Tải bản đầy đủ - 0trang
V. Technologies for Increasing Plant-Available Water
B. A. STEWART AND C. A. ROBINSON
centuries, new technologies and strategies have been developed that increase crop
production in water-deficient areas. Some of these technologies and the principles
on which they are based are presented in the following sections.
One of the oldest, and most controversial, technologies for increasing plantavailable water is lengthening the fallow period. This is generally called summer
fallow, defined as a practice wherein no crop is grown and all plant growth is controlled by cultivation or chemicals during a season when a crop might normally be
grown. Proponents have emphasized the water-conserving, weed-controlling, and
crop yield-stabilizing virtues, whereas critics have emphasized the inefficiency in
soil water storage and the wind and water erosion and declining organic matter
problems associated with fallow. Fallow efficiency is defined as the percentage of
precipitation occurring during the fallow period that is stored in the soil profile at
the end of the fallow period. Historically, fallow efficiencies have been in the range
of 15-20%. The remainder of the precipitation is lost as runoff and evaporation
and, on some soils, as drainage below the root zone. Stewart et al. (1 994) summarized long-term data from different cropping systems at Bushland, Texas, to illustrate the effect that lengthening the fallow period has on increasing soil water
storage (Table I). The cropping systems were (i) continuous wheat, where winter
wheat is seeded each year in October and harvested in late June or early July;
(ii) wheat-sorghum-fallow (two crops in 3 years), where the land is fallowed for
approximately 11 months following wheat harvest and the land is then seeded to
grain sorghum in June and harvested in November and then fallowed for 1 1 months
until the following October when the land is again seeded to winter wheat; and
(iii) wheat-fallow (one crop in 2 years), where the land is fallowed for approximately 16 months from wheat harvest in July to October of the following year
when wheat is seeded again. The amount of precipitation lost during the rotation
as evaporation during the nongrowing (fallow) season ranges from 36% for the
continuous wheat system to 61% for the wheat-fallow system. In the wheatfallow system, where the fallow period is 15 or 16 months, only approximately
15% of the precipitation that occurs during the fallow period is stored in the soil
for later use by a growing crop. Even under the continuous wheat system, where
the fallow period is only 3 or 4 months, the fallow efficiency is less than 20%.
However, despite the very low efficiency of water storage during fallow periods,
the wheat-fallow system is practiced widely because of the importance of sustaining crop yields under dry farming conditions. As illustrated in Fig. 1, there is
no period during the year at Bushland, Texas, when average precipitation is as
much as 50% of the potential evapotranspiration. Therefore, without a substantial
amount of stored soil water present at time of seeding, the chance of a crop fail-
AGROECOSYSTEMS SUSTAINABLE IN SEMIARID REGIONS? 207
ure increases sharply. The annual precipitation in the area ranges from a low of approximately 50% of average to a high of approximately200% of average, and crop
yields range from 0 to approximately three times the average. Again, referring to
the data in Table I, soil water storage during a 4-month fallow period is only 37
mm compared to approximately 80 mm for an 11-month fallow period and almost
100 mm for a 15-month fallow period. The additional stored water leads to a significant increase in crop yield. Stewart and Steiner (1990) summarized long-term
grain sorghum studies at Bushland and found that there is a threshold of approxi-
Water Balance for Various Cropping Systems at Bushland, Texas"
One crop annually" (mm)
Soil water change
Evaporation (and runoff)
Two crops in 3 years' (mm)
Soil water change
One crop in 2 yearsd (mm)
Soil water change
Evaporation (and runoff)
"Adapted from 0. R. Jones (personal communication) and Johnson and Davis (1972).
bF~llowperiod between crops is 3 or4 months. Runoff was not measured but would be minimal under annual cropping.
'Fallow periods between crops are approximately 1 I months.
"Fallow period between crops is 15 or 16 months. Runoff was not measured but was a minor portion of the total.
