Abstract
Escalating production costs, heavy reliance on non-renewable resources, reduced biodiversity, water contamination, chemical residues in food, soil degradation and health risks to farm workers handling pesticides all bring into question the sustainability of conventional farming systems1, 2, 3, 4. It has been claimed5, 6, however, that organic farming systems are less efficient, pose greater health risks and produce half the yields of conventional farming systems. Nevertheless, organic farming became one of the fastest growing segments of US and European agriculture during the 1990s7, 8. Integrated farming, using a combination of organic and conventional techniques, has been successfully adopted on a wide scale in Europe9. Here we report the sustainability of organic, conventional and integrated apple production systems in Washington State from 1994 to 1999. All three systems gave similar apple yields. The organic and integrated systems had higher soil quality and potentially lower negative environmental impact than the conventional system. When compared with the conventional and integrated systems, the organic system produced sweeter and less tart apples, higher profitability and greater energy efficiency. Our data indicate that the organic system ranked first in environmental and economic sustainability, the integrated system second and the conventional system last.
Organic management practices combine traditional conservation-minded farming methods with modern farming technologies but exclude such conventional inputs as synthetic pesticides and fertilizers, instead putting the emphasis on building up the soil with compost additions and animal and green manures, controlling pests naturally, rotating crops and diversifying crops and livestock10. Organic farming systems in the US range from strict closed-cycle systems that go beyond organic certification guidelines by limiting, as much as possible, external inputs to more standard systems that simply follow organic certification guidelines. Integrated farming systems reduce the use of chemicals by integrating organic and conventional production methods.
Just because a system is organic or integrated does not ensure its sustainability; nor does sustainability, an inherently complex concept11, readily lend itself to quantification. To be sustainable, a farm must produce adequate yields of high quality, be profitable, protect the environment, conserve resources and be socially responsible in the long term12. But under conventional economic systems, market and social forces can change the viability of a production system independent of its environmental sustainability13. It has been proposed that ecological and economic systems should be linked so that ecosystem services are accounted for in commercial markets, thereby making sustainable land management a prerequisite for economic sustainability14, 15.
A crucial step in developing such ecological–economic links is to assess the effects of agricultural systems on specific, measurable properties that are important indicators of sustainability. We measured the effects of an organic, a conventional and an integrated apple production system on the sustainability indicators of soil quality, horticultural performance, orchard profitability, environmental quality and energy efficiency. Perennial food crops such as apples may prove to be more sustainable to produce over the long term than annual crops16, and they currently comprise a significant portion of the world's agricultural production. For example, globally, nearly 5.6 million hectares of apples were harvested in 2000 (ref. 17). In the USA alone, apples and other high-value perennial food crops constituted 16% of the total value of food crops in 1998 (ref. 18).
We measured soil quality by analysing physical, chemical and biological soil properties and incorporating the data into a soil quality index19. Soil quality is the capacity of a soil to sustain biological productivity, maintain environmental quality and promote plant and animal health20. We evaluated soil quality in terms of four soil functions: accommodating water entry; accommodating water movement and availability; resisting surface structure degradation; and supporting fruit quality and productivity. Soil quality ratings in 1998 and 1999 for the organic and integrated systems were significantly higher than those for the conventional system (Table 1), largely owing to the addition of compost and mulch in 1994 and 1995. Organic matter has a profound impact on soil quality, enhancing soil structure and fertility and increasing water infiltration and storage21. Because of poorer ability to accommodate water entry and to resist surface structure degradation, the conventional system (no organic amendments added) scored lowest overall in soil quality.
We assessed horticultural performance by measuring fruit yields, size and grade; tree growth; leaf and fruit mineral contents; fruit maturity; and consumer taste preference. There were no observable differences in pests, disease or physiological disorders among plots during each growing season. Differences in annual fruit yields were inconsistent among the three systems (Fig. 1). Cumulative yields were similar for all three systems. In 1995–1997, fruit size was similar across systems, except in 1996 when apples were larger in the integrated system (see Fig. A1 in Supplementary Information). In 1998 and 1999, the organic system produced smaller fruit (see Figs A1 and A2 in Supplementary Information). In 1995–1997, all marketed fruit produced from the three systems was sold for processing because it was downgraded primarily owing to skin russetting, a physiological skin disorder that reduces the fruit's visual appeal but not its taste or other attributes. (Although russetted Golden Delicious apples are not sold as fresh fruit in the US marketplace, Italy domestically markets a fully russetted Golden Delicious apple, and in the world market fully russetted Bosc pears are preferred to non-russetted ones.) The low landscape position of the experimental site in the orchard resulted in early season cool, humid conditions that contributed to the unusually high level of russetting. Fruit damage due to other physiological disorders, pests and diseases were minimal and equal for each of the three systems. In 1998 and 1999, marketable fruit not graded as Washington Extra Fancy or Fancy was sold for processing (see Table A1 in Supplementary Information).
Figure 1: Fruit yields of three apple production systems.
Differences between values in a year followed by different letters are significant at the 0.05 level (least significant difference).
