February 2012 Bulletin Number 891 Soil Compaction Thresholds for the M1A1 Abrams Tank: Field Study at Camp Minden, La. Michael R. Lindsey, H. Magdi Selim, Jerry Daigle, Charles Guillory, Tamer A. Elbana, Marc Bordelon and Mitchell Mouton Tank exiting plot. LSU AgCenter Research Bulletin 891 - Soil Compaction Thresholds for the M1A1 Abrams Tank 1 Table of Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Material and Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Soils of the Study Site . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Experimental Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Soil Texture and Liquid and Plastic Limits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Bulk Density . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Penetration Resistance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Soil Moisture Retention . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Soil Texture and Liquid and Plastic Limits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Bulk Density . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Soil Penetration Resistance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 Residual Effect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Soil Moisture Retention . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Summary and Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 M1A1 tank traversing through study plot. 2 LSU AgCenter Research Bulletin 891 - Soil Compaction Thresholds for the M1A1 Abrams Tank Introduction Training in accordance with accepted standards and et al. (2005), suggest that a number of knowledge gaps under realistic combat conditions is necessary to produce exist related to the effects of military vehicles on natural military forces of the highest quality and thus ensure the resources. They also indicate that the bulk of the research national defense. In recent years, increased environmental to date had been conducted on military lands in the effects on many U.S. military installations can be attributed southwestern United States, while other regional areas like to a variety of factors including increased mechanization, the Southeast and Northeast remain largely understudied. heavier and faster vehicles, combined arms exercises, testing Due to significant regional ecosystem differences, it is requirements for advanced weapon systems and more unlikely that study results from one region will directly concentrated training because of base realignments and apply to others. As such, the environmental effects of closures. Military training activities frequently result in military tank maneuvers on training lands’ soils and land degradation that can negatively affect long-term use of vegetation are identified as a priority issue at military the land for training, as well as a broad range of damaging installations across the country (Althoff and Thien, 2005). environmental and ecosystem effects. An opportunity to further the study of the effects of Military training exercises using heavy tracked vehicles military tank traffic in the southeastern United States is an intensive land use activity that results in vegetation arose in 2002 when the Louisiana Army National Guard’s disturbance and soil compaction, which can have long- Camp Minden Training Site was chosen to serve as an lasting environmental effects (Althoff and Thien, 2005; M1A1 Abrams battle tank training facility. Approximately Johnson and Bailey, 2002; Palazzo et al., 2003; 2005; 50 M1A1 tanks were scheduled for detailed training and Fehmi et al., 2001; Diersing and Severinghaus, 1984). maneuvers at this facility. Camp Minden officials sought to Continuous long-term or intense short-term traffic by implement a soil and vegetation resilience study to comply military tanks can cause soil compaction and changes in with Department of Defense Integrated Training Area soil bulk density and soil strength that adversely affect a Management program’s regulations designed to maintain soil’s ability to sustain those functions considered to be training lands in a condition that accommodates future indicative of a soil in good condition. Furthermore, these long-term sustainability. changes may remain virtually invisible until secondary