Crop Water Stress

The SPAW model focus is on hydrologic water budgets of agricultural fields as compared with emphasis on crop growth and production. However, the cropping plants play a major role in water budgeting, and if they are impacted by water availability, that effect needs to be estimated. Having estimated daily plant transpiration, it is a logical extension to estimate the magnitude of transpiration-related water stress and its effects on growth, phenological development, root distribution and yield. Only grain yield does not in turn affect the plant status and future transpiration, therefore these effects can be considered as feedback loops as shown in Figure 5.

Daily plant stress was defined as:

Stress = 1 - (Actual Transpiration/Potential Transpiration)

Or simply, if the daily AT was only 80% of PT, the daily stress value would be 0.20. With AT/PT defined for each soil layer, the total plant stress is the weighted summation of all active soil layers, the weighting being the proportionate share of PET assigned to that layer according to root density and layer thickness.

Water stress effects on canopy growth and crop greenness cause reductions from the entered canopy and greenness values which represent non-stress conditions. The stress was classified into three different zones as depicted in Figure 8. For canopy development, stress zone A (where AT/PT is high) assumes normal growth conditions, Zone B computes a linear decline of growth from full normal daily growth to no growth, and no growth occurs if the AT/PT ratio falls in zone C. The ranges and number of zones have been fixed to default values. The increment of expected daily growth is computed from the entered canopy curve, then this amount is reduced by multiplying the factor for the stress zone encountered.

Actual ET Zones
Figure 8: Crop water stress zone effects on crop canopy and greenness.

Similarly, crop greenness is reduced by defined stress zones. The values associated with each zone are actual percentages of greenness which are subtracted from those entered (not a multiplied percent as for canopy). More stress for greenness than growth is required before permanent damage and loss of transpirability result. Thus, non-stress conditions are assigned to a stress value of 0.5, the increasing amounts of phenology loss are applied. Values derived by calibration with observations from several stress years were made default. Other crops and locations may require adjustments, or additional data may suggest refinement.

The effect of crop stress and non-uniform soil water profiles on root distribution and abstraction has not yet been included in the SPAW model. This is an obvious refinement which may improve accuracy in some specific cases.

Water stress effects on crop yield can be significant and but difficult to evaluate. There are many individual effects which integrate to result in any specific crop yield, and all but water availability are outside the bounds of this model. Yet, for any particular crop and locality under non-irrigated farming, available soil water is often the single most significant yield determinant. Therefore, assuming all other effects are near normal, an accurate estimate of daily crop water stress and how this stress is integrated throughout the growing season to result in a final yield could provide a strong predictor of yield reduction. This capability has been incorporated into the SPAW model.

It is well known that water stress affects grain yields more severely if it occurs during the fruiting period than during the vegetative phase. Some crops appear to be more susceptible to this fruiting period stress than others. Corn is a good example. Some crops are quite stress resistant compared with others, such as some sorghum varieties.

To represent this relative water stress susceptibility in time and among crops, susceptibility relationships were developed as shown for corn and soybeans in Figure 9. The susceptibility numbers are arbitrary to provide relative values. Highest values occur during times involved with pollination and seed production. The time scales are begun at planting to allow some flexibility. However, it is more critical to center the maximum values when the crop actually flowers, thus the entered susceptibility curve may need to be modified by stretching or shrinking the time scale. The yield susceptibility data are entered as time-value data pairs to form linear segments representing the calendar year.

Corn Yield SusceptibilitySoybean Yield Susceptibility
Figure 9: Example yield susceptibility curves for corn and soybeans.

Each day's plant stress (1 - AT/PT) is multiplied by the corresponding susceptibility value and these values accumulate to the end of the growing season. These seasonal values, of course, are only relative water stress indices. They only become meaningful when correlated with observed crop yields from the site or region. Many such correlations were made for corn in a regional study by Examples are shown in Figure 10 (Saxton and Bluhm, 1982; Saxton et al, wheat??). These results show that the cumulative effects of crop growth, soil water availability, atmospheric demand, and crop stage can provide a reasonably good estimate of crop water stress effects on yield under rain-fed conditions.

Soybean Water StressWheat Water Stress
Figure 10: Correlations of computed annual crop water stress index with grain yields.

There is evidence that vegetative growth is closely allied with plant transpiration (Dewit 1958). Thus, the estimated transpiration from SPAW can also provide water stress effects on vegetative crops such as forage and grassland.

Field Hydrologic Processes | Precipitation | Infiltration | Potential ET
Interception | Soil Water Evaporation | Plant Transpiration
Root Water Uptake | ActualET | Soil Water Redistribution | Irrigation