soil& Tillage Research

Soil & Tillage Research 41 (1997) 25-42

Soil physical properties, water depletion and crop development under traditional and conservation

tillage in southern Spain

F. Moreno aj * , F. Pelegrh b, J.E. Ferniindez a, J.M. Murillo a a lnstituto de Recursos Nuturales y Agrobiologh de Sevillu fCSIC), P.O. Box 1052,41080 Seville, Spain

b Escuela Universitaria de lngenieriu Tknica Agricolu, University oj’seville, Seville, Spain

Accepted 3 September 1996

Abstract

Tillage methods affect soil physical properties and, thus, have a direct influence on the replenishment and depletion of soil water storage and crop performance. This study was conducted to determine the effects of traditional and conservation tillage on soil physical properties, soil water replenishment and depletion, and crop development and yield under southern Spanish conditions. The experiments were carried out from 1992 to 1995 in a sandy clay loam soil (Xerofluvent). The traditional tillage (TT) method consisted mainly of the use of mouldboard ploughing, and the conservation tillage (CT) was characterized by not using mouldboard plough- ing, by reduction of the number of tillage operations and leaving the crop residues on the surface as mulch. In both tillage treatments a wheat (Triticum aestivum, L.)-sunflower (Helianthus annuus, L.) crop rotation was established. In each treatment, systematic measurements of bulk density, resistance to penetration, infiltration rate and hydraulic conductivity (using tension disc infiltrometers) in the soil top layer were carried out. Changes in water profiles through the experimental period were also followed using a neutron probe. Crop development and yield were determined. The soil bulk density in the 0 to 20 cm layer was significantly higher in the CT than in the ?T treatment, mainly after tillage operations (between 10% and 24% higher in CT than in ‘IT). After 3 years of continuous tillage treatments, the soil bulk density did not increase. The resistance to penetration at any time was higher in the CT than in the 7T treatment, but not always significantly different. Infiltration rates were significantly higher in the ‘IT than in the CT treatment (about 35% higher in TT than in CT). After 3 years of the tillage treatments the hydraulic conductivity of the soil surface layer, at a pressure head of 0 mm, was significantly higher in the CT (124 mm h-l) than in the TT (66 mm h-l). This is related to the existence of

* Corresponding author. Tel: + 34-5-46247 1 1; Fax: + 34-5-4624002; E-mail: FMORENO@IRNASE.CSIC.ES.

0167-!987/97/$17.00 Copyright 0 1997 Elsevier Science B.V. All rights reserved. PII SO167-1987(96)01083-5

26 F. Moreno rr d/Soil C? Tilluge Research 41 (1997) 25-42

preferential paths created by an increase of the earthworm population in the CT treatment. Soil water profiles showed higher replenishment of soil water storage in the CT than in the 7T treatment, particularly from October 1994 to June 1995 when the lowest precipitation of the experimental period was recorded. Plant height, leaf area index (LAI) and root length density (RLD) of the first sunflower crop were significantly higher in the TT than in the CT treatment. Nevertheless, the seed yield was slightly higher in the CT than in the TT treatment. In the second sunflower crop, plant height, LAI, RLD were significantly higher in the CT than in the TT treatment except early in the season, and yield was significantly (and extremely) higher in the CT (1521 kg ha-‘) than in the ‘IT (473 kg ha-‘) treatment. During the wheat crop season, plant height and RLD were higher in the TT than in the CT treatment, but grain yield was again slightly higher in the CT than in the TT treatment. The conservation tillage applied seems to be highly effective in enhancing soil water recharge and water conservation, particularly in years with much lower than average precipitation.

Keywords: Conservation tillage; Traditional tillage; Soil physical properties; Water depletion; Crop develop- ment; Crop yield; Wheat, Sunflower

1. Introduction

The increasing interest in the conservation of soil and water resources under rainfed conditions over the last two decades has prompted increased research on effects of different tillage systems on soil properties and crop development and yield. This interest in the conservation of soils and water has favoured the development of tillage practices grouped under the term conservation tillage. Basically, conservation tillage comprises reducing the number of tillage operations and maintaining crop residues on the soil surface (Unger, 1984; Sprague and Triplett, 1986). Conservation tillage is a term covering a range of tillage practices that have, as a common characteristic, the potential to reduce soil and water loss compared with conventional tillage (Mannering and Fenster, 1983). The effects of tillage systems can vary over a wide range of soils and climatic conditions.

Effects of conservation tillage on crop yields have been extensively studied under different conditions of soil and climate. In contrast, there is limited information about its effect on soil physical properties (Heard et al., 1988) and water storage in the soil. This, together with the increase of process modelling of water balance in tillage experiments, has imposed a demand of accurate measurements of soil physical properties, crop development and crop yield. As conservation tillage is also specific to the soil and conditions of the site, no single blueprint of cultural practices can be universally applicable (Lal, 1989). The dependence of conservation tillage on the soil and climatic conditions makes the study of its effects on soil physical properties, water storage and depletion and crops necessary for different areas of the world.

In semi-arid areas, such as southern Spain, water availability is the most important limiting factor in agriculture under rainfed conditions. In soils of the Andalusian Plain (southern Spain), climatological conditions, characterized by the concentration of rain- fall in the autumn-winter period, lead to replenishment of water storage capacity at the end of the winter. The subsequent shortage of precipitation and very high temperatures are responsible for the water depletion observed at the end of summer.

F. Moreno et d/Soil & Tillage Research 41 (1997) 25-42 21

Work has been carried out in the last few years to determine the replenishment and water conservation in the profile of soils under different tillage methods in western Andalusia and their influence on the development of dry farming crops (Moreno et al., 1982, Moreno et al., 1986; Garcia et al., 1986; Pelegrin et al., 1990). Although these experiments show some relationship between soil physical properties and water deple- tion by crops under traditional tillage and no-tillage, more detailed studies are needed on conservation tillage practices.

The objectives of this study were to evaluate the effects of conservation and traditional tillage, under rainfed agriculture in southern Spain, on (i) soil physical properties, (ii) water profile replenishment and depletion by crops in a wheat-sunflower crop rotation, and (iii) crop development and yield.

