Soil sulfur availability due to mineralization: Soil amended with biogas residues

Sulfur, an essential element for plant growth, has received lesser attention than it deserves. Current inputs of sulfur to agricultural soils from atmospheric deposition have reduced to less than the amount sulfur required by most crops. In such soils the release of sulfur from organic matter is vital for the supply of sulfur. A pot experiment without plant (incubation study) in a green house was conducted at the experimental station Dürnast TU Weihenstephan to investigate the potential of biogas residues as sources of available sulfur. The pot experiment comprised 20 different fertilization variants; 16 biogas residues and 3 mineral sulfur fertilizer variants, that is, S- 30, S-60, and S- 90 mg/pot, and a control (S-0), replicated trice. Pots were arranged in completely randomized design. Soil SO 42- -S and NO 3- -N contents upon were measured trice during the study period. Soil SO 42- -S and NO 3- -N content were significantly influenced by fertilizer treatments at all times. Generally, liquid biogas residues tend to show higher soil SO 42- -S and NO 3- -N content. Soil SO 42- -S content varied from 26.1 to 100.20 mg pot -1 (625-S, S-90); 30.4 to 100.4 mg pot -1 (621-S, 626-L) and 31.4 to 98.2 mg pot -1 (S-0, 626-L), in first, second and third sampling times, respectively. Soil NO 3- -N content also varied from 114.5 to 526.5 mg pot -1 (629-S, 616-L); 99.0 to 1054 mg pot -1 (620-S, 616-L); 114.8 to 1045.4 mg pot -1 (620-S, 616-L), in first, second and third sampling times, respectively. Biogas residues containing more than 0.1% S t in fresh matter and C org : S ratio lower than 30 could replace short term sulfate limitation.


INTRODUCTION
Sulfur, an essential element for plant growth, has received lesser attention than it deserves. In recent years, deficiency of sulfur in crops has increased worldwide and has been recognized as a constraint in crop production (Eriksen et al., 2004;Girma et al., 2005;Mascagni et al., 2008). The main reasons are: (1) the environmental control of sulfur dioxide emissions in industrial areas, (2) the increasing use of high-analysis, sulfur free fertilizers, (3) the decreasing use of sulfurcontaining pesticides and fungicides, (4) the adoption of high-yielding crop cultivars which demand a high fertility level and results in greater exploitation of soil reserve nutrients and removal of much larger quantities of nutrients in the harvested crop, and (5) the increased cropping intensity (intensive cultivation) (Scherer, 2001). microfauna, have been widely documented (Heimann Current inputs of sulfur from atmospheric deposition are less than 10 kg ha -1 in most Western European countries (Hu et al., 2005), which is less than the amounts of sulfur required by most crops (McGrath et al., 2002 as cited in *Corresponding author. E-mail: mewaelk@yahoo.com Scherer, 2009). In Western Europe, incidence of sulfur deficiency has increasingly been reported in Brassica over the last decade (Scherer, 2001). In Bavaria, in the last 15 years, sulfur emissions were reduced from 30 to 60 kg ha -1 (Guster and Tucher, 2001). Therefore, demand based sulfur fertilization for specific crops (20-70 kg ha -1 ) is becoming increasingly important.
In agricultural systems, where sulfur inputs from fertilizer and atmospheric deposition are low the release of sulfur from organic forms is important for the supply of sulfur to plants and a valuable parameter is the measurement of the potential contribution that the organic sulfur pool makes to plant-available sulfur, especially following the addition of organic matter to the soil. According to Bermuth (2008), despite the steadily growing number of biogas plants which produce biogas residues which are heterogeneous in their composition and used as organic fertilizers, the effect of the biogas residues as a source of available sulfur is largely unknown. Moreover, biogas residues are often mechanically separated into a liquid and a solid fraction. The effect of separation on the availability of S is largely unknown. This study was, therefore, conducted to determine available sulfur in an incubation experiment of soil amended with 16 different biogas residues by way of mineralization.

MATERIALS AND METHODS
Pot experiment in a green house was conducted at the experimental station Dürnast TU Weihenstephan to investigate the potential of biogas residues as sources of available sulfur.

Experimental set up
Top soil was collected from a permanent arable land near the experimental station Dürnast TU Weihenstephan. The soil was sandy loam (58% sand, 27% silt and 15% clay), pH 5.7, and contained Ct 0.80%; Nt 0.09%; P CAL 3.5 mg 100 g -1 ; kcal 9.6 mg 100 g -1 ; Mg CaCl2 11.2 mg 100 g -1 , which was percolated to wash out SO4 2--S and was air dried and mixed to create homogenous soil conditions which was then filled up to a 6 plus kg pot containing holes that ensure free drainage to impose treatments. The experiment comprised 20 different fertilization variants replicated trice setup in a Completely Randomized Design (CRD) in a green house. Sixteen of which included biogas residues from different provenances. Three of the variants were mineral sulfur fertilizer different rates vis., S-30, S-60, and S-90 mg/pot.

