Evaluating soil mesofauna as indicators of soil health across agricultural and forestry systems

Abstract

Introduction:

Assessing whether land-use practices promote soil sustainability is increasingly important as agricultural technologies aim to support long-term soil health. Soil mesofauna provide valuable indicators of these patterns, yet species-level assessment is labor-intensive and requires specialist expertise. Trait-based approaches such as QBS-ar offer a simplified alternative, but their performance across contrasting land-use systems remains insufficiently understood.

Methods:

We examined mesofaunal communities in two experiments applying conservation-oriented management: conventional versus regenerative agriculture, and diverse forestry treatments contrasted with continuous-cover forests. Taxonomic indicators (species richness, diversity, density) were assessed alongside trait- and ecomorphology-based measures (QBS-ar, life-form traits, and ecomorphological groups) to determine which metrics most effectively reflect management impacts.

Results:

Mesofaunal responses varied markedly between land-use types. In intensively managed agricultural systems, species richness, density, and functional responses derived from ecomorphological groups and life-form traits were sensitive to management intensity, whereas in forest environments, although initial disturbances were strong, responses were weaker and more variable because recovery processes were already underway. Seasonal sampling was essential, as indicator sensitivity varied across taxa.

Discussion:

Overall, traditional metrics such as density and species richness may fail to detect management effects in structurally complex or lightly disturbed systems. In contrast, QBS-ar provided a rapid and broadly applicable assessment of disturbance, while other trait-based measures captured subtle functional responses across seasons and land-use types, highlighting their value as complementary tools for soil biodiversity monitoring.

1 Introduction

In both agricultural and forestry systems, the widespread adoption of intensively managed monocultures has led to the degradation of the physical, chemical, and biological properties of soils (1). At present, these practices are no longer sustainable. An increasing number of land managers recognize that alternative approaches are required to improve soil health (2). Soil health emphasizes the functional condition of soil as a living system, describing how effectively it performs key functions such as supporting plant growth and regulating ecosystem processes (2, 3). Enhanced soil health ensures the long-term productivity of ecosystems and supports higher levels of biodiversity, which in turn contributes to ecosystem resilience under changing climatic conditions. In addition, it improves several ecosystem services (4).

Over recent decades, several sustainable management techniques have been developed in both agriculture (e.g., no-till and conservation tillage) and forestry (e.g., continuous cover forestry) to improve environmental conditions, restore ecosystem services, and maintain yields (5–7). In both systems, continuous plant cover, reduced soil disturbance, and diversification of biota are prioritized (8–10). In both forestry and agricultural systems, healthy soils have a greater capacity to store water and nutrients, enhance carbon sequestration, and exhibit increased resistance to erosion due to continuous ground cover (2). These characteristics also contribute to high ecological resilience, buffering soils against environmental extremes (11, 12).

Reliable assessment of soil health recovery can only be achieved through the use of appropriately selected soil indicators (13). Despite substantial efforts devoted to soil monitoring, a comprehensive, up-to-date framework for defining and identifying healthy soils remains lacking. Consequently, no universally accepted methodology or standardized protocol for soil health assessment has yet been established (14). Numerous physical, chemical, and biological indicators have been proposed for this purpose (14, 15). In parallel with the advancement of sustainable management practices, soil health indicators must also be refined to more effectively evaluate and compare current land-use strategies. In agro-ecosystems, frequent assessment of soil health is often required; however, the use of multiple indicators is time-consuming and financially demanding. Considerable efforts have been made to identify soil health indicators that are robust and broadly applicable across land-use types, including both agricultural and forest ecosystems, while also meeting key criteria such as sensitivity, ease of measurement, and cost-effectiveness (16–18). In addition to the automation of the current measurements (19–21), there is a growing need for simplified soil health indicators. Such indicators should be applicable across a wide range of soil types, sensitive to environmental and management-induced changes, and capable of capturing temporal soil dynamics, including degradation and recovery processes (22–24).

Soil biological indicators are often considered the most informative measures of soil health. These include microbial activity, microbial biomass, enzymatic activity, soil respiration, and soil fauna, all of which reflect the biological functioning of soils (24, 25). Among soil fauna, mesofauna are particularly abundant and diverse, with Collembola and Acari being the most numerous groups (26). Mesofauna play a critical role in decomposition, nutrient cycling, and other ecosystem services (27, 28). They are generally sedentary, highly fecund, and respond rapidly to environmental changes (29–32), often faster than chemical or physical soil parameters (33, 34). This sensitivity makes them excellent bioindicators of soil health (35), serving both as supporters of soil ecosystem functions and as indicators of soil condition (36).

Although soil mesofauna meet many of the criteria for effective bioindicators, they remain relatively underutilized in soil health assessments, largely due to the labor-intensive nature of their analysis. Conventional taxonomy-based methods require substantial taxonomic expertise and are time-consuming because of the need to identify and enumerate individuals, resulting in high labor costs and limited practical applicability (24, 37). To reduce the analytical effort of taxonomic methods, several soil mesofauna-based indices have been proposed, such as trait-based and composite metrics (38, 39).

High density and taxonomic diversity of soil mesofauna are widely recognized as indicators of soil fertility and overall ecosystem condition (40). As a result, abundance and diversity metrics are among the most commonly applied variables in soil monitoring studies. Both typically decline as land-use intensity increases (40). Mesofaunal groups such as Acari and Collembola are frequently targeted because of their high abundance and sensitivity to management practices, particularly under no-till systems (41–43). In well-functioning soils, these organisms generally occur at high densities, reflecting favorable soil properties including organic matter content, pore structure, and moisture availability (41, 44, 45).

In recent decades, trait-based approaches have become increasingly important for simplifying ecological patterns and improving the interpretation of biological data (46, 47). Traits are measurable morphological, physiological, phenological, or behavioral characteristics that reflect environmental conditions and influence organismal fitness (48, 49). Soil mesofauna are vertically structured within the soil profile according to their ecomorphological characteristics, which correspond to distinct ecological functions. This vertical stratification is often illustrated using Collembola, which can be readily classified into life forms reflecting broader mesofaunal patterns (50–52). Atmobiotic and epigeic species typically inhabit the soil surface and vegetation layers, where they primarily contribute to the fragmentation of plant residues and interact with surface-associated microbial communities, while hemiedaphic and euedaphic species occupy litter and deeper mineral soil layers, where they are more closely associated with microbial grazing, nutrient mineralization, and the regulation of soil organic matter turnover (53). These ecological roles are generally associated with characteristic morphological adaptations. For example, epigeic forms often exhibit traits linked to disturbance resistance and dispersal capacity, such as pigmentation, elongated legs, and longer antennae, whereas euedaphic species tend to possess soil-adapted traits, including reduced appendages, depigmentation, and streamlined body structures that facilitate movement through soil pores (54, 55). Different species may respond to environmental factors through different strategies, and distinct traits can contribute to the same ecological function. Therefore, analyzing sets of traits jointly, rather than individually can provide more integrative and ecologically meaningful results (32, 48, 52, 56). Such trait sets, often based on the combination of morphological traits, allow a more objective and fine-scale classification of organisms than traditional, experience-based categorizations (57). In many cases where land use was evaluated, a trait-based approach has proven to be more useful (32, 58).

From the various arthropod-based indices developed to assess soil quality and biological health, the QBS-ar (Soil Biological Quality–arthropods) index evaluates the presence or absence of soil arthropod groups and assumes that soils harboring a greater number of taxa well adapted to the soil environment indicate higher soil quality (25, 38). The index relies on simple morphological criteria reflecting the degree of soil adaptation, allowing a rapid and cost-effective evaluation of soil conditions. QBS-ar has proven to be a robust tool for detecting differences in soil quality under a range of land-use and management practices, including conservation, reduced-tillage, and forestry systems (59–63). Several modified versions have also been proposed, including a Collembola-specific variant (QBS-c) and an abundance-weighted form (QBS-ab) (61, 64, 65). However, the original QBS-ar has been reported in the literature as the most consistent and robust version for assessing soil biological quality compared to its modified variants.

Although alternative management systems are expected to influence soil biota, their effects on soil mesofauna remain poorly understood, particularly when comparing multiple biological indicator approaches across contrasting land-use systems (66). This gap is especially evident in forestry, where mesofaunal responses to conservation-oriented management remain insufficiently studied. To address this, we investigated soil mesofauna, with a focus on Collembola, in agricultural and forest ecosystems representing distinct disturbance gradients. We applied multiple indicator approaches, including taxonomic metrics (species richness, diversity, density), trait-based measures, and composite indices such as QBS variants, life-form trait classifications, and ecomorphological groups (Table 1). To support indicator evaluation, we developed a conceptual framework linking management-related drivers, such as soil disturbance, vegetation removal, habitat complexity, and organic matter retention, to expected mesofaunal responses (Table 2). Within this framework, soil health is interpreted as the ability of indicators to consistently reflect ecological degradation or recovery under differing management intensities.

The primary aim of this study was to determine which mesofaunal indicators most effectively capture management impacts across agricultural and forest systems, while identifying metrics that remain robust across ecosystems and seasonal variation. The novelty of this study lies in the explicit comparison of multiple soil mesofauna indicator approaches across agricultural and forest ecosystems, allowing evaluation of their sensitivity and transferability under contrasting disturbance regimes. We hypothesized that increasing disturbance would reduce mesofaunal density, species richness, and diversity in both ecosystems, and that these patterns would also be reflected in life-form traits and ecomorphological group composition. QBS-based indices were similarly expected to decline under higher management intensity. We further hypothesized that trait-based and composite indicators, which are less frequently applied in forestry, would provide more sensitive and transferable assessments of soil condition than traditional taxonomic metrics. By integrating multiple indicator frameworks across contrasting ecosystems, this study advances both the conceptual and applied understanding of soil biological assessment under alternative land management systems.

