HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via ISI Web of Science (10) Citing Articles via Google Scholar Google Scholar Articles by Aubertot, J.N. Articles by Richard, G. Search for Related Content PubMed Articles by Aubertot, J.N. Articles by Richard, G. Agricola Articles by Aubertot, J.N. Articles by Richard, G.

DIVISION S-6-SOIL & WATER MANAGEMENT & CONSERVATION

Characterization of Sugar Beet Seedbed Structure

^a INRA, Unité d'Agronomie de Laon-Péronne, rue Fernand Christ, 02007 Laon Cedex, France
^b INRA, Unité d'Agronomie de Laon-Péronne, rue Fernand Christ, 02007 Laon Cedex, France
^c INRA, Unité d'Agronomie de Laon-Péronne, rue Fernand Christ, 02007 Laon Cedex, France
^d INRA, Unité de Biométrie, route de Saint-Cyr, 78026 Versailles Cedex, France

durr{at}laon.inra.fr

	ABSTRACT

TOP ABSTRACT INTRODUCTION Methods Results Discussion REFERENCES

 
Little information is available on the distribution and shape of aggregates or their spatial arrangement within seedbeds, although it influences crop emergence. The aim of this study was to obtain quantitative information on seedbed structure by sampling soil layers, sieving, surface photographs, and stereological analysis of embedded sample sections. Seedbeds prepared with a combined implement (spring tine cultivator and rollers) in a silt loam soil (Typic Hapludalf) with two seeders (K and Pl) on plowed (P), not plowed (NP), and not plowed and compacted (NPC) plots, were analyzed. Aggregates were described by measurements of L, l, h, the longest, intermediate and shortest axes and visually classified according to their surface roughness (smooth, 0; to rough, 4). The D fitting parameter of the power-law aggregate size distribution after seeding was significantly smaller (3.4) on NPC than on the other plots (3.8–4.0). The aggregate aspect ratios (l/L and h/L) did not vary with aggregate size or plot, their overall means were 0.77 and 0.55 respectively. Aggregates of roughness classes >2 were 28 to 32% on NPC and NP plots, compared with 74% on P plot. Seeders reduced the number of aggregates >10 mm by one-third compared with the values before seeding. Seeder K placed over 60% of the aggregates > 20 mm on the soil surface. Aggregates were uniformly distributed across the row by K seeder and away from the center of the row by Pl seeder. The information obtained will be used in a computerized seedbed generator.

Abbreviations: MWD, mean weight diameter • NP, not plowed • NPC, not plowed and compacted • P, plowed

	INTRODUCTION

TOP ABSTRACT INTRODUCTION Methods Results Discussion REFERENCES

 
THIS STUDY was carried out to characterize seedbed structure, particularly for sugar beet (Beta vulgaris L.) crops, to obtain relevant variables that would be both input variables for incorporation into a model predicting seedling emergence (Boiffin et al., 1994) and output variables describing the effects of tillage operations. Few data are available on seedbed structure, defined as the spatial organization, size, and shape of aggregates, although crop establishment can vary widely with these characteristics (Durrant et al., 1988; Dürr et al., 1992). Seedbed structure depends on the soil features (texture, bulk density, and water content), the climate, tillage, and sowing operations. It influences the path the seedling takes to the soil surface. The probability of the seedling failing to emerge depends on the aggregate size and position in the seedbed (Bouaziz and Bruckler, 1989; Souty and Rode, 1993; Boiffin et al., 1994) and roughness (Richard and Dürr, 1997). One reason for the limited amount of information available on seedbed structure is that these variables are not easy to describe. Sieving gives the fraction of the total sample weight in each size range (Kemper and Chepil, 1965). Kritz (1983) measured the mass proportion of aggregates within sublayers in sugar beet seedbeds. Aggregates over 5 mm were more frequent in the upper layer (0–3 cm) than in the sublayers (3–10 cm). Sandri et al. (1998) found correlations between the results obtained by sieving, image analysis of seedbed surface photographs, laser profile metering, and counting the visible clods with a diameter >40 mm. Their aim was to identify parameters with which to quantify seedbed cloddiness, but the techniques used gave no information on the spatial distribution of aggregates in the seedbed. Seedbed spatial organization could also be described by impregnating soil samples with resin. Image analysis of the embedded sample sections has been used to measure the size, shape, and distribution patterns of voids and pedological features (Protz et al., 1987; Bresson and Boiffin, 1990; Dexter, 1991). X-ray tomography has also been used to determine three-dimensional changes in soil bulk density, porosity, water content, solute concentration, macropore size, and fracture width (Steude and Hopkins, 1994). However, these techniques were not used to determine the number or spatial distribution of aggregates.

