Printer Friendly

Suboptimal fertilisation compromises soil physical properties of a hard-setting sandy loam.


During recent decades, a more comprehensive concept of crop production sustainability has evolved within the European Union (Jones et al. 2012), emphasising that sustainable management involves protection of the fundamental and non-renewable soil resource. The overall target of nutrient management in agroecosystems is to achieve sustainable crop production by optimising fertiliser use efficiencies and minimising nutrient losses to the environment while maintaining soil quality in the long term. This involves sustaining soil physical properties related to soil tilth. Soil physical properties are affected by nutrient management via inputs and turnover of soil organic matter (SOM) from aboveground and belowground crop residues and from organic amendments (e.g. animal manures).

Previous studies have examined the physical properties of soils under different fertilisation regimes, but generally only one or two physical properties have been considered (Haynes and Naidu 1998; Edmeades 2003; Bronick and Lai 2005). Additions of animal manure tend to decrease soil bulk density (BD) and increase porosity, aggregate stability and hydraulic conductivity relative to soil dressed with mineral fertiliser (Edmeades 2003). Research on fertilisation effects on soil physical properties has often focussed on fertiliser type, whereas effects of fertiliser rate have largely been neglected. One exception is Schjonning et al. (2005), who found increased hydraulic conductivity and macro-porosity in the subsoil with increased fertiliser rate, irrespective of fertiliser type (mineral fertiliser or animal manure). Studies on long-term effects of fertiliser type typically lack field replicates (Edmeades 2003) or are flawed by having confounding effects of other management elements (e.g. crop rotation, residue management and tillage system).

Since soil physical properties are intimately linked to SOM levels and turnover, experiments with long-continued nutrient management are indispensable as changes in SOM occur slowly and over long periods (Christensen and Johnston 1997). Thus, long-term field experiments represent a unique research platform for assessing relationships between nutrient management and soil physical properties related to soil tilth.

This study investigates the long-term effects of fertiliser rate and type on key soil structural, mechanical and chemical properties. Soil was retrieved from the Lermarken site of the Askov Long-Term Experiment on Animal Manure and Mineral Fertilizers (Askov-LTE), a sandy loam with little spatial variation in topsoil texture and a gradient in soil organic carbon (SOC) that has developed during 120 years of contrasting nutrient management without the confounding effects of variability in other management practices.

Materials and methods

The Askov-LTE and treatments

The Askov-LTE was established in 1894 on a sandy loam soil (Table 1) at Askov Experimental Station, Denmark (55[degrees]28'N, 09[degrees]07'E; 63 m above sea level). The soil is terminal morainic deposits from the Weichselian glaciation (Nielsen and Moberg 1984) and classified as an Aric Haplic Luvisol (1USS Working Group WRB 2015) and Ultic Hapludalf (Soil Survey Staff 2014). Average annual temperature and precipitation are 7.7[degrees]C and 862 mm respectively (means of 1961-1990).

The Askov-LTE Lermarken site includes four field blocks, termed B2-B5, and carries a four-course rotation of winter wheat (Triticum aestivum), silage maize (Zea mays) and spring barley (Hordeum vulgare), and a grass-clover mixture (Trifolium hybridum, Medicago sativa, Lotus corniculatus, Lolium perenne, Festuca pratensis and Phleum pratense) that is used for cutting in the following production year. Soil pH is kept in the range 5.5-6.5 by adding magnesium-cnriched lime every four years.

In B5, we selected four different nutrient treatments: unfertilised (UNF), 1/2 mineral fertiliser (V2NPK; initiated in 1923), 1 mineral fertiliser (1NPK.) and 1 1/2 animal manure (1 1/2AM). Since 1973, 1NPK and 1AM have corresponded to 100 kg total-nitrogen (N) [ha.sup.-1], 20 kg phosphorus (P) [ha.sup.-1] and 90 kg potassium (K) ha 1 (annual input across the rotation) and animal manure has been cattle slurry with 60-70% of its total-N being ammoniacal-N. The treatments were selected to test the effect of (i) fertiliser type by comparing 1NPK. and 1 1/2AM and (ii) mineral fertiliser rate by comparing UNF, 1/2NPK and 1N PK. Based on SOC data obtained for all plots in 2012, the 12 plots used in the present study were organised into three blocks (Fig. 1) to account for the spatial variation in background SOC level. More details on the Askov-LTE are given by Christensen and Johnston (1997) and Christensen et al. (2006). For the treatments involved in the present study, unpublished crop yields for the period 2006-2015 were supplied by Askov Experimental Station.

Soil sampling

Soil was sampled in the central part of the plot (Fig. 1) in September 2014 following a winter wheat crop and at a soil water potential of-250 hPa. Undisturbed cores (100 [cm.sup.3]) were extracted from the 6-10-cm soil layer, and minimally-disturbed soil cubes (4000 [cm.sup.3]) were sampled from the 6-15-cm layer. Six cores and four cubcs were extracted within each plot and stored at 2[degrees]C. Soil from two cubes per plot were spread out on a table at room temperature, carefully fragmented by hand along natural planes of weakness and finally left to air-dry.

Basic chemical and physical analysis

Portions of air-dry soil from each plot was crushcd to <2 mm and SOC content determined by dry combustion using ball-milled aliquots and a Thermo Flash 2000 NC Soil Analyzer (Thermo Fisher Scientific, Waltham, MA, USA). Soil texture was determined with a combined hydrometer/sieve method after removal of SOM with hydrogen peroxide (Gee and Or 2002). Soil pH was determined with a glass electrode using 10 g of soil in 25 mL of 0.01 M calcium chloride (Ca[Cl.sub.2]) solution, and specific surface area was determined by the ethylene glycol monocthyl ether method (Petersen et al. 1996).

