Comparing Passive Wick Lysimeter Against Soil Water Balance and Simulation Models in High and Perched Water Tables and Coarse Sandy Soils

Bibek Acharya¹, Vivek Sharma^1,*, Charles Barrett², Michael D. Dukes¹, Lincoln Zotarelli²

Published in Journal of Natural Resources and Agricultural Ecosystems 3(1): 37-52 (doi: 10.13031/jnrae.16188). Copyright 2025 American Society of Agricultural and Biological Engineers.

¹ Agricultural and Biological Engineering Department, University of Florida, Gainesville, Florida, USA.

² Department of Horticultural Sciences, University of Florida, Gainesville, Florida, USA.

^* Correspondence: vsharma1@ufl.edu

The authors have paid for open access for this article. This work is licensed under a Creative Commons Attribution NonCommercial NoDerivatives 4.0 International License https://creative?commons.org/licenses/by-nc-nd/4.0/

Submitted for review on 12 September 2024 as manuscript number NRES 16188; approved for publication as a Research Article by Associate Editor Dr. Sumon Datta and Community Editor Dr. Kati Migliaccio of the Natural Resources & Environmental Systems Community of ASABE on 17 December 2024.

Citation: Acharya, B., Sharma, V., Barrett, C., Dukes, M. D., & Zotarelli, L. (2025). Comparing passive wick lysimeter against soil water balance and simulation models in high and perched water tables and coarse sandy soils. J. Nat. Resour. Agric. Ecosyst., 3(1), 37-52. https://doi.org/10.13031/jnrae.16188

Highlights

• Passive wick lysimeter performance was tested against soil water balance and simulation models under high and perched water table conditions.
• Passive wick lysimeters overestimated deep drainage as compared to soil water balance and simulation models.
• Lysimeter data inflated NO3-N leachate estimates, questioning the reliability in similar field conditions.
• The out-row placed lysimeters recorded higher drainage volumes than in-row placed.

Abstract. Understanding soil water and nutrient dynamics is pivotal for agricultural and environmental research, particularly the accurate measurement of nutrient leaching. This study critically compares the performance of passive-wick drainage lysimeters to the soil water balance approach and simulation models such as DSSAT (Decision Support System for Agrotechnology Transfer) and HYDRUS-1D (1-dimensional) in estimating deep drainage and nitrate-nitrogen (NO₃-N) leaching under field conditions marked by high and perched water tables and coarse sandy soils. The study was conducted at the North Florida Research and Education Center, Live Oak, FL, over four years (2019–2022). Forty passive-wick drainage lysimeters, with 25 in-row and 15 out-row placements, were installed 61 cm below the soil surface to record deep drainage measurements. The results showed lysimeters exhibited poor correlation (R² = 0.04) with soil water balance estimates during the crop growing season, alongside a Root Mean Square Error (RMSE) of 7.9 liters, indicating substantial measurement discrepancies. The average deep percolation from the lysimeter was 2221 mm (SD = 1490 mm), which was 841% higher than the soil water balance method, 741% higher than the DSSAT simulated deep drainage, and 774% higher than the HYDRUS-1D simulation. During the 2020 maize growing season under a peanut-maize rotation (PC), lysimeter-derived NO₃-N leachate was notably higher at 749 kg ha?¹. In contrast, soil water balance estimates were 110 kg ha?¹, the DSSAT model predicted 127 kg ha?¹, and the HYDRUS-1D model simulated 81 kg ha?¹. Furthermore, lysimeters placed out-row exhibited a higher mean drainage volume compared to those placed in-row, with a percent bias (%bias) of 8.3. These findings highlight the challenges and potential limitations in utilizing passive-wick drainage lysimeters for deep drainage measurement under perched, high-water table, and coarse sandy soil conditions. The substantial overestimation of drainage volumes by lysimeters, particularly evident across various crops, calls for a cautious interpretation of lysimeter data in soil water and nutrient flux studies, especially in environments with similar soil and hydrological properties.

Keywords.Deep percolation, Drainage lysimeter, DSSAT, HYDRUS-1D, Nitrate leaching.

Nitrate-nitrogen (NO3-N) leaching from agricultural fields is a global environmental concern and one of the most pervasive non-point source (NPS) pollutants. Its impacts are evident worldwide, contributing to over 400 hypoxic zones in coastal waters (Diaz and Rosenburg, 2008). One of the most notable examples is the hypoxic zone in the Gulf of Mexico, primarily caused by nitrogen runoff from agricultural fields in the Midwest (McIssac et al., 2001; David et al., 2010). Beyond ecological damage, NO3-N contamination in drinking water poses significant health risks, such as methemoglobinemia and potential links to certain cancers, leading the USEPA to establish a maximum contaminant level of 10 mg L?¹ (USEPA, 2024). These global challenges are mirrored at a regional level in the Suwannee River Basin (SRB) of northern Florida. The SRB receives substantial rainfall and has seen an increase in irrigated farmlands over sandy soils that are hydraulically connected to the karstic Upper Floridan Aquifer (UFA). The region also hosts the highest density of large artesian springs in the world, with approximately 300 springs feeding the Suwannee River, which drains into the Gulf of Mexico. Under these hydrological conditions, most soil nitrogen is rapidly converted to NO3-N, which moves with the wetting front and leaches into the groundwater, creating environmental and water quality challenges. For example, NO3-N concentrations in SRB springs have risen dramatically over the past 40 years, from =0.1 mg L?¹ to more than 5 mg L?¹ (Katz, 2004; Heffernan et al., 2010; Rath et al., 2021). These challenges have prompted state and local governments to implement long-term strategies to monitor and mitigate NO3-N leaching. The Florida Department of Environmental Protection (FDEP) has established a Total Maximum Daily Load (TMDL) target of 0.35 mg L?¹ for NO3-N and developed Basin Management Action Plans (BMAPs) to introduce Best Management Practices (BMPs). These efforts aim to reduce nutrient losses while maintaining crop yields and supporting the competitiveness of farming operations in the region (FDEP, 2013; Zamora Re et al., 2020).

Under field conditions, quantifying NO₃-N losses is complex and challenging because it is intrinsically linked with the soil water dynamics and driven by site-specific parameters such as soil type, rainfall, irrigation, soil water content, soil temperature, presence of active roots, N source, application rates, crop N uptake capacity, water displacement within the crop root zone (Zotarelli et al., 2007). In general, the measurement of NO₃-N leachate requires the knowledge of soil solution flux (leachate) through the soil profile below the root zone and precise quantification of NO₃-N concentration. Thus, to understand the NO₃-N dynamics, quantification of the components of soil water balance—precipitation (P), irrigation (I), runoff (RO), soil moisture (S), evapotranspiration (ET), and particularly deep drainage (D)—becomes paramount (Li et al., 2019). Over the years, different direct and indirect methods, ranging from lysimeters, porous suction cups, and soil cores to mechanistic modeling tools, have been employed to quantify NO₃-N leaching (Webster et al., 1993; Prasad and Hochmuth, 2016; Zotarelli et al., 2007; Acharya et al., 2023). Among various methodologies developed to measure soil water and chemical flux in the unsaturated zone, no standard method has been universally accepted (Gee et al., 2009). Previous studies have critically reviewed various methods to measure/estimate soil water drainage and NO₃-N leaching and discussed the uncertainties in these methodologies (Mason et al., 1983; Gee and Hillel, 1988; Benson et al., 2001; Allen et al., 2011). Among many methods, lysimeters are devices that help in-situ measurement (direct) of deep drainage and nutrient flux in unsaturated soil zones (Benson et al., 2001). They can vary from simple pan/zero tension lysimeters, static tension lysimeters, controlled tension lysimeters, passive wick drainage lysimeters to sophisticated large-scale weighing lysimeters (Abdulkareem et al., 2015). Despite the precision measurement in weighing-type lysimeter, the high cost associated with the installation and maintenance of these lysimeters has led to exploring alternatives such as the use of passive wick drainage lysimeters (Payero and Irmak, 2008). The passive wick drainage lysimeters have been applied in solute leaching studies (Allen et al., 1991, Acharya et al., 2023). The collection efficiency (lysimeter drainage divided by soil water balance drainage) of passive wick drainage lysimeters has been reported to be 100% and sometimes higher (Boll et al., 1991; Zhu et al., 2002), and its efficacy depends on the wicking material, wick's length, soil type, and installation accuracy (Boll et al., 1992; Gee et al., 2002; Meissner et al., 2008; Acharya et al., 2023). Despite their advantages, Derby et al. (2002) reported the inaccuracy resulting from installation error while extracting undisturbed soil monoliths. In addition, discrepancies in water flow measurements can arise when the suction applied by the passive wick lysimeter’s hanging water column does not match the natural soil water suction. The lysimeter collects drainage water through the Divergence Control Tube (DCT), which directs water to a fiberglass wick and subsequently to a reservoir. Water enters the lysimeter when the soil suction at the top of the DCT is equal to or greater than the surrounding soil suction. However, when soil suction at the DCT top ranges between 11 kPa and 33 kPa, water may bypass the lysimeter, leading to flux divergence. Conversely, when soil suction is less than 11 kPa, excess water may flow into the lysimeter, resulting in flux convergence. These phenomena, influenced by mismatches in suction, can affect the accuracy of drainage measurements (Brandi-Dohrn et al., 1996; Acharya et al., 2023). Moreover, the shape and size of the lysimeter need to be considered during installation in relation to crop geometry, water table level, and soil texture to ensure accuracy and representativeness of the drainage volume collected (Shukla et al., 2006). The study by Filipovic et al. (2023), conducted at the Suprehill Critical Zone Observatory in Croatia, investigated nutrient leaching in a sloped vineyard using 18 in situ wick lysimeters installed at a depth of 40 cm. These lysimeters, strategically placed in both intra-row and inter-row locations, collected data monthly over a two-year period. They revealed that NO₃-N concentrations were significantly higher at the foot slope as compared to the hilltop and backslope, but notably, this pattern was exclusively observed in the intra-row areas. This precise placement and analysis of lysimeters not only provided deeper insights into the spatial variability of nutrient leaching but also underscored the critical need for targeted management practices in vineyard operations to mitigate nutrient loss, particularly in sloped terrains.

