| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Dep. of Agricultural Economics, Krannert Building, Purdue Univ., West Lafayette, IN 47907-1145
* Corresponding author (lowenberg-deboer{at}purdue.edu)
Received for publication October 15, 2001.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONPREVIOUS RESEARCHMATERIALS AND METHODSNOTESRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONPREVIOUS RESEARCHMATERIALS AND METHODSNOTESRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
PREVIOUS RESEARCH |
|---|
|
TOPABSTRACTINTRODUCTIONPREVIOUS RESEARCHMATERIALS AND METHODSNOTESRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Agronomic research also indicates that soybean produced in narrow rows can reduce weed competition (Yelverton and Coble, 1991), and reduce the amount of herbicide needed to control weed growth. Mickelson and Renner (1997) found that planting soybean in 19-cm rows reduced the frequency of needed herbicide applications and increased crop profitability. They concluded that soybean planted in narrow rows out-compete weeds for space and light. Nelson and Renner (1998)(1999) provide evidence that, in some situations, soybean produced in narrow rows decrease the amount of herbicide needed, thereby increasing economic benefits to narrow row soybean.
Although there is ample literature with reference to the yield benefits of narrow row soybean, relatively few studies have documented economic benefits. Oriade et al. (1997) found estimated net returns for narrow row (48 cm) superior to wide row (96 cm) spacings for irrigated and nonirrigated soybean. They estimated returns to nonirrigated, narrow row soybean to range from $39 to $77/ha, and net returns to nonirrigated soybean grown in wide rows to range between $23 and $66/ha. Oplinger (1980) found a $13/ha increase in net returns with row width reduction from 76 to 25 cm. Lambert and Lowenberg-DeBoer (2001) estimated net returns per hectare to soybean planted in 38- and 50-cm row spacings $13/ha more than soybean planted in 76-cm rows. Combining soybean production data from Illinois, Indiana, Ohio, Iowa, and Nebraska, Paszkiewicz (1998) found that soybean planted in 19- and 38-cm rows yielded 144 kg/ha (6%) over 76-cm row spacings. In that report, net returns from narrow row soybean (19 and 38 cm) were superior to soybean produced in 76-cm rows ($121, $120, and $114/ha, respectively).
Reported Agronomic and Economic Benefits of Narrow Row Corn
Agronomic research on corn row width has been going on since at least the early 20th century (Hume et al., 1908), and potential yield benefits of narrow row corn were recognized in the 1940s (Bryan et al., 1940). Lambert and Lowenberg-DeBoer (2001) reported that yields of corn grown in narrow rows was 3.2% greater than corn produced in wide rows in the Corn Belt region, and in northern Iowa, corn grown in narrow rows increased yields by 4.2%. However, Lambert and Lowenberg-DeBoer (2001) concluded that corn yield increases due to narrow row spacing are not consistent, and that frequency of stalk breakage increases when corn is planted in narrow rows. They also suggested that insecticide cost for narrow row corn increase substantially because it is applied on a linear foot basis.
Hallman and Lowenberg-DeBoer (1999a) found that equipment costs increased with narrow row corn by about $3/ha. They found that moving from wide (76 cm) to narrow (50 cm) rows increased rootworm insecticide costs by 50%, or $3/ha. For a producer in the northeastern Corn Belt region who had a planter that was used for corn only, they concluded that narrow row corn could generate on average $3.54/ha net returns if rootworm insecticide was not needed. In the northwestern Corn Belt region, estimated net returns to corn grown in narrow rows was $0.81/ha greater than conventional, 76-cm rows. Narrow row corn resulted in lower net returns in the central and southern Corn Belt. They estimated that producers who switched from 76-cm row corn and drilled soybean to a narrow row planter for both crops would reduce equipment costs by as much as $4/ha for soybean and $1.31/ha for corn. They also estimated that a producer who currently uses a 76-cm planter for corn and row splitters for 38-cm row soybean would incur an extra annual equipment cost of about $2.46/ha when moving to 38-cm row corn because of the high cost of custom-built 38-cm row corn heads.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONPREVIOUS RESEARCHMATERIALS AND METHODSNOTESRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
This paper uses publicly available yield data for nonirrigated soybean and corn (Table 1). These data are found in refereed journals, trade magazines, or on the World Wide Web. The corn and soybean data sets can be downloaded at www.agecon.purdue.edu/research/pub_data (verified 20 Feb. 2003). They are the public data available to producers, agricultural consultants, and field agronomists who are faced with the practical problems of row width choice.
