<b>The nature and extent of freeze/thaw events influences trace gas emissions from U.S. Corn Belt soils - Experiment 1</b>

Soil freezing and thawing events have been shown to be important drivers of episodes of high emissions of greenhouse gases (CO₂, N₂O, CH₄) from soil to the atmosphere. Here we show that these events also stimulate emissions of NO, a precursor of atmospheric pollution (ozone, particulate matter). The intensity of these episodes varies with the nature and extent of the freezing and thawing event. There is a clear need for improved measurement and modeling of freeze/thaw events to account for winter fluxes of trace gases and assessment of how they are changing with climate.<b>Methods</b>Soils were sampled in November 2022 with a hand spade to a depth of approximately 25 cm from approximately 20 locations along transects along the north and south sides of a field cropped to maize. Soils were not excessively wet (no recent irrigation) or in a drought condition at sampling. Soils were shipped to the Cary Institute of Ecosystem Studies in Millbrook, NY USA where they were packed into PVC cores. Sufficient soil to pack 45 cores to a depth of 12 cm and a bulk density of ~1.2 g/cm³ was placed into a large bin. Soil was homogenized by hand, large clods were manually broken up, and large roots and rocks were removed. Soil was weighed out for each core, then poured, 1/5 at a time, into the PVC core. As each section of soil was added it was compacted with a section of smaller diameter PVC which was capped with a rubber stopper. Each core was labelled with its assigned treatment. A sample was taken from the bulk soil for initial moisture content. Soil moisture was also measured after each destructive sampling event (described below).Replicate cores (enough to allow for destructive sampling at specific intervals), were assigned to different freezing temperatures and freezing degree days (FDD) treatments for 15 days at a) -1 ^oC (15FDD) (five cores), b) -3 ^oC (45FDD) (five cores), and c) -5 ^oC (75FDD) (fifteen cores), after which the cores were thawed to 5 ^oC for 14 days. A set of replicate cores was kept at -3 ^oC and thawed after 5 days (15FDDFAST) (five cores). Control cores were maintained at 5 ^oC (fifteen cores). Treatments were applied by placing cores in four laboratory incubators (Percival Scientific model E36VL) set at different temperatures (-1 ^oC , -3 ^oC, -5 ^oC, +5 ^oC). Air, but not soil temperatures in the incubators were monitored continuously. To ensure that soil frost started at the surface and then extended to deeper soil as occurs in the field, cores were housed in fitted polystyrene blocks. Visual observations confirmed that soil cores froze from top to bottom. After treatment, cores were thawed to 5 ^oC by increasing temperatures by 0.5 C hr^-1. Cores were also frozen gradually, 0.5 ^oC/hr, with the exception of the 15FDDfast treatment, which was immediately placed in a -3 ^oC incubator.Gas flux measurements were made after 0, 3, 6, 9, and 14 days. The number of replicate core flux measurements varied due to destructive sampling (described below) of one core from each treatment after 0 and two cores each after 6 and 14 days. The number of replicate soil core gas flux measurements was 5 on day 1, 4 on days 4 and 7, and 2 on days 11 and 14. To measure gas fluxes, cores were removed from the incubator and from the polystyrene block and placed inside a quart mason jar and then capped with a lid containing a butyl septum. Immediately after the jars were capped, 3 replicate samples were taken from the ambient air inside each incubator to act as time zero samples for all of the cores in each treatment. These time zero samples for CO₂, CH₄ and N₂O were taken with a 10 ml syringe and consisted of 8 ml of ambient air injected into evacuated 6 ml Labco exetainer vials. The time zero samples for NO were taken with a 60 ml syringe fitted with a 3-way stopcock. The jars were then sampled at ~30 minutes and ~ 60 minutes after the time zero samples were taken. To do this, first 8 ml of headspace was pulled from each jar and injected into an evacuated 6 ml exetainer for later analysis of CO₂, CH₄ and N₂O. After this sample was taken 60 ml of He was injected into the jar and mixed with the headspace by pumping the syringe a few times. Then the 60 ml syringe was filled with this mixed headspace, the stopcock closed, and the sample was set aside for analysis. The mason jars remained in their respective incubators for the duration of the gas sampling period. All NO samples were analyzed immediately (within 15 minutes) by injecting directly from the 60 ml syringe into a Sievers Nitric Oxide Analyzer with a chemiluminescence detector. The CO₂, N₂O and CH₄ samples were stored at ambient temperatures until analysis on a Shimadzu GC2014.One core for each treatment was destructively sampled after 0, 6, and 14 days. These measurements include inorganic N (nitrate (NO₃^-) and ammonium (NH₄⁺)), potential net N mineralization and nitrification rates, microbial biomass C and N content using the chloroform fumigation-incubation method (Jenkinson and Powlson 1976) and denitrification enzyme activity using the acetylene inhibition method (Smith and Tiedje 1979) as described by Groffman et al. (1999). Soil moisture was determined by drying at 105 ^oC and concentrations of NO₃^- and NH₄⁺ were determined colorimetrically on a Lachat QuikChem autoanalyzer (Loveland, CO) following extraction in 2M KCl solution. Potential net N mineralization and net nitrification were determined as the difference in starting and ending total inorganic N (mineralization) or NO₃^- (nitrification) concentrations in a 10-day incubation at laboratory temperature.One core was destructively sampled after 6 days of incubation for measurement of N₂, N₂O and CO₂ flux from an intact core by nitrogen free atmosphere recirculation method (NFARM) as described in Burgin and Groffman (2012). Briefly, this method allows for an intact core to be incubated in a helium/oxygen headspace. The headspace can then be sampled at intervals for N₂, N₂O and CO₂, and a flux rate calculated. Results were analyzed with analysis of variance (ANOVA) with time and freezing treatment as main effects, with interactions. Duncan’s multiple range test was used to detect differences between specific treatments over all sample dates and on a date-by-date basis.<b> </b><br>