Introduction
Organic retail food sales have consistently grown approximately 20% annually since 1990Reference Dimitri, Green and Wellson1, although certified organic cropland still only accounted for only 0.5% of agricultural lands in the US in 20052. Fresh produce is the top-selling category in organic food production, and accounted for 42% of all organic food sales in 2001, with sales increasing 51.4% from 1999 to 2000Reference Dimitri, Green and Wellson1. In 2005, Georgia had 439,660 ha in vegetable productionReference Boatright and McKissick3, and ranks in the top four states nationally for fresh market vegetables in area harvested, production and value4. Additionally, the southern and western regions of the US have been identified as the two fastest-growing organic markets in the countryReference Stevens-Garmon, Huang and Lin5. Despite the market potential for organic production in the state, in 2005, Georgia ranked 31st in the country in the number of certified operations and 43rd in the total number of certified organic acres, with only 53 certified organic operations in the state, totaling 2413 ha2.
Economic decision-making tools may help alleviate the risk associated with converting to new production practices; enterprise budgets are one such tool. Enterprise budgets estimate the costs and returns associated with the production of a commodity, or enterpriseReference Chase, Smith and Delate6. While common for conventionally produced commodities, few enterprise budgets exist for organic agriculture, and even fewer for the Southeast. There are enterprise budgets for organic commodity crops in the Midwest, and for vegetables in California, Wisconsin and New JerseyReference Born7. Some production cost information for large-scale organic vegetable production has been assembled for the southeastern US8, and is applicable to the large-scale farms in the southern coastal plains. Only one production cost study was found for the region that would be applicable to the small, diversified organic farms of northern Georgia PiedmontReference Estes, Kleese and Lauffer9.
Although enterprise budgets allow the comparison of input costs and returns between systems, they do not account for changes in the environmental quality associated with production systems. Centuries of tillage-intensive agriculture have left the soils of Georgia Piedmont severely eroded, lacking topsoil, soil organic matter and any appreciable nutrientsReference Rhoades, Nissen and Kettler10. In order to restore sustained production to degraded soils, farmers must also restore soil quality, specifically the soil organic matterReference Jordan11. These challenges are intensified by the region's subtropical climate, which contributes to rapid pest and disease outbreaks, high weed pressure and rapid decomposition of soil amendments.
The purpose of this work was to provide an economic analysis of a field experiment that examined the ability of two experimental ecological agricultural systems to restore degraded soils in the southeastern USReference Jacobsen and Jordan12. A 3-year field study was conducted to assess the effects of organic farming systems on soil characteristics, crop production and weed biomass. The two experimental systems were based on best management practices and recommendations from previous agroecological research in the region, including conservation tillage, incorporation of perennial legumes, crop rotation and use of winter cover crops and composts. This analysis was intended to assess the economic costs and benefits of the experimental systems on degraded soils using an enterprise budget approach and to discuss the results from a variety of crops in the rotation using a stochastic dominance analysis. Production economics for these experimental techniques is not intended to be representative of organic conservation tillage systems on fertile soils, but rather to contribute to the discussion of the potential profitability of experimental technologies on degraded lands. Enterprise budgets are presented here to contribute to basic production economics information that is lacking for the region as an aid in the grower's decision-making. Stochastic dominance analysis is employed to compare the risk-return trade-offs of the 10 production systems from crop productivity, economic profitability, to identify risk-efficient systems and discuss these results in the light of desired changes in parameters related to soil quality (soil carbon).
Production practices and methods
The study site was located on the Spring Valley Ecofarm, near Athens, Georgia, USA (33°57′N latitude, 83°19′W longitude). This historic farm had been in cultivation since 1864, with cotton, cattle, sorghum and soybean previously grown on the site until 1993. At this time, the site was removed from extensive cultivation and management shifted to a mowed fallow of pasture grasses and weeds. An alley cropping (AC) system utilizing perennial legumes was established on the site in 2001. AC is a technique where hedgerows of trees or shrubs are planted between rows, or alleys, of crop plantsReference Kang, Ghuman, Moldenhauer, Hudson, Sheng and Lee13. This technique is often practiced in the tropics where leguminous hedgerow species are used as a perennial source of crop fertility, animal fodder and erosion controlReference Rao, Ong, Pathak and Sharma14–Reference Long and Nair16. In these systems, shoots of hedgerow plants are coppiced, and the pruned branches applied as a mulch and green manure for adjacent crop plants.
