|
IRRIGATION PROJECTIONS
IN GEORGIA'S ACT/ACF BASIN: 1995-2020
A proposal for research to the
Northwest Florida Water Management District
Route 1 Box 3100
Havana, FL 32333-9700
by
Charles Moss, Professor
Food and Resource Economics Department
Institute of Food and Agricultural Sciences
University of Florida
P.O. Box 110240
Gainesville, FL 32611-0240
Tel. (352) 392-1845
Fax. (352) 392-3646
e-mail: cmoss@fred.ifas.ufl.edu
July 30,1998
IRRIGATION PROJECTIONS
IN GEORGIA'S ACT/ACF BASIN: 1995-2020
1. Introduction
Differences between geological and political boundaries in the United States often
cause resource use in one state to directly affect resource use in another. While this
conflict may raise difficulties for several resources, the most severely impacted may be
water. In the southeastern United States, Alabama, Florida, and Georgia share water
resources in the ACT/ACF Basins. In total, farmers in these basins were estimated to use
400 million gallons per day (mgd) in 1992 (USDA, 1995). Georgia's farmers were
estimated to use 72 percent of that total; Alabama's farmers, 21 percent; and Florida's
farmers, 7 percent. Further, the majority of water use (78.6 percent, or 226 mgd) was
estimated to be used for crop irrigation.
In response to this sharing of water resources through an Interstate Compact,
these states, through an inter-state agency, have entered into negotiations to allocate the
water in the ACT/ACF Basins among the states. Prior to this negotiation process, the U.S.
Department of Agriculture's (USDA) Natural Resources Conservation Service issued a
study on agricultural water demand that created a baseline of that demand in 1992 and
projected future developments in 1995, 2000, 2010, 2020, and 2050. This study revisits
those projections and updates this analysis, focusing on the input and output price
assumptions implied in the previous study.
The next section of this study reviews the major projections for crop irrigation
demands in Georgia from the USDA study. Following its review, we present an empirical
model for crop acreage based on Coyle (1993), and Moss and de Bodisco (forthcoming).
The empirical model is followed by a section discussion of the data used to estimate this
empirical model. Next, the results of this estimation are presented in comparison with the
baseline from the Food and Agricultural Policy Research Institute. We then conclude
with a discussion of the implications of these results and proposed modifications to this
study.
2. Results of Previous Research
The USDA's report identifies five crops (peanuts, corn, cotton, vegetables, and
orchards) that account for 94 percent of the water used in irrigation in the state of Georgia
in 1992. These uses are presented in table 1. In total, this demand represents a water use
of 226.25 mgd in 1992. Table 1 also estimates that this demand is projected to grow to
413.58 mgd in 2010 and 556.95 mgd in 2050.
In addition to the changes in the overall level of water demand, the USDA also
indicates some subtle changes in the allocation of water use. Specifically, peanuts, corn,
and cotton will continue to be the largest demand, but there is a significant leveling
among their relative requirements. Also, the irrigation demand for orchards is projected
to increase rapidly.
These changes in water demand can be traced to three factors: change in the
number of acres in each crop, changes in the share of each crop irrigated, and changes in
the intensity of irrigation of each crop. Table 2 presents the historical and projected
number of acres in each crop, the number of acres irrigated in each crop, and the share of
each crop irrigated. These figures indicate that the total acreage of peanuts has been fairly
consistent over the past 25 years. This result is consistent with the government program
for peanuts. Extending the trend, the total acreage of peanuts is projected to remain fairly
consistent over the projected period.