2. Materials and methods

Field experiments were carried out on a sandy clay loam soil (Xerofluvent) at the experimental farm of the Instituto de Recursos Naturales y Agrobiologia de Sevilla (IRNAS-CSIC) located 13 km southwest of the city of Seville (Spain). Some general characteristics of the soil are shown in Table 1. An area of about 2500 m2 was selected to establish the experimental plots. Experiments started in 1991. During the autumn of that year the general experimental area (2500 m*) was cropped with wheat under rainfed conditions. The tillage operations applied were the traditional ones used in the region. After harvest of the wheat in June 1992, the area was divided into six plots each of approximately 300 m2 (22 X 14 m’). Two tillage treatments were applied: traditional tillage (TT) used in the area for rainfed agriculture and a conservation tillage (CT). Tillage and other agronomical operations given during the experimental period (1992- 1995) are summarized in Table 2. Three replications per treatment were used, distributed in random blocks. A wheat-sunflower crop rotation was established in each treatment.

Changes in the water profile during the experimental period were monitored using a neutron probe (Troxler 3333). Two access tubes for the neutron probe were installed in each individual plot of each treatment, to a depth of 2.3 m. Measurements were carried out at 0.1-m intervals, every 6-10 days during the crop seasons, and at variable intervals during the bare soil period. During the crop season of 1995 (sunflower), changes of

Table I

General characteristics of the soil

Depth Cm) Soil particle size ( pm)

(g per 100 g)

> 20 20-2 <2

CaCO, tg per 100 g)

Organic matter

(gper100g)

0 -0.1 56.6 18.6 24.8 27.0 0.93

0.1-0.3 58.2 17.6 24.2 28.0 0.71 0.3-0.6 45.7 20.7 33.6 26.5 n.d. a > 0.6 50.7 15.0 34.3 n.d. n.d.

a Not determined.

28 F. Moreno et d/Soil & Tillage Research 41 (1997) 25-42

Table 2 Tillage and agronomic operations carried out in the experimental plots

Date Treatments

‘lT (traditional tillage) CT (conservation tillage)

15-07- 1992 Burning the straw of the wheat crop

17-07- 1992 11-10-1992

03-l 1-1992

28-1 l-1992

Mouldboard ploughing 25-30 cm depth Cultivator application 15-20 cm depth

Disc harrowing 15 cm depth

25-02-1993

26-02- 1993

Cultivator application 15-20 cm depth Seeding of sunflower (cultivar ‘florasol’) by

precision line seeder (seed density - 54000

ha- ’ ). Dry-farming sunflower is not usually fertilized in this zone

lo-03- 1993

08-04 1993

20-04 1993

28.07- 1993 3 l-07-1993

25- lo- 1993 12-1 I-1993

Irrigation for emergence (10 mm) Cultivator application between rows

(12-15 cm depth) Cultivator application between rows

(12-15 cm depth) Harvesting

Burning the straw Mouldboard ploughing 25-30 cm depth

Disc harrowing 15 cm depth

19-11-1993

22-11-1993

Fertilization with 400 kg ha- ’ of a complex fertilizer 15- 15- 15 Sowing winter wheat (cultivar ‘ yavaros’,

C-79 R-2) by boot seeder ( - 350000 kernels ha-‘) and disc harrowing 5 cm depthsowing

winter wheat (cultivar ‘yavaros’, C-79 R-2) by boot seeder ( - 350000 kernel

ha-‘)

25-l l-1993 Weed control by weedkiller (trifluraline, Weed control by weedkiller (trifluraline,

1.5 1 ha-‘) 1.5 1 ha- ‘)

16-12-1993

07-06- 1994 15.07- 1994

16-07-1994 20-07-1994 1 l-10-1994

18-l l-1994 21-02- 1995 22.02- 1995

27.02- 1995

N fertilization: 200 kg ha- ’ urea (46% N) N fertilization: 200 kg ha- ’ urea (46% N)

Harvesting Harvesting

Burning the straw

Mouldboard ploughing 25-30 cm depth Cultivator application 12-15 cm depth Disc harrowing 15 cm depth

1 l-04-1995

10-07-1995

Disc harrowing 12 cm depth

Seeding of sunflower (‘florasol’) by precision line seeder (seed density - 54000 ha-‘)

Cultivator application between rows 12-15 cm depth Harvesting

Leaving the straw of the wheat crop on the surface as mulch

Chiseling 25-30 cm depth

Disc harrowing 5-7 cm depth

Weed control by weedkiller (trifluraline, 1.5 1 ha-‘)

Seeding of sunflower (cultivar ‘florasol’) by

precision line seeder (seed density - 54000 ha- ‘) application of terbutryn (2 I ha- ‘)

Irrigation for emergence (10 mm)

Harvesting Leaving the straw on the surface as mulch

Disc harrowing 5-7 cm depth Weed control by weedkiller (glyphosate,

41 ha-‘) Fertilization with 400 kg ha- ’ of a

complex fertilizer 15 15- 15

Leaving the straw on the surface as mulch

Weed control by weedkiller (glyphosate, 4 1 ha-

Weed control by weedkiller (trifluraline, 2 1 ha- Disc harrowing 5-7 cm depth

Seeding of sunflower (‘florasol’) by precision line seeder (seed density - 54000 ha- ‘1

Harvesting

F. Moreno et d/Soil & Tillage Research 41 (1997) 25-42 79

water content in the surface layer (O-15 cm) were monitored using a time-domain-re- flectometry (TDR) with a Tektronix Model 1502C. The TDR waveguides comprised three parallel stainless rods, 2 mm in diameter, and 0.15 m long. A portable computer was used to record and analyse the TDR wave-forms using an analysis similar to that of Baker and Allmaras (1990).

Measurements of water infiltration into the soil were carried out on selected dates of the experimental period using a double-ring infiltrometer (Bouwer, 1961), with three replications per treatment.

