Development of sustainable vegetable production systems for south Florida and Virginia based on use of cover crops and precision irrigation

Florida Component:
No.1. Biologically-based tomato production system*.
*Note: This experiment was initiated in 2001 without the benefit of support from SARE Project# LS03-148. When the latter grant was approved, this project was continued and expanded.

Field experiments were conducted for three years at Homestead, Florida. In the first two years, the experiments were conducted at the Tropical Research and Education Center (TREC), University of Florida, Homestead, and in the third year at Pine Island Farms (PIF), a commercial vegetable production farm about 22 miles northeast of TREC. The soil at TREC is Krome, very gravelly loam (loamy-skeletal, carbonatic hyperthermic lithic Udorthents) and consists of about 33% soil and 67% pebbles (>2mm). The soil at PIF is Opalocka (sandy, siliceous, hyperthermic lithic Udorthents). The experiment in year 1 consisted of three treatments: two cover crop treatments, ‘Iron Clay’ cowpea and velvetbean (cultivar not specified), and one methyl bromide/ chloropicrin (MC-33) treatment at 392 kg ha-1 which contained 330g kg-1 chloropicrin (Helena Chemical Co., Florida City, F La.). In year 2, ‘Tropic Sun’ sunn hemp was added to the year 1 treatments as a third cover crop treatment. Each year the fields were disked three times and raised beds 15-cm high were formed. The experiments were laid out in a randomized complete block design. Each treatment was applied to three raised beds each 13.5-m long by 1.8-m wide (center-to-center). A distance of 5m separated the various treatments. Plots 6-m long were randomly designated from the middle row for yield determination. There were four replications per treatment in each year. In order to suppress weeds in the aisles between the raised beds S-metolachlor and metrabuzin were incorporated into the soil after the beds had been formed. If needed one or more additional in-season applications of metrabuzin were made. Fallow plots were maintained weed-free by rototilling.
In year 3, the experiment at PIF consisted of four treatments—sunn hemp, velvetbean, an MC-33 soil fumigation treatment following a summer sorghum sudangrass cover crop, and a fallow treatment which was kept weed free. The field was disked as in the previous two years and the raised beds were formed as described earlier. The experiment was laid out in a randomized complete block design with four replications for each treatment. Each treatment consisted of a double-bed 26-m long with plants spaced 0.6 m within the row. Because of some differences in management practices and dates of various operations from year to year, some imposed by the grower, each year will be described separately to point out these differences.

Year1. Rhizobium-treated seeds of the leguminous cover crops and sorghum sudangrass were seeded in mid-June using a Tye no-till drill (AGCO Corp., Lawrenceville, GA). Seeding rates for cowpea and velvetbean were 112 kg ha-1 and 34 kg ha-1, respectively. The cover crops received no fertilizer during the growing season. They were flail-mowed on 16 Aug. Above-ground samples of biomass were taken and biomass yield was determined. The beds were reseeded and biomass samples were taken again for biomass determination before the crops were flail-mowed on 12 Oct. The cover crop residues were incorporated into the soil using a spader (Imants Heavy Duty, Imants USA/ Autrusa Co., Perkeomenville, PA) during the second week of Oct. During the first week of Dec., dry fertilizer (6N-2.6P-10K) was applied at the rate of 1123 kg ha-1 for the MC-33 treatment and 392 kg ha-1 for the two cover crop treatments. The fertilizer was banded 25cm each side of the bed center and roto-tilled into the soil for all treatments. The beds were then reformed. Immediately thereafter, MC-33 was injected into the appropriate beds at 392 kg ha-1 of MeBr, and two drip lines and white-on-black plastic mulch were laid on all beds. Nematode-tolerant ‘Sanibel’ tomato seedlings were transplanted into the beds on 19 Dec. Beginning 15 Jan., an additional 180 kg ha-1 of N was applied as liquid 4N-0P-6.6K to all treatments through the drip lines twice a week starting at a rate of 1 kg N ha-1 day-1, and beginning 8 Feb., it was increased to 3.6 kg N ha-1 day-1 up to the first harvest. The rate was thereafter reduced to 1 kg N ha-1 day-1 up to one week before the final harvest. This made a total of 247 and 214 kg N ha-1 for the MC-33 and the cover crop treatments, respectively. Due to low price during that season, only two tomato harvests were made, one harvest on 30 March and the other on 13 April.

Year 2. As in year 1, the experiment was laid out in a randomized complete block design with four replications. Rhizobium-treated seeds of cowpea, velvetbean and sunn hemp were seeded on 9 May. Seeding rates for cowpea and velvetbean were as in the previous year and for sunn hemp, it was 56 kg ha-1. On 13 July, the cowpea was flail-mowed at ground level and the sunn hemp was flail-mowed at 75 cm above ground level to induce profuse branching (Abdul-Baki et al., 2001). Residues from the first mowing were left to decompose in the field. The cowpea plots were reseeded on 18 July. Subsequently, on 1 Oct., all three cover crops were flail-mowed and incorporated into the soil. As in the previous year, biomass samples were taken from each treatment before mowing, dried, and total biomass determined. In mid-October the raised beds were reformed and 1235 kg ha-1 of 6 N-2.6 P-10 K fertilizer was applied and roto-tilled into the soil of all treatments. Two drip lines were installed into each bed, MC-33 was applied to the appropriate treatment as before and the beds were immediately covered with polyethylene mulch. Five-week-old seedlings of ‘Leila’, a nematode-susceptible cv. were transplanted into the beds on 9 Nov. As in year 1, an additional 180 kg ha-1 was applied as liquid 4N-0 P-6.6 K to all treatments through the drip line thus resulting in a total application of 247 kg N ha-1 for all treatments following the application regime of year 1. Three harvests were made: 6 Feb., 5 March, and 25 March.

Year 3: The experiment in the third year was conducted at Pine Island Farms (PIF), a commercial production farm described by the grower to have moderate-to-high nematode population densities. Two cover crop treatments were velvetbean and sunn hemp. Two additional treatments were fallow (no cover crop or soil fumigation), and summer sorghum sudangrass followed by MC-33. This last treatment is commonly used by large-scale conventional growers in south Florida. The cover crops (including sorghum sudangrass, which was part of the MC-33 treatment) were seeded on 9 June and an irrigation system was installed to deliver water as needed. Seeding rate for sorghum sudangrass was 45 kg ha-1 and for the other cover crops, as in previous years.
The fallow plots were kept weed free. On 22 Aug. the cover crops were prematurely plowed under before the biomass samples could be taken. Fertilizer was applied to all treatments as N6-P6-K12 at an N rate of 247 kg ha-1 and incorporated into the soil. The beds were reformed and MC-33 was injected into the appropriate beds as in previous years. Two drip lines and white-on-black plastic mulch were laid on all beds. On 11 Nov., five-week-old seedlings of nematode-resistant ‘6153 Rogers’ (Rogers Seed Co., Greensboro, NC) were transplanted. Additional liquid fertilizer (N13-P0-K46) was applied twice-a-week for five weeks starting at fruit set at a rate of 8 kg per application thus bringing the N application to a total of 300 kg ha-1. All other cultural practices were the same as in the previous two years.
Eight on-vine-ripe fruit harvests were made at the breaker to pink stages between 17 Feb. and 31 March. The last harvest included unripe marketable fruits. The fruits were graded in accordance with Florida Tomato Committee standards (Brown, 2000) and separated into extra-large, large, and medium. Market prices during that year were favorable even for medium-size fruits. All cultural operations were done by the grower in the same manner he managed the rest of his large-scale tomato field. Pesticides were applied according to standard growers’ practices (Maynard and Olson, 2000).

