Effect of organic residues on soil properties of loamy topsoil of haplic Luvisol in Northern Germany

For the laboratory research, disturbed soil material was sampled from the Ap horizon (0–0.3 m in depth) of an agricultural-used and glacial till-derived haplic Luvisol (horizon sequence: Ap/E/Bt/Bw/C) (IUSS Working Group WRB, 2014), located at the research farm in Hohenschulen (54°31′28″N, 9°98′35″E) in Northern Germany. The silage from maize, sugar beet, and winter wheat were derived from the biogas plant in Schleswig-Holstein. The liquid digestates (i) 80% maize and 20% sugar beet (80m20b), (ii) 20% maize and 80% sugar beet (20m80b), and (iii) 20% winter wheat and 80% sugar beet (20w80b) were generated in a batch fermentation process. The compost (com) was produced by the local composting facility in Schleswig-Holstein, made out of shrub chippings (Beck-Broichsitter et al., 2018), whereas the sewage sludge (sl) and the associated chemical properties were derived from the municipal waste-water treatment plant in Schleswig-Holstein.

The texture of the soil material was classified as loam (FAO, 2006) consisting of 550 g kg⁻¹ sand, 300 g kg⁻¹ silt, and 150 g kg⁻¹ clay with an OC content of 11.96 g kg⁻¹, and pH of 6.79. The organic residues indicate pH values between 7.47 and 8.08 and, for example, total nitrogen contents between 1.2 and 4.0 kg m⁻³ organic mass⁻¹ (Table 1).

The organic residues were air dried, sieved (≤2 mm), and mechanically mixed with the loam (θ of approx. 0.1 cm³ cm⁻³) to simulate the annual application rates for fertilizer in 0.3-m Ap horizon with 30 Mg dry mass ha⁻¹ for compost and 30 m³ moist mass ha⁻¹ for digestates and sewage sludge. After the application of the organic residues, particle density (ρ_s; g cm⁻³), using pycnometer method; OC (g kg⁻¹), using coulometric carbon dioxide (CO₂) measurement; potential cation exchange capacity, CEC_pot (cmolc kg⁻¹), including natrium (Na), potassium (K), magnesium (Mg), and calcium (Ca), using barium chloride method; texture, using combined sieve and pipette method; soil pH values (in 0.01M CaCl₂ solution); and saturated hydraulic conductivity, Ks (cm d⁻¹), using by steady-state flow method were analysed following Hartge and Horn (2016).

Furthermore, the loam and the residue mixtures were compacted to a dry bulk density, ρ_b (g cm⁻³) of approximately 1.45 g cm⁻³ by a load frame (Instron 8871, Norwood, USA) with a pressing force of 5 kN, resulting in 8–10 soil cores (diameter: 5.5 cm; height: 4 cm) each.

The volumetric water content (θ_v) for the different drying stages was from the soil cores with 8–10 replicates per depth by a combined pressure plate (saturated, −60 hPa, and −300 hPa) and ceramic vacuum outflow method (−15,000 hPa) as well as oven dried for 24 hours at 105°C. The air-filled porosity, ε_a, was calculated as follows: 1εa=[(1−ρbρs)−θv]\varepsilon _{\rm{a}} = \left[ {\left( {1 - {{\rho _{\rm{b}} } \over {\rho _{\rm{s}} }}} \right) - \theta _{\rm{v}} } \right] where ρ_b is the dry bulk density (g cm⁻³), ρ_s is the particle density (g cm⁻³), here a value of 2.65 g cm⁻³ for ρ_s was assumed for the quartz-dominated soil, and θ_v is the volumetric water content (cm³ cm⁻³).