The tension-disc infiltrometer was used to determine in situ the hydraulic conductiv- ity and the sorptivity in the range near saturation (Perroux and White, 1988; Smettem and Clothier, 1989). Experiments were carried out with disc infiltrometers of 125 mm (Sl) and 40 mm (S2) radius. Infiltration tests were carried out on the undisturbed soil surface of both treatments at the end of the crop season (wheat) in 1994, and at the beginning of the crop season (sunflower) in 1995. A thin layer (2-3 mm depth) of fine sand, with a radius corresponding to the disc infiltrometer, was used to ensure a good contact between the membrane of the infiltrometer and the soil surface. The pressure heads (Go) chosen were - 100, - 30, and 0 mm. Triplicate flux measurements were made with each disc for the different heads in each treatment. The hydraulic conductiv- ity, K, = K(@,), and the sorptivity, S, = S(+O), were obtained using the multi-disc approach as described by Smettem and Clothier (1989). Following the procedure of Clothier et al. (1992), the disc infiltrometers were also used for direct measurement of the mobile water content. When water infiltration from disc Sl (rl = 125 mm and at -30 mm of water pressure head) reached steady state, the disc reservoir was quickly filled with a tracer solution of 0.1 M KCl. The infiltration of the tracer then proceeded at the same location. The total solute in the soil can be expressed by

BC” = e,c, + eimci, (1)

where C * is the solute concentration of the entire soil solution (6 = 0, + e,,), and C, and C, are the concentrations in the mobile (0,) and in the immobile (0,,) water fractions, respectively.

Assuming steady-state pattern for both water content and pore water velocity immediately under the disc (Philip, 1986), and neglecting diffusion of applied solute from the mobile into the immobile phase over short times scales, then the invading tracer must be only in the mobile phase at the time of sampling. The chloride present in the soil before the infiltration of the KC1 solution was negligible, so Eq. (I) reduces to

*

Om=*g (2) m

where both C* and 13 can be easily measured, and C, is known, being the solute concentration in the reservoir of the disc.

The initial volumetric water content (6,) was determined from soil cores. The initial dry bulk density (D,) of the soil was determined on undisturbed cores of 200 cm3 vohune. The vohtmetric water content at I&(@,) was calculated from the water content of shallow samples scraped from the soil surface under the disc. These samples were taken immediately after removal of the solute-filled disc infiltrometer. Samples were

30 F. Moreno et al./Soil & Tillage Reseurch 41 (1997) 25-42

also taken at different locations at a shallow depth by excavating a face across the diameter of where the disc had been. The gravimetric water content was determined for each sample, and the chloride concentration was measured in the extracted soil solution by using spectrophotometry (Florence and Farrar, 1971).

Bulk density was determined from the ratio mass/volume of soil cores taken with stainless-steel cylinders of 8 cm diameter and 4 cm height. These samples were taken at different depths in both treatments on selected dates during the experimental period. Samples were taken in six replicates per treatment.

The resistance to penetration was determined by using a falling cone penetrometer of 1.5 cm* conical section and an angle of 30” (Hoyas, 1994) to a depth of 40 cm with ten replicates per treatment, deducing a medium curve for each application period.

Moisture-retention curves were determined in the laboratory on undisturbed soil samples taken in cylinders of 8 cm diameter and 2 cm height by suction in fritted-glass plates (Vomocil, 1965) and by pressure in ceramic plates (Richards, 1948). Samples were taken in three replicates per treatment.

Plant height and leaf area index were determined during the crop seasons. Root length density was determined by the method of Newman (1966) in samples taken by auger, once in the season when the crop was fully developed. Stomata1 conductance (Model LI-6200, Li-Cor, Inc., NE, USA) and leaf water potential (Pressure Chamber, Soil Moisture Equipment Corp., Santa Barbara, USA) were measured in sunflower during the crop season of 1995.

Rainfall and other climatic parameters were obtained from the weather station located at the experimental farm.

3. Climatic conditions of the experimental zone

Climate of the zone in which the experiment was conducted is typically Mediter- ranean, with mild rainy winters (500 mm mean rainfall) and very hot, dry summers. The experimental period (1992-1995) was characterized by a severe drought. Table 3 shows rainfall during the periods September-August (hydrological year) and September-De- cember of every year of the experimental period together with the averages of 1971-1991.

Table 3 Comparison of rainfall during the experimental period with the 20 year average in the zone

Period

(September- August)

1991-1992 1992- 1993 1993-1994

1994-1995

Rainfall

(mm)

446 353 328

245

Average for the

period September- August (1971- 1992), (mm)

500

Period (September- December)

1992 1993

1994

Rainfall

(mm)

160 147

111

Average for the period September- December ( 197 l-

1992) (mm)

247

F. Moreno et al./ Soil & Tilluge Research 41 11997) 25-42 31

4. Results and discussion

4.1. Bulk density

Results of bulk density (D,) on selected dates of the experimental period are shown in Table 4. In the first year of the experiment at the planting date of the sunflower crop, bulk density was significantly different in the two treatments, being higher in the CT than in the TT treatment. This is related to the cultivator application given in the TT treatment (Table 2) before planting. In contrast, on 2.5-05-1993 (when the sunflower crop was at flowering) bulk density had increased in the TT treatment and was not significantly different from the value observed in CT.

During the second crop of the rotation (wheat), when the crop was 20 cm height (14-Ol-1994), D, was practically the same in both treatments. This was due to the same pattern of soil consolidation on both treatments before sowing. On 25-05-1994 (close to harvesting) D, had increased in both treatments, and was significantly higher in CT than in TT in the O-10 cm soil layer depth.

In 1995 at the planting date (23-02-1995) of the sunflower crop (third crop of the rotation), once again the two treatments showed significantly different D, values, being higher in the CT than in the TT treatment. As in the first sunflower crop, D, values at flowering were not significantly different between treatments, although they were slightly higher in CT than in TT.