2 Materials and methods

2.1 Study sites and design

To evaluate soil mesofaunal responses under contrasting management regimes, we analyzed two case studies: one from a conservation agriculture experiment (CAGR) and one from a forestry treatment (FOR), representing alternative approaches to conventional land use.

2.1.1 Conservation agriculture experiment

The conservation agriculture experiment was conducted near the village of Dióskál (46°42′15″ N, 17°02′50″ E), western Hungary. A detailed description of the site and study design is presented in Juhos et al. (67). Briefly: the long-term experimental field established in 2003 to investigate the effects of conventional plow-based tillage (P) and conservation tillage (regenerative; R) systems. The site is located at an elevation of 176–206 m above sea level, within a hilly landscape representative of Central European arable regions. The climate of the area is warm-summer humid continental, with a mean annual temperature of approximately 11 °C and a mean annual precipitation of 600–700 mm. The parent material is loess, and soils are classified as haplic Luvisols according to the World Reference Base. In the 0–15 cm soil layer, the soil is slightly acidic, with an average pH of 5.46 (67).

The 32 ha experimental area was divided into eight plots (~4 ha each), arranged as four paired replicates of R and P treatments (Figure 1). All plots followed the same crop rotation, crop types, fertilization regime, and plant protection practices. The P system involved annual mouldboard plowing to a depth of 25–30 cm, followed by harrowing and seedbed preparation. The R system was based on non-inversion tillage using discs and cultivators, with a reduced number of operations and approximately 30% of the soil surface left covered by crop residues. Long-term application of conservation tillage at the site has been shown to reduce soil organic matter loss and enhance both soil biological activity and structure. During the study year, sunflower was grown during the main vegetation period, followed by autumn-sown winter wheat. A nearby 10-year-old meadow (M), established on former arable land, served as a reference system. Prior to conversion, the site was managed under conventional crop rotation (wheat, maize, spring barley, and sunflower). The area was subsequently restored using a grass seed mixture and has since been managed through grazing by cattle and periodic mowing. The soil type is consistent with the experimental fields (Luvisol). Previous investigations at the study site demonstrated clear differences in soil properties between R and P systems (67). Long-term R resulted in substantially higher topsoil organic carbon stocks compared with P, reflecting enhanced carbon sequestration and reduced organic matter mineralization under non-inversion management. These changes were accompanied by improved soil biological activity, including greater microbial biomass, higher earthworm abundance (all taxa combined), and enhanced soil aggregation. In contrast, P was characterized by lower organic carbon levels, attributed to intensified aeration and accelerated decomposition processes. Overall, the results of Juhos et al. (67) indicate that long-term conservation tillage at the site improved soil structural stability, biological functioning, and nutrient dynamics relative to conventional plowing.

Figure 1

2.1.2 Forestry treatments

The forestry treatment experiment was conducted at Hosszú-hegy in the Pilis Mountains, Hungary (47°40′ N, 18°54′ E), at elevations ranging from 370 to 470 m above sea level. A detailed description of the site and study design is presented in Kovács et al. (68). Briefly: the regional climate is characterized by a mean annual temperature of 9.0–9.5 °C and an average annual precipitation of approximately 600–650 mm. Soils are classified as Luvisols (lessivated brown forest soils), developed on limestone and red sandstone bedrock, with loess-derived material forming the upper soil layers.

The experiment was established in a managed, even-aged oak–hornbeam forest covering approximately 40 ha. At the onset of the study in 2014, the stand was approximately 80 years old. The canopy is dominated by sessile oak (Quercus petraea), with an average tree height of 21 m and a mean diameter at breast height of 28 cm (68). Additional canopy species include turkey oak (Quercus cerris), European beech (Fagus sylvatica), and wild cherry (Prunus avium). The sub-canopy layer is mainly composed of hornbeam (Carpinus betulus) and manna ash (Fraxinus ornus).

Forestry treatments were applied between December 2014 and January 2015 using a randomized complete block design with six replicate blocks. In this study from five experimental treatments (Figure 1), we investigated 3: (i) control (C), representing closed-canopy forest without intervention; (ii) clear-cutting (CC), involving complete tree removal within a circular area of 0.5 ha (80 m diameter) to simulate high-intensity disturbance; (iii) gap-cutting (G), where small circular gaps (20 m diameter) were created within the closed-canopy stand, reflecting selective harvesting practices typical of continuous cover forestry. The other two treatments (preparation cutting and retention tree group) showed negligible changes in the case of soil mesofauna (69–71), so we did not include them in the analyses.

Previous studies conducted at the experimental site documented clear differences in environmental conditions among the C, CC, and G treatments based on measurements from 2020 (69, 70). Vegetation structure differed markedly among treatments, with herbaceous cover and sapling density being lowest in the C plots and higher in the G- and CC treatments. Plant species richness was reduced in control relative to managed plots, while litter mass showed the opposite pattern, being greatest in the C and lowest under CC, with G exhibiting intermediate conditions. Despite these pronounced differences in surface organic inputs and vegetation structure, soil chemical properties, including pH and soil carbon and nitrogen contents, did not differ significantly among treatments. In microclimatic conditions, CC produced the most distinct conditions, particularly in terms of relative air humidity and soil temperature. G plots were generally associated with higher soil moisture and sapling density compared with the C, while CC showed intermediate or more variable responses. Overall, the 2020 data indicate that forest management treatments primarily altered vegetation structure, litter dynamics, and microclimatic conditions, while short-term changes in soil chemical properties remained limited (69–71).

2.3 Fauna sampling

At both sites, soil samples were collected using the same method with a cylindrical soil corer (400 cm³; 8 cm diameter × 8 cm depth, sampling area (0.00503 m²)) during the pre-vegetation (spring) and post-vegetation (autumn) periods. At the conservation agriculture experiment, three soil cores were collected from each subplot in 2023 (25 May and 13 November), resulting in 12 subplots per treatment and three meadow subplots as controls, for a total of 27 subplots. This yielded 81 soil cores per season, which were averaged at the subplot level prior to statistical analysis, resulting in 27 independent subplot-level observations per season and thereby avoiding pseudoreplication. In the forestry treatment, one soil core was collected per plot in 2020 (2 April and 10 October) across six blocks and three treatments, resulting in 18 independent plot-level observations per season. Accordingly, the experimental unit used in the statistical analyses was the subplot in the conservation agriculture experiment and the plot in the forestry experiment. All samples were transported immediately to the laboratory and extracted for one week in 70% ethanol using Berlese extractors. Animals were sorted and identified into the main taxonomic groups under a Nikon SMZ25 digital microscope. For detailed identification of Collembola, specimens were examined under a Nikon Eclipse Ts2R microscope and identified to species level using identification keys (72–78).

2.4 Indices, indicators

We chose a total of 14 indicators. Most of them were mainly based on Collembola, like species richness, Shannon–Wiener diversity, total Collembolan density, or density of different Collembola ecomorphological groups (epigeic, hemiadaphic, and euedaphic) and life-form (LF) traits (see later), and specific, modified QBS-ar indices modified for Collembola (QBS-cab, QBS-cpres; see description later). Besides these, the density of Acari was also shown, and a mesofauna-based index, the QBS-ar and its only presence-based variant, the QBS-pres.

Collembola species richness and the Shannon–Wiener diversity index were calculated using the PAST program (79). Ecomorphological group characterization was assigned to each Collembola species (Supplementary Table 1), based on literature (53, 80–82).

Trait-based indices were derived using species-level characteristics for each Collembola taxon, rather than traits measured at the individual level (57). In total, eight morphological traits were considered: presence of a furca, number of ocelli, presence of scales or dense macrosetae (e.g., clavate or ciliate setae), presence of a post-antennal organ (PAO), and body, antenna, and leg length (see Supplementary Table 1). Trait information was compiled from the above-mentioned indication keys. When species-specific data were unavailable, trait values were inferred from the closest related taxa or verified directly from collected specimens (e.g., presence of dense setae, scales, pigmentation, or body size), as indicated in the Supplementary Table 1. Based on these morphological traits, three life-form (LF) traits were derived following Vandewalle et al. (56). In accordance with their approach, these LF traits represent composite variables calculated from combinations of individual morphological traits and describe adaptation to the soil environment, dispersal ability, and resistance to disturbance. Each morphological trait was assigned a numerical score ranging from 0 (weak association) to 4 (strong association) with the corresponding LF. Different combinations of functional traits and scores were used to compute each LF trait (Supplementary Table 2). Trait scoring followed established approaches in the literature to ensure consistency and comparability among studies (32, 52, 56, 83, 84). LF indices were calculated for each species by summing the relevant functional trait scores [see in Supplementary Table 1 (57)]. Higher LF values indicate stronger expression of the corresponding ecological strategy. For each sampling plot, community-weighted mean (CWM) values were calculated for all LF traits following Ricotta and Moretti (85). Species relative abundances were obtained by dividing the abundance of each species by the total abundance recorded within the plot. These relative abundances were then multiplied by the corresponding species-level LF scores, and the resulting values were summed across species. The resulting CWM values were used as response variables in subsequent statistical analyses.

We calculated several variants of the QBS-ar index. In addition to the standard QBS-ar score for the entire mesofauna (38), we recorded the number of QBS-ar groups present (QBS-pres), i.e., the total number of distinct biological trait groups detected in the sample, regardless of their individual scores. For Collembola, we also assessed the presence of the seven QBS ecological categories (QBS-cpres) (59). Following the QBS-ab approach of Mantoni et al. (61), we calculated an abundance-weighted Collembola QBS index (QBS-cab), in which the score assigned to each QBS category was multiplied by the density of specimens belonging to that category and then summed for each sample.

2.5 Statistical analyses

The two case studies were analyzed separately, but the same variables were studied in both cases, considering the whole study year and separately for the pre- and post-vegetation periods. Densities were standardized to individuals per square meter using the sampling area. In case of right-skewed distribution, data were log-transformed using log (x + 1) before analysis.

