Published online May 27, 2005

Soil Fragment Size Distribution and Compactive Effort Effects on Maize Root Seedling Elongation in Moist Soil M. Dı´az-Zorita,* J. H. Grove, and E. Perfect

Reproduced from Crop Science. Published by Crop Science Society of America. All copyrights reserved.

ABSTRACT

M. Dı´az-Zorita, CONICET and Dep. of Plant Production, Faculty of Agronomy, Univ. of Buenos Aires, (1417) Av. San Martı´n 4457, Buenos Aires, Argentina, and Nitragin Argentina S.A., Calle 10 y 11, Parque Industrial Pilar 1629, Pilar, Buenos Aires, Argentina; J.H. Grove, Dep. of Agronomy, Univ. of Kentucky, Lexington, KY 405460091; E. Perfect, Dep. of Earth and Planetary Sciences, Univ. of Tennessee, Knoxville, TN 37996-1410. Received 14 Dec. 2003. *Corresponding author ([email protected]).

according to the degree of soil fragmentation after applying a disruptive stress. Model parameters are then derived from the resulting mass-size distribution of soil fragments or aggregates. The distribution of soil structural units controls the availability of oxygen, water, and the resistance to penetration by shoots and roots in seedbeds created by tillage (Hadas and Russo, 1974; Taylor, 1974; Schneider and Gupta, 1985; Nasr and Selles, 1995). Braunack (1995) described an earlier and greater emergence of maize or soybean [Glycine max (L.) Merr.] seedlings when planted in fine (aggregate size between 1 and 2 mm) vs. coarse (aggregate size between 5 and 15 mm) seedbeds. Soybean emergence was delayed in the presence of structural units ⬍ 1 mm and ⬎ 4 mm (Nash and Baligar, 1974). In the same experiment, these authors concluded that optimal root growth occurs in the presence of soil structural units ⬎ 0.5 mm. Seed-soil contact, imbibition of water, and germination all depend on the size and packing of soil aggregates (Brown et al., 1996). Haque et al. (1992) observed a greater emergence rate of rice (Oryza sativa L.) in seedbeds with coarse aggregates but concluded that complete crop establishment was not affected by soil structure. Wuest et al. (1999) concluded that vapor pressure, rather than seed–soil contact, may be the most important factor for imbibition and wheat (Triticum aestivum L.) seed germination. Measurements of fragment size distribution are most relevant to the germination and early growth of plants in soils that are tilled, structured, and uncompacted by traffic (Kay and Angers, 1999). Root growth pathways, based on an experiment conducted with artificially constructed aggregates, differ depending on aggregate density and strength (de Freitas et al., 1999). Voorhees et al. (1971) observed that root growth is restricted to the periphery of aggregates having densities of 1.8 Mg m⫺3 or greater. The proportion of the total root length penetrating aggregates decreases with increasing size and strength of the aggregates (Misra et al., 1988). Dexter (1978) concluded that optimal soil structure for maximum root growth depends on the strength of soil aggregates. Increasing the spread in aggregate strength resulted in an increase in nutrient availability proportional mostly to the presence of small aggregates (Hewitt and Dexter, 1979). Furthermore, the effects of aggregation and water availability on root growth cannot be clearly separated because these two soil properties are interrelated (Vepraskas and Wagger, 1990).

Published in Crop Sci. 45:1417–1426 (2005). Crop Ecology, Management & Quality doi:10.2135/cropsci2003.0670 © Crop Science Society of America 677 S. Segoe Rd., Madison, WI 53711 USA

Abbreviations: ␳b, soil bulk density; ␳br, relative bulk density; AFP, air-filled porosity; CE, compactive effort; GMD, geometric mean diameter; GWC, gravimetric soil water content; LogGSD, log of the geometric standard deviation; Pi, intrafragment porosity; RLD, root length density; TOC, total organic carbon.

Distributions of soil fragments (a mixture of primary aggregates resulting from tillage fragmentation) in seedbeds are known to influence emergence and early shoot and root growth of crops. However, it is not clearly understood which distribution model parameters the roots are responding to when water imbibition and nutrient availability are not limiting factors. The objective of this study was to determine the effect(s) of variation in geometric mean diameter (GMD) and log of the geometric standard deviation (LogGSD) taken from a lognormal model of soil fragment size distribution on maize (Zea mays L.) root elongation over a range of soil bulk density (␳b) and air-filled porosity (AFP) levels. Root growth, determined 48 h after seedling emergence, was evaluated in a greenhouse experiment with artificially packed soil fragments sieved from a Maury silt loam (fine, mixed, semiactive, mesic Typic Paleudalf) under sod. Two experiments were conducted. The first consisted of a complete factorial combination of four compactive efforts (CEs) (0.0, 26.8, 53.5, and 107.0 kJ m⫺3) applied over four GMDs (4.3, 5.1, 6.8, and 9.0 mm) at a uniform LogGSD of 0.22. The second experiment was a complete factorial combination of the same four CE treatments applied over three LogGSD (0.22, 0.31, and 0.48) values at a uniform GMD of 5.1 mm. Increasing CE, GMD, or LogGSD caused the ␳b to increase. Maximum root elongation occurred at intermediate (5.1–6.8 mm) GMD values when 26.8 kJ m⫺3 of CE was applied, corresponding to an average ␳b of 1.15 Mg m⫺3. There was no direct effect of the spread in soil fragment size (LogGSD) on root elongation. Total root length density (RLD) showed a quadratic response to ␳b, relative bulk density (␳br), bulk density divided by maximum Proctor density), or AFP levels reaching maximum elongation at values of 1.12 Mg m⫺3, 0.78 m3 m⫺3, or 0.187 m3 m⫺3, respectively. These results suggest that maximum radicle elongation depends more on the size of soil fragments rather than on the spread in their size distribution. Bulk density seems to be a more relevant parameter than size or distribution of soil fragments in characterizing compacted seedbeds. Loose or highly compacted seedbeds are inadequate for maximal early growth of maize roots.