B. A. STEWART AND C.A. ROBINSON
mately 1 10mm seasonal evapotranspirationrequired before any grain is produced,
and for every mm of additional evapotranspiration, 15.5 kg ha-' of grain is produced (Fig. 6). This fact, coupled with the water use data for grain sorghum shown
in Table I, suggests that grain sorghum yield would be 1.61 mg ha-' if only seasonal precipitation was available for evapotranspiration.The use of 72 mm stored
soil water would increase the yield by 1.12 rng ha-'. Musick et al. (1994) developed a similar grain yield-seasonal evapotranspiration relation for winter wheat
(Fig. 7) that indicates that about 200 mm of seasonal evapotranspirationis required
before any grain is produced and that 12.2 kg ha-' of grain is produced for each
mm above that threshold amount. The relations shown in Figs. 6 and 7 show the
great importance of having substantial amounts of stored soil water at the time of
seeding in areas where precipitation during the growing season is lacking.
The relation between grain yield and evapotranspirationwill differ among semiarid locations because of climatic differences. The water balance values for annual cropping of winter wheat at three semiarid locations are presented in Table 11.
The percentage of total rainfall that was used for seasonal evapotranspirationwas
similar for all three locations-approximately 65%. Seasonal evapotranspiration
is the combined loss of water from transpiration from the growing crop and evaporation from the soil surface during the period when the crop is growing. The fallow period is the time between harvesting the crop and seeding the subsequent
Y = 0.0155X - 1.97
SEASONAL EVAPOTRANSPIRATION - mm
Figure 6 Relation between yield of grain sorghum and seasonal evapotranspiration at Bushland,
TX (from Stewart and Steiner, 1990).
AGROECOSYSTEMS SUSTAINABLE IN SEMIAIUD REGIONS? 209
l RRl GAT ED
R 2 = 0.74
SEASONAL EVAPOTRANSPIRATION, mrn
Figure 7 Relation of wheat grain yield to seasonal evaportranspiration at Bushland, TX (from Musick et d.,1994).
crop. For the data presented in Table 11, evapotranspiration values were calculated by adding the growing season precipitationamounts to the change in the amount
of available water held in the soil at seeding time and at harvest time. In all locations, soil water was decreased significantly during the growing season and increased during the fallow period. However, the change was considerably less for
the Texas location. There was less precipitation at this location during the fallow
period, and the potential evapotranspiration was very high, resulting in only 37
mm of storage compared to storage at the other two locations of 80 mm or more.
The Texas location is the most arid of those presented in Table 11. Although total
precipitation was more for the Texas site than for the China site, the amount of actual evapotranspiration was only 26% of the potential evapotranspiration for the
Texas location compared to 56% for the China site. The China location had a much
Water Balance Values for Annual Cropping of Wheat at Three Semiarid Locations"
Soil water change
Evaporation and runoff (mm)
Water use efficiency kg m-3
New South Wales, Australia
"Adapted from 0. R.Jones, unpublished data, Lun et al. (1992). and Cornish and Pratley (1991).
AGROECOSYSTEMS SUSTAINABLE IN SEMIARID REGIONS? 2 1 I
higher yield and a water use efficiency of 0.47 kg mP3 compared to 0.33 kg mP3
for the Texas location. The yield and water use efficiency values were low for both
sites, but the Texas site values were extremely low. Water use efficiency values for
wheat grown in humid regions or under irrigation often exceed 1.25 kg m-3and values as high as 1.9 kg mP3 are reported in the literature (Musick and Porter, 1990).
The data presented in Figs. 6 and 7 and Tables I and I1 discussed previously
clearly show the importance of stored soil water and explain why fallowing is so
widely practiced even though the storage efficiency of precipitation occurring during the fallow period is very low. Perhaps the biggest concern about summer fallowing is its effect on soil degradation. Until herbicides became available in recent decades, tillage was the only means of controlling vegetative growth during
the fallow period. Consequently, it was not uncommon for a field to be tilled 8-10
times during the fallow period. Intensive and frequent tillage buries most of the
crop residues and hastens the decomposition of crop residues and soil organic matter. Cultivation increases biological activities, often as a result of better aeration.
Cultivation also exposes fresh topsoil to rapid drying and after each drying a burst
of biological activity occurs for a few days following rewetting (Allison, 1973).
This is because the drying process releases organic compounds, probably from
the breakdown of soil aggregates that are bound together by humic substances.
