This study was supported by funding from a US Department of Agriculture Evans-Allen Grant (No. 1005761). The author thanks staff members at the TSU&#39;s Main Campus AREC in Nashville, Tennessee for their assistance. Maggie Syversen helped by reading the early version of the manuscript. The author appreciates the anonymous reviewers for their constructive comments and suggestions.

1. Clustered Sampling Design in a Plot
<ol>
	<li>Identify sampling zones within a research plot. Determine the number of square grids with equal length (i.e., Figure 1; Figure 3). Based on the size and shape of the research plot, the target number of square grids is expected to be six to ten so that the total number of soil samples is controlled below 30 within a plot (see step 1.3).</li>
	<li>Mark the center of each square grid (i.e., centroid) and create a circular sampling area with a diameter equal to the side length of the square grid.</li>
	<li>Stand on the centroid in the circular zone with closed eyes and throw a small stone (or another object with weight) in a random direction and distance from the centroid.
	<ol>
		<li>If the stone is dropped outside of the circular area, do it again until the first sampling location is identified.</li>
	</ol>
	</li>
	<li>Repeat step 1.3 until three random sampling locations are obtained in the circular zone.</li>
	<li>Put flags on the three sampling locations and number each flag (i.e., 1, 2, and 3).</li>
	<li>Repeat steps 1.3 - 1.5 in all other circular sampling zones until all locations are determined and numbered in a sequential order (i.e., 4, 5, 6, etc.).</li>
</ol>
2. Distance Measurements and Soil Collection in a Plot
<ol>
	<li>Choose one corner point and identify it as the origin for the sampling area in the plot.</li>
	<li>Measure horizontal and vertical distances of each flagged location relative to the origin and record them in a field notebook as x and y coordinates.</li>
	<li>Use a soil auger to take a soil core (0 - 15 cm) from each flagged location and label the bag based on the flag number. Repeat this step until soil cores are taken at all flagged locations.</li>
	<li>To minimize the influence of sampling (e.g., trampling on plants and soil in the plot), ensure that the bags with the soil samples inside stay with their respective flag until assembling all bags in the plot at once at the end of the collection.</li>
	<li>Transport the soil samples in coolers to the laboratory and process each soil core on the same day.</li>
	<li>Remove roots from each core, sieve it through a 2 mm soil sieve, and thoroughly homogenize each core sample prior to any analysis.</li>
	<li>Determine soil moisture content in each sample by oven-drying subsamples for 24 h at 105 &#176;C and ground the air-dried soil subsamples to a fine powder for a total carbon (C) analysis using an elemental analyzer4. SOC is derived based on the moisture and C contents.</li>
	<li>Weigh fresh soil subsamples (of 10 g each) and quantify the soil MBC by chloroform fumigation-K2SO4 extraction and potassium persulfate digestion methods5.</li>
	<li>Combine the SOC and MBC dataset with x and y coordinates based on the flag numbers in the plot.</li>
</ol>
3. Descriptive and Geostatistical Analyses in a Plot
<ol>
	<li>For each variable of SOC and MBC, calculate the minimum, maximum, mean, median, and standard deviation, as well as the coefficient of variation (CV).</li>
	<li>For each variable, conduct a set of geospatial analysis (i.e., trend surface analysis, autocorrelation, and kriging map) to depict the primary surface pattern, fine-scale variability, and spatial distribution. The details of the approaches of geostatistical analyses can be found in former publications4,5.</li>
</ol>
4. Exploration of SSR and the Associated Sampling Accuracy in a Plot
<ol>
	<li>Plot the SSR and relative error (&#947;) based on the CV obtained in a plot. Within each plot, the log-transformed SSR and relative error (&#947;) have a negative linear relationship (equations 1 - 3). Based on the relationship (equation 3), the number of samples required for the specified accuracy can be determined: 
	<img alt="Equation 1" src="/files/ftp_upload/58519/58519eq2.jpg" />&#160; 
	<img alt="Equation 2" src="/files/ftp_upload/58519/58519eq3.jpg" />&#160; 
	<img alt="Equation 3" src="/files/ftp_upload/58519/58519eq4.jpg" /> 
	Here, CI, <img alt="Equation 6" src="/files/ftp_upload/58519/58519eq6.jpg" />, s, n, N, CV, and denote the confidence interval, plot mean, plot standard deviation, sample number, coefficient of variation, and relative error, respectively; t0.975 = 1.96. The log-transformed sample size requirement (N) has a negative linear relationship (i.e., slope = -2) with the log-transformed relative error (&#947;).</li>
	<li>Apply the above relationship for future sampling in a plot by calculating N in equation 3 under a desired accuracy (e.g., relative error [&#947;]). Or, for a given number of soil samples already collected in a plot, apply the relationship to derive at the associated accuracy.</li>
</ol>

