Seeds of a Wyoming native perennial herb, B. stricta, were originally collected from the Snowy Range Mountains (41.32971759902109 N, −106.50515422710646 W) and grown for one generation in the greenhouse to increase seed numbers and minimize maternal effects. Before planting, seeds were surface sterilized by rinsing for 1 min in a mixture of 70% ethanol, 0.1% Triton, and 30% RO water, followed by rinsing in RO water, and then rinsing for 12 min in a mixture of 10% bleach, 0.1% Triton, and 90% water. Seeds were then rinsed three more times with RO water before being placed on sterile filter paper for ease of planting (adapted from Lundberg et al., 2012).

Field soil was collected from unvegetated sites adjacent to a field location with a native B. stricta population, referred to hereafter as the Crow Creek field site (CRW), and sieved to 4 mm to remove large debris. The soil was then autoclaved three times for 30 min, with mixing between autoclaving steps. The soil was then mixed with autoclaved potting soil [Redi-Earth Potting Mix (Sungro Horticulture, Agawam, MA, USA) in a 9:1 ratio]. We included this small percentage of potting mix because its greater water-holding capacity relative to the field soil improves the overall rate and synchrony of seed germination. This soil mixture was next inoculated with a 4% v/v of non-autoclaved field soil inoculum. Although autoclaving reduces the similarity of the study system to the natural field soil environment, it reduces variability between replicates and removes any pest organisms within the soil. Inoculum from field soil was added to restore the autoclaved soil to a similar but not identical microbial soil ecosystem.

This soil mixture was used to plant the surface-sterilized seeds. In addition to pots planted with seeds, which would be used to characterize the rhizosphere microbiome, we prepared soil-filled pots without plants, which would be used to estimate the microbiome of bulk soil or unvegetated microsites, referred to hereafter as the “control soil.” Seeds were germinated in 2″ mesh net pots (2″ Inch TEKU Net Slit Pots for Hydroponic Aeroponic Use) with the bottom surrounded by sterilized coffee filters and plastic cups to prevent soil erosion. Plants were allowed to grow for 4 weeks under greenhouse conditions (UW Laramie Research and Extension Center, Laramie, WY) with ambient day/night light and temperature cycles before field transplanting. Mesh pots were placed in a randomized checkerboard array in tray blocks, such that no pot was directly adjacent to another. Plastic covers were placed over all trays to retain humidity and promote germination. All materials for planting, such as bench tops, trays, pots, and covers, were bleached and rinsed before use. Pots were individually watered, initially via subirrigation and after 2 weeks of growth via overhead misting. Covers were removed 2 weeks after germination. Three weeks after germination, all pots were acclimated to the outside environment via 2-h field exposures.

Four weeks after plant germination in May of 2022, plants were transferred to the Crow Creek field site, which has a naturally occurring population of B. stricta. The CRW study site is located in the Medicine Bow-Routt National Forest in southeastern Wyoming (41.227318 N, −105.383343 W), at an elevation of ~2,560 m. Six 26 cm × 140 cm plots were cleared of plants and debris. Mesh pots were randomly assigned to each plot and planted approximately 10 cm apart in two rows of 24 pots. Mesh pots were removed from filters and cups and placed directly in field site plots to minimize root damage while transplanting and to facilitate the collection of plant rhizospheres. Plant germination time, rosette size, and true leaf number were measured weekly to estimate plant performance. All pots were surface watered every other day using RO water and checked for insect damage.

Before planting in the field, while the plants were acclimating to the environment outside the greenhouse, we collected chlorophyll fluorescence data as a measure of photosystem II activity (Murchie and Lawson, 2013) over 5 days. Specifically, for each plant, we measured chlorophyll fluorescence every 3 h starting at 4:30 AM before sunrise, with an additional timepoint at 5:30 AM immediately after sunrise, and continuing until 8:30 PM after sunset. This data was used to choose two timepoints where the inferred physiological activity of the plant was most divergent from one another, which we hypothesized could reflect timepoints when the microbial community was most divergent if plant physiology were affecting the rhizosphere environment. We selected a pre-dawn timepoint of 3 AM and an early afternoon timepoint of 2 PM (Figure 1).

