Plant Soil DOI 10.1007/s11104-015-2789-6

REGULAR ARTICLE

Sulfur dynamics in sub-tropical soils of Australia as influenced by long-term cultivation Peter M. Kopittke & Ram C. Dalal & Neal W. Menzies

Received: 5 November 2015 / Accepted: 28 December 2015 # Springer International Publishing Switzerland 2016

Abstract Background and aims Deficiencies of S in agricultural crops are becoming more common but comparatively little is known regarding the kinetics of cultivationinduced long-term changes in soil S pools. Methods In the present study, six soils were examined with samples collected from 82 sites – this allowing examination of the effects of cultivation for up to 70 years. Results For four of the six soils, long-term cultivation resulted in significant decreases in total S concentrations, with calculated decreases ranging from 35 to 51 %. This decrease was due to mineralization of organic S, with half-life values ranging from 1.0 to 11 years. Generally, organic S decreased linearly with organic C concentrations. For newly-cultivated soils, the mineralization of organic S was sufficient to replace the S harvested in wheat (Triticum aestivum) grain, but after prolonged cultivation, the rate of S mineralization decreased by several orders of magnitude. Conclusions The data presented here provide important information on the effects of cultivation on S dynamics

Responsible Editor: Zucong Cai. Electronic supplementary material The online version of this article (doi:10.1007/s11104-015-2789-6) contains supplementary material, which is available to authorized users. P. M. Kopittke (*) : R. C. Dalal : N. W. Menzies School of Agriculture and Food Sciences, The University of Queensland, St Lucia, QLD 4072, Australia e-mail: [email protected]

within sub-tropical soils – this being required to effectively and sustainably manage these systems. Keywords Biochemical mineralization . Biological mineralization . Cultivation . Sulfur

Introduction Sulfur (S) nutrition is of increasing importance in agricultural systems worldwide. Historically, inputs of S into soils from the atmosphere have generally been sufficient to meet plant demands, particularly within heavily industrialized areas of the northern hemisphere (Eriksen 2008). However, continuing decreases in these atmospheric inputs, coupled with decreased additions of S within higher-analysis P fertilizers, have increased the occurrence of S deficiency (Srinivasarao et al. 2008). Although it is well-known that cultivation of agricultural soils results in the substantial loss of soil organic C with concomitant decreases in organic N (Dalal and Mayer 1986b; David et al. 2009; Tiessen et al. 1982; Wang et al. 2011), surprisingly few field studies have examined the effects of long-term cultivation on soil S concentrations and on mineralization of organic S (Zhao et al. 1996). Furthermore, most studies have examined temperate systems, with comparatively little known regarding S dynamics in tropical and sub-tropical systems (Möller et al. 2002). Wang et al. (2006) found that longterm cultivation decreased soil S concentrations by 30 %, Bettany et al. (1980) found that cultivation for 65 years decreased soil S concentrations by 38 %, and

Plant Soil

Bhupinderpal et al. (2004) found that cultivation for 30 years reduced S concentrations by 50 %. However, studies such as these generally compare single pairedsites and the authors of the present study are unaware of any experiments examining the kinetics of changes in organic S caused by long-term cultivation. Within soils, S may be present as a range of inorganic and organic compounds. In well-drained soils, organic S comprises > 90 % of the total soil S whilst inorganic SO4-S dominates the inorganic fraction (Schoenau and Germida 1992). Commonly, ca. 1 to 5 % of the organic S within the soil is mineralized within a growing season (Schoenau and Malhi 2008; Zhao et al. 1996), with the inorganic SO4-S released subsequently available for plant uptake. It has been proposed that this mineralisation of organic S occurs through both biological mineralization (which occurs indirectly due to release of S from C-bonded S following the mineralisation of organic C by microorganisms) and by biochemical mineralization (which occurs due to extracellular hydrolysis of ester-SO4 S) (McGill and Cole 1981). For the biochemical model, the mineralisation of S is controlled by the supply (and demand) of S rather than the need for energy (C) (McGill and Cole 1981). However, there remains uncertainty as to the pool of organic S that is most easily mineralized. For example, Solomon et al. (2001) and Möller et al. (2002) found that the major source of mineralizable S in soils was C-bonded S rather than from ester-SO4 S. In contrast, Wang et al. (2006) found that long-term cultivation of grasslands resulted in a significant decrease in both C-bonded S and ester-SO4 S, with other studies also demonstrating the importance of ester-SO4 S for mineralization (Bettany et al. 1979; Schoenau and Malhi 2008). In the present study, we examined the effects of longterm cultivation (up to 70 years) on soil S concentrations for six soils from southeast Queensland, Australia. For each soil, a series of paired-sites enabled analysis of the kinetics of these changes, with examination of total S, organic S, and inorganic (SO4-) S. The changes in soil S concentration were also related to changes in organic C levels in order to assist in understanding the process of S mineralisation. Finally, the removal of S in wheat grain was compared to the mineralisation of organic S. This study provides important information on the effects of long-term cultivation on S dynamics in sub-tropical agricultural systems and for management of S nutrition for crops.

