SCOPE 21 -The Major Biogeochemical Cycles and Their Interactions 

ABSTRACT

Interrelationships of C, N, S, and P cycles are dealt with in terms of specific effects on specific cycles; stoichiometry of C, N, S, and P; mineralization and immobilization relations; and nutrient distribution and compartment transfers. A concluding integrating concept involving soil organic matter has been used to tie these together.

The P cycle is altered by those of C, N, and S primarily through alterations of soil microbial and plant environments (by N and S) and alterations of demand for P (by C and N). The S and N cycles are intertwined due to influences of both elements on the soil environment and the tendency of each element to become directly involved with the other in organisms or in enzymes. Examples are provided and include concurrent N reduction and S oxidation by Thiobacillus denitrificans and inhibition by reduced S compounds of NO₃^-, N₂O, or NO reduction processes.

Stoichiometric relations among the cycles are shown to be variable in response to the ability of organisms to differentially absorb or exclude elements, on one hand, and to fundamentally different stabilization and mobilization mechanisms, on the other. The challenge remains to understand and quantify the balance between mineralization rate, mineral nutrient absorption rate, microbial use, and net primary production.

Within the soil organic component, a distinction is made between elements stabilized through association with C, and elements stabilized through reaction of the non-carbonaceous component of the molecule. The first group is typified by N and possibly C-bonded S; organic phosphate esters, and possibly ester sulphates, are represented in the second group. Two mineralization mechanisms are discussed. The first, termed biological mineralization, is release of NH₃ or SO₄^2-, etc. as waste following oxidation of N- or S-containing organics to CO₂ for energy. The second mineralization process has been termed biochemical mineralization to represent those processes external to the cell membrane resulting from activities of hydrolytic enzymes produced in response to a demand for the end product.

Whereas in the atmosphere the four cycles are linked by chemical reactions, in soils they are primarily linked through growth processes of soil organisms, including plants. Further influences and interactions result from effects of the atmospheric N and S cycles (and to a lesser degree the C cycle) on the terrestrial environment.

9.1 INTRODUCTION

The purpose of this paper is to examine in a broad way some of the interrelations of C, N, S, and P cycles in terrestrial systems and the principles controlling those interrelationships. The topics covered will include specific effects on specific cycles; stoichiometry of C-, N-, S-, and P-mineralization and immobilization (stabilization) relations; nutrient distribution and compartment transfers on the short term and over longer time scales; and finally some comments on a conceptulization of how the cycles are integrated through soil organic matter.

9.2 SPECIFIC EFFECTS ON SPECIFIC CYCLES

Soil organisms integrate C, N, S, and P cycles through biomass production. They can, however, alter the relative amounts of C, N, P, and S in terrestrial systems by altering the turn-over rate of individual elements. Soil organisms can also alter these cycles by modifying the soil environment. Changes in soil enzyme activities may indicate the responses of soil organisms to defined stresses or stimuli. Interpretation of such data, however, requires that information be available on how microbes respond to similar perturbations in simple systems. The potential of alterations in one element to affect others may be interpreted in part from changes in activity of specific soil enzymes following addition or removal of one or more elements. Many of the specific effects observed have been recorded at the microbe level.

9.2.1 Phosphorus Cycle

Phosphorus cycling is complex and involves the storage of P in biological, organic and inorganic forms. Increases in C and N supply can increase demand for P as indicated by production of soil acid phosphohydrolase activity following C and N addition to soil (Ladd and Paul, 1973; Spiers and McGill, 1979). Spiers and McGill (1979) showed that production of phosphohydrolase activity is repressed and activity of existing phosphohydrolases inhibited by adding orthophosphate. These responses to C, N, and P additions have also been demonstrated with soil alkaline phosphatase activity (Figure 9.1), although changes in alkaline phosphatase activity were small. The ability to store P as polyphosphate (Harrold, 1966) may buffer the P cycle against either decreases in external P availability or increases in C, N, or S supply.

Figure 9.1 Effect of C, N, and P additions (glucose10 mg C g^-1 soil; NH₄NO₃875 µg N g^-1 soil; KH₂PO₄500 µg P g^-1 soil) on phosphohydrolase activity of a Grey Luvisolic Ap horizon (Bergstrom and McGill, unpublished data)

Inorganic P dynamics can be altered by C, N, and S cycles. Following the report of Gerretson (1948), many researchers examined the relationships between production of organic acids in the rhizosphere and plant uptake of P from mineral sources. Whether the effects observed were caused by increased P solubility or hormonal effects is not clear. Organic acids have been widely considered to act as calcium chelators (Duff et al., 1963; Stevenson, 1967), thus increasing P solubility. One of the most often mentioned organic acids in this regard is 2-ketogluconic acid, which is found abundantly in rhizospheres (Moghimi et al., 1978). Moghimi and Tate (1978), however, showed that, although 2-ketogluconic acid is among the strongest monobasic carboxylic acids (pK^B_a = 2.66), its calcium stability constant is negligible between pH 2.4 and 6.4. They concluded that P dissolution by such acids resulted from localized lowering of pH and not from chelation.

Figure 9.2 Distribution of P removed from synthetic hydroxyapatite (3 Ca(PO₄)₂ Ca(OH)₂) by a phosphate dissolving bacterium, in 50 ml of liquid culture containing added CaCO₃. P as cells and soluble organic matter (P_o); P as soluble orthophosphate P (Hmeidan and McGill, unpublished data)

The literature has concentrated on increases in inorganic P (P_i) in solution in the presence of 'phosphate-dissolving bacteria'. In highly buffered soils, changes in pH are difficult to effect, and P_i cannot be maintained at a high level due to reprecipitation and adsorbtion. Hmeidan and McGill (unpublished data) have examined P incorporation into bacterial cells and soluble organic matter during growth on synthetic hydroxyapatite. Results of these studies (Figure 9.2) show that large quantities of P_i can be maintained in unbuffered liquid cultures (i.e. no CaCO₃), but in systems buffered with CaCO₃, soluble P_i drops rapidly while soluble organic P and biomass P remain relatively constant. The evidence for direct microbial dissolution of P-containing minerals in soil is not strong. Possibly the more significant method of microbial influence on mobilization of P from primary minerals in soil systems is through the incorporation of P_i into microbial biomass and organic forms and hence to plant roots. Although this is a longer route, it may be more feasible, considering the greater mobility of organic P than of inorganic P in soils (Rolston et al., 1975).

Azcon et al. (1976) observed that plants innoculated with Endogone mycorrhiza and a phosphate-dissolving bacterium took up more P from rock phosphate than did plants with either bacteria or Endogone mycorrhiza alone. They suggested that the role of the bacterium may include either production of plant growth hormones or dissolution of rock phosphate, or both. The dissolved P_i may then have been taken up more rapidly by mycorrhizal roots than by uninfected roots. Alternatively, they suggested that the extra P could have come from dead bacterial cells. Turn-over of bacteria was greater around mycorrhizal than non-mycorrhizal roots. Raj et al. (1981), using ³²P-labelled superphosphate and tricalcium phosphate, confirmed the synergistic effect of phosphate-dissolving bacteria and mycorrhizae but did not clearly define the mechanisms involved. Regardless of whether the mechanism involves vitamin production (Baya et al., 1981), hormone activity, or dissolution of insoluble P minerals by the bacterium, ultimately the effects seem related to C supply to bacteria in the rhizosphere.

