Printer Friendly

Economic impact of shelterbelts on groundnut (Arachis hypogaea L.) production in therilands: a decomposition analysis.

The study's objective was to assess impact of shelterbelts on input productivity, economic input use efficiency, and net input savings on southern Indian commercial farms threatened by sand dunes. Groundnut production functions were estimated from farm survey data in areas with and without shelterbelts. After accounting for land removed from production, groundnut yields increased by 15 percent (12 percent attributed to shelterbelts, three percent attributed to input use changes) in areas with mature shelterbelts. Mature shelterbelts induced estimated annual input savings of 1,074 rupees/ha (435 rupees/acre). High establishment costs and delayed benefits discourage private shelterbelt investments, suggesting a role for public investment.

Keywords: Groundnut, India, shelterbelt, and wind erosion

From 1945 to 1990, almost 11 percent of the earth's agricultural land was moderately or strongly degraded, causing substantial reductions in productivity (Oldeman et al., 1990). The result is temporary or permanent reduction in agricultural productivity of land due to chemical, physical, or biological processes. In a developing country like India, wind and water erosion are among the most severe environmental problems, affecting 141 million hectares (348.27 ac) of land (Government of India, 1998). The per capita availability of cultivable land in India declined from 0.48 hectare (1.19 ac) in 1951 to 0.15 hectare (0.37 ac) in 2000, due to increasing population, intensive cultivation, soil degradation and lack of conservation strategies.

In the southern state of Tamil Nadu, 3.82 million hectares (9.44 million ac), constituting 29 percent of the state's geographical area, are affected by soil erosion and other kinds of land degradation (Government of India, 1994). In Tirunelveli and Tuticorin districts, nearly 200,000 hectares (0.49 million ac) are deemed degraded. The coastal areas of the Tirunelveli and Tuticorin districts suffer from wind erosion and moving sand dunes, promoted by a lack of canopy Deposition of sand eroded from other sites is the primary threat to productivity.

Areas totaling 20,170 hectares (49,820 ac) contain wind-eroded soil commonly known as "theri" soil, and these areas are thus referred to as "theries" or "therilands." Therilands are characterized by sandy, neutral-to-acidic, red or dark red soil that is non-sticky non-plastic, and non-calcareous (Manoharan and Kombairaju, 1995). As per the U.S. Department of Agriculture soil taxonomy, theri soils in the study area are classified as Typic Ustipsaments (14,220 hectares or 35,123.4 ac), Typic Ustropepts (1,573 hectares or 3,885.3 ac), Typic Rhodustalfs (3,308 hectares or 8,170.76 ac), Rhodic Palenstalfs (625 hectares or 1,543.75 ac), and Grossarenic Palenstalfs (444 hectares or 1,096.68). Climatically, 50-year mean annual rainfall in the study area is 648 mm per year, most of which is received during October December (431 mm or 16.97 in). The mean temperature range is 32[degrees] C to 40.5[degrees]C, average relative humidity is 67 percent, and the groundwater depth range is 10 to 20 m (32.8 to 65.6 ft). The ave rage wind velocity is 17 km per hour (10.56 mph) with a maximum of 70 km per hour (43.5 mph) (in June).

Soil degradation imposes on-farm and off-farm costs. On-farm impacts include declining yield levels, subsequent higher application rates of chemical inputs, shifts in cropping patterns, and in extreme cases, abandonment of cultivation. Off-farm costs include deterioration of water and air quality, and welfare loss from declining farm productivity (Ashok, 1997).

Shelterbelts formed by planting rows of closely spaced trees are a recommended wind erosion control technique. Public investment funds most wind erosion control programs in India. The State Department of Agricultural Engineering, with funding from the Government of India, first formed shelterbelts, as part of Wind Erosion Control Project, in Tamil Nadu's therilands in 1978-79, and covered 1150.03 hectares (2,840.57 ac) spread over three villages, which is referred to as Phase I in this study. Subsequently, with Danish International Development Agency (DANIDA) assistance, the Department of Agricultural Engineering undertook a Comprehensive Watershed Development Program (CWDP) that included additional shelterbelt formation. Shelterbelts created from 1991 to 1994 are referred to as Phase II in this study Phase II covered nearly 5900 hectares (14,573 ac) spread over 11 villages. Additional program activities, for which data are not yet available, extend from 1995 to 2004.

