Tải bản đầy đủ - 0trang
II. Comparison of Environmental Conditions in Tilled and Untilled Soils
depends on the amount of mulch present, the activity of soil animals and
the prevailing weather conditions. Where such a top layer of biologically
stabilized crumb structure has been established, slaking of silt material and,
consequently, formation of a dense crust are rarely observed on zero-tilled
The structure and size of deeper layers depend mainly on their soil texture and on the texture dependent reaction to changes in soil moisture
and temperature. In 'hit soils, zero-tillage induces a platy structure (Bulfin,
1967). This type of frost structure is unstable in tilled soils owing to excessive water in the top layers during the thawing process; however, it normally remains visible throughout the year in zero-tilled silty loam soils.
A polyhedric structure is typical for soils with a high clay content, low
capillary water conductivity and distinct swelling and shrinking properties.
Since swelling of the clay after remoistening will close every cleavage again
if the soil has not been previously mechanically loosened, tilth induced
by frost or drought may be a transient phenomenon in zero-tilled clay soils
B. SOIL FLORA
Changes in soil flora and fauna can be expected when zero-tillage practices are introduced. Suitable information is lacking, especially with regard
to reactions of microorganisms to zero-tillage effects per se. Indirect evidence that zero-tillage changes microbial activity is derived from tests of
cellulose decomposition under field conditions. On zero-tilled soils, higher
decomposition rates than on plowed soils were observed by Herzog et al.
(1969), whereas Bender and Adamczewski (1970) found the reverse.
However, these results reflect more the prevailing soil conditions than possible changes in microbial populations.
In two field experiments with continuous wheat in England (Corbett
and Webb, 1970), the total number of nematodes was sometimes larger
with zero-tillage, while migratory parasitic nematodes were usually less numerous on untilled soils. Small-sized species of nematodes seemed to be
favored on naturally compacted soils; inadequate observations do not allow
Although earthworms form the most conspicious group of soil-inhabiting
animals, little information is available about the changes in weight and
number of earthworms upon introduction of a no-tillage system. For sampling, all investigators used vermifuges, which do not allow complete recovery of existing worms; therefore, only relative values can be reported.
In West Germany, Schwerdtle (1969) found on the average a 12-fold increase in number and a 16-fold increase in weight of earthworms collected
K. BAEUMER AND W. A. P. BAKERMANS
on zero-tilled plots after three years’ cropping with corn. In England, Wilkinson (1967) reported less spectacular increases, e.g., on cereal stubble
fields a mere 1.6-fold weight increase. On former leys and permanent pastures, he did not observe any difference in the weight of earthworm populations resulting from tillage treatments. On three sites of former permanent
pastures in the Netherlands, we found about half as many earthworms on
tilled as compared to untilled plots after seven years.
More than in any changes in abundance of earthworm populations, the
agronomist is interested in their activities which alter the ecological conditions of naturally compacted soils. On loamy sand in West Germany, Graff
(1969) measured the rate of casting from September to May over a period
of three years. On untilled, mulched plots, between 2 and 4.5 kg dry matter
per m2, which is within the range of values encountered normally on old
pastures (Evans and Guild, 1948), were deposited on the soil surface.
Graff (1969) observed 20- to 25-fold increases in rate of casting on untilled plots as compared to turnplowed barley stubble. Normally, most
earthworms deposit their castings in the soil, not on top of it. In compacted
soils, however, most castings are deposited on the soil surface. Still, even
on zero-tilled soils, considerable soil mixing is to be expected.
Earthworm tunnels which open to the soil surface may influence the
rate of water infiltration. Since small tunnels are difficult to distinguish
from soil cracks, the number of earthworm “mittens” may serve as a first
approximation. On untilled cereal stubble fields, we found an average of
55 mittens per m2 soil surface. One centimeter below the soil surface, an
average of 68 tunnels (diameter 2-10 mm) was observed on no-tilled stubble as compared to 15 on plowed stubble (W. Ehlers, personal communication, 1972).
Moles are predators of earthworms and increase in number when fields
are left undisturbed for a prolonged period. In Switzerland, Vez (1969)
counted 10 to 12 molehills per are on zero-tilled plots as compared to
1 to 2 on conventionally tilled plots. Similar differences can be observed
in burrows of voles and mice; these have not yet been quantified.
