|
|
||||||||
a Institute of Soil and Water Conservation, Chinese Academy of Sciences and Ministry of Water Resources, Yangling, Shaanxi Province 712100, People's Republic of China
b Institute of Soil, Water and Environmental Sciences, A.R.O., the Volcani Center, Bet Dagan 50250, Israel
meni{at}agri.gov.il
| ABSTRACT |
|---|
|
|
|---|
Abbreviations: CV, coefficient of variance EC, electrical conductivity LEPAS, low-energy precision application socks MIS, moving irrigation system SAR, Na adsorption ratio
| INTRODUCTION |
|---|
|
|
|---|
Moving irrigation systems have become increasingly popular in recent years (Anonymous, 1993). However, an MIS is designed to apply given amounts of water within relatively short periods; therefore, it is characterized by a relatively high instantaneous rate of water application (Gilley and Mielke, 1980). When the water application rate exceeds the soil infiltration rate, runoff will occur. For example, toward the outer end of a 53-ha center-pivot, Addink (1975) found runoff values as high as 65% of the applied water on a very fine sandy soil. Likewise, Ben-Hur (1994) found that under irrigation with MIS (average application rate of
100 mm h-1), the runoff percentage of irrigation water from 3 by 5 m plots in cotton (Gossypium hirsatum L.), corn, and peanut (Arachis hypogaea L.) fields was
40% and in potato (Solanum tuberosum L.) field
60%.
According to the field characteristics (slope and soil surface roughness), the runoff may flow out of the field and/or accumulate in small depressions within the field. Runoff from a cultivated field is usually lost to crop production, and it accelerates soil erosion and fertilizer depletion. Local runoff within the field leads to nonuniform water distribution and reduces irrigation efficiency and crop yield significantly (Letey et al., 1984; Ben-Hur et al., 1995).
The runoff potential under irrigation with MIS is especially high in soils that are characterized by low permeability (Kincaid et al., 1969). Ben-Hur et al. (1989) indicated that seal formation at the soil surface during irrigation with sprinkler MIS is the main cause for the reduction of soil permeability in a semiarid region.
Soil tillage practices, such as microbasins (pitting) and dikes across the furrows, increase the surface storage capacity of the field, which in turn decreases runoff formation within the field. Morin et al. (1984) observed that these practices increased the surface storage in a wheat (Triticum aestivum L.) field under rainfall conditions in arid and semiarid regions. Ben-Hur et al. (1995) found that dikes increased the water uniformity within a peanut field irrigated with a sprinkler MIS. The average peanut pod yield in the dike treatment was 880 kg ha-1 higher than in the control treatment where runoff was allowed to flow downhill. Stern et al. (1992) studied the effects of surface treatments on runoff amount and wheat crop productivity in small plots (5 by 2 m) under irrigation with a sprinkler MIS. Runoff percentages from plots with no treatment and with dikes plus 5 Mg ha-1 phosphogypsum were 36.1 and 1.1%, respectively, and the corresponding grain yields were 2.12 and 3.66 Mg ha-1.
The effect of surface runoff on the distribution of water content in a sloping field has been studied mostly under rainfall or irrigation with fresh water of high quality. However, conventional sources of fresh water in arid and semiarid regions are scarce, and this has led to increased use of saline and sodic water for irrigation (Bresler et al., 1982).
Irrigation with saline and sodic water via a sprinkler MIS can affect the water and salt regimes in the root zone, and the crop yield in three main ways:
These effects of the water application methods and tillage practices on the distribution of water and salt in a sloping field irrigated with saline and sodic water using an MIS have not been well documented.
The objectives of our study were (i) to determine the effects of runoff on the distributions of soil water content, soil salt concentration, and corn yield under irrigation with saline water applied by MIS and (ii) to study the interaction between tillage practices (microbasin and dike) and water application methods (sprinkler and flooding MIS) in terms of water, salt, and yield distributions in a sloping corn field.