Shelterbelts were formed by planting three rows of trees, one each of acacia (Acacia nilotica L.), cashew (Anacardium occidentale L.) and neem (Azadirachta indica A. Juss), using intra- row spacing of three meters (10 ft), inter-row spacing of three meters (10 ft), and inter-belt spacing of 160 meters (525 ft). Shelterbelts occupy an average of 5.6 percent of each hectare under cultivation. In all cases, shelterbelts were oriented to inhibit sand dune movement from the southeast. Shelterbelts are expected to improve groundnut yields primarily by inhibiting sand dune movement that encroaches on the short-stature crop, and secondarily by providing windbreaks. The portion of benefits from arresting sand dune movement is independent of a field's distance from the shelterbelt. Yields on land closest to shelter-belts benefit from greater wind protection, but may suffer from shelterbelts' competition for scarce water and nutrients (Lyles et al., 1984).

Wind erosion control via shelterbelts potentially offers significant economic benefits in many locations throughout the world. Kort (1988) summarized over three-dozen shelterbelt studies, only three of which reported yield decreases for sheltered crops. Net present value analyses in the Great Plains have been mixed. McMartin, et al. (1974) found that wheat yield increases were insufficient to compensate for land taken out of production, but Brandle, et al. (1984) and Brandle, et al. (1992) found positive net present values for Nebraska wheat over a range of conditions. Economic evaluations are rare in desert environments. This study contributes an analysis of primary data from a long-running shelterbelt program in a commercial agricultural setting. As empirical evidence from site-specific, crop-specific applications accumulates, decision makers in similar environments will be better equipped to assess the benefits of shelterbelts.

The overall objective of the study was to determine if shelterbelts induce a statistically and economically significant increase in groundnut productivity in therilands. The analysis addressed groundnut production because it is the predominant crop in the study area, accounting for 34 percent of total crop value. Specific objectives of the study were to:

1. decompose total changes in groundnut yield into selected components that include shelterbelt status and input use,

2. estimate the value of inputs saved in groundnut cultivation with shelterbelts, and

3. based on the results, provide recommendations regarding shelterbelt establishment in areas threatened by wind erosion and shifting sand dunes.

While data allow analysis of private benefits, objective three was constrained by limited information on shelterbelt costs. Estimation of public benefits from controlling erosion and desertification was beyond the scope of this study

Methods and Materials

Tuticorin district was selected as the study area because its therilands contain the largest area protected by shelterbelts under the CWDP. Three villages were selected, one in the Phase I area (647 hectare or 1,598.1 ac)) where the shelterbelts were established in 1978-79, one in the Phase II area (1258 hectare 3,107.26 ac) where shelterbelts were established in 1991-94, and one in an area not protected by shelterbelts (1301 hectare or 3,213.47 ac). The villages were chosen to maximize the similarity of agronomic, climatic, and cultural factors other than shelterbelt status, because farm-level data were not available with which to control for these variables. Thirty respondents from each village were randomly selected to participate in the study. Data collection from the sample respondents occurred during January and February 2000, by which time the groundnut harvest was complete. In therilands, groundnut crops are rain fed, with a ten-year average annual rainfall of 721 mm (28.4 in).

Primary data were collected from respondents in pre-tested interviews. The purpose of the study was explained to farmers to ensure their cooperation and encourage accurate responses. Primary data included farmers' age, education, land holding, cost and returns of groundnut cultivation, input use, yield, credit availability, and marketing techniques. The study differs from most shelterbelt studies in that the data collection effort emphasized economic production relationships, and agronomists will be disappointed at the lack of agronomic data available for analysis. Errors in data collection, such as recall bias, were minimized through crosschecks.