With zero-tillage, plant residues remain on the soil surface. This is essential when soil erosion limits successful farming. Where erosion presents
no problem, a mulch cover may be desirable to create a favorable soil
tilth. The highest permissible level of mulch is determined by conditions
governing the effective performance of drilling and weed control operations. In the Netherlands, 3000 to 4000 kg straw (dry matter) per hectare
can remain on the ground only when it is chopped up into small pieces
and evenly distributed. For dryland farming, an amount of 3500 kg straw
per hectare at harvest is considered to provide adequate soil protection
from erosion without presenting problems with seeding, weed control and
Brown and Dickey (1970) determined losses of wheat straw buried in
the soil, placed on the soil surface and exposed above the soil surface to
simulate a standing stubble. At two locations in Montana, they found that
the rate of dissipation increased with greater contact between soil and plant
material and decreased when rising amounts of straw were applied. Lower
mean annual temperature, though in combination with higher precipitation,
retarded straw decompositibn.
During the first 3 months of exposure, the weight of above-surface and
on-surface straw increased by as much as 1 3 % due to an accumulation
of soil particles inside the hollow straws. Some soil probably is moved by
wind or raindrop splash and, near the ground, by the activities of soil-inhabiting animals.
The experiment of Brown and Dickey was begun in May. After 18
months’ exposure, only 22 to 40% losses were measured on and above
soil as compared to 93 to 98% buried in the soil. In the more humid climate of Germany, higher rates of wheat straw decomposition were observed, e.g., 40% during the period September through July (K. W.
Becker, personal communication, 1972). The weight of adhering soil particles equaled the amount of straw remaining after 1 1 months’ exposures.
These figures indicate that a slow surface accumulation of straw residue
can occur if the above-mentioned process of incorporation into the soil
is not effective.
Plant residues with higher N content decompose more rapidly. Cornstalks applied to the surface of a cornfield in Iowa in May lost 50% of
the initial weight after 20 weeks’ exposure (Parker, 1962). Sugar beet tops
left on the ground in Western Europe are completely disintegrated by July.
Hence, leafy and succulent plant material presents no problems in mulch
In Fig. 1, some results of trials on former grassland and recently plowed
soils concerning the distribution of organic matter in soil are summarized.
No differences in organic matter concentration were found in regularly
plowed soil layers. In undisturbed soils, the concentration was highest near
the soil surface and declined steadily to subsoil values below those on conventionally tilled soils. The gradient of organic matter content was more
pronounced on former grassland soils, where zero-tillage presumably preserved the original distribution of organic matter.
No thorough investigation has yet been published as to whether zerotillage results in an accumulation of total soil organic matter. Available
K. B A E U M E R A N D W. A. P. B A K E R M A N S
organic matter ( % . d r y s o i l )
arable l a n d
N ( '/. )
8 9 1011
FIG.1. Average distribution of organic matter and N in tilled (-)
(- - - ) soil. (From Bakermans and De Wit, 1970, grassland; Buhtz et al., 1970,
data from long-term trials (Moschler et al., 1972; Buhtz el al., 1970) suggest that zero-tillage increases the total organic matter of the soil.
Whether the observed differences in accumulated organic matter are
caused by restricted decomposition andlor higher production of organic
matter on zero-tilled soils is not yet known. As compared to a tilled
chernozem soil in East Germany, concentrations of CO, in the atmosphere
near the soil surface were lower on untilled soil (Buhtz, 1972). These
observations suggest a reduced rate of mineralization in naturally compacted soils.
OF SOIL STRUCTURE
With zero-tillage, soils are loosened only locally and superficially; yet
they have to bear the normal load of traffic in the field. Hence natural
consolidation and mechanical compaction will cause a denser packing of
zero-tilled topsoils. The average decrease in total porosity was found to
vary between 0 and 6% (v/v) (Czeratzki and Ruhm, 1971; Herzog and
Bosse, 1969; Vez and Vullioud, 1971a,b). A few exceptions were noted
on heavy river clay rich in organic matter (Van Ouwerkerk and Boone,
1970), on two sites with silt loam-chernozem soil (Buhtz et al., 1970)
and on silty clay (Bachthaler, 1971) where values of total porosity were
lower on tilled than on zero-tilled plots, probably as a result of compaction
caused by tillage operations.
In general, the differences in total porosity were greatest in the soil layer
which is loosened by plowing, but not compacted by seedbed preparation
and cultivation ( 10-1 8 cm) . In deeper soil layers, the differences .tended
to diminish. Near the soil surface, they varied with the effects of tillage
operations, weather, and biological activities.
