|Main authors:||Else K. Bünemann, Giulia Bongiorno, Zhanguo Bai, Rachel E. Creamer, Gerlinde De Deyn, Ron de Goede, Luuk Fleskens, Violette Geissen, Thom W. Kuyper, Paul Mäder, Mirjam Pulleman, Wijnand Sukkel, Jan Willem van Groenigen and Lijbert Brussaard|
|Source document:||Bünemann, E. K. et al. (2018) Soil quality - A critical review. Soil Biology and Biochemistry, Volume 120, May 2018, pp 105-125
Adoption of additional or novel soil quality indicators into minimum datasets is of interest if they have clear added value from the perspective of the management goals for a particular situation. Recent developments in soil science, especially in soil biology, but also in spectroscopy and other fields, hold promise for future soil quality assessment schemes. Below, we briefly review these developments, from biological and biochemical indicators to data capture and high-throughput
Supplementary Table 6. Groups of soil organisms as indicators: relation to main soil functions, mechanisms involved and soil-based ecosystem services and ease of application
|Soil organism||Main soil functions||Mechanisms involved||Soil-based ecosystem services||Ease of application||References|
|Earthworms (macrofauna)||Soil structure maintenance, decomposition, organic matter and water cycling, habitat provision||Burrowing, fragmentation of litter, soil aggregation, humification, organic matter distribution||Biomass production, erosion control, water supply, climate regulation, biodiversity conservation||Easy to sample but not ubiquitous||(Blouin et al., 2013; Lavelle et al., 2006)|
|Nematodes (microfauna)||Element cycling, decomposition, biological population regulation||Grazing on microorganisms, root herbivory, predation||Biomass production, pest and disease control||Identification via morphology currently only by specialists, but facilitated by molecular tools in the future. Ubiquitous, easy to sample, abundant, sensitive. Key role in soil food web. Information about feeding preferences and life strategy.||(Mulder et al., 2005; Neher, 2001; Schloter et al., 2003)|
|Protists (microfauna)||Element cycling, biological population regulation||Grazing on microorganisms||Biomass production||Poorly defined taxonomically, difficult to isolate and identify. Variable in space and time.||(Foissner, 1999; Riches et al., 2013)|
|Collembola (mesofauna)||Decomposition, element cycling, biological population regulation||Grazing on fungi||Biomass production, pest and disease control||Cumbersome to sample and isolate, difficult to identify||(Brussaard et al., 2004; Cardoso et al., 2013; Pulleman et al., 2012; Ruf et al., 2003)|
|Enchytraeids (mesofauna)||Decomposition, soil structure maintenance||Burrowing, fragmentation of litter, soil aggregation, decomposition, humification, organic matter distribution||Water supply, climate regulation||Easy to sample but difficult to identify|
|Mites (mesofauna)||Decomposition, element cycling, biological population regulation||Grazing on bacteria and fungi, fragmentation of residues||Biomass production, pest and disease control||Cumbersome to sample and isolate, difficult to identify|
|Macroarthropods (macrofauna)||Soil structure maintenance, biological population regulation||Burrowing, root herbivory, predation, grazing on bacteria and fungi||Biomass production, pest and disease control, biodiversity conservation||Relatively easy to sample, taxonomically very diverse|
|Bacteria||Element and organic matter cycling, decomposition, biological population regulation||Symbiotic association (nitrogen fixing bacteria), production of antibiotics, transformation and mineralization of organic material||Biomass production, pest and disease control, climate regulation||Spatially and temporally variable. Taxonomically very diverse and difficult to classify.||(Barrios, 2007; Lehman et al., 2015; Schloter et al., 2017)|
|Fungi||Element, organic matter and water cycling, soil structure maintenance, decomposition, biological population regulation||Symbiotic association (mycorrhizae), production of antibiotics, transformation and mineralization of organic material||Biomass production, water quality and supply, erosion control, pest and disease control, climate regulation|
Soil organisms play a central role in soil functioning (Supplementary Table 6). Therefore, adding biological and biochemical indicators can greatly improve soil quality assessments (Barrios, 2007). Moreover, the assessment of biological indicators of soil quality is required to connect abiotic soil properties to (changes in) soil functions in terms of biochemical and biophysical transformations and (potential) aboveground vegetation performance (Lehman et al., 2015). Nevertheless, soil biological indicators are still underrepresented in soil quality assessments and mostly limited to black-box measurements such as microbial biomass and soil respiration (»Frequently proposed sol quality indicators Figure 4, Table 4). Despite clear potential, more specific indicators such as those based on nematodes (Stone et al., 2016b), (micro)arthropods (Rüdisser et al., 2015) or a suite of soil biota (Velasquez et al., 2007) have rarely been suggested, possibly because they require specific knowledge and skills. This situation is unfortunate because soil biota are considered the most sensitive indicators of soil quality due to their high responsiveness to changes in environmental conditions (Bastida et al., 2008; Bone et al., 2010; Kibblewhite et al., 2008a; Nielsen and Winding, 2002). In particular, there is an urgent need for indicators of soil–borne diseases (Kyselková et al., 2014; Liu et al., 2016; Trivedi et al., 2017). In this context, soil suppressiveness, defined as the property of a soil to naturally reduce plant disease incidence (Hornby, 1983), is of interest. Specific soil suppressiveness is the result of the presence of specific antagonists to pathogens, while general soil suppressiveness is based on the collective capacity of soil and plant microbiomes to act complementarily against pathogens (Schlatter et al., 2017). Both combined are governing soil suppressiveness as a whole (Yadav et al., 2015). Several soil abiotic and biotic parameters have been suggested to underlie suppressiveness, such as soil pH, specific cations such as Mg and K, soil total N content, microbial biomass and activity, diversity and structure of microbial communities and specific microbial taxa in the case of specific suppressiveness (Janvier et al., 2007; Wu et al., 2015), but without validation.