T

he spatial heterogeneity of soil components, that is, soil structure, is a complex condition related to many biogeochemical processes (Dexter, 1988; Kay, 1990; Dexter, 1997; Angers and Caron, 1998; Kay and Angers, 1999). Although there is no universally accepted way to measure soil structure, it is commonly characterized

1417

Reproduced from Crop Science. Published by Crop Science Society of America. All copyrights reserved.

1418

CROP SCIENCE, VOL. 45, JULY–AUGUST 2005

Alexander and Miller (1991), as well as others (Taylor, 1974; Logsdon et al., 1987; Donald et al., 1987; Misra et al., 1988) observed decreasing root dry matter production and length with increasing aggregate size when water and/or nutrients were not limiting. In contrast, Nash and Baligar (1974) observed that optimal soybean root growth occurred in the presence of peds larger than 0.5 mm, with increasing vertical root penetration as aggregate size increased. The pore system associated with a particular arrangement of soil aggregates or fragments has major effects on aeration, available water, and soil strength, consequently on root and shoot growth, and perhaps ultimately on crop productivity. Several indices have been proposed to qualitatively describe the structural state of the soil. Results from Scandinavia (Hakansson and Lipiec, 2000) and eastern Canada (Carter, 1990) suggest that crop yields seem to follow a curvilinear relationship with respect to relative compaction levels. Normalizing ␳b with respect to a reference ␳b value, for example, maximum value under a standard compaction treatment, eliminated the influence of clay and organic matter contents and their interaction on ␳b (Da Silva et al., 1997). The ratio between actual ␳b and the reference ␳b reflects the ability of soil’s structure to control the availability of air and water to plants as well as the soil’s resistance to penetration by roots. The log-normal distribution model is commonly used to describe soil fragment size distribution and, indirectly, soil structure. However, research describing the independent effects of the log-normal distribution parameters, GMD and LogGSD, on root growth is limited. Most soil structure research related to root growth has focused on the effect of fragment sizes without relevant discussion of fragment distribution (LogGSD) effects on root growth. Furthermore, only a few studies suggest that the packing of soil fragments, not necessarily their size distribution, has relevant effects on seedling emergence (Bouaziz and Bruckler, 1989; Souty and Rode, 1993). The reduction in soil void space after the application of mechanical stress has major and mostly negative effects on root growth and has been reviewed extensively (Russell, 1977; Bennie, 1996). Soil aggregation is an important factor determining soil compressibility, that is, soil resistance to volume decrease or ␳b increase when the soil is subjected to a mechanical load (Horn and Lebert, 1994). In general, mechanical impedance to root growth in structured soils is mostly due to pore width between aggregates and the size of the individual aggregates (Bennie, 1996). However, few studies have reported on the combined effects of soil fragment size distribution and compaction on root growth. The adaptation of root systems to soil characteristics is complex, involving the interaction of multiple factors such as water and/or nutrient availability and absorption, aeration, and temperature. Until the four-leaf stage (four leaves visible), most of the resources for seedling growth are supplied by the reserves in maize seed (Tollenaar and Dwyer, 1999). Seedling growth depends mainly on water imbibition, temperature, soil aeration,

and the reserves stored in the kernel. Seedling emergence is independent of the nutrient status of the soil, and depends totally on seed reserves until the seedling becomes autothropic. Soil fragment size distribution has been introduced as a major factor controlling soil porosity and related soil processes in the seedbeds (Braunack and Dexter, 1989a, 1989b; Kay and Angers, 1999). Whiteley and Dexter (1984) concluded that root displacement of soil fragments ⬎ 1 mm in diam. rarely occurs. Thus, it is assumed that for seedbeds formed by soil fragments ⬎ 1 mm in diam., differences in early root elongation from similar sized seeds will be primarily related to differences in soil fragment size distribution and porosity of the seedbeds. The objective of this study was to determine the effects of soil fragment size distribution parameters (GMD and LogGSD) on maize root elongation over a range of total porosity levels created by the application of a range of CEs under nonlimiting water and nutrient conditions. MATERIALS AND METHODS The soil used in this experiment was collected from the top 100-mm layer of a Maury silt loam under fescue (Festuca arundinacea L.) sod near Lexington, KY. Field-moist soil was sieved into four soil fragment size fractions (Table 1) using a vibratory Gilson testing screen model TS-1 for a sieving duration of 30 s and then air dried for 30 d. The soil water content at sampling was of 203.0 g kg⫺1 and after air-drying was 14.6 g kg⫺1. The particle size distribution (clay, silt, and sand) was determined on air-dried samples sieved through a 2-mm screen using the pipette method (Gee and Bauder, 1986). Dry combustion with a TOC-5000A total organic carbon (TOC) analyzer (Shimadzu Corporation, Kyoto, Japan) was used for measuring TOC contents. Apparent porosity of individual air-dried fragments, or intrafragment porosity (Pi), was determined for each of the size fractions using kerosene as the saturating fluid (Henin et al., 1969). The fragments were submerged in kerosene for 24 h, removed, and placed on filter paper (Whatman no. 42) to drain. After approximately 10 min, when the surface of the soil fragments no longer appeared shiny, the fragments were weighed and suspended in kerosene and weighed again. The intrafragment porosity was calculated assuming the density of kerosene to be 0.783 Mg m⫺3 (Donald et al., 1987). The particle density of each fragment size fraction was determined using the pycnometer method (Blake and Hartge, 1986). Seven fragment size distribution treatments were prepared by combining the four air-dried fragment size fractions in different proportions. Four treatments consisted of GMD values between 4.3 and 9.0 mm with a constant LogGSD value of 0.22. The other three treatments consisted of LogGSD Table 1. Mean soil physical properties of the different fragment size fractions. TOC ⫽ total organic carbon, PD ⫽ particle density, Pi ⫽ intrafragment porosity. Fragment size mm 12.50–9.50 9.50–6.30 6.30–3.35 3.35–2.00