Several indices have been calculated to summarize the results on soil structure. Some are simply descriptive, like the mean weight diameter (MWD, Van Bavel, 1949) which is expressed as:

assuming that the aggregates are graded into n size fractions, with _i being the mean diameter of each size fraction and w_i, the proportion of the total sample weight occurring in the corresponding size fraction. The MWD increases with the coarseness of the soil samples. Fractal analysis, based on a physical approach of fragmentation (Turcotte, 1986), has been used increasingly in recent years to describe variations in soil structure (Perfect, 1997). Relationships can be calculated between the mass or number of fragments and their size (Turcotte, 1986; Perfect et al., 1992). The number-size distribution can be written as , where N represents the cumulative number of aggregates whose equivalent radii are greater than l, l₀ the unit length, N₀ the number of aggregates greater than the unit length, and D a dimensionless coefficient called the fractal dimension of the distribution. A power-law number-size relationship indicates that the fragmentation process does not vary over a wide range of scales (Turcotte, 1986). The shape of aggregates should also be independent of the scale if soil fragmentation has a fractal behavior. The shapes of aggregates have been described by measuring their longest L, intermediate l (the longest axis in a plane perpendicular to the longest axis L), and shortest h (the longest axis in a plane perpendicular to the L and l axes) principal axis and by calculating the aspect ratios (L/L: l/L: h/L, Dexter, 1985; Perfect et al., 1997). They were independent of aggregate size (Perfect et al., 1997), but were correlated with permanent soil features, like the soil organic and clay contents (Dexter, 1985).

The D values calculated from power-law fitting can be used to compare soil fragmentation under the action of tillage and climate. D increases with the degree of soil fragmentation (Perfect and Blevins, 1997). Another advantage of developing a continuous relationship between the numbers and sizes of aggregates is that it gives a more precise description of the aggregate size distribution than using grades. A closer fit to the real aggregate size distribution, for wide size ranges, can be obtained by a four-parameter distribution (Wagner and Ding, 1994).

Experiments were carried out to characterize the effect of the initial state of the cultivated soil layer before seeding and the influence of the seeder on seedbed structure. The variables for incorporation into a three-dimensional seedbed generator (Boiffin et al., 1994) were aggregate numbers per grade, shape, and spatial arrangement. They were determined by several methods on the same plots: soil sampling with different layers separated and sieved, analysis of soil surface photographs, and stereological analysis of embedded sample sections. The results obtained were compared whenever possible. The relationship between the number and size of aggregates, the MWD, aspect ratios, and aggregate roughness were all determined to compare the fragmentation obtained in the different plots.

	Methods

TOP ABSTRACT INTRODUCTION Methods Results Discussion REFERENCES

 
The experiments were carried out in March through May 1996 at Mons-en-Chaussée (northern France) in a silt loam soil (Typic Hapludalf, Luvisol Orthique, 0.74 g g^-1 silt, 0.20 g g^-1 clay, 0.04 g g^-1 sand, in the 0–0.3 m plowed layer). The preceding crop was wheat with straw removed. The soil before seedbed preparation was either not plowed and compacted (NPC, April 1996), not plowed (NP), or plowed (P, December 1995), providing three different initial states. Soil was compacted under wet conditions (0.22 g g^-1 mean soil water content in the 0–0.3 m soil layer) with a tractor (8.0 Mg) running wheel tracks over wheel tracks, with tires inflated at 200 kPa. The initial bulk density was 1.25 g cm^-3 in the 0-0.3 m soil layer on NP plot and reached 1.55 g cm^-3 on NPC plot. The seedbed was prepared by a single pass of combined implements (spring tine cultivator and skeleton rollers) at a soil water content of 0.19 to 0.21 g g^-1 in the 0 to 0.1 m soil layer. Two seeders (K and Pl), representing the two main types of sugar beet seeders, were compared. Small skeleton rollers, situated behind Seeder K produced almost flat rows, while the two opposed and inclined covering wheels of Seeder Pl produced set-up rows. Each plot (12 rows, 5.4 m wide, 30 m long) was replicated two times. The methods used to describe seedbed structure were compared for two variables (Table 1) : total number of aggregates per volume (sieving analysis compared with stereological analysis of embedded sample sections) and the number of visible aggregates per surface area (sieving analysis compared with image analysis of soil surface photographs).

