Irrigation water management for cotton on Vertisols in the middle Awash region of Ethiopia

G. Haider, Tilahun Hordofa and Endale Bekele

Melka Werer Research Centre Institute of Agricultural Research (JAR)
PO Box 2003, Addis Ababa, Ethiopia

Abstract
Introduction
Materials and methods
Results and discussion
Conclusions
Acknowledgements
References

Abstract

Improved water management practices for cotton, planted on Vertisols in the Middle Awash region of Ethiopia, were developed at Melka Werer Research Centre of the Institute of Agricultural Research.

The optimum irrigation furrow length on Vertisols is 200 m. The optimum initial stream flow rate for a furrow of this length with a slope range of 0.005-0.008% is 3.5 litres sec^-1 with a cut-back stream flow rate of 1.5 litres sec^-1: for a slope of about 0.015%, the optimum initial stream flow rate is 2.13 litres sec^-1 and cut-back stream flow rate 1.61 litres sec^-1.

The recommended irrigation schedule for cotton includes irrigations of 75 mm at 2-week intervals or 125 mm at 3-week intervals. However, according to a computed irrigation schedule for cotton planted on 15 May on Vertisols (available water: 220 mm m^-1), the crop needs one irrigation of 60 mm at establishment, two irrigations of 75 mm each at 12-day intervals during the vegetative stage, four irrigations of 105-110 mm each at 2-week intervals during yield formation, and one irrigation of 140 mm towards ripening, in addition to 150 mm at planting. For maximum crop production, 75-mm irrigations should be applied at 2-week intervals in addition to 200 mm at planting.

To optimise production per unit of water applied, three irrigations, one of 200 mm at planting, followed by two of 150 mm each at flowering and boll formation, are adequate to obtain a yield similar to that obtained with 75-mm irrigations at 2-week intervals, provided rainfall is normal and well distributed during the season. This will save 40-50% of the irrigation water. Cotton can extract considerably more water from the lower soil depths when irrigated at 4-week intervals than when irrigated at 2-week intervals.

Introduction

The Middle Awash region of Ethiopia is in the semi-arid climatic zone with a long hot summer and a short mild winter. Annual rainfall amounts to 200-500 mm. Irrigation is therefore vital to ensure crop production.

Accurate information on soils in the area is sparse. However, a soil survey of 200 ha at the farm in Melka Werer Research Centre showed that Vertisols constitute about 70% of the farm area.

In this area, cotton is the main crop and is grown on about 13 000 ha, but wheat is gaining popularity. Soil, water and climatic conditions are suitable for growing many other lowland crops such as maize, groundnut, sesame, kenaf, fruits and vegetables.

Mismanagement of irrigation on poorly drained Vertisols under semi-arid conditions could lead to waterlogging and soil salinisation. These problems have resulted in 30% of the 3500-ha Melka Sadi State Farm being seriously affected by waterlogging and salinity after only 15-20 years.

Melka Werer Research Centre (MWRC) of the Institute of Agricultural Research (JAR) has been engaged in research to develop improved water management practices. Research topics include evaluation of stream size in relation to furrow length, determination of optimum irrigation frequency and irrigation depth, soil moisture extraction patterns by different crops, and evaluation of the effect of drought stress on growth and yield of important crops. This paper summarises the results of some of the experiments conducted at MWRC for developing efficient water management practices for cotton grown on Vertisols in the Middle Awash region of Ethiopia.

Materials and methods

Chemical and physical characteristics of Vertisols

To determine chemical and physical characteristics of Vertisols, a representative soil profile to 2 m was described and soil samples taken. The infiltration rate of the soil was measured near the sampling site using the standard double ring method. Field capacity and permanent wilting point were determined using a pressure membrane apparatus.

Soil moisture samples were taken separately from 19 different locations representing Vertisols at the MWRC and the Amibara Irrigation Project. Field capacity and permanent wilting point of each sample were determined. Soil density measurements were taken at each of the 19 sites and available soil moisture was calculated. Regression analyses of clay content on field capacity on permanent wilting point were carried out.

