Global Change Consulting Consortium, Inc.

Irrigation management

   The role of irrigation in the global food supply is indisputable; 8% of arable land that is irrigated supplies 33% of our food.  Yet, irrigation is challenging to maintain against two major problems:

            Consider surface waters first as sources of irrigation water.  The rise in river salinity with irrigation is at heart a simple matter of mass balance.  In the absence of irrigation, rivers carry salt loads to the sea, where it is ultimately disposed to sediments and recycled by tectonic action.  In carrying the load, rivers attain mean salt concentrations that are set as input rates (largely from weathering) divided by mean water flow.  Agriculture does not add sodium salts in any quantity, so the effect of diverting river flows to crops, thus increasing the loss of water to the atmosphere, is to decrease river flow and thus decrease the dilution of salts.      

            Increased river salinity can be modest.  It is, however, not the only determinant of soil salinity.  If surface waters are applied in merely sufficient quantities for crop growth, all the water is lost and all the salt remains, thereby building up progressively.  Flushing the salts, at least periodically by overwatering, returns the salts to the river source.  On a long-term average, salt loads in the river increase as 1/(fraction not lost to ET).  This may be tolerable on average, but on the lowest reaches the salinity is highest and can be lethal to crops.  The river may even run dry.  A number of rivers in the US West and China match this description.    

                        Mitigation methods include:

   In light of growing problems with sufficiency of irrigation water, a number of research groups have attempted to develop management strategies for growers, including:

o       More efficient irrigation methods.  Subsurface / drip irrigation reduces the exposure of applied water to surface evaporation.  The reduction in total water use is apparent mainly for crops that have lengthy stages of low leaf cover, such as row crops; for orchards, the savings are considerably smaller.  The need remains to irrigate to excess periodically to flush inevitable salt accumulations to drainage areas. 

                 The complete irrigation system is also scrutinized.   Water delivery to the point of use is improved by lining irrigation canals.  There are some downsides, including reduction of canal-side wild vegetation that may support insects and birds that offer pest control.

o       Better decisions on which crops to grow, partly on the basis of economic yield per unit water use.  Decisions can be implemented by individual growers.  On larger scales, the need for irrigation water can be reduced by changes in the agricultural system.   Large decreases in crop production required to sustain a population and thus in water use attend the switch from using grain to feed livestock to using grain and other crops directly in the human diet.  Among livestock, ruminants such as cattle are far less efficient in converting crop biomass to edible animal products.  Regrettably, current global trends are in the opposite direction, particularly in the large and growing economics of China and India.

o       Identifying stages of crop development at which reductions in water supplies have minimal or no effect on final yield and on the potential for future yield.  The full, quantitative optimization of such deficit irrigation, as it is termed, is a project in which we are currently involved.  The research is being carried out by a consortium of 13 investigators at three universities in CA, NM, and TX.  We offer a few details here.

·        One of the investigators, Ted Sammis at New Mexico State Univ., has engaged our consulting firm to develop:

o       A detailed model of pecan photosynthesis and transpiration, under varying degrees of water stress, using all that we know about the physiology and development of the plants, the soil water balance and water transport, soil and plant N dynamics, and the weather drivers

o       Simpler, more empirical models that growers could ultimately use.  The holy grail is a model that predicts performance (photosynthesis, growth, nutfill, water use) as a function of the degree of water stress imposed.  In turn, the simplest measure of imposed stress is the fraction of unstressed ET allowed by managing irrigation inputs.

o       Consequently,  prediction of the optimal stress to impose at various growth stages.

o       Ideally, the responses of ET, photosynthesis, growth, and yield to low-nitrogen stresses.  We could then develop joint water and N management strategies.  We won't present these ideas here.

·        The detailed model predicts many plant responses that can be tested, to see if the model is accurate enough:

o       ET, of course, at all times of day and of the season

o       Photosynthetic rate, ditto - Measure both this and ET with eddy covariance, on selected sites (here and in California; the funding comes from a grant to a consortium in CA, NM, and TX)

o       Additional physiological responses:

§         Stomatal conductance at all times and in all local environments on the tree (measured by gas exchange)

§         Photosynthetic capacity, which is the potential peak rate of photosynthesis at high light and high CO2 level.  It is modulated by actual light and CO2 levels, with the latter controlled by stomtal conductance and the CO2 level in ambient air (this capacity is measured by gas exchange)

o       Leaf water stress at all times, which can be measured by pressure bomb and also by remote sensing of leaf reflectance

o       Leaf N content over the season, which can be sampled

o       Soil water balance, ditto

              We already have a detailed model that predicts:

·        Light interception on all leaves (statistically), using canopy architecture (including neighboring trees) and solar angles at all times (in time steps of one's choice - e.g., hourly)

·        The rest of the leaf environment - canopy air temperature and humidity (including changes induced by canopy  ET and PS), windspeed, CO2 level

·        Resulting transpiration and photosynthesis of all leaves (again statistically), including:

o       The stomatal response to the immediate (aerial) environment of the leaf and to achieved leaf PS rate, and to leaf (or root) water potential, using an extended version of the well-verified Ball-Berry model

o       The photosynthetic rate, from the robust Farquhar-von Caemmerer-Berry enzyme-kinetic formulation

o       Leaf temperature from the energy balance on each leaf (inputs of solar radiation in both the PAR and NIR; thermal IR from the sky and from surrounding vegetation; thermal IR output radiated by the leaf; transpirational cooling; and convective heat transfer between leaf and air)

We also have generated a simpler, more empirical model of whole-tree ET, PS, and nut growth, using:

·        The predicted changes in stomatal conductance (gs) from water stress

o       The calculation is iterative.  Water stress depends on soil water withdrawal, hence, on ET and gs themselves.  At each time step, gs is adjusted to maintain the desired fraction of unstressed water use.

·        The consequent changes in PS rate, with the assumption that PS capacity is not changed by water stress but only by leaf N content

·        The predicted seasonal trend in leaf N content, assuming that leaf N is progressively withdrawn to support nut growth.  (This need work.)

·        An estimate of how much the final leaf area is reduced by water stress during leafout. (Also needs work)

This model makes some interesting predictions:

·        In short, if one has only 50% of normal irrigation water available, one should apply irrigation to support 36% of normal unstressed ET during leafout and 60% of normal unstressed ET during nutfill.  One expects 67% of normal nut yield. 

·        The early-season stress reduces leaf area development to 71% of normal.

o       Combined with reduced stomatal conductance, the water use becomes 36% of what a full-area, unstressed canopy would have.  The reductions in area and in ET rate per unit leaf area are both important.

o       During nutfill, the canopy uses less water, computed as leaf area multiplied by transpiration rate per unit leaf area.  Leaf area is reduced to 71% of normal, and the leaves are stressed to use less water per area.   As a saving grace, the leaves have far smaller proportional reduction in photosynthesis per area

            Other schedules of stress are worse for yield.  This is a preliminary model; it needs improvement to resolve more growth stages.  It also needs better data on leaf area development.

            The results are expected to be applied in agricultural extension in the three states.  There is a niche for using these developments more widely.  Provided that the application is effective in the original crops (pecans, almonds, pistachios) and states, the method will have credibility in application to other states and in being developed for other crops.  Decision support can also be offered to irrigation districts and state water boards.