An important feature of this type of work is the continually changing water status of soil-grown plants. In contrast to marine environments, where the water supply is essentially infinite and salinity often varies in a moderate range, water in soil is finite and gradually depleted. The available water changes from day to day and even during the day. Metabolism continually changes, and the yield reflects these changes. Keeping plant conditions constant and reproducible becomes a challenge. The capacity to control the timing and intensity of plant water deficit is paramount for designing meaningful genomic and metabolomic studies. Changes in gene expression can occur within minutes in response to a change in environmental conditions (Seki et al., 2002).
The finite water supply is traceable to the particulate nature of soil. The pores between particles fill with water by capillarity, and surface tension in the water prevents the water from spreading to nearby unfilled pores. As a result, partial rewatering fills a few pores and wets only a local volume of soil. The roots in the wetted part are very wet while others remain dry in the rest of the soil. Maintaining a steady, uniform soil water deficit while the plant is removing water is impossible.
Refocusing on the plant suggests a way to maintain plant water status, however. By adding in the morning only the amount of water used the previous day, the added water wets a small soil volume and enters slowly through a few roots, preventing the plant w from declining during the day. The added water is depleted by the end of the day, and the soil water content returns to that of the previous night. Because transpiration is minimal at night, the plant water status continues to be constant through the night. The water addition is repeated at first light the next morning. With this procedure, plant w can be held essentially stable for weeks. Boyer and McPherson (1975), McPherson and Boyer (1977), Westgate and Boyer (1985), and Setter et al. (2001) used this approach.
The efficacy of the method can be tested by measuring w. When Boyer and McPherson (1975) conducted maize experiments in controlled environments with irradiances similar to full sun, w measured in the upper leaves were essentially stable throughout the day and night (JS Boyer and HG McPherson, unpublished data) and for weeks afterward (Boyer and McPherson, 1975; McPherson and Boyer, 1977). In subsequent field experiments, the results were more variable because of variation in climatic conditions (Jurgens et al., 1978).
The measurements have the additional benefit of thermodynamically based data with a physically defined reference for precisely repeating conditions in subsequent crops. Working in controlled environments, this ability to repeat conditions allows progressive probing of molecular mechanisms, and comparison of the results with those from other scientists. Because differences in w drive water through the plant and thus vary from place to place and time to time, the measurements need to include sufficient conditions to be informative. For example, knowledge of pollen and floret w was needed throughout the day in order to interpret the cross-pollination experiments of Westgate and Boyer (1986b).
Split root systems are an extension of this approach and involve growing part of the root system in one compartment supplied with water and the other in another compartment containing dehydrated soil. The split roots provide an experimental way to separate the direct effects of low tissue w from those that develop indirectly from root-derived plant growth regulators, for example. The plant w varies diurnally because water is continually available to the wet part of the root system, but the time spent at high w tends to be shorter each day than in a plant with water around the entire root system. Split root systems have been used to study the effect of drying soil on a number of plant functions, especially root signals sent to the shoots (Davies and Zhang, 1991; Dembinska et al., 1992).