50 Proceedings of The South African Sugar Technologists' Association — June 1969
THE EFFECT OF DIFFERENT METHODS OF FURROW
TABLE I
Available moisture in PE 1 sandy clay loam
Part I
The experiment consisted of twelve treatments, arranged as a 6 x 2 factorial, in 4 replications, plot size being 6 rows (5 ft. apart) x 290 ft. The following six systems of furrow irrigation were compared, starting in August, 1967.
R S Irrigated in cane row: standard practice.
R R Irrigated in cane row with the inter-row ridged up, to give a wide furrow.
RC Irrigated in cane row: standard practice with checkdams.
I T Irrigated in alternate inter-rows; a broad flat furrow with trash piled on alternate inter-row.
I B Irrigated in alternate inter-rows; trash burnt, tops re-burnt.
I N Irrigated every inter-row with narrow V-shaped furrows and ridges.
Tractors and mules were both used to achieve the furrow shapes desired; these are shown below.
Ripping to a depth of 18 in. was compared with non-ripping in all the above systems of irrigation.
Water was measured by means of a Parshall flume with recorder just before entering the experiment.
The flow was generaly about 1 1/3 cusecs and was diverted by shovel into three cane rows (or inter- rows). Water was cut off just before reaching the bottom of the furrows which were closed to prevent loss of tailwater. The quantity applied at each irriga- tion was thus a function of soil conditions and fur- row shape, since other factors (furrow length, grade and stream size) were held constant. Observation wells were installed in every plot to follow fluctua- tions in the ground water table.
Each plot was harvested separately using normal field methods, and cane weights and sucrose % cane were obtained from weighbridge data and mill samples.
Part II
After harvest at 13 months in September 1968, the experiment was converted to siphon irrigation by raising the feeders. A new layout was installed which included the following treatment in 4 repli- cations.
R S Irrigated in cane row (as in previous ratoon).
I W Inter-row wide furrow: contact width about 3 ft. (as in previous ratoon I B, but irrigated every row).
I M Inter-row medium furrow: contact width about 2 ft.
I N Inter-row narrow furrow: contact width about 1 ft. (as in previous ratoon).
After a number of preliminary irrigations, record- ing was commenced in January 1969.
Results: Part I Irrigation measurement
Several irrigations were carried out in September and October, 1967, during which time the treatments were being imposed and soil conditions stabilized.
Useful data were obtained from 10th November 1967 and continued up to 1st July 1968 after which the field was dried off for harvest. The data for total irrigation included all irrigation applied; how- ever the figures for mean application refer only to the period 10th November 1967 to 1st July 1968.
It was soon observed that far less water was applied on all treatments where irrigation was applied between the rows than where irrigation was applied down the row, presumably because of the lack of obstruction to flow caused by the cane stools. Con- sequently the inter-row treatments had to be irrigated more frequently than the row treatments.
The mean water application in inches per irriga- tion over 14 irrigations for the row treatments and 18 for the inter-row treatments as shown in Table 2.
TABLE 2
Effect of treatment on water application
The water applied at each irrigation for all the treatments irrigated down the row was far greater than for those irrigated in the inter-row. The appli- cation in the row treatments was clearly excessive, since the total available moisture in the effective rooting depth of 21 ft. is only 4 in. Using a 75%
depletion figure, it is considered desirable to apply 3 in. per irrigation on this soil. Where checkdams were employed (treatment R C) some 8 in. water was applied at each irrigation and at least half must have been wasted. Little difference was observed between the other in-row treatments, R S and R R.
52 Proceedings of The South African Sugar Technologists' Association — June 1969
The application in the inter-row treatments was below optimum, especially in the narrow V-shaped furrow of the I N treatment (1.7 in.). The treatments irrigated in alternate inter-rows (1 T and I B) also gave too low applications (1.5 in.). However if these had been irrigated every inter-row, the quantity applied, 3 in., would have been about optimum, and this treatment was clearly the best for the conditions in the experimental field.
No differences were observed in water application between the ripping and no-ripping treatments. This is perhaps surprising, but it would appear that rip- ping is unnecessary on this soil.