In 2004, an experiment was initiated that compared three cropping systems over a 3-year vegetable crop rotation for their effects on soil characteristics, plant production and yieldsReference Jacobsen and Jordan12. These systems included two organic systems in conservation tillage, and one tillage-intensive conventional system, outlined further below. The subheadings below the cropping systems describe experimental treatments which employed three levels of compost for nutrient supply (0, 22.4 and 44.8 Mg ha−1 yr−1), as well as two levels of mulch (with or without), for weed suppression. For a complete description of the experimental design and rationale, see Jacobsen and JordanReference Jacobsen and Jordan12.
The three cropping systems and their treatments are as follows:
1. AC with organic vegetables using strip tillage (AC treatments) are as follows:
(a) AC1: AC (including hedgerow prunings) with winter cover crops.
(b) AC2: AC1+straw mulch.
(c) AC3: AC1+spring compost.
(d) AC4: AC1+spring compost+mulch.
(e) AC5: AC1+fall compost+spring compost.
(f) AC6: AC1+fall compost+spring compost+mulch.
2. Organic vegetables using strip tillage [no AC, organic strip tillage (OST) treatments] are as follows:
(a) OST1: winter cover crops+mulch.
(b) OST2: OST1+spring compost.
(c) OST3: OST1+fall compost+spring compost.
3. Conventionally fertilized, conventionally tilled vegetables [conventional tillage (CT) treatment].
A general outline of the annual management practices for each treatment is detailed in Table 1. In AC treatments, leguminous perennial hedgerows consisted of Albizia julibrissin planted in hedgerows 5 m apart with plants 0.5 m apart within the hedgerow. Hedgerows were coppiced using a hedge trimmer 1 to 3 times per summer when the leaves began to shade the adjacent crop plants. These prunings were applied by hand to adjacent cropped areas as a green manure in all AC treatments. Winter cover crops and crop plants were terminated using a commercial grass roller, or roller crimper, in the AC and OST treatments. This implement was used to flatten and kill cover crops before seed maturation and crop residue after crop harvest. All amendments and cover crop seeds were broadcast and spread by hand to ensure uniformity of application. Compost feedstock was primarily poultry litter and wood chips (2% N, C:N ratio=12.5), and was spread at a rate of 22.4 Mg ha−1 per application. Planting beds in both the AC and OST systems were prepared using strip tillage to create two planting furrows per bed 45 cm apart, with 90 cm between each bed. All vegetables were drip irrigated, and were either direct seeded (Years 1 and 3) or transplanted by hand (Year 2) approximately 2 weeks after cover crop termination. Once weeds began to emerge from the cover crop residue layer, generally 4–8 weeks from planting, plots were hand weeded and a 2.5 cm thick layer of wheat straw (Year 1) or pasture hay (Years 2–3) was applied as mulch to designated treatments. Weeding continued as needed throughout the summer.
The timing of CT treatments was similar, with tillage conducted at the same time as spring roller crimping in the other treatments. Fertilizer was applied by hand as a side dressing in bands adjacent to crop plants according to the University of Georgia Cooperative Extension fertilizer guidelines for specific cropsReference Colditz and Vavrina17–Reference Kelley, Boyhan, Granberry and Kelley19 at the time of planting. Inorganic fertilizer was used on the CT plots only. Herbicides and pesticides were not used on any of the plots. Thus, the input costs for the CT plots are an underestimate, since conventional farmers presumably would use these inputs. However, labor costs may be overestimated, as time spent weeding CT plots would have been lower than if an herbicide were used.
Table 1. General management timeline for the ten treatments in the field study. For a complete management description, see Jacobsen and Jordan12.