Two samples were carefully extracted with a spoon from 6 of the 12 rows in each plot replicate. They were air dried and sieved with a gently shaking machine (30 s, 150 shakes/min, 3-mm amplitude). Grades were <2, 2–5, 5–10, 10–20, 20–30, 30–40 and >40 mm. Painted and non painted aggregates were separated, weighed, graded, and counted. Three classes of aggregate burial were assessed visually from the painted proportion of the aggregate surface: mostly buried aggregates (less than 50% of the periphery painted), superficially buried aggregate (more than 50% of the periphery painted), and aggregates laid on the surface (the aggregate surface was almost totally painted).

Subsamples of aggregates (n = 20–80) were taken for grades over 20 mm to measure their principal axes with slide calipers (L, l, h) and the aspect ratios (L/L: l/L: h/L) were calculated. Their surface roughness was visually classified as zero (completely smooth) to four (very rough).

Photographs of the Soil Surface
Photographs of the soil surface were taken after the samples had been delimited for sieving. Aggregates with an L axis over 5 mm were outlined manually on photographs. The aggregates were assigned to a grade according to their L axis. Three zones (20 mm wide) were delimited on photographs to determine the aggregate positions across the row: central, intermediate and external (Fig. 1) . The assignment to a zone depended on the position of the aggregate center of gravity.

View larger version (52K): [in this window] [in a new window]   Fig. 1 Definition of three zones across the row for analysis of soil surface photographs and nine cells for analysis of embedded sample sections

where j corresponds to the grade of objects whose radii are between ^-_j and ⁺_j and which are represented by _j, N_A(j) the number of discs from the Grade j, N_V(j).dy the number of spheres from the Grade j in the elementary volume S.dy. This relationship can be inverted to predict the size distribution of spheres in the volume from the size distribution of discs observed on the sections. Each section was analyzed independently and the results of the 15 sections were averaged. Wicksell's method was tested to determine whether it could be applied to non spherical objects, since aggregates are not spherical. A three-dimensional numerical seedbed was created in which aggregates were represented by ellipsoids. The ellipsoid shapes were obtained using the theoretical values of aspect ratios established by Dexter (1985). The method was used to analyze the same number of virtual samples on the same number of sections (three samples, 15 sections per sample, V = 200 by 80 by 45 mm³). The results indicate that the method can be applied with good results to ellipsoids and that the number of sections are sufficient to describe the size distribution (Table 2) . The analysis predicted that there should be no aggregate of the grade 20 to 30 mm in the analyzed volume though two ellipsoids were placed in each analyzed volume. This happened because the profiles obtained depended on the distance of the sections from the centers of the ellipsoids and their relative orientations. As there were only two ellipsoids of Grade 20 to 30 mm, the probability that the sections gave profiles belonging to that grade was very small and the profiles obtained were assigned to smaller grades. This indicates that the spacing between the sections used may underestimate the number of larger aggregates.

	Results

TOP ABSTRACT INTRODUCTION Methods Results Discussion REFERENCES

for each sample where N_i represents the number of aggregates of Grade i, i ranging from 1, the largest grade, to n, the considered grade, l₀ the unit length, N₀ the number of aggregates greater than the unit length and D a dimensionless coefficient. The length l_i is the lower sieve aperture diameter of Grade i. The correlation coefficients were good (r² always >0.97). D had values over three and increased with the degree of soil fragmentation: D was statistically higher on NP and P than on NPC. Mean weight diameter varied in the opposite manner and the differences between plots were the same (Table 7) .

