Final Report for SW03-040
We evaluated the effects of cover crop, amendment, and tillage strategies on soil quality in organic vegetable and tree fruit production. In the vegetable experiment, amendment had the greatest effect across a range of soil properties. Tillage frequency affected nematode and collembola ecology. Spader tillage reduced subsurface compaction. Amendment, cover crop, and tillage all affected crop yield in at least one year. In the orchard system, wood chips led to better tree performance and fruit production than living mulch or cultivation, but not better soil quality. The economic benefit covered the cost of wood chip application. Tillage degraded soil quality.
- Compare the effects of selected organic row crop and orchard management systems on soil quality as measured by field and laboratory tests and field observations.
Evaluate the NRCS Soil Conditioning Index as a tool for soil quality assessments.
Develop underseeded cover crop guidelines for humid Northwest environments.
Share findings with organic vegetable and fruit producers and other producers interested in sustainable practices in the Northwest.
Washington State ranks number one in the nation in certified organic tree fruit production and number two in certified organic vegetable production (Granatstein and Kirby, 2007). Vegetable and tree fruit production account for nearly half of the certified organic acreage in Washington State. Gross farm sales in the organic sector increased by 31% in 2005 over the previous year. Direct sales of all produce at farmers markets increased by 20% over the same period. Strong demand by consumers has led to rapid growth of the organic sector.
Organic vegetable and fruit production require intensive management, presenting challenges to maintaining and building soil quality and productivity while building profits in the short and long term. Growers need reliable information on how different organic production systems affect short and long term soil quality, land productivity, and profitability. This project compares different organic management systems for tree fruit and vegetable crop production, with the overall goal of identifying management strategies that build soil quality within intensive production systems.
Objective 1. Compare the effects of selected organic row crop and orchard management systems on soil quality as measured by field and laboratory tests and field observations.
This project includes soil quality measurements in an organic vegetable crop systems experiment at WSU Puyallup (established 2003), two orchard weed and soil management experiments established in Wenatchee in 2004 and 2005, and six on-farm sites in western Washington. Five of the cooperating farmers grow organic vegetables, while the sixth farmer grows organic vegetables and small fruit (raspberry, strawberry, and blueberry). Two farms also had small poultry flocks for egg and/or meat production.
Vegetable crop systems experiment.
The vegetable crops systems experiment at WSU Puyallup compares 12 combinations of organic management systems, consisting of three cover crop treatments, two tillage treatments, and two soil amendment treatments arranged in a split-split plot design. Cover crops are the main plots, tillage is the first split and amendments the second split (Fig 1). The experimental area consists of four replicate blocks, each measuring 250 x 50 ft. Main plots (cover crops) within the blocks are 80 x 50 ft. Mowed tall fescue strips buffer the treatment area. The entire field, including buffers, is organically certified and measures 570 x 210 ft. The soil is a Puyallup fine sandy loam, developed from recent alluvium in the Puyallup valley of western Washington. The surrounding area is a mosaic of annual and perennial crops, grass pastures, and wetlands.
We chose treatments based on input from farmers through focus groups, surveys, meetings, and farm visits. They represent a range of strategies of interest to organic farmers, and provide contrasts in organic carbon additions, tillage intensity, and duration of the cover crop cycle (Fig. 2). Cash crops chosen for the rotation are commonly grown on local organic fresh market farms and cover a range in growth habits and management requirements.
Cover crop treatments include 1) a fall-seeded cereal rye-hairy vetch mix, 2) relay-intercropped hairy vetch or red clover planted into the cash crop, and 3) a short-term annual ryegrass-perennial ryegrass-red clover pasture (ley) crop. The value of these cover crops includes N fixation, ground cover, weed suppression, and biomass.
Relay planting offers benefits of early establishment of legumes, an N-rich green manure, and reduced tillage. Disadvantages include a smaller window for weed cultivation, and risk of reduced stands from crop and residue competition and winter damage.
The ley rotation provides an extended period of weed suppression and reduced tillage, and is attractive to farmers with livestock or who are not limited by land. The land is in pasture more than 75% of the cropping cycle and cash crops are planted only in alternating years. Poultry (broilers) are raised on the ley plots for 6-8 weeks during the pasture phase. We have transitioned the ley rotation into a low-input treatment, with no amendments applied since 2005.
Organic amendments include a low-carbon treatment (broiler litter, mean total N = 3.6%; C: N = 11:1) and a high-carbon treatment (mixed on-farm compost, mean total N = 1.8%; C:N = 15:1). The broiler litter is self-heated and turned 5 times to meet organic pathogen reduction standards. The on-farm compost is actively composted for one month in an aerated static pile, followed by a 5-month curing period. Application rates ranged from 1.8 to 2.7 dry tons/acre (128 to 216 lb total N/acre) for the broiler litter and 10 to 17 dry tons/acre (440 to 580 lb total N/acre) for the on-farm compost. Rates are based on a 2-year field and laboratory assessment of available N from organic amendments (Gale et al., 2006). We did not met time-temperature requirements for pathogen reduction all years, and when that occurred, we observed 90 or 120 day waiting periods between amendment application and crop harvest, depending on the crop.
The tillage regimes are conventional tillage (plow, disc, rototill) and modified tillage (low-speed rotating spader). The spader is designed to simulate hand digging with a shovel. It can incorporate amendments and produce a seedbed in a single pass in light-textured soils.
The crop rotation for the first three years was snap bean (2003), spinach following summer cover (2004), and winter squash (2005). In 2006 and 2007, 4 rotation crops were grown in each plot each year, by dividing the plots into narrow beds, typical of those used by local organic mixed vegetable growers. The rotation across beds is: 1) double crop of spring broccoli followed by fall spinach, 2) snap bean, and 3) winter squash.
The plots in the ley treatment were in cash crops each odd-numbered growing season (2003, 05, 07) and in pasture at all other times. Table 1 shows the timing of key field activities in each of the cover crop treatments from 2003-2007.