Materials and methods Soil and plant sampling An experiment was conducted to investigate the effects of long-term cultivation on the S status of six soils from southern Queensland, Australia. Soil samples were collected from fields that had been cultivated for 0.5 to 70 years and from adjacent areas that were undisturbed and contained the original native vegetation. Across the six soils, a total of 83 sites were sampled (ranging from 12 to 16 sites per soil series), with each of the sites representing differences in the period of cultivation (Table 1). At each of the 83 sites, samples were collected across a 0.1 ha area using a 5 × 8 m grid at a depth of 0 to 0.1 m. A total of 25 individual samples were collected from each of the 83 sites, with five of these samples then mixed to form a single composite sample (thus yielding five composite samples for each of the 83 sites). Samples of wheat (Triticum aestivum) straw and grain were collected, dried at 80 °C for 24–48 h, weighed, and ground to pass a 1 mm sieve. The six soils have been described previously (Dalal and Mayer 1986a) with some of their physicochemical properties given in Table 1. All soils (other than Riverview) were Vertisols, with the Riverview soil having the highest sand content (73 %) and lowest clay content (18 %). The soils were generally slightly alkaline (pHH2O values ranging from 6.4 to 8.1) and received either no or low rates of P fertilizers.

Analyses The total soil S concentration was determined using X-ray fluorescence spectrometry with the organic C concentration determined using the Walkley Black method (Sims and Haby 1971). Concentrations of inorganic S (SO4-S) were determined as calcium phosphate extractable S (Rayment and Lyons 2011), measured using inductively coupled plasma optical emission spectrometry (ICP-OES). Organic S concentrations were calculated as the difference between total S and SO4-S. Although calcium phosphate can also extract some labile forms of organic S, it is dominated by inorganic forms of S (Rayment and Lyons 2011). Indeed, in the present study, calcium phosphate extractable S accounted for an average of 5 % of the total S, with organic S accounting for 95 % (see Results). After

Plant Soil Table 1 The six soils investigated in the present study, including the period of time for which the soils have been cultivated, the physicochemical properties of the undisturbed soils (0–10 cm depth), and the rate of P-addition in superphosphate fertilizer Soil series

Classificationa Period of Number pH EC (dS/m) cultivation of sites (1:5 water) (1:5 water) (years)

Organic C (%)

Sand Silt Clay Rate of P Bulk density (%) (%) (%) fertilizer (kg/ha/y) (g/cm3) Range Mean

Billa Billa

Vertisol

0.5–25

14

7.4 (7.9)

0.20 (0.17)

1.5 (0.99) 0.95

48

18

34

0–20

2.0

Cecilvale

Vertisol

3–35

12

7.4 (7.9)

0.11 (0.14)

1.7 (1.1)

1.0

44

15

41

0–33

7.7

Langlands- Vertisol Logie Riverview Ultisol

0.5–45

12

7.4 (8.0)

0.25 (0.17)

2.2 (1.2)

0.99

34

16

50

0

0

73

9.3 18

0

0

18

23

60

0

0

13

14

72

0–8

1.0

0.5–20

13

6.4 (6.3)

0.035 (0.048)

Thallon

Vertisol

2–23

16

7.2 (7.7)

0.083 (0.091) 0.78 (0.62) 0.95

Waco

Vertisol

1–70

16

8.1 (8.3)

a

0.13 (0.18)

1.2 (0.87) 1.3 1.7 (1.2)

0.85

(Soil Survey Staff 2003)

For pH, electrical conductivity (EC), and organic C, the numbers in brackets correspond to the average values for cultivated soils

drying at 65 °C for 48 h, plant tissues were digested in concentrated HNO3 before analysis using ICP-OES.