TOC

Clay

Silt

10.8 13.4 15.8 15.1

g kg⫺1 142 770 120 785 89 800 47 833

Sand

Particle density

Pi

88 95 111 120

Mg m⫺3 2.57 2.51 2.42 2.39

m 3 m ⫺3 0.35 0.34 0.35 0.33

1419

Reproduced from Crop Science. Published by Crop Science Society of America. All copyrights reserved.

DI´AZ-ZORITA ET AL.: EFFECTS ON ROOT ELONGATION IN MOIST SOIL

values of 0.22, 0.31, and 0.48 with a constant GMD of 5.1 mm. The properties of these treatments are presented in Table 2. Soil fragments, in weight proportions corresponding to each of the treatments, were mixed in a plastic bag such that the final quantity of the mixture was 120 g. The mixing was accomplished by pouring the fragments, sequentially from largest to smallest sizes, into a polyethylene container (58 mm in length and 90 and 71 mm in upper and lower diameters, respectively) and dropping the container 50 times from a height of approximately 125 mm. Then the fragment mixtures were poured into cylindrical aluminum pots (60 mm deep with a 54-mm diam.), the bottom was covered with cheesecloth and packed by dropping the container 50 times from a height of approximately 125 mm. Preliminary measurements suggested that with 50 drops, the segregation of the fragments packed in these containers was minimal (data not shown). Pots were placed on trays at room temperature (approximately 20⬚C) and wetted for 72 h from below by capillarity using a 1-cm-deep water table. Then, pots were covered with parafilm and placed onto paper towels at room temperature (approximately 20⬚C) until drainage stopped (approximately 96 h). Four compaction treatments were imposed on each set of fragment size distribution treatments by dropping the rammer used for the Proctor test procedure (mass ⫽ 2.50 kg and diam. ⫽ 51 mm, American Society for Testing Materials, 2000) onto the top of the pots from 0-, 150-, 300-, and 600-mm heights. The rammer was placed over the pots with the walls of the driving cylinder setting on the walls of the pots. The applied compactive efforts (CE) were 0.0, 26.8, 53.8, and 107.6 kJ m⫺3. Similar-sized (10 ⫾ 1 mm) maize cultivar Pioneer 33G26 seeds were germinated in a growth chamber at an 8-h daytime temperature of 30⬚C and a 16-h nighttime temperature of 20⬚C for 32 h. Two germinated seeds, with radicles approximately 1 mm long, were planted at a depth of 10 mm below the surface of the soil in each of three pots for each of the 28 treatments; that is, 7 fragment size distributions by four CEs, and the pots were arranged in three blocks within a growth chamber maintained at 28⬚C with a 16-h photoperiod. Pots were watered by capillary rise with the pots placed on a tray with a 10-mm-deep water table. The resulting total water potential varied from 0.10 kPa at the bottom of the pots and, in agreement with soil height differences after compaction, ⫺0.35 to ⫺0.50 kPa at the top of the pots for highly compacted to uncompacted treatments, respectively.

on a 1-mm screen until all soil was washed through the screen. The roots and seeds were placed in plastic bags and refrigerated at approximately 8⬚C for 12 h before further analysis. Roots were characterized by measuring length and average diameter of the radicle and lateral seminal roots (Arsenault et al., 1995). The measurements were made with an STD-1600 EPSON scanner and the images analyzed with the WinRhizo v. 4.0b software (Regent Instruments Inc., Quebec, Canada). Root length density was calculated by dividing the radicle, lateral, or total (radicle ⫹ laterals) root length by the total soil volume (soil volume ⫽ 120 g per dry ␳b).