Stream flow rate and furrow length evaluation

Six alternate furrows, each 250 m long, were selected on Vertisols at MWRC, Melka Sadi and Amibara State Farms. Furrow spacing was 80-90 cm, depth was 25-40 cm, field slope was 0.005-0.015%.

Furrows were staked at 25-cm intervals along the entire 250-m length. Six stream flow rates ranging from 0.60 to 4.42. litres sec^-1 were tested. Flow rates were calibrated and measured volumetrically before the test. As water was led into each furrow, the time taken for the water to reach each station was recorded, and Parshall flumes were used to measure flow during the test. Each furrow was inspected for overflow or erosion, out-flow at the end of each furrow was observed, and excess flow was cut back and the cut-back flow rate recorded. At the end of the test, the time for the water to recede at each station was recorded. Forty-eight hours after the test, water distribution at the upper, middle and lower end of each furrow was estimated from soil samples.

Quantities of water needed to refill the soil moisture reservoir to field capacity were computed, and the time necessary to refill the soil moisture reservoir was determined using the water intake rate measured at each site. Stream advance time and total irrigation time were determined. From the advance rate, optimum stream flow rate and optimum furrow length for each stream flow rate were determined. Water distribution and irrigation application efficiency were recorded.

Irrigation frequency and depth

Cotton varieties Acala 1517/70 and Acala 1517/70C were planted as test crops. Different irrigation frequencies and irrigation depths were studied. Two or three irrigations during the crop establishment stage were common to all treatments, after which differential irrigation frequencies and depths were applied to the respective treatments. Each treatment was replicated at least four times. Irrigation applied was supplemented to rainfall during the season (Table 1). Each irrigation application was measured and seasonal irrigation applications and rainfall were recorded. All other cultural practices were standard and common to all treatments. The characteristics of the Awash river water used for irrigation are given in Table 2.

Computed irrigation schedule

The irrigation schedule for cotton planted on Vertisols (available water: 220 mm m-1) in the Middle Awash area was calculated for 15 May planting. Growth stages, root depth and distribution, crop water requirements, and irrigation application at 60% depletion of available soil moisture were considered.

Drought stress vs yield

Six previously screened stress-tolerant cotton varieties were tested during 1983-1986. Different levels of drought stress were imposed by the four treatments:

A. One 200-mm irrigation at planting;

B. One 200-mm irrigation at planting and a second irrigation of 150 mm at peak flowering;

C. One 200-mm irrigation at planting and two subsequent irrigations of 150 mm each at peak flowering and boll formation; and

D. One 200-mm irrigation at planting and nine subsequent irrigations of 75 mm each at 2-week intervals up to 126 days after planting (control treatment).

Table 1. Rainfall during cotton growing season.

Month	Rainfall (mm)
	1968	1969	1974	1975	1983	1984	1986
May 15-30	1.0	17.8	25.6	0.0	30.9	70.3	53.1
June	55.8	6.1	53.8	57.0	17.8	33.7	59.1
July	130.7	75.0	151.1	186.0	107.5	144.6	107.4
August	145.8	84.6	28.8	169.0	149.2	63.3	67.1
September	46.4	18.0	72.8	57.4	24.4	70.0	114.6
October	12.0	0.4	4.2	4.0	12.2	0.0	5.2
November	53.1	0.8	0.0	0.0	0.0	2.0	0.0
Total	444.8	202.7	336.3	473.4	342.0	383.9	406.5
Effective rainfall (70%)	311.4	141.9	235.4	331.4	239.4	268.7	284.6

Table 2. Characteristics of Awash river water at Melka Werer (1985).

Time of sampling	pH	EC (mmhos cm-1)	CO32-	HCO3-	Cl-	Ca2+ and Mg2+	Na+	K+	SAR
July	8.1	0.20	0.24	0.24	0.22	0.46	1.09	0.15	2.27
October	8.5	0.70	0.40	0.76	1.34	0.58	6.03	0 31	11.17

These four treatments were replicated four times in a randomised complete block design. Irrigation applied was supplemental to rainfall during the season (Table 1). All other cultural practices were standard and common to all treatments.