Considerable organic nitrogen is mineralized as ammonia and later oxidized in
large part to nitrates. Other nutrients are also made available from the decomposition of organic matter. This is particularly true for phosphorus because much of
the phosphorus in soils is present in organic forms. The nutrients released as a result of tillage are readily available to growing plants and increased yields are generally obtained. Therefore, in addition to increasing water storage, summer fallowing increases available soil nutrients. However, unless the organic matter
supply is replenished by plant residues or manures, the system is not sustainable.
This is the situation for many soils of the world located in arid and semiarid regions and increased attention to the problem is critical. It is also the underlying
principle that resulted in the infamous “Dust Bowl” that occurred in the U.S. Great
Plains during the drought years of the 1930s and considered by many to be the
worst ecological disaster ever exacerbated by man.
The U.S. Great Plains region was largely settled in the early 1900s by farmers
who migrated from the humid areas of the eastern United States and brought with
them their clean-tillage tools and experiences. These worked well for the first few
years after cropping began because the native soil organic matter content was high
and the precipitation during the period of the “big plowout” was above average.
However, when annual precipitation decreased to average and below, the annual
net loss of soil organic matter accelerated and led to increased vulnerability to wind
The moldboard plow, and many other intensive tillage implements, was developed in Europe, where soil organic matter content of soils is high and the organic
matter level can be maintained at a high level because of relatively high precipi-
B. A. STEWART AND C. A. ROBINSON
tation amounts that produce large amounts of biomass and cool temperatures that
slow the rate of decomposition. In arid and semiarid regions, high temperatures
accelerate the rate of decomposition and the lack of precipitation severely limits
biomass production so that organic matter loss can be rapid and severe.
In certain situations, salinity problems can be very significant in semiarid regions. Summer fallow, particularly during years of above-average precipitation,
can infiltrate more water than can be stored in the profile. This can result in substantial amounts of water moving through the profile removing nutrients and, if
salts are present, they will be leached and cause saline seeps in certain situations.
A saline seep occurs when water in excess of that required by plants percolates below the root zone and, upon encountering some type of barrier or restricting layer, movers laterally downhill and emerges in a seepage area, having picked up dissolved solids in transit. This has been a significant problem in parts of the northern
Great Plains of the United States where spring wheat is the dominant cropping system (Halvorson, 1990). The length of the fallow period in a spring wheat-fallow
cropping system is approximately 20 months during each 2-year cycle. Saline seep
problems are also widespread in some semiarid regions of Australia (Sharma and
Summer fallowing has also been used extensively in Australia and China for increasing soil water storage. Cornish and Pratley (1991) stated that fallows have
had a long and often sorry history in Australia. Summer fallowing in the United
States was so successful when it was first introduced that the practice was imported
by Australian farmers in the early 1900s. The primary practices involved deep
plowing and frequent harrowing to produce a dust mulch. The plowing was
thought to increase the water-holding capacity of the soil, whereas the dust mulch
supposedly prevented water rising to the soil surface by capillary action and evaporating. Subsequent research showed that the major loss of water from soils was
through transpiration by weeds and that the benefits of dust mulching were largely due to weed control. These technologies were used to extend the limits of wheat
growing into the marginal 250- to 400-mm rainfall zone of the South Australian,
Victorian, New South Wales, and Western Australian Mallee. Long fallow periods
(15 months) were used and the frequent cultivation of these light-textured soils resulted in soil structural breakdown, fertility decline, and, ultimately, catastrophic
erosion. By the mid- 1930s wheat farming in the Mallee was not a paying proposition and was replaced by sheep production. By the early 1950s wheat production
occurred in rotation with pastures. Support for a 15-month fallow also waned as
results showed that fallows of 8-11 months produced yields equal to those of a
longer fallow. The cost in terms of loss of winter grazing and soil erosion was also
significant. It was further shown that much of the benefit of fallowing was from
nitrogen mineralization and not water conservation as originally supposed. Cornish and Pratley (1991) stated that two lessons emerged from the fallowing experiences in Australia. First, widely accepted farming practices can be based on poor
foundations: Long fallowing very often gave no appreciable increase in water stor-
AGROECOSYSTEMS SUSTAINABLE IN SEMIARID REGIONS?
age over short fallows, whereas an unrecognized response to fallowing was nitrogen mineralization. It is therefore advisable to consider the outcome of farming
practices in terms of specific effects on soils and plants and not yield alone. The
second lesson is that practices can have effects well beyond those intended, emphasizing the need to understand how the specific effects interact with one another and the environment.
