Tải bản đầy đủ - 0trang
VII. Sulfate Retention in Soil
N. S. PASRICHA AND R. L. FOX
Extractable Sulfate Sulfur, Solubility of Sulfate, and Calculated Sulfate Adsorption Maxima
of Some Highly Weathered Soils of Puerto Rico'
0- 120 cm
(pg S g-' soil)
Adapted from Fox (1982) with permission.
Sulfate extracted with 500 pg ml-1 P.
Sulfate in saturation extract.
Maximum sulfate adsorbed as determined from an isotherm based on the Langmuir equation.
an abundance of sorbed sulfate. This is not always evident from published
data because most investigations of sulfate in tropical soils have been
restricted to relatively shallow depths. Also, the extraction procedures most
frequently used have not effectivelyextracted sulfate, nor has the extracted
SO, been effectively determined (Fox et al., 1987). Nevertheless, in assess-
PLANT NUTRIENT SULFUR
Phosphate-Extractable SO,-S in Profiles of Selected Soils from the Humid
Depth increment (cm)
a Adapted from Fox and Blair (1986). Values in table body are in kilograms
ing the available SO4supply, the quantity of sorbed SO, in the subsoil, in
addition to that contained in the surface horizons, should be considered,
because adsorbed SO, is a feature of subsoils in most highly weathered soils
of the tropics. However, deep root penetration is a precondition for adsorbed SO, utilization in such soils (Probert and Jones, 1977).
Sulfate accumulates to very high levels in some soils, especially in
weathered Andepts, which contain gellike materials, and acid soils, which
contain oxides of iron and aluminum. This is illustrated in Table IV.
Phosphate-extractable S from 19 New Zealand soil profiles to a depth of
2 m was as follows:
SO,-S range (kg ha-')
Number of soils
N. S. PASRICHA AND R. L. FOX
All soils that contained more than 10,000 kg SO4-S ha-' developed in
volcanic ash and cinders. All were from the North Island and were highly
weathered. Incipient S deficiency is encountered in the area. Soil profiles
that contained less than lo00 kg S0,-S ha-' were from the South Island,
from areas where S deficiency is a major nutritional problem for white
clover -ryegrass pasture production. Sulfate does not accumulate in large
amounts in profiles developed in volcanic ash until weathering is well
Coarse textured sandy soils of the subtropics, especially those developed
on alluvium with neutral to alkaline reaction, adsorb little sulfate (Bahl
and Pasricha, 1984). In general, surface soil materials adsorb less sulfate
than do subsoil materials. No doubt, eluviation of clay-sized materials is
partially responsible, but organic matter and phosphate accumulations in
surface horizons, which block sulfate sorption sites, are major factors. Even
when eluviation of clay is of little consequence, the relationship is clearly
seen, as in the case of Oxisols. Sulfate from surface horizons is easily
leached to lower depths, where it accumulates in large quantities, especially
in highly weathered Oxisols and Dystrandepts (Fox, 1974). Sulfate distribution in the Nipe profile is typical of highly weathered Oxisols that have
been examined (Table IV). The fact that SO, solubility continued to
increase with increasing depth, even though saturation percentage decreased, suggests that adsorbed SO, and saturation percentage are not the
only factors controlling SO, availability. For Andepts in Hawaii and a
diverse group of Puerto Rico soils, sulfur solubility was about 5 pg ml-'
when saturation was 60-80% (Hasan et al., 1970; Fox, 1982). Sulfur
nutrition is probably borderline deficient for some crops when soil solutions contain 3 - 5 mg liter-' SO,-S. Upland soils of the Llanos of Columbia contain little adsorbed SO,. The mean is about 150pg g-I and sulfate
solubility is low (Fox, 1974).These soils require about 60 pg S g- of soil to
attain 5 pg ml-' S in solution (Fig. 5).
Sulfate is readily desorbed from soils by phosphate and hydroxide. Organic anions also compete for adsorption sites with sulfate and in this way
may account for the low quantity of sulfate found in surface soils. Desorption of SO, results in decreased concentration of SO, in solution. Repeated
desorption produces a relatively smooth desorption curve over the concentration range usually encountered in leached tropical soils (low solution
concentration). This, we believe, is evidence for sulfate (ligand) adsorption
as the principal mechanism for SO, retention by such soils. At higher
concentration, other mechanisms of SO, retention probably operate, including electrostatic adsorption associated with positive charge.