Table 1. Projected Irrigation Demand for Georgia, 1992-2050
Year
Crop 1992 1995 2000 2010 2020 2050
Total Water Use
Peanuts 72.72 92.81 86.00 95.13 101.88 129.03
Corn 64.73 74.50 80.50 90.70 93.73 109.71
Cotton 39.23 53.21 60.20 75.65 82.42 111.55
Vegetables 22.85 30.57 35.31 44.96 46.37 54.44
Orchards 12.78 22.51 31.07 51.05 58.42 88.94
Other 13.94 37.11 43.45 56.09 56.87 63.28
Total Demand 226.25 310.71 336.53 413.58 439.69 556.95
Percent of Water Use
Peanuts 32.1 29.9 25.6 23.0 23.2 23.2
Corn 28.6 24.0 23.9 21.9 21.3 19.7
Cotton 17.3 17.1 17.9 18.3 18.7 20.0
Vegetables 10.1 9.8 10.5 10.9 10.5 9.8
Orchards 5.6 7.2 9.2 12.3 13.3 16.0
Other 6.2 11.9 12.9 13.6 12.9 11.4
Table 2. Historical and Projected Cropping and Irrigation in Georgia
Historical Projected
Crop 1974 1980 1986 1992 1995 2000 2010 2020 2050
Acres In Crop
Peanuts 322,700 327,400 378,400 455,750 403,800 351,900 351,900 351,900 351,900
Corn 557,800 490,600 275,300 189,500 191,100 192,650 195,800 199,000 208,400
Cotton 94,900 60,100 46,650 54,550 54,950 55,350 56,150 56,950 59,350
Vegetables 3,000 5,700 7,550 23,850 23,900 23,950 24,000 24,100 24,350
Orchards 8,750 13,000 15,050 15,700 15,700 15,700 15,700 15,700 15,700
Other 495,500 1,195,550 839,200 589,900 582,350 582,350 582,350 582,200 581,900
Acres Under Irrigation
Peanuts 21,450 148,750 215,950 225,400 208,200 188,850 200,350 211,850 246,350
Corn 34,300 199,400 137,900 143,200 144,300 145,500 147,850 150,300 157,350
Cotton 1,050 10,450 22,250 27,850 29,400 30,950 34,100 37,300 47,750
Vegetables 100 2,400 5,150 17,350 17,750 18,100 18,850 19,600 21,900
Orchards 200 5,200 4,150 6,000 6,850 7,050 8,100 9,150 12,300
Other 5,250 121,000 103,350 94,450 94,400 94,350 94,250 94,150 93,900
Share of Crop Irrigated
Peanuts 6.6 45.4 57.1 49.5 51.6 53.7 56.9 60.2 70.0
Corn 6.1 40.6 50.1 75.6 75.5 75.5 75.5 75.5 75.7
Cotton 1.1 17.4 47.7 51.1 53,5 55.9 60.7 65.5 79.9
Vegetables 3.3 42.1 68.2 7.27 74.3 75.6 78.5 81.3 89.9
Orchards 2.3 40.0 27.26 38.2 41.7 44.9 51.6 58.3 78.3
Other 1.1 10.1 12.3 16.0 16.1 16.2 16.2 16.2 16.1
Geometric Mean Share Irrigated
5.3 29.2 39.3 45.8 46.8 47.7 49.6 51.6 57.7
The data also indicate a significant decline in the two remaining agronomic row
crops (corn and cotton). These crops are projected to level off at their 1992 levels. This
result is somewhat consistent with the projections of Moss and de Bodisco who project a
further decline in corn and cotton until the year 2020.
The most drastic change in the demand for water, however, is the significant
increase in share of each crop under irrigation. The share of peanuts under irrigation
increases from 6.6 percent in 1974 to 45.4 percent in 1980; irrigated corn increases from
6.1 percent in 1974 to 40.6 percent in 1980; irrigated vegetables increases from 3.3
percent in 1974 to 42.1 percent in 1980; and irrigated orchards increases from 2.3 percent
in 1974 to 40.0 percent in 1980. Overall, the geometric average of the share of crops
irrigated increases from 5.3 percent in 1974 to 29.2 percent in 1980. This overall growth
rate slows to 45.8 percent in 1992. Thus, a significant portion of the increase in the
demand from irrigation in the state of Georgia may be traced to increases in the share of
each crop irrigated.
The final factor that may lead to a change in water demand is the change in
irrigation intensity (water applied per acre). Table 3 presents the irrigation rate in acre-
feet per acre implied by the baseline and projected water use in the USDA's report. Also
presented in table 3 is the Census of Agriculture's water application rate for the 1988 and
1994 Census. The USDA projects a significant increase in irrigation intensity for each
crop. Dividing the imputed rate of irrigation by its census values then gives an index of
water use. In 1992, corn and cotton projections are close to its Census figures. Similarly,
peanuts' and vegetables' water uses are within 10 percent of the Census numbers.
However, the USDA's estimates for orchards and other crops are far below the Census
estimates. Overall, the geometric area of water use is 77.8 percent of that predicted by its
Census numbers. However, the estimated values rise rapidly, reaching 112.5 percent of
the Census estimates in 1995 and 144.5 percent of the Census estimates by 2020. Thus,
later changes in the share of irrigation and changes in the intensity of irrigation are
projected to increase rapidly over the projected period.