Shengxiu and Ling (1992) summarized some of the results from the Loess
Plateau region of China and concluded that fallowing was a good practice for the
drylands. Summer fallowing was usually combined with summer deep plowing for
controlling weeds, keeping the soil loose, and increasing soil infiltration. As a result, soil water was increased. In addition, available plant nutrients, especially nitrates, accumulated in the profile and stimulated growth of the subsequent crop.
They did not stress the negative effects such as organic matter decline and deteriorating soil structure.This may be because Chinese farmers have historically used
organic wastes on their fields and this may offset the otherwise negative effects.
The steppe area of northern Kazakhstan is another region where fallow has been
widely practiced. However, Souleimenov (1992) concluded that fallow in this region of about 350 mm annual precipitation was not justified. Research showed that
the available water storage prior to seeding wheat was only slightly higher for the
fallow fields than those for wheat after fallow or for continuous wheat. Fallow was
adopted in this region in 1966 based largely on some selected data of the state
farms for extremely dry years (1962, 1963, and 1965). The decision was also influenced by data and experiences from the Canadian prairies. Souleimenov (1 992)
recommended that most fallow be discontinued with the more marginal lands being returned to grass and the better lands cropped annually. He also pointed out the
benefits that such a system would have on the environment.Weed infestation, wind
erosion, and other soil degradation processes have been widely experienced in the
region where fallow systems were the dominant practice.
The previous discussion clearly indicates that although summer fallow does increase soil water storage and greatly stabilizes grain yields and reduces risk, there
are many disadvantages. The biggest disadvantage, particularly when tillage is
used during the fallow period, is the decline in soil organic matter. A fallow system accelerates carbon oxidation, increases carbon removal by erosion, and adds
less carbon back to the soil as crop residues. Fallow systems are best used on loam
and clay loam soils; they are not effective on sandy soils because of low soil
water storage capacities.
The Dust Bowl of the 1930s in the U.S. Great Plains led to the development of
stubble mulching. Stubble mulching uses V-shaped sweeps or blades that are
pulled flat approximately 10 cm beneath the soil surface. This operation cuts plant
B. A. STEWART AND C. A. ROBINSON
roots and kills the weeds but does not invert the soil. Therefore, most of the crop
residue is left on the surface where it can serve as a mulch to prevent wind and
water erosion and slow evaporation losses. Only approximately 15-20% of the
residue is buried by a sweep tillage operation, so there is substantial residue remaining on the surface even after three or four operations, which are commonly
done between the time a crop is harvested and subsequent crop is seeded. At least
this many tillage operations are performed when summer fallow is used and the
length between crops is 11 months or longer. A rodweeder, a square rod approximately 25-mm thick that turns approximately 5-10 cm beneath the surface as it is
pulled, is another tool that is sometimes used to kill weeds without intensively tilling the soil. A rodweeder operation can sometimes bury less than 10% of the crop
residue present on the soil surface. In recent years, herbicides have been used to
replace some tillage operations in reduced tillage systems and completely replace
tillage when no-tillage systems are used. The sweep plow that was developed for
stubble mulching is often referred to as “the plow that saved the plains” because
it played a key role in controlling wind erosion that was so devastating during the
Dust Bowl era. The Dust Bowl was one of the worst ecological disasters ever
Although stubble mulching was developed to address the wind erosion problem, it soon became evident that mulches had beneficial effects on soil water storage. The increase in soil water storage generally is attributed to increased infiltration and reduced evaporation. Johnson et al. (1974) reported on studies from
Bushland, Texas, conducted from 1942 to 1969 that compared stubble mulch
tillage to one-way disk tillage (Table 111). For continuous wheat, which had only
3 or 4 months between crops, the plant-available soil water stored in 1.8 m profile
at seeding time averaged 103 mm for the stubble mulch plots compared to 91 mm
for the one-way disked plots. For the wheat-fallow system, which had 15 or 16
months between wheat crops, the plant-available soil water values at seeding time
were 154 and 128 mm for the stubble mulch and one-way disk treatments, respectively. The yields were also increased approximately 100 kg ha-’ in both
treatments as a result of the increased soil water storage.
