Transport of Nutrients by Overland Flow on Sites Containing Beef Cattle Manure

Gilley, John E.; McGehee, Ryan P.; Wacha, Kenneth M.

doi:10.1061/JOEEDU.EEENG-7838

Open access

Technical Papers

Jan 11, 2025

Transport of Nutrients by Overland Flow on Sites Containing Beef Cattle Manure

Authors: John E. Gilley, Ph.D., P.E. https://orcid.org/0000-0002-8167-5362 [email protected], Ryan P. McGehee, Ph.D. https://orcid.org/0000-0003-0464-9774 [email protected], and Kenneth M. Wacha, Ph.D. [email protected]Author Affiliations

Publication: Journal of Environmental Engineering

Volume 151, Issue 3

https://doi.org/10.1061/JOEEDU.EEENG-7838

PDF

Abstract

In an inductive metastudy, we investigated the effects of varying rates of overland sheet flow on the transport of phosphorus (P) and nitrogen (N) from sites containing beef cattle manure. The analyzed data were collected during previously reported field rainfall simulation investigations conducted on cropland sites and a beef cattle feedlot in the US state of Nebraska. During the experiments, inflow was incrementally added to the top of the experimental plots to simulate runoff rates occurring at greater downslope distances. Runoff rates on the experimental plots ranged from 2.9 to

22.9 L \min^{- 1}

, and the maximum equivalent downslope distances varied from 5.3 to 42.3 m. The findings revealed that P and N transport rates increased linearly with runoff rate on sites where N was applied at rates

\leq 151 kg {ha}^{- 1}

(Condition 1). Nutrient transport rates were influenced by the quantity of nutrients released to overland flow and the amount of runoff available to transport these nutrients. Interestingly, on sites where N was added at rates

> 151 kg {ha}^{- 1}

(Condition 2) and on beef cattle feedlots (Condition 3), maximum transport rates for P were not affected by runoff rate, but N transport rates again increased linearly with runoff rate. The following hypotheses were formulated: Hypothesis 1 states that P transport rates can be directly related to runoff rates for Condition 1; Hypothesis 2 states that for Condition 2, the maximum rate at which beef cattle manure can release P to overland flow can be estimated from the P content of the applied manure; and Hypothesis 3 states that transport rates for total N can be directly related to runoff rates for Conditions 1, 2, and 3. Hypothesis testing using the student’s

t - test

affirmed each hypothesis. However, further assessment of these hypotheses is needed at other field sites with varying soil, cropping, and management conditions.

Practical Applications

The land application of manure to cropland areas has the dual benefit of reducing fertilizer expenses and improving soil health. However, it is essential to consider water quality concerns that may arise if nutrients in manure are transported by runoff to receiving water bodies. In this investigation, rainfall simulation data were used to identify the effects of varying runoff rates on nutrient transport from sites containing beef cattle manure. The study revealed the following findings: (1) P transport rates increase linearly with runoff rate on sites where beef cattle manure was applied at N application rates

\leq 151 kg {ha}^{- 1}

, which is approximately the annual N requirement for corn; (2) P transport rates do not vary with runoff rate when N is added at rates

> 151 kg {ha}^{- 1}

; (3) the maximum P transport rate is directly related to the P content of the applied manure; and (4) N transport rates increase linearly with runoff rate on sites containing beef cattle manure. These findings suggest that if nutrient transport rates can be related to runoff rates, rainfall simulation data obtained from small plots could potentially be extrapolated to greater downslope distances.

Introduction

Nutrient accumulation from excess fertilization and land application has become an environmental issue in many watersheds (Chen et al. 2018). A surplus of phosphorus (P) and nitrogen (N) in soil may eventually reach aquatic ecosystems (Shi and Huang 2021). Overenrichment of nutrients may seriously degrade water bodies and result in eutrophication (Sharpley et al. 2017). Selected management practices are available to reduce the transport of nutrients to surface waters (Kamrath and Yuan 2023). Currently, these practices are assessed and implemented based on either in situ monitoring data, which is uncommon, or hydrologic models, which are based on relatively limited empirical representations of complex physical, chemical, and biological processes (Hollaway et al. 2018). Best management practices and technologies could be better implemented if the factors influencing nutrient transport on upland areas could be identified and quantified. Although additional understanding of all overland flow factors would be ideal, this paper is focused on overland flow, which is a precursor to shallow concentrated flow and channel flow in the hydrologic cycle.

The US National Phosphorus Research Project was initiated to identify management practices to reduce the transport of P from watersheds (Macrae et al. 2024). Several federal and state government agencies and universities have participated in this project. Rainfall simulation tests have been used to develop indexing tools to assess and rank site vulnerability to P loss (Flaten et al. 2024). However, integration of information collected from relatively small field plots into a unified nutrient management decision-making process applicable at a field scale has not yet been achieved (Kleinman et al. 2015).

Small-scale rainfall simulators have been reported to be of limited use in predicting nutrient transport on upland areas (Nash et al. 2021). Larger plots that avoid concentrated flow along the borders can be difficult to establish (Verbree et al. 2010). To address some of the inherent limitations associated with rainfall simulators, researchers have introduced inflow at the top of experimental plots to simulate conditions occurring at greater flow rates (Elliot and Flanagan 2023; Robichaud et al. 2010). By adding inflow, nutrient transport data can potentially be extrapolated from these small plots and used to estimate nutrient contributions along the hillslope gradient. Lastly, existing mathematical procedures for routing overland flow along hillslopes could be modified to include nutrient constituents, providing a more comprehensive assessment of nutrient transport.