Mean values of total pore space average over sampling dates, crops,
and locations, eliminate extreme values, which may be decisive for plant
growth and farming operations in critical situations. The lowest sampling
means of the porosity data published were found with values near 35 and
38% (v/v) on untilled plots on sandy soil (22-27 cm) and clay soil
(15-20 cm), respectively (Czeratzki and Ruhm, 1971). The very high
density of the soil layer 22-27 cm on sand merits special attention, as it
is probably induced by mechanical compaction and perhaps by downward
displacement of finer soil particles, which by turnplowing are redistributed
to upper soil layers. This could be proved by particle size and pore size
distribution analysis, but no information is yet available.
The observed minimum values did not mark the final stage of soil density on zero-tilled soils. At subsequent dates, porosity increased again, especially on stable soils with medium to high clay content. The above-stated
lower levels of pore space were reached within two to three years of zerotillage, after which time seasonal fluctuations of total porosity tended to
be smaller as compared with conventionally tilled soils (Van Ouwerkerk
and Boone, 1970).
Similar results are shown in Fig. 2, which contains a time series of
porosity measurements in the top 2-6 cm layer of arable silt loam soil derived from loess (Ehlers, 1973). Zero-tillage resulted in a smaller total
porosity but also in reduced variability of- the sampling means; consetotal
0.2 - 3.0p m
un ti lied
FIG. 2. Changes of total pore space and pore size distribution with time at
a depth of 2-6 cm on tilled and untilled silt loam soil. (From Ehlers, 1973.)
K. BAEUMER AND W. A. P. BAKERMANS
quently, homogeneity in time increased in naturally compacted soils. The
remaining fluctuations of total porosity presumably result from the combined effect of seasonal changes in climate and soil cover on the activities
of soil flora and fauna.
The example in Fig. 2 shows further that changes in total porosity were
accompanied by concomitant changes in other pore size fractions. It can
be concluded, therefore, that mechanical loosening effects mainly the fraction of large pores and that dense parts of the soil remain more or less unchanged. As compared to the fraction of large pores, the other pore size
fractions fluctuated to a smaller extent.
p o r e space
FIG. 3. Changes of total pore space and pore size distribution with depth on
tilled (right panel) and untilled (left panel) silt loam soil. (From W. Ehlers,
personal communication, 1972.)
Figure 3 shows the vertical pore size distribution of a silt loam soil (W.
Ehlers, personal communication, 1972). Although in this case the total
porosity and the fraction of large pores did not differ much between zerotilled and conventionally tilled soils, the pattern of porosity reveals an important difference: on undisturbed soil, the relative space occupied by each
pore size fraction varied less than on the ploughed soil, where the layers
at 0 to 15 cm and 25 to 30 cm were compacted as compared to the layer
at 15 to 20 cm. The compaction at 25-30 cm is presumably caused by
pressure and smearing actions during plowing. It resulted not only in an
absolute reduction in large and medium pores-based on volume as well
also in a relative increase in small and very small
as on weight-but
pores-based on volume only, as discussed by Ehlers (1973). Untilled
soil, though generally denser, may also exhibit more structural homogeneity in space as compared to conventionally tilled soils.
A relatively higher amount of smaller pores, but greater homogeneity
in time as well as in space are thus the dominant changes in porosity when
a soil remains untilled for a long period. Another feature may be connected
with the continuity of pores. Since earthworm tunnels can be regarded as
primarily continuous pores, an estimate of the relative pore space occupied
by them may serve as a first approximation. Figure 4 shows that the space
occupied by large pores with presumably uninterrupted connections to the
atmosphere is more than doubled near the soil surface and in the top 20
cm of zero-tilled soil as compared to plowed soil (W. Ehlers, personal
communication, 1972) .
p o r e space (% v/v)
FIG. 4. Pore space occupied by
rainworm tunnels on tilled (O---O)
and untilled (0-0) silt loam soil
(From W. Ehlers, personal communications, 1972.)
Other composite parameters of soil structure are resistance to penetration and shear stress, which are highly dependent on texture, soil moisture,
and porosity. In general, larger resistance to a cone-shaped probe forced
into the soil was observed on zero-tilled soil (Buhtz et al., 1970). On sand
soil, J. M. Houben (personal communication, 1972) found no rooting
when penetrometer resistance exceeded 40 kg/cm2. In one case, we observed that continuous zero-tillage on sandy soil produced a comparable
compaction in layers between 5 and 30 cm.