Decomposition analysis, allocative efficiency, and value of inputs saved. Determining the differential impact of shelterbelts on yield required a statistical decomposition of factors affecting yield into distinct components. Decomposition techniques have been used in a variety of agricultural applications, including decomposition of output variance (Hazell, 1984), influence of technical change (Kalirajan et al., 1996), impact of seed variety adoption (Kiresur, 1995), and impacts of policy reforms (Fan et al., 1997). For the present study, the decomposition model suggested by Bisaliah (1977, 1978) was most appropriate. A production function was first specified by expressing groundnut yield Y as a function of land, labor, fertilizer, and plant protection chemical input quantities.

A variety of production functions, including translog, quadratic, and log-linear, displayed similar explanatory power. The log-linear production function was selected for subsequent analysis because it returned robust estimates in all models and was the most tractable. In a log-linear model, parameter estimates represent input elasticities, e.g., in equation (1) below, [a.sub.1] denotes the percentage change in yield ([Y.sub.1]) given a one percent increase in labor ([L.sub.1]). The production function was specified on a per hectare basis. Fertilizer price is regulated by the government, so the per unit cost was equal for all farmers. Wages paid on a per day basis were different for men and women; hence the wage rate was rescaled to reflect mandays.

ln [Y.sub.1] = ln[A.sub.1] + [a.sub.1]ln[L.sub.1] + [b.sub.1]ln[F.sub.1] + [c.sub.1]ln[P.sub.1] + [u.sub.1] (1)

ln [Y.sub.2] = ln[A.sub.2] + [a.sub.2]ln[L.sub.2]+[b.sub.2]ln[F.sub.2]+ [c.sub.2]ln[P.sub.2] + [u.sub.2] (2)

ln[Y.sub.3] = ln[A.sub.3] + [a.sub.3]ln[L.sub.3] + [b.sub.3]ln[F.sub.3]+ [c.sub.3]ln[P.sub.3] + [u..sub.3] (3)


Y = yield (kg/ha.)

L = labor (mandays/ha.)

F = fertilizer (rupees/ha.)

P = plant protection chemicals (rupees/ha.)

A = scale parameter

a, b & c regression parameters (input elasticities)

u = random disturbance term.

Equations (1), (2), and (3) represent the production relationship during 1999-2000 in areas where shelterbelts were formed in 1978-79, 1991-94, and without shelterbelts, respectively.

Algebraic manipulation of (1) and (3), comparing Phase I to no shelterbelts, implies:


ln [Y.sub.1] - ln [Y.sub.3] = (ln [A.sub.t]-ln [A.sub.3]) + ([a.sub.1] ln [L.sub.1] - [a.sub.3] ln [L.sub.3] + [a.sub.1] ln [L.sub.3] - [a.sub.1] ln [L.sub.3] - [a.sub.1] ln [L.sub.3]) + ([b.sub.1] ln [F.sub.1] - [b.sub.3] ln [F.sub.3] + [b.sub.1] ln [F.sub.3] -[b.sub.1] ln [F.sub.3] + ([c.sub.1]ln [P.sub.1] - [c.sub.3] ln [P.sub.3] + [c.sub.1] ln [P.sub.3] - [c.sub.1] ln [P.sub.3] + ([u.sub.1] - [u.sub.3])

Further rearrangement yields:


ln ([Y.sub.1]/[Y.sub.3]) = ln ([A.sub.1]/[A.sub.3]) + [([a.sub.1] - [a.sub.3]) ln [L.sub.3] + ([b.sub.1]-[b.sub.3] ln [F.sub.3] + ([c.sub.1] - [c.sub.3]) ln [P.sub.3])] + [([a.sub.1] ln ([L.sub.1]/[L.sub.3]+ [b.sub.1] ln ([F.sub.1]/[F.sub.3]) + [c.sub.1] ln ([P.sub.1]/[P.sub.3])] + ([u.sub.1] - [u.sub.3]

The equation for ln([Y.sub.2]/[Y.sub.3]) is analogously expressed.