Weathering of soil minerals by H₂CO₃ and HCO₃^- derived from biologically produced CO₂ is generally accepted as fundamental to soil genesis. A role of such CO₂ in P cycling should therefore follow. Parker (1924), however, found no relation between CO₂ production and P uptake. Johnston and Olson (1972) found that dissolution of apatite and P uptake from nutrient solution by roots of several plant species was unaffected by removal of CO₂. These data suggest that, although CO₂ may be important in soil mineral weathering on a geological time scale, it is likely of little importance in controlling the dynamics of P on a biological time scale.

Effects of S on P cycling have long been recognized and exploited. Lipman et al. (1916) showed that addition of elemental S to soil increased P availability. Brown and Gwinn (1917) used elemental S (50 g m^-2) and manure (2400 g m^-2) to increase availability of P from rock phosphate. More recently Bromfield (1975) demonstrated how a transformation in the S cycle could be used to influence P cycling and hence the N cycle. He mixed elemental S with rock phosphate and added the mixture to S- and P-deficient Nigerian soils before planting groundnuts (Arachis hypogaea). The S was oxidized, and apparently acidified the rock phosphate, thus increasing plant uptake of P (Table 9.1). It is not clear if the increased P uptake was from dissolution of rock phosphate or from the soil due to better root development caused by the additional S. Groundnuts are capable of N₂ fixation. Therefore, all three cycles (N, S, and P) had been integrated and manipulated.

9.2.2 Sulphur Cycle

Effects on the S cycle operate through organic matter transformations and through alterations of S oxidation/reduction reactions. Effects of C and N on organic S cycling can be related to the nature of organic S in soil. McGill and Cole (1981) have proposed a model that suggests C-bonded S cycles through soil organic matter in a manner analogous to organic N, whereas HI-reducible S (sulphate and compounds containing C-O-S, N-S and N-O-S linkages (see Stewart et al., chapter 8 (section 8.4.2), this volume)) is mobilized by a different mechanism Sulphohydrolases are involved in release of SO₄^2- from ester sulphates (Dodgson and Rose, 1975; Fitzgerald, 1978). McGill and Cole (1981) introduced the term biochemical mineralization to indicate this process in soil and to distinguish it from biological mineralization, which, they conclude, relates to release of NH₃ and C-bonded S during microbial oxidation of N- and S-rich organics to provide energy for growth. The fate and role of SO₄^2- so released has not been fully clarified.

Table 9.1 Uptake of sulphur and phosphorus by mature groundnuts following additions of ground rock phosphateelemental S mixtures at time of planting (Bromfield, 1975). Reproduced by permission of Cambridge University Press

	Uptake (g m-2)
	S		P
	GRP*	GRPS+	GRP	GRPS
Haulms	0.17	0.42	0.40	0.53
Shell	0.02	0.04	0.04	0.05
Kernel	0.16	0.32	0.66	0.87
Total	0.35	0.78	1.10	1.45
*GRP = Ground rock phosphate.
+GRPS = Ground rock phosphate elemental S mixture.

Nitrogen may influence organic S cycling either by creating a demand for S and stimulating production of sulphohydrolases to obtain it or, in some organisms, by repressing sulphohydrolase activity. Okamura et al. (1977), working with Klebsiella aerogenes, showed that NH₄Cl repressed aryl sulphohydrolase production in liquid culture (Table 9.2). They also demonstrated catabolite repression with glucose and derepression by tyramine. Henderson and Milazzo (1979), however, found that arylsulphatase synthesis was not repressed by NH₄Cl in Salmonella typhimurium. The observations of Okamura et al. (1977) may, therefore, not be completely general. Effects of NH₄⁺ on arylsulphohydrolase synthesis have not been examined in soil systems. 

Table 9.2 Repression of arylsulphodydrolase production by NH₄Cl (Okamura et al., 1977). Reproduced by permission of American Society for Microbiology

Orthophosphate has been shown to inhibit arylsulphohydrolase activity in soil (Al-Khafaji and Tabatabai, 1979) at 25 µmoles g^-1 of soil. The inhibitory effect caused an increase in apparent Km (Table 9.3) and was consequently concluded to be competitive. They found no effect of SO₄^2- on arylsulphohydrolase activity.

As with their effect on P uptake, mycorrhizae increase S uptake by infected plants (Gray and Gerdeman, 1973). Phosphorus has also been shown to affect S uptake by vesiculararbuscular mycorrhizae of onion plants (Rhodes and Gerdeman, 1978). These workers further showed that the rate of S uptake was increased by P addition to non-mycorrhizal roots (Table 9.4). P supply in the soil appeared to have a greater effect on S uptake than did mycorrhizal infection.

Choline sulphate is formed by both fungi and bacteria (Fitzgerald, 1978). It appears to serve a storage function (Spencer et al., 1968) and to be hydrolysed to yield SO₄^2- for growth during times of S stress. The enzyme involved is repressed by S-containing amino acids (Scott and Spencer, 1968). Consequently, S may be stored in this manner in soils when C, N, and P are inadequate for growth. This hypothesis has not been rigorously tested in soil systems. Yeung (1980) showed that K₂SO₄ added to three soils in the Athabasca Tar Sands area of Alberta was largely converted to organic form during 1 year in the field (Table 9.5). Saggar et al. (1981a) reported that, although net N mineralization occurred during incubation of two soils in the laboratory, net immobilization of S was observed. Whether or not this observed conversion is related to any sort of biological storage mechanism is not known.

	Net S recovery (g m-2 to 90 cm)
Site	As SO42-	As total S	Gain or loss
AOSERP Camp	1.2	4.3	1.3
(Pine Sand)
Ruth Lake	0.5	4.4	1.2
(Pine Sand)
Thickwood Hills	0.5	4.8	0.8
(Luvisol)

9.2.3 Nitrogen

Nitrogen cycles interact with C, S, and P cycles. Cole and Heil (1981) have summarized the effects of P on the N cycle. They stress that biologically active P, not total P, controls the N cycle. Further, they conclude that microbial growth processes are the main points at which N cycling is adjusted to the P supply in terrestrial systems. Nitrogen cycling processes shown to be affected by P may be listed as (Cole and Heil, 1981):

1. Multiplication and activity of Rhizobium. 
2. Root infection and growth of nodules.
3. N₂ fixation and other biochemical processes within the plant.
4. Blue-green algal proliferation and influence on rice in paddy soils. 
5. Initial colonization by N₂-fixing lichens.
6. N mineralization rate in some Andepts. 
7. Nitrification in some soils.
8. Decomposition rate of organic N in systems where microbial death is an important return mechanism for available P.

Nitrogen dynamics follow closely those of carbon (McGill et al., 1975). McGill et al. (1981) used C cycling as a driving mechanism in developing a simulation model of N cycling. Energy and electrons to reduce N for fixation and denitrification, which respectively add N to and remove it from terrestrial systems, are provided largely by reduced C. Internal N cycling interacts at the organism level with that of C most intimately through mineralization and immobilization relations. The balance between these two processes is sensitive to C utilization efficiency and to the C:N ratio of soil organisms. Immobilization is favoured by high efficiencies and low microbial C:N ratios, whereas the reverse conditions favour N mineralization. McGill et al. (1981) suggest that microbial C:N ratios in soil are not constant. They show that the fluctuation in C:N ratio over time in response both to shifting substrate quality and suite of organisms present influences N availability. The main effect is to provide some buffering in the N supply to plants in the system.