The purpose of this study was to establish critical soil indicators start to appear (Horn et al., 1995). These compaction thresholds for M1A1 Abrams battle tank secondary indicators are most often expressed as reduced traffic in an effort to minimize soil physical properties that soil structure and porosity, altered soil-water relationships, adversely affect vegetation regeneration. The hypothesis reduced aeration, increased runoff and soil erosion, reduced was that management of M1A1 training maneuver vigor in plant growth, impaired vegetation regeneration timing and intensity levels, as determined by soil moisture capabilities, altered plant community composition and conditions and traffic rates, could effectively reduce soil diversity and altered bird and mammal species diversity and compaction levels and the associated harmful effects on the distribution (Palazzo et al., 2003; Brady and Weil, 2002; overall soil quality and vegetation regeneration capabilities. Ayers, 1994; Diersing and Severinghaus, 1984; Goran et For this purpose, two main treatments were investigated: al., 1983). (1) soil moisture content and (2) tank traffic rates during Soil compaction and the associated negative effects on training maneuvers. The effect of soil moisture content other soil physical, chemical, biological and hydrologic and traffic rate on soil bulk density and soil penetration properties are widely recognized as the primary factors resistance measured before and after tank traffic were in reduced soil quality and function where tank training assessed. Furthermore, the influence of the resulting soil activities occur (Prose and Wilshire, 2000). In a review of compaction on soil moisture retention was investigated. the relevant military vehicle impact literature, Anderson LSU AgCenter Research Bulletin 891 - Soil Compaction Thresholds for the M1A1 Abrams Tank 3 Material and Methods Soils of the Study Site compaction. Because of this high susceptibility to wetness and the associated negative effects of soil compaction The location selected for the study was the Camp that would result from heavy mechanized maneuvers, our Minden Training Site, which is the Louisiana Army investigation was limited to this soil type. National Guard’s second-largest training site. It is located Experimental Design 16 miles east of Bossier City, La., Bossier/Webster parish line and covers approximately 13,682 acres (Figure 1). In March of 2003, 48 plots (5 by 5 meters each) were Camp Minden is located in the Western Coastal Plain established in the intermound areas of the selected study Major Land Resource Area (MLRA 133B) and in the site, which was in a managed pine forest stand. The plots Coastal Plain Province physiographic region. Camp were distributed over an area of approximately 2.6 hectares Minden is situated on Quaternary geologic sediments (6.4 acres) and were permanently located by driving 1.5 derived from braided stream terrace deposits of ancient meter by 1.6 centimeter diameter steel rebar rods into the river systems. ground at the plot corners. Subsequently, between March 2003 and July 2003, trees were harvested by chainsaw. Camp Minden Trees were removed from the site by skidder while avoiding traffic on research plots to minimize compaction or other disturbance. The site was not replanted, and it remained undisturbed for four years (until June 2007) to allow natural establishment of early succession vegetation. The experimental design was a completely randomized factorial design to evaluate the effects of soil moisture content (factor 1) and tank traffic rates (factor 2) on soil Camp Minden compaction and soil strength in the soil profile. Each treatment combination was replicated three times, resulting Study Site in a total of 27 experimental plots. Based on soil moisture Camp Ball determinations, three levels of soil moisture content were Camp Beauregard Camp Minden selected – low, medium and high. The effect of tank passes Camp Villere (factor 2) was split into three levels: (i) three; (ii) six; and Louisiana (iii) nine passes with the M1A1 battle tank in crisscross Carville Training Site Jackson Barracks configuration to achieve complete coverage of each plot. Treatment combinations were