Many of the current generation of agricultural models assume that soluble nutrient transport in surface runoff is linked to a constant ratio between the concentration of pollutants in the runoff and the concentration in the uppermost soil layer. However, studies revealed that the enrichment ratio (ER) is highly dynamic and can vary significantly due to storm events, soil properties, and management practices (Papanicolaou et al. 2015). Despite this knowledge, widely applied modeling frameworks such as the Soil and Water Assessment Tool (SWAT) model (Abbaspour et al. 2015) and the enhanced water quality version of the process-based Water Erosion Prediction Project (WEPP) model (McGehee et al. 2023a, b) continue to assume fixed ER values. This oversimplification could lead to significant overpredictions or underpredictions of nutrient loss and may even mask the benefits of certain conservation practices.

In response, we propose an alternative formulation based on simple, empirical, first-order relations between nutrient transport rates and runoff rates. Using data from previously reported small-scale rainfall simulation experiments conducted on sites containing beef cattle manure, we analyzed P and N transport as affected by varying runoff rates. By incrementally introducing inflow at the top of experimental plots, we scaled up runoff rates, making them applicable to greater downslope distances. Our findings suggest that if nutrient transport rates can be explicitly related to runoff rates, data obtained from small plots could potentially serve as predictors of P and N losses on upland areas. The objective of this study was to identify the effects of varying runoff rates on nutrient transport by overland flow on sites containing beef cattle manure.

Methods

Study Site Characteristics

Data from three rainfall simulation investigations conducted following the land application of beef cattle manure were examined in the present investigation (Table 1). The cropland sites were located at the University of Nebraska Rogers Memorial Farm, located 18 km east of Lincoln, Nebraska. The investigation by Thayer et al. (2012) was the only study we found where nutrient transport/flow rate data was available on a site where beef cattle manure had been applied at a N application rates

\leq 151 kg {ha}^{- 1}

, which is approximately the annual N requirement for corn. We felt that analysis at two sites where cattle manure was added at a rates

> 151 kg {ha}^{- 1}

would suffice so we also examined data obtained by Gilley et al. (2008, 2011).

Table 1. Studies where the effects of varying runoff rates on DP, TP,

{NO}_{3} - N

, and TN transport rates were measured on sites containing beef cattle manure

Condition (observation)	References	Nutrient content of beef cattle manure or beef cattle feedlot	Nutrient loads ( $g {ha}^{- 1} \min^{- 1}$ )	Runoff rates ( $L \min^{- 1}$ )	Equivalent downslope distance (m)	Transport equation and nutrient transport ( $g {ha}^{- 1} \min^{- 1}$ ), runoff rate ( $L \min^{- 1}$ )	$R^{2}$
Surface application of cattle manure at N rate $\leq 151 kg {ha}^{- 1}$ (Observations 1 and 2)	Thayer et al. (2012)	Total available P in $manure = 0 - 103 kg {ha}^{- 1}$ and total available N in $manure = 0 - 151 kg {ha}^{- 1}$	$DP = 2.5 - 5.2$ , $TP = 7.4 - 27.2$ , ${NO}_{3} - N = 1.7 - 9.0$ , and $TN = 35 - 127$	4.4–22.9	7.7–39.8	$DP = 0.268$ runoff rate, $TP = 1.22$ runoff rate, ${NO}_{3} - N = 0.413$ runoff rate, and $TN = 6.87$ runoff rate	0.941, 0.996, 0.989, and 0.942
Surface application of cattle manure at N rate of $204 kg {ha}^{- 1}$ (Observations 3 and 4)	Gilley et al. (2008)	Total available P in $manure = 191 kg {ha}^{- 1}$ and total available N in $manure = 204 kg {ha}^{- 1}$	$DP = 41$ (constant), $TP = 110$ (constant), ${NO}_{3} - N = 613 - 2,400$ , and $TN = 670 - 2, 460$	2.9–16.5	5.3–30.0	${NO}_{3} - N = 152$ runoff rate and $TN = 157$ runoff rate	0.994 and 0.992
Surface application of cattle manure at N rate $> 151 kg {ha}^{- 1}$ (Observations 5 and 6)	Gilley et al. (2011)	Total available P in $manure = 92 - 367 kg {ha}^{- 1}$ , and total available N in $manure = 151 - 604 kg {ha}^{- 1}$	$DP = 22$ (constant), $TP = 28$ (constant), ${NO}_{3} - N = 7.5 - 11.0$ , and $TN = 83 - 143$	4.0–11.1	15.4–42.3	${NO}_{3} - N = 1.03$ runoff rate and $TN = 13.8$ runoff rate	0.959 and 0.982
Beef cattle feedlot (Observations 7 and 8)	Gilley et al. (2010)	Bray- $1 P = 777 mg {kg}^{- 1}$ , $WSP = 170 mg {kg}^{- 1}$ , and $TN = 13.3 g {kg}^{- 1}$	$DP = 76$ (constant), $TP = 113$ (constant), ${NO}_{3} - N = 90 - 392$ , and $TN = 300$ –646	4.4–17.2	7.7–29.9	${NO}_{3} - N = 22.4$ runoff rate and $TN = 41.7$ runoff rate	0.998 and 0.982
Beef cattle feedlot (Observations 9 and 10)	Gilley et al. (2012)	Bray- $1 P = 662 mg {kg}^{- 1}$ , $WSP = 161 mg {kg}^{- 1}$ , and $TN = 15.1 g {kg}^{- 1}$	$DP = 105$ (constant), $TP = 127$ (constant), ${NO}_{3} - N = 143 - 510$ , and $TN = 392 - 767$	5.0–15.3	10.0–30.6	${NO}_{3} - N = 32.9$ runoff rate and $TN = 55.7$ runoff rate	0.998 and 0.982

Note: DP = dissolved P; TP = total P; TN = total N; WSP = water soluble P; and $R^{2}$ = coefficient of determination.