Water table
Readings from weekly well observations on depth of water table showed that there was no treatment effect in spite of the very large differences in water applied to each treatment. This was presumably due to lateral movement of water resulting in a uniform water table over the relatively narrow plots (6 rows).
Over the whole period of the experiment there was an average depth of about 3 ft. - 3 ft. 6 in. to the water table. Figure 2 shows the fluctuation in mean water table depth over the whole experiment from January to August 1968. When wells were dry, they
TABLE 3
Effect of treatment on yield and irrigation efficiency
FIGURE 2: Mean depth of water table in experiment.
were recorded as having a water table at 4 ft., since they were only installed to this depth and it was necessary to have a figure for calculating means.
Thus where the mean water table was 3 1/2 ft., a large number of wells were, in fact, dry.
Yield data
Yield data from the various treatments together with efficiency of water use expressed as tons cane/
acre-inch of irrigation water are shown in Table 3.
Considering the very large difference between the row and inter-row treatments in total water application, it is surprising that there was not a greater difference in yield between these treatments.
This is probably because the water table contributed a significant amount of water in the inter-row treat- ments. The efficiency of water use expressed as tons cane/acre-inch of water applied rose from 0.56 for the R C treatment to 1.7 for inter-row treatments, the latter figure also being inflated because of the effect of the water table. The figures of 0.56 to 0.82
for the row treatments should not have been affected, however, and reflect an inefficient use of water com- pared with figures of around 1 ton cane/inch of water generally obtained e.g. Thompson and Boyce1.
No difference between the sucrose contents of the various treatments could be detected, and the means of the row and inter-row treatments were identical (18.3%). Consequently yields of tons sucrose/acre followed identical trends with those of tons cane/
acre.
Ripping had ho significant effect on yield or sucrose content.
Results: Part II Water application
Preliminary studies indicated that the optimum pipe size for trickle flow in the narrow, medium and wide furrows was 3/4 in., I in. and 1 1/4 in. respectively giving flows of 6, 10 and 14 gal/min from a head of 6 in. For the flush flow it was found that 90 gal/min from 2 in. siphons was the largest flow which did not cause erosion. Mean applications over two irriga- tions in January-February 1969 are shown in Table 4.
The effect of the shape of the furrow on water application is well demonstrated; with a flush flow
ST-D
of 90 gal/min only 0.80 in. was applied on a narrow inter-row furrow, while 2.11 in. was applied on a wide furrow and 7.88 in. was applied irrigating down the cane row. Trickle flow was found to be imprac- ticable for this last treatment; in the other treatments one would allow the trickle flow to run for sufficient time to apply the desired quantity of water. This was achieved in the I M treatment, but too little was applied in I N treatment. This was later corrected by trickling for 4 hours instead of 3, which applied
1.84 in. instead of 1.38.
Discussion
Given an existing furrow layout with fixed grade, furrow length and row spacing, it is possible to vary the amount of water applied per irrigation within very wide limits by manipulation of furrow shape and by the use of trickle flow. It was consis- tently found in this experiment that irrigating down the cane row resulted in far greater (two to fourfold) applications of water than irrigation in the inter-row.
Considerable variation in application was also achieved by varying the shape and effective width of the inter-row furrow; a wide furrow resulting in a much greater application than a harrow furrow.
Applications greater than 4 in. were wasteful on the soil in question and the efficiency of utilization of irrigation (expressed as tons cane or lb. sucrose per acre-inch of water applied) was greatly increased by inter-row irrigation compared with in-row.
The preliminary data obtained from the experi- ment have highlighted the need for much further investigation of the effects of grade, furrow length and stream flow on total intake and uniformity of application down the row on various soil types.
Acknowledgement
Thanks are due to Mr. R. Yeatman for suggesting the experiment, for providing the land required and co-operation in management of the trial, and to the Chiredzi Research Station for carrying out the wilt- ing point determination.
Reference
1. Thompson, G. D. and Boyce, J. (1968). T h e plant crop results of two irrigation experiments at Pongola. Proc.
S.A. Sug. Tech. Assn. 42: 143.
