Performing Department
Plant Science
Non Technical Summary
Soils are the foundation of civilization and provide food, fiber, and fuel, along with numerous environmental services. Improper soil management has led to some of the most disastrous times for humanity, such as the dust bowl era in the central United States. In order to properly manage soils, we need to understand their basic properties and the processes and factors that led to their formation. This informs land managers about the capacity of a specific soil to be used for various purposes, as well as how the soil will react to changes in land use. Soil mappers and pedologists use intricate knowledge of how soils form, and how they respond to environmental factors, to predict soil properties at any given location, often without having to physically analyze the soil.Traditionally, soil maps were created at the field level, with a relatively coarse degree of resolution that showed major changes in soil types across the field. This level of resolution was matched by crop production practices, which typically applied seed and other crop inputs at a uniform rate across fields to match the "average" soil type on that field. Increasingly rapid development and adoption of precision agriculture technologies now allows farmers to more precisely apply crop inputs (1 meter or less resolution) to match soils and landscape positions within fields. To keep up, soil sampling has become more precise, with producers using grid sampling and other approaches to collect soil information at a greater degree of resolution than whole field or even per acre scales. However, precision technologies are often limited by the relative lack of resolution currently available in soil maps.The overarching goal of this project is to accelerate the science behind digital soil mapping and use this information to improve precision agriculture. Digital soil mapping allows us to produce soil maps at a finer degree of resolution, while covering more extensive areas. As a result, state and county soil surveys can be updated more frequently, and provide precise soil information to land managers, policy makers, and researchers in a more useable format. Moreover, digital soil mapping can be integrated with machine learning to develop predictive correlations relating soils and their properties to various uses and climatic changes.South Dakota, like most of the world, needs to prepare for the impacts of climate change. Climate predictions are for South Dakota to become warmer and wetter, which has the potential for both benefits and adverse consequences. Warming temperatures mean longer growing seasons and potentially increased spring soil temperature, but also more intense summer heat stress. Increased precipitation could potentially expand the area of crop production farther west, where it is often too dry for row crops. However, the current trends and predictions are for this precipitation to occur in less frequent and more intense events. This would likely cause more flooding and erosion, with drought stress between precipitation events. One of South Dakota's more unique soil issues, exacerbated by climate change, is saline soils. Sodium is naturally weather out of the parent material and typically moves down into ground water and away from plant roots. Recent shifts in precipitation have raised the water table up into the rooting zone, bringing sodium with it.This project will expand the types of soil properties that we can measure using drones, including subsurface properties, soil carbon stock, and total water holding capacity. Using drones, we will be able to create high resolution soil maps which should improve the efficiency of precision agricultural practices by reducing over-fertilizing and creating more efficient planting regimes. We will also assess impacts of land use and climate change on soils to predict potential shifts in soil carbon, while informing land managers and policy makers about practices that will reduce soil erosion and development of saline affected soils. Predictive models of these soil issues will be created using the updated soil maps produced by this project.
Animal Health Component
50%
Research Effort Categories
Basic
10%
Applied
50%
Developmental
40%
Goals / Objectives
The goal of this project is to create a nationally recognized program centered around the advancement of digital soil mapping, soil development, remotely sensed soil properties, and assessing the effects of climate change on South Dakota's soil resources. The short term (1-2 year) focus will be to create a soil property database and soil products for use in precision agriculture. The long term (2-5 year) focus will be to assess climate change impacts to soil and agriculture using models based on the initial soil property database.Objective 1: Create a high-resolution soil property database for South Dakota using a combination of available data and newly collected soil samples.Sub-Objective 1A: Compile baseline soil data for South Dakota.Sub-Objective 1B: Asses the use of spectral and other remotely sensed data in predicting soil properties.Sub-Objective 1C: Create regional coefficients and relationships between soil properties and remotely sensed data for use in area mapping projects.Sub-Objective 1D: Create high resolution soil products for precision agriculture and collaborations to assess the potential impacts on yield, efficiency, and environmental degradation.Sub-Objective 1E: Update county soil survey boundaries and create an updated state soil map for South Dakota.Objective 2: Assess the effects of climate change on South Dakota soil resourcesSub-Objective 2A: Create a high-resolution soil carbon stock map to monitor and predict changes in soil carbon caused by climate change.Sub-Objective 2B: Assess the current distribution of sodium affected soil in South Dakota and create a risk map for the increased distribution and concentration of sodium in soils caused by predicted changes in precipitation and ground water.Sub-Objective 2C: Assess and predict the distribution of soils that will become more suited to crop production due to increased temperature and/or precipitation.Sub-Objective 2D: Create a predictive map of increased erosion susceptibility due to land-management changes and an increasing intensity of precipitation events.
Project Methods
Objective 1. Create a high-resolution soil property database for South Dakota using a combination of available data and newly collected soil samples. 1A: Compile baseline soil data for South Dakota. Each of the subsequent sub-objectives require baseline data for common soil types throughout the state. An assessment of available soil data and related information for SD will be conducted. This includes the Soil Survey Geographic Database (SSURGO) published by the NRCS, soil profile data for individual locations across the state, and statewide information on soil covariates (DEM, geology, vegetation, etc.). Test plots for common soils across the state will be extensively sampled to create high-resolution (1-2 m) maps of soil properties. Properties will be comparable to standard NRCS soil surveys and include both physical (texture, water holding capacity, hydraulic conductivity, drainage, etc.) and chemical (carbonates, cation exchange capacity, pH, electrical conductivity, etc.) properties. These properties will be used to determine secondary properties such as farmland classification, land use suitability ratings, and soil taxonomic classification.1B: Asses the use of spectral and other remotely sensed data in predicting soil properties. Unmanned Aerial Vehicles (UAVs) are being utilized in crop production and to collect a small variety of surficial soil properties, but have few known applications for deep soil properties. UAVs will be flown over each test site to collect and map various spectral properties (color, infrared, near-IR, etc.) along with a decimeter scale Digital Elevation Model (DEM). Sensors utilized by UAVs will include all currently used in soil science applications, as well as a variety of exploratory options to assess potential new relationships.1C: Create regional coefficients and relationships between soil properties and remotely sensed data for use in area mapping projects. Data from objective 1A should yield coefficients of relationships between soil properties and remotely sensed data that can be collected with UAVs. Data from individual test sites will be assessed as either single plots, clusters based on geographic proximity, or clusters based on soil type to see which yields the most statistically significant relationships. The goal for this objective is to produce and publish relatively simple relationships for single soil properties so that interested researchers, producers, or conservation district agents could measure soil properties using UAVs.1D: Create high resolution soil products for precision agriculture and collaborations to assess the potential impacts on yield, efficiency, and environmental degradation. In collaboration with precision agriculture researchers at SDSU, field trials will be established to determine if high resolution soil data can provide significant improvements to crop yield, input use efficiency, or pollutant loss from the soil system. Field trials will be conducted to evaluate the use of high resolution images and soil data in the production of corn, soy, and wheat. Multiple trials will be set up at experiment stations across South Dakota to determine soil types and climates most benefitted from high precision soil property information.1E: Update county soil survey boundaries and create an updated state soil map for South Dakota. Workflow for this objective will be partially determined by the strategies of the NRCS Digital Soil Mapping Team. Collaboration with this team will ensure that the product created will be compatible with the nationwide effort to update SSURGO and will assist with the team workload. Digital Soil Mapping (DSM) techniques will be applied to create a modern, raster format soil survey for South Dakota. DSM uses machine learning to understand and predict soil properties based on an extensive list of covariate data. The database created in sub-objective 1A will be used as training data for the machine learning application. The resulting product will be a seamless statewide soil survey which no longer has mapper biases and unnatural features along county (survey area) lines.Objective 2. Assess the effects of climate change on South Dakota soil resources2A: Create a high-resolution soil carbon stock map to monitor and predict changes in soil carbon caused by climate change. SD land managers have been effectively sequestering carbon in soil in recent decades, however the state lacks a comprehensive and current catalog of the total carbon stored in soils. The foundation of this project is within the soil property database and updated soil survey of objective 1. Those resources will be used to predict the total soil carbon across SD. A model will then be produced predicting the net flux of soil carbon based on published studies of the effects of warming on carbon cycling and predicted changes in climate for South Dakota. If there are no published studies applicable to this region then one will be designed and carried out with collaborators from the SDSU Agronomy, Horticulture, and Plant Science Department.2B: Asses the current distribution of sodium affected soil in South Dakota and create a risk map for the increased distribution and concentration of sodium in soils caused by predicted changes in precipitation and ground water. Sodium affected soils (SAS) are a growing concern in SD due to increases in precipitation and local water table height. To assess the extent and severity of the issue, a survey will be sent to SD land managers and crop producers. Local NRCS and county conservation districts will be contacted to gain information on SAS in their district areas. From this information a preliminary map of SAS will be created. This map will be used to find correlative soil data which should help predict other areas affected or susceptible to salinity issues. Representative areas of SAS will sampled and described to characterize their exact extent and severity.2C: Asses and predict the distribution of soils which will become more suited to crop production due to increased temperature and/or precipitation. Assessing changes in land use potential will mostly rely on models of climate change predictions and soil properties. Assessments will be modelled based on three climate change prediction scenarios: 1) 'business as usual', 2) slight reductions in greenhouse gas (GG) emissions, and 3) significant reductions in GG emissions. Areas where future precipitation and frost-free days are favorable to crop production will be compared to current areas of crop production to show the potential area of expanded land-use. Within this area of potentially expanded crop production, soil data will be used to eliminate areas unsuitable to for production due to slope, flooding, rock content, low fertility, etc. The remaining areas will then be ranked for crop production based on soil properties and land capability class.2D: Create a predictive map of increased erosion susceptibility due to land-management changes and an increasing intensity of precipitation events. Erosion rates have longstanding computer models but are usually only performed at the site scale. This project will use SDSU's High Performance Computing capabilities to develop a statewide erosion model and compare it with erosion potential data from published soil data. The erosion model will be tested under varying precipitation levels for three climate change predictions. Soils that show the most potential for erosion will be selected for a field study. The field study will monitor actual erosion amounts under various precipitation events, with a variety of ground cover options. These measurements will help validate the erosion model, as well as inform land managers of the most erosion sensitive soils. Thus, any expansion of row crops driven by climate change will not cause extensive and damaging erosion.