Nematode and root health evaluation. Root-knot nematode gall ratings were evaluated on 250 cm3 soil samples taken at three time intervals during the production season: before seeding the cover crops; before transplanting the tomatoes; and at the end of tomato harvest. Likewise, tomato root health was determined at the end of harvest by examining necrosis caused by root rot (primarily Rhizoctonia solani), and gall formation by the root knot nematode. In each case a 0 to 5 scale was used as proposed by Taylor and Sasser (1978) for nematode density and by Coyler (1988) for root rot.

Statistical Analysis. Data were analyzed statistically using analysis of variance (ANOVA) and Duncan’s multiple range test using SAS (Version 8.1, SAS Inst. Inc., Cary, NC, USA).

No. 2. Use of summer cover crops and organic mulch to improve tomato yields and soil fertility.
Site description
The experiment was conducted at the Tropical Research and Education Center, University of Florida, Homestead, Florida. At this site the average annual rainfall is 1,499 mm of which an average 76% falls between June and October, the annual temperature averages 23.9 oC and usually ranges from a maximum of 35 oC in June to a minimum of 5 oC in January (Wang et al., 2002a). The soil is a Krome very gravelly loam (loamy-skeletal, carbonatic, hyperthermic Lithic Udorthents), which consist of 58.8% gravel (> 2 mm), and a gravel-free fraction. The latter consists of 48.4% sand, 30.3% silt, and 21.3% clay. CaCO3 comprises about 60% of the soil, organic C is 11.3 to 18.0 g kg-1, total N is 0.9 to 1.3 g kg-1, the soil pH is 7.6 to 7.8 (water), ABDTPA (ammonium bicarbonate-diethylene triaminepentaacetic acid) extractable phosphorus (P) and potassium (K) are 16.5 to 26.8 mg kg-1 and 123.5 to 132.6 mg kg-1, respectively.

Experimental design and management
The field experiment was conducted in three consecutive years from 2002 through 2005 using a randomized split-plot design. The main plots consisted of four cover crops, sunn hemp, velvetbean, cowpea and sorghum sudangrass, grown from June to September, and fallow as a control, and all plots were on raised beds. In the first year compost was applied as an organic mulch (OM) to some plots at the single rate of 50 t ha-1 dry weight, while white on black plastic mulch (PM) was installed on others. In the second and third years, the 50 t ha-1 treatment was applied to the very same plots as in the first year of the experiment, and three additional compost treatments were made, i.e., 25 t ha-1, 50 t ha-1, and 75 t ha-1. Thus the four organic mulch treatments are denoted as OM1-25, OM1-50, OM1-75, and OM2-50 in the second year and as OM2-25, OM2-50, OM2-75, and OM3-50 in the third year, respectively.
The seeding rates of cover crops were 56 kg ha-1 for sunn hemp, 112 kg ha-1 for cowpea, 45 kg ha-1 for both velvetbean and sorghum sudangrass. In order to promote branching and increase the production of high quality biomass, the apical dominance of the sunn hemp was overcome by cutting the stems at 30 cm above the ground in the middle of August (Abdul-Baki et al. 2001.). The cover crops were terminated by flail-mowing, incorporated into the soil in mid September each year. Early in October, 1120 kg ha-1 of dry fertilizer (6 N-6 P2O5-12 K2O) was rototilled into the soil together with the cover crop residues. Then the raised beds 15-cm high, 91-cm wide and 182-cm between centers of two adjacent beds were re-formed. Compost was supplied to the appropriate plots by means of tractor-powered calibrated spreader. Each plot consisted of a 15-m long section of a raised bed, which was 27 m2. Two parallel T-tapes, each 18 cm from the center of the bed were installed and connected to a layflat hose, which provided the water for drip irrigation. Tomato (Lycopersicon esculentum (Mill.)) ‘Sanibel’ in 2002, and ‘Florida 47’ in 2003 and 2004, respectively. Seedlings were transplanted into in a single row on each bed with 50-cm between plants. Practices to control foliar insects and diseases were applied according to Maynard and Olson (2000). Each year starting from the sixth week after tomato transplanting, 328 l ha -1 of liquid fertilizer consisting of 4 N-0 P2O5-8 K2O was delivered through the drip lines once per week until the second harvest.

Harvests and grading
Tomato fruits were harvested three times each year. The first harvest was conducted when about 10% of fruits had turned red, and only red or pink, and extra large green and large green fruits (“mature green” fruits) were harvested during this time. Two or three weeks later, the second harvest was similarly performed. Finally 3 or 4 weeks after the second harvest, the third harvest was conducted, which included all fruits equal or larger than the small size. Harvested fruits were separated by machine (Kerian speed sizer, Kerian Machine, Inc., Grafton, ND) according to the Florida Tomato Committee Standards (Brown, 2000) into extra-large, large, and medium and small sizes. Defective fruits were culled manually. The fruits in each size category as well as the culls were weighed. Extra-large, large, and medium fruits constituted the marketable yield, and the small and culled fruits constituted the non-marketable yield.

Sampling and analysis
Soil samples (0-10 cm in depth) were collected before the cover crops were seeded, after the cover crops had been soil-incorporated but before tomato seedlings were transplanted and at tomato flowering. Above ground biomass samples of the cover crops were collected each year before they were terminated by flail-mowing. At the flowering, the height and canopy diameter of each of 5 randomly selected tomato plants from each plot were measured and the canopy area was calculated. The chlorophyll content of the third fully expanded leaf from the top of each of these same tomato plants was determined with a SPAD-502 meter, and then this leaf was collected for further analysis.
Soil samples were air dried and ground to pass through a < 2 mm mesh sieve. Plant biomass samples were oven dried at 70 oC for >72 hr before being ground to pass through a <1 mm mesh sieve for chemical analysis. Total N of soil and plant samples was determined with a CNS auto-analyzer (Vario Max Elementar, Hanau, Germany). Soil organic carbon was determined by the weight loss-on-ignition (WLOI) method (Schulte and Hopkins, 1996; Jolivet et al., 1998). In order to determine the total nutrient elements other than N, the samples were digested with concentrated nitric acid-hydrogen peroxide-hydrochloric acid (HNO3-H2O2-HCl) according to the USEPA method, 3050A (USEPA, 1990), then analyzed via inductively coupled plasma optical emission spectroscopy [ICP-OES (Ultima 2C) Horiba Jobin Yvon Inc., Edision, N.J.]. Soil sub-samples were extracted with AB-DTPA for colorimetric determination of extractable P, and for ICP-OES determination of extractable K. The composition and properties of compost used in the experiment were determined with the methods used for the soil samples, and the data were summarized in Table 1. At the end of the third year, soil physical properties, e.g., soil bulk density (Blake and Hartge, 1986), field capacity and hydraulic conductivity (Jury and Horton, 2004), etc. were determined. Statistics
By means of the SAS program (SAS Institute, version 8.1, Cary, N.C.), the data were subjected to ANOVA and the means were separated using Duncan’s multiple range test.

