SURVEY OF IRRIGATION SCHEDULING PRACTICES IN

THE SOUTH AFRICAN SUGAR INDUSTRY

F OLIVIER and A SINGELS

South African Sugar Association Experiment Station, P/Bag X02, Mount Edgecombe, 4300,
South Africa. E-mail: [email protected]

Keywords: sugarcane, irrigation scheduling, survey, irrigation scheduling tools, irrigation

system, water use efficiency, neutron water meter

Introduction

Effective irrigation scheduling is a prerequisite to optimising the use of expensive irrigation

water for sugarcane production in South Africa. Accurate irrigation scheduling is not widely
practised in the sugar industry despite the many scheduling tools available to sugarcane
growers. This leads to very low water use efficiencies (WUE) of approximately 50% of what
could be achieved theoretically (Olivier and Singels, 2003). Irrigation scheduling can be
defined as a programme of irrigation determining the amount of water and timing of
application. Schedules can be classified as fixed (amount and cycle fixed for entire growing
season), semi-fixed (amount and cycle are changed a few times to accommodate rainfall and
significant seasonal and crop age induced changes in water demand) or flexible (amount and
cycle are changed daily or weekly according to calculated water budget based on recent crop
and weather conditions). Scheduling methods aim to maintain the soil water content in the
optimal range by direct monitoring of soil water content or by estimating the soil water
content through water budgeting.

The objectives of this study were to determine (i) which irrigation scheduling methods and
tools are being used within the industry and to what extent, (ii) reasons for adoption or
non-adoption of specific scheduling methods and (iii) practical limitations that prevent the
successful implementation of scheduling techniques. The information will be used to identify
priorities for future research and technology transfer to improve irrigation water use
efficiency.

Methods

Interviews were conducted with 10 growers in each of the major irrigation areas of Malelane,
Komatipoort, Pongola and Mtubatuba.

A comprehensive questionnaire was completed that covered the following aspects:

General farm information (irrigation system, system design and maintenance, water
source and cost and computer skills).
Scheduling practices (scheduling methods, scheduling tools used and reasons for use).
Perceptions (irrigation scheduling, advice and training).

South African Sugar Association Experiment Station Extension Officers also completed a
shortened version of the questionnaire to summarise the situation in each extension area. An
Excel spreadsheet was used to run queries and analyse the data sets of both the individual and
area questionnaires. Due to the small sample size of the study (Table 1), results from
individual questionnaires (irrigation systems and scheduling methods and instruments) were

239 Proc S Afr Sug Technol Ass (2004) 78

compared with that of the area questionnaires to confirm trends.

Table 1. General information obtained from the questionnaires. Irrigation and crop
water requirements were calculated using the Canesim crop model (Singels et al., 1998).
Survey area Industry
Komatipoort Malelane Pongola Mtubatuba totals
Number of interviews
Commercial 9 10 9 10 38
Small-scale − − − − −
Number of growers
Commercial 93 102 160 101 456
Small-scale 836 452 375 - 1 663
Irrigated cane (ha)
Commercial 15 899 18 442 16 009 6 318 56 668
Small-scale 5 734 3 331 2 121 - 11 186
Average farm size (ha)
Commercial 574 606 300 220 425
Small-scale 7 7 10 10 9
Average yield (t/ha)
Commercial 105 97 103 95 100
Small-scale 80 70 70 65 71
Net irrigation
requirement 970 970 840 660 −
(mm/annum)
Crop water
requirement 1300 1300 1200 1100 −
(mm/annum)
Water quota 995 1300 1000 No limit −
(mm/annum) (Komati) (Crocodile) (Pongola) (Umfolozi)
1300 1250 1000
(Water source in (Crocodile) (Noord- (Hluhluwe)
brackets) 850 Kaap)
(Lomati)
Water and electricity
cost (R/ha/annum) 1640 1667 3048 632 −

Results

Although the questionnaires covered a vast number of issues related to irrigation scheduling,
this paper highlights only the most relevant findings (Table 2).

Overhead dragline irrigation systems are used the most (67%), followed by drip irrigation
(18%), centre pivot (12%), flood (3%) and floppy (1%), (Table 2). The Komatipoort area has
less dragline irrigation than the other areas. Southern irrigated areas (Pongola and Mtubatuba)
have less drip irrigation (18%) than the northern irrigated areas (Komatipoort and Malelane).
These results were confirmed by the area survey. With larger farms and higher yield
potentials in the northern irrigated areas (Table 1), growers are investing in automated
irrigation systems such as drip.

