}) at six different potentials (-1.0, -3.2, -6.3, -10, -100, and -1585 kPa). Some simple transformations of these variables, as well as the sum of Al and Fe and the sums of several measured textural fractions (2–0.2 mm coarse fraction, <0.02-mm fine fraction) were also tested (Table 1) . The number of predictor variables in the model was limited by an F-test value required to be 3 or more to include or exclude a variable in the regression model. In addition to the near-saturated K models, the correlations of the structure parameter {alpha} with the measured soil properties were calculated.


View this table:
[in this window]
[in a new window]
  		Table 1 Textural, structural, and pedogenic soil properties. (Mean values and coefficients of variation, in parentheses; n = 7–8)

 
To distinguish the effects of pedological processes on K from textural effects, the regression models were applied separately to (i) all measurements , (ii) all measurements except those from the upper illuvial horizon with average depths of , and (iii) measurements of parent material soil at a depth of . These models are named Models 1, 2, and 3, respectively. Two types of models were constructed: the first was the outcome of the stepwise multiple regression analysis (1a, 2a, and 3a) and the second was a model in which only the best predicting variable, , was used (1b, 2b, and 3b).


   Results and discussion
TOP
 ABSTRACT
 INTRODUCTION
 Materials and methods
 Results and discussion
Conclusion
 REFERENCES
 
Structural, Textural, and Pedogenic Soil Properties
The variation in soil structural, textural, and pedogenic properties (Table 1) covered a large range of their known variation in these soils, and the vertical distribution was typical of podzolic profiles (e.g., Scheffer and Schachtschabel, 1989). The {theta}s variable correlated highly with C content and but not with the coarse or fine textural fractions. The {theta}(-1.0) correlated moderately with the coarse textural fraction . At potentials between -3.2 kPa and -10 kPa, {theta}({psi}) correlated most highly with the coarse textural fraction ; and at -100 kPa, {theta}({psi}) correlated most highly with the fine textural fraction .

Since {theta}s did not correlate with texture parameters, while {theta}(-1.0) and {theta} of lower potentials did, this suggests that there were at least two pore systems: one that is dependent on texture and the other that is not. The pore-diameter limit between these systems was 0.3 to 3 mm, corresponding to potentials between -1.0 and -0.1 kPa. When calculating K from infiltration flux values by applying the one-{alpha} model, we approximated that the supply potential range that was used, from -0.35 to -1.10 kPa, corresponded to a single-pore system. The macropores were thus expected to be >1 mm in diam.

Soil Hydraulic Conductivity
The near-saturated hydraulic conductivity K is needed in bimodal water-transport models (Beven and Germann, 1981). The K value at the measured potential range from -0.35 to -0.50 kPa, which is often regarded as a boundary limit between macropores and mesopores, is then an important soil parameter (e.g., Jarvis et al., 1991). The near-saturated K should also be used as a matching point for the predicted unsaturated K({psi}) curve instead of Ks (e.g., Wessolek et al., 1994).

The K(-0.35) values varied by a factor of 500 (0.13–68.38 cm h-1), but the variation was essentially systematic between horizons and sites (Table 2 a). The effect of texture on K(-0.35) was pronounced when comparing the single-grained parent material (>50-cm depth) of the three sites. The geometric mean of K(-0.35) decreased from 40.98 to 0.46 cm h-1 from Site 1 on very coarse sand to Site 3 on clay-rich sand.


View this table:
[in this window]
[in a new window]
  		Table 2 Tension infiltrometer results of the near-saturated (-0.35 and -0.70 kPa) conductivities at the four measurement depths (geometric mean and range)

 
The effect of podzolization on soil structure was very clear in the coarse and medium-coarse sand soil at Sites 1 and 2. At the two uppermost horizons of these sites, K(-0.35) was of the same magnitude, and the geometric mean measured in the parent material had decreased from 0.48 to 0.69 cm h-1. The effect of parent material on K could be seen in these horizons at all three sites, but while the ratio between the highest and lowest mean K(-0.35) was 1:90 in the parent material, it was 1:2 in the two uppermost horizons. The differences between the sites in the two uppermost horizons were not statistically significant, whereas at Sites 1 and 2 the differences between the two uppermost and two lower layers were clearly significant. The K values at (Table 2b) followed the same pattern as at . In the very coarse soil compared with fine-textured soils, the decrease in K was steeper with decreasing potential, indicating a higher {alpha} value and a larger number of pores emptied between these potentials.