The surface painting and sieving results were used to compare NP and NPC plots. The proportion of visible aggregates always increased with the aggregate size, for both plots (Fig. 2) . Aggregates over 40 mm were mostly visible. This could result from the spatial arrangement in a given volume and/or tillage action. A simple model can be used to assess the effect of the geometrical arrangement in itself. Assuming that an infinite number of spheres are randomly and uniformly distributed within an infinite volume of height H, the mean proportion of visible spheres can be calculated as follows (Fig. 3) . The reference level is at the surface . The altitude of the center of a sphere is called z_c. Assuming that spheres of radius R are placed between the top and the bottom , the mean proportion of visible spheres over the total number of spheres equals 2R/(H + R). The proportions of visible aggregates observed were always greater than the proportions obtained in the geometrical model (Fig. 2). This indicates that granulometric sorting had occurred, with bigger aggregates being placed at the surface by the seeder.

View larger version (24K): [in this window] [in a new window]   Fig. 2 Fraction of aggregates visible on the soil surface. An asterisk (*) denotes a significant difference (P = 0.05) between the fraction observed on a treatment and a fraction predicted by a geometrical model. The observed value (1.0) is not shown because only one clod was observed

View larger version (9K): [in this window] [in a new window]   Fig. 3 Spheres of radius R randomly and uniformly distributed within a volume of height H between the top and the bottom , the mean proportion of visible spheres on the total number of spheres equals 2 R/(H + R)

View larger version (27K): [in this window] [in a new window]   Fig. 4 Observed (a) and theoretical (b) degrees of aggregate burial, for spheres randomly and uniformly placed sectioning an horizontal plane (not plowed [NP] and not plowed and compacted [NPC] plots). The numbers above the bars are the numbers of observed aggregates for each grade. An asterisk (*) denotes a significant difference (P = 0.05) between the observed degree of aggregate burial and the theoretical value, as determined by a 2 test

View larger version (10K): [in this window] [in a new window]   Fig. 5 Degrees of burial of randomly and uniformly distributed spheres of radius R intersecting the plane z = 0. A sphere is assumed as being laid on the surface when the area of the disk created by its intersection with the plane is less than 20% of the maximum intersection (i.e., R2). A sphere is considered to be laid on the surface if zlim zc R, superficially buried if 0Ê < zc < zlim and deeply buried if -R zc 0

View larger version (48K): [in this window] [in a new window]   Fig. 6 Numbers of aggregates obtained by image analysis and sample sieving (>20 mm, not plowed [NP] and not plowed and compacted [NPC] plots)

View larger version (27K): [in this window] [in a new window]   Fig. 7 Distributions of aggregates across the row for two seeders on the plowed plot (P). Characterization of sugar beet (Beta vulgaris L.) seedbed structure

	Discussion

TOP ABSTRACT INTRODUCTION Methods Results Discussion REFERENCES

The use of several methods raised the problem, as mentioned by Perfect et al. (1997), of comparing the numbers of aggregates when the aggregates are described by sieve apertures or their measured axes. The most frequent way used to represent a given grade (min, max) is the arithmetic mean of the sieve upper and lower apertures (min + max)/2. Wang and Komar (1985), Kozak et al. (1996), Perfect et al. (1997) proposed several methods leading to different representations for a given grade. The intermediate axis of an aggregate appeared to govern the assignment to a grade during sieving. An aggregate can swivel and the assignment to a grade is based on the surface perpendicular to the main axis, assuming that the aggregate has an ellipsoidal shape and that its longest axis is not too great to allow swiveling, i.e., L is less than two times the diameter of the aperture. A grade [min, max] appeared to include aggregates with a mean intermediate axis close to (3min + max)/4. This value gives a simple and more accurate representation of the grade than the arithmetic mean (min + max)/2.

When the length of only one axis is known, usually the longest one, the aspect ratios can be used to reconstruct the aggregate shape and assign it to a grade. The values of aspect ratios measured in this study were close to those given in the literature (Addiscott and Dexter, 1994, Perfect et al., 1997). The aspect ratios were rather constant, especially when comparing several sizes, as reported by Perfect et al. (1997). The observed aspect ratios were similar for a given grade on the different plots: they appeared to be independent of the initial state of the soil layer before fragmentation. The mean values of measured aspect ratios in this study (1:0.77:0.54) were close to those calculated by Dexter (1985)(1: ) but the h/L ratio was significantly smaller. Theoretical values of aspect ratios were calculated assuming that smaller fragments are formed by breaking larger ones in the middle of their longest axis. The theoretical aspect ratios calculated under a cubic hypothesis become: 1: (Perfect et al., 1997). The experimental aspect ratios were close to those values, suggesting that aggregate shape could refer to the reiterative breaking process used in the theoretical calculations.

