swat – Livestock and Poultry Environmental Learning Community

Purpose

The dairy industry in Idaho has grown substantially over the past 30 years and is the state’s largest agricultural commodity, accounting for $3.7 billion in sales in 2022. Roughly 500,000 of Idaho’s 660,000 dairy cows reside in a six-county region known as the Magic Valley, a name originating in the early 1900s when large canal irrigation projects turned a dry landscape into verdant farmland. The Magic Valley is semi-arid, receiving around 254 mm of precipitation each year and requiring cropland to be irrigated throughout the growing season. Due to a limited amount of water available for irrigation each season cropland area has not expanded since the 1980s.

The large number of dairy cows in the Magic Valley has shifted crop production towards forage crops, predominantly silage corn and alfalfa. For example, between 1992 and 2022 the number of dairy cows in Twin Falls County increased from 18,000 to 108,000. During this same timespan corn silage and alfalfa saw a 14,000 and 5,000 hectare increases in land cover, respectively (Figure 1). This change in land cover has potentially increased consumptive water use within the region through the replacement of crops with shorter irrigation seasons (e.g., wheat and beans) with forage crops. In addition to changes in water use, the increase in dairy cattle has resulted in greatly increased manure applications to surrounding fields. It is typical for cropland to receive manure at rates of 52 Mg ha-1 year-1, which can input high amounts of nitrogen and phosphorus beyond what is removed by the crop. Over time, this could result in soil phosphorus enrichment and the leaching of nitrate to groundwater.

Figure 1. Population of dairy cows in Twin Falls County from 1992 to 2022 along with total hectares of corn silage and alfalfa.

What Did We Do?

The study area for this project was the Twin Falls Canal Company, a large irrigation project in southern Idaho. Investigation into potential changes in water quality and quantity brought about by the growing dairy agriculture in southern Idaho was carried out using the Soil and Water Assessment Tool (SWAT) model. SWAT is a physically based geospatial watershed-scale hydrologic model that incorporates climate, topography, soils, land cover, and management practice data. Model scenarios included examining changes in consumptive water use over time, effects of irrigation practices on the leaching of water and nutrients, and the impact of continuous manure applications on the buildup and leaching of nutrients. Nutrient cycling and crop nutrient uptake were calibrated in the model using two USDA-ARS eight-year studies. The first study applied manure under a corn-barley-alfalfa rotation only when soil nutrient concentrations were deficient, and the second study applied manure on a yearly basis in the spring at a rate of 52 Mg ha-1 under a barley-sugar beet-wheat-potato rotation.

Table 1. Crop areas and percentages under the 1992 and 2022 scenarios.

	1992 km2 (%)	2022 km2 (%)
Alfalfa	189 (25.3)	244 (32.8)
Barley	104 (13.9)	132 (17.7)
Beans	169 (22.7)	60 (8.0)
Corn Silage	55 (7.4)	191 (25.7)
Potatoes	35 (4.6)	34.5 (4.6)
Sugar Beets	46 (6.2)	26 (3.5)
Wheat	148 (19.8)	57 (7.6)

Table 1. Crop areas and percentages under the 1992 and 2022 scenarios.

Consumptive water use within the Twin Falls Canal Company was compared between two distinct time periods: pre-dairy and present. 1992 was selected as the pre-dairy benchmark due to being before large increases in dairy cattle numbers. Modeled crops were alfalfa, barley, beans, corn silage, potatoes, sugar beets, and wheat, which account for over 95% of irrigated cropland within the TFCC. Land cover in 2022 was used as the present scenario, and crop distributions were altered for the 1992 scenario based on USDA agricultural census data (Table 1). The model was run using climate data from 2002 to 2022 to have consistency between the two scenarios and to allow for year-to-year variability weather patterns. Automatic irrigation routines were used in the model, with a 9.1 mm irrigation event being triggered when soil water content dropped 5 mm below field capacity. 9.1 mm was chosen as the daily irrigation amount because it is roughly equivalent to the flow rate of an 850 gallon per minute center pivot. Irrigation schedules varied by crop within the April 15th – October 31st irrigation season (Table 2).

Table 2. Irrigation seasons for modeled crops.

	Irrigation Season
Alfalfa	April 15th – October 9th
Barley	April 15th – July 25th
Beans	June 26th – September 10th
Corn Silage	May 25th – September 18th
Potatoes	May 15th – September 1st
Sugar Beets	April 20th – September 25th
Wheat	April 15th – July 16th

What Have We Learned?

