A Protocol for Collecting and Constructing Soil Core Lysimeters

The Delmarva Peninsula borders the eastern shore of the Chesapeake Bay, and is the home to one of largest poultry production regions in the US. Roughly 600 million chickens and an estimated 750,000 tons of manure are generated from the production of these birds each year¹. Most of the manure is used locally as a fertilizer amendment on agricultural fields. Because of historically high rates of manure application, nutrients such as nitrogen and phosphorus have accumulated in the soil and are now susceptible to off-site losses via subsurface leaching². Much of the groundwater flow is directed toward an extensive network of ditches that ultimately drain to the Chesapeake Bay³. The nutrients carried to the Bay are linked to the decline in the Bay's health due to eutrophication⁴.

Connecting nutrient management with off-site losses of nutrients requires specialized tools to monitor hydrologic flows and associated nutrient transfers. Lysimeters represent a major category of instruments used to characterize and quantify the movement of nutrients through soils. Lysimeters have a long history of use in monitoring nutrient flow in percolating water^5-7, from tension lysimeters that can be adjusted to counter soil matrix potential so that they better estimate plant available water, to zero-tension lysimeters that are more representative of processes occurring during free drainage. All approaches to lysimetery present inherent biases. For instance, some lysimeters are too small to fully represent spatially complex processes in natural soils, or are too large and expensive to provide good statistical replication of heterogeneous soils⁸. Further, pan lysimeters require soils above them to be saturated to collect leachate and are inefficient compared to tension lysimeters at measuring matrix flow⁹.

Closed lysimeter systems, such as zero-tension soil core lysimeters (also known as soil monolith lysimeters), greatly improve the confidence with which water budgets and associated pollutant budgets (e.g., nutrient budgets) are carried out¹⁰. These lysimeters are most representative when they contain intact cores of soil; lysimeters filled with repacked soils do not maintain the original structure, horizons and macropore connections that influence the transport of solutes and particulate compounds alike^11,12. From an experimental stand point, approaches that facilitate greater replication of undisturbed soil conditions are advantageous, given the inherent spatial variability that exists in soil physical and chemical properties¹³.

Two preferred methods have been used for collecting intact soil core lysimeters: drop hammer and cutting head. The former has been more commonly performed, as it can be achieved with devices as simple as a sledge hammer (smaller lysimeters). When performed properly, soil core collection with a drop hammer has been shown to be relatively cost effective, especially when compared with other coring techniques. However, the sheer forces imposed by driving a lysimeter casing into the ground can cause smearing and compaction, producing conditions inside the lysimeter that are not representative of native soil and may even favor certain types of water movement (e.g., bypass flow, or flow along the soil core edge). As a result, some researchers have preferred the use of corers that cut away an intact soil with a drilling apparatus or other excavation device⁵.

Various materials have been used as casings for soil core lysimeters. Steel pipes and boxes are comparatively low cost, durable and readily available and can be used to collect larger lysimeters due to their strength^14-17. However, while steel is satisfactory for evaluating the leaching of relatively unreactive compounds such as nitrate, the iron in steel reacts with phosphate and must therefore be coated or otherwise treated for the study of phosphorus leaching. Commonly, plastic casings are used to study phosphorus leaching, such as thick walled (Schedule 80) PVC pipe that can withstand the impact of a drop hammer (if used) and retain its structure when larger diameter soil cores are obtained (e.g., ≥30 cm)^18-22.

In general, soil core lysimeters are analyzed ex situ. Once collected, soil core lysimeters may be installed in outdoor "lysimeter farms" where surrounding soil and above ground climates represent natural field conditions. For instance, in Sweden, the Swedish Agricultural University has maintained three separate lysimeter farms over the past three decades, analyzing pesticide fate-and-transport, long-term soil fertility trials, and management practices that can be scaled to 30 cm diameter intact cores²³. Soil core lysimeters have also been subjected to indoor leaching experiments where there is greater control of climatic conditions^24,25. Liu et al. used a rainfall simulator to regularly irrigate soil core lysimeters under an array of catch crops²⁶. Kibet and Kun all employed hand irrigation techniques to study arsenic and nutrient leaching through soil cores^27,28.