Soil Characterization In each pot, dry soil ␳b was estimated from the ratio between soil air-dry mass and soil volume (height of the soil measured with a ruler ⫻ the circular surface area of the pots) determined after compaction. The mass of dry soil in each pot was known, and no soil was removed during compaction. In the case of the uncompacted treatments, the soil volume was equal to the volume of the pots. In these treatments, some fragments were observed to lie above the surface of the pot, leading to overestimation of ␳b. Maximum ␳b of the treatments was measured following the procedure described by Dı´az-Zorita et al. (2001). Basically, 120 g of air dry soil corresponding to each fragment size distribution treatment was placed in a plastic bag and sprayed with tap water to give five different water contents. Then, the samples were manually mixed, avoiding disruption of the aggregates, and the bags were closed and left to equilibrate at room temperature (approximately 20⬚C) for 24 h. The moist soil was carefully transferred into aluminum cylinders, the same as described previously, and 545 kJ m⫺3 of CE was applied by dropping a Proctor rammer 10 times from a height of 305 mm. After compaction, the moist soil was weighed, oven dried (105⬚C, 24 h), and then weighed again to determine the gravimetric soil water content (GWC) and ␳b. The maximum bulk density (␳bmax) was estimated after fitting a quadratic model to the relationship between ␳b and GWC (the laboratory soil compaction curve) using procedure NLIN of PC-SAS (SAS Institute, 1997). From the ratio between ␳b and ␳bmax, ␳br was calculated. Mean AFP was estimated from Eq. [1],

AFP ⫽ TP ⫺ VWC

[1]

where TP is total porosity estimated from Eq. [2] (e.g., Hillel, 1998),

Root Characterization After 48 h, the soil within the pots was carefully removed, exposing the two seeds and roots. The duration of the growth period was selected to minimize direct interactions between the elongating radicle and the wall of the pot, and considering that radicle extension rates can vary between 1 and 4 cm d⫺1 (Feldman, 1994). Roots were separated from soil by gently spraying water

TP ⫽ 1 ⫺ ␳b/PD

[2]

and PD is the treatment particle density estimated from the particle density of individual fragment size fractions used for each treatment. The volumetric water content (VWC) of the topsoil was estimated from

Table 2. Mean soil properties for the different fragment size distribution treatments. GMD ⫽ geometric mean diameter, LogGSD ⫽ log of geometric standard deviation, TOC ⫽ total organic carbon, ␳b ⫽ packed , ␳bmax ⫽ maximum, AFP ⫽ air-filled porosity, Pe ⫽ interfragment porosity estimated from weighted mean of fragment size fractions from Table 1. Treatment A B C D E F G

GMD mm 9.0 6.7 5.1 4.3 5.1 5.1 5.1

LogGSD

Clay

TOC

␳b

g kg⫺1 0.22 0.22 0.22 0.22 0.48 0.31 0.22

130.6 111.8 92.9 84.8 100.0 96.5 92.9

␳bmax

AFP

Mg m⫺3 11.8 13.9 15.3 15.5 13.5 14.6 15.3

0.88 0.88 0.87 0.88 0.87 0.87 0.87

Pe m3 m⫺3

1.47 1.45 1.40 1.38 1.41 1.39 1.40

0.335 0.339 0.321 0.302 0.318 0.314 0.321

0.309 0.308 0.308 0.310 0.313 0.314 0.308

Reproduced from Crop Science. Published by Crop Science Society of America. All copyrights reserved.

1420

CROP SCIENCE, VOL. 45, JULY–AUGUST 2005

Fig. 1. Interaction of compactive effort and (a) geometric mean diameter (GMD) or (b) log geometric standard deviation (LogGSD) on of a Maury silt loam. Bars are standard error of the least square mean.

VWC ⫽ GWC ⫻ ␳b

[3]

where GWC was measured when removing the roots from the pots before spraying them with water. The Pi of each treatment was estimated from individual fragment porosity values used for each treatment. The interfragment porosity (Pe) was estimated from the difference between the TP and the weighed Pi values.

Statistical Analysis Data was analyzed according to a completely randomized block design. In each experiment (i.e., constant GMD or constant LogGSD), ANOVA and LSD tests were performed on data sets using the GLM procedure of PC-SAS (SAS Institute, 1997). The normal distribution of the error was evaluated

Fig. 2. Interaction of compactive effort and (a) geometric mean diameter (GMD) or (b) log geometric standard deviation (LogGSD) on relative bulk density of a Maury silt loam. Bars are standard error of the least square mean.

for each of the root parameters by the Shapiro-Wilk test for normality, calculated by the UNIVARIATE procedure of PCSAS (SAS Institute, 1997). Because root length and diameter were the only variates not normally distributed, log transformation of these variables was used for further analysis. Regression and correlation analysis were also performed using PCSAS procedures (SAS Institute, 1997).

RESULTS Fragment Size Distribution Effect on Soil Application of mechanical stress to the dry fragment size distributions caused ␳b (0.87–1.41 Mg m⫺3) and ␳br (0.60–0.99 Mg m⫺3) to vary, with significant interactions between treatments. In each GMD or LogGSD treat-

Reproduced from Crop Science. Published by Crop Science Society of America. All copyrights reserved.

DI´AZ-ZORITA ET AL.: EFFECTS ON ROOT ELONGATION IN MOIST SOIL

Fig. 3. Interaction of compactive effort and (a) geometric mean diameter (GMD) or (b) log geometric standard deviation (LogGSD) on interfragment porosity of a Maury silt loam. Bars are standard error of the least square mean.

ment, increasing the mechanical stress input caused either ␳b or ␳br to increase according to a diminishing returns behavior (Fig. 1 and 2). Both properties were positively and strongly related (␳br ⫽ 0.02 ⫹ 0.69 ⫻ ␳b, r 2 ⫽ 0.98, n ⫽ 6, P ⬍ 0.01). This behavior was expected because of the relatively narrow range in ␳bmax among treatments (Table 2). Maximum ␳b was positively related to GMD (␳bmax ⫽ 1.30 ⫹ 0.0194 GMD (mm), r 2 ⫽ 0.94, n ⫽ 4, P ⬍ 0.01). There was no significant relationship between LogGSD and ␳bmax. Increasing clay content or decreasing soil organic carbon was associated with greater ␳bmax values, ␳bmax ⫽ 1.21 ⫹ 0.0021 Clay (g kg⫺1), r 2 ⫽ 0.94, n ⫽ 6, P ⬍ 0.01

1421

Fig. 4. Interaction of compactive effort and (a) geometric mean diameter (GMD) or (b) log geometric standard deviation (LogGSD) on air-filled porosity of a Maury silt loam. Bars are standard error of the least square mean.