<ol>
	<li>Young, I. M., Crawford, J. W. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Interactions+and+Self-Organization+in+the+Soil-Microbe+Complex.">Interactions and Self-Organization in the Soil-Microbe Complex.</a> Science. 304 (5677), 1634-1637 (2004).</li><li>Masoom, H., et al. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Soil+Organic+Matter+in+Its+Native+State:+Unravelling+the+Most+Complex+Biomaterial+on+Earth.">Soil Organic Matter in Its Native State: Unravelling the Most Complex Biomaterial on Earth.</a> Environmental Science and Technology. 50 (4), 1670-1680 (2016).</li><li>Tan, K. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=.">.</a> Soil Sampling, Preparation, and Analysis. , (2005).</li><li>Li, J. W., Richter, D. D., Mendoza, A., Heine, P. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Effects+of+land-use+history+on+soil+spatial+heterogeneity+of+macro-+and+trace+elements+in+the+Southern+Piedmont+USA.">Effects of land-use history on soil spatial heterogeneity of macro- and trace elements in the Southern Piedmont USA.</a> Geoderma. 156 (1-2), 60-73 (2010).</li><li>Li, J., et al. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Nitrogen+Fertilization+Elevated+Spatial+Heterogeneity+of+Soil+Microbial+Biomass+Carbon+and+Nitrogen+in+Switchgrass+and+Gamagrass+Croplands.">Nitrogen Fertilization Elevated Spatial Heterogeneity of Soil Microbial Biomass Carbon and Nitrogen in Switchgrass and Gamagrass Croplands.</a> Scientific Reports. 8 (1), 1734 (2018).</li><li>Chung, C. K., Chong, S. K., Varsa, E. C. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Sampling+Strategies+for+Fertility+on+a+Stoy+Silt+Loam+Soil.">Sampling Strategies for Fertility on a Stoy Silt Loam Soil.</a> Communications in Soil Science and Plant Analysis. 26 (5-6), 741-763 (1995).</li><li>Luo, Y. Q., et al. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Toward+more+realistic+projections+of+soil+carbon+dynamics+by+Earth+system+models.">Toward more realistic projections of soil carbon dynamics by Earth system models.</a> Global Biogeochemical Cycles. 30 (1), 40-56 (2016).</li><li>Li, J., Wang, G., Allison, S., Mayes, M., Luo, Y. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Soil+carbon+sensitivity+to+temperature+and+carbon+use+efficiency+compared+across+microbial-ecosystem+models+of+varying+complexity.">Soil carbon sensitivity to temperature and carbon use efficiency compared across microbial-ecosystem models of varying complexity.</a> Biogeochemistry. 119 (1-3), 67-84 (2014).</li><li>Conant, R. T., Ogle, S. M., Paul, E. A., Paustian, K. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Measuring+and+monitoring+soil+organic+carbon+stocks+in+agricultural+lands+for+climate+mitigation.">Measuring and monitoring soil organic carbon stocks in agricultural lands for climate mitigation.</a> Frontiers in Ecology and the Environment. 9 (3), 169-173 (2011).</li><li>Wieder, W. R., Bonan, G. B., Allison, S. D. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Global+soil+carbon+projections+are+improved+by+modelling+microbial+processes.">Global soil carbon projections are improved by modelling microbial processes.</a> Nature Climate Change. 3 (10), 909-912 (2013).</li><li>Swenson, L. J., Dahnke, W. C., Patterson, D. D. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=.">.</a> Sampling for Soil Testing. , (1984).</li><li>Jones, J. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=.">.</a> Laboratory Guide for Conducting Soil Tests and Plant Analysis. , (2001).</li></ol>

The author has nothing to disclose.

The traditional soil sampling method lacked a quantitative basis and led to unknown accuracy, whereas the more advanced sampling strategies involved intensive soil collections and induced unaffordable costs for most soil research at the field plot scale. A simple, efficient, and reliable sampling design should be a useful tool to balance both aforementioned methods and, more importantly, inform a quantitative way to determine the number required under certain accuracy for the sake of future sampling needs. However, such a sampling design is still missing. Here, a method for manipulating a clustered sampling procedure to quantify soil spatial heterogeneity was presented and, relying on this design, to inform the number of soil samples required for future sampling under specific accuracy. There are two critical steps within the protocol. The first is to determine the sampling area and identify the sampling zone in a given plot area. Because the dimension and shape of a specific research plot can vary from one study to another, the number and length of the square grid representing the sampling zone should be modified to best fit the plot characteristics and cover the plot area as much as possible. In general, the number of square grids should be limited to eight to ten so that 24 - 30 soil samples will be collected in a given plot. This less intensive sampling requirement is acceptable for a pilot study in a plot. The second critical step is to determine the sample number required under specific accuracy. Although the number of soil samples under a desired accuracy can be derived at based on the pilot sampling strategy, other available resources need to be accounted for (e.g., labor, cost, and personnel). If the number of soil samples required for a desired accuracy exceeds the affordability, the desired accuracy should be lowered so that the number of soil samples can be recalculated. The recalculations should be repeated until the best fit is achieved to balance the desired accuracy and the available resources.
The protocol can be easily modified to fit the specific shape, area, and location of a research plot. Even within an irregular plot or a very large or small plot area, the procedure can be performed by controlling the size of the square grid to cover most of the plot area. On the other hand, when soil samples are collected beyond the circular sampling zone in the plot, they can be still accounted for in the descriptive and geostatistical analysis. The flexibility of the protocol in this regard is advantageous as it can, thus, reduce the cost of sampling.
An important limitation of this method is that the number of soil samples required for certain accuracy will depend upon the plot level CV determined by a group of 24 - 30 soil samples in the pilot soil sampling. For a highly heterogeneous plot, 30 samples or less can produce a larger CV than that based on a greater number of samples (&#62; 30). As a result, the number of soil samples calculated with the same accuracy will be larger. That is, the number of soil samples required for the same accuracy will be overestimated in the plot. For a highly homogenous plot, a smaller number of samples will produce a plot level CV similar to 30 samples, thus, resulting in an overestimation of the resource need. Therefore, for these extremely heterogeneous or homogeneous plots, the soil sample number (i.e., 30 or less) proposed in the pilot sampling design may cause unnecessary investment either in the pilot sampling stage or in future sampling.
We demonstrate significant advantages of the clustered soil sampling strategy. It provides a reliable and affordable soil sampling strategy to obtain soil spatial heterogeneity and offers a quantitative way to derive the number of soil samples required for a certain desired accuracy. Although the intensive strip or stratified sampling may provide a better description of spatial variation, the cost of conducting such sampling is too high for most soil studies. The traditional sampling is arbitrary and lacks any quantitative basis for sampling accuracy. The current protocol is superior due to its less intensive sampling requirement, ease in operating it in the field, power to reveal spatial patterns using rigorous geostatistical analysis methods, and capacity to quantitatively determine the sample size given any desired accuracy. The knowledge of the sample size required for a specific sampling accuracy will allow researchers to strategize their investment in soil sampling efforts.
Employing the efficient clustered sampling procedure allows rigorous testing of soil spatial heterogeneity and improves scientists&#39; capacity to conduct soil sampling with accuracy. The less intensive and quantitative nature of the soil sampling strategy will enable its wide application in soil research communities. Given the likely altered soil spatial heterogeneity under rapid global changes, the soil sample requirement for the same sampling accuracy in a research plot may vary over time. The proposed sample number in the pilot sampling design may vary with different soils and ecosystems. Future applications that could emerge from this work include determining the sample number for specific soils or ecosystems. Thus, further empirical work is needed on the application and identification of the method in various soils and ecosystems. Long-term and wide applications may help identify a generic sample size requirement for specific ecosystems, which can be recommended for soil researchers.

