The effect of applications of various forms of sulfur on the yields and quality of grass forage

Abstract

The objective of our study was to assess the effect of soil application of sulfur in ammonium sulfate and gypsum and of either soil or foliar application of elemental sulfur on yield of grass forage and its qualitative parameters. The effect of various forms of sulfur on the yields of grass forage and its qualitative parameters was explored in the form of a small plot experiment in the Bohemian-Moravian Uplands in 2004–2006 involving the following variants: 1) sulfur unfertilised control; 2) ammonium sulfate; 3) elemental sulfur; 4) gypsum; 5) foliar elemental sulfur. Sulfurous fertilisers and foliar elemental sulfur were applied to the soil in doses of 40 kg and 8 kg S per ha, respectively. Nitrogen applied in ammonium sulfate was added to all the other variants in the form of ammonium nitrate. Application of the fertilisers was repeated at the beginning of each vegetation season. The stand was cut twice in the course of vegetation. In the harvested biomass we assessed the content of sulfur, nitrogenous substances, and net energy of lactation.

The effect of various forms of sulfur on the grass biomass yields was not significant in either of the two cuts. Yields increased after sulfur fertilisation only in the 1st cut, especially after the application of sulfate sulfur and gypsum, and/or after foliar application of elemental sulfur. In the individual years the sulfur content in the biomass gradually increased significantly (0.17–0.23–0.29%). In the first year the sulfur content did not reach the critical deficiency limit (0.2%). Sulfur fertilisation increased the S concentration in grass forage in all the fertilised variants; the highest S content was detected in the variant where gypsum was applied (0.27%). No significant correlation was established between the values of water-soluble sulfur in the soil and the sulfur content in grass forage. Sulfur fertilisation had no significant effect on the N/S ratio, but was the highest in the variant not fertilised with sulfur and the lowest in the gypsum variant. The content of nitrogenous substances and net energy of lactation were significantly the highest after fertilisation with elemental sulfur and was related, among others, to the lowest yields of this variant. Sulfurous fertilisers did not significantly affect the exchangeable soil reaction and the highest content of water-soluble sulfur after the 2nd cut was seen in the gypsum-applied variant.

Keywords:

• Czech Republic
• grass forage
• nitrogen substance
• N/S ratio
• quality
• sulfur
• yields

Introduction

Sulfur deficiency has become widespread in many countries in Europe since the 1980s. This is due to a massive decrease in the atmospheric deposition of S, a change in the use of fertilisers that contain low S, and increased crop yields. As early as the 1960s and 1970s, sulfur deficiency appeared sporadically, at first in areas of low sulfur depositions, such as Norway and Ireland (Murphy & Boggan, 1988; Singh, 1994). In the early 1980s, S deficiency became common in intensive grass-production systems in Ireland (Murphy & Boggan, 1988; Murphy & O'Donnell, 1989), Scotland (Scott et al., 1983), and also later in England (Syers et al., 1987; Richards, 1990). In the 1980s sulfur deficiency became common in winter rape in a number of West European countries like France, Germany, Sweden, Denmark, and Great Britain (Merrien, 1987; Schnug, 1991; Knudsen & Pedersen, 1993; McGrath et al., 1996); in some countries at the same time and then especially in the 1990s, particularly in Germany and Great Britain, sulfur deficiency was also seen in cereals (Haneklaus et al., 1995; Zhao et al., 1999).

In the last 15 years a rapid decrease in the input of sulfur into the agri-ecosystems has been observed also in the Czech Republic (CR), particularly as a result of the great reduction in atmospheric depositions due to desulfurisation of the most important emitters of sulfur dioxide into the atmosphere. In recent years, after desulfurisation of thermal power plants, which fired brown coal of poor quality, and after the installation of gas in towns and villages, attention has been devoted to desulfurisation of fuel during oil refining.

In relative terms the decrease in sulfur depositions in the Czech Republic is almost 90%; in 1988 sulfur fall-out per hectare was 131 kg, while in 2003 it was only 14 kg of sulfur per hectare. According to Matula (2001) 69% of soils in the Czech Republic show various degrees of sulfur deficiency.

