|
A Technical Guide for Wine Producers
|
|
RECENT ARTICLES | WYNBOER HOME
An investigation into possible water savings using sub-surface irrigation (Part I) - Irrigation quantities, wetting patterns and root distribution
Philip Myburgh, ARC Infruitec-Nietvoorbij, Stellenbosch
Key words: Grapevine, irrigation, sub-surface, root distribution, clogging.
Introduction
Although the Lower Orange River region is situated in a summer rainfall region, the rainfall is too little to meet the water demands of grapevines. Viticulture is therefore dependent on irrigation quotas varying from 15 000 m3 per ha per annum in the Upington area, to 18 000 m3 in the Kakamas area and further along the river. During abnormally dry seasons in the eastern parts of the country rainfall is insufficient to fill the dams supplying the vineyards along the Lower Orange River. When this happens water restrictions may be imposed. Consequently there is enormous pressure on the viticultural industry to develop irrigation technologies to ensure sustainable production and the best possible grape quality using less water. In this regard it has already been proven that the same quantity of grapes can be produced using less water if the traditional full surface flood irrigation is replaced by irrigation in alternating rows or furrows (Myburgh, 2003). The question is whether even more water may be saved in the region if vineyards are irrigated using permanent irrigation systems, for example drip. As a considerable amount of water evaporates from the wet surface after irrigation, sub-surface drip should result in further savings if the surface soil does not get completely wet.
Sub-surface drip irrigation is already being used commercially in vegetable crops in the USA, sugar cane in Swaziland and pastures in South Africa (C. Malan, personal communication). In Australia 50 mm drainage pipe is being used below the surface to irrigate vineyards that were previously subject to flood irrigation. The pipes are installed mechanically next to the vineyard rows. Heavier soils proved to be more suited to the drainage pipe than sandy soils (P. Clingeleffer, personal communication). Irrigation can also be applied sub-surface using porous pipe which is produced from recycled rubber. ARC Infruitec-Nietvoorbij tried to irrigate a vineyard near Oudtshoorn using the so-called Leaky PipeŠ, but without success. Tests showed that the water delivery per unit length pipe was uneven and that the flow decreased when the pressure was increased to 100 kPa (Myburgh, unpublished data). The pores in the pipe walls are possibly compressed when the water pressure increases.
The purpose of this trial was to compare sub-surface with traditional above-ground irrigation methods, with a view to possible water savings, without reducing production levels and grape quality. In this article the practical aspects of application methods, wetting patterns and root distribution are discussed. Grapevine reaction with regard to growth, production and quality are discussed in a follow-up article.
Material and methods
The investigation was conducted near Upington using Sultanina (clone H5)/143B Mgt, which is cultivated for drying purposes. The vineyard was planted in 1998 in layered alluvial soil, classified as a Dundee form (Soil classification working group, 1991). The soil was deep ploughed before establishment using a wheel tractor to mix the layers, which usually restrict root distribution. The plant spacing was 3 m by 1.8 m and the vineyard was trained onto a gable trellis (Zeeman, 1981). Various methods of applying water sub-surface were compared to furrows and above-ground drip (Table 1). Although previous attempts to irrigate vines with porous pipe had failed, it was decided to run the test once again. Instead of using Leaky PipeŠ as in the previous trial, this time Porous PipeŠ was used for the trial. The latter was slightly sturdier, thereby reducing the possible compression of the pores. The wetting effectiveness of drippers with regard to flow rate, as well as the depth at which they are installed, also formed part of the investigation. All sub-surface irrigation lines were installed approximately 20 cm from the vine row. Each irrigation treatment was repeated four times in a randomised block design. On each plot the 12 experiment grapevines were bordered by two grapevines on each side of the experiment rows, and on each side by two rows to prevent overlapping of treatment effects. The respective treatments were irrigated from the 1999/2000 season onwards, in other words just after establishment, until the 2003/04 season, using the different methods.