Results Changes in soil S concentrations with cultivation

Statistical analyses All statistical analyses were undertaken using SYSTAT 13.1 (Cranes Software, India). To examine changes in S concentrations over time, regressions were fitted using the first-order kinetic model, similar to that for N mineralization (Stanford and Smith 1972), of the form St ¼ Se þ ðS0 – Se Þexpð− ktÞ

ð1Þ

where St is the concentration of S after t years, Se is the concentration of S at equilibrium following a long period of cultivation, S0 is the S concentration prior to disturbance, and k is the rate of S loss per year. For Eq. 1, the half-life (t½) can be calculated as: ln2/k. Where the 95 % confidence intervals encompassed zero for any given coefficient, changes in S were examined using a linear regression St ¼ S0 – at

ð2Þ

where a is the average annual decrease in S. No coefficients are reported (and regressions are not plotted) where the 95 % confidence interval encompassed zero.

For undisturbed soils at a depth of 0–10 cm, total concentrations of S ranged from 114 mg/kg for Riverview to 373 mg/kg for Langlands-Logie soil (Fig. 1). That the lowest concentration of S was found in the sandiest soil (Riverview) is not unexpected, with S concentrations generally being lowest on light-textured soils due to their lower organic matter content (Dick et al. 2008). For all six soils, cultivation resulted in a decrease in the total S concentration. Indeed, when the total S concentrations were averaged across all cultivation periods and compared to undisturbed sites, total S concentrations decreased an average of 34 % for Thallon, 29 % for Cecilvale, 28 % for Waco, 21 % for both Billa Billa and Langlands-Logie, and 1.5 % for Riverview. However, fitting a first-order decay regression, significant relationships were not found for either Billa Billa or Riverview (Fig. 1a, b and Table 2). Based upon these regressions, it was apparent that the largest decrease in total S occurred in the Langlands-Logie soil, with calculated total S concentration decreasing from 436 mg/kg in the undisturbed soil to 207 mg/kg after cultivation for 45 years (Fig. 1a and Table 2) – this decrease of 229 mg/kg after 45 years representing a 53 % decrease. In contrast to total S, concentrations of inorganic S (SO4-S) remained relatively constant, with cultivation causing a significant (linear) decrease for only the Langlands-Logie soil (Fig. 1e and f). For this soil,

Plant Soil Fig. 1 Effect of the period of cultivation on concentrations of total S (a, b), organic S (b, c), and SO4-S (e, f) at a depth of 0–10 cm in six soils. Where significant relationships were found, data were plotted on the left (a, c, e) and the coefficients presented in Table 2

Billa Billa Cecilvale Langlands-Logie Riverview Thallon Waco

600

Total S (mg/kg)

500

(a)

(b)

(c)

(d)

(e)

(f)

400 300 200 100 0 600

Organic S (mg/kg)

500 400 300 200 100 0 40

SO4-S (mg/kg)

30

20

10

0 0

20

40

60

Period of cultivation (y)

Table 2 Initial values (So), equilibrium values (Se), overall rate of loss (kc) and half-life of loss (t½) of total S and organic S from the top layer (0–0.1 m) of six soils Soil

Total S Cecilvale Langlands-Logie Thallon Waco Organic S Cecilvale Langlands-Logie Thallon Waco

So (% S)

Se (% S)

k (year−1)

t½ (year)

R2

216 436 281 212

141 207 160 128

0.154 0.075 0.673 0.064

4.5 9.2 1.0 11

0.547 0.635 0.572 0.860

204 413 274 200

130 195 152 116

0.176 0.076 0.677 0.064

3.9 9.1 1.0 11

0.573 0.622 0.604 0.850

80

0

20

40

60

80

Period of cultivation (y)

calculated concentrations decreased by 48 % (from 21 to 11 mg/kg), but SO4-S accounted for < 10 % of the total S and hence this decrease in SO4-S in Langlands-Logie soil (Fig. 1e) could not account for much of the overall decrease in total S (Fig. 1a). Rather, the observed decrease in total S concentration was almost entirely due to a decrease in organic S (Fig. 1c). Indeed, at a depth of 0– 10 cm, an average of 95 % of S was present in organic forms (with only 5 % as SO4-S), ranging from 89 % for Riverview to 98 % for Thallon. For the LanglandsLogie soil, the calculated concentration of organic S (Fig. 1c and Table 2) decreased by 211 mg/kg after cultivation for 45 years (from 413 to 202 mg/kg), this being a 51 % decrease (c.f. the corresponding decrease in total S of 221 mg/kg, Fig. 1a).

