Does Management Intensive Grazing Protect Groundwater by Denitrification?

2003 Annual Report for LNC01-181

Project Type: Research and Education
Funds awarded in 2001: $90,340.00
Projected End Date: 12/31/2004
Matching Federal Funds: $19,200.00
Matching Non-Federal Funds: $17,813.00
Region: North Central
State: Wisconsin
Project Coordinator:
Nancy Turyk
University of Wisconsin Stevens Point

Does Management Intensive Grazing Protect Groundwater by Denitrification?

Summary

Denitrification is being assessed at four farms in Wisconsin: three management intensive grazing farms and one conventional farm. Assessment is being accomplished by collecting groundwater from monitoring wells, sampling soil, and measuring soil and groundwater gases.

Preliminary data suggest that denitrification differs between the MIG and conventional sites, with the MIG site showing less total dissolved nitrogen, nitrate, and nitrous oxide then the conventional site. Dissolved carbon dioxide and organic carbon was higher in groundwater at the MIG site, suggesting higher microbial respiration in groundwater and greater source of electron donor for the denitrification process. Nitrate groundwater concentrations were greater at the conventional site and soluble reactive phosphorus was also elevated compared to the grazing sites.

Objectives/Performance Targets

The primary objective is to determine whether denitrification is higher in soil and groundwater under MIG than annual cropping. We are focusing on coarse- and medium-textured soils, where nitrate loading potential is higher than fine-textured soils.

Accomplishments/Milestones

The sites that were selected in year one of this study were sampled in year two. They include four farms in Wisconsin. The management intensive grazing (MIG) farms include the Bestul Farm in Waupaca County, the Onan Farm in Portage County, and the Brenneman farm in Sauk County. The conventional cropped field (currently in corn) is near the Bestul Farm in Waupaca County and is being farmed by the Rambos. During the first year of study, groundwater monitoring wells were installed near the paddocks/field that did not already have them and groundwater and soil samples were collected for analyses. The soil/groundwater gas component of the study was conducted at the Rambo and Bestul farms.

This denitrification study was discussed at the Onan and Bestul farms at pasture walks in June and July. Participants included farmers and agency personnel, and a group of 25 individuals from around the world who were participating in an International Watershed Management Seminar through UWSP. Presentations were made to professional audiences at the Wisconsin Chapter of the American Water Resource Association Annual Meeting, American Society of Agronomy meetings, Denver, CO, and the Minnesota-Wisconsin Nutrient Management Research and Planning Workshop for Grazing Systems, River Falls, WI.

All soil and gas samples for the study have been collected, and most lab analyses have been conducted. One more round of samples from the groundwater monitoring wells will be collected in late winter/early spring; all other lab analyses have been completed.
Following are specific details regarding the work that has been conducted over the last two years in the three primary study components.

Soils

Field sampling. Analysis of samples collected in 2002 under apparent urine spots, dung pats, and background areas of three grazing farms and from a cornfield on the fourth farm showed no discernable differences among treatments. This result is likely due to the high horizontal and vertical variability present within pasture soils and our inability to correctly identify urine spots and background areas. In a short-term incubation trial, we found that both dung and urine have high concentrations of DOC and that urine increased DOC in soil 5- to 10-fold over at least a 3-week period, supporting our hypothesis that excreta may increase the supply of the two substrates (N and C) needed for denitrification.

In 2003, we discontinued frequent sampling of the fields, and took only one set of samples in late autumn. These samples have not yet been analyzed. In place of the planned field samples, we decided to investigate the effect of dung and urine deposition under controlled conditions.

Intact core study. This experiment was designed to obtain more reliable information on the formation of nitrate and presence of DOC in pasture soil affected by dung and urine deposition. Two sets of intact 6-cm diameter cores about 60 cm deep were obtained from Bestul Farm in June and November 2002 (to be used in a follow-up experiment). All cores are encased in plastic sleeves. After several months of equilibration to reduce variations in N supply among the June cores, the lower 10 to 20 cm was saturated to mimic a high water table, and cores received typical rates of fresh dairy cow urine or dung. A subset of cores received no treatment. Four replicates were sealed to allow ammonia collection from the headspace by continuous air flow (about 100 mL/min) through phosphoric acid traps. We simulated a 2 cm rainfall four days after application. At intervals, cores were separated into three depth increments (0 to 15 cm, 15 cm to the water table, and the saturated zone). Soil water content, bulk density, inorganic and total N, DOC, and pH are being determined on each increment. The headspace has been sampled occasionally for nitrous oxide.

The first experiment is underway at the time of writing and we have only a few early results to report. In the first few days after addition, soil inorganic N concentrations were slightly elevated by dung, but increased by urine to over 300 ppm ammonium in the topsoil, 180 ppm in the intermediate zone, and 20 ppm in the saturated zone. Little net nitrate production was apparent in the first few days after excreta addition. Soil pH was not affected by dung addition, but increased by more than 1 pH unit in the non-saturated soil and by 0.5 unit in the saturated soil after urine treatment. By 26 days after urine application, pH had declined to the background level (6.9) in the upper 15 cm. These responses are similar to those reported elsewhere, but of lesser magnitude. In poorly buffered topsoils, pH can rise by 3 units. Surface pH > 7.5 increases ammonia volatilization loss. Ammonia losses were very small (about 3 g/ha/d) from untreated soil, but up to 0.25 kg/ha/d from dung and 1.5 kg/ha/d from urine during the first four days after treatment. No other analyses are available at this time, but will be reported at a later date. The preliminary early results appear to meet our expectations, based on the literature. Data on DOC and denitrification are the keys to testing our hypothesis.

