Four cotton tillage systems were evaluated for PM10 emissions, energy requirements, field operation times, yield, grower adoption and economic analysis. PM10 emissions, fuel usage and field operation times reduced as the number of field operations decreased. PM10 emissions and total fuel usage both reduced by a factor of four when comparing conventional tillage to the single-pass Pegasus system. Field operation times were reduced by a factor of five when comparing these same tillage systems. Replacing the Conventional system with the Pegasus system produced an additional $25.91 per acre while the Sundance and Paratill systems increased income by $16.76 and 16.06 per acre, respectively. Growers are slow to adopt minimum tillage systems even though there is an increase in revenue and there is benefit to reducing the amount of fugitive dust. However, the survey results showed that growers are reducing the number of field operations for the current conventional tillage systems.
- Compare three minimum tillage/controlled traffic systems to a conventional system in terms of profitability/efficiency, sustainability and particulate generation.
Evaluate commercially viable methods of reducing emissions from tillage operations.
Demonstrate minimum tillage systems at field days, thereby permitting first hand observation by growers, extension personnel, etc. of their functionality and performance.
Disseminate results through the popular press, Extension bulletins and technical manuscripts.
The majestic plume rising behind the tractor and implement advancing across the cotton field (Photo 1) has long been a common site in the American farming scene. What was in it? Who Knew? How could we lessen it? Who cared? The rising awareness of air pollution and the consequences of breathing airborne particulates thrust environmental conditions into the public eye. The passage of the 1990 Clean Air Act by the United States Congress provided the impetus to monitor and control air borne particulates of less than 10 micrometers aerodynamic diameter (PM10). Particulate matter 10 micrometers or less in diameter (0.0004 inches or one-seventh the width of a human hair) can pass through our natural defense mechanisms and collect throughout our respiratory system. This new standard considers smaller particles to be more responsible for affecting ones health. The U.S. Environmental Protection Agency’s (EPA) national air quality health standard for PM10 is 50 micrograms/cubic meter measured as an annual mean and 150 micrograms/cubic meter measured as a daily concentration. Exposure to these levels of fugitive dust can cause health problems, especially to those with pre-existing respiratory or cardiovascular diseases.
The Federal Clean Air Act requires that fugitive dust emissions from all significant sources where the ambient air quality standards are not being met must develop programs to address these areas. The Arizona Department of Environmental Quality (ADEQ) determined in 1995 that agriculture activities as well as other sources were contributing to the production of particulate matter (PM). EPA designated Maricopa County as a high PM10 area in 1996 and required that programs be developed to address the PM10 problems in this county. Kennedy and Wilson (2003) described in detail the chronology of events surrounding Arizona’s PM10 non-attainment designation by EPA. Arizona’s governor created the Agricultural Best Management Committee to develop an agricultural PM general permit to address these issues. Agriculture Best Management Practices for Arizona were written and published on February 2001. Anyone who engages in agricultural activities must comply with this general permit. Further more, anyone engaged in agricultural activities who isn’t in compliance will go through a review that could result in the loss of his general permit and be required to obtain an individual fee-based permit. There is a general feeling that these regulations could easily be applied to other Arizona counties. This same scenario is happening in other states. In addition, EPA is considering setting new standards for PM2.5 (particulate matter 2.5 micrometers or less in diameter) and is reviewing revisions to the current PM10 standards.
Arizona has a mandated cotton stalk plow down program that is intended to assist in the control of pink bollworm. This program requires that cotton producers must have all of their cotton stalks plowed down by a certain date and cannot plant a new cotton crop until a later date. These mandated dates vary depending on the climate and typical planting dates of each region. All cotton stalks, including the major root system, must be completely removed from the soil, thus preventing the cotton plants from growing. Three dollars is assessed for every cotton bale until each field is inspected and certified to meet plow down criteria before the money is returned to the grower. Failure to meet plow down will result in the loss of the cotton bale assessment and the money retained would be used to bring those fields into compliance. Because of this mandate, growers must engage in tillage activities during the driest portion of the year. The resultant dust particulates put into the ecosystem during this period may have an effect on the failure to maintain levels below the maximum allowable PM10 amount as determined by EPA.
Fugitive dust has always been a common sight in the desert Southwest. The urban areas are moving into traditional agricultural areas and the effects of dust are becoming a major problem. Urban development will continue in agricultural areas because this is the only source of water for development. Whether the dust originates in native desert lands or agricultural fields has long been discussed. Large dust plumes caused by summer monsoons, thousands of feet high and miles wide, have been photographed as they passed over agricultural and native desert lands on their way to urban areas. However, there is no question that dust plumes are created by agricultural equipment as farmers prepare the soil for planting the next crop or harvesting the current crop. Desert soils receive limited rainfall and agriculture must depend on irrigation to meet plant consumptive use. Use of valuable irrigation water is limited to critical periods of crop establishment and growth. Thus, conditions are optimum for creating dust when implementing agricultural cultural practices. Saxton (1996) discussed in detail the hazards and controls of PM10 particulate material emitted by wind erosion on Washington State agricultural lands in the Columbia Plateau. He also pointed out that there is evidence that a significant amount of particulate matter was generated by wind erosion on upwind agricultural fields that impacted downwind urban areas. However, there is currently no instrumentation or ability to quantify what these wind soil erosion emissions might be or how much is transported to the urban areas.
Quantifying the levels of PM10 caused by agricultural implements is very difficult because they are constantly moving across the field. Only the width of the implement that is in contact with the soil is causing fugitive dust to become airborne. Where do you locate a stationary PM10 monitoring equipment to determine the quantity? How do you quantify the amount of PM10 if the wind increases, decreases or changes direction? Coates (1997) compared five different cotton tillage systems, Conventional, Sundance, USM, Modified Conventional and Puller. The Conventional system was the typical Arizona cotton grower practice with seven different field operations that included shredding, ripping, three passes with a disk, listing and mulching. The other tillage systems were considered reduced tillage: Sundance four field passes, USM three field passes, Modified Conventional four field passes and Puller four field passes. The dust particles collected in this study were the mass concentration of total suspended particulate matter (TSP) in ambient air. This was the EPA standard prior to 1987 when they replaced the TSP air quality standard with a PM10 standard. Coates’ results showed that the Sundance uprooter generated the greatest amount of particulates compared to all the other implements. There was also a trend that showed the amount of particulates reduced as the number of field passes decreased.