Nutrient transport from a beef cattle feedlot was thought to represent an upper limit on nutrient transport. Therefore, data collected by Gilley et al. (2010, 2012) during two rainfall simulation tests performed within beef cattle feedlot pens located at the US Meat Animal Research Center near Clay Center, Nebraska, were also examined (Table 1). Overland flow drained uniformly from the experimental plots, which were in upslope pen locations.

Abstracts from each of the five studies from which data were examined are provided in the Supplemental Materials. Also included is information on study site characteristics, rainfall simulation procedures, addition of inflow, and feedlot soil and runoff sample collection. The rainfall simulation equipment (Fig. S1) and inflow tests (Fig. S2) are shown. Finally, figures developed during hypothesis testing are also included in the Supplemental Materials (Figs. S3–S6).

The plots examined in this investigation were 0.75 m wide and either 2.0 or 4.0 m long. Simulated overland flow was applied at the upslope end of each plot in four successive increments after the third simulation run, and rainfall application continued at a rate of approximately

70 mm h^{- 1}

. There was a concern that the hydrologic processes occurring at larger flow rates could not be adequately represented using 2.0- or 4.0-m-long plots. Therefore, a flow rate that did not cause transport of residue materials on the cropland sites or rilling on the feedlot pens was selected as the largest flow rate for a given rainfall simulation investigation. The other three inflow quantities were then chosen to represent an intermediate range of flow rates.

Statistical Analyses

Data from replicated treatments where beef cattle manure was applied to three cropland sites and two beef cattle feedlots were examined in this investigation. Mean nutrient transport rates were first related to runoff rates using linear regression analysis [Eqs. (1)–(12)] (Figs. 1–5). The goodness of fit of the linear regression models was identified using the coefficient of determination,

R^{2}

(Table 1). Dissolved P (DP), total P (TP),

{NO}_{3} - N

, and total N (TN) transport rates were also related to runoff rates on individual plots using linear regression analysis [Eqs. (13), (14), and (21)] [Figs. S3 and S6(a)]. Nutrient transport rates obtained from the regression relationships were then compared with measured values [Eqs. (15), (16), and (22)] [Figs. S4 and S6(b)]. The hypothesis that the regression coefficient of the predicted versus measured nutrient transport values equals one at the 95% confidence level was evaluated using the student’s

t - test

(Table 2).

Fig. 1. Transport rates for (a) dissolved phosphorus (DP) and total phosphorus (TP) (Observation 1); and (b) nitrate nitrogen ( ${NO}_{3} - N$ ) and total nitrogen (TN) (Observation 2) versus runoff rates for the study performed by Thayer et al. (2012).

Fig. 2. Transport rates for (a) DP and TP (Observation 3); and (b) ${NO}_{3} - N$ and TN (Observation 4) versus runoff rates for the study performed by Gilley et al. (2008).

Fig. 3. Transport rates for (a) DP and TP (Observation 5); and (b) ${NO}_{3} - N$ and TN (Observation 6) versus runoff rates for the study performed by Gilley et al. (2011).

Fig. 4. Transport rates for (a) DP and TP (Observation 7); and (b) ${NO}_{3} - N$ and TN (Observation 8) versus runoff rates for the study performed by Gilley et al. (2010).

Fig. 5. Transport rates for (a) DP and TP (Observation 9); and (b) ${NO}_{3} - N$ and TN (Observation 10) versus runoff rates for the study performed by Gilley et al. (2012).

Table 2. Linear regression analyses of predicted versus measured DP, MDP, MTP, TP, and TN transport rates (

g {ha}^{- 1} \min^{- 1}

) determined for selected studies

References	Regression equation	Coefficient of determination, $R^{2}$	$F$	$β_{1}$		$β_{0}$
References	Regression equation	Coefficient of determination, $R^{2}$	$F$	Student’s $t - test$	$P - value$	Student’s $t - test$	$P - value$
Thayer et al. (2012)	$DP = 1.23 x - 0.021$	0.645	23.6	4.85	$3.2 \times 10^{- 4}$	$- 0.028$	0.97
Thayer et al. (2012)	$TP = 1.07 x + 2.60$	0.700	32.0	5.66	$5.9 \times 10^{- 5}$	0.856	0.41
Gilley et al. (2011)	$MDP = 0.804 x + 1.97$	0.986	290	17.0	$7.0 \times 10^{- 5}$	0.994	0.38
Gilley et al. (2011)	$MTP = 0.806 x + 1.60$	0.982	224	15.0	$1.2 \times 10^{- 4}$	0.582	0.59
Gilley et al. (2011)	$TN = 0.728 x + 46.9$	0.681	46.9	6.85	$7.1 \times 10^{- 7}$	2.61	0.02

Note: MDP = maximum dissolved P; MTP = maximum total P; $R^{2}$ = coefficient of determination; $F = F$ ratio; $β_{1}$ = slope of the regression line; and $β_{0}$ = intercept of the regression line.

Mean maximum transport rates for DP (MDP) and TP (MTP) were determined for selected plots having variable amounts of total available P (TAP) (Table 3). Linear regression equations relating MDP and MTP transport rates to TAP in manure were next identified [Eqs. (17) and (18)] (Fig. 6). MDP and MTP transport rates on additional plots were estimated using the regression equations, and the predicted values were compared with field measurements [Eqs. (19) and (20)] (Fig. S5). Finally, the hypothesis that the regression coefficients equal one at the 95% confidence level was evaluated using the student’s

t - test

(Table 2).

Table 3. MDP and MTP transport rates for plots containing selected amounts of TAP in the study performed by Gilley et al. (2011)

Plot	Transport rate ( $g {ha}^{- 1} \min^{- 1}$ )	TAP ( $kg {ha}^{- 1}$ )
MDP
37	$MDP = 67.8$	367
39	$MDP = 16.2$	92
41	$MDP = 16.5$	92
43	$MDP = 40.9$	184
45	$MDP = 0.93$	0
47	$MDP = 67.8$	367
MTP
37	$MTP = 83.5$	367
39	$MTP = 18.9$	92
41	$MTP = 18.6$	92
43	$MTP = 48.4$	184
45	$MTP = 2.84$	0
47	$MTP = 80.7$	367

Note: TAP = total available phosphorus.