There are some parallels between N and S cycles at the organism level with respect to the inorganic components; organic components of N and S do not always share the same parallels (McGill and Cole, 1981). Both elements undergo oxidation and reduction reactions in soil. Both elements may be lost from the soil system by reduction processes if they proceed far enough in an appropriate environment.

Transfers of C, S, and N gases from the atmosphere to soil systems help complete the global cycles of these elements. Soils have been shown to sorb NO₂ (Bremner and Nelson, 1967), C₂H₄ (Abeles et al., 1971), CO (Inman and Ingersoll, 1971), phosphine (Burford and Bremner, 1972), SO₂ (Seim, 1970) and NH₃ (Malo and Purvis, 1964). Smith et al. (1973) showed that several S-containing gases are adsorbed by both sterile and non-sterile soils. Sorbtion of SO₂ was more rapid than that of H₂S or CH₃SH. They also reported that whereas micro-organisms appeared not to be required for sorbtion of the above S-containing gases, they were responsible for sorbtion of CO, C₂H₄, and C₂H₂. Dry soils examined sorbed up to 65 mg of H₂S/g Soil. Moist soils absorbed up to 63 mg H₂S/g soil and up to 67 mg SO₂/g soil indicating that soil is an important sink for atmospheric gases.

On the basis of steam-sterilized soil experiments, Smith et al. (1973) showed that abiotic processes in soils may in some cases also be a source of atmospheric gases. A gas originating from soil systems that may influence atmospheric chemistry is N₂O (Crutzen, Chapter 3, this volume). Its production may result from either nitrification (Hutchinson and Mosier, 1979) or denitrification (Knowles, 1981) and seems to vary with degree of aeration, nitrate supply and soil pH (Focht, 1974; Blackmer and Bremner, 1978; Knowles, 1981). Interactions of N and S gases in soil occur during oxidation and reduction of N and S. These interactions are integrated at the microbial level rather than the level of chemical reactions in soil. Ghiorse and Alexander (1976) confirmed that NO₂, as well as SO₂, was adsorbed by sterile soil, thus showing that their uptake by soil is not strictly biologically mediated. They further showed that whereas soil micro-organisms were not directly implicated in SO₂ or NO₂ sorbtion by soil or in conversion of sorbed SO₂ to either SO₄^2- or organic S, nitrifying organisms were implicated in the oxidation of sorbed NO₂ to NO₃. Labeda and Alexander (1978) extended this work and showed that nitrification in a neutral loam soil was not affected by continuous exposure of soil to 0.5 ppm of SO₂. When an already acidic soil was similarly treated, the nitrification rate was reduced, however. Oxidized S compounds appear to alter N oxidation primarily by altering the environment of soil organisms. Wainwright (1980) showed that, although atmospheric pollution altered the soil environment, it had little effect on the organisms and biological processes investigated.

Although sterile soils will sorb SO₂, Cracker and Manning (1974) showed that fungi isolated from soil incorporated ³⁵S into hyphae upon exposure to ³⁵SO_2.

Concurrent oxidation of S and reduction of NO₃^- under conditions of reduced aeration by organisms such as Thiobacillus denitrificans is one classical example of interactions of the N and S cycle. The following reaction generalizes the overall process:

The above scheme assumes complete N reduction and S oxidation. This process has been evaluated as a potable water treatment method to reduce the NO₃^- content of ground-waters (Sikora and Keeney, 1976). Reductions in N availability to growing crops have also been observed following elemental S addition to soils and have been attributed in part to enhanced denitrification (Martin and Ervin, 1953).

Although reduced S compounds in the presence of Thiobacillus denitrificans will cause denitrification, there is evidence that highly reduced forms of S (especially S^2-) inhibit denitrification (Knowles, 1981). At about 8 µmol/g soil, sulphide delays N₂O reduction, but thiosulphate has no such effect (Tam and Knowles, 1979). The effect of sulphide may also extend to reduction of NO (Sorensen et al., 1980). Interestingly, Kowalenko (1979) reported that thiosulphate slowed the rate of NO₃ reduction. Gould and McCready (1982) similarly showed that SO₃^2- and S₂O₃^2- both delayed NO₃ reduction. They postulate that in soils, partially oxidized S intermediates such as SO₃^2- and S₂O₃^2- are reduced under anaerobic conditions, generating S^2- that inhibits the reductases of N₂O and NO and causes feedback inhibition of denitrification, with attendant accumulation of NO₂^- .

The preceding has provided some examples of the ways C, N, S, and P interact and some of the controls on that interaction. Frequently, however, the stoichiometric relations among the cycles is used to predict the behaviour of several elements on the basis of one or two. Is such an approach valid? The following section will discuss this question.

9.3 NITROGEN, SULPHUR AND PHOSPHORUS STOICHIOMETRIC RELATIONSHIPS 

The frequent reporting of average nutrient concentrations for organisms and soils leads to a conclusion that these values are rather constant. Indeed, they are not. The C contents of bacteria summarized by McGill et al. (1981) ranged between 45 and 55%, and the N contents ranged from 5 to 19% in response to growth conditions. For fungi the corresponding ranges were 44 to 63% C and 1.3 to 10% N. Sulphur contents of bacteria and fungi can range between 0.09 and 1% and 0.09 and 0.3% respectively (Ribbons, 1970; Coughenour, 1978; Saggar et al., 1981b). Phosphorus contents of microbes similarly vary widely with values ranging from 0.2 to 5.0% for bacteria and 0.2 to 1.0% for fungi (Kowalenko, 1978; Hmeidan, personal communication).

In soils, N contents in surface mineral horizons range between 0.1 and 0.6% N with between 10 and 20 g C/g N over a broad range of soils. The ratios of N: S: P tend to be about 10:1.11.4:0.82.4 (Kowalenko, 1978).

Natural plant communities differ in their tissue nutrient concentrations, nutrient stocks, and nutrient turn-over rates in response to edaphic and climatic factors. Garten (1978), using principal component analysis to correlate the concentration of many elements reported for 110 North American plant species, showed that the concentration of elements such as N, P, and S in plant tissues were correlated; together they formed a set of elements related to the metabolism of nucleic acids and proteins.

Of immediate interest is to determine what balance exists over time between N, P, and S with increasing live biomass production by the plant community. On a red earth soil (Luvic yermosol) where deficiencies (in decreasing order of magnitude) of P, N, and S had been recorded, the uptake of each of these elements declined after about 8 weeks, with soil water non-limiting and with continuous growth, for both a C₃ (Thyridolepis mitchelliana) and a C₄ (Cenchrus ciliaris) grassland community (Christie, 1978, 1979). Moreover, although both communities differed in photosynthetic potential, the pattern of nutrient balance remained fairly constant over time (Figure 9.3). The mean value for N:S:P in the green shoots was around 18:1.5:1, irrespective of standing biomass crop. On a fertile grey cracking clay (Chromic vertisol), however, where nitrogen was the only limiting soil nutrient for growth, uptake of nitrogen by a natural Astrebla lappacae grassland declined over time, whereas phosphorus uptake remained fairly constant (Christie, 1982). A comparison of changes in value for the P:N ratio in the living shoots, together with critical values for the P:N ratio (Penning de Vries et al., 1980), for the Thyridolepis and Astrebla grasslands suggest that phosphorus is the major limiting nutrient for growth of the former community, whereas nitrogen is the major limiting nutrient for the Astrebla grassland (Figure 9.4). These results, although site-specific, suggest that one adaptive mechanism to account for the growth and distribution of these two natural grassland communities is associated with their differences in efficiency of absorption and utilization of limiting soil nutrients. Also, the N:S:P or N:P ratio appears to be specific for any one soil/vegetation system.