High resolution image and legend (20K)Tree growth was similar in all three systems (see Fig. A3 in Supplementary Information). Although there were some differences in leaf nutrient contents among the three systems, analyses indicated satisfactory levels of nutrients22, 23 (see Table A2 in Supplementary Information). Fruit tissue nutrient analyses indicated some inconsistent differences (see Table A3 in Supplementary Information).
Mechanical analysis of fruit firmness at harvest and after storage in 1998 and 1999 showed that organic fruit was firmer (a positive consumer attribute for apples) than or as firm as conventional and integrated fruit (see Table A4 in Supplementary Information). Ratios of soluble solids (sugar) content to acidity (tartness), an indication of sweetness, were most often highest in organic fruit. These data were confirmed in taste tests by untrained sensory panels that found the organic apples to be sweeter after six months of storage than conventional apples and less tart at harvest and after six months storage than conventional and integrated apples (see Table A5 in Supplementary Information). The same taste tests, however, could not discern any difference in firmness among apples in the three systems at harvest or after storage. Taste tests also indicated that the integrated apples had a better flavour after six months storage but found no differences among organic, conventional and integrated apples in texture or overall acceptance.
Enterprise budgets were generated each year to calculate net returns from total costs and gross receipts. Receipts for the integrated system were estimated using prices for conventionally produced fruit, as unlike organic fruit there was no price premium for integrated fruit. Receipts for the organic system were estimated using prices for conventionally produced fruit in the first three years (1994–1996), the number of years necessary to convert from conventional to certified organic. The price premium to the grower for each grade of organic fruit in the next three years (1997–1999) averaged 50% above conventional prices.
The three systems did not show a net annual profit until 1999 under measured fruit quality conditions (with skin russetting) (Table 2). When we adjusted the economic analysis by eliminating the effects of russetting but maintained the estimated crop loss of 15% due to other factors and the measured size, grade and firmness of fresh fruit in this study, the organic system was more profitable than the conventional and integrated systems in 1997 and 1998. Higher production costs for the organic system in 1995 and 1997 (under measured and non-russetted fruit quality conditions) were largely due to differences in weed control practices, fruit thinning and compost applications. Production costs in 1999, however, were significantly lower for the organic system than for the other two systems due to reduced carryover interest costs resulting from faster repayment of the original investment.
The breakeven point, when cumulative net returns equal cumulative costs, can vary depending on several factors, such as fruit prices, input costs, yields and fruit quality. The breakeven point in this study is projected to occur nine years after planting (in 2002) for the organic system under measured fruit quality conditions. The conventional and integrated systems would break even 15 and 17 years after planting, respectively, under measured conditions. Under non-russetted fruit quality conditions, the breakeven point would occur six, eight and nine years after planting for the organic, conventional and integrated systems, respectively. Assuming similar non-russetted fruit quality conditions, estimated breakeven points for conventional apple orchards in central Washington range from 8 to 11 years from planting24.
Without price premiums for organic fruit, the conventional system would break even first, the integrated second and the organic third under measured or non-russetted fruit quality conditions. For breakeven points of the organic and integrated systems to occur in the same year as the conventional system, price premiums of 12% for the organic system and 2% for the integrated system would be necessary under measured fruit quality conditions. Under non-russetted fruit quality conditions, premiums of 14% for the organic system and 6% for the integrated system would be necessary to match the breakeven point of the conventional system.
We assessed the environmental impact of the three production systems by using a rating index25 employed by scientists and growers to determine the potential adverse impact of pesticides and fruit thinners, including naturally occurring certified organic products. The higher the rating, the greater the negative impact. As only 35% of conventional Washington apple growers use pheromone-mating disruption (PMD), an environmentally benign biological control used in our conventional treatment, we also included a conventional system in which synthetic pesticides were used in place of PMD. The total environmental impact rating of our conventional system was 6.2 times that of the organic system, whereas the integrated system rating was 4.7 times greater and the non-PMD conventional system rating was 7.7 times greater (Fig. 2).
Figure 2: Environmental impact ratings of four apple production systems: Organic, conventional, integrated and non-PMD conventional.
Higher ratings indicate greater potential for negative environmental impact. For a listing of chemicals used and their impact ratings for the four production systems, see Table A6 in Supplementary Information.
High resolution image and legend (26K)Energy accounting was divided into inputs (labour, fuel, fertilizers and so on), output (yield) and output/input ratios (energy efficiency). Cumulative energy inputs and output for the six-year study period were lower for the organic system than for the conventional and integrated systems (Table 3). The output/input ratio for the organic system during the six-year study period, however, was 7% greater than that for the conventional system and 5% greater than that for the integrated system, making the organic system the most energy efficient.