As a result of limited infrastructure, all the low flow rate drip treatments initially had to be irrigated in two instalments over two days. The storage dam for the irrigation water used in the trial, which was pumped from a concrete canal on the experiment farm, was increased by about 30% during March 2002. Over the last two seasons it was therefore possible to wet all the low flow rate drip treatments in one irrigation instead of in two instalments. Despite the bigger storage capacity it was not possible to irrigate all the treatments on the same day. Soil water content was measured using a neutron moisture probe at 30 cm, 60 cm and 90 cm depths. The probe was calibrated against gravimetric soil water content. The irrigation volumes were monitored using water meters. In the case of the furrows and drainage pipe the water levels in the dam were measured before and after irrigation to calculate the volume of water applied per irrigation. At the end of the trial samples of drippers were collected at all the drip irrigated plots. These drippers were subjected to flow measurement tests at the factory where they were manufactured. Intensive penetrometer measurements were taken five years after soil preparation, on all the experiment plots to characterise the physical condition of the soil. The measurements were done two days after irrigating the entire trial over the full surface. During the penetrometer study both the bulk density of the soil and the gravimetric water content were measured. Root distribution and density across the vineyard rows were determined for each site at the end of the trial.
Results and discussion
Soil physical condition: There were no differences in penetrometer resistance among the respective irrigation treatments. The slight compaction occurring between 20 cm and 30 cm, resulted from the cultivation required to level the beds for flood irrigation once the soil had been deep ploughed (Fig. 1). The average penetrometer resistance exceeded 2.0 MPa at 70 cm depth. This is in line with the soil preparation depth that is usually achieved by wheel tractors. However, the critical resistance of 2.5 MPa at which root penetration starts to be restricted (Van Huyssteen, 1983 and references therein), was not exceeded up to 90 cm. During the penetrometer study the average soil water content was 13.7%, 13.9% and 12.5% respectively over the 0-30 cm, 30-60 cm and 60-90 cm depth increments. As the subsoil was slightly drier than the upper layers, this could have increased the penetrometer resistance compared to the upper layers. There was nevertheless a good correlation between penetrometer resistance and the bulk density of the soil (Fig. 2). The fact that the bulk density over the biggest part of the root depth was below 1.5 g/cm3, also indicates that there were no soil physical restrictions to root penetration up to 90 cm (Van Huyssteen, 1988).
Pipe and dripper clogging: The porous pipe was entirely clogged after three seasons. Consequently the grapevines of this treatment were growing extremely poorly, and on the verge of dying back. It was decided to replace the porous pipe with 3.5 L/hour above-ground drip to prevent the grapevines from dying. The treatment was changed in November 2001, and could therefore no longer be considered part of the field trial. Tests at the dripper factory showed that no clogging occurred in the above-ground drippers, while 12.5% of the sub-surface drippers were clogged (Table 2). There is no explanation why no dripper clogging occurred in drip lines at 45 cm depth (B5). Microscopic investigations showed that the clogging was caused by grains of sand or root penetration. The latter occurred despite the application of trifluralin through the filter system, it being released into the water on an ongoing basis. In the drippers that did not become clogged, however, there was no significant decrease in the flow rate (Table 2). Under the given circumstances clogging will definitely decrease the lifespan of sub-surface drip compared to conventional above-ground drip. It was noted that the drainage pipe had a varying degree of silt accumulation. Clogging as a result of the silt may explain why the water bubbled to the surface during irrigation on some plots after three seasons.
Irrigation quantities: The average monthly irrigation quantities indicated in Table 3, are based on the full surface. In the case of sub-surface irrigation using 2.3 L/hour drippers, an average of 3.7% less water was applied compared to above-ground drip. Since all four treatments irrigated with 2.3 L/hour drippers, were irrigated from the same valve, the lower application was possibly caused by a measure of dripper clogging, which could have reduced the flow rate per unit surface. As a result of practical problems it was not possible to apply the same amount of water using the 3.5 L/hour as with the 2.3 L/hour drippers. Over the three seasons an average of 16% less water per annum was applied sub-surface using the 2.3 L/hour drippers compared to the 3.5 L/hour drippers (Table 3).
Soil water content: An example of the seasonal changes in the soil water content is indicated in Figure 3. For all treatments there was a gradual drying out over the warmest part of the season. The existing infrastructure and the fact that water was obtained on demand from a canal system, did not allow more than one irrigation per week. More regular irrigations would probably have prevented the gradual drying out. It was also clear that no water flowed through the sub-surface porous pipe during the first half of the particular season (Fig. 3). The soil water content of this treatment only started to increase once above-ground drip was installed towards the end of November 2001. As it was not practically possible to irrigate all the treatments on the same day, it was not possible to measure the soil water content just before and after the irrigations. The soil water content of the respective irrigation treatments could therefore not be properly compared. The soil water content of the drip treatments that were irrigated sub-surface, tended to be drier than the above-ground drip. This was probably due to clogging as discussed above. Compared to the drip treatments, there was also a tendency towards higher water content with furrows.