The dependent variable in equation (5) represents the difference in the log of yields between areas with and without shelter- belts. Yield differences may stem from three sources: shelterbelts' direct ability to increase yield, shelterbelts' indirect ability to raise yield by making inputs more productive, and yield changes unassociated with shelterbelts but due merely to differences in input use.

The first term is the difference in logged scale parameters, indicating a vertical shift in the production function attributable to shelterbelts and any omitted variables correlated with shelterbelt status. If the term is positive, it indicates an upward shift in the production function that suggests shelterbelts directly improve yields at any given level of input use. The following bracketed expression is the sum of changes in input elasticities, each weighted by the log of input quantity used in the control scenario without shelterbelts. If the term is positive, it suggests that shelter- belts (and omitted variables correlated with shelterbelt status) improve the productivity of inputs themselves, as evidenced by higher input elasticities. The last bracketed expression is the sum of differences in the log of input quantities between areas with and without shelterbelts each weighted by the input's elasticity corresponding to the treatment scenario with shelterbelts. This term represents change in yield due t o changes in the per hectare quantities of labor, fertilizer and plant protection chemicals used (i.e., not attributable to shelterbelts).

Note that the first two terms on the right-hand side of (5) measure the contribution of shelterbelts as well as omitted variables correlated with shelterbelt status. Omitted variables are almost a certainty in any statistical model, and this model omits numerous agronomic, climatic, and cultural variables that affect yields. However, none are expected to be highly correlated with shelterbelt status, as prevailing wind direction rather than soil type was the chief criterion in shelterbelt placement, and the close proximity of the study areas subjected them to the same weather patterns. If the shelterbelts induced changes in cultural practices, the impact of such changes should properly be attributed to the shelterbelts. The importance of agronomic factors for which data were unavailable should not be discounted, but such omissions affect the estimated variance more than the estimated level of shelterbelt impacts.

Allocative efficiency implies that farmers should continue using inputs as long as the additional revenue from using another unit of input equals the cost of buying that unit. In areas with and without shelterbelts, allocative efficiency can be evaluated using the production function parameter estimates. Assuming farmers are price takers in input and output markets, the efficiency condition requires a unitary ratio of the value of marginal product (VMP) and input price (iv). Intuitively, the VMP is the revenue gained by using one more unit of a given input. Parameter estimates represent input elasticities ([epsilon]), i.e., the percentage change in yield given a one percent increase in input quantity. The efficiency condition [VMP.sub.i]/[w.sub.i] = 1 for input [x.sub.i] is algebraically equivalent to [[epsilon].sub.i]Y [P.sub.y]/x [w.sub.i] 1, where Y denotes yield, and [P.sub.y] denotes the price of the output Statistically significant deviations from unity suggest allocative inefficiency.

The value of inputs saved by planting shelterbelts was estimated by first calculating the value of inputs required to produce, on unprotected land, the yields that were observed in the shelterbelt areas. The difference between this value and the value of inputs actually used in protected lands represents the value of inputs saved due to technical change induced by shelterbelts (Schultz, 1953). Technical change refers to production process changes, or introduction of new products, such that more and/or improved output can be obtained from the same bundle of inputs (Samuelson and Nordhaus, 1995). The following expression allows estimation of the value of inputs saved in areas with shelterbelts:

[]=(1+[gamma]/100) []-SMC (6)

[S.sub.r]=([gamma]/100) []-SMC, (7)


[]=value of inputs used to produce the yield observed on protected land, accounting for land taken out of production.

[] = value of inputs required to produce, on unprotected land, the yield that was observed on protected land,

[gamma] = percent yield increase attributed to shelterbelts, accounting for land taken out of production,

[S.sub.r] = value of inputs saved attributed to shelterbelts, and

SMC = shelterbelt maintenance costs.