The soil porosity, ε, was calculated as: 2ε=1−ρbρs\varepsilon = 1 - {{\rho _{\rm{b}} } \over {\rho _{\rm{s}} }}

The AC (cm³ cm⁻³) and the plant AWC (cm³ cm⁻³) were calculated using the following equations: 3AC=ε−θ−60hpa{\rm{AC}} = \varepsilon - \theta _{ - 60\;{\rm{hpa}}} 4AWC=θ−60hpa−θ−15000hpa{\rm{AWC}} = \theta _{ - 60\;{\rm{hpa}}} - \theta _{ - 15000\;{\rm{hpa}}} where θ_{−60 hPa} and θ_{−15000 hPa} correspond to the water content at pressure heads, h, of −60 and −15,000 hPa, respectively. The German soil classification system (Ad-hoc-AG-Boden, 2005) was used to classify the AC and AWC values.

The air conductivity, k_l, was simultaneously determined from the same samples (100 cm³) as used for measuring soil water retention characteristics at −60, −300, and −15,000 hPa and in dry stage (105°C, 24 h) using an air flow meter with different scales between 0.1 and 10 L m⁻¹ (Key Instruments, Trevor, USA) (for more details, see Zhai and Horn, 2018): 5kl=ρ1⋅g⋅ΔV⋅ΔlΔt⋅Δp⋅ΔA{\rm{k}}_{\rm{l}}\, = \rho _1 \cdot {\rm{g}} \cdot {{\Delta {\rm{V}} \cdot \Delta {\rm{l}}} \over {\Delta {\rm{t}} \cdot \Delta {\rm{p}} \cdot \Delta {\rm{A}}}} where ρ_l is the air density (kg m⁻³), Δρ is the flow pressure (hPa), ΔV is the volume of air flow (m³) through the soil sample during time Δt (min), and g is the acceleration of gravity (m s⁻¹). The air density, ρ_l, was determined to calculate the kl values, whereas the atmospheric pressure and the temperature were obtained simultaneously (Dörner and Horn, 2006): 6ρl=ρn*273.15*ρL1013*(273.15+T)\rho _{\rm{l}} = \rho _n *{{273.15*\rho _{\rm{L}} } \over {1013*\left( {273.15 + {\rm{T}}} \right)}} where ρ_l is the air density (kg m⁻³), ρ_n is density of the air (1.293 kg m⁻³), ρ_L is air pressure (mbar), and T is laboratory temperature (°C) for each measurement time.

In addition, the air permeability, k_a, was calculated based on the air conductivity, k_l (cm s⁻¹), considering Darcy's law with the following formula (Ball et al., 1981; Dörner and Horn, 2006): 7ka=kl⋅ηρl⋅g{\rm{k}}_{\rm{a}} = {{{\rm{k}}_{\rm{l}}\, \cdot \eta } \over {{\rm{\rho }}_{\rm{l}} \,\cdot {\rm{g}}}} where η is the air viscosity (g s⁻¹ cm⁻¹).

The pore continuity was determined using the c₂ and c₃ indices as proposed by Groenevelt et al. (1984) and Zhai and Horn (2018). The indices are calculated as relationship between k_a and ε_a and also ε_a² in the following form (Ball et al., 1988; Dörner and Horn, 2006): 8c2=kaεa{\rm{c}}_2 = {{{\rm{k}}_{\rm{a}} } \over {\varepsilon _{\rm{a}} }}9c3=kaεa2{\rm{c}}_3 = {{{\rm{k}}_{\rm{a}} } \over {\varepsilon _{\rm{a}}^2 }}

Furthermore, soils with similar pore size distribution and porosity continuity have analog c₂ values, whereas soils with similar c₃ values only have analog pore size distribution (Groenevelt et al., 1984). In additionally, differences between c₂ and c₃ are related to differences in the pore continuity, independent from the pore size distribution (Dörner and Horn, 2006).