Abbreviations for management activity: CC, cover crop planting; CP, compost application; CT, conventional tillage; H, harvest; M, straw mulch application; P, hedgerow pruning; RC, roller crimping; ST, strip tillage; W, weeding.
Costs and returns
Fixed farm costs are presented in Table 2, and were calculated for 4 ha of production. Fixed farm costs included only equipment used in this work, and assumed straight-line depreciation and 10% salvage value. The small-scale no-till rig used in this work was fabricated by the University of Georgia machine shop, and consisted of a series of two discs followed by a 10 cm sweep. This cost was not reflective of a no-till seed drill or other commercially available no-till equipment, but could be representative of a rig appropriate for small-scale diversified operations in the region. Fixed farm irrigation costs included a pump and infrastructure for the drip irrigation system. Land rent cash value was an average paid for irrigated vegetable production in the state of Georgia in 2007Reference Escalante20.
Table 2. Fixed costs for a 40 ha (100 acre) farm. Total fixed farm costs per hectare are calculated for 4 ha (10 acres) in production.
While the site was not certified organic, it was organically managed and production costs reported here were for USDA Certified Organically approved materials. They did not include the cost of organic certification or record keeping. Budgeting periods for each crop began with the sowing of the winter cover crop seed in October of the fall preceding the summer crop and ended with the final harvest of the summer crop. All costs and returns were calculated on a 1 ha scale, and accounted for a 25% loss of production land to hedgerows in AC treatments.
Production input costs
Production input costs that are consistent across all enterprises are presented in Table 3. Organic crop and cover crop seed prices reflect 2007 costs from the average price of three common organic farming supply sources, but did not include shipping and delivery charges. Estimates from the Georgia Vegetable Budgets21 were used for conventional corn and okra seed costs. Conventional winter squash and pepper seed costs were the average price from three common farm supply sources for the region, as no data were found for the region with these figures. Irrigation costs for each crop included an annual purchase of drip tape, and a lifespan of mainline and connectors of 3 years. Connector and mainline costs were distributed evenly over 3 years.
Table 3. Production input costs for a 3-year vegetable crop rotation in alley cropping (AC), organic strip-tillage (OST) and conventionally tilled cropping systems. Input quantities and prices are calculated on a per hectare basis.
Fuel costs for irrigation and tractor operations of each cropping system are presented in Table 4. Costs were based on an average diesel price of US$0.79 per liter and a fuel consumption rate of 3.8 liters h−1 for the 40 hp tractor and 1.9 liters h−1 for the irrigation pump. Lubrication costs were calculated as a standard 15% of fuel costsReference Born and Baier22.
Table 4. Fuel costs for tractor and irrigation operations for each cropping system.
1 Operation times were calculated as annual means from 3 years of operations.
2 Fuel costs assume a gasoline price of US$3.00 gallon−1 plus 15% lubrication cost.
Fertility costs were treatment-specific and are presented in the enterprise budgets (Tables 6–8). Compost costs were calculated from the only organically approved compost provider in the state, and do not include delivery costs, as these would vary by distance from the supplier. Straw mulch costs were the local prices for an approximately 1 short ton round bale, delivery included. Costs for conventional fertilizers were based on estimates from the Georgia Vegetable Budgets21.
Labor
Labor time for each task was recorded in every experimental plot (25–45 m2) for the duration of the 3-year study, and the mean value for each treatment was converted to hours per hectare estimates. Due to the experimental nature of this work, practices such as the application of compost, fertilizer and hedgerow pruning, as well as seeding, weeding and harvest were all conducted by hand to ensure consistency in all plots. Labor for general, non-harvest operations were averaged over the 3-year experiment to reduce year-to-year variation (Table 5). A 15-min set-up time was assigned to each tractor operation (the average from this study) to allot for time spent changing equipment, refueling, etc., and spread evenly across all treatments that task was performed upon. Harvest and weeding labor were treatment-specific, and are presented in the enterprises budgets (Tables 6–8).