Reduced or minimum cotton tillage systems can offer other advantages besides reducing particulate matter emissions, such as reduced energy requirements and less time spent in the field with no loss of yield (Coates et al., 1994). Coates (1996) reported that the reduced tillage systems of Sundance, Modified Conventional and Puller used less than 50% of the energy compared to the Conventional system. The USM energy required was about 75% that of the Conventional system. Growers could save energy (fuel) if they were to convert from Conventional tillage systems to a reduced tillage system. Time spent in the field was reduced to 58% for the USM and Sundance reduced tillage systems compared to the Conventional system (Coates and Thacker, 1994). It was also reported that there were no differences in cotton yields for all tillage systems (Coates and Thacker, 1997) and soil compaction levels weren’t significantly different among the three tillage systems of Conventional, USM and Sundance (Coates, 1990).
Dust samples caused by agricultural implements were collected from 29 farming operations performed in furrow-irrigated crops near the University of California at Davis (Clausnitzer and Singer, 1996). The samples collected were defined as respirable-dust (particles smaller than 4 micrometers diameter). The respirable-dust fraction can reach the alveolar region of the lungs and cause serious health problems. Personal cyclone samplers with filters were mounted 38 inches above the ground on each implement. The cyclone’s cylindrical design results in a spiraling motion that separates fine and course particles depending on the vacuum flow rate. The respirable-dust samples were collected very close to their source and therefore shouldn’t be used to determine dust concentrations over distance and space or interpret off-site environmental effects. The highest respirable-dust concentrations were produced by intensive land preparation operations that have a high level of soil contact (land planning, ripping, plowing and disking), while those field operations that have a lower level of soil contact (harrowing, cultivation, seeding and listing) have lower levels of concentrations (Clausnitzer and Singer, 1997). They also reported that air temperature was more correlated with the respirable-dust collected compared to all the other independent variables (soil moisture, tractor speed, sampler height, wind speed and wind direction relative to the driving direction). Although, soil moisture, wind and tractor speed each were found to have a significant effect on respirable-dust production. Carvacho et al. (1999) reported results from the San Joaquin Valley, California, that cotton harvesting had much higher dust concentrations (340 kg/km2) than the shredding of the cotton stalks (258 kg/km2).
The California Air Resources Board (1988) suggested that more than one-third of the PM10 particulates in the San Joaquin Valley originate from paved and unpaved roads and another one-third from farming operations. It is difficult to trace the source or the amount of these non-point sources using stationary dust sampling stations. Holmen et al. (1998) proposed using a field portable remote sensing technique, light detection and ranging (lidar). The fast scanning lidar shows great potential for measuring PM10 emissions from agricultural field operations. The advantages of the lidar system are: (1) plume dynamics could be described in detail, (2) measurements of average wind speed and direction over 50 to 100-meter scales, (3) quantitative determination of the dust fraction missed by point sampling arrays and (4) currently provide unparalleled information on non-point source emission variability, both temporally and spatially. As this technology becomes commercially available, data can be collected that more accurately define the dust emissions from non-point agricultural sources.
Two sites (Marana Agricultural Center located near the town of Marana, Arizona, and Maricopa Agricultural Center located between Phoenix and Casa Grande, Arizona) were selected for this study. These two sites are approximately 70 miles apart. The twenty-acre Marana site was identified as Fields E-1 and D-1 by the Marana Agricultural Center. The soil at this site was classified as a Pima Clay Loam (fine-silty, mixed, thermic and coarse loamy, mixed thermic Typic Torrifluvents). Each field was divided into three replicates and randomized for each of the four different tillage treatments, resulting in six replicates. Each treatment consisted of twelve rows divided by a two-row border to help delineate treatments. The Conventional tillage treatment consisted of 28 rows rather than the twelve rows so the Conventional tillage equipment could be used. The cotton rows were oriented in a north-south direction and were about 600 feet long.
The twenty-acre Maricopa site was identified as the south half of Field 22 by the Maricopa Agricultural Center. This field was laser-leveled to zero slope producing two benches because of the side fall. The soil at this site was classified as a Casa Grande Sandy Clay Loam (fine-loamy, mixed, hyperthermic Typic Natraargids). The field was divided into five replicates and each was randomized for the four different tillage treatments (the same tillage treatments as the Marana site). Each treatment consisted of twelve rows divided by a two-row border to help delineate the treatments. The Conventional tillage treatment consisted of 36 rows rather than the twelve rows so the Conventional tillage equipment could be used. Field 22 Bench One contained the first two Conventional tillage replicates and the first three replicates of the three minimum tillage treatments (Rototill/Paratill, Pegusus and Sundance). The remaining Conventional tillage replicates and minimum tillage treatment replicates were located in Field 22 Bench 2. The cotton rows were oriented in an east-west direction and were 660 feet in length.
All field implements utilized in this study were set for four equally spaced 40-inch rows. The cotton was planted in dry soil to a population of approximately 40,000 plants per acre and irrigated. This reduced the time and energy requirements of additional tillage practices to disk in the herbicide application, pre-irrigate and mulch before planting the cotton in moisture as practiced by some growers. The short staple cotton variety that was grown at both locations for all years was DP448B; which is a transgenic Bt type. The cotton cultural practices of irrigation, fertilization, cultivation, insecticide, herbicide and defoliation applications, etc. were typical grower practices and were identical for all treatments at each location. The cultural practices varied between locations because of environmental differences. The Marana cotton planting dates were April 18, 2000, April 17, 2001, April 11, 2002 and April 23, 2003. The 2003 cotton had to be replanted on May 7 because of a poor stand due to cold soil temperatures that reduced plant germination and vigor. The Maricopa Cotton planting dates were April 18, 2000, April 30, 2001, April 16, 2002, and April 10, 2003. The Marana cotton plots were harvested on September 29, 2000, October 24, 2001, and October 15, 2002. The Maricopa cotton plots were harvested on October 2, 2000, September 25, 2001, and September 28, 2002. After all treatments had been harvested and yield data obtained, the different tillage systems were imposed to each replicated set of plots and the data recorded. Yield data were only collected from four center rows using a two-row cotton picker. A different two-row cotton picker was used at the two different test sites. The fields remained fallow until they were planted the following April.