Soils are highly heterogeneous biomaterials1,2. Soil sampling is conducted to collect the most representative samples and characterize the nutrient status of a field as accurately and inexpensively as possible. Variability in a soil lies in soil spatial heterogeneity and accuracy of quantification. When spatial variation in soil is not taken into consideration, typical soil sampling can result in a substantial departure from the true mean value of a soil variable, even if the soil analysis itself is highly accurate3. For a heterogeneous research plot, variability is frequently of more importance than means3; that is, a sampling design that can accurately measure both variability and mean will be preferred.
When soil spatial variation is further altered due to land management practices4,5,6, it is more difficult to conduct soil sampling in an accurate manner. Nevertheless, concerns also arise with regard to the large variations in key soil variables (e.g., SOC and MBC)7 that are propagated to cause poor constraints of key model parameters which are critical for long-term global soil model projections under climate change8,9,10. As the cost of soil sampling to characterize field variability is a key problem, a simple, reliable, and efficient soil sampling strategy is sought.
There are many different approaches to collecting representative soil samples in a research plot, and their advantages and disadvantages are summarized in Table 1. In a traditional soil sampling (i.e., simple and random sampling), a random collection of a few to more than 10 soil samples is performed in a research plot. In particular, the number of samples in a traditional soil sampling design is always determined arbitrarily and the associated sampling error (i.e., accuracy) remains unknown.
<table fo:keep-together.within-page="1">
	<tbody>
		<tr>
			<td>Sampling design</td>
			<td>Advantage</td>
			<td>Disadvantage</td>
		</tr>
		<tr>
			<td>Simple and random sampling</td>
			<td>Cost effective, quick and inexpensive, widely adopted, easy operation, optimal in homogenous site</td>
			<td>Low accuracy and high variation, &#60;5 samples</td>
		</tr>
		<tr>
			<td>Systematic sampling</td>
			<td>High accuracy and known variation, optimal in large scale heterogeneous site</td>
			<td>Cost ineffective, large sample number</td>
		</tr>
		<tr>
			<td>Stratified sampling</td>
			<td>&#160;Accurate mean estimate, relatively easy operation, optimal for clustered and stratified region</td>
			<td>Cost ineffective, large sample number (usually less than systematic/grid sampling)</td>
		</tr>
		<tr>
			<td>Compositing</td>
			<td>Cost effective, accurate mean estimate, easy operation, optimal in heterogeneous site</td>
			<td>Unknown field variation, &#62;3 samples for composite</td>
		</tr>
	</tbody>
</table>
Table 1: The advantages and disadvantages of major soil sampling designs adopted in the soil research community. The table has been summarized from Tan et al.3, Jones12, and Swenson et al.11
As compared with simple and random sampling or compositing, systematic and stratified sampling designs can achieve means with high accuracy along with associated variability (Table 1). However, they will require intensive soil sampling (e.g., a few 100 samples). Although the accuracy of, and confidence in, a soil test level increases with more soil samples collected per plot11, the requirement for a large number of soil samples is generally only applicable for a large-scale study5,11; it is well beyond the affordability of most soil research projects conducted at the scale of field plots due to constraints in resources. A sampling design is preferred to balance the tradeoffs of these different methods.
A key issue for a soil sampling design is to determine the number of soil samples required and the associated accuracy given the research questions and field conditions. For instance, a reduction in the number of soil samples is possible in less disturbed sites while still achieving the same level of precision6, suggesting a need to explicitly quantify the spatial heterogeneity (i.e., nature and occurrence of soil variability) prior to soil sampling3. In fact, no such pilot sampling is recommended in most soil sampling designs. Field scientists frequently fail to recognize the importance of estimating statistical power when they design experiments.
To improve the experimental rigor in soil sampling, a simple and efficient sampling method is presented in this study. The new design shall not only enable the accurate characterization of soil nutrient levels and variability but also, by accounting for soil spatial heterogeneity, provide a quantitative way to inform the number of soil samples and the associated sampling accuracy for future research. The new soil sampling design should help researchers identify optional strategies that fit their sampling and research needs. The overall goal of this method is to provide soil biogeochemists and ecologists with a quantitative and manipulative approach to optimize soil sampling strategies in the context of field research.

