Technical Summary:

This proposal is a research to develop and validate the methodology of a very high speed (sub seconds) automated soil sampling system incorporating robotics for the collection of soil samples. This research will use nematode soil sampling as a model system.  

Precision agriculture for nematode management will depend on numerous sample collections through out the growing season.  Intensive soil sampling is required to accurately detect these pests in that specific symptoms are not always associated with nematodes. This is labor intense, costly and often times limited by the stage of plant growth and existing weather conditions. A fast, mechanized robotic system would help both research and agribusiness alike. This type of system would reduce the labor required to collect samples necessary for precision agriculture, enable all-weather and all-plant growth stage sampling, and enable agribusinesses to enhance their capabilities to a point where timing for collection of soil samples can be the primary driving factor rather than the availability of personnel to favorable weather conditions. This would improve the efficiency and accuracy of ground truthing remote sensing data across all areas of pest control. 

Objectives:

1. Develop a novel very high speed mechanized soil sampling system incorporating the use of robotics.
    

2. Evaluate the accuracy of the system using the temporal grid sampling method required in nematology as a model system under different soils types and moisture conditions.

Procedures:

Objective 1: This objective will be done in two sequential stages: the development of the core mechanical components, and the incorporation of robotics into the automation. Stage1: Three core components of the automated soil sampler will be developed in-house: a pneumatically powered probe injector mechanism that will perform the insertion of probes into the ground; a probe extractor unit that will extract probes from the ground; and an electromechanical probe cleaner/soil collector that will collect soil samples into containers. 

The system will be configured in a metal framework that will initially be mounted on a miniature trailer for initial field testing.

Objective 2: Cotton fields naturally infested with either the reniform or root-knot nematodes will be selected for the study. Large plots will be established by dividing the field into four 10 acre blocks. Each block will be mapped and georeferenced on 1 and 1/2, acre grids. Georeference points will be used to direct the robotic sampler across the test plots. Nematode samples will be collected at each grid intersect using the high speed automated sampler and manually with an Oakfield soil sample tube. Each sampler will collect four, 2.5-cm diameter X 20-cm deep soil cores from an area within 1 foot of each reference point. Nematodes will be separated from the soil using the gravity screening and sucrose centrifugation techniques for both methods. Data will be subjected to analysis for spatial dependence and for the sampling design using the appropriate analysis. Samples will be collected at the time of planting without any plant canopy and again at mid-season with a full plant canopy. Soil separates and moisture will be recorded. Results from this part of the study will validate the accuracy of the automated system to current soil sample practices. 

Potential Pitfalls

The proposed project has been designed as a multi-phase project, and this is the first phase. The technical detail of the system is straight forward, but a potential pitfall that we can envision at this time is the safety factor. Because of its high speed, many provisions must be developed to eliminate the danger to operator. As in any systems there will be tradeoffs that must be made for the system to be economical.

The timing of the project and the timing of growing season may not match perfectly to allow orderly sequence of project progression. It is likely that parts of the first year’s tests will be done at the incremental completion times of the design and construction phase of the system, and it is possible that some of the data collected will consist of only soil volumetric information for comparison between the manual and automated methods. It is possible that the actually nematology soil sampling will fall into the next project year. 

Limitations to the Procedures

Although the technology is applicable to general soil sampling, the procedures are designed around the topic of nematology, which require the hardware design be narrowed and somewhat tailored to do soil sampling for nematodes.
 

Justification:

Soil Sampling is Critical, Faster-Cheaper-Better Soil Sampling is Needed

Soil sampling is necessary regardless of the type of crops. It is a means by which the needs of the crop are assessed, and cotton is the commodity of focus in this collaborative project. Cotton is a critical cash crop to Mississippi economy producing more than 1.81 bales per acre on over 1.1 million acres in 1999. Cotton production is highly dependent on pesticides including herbicides, insecticides, and nematicides to produce economical yields.  Producers have taken advantage of new technologies and have incorporated transgenic varieties into their production practices.  But plant-parasitic nematodes are still a major economic concern to cotton production.

In 1998, the estimated yield loss from nematodes in Mississippi was estimated to be about $27,209,130 based on a price point of $390/bale, and it represented a 5% yield loss. The annual average loss of yield due to nematode damage from 1992 to 1998 was about 4.14% (Kirkpatrik 2002). It has been estimated that the reniform nematode will become the most serious nematode threat to cotton in the southeast U. S.  There are currently no commercially available cotton cultivars with resistance to this nematode.  Therefore the application of chemical nematicides are still the most frequently utilized method of management.  These chemicals may provide economic nematode control during a single season but they do not provide long term management and must be used annually.   

In order to optimize the use of nematicides, a site specific technology is desired, where the spatial distribution of the in-field nematode population levels can be determined and a proper prescription can be administered. The population level of nematodes in a field is determined by taking timely soil samples. The density and frequency of the sampling is actually determined by necessity and to some extent, cost. Site specific nematodes population determination is done to identify regions of the field which is experiencing nematodes stress. The ideal time for taking soil samples for nematodes population determination is currently not known because the amount of resources required to gain this knowledge is significant. Repeated soil sampling over a season is required and depending on the volume of samples taken, analysis of the samples is costly. Depending on the region, the cost for analyzing soil samples for nematodes ranges from $6 to $25 per sample. 