Biogas residues and fertilization
Biogas residues (BGR) with a broad spectrum of substrates composition were collected from different provenances. BGR were either used directly after the fermenter, that is, raw non-separated (NS) (AD7=628-NS) or mechanically separated by a screw separator into a liquid fraction (614-619+627) and a solid fraction (620-625+629). Biogas residue 626 was derived from the liquid fraction 627 after reverse osmosis. In all biogas plants the purification of the biogas from SO2 was performed by the addition of iron chloride and/or the injection of air. BGR were analyzed for dry matter content, St, SO4 -2 -S, Nt, NH4-N, P, K, Mg and Cl on fresh matter basis. Since BGR contained very different amounts of sulfur and had different dry matter, accurate adjustments were necessary to 90 mg S per pot application.BGR already contained certain amounts of N, P, K and Mg, which varied with residue types and thus based on pak choi requirement (experimental plant of parallel experiment-result not presented here) all pots received adequate amount of NH4NO3, Ca (H2PO4)2.H2O, KNO3 and Mg (NO3)2.6H2O to level out N (900 mg/pot), P (100 mg/pot), K (800 mg/pot), and Mg (100 mg/pot).
In order to impose mineral sulfur fertilization variant treatments (S-30, S-60 and S-90, mg of sulfur), sulfur was applied in the form of MgSO4 along with KNO3 as a compensatory fertilization of nitrogen for shooting. Each of the sixty pots were filled with approximately 6 kg of soil portion in such a way that the bottom layer contained 2.5 kg of soil, mechanically shaked to adjust the bulk density. The next layer (middle layer) was filled with another 2.5 kg of soil to which the fertilizer treatments were added and thoroughly mixed and was again shaked. The top layer was only 1 kg of soil, which was not shaked, to cover the fertilized layer.
Pots were regularly watered with distilled H2O according to 60 to 70% of the maximum water holding capacity (100% water holding capacity: 29.3% gravimetric soil water content) corresponding to a gravimetric soil water content of 17.6 to 20.5%. Pots were monitored for their SO4 2--S and NO3content.

Analysis of the biogas residues
Liquid, solid and raw BGR were analyzed for dry matter content (drying at 100°C after the addition of sand for the liquid residues), for pH value (with pH sensitive electrode after the addition of CaCl2 for the solid residues), for total contents of S, P, K and Mg after wet digestion with HNO3 and perchloric acid with an ICP-OES (Liberty, Varian, Mulgrave, Australia) using the lines of 182.034 nm for S, 213.618 nm for P, 769.896 nm for K, 285.213 for Mg), for total N after Kjeldahl digestion, steam distillation and titration with H2SO4, for NH4-N by steam distillation and titration with H2SO4, for Cland SO4 2--S after extraction with H2O dist. by ion chromatography (for further details see "soil analysis"). Organic carbon was determined by subtracting carbonate-C (measured following the Scheibler procedure, that is gas-volumetric measurement of CO2 release after the addition of HCl) from total C which was measured after freezedrying using an isotopic ratio mass spectrometer (Anca, SL 20-20, Europa Scientific, Crewe, UK).

Soil analysis
Soil samples were air dried and ground to pass a 2 mm sieve. All samples were extracted in 100 ml of CaCl2 (0.01mol/l) by shaking for 60 min on an end-over-end shaker. Extracts were filtered through a filter paper and were stored at -15°C prior to analysis. Thawed extracts were taken directly for SO4 2--S and NO3 --N determination by ion chromatography. Chromatographic conditions were: column AS4A; eluent: 1.8 mmol L -1 Na2CO3 / 1.7 mmol L -1 NaHCO3 pH 8-9; flow: 2 ml min -1 ; detection by conductivity.

Statistical analysis
Statistical significance of differences between treatments were determined by Tukey HSDa,b mean separation test for multiple comparisons after analysis of variance (ANOVA), using IBM SPSS 19. All tests were performed at a significance level of P < 0.05. Correlation analysis was also employed.