The type of irrigation system being used largely determines the way irrigation is scheduled.
Dragline irrigation systems are mostly (73%) irrigated according to a semi-fixed irrigation
schedule, whereas centre pivot and drip systems are mostly (71%) scheduled according to
flexible schedules (Table 2). Current dragline system practices suggest that flexible irrigation
scheduling is impractical, due mainly to labour constraints and design limitations. In 95% of
cases, schedules are adjusted for significant rainfall (>30 mm), 92% for season (winter and
summer) and 55% for crop age.

Proc S Afr Sug Technol Ass (2004) 78 240

Flexible schedules are used on 80% of centre pivot irrigation systems and 64% of drip
irrigation systems (Table 2). Of the growers that make use of flexible schedules, 85% prefer
direct soil water content monitoring and 15% prefer water budget calculations based on crop
water use data (Figure 1).

Table 2. Number of growers (percentage given in brackets) that schedule irrigation

according to a fixed, semi-fixed or flexible irrigation schedule
for different irrigation systems and areas.
Irrigation Survey area Industry totals
system Komatipoort Malelane Pongola Mtubatuba
Flood 0 0.25 (0.7%) 0.34 (0.9%) 0.4 (1.1%) 0.99 (2.6%)
-Fixed - 0.25 (0.7%) 0.08 (0.2%) - 0.33 (0.9%)
-Semi-fixed - - 0.26 (0.7%) 0.4 (1.1%) 0.66 (1.7%)
-Flexible - - - - -
Dragline 4.18 (11.0%) 6.53 (17.2%) 7.21 (19.0%) 7.46 (19.7%) 25.38 (66.8%)
-Fixed 1.15 (3.0%) 2.04 (5.4%) - 3.60 (9.5%) 6.79 (17.9%)
-Semi-fixed 3.03 (8.0%) 4.49 (11.8%) 7.21 (19.0%) 3.86 (10.2%) 18.59 (48.9%)
-Flexible - - - - -
Floppy 0.03 (0.1%) 0.04 (0.1%) 0 0.22 (0.6%) 0.29 (0.8%)
-Fixed - - - - -
-Semi-fixed 0.03 (0.1%) 0.04 (0.1%) - 0.22 (0.6%) 0.29 (0.8%)
-Flexible - - - - -
Centre pivot 2.36 (6.2%) 0.18 (0.5%) 1.38 (3.6%) 0.78 (2.1%) 4.70 (12.4%)
-Fixed - - - - -
-Semi-fixed 0.57 (1.5%) - - 0.39 (1.0%) 0.96 (2.5%)
-Flexible 1.79 (4.7%) 0.18 (0.5%) 1.38 (3.6%) 0.39 (1.0%) 3.74 (9.8%)
Drip 2.43 (6.4%) 3.00 (7.9%) 0.07 (0.2%) 1.14 (3.0%) 6.64 (17.4%)
-Fixed - - - - -
-Semi-fixed 0.41 (1.1%) 1.36 (3.6%) - 0.60 (1.6%) 2.37 (6.2%)
-Flexible 2.02 (5.3%) 1.64 (4.3%) 0.07 (0.2%) 0.54 (1.4%) 4.27 (11.2%)

Total: 9 (23.7%) 10 (26.3%) 9 (23.7%) 10 (26.3%) 38 (100%)

The neutron water meter is used the most (75%) for direct soil water measurement, followed
by tensiometers (9%) and the wetting front detector (1%) (Figure 1). Consultants mostly
make use of a neutron water meter to provide a scheduling service. Growers are willing to
pay for such a service. The neutron water meter is accurate, relatively easy to use and
requires little maintenance compared with tensiometers. It should however be noted that the
initial calibration of the neutron water meter and interpretation of data can be complicated
and time consuming.

Water budgeting requires evapotranspiration (ET) data (mostly obtained from automatic
weather stations) and crop water use coefficients (Singels et al., 1999). Real-time ET data is
used more (10%) than long-term ET data (5%) (Figure 1). Computer availability (84%) and
internet access (74%) are not limiting factors. However, appropriate crop coefficients are not
readily available. Crop factors, meant for use with A-pan evaporation figures (George, 1988),
are wrongfully being applied to Penman-Monteith reference ET figures. Appropriate crop
coefficients have been calculated for all irrigated areas (Olivier and Singels, 2001) and will
soon be available.