|
Another problem using publicly available data is representation of observations across time and space. Yields are not available for each strategy for each year: only 11 out of the 34 yr covered (1966–2000) have both corn and soybean yield points. Additionally, observations may occur more frequently in certain regions than others, and representative sample sizes for different row widths and different crops are unbalanced. For example, for soybean grown in 25-cm rows or less, there are 235 observations (n = 235), whereas for soybean grown in 38-cm widths, n = 78 (Table 1). For the corn studies, n = 122, 52, and 70 for 76-, 38-, and 50-cm corn, respectively. Net yields for corn grown at 38- and 50-cm row widths were not statistically different. Following Hallman and Lowenberg-DeBoer (1999a), narrow row corn data for 38- and 50-cm row widths was combined to overcome the problems of unbalanced sample sizes, limited numbers of observations, and representation across time and regions. Corn and soybean net revenue distributions for each row width were estimated using the data reported in the literature, then combined, resulting in an estimated distribution. Mathematically, the combined distributions for a rotation are in fact a weighted average of corn and soybean distributions at every probability level. Because of the way the corn and soybean CDFs were estimated and because the crops do not have equal numbers of observations, it was necessary to impose a common probability scale and estimate net revenues on that probability scale. The simplest probability scale was one with equal increments; 1% increments were used so the resulting CDF for a system has 100 points. Further details are found in Lambert and Lowenberg-DeBoer (2001). Despite these drawbacks, the publicly available data are the best available to producers and agribusinesses.
Soybean net returns are combined across regions since there were no readily identifiable regional differences between row classes (Lambert and Lowenberg-DeBoer, 2001). None of the soybean trials explicitly include the effects of white mold (Sclerotinia sclerotiorum) or other plant diseases, in spite of the hypothesis that these problems are linked to row width (Edwards et al., 1999; Grau and Radke, 1984; Dorrance et al., 1998). However, limited evidence suggests that even when there are moderate yield losses caused by white mold, yields from soybean produced in narrow rows are still superior to wide row spacings (Butzen, 1998).
The corn part of this analysis assumes that insecticide is not used on corn. Hallman and Lowenberg-DeBoer (1999a) found that added insecticide costs offset any economic benefits from narrow row corn.
Economic profits are distinguished from net returns as the profits from operations less rent to capital and labor. In this analysis, net returns are determined using a partial budget. A partial budget focuses only on those costs and revenues that change when adopting a new practice (Swinton and Lowenberg-DeBoer, 1998). Thus, economic profit is not measured in this analysis because all costs are not included. Net returns to soybean yield were determined using the mean of the prices from the states included in the data set between 1988 and 1999 ($0.11/kg or $5.97/bu, USDA-NASS, 2000). Hauling charges ($0.004/kg or $0.20/bu) and drying charges ($0.002/kg or $0.13/bu, Pierce, 2000) were subtracted from the sale price. Seed costs for drilled soybean were estimated to be $2.48/ha more than the cost of seeding for all other row classes. This extra cost is based on the assumption that extra seed is needed for drilled soybean to cover seed breakage and misplacement (hence lower germination rates) caused by the inaccurate placement of the seed in rows by the drill. Soybean seed was estimated to cost $4.80/50 000 (Lambert and Lowenberg-DeBoer, 2001), or $13.44 for 22 kg (one 50-lb bag) of seed. Glyphosate-resistant soybean seed cost was estimated to be $24.25 for a 22-kg bag of seed (Ohio State Univ., 2000). A seeding rate of 49 to 64 kg/ha at 6160 seed/kg was assumed for 38-, 50-, and 76-cm soybean, while a seeding rate of 83 kg/ha at a similar seed count was assumed for drilled soybean (Purdue Crop Diagnostic Training and Research Center, 1999). Corn prices and regional categories used by Hallman and Lowenberg-DeBoer (1999a) are adopted, where the Ohio Valley Corn Belt region identified in that report is combined with corn yields from the Central region. Corn prices ($/kg) for each region were: NW, $0.042 ($2.33/bu); NE, $0.043 ($2.40/bu); and CE, $0.045 ($2.53/bu). Corn prices are the average within regions from 1988 to 1999.