Pore characteristics

The six undisturbed soil cores from each plot were placed on sandboxes and slowly wetted to saturation from beneath. The soil water content was then sequentially adjusted to -10, -30, -100 and -300 hPa using sandboxes and ceramic plates (Dane and Hopmans 2002). Finally, the soils were oven-dried at 105[degrees]C for 24 h. Soil sample weight was recorded at field moisture level, at each matric potential, and again after oven-drying. Soil porosity was estimated from BD and particle density. Particle density was estimated according to Schjonning et al. (2017). Volumetric water content at each matric potential was calculated from the weight loss upon oven-drying. Air-filled porosity ([[epsilon].sub.a]) was determined as the difference between total porosity and water retained at the corresponding matric potential. Water retention at -1.5 MPa (permanent wilting point) was determined at plot level using a WP4-T Dewpoint Potentiometer (Scanlon et al. 2002) and based on air-dry <2-mm soil. Air permeability ([k.sub.a]) was measured by the steady-state method of Iversen et al. (2001) using the cores adjusted to -30 and -100 hPa. Specific permeability (SP) was calculated (Groenevclt et al. 1984) from air permeability and air-filled porosity:

SP = [k.sub.a]/[[epsilon].sub.a] (1)

Shear strength

Shear strength was determined on the soil cores adjusted to -300 hPa using an annulus shear test (Schjonning 1986). The six soil cores per plot were tested at one of the following normal loads: 30, 60, 90, 120, 150 and 180 kPa. The inner and outer diameters of the shear annulus carrying the load were 18 and 40 mm respectively. Thus, 10 [cm.sup.2] of the ~30-[cm.sup.2] core surface area was loaded and sheared. Displacement of the annulus was recorded after the load had been applied, before and after shearing. The shearing rate at the mean shear radius was 45.6mm [min.sup.-1]. Soil cohesion and internal friction were estimated for each plot as the intercept and slope respectively, of a regression of peak shear stress against normal load. The displacement (s) before shear may be taken as the result of a uniaxial, semi-confined compression test. The strain ([epsilon]) was calculated as the ratio between s and the height of the soil core (H):

[epsilon] = s/H (2)

Clay dispersibility and water-stable aggregates

Clay dispersibility (CD) and water-stable aggregates (WSA) were determined on field-moist soil as outlined by Vendelboe et al. (2012) and originally suggested by Pojasok and Kay (1990). Soil was retrieved from the minimally-disturbed 4000-[cm.sup.3] soil cubes using a small corer (22-mm diameter) and gently crumbled by hand to pass an 8-mm sieve. Two replicates per plot were made. Briefly, CD was determined by adding artificial rainwater to cylindrical plastic bottles containing 10 g of soil in order to obtain a soil: water ratio of 1:8 by weight. The bottles were rotated end-over-end (33 rpm; 23-cm diameter rotation) for 2 min, left to stand for 230 min, after which the upper 50 mm (60 mL) containing particles [less than or equal to]2 [micro]m was siphoned off. The weight of dispersed clay was determined after oven-drying at 105[degrees]C for 24 h. The sediment residing in the plastic bottles was wet-sieved (2 mm) in order to relate the dispersed clay to stone-free soil. Percentage WSA was determined by transferring 30 g of soil to a 20-cm sieve with 250-[micro]m openings. The sieve was mounted in a sieving apparatus containing 5 L of artificial rainwater. In its upper position the sieve was just touching the water surface. After 30 s of capillary wetting, the sieve was moved up and down for 2 min (52 cycles min 1; stroke length 28 mm). WSA was calculated as the fraction of soil remaining on the sieve and corrected for mineral particles >250 [micro]m isolated by chemical dispersion. A 10-g subsample of soil was oven-dried to determine water content to allow expressing results on a dry-weight basis.

Aggregate tensile strength

The soil remaining from the field-moist soil cubes was carefully fragmented by hand, and 8-16-mm aggregates were isolated by sieving. These aggregates were divided into three groups of different moisture status: (i) natural air-dry aggregates (NAD); (ii) natural air-dry aggregates re-wetted to field capacity, i.e. -100 hPa (NFC; Munkholm et al. 2002); and (iii) natural field-moist aggregates (NFM). Tensile strength was tested for 15 randomly-selected aggregates for each of the three moisture states (12 field plots; 540 tests). Individual aggregates were placed between two parallel plates that were compressed at a constant rate of 0.03 mm s 1. The compressive force was measured with a load cell (0-100 N), and aggregate failure was detected when aggregates cracked or a sudden drop in compression force was observed. Subsequently, aggregate water content was determined by oven-drying subsamples of soil aggregates at 105[degrees]C for 24 h. Aggregate tensile strength (y) was calculated by the equation (Dexter and Kroesbergen 1985):

Y = 0.576F/[d.sup.2] (3)

where F (N) is the compressive force needed to fracture the aggregates and d (m) is the effective diameter of each aggregate. To obtain the effective diameter of each aggregate, aggregate diameters were adjusted according to their mass (Dexter and Kroesbergen 1985):

d = d,[(m/[m.sub.0]).sup.1/3] (4)

where [d.sub.1] is the mean aggregate diameter (0.012 m; average of the upper (0.016 m) and lower (0.008 m) mesh size), w (g) is the dry mass of the aggregate and [m.sub.0](g) is the mean dry mass for batches of 15 aggregates.

Remoulded aggregates based on soil from 30 natural aggregates (8-16 mm) were used to simulate the effect of total loss of soil structure (Watts et al. 1996). Initially, one group of natural aggregates was air-dried while another group remained field-moist. Distilled water was added to the aggregates until they became liquid, and then they were stirred for some time and rolled by hand to form spherical 12-mm aggregates. The remoulded air-dry (RAD) and field-moist (RFM) aggregates were air-dried and tested for Y as described for natural aggregates. In contrast to natural aggregates, the diameter of remoulded aggregates was estimated from their average height and width.

The rupture energy (E) was derived from the area under the stress-strain curve (Vomocil and Chancellor 1969):

E [approximately equal to] [[summation].sub.i]F([s.sub.i])[DELTA][S.sub.i] (5)

where F([s.sub.i]) is the mean force at the ith subinterval and [DELTA][s.sub.i], is the displacement length of the ith subinterval. The mass-specific rupture energy ([E.sub.sp]) was calculated as:

[E.sub.sp] = E/m (6)

where m is the mass of the aggregate.

