An Integrated System of Organic Food Production and Urban Food Waste Recycling Using On-Farm Anaerobic Digestion and Fertigation

Research Plan

Research questions addressed in this project were logistical, financial, health, agronomic, and technical questions concerning collection of food waste, anaerobic digestion, and fertigation. The technical feasibility of anaerobically digesting food waste already has been demonstrated using several different digester designs including a design developed by FCSI (Full Circle Solutions, Inc., 1997). However, to ensure the system was functioning as intended and to validate research results pertaining to use of the liquid fertilizer, some basic digester parameters (pH, effluent solids, effluent electrical conductivity) were monitored throughout the project.
This was a three year project. During the start-up of approximately six months, a food waste collection program was implemented and the anaerobic digester and irrigation systems will be designed, built and tested. Irrigation and agronomic research began soon after start up of the digester.

Logistics of Food Waste Collection and Anaerobic Digestion

The original intent of this project was to collect food waste from downtown Gainesville Florida restaurants and truck it a short distance (1.5 miles) to an organic blueberry farm where the food waste would be converted via anaerobic digestion to a ammonium rich liquid fertilizer ideal for blueberry nutrition and application using an irrigation system. Between the writing of the proposal and the actual start of the project the blueberry farm was put on the market. As a result we decided to locate the digester at Possum Hollow Farm, which is more than 15 miles from the center of downtown Gainesville. The blueberry farm has not yet been sold and is still producing blueberries organically. Early in the project we surveyed the restaurants in downtown Gainesville and found widespread interest in participation in the project. Of twelve restaurants in a four block area in the center of downtown Gainesville, two were not interested food waste recycling, two were already doing it, and the eight others were all willing to participate in the project. In order for the waste hauler (Cherry Tree Recycling) to cost effectively collect the waste from these eight restaurants they needed to make one or two stops downtown. We therefore needed to establish one or two collection points for the food waste. This proved to be a unworkable for a variety of reasons that could have eventually overcome but not in the time frame required for this project. Plan “B” was to collect fish processing waste from a wholesale fish market. We collected fish waste from this facility for several months but found we were being overwhelmed by the quantity of waste and the large number of large bones, shells, and enormous fish heads. Plan “C” was to prove to be the most workable situation and the one we used for the remainder of the project. This involved collecting kitchen waste from a large institutional kitchen located at a state run residential facility for disabled people. Food waste from this facility was picked up in 35 gallon wheeled containers using a flatbed truck. Approximately eight of these containers were brought to the farm every week with the remainder going to a hog farm and a composter located at the local transfer station. This system gave us some flexibility as we could divert food waste to the other recycling options if necessary. In addition, to the food waste we also used the digester to convert Air Potato propagules, an exotic invasive plant, into fertilizer and gas the first two years of the project. This was as a favor to the City Of Gainesville and while effective as a disposal method for the propagules, produced a poor quality fertilizer so we did not offer the service the third year of the project Since the termination of the project the farmer (Joe Durando) continues to use the digester to provide fertilizer for the farm. He collects food waste from one of his customers, a local restaurant. For further discussion of the logistics and economics of this type of recycling can be found in the financial analysis section of this report.

The digester leaching compartments were loaded twice a week. Data from analyses of food waste, solid residues, gas and liquid byproducts were used to evaluate process performance, and perform nutrient and solids balances

Materials Balance, Solid Residue Analysis, and Residue Composting

After 15 months of operation the digester was opened, and cleaned out. Over this period approximately 50,000 pounds or 6,400 gallons of food waste were treated by the digester. An approximate materials balance based on volume can be found in Table A20. About 25% of the added material was removed during the cleanout, 69% was converted to fertilizer or gas, while 6% was left in the digester as active biomass. Biogas was sampled four times in November and December of 2001 and analyzed for CO2 and CH4 using gas chromatography. The gas was %64.4 (stdev 5.7) CH4 and %34.3 (stdev 6.3) CO2. Most of the removed solid residue was applied to a field and disked in. The field was planted a few weeks after disking with plants for cut flowers. Some of the solid residue was composted. Both the solid residue and composted solid residue were analyzed. The stability of the solid, organic residue from the digester was tested by measuring reheating temperatures while the material was in a garden-scale composting system. The solid residue was too wet to compost initially. Because adding bulking agents would have effected the characteristics of the final compost we chose to wait until the material dried out enough to begin compost. This took several months. When the residue began to heat above ambient temperature we assumed it was dry enough to compost and began periodically turning the residue to enhance the composting processes. When the residue does not heat to greater than 10oC over ambient temperature after turning it was considered stable. This took five weeks. We also measured nutrient content, pH, and electrical conductivity of the solid residue and finished compost. These results are summarized in Table A21. While the residue could be processed into a suitable compost, there is little reason to do so as the unprocessed residue is can be used directly and is actually higher in nutrients on a dry weight basis than the finished compost.

Agronomic Trials

The agronomic value of the digester liquid fraction was tested in field plots and a container plant bioassay. Field trials were conducted at Possum Hollow Farms , Bellevue Gardens Organic Farms, Rosie’s Organic Farm, and the Florida Agricultural and Mechanical University (FAMU) research farm in Quincy, Florida. The FAMU trials are discussed in a separate section of the report. In addition, a green house study performed at Rosie’s Organic Farm is discussed in a separate section. Liquid fertilizer for all agronomic trials was applied by hand using 5-gallon containers. The application rates of organic fertilizer materials were based on equivalent nitrogen values of the materials The digester liquid fraction was tested twice for the presence of fecal coliform to ensure that it met Federal 503 regulatory standards. Both tests indicated the effluent met these standards (less than 200 colonies per gram solids). Given that we were not adding manure to the digesters these findings were not surprising. Effluent properties for the experiments discussed in this section are summarized in Table A17 in the Appendices. Fertilizer label information for the commercial organic fertilizer, and fish emulsion are summarized in Table A18. Properties for the chicken manure used in the lettuce trial at Rosie’s Organic Farm are listed in Table A19.