| Materials and methods |
|---|
|
|
|---|
300 mm, which falls only in the winter. There was no rainfall during the experiment. The soil in the corn field is deep sandy loam loess (Calcic Haploxeralf). The soil texture is 15.7% clay, 18.2% silt, and 66.1% sand, and the dominant clay is montmorillonite. The soil contained 13.5% CaCO3 and had a cation-exchange capacity of 11.5 cmol kg-1 and total porosity of 42.7%. This soil type had previously been found to be sensitive to seal formation under rainfall and sprinkler irrigation conditions (Mualem and Assouline, 1991; Ben-Hur, 1994). The previous crop in this field was potato, and it was irrigated with fresh water [EC =
1 dS m-1 and SAR =
2.5 (mmol L-1)0.5]. The average gravimetric water content and the average EC in the saturation soil paste in the field after the potato harvest were 0.13 kg kg-1 and 1.5 dS m-1, respectively, in the 0- to 0.6-m soil layer, and 0.12 kg kg-1 and 0.95 dS m-1, respectively, in the 0.6- to 1.5-m soil layer. No trend of these two parameters along the slope was found. The field was plowed, disked, and leveled with a roller to provide a smooth, circular seedbed. The corn was planted in circular ridges on 10 July 1998 with a row spacing of 0.96 m and a population of 88900 plants ha-1. The field was routinely irrigated twice weekly with a center pivot MIS comprising six spans, each 60 m long. The irrigation was started on 12 July 1998 and continued through 24 Sept. 1998. The total irrigation amount during the entire growing season was 356 mm, with intervals of 3 or 4 d between the applications. The EC of the irrigation water was 4.7 dS m-1, and the SAR was 21.6 (mmol L-1)0.5. The soil was fertilized with NH4NO3 (200 kg N ha-1) applied with the irrigation water.
The experimental site was 120 m wide (below the second and third spans of the MIS) by 100 m long, and the long axis was a slope with a fairly constant gradient of
5%. The crop was planted in rows parallel to the long axis, and the MIS traveled along it. Each irrigation event began at the uphill position and moved in an arc downhill, while the runoff flowed along the long axis. At the upper boundary of the experimental site, a ditch was dug to prevent entry of surface runoff from outside the experimental site. Likewise, the ridges prevented runoff from crossing the plant rows into adjacent areas.
The experiment included three tillage practice treatments: (i) conventional tillage (control); (ii) microbasin; and (iii) diked furrows (dike) (Fig. 1) . In the control treatment, no other cultivation was conducted after ridging, and the runoff was allowed to flow freely downhill. In contrast, in the other two treatments, microbasins and dikes were constructed in the furrows after ridging using special machinery (Morin et al., 1984). Microbasins were constructed in furrows, and dikes in every second furrow. The depth of the microbasins was 0.15 m and that of the basins delimited by the dikes was 0.4 m (Fig. 1). In the microbasin and the dike treatments, runoff was trapped on the soil surface during the irrigation and infiltrated into the soil later. The treatment plot was 3.9 m wide by 100 m long and included four plant rows; the two central rows were used for sampling and the other two rows for borders. The control and microbasin plots were irrigated with sprinkler MIS, and the dike plots with flooding MIS.
|
For the flooding method, the third span in the MIS was equipped with low-energy precision application socks (LEPAS) spaced 2 m apart, which delivered the water directly to every second furrow with minimal loss. In this irrigation method, the application rate per wetted area was much higher than in sprinkler MIS. Irrigation by LEPAS under conventional tillage forms a large amount of runoff that could seriously damage the field. Therefore, the whole area under the flooding MIS was cultivated with dikes. Six plots were treated, between the 13th and the 25th LEPAS, and their discharges ranged from 635 to 738 L h-1. Three dike plots of the six were selected randomly for testing. The control treatment under irrigation with flooding MIS was not tested. It should be emphasized that all the plots in the various treatments received the same amount of water during the irrigation, although the application rate per wetted area in the LEPAS was higher.
Within each plot, in all the treatments, five subsampling sites were located 25 m apart along the slope. Crop yield was determined for each sampling site on the two central plant rows at the end of the growing season (18 Oct. 1998). In each sampling site, three plants from each row were cut by hand 3 cm above the soil surface, and the cob and canopy weights were determined after 48 h of drying at 60°C. Total Na and Ca contents in the plant canopy were determined by flame photometry for Na and atomic absorption spectrometry for Ca after HNO3 digestion.
Soil water content was determined to a depth of 1.5 m at 0.3-m intervals using neutron scattering method (503 DR Hydroprobe, CPN Co., Martinez, CA). Access tubes were installed in two of three plots of each treatment. These tubes were located in the center between two adjacent plant rows in each sampling site in all the treatments (Fig. 1).
Soil samples to a depth of 1.5 m at 0.3-m intervals were taken from each sampling site in two of the three plots of each treatment on 25 Oct. 1998. Each soil sample was taken from a point
1 m from the access tube in the downhill direction (Fig. 1) from which EC and Na and Ca concentrations in saturated soil paste were determined.
| Results |
|---|
|
|
|---|
|
Average gravimetric water content in different soil layers for the various treatments and at three sampling sites along the slope, as determined 1 d before the last irrigation (23 Sept. 1998), are presented in Fig. 3 . The patterns of the soil water content distribution on the various sampling dates were similar to those in Fig. 3. The sampling of 23 September was presented because it expresses the cumulative effect of the whole irrigation season and growing period on the soil water content. Three sampling sites, upslope, midslope, and downslope, were chosen to represent the effect of the slope on the uniformity of the soil water content.