Table 4: Soil biological indicators, methodologies, related main soil functions, and advantages/disadvantages at different scales.
Table compiled from (Bastida et al., 2008; Blagodatskaya and Kuzyakov, 2013; Bloem et al., 2009; Bouchez et al., 2016; Brussaard, 2012; Brussaard et al., 2004; Cardoso et al., 2013; de Groot et al., 2014; Gil-Sotres et al., 2005; Lehman et al., 2015; Neher, 2001; Nielsen and Winding, 2002; Orgiazzi et al., 2015; Parisi et al., 2005; Rocca et al., 2015; Saleh-Lakha et al., 2005; Schloter et al., 2017; Stenberg, 1999; Torsvik and Ovreas, 2002; Trasar-Cepeda et al., 2008; Visser and Parkinson, 1992; Watzinger, 2015).
|INDICATOR||METHODOLOGY||MAIN SOIL FUNCTIONS||MAIN PROS||MAIN CONS|
|Individual, population and community level|
|Presence, richness, abundance of individual soil organisms(for details see Supplementary Table 6).||Traditional handsorting and microscopic methods; molecular quantitation (qPCR).||Element, organic matter and water cycling, biological population regulation, soil structure maintenance.||Taxonomic and functional level.||Not always linked directly with functions. Difficult to apply to fauna, e.g. protozoa, mites and collembola.|
|Microbial biomass and fungal biomass, fungal:bacteria ratio.||Direct counting, chloroform fumigation extraction, SIR, PLFA, molecular quantitation.||Element and organic matter cycling, decomposition, soil structure maintenance.||Sensitive and well related with other soil quality indicators.||Spatially variable, difficult interpretation, contradictory results. Unclear direct link to functionality.|
|Indices based on faunal communities (e.g. Maturity Index, Enrichment Index, Channel Index, Structural Index for nematodes).||Counting and identification of specific groups of organisms.||Element and organic matter cycling, biological population regulation, decomposition||Sensitive. Taxonomic and functional level.||Time-consuming and costly. Specialist required for morphological identification.|
|Community composition||Manual counting and identification||Element and organic matter cycling, biological population regulation, habitat provision, decomposition, soil structure maintenance||Division in functional groups can give an indication of functions.||Time-consuming, expertise required. Not indicative of active biota.|
|PLFA||Correlated with other measurements. Good indicator of active microbial biomass. Integrated information on the microbial community.||Time-consuming. No direct link with functions. Coarse resolution.|
|Fingerprinting methods (e.g. DGEE, T-RFLP, A-RISA, ARDRA, TGGE), microarrays||Greater phylogenetic resolution.||No direct link with function. Difficult comparison between studies due to great variety in methods. Difficulties to extract and amplify DNA.|
|Sequencing (metabarcoding)||Detailed view of diversity. Enormous amounts of data. Detects less abundant organisms. Permits discovery of new diversity.||Taxonomic genes no direct link with functions. Difficulties to extract and amplify DNA. Costly. Problems related with handling of large datasets and analyses. Dependent on libraries. No standard methodology.|
|Community Level Physiological Profiling (Biolog™, MicroResp)||Element and organic matter cycling, decomposition, habitat provision||Insight into functionality of the community. MicroResp closer to in situ conditions, shorter time of measurements.||Many replicates needed because of variability.|
|Soil respiration, nitrogen mineralization, denitrification, nitrification||CO2 evolution, N2O emission, NO3 produced.||Element, organic matter and water cycling, decomposition, habitat provision
||Sensitive and ecologically relevant.||Highly variable and fluctuating. Relatively laborious.|
|Potentially mineralizable nitrogen||Anaerobic incubation.||Good correlation with MB and total soil N.||Relatively laborious.|
|Metabolic quotient (qCO2), microbial quotient (MicrC/SoilC)||Thymidine and leucine DNA incorporation.||Sensitive, simple and inexpensive.||Difficult interpretation: confounds disturbance with stress.|
|DNA and protein synthesis.||Reflection of active microbial biomass.||No standardized procedure.|