Results and Discussion

The sample in each of the three study areas (30 observations each) was divided at the median farm size value into two sub samples. Farms with area of fewer than four hectares (10 ac) were categorized as small farms and those with more than four hectares (10 ac) were categorized as large farms. Farm sizes ranged from 0.5 hectare (1.24 ac) to almost 20 hectares (50 ac) with an average area of 5.3 hectares (13.1 ac), 4.8 hectares (11.9 ac) and 4.6 hectares (11.4 ac) in the phase I, phase II and unprotected areas, respectively The inputs used and yield of groundnut in the sample farms are shown in Table 1; seeds and labor account for the major share in the cost of groundnut production. Coefficients of variation indicate that, with the exception of plant protection chemicals, input use and yields were not highly variable across the farms within each shelterbelt regime. Yields were highest, as expected, in phase I areas at 1410.48 kg/ha (1,256.30 lb/ac), and lowest at 1135.93 kg/ha (1,011.76 lb/ac) in unprotected a reas. Total profits were 7,672 rupees, 5,661 rupees, and 3,531 rupees per hectare respectively (3,106, 2,292, and 1,430 rupees per ac respectively) in phase I, phase II and unprotected areas. F-values from two-factor ANOVA for the size groups and the three study areas were 2.70 and 24.58 respectively The tests indicated that farm size did not affect yield, but shelterbelt status did. The joint equality of production function coefficients across shelterbelt regimes was tested via Chow tests. F-values of 200.07 (Phase I vs. unprotected), 213.41 (Phase II vs. unprotected) and 31.23 (Phase I vs. Phase II) indicated significant differences between estimated coefficients (i.e., input elasticities) in each pair of regression equations.

As shown in Table 2, the estimated production functions explained 60, 40, and 58 percent of yield variation in the Phase I shelterbelt areas, Phase II shelterbelt areas, and unprotected areas, respectively Explanatory power was lower than would be expected in experimental plots, most likely due to lack of farm-specific agronomic data and potential measurement error in farmer--supplied data. All statistically significant input elasticity estimates were positive, as expected. Labor was significant in all three regressions, and was substantially more elastic in unprotected areas (0.45) than in protected areas (0.16 and 0.06). Fertilizer was significant in the two areas protected by shelterbelts, with input elasticities of 0.24 and 0.26. Plant protection chemical inputs were significant only in the Phase I area, with the longest period of shelterbelt protection, and the estimate was highly inelastic.

Allocative efficiency tests, shown in Table 3, indicate that consistency with the efficiency condition ([VMP.sub.i]/[w.sub.i] = 1) was the exception rather than the rule. Labor in areas without shelterbelts appeared to be the only input used in an allocatively efficient quantity. The remaining inputs with significant production function parameter estimates appeared to be either over-utilized (labor) or under-utilized (fertilizer and plant protection chemicals). One can expect such a result in areas characterized by binding input supply and/or liquidity constraints, and few employment opportunities.

Table 4 presents results of the decomposition analysis, suggesting that shelterbelts were the dominant cause of productivity differences. After accounting for 5.6 percent of land removed from production for shelterbelt plantings, estimated Phase I area yields were 15.3 percent higher than in unprotected areas, and estimated Phase II area yields were 6.2 percent higher than in unprotected areas. Changes in labor, fertilizer, and plant protection chemical input use accounted for only 3.2 percent and 1.7 percent of the Phase I and Phase II productivity increases, respectively. Of these inputs, increased fertilizer use appeared to be the most influential, consistent with its relatively high factor elasticity. The remaining 12.1 percent (Phase I) and 4.5 percent (Phase II) yield increases were attributed to shelterbelts. Note that shelterbelt-induced productivity increases reflect 17.8 (Phase I) and 10.1 (Phase II) percent productivity increases on land that was cultivated, reduced by the 5.6 percent of land devot ed to shelterbelts.