A variety of edaphic and hydrologic processes can be inferred from soil core lysimeters. Kun et al. (2015) used 30 cm diameter PVC column lysimeters to investigate nitrogen leaching after urea application²⁸. By collecting leachate at different time intervals following an irrigation event, they were able to differentiate between rapid and gradual flows, with the former assumed to be dominated by macropore flow, and the later assumed to be dominated by matrix flow. Since urea is readily hydrolyzed when in contact with soil, they interpreted the presence of elevated urea concentrations in leachate collected shortly after urea application as evidence of macropore transport that bypassed the soil matrix. Over time, they detected elevated concentrations of different forms of nitrogen in leachate, tracking the transformation of applied urea to ammonium after initial hydrolysis, then the transformation of ammonium to nitrate with nitrification.

To illustrate considerations in designing, conducting and interpreting soil core lysimeter experiments, we carried out an investigation of four different soils found in the mid-Atlantic coastal plain of USA. The study measured leaching concentration and loss of nitrate before and after application of dry poultry manure (i.e., poultry "litter")²⁸. Nutrient losses from the application of poultry litter to soils are a key concern to the health of the Chesapeake Bay, and understanding the interaction of applied poultry litter and agricultural soil properties is needed to improve nutrient management recommendations. We present here a detailed method for extracting intact soil core lysimeters, tracking soil moisture, and interpreting differential nitrate leaching losses from these soils.

This experiment is part of a larger study conducted to assess nutrient leaching from agricultural soils of the Delmarva Peninsula, USA^27,28. Soil core lysimeters were collected from sites in Delaware, Maryland and Virginia in 2010. Here we present unpublished results from these studies. Although initial experiments were conducted to assess phosphorus leaching, nitrate leaching from theses soils was also monitored.

Four common agricultural soils from the Atlantic coastal plain of the Chesapeake Bay Watershed were sampled: Bojac (coarse-loamy, mixed, semiactive, thermic Typic Hapludult); Evesboro (mesic, coated Lamellic Quartzipsamment); Quindocqua (fine-loamy, mixed, active, mesic Typic Endoaquult); Sassafras (fine-loamy, siliceous, semiactive, mesic Typic Hapludult). For each soil, horizon morphology was described from the profiles exposed by the excavation of the columns (Table 1). Surface textures of the soils ranged from sand (Evesboro) to loamy fine sand/sandy loam (Bojac and Sassafras) to silt loam (Quindocqua). Although all soils had been historically fertilized with poultry litter, none had been applied in the 10 months prior to the study. All soils had been in no-till corn production for at least one season prior to soil core lysimeter collection.

Following collection, soil core lysimeters were transported to the USDA-ARS simulatorium facility in State College, PA. There they were subject to indoor irrigation experiments (22-26 °C) to assess nutrient leaching related to poultry litter application. Specifically, lysimeters were irrigated with 2 cm of water weekly for 8 weeks until nitrate in percolate was equilibrated between the soils. Poultry litter (dry poultry manure) was then applied to the surface of all soils at the rate of 162 kg ha^-1 of total N. Irrigation was continued for 5 more weeks. Moisture sensors recorded volumetric moisture content at 5 minute intervals continuously, throughout the irrigation and leaching cycle. Leachate was collected after 24 hr and again 7 days later immediately prior to irrigation.

Leachate data from the soil core lysimeters were analyzed using simple descriptive statistics to illustrate differences in leachate quantity and quality between soils, as well as differences before and after litter application. Because soil moisture sensors were placed in only two of the replicate soil core lysimeters for each soil (Evesboro, Bojac, Sassafras, Quindocqua), statistics for soil moisture content were based on N = 2, whereas statistics for leachate depth, nitrate-N concentration and nitrate-N flux were derived from 10 soil core lysimeters for Evesboro, Bojac and Sassafras and 5 soil core lysimeters for Quindocqua. To assess the importance of replication within soils, coefficients of variation (CV) for leachate depth were calculated for different replicate numbers. A Monte Carlo simulation approach was used to repeatedly sample a subset of soil core lysimeters (N = 3) from the total number of replicates within each soil group (10 for Evesboro, Bojac, Sassafras; 5 for the Quindocqua).