Measurements of nutrient concentrations in the soil environment alone are poorly correlated with the chemical composition of plant species because of physiological mechanisms that allow plant species to differentially absorb or exclude elements (Gerloff et al., 1966). Differences between species or communities in nutrient absorption will also be reflected in utilization of the element in biomass production. A comparison of efficiency of nutrient conversion into above-ground live biomass (ratio of Annual Carbon Production to Annual Mineral Uptake) for a number of unfertilized vegetation systems occurring over a wide range of edaphic and climatic conditions is made in Table 9.6. These dates indicate that the efficiency of utilization of any one nutrient varies, reflecting differences both in vegetation type and in soil fertility. For example, the highest values for efficiency of phosphorus use were found for Eucalyptus socialis and Acacia aneura, two woodland communities that occur naturally on soils grossly deficient in phosphorus.

Figure 9.3 Sequential changes in total biomass production (), nitrogen (), sulphur () and phosphorus (O) uptake over a 12 week continuous summer growing season for: (a) a C₃ Thyridolepis mitchelliana and; (b) a C₄ Cenchrus ciliaris grassland. (plotted from data in Christe, 1978, 1979; and sulphur data from Christie, unpublished data)

Figure 9.4 Sequential changes in the phosphorus: nitrogen ratio for natural Thyridolepis mitchelliana () and Astrebla lappaccea () grassland communities. Number in parentheses indicates the length of the continuous summer growing season in weeks. (plotted from data in Christie, 1981)

9.4 MINERALIZATION AND IMMOBILIZATION RELATIONS

9.4.1 Carbon and Nitrogen

Soil organic C and N are present in several distinct formssuch as proteins, peptides, carbohydrates, lignin, organic acids, aromatics, lipids, some hydrocarbons, nucleic acids, and amino sugars (Kononova, 1966; Bremner, 1967; Paul, 1970; Flaig, 1971; Flaig et al., 1975). The character and stability of soil organic matter cannot, however, be represented as a simple sum of the properties of these components (Minderman, 1968; Sorensen and Paul, 1971; McGill et al. ,1974; McGill and Paul, 1976; Paul and McGill, 1977). The humified component of soil organic matter has a role in stabilizing these elements but its chemical nature does not totally explain stability (McGill and Cole, 1981). Stability in soil is due to a combination of factors, including chemical recalcitrance, heterogeneity of components available for attack, physical protection, interaction with polyvalent cations, and adsorption to soil inorganic colloids (Ladd and Butler, 1975; Jenkinson and Rayner, 1977; Anderson, 1979; McGill et al., 1981). A close correspondence has been shown between C and N cycling through soil organic matter (McGill et al., 1975; McGill et al., 1981).

Mineralization or mobilization of N occurs as the C to which it is attached is oxidized to CO₂. This process occurs internally and is strictly catabolic. Therefore, N mineralization occurs when soil organisms use N-rich materials as an energy substrate. Consequently it is primarily the need for C rather than the need for N that causes N mineralization. Similar controls should apply to S mineralization from C-bonded S.

Table 9.6 Efficiency of nutrient use for a number of vegetation types 

	Total Biomass		Efficiency of nutrient use
	(g C m-2)a		(g C/g mineral nutrient)
Vegetation type/Location			Nitrogen		Sulphur		Phosphorus
Thyridolepis mitchelliana	135		29		322		555
Grassland, Charleville, Australiab
Cenchrus ciliaris	305		33		400		500
Grassland, Charleville, Australiab
Astrebla lappacea	360		50				250
Grassland, Charleville, Australiac
Eucalyptus socialis Woodland	2310		110				1670
Rankins Springs, Australia d
Acacia aneura	3370		35				1250
Woodland, Charleville, Australiad
Quercus robur	7960		48		345		625
Forest Virelles, Belgiume
European Forestsf			70125				8301670
Agricultural Cropsf			27 45				150230
aThroughout this paper C contents of the vegetation compartments have been derived following Ajtay et al. (1979).
bChristie (1978, 1979).
cChristie (1982).
dBurrows (1976).
eDuvigneaud and Denaeyer-De Smet (1970).
fTrom Bakuzis (1969) based on data of Ehwald (1957).

9.4.2 Sulphur

The literature on organic S has been reviewed by Freney (1967), Biederbeck (1978), and Fitzgerald (1978) and is discussed in Stewart et al. (chapter 8, this volume). In most non-saline soils organic S accounts for 90% or more of the total S. Two broad groups of compounds can be distinguished: HI-reducible S and C-bonded S. HI-reducible S is generally considered to be mainly sulphate esters (C-O-S) and sulphamates (C-N-S) with ester sulphates predominating (Fitzgerald, 1978). These two groupings are of significance because S bonded directly to C is likely to be incorporated into humic materials as components of amino acids etc., whereas sulphate esters may not.

1. C-bonded S appears to be part of the central humic core, whereas HI-reducible S may be restricted to the periphery of humic complexes. 
2. Potassium phosphate extraction at pH 7 seems to preferentially remove HI-reducible S.
3. Sulphate esters and SO₄^2- appear to exhibit somewhat similar adsorption characteristics.
4. HI-reducible S is correlated with C and N contents in acid soils but not in calcareous soils; whereas C-bonded S is correlated to C and N in acid and calcareous soils.
5. The relative proportions of HI-reducible and C-bonded forms of S frequently change down a soil profile.

Correlations of HI-S and C-bonded S with each other and with N and P_o have been calculated from several sources of data (Table 9.7).

Although mineralization of S has frequently been assumed to follow N in proportion to its abundance in soil organic matter, data summarized by Biederbeck (1978) show no consistent relationship between N and S mineralization. Where N and S mineralization are correlated, mineralization is often unrelated to the relative amounts of N and S in the soil. Tabatabai and Al-Khafaji (1980) found that although cumulative amounts of N and S mineralized were significantly correlated (r = 0.95 and 0.79 at 20 and 35°C, respectively), they were unrelated to organic C, total N or total S. Similarly Kowalenko and Lowe (1975) found that the relative amounts of N and S mineralized from four Canadian soils could not be predicted from ratios of C:N, C:S, N:HI-S or N:C-bonded S. They concluded: `although microbial mineralization of soil sulphur was closely related to mineralization of soil C and nitrogen, it did not parallel these elements'.