Four weeks after transplanting the plants to the field, when the plants were 8 weeks past germination, soil and plant samples were harvested over three consecutive days, from June 22 to June 24, 2022. Sampling occurred during pre-dawn hours from 3:00 to 3:30 AM and in the early afternoon from 2:00 to 2:30 PM. Samples were randomly collected, ensuring an equal number of samples per treatment of rhizosphere and control soils, with samples taken at both AM and PM. Samples from the two treatment types were handled as follows: (1) For the pots without plants, soil was stored in whirl packs (Whirl-Pak^® Standard Sterilized Sampling Bags, Nasco Sampling LLC., Pleasant Prairie, WI, United States), and smaller subsamples of these soils were taken and stored in 2 mL test tubes for later nucleic acid extraction; (2) For collection of rhizosphere soil from the pots with plants, mesh pots were removed from the soil. Care was taken to ensure that roots growing out of the mesh pots were damaged minimally; plants were then removed from the pots and shaken to remove excess loose soil. Soil that adhered closely at approximately 1 mm to the root mass was considered rhizosphere soil. Roots and adhered soil were separated from plant leaves and stems using flame-sterilized scissors and placed in Falcon tubes containing approximately 200 mL of PBS buffer (200 μL Silwet, 900 mL RO water, 100 mL 10 × PBS; adapted from Bulgarelli et al., 2012). Samples were stored on ice during the fieldwork. Immediately upon returning to the lab, the control soil samples were transferred to a −80 °F freezer, and the rhizosphere soil samples were processed. Falcon tubes containing rhizosphere soil and plant root tissue were defrosted, then lightly vortexed to remove adhered soil from root tissue. Root tissue was removed using sterile forceps. The soil slurry was then vacuum-filtered through a 0.2 nm filter. The resulting soil and filter were transferred to a 0.5 mL centrifuge tube and flash-frozen with liquid nitrogen before being stored long-term at −80 °C.

This study sought to eliminate biases that might artificially inflate the differences between DNA- and RNA-derived community composition measurements. Methodological biases were minimized via the simultaneous extraction of nucleic acids from a single soil sample, so only one community profile from one soil aliquot in both DNA and RNA was analyzed (Gentile et al., 2006; McCarthy et al., 2015; Moeseneder et al., 2005; Morgan et al., 2010). Different methods of RNA and DNA extraction, such as RNA having a reverse transcription step, whereas DNA does not, cannot be corrected for (Zhen et al., 2015). Control and rhizosphere soil RNA and DNA samples were simultaneously extracted using the methods described in the RNeasy PowerSoil Total RNA Kit and RNeasy PowerSoil DNA Elution Kit (https://digitalinsights.qiagen.com). After extraction, RNA was reverse-transcribed into cDNA using the QuantiTect Reverse Transcription Kit (https://digitalinsights.qiagen.com). Negative control blank samples were included for extractions and reverse transcription. Samples were stored at −20 °C until further processing.

Both 16S-rRNA-gene (DNA) and 16S-rRNA-transcript (RNA) amplicons were sent to The University of Chicago Marine Biological Laboratory in Woods Hole for sequencing and taxonomic identification (https://vamps.mbl.edu). Illumina MiSeq was used with bacterial v4v5 primers, and taxonomy was assigned using the SILVA 119 database on the VAMPS (Visualization and Analysis Microbial Population Structures) website (Huse et al., 2014). The SILVA 119 database was used so that a direct comparison of taxa between this and previous studies (i.e., Ceretto and Weinig, 2025) would be equivalent.

Sequences, read counts, taxonomic IDs, and sequence ID names were downloaded from the MBL VAMPS website portal (Huse et al., 2014) for OTU table generation and analysis. RNA samples underwent an extra round of PCR compared to DNA samples, as cDNA is single-stranded and requires an extra step to compare with double-stranded DNA.