From these results it seems that the bulk density was always higher in the CT treatment than in the TT, as has been observed by other authors who have compared no-tillage or reduced tillage and conventional tillage (Hill and Cruse, 1985) and by ourselves (Pelegrin et al., 1990). These results clearly show that in both treatments D,

Table 4 Bulk density CD,) and water content (0) for both treatments on selected dates of the experimental period (TT,

traditional tillage; CT, conservation tillage)

Date Depth (cm) Treatments

T-r CT

D, (Mg m-3) tJ (m3 m-3) D, (Mg m-3) 0 (m3 mm3)

26-02- I993 O-10 1.19a 0.190‘ 1.48b 0.161 1 O-20 1 SOa 0.192” 1.65b 0.159

25-05-1993 O-10 1.46a 0.087 )L 1Sla 0.106 * 10-20 1.58a 0.140’ 1.62a 0.132’

14-01-1994 O-10 1.35a 0.247 * 1.39a 0.275 ’ IO-20 1 S4a 0.297 * 1.60a 0.293 *

25-05- 1994 O-IO 1.53a 0.258 * 1.65b 0.292 10-20 1.65a 0.246 * I .68a 0.229 *

23-02-1995 O-10 1.26a 0.162” 1.39b 0.195 1 O-20 1.40a 0.208 * 1.62b 0.221*

I&05- 1995 O-10 1.36a 0.091* 1.43a 0.086 * 10-20 1.57a 0.111- 1.65a 0.119’

Values of bulk density CD,) in each line followed by the same letter are not significantly different (p < 0.05). Values of 8 in each line followed by X are not significantly different (P < 0.05).

32 F. Moreno et d/Soil & Tilluge Research 41 (1997) 25-42

Penetration resistance (MPa)

012345678

J

Penetration resistance (MPa)

0 1 2 3 4 5 6 7 8

24/02/93 I OTT . CT

- 1.s.d.

Fig. 1. Penetration resistance on selected dates during the sunflower crop season in 1993 and 1995. TT, traditional tillage; CT, conservation tillage. First and second date in each year were planting and flowering

time, respectively.

did not increase with the time after the establishment of the two tillage systems. In our case, 3 years of continuous tillage treatments may have been sufficient time to attain a stabilized range of D, values during the crop season. This is in agreement with results of Pidgeon and Soane (1977) which showed that 3 years were required in soils under no-tillage to reach an equilibrium D,.

4.2. Penetration resistance

Fig. 1 shows the results of penetration resistance (PR) for both treatments with sunflower crop during the first and third year after the establishment of the tillage systems. At planting date in the first year, PR was higher in the CT than in the TT treatment between 0.1 and 0.4 m in depth. The differences between treatments were not significant except at 0.25 m depth. The water content at this date was higher in TT (0.196 cm3 cm- 3, than in CT (0.160 cm3 cmm3) in the 10-40 cm soil layer depth. In the third year at planting date, PR values in the CT treatment were significantly higher than in the TT treatment for the depths between 10 and 25 cm. At this time the water content was practically the same in both treatments for these depths (0.208 cm3 cmp3 in TT and 0.221 cm3 crne3 in CT). An increase of PR similar to that in the CT treatment

F. Moreno et d/Soil & Tillage Research 41 (1997) 25-42 33

600 1993 1995

-T-r VT-r 5co 1 l CT o CT

0 50 100 150 Time (min)

Fig. 2. Infiltration rates (i) in the traditional tillage (‘IT) and conservation tillage (CT) treatments at planting

date of sunflower crop in 1993 and 1995.

has been observed by other authors in reduced tillage and no-tillage experiments under different conditions of soils and climate (Lindstrom et al., 1984; Tollner et al., 1984; Hill, 1990).

In our study the higher PR in the CT treatment than in the TT treatment in the third year (about 2 MPa) could be a limiting factor for root growth. Although the soil is not a vertisol, we observed some small fissures, with water content lower than 0.23 cm3 cm -3, due to shrinking processes; in the clay fraction the smectite content is about 20-30%. These fissures allow root growth, as has been shown by others (Whiteley and Dexter, 1983).

Fig. 1 also shows the PR in both treatments at the flowering stage in the sunflower crop during 1993 and 1995. In both years, PR increased compared with the values observed at planting date. In the first year (1993) PR values for depths between 20 and 45 cm were significantly higher in the CT than in the ‘IT treatment. In contrast, in the third year (1995) no significant differences were found between treatments, although there was a tendency for higher PR in CT in the top 30 cm layer.

4.3. Water injiitration, hydraulic conductivity and sorptivity

Infiltration rates measured with the double ring infiltrometer in both treatments are shown in Fig. 2. The measurements were made at planting dates of sunflower crop in the first (1993) and third (1995) years after the establishment of tillage systems. Infiltration rates were significantly higher in the ‘lT than in the CT treatment in both years. Differences observed in the TT treatment between 1993 and 1995 were significant only for the infiltration rates obtained in the first 30 min. These differences can be due to a different pore size distribution in the surface layer (first 2 or 3 cm). Initial soil water content at O-10 cm in depth was practically the same in both years (0 = 0.160-0.180 cm3 cme3). When steady state was reached, no significant differences were observed between the two dates in the TT treatment. In the CT treatment, infiltration rates decreased between 1993 and 1995. This decrease was statistically significant. The decrease in the infiltration rate in CT after 3 years is not in accord with D, results,

34 F. Moreno et d/Soil & Tillage Research 41 (1997) 25-42

Table 5

Hydraulic conductivity (K) and sorptivity (S) of the soil surface measured at harvest of the wheat crop (June 1994) and at the beginning of growth of the sunflower crop (March 1995) in the Tf (traditional titlage) and

CT (conservation tillage) treatments

+bm>

June 1994

0

-30

- 100

Treatments

T-r

K(mmh-‘)

24.5 a

9.4 a

1.5 a

S (mm h - o.5)

39.0 *

30.0 *

16.2*

CT

K(mmh-‘)

4.0 b

2.2 a

0.4 a

S (mm h- o.5)

40.2 *

36.6 *

23.4’

March 1995

0 -30

-100

66.6 a 49.2 - 124.6 b 39.0 r 48.2 a 40.2 x 16.6a 30.0 -

4.3 a 22.2 * 3.6 a 10.8

Values of K in each line followed by the same letter are not significantly different (P < 0.05).

Values of S in each line followed by * are not significantly different (P < 0.05).

which did not show an increase. Thus the reduction in infiltration rates could be due to a different pore size distribution rather than to the total porosity in the surface layer.

Infiltration rates measured during the late stage of the wheat crop (May 1994) in the second year were practically the same in both treatments (data not shown). This could be due to a similar soil consolidation in the surface layer.