Another widely accepted method to quantify NO₃-N leaching in unsaturated soils is by direct soil sampling below the root zone. Although this method is simple and relatively inexpensive, it is time-consuming, and it only provides a distribution of NO₃-N concentration within the soil solution in the profile, but no information on soil solution flux. In these cases, the measured solute NO₃-N values are generally combined with physically based models that estimate soil water balance components based on the principle of conservation of mass in the effective root zone to quantify the NO₃-N leaching below the soil surface (Filipovic et al., 2016). In addition to the aforementioned soil water balance, various crop growth and hydrological models such as DSSAT (Hoogenboom et al., 2021) and HYDRUS-1D (Šimunek et al., 2008) can be used to estimate drainage. DSSAT is a widely used model capable of simulating crop growth and development, soil water, nitrogen, and carbon cycling. It estimates the one-dimensional flow of soil water using the simplified Ritchie’s tipping bucket approach (Ritchie, 1998). On the other hand, HYDRUS-1D is a vadose zone mechanistic model that simulates water flow using Richard’s equation (Šimunek et al., 2008). Although DSSAT and HYDRUS 1D differ in their approaches to simulating water flow in the soil, various studies have applied both models to quantify drainage (Hu et al., 2007; Zamora Re et al., 2020; Kanzari et al., 2018). Studies have reported that the soil water balance approach in DSSAT is intrinsically less accurate compared to hydrological models like HYDRUS 1D (Scanlon et al., 2002; Shelia et al., 2018). Suleiman (2008) used the HYDRUS 1D model as a benchmark to test the performance of a new simulation model on deep drainage. Beyene et al. (2018) used the HYDRUS 1D model in irrigated soils of Northwest Ethiopia and reported comparable results between HYDRUS 1D (99 mm average drainage across 9 fields) and a simple bucket-type soil water balance model (135 mm average drainage across 9 fields) in estimating deep drainage. Similarly, Bethune et al. (2008) used the HYDRUS 1D model to test the effectiveness of a conceptual deep percolation model. Soldevilla-Martinez et al. (2014) compared DSSAT and a mechanistic WAVE (Water and Agrochemicals in soil, crop and Vadose Environment) model against weighing lysimeter drainage readings in Spain. Van der Laan et al. (2010) compared NO₃-N concentration measured using the wetting front detector (passive sampler) and ceramic suction cups (active sampler) with NO₃-N concentration simulated drainage using the soil water balance model in a sandy clay loam condition. They observed suction cup concentrations were closely correlated with soil water concentrations, while wetting front detector concentrations aligned closely with draining soil water NO₃-N concentrations. Zotarelli et al. (2007) compared the effectiveness of three different methods for monitoring and quantifying NO₃-N leaching. They observed consistent patterns between drainage lysimeter and soil cores across different mulched vegetable systems. The knowledge gleaned from such studies is crucial for refining crop management practices with the goal of reducing nutrient loss and lessening environmental impacts.

Thus, this paper aims to evaluate the performance of passive-wick drainage lysimeters in measuring deep drainage and NO3-N leaching under field conditions, comparing their efficacy with the soil water balance method and simulation models such as DSSAT and HYDRUS-1D. Additionally, it examines the impact of lysimeter placement (in-row vs. out-row) on drainage volume collection and NO3-N leachate. This study provides critical insights into the performance of lysimeters under challenging hydrological conditions, such as high and perched water tables and coarse sandy soils, where accurate water and nutrient flux monitoring is essential. The findings will not only offer practical recommendations for lysimeter application but also contribute to improving water management and nutrient monitoring strategies in agricultural systems, with implications for sustainable farming and environmental protection globally.

Material and Methods

Study Site and Field Management

The study was conducted at the University of Florida, North Florida Research and Education Center–Suwannee Valley (NFREC-SV) in Live Oak, FL (82°54’20.96” W, 30°18’16.69” N) (fig. 1a). The site featured two predominant soil types: (1) one was a mixture of Hurricane, Albany, and Chipley soils, characterized by 0%–3% slopes, and (2) a blend of Blanton, Foxworth, and Alpin soils, with 0%–5% slopes. In terms of soil taxonomy, the Hurricane soil type was classified as Sandy, siliceous, thermic Oxyaquic Alorthods, and the Blanton soil as Loamy, siliceous, semiactive, thermic Grossarenic Paleudults (Weatherspoon, 2006). Table 1 presents the different soil types present in the study site, corresponding crop rotation, slope, proportion of sand (measured in situ), drainage class, hydrological soil group, saturated hydraulic conductivity (mm s^-1), and the depth to the water table (cm) of the study area (NRCS, 2024).

This study is a part of a 16-year-long sod-based rotational production study. Sod-based rotation is an agricultural system that alternates sod-forming perennial grasses, such as bahiagrass, and legumes with row crops and cereal grains, breaking the row crop cycle for more than one year. (Wright et al., 2005). A 16 ha plot with a center pivot irrigation system was divided into four 4 ha sections to investigate various crop rotation strategies. Two of these sections were set aside for sod-based rotation (BBCP and BBCAP). The other two sections followed conventional rotation: one with continuous maize and peanuts and the other incorporating carrots into the maize-peanut rotation (CAP). To maintain the yearly cultivation of maize and peanuts, one of the conventional sections was split into two smaller parcels of 2 ha each of peanut-maize (PC) and maize-peanut rotation (CP). Over four years, each section completed its rotation cycles, with the sod-based rotation concluding one cycle and the conventional rotation completing two. Additional crop management data on vegetation cover, fertilizer application, irrigation amount, and rainfall received are presented in table 2. This study analyzed the data from the first four years (2019-2022) of rotational production.

Figure 1. Overview of the study site and field condition. (a) Map of study area located in NFREC, Live Oak, and the nearby reservoirs (Google Maps, n.d.) and (b) Shallow and perched water table visible during lysimeter installation in the field. Rotation include PC (peanut-maize), CP (maize-peanut), CAP (maize-carrot-peanut), BBCP (bahiagrass-bahiagrass-maize-peanut), and BBCAP (bahiagrass-bahiagrass-maize-carrot-peanut).