To express the relationship between {alpha} and texture more clearly, the measurement points of Site 3 were divided into two subgroups according to the particle-size distribution of the parent material (Table 3) . In Site 3a (five measurement points), the average clay percentage at depths of 23 and 50 cm was 9%, while at Site 3b (three measurement points) it was 5%. Site 3a had exceptionally dark and compact soil with horizontal cracks and a layered, clay-type structure, while Site 3b had typical loose soil of fine textured podzol . The influence of parent material texture on the average value of {alpha} could be seen at Sites 1, 2, and 3b, while Site 3a had higher {alpha} values than Site 2.


View this table:
[in this window]
[in a new window]
  		Table 3 Exponential coefficient of unsaturated conductivity, {alpha} for Sites 1, 2, 3a, and 3b (arithmetic mean and standard deviation)

 
When all the samples of the eluvial horizon (0- to 5-cm depth) and the samples of Site 3a were excluded from the correlation analysis, the logarithm of the structure parameter, {alpha}, correlated with the coarse textural fraction ( ; Fig. 1) . When these samples were included, r was 0.56. In the eluvial horizon and at Site 3a (Table 3), there appeared to be an effect resulting from a more complicated soil structure, that is, root channels in the eluvial horizon and macropores due to cracking at Site 3a; although {alpha} values in the sampling points were high, the K values were not.



View larger version (18K):
[in this window]
[in a new window]
  		Fig. 1 The coefficient of the exponential K({psi}) relationship, {alpha}, plotted against the coarse textural fraction. The eluvial horizon measurements and the measurements of Site 3a are excluded

 
Reynolds and Elrick (1991) measured {alpha} values for sand, loam, and clay of 0.36, 0.12, and 0.04 cm-1, respectively. The trend of decreasing {alpha} with the finer texture agrees with the results found for the present study. In the study of Messing and Jarvis (1993), high {alpha} values were found for cracking clays, typical for Site 3a.

The Models
In the regression analysis, the water-retention properties of lower potentials explained log(K) better than any single particle fraction, since in all models a function of {theta}(-100) became the first selected variable (Table 4) ; {theta}(-10) also explained log(K) well. The P value of the t statistics in Table 4 gives the risk level by which the coefficient of a single variable is zero, when all other variables of the model are included. Pmax is the highest P value for a single variable in the model.


View this table:
[in this window]
[in a new window]
  		Table 4 Coefficients of regression equations of the three models for prediction of log[K(-0.35)] (cm h-1). The degrees of explanation, standard deviations, and the highest P values of coefficient of single variables, Pmax, for each model are also given.{dagger}

 
For Models 1a and 2a, the second choice was log(C) (Table 4), which showed that the accumulation of C efficiently retarded the water flow. Even if C and {theta}(-100) were correlated with each other ( in logarithmic scales), {theta}(-100) was selected with C rather than any of the noncorrelated texture variables. Due to its colloidal nature, low density, and varying volume, even a small amount of organic material efficiently changes the pore-size distribution by filling and blocking the flow channels (Bouma and Anderson, 1973). In the study of Schuh and Bauder (1986), the slowing-down effect on unsaturated water flow, K(-1.5 kPa), was suspected to be due to high organic C content in altered topsoil materials that lacked pore continuity and in which interpedal spaces were filled.

Model 1a, where all soil layers were included (Fig. 2a) , gave an explanation with . Even if Model 1b (Fig. 2b) had only one predictor variable log{theta}[(-100)], it gave an explanation with . Neither of the models were accurate at small K values. Due to the strong dependency of K on pore radius, the correspondence between measured K(-0.35) and parameters connected to pore-size distribution was more obvious in coarse soils than in soils of fine particles.




View larger version (39K):
[in this window]
[in a new window]
  		Fig. 2 Near-saturated hydraulic conductivity K(-0.35) as predicted by Model 1, when all four horizons are included: (a) Model 1a, all four predictor variables are used; (b) Model 1b, only the best predictor variable is used. For example, regression equation of Model 1a is , where a, b, c, d, and e are the constants given by Table 4

 
Excluding the upper illuvial horizon (5–12 cm) measurements (Model 2; Fig. 3a and b) increased the degree of explanation, since the secondary structure was most developed in this horizon and the system of pores was most irregular. Furthermore, the vertical variation in soil properties was highest in the upper illuvial horizon, and the infiltration and water-retention measurements may not have represented identical soil volumes. Application of {theta}(-100) as the only predictor in Model 2b (Fig. 3b) caused a very big difference in the accuracy of the estimates between small and high K values.




