Revisiting the Pedocal/Pedalfer boundary and Soil Moisture Regimes using the javaNewhall simulation model and PRISM data

Highlights

• The zone where Annual Water Balance = 0 decreases from North to South in CONUS

• Annual Water Balance = 0 estimated from Newhall Model approximates Pedocal line.

• Soil Moisture Regimes are dynamic with the Udic being most stable and Ustic least.

• Weak relationship of organic carbon and CaCO₃ content in relation to Water Balance

Abstract

We examine the climatic record of the conterminous United States from 1895 to 2014 as expressed through the PRISM dataset and the jNewhall model. Specifically, the zero line of the Annual Water Balance (AWB) and the respective oscillation zone through its longitudinal extent is examined. The zero line corresponds to the Pedocal/Pedalfer line defined by C.F. Marbut, considered outdated in current pedological circles, but conceptually powerful in denoting regional negative vs. positive AWB. Soil Moisture Regimes are reviewed and a means of expressing Soil Moisture Regime variability is introduced. Results indicate a difference in the width of the AWB oscillation zone from South to North with a demarcation approximating 40 degrees North Latitude. PRISM data is verified from a select set of National Weather Service station data to assess the utility of using readily accessible PRISM data for performing similar work by others. The effect of climate variables on organic carbon (OC) stock and depth of maximum Calcium Carbonate concentration is examined for a suite of soils along a climo-sequence from North Dakota to Central Iowa and found to account for a mild amount of the variability of both variables.

Introduction

The spatial climate datasets developed by the Oregon State University PRISM Climate Group are widely used, providing multi-temporal and multi-resolution data in a readily available and usable raster format. PRISM data has been developed and refined since 1991 (Daly et al., 1994) using a combination of mathematical models and expert knowledge and serves as the official spatial climate datasets for the United States Department of Agriculture. The continuous surface representation of common climate variables allows users to overcome the limitations of using discrete point data from irregularly distributed weather stations when performing GIS and modeling operations. PRISM data has been shown to be a reasonable representation of temperature and precipitation, specifically in the low relief Central and Eastern regions of the country with cross-validated Mean Absolute Errors of 4–5% for annual precipitation and approximately 0.7 °C for monthly minimum and 0.5 °C for monthly maximum temperatures (Daly et al., 2008).

Soil Moisture Regimes (SMR) reflect variations in seasonal soil moisture status and are used to differentiate soil classes in Soil Taxonomy from the Order to subgroup categories (Soil Survey Staff, 2014). The Newhall Model was developed in 1972 to apply data from weather stations to assist in classification of Soil Moisture Regimes in the application of Soil Taxonomy (Newhall and Berdanier, 1996). Any weather station with sufficient mean monthly precipitation and temperature data is suitable for this model, with the standard practice of using a 30 year normal for estimation of soil moisture or temperature regime (Waltman, 1997). The model is designed for regional rather than field specific spatial extents where detailed climate data is not available and relies on basic assumptions regarding water inputs and outputs and calculation of evapotranspiration (Thornthwaite, 1948). The Newhall model has been used to study trends in climate variability and drought in the Central Plains, USA (Waltman et al., 2004) and the Great Plains, USA (Salley et al., 2016), Soil Climate regimes in Pennsylvania, USA (Waltman, 1997), the Soil Climate Regimes in the United States (Soil Survey Quality Assurance Staff, 1994; Winzeler et al., 2013), South America (Van Wambeke, 1981), Africa (Van Wambeke, 1982) and Asia (Van Wambeke, 1985). A comparison of the Newhall model with physically based models reflects the strength of the Newhall model in classifying broad, pedoclimatic regions, with shortcomings for more specific estimates of soil moisture dynamics and evapotranspiration (Bonfante et al., 2011).