Reads that were not assigned to the kingdom of bacteria were removed. Reads that appeared in the negative control samples and were present at low frequency in other samples (i.e., less than 100 reads per sample) were considered contamination and subsequently removed. Sequences that occurred five or more times in all samples were considered potential OTUs. Samples with less than 25,000 total reads were removed from analysis, and samples were filtered so each RNA-derived sample had a corresponding DNA sample that was extracted from the same collected soil. OTUs that occurred in less than 5% of all samples with less than two reads were removed.

The final data set consisted of 234 total samples: 117 RNA samples, comprising 63 from rhizosphere soil and 54 from control soil, and 117 DNA samples, comprising 63 from rhizosphere soil and 54 from control soil. These samples contained 11,402 unique OTUs and 11,137,182 total reads (63% RNA reads and 36% DNA reads). Notably, selecting an even number of samples at random from each treatment rhizosphere, or control, did not alter the conclusions of the analysis; therefore, we present results based on all 234 samples.

In R [version 4.3.1 (2023-06-16)], samples were rarefied to 25,491 reads per sample, and alpha diversity was estimated using the phyloseq package “estimate_richness,” with differences between alpha diversity categories determined using Student's T-test. Statistical differences between community nucleic acid derivation and soil type were determined using ANOVA (analysis of variance). To calculate differences between the RNA- and DNA-derived communities, the model OTU abundance (within the RNA- or DNA-derived community) = Soil Type ^* Collection Date ^* Field Block, where soil types were rhizosphere and bulk-control soils, collection date was 1 of 3 days where samples were collected, and field block is the approximately two by 1.0 m rectangular field block at the CRW site which contained study samples. The model Soil Type ^* Collection Date ^* Field Block was used to calculate significance for the RNA- and DNA-derived communities separately.

We calculated the significance of the collection timepoint (AM or PM). To remove the confounding effect of collection date on collection time, samples were first separated into three groups based on the date of collection and whether they were derived from RNA- or DNA-derived communities. The model OTU abundance (within the RNA- or DNA-derived community) = Soil Type ^* Collection Time ^* Field Block was used to calculate significance, where Collection Time represented the AM and PM timepoints. Samples were separated by RNA- and DNA-derived communities and soil type, but kept together based on collection day to calculate significance again using the model OTU abundance (within RNA- or DNA-derived community) = Collection Time.

To calculate beta diversity, we used the package Phyloseq v1.30.0 (McMurdie and Holmes, 2013). Reads assigned to an OTU in a sample were divided by the total number of reads per sample to calculate within-site proportional abundances. PCA plots were used to visualize Bray-Curtis pairwise dissimilarities in community data. Significant differences were detected between RNA- and DNA-derived communities, as well as between the rhizosphere and control soils within and between communities using PERMANOVA (pairwise Adonis testing with Bonferroni correction). Barplots comparing taxa within the samples were generated using the proportional abundances of reads, with pairwise t-tests used to determine significance between groups. The corncob package (Martin et al., 2020) was used to estimate differential abundances between the RNA- and DNA-derived communities, as well as between the RNA-derived and DNA-derived rhizosphere and control soil samples, and between the RNA-derived and DNA-derived rhizosphere and control soil samples by collection time of 3 AM and 2 PM, using the absolute abundance of reads, rather than proportional abundances.

Chlorophyll fluorescence data showed a change in mean Fv′/Fm′ from dawn to dusk, with the lowest mean Fv′/Fm′ around 2 PM. A high mean Fv′/Fm′ was observed at timepoints when it was dark, pre-dawn at 4:30 AM and post-sunset at 8:30 PM (Figure 1). Although the highest mean Fv′/Fm′ occurred at 5:30 AM just after sunrise, we chose the pre-dawn timepoint, because it provided a large window of time to collect samples in the dark instead of rushing to finish sampling during the brief hour of dawn. It also allowed us to collect samples earlier, before sunrise, when it was still dark, if more time was needed for collection. We chose not to collect samples immediately after sunset, as we felt giving the plants longer to acclimatize to night conditions would better magnify any diel changes mediated by the plant in the rhizosphere community.

Between the RNA- and DNA-derived communities, some phantom taxa and reads from possibly inactive or dead cells were detected but made up less than 1% of the total number of reads. After data QA/QC, 394 taxa, represented by 1,018 reads, were unique to the RNA-derived community profile, and 44 taxa, consisting of 222 reads, were unique to the DNA-derived community.