Wetting patterns: Measurement of soil water content directly after an irrigation on the drip line, as well as at a distance of 30 cm and 60 cm away in the work row, indicated that most of the water was applied in a 60 cm wide strip with furrows as well as drippers. This distribution pattern was especially noticeable in the case of the drip treatments (Fig. 4). It was also clear that the soil water content was higher where irrigation was applied at 30 cm and 45 cm depths using drip lines. Wetting front detectors and tensiometers that were installed in a furrow and an above-ground drip plot at 30 cm and 60 cm depths respectively during the last season (2003/04), showed that during each irrigation a well-defined wetting front moved down to 30 cm at the furrows (data not shown). According to the tensiometer readings soil wetting occurred down to 60 cm at the drip plot, but even after an 18 hour irrigation, neither of the two depths had an obvious wetting front to which the detectors could react.
Figure 4. Wetting patterns around sub-surface drip at different depths compared to above-ground drip lines in alluvial soil as measured at the pea size berries during the 2001/02 season at Upington.
Root distribution: The root systems in all the treatments consisted mainly of fine roots. Despite limited wetting as discussed above, the fine roots occurred in all the treatments up to the middle of the work row (Fig. 5). Just after planting, two full surface flood irrigations were applied to ensure that the young grapevines grow properly. This could have caused the root systems to develop up to the middle of the work rows. Wetting of the total available soil volume by rain could also have contributed to the unexpectedly wide lateral root distribution. The absence of physical restrictions as discussed above, clearly did not restrict lateral, as well as vertical root distribution up to 90 cm depth. However, the root concentration was higher away from the vine row where the sub-surface drip lines were installed (Fig. 5c). Root concentrations were considerably lower compared to the 3000 roots/m2 in the small wetted soil volumes where drip irrigation was applied daily in pulses according to open hydroponic principles (Myburgh & Howell, unpublished data).
The root distribution patterns of all the replications of a specific treatment reacted the same. In the case of furrows, root concentration occurred to some extent in the wetted soil volume (Fig. 5a). A lesser amount of root concentration occurred where the above-ground drip (Fig. 5b) was used, compared to the sub-surface drip (Fig. 5c). There is no acceptable explanation for the absence of root concentration under the above-ground drip lines as well as around the drainage pipe (Fig. 5d). Where sub-surface drip was used to irrigate at 15 cm depth, there was not a high concentration of roots around the drip lines as at 30 cm and 45 cm depths (Fig. 6) either. As trifluralin was also applied by the above-ground drip, this substance may have enjoyed more widespread distribution in the top soil layer than in the subsoil, thereby limiting root development further away from the dripper lines. Higher drip flow rate (3.5 L/hour) did not result in a significantly higher root concentration around the drip line compared to the 2.3 L/hour drippers, installed at the same depth (Fig. 6b and 6d).
Conclusions
Acknowledgements
The ARC for infrastructure and other resources, Winetech for partial funding as well as Leon van der Walt and the Soil Science personnel at ARC Infruitec-Nietvoorbij for technical assistance.
For more information contact Philip Myburgh at myburghp@arc.agric.za.
Literature references
Soil classification working group, 1991. Soil classification - 'n Taxonomic system for South Africa. Dept. Agricultural Development, Private Bag X144, Pretoria 0001.
Myburgh, P.A., 2003. Possible flood irrigation technologies to reduce water use of Sultanina grapevines in a hot, arid climate. S. Afr. J. Plant Soil 20, 1 - 8.
Van Huyssteen, L., 1983. Interpretation and use of penetrometer data to describe soil compaction in vineyards. S. Afr. J. Enol. Vitic. 2, 59 - 65.
Van Huyssteen, L., 1988. Soil preparation and grapevine root distribution - A qualitative and quantitative assessment. In: Van Zyl, J.L. (red). The Grapevine root and its environment. Technical communication No. 215. Dept. Agriculture & Water Supply, Pretoria, South Africa. pp. 1 - 15.
Zeeman, A.S., 1981. Oplei. In: Burger, J.D. & Deist, J. (eds). Wingerdbou in Suid-Afrika. Nietvoorbij, Stellenbosch, South Africa. pp. 185 - 201.
Summary
|
|
Wynboer is incorporated in WineLand, magazine of the SA wine producers.
Visit our sister sites:
|
| UP |
COPYRIGHT (C) 2000 WineLand |