Estimation of Deep Drainage and Nitrate-Nitrogen Leaching

Between 11 and 27 January 2019, a total of forty passive wick drainage lysimeters (Drain Gauge G3, Decagon Devices, Pullman, Washington, USA) were installed in the study field. These lysimeters extend 147 cm in length and are designed for the accurate (± 1.4 mm/observation) long-term observation of vertical movements in soil water and the flux of chemicals (Decagon Devices, Inc., 2018). Out of 40 lysimeters, 25 were placed in-row (within the rows where crops are planted) and 15 were placed out-row (in the rows where crops are not planted) (as shown in table S1 of the supplementary material). Considering the shallow water table and perched water tables of the study area (fig. 1b), the DCT opening was placed 61 cm below the soil surface. A key process during the installation was selecting a representative site with minimal soil disturbance to maintain natural conditions. The DCT requires an intact soil monolith that corresponds to the crop root zone's depth. Wooden boards were used to prevent direct damage to the DCT during installation. While refilling the soil after installing the lysimeter setup, effort was made to ensure the compactness of the replaced soil matched that of the surrounding ground to maintain uniformity. To address shallow and perched water tables, the installation involved a unique step of inserting a steel pipe into an augured hole to prevent soil from collapsing. This pipe was then removed after the drain gauge was positioned, ensuring its stability within the soil. Five potentiometric wells were installed to monitor groundwater levels beneath various crop rotation systems. Over the four years, the groundwater levels were monitored, and instances where the groundwater levels exceeded the DCT (positioned 61 cm below the soil surface) were carefully documented. It is important to note that the lysimeter setup measures drainage below a depth of 122 cm (61 cm from the soil surface to the Divergence Control Tube opening, plus 61 cm for the DCT tube length). To prevent root interference with measurements, a root growth inhibitor is installed at the junction of the DCT's bottom and the reservoir opening. This consistent drainage setup was applied across all lysimeters, including soil water balance, DSSAT, and HYDRUS-1D simulations. For comprehensive detail on passive wick drainage lysimeter components, their working principles, and the specific installation techniques used in this research, please refer to Acharya et al. (2023).

Lysimeter water sampling was carried out following methods described in FDEP SOPs FD 1000, FQ 1000, FS 1000, and FS 2000 (FDEP, 2010). For 44 months, every week depending on the rainfall, the leachate collected from the lysimeter was first extracted into a 1-liter sampling jar to obtain representative samples for analysis. From this 1-liter jar, 500 ml and 250 ml samples were transferred into two smaller plastic bottles. The remaining leachate was collected in a larger 20-liter sampling jar for leachate volume measurement. These bottles were distinctly labeled to prevent any mix-up of samples. To preserve the samples, a few drops of sulfuric acid were added to the 500 ml bottle and phosphoric acid to the 250 ml bottle, with the quantity of acid based on achieving a very low pH for the samples. This further effectively inhibits microbial activity and chemical transformations, which could otherwise alter the concentrations of analytes such as nitrates and ammonium. A cooler with ice was used in the field to preserve the integrity of the samples immediately after collection. Once transported from the field, the samples were promptly transferred to a refrigerator until their delivery to the testing laboratory (as shown in figure S1 of the supplementary material). This procedure ensured that the samples remained at an appropriate low temperature throughout the process, preventing any degradation of analytes. Additionally, field duplication samples were taken randomly from the same original sampling bottle and analyzed separately to assess the variability and potential errors, including contamination. These samples were then securely placed in an ice cooler, clearly labeled, and promptly transported to the FDACS Agricultural and Environmental Services Lab in Tallahassee, FL, a facility accredited by the National Environmental Laboratory Accreditation Program (NELAP), for detailed analysis of NO₃ by EPA method 353.2 (USEPA, 1993a), NH₄ by EPA method 350.1 (USEPA, 1993b), total Kjeldahl N (TKN) by EPA method 351.2 (USEPA, 1993c), and total organic carbon by EPA method 9060A in SW-846 (USEPA, 2004). This analysis helps in calculating the chemical fluxes in the soil by using data on the drainage volume and chemical concentration of the water that has drained deep into the soil (eq. 1). Likewise, the collection efficiency of the lysimeter was computed as a ratio of drainage volume collected to drainage volume expected (soil water balance) (eq. 2).

(1)

(2)

where cross sectional inner surface area of DCT is 506.7 cm².

Soil Water Balance

The soil water balance considers the inflow, storage, and outflow of water within the soil system. Water volume inputs were from rainfall (P) and irrigation practices (I), while water volume exiting the system through crop evapotranspiration (ET), surface runoff (RO), and deep percolation (DP) (table 3). Groundwater contribution to soil water balance was not accounted for in the study. This decision was based on the coarse sandy soils in the experimental field, where capillary rise is minimal due to large pore spaces that limit upward water movement (Cao et al., 2022). Furthermore, groundwater levels were monitored using five potentiometric wells, and days when groundwater levels exceeded the lysimeter's Divergence Control Tube (DCT) were excluded from the analysis to minimize the influence of direct groundwater contributions (fig. 2). Rainfall measurements were obtained from the Florida Automated Weather Network (FAWN, 2022), located approximately 500 m from the experimental field. FAWN uses tipping bucket rain gauges that measure in increments of 0.0254 cm and report data every 15 minutes. Water applied through irrigation was measured on the farms. The calculation of runoff was based on the NRCS Curve Number (CN) method, which evaluates the impact of land use, soil type, and the initial soil moisture condition, classified into three categories (I, II, or III) to represent different levels of soil moisture (NRCS, 1985) (eq. 4). The reference evapotranspiration (ETo) was determined using the FAO Penman-Monteith equation (Allen et al., 1998, eq. 5). Daily climatic data such as maximum and minimum temperatures (°C), humidity levels (%), wind speed (ms^-1), solar radiation (Wm^-2), and rainfall (mm) were collected from FAWN and used in this calculation (FAWN, 2022). Crop evapotranspiration (ETc) was estimated by a two-step method that involves multiplying ETo by crop-specific coefficients, following the guidelines of Allen et al. (1998) (eq. 6). Changes in soil moisture (?S) were tracked using time-series capacitance probes installed at depths of 15 cm, 30 cm, and 45 cm. Deep percolation (DP) was derived indirectly as a residual of the water balance which considers soil properties and moisture content using equation 3.

Figure 2. Groundwater table depth (mm) below the soil surface and daily rainfall (mm; plotted on the secondary axis) observed during 2019–2022 active crop growing season at North Florida Research and Education Center (NFREC), Live Oak, FL. The red dashed line indicates the lysimeter DCT opening, located 61 cm below the soil surface.

(3)

(4)

(5)

(6)

(7)

(8)

where

DP_i = measurement of deep percolation (mm) on day i

I_i = irrigation depth (mm) on day i

P_i = amount of precipitation (mm) received on day i

RO_i = runoff depth (mm) from the soil surface on day i

fw = fraction of soil surface wetted by irrigation or precipitation, ranging from 0.01 (minimal wetting) to 1 (complete wetting), depending on irrigation method and extent of wetting

ET_c,i = crop's evapotranspiration (mm) on day i

D_r = root zone depletion or the water deficit relative to field capacity

D_(r,i-1) = total decrease in moisture within the crop's root zone (mm) by the end of day i-1

I_a = initial abstraction, including surface storage, interception, and infiltration prior to runoff, commonly approximated as 0.2 × S

S = potential maximum retention after runoff begins (mm), which is a function of the CN

ET_o = reference evapotranspiration (mm d?¹)

? = slope of the vapor pressure curve (kPa °C?¹)

R_n = net radiation at the crop surface (MJ m?² d?¹)

G = soil heat flux density (MJ m?² d?¹)

? = psychrometric constant (kPa °C?¹)

T = mean daily air temperature at 2 m height (°C)

u₂ = wind speed at 2 m height (m s?¹)

es = saturation vapor pressure (kPa)

ea = actual vapor pressure (kPa)

K_cb = basal crop coefficient

K_e = soil evaporation coefficient

K_c-mid = mid-stage crop coefficient, adjusted based on relative humidity and wind speed

K_c-end = end-stage crop coefficient, adjusted based on relative humidity and wind speed

RH_min = daily average minimum relative humidity during respective growth stage (%), valid within 20% = RHmin = 80%

h = average plant height during the respective growth stage (m), valid for 0.1 m < h < 10 m.