All statistical analyses were conducted in R version 4.5.1 (86). Treatment effects on soil mesofaunal variables were analyzed using linear mixed-effects models (LMM) fitted with the lmer function of the ‘lme4’package (87). For count-based response variables (species richness), generalized linear mixed-effects models (GLMM) were fitted using the glmer function of the ‘lme4’ package, assuming a Poisson distribution. In all models, the response variable was the measured variable, treatment was included as a fixed effect, and block was included as a random intercept to account for the hierarchical experimental design. This approach ensured independence at the level of experimental units and minimized the risk of pseudoreplication. Additionally, sampling units within blocks were spatially separated (minimum 100 m), reducing the likelihood of spatial autocorrelation.

Factor levels of treatment were M/R/P for the conservation agriculture experiment and C/G/CC for the forestry experiment. For models covering both sampling periods, season and its interaction with treatment were included as fixed effects. Separate seasonal models (pre- and post-vegetation) were fitted without the season term. Model assumptions were assessed using residual diagnostics and checks for overdispersion where appropriate. Pairwise differences among treatments were evaluated using estimated marginal means with Tukey-adjusted multiple comparisons, implemented with the emmeans function of the ‘emmeans’ package (88).

Effect sizes were calculated to quantify the magnitude of management effects across indicators. For variables analyzed using linear mixed-effects models (LMMs) with a Gaussian distribution, standardized mean differences (Cohen’s d) were computed from estimated marginal means using the eff_size function of the ‘emmeans’ package. For count-based variables (species richness) analyzed using generalized linear mixed-effects models (GLMMs) with a Poisson distribution, effect sizes were expressed as log response ratios (LRR). These were derived from back-transformed estimated marginal means and represent the natural logarithm of the ratio between treatment means. Confidence intervals were calculated on the log scale, and effect sizes whose intervals did not overlap zero were considered statistically meaningful. To facilitate interpretation and comparison across indicators, seasons, and study sites, effect sizes were visualized using heatmaps showing only contrasts with large and statistically supported responses (|effect size| > 0.8; 95% confidence intervals not overlapping zero). Effect sizes were interpreted following Cohen (89), where values of |d| ≥ 0.8 are considered large, while acknowledging that these thresholds represent general guidelines and should be interpreted in the ecological context of the study.

Indicator sensitivity to different levels of disturbance was quantified using standardized effect sizes. Effect size estimates from all management contrasts were compiled across study systems, seasons, and response variables. For each indicator, sensitivity was calculated as the mean absolute effect size across all contrasts, using absolute values to quantify response magnitude irrespective of direction and to avoid cancellation of opposing effects. Indicators were ranked according to mean absolute effect size to provide a quantitative measure of relative sensitivity. Sensitivity was also examined separately for each study system to assess consistency of responses across systems.

To evaluate redundancy among QBS-derived indices (QBS-ar, QBS-cab, QBS-pres, and QBS-cpres), correlation analyses were conducted separately for the two study sites (FOR and CAGR). Spearman’s rank correlation coefficient (ρ) was used to assess pairwise relationships among indices. Correlation matrices were calculated using the cor function with the method set to “spearman”, and statistical significance of correlations was assessed using the rcorr function from the ‘Hmisc’ package (90).

3 Results

3.1 Densities

Total Acari density differed among management regimes in the conservation agriculture experiment (CAGR), with the highest values observed in meadow (M), intermediate values under conservation tillage (R), and the lowest values under conventional plow-based tillage (P) (Figure 2; Supplementary Figure 1A). Seasonal analyses revealed that Acari density was highest in the R treatment during the pre-vegetation period, whereas post-vegetation patterns followed the hypothesized gradient, with decreasing densities from M to P. In the forestry experiment (FOR), both disturbed treatments (gap-cutting, G, and clear-cutting, CC) exhibited reduced Acari densities compared to the control (C). These differences were not significant during the pre-vegetation period (Supplementary Table 3), while in the post-vegetation season, CC plots showed the lowest densities (Figure 2A; Supplementary Figure 2A).

Figure 2

Patterns in total Collembola density largely mirrored those observed for Acari in the agricultural system, with higher values under R management relative to P at the annual scale (Figure 2B; Supplementary Figure 1B). Seasonal analyses revealed a shift in treatment differences between sampling periods. During the pre-vegetation period, Collembola density was significantly higher in the meadow than in both agricultural treatments, whereas no significant difference was detected between P and R. In contrast, during the post-vegetation period, Collembola density was significantly higher under R than P, while densities in M were intermediate and did not differ significantly from either treatment (Figure 2B; Supplementary Figure 1B). In the FOR, treatment-related differences were detected only during the pre-vegetation period, when G plots exhibited lower values than C and CC treatments (Figure 2B; Supplementary Figure 2B).

3.2 Ecomorphological groups

Collembola ecomorphological groups revealed contrasting responses. In the CAGR, epigeic Collembola showed significant differences primarily in seasonal analyses and were most sensitive to management intensity, with the highest densities recorded in M sites (Figure 2C; Supplementary Figure 1C). Hemiedaphic taxa generally followed the expected disturbance gradient. At the annual scale, densities were significantly higher in M and R treatment than under P. A similar, although non-significant, trend (M > R > P) was observed during the pre-vegetation period, whereas in the post-vegetation period meadow and regenerative treatments again exhibited significantly higher densities than conventional plowing (Figure 2D; Supplementary Figure 1D; Supplementary Table 3). In both the epi- and hemiedaphic groups, the trends were the same throughout the seasons. Euedaphic Collembola reached their highest densities under R management, only in the post-vegetation period (Figure 2E; Supplementary Figure 1E). In the FOR, differences among ecomorphological groups were detected only during the pre-vegetation period. Epigeic and euedaphic Collembola showed their highest densities in CC plots, whereas hemiedaphic taxa were most abundant in C plots. G treatments consistently exhibited the lowest densities across groups (Figures 2C–E; Supplementary Figures 2C–E).

3.3 Trait-based approach

Trait-based analyses revealed clear differences among agricultural management types. Traits associated with dispersal ability and disturbance resistance were most pronounced in M sites, whereas traits related to adaptation to the soil environment were more strongly expressed in cultivated fields (R, P) (Figure 2F–H; Supplementary Figures 3A–C). These patterns were consistent across annual and seasonal analyses. In contrast, trait-based indices did not reveal clear or consistent differences among forestry treatments, regardless of temporal scale (Figures 2F–H; Supplementary Figures 4A–C).

3.4 Diversity and richness

Species richness in the CAGR followed the hypothesized disturbance gradient, with the highest values in M sites and the lowest under P (Figure 3A; Supplementary Figure 3D). However, no significant difference was detected between M and R treatments (Supplementary Table 3). In the FOR, species richness was the lowest in G and the highest in the C sites. CC sites maintained intermediate values (Figure 3A; Supplementary Figure 4D). In both studies, the effects were consistent across seasons. In contrast, Shannon–Wiener diversity did not differ significantly among treatments in either experimental system (Figure 3B; Supplementary Figures 3D, 4D).

Figure 3

3.4 QBS approach

In the CAGR, QBS-ar values followed patterns similar to those observed for total Acari and Collembola densities, although differences in statistical significance were noted among temporal scales (Figure 3C; Supplementary Figure 5A; Supplementary Table 3). In the FOR, QBS-ar values showed a pattern similar to that observed for Collembola density. Although seasonal differences were not statistically significant, the lowest values were recorded in G and the highest in C plots (Figure 3C; Supplementary Figure 6A).

Collembola density-weighted QBS index (QBS-cab) differentiated management types only in the CAGR. No significant differences were observed during the pre-vegetation period, whereas at the annual scale and during the post-vegetation period, R treatments exhibited the highest values and P the lowest (Figure 3D; Supplementary Figure 5B). In the FOR, no significant treatment effect was found (Supplementary Table 3). Presence-based QBS index (QBS-pres) consistently followed the expected disturbance gradient in the agricultural system. Significant differences among management types were observed in both the annual and pre-vegetation analyses, whereas during the post-vegetation period only P showed significantly lower values compared to the other treatments, although the same trend was maintained (Figure 3E; Supplementary Figure 5C; Supplementary Table 3). In contrast, the response of the Collembola presence-based QBS index (QBS-cpres) was weaker. At the overall level, effects were limited; however, significant differences between R and P were detected in the annual and post-vegetation analyses, while other comparisons were not statistically significant (Figure 3F; Supplementary Figure 5D; Supplementary Table 3). In the forest system, both QBS-pres and QBS-cpres showed patterns similar to QBS-ar, with G plots generally exhibiting the lowest values (Figures 3E, F; Supplementary Figures 6C, D).

Spearman rank correlation analysis revealed differing patterns of association among QBS-derived indices between the two study sites (FOR and CAGR). In the FOR dataset, correlations among indices were generally weak to moderate, with only a few strong and statistically significant relationships. The strongest correlation was observed between QBS-ar and QBS-pres (Table 3), indicating a high degree of redundancy between these two indices. A moderate positive correlation was found between QBS-cab and QBS-cpres. Other relationships were weak and not statistically significant. In contrast, the CAGR dataset showed consistently stronger and more widespread correlations among indices (Table 3). The relationship between QBS-ar and QBS-pres remained strong, similar to FOR. However, additional moderate to strong correlations were observed, including QBS-cab and QBS-pres, and QBS-ar with QBS-cab (Table 3). Moreover, all pairwise correlations were statistically significant, indicating a high level of coherence among indices in this dataset.