Soil friability for each field plot was estimated from the variation of Y and [E.sub.sp] in relation to their means respectively (Watts and Dexter 1998):

[F.sub.Y] = [[sigma].sub.Y]/[bar.Y] (7)



where [F.sub.Y] and [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] are the friability, [[sigma].sub.Y] and [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] are the standard deviation and [bar.Y] and [[bar.E].sub.sp] are the mean.

The workability index (W) developed by Arthur et al. (2014) was calculatcd based on both Y and [E.sub.sp]:

[W.sub.Y] = [F.sub.Y](1/[bar.Y]) (9)



Thus, W combines friability and the energy needed to fragment the aggregates. High values of W indicate greater workability and vice versa.


The experimental design was a block design with three field replicates. The statistical analysis applied the R-Project software package Version 3.1.1 (R Foundation for Statistical Computing). Treatment effects were analysed with a linear model including block as a fixed effect. Clay was included as a co-variable if it was significant in itself and made the treatment effect significant. Logarithmic (In) transformation was performed on Y, [E.sub.sp], [k.sub.a] and SP to yield normality. The criterion used for statistical significance of treatment effects was P < 0.05. When an ANOVA test showed the treatment effect to be significant, further analyses were made to isolate differences between treatments (pairwise comparisons) using the general linear hypotheses (glht) function implemented in the R multcomp package. The univariate test applied was Fischer's protected t-test. A paired t-test was used to investigate if Y and [E.sub.sp] were affected by the different initial water contents of the remoulded aggregates.


Basic soil characteristics and crop yields

Concentrations of SOC differed significantly in the order 1 1/2AM > 1NPK = 1/ 2NPK > UNF (Table 1). Contents of clay, silt and sand did not differ among treatments and the effects of fertilisation could be examined without confounding effects related to soil texture. The texture analysis showed that the Askov soil is well graded and coarse. Thus the soil can be considered hard-setting (Schjonning and Thomsen 2013). Soil BD was greater for UNF and 1/2 NPK than for 1NPK and 1 1/2 AM treatments. Dry matter (DM) yields of crop biomass were significantly higher for I 1/2AM compared with 1NPK, although the difference was numerically small. In contrast, biomass yields were smaller for 1/2 NPK and very small for UNF. Soil specific surface area was not significantly affected by fertilisation. Mineral fertiliser tended to lower soil pH reflecting the acidifying nature of mineral fertiliser.

Soil stability and strength

The amount of CD was significantly higher and the percentage of WSA significantly lower for UNF than for 1NPK and 1 1/2AM, with the 1/2NPK. treatment showing intermediate soil structural stability (Fig. 2).

Generally, fertilisation did not affect aggregate strength parameters (Table 2). The only significant differences were for Y of NAD aggregates and the ratio between NAD and NFM aggregates. Y increased nearly 14-fold for UNF, V2NPK and 1NPK but just 8-fold for 1/2AM when going from NFM to NAD aggregates.

For RAD and RFM aggregates, the geometric mean Y for all plots was 233 and 259 kPa respectively, with corresponding standard deviations of 39 and 32 kPa. The initial water content had a significant effect on Y (P=0.03). The same was true for [E.sub.sp] (P = 0.01), the geometric means for RAD and RFM aggregates being 3.86 and 4.54 J [kg.sup.-1] respectively, with corresponding standard deviations of 0.80 and 0.60 J [kg.sup.-1]. Thus, remoulded initially air-dry aggregates were weaker than remoulded initially field-moist aggregates.

Soil cohesion was significantly higher for the UNF and 1/2NPK compared with 1NPK and 1 1/2AM (Table 3). The differences in soil cohesion were reflected by differences in BD, and BD explained 67% of the variation in soil cohesion (P= 0.001). Fertilisation did not affect soil friction and soil strain.

Pore characteristics

Plant-available water (PAW; water retained in 0.2-30 [micro]m pores) was very similar for 1/2NPK and 1NPK. (Table 4). PAW was higher for 1 1/2AM and lower for UNF. The PAW was 7-18% higher for fertilised soils compared with the UNF soil; corresponding to 2-5 mm water in the 0-20 cm soil layer. The volume of coarse pores >30 [micro]m was significantly higher for 1NPK than for UNF and 1/2NPK, with an intermediate value for 1 1/2AM.

At a matric potential of-30 hPa, [k.sub.a] and SP were significantly higher for 1NPK. and 1 1/2AM than for V2NPK (Table 4). This indicates that for these two treatments, soil pores >100 [micro]m were more effective at conducting air by convection and had a higher degree of pore continuity. The same trend was seen at -100 hPa (>30 [micro]m pores), although differences at this potential were not significant.


Fertiliser rate

The contrasting rates of mineral fertilisers induced significant differences in soil properties. The UNF soil had less SOC, higher BD, reduced aggregate stability, less PAW and higher CD and cohesion than soil dressed with 1NPK. This is consistent with previous studies using the Askov-LTE (Schjonning et al. 1994, 2005; Scanning 1995). The 1/2NPK soil had intermediate values for CD and aggregate stability but was similar to UNF with regard to cohesion and BD. The ka and SP at -30 hPa were significantly lower for 1/2NPK than for 1/2NPK while Y of NAD aggregates was significantly higher for V2NPK compared with the other treatments. Higher CD and decreased wet-stability of aggregates increase the potential for surface runoff and downward preferential transport of fine particles carrying pollutants to the environment (Norgaard et al. 2013). Migrating clay particles may be deposited in macropores in the soil profile (Kjaergaard et al. 2004) causing internal crusting and reducing the drainage capacity of the soil. Dispersed clay may also be deposited at the soil surface causing soil cementing and non-friable aggregates (Schjenning et al. 2012) with negative effects on soil tilth, aeration, crop establishment and early growth. Clearly, several vital soil physical properties were compromised at suboptimal fertilisation rates.

Fertiliser type

The effect of fertiliser type on soil physical properties was tested by comparing 1NPK and 1 1/2AM. The 1 1/2AM soil received more total-N than 1NPK to compensate for its smaller content of plant-available N. Although crop biomass yields differed significantly between 1 NPK. and 1 1/2AM, the differences were small (Table 1). The main difference between 1NPK and 1 1/2AM is ascribed to the organic matter added with animal manure and reflected in the significantly higher SOC content of the 1 1/2AM soil.

