Janine de Wit

Fresh water is needed worldwide for nature, agriculture and industrial activities. The water demand will increase in the future due to economic growth, urbanization, and food production, while freshwater supplies are decreasing. In addition, weather extremes (both more dry and more wet periods) will occur more frequently and more intensively due to climate change. Therefore, measures are necessary to anticipate the imbalance of water demand and supply. One of the measures could be to convert the historically existing drainage systems (in the Netherlands currently at approximately 34% of the agricultural fields) to controlled drainage with subirrigation (CDSI) systems to discharge, retain, and recharge water.

The Dutch water system has traditionally been designed to drain water efficiently to prevent flooding. Water is quickly drained from agricultural fields to surface water and then discharged from the area via canals and rivers. After World War II, the population growth led to an increase in food demand. Therefore, agricultural fields were drained with subsurface pipes to avoid too wet circumstances and thereby increase crop production. In addition to intensive drainage, land consolidation took place to enlarge agricultural fields and improve their accessibility. Pipe drainage, land consolidation, urbanization and groundwater extractions, among other things, caused a systematic decline in the groundwater levels. As a result, the agricultural fields in the Dutch Pleistocene uplands became more vulnerable for the effects of droughts. Parallel to the social-economical changes in society, pipe drainage systems also developed over time. Conventional pipe drainage aimed to discharge water, but developed into controlled drainage (CD) which aimed to both discharge and retain water. Nowadays, CD develops into controlled drainage with subirrigation (CDSI) with three aims: i) discharge water to avoid flooding or waterlogging, ii) retain water in the soil, and iii) recharge water to prevent agricultural and hydrological drought.

In general, a CDSI system consists of a control pit where external water for subirrigation is supplied. The control pit is connected to a collector pipe, which is connected to the infiltration/drainage pipes perpendicular to the collector pipe. The control pit contains a small pipe which determines the maximal water level (‘crest height’). A float can be used to control the water supply. If the hydraulic head in the control pit and drains exceeds the groundwater level, water flows through the pipes and infiltrates into the subsurface through the perforated pipe holes. As a result, groundwater level increases and soil moisture in the rooting zone of plants increases due to capillary rise from the groundwater. Internationally, countries as United States of America, Canada, Australia, Sweden and Iran have experience with CDSI.

In recent years, CDSI has been investigated as a technological measure to address the imbalance between water demand and water supply in the Dutch Pleistocene uplands. Four field pilots were conducted in America, Haaksbergen, Lieshout and Stegeren, which are all located on the higher sandy soils of eastern and southern Netherlands. The field pilots differed in soil profiles and geohydrological characteristics, but had similar measuring equipment. The field measurements show that CDSI raises groundwater levels, which increases the soil moisture content in the root zone through capillary rise. This increases water availability to the crop and thus increases plant transpiration and dry matter production. A simulation study with the Soil-Water-Atmosphere-Plant model (SWAP), calibrated to the field measurements, shows that only a minor part of the supplied water contributes to increased plant transpiration. Most of the supplied water leaves the system as ditch drainage and as downward seepage to the deeper groundwater. The distribution between these components largely depends on the adjacent surface water level and geohydrological characteristics.

The required water supply (subirrigation) was high in the field experiments, because a fixed and relatively high groundwater level was pursued throughout the growing season. The combination of a fixed crest level and the possibility of continuous water supply resulted in a water supply ranging between 500 and 1,000 mm per year. Water supply can be reduced by more smart water applications, whereby the CDSI system contains a dynamic crest level and a dynamic pump management. This allows an appropriate water level in the control pit on a daily basis, calculated via an algorithm based on actual soil moisture conditions and weather forecasts. A dynamically controlled CDSI system saves more water in a wet growing season with a small plant water deficit and herewith a relatively low water demand for subirrigation than in a drier growing season with a larger plant water deficit and therefore a relatively high water demand for subirrigation. Besides the active control of the CDSI system, water can also be saved if the resistance to downward seepage is higher (location characteristics), a higher ditch water level is maintained around an agricultural field or region (management) or a crop with deeper roots is cultivated (crop specific).

Finally, this thesis shows that transpiration can increase (less transpiration reduction) through water supply and higher groundwater levels, but a situation may occur where more water is supplied while the transpiration remains equal. This means that the extra supplied water drains to the ditch or recharges regional groundwater system. Thus, in this situation, CDSI does not serve as a measure to increase crop production. The data obtained and SWAP-modeling procedure developed in this thesis can be used to determine the effects of CDSI under different conditions, which supports CDSI implementation and management on field scale.

Although the water supply can be reduced through applying water in a smarter way, a CDSI system requires water, so this local measure affects the regional water system. Therefore, it is important that water system stakeholders agree on the extent to which CDSI can be applied. A system dynamics model (SDM) was built which is relatively easy to understand, but includes all key factors and non-linear interactions of a physical CDSI system. The SDM is based on the four field experiments and the calibrated SWAP model. In this SDM, CDSI is scaled in percentages of a total area where subirrigation is applied. As surface water is not infinitely available, the challenge is to balance water use for nature and across societal sectors. Results show that CDSI upscaling propagates non-linearities in hydrological fluxes, with three critical phases related to regional water availability determining the potential for successful CDSI upscaling. It is crucial to identify which phase the effects of upscaling correspond to: from Phase 1, with sufficient regional water availability; through Phase 2, a critical zone where water demand begins to limit regional availability; to Phase 3, where high water demand significantly impacts the regional water system and CDSI is no longer beneficial for the crop. All in all, CDSI could be one of the measures to consider for a climate robust soil-water system. However, CDSI, as a local measure, may increase pressure on surface water availability during drier periods. This requires embedding field measures such as CDSI into regional water management strategies.

The design, construction, and management of a CDSI-system should include the water quality of the water supply source to prevent clogging. Additionally, it is important to include stakeholders both for knowledge exchange as well as to build support for responsible implementation. Although the methodology in this study is applied to field experiments in the Netherlands, the modeling approach is generally applicable, allowing the analysis to be extended to other areas in the world.

As conclusion, historically, drainage systems were a measure for removing excess water. Therefore, drainage pipes were installed at approximately 34% of the Dutch agricultural fields. In the short term, it is a relatively small investment to convert traditional drainage systems to CD. In this way, the water strategy at these fields shifts from drainage to drainage when necessary and retention when possible, which is already a step towards a more climate robust soil water system. Then, if CDSI fits within the regional water management, water can be recharged. The effectiveness of CDSI systems depends, among other things, on the geohydrological characteristics (retention of water), meteorological and crop characteristics (water demand for subirrigation and effect on crop growth) and the implementation of CDSI within the regional water management (water availability for subirrigation).

CDSI is one of the measures that can contribute to the transition towards a more climate-robust soil-water system in the Netherlands. In this transition period, there is a need to rethink the current landscape design, taking into account the physical interactions in the soil-water system. This might mean that geohydrological conditions are more leading in choices about land and water use, which could also mean accepting that not all types of land use are feasible in every location at all times. For example, cultivate more drought-resistant crops in drier areas and cultivate more flood-tolerant crops in wetter areas. The measure CDSI could then be a technical measure to be used on agricultural fields where soil and water were leading in the decision to grow crops suitable for that location, but the field may still experience some dry and/ or wet periods, where CDSI could be used to mitigate the water stresses. The field experiments and related modeling approaches in this study have increased the hydrological understanding of CDSI, which can support decision-making regarding implementation by farmers and water management authorities.