Plant Soil

Changes in yield and grain S concentrations

The importance of the loss of organic S in regulating the total soil S concentration could also be seen through examining the rate of change in these properties. For the four soils for which a significant relationship was found for total S (Fig. 1a), values for the half-life ranged between 1.0 years for Thallon and 11 years for Waco (Table 2). These half-life values for total S were markedly similar to those for organic S (Fig. 1c) which also ranged from 11 years for Waco to 1.0 years for Thallon (Table 2).

Averaged across all sites, yield was highest in Waco (2.4 t/ha) and lowest in Riverview (0.55 t/ha), but grain yield tended to decrease with increasing period of cultivation (see Dalal and Mayer (1986a) for more information). Averaged across all soils, the period of cultivation was found to cause a slight reduction in the grain S concentrations, with cultivation for 70 years calculated to decrease grain S from 0.16 to 0.13 % (although the relationship was poor, R2 of 0.082, the 95 % confidence intervals for the coefficients did not encompass zero) (Supplementary Fig. S2a). Although cultivation resulted in a small (but significant) decrease in grain S concentrations, no relationship was found between the S concentration in the grain and either the total or organic S concentration in the soil (Supplementary Fig. S2b and d). However, decreasing concentrations of inorganic S in the soil were found to result in a slight decrease in the concentration of S in the grain (Supplementary Fig. S2c). Straw tissue S concentrations were also examined, although it was not related to period of cultivation or soil S concentrations for any of the six soil types (Supplementary Fig. S3). Due to the comparatively low grain yields (0.55 to 2.4 t/ha/y), the rate of S removal in the grain was also comparatively low, ranging from 1.1 to 3.7 kg/ha/y (Table 3). As discussed earlier, for four soils, this removal of S in the grain corresponded to a significant decrease in organic S concentrations (no significant relationships were found for either Billa Billa or

Changes in the C:S ratio and relationship between organic C and organic S For undisturbed soils, the C:S ratio ranged from 29:1 in the Thallon soil to 118 in the Riverview soil (Fig. 2). Although cultivation was found to cause a substantial decrease in the organic S concentration for four of the six soils (Fig. 1c and d), it generally had no significant effect on the C:S ratio (Fig. 2). The only exception was for Thallon in which the C:S ratio decreased significantly with cultivation. For Cecilvale, Langlands-Logie, and Waco, a significant linear relationship was found between organic S and organic C, although no relationship was found for Billa Billa, Riverview, and Thallon (Fig. 3). For these linear regressions relating organic S and organic C concentrations, the slopes ranged from 96 (Cecilvale) to 150 (Langlands-Logie), indicating the mineralisation of S at a constant rate of 96 to 150 mg/kg soil for every 1 % decrease in organic C.

Fig. 2 Effect of the period of cultivation on C:S ratios. Where significant relationships were found, data were plotted on the left (a)

Billa Billa Cecilvale Langlands-Logie Riverview Thallon Waco

160

(a)

120 C:S ratio

(b)

y = 0.879x + 48.3 R2 = 0.296

80

40

0 0

20

40

60

Period of cultivation (y)

80

0

20

40

60

Period of cultivation (y)

80

Plant Soil 600

(a) 500 Organic S (mg/kg)

Fig. 3 Change in organic S with organic C for six soils cultivated for up to 70 years. No significant relationships were found for the three soils in (d)

Cecilvale

(b)

Langlands-Logie

y = 95.8x + 37.0 R2 = 0.761

y = 151x + 86.6 R2 = 0.775

(c)

(d)

400 300 200 100 0 600

Organic S (mg/kg)

500

Waco

Billa Billa Riverview Thallon

y = 125x - 2.7 R2 = 0.803

400 300 200 100 0 0.0

0.5

1.0

1.5

2.0

2.5 0.0

Organic C (%)

Riverview) (Fig. 1). For these four soils, averaged across the entire periods of cultivation (23 years for Thallon to 70 years for Waco), this removal of S in the grain corresponded reasonably well to the decrease in organic S, which ranged from 1.2 to 5.4 kg/ha/y. However, this decrease in organic S occurred at a substantially higher rate upon initial cultivation of the undisturbed soil, with rates ranging from 60 to 5.2 kg/ha/y. Conversely, after prolonged cultivation, the rate of decrease in organic S was markedly lower, with the decrease being < 1 kg/ha/y for all four soils (Table 3).