␳bmax ⫽ 1.72 ⫺ 0.021 TOC (g kg⫺1), r 2 ⫽ 0.72, n ⫽ 6, P ⬍ 0.01 Horn and Lebert (1994) also observed that in agricultural soils with similar ␳b and water contents are more compressible the higher the clay content and the lower the soil organic matter level. This behavior can partially be explained by changes in fragment strength in relation to clay content. The high the clay content, the weaker soil fragments become (Horn and Lebert, 1994). In general, high ␳b values were found in the compacted treatments with greater GMD values when LogGSD was held constant (Fig. 1a). Seedbeds with the greatest LogGSD values (0.48) showed a significant difference in ␳b levels compared with the other LogGSD treatments

1422

CROP SCIENCE, VOL. 45, JULY–AUGUST 2005

Table 3. Results of ANOVAs for several maize seminal root elongation parameters in a Maury silt loam under four compaction effort (CE) and four geometric mean diameter (GMD) or three log geometric standard deviation (LogGSD) treatments. L ⫽ length, LD ⫽ length density.† F values Radicle

Reproduced from Crop Science. Published by Crop Science Society of America. All copyrights reserved.

Source GMD CE GMD ⫻ CE LogGSD CE LogGSD ⫻ CE

Total roots

L

LD

L

2.3‡ 3.8* 1.0 1.3 7.5* 1.4

2.3‡ 3.4* 1.2 1.1 3.0‡ 1.9

2.2‡ 4.1* 0.6 0.8 6.9* 0.6

* Significant at the 0.05 level of probability. † Nonsignificant factors: Radicle diameter, lateral roots diameter, length and length density, and total root length density. ‡ Significant at the 0.10 level of probability.

only after the application of 53.5 kJ m⫺3 of CE (Fig. 1b). No significant differences in ␳b values due to fragment size distribution parameters were observed at zero compactive effort. This effect was expected because all fragment size distribution mixtures were packed into pots of equal volume. The intrafragment porosity includes pores of smaller diameter than that of the interfragment pore system (Horn, 1990) and remains mostly unaffected by compaction (Hillel, 1998). Thus, assuming similar intrafragment porosity among the CE treatments, greater changes in porosity were related to changes in interfragment porosity or macroporosity (Fig. 3). Interfragment porosity was similar in uncompacted seedbeds (Table 2), suggesting that the change in macroporosity was mostly the consequence of mechanical disruption of soil fragments rather than differences in the initial porosity state of the treatments. Seedbeds formed mostly by large soil fragments have been described as having low strength and small resistance to crushing compared with those formed by small sized fragments. Misra et al. (1988) found that the larger the soil aggregates, the greater the size of the voids. The application of compactive stress caused greater disruption and void reduction between large-sized fragments. Horn et al. (1995), reporting on several experiments with different soil types, found that aggregate tensile strength decreased with increasing aggregate size in response to greater sizes of the pores. In this study, seedbeds with equal GMD values exhibited a significant reduction in interfragment porosity that occurred after the application of 26.8 and 53.7 kJ m⫺3 of CE to the treatment with a LogGSD of 0.48 as compared with a LogGSD of 0.31 or 0.22 (Fig. 3b). Under a relatively uniform distribution of soil fragments (similar LogGSD) and without compaction, decreasing fragment GMD caused the ␳br values to significantly increase (Fig. 2a). After the application of a CE of 26.8 kJ m⫺3, significantly greater ␳br values were generally observed in the treatments with GMD values of 9.0 or 6.8 mm than in those formed mostly by small soil fragments (Fig. 2a). This behavior confirms the occurrence of lower interfragment porosity in the presence of small-sized soil fragments and the potential overesti-

Fig. 5. The effect of soil fragment geometric mean diameter on maize seedling root elongation parameters. Values are averaged over the four compactive efforts. Bars are standard error of the least square mean.

mation of ␳b values after packing described in the Materials and Methods section. When comparing seedbeds of equal GMD, but unequal LogGSD, significantly greater ␳br values resulted from the application of 26.8 or 53.9 kJ m⫺3 CE to the treatment with a LogGSD value of 0.48 than in the treatments with LogGSD values of 0.22 or 0.31 (Fig. 2b). In both experiments, regardless of GMD or LogGSD treatment, there were no significant differences in ␳br values after the application of 107.0 kJ m⫺3 (Fig. 2), suggesting that at this level of mechanical disruption most of the fragments had compressed. Air-filled porosity values ranged between 0.02 and 0.36 m3 m⫺3, decreasing with increasing mechanical

Reproduced from Crop Science. Published by Crop Science Society of America. All copyrights reserved.

