Increased uncertainty in soil carbon stock measurement with spatial scale and sampling profile depth in world grasslands: A systematic analysis
Citation
Maillard, É., McConkey, B.G., Angers, D.A. (2017). Increased uncertainty in soil carbon stock measurement with spatial scale and sampling profile depth in world grasslands: A systematic analysis. Agriculture, Ecosystems and Environment, [online] 236 268-276. http://dx.doi.org/10.1016/j.agee.2016.11.024
Plain language summary
There is an important need to better understand how the amount of soil organic carbon (SOC) affects the concentration of the greenhouse gas, carbon dioxide, in the atmosphere. Grasslands represent 40% of the earth's land surface but efficient measurement of the amount and changes of SOC in grasslands is a challenge. Although we know a lot about the factors that affect the grassland SOC, the variability of SOC over the grassland area is not well understood. This variability affects our ability to measure changes in SOC. Knowledge of the variability of SOC is needed to design effective measurement procedures for grasslands. The variability of SOC has not been described yet in the scientific literature. We determined the variability of grassland SOC from analysis of 51 research articles. The articles included data for 177 grasslands from 19 countries. We found that variability was proportional to the amount of SOC. We found that the size of the measurement area and the depths of sampling were important factors affecting the variability of SOC. The variability increases with the size of the area of grassland being measured. It also increases with the measurement depth in the soil. The variability of SOC affects our capability to detect changes in SOC. For example, with 50 samples at each measurement time, it would not be usually possible to detect SOC changes over time of less than 14% for an area of 10 hectares (the equivalent area of about 12 Canadian football fields). For measurements with 50 samples over a 100 km by 100 km area, it would not usually be possible to detect SOC changes of less than 29%. Regarding the effect of depth of measurement, it would not be usually possible to detect differences of 12% of SOC to depth of 10 cm using 50 samples over a 10 ha area. If the depth of measurement was increased to 100 cm, it would not usually be possible to detect changes of 18% of SOC. This study is important for designing sampling methods to detect changes in SOC in grasslands. Changes in grassland SOC provide information needed to manage the concentration of carbon dioxide in the atmosphere.
Abstract
There is an important need to better understand how the soil organic carbon (SOC) stocks affect the global atmospheric C balance. Grasslands represent 40% of the earth's land surface but efficiently and effectively quantifying their SOC stocks and their changes is a challenge. Although factors influencing the variability of grassland SOC stocks such as climate, management, or topography have been investigated, very few studies have quantified the uncertainty of SOC stock measurement. Quantifying this uncertainty is critical to determine our ability to detect changes in space or time, and ultimately to develop guidance to help designing appropriate measurement strategies for quantifying carbon stocks and stock changes. SOC stock measurement may be performed at various spatial scales, from local (≤1 km2) to broader scale (≥100 km2) applications. In addition, the recommended sampling depth for SOC measurement varies according to project purposes, national circumstances, and land use. To our knowledge, the assessment based on world literature of the effect of spatial scale (i.e. size of the measured area) and sampling depth on the uncertainty of SOC stock measurement in grasslands has never been performed. We quantified the uncertainty of SOC stock measurement from a global analysis of 51 research articles meeting strict requirements of rigour of SOC measurements, totaling 177 grasslands from 19 countries, to assess the effects of the spatial scale and the sampling depth of soil profile on this uncertainty, and to explore the implications for SOC stock change detection involving future sampling. We observed that spatial scale and soil profile depth combined to explain 44% of the variability of SOC stock uncertainty, as measured by the coefficient of variation (CV). More specifically, the CV increased with the spatial scale and soil profile depth measured. However, the sampling depth effect is less certain due to lack of data at deeper depths. This uncertainty associated with SOC stock measurements has ramifications for SOC stock change detection. For example, at a fixed 0–30 cm soil profile depth, the minimum detectable change (MDC) of SOC stocks between two sampling dates (50 samples) was 14, 20 and 29% at 0.1, 10 and 10 000 km2, respectively. At a fixed spatial scale of 0.1 km2, the MDC was 12, 14 and 18% for soil profiles of 0–10, 0–30 and 0–100 cm, respectively. Finally, this study provides global estimates of the uncertainty that will be useful for planning sampling strategies for SOC stock measurement in various projects across the world and to evaluate the feasibility of SOC stock measurement for different investment levels and timescales.