0.5

1.0

1.5

2.0

2.5

Organic C (%)

Discussion Long-term cultivation (up to 70 years) was found to result in substantial decreases in soil S concentrations, with significant relationships found between total S concentration and the period of cultivation for four of the six soils examined. Given that an average of 95 % of the S was present in organic forms, this decrease in total soil S concentration resulted almost entirely from a decrease in soil organic S concentrations. Upon the initial cultivation of undisturbed land, the rate of decrease in organic S exceeds the rate at which S is

Table 3 Removal of S in wheat grain and the corresponding decrease in soil S (assuming uptake from 0 to 10 cm only) Removal of S

Decrease in soil organic S (kg/ha/y)

Soil

Average grain yield (t/ha/y)

S removed in grain (kg/ha/y)

Average

Initial

Final

Billa Billa

1.4

2.4

NS

NS

Cecilvale

1.9

3.0

2.1 (1–35)

12 (1)

0.030 (35)

Langlands-Logie

1.2

1.8

4.6 (1–45)

16 (1)

0.56 (45)

Riverview

0.55

1.1

NS

NS

Thallon

1.3

1.9

5.0 (1–23)

57 (1)

0.000019 (23)

Waco

2.4

3.7

1.0 (1–70)

4.4 (1)

0.053 (70)

NS

NS

Values are the average annual decrease calculated from the regressions (Fig. 1) where significant differences were found, using the bulk density values (Table 1) and assuming a depth of 10 cm. For the decrease in soil organic S concentrations, the values in parentheses indicate the number of years cultivation from which the decrease was calculated using the equation

Plant Soil

removed in the harvested grain. However, after prolonged cultivation, the rate at which organic S mineralized decreased substantially, with removal of S in grain exceeding the decrease in organic S by at least an order of magnitude although some S may have also been taken up by plants from deeper soil layers. Effects of long-term cultivation on soil total S concentration For the six undisturbed soils, total S concentration ranged from 114 to 373 mg/kg, which is similar to values reported for a wide range of soils (Zhao et al. 1996). However, cultivation significantly decreased total S concentration in four of the six soils (for the two other soils, S concentrations decreased somewhat, but no significant relationship was found). For the Langlands-Logie soil, cultivation for 45 years was calculated to result in a 51 % decrease in total S, with corresponding values being 43 % for Thallon after 23 years, 39 % for Waco after 70 years, and 35 % for Cecilvale after 35 years (Fig. 1). These decreases in organic S are similar to those reported previously for long-term cultivation. For example, studying Mollisols (Udic Haploborolls) on the Canadian prairie following cultivation for 65 years, Bettany et al. (1980) found that S concentration decreased by 38 %. Similarly, studying a range of grassland soils from the Great Plains of North America, Wang et al. (2006) found that an average of 30 % of the organic S was lost with long-term cultivation. Finally, in New Zealand, Bhupinderpal et al. (2004) found that cultivation for 30 years decreased total soil S concentration by ca. 50 % compared to permanent pasture. Mineralisation of organic S The 35–51 % decrease in soil total S concentration observed in the present study (Fig. 1a and b) occurred due to the mineralisation of organic S, with organic S accounting for an average of 95 % of total soil S in the undisturbed soils. However, the rate at which organic S decreased varied markedly between soils (Fig. 1c and d), with half-life values ranging from 1.0 years (Thallon) to 11 years (Waco) (Table 2) [for two soils there was no significant relationship between organic S concentration and period of cultivation (Fig. 1d)]. This variability in the rate at which organic S was lost was not related to the soil texture (clay content, silt content, or silt + clay content, data not presented), with Thallon having the