DI´AZ-ZORITA ET AL.: EFFECTS ON ROOT ELONGATION IN MOIST SOIL

Fig. 6. Compactive effort effect on maize seedling root elongation parameters. Values averaged over the four fragment geometric mean diameter treatments. Bars are standard error of the least square mean.

stress input (Fig. 4). After the application of a CE of 26.8 and 53.5 kJ m⫺3, lower AFP occurred in treatments with a GMD of 9.0 mm (Fig. 4a). These results are in agreement with the general observation of smaller changes in macroporosity with mechanical load to seedbeds with small GMD as compared with seedbeds formed by larger fragments.

Compactive Effort and Fragment Size Distribution Effects on Root Elongation The fragment size distribution and CE treatments had only small effects on root elongation. Radicle diameter (1.08–2.04 mm), as well as the diameter (0.64–1.39 mm), length (8–107 mm), and length density (0.009–0.107 cm

1423

Fig. 7. Compactive effort effect on maize seedling root elongation parameters. Values averaged over the three log geometric standard deviation treatments. Bars are standard error of the least square mean.

cm⫺3) of lateral roots and the total RLD (0.021–0.152 cm cm⫺3) were not significant in the ANOVAs at the 0.05 significance level (data not shown). There were significant effects for other root parameters (Table 3). For most of those parameters, there were no significant (P ⬎ 0.10) interactions between fragment size distribution and CE treatments. The length (19–68 mm), the length density (0.021– 0.062 cm cm⫺3) of maize radicles and the total root length (19–167 mm) were significantly influenced by the GMD, when the LogGSD value was held constant (Table 3). No significant effect of the LogGSD, when the GMD value was held constant, on root elongation parameters was observed (Table 3). Maximum radicle length and radicle length density, when averaged across the four CE treatments, occurred in seedbeds with GMD values of 5.1 and 6.8 mm (Fig. 5).

1424

CROP SCIENCE, VOL. 45, JULY–AUGUST 2005

Reproduced from Crop Science. Published by Crop Science Society of America. All copyrights reserved.

Maximum total root length occurred when the GMD was 5.1 mm (Fig. 5). The application of a mechanical stress, independent of GMD or LogGSD values, caused significant differences in the length and length density of the radicle and the total root length (Table 3). Maximum radicle and total root length occurred at the 0.0 or the 26.8 kJ m⫺3 CE treatments, while the greatest radicle length density was measured after the application of 26.8 kJ m⫺3 of mechanical stress (Fig. 6 and 7).

DISCUSSION

Fig. 8. Bulk density, air-filled porosity, and relative bulk density effects on total root length density of maize seedlings growing in the 28 treatments (seven dry fragment size distribution treatments by four compactive effort treatments).

Maize root elongation responded to dry fragment GMD as well as to differences in total porosity and AFP created by the input of mechanical stress. The spread in fragment sizes, as described by the LogGSD, did not significantly affect maize root elongation. Root growth impedance, measured by the reduction in radicle elongation, occurred mostly in the presence of either the largest- (GMD of 9.0 mm) or the smallestsized (GMD of 4.3 mm) fragments (Fig. 5). Displacement of soil aggregates in the upper layer of seedbeds is only important in the presence of roots with a diameter ⬎ 0.5 mm and aggregate diameters ⬍ 1 mm (Whiteley and Dexter, 1984). In this study, all soil fragments were ⬎2.0 mm, so most of the root growth occurred through void space between and within fragments rather than by pushing them apart. Goss and Russell (1980) showed that maize root elongation is slowed for approximately 10 min before deflection, just when the root tip makes contact with a glass bead. A thickened diameter of the roots at the point of deflection has been reported by several researchers (Whiteley et al., 1982; Logsdon et al., 1987). In this study, mean radicle diameter in the 9.0- and 4.3-mm seedbeds was 1.51 (⫾0.3) and 1.47 (⫾0.31) mm, respectively. In the 6.8- and the 5.1-mm seedbeds, mean radicle diameter was 1.30 (⫾0.15) and 1.34 (⫾0.17) mm, respectively. Thus, the reduction in radicle length observed in the 9.0-mm seedbeds may be a result of the frequent interception of maize root tips by the surface of fragments, reducing the root elongation rate. Similar results have been already reported, and attributed to the low penetrability of large soil fragments (e.g., Logsdon et al., 1987). The lowest AFP values occurred in the seedbeds with a fragment GMD of 4.3 mm (Table 2), suggesting a greater probability of interfragment pores being watersaturated, consequently reducing the total void space with pore sizes large enough for roots to grow through them. Bulk density, as well as the ␳br, and AFP partially explained the variability in total RLD according to a quadratic response model (Fig. 8). No improvement in model fit was obtained by discriminating between the different fragment size distribution treatments. From the first derivative of the fitted quadratic models, it can be observed that maximum root elongation per unit of soil volume occurred at a ␳b of 1.12 Mg m⫺3, a ␳br of 0.78 and/or an AFP of 0.187 m3 m⫺3. Under greater ␳b values, root growth was reduced because of mechanical impedance and greater possibility of anaerobic condi-

Reproduced from Crop Science. Published by Crop Science Society of America. All copyrights reserved.