The four tillage systems selected for this study were Conventional, Rototill/Paratill, Pegusus and Sundance. The Conventional system was determined based on typical grower practices in Central Arizona. The Rototill/Paratill, Pegusus and Sundance tillage systems were selected since they are commercially available minimum tillage implements that have been or could be incorporated into commercial grower practices. The Conventional system resulted in establishing new beds and furrows each growing season while the minimum tillage systems used the same bed-furrow arrangement for each year. Only the cultural practices that occur after cotton harvesting and before cotton planting next spring were evaluated in this study. The planting, cultivation and harvesting procedures were not part of this study. The same Case-IH 7250 MFWD tractor was used for each implement and tillage system for the second and third years of this study. A Case-IH 7140 MFWD was used the first year. The tractors are very similar in attributes, the main difference being that the Case-IH 7250 model tractor engine has a higher horsepower rating. The field performance data were commingled from both tractor models because of their similar characteristics. The same field implements were used at both test sites.
The Conventional system consisted of a IHC 14-foot medium offset disk with 24″ blades (Photo 2), a Dandl four-row stock shredder (flail type knives) (Photo 3), a three straight shank ripper set on 40″ centers and capable of ripping to a 18-inch depth (Photo 4) and a five-bottom Allis-Chalmers lister set on 40″ centers and equipped with an Accu-Trac mechanical guidance system (Photo 5). New wear parts were replaced on all implements to ensure that they were in excellent condition prior to initiating this study. After the cotton crop was harvested, the standing cotton stalks were shredded to a height of 6 inches. The field was then disked parallel to the rows, which started to flatten the bed-furrow arrangement. The fields were then ripped 18 inches deep at a diagonal to the cotton rows to reduce the compaction that had occurred during the production season. The fields were disked a second time diagonally opposite to the ripping direction. At the completion of this operation the fields were flat and the original cotton rows were not visible. Each Conventional tillage plot was listed to form new 40″ beds for planting the cotton the next season. A total of five different implement passes was made to each Conventional tillage plot.
The Pegasus tillage system (Photo 6) was designed and patented by Mr. Gary Thacker who sold the patent rights to Rome Plow, Inc. This system is designed as a once-over machine that can bury standing cotton stalks in the current bed. The machine uses a slip type plow to open a furrow in the center of the bed and force the standing cotton stalks into the bed at about a 6-inch depth. Disks follow this plow and shape the furrow and beds so that the bed is ready for planting. Rome Plow Inc. provided a 4-row implement the first two years of this study, however, a unit had to be rented the third year. This machine required great care in field setup and adjustment in order to perform as designed. The Pegasus tillage implement has to be exactly centered on the cotton rows in order to bury the cotton stalks and prepare the bed for planting. If not, the implement won’t bury the stalks or make a suitable bed for planting. This is a single-pass tillage implement that does not require the cotton stalks to be shredded beforehand.
The Sundance system was designed and patented by Mr. Howard Wuertz, and has gained wide acceptance for drip-irrigated fields. This tillage system does require that the standing cotton stalks be shredded to a height of 6 inches before the implements are utilized. This system consists of two parts, a root puller (Photo 7) and the ripper/disk lister (Photo 8) or Level Bed Disc (Photo 9). The Level Bed Disc wasn’t available for the first year of this study because of manufacturing problems. As a result, a two-row ripper/disk lister was used. Typically growers would use these two parts by attaching the root puller to the front of the tractor and either the ripper/disk lister or the Level Bed Disc on the rear of the tractor. Thus, one would be able to complete this job as a once-over operation after the cotton stalks had been shredded. The root puller and ripper/disk lister or Level Bed Disc were run separately for this study so measurements of the power and dust levels could be determined for each implement. The root puller centers itself on the stalk row by design and uses two converging disk blades per row that pinch the stalk and lift it out of the soil. The ripper/disk lister used cone disks that first opened the bed and then used a reversed set of cone disks to build the bed back. The Level Bed Disc used small disks on each bed to till the bed and cone disks followed to build the bed back. Both the ripper/disk lister and Level Bed Disc utilized a straight ripper shank in the bottom of each furrow. The ripper shank was inserted about 12 inches. The cone disks are designed to rebuild the bed to its original shape and prepare it to plant the next crop. The Level Bed Disc implement did a much better job of covering the cotton residues compared to the ripper/disk lister. Compared to the Pegusus tillage equipment not all of the cotton residues were buried using the Sundance Level Bed Disc implement. A total of three implement passes were made with this system. The Sundance equipment used in this study was provided by Arizona Drip Systems of Coolidge Arizona.
The Rototill/Paratill system was a combination system. Northwest Rotovator provided a four-row production implement (Photo 10) that was used for the first year of this study. This implement is basically a heavy duty roto-tiller for each cotton bed. It was designed to till each cotton bed to a 6-inch depth and shape the bed for planting. This system also required that the standing cotton stalks be shredded to a height of 6 inches prior to tilling each bed. The machine wasn’t capable of completely uprooting all of the shredded cotton stalks as required by the Arizona cotton plow down rules. The forward travel speed of this implement was quite slow. The beds built by this implement were clean of cotton residue and shaped for planting. The Northwest Rotovator Company decided not to pursue any implement sales in Arizona and this system wasn’t available for the second or third years of this study. A Bigham Brothers Paratill (Photo 11) was substituted for the second and third years. The Paratill system utilizes a bent leg ripper shank that is inserted in the bed center and cone disks follow that reshape the bed-furrow arrangement. The Paratill tillage system also requires that the standing cotton stalks be shredded. Ripping the center of the bed did help uproot the shredded cotton stalks and roots. The Bigham Brothers Paratill implement was available at the Maricopa Agricultural Center and was also used at the Marana site. A total of two implement passes was required for either the Rotovator system or the Paratill system.
A dust-sampling unit with a total of 16 inlets or air/dust intakes was designed and constructed. The air/dust intakes were equally spaced 24 inches on center vertically and 48 inches horizontally. At each elevation the four inlets were equally spaced and each of the four-air/dust intakes connected to a collector box (8″x8″x12″) located in the center (Sketch 1). Calculations of the expected airflow friction coefficients with application to the Reynolds Number in turbulent pipe flow for 100 m3/hr total airflow were conducted. It was determined that a 4:1 ratio of pipe lengths of 4″ ABS and 3″ABS pipe diameters would result in approximately equal flow rates through each pipe section. The center two-air/dust intake pipes were 3 inches diameter and the pipes were 15 inches long. Including the two 4-inch 90-degree plastic elbows and connectors these air/dust intakes were spaced 48 inches apart. The outside two air/dust intake pipes were 4 inches diameter and the pipes were each 60 inches long. Including the two 4-inch 90-degree plastic elbows and connectors these air/dust intakes were spaced 144 inches apart. The elbows attached to the end of each plastic pipe were oriented downward to collect fugitive dust and reduce the possibility of collecting larger particles. A high volume dust sampler (General Metal Works, Inc. model number 76-100) was connected to each of four collector boxes to create the desired airflow for collecting the dust samples from each sampling elevation.