We aimed to identify indicators showing consistently large, precise, and directionally coherent responses across management gradients, seasons, and land-use types. For visualization in the heatmaps, contrasts with low effect sizes or confidence intervals overlapping zero were excluded; however, all indicators and contrasts were retained in sensitivity and consistency analyses (Supplementary Table 4). Overall, the strongest and most frequent treatment effects were observed in agricultural systems, whereas fewer strong responses were detected among forestry treatments (Figures 4, 5). In CAGR, effect sizes were generally larger during post-vegetation periods, whereas in FOR, stronger effects tended to occur in pre-vegetation periods (Figure 4).

Figure 4

Figure 5

Across both studies, Acari and Collembola density showed the highest sensitivity, followed by QBS-pres and QBS-ar, while Shannon-Wiener diversity exhibited weaker responses. Acari density responded strongly and consistently across management gradients, with large effect sizes in nearly all contrasts (Figure 4; Supplementary Table 4). In FOR, it was the only consistently highly sensitive variable. Collembola density showed particularly strong responses in agricultural comparisons. Among ecomorphological groups, hemiedaphic Collembola exhibited the most pronounced responses, especially in post-vegetation periods across both agricultural and forestry systems. In contrast, epigeic and euedaphic Collembola showed more variable and generally weaker responses. Epigeic groups occasionally displayed increased sensitivity under R and G treatments, while euedaphic groups were characterized by smaller and less consistent effect sizes across contrasts and sites. Among life-form trait indices, disturbance and adaptation to the soil-environment showed strong but opposing responses along the agricultural gradient, effectively distinguishing meadow from managed agricultural treatments. In contrast, QBS-ar was particularly effective in separating conventional tillage from other management types (Figures 4, 5). QBS-pres emerged as the most sensitive indicator, but only under agricultural management. Analyses based on log response ratios (LRR) indicated that species richness exhibited large proportional changes (mean absolute LRR ≈ 35%) and a proportion of strong responses comparable to several other indicators (Supplementary Table 4).

Indicator consistency, expressed as the proportion of contrasts showing strong effects, varied markedly among variables (Table 4). Acari density and QBS-pres were the most consistent indicators, each exhibiting strong effects in 75% of contrasts, followed by Collembola density and several trait-based indices (QBS-ar, QBS-cpres, hemiedaphic Collembola) showing intermediate consistency. Shannon–Wiener diversity displayed the lowest consistency. Species richness, assessed using log response ratios, showed a moderate proportion of strong responses (0.42), comparable to several Gaussian-based indicators but lower than the most responsive ones.

4 Discussion

4.1 Treatment effects on microarthropods

4.1.1 Conservation agricultural comparisons

As with previously reported environmental parameters (67), soil microarthropod communities responded consistently to the management-related disturbance gradient, supporting the hypothesis that reduced disturbance and increased structural complexity enhance belowground biological functioning (91). Although concurrent physicochemical measurements were not directly included in the present study, available site-specific environmental data from the same location and timeframe consistently indicated improved soil conditions under regenerative management compared to conventional systems. Across all indicators, meadow and conservation treatments supported higher densities, Collembola species richness, and functional diversity than conventional plow-based tillage, in line with previous studies (41, 42, 45, 61). Nevertheless, the absence of direct physicochemical measurements within the meadow site limits mechanistic interpretation and should be considered when linking biological responses to specific soil properties.

Meadow sites, characterized by minimal physical disturbance and continuous plant cover, exhibited the highest or near-highest mesofauna densities and species richness, confirming their role as reference systems for soil biological functioning. Conservation management frequently approached these values, particularly for Acari and Collembola densities and species richness, suggesting that reduced tillage combined with organic inputs can promote the recovery of soil biota under agricultural use (43, 45, 92). The high density of euedaphic Collembola under conservation management, is consistent with improved soil structure and pore availability compared with plowed fields (42, 93). Interestingly, the soil-adaptation LF trait did not follow the same pattern as euedaphic Collembola density and reached its lowest values in the meadow treatment. This suggests that the index reflects not only soil disturbance but also vegetation structure and seasonal habitat conditions. Continuous vegetation cover in meadows likely favors a greater proportion of epigeic species, whereas in agricultural fields, in non-vegetated seasons, the relative contribution of soil-dwelling forms increases (57). Consequently, the soil-adaptation LF trait should be interpreted in the context of both disturbance and vegetation dynamics rather than as a direct indicator of soil quality alone. In contrast, conventional plow-based tillage showed consistently lower microarthropod densities, reduced Collembola species richness, and the lowest QBS values, indicating that repeated mechanical disturbance constrains both population size and functional composition, even where some soil-adapted life-form traits persist. Although contrasting patterns have occasionally been reported (94, 95), such responses are widely documented under intensive agriculture and are generally attributed to soil disruption, reduced organic matter inputs, and unfavorable microclimatic conditions (91, 96).

Overall, the convergence between meadow and conservation treatments, particularly in Collembola species richness, suggests that regenerative practice in this study can recover key components of soil biological diversity, even if community composition or functional structure does not fully match that of semi-natural systems. These findings highlight the potential of conservation management to mitigate the negative impacts of intensive tillage on soil mesofauna.

4.1.2 Forestry management comparisons

The strong and coherent responses observed in agricultural treatments contrast with the weaker and more variable effects previously reported under forestry management (69–71). This difference likely reflects fundamental contrasts in disturbance regimes and environmental filtering between systems. In the forestry experiment, soil chemical properties, including soil C, N, and pH, did not differ significantly among treatments (57), suggesting that management effects were driven less by direct edaphic changes and more by aboveground structural alterations. In contrast to relatively stable soil properties, vegetation structure and litter characteristics responded more strongly to forestry treatments. Herbaceous cover and sapling density increased substantially in managed stands, and litter mass remained greatest in control stands while declining markedly in clear-cutting areas (69, 70). Such indirect and structurally mediated responses may explain the lower consistency and weaker magnitude of biological indicator responses in forestry relative to the more direct soil disturbance gradients observed in agricultural systems.

Acari (mainly Oribatida) densities generally followed the expected disturbance gradient, whereas Collembola showed no consistent response. As discussed by Flórián et al. (71), this contrast likely reflects fundamental differences in life-history strategies between the two groups (i.e., r- versus K-strategists), as well as the long-term nature of the experiment, in which vegetation regeneration and canopy closure are already underway. Additional previously reported differences among Collembola ecomorphological groups, detectable only during the pre-vegetation period, further indicate that forestry-related responses are temporally restricted and context dependent rather than consistently driven by management intensity (71). Against this background, we focused on assessing how additional variables, some of which were previously never applied in forestry studies, respond to the management-induced changes.

Control forest stands maintained high QBS-ar values, indicating relatively undisturbed soil environments, which derive from stable microclimatic conditions and continuous litter inputs (68). This confirms the role of closed-canopy forests as reservoirs of soil functional integrity. Among disturbed treatments, gap-cutting produced the strongest negative responses, particularly in spring Acari density, Collembola species richness, and QBS indices. This suggests that even moderate canopy opening can disrupt litter continuity and microclimatic stability, with negative consequences for arthropod communities. In contrast, clear-cutting sites showed intermediate responses, with QBS values often exceeding those observed in gap-cutting plots, which reflects more homogeneous post-disturbance conditions that facilitate partial recolonization observed in earlier studies (57, 71). Overall, the weaker and less consistent responses observed in forest systems compared to agriculture may partly reflect the inherently greater structural complexity of forest habitats. Even under intensive forestry treatments, residual vegetation cover, litter presence, and moderated microclimatic conditions likely remain more favorable than those in heavily disturbed agricultural systems, buffering mesofaunal communities against more severe ecological disruption. The absence of strong trait-based differentiation further suggests that functional responses in forest soils may be buffered, indirect, or delayed relative to the more immediate impacts observed under agricultural disturbance gradients.

4.2 Sensitivity and consistency of indicators

A central aim of this study was to identify indicators that show strong, precise, and directionally consistent responses across management gradients, seasons, and land-use types. Across both case studies, indicators varied markedly in their sensitivity, consistency, and temporal stability. In agroecosystems, the pool of suitable faunal indicators is limited, as most species occur at low abundance and diversity (41). Nevertheless, when detectable responses occur, they are often stronger than those observed in forestry systems, where faunal groups are generally represented by a higher number of species. In contrast, potential indicator groups from forest ecosystems, such as Protura and Formicidae (57, 97), are typically too rare in agricultural fields to allow robust statistical comparisons.

Acari and Collembola have long been proposed as effective bioindicators of soil quality change, partly because of their high abundance and diversity in agroecosystems (41, 42, 98). Our results support this view, as the densities of Acari and Collembola were among the most sensitive and consistent indicators across both study systems. Acari density showed strong and directionally coherent responses to management intensity in both agricultural and forestry contexts, making it the most robust single indicator identified in this study. It exhibited the largest effect sizes and the highest consistency across contrasts, remaining responsive even in forestry systems where most other indicators weakened. These findings are in line with previous studies identifying Acari as sensitive integrators of soil disturbance, microhabitat alteration, and resource availability (99–101). Collembola density ranked second in overall performance, with particularly strong responses in agricultural systems, where it effectively distinguished conventional tillage from less disturbed management. In forest ecosystems, responses were more variable, separating gap-cutting from other interventions and indicating system-dependent sensitivity related to the rapid response dynamics of Collembola and the long-term nature of the study (71).

From a practical perspective, both Acari and Collembola also represent relatively feasible indicators in terms of analytical effort. Their separation from other soil invertebrates at the class or subclass level does not require advanced taxonomic expertise, making their identification accessible for routine monitoring. The primary workload associated with these indicators lies in the counting of individuals, which remains time-consuming. However, ongoing methodological developments and emerging automated image-based identification approaches (e.g., 102–104) offer promising prospects for substantially reducing processing time and increasing the applicability of these indicators in large-scale monitoring programs.

Among Collembola ecomorphological groups, hemiedaphic taxa showed the strongest and most consistent responses to management across agricultural and forestry systems, particularly during post-vegetation periods when effect sizes were highest. Their intermediate position in the soil profile likely increases sensitivity to changes in soil structure and organic matter distribution, enabling reliable tracking of disturbance gradients and distinguishing highly from moderately disturbed systems. This integrative position supports their value as robust disturbance indicators.