DI´AZ-ZORITA ET AL.: EFFECTS ON ROOT ELONGATION IN MOIST SOIL

tions. Carter (1990) described a similar curvilinear relationship between cereal yields and ␳br, obtaining maximum yields at similar values, i.e., 0.81. Increasing soil compaction increases soil ␳b and penetration resistance levels and reduces root growth (Panayiotopoulos et al., 1994). Crop productivity is also reduced because of poor aeration in compacted soils (Hakansson and Lipiec, 2000). In the more compacted treatments (CE ⫽ 53.5 and 107.0 kJ m⫺3), average AFP values were lower than 0.12 m3 m⫺3, suggesting greater potential for inadequate oxygenation, mostly at the bottom of the pots. Hakansson and Lipiec (2000) reported that critical values of AFP for root growth vary with soil texture and stability and the continuity of macropores. For many soils, oxygen diffusion approaches zero at an AFP of 0.10 m3 m⫺3. This value is close to that observed at greater CE levels used in this experiment. Under the conditions of this experiment, with mean matric potentials approaching zero, both soil strength and oxygen limitations controlled maize root elongation at the greater CE levels. In the compacted soils, root elongation was much more sensitive to soil properties such as total, air-filled, and interporosity that resulted from the combination of fragments with different sizes and distribution rather than size of distribution itself.

CONCLUSIONS Maize radicle elongation is affected by changes in total porosity and AFP caused mainly for the application of mechanical stress. The GMD of soil fragments changes the average expression of that stress by not interacting with it over the levels of stress imposed in this work. Increasing CE caused the interfragment porosity to be reduced, more in seedbeds formed by large fragments, causing a reduction in the radicle length. Maximum total RLD occurred in seedbeds with GMDs of 5.1 to 6.8 mm and after the application of 26.8 kJ m⫺3, resulting in a mean dry ␳b of 1.15 Mg m⫺3. Variations in the LogGSD value (spread of the soil fragment sizes), between 0.22 and 0.41 caused little and no significant effect on root elongation. REFERENCES Alexander, K.G., and M.H. Miller. 1991. The effect of soil aggregate size on early growth and shoot-root ratio of maize (Zea mays L.). Plant Soil 138:189–194. American Society for Testing Materials. 2000. Standard methods for laboratory compaction characteristics of soil using standard effort (12,400 ft-lb f/ft3(600 kN/m3)). p. 1–11. In American Society for Testing Materials (ed.) ASTM Standard in Building Codes. ASTM. Angers, D.A., and J. Caron. 1998. Plant-induced changes in soil structure: Processes and feedbacks. Biogeochemistry 42:55–72. Arsenault, J.L., S. Poulcur, C. Messier, and R. Guay. 1995. WinRHIZOTM, a root-measuring system with a unique overlap correction method. HortScience 30:906. Bennie, A.T. 1996. Growth and mechanical impedance. p. 453–470. In Y. Waisel et al. (ed.) Plant roots: The hidden half. Marcel Dekker, New York. Blake, G.R., and K.H. Hartge. 1986. Particle density. p. 377–382. In A. Klute (ed.) Methods of soil analysis: Part 1—Physical and mineralogical methods. 2nd ed. Agron. Monogr. 9. ASA and SSSA, Madison, WI.

1425

Bouaziz, A., and L. Bruckler. 1989. Modeling wheat seedling growth and emergence: II. Comparison with field experiments. Soil Sci. Soc. Am. J. 53:1838–1846. Braunack, M.V. 1995. Effect of aggregate size and soil water content on emergence of soybean (Glycine max L. Merr.) and maize (Zea mays L.). Soil Tillage Res. 33:149–161. Braunack, M.V., and A.R. Dexter. 1989a. Soil aggregation in the seedbed: A review. I. Properties of aggregates and beds of aggregates. Soil Tillage Res. 14:259–279. Braunack, M.V., and A.R. Dexter. 1989b. Soil aggregation in the seedbed: A review. II. Effect of aggregate sizes on plant growth. Soil Tillage Res. 14:281–298. Brown, A.D., A.R. Dexter, W.C.T. Chamen, and G. Spoor. 1996. Effect of soil macroporosity and aggregate size on seed-soil contact. Soil Tillage Res. 38:203–216. Carter, M.R. 1990. Relative measures of soil to characterize compaction in tillage studies on fine sandy loams. Can. J. Soil Sci. 70: 425–433. Da Silva, A.P., B.D. Kay, and E. Perfect. 1997. Management versus inherent soil properties effects on and relative compaction. Soil Tillage Res. 44:81–93. de Freitas, P.L., R.W. Zobel, and V.A. Snyder. 1999. Corn root growth in soil columns with artificially constructed aggregates. Crop Sci. 39:725–730. Dexter, A.R. 1978. A stochastic model for the growth of root in tilled soils. J. Soil Sci. 29:102–116. Dexter, A.R. 1988. Advances in characterization of soil structure. Soil Tillage Res. 11:199–238. Dexter, A.R. 1997. Physical properties of tilled soils. Soil Tillage Res. 43:41–63. Dı´az-Zorita, M., J.H. Grove, and E. Perfect. 2001. Laboratory compaction of soil using a small mold procedure. Soil Sci. Soc. Am. J. 65:1593–1598. Donald, R.G., B.D. Kay, and M.H. Miller. 1987. The effect of soil aggregate size on early shoot and root growth of maize (Zea mays L.). Plant Soil 103:251–259. Feldman, L. 1994. The maize root. p. 29–37. In M. Freeling and V. Walbot (ed.) The maize handbook. Springer-Verlag, New York. Gee, G.W., and J.W. Bauder. 1986. Particle-size analysis. p. 383–411. In A. Klute (ed.) Methods of soil analysis. Part 1. Physical and mineralogical methods. 2nd ed. Agron. Monogr. 9. ASA and SSSA, Madison, WI. Goss, M.J., and R.S. Russell. 1980. Effects of mechanical impedance on root growth in barley (Hordeum vulgare L.) III. Observations on the mechanisms of response. J. Exp. Bot. 31:577–580. Hadas, A., and D. Russo. 1974. Water uptake by seeds as affected by water stress, capillary conductivity, and seed-soil water contact. II. Analysis of experimental data. Agron. J. 66:647–652. Hakansson, I., and J. Lipiec. 2000. A review of the usefulness of relative bulk density values in studies of soil structure and compaction. Soil Tillage Res. 53:71–85. Haque, A., A. Hamid, M.T. Islam, and M. Mohiuddin. 1992. Seedbed characteristics and seeding depth effects on emergence and seedling vigour of upland rice. J. Agron. Crop Sci. 168:61–64. Henin, R., R. Gras, and G. Monnier. 1969. Le profil cultural l’etat physique du sol et ses consequences agronomiques. 2nd ed. Masson, Paris. Hewitt, J.S., and A.R. Dexter. 1979. An improved model of root gowth in structured soil. Plant Soil 52:325–343. Hillel, D. 1998. Environmental soil physics. Academic Press, San Diego, CA. Horn, R. 1990. Aggregate characterization as compared to soil bulk properties. Soil Tillage Res. 17:265–289. Horn, R., K.R. Baumgartl, and S. Baasch. 1995. Effect of aggregate strength on strength and stress distribution in structured soils. p. 31–52. In K.H. Hartge and B.A. Stewart (ed.) Soil structure: Its development and function. CRC Lewis Publ., Boca Raton, FL. Horn, R., and M. Lebert. 1994. Soil compactability and compressibility. p. 45–69. In B.D. Soane and C. van Ouwerkerk (ed.) Soil compaction in crop production. Elsevier Science, the Netherlands. Kay, B.D. 1990. Rates of change of soil structure under different cropping systems. Adv. Soil Sci. 12:1–52. Kay, B.D., and D.A. Angers. 1999. Soil structure. p. 229–276. In