The problem of replenishing the missing sulfur in the CR was first handled in the more demanding crops, namely rape, later in wheat; it is gradually becoming relevant also in other agricultural crops, like for instance sugar-beet and barley. Little is known about the sulfur requirements of grasslands in the Czech Republic.

Sulfur deficiency in grasslands restricts the synthesis of protein in plants, and the nitrogen in plants accumulates as nonprotein N (NPN), such as nitrate, free amino acids, and amides in plant tissues (Kaiser & Weiss, 1997). Owing to sulfur deficiency both the quality of the harvested forage and the digestibility are poor (Pierzynski et al., 2000).

Sulfur may be applied either in the form of classical mineral fertilisers where sulfur is an accompanying anion (e.g., ammonium sulfate) or in the form of by-products from the chemical industry (elemental sulfur from oil refineries, gypsum from the production of titanium white etc.)

The objective of our study was to assess the effect of various forms of sulfur, namely sulfate in ammonium sulfate and gypsum, and elemental sulfur in the soil and as foliar applications on yields of grass forage and its qualitative parameters.

Material and methods

Site description

The experiment was carried out in the Bohemian-Moravian Uplands (Czech Republic) in 2004 at an altitude of 553 m. In the period 1971–2000, the mean annual air temperature was 6.9°C and the mean annual sum of precipitation amounted to 617 mm. Mean annual air temperatures in 2004, 2005, and 2006 were 6.8, 6.9, and 7.0°C, respectively. Mean annual sum of precipitation in 2004, 2005, and 2006 was 658, 566, and 625 mm, respectively (Figure 1). Soil profile: A_p–E_N–B_M–B/C–C. Particle-size Class: loamy to silt loam. Soil unit: Dystric Planosol. Table I shows the agrochemical properties of the soil.

Figure 1.  Climadiagram for the experimental station.

Table I. Agrochemical properties of the soil before establishing the experiment (0–20 cm).

	Cox (%)			Content of available nutrients (mg kg^-1)
Soil category	0–10 cm	10–20 cm	pH/ CaCl2	P	K	Ca	Mg	Swater-soluble
Moderately heavy soil	5.06	3.40	5.6	7	111	2777	315	14.2

The species composition of the grass stand under study in 2004, expressed as the proportions of dominant herb species in dry matter, was as follows: Arrhenatherum elatius (29%), Elytrigia repens (21%), Taraxacum officinale (15%), Trisetum flavescens (8%), Phleum pratense (7%), Agrostis capillaris (6%), Poa ssp. (4%), Festuca rubra (4%), and Dactylis glomerata (3%).

The soil showed weakly acid soil reaction. In this locality the content of available phosphorus is low, potassium is satisfactory, calcium good, and magnesium very high. According to Dutch criteria the level of water-soluble sulfur is satisfactory (Postma et al., 1999).

Experimental design

The experiment was established in three replications. The size of the experimental plots was 2 m×10 m (20 m²).

In this experiment, the following factors were evaluated: year (2004, 2005, and 2006), used fertilisers (unfertilised control, ammonium sulfate, elemental sulfur, gypsum and foliar application of elemental S) and order of cuts (first and second) – see Table II.

Table II. Experimental factors evaluated and factor level.

Experimental factor	Level of factor
Years	2004
	2005
	2006
Variants	Control unfertilised with sulfur
	Ammonium sulfate
	Elemental sulfur
	Gypsum
	Foliar elemental sulfur
Cuts	First
	Second

The small-plot field trial was established on 29 April 2004. In the two following years the S fertilisers were applied again in the same doses as in the first year, i.e. on 29 April 2005 and 28 April 2006. Basic fertilisation before establishment of the experiment was 50, 30, and 60 kg N, P, and K, respectively. Sulfurous fertilisers were applied in doses corresponding to 40 kg S per hectare and 8 kg per hectare of foliar elemental sulfur in the form of fertiliser ORIN basis S (defined below). The amount of nitrogen applied as ammonium sulfate corresponding to 35 kg ha⁻¹ was supplemented in all the variants in the form of ammonium nitrate. The dose of sulfur was on the upper limit recommended for European countries (Walker & Dawson, 2003).

Used fertilisers

Ammonium nitrate is a fertiliser with two forms of nitrogen. It contains 34% N, one half in the form of nitrate and one half as ammonium. It is physiologically neutral, well soluble in water, very hygroscopic, and was applied in the form of granules.