Unger (1978), also working at Bushland, Texas, on a clay loam soil found very
significant increases in soil water storage when crop residues were maintained on
the soil surface. The residues enhanced water infiltration and suppressed evaporation, thus providing more water for the subsequent crop (Table IV). This study was
conducted on small plots and the straw was added from an outside source rather
than being the residue from a crop produced on the plot. The straw rates ranged
from 0 to 12 Mg ha-’. Under semiarid conditions, the amount of straw remaining
after harvesting wheat would rarely exceed 5 or 6 Mg ha-’ and in many cases is
only 1 or 2 Mg ha-’. Unger (1978) showed that even these small amounts, however, significantly increased water storage during fallow, grain yield, and water use
efficiency (Table IV).
AGROECOSYSTEMS SUSTAINABLE IN SEMIARID REGIONS? 2 15
Effect of Various Cropping Systems on Soil Organic Matter (SOM), Soil Water Content at
Seeding (PAW), and Yield of Winter Wheat at Bushland, Texasa
One-way disk tilled
Subtilled with sweeps
One-way disk tilled
Subtilled with sweeps
“Adapted from Johnson et al. (1994).
”Nitrate-N in 180 cm soil profile at end of experiment.
“Average plant-available water in 180 cm soil profile at seeding time.
%inter wheat seeded annually approximately October I and harvested approximately July I the
‘Winter wheat seeded approximately October 1 every second year; approximately 15 months fallow between crops and yields shown must be divided by 2 to indicate annual land production.
’Tillage was delayed for approximately 10 months following wheat harvest; weeds and volunteer
wheat were allowed to grow during the 10-month period.
Straw Mulch Effects on Soils Water Storage during an 11-Month Fallow, Water Storage
Efficiency, and Dryland Grain Sorghum Yield at Bushland, Texasa
“Adapted from Unger (1978).
”Water storage determined to 1.8-mdepth. Precipitation averaged 318 mm.
‘Water use efficiency (WUE) based on grain produced, growing season precipitation. and soil
“Column values followed by the same letter are not significantly different at the 5% level (Duncan’s
multiple range test).
B.A. STEWART AND C. A. ROBINSON
Cornish and Pratley (1991), working on clay soils in Australia, found that plant
residues on the soil surface increased fallow efficiencies from 21 to 29%. They attributed the increased soil water primarily to increased infiltration. When plant
residues were not on the soil surface, there was a major increase in runoff due to
crusting caused by raindrop action. They reported that approximately 4 Mg ha-'
of wheat straw was sufficient to achieve the maximum benefit on infiltration. They
further reported that the average grain yield of wheat in the region was 1.5 Mg
ha-'. The straw to grain ratio for wheat is generally from 1.5 to 2, so the maximum benefit was probably not achieved in most years.
Shengxiu and Ling (1992) summarized studies from China and results are similar to those reported previously. They concluded that straw mulch increased infiltration and reduced water loss by evaporation, thereby increasing water storage
in both summer and winter. One of the studies they summarized was conducted by
Siming et al. (1988) and some of their data are presented in Table V. Soil water
storage was increased at all soil depths and with all levels of mulch. The researchers stated that in addition to increasing soil water storage, the mulch decreased bulk density and increased the number of earthworms and soil organic
matter. Bulk density in the top 0-10 cm soil layer was 1.36 Mg mP3 when no mulch
was present compared with 1.29 Mg mP3 with 4.5 Mg ha-' mulch and 1.23 Mg
m-3 with 6.0 Mg- I ha-' mulch. The number of earthworms per square meter in
the top 15 cm was 2 when no mulch was present but 12,32, and 34 when straw
mulch was present at rates of 3.0,4.5, and 6.0 Mg ha-', respectively. Organic matter content in the top layer was 1.61% with no straw mulch and 1.67 and 1.76%
with 4.5 and 6.0 Mg ha-', respectively. Soil temperature was also cooler in the
summer and warmer in the winter when there was straw mulch on the surface.
Although crop residues on the soil surface clearly reduce runoff and evapora-
Amount of Water Stored in Different Soil Layers with Different
Amounts of Straw Mulch"
Mulch (kg ha-')
Amount of water stored (mm)
Note. Two years' average values in two locations.
"Adapted from Siming er al. (1988).