No. 3. Automated irrigation/fertigation system.
Development of the soil water controller
A quantified irrigation control (QIC) system was developed around a custom-built integrated circuit (IC) board and a commercially available capacitance based soil water probe (0.20 m ECH20 probe, Decagon Devices, Inc., Pullman, WA). Most materials are readily available electronic components except for the custom-built IC board. The soil water probe used in this research could be replaced with any type of sensor that has a predictable voltage response to variation in soil water or tension. The total cost of the controller components without the sensor, not including shipping of components or labor for assembly was US$64 (US$124 with the sensor). This compares favorably with existing commercial alternatives including switching tensiometers and other dielectric probe based controllers.
The programmable microcontroller used in this system (MSP430-FETI49, Texas Instruments Inc., Dallas, Texas) contained a 16 bit timer with a 12 bit analog to digital (A/D) converter and was developed for ultra low power applications (0.1-250 A). The 12 bit A/D converter allows the 250 (dry) - 1000 (wet) mY signal sent from the probe to be resolved to 0.18 mY, which translates into <0.01 % volumetric soil water in South Florida soils.
The QIC microcontroller queries the probe at user set intervals by sending a 25 millisecond 2.5 YDC excitation signal to the capacitance probe. Normally, this comparison occurs every minute (adjustable via a software interface with the microcontroller). If the voltage returned from the probe is below a user set threshold (potentiometer), then the controller allows the time clock 24 YAC signal to power the irrigation solenoid valve via the on board latching relay. When the signal is higher than the set point the relay opens to stop irrigation. Since several 1 minute sampling cycles are possible within an irrigation event, it is possible for the system to irrigate for periods shorter that the duration of the scheduled event, i.e. there is no minimum irrigation time per se, other than the querying and sensor response lag time.
The QIC can be powered by either a 9V DC battery or by the power from the controller (24 V AC when the time clock sends a signal), both of which are transformed to 5V DC. The advantage of using a battery is that the voltage of the soil water probe can be checked while in the field without need for the controller to power the irrigation zone containing the QIC in question. The QIC can also be used in the place of a time based irrigation controller for a single irrigation valve and a 24V AC power supply. Although this configuration was not used in the experiment, it would allow for complete irrigation control based on soil water conditions. By connecting the on-board jumper the QIC will query the probe every four seconds to determine the probe output directly from the IC board.
The potentiometer can be adjusted until the LED activates, which establishes the QIC set point. This provides the alternative to select the set point in the field when the soil water level is at optimal conditions such as field capacity without the need for a specific soil calibration.

Site selection and probe calibration
The first experiment was conducted at the Tropical Research and Education Center in Homestead, FL on a Krome gravelly-loam "rock-plowed" soil (loamy-skeletal, carbonatic, hyperthermic Lithic Udorthents) of 0.30 m average depth (with a range of 0.10-0.40 m) overlaying porous limestone bed rock, (USDA, 1996). The physical properties at the site are summarized in Munoz-Carpena et al., (2002).
Although the manufacturer of the soil water probe used here provides a general linear calibration equation (Decagon Devices, 2002), a specific calibration was developed for soil from our test field. The soil was sampled at three different field locations from a depth of 0 to 0.21 m with care to obtain the gravel and fine portions as present in the field. After collection, each sample was homogenized independently by tumbling in a bucket with a lid for 15 seconds, and then hand-packed in 3 PYC cylinders (0.10 x LO.21 m) bounded at the bottom with a stainless steel fine wire mesh held by a metal bracket around the tube. To achieve the original field bulk density the packing was done in four layers of equal thickness. A capacitance probe was inserted vertically in the center of each core and the samples saturated from the bottom up during 24 hr using a solution of 0.005 M CaS04 saturated with thymol (Klute and Dirksen, 1986). Once saturated the probe outputs (mY) were read with a handheld reader (ECH2O Check, Decagon Devices, Inc., Pullman, WA) and the cores weighted on a laboratory scale of 0.000 I kg of resolution over the 0-8.0 kg range. The cores were then placed on a wire screen to allow free drainage and air-drying while frequent probe readings (mY) and weights (kg) were recorded. When the volumetric soil water reached 17%, value below those typical of irrigated field conditions, the probe was removed and the soil dried in a laboratory oven to obtain the dry weight needed to calculate the volumetric water content for each reading. A total of 25 paired probe readings and volumetric water content data collected from the three cores were fitted to a straight line. Thus, an average calibration over the entire field was applied to all probes use for irrigation control rather than a specific calibration of each probe location.

Field testing of the controller
A field at the University of Florida's Tropical Research and Education Center in Homestead, FL in which sorghum sudangrass had been grown as a summer cover crop was utilized for this experiment. Tomatoes were cultured according to local horticultural practices. A fumigant (66:33 volumetric mix of methyl-bromide:chloropicrin, MC-33) was injected into the soil at 392 kg/ha during the formation of the raised beds, and immediately thereafter the drip lines and plastic mulch were installed. Pre-plant dry fertilizer (6-6-12) at 1867 kg/ ha was roto-tilled into the bed. The tomato seedlings (cultivar 'FL 47') were transplanted on November 20, 2003 into plastic mulched raised beds spaced 1.8 m apart in one row per bed with plants spaced 0.46 m apart. Each plot was 16.7 m long. Irrigation was supplied with dual drip irrigation lines (T-TAPE TSX 508-12-450, T-Systems International, Inc., San Diego, CA with 0.015 m internal diameter, 0.30 m emitter spacing, 1.0 Uh emitter discharge at 69 kPa, and 0.002 m thickness) under the plastic mulch and approximately 0.30 m apart on either side of the tomato row. Dissolved fertilizer (4-0-8) at 19.6 kg N/ha was applied manually to each individual plant only during each of the final five weeks prior to harvest. Tomatoes were harvested four times during the period March 1-18, 2004 from a 6.1 m section within each plot. Harvested fruits were graded following Florida Tomato Committee standards (Brown, 2000). Irrigation treatments were implemented at 20 days after planting to allow the transplants time to become established. Prior to that date, all experimental treatments were irrigated at least once each day on the same schedule. Thereafter irrigation treatments were established in a completely randomized design with four replications. Irrigation scheduling mechanisms consisted of switching-tensiometer, historical weather, and practices used by local growers.. The tensiometer, QIC, and weather-based methods were set to irrigate a maximum of five times each day for one hour total. The grower-based treatment was irrigated once each day for one hour similar to practices in the region. Tensiometer and QIC methods allowed irrigation only if soil tension exceeded set points for tensiometer treatments, or if soil water was below set points for QIC treatments, respectively. The tensiometer treatments consisted of switching tensiometers (low tension model TGA-LT, Irrometer Co., Riverside, CA) set at 10 kPa and 25 kPa for the two treatments, respectively. The QIC treatments were set at two thresholds of 425 mY (95% confidence interval of 420-430 mY) and 450 mY (95% confidence interval of 445-455 mY), corresponding to the soil water status at 25 kPa and 10 kPa obtained using the soil water release curve given by Al-Yahyai et al., (2006) for the gravelly-loam soil of this site. This voltage threshold was used previously for the gravelly-loam soil (Munoz-Carpena et al., 2005a) and verified at the beginning of the experiment. Weather-based treatments were irrigated according to calculated crop evapotranspiration (ETc) that was calculated by the local historical daily average reference ET (ETo) multiplied by the published crop coefficient (Ke). For our crop and area the Kc values used were: 0.3 (from 11/20-12/3), 0.6 (12/4-12/23), 1.15 (12/24-2/12) and 1.00 (2/13-3/18) (Simonne et al., 2004). Historical average seasonal ETc in this area is 291 mm (Simonne et al., 2004). The historical weather-based treatment was irrigated according to the average long-term maximum daily ETo throughout the season. The grower practice irrigation treatment consisted of one hour of irrigation per day throughout the season (3.7 mm/d).
Water use in each plot was continuously and independently recorded by a positive displacement
water meter equipped with a magnetically actuated reed switch (PSM- T 0.016 X 0.013 m (5/8" X
1/2"), ABB Water Meters, Inc., Ocala, FL) connected to an event data logger (H7-002-04, Onset
Computer Corporation, Bourne, MA). Weekly readings were also manually taken from the counters in each water meter. The water meters were installed at the inlet of each plot upstream of the pressure regulator and a solenoid valve.
For each treatment irrigation water use efficiency (IWUE, kg/m2) was calculated by dividing the sum of the marketable yield by the total amount of water applied during the season. One-way analysis of variance and comparison of means using Tukey Kramer HSD were performed on irrigation water, yield and IWUE (JMP 6, SAS Institute, Cary, NC, 2005). This test controls the type 1 errors of all comparisons simultaneously, rather than a pair of means at a time.
Weather parameters such as temperature, relative humidity, incoming solar radiation, wind speed, and precipitation were measured on-site by the Florida Automated Weather Network (FAWN) system. Daily ET was calculated by the modified Penman method as described in Jones et al., (1984).