In general, 1NPK and 1 1/2AM showed similar physical soil properties. However, the difference in Y between NAD and NFM aggregates was significantly smaller for 1 1/2AM than for 1 NPK. This was due to a lower Kofthe NAD aggregates isolated from the 1 1/2AM soil, implying a wider range in optimum water content for tillage in manured soil (Munkholm et al. 2002). The distribution of soil porosity in pore size classes showed that the use of AM increased the amount of PAW and tended to reduce the fraction of coarse pores >30 [micro]m (P=0.09). Minor differences in soil pore characteristics between manured and mineral-fertilised soil have also been reported in other studies (Schjonning et al. 1994, 2005; Eden et al. 2011).

The overall effect on soil physical properties caused by manure or mineral fertiliser added at crop-yield-optimised rates is consistent with results reviewed by Edmeades (2003), whereas Munkholm et al. (2001) and Abdollahi et al. (2014) reported poorer tilth conditions for mineral-fertilised soil compared with manured soil. These conflicting results may be ascribed to differences in management factors, such as crop rotation and residue management. We embedded our study in a long-term field experiment with contrasting nutrient regimes, which eliminated confounding effects derived from differences in soil type, soil texture, cropping history, crop rotation, soil tillage system and climate.

Critical carbon levels?

Significant differences in SOC between unfertilised, mineral-fertilised and manured soil have previously been reported for the Askov-LTE (e.g. Schjonning et al. 1994). In relation to soil structural properties, the soils can be divided into two groups in which the 1NPK and 1 1/2AM soils have better soil structure than V2NPK and UNF soils. This may be ascribed to differences in the presence of organic binding and bonding material such as soil polysaccharides and plant roots (Degens 1997). Crop biomass yields--and thus supposedly the amounts of roots--differed little between 1NPK and P/2AM, whereas yields were lower in 1/2NPK and markedly lower in UNF. Debosz et al. (2002) found similar levels of hot-water-extractable C in soils with crop-yield-optimised additions of mineral fertiliser and animal manure; whereas it was significantly lower in unfertilised soil. This may explain why there were similar levels of CD and WSA for 1 NPK and IV2AM despite their different contents of total SOC, and why CD increased and the WSA percentage decreased linearly with decreasing fertiliser rate. It is noteworthy that CD was 36% higher and WSA 11 % lower for V2NPK soil compared with 1NPK soil, whereas CD was 62% higher and WSA 28% lower when comparing UNF to 1NPK.

This illustrates that even small differences in SOC contents may have large impacts on soil structural stability, but also that concentrations of total SOC are less illustrative of the soil structural status. The deprived structural properties in UNF and 1/2NPK soils may reflect that the fraction of SOC associated with clay (<2 [micro]m) has become critically low. Dexter et al. (2008) showed that for selected arable soils it was not the SOC per se that controlled CD and BD but the amount of SOC interacting with clay. They suggested a critical clay-to-SOC ratio of ~10 corresponding to a soil where the clay particles are complex-bound with SOC. At values above 10 they found a decline in CD and BD. Schjonning et al. (2012) later proposed a critical clay+silt (<20 [micro]m)-to-SOC ratio of ~20. The clay-to-SOC and clay+silt-to-SOC ratios for UNF and 1/2NPK soils were 9-10 and 19-21 respectively (Table 1), and for these soils CD, WSA and BD declined. Thus our study corroborates the suggested soil mineral particles-to-SOC ratios as a defining factor for critical soil physical properties.

Shear strength was related to field tillage by calculating the shear strength at 100 kPa normal load ([[tau].sub.100]) using cohesion and friction estimates for each plot. Linear regression revealed a decrease in [[tau].sub.100] when porosity increased (Fig. 3), indicating that less energy is needed for tillage operations. The UNF and 1/2NPK treatments had low porosities and high [[tau].sub.100] compared with 1NPK and 1 1/2AM, confirming the differences in soil physical properties.

The degree of compaction was suggested by Hakansson (1990) as an indicator of the physical status of soil. The degree of compaction is the actual BD divided by the reference bulk density (RBD). RBD is defined as the BD obtained after structurally-disturbed and wet soil has been subjected to uniaxial confined compression at 200 kPa for around one week. The RBD was predicted from the Rosin-Rammler texture parameters (Table 1; Rosin and Rammler 1933) and the SOC concentration by applying eqn 14 of Keller and Hakansson (2010). The RBD differed little between treatments, indicating that the degree of compaction was dominated by BD which, in turn, was closely related to SOC content (r=-0.81, P=0.001). The degree of compaction for UNF, 1/2NPK, 1NPK and 1 1/2AM was 91, 90, 84 and 85% respectively. Hakansson (2005) estimated the optimal degree of compaction for spring barley, winter wheat, and rye and winter rape to be 87, 84 and 81% respectively. The degree of compaction was therefore considered optimal for INPK and 1 1/2AM but too high for UNF and 1/2NPK soils.

Comparison of friability indices

The friability indices ([F.sub.Y] and [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]) were [log.sub.10]- transformed to obtain homogeneity of the variances. The friability was rated higher when using [E.sub.sp] than when using Y independent of the water content, and when aggregates were remoulded (Fig. 4). This was due to a higher variation in [E.sub.sp] than Y. Taking the geometric mean values of [F.sub.Y] and [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] for natural and remoulded aggregates, the following linear relationships were derived:



This is in agreement with the study of Perfect and Kay (1994) who used air-dry aggregates and the Weibull model to estimate friability. Munkholm and Kay (2002) also found higher values of [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] using aggregates with different moisture status and scaling of aggregate strength with aggregate size to estimate friability. However, Schjonning et al. (2012) found similar friability indices when assessing the same aspect of friability for air-dry aggregates. Hence the values obtained for [F.sub.Y] and [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] appear to reflect the method used to assess friability and aggregate moisture status, but generally values for [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] are higher.