Soil sulfate (SO 4 2--S) content
Determination of available soil sulfur from biogas residue fertilizers is very important to gain better understanding of their mineral equivalence capability. SO 4 2--S content was significantly influenced by fertilizer treatments at all times. Generally, liquid biogas residues tend to show higher SO 4 2--S content. At first time sampling, that is 4 days after fertilizer application, SO 4 2--S content was significantly influenced by fertilizer treatments. Significantly highest SO 4 2--S content (100.2 mg pot -1 ) was recorded with S-90 mineral fertilization, which was on par with pots treated with 626-L and S-60 followed by pots treated with 627-L, 629-S, 628-NS, and S-30 where as the lowest SO 4 2--S content (26.0 mg pot -1 ) was recorded in pots fertilized with 625-S, which was on par with pots treated with 621-S , S-0, 624-S, 620-S, 623-S, and others.
Similarly, at second time sampling, that is, 22 days after fertilizer application, SO 4 2--S content was significantly influenced by fertilizer treatments. Significantly highest SO 4 2--S content (101.4 mg pot ) was recorded in pots treated with S-0 which was on par with pots treated with 621-S, 620-S, and others.
The relatively higher CaCl 2 -extractable SO 4 2--S status in 626-L and 627-L fertilized pots could be attributed to higher totals sulfur and lowers C org :S ratio of biogas residues. Table 2 was worked out from Table 1.
At each sampling time there was highly significant positive correlation between SO 4 2-S contents in soil and the S t contents of the biogas residues and a highly significant negative correlation to the C org : S t ratio of the biogas residues. Figure 1 depicts that changes in the soil SO 4 2--S content did not show clear pattern of mineralization in the course of the study period. However, 67% (622-S) up to 95% (615-L) of the SO 4 2--S found at the 3 rd sampling 30 days after fertilizer application were already present at the 1 st sampling 4 days after fertilizer application. The increase in soil SO 4 2--S content from the 2 nd (22 days after fertilizer application) to the 3 rd sampling was low and seemed to be more pronounced in the group of the solid biogas residue fractions (e.g. 620-S, 621-S, 623-S and  625-S). Figure 2 depicts the amounts of SO 4 2--S recovered as the proportion of S present as SO 4 2--S from the biogas residues (that is, minus the unfertilized control) in relation to the S applied as fertilizer (90 mg) from biogas treatments variants using 10 mM CaCl 2 in relation to C org :S t ratio. It is obvious that the relation was non-linear. None of the samples showed values significantly lower than zero.

Soil nitrate (NO 3 --N) content
Soil nitrate (NO 3 --N) content was significantly influenced by fertilizer treatments at all times. Similar to SO 4 2--S content, liquid biogas residues tend to show higher NO 3 --N content.
At first time sampling, NO 3 --N content was significantly influenced by fertilizer treatments. Significantly highest NO 3 --N content (526.6 mg pot -1 ) was recorded with 619-L fertilization, which was on par with pots treated with 618-L and 617-L followed by pots treated with 616-L where as the lowest NO 3 --N content was recorded in pots fertilized with 629-S, which was on par with pots treated with 620-S, 621-S, 626-L, 622-S, and 623-S. Also during second time sampling, NO 3 --N content was significantly influenced by the fertilizer treatments. Significantly highest NO 3 --N content was recorded with 616-L fertilization, which was on par with pots treated with 618-L, 619-L, 614-, 615-L and 617-L followed by pots treated with 627-L where as the lowest NO 3 --N content (99.0 mg pot -1 ) was recorded in pots fertilized with 620-S, which was on par with pots treated with 621-S, 629-S, 623-S, and others.
Similarly, in third time sampling, NO 3 --N content was significantly influenced by the fertilizer treatments.

