Gardner and Young
Salinity Control Evaluation
“average” benefits for the range of 800
mg∕liter to 1,100 mg∕liter of dissolved
solids. This formulation serves to overstate
benefits (to an unknown degree) for two
reasons. The first is that because of poten-
tials for alterations in crop mix and water
use technology as salinity increases, the
damage function is most probably nonlin-
ear, increasing at an increasing rate. Sec-
ond, the likely average level of salinity
over the future planning period without
the control program is less than the 950
mg∕liter implicitly assumed in the federal
analysis. Hence, the appropriate base from
which to measure benefits, even before the
high flows which diluted salinity in 1983
and 1984, would be in the 800-900 mg∕
liter range. Thus we believe the marginal
agricultural damages are actually some-
what less than reported here.
The model. Agriculture in the Imperial
Valley is an extremely diversified, year-
round enterprise. More than one-fourth of
the 450,000 acres of cropland are double-
cropped. The extremely arid region di-
verts over 2.5 million acre-feet of water
annually from the Colorado River. Each
linear program consisted of 76 crop pro-
duction activities spread over five field
crops, nine vegetable crops, two soil drain-
age conditions, and two irrigation fre-
quencies. The predominant doublecrop-
ping options were each combined into
single activities.
Crop budgets and water use estimates
were adapted from the Imperial County
Cooperative Extension Service (1982).
Harvest costs were varied with yield. Crop
prices are real 1977-81 averages, ex-
pressed in 1982 dollars.
Proportional constraints were imposed
on the degree to which each crop could
be produced on well-drained soils. This
added realism by simulating heterogene-
ity of soil distributions and cropping rigid-
ities caused by marketing contracts. These
constraints limit the maximum response
farmers in the model can make to in-
creased salinity. This is thought to be more
realistic than earlier approaches which al-
lowed switching the higher valued crops
entirely to well-drained soils. (The general
format of the model is given in Table 1.)
Crop yields. Since the model reflects
productivity on two soil types and under
two irrigation frequencies, published
1977-81 District average yields were ad-
justed according to agronomists’ judg-
ments to derive the expected yield under
each of these four alternatives.
In addition, yield declinations were es-
timated for an increase in irrigation water
salinity from 800 to 1,100 mg∕liter. The
yield declinations were calculated from
the effective soil saturation extract con-
ductivities by salinity level and soil type
(as reported by Robinson, 1978:90). They
are “effective” conductivities in that they
were adjusted downward to reflect the
significant amounts of gypsum (calcium
sulfate) in the soil and water. (Gypsum
contributes to measured salinity, but is
relatively less detrimental to plant growth.)
These conductivities were used with the
expected yield decrements from increas-
ing soil salinity estimated by Kleinman and
Brown (p. 121).
This method allows the estimation of
yield decrements on poorly drained soil at
a given salinity level, as well as decre-
ments from increased salinity. Yields on
each soil type were estimated using five-
year Imperial County average yields, to-
gether with the expected yield decre-
ments and the crop distribution on soil
types reported by Robinson (p. 89). Alfal-
fa is moderately sensitive to salinity, while
the other field crops are more tolerant.
Vegetable crops generally are more salt
sensitive. Lettuce yields, in particular, de-
cline rapidly as soil salinities increase.
Carrots and onions are also sensitive to sa-
linity, but are grown exclusively on well-
drained soils.
The model was validated by comparing
projected crop acreage and water use in
the 800 mg∕liter model with recent ex-
perience. Irrigation water requirements