The baseline values for K_c-mid(Table) and K_c-end(Table) are provided in Allen et al. (1998). The NO₃-N leaching using soil water balance approach was calculated using the soil water balance estimated drainage (eq. 3) coupled with NO₃-N concentration analysis from water samples collected from the drainage lysimeters.

DSSAT

The DSSAT model utilized DSSAT version 4.8 (Hoogenboom et al., 2021), which employs Ritchie’s one-dimensional "tipping bucket" method to model the soil water balance (Ritchie, 1998). This approach requires input parameters such as the drained upper limit (DUL), lower limit (LL), and saturated water content (SAT) for each soil layer to calculate available soil water. Drainage occurs when the water content in a soil layer exceeds the DUL. In DSSAT, nitrogen transport within the soil profile is driven by simulated water flux values from the soil water module. The model calculates nitrate movement within and between soil layers by considering factors such as the adsorption coefficient, bulk density, and volumetric water content at the drained upper limit. Nitrate transport is closely linked to water flux, with drainage events causing nitrate to move from one layer to the next.

DSSAT was calibrated using field data from the 2019 peanut growing season and the 2020 maize growing season within the peanut-maize (PC) rotation, as these periods provided optimal conditions for maximizing growth and yield. Sequential adjustments were made to cultivar coefficients within their uncertainty ranges to align simulated results with observed maturity dates and yields for maize and peanuts. Statistical metrics, such as Root Mean Square Error (RMSE) and Normalized RMSE (nRMSE), were employed to assess model accuracy. Calibration results for crop yield showed an RMSE of 940 kg ha?¹ (nRMSE = 12.8%) for peanuts and an RMSE of 32 kg ha?¹ (nRMSE = 0.2%) for maize. During validation, RMSE values ranged from 462 to 1883 kg ha?¹ (nRMSE = 3.5%–13.3%), highlighting the model's reliability in estimating crop growth and yield. The DSSAT calibration and validation results are provided in table S2 of the supplementary materials.

HYDRUS-1D

HYDRUS 1D numerically solves Richard’s equation (eq. 9) to simulate water flow in the vadose zone, with water flowing below the root zone (bottom flux) referred to as deep drainage (Šimunek et al., 2008). Likewise, HYDRUS-1D models the simultaneous transport of multiple solutes, including ammonium (NH₄-N) and nitrate (NO₃-N), with first-order degradation. The transport equations were solved using a Galerkin-type linear finite element method. The partial differential equations governing the one-dimensional advective-dispersive transport of nitrogen, ammonium, and nitrate as shown in equation 8 (Šimunek et al., 2008). In this context, T represents the volumetric water content (cm³ cm^-³), t is time (day), z is the spatial coordinate (cm), h is the pressure head of water (cm), K is the unsaturated hydraulic conductivity (cm day^-1), and S represents root water uptake (cm³cm^-³day^-1), c is the concentration of solute (soil NO₃) in the liquid phase, µw and µ'w are first-order rate constants for solute and chain reactions, respectively, c₂ is the concentration of NH₄⁺-N in the liquid phase, D is the dispersion coefficient, and q is the volumetric flux density. The HYDRUS-1D models nitrogen transformations, including the hydrolysis of nitrogen species into NH₄?-N and the nitrification of NH₄?-N into NO₃?-N. NH₄?-N considered in both adsorbed and dissolved phases, while NO₃?-N is only in the dissolved phase.

(9)

(10)

HYDRUS-1D version 4.17.0140 was used in this study and calibrated using soil moisture data from the 2019 peanut growing season in the peanut-maize (PC) rotation. The soil hydraulic parameters, such as saturated and residual soil moisture (?_s, ?_r ?), the soil hydraulic conductivity (K_s), and shape parameters (a, n), were first adjusted based on observed soil moisture content at 15 cm, 30 cm, and 45 cm depth. These calibrated parameters were then applied to the remaining growing seasons to validate the model's performance. Statistical metrics, including Root Mean Square Error (RMSE) and Normalized RMSE (nRMSE), were used to evaluate the accuracy of the model. The model's initial conditions were defined by the soil's water content and nitrogen concentration, with boundary conditions set for both water and solute transport. For water, the upper boundary was defined by atmospheric conditions with runoff, while the lower boundary had free drainage. For solute transport, the upper boundary was set as concentration flux, and the lower boundary had a zero-concentration gradient. The time-variable boundary conditions for water and solute transport accounted for precipitation, potential evaporation, potential transpiration, and N fertilizer applied during the growing season. An hCritA value of 100,000 cm was used, representing the absolute minimum pressure head at the surface. Root water uptake followed the method developed by Feddes et al. (1982), using parameters from the HYDRUS-1D database (h1 = -10 cm, h2 = -25 cm, h3 = -200 cm, and h4 = -8000 cm) for both maize and peanut. The root growth parameters were defined assuming that half of the maximum rooting depth was reached by mid-season with the initial rooting depth set at 3.8 cm and the maximum rooting depth of 90 cm. The calibration results showed strong agreement between simulated and observed soil moisture content, with RMSE values of 0.02 cm³ cm?³ and nRMSE values of 20%, 18%, and 19% at soil depths of 15 cm, 30 cm, and 45 cm, respectively. During validation (pooled data across the growing season), RMSE was 0.04 cm³ cm?³, and nRMSE values ranged from 38% to 42% across all depths. The calibration and validation results are detailed in table S3 in the supplemental material.

The use of three models—soil water balance, DSSAT, and HYDRUS-1D—was intended to leverage their complementary strengths and provide a comprehensive evaluation of drainage and NO3-N leaching under field conditions. The soil water balance approach offered a practical framework for estimating drainage based on measured water inputs and outputs, while DSSAT provided insights into crop-soil interactions, including nutrient uptake and yield, by simulating nitrogen transport linked to water flux. HYDRUS-1D added a mechanistic dimension, accurately modeling water flow and solute transport in the vadose zone, including nitrogen transformations. This combination allowed us to compare and validate results across approaches, ensuring robust and reliable findings.

Statistical Analysis

Statistical analysis was conducted using RStudio Version 2023.12.1. Box plots and histograms were utilized to visually assess the data distribution, highlighting central tendencies and potential outliers. Measures of central tendency, specifically the mean and median, along with standard deviation, were employed to analyze the non-normal data. The Shapiro-Wilk test was used to evaluate the normality of the data distribution because of its robustness and reliability, particularly for small to medium sample sizes, which aligns with the study's dataset. This test provided a foundation for selecting appropriate descriptive statistical methods. Lysimeter drainage data collected during the 2019–2022 crop growing seasons were analyzed to evaluate drainage patterns and frequency distributions. A total of 2781 observations were grouped into 2 L bins, ranging from 0.1 L to 27 L, to facilitate pattern visualization. Descriptive statistics, including means and frequencies, were used to summarize the data. In addition, statistical metrics, namely root mean square error (RMSE) (eq. 11) and normalized root mean square error (nRMSE) (eq. 12), were used to evaluate DSSAT and HYDRUS-1D models, where O_i is the i^th observed value, S_i is the i^th simulated value, is the average of observed and n is number of samples.

Figure 3. Daily and annual variation of precipitation (P) at the North Florida Research and Education Center-Suwannee Valley, Live Oak, Florida.

(11)

(12)

Results and Discussion

Precipitation and Leaching Rain Event

In 2021, the study area experienced the highest precipitation out of the study period, totaling 1417 mm, followed by 1365 mm in 2020—both exceeding the historical annual average of 1243 mm, calculated from 2003 to 2022. Conversely, 2019 and 2022 received the lowest totals of 1091 mm and 1130 mm, respectively (fig. 3). The highest single-day precipitation during the study period was recorded on 7 June 2020, with 205.2 mm. Between 2019 and 2022, 18 leaching rain events were documented. These events, as defined by the Florida Department of Agriculture and Consumer Services (FDACS, 2015), involve either at least 76 mm of precipitation over three days or 102 mm over seven days. During the study, there were two three-day criterion events in 2019, three each in 2020 and 2021, and one in 2022. For the 7-day criterion, two events were noted in 2019, 2020, and 2022, and three in 2021. These events are crucial for assessing the impact of heavy rainfall on soil nutrients and guiding crop management practices.