Cover Crop Trials

At Possum Hollow Farm the farmer (Joe Durando) uses living mulches and green manures as an important part of his cropping system so several trials were done testing effluent on cover crops including sorghum-sudan grass (Sorghum halapanse), buckwheat (Fagopyrum esculentum)), and Rye (Lolium Perenne). Three treatments and a nonfertilized control were compared: a chicken manure based commercial organic fertilizer, fish emulsion, and the digester liquid fraction. The application rates were be based on equivalent nitrogen values of the materials. Plots were six by four feet and each treatment was replicated four times. Cover Crops were fertilized at planting and after the first cutting with the equivalent of 30 lbs N per acre. Yields were determined by harvesting a two-foot square area in each plot. We measured yield, tissue nutrient content (N,P,K), soil organic matter, soil pH, and indirectly, salt accumulation in the soil (changes in soil EC).

Commodity Trials

In addition to the cover crop trials, at each commercial farm field trials were done on crops of particular interest to the grower. The fertilizer trials compared the current organic fertilizer source used at each farm to the digester liquid fraction. Bellevue Gardens Organic Farm grows all crops except watermelon on residual fertilizer from chicken manure after the water melon crop is harvested. We did two trials in two succeeding years comparing residual chicken manure and the liquid fraction on Dikon Radishes (Raphanus sativus). Rosie’s Organic Farm performed a green house trial on lettuce starts that is discussed in a separate section of this report. In addition, a field trial was performed comparing the use of manure, effluent, and a nonfertilized control on lettuce (Lactuca sativa) yields. Possum Hollow Farm compared a chicken manure based commercial organic fertilizer, the digester effluent, and a nonfertilized control on Lettuce, Basil, and Kale. Possum Hollow Farms also has been using the digester liquid fraction as a fertilizer for its vegetable "starts" as an alternative to their practice of using an organically fortified potting mix. All the field trails used a completely randomized block design. The digester liquid fraction and the currently used organic fertilizer were applied on an equivalent N basis. Treatment fresh yields measured at harvest and qualitative differences were documented if observed. In addition, soil from some trials was analyzed for soil organic matter, ec, pH, and residual fertilizer at the final harvest to determine if there are any residual effects of the organic fertilizers on soil properties.

The field study conducted was conducted FAMU will determine the environmental impact of applying digester effluent to cropland. This effort is documented in a separate section of this report.

A container, plant bioassay was performed to assess potential phytotoxic effects of the liquid fraction as well as study the release characteristics of the nutrients. Plants received five liquid fraction application rates, from growth promoting to phytotoxic. Controls consisted of a zero level of fertilizer and a water soluble mix of diammonium phosphate ammonium nitrate, and potassium nitrate formulated to match the N P K ratio (19-1-6) of the digester liquid. The replicated trial used sorghum-sudan grass (Sorghum halapanse) as the test plants. Top growth was harvested at two and four months. Plant shoot dry weight and shoot nutrient (N, P, K,) content were measured. The container experiment was a completely randomized block design. Pots were place outside on black plastic mesh and watered only if rainfall was insufficient. Analysis of variance was used to analyze the data. An attempt was made to fit significant treatment effects to the Von Bertalanffy function (Draper and Smith, 1981) that describes full range of plant response to fertilizer additions, including phytotoxicity. Fit was poor so fertilizer response was fit to two linear regression equations, one within the growth promoting range and one within the phytotoxic range.

Results

Pot Study

At the first harvest digester effluent produced growth promoting effects up to a rate of 500 lbs nitrogen per acre (561 kg/ha) and phytotoxic effects at nitrogen application rates of 1,000 and 2,000 lbs nitrogen per acre (1,121 and 2,242 kg/ha). Figures 1a and 1b illustrate these findings. At the second harvest there were no significant (P< 0.05) treatment differences in yields suggesting there were no residual effects, either positive or negative, of the digester effluent. A complete summary of the results of this experiment can be found in Table A1 in the Appendices. Exposure to the heavy downpours typical of Florida summers apparently leached both the digester effluent and the simulated effluent from the pots sometime in the first two months of the experiment. Within the growth promoting range there was no difference between the digester effluent and the simulated treatments. The digester effluent had no effect on seed germination but seedlings at the phytotoxic levels wilted within a few days of emergence. Most of these wilted plants subsequently died. Figures 1a and 1b. The growth promoting and phytotoxic effects of digester effluent compared to application of a simulated (reagent grade chemical equivalent N-P-K value) fertilizer blend on sorghum-sudan grass (Sorghum halapanse)
Cover Crop Studies

Sorghum-Sudan Grass

The data for this trial is summarized in Table A2 found in the Appendices. The plots were seeded August 12, 2000 and harvested September 24 and November, 22. There were no treatment differences in yield at either harvest. At the second harvest the effluent treatment produced higher Nitrogen and Potassium concentrations in the tissue than either the control or other organic fertilizers. The results of this trial suggest that the initial fertility of these plots, built up over years of mulching and cover cropping, was sufficient to grow an adequate cover crop without fertilization.

Buckwheat

This crop went to seed within six weeks and as a result was harvested only once at six weeks. The crop was planted October 10, 2001, and harvested November 16. The plant data are summarized in Table A3 in the Appendices. The fish emulsion gave significantly better growth than either the Fertrell, effluent, or control. However, the Fertrell plots produce higher concentrations of N than all but the effluent plots. This suggests that some element than N, P, or K was limiting growth in these plots. The data suggest that whatever this limiting element was, it was supplied by the fish emulsion.