randomly assigned to 27 plots with eight additional plots available as controls for Figure 1 . Map of Louisiana Army National Guard facilities follow-up evaluations. In Table 1, measured soil moisture and the Camp Minden tank traffic and soil resilience study contents in the top 50 centimeters are presented in three site, near Bossier City, La . separate groups illustrating the differences in their values. To achieve this, a soil core, 1.9 centimeters in diameter and 50 centimeters long, was collected from the center of each The training site is situated in an area with nearly level plot prior to tank runs. The bulk sample was oven-dried, topography dominated by soils mapped as Kolin silt loam. and its volumetric moisture content was quantified. The intermound areas of these soils have been identified To arrive at different moisture levels in the soil profile, as Wrightsville inclusions and typically are level to sampling as well as tank passes were carried out at different depressional in nature. The Wrightsville series classifies as times during the year from August 2007 through October fine, mixed, active, thermic Typic Glossaqualfs. It is in land 2007. From the results in Table 1, three soil moisture levels capability subclass IIIw and as such has severe limitations were delineated for the different experimental plots. due to wetness that reduce the choice of plants or that require special conservation practices or both. This soil is The soil moisture distributions for low moisture level used mainly as woodland and is moderately well suited ranged from 0.07 to 0.18 cm3/cm3 and 0.14 to 0.21 cm3/ as pine woodland. The main concerns in producing and cm3 in the surface 25 centimeters and the 25-50 centimeter harvesting timber are severe equipment use limitations and layers, respectively. For the medium moisture level, the severe seedling mortality caused by wetness. When the respective soil moisture ranged from 0.24 to 0.29 and 0.20 soil is moist, timber harvesting methods that use standard to 30 cm3/cm3 in the surface 25 centimeters and the 25-50 wheeled and tracked vehicles often cause rutting and soil centimeter layers, respectively. For the high soil moisture 4 LSU AgCenter Research Bulletin 891 - Soil Compaction Thresholds for the M1A1 Abrams Tank Table 1 . Soil moisture content during 2007 at two depths and during 2009 at one depth for low, medium and high moisture plots . Soil Moisture Content Number of Plot and Treatment Sampling Date Tank Passes (cm3/cm3) 2007 Sampling 2009 Sampling 0 to 25 cm 25 to 50 cm 0 to 50 cm Low Soil Moisture Plots P15-L9 9 11-Oct-07 0 .08 0 .17 0 .49 P46-L9 9 11-Oct-07 0 .07 0 .14 0 .41 P47-L9 9 11-Oct-07 0 .13 0 .17 0 .46 P48-L6 6 20-Sep-07 0 .12 0 .19 0 .45 P23-L6 6 11-Oct-07 0 .12 0 .18 0 .48 P47-L6 6 11-Oct-07 0 .13 0 .17 NA P12-L3 3 11-Oct-07 0 .11 0 .14 0 .44 P09-L3 3 11-Oct-07 0 .18 0 .21 0 .49 P47-L3 3 11-Oct-07 0 .13 0 .17 NA Average Moisture Content 0 .12 0 .17 0 .46 Medium Soil Moisture Plots P08-M9 9 20-Sep-07 0 .24 0 .28 0 .5 P40-M9 9 11-Oct-07 0 .28 0 .20 0 .44 P41-M9 9 19-Sep-07 0 .23 0 .27 0 .45 P35-M6 6 29-Aug-07 0 .29 0 .25 0 .48 P21-M6 6 11-Oct-07 0 .24 0 .22 0 .45 P44-M6 6 11-Oct-07 0 .29 0 .23 0 .52 P06-M3 3 20-Sep-07 0 .27 0 .3 0 .53 P21-M3 3 11-Oct-07 0 .24 0 .22 NA P05-M3 3 11-Oct-07 0 .29 0 .22 0 .54 Average Moisture Content 0 .26 0 .24 0 .49 High Soil Moisture Plots P14-H9 9 21-Aug-07 0 .36 0 .41 0 .55 P13-H9 9 22-Aug-07 0 .38 0 .38 0 .63 P33-H9 9 22-Aug-07 0 .36 0 .41 0 .51 P07-H6 6 21-Aug-07 0 .35 0 .37 0 .54 P16-H6 6 22-Aug-07 0 .38 0 .37 0 .57 P17-H6 6 22-Aug-07 0 .4 0 .36 0 .52 P01-H3 3 21-Aug-07 0 .34 0 .35 0 .46 P32-H3 3 22-Aug-07 0 .37 0 .37 0 .45 P34-H3 3 22-Aug-07 0 .36 0 .37 0 .46 Average Moisture Content 0 .37 0 .38 0 .52 level, the respective soil moisture ranged from 0.36 to 0.40 top 25 centimeter depth when compared to the 25-50 cm3/cm3 and 0.36 to 41 cm3/cm3, respectively. The average centimeter soil depth – 0.12 versus 0.17 cm3/cm3. soil moisture content for the top 25 centimeter depths were Soil bulk density and penetrometer resistance 0.12, 0.26 and 0.37 cm3/cm3, for the low, medium and high measurements before and after tank passes were performed moisture level plots, respectively. For the 25-50 centimeter