A frame was designed and constructed so that the air/dust intake pipes, collector boxes and high volume dust samplers could be connected. This frame was also designed so that it would easily attach to common mounting brackets. Individual mounting brackets were built and attached to each individual tillage implement. This allowed the dust-sampling unit to be easily moved from implement to implement while maintaining the same sampling grid (Photo 12). The first elevation of air/dust intakes was located approximately 24 inches above the field elevation.
The electrical power required for the high volume dust samplers and other electronics was supplied with a 5500-watt generator mounted on the front tractor weights. An electrical extension cord was made to connect the generator and the samplers with a control box located inside the tractor cab to control the generator and samplers.
A laptop computer equipped with a 400 MHz processor and 6 Gbyte hard drive was used to process data in real time. The National Instruments graphical programming language and virtual instrument (vi) software package LabVIEW version 6i, was used for collecting and processing data from all instruments. A National Instruments DAQ CArd-AI-16E Multifuntion I/O with an SCB068 Shielded I/O connector block was installed to connect the instruments to the computer. The computer was mounted inside the tractor cab
Fuel usage and ground speed were recorded using a Blackwatch Marine fuel flow transducer, AD590 current loop temperature transducer and the Case-IH ground-sensing radar all connected to the laptop computer. Atmospheric pressure and high flow sampler outlet pressures were measured with SenSym ICT ASCX family of differential/gauge pressure transducers that were connected to the I/O connector block. Atmospheric pressure was measured with a 0-100 kPa gauge transducer while outlet pressures at each of the four high volume-sampling pumps were measured with 0-7 kPa differential transducers. Air temperatures at the four-air/dust intake elevations were measured with AD590 current loop transducers. All sensors were wired from their location to the tractor cab and the laptop computer.
It was desired to determine the dust expressed in mass of particulates per unit area of the tilled field (i.e., g/ha) for each of the four sampling elevations in the dust plume. This may be calculated as the amount of dust particulates collected on a filter divided by the sampled plot area (length of row x width of implement x vertical spacing of the sampling inlets). If a filter were to partially plug during a dust collection pass of a tillage operation, a reduction from the expected airflow through the filters would result. Consequently, the reduced airflow through the filters would result in a less-than-expected volume of dust collected by the filter. This would result in an under reporting of actual volume of the dust plume. To address this issue, LabVIEW software and electronic pressure sensors were used to calibrate the high volume sampler outlet orifice differential pressure values. A mechanical vane airflow measurement anemometer was used as a calibration reference. Outlet orifice pressure differential would then be used to model resulting airflow for each of the four high volume sampling pumps. The resulting chart and conversion coefficients are shown in Figure 1.
The air volume passing through each of the high flow samplers can be found for any row segment in the field using the distance traveled and the airflow rate. A tractor operating at a known speed for a fixed time interval will travel a known distance. Thus, the amount of air drawn through a sampler filter in a fixed time interval can be determined. The implement width, the row length and the vertical spacing between sampling inlets determine the total volume of dust plume that is being sampled. Dividing the sampled filtered airflow volume by the total dust plume volume will give a proportion of the actual dust plume being sampled. Multiplying the mass of the dust particulates found on the filter by the inverse of this proportion will result in the amount of dust particulates created in the tested region during the sampling test. Dividing the amount of particulates computed during the test by the test plot area will give the particulate amount per unit tilled soil area.
The filter papers in the high flow samplers were very cumbersome to install and retrieve. The Arizona Department of Environmental Quality provided the high volume samplers and filters and measured the dust collected on the filters. They noted that errors could occur when filters are mishandled, torn or touched with dirty fingers in the installations and removal process. It is nearly impossible to handle the filter papers correctly under field conditions with the high volume samplers attached to each implement. There is also evidence to suggest that the filter papers contain particle sizes greater than 10 microns. The separators for the high volume samplers could not be used in this application. Particle sizes greater than 80 microns rarely stay in suspension because they are too heavy. Thus, particles greater than 10 microns and possibly less than 80 microns could easily be collected on the paper filters. The dust values collected the first year were Total Suspended Particulates (TSP). Since EPA no longer uses the TSP standards for measuring air quality, the filter papers were abandoned in favor of the electronic particle counters that would report air quality as PM10 values. Instead, four electronic particle counters (Met One Instruments, model GT-640 Logger Particle Monitor) were acquired and used for the second and third seasons. The high volume samplers were still used to collect the fugitive dust through each air/dust intake; however the filter papers weren’t installed. The electronic particle counters were attached after the collector box near the filter location to obtain a air/dust sample for measurement. These samplers are equipped with a separator to eliminate those soil particles greater than 10 microns. Since EPA no longer uses the TSP standards for measuring air quality, the filter papers were abandoned in favor of the electronic particle counters that would report air quality as PM10 values.
The Met One electronic samplers are complete ambient air samplers using a forward light scattering detector and built-in data logger. A laser optical sensor is used to detect and measure particle concentrations up to 10 milligrams per cubic meter. The continuous flow optical sensor is combined with purge air to ensure accurate measurements in adverse environments. The sensor is enclosed with matched electronics and sensor display to provide a complete modular unit. Built-in calibration functions are included. These particle counters report the actual number of either PM10 or PM2.5 particles that pass through the sensor per unit time, which gives real time data and is recorded directly in the computer. The pressure transducers on each high volume sampler motor were eliminated since the paper filters were no longer used.
The energy required to pull each implement used in meeting the cotton plow down requirements is of importance to the cotton producer. The basic energy unit applicable to field operations such as cotton plow down is horsepower-hours per acre. This unit is somewhat difficult to measure directly and easily. However, a directly related unit that is simply obtained is gallons of fuel consumed per acre. Fuel consumption is an easily understood concept that cotton producers and others involved in mechanized agriculture do understand. Direct conversion to cost per acre is easily calculated by multiplying the gallons of fuel per acre by the current fuel price. The tractors used in this study used diesel fuel.