Fig. 6. MDP and MTP transport rates versus TAP for Plots 37, 39, 41, 43, 45, and 47 of the study performed by Gilley et al. (2011).

Experimental Treatments and Conditions

Transport of P and N Following Manure N Application $\leq 151 kg {ha}^{- 1}$

Thayer et al. (2012) evaluated the effects of varying manure application rates on P loads following the addition of beef cattle manure at N application rates

\leq 151 kg {ha}^{- 1}

. The quantity of total available P and N that were applied varied from 0 to 103 and 0 to

151 kg {ha}^{- 1}

, respectively. In the present study, data were only evaluated on those plots where manure was not incorporated:

•

Observation 1: P transport rates increased in a linear fashion with the runoff rate [Fig. 1(a)]. The regression equations that were obtained relating DP and TP transport rates (

g {ha}^{- 1} \min^{- 1}

) to the runoff rate (

L \min^{- 1}

) are

DP = 0.268 runoff rate R^{2} = 0.941

(1)

TP = 1.22 runoff rate R^{2} = 0.996

(2)

•

Observation 2: Transport rates for

{NO}_{3} - N

and TN also increased in a linear fashion with the runoff rate [Fig. 1(b)]. The regression equations that were identified relating

{NO}_{3} - N

and TN transport rates (g

{ha}^{- 1} \min^{- 1}

) to the runoff rate (L

\min^{- 1}

) are

{NO}_{3} - N = 0.413 runoff rate R^{2} = 0.989

(3)

TN = 6.87 runoff rate R^{2} = 0.942

(4)

Transport of P and N Following the Addition of Manure at N Application Rates ${> 151 kg ha}^{- 1}$

Gilley et al. (2008) measured the effects of varying runoff rates on P transport following the application of total available P and N at rates of 191 and

204 kg {ha}^{- 1}

, respectively. The plots were either disked or maintained in a no-till condition following manure application:

•

Observation 3: The transport rates of DP and TP were similar at the three largest inflow rates and averaged 41 and

110 g {ha}^{- 1} \min^{- 1}

[Fig. 2(a)]. Because the transport rates for DP and TP were comparable at the largest inflow rates, trendlines were not included in Fig. 2(a).

•

Observation 4: The transport rates for

{NO}_{3} - N

and TN increased in a linear fashion with runoff rate [Fig. 2(b)]. The regression equations that were obtained relating

{NO}_{3} - N

and TN transport rates (

g {ha}^{- 1} \min^{- 1}

) to the flow rate (

L \min^{- 1}

) are

{NO}_{3} - N = 152 runoff rate R^{2} = 0.994

(5)

TN = 157 runoff rate R^{2} = 0.992

(6)

The effects of varying manure application rates on P loads following the addition of manure at N application rates

> 151 kg {ha}^{- 1}

were identified by Gilley et al. (2011). The quantity of total available P and N that was applied varied from 92 to 367 and 151 to

604 kg {ha}^{- 1}

, respectively.

•

Observation 5: The transport rates of DP and of TP were similar at the four inflow rates and averaged 22 and

28 g {ha}^{- 1} \min^{- 1}

[Fig. 3(a)] (Table 1). Trendlines were not included in Fig. 3(a) because the transport rates for DP and TP were comparable for each inflow rate.

•

Observation 6: The transport rates of

{NO}_{3} - N

and TN increased in a linear fashion with flow rate [Fig. 3(b)]. The regression equations that were obtained relating

{NO}_{3} - N

and TN transport rates (

g {ha}^{- 1} \min^{- 1}

) to the flow rate (

L \min^{- 1}

) are

{NO}_{3} - N = 1.03 runoff rate R^{2} = 0.959

(7)

TN = 13.8 runoff rate R^{2} = 0.982

(8)

Transport of P and N from Beef Cattle Feedlots

P transport in runoff from feedlot surfaces as affected by corn-based and wet distillers grain diets was examined by Gilley et al. (2010). The mean Bray-1 P, water-soluble P (WSP), and TN of material obtained from the 0–1.5-cm depth on the feedlot surface was 777, 170, and

13,300 mg {kg}^{- 1}

, respectively:

•

Observation 7: The transport rates of DP and of TP were similar at the three largest inflow rates and averaged 76 and

113 g {ha}^{- 1} \min^{- 1}

[Fig. 4(a)]. Because the transport rates for DP and TP were comparable at the largest inflow rates, trendlines were not included in Fig. 4(a).

•

Observation 8: The transport rates for

{NO}_{3} - N

and TN increased in a linear fashion with flow rate [Fig. 4(b)]. The regression equations that were identified relating

{NO}_{3} - N

and TN transport rates (

g {ha}^{- 1} \min^{- 1}

) to the flow rate (L

\min^{- 1}

) are

{NO}_{3} - N = 22.4 runoff rate R^{2} = 0.999

(9)

TN = 41.7 runoff rate R^{2} = 0.976

(10)

Runoff P losses from feedlot surfaces containing varying amounts of unconsolidated surface materials were examined by Gilley et al. (2012). The mean Bray-1 P, WSP, and TN of material obtained from the 0–1.5-cm depth on the feedlot surface was 662, 161, and

15,100 mg {kg}^{- 1}

, respectively.

•

Observation 9: The transport rates of DP and TP were similar at the four inflow rates and averaged 105 and

127 g {ha}^{- 1} \min^{- 1}

[Fig. 5(a)]. Trendlines were not included in Fig. 5(a) because the transport rates for DP and TP were comparable for each of the inflow rates.