<table><tbody><tr><td>Soil auger</td><td>AMS</td><td>350.05</td><td>For soil collection</td></tr><tr><td>Screwdriver</td><td>Fisher Scientific</td><td>19-313-447</td><td>For soil collection</td></tr><tr><td>Rope</td><td>Fisher Scientific</td><td>19-313-429</td><td>For delineating sampling zone</td></tr><tr><td>FatMax 35 ft. Tape Measure</td><td>Home Depot</td><td>#215880</td><td>For measuring distances</td></tr><tr><td>Marking flag</td><td>Fisher Scientific</td><td>S99537</td><td>For marking sampling locations</td></tr><tr><td>Plastic Zipper Seal Storage Bag</td><td>Fisher Scientific</td><td>09-800-16</td><td>For soil collection</td></tr><tr><td>Sharpie</td><td>Fisher Scientific</td><td>50-111-3135</td><td>For soil collection</td></tr><tr><td>Marking pencil</td><td>Fisher Scientific</td><td>50-294-45</td><td>For recording data in field</td></tr><tr><td>Lab notebook</td><td>Fisher Scientific</td><td>11-903&amp;nbsp;</td><td>For recording data in field</td></tr><tr><td>ArcGis 10.3</td><td>ESRI</td><td></td><td>For producing kriging map</td></tr><tr><td>Sieve</td><td>Fisher Scientific</td><td>04-881G&amp;nbsp;</td><td>For sieving soil sample</td></tr></tbody></table>

sampling soils in a heterogeneous research plot

Soils are highly heterogeneous. In general, the number of soil samples required for soil research has always been determined arbitrarily and the associated accuracy is unknown. Here, we present a detailed protocol for efficient and clustered soil sampling in a research plot and, relying on a pilot sampling using this design, for demonstrating soil spatial heterogeneity and informing reasonable sample sizes and associated accuracy for future study. The protocol mainly comprises four steps: sampling design, field collection, soil analysis, and geostatistical analysis. The step-by-step procedure is modified according to former publications. Two examples will be presented to demonstrate contrasting spatial distributions of soil organic carbon (SOC) and soil microbial biomass carbon (MBC) under different management practices. In addition, we present a strategy to determine the sample size requirement (SSR) given a certain level of accuracy based on the plot-level coefficient of variation (CV). The field sampling protocol and the quantitative determination of the sample size will assist researchers in seeking feasible sampling strategies to meet research needs and resources' availability.