Table 9.7 Correlation coefficients for the relationships of S Fractions with each other and with N and P

		Correlation coefficients and reference
		Scott and	Lowe	Tabatabai and Bremner (1972)		Lee and Speir (1979)
		Anderson	(1965)	Surface	Subsurface	Without	With
		(1976)	chernozemics			strafford	strafford
						data	data
Organic
sulphate versus
	N	0.58***	0.66**	0.91**	0.80**	0.98***	0.99***
	Po	0.47**
C-bonded S versus
	N	0.85***	0.98***	0.81**	0.87***	0.65*	0.83***
	Po	0.18(NS)
	HIS	0.69***	0.63**			0.52(NS)	0.83***
*Statistically significant at P > 0.95
**Statistically significant at P > 0.99
***Statistically significant at P > 0.999
NS, Not Significant.

Sulphohydrolase activity, and changes in it, may indicate changes in either microbial or plant demand for S, or both, and may reflect changes in mode of S release over time.

C-bonded S compounds appear capable of repressing sulphohydrolase synthesis. It seems likely that sulphohydrolase release of SO₄^2- from ester sulphates is controlled also by the need for S and not strictly the need for C as occurs with N. Dodgson and Rose (1975) concluded that:

Studies on the production of arylsulphohydrolase activity in micro-organisms provide strong evidence that the appearance of the enzymes in fungi and bacteria is a reflection of the need to acquire sulphate for growth purposes. Expressed in the most simple terms, in times of sulphur sufficiency it seems probable that the presence of sulphate or sulphur-containing intermediates on the pathway to cysteine are sufficient to repress bacterial arylsulphohydrolases.

A similar conclusion was reached regarding glycerosulphohydrolases. Changes in sulphohydrolase activity may therefore reflect changes in demand for, and mineralization potential of, S.

9.4.3 Phosphorus

Barrow (1961), Anderson (1967), Hayman (1975), Halstead and McKercher (1975), Dalal (1977), and Kowalenko (1978) have reviewed the literature on soil organic P. Williams et al. (1960) observed: `the correlations of organic phosphorus with carbon and nitrogen are much lower than for sulphur, and it appears to be a less integral part of the organic matter.' In summarizing data on the effect of P, supply on C:P<sub>o</sub> ratio, Barrow (1961) concludes: `The data suggest that neither the rate of addition of organic materials nor the P, content of those organic materials has much effect on the P, content of the soil organic matter. They therefore imply the existence of some other powerful determinants.'

1. Changes in C:N were not accompanied by similar C:P_o changes upon cultivation of savannah soils.
2. Most P_o, is linked to C as an ester (C-O-P) bond, whereas N is usually directly bonded to C.
3. P_o appears to be more readily dispersed than the bulk of soil organic C and N.
4. Results of several studies have shown that P_o does not appear to be intimately associated with the bulk of soil organic matter.
5. The phosphate group, rather than the C atom, appears to be responsible for controlling stability of P_o in soil.
6. P_o content of many soils appears to be more closely related to P_i-sorbing power than to soil organic C or N.
7. P_o appears to be more variable in soil organic matter than are N or S.

Measurements of P mineralization are difficult because of either P adsorption or precipitation, or both. Most available data are indirect. Taken in total, however, one would conclude that P_o is a significant source of P for plants but that its rate of mineralization from soil organic matter, relative to that of C or N, is not strictly in proportion to its relative abundance or stoichiometry (McGill and Cole, 1981).

It may be concluded from the above that mechanisms controlling mineralization of N, S, and P are specific to each element and that the supply of organic substrate further controls the mineralization rate separately for each element. Stabilization of organic materials containing N, S, or P in soil is also seen to be influenced by the inorganic matrix. This involvement of the inorganic matrix in modifying elemental cycling is unique to soils and sediments; it is absent in marine or fresh-water environments beyond the direct influence of bottom sediments. Consequently, biomass stoichiometry has a fundamental control over nutrient cycling in marine systems but shares control with overriding physico-chemical mechanisms such as differential adsorptiondesorption, humification, and other phenomena in terrestrial systems.

Another feature unique to terrestrial systems pertains to litter decomposition. Litter forms the habitat for many groups of decomposers. Consequently, organisms decomposing litter are frequently decomposing their own habitat, thereby creating a constantly changing environment that in turn alters the suite of organisms present and subsequent decomposition dynamics. Such a situation is less pronounced, if observable at all, in marine systems.

The plant component, with its wide range in efficiency of nutrient conversion to above-ground biomass, permits further fluctuations (within limits) of elemental stoichiometries in response to their relative availabilities.

9.5 NUTRIENT DISTRIBUTION AND COMPARTMENT TRANSFERS 

Emphasis here is given to N, S, and P in the vegetative components. Flows in grassland soils are further dealt with by Stewart et al. (Chapter 8, this volume).

Detailed analysis of the size and distribution of the various nutrient pools found in any ecosystem provides a datum for the study of cycling processes (Richards and Charley, 1977). Three unfertilized natural Australian plant communities, differing widely in the size of storage pools, have been selected as examples for comparing nutrient distribution and turn-over (Table 9.8) viz. a Eucalyptus diversicolor forest (density 440 stems ha^-1; basal area 26 m² ha^-1), an Acacia aneura woodland (density 866 stems ha^-1; basal area 14.1 m² ha^-1), and a natural Thyridolepsus mitchelliana perennial grassland (basal area 560 m² ha^-1) found in cleared A. aneura woodlands. Values for the soil nutrient pools, of necessity, have been based on total pool because of data limitations and inherent problems in estimating the available soil fraction. In all systems, the soil compartment contained much greater total amounts of nutrients than the vegetation compartments. In all vegetation types, but especially for the woodland and forest, a higher proportion of the nitrogen than the sulphur or phosphorus pools was held in the organic material. Differences in biomass compartment size were also reflected in the relative distribution of nutrients in each system, except for nitrogen in the above-ground vegetation compartment of A. aneura. The latter value is the result of a higher stem nitrogen content in the leguminuous A. aneura than in E. diversicolor. Mean values for tissue nutrient content of E. diversicolor and A. aneura, however, are only one-third the nitrogen, one-fifth the sulphur, and one-sixth to one-tenth the P content of values recorded by Woodwell et al. (1975) for northern hemisphere temperate forest species.

Table 9.8 Nutrient distribution, annual cycle and compartment transfer for three Australian vegetation types 

Compartment/	Eucalyptus Diversicolor Foresta				Acacia Aneura Shrublandb				Thyridolepis Mitchelliana Grasslandc
Process	Carbon	Nitro-	Sul-	Phos-	Carbon	Nitro-	Sul-	Phos-	Carbon	Nitro-	Sul-	Phos-
		gen	phur	phorus		gen	phur	phorus		gen	phur	phorus
	(gCm-2)	(gNm-2)	(gSm-2)	(gPm-2)	(gCm-2)	(gNm-2)	(gSm-2)	(gpm-2)	(gCm-2)	(gNm-2)	(gSm-2)	(gpm-2)
Above-ground	10800	18.9	1.8	1.8	3585	42.1		1.4	82	3.4	0.30	0.18
vegetation
Litter	1365	22.4	2.8	0.7	695	10.9		0.3	12	0.25	0.02	0.01
Root					1270	22.4		0.62	53	0.9	0.05	0.02
Soil (0-100cm)	15050	780	305	180	7680	480	136	295	7790	540	136	379
Annual nutrient						6.16		0.19		2.3	0.20	0.12
uptake
Annual nutrient						3.11		0.11
retention
Annual litter					113	3.06		0.09	46	0.9	0.06	0.044
production
Annual litter						1.57		0.033		0.4	0.03	0.02
turnover
Annual root						1.82		0.051		0.3	0.02	0.11
turnover
Above-ground/	72.0	2.4	0.6	1.0	40.0	8.4		0.5	1.0	0.6	0.2	0.05
below ground (%)
Annual turn-over/						55		42		30	25	29
annual uptake (%)
aHingston et al. (1979)
bBurrows (1976)
cChristie (1978, 1979)