Field installation and operation of the controller
Switching tensiometers and ECH2O probes were installed vertically 3 m away from the solenoid valves in the center of the bed, between the paired irrigation lines and two consecutive tomato plants of each experimental plot. The ECH2O dielectric probes were inserted in the top 0.20 m, roughly equivalent to the total bed depth, whereas the porous cup of the switching
tensiometers was placed at the midpoint of the bed depth, i.e. 0.10111. Thus an average soil water status was obtained for entire bedded soil profile of the tomato plants with both types of probes. The QIC was placed nearby on top of the vegetable bed. Although the QIC was designed within a waterproof metal housing, the entire apparatus was placed within a plastic food storage container. Both the QIC and plastic container contained a desiccant to prevent
condensation on the inside of the containers. For the first week of QIC operation, readings were
collected from manual tensiometers buried 0.10 m deep near the QICs. Using these readings the set point of the QIC potentiometer was verified to match the target set points of 10 and 25 kPa. It is important to note that QICs were checked weekly for proper operation but the set-points were not adjusted for the remainder of the season.
An independent set of dielectric probes (0.20 m ECH20 probe, Decagon Devices, Inc., Pullman, WA) connected to individual dataloggers (HOBO H08-006-04, Onset Computer Corporation, Pocasset, MA) was installed vertically in the top 0.20 m next to the switching tensiometers and QIC probes. Soil water content was recorded hourly from these 16 probes and average seasonal values for each treatment were calculated.

No. 4. Influence of various crops in rotation on population of soil nematodes.
Experimental conditions
The pot experiment was conducted in a screen house from June 2002 to February 2003 at Tropical Research and Education Center, University of Florida, Homestead, Florida, which is in a subtropical region. The average annual rainfall is 1,499 mm of which 76% falls between June and October; the annual temperature averages 23.9 oC and usually ranges from a maximum of 35 oC in June to a minimum of 5 oC in January (Wang et al., 2002). Between June 2002 and February 2003, the air temperature ranged from a minimum of 2.2 oC to an average monthly maximum of 34.6 oC; the monthly average maximum temperature was 35.3 oC in August 2002 but only 28.5 oC in January 2003; and the relative humidity ranged between 75% in January to 84% in June (Wang et al., 2006).
The soil (loamy-skeletal, carbonatic, hyperthermic Lithic Udorthents) contained 58.8% gravel (> 2mm), and the non-gravel fraction had a distribution of soil particles of 48.4% sand, 30.3% silt and 21.3% clay. The soil also consisted of 60% calcium carbonate (CaCO3), pH 7.8, soil organic C 28 g kg-1, total N 1.1 g kg-1, and ammonium bicarbonate-diethylene triaminepentaacetic acid (AB-DTPA) extractable phosphorus (P) 22.7 mg kg-1 and potassium (K) 129 mg kg-1, respectively.

Experimental design and management
A complete randomized block design with 8 different rotation schemes of the given crops was adopted. The 8 crops were marigold I (MI), marigold II (MII), radish, Raphanus sativus L. var. sativus (RD), Indian mustard, Brassica juncea (L.) (IM), sunn hemp (SH), cowpea [Vigna unguiculata (L.)] (CP), velvetbean (VB) and okra (OK). There were three rotations in each scheme, i.e. MI-RD-MI, MII-CP-MII, IM-VB-IM, RD-MI-RD, SH-OK-SH, VB-IM-VB, CP-MII-CP, and OK-SH-OK, respectively. Each treatment was replicated 3 times. Before the experiment started and after each rotation, a 250 ml soil sample was collected from each pot to identify and count the nematodes.
Prior to the experiment, tomato roots replete with root galls caused by the root-knot nematode were collected from a commercial farm, cut into small pieces (< 1 cm) and this inoculum was blended into the potting soil. A very nematode-susceptible plant, okra was seeded then and grown for 4 weeks to increase the root-knot nematode population since the initial population of root-knot nematodes in this soil were fairly low (Wang et al., 2003). After 4 weeks the okra plants were harvested and their roots replete with root galls produced by the female root-knot nematode, Meloidogyne incognita, were cut into pieces 1-cm long and evenly distributed into the soil. The soil was mixed well and 3 kg of it was placed into each pot, which was 6 cm in diameter and 15 cm high with a capacity of 1.5 liter. The soil was sieved to remove large gravel (> 1 cm). The crops were first seeded on June 6, 2002 to produce 3 plants per pot for velvetbean, 5 for cowpea, radish, Indian mustard, and 10 for the others. Drip irrigation was installed and adjusted to deliver 2 L water per hr. and a clock timer was used to control automatically the irrigation duration and frequency based on the plant growth stages. Each succeeding rotation was started immediately after the roots had been rated and the roots, stems and leaves cut into 2-cm long pieces and incorporated into the soil in same pot used for the previous rotation. Each rotation lasted about 3 months.