The above approach has been employed in two case studies, one in a Southern US rural region and another in Middle Tennessee.
In the rural Southern Piedmont region, three land-use types were selected, including 1) uncultivated oak-hickory hardwood forests, 2) cultivated fields where conventional tillage and fertilization are used annually to produce wheat, sorghum, and corn, and 3) old-field pine forests that are each about 50 years in age since the last cultivation4. Three independently replicated 30 x 30 m plots were identified from the area for each land use. In each plot, a cluster soil sampling design was applied (Figure 1). Each circular zone had a 5 m radial distance from each centroid. Twenty-seven cores were collected from each of the nine plots, 81 cores per land use, and 243 cores in total. SOC was quantified by a CHN analyzer. The major finding was that cultivated land substantially homogenizes the spatial heterogeneity of SOC and other variables4. The SSR differed among land uses with a generally ascending order as old-field forest &#62; regenerated pine forest &#62; cultivated cropland (Figure 2). Exceptions are that one hardwood forest plot had an SSR as small as the cultivated plot, and one pine plot had an SSR as large as the hardwood plot (Figure 2). Taking &#947; = 0.1 or 10% as an example, SSR was 4, 10, and 30 (cultivated cropland), 80, 85, and 300 (pine forest), and 25, 200, and 350 (hardwood). If only three soil samples were collected in all plots, the relative error would have been ~10% - 30% (cultivated cropland), ~50% - 80% (pine forest), and ~28% - 100% (hardwood).
<img alt="Figure 1" class="xfigimg" src="/files/ftp_upload/58519/58519fig1.jpg" /> 
Figure 1: An illustration of a clustered random sampling design within a 30 x 30 m research plot at the Calhoun Experimental Forest, SC, USA4. The filled circles represent centroids (n = 9). The large dashed circle represents the sampling area around one centroid (radius = 5 m). Xs represent sample locations determined from randomly chosen directions and distances from a centroid. This figure has been modified from Li et al.4. <a href="https://www-jove-com.remotexs.ntu.edu.sg/files/ftp_upload/58519/58519fig1large.jpg" target="_blank">Please click here to view a larger version of this figure.</a>
<img alt="Figure 2" class="xfigimg" src="/files/ftp_upload/58519/58519fig2.jpg" /> 
Figure 2: Plot of the sample size requirement (SSR) and relative error (&#947;) for SOC of hardwood forest, pine forest, and cultivated cropland. The log scale was applied on both axes. The dotted lines represent cultivated soils, the grey lines pine-forest soils, and the dark lines hardwood forest soils. Three different lines for each land use correspond to three replicate plots. This figure has been modified from Li et al.4 <a href="https://www-jove-com.remotexs.ntu.edu.sg/files/ftp_upload/58519/58519fig2large.jpg" target="_blank">Please click here to view a larger version of this figure.</a>
In the Tennessee State University (TSU) Main Campus Agriculture Research and Extension Center (AREC) in Nashville, TN, USA (36.12&#176; N, 36.98&#176; W, elevation 127.6 m) in 2011, a field switchgrass experiment was established with three nitrogen (N) fertilization treatments in a randomized block design5. The crop type is of the &#39;Highlander&#39; variety of eastern &#39;Alamo&#39; switchgrass (Panicum virgatum L.). The three N treatments included no N fertilizer input (NN), low N fertilizer input (LN: 84 kg of N ha-1 in urea), and high N fertilizer input (HN: 168 kg of N ha-1 in urea). Within each plot, a rectangular area of 2.75 x 5.5 m zone was identified and further divided into eight square grids of 1.375 x 1.375 m. Within each circular zone, a centroid was identified, and three cores were collected with a random direction and distance relative to each centroid (Figure 3). A total of 24 cores were thus collected from each of 12 plots, yielding 288 soil cores. The MBC in each core was quantified by chloroform fumigation-K2SO4 extraction and potassium persulfate digestion methods. The major finding was that the N fertilization generally enhanced the spatial heterogeneity of MBC in the switchgrass cropland. The SSR was generally greater with fertilization (Figure 4). One exception is that the SSR for an HN plot was lower than that of the NN plot (Figure 4). Taking &#947; = 0.1 or 10% as an example, SSR was 10 and 20 in two replicated plots (NN), 30 and 50 (LN), and 15 and 70 (HN). If only three soil samples were collected in all plots, the relative error would have been ~20% - 25% (NN), ~26% - 35% (LN), and ~20% - 40% (hardwood).
<img alt="Figure 3" class="xfigimg" src="/files/ftp_upload/58519/58519fig3.jpg" /> 
Figure 3: Illustration of a clustered random sampling design within a 2.75 x 5.5 m plot in a fertilization experimental site at the Tennessee State University (TSU) Agricultural Research Center in Nashville, TN, USA. The filled circles represent centroids (n = 8) and each plot consisted of eight centroids in each square grid (of 1.375 x 1.375 m). In each subplot, a circular area was determined for soil sampling. Xs represent sample locations determined from random directions and distances from a centroid within each circular sampling area (dashed circle). This figure has been modified from Li et al.5 <a href="https://www-jove-com.remotexs.ntu.edu.sg/files/ftp_upload/58519/58519fig3large.jpg" target="_blank">Please click here to view a larger version of this figure.</a>
<img alt="Figure 4" class="xfigimg" src="/files/ftp_upload/58519/58519fig4.jpg" /> 
Figure 4: Plot of the sample size requirement (SSR) and relative error (&#947;) for MBC under three fertilization treatments. The log scale was applied on both axes. The dotted lines represent cultivated soils, the grey lines pine-forest soils, and the dark lines hardwood forest soils. NN = no N fertilizer input; LN = low N fertilizer input; and HN = high N fertilizer input. Two different lines for each land use correspond to two replicate plots. This figure has been modified from Li et al.5. <a href="https://www-jove-com.remotexs.ntu.edu.sg/files/ftp_upload/58519/58519fig4large.jpg" target="_blank">Please click here to view a larger version of this figure.</a>

Watch this Scientific Journal Video about Sampling Soils in a Heterogeneous Research Plot at JoVE.com

Sampling Soils in a Heterogeneous Research Plot

The traditional soil-sampling procedure determines the number of soil samples arbitrarily. Here, we provide a simple yet efficient clustered soil-sampling design to demonstrate soil spatial heterogeneity and quantitatively determine the number of soil samples required and the associated sampling accuracy.

Soils are highly heterogeneous. In general, the number of soil samples required for soil research has always been determined arbitrarily and the ...

sampling-soils-in-a-heterogeneous-research-plot

Nanyang Technological University

Research

Education

JoVE Journal

JoVE Core

Environment

Statistics

Civil Engineering

Unit 1

Introduction to Surveying

Understanding Statistics

SCIENTISTS IN ACTION

34.3K Views.  Tennessee State University. The traditional soil-sampling procedure determines the number of soil samples arbitrarily. Here, we provide a simple yet efficient clustered soil-sampling design to demonstrate soil spatial heterogeneity and quantitatively determine the number of soil samples required and the associated sampling accuracy.