The power-law relationships observed between number and size of aggregates were consistent with the theoretical assumption of a reiterative fragmentation. However, D values were greater than three and were not consistent with Turcotte's model of fragmentation which imposes 0 D 3 (Kozak et al., 1996; Pachepsky et al., 1997; Young et al., 1997). Turcotte (1986) suggested that the fractal dimension D is a measure of the fracture resistance of the material to the process causing fragmentation. D appeared to have decreasing values with increasing initial soil compaction. Perfect and Blevins (1997) had consistent results: they indicated that D increased with increased tillage intensity.

Aggregate roughness varied more than the aggregate aspect ratios and appeared to be also correlated with the degree of compaction of the plot before planting. Aggregate facets could be identified with fracture faces: thus they could be influenced by the initial state of compaction, being smoother when the soil was initially more resistant to fragmentation and perhaps by the way fragmentation was obtained (tillage, climate and/or roots actions).

The methods used in this study provided precise information on the variations in seedbed structures and on the changes caused by soil initial state and implements on aggregate shape, number, and spatial distribution. The seedbed was coarser when the soil was not plowed even if the seeder reduced the number of larger aggregates. Aggregates were also smoother when the soil layer was not plowed. Aggregates were not uniformly distributed within the seedbed. Larger aggregates were mostly visible on the soil surface because of the sorting action of the tillage implements and seeder. The distribution across the row also depended on the seeder, especially on its last implement: the small skeleton rollers divided up aggregates all across to the row, whereas the V wheel ejected them to the edges of the row or crushed those in the row. The degree to which visible aggregates were buried was, in contrast, consistent with a simple geometrical model, the implements having no influence.

The variables described in this study influence seedling emergence and must be taken into account when predicting crop establishment. However, seedbed structure has complex effects on emergence:aggregate roughness, size, and position vary together, they influence the sowing depth distribution, and their effects depend on the climatic conditions following sowing. These effects are difficult to study in conventional field experiments. Standard recommendations for sowing operations are not possible because of the number of situations the growers have to cope with. A simulation tool that would allow numerical experiments would be useful to help in decision making. Information on seedbed structure (numbers, shape, and roughness of aggregates in each grade, and their position) will be used to generate numerical seedbeds and study the effects of soil structure on seedling emergence for a wide range of sowing conditions.

	ACKNOWLEDGMENTS

 
The authors thank the Institut Technique de la Betterave and the Biopôle Végétal Picard for financial support. They thank L.M. Bresson and P. Guilloret, A. Kretzchmar and A. Pierret, J.C. Fiès for their help with sample embedding and sawing methods. C. Leforestier, P. Regnier and C. Dominiarzick provided technical assistance and Owen Parkes checked the English text.

Received for publication August 6, 1998.

	REFERENCES

TOP ABSTRACT INTRODUCTION Methods Results Discussion REFERENCES

 