3. An Empirical Model of Crop Selection
The USDA projection of irrigation demand discussed above points to a three-step
process of estimating water demand: (1) estimation of crop acreages; (2) estimation of the
share of each crop irrigated; and (3) estimation of the intensity of irrigation. As a first
step in this process, we estimate the level of crops and assume that the percentage of each
crop irrigated and level or irrigation remains at the USDA projected levels.
To estimate the crop acreage of each crop, we assume that farmers maximize
profit by allocating their land across various cropping alternatives. Specifically, they
allocate land across crop alternatives zI, (where i= 1,...m crops) according to output
prices p' and input prices w' (k=l,...n inputs). Farmers also take into account the level
of other quasi-fixed factors K. Given this formulation, Coyle derives an acreage response
equation:
(1) zA =a + a + aik + k K,+ ak+2 ZA, + ak+3 ZAt-I + ak-+4
Table 3. Changes in Irrigation Intensity
Crop 1992 1995 2000 2010 2020 2050 Census
Acre-feet Per Acre
Peanuts .36 .50 .51 .53 .54 .59 .40
Corn .51 .58 .62 .69 .70 .78 .50
Cotton .42 .55 .60 .70 .72 .80 .40
Vegetables .44 .55 .60 .70 .72 .80 .40
Orchards .37 .60 .77 1.12 1.14 1.31 .90
Other .17 .44 .52 .67 .68 .75 .60
Percent of Census
Peanuts 90.3 124.8 127.4 132.9 134.6 146.6
Corn 101.2 115.6 123.9 137.4 139.6 155.7
Cotton 105.9 138.1 150.4 176.0 179.0 200.3
Vegetables 109.6 145.8 167.5 210.8 215.0 244.0
Orchards 41.1 66.4 85.5 123.9 126.4 145.9
Other 27.5 73.3 85.9 111.0 112.7 125.7
Overall 77.8 112.5 122.7 142.2 144.5 161.2
In addition, Coyle uses the summing-up constraint to impose a homogeneity condition
across dual equations.
We depart from Coyle in two ways. First, our focus on an irrigation basin
eliminates the summing-up condition. We acknowledge that the total amount of land in
the basin is fixed, but we assume that sufficient pastureland, forest, etc. exists to allow
expansion or contraction of the land base. Second, to eliminate the effect of lagged land
allocation, we use lag-difference. Hence, we estimate the model
6
(2) DzA, = ao + ,c Dp, + jfk DW, + ekt
j=1 k=l
where DzA, is the logarithmic change in acreage planted to crop i in year t, Dpi, is the
logarithmic change in the price of output j, Dwk, is the logarithmic change in the price
of input k, and ekt is the error in equation k. The equation can be estimated using
maximum likelihood.
These estimates, along with the price sequence for future input and output prices,
can be used to generate a future series of acreage for each crop. Specifically,
(3) ZA,,+I = z exp(Dz,),
where exp () is the controlling operative. These acreages can then be used to estimate
future water demand in Georgia. Specifically, Georgia's water demand in the ACT/ACF
basin is then estimated as
(4) A,tu n+ z, (P)S = t A, Prt)SIAU + ZAA,t (P)S'UA
+ zU,,(P,)s: + z+,,(P,)sU4+ z ,(P)s u,
where w, ,(,, ) is the total water use, conditional on the price adjusted Pr,; sA is the share
of crop i irrigated; and u' is the water usage per acre. Taking s', from table 2 and u,
from table 3, we can then generate the projected water uses in Georgia consistent with
table 1.
The analysis of the price assumptions implicit in the water forecast in table 1 is
involved in choosing a set of output prices that make the projection error between the two
forecasts small. One approach would be to choose a constant geometric adjustment to the
price series to
rain T
(5) (n At ((P exp(8t))- W At2
t=1
If = = 0, then the projected irrigation demand is consistent with the acreage response
model. The distribution of irrigation demand will then be generated from this formulation
by applying Efron's bootstrapping to the price series.