Sampling and analysis
Roots were washed free of soil and examined for galling and root knot infection. Some roots showing lesions were cut into small pieces and left in water for 36 to 48 hr to reveal the presence of any lesion and other nematodes. Root galling was rated after each rotation on a 0-5 scale (Taylor et al., 1978). Each 250-ml soil sample was collected from 3 to 12 cm below the surface in each pot. Nematodes from each sample were extracted by means of Cobb’s sieving and decanting technique (Cobb, 1918), followed by a modified Baermann funnel method (Hopper, 1986). Sieves used in nematode extraction were U.S. Standard Sieve Series of 100, 200, and 325 mesh with openings of 149, 74, and 44 m, respectively. Nematodes were fixed in hot 30 ml/L formaldehyde solution, identified to genus level, and counted using a stereoscope. Some fixed specimens were processed with anhydrous glycerin (Seinhorst, 1959), and examined under a compound microscope for species identification. Nematode identifications were based on the morphology of adult and larval forms and their identities were confirmed with recent taxonomic keys (Handoo et al., 1992; Mai et al., 1996; Maqbool, 1982; Robinson et al., 1997; Sher, 1996). Nematode density (number in 250 ml of soil) was determined for each species and recorded.

Data analysis
The data were subject to the analysis of variance (ANOVA) and Duncan’s multiple range test with a general linear model (GLM) for significant differences using SAS (version 8.1, SAS Inst. Inc., Cary, NC, USA).

No. 5. Influence of cover crops in rotation on improving okra (Abelmoschus esculentus L.) yield and suppressing parasitic nematodes.
Site description
The experiment was conducted at the Tropical Research and Education Center, University of Florida, Homestead, Florida. At this site the average annual rainfall is 1,499 mm of which an average 76% falls between June and October, the annual temperature averages 23.9 oC and usually ranges from a maximum of 35 oC in June to a minimum of 5 oC in January (Wang et al., 2002a). The soil is a Krome very gravelly loam (loamy-skeletal, carbonatic, hyperthermic Lithic Udorthents), which consist of 58.8% gravel (> 2 mm), and a gravel-free fraction. The latter consists of 48.4% sand, 30.3% silt, and 21.3% clay. CaCO3 comprises about 60% of the soil, organic C is 11.3 to 18.0 g kg-1, total N is 0.9 to 1.3 g kg-1, the soil pH is 7.6 to 7.8 (water), ABDTPA (ammonium bicarbonate-diethylene triaminepentaacetic acid) extractable phosphorus (P) and potassium (K) are 16.5 to 26.8 mg kg-1 and 123.5 to 132.6 mg kg-1, respectively.

Experimental design and management
The field experiment was conducted in three consecutive years from 2002 through 2005 using a randomized split-plot design. The main plots consisted of four cover crops, sunn hemp, velvetbean, cowpea and sorghum sudangrass, grown from June to September, and fallow as a control, and all plots were on raised beds. In the first year compost was applied as an organic mulch (OM) to some plots at the single rate of 50 t ha-1 dry weight, while white on black plastic mulch (PM) was installed on others. In the second and third years, the 50 t ha-1 treatment was applied to the very same plots as in the first year of the experiment, and three additional compost treatments were made, i.e., 25 t ha-1, 50 t ha-1, and 75 t ha-1. Thus the four organic mulch treatments are denoted as OM1-25, OM1-50, OM1-75, and OM2-50 in the second year and as OM2-25, OM2-50, OM2-75, and OM3-50 in the third year, respectively.
The seeding rates of cover crops were 56 kg ha-1 for sunn hemp, 112 kg ha-1 for cowpea, 45 kg ha-1 for both velvetbean and sorghum sudangrass. In order to promote branching and increase the production of high quality biomass, the apical dominance of the sunn hemp was overcome by cutting the stems at 30 cm above the ground in the middle of August (Abdul-Baki et al. 2001.). The cover crops were terminated by flail-mowing, incorporated into the soil in mid September each year. Early in October, 1120 kg ha-1 of dry fertilizer (6 N-6 P2O5-12 K2O) was rototilled into the soil together with the cover crop residues. Then the raised beds 15-cm high, 91-cm wide and 182-cm between centers of two adjacent beds were re-formed. Compost was supplied to the appropriate plots by means of tractor-powered calibrated spreader. Each plot consisted of a 15-m long section of a raised bed, which was 27 m2. Two parallel T-tapes, each 18 cm from the center of the bed were installed and connected to a layflat hose, which provided the water for drip irrigation. Tomato (Lycopersicon esculentum (Mill.)) ‘Sanibel’ in 2002, and ‘Florida 47’ in 2003 and 2004, respectively. Seedlings were transplanted into in a single row on each bed with 50-cm between plants. Practices to control foliar insects and diseases were applied according to Maynard and Olson (2000). Each year starting from the sixth week after tomato transplanting, 328 l ha -1 of liquid fertilizer consisting of 4 N-0 P2O5-8 K2O was delivered through the drip lines once per week until the second harvest.

Harvests and grading
Tomato fruits were harvested three times each year. The first harvest was conducted when about 10% of fruits had turned red, and only red or pink, and extra large green and large green fruits (“mature green” fruits) were harvested during this time. Two or three weeks later, the second harvest was similarly performed. Finally 3 or 4 weeks after the second harvest, the third harvest was conducted, which included all fruits equal or larger than the small size. Harvested fruits were separated by machine (Kerian speed sizer, Kerian Machine, Inc., Grafton, ND) according to the Florida Tomato Committee Standards (Brown, 2000) into extra-large, large, and medium and small sizes. Defective fruits were culled manually. The fruits in each size category as well as the culls were weighed. Extra-large, large, and medium fruits constituted the marketable yield, and the small and culled fruits constituted the non-marketable yield.

Sampling and analysis
Soil samples (0-10 cm in depth) were collected before the cover crops were seeded, after the cover crops had been soil-incorporated but before tomato seedlings were transplanted and at tomato flowering. Above ground biomass samples of the cover crops were collected each year before they were terminated by flail-mowing. At the flowering, the height and canopy diameter of each of 5 randomly selected tomato plants from each plot were measured and the canopy area was calculated. The chlorophyll content of the third fully expanded leaf from the top of each of these same tomato plants was determined with a SPAD-502 meter, and then this leaf was collected for further analysis.
Soil samples were air dried and ground to pass through a < 2 mm mesh sieve. Plant biomass samples were oven dried at 70 oC for >72 hr before being ground to pass through a <1 mm mesh sieve for chemical analysis. Total N of soil and plant samples was determined with a CNS auto-analyzer (Vario Max Elementar, Hanau, Germany). Soil organic carbon was determined by the weight loss-on-ignition (WLOI) method (Schulte and Hopkins, 1996; Jolivet et al., 1998). In order to determine the total nutrient elements other than N, the samples were digested with concentrated nitric acid-hydrogen peroxide-hydrochloric acid (HNO3-H2O2-HCl) according to the USEPA method, 3050A (USEPA, 1990), then analyzed via inductively coupled plasma optical emission spectroscopy [ICP-OES (Ultima 2C) Horiba Jobin Yvon Inc., Edision, N.J.]. Soil sub-samples were extracted with AB-DTPA for colorimetric determination of extractable P, and for ICP-OES determination of extractable K. The composition and properties of compost used in the experiment were determined with the methods used for the soil samples, and the data were summarized in Table 1. At the end of the third year, soil physical properties, e.g., soil bulk density (Blake and Hartge, 1986), field capacity and hydraulic conductivity (Jury and Horton, 2004), etc. were determined. Statistics
By means of the SAS program (SAS Institute, version 8.1, Cary, N.C.), the data were subjected to ANOVA and the means were separated using Duncan’s multiple range test.

