Balancing machine learning and artificial intelligence in soil science with human perspective and experience

Machine learning and artificial intelligence continue to evolve at a rapid pace, with many potential applications related to soil science. Even so, human experience and perception play an invaluable role in characterizing soil properties, especially qualitative properties that may elude sensing/computer-based modeling approaches. The elegant solution to this conundrum relies on the synthesis of computer-aided predictive modeling with human insight and knowledge.As global population surpasses 8 billion, the importance of optimized agronomic production to feed a hungry world has never been more important.

Characteristics of water erosion and conservation practice in arid regions of Central Asia:Xinjiang,China as an example

Located in the inland arid area of Central Asia and northwest China,Xinjiang has recently received heightened concerns over soil water erosion,which is highly related with the sustainable utilization of barren soil and limited water resources.Data from the national soil erosion survey of China(1985-2011)and Xinjiang statistical yearbook(2000-2010)was used to analyze the trend,intensity,and serious soil water erosion regions.Results showed that the water erosion area in Xinjiang was 87.6
103 km^(2) in 2011,mainly distributed in the Ili river valley and the northern and southern Tian Mountain.Soil erosion gradient was generally slight and the average erosion modulus was 2184 t/(km^(2) a).During the last 26 years,the water erosion area in Xinjiang decreased by 23.2%,whereas the intensity was still increasing.The driving factors from large to small impact included:population boom and human activities4vegetation degradation4rainfall and climate change4topography and soil erodibility4tectonics movement.Soil water erosion resulted in eco-environmental and socioeconomic losses,such as destroying farmland and grassland,triggering floods,sedimentation of reservoirs,damaging transportation and irrigation facilities,and aggravating poverty.A landscape ecological design approach is suggested for integrated control of soil erosion.Currently,an average of 2.07×10^(3) km^(2) of formerly eroded area is conserved each year.This study highlighted the importance and longevity of soil and water conservation efforts in Xinjiang,and offered some suggestions on ecological restoration and combating desertification in arid regions of Central Asia.&2015 International Research and Training Center on Erosion and Sedimentation and China Water and Power Press.Production and Hosting by Elsevier B.V.This is an open access article under the CC BY-NC-ND license(http://creativecommons.org/licenses/by-nc-nd/4.0/).

Assessment of Potential Nutrient Release from Phosphate Rock and Dolostone for Application in Acid Soils

Finding alternative local sources of plant nutrients is a practical, low-cost, and long-term strategy. In this study, laboratory column experiments were conducted in a completely randomized design to evaluate the feasibility of using phosphate rock and dolostone as fertilizers or acid-neutralizing agents for application in tropical acid soils. The dissolution rates of different particle-size fractions(0.063–0.25, 0.25–0.5, and 0.5–2 mm) of both rocks were studied by citric acid solution at p H 4 and 2 and water, with extraction times of 1, 3, 5, 7, 12, 24, 72, 144, 240, and 360 h. The results showed that the dissolution of both rocks depended on the particle size,leaching solution, and extraction time. The dissolution rate of rock-forming minerals increased as the specific surface area increased,corresponding to a decrease in particle size. In all cases, the release kinetics was characterized by two phases: 1) a first stage of rapid release that lasted 24 h and would ensure short-term nutrient release, and 2) a second stage of slow release after 24 h, representing the long-term nutrient release efficiency. Both rocks were suitable as slow-release fertilizers in strongly acid soils and would ensure the replenishment of P, Ca, and Mg. A combination of fine and medium particle-size fractions should be used to ensure high nutrient-release efficiency. Much work could remain to determine the overall impact of considerable amounts of fresh rocks in soils.

Proximal sensor-enhanced soil mapping in complex soil-landscape areas of Brazil

Portable X-ray fluorescence(pXRF) spectrometry and magnetic susceptibility(MS) via magnetometer have been increasingly used with terrain variables for digital soil mapping. However, this methodology is still emerging in many countries with tropical soils. The objective of this study was to use proximal soil sensor data associated with terrain variables at varying spatial resolutions to predict soil classes using the Random Forest(RF) algorithm. The study was conducted on a 316-ha area featuring highly variable soil classes and complex soil-landscape relationships in Minas Gerais State, Brazil. The overall accuracy and Kappa index were evaluated using soils that were classified at 118 sites, with 90 being used for modeling and 28 for validation. Digital elevation models(DEMs) were created at 5-, 10-, 20-, and 30-m resolutions using contour lines from two sources. The resulting DEMs were processed to generate 12 terrain variables. Total Fe, Ti, and SiO_(2) contents were obtained using pXRF, with MS determined via a magnetometer. Soil class prediction was performed using the RF algorithm. The quality of the soil maps improved when using only the five most important covariates and combining proximal sensor data with terrain variables at different spatial resolutions. The finest spatial resolution did not always provide the most accurate maps. The high soil complexity in the area prevented highly accurate predictions. The most important variables influencing the soil mapping were MS, Fe, and Ti. Proximal sensor data associated with terrain information were successfully used to map Brazilian soils at variable spatial resolutions.