The soybean production analysis considered weed control programs and their expected costs. For the conventional 76-cm row width, herbicide material and application costs included a preplant treatment of a chlorimuron ethyl/sulfentrazone co-pack (Canopy XL) at $4.91/ha, and 1.5 (half-rate) to 2 postplant applications of fomesafen (Flexstar) at $6.15/ha. Soybean grown in row widths <25 cm, 38, and 50 cm assumed a pre/postplant treatment using the same materials, except at a 1:1 ratio pre/postplant application rate, assuming narrower rows suppress weed growth. In a sensitivity analysis, an herbicide program was assumed for glyphosate-resistant (GR) soybean grown in each row class. Herbicide costs for GR soybean cultivated in conventional row widths assumed at least two postplant applications of glyphosate (2 x $2.87/ha), while GR soybean grown in the other narrow row class widths received only one postplant treatment.
Long run equipment costs for various systems are compared. Transition costs of retrofitting equipment are outside the scope of this analysis. For corn, harvesting equipment must be changed for narrow rows. Narrow row soybean is harvested with conventional equipment. Presently, there is no mass-produced 38-cm row corn head commercially available. The 38-cm corn heads must be custom-built, and hence they are relatively high cost. However, 50-cm row heads are commercially available. Equipment costs for wide and narrow row (50-cm) corn were determined following Hallman and Lowenberg-DeBoer (1999a), and equipment costs for a custom 38-cm corn head were estimated following Hallman and Lowenberg-DeBoer (1999b).
To avoid the issue of differences in timeliness, all planting equipment was assumed to cover a 12.19-m swath for the baseline analysis, and the analysis of regional differences between technologies. Per-hectare equipment cost estimates assume that planting time is the limiting factor and that a producer plans to complete corn and soybean planting in 10 working days (Doster et al., 1997). For a mixed, 50/50, 729-ha corn–soybean system, this implies the capacity to finish planting soybean in 5 working days, assuming use of a 12.19-m wide planter for 38- or 50-cm corn and soybean row widths, or 76-cm corn. For drilled soybean, two 6.1-m (20-ft) wide 25-cm drills pulled with a dolly hitch at an estimated planting rate of 6.2 ha/h (Doster, 1996), and a 12-h work day, 364-ha (900-acres) of soybean could be planted in 5 working days.
Equipment costs for the 10S + 30C system included the cost of a 25-cm soybean drill plus a dolly hitch, combine and tractor tires, a 76-cm corn head, and a 16-row 76-cm planter. Equipment costs for the C/S 15, 20, and 30 systems included the costs of a planter and corn head following each row width specification, and combine and tractor tires following row width specification. A 12.19-m swath is also assumed.
A partial budget spreadsheet was developed to calculate expected net returns between planting technologies for corn and soybean yield data sets. Costs not included in the partial budget include other operating costs (following USDA-ERS, 2002; fuel,1 lube, electricity, custom operations, soil conditioners, and other chemicals), allocated overhead (labor, land rental rates, and general farm overhead). Costs included are fertilizer and herbicide costs, seed costs, drying and hauling costs, and equipment costs. Thus, net returns are returns to the items excluded from the partial budget (land, labor, fixed costs, general overhead, and other operating costs listed above). Equipment cost estimates used commercially available machine list prices (Heartland Ag-Business Group, 2001). Unlike the costs associated with switching from wide (>76 cm) row corn to narrow row corn (<50 cm, Hallman and Lowenberg-DeBoer, 1999a), cost of switching from wide to narrow row widths for soybean would be primarily attributable to planting equipment. A switch to drilled narrow row soybean can cut equipment costs since drills are usually less expensive than planters. Planters designed to plant at 38 or 50 cm are expected to be slightly more expensive than 76-cm planters because they have relatively more moving parts. Equipment costs from wide to narrow row widths assumed the purchase of a drill or the purchase of a planter capable of planting either corn or soybean in 38- or 50-cm row widths.
A sinking fund approach (Hunt, 1995) estimated annual equipment costs with a 10-yr useful life and a 10% interest rate. Insurance and annual property tax was estimated at 0.91%, and repairs were estimated at 4.66% of the list price per 150 h of use for seeding equipment (Doster, 1996). As suggested by Hunt (1995), a salvage value factor of 0.42 was used to estimate depreciation over a 10-yr use period. To account for the fact that farmers often negotiate equipment prices, actual purchase price was assumed to be 85% of the list price.