Ammonium sulfate is a fine crystalline fertiliser, particles mostly smaller than 2.0 mm, obtained during crystallisation and containing 20.88% nitrogen; 20.78% of the salt is ammonium, and 23.7% sulfur. For the experiment we used ammonium sulfate not treated superficially from Spolana Neratovice obtained as a by-product of caprolactam production.

Elemental sulfur is a dry finely ground sulfur from Poland. It contains 98% sulfur and 95.95% of the particles are smaller than 0.063 mm.

Pregips H (gypsum) is a pulverised fertiliser containing sulfur and calcium (14%S and 25% CaO); it is hydrated calcium sulfate (CaSO₄ ^.2 H₂O). It is produced by neutralisation of acid water removed from the production of titanium white by water suspension of very finely ground limestone and is slightly soluble in water, i.e. 2.6 g in one litre.

ORIN basis S comprises 80% of elemental sulfur in the form of a dispersible microgranulate with bentonite, making up an easily applicable suspension.

Measurements

The plots were harvested twice a year (Table III) with a lawn mower Model MF 120 (working width 1.2 m) with the harvested area being 12 m². Stubble height was 7 cm. All harvested biomass was weighed.

Table III. Harvest dates in the three years of the experiment.

Year	First cut	Second cut
2004	2 June	18 August
2005	15 June	18 August
2006	20 June	30 August

All soil samples were collected from the arable layer (0–20 cm). Exchangeable soil reaction was specified in a soil extract of 0.01 M CaCl₂ by potentiometer with a glass electrode against a reference calomel electrode (Zbiral, 2002). The content of available nutrients in the soil was determined in a soil extract obtained by Mehlich III extraction agent [CH₃CO₂H, NH₄NO₃, NH₄F, HNO₃, and ethylenediaminetetraacetic acid (EDTA)]. The amount of available phosphorus in the extract was determined by spectrophotometry and the content of available potassium, magnesium, and calcium was assessed by atomic absorption spectrophotometry (Zbiral, 2002). Water extraction of a 1/5 ratio for 16 hours on a rotary shaker preceded the determination of the content of water-soluble sulfur in the soil. The measurement proper was conducted using the Inductively Coupled Plasma-Optical Emission Spectroscopy (ICP-OES) method (Zbiral, 2002).

The herbage samples for chemical analyses consisted of stems and leaves of all the plants in the sward. To assess the sulfur content in grass forage the plant matter was mineralised with nitric acid and hydrogen peroxide. The measurement proper of the sulfur content was conducted using the method of optic emission spectrophotometry in induction-bound plasma (Zbiral, 2005).

To assess the content of nitrogenous substances and nett energy of lactation (NEL) the ground sample of dried forage was scanned on the NIRSystems 6500 apparatus in small round cells using the WINISI II program. Scanning was conducted in the reflectance regime and wavelength range of 1100–2500 nm and step 2 nm (Mika et al., 1998).

Statistical analysis

The results were statistically analysed by analysis of variance (ANOVA) and Fisher test operated under Statistika 6.0.

Results and discussion

The dry-matter yields (sum of two cuts) ranged between 3.56 and 10.75 t ha⁻¹. The weather conditions significantly affected the dry matter yields (see Table V). The differences between the years were significant (P>0.95). The yields were highest in 2006 and ranged between 8.13 and 10.75 t ha⁻¹ as precipitation in spring and summer was favourable; yields were lowest in 2005 when the sum of all cuts ranged between 3.56 and 5.09 t ha⁻¹. The factor year comprises both weather conditions and multiple effect of annual application of nitrogen and, particularly, sulfur. The share of the 1st and 2nd cuts in the total yields was 63 and 37%, respectively. The difference between the 1st and 2nd cut was statistically significant (P>0.95). Klessa et al. (1989) pointed out that deposition and some mineralisation in the winter months would increase the availability of S for sward performance in the 1st cut. In the summer months, the occurring dry period and missing mineralisation would lower the sward performance (Taube et al., 2000). The effect of sulfur fertilisation on dry-matter yields was not significant (P>0.95) and corresponds with the results of Hahtonen and Saarela (1995) in Scotland, who detected plant growth responses to supplementary sulfur as small, inconsistent, and statistically insignificant. According to Kowalenko (2004) it is not possible to monitor the results of yields in every cut. In agreement with results in the Republic of Ireland, the 2nd and 3rd cuts of silage tended to be more responsive to S than was the 1st cut. However, by the mid-1990s, even the 1st-cut silage in England was highly responsive to S (Zhao et al., 2003). In our field experiment the effect of different forms of sulfur on yields of grass biomass was insignificant (P>0.95) in both cuts (Table IV). Insignificant effects of interactions prove that the multiple effect of annual N and S fertilisation did not lead to significant differences among variants fertilised with various forms of sulfur (see Table V). Yields were seen to increase after sulfur fertilisation only in the 1st cut, i.e. always after the application of fertilisers, namely after the application of sulfate sulfur in ammonium sulfate and gypsum, and/or after foliar application of elemental sulfur.