on the plots having the different soil moisture levels and depth, the respective soil moisture values were 0.17, 0.24 tank traffic rates. Penetration resistance measurements and 0.48 cm3/cm3. were carried out using Field Scout SC-900 cone For plots at medium and high moisture levels, uniform penetrometer (Spectrum Technologies Inc., Plainfield, Ill.). moisture contents were realized. Only plots at low moisture Therefore, in the subsequent discussion, designations low, levels showed significantly lower moisture content at the medium and high soil moisture levels refer to the soil- LSU AgCenter Research Bulletin 891 - Soil Compaction Thresholds for the M1A1 Abrams Tank 5 moisture content of the different plots as measured during the original core samples during August-December 2007. 2007 and given in Table 1. In total, 12 bulk density cores were extracted from each plot. We also investigated the influence of the tank traffic more than one year following the tank traffic. Our goal Penetration Resistance was to ascertain whether issues related to tank traffic Initial cone penetration resistance measurements manifested themselves over time. were taken at 5 centimeter depth intervals to a depth Specifically, we quantified the residual effect of the of 45 centimeters using a Spectrum Technologies Inc. tank traffic on soil penetrometer measurements on Jan., (Plainfield, Ill.) Field Scout SC-900 cone penetrometer. 13 2009, some 14-16 months following application. The The penetration resistance measurements were taken in date for these subsequent measurements was selected August, September and October 2007, when tank traffic when the soil moisture across all plots was near saturation was applied to individual plots. A total of 18 penetration and relatively uniform throughout the soil profile. Since resistance measurements were taken in each of the 27 moisture saturation in the winter months often is attained, experimental treatment plots. Nine measurements soil-moisture measurements were carried out only for were taken immediately preceding and nine were taken the surface 5 centimeters. The moisture content values immediately after tank passage to minimize possible corresponding to the low, medium and high soil moisture temporal effects related to soil moisture change and plots were 0.48, 0.49 and 0.53 cm3/cm3, respectively (Table possible disturbances. The measurements before and 1). Additional measurements were carried out on eight after tank traffic were taken along a diagonal transect in plots that were not subjected to tank traffic and are referred predetermined 1 meter grid sections within each plot. Cone to here as control plots. The average soil moisture content penetration rate of 2 centimeters per second, as specified for the control plots was 0.44 cm3/cm3 (data not given). in the ASAE standards, was followed. The measurements were taken under variable soil moisture levels as previously Soil Texture and Liquid and Plastic Limits outlined in Table 1. The soil particle size distribution, particle densities and In addition, follow-up penetration resistance was USDA textural classes were determined for each plot. measured at 1 centimeter depth intervals to a total depth Soil particle size distribution was determined using the of 60 centimeters using a Penetrologger cone penetrometer hydrometer method (Gee and Bauder, 1986, and particle (Eijkelkamp Agrisearch Equipment, Giesbeek, density was determined using the pycnometer method Netherlands). This was carried out in January 2009 when (ASTM D854-00 Standard Test Methods for Specific all plots were at relatively uniform soil moisture and near Gravity of Soil Solids by Water Pycnometer, 2000). The saturation (0.40 ± 0.05 cm3/cm3). The programmed Unified Soil Classification System classes and Atterberg penetration rate of 2 centimeters per second, as specified liquid and plastic limits for the less than 2 millimeter in the ASAE standards, was followed. Seven penetration particle size fraction were determined using ASTM 4318- resistance measurements were taken from each plot. An 00 Standard Test Methods for Liquid Limit, Plastic Limit additional eight randomly selected control plots were used and Plasticity Index of Soils (2000). to compare residual soil compaction effects on plots that Bulk Density had tank traffic and the undisturbed control plots. The penetration resistance measurements were taken along two Three soil bulk density core sample replicates were taken diagonal transects in predetermined 1 meter grid sections from each experimental plot at the 20 centimeter and within each plot. A total of 224 penetration resistance 50 centimeter depths prior to tank traffic. Soil depths of measurements were taken for a total of 13,440 data points. 