Our results show that organic and integrated apple production systems in Washington State are not only better for soil and the environment than their conventional counterpart but have comparable yields and, for the organic system, higher profits and greater energy efficiency. Although crop yield and quality are important products of a farming system, the benefits of better soil and environmental quality provided by the organic and integrated production systems are equally valuable and usually overlooked in the marketplace. Such external benefits come at a financial cost to growers. Currently, growers of more sustainable systems may be unable to maintain profitable enterprises without economic incentives, such as price premiums or subsidies for organic and integrated products, that value these external benefits. Equally important, upon incorporation of external costs into economic assessments of farming systems, we may find that many currently profitable farming systems are uneconomical and therefore unsustainable. The challenge facing policymakers is to incorporate the value of ecosystem processes into the traditional marketplace, thereby supporting food producers in their attempts to employ both economically and environmentally sustainable practices.
Methods
Study area
In May 1994, we planted four replicate plots for each of the three apple production systems with 'Golden Delicious' apples (Malus 
domestica Borkh.) on EMLA.9 rootstocks on 1.7 ha in a randomized complete block design. The experiment was part of a 20-ha commercial apple orchard in the Yakima Valley, Washington. See Fig. A4 of Supplementary Information for plot layout and orchard system.
Farming systems
In cooperation with the farmers, professional consultants and extension agents, we chose appropriate management practices for the three systems (see Table A7 in Supplementary Information). The organic system included compost and foliar sprays. In the first three years (1994–1996), bark mulch and landscape fabric controlled weeds; thereafter, cultivation and mowing were used for weed control. Organically certified biological controls, including applications of Bacillus thuringiensis and PMD to control codling moth (Cydia pomonella L.), were used for pest management. Fruit thinning was by hand. The conventional system included synthetic soil fertilizers and foliar sprays, pesticides, chemical fruit thinners and PMD. The integrated system used both compost and synthetic fertilizers and controlled weeds with both bark mulch and herbicides. Pest management and fruit thinning were similar to those of the conventional system. The three systems had similar total soil nitrogen inputs. Pests, diseases and physiological disorders were monitored throughout each growing season by the farmers and professional consultants, who recommended organic, conventional or integrated treatments for their control.
Soil analyses
All soil samples were taken from the inner two rows of each experimental plot to minimize edge effects, excluding the first 20 trees from each end of these sample rows. Samples were collected midway between trees within tree rows. In 1998 and 1999, soil analyses included bulk density, water content, total nitrogen, nitrate-nitrogen, extractable phosphorus, cation-exchange capacity, pH, electrical conductivity, organic carbon content, aggregate stability, microbial biomass carbon and nitrogen, and earthworm populations. Details of analytical procedures are described elsewhere17.
Horticultural performance
All components of horticultural performance were measured from trees, leaves and fruit sampled in the middle two rows of each plot, excluding the first 20 trees from each end of these sample rows. We recorded yields and size (average mass) of fruit at harvest in 1995–1999. The proportions of fruit suitable only for processing (due to small size or defects) and fruit suitable for fresh market and divided by grade were also recorded. Grading of fresh fruit was based on Washington State's apple industry standards26. We used trunk cross-sectional area to estimate unit tree growth. We analysed leaf mineral contents (N, P, K, S, Ca, Mg, B, Zn, Mn, Cu and Fe) in 1994–1999 from pooled samples of mid-shoot leaves taken randomly from each plot in midsummer. We analysed fruit flesh mineral contents (N, P, K, Ca, Mg, B and Zn) in 1995 and 1997–1999 from pooled samples of uniformly sized fruits taken three weeks before harvest. Mineral analyses were carried out according to standard methods23. We assessed fruit maturity parameters, including flesh firmness, soluble solids and acidity, according to standard procedures27 at harvest and after three and six months of controlled atmosphere storage in 1998 and 1999. Untrained sensory panels were used to determine preferences for overall acceptance, texture, flavour, firmness, sweetness and tartness of 1999 fruit from each production system and storage treatment.
Economic analyses
We calculated gross receipts using farmgate prices paid by packing houses to farmers for apples sold at harvest or after storage. Prices for the specific size, grade and firmness of 'Golden Delicious' organic and conventional apples from our study were based on prices from Washington Growers Clearing House Bulletins and fruit packing houses in Washington State. Total costs included non-harvest variable costs (fertilizers, pesticides, fuel, labour and water), harvest variable costs (picking, grading, packing and storage) and fixed costs (machinery, interest and taxes). Projected returns for 2000 and beyond were estimated from average fruit sizes (1998–1999) and yields (1997–1999) for each treatment, assuming a 15% cullage rate for all three treatments and a 50% price premium for organic fruit.
Environmental impact assessment
We determined environmental impact ratings for each farming system using an index developed by Stemilt Growers, Inc. of Wenatchee, Washington, as part of their 'Responsible Choice' program25. Similar to Cornell University's Environmental Impact Quotient28 but updated to include fruit thinners and certified organic products, the index takes into account chemical efficacy, potential worker and consumer exposure, leaching potential, soil sorption index, chemical half-life and the effects of chemicals on beneficial organisms, all based on toxicological studies and chemical characteristics of each product. The active ingredient of each pesticide and the dose and frequency of application were used to calculate the environmental impact ratings.