Soil quality measurements. Aggregate stability, infiltration, and soil compaction measurements were made before planting the cash crop (2-3 weeks after incorporation of cover crops and amendments) and again in mid-season (beginning of canopy closure) each year. Samples for bulk density were collected mid season. We measured aggregate stability on composite samples from the 0 to 10 cm depth of each plot using wet sieving (4.25, 2, 1, and 0.25 mm sieves) and evaluated data as mean aggregate diameter and aggregate size histograms (Nimmo and Perkins, 2002). Soil compaction measurements were made with a Rimik recording penetrometer (9 readings per plot, 0 to 30 cm depth), and bulk density done with a hammer driven core sampler (0 to 8 cm depth, 3 per plot). Infiltration rate was measured using a simple single-ring falling-head method (Soil Quality Institute, 1999) at 5 locations/plot. Enzyme activities (dehydrogenase and b-glucosidase) (Tabatabai, 1994) and POM C (53 um + fraction; Cambardella and Elliot, 1992).were measured in samples collected before planting the cash crop and at mid-season in samples collected from the 0 to 10 cm depth.
Collembola were isolated from three replicate soil samples per plot (5 cm depth, 0.45 L per sample). Soil samples were weighed, placed in Berlese-Tullgren funnels, and collembolans extracted by heating and drying the soil, based on the method of Moldenke (1994). Collembola were identified at the family level.
Nematodes were extracted from fresh soil collected from the 0 to 15 cm depth (15 cores/plot) using a sieving-Baermann funnel procedure. Prior to routine counting and identification of nematodes in samples from experimental plots, the dominant genera at the site were identified by observing >500 individual specimens representing all treatments at 400X to 1000X with a compound microscope equipped with Nomarski DIC. Thereafter, routine analyses of experimental samples were performed by observing samples in a counting dish with an inverted microscope. The number of nematodes was determined in each sample (at 40X) and then 100 randomly chosen individuals were identified to the level of genus when possible (at 400X). The data on relative and absolute abundance of the taxa were used in construction of ecological indices as described previously (Forge et al., 2005; 2003; Ferris et al., 2001; Bongers and Ferris, 1999). Samples for collembola and nematodes were collected once in 2005 and three times (pre-plant, mid-season, and post-harvest) in 2006 and 2007 from a subset of 24 plots plus 4 areas in the grassed alleyways between plots. The pre-plant 2007 samples were collected in early spring before disturbance by amendment application and tillage.
On-farm cooperators. In 2005 we measured bulk density, aggregate stability, infiltration rate, penetrometer resistance, and texture at multiple points on each of the six cooperating farms. At two of the farms we made measurements at two separate locations under their management. We also collected composite samples for soil organic matter, pH, Bray P and exchangeable K. Each sampling location was logged into a GPS system so that we can return to the same points in subsequent years. In 2006 we completed measurements at four of the six farms.
Orchard mulches and weed management experiments.
The objective of these experiments is to evaluate soil quality differences related to in-row weed control methods in orchards including mechanical tillage (different rates and types), wood chip mulch, living mulch or control treatments applied to the weed strip. Organic orchardists predominantly rely on repeated tillage to control weeds. Repeated tillage without countervailing practices (e.g. compost application) can easily lead to depletion of soil carbon and loss of soil structure. However, lack of weed control will lead to poor tree performance due to competition for nutrients (especially nitrogen) and water. Our overarching hypothesis for these two trials is that the wood chip mulch improves soil physical properties, and tillage degrades them relative to a bare or mowed control. Additionally, we hypothesized that a living mulch planting would also increase soil quality over tillage or control treatments. Ultimately, we seek to understand how to balance weed control, fertility management, and soil quality for optimal tree performance and fruit yield and quality in organic orchards. The National Organic Standards explicitly require growers to maintain or improve soil quality. Therefore, we want to better understand whether practices to ensure optimal fruit production (e.g. tillage for weed control) are in direct conflict with the soil quality standard, and what alternatives might be available to resolve this conflict.
We conducted two experiments during the course of this project, referred to as Tillage Comparison Trial and Integrated Mulch Trial.
Tillage Comparison Trial. In April 2004, a 2-year trial was set out in an 8-yr old block of Gala/M26 in transition to organic certification at the Wenatchee Valley College teaching orchard near East Wenatchee, WA. Treatments included control (no tillage, mowing to keep weeds down), wood chip mulch (applied 6” thick,), Cultivator Y (3x), Cultivator Z (2x), Cultivator Z (3x), and Cultivator Z (4x), with 5 replicates, where “x” denotes the number of times plots were tilled. In Year 2 Cultivator Y treatment was tilled once only, in early August. Cultivator Y is a hydraulically driven unit with a vertical axis cultivating head. Cultivator Z is a ground-driven rolling cultivator with a spring blade that works in between the trees. Infiltration and soil penetration resistance data were collected in both years, and in 2005 soil samples were taken for organic matter and microbial tests. Resistance data were collected using a Rimik® recording cone penetrometer with readings from 0-425 mm depth, at 15 mm intervals, and nine measurements were made per plot. Infiltration was collected using brass rings in 2004, and Mini Disk Infiltrometers (Decagon, Inc. Pullman, WA) in 2005.
Integrated mulch trial. In April 2005, a trial was set up in a new apple (Pinata/M7) block to test the following understory treatments during orchard establishment: bare ground control, bare ground with Brassica meal, wood chip mulch, cultivation, Sandwich system (tillage each side of the tree line and a 45-cm strip of living mulch in the tree line, see Fig. 3), and a Living Mulch (LM) treatment filling the 150 cm wide tree row. The living mulch and sandwich had separate legume and non-legume treatments to test the nitrogen impacts. The LM and Sandwich legume treatments are a species mix of Mt. Barker subclover, black medic, burr medic, birdsfoot trefoil, and bentgrass. The Sandwich non-legume treatment is transplanted sweet woodruff and Corsican mint, while the living mulch non-legume mix contains sweet alyssum, five spot, mother of thyme and bentgrass. Experimental design is a randomized complete block with 5 replicates. Tillage rates were uniform for all tilled plots, and as needed throughout the growing season. Soil measurements included volumetric soil moisture using both a Moisture Meter® and Hydrosense® meter (Decagon, Pullman WA), resistance, and nutrient and carbon status. Resistance data were collected as in the tillage trial. Infiltration data were collected with Mini-Disk® infiltrometers (Decagon).