Hydraulic conductivity and sorptivity deduced from measurements with the tension disc infiltrometer are shown in Table 5. At harvest of the second crop of the rotation, hydraulic conductivity in the TT treatment was higher than in CT treatment for the three pressure heads applied (rj~~ = 0, - 30, - 100 mm). Sorptivity was similar under both treatments. Differences in hydraulic conductivity between the two treatments can be attributed to a different soil consolidation, in agreement with the different soil bulk density observed at this time in the two treatments (significantly higher in CT than in TT, Table 4). The mobile water content was also higher in the TT treatment (0.190 cm3 cm-3) than in the CT treatment (0.158 cm3 cmm3). In contrast, the hydraulic conductiv- ity at r,!~a = 0 was significantly higher in CT than in TT at the beginning of the third crop. At this time the mobile water content was 0.210 cm3 cmW3 and 0.189 cm3 cmm3 in CT and TT, respectively. This could be related with the existence of preferential paths created by an increase of the earthworm population in the CT treatment that was visually observed during a soil sampling at this time. Gantzer and Blake (1978) reported an increased earthworm population in no-tillage compared with conventional tillage. The effect of these preferential paths due to biochannels in the infiltration has been reported by Ehlers (1975). Douglas et al. (1980) found macropores of earthworm channels to be the prime cause for the difference between cultivated and direct-drilled soil. Their direct-drilled soil had the greatest hydraulic conductivity. At @a = - 30 mm the hydraulic conductivity was higher in the TT treatment than in the CT, but practically the

F. Moreno et (11. /Soil & Tillage Research 41 (1997) 25-42 35

same at $a = - 100 mm (Table 5). Sorptivity was higher in TT than CT at all pressure heads applied.

Comparing hydraulic conductivity values of the two measurements dates we observed an increase from the measurements made at harvest in the second year to the beginning of the crop season in the third year in both treatments. This is due to the tillage applied before sowing (Table 2).

On both measurement dates the hydraulic conductivity at qGO = - 100 mm was practically the same in TT and CT treatments. This result is similar to that reported by Sauer et al. (1990) for mouldboard plough and no-till using the tension-disc infiltrome- ter. Their results show differences between treatments at I,!J~ > - 100 mm as found in our experiments (Table 5). In our case this difference was statistically significant only at & = 0 mm. In June 1994, for the TT treatment, the hydraulic conductivity rose greatly between & = - 30 mm and saturation (& = 0), whereas the sorptivity was not greatly changed. As has been reported by Sauer et al. (1990) such a relative change implies macropores with large cross-sectional area since they rapidly transmit water (hydraulic conductivity) without significant transient sorption into the walls (sorptivity). The same was observed in March 1995 in the CT treatment.

4.4. Soil water content and its depletion

Changes in the soil water profile during the first crop of the rotation (sunflower) in 1993 are shown in Fig. 3a and Fig. 3b. Soil water profiles 52 days after sowing (20-4-1993) were similar in both treatments down to 1 m depth. Below this depth the soil water content in the CT treatment was higher than in the TI treatment. This could be due to a better recharge of the profile in the deeper layers in the CT than in the TT treatment. A similar pattern was observed by Pelegrfn et al. (1990) for a rainfed sunflower crop in this area when comparing two tillage treatments (mouldboard plough- ing and no-tillage). Changes of the water profile shown in Fig. 3a and Fig. 3b seem to indicate a similar water depletion by the crop in both treatments. Below a depth of 2 m, water depletion by the crop was slightly greater in the CT than in TT treatment. Water depletion by sunflower crop below 2 m depth has been observed by Unger et al. (1976), and in our region by Berengena et al. (1985) and Pelegrin et al. (1990). Rachidi et al. (1993) reported that a rainfed sunflower under semi-arid conditions depleted a signifi- cant amount of water at a depth of 2.5 m.

The changes of water profile during the second sunflower crop (1995) are shown in Fig. 3c and Fig. 3d. In that year the water storage replenishment of the soil in the TT treatment took place only down to 1 m depth. In contrast, replenishment of the soil water profile in the CT treatment was observed down to 1.4 m depth. A few days before sowing (I 6-2- 1995), the water content of the soil layer at the 0- 1.4 m depth was higher in the CT than in the TI treatment. It seems that the water recharge of the profile during the autumn and early in the winter (bare soil period) was more effective in the CT than in the TT treatment. These results showed that, under a situation with much lower rainfall than the average during autumn and winter (Table 3), the CT treatment was able to recharge the profile much better than the TT. The presence of residues of the preceding crop in the soil surface in the CT treatment seems to be highly effective in

36 F. Moreno et al./Soil & Tilluge Research 41 (1997) 25-42

0 x-04-93 A 05-05-93 0 19-05-93 l 25-05-93

l 15-06-93 7 30-06-93

2.5 ’ ’ 2.5 I

B (cm3 Cm-3) e (cm3 crL3)

U.U r

l 16-02-95 A 24-03-95 9 10-04-95 0 21-04-95 A 04-05-95 Cc) 0 19-05-95 + 09-m-95

2.5 -

i

2.5 '

Fig. 3. Water profiles in the traditional tillage (Ti’) and conservation tillage (CT) treatments during the sunflower crop season in 1993 and 1995.

0.3

;- 0.2

$

2

; 0.1

0.0

o TT l TT

a CT

A CT

Soil layer: O-15 cm

ulant row between plant rows

I , / / , , /

0 20 40 60 80 100 120 Days after sowing (day 0=27-2-95)

Fig. 4. Change of the soil water content in the soil layer O-15 cm depth in traditional tillage (TT) and

conservation tillage (CT) treatments during the sunflower crop season in 1995.

F. Mnreno et d/Soil & Tillnge Reseurch 41 (1997) 25-42 31

e (cm3 Cm-3) B (cm 3 -3

cm )

0.0 0.1 0.2 0.3 0.0 0.1 0.2 0.3

1.5

2.0

- TT

J

Fig. 5. Water profiles in the traditional tillage (‘IT) and conservation tillage (CT) treatments during the wheat

crop season in 1993- 1994.

enhancing water infiltration and water conservation under dry weather conditions, as has been reported by Unger et al. (1991) among others. During 1995 the sunflower depleted water down to 1 m and 1.6 m depth in the TT and CT treatments, respectively.