The fuel rate consumption, implement width and tractor speed values are necessary to calculate the diesel fuel consumption per acre. The implement widths in this study were all 160 inches (four 40-inch rows), except the ripper that was 120 inches wide and Sundance ripper/disk lister that was 80 inches wide or 2-rows. The tractor radar unit measured the ground speed, and fuel consumption was measured in gallons per hour using an electronic fuel meter. The fuel usage was recorded twice for each implement, once operating in the ground and once pulled across the field. Energy usage was defined as gross energy, the total fuel consumed by the tractor with the implement operating in the ground. The operating energy required for the tractor and each implement being pulled (not operating in the ground) was also recorded. Net energy was calculated by subtracting the operating energy from the gross energy. Thus, net energy is the amount of fuel required to power each implement as it is operating in the ground. All energies were recorded as gallons of diesel fuel consumed per hour. These data were converted to gallons of diesel fuel used per acre using tractor speed and implement width.
The gallons of diesel fuel used per acre for each implement for year 2000 and 2001 are shown in Table 1. A different tractor was used each year because of availability. A Case-IH 7140 MFWD was used in year 2000 and a Case-IH 7250 MFWD was used in year 2001. Both tractors were very similar except that the Case-IH 7250 model had greater engine horsepower. Fuel data were not collected in year 2002 because of electronic problems that couldn’t be solved during the fall study. Gross fuel usage increased from 2000 to 2001 for the Conventional and Sundance tillage systems. The 4-row Sundance Level Bed disc was used in 2001, which required more power compared to the 2-row Sundance ripper/disk lister used in 2000. The Rototill/Paratill gross fuel usage decreased because the Rototill implement was used in 2000 and the Paratill implement was used in 2001. The Rototill required much more power to operate than the Paratill. It is obvious that the tractor power requirements far exceed the actual energy expended in each tillage operation. Selecting the correct horsepower tractor for each tillage operation could reduce gross fuel usage. A much smaller horsepower tractor could be used for the shredding and root pulling compared to the Pegasus or Rototill implements. However, for this study it wasn’t feasible to select the correct horsepower tractor for each implement. Therefore, the larger horsepower tractor was equipped with all of the instrumentation for collecting the field data.
Shredding the cotton stalks required about 0.08 gallons of diesel per acre, which was the least compared to all other implements used in this study (Table 1). The Sundance root puller was slightly higher at about 0.13 gallons of diesel per acre. The amount of fuel used for the Sundance root puller was about one-half the amount of fuel used compared to the disking, listing and Sundance ripper/disk lister implements. The implements that required the most fuel per acre were the ripper, Sundance Level Bed Disc, Rototill, Paratill and Pegasus. The Conventional tillage implements didn’t require the most energy individually, however, collectively their sums exceed all of the minimum tillage systems. In fact, the minimum tillage systems evaluated required about one-half the amount of fuel compared to the total used by the Conventional system. The average 2001 net energy for both locations showed a trend of decreasing fuel consumption when decreasing the number of trips across the field as shown in Table 2.
Also associated with the savings in reduced fuel costs is the amount of time spent in the field with the tractor and operator. It wasn’t feasible to measure actual tillage rates in this study because of the small plots and time required at the turn rows. Tractor speeds were estimated from large field plots and discussions with tractor operators from active farms. Speeds can vary because of tractor power available, the hardness and roughness of the soil surface, the condition of the implement, size of the implement (4-row, 6-row, etc.) and other related factors. Table 3 shows the normal speed in miles per hour for each tillage implement, the implement widths used in this study and a calculated tillage rate in hours per acre. The calculated tillage rate assumes that no time is spent turning the tractor and implement around and all time is spent working the field. The total hours to prepare each acre for planting are also shown in Table 3. The Conventional tillage system requires almost twice the amount of time compared to the Sundance and Paratill tillage systems and 5 times more time compared to the Pegasus. There may be economic savings not only in fuel costs but also in operator time, depreciation and maintenance costs for the tractor and implements.
Cotton yields were recorded for each treatment and replication for each year of this study by collecting and measuring the seed cotton from 4 rows located in the center of each plot. The seed cotton yield was converted to bales per acre by using the ginning turnout of 35% for both test sites and a cotton bale weight of 495 pounds. The 2001 and 2002 cotton yields for Marana and Maricopa are presented in Table 4. There were no differences at the 5% significance level for treatments at either location; except for the 2000 Marana average yields. However, the differences were small and it is not clear that these differences were due to the tillage treatments. Typically cotton yields do vary between replications for the same treatments because of soil differences that can affect available moisture and fertility. The difference between locations can be attributed to the climatic environmental conditions.
Measurements of Total Suspended Particulate (TSP) were made the first year of this study for each implement and site, except for the Sundance ripper/disk lister implement. Manufacturing problems delayed delivery of a four-row unit so a two-row unit was substituted. It wasn’t possible to mount the dust monitoring system on the two-row unit so data couldn’t be obtained. The Sundance Level Bed Disc, an improved version of the Sundance ripper/disk lister, was available the second and third years and was used instead of the Sundance ripper/disk lister. Tables 5 and 6 present the 2000 TSP levels in micrograms/cubic meter for the Marana and Maricopa test sites, respectively. Also shown are the average TSP levels for each implement over all four sampling elevations. The TSP levels were summed for each sampling elevation and averaged to compare the total amount of particulates generated from each tillage system. Soil moisture is definitely related to how much fugitive dust is generated by those implements that are disturbing the soil. Soil samples were collected from each tillage treatment and the soil moisture was determined. The soil samples were collected from the top 6 inches, weighed, dried at 105oF for 24 hours and reweighed to determine the moisture loss. The soil moisture (SM) values based on a dry weight basis are presented in the last column of these Tables. The soil moisture was fairly consistent at each site varying from 17.3% to 18.6% at Marana to 12.3% to 15.4% are Maricopa.
The 2000 data showed some interesting results. The shredding operation generated very large TSP rates at the 24-inch elevation compared to other sampling heights or other implements. That was not unexpected since this implement chopped the cotton stalks into smaller pieces, and expelled them from under the implement. The data also suggest that these particles could have been greater than 10 microns since they weren’t measured at the higher elevations. The shredding operation accounted for 50 to 97% of the TSP values for all tillage systems at the 24-inch elevation. It became obvious after the first year that PM10 values needed to be collected in order to evaluate the tillage systems. The TSP rates can’t be used to determine if a tillage system would help improve air quality by reducing the amount of PM10 being put into the air.