shortest half-life but also having a high clay content (Table 1). Nevertheless, the rate at which organic S decreased in the present study is similar to that reported by Jackman (1964) for a range of soils from New Zealand, with half-life values ranging from 2 to 42 years. Similarly, for the present study, the rate at which organic S is mineralised in newly-cultivated soils (5.2 to 60 kg/ ha/y) is similar to that reported previously (Schoenau and Germida 1992), although the rate of S mineralisation after prolonged cultivation was observed to decrease by several orders of magnitude (Fig. 1b and c). The C:S ratio was found to remain relatively constant, with cultivation resulting in a decrease in C:S ratio for only the Thallon soil (Fig. 2). Indeed, for three of the four soils in which organic S decreased significantly with period of cultivation (Fig. 1), a linear relationship was found between organic S and organic C (Fig. 3). The observation that the C:S ratio generally remained constant (and that organic S decreased linearly with organic C) is in contrast to that typically reported previously. For example, in prairie soils of North America, Wang et al. (2006) found that although organic S concentrations decreased by 30 %, this was less than the decrease in organic C (49 %). Bettany et al. (1980) reported that losses of organic S (38 %) were slightly less than for organic C (44 %) following cultivation for 65 years. Similarly, Walker et al. (1959) found that C:S ratio was altered from 250:1 in an undisturbed soil to 92:1 after 25 years under pasture (although the application of superphosphate at a rate of 330 kg/ha/y would have resulted in the substantial addition of inorganic S also). The observation of a close relationship between organic S and organic C for three of the six soils (Fig. 3) suggests that S is stabilized in close association with the C in the soil organic matter and hence that biological mineralization is of importance for the long-term mineralisation of S in these soils. Indeed, whilst the biochemical mineralization of S depends upon the demand (and supply) for S, for biological mineralization, release of S depends upon demand for energy (C) with the C-bonded S released due to mineralization of organic C (McGill and Cole 1981). Thus, for these three soils, this data is in agreement with McLaren and Swift (1977) who reported that C-bonded S is of greater importance for the longer-term mineralization of S. Similar findings regarding the importance of C-bonded S for long-term mineralization have been reported by Möller et al. (2002); Freney et al. (1975), and Solomon et al. (2001). For the other three soils in the present study,

Plant Soil

the absence of a significant relationship between organic S and organic C (Fig. 3) is not surprising for Billa Billa and Riverview given that organic S did not decrease significantly with period of cultivation (Fig. 1c and d). However, for Thallon, the lack of a relationship between organic S and organic C, coupled with observations that organic S decreased significantly (Fig. 1c and d) and that the C:S ratio was altered (Fig. 2) suggests that much of the mineralization of S from the Thallon soil was not directly related to the mineralization of C (biological mineralization) but that it was due to biochemical mineralization. However, S speciation was not measured in the present study and so care needs to be taken when interpreting this data.

and crop response to S fertilization has been demonstrated (Bell et al. 2012). Regardless, continued mineralisation of organic S is insufficient to meet crop removal. Indeed, on similar soils with much longer cultivation history, for example the Vertisol soils of India, responses to S are widespread and S fertilizer more commonly used (Ganeshamurthy and Sammi Reddy 2000; Srinivasarao et al. 2008). Many of the soils considered here have relatively high exchangeable Na levels, and demonstrate the attendant poor surface structural behaviours of dispersion and surface sealing. Application of gypsum to address the soil surface structural problems (Dang et al. 2010) would also ensure adequate S nutrition of crops.

Balance between removal of S in grain and the mineralisation of organic S

Conclusion

On average, the removal of S in grain accounted for 1.1 to 3.7 kg/ha/y – this being substantially lower than that reported previously for wheat (Dick et al. 2008) due to the low yields in the present study (Table 3). During initial cultivation when mineralisation exceeded crop removal, concentrations of SO4-S did not increase (Fig. 1e and f) despite the rapid mineralization of organic S (Table 3). This is presumably the result of leaching of inorganic S beyond the 0–10 cm layer of interest in the present study, although it is likely that the leaching of SO4-S to deep within the soil profile was comparatively small (David and Mitchell 1987; Donn and Menzies 2005). Although initial mineralization of S was high, with extended periods of cultivation, the rate of S mineralization became much lower than crop removal (Table 3). However, consideration should also be given to other inputs and losses not measured in the present study. For example, for all six soils, additional inputs of S would have occurred through atmospheric inputs [which were likely to be ca. 1–5 kg/ha/y (Dentener et al. 2006; Langner and Rodhe 1991)] and from the modest addition of P fertilizers for Waco, Cecilvale, and Billa Billa (Table 1). Assuming the use of single superphosphate (single superphosphate is 11 % S, triple superphosphate is 1.5 % S), the rates of P application would correspond to S inputs of 1.3 kg/ha/y for Waco, 9.6 kg/ha/y for Cecilvale, and 2.5 kg/ha/y for Billa Billa. However, the widespread use of high-analysis P fertilizers (mono- and di-ammonium phosphate), rather than S-containing superphosphate, has further limited S availability to crops,