Rye

The Rye crop was planted and fertilized November 16, 2001, fertilized a second time on December 7, and harvested February 15, 2002. The fertilizer rate was doubled for this trial to 60 lbs per acre at each application. The plant data are summarized in Table A4 in the Appendices. Digester effluent and the Fertrell gave significantly higher (P< 0.05) yields than the fish emulsion treatment which was no different from the control. The concentration of nitrogen and potassium in the plant tissue was highest for the effluent fertilized plots, followed by Fertrell plots. The fish emulsion treatment was no different from the control in this regard. Tissue phosphorus was highest in the effluent fertilized plots while the other two fertilizer treatments were not different from the control.
Commodity Trials

Possum Hollow Farm

Four trials were carried out at Possum Hollow Farm during the project. Both a lettuce trial and a basil trial were destroyed by grasshoppers without any meaningful data being collected. An interesting observation; the burning of young lettuce transplants by contact of the you plants with digester effluent, may have resulted in decreased yields but the experiment did not progress far enough to establish this. The third trial using Kale was more successful. Kale transplants were transferred to raised bed on October 10, 2001 and fertilized. Plants were fertilized again on November 6, and the plants were harvested on December 10. Both the fertilized treatments produced yields almost twice a large as the control. There was no yield differences between the effluent and Fertrell. The effluent plots had higher concentrations of nitrogen than either the Fertrell or control plots. Plant data for this trial are summarized in Table A5. The final commodity trial on Possum Hollow farm was a lettuce trial planted on November 26, 2001. In this trial the digester effluent severely burned the lettuce transplants. Plants were fertilized only at planting and harvested on January 10, 200 and a second harvest was done approximately one month later. The effluent fertilized plots gave the lowest yields due to the initial fertilizer burning. The Fertrell fertilized plots produced the highest yield. At the second harvest there were no significant treatment effects on yield. Plant data for this trial are summarized in Table A6.

Bellevue Organic Gardens

Four trials were done at Bellevue Organic Gardens, two in the Fall,2000 and two in the Fall, 2001. All the trials were done with Dikon Radish, a crop that this farm grows in the winter for local markets. The radishes were fertilized once at planting and again at approximately 6 weeks. The radishes were grown in double rows on raised beds. Each row was divided into 10 foot plots and treatments were randomly assigned to the plots. Fertilizer was applied at a rate of 90 lbs per acre by pouring the effluent into trenches formed in the beds. Whole plots were harvested and fresh weights were determined when the radishes were ready for market. During the first season one of the trials was destroyed by grasshoppers. The second trial, located over a mile from the first trial, did not have grasshopper problems and the effluent treated plots gave significantly higher yields than the unfertilized plots. During the second year there were no significant differences between the treatments in either trial. The plant data for the Bellevue Organic Gardens trials are summarized in Table A7.

Rosie’s Organic Farm: Lettuce Trial

Lettuce seedlings were transplanted to the field on October 21, 2001. Just prior to this manure had been applied to the appropriate plots to give a nitrogen application rate of 150 lbs per acre. Immediately after planting effluent was applied to the appropriate plots at a rate of 50 lbs nitrogen per acre. A second application of 50 lbs N was applied to the effluent treated plots on November 12. A third application was to be applied but the crop matured earlier than anticipated. As a result the effluent plots received 50 lbs per acre less nitrogen than the manure plots. The crop was harvested December 11, 2001. No significant treatment effects were seen for yield or tissue concentrations of N, P, or K. Both the manure and effluent treatments showed significantly higher concentrations of Na in the plant tissues. This is discussed in the next section. The plant data for this trial is summarized in Table A8.

Organic Fertilizer Effects on Soil Properties

The soils data for the trials described above are summarized in Tables A10 through A16. For trials done at Possum Hollow Farm and Rosie’s Organic Farm each plot was sampled at harvest to determine treatment effects on soil properties. The native soils at both farms were sandy and highly weathered. The Possum Hollow soil is a Gainesville Sand (hyperthermic, coated Typic Quartzipsamments). The soil at Rosie’s organic farm is a Millhopper Sand (loamy, siliceous, hyperthermic, Gross arenic Paleudolt). The most common treatment effect was a slight but statistically significant increase in conductivity (ec) associated with fertilization with Fertrell. This was evident in each trial where Fertrell was used with the exception of the lettuce trial at Possum Hollow farm where the control plots had a slight increase in conductivity over both the Fertrell and effluent treated plots. Effluent increased the conductivity in only one trial; the lettuce trial at Rosie’s Organic Farm where both manure and effluent raised the conductivity by almost a factor of 3. This is an important finding as the conductivity of digester effluent ran as high as 18 mmhos/cm at times and there was some concern that this might cause lingering soil conductivity (salt) problems. This is apparently not the case. In the Kale and Rye Cover crop trials Fertrell also increased the post harvest potassium levels in the soil. Post harvest Organic matter was increased by effluent and Fertrell addition over both the control and fish emulsion in the Sudan Grass cover crop trial. Post harvest organic matter was increased by both manure and effluent application in the Lettuce trial at Rosie’s Organic Farm. This was accompanied by an increase in nitrogen by both manure and effluent applications. Manure application increased post harvest potassium levels in this trial as well. In this trial the effluent application was two thirds of the intended rate because the lettuce matured before the third application could be done. As a result it is likely that an even more pronounced effect on post harvest soil qualities may have been achieved with the full application. The effect of organic soil amendments on soil for Rosie’s Organic Farm appeared to be more pronounced than at Possum Hollow Farm. This is most likely due to the higher background levels of organic matte; 3 to 5% for Possum Hollow Farm compared to 1.3% for Rosie’s Organic Farm, and nitrogen; 1,300 to 2,000 mg/kg N for Possum Hollow farm compared to 700 mg/kg N for Rosie’s Organic Farm. Years of mulch application and cover cropping at Possum Hollow Farm has clearly improved these soils.