Objective 3. Develop underseeded cover crop guidelines for humid Northwest environments.
Relay planting or interseeding of legume cover crops is designed to overcome some of the drawbacks of legumes, such as slow establishment in the fall, poor winter ground cover, and poor weed suppression, while improving N accumulation. Relay-cropping presents its own challenges, however, including shading by the cash crop, competition with the cash crop, weed management, and effects on winter hardiness. We are determining which cash crops, cover crops, and conditions are most amenable to successful relay cropping of legumes.
This experiment began in 2003 using other funding, and has been jointly funded by SARE and the WSU Center for Sustaining Agriculture and Natural Resources since 2004. Legume cover crops are interseeded into snap bean (Jade) and sweet corn (Blaze) approximately 30 days after planting the cash crop. Following cash crop harvest, the residue is mowed and the cover crop must grow through the residue. Specific treatments have changed each year as we have narrowed the focus of our experiment based on previous years’ results. Measurements include cash crop fresh yield, cover crop stand at intervals during the winter and spring, cover crop biomass at the time of incorporation, and soil inorganic N three weeks after cover crop incorporation.
In 2005 we moved the location of the experiment to increase the size of the plots and alleys. New plot sizes are 30 ft by 15 ft with 20 ft alleys between each plot along the direction of planting. Each experiment has five treatments and three replicates.
Objective 1. Compare the effects of selected organic row crop and orchard management systems on soil quality as measured by field and laboratory tests and field observations.
Vegetable crop systems experiment.
Bulk density. Soils with low bulk density are generally more porous and contain more organic matter than soils with high bulk density. Bulk density measured in mid season of 2005, 2006, and 2007 was significantly lower in plots receiving the high C amendment (on-farm compost) compared with the low C amendment (broiler litter). Cover crop affected bulk density in 2004 and 2006, with higher bulk density in the ley treatment. There was no difference among cover crop treatments in 2005 and 2007, when the ley treatment was in cash crop production. In 2006 there was an interaction between cover crop and amendment, with a larger amendment effect in the relay and post-harvest treatments than in the ley treatment. This did not appear to be related to the difference in cumulative amendment rates between the ley and other treatments, because the interaction was not present in 2007. The compaction that occurs in the pasture phase of the ley treatment may have masked some of the differences from amendment in 2006. Tillage system did not affect bulk density. Amendment and cover crop effects from 2004-2007 are shown in Table 2.
Soil compaction (0 to 40 cm depth) was measured twice each year using a recording cone penetrometer. Soils have been consistently less compact at the 15 to 30 cm depth in the plots tilled with the spader, in part because of the greater depth of tillage done by the spader. Soil in the ley treatment was significantly more compact in the upper part of the profile in 2004 and 2006 (pasture years), but not in 2005 (cash crop year). There was a cover crop x tillage interaction in 2004 and 2006. In 2004 there was no tillage effect in the ley treatment, while in 2006 there was a smaller tillage effect in the ley treatment than observed in the other cover crops. The tillage effect in the 2006 ley treatment was noteworthy, because the plots had not been tilled since the fall of 2005, indicating tillage effects persisting more than one growing season. We measured a small amendment effect in 2006, with less compaction in the high C treatment in the upper part of the profile. This was the first year that an amendment effect was seen. Fig. 4 compares tillage, cover crop, and amendment effects on compaction for the mid-season measurement in 2004, 2005, and 2006. Data were averaged over the 0-15, 15-30 and 30-40 cm depths and statistical comparisons made over those depth intervals. Results were similar for the early season measurements. Analysis of 2007 data is not complete.
Infiltration measurements were made at planting and mid season in 2005, 2006, and 2007 using a simple single ring infiltrometer with falling head. Infiltration rate was greater with the high C amendment in the spring of 2005 and 2006, but the differences were no longer significant by mid season. In 2007 the amendment effect on infiltration rate was significant in both the early and mid-season measurement, suggesting the amendment effect was persisting later into the growing season (Table 3). Cover crop treatment had no effect on infiltration rates in 2005 and 2007 when all plots were in cash crops during the growing season. The infiltration rate was much lower in the ley treatment in 2006, when it was in the pasture phase. The significant cover crop x amendment interactions in 2006 and 2007 were the result of larger amendment differences for the relay and post-harvest treatments than for the ley treatment, which received no amendments since 2005.
Aggregate stability. No differences in aggregate stability were observed among treatments analyzed through 2006. The lack of difference may reflect difficulty in improving aggregation in the fine sandy soils in our study, or may be because the technique used is not good at differentiating aggregation under the conditions of our study.
Particulate organic matter (POM) C was consistently greater in the plots receiving the high C amendment, indicating the immediate effect of amendment addition on soil C when large amounts of amendments are applied (Table 4). Neither cover crop nor tillage affected POM C.
b-glucosidase and dehydrogenase. Dehydrogenase activity was greater in plots treated with the high C amendment compared with the low C amendment. Differences were significant on four of six dates analyzed between 2004 and 2007 (Table 4). Amendment affected b-glucosidase activity on only one date. Dehydrogenase activity was also affected by tillage. In May 2005 and 2007 activity was lower under conventional tillage, while in 2006 activity was greater under conventional tillage (Table 5). The conventional plots were plowed (moldboard plow) in the spring of 2005 and 2007, suggesting plowing could temporarily reduce dehydrogenase activity. The plots were not plowed in 2006. Additional data will be needed to see if this trend continues. B-glucosidase activity was greater under spader tillage on two of the six dates, with no clear trends emerging.
Nematodes and collembola. Nematode community analyses completed thus far have indicated a significant and consistent effect due to amendment, no significant effect due to tillage, and major changes to the nematode community due to cover crop rotation only during the pasture phase of the ley treatment.
Nematode populations were higher with the low-C broiler litter amendment than with the high-C on-farm compost at mid-summer in 2005, 2006, and 2007. Community analyses for 2005 and 2007 (2006 not completed yet) indicated greater bacterial-feeding nematodes with broiler litter at both of these samplings, but no differences in fungal-feeding, omnivorous/predacious, or plant pathogenic nematodes (Figs. 5 and 6). While the on-farm compost supported less bacterial-feeding nematodes, the community structure was not shifted significantly toward fungal-feeding nematodes or a fungal-based food chain as evidenced by the lack of difference in the channel index (CI).