Soil aeration depends on porosity and water content. Hence when a soil
is water saturated to field capacity (soil moisture tension: 0.1 bar A p F
2), the extent of the remaining pore space filled with air (air capacity)
may be critical for the maintenance of soil aeration. A minimum volume
of 10% is thought to be necessary for adequate gas exchange between
the soil air and the free atmosphere.
Though zero-tillage generally causes a decrease of large, mostly air filled
pores (diameter > 30 pm) and thus reduced aeration, air capacity at p F
K. BAEUMER AND W. A. P. BAKERMANS
2 was observed only on medium- to heavy-textured soil to be below 10%
(v/v) (Van Ouwerkerk and Boone, 1970; Czeratzki and Ruhm, 1971;
W. Ehlers, personal communication, 1972). Impeded aeration, if caused
by zero-tillage, may provide a serious objection to the application of this
system on heavy soil in humid regions. With regard to this point, an evaluation of large, continuous pores, such as earthworm tunnels, would be of
The observed relative increase in the amount of medium to small pores
caused by zero-tillage has consequences for the water-holding capacity of
the soil. Plowing up grassland results in the redistribution of organic matter
( Fig. 1 ) ; zero-tilled sod retains its original accumulation of organic matter
near the soil surface. Water-holding capacity is related to organic matter
content, especially on sandy soils; this was confirmed by Van Ouwerkerk
and Boone (1970), who found that water content at p F 2 changed more
in conjunction with organic matter content than with soil porosity. Hence,
in the top 6 cm of the zero-tilled soil, a higher water content at p F 2
was found than in the plowed soil, whereas the reverse was true in the
layer at 11-1 6 cm.
Thus, beginning with a permanent pasture, changes in soil behavior
caused by different tillage systems cannot be ascribed solely to differences
in porosity. On arable land, the situation is less complicated since waterholding capacity generally increases with increasing pore space of an
equivalent diameter <30 pm. On zero-tilled soils, therefore, a relatively
larger part of the pore space ( % v/v) was found to be water filled, though
this is of no consequence for the amount of available water, i.e., the difference in moisture content in % (w/w) at p F 4.2 and p F 2.0 (Van
Ouwerkerk and Boone, 1970; W. Ehlers, personal communication, 1972).
Information about the energy associated with soil water is essential for
understanding its movement in soil and its availability for plant growth.
W. Ehlers (personal communications, 1972) investigated changes in gravimetric water content and matric potential in time and down the profile
of a silt loam derived from loess. Differences between tilled and untilled
soils are shown for four situations (Fig. 5, a-d). These are characterized
by beginning and advanced stages of either depletion (4-8 and 6-7, respectively) or recharge of soil moisture (6-10 and 6-21, respectively).
Most remarkable is the observation that soil water tensions were effected
by the tillage system down to a soil depth of 220 cm. This demonstrates
the consequences of changing the structure of one soil layer only for the
moisture regime of a whole profile. Differences in soil water content between tilled and untilled soils were relatively small and inconsistent compared to differences in soil water tension and hydraulic potential. Zerotilled soil with a similar water content generally had a lower soil water
(cm water column)
(cm water column)
9 6 - 21 -1971
FIG.5 . Matric potential, hydraulic potential, and water content of tilled (O---O)
and untilled (0-0)
silt loam soil during phases of depletion (a, b ) and recharge of soil moisture (c, d ) . Arrows indicate situations where gradients of hydraulic potential are zero. (From W. Ehlers, personal communication, 1972.)
tension, which indicates a smaller resistance to water uptake by plant roots
and a higher conductivity of soil water.
The largest differences between tilled and untilled soils were observed
during the rewetting phase (Fig. 5 , c, d ) . The rain water infiltrated rapidly
into the plowed layer, but slowly into the subsoil of the tilled plots. Thus,
soil water tensions were reduced to near zero in the upper soil layer, but
continued almost unchanged at greater depth. Yet, in the untilled soil, the
soil water tensions remained at a higher level near the soil surface and
decreased rapidly in the subsoil.
K. BAEUMER AND W. A. P. BAKERMANS
This behavior indicates a lower resistance of the zero-tilled soil to infiltration, which may be typical for light to medium loam soils, where earthworms or other soil animals construct a continuous set of large pores connecting the soil surface with the subsoil. Decaying roots which remain
undisturbed in place could provide ready avenues for water infiltration into
the soil profile as well (Barley, 1954).