The following exponential relation between k_a and ε_a was suggested by Ball et al. (1988): 10log⁡(ka)=log⁡(M)⋅Nlog⁡(εa)\log ({\rm{k}}_{\rm{a}} ) = \log ({\rm{M}}) \cdot {\rm{N}}\log (\varepsilon _{\rm{a}} ) where M is an empirical parameter. N is the pore continuity index, which reflects the increase in k_a with increasing ε_a or the decrease in the pore tortuosity and surface area with increasing fraction of pores available to flow (Ball et al., 1988). In addition, the blocked air-filled pore space (ε_b, cm³ cm⁻³), is defined as follows: 11εb=10(−log⁡M)N\varepsilon _{\rm{b}} = 10{{( - \log {\rm{M}})} \over {\rm{N}}}

The statistical software R (R Development Core Team, 2014) was used to evaluate the data. The data were tested for normal distribution and heteroscedasticity on the Shapiro–Wilk test and graphical residue analysis. An analysis of variance (ANOVA) was conducted with p<0.05 followed by Tukey's HSD (honestly significant difference) test (*p≤0.05, **p≤0.01, ***p≤0.001, and ****p≤0.0001) following Hasler and Horton (2008) to evaluate the differences between loam and the loam residue mixture in basic soil characteristics (see Table 2), AC, plant AWC (see Table 3), and their interaction terms (twofold and threefold), respectively. The coefficient of determination (r²) is an index for the goodness of fit.

The results show that the OC content of the loam was improved through application of organic residue (12.4 up to 14.3 g kg⁻¹). There are very small differences in weak acidic pH values between 6.59 and 6.87, whereas application of compost (com) increases and application of digestate decreases the pH value (Table 2). The calcium (Ca) and especially the potassium (K) content significantly increased through application of organic residue from 0.64 up to 1.24 cmolc kg⁻¹. The C/N ratios between 9 and 10 indicate easily decomposable organic substances and, therefore, a rapid available nutrient source for plants.

The results indicate medium to high total porosities and the application of organic residues slightly decreased the porosity values, except for 20m80b that in turn increased the AC value of up to 0.191 cm³ cm⁻³ and decreased the AWC value of down to 0.111 cm³ cm⁻³ compared to loam, while 20w80b shows an opposite trend (Table 3). The AC values can be classified as medium to high, while the AWC values are at medium to low level. The Ks values between 40 cm d⁻¹ and 79 cm d⁻¹ can be classified as high without any significant changes compared to the loam.

In the range of ε_a values between 0.14 and 0.22 cm³ cm⁻³, the k_a values are nearly identical, whereas the k_a values increase up to 433 μm² for compost (com) at the dry stage between ε_a values of 0.45 and 0.48 cm³ cm⁻³ (Figure 1).

Figure 1

The c₂ and c₃ values in the range between −60 and −15,000 hPa show significant differences between the loam and the loam residue mixtures resulting in differences in pore size distribution and porosity continuity (Table 4). On the other side, the c₂ and c₃ values in the dry stage (105°C) are not very pronounced, resulting in a similar pore continuity and pore size distribution.

For better understanding of the model parameters, M, N, and ε_b, the k_a and ε_a values were fitted with Eq. 10 and ε_b was estimated with Eq. 11. The fitted parameters for loam and organic residue mixtures are listed in Table 5. A positive linear k_a/ε_a relationship (r²: 0.96–0.99) was found, and the blocked air-filled pore space, ε_b, of the loam with 0.0079 cm³ cm⁻³ was not significantly affected through the application of organic residue.