View larger version (38K):
[in this window]
[in a new window]
  		Fig. 3 Near-saturated hydraulic conductivity K(-0.35) as predicted by Model 2, when all horizons except upper illuvial horizon are included: (a) Model 2a, all four predictor variables are used, (b) Model 2b, only the best predictor variable is used

 
In the single-grained parent soil material deeper than 50 cm (Model 3; Fig. 4a and b) , various textural parameters explained the variation in K(-0.35) to a high degree; for example, with two textural parameters, silt and sand, r2 was 0.96. It appeared that podzolization, together with other soil-forming processes, changes soil structure so drastically that simple texture-based models cannot provide sufficient K estimates in the upper soil horizons.




View larger version (32K):
[in this window]
[in a new window]
  		Fig. 4 Near-saturated hydraulic conductivity K(-0.35) as predicted by Model 3, when only the parent soil measurements are included: (a) Model 3a, all four predictor variables are used, (b) Model 3b, only the best predictor variable is used

 
K(-0.35) was not dependent on the small amount of >2-mm particles. When Models 1, 2, and 3 were calculated with mass proportions excluding gravel, similar models with only slightly different coefficients were found. The models constructed for other near-saturated K values at the measured supply potential range and for Kse gave approximately the same degree of explanation as the models presented here. While it is known that Ks often correlates with drainable porosity, defined as the difference between {theta}s and {theta}(-10), it appeared that near-saturated K can correlate with soil properties that are related to a matrix-scale structure.

Puckett et al. (1985) were able to use particle-size fractions to predict the soil hydraulic properties; for example, ln(Ks) could be predicted with clay content . Their material was collected from an area with one type of genesis and highly varying texture that also included medium-coarse sand (1.4–42% clay content). The small {theta}s values may have contributed to successful modeling with texture, since in their study the predicted values of three samples with the highest {theta}s (>40%) deviated considerably from the measured values. In boreal podzols, the high {theta}s values of the two uppermost horizons are connected to various soil-forming processes and thus to a more complicated structure, which is lowering the explanation capacity of texture alone.

Espeby (1990) performed a statistical analysis of 100 soil core samples taken from six soil pits and four different mineral soil horizons in a forested glacial till slope. The correlation of Ks with other soil properties was highest with drainable porosity . When excluding the samples of the eluvial horizon, the correlation increased to . This was explained by the fact that the upper horizon had a higher proportion of macropores and that the soil is more affected by pedological processes than in the lower horizons. The effect of pedological processes was also pronounced in our study, and the effect of accumulated organic C appeared to be the most distinctive.


   Conclusion
TOP
 ABSTRACT
 INTRODUCTION
 Materials and methods
 Results and discussion
 Conclusion
REFERENCES
 
The near-saturated K, which was measured by TI on different horizons of the three podzolic forest soils, had clearly decreased by podzolic and other forest soil processes in the two topsoil horizons of coarse and medium-coarse sand soil, but in the clay-rich sandy soil the near-saturated K remained nearly unaltered. In the parent soil, the texture considerably affected near-saturated K at all three sites.

In these soils the near-saturated K, as well as many other related soil parameters, varied more than in many of the soils used in earlier studies of K, and the correlations among these parameters made model prediction possible. Since the proportion of the coarse textural fraction (2–0.2 mm) was high (23–98%), varying in a regular manner and over a wide range throughout the data material, there was a correspondence between pores and grains of the fabric. In this relatively regular pore structure the near-saturated flow of water, which highly depends on the amount of mesoscale pores, was inversely proportional to the amount of retained water in small pores. In addition, the high amount of organic material in the surface horizons strongly retarded the water flow. Small superimposed effects of the fine material and sesquioxides on K could be estimated in these soils, while in fine soils different variables would be more difficult to separate from each other. Gradually excluding horizons according to the effect of podzolic and other forest soil processes on structure increased steadily the degree of explanation of the models. In the model of only single-grained parent soils, the high degree of explanation supports the measurement results of these soils. At least in limited areas, some measurements of conductivity by TI connected to modeling can be an alternative to intensive and laborious measurement of hydraulic conductivity.


   ACKNOWLEDGMENTS
 
The authors would like to thank Silja Aho, M.Sc., for laboratory assistance and Jukka Pumpanen, M.Sc., for help in the field work. The work was supported by research funds of the University of Helsinki.

Received for publication January 26, 1999.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 Materials and methods
 Results and discussion
 Conclusion
 REFERENCES
 





This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via ISI Web of Science (6)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Mecke, M.
Right arrow Articles by Ilvesniemi, H.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Mecke, M.
Right arrow Articles by Ilvesniemi, H.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Mecke, M.
Right arrow Articles by Ilvesniemi, H.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Crop Science
Vadose Zone Journal Journal of Plant Registrations
Journal of Natural Resources
and Life Sciences Education		 Journal of
Environmental Quality