Sodium Accumulation

In four of the trials harvested plants were analyzed for a metals other than N, P, and K. In the lettuce trial performed at Rosie’s Organic Farm both the manure and effluent fertilized plots had concentrations of sodium in the plant tissues three times higher than the control plots. In the Kale trial done at Possum Hollow Farm the effluent treated plants had sodium concentrations more then ten times the control. In this trial the Fertrell fertilized plots had sodium levels approximately four times higher than the controls but the differences were not statistically significant. On the other hand, in the cover crop trials using buckwheat and rye at Possum Hollow Farm, there was no effect of fertilizer use on sodium levels in the plant tissue. In fact, the average sodium level in the rye trials for all treatments, including the control, was similar (1,000 mg/kg) to the highest, fertilizer induced levels found in the Kale and Lettuce trials. The sodium data is summarized in Table A9.

Field Trial Study-Florida A&M University

Dr. Cassel Gardner, Florida A&M University

Abstract:

Field studies were conducted at the FAMU research farm at Quincy, Florida to determine the effect of food waste effluent (WE) on soil pH and dry-matter (DM) yield of Sorghum sudangrass and ryegrass. The experimental design was a Randomized Complete Block (RCB) with 4 replications. Four levels (treatments) of food waste effluent were applied on plots cropped with sorghum sudangrass during summer and ryegrass during fall. Plots were 6 ft x 10 ft. WE supplied N treatments were 0, 100, 200 and 300 lbs N acre-1. Half the recommended rates was applied at planting and half applied 4 weeks after planting. Application was done by means of a calibrated watering can. Bromine (Br) in the form of sodium bromide (NaBr) was applied, at a rate of 40 lbs acre-1 as a N tracer. To determine initial soil N status, soil samples were taken prior to planting at depth intervals of 0 to 12 and 12 to 24 ins. During crop growth Soil samples were taken monthly and analyzed for NO3-N using Ion chromatography. Above ground crop samples were taken simultaneously with soil sampling and used to determine dry matter accumulation. Results showed that dry matter yield was significantly higher (P < 0.05) for treatments that received the higher rates of WE. Sudangrass cover crop recovered up to 34 lbs N acre-1 at 28 DAP and up to 378lbs N acre-1 at 56 DAP from plots that received 300 lbs N acre -1.
Introduction

Farming practices that are in concert with environmental preservation and natural resource conservation, will simultaneously meet the mandate of the 1990 farm bill for sustainable agriculture, if they conserve water and soil and add on-farm waste material back to the soil environment. Food waste is generally difficult to collect and dispose of due to its high moisture content and rapid breakdown rate. Disposal is generally, through dumping into landfills or by using it as animal feed. However, comparative energy studies have shown that converting food waste to biogas or fertilizer may be more cost effective (Skajaa 1989, Edelmann and Engeli 1993). Organic farmers are concerned with building soil quality to the level of providing sufficient organically derived nutrients for producing crops. A concern which is critical in areas with coarse textured (sandy) soils, warm temperatures and high rainfall. The end result is an acceleration of the mineralization of organic matter, thus facilitating leaching of nutrients form the root zone.

Concomitant with the waste disposal situation is the fact that organic farmers are constrained by price and availability of organic soil amendments for growing their crops. This study represents a part of a project to recycle nutrients and organic matter by anaerobically digesting food waste material and applying the liquid effluent as a soil amendment in cropping systems. The study being reported was conducted at Florida A & M University research and extension farm, located in Quincy, Florida.

The specific purpose of the study was to determine the effect of liquid effluent from digested food waste on the dry-matter (DM) yield of summer grown sorghum sundan-grass (Sorghum bicolor) and winter grown ryegrass (Lolium multifolium (Lam.)

Materials and Methods:

Studies were conducted during July 2000 and October 2001 to evaluate the effect of WE on Sudangrass (Summer) and Ryegrass (Winter) respectively. The experimental design was a Randomized Complete Block (RCB) with 4 replications. Four levels of WE supplied N were applied. Plot sizes were 60 sq. ft. (Fig. 1). Treatment levels were 0, 100, 200 and 300 lbs of WE supplied N acre-1 ( See Table 1 for WE equivalents of lbs N acre-1). Half the rate was applied at planting and half applied 4 weeks after planting. Application was done using a calibrated watering can. Bromine (Br) in the form of sodium bromide (NaBr) was applied at a rate of 40 lbs acre-1 as a N tracer. In order to determine initial soil N status, soil core samples were taken from all plots at planting. During crop growth core samples were taken once each month (July to October for the Summer crop and December to March for the winter crop) from 0 to 1 ft. and 1 to 2 ft. soil depths. Soil samples were analyzed for NO3-N and Br, using Ion chromatography. Above ground crop tissue samples were taken simultaneously with the soil samples until the termination of the study. Crop samples were oven dried to constant weight at 70oC, weighed to determine biomass accumulation following which they were ground in a Wiley mill, tested for N content and used in association with dry matter yield to determine crop N uptake.

