Blind spots in global soil biodiversity and ecosystem function research

Soils are vital ecosystems, housing a significant portion of global biodiversity and driving various crucial ecosystem functions and services. This biodiversity includes microorganisms (bacteria), microfauna (*Nematoda*), mesofauna (*Collembola*), and macrofauna (*Oligochaeta*), all of which play critical roles in climate regulation, nutrient cycling, and food production. Recent studies, both experimental and observational, have highlighted the importance of soil biodiversity for maintaining ecosystem multifunctionality—the ability of ecosystems to simultaneously provide multiple services. However, despite this importance, global soil biodiversity-ecosystem function relationships haven't been thoroughly examined from a macroecological perspective. Unlike aboveground biodiversity, which has extensive global datasets, soil macroecology suffers from significant data limitations, especially concerning temporally explicit data. Existing studies reveal major geographic and taxonomic biases, with gaps in our understanding of soil biodiversity, particularly across ecological gradients and specific taxa. The lack of data on temporal patterns at larger scales further hinders comprehensive assessments. This limited knowledge base restricts our ability to identify ecological preferences of soil taxa, assess vulnerabilities to global change, and understand causal links between biodiversity, ecosystem functioning, and services. Furthermore, the absence of soil biodiversity and function data in land management and conservation debates reflects a critical oversight in environmental policy. The Global Biodiversity Information Facility (GBIF), a crucial global data hub for aboveground biota, poorly represents soil organisms. Data are scattered across various platforms, hindering aggregation and comprehensive analysis. Major issues remain regarding the spatial and temporal representativeness of available data, including underrepresentation of tropical systems and taxonomic groups. Even for well-represented taxa (bacteria and fungi), concerns exist regarding taxonomic depth. This lack of representativeness hinders the prioritization of future monitoring efforts and prevents stakeholders from implementing effective management actions to preserve crucial ecosystem services. This research aims to identify these blind spots in global soil macroecological knowledge, assess their causes, and propose solutions to overcome them. The study goes beyond simply identifying data gaps; it examines how well existing macroecological studies cover Earth's environmental conditions (soil properties, climate, topography, land cover) to pinpoint critical ecological and geographical blind spots.

The authors conducted a comprehensive literature review to identify existing macroecological studies on soil biodiversity and ecosystem functions. Their search strategy included keywords related to global patterns, soil characteristics, biodiversity metrics (such as diversity, abundance, biomass), and ecosystem functions (decomposition, respiration, nutrient cycling). The search encompassed a wide timeframe and multiple databases. The inclusion criteria for the studies were stringent: studies had to cover more than one continent, extend across at least one entire continent, and present spatially explicit data at the sampling site level. After screening, 62 studies fulfilled these criteria, covering diverse taxa (bacteria, archaea, fungi, protists, nematodes, rotifers, collembolans, acarids, formicoids, oligochaetes) and functions (decomposition, soil respiration, nutrient cycling, water infiltration, bioturbation, soil aggregate stability). This literature review highlighted the uneven distribution of research efforts, with certain taxonomic groups and functions being overrepresented in comparison to others.

The authors collected data from 62 studies that met specific criteria, focusing on macroecological patterns of soil biodiversity and ecosystem functions across global gradients. These criteria ensured the studies encompassed multiple continents, spanned at least one entire continent, and provided spatially explicit sampling site data. The dataset included information on ten different soil taxa and five ecosystem functions relevant to biogeochemical cycles and ecosystem services. The locations of sampling sites were georeferenced and projected to WGS84. The authors analyzed the spatial distribution of sampling sites to identify geographic biases, using a 1° grid resolution. They compared the distribution of sampling sites with the global distribution of environmental variables such as soil properties (carbon content, pH, texture), climate variables (temperature, precipitation, seasonality, aridity, potential evapotranspiration), and land cover types. To assess the extent of global coverage for each variable, they created histograms and used a Jenks natural breaks classification to group similar values. The comparison of the study-specific distributions with the global distribution of these variables allowed them to identify over- and underrepresented conditions in soil macroecological studies. They further analyzed environmental combinations by combining land cover, soil, and climate variables. A Mahalanobis distance was used to calculate multidimensional distance for all continuous environmental variables from each study, allowing detection of outliers. This method incorporated multiple variables to evaluate how well the studies represented the range of environmental conditions. The authors also analyzed the overlap between biodiversity and function data in studies and highlighted the near complete absence of temporally explicit data.

The analysis of 17,186 sampling sites from 62 studies revealed significant biogeographical biases. Bacteria, fungi, and Formicoidea were the most represented taxa, while Rotifera, Collembola, and Acari had substantially fewer sampling sites and more scattered distributions. Temperate biomes were significantly overrepresented, whereas tropical and subtropical regions, montane grasslands, and hyper-arid areas were underrepresented. Soil respiration was the most frequently studied function, comprising 78.8% of all function records. A striking finding was the extremely low overlap (0.3%) between biodiversity and function datasets, with only 67 sites having both types of data. The majority of studies lacked repeated measurements over time, leading to a near absence of temporally explicit data on soil biodiversity and functions. Analysis showed that the studies tended to cluster geographically in temperate regions. The range of environmental conditions covered in the studies was limited, particularly in terms of extreme soil carbon levels, soil types, climatic conditions (e.g., low and high potential evapotranspiration areas, high climate seasonality regions), and land cover types (e.g., urban areas were overrepresented compared to shrublands and lichen-dominated areas). Most importantly, many climate and soil property combinations were not covered, leading to gaps in our understanding of these relationships. The analysis using Mahalanobis distance revealed that most studies had less than 50% environmental coverage across various regions, with mega-diverse areas often underrepresented.

The findings highlight substantial gaps and biases in global soil macroecological research. The limited spatial coverage, particularly the underrepresentation of tropical and subtropical regions, hinders our understanding of soil biodiversity patterns and their response to global change. The lack of temporal data prevents the assessment of long-term trends and the impacts of environmental change on soil ecosystems. The low overlap between biodiversity and function studies reveals a significant limitation in our understanding of the links between these two critical aspects of soil ecology. The uneven distribution of studies across environmental conditions, particularly the underrepresentation of extreme conditions, suggests potential biases in our knowledge of soil ecosystem functioning. Addressing these limitations requires multi-faceted efforts, from improved data mobilization and standardization of sampling protocols to coordinated global monitoring initiatives and enhanced funding for soil ecological research. Such initiatives should focus on addressing the spatial, taxonomic, and functional gaps identified in this study.

This study reveals significant blind spots in global soil macroecological research, characterized by biogeographical biases, limited taxonomic and functional coverage, and a near absence of temporally explicit data. These gaps restrict our capacity to understand biodiversity-ecosystem function relationships and predict the impacts of global change on soil ecosystems. Overcoming these limitations requires a concerted effort to improve data availability through data mobilization initiatives, develop standardized protocols for sampling and analysis, and implement coordinated global monitoring programs that prioritize underrepresented regions and extreme environmental conditions. Future research should prioritize integrating biodiversity and ecosystem function assessments in the same locations and, ideally, from the same soil samples, expanding the temporal and spatial coverage of global soil biodiversity data.

The study relies on existing published literature, and the results reflect the inherent biases and limitations of the available data. The inclusion criteria for studies, while stringent, may have inadvertently excluded relevant research. The authors acknowledge that the lack of standardized protocols for sampling and analysis across the included studies limits the possibility of making some comparisons. The absence of detailed information on specific sampling methods in some studies also introduces uncertainty into the analysis. Finally, the reliance on publicly available datasets restricts the investigation of specific regions or taxa.