HOME The Trevey Home Table of Contents The Finder Glossary

Soil Sampling

From
The Trevey, A Dynamic Planning System for Rural Areas

Rural System staff seeks representative samples of soils that are highly varied, a difficult and costly task. We are engaged in decision making and, though complex, it must be done cost effectively. We observe and we adjust. Some people call our work an aspect of "adaptive management."

We use expert-informed, constrained, area-proportional sampling. If we are to analyze a 600 acre tract, we develop a GIS map and get data in amounts such as from the central table below:

Field Form Sample # and name GPS location Time/date Observer Level 1,2,3,4 Thickness of level 1,2, and 3(measured. cm.) Likely parent type (granite, sandstone, shale) Thickness of Level 3 / level 2 Thickness of Level 1/ level 2

Elements of a javascript unit to present the number of samples needed Fixed budget amount Price of analysis per sample Cost of taking one sample Costs of collection, supplies, bags, boxes, labels You need xxx samples (pooled for a sample or not) per stand or major area. Number of stands or major areas You need at least ___samples at a cost of ___ not including analyses and writing of reports on the reported findings.

	Acres	Percent
Type A	300	50
Type B	200	40
Type C	100	10

Based on a fixed budget of $240 for soil sample analyses, at $8.00 per sample, then 30 samples may be taken. Of these, 15 are taken in A, 12 in B, and 0nly 3 in Type C. These are separate samples; conventional statistical analyses can be performed on them; they are not pooled as done for average fertilizer applications to farm fields.

Each is given a GPS location and used as a GIS training site, i.e., to map areas with the same characteristics of slope, aspect, elevation, Landsat vegetative coverage, terrain position, and parent material.

Each sample is "representative" of the site. All sites are considered "non-uniform." The sampler is an expert, attempting to develop a best-possible description of a unit of land such as a forest stand or pasture. Unusual conditions are excluded; landslides, soil slumps, and over-turned tree roots are not sampled; rare sites are excluded; uniform conditions are accounted; boundary conditions (roads, etc.) are avoided. Points are at least 3 meters apart. "We want a good median estimate of the stand or land unit soil factors.We want the median and high and low values."

We hold that experts can provide that better and more cost effectively with guided representative sampling than by random sampling. Tests of this hypothesis are underway, but the premise is used untildisprovenn.

How representative is each sample remains our question. A sample may represent texture very well, pH poorly. De we compute sample size based on the factor with highest variance?

We have computed sample sizes and costs based on perceived variance and acceptable levels of accuracy and precision.

A "sample" is assumed to be 100 grams from layer #2 and the recorded thickness of layer 1 and 2. An estimate of the thickness of layer 3 and the type of layer 4 are requested, but not essential.

Samples are labeled, placed in a plastic bag, and the bag placed in a padded rigid box in a knapsack to avoid crushing or introducing limestone particles into the fine-earth fraction.

We make a preliminary or before-treatment baseline survey. Subsequent samples are compared (usually with an un-paired t test) to the base and to previous and adjacent samples to note significant change. In part, we are protecting us and the owners against "degradation" claims; monitoring to assure sufficient conditions for positive net primary production (growth).In part we are developing the data base that assures everyone (clients and observers) that we are producing on our areas no non-point pollution, action that will violate the Clean Water Act. The sampling procedures developed assures them that on the area the best management practices (BMPs) are working, or that, if not, what adjustments will be needed to accommodate unique site conditions.

We do not attempt to return to the exact sample sites. The costs are too great, variations too high, and smaple sizes required too great to justify such a procedure.

The usual level of confidence used for analyses is an alpha of 0.15.

We seek to create a full-scale, cost-effective analysis of soils within units where plant production is likely to be feasible. We characterize other sites (slumps, stream banks, tallus) in other ways, avoiding the costs of one type of analysis and commenting on the other potentials of a site with area unlikely for wood or forage production for profit in over 150 years.

Using models, strategic double sampling, and historical data, we provide a comprehensive analysis and description of the soils, the surface lithosphere of each stand or land unit, and a summary area analysis. Projections based on global warming and acid rain scenarios are also provided.

Probable soil productivity for trees forage biomass and water retention over the next 5 years is estimated. No extra comments are made about animal life because it is so positively correlated with thickness of layers 2 and 3.

Forms being developed

Notes 2007

Soil pH, Drought and Lime Requirements Fall 2007

Mark Alley and Rory Maguire Crop & Soil Environmental Sciences Dept. VA Tech Blacksburg, VA

Most of Virginia continues to experience drought conditions, and as corn, soybean and forage harvests are completed, growers are beginning to take soil samples. Dry soil conditions for extended periods of time can affect soil pH and associated lime recommendations, but dry soil conditions can also influence the quality of soil samples.

When soils are very dry and hard, it is difficult to take quality soil samples simply because the soil is too hard to push the soil probes to the needed depth. Samples that are taken from too shallow depths and/or at different depths at different locations in the field can result in non-representative samples. Such samples do not provide an accurate representation of the soil chemical status of the field. High quality soil samples require that each core of soil taken within a field be the same depth. In particular, very shallow samples, i.e. 1 inch, can show very low pH values due to surface acidity that has developed from surface fertilizer applications and decomposition of crop residues, especially in no-till. If the soils are very dry, a rubber mallet can be used to drive the soil probe to the required depth necessary to obtain a representative soil sample.

Soil pH levels following extended drought can also be influenced by the "salt effect." The "salt effect" results from the lack of dilution and leaching of fertilizer salts and organic matter decomposition products. Soil pH values may be as much as 0.5 units lower than when soils have received more average rainfall during the growing season. If only the soil pH value is considered in these situations, higher than needed lime recommendations may result. Growers and advisers should be certain to consider the time and amount of the last lime application, the historical lime requirements in the crop rotation for the specific field, and the results from the "buffer index" on the soil test report. The buffer index is the laboratory test to predict soil lime requirement. The buffer index is not affected by the "salt effect" and will result in a more accurate lime recommendation than just using the soil pH value alone.

Lime and fertilizer inputs have become more expensive in the past 3 years due to increased demand and fuel costs. Soil acidity must be neutralized for efficient utilization of fertilizer nutrients. Soil pH values may be lower than expected in fields that have been subject to drought in 2007. Take quality soil samples that are representative of the field in order to make proper recommendations. It is especially important to utilize historic knowledge of lime requirements for particular crop rotations and fields, as well as the soil buffer index test on the soil test report to determine appropriate lime needs in fall 2007. Proper lime use is the foundation of an efficient fertilization program.

Go to top of page.

This Web site is maintained by R. H. Giles, Jr.
Last revision September 22, 2006.