The water content in the surface layer of the soil (O-15 cm) was monitored in detail during the crop season (sunflower) of 1995. The results are shown in Fig. 4. From the beginning of measurements the soil water content was significantly higher in the CT than in the TT treatment. A few days after emergence of plants, no difference in the soil water content was observed between the position in the plant row and the centre between rows for the TT treatment. In contrast, for the CT treatment a different soil water content was observed between the two positions of measurement. This difference was maintained throughout the season, being maximum 50 days after sowing. It can be due to both higher evaporation from the soil surface (the soil surface was altered during sowing) and higher water uptake by roots in the plant row position than in the centre between rows. These results clearly show that the CT treatment improved the storage and conservation of water in the surface layer of the soil.

Water depletion by the wheat crop for selected dates of the growing season in both treatments is shown in Fig. 5. The changes of water profiles show a similar water depletion in both treatments. The crop depleted more water in the CT than in the TT treatment for the soil layer between 90 and 125 cm depths.

4.5. Crop development and crop yield

Plant height during the growing season of the first sunflower crop (1993) was significantly higher in the TT than in the CT treatment before flowering (Table 6). At flowering (20-27 May) the crop reached the same height in both tillage treatments. The leaf area index (LAI) was also significantly higher in the TT than in the CT treatment during the whole growing season, These differences in both parameters could be related with the differences observed in the soil physical properties (i.e. bulk density and penetration resistance) during the first year after the establishment of both tillage treatments. In contrast, during the second sunflower crop (third year after the establish-

38 F. Moreno et al./ Soil & Tilluge Research 41 (1997125-42

Table 6 Plant height, leaf area index (LAI) and root length density (RLD) for the sunflower and wheat crops in both

treatments (DAS, days after sowin,, 0. ‘IT, traditional tillage; CT, conservation tillage)

Date Plant height ’ km) LA1 ’ RLD ’ (cm cm- 3,

DAS TT CT 7-r CT Depth km) IT CT

Sunflower 1993

21 April

29 April 7 May

14 May

20 May 3 June

17 June

Sunflower 1995 6 April

17 April

3 May

16 May 29 May

Wheat 1994

14 January 21 February 12 March

14 April

54

62 70

77

83 97

111

37.4a

44.0a 74.la

117.3a

143.la 171.8a

26.5b

39.0b

67.9b 107.8b

139.4a

167.8a

38 ll.la 8.7a 49 31.3a 29.2a

65 54.0a 70.2b 78 66.6a 109.0b 91 68.7a 114.7b

53

91 110

143

20.2a

58.0a 74.0a

89.4a

17Sb 55.2b

69.8b 82.9b

0.53a 0.36b

1.23a 1.03b 2.19a 2.01b 2.25a 2.17b

0.17a O.llb 0.58a 0.36b 0.72a 0.92a 0.67a 1.60b

O-18

18-36 36-54

O-18 18-36

36-54

O-10 10-20

20-30 30-40 40-50

(3 June) 1.02a 0.77b

0.23a 0.21a

0.20a 0.22a

(29 May) 0.62a 1.53b

0.19a 0.33b

0.16a 0.31b

(8 April) 3.72a 4.42a 3.15a 2.12b

2.03a 1.99a 1.98a I .67a

1.17a 1.07a

’ Treatment means for a date and parameter followed by the same letter are not significantly different

(P < 0.05). ’ Treatment means in the same row followed by the same letter are not significantly different (P < 0.05).

ment of the tillage treatment) plant height and LA1 were significantly greater in the CT than in the TT treatment. This could be due to a better soil water storage at planting date and during the crop season in the CT than in the TT treatment, as has been shown before (Fig. 3c and Fig. 3d). These results seem to indicate that 3 years of conservation tillage improved water infiltration and redistribution in the soil compared with the traditional tillage, particularly in the year with the lowest rainfall, during the preceding autumn and winter (Table 3), of the experimental period.

Although in 1995 (second sunflower crop) less water was available in total in the TT treatment because of smaller replenishment as compared with the CT treatment, plants developed more rapidly in the TT than in the CT treatment early in the season (from seeding to the middle of April, see plant height and LA1 in Table 6). This was due, in part, to a higher N availability in the TT than in the CT treatment (data not published). In this year of low rainfall and water recharge, higher water use by sunflower plants early in the season (height, LAI, see Table 6) in the TT treatment, leaving not enough water for the seed-fill season. On the other hand, in the CT treatment the sunflower development was slightly retarded and there was therefore (and because of higher water

F. Morerw et cd./ Soil & Tilluge Research 41 (1997) 25-42 39

replenishment) more water available during the seed-filling stage. The evapotranspira- tion of sunflower during the early season (from seeding to the middle of April) was 65 mm in the TT treatment and 48 mm in the CT. In contrast, from the middle of April to the end of the season, the evapotranspiration amounted to 99 mm and 150 mm in the TT and CT treatments, respectively. This behaviour was also observed during the first crop of the rotation, although the effects on crop development were not so drastic as in 1995 due to higher water recharge in accordance with the rainfall (Table 3). In 1993 the evapotranspiration of sunflower during the early season (from seeding to the middle of April) was 44 mm in the TT treatment and 21.5 mm in the CT. In contrast, from the middle of April to the end of the season, the evapotranspiration amounted to 209.5 mm and 207.5 mm in the TT and CT treatments, respectively. The evapotranspiration of the wheat crop during the early season (from sowing to the middle of February) was 61.5 mm in the TT treatment and 67 mm in CT. From the middle of February to the end of the season the evapotranspiration reached 168 mm and 174 mm in the TT and CT treatments, respectively.

The root length density (RLD) (Table 6) of the first sunflower crop (1993) at the flowering stage, was significantly higher in the TT than in the CT treatment in the soil layer O-18 cm in depth. In the soil layers 18-36 cm and 36-54 cm in depth, the root length density was practically the same in both treatments. In contrast, in the second sunflower crop (1995) the RLD was significantly greater in the CT than in the TT treatment (Table 6). In that year, the improved soil water storage and the presence of a great number of channels produced by the increase of the earthworm population in the CT treatment allowed a better root development than in the TT treatment.