Video: Sampling Soils in a Heterogeneous Research Plot

Experimental Protocol for Manipulating Plant-induced Soil Heterogeneity

Coexistence theory has often treated environmental heterogeneity as being independent of the community composition; however biotic feedbacks such as plant-soil feedbacks (PSF) have large effects on plant performance, and create environmental heterogeneity that depends on the community composition. Understanding the importance of PSF for plant community assembly necessitates understanding of the role of heterogeneity in PSF, in addition to mean PSF effects. Here, we describe a protocol for manipulating plant-induced soil heterogeneity. Two example experiments are presented: (1) a field experiment with a 6-patch grid of soils to measure plant population responses and (2) a greenhouse experiment with 2-patch soils to measure individual plant responses. Soils can be collected from the zone of root influence (soils from the rhizosphere and directly adjacent to the rhizosphere) of plants in the field from conspecific and heterospecific plant species. Replicate collections are used to avoid pseudoreplicating soil samples. These soils are then placed into separate patches for heterogeneous treatments or mixed for a homogenized treatment. Care should be taken to ensure that heterogeneous and homogenized treatments experience the same degree of soil disturbance. Plants can then be placed in these soil treatments to determine the effect of plant-induced soil heterogeneity on plant performance. We demonstrate that plant-induced heterogeneity results in different outcomes than predicted by traditional coexistence models, perhaps because of the dynamic nature of these feedbacks. Theory that incorporates environmental heterogeneity influenced by the assembling community and additional empirical work is needed to determine when heterogeneity intrinsic to the assembling community will result in different assembly outcomes compared with heterogeneity extrinsic to the community composition.

Understanding the role of environmental heterogeneity in species coexistence has typically focused on types of heterogeneity that are extrinsic to the community&#8217;s species composition. We provide novel detailed methods for creating soil heterogeneity treatments using soils subject to plant-soil feedback conditioning, or heterogeneity intrinsic to the community composition.

Coexistence theory has often treated environmental heterogeneity as being independent of the community composition; however biotic feedbacks such as ...

Integrated Field Lysimetry and Porewater Sampling for Evaluation of Chemical Mobility in Soils and Established Vegetation

Potentially toxic chemicals are routinely applied to land to meet growing demands on waste management and food production, but the fate of these chemicals is often not well understood. Here we demonstrate an integrated field lysimetry and porewater sampling method for evaluating the mobility of chemicals applied to soils and established vegetation. Lysimeters, open columns made of metal or plastic, are driven into bareground or vegetated soils. Porewater samplers, which are commercially available and use vacuum to collect percolating soil water, are installed at predetermined depths within the lysimeters. At prearranged times following chemical application to experimental plots, porewater is collected, and lysimeters, containing soil and vegetation, are exhumed. By analyzing chemical concentrations in the lysimeter soil, vegetation, and porewater, downward leaching rates, soil retention capacities, and plant uptake for the chemical of interest may be quantified. Because field lysimetry and porewater sampling are conducted under natural environmental conditions and with minimal soil disturbance, derived results project real-case scenarios and provide valuable information for chemical management. As chemicals are increasingly applied to land worldwide, the described techniques may be utilized to determine whether applied chemicals pose adverse effects to human health or the environment.

Field lysimetry and porewater sampling allow researchers to evaluate the fate of chemicals applied to soils and established vegetation. The goal of this protocol is to demonstrate how to install required instrumentation and collect samples for chemical analysis during integrated field lysimetry and porewater sampling experiments.

Potentially toxic chemicals are routinely applied to land to meet growing demands on waste management and food production, but the fate of these chemicals ...

A Protocol for Collecting and Constructing Soil Core Lysimeters

Leaching of nutrients from land applied fertilizers and manure used in agriculture can lead to accelerated eutrophication of surface water. Because the landscape has complex and varied soil morphology, an accompanying disparity in flow paths for leachate through the soil macropore and matrix structure is present. The rate of flow through these paths is further affected by antecedent soil moisture. Lysimeters are used to quantify flow rate, volume of water and concentration of nutrients leaching downward through soils. While many lysimeter designs exist, accurately determining the volume of water and mass balance of nutrients is best accomplished with bounded lysimeters that leave the natural soil structure intact. 
Here we present a detailed method for the extraction and construction of soil core lysimeters equipped with soil moisture sensors at 5 cm and 25 cm depths. Lysimeters from four different Coastal Plain soils (Bojac, Evesboro, Quindocqua and Sassafras) were collected on the Delmarva Peninsula and moved to an indoor climate controlled facility. Soils were irrigated once weekly with the equivalent of 2 cm of rainfall to draw down soil nitrate-N concentrations. At the end of the draw down period, poultry litter was applied (162 kg TN ha-1) and leaching was resumed for an additional five weeks. Total recovery of applied irrigation water varied from 71% to 85%. Nitrate-N concentration varied over the course of the study from an average of 27.1 mg L-1 before litter application to 40.3 mg L-1 following litter application. While greatest flux of nutrients was measured in soils dominated by coarse sand (Sassafras) the greatest immediate flux occurred from the finest textured soil with pronounced macropore development (Quindocqua).

A detailed method for extraction and assembly of intact soil core lysimeters and their use for study of leachate and associated loss of nutrients from surface applied poultry litter is demonstrated.

Leaching of nutrients from land applied fertilizers and manure used in agriculture can lead to accelerated eutrophication of surface water. Because the ...