In contrast, epigeic and euedaphic taxa exhibited weaker and context-dependent responses, reflecting opposing effects of surface disturbance and soil buffering. In conservation agriculture, epigeic Collembola declined consistently along the disturbance gradient, with the highest densities in meadow sites and strong seasonal differentiation, likely due to their dependence on surface litter, vegetation cover, and microclimatic stability. However, this pattern did not generalize to forestry systems, where responses reflected post-disturbance regeneration, including increased densities in clear-cuts five years after intervention (71). A similar regeneration-related pattern was observed for euedaphic taxa. In agriculture, euedaphic Collembola distinguished between sustainable and conventional systems but showed unexpectedly low densities in meadow sites, possibly due to soil compaction. These contrasting patterns confirm that the indicator value of epigeic and euedaphic taxa varies across ecosystems and along disturbance–recovery gradients.

Beyond their value as disturbance indicators, epigeic forms are more strongly associated with surface litter decomposition and organic matter incorporation, while hemiedaphic and euedaphic taxa contribute more directly to microbial regulation, nutrient turnover, and deeper soil food-web functioning (53). However, the ecological interpretation of Collembola ecomorphological groups has long been debated (53, 105, 106). Although classical concepts associate higher proportions of euedaphic forms with improved soil quality (38) and link epigeic traits to disturbed habitats (55), our results and recent studies (57, 92) indicate that, when analyzed separately, epigeic dominance is more closely associated with favorable habitat conditions. In our systems, epigeic densities declined with increasing disturbance, contrary to Ponge et al. (95). Their persistence was consistently linked to maintained surface cover (crops, residues, litter, understory vegetation), greater microclimatic stability, moisture retention, and resource availability, key components of soil health (2). Accordingly, epigeic taxa appear to be more informative indicators of favorable soil conditions than euedaphic forms alone, whereas euedaphic traits tend to dominate under conditions unfavorable for surface-dwelling species (57, 83). Trait analyses further supported this pattern, showing greater representation of disturbance-resistant traits in systems with continuous vegetation cover.

Overall, all three ecomorphological groups provided complementary information. Hemiedaphic taxa consistently responded to cumulative disturbance, particularly by distinguishing conventionally plowed fields from less disturbed management systems, while epigeic taxa were particularly informative for distinguishing meadow from cultivated habitats. Overall, our results indicate that Ecomorphological classification provides a useful framework for evaluating disturbance responses across ecosystems. Nevertheless, its application remains subject to limitations related to subjective group assignment (if it is not literature-based) and the labor-intensive nature of density-based, species-level identification.

In contrast to literature- or species-based ecomorphological classifications, we applied life-form (LF) trait indices, which provide a finer-resolution, measurement-based representation of soil adaptation traits (44). Many LF traits correspond to traditional ecomorphological groups, but this approach does not require species-level identification, as traits are assigned using observable morphological characteristics. Although LF trait analysis remains density-based and requires counting individuals, it is faster and demands less taxonomic expertise than species-level methods, and offers strong potential for automation through AI-based recognition of morphological features (37). Previous studies have shown that LF traits can provide clearer mechanistic insight in long-term agricultural systems (57). In the present study, LF trait indices distinguished managed agricultural fields from natural reference sites but were less effective in resolving differences among sustainability levels. Trait-based signals were consistent across seasons, suggesting they reflect underlying ecological processes rather than short-term population dynamics. In forestry systems, LF indices performed more weakly, possibly due to functional redundancies or slower trait turnover. Notably, this contrasts with the clear management responses observed for the original ecomorphological groups, indicating that aggregating groups into composite indices may obscure treatment-specific signals, particularly in forests. Although traits were derived from species-level identifications here, assigning traits directly at the individual level, while more time-consuming, may yield ecologically more robust results (92).

Collembola species richness performed well as an indicator, effectively distinguishing among agricultural treatments and between forestry treatments, particularly control and managed stands. However, it was generally less responsive than density-based metrics. Similar to LF traits, seasonal stability in species richness indicates a dominant management signal rather than short-term fluctuations, despite its lower sensitivity to impact magnitude. From a practical perspective, species richness does not require individual counts, but reliable identification remains taxonomically demanding. Although morphospecies assignment can reduce effort, the approach remains knowledge-intensive.

In contrast, Shannon–Wiener diversity was among the least sensitive indicators, showing weak and inconsistent responses across treatments and land-use systems. This aligns with previous findings that diversity indices are highly variable in cropping systems (92), and that Shannon diversity often correlates weakly with abundance or richness (61). Moreover, because it requires both counting and taxonomic identification, it is among the most labor-intensive metrics while providing limited additional explanatory value in this context.

QBS-ar is not density-based and therefore requires substantially less analytical effort than abundance-based indices. Although the scoring procedure involves some subjectivity, it can be applied consistently with limited training, making it a rapid assessment tool that demands less taxonomic expertise than species-level approaches. Its effectiveness in detecting management effects has been widely demonstrated (61), and our results similarly identified QBS-ar as one of the most responsive indicators. It consistently outperformed Shannon–Wiener diversity and showed comparable performance to Collembola species richness in detecting management-related differences (61, 62). While several refinements of the QBS method have been proposed, the original QBS-ar formulation remains the most practical and widely applied (61). Correlation analyses revealed strong redundancy between QBS-ar and QBS-pres across both agricultural and forestry systems, indicating that these indices capture highly similar ecological patterns despite the simpler calculation of QBS-pres. Because QBS-ar requires only minimal additional scoring effort beyond presence-based group identification, its broader standardization and established application may justify its continued preferential use. In forestry systems, where management responses were less distinct, QBS-ar and QBS-pres provided largely overlapping information, supporting QBS-ar as a practical primary indicator. In agricultural systems, QBS-pres showed slightly greater sensitivity, while stronger correlations among all QBS-derived indices reflected clearer disturbance gradients overall. Although density-weighted indices (QBS-cab and QBS-cpres) offered some additional sensitivity, these benefits were generally modest relative to their substantially greater analytical demands. Therefore, QBS-ar remains the most balanced and broadly applicable index for routine soil disturbance assessment, while QBS-pres may be advantageous where maximal simplicity or slightly improved sensitivity is prioritized, and abundance-based variants are best reserved for specialized research contexts.

Beyond the choice of indicators, sampling timing critically influences both the detectability and interpretation of biological responses (62), particularly in tillage systems where soil disturbance interacts with seasonal dynamics (45). Across both study systems, indicator sensitivity exhibited clear seasonal structuring: in agriculture, management effects were strongest during the post-vegetation period, whereas in forestry, they were more pronounced in the pre-vegetation season. Individual indicators also displayed distinct seasonal responses. Collembola reacted most strongly in the pre-vegetation season, while Acari density showed hypothesis-consistent responses mainly during post-vegetation periods. These differences likely reflect contrasting life-history strategies and resource use. Collembola, with short generation times, rapidly rebuild populations in spring and exploit decomposed organic matter accumulated in autumn (107, 108). In contrast, Acari represent a highly heterogeneous group with diverse feeding habits and life-history strategies (109), increasing the likelihood that disturbance-sensitive taxa are present in any given season. Although some indicators (epigeic and hemiedaphic Collembola densities, life-form trait indices, QBS-pres) showed limited seasonal variation, our results suggest that, within the studied agricultural and forestry systems, sampling across at least two seasons is necessary to capture management-related ecological processes adequately (62).

One limitation of this study is that forestry management represents a long-term disturbance regime, and our assessments were conducted several years after the treatments, once vegetation had already undergone substantial regeneration (71). As a result, some indicators suggested that clear-cutting stands currently provide more favorable conditions for soil microarthropods than gap-cutting stands, although this pattern was not consistent across all measures. Assessing these systems immediately after management, and over a broader temporal scale, would likely reveal different responses and improve identification of indicators most sensitive to early disturbance effects.

A further limitation is the lack of site-specific baseline measurements for the meadow reference sites. While meadows generally represent more natural and less intensively disturbed soil environments compared to conventionally managed fields (44, 57, 110), the absence of prior background data prevents full verification of the meadow’s initial ecological condition relative to the agricultural systems. Consequently, although the meadow served as a reasonable low-disturbance reference, this uncertainty limits the strength of causal inference. Future studies would benefit from baseline surveys and long-term monitoring to better distinguish pre-existing differences from management-driven effects.

In current soil monitoring, the key question is no longer whether soil biology should be measured, but how and at what level of detail. Although our findings are based on two case studies within a single region, collected in different years and with different sample sizes, they are therefore context-dependent and should be interpreted in the context of these differences when making direct cross-system comparisons. In particular, differences in sampling intensity may influence statistical power and should be considered when comparing indicator sensitivity across systems. In forestry systems, indicator responses should also be interpreted in relation to the timing of sampling, as data were collected several years after management interventions, when ecological recovery processes may already have attenuated disturbance signals. Future work should therefore incorporate balanced sampling designs and short-, medium-, and long-term dynamic monitoring following disturbance to strengthen temporal datasets and improve comparability in soil biological responses. Nevertheless, the results provide a valuable starting point for future multi-ecosystem investigations. Expanding this approach across diverse environmental settings will help evaluate the broader applicability and robustness of existing biological parameters and indices.

5 Conclusion

Soil microarthropods are sensitive and informative indicators of management-related disturbance across contrasting land-use systems, although their performance is strongly context-dependent. Density-based metrics, particularly Acari density, proved the most robust across systems and seasons, while Collembola, hemiedaphic groups, and QBS indices provided complementary information. These indicators offer a relatively rapid method requiring limited specialist expertise for evaluating the quality of different management practices. Trait-based approaches added mechanistic insight but were less discriminating in forestry systems. Seasonal dynamics emphasize the need for multi-season sampling. Overall, in our study, no single indicator captures all dimensions of management impact, highlighting the value of multi-indicator frameworks tailored to ecosystem context and monitoring objectives. These findings underline the importance of prioritizing indicators that balance sensitivity, consistency, and seasonal robustness when assessing soil biodiversity responses to land-use change and management intensity.