Table IV. Yields of dry matter of grass forage in the individual cuts (t ha⁻¹)

	First Cut	Second Cut
	±s x	±s x
Year
2004	4.01^a±0.75	2.74^b±0.73
2005	2.78^b±0.65	1.40^a±0.51
2006	6.17^c±1.94	3.38^c±0.97
Variant
Control	4.02^a±1.72	2.84^a±1.38
Ammonium sulfate	4.59^a±2.68	2.54^a±1.47
Elemental sulfur	3.89^a±1.12	2.21^a±0.77
Gypsum	4.51^a±2.13	2.47^a±1.16
Foliar elemental sulfur	4.59^a±1.71	2.47^a±0.78
^aAverage values of the individual variants do not differ significantly (p<0.95) if they have the same superscript.

Table V. Analysis of variance of the characters.

		Mean square
Factor	D.f.	Yield of dry matter		Sulfur content in dry matter		N/S ratio		Yield of sulfur		Yield of nitrogen		Nitrogen substances content		Nett energy lactation content
Year	2	54.191	***	0.096031	***	17.467		683.341	***	28070.2	***	16339	***	1.568	***
Cut	1	74.085	***	0.027317	*	341.763	***	269.621	**	64930.5	***	15199	***	3.575	***
Variants of fertilisation	4	0.775	NS	0.012704	*	7.191	NS	47.509	NS	94.6	NS	590		0.131	NS
Year ×cut	2	5.408	*	0.006189	NS	69.067	***	47.711	NS	6001.1	***	9955	***	0.525	***
Year×variants of fert.	8	1.220	NS	0.007553	NS	8.179	NS	36.904	NS	712.8	NS	513	NS	0.070	NS
Cut×variants of fert.	4	0.709	NS	0.007731	NS	16.323	*	42.387	NS	153.3	NS	74	NS	0.033	NS
Year×cut×variant of fert.	8	0.432	NS	0.003714	NS	4.843	NS	29.255	NS	242.6	NS	51	NS	0.004	NS
Error	60	1.193		0.004856		6.288		25.239		574.1		260		0.063
D.f.=Degree of freedom. DM = Dry matter. * = α ≤ 0.05. ** = α ≤ 0.01. *** = α ≤ 0.001. NS = insignificant.

In the individual years (Table VI) the content of sulfur in the harvested biomass gradually increased significantly (0.17–0.23–0.29% in dry matter) and can be attributed, among other things, to the annually repeated sulfur fertilisation, i.e. its residual effects. The detected concentrations of sulfur in the plants correspond to results of analyses of grass forage in Finland (Hahtonen & Saarela, 1995). In 2004 the sulfur content in plants was even lower than 0.2% in dry matter; according to some authors (e.g., Lancaster et al., 1971; Bolton et al., 1976; Scott et al., 1983; Stevens, 1985; Keer et al., 1986) this is the critical limit of sulfur insufficiency. Comparisons of the variants showed that the sulfur contents in grass forage of the unfertilised control variant were on this level. Sulfur fertilisation increased its concentration in grass forage of all the fertilised variants, as Kowalenko (2004) and Hahtonen and Saarela (1995) have also reported. The significantly highest sulfur content was discovered in the gypsum-applied variant (0.27% in dry matter) and corresponds with the results of Spears et al. (1985). On the other hand, the variant where elemental sulfur was applied to the soil provided the lowest yields. Well-timed supply of available sulfate sulfur to the plants is dependent on the speed of oxidation of the elemental sulfur. On the experimental site oxidation could have been slowed down by a higher level of ground water and thus lack of air for the oxidation bacteria. The second factor was the very low level of phosphorus in the soil; according to Sholeh et al. (1997) phosphates in the soil are essential for oxidation of elemental sulfur. No significant correlation was established between the value of water-soluble sulfur in the soil and the sulfur content in grass forage as reported by Hahtonen and Saarela (1995).