|
40%. The higher soil water content at the downslope site in the control treatment (Fig. 3) was probably due to infiltration of downward flow of surface runoff generated further uphill. This runoff increased the infiltration duration in the downhill sites, which in turn increased the soil water content in this direction. In contrast, in the microbasin and dike treatments, the tillage practices prevented the flow of surface runoff downhill, which, in turn, ensured more uniform distribution of the soil water content along the slope. In the microbasin treatment, the water contents in the 0- to 0.6-m soil layer were higher than in the deeper soil layers for each sampling site along the slope (Fig. 3). The average water content of the 0- to 0.6- and 0.9- to 1.5-m soil layers at the three sampling sites for this treatment was 0.16 and 0.11 kg kg-1, respectively. This water content of the 0- to 0.6-m soil layer was significantly higher than in the upslope and midslope site of the control treatment.
In the dike treatment, the average water content of the 0- to 0.6- and 0.9- to 1.5-m soil layers for the upslope and downslope sites was 0.15 and 0.12 kg kg-1, respectively. In the midslope site, it was 0.15 kg kg-1 for the whole profile (Fig. 3). These results suggest that wetting extended deeper than in the microbasin treatment. The main reason is that the water application per wetted area of the flooding MIS applied to every second furrow in the dike treatment was much higher than that of the sprinkler MIS applied to the whole field surface in the microbasin one.
Average EC values in the saturated soil paste extractions from different soil depths, in various treatments and at three sampling sites along the slope are presented in Fig. 4 . The soil samples for these EC measurements were taken on 25 Oct. 1998 after harvest. Hence, the EC values in this figure represent the cumulative effects of the whole irrigation season and growing period on the salt distributions in the field for the different treatments. In the three treatments, significant salt accumulation was observed in the 0- to 0.6-m soil layer in most of the sampling sites along the slope (Fig. 4), with the largest accumulation obtained in the 0- to 0.3-m soil layer. In contrast, the EC values in the various treatments and sampling sites were practically similar for the 0.6- to 1.5-m soil layer, and the average value of these EC values was 0.9 dS m-1. This average value is similar to that measured in the field before starting the present study. According to Shalhevet et al. (1981), 70% of the water uptake by the corn plants in southern Israel conditions comes from the 0- to 0.6-m layer of the soil profile. Therefore, the accumulation of the salt should occur mostly in this layer, which is in agreement with the results presented in Fig. 4.
|
|
The mean EC of the 0- to 0.3-m soil layer in the dike treatment was higher than in the microbasin treatment (Fig. 5). The average EC at the five sampling sites along the slope was 3.4 dS m-1 in the microbasin treatment and 7.0 dS m-1 in the dike treatment. The main explanation for this difference in the EC lies in the higher water application per wetted area in the dike treatment than in the microbasin treatment, which led to greater salt load and salt accumulation in the upper soil layer.
Another factor that could affect the corn yield was the spraying of saline water over the plant canopy under irrigation with the sprinkler MIS. Concentrations of Na and Ca within the corn canopy for the various treatments are presented in Table 1 . Because no significant differences in these ion concentrations were found among the sampling sites along the slope, average ion concentrations for each treatment are presented in Table 1. The Na concentration within the corn tissue under flooding MIS irrigation (dike treatment) was 2.7 g kg-1 (Table 1). Because no direct contact occurred between the plant canopy and the irrigation water under this water application method, the Na concentration within the corn canopy was mainly a result of Na uptake from the soil solution by the plant roots. In a pot experiment with a silty loam irrigated with water of various qualities applied to the soil surface (with no direct contact between the plant canopy and the irrigation water), Bar-Tal et al. (1991) found that the Na concentration within the corn tissue was 0.5 g kg-1. The higher concentration of Na in the corn tissue under the flooding MIS irrigation in our study was probably a result of the higher concentration of Na in the soil solution in this treatment. The average Na concentration in the 0- to 0.3-m soil layer (the most active root zone) for the whole slope in the dikes treatment was 79.2 mol m-3 compared with 1.66 mol m-3 in the pot experiment with fresh water used for irrigation (Bar-Tal et al., 1991).
|
No significant differences in the Ca concentrations in the corn tissue were observed among the various treatments (Table 1). These results indicate that the absorption or adsorption of Ca through the corn canopy was minor, probably because of the low transportation of Ca in the plant (Hanson, 1982).