References

• 1

  SmithPHouseJIBustamanteMSobockáJHarperRPanGet al. Global change pressures on soils from land use and management. Global Change Biol. (2016) 22:1008–28. doi: 10.1111/gcb.13068

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 2

  LehmannJBossioDAKögel-KnabnerIRilligMC. The concept and future prospects of soil health. Nat Rev Earth Environ. (2020) 1(10):544–53. doi: 10.1038/s43017-020-0080-8

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 3

  KarlenDLVeumKSSudduthKAObryckiJFNunesMR. Soil health assessment: Past accomplishments, current activities, and future opportunities. Soil Tillage Res. (2019) 195:104365. doi: 10.1016/j.still.2019.104365

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 4

  BardgettR. The Biology of Soil: a Community and Ecosystem Approach. Oxford, UK: Oxford University Press (2010). doi: 10.1093/acprof:oso/9780198525035.001.0001

  • CrossRef
  • Google Scholar
  • View reference in article

• 5

  AszalósRThomDAakalaTAngelstamPBrūmelisGGálhidyLet al. Natural disturbance regimes as a guide for sustainable forest management in Europe. Ecol Appl. (2022) 32:e2596. doi: 10.1002/eap.2596

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 6

  JayaramanSDalalRC. No-till farming: Prospects, challenges - Productivity, soil health, and ecosystem services. Soil Res. (2022) 60:435–41. doi: 10.1071/SR22119

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 7

  Palomo-CampesinoSGarcía-LlorenteMHeviaVBoeraeveFDendonckerNGonzálezJA. Do agroecological practices enhance the supply of ecosystem services? A comparison between agroecological and conventional horticultural farms. Ecosyst Serv. (2022) 57(101474). doi: 10.1016/j.ecoser.2022.101474

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 8

  KnowlerDBradshawB. Farmers’ adoption of conservation agriculture: A review and synthesis of recent research. Food Policy. (2007) 32:25–48. doi: 10.1016/j.foodpol.2006.01.003

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 9

  GustafssonLBauhusJAsbeckTAugustynczikALDBasileMFreyJet al. Retention as an integrated biodiversity conservation approach for continuous-cover forestry in Europe. AMBIO. (2020) 49(1):85–97. doi: 10.1007/s13280-019-01190-1

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 10

  MasonWLDiaciJCarvalhoJValkonenS. Continuous cover forestry in Europe: Usage and the knowledge gaps and challenges to wider adoption. Forestry. (2022) 95:1–12. doi: 10.1093/forestry/cpab038

  • CrossRef
  • Google Scholar
  • View reference in article

• 11

  GhimireyVChaurasiaJAcharyaN. Understanding soil carbon sequestration: mechanistic insights, management approaches, and future challenges. Discover Soil. (2025) 2:101. doi: 10.1007/s44378-025-00133-5

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 12

  KabatoWHailegnawNMutumLMolnarZ. Managing soil health for climate resilience and crop productivity in a changing environment. Sci Total Environ. (2025) 1000:180460. doi: 10.1016/j.scitotenv.2025.180460

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 13

  HeneghanLMillerSPBaerSCallahamMAMontgomeryJPavao-ZuckermanMet al. Integrating soil ecological knowledge into restoration management. Restor Ecol. (2008) 16:608–17. doi: 10.1111/j.1526-100X.2008.00477.x

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 14

  Page-DumroeseDSSanchezFGUdawattaRPHobie PerryCGonzálezG. Soil Health Assessment of Forest Soils. In: Approaches to Soil Health Analysis, Volume 1. Hoboken, NJ, USA: John Wiley & Sons (2021). p. 100–38. doi: 10.1002/9780891189817.ch6

  • CrossRef
  • Google Scholar
  • View reference in article

• 15

  AllenDESinghBPDalalRC. Soil Health Indicators Under Climate Change: a Review of Current Knowledge. Berlin, Germany: Springer-Verlag (2011). pp. 25–45. doi: 10.1007/978-3-642-20256-8_2

  • CrossRef
  • Google Scholar
  • View reference in article

• 16

  DaleVHBeyelerSC. Challenges in the development and use of ecological indicators. Ecol Indic. (2001) 1:3–10. doi: 10.1016/S1470-160X(01)00003-6

  • CrossRef
  • Google Scholar
  • View reference in article

• 17

  GriffithsBSRömbkeJSchmelzRMScheffczykAFaberJHBloemJet al. Selecting cost effective and policy-relevant biological indicators for European monitoring of soil biodiversity and ecosystem function. Ecol Indic. (2016) 69:213–23. doi: 10.1016/j.ecolind.2016.04.023

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 18

  DynarskiKAAdelekeEBaumhardtRLBurkeJCarterTDeLaunePet al. Soil health indicator variability and management sensitivity across soils, bioregions, and agricultural systems. Agric Ecosyst Environ. (2026) 399:110203. doi: 10.1016/j.agee.2025.110203

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 19

  SudduthKAMyersDBKitchenNRDrummondST. Modeling soil electrical conductivity-depth relationships with data from proximal and penetrating ECa sensors. Geoderma. (2013) 199:12–21. doi: 10.1016/j.geoderma.2012.10.006

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 20

  KubiakKSpiralskiMPompeuJLevavasseurVWawerR. Advances in remote sensing for monitoring soil conditions in forest ecosystems: Techniques, challenges, and applications. Trans Aerospace Res. (2024) 2024:1–13. doi: 10.2478/tar-2024-0019

  • CrossRef
  • Google Scholar
  • View reference in article

• 21

  ThingujamUPrabhaDGhosh BagAThingujamVDarshanNPDuttaSet al. From point sensing to intelligent systems: a comprehensive review on advanced sensor technologies for soil health monitoring. Discover Sensors. (2025) 1:27. doi: 10.1007/s44397-025-00028-8

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 22

  KnoeppJDColemanDCCrossleyDAClarkJS. Biological indices of soil quality: An ecosystem case study of their use. For Ecol Manage. (2000) 138:357–68. doi: 10.1016/S0378-1127(00)00424-2

  • CrossRef
  • Google Scholar
  • View reference in article

• 23

  RinotOLevyGJSteinbergerYSvorayTEshelG. Soil health assessment: A critical review of current methodologies and a proposed new approach. Sci Total Environ. (2019) 648:1484–91. doi: 10.1016/j.scitotenv.2018.08.259

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 24

  Gatica-SaavedraPAburtoFRojasPEcheverríaC. Soil health indicators for monitoring forest ecological restoration: a critical review. Restor Ecol. (2023) 31:e13836. doi: 10.1111/rec.13836

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 25

  ParisiVMentaCGardiCJacominiCMozzanicaE. Microarthropod communities as a tool to assess soil quality and biodiversity: A new approach in Italy. Agric Ecosyst Environ. (2005) 105:323–33. doi: 10.1016/j.agee.2004.02.002

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 26

  GeorgePBLKeithAMCreerSBarrettGLLebronIEmmettBAet al. Evaluation of mesofauna communities as soil quality indicators in a national-level monitoring programme. Soil Biol Biochem. (2017) 115:352–60. doi: 10.1016/j.soilbio.2017.09.022

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 27

  FilserJFaberJHTiunovAVBrussaardLFrouzJDe DeynGet al. Soil fauna: Key to new carbon models. SOIL. (2016) 2:565–82. doi: 10.5194/soil-2-565-2016

  • CrossRef
  • Google Scholar
  • View reference in article

• 28

  ZhangKMaltais-LandryGLiaoHL. How soil biota regulate C cycling and soil C pools in diversified crop rotations. Soil Biol Biochem. (2021) 156:108219. doi: 10.1016/j.soilbio.2021.108219

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 29

  OjalaRHuhtaV. Dispersal of microarthropods in forest soil. Pedobiologia. (2001) 45:443–50. doi: 10.1078/0031-4056-00098

  • CrossRef
  • Google Scholar
  • View reference in article

• 30

  ThakurMPReichPBHobbieSEStefanskiARichRRiceKEet al. Reduced feeding activity of soil detritivores under warmer and drier conditions. Nat Clim Change. (2018) 8(1):75–8. doi: 10.1038/s41558-017-0032-6

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 31

  MentaCRemelliS. Soil health and arthropods: From complex system to worthwhile investigation. Insects. (2020) 11(1):54. doi: 10.3390/insects11010054

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 32

  JoimelSSchwartzCBonfantiJHeddeMKroghPHPérèsGet al. Functional and Taxonomic Diversity of Collembola as Complementary Tools to Assess Land Use Effects on Soils Biodiversity. Front Ecol Evol. (2021) 9. doi: 10.3389/fevo.2021.630919

  • CrossRef
  • Google Scholar
  • View reference in article

• 33

  Van StraalenNM. Evaluation of bioindicator systems derived from soil arthropod communities. Appl Soil Ecol. (1998) 9:429–37. doi: 10.1016/S0929-1393(98)00101-2

  • CrossRef
  • Google Scholar
  • View reference in article

• 34

  AndrésPMateosE. Soil mesofaunal responses to post-mining restoration treatments. Appl Soil Ecol. (2006) 33:67–78. doi: 10.1016/j.apsoil.2005.08.007

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 35

  GerlachJSamwaysMPrykeJ. Terrestrial invertebrates as bioindicators: An overview of available taxonomic groups. J Insect Conserv. (2013) 17(4):831–50. doi: 10.1007/s10841-013-9565-9

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 36

  RemelliSGhobariHOliveira FilhoLCI. Editorial: The role of soil mesofauna as indicators of sustainable ecosystem management plans. Front Ecol Evol. (2024) 12:1400232. doi: 10.3389/fevo.2024.1400232

  • CrossRef
  • Google Scholar
  • View reference in article

• 37

  RemelliSGrussIMentaC. Advances in soil arthropod identification: Integrating morphological, molecular, and AI-based approaches. Eur J Soil Biol. (2026) 128:103812. doi: 10.1016/j.ejsobi.2026.103812

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 38

  ParisiVMentaCGardiCJacominiC. Evaluation of soil quality and biodiversity in Italy: the biological quality of soil index (QBS) approach. In: Impacts on Soil Erosion and Soil Biodiversity. Francaviglia R (2003) pp. 541–50.