The 2001 Marana and Maricopa PM10 rates in micrograms/cubic meter are presented in Tables 7 and 8, respectively. Also presented in these tables are the average and total PM10 rates and the soil moisture in the top 6 inches on a dry weight basis. The 2001 Marana soil moisture values were about 6% lower than the 2000 Marana values, except for the shredding and Pegasus implement operations. The soil moisture values for the 2001 Maricopa test site averaged about 6.5%, which were about one-half those of the Maricopa site in 2000. The average PM10 rates decreased as the number of field operations decreased. These trends show that the total PM10 rates for the Conventional tillage system was always the highest and the Pegasus system was always the least. The PM10 rates for each implement and elevation were more consistent compared to the 2000 TSP rates for both test sites.
Tables 9 and 10 present the 2002 Marana and Maricopa PM10 rates in micrograms/cubic meter, the total and average values and the percent soil moisture on a dry weight basis. The shredding and Paratill tillage operations were conducted in Marana when the soil moisture was only about 4%. However, a rain occurred and returned the soil moisture to similar levels as the 2001 season. The 2002 Maricopa soil moisture levels were very similar to (2% less) the 2001 soil moisture levels. Again the average PM10 rates decreased with a decreasing number of field operations, similar to the 2001 results. If the average PM10 levels for each tillage system at each location for years 2001 and 2002 are grouped, a simple analysis of variance can be performed to test the hypothesis that the means from each location and year are equal (drawn from populations with the same mean). The results indicate that at a 5% significant level, related to the probability of having a type I error (rejecting a true hypothesis), that the means of the samples are not the same. Thus, the PM10 production is different for the four tillage systems.
A dust-sampling unit was designed and constructed that could be easily attached to common mounting brackets on each individual tillage implement used in this study. This design allowed the dust sampling unit to maintain the same sampling grid for each implement as it was moved between each implement to collect the fugitive dust information. This unit also allowed the grid and sampling elevations to remain constant at each test site and for each year.
The dust results collected the first year of this study can’t be used to compare with the EPA standards for PM10. However, the second and third year do provide results at four elevations that show the rates in micrograms/cubic meter. These values can be compared directly with the EPA PM10 standard reading of 150 micrograms/cubic meter measured as a daily concentration. Comparing the average PM10 values for all four elevations (Tables 7-10) shows that in 2001 there were several implements that exceeded the 150 micrograms/cubic meter EPA PM10 limit. In general they were shredding, listing, root pulling and Level Bed Disc implements. In addition disk, ripper and Pegasus nearly reached this limit. In 2002 only the Level Bed Disc at Maricopa exceeded the 150 micrograms/cubic meter limit. Most of the remaining implements in 2002 produced only 100 micrograms/cubic meter or less. This does indicate that these agricultural implements could potentially exceed the EPA PM10 daily concentration, which might result in a violation.
It is also obvious that the average PM10 rates decreased as the number of field operations decreased. The total PM10 rates for the Conventional tillage system were always the highest and the Pegasus system were always the least. Increasing the number of tillage operations has the potential of creating more fugitive dust, which could lead to violations of the EPA standard for PM10 air quality. In addition to reducing the number of field operations and the total PM10 rates, a grower could also save time and fuel by reducing the number of field operations. The Pegasus, a single-pass implement, can reduce the time spent per acre by a factor of 5 times compared to the Conventional tillage system. The economic savings could include reduced equipment operator man-hours, reduced maintenance and depreciation of the tractor and equipment, while reducing fuel consumption to just one-fourth that of the Conventional tillage system. No differences in cotton lint yields were measured for the Conventional, Sundance, Rototill/Paratill or Pegasus tillage systems.
Educational & Outreach Activities
- “Cotton Tillage Practices and Fugitive Dust”, Frank E. Eaton, University of Arizona 2000 Cotton Field Day, Maricopa, Arizona, October 11, 2000.
“Minimum Tillage Practices and Equipment Demonstration”, Frank E. Eaton, University of Arizona 2001 Cotton Field Day, Maricopa, Arizona, October 10, 2001
“Cotton Tillage Systems and Equipment Demonstration”, Frank E. Eaton, University of Arizona 2002 Cotton Field Day, Maricopa, Arizona, October 9, 2002.
“Cotton Tillage Implements and Study”, Frank E. Eaton, University of Arizona Marana 2nd Annual Cotton Field Day, Marana, Arizona, October 3,2000.
“Results of the Cotton Tillage Study”, Frank E. Eaton, University of Arizona Marana 3rd Annual Cotton Field Day, Marana, Arizona, October 4, 2001.
“Results of 2001 Marana Agriculture Center Minimum Tillage Field Experiments”, Frank E. Eaton, University of Arizona Marana 4th Annual Cotton Field Day, Marana, Arizona, October 2, 2002.
Cotton Tillage Systems: PM10 Production and Energy Consumption, F.E. Eaton, J.L. Walworth, R.L. Roth, S. Husman, and P. Wilson. Poster presented at University of Arizona Cotton Field Days.
“UA Researcher Seeks Methods to Control Dust, Gain Efficiency”. 2001. Written by Alan Lavine. Casa Grande Dispatch, Tri-Valley Edition. October 10 & 11. Cotton & Ag. Section. Casa Grande Arizona. Pages 2B, 6B, and 7B.
“Cleaner Air in the Desert”. January 2003. Written by Joanne Littlefield. 2002 Arizona Agricultural Experiment Station Research Report. College of Agriculture and Life Sciences. The University of Arizona. Tucson, Arizona. Pp. 17.
- Eaton, F.E., R.L. Roth, J. Walworth, P.N. Wilson, and S.H. Husman. 2002. Computer Manipulation of Modeling Variables in Dust Particulates Study. Pp. 463-469 In: Proceedings of the World Congress of Computers in Agriculture and Natural Resources (13-15 March, 2002, Iguacu Falls, Brazil) eds. F.S. Zazueta and J. Xin., Pub. Date 13 March 2002. ASAE Pub. No. 701P0301.
Kennedy, Ana and Paul N. Wilson. 2003. Reduced tillage as a Potential Economic Response to Clean Air Regulations. Presented at the annual meeting of Western Agricultural Economics Association and the Western Economic Association International. July 15, 2003. Denver, Colorado.