Community analysis of nematode samples taken mid-summer 2006, when the ley treatment was in pasture, is still underway. However, community analysis has been completed for a sampling done April 2007, before spring tillage. Therefore, this sampling reflects the rotations in place in 2006. At the time of this sampling, the ley had not been tilled for 18 months, the relay treatment for 9 months, and the post-harvest treatment for about 6 months. All of the treatments indicated a greater proportion of fungal-feeding nematodes (greater CI value) than during the summer 2005 and 2007 samplings (Table 6). Furthermore, the difference between the ley treatment and the other two treatments was highly significant, indicating that the pasture phase promoted much more fungal activity and more of a fungal-based food web. The ley had a higher maturity index (MI) than the post-harvest treatment, indicating a more stable ecosystem. Conversely, the post-harvest treatment trended toward a higher enrichment index (EI) (a=0.09) than the ley and relay treatments. A higher EI indicates more dominance in the population of early colonizing bacterial- and fungal-feeding nematodes, and this enrichment in the post-harvest treatment was likely due to the relatively recent tillage and incorporation of organic matter during the previous fall.
Collembola numbers were most strongly affected by tillage frequency and unaffected by tillage type, while amendment had a variable effect. In the late fall of 2005 and 2006 the relay cover crop treatment had greater collembola numbers than the post-harvest cover crop treatment, reflecting the more recent tillage that occurred in the post-harvest cover treatment prior to planting cover crops. During 2006, when the ley treatment was in the pasture phase and undisturbed by tillage, it had much greater collembola numbers than the other treatments (Fig. 7). Collembola population in the pasture in 2006 was similar to that in the adjacent grass alleys, which have been untilled since 2003.
Soil pH, exchangeable K, and organic matter (0-30 cm depth) were higher with the High C amendment (on-farm compost) compared with the Low C amendment (broiler litter) (Table 7). Exchangeable K levels showed increasing trends for both amendment treatments over time. Bray-P levels are not shown, because they were very high (> 100 mg/kg) at the start of the experiment, due to historical (pre-1990) applications of P fertilizer and chicken manure. The ley treatment, which received biennial rather than annual amendment applications, had slightly but significantly higher pH in 2005 and 2006, and significantly lower exchangeable K in 2006. Significant cover crop x amendment interactions in 2005 and 2006 were the result of smaller amendment effects in the ley treatment, an expected result of less frequent amendment applications.
Crop yield. Treatments affected crop yields in a few instances during 2003-2006 (Table 8). Amendment effects on snap bean in 2003 and broccoli in 2006 appeared to be related to the amount and timing of available N. Amendment and cover crop differences in spinach in 2004 were related to seedbed condition, resulting in differences in spinach stand.
In 2007 both cover crop treatment and tillage had significant effects on crop yield. Snap bean, broccoli, and spinach following the post-harvest cover crop all had greater yields than following either the relay cover crop or pasture. Moderate to poor stands of the relay vetch would have reduced N contribution from the cover crop compared with the more vigorous stands of the post-harvest rye-vetch blend. The ley treatment received no amendments and relied solely on N from the plowed down pasture. Interpretation of yield responses in terms of soil available N status in the cover crop treatments will be done when 2007 N analyses are complete. Spader tillage increased yields of broccoli, spinach, and squash compared with conventional tillage. The reason for the yield benefit from spader tillage is not yet apparent. We have observed few measurable differences in physical and biological soil quality between the spader and conventional tillage treatments, with the exception of soil compaction in the subsurface. Differences reflected in the yields may be evolving as each system matures.
Post-harvest soil nitrate showed different trends each year from 2004 through 2006 (Table 9). In 2004 nitrate-N was much greater in the post-harvest and relay cover crop treatments than the ley treatment, a reflection of management during that year. The ley treatment was in the pasture phase during 2004, receiving no amendments and having an efficient root system scavenging nitrogen, resulting in a very low residual. The higher nitrate levels in the other treatments indicate some unutilized available N at the end of the crop season. The amendment treatments had similar residual nitrate levels, indicating similar levels of N availability to the crops. In 2005, when all treatments were planted to the cash crop, all post-harvest nitrate levels were low, indicating nearly complete use of available N by the cash crop. Squash yields in 2005 were lower than those observed in 2006 and 2007, and the plots may have had sub-optimal levels of available N. However, in-season levels of nitrate-N (see below) appeared adequate for agronomic yields. The ley treatment was back in the pasture phase in 2006, and had very low residual nitrate-N levels as expected. Nitrate levels following application of broiler litter (low C) were greater than those following application of on-farm compost (high C), indicating application of the broiler litter at levels greater than crop need. Nitrate levels were also higher following spader tillage than conventional, although the reason for this difference is not clear.
In-season soil nitrate-N was measured several weeks after amendment application in 2005 and 2006 and mid-season in 2005. Results showed significant available N for all treatments in 2005 and all treatments except the pasture in 2006 (Table 10). The low C broiler litter supplied more early season N than the compost, as expected. The ley provided the least early season N in 2005, but mid season N levels indicated adequate levels for all treatments. The low level in the pasture in 2006 was a result of no amendment application, and efficient use of available N during the pasture phase. Data for 2007 have not been analyzed.
Conclusions. Amendment had the greatest effect on soil properties. The High C amendment (on-farm compost) decreased bulk density and increased POM C, total OM, dehydrogenase activity, soil pH, and exchangeable K compared with the low C amendment (broiler litter). All of these effects were measured each year, and were the result of an immediate response to incorporation of the compost. The High C amendment also increased infiltration rate, with a transient increase observed in 2005 and 2006 and a more lasting increase in 2007. Amendment did not affect compaction until 2006, when a significant decrease in compaction in the surface 15 cm was observed with the High C amendment. Amendment has not affected aggregate stability, while effects on b-glucosidase activity and collembola are not yet clear. Total and bacterial-feeding nematodes were increased by the low C amendment (broiler litter), probably as a response to N rather than C. Crop yield also showed occasional responses to amendment, which usually appeared to be related to N. The amendment effects were greater in the relay and post-harvest cover crop treatments, which received amendments every year, compared with the ley treatment, which received amendments only in 2003 and 2005..