Table 5. General operations labor, hours per hectare, expressed as mean values from 2005 to 2007.
Table 6. Enterprise budget for okra, grown in degraded soil in Georgia Piedmont.
1 Labor costs based on $10.00 hourly wage.
2 Typical marketing unit=½ bushel.
Table 7. Enterprise budget for hot pepper, grown in degraded soil in Georgia Piedmont.
1 Labor costs based on $10.00 hourly wage.
Table 8. Enterprise budget for corn and winter squash intercrop grown in degraded soil in Georgia Piedmont.
1 Labor costs based on $10.00 hourly wage.
Labor arrangements on organic farms in the region are highly variable, consisting of family, paid farm workers, interns and volunteer labor under a heterogeneous blend of compensation schemesReference Estes, Kleese and Lauffer9. Rather than applying a standard hourly wage to the total labor (man) hours calculated for all operations, a US$10.00 per hour hired labor wage was included only when labor exceeded 80 h for any 7-day period throughout the season. This rationale was based on the 2004 survey results from the Organic Farming Research Foundation's (OFRF) National Organic Farmers Survey. The OFRF reported 67% of organic farmers worked on the farm full-time, with an average of two full-time, year-round household employees. Our hired labor costs were thus an estimated net of the 80 h weekly family labor limit, derived under the assumption that each full-time household on-farm employee would work 40 h per week. In reality, most farmers would elect to work more than 40 h week−1 person−1 to the greatest extent possible. Thus our labor values were likely to be overestimates of labor costs on such a farm. Labor and production costs did not include marketing costs.
Yields and returns
Organic crop prices for okra (Table 6), hot pepper (Table 7) and winter squash (Table 8) are the average of market prices from three organic farms in the Athens, Georgia area and a certified organic market in Atlanta, Georgia, the nearest major metropolitan market. The organic corn price was an average value for No. 2 yellow corn from nine nationwide markets reported on the New Farm Organic Price Index23 for the 2007 crop. Conventional prices for okra and corn were prices reported in the University of Georgia Vegetable Budgets21. No prices for conventional winter squash or hot pepper were found for the state. Instead, winter squash and cayenne pepper prices from the Louisiana State University Research and Extension Ag Center24 were used.
Identifying dominant, preferred production methods
Our study employed a stochastic dominance analysis model to evaluate the risk and return structures of the ten treatments in this work over the course of the 3-year study. Stochastic dominance analysis is a risk-efficiency criterion for determining the risk-efficient set of alternatives available to producers when faced with uncertain outcomes. More than just examining the return structures of (or level of payoffs from) different production alternatives, the stochastic dominance framework evaluates trade-offs between such levels and variability (riskiness) of the payoff variable.
Stochastic dominance is a useful, appropriate tool for analyzing decisions in agriculture, as outcomes of production decisions and methods can be influenced by a wide range of risk factors (such as weather, climate and pests). In making production decisions, a farmer thus must not rely solely on absolute figures, such as average yields or total net farm incomes, as these could be misleading and thus produce less optimal decisions. Specifically, without accounting for risk or variability, a farmer may quickly consider a specific production alternative as already adequately profitable merely based on a high average yield result. The stochastic dominance approach looks beyond total and average values, and also considers the variability of the year-to-year (or seasonal) production conditions that produce such figures. This approach helps farmers determine more favorable production alternatives that could ensure more stable or less risky outcomes.
The risk component of this analysis recognized that uncertainty, which is especially a significant factor in farming situations, can affect realization of payoff. Thus, a decision-maker evaluates trade-offs between higher payoffs and lower risk, with the choice determined by his/her level of risk aversion. In this study, the application of the model and the stochastic dominance framework allows for a ranking of alternatives based on producers' risk preferences for the most risk-efficient set of yields and returns to management.
Researchers have developed multiple variations of stochastic dominance, but its two basic criteria are first- and second-degree stochastic dominance. We employed second-degree stochastic dominance analysis, which eliminates dominated or inefficient distributions from the first-degree stochastic dominance setReference Huang and Litzenberger25. This was accomplished by adding the assumption of risk aversion to the decision-making process with respect to the farmer or land managers' preferences.