Kennedy, A.M., P.N. Wilson, E. Eaton, S. Husman, J. Walworth, and R. Roth. 2003. “The Economics of Single Pass Multiple Operation Equipment for Dust Mitigation”. Unpublished Manuscript, Department of Agricultural and Resource Economics, University of Arizona, Tucson.
Kennedy, A.M. 2003. The economics of dust mitigation in tillage operations. Master Thesis. Department of Agricultural and Resource Economics. University of Arizona, Tucson.
The economic analysis of reduced tillage systems involved two types of evaluation: (1) a partial budget analysis of the profitability of the three reduced tillage systems relative to Conventional tillage and (2) a survey-based study of the willingness to adopt (WTA) reduced tillage systems in central Arizona agriculture. The current section of the final report will summarize the findings from the partial budget analysis while the “Farmer Adoption” section will provide the highlights of the WTA evaluation.
The Sundance, Paratill and Pegasus systems (henceforth called single-pass multiple operation equipment, or SPMOE) can either replace the entire Conventional tillage system or reduce the number of operations through the use of this alternative equipment. Table 11 presents the nature of the SPMOE systems and the Conventional tillage system used in this analysis. These systems have been discussed earlier in this report.
Partial budgets were constructed to evaluate the change in net returns of switching from the Conventional system to one of the three SPMOE systems. Partial budgeting is an analytical tool that compares the costs and returns of alternative plans or proposed changes to an ongoing business that do not impact the total business operation. Only those costs and returns associated with the proposed change are included in the partial budget model. The prediction accuracy of partial budgets depends on the accuracy of the information and estimates they contain. The partial budgets constructed for this study were not for a particular farm so the cost information was drawn from cotton crop budgets and machinery cost estimates prepared by the University of Arizona. Detailed descriptions of the assumptions utilized in this analysis are contained in Kennedy (2003).
Partial budget models have four sections: (1) added annual returns, (2) reduced annual costs, (3) added annual costs and (4) reduced annual returns. Added annual returns include any additional income realized due to switching from the Conventional tillage system to any one of the SPMOE systems. The existing literature on reduced tillage systems and their impact on cotton yields, and the limited yield information from this present study, do not indicate any significant increase or decrease in yields over the Conventional system. As a result, it was assumed that no added income would result from the adoption of SPMOE.
The reduced annual costs are the variable and fixed costs of the Conventional tillage system eliminated with the adoption of SPMOE. The added cost component of the partial budget model has two components: (1) investment/ownership costs and (2) added annual operating costs of SPMOE systems. The final component of the partial budget model is reduced annual income. As noted above, no statistically significant change, either positive or negative, in cotton income is associated with the adoption of SPMOE.
The change in net income and the return on investment (ROI) are measures used to evaluate the economic performance of the SPMOE systems. Changes in net income are reported in two manners: profitability and cash flow. Profitability is the change in net income that includes both cash and non-cash expenses. The cash flow computation includes only cash expenses. ROI is calculated by dividing the change in net returns (profitability) by the total added SPMOE investment.
Due to growers’ concern regarding yield reductions with SPMOE, a sensitivity analysis was prepared under the condition of small reductions in yield. An additional sensitivity analysis was conducted to determine the impact of a state tax credit on the economics of a SPMOE investment. Arizona tax law (ARS §§ 43-1081 and 43-1170) provides a tax credit for equipment used to control or prevent pollution. The amount of the tax credit is 10% of the purchase price and is applied to an individual’s or business’ state income tax return.
The results of the partial budget analysis indicate that replacing the Conventional tillage system with any of the three SPMOE systems result in an increase in net income (Table 12). The Pegasus system had the highest profitability under our assumptions, followed by the Sundance and the Paratill systems. Replacing the Conventional system with the Pegasus system produced an additional $25.91 per acre while the Sundance and Paratill systems increased income by $16.76 and $16.06 per acre. The ROI on all three systems is high: 115%, 76% and 70% for the Paratill, Pegasus and Sundance systems, respectively.
These results, however, may be overstated because the SPMOE systems do not adequately replace the ripping operation of the Conventional system. Only the Paratill system partially approximates the conventional ripping operation. Therefore, the results were re-estimated for the SPMOE assuming the new system did not replace the Conventional ripping operation. These modified results also are reported in Table 12 and show a substantial decline in all three economic measures. Nevertheless, the profitability, repayment capacity and ROI all remain positive.
Given that the Pegasus system ranked highest in profitability and repayment capacity, we would predict that cotton growers would favor this system. However, despite the estimated $25.91 per acre increase in net income, the Pegasus has not been widely adopted in Arizona. Several reasons explain this economic decision not to adopt. First, the Pegasus system does not perform well in all soil types and moisture conditions. In soils with high clay content the Pegasus can leave large dirt clods that must be worked with additional tillage operations. Secondly, the implement does not destroy voluntary cotton stalks and struggles to stay within the track of existing cotton beds. Failure to destroy all cotton stalks and root systems can place a grower in violation of cotton plow-down regulations.
The perceived reduction in cotton yields by growers also may contribute to why SPMOE has not been adopted in general. The sensitivity analysis of a 5% reduction in yield reveals a significant negative change in net income (-$20.36 for the Pegasus, -$29.51 for the Sundance system, and -$30.21 for the Paratill). This risk-return tradeoff with SPMOE may determine growers’ willingness to adopt reduced tillage systems.
A mailing list of all cotton growers in Maricopa and Pinal Counties was assembled from pesticide permit records from the Arizona Department of Agriculture (ADA) and cross-referenced with extension mailing lists utilized by the cotton specialists in each county. Growers in Pinal County were included in the surveyed group because part of the Maricopa non-attainment area extends into Pinal County and growers in this largely agricultural county south of Phoenix would act as a control group for the adoption decisions. Two hundred and thirty seven growers were surveyed: 119 in Maricopa County and 118 in Pinal County.
A mail questionnaire was developed to gather information on current farming operations, perceptions of dust as an environmental problem, reduced current tillage operations (RCTO), estimated benefits of RCTO and SPMOE, the willingness to adopt SPMOE and basic socioeconomic information including the distance of the center of the farm to the edge of the nearest residential housing area. The design of the WTA questions followed the general recommendations for contingent valuation studies. The questionnaire was reviewed several times by the Pinal County cotton specialist and pre-tested with four cotton growers, two from each county.
