Initial Cell Number Regulates Microscale Community Assembly and Structure

Technical Summary: Initial Cell Number Regulates Microscale Community Assembly and Structure

Problem Statement
Soil microbial communities exhibit pronounced spatial heterogeneity at the microscale, existing as small aggregates of tens to hundreds of cells within biofilms. While the roles of abiotic drivers and biotic interactions in soil community assembly are well-documented, the influence of the initial cell number (founding population size) as a fundamental constraining factor remains severely understudied. Theoretical frameworks suggest that low initial densities may amplify stochastic processes (e.g., ecological drift, priority effects), while higher densities may favor deterministic processes (e.g., competition, cooperation). However, few studies have systematically examined how initial cell number impacts the evolutionary dynamics and assembly of low-complexity communities across varying microhabitat conditions. This study addresses the gap in understanding how the starting population size in microscale aggregates drives community assembly trajectories, structural convergence, and functional differentiation.

Methodology
The researchers employed a high-throughput, flow cytometry-based approach to precisely control the initial inoculation of soil-derived microorganisms.

• Sample Preparation: Soil microorganisms were extracted from surface soil using a modified Nycodenz gradient centrifugation method. Viable cells were identified via Propidium Iodide (PI) staining to exclude those with compromised membranes.
• Experimental Design: Using flow cytometric sorting, the team constructed approximately 8,000 parallel random communities in 96-well microcosms. They precisely controlled initial cell numbers at four levels: 10, 50, 100, and 250 cells per well.
• Environmental Variables: Four distinct carbon sources (cellobiose, glutamic acid, trehalose, and mannitol) were used as sole carbon sources to simulate nutrient heterogeneity found in natural soil microhabitats.
• Phenotypic and Genotypic Analysis:
  • Phenotypes: Growth (OD₆₀₀) and biofilm formation (crystal violet staining) were quantified over 14 days. Communities were categorized into four phenotypic groups based on high/low growth and biofilm levels.
  • Genomics: Representative samples (N=384) underwent 16S rRNA gene amplicon sequencing (V4 region).
  • Bioinformatics: Data analysis included diversity metrics (Shannon, Chao1), source tracking (FEAST), differential abundance testing (ANCOM-BC2), and microbiome genus set enrichment analysis (mgsea).
  • Network Analysis: Sparse Latent Factor Regression (SLR) was used to construct microbial association networks, separating the direct effects of initial cell number from putative microbial interactions.

Key Results

1. Phenotypic Regulation: Initial cell number exerted carbon-source-dependent effects on growth but showed a consistent, strong positive correlation with biofilm formation across all carbon sources. Higher founding densities promoted biofilm matrix production independent of total biomass accumulation, suggesting a mechanism linked to quorum sensing and Allee-like effects.
2. Structural Convergence: Post-incubation communities underwent significant simplification, converging toward structures dominated by Pseudomonas and Klebsiella. Source tracking revealed that while incubated communities originated largely from the soil extract, the cultivation process imposed intense environmental filtering, retaining only a tiny fraction of native soil diversity.
3. Diversity and Evenness: Shannon diversity increased significantly with initial cell number, driven primarily by improved evenness of species abundances rather than an increase in richness (Chao1 or observed ASVs remained stable).
4. Dominant Taxa Switching: A distinct, cell-number-dependent switching of dominant taxa was observed:
   • Low initial cell numbers (10–50): Strongly enriched Pseudomonas species.
   • High initial cell numbers (100–250): Increased abundance of Klebsiella species.
   • This reciprocal pattern was robust across carbon sources and exceeded the effect size of carbon source type itself.
5. Intra-Genus Niche Differentiation: The cell-number preference extended to the Amplicon Sequence Variant (ASV) level. Different ASVs within the same genus (e.g., Pseudomonas) exhibited heterogeneity in their preference for initial cell numbers, indicating fine-scale niche partitioning.
6. Association Networks:
   • Low cell-number communities: Harbored more unique ASVs and were characterized by negative-correlation networks, reflecting amplified stochastic processes and local extinction risks.
   • High cell-number communities: Promoted shared ASVs and positive-correlation networks, suggesting deterministic processes and cooperative interactions.
   • SLR analysis identified initial cell number as the primary latent variable driving variation in microbial association networks.

Significance and Claims
The paper claims that initial cell number is a critical driver of microscale soil microbial community assembly. The study provides direct experimental evidence that:

• Initial cell number acts as a core constraining factor that shapes community structure, diversity, and function, often overriding the influence of specific carbon sources.
• It regulates the competition–coexistence dynamics of dominant taxa and drives intra-genus niche differentiation at the ASV level.
• It fundamentally alters microbial association networks, shifting communities from stochastic, negative-interaction dominated states (at low densities) to deterministic, positive-interaction dominated states (at high densities).
• These findings offer a mechanistic explanation for the spatial heterogeneity of soil microbiomes and provide a theoretical basis for predicting community responses to environmental changes and designing microbial engineering strategies (e.g., biofilm-enhanced remediation) based on initial cell density.

The authors emphasize that while genus-level trends are observable, the core mechanism driving these dynamics lies in the ASV-level heterogeneity and the specific preferences of strains for founding population sizes, which cannot be inferred from higher taxonomic ranks alone.