Results and Discussion:

Observations on crop growth for both species (Sudangrass and Ryegrass) showed lush green foliage within plots that received 200 and 300 lbs. acre-1 of WE-N. Plant growth in these plots were also much more vigorous compared to the 0 and 100 lbs. N acre-1treatments. At 56 days DAP, dry matter yield from the 200 and 300 lbs. N acre-1,treatments were significantly higher (P < 0.05) compared to dry matter yield from the 0 and 100 lbs. N acre-1 treatments (Tables 2 and 4). However, with the exception of the control plots, dry mater yield appeared to even out in all Sudangrass plots as the plants approached maturity (Table 2). This trend was not evident in ryegrass study since the 200 and 300 pound acre-1 rates continued to yield significantly higher amounts of dry matter compared to the control and 100 pound acre-1 treatments (Table 4). Sudangrass cover crop recovered up to 34 lbs N acre-1 at 28 DAP and up to 378 lbs N acre-1 at 56 DAP from plots that received 300 lbs. N acre-1 (Table 3). Table 1. Conversion table used for the experiment Table 2. Dry matter yield of sudangrass between 28 and 120 days after planting Table 3. N uptake by sudangrass between 28 and 120 days after planting Table 4. Dry matter yield of ryegrass between 28 and 120 days after planting
Figure 1. Field plan for WE experiment

Field Plan of Waste Effluent Experiment at FAMU research, Quincy, Florida

Green House Study: Rosie’s Organic Farm

Developing an Organically Approved Soil Mix for Use in Vegetable Transplant Production

Rachel Seman (Botany Department, University of Florida), Rosalie Koenig (Producer, Rosie’s Organic Farm), Charles Vavrina (Associate Professor, Department of Horticultural Sciences, University of Florida), Robert Hochmuth (Multi-County Extension Specialist, University of Florida) and Nicholas Batty (Nursery Operator, Naples, Florida).

This Portion of the Project Funded Primarily by SARE Producer Grant FS 99-94

Introduction:

The USDA’s National Organic Program (NOP) was implemented on October 21, 2002. As a result, certified organic growers throughout the United States will be required to follow all of the provisions of the NOP’s Final Rule (7 CFR Part 205). Section 205.204, Seeds and planting stock practice standard, requires that growers utilize organically produced annual seedlings if they are commercially available. Organically produced annual transplants cannot be treated with prohibited substances which includes all synthetic substances unless they are specifically listed in Section 205.601 - Synthetic substances allowed for use in organic crop production. A constraint to the production of vegetable transplants that meet organic regulations is the lack of locally available, organically-approved commercial soil mixes or a specific and reliable formulation and methodology for an approved organic mix that could be produced on-farm. Commercial soil mixes are used because of the difficulty of producing an acceptable on-farm mix that is economical, consistent, uniform and which promotes good germination and subsequent plant growth. In California, where the largest concentration of organic production is located, there are companies that manufacture organic soil mixes for the organic market. However, shipment of such products to the Southern Region is cost-prohibitive. In Florida, as in much of the Southern Region, there is no locally available, organically-approved commercial soil mix. This research project addressed the needs of organic producers in the Southern Region by developing formulation recommendations of soil mixes that could be applied to various on-farm operations. We developed a number of formulations through a series of experiments at the University of Florida and on Rosie’s Organic Farm which identified optimal growth media for a number of vegetable species. Organic farmers throughout the Southern Region should test these formulations on their farm to determine those that work optimally in their operations. Growers may have to adjust the concentration of organic amendments in the mixes to obtain the type of transplant growth that they desire. However, these formulations provide the framework upon which an optimal organic transplant production system can be obtained.

Materials and Methods:

There were three phases of the experiment to develop optimal formulations. This report addresses only the third phase in which digester effluent was included as a treatment Formulations for the third phase of experiments were developed in the spring of 2000 at the Suwannee Valley Research and Education Center in Live Oak, Florida by Robert Hochmuth. The final experiment was conducted by Rachel Seman-Varner and Rose Koenig from December to February 2000 at Rosie’s Organic Farm in Gainesville, Florida. All of the treatments with the exception of the anaerobic digester liquid represented the best formulations from the previous experiments. These formulations were tested a final time to determine their worthiness as a general transplant medium for different plant species. The results of this experiment will be detailed below because they represent the best formulations determined from Phase 1 and 2 of the research project

Seeds of collards, lettuce and tomato were planted in flats on December 2, 2000 and grown in a greenhouse. There were three replications of each treatment per species. When the plants developed true leaves the nutrient treatments were applied every 7 days for a total of 6 weeks beginning on December 20, 2000.

A base soil mix was 70% peat, 30% vermiculite and approximately 12g/gal lime, to bring the pH to 6.0. Nine soil amendment treatments were prepared as follows:
1. Base with a top dressing (TD) fish emulsion (1 tablespoon/gal water)
2. Base with poultry manure (17.2 g manure/gal base)
3. Base with poultry manure (11.5 g manure/gal base) with TD fish emulsion
4. Base (80%) and mushroom compost (20%) and Fertrell (45g/gal)
5. Base with TD of hydroponic nutrient (1 tbsp mix A and 1 tbsp mix B/12.5 cups water)
6. Base (80%) and mushroom compost (20%) with TD of fish emulsion
7. Base only
8. Base with poultry manure (11.5 g manure/gal base) and Fertrell (45g/gal)
9. Base with TD digester liquid (.25 cups liquid/ 4.75 cups water)

Comparative photos were taken for each species and each treatment on January 31, 20001. On February 1, 2001 ten tomato plants per replication were harvested above the soil line and the tissue sent to A&L Southern Agricultural Laboratories for composited nutrient analysis that included nitrogen, sulfur, phosphorous, potassium, magnesium, calcium, sodium, iron, aluminum, manganese, boron, copper and zinc. On February 3, 2001 the lettuce plants were harvested and dried. Dry weight data was collected on the plant tissue after the roots were removed.