Shelterbelt productivity gains increased with age, as the shelterbelts provided better protection against shifting sand dunes. The more established Phase I areas enjoyed an 8.8 percent yield advantage over areas containing the more recently established Phase II shelter-belts. Maturation of the shelterbelts contributed a 7.7 percent yield increase, and input use changes accounted for the remaining 1.1 percent increase.

After accounting for 5.6 percent of land taken Out of production by shelterbelt plantings, the values of inputs used to produce the average observed yield in Phase I and Phase II areas were 9,726 rupees/ha (3,938 rupees/ac) and 10,182 rupees/ha (4,122 rupees/ac), respectively. Shelterbelt maintenance costs were estimated at 1,906 rupees per hectare (772 rupees/ac) of shelterbelt. Costs consisted of shelterbelt repair (824 rupees/ha or 334 rupees/ac), new bore-well installation and repair of existing wells (635 rupees/ha or 257 rupees/ac), observation and research trials (64 rupees/ha or 26 rupees/ac), and overhead charges (383 rupees/ha or 155 rupees/ac). Shelterbelts occupied 5.6% of each cultivated hectare, on average, thus implying maintenance costs of 107 rupees per cultivated hectare.

Applying the productivity estimates (from the first line of Table 4) and maintenance costs to equation (7) implies that, relative to unprotected areas, shelterbelts induced input savings valued at 1,074 rupees/ha (435 rupees/ac) in Phase I areas and 348 rupees/ha (141 rupees/ac) in Phase II areas. Regarding Phase I versus Phase II, 15 years of shelterbelt maturation allowed annual input savings valued at 648 rupees/ha (262 rupees/ac).

Summary and Conclusion

Policy implications. The results of this study contribute much, but not all, of the information needed to make a strong argument for shelterbelt establishment in semi-arid areas characterized by wind erosion and shifting sand dunes. Shelterbelts substantially enlarged the production possibilities set and allowed economically significant input savings. With much of the world's arable land declining in productivity, the double-digit percentage gains in yield offered by mature shelterbelts present an attractive management alternative. Unlike new seed varieties, manufactured chemical inputs, or mechanized equipment, shelterbelts represent a technology that is locally accessible worldwide. Shelterbelts are a flexible technology that can be adapted to diverse conditions via indigenous species selection. Where feasible, fruit and nut trees planted in shelterbelts can themselves yield economic returns. Shelterbelts can be essentially permanent structures that require initial investments followed by declining maintena nce costs over time, whereas the benefits grow as the trees mature and extend for decades until replanting becomes necessary.

While the data used in this study offered a rare opportunity to assess the private benefits of shelterbelts, currently unavailable data on capital costs and the rate of decline in maintenance costs are necessary to make definitive statements about the expected net present value of establishing shelterbelts in similar environments. In cases where capital and initial maintenance costs are paid by governments or international aid organizations, farmers and local economies clearly face incentives to support shelterbelt establishment. Absent outside investment, however, the initial years of negative net benefits will discourage private investment, particularly if discount rates are high. The failure to renovate Great Plains shelterbelts established during the Dust Bowl years (Annou and Pederson, 2000) suggests this is often the case.

Shelterbelts appear to induce sufficient productivity gains in environments featuring wind erosion and moving sand dunes to warrant serious consideration as public investments. Shelterbelts offer public, as well as private, benefits. Foremost among these is maintaining soil productivity for future generations. Public discount rates are lower than private discount rates (some economists argue that the public discount rate should be zero), implying a lower net present value "hurdle" that must be cleared to justify investment. A public net present value calculation would require arbitrary assumptions about the human costs of land degradation and desertification, and reasonable people might well assume values that justify considerable initial investments to avoid such costs.
Table 1

Mean input use and groundnut yield per cultivated hectare (a), by level
of shelterbelt protection

                                    Phase I                  Phase II
                                    (n = 30)                 (n = 30)