The total nutrient pool size must be complemented with information on short term requirements for vegetative production, because the rates at which nutrients pass through the soil litter sub-system regulates the productivity of the whole system, particularly on infertile soils (Charley and Richards, 1974). Compartment transfer rates for the A. aneura woodland and the T. mitchelliana grassland, both growing on the same highly infertile soil type and in the same climatic environment, were based on measured above- and below-ground biomass production rates, litter production rates, and root and litter turn-over rates over a 2 year period. The higher turn-over: uptake ratio for the woodland than for the grassland community, together with its greater annual litter production rate, was further evidence for a more efficient nutrient cycle for the woodland. This efficiency in turn would contribute to its advantage in efficiency of phosphorus use (Table 9.6). The N: S: P ratio for the grassland community for the processes of annual uptake, annual litter production, and annual litter and root turn-over was fairly constant, having a value of around 20:1.5:1. The ratio of N:S:P for an annual balance for a mixed-oak forest in Virelles, Belgium, for the processes of nutrient uptake and nutrient return were also fairly similar; the value ranged around 13:1.9:1 (Duvigneaud and Denaeyer-De Smet, 1970). Values for N:S:P ratio for the total nutrients incorporated annually into total net primary productivity (excluding amounts returned to the soil by leaching) for an Oak-Pine forest at Brookhaven, New York and a Northern Hardwoods forest at Hubbard Brook, New Hampshire (Whittaker et al., 1974) were 8:1.8;1 and 9:0.8:1 respectively (following Woodwell et al., 1975). This type of analysis now needs to be extended to a wider range of vegetation structural forms, with particular emphasis on the N:S:P balance in the biomass for the processes of annual uptake, retention, return and turnover.

There were a number of limitations in the Australian studies reported in Table 9.8. First, the nutrient flows into live and dead roots were not monitored separately. In addition, the rate of soil mineral change following mineral release from decaying organic residues also needs to be carefully quantified; for example, the relative amounts of phosphorus moving to the insoluble and soluble pools need to be determined for soils having a high sorption capacity. Nevertheless, for comparative purposes, this type of data analysis highlights the necessity to study transfer rates between compartments, residence times and interrelationships between compartments if nutrient dynamics in terrestrial communities are to be clearly understood.

9.6 LONG TERM NUTRIENT CYCLES AND CARBON PRODUCTION

Most reported nutrient budgets are annual, but as Woodmansee et al. (1981) point out, year to year variability may also be great, so that annual budgets may be quite different from long term balances. Ideally, the water cycle must be linked to both nutrient fluxes between ecosystem compartments and carbon production. Stable ecosystems must maintain relatively balanced nutrient transfers within their actively cycling portion over long periods of time.

The initial step in the nutrient cycle, nutrient absorption rate, is a specific function of the soil and plant community type. In addition, rates vary over time as biomass increases. The sequential changes in nutrient absorption rate (I) for two perennial grasslands, viz. C₃ T. mitchelliana and C₄ C. ciliaris communities (Figure 9.5), were derived following Williams (1948) as

where W_R is the community root weight and M the absolute nutrient content of the total biomass at time t. Root weight of these two species was well correlated with root surface area. The interrelationships between rainfall, above-ground live biomass production, nutrient uptake, and litterfall for a grazed (moderate grazing pressure) T. michelliana grassland community was derived from the net primary producution model of Hughes and Christie (1982) with historic rainfall figures for Charleville, Australia, for the period 19541975. There was an almost four-fold difference between years in carbon production in live biomass (mean 26 g C m^-2 yr^-1) and nutrient uptake (mean value: 1.15 g N m^-2 yr^-1, 0.13 g S m^-2 yr^-1, 0.06 g P m^-2 yr^-1) (Figure 9.6a, b). Furthermore, the tissue nutrient balance for N:S:P remained around 19:2:1, irrespective of total rainfall and live biomass production for any one year. Annual variations in rainfall and above-ground live biomass production were also reflected in litterfall but to a much smaller extent; values fluctuated around 18 g C m^-2 yr^-1. The long term mean values for mineral nutrient content of the litterfall were 0.38, 0.046, and 0.023 g m^-2 yr^-1 for nitrogen, sulphur and phosphorus, respectively, or around 35% of the mineral quantity absorbed (Figure 9.6c). This highlights the importance of the root compartment, soil organic, and primary mineral pools, as well as other external sources, for balancing the mineral nutrient cycles of stable, grazed grassland systems.

Figure 9.5 The pattern of above-ground biomass production rate and nutrient absorption rate for a Thyridolepis mitchelliana () and a Cenchrus ciliaris () grassland over the summer growing period. (a) Above-ground biomass production rate; (b) Nitrogen absorption rate; (c) Phosphorus absorption rate; (d) Sulphur absorption rate (plotted from data in Christie (1978, 1979) and sulphur data from Christie, unpublished data)

Figure 9.6 Interrelationships between: (a) total summer rainfall; (b) total summer above-ground biomass production and nutrient uptake; (c) annual litterfall and mineral content; and (d) standing litter crop in relation to grazing pressure, for a natural Thyridolepis mitchelliana grassland. (Analysis based on an analysis of historic daily rainfall records for the 20 year period 19541975 for Charleville, Australia). (Plotted from data in Hughes and Christie, 1982)

Overgrazing of natural grazing lands can lead to diminishing ecosystem surfaces and vegetation cover and eventually desertification of the land (e.g. Atjay et al., 1979). Values for litter yield for the natural T. mitchelliana grassland were derived in relation to moderate (or conservative use) and heavy grazing pressure by herbivores (sheep) over the same period of time. Litter yield values are dependent on the relative difference in the litterfall and litter decomposition rates. Litter yield values for the heavily grazed system fell rapidly within 1 year and thereafter remained at a new lower level, some one-third lower than the moderately grazed, stable system (Figure 9.6d). Disruptions to the mineral cycle can be induced quickly in overgrazed natural systems.

Two issues arise from this type of analysis. As Cole et al. (1977) point out, uncertainty exists in understanding precisely the source of, and relative contributions made by, litter pools of various ages and by soil humic material to the plant's annual nutrient uptake. Furthermore, the plant community has a finite capacity over time for absorbing minerals released from organic residues. What balance exists, then, between mineralization rate, mineral nutrient absorption rate, microbial use, and net primary production rate, in the short term and at different times of the year?