Groundwater Level

The average groundwater level measurements across four rotational systems throughout the study years with lysimeter's divergence control tube (DCT) positioned 61 cm below the soil surface as a reference are presented in table 4. Over the four-year study period, the groundwater level generally remained below the DCT reference point. The peanut-maize (PC) rotation had the highest average groundwater level of 1505 mm, with no instances of groundwater surpassing the DCT threshold during either the study period or active growing seasons (from planting to harvest). The maize-peanut (CP) rotation averaged 812 mm, exceeding the DCT threshold 187 times during the study period and 25 times during active growing seasons. The maize-carrot-peanut (CAP) rotation recorded an average of 1044 mm, with 58 exceedances during the study period, of which 8 occurred during the active growing seasons. For the sod-based rotations, BBCP-1 had an average groundwater level of 1163 mm, with two instances above the DCT threshold during both the study period and active growing seasons. BBCP-2 had an average groundwater level of 1018 mm, with 210 exceedances during the study period, 40 of which occurred during the active growing seasons. It is important to note that data points where the groundwater level exceeded the DCT threshold were excluded from NO₃-N concentration estimation from the drainage samples.

Lysimeter and Soil Water Balance Drainage Distributions

Lysimeter drainage measurements were collected during the 2019-2022 crop growing seasons, resulting in 2781 observations (fig. 4a). The drainage volumes were grouped into 2 L bins, ranging from a minimum of 0.1 L to a maximum of 27 L. The mean drainage volume was 8 L, with a standard deviation of 4.6 L. The most frequent drainage volume was in the 8–10 L range, with 923 occurrences. Notably, 1412 observations were less than or equal to the lysimeter reservoir capacity of 8.2 L, while 1369 exceeded this capacity. In contrast, the distribution of soil water balance drainage volumes during the same periods showed most values concentrated at the lower end of the spectrum (fig. 4b). Out of 181 measurements, 126 fell within the 0–2 L range, with a mean of 1.7 L and a standard deviation of 1.7 L. Larger drainage volumes were less common, with values above 6 L being particularly rare. The maximum recorded value was 9.6 L, and the minimum was 0.02 L. Notably, lysimeter drainage measurements and soil water balance drainage were compared exclusively for the 2019–2022 crop growing seasons, excluding fallow periods. This resulted in 181 drainage measurements using the soil water balance method. The accuracy of lysimeter measurements can be affected by several factors, including lysimeter geometry, local groundwater levels, nearby reservoirs, the presence of perched water tables within the soil profile, and issues related to installation and operation (Aboukhaled et al., 1982; Abichou et al., 2006; McLean et al., 2018; Riedl et al., 2022; Radovanovic et al., 2021), all of which can significantly impact the reliability of drainage measurements.

Efficiency and Performance of Passive Wick Drainage Lysimeters

Figure 5 shows the distribution of lysimeter collection efficiency percentages, calculated relative to soil water balance drainage estimates. The data clustered in two distinct areas: one within the 300%–400% range, with 26 observations, and another at the higher end, with 60 observations exceeding 1000%. In contrast, the 0%–100% range contained only a single observation. The frequent recording of drainage volumes by lysimeters that exceeded soil water balance estimates suggests inefficacy in the study area. This overestimation can be attributed to factors such as poor evaporation capture, root water uptake, high groundwater levels, perched water tables, and inaccuracies in lysimeter installation and design (Hyde and Ford, 1989; Gee et al., 2009; Meissner et al., 2010; Simonne et al., 2024). Throughout the 2019–2022 growing seasons, all 40 lysimeters consistently recorded drainage volumes that exceeded their 8.2 L maximum capacity. Figure 6 illustrates this trend during the 2020 maize growing season in the maize-peanut (CP) rotation, where lysimeter readings frequently reached or exceeded capacity, even during periods without rainfall. This pattern was observed in maize, peanut, and carrot crops.

Supporting evidence from Gee et al. (2009) highlights similar limitations of lysimeters in determining drainage volume. In their study of passive wick fluxmeters in a squash plantation in Tonga, over a 60-day period with 340 mm of rainfall, the fluxmeters recorded 482 mm of drainage, resulting in a collection efficiency of 223%, which was significantly higher than the water balance estimate of 217 mm. This suggests convergent flow and overestimation of drainage, emphasizing the need to match lysimeter design to specific site conditions—such as soil hydraulic properties and rainfall patterns—to improve measurement accuracy.

An analysis comparing lysimeter and soil water balance drainage measurements, based on 181 observations recorded over the 2019–2022 crop growing seasons, revealed significant deviations between the two methods. Preliminary statistical tests, including the Shapiro-Wilk test, indicated non-normality in both datasets. The median lysimeter-measured drainage volume was 8.2 L, while the median soil water balance estimate was 1.4 L, highlighting a substantial overestimation by the lysimeters. This overestimation, with an average difference ranging from 6.3 to 7.54 L, was further quantified by a Root Mean Square Error (RMSE) of 7.9 L, a Mean Bias Error (MBE) of 6.9 L, and %BIAS of 401%, underscoring the significant discrepancy. Additionally, the Index of Agreement at 0.24 and the weak coefficient of determination (R² = 0.04) (fig. 7) further confirmed the misalignment between lysimeter and soil water balance measurements.

(a)

(b)

Figure 4. Frequency distribution [Bin width = 2 liter (L)] of deep drainage. (a) lysimeter measurements and (b) soil water balance estimated.

Figure 5. Collection efficiency (%) of drainage lysimeter as compared to soil water balance.

Figure 6. Lysimeter drainage and soil water balance (SWB) drainage volume during 2020 maize growing season in maize-carrot-peanut (CAP) rotation. Rainfall received during the duration is plotted in secondary axis.

Figure 7. Correlation between drainage volume measured by lysimeter and soil water balance.

Comparison of Drainage Volume and No₃-N Leaching Estimates

Table 5 lists average deep percolation data measured from drainage lysimeters, soil water balance, DSSAT model, and HYDRUS 1D model during the 2019–2022 crop growing seasons for peanut-maize (PC) and maize-peanut (CP) crop rotations. The average deep percolation from the soil water balance was 236 mm (SD = 126 mm), from the DSSAT model was 264 mm (SD = 89 mm), from the HYDRUS 1D model was 254 mm (SD = 144 mm), while the lysimeter measurements were significantly higher at 2221 mm (SD = 1490 mm).

Similarly, figure 8 compares lysimeter-measured drainage with soil water balance (SWB) estimates, as well as DSSAT and HYDRUS 1D model simulations, during the 2020 peanut growing season in the maize-peanut (CP) rotation. Due to the high drainage volumes recorded by the lysimeters, the NO₃-N leachate measurements were higher than expected. These discrepancies may be attributed to various factors affecting lysimeter accuracy, such as geometry, groundwater levels, nearby reservoirs, perched water tables, and potential issues with installation and operation (Aboukhaled et al., 1982; Abichou et al., 2006; McLean et al., 2018; Riedl et al., 2022; Radovanovic et al., 2021). These factors should be considered when interpreting drainage volumes, as they can significantly impact the accuracy of soil water balance estimates.

Lysimeter drainage values, combined with NO₃-N concentration data, were used to calculate NO₃-N leaching, revealing significant discrepancies between methods. Lysimeter results indicated substantial NO₃-N leachate, ranging from 124 to 665 kg ha?¹ in maize fields during the 2019-2022 study years, while soil water balance estimates were much lower, ranging from 11 to 109 kg ha?¹. Similarly, lysimeter measurements of NO₃-N leaching in peanut fields ranged from 84 to 898 kg ha?¹, whereas soil water balance estimates ranged from only 6 to 133 kg ha?¹. These measurements were also compared with simulation models (fig. 9). During the 2020 maize growing season under a peanut-maize rotation (PC), lysimeter-derived NO₃-N leachate was notably higher at 749 kg ha?¹. In contrast, soil water balance estimates were 110 kg ha?¹, the DSSAT model predicted 127 kg ha?¹, and the HYDRUS-1D model simulated 81 kg ha?¹. These stark differences underscore the limitations of lysimeters as a reliable method for assessing NO₃-N leaching under field conditions. Improving the accuracy of deep drainage estimates may require integrating lysimeters with additional data sources, such as hydrological modeling (Zhang et al., 2019; Yu et al., 2023), soil moisture sensors (Zhang et al., 2019), machine learning (Wohling et al., 2012; Yu et al., 2023), and satellite imagery (Yu et al., 2023) to address discrepancies and enhance the reliability of measurements.