DISCUSSION
The study of sulfur transformation is important for sulfur fertilizer optimization. It has been suggested that, the mineralization of sulfur involves both biological and biochemical mineralization (McGill and Cole, 1981). Biological mineralization is driven by the microbial need for organic C to provide energy, and S released as sulfate is a by-product when microorganisms break down organic compounds to acquire C for their energy metabolism. Biochemical mineralization is the release of sulfate from sulfate-esters through enzymatic hydrolysis.
Whereas the mineralization of C-bonded S is strictly dependent on microbial activity, sulfate-esters can be readily hydrolyzed by sulfatase enzymes in the soil, and biochemical mineralization is, therefore, controlled by the supply of S rather than the turnover of organic matter. The experiment showed biogas residues' clear differences presumably attributed to their different substrate composition and/or the efficiency of the biogas process. The substrate compositions were, actually, subject to different degradation rates during the fermentation process depending on the length of time, temperature and mixing ratio in the fermenter. The C org :S ratio of various biogas residues fertilizers ranged from 1 to 234. SO 4 2--S content measured in the unplanted soil was significantly influenced by the fertilizer treatments at all times. Generally, liquid biogas residues tended to show higher SO 4 2--S content. Among the biogas residue variants the mean available sulfur content varied from 26 to 86.3 mg pot -1 in first sampling, 30 to 101 mg pot -1 in second sampling, and 31-98 mg pot -1 in the third sampling.
About 67 to 95% of the SO 4 2--S measured in the third sampling were already present 4 days after the application of the biogas residues. As the SO 4 2--S applied by biogas residues was marginal (up to about 1 mg SO 4 2--S per pot), the rapid increase in soil SO 4 2--S may be due to oxidation of S 0 or S 2derived from the biogas purification process (Abatzoglou and Boivin, 2009). In soils under aerobic conditions these S forms will be readily oxidized to sulfate (Janzen and Ellert, 1998;Scherer, 2009). Although there is no information on the S forms present in the biogas residues tested here, it is interesting that sulfate formation seemed to be completed within 3 to 4 weeks (Figure 1).
The overall mineralization of SO 4 2--S 4 days, 22 days and 30 days after fertilizer application was strongly correlated with the organic C and total S (S t ) of the biogas residues (BGRs). Table 2 shows the extent to which available sulfur mineralized was strongly correlated with the total sulfur, that is, 0.68**, 0.70**, 0.74**; and Corg:S, that is, -0.83**, -0.82**, and -0.79** at first, second and third sampling, respectively. The fact that there exist significant negative correlation between the short term S availability and C: S ratio of the organic fertilizers agrees with the finding by Guster and Tucher (2001).
The biogas residue (BGR) L-626 happened to consistently record highest SO 4 2--S content (when compared with rest of BGRs) which was on par with the apparently highest SO 4 2--S recorded in the S-90 treated pots (86.3,100.2;101.4,96.2 and 98.2, 2--S content, respectively across the three samplings). The relatively higher S mineralization of the BGR L-626 was presumably attributed to its low C org content and high S t content (1.4 g/100 g, 0.169 g/100 g) ( Table 1). L-627, S-628 and S-629 also showed consistent and comparable high SO 4 2--S content throughout the study period.
Moreover, data for the amounts of SO 4 2--S recovered from all of the treatments using 10 mM CaCl 2 in relation to C org :S t ratio depicted in Figure 2 reveals that Srecovery increased with smaller C org :S t ratio at all times. This again substantiated the fact that mineralization of organic fertilizers increase as its C org :S ratio decreases. None of the samples showed values significantly lower than zero which implies there was no immobilization. Taking the third sampling time into consideration, recovery was from as high as 74% in L-626 to as low as only 4 % in 621-S. When compared against S-90 (81, 73 and 60%), except in the first sampling, S-recovery was higher in 626-L (66, 79 and 74.5%) which implies some organic biogas residue fertilizers could replace mineral fertilizer equivalent for sulfur nutrition or that at least no immobilization could indicate a potential source of available sulfur. S-recovery showed most of the solid BGRs almost did not release SO 4 2--S. Earlier studies have shown that the incorporation of organic residues with narrow C org :S ratio (< 200:1) can increase the contents of SO 4 2--S in the soil, whereas the incorporation of those organic residues which have wide C org :S ratio (>400: 1) may result in a net immobilization of soil inorganic-S (Barrow, 1960;Stevenson, 1986). In this study, despite the non-equivalency of the above reported cut off points for mineralization and immobilization the tendency matches. In the treatments using biogas residues, the C org :S ratio in the pot 626-L which recorded highest SO 4 2--S had a very low C org :S ratio (that is, organic carbon was by far lower in 626-L which is the biogas residue which was improved by reverse osmosis after the separation of the liquid fraction) whereas the pot with the least SO 4 2--S record had C org :S ratio of 234 (623-S) followed by 168 (624-S). Moreover, the N/S ratio showed significant inverse relationship with sulfur mineralization of r= -0.548 * , r= -0.575 * , r= -0.654 ** at first sampling, second sampling and third sampling, respectively. In agreement to this result, Kowalenko and Lowe (1975) reported that a high N: S ratio resulted in a decrease in mineralization of sulfur in soil sample during incubation.
Mineralized sulfur probably originates from both the sulfate-ester and the C-bonded S pool (Freney et al., 1975;McLaren et al., 1985) although C-O-S is considered more labile than C-S. The pot experiment without plants indicated mineralization-immobilization processes were occurring simultaneously resulting in minimal net changes in CaCl 2 -extractable SO 4 2--S concentrations. This is presumably because short term sulfur cycling evaluation involves small fraction of potentially available sulfur when mineralization may need longer time to show its effect. This goes in line to the finding by Saggar et al. (1981) that the short term supply of available sulfur upon mineralization is low.
There was significant difference in NO 3 --N content among fertilizer treatments at all times. BGRs L-619, L-618, L-617 and L-616 showed consistently higher NO 3 --N content. This, however, did not seem to relate to the Sfertilization variation. Higher NO 3 --N records tend to show strong positive correlation (r=0.851 ** ) to the amount of supplemental mineral NH 4 -N fertilizer applied to bring treatments to equal bases. And, thus, the increased changes in NO 3 --N content with time in many cases is seemingly due to nitrification rather than mineralization.
In conclusion it gives the impression sulfur limitation from less atmospheric deposition can in the short term be replaced by some biogas residues, given that the biogas residues are high in total sulfur content i.e., higher than about 0.1% S t in fresh matter and in C org :S ratio lower than 30.