4. Data
This model uses data on the total harvested acres and season average prices for
peanuts, corn, cotton, fresh vegetables, and orchards in Georgia. These are provided by
the National Agricultural Statistical Service (NASS) and the Georgia Agricultural
Statistical Service (GASS) for the years from 1970 through 1995. Due to the
unavailability of some of these numbers in the 1980s, tomatoes are used as a proxy for
vegetables, and peaches are used for orchards. Input prices are Divisia price indices based
on USDA historical data for the United States. The specific categories are fuel, labor, and
other. With the exception of labor, which is directly obtainable from the USDA numbers,
each of these categories is a Divisia price index constructed from the much larger USDA
input price index. The weights in each of these smaller categories consist of the published
weight of each price in the USDA index normalized over the sum of the weights of the
prices included in each constructed category.
5. Results
Table 4 presents the coefficients, the standard errors, the t-scores, and the p-values of
the maximum likelihood estimates of the acreage equations DZiAt from equation (3). The
variables i in the left-hand column are prices in each equation. Because of the small
sample size and the relatively large number of parameters being estimated, none of the
coefficients are significant at the 5 percent level.
The future acreage in each crop is calculated by multiplying the estimated crop
response (percent change) by the previous year's acreage. For comparison purposes, we
use the USDA estimate for crop acreage in 1992 as our base acreage. Thus, in equation
(4), zjAt is harvested acres in crop j as determined by the USDA analysis. Total irrigated
acreage in the district by crop is determined using the USDA's projections for the
fraction of each crop that is irrigated. Total and irrigated acreage in the district by crop is
presented in table 5. Since this table is directly comparable to table 2, the percentage
difference between each of our estimates and the USDA estimates is included in table 6.
Table 6 presents total projected water use in the district in millions of gallons per day.
These values are determined by multiplying irrigated acres by the acre-feet of water used
per crop in each year. (These numbers are converted to millions of gallons per day in
table 6.) Once again, we use the USDA projections for the acre-feet of water used for
each crop in future years. As a result, our total water use figures will differ from the
USDA estimates in exactly the same proportions as irrigated acreage. These percentage
differentials are presented in table 6.
As is seen in the percent differential section of table 6, our estimates are fairly close
to the USDA estimates for water use over the next 20 years. For the year 1995, our
estimates are 9.1 percent lower than the USDA estimates; in 2000, they are 1 percent
lower; and in 2010 and 2020, they are 15.7 percent and 11 percent higher, respectively.
Table 4. Estimated Coefficients of Acre Response
Standard
Peanut Acreage
Constant 0.0185 0.0638 0.2906 0.3873
Peanut 0.3276 0.7946 0.4123 0.3425
Corn 0.0964 0.2415 0.3989 0.3473
Cotton -0.1632 0.2801 -0.5827 0.7163
Tomatoes 0.1277 0.3165 0.4033 0.3457
Peaches -0.0828 0.2128 -0.3889 0.6490
PFuel -0.0673 0.6812 -0.0988 0.5388
PLabor -1.2164 3.2157 -0.3783 0.6452
POther 0.0404 0.8647 0.0467 0.4816
Corn acreage
Constant -0.0563 0.1064 -0.5293 0.6985
Peanut -0.1730 1.3267 -0.1304 0.5512
Corn -0.3394 0.4033 -0.8416 0.7945
Cotton 0.0404 0.4700 0.0861 0.4662
Tomatoes -0.0585 0.5269 -0.1110 0.5436
Peaches 0.1767 0.3548 0.4981 0.3122
PFuel 0.1965 1.1390 0.1725 0.4325
PLabor 2.4706 5.3635 0.4606 0.3253
POther 0.0586 1.4069 0.0417 0.4836
Cotton Acreage
Constant 0.0393 0.1399 0.2807 0.3911
Peanut -0.7165 1.7410 -0.4116 0.6572
Corn -0.0894 0.5301 -0.1687 0.5660
Cotton -0.0856 0.6145 -0.1393 0.5546
Tomatoes -0.0714 0.6950 -0.1028 0.5404