Risk-averse agents seeking to maximize utility would never prefer a dominated distributionReference Anderson, Dillon and Hardaker26. Therefore, a second-degree stochastically efficient set of alternatives would be comprised of only non-dominated distributions, and any further reduction of this set would require additional assumptions concerning risk preferencesReference Anderson, Dillon and Hardaker26.
This analysis ranked the risk and return efficiencies of the ten treatments based on the following parameters:
1. Production yields: this measure isolated the production efficiency from other factors that can affect the risk and return profiles of the production methods. In order to standardize yields across all years and reduce variation, this analysis used the proportion (percent) of the experimental yields to conventional yield estimates for Georgia27, instead of the absolute yield. Yields for the first year were based on a conventional okra yield of 10,946 kg ha−1 (325 bushels acre−1); Year 2 yields were based on average conventional cayenne pepper yield of 10,105 kg ha−1 (9000 lbs acre−1); and Year 3 yields were based on a 50% corn/50% winter squash planting area distribution assumption, with an average corn yield of 3.9 Mg ha−1 (62.5 bushels acre−1) and average squash yield of 3521 kg ha−1 (112 bushels acre−1).
2. Net returns per man hour of family labor: the combined effects of production and cost efficiencies are captured by this measure of net return. Profitability is also related to the variable rates of labor intensiveness of the different production methods by calculating the net return per man hour of residual family labor invested in farm work that supplemented the assumed hired labor requirements. This measure thus represented the farm owner's family's compensation for their collective labor hour-investment in the farm operations.
Results and discussion
As discussed previously, budgetary information for organic, conservation tillage systems with which to compare our results was sparse. Some yield and production cost information existed for either organic production or conservation tillage, but rarely both. The few exceptions found used living mulches interplanted with organically managed vegetablesReference Infante and Morse28–Reference Carrera, Morse, Hima, Abdul-Baki, Haynes and Teasdale30. Although yields were not statistically different between treatments due to high variation within treatmentsReference Jacobsen and Jordan12, the resulting differences in returns may be of economic significance to the farmer.
Experimental yields were consistently extremely low compared to conventional production estimates for the state. All yield data in this study must be interpreted in the light of the degraded soils in which the experiment took place. The soils in this study were a highly eroded Pacolet sandy clay loam that was devoid of an A horizon and largely consisted of the B (subsoil) horizon. The site was chosen, because it was the worst soil on the experimental farm, and is characteristic, if not worse, than the sandy clay loams commonly found in the region. Additionally, each crop was only grown for one year in the rotation, and thus yields may not be representative of long-term yield potential. Okra yields were in the ‘worst production category’Reference Flanders and Flanders31, and corn was 14% of the average yield for strip tilled corn in South Georgia32, after accounting for the production area occupied by winter squash. In addition to the degraded soil on the site, a severe drought in 2007 limited the germination of the heirloom corn variety used in this work that was not treated with fungicide or drought tolerant. Although the yields were lower than conventional averages for the state, okra and hot pepper yields were comparable to experimental yields in an organically managed, conservation tillage experimental study in the MidwestReference Biazzo and Masiunas29.
Organic okra and hot pepper production had the highest net returns to management, although the harvest labor requirements for these crops were 10–15-fold higher than the corn/winter squash intercrop. Both okra and hot pepper are high-yielding, labor-intensive crops, the former with a significant organic price premium. Okra production did not require additional hired labor for any of the treatments, but the hot pepper did require additional harvest labor. However, the greater returns to management justified the expense of hired harvest labor.
Previous research in the region suggested that AC may be best suited to high-value horticultural crops due to higher labor and land requirements needed to manage the hedgerowsReference Jordan11. However, the labor associated with hedgerow management averaged only 30 h pruning event−1 ha−1. This time included the precise application of residues for experimental purposes, and was probably an overestimate of actual labor requirements.