Figure 8. A comparison between lysimeter drainage, soil water balance (SWB) drainage, DSSAT, and HYDRUS 1D simulated drainage during 2020 peanut growing season in maize-peanut (CP) rotation. Rainfall during the duration is plotted in secondary axis.

Figure 9. A comparison between nitrate-nitrogen leachate (NO3-N) leachate computed using lysimeter drainage, soil water balance (SWB) drainage, DSSAT, and HYDRUS 1D simulation during 2020 maize growing season in peanut-maize (PC) rotation.

In-Row vs Out-Row Lysimeter Placement

The analysis of lysimeter placement data delineated as "in-row" and "out-row,” revealed distinct differences in drainage volume. The 'out-row' placements exhibited a mean drainage volume of 8.6 L, compared to 7.9 L for the 'in-row' placements (fig. 10). This was based on 2835 'in-row' observations from 25 lysimeters and 1744 'out-row' observations from 15 lysimeters, covering both crop growth and fallow periods. Despite the differences in mean drainage volumes, both 'in-row' and 'out-row' groups displayed a similar spread of data, with standard deviations of 4.7 L and variances approximately equal to 22. The maximum volume collected by the 'in-row' lysimeters reached 28.1 L, while the 'out-row' had a slightly lower maximum of 26.7 L. The minimum recorded for the 'in-row' was marginally lower at 0.05 L compared to 0.1 L for the 'out-row'. Statistical analysis confirmed that both 'in-row' and 'out-row' datasets significantly deviated from a normal distribution, as indicated by the Shapiro-Wilk test results (p-values < 0.001). 'In-row' placed lysimeter had a median of 8.2 L and ‘out-row’ placed lysimeters had a median of 8.35 L. These insights are critical for optimizing lysimeter placement in future research, ensuring accurate and efficient volume collection (Bergstrom et al., 1990).

Furthermore, the investigation into nutrient leachate levels across maize and peanut crops revealed that neither the maize nor peanut data followed a normal distribution (fig. 11). For maize, the in-row mean leachate level was 8.03 kg ha^-1 with an SD of 11.8 kg ha^-1 from 237 observations, ranging from 0.004 to 99.3 kg ha^-1. The out-row placements had a mean of 8.3 kg ha^-1, SD of 10.7 kg ha^-1, with 142 observations, ranging from 0.005 to 49.6 kg ha^-1. The median for nutrient leachate levels in maize from ‘in-row’ placed lysimeters was 3.75 kg ha^-1 while from ‘out-row’ placed lysimeters was 4.62 kg ha^-1. For peanuts, the in-row mean leachate level was 8.3 kg ha^-1 with an SD of 19.5 kg ha^-1 from 254 observations, ranging from 0.007 to 226.6 kg ha^-1. The out-row placements had a mean of 9.1 kg ha^-1, SD of 20 kg ha^-1, with 161 observations, ranging from 0.031 to 151 kg ha^-1. The median for nutrient leachate levels in peanut from ‘in-row’ placed lysimeters was 2.2 kg ha^-1 while from ‘out-row’ placed lysimeters was 2.1 kg ha^-1. Filipovic et al. (2023) observed higher concentrations of NO₃-N, orthophosphate, and potassium at the vineyard footslope, particularly within intra-row areas. While Filipovic et al. (2023) and this study underscores the importance of lysimeter placement in understanding nutrient leaching dynamics, this study’s results diverge in the specific patterns of nutrient concentration, suggesting that site-specific factors such as soil properties, cropping system, and climatic conditions may significantly influence nutrient leaching.

(a)

(b)

Figure 10. Box plots display the central tendency, spread, potential outliers, quartiles and the whiskers for lysimeters drainage (L) in In-row vs out-row lysimeter placement. (a) in-row and (b) out-row placed.

Figure 11. Box plots display the central tendency, spread, potential outliers, quartiles and the whiskers for NO3-N leachate (kg ha-1) in In-row vs. out-row lysimeter placement. (a) in-row (maize), (b) out-row (maize), (c) in-row (peanut), and (d) out-row (peanut).

Conclusion

This study evaluated the performance of passive wick drainage lysimeters compared to the soil water balance method and simulation models, specifically DSSAT and HYDRUS-1D, in estimating deep drainage and NO3-N leaching under high and perched water table conditions. Additionally, it examined the impact of lysimeter placement (in-row vs. out-row) within the field. The findings revealed notable discrepancies between lysimeter-derived measurements and other methods, with lysimeters consistently overestimating drainage volumes and nitrate-nitrogen (NO3-N) leaching. For instance, during the 2020 maize growing season in a peanut-maize rotation (PC), lysimeter-derived NO3-N leachate was significantly higher (749 kg ha?¹) than estimates from the soil water balance (110 kg ha?¹), DSSAT (127 kg ha?¹), and HYDRUS-1D (81 kg ha?¹). These results highlight the importance of considering local soil and hydrological conditions when interpreting lysimeter data, particularly in fields with shallow ground or perched water tables. The study underscores the need for caution when relying solely on lysimeter data and advocates for complementary methods to enhance the reliability of deep drainage and nutrient flux measurements. Future research could refine these approaches by incorporating adjustments such as accounting for capillary rise in the soil water balance or employing alternative tools like suction cup or weighing-type lysimeters to improve accuracy under challenging field conditions.

Supplemental Material

The supplemental materials mentioned in this article are available for download from the ASABE Figshare repository at: https://doi.org/10.13031/28223027

Acknowledgments

We extend our sincere gratitude to the Florida Department of Agriculture and Consumer Services - Office of Agricultural Water Policy (FDACS-OAWP) for their generous support and funding (Award # AWD03374). Special thanks to the dedicated team at the North Florida Research and Education Center for their invaluable assistance and collaboration throughout the project.

References

Abdulkareem, J. H., Abdulkadir, A., & Abdu, N. (2015). A review of different types of lysimeter used in solute transport studies. Int. J. Plant Soil Sci., 8(3), 1-14. https://doi.org/10.9734/IJPSS/2015/18098

Abichou, T., Liu, X., & Tawfiq, K. (2006). Design considerations for lysimeters used to evaluate alternative earthen final covers. J. Geotech. Geoenviron. Eng., 132(12), 1519-1525. https://doi.org/10.1061/(ASCE)1090-0241(2006)132:12(1519)

Aboukhaled, A., Alfaro, A., & Smith, M. (1982). FAO Irrigation and Drainage Papers 39: Lysimeters. Rome, Italy: United Nations FAO.

Acharya, B., Sharma, V., Barrett, C., Sidhu, S. S., Zotarelli, L., & Dukes, M. D. (2023). Methods to quantify in-field nutrient leaching: AE581/AE581, 12/2022. EDIS, 2022(6). https://doi.org/10.32473/edis-ae581-2022

Allen, R. G., Howell, T. A., Pruitt, W. O., Walter, I. A., & Jensen, M. E. (1991). Lysimeters for evapotranspiration and environmental measurements. Proc. International Symposium on Lysimetry (pp. 79-84). New York: American Society of Civil Engineer.

Allen, R. G., Pereira, L. S., Howell, T. A., & Jensen, M. E. (2011). Evapotranspiration information reporting: I. Factors governing measurement accuracy. Agric. Water Manag., 98(6), 899-920. https://doi.org/10.1016/j.agwat.2010.12.015

Allen, R. G., Pereira, L. S., Raes, D., & Smith, M. (1998). Crop ET: Guidelines for computing crop water requirements - FAO Irrigation and Drainage Paper No. 56. Rome, Italy: United Nations FAO.