Soil Lysimeter Excavation for Coupled Hydrological, Geochemical, and Microbiological Investigations

Studying co-evolution of hydrological and biogeochemical processes in the subsurface of natural landscapes can enhance the understanding of coupled Earth-system processes. Such knowledge is imperative in improving predictions of hydro-biogeochemical cycles, especially under climate change scenarios. We present an experimental method, designed to capture sub-surface heterogeneity of an initially homogeneous soil system. This method is based on destructive sampling of a soil lysimeter designed to simulate a small-scale hillslope. A weighing lysimeter of one cubic meter capacity was divided into sections (voxels) and was excavated layer-by-layer, with sub samples being collected from each voxel. The excavation procedure was aimed at detecting the incipient heterogeneity of the system by focusing on the spatial assessment of hydrological, geochemical, and microbiological properties of the soil. Representative results of a few physicochemical variables tested show the development of heterogeneity. Additional work to test interactions between hydrological, geochemical, and microbiological signatures is planned to interpret the observed patterns. Our study also demonstrates the possibility of carrying out similar excavations in order to observe and quantify different aspects of soil-development under varying environmental conditions and scale.

This study presents an excavation method for investigating subsurface hydrological, geochemical, and microbiological heterogeneity of a soil lysimeter. The lysimeter simulates an artificial hillslope which was initially under homogeneous condition and had been subjected to approximately 5,000 mm of water over eight cycles of irrigation in an 18-month period.

Studying co-evolution of hydrological and biogeochemical processes in the subsurface of natural landscapes can enhance the understanding of coupled ...

Assessment of Labile Organic Carbon in Soil Using Sequential Fumigation Incubation Procedures

Management practices and environmental changes can alter soil nutrient and carbon cycling. Soil labile organic carbon, a readily decomposable C pool, is highly sensitive to disturbance. It is also the primary substrate for soil microorganisms, which is fundamental to nutrient cycling. Due to these attributes, labile organic carbon (LOC) has been identified as an indicator parameter for soil health. Quantifying the turnover rate of LOC also aids in understanding changes in soil nutrient cycling processes. A sequential fumigation incubation method has been developed to estimate soil LOC and potential C turnover rate. The method requires fumigating soil samples and quantifying CO2-C respired during a 10 day incubation period over a series of fumigation-incubation cycles. Labile organic C and potential C turnover rate are then extrapolated from accumulated CO2 with a negative exponential model. Procedures for conducting this method are described.

Labile organic carbon (LOC) and the potential carbon turnover rate are sensitive indicators of changes in soil nutrient cycling processes. Details are provided for a method based on fumigating and incubating soil in a series of cycles and using the CO2 accumulated during the incubation periods to estimate these parameters.

Management practices and environmental changes can alter soil nutrient and carbon cycling. Soil labile organic carbon, a readily decomposable C pool, is ...

Methods of Soil Resampling to Monitor Changes in the Chemical Concentrations of Forest Soils

Recent soils research has shown that important chemical soil characteristics can change in less than a decade, often the result of broad environmental changes. Repeated sampling to monitor these changes in forest soils is a relatively new practice that is not well documented in the literature and has only recently been broadly embraced by the scientific community. The objective of this protocol is therefore to synthesize the latest information on methods of soil resampling in a format that can be used to design and implement a soil monitoring program. Successful monitoring of forest soils requires that a study unit be defined within an area of forested land that can be characterized with replicate sampling locations. A resampling interval of 5 years is recommended, but if monitoring is done to evaluate a specific environmental driver, the rate of change expected in that driver should be taken into consideration. Here, we show that the sampling of the profile can be done by horizon where boundaries can be clearly identified and horizons are sufficiently thick to remove soil without contamination from horizons above or below. Otherwise, sampling can be done by depth interval. Archiving of sample for future reanalysis is a key step in avoiding analytical bias and providing the opportunity for additional analyses as new questions arise.

Repeated soil sampling has recently been shown to be an effective way to monitor forest soil change over years and decades. To support its use, a protocol is presented that synthesizes the latest information on soil resampling methods to aid in the design and implementation of successful soil monitoring programs.

Recent soils research has shown that important chemical soil characteristics can change in less than a decade, often the result of broad environmental ...

Microplot Design and Plant and Soil Sample Preparation for 15Nitrogen Analysis

Many&#160;nitrogen fertilizer studies evaluate the overall effect of a treatment on end-of-season measurements such as grain yield or cumulative N losses. A stable isotope approach is necessary to follow and quantify the fate of fertilizer derived N (FDN) through the soil-crop system. The purpose of this paper is to describe a small-plot research&#160;design utilizing non-confined 15N enriched microplots for multiple soil and plant sampling events over two growing seasons and provide sample collection, handling, and processing protocols for total 15N analysis. The methods were demonstrated using a replicated study from south-central Minnesota planted to corn (Zea mays L.). Each treatment consisted of six corn rows (76 cm row-spacing) 15.2 m long with a microplot (2.4 m by 3.8 m) embedded at one end. Fertilizer-grade urea was applied at 135 kg N&#8729;ha-1 at planting, while the microplot received urea enriched to 5 atom % 15N. Soil and plant samples were taken several times throughout the growing season, taking care to minimize cross-contamination by using separate tools and physically separating unenriched and enriched samples during all procedures. Soil and plant samples were dried, ground to pass through a 2 mm screen, and then ground to a flour-like consistency using a roller jar mill. Tracer studies require additional planning, sample processing time and manual labor, and incur higher costs for 15N enriched materials and sample analysis than traditional N studies. However, using the mass balance approach, tracer studies with multiple in-season sampling events allow the researcher to estimate FDN distribution through the soil-crop system and estimate unaccounted-for FDN from the system.