Results and discussion

Physical and chemical characteristics of Vertisols

A typical Vertisol is characterised in Table 3. From data in Table 4, the regression analysis of clay content (C) on field capacity (FC) on permanent wilting point (POOP) gave the following equations:

FC = 35.74 + 0.37 C,	r = 0.76
PWP = 12.42 + 0.41 C,	r = 0.74

The predicted and measured values matched reasonably well. The above regression equations are valid under the following limiting conditions:

· clay content of 40-84%,

· textural group of clay loam to clay, and

· negligible organic matter content.

Table 3. Chemical and physical characteristics of a Vertisol.

Depth (cm)	ECe pH (mmhos cm-1)		Colour	Clay (%)	Structure	Water content		Bulk density (g cm-3)	Available water (mm m-1)
						Fca (%)	PWPb (%)
0-32	8.5	1.8	10YR 3/2	69.5	Sub-angular blocky	53.9	36.4	1.31	229.9
32-48	8.2	1.7	10YR 4/2	65.5	Sub-angular blocky	56.5	35.1	1.44	308.6
48-83	8.1	1.2	7.5YR 3/2	67.5	Prismatic	55.5	35.8	1.38	271.9
83-140	8.0	3.5	10YR 2/2	63.5	Sub-angular blocky	53.3	35.2	1.45	262.0
140-200	7.3	4.5	10YR 6.3	47.5	Massive	48.4	32.1	1.27	207.1

a. Field capacity.
b. Permanent wilting point.

Source: G. Haider, Endale Bekele and Tilahun Hordofa (JAR, Addis Ababa, Ethiopia, unpublished data).

Table 4. Moisture characteristics of 19 Vertisols

Soil No.	Clay (%)	Bulk density (g cm-3)	Water content		Available water (mm m-1)
			Fca (%)	PWPb (%)
1.	64	1.2	56.53	34.24	222.9
2.	48	1.3	53.89	32.42	214.7
3.	40	1.3	51.26	28.61	226.5
4.	76	1.15	65.64	43.53	221.1
5.	80	1.15	64.20	41.52	226.8
6.	54	1.25	60.40	40.64	197.6
7.	60	1.25	65.98	47.04	189.4
8.	52	1.25	55.48	32.73	227.5
9.	58	1.25	62.89	42.01	200.8
10.	48	1.30	52.53	29.38	231.5
11.	38	1.35	42.46	20.95	213.1
12.	56	1.25	53.76	33.63	201.3
13.	78	1.15	65.04	40.88	241.6
14.	72	1.15	56.44	35.48	209.6
15.	56	1.25	52.36	32.30	200.6
16.	46	1.30	58.01	35.58	224.3
17.	52	1.25	57.56	37.33	202.3
18.	62	1.20	52.66	31.25	214.1
19.	84	1.15	68.23	53.50	247.3

a. Field capacity.
b. Permanent wilting point.

Source: Kandiah, JAR, Addis Ababa, unpublished data.

Stream flow rate and furrow length

Soil moisture characteristics at three selected sites are given in Table 5. Optimal stream flow rates and furrow lengths at three sites are given in Table 6. In general:

· a furrow length of 200 m is optimum for Vertisols;

· with a furrow slope range of 0.005-0.008%, the optimum initial stream flow rate for a 200-m long furrow is 3.5 litres sec^-1 with a cut-back stream flow rate of 1.5 litres sec^-1

· for a field slope of about 0.015%, the optimum initial stream flow rate is 2.13 litres sec^-1 with a cut-black stream flow rate of 1.61 litres sec^-1; and

· the application efficiency is around 70% (Kandiah, 1981).

Irrigation frequency and irrigation depth

The first irrigation frequency trials were conducted during 1968 and 1969. The crop was irrigated at 2-, 3- and 4-week intervals. There were no significant differences in yield among the irrigation interval treatments (Table 7). However a consistently higher yield (combined seed and fibre) was obtained when the crop was irrigated every 2 weeks. The higher yield levels for 1968 could be due to higher rainfall and a better rainfall distribution (Table 1). Plants that were irrigated at 4-week intervals had deeper roots that extracted considerably more water from the lower soil depths than plants that were irrigated at 2-week intervals (Table 8) (MWRS, 1968; 1969).