Reproduced from Crop Science. Published by Crop Science Society of America. All copyrights reserved.

1426

CROP SCIENCE, VOL. 45, JULY–AUGUST 2005

M.E. Sumner (ed.) Handbook of soil science. CRC Press, Boca Raton, FL. Logsdon, S.D., J.C. Parker, and R.B. Reneau. 1987. Root growth as influenced by aggregate size. Plant Soil 99:267–275. Misra, R.K., A.M. Alston, and A.R. Dexter. 1988. Root growth and phosphorus uptake in relation to the size and strength of soil aggregates. I. Experimental studies. Soil Tillage Res. 11:103–116. Nash, V.E., and V.C. Baligar. 1974. The growth of soybean (Glycine max) roots in relation to soil micromorphology. Plant Soil 41:81–89. Nasr, H.M., and F. Selles. 1995. Seedling emergence as influenced by aggregate size, and penetration resistance of the seedbed. Soil Tillage Res. 34:61–76. Panayiotopoulos, K.P., C.P. Papadopoulou, and A. Hatjiioannidou. 1994. Compaction and penetration resistance of an Alfisol and Entisol and their influence on root growth of maize seedlings. Soil Tillage Res. 31:323–337. Russell, R.S. 1977. Plant root systems: Their function and interaction with the soil. McGraw-Hill, New York. SAS Institute. 1997. SAS/STAT威 software: Changes and enhancements through Release 6.12. SAS Inst., Cary, NC. Schneider, E.C., and S.C. Gupta. 1985. Corn emergence as influenced

by soil temperature, matric potential, and aggregate size distribution. Soil Sci. Soc. Am. J. 49:415–422. Souty, N., and C. Rode. 1993. Emergence of sugar beet seedlings from under different obstacles. Eur. J. Agron. 2:213–221. Taylor, M.S. 1974. The effect of soil aggregate size on seedling emergence and early growth. East Afr. Agric. For. J. 40:204–213. Tollenaar, M., and L.M. Dwyer. 1999. Physiology of maize. p. 169–204. In D.L. Smith and C. Hamel (ed.) Crop yield physiology and processes. Springer-Verlag, Berlin. Vepraskas, M.J., and M.G. Wagger. 1990. Corn root distribution and yield response to subsoiling for Paleudults having different aggregate sizes. Soil Sci. Soc. Am. J. 54:849–854. Voorhees, W.B., M. Amemiya, R.R. Allmaras, and W.E. Larson. 1971. Some effects of aggregate structure heterogenity on root growth. Soil Sci. Soc. Am. Proc. 35:638–643. Whiteley, G.M., and A.R. Dexter. 1984. Displacement of soil aggregates by elongating roots and emerging shoots of crop plants. Plant Soil 77:131–140. Whiteley, G.M., J.S. Hewitt, and A.R. Dexter. 1982. The buckling of plant roots. Physiol. Plant. 54:333–342. Wuest, S.B., S.L. Albrecht, and K.W. Skirvin. 1999. Vapor transport vs. seed-soil contact in wheat germination. Agron. J. 91:783–787.