This study has shown that long-term cultivation of soils (up to 70 years) can result in substantial decreases in soil total S concentrations. For four of the six soils studied, cultivation decreases S concentrations by 35 to 51 % due to the mineralization of organic S. Generally, this decrease in organic S was linear with organic C concentrations, suggesting that biological mineralization of C-bonded S is important in long-term mineralization of organic S (although further work is required to verify this hypothesis). Although this mineralization of organic S was sufficient to replace the S removed in the harvested grain, after prolonged cultivation, the rate of S mineralization decreased by several orders of magnitude and was insufficient to replace the S removed. This data provides important information on S dynamics in sub-tropical soils in agricultural systems. Acknowledgments Support was provided to Dr. Kopittke as a recipient of an Australian Research Council (ARC) Future Fellowship (FT120100277). The assistance of Maria HernandezSoriano in preparing the sample location map is acknowledged.

References Bell M, Lester D, Smith L, Want P (2012) Increasing complexity in nutrient management on clay soils in the northern grain belt – nutrient stratification and multiple nutrient limitations. In: Proceedings of 16th Australian Agronomy Conference (14–18 October 2012), Armidale, Australia Bettany JR, Stewart JWB, Saggar S (1979) The nature and forms of sulfur in organic matter fractions of soils selected along an environmental gradient. Soil Sci Soc Am J 43:981–985

Plant Soil Bettany JR, Saggar S, Stewart JWB (1980) Comparison of the amounts and forms of sulfur in soil organic matter fractions after 65 years of cultivation. Soil Sci Soc Am J 44:70–75 Bhupinderpal S, Hedley MJ, Saggar S, Francis GS (2004) Chemical fractionation to characterize changes in sulphur and carbon in soil caused by management. Eur J Soil Sci 55:79–90 Dalal RC, Mayer RJ (1986a) Long-term trends in fertility of soils under continuous cultivation and cereal cropping in southern Queensland. 1. Overall changes in soil properties and trends in winter cereal yields. Aust J Soil Res 24:265–279 Dalal RC, Mayer RJ (1986b) Long-term trends in fertility of soils under continuous cultivation and cereal cropping in southern Queensland. 2. Total organic-carbon and its rate of loss from the soil-profile. Aust J Soil Res 24:281–292 Dang YP, Dalal RC, Buck SR, Harms B, Kelly R, Hochman Z, Schwenke GD, Biggs AJW, Ferguson NJ, Norrish S, Routley R, McDonald M, Hall C, Singh DK, Daniells IG, Farquharson R, Manning W, Speirs S, Grewal HS, Cornish P, Bodapati N, Orange D (2010) Diagnosis, extent, impacts, and management of subsoil constraints in the northern grains cropping region of Australia. Aust J Soil Res 48:105–119 David MB, Mitchell MJ (1987) Transformations of organic and inorganic sulfur: importance to sulfate flux in an Adirondack forest soil. JAPCA 37:39–44 David MB, McIsaac GF, Darmody RG, Omonode RA (2009) Long-term changes in mollisol organic carbon and nitrogen. J Environ Qual 38:200–211 Dentener F, Drevet J, Lamarque JF, Bey I, Eickhout B, Fiore AM, Hauglustaine D, Horowitz LW, Krol M, Kulshrestha UC, Lawrence M, Galy-Lacaux C, Rast S, Shindell D, Stevenson D, Van Noije T, Atherton C, Bell N, Bergman D, Butler T, Cofala J, Collins B, Doherty R, Ellingsen K, Galloway J, Gauss M, Montanaro V, Müller JF, Pitari G, Rodriguez J, Sanderson M, Solmon F, Strahan S, Schultz M, Sudo K, Szopa S, Wild O (2006) Nitrogen and sulfur deposition on regional and global scales: a multimodel evaluation. Glob Biogeochem Cycles 20:GB4003 Dick WA, Kost D, Chen L (2008) Availability of sulfur to crops from soil and other sources. In: Sulfur: a missing link between soils, crops, and nutrition. American Society of Agronomy, Crop Science Society of America, Soil Science Society of America, Madison, WI, pp 59–82 Donn MJ, Menzies NW (2005) The effect of ionic strength variation and anion competition on the development of nitrate accumulations in variable charge subsoils. Aust J Soil Res 43:43–50 Eriksen J 2008 Soil sulfur cycling in temperate agricultural systems. In: Sulfur: a missing link between soils, crops, and nutrition. American Society of Agronomy, Crop Science Society of America, Soil Science Society of America, Madison, WI, pp 25–44 Freney JR, Melville GE, Williams CH (1975) Soil organic matter fractions as sources of plant-available sulphur. Soil Biol Biochem 7:217–221 Ganeshamurthy AN, Sammi Reddy K (2000) Effect of integrated use of farmyard manure and sulphur in a soybean and wheat cropping system on nodulation, dry matter production and chlorophyll content of soybean on swell-shrink soils in central India. J Agron Crop Sci 185:91–97