Effects of the initial water content on the strength of remoulded aggregates

The RFM aggregates were stronger than the RAD aggregates probably because more clay was released to the water when aggregates were initially moist (Dexter et al. 2011). Enhanced clay dispersion is expected to lead to more cementation and stronger aggregates. Consequently, the water status of aggregates when remoulded influences their strength, suggesting that it will be more relevant to remould aggregates at field moisture content when mimicking tillage operations that occur at water contents above the wet limit for tillage. Applying a multiple linear regression to explain Y of RFM and RAD aggregates by clay and SOC contents showed a significant coefficient of determination for RFM ([R.sup.2]-adj. = 0.63, P=0.005) but not for RAD aggregates ([R.sup.2]-adj. =0.00, P = 0.42), indicating a higher relevance of remoulding initially field-moist aggregates.

General considerations and perspectives

One may speculate why 120 years of no fertilisation did not degrade the soil physical properties to a greater extent. This could possibly be explained by the versatile crop rotation that includes grass-clover. Grass-clover adds organic binding and bonding material to the soil and generally upholds the soil structure (Riley et al. 2008) and thus reduces the exhaustive effect of non-fertilisation. This is also indicated by the smaller yield reduction for the grass-clover crop (49%) than for the other crops in the rotation (70-81%) when comparing UNF and INPK (Table 1). The literature analysed by Schjonning et al. (2007) also indicates that crop rotations appear to better sustain the SOC levels and presumably the soil physical properties than different fertilisation strategics.

The lack of dramatic differences in soil physical properties between treatments might also be explained by the soil type being coarse, graded and therefore susceptible to densification if not frequently tilled (Schjonning and Thomson 2013). Thus, the hard-setting behaviour of the soil gives rise to a high natural density which reduces differences in treatment effects. The treatment differences would probably also be more pronounced in soils richer in clay as these would require more SOM to stabilise soil structure (Dexter et al. 2008).


This study shows that application of mineral fertiliser at half the standard rate and especially the omission of fertilisation cannot sustain crucial soil physical properties. In contrast, mineral fertiliser and animal manure applied at a standard rate of plant-available N are almost equally able to sustain soil physical properties. The degraded physical properties at suboptimal fertilisation may be ascribed to a smaller amount of active organic binding and bonding materials due to a lower input of labile organic matter and critically low concentration of SOC in relation to contents of reactive minerals. Our results corroborate the suggested critical clay (<2 [micro]m)-to-SOC ratio of 10 and clay+silt (<20 [micro]m)-to-SOC ratio of 20 for CD, wet aggregate stability and BD.


We gratefully acknowledge the technical assistance of Stig T. Rasmussen, Bodil B. Christensen, Jorgen M. Nielsen and Palle Jorgensen and the staff at Askov Experimental Station. We thank Kristian Kristcnsen for statistical advice. The text was linguistically improved by Margit Schacht. The study was financially supported by the Ministry of Environment and Food of Denmark (GUDP projects OptiPlant and OptiTill).


Abdollahi L, Schjonning P, Elmholt S, Munkholm LJ (2014) The effects of organic matter application and intensive tillage and traffic on soil structure formation and stability. Soil & Tillage Research 136, 28-37. doi: 10.1016/j.still.2013.09.011

Arthur E, Schjonning P, Moldrup P, Razzaghi F, Tuller M (2014) Soil structure and microbial activity dynamics in 20-month field-incubated organic-amended soils. European Journal of Soil Science 65, 218-230. doi:10.l 111/ejss. 12121

Bronick CJ, Lai R (2005) Soil structure and management: a review. Geoderma 124, 3-22. doi:10.1016/j.geoderma.2004.03.005

Christensen BT, Johnston AE (1997) Soil organic matter and soil quality--lessons learned from long-term experiments at Askov and Rothamsted. In 'Developments in soil science. Vol. 25'. (Eds EG Gregorich, MR Carter) pp. 399-430. (Elsevier)

Christensen BT, Petersen J, Trentemoller LIT (2006) The Askov long-term experiments on animal manure and mineral fertilizers: the Lermarken site 1894-2004. DIAS report Plant Production no. 121. Danish Institute of Agricultural Sciences, Tjele, Denmark.

Dane JH, Hopmans JW (2002) Water Retention and Storage, in 'Methods of soil analysis. Part 4. Physical methods'. (Eds JH Dane, GC Topp) pp. 671-720. (Soil Science Society of America, Inc.: Madison, WI)

Debosz K, Vognsen L, Labouriau R (2002) Carbohydrates in hot water extracts of soil aggregates as influenced by long-term management. Communications in Soil Science and Plant Analysis 33,623-634. doi: 10. 1081 /CSS-120002768

Degens BP (1997) Macro-aggregation of soils by biological bonding and binding mechanisms and the factors affecting these: a review. Australian Journal of Soil Research 35, 431-460. doi: 10.1071/S96016

Dexter AR, Kroesbergen B (1985) Methodology for determination of tensile strength of soil aggregates. Journal of Agricultural Engineering Research 31, 139-147. doi: 10.1016/0021-8634(85)90066-6

Dexter A, Richard G, Arrouays D, Czyz E, Jolivet C, Duval O (2008) Complexed organic matter controls soil physical properties. Geoderma 144, 620 627. doi:10.1016/j.geoderma.2008.01.022

Dexter AR, Richard G, Davy J, Hardy M, Duval O (2011) Clay dispersion from soil as a function of antecedent water potential. Soil Science Society ofAmerica JournalTi, 444 455. doi:10.2136/sssaj2010. 0088

Eden M, Schjonning P, Moldrup P, De Jonge LW (2011) Compaction and rotovation effects on soil pore characteristics of a loamy sand soil with contrasting organic matter content. Soil Use and Management 27, 340-349. doi:10.1111/j.1475-2743.2011,00344.x

Edmeades DC (2003) The long-term effects of manures and fertilisers on soil productivity and quality: a review. Nutrient Cycling in Agroecosystems 66, 165 180. doi:10.1023/A:t023999816690

Gee GW, Or D (2002) Particle-size analysis. In 'Methods of soil analysis. Part 4. Physical methods'. (Eds JH Dane, GC Topp) pp. 255-294. (Soil Science Society of America, Inc.: Madison, WI)

Groenevelt PH, Kay BD, Grant CD (1984) Physical assessment of a soil with respect to rooting potential. Geoderma 34, 101-114. doi:10.1016/ 0016-7061(84)90016-8

Hakansson I (1990) A method for characterizing the state of compactness of the plough layer. Soil & Tillage Research 16, 105-120. doi: 10.1016/ 0167-1987(90)90024-8

Hakansson 1 (2005) Machinery-induced compaction of arable soils: incidence consequences--counter-measures. Reports from the division of soil management No. 109, 1-153. (Swedish University of Agricultural Sciences, Department of Soil Sciences, Sciences: Uppsala).