Soil and Groundwater Gases

We completed a second field season at Bestul’s (MIG site) and Rambo’s (conventional cropping) farms.

1. Miniature soil gas and groundwater wells at the Bestul Farm (MIG site) lost to attrition over the winter were replaced and resurveyed in the spring, reestablishing 91 grid sampling nodes on 50 x 100 ft spacing.

2. A completely new grid network of miniature wells was installed and surveyed and the Rambo farm (n=35, conventional corn cropping site) to replace wells lost during cultivation and overwintering. The wells were installed in the same pattern (100 ft x 100 ft grid spacing) and approximate locations as the first year’s grid to maximize consistency with the first field season.

3. Ground water and soil gas samples were collected from the grid network of wells at the Bestul and Rambo farms during June, July, and August.

4. Field measurements included depth to water, pH, specific conductance, temperature, oxidation reduction potential, temperature, and total dissolved gas pressure. Soil gas samples were collected into septum sealed evacuated bottles and werek overpressurized for gas chromatography in the lab. Dissolved gases were collected in the field from groundwater using pumping induced ebullition. Water samples for dissolved solids measurements were filtered and preserved in the field. Most laboratory analyses have been completed. Gas samples were analyzed for nitrous oxide, carbon dioxide, methane, nitrogen, argon, and oxygen by gas chromatography. Water samples were analyzed for nutrients and major ions. Metal analyses remain to be completed.

5. Static chambers were used to measure nitrous oxide fluxes to the atmosphere at each selected grid node at each farm.

Groundwater Monitoring Wells

Samples from the monitoring wells were collected five times in between June and September 2003. The samples were analyzed for NO2+NO3-N, chloride during all sample periods and once for NH4, SRP, temperature, pH, conductivity, alkalinity, total hardness, and dissolved oxygen. One additional round of samples will be collected during late winter 2004.

A pressure transducer was installed in a shallow well at the Brenneman site to measure daily fluctuations in groundwater elevations.

Impacts and Contributions/Outcomes

Soil and Groundwater Gasses

Data analysis is ongoing. However, several major differences between MIG and conventional agriculture are evident in the data at this point. Consistent with the first year’s data, the following data highlights indicate that denitrification improves groundwater quality to a much greater extent under MIG than annual cropping:

Total dissolved nitrogen, nitrate and nitrous oxide were seven to 10 times higher in concentration in shallow groundwater at the conventional site than at the MIG site.

Groundwater was generally oxygenated at both study sites. However the incidence rate of anaerobic grid nodes, conducive to denitrification, was higher at the MIG site than at the conventional site. Overall dissolved oxygen was slightly lower at the MIG site than at the conventional study site, suggesting greater microbial respiration in the soil and groundwater there than at the conventional site.

Dissolved carbon dioxide was higher in groundwater at the MIG site than at the conventional site, again suggesting higher microbial respiration in soil and groundwater at the MIG site than at the conventional site.

Dissolved organic carbon was much higher in groundwater at the MIG site than at the conventional site suggesting a greater supply of electron donor for denitrification in groundwater.

Denitrified N (excess N2-N) was similar to but slightly higher in groundwater at the MIG site than the conventional site. The concentration of denitrified N appeared to be strongly a function of recharge temperature at the MIG site but unrelated to temperature at the conventional site, suggesting different processes control denitrification at the two sites.

Denitrified N as a percentage of total nitrate (NO3-N + excess N2-N) was about tenfold greater at the MIG site than the conventional site (based on the median denitrification rates of about 60% of total nitrate at the MIG site and 10% of the total nitrate at the conventional site).

Nitrous oxide flux measurements have not yet been analyzed.

Groundwater Monitoring Wells

Significantly more statistical analysis will be conducted on the groundwater samples following the final sampling round. However, preliminary data analysis has been conducted and some observations can be made. Average and maximum concentrations of NO2+NO3-N and SRP were higher in the groundwater acquired at the conventional site R (Table 1). Although Site O is a grazing site, concentrations of NO2+NO3-N were most similar to those measured at site R. This may be a function of a greater depth to groundwater (poorer denitrification potential), different soil type, groundwater mixing with upgradient groundwater that originates in a cornfield, or a combination of these factors. Concentrations of NO2+NO3-N were similar in the groundwater at the other grazing sites BE, BR, BW (Figure 1). Overall, ammonium concentrations were low. SRP concentrations were similar in groundwater sampled at the grazing sites compared to the conventional site (Figure 2). Concentrations of SRP vary throughout the year and were greatest in the samples obtained in June.

Table 1. Average and maximum concentrations in groundwater samples obtained at each study site.
Site

NO2+NO3-N Cl SRP NH4
mg/L mg/L mg/L mg/L

Conventional
R Average 18.4 10 .086 0.10
R Max 41.4 26 .388 0.26

Grazing
O Average 15.4 22 .018 0.23
O Max 30.2 71 .049 0.44

BE Average 2.2 6 .007 0.04
BE Max 5.6 25 .028 0.19

BW Average 2.5 2 .010 0.20
BW Max 8.6 22 .030 1.31

BR Average 4.5 6 .025 0.23
BR Max 20.9 24 .071 0.79

Figure 1. Distribution of NO2+NO3-N (mg/L) in samples collected in the upper 18 inches of aquifer at each study site.
Figure 1. Distribution of SRP (mg/L) in samples collected in the upper 18 inches of aquifer at each study site.

Collaborators:

Michael Russelle

USDA ARS St. Paul, MN
MN 55108
Bryant Browne

bbrowne@uwsp.edu
Assoc. Professor of Soil and Water Resources
UW-Stevens Point
College of Natural Resources - 276
Stevens Point, WI 54481
Office Phone: 7153464190