Benson, C., Abichou, T., Albright, W., Gee, G., & Roesler, A. (2001). Field evaluation of alternative earthen final covers. Int. J. Phytorem., 3(1), 105-127. https://doi.org/10.1080/15226510108500052

Bergström, L. (1990). Use of lysimeters to estimate leaching of pesticides in agricultural soils. Environ. Pollut., 67(4), 325-347. https://doi.org/10.1016/0269-7491(90)90070-S

Bethune, M. G., Selle, B., & Wang, Q. J. (2008). Understanding and predicting deep percolation under surface irrigation. Water Resour. Res., 44(12). https://doi.org/10.1029/2007WR006380

Beyene, A., Cornelis, W., Verhoest, N. E., Tilahun, S., Alamirew, T., Adgo, E.,... Nyssen, J. (2018). Estimating the actual evapotranspiration and deep percolation in irrigated soils of a tropical floodplain, northwest Ethiopia. Agric. Water Manag., 202, 42-56. https://doi.org/10.1016/j.agwat.2018.01.022

Boll, J., Selker, J. S., Nijssen, B. M., Steenhuis, T. S., Van Winkle, J., & Jolles, E. (1991). Water quality sampling under preferential flow conditions. In R. G. Allen, T. A. Howell, W. O. Pruitt, I. A. Walter, & M. E. Jensen (Eds.), Lysimeters for evapotranspiration and environmental measurement: Proceedings of the ASCE International Symposium on Lysimetry (pp. 290-298). ASCE.

Boll, J., Steenhuis, T. S., & Selker, J. S. (1992). Fiberglass wicks for sampling of water and solutes in the vadose zone. Soil Sci. Soc. Am. J., 56(3), 701-707. https://doi.org/10.2136/sssaj1992.03615995005600030005x

Brandi-Dohrn, F. M., Hess, M., Selker, J. S., & Dick, R. P. (1996). Field evaluation of passive capillary samplers. Soil Sci. Soc. Am. J., 60(6), 1705-1713. https://doi.org/10.2136/sssaj1996.03615995006000060014x

Cao, J., Mao, Z., Hu, L., Dai, Y., Luo, Y., & Zhou, W. (2022). Analysis and prediction of stable height of capillary water rising under unsaturated subgrade soil. Adv. Civil Eng., 2022(1), 5993984. https://doi.org/10.1155/2022/5993984

David, M. B., Drinkwater, L. E., & McIsaac, G. F. (2010). Sources of nitrate yields in the Mississippi River Basin. Journal of environmental quality, 39(5), 1657-1667.

Decagon Devices, Inc. (2018). Drain Gauge G3: Operator’s Manual. Pullman, WA.

Derby, N. E., Knighton, R. E., & Montgomery, B. R. (2002). Construction and performance of large soil core lysimeters. Soil Sci. Soc. Am. J., 66(5), 1446-1453. https://doi.org/10.2136/sssaj2002.1446

Diaz, R. J., & Rosenberg, R. (2008). Spreading dead zones and consequences for marine ecosystems. science, 321(5891), 926-929.

FAWN. (2022). Florida Automated Weather Network: Data access. Retrieved from https://fawn.ifas.ufl.edu/data/reports/

FDACS. (2015). Water Quality/Quantity BMPs for Florida Vegetable and Agronomic Crops: 2015 Edition, FDACS-P-01266. Tallahassee, FL: Florida Department of Agriculture and Consumer Services.

FDEP. (2010). Florida Administrative Code, Quality Assurance, Rule 62-160. Retrieved from https://www.flrules.org/gateway/?ChapterHome.asp?Chapter=62-160

FDEP. (2013). Florida Administrative Code, Surface water quality criteria, Rule 62-302.530. FDEP. Retrieved from https://www.flrules.org/gateway/RuleNo.asp?ID=62-302.530

Feddes, R. A. (1978). Simulation of field water use and crop yield. In Simulation of plant growth and crop production (pp. 194-209). Pudoc.

Filipovic, L., Krevh, V., Chen, R., Defterdarovic, J., Kovac, Z., Mustac, I.,... Filipovic, V. (2023). Quantification of intra- vs. Inter-row leaching of major plant nutrients in sloping vineyard soils. Water, 15(4), 759. https://doi.org/10.3390/w15040759

Filipovic, V., Ondrašek, G., & Filipovic, L. (2016). Modelling water dynamics, transport processes and biogeochemical reactions in soil vadose zone. In M. S. Javaid (Ed.), Groundwater - Contaminant and resource management (p. Ch. 7). Rijeka: IntechOpen. https://doi.org/10.5772/63496

Gee, G. W., & Hillel, D. (1988). Groundwater recharge in arid regions: Review and critique of estimation methods. Hydrol. Process., 2(3), 255-266. https://doi.org/10.1002/hyp.3360020306

Gee, G. W., Newman, B. D., Green, S. R., Meissner, R., Rupp, H., Zhang, Z. F.,... Salazar, J. (2009). Passive wick fluxmeters: Design considerations and field applications. Water Resour. Res., 45(4). https://doi.org/10.1029/2008WR007088

Gee, G. W., Ward, A. L., Caldwell, T. G., & Ritter, J. C. (2002). A vadose zone water fluxmeter with divergence control. Water Resour. Res., 38(8), 16-1-16-7. https://doi.org/10.1029/2001WR000816

Google Maps. (n.d.). Google map of study area located in NFREC, Live Oak. Retrieved from https://www.google.com/maps/search/NFREC,+LIVE+OAK/@30.3047217,-82.9072913,418m/data=!3m1!1e3?entry=ttu

Heffernan, J. B., Liebowitz, D. M., Frazer, T. K., Evans, J. M., & Cohen, M. J. (2010). Algal blooms and the nitrogen-enrichment hypothesis in Florida springs: Evidence, alternatives, and adaptive management. Ecol. Appl., 20(3), 816-829. https://doi.org/10.1890/08-1362.1

Hoogenboom, G., Porter, C. H., Shelia, V., Boote, K. J., Singh, U., Pavan, W.,... Jones, J. W. (2021). Decision Support System for Agrotechnology Transfer (DSSAT) Version 4.8.2. Gainesville, FL: DSSAT Foundation. Retrieved from www.DSSAT.net

Hu, K., White, R., Chen, D., Li, B., & Li, W. (2007). Stochastic simulation of water drainage at the field scale and its application to irrigation management. Agric. Water Manag., 89(1), 123-130. https://doi.org/10.1016/j.agwat.2006.12.010

Hyde, A. G., & Ford, R. D. (1989). Water table fluctuation in representative immokalee and Zolfo soils of Florida. Soil Sci. Soc. Am. J., 53(5), 1475-1478. https://doi.org/10.2136/sssaj1989.03615995005300050029x

Kanzari, S., Ben Nouna, B., Ben Mariem, S., & Rezig, M. (2018). Hydrus-1D model calibration and validation in various field conditions for simulating water flow and salts transport in a semi-arid region of Tunisia. Sustain. Environ. Res., 28(6), 350-356. https://doi.org/10.1016/j.serj.2018.10.001

Katz, B. G. (2004). Sources of nitrate contamination and age of water in large karstic springs of Florida. Environ. Geol., 46(6), 689-706. https://doi.org/10.1007/s00254-004-1061-9

Li, Z., Liu, H., Zhao, W., Yang, Q., Yang, R., & Liu, J. (2019). Quantification of soil water balance components based on continuous soil moisture measurement and the Richards equation in an irrigated agricultural field of a desert oasis. Hydrol. Earth Syst. Sci., 23(11), 4685-4706. https://doi.org/10.5194/hess-23-4685-2019

Mason, W. K., Meyer, W. S., Smith, R. C., & Barrs, H. D. (1983). Water balance of three irrigated crops on fine-textured soils of the Riverine Plain. Aust. J. Agric. Res., 34(2), 183-191. https://doi.org/10.1071/AR9830183

McIsaac, G. F., David, M. B., Gertner, G. Z., & Goolsby, D. A. (2001). Nitrate flux in the Mississippi River. Nature, 414(6860), 166-167.