Cover crop effects were often related to the frequency and timing of tillage in each cover crop rotation. The post-harvest cover crop treatment was fully tilled every spring and fall, while the relay treatment was tilled in the spring only (except for 2004), and the ley was tilled spring and fall every other year. During the pasture phase, the ley treatment had higher bulk density, greater compaction in the surface soil, slower infiltration, and fewer nematodes, but greater nematode diversity and greater numbers of collembola. Collembola responded negatively to tillage, but numbers rebounded within months when the soil was not disturbed. In 2007 yields were affected by the previous cover crop, a likely result of soil nitrogen status.
The type of tillage affected soil compaction, with consistent differences between spader and conventional tillage in the 15-30 cm depth occurring every year. Plots tilled with the spader also showed occasional, transient differences in other soil properties, and in 2007 showed improved crop yields.
On-farm cooperators. Mean bulk density, infiltration, aggregate stability, and soil test values are compared across on-farm sites in Tables 11-13. Nutrient analyses for 2006 were similar to 2005, and only 2005 data are shown.
Soil test values showed a wide range in organic matter, reflecting native soil properties, current and past management, and length of time in organic production. P and K levels also varied widely, with four locations having excessive levels of P. Post-harvest nitrate levels were measured at only three farms in 2005 and four farms in 2006, but all suggest conservative N management, with low levels of nitrate-N available for leaching at the end of the growing season. The farm with the lowest soil N levels showed visible signs of N deficiency in the crops. Infiltration rates reflected soil texture and to a lesser extent management. Infiltration data for 2005 (not shown) and 2006 were similar. Penetration resistance had different patterns among farms, but was generally within the range of that observed in the systems experiment (Fig 8).
Orchard mulches and weed management experiments
Over the three years of the Tillage Comparison Trial, tree performance and fruit size tended to be better in the wood chip plots and worse in the tilled plots relative to the mowed control. This was an established block of apples, where some weed competition can be tolerated. Based on the results, we hypothesize that the poorer tree performance with tillage is likely due to root pruning. No root sampling was done, and the trees are on a wire trellis and therefore not subject to increased leaning with root pruning. The tillage implements used worked at 5 to 10 cm depth. If more shallow tillage was possible, this problem might be avoided. To that end, we have purchased a new tillage implement, the Spedo blade, to test this idea in future experiments.
Results from the Integrated Mulch Trial were more mixed. This was a new planting of apple, and weed pressure was very high every year. Overall tree performance in Year 1 was poor, due to a combination of stressed nursery stock, high winds, and tree shock from late heading. Still, treatment effects were clear, with tilled plots and wood chip plots performing the best, and plots with living mulches clearly stressed for nutrients. In Year 2, all trees grew very well, and again treatment differences were apparent in a manner similar to Year 1. However, the sandwich non-legume plots performed as well as the tilled and wood chip, suggesting that this system could provide an acceptable compromise between tree growth and soil quality maintenance. The increase in overall fertilization in Year 2 clearly contributed to the improved growth. Tree performance in Year 3 was mixed, with periods of obvious stress during the season and more spatial variation likely due to soil differences. Trees were pruned in winter but not supported, due to the phasing out of this planting. Fruit load was adjusted based on the trunk cross-sectional area. While final tree size measurements have not yet been taken, it appears that the fruit load may have been excessive for a number of treatments and held back growth. In addition, the lack of tree support led to much branch bending and some tree leaning that also compromised vegetative growth. Ironically, the unfertilized control trees (with virtually no fruit) often looked best visually compared to the other plots, which was not the case in the first two years.
Soil Organic Matter. Soil organic matter was measured by Loss on Ignition in the Tillage Comparison Trial in Year 3 (0-15 cm). Neither the wood chips nor the tillage were significantly different than the control, but they were significantly different from each other. Surprisingly, the tilled plots showed higher organic matter than the wood chips (Table 14.). Analysis for active carbon (0-15 cm) showed no difference among treatments (Figure 9).
In the Integrated Mulch Trial, total soil C was measured in September 2005 and 2006 (0-10cm) using dry combustion (Leco C-N Analyzer) (Table 15). There were few significant differences among treatment in 2005. The fertilized control (CTL1) was not different from the tillage treatments (WW). All plots showed an increase in total C from 2005 to 2006 (not all significant). The effect of compost rate became more apparent, with significant increases for living mulch legume (LML), wood chip (WC), and tillage (WW). In 2006, the fertilized control (CTL1) was greater than tilled (WW.5, WW1) but similar to WW1.5, suggesting that the additional C from the higher compost rate was compensating for increased C losses due to tillage. Living mulches had the highest total soil C, which exceeded the fertilized control. This indicates an additional carbon contribution from the living mulch tops and roots. The wood chip mulch did not show a similar elevation and carbon levels were similar to the tilled plots. This was surprising as the wood chips experienced rapid decomposition each year and needed to be renewed each spring. The unfertilized control (CTL0) showed little change in soil C from 2005 to 2006. The low and medium compost tilled plots were significantly lower than CTL0 in 2006, again illustrating the potential for loss of soil C with tillage.
Particulate organic matter was highest in the living mulches and lowest in the wood chip, tilled, and unfertilized control treatments. C mineralization over 54 days followed a similar pattern (Figure 10), as did soil dehydrogenase. Nematode abundance in a summer soil sample was also significantly higher in the living mulches than the other treatments.
Together, these results all support our hypothesis that the living mulches would improve soil quality over the control and the tilled plots, but at the expense of optimal tree growth. In a new planting, tillage resulted in tree growth equal to wood chips, with soil quality also similar. This contrasts with the measured tree decline with tillage in an established orchard compared to wood chip mulch.