Mulch physically absorbs raindrop impact energy. Thus, slaking and
sealing of the soil surface is prevented or at least retarded. Therefore,
zero-tillage generally reduces surface runoff. On silt loams with an 8-10%
slope planted with row crops, reduction ranged from one-half to one-sixth
of the amount observed on clean tilled land (Harrold et al., 1967; Shanholtz and Lillard, 1969; Jones et al., 1969).
On a silt loam in Ohio, Triplett et al. (1968) found an increase of both
the infiltration rate and total infiltration with increasing soil cover by cornstalks, the zero-tillage normal residue treatment resulting in higher values
than the conventional tillage treatment. Partition of the mulch effect due
to physical protection and structural stability of the soil showed that in
this case the accumulation of soil stability was more effective than physical
protection of the soil surface.
The purely protective effect of residue cover may influence the rate of
soil water evaporation too. During the initial constant rate of evaporation,
when rates depend solely on the saturated hydraulic conductivity of the
soil and the evaporative demand of the atmosphere, Bond and Willis
( 1969) observed decreased evaporation with increasing residue rates. During the following stage of falling rates of evaporation, when the soil surface
also dried underneath a mulch cover, no differences in evaporation rate
between bare and mulched soil were found.
Most of the evaporation losses in row cropped soil should occur before
the closing plant canopy reduces the incident radiation and thereby evaporation at the soil surface. Hence differential gains in soil moisture content
by means of zero-tillage as induced by mulch protection are to be expected
mainly during the early growth stages of row crops, when evaporation rates
of the more or less saturated soil are high.
Generally, zero-tillage resulted in higher mean volumetric moisture content in the upper 30-60 cm soil layer than conventional tillage on soils
situated in the subhumid regions of North America on gently sloped silt
loams planted with row crops (Harrold et al., 1967; Amemiya, 1968;
Blevins et al., 1971; Jones et al., 1969). The greatest differences in total
available water always occurred early in the growing season (Shanholtz
and Lillard, 1969; Van Doren and Triplett, 1969).
Zero-tillage methods could be expected to increase water conservation
in dry-farming regions. As compared to conventional stubble mulch fallow,
complete chemical fallow resulted in lower moisture conservation at one
location in the semiarid Great Plains (Black and Power, 1965), whereas
water storage gains were observed at two other locations (Smika and
Wicks, 1968; Army et al., 1961). These contradictory results could perhaps be explained in part by the differential reduction of evaporation by
a mulch cover during constant and falling rate drying phases. After prolonged falling rate drying, the cumulative evaporation from a mulched or
bare soil was nearly equalized (Bond and Willis, 1969).
Greater soil water storage may be attained with chemical fallow only
under the following conditions: first, when cumulative evaporation during
drying intervals is less with an undisturbed upright standing stubble than
with a residue cover knocked down by subtillage; second, when more frequent rains are prevalent, since mulches are of little value for water conservation during extended dry periods.
Soil temperature depends on the thermal conductivity and volumetric
heat capacity of a soil, and on the amount of heat that enters or leaves
the soil surface. Hence the amount of soil cover and the water and air
content of the various soil layers are decisive factors for the temperature
regime of soil.
Van Duin (1956), who used the results of a theoretical investigation,
showed that loosening the upper soil layer by mechanical means should
increase the diurnal temperature variation near the surface of a clean-tilled
soil, but decrease the amplitude in the deeper, undisturbed layers. Thus,
the loosened zone acts as an insulating layer. During phases of rising soil
temperature, a tilled soil should be warmer near the surface but cooler
in the subsoil than in undisturbed soil. The reverse is true during periods
of falling temperature.
Since the effect of a mulch cover is similar to that of a loosened soil
layer, differences in soil temperature between conventionally tilled and
zero-tilled soil will become larger with increasing amounts of cover.
During the growing season, untilled or merely mulched soil were observed to be cooler than clean tilled soil (Van Wijk et al., 1959; Parker
and Larson, 1962; Shanholtz and Lillard, 1969). At a depth of 10 cm,
the average difference in maximum soil temperature ranged from l o to
3 O C , which was greater than differences in minimum temperature, which
were less than 1OC.
During cool periods in May, Moody et al. (1963) observed higher soil
temperatures on mulched soil, confirming the postulated reversal of differences during periods of falling temperatures (Van Duin, 1956). Similar