A second trial was carried out in 1974 and 1975. Four irrigation frequencies (2-, 3-, 4- and 5-week intervals) and three amounts of irrigation water (75, 125 and 175 mm) were replicated four times in a split-plot design. Irrigation at 2-week intervals during 1974 gave a significantly higher yield than irrigation at 3-,4- and 5-week intervals (Table 9). The difference in yield between 4- and 5-week intervals was not significant. During 1975, irrigating every 2 or 3 weeks resulted in significantly higher yields than irrigating at 4- or 5- week intervals. In 1974 plants that received 175-mm irrigations gave higher yields than plants which received 125- and 75-mm irrigations (Table 9). In 1975, irrigation depth had no significant effect on yield. Interaction between irrigation frequency and irrigation depth was not significant in either year. However, water use efficiency was greatest when a 75-mm irrigation was applied at 2-week intervals or when a 125-mm irrigation was applied at 2- or 3-week intervals.

Table 5. Soil moisture characteristics of selected Vertisol sites.

Characteristics	Location I	Location II	Location III
Slope (%)	0.008	0.015	0.005
Water content at field capacity (%)	42.0	42.0	42.0
Bulk density (g cm-3)	1.2	1.2	1.2
Average initial soil moisture (%)	27.6	27.34	27.44
Water required to refill 75 cm of root zone (cm)	12.06	13.20	13.10
Refill time (hours)	3.99	3.62	4.65
Stream advance time (hours)	1.00	0.90	1.16
Total irrigation time (hours)	4.99	4.52	5.81

Table 6. Optimum stream flow rates and furrow lengths.

Characteristics	Location I	Location II	Location III
Optimum initial stream flow rate (litres sec-1)	3.44	2.34	3.50
Cut-back stream flow rate (litres sec-1)	0.72	1.72	1.42
Optimum furrow length (m)	205.00	190.00	210.00
Application efficiency (%)	97.00	78.00	62.00

Table 7. Cotton yield (combined seed and fibre) for different irrigation intervals.

Irrigation interval (week)	Yield (t ha-1)
	1968	1969
2	2.15	1.72
3	2.02	1.68
4	1.94	1.57

None of the values were significant at the P=0.05 level.

Table 8. Soil moisture extraction pattern.

Soil depth (cm)	Percent moisture depletion
	2-week intervals		4-week intervals
	1968	1969	1968	1969
0-30	64.5	60.8	46.5	44.1
30-60	20.7	23.2	30.0	29.7
60-90	14.8	16.0	23.5	26.2

Source: IAR/MWRS (1968; 1969).

The higher yields in 1975 could have been due to well-distributed higher rainfall (Table 1), which could also be responsible for the lack of yield differences among the treatments during 1975 (IAR/MWRS 1974; 1975).

Effective rainfall during any irrigation interval should be deducted from the irrigation quantity in the computed schedule shown in Table 10 (Haider, IAR/MWRC, Addis Ababa, unpublished data).

Drought stress vs yield

Cotton yield was affected significantly by different drought stress treatments. Yields from treatments C and D were significantly higher than from treatments A and B (P=0.05) (Table 11). Differences in yields from treatments C and D were non-significant in two out of the three seasons. Treatment A gave a significantly lower yield than the other treatments except in 1987 when the differences in yield from treatments A and B were not significant. Higher yields in 1987 could have been due to higher and better distributed rainfall during the season. Treatments C and D consistently used water more efficiently than treatments A and B.

Table 9. Effect of irrigation interval and depth on cotton yield.

Year	Mean cotton yield (t ha-1)
	Irrigation interval (weeks)a				Irrigation depth (mm)a
	2	3	4	5	75	125	175
1974	5.70a	4.97b	3.64c	3.43c	3.73a	4.41b	5.16c
1975	6.41a	6.42a	6.06b	6.19b	6.39a	6.12a	5.90a

a. Within factors and years, values followed by the same letter are not significant at the P=0.05 level.