In any case, adsorption -desorption of sulfate proceeds relatively rapidly
as compared with phosphate. For example, a 24-hr equilibration time was
PLANT NUTRIENT SULFUR
sod-s in solution (pg m ~ ' )
Figure 5. Sulfate sorption by some diverse soils from tropical America: 0, Typic Dystrandept, Costa Rica;X, Nipe, Puerto Rico; A, Llanos, Columbia; 0, Terra Roxa, Brazil; A,
Eutropept, Costa Rica; 0, 1963-1965 ash, Costa Rica (Fox, 1974; by permission of the
publishers, Buttenvorth-Heinnemann Ltd. 0 )
more than adequate for obtaining a constant concentration of sulfate in
solution (Hasan ef al., 1970), whereas, for phosphate, up to a 6-day equilibration time was required (Fox and Kamprath, 1970). Data in Fig. 6
suggest that at levels of extractable SO,- S above 1000 mg/kg, the solubility
of SO, decreases abruptly. Each of the two points represent five soil samples (all subsoils), two from a highly weathered ash layer (60- 100 cm) and
the remainder developed in weathered basalt scoria. These data suggest one
or more of the following criteria:
1. Up to approximately 100 mg S kg-' soil, availability is controlled by
SO4- S in solution by weakly adsorbed SO4.
2. In the range of 100- lo00 mg S kg-' soil, SO, adsorption assumes an
important role and availability increases with quantity adsorbed.
3. In the range of 1000-3500 mg S kg-' soil, availability is considerably
Perhaps decreased availability at high levels of extractable SO, is because
adsorption capacity increased more than SO, quantity, or perhaps because
of precipitated iron or aluminum sulfate compounds of low solubility
(Wolt and Adams, 1979; Wolt ef al., 1992).
N. S. PASRICHA AND R. L. FOX
Mean extractable SO., -S (mg S/kg)
Figure 6. Sulfur contents of ryegrass indicator plants in relation to SO, extracted with
phosphate from New Zealand soils (R. L. Fox, P. M. Cooper, and W. M. H. Saunders,
Indicated yield potential of ryegrass at 0.175% plant S is approximately
80% of the maximum attained (approximately 0.22% for maximum yield).
What do these data indicate about S solubility? Information is not sufficient for us to make a firm statement about ryegrass, but if data for maize
apply here, approximately 5 mg liter-' of S in solution is adequate for
maximum yield and only 1 to 2 mg liter-' should be adequate for 80%
yield. Thus, it appears that these soils are undersaturated with SO, (assuming that solution concentration is being controlled by adsorbed SO,).
Given proper environmental conditions and time for precipitating aluminum hydroxy sulfate, lower concentrations may be expected, thus, greater
quantities of extractable SO,.
Adsorption curves are useful for describing, studying, and managing the
SO, status of soils. They integrate and reflect many aspects of mineralogy,
chemistry, and management history of soils. The concentration of SO, in
solution, as predicted by sulfate sorption - desorption curves (the equilibrium concentration when SO, is neither sorbed or desorbed), provides
valuable information on plant nutrition. It indicates the immediate concentration at which sulfate should be available to plants and the concen-
PLANT NUTRIENT SULFUR
tration of sulfate in water that drains from that horizon (Fox, 1982). It also
provides an explanation for the fact that crops may be adequately supplied
with S, which usually is 3-5 mg SO4-S liter1, even though the soil is
being leached with rainwater that contains much less S than that. An
example of this has been reported for Hawaii by Hasan et al. ( 1970). Akaka
surface soil materials equilibrated at 5 mg SO,-S liter-' and subsoil material at 3 mg SO4-S liter-' even though predicted SO,-S in rainfall at that
location is approximately 0.35 mg liter1.Annual rainfall is approximately
4500 mm and contains 16 kg S ha-'. Sugarcane and pasture vegetation
growing at this and similar locations give no obvious evidence of S deficiency.
Evapotranspiration probably does not concentrate the leachate more
than 50%. There is, therefore, a 10-fold difference in the concentration of
rainwater and predicted soil solution concentration. A probable explanation for this discrepancy is that SO, in rainwater is augmented by S leached
from vegetation, from leaves, stems, and roots, and from dry deposition in
the plant canopy and on the soil. Further reference to this effect was made
in Section IV dealing with S cycling.