|Enzymatic activities||Extraction of enzymes in the soil and incubation with various substrates.||Element and organic matter cycling, decomposition, biological population regulation.||Closely related to important soil quality parameters. Very sensitive. Simple and inexpensive methods.||Standard procedure not available. Contradictory results, complex behaviour and variable for each enzymes. Potential activity.|
|Functional genes and transcripts||FISH, Microarrays, meta-transcriptomic, qPCR, metagenome analysis.||Closer link to functionality. FISH and microarrays can give an idea of active microorganisms. High sensitivity and throughput.||Restricted to known gene sequences. Genes and transcripts might not be expressed. Difficulties linked with RNA extraction. Costly.|
|Metabolomics and metaproteomics||Assessment and quantitation of metabolites and proteins in the soil.||Element and organic matter cycling, decomposition, biological population regulation, soil structure maintenance||Closer link to functionality.||Field in development. Difficult extraction of metabolites and proteins.|
|Stable isotope probing||Incorporation of 13C- or 15N-labelled substrates into DNA, RNA, PLFA, proteins||Element and organic matter cycling, decomposition||Permit to establish link between biodiversity and functions. Allow in situ analysis of active microbial population.||Field in development. Time involved in the assimilation of the substrates.|
Recent rapid developments in soil biology have prompted the feasibility of indicators based on genotypic and phenotypic community diversity (Hartmann et al., 2015; Kumari et al., 2017; Nielsen and Winding, 2002; Ritz et al., 2009). Molecular methods focusing on DNA and RNA hold great potential to perform faster, cheaper and more informative measurements of soil biota and soil processes than conventional methods (Bouchez et al., 2016). Consequently, they may yield novel indicators that could substitute or complement existing biological and biochemical soil quality indicators in regular monitoring programs (Hartmann et al., 2015; Hermans et al., 2017). In the participatory approach used by Stone et al. (2016a), seven out of ten selected indicators were indeed based on molecular methods, with ‘molecular bacteria and archaea diversity’ on top. In addition, recent data analysis approaches such as network analysis, structural equation modelling and machine learning could facilitate the establishment of links between indicators and functions (Allan et al., 2015; Creamer et al., 2016). For example, Karimi et al. (2017) proposed microbial networks as integrated indicators of environmental quality that can overcome the lack of sensitivity and specificity of taxonomic diversity indicators. However, the prediction of process rates from the presence and quantity of genes and transcripts is yet to be clearly established (Rocca et al., 2015). Results gathered with these molecular techniques are also faced with biases introduced by sample contamination, PCR reaction, choice of primers and OTU definition and taxonomic assignment techniques (Abdelfattah et al., 2017; Hugerth and Andersson, 2017; Schloter et al., 2017). The analysis of the “big data” generated with sequencing also poses a serious challenge in terms of time, computing capacities and interpretation, since a large proportion of soil organisms yet remains to be characterized in taxonomic and functional terms (Schloter et al., 2017; Bouchez et al., 2016). Other molecular techniques such as metabolomics (Vestergaard et al., 2017) and metaproteomics (Simon and Daniel, 2011) may yield potentially suitable soil quality indicators because the measurements are directly linked to ecosystem processes (Bouchez et al., 2016). These technologies have benefits but are limited in their application by the difficulty to extract metabolites and proteins from soil and to choose representative samples (Bouchez et al., 2016). Stable Isotope Probing (SIP) in conjunction with phospholipid fatty acid analysis (PLFA) and DNA probing could also help to link soil biodiversity to soil processes (Wang et al., 2015; Watzinger, 2015). Finally, for a meaningful integration of indicators based on molecular methods into soil quality assessments, standardized techniques and a reference system are still lacking and will have to be established (Bouchez et al., 2016).