Modeled changes in land use within the Twin Falls Canal Company towards forage crops for dairy cattle have increased consumptive use during the year by 9% on average. June, August and September showed the greatest average increases in evapotranspiration (ET) (Figure 2). Irrigation amounts increased under the 2022 land use scenario for all months except April. Percolation under the 2022 scenario also increased to an average of 155 mm each year, up from 132 mm in the 1992 land use scenario.

Figure 2. Modeled monthly average cropland ET for the pre-dairy (1992) and post-dairy (2022) land cover scenarios.

Typical yearly water diversions for the Twin Falls Canal Company were sufficient to meet the current and future irrigation demand. Diversion reductions in August and September are common depending on reservoir storage and the timing and volume of snowmelt. A shift towards greater cropland area irrigated during those months could require deficit irrigation during extreme drought years, which are likely to become more common given climate change projections indicating reduced snowpack and earlier snowmelt runoff.

SWAT was able to reasonably represent manure nitrification, including the increases in nitrification during the year following sugar beet and potato residue being left on the field (Table 3). Crop nutrient uptake in the two USDA-ARS studies was also able to be accurately modeled after adjusting nutrient uptake parameters. Modeled soil nitrate and plant-available phosphorus concentrations were similar to field samples. Changes to SWAT source code was necessary to better partition “fast” and “slow” organic nitrogen fractions in manure between the two pools and limit mineralization when the air temperature is below 6 degrees Celsius. Under a manure application rate of 52 Mg ha-1 soil plant-available phosphorus levels exceed the allowed maximum of 40 mg kg-1 in just two years. Applying manure only when needed to satisfy crop nutrient requirements did not result in soil plant-available phosphorus approaching or exceeding the 40 mg kg-1 threshold. In addition to high soil phosphorus levels, nitrogen mineralization from yearly applications of manure resulted in high soil nitrate levels. Modeled percolation using actual irrigation amounts over the eight-year study totaled 1,176 mm and resulted in 1,256 kg ha-1 of leached nitrogen. This highlights the risk that yearly manure applications can have to water quality, especially if water is applied in excess of crop needs when also accounting for soil moisture. In addition, high variability in manure nitrogen and phosphorus concentrations suggests yearly fixed-rate applications are not the ideal for managing nutrient budgets.

Table 3. Yearly and in-season manure nitrogen mineralization from the SWAT model output compared to in-season nitrogen mineralization collected from field samples during the long-term manure study. Asterisks denote years in which sugar beet or potato residue was left on the field, resulting in greater N mineralization the following year.

Year	SWAT N Mineralization	SWAT In-Season N Mineralization	Field In-Season Mineralization
	kg ha-1	kg ha-1	kg ha-1
2013	211	117	180
2014*	287	192	110
2015	442	308	280
2016*	321	205	190
2017	399	242	250
2018*	297	197	150
2019	393	285	230
2020	357	145	150
Total	2,707	1,690	1,540

Future Plans

Now that the SWAT model has been fully calibrated, the next step will be to test various scenarios in which yearly manure application amounts, crop rotations, and irrigation schedules are adjusted. Typical regional dairy crop rotations include silage corn, alfalfa, wheat, barley, triticale, and occasionally potatoes or sugar beets. Manure is not applied to alfalfa, possibly allowing for a drawdown of phosphorus that has accumulated over previous years. Changing irrigation schedules will alter the timing and quantity of percolated water which will change nutrient export characteristics. Incorporating these scenarios over a large irrigation district with variable soils should identify areas that are more at risk of nutrient losses through runoff or leaching. Results from this research will be used to inform management agencies on the water use and water quality implications of crop rotations, manure applications, and irrigation schedules in southern Idaho.

Authors

Presenting & corresponding author

Galen I. Richards, PhD Candidate, University of Idaho, grichards@uidaho.edu

Additional authors

Erin Brooks, Professor, Department of Soil and Water Systems, University of Idaho

Linda Schott, Assistant Professor and Nutrient & Waste Management Extension Specialist, Department of Soil and Water Systems, University of Idaho

Kossi Nouwakpo, Research Soil Scientist, USDA-ARS Northwest Irrigation and Soils Research Station

Daniel Strawn, Professor, Department of Soil and Water Systems, University of Idaho

Additional Information

https://www.uidahoisaid.com/

Acknowledgements

This research was funded under the University of Idaho Sustainable Agriculture Initiative for Dairy (ISAID) grant USDA-NIFA SAS 2020-69012-31871

I would like to thank USDA-ARS researchers April Leytem, Robert Dungan, and Dave Bjorneberg at the Northwest Irrigation and Soils Research Station in Kimberly, ID for providing me with data from their long-term research studies and general assistance in accurately modeling regional agricultural practices.