Measurements of the midday leaf water potential (MLWP) made in sunflower plants during 1995 did not show significant differences between treatments. The MLWP measured when the crop was at the flowering stage (19-5- 1995) reached - 0.57 and - 0.6 1 MPa in the CT and TT treatment, respectively. On the same date, the maximum stomata1 conductance was significantly higher in the CT treatment (0.76 cm s- ’ ) than in the TT treatment (0.29 cm s-l). These results are in accord with a higher soil water content, below a depth of 30 cm, in the CT than in the TT treatment (Fig. 3c and Fig. 3d) on the measurement date (19-5-1995). It seems that under low available soil water (TT treatment), the plant reduced the stomata1 conductance to maintain a leaf water potential similar to that observed in the CT treatment. This behaviour has been reported by Gollan et al. (1986).

In 1993 the seed yield of the sunflower was similar in both treatments (Table 7). Although the seed yield was slightly higher in the CT than in the TT treatment, the difference was not statistically significant. The weight of 1000 seeds was also not significantly different between treatments. Seed yield and weight of 1000 seeds in 1995 were much higher in the CT than in the TT treatment. Yields of the second sunflower crop are in accord with the differences in the soil water profile observed between treatments.

In 1995, as mentioned before, the higher water use by the sunflower crop in the TT treatment than in CT early in the season - leaving not enough water for the seed-fill season - was the reason for the low seed yield and seed weight in the TT treatment (Table 7). This may indicate that the sunflowers in the TT treatment had ripened

40 F. Moreno et al./Soil C? Tillage Research 41 (1997) 28-42

Table 7 Yield of sunflower and wheat crops under both tillage treatments (TT, traditional tillage; CT, conservation

tillage)

Treatments Sunflower Wheat

1993 1995 1993-1994

Seed yield 1000 seed Seed yield 1000 seed Grain yield 1000 grain (kg ha- ‘) weight(g) (kgha-‘) weight (g) (kg ha- ‘) weight(g)

l-r CT

2143a

2463a 44Sa 43.9a

473a 1521b

23.8a

40.5b 3069a 3266a

41.5a 43.5a

Values per columns followed by the same letter are not significantly different (P < 0.05).

prematurely and imperfectly. This behaviour was much less important during the sunflower crop in 1993.

The wheat crop (second crop of the rotation) grown in 1993-1994 showed signifi- cantly higher plant height in the TT than in the CT treatment at any time during the growing period (Table 6). The RLD, when the crop was fully developed (Table 6), was greater (not statistically significant) in the CT than in the TT treatment for the soil layer O-10 cm in depth. In contrast, for the soil layer lo-20 cm in depth, the RLD was significantly greater in the TT than in the CT treatment. Below a depth of 20 cm the RLD was slightly greater in the TT than in the CT treatment, although the difference was not statistically significant.

The yield of the wheat crop (Table 7) was 3266 kg ha-’ and 3069 kg ha- ’ in the CT and TT treatments, respectively. As in the case of the sunflower, the wheat yield was higher in the CT than in the TT treatment, although these values were not significantly different. The weight of 1000 grains was also higher in the CT than in the TT treatment, but not statistically different.

Water use efficiency expressed in terms of seed or grain yield produced by water used by the crop was 8.5 kg ha-’ mm -’ in the TT treatment and 10.7 kg ha- ’ mm-’ in CT, for the sunflower crop in 1993. In contrast, for the sunflower crop in 1995 the water use efficiency amounted to 2.9 kg ha-’ mm-’ and 7.7 kg ha-’ mm-’ in the TT and CT treatments, respectively. For the wheat crop the water use efficiency was practically the same in both treatments: 13.4 kg ha- ’ mm-i and 13.6 kg ha-’ mm-’ in TT and CT treatments, respectively.

5. Conclusions

Soil bulk density, measured after tillage operations, was significantly higher in the conservation tillage treatment than in the traditional tillage treatment. At the end of the crop seasons, the soil bulk density was similar in both treatments. After 3 years from establishment of tillage treatments, the soil bulk density did not increase.

The penetration resistance in the third year, measured at planting date, was signifi- cantly higher in the conservation tillage treatment than in the traditional tillage treatment for the soil layer 0.1-0.25 m. Although the penetration resistance values in the conservation tillage treatment could be a limiting factor for the root growth, the

F. Moreno et al./ Soil & Tillage Research 41 (1997) 25-42 41

existence of small fissures and the increase of earthworm channels in the soil allowed root development.

The hydraulic conductivity at the beginning of the third crop season, at qVO = 0 mm, was significantly higher in the conservation tillage treatment than in the traditional tillage treatment. The mobile water content was also higher in the conservation tillage treatment than in the traditional tillage treatment.

Soil water depletion by crops, in the soil layer at the O-O.9 m depth, was similar in the two treatments for the first and second growing seasons. In the third crop season, water depletion was different in the two treatments due to a better soil water recharge, during the previous autumn and winter, in the conservation tillage treatment than in the traditional tillage treatment.

Under the conservation tillage applied the crop saved water, early in the season, leaving more water for the seed or grain filling season. Because of this the yield was slightly higher in the conservation tillage treatment than in the traditional tillage for the first and second crops, and extremely higher in the third crop. Water use efficiency was always higher in the conservation tillage treatment than in the traditional tillage.

The conservation tillage applied seems to be highly effective in enhancing soil water recharge and water conservation, particularly in years with much lower than average precipitation.

Acknowledgements

We thank Prof. Dr. M.B. Kirkham for reviewing an early draft of the paper and the helpful comments. Thanks are also due to A. Hoyas, D. Mor&r and J. Rodriguez for help with measurements in the field. This study was supported with funds of the Spanish CICYT, project AGF93-0613-C02-01, and the Junta de Andalucia (Research Group 2042).

References

Baker, J.M. and Allmaras, R.R., 1990. System for automating and multiplexing soil moisture measurements by

time-domain reflectometry. Soil Sci. Sot. Am. J., 54: l-6. Berengena, J., Jimtnez, C., Alba, A. and Fereres, E., 1985. The role of the root system in the drought

resistance of sunflower. In A. Penier and C. Riou (Editors), Le besoins en eau de cultures. Conference

International, Paris 1984, pp. 385-395. Bouwer, H., 1961. A double tube method for measuring hydraulic conductivity in sites above a water table.