Adsorption curves can also be used to follow the course of soil development. A good example is sulfate adsorption by highly weathered soils on
volcanic ash in the humid tropics, where soil materials and climatic conditions promote rapid removal of silicon and accumulation of hydrated iron
and aluminum oxide. Sulfate concentration in solution is related to the
degree of saturation of the exchange complex with sulfate or with other
specifically adsorbed anions.
Sorption maxima for highly weathered soils are usually associated with
10-2Opg SO,-S ml-I (Fox, 1982). Studies of SO, adsorption by some
West Indian soils by Haque and Walmsley (1974) indicate that higher
solution concentrations do not conform to the Langmuir equation. Therefore, SO4-S concentrations of approximately 10 pg ml-* or less should be
used for calculating adsorption maxima in such soils.
Some specifically adsorbed anions and structural anionic impurities can
give oxide surfaces a net negative charge, thus shifting the point of zero net
charge (PZNC) to lower pH values. The PZNC of synthetic hematite,
corundum, boehmite, and geothite occurs at pH values greater than 7 and
hydration can increase the pH at zero net charge (Parks, 1965). The PZNC
of surface horizons of ferruginous soils is, however, usually below pH 5.0.
Mekaru and Uehara ( 1972) demonstrated that phosphate adsorption increased the CEC of highly weathered soils, resulting in negative adsorption
of NO; and CI-. Such variable-charge soils, in which the sign and magnitude of charge on the solid phase are determined by the chemical environment, adsorb significant amounts of Sot- (Barrow, 1967; Fox, I980b).
N. S. PASRICHA AND R. L. FOX
Variable-charge colloids dominate the mineralogy of most soils of the
humid tropics because weathering and leaching processes dedicate soil,
forming minerals enriching them in hydrated oxides of Fe and Al. Materials such as these, together with kaolin clays and organic matter, give rise to
In soils in which permanent negative charge predominates, practically all
of the soil SO, is in the soil solution. But in variable-charged soils, especially if they are sufficiently acid to be near the point of zero net charge,
most of the SO, is sorbed. Thus, SO, sorption by soils requires positive
charge. Net positive charge is not required. To attain net positive charge in
whole soils requires variable-charge colloids and lower pH than is common
to highly weathered soils. The final result, however, depends on at least
three possibilities: (1) the point of zero net charge for certain components
of the soil system may be sufficiently high for net positive charge to
develop on that component at soil pH values near 7.0, (2) adsorbed SO,
may be held as ligands or by mechanisms other than electrostatic attractions, or (3) sulfate may be retained in compounds or complexes that are
relatively insoluble at low pH but are soluble in phosphate extractants.
Soils derived from volcanic ash may adsorb large amounts of SO, owing
to hydrous oxides of Fe and Al and allophanic clays (X-ray amorphous A1
silicates) present in them (Hingston el al., 1972). Rajan (1979) observed
that SO, is adsorbed on a net positive charged surface as a bidentate
forming a six-membered ring, displacing either two aquo or hydroxyl
ligands. On neutral or negative surfaces, SO, is adsorbed as a monodentate,
displacing one aquo or one hydroxyl ligand. This makes the surface charge
Organic ligands play an important role in determining the SO, adsorption capacity of soil and subsequent amount of OH released (Fuller ef al.,
1985). Evans (1986) noted that increases in dissolved organic carbon resulted in increased SO, transport in soil columns at pH 4.6. Similarly,
Gobran and Nilsson ( 1988) found that forest floor leachates containing
dissolved organic ligands inhibited SO, retention by a Spodosol soil at SO4
levels less than 7 mM. The ability of soils to adsorb SO, is an important
factor in determining the effect of acidic deposition on the transport of H+
and cations in terrestrial ecosystems. Inskeep ( 1989) showed that inhibition of SO, adsorption was related to the quantity of oxygen-containing
functional groups rather than soluble C. These results indicate that organic
PLANT NUTRIENT SULFUR
acids compete for SO, adsorption sites and that the presence of organic
acids in soil solution will influence SO, adsorption capacities.
The input of acidic deposition containing mobile anions like SO, may be
responsible for increased cation leaching from soils to ground and surface
waters in some forest ecosystems (Seip, 1980; Van Breeman ef al., 1983).