Mean-Variance and Stochastic Dominance Analysis
Empirical distributions of net returns to production systems were compared using a mean-variance criterion (Richardson, 2002) and stochastic dominance (Mas-Collel et al., 1995). The mean-variance rule assumes that the dominant alternative must have either a higher mean for a given variance or a lower variance for a given mean. The mean-variance criterion assumes no specific levels of risk preference. The stochastic dominance analysis used the spreadsheet approach outlined by Lowenberg-DeBoer et al. (1990) and Hien et al. (1997). Net return per hectare was estimated for each data point using the cost and price assumptions outlined above. The analysis assumed that each observation in the dataset had equal probability of occurring.
Stochastic dominance compares cumulative distributions of outcomes based on two observations about human nature: (i) most people prefer more to less, and (ii) most people prefer to avoid low value outcomes. Observation (ii) implies that humans are generally risk-averse, but is not the same as saying that individuals avoid variability. Most people enjoy upside variability so long as they benefit from the outcomes (e.g., higher yields, higher crop prices, higher profits), but are risk averse to downside variability. Those preferring more to less, but who do not seek to avoid variability are characterized as risk neutral.
These observations are quantifiable in terms of empirical distributions using two decision rules that correspond to the two assertions made above regarding human behavior: first-degree stochastic dominance (FDSD) and second-degree stochastic dominance (SDSD) rules. FDSD assumes decision makers prefer more to less, and states that an alternative is preferred over others if it provides a higher outcome at every level of probability. Expressed graphically, the preferred distribution is always to the right of other distributions.
Second-degree stochastic dominance assumes risk aversion. The area under the empirical distribution is a measure of the propensity of an alternative to have low-value outcomes. An alternative is dominant over others if the area under its empirical density curve is smaller at every outcome level. First-degree stochastic dominance is easier to visually identify than SDSD. In the simple case of a cumulative distribution starting to the right of an alternative distribution and crossing over only once, the distribution to the right at the horizontal axis dominates if the area between the distributions below the crossover is greater than the area between the distributions above the intersection. Anderson (1974) provides discussion and examples of stochastic dominance use in interpreting agronomic research data. Stochastic dominance tests were performed using a spreadsheet developed in Microsoft Excel.
Certainty Equivalents
Certainty equivalents for each system were estimated at different risk tolerance levels. A certainty equivalent value is the amount of money (as net returns per hectare) for which a producer is indifferent between an uncertain outcome and receiving an amount of money-for-certain. Certainty equivalents for risk-averse individuals are always less than the expected payoff of an uncertain project. When faced with alternatives amongst projects with different, uncertain returns, a risk-averse individual would always choose the project with the highest certainty equivalent.
For risk-averse individuals, attitudes and beliefs toward uncertainty are aptly described by concave functions. To estimate certainty equivalents (CE), a concave, constant absolute risk aversion (CARA) utility function U(x) = -e-
x is assumed, where is the Arrow-Pratt absolute risk coefficient (Moschini and Hennessy, 2001). The certainty equivalent (
) is derived with the following equation:
 |
is the average net returns per hectare of a production alternative, and 
is the coefficient of variation of net returns to a technology. This equation is the result of a Taylor series expansion around a mean represented by the utility function U(0 + z), where U(·) is the assumed utility function, 0 is the estimated net revenue of a given alternative, and z is the deviation around that mean. Risk coefficient values assessed here are 0.001, 0.01, 0.10, and the values that drive certainty equivalents to zero. Higher values of represent more risk-averse agents. |
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONPREVIOUS RESEARCHMATERIALS AND METHODSNOTESRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
Risk Analysis
Combined Analysis of Corn–Soybean Systems: Baseline Results
When data are pooled over the Corn Belt, the 10S + 30C system had the highest estimated net return ($96.90 ± 27.01/ha, mean ± standard deviation, Table 3). Mean net return per hectare for the C/S 20 system was slightly lower at $96.09 ± 26.10/ha followed by the C/S 15 ($95.51 ± 25.13/ha), and then the 15S + 30C system ($94.98 ± 25.05/ha). The estimated net return for the C/S 30 system was $88.42 ± 26.12/ha. When the mean-variance rule is applied, the C/S 30 system is dominated by the C/S 15, C/S 20, and 15S + 30C systems. The 10S + 30C system did not dominate any other systems by the mean-variance rule since the relatively high variance of this system.