                            Quantity      Value     Quantity     Value
                             kg/ha        rs/ha      kg/ha       rs/ha

Seeds                        116.42        3260.00   124.82      3495.00
                                        (4.37) (b)                (4.39)
Labor (mandays)              112.00        5600.00   116.00      5800.00
                                           (11.01)                (6.05)
  (i) nitrogen                17.15         241.09    18.72       253.24
 (ii) phosphorus              43.83         616.18    47.85       607.31
(iii) potassium               31.38         169.98    36.00       195.00
 (iv) gypsum                 347.78         260.84   327.90       245.93
Total fertilizer                 --        1288.09       --      1301.48
                                            (5.65)                (3.62)
Plant protection chemicals       --         157.89       --       191.95
                                           (53.41)               (29.00)
Total input cost                          10305.98              10788.43
Yield (pods)                1410.48       17977.50  1290.15     16449.40
                                            (4.85)                (3.27)

                               Without shelterbelts
                                     (n = 30)

                            Quantity       Value
                             kg/ha         rs/ha

Seeds                        122.50        3430.00
Labor (mandays)              118.00        5900.00
  (i) nitrogen                25.74         332.99
 (ii) phosphorus              40.21         584.47
(iii) potassium               41.42         224.36
 (iv) gypsum                 342.43         256.82
Total fertilizer                 --        1398.64
Plant protection chemicals       --         223.65
Total input cost                          10952.29
Yield (pods)                1135.93       14483.10

(a) Values are expressed on a "cultivated hectare" basis to enable
comparisons and for use in further analysis; actual values are 5.6
percent lower in Phase I and Phase II areas due to land taken out of
production by shelterbelts.

Table 2

Log-linear production function coefficient estimates.

                            Dependent variable: In (yield)

                  Phase I            Phase II         shelterbelts
                  eqn. (1)           eqn. (2)         eqn. (3)

Constant          4.0856 **          4.3935 **        4.8938 **
                  (0.7474) (a)       (0.8370)         (0.7730)
Labor (mandays)   0.1642 **          0.0604 *         0.4536 **
                  (0.0608)           (0.0269)         (0.0935)
Fertilizer (Rs)   0.2431 *           0.2636 *         -0.1030
                  (0.1152)           (0.1312)         (0.1553)
Plant protection  0.0224 **          0.0005           0.0287
chemicals (Rs)    (0.0050)           (0.0031)         (0.0149)
[R.sup.2]         0.60               0.40             0.58

* and ** denote statistical significance at the .05 and .01 levels,

(a)standard errors in parentheses.

Table 3

Value of marginal product and factor price ratio

                            Phase I  Phase II  shelterbelts

Labor                       0.53 **  0.17 **       1.11
Fertilizer                  3.39 **  3.33 **        --
Plant protection chemicals  2.55 **     --          --

** Denote significantly non-unitary values at the .01 levels.

Table 4

Decomposition of productivity differences

                                Phase I    Phase II --  Phase I --
Sources of difference          -- WS (a)       WS        Phase II

Due to shelterbelts             12.15 (b)     4.48 (b)     7.66
Due to difference in input
 use level
(a) Labor                        0.77         0.08         0.57
(b) Fertilizer                   1.62         1.63         0.11
(c)Plant protection chemicals    0.77         0.00         0.43
Total estimated difference in
 productivity                   15.30         6.18         8.77

(a)WS denotes area "without shelterbelts."

(b)Accounts for 5.63 percent of land taken out of production by
shelterbelt plantings; productivity difference on the cultivated portion
is 17.77 percent (Phase I) and 10.10 percent (Phase II).

References Cited

Ashok, K.R. 1997. An Economic Analysis of Soil Conservation - Productivity, Sustainability, and Constrains. Unpublished Ph.D., thesis, Tamil Nadu Agricultural University, India.

Annou, M. and G. Pederson, 2000. The Economics of Windbreak Renovation. Agricultural Economics Newsletter. University of Minnesota Extension Service, Spring 2000 edition.