241 Proc S Afr Sug Technol Ass (2004) 78

 5%

10%

Neutron water meter

1% Tensiometer
Wetting front detector
Real-time ET
9% Long-term ET

75%

Figure 1. Percentage of growers with flexible irrigation schedules that make use of
direct monitoring of soil water content (neutron water meter, tensiometer or wetting
front detector) or by soil water budgeting (real-time or long term
evapotranspiration (ET)).

Budgeting calculations are mostly performed on in-house developed spreadsheets. Generally,

growers are not willing to spend more than one hour per day on irrigation scheduling. Water
budget calculations are seen as too much trouble. Only one grower interviewed is using a
crop model for irrigation scheduling. Crop models are perceived to be too complex and
difficult to use.

Effects of age of individual growers, farm size, average yield, irrigation cost and irrigation
demand on irrigation scheduling strategy were also investigated. No clear trends could be
observed.

Although most growers realise the importance of accurate irrigation scheduling, they are
often unable to successfully apply the scheduling information in the field. This is mainly due
to the inflexibility of the dragline irrigation systems, problems with labour, theft, farm
security and complexity of the farming enterprise (more than one kind of crop). In many
instances, dragline systems are not capable of applying the required irrigation amounts and
growers are consequently forced to give up on their efforts to schedule irrigation accurately.

Conclusions

Key findings from this work were:

The dragline irrigation system is used by most growers (67%), followed by drip (18%)
and centre pivot (12%) systems.
Scheduling practices are highly dependent on the irrigation system. Semi-fixed
scheduling is predominant in dragline systems, while flexible methods are used the most
in centre-pivot and drip systems.
It is estimated that over 12 000 ha of dragline irrigated sugarcane could potentially be
converted from fixed to semi-fixed scheduling, while 6000 ha of centre pivot and drip
irrigated sugarcane could be converted from semi-flexible scheduling to flexible
scheduling. This will bring about considerable improvement in irrigation WUE.

Proc S Afr Sug Technol Ass (2004) 78 242

 Direct soil water content monitoring is favoured (85%) above water budget calculations
based on crop water use data (15%). The neutron water meter is the most popular method
of soil water content monitoring, followed by tensiometers.

Recommendations

Attempts should be made to promote a change from fixed to semi-fixed scheduling for
dragline systems. Tables with appropriate cycle and stand times could be generated by
applying crop models to historic climate data.

More centre pivot and drip irrigation systems should be converted to flexible schedules.
Irrigation scheduling based on water budgeting could be promoted by providing simple, easy
to use spreadsheets or crop models and appropriate crop coefficient data. Effective soil water
monitoring should also be encouraged.

The survey has also identified the need for information and advice on available irrigation
scheduling tools and methods. A concerted effort is required to promote and implement
scheduling technology to improve irrigation water use efficiency in sugarcane production.
Agricultural extension and demonstration trials are possible options.

Acknowledgements

The authors would like to thank the following Extension Officers for their help during the
study: Banie Swart, Francois Botha, Sarel van der Walt and Marius Adendorff. We would also
like to thank all growers that were willing to participate in the study. We appreciate the
suggestions by Mike Clowes (Swaziland Sugar Association) on the design and contents of the
questionnaire.

REFERENCES

George BRF (1988). A simple method of scheduling irrigation. Proc S Afr Sug Technol Ass
62: 149-151.

Olivier F and Singels A (2001). A database of crop water use coefficients for irrigation
scheduling of sugarcane. Proc S Afr Sug Technol Ass 75: 81-83.

Olivier F and Singels A (2003). Water use efficiency of irrigated sugarcane as affected by
row spacing and variety. Proc S Afr Sug Technol Ass 77: 347-351.

Singels A, Kennedy AJ and Bezuidenhout CN (1998). Irricane: A simple computerised

irrigation scheduling method for sugarcane. Proc S Afr Sug Technol Ass 72: 117-122.

Singels A, Kennedy AJ and Bezuidenhout CN (1999). Weather based decision support

through the internet for agronomic management of sugarcane. Proc S Afr Sug Technol Ass
73:30-32.

243 Proc S Afr Sug Technol Ass (2004) 78

Proc S Afr Sug Technol Ass (2004) 78 244