No. 6. Influence of cover crops and soil amendments on okra production and soil nematodes.
Pot size and soil properties
The growth media were placed in 8.3-L plastic pots each 23 cm in diameter and 20 cm high with a capacity of 8 kg of soil per pot. The soil was Krome very gravelly loam (loamy-skeletal, carbonatic, hyperthermic Lithic Udorthents) collected from a field in the Tropical Research and Education Center. The soil contained 58.8% gravel (>2mm), had a distribution of soil particles of 48.4% sand, 30.3% silt and 21.3% clay. Also the soil contained 60% CaCO3, pH 7.8 (water), soil organic C 28 g kg-1, total N 1.1 g kg-1, and AB-DTPA extractable P 22.7 and K 129 mg kg-1, respectively. Since the densities of Meloidogyne incognita root-knot nematode population in fields of the Tropical Research and Education Center were known to be quite low (Mannion et al. 1994), roots of tomatoes with galls were collected from several farms in Miami-Dade County, and pieces of these roots were added to the soil in the pots.

Cover crops and organic soil amendments
Summer cover crops, sunn hemp (Crotalaria juncea L.), and sorghum sudangrass [Sorghum bicolor × S. bicolor var. sudanense (Piper) Stapf.] were grown to compare their effects on okra yield and nematode population densities with those of fallow as a control. Organic amendments incorporated into the soil were biosolids, N-viro soil (1:1 mixture of coal ash and biosolids), coal ash, co-compost (3:7 mixture of biosolids and yard wastes), yard waste compost (YW-compost), control (cover crop grown and incorporated into the soil, but without any other organic amendment), cover crop removal (the cover crop was grown until the flowering stage when the aerial parts and most of the roots were removed and discarded), and cover crop + MC-33 (the cover crop was grown and incorporated into the soil, and subsequently fumigated with a mixture of 33% of methyl bromide and 67% of chloropicrin). The amount of each organic amendment added to the soil in each pot was 205 g of dry weight equivalent; which was equivalent to 50 Mg ha-1. After the cover crops residues and/or organic amendments had been incorporated into the soil, the amended soil was allowed to equilibrate for 10 days before the okra was seeded. The composition and properties of the various organic amendments were determined and are summarized in Table 1. More detailed information can be found in Wang et al. (2003b). The experiment was carried out from June 2002 to February 2003 as the first year (Year-1), and June 2003 to February 2004 as the second year (Year-2).

Experimental design and management
In order to determine the effectiveness of sorghum sudangrass, and sunn hemp relative to fallow in suppressing various taxa of nematodes an experiment was conducted using a randomized complete block design with three replicates. Subsequently to determine the relative efficacy of combinations of cover crops and various inanimate organic amendments a split plot design with 3 replicates was employed in which the two cover crops were the main plots, and the organic amendment were subplots. In order to keep the size of the experiment manageable, fallow was not included as a main plot. When the cover crop had developed to the onset of flowering, the aerial parts were cut into pieces (< 2 cm) and then thoroughly mixed into the soil from the pot together with the appropriate organic amendment if any. This mixture was returned into the same pot. In the MC-33 treatment, the pots in which the cover crop had been soil incorporated were fumigated by placing them into a chamber and using liquid N to introduce the appropriate amount of the 33% methyl bromide-67% chloropicrin mixture. The chamber was kept hermetically sealed for 72 hours. Thereafter the pots were exposed to the atmosphere for an entire week, and then okra (Abelmoschus esculentus L.) was seeded in each pot. The okra planting was thinned at the 3-leaf stage to 6 plants per pot. Drip irrigation was installed and adjusted to deliver 2-L hr--1 of water. A clock timer was used to control the irrigation automatically. The period and frequency of irrigation were adjusted according to needs of the successive plant growth stages. Okra fruits were harvested when they attained full development. Sampling and analysis
Each cover crop was sampled for biomass determination at the time of termination. Okra fruits were harvested to obtain the fruit yield. The okra yields and cover crop biomass weights were converted to metric tons per hectare (Mg ha-1) equivalent according to the surface area of the pot. For determining the identities and densities of the various nematode taxa, soil samples were collected before the cover crops were seeded and after their termination, and also after the okra harvest. A soil sample of 250 ml was taken from each pot. Nematodes from each sample were extracted by means of Cobb's sieving and decanting technique (Cobb, 1918), and the modified Baermann funnel method (Hooper, 1986). Sieves used in nematode extraction were U.S. Standard Sieve Series of 100-, 200-, and 325-mesh with openings of 149, 74, and 44 m, respectively. Nematodes were fixed in hot 30ml•L-1 formaldehyde solution, identified to genus level, and counted using a stereoscope. Some fixed specimens were processed with anhydrous glycerin (Seinhorst, 1959), and examined under a compound microscope for species identification. Nematode identifications were based on the morphology of adult and larval forms and their identities were confirmed with recent taxonomic keys (Handoo and Golden, 1992; Mai et al., 1996; Sher, 1966; Handoo, 2000). Roots were washed free of soil and examined for galling and root knot infection. Some roots showing lesions were cut into small pieces and left in water for 36 to 48 h to reveal the presence of any lesion and other nematodes. Nematode density (number in 250 ml of soil) was determined for each species and recorded. The okra plants were individually rated for root galls and egg masses on a 0-5 scale: 0 = 0 galls, 1 = 1-2, 2 = 3-10, 3 = 11-33, 4 = 31-100, 5> 100 galls or egg masses (Taylor and Sasser, 1978). The data were subjected to the analysis of variance (ANOVA) and Duncan’s multiple range tests for a significant difference by means of SAS (version 8.1, SAS Inst. Inc., Cary, NC, USA).