McLean, D. C., Koeser, A. K., Shober, A. L., Qin, Z., Hasing, G., & Beeson, R. C. (2018). Incorporating woody ornamentals into residential landscapes to reduce nutrient leaching. ENH1242. Gainesville, FL: University of Florida Institute of Food and Agricultural Sciences. Retrieved from https://journals.flvc.org/edis/article/download/131616/135221/236703

Meissner, R., Rupp, H., & Seyfarth, M. (2008). Advances in out door lysimeter techniques. Water Air Soil Pollut. Focus, 8(2), 217-225. https://doi.org/10.1007/s11267-007-9166-2

Meissner, R., Rupp, H., Seeger, J., Ollesch, G., & Gee, G. W. (2010). A comparison of water flux measurements: Passive wick-samplers versus drainage lysimeters. Eur. J. Soil Sci., 61(4), 609-621. https://doi.org/10.1111/j.1365-2389.2010.01255.x

NRCS. (1985). Urban hydrology for small watersheds, Technical Release 55 (TR-55). USDA-NRCS.

NRCS. (2024). Web Soil Survey. Retrieved from https://websoilsurvey.nrcs.usda.gov/app/WebSoilSurvey.aspx

Payero, J. O., & Irmak, S. (2008). Construction, installation, and performance of two repacked weighing lysimeters. Irrig. Sci., 26(2), 191-202. https://doi.org/10.1007/s00271-007-0085-9

Prasad, R., & Hochmuth, G. J. (2016). Environmental nitrogen losses from commercial crop production systems in the Suwannee River Basin of Florida. PLoS One, 11(12), e0167558. https://doi.org/10.1371/journal.pone.0167558

Radovanovic, J., Bean, E. Z., & Reisinger, A. J. (2021). Design, construction, and installation of a drainage lysimeter for use on sandy, well-drained soils under turfgrass: AE554/AE554, 02/2021. EDIS, 2021(1). https://doi.org/10.32473/edis-ae554-2021

Rath, S., Zamora-Re, M., Graham, W., Dukes, M., & Kaplan, D. (2021). Quantifying nitrate leaching to groundwater from a corn-peanut rotation under a variety of irrigation and nutrient management practices in the Suwannee River Basin, Florida. Agric. Water Manag., 246, 106634. https://doi.org/10.1016/j.agwat.2020.106634

Riedl, A., Li, Y., Eugster, J., Buchmann, N., & Eugster, W. (2022). Technical note: High-accuracy weighing micro-lysimeter system for long-term measurements of non-rainfall water inputs to grasslands. Hydrol. Earth Syst. Sci., 26(1), 91-116. https://doi.org/10.5194/hess-26-91-2022

Ritchie, J. T. (1998). Soil water balance and plant water stress. In G. Y. Tsuji, G. Hoogenboom, & P. K. Thornton (Eds.), Understanding options for agricultural production (pp. 41-54). Dordrecht, Netherlands: Springer. https://doi.org/10.1007/978-94-017-3624-4_3

Scanlon, B. R., Christman, M., Reedy, R. C., Porro, I., Simunek, J., & Flerchinger, G. N. (2002). Intercode comparisons for simulating water balance of surficial sediments in semiarid regions. Water Resour. Res., 38(12), 59-1-59-16. https://doi.org/10.1029/2001WR001233

Schwen, A., Zimmermann, M., & Bodner, G. (2014). Vertical variations of soil hydraulic properties within two soil profiles and its relevance for soil water simulations. J. Hydrol., 516, 169-181. https://doi.org/10.1016/j.jhydrol.2014.01.042

Shelia, V., Šimunek, J., Boote, K., & Hoogenbooom, G. (2018). Coupling DSSAT and HYDRUS-1D for simulations of soil water dynamics in the soil-plant-atmosphere system. J. Hydrol. Hydromech., 66(2), 232-245. https://doi.org/10.1515/johh-2017-0055

Shukla, S., Srivastava, S., & Hardin, J. D. (2006). Design, construction, and installation of large drainage lysimeters for water quantity and quality studies. Appl. Eng. Agric., 22(4), 529-540. https://doi.org/10.13031/2013.21221

Simonne, E. H., Liu, G., & Morgant, B. (2024). Denitrification in seepage-irrigated vegetable fields in South Florida: HS1004/HS248, 1/2005. EDIS, 2005(1). https://doi.org/10.32473/edis-hs248-2005

Šimunek, J., Sejna, M., Saito, H., Sakai, M., & van Genuchten, M. T. (2008). The HYDRUS-1D software package for simulating the one-dimensional movement of water, heat, and multiple solutes in variably-saturated media version 4.0. Riverside, California: Department of Environmental Sciences, University of California.

Soldevilla-Martinez, M., Quemada, M., López-Urrea, R., Muñoz-Carpena, R., & Lizaso, J. I. (2014). Soil water balance: Comparing two simulation models of different levels of complexity with lysimeter observations. Agric. Water Manag., 139, 53-63. https://doi.org/10.1016/j.agwat.2014.03.011

Suleiman, A. A. (2008). Modeling daily soil water dynamics during vertical drainage using the incoming flow concept. CATENA, 73(3), 312-320. https://doi.org/10.1016/j.catena.2007.12.002

USEPA. (1993a). Method 353.2: Determination of nitrate-nitrite nitrogen by automated colorimetry. Revision 2.0. Cincinnati, OH: Environmental Monitoring Systems Laboratory, Office of Research And Development, U.S. Environmental Protection Agency.

USEPA. (1993b). Method 350.1: Nitrogen, ammonia (colorimetric, automated phenate) (Revision 2.0). Cincinnati, OH: Environmental Monitoring Systems Laboratory, Office of Research And Development, U.S. Environmental Protection Agency.

USEPA. (1993c). Method 351.2: Determination of total Kjeldahl nitrogen by semi-automated colorimetry (Revision 2.0). Cincinnati, OH: Environmental Monitoring Systems Laboratory, Office of Research And Development, U.S. Environmental Protection Agency.

USEPA. (2004). Method 9060A: Total organic carbon (Revision 1.0).

USEPA. (2024). Safe Drinking Water Act (SDWA). In EPA.gov. Retrieved May 13th, 2024, from https://www.epa.gov/sdwa

Weatherspoon, R. L. (2006). Soil survey of Suwannee County, Florida. United States Department of Agriculture, Natural Resources Conservation Service.

Webster, C. P., Shepherd, M. A., Goulding, K. W., & Lord, E. (1993). Comparisons of methods for measuring the leaching of mineral nitrogen from arable land. J. Soil Sci., 44(1), 49-62. https://doi.org/10.1111/j.1365-2389.1993.tb00433.x

Wohling, D. L., Leaney, F. W., & Crosbie, R. S. (2012). Deep drainage estimates using multiple linear regression with percent clay content and rainfall. Hydrol. Earth Syst. Sci., 16(2), 563-572. https://doi.org/10.5194/hess-16-563-2012

Wright, D., Marois, J., Katsvairo, T., Wiatrak, P., & Rich, J. (2005). Sod-based rotations - the next step after conservation tillage. In W. Busscher, & J. Frederick (Ed.), Proc. 27th Southern Conservation Tillage Systems Conf. (pp. 5-12). North Florida Research and Education Center, University of Florida.

Yu, H., Cui, Y., Li, S., Kang, S., Yao, Z., & Wei, Z. (2023). Estimation of the deep drainage for irrigated cropland based on satellite observations and deep neural networks. Remote Sens. Environ., 298, 113819. https://doi.org/10.1016/j.rse.2023.113819

Zamora Re, M. I., Rath, S., Dukes, M. D., & Graham, W. (2020). Water and nitrogen budget dynamics for a maize-peanut rotation in Florida. Trans. ASABE, 63(6), 2003-2020. https://doi.org/10.13031/trans.13916

Zhang, Y., Zhao, W., Ochsner, T. E., Wyatt, B. M., Liu, H., & Yang, Q. (2019). Estimating deep drainage using deep soil moisture data under young irrigated cropland in a desert-oasis ecotone, Northwest China. Vadose Zone J., 18(1), 180189. https://doi.org/10.2136/vzj2018.10.0189

Zhu, Y., Fox, R. H., & Toth, J. D. (2002). Leachate collection efficiency of zero-tension pan and passive capillary fiberglass wick lysimeters. Soil Sci. Soc. Am. J., 66(1), 37-43. https://doi.org/10.2136/sssaj2002.3700

Zotarelli, L., Scholberg, J. M., Dukes, M. D., & Muñoz-Carpena, R. (2007). Monitoring of nitrate leaching in sandy soils. J. Environ. Qual., 36(4), 953-962. https://doi.org/10.2134/jeq2006.0292