TABLE 4
Water application in flush and trickle flow
Discussion
Dr. Thompson (in the chair): Furrow irrigation is not highly thought of in Natal but the reasons are mainly our topography, soil variations over small distances, and paucity of water. We therefore favour sprinkler rather than furrow irrigation on hillsides. But we have always recognised a place for furrow irrigation in the northern areas. Never- theless, I do see some difficulty in persuading farmers to measure water and to apply it evenly.
Dr. Gosnell: Those are certainly the main difli- culties of furrow irrigation but with various devices, e.g. gates, it is not too difficult to measure water application per field. Most water measurement in the Lowveld is done once only onto the farm but it is not an expensive process to install gates at each field and calibrate them to get an accurate measure of water.
Using siphon irrigation and trickle flow we should obtain far greater uniformity of application.
There will always be difficulties in achieving uni- formity from top to bottom of a furrow but this could be brought up to 75% by efficient design.
Mr. Boyce: Does Dr. Gosnell feel that in the North we should supplement the work he is doing in Rhodesia?
Dr. Gosnell: I was amazed when 1 started how little work has been done by anyone on furrow irrigation, so any work done at Pongola or else- where will be most valuable.
Dr. Thompson: To what extent will you be able to extrapolate your results into the field?
Dr. Gosnell: A lot of the work cannot be done as statistical experiments and must be clone as ex- ploratory trials on farms.
This particular experiment was done on a settler's farm as we did not have suitable land available.
Mr. Moberly: Dr. Gosnell seemed to get very high sucrose figures. Could he please explain the drying-off techniques?
Dr. Gosnell: The drying-off process was simply to close off water.
T h e figures for sucrose are higher than any others we obtained on the Experiment Station and are somewhat puzzling as in some plots there was a water table right through the harvest.
Mr. du Toit: Was the application of water, 8"
per irrigation, in the row and, if so. as the soil could only hold 4", what happened to the rest?
Dr. Gosnell: It went to the water table.
Dr. Thompson: With in-row irrigation, the crop did not have to resort to the water table because the applications were in excess of its requirements.
But with inter-row irrigation these applications were probably less than the crop requirements so
it almost certainly did resort at times to the water table.
Mr. Rogers: How was rainfall accounted for?
Dr. Gosnell: During the crop 10" of rainfall was measured: it was not included in the calculations, although tons cane per inch total water was cal- culated.
Mr. Ranger: A row length of 290 feet is short and I think this could be extended. By irrigating in the inter-row, putting a large volume of water down and choosing the correct furrow shape it should be possible to double or treble the length which is irrigated, still applying three inches.
Dr. Gosnell: In the new ratoon experiment we have done this and there are furrows 300, 600 and 900 feet long.
On this particular soil we cannot irrigate 900 feet as the water application is too high—well over ten inches.
Mr. Ranger: The diameter of the siphon, one inch, seems small as we use three inches.
Dr. Gosnell: For the flush flow we use 2-2 1/2 inch siphons but for the trickle flow only three quarter, one or one and a quarter inches.
Dr. Thompson: In Hawaii the longest furrows were 225 feet and they were worried about distri- bution.
Mr. Ranger: Our soil conditions at Nambala are probably different from Hawaii and our basic infil- tration rates are comparable to the figures men- tioned by Dr. Gosnell.
Our row lengths for irrigation are 1,200 feet and grades are 1 in 400 to 1 in 500. Average applica- tions, measured through flumes, are about two and a half inches net, assuming an irrigation efficiency of 75 per cent.
Dr. Hill: I think the main differences are due to infiltration rates.
Dr. Gosnell: Yes, our soils are sandy loams to sandy clay loams and the mean infiltration rates are probably much higher than the soils of Nakambala, which are very heavy.
Mr. Wilson: I think it would be easier to apply modern irrigation systems to the newer areas rather than to the established areas so that the labour force can be trained correctly from the start.
Dr. Gosnell: It is possible to convert to the modern systems and we have done so in a number of fields.
A labourer previously handling a third to a half a cusec is now handling one to one and a quarter.
Some maize farmers in Rhodesia are handling two and a half cusecs per labourer.
Mr. Andries: We must not overlook the import- ance the soil plays in the choice of system and length of line. It is not possible to generalise.
54 Proceeding of The South African Sugar Technologists' Association - June 1969