Table VI. Average values of the characters and significance of their differences evaluated by Fisher's Test.

			Yield of dry matter (t ha^−1)	Sulfur content in dry matter (%)	N/S ratio	Yield of sulphur (kg ha^−1)	Yield of nitrogen (kg ha^−1)	Nitrogen substances content (g kg^−1 of dry matter)	Nett energy lactation content (MJ kg^−1 of dry matter)
Factor	Level of factor	n	±s x	±s x	±s x	±s x	±s x	±s x	±s x
Year	2004	30	3.37^b±0.97	0.17^a±0.02	9.2^ab±3.86	5.8^a±1.71	55.3^a±31.60	128.54^a±29.65	5.22^a±0.36
	2005	30	2.09^a±0.91	0.23^b±0.06	9.7^b±3.04	4.7^a±1.79	43.2^a±17.48	131.60^a±11.69	5.67^b±0.16
	2006	30	4.78^c±2.07	0.29^c±0.11	8.2^a±3.49	13.4^b±9.45	101.2^b±53.24	170.40^b±31.31	5.43^c±0.42
Variants of fertilisation	Unfertilised control	18	3.43^a±1.63	0.20^a±0.05	10.0^a±3.59	6.6^a±3.44	65.2^a±41.81	138.09^a±25.50	5.36^a±0.34
	Ammonium sulfate	18	3.56^a±2.35	0.22^a±0.05	8.7^a±2.73	7.7^ab±5.35	69.8^a±58.24	144.31^ab±32.83	5.42^ab±0.34
	Elemental sulfur	18	3.05^a±1.27	0.23^ab±0.07	9.2^a±3.33	6.9^a±3.19	63.8^a±37.80	152.65^b±44.73	5.58^b±0.54
	Gypsum	18	3.49^a±1.97	0.27^b±0.15	8.5^a±4.74	10.7^b±12.86	67.5^a±49.43	143.18^ab±28.46	5.42^a±0.36
	Elemental sulfur foliar	18	3.53^a±1.69	0.23^ab±0.06	8.5^a±2.93	7.9^ab±4.02	66.4^a±36.44	139.33^a±25.48	5.42^ab±0.25
Cut	First	45	4.32^b±1.63	0.21^a±0.10	10.9^b±2.91	9.7^b±8.71	93.4^b±46.21	156.51^a±27.76	5.64^a±0.26
	Second	45	2.51^a±2.35	0.25^b±0.07	7.0^a±2.93	6.2^a±3.46	39.7^a±19.82	130.52^b±25.31	5.24^b±0.38
Average values of the individual variants do not differ significantly (p>0.95) if they have the same superscript. n=number of observations.

When expressing the yield of sulfur (see Table VI) we can see similar trends, with the exception of the variant with elemental sulfur where the yields of grass forage were depressed. The highest yield of sulfur by the harvested biomass of the grassland was again seen in the gypsum variant.

The observed N/S ratio as an indicator of the S supply in plant material is in agreement with the identification of S deficiency (Dijkshoorn & van Wijk, 1967; Stevens & Watson, 1986). The N/S ratio in the herbage is considered to be a better indicator of the sulfur status than is the S percentage itself (Siman, 1996). The highest yields are usually obtained with ratios of about 10, while values of more than 20 indicate sulfur deficiency (Bolton et al., 1976). Because there is an intrinsic correlation between the N and S supply in plants for amino acid and protein synthesis, it is suggested that an N/S ratio above 16 could be used as a diagnosis of S deficiency in grasslands. The critical N/S ratio has been reported to be 12–14 (Metson, 1973; Stevens & Watson, 1986) and 14 (Bolton et al., 1976; Scott et al., 1983; Stevens, 1985) for silage grasses. According to Murphy and O'Donnell (1989) or Richards (1990) sulfur-deficient grass is usually associated with an N/S ratio higher than 17:1. The nutritional quality of S-deficient silage is poor. Our N/S ratio was significantly the highest in 2005 (9.7). Sulfur fertilisation had no significant effect (Table V), but the highest N/S ratio was achieved in the variant not fertilised with sulfur (10.0) and the lowest in the gypsum variant. The differences among the cuts were significant and the N/S ratio in the 2nd cut was lower (only 7.0, see Table VI) also because, after the 1st cut, nitrogen fertilisation was not applied. Zhao et al. (2003) reported that for animal feeding, an N/S ratio below 15:1 was considered satisfactory.

