Microscale Heterogeneity of the Spatial Distribution of Organic Matter Can Promote Bacterial Biodiversity in Soils: Insights From Computer Simulations

Abstract

There is still no satisfactory understanding of the factors that enable soil microbial populations to be as highly biodiverse as they are. The present article explores in silico the hypothesis that the heterogeneous distribution of soil organic matter, in addition to the spatial connectivity of the soil moisture, might account for the observed microbial biodiversity in soils. A multi-species, individual-based, pore-scale model is developed and parameterized with data from 3 Arthrobacter sp. strains, known to be, respectively, competitive, versatile, and poorly competitive. In the simulations, bacteria of each strain are distributed in a 3D computed tomography (CT) image of a real soil and three water saturation levels (100, 50, and 25%) and spatial heterogeneity levels (high, intermediate, and low) in the distribution of the soil organic matter are considered. High and intermediate heterogeneity levels assume, respectively, an amount of particulate organic matter (POM) distributed in a single (high heterogeneity) or in four (intermediate heterogeneity) randomly placed fragments. POM is hydrolyzed at a constant rate following a first-order kinetic, and continuously delivers dissolved organic carbon (DOC) into the liquid phase, where it is then taken up by bacteria. The low heterogeneity level assumes that the food source is available from the start as DOC. Unlike the relative abundances of the 3 strains, the total bacterial biomass and respiration are similar under the high and intermediate resource heterogeneity schemes. The key result of the simulations is that spatial heterogeneity in the distribution of organic matter influences the maintenance of bacterial biodiversity. The least competing strain, which does not reach noticeable growth for the low and intermediate spatial heterogeneities of resource distribution, can grow appreciably and even become more abundant than the other strains in the absence of direct competition, if the placement of the resource is favorable. For geodesic distances exceeding 5 mm, microbial colonies cannot grow. These conclusions are conditioned by assumptions made in the model, yet they suggest that microscale factors need to be considered to better understand the root causes of the high biodiversity of soils.

Keywords: agent-based modeling; bacteria; biodiversity; organic matter; pore scale; resource allocation; soil.

FIGURE 1

General workflow of the Ib-LBioS-Comp model.

FIGURE 2

General workflow of the bacterial agents loop of the Ib-LBioS-Comp model.

FIGURE 3

2D view of the POM initialization scheme used in the scenarios. In the figure, the POM (red), the DOC (yellow), the soil solid matrix (black), and the pore space (white) are depicted. In (a), POM disaggregated, four parallelepiped POM fragments of 1 × 1 × 3 image voxels, were assumed to be present and connected to the pore space through a single voxel. In (b), POM aggregated, a single 3D POM fragment of 4 × 4 × 3 was assumed to be present and connected to the pore pace through 4 of its voxels.

FIGURE 4

Global biodegradation kinetics (left) and strains abundance (right) of scenarios with dispersed POM fragments (S1) and aggregated POM fragments (S2). M/M₀ is the ratio of the carbon mass (mg C) of each output over the total initial carbon mass M₀.

FIGURE 5

Global biodegradation kinetics (left) and strains abundance (right) of the scenario with the C available as DOC and a water saturation level of 0.5 (S3). M/M₀ is the ratio of the carbon mass (mg C) of each output over the total initial carbon mass M₀.

FIGURE 6

Biomass growth kinetics of the three strains (3R, 9R, and 7R) of each of the 690 spots in scenarios of dispersed POM fragments (S1, Top figures) and aggregated POM fragments (S2, Bottom figures). M/M₀ is the ratio of the carbon mass (mg C) of the cells over the total initial carbon mass M₀. For each scenario, the replicate having the minimal, mean and maximum biomass growth is displayed from left to right.

FIGURE 7

Total biomass growth of each of the 690 spots against the geodesic distances between the spots and the POM fragments of scenarios S1 (Top figures) and S2 (Bottom figures). M/M₀ is the ratio of the carbon mass (mg C) of the cells over the total initial carbon mass M₀. For each scenario, the replicate having the minimal, mean and maximum biomass growth is displayed from left to right.

FIGURE 8

Global biodegradation kinetics (left) and strains abundance (right) of replicate 2 of the scenario with aggregated POM fragments (S4) in which the three strains are initially alone in the spots (S2r2, solid lines) or gathered in the spots (S4a dotted lines and S4b dashed lines). M/M₀ is the ratio of the carbon mass (mg C) of each output over the total initial carbon mass M₀. Simulation outputs obtained in the second repetition of the scenario S2 (solid lines) are compared to the two simulations of the scenario S4 (dashed lines).

FIGURE 9

Global biodegradation kinetics (left) and strains abundance (right) of the scenario with aggregated POM fragments and a water saturation level of 1.0 (S5). M/M₀ is the ratio of the carbon mass (mg C) of each output over the total initial carbon mass M₀.

FIGURE 10

Global biodegradation kinetics (left) and strains abundance (right) of the scenario with aggregated POM fragments and a water saturation level of 0.25 (S6). M/M₀ is the ratio of the carbon mass (mg C) of each output over the total initial carbon mass M₀. The repetition number generating the trends depicted are detailed in the figure.