Mulches are frequently used in organic farming systems to suppress weeds, control erosion and retain soil moisture. In 2005 and 2006, mulched AC treatments required 23% less labor than non-mulched treatments, due to effective weed suppression. In 2007, mulch only reduced labor requirements by 4%, due to a lack of weed pressure in the extreme drought. All the OST treatments received mulch; thus, mulched versus non-mulched labor requirements could not be compared in this system. Although mulch applications generally reduced weeding requirements, they also resulted in significantly lower soil carbon levels than non-mulched treatments receiving the same level of compost (Table 9). In the hot, subtropical climate of the southeastern US, evapotranspiration rates in the summer are high and soils dry down rapidly. Increased soil moisture under mulch could have led to increased decomposition rates of organic amendments and soil organic matter in mulched treatments; thus, decreased labor requirements also decreased potential environmental benefits to the system (for a complete discussion of the soil carbon dynamics in this work, see Jacobsen and Jordan12).
Table 9. Changes in total soil carbon levels over the course of the accompanying field study. For complete discussion and presentation of results, see Jacobsen and Jordan12.
1 Indicates significant change between Year 1 and end of Year 3 at the P⩽0.10 level.
2 Indicates significant change between Year 1 and end of Year 3 and the P⩽0.05 level.
Additionally, mulched treatments had inconsistent effects on yields in the 3 years. Okra, corn and winter squash yields in the mulched treatments were higher than non-mulched treatments receiving the same level of compost. These crops benefited from the increased moisture under the mulch, as the okra is fungal-disease resistant and the corn and winter squash were grown in a drought year. However, the increased moisture under the mulch led to an outbreak of Fusarium wilt in the pepper plants in mulched treatments, decreasing yields. This indicates that mulch in addition to the roller crimper killed cover crop residue may be best suited for dry years and for plants that are not susceptible to soil-borne pathogens.
The effects of compost additions were less straightforward and were highly variable in this relatively short study period. The greater compost application rate (44.8 Mg ha−1) did not have consistently greater yields than treatments with 22.4 Mg ha−1 or no compost. This indicates that the ecological interactions in restorative agroecosystems on degraded soils are more complicated than what would be predicted by simply adding more amendments. Generally, the organic treatments had higher yields than the conventional treatment. When accounting for organic price premiums, the organic treatments had consistently higher net returns to management than the conventional treatment.
Treatments AC2, OST2, OST1 and AC3 had the highest net returns per family labor man hour. In this analysis, the first three of these four treatments had the lowest labor requirements due to the presence of mulch, which suppressed weeds and reduced weeding labor. Additionally, these treatments ranked in the top four in net returns, exclusive of labor costs (Table 10). Altogether, the reduction in labor and high net returns produced better risk-efficiency rankings for these alternatives in terms of net returns to family labor (Table 11). The existing literature presents contrasting results regarding the profitability of organic farming systems when considering both the increased labor requirements and organic price premiums. Some have claimed that reduced input costs, high price premiums and endurance under drier conditions have enhanced organic farms' relative profitabilityReference Jans and Fernandez-Cornejo33, Reference Barkley34. Other studies, however, have contested the advantage due to higher labor costs. As reviewed by FriedmanReference Friedman35, the production costs of organic apples in California were 10–25% higher than conventional farms as a result of higher material and labor costs. In contrast, in a potato study in Idaho involving 18 conventional and organic farming systems, the average material costs were lower among organic farms while labor costs were higher. In this study, the stochastic dominance analysis of the net returns to family labor parameter ranked most of the organic systems consistently higher than the conventional system on this highly degraded soil.
Table 10. Derivation of net return per hour of family labor. Derivations were based on raw data from individual plots across the three crops (okra, hot pepper and corn/winter squash intercrop) grown in the study. Variation reflects differences in labor requirements and net returns between the three enterprises.
Table 11. Stochastic dominance results based on production yields, revenues and changes in soil carbon levels. Rankings are expressed from low to high, where 1 is the most dominant (preferable) and 10 is the least.