The proposed project will have the potential of producing a very high speed automated soil sampling system and methodology that will positively impact the timing of soil samples taking, and potentially it can lower the cost of taking soil samples. The term "very high speed" is used here to represent a sub-seconds speed of injecting probes into the ground, a potential speed of injecting and extracting probes at an order of magnitude faster than any existing soil sampler in operation to date. The longer-term objectives of the proposed project are to include an onboard partial processing of soil samples that will take some measurements and will reduce the number of steps that must be performed in the laboratory. 

The proposed project, if developed and validated, will help research and agribusinesses in the following ways: 

·        Help identify the time at which nematodes population level will begin to cause yield loses, thus facilitating the identification of optimal time for taking soil samples in a season. The high speed way of taking soil samples and the automated nature of the methodology will enable frequent soil samples taking in all-weather conditions (possibly autonomous if we get to the next phase).

·        Help reduce the amount of labor required to take a large number of soil samples. The insertion, extraction of probes, and the dispensing of soil samples from the probes con containers will be automated thus minimizing the need for manual intervention. Under certain conditions, multiple systems could be operated autonomously on the field without the need of additional operators. In addition, if made autonomous, weather will become a lesser factor in preventing soil sampling from being scheduled.

·        Enhance or enable agribusinesses to improve their capabilities to a point where timing for taking soil samples can be made the primary driving factor rather than the availability of personnel or the favorability of weather condition. This is probably the most important justification for this project. If the technology is developed and validated, it can reduce the cost of taking soil samples, and it has the potential of developing a more robust soil sampling service businesses for a variety of applications.

·        Help improve the efficiency and accuracy of validating remote sensing nematode reflectance studies. Remote sensing holds the promise of nematodes sensing in a larger scale. Extensive ground truth data must be obtained to enable scientists to properly correlate reflectance data to nematodes population levels. The proposed project can help in reducing the cost of taking ground truth data. 

Literature Review:

Remote Sensing, Geographical Information System (GIS), Global Positioning (GPS), and Variable Rate Technology (VRT) are becoming an integral part of many agricultural practices today. They are an integral part of farm management practices: they are used to help making the farms more profitable. Soil types and surface conditions (Everitt et al., 1989, 1994), insect infestations  (Everitt et al., 1994), weed infestations (Nixon et al., 1985), and  virus disease development (Nelson, 1994) are some of the aspects that are closely monitored.   The use of new technologies in nematology has been limited to infrared remote sensing and they are used to detect plant injury due to nematodes in cotton (Heald et al., 1972). This fact highlights the need for bringing technologies into this aspect of agriculture.  It was reported that yield losses in cotton due to nematode damage in Mississippi were in an average of 4.14% per year from 1992 to 1998, which translates to an average of 22 million dollars per year (Kirkpatrik, 2002).

Soil sampling has been a means by which the level of nematode stress is determined. The necessity of soil sampling in agriculture setting is well documented (Crouse, Keith, 1994), and the cost of soil sampling varies with the geographical region.

Our investigation into currently available technology for soil sampling shows that manual soil sampling is still the predominant method of soil sampling, although most are augmented with the use of some kind of motor vehicle. There are many different brands and types of soil samplers currently available.  The most commonly marketed soil samplers are core-cutting samplers.  According to C. A. Shapiro, core cutting samplers take into consideration the stratification of nutrients and banded fertilizer applications (K. A. Janssen et al, 1998).  Companies such as Case International, Geoprobe Systems, Concord Environmental Equipment, and En Core market core cutting samplers. The fastest coring device is the Case IH AFS Soil Sampler, which takes a sample in twelve seconds (Case, 2000).  

References:
 

Case International Harvester Website. 11 Aug. 2000 http://www.casecorp.com/.soil 

Everitt, J. H. , D. E. Escobar, M. A. Alaniz and M. R. Davis. 1989. Using multi spectral video imagery for detecting soil surface conditions.  Photogrammetric Engineering and Remote Sensing 55(4):467-471. 

Everitt, J. H. , D. E. Escobar, K. R. Sunny and M. R. Davis. 1994.  Using airborne video global positioning systems and geographical information system technology for detecting and mapping citrus blackfly infestations. Southwestern Entomologist Vol. 19 No. 2: 129-138. 

Heald, C. M.., W. H. Thames and C. L. Wiegand. 1972. Detection of Rotylenchulus reniformis Infestations by Aerial Infrared Photography.  Journal of Nematology, Vol. 4, No. 4:298-300. 

Janssen, K. A., J.L. Havlin, and E. H. Horstick. “Hydraulic Device for Transect Sampling Soil.”, Soil Science Society American Journal. Mar.-Apr. 1998: 480-6

Keith Crouse, Soil Testing for Farmers, Mississippi State University Extension Service, Information Sheet 346, 1994. 

Kirkpatrick, Terry, Nematode Survey and Education Program, The Cotton Foundation, March 19, 2002 (http://www.cotton.org/CF/nematodes/survey-3.cfm). 

Nixon, P. R., D. E. Escobar and R. M. Menges. 1985. A  mulitband video system for quick assessment of vegetal condition and discrimination of plant species.  Remote Sens. Environ. 17:203-208 

Current Research:

This project is spawned from an exploratory project performed by a group of senior students at the Biological Engineering Department in 2001. The project produced a one-shot handheld probe injector using a miniature probes (rifle bullet shells) to prove the concept.

Back to Top

Back to Menu