|
|
Although no instances of stochastic dominance were detected between the C/S 30 and 15 alternatives, net returns per hectare are higher for C/S 15 except under the poorest growing conditions (Fig. 1). Following several crossing of the distributions in the lower tail, net returns for the C/S 15 system surpass returns from the C/S 30 system at the 5th percentile ($36/ha) and are to the right throughout the remainder of the comparison (Fig. 1). Estimated empirical distributions of the C/S 15 and 15S + 30C systems were nearly identical (Lambert and Lowenberg-DeBoer, 2001), and stochastic dominance was not detected when C/S 30 and 15S + 30C were compared. However, like the comparison of the C/S 30 and 15 distributions, the 15S + 30C net returns per hectare dominate the C/S 30 system at all probability levels after several lower tail crossings.
With CEs evaluated at different risk coefficients, results are similar to stochastic dominance rankings. In general, any system using narrow row corn and soybean technologies is superior to the C/S 30 system. As expected, as the level of risk aversion was increased during the analysis, systems producing the most variable returns became less desirable with respect to certainty equivalents. In the combined analysis, the 10S + 30C system was preferred more than the other alternatives assuming low levels of risk aversion (Table 4). The C/S 15 CE value was highest assuming a 10-fold increase in the risk coefficient. Assuming risk aversion levels between 0.25 and 0.26, all systems except the C/S15 and the 15S + 30C systems were no longer attractive to risk-averse individuals. At = 0.26 to 0.30, the 15S + 30C system is the preferred alternative.
|
Sensitivity Analysis
Regional Analysis
Overall, returns to narrow row technology by regions were higher than the combined, baseline results, except in the northwestern Corn Belt region (Table 3). Like the baseline results, the net returns to all systems including narrow row soybean were similar. For the central Corn Belt, rankings of net returns for each technology were the same as the baseline results. However, mean-variance rankings change slightly. Although all narrow row alternative net returns are higher than the C/S 30 system ($92.83 ± 27.60), only the 15S + 30C system is dominant. The others have a higher variance than C/S 30. By the mean-variance criterion, 10S + 30C dominates the C/S 20 system. The 10S + 30C system ($101.31 ± 28.56) maintained FDSD over the C/S 30 alternative, but the baseline FDSD ranking of C/S 20 ($99.73 ± 28.88) over C/S 30 switched to SDSD. In the central Corn Belt region, the 15S + 30C ($99.40 ± 26.54) system was preferred to the C/S 15 ($99.15 ± 27.71) system by the SDSD rule.
Rank order of net returns per hectare, mean-variance, and stochastic dominance results were different in the northeast and northwestern Corn Belt regions than the baseline analysis (Table 3). In the Northwest region, the rank order of the baseline results were maintained, except that C/S 15 and 15S + 30C exchange places at Ranks 3 and 4. However, mean-variance criterion rankings are the same as the baseline order. The drilled soybean/76-cm corn system maintained highest returns ($93.55 ± 28.87), but did not dominate any of the other alternatives by the mean-variance criterion because of its relatively high variance. The remaining combinations with narrow row soybean dominated the conventional C/S 30 system by the mean-variance rule. The 10S + 30C system maintained FDSD over the C/S 30 alternative, but unlike the baseline results, the C/S 20 system ($92.72 ± 27.27) dominated the conventional 76-cm row width by the SDSD order.
In the northeastern Corn Belt region, the C/S 20 system had the highest mean return, but the largest difference between net returns per hectare for the C/S 20, C/S 15, and 15S + 30C systems was $1.97/ha (Table 3). These systems were about $8/ha higher than returns from the wide row alternative. Net returns for the 10S + 30C system were $99.45 ± 27.53/ha. Similar to the Northeast and Northwest results, the FDSD ranking of the C/S 20 strategy over the C/S 30 alternative switched to SDSD. Likewise, the planter with row splitter system dominated the C/S 15 alternative by SDSD in this region. The 10S + 30C system maintained FDSD over the conventional row width alternative. In terms of mean-variance, the C/S 20 system dominated the drilled soybean/76-cm corn system, but not the other alternatives.