The mail survey was implemented following the Dillman’s total design method. The research team and the appropriate cotton specialist signed each cover letter. Follow-up procedures included a reminder postcard and a second mailing of the survey packet. One hundred and twenty eight respondents (54%) returned their “completed” questionnaires. Thirty-three growers responded that they were no longer farming cotton. Twenty additional responses were eliminated from the data set due to incomplete or inconsistent responses to specific questions.
Sixty-two percent of the responding cotton growers acknowledged that dust generated by tillage operations can create an environmental nuisance. A majority of these respondents, however, believed this nuisance was not a severe problem. Only seven percent of the growers labeled the problem as very severe or severe. Thirty-six percent of all respondents classified their tillage operations as not creating a dust problem while the remaining three percent did not know if tillage activities created an environmental nuisance or not.
Table 13 summarizes the basic information about the respondents based on their adoption decisions. Among the adopters of some form of reduced tillage, two growers reported adopting SPMOE only, 22 adopted RCTO only and 28 reported adopting both SPMOE and RCTO. RCTO adopters reported eliminating some disking, ripping and listing operations. One or more disking operations were eliminated by 84 percent of the RCTO adopters, and 52 percent reported eliminating ripping their fields. Thirty-two percent of the RCTO adopters eliminated one listing operation. Smaller operations and older operators tended to rely on RCTO only to compete in the current economic and regulatory environment. Only education differentiates the adopter and non-adopter groups with adopters having a statistically significant higher level of post-secondary education. The two reasons most commonly chosen for both SPMOE and RCTO adopters were “to reduce costs” and “to reduce cost and dust.” The two sole SPMOE adopters indicated that reducing costs was their only reason for adopting these tillage systems. None of the survey respondents chose “to reduce dust” as a reason for adopting reduced tillage systems.
The growers who adopted SPMOE and/or RCTO reported their estimated realized long-term net benefits per acre associated with their decision. In addition, the respondents estimated the percent of their cotton acreage on which the reduced tillage systems were used. Growers who had not adopted SPMOE were asked to estimate the expected long-term net benefit per acre that would induce them to adopt any of these three reduced tillage systems. These growers also were asked to estimate the expected percent of their cotton acreage they would farm with this technology.
All three estimates (ex post SPMOE, ex post RCTO, ex ante SPMOE) were analyzed using a semi-log function to evaluate the relationship between technology benefits and the percentage of cotton acreage on which the technology was or would be applied. In this formulation the slope coefficient of the explanatory variable, net benefits, measures the absolute change in the percentage of cotton acreage relative to a change in net benefits. So a one percent increase in net benefits associated with SPMOE produces a 0.0028 increase in the percentage of cotton acreage using SPMOE. Willingness to adopt “elasticity” for SPMOE exceeds the realized measures of adoption elasticity for either SPMOE or RCTO. This result is analogous to contingent valuation studies that have shown the tendency of respondents to overestimate willingness to pay measures relative to actual financial payments.
Figure 2 illustrates the estimated adoption functions. The expected SPMOE adoption function exceeds the two realized functions over most estimates of long-term net benefits. Acreage on which RCTO is applied exceeds the SPMOE acreage for the same level of new benefits. We hypothesize that this realized difference is due to the uncertainty associated with the new technology. Equivalent cost savings for the risk averse grower can be obtained from reducing conventional practices as opposed to adopting an unfamiliar new technology requiring a learning period.
The survey results also show that RCTO have been adopted widely by cotton growers to reduce costs, sometimes in combination with SPMOE. Any dust reduction benefits emerging from fewer tillage operations become an added bonus to the grower and society. SPMOE alone does not represent a panacea to the agricultural dust problems in the Phoenix non-attainment area however. Only two growers solely adopted SPMOE to reduce costs and control dust. At estimated long-term net benefits of $12-26 per acre, our analysis indicates that SPMOE alone may be adopted on 10-35 percent of the cotton acreage in Maricopa and Pinal counties. However, a combination of RCTO and SPMOE has the potential to significantly reduce costs and reduce dust emissions on a significant percentage of acreage in both counties.
Areas needing additional study
Very little information has been collected relative to the fugitive dust generated by different agricultural implements. The dust-sampling unit developed in this study could be adapted to other agricultural implements and tillage systems that would be used in different crop production systems. This unit could also be used to collect fugitive dust from other areas that generate fugitive dust, such as dirt roads, construction sites, non agriculture soils, urban areas etc. It would also be easy to collect data relative to the PM2.5 standards that might cause even greater health risks.
Current cotton research projects are evaluating the advantages of planting twin-line cotton rather than the standard single row cotton on 40-inch beds. Twin-line cotton is two cotton rows spaced 7 inches apart on the same 40-inch bed. The data to date from twin-line cotton research sometimes show a slight yield increase but do show improved cotton quality by reducing the fiber microaire values. The minimum tillage systems used in this study wouldn’t work on twin-line cotton systems without additional modification. More research would have to be completed to determine how the minimum tillage systems could be modified.
The agricultural implements used in this study provided new insights to other machinery that might be more efficient in preparing the soil for planting, reducing the number of field passes and easier to operate. The Pegasus implement can prepare the cotton beds if the soil conditions are correct, the implement is properly set for the field and it can track the cotton rows. Most growers have difficulty operating this implement. Precision autopilot tractor guidance systems are now available that use GPS technology that can auto-steer a tractor to sub-inch accuracy. This technology wasn’t available a few years ago. Growers are adapting this technology and using it for listing, planting, cultivating and harvesting of crops. This system will help growers to field set the Pegasus so that it will track the cotton stalk rows.
It also was apparent that the Sundance Level Bed Disc did a very good job in covering the cotton residues compared to the Paratill minimum tillage system. However, it was designed to operate in drip-irrigated fields where the beds can’t be ripped because the drip tape is buried there. Instead, the Sundance Level Bed Disc rips the furrow bottom. It is suggested that the furrow bottom rippers be replaced with the Bigham Paratill bent leg rippers that would be mounted to rip the bed center. This would reduce the soil compaction in the bed and help lift any remaining shredded cotton stalks so that the bed discs could incorporate the residue into the bed. The result would be a cotton bed that would have less surface trash allowing the planters to operate more efficiently.
It is felt that more data are needed from different soil types and moistures so that models could be developed to predict the amount of PM10 material that would be generated by different tillage systems. Thus, growers would be able to schedule their field operations to minimize the amount of particulates that would be generated from each implement.