Microplot Design and Plant and Soil Sample Preparation for 15Nitrogen Analysis

A microplot design for 15N tracer research is described to accommodate multiple in-season plant and soil sampling events. Soil and plant sample collection and processing procedures, including grinding and weighing protocols, for 15N analysis are put forth.

Many&#160;nitrogen fertilizer studies evaluate the overall effect of a treatment on end-of-season measurements such as grain yield or cumulative N losses. ...

Monitoring Pedogenic Inorganic Carbon Accumulation Due to Weathering of Amended Silicate Minerals in Agricultural Soils.

The present study aims to demonstrate a systematic procedure for monitoring inorganic carbon induced by enhanced weathering of comminuted rocks in agricultural soils. To this end, the core soil samples taken at different depth (including 0-15 cm, 15-30 cm, and 30-60 cm profiles) are collected from an agriculture field, the topsoil of which has already been enriched with an alkaline earth metal silicate containing mineral (such as wollastonite). After transporting to the laboratory, the soil samples are air-dried and sieved. Then, the inorganic carbon content of the samples is determined by a volumetric method called calcimetry. The representative results presented herein showed five folded increments of inorganic carbon content in the soils amended with the Ca-silicate compared to control soils. This compositional change was accompanied by more than 1 unit of pH increase in the amended soils, implying high dissolution of the silicate. Mineralogical and morphological analyses, as well as elemental composition, further corroborate the increase in the inorganic carbon content of silicate-amended soils. The sampling and analysis methods presented in this study can be adopted by researchers and professionals looking to trace pedogenic inorganic carbon changes in soils and subsoils, including those amended with other suitable silicate rocks such as basalt and olivine. These methods can also be exploited as tools for verifying soil inorganic carbon sequestration by private and governmental entities to certify and award carbon credits.

The verification method described here is adaptable for monitoring pedogenic inorganic carbon sequestration in various agricultural soils amended with alkaline earth metal silicate-containing rocks, such as wollastonite, basalt, and olivine. This type of validation is essential for carbon credit programs, which can benefit farmers that sequester carbon in their fields.

The present study aims to demonstrate a systematic procedure for monitoring inorganic carbon induced by enhanced weathering of comminuted rocks in ...

In Situ Soil Moisture Sensors in Undisturbed Soils

Soil moisture directly affects operational hydrology, food security, ecosystem services, and the climate system. However, the adoption of soil moisture data has been slow due to inconsistent data collection, poor standardization, and typically short record duration. Soil moisture, or quantitatively volumetric soil water content (SWC), is measured using buried, in situ sensors that infer SWC from an electromagnetic response. This signal can vary considerably with local site conditions such as clay content and mineralogy, soil salinity or bulk electrical conductivity, and soil temperature; each of these can have varying impacts depending on the sensor technology.
Furthermore, poor soil contact and sensor degradation can affect the quality of these readings over time. Unlike more traditional environmental sensors, there are no accepted standards, maintenance practices, or quality controls for SWC data. As such, SWC is a challenging measurement for many environmental monitoring networks to implement. Here, we attempt to establish a community-based standard of practice for in situ SWC sensors so that future research and applications have consistent guidance on site selection, sensor installation, data interpretation, and long-term maintenance of monitoring stations.
The videography focuses on a multi-agency consensus of best-practices and recommendations for the installation of in situ SWC sensors. This paper presents an overview of this protocol along with the various steps essential for high-quality and long-term SWC data collection. This protocol will be of use to scientists and engineers hoping to deploy a single station or an entire network.

In Situ Soil Moisture Sensors in Undisturbed Soils

The determination of soil water content is a critical mission requirement for many state and federal agencies. This protocol synthesizes multi-agency efforts to measure soil water content using buried in situ sensors.

Soil moisture directly affects operational hydrology, food security, ecosystem services, and the climate system. However, the adoption of soil moisture ...

Soil Core Sample Collection for Water Retention Property Analysis

This video demonstrates the technique for collecting a soil sample for hydraulic property analysis. The sampling area is cleared, and the top crust is evacuated. The sampling ring is then placed into the soil to collect the sample. The ring with the sample is then cleaned, labeled, packed, and transferred undisturbed for analysis.

This video demonstrates the technique for collecting a soil sample for hydraulic property analysis. The sampling area is cleared, and the top crust is ...

Sampling Soils in a Heterogeneous Research Plot

Podsumowanie

Przeglądaj więcej filmów

Rozdziały w tym wideo

Experimental Protocol for Manipulating Plant-induced Soil Heterogeneity

Integrated Field Lysimetry and Porewater Sampling for Evaluation of Chemical Mobility in Soils and Established Vegetation

A Protocol for Collecting and Constructing Soil Core Lysimeters

Soil Lysimeter Excavation for Coupled Hydrological, Geochemical, and Microbiological Investigations

Assessment of Labile Organic Carbon in Soil Using Sequential Fumigation Incubation Procedures

Methods of Soil Resampling to Monitor Changes in the Chemical Concentrations of Forest Soils

Microplot Design and Plant and Soil Sample Preparation for ¹⁵Nitrogen Analysis

Monitoring Pedogenic Inorganic Carbon Accumulation Due to Weathering of Amended Silicate Minerals in Agricultural Soils.

In Situ Soil Moisture Sensors in Undisturbed Soils

Soil Core Sample Collection for Water Retention Property Analysis