Hardwood Tree Growth Aligns With Soil Texture and Specific Bacteria

A multi-site field evaluation across temperate hardwood plantations in Indiana and Michigan has revealed that tree growth and structural development correspond far more directly to local soil physical texture and discrete bacterial taxa than to overall microbial community diversity. Researchers investigated how soil chemistry, physical properties, and belowground prokaryotic assemblages influence the growth performance of northern red oak (Quercus rubra) and black walnut (Juglans nigra), two economically and ecologically vital timber species within the Central Hardwood Region of the United States. Across three established experimental forest plantations, researchers found that overall bacterial alpha and beta diversity metrics failed to correlate significantly with divergent growth categories. Instead, structural tree development—evaluated through continuous measurements of stem diameter at breast height and mean dominant canopy height—was tied to specific edaphic properties and distinct amplicon sequence variants. The investigation underscores that hardwood development is governed by intricate, species-specific plant-soil-microbe feedbacks, indicating that individual bacterial indicators rather than community-wide taxonomic richness serve as the primary microbial correlates of silvicultural performance.

Natural and managed temperate hardwood forests support critical regional biodiversity and ecological stability while contributing substantially to local and national timber economies. The sustainable development and long-term productivity of hardwood species depend upon belowground processes, where root networks interact bidirectionally with surrounding edaphic matrices and resident prokaryotic populations. Northern red oak and black walnut occupy contrasting physiological and environmental niches across the central hardwood landscape, exhibiting divergent soil adaptations and nutritional requirements. Black walnut typically demands fertile, moderately textured, well-aerated soils such as loam or sandy loam, accompanied by substantial calcium, magnesium, and potassium availability, while displaying marked vulnerability to waterlogging and compaction. Conversely, northern red oak demonstrates broader tolerance across acidic, nutrient-deficient, and variable soil structures. These differing physiological baselines establish contrasting belowground environments, modulating the composition, ecological network structure, and functional dynamics of tree-associated microbial consortia.

To rigorously examine these multi-trophic interactions under field conditions, the investigative team utilized a replicated multi-site field trial established between 2002 and 2008 across three distinct geographic locations: Western Michigan, Southern Indiana, and Western Indiana. Soils were gathered from six species-specific plantations of northern red oak and black walnut arranged in uniform randomized complete block designs with standardized planting densities of approximately three meters by three meters. The plantations were situated on Rockfield, Crosby–Miami, Blount, and Ryker–Muscatatuck silt loam soils with slope gradients ranging from 1% to 6%. Macroclimatic parameters varied systematically among locations: the Western Michigan site exhibited an average annual temperature of 9.5 degrees Celsius, approximately 940 millimeters of annual precipitation, and 350 millimeters of annual snowfall; the Western Indiana site recorded a mean annual temperature of 10.5 degrees Celsius, 1025 millimeters of rainfall, and 500 millimeters of snow; and the Southern Indiana site registered a mean annual temperature of 12.5 degrees Celsius, 1041 millimeters of precipitation, and 508 millimeters of annual snowfall. Importantly, no chemical fertilizers, artificial irrigation, or commercial microbial amendments were introduced into any plot, ensuring all observations reflected natural pedological and biological mechanisms.

Long-term vegetative growth assessments culminated in comprehensive stem diameter at breast height and total height measurements completed across all plots in November 2022. Across the entire experimental cohort, mean dominant height and stem diameter exhibited a strong, highly significant linear correlation, yielding a coefficient of determination of 0.8046. Individual trees were categorized into divergent growth performance tiers—designated as high-performing healthy phenotypes and low-performing poor phenotypes—by isolating specimens occupying the top and bottom 20% of block-level height and diameter distributions while strictly excluding the intermediate 60% of trees to avoid phenotypic overlap. Stand-level productivity metrics differed markedly across geographic boundaries, with substantially larger overall dimensions occurring in Southern and Western Indiana than in Michigan. Walnut generally surpassed oak in Indiana, whereas oak exhibited superior relative growth in Michigan. Healthy trees demonstrated 25% to 40% higher absolute growth rates and 10% to 20% higher relative growth rates across study sites compared to low-performing trees, while low-performing specimens accounted for 23% and 50% of walnut, and 12% and 35% of oak trees, in Michigan and Indiana, respectively.

Pedological investigations were performed on ninety-six composite soil samples gathered concurrently with final tree assessments in November 2022 across six established twenty-by-twenty-meter blocks separated by at least fifty meters. Within each block, twenty-five mineral soil cores spanning zero to ten centimeters depth were collected beneath cleared surface litter, combined, ground manually, and sieved through two-millimeter mesh. Physicochemical analyses performed at the University of Missouri Soil Health Assessment Center quantified soil particle size distributions via laser diffraction, measured soil pH in water and calcium chloride solutions, quantified electrical conductivity, and determined organic carbon and nitrogen through elemental combustion analysis. Cation exchange capacity was evaluated through neutral ammonium acetate extraction coupled with inductively coupled plasma optical emission spectrometry, and bulk density was evaluated using standardized core sampling and computed tomography scanning. Moisture contents were evaluated via time- and frequency-domain reflectometry, capturing detailed edaphic profiles underlying healthy and poor growth blocks across all three regional research plantations.