9.7 INTEGRATION OF THE CYCLES THROUGH SOIL ORGANIC MATTER

The plant component of the system provides the C used as an energy source and receives, in return N, S, and P. Control over N is exercised only through biological N₂ fixation; otherwise, the plant community is dominated by supply rate for N. Some control by plant roots over S and P supply through production of sulphohydrolases and phosphohydrolases has been suggested. Whether their activities increase supply from soil organic matter or increase the rate of root S or P recycling or both is still under investigation. Consequently the soil and the plant parts of the system both interact with the cycles of these elements. The patterns of C, N, S, and P cycling through humus because of the size of this pool profoundly affect the whole system. McGill and Cole (1981) proposed a dichotomous system to describe the comparative aspects of C, N, S, and P cycling through soil organic matter. The pertinent concepts of their system are listed below:

1. Elements stabilized in direct association with C (N and C-bonded S) are mineralized as a result of C oxidation to provide energy; this process is termed biological mineralization.
2. Elements existing as esters and related structures and not stabilized as a direct result of association with C are considered to be mobilized by extracellular or periplasmic hydrolases controlled by end product supply. Such a process is termed biochemical mineralization because of its operation outside the membrane.
3. By implication specific independent controls regulate each element. They are not regulated solely by a required stoichiometry.
4. Three concentric but independent cycles are envisaged (Figure 9.7).

Figure 9.7 Schematic illustration of interrelations of C, N, S and P cycling within soilplant systems (from McGill and Cole, 1981). Reproduced by permission of Elsevier Scientific Publishing Co.

Extrapolations from these concepts would suggest:

1. There may be preferential hydrolysis of P_o and HI-reducible S by both plant roots and micro-organisms if a specific need exists for those elements and if the organic substrates are available.
2. C-bonded S would closely follow the dynamics of organic C, and supply of S to plants from this source would require a microbial vector.
3. Terrestrial systems should exhibit increasing variability in ratios of C:N, C: S, and C: P respectively.
4. Accumulation of P_o and HI-reducible S should be possible even if C is not accumulating or if it is accumulating very slowly. Consequently, terrestrial systems may serve as expanding sinks for S and P. If such expansion is possible, what is its limit? Do similar concepts apply to sediments?

Predictions from this conceptual model have been examined by McGill and Cole (1981). They appear consistent with a wide range of data sources, but much more research is needed to define the limits of the model.

9.8 CONCLUSIONS

In soilplant systems, reduced C normally provides both the energy to drive nutrient cycles at the organism level and the skeleton on which to store many elements in organic form at the system level. Phosphorus is provided by the parent geologic material. The supply and rate of cycling of P, however, control the size of the biological and organic components of the system. Nitrogen and sulphur, both available from the atmosphere (or parent material-S), link the source of energy (C) with the vehicle controlling its cycling rate (P). At the ecosystem level, therefore, N and S are functions of the system and respond to fundamental controls such as P supply, climate, etc. They are not of themselves fundamental controls. They alter the system, however, by altering the environment (e.g. pH) of organisms essential to the system.

The stoichiometry of C, N, S, and P in soilplant systems is a function of the system, not of one particular group of organisms; this reflects element-specific mechanisms of stabilization, mobilization, and uptake. Involvement of the inorganic soil matrix in altering elemental cycling is fundamental to soilplant systems and sets them apart from the patterns of cycling observed in marine systems.

More research is needed to further define mechanisms of control and their quantitative effects on elemental cycling in soilplant systems.

9.9 REFERENCES

Ajtay, G. L., Ketner, P., and Duvigneaud, P. (1979) Terrestrial primary production and phytomass, in Bolin, B., Degens, E. T., Kempe, S., and Ketner, P. (eds) The Global Carbon Cycle, Chichester, Wiley, 129-182.

Anderson, D. W. (1979) Process of humus formation and transformation in soils of the Canadian Great Plains, J. Soil Sci., 30, 77-84.

Azcon, R., Barea, J. M., and Hayman, D. S. (1976) Utilization of rock phosphate in alkalin soils by plants inoculated with mycorrhizal fungi and phosphate-solubilizing bacteria, Soil Biol. Biochem., 8, 135-138.

Barrow, N. J. (1961) Phosphorus in soil organic matter, Soils Fert., 24, 169-173.

Biederbeck, V. O. (1978) Soil organic sulfur and fertility, in Schnitzer, M., and Khan, S. U. (eds) Soil Organic Matter. Developments in Soil Science 8. Amsterdam, Elsevier, 273-310.

Bremner, J. M. (1967) Nitrogenous compounds, in McLaren, A. D., and Peterson, G. H. (eds) Soil Biochemistry, New York, Dekker, 19-66.

Bromfield, A. R. (1975) Effect of ground rock phosphate-sulphur mixture on yield and nutrient uptake of groundnuts (Arachis hypogaea) in northern Nigeria, Exp. Agric., 11, 265-272.

Burford, J. R., and Bremner, J. M. (1972) Is phosphate reduced to phosphine in waterlogged soils?, Soil Biol. Biochem., 4, 489-495.

Charley, J. L., and Richards, B. N. (1974) Energy and nutrient dynamics of forest floors, Proc. Aust. Forest. Res. Conf., Coloundra, Queensland, 25-34.

Christie, E. K. (1979) Ecosystem processes in semiarid grasslands. II: Litter production, decomposition and nutrient dynamics, Aust. J. Agric. Res., 30, 29-42.

Christie, E. K. (1982) Biomass and nutrient dynamics in a C₄ semiarid Australian grassland community, J. Appl. Eco1., 18, 907-918.

Cole, C. V., and Heil, R. D. (1981) Phosphorus effects on terrestrial nitrogen cycling, in Clark, F. E., and Rosswall, T. (eds) Terrestrial Nitrogen Cycles, Ecol. Bull. (Stockholm), 33, 363-374.

Cracker, L. E., and Manning, W. J. (1974) SO₂ uptake by soil fungi, Environ. Pollut., 6, 309-311.

Dalal, R. C. (1977) Soil organic phosphorus, Adv. Agron., 29, 83-113.

Duff, R. B., Webley, D. M., and Scott, R. O. (1963) Solubilization of minerals and related materials by 2-ketogluconic acid-producing bacteria, Soil Sci., 95, 105-114. 

Duvigneaud, P., and Denaeyer-de Smet, S. (1970) Biological cycling of minerals in temperate deciduous forests, in Reichle, D. E. (ed.) Analysis of Temperate Forest Ecosystems, Berlin, Springer-Verlag, 199-255.

Fitzgerald, J. W. (1978) Naturally occurring organosulphur compounds in soil, in Nriagu, J. O. (ed.) Sulphur in the environment. Part II. Ecological Impacts, New York, Wiley, 391-443.

Flaig, W., Beutelspacher, W. H., and Rietz, E. (1975) Chemical composition and physical properties of humic substances, in Gieseking, J. E. (ed.) Soil components, Vol. 1, New York, Springer-Verlag, 1-212.

Freney, J. R. (1967) Sulfur containing organics, in McLaren, A. D., and Peterson, G. H. (eds) Soil Biochemistry, New York, Dekker, 229-259.

Gerloff, G. C., Moore, D. G., and Curtis, J. T. (1966) Selective absorption of mineral elements by native plants of Wisconsin, Plant Soil, 25, 393-405.

Ghiorse, W. C. and Alexander, M. (1976) Effect of micro-organisms on the sorption and fate of sulphur dioxide and nitrogen dioxide in soil, J. Environ. Qual., 5, 227-230. 

Gould, W. D., and McGready, R. G. L. (1982) Denitrification in several Alberta soils: Inhibition by oxidized sulfur intermediates, Can. J. Soil Sci. (in press).