The authors are solely responsible for the content of these proceedings. The technical information does not necessarily reflect the official position of the sponsoring agencies or institutions represented by planning committee members, and inclusion and distribution herein does not constitute an endorsement of views expressed by the same. Printed materials included herein are not refereed publications. Citations should appear as follows. EXAMPLE: Authors. 2025. Title of presentation. Waste to Worth. Boise, ID. April 7-11, 2025. URL of this page. Accessed on: today’s date.

Purpose

To assess and advance site assessment for P based management. This presentation is part of a symposium addressing the P Index.

What did we do?

The revision of USDA-NRCS’s standard for nutrient management coincided with significant assessment of the performance of Phosphorus (P) Indices in the six states that are tied to the Chesapeake Bay watershed. The 64,000 square mile watershed is the focus of unprecedented activity around nutrient management as a result of a 2011 Total Maximum Daily Load for P, nitrogen (N), and sediment under the Clean Water Act. In addition, the state of Maryland had required updates to it’s original P Index, resulting in broad scrutiny by various interest groups. Within this setting, USDA-NRCS funded a multi-state project to help advance the testing and harmonization of P-based management in the Chesapeake region.

States in the Chesapeake Bay region have had a long history of collaboration over P management. As a result, participants in the project decided to employ a unique approach to evaluating site assessment and P-based management on the basis of physiographic groupings. Expert panels were established for each of the major physiographic provinces in the region: (1) the Allegheny Plateau (NY, PA and WV); (2) the Appalachian Valley and Ridge (PA, MD, VA, WV) and the Appalachian Piedmont (PA, MD, VA); (3) and the Atlantic Coastal Plain (DE, MD, VA). Panels were comprised with representatives from the nutrient management planning community, the conservation community, the action agency community (state and federal), the farming community and researchers.

At the beginning of the project, each physiographic panel was charged with helping to identify the primary P management concerns for a province and the range of conditions that are most appropriate for site assessment. As the project evolves, panels will be expected to provide feedback on results, look for consistencies and differences between state approaches to site assessment, and, for states working to modify they site assessment approaches, offer input on suitable options.

What have we learned?

Modeling P runoff plays a fundamental role in the Chesapeake project. For each physiographic region, watersheds were identified where the SWAT model and the local P Index would be run to compare results. This objective has proven challenging to the SWAT model, as it requires a common scale of spatial inference with the P Index, i.e., the field. At the start of the project, the performance of SWAT was evaluated with regard to its representation of hydrologic and P cycling processes. It was determined that a version of SWAT that better represents variable source area hydrology, TOPO-SWAT, is best suited to the uplands of the Chesapeake Bay region. In addition, it was determined that the original P routines in SWAT are insensitive to the key nutrient management factors considered by the P Index (rate, timing, method and form of P application). Therefore, a new set of P routines was developed to correct the problem.

To-date, SWAT has been applied to watersheds in all of the physiographic regions of the Chesapeake Bay watershed: Factory Brook (NY) and Dressler Run (PA) in the Allegheny Plateau; Spring Creek (PA), Mahantango Creek (PA), Conewago Creek (PA), Upper Potomac (VA) in the Appalachian Mountains (Valley and Ridge as well as Piedmont). On the Atlantic Coastal Plain, where subsurface P losses to artificial drainage dominate, existing models were deemed inadequate. Therefore, an empirical approach to describing P transfers has been used that includes the application of geophysical imagining techniques to describe the flow paths connecting fields with drainage ditches.

Future Plans

Next steps of the project are to begin comparing P Index output with SWAT output for select field conditions under an array of management scenarios. At that point it is expected the physiographic panels will be re-engaged to work through the assessment of P Index performance and generate insight into the potential for regionally consistent approaches to P-based management.

Authors

Peter Kleinman, Research Leader, USDA-ARS Pasture Systems and Watershed Management Research Unit peter.kleinman@ars.usda.gov

D. Beegle, Z. Easton, A. Collick, Q. Ketterings, D. Fuka, T. Veith, A. Shober, S. Cela, M. Reiter, A. Allen, J. Liu and T. Basden

Additional information

Information on this project can be obtained by contacting the corresponding author.

Acknowledgements

This project is funded, in part, by a USDA-NRCS Conservation Innovation Grant.

The authors are solely responsible for the content of these proceedings. The technical information does not necessarily reflect the official position of the sponsoring agencies or institutions represented by planning committee members, and inclusion and distribution herein does not constitute an endorsement of views expressed by the same. Printed materials included herein are not refereed publications. Citations should appear as follows. EXAMPLE: Authors. 2015. Title of presentation. Waste to Worth: Spreading Science and Solutions. Seattle, WA. March 31-April 3, 2015. URL of this page. Accessed on: today’s date.