Apart from the absolute content of N substances in grass forage we also explored the nitrogen yield (Table VI), which was significantly the highest in the last experimental year (2006) and in the 1st cut preceded by nitrogen fertilisation. The effect of variants fertilised with sulfurous fertilisers was not significant, but the average nitrogen yields were higher after the application of sulfate forms of sulfur.

The content of nitrogenous substances was lowest in the control variant (138.09 g kg⁻¹ of dry matter). The level of nitrogenous substances increased after fertilisation (139.33–152.65 g kg⁻¹ of dry matter). The content of nitrogenous substances was significantly the highest (P>0.95) when elemental sulfur was applied (152.65 g kg⁻¹ of dry matter). Elemental sulfur increased the content of nitrogenous substances in the aboveground plant biomass, not only when compared with the control variant, but also with the variant where elemental sulfur was applied foliarly. The content of nitrogenous substances was significantly higher (P>0.95) in the 1st cut (156.51 g kg⁻¹ of dry matter) than in the 2nd (130.52 g kg⁻¹ of dry matter).

In the course of the three experimental years the NEL increased when compared with the control (5.36 MJ kg⁻¹ of dry matter) in all variants of fertilisation (5.42–5.58 MJ kg⁻¹ of dry matter). The effect of elemental sulfur was significant (P>0.95) (5.58 MJ kg⁻¹ of dry matter). A significant (P>0.95) difference was also seen between the NEL content after the application of elemental sulfur (5.58 MJ kg⁻¹ of dry matter) and gypsum (5.41 MJ kg⁻¹ of dry matter). The NEL content is strongly affected by the year and weather (Table VI). The significant (P>0.95) effect of the year and the cut on the NEL content is in accord with the conclusions of Holubek et al. (2001) who proved that the quality of the stand is considerably affected by the age of the stand and stage of development. The NEL content was significantly (P>0.95) the lowest in 2004 (5.22 MJ kg⁻¹ of dry matter) when also the differences among the individual variants of fertilisation were minimal, while it was significantly the highest in 2005 (5.67 MJ kg⁻¹ of dry matter). In 2006 the effect of fertilisation on increasing the NEL content was most pronounced and reached 6.01 MJ kg⁻¹ of dry matter when elemental sulfur was applied. As for the content of nitrogenous substances, this increase is also due to lower yields of this variant and thus lesser dilution of nitrogenous substances, or NEL content. In 2006 the sum of precipitation was above average, but a period of physiological drought prevailed at harvest when the plants lacked water. The clima diagram in Figure 1 documents these fluctuations. Water deficiency could be reflected in the reaction of the plants to accumulation or release of reserve energy substances rather than application of fertilisers.

Table VII presents the values of soil reactions and nutrient reserves discovered after the 2nd cut in the respective years. Since only sulfur and nitrogen were applied as fertilisers, over the three years the nutrient (P, K, Ca, Mg) reserves in the soil decreased; the decrease of phosphorus and magnesium was significant (P>0.95). On the other hand, the application of various forms of sulfur had no significant effect on the macroelement reserves. In the individual years the soil reaction did not change significantly. The various sulfurous fertilisers did not significantly affect the value of the exchangeable soil reaction either, although some authors mentioned the acidification effect of elemental sulfur, and/or ammonium sulfate (Besharati & Rastin, 1999; Kayser et al., 2001).