Annual Water Balance is an additional output of the Newhall model, which allows investigation of the traditional pedocal-pedalfer regions as proposed by Marbut (1935). Marbut was building on the concepts of Dokuchaiev that soil was the product of the factors of climate, vegetation, relief, parent material and age, rather than the previously held theory that soil was strictly a product of parent material. He proposed grouping soil types into two great groups, Pedocals and Pedalfers. Pedocals are noted by the accumulation of calcium and magnesium in parts or all of the soil profile, while Pedalfers are noted by a lack of calcium or magnesium and often by an accumulation of iron and/or aluminum. Marbut noted the correspondence of these two groups to climatic zones, with humid regions providing optimal conditions for Pedalfers and all drier areas, approximately 635 mm or less precipitation in the Continental United States (CONUS), providing optimal conditions for Pedocals (Kellogg, 1938). The subsequent classification scheme grouped soils into three broad orders, zonal, intrazonal and azonal. Zonal soils were defined as those effected solely from the actions of climate and organisms (primarily vegetation), while intrazonal soils have parent material and/or relief providing an overriding influence of climate or vegetation and azonal soils being unchanged by any factor (Kellogg, 1938). The terms Pedocal and Pedalfer were maintained in the description of the 1938 classification scheme as connotative subdivisions of zonal soils, broadly defined as soils of the grasslands (Pedocal) and soils of the forest and forest-grassland transition (Pedalfer). While Pedocal and Pedalfer are not in common use in the current soil science community, the concept and geographic representation of areas of moisture deficit and surplus (sub-humid and humid) is useful for investigating pedogenic and ecologic trends along this important climo-sequence. The current SMR classes of Udic and Ustic reflect the continued recognition of the humid/sub-humid climate regions with a continuing agricultural focus, with Udic areas requiring no irrigation for crop production and Ustic areas requiring drought tolerant/resistant crops.

The ready availability of PRISM data and the minimal data input requirements of the java Newhall simulation model (jNSM) (Soil Survey Staff, 2012) have spurred the use of PRISM data as a proxy for weather station data input files. Winzeler et al. (2013) compared the Soil Moisture Regimes for the CONUS (Soil Quality Assurance Staff, 1994) with results from Soil Moisture Regimes estimated from PRISM data and the Available Water Holding Capacity (AWC) parameterized from SSURGO (Soil Survey Staff, 2015) and STATSGO (Soil Survey Staff, 2006) data as inputs to the jNSM. They found class to class agreement with the 1994 map (~76%) and noted greater agreement with weather station data for the PRISM derived results compared to the 1994 map. They also note the ease of production and repeatability when using a standardized PRISM dataset and Newhall algorithm to generate Soil Moisture Regime maps compared with the time-consuming, imprecise, knowledge-based techniques required to create the 1994 map.

Salley et al. (2016) utilized PRISM data for the period 1895–2014 and AWC parameterized from STATSGO2 (Soil Survey Staff, 2006) as inputs to the jNSM to study drought trends, climate variability and climatic influence on soil properties in the Great Plains, USA. They plotted the pedocal-pedalfer boundary based on a constant AWC and the STATSGO2 based AWC and noted general agreement along the longitudinal extent with several sizable differences in Nebraska and Texas they attribute to a combination of AWC, AWB or elevation. They noted areas with soils of higher AWC exhibit an expected resiliency to dryer periods of short to moderate duration. They demonstrated correlation of calcic soil characteristics and AWB, but noted AWB as only one of the explanatory factors and suggested the limitations of using data from the National Soil Information System database for evaluating soil climate relationships.

The climatic parameter outputs from jNSM; Annual Water Balance, summer water balance, growing season water balance and biological windows at 5 °C and 8 °C are useful for mesoscale modeling, particularly when generated as a continuous surface. The relationship between mean annual precipitation and soil properties has a long history. Jenny and Leonard (1934) investigated the depth to carbonates in relation to precipitation in the Great Plains and noted a reasonable correlation. Ruhe (1984) emphasized the factors of stratigraphy and time as major explanatory variables determining depth of carbonates that needs to be accounted for in addition to precipitation. Retallack (1994) constrained variables to control for parent material, time, land use and landform and found a reasonable correlation (r² = 0.64) between mean annual precipitation and depth to calcic horizons. Salley et al. (2016) used gSSURGO data as inputs and confirmed Ruhe's point of precipitation being one of many variables controlling calcic horizon development. In a broad, global context, soil OC content increases with increasing precipitation, with a corresponding inverse relationship with temperature for matching amounts of precipitation (Post et al., 1982). Nichols (1984) in the Southern Great Plains and Sims and Nielsen (1986) in the Northern Great Plains of Montana, noted reasonable relationships (r² = 0.45–0.55) between soil OC content and precipitation.

The objectives of this investigation are to 1) identify the long-term ZL location and document the variability about the long-term line, 2) examine SMR distribution and variability over time and 3) examine organic carbon stock and calcium carbonate concentrations in relation to AWB. We hypothesized that 1) the ZL is in proximity to the ZL proposed by Marbut and varies uniformly along its extent, 2) SMR are dynamic and, 3) water balance has a moderate effect on depth to calcium carbonate concentration and OC stock.