Haynes RJ, Naidu R (1998) Influence of lime, fertilizer and manure applications on soil organic matter content and soil physical conditions: a review. Nutrient Cycling in Agroecosystems 51, 123-137. doi: 10.1023/ A: 1009738307837

IUSS Working Group WRB (2015) World Reference Base for Soil Resources 2014, update 2015. International soil classification system for naming soils and creating legends for soil maps. World Soil Resources Reports No. 106. FAO, Rome.

Iversen BV, Schjonning P, Poulsen TG, Moldrup P (2001) In situ, on-site and laboratory measurements of soil air permeability: boundary conditions and measurement scale. Soil Science 166, 97-106. doi: 10.10 97/00010694-200102000-00003

Jones A, Panagos P, Barcelo S, Bouraoui F, Bosco C, Dewitte O, Gardi C, Erhard M, Hervas J, Hiederer R, Jeffery S, Lukewille A, Marmo L, Montanarella L, Olazabal C, Petersen JE, Penizek V, Strassburger T, Toth G, Van Den Eeckhaut M, Van Liedckerke M, Verheijen F, Viestova E, Yigini Y (2012) The state of soil in Europe. A contribution of the JRC to the European Environment Agency's Environment State and Outlook Report--SOER 2010, Report EUR 25186 EN. 1-71. Ispra, Italy.

Keller T, Hakansson I (2010) Estimation of reference bulk density from soil particle size distribution and soil organic matter content. Geoderma 154, 398 406.

Kjaergaard C, Hansen HCB, Koch CB, Villholth KG (2004) Properties of water-dispersible colloids from macropore deposits and bulk horizons of an Agrudalf. Soil Science Society of America Journal 68, 1844-1852. doi: 10.2136/sssaj2004.1844

Munkholm LJ, Kay BD (2002) Effect of water regime on aggregate-tensile strength, rupture energy, and friability. Soil Science Society of America Journal 66, 702-709. doi:10.2136/sssaj2002.7020

Munkholm LJ, Schjonning P, Petersen CT (2001) Soil mechanical behaviour of sandy loams in a temperate climate: case studies on long-term effects of fertilization and crop rotation. Soil Use and Management 17,269-277. doi: 10.1079/SUM200190

Munkholm LJ, Schjanning P, Dcbosz K, Jensen HE, Christensen BT (2002) Aggregate strength and mechanical behaviour of a sandy loam soil under long-term fertilization treatments. European Journal of Soil Science 53, 129-137. doi: 10.1046/j.1365-2389.2002.00424.X

Nielsen ID, Maberg JP (1984) Classification of 5 soil profiles at Danish State Research Stations. Tidsskriftfor Planteavl 88. 155-168. [In Danish with an English summary]

Norgaard T, Moldrup P, Olsen P, Vendelboe AL, Iversen BV, Greve MH, Kjaer J, de Jonge LW (2013) Comparative mapping of soil physical-chemical and structural parameters at field scale to identify zones of enhanced leaching risk. Journal of Environmental Quality 42, 271 283. doi: 10.2134/jeq2012.0105

Perfect E, Kay BD (1994) Statistical characterization of dry aggregate strength using rupture energy. Soil Science Society of America Journal 58, 1804 1809. doi:10.2136/sssaj1994.0361599500580006 0032x

Petersen LW, Moldrup P, Jacobsen OH, Rolston DE (1996) Relations between specific surface area and soil physical and chemical properties. Soil Science 161, 9-21. doi: 10.1097/00010694-199601000-00003

Pojasok T, Kay BD (1990) Assessment of a combination of wet sieving and turbidimetry to characterize the structural stability of moist aggregates. Canadian Journal of Soil Science 70, 33-42. doi: 10.4141/ cjss90-004

Riley H, Pommeresche R, Eltun R, Hansen S, Korsaeth A (2008) Soil structure, organic matter and earthworm activity in a comparison of cropping systems with contrasting tillage, rotations, fertilizer levels and manure use. Agriculture, Ecosystems & Environment 124, 275-284. doi:10.1016/j.agee.2007.11.002

Rosin P, Rammler E (1933) Laws governing the fineness of powdered coal. Journal of the Institute of Fuel 7, 29-36.

Scanlon BR, Andraski BJ, Bilskie J (2002) Miscellaneous methods for measuring matric or water potential. In 'Methods of soil analysis. Part 4. Physical methods'. (Eds JH Dane, GC Topp) pp. 643-670. (Soil Science Society of America, Inc.: Madison, WI)

Schjonning P (1986) Shear strength determination in undisturbed soil at controlled water potential. Soil & Tillage Research 8, 171 179. doi: 10.1016/0167-1987(86)90332-6

Schjonning P (1995) Measurements of soil physical properties in the Askov long-term experiments. In 'The Askov long-term experiments on animal manure and mineral fertilizers'. (Eds BT Christensen, UT Trentemoller) pp. 125 135. (Danish Institute of Plant and Soil Science: Tjele, Denmark)

Schjonning P, Thomsen IK (2013) Shallow tillage effects on soil properties for temperate-region hard-setting soils. Soil & Tillage Research 132, 12-20. doi: 10.1016/j.sti 11.2013.04.006

Schjonning P, Christensen BT, Carstensen B (1994) Physical and chemical properties of a sandy loam receiving animal manure, mineral fertilizer or no fertilizer for 90 years. European Journal of Soil Science 45, 257 268. doi: 10.1111/j.1365-2389.1994.tb00508.x

Schjonning P, Iversen BV, Munkholm LJ, Labouriau R, Jacobsen OH (2005) Pore characteristics and hydraulic properties of a sandy loam supplied for a century with either animal manure or mineral fertilizers. Soil Use and Management 21, 265-275. doi: 10.1111 /j.1475-2743.2005. tb00398.x