| Discussion |
|---|
|
|
|---|
Free flow of the runoff increased the infiltrated water amount and the salt load in the downhill sites in the control treatment. The net effect of this increase of the water and salt content in the downhill direction in this treatment was an increase of the corn yield (Fig. 2). The EC in the 0- to 0.3-m soil layer in the downslope site in the control treatment was 4.2 dS m-1 (Fig. 5), which is lower than the average EC (7.0 dS m-1) for the whole slope in the dike treatment. The salinity threshold for yield reduction of corn is 1.8 dS m-1 with decreasing rate of 7.4% dS m-1 (Bresler et al., 1982). Therefore, a harmful effect of the soil salinity on yield is expected to be larger in the dike treatment than in the control. However, the corn yield in the downhill site of the control treatment was significantly lower than the average yield in the dike treatment. The gravimetric water content at field capacity of 0- to 0.6-m soil layer in the experimental field is 0.16 kg kg-1. During most of the irrigation season, the gravimetric water content in the 0- to 0.6-m soil layer of the downslope site of the control treatment a day before irrigation, ranged from 0.17 to 0.20 kg kg-1, with an average value of 0.19 kg kg-1. In the other treatments, the corresponding gravimetric water content for all the sampling sites ranged from 0.13 to 0.16 kg kg-1, with an average value of 0.15 kg kg-1 in the microbasins, and from 0.14 to 0.16 kg kg-1, with an average value of 0.15 kg kg-1 in the dikes. Consequently, only in the downslope site in the control treatment, irrigation took place while the water content of the soil was above the field capacity. Therefore, the relative lower yield of the downslope site in the control treatment compared with the microbasin treatment could be explained by the effect of the high soil water content in this site, which probably led to low aeration in the root zone. Hence, under irrigation by sprinkler MIS with saline water and conventional tillage, a lower yield would be expected due to a reduced amount of available water in the root zone in the uphill area and the low aeration and accumulation of salt in the downhill area.
Storage of the saline water between the dikes in alternate furrows in the field irrigated using flooding MIS led to deeper water percolation (Fig. 3) and to greater salt accumulation in the 0- to 0.3-m soil layer (Fig. 4 and 5) than under microbasin tillage and irrigation with sprinkler MIS. The water percolating deeply below the effective root zone in the dike treatment was lost for plant production. Therefore, the combined effect of less available water and high EC in the root zone in the dike treatment led to a lower yield in this treatment than in the microbasin treatment. In the latter treatment, the irrigation water was sprayed over the whole soil surface in the field; this minimized the accumulation of the salts in the soil and the percolation of the irrigation water into the deep layers, which in turn led to a higher corn yield.
In spite of the high concentration of Na in the corn canopy under irrigation with sprinkler MIS (Table 1), the corn yield in this treatment was higher than under flooding irrigation (Fig. 2), where there was no direct contact between the saline irrigation water and the plant canopy. The relatively high concentration of Na accumulated in the plant canopy (
10.0 g kg-1 of the dry material, Table 1) was apparently less important than the combined effect of the water and salt regimes in the root zone that characterized each of the different tillage and water application practices.
| Summary and conclusions |
|---|
|
|
|---|
In all the tillage treatments, salt accumulation was observed in the 0- to 0.3-m soil layer. The mean EC of the 0- to 0.3-m soil layer (for the whole slope) in the control, microbasin, and dike treatments were 3.6, 3.4, and 7.0 dS m-1, respectively. The higher water application per wetted area in the flooding MIS of dike alternate furrows than in the control and microbasin treatments led to greater salt load and salt accumulation in the upper soil profile.
The tillage and water application practices affected the yield. The average dry-canopy yield of the corn (for the whole slope) in the control, dike, and microbasin treatments were 21.7, 25.3 and 30.6 Mg ha-1, respectively. In the control treatment, an increasing trend of the yield was observed in the downslope direction. In this treatment, the low yield of the upslope sites was mainly a result of the low water content in this area, which decreased the available water for the plants. In the microbasin and dike treatments, the uniformity of the yield along the slope was relatively high.
The lower yield in the dike treatment compared with that obtained in microbasin tillage was mainly a result of the greater salt accumulation in the upper soil layer in the former treatment. The Na concentration in the corn tissue under irrigation with flooding MIS was 25% of that obtained under sprinkler MIS. However, this difference in Na concentrations had a minor effect on the yield.
| ACKNOWLEDGMENTS |
|---|
| NOTES |
|---|
|
|
|---|
Received for publication January 4, 2000.
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
S. Assouline and M. Ben-Hur Effects of Water Applications and Soil Tillage on Water and Salt Distribution in a Vertisol Soil Sci. Soc. Am. J., May 1, 2003; 67(3): 852 - 858. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Ben-Hur and S. Assouline Tillage Effects on Water and Salt Distribution in a Vertisol during Effluent Irrigation and Rainfall Agron. J., November 1, 2002; 94(6): 1295 - 1304. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Crop Science | |||
| Journal of Natural Resources and Life Sciences Education |
Vadose Zone Journal | ||||
| Journal of Plant Registrations | Journal of Environmental Quality |
The Plant Genome | |||