20 centimeters and 50 centimeters were chosen to yield information on the epipedon (A and EBg horizons) and Soil Moisture Retention the argillic subsoil (Btg/E horizons), respectively. Soil Soil moisture retention curves were developed for a core samples were facilitated by excavating a 30 centimeter subset of field extracted soil cores using the pressure diameter hole to a depth of approximately 60 centimeters plate method. The moisture retention curves were used deep at the center of each plot. The cores were taken by to evaluate changes in pore size distribution of the soils driving a 68.7 cm3 (3 centimeters long by 5.4 centimeters resulting from tank traffic induced soil compaction. Twelve diameter) brass cylinder horizontally into the bore hole’s soil cores (2 centimeters long by 5.08 centimeters diameter) wall. These cores were used to establish the pre-traffic soil were extracted from the 20 and 50 centimeter depth bulk densities of the individual plots in June-August 2007. intervals of two high moisture, nine pass treatment plots Post-traffic soil bulk densities were determined subsequent and adjacent nontrafficked control areas on Sept. 15-16, to tank passes by excavating the original bore hole and 2009. taking an additional three cores within 30 centimeters of 6 LSU AgCenter Research Bulletin 891 - Soil Compaction Thresholds for the M1A1 Abrams Tank The extracted cores were wrapped in cellophane to contact and were allowed to saturate for five to seven prevent moisture loss during transport to the laboratory. days. The moisture retention or characteristic curves were Prior to placement on the ceramic pressure plates, the developed using the following pressures 0.1, 0.3, 0.5, 1.0, cores were shaved at both ends to ensure maximum surface 3.0, 5.0, 10.0, and 12.5 bars. Results and Discussion Soil Texture and Liquid and Plastic Limits with low liquid limit values less than 50. The Unified Soil Classification System designations and Atterberg limits of Soil texture results were grouped and averaged in four the Camp Minden soils are given in Table 3. groups referred to here as site areas S1 to S4. Those areas Bulk Density correspond to centralized data loggers around which individual plots were distributed (Table 2). The soils Analysis of variance and Tukey-Kramer HSD (honest of the study site are considered fine-grained soils and significant differences) statistical analysis were used for consist of various percentages of silt and clay with smaller treatment means comparisons. Data analysis after tank percentages of sand. disturbance indicated no significant treatment effects Of potential relevance to this study was the identification (P ≤ 0.05) for changes in bulk density at the 20 centimeter or the 50 centimeter depth. Table 4 shows the average of soil textures in the A, EBg, and Btg/E horizons. post-tank bulk density values as grouped by moisture level Generalized USDA soil textures were as follows: (i) A and traffic rate. The table illustrates the average trends of horizon – silt loam; (ii) EBg horizon – silt loam and silty the treatment levels without consideration of treatment clay loam; and (iii) Btg/E horizon – silty clay loam and silt interactions and is presented as a simplified overview of the loam. tank traffic experiment results. Furthermore, classification of soils under the Unified At the 20 centimeter depth, moisture treatment effect Soil Classification System uses a combination of letters to followed the trend high > medium > low moisture levels describe soil properties that primarily affect engineering with average bulk densities of 1.65, 1.61 and 1.57 g/cm3, properties. The soils at the Camp Minden study site are respectively (Table 4). At the same