Similar rankings of net returns per hectare across all regions were obtained when a small, 364-ha farm using smaller planting and harvesting equipment are used (Lambert and Lowenberg-DeBoer, 2001). The 10S + 30C system was preferred to the other strategies in the Central, Northwest, and combined analysis. In the Northeast region, the C/S 20 system was the preferred alternative.
In the Northwest region, the 10S + 30C system produced the highest certainty equivalents assuming risk coefficients of 0.001 and 0.01 ($93.13 and $89.38/ha, respectively, Table 4). However, individuals that are more risk-averse would prefer the C/S15 system in the Northwest region up to a risk aversion coefficient of = 0.25.
In the Northeast region, the C/S20 alternative was preferred over the other alternatives assuming lower risk coefficient levels ( = 0.001 to 0.01). The C/S15 system was preferred over all other systems. The 10S + 30C system was never preferred when evaluated at different risk level coefficients.
On the other hand, the 10S + 30C alternative produced the highest CE values when evaluated at = 0.001 and 0.01 ($100.90 and $97.23/ha, respectively) in the Central region. At higher risk aversion levels ( = 0.10 to 0.26), the 15S + 30C system was the preferred alternative.
Net Returns at Loan Rates
Net revenue per hectare for corn and soybean systems were recalculated based on 2000 loan rates (Farm Service Agency, USDA, 2000). Estimated mean of loan rates for corn and soybean across states included in this report were $0.03/kg ($1.78/bu) and $0.09/kg ($5.07/bu), respectively. Loan rates are determined by county. The minimum loan rate for each state was used in the average. The rank order of the mean net returns per hectare for the loan rate sensitivity analysis was different than the corn–soybean baseline analysis. Ranking between the 10S + 30C ($69.53 ± 20.30/ha), C/S 20 ($68.10 ± 19.60/ha), and C/S 30 ($61.98 ± 13.8/ha) remained the same, but ranking between the 15S + 30C and the CS 15 systems switched, with returns from the 15S + 30C system higher than the C/S 15 alternative by about $0.70/ha.
At loan rate corn and soybean prices, the choice pattern was identical to the Northwest region (Table 4). The 10S + 30C system produced the highest CE values at two lowest risk aversion levels evaluated, except that at loan rates the value end point was 30% greater than that of the Northwest region C/S 15 system. The C/S15 system was preferred when evaluated at all other risk aversion levels.
Net Returns with Glyphosate-Resistant Soybean
When glyphosate-resistant soybean was considered in combination with corn, net returns to the C/S 20 system were highest ($95.97 ± 26.10/ha) compared with 15S + 30C, C/S 15, and 10S + 30C systems ($94.86 ± 25.05, $95.39 ± 25.13, and $91.87 ± 27.02/ha, respectively, Table 3). Estimated returns to the C/S 30 system were lowest at $88.50 ± 26.12/ha. With GR soybean the drill system drops to the fourth place because it requires the most soybean seed, and glyphosate-resistant seed is expensive. The C/S 20 displayed SDSD over the 10S + 30C and C/S 30 systems. No other instances of stochastic dominance were detected between remaining production systems.
The CS20 system produced the highest CE levels when evaluated at = 0.001 and 0.01 ($95.63 and $92.56/ha, respectively, Table 4). The CS15 system produced higher CE as the risk coefficient level increased, until about = 0.28. At this risk aversion level, the 15S + 30C alternative became the preferred choice.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONPREVIOUS RESEARCHMATERIALS AND METHODSNOTESRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Results calculated using 2000 loan rates for corn and soybean produced similar findings (Table 5). The 10S + 30C strategy ranks highest, closely followed by the C/S 20 strategy. The net return for the C/S 20 glyphosate-resistant (GR) system is $0.93/ha less than the 10S + 30C net returns in the baseline study (Table 5). If a producer considers using GR soybean in his or her operation while moving from a wide row corn–soybean system to narrow row combinations, the corn–soybean planting strategy that yields the highest expected net returns per hectare is the 50-cm alternative. The second highest returns are afforded by the C/S 15 alternative. Returns are highest for these planter systems because relatively less soybean seed is needed for the 20 and 38-cm soybean system than the 25-cm drill alternative.