Nitrogen Dynamics. Organic orchards in Washington State typically struggle to provide enough N to the trees without excess losses to the environment. One soil quality goal is to increase the internal capacity of the orchard to supply N at the needed times and amounts. Our Integrated Mulch Trial incorporated legume treatments expressly for this purpose. We were able to produce up to 3.3 MT dry matter/ha in the living mulch legume treatments for a single cutting. This contained 120 kg N/ha in the aboveground biomass, which based on previous experiments would have about 48% mineralization over the following 4 weeks. Two or three cuttings would be possible. Thus the living mulch has the potential to provide most of the N needed for organic apples (60-80 kg N/ha/yr). However, appropriate timing is more difficult to achieve, with a high N demand in early spring prior to the onset of shoot extension, and another optimal application period just prior to harvest. The legume living mulch cannot produce enough biomass early in the season to meet that demand; however, those plots did appear to conserve more N over winter and to show the greatest increase in soil N over time. The legume could be managed to supply late summer N for a pre-harvest application.
Based on our monitoring of soil and tree N, the living mulches appeared to outcompete the trees for inputs of available N. Further study is needed to determine how much N from decomposing legume residue is simply recycled back into the legume versus taken up by tree roots. A mow and blow system with legumes planted in the alley and blown on to the tree row would help solve this problem. Living mulch legume had significantly higher available N than all other treatments in summer 2006 (Figure 11), and trees in those plots had among the highest leaf nitrogen (July 2006). But tree growth in these plots was among the lowest (Table 16). Again, while the living mulches offer promise to enhance the soil N, in a new orchard, they are too competitive for the young trees.
We experimented with direct soil EC measurement as an indicator of soil nitrate status using a probe from Spectrum Technologies. Several problems were encountered. First, the probe is sensitive to soil moisture content, and thus providing uniform conditions for this during a field sampling proved difficult. We tried to measure at the same point after an irrigation, but moisture content changes during the course of a day and the irrigation pattern is not all that uniform. Second, we were trying to monitor nitrate release from the compost applied, and it was applied only around the base of the tree (generally on top of weed fabric). We attempted measuring right under the fabric as well as next to it, and also measured between trees where no compost was used. But this non-homogeneous pattern, added to typical soil spatial variability, greatly increased the noise and made it difficult to see any treatment effects. Third, the probe tip, which has steel electrodes embedded in a plastic matrix, wore out extremely fast when we used it directly in the soil to 60 cm depth in an attempt to monitor nitrate movement. Thus, we are uncertain how much readings change as the tip wears. After replacing the original probe, we switched to a method where we made a pilot hole first with a steel rod, and then inserted the probe tip into the hole for our reading, to minimize wear. This appeared to work, but greatly slowed down the process. Fourth, we do not know how much other anions might be influencing the readings, as we were not in a position to test this.
However, looking at the results from the June 2006 measurements, the EC probe did appear to provide a relative relationship among treatments that was plausible (Fig. 12). The readings for the unfertilized control were consistently the lowest of all treatments, and did not change at all with depth. And they were similar next to tree and between tree, suggesting that they represented background soil EC. The effect of increasing compost rate can be seen among the tillage treatments (WW). Tillage led to higher readings than other treatments, perhaps because more compost was incorporated into the soil and in contact with organisms, and perhaps due to some additional soil N mineralization.
We did compare the direct soil EC reading with a lab EC reading and a lab nitrate reading for a set of the same samples. The correlations were reasonable (Figs. 13a-d). However, for the November 2006 measurements, the correlation was poor (R2=0.29) (Fig. 14). For this comparison, the soil core was removed from the plot, mixed and placed in a bag, and the EC reading then taken from the bagged soil. A subsample was then analyzed for available N.
A direct tree measurement of N status will be more useful than a soil indicator. We have used leaf greenness (SPAD) with good results, and new technology such as the Green Seek may offer better options for quick N status evaluation. Yet, inexpensive soil N monitoring would be useful for organic growers in helping them better understand the release rate of N from organic fertilizers and to monitor the fate of the N, especially in irrigated systems and on coarse-textured soils. Loss of N through leaching not only represents a potential environmental cost but a real cost to the producer for wasted fertilizer.
Soil Physical Properties. Water infiltration was measured each year in the Tillage Comparison Trial, using a single ring the first year and a tension infiltrometer the other two years. No trend in decreasing infiltration with increasing tillage frequency was observed (Table 17). For Years 2 and 3, the mowed control plots had significantly greater infiltration at both tensions. We can speculate that the weeds that did grow in these plots left root channels that improved infiltration. However, we expected better infiltration with wood chip mulch than tillage but this was not the case.
Infiltration was also measured in the Integrated Mulch Trial in summer 2006 and 2007 using a tension infiltrometer. Living mulches tended to have the highest infiltration in 2006, while there were no significant differences in 2007. However, the wood chip plots again exhibited lower infiltration rates than other treatments in both years (Table 18).
Another indicator of changing soil physical condition was the significant increase in tree leaning from tillage in the Integrated Mulch Trial. How much was due to root pruning versus soil loosening by tillage is not known.
Soil resistance was measured in both trials in 2005 but no clear conclusions could be drawn. Measurements are planned for the Integrated Mulch Trial this fall.
Objective 2. Evaluate the NRCS Soil Conditioning Index as a tool for soil quality assessments.
We ran a series of scenarios based on field data (crop yield, cover crop biomass, amendment application rates, and field operations) for the different systems in the WSU Puyallup organic systems experiment. We evaluated 1-year (2005) and 2-year (2004-05) simulations for the 6 cover crop and amendment combinations, but did not compare tillage systems, because the Soil Conditioning Index (SCI) database does not contain field operation inputs for spaders.
Preliminary results indicate a wide range of soil conditioning indices among the experimental treatments. One-year results ranged from –0.13 to +0.12, while 2-year results ranged from –0.02 to +0.36. The two-year results were higher because they included a year with more time spent in cover crops and less in cash crops. This range of values suggests that relatively large differences in soil quality may become apparent over time in the experiment. A problem with adapting the SCI to use with small-scale organic farms is that many of the common equipment and tillage operations on these farms are not part of the current SCI database. Also, it is clear that accurate record keeping is needed to gain confidence in the SCI results.
The NRCS is now using the RUSLE model as a basis for estimating changes in soil organic matter rather than using the SCI. We have begun to work with the local NRCS to test the RUSLE model with our cropping systems, but we are still in the early stages of work with RUSLE, and do not have additional results to report.