Using rarefied communities, observed alpha diversity significantly differed between the RNA- and DNA-derived communities in the Control AM (p < 0.05), Rhizosphere PM (p < 0.01), and Rhizosphere AM (p < 0.01) soils. Only the DNA control soil AM and RNA control soil AM differed in Shannon (p < 0.05) and Simpson (p < 0.01) diversity (Figure 2). This indicates that within a sample, the diversity of what bacteria are present differs from the diversity of which bacteria are contributing to the community PSP profile. Visualization of the relative abundance of phyla (Figure 3) between the RNA- and DNA-derived communities showed significant differences between the eight soil-time sample types in all but 2 (Elusimicrobia, Nitrospirae) of the 18 total number of phyla identified in this study, indicating that microbial membership and community PSP profile differed for all soil-time sample groups. Three phyla showed minimal to no differences between the RNA- and DNA-derived communities, indicating that, in terms of the number of microbial members in the community, they contribute equivalently to the community PSP profile. Chloroflexi only showed significant (^*p < 0.05) differences between the RNA- vs. DNA-derived rhizosphere soils collected at the AM timepoint. Furthermore, Actinobacteria (DNA vs. RNA rhizosphere PM collection p > 0.05 NS) and Verrucomicrobia (DNA vs. RNA control soils PM collection p > 0.05 NS; Figure 3).

Beta diversity analysis via ANOVA model OTU abundance (within RNA- or DNA-derived community) = Soil Type ^* Collection Date ^* Field Block on the RNA- and DNA-derived communities together showed RNA (nucleic acid of origin) to be significant (^***P < 0.001; R² = 0.20436) and (Figure 4A), as well as soil type (rhizosphere or bulk-control; ^***P < 0.001; R² = 0.04486), field block (^**P < 0.006; R² = 0.00766), and Soil Type: Collection Date (^**P < 0.002; R² = 0.01342). The low R² effect size suggests that the nucleic acid of origin has the most significant impact on microbial community composition, though even that effect is small. Other factors, such as the spatial scale of the field block or stochastic differences of collection day, had minimal to no effects.

Differential abundance analysis via corncob identified 12 phyla (202 genera) that differed significantly (p < 0.01) between DNA- and RNA-derived communities. Of these, 108 were more abundant in the RNA-derived community compared to the DNA-derived community (Figure 5A), indicating a significantly higher PSP for the number of cells present.

Using rarefied communities, alpha diversity differed significantly between the rhizosphere and control soils in the DNA-derived microbial community in the observed richness (DNA Control PM to DNA Rhizosphere PM p < 0.05; DNA Control AM to DNA Rhizosphere AM p < 0.01), Shannon (DNA Control PM to DNA Rhizosphere PM p < 0.01; DNA Control AM to DNA Rhizosphere AM p < 0.0001), and Simpsons (DNA Control PM to DNA Rhizosphere PM p < 0.001; DNA Control AM to DNA Rhizosphere AM p < 0.0001) diversity indices. No significant differences in alpha diversity were observed between the different soil types in the RNA-derived communities, except in Shannon diversity between the RNA rhizosphere and control soils collected at 3 AM (p < 0.5; Figure 2), indicating that while microbial membership differed within samples, the diversity of what bacteria were contributing to the PSP profile within a sample did not.

Separately analyzing the RNA- and DNA-derived communities from one another, rhizosphere and control soils differed significantly from each other (^*p < 0.05) for 7 (Acidobacteria, Bacteroidetes, Chlorobi, Deinococcus-Thermus, Elusimicrobia, Fibrobacteres, Verrucomicrobia) of the 18 total phyla identified in this study. Three unique phyla (Actinobacteria, Armatimonadetes, Planctomycetes) showed significant (^*p < 0.05) differences only in the DNA-derived rhizosphere and control soil microbial communities, while three unique phyla (Chlamydiae, Firmicutes, Gemmatimonadetes) showed significant (^*p < 0.05) differences only in the RNA-derived microbial communities (Figure 3). This indicates the differences in microbial membership and microbial contribution to community PSP profile differ according to phyla within the rhizosphere and control soils. Using the model OTU abundance (within RNA- or DNA-derived community) = Soil Type ^* Collection Date ^* Field Block showed significant differences in community composition within the RNA- and DNA-derived samples. In the RNA (Figure 4B), soil type (P < 0.001^***; R² = 0.05962), field block (P < 0.011^**; R² = 0.01990), and soil type: collection date (P < 0.037^*; R² = 0.03288) were significant. In the DNA (Figure 4C) only the soil type was significant (P < 0.001^***; R² = 0.094). However, the low R² effect size suggests a very small effect of these factors on microbial community composition.