Schjonning P, Munkholm LJ, Elmholt S, Olesen JE (2007) Organic matter and soil tilth in arable farming: management makes a difference within 5-6 years. Agriculture, Ecosystems & Environment 122, 157 172. doi: 10.1016/j.agee.2006.12.029

Schjonning P, de Jonge LW, Munkholm LJ, Moldrup P, Christensen BT, Olesen JE (2012) Clay dispersibility and soil friability--testing the soil clay-to-carbon saturation concept. Vadose Zone Journal 11, 1. doi: 10.2136/vzj2011.0067

Schjonning P, McBride RA, Keller T, Obour PB (2017) Predicting soil particle density from clay and soil organic matter contents. Geoderma 286, 83-87. doi:10.1016/j.geoderma.2016.10.020

Soil Survey Staff (2014) 'Keys to soil taxonomy.' 12th edn (USDA-Natural Resources Conservation Service: Washington, DC)

Vendelboe AL, Schjanning P, Moldrup P, Jin Y, Merbach I, de Jonge LW (2012) Colloid release from differently managed loess soil. Soil Science 177, 301-309. doi: 10.1097/SS.0b013e3182506dd3

Vomocil J A, Chancellor WJ (1969) Energy requirements for breaking soil samples. Transactions of the ASAE 12, 375-388. doi: 10.13031/ 2013.38846

Watts CW, Dexter AR (1998) Soil friability: theory, measurement and the effects of management and organic carbon content. European Journal of Soil Science 49, 73 84. doi:10.1046/j.1365-2389.1998.00129.x

Watts CW, Dexter AR, Dumitru E, Canarache A (1996) Structural stability of two Romanian soils as influenced by management practices. Land Degradation & Development 7, 217 238. doi: 10.1002/(SICI) 1099-145X(199609)7:3<2I7::A!D-LDR226>3.0.CO;2-B

Johannes Lund Jensen (A,B), Per Schjonning (A), Bent T. Christensen (A), and Lars Juhl Munkholm (A)

(A) Department of Agroecology, Aarhus University, Blichers Alle 20, DK-8830 Tjele, Denmark.

(B) Corresponding author. Email:

Caption: Fig. 1. Distribution of plots in the B5 field and sampling area within a plot (distances in meters). The subdivision in blocks is indicated with dotted lines. For key to treatments, see Table 1.

Caption: Fig. 2. Long-term fertilisation effects on (A) clay dispersibility and (#) percentage water-stable aggregates. The analysis was performed on field-moist soil (-250hPa). Letters denote statistical significance at P<0.05. For key to treatments, see Table 1.

Caption: Fig. 3. Shear strength at 100 kPa normal load ([[tau].sub.100]) as a function of soil porosity for the 12 experimental plots. The shear strength analysis was performed on soil cores drained to 300 hPa. The linear regression line, coefficient of determination and the significance level for all plots are indicated. For key to treatments, see Table 1.

Caption: Fig. 4. Log-transformed indices of friability based on either tensile strength ([F.sub.Y]) or specific rupture energy [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] of remoulded aggregates which were field-moist (RFM) or air-dry (RAD) when re-moistened, natural aggregates in a field-moist state (NFM), at field capacity (NFC) or air-dry (NAD) for each individual plot. The linear regression line, coefficient of determination and the significance level for natural aggregates (long dashed line) and remoulded aggregates (dotted line) are indicated. The y = x line is indicated.
Table 1. Treatment effects on basic soil characteristics and dry matter
(DM) yield of crop biomass. Within rows, letters denote statistical
significance at P<0.05. UNF, unfertilised; 1/2 NPK, 1/2 mineral
fertiliser; 1NPK, 1 mineral fertiliser; 1 1/2 AM, 1 1/2 animal manure

                                               UNF      1/2 NPK

Soil organic carbon                           0.92a      1,04b
  (g 100 [g.sup.-1])
Texture (A)
  Clay <2[micro]m (g 100 [g.sup.-1])           9.2        9.5
  Silt 2 20[micro]m (g 100 [g.sup.-1])         9.4        10.0
  Sand 20 63[micro]m (g 100 [g.sup.-1])        16.8       17.1
  Sand 63 125[micro]m (g 100 [g.sup.-1])       14.6       13.6
  Sand 125 200[micro]m (g 100[g.sup.-1])       13.6       13.6
  Sand 200 500[micro]m (g 100 [g.sup.-1])      28.9       28.9
  Sand 0.5-2mm (g 100 [g.sup.-1])              7.5        7.3
Rosin Rammler parameters (B)
  [alpha] ([micro]m)                           185        182
  [beta]                                       0.77       0.74
Mineral-to-carbon ratios
  Clay (<2 [micro]m)/SOC                      10.1b       9.2b
  Clay 1 silt (<20[micro]m)/SOC               20.5b      18.9b
Bulk density (gem                             1.54a      1.51a
Specific surface area                          19.5       23.2
  ([m.sup.2] [g.sup.-1])
pH (Ca[Cl.sub.2])                              6.3c      5.8ab
DM yield of crop biomass1
    (Mg [ha.sup.-1])
  Winter wheat                                 3.Id       9.0c
  Silage maize                                 2.0d       7.8c
  Spring barley                                2.0d       5.4c
  Grass-clover                                 4.0d       7.0c
  Annual average                               2.8d       7.3c