Bisaliah, S. 1977. Decomposition analysis of output change under new production technology in wheat farming. Indian Journal of Agricultural Economics 23(3): 193-201.

Bisaliah, S. 1978. Decomposition analysis of employment change under new production technology in Punjab agriculture. Indian Journal of Agricultural Economics 33(2):70-80.

Brandle, J.R., B.B. Johnson, and T. Akeson. 1992. Field windbreaks: Are they economical? Journal of Production Agriculture 5:393-398.

Brandle, J.R., B.B. Johnson. and D.D. Dearmont. 1984. Windbreak economics: The case of winter wheat production in Eastern Nebraska. Journal of Soil and Water Conservation 39(5):339-343.

Fan, S., E.J. Wailes, and K.B. Young. 1997. Policy Reforms and Technological Change in Egyptian Rice Production: A Frontier Production Function Approach. Journal of African Economics 6(3):391-411.

Government of India. 1994. Indian Agriculture in Brief. (25 ed.) Directorate of Economics and Statistics, Department of Agriculture and Co-operation. 25.

Government of India. 1998. Indian Agriculture in Brief. (26 ed.) Directorate of Economics and Statistics, Department of Agriculture and Co-operation.

Hazell. P.B.R. 1984. Sources of Increased Instability in Indian and US Cereal Production. American Journal of Agricultural Economics 66(3):302-311.

Kalirajan, K.P, M.B. Obwona, and S. Zhao. 1996. A decomposition of TFPG: The ease of Chinese agricultural growth before and after reforms. American Journal of Agricultural Economics 78(2):331-338.

Kiresur. 1995. Technological change in sorghuna production: An economic study of Dharwad farms in Karnataka. Indian Journal of Agricultural Economics. 50(2):185-192.

Kort, J. 1988. Benefits of windbreaks to Field and Forage Crops. Agriculture, Ecosystems and Environment 22/23:165-190.

Lyles, L., J. Tatarko, and J.D. Dickerson. 1984. Windbreak Effects on Soil Water and Wheat Yield. Transactions of the American Society of Agricultural Engineers 27:69-72.

Manoharan, M. and S. Kombairaju. 1995. ITK Suits Transported Sandy Soils. Indigenous Knowledge and Development Monitor 3,1: April online issue.

McMartin, W., A.B. Frank, and R.H. Heintz. 1974. Economics of shelterbelt influence on wheat yields in North Dakota. Journal of Soil and Water Conservation. 29(2):87-91.

Oldeman, L.K, R.T.A. Hakkeling. and W.G. Sombroek. 1990. World Map of the Status of Human-Induced Soil Degradation: An Explanatory Note, 2nd ed. Wageningen, Netherlands: International Soil Reference and Information Centre.

Samuelson, P.A and W.D. Nordhaus. 1995. Economics. New York: McGraw Hill.

Schultz, T.W. 1953. Economic Organization of Agriculture. New York: McGraw Hill.

Venkat N. Veeramani is a graduate research assistant in the Department of Agricultural Economics at the university of Kentucky in Lexington. Kentucky. Leigh J. Maynard is an assistant professor in the Department of Agricultural Economics at the university of Kentucky in Lexington, Kentucky. chandrasekaran Murugappan is an associate professor in the Department of Agricultural Economics at the Tamil Nadu Agricultural university in Tamil Nadu, India.
COPYRIGHT 2003 Soil & Water Conservation Society
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2003 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Veeramani, V.N.; Maynard, L.J.; Murugappan, C.
Publication:Journal of Soil and Water Conservation
Geographic Code:9INDI
Date:Mar 1, 2003
Previous Article:Applications of land evaluation and site assessment (LESA) and a geographic information system (GIS) in East Park County, Wyoming. (Research).
Next Article:Soil information for a changing world: a viewpoint from the soil and water conservation society's executive director.

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters