Tải bản đầy đủ - 0trang
II. Extent of Sulfur Deficiency
N. S. PASRIClIA AND R. L. FOX
strated that incipient S deficiencies do abound. This is related to low
concentrations of S in rainwaters and generally low levels of organic S in
soils (Fox and Blair, 1986). Organic matter contents of tropical soils are
not so low as is sometimes supposed. This fact is evidence for the stability
of such organic matter that persists. It should not be assumed, however,
that the availability of S in the resistant organic matter will be as readily
available to plants as that which has already decomposed. The implications
of this for the future are clear; organic S will be decreasingly important with
increasing time of arable agriculture. To complicate the problem, some of
the soils of the humid tropics adsorb S q - strongly, especially in subsurface
horizons. These soils may be S deficient even though they contain much
SO, (Hue et al., 1990). Sulfate solubility, and presumably its availability to
plants, is related to the quantity of adsorbed SO, in relation to the capacities of soils to adsorb SO,. Most soils developed in weathered volcanic ash
adsorb SO, strongly.
Widespread S deficiencies in groundnut in the West Africa savanna have
been recognized for 40 years and are well documented. Fox (1980a) estimated that exported S in 750,000 tons of groundnut kernels would amount
to 1500 tons S per year, assuming a nominal content of 0.2% S in the
kernels. Although this is not a large quantity by the standards of S reserves
in most temperate zone soils, it is a serious drain for a system in which S
deficiency has been chronic. The magnitude of the problem is placed in
better perspective by considering that the total expected S content of rain
for the groundnut-growing belt of northern Nigeria ( 1.2 million hectares)
amounts to about 1400 tons, essentially equivalent to S being exported
Kang et al. (198 1) reported results of laboratory and greenhouse expenments comparing the S status of savanna and forest upland soils of 30
surface soil samples (0- 15 cm) from Nigeria. The S status was not related
to parent material or soil type. Total and extractable S levels were in the
following order: forest zone > derived savanna > Guinea savanna. Observed sulfur deficiency was most acute in soils from Guinea savanna and
least acute from the forest zone. Sulfur responses are most frequently
observed in the savanna zone (Enwezor, 1976).
Sulfur deficiencies have been reported from several locations in the
humid tropics and a few studies on the sulfur status of soils there have been
reported (Bornemisza and Llanos, 1967). However, much more needs to
be done before accurate predictions can be made on the magnitude of S
problems, or probable requirements and effectiveness of S fertilizers. Many
Central American soils, particularly those under more or less permanent
crops such as coffee, cocoa, sugarcane, or pastures, have accumulated
organic matter in the surface and consequently contain high organic S
levels (Bornemisza et al., 1978; Burbano and Blasco, 1975; Granados,
PLANT N U T R I E N T SULFUR
1972; Hardy and Bazan, 1966). However, a favorable C :S ratio is more
important than a large store of organic matter for S response under tropical
conditions (Bromfield et af.,1982). This is because the ratio influences
amounts of S that will be mineralized (or immobilized) during microbial
decomposition of organic matter.
Bornemisza ( 1990) reported detailed information on S distribution and
S problems in Central America. About half of the soils in the Entisol and
Inceptisol orders respond to S, especially after N and P deficiencies have
been corrected. Soils that are coarse textured are most responsive because
of the leaching that is associated with them. Cordero et af. (1986) reported
that in some of these soils, S deficiencies are the main production-limiting
factor. Andisols are important in the area because of widespread volcanic
ash influence, and research on S problems of these soils has been reported
(Jimenez and Cordero, 1988).The more highly weathered Andisols adsorb
large amounts of SO, on exchange sites below the A horizon (Gebhart and
Coleman, 1974). However, SO, can be displaced by phosphate fertilizers,
which can result in deficiencies, as has occurred in El Salvador (Muller,
1965). Sulfur deficiencies have been reported on Mollisols particularly
after the P status was improved by P fertilization (Kass et af., 1984).
Research in Panama and Costa Rica indicates considerable variation in S
status of Alfisols. Research in Panama has confirmed S problems in these
soils both in pot studies and in field conditions. Comparable soils in Brazil
have shown strong pH-dependent SO, adsorption capacity (Couto ef al.,
1979). In the Pacific lowland of Central America, S deficiency in Vertisols
becomes evident only after correcting P deficiencies and consequent leaching of SO, (Bornemisza et al., 1978). In Brazil, accumulations of SO,
results in increased cation retention, which can partially compensate for
low exchange capacities of soils. Insufficient S usually becomes a problem
for sugarcane after problems of N, P, and sometimes K have been
Williams and Andrew (1970) were unable to outline any major occurrence of S deficiency in tropical Australia. It cannot be inferred, however,
that S is adequate on all soils and one must look further for an explanation
of this apparent sufficiency of S. The accession of S by the plant/soil system
in this nonindustrialized area, if indeed it does accrue, can be considered a
redistribution rather than a gain (Wetselaar and Hutton, 1963), and in any
case must be low. At Townsville (near the sea) and Woodstock (20 km
inland), depositions of 5.7 and 2.6 kg S ha-' annum-', respectively, have
been measured (Jones et af., 1975). Other records are available from
coastal stations further south in Queensland (Sedl, 197 l), where annual
deposition of 3.7 to 25.2 kg S ha-' was measured. The higher figures were
registered during periods of intense rainfall associated with cyclones.