Halstead, R. L., and McKercher, R. B. (1975) Biochemistry and cycling of phosphorus, in Paul, E. A., and McLaren, A. D. (eds) Soil Biochemistry, Vol. 4, New York, Dekker, 31-63.

Hayman, D. S. (1975) Phosphorus cycling in soil micro-organisms and plant roots, in Walker, N. (ed.) Soil Mircrobiology: A critical review, London, Butterworths, 67-91. 

Henderson, M. J., and Milazzo, F. H. (1979) Arysulphatase in Salmonella typhimurium. Detection and influence of C source and tyramin on its synthesis, J. Bacteriol., 139, 79-87.

Hutchinson, G. L., and Mosier, A. R. (1979) Nitrous oxide emission from an irrigated cornfield, Science, 205, 1125-1127.

Inman, R. E., and Ingersoll, R. B. (1971) Uptake of carbon monoxide by soil fungi, J. Air Pollut. Control Assoc., 21, 646-647.

Johnston, W. B., and Olson, R. A. (1972) Dissolution of fluorapatite by plant roots, Soil Sci., 114, 29-36.

Kononova, N. M. (1966) Soil Organic Matter, London, Pergamon Press.

Kowalenko, C. G., and Lowe, L. E. (1975) Mineralization of sulphur from soils and its relationship to soil carbon, nitrogen and phosphorus, Can. J. Soil Sci., 55, 9-14. 

Kowalenko, C. G. (1978) Organic nitrogen, phosphorus and sulphur in soils, in Schnitzer, M., and Khan, S. U. (eds) Soil Organic Matter. Developments in Soil Science 8. Amsterdam, Elsevier Scientific Publ. Co., 95-136.

Ladd, J. N., and Butler, J. H. A. (1975) Humus-enzyme systems and synthetic organic polymer-enzyme analogs, in Paul, E. A., and McLaren, A. D. (eds) Soil Biochemistry, Vol. 4, New York, Dekker, 143-194.

Lee, R., and Speir, T. W. (1979) Sulphur uptake by ryegrass and its relationship to inorganic and organic sulphur levels and sulphatase activity in soil, Plant Soil, 53, 407-425.

Lowe, L. E. (1965) Sulphur fractions of selected Alberta soil profiles of the Chernozemic and Podzolic Orders, Can. J. Soil Sci., 45, 297-303.

McGill, W. B., Hunt, W. H., Woodmansee, R. G., and Reuss, J. O. (1981) PHOENIX: A model of the dynamics of carbon and nitrogen in grassland soils, in Clark, F. E., and Rosswall, T. (eds) Terrestrial Nitrogen CyclesProcesses, Ecosystem Strategies and Management Impacts, Ecol. Bull. (Stockholm), 33, 49-115.

McGill, W. B., Paul, E. A., and Sorenson, H. L. (1974) The role of microbial metabolites in the dynamics of soil nitrogen, Tech. Rept. No. 46. Canadian Committee for the International Biology Program Matador Project, University of Saskatchewan, Saskatoon, Saskatchewan.

Malo, B. A., and Purvis, E. R. (1964) Soil absorption of atmospheric ammonia, Soil Sci., 97, 242-247.

Minderman, G. (1968) Addition, decomposition and accumulation of organic matter in forests, J. Ecol., 56, 355-362.

Moghimi, A., Tate, M. E., and Oodes, J. A. (1978) Characterization of rhizosphere products especially 2-ketogluconic acid, Soil Biol. Biochem., 10, 283-287.

Okamura, H., Murooka, Y., and Harada, T. (1977) Tyramine oxidase and regulation of arylsulphatase synthesis in Klebsiella aerogenes, J. Bacteriol., 129, 59-65.

Paul, E. A. (1970) Plant components and soil organic matter. Recent Adv. Phytochem., 3, 59-104.

Penning de Vries, F. W. T., Krul, J. M., and Van Keulen, H. C. (1980) Productivity of Sahelian rangelands in relation to the availability of nitrogen and phosphorus from the soil, in Rosswall, T. (ed.) Nitrogen Cycling in West African Ecosystems, Stockholm, NFR, 95-113.

Rhodes, L. H., and Gerdeman, J. W. (1978) Hyphol translocation and uptake of sulfur by vesicular-arbuscular mycorrhizae of anion, Soil Biol. Biochem., 10, 355-360. 

Ribbons, D. W. (1970) Quantitative relationships between growth media constituents and cellular yield and composition, in Norris, J. R., and Ribbons, D. W. (eds) Methods in Microbiology, Vol. 3A, New York, Academic Press, 297-304. 

Richards, B. N. and Charley, J. L. (1977) Carbon and nitrogen flux through native forest floors, Proc. Conf. Nutrient Cycling in Indigenous Forest Ecosystems, Como, Western Australia, 1-16.

Saggar, S., Bettany, J. R., and Stewart, J. W. B. (1981a) Sulfur transformations in relation to carbon and nitrogen in incubated soils, Soil Biol. Biochem., 13, 499-511. 

Saggar, S., Bettany, J. R., and Stewart, J. W. B. (1981b) Measurement of microbial sulfur in soil, Soil Biol. Biochem., 13, 493-498.

Scott, J. M., and Spencer, B. (1968) Regulation of choline sulphatase synthesis and activity in Aspergillus nidulans, Biochem. J., 106, 471-477.

Sikora, L. J. and Keeney, D. R. (1976) Evaluation of a sulphur- Thiobacillus denitrificans nitrate removal system, J. Environ. Qual., 5, 298-303.

Sorensen, L. H., and Paul, E. A. (1971) Transformation of acetate carbon into carbohydrate and amino acid metabolites during decomposition in soil, Soil Biol. Biochem., 3, 173-180.

Spencer, B., Hussey, E. C., Orsi, B. A., and Scott, J. M. (1968) Mechanism of choline-O-sulphate utilization in fungi, Biochem. J., 106, 461-469.

Stevenson, F. J. (1967) Organic acids in soil, in McLaren, A. D., and Peterson, G. H. (eds) Soil Biochemistry, New York, Dekker, 119-146.

Tabatabai, M. A., and Al-Khafaji, A. A. (1980) Comparison of nitrogen and sulfur mineralization in soils, Soil Sci. Soc. Amer. J., 44, 1000-1006.

Tam, T. Y., and Knowles, R. (1979) Effects of sulfide and acetylene on nitrous oxide reduction by soil and by Pseudomonas aeruginosa, Can. J. Microbiol., 25, 1133-1138.

Whittaker, R. H., Bormann, F. H., Likens, G. E., and Siccama, T. G. (1974) The Hubbard Brook ecosystem study: Forest biomass and production, Ecol. Monogr., 44, 233-254.

Williams, R. F. (1948) The effects of phosphorus supply on the rates of intake of phosphorus and nitrogen, and upon certain aspects of phosphorus metabolism in gramineous plants, Aust. J. Sci. Res., B1, 331-361.

Woodwell, G. M., Whittaker, R. H., and Houghton, R. A. (1975) Nutrient concentrations in plants in the Brookhaven oak-pine forest, Ecology, 56, 319-332.

Yeung, P. Y. (1980) Soil acidification by emitted sulphur in the Athabasca Oil Sands area. M.Sc. Thesis. Dept. Soil Science, University of Alberta, Edmonton, Alberta, Canada.