Microcosm Construction
Microcosms consisted of 7 g sterile sand, 0.1 g (dry weight) of 1 cm cut blue grama litter. Sand was sterilized by autoclaving 3 times for 1 hour with at least 12 hours between cycles. Microbial communities were chosen based on their functional performance in previous experiments measuring DOC concentrations from microbial decomposition of blue grama litter. We extracted microbial communities from soil samples (n=12) on the day of inoculation by suspending 2 g of soil into autoclaved milliQ water then transferring 5ml of soil dilution to either 5ml of filter-sterilized NH4NO3 solution with a concentration of 3.85 mg/ml or 5ml of autoclaved milliQ water to create a 1:20 dilution. For negative controls, autoclaved milliQ water or sterilized NH4NO3 solution was added to litter without a soil inoculum. All microbial communities had 12 replicates with 6 having the high nitrogen background to test the effect of inorganic nitrogen availability on carbon cycling. Microcosms were sealed with Teflon-lined crimp caps and incubated at 25ºC in the dark for 48 days with periodic replacement of air in the headspace.
Respiration Measurements
On days 3, 8, 14, 21, 28, 35, 42, and 48, respiration was measured by gas chromatography using an Agilent Technologies 490 Micro GC (Santa Clara, CA, United States). After each measurement, the headspace air was evacuated with a vacuum pump and replaced with sterile-filtered (0.2µm Millpore filters) air.
Dissolved Organic Carbon and Dissolved Total Nitrogen
To test the effect of nitrogen deposition and precipitation frequency on carbon cycling, 3 microcosms from each microbial community with nitrogen addition and 3 without nitrogen addition had DOC removed weekly while the other half only had DOC removed at the end of the experiment (day 48). When DOC was removed weekly, we added 10 ml of sterilized milliQ water, mixed by gently swirling, then collected the water with sterile pipets. The collected liquid was then filter-sterilized using a 0.2 µm filter. Filtered extracts were analyzed for non-purgeable organic carbon and total nitrogen (TN) using a Shimadzu TOC-LCSH/TNM-L Carbon/Nitrogen Analyzer (Shimadzu Scientific Instruments, Inc. Columbia, MD) employing a catalytic combustion oxidation method and non-dispersive infrared detector (NDIR) for carbon and chemiluminescent detector for nitrogen. Samples were acidified with 5N HCL and purged with air to remove inorganic carbon. A sample aliquot was then injected onto a palladium catalytic bed held at 720 °C. The CO2 and NO combustion products were then measured serially with the two detectors. The instrument was calibrated daily using a NIST-traceable standard of potassium hydrogen phthalate and potassium nitrate. Reagent water blanks were run periodically along with the check standards to monitor carry-over and instrument drift. During methods development a representative set of samples was analyzed by the method of standard additions to verify complete recovery of organic carbon and total nitrogen under the analysis conditions.
Destructive Sampling
On the final day (day 48) of the experiment after all the respiration measurements were taken, microcosms were divided into three subsamples. The procedure included weighing each tin dish prior to splitting the sample. Dividing the cut litter between each of the three subsamples to ensure all samples had equal plant litter, then adding sand until the entire sample was split into 3 subsamples. Then, each tin dish was measured and the weight recorded again. One subsample was used to determine moisture content. The other two subsamples were used for microbial biomass measurement.
Moisture Measurement
For each microcosm, the subsample for moisture measurement was placed in a drying oven at 105°C. Samples were dried overnight (14 hours). The samples were then cooled to room temperature and reweighed. The dry weights were recorded and the moisture content (%) was determined using this equation: ((wet weight (g) – dry weight (g))/ dry weight (g)) * 100.
Microbial Biomass Carbon and Nitrogen
For each microcosm, the other two subsamples of the sand-litter mixture were used to determine microbial biomass carbon and nitrogen. Both subsamples were air-dried for 24 hours. One sample was fumigated as described previously (Allison et al. 2008) with some alterations. The subsample for fumigation was placed in a desiccator with a 50 ml beaker containing 20 ml of chloroform and boiling chips. The desiccator was evacuated four times until the chloroform boiled and then vented, except the desiccator was not vented the final time. The desiccator and samples incubated in the dark for 24 hours to ensure complete fumigation of the samples. After 24 hours, the vacuum was released and the chloroform beaker removed. Any excess chloroform fumes were removed through vacuum and venting prior to sample removal. After removing chloroform vapor, we extracted the DOC from the subsample with 10 ml of 0.5M K2SO4 by inverting 5 times and vortexing for 30 seconds at full speed. The supernatant was filtered through a 0.2 µm syringe filter and stored at -20ºC until analyzed for DOC and TN concentrations as described above. The non-fumigated subsample was processed like above except no chloroform fumigation occurred. Specifically, non-fumigated samples were air dried, DOC was extracted with 10 ml of 0.5M K2SO4, filtered, and stored at -20ºC.
Carbon Use Efficiency and Microbial Metabolic Quotient
Carbon use efficiency was calculated using the microbial biomass measured by chloroform fumigation method (above) and the cumulative respiration measured throughout the experiment using gas chromatography and using the following equation:
The microbial metabolic quotient calculation was determined using the following equation:
Microbial Community Sequencing and Analyses
A smaller microcosm experiment using a subset of soil communities (n=4) used in this experiment that were previously characterized as producers of low or high DOC concentrations from litter decomposition (2 each) were set up as described above except all of these received the nitrogen treatment. At the end of the experiment, the samples were analyzed for microbial community composition using amplicon sequencing described in detail previously (Kroeger et al. 2021). Data was rarefied once to total OTU counts of 1562 for bacteria, and 1776 for fungi using the package vegan::rarefy (Oksanen et al. 2020). Samples S39.DOC.5 (8 total OTUs), S49.DOC.5 (628 total OTUs), S49.Norm.4 (280 total OTUs), and S17.Norm.5 (208 total OTUs) for bacteria were not analyzed because they had a total OTU count less than 1000; all fungal sequencing samples were kept. Data was then log-transformed using the package vegan::decostand. For Figure 7, NMDS distances were calculated using the package vegan::metaMDS, based on Bray-Curtis dissimilarity. 95% confidence ellipses were calculated using vegan::ordiellipse on the calculated NMDS distances. For Figure 8, indicator species analysis was performed using the indicspecies::multipatt package (De Cáceres and Legendre 2009; De Cáceres et al. 2010). Only species with a P value less than or equal to 0.05 were kept. All sequences are deposited in the NCBI Sequence Read Archive under Bioproject ID PRJNA851447.
Statistics
All statistical analyses were conducted in RStudio (v1.4.1717) with R (v4.1.1) (RStudio Team 2020; R Core Team 2021). All metrics were checked for normality using the Shapiro-Wilk test. Since the data were not normal, Kruskal-Wallis tests were used for comparing the medians between treatment groups. The Ansari-Bradley test was used to compare variance between treatments. Spearman correlations were used to assess the relationship between measured metrics. Figures were made using packages ggplot2 and ggpubr (Wickham. 2016; Kassambara 2020).