  • Google Scholar
  • View reference in article

• 39

  YanSSinghANFuSLiaoCWangSLiYet al. A soil fauna index for assessing soil quality. Soil Biol Biochem. (2012) 47:158–65. doi: 10.1016/j.soilbio.2011.11.014

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 40

  TsiafouliMAThébaultESgardelisSPde RuiterPCvan der PuttenWHBirkhoferKet al. Intensive agriculture reduces soil biodiversity across Europe. Global Change Biol. (2015) 21(2):973–85. doi: 10.1111/gcb.12752

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 41

  BokovaAIPaninaKSDridigerVKGadzhiumarovRGKuznetsovaNAPotapovMB. Soil-dwelling springtails as indicators of the efficiency of No-till technologies with different amounts of mineral fertilizers in the crop rotation on chernozem soils. Soil Tillage Res. (2023) 232:105760. doi: 10.1016/j.still.2023.105760

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 42

  LiuYSongLWuDAiZXuQSunXet al. No tillage increases soil microarthropod (Acari and Collembola) abundance at the global scale. Soil Ecol Lett. (2024) 6:230208. doi: 10.1007/s42832-023-0208-0

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 43

  AiZChenXWangBWuDZhangWAliAet al. How no-tillage and earthworms promote Collembola abundance by increasing large macropores. Soil Ecol Lett. (2025) 7:250302. doi: 10.1007/s42832-025-0302-6

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 44

  MachadoJSFilhoLCIOSantosJCPPaulinoATBarettaD. Morphological diversity of springtails (Hexapoda: Collembola) as soil quality bioindicators in land use systems. Biota Neotrop. (2019) 19:e20180618. doi: 10.1590/1676-0611-BN-2018-0618

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 45

  CoulibalySFMAubertMBrunetNBureauFLegrasMChauvatM. Short-term dynamic responses of soil properties and soil fauna under contrasting tillage systems. Soil Tillage Res. (2022) 215:105191. doi: 10.1016/j.still.2021.105191

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 46

  ReisFCarvalhoFMartins Da SilvaPMendesSSantosSAPSousaJP. The use of a functional approach as surrogate of Collembola species richness in European perennial crops and forests. Ecol Indic. (2016) 61:676–82. doi: 10.1016/j.ecolind.2015.10.019

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 47

  MorettiMDiasATCde BelloFAltermattFChownSLAzcárateFMet al. Handbook of protocols for standardized measurement of terrestrial invertebrate functional traits. Funct Ecol. (2017) 31(3):558–67. doi: 10.1111/1365-2435.12776

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 48

  PeyBNahmaniJAuclercACapowiezYCluzeauDCortetJÔet al. Current use of and future needs for soil invertebrate functional traits in community ecology. Basic Appl Ecol. (2014) 15(3):194–206. doi: 10.1016/j.baae.2014.03.007

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 49

  JoimelSPotapovAPeyBBonfantiJCortetJAlmeidaTDet al. Trait concepts, categories, and databases in soil invertebrate ecology - ordering the mess. Soil Org. (2024) 96:151–66. doi: 10.25674/431

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 50

  GisinH. Ökologie Und Lebensgemeinschaften Der Collembolen Im Schweizerischen Exkursionsgebiet Basels: Inauguraldissertation. Vorgelegt Der Philosophisch-Naturwissenschaftlichen Fakultät Der Universität Basel Von Hermann Gisin. Basel, Switzerland: Revue Suisse de Zoologie (1943).

  • Google Scholar
  • View reference in article

• 51

  PetersenH. General aspects of collembolan ecology at the turn of the millennium. Pedobiologia. (2002) 46(3-4):246–60. doi: 10.1078/0031-4056-00131

  • CrossRef
  • Google Scholar
  • View reference in article

• 52

  de FilhoLCIOKlauberg FilhoOBarettaDTanakaCASSousaJP. Collembola community structure as a tool to assess land use effects on soil quality. Rev Bras Ciencia do Solo. (2016) 40:1–18. doi: 10.1590/18069657rbcs20150432

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 53

  PotapovAASemeninaEEKorotkevichAYKuznetsovaNATiunovAV. Connecting taxonomy and ecology: Trophic niches of collembolans as related to taxonomic identity and life forms. Soil Biol Biochem. (2016) 101:20–31. doi: 10.1016/j.soilbio.2016.07.002

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 54

  HopkinSP. Biology of the Springtails (Insecta: Collembola). Oxford: OUP (1997).

  • Google Scholar
  • View reference in article

• 55

  SalmonSPongeJFGachetSDeharvengLLefebvreNDelabrosseF. Linking species, traits and habitat characteristics of Collembola at European scale. Soil Biol Biochem. (2014) 75:73–85. doi: 10.1016/j.soilbio.2014.04.002

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 56

  VandewalleMde BelloFBergMPBolgerTDolédecSDubsFet al. Functional traits as indicators of biodiversity response to land use changes across ecosystems and organisms. Biodivers Conserv. (2010) 19(10):2921–47. doi: 10.1007/s10531-010-9798-9

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 57

  FlóriánNGergócs-WinklerVSzili-KovácsTDombosM. Effects of different forestry management practices on soil mesofauna in the regeneration phase. For Ecol Manage. (2025) 598:123196. doi: 10.1016/j.agee.2025.109810

  • CrossRef
  • Google Scholar
  • View reference in article

• 58

  BrousseauPMGravelDHandaIT. On the development of a predictive functional trait approach for studying terrestrial arthropods. J Anim Ecol. (2018) 87(5):1209–20. doi: 10.1111/1365-2656.12834

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 59

  MentaCContiFDPintoSBodiniA. Soil Biological Quality index (QBS-ar): 15 years of application at global scale. Ecol Indic. (2018) 85:773–80. doi: 10.1016/j.ecolind.2017.11.030

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 60

  GalliL. An User Friendly Tool to Assess the Effects on Agricultural Soils of Different Practices: The QBS-Ar Index. Modern Concepts Developments Agron. (2020) 6(5):680–2. doi: 10.31031/mcda.2020.06.000648

  • CrossRef
  • Google Scholar
  • View reference in article

• 61

  MantoniCPellegriniMDapportoLDel GalloMMPaceLSilveriDet al. Comparison of soil biology quality in organically and conventionally managed agro-ecosystems using microarthropods. Agric (Switzerland). (2021) 11(10):1022. doi: 10.3390/agriculture11101022

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 62

  GalleseFGismero-RodriguezLGovednikAGiagnoniLLuminiESuhadolcMet al. Soil Microarthropods as Tools for Monitoring Soil Quality: The QBS-ar Index in Three European Agroecosystems. Agric (Switzerland). (2025) 15:89. doi: 10.3390/agriculture15010089

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 63

  Castillo-AvilaCCastillo-FigueroaD. First application of the QBS-ar index in the tropical montane Andes reveals soil biological quality across agroforestry systems and forest succession. Environ Monit Assess. (2026) 198:149. doi: 10.1007/s10661-025-14926-2

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 64

  MentaCLeoniABardiniMGardiCGattiF. Nematode and microarthropod communities: Comparative use of soil quality bioindicators in covered dump and natural soils. Environ Bioindic. (2008) 3:35–46. doi: 10.1080/15555270701885762

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 65

  NagličVŠibancNGrebencTBertonceljI. Soil mesofauna diversity in agricultural systems of Slovenia using the QBS index and its modifications. Acta Biologica Slovenica. (2025) 81(1):104–17. doi: 10.14720/abs.68.01.19787

  • CrossRef
  • Google Scholar
  • View reference in article

• 66

  LoriMLeitaoRFelixDImbertCCortiACunhaLet al. Response of soil biota to agricultural management practices: A systematic quantitative meta-data-analysis and method selection framework. Soil Biol Biochem. (2025) 207:109815. doi: 10.1016/j.soilbio.2025.109815

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 67

  JuhosKNugrohoPAJakabGPrettlNKotroczóZSzigetiNet al. A comprehensive analysis of soil health indicators in a long-term conservation tillage experiment. Soil Use Manage. (2024) 40:e12942. doi: 10.1111/sum.12942

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 68

  KovácsBTinyaFGubaENémethCSassVBidlóAet al. The short-term effects of experimental forestry treatments on site conditions in an oak-hornbeam forest. Forests. (2018) 8(7):406. doi: 10.3390/f9070406

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 69

  FlóriánNGergócs-WinklerVKovácsBAszalósRBidlóAÓdorP. Seasonality and crop type override the effects of long-term agricultural management on collembolan communities and traits. Agric Ecosyst Environ. (2025) 393:123196. doi: 10.1016/j.agee.2025.109810

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 70

  Gergócs-WinklerVFlóriánNKovácsBAszalósRBidlóAÓdorP. Leaf litter quantity shapes oribatid mite assemblages five years following forestry interventions in a European deciduous forest. Eur J Soil Biol. (2025) 127:103790. doi: 10.1016/j.ejsobi.2025.103790

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 71

  FlóriánNGergócs-WinklerVKovácsBAszalósRBidlóAÓdorP. Intermediate disturbance and regeneration: Long-term Collembola responses to forest management in an oak-hornbeam forest. Soil Ecol Lett. (2026) 8:260463. doi: 10.1007/s42832026-0463-y

  • CrossRef
  • Google Scholar
  • View reference in article

• 72

  StachJ. The Apterygotan Fauna of Poland in Relation to the World-Fauna of This Group of Insects. Kraków, Poland: Zoological Institute (1960).

  • Google Scholar
  • View reference in article

• 73

  FjellbergA. The Collembola of Fennoscandinavia and Denmark. 1. Poduridae. Fauna Entomological Scandinavica. (1998) 35:1–184. doi: 10.1163/9789004261051_007

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 74

  BretfeldGDungerW. Synopses on Palaearctic Collembola: Symphypleona. Abhandlungen und Berichte Des Naturkundemuseums Gorlitz. (1999) 71:1–318.

  • Google Scholar
  • View reference in article

• 75

  PotapovMB. Synopses on Palaearctic Collembola: Isotomidae. Abhandlungen und Berichte Des Naturkundemuseums Gorlitz. (2001) 73:1–603.

  • Google Scholar
  • View reference in article

• 76

  ThibaudJ-MSchulzH-Jda Gama AssalinaMM. Synopses on Palaearctic Collembola: Hypogastruridae. Görlitz: Sttatlisches Museum für Natürkunde Görlitz (2004).

  • Google Scholar
  • View reference in article

• 77

  HopkinSP. A Key to the Collembola (Springtails) of Britain and Ireland. Shrewsbury: Field Studies Council (2007).

  • Google Scholar
  • View reference in article

• 78

  JordanaR. Capbryinae & Entomobryini. Görlitz: Senkenberg Museum of Natural History Görlitz (2012).

  • Google Scholar
  • View reference in article

• 79

  HammerØHarperDAT. Past: Paleontological statistics software package for education and data analysis. Palaeontol Electron. (2001) 4(1):9. doi: 10.1002/9781119933960

  • CrossRef
  • Google Scholar
  • View reference in article

• 80

  DányiLTraserG. An annotated checklist of the springtail fauna of Hungary (Hexapoda: Collembola). Opuscula Zoologica. (2008) 38:3–82. Available online at: http://opuscula.elte.hu/PDF/Tomus38/Opuscula38.pdf (Accessed July 25, 2026).

  • Google Scholar
  • View reference in article

• 81

  SławskiMSławskaM. Seven decades of spontaneous forest regeneration after large-scale clear-cutting in Białowieza forest do not ensure the complete recovery of collembolan assemblages. Forests. (2019) 10(11):948. doi: 10.3390/f10110948

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 82

  YinREisenhauerNAugeHPurahongWSchmidtASchädlerM. Additive effects of experimental climate change and land use on faunal contribution to litter decomposition. Soil Biol Biochem. (2019) 131:141–8. doi: 10.1016/j.soilbio.2019.01.009

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 83

  YinRKardolPThakurMPGrussIWuGLEisenhauerNet al. Soil functional biodiversity and biological quality under threat: Intensive land use outweighs climate change. Soil Biol Biochem. (2020) 147:107847. doi: 10.1016/j.soilbio.2020.107847

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 84

  PolletSChabertABurgeonVCornélisJTFouchéJGersCet al. Limited effects of century-old biochar on taxonomic and functional diversities of collembolan communities across land-uses. Soil Biol Biochem. (2022) 164:108484. doi: 10.1016/j.soilbio.2021.108484

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 85

  RicottaCMorettiM. CWM and Rao’s quadratic diversity: A unified framework for functional ecology. Oecologia. (2011) 167(1):181–8. doi: 10.1007/s00442-011-1965-5

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 86

  R Core Team. R: A language and environment for statistical computing. R Foundation for Statistical Computing (2025). Available online at: https://www.r-project.org/ (Accessed July 25, 2026).

  • Google Scholar
  • View reference in article

• 87

  BatesDMächlerMBolkerBMWalkerSC. Fitting linear mixed-effects models using lme4. J Stat Software. (2015) 67:1–48. doi: 10.18637/jss.v067.i01

  • CrossRef
  • Google Scholar
  • View reference in article

• 88

  LenthRPiaskowskiJ. emmeans: Estimated Marginal Means, aka Least-Squares Means. R package version 2.0.0. (2025). [Preprint]. doi: 10.32614/cran.package.emmeans

  • CrossRef
  • Google Scholar
  • View reference in article

• 89

  CohenJ. Statistical Power Analysis for the Behavioral Sciences. New York, NY: Routledge Academic (1988).

  • Google Scholar
  • View reference in article

• 90

  HarrellF. _Hmisc: Harrell Miscellaneous_. (2026) 235–6. doi: 10.32614/cran.package.hmisc

  • CrossRef
  • Google Scholar
  • View reference in article

• 91

  ChiapperoMFRossettiMRMorenoMLPérez-HarguindeguyN. A global meta-analysis reveals a consistent reduction of soil fauna abundance and richness as a consequence of land use conversion. Sci Total Environ. (2024) 946:173822. doi: 10.1016/j.scitotenv.2024.173822

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 92

  ChassainJJoimelSVieublé GonodL. Collembola taxonomic and functional diversity in conventional, organic and conservation cropping systems. Eur J Soil Biol. (2023) 118:103530. doi: 10.1016/j.ejsobi.2023.103530

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 93

  DridigerVKIvanovALBelobrovVPKutovayaOV. Rehabilitation of Soil Properties by Using Direct Seeding Technology. Eurasian Soil Sci. (2020) 53:1293–301. doi: 10.1134/S1064229320090033

  • CrossRef
  • Google Scholar
  • View reference in article

• 94

  SchraderSKienleJAndersonTHPaulsenHM. Development of collembolans after conversion towards organic farming. Aspects Appl Biol. (2006) 79:181–6. Available online at: http://orgprints.org/9310/1/ABB06_Collembolans.pdf (Accessed July 25, 2026).

  • Google Scholar
  • View reference in article

• 95

  PongeJFPérèsGGuernionMRuiz-CamachoNCortetJÔPerninCet al. The impact of agricultural practices on soil biota: A regional study. Soil Biol Biochem. (2013) 67:271–84. doi: 10.1016/j.soilbio.2013.08.026

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 96

  MentaCContiFDLozano FondónCStaffilaniFRemelliS. Soil Arthropod Responses in Agroecosystem: Implications of Different Management and Cropping Systems. Agronomy. (2020) 10(6):982. doi: 10.3390/agronomy10070982

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 97

  GrevéMEHagerJWeisserWWSchallPGossnerMMFeldhaarH. Effect of forest management on temperate ant communities. Ecosphere. (2018) 9:e02303. doi: 10.1002/ecs2.2303

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 98

  TabaglioVGavazziCMentaC. Physico-chemical indicators and microarthropod communities as influenced by no-till, conventional tillage and nitrogen fertilisation after four years of continuous maize. Soil Tillage Res. (2009) 105(1):135–42. doi: 10.1016/j.still.2009.06.006

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 99

  BedanoJCCantúMPDoucetME. Influence of three different land management practices on soil mite (Arachnida: Acari) densities in relation to a natural soil. Appl Soil Ecol. (2006) 32(3):293–304. doi: 10.1016/j.apsoil.2005.07.009

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 100

  de GrootGAJagers op AkkerhuisGAJMDimmersWJCharrierXFaberJH. Biomass and diversity of soil mite functional groups respond to extensification of land management, potentially affecting soil ecosystem services. Front Environ Sci. (2016) 4:15. doi: 10.3389/fenvs.2016.00015

  • CrossRef
  • Google Scholar

• 101

  WehnerKSimonsNKBlüthgenNHeethoffM. Drought, windthrow and forest operations strongly affect oribatid mite communities in different microhabitats. Global Ecol Conserv. (2021) 30:e01757. doi: 10.1016/j.gecco.2021.e01757

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 102

  FlóriánNGrániczLGergócsVTóthFDombosM. Detecting soil microarthropods with a camera-supported trap. Insects. (2020) 11(4):244. doi: 10.3390/insects11040244

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 103

  OriolTPasquetJCortetJ. Automatic identification of Collembola with deep learning techniques. Ecol Inf. (2024) 81:102606. doi: 10.1016/J.ECOINF.2024.102606

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 104

  ZhuangXWangBXieZQiaoZFengJBehmJEet al. Deep learning method for collembola identification using single species and community combinations images. Soil Ecol Lett. (2025) 7:250352. doi: 10.1007/s42832-025-0352-9

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 105

  PolliererMMScheuS. Driving factors and temporal fluctuation of Collembola communities and reproductive mode across forest types and regions. Ecol Evol. (2017) 7(12):4390–403. doi: 10.1002/ece3.3035

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 106

  GrussIYinRJuliaSEisenhauerNSchädlerM. The responses of Collembola biomass to climate and land-use changes vary with life form. Soil Tillage Res. (2023) 225:105541. doi: 10.1016/j.still.2022.105541

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 107

  KlironomosJNKendrickB. Relationships among microarthropods, fungi, and their environment. Plant Soil. (1995) 170:183–97. doi: 10.1007/BF02183066

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 108

  ChauvatMZaitsevASWoltersV. Successional changes of Collembola and soil microbiota during forest rotation. Oecologia. (2003) 137:269–76. doi: 10.1007/s00442-003-1310-8

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 109

  KrantzGWWalterDE. A Manual of Acarology. Lubbock, TX, USA: Texas Tech University Press (2009). doi: 10.1111/j.1439-0469.2009.00546.x

  • CrossRef
  • Google Scholar

• 110

  GhimireRThapaVRCanoAAcosta-MartinezV. Soil organic matter and microbial community responses to semiarid croplands and grasslands management. Appl Soil Ecol. (2019) 141:30–7. doi: 10.1016/j.apsoil.2019.05.002

  • Pubmed Abstract
  • CrossRef
  • Google Scholar