No. 7. Summer Cover Crops and Soil Amendments to Improve Growth and Nutrient Uptake of Okra.
Soil properties and experimental design.
The experiment was conducted in pots of 8.3-L each, 23 cm in diameter and 20 cm high with a soil capacity of 8 kg/pot. The soil was a Krome very gravelly loam (loamy-skeletal, carbonatic, hyperthermic Lithic Udorthents) collected from a field in the Tropical Research and Education Center, University of Florida, Homestead, FL. The soil contained 58.8% gravel (>2 mm sieve), had a distribution of soil particles of 48.4% sand, 30.3% silt and 21.3% clay. Also the soil contained 60% calcium carbonate (CaCO3), pH 7.8 (water), soil organic C 28 g•kg-1, total N 1.1 g•kg-1, and ammonium bicarbonate-diethylene triaminepentaacetic acid (AB-DTPA) extractable phosphorus (P) 22.7 and potassium (K) 129 mg•kg-1, respectively.
A split plot design was used in which the main plots were four summer cover crops and fallow, and the subplots were five organic soil amendments along with a no amendment control. When the cover crop, sunn hemp, reached the optimum stage for termination (flowering), the stems of all the cover crops were cut off at ground level and cut into pieces about 2 cm long and thoroughly mixed into the soil together with the appropriate organic amendment if any. After these soil amendments were applied, the materials in the pots were allowed to equilibrate for 10 d before the okra was seeded. The experiment was carried out in a screen house from Jun. 2002 to Feb. 2003 of the first year (2002-03), and Jun. 2003 to Feb. 2004 of the second year (2003-04).
The main treatments were summer cover crops, sunn hemp, cowpea, velvetbean, and sorghum sudangrass and a weedy natural fallow as the control. The subplot treatments were soil amendments: biosolids, N-Viro Soil (N-Viro International Corporation, Toledo, Ohio), coal ash (a byproduct from power plants in Fla.), co-compost (3:7 of biosolids to yard wastes, West Palm Beach, Fla.), yard waste compost (mainly from leaves and branches of trees and shrubs, and grass clippings), and a control (soil-incorporated cover crop residues but without any organic amendment). Each soil amendment was applied at 205 g/pot of dry weight equivalent, which was equivalent to 50 Mg•ha-1. The composition and properties of the various organic amendments were determined and are summarized in Table 1.
Okra seeds were sown on Aug. 20 and 25 of years 2002 and 2003, respectively, and thinned at the three-leaf stage to six plants per pot. Drip irrigation was installed and adjusted to deliver 2 L•h-1 of water. A clock timer was used to control the irrigation automatically. The period and frequency of irrigation were adjusted according to needs of the successive plant growth stages. Okra fruits were harvested twice each week for seven weeks started on Jan. 5, 2003, and Jan. 10, 2004, respectively. At the end of the experiment, to obtain the total biomass, the remaining fruits, shoots and roots washing free of soil were harvested. Plants were separated into fruits, stems and roots, and random samples of these plant parts were taken for chemical analysis to determine concentrations of B, C, Cd, Cu, Fe, K, Mg, manganese (Mn), Mo, N, P and Pb. Carbon and N were analyzed with a carbon-nitrogen-sulfur (CNS) Auto-analyzer (vario Max Elementar, Hanau, Germany), and all the other elements were analyzed via inductively coupled plasma optical emission spectroscopy (ICP-OES, Ultima 2C, Horiba Jobin Yvon Inc., Edison, N.J.) after the samples had been digested with concentrated nitric acid-hydrogen peroxide-hydrochloric acid (HNO3 – H2O2 – HCl) according to the U.S. EPA Method 3050A (1990).
After the okra plants were harvested and removed, sorghum sudangrass was grown in the same pots to study the residual effects of the cover crops and soil amendments. Twenty five sorghum sudangrass seedlings were grown in each pot and irrigated as above. The sorghum sudangrass was grown from Nov., 2002 to Jan., 2003, then removed from the pots, dried at 70 oC for 72 h and weighed to obtain the aboveground dry biomass.

Sampling and analysis.
The cover crops were sampled for biomass determination at the time of their termination. Also a representative okra plant from each pot was sampled at the visible bud (pre-flowering) stage to determine the dry weight of its biomass. Okra fruits, shoots and roots (washed free of soil) were sampled. Soil samples were collected before and after growing the cover crops and also after the okra harvest.
The plant samples were dried at 70 oC for 72 h and ground to pass through a 1-mm sieve. Soil samples were air-dried, plant materials were removed, and the samples were gently ground by hand to pass through a 2-mm sieve. Soil samples were extracted with AB-DTPA for P, K, Ca, Mg, and micro-nutrient element analysis via ICP-OES. Total N and C in both soil and plant samples were determined via a CNS Auto-analyzer. Soil organic carbon was determined by the weight loss-on-ignition (WLOI) method (Schulte and Hopkins 1996).
The data were subjected to the analysis of variance (ANOVA) and Duncan’s multiple range tests for significant differences using SAS (version 8.1, SAS Inst. Inc., Cary, NC, USA).

No. 8. Conservation of soil and nutrients by cover crops under high and low rates of simulated rainfall.
Cover crops chosen and soil properties
Three legume summer cover crops, sunn hemp (Crotalaria juncea), cowpea (Vigna unguiculata), velvetbean (Mucuna deeringiana) and a non-legume, sorghum sudangrass (Sorghum bicolor × S. bicolor var. ‘sudanense’) were cultivated in pots with Krome very gravelly loam soil (loamy-skeletal, carbonatic, hyperthermic Lithic Udorthents) in a screenhouse from Jul. through Oct., 2002. The soil contained 29.2g kg-1 of organic C, 982.3 mg kg-1 of total N, 46.8 mg kg-1 of AB-DTPA extractable P and had a pH (KCl) of 8.2 and a CaCO3 content of 682 g kg-1. The soil contained 58.8% of gravel (>2 mm) and particle distributions in the gravel free soil were 40.8% sand, 39.7% silt, and 19.5% clay. Each pot was 18 cm in diameter, 30 cm deep and had a capacity of 8 kg soil. The soil was thoroughly mixed and then placed into each pot. Plant density per pot was as follows: 30 sunn hemp plants, 5 sorghum sudangrass plants, 5 velvetbean plants or 5 cowpea plants. Each treatment was replicated four times.

Water supply and leachate collection
Sprinkler heads with a tubing system to deliver water at 4 L hr-1 (high rate) and 2 L hr-1 (low rate), equivalent to average rainfall amounts of 5.8 mm week-1 and 11.6 mm week-1, respectively, were installed. Graduated beakers were kept among the pots for monitoring the amount of water actually delivered, and hence, facilitate efforts to ensure uniform delivery to each pot. From the third week after cover crops had been sown, a plastic saucer was placed under each pot to receive the leachate (gravitational water) and, a PVC cylinder, 10 cm in diameter and 5 cm high, was placed between the pot and the saucer to prevent the percolation from soaking the soil at the bottom of the pot. The amount of water and frequency of application were adjusted according to plant growth periods, and the system was automatically controlled with a clock timer (Hose Faucet, model 62001/62401) installed on a water tap. The amounts of water supplied and leachate collected were recorded daily. Leachate samples were collected weekly for chemical analysis. To enable the sorghum sudangrass to establish but to keep each cover crop uniform in all treatments, fertilizer (10 N-10 P2O5-10 K2O) at a rate equivalent to 771 kg ha-1 was applied to each pot at the end of the first week of leachate collection. The fertilizer contained 2.0% NO3-N, 6.0% NH4-N, 1.8% other /water soluble N or urea N and 0.2% water insoluble N, 10% available phosphate (P2P5), 10% soluble potash (K2O) and some microelements.

Sampling and analysis
Soil samples were taken before the experiment, ground to pass < 2 mm sieve for chemical analysis. The stems of all the cover crop plants were cut off at soil surface when the sunn hemp reached its blossom stage. The roots were washed free of soil, and the stems were separated from the leaves. The plant parts were dried at 70 oC for 72 hours to obtain the dry weight of the biomass. The leachates were filtered before the chemical analyses were carried out using a colorimetric method via an Auto-Analyzer-3 (BRAN-LUEBBE, Germany). The detection limits for the equipment were 0.02 mg l-1 for NO3-N, 0.01 mg l-1 for NH4-N, and 0.002 mg l-1 for inorganic P, respectively. Soil and plant N were analyzed by means of CNS Auto-analyzer (vario Max Elementar, Hanau, Germany).
To investigate the movement of soil particles and their redistribution as influenced by various cover crops under low- and high-rates of simulated rainfall, an intact core column sample was collected from the center of each pot after the harvest of the cover crops. Each core column was sliced into 0-10 cm, 11-20 cm, and 21-30 cm segments, respectively. The soil samples were air dried and the particle distribution analysis was conducted using the micro-pipette approach (Miller and Miller, 1987). The same procedures were conducted on the soil mixture initially used to fill the pots.