In tropical regions with little industrial activity, SO4-S concentrations in
N. S. PASRICHA AND R. L. FOX
Sulfate S in Soils along Tmsects in Rwanda as Determined
by Repeated Phosphate Extraction"
Solyinyo ( I )
( M g-')
6 - I36
4 - I80
4 - 120
6 - 1 372b
Adapted from Vander Zaag ef a[. (1984).
bvalues > 200 not included in the average.
rainwater are usually less than 1 pg m1-I. In Rwanda, SO, was generally a
little more abundant in subsoil samples than in 0- to 15-cm samples (Table
I). Information on availability of subsoil sulfate is limited, but deep-rooted
crops are likely to benefit from subsoil sulfate. Sulfate levels determined in
these soils appear to be deficient or near deficient. For soils from volcanic
ash (Dystrandepts) 25 pg g-l was inadequate (Fox et al., 1965), but 810 pg g-l may be adequate for soils that do not adsorb sulfate in appreciable quantities (Kang and Osiname, 1976).
Sulfur deficiency is widespread in rice fields in Bangladesh (Hoque and
Hobbs, 1978; Hussain, 1990). About 44% of the cultivated land in Bangladesh is estimated to be S deficient. Crops grown on these soils respond to S
application. The major cause of S deficiency seems to be extremely low
redox conditions in wetland rice, although SO, is the last compound to
undergo reduction, after NO3-, Mn (IV), and Fe (111) compounds, when
reducing conditions set in. In soils with fine texture and appreciable decomposable organic matter, SO, reduction to H,S is likely.
Decreased incidental addition of S is a probable reason why S deficiency
is now appearing more frequently than formerly. Besides smaller accretions of S by rain, a drastic decrease in incidental S additions in fertilizers is
PLANT NUTRIENT SULFUR
Figure I . Trends in consumption of S-bearing N and P fertilizers in India during 19501990 (Pasricha and Aulakh, I99 I ; by permission of The Sulphur Institute, Washington, D.C.)
a probable cause for the appearance of S deficiency in crops in India
(Pasricha and Aulakh, 199I). In 1950, sulfur-containing fertilizers were
commonly used sources of N and P. At that time 28,900 tons of N as
ammonium sulfate and 4279 tons of P as single superphosphate were
applied. The consumption of fertilizer N and P has increased sharply since
then to 7,396,000 tons of N and 1,315,000 tons of P in 1990. However, use
of N and P fertilizers that contain S has relatively decreased, thus resulting
in a drastic decrease in the use of S (Fig. 1). An analysis of 1164 soil
samples collected from all over India indicated S deficiency to the extent of
4 1 % (Singh, 199I). In Thailand, Hoult et al. ( 1983) estimated that 50% of
the northeast plateau, 30-40% of the southeast coast, and 30-40% of the
northern highland area are S deficient and would respond to S fertilization.
111. FORMS OF SULFUR IN SOIL
Sulfur is continuously cycled between inorganic and organic forms.
Three broad fractions of organic S have been identified: (1) ester sulfate, (2)
C-bonded S (mainly amino acids), and (3) residual S (Tabatabai, 1982).
The nature of the compounds formed and their transformations are
N. S. PASRICHA A N D R. L. FOX
strongly influenced by biologically mediated processes, which in turn are
affected by environmental conditions. Perhaps 95% or more of the total S
in arable soils in temperate regions is organic S. This generalization does
not apply in the tropics.
Inorganic SO, vanes seasonally in soils. Castellano and Dick ( 1991)
observed that during rainy winter seasons, SO, levels ranged from 7 to
13 mg kg-' in gypsum-treated plots compared with 2 to 7 mg kg-' in
control plots. In the months from March to May, biomass S increased and
SO, level decreased (<6 mg SO4-S kg-' soil), indicating that S immobilization occurred in the spring.
Activity of aryl sulfatase (an enzyme that releases SO, from aromatic
SO, esters) increased significantly in cropped plots where plant activity was
greater (Tabatabai and Bremner, 1970; Castellano and Dick, 1991 ). Carbon-bonded s, which is mainly a measure of the amino acids cystine and
methionine, correlates with biomass S. Generally these compounds do not
accumulate in soil because they readily decompose (Fitzgerald, 1986).
Residual S is resistant to hydrolysis by strong acids or bases, yet it varies
with the season (Freney, 1986). However, this variation may be due to
measurement errors associated with other fractions, because residual S is
determined by the difference between total S and the sum of all other S
fractions. Assuming the data are reliable, seasonal variation follows a
pattern that appeared to be influenced by biological activity during 2 years.
With increasing moisture in the autumn, residual S increases, whereas in
the cool winters and springs, there is a general decrease in residual S.
Residual S is least in the dry summer.
Ester sulfate accumulation in the soil was associated with incorporation
of inorganic SO, by Saggar et al. (198 1). Ester sulfate constituted from 20
to 65% of the total S in a group of six Brazilian soils (Neptune et al., 1975).
Several short-term laboratory/greenhouse studies using 35S0, have found
that ester sulfate is more transitory than C-bonded S (McLaren ef al., 1985;
McLaren and Swift, 1977; Freney el al., 1975). These studies have shown
that, typically, 60-90% of 35S0, added to soil is quickly incorporated into
the ester sulfate fraction. A large proportion of S taken up by plants comes
from the ester sulfate pool and not the C-bonded S pool. There is a
tendency for the initial incorporation of applied S into ester sulfate. With
time and with further S cycling, a large portion is redistributed into Cbonded and residual S pools.
Many soils of the humid and subhumid temperate zones contain only
small quantities of inorganic S. In such soils, plants quickly utilize S