Table VII. Average value of exchangeable soil reaction and average content of available nutrients in soil after the 2nd cut (mg kg⁻¹) and significance of their differences evaluated by Fisher's Test.

			pH	P	K	Ca	Mg	S
Factor	Level of factor	n	±s x	±s x	±s x	±s x	±s x	±s x
Year	2004	15	5.37^a±0.20	8.21^ab±2.00	95.04^a±14.51	2715.0^a±357.54	303.82^b±36.49	31.10^b±8.46
	2005	15	5.50^a±0.21	9.76^b±5.84	88.14^a±10.60	2539.7^a±251.67	269.62^a±26.93	18.42^a±2.69
	2006	15	5.40^a±0.24	5.72^a±1.18	73.81^a±8.30	2541.2^a±227.37	260.31^a±29.31	18.37^a±3.12
Variants of fertilisation	Unfertilised control	9	5.43^a±0.17	9.36^a±6.32	82.46^a±13.09	2596.4^a±194.66	277.93^a±36.98	18.91^a±4.01
	Ammonium sulfate	9	5.46^a±0.23	7.93^a±4.49	83.14^a±11.72	2672.7^a±230.23	274.63^a±36.55	24.32^b±9.85
	Elemental sulfur	9	5.38^a±0.17	7.48^a±3.54	83.77^a±14.06	2548.3^a±241.09	292.70^a±38.09	23.50^b±2.56
	Gypsum	9	5.37^a±0.23	7.06^a±1.61	84.24^a±13.44	2506.3^a±253.04	274.82^a±36.48	27.14^c±12.43
	Elemental sulfur foliar	9	5.48^a±0.29	7.66^a±2.45	94.70^b±17.93	2669.3^a±475.20	269.51^a±35.19	19.26^a±5.22
Average values of individual variants do not differ significantly (P>0.95) if they have the same superscript. n=number of observations.

In the individual years also the sulfur content gradually decreased significantly (P>0.95). The application of sulfurous substances to the soil (variants 2, 3, and 4) increased the S content in the soil significantly (P>0.95); the significantly highest content of water-soluble sulfur was detected in the gypsum variant. The results of Scott et al. (1983) showed that 10 mg of available sulfur in 1 kg of soil represented adequate reserves for grasses. Our values ranged between 13.8 and 43.1 mg kg⁻¹ and they meet this demand also in variants where sulfur was not applied to the soil.

Compared with northern countries there is less precipitation in the Czech Republic, i.e. lesser washout of sulfates, the climate is warmer and leads to more intensive mineralisation of organic matter in the soil with still a sufficient supply of available sulfur in the soil for grasses in spite of the radical reduction of atmospheric deposition. On the other hand experience from Germany, where the conditions are closer to ours and which desulfurised the emitters of SO₂ about 10 years earlier, proves that grasslands are beginning to respond positively to sulfur fertilisation (Zhao et al., 2003).

Conclusions

In three-year small-plot trial we explored the effect of applications of various forms of sulfur on yields and quality of grass forage, achieving the following conclusions.

Sulfur application in the two cuts did not significantly affect the yields of the grass biomass. Yields increased after sulfur fertilisation only in the 1st cut, especially after the application of sulfate sulfur in ammonium sulfate and gypsum, and/or foliar application of elemental sulfur. In the individual years the content of sulfur in the harvested biomass gradually increased significantly (0.17–0.23–0.29%). In the first year the sulfur content did not reach the limit of critical deficiency (0.2%). Sulfur fertilisation increased the S level in grass forage in all the fertilised variants; the significantly highest content was discovered in the gypsum variant (0.27%). No significant correlation was established between the value of water-soluble sulfur in the soil and the sulfur content in grass forage. Sulfur fertilisation had no significant effect on the N/S ratio, but was highest in the variant not fertilised with sulphur, and lowest in the gypsum variant. The content of nitrogenous substances and nett energy lactation were significantly the highest after fertilisation with elemental sulfur and are connected, among others, with the lowest yields of this variant.

Sulfurous fertilisers did not significantly influence the exchangeable soil reaction and the highest content of water-soluble sulfur after the 2nd cut was seen in the gypsum-applied variant.

Acknowledgements

This study was supported by the Research plan No. MSM6215648905 ‘Biological and technological aspects of sustainability of controlled ecosystems and their adaptability to climate change’, which is financed by the Ministry of Education, Youth and Sports of the Czech Republic.