However, mobility of SO, has received considerable attention because net
SO, retention by soils can result in decreased cation leaching. Where an
equivalent displacement of OH or other anions does not occur during the
SO, retention process, cations are coadsorbed with SO, (Khanna and
Beese, 1978; Singh ef al., 1980). Highly weathered tropical soils, which
adsorb substantial amounts of SO,, are considered to be resistant to accelerated cation leaching (Huete and McColl, 1984).
The mechanism of SO, adsorption involves an exchange of OH coordinated with an Fe atom of the oxide structure, resulting in increases in pH at
higher levels of sorption. When SO, is specifically sorbed by goethite or
hematite, negative charge on the surface increases, increasing the ability of
the material to retain cations. In tropical soils, this aspect of SO, sorption
could be valuable, given the extremely low cation exchange capacities at
the low pH of many of these soils. Couto et al. (1979), however, did not
observe a constant increase in pH per unit SO., sorbed for different soil
samples. The Ap horizons sorbed small amounts of SO, and the effect on
pH was not marked, whereas B2 horizons sorbed high amounts of SO,,
and sorption clearly decreased as the pH of the equilibrium solution
increased. The differences between horizons were probably due to the
higher organic matter content of the Ap horizons, because no clear changes
in mineralogical composition were observed. Blocking of positively
charged sites may occur as a result of occupation of these sites by organic
anions or coating of oxide surfaces by organic matter. Phosphate quantity
and intensity also are almost invariably greater in surface horizons. We
may well suppose that organic matter and its decomposition products are
involved. Adams and Rawajfih ( 1977) have suggested that SO, is precipitated as insoluble basic A1 and Fe sulfates in these kinds of soils.
Acid soils that have been limed or those that have naturally high pH,
such as the Tropudalf studied by Couto ef al. (1979) and alluvial soils
studied by Bahl and Pasricha (1984), may require SO, application as a
consequence of their lowered SO, sorption capacities. In addition, liming
acid soils enhances movement of SO, out of the limed zone of soils.
Sulfate retention involves more than one mechanism. Our results (Fox,
1984) suggest that mechanisms of SO, sorption and phosphate sorption are
similar, and that both ions compete for the same sorption sites, although
sorbed SO, does not compete strongly with phosphate (Gebhart and Coleman, 1974). As a consequence of phosphate fertilization, sulfate may be
N. S. PASRICHA AND R. L. FOX
so4 in supernatant (vg mi-’)
Figure 7. The residual influence of phosphate fertilizers on SO, sorption by an Oxisol in
Hawaii (Fox ef a!.. 1971).
displaced from some soils. This effect may persist for several years as is
demonstrated by the shift in a SO, sorption curve that resulted from a
massive phosphate application 12 years earlier (Fig. 7).
Phosphate can extract relatively large amounts of SO, from variablecharge soils, although this may not be obvious if only surface horizons are
investigated (Neptune ef al., 1975). Even CaC1, extraction in a wide
soil :solution ratio may desorb substantial quantities of SO, from highly
buffered soils without depressing SO, concentration below an adequate
level, as indicated by vigorously stirred solution culture experiments. Thus
the external SO, requirement of plants growing in acid, variable-charge
soils and permanent-charge soils may be distinctly different.
Phosphate-extractable SO,- S in profiles of selected soils from the humid
tropics are presented in Table V. It is obvious from such diverse data that it
is unwise to generalize about the quantity of SO, in soils of the humid
tropics, except to say that somewhere in the profile of most of these soils
there is a significant quantity of SO,. Whether such sulfate is available to
plants cannot be inferred from the quantity of SO, alone. Growth of
stylosanthesis on leached sandy soils of tropical Queensland has indicated
that some species can utilize sulfate from considerable depth in such soils
PLANT NUTRIENT SULFUR
VIII. DIAGNOSIS OF SULFUR NEEDS
1 . Extractable Sulfur
Estimates of plant-available S in soils have been attempted by several
scientists using a variety of methods. Phosphate-extractable S using monocalcium phosphate (Fox et al., 1964) or monopotassium orthophosphate
( KH2P04)(Ensminger and Freney, 1966) extracts soluble plus adsorbed
SO,. Repeated extraction is required for quantitative removal of sulfate
from many highly weathered soils (Hasan et al., 1970). Hot-water-soluble
sodium acetate (NaOAc, pH 4.8) and sodium bicarbonate extracts include
some organic S also. Phosphate extracts correlate best with crop yields or S
uptake by plants (Barrow, 1967; Scott, 1981) and soil tests appear to be
most valuable on sandy soils. Beaton et al. (1985) reported that soils with
the same characteristics but with soil sulfur concentrations between 7 and
15 kg ha-' would be expected to respond to applied S in dry years but not
in wet years. Brogan and Murphy (1980) did not find soil testing for S to be
promising. They observed, however, that soils with more than 50% sand
and less than 3% organic carbon would likely respond to applied S.