•

Observation 10: The transport rates for

{NO}_{3} - N

and TN increased in a linear fashion with runoff rates [Fig. 5(b)]. The regression equations that were obtained relating

{NO}_{3} - N

and TN transport rates (

g {ha}^{- 1} \min^{- 1}

) to the flow rate (

L \min^{- 1}

) are

{NO}_{3} - N = 32.9 runoff rate R^{2} = 0.998

(11)

TN = 55.7 runoff rate R^{2} = 0.982

(12)

Results

Hypotheses

The observations that nutrient transport rates can be related to runoff rates on sites containing beef cattle manure resulted in three hypotheses:

•

Hypothesis 1 (Observation 1): The P transport rate can be related in a linear fashion to runoff rate on sites where manure is added at N application rates

\leq 151 kg {ha}^{- 1}

.

•

Hypothesis 2 (Observations 3, 5, 7, and 9): On sites where manure is added at N application rates

> 151 kg {ha}^{- 1}

, the maximum rate at which beef cattle manure can release P to overland flow can be estimated from the total P content of the applied manure.

•

Hypothesis 3 (Observations 2, 4, 6, 8, and 10): Transport rates for TN can be related in a linear fashion to runoff rate on sites containing beef cattle manure.

Testing of Hypothesis 1

Data used to test this hypothesis were obtained from the study conducted by Thayer et al. (2012). It was observed on the plots where manure had been added to meet the 4-year P-based requirements for corn that the maximum rate at which manure can release P to overland flow had been reached and the P transport rate was constant at varying runoff rates. Therefore, only the plots on which manure was applied to meet less than the 4-year P-based crop requirements were used in the subsequent analyses.

P transport rates (

g {ha}^{- 1} \min^{- 1}

) were first related to runoff rates (

L \min^{- 1}

) on selected sites (Plots 1, 4, 19, and 21) using linear regression analysis. The regression equations obtained for DP and TP transport rates using data combined from each of the four plots (Fig. S3) were

DP = 0.237 runoff rate R^{2} = 0.709

(13)

TP = 1.27 runoff rate R^{2} = 0.937

(14)

These equations were then used to relate predicted DP and TP transport rates obtained at varying runoff rates to measured DP and TP transport rates for additional sites (Plots 8, 10, 16, and 23). The resulting equations (Table 2 and Fig. S4) are

DP (predicted) = 1.23 DP (measured) - 0.021 R^{2} = 0.645

(15)

TP (predicted) = 1.07 TP (measured) + 2.60 R^{2} = 0.700

(16)

The student’s

t - test

was used to evaluate the hypotheses that the regression coefficients equal one at the 95% confidence level. The regression coefficients were found to be not significantly different from one. Analyses of the experimental data collected by Thayer et al. (2012) suggest that P transport rates can be estimated from measurements of runoff rates on sites where manure is added at N application rates

\leq 151 kg {ha}^{- 1}

and, thus, Hypothesis 1 is affirmed.

Testing of Hypothesis 2

This hypothesis was tested using data from a study performed by Gilley et al. (2011). MDP MTP were first determined for selected sites (Plots 37, 39, 41, 43, 45, and 47) (Table 3). Linear regression equations relating MDP and MTP transport rates to total applied P (TAP) in manure were next identified (Fig. 6) as follows:

MDP = 0.184 (TAP) + 1.17 R^{2} = 0.990

(17)

MTP = 0.222 (TAP) + 1.36 R^{2} = 0.990

(18)

The regression equations were then used to predict MDP and MTP transport rates on additional sites (Plots 38, 40, 42, 44, 46, and 48). The resulting equations (Table 2 and Fig. S5) are

MDP (predicted) = 0.804 MDP (measured) + 1.97 R^{2} = 0.986

(19)

MTP (predicted) = 0.806 MTP (measured) + 1.60 R^{2} = 0.983

(20)

The student’s

t - test

was also used to examine the hypotheses that the regression coefficients equal one at the 95% confidence level. The regression coefficients were determined to be not significantly different from one. Analyses of the experimental data collected by Gilley et al. (2011) suggest that on sites where beef cattle manure is added at N application rates

> 151 kg {ha}^{- 1}

, the maximum rate at which manure can release P to overland flow can be estimated from the total P content of the manure and, therefore, Hypothesis 2 is affirmed.

Testing of Hypothesis 3

Hypothesis 3 was that transport rates for TN can be related in a linear fashion to runoff rates. Data to test this prediction were obtained from a previously described study conducted by Gilley et al. (2011) to measure nutrient transport from plots on which selected amounts of beef cattle manure had been applied to a cropland site. The following regression equation was first derived relating combined TN transport rates (

g {ha}^{- 1} \min^{- 1}

) to runoff rates (

L \min^{- 1}

) on selected sites (Plots 37, 39, 41, 43, 45, and 47) (Fig. S6):

TN = 15.5 runoff rate R^{2} = 0.950

(21)

The regression equation was then used to relate predicted TN transport rates obtained at varying runoff rates to measured TN transport values on additional sites (Plots 38, 40, 42, 44, 46, and 48) (Fig. S6). The resulting equation (Table 2) is

TN (predicted) = 0.728 TN (measured) + 46.9 R^{2} = 0.681

(22)

The hypothesis that the regression coefficient equals one at the 95% confidence level was evaluated using the student’s

t - test

. The regression coefficient was found to be not significantly different from one (Table 2). Analyses of the experimental data collected by Gilley et al. (2011) suggest that the TN transport rates can be estimated from measurements of runoff rates on sites containing beef cattle manure, and, therefore, Hypothesis 3 is affirmed.