depth interval, the thus classified as ML, CL and CL-ML. The study area is traffic rate treatment effect followed the trend 6 >3 = 9 dominated by CL and to a lesser degree ML soils, where passes with average bulk densities of 1.63, 1.60 and 1.60 C equals fine-grained soils with plastic characteristics; g/cm3, respectively. At the 50 centimeter depth interval, M equals fine-grained soils with nonplastic to slightly the moisture treatment effect followed the trend medium plastic characteristics; and L equals fine-grained soils Table 2 . Mean particle size fractions, particle density and USDA textural class . Site Area† and Clay Silt Sand Particle Density USDA Texture Soil Depth (< 2µm)‡ (2-50µm)‡ (>50µm)‡ -------------------- % --------------------- g cm-3 S1-20cm 28±3 66±5 6±2 2 .69 SiL, SiCL S2-20cm 24±3 68±3 8±2 2 .69 SiL S3-20cm 23±2 62±3 15±3 2 .69 SiL S4-20cm 24±3 61±4 15±4 2 .69 SiL S1-50cm 28±7 62±2 10±8 2 .69 SiCL, SiL S2-50cm 26±3 66±4 8±5 2 .69 SiL, SiCL S3-50cm 27±4 57±4 16±6 2 .69 SiCL, SiL S4-50cm 27±6 56±2 17±6 2 .69 SiCL, SiL † Site area denotes plots associated with data loggers S1 to S4 and depth (cm). ‡ Values following ± represent standard deviation. LSU AgCenter Research Bulletin 891 - Soil Compaction Thresholds for the M1A1 Abrams Tank 7 Table 3 . Atterberg limits expressed as gravimetric soil moisture content (g/100g soil) and Unified Soil Classification System class for shallow (20 centimeter) and deep (50 centimeter) horizons for the different plots . Site Area† Liquid Limit Plastic Limit Plasticity Index USCS Class and Soil Depth (g/100g) (g/100g) (g/100g) S1-20cm 28 17 .4 10 .6 CL S2-20cm 26 .1 20 .5 5 .6 CL-ML S3-20cm 24 .2 19 .9 4 .3 CL-ML S4-20cm 25 .9 18 .6 7 .3 CL, CL-ML S1-50cm 29 .1 18 .5 10 .6 CL S2-50cm 30 .7 17 .7 13 .0 CL S3-50cm 29 .6 18 .4 11 .2 CL S4-50cm 29 .5 18 .7 10 .8 CL † Site Area denotes plots associated with data loggers S1 to S4 Table 4 . Average soil bulk density at 20 and 50 centimeter depths after tank passes for the different moisture plots . Values in parenthesis are the standard deviation . Mean Soil Bulk Density (g/cm3) by Moisture Level and Number of Tank Passes Moisture Number Tank Passes Means by Plots 3 6 9 Moisture Level 20 cm Depth Low 1.58 (0.08) 1.59 (0.11) 1.55 (0.10) 1.57 (0.09) Medium 1.58 (0.11) 1.68 (0.07) 1.58 (0.09) 1.61 (0.09) High 1.64 (0.07) 1.63 (0.06) 1.67 (0.08) 1.65 (0.07) Means by Tank Passes 1.60 (0.08) 1.63 (0.08) 1.60 (0.08) 50 cm Depth Low 1.58 (0.06) 1.61 (0.05) 1.48 (0.08) 1.56 (0.06) Medium 1.61 (0.05) 1.60 (0.05) 1.62 (0.05) 1.61 (0.05) High 1.64 (0.07) 1.61 (0.04) 1.57 (0.05) 1.60 (0.05) Means by Tank Passes 1.61 (0.06) 1.61 (0.05) 1.56 (0.06) > high > low, with bulk densities of 1.61, 1.60 and 1.56 measured for the medium moisture plot after six passes. In g/cm3, respectively (Table 4). The traffic rate treatment contrast, lowest increase in bulk density was observed for effect in the 50 centimeter interval was 3 = 6 > 9 with bulk the low moisture plots after nine tank passes. We recognize densities 1.61, 1.61 and 1.56 g/cm3, respectively. the extensive heterogeneity of the soil profile across the landscape as a contributor to the variability in the observed Increases in soil bulk density as a result of tank traffic bulk density. Nevertheless, such increases indicate soil was observed for all experimental plots regardless of the moisture is a significant soil parameter in changes of bulk number of tank passes (see Figures 2 and 3). Overall soil density resulting from tank traffic. bulk density increases throughout the soil profile were 0.04, 0.07 and 0.04 g/cm3 for the low, medium and high The lack of statistical significance observed among moisture plots, respectively. Such increases in bulk density treatment combinations using analysis of variance and appear similar for all moisture levels. Nevertheless, the Tukey’s HSD tests at P = 0.05 can be attributed to low largest increase due to compaction from tank passes was sample replicate numbers and high soil heterogeneity at 8 LSU AgCenter Research Bulletin 891 - Soil Compaction Thresholds for the M1A1 Abrams Tank Soil Penetration Resistance A primary goal of this study was to assess the influence of different levels of soil moisture on changes