Results & Discussion

The germination test showed that 93% of the collard seeds, 96% of the lettuce and 93% of the tomato seeds germinated. The average days to 50% germination for collard, lettuce and tomato were 6.25 (standard deviation, 0.18), 6.6 (0.5) and 12.2 (0.3) respectively. Based on the days to 50% germination, each species germinated at similar rates for each treatment.
After six weeks of treatment the dry weights for the lettuce seedlings was recorded. Treatment 7 and treatment 5 were used as the controls and produced dry weights of 0.3 (0.2) and 2.3 (0.4) respectively. Treatment 6 (20% mushroom compost with TD of fish emulsion) had the highest dry weight at 3.8 (0.5) while treatment 9 (TD of digester liquid) had the lowest and was even less than the control treatment of no added nutrients at 0.3 (0.2). The dry weights for treatment 2 (base with poultry manure at 7.5X) and treatment 3 (base with poultry manure at +5X with the TD of fish emulsion) were 2.4 (0.5) and 3.1 (0.3) respectively.

Focusing on the above ground dry weight of the lettuce seedlings as a measure of productivity and using the non-organic hydroponic nutrient as the control, treatment 6, using 80% base soil mix, 20% mushroom compost and a top dressing of fish emulsion, had the best results. Treatment 4, using the same soil and mushroom compost mixture and including Fertrell was the second most productive. Treatment 3, poultry manure with a TD of fish emulsion was the third most productive and significantly better than treatment 5, the control amendment. Treatment 9 was a TD of digester liquid and had very low dry weight. Relatively lower levels of P and higher levels of Mg and Na compared to the standard mix control with nutrient solution likely resulted in it’s overall poor performance. It is apparent from the nutrient concentrations in the plant tissue that the effluent treated plants were phosphorus limited. Nitrogen and potassium levels were not low compared to other treatments. The application rate for the effluent was based on the nitrogen level in the effluent. As the phosphorus level is proportionately lower in the effluent than either nitrogen or potassium, it is probably more appropriate to use the phosphorus concentration of the effluent when determining application rates in artificial media..

Figure 1: Average dry weight of lettuce seedlings after 6 weeks of treatment with nine organic soil mixtures and organic and non-organic soil amendments

Figure 2: Nitrogen content of tomato transplants after 6 weeks of treatment with nine organic soil mixtures and organic and non-organic amendments

Figure 3: Potassium content of tomato transplants after 6 weeks of treatment with nine organic soil mixtures and organic and non-organic amendments

Figure 4: Sodium content of tomato transplants after 6 weeks of treatment with nine organic soil mixtures and organic and non-organic amendments

Figure 5: Phosphorous content of tomato transplants after 6 weeks of treatment with nine organic soil mixtures and organic and non-organic amendments

Irrigation trials

Performance Of Line Source Emitters Under Fertigation Using Leachate From An On-Farm Anaerobic Digester.

Final Report

Dorota Z. Haman
University of Florida
Gainesville, FL 32611﷓0570

David M. O'Keefe
Anne W. Barkdoll
Full Circle Solutions, Inc.
Gainesville, FL

ABSTRACT

With appropriate treatment, it is possible to inject the effluent from an anaerobic digester directly into a drip system for fertigation. Three types of drip irrigation lines under seven treatments for plugging control were evaluated during two seasons on an organic farm that is currently receiving food waste and recycling it using anaerobic digestion and fertigation using the liquid effluent from the digestion process. The following treatments were used to prevent emitter plugging: T1: filtration, T2: filtration and chlorine, T3: filtration and acid, T4: filtration, acid and chlorine, T5: ozone, T6: well water (no effluent), T7: well water and chlorine. The change in uniformity and in flow rate with time was evaluated. Two of the drip tapes DripIn and Chapin were used in both seasons, however TigerTape was substituted with Queen Gil in the second season. There was a difference in performance among the drip lines tested. Sand media filtration without chemical injection was not sufficient to prevent plugging of all three types of drip line, especially during the first season. The quality of the effluent was much better in the second season resulting in less plugging problems in all treatments.

Introduction

There are very few liquid, organic fertilizers currently available. Most organic forms of fertilizer are not sufficiently soluble in water to be suitable for fertigation. An exception is fish emulsion, which however, is ten times more expensive than comparable forms of soluble fertilizer (Burt et al., 1995). Fish emulsion is also often so low in nutrients that it has little fertilizer value (Huntley et al., 1997). Based on FCSI research to date, the liquid fertilizer from anaerobic digestion should be less costly than fish emulsion which is imported into the region from as far away as Alaska (Full Circle Solutions, Inc., 1997). Effluent was injected into a drip irrigation system during two vegetable growing seasons in 2001 and 2002. Three types of drip irrigation lines under seven treatments for plugging control were evaluated in each season.

Some of the characteristics of the digester liquid fertilizer may contribute to clogging of micro-irrigation emitters. Emitter clogging is still a major problem and is related to the quality of the irrigation water (Gilbert and Ford, 1986). Factors such as microbial activity, suspended solids, and chemical activity determine the type of water treatment required to prevent clogging (Gilbert and Ford, 1986). Suspended solids in the range of 50 – 100 ppm and bacterial populations of 10,000 – 50,000 per L can cause moderate clogging problems (Burt et al., 1995). Other than using high quality water sources, methods to prevent clogging include water filtration, flushing and chemical treatment. Chlorine, acids, and ozone are some of the chemical treatments used to prevent clogging (Burt et al., 1995; Burt and Styles, 1994).

Anaerobic digester effluent was applied using a microirrigation system (Haman et al., 1997). Emitter clogging in micro systems can be the biggest problem with fertigation. Usually, sodium hypochlorite (chlorine) is used for periodic cleaning of irrigation lines and emitters. Currently, chlorine is still permitted for irrigation cleaning purposes on organic farms. Other methods, such as ozone treatment, are more expensive but promising for organic production. Clogging problems can also be minimized through careful selection of irrigation and filtration equipment. Five types of treatment (filtration, chlorination with filtration, acid injection, acid combined with chlorine, and ozone treatment with filtration and flushing) were compared to two control treatments; direct well water and chlorinated water. The systems were evaluated for clogging and changes in application uniformity by using a statistical uniformity coefficient (Bralts and Kesner 1983, Haman et al., 1997).