Discussion

The key observations of the experimental studies are shown in Fig. 7. It was found (Observation 1) that P transport rates can be related in a linear fashion to runoff rates on sites where manure is added at N application rates

\leq 151 kg {ha}^{- 1}

. It was also determined (Observations 2, 4, 6, 8, and 10) that transport rates for TN can be related in a linear fashion to runoff rates on sites containing beef cattle manure. The constraint for these two conditions may have been the amount of overland flow available to transport nutrients. P and N transport rates were influenced by the quantity of nutrients released to overland flow and the amount of runoff available to transport these nutrients.

Fig. 7. Relationships between runoff ( $Q_{out}$ ) and transport rates of DP, TP, ${NO}_{3} - N$ , and TN for sites with (a) high manure N applications $> 151 kg {ha}^{- 1}$ ; and (b) low manure N applications $\leq 151 kg {ha}^{- 1}$ .

Nutrient transport by overland sheet flow on sites containing swine slurry was examined by Gilley (2024). In contrast to beef cattle manure, which contains very little water, swine slurry is predominately liquid. Nutrients present in swine slurry are readily transported by overland flow. Both P and N transport rates on sites containing swine slurry were found to increase in a linear fashion with runoff rate.

It was also found that on sites where manure was added at N application rates

> 151 kg {ha}^{- 1}

(Observation 3) and on a beef cattle feedlot (Observation 7), P transport rates appeared to vary in a linear fashion with runoff rate at smaller runoff rates. As flow rates increased, the P transport rate–runoff rate relation changed from a linear equation to a constant value. For Observations 3, 5, 7, and 9, the maximum rate at which soil or feedlot surface material could release P to overland flow was reached (point of inflection for P transport), and increasing runoff rates did not influence P transport rates.

The following procedure was used to estimate the point of inflection (POI) for DP and TP transport in the study conducted by Gilley et al. (2008). The nutrient transport coefficient in the linear equation relating nutrient transport rate to runoff rate was first identified using the origin and data from the initial runoff point (

9.8 g {ha}^{- 1} L^{- 1}

for DP and

20.7 g {ha}^{- 1} L^{- 1}

for TP). At POI, nutrient loads for DP and TP were 41 and

110 g {ha}^{- 1} \min^{- 1}

. These nutrient load values were then divided by the previously determined nutrient transport coefficients to determine the flow rates at the POI which for DP and TP were 4.2 and

5.3 L \min^{- 1}

, respectively. A mean overland flow rate of

1.1 L \min^{- 1}

was measured without the addition of simulated overland flow on these 2-m-long plots. Thus, the POI for DP and TP was estimated to be 7.6 and 9.7 m, respectively. The POI for DP and TP in the study performed by Gilley et al. (2010) were found in a similar manner to be 8.9 and 7.5 m, respectively. Procedures will need to be identified for estimating the point of inflection for P transport under varying manure management conditions.

The quantity of N released to overland flow may also reach an upper limit (point of inflection for N transport) at flow rates much larger than those examined in the present investigation. Under this condition, N transport rates become constant as runoff rates continue to increase. The maximum rate of N release by soil or feedlot surface material to overland flow serves as the constraint for this situation.

Temporal changes in nutrient transport following the application of beef cattle manure to a cropland site were examined by Gilley et al. (2007). Substantial reductions in the transport of nutrients were measured during the year following manure addition. The smallest nutrient concentrations usually occurred on the final sampling date. Temporal changes in nutrient transport must be identified to accurately estimate annual nutrient delivery from land application areas.

P fluxes are influenced by several factors including hydraulics, hydrology, geomorphology, and land management (Sharpley et al. 2013). Subsurface hydrologic pathways including interflow were not addressed in the present investigation. Therefore, the observations made, hypotheses developed, and predictions advanced are only applicable for overland flow conditions.

It may be possible to extrapolate the experimental results obtained from previous rainfall simulation experiments conducted on small plots to larger slope lengths if generalized nutrient transport rate–runoff rate relationships are shown to be accurate. The effects of varying soil, cropping, and management conditions on nutrient transport by overland flow could then be estimated using data obtained in other rainfall simulation investigations including the US National Research Project described by Osmond et al. (2024).

Conclusions

In this study, we analyzed previously collected rainfall simulation data obtained from either 2- or 4-m-long plots located on cropland areas following manure application and from beef cattle feedlots located in southeast Nebraska. During these investigations, inflow was incrementally added to the top of experimental plots to simulate runoff rates occurring at greater downslope distances. The runoff rates on the experimental sites ranged from 2.9 to

22.9 L \min^{- 1}

, and equivalent downslope distances varied from 5.3 to 42.3 m. Our findings revealed that P transport rates increased linearly with runoff rates when beef cattle manure was applied at N application rates

\leq 151 kg {ha}^{- 1}

, which is approximately the 1-year N requirement for corn.

The P transport rates for this condition were influenced by two factors: (1) the quantity of P released by manure at a particular runoff rate; and (2) the amount of overland flow available to transport the released P. Interestingly, when beef cattle manure was applied at rates

> 151 kg {ha}^{- 1}

and within beef cattle feedlots, P transport rates remained similar at larger runoff rates. The maximum rate at which manure can release P to overland flow was reached, resulting in an approximately constant P transport rate. We estimated this maximum P transport rate based on the total P content of the applied manure.

Additionally, N transport rates increased linearly with runoff rate on sites where beef cattle manure was applied and within beef cattle feedlots. The N transport rates for this condition were again influenced by two factors: (1) the quantity of N released by manure at a particular runoff rate; and (2) the amount of overland flow available to transport the released N. However, further testing of the predictions is necessary on additional locations with varying soil types and cropping systems. If nutrient transport rates can be linked to runoff rates, it may be possible to extrapolate the results for P and N delivery obtained on small plots to greater downslope distances. Nutrient constituents could then be incorporated into existing process-based models used to route overland flow along a hillslope.

Supplemental Materials

File (supplemental_materials_joeedu.eeeng-7838_gilley.pdf)

Download
1.74 MB

Data Availability Statement

Some or all data, models, or code that support the findings of this study are available from the corresponding author upon reasonable request.

References

Abbaspour, K. C., E. Rouholahnejad, S. Vaghefi, R. Srinivasan, H. Yang, and B. Klove. 2015. “A continental-scale hydrology and water quality model for Europe: Calibration and uncertainty of a high-resolution large-scale SWAT model.” J. Hydrol. 524 (May): 733–752. https://doi.org/10.1016/j.jhydrol.2015.03.027.

Google Scholar

Chen, D., H. Shen, M. Hu, J. Wang, Y. Zhang, and R. A. Dahlgren. 2018. “Chapter Five—Legacy nutrient dynamics at the watershed scale: Principles, modeling, and implications.” Adv. Agron. 149 (Jan): 237–313. https://doi.org/10.1016/bs.agron.2018.01.005.

Google Scholar

Elliot, W. J., and D. C. Flanagan. 2023. “Estimating WEPP cropland erodibility values from soil properties.” J. ASABE 66 (2): 329–351. https://doi.org/10.13031/ja.15218.

Google Scholar

Flaten, D. N., P. J. A. Kleinman, and D. L. Osmond. 2024. “Balancing agriculture and environment: Andrew Sharpley’s nutrient, soil, and water management legacy.” J. Environ. Qual. 2014 (Feb): 1–7. https://doi.org/10.1002/jeq2.20547.

Google Scholar

Gilley, J. E. 2024. “Nutrient transport by overland sheet flow on sites containing swine slurry.” Int. Soil Water Conserv. Res. https://doi.org/10.1016/j.iswcr.2024.04.001.

Google Scholar

Gilley, J. E., L. M. Durso, R. A. Eigenberg, D. B. Marx, and B. L. Woodbury. 2011. “Narrow grass hedge control of nutrient loads following variable manure applications.” Trans. ASABE 54 (3): 847–855. https://doi.org/10.13031/2013.37110.

Google Scholar

Gilley, J. E., B. Eghball, and D. D. Marx. 2007. “Nutrient concentrations of runoff during the year following manure application.” Trans. ASABE 50 (6): 1987–1999. https://doi.org/10.13031/2013.24102.

Google Scholar

Gilley, J. E., W. F. Sabatka, B. Eghball, and D. B. Marx. 2008. “Nutrient transport as affected by rate of overland flow.” Trans. ASABE 51 (4): 1287–1293. https://doi.org/10.13031/2013.25245.

Google Scholar

Gilley, J. E., J. R. Vogel, E. D. Berry, R. A. Eigenberg, D. B. Marx, and B. L. Woodbury. 2010. “Nutrient transport in runoff from feedlots as affected by wet distillers grain diet.” Trans. ASABE 53 (2): 545–552. https://doi.org/10.13031/2013.29569.

Google Scholar

Gilley, J. E., J. R. Vogel, R. A. Eigenberg, D. B. Marx, and B. L. Woodbury. 2012. “Nutrient losses in runoff from feedlot surfaces as affected by unconsolidated surface materials.” J. Soil Water Conserv. 67 (3): 211–217. https://doi.org/10.2489/jswc.67.3.211.

Google Scholar

Hollaway, M. J., et al. 2018. “The challenges of modeling phosphorus in a headwater catchment: Applying a ‘limits acceptability’ uncertainty framework to a water quality model.” J. Hydrol. 558 (Mar): 607–624. https://doi.org/10.1016/j.jhydrol.2018.01.063.

Google Scholar

Kamrath, B., and Y. Yuan. 2023. “Effectiveness of nutrient management for reducing phosphorus losses from agricultural areas.” J. Nat. Resour. Agric. Ecosyst. 1 (2): 77–88. https://doi.org/10.13031/jnrae.15572.

Google Scholar

Kleinman, P. J. A., A. N. Sharpley, P. J. A. Withers, L. Bergstrom, L. T. Johnson, and D. G. Doody. 2015. “Implementing agricultural phosphorus science and management to combat eutrophication.” Supplement, Ambio 44 (S2): 297–310. https://doi.org/10.1007/s13280-015-0631-2.

Google Scholar

Macrae, M. L., P. J. A. Kleinman, D. Osmond, A. Shober, and N. Nelson. 2024. “The importance of consensus science to managing phosphorus in the environment: SERA 17 and the legacy of Andrew Sharpley.” J. Environ. Qual. 2024 (Feb): 1–9. https://doi.org/10.1002/jeq2.20546.

Google Scholar

McGehee, R. P., D. C. Flanagan, and B. A. Engel. 2023a. “A WEPP-Water Quality model for simulating nonpoint source pollutants in nonuniform agricultural hillslopes: Model development and sensitivity.” Int. Soil Water Conserv. Res. 11 (3): 455–469. https://doi.org/10.1016/j.iswcr.2023.02.002.

Google Scholar

McGehee, R. P., D. C. Flanagan, B. A. Engel, and J. E. Gilley. 2023b. “A validation of WEPP water quality routines in uniform and nonuniform agricultural hillslopes.” Int. Soil Water Conserv. Res. 12 (3): 487–505. https://doi.org/10.1016/j.iswcr.2023.11.005.

Google Scholar

Nash, D., A. Weatherley, P. Kleinman, and A. Sharpley. 2021. “Estimating dissolved phosphorus losses from legacy sources in pastures: The limits of soil tests and small-scale rainfall simulators.” J. Environ. Qual. 50 (5): 1042–1062. https://doi.org/10.1002/jeq2.20265.

Google Scholar

Osmond, D. L., P. J. Kleinman, F. Coale, N. O. Nelson, C. H. Bolster, and J. McGrath. 2024. “A short history of the phosphorus index and Andrew Sharpley’s contributions from inception through development and implementation.” J. Environ. Qual. 2023 (Dec): 1–10. https://doi.org/10.1002/jeq2.20535.

Google Scholar

Papanicolaou, A. N., K. M. Wacha, B. K. Abban, C. G. Wilson, J. L. Hatfield, C. O. Stanier, and T. R. Filley. 2015. “From soilscapes to landscapes: A landscape-oriented approach to simulate soil organic carbon dynamics in intensively managed landscapes.” J. Geophys. Res. Biogeosci. 120 (11): 2375–2401. https://doi.org/10.1002/2015JG003078.

Google Scholar

Robichaud, P. R., J. W. Wagenbrenner, and R. E. Brown. 2010. “Rill erosion in natural and disturbed forests: 1. Measurements.” Water Resour. Res. 46 (10). https://doi.org/10.1029/2009WR008314.

Google Scholar

Sharpley, A., et al. 2017. “Evaluation of phosphorus site assessment tools: Lessons from the USA.” J. Environ. Qual. 46 (6): 1250–1256. https://doi.org/10.2134/jeq2016.11.0427.

Google Scholar

Sharpley, A., H. Jarvie, A. Buda, L. May, B. Spears, and P. Kleinman. 2013. “Phosphorus legacy: Overcoming the effects of past management practices to mitigate future water quality impairment.” J. Environ. Qual. 42 (5): 1308–1326. https://doi.org/10.2134/jeq2013.03.0098.

Google Scholar

Shi, W., and M. Huang. 2021. “Predictions of soil and nutrient losses using a modified SWAT model in a large hilly-gully watershed of the Chinese Loess Plateau.” Int. Soil Water Conserv. Res. 9 (2): 291–304. https://doi.org/10.1016/j.iswcr.2020.12.002.

Google Scholar

Thayer, C. A., J. E. Gilley, L. M. Durso, and D. B. Marx. 2012. “Wheat strip effects on nutrient loads following variable manure applications.” Trans. ASABE 55 (2): 439–449. https://doi.org/10.13031/2013.41393.

Google Scholar

Verbree, D., S. Duiker, and P. J. A. Kleinman. 2010. “Runoff losses of sediment an phosphorus from no-till and cultivated soils receiving dairy manure.” J. Environ. Qual. 39 (5): 1762–1770. https://doi.org/10.2134/jeq2010.0032.

Google Scholar

Information & Authors

Information

Published In

Journal of Environmental Engineering

Volume 151 • Issue 3 • March 2025

Copyright

This work is made available under the terms of the Creative Commons Attribution 4.0 International license, https://creativecommons.org/licenses/by/4.0/.

History

Received: May 17, 2024

Accepted: Sep 16, 2024

Published online: Jan 11, 2025

Published in print: Mar 1, 2025

Discussion open until: Jun 11, 2025

Authors

Affiliations

John E. Gilley, Ph.D., P.E. https://orcid.org/0000-0002-8167-5362 [email protected]

Research Agricultural Engineer, Agroecosystems Management Research Unit, USDA — Agricultural Research Service, 3720 East Campus Loop South, Lincoln, NE 68583 (corresponding author). ORCID: https://orcid.org/0000-0002-8167-5362. Email: [email protected]

View all articles by this author

Ryan P. McGehee, Ph.D. https://orcid.org/0000-0003-0464-9774 [email protected]

Assistant Professor, Agricultural and Biosystems Engineering, Iowa State Univ., Ames, IA 50011. ORCID: https://orcid.org/0000-0003-0464-9774. Email: [email protected]

View all articles by this author

Kenneth M. Wacha, Ph.D. [email protected]

Research Hydrologist, National Laboratory for Agriculture and the Environment, USDA — Agricultural Research Service, 1015 North University Blvd., Ames, IA 50011. Email: [email protected]

View all articles by this author

Metrics & Citations

Metrics

Citations

Download citation

If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download.

Transport of Nutrients by Overland Flow on Sites Containing Beef Cattle Manure

Abstract

Practical Applications

Introduction

Methods

Study Site Characteristics

Statistical Analyses

Experimental Treatments and Conditions

Transport of P and N Following Manure N Application $\leq 151 kg {ha}^{- 1}$

Transport of P and N Following the Addition of Manure at N Application Rates ${> 151 kg ha}^{- 1}$

Transport of P and N from Beef Cattle Feedlots

Results

Hypotheses

Testing of Hypothesis 1

Testing of Hypothesis 2

Testing of Hypothesis 3

Discussion

Conclusions

Supplemental Materials

Data Availability Statement

References

Information & Authors

Information

Published In

Copyright

History

Authors

Affiliations

Metrics & Citations

Metrics

Citations

Download citation

View Options

Figures

Tables

Media

PREVIOUS ARTICLE

Verify Phone

Congrats!

Abstract

Practical Applications

Introduction

Methods

Study Site Characteristics

Statistical Analyses

Experimental Treatments and Conditions

Transport of P and N Following Manure N Application ≤151 kg ha−1

Transport of P and N Following the Addition of Manure at N Application Rates >151 kg ha−1

Transport of P and N from Beef Cattle Feedlots

Results

Hypotheses

Testing of Hypothesis 1

Testing of Hypothesis 2

Testing of Hypothesis 3

Discussion

Conclusions

Supplemental Materials

Data Availability Statement

References

Information

Published In

Copyright

History

Authors

Affiliations

Metrics

Citations

Download citation

Share

Copy the content Link

Share with email

Share

Request Username

Create a new account

Change Password

Password Changed Successfully

Verify Phone

Congrats!

Transport of P and N Following Manure N Application $\leq 151 kg {ha}^{- 1}$

Transport of P and N Following the Addition of Manure at N Application Rates ${> 151 kg ha}^{- 1}$