|
A regional-level analysis produced generally similar rankings, except that the C/S 20 estimated returns were highest in the northeastern Corn Belt region in the baseline Corn Belt wide analysis. These findings are in agreement with Hallman and Lowenberg-DeBoer's (1999a) findings in their narrow row corn study. In the northern Corn Belt region, they found a slight advantage in net returns per hectare of corn grown in narrow (50-cm) rows in the Northwest region ($5/ha), and in the Northeast region ($21/ha). They anticipated that producers who switch from 76-cm row corn and drilled soybean to a system where the same planter was used for corn and soybean, narrow row corn could gain competitive ground if soybean yields could be maintained. The present results suggest that producers in the northeastern Corn Belt region may benefit from adopting a narrow row corn–soybean (50-cm) production strategy in regions where rootworm on first year corn or white mold is not a problem. In other regions, the best alternative may be purchasing a soybean drill, and maintaining a conventional, 76-cm corn row width.
One limitation of the corn–soybean analysis is that it assumed a 50:50 corn–soybean rotation. This assumption overlooks the possibility of combining corn and soybean over the whole farm in different ratios. In this analysis, net returns to corn and soybean systems are sensitive to equipment costs and prices, and changes in costs may change preferred rankings of production alternatives. Another limitation is that only one herbicide program was considered in the estimates. Combinations of herbicides, varied application schedules, and competitive pricing will affect the cost per narrow row system, and ultimately net returns to a particular alternative. Another limitation of this study is that data were not available to quantify yields under pressure from white mold. If narrow row soybean yields are more susceptible to white mold, this may favor the 38- and 50-cm planting systems over use of a drill.
Another limitation of this study is that it does not deal with timeliness of field operations. Timeliness could be assessed with a representative farm linear programming model, but that would be another study requiring development and validation of the farm model. By assuming that all the equipment has the same 12-m width, that it is operated at the same speed, and that only one seeding operation can happen at a time (e.g., because of tractor or labor constraints) this analysis attempts to hold timeliness constant. The issues that would arise if timeliness issues were included are: (1) speed of operation of planters and drills (in come circumstances drills may be operated at higher speeds than planters), and (ii) is there a tractor and labor available to simultaneously plant corn and drill soybean? If drills are operated at higher speeds than planters and/or if it is possible to drill soybean and plant corn at the same time, the results would favor drills more strongly than is currently the case. Additionally, the analysis sidesteps other issues about row choice. For instance, the study essentially assumes conventional tillage. Certain types of reduced till do not work well with narrow rows because they depend on moving most of the crop residue to the row middles to open a space for planting the new crop.
The findings suggest that producers will benefit by switching from wide to any narrow row system. However, it is less clear which narrow row system is in fact superior. In most cases, differences in net returns for any particular narrow row system are <$2/ha. Risk related to resale of narrow row equipment is largely limited to corn; it is not substantial for soybean because no special harvesting equipment is required for narrow row soybean and because narrow row planting of soybean is well established. If narrow row corn systems do not become standard practices, the business risks of adopting the C/S 20 alternative may increase if equipment, especially narrow row cornheads, cannot be resold at a reasonable salvage value, even though this alternative dominated others in terms of risk. In this case, the best alternative remains the 10S + 30C alternative. In all comparisons, systems with narrow row soybean result in higher net returns per hectare than the conventional 76-cm corn–soybean system.
|
NOTES |
|---|
|
TOPABSTRACTINTRODUCTIONPREVIOUS RESEARCHMATERIALS AND METHODSNOTESRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONPREVIOUS RESEARCHMATERIALS AND METHODSNOTESRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. L. De Bruin and P. Pedersen Effect of Row Spacing and Seeding Rate on Soybean Yield Agron. J., May 7, 2008; 100(3): 704 - 710. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. F. Stanger, J. G. Lauer, and J.-P. Chavas The Profitability and Risk of Long-Term Cropping Systems Featuring Different Rotations and Nitrogen Rates Agron. J., January 11, 2008; 100(1): 105 - 113. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. S. Poag, M. Popp, J. Rupe, B. Dixon, C. Rothrock, and C. Boger Economic Evaluation of Soybean Fungicide Seed Treatments Agron. J., November 17, 2005; 97(6): 1647 - 1657. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS |
from publicly available data.
for corn–soybean rotation systems at different row spacing.