Peaches 0.2388 0.4661 0.5124 0.3073
PFuel -0.2134 1.5020 -0.1421 0.5557
PLabor 4.6200 7.0414 0.6561 0.2600
POther -2.4851 1.9088 -1.3019 0.8953
Tomato Acreage
Constant 0.0215 0.0827 0.2602 0.3988
Peanut -0.0249 1.0161 -0.0245 0.5096
Corn 0.1298 0.3125 0.4155 0.3413
Cotton 0.0978 0.3594 0.2722 0.3943
Tomatoes -0.2547 0.4073 -0.6254 0.7302
Peaches 0.0450 0.2755 0.1632 0.4361
PFuel -0.1924 0.8757 -0.2197 0.5857
PLabor 0.2116 4.1962 0.0504 0.4802
POther -0.0238 1.0425 -0.0229 0.5090
Orchard Acreage
Constant 0.0085 0.0147 0.5816 0.2840
Peanut -0.0496 0.1830 -0.2710 0.6052
Corn -0.0459 0.0556 -0.8265 0.7903
Cotton -0.0731 0.0646 -1.1328 0.8639
Tomatoes 0.0531 0.0727 0.7304 0.2373
Peaches 0.0374 0.0489 0.7642 0.2273
PFuel 0.1268 0.1571 0.8066 0.2152
PLabor 0.4409 0.7392 0.5964 0.2792
POther 0.0459 0.1976 0,2325 0.4094
Table 5. Total and Irrigated Acres in Geo
1995
Peanuts
Corn
Cotton
Vegetables
Orchards
Other
Peanuts
Corn
Cotton
Vegetables
Orchards
Other
451,536
145,802
91,173
82,541
93,651
585,900
232,813
110,095
55,077
71,148
40,779
94,400
rgia
2000 2010
Total Harvested Acres
467,154 545,395
132,194 86,769
115,826 124,633
94,241 111,440
101,495 101,007
582,350 582,350
Total Irrigated Acres
250,702 310,514
99,840 65,520
72,070 81,953
81,349 172,804
47,392 53,477
94,350 94,250
Table 6. Total Annual
Peanuts
Corn
Cotton
Vegetables
Orchards
Other
Total
Peanuts
Corn
Cotton
Vegetables
Orchards
Other
Peanuts
Corn
Cotton
Vegetables
Orchards
Other
Total
Water Use in Georgia (mgd)
1995 2000 2010
Total Water Use
103.782 114.167 147.438
56.841 55.238 40.194
27.186 38.737 51.536
37.053 48.685 130.138
21.778 32.577 53.269
37.110 43.450 56.090
283.749 332.854 478.664
Percent Water Use
37 40 52
20 19 14
112
76
51
121
97
100
91
11 19
15 20
Percent Differential from USDA Estimates
133 155
69 44
64 68
138 289
105 104
100 100
99 116
2020
566,320
74,874
199,255
114,870
109,840
582,200
340,935
56,550
138,197
100,908
65,246
94,150
2020
163.958
35.266
88.365
77.533
66.290
56.870
488.281
58
12
31
27
23
20
161
38
107
167
113
100
111
Given that our study uses the USDA's figures for the fraction of each crop irrigated and
for the amount of irrigation applied per acre for each crop, this result is not remarkable.
Of more interest, however, are the large differences in projected cropping patterns. On
the basis of projected changes in crop prices and input prices, we determine that
vegetables and peanuts will receive a much larger fraction of total irrigation while corn
will receive much less than is predicted by the USDA. From table 6, our study shows
that, in 1995, 12 percent more water is applied to peanuts than is reflected in USDA
estimates, and 61 percent more is applied in 2020. Vegetables also show an increase in
irrigation in comparison to the USDA figures: 21 percent more in 1995, rising to nearly
three times the amount in 2010, and stabilizing at roughly two-thirds more in 2020. Corn,
on the other hand, is predicted to use much less water than is reflected in USDA
estimates: 24 percent less in 1995 and 62 percent less in 2020.
Table 7 presents our estimate of the amount of variation that it is reasonable to
expect in our water use projections due to weather and other random effects. This
variation is based on standard deviations determined by Moss and de Bodisco
(forthcoming). Using direct pumping data from the NWFWMD and disaggregating total
water into monthly flows, Moss and de Bodisco derive the monthly water requirement by
crop for northwest Florida. Specifically, a system of equations is estimated such that
oj = P1 aij + Pza2j + +Pimaj +ij
(6)j = P21lalj + 22a2j +* ++P2mamj E2j
q2j =P12ia1 +i122a2j +...+P12mamj +Ei2j,
where oij represents total water use per permit per month. The ij coefficients then yield
the average monthly flow by crop. Further, the coefficients for each crop are restricted to
sum to the previously estimated annual levels. Thus, the expected monthly sums equal
the annual water requirements.