Quantitative soil physicochemical analyses revealed that organic carbon, organic nitrogen, and cation exchange capacity varied significantly between tree species, yet did not differ across geographic locations or growth performance classes. Available soil phosphorus showed no statistically significant differences across any grouping factor. Soil pH was significantly lower in soils supporting northern red oak and among plots containing poorly performing trees, while base saturation and exchangeable magnesium varied significantly across species, locations, and growth classes. Exchangeable calcium was markedly more abundant in black walnut soils, while potassium was higher in Indiana than Michigan soils. Texture analyses demonstrated pronounced geographical divergences: Michigan soils were characterized by high sand fractions, whereas Indiana soils were dominated by silt. Pearson correlation analyses indicated that northern red oak growth was negatively associated with calcium, magnesium, organic nitrogen, soil pH, and sand percentage, but positively associated with silt content. In contrast, black walnut development was positively correlated with soil pH and silt, while showing strong negative associations with sand content, reflecting species-specific edaphic sensitivities.

Prokaryotic community profiling was conducted using high-throughput 16S ribosomal RNA amplicon sequencing targeting the V4 hypervariable region with universal primers 515F and 806R on an Illumina MiSeq platform, generating read depths between 3025 and 5581 reads per sample. Following sequence quality filtering, primer trimming, and chimera removal via the DADA2 bioinformatic pipeline, amplicon sequence variants were aligned against the curated SILVA reference database and rarefied to 3025 reads per sample for alpha diversity evaluations. Permutational multivariate analysis of variance demonstrated that bacterial and archaeal community structure differed significantly by tree host species, explaining 10.7% of total variance, geographic location, explaining 9.4%, and their mutual interaction, explaining 7.57%. However, overall prokaryotic community composition did not differ significantly between healthy and poor growth states. Non-metric multidimensional scaling ordination separated samples by tree species along the first ordination axis and by geographic location along the second axis, while dominant phyla across all samples comprised Proteobacteria, Acidobacteria, and Actinobacteria, with lower taxonomic levels dominated by Gp1, Gp3, Gp6, Gp16, Spartobacteria incertae sedis, and Bradyrhizobium.

Bacterial alpha diversity, quantified using the Shannon diversity index, was significantly elevated in black walnut soils compared to northern red oak soils, and was higher in Western Indiana than Southern Indiana. However, statistical correlations revealed no significant association between Shannon diversity (or related richness metrics) and dominant tree height or stem diameter. Ecological co-occurrence networks constructed using the SpiecEasi pipeline and sparCC correlations revealed profound structural divergences driven by tree host identity. Black walnut supported markedly larger, denser co-occurrence networks with significantly higher node counts, edge numbers, and average node degrees than red oak networks, likely reflecting enhanced carbon and nitrogen availability in walnut rhizospheres. Furthermore, topological network metrics revealed divergent trends across performance classes: high-performing red oak individuals exhibited lower modularity, mean distance, and edge density, whereas high-performing black walnut trees showed the opposite structural patterns, indicating that host growth performance associates with contrasting microbial organizational architectures in different hardwood species.

Despite the absence of broad community-level shifts between performance classes, targeted Mantel tests demonstrated that fine-scale prokaryotic community structure correlated significantly with stem diameter at breast height in both black walnut and northern red oak, as well as with dominant tree height in walnut. Differential abundance analyses identified 88 growth-associated amplicon sequence variants in black walnut and 122 in red oak, with 59 and 45 variants enriched in low-performing walnut and oak trees, respectively. Low-performing walnut trees exhibited twice the relative abundance of Betaproteobacteria compared to high-performing trees, whereas high-performing oak trees were enriched in Acidobacteria subgroups GP1, GP2, and GP3, while GP6 and GP7 dominated low-performing individuals. Specific genera with over 50% prevalence showed significant positive correlations with growth metrics, including Bacillus, Arthrobacter, Aciditerrimonas, unclassified Chitinophagaceae, Gp1, and Gp3. Random Forest machine learning models confirmed that soil sand and silt percentages remained the primary predictors of hardwood development, while microbial taxa such as Bacillus, Arthrobacter, GP1, and GP3 accounted for meaningful proportions of variance in tree height and stem diameter.

The study authors emphasize several vital methodological and biological limitations, explicitly cautioning against premature causal interpretations. Because this research relied on correlational and observational field designs, direct causal mechanisms linking individual bacterial taxa or soil physical parameters to hardwood growth rates cannot be established. Furthermore, comparisons were constrained by a sample size of eight composite replicates per group, which limits statistical power, widens uncertainty bounds in effect size estimations, and elevates the risk of false-positive discoveries in complex machine learning, ordination, and differential abundance models. Biochemical processes that might mechanistically govern these associations—such as root exudate chemistry, allelopathic juglone dynamics in walnut, phenolic and lignin litter profiles in oak, or bacterial phytohormone synthesis—were not directly quantified. Consequently, the researchers conclude that incorporating microbial indicators into silvicultural management, reforestation protocols, or site suitability indices remains exploratory and requires targeted experimental validation across varied soil horizons, developmental stages, and controlled silvicultural interventions.

Subject of Research: Soil properties and prokaryotic community associations with hardwood tree development

Article Title: Evaluating the role of biotic and abiotic factors in hardwood tree development in the Central Hardwood Region (CHR)

Article References: Lawson, S. S., Frene, J. P., & Lue Sue, N. D. (2026). Evaluating the role of biotic and abiotic factors in hardwood tree development in the Central Hardwood Region (CHR). Environmental Microbiome, 21(1), Article 120. https://doi.org/10.1186/s40793-026-00935-2

Image Credits: AI Generated

DOI: 10.1186/s40793-026-00935-2

Keywords: Juglans nigra, Quercus rubra, Central Hardwood Region, soil microbiome, 16S rRNA sequencing, soil texture, Bacillus, Arthrobacter, Acidobacteria, diameter at breast height, dominant tree height, Random Forest modeling, silviculture, microbial co-occurrence, cation exchange capacity, soil pH, forest ecology