Determining SO4in soil extracts is fraught with difficulties. The methylene blue method of Johnson and Nishita (1952), although sensitive, has
many disadvantages (Lee ef al., 1981; Fox et al., 1987). Ion chromatographic methods are sensitive and more reliable (Kalbasi and Tabatabai,
1985) and are advantageous for simultaneous measurement of other
anions. However, equipment is expensive. Turbidimetric methods have
been widely used for SO, assay of soil extracts since Chesnin and Yien
(1950) introduced one such procedure. Hesse ( 1957) used a sodium acetate
extractant. He attributed interference to colloidal organic matter. Fox el al.
(1964) attempted to overcome this difficulty by extracting with
Ca(H,PO,), instead of NaOAc and digesting the extracts with oxidizing
agents such as nitric-perchloric acid; but even this may not eliminate the
problem of BaSO, precipitate suppression in extracts of some soils (Vander
Zaag et al., 1984). Introduction of BaSO, seed crystals as proposed by
Tabatabai (1974) and adding additional Ba as solid BaCI, in phosphate
extracts that have been evaporated and digested with nitric-perchloric
acid produced higher values. Searle ( 1979) also discovered soil extracts that
failed to yield satisfactory precipitates. Absorbance values less than the
zero standard were obtained. Additions of a SO, spike to these extracts
produced spuriously low results and several activated carbon treatments
N. S. PASRICHA AND R. L. FOX
did not overcome the interference. Fox et al. (1987) proposed a more
reliable, although more complicated, turbidimetric method for determining phosphate-extractable sulfate in tropical soils. This method consistently yielded more SO, than other turbidimetric procedures.
Inconsistencies between soil tests for S and crop performances have been
reported widely. These inconsistencies may result from seasonal effects on
extractable S. Castellano and Dick (1991) observed great seasonal variability in SO, levels even in control plots that have not received any S applications for at least 20 years.
Critical levels as high as 10 mg SO,-S kg-' soil have been reported for
the production of winter rape in the Pacific Northwest (Murray and Auld,
1986). One might reach different conclusions, depending on the time of
sampling. Current S soil test recommendations are usually based on samples derived from surface horizons. In environments where there are periods of significant evaporation, movement of SO, from subsurface to
surface may affect soil test results, depending on time of sampling. Additionally, plants usually develop roots below 60 cm and subsoils vary in
their SO, contents. Leaching is an even greater problem. So also is knowing
how to evaluate adsorbed sulfate and subsoil sulfate, i.e., whether it is
adsorbed or not.
2. Optimum N :S Ratio in Soil
Because S is an important component of protein, balanced N :S fertilization is important in obtaining optimal yields and protein contents. If the
N : S ratio is too great, protein synthesis may be restricted and N may
accumulate in plants in nonprotein forms (Pasricha and Randhawa, 1975).
Applications of N to soils deficient in S may lead to decreased yields
(Janzen and Bettany, 1984; Nyborg et al., 1974). The optimal fertilizer
N :S ratio vanes among soils because of differences in available soil N and
S levels. However, one estimate of a suitable available N: S ratio [(soil NO,
N fertilizer N)/(soil SO4-S fertilizer S)] is approximately 7 for upland
conditions (Janzen and Bettany, 1984).
This probably exceeds S requirements in the tropics. A suitable N:S
ratio in solution for sugarcane growth is approximately 10: 1 (Fox, 1976).
As a first approximation, this ratio could be a guide to fertilizer applications for nonlegumes in those areas of the tropics where S removal by crops
is great in comparison with soil and rainfall S (Fox and Blair, 1986).
However, because SO, is not so easily leached or so easily reduced as NO,,
and because the internal N :S ratio is greater than 10, it seems reasonable
to suppose that a 10: 1 N:S ratio in fertilizer, consistently applied, will
more than adequately meet the S requirement of nonlegumes.