The estimated vector of ps in equation (7) possesses an estimated covariance
matrix V(l). Implicitly, this variance matrix contains the variance coefficients for water
use by crop by month. The variance of a particular water use can then be derived as
(7) X'V(p)X=o,
Table 7. Total Random Variation by Crop by Year
Multiple of
Variation 1992 1995 2000 2010 2020
Peanuts 1.28 0.41 0.38 0.33 0.30 0.30
Corn 1.22 0.35 0.29 0.29 0.27 0.26
Cotton 1.29 0.22 0.22 0.23 0.24 0.24
Vegetables 1.11 0.11 0.11 0.12 0.12 0.12
Orchards 1.09 0.06 0.08 0.10 0.14 0.15
Other 2.01 0.12 0.24 0.26 0.27 0.26
Total Annual Variation 1.28 1.33 1.33 1.33 1.32
where X is a vector of crop levels. Substituting crop levels for the district, region, or
country for a particular year into X then yields an estimate of a monthly, seasonal, or
annual variance for water use. Summing the monthly standard deviations calculated
above gives an annual measure of the standard variation. Multiplying this by two and
adding it to the average water use per crop gives a measure of the total annual variation in
water use due to random factors for each crop. The variation for each crop is presented in
the first column of table 7. The product of this variation and each crop's share of water
use equals the variation in total water use attributable to each crop. Summing over all
crops gives the total variation in water use due to random events.
6. Conclusions and Suggestions
The results presented in the section above can be explained by two factors. The
first factor is peanuts. As is seen in table 6, our model is strongly driven by increased
water use in peanuts. This is especially true in comparison to the USDA's projections.
For example, in our study, water applied to peanuts increases 10 mgd between 1995 and
2000. This is larger than the increase that the USDA predicts through the year 2020. Yet
our analysis does not take into account the fact that peanuts are produced under federal
allotment. Should the allotment program end, peanut prices and acreage would be
expected to fall. Should the allotment continue, it is reasonable to assume that peanut
acreage would not change significantly in the future. One proposed modification of this
study would be to reestimate the model, treating peanut acreage as fixed or quasi-fixed.
This should decrease our projections of total water use.
The second factor that explains large increases in water use in Georgia is the
dramatic increases in the fraction of each crop that is irrigated and the increased amount
of water applied per crop per acre. Since our analysis uses the USDA's projections for
these numbers, there is no basis for comparison in this study. Our analysis of water
demand in the North West Florida Water Management District (NWFWMD) found a
close fit between the Census estimates of acre-feet of water applied per acre and the
water management district's permit data on water use per acre per crop. The USDA
predicts that water use per acre will increase an average of 44 percent over Census
estimates by 2020.
While data limitations currently preclude the direct estimation of acre-feet of
water use per crop in the future, a second proposed modification of this study would
provide an estimate of the percentage change in a crop's price that is necessary to make
the USDA's estimated water use economically feasible for the farmer. Specifically, using
the modified model suggested above, we would generate a set of constants that would
make our econometric projections identical to the USDA baseline.
(8) Pi = P, exp(yt)such that Min I (zt z ),
Yi ti
where P*i are the adjusted prices for each crop i, PUi are the baseline prices (assumed by
USDA), the zi are the acreage responses estimated in our model, and ziu are the acreage
responses in the USDA model. USDA acreage responses, ziU, depend on prices P"i,, and
zi depend on P*i The constant yi shows the percent increase or decrease in the price of
the crop necessary for our estimates of crop acreage to be consistent with the USDA
estimates. For example, in our current study the y for corn and cotton in 2010 would be
positive (corn and cotton prices would have to be higher than we predict for our estimates
to match the USDA estimates) while the y for vegetables and peanuts would be less than
zero.
8. Proposals and Costs
A basic time frame for each of the modifications suggested above would be as follows:
1. Create baseline projections, using the acreage response equation and treating
peanuts as a quasi-fixed variable, by September 1.
2. Solve for change in output prices required to equate our acreage response with the
USDA estimates by September 15.
3. Create a variation consistent with the new baseline by September 20.
4. Final report by September 30.
References
Coyle, Barry T. 1993. "On Modeling Systems of Crop Acreage Demands." Journal of
Agricultural and Resource Economics 18(1): 57-69.
Moss, Charles, and C. de Bodisco. Forthcoming. "Projected Water Demand by
Agriculture in the Northwest Florida Water Management District." Economic
Information Report. University of Florida, Food and Resource Economics
Department, Gainesville, FL.
USDA (U.S. Department of Agriculture). 1995. ACT/ACF River Basins Comprehensive
Study: Agricultural Water Demand. Soil Conservation Service, Washington, DC.
|