1 To standardize across yields for different crops, yields are expressed as a percentage of average conventional yields. Proportional yields are based on the following average conventional yields per hectare: 10,946 kg ha−1 (325 bushels acre−1) of okra (Year 1), 10,105 kg ha−1 (9000 lbs acre−1) of cayenne pepper (Year 2) and a 50/50 acreage allocation with 62.7 kg ha−1 (62.5 bushels acre−1) of corn and 3521 kg ha−1 (112 bushels acre−1) of squash (Year 3).
Although the year-to-year variation was important for understanding the utility of each system and certain production practices to specific enterprises, the stochastic dominance analysis allowed us to assess variations in system performance over time. These results can be compared with ecological parameters across all enterprises. Treatments AC3 (AC+22.4 Mg ha−1 yr−1 compost), AC2 (AC+mulch), OST2 (OST+22.4 Mg ha−1 yr−1 compost+mulch) and OST3 (OST+44.8 Mg ha−1 yr−1+mulch) ranked highest in proportional yields across all years, but averaged only 42–45% of the statewide average27 (Table 11). These four treatments also produced the most risk-efficient yield structures, given their stochastic dominance rankings.
AC3 and AC5 (AC+44.8 Mg ha−1 mulch yr−1) were the only treatments with net increases in soil carbon over the course of the 3-year study, although the latter was the only treatment with a significant increase. It is important to note that appreciable changes in soil carbon levels requires long-term study, and that soil carbon results presented in this work can be considered indicators of trends, but not definitive, long-term data. However, these results suggested that in highly degraded soils in the subtropics compost additions may be necessary to build soil carbon levels even in systems incorporating winter cover crops and perennial legumes. In addition, while the absence of mulch increased weeding labor, only treatments without mulch increased in soil carbon in the AC system. These results demonstrate that while a few treatments may be preferable for a single parameter, no treatment emerged as dominant for all parameters. However, the AC treatment receiving 22.4 Mg ha−1 yr−1 (AC3) ranked in the top four for all parameters in the stochastic dominance analysis, had significant increases in soil carbon levels and required very little additional labor for hedgerow management, as previously discussed.
Conclusions
Organic farming and other ecological approaches to agriculture employ long-term systems approaches to nutrient and pest management. These systems frequently incorporate the use of winter cover crops, fallow periods and organic amendments to increase soil organic matter and thus increase the long-term productive capacity of the soilReference Altieri36. The goal of this work was to gain a general understanding of the economic costs and benefits of experimental agroecosystems designed to restore highly degraded soils in the Georgia Piedmont using a suite of these techniques in combination.
The organic conservation tillage treatments had less tractor and labor costs than the tillage-intensive conventional treatment. The application of mulches effectively suppressed weeds and reduced weeding labor requirements by an average of 23% during non-drought years. However, the presence of mulch increased the decomposition of compost and soil organic matter in this 3-year study, highlighting the need for longer term research on potential soil carbon sequestration benefits of organic conservation tillage-based systems. Yields in all experimental treatments were lower than in conventional studies found for the region. This was expected due to the nature of the soil at the study site, which had been abandoned for conventional row crop production due to lack of productivity for a number of years. However, returns on high-labor, organic crops were over US$30,000 ha−1 in some treatments. Our stochastic dominance results demonstrate that no one treatment maximized yields, net returns per family hour of labor or soil carbon increases. However, when these parameters were viewed together, AC systems receiving 22.4 Mg ha−1 yr−1 of compost were an optimal risk-efficient choice for all parameters and demonstrated significant gains in soil carbon, a key management challenge in the region. These results indicate that some organic, conservation tillage systems could restore soil productivity and command high returns per land area across multiple enterprises, allowing land to remain in cultivation while improving soil quality.
Acknowledgements
This work was funded by a United States Department of Agriculture Sustainable Agriculture Research and Education Program grant. The figures for costs of organic materials were greatly improved by the contributions from local organic farmers in the Athens, Georgia area and the farmers of the Morningside Farmer's Market in Atlanta, Georgia.