Table 10. Computed irrigation scheme for different growth stages of cotton. Total quantity of irrigation water is 930 mm.

	Planting	Establishment	Vegetative stage		Yield formation stage				Ripening
Irrigation interval (days)	0	21	12	12	14	14	14	14	25
Cumulative days after planting	0	21	33	45	59	73	87	101	126
Irrigation quantity (mm)	150	60	75	75	110	110	105	105	140

Source: Haider, IAR/MWRC, Addis Ababa, unpublished data.

Table 11. Cotton yield for different treatments.

Treatment	Number of irrigations	Mean cotton yield (t ha-1)a
		1983	1985	1987
A	1	0.91c	0.47d	3.56b
B	2	1.83b	0.72c	3.88b
C	3	2.83a	1.11b	4.26a
D	10	3.13a	3.87a	4.32a

a. Values in the same column followed by the same letter are not significantly different at the P=0.05 level.

Source: IAR/MWRC, Addis Ababa, Ethiopia, unpublished data.

To maximise cotton yield, nine irrigations of 75 mm each should be applied at 2-week intervals up to 126 days after planting, in addition to one 200-mm irrigation at planting. However, if irrigation water is inadequate, three irrigations of 200, 150 and 150 mm at planting, peak flowering and boll formation are enough to obtain reasonably good yields. If rainfall is well distributed during the season, then yields comparable to the treatment of 10 irrigations can be expected from the three irrigations, in addition to reducing water use by 40% (IAR/MWRC, Addis Ababa, Ethiopia, unpublished data).

Conclusions

The clay content of Vertisols in the Middle Awash area is 40-84% with an available soil moisture range of 180-250 mm m^-1. There are positive relationships between soil clay content and field capacity and permanent wilting point.

The optimum irrigation furrow length for Vertisols is 200 m, with an optimum initial stream flow rate for a slope range of 0.005-0.008% of 3.5 litres sec^-1, with a cut-back stream flow rate of 1.5 litres see. For a furrow slope of about 0.015%, the optimum initial stream flow rate is 2.13 litres sec^-1 with a cut-back stream flow rate of 1.61 litres sec^-1.

The recommended irrigation schedules for cotton in the Middle Awash region are: 75-mm irrigations at 2-week intervals, or 125-mm irrigations at 3-week intervals.

According to the computed irrigation schedule, the crop needs one irrigation of 60 mm during the crop establishment stage, two irrigations of 75 mm each, 12-days apart, during the vegetative stage, four irrigations of 105-110 mm each at 2-week intervals during yield formation, and one irrigation of 140 mm towards the ripening stage.

For maximum production, cotton should be irrigated at 2-week intervals with 75 mm of water. However, to optimise the production per unit of water, three irrigations, one of 200 mm at planting, followed by two of 150 mm each at flowering and boll formation, are adequate to obtain a reasonably good yield. This yield would be similar to that from 75-mm irrigation applications at 2-week intervals, provided rainfall is normal and well distributed during the growing season, and will also save 40% of the irrigation water.

Acknowledgements

The authors are grateful to all staff members at the Melka Werer Research Centre who were involved in carrying out the work reported in this paper.

References

IAR/MWRS (Institute of Agricultural Research/Melka Werer Research Station). 1968. Progress report for the period April 1967 to March 1968. JAR, Addis Ababa, Ethiopia.

IAR/MWRS (Institute of Agricultural Research/Melka Werer Research Station). 1969. Progress report for the period April 1968 to March 1969. JAR, Addis Ababa, Ethiopia.

IAR/MWRS (Institute of Agricultural Research/Melka Werer Research Station). 1974. Progress report for the period April 1973 to March 1974. JAR, Addis Ababa, Ethiopia.

IAR/MWRS (Institute of Agricultural Research/Melka Werer Research Station). 1975. Progress report for the period April 1974 to March 1975. JAR, Addis Ababa, Ethiopia.

Kandiah A. 1981. Evaluation of furrow irrigation system for cotton. ETH/82/004 Project Bulletin No. 2. IAR (Institute of Agricultural Research Addis Ababa, Ethiopia.