• Addiscott T.M., Dexter A.R. Tillage and crop residue management effects on losses of chemicals from soils. Soil Tillage Res. 1994;30:125-168.
• Boiffin, J., P. Dubrulle, C. Dürr, G. Richard, and Y. Duval. 1994. Modelling sugarbeet seedling emergence and early growth. p. 1143–1148. In Proc. International Soil Tillage Research Organization (ISTRO) Conf., 13th, Aalborg, Denmark. 24-29 July 1994. ISTRO, Wageningen, the Netherlands.
• Bouaziz A., Bruckler L. Modeling wheat seedling growth and emergence: II. Comparison with field experiments. Soil Sci. Soc. Am. J. 1989;53:1838-1846.[Abstract/Free Full Text]
• Bresson L.M., Boiffin J. Morphological characterization of soil crust development stages on an experimental field. Geoderma 1990;47:305-325.
• Dexter A.R. Shapes of aggregates from tilled layers of some dutch and australian soils. Geoderma 1985;35:91-107.
• Dexter, A.R. 1991. Statistical structure of tilled soil as determined by image analysis of sections. Div. Note DN 1609, AFRC Inst. Eng. Res., Agriculture and Food Research Center, Silsoe, UK.
• Dürr C., Boiffin J., Fleury A., Coulomb I. Analysis of the variability of sugar beet (Beta vulgaris L.) growth during the early stages. II. Factors influencing seedling size in field conditions. Agronomie 1992;12:527-535.
• Durrant M.J., Dunning R.A., Jaggard K.W., Bugg R.B., Scott R.K. A census of seedling establishment in sugar-beet crops. Ann. Appl. Biol. 1988;113:327-345.
• Kemper W.D., Chepil W.S. Size distribution of aggregates in methods of soil analysis. In: Black C.A., et al. , ed. Methods of soil analysis. Madison, WI: ASA, 1965:499-510.
• Kozak E., Pachepsky Ya.A., Sokolowski S., Sokolowska Z., Stepniewski W. A modified number-based method for estimating fragmentation fractal dimensions of soils. Soil Sci. Soc. Am. J. 1996;60:1291-1297.[Abstract/Free Full Text]
• Kritz, G. 1983. Sbäddar för vrstrsäd. En stickprovsundersökning. Division of Soil Management, Report No. 65, Swedish Univ. Agric. Sci., Uppsala.
• Pachepsky Ya., Giménez D., Logsdon S., Allmaras R.R., Kozak E. On interpretation and misinterpretation of fractal models: A reply to "Comment on number-size distributions, soil structure, and fractals.". Soil Sci. Soc. Am. J. 1997;61:1800-1801.[Free Full Text]
• Perfect E. Fractal models for the fragmentation of rocks and soils: a review. Eng. Geol. 1997;48:185-198.
• Perfect E., Rasiah V., Kay B.D. Fractal dimensions of soil aggregate-size distributions calculated by number and mass. Soil Sci. Soc. Am. J. 1992;56:1407-1409.[Abstract/Free Full Text]
• Perfect E., Blevins R.L. Fractal characterization of soil aggregation and fragmentation as influenced by tillage treatment. Soil Sci. Soc. Am. J. 1997;61:896-900.[Abstract/Free Full Text]
• Perfect E., Zhai Q., Blevins R.L. Soil and tillage effects on the characteristic size and shape of aggregates. Soil Sci. Soc. Am. J. 1997;61:1459-1465.[Abstract/Free Full Text]
• Protz R., Shipitalo M.J., Mermut A.R., Fox C.A. Image analysis of soils — present and future. Geoderma 1987;40:115-125.
• Richard G., Dürr C. Conséquence de l'état du profil cultural sur les peuplements végétaux: réussite de la levée en relation avec l'état du lit de semences. In: Pirot R., et al. , ed. Le travail du sol dans les sytèmes mécanisés tropicaux, Montpellier. 11–12 Sept. 1996. Montpellier, France: CIRAD, 1997:49-54.
• Sandri R., Anken T., Hilfiker T., Sartori L., Bollhader H. Comparison of methods for determining cloddiness in seedbed preparation. Soil Tillage Res. 1998;45:57-64.
• Souty N., Rode C. Emergence of sugar beet seedlings from under different obstacles. Eur. J. Agron. 1993;2(3):213-221.
• Steude, J.S., and F. Hopkins. 1994. Industrial X-ray computed tomography applied to soil research. In Tomography of soil-water-root processes. SSSA Spec. Publ. 36. SSSA, Madison, WI.
• Turcotte D.L. Fractals and fragmentation. J. Geophys. Res. 1986;91:1921-1926.
• Van Bavel C.H.M. Mean weight diameter of soil aggregates as a statistical index of aggregation. Soil Sci. Soc. Am. Proc. 1949;14:20-23.
• Wagner L.E., Ding D. Representing aggregate size distributions as modified lognormal distributions. Trans. ASAE 1994;37:815-821.
• Wang C., Komar P.D. The sieving of heavy mineral sands. J. Sediment. Petrol. 1985;55:479-482.[Abstract/Free Full Text]
• Wicksell S.D. The corpuscule problem. A mathematical study of biometric problem. Biometrika 1925;17:84.[Free Full Text]
• Young I.M., Crawford J.W., Anderson A., McBratney A. Comment on number-size distributions, soil structure and fractals. Soil Sci. Soc. Am. J. 1997;61:1799-1800.[Free Full Text]

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via ISI Web of Science (10) Citing Articles via Google Scholar Google Scholar Articles by Aubertot, J.N. Articles by Richard, G. Search for Related Content PubMed Articles by Aubertot, J.N. Articles by Richard, G. Agricola Articles by Aubertot, J.N. Articles by Richard, G.