                                               INPK     1 1/2 AM

Soil organic carbon                           1,09b      1.29c
  (g 100 [g.sup.-1])
Texture (A)
  Clay <2[micro]m (g 100 [g.sup.-1])           9.4        9.6
  Silt 2 20[micro]m (g 100 [g.sup.-1])         9.4        9.8
  Sand 20 63[micro]m (g 100 [g.sup.-1])        17.7       16.8
  Sand 63 125[micro]m (g 100 [g.sup.-1])       14.3       14.0
  Sand 125 200[micro]m (g 100[g.sup.-1])       13.6       13.3
  Sand 200 500[micro]m (g 100 [g.sup.-1])      28.1       28.9
  Sand 0.5-2mm (g 100 [g.sup.-1])              7.6        7.7
Rosin Rammler parameters (B)
  [alpha] ([micro]m)                           181        184
  [beta]                                       0.75       0.74
Mineral-to-carbon ratios
  Clay (<2 [micro]m)/SOC                      8.7ab       7.5a
  Clay 1 silt (<20[micro]m)/SOC               17.4ab     15.1a
Bulk density (gem                             1.41b      1.42b
Specific surface area                          21.0       21.5
  ([m.sup.2] [g.sup.-1])
pH (Ca[Cl.sub.2])                              5.5a      6.0bc
DM yield of crop biomass1
    (Mg [ha.sup.-1])
  Winter wheat                                1 L8a      10.7b
  Silage maize                                10.4b      12.4a
  Spring barley                                6.7b       7.1a
  Grass-clover                                 7.9b      10.6a
  Annual average                               9.2b      10.2a

(A) g 100g 1 of mineral fraction.

(B) Calculated by cqn 2 in Rosin and Rammler (1933). a indicates the
coarseness of particles and [beta] indicates the spread of particle

(C) DM yield of crop biomass in the period 2006 2015. Grass-clover and
spring barley are average of two years, whereas silage maize and winter
wheat are average of three years.

Table 2. Long-term fertilisation effects on aggregate strength
parameters: tensile strength, Y (kPa), rupture energy, [E.sub.sp] (J
[kg.sup.-1]), friability (F) and workability (W) based on either Y or
[E.sub.sp] of natural aggregates at field capacity (NFC), Held-moist
(NFM) or air-dry (NAD) and remoulded aggregates which were
field-moist (RFM) or air-dry (RAD) when re-moistened

Y and [E.sub.sp] were calculated as geometric means. For key to
treatments, see Table 1. Within rows, letters denote statistical
significance at P< 0.05

Strength parameter    STATE      UNF     1/2NPK    1NPK    1 1/2 AM

Y                      NFC       4.5      5.7      5.6       6.2
                       NFM       5.5      7.0      5.8       7.2
                       NADA      75ab     103c     76b       57a
                       RFM       270      280      250       232
                       RAD       251      271      211       202
                     NAD/NFMA    13.%    15.1b    13.6b     7.9a
[E.sub.sp]             NFC       0.15     0.24     0.22     0.27
                       NFM       0.29     0.38     0.39     0.39
                       NAD       2.5      3.6      2.6       2.5
                       RFM       4.6      5.0      4.2       4.4
                       RAD       4.1      4.7      3.5       3.4
F(Y)                   NFC       0.46     0.40     0.37     0.42
                       NFM       0.32     0.36     0.39     0.41
                       NAD       0.37     0.50     0.53     0.55
                       RFM       0.11     0.10     0.10     0.11
                       RAD       0.07     0.09     0.10     0.10
[E.sub.sp]             NFC       0.66     0.61     0.58     0.62
                       NFM       0.69     0.55     0.53     0.58
                       NAD       0.60     0.64     0.69     0.92
                       RFM       0.14     0.14     0.18     0.18
                       RAD       0.13     0.13     0.18     0.16
W(Y)                   NFC       0.11     0.07     0.07     0.07
                       NFM       0.06     0.05     0.07     0.06
                       NAD      0.006    0.005    0.007     0.009
                       RFM      0.0004   0.0004   0.0004   0.0005
                       RAD      0.0003   0.0003   0.0005   0.0005
W([E.sub.sp])          NFC       5.1      2.7      2.7       2.4
                       NFM       2.6      1.4      1.4       1.5
                       NAD       0.27     0.17     0.27     0.39
                       RFM      0.032    0.029    0.044     0.042
                       RAD      0.032    0.029    0.053     0.049

(A) Ratio between Y of NAD and NFM aggregates. Clay is included as a
co-variable because it is significant in itself and makes the
treatment effect significant.

Table 3. Long-term fertilisation effects on soil cohesion, friction and
strain estimated from the annulus shear strength data

Soil water content was adjusted to 300 hPa. For key to treatments, see
Table 1. Within rows, letters denote statistical significance at P<0.05

                                  UNF    1/2NPK   1NPK    L 1/2AM

Cohesion (A) (kPa)               72.2b   63.8b    41.1a    41.9a
Internal friction (tan [phi])    0.42     0.47    0.51      0.5
Strain (%)                       5.95     6.01    6.06     6.42

(A) Clay is included as a co-variable because it is significant in
itself and makes the treatment effect significant.

Table 4. Long-term fertilisation effects 011 porosity in three pore
size classes, total porosity, air permeability

([k.sub.a]) and specific permeability (SP) [k.sub.a] and SP were
calculated as geometric means. For key to treatments, sec Table 1.
Within rows, letters denote statistical significance at P<0.05

Treatment                   Porosity in pore size classes
            <0.2 ([micro]   m 0.2-30[micro]m   >30[micro]m (A)

UNF             0.066            0.150a            0.193a
1/2NPK          0.069            0.165b            0.192a
1NPK            0.064            0.161b            0.239b
1 1/2AM         0.072            0.177c            0.214ab

Treatment                  [k.sub.a]             SP
            Total    30 hPa    100 hPa    30 hPa    100 hPa
                     [micro][m.sup.2])    ([micro][m.sup.2])

UNF         0.414a    8.1ab      13.9      92ab        89
1/2NPK      0.426a    5.2a       10.0       60a        69
1NPK        0.464b    26.9b      39.8      229b       209
1 1/2AM     0.459b    22.4b      33.9      233b       206

Clay is included as a co-variable because it is significant in itself
and makes the treatment effect significant.
COPYRIGHT 2017 CSIRO Publishing
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2017 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Jensen, Johannes Lund; Schjonning, Per; Christensen, Bent T.; Munkholm, Lars Juhl
Publication:Soil Research
Article Type:Report
Geographic Code:4EUDE
Date:Jul 1, 2017
Previous Article:Prediction of soil properties by using geographically weighted regression at a regional scale.
Next Article:The interaction between soil pH and phosphorus for wheat yield and the impact of lime-induced changes to soil aluminium and potassium.

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters