Global meta-analysis of the relationship between soil organic matter and crop yields

The researchers investigated the quantitative relationship between **soil organic matter (SOM)**, specifically measured as **soil organic carbon (SOC) concentration**, and **crop yields** for maize and wheat. They aimed to understand how changes in SOC levels impact crop productivity.

They also examined how other factors might influence this relationship, including:

**Nitrogen (N) fertilizer input rate:** The amount of nitrogen fertilizer applied to the soil.

**Irrigation:** Whether the crops were irrigated or rain-fed.

**Soil pH:** The acidity or alkalinity of the soil.

**Soil texture (% clay):** The proportion of clay particles in the soil, which affects water retention and nutrient availability.

**Aridity:** The dryness of the climate.

**Crop type:** Specifically differentiating between maize and wheat.

**Latitude:** Used as a proxy for growing-season day length, which influences plant growth.

The primary outcome measure was **crop yield** (e.g., bushels per acre or tons per hectare). Secondary outcomes included the **potential for reducing nitrogen fertilizer reliance** and the **potential for closing global yield gaps** (the difference between actual and potential yields).

This study was a meta-analysis, meaning it did not involve new experiments with human or animal subjects. Instead, it synthesized data from a large collection of **published empirical studies** conducted globally.

**Sample size:** The study analyzed data from numerous published empirical studies, though the exact number of individual studies included is not specified in the abstract. The authors state that "Ninety-one percent of the published studies used for our analysis were carried out in fields with less than 2% SOC, with a mean of 1.1%," indicating a substantial dataset of field observations.

**Population:** The data represented agricultural fields cultivating **maize (corn)** and **wheat** across diverse geographical locations and climatic zones worldwide. These two crops were chosen because they are staple foods, constituting two-thirds of the energy in human diets.

**Setting:** The studies included in the meta-analysis were conducted in various agricultural settings globally, reflecting a wide range of management practices, soil types, and climates. This global scope allowed the researchers to identify general relationships between SOC and yield that transcend local variability.

As a meta-analysis, the researchers did not directly perform measurements but rather extracted and standardized data from existing published studies.

**Soil Organic Carbon (SOC):** The primary variable of interest, SOC, was measured as a percentage of the soil's dry weight. SOC is a common and widely accepted proxy for total soil organic matter (SOM). The specific laboratory methods used to determine SOC would have varied across the original studies (e.g., dry combustion, Walkley-Black method), but the meta-analysis aggregated these reported values.

**Crop Yield:** Yields for maize and wheat were reported in various units (e.g., kg per hectare, tons per acre) in the original studies and were standardized for the meta-analysis to allow for comparison across different studies and regions.

**Co-varying factors:**

* **Nitrogen (N) input rate:** Reported in kilograms of nitrogen per hectare per year (kg N ha−1 yr−1).

* **Irrigation:** Categorized as either present or absent (rain-fed).

* **Soil pH:** Measured on a standard pH scale.

* **Soil texture (% clay):** Reported as the percentage of clay particles in the soil.

* **Aridity:** A measure of dryness, likely derived from climatic data associated with the study locations.

* **Crop type:** Categorical variable (maize or wheat).

* **Latitude:** The geographical latitude of the study site.

The researchers also utilized globally gridded data on crop yield (from Monfreda et al., 2008) and SOC (to a depth of 15 cm, from Hengl et al., 2014) to assess global patterns of SOC distribution in cultivated lands, comparing their meta-analysis findings to broader global agricultural realities.

This study employed a **global meta-analysis** design, which is a statistical method used to combine the results of multiple scientific studies. The goal was to synthesize existing empirical data to develop a **quantitative model** that describes the relationship between soil organic carbon (SOC) and crop yields, while accounting for other influential factors.

**How they ran the study:**

1. **Data Collection:** The researchers systematically gathered data from a wide range of published empirical studies that reported on SOC levels and maize or wheat yields, along with other relevant co-varying factors (N input, irrigation, pH, soil texture, aridity, crop type, latitude). This involved a comprehensive literature search to identify relevant studies.

2. **Data Standardization:** Data extracted from different studies, which might have used varying units or reporting formats, were standardized to ensure comparability.

3. **Quantitative Model Development:** A **multiple-regression model** was developed. This statistical approach allowed the researchers to simultaneously assess the influence of multiple independent variables (SOC, N input, irrigation, pH, etc.) on the dependent variable (crop yield). The model aimed to isolate the effect of SOC on yield while controlling for the confounding effects of other factors. They specifically included a quadratic term for SOC (SOC^2) to capture potential non-linear relationships, such as a leveling-off effect at higher SOC concentrations.

4. **Global Pattern Assessment:** To contextualize their findings, the researchers also used existing globally gridded datasets for crop yield and SOC content (to a depth of 15 cm) to estimate the proportion of global maize and wheat lands currently below certain SOC thresholds.

5. **Scenario Modeling:** Using their developed regression relationship, they then estimated the potential impacts of increasing SOC concentrations to regionally specific targets. This involved calculating potential reductions in N fertilizer requirements and potential increases in yield, and how these might contribute to closing global yield gaps.

**Why this design matters:**

**Synthesizing diverse evidence:** A meta-analysis is crucial for drawing general conclusions from a large body of research that might otherwise appear contradictory due to local variations. By combining data from numerous studies across different climates, soil types, and management practices, it helps identify overarching patterns and quantify average effects.

**Controlling for confounds:** The multiple-regression approach is vital because the relationship between SOC and yield is complex and influenced by many interacting factors. By including variables like N input, irrigation, pH, and climate in the model, the researchers attempted to statistically "control" for these co-varying factors, allowing them to better isolate the potential effect of SOC itself on yield. This helps to address the challenge that "local-scale differences in soils, climate, and farming systems" can confound the SOM-yield relationship.

**Quantifying relationships:** Unlike qualitative reviews, this meta-analysis aimed to provide specific quantitative estimates (e.g., "yields are 1.2 times higher at 1.0% SOC than 0.5% SOC"), which are essential for setting targets and informing policy.

**What this design can and cannot prove:**

**Can prove:** This meta-analysis can establish **strong correlative relationships** between SOC and crop yields at a global scale, providing a robust average estimate of the effect size. It can identify thresholds (like the ~2% SOC leveling-off point) where the relationship changes. By controlling for multiple variables, it strengthens the argument for a relationship, even if not strictly causal in every individual study. The asymptotic relationship observed (yields leveling off) provides some evidence that building SOC *causes* yield increases, rather than just being an *outcome* of higher yields (which would likely show a more linear relationship).

**Cannot prove:** While the study provides strong evidence for a relationship, it **cannot definitively prove direct causation** in all cases. Many of the underlying studies included in the meta-analysis might be observational or correlational themselves. The authors explicitly acknowledge the "correlative nature of the database we assembled." While they note that "experimental evidence showing that building SOM positively affects yield" exists, their own model is built on aggregated data that may include both experimental and observational studies. Therefore, while the findings suggest that increasing SOC *leads to* higher yields, it's important to remember that high yields can also contribute to higher SOC (through greater plant biomass inputs). Disentangling these bidirectional causal effects fully would require more controlled, long-term experimental studies specifically designed for that purpose.

**Major