Gillins, Elizabeth

							                              Gillins, Elizabeth
     Reducing the Cost of Soil Mapping In Precision Agriculture
                      Faculty Mentor: Ruth Kerry, Geography Department

Using electrical conductivity data (ECa) in soil survey is related to several soil properties.
Particular use has been made of ECa data in precision agriculture because it is related to
properties that could affect crop yield. Although theoretically sound reasons exist to explain
these relationships, they vary spatially, making ECa data interpretation difficult, even within
fields.

Some soil scientists have attempted to establish relationships with ECa for which there is no
theoretical basis. McBratney et al. (2005) suggest that theoretically sound relationships
developed between high frequency devices can be used to inform interpretation of the lower
frequency ECa data. A three part theory was presented:

           1. If the soil is hyper-electrolytic, ECa measures salinity.
           2. If the profile thickness is thinner than the effective depth of measurement and the
              underlying material has a smaller ECa, ECa indicates soil depth.
           3. If there is no compaction and the electrolyte concentration balances with soil
              charge, ECa represents variations in clay and moisture content.

Our aim was to determine whether the following model provides insights into the causes of
patterns of ECa in several fields with soil likely to meet the different conditions mentioned above.

Data from several fields in the United Kingdom collected by Dr. Kerry were statistically
analyzed to test parts two and three of the theory above and a local field with salinity problems
was sought to test part one of the theory.

A suitable testing field site located in Lakeshore, Utah, was identified through contacts in the
Brigham Young University Plant and Animal Sciences Department and at Utah State University
Soil Extension Office. The use of the field and equipment was coordinated with the help of a
professor in the Extension Office.

One hundred stakes were placed 15 meters apart across the field using a GPS unit. Each stake
marked a collection point for a soil sample. At each point, six samples from within one meter of
the stake were collected and mixed in a bucket. The mixed sample was placed in a bag and
marked according to location in the field.

Each soil sample was tested for pH and electrical conductivity in the soil lab in the Plant and
Animal Sciences Department at BYU. The texture of the soil samples was determined in
connection with a course Dr. Kerry taught in fall semester. Electrical conductivity data and soil
sample locations were collected using the EM38 instrument connected to a GPS.

For each data point where soil had been collected, the lab results were used to make a
determination of which the condition of the theory above was met. Moving correlation analysis
was then used to determine the relationship between ECa and relevant soil properties in the
vicinity of each data point.

Each step in this process was a challenging experience. I made several trips to the field site only
to find the site flooded with irrigation water. Precise conditions were required to collect the
samples. I learned how to use the electrical conductivity equipment and GPS unit after several
attempts to complete the data collection. Lab work took several weeks to finish by myself and I
learned the necessity of careful, accurate analysis.

Results showed that where a condition from the theory was met, there was a strong relationship
between ECa and the relevant soil property mentioned in the theory. However, within most fields
different conditions from the model apply in different sections of the field. Therefore, assuming
that a site is salty or has shallow soil etc. will not give a true interpretation of what EC a is
measuring in the entire field.

The cost of soil mapping cannot be reduced based on the model proposed by McBratney et al.
(2005) alone because of the inconsistencies in which soil properties are correlated with ECa data
within a field. More cost could be incurred by incorrectly mapping a field based on the model. A
promising approach seems to be that the ECa data can be used to divide the field into regions
with similar ECa values. Six soil samples can then be collected in each region and be used to
determine which conditions of the model are met in that region. However, this needs some
further investigation.

The preliminary results of this research were presented by myself at the Western Society of Soil
Science Regional Meeting in Park City, Utah in June 2006. I received an award for my
presentation and was the only undergraduate among graduate students and professors presenting
at the conference. This research project will soon be fully written up as a journal article and
submitted to the Precision Agriculture Journal where I will be listed as a co-author.

My personal growth from this project has been educational both academically and personally.
Previous to this project, I had little experience with soil science. However, at the research
conference, I discovered I understood the research other soil scientists were presenting. Our
professorial contact at Utah State University was impressed at my progress in soil science and
even offered me a stipend-paid Master’s student slot at USU.

BYU received positive exposure from my participation in the conference. Many conference
attendees were unfamiliar with the university, while several other scientists were familiar and
quite impressed with the research presented. The ORCA experience was positive for me also. I
grew and was stretched as I prepared and presented the research, then interacted with
professionals in the field in an academic setting.

Reference
McBratney, A. B., Minasny, B. & Whelan, B. M. (2005) Obtaining ‘useful’ high-resolution soil
data from proximally-sensed electrical conductivity/resistivity (PSEC/R) surveys. In: J. V.
Stafford ed. Precision Agriculture ’05. p. 503-511. Wageningen Academic Publishers,
Wageningen, The Netherlands.