Subsurface drip irrigation (SDI) is a highly efficient irrigation technology that, while commonly used in arid agriculture, has had slow uptake in New Zealand. The Canterbury region’s water resources are approaching or have reached full allocation, and water usage is a focal point of societal tensions around agricultural pollution and usage of natural resources. SDI has potential to improve water use efficiency on farms and to minimise nutrient loss, but it is currently more expensive to install per-hectare than spray irrigation systems that are typically used in Canterbury. To invest in SDI, farmers require confidence that this technology is proven and that it will be cost-effective. There is a scarcity of documentation on the effectiveness of SDI in Canterbury soils.

The overall objective of this study is to document the efficacy of SDI for Canterbury arable agriculture. Three different lateral spacings of SDI and one spray treatment were installed in a repeating block trial in Lincoln, Canterbury. The three lateral spacings were 0.6, 0.9, and 1.2 m, and the drip tapes were installed at a depth of 25 cm. A wider lateral SDI spacing requires less drip tape per area, but has a lower water emitter resolution. Hence, there may be a trade-off between the installation cost and yield. A bushing Monaro pea crop was grown from spring 2023 to summer 2024 to compare the irrigation treatments. This was the first season of a multi-year trial to evaluate the efficacy of SDI on this site.

There were two end points to this research. One was a comparison of yield and water productivity between treatments. Plant growth and development were monitored to evaluate differences in water stress. Disease incidence was also monitored, as foliar splash dispersal, a mechanism for pathogen transfer in plants, is minimised under drip irrigation. The other end point was an investigation into subsurface water movement. Optimal SDI design is highly dependent on local soil characteristics, so documenting the behaviour of water from buried emitters in a Canterbury soil typology can provide a knowledge base to inform decision making.

This project had unexpected resource limitations. Nonetheless, thanks to the efforts of various stakeholders, volunteers, and substantial donations of time and equipment, the trial was installed before crop irrigation was required. However, system validation was not possible pre-season and there were ongoing equipment reliability issues. A consistent 22% more water was applied to the spray treatment blocks than to the three drip tape treatment blocks due to a mis-recording of installed spray-system specifications. There was an unexpectedly wet summer (214 mm over the growing season) and only 132 mm of irrigation water was scheduled. This led to 161 mm of irrigation being applied to the spray treatment, and a total of 8% more total water being input to the spray treatment over the season.

The trial was irrigated at a slight soil moisture deficit over the season. Nonetheless, there was a good yield of an average of 5.3 t/ha across the treatments. There were no significant differences in yield between treatments (p = 0.77). Despite differences in water application volumes, there were no differences observed in average soil water storage between treatments. As the spray treatment received less water than the SDI treatments (346 mm versus 375 mm total water for spray and drip treatments, respectively), the spray treatment had a lower water productivity than the SDI treatments (p = 0.003).

There were no consistent, significant differences in measured plant stress responses over the season. Site observations showed differences in plant height developing over the season relative to lateral distance from the drip tapes. However, this difference was not apparent in the dry matter measurements made by quadrat cuts throughout the season. There were also minimal significant differences in plant count, pod mass, leaf area, and specific leaf area, which could provide indication of water stress. If plant stress was present, as site observations implied, the sampling procedures used in this trial were not sufficiently sensitive to detect any statistically significant differences in plant response. Any water stress present did not affect final yield or thousand seed weight between treatments, implying that plant stress was likely minimal across all treatments.

Soil moisture sensing by neutron probe showed less soil moisture retention in the gap between drip tapes than at the drip tape. Buried electromagnetic sensors showed that, for a 10 mm irrigation cycle, the acute soil moisture response was highly localised, and did not reach the gap between tapes even at the narrowest spacing. The equivalent yields between treatments demonstrate that, even at a 1.2 m spacing, the pea plants could access sufficient water. The implication of this is that the 1.2 m drip tape spacing is the most cost-effective of the three spacings. However, more trials of different crops on this site will be required to determine if the wider spacing is generally non-limiting.

This trial has shown that SDI irrigation can be effective in a Canterbury arable agricultural context. The installed irrigation regime has an expected lifetime of twenty years, so this work has facilitated follow-on trials on different crops under different weather conditions and irrigation operation regimes. A season with less rainfall will be necessary to investigate the comparative effect of SDI and spray irrigation on disease incidence. Further investigation of the hydraulic behaviour of SDI irrigation in the site’s Wakanui silt loam soil is also recommended. This research can provide useful information for modelling environmental effects, and for an economic analysis of SDI for Canterbury arable farmers.