Statistics
The data were subjected to the analysis of variance (ANOVA) and Duncan’s multiple range tests via SAS program version 8.11, SAS Inst. Inc., Cary, NC, USA.

No. 9. Joint use of organic and plastic mulch to improve bell pepper production in a subtropical region (manuscript not prepared).
Experimental site and primary conditions
Experiments were conducted at commercial farm, Pine Island Farms (PIF), located 22 miles northeast of Homestead, FL and at the Tropical Research and Education Center (TREC), University of Florida, Homestead, FL, The soil at PIF is a very sandy and calcareous soil and that TREC is Krome very gravelly loam soil with low soil fertility and is pH 7.8 (Table 1).

Experimental design and treatments
At Pine Island Farms, the land was used for commercial winter fresh market vegetable production. The previous crop was tomato grown from October through March, and the TREC site had been fallow. On both sites a conventional cover crop, sorghum sudangrass was seeded in a rate of 45 kg/ha and grown from June through August. After the cover crop had been flail-mowed and incorporated into the soil, raised beds each 3 ft wide, 8 in high with 3 ft-wide aisles between adjacent beds were made.
The following five treatments at both sites were 1) organic mulch plus plastic mulch, 2) herbicides, Devrinol and Dual Magnum, applied to organic mulch on the bed, which was then covered with plastic mulch, 3) herbicides, Devrinol and Dual Magnum, applied to the organic mulch on the bed, 4) fumigation of the bed with a mixture of methyl bromide (33%) and chloropicrin (67%) (MC33); the bed was then covered with plastic mulch, and 5) plastic mulch alone as a control. Dual Magnum at 1.78 l/ha and Devrinol powder at 4.5 kg/ha were sprayed on the surface of corresponding beds. Soil fumigation was accomplished by injecting 312 kg/ha of MC-33 into the appropriate beds. Compost at 50 t/ha dry weight equivalent was delivered and evenly spread on the surface of the appropriate beds. Two drip irrigation lines were installed just before covered the plastic mulch.
Bell pepper (Capsicum annuum, cv. ‘Cascade’) was transplanted in two rows on each bed with 30 cm between two adjacent and 45 cm between the two rows. Liquid fertilizer with a formula of 4-0-8 (N-P2O5-K2O) was injected through the drip lines twice per week at 1.1 kg/ha/day prior to fruiting and at 2.2 kg/ha/day after the onset of fruiting.
Peppers were harvested five times from each plot and graded according to the USDA standard.

Virginia Component
No. 10. Using high-residue cover crop mulch for weed management in organic no-till potato production systems.

Research. In order to develop the potato production systems, a split plot, randomized block experiment with four replications will be conducted on 16 raised beds, 40-m long at the Kentland Agricultural Research Farm (KARF), located near Blacksburg, VA. Main plots will be 10-m sections across four beds consisting of ground covers: A = no cover (control); B = sorghum sudangrass; C = a mixture of sorghum sudangrass and sunn hemp; and D = a mixture of spring oats and cowpeas. Subplots within each main plot will be weed management methods: 1 = no herbicides applied; and 2 = pre-emergence herbicides (S-metolachlor and Linuron) applied.
Main plot ground covers will be drilled on preformed raised beds (2 m wide and 20 cm high) in August 2003 and 2004. Appropriate organic fertilizers and irrigation water (if needed) will be applied to achieve production of high levels of weed-free ground covers. The winter-killed cover crops (treatments B, C, and D) will be left undisturbed until potato seed pieces are planted in early April 2004 and 2005. Healthy 40-60 g seed pieces will be placed 10-12-cm deep in the raised beds using the Subsurface Tiller-Transplanter (SST-T) (Morse et al., 1993). Drip tubing will be laid and organic fertilizer applied in row at planting with the SST-T. Additional soluble organic fertilizer (foliar or applied through the drip tubing) will be applied as needed based on leaf-sap nitrate testing during the growing season.
Data collected will include the following: tuber yield, graded according to USDA standards; weed biomass taken at 5 and 8 weeks after planting; and nematode and Colorado potato beetle populations taken at 6, 9 and 12 weeks after planting.

Dissemination (outreach) plan. Four methods will be used to disseminate the knowledge and results gained from this study and previous research at KARF. First, two on-farm grower demonstration plots will be established each year of this study. Second, field days will be hosted at the on-farm demonstration sites. Third, presentations will be given at county and multi-county extension meetings and at the annual meeting of the Virginia Association for Biological Farming (VABF). Fourth, various materials including an extension leaflet will be prepared and distributed electronically as well as in printed form.
Demonstration plots will be established on the farms of two organic growers (Warren LaForce and James Fannon). In August of 2003 and 2004, raised beds will be erected and seeded with an appropriate cover crop mixture, as determined in consultation with the grower and from previous research conducted at KARF during the past eight years. In April of 2004 and 2005, 'Kennebec' potato seed pieces will be planted with the SST - T; production practices and data collection will be followed similar to that used in the research at KARF.

Relevant Literature
Belair, G. and L.E. Parent. 1996. Using crop rotation to control Meloidogyne hapla Chitwoud and improve marketable carrot yield. HortScience 31(1): 106-108.
Bellinder. R.R. and R.W. Wallace. 1991. An integrated production management approach to weed control in potatoes, p. 677-687. In: D. Pimental (ed.). CRC Handbook of pest management in agriculture, 2nd Edition, Vol. III.
McSorley, R. 1998. Alternative practices for managing plant parasitic nematodes. American Journal of Alternative Agriculture 13 :98-1 04.
Morse, R.D. 1999. Mechanical methods of killing cover crops for high-residue/no-till production of transplanted broccoli (Brassica oleracea L. gp. 1talica). Acta Horticulturae, 504: 121-128.
Morse, R.D. 1997. No-till production of Irish potato on raised bed, p. 117-121. In: R.N. Gallaher and R.M. McSorley (eds.). Proc. Southern Conservation Tillage Conference for Sustainable Agriculture, Gainesville, FL.
Morse, R.D., D.H. Vaughan and L.W. Belcher. 1993. Evolution of conservation tillage systems for transplanted crops; potential role of the Subsurface Tiller Transplanter (SST - T), p. 145-151. In: Billich P.K. (ed). Proc. Southern Conservation Tillage Conference for Sustainable Agriculture, Monroe, LA.
Wyman, J .A., J. Feldman and S.K. King. 1994. Cultural control of Colorado potato beetle: Off-crop management. p. 376-385. In: G.N. Zehnder, M.L. Powelson, R.K. Jansson, and K.V. Raman (Eds.). Advances in Potato Pest Biology and Management. APS Press, St. Paul, Minnesota.