Soil Sci. Sot. Am. Proc., 25: 334-339. Clothier, B.E., Kirkham, M.B. and McLean, J.E., 1992. In situ measurement of the effective transport volume

for solute moving through soil. Soil Sci. Sot. Am. J., 56: 733-736. Douglas, J.T., Goss, M.J. and Hill, D., 1980. Measurements of pore characteristics in a clay soil under

ploughing and direct drilling, including use of a radioactive tracer (144Ce) technique, Soil Tillage Res., 1: 11-18.

Ehlers, W., 1975. Observations on earthworm channels and infiltration on tilled and untilled loess soil. Soil Sci., 119: 242-249.

Florence, T.M. and Farrar, Y.J., 1971. Spectrophotometric determination of chloride at the parts-per-billion

level by mercury(E) thyocyanate method. Anal. Chim. Acta, 54: 373-377.

Gantzer, C.J. and Blake, G.R., 1978. Physical characteristics of the Le Seur clay loam soil following no-till and conventional tillage. Agron. J., 70: 853-857.

42 F. Moreno et al./Soil & Tillage Research 41 (1997) 25-42

Garcia, M., Giraldez, J.V., Gonzalez, P. and Fereres, E., 1986. In: A. Pulido (Editor), Aprovechamiento de1

agua en el secano de la campiiia andaluza. Proceedings II Simposio El Agua en Andalucia, Granada, Spain, Vol. II, pp. 653-660.

Gollan, T., Passioura, J.B. and Munns, R., 1986. Soil water status affects stomata conductance of fully turgid

wheat and sunflower leaves. Aust. J. Plant Physiol., 13: 459-464. Heard, J.R., Kladivko, E.J. and Mannering, J.R., 1988. Soil macroporosity, hydraulic conductivity, and air

permeability of silty soils under long-term conservation tillage in Indiana. Soil Tillage Res., 11: l-18.

Hill, R.L., 1990. Long-term conventional and no-tillage effects on selected soil physical properties, Soil Sci.

Sot. Am. J., 54: 161-166. Hill, R.L. and Cruse, R.M., 1985. Tillage effects on bulk density and soil strength of two mollisols. Soil Sci.

Sot. Am. J., 49: 1270-1273. Hoyas, A., 1994. Estudio monograftco de las propiedades fisicas y su evolucidn en un suelo con una rotaci6n

de cultivos en secano (trig0 y girasol), bajo dos sitemas de laboreo (laboreo traditional y laboreo de

conservacibn). Trabajo fin de Camera, EUITA, University of Seville (Spain), 144 pp. Lal, R., 1989. Conservation tillage for sustainable agriculture: tropics versus temperate environments. Adv.

Agron., 42: 85-197. Lindstrom, M.J., Voorhees, W.B. and Onstad, C.A., 1984. Tillage system and residue cover effects on

infiltration in northwestern corn belt soils. J. Soil Water Conserv., 39: 64-68. Mannering, J.V. and Fenster, C.R., 1983. What is conservation tillage?. J. Soil Water Conserv., 38: 140-143. Moreno, F., Muriel, J.L. and Arrue, J.L., 1982. Effect of soil physical conditions on cereal growth in SW

Spain. Proceedings Ninth Conference of the International Soil Tillage Research Organization, Osijek,

Yugoslavia, Vol. I, pp. 363-369. Moreno, F., Martin-Aranda, J., Konstankiewicz, K. and Gomez-Rojas, F., 1986. Changes in soil physical and

mechanical properties due to transformation from dry farming (olives) to irrigation (annual crops). Soil

Tillage Res., 7: 75-84. Newman, El., 1966. A method of estimating the total length of root in a sample. J. Appl. Ecol., 3: 139-145.

Pelegrin, F., Moreno, F., Martin-Aranda, J. and Camps, M., 1990. The influence of tillage methods on soil

physical properties and water balance for a typical crop rotation in SW Spain. Soil Tillage Res., 16: 345-358.

Perroux, K.M. and White, I., 1988. Designs for disc permeameters. Soil Sci. Sot. Am. J., 52: 1205-1215.

Philip, J.R., 1986. Linearized unsteady multidimensional infiltration. Water Resour. Res., 22: 17 17- 1727. Pidgeon, J.D. and Soane, B.D., 1977. Effects of tillage and direct drillin, * on soil properties during the growing

season in a long-term barley mono-culture. system. J. Agric. Sci., 88: 431-442. Rachidi, F., Kirkham, M.B., Stone, L.R. and Kanemasu, E.T., 1993. Soil water depletion by sunflower and

sorghum under rainfed conditions. Agric. Water Manage., 24: 49-62. Richards, L.A., 1948. Porous plate apparatus for measuring moisture retention and transmission by soil. Soil

Sci., 66: 105-l IO. Sauer, T.J., Clothier, B.E. and Daniel, T.C., 1990. Surface measurements of the hydraulic properties of a tilled

and untilled soil. Soil Tillage Res., 15: 359-369. Smettem, K.R.J. and Clothier, B.E., 1989. Measuring unsaturated sorptivity and hydraulic conductivity using

multi-disc permeameters. J. Soil Sci., 40: 563-568. Sprague, E.M. and Triplett, G.B., 1986. No-Tillage and Surface-Tillage Agriculture. Wiley, New York.

Tollner. E.W., Hargrove, W.L. and Langdale, G.W., 1984. Influence of conventional and no-till practices on soil physical properties in the southern Piedmont. J. Soil Water Conserv., 39: 73-76.

Unger, P.W., 1984. Tillage systems for soil and water conservation. FAO Soils Bulletin NO. 54, 277 pp.

Unger, P.W., Allen, R.R., Jones, O.R., Matter, A.C. and Stewart, B.A., 1976. Sunflower research in the

southern High Plains. Proc. Sunflower Forum, I: 24-29. Unger, P.W., Stewart, B.A., Parr, J.F. and Singh, R.P., 1991. Crop residue management and tillage methods

for conserving soil and water in semi-arid regions. Soil Tillage Res., 20: 219-240. Vomocil, J.A., 1965. Porosity. In: C.A. Black (Editor), Methods of Soil Analysis. Part I. Agronomy. Am. Sot.

Agron., Madison, WI, pp. 299-314. WhiteIcy, G.M. and Dexter, A.R., 1983. Behaviour of roots in cracks between soil peds. Plant Soil, 74:

153- 162.