A global database of land management, land use change and the effects of climate change on soil organic carbon

Data collection

A literature search was conducted on 09 January 2020 (Fig. 1). The following search equations were used: (“meta*analysis” OR “systematic review”) AND (“soil organic carbon” OR SOC OR “soil organic matter” OR SOM OR “soil carbon”) in “topic words”, that is titles, abstracts and keywords from the following databases:

  • Web of Science, New York, USA, http://apps.webofknowledge.com, includes 12,000 journals and 160,000 conference proceedings.

  • Scopus, USA, https://www.scopus.com/search. The Scopus database includes more than 41,000 reference journals.

  • OVID. Publisher, USA. https://www.ovid.com. The Ovid database includes more than 10,000 titles of scientific journals, books, and proceedings (for Cab Abstracts on Ovid).

  • Google Scholarship. https://scholar.google.com/. Publisher: Google. It contains both, a multidisciplinary peer review and gray literature. We selected the first 150 search results, arranged by relevance, because this engine is very precise for the first page of results displayed23.24.

Figure 1

The methodological framework used to identify and characterize the data included in the database. The study selection criteria were (1) duplicates removed; (2) only studies published in English (non-English studies: n = 7) with available text (studies with unavailable text: n = 7) were considered; (3) studies not related to SOC were excluded; (4) only meta-analyses were included; (5) the meta-analysis presents at least one effect measure, i.e. a quantitative measure on SOC (or one described effect size for different levels of SOC content, i.e. SOC as a covariable). The hexagons represent the different characteristics analyzed in the meta-analysis or in the main study.

Following the recommended systematic review gold standard25,26, we used these disparate databases with varying journal coverage to obtain a complete literature search and to avoid potential bias. Search is supported by librarians to further help reduce the possibility of bias and improve the quality of the overall search strategy27.

No publication year limit is applied. All climatic zones and countries are considered. Sensitivity is preferred over specificity. Sensitivity implies that the emphasis in the search procedure is placed on gathering the largest selection of potentially relevant studies at risk as well as obtaining a large number of irrelevant studies (thus increasing the duration of the screening step). In addition to database searches, a number of other potentially relevant meta-analyses were added by the authors of this study.

A literature search identified 1,535 studies (of which 1,008 were unique). These studies were compiled in a database and then screened to identify relevant studies based on the following inclusion and exclusion criteria: (i) only studies published in English with available full texts were considered; (ii) this study presents a formal quantitative analysis of several previous empirical studies, that is meta-analysis (we did not consider vote-counting studies and narrative reviews were also excluded); (iii) the meta-analysis presents at least one effect size, that is a quantitative measure of magnitude, the SOC variable, either as a primary variable, or as a co-variable.

The studies were screened first by title and abstract, and if necessary, the entire manuscript was read. Each study was screened by the two authors of this article. Rejected studies are compiled in a database exception sheet, citing exceptions. Finally, 217 meta-analyses met our inclusion/exclusion criteria, 18 of them with SOC as a co-variable.

Meta-analysis characterization

Our database reports meta-data (name of author and affiliation, name of journal, keywords, date of publication, and countries covered in the meta-analysis) for the meta-analyses considered (Fig. 1). Transparency and reproducibility of each of the 217 meta-analyses were assessed based on criteria related to literature search, data extraction, data analysis, and interpretation. These criteria are adaptations of those proposed in several other studies covering various research fields28,29,30,31. When satisfied, the criteria are given a score of 1, and 0 otherwise. A global quality score is given by calculating the proportion of criteria that are met.

We also classified each meta-analysis for its scope, describing whether they addressed the dimensions of “climate change mitigation”, “adaptation to climate change” and “food security”. Dimension definition based on IPCC glossary32and goals of the 4p1000 Initiative15 (https://www.4p1000.org/). The keywords associated with each dimension (Table 1) were defined for the classification, which was performed manually by two different authors of this study. Titles and abstracts are filtered, and the full text is studied if necessary. Consistency between reviewers was checked on a sample of 30 studies. The final database consisted of 199, 28 and 54 meta-analyses, respectively, analyzing the dimensions of mitigation, adaptation and food security.