Objective 3. Develop underseeded cover crop guidelines for humid Northwest environments.
Treatments each year and key management dates are shown in Table 19
Summary of 2003-04 experiment. Hairy vetch and woolypod vetch were well established by late October. Red clover had a fair stand, but common vetch and crimson clover performed poorly and were dropped from the trial. Some winter damage occurred to the vetch in January during a week-long period of snow cover and cold temperatures, but the vetch recovered by spring. Cover crops were harvested in early April 2004 and biomass was similar in the corn and bean experiments. Hairy vetch had the greatest biomass (2170 lb/acre dry weight) followed by woolypod vetch (1760 lb/acre) and red clover (720 lb/acre). Hairy vetch significantly increased early season inorganic N over the control treatment, but red clover did not. Later incorporation would likely have substantially increased biomass and N.
2004-05 experiment. Establishment of the cover crops was poorer than the previous year. Heavy crop residue from the corn, followed by early rains that compacted the crop residue reduced the fall stand of the interseeded cover crops. Cover crops partially recovered during the fall, and by January ground cover averaged 40-50% for hairy vetch and 30-35% for red clover in both corn and bean plots. Cold night temperatures (10 consecutive nights between 20 and 25 F) and lack of insulation by snow in February reduced ground cover to 10 to 20% for hairy vetch by early March, while red clover ground cover was not affected. This illustrates one of the potential problems with interseeded hairy vetch – reduced winter hardiness when planted in mid-summer.
Red clover had the greatest biomass at incorporation (12 May), averaging 1980 lb/acre dry matter (mean of corn and bean experiments), while average hairy vetch yields were only 900 lb/acre. Post-harvest rye-vetch (planted 4 October as a comparison to the relay plantings) yielded nearly 6000 lb/acre of dry matter.
2005-06 experiment. The design was modified to make a direct comparison of interseeded vs. post-harvest plantings of hairy vetch and red clover. Fall stand of the interseeded hairy vetch and red clover was much improved over the poor stands 2004-05, and similar to 2003-04. Stand ratings over the winter showed trends similar to those observed in previous years: better initial stands with hairy vetch compared to red clover, better initial stands following beans than corn, a decline in the hairy vetch stand during the winter, and a recovery in the spring. Post-harvest plantings of hairy vetch and red clover (done on October 4) had weak stands and never caught up to the interseeded plantings. Biomass yield at harvest averaged 2800 lb/acre dry wt for hairy vetch and 1900 lb/acre for red clover. Soil nitrate-N at 2 weeks after cover crop incorporation greatest for interseeded hairy vetch (9 mg/kg) and 3-4 mg/kg for other treatments. The trend was similar in mid July, with only interseeded hairy vetch (26 mg/kg) approaching an adequate level for a pre-sidedress nitrate test. No differences in weed pressure were observed during the 2006 cash crop season. Neither were differences in cash crop yield observed, although this may have been the result of high yield variability across the plots. Cover crop stand data are summarized in Table 20, and cover crop biomass and soil nitrate are summarized in Tables 21 and 22.
2006-07 experiment. Fall and winter stands of relay planted hairy vetch and red clover were similar to those observed in 2005-06. Biomass was much lower than the previous year, however, averaging only 940 lb/acre for hairy vetch and less than 500 lb/acre (stand not harvested) for red clover. The red clover was inadvertently drilled instead of broadcast, which reduced stand and yield. Fall seeded legumes were not planted until late October, and produced too little biomass to harvest. The reason for the reduced biomass in the relay plantings was not clear. Some winter damage of the hairy vetch occurred, but it was not more severe than in previous years. In contrast, hairy vetch that was relay planted between rows of spinach in early September in an adjacent experiment (organic systems experiment described in Objective 1) yielded 1900 lb biomass/acre. Samples were collected for soil nitrate in late June, but analyses are not complete.
Objective 4. Share findings with organic vegetable and fruit producers and other producers interested in sustainable practices in the Northwest.
We held 5 field days at WSU Puyallup and one at Full Circle farm focused on organic vegetable production systems, cover cropping, and soil quality. Field day attendance ranged from 40 to 90. A 1-evening soils module was taught at 3-5 locations per year as part of WSU classes for small farms operators. Attendance per class ranged from 10 to 30. Workshops on soil management were held for Hmong farmers in western Washington and Latino farmers in eastern Washington, and at the Western Washington Horticultural Association annual meeting. In addition we presented two posters at the American Society of Agronomy national meetings, which brought our work to a national audience. We will make two presentations at the Washington Tilth Producers meeting in November, 2007. Information on our soil quality and organic systems research is available on our web site: http://www.puyallup.wsu.edu/soilmgmt/SusAg.htm. The website is revised each January with current information, including portions of this report. Presentations on orchard understory management were given at 8 grower meetings, the Washington Horticulture Association annual meeting, the American Society of Horticultural Science annual meeting, and the International Horticulture Congress, reaching over 400 growers, researchers, and consultants. Project reports are posted at http://organic.tfrec.wsu.edu/OrganicIFP/OrchardFloorManagement/Index.html . The PIs were participants in the Building Soils for Better Crops workshop and field day in the Columbia Basin in February and July 2006 respectively.
Bongers, T., and H. Ferris. 1999. Nematode community structure as a bioindicator in environmental monitoring. Trends Ecol. Evol. 14:224-228.
Cambardella, C. A. and E. T. Elliott. 1992. Particulate soil organic-matter changes across a grassland cultivation sequence. Soil Sci. Soc. Am. J. 56:777-788.
Ferris, H., T. Bongers, and R.G.M. de Goede. 2001. A framework for soil food web diagnostics: extension of the nematode faunal analysis concept. Appl. Soil Ecol. 18:13-29.
Forge, T.A., S. Bittman, and C.G. Kowalenko. 2005. Responses of grassland soil nematodes and protozoa to multi-year and single-year applications of dairy manure slurry and fertilizer. Soil Biol. Biochem. 37: 1751-1762.
Forge, T. A., E. Hogue, G. Neilsen, and D. Neilsen. 2003. Effects of organic mulches on soil microfauna in the root zone of apple: Implications for nutrient fluxes and functional diversity of the soil food web. Appl. Soil Ecol. 22:39-54.
Gale, E.S., D.M. Sullivan, C.G. Cogger, A.I. Bary, D.D. Hemphill, and E.A. Myhre. 2006. Estimating plant available nitrogen release from manures, composts, and specialty products. J. Env. Qual. 35:2321-2332.
Granatstein, D. and Kirby, E. 2007. Profile of organic crops and livestock in Washington State – 2006. Washington State University Center for Sustaining Agriculture and Natural Resources. http://organic.tfrec.wsu.edu/OrganicStats/WA_CertAcres_06.pdf
Moldenke, A.R. 1994. Arthropods. pp. 517-542. In R. Weaver, J. S. Angle et al. (eds). Methods of Soil Analysis: Microbiological and Biochemical Properties, Soil Sci. Soc. Am. Book Series, no. 5. Madison, WI.
Nimmo, J. R., and K. S. Perkins. 2002. Aggregate stability and size distribution. p. 317 – 329. In J. H. Dane and G. C. Toppe (eds) Methods of Soil Analysis Part 4. Physical Methods. Soil Sci. Soc. Am. Book Ser. 5. SSSA, Madison, WI.
Soil Quality Institute. 1999. Soil quality test kit guide. United States Department of Agriculture. http://soils.usda.gov/sqi/files/KitGuideComplete.pdf
Tabatabai, M.A. 1994. Soil enzymes. pp.775-833. In R.W. Weaver, J.S. Angle et al. (eds). Methods of Soil Analysis. Part II: Microbiological and Biochemical Properties. Soil Sci. Soc. Am., Madison, WI.
Vegetable project: Our results are leading to strategies for relay planted cover crops in irrigated vegetable production systems, although our recommendations are still preliminary. Organic growers at our field days have responded enthusiastically to our work on relay planted cover crops. Our research site gives growers an opportunity to compare different relay cover crop strategies. Growers are eager to obtain additional information from our research, to make informed decisions on adopting relay-planted cover crops. Three of our cooperating farmers have integrated relay cover crops into their farming systems on a trial or full-scale basis.
Our soil quality research is still in early stages, so we do not have on-farm impacts to report yet. We have demonstrated that different organic management systems affect soil quality in different ways in the short run; we plan to continue this research to gain a better understanding of long-term effects. We expect that long term assessment of soil quality in our experiment and on participating farms will encourage farmers of intensively managed high-value organic crops to adopt practices that provide more long term benefits to soil quality. Even though the measurable effects of modified tillage using the spader are small at this point, growers are impressed by the quality of seedbed that it can produce after a single pass. The spader appears to be most suitable for small farms and light to medium textured soils. We have applied for funding to conduct economic analyses of the different cropping systems in our experiment so that long-term soil quality gains can be better assessed in terms of economic costs and benefits.
Orchard project: Orchardists are keenly interested in soil quality, especially as the number and acreage of organic orchards increases. They typically rely on tillage for weed control and incorporation of fertility amendments. However, a few more growers each year begin using mulches, particularly on low performing parts of the orchard, and report favorable results. Our two years of fruit yield results showed that the wood chip mulch led to a greater economic return than tillage, even after including the cost of a mulch application (about $900/acre, which could last 3 years before renewal). Previous studies in British Columbia found up to a 50% increase in fruit yield compared to an herbicide control, but no tillage treatment was included in the trial. More orchardists are moving away from tillage for weed control, and at present are trying flame weeding as a cost effective alternative that should not have the potential soil quality loss of tillage. These same growers are interested in how they might grow a portion of their nitrogen need internally. Several are testing legumes in the drive alleys and using a mow-and-blow strategy.
Education and Outreach
Pritchett, K. 2006. MS. Management effects on soil quality in organic vegetable cropping systems in western Washington.
Collins, D.P. 2008. PhD. (in progress). Cover crop, amendment, and tillage effects on collembolan, nematode, and decomposer communities in an organic vegetable system.
Journal articles (in preparation)
Collins, D.P., C.G. Cogger, A.C. Kennedy, T. Forge, H.P. Collins, A. I. Bary, E. Maki
Cover crop, amendment, and tillage effects on collembolans and nematodes in an organic vegetable system.
Cogger, C.G., Collins, D.P., Pritchett, K., Bary, A.I., and Kennedy, A.C. Soil quality and productivity in intensive organic vegetable crop systems.
Outreach programs and events
Field days: Five at the WSU Puyallup Organic Systems experiment and one at Full Circle Farm focused on systems research, cover cropping, soil quality, and pastured poultry production. (> 350 participants)
Classes: 3-5 soils classes and 2-3 composting classes per year for Cultivating Success and Living on the Land series for small acreage farmers. (50-80 participants/year)
Workshops on soil management for immigrant farmers: One for Hmong farmers in western Washington and one for Latino farmers in eastern Washington. (30 participants)
Building soils with better crops workshop and field day (> 200 participants)
Orchard understory management presentations to growers (8 presentations, > 400 participants)
Field tours for growers, students (high school, technical college, university, and graduate students), state senators and representatives, congressional staff, and international visitors from Eritrea, Peru, and Ecuador (18)
Presentations at regional, national, and international meetings (6)
Education and Outreach Outcomes
Areas needing additional study
The organic vegetable cropping systems experiment is designed to be a long-term project, and our data show that differences among systems are evolving over time. Our goal is to be able to run the experiment for a minimum of 12 years to assess longer term effects of the systems on soil quality and productivity.
An economic analysis of the vegetable farming systems experiment is needed, comparing costs, benefits and risks. We are actively seeking funding to carry out this analysis.
We are initiating a new project to assess available N contributions from the cover crop systems in the vegetable experiment, and to develop guidelines to assist farmers in predicting cover crop N contribution to the next cash crop.
A detailed farmer survey is needed to assess knowledge changes, adoption of practices, and impacts. We will incorporate a proposal to conduct this survey into grant proposals written during the next year.
Additional research on species selection and management for living mulches and cover crops is needed to improve their utility in organic orchards.