The results of the study presented in Table 2 indicate lower nitrogen contents, N_total, of 1.3 g kg⁻¹ after the application of compost and sewage sludge compared to the digestates with N_total values between 1.5 and 1.6 g kg⁻¹ through treatment-derived NH₃ and NH₄⁺ losses (Möller and Müller, 2012; Stoknes et al., 2016). The nutrient contents (Na, K, Mg, and Ca) and, therefore, the cation exchange capacity, CEC_pot, were higher after the application of organic residues with values between 9.6 and 11.1 cmol_c kg⁻¹ than for the loam with 9.4 cmol_c kg⁻¹ as also proposed by Ojeda et al. (2015). The increase in pH values through the application of compost considering the initial pH of 8.1; thus, this liming effect was confirmed in several studies (Ojeda et al., 2015), also for other organic residues (e.g., biochar). The composition of organic matter was not directly investigated, but Tambone et al. (2010) proposed that the chemical composition did not greatly contribute to compare digestates with compost or sewage sludge. It should also be taken into account that the dose of the applied organic residues affects its use as soil conditioner and fertilizer each (Govasmark et al., 2011) and the additional supply of OC that contains different numbers of polar and non-polar functional groups can negatively affect the wettability of the loam. Thus, also the local connectivity of flow pathways is reduced (Beck-Broichsitter et al., 2020b). Furthermore, the potential risk of soil acidification after organic residue amendment (Voelkner et al., 2015a) and the effect on wettability and dispersion (Voelkner et al., 2015b) of aggregated structured soils compared to the initially homogenized loam used in this study should be validated under field conditions.

The results presented in Table 3 indicate that the application of organic residues increased the ACs and decreased the plant AWC especially in case of digestate containing 20% maize and 80% sugar beet (20m80b). The results regarding the AWC values are contrary to the initial hypotheses and findings of Ojeda et al. (2015) and Beck-Broichsitter et al. (2018), except for digestate containing 20% wheat and 80% sugar beet, that slightly increased the AWC of the loam. Thus, the application of organic residues can potentially compensate low air capacities (Jasinska et al., 2006), whereas the decrease in AWC values can be attributed to the decrease in volume fraction of medium pores (−300 < h < −15,000 hPa) through homogenization as stated by Hu et al. (2009). However, wheat-containing digestate (20w80b) seems to slightly improve or not deteriorating the AWC as hypothesized before. It should be noted that the increase in the air capacities indicates an increase in the volume fraction of pores available for air flow or soil aeration, but it is not identical with the accessibility of oxygen for plant roots (Reszkowska et al., 2011; Beck-Broichsitter et al., 2020a).

The application of organic residues, however, significantly lowered the pore continuity at pressure heads of −60 and −300 hPa, except digestate 20w80b. The air-filled coarse pores may not contribute to the flow when they are discontinuous (Lipiec et al., 2003; Dörner and Horn, 2006). In the dry stage, the pore continuity indices presented in Table 4 show no significant differences; thus, the formation of shrinkage cracks seems to become more important for the connectivity of the pore system (Beck-Broichsitter et al., 2020b). However, compared to aggregated, structured soils in the field, the initially homogenized samples in this study tend to a higher shrinkage crack formation potential (Beck-Broichsitter et al., 2018) resulting in a comparatively higher air-filled porosity, more pronounced preferential flow paths and, therefore, overestimated air permeabilities. By the way, the slope factor N derived by Eq. 10 was significantly lower for digestate 20w80b than for loam and other residues; thus, the pore connectivity increases slowly with increase in air-filled pore porosity (Ball et al., 1988).

The blocked air-filled porosities presented in Table 5, between 0.0072 and 0.0084 cm³ cm⁻³, are nearly equal and on a very low level compared to the porosities in Table 2; thus, almost all pores of loam and loam residues mixtures are available for gas flow as also proposed for tilled topsoils (Dörner et al., 2012). This may be related to the increase in OC content that contributes a better quality of porous media (Assis et al., 2016). On the other side, tillage-induced homogenization can interrupt the functional pore system (Petersen et al., 2008), and compared to non-tillage conditions, the blocked air-filled porosity could be higher (Dörner and Horn, 2006; Dörner et al., 2012). In terms of the study results, the initial re-compaction of the soil cores to 1.45 g cm⁻³ may positively affected the rearrangement of the soil particles and pore connectivity (Horn et al., 2014) that could be another reason for the very low ε_b values.