Jackman RH (1964) Accumulation of organic matter in some New Zealand soils under permanent pasture. N Z J Agric Res 7: 445–471 Langner J, Rodhe H (1991) A global three-dimensional model of the tropospheric sulfur cycle. J Atmos Chem 13:225–263 McGill WB, Cole CV (1981) Comparative aspects of cycling of organic C, N, S and P through soil organic matter. Geoderma 26:267–286 McLaren RG, Swift RS (1977) Changes in soil organic sulphur fractions due to the long term cultivation of soils. J Soil Sci 28:445–453 Möller A, Kaiser K, Kanchanakool N, Anecksamphant C, Jirasuktaveekul W, Maglinao A, Niamskul C, Zech W (2002) Sulfur forms in bulk soils and alkaline soil extracts of tropical mountain ecosystems in northern Thailand. Aust J Soil Res 40:161–175 Rayment GE, Lyons DJ (2011) Soil chemical methods: Australasia. CSIRO Publishing, Collingwood, Australia, p 495 Schoenau JJ, Germida JJ (1992) Sulfur cycling in upland agricultural systems. In: Howarth RW (ed) Sulfur cycling on the continents. Wiley, New York, pp 261–277 Schoenau JJ, Malhi SS (2008) Sulfur forms and cycling processes in soil and their relationship to sulfur fertility. In: Sulfur: a missing link between soils, crops, and nutrition. American Society of Agronomy, Crop Science Society of America, Soil Science Society of America, Madison, WI, pp 1–10 Sims JR, Haby VA (1971) Simplified colorimetric determination of soil organic matter. Soil Sci 112:137–141 Soil Survey Staff (2003) Keys to soil taxonomy. United States Department of Agriculture, Washington, p 332 Solomon D, Lehmann J, Tekalign M, Fritzsche F, Zech W (2001) Sulfur fractions in particle-size separates of the sub-humid Ethiopian highlands as influenced by land use changes. Geoderma 102:41–59 Srinivasarao C, Wani SP, Sahrawat KL, Rego TJ, Pardhasaradhi G (2008) Zinc, boron and sulphur deficiencies are holding back the potential of rainfed crops in semi-arid India: experiences from participatory watershed management. Int J Plant Prod 2: 89–99 Stanford G, Smith SJ (1972) Nitrogen mineralization potentials of soils. Soil Sci Soc Am J 36:465–472 Tiessen H, Stewart JWB, Bettany JR (1982) Cultivation effects on the amounts and concentration of carbon, nitrogen, and phosphorus in grassland soils. Agron J 74:831–835 Walker T, Thapa B, Adams A (1959) Studies on soil organic matter. 3. Accumulation of carbon, nitrogen, sulfur, organic and total phosphorus in improved grassland soils. Soil Sci 87: 135–140 Wang J, Solomon D, Lehmann J, Zhang X, Amelung W (2006) Soil organic sulfur forms and dynamics in the Great Plains of North America as influenced by long-term cultivation and climate. Geoderma 133:160–172 Wang Y, Tu C, Cheng L, Li C, Gentry LF, Hoyt GD, Zhang X, Hu S (2011) Long-term impact of farming practices on soil organic carbon and nitrogen pools and microbial biomass and activity. Soil Tillage Res 117:8–16 Zhao F, Wu J, McGrath S (1996) Soil organic sulphur and its turnover. In: Piccolo A (ed) Humic substances in terrestrial ecosystems. Elsevier, Amsterdam, pp 467–506