Although total soil organic matter is ubiquitous as a soil quality indicator (Figure 4), changes in response to management and land use are difficult to detect since the total pool is large (Haynes, 2005). Moreover, due to the structural and functional heterogeneity of total soil organic matter, its relevance in soil processes is not unequivocal. Therefore, qualitative information on soil organic matter may be more informative in soil quality assessments. Pools of soil organic matter such as labile or active carbon are typically more sensitive to disturbance than total soil organic matter and can give a better indication about soil processes (Gregorich et al., 1994). Suggestions to measure this fraction include: particulate organic matter (Cambardella and Elliott, 1992), permanganate-oxidizable carbon (Weil et al., 2003), hot water-extractable carbon (Ghani et al., 2003) and water-soluble carbon, also called dissolved organic carbon (Filep et al., 2015). Despite their sensitivity to management and strong correlations to other parameters that are more difficult to measure, their relationship with soil processes is not well understood, partly because it is not clear which part of the organic matter they represent. Other methods to characterize (quality and quantity) of total soil organic matter such as thermal and spectroscopic methods are rapidly developing (Clemente et al., 2012; Derenne and Quénéa, 2015; Mouazen et al., 2016) and hold promise for soil quality assessments.
Additionally, soil sensing approaches such as spectroscopic techniques, e.g. near-infrared spectroscopy and remote sensing, offer the opportunity to measure various soil chemical, physical and biological parameters in a fast and inexpensive way (e.g. Cecillon et al., 2009; Gandariasbeitia et al., 2017; Kinoshita et al., 2012; Paz-Kagan et al., 2014). Sensors can be used directly in the field or in the laboratory (McKenzie et al., 2003), and commercial providers increasingly offer spectroscopy-based analyses (e.g. »www.soilcares.com, www.eurofins.com). Combining laboratory-based visible and near-infrared spectroscopy with in situ measurements such as electrical conductivity and penetration resistance may be particularly useful (Veum et al., 2017). Spectroscopic techniques, however, also face limitations that hamper their routine use in soil quality assessment. First, when applied to the soil surface in the field, information is gained only about the first millimeters of the soil. Second, sample characteristics such as moisture content, particle size distribution and roughness of the soil surface can influence the outcome of the analysis (Baveye and Laba, 2015; Stenberg et al., 2010). Third, a calibration step is used to relate the spectral information to soil characteristics (Gandariasbeitia et al., 2017) and the prediction is as good as the calibration data set. Several studies showed that calibration efficiency varies between studies and parameters considered (Islam et al., 2003); Kinoshita et al., 2012). Through their nature, spectroscopic estimates are always less precise than traditional analytical methods (Islam et al., 2003). Creation of freely-available databases that can be used for proper calibration and prediction of soil properties are essential for realizing the full potential of these techniques. These databases should involve both NIR spectra and results from wet chemistry and biological methods.
X-ray tomography is another non-destructive technique that can be used for soil structural analysis and can shed light on processes integrating soil physical and biological properties (Helliwell et al., 2013). It avoids some drawbacks of spectroscopic techniques, namely the fact that it scans a 3D image of the soil instead of only scanning its surface. Nevertheless, this technique is still a long way from routine application for soil quality assessment.
Such novel indicators potentially allow a more detailed assessment of soil processes. At the same time, some of the techniques may be developed into high-throughput soil analysis to shed light on the spatial and temporal variability of soil parameters and determine soil quality across different scales for application in precision agriculture, monitoring programs and life cycle assessments (Ge et al., 2011; Viscarra Rossel et al., 2017). The rapid evolution of these techniques and the decreasing costs associated with them will facilitate this development. However, the practical operability of these indicators by different stakeholders needs to be taken into account. The various limitations described above still seriously hamper application of such novel indicators in routine soil quality assessments. In addition, the absence of standard operating procedures (SOPs) and accepted threshold values, especially for molecular methods, make the comparison and the interpretation of the results challenging (Callahan et al., 2016). The final and most important limitation to the interpretation of these novel soil quality indicators is the lack of functional linkages with soil processes and management implications. Although use of novel indicators directly by farmers would be an advantage, most farmers are willing to send samples to the laboratory as long as the analysed indicators are meaningful and responsive to management (Bouchez et al., 2016). For policy makers operating or setting up soil quality monitoring schemes, the introduction of novel indicators would also be aided by relating them to existing ones that may be phased out when performance (or cost-efficiency) of novel indicators is superior.
At the moment, however, most novel soil quality indicators still belong to the research domain, and many technological, practical and interpretation related issues need to be overcome.
Note: For full references to papers quoted in this article see