Irrigation and Fertigation Treatments

Effluent was injected into the drip irrigation system during two vegetable growing seasons; one in 2001 and one in 2002. Three types of drip irrigation lines under seven treatments for plugging control were evaluated in each season. Continuous fertigation using the liquid fraction from the digestion process was used during each growing season using a peristaltic injection pump (Masterflex). The injection rate of the effluent was 1 gallon per minute (gpm).

Approximately, 60 gallons of effluent were injected into 5 irrigation treatments during each irrigation cycle.
The test included seven treatments. Five included effluent injection and two controls were used. One control included typical chlorine treatment of well water often used by the growers and one control was well water without any chlorine.

The treatments are summarized below:

Summary of prevention treatments for emitter clogging
Treatment symbol,Treatment description
T1,Effluent + sand and screen filtration
T2,Effluent + sand and screen filtration + chlorine
T3,Effluent + sand and screen filtration + acid
T4,Effluent + sand and screen filtration + acid + chlorine
T5,Effluent + sand and screen filtration + ozone
T6,Well water
T7,Well water + chlorine

All sand filtration was done using a media filter with a #20 media followed by 200 mesh screen.

Each manifold (one for each treatment) included a pressure regulator, a flow meter, a pressure gauge and a 200- mesh screen filter. The layout of the microirrigation system control head and fertigation treatments is presented in Figure 1.

Figure 1. Control head and injection schematic for irrigation/fertigation treatment

Injection of acid and chlorine were done using Masterflex peristaltic pumps with a flow rate of approximately 1 gal/ hour (gph). Acid (hydrochloric, 35%) and ozone were injected continuously during irrigations (treatments 3,4,5). Chlorine injection was done once a week using 10% household bleach at the approximate rate of for one hour. Initially, we attempted to inject chlorine continuously at a rate that would maintain the concentration of 2 ppm at the end of the farthest lateral line. However, due to varying quality of effluent and changing amount of organic matter in the lines it was decided to chlorinate at a high concentration once a week.

The objective of acid injection was to lower the pH of the water to inhibit bacterial growth and to increase the activity of chlorine in the treatment where chlorine was injected. Again, due to variation in effluent quality and very high buffering capacity of the effluent in the first season frequent adjustment was necessary.

The ozone was generated using model CS-4 ozone generator (Ozonology Inc. Northbrook, IL) and injected into the system using a venturi (Mazzei) injector. The rate of air intake into the ozone generator was approximately 2-3 cubic feet per hour. The average rate of ozone that could be detected at the end of the line was approximately 0.2 ppm.

The water collected at the end of the lines was periodically tested for pH, free chlorine and ozone, depending on the treatment.

Three blocks with different drip tape were tested. Seven treatments were completely randomized throughout the block and replicated three times. Each replication consisted of two 50-ft long deep lines. The layout of the drip tape is presented in Figure 2.

All three tapes were 8 mil thick with 8 inch spaced emitters and with very similar flow rates. In 2001 the following tapes were evaluated:

TigerTape – 40 gph flow rate
RoDrip – 40 gph flow rate
Chapin – 39 gph flow rate

In the 2002 season the Tigertape was substituted with Queen Gil with approximately 40 gph flow rate due to the poor performance of the Tiger Tape in the first season. Queen Gil was selected since it has a very different design and emitter flow pattern and there is a lot of interest among vegetable growers in this new drip tape.

Figure 2. The layout of drip tape treatments in the field.

Water was applied daily for one hour. All treatments were watered at the same time. Effluent was injected into T1-T5 whenever the irrigation system was on. Water application to each treatment was recorded using flow meters. The system was turned off after major rainfall and on some days during winter months in the second season. This was controlled by the farmer.

Results

Irrigation System Performance

The change in uniformity and in flow rate with time was evaluated. Two tests were performed in 2001 season and three were performed in 2002. The number of uniformity tests was increased in 2002 since the tape was installed earlier (November 2001) and the fertigation trials started before crops were planted at the beginning of 2002. At this farm plantings are staggered for marketing purposes so there is no specific day of planting.

Table 2. Statistical uniformity (%) of three drip tapes at the beginning of 2001 season
,02/13/01

Table 3. Statistical uniformity (%) of three drip tapes at the end of 2001 season 04//24/01

Table 4. Statistical uniformity (%) of three drip tapes at the beginning of 2001/2002 season11/08/01

Table 5. Statistical uniformity (%) of three drip tapes in the middle of 2001/2002 season
, 03/07/02

Table 6. Statistical uniformity (%) of three drip tapes at the end of 2001/2002 season
, 04/26/02

The flow rate to each individual treatment was recorded using ¾” Kent flowmeters every week or more often. The changes of low throughout the season are presented in Figures 3 and 4.

Figure 3. The changes of flow rates to the individual treatments during the first season.

Figure 4. The changes of flow rates to the individual treatments during the second season.

In the first season the flow rates were reduced at the end of the season in all treatments where effluent was injected. Only two treatments without injection maintained approximately the same flow rate (300 gph). Treatment T1 (effluent injection without any chemical treatment) was showing the lowest flow rate at the end of the season. Media filtration followed by 200 mesh screen without chemical treatment of chlorine, acid or ozone was not sufficient to prevent significant clogging of emitters. Treatments 2,3,4,and 5 had the flow rates reduced by approximately 50% but the uniformity was still high at the end of the season (88%, 85%, 89%, and 74% respectively). This indicates “relatively uniform clogging” along the lines. To deliver the required amount of water to the plants, due to the flow rate reduction, the watering time would have to be double by the end of the season.

In the second season there were no significant differences in the flow rates between the beginning and the end of the season (no significant emitter clogging) and among the treatments.
Based on the results, it can be concluded that with appropriate chemical treatment, it is possible to use the effluent from the anaerobic digester for injection directly into the drip system with minimal loss of uniformity throughout the growing season.

Differences in the results between the first and second season are probably attributable to improved management and operation of the digester during the second season. Improved management and operation of the digester led to more consistent effluent properties and more thoroughly treated effluent. During the first season effluent was being drawn from the leach bed portion of the digester due to plumbing difficulties associated with the choice of high-rate centrifugal pumps for recirculating effluent. This problem was rectified by the second seasons trial by using low-rate peristaltic pumps. As a result we were able to use the effluent from the second stage (pack bed) that was lower in both total and volatile solids. Average total and volatile solids for effluent drawn from the leach bed portion of the digester were 1.3% and 56.6% respectively, while for effluent drawn from the packed bed portion of the digester they were 0.9% and 50.4%. As a result, effluent drawn from the leach bed portion of the digester had higher inert particulates and carbohydrates to encourage bacterial growth. Both of these factors can increase fouling of emitters. In addition, the during the irrigation trials the first season the digesters were cleaned out and restarted using air potatoes. As a result the effluent was affected by both the change in digester feed stocks and effluent changes during the digester startup after the cleanout. During the second trial the digester had been running for over six months at “steady state” conditions treating only food waste. This change may have contributed to the lower plugging problems in the second season (see tables 2-6). We consider the first season’s trial conditions to be close to a worst-case scenario for effluent quality and the second season conditions to be normal..

Irrigation Water Quality

Periodically the pH and Electrical Conductivity (EC) was measured at the plots. These results are presented in table 7.

Table 7. Electrical conductivity and pH of water in different irrigation treatments throughout the second season.

A sample of effluent was tested every time the tank was filled. The results of the tests are presented in table 8. There was variation in nutrient content of the effluent throughout the season.

Table 8. Effluent analysis

The well water was tested twice during the second season to provide a base line for nutrient analysis. The results of these tests are presented in Table 9.

Table 9. Well water analysis

Table 10 presents the nutrient analysis of two groups of treatments (with effluent and without effluent). These are the nutrient concentrations that actually were applied to the field. The tests were performed three times during the second season. The concentration of nutrients for T1 through T5 was approximately the same since the injection was done with one pump injecting the effluent into the submain line supplying the water to these treatments (see Figure 1). Each number in the second column represents an average value for five treatments and three different dates. In the last column the two treatments were averaged over three sampling dates.
Table 10. Average content of nutrients per treatment

Soil and Plant Testing

Two plots were selected for soil testing. One of them was under T6 treatment for two seasons (pure water application) and the other was T1 (effluent only). Soil was sampled on April 5, 2002. The results of the tests are presented in Table 11. There was no statistically significant difference between the treatments.

Table 11. Soil sample analysis of two treatments T1 and T6.

Yield Results

Two varieties of squash were planted on these plots. The yield is presented in Tables 12 and 13 respectively. The treatments were not replicated so we cannot say that statistically the yield was increased. However, the number of fruit and the fruit weight were higher in T1 (effluent applied) then in T6. For Lebanese variety the number of marketable fruit was 118 for effluent fertigation treatment (T1) as compared to 84 without the effluent application (Table 12). For the Zepher variety the crop produce with effluent fertigation was twice as high (number of fruit and weight) and most of the fruits were sellable. The total number of fruit produced using the effluent was 128 as compared to 111 in the treatment without the effluent. The biggest influence of effluent application was on production of marketable fruit . Comparing sellable fruit during the season, T1 produced 115 as compared to 69 in T1 (Table 13).

Table 12. Yield of Lebanese squash

Table 13. Yield of Zepher squash.

Periodically the pH and Electrical Conductivity (EC) was measured at the plots.

Conclusions

It can be concluded that effluent can be injected into the drip line if appropriate clogging prevention methods are used to prevent the decrease of uniformity. The quality of effluent is very important in drip tape performance. Drip tape selection is an important factor in maintaining high application uniformity throughout the season.

References

Burt, C.M., and S.W. Styles. 1994. Drip and Microirrigation for Trees, Vines and Row Crops. Irrigation Training and Research Center, San Luis Obispo, CA pp. 261

Burt, C., O’Connor, K. and T. Ruehr. 1995. Fertigation. Irrigation Training and Research Center, San Luis Obispo, CA. pp. 295.

Bralts, V. F. and C. D. Kesner. 1983. Drip Irrigation Field Uniformity Estimation. Transactions of the Amer. Soc. Ag. Eng. 26(5):1369-1374.

Full Circle Solutions. 1997. Project Final Report: Percolating-bed anaerobic composting of food waste. Submitted to: The Cooperative Research, Education, and Extension Service, United States Department of Agriculture, Washington, DC.

Gilbert, R.G. and H.W. Ford.. 1986. Emitter Clogging. In: Trickle Irrigation for Crop Production: Design, Operation and Management. Eds: F.S. Nakayama and D.A. Bucks. pp.142-163. Elsevier.

Haman, D.Z., R.T. Pritchard, A.G. Smajstrla, F.S. Zazueta, and P.M. Lyrene. 1997. Evapotranspiration and crop coefficients for young blueberries in Florida. Applied Eng. in Agri. 13:209-216.

Haman, D. Z., A. G. Smajstrla and D. J. Pitts. 1997. Uniformity of Sprinkler and Microirrigation Systems for Nurseries. Bulletin 96-10:1-7, Fla. Coop. Ext. Ser. Univ. of Fla., Gainesville, Florida.

Huntly, E.E., Barker, A.V., and M.L. Stratton. 1997. Composition and uses of organic fertilizers. pp.120-139.