Lignin PhenolsData file with results from lignin phenol extraction of soil samples, and plant tissue end-members (i.e. shoots, rhizomes and roots of Bahiagrass from inside and outside grazing exclosure). The major families are summarized as v (vannilyl), c (cinnamyl) and s (syringyl). Column 'vsc' represents the sum of v,s, and c, standardized per 100 mg organic carbon, whereas 'vsc.sed' is standardized per 1 g of soil weight. 'adal.v' is the acid-aldehyde ratio of the vanillyl family.gcb_lignin.csvSoils EA/IRMSFile with results of elemental analysis and isotope ratio mass spectrometry. Note that plot number needs to be paired with treatment ('trt') to generate a unique ID. 'Back' column identifies whether data is from background survey of exclosures, or from pulse-chase plots.Soils_EAIRMS.csvPulse Chase Vegetation EA/IRMSData with results of EA/IRMS analysis on plant samples from pulse chase experiment. Note that plot number and treatment must be combined to generate a unique plot ID. Harvest identifies time post pulse (2 days, 7 days, or 32 days). Pool identifies whether it is shoots ('Ag Veg'), roots or rhizomes ('Rh'). 'Sub' represents replicated harvested swaths ('A' or 'B') per harvest date (for 2 day and 7 day only, hence 32 day is identified as AA). One sample was accidentally combined in the field and is identified with A/B. Biomass is reported only for the t = 0 initial harvests of aboveground material.PulseChase_VegDataALL.csvPulse Chase Vegetation MassContains dry weight biomass for all plant samples collected in pulse chase experiment. Note that treatment and plot number must be combined to generate unique plot ID. 'Sub' identifies replicate harvest swath collected at each harvest date (2 day and 7 day only, 32 day only had one swath and is thus identified as AA, as is the t=0 harvest of aboveground tissue immediately post pulse). 'Harvest' identifies time of harvest post pulse and is either 0 (immediately post pulse, aboveground tissue only), '2d' (2 day), '7d' (7 day), or '32day' (32 days). 'Pool' identifies the plant tissue and is either shoots ('Ag Veg'), roots ('Root') or rhizomes ('Rh'). Biomass is in grams. 'Standing dead' represents the senesced tissue sorted out of the sample prior to analysis, and is also in grams.PlantData_MassALL.csvPulse Chase Microbial DataMicrobial biomass and isotope ratio data. Note that treatment and plot number must be combined to generate a unique sample ID. Harvest indicates time since post pulse: '48' represents 2 days, '336' represents 7 days, and 4 represents 32 days. 'Sub' represents replicated harvest swath within a given harvest date (2 day and 7 day only). Column 'Rep' can be ignored. 'mgC/dry_mass_soil(g)' represents the carbon concentration of the extract, standardized per gram of soil extracted. 'd13c (permil, versus VPDB)' is the standard isotope delta 13C value, and 'Fumigated' identifies fumigated extracts (containing lysed microbial cell contents in addition to dissolved organic carbon, DOC) and 'unfumigated' extracts (which contain just DOC). Formulae for calculating microbial biomass and isotope enrichment are in the main paper, and are also explained in the open source code used to process and analyze data available at https://github.com/chwilson/GCB_2018).PulseChase_MicrobialData.csv

Despite the large contribution of rangeland and pasture to global soil organic carbon (SOC) stocks, there is considerable uncertainty about the impact of large herbivore grazing on SOC, especially for understudied subtropical grazing lands. It is well known that root system inputs are the source of most grassland SOC, but the impact of grazing on partitioning of carbon allocation to root tissue production compared to fine root exudation is unclear. Given that different forms of root C have differing implications for SOC synthesis and decomposition, this represents a significant gap in knowledge. Root exudates should contribute to SOC primarily after microbial assimilation, and thus promote microbial contributions to SOC based on stabilization of microbial necromass, whereas root litter deposition contributes directly as plant-derived SOC following microbial decomposition. Here we used in situ isotope pulse-chase methodology paired with plant and soil sampling to link plant carbon allocation patterns with SOC pools in replicated long-term grazing exclosures in subtropical pasture in Florida, USA. We quantified allocation of carbon to root tissue and measured root exudation across grazed and ungrazed plots and quantified lignin phenols to assess the relative contribution of microbial versus plant products to total SOC. We found that grazing exclusion was associated with dramatically less overall belowground allocation, with lower root biomass, fine root exudates, and microbial biomass. Concurrently, grazed pasture contained greater total SOC, and a larger fraction of SOC that originated from plant tissue deposition, suggesting that larger root litter deposition under grazing promotes greater SOC. We conclude that grazing effects on SOC depend on root system biomass, a pattern that may generalize to other C4-dominated grasslands, especially in the subtropics. Improved understanding of ecological factors underlying root system biomass may be the key to forecasting SOC and optimizing grazing management to enhance SOC accumulation.