in penetration resistance and soil bulk density as a result of traffic by the A1M1 Abraham tank. Penetration resistance before and after tank passes are shown in Figures 4-6. These penetration resistance measurements were carried out during 2007 on plots having low, medium and high moisture levels as discussed under the Methods section (Table 1). For plots with low soil moisture content profiles, prior to tank passes, high penetration resistance values were measured for all plots as shown in Figure 4. The penetration resistance values ranged from 3 to 4 MPa Figure 2 . Mean bulk densities before and after tank traffic throughout the soil profile. Such penetration resistance and ± 95 percent confidence intervals at 20 centimeters results were not unexpected, particularly near the soil depth . surface where low moisture contents in the range of 0.07 to 0.18 cm3/cm3 were encountered. Following three, six and nine tank passes, consistent increases in penetration resistance measurements were observed compared to penetration resistance measurements before tank passes. Such an observation was consistent throughout the soil profile. After six and nine tank passes, penetration resistance values exceeded 5 MPa near the surface compared to plots after three tank passes where lower values were observed (< 5MPa). Nevertheless, the effects of the number of passes appear somewhat inconsistent when low moisture contents were dominant in the soil profile. For plots with medium moisture levels, initial penetration resistance distributions before tank passes were about 3 MPa throughout the soil profile as shown in Figure 5. The only exception was for soil depths below Figure 3 . Mean bulk densities before and after tank traffic 20 centimeters where a maximum occurred. An increased and ± 95 percent confidence intervals at 50 centimeters penetration resistance was measured for all plots following depth . tank passes. Such a penetration resistance increase was about 2 MPa and was observed throughout the soil profile regardless of the number of tank passes. Moreover, regardless of the number of tank passes, penetration resistance distributions followed the overall trend of the the site. Trends in the data strongly suggested a moisture initial penetration resistance distributions – those prior to treatment effect, however. Considering individual treatment tank trafficking. Such trends in penetration resistance prior factor only, “moisture” was a stronger determinant of to and following tank passes may be somewhat unique and final bulk density than was “traffic rate” at both the 20 were not found in the other experimental plots of high or centimeter and the 50 centimeter depths. The Leverage low soil moisture levels. plots indicated that, at the 20 centimeter depth interval, moisture treatment was significant (a = 0.05) but that For plots with the highest moisture level, initial (prior neither traffic rate level or the interaction between moisture to traffic) penetration resistance values versus soil depth treatment and traffic rate were significant (P = 0.05). At were lowest among all measured plots as shown in Figure the 50 centimeter depth interval, Leverage plots indicated 6. Penetration resistance values did not exceed 3 MPa that moisture treatment, traffic rate and their interactions throughout the soil profile. The effect of tank traffic was all were borderline significant at a confidence level of 0.05, largely concentrated in the surface 15 centimeters with an as indicated by confidence interval curves. average increase of 1 MPa regardless of the number of tank LSU AgCenter Research Bulletin 891 - Soil Compaction Thresholds for the M1A1 Abrams Tank 9 Figure 4 . Penetration resistance versus soil depth as measured during 2007 on the low mois- ture plots before and after three, six and nine tank passes . Figure 5 . Penetration resistance versus soil depth as measured during 2007 on the medium moisture plots before and after three, six and nine tank passes . Figure 6 . Penetration resistance versus soil depth as measured during 2007 on the high moisture plots before and after three, six and nine tank passes . 10 LSU AgCenter Research Bulletin 891 - Soil Compaction Thresholds for the M1A1 Abrams Tank