We used the corncob Bayesian analysis to identify individual taxa that differed between the rhizosphere and control soils. In the DNA-derived community, 10 phyla (114 genera) were significantly different (p < 0.01) between control and rhizosphere soils. Of these, 35 genera were more abundant in the rhizosphere soil compared to the control soil community (–log10 p-value above zero; Figure 5B). In the RNA-derived community, corncob differential abundance analysis identified 10 phyla (96 genera) that differed significantly (P < 0.01) between control and rhizosphere soils. Of these, 39 genera were more abundant in the rhizosphere soil compared to the control soil community (Figure 5B). Between the DNA- and RNA-derived communities, 58 genera (7 phyla) overlapped, with 23 genera being above zero and 34 below zero. In the DNA-derived community, 57 genera were unique, with 12 having more than zero and 45 having fewer than zero. In the RNA-derived community, 39 genera were unique to the corncob analysis, with 16 genera having a positive value and 23 having a negative value.

Using rarefied communities, alpha diversity did not differ significantly between the two collection timepoints. Visualization of the relative abundance of phyla between the AM and PM collection timepoints showed no significant differences, except in two phyla between the DNA-derived control soils (Deinococcus-Thermus, ^*p < 0.05), where the AM timepoint was greater, and the RNA-derived control soils (Chlamydiae, ^*p < 0.05), where the AM timepoint was greater (Figure 3).

To remove the confounding effect of collection date, samples were separated into three groups corresponding to the date of collection, in addition to being separated by their nucleic acid type, and the model OTU abundance (within RNA- or DNA-derived community) = Soil Type ^* Collection Time ^* Field Block was used to calculate significance. Community composition was significantly different between soil type (^**P < 0.004; R² = 0.06336) and field block (^**P < 0.008; R² = 0.0542) on day one and soil type only on day 2 (^***P < 0.001; R² = 0.08644) and day 3 (^***P < 0.001; R² = 0.11179). In the DNA community, soil type (^***P < 0.001; R² = 0.09588) and field block: sample type (^*P < 0.021; R² = 0.05150) on day one, and soil type only on day 2 (^***P < 0.001; R² = 0.11893) and day 3 (^***P < 0.001; R² = 0.13241). The collection timepoint was never significantly different between the RNA- nor DNA-derived microbial communities (Figures 4B, C). However, the low R² effect size suggests a very small effect of these factors on microbial community composition, though the effect of collection day varied from a very small to small effect.

Samples were pooled once more by collection date, then separated based on soil type and RNA- and DNA-derived community to calculate significance using the model OTU abundance (within RNA- or DNA-derived community) = ~Collection Time. Collection Time was not significant in either rhizosphere (RNA: P = 0.04; R² = 0.0085; DNA: P = 0.04; R² = 0.0088) or in control soils. Here again, the R² effect size is very small, indicating that this result is insignificant.

Corncob differential abundance analysis identified five genera that differed significantly (^*P < 0.05) between the 3 AM and 2 PM collection timepoints of RNA-derived rhizosphere soil microbial communities (Figure 5C). No genera were significantly (^*P < 0.05) different between the RNA-derived control soils, or the DNA-derived rhizosphere or control soils. Observing the absolute abundance of reads of the genera determined to be significantly different from the corncob analysis shows only one genus, Curtobacterium, to be significantly (^*P < 0.05) different between collection timepoints (Figure 6A), but the amplitude of the signal was not significant when collection time was separated within collection days (Figure 6B).