Table 1 Keywords associated with each of the 4 per 1000 dimensions.

Effect size characterization

We extracted all quantitative data related to the effects reported in the meta-analyses taken: effect sizes, dispersion indicators (confidence intervals, and/or standard deviations or quantiles), significance (P-values) and the amount of data for which effects were calculated (Fig. 2). The database contains an effect size that measures the direct effect of the intervention on SOC, and an effect size that assesses the effect on other outcomes but with SOC as a co-variable (indirect effect). Data were collected from tables or from figures using the WebPlot Digitizer software (www.automeris.io/WebPlotDigitizer/). We also describe the types of metrics associated with each effect size (for example mean difference, ratio, hedge d).

Figure 2.
Figure 2

Number of effect sizes (left) and number of meta-analyses (right) available in the database per intervention type and land use. The total number of effect sizes reported in the database is presented, along with the total number of effect sizes on SOC (dark shade) or on other results (light shade). For land use change interventions, initial and final land uses are considered in this plot. A meta-analysis may consist of different interventions or land uses.

We did not extract data from meta-regressions, correlations, and their associated characteristics due to the difficulty of synthesizing these types of results across studies. However, when a subgroup analysis was performed (for example, based on soil characteristics or climatic zones – Table 2), we extracted moderator effects to analyze and understand the variability of SOC values.

Table 2 Co-variables reported in the meta-analysis.

Interventions related to effect sizes were grouped into main categories: land management, land use change, and global change (Fig. 2, 3). We consider land use types as defined in the IPCC Guidelines for National Greenhouse Gas Inventories33: agricultural land, forest land, grasslands, wetlands and other lands. We consider land use change as a conversion from one of the land uses mentioned above to another. We consider land management as an intervention carried out on one of the above mentioned land uses (for example forest harvesting, wetland restoration, mineral fertilization). We define global change as planetary scale change other than land use change (for example climate change). The database presents broad categories of interventions, but also more detailed interventions reported in each study (Fig. 3). The number of effect sizes and meta-analyses is mostly dominated by land management studies, especially for agricultural land. This is followed by land use change interventions and global types of change, where the distribution of land use is more balanced.

Figure 3
picture 3

Categories used to characterize interventions for global change, land use change and management.

Outcomes other than those related to SOC are grouped into seven broad categories: soil chemistry, crop productivity, soil physics, soil biology, greenhouse gases, water quality, and others. The first three categories represent nearly 20-30% of the reported effect sizes. Each category is further refined into 2 to 11 subcategories (Fig. 4). The subcategories of soil nutrients and aboveground biomass alone represent almost 15-20% of the effect size.

Figure 4
picture 4

The main categories and other effect size subcategories were drawn in 217 meta-analyses and studied concurrently with SOC. The area is proportional to the number of effect sizes in the database.

Main study characterization used in 217 meta-analyses

We retrieved all available references (main study) used by the 217 meta-analyses, by searching through the available reference lists, supplementary materials and databases associated with each meta-analysis.

The main studies were characterized by their meta-data (for example DOI, author, publication date, journal). Based on the title and abstract, if necessary, we also manually extracted the types of interventions and outcomes associated with the main study. Manual classification into the same intervention and outcome categories as previously described, was facilitated by automatic classification by keyword (Supplementary Table 1). The final database consisted of 13,632 unique primary studies (Fig. 5). 9,130 ​​primary studies were used in multiple meta-analyses. Geographic distribution shows the highest number of these main studies in the United States and China, followed by Brazil and Canada, and then Australia, India and several European countries (UK, Germany, Spain, and Italy). The regional distribution in the five countries with the largest number of studies shows large regional disparities. Africa is the least investigated continent; no major studies have been conducted in several African countries.

Figure 5
number 5

The geographic distribution of the 13,632 main studies included in the 217 meta-analyses (a), with details of the distribution of provinces/states for (b) United States of America, (c) China, (d) Brazil, (e) Australia and (f) India.

Leave a Comment

%d bloggers like this: