What if one simple measurement could reveal how agricultural soil is functioning—and the best clue turned out to be soil moisture? This study tests whether microbial activity can serve as a fast biological health indicator. Long-term agricultural activity may lead to substantial changes in chemical, physical, and biological soil properties and to overall soil health status.

Reliable indicators are needed to estimate those changes, and they must be sensitive to changes in land use and land management while also being cost-effective. The challenge is even larger for biological soil health indicators because they represent soil biota activity, composition, density, diversity, and trophic interactions.

Biological soil health indicators therefore face two questions: which basic biological soil properties are most sensitive to changes in land use, and what is the best season for soil sampling when seasonality affects soil biota composition and activity? The objective is to determine a simple and cost-effective biological indicator sensitive to changes in land use.

The indicator combines a biological component with abiotic variables previously reported as most sensitive for detecting soil variation caused by agricultural management. The indicator was tested in two seasons. Soils were sampled in two climatic regions of Israel during Spring, or late winter, and Autumn, or late summer.

The regions contained wide and intensive agricultural systems with three common management practices, but differed in climate and soil texture. The Yizrael valley had a Mediterranean climate and clayey soils, while the north-western Negev had a semi-arid climate and silty to sandy soils.

Figure one maps the six Israeli study sites across two climatic regions. Ein Harod, Geva, and Nir Yafe are in the Mediterranean Yizrael Valley, characterized by clayey soils, while Magen, Chavat HaShikmim, and Talmey Yosef are in the north-western Negev.

This geographic layout matters because the authors compare biological soil indicators across contrasting climates and soil textures, and sample each region in spring and autumn. Soil samples were collected at three depths along the one-meter soil profile: zero to ten centimeters, ten to thirty centimeters, and thirty to sixty centimeters, using three samples from each of eighteen sampling pits.

Approximately one kilogram from each sample was immediately sieved by a two-millimeter sieve, sealed, and placed in a cooling box at four degrees Celsius for further analyses. The sampling produced a dataset of one hundred sixty-two observations for each season.

Community-level physiological profiling, or CLPP, measured substrate utilization profiles in the soil microbial communities. The modified MicroResp method constructed sole-carbon-source utilization profiles of those communities. Fifteen carbon sources were added to soil samples in deep-well plates, covering aromatic carboxylic acids, carboxylic acids, carbohydrates, and amino acids.

The statistical analysis consisted of canonical correlation analysis followed by regression analysis. Canonical correlation analysis provides a multivariate interpretation between sets of variables, rather than relating individual variables one at a time.

Set one contained soil abiotic components, including soil moisture, organic matter, electrical conductivity, pH, total nitrogen, organic carbon, and carbon-to-nitrogen ratio. Set two contained observations on substrates: carbon dioxide evolution, the Shannon-Weaver index, microbial biomass, and CLPP, identified as potential response variables.

The statistical analyses were applied separately to observations made in Spring and Autumn, rather than combining the two seasonal datasets. Figure two compares autumn and spring soil properties using box plots, with Set one variables describing substrate characteristics and Set two capturing abiotic and biotic measures.

The Set two panels show especially tight spring distributions for carbon dioxide evolution, microbial biomass, and CLPP, while autumn generally has broader spreads and more separated values. The visual also shows many outliers, and the authors note a strong spring correlation between organic carbon and total nitrogen, prompting the exclusion of organic carbon from Set one.

The box plot analysis shows large differences in the Spring data for the Set two variables. In Spring, carbon dioxide evolution, microbial biomass, and CLPP have little variation, with similar first, second, and third quartiles. Autumn shows higher variation, indicating that substrate characteristics are more distinguished in the dry season than in the wet season.

Table One reports canonical correlations between Set One and Set Two for the spring data. The first canonical variable has a correlation of zero point four nine one and is significant at less than zero point zero zero one, while the second and third correlations, zero point two three five and zero point two zero two, are non-significant, with significance values of zero point three eight six and zero point four eight two.

The authors therefore focus interpretation on CV1, linking higher soil moisture with lower values of MB, CLPP, and H. Table Three reports the canonical correlations between the two variable sets in the autumn data. The first canonical variable, or C-V one, has a correlation of zero point four two three and a significance value of zero point zero one four, indicating the only statistically significant association.

C-V two and C-V three have significance values of zero point four one five and zero point seven seven nine, respectively, so the table identifies them as non-significant. Soil moisture non-linearly affects all four microbial properties in Spring, and carbon dioxide evolution and CLPP in Autumn.

In every case, activity is minimal at intermediate soil moisture and increases at high or low soil moisture. Figure four shows model-predicted relationships between soil moisture, or SM, and four microbial properties across spring and autumn. In spring, H-prime, MB, and CLPP decline as SM increases, while MB shows a curved response with activity lowest at intermediate moisture; in autumn, both CO2 evolution and CLPP follow curved patterns, with values increasing toward high or low SM.

This matters because the figure illustrates the reported non-linear moisture effects and indicates that some seasonal property combinations had no supported SM effect. The study concludes that soil moisture drives different processes and serves as a proxy for biological soil indicators.

Canonical correlation analysis and regression analysis indicate strong relations between soil moisture and CLPP alone in Autumn, and between soil moisture and CLPP, microbial biomass, and the Shannon-Weaver index in Spring. Soil moisture significantly affects CLPP in both seasons.

Because no other variables were selected as explanatory variables, CLPP is identified as a sensitive soil biological indicator in regions with Mediterranean climate. Further analysis is required, with attention to more variables, more locations, or more controlled conditions.

Across the tested Mediterranean and semiarid agricultural soils, soil moisture was consistently linked to microbial activity, especially CLPP. The results point to autumn CLPP as a useful rapid indicator, while also calling for broader testing.