BYE, Circuit Judge.

Thomas A. Lawler, Parkersburg, IA, for Plaintiffs–Appellants.

Cheryl Schrempp DuPris, AUSA, Pierre, SD, for Defendant–Appellee.

Before SMITH, BYE, and BENTON, Circuit Judges.

BYE, Circuit Judge.

Arlen and Cindy Foster brought this action to challenge the United States Department of Agriculture's (USDA's) determination that a portion of the Fosters' farmland is a wetland within the meaning of the pertinent federal statutes and regulations. The district court granted summary judgment in favor of the USDA after concluding the agency's final decision was not arbitrary, capricious, or contrary to the law. We affirm.

I

Arlen and Cindy Foster own and farm land in Miner County, South Dakota. Miner County is located within what is generally referred to as the Prairie Pothole Region covering some of South Dakota, North Dakota, Minnesota, and parts of three Canadian provinces. The USDA uses its own nomenclature to describe various land areas within the United States; under that nomenclature larger Land Resource Regions (LRRs) are subdivided into Major Land Resource Areas (MLRAs). The Fosters' farm is located within LRR F, or the Northern Great Plains Spring Wheat Region, and more specifically within an MLRA called the Southern Black Glaciated Plains. For purposes of this appeal, the MLRA where the Fosters' farm is located is relevant for determining the types of soils found within the MLRA, which in turn is relevant for determining what types of vegetation would exist when a particular soil is in its natural state, including vegetation which would naturally be found in a wetland.

In 1985, Congress passed the Food Security Act of 1985 which contains “Swampbuster provisions authoriz[ing] the USDA to make determinations as to whether certain lands qualify as wetlands and whether wetlands that have been manipulated qualify as converted wetlands.” Clark v. United States Dept. of Agric., 537 F.3d 934, 935 (8th Cir.2008). Swampbuster was passed “[i]n order to combat the disappearance of wetlands through their conversion into crop lands[.]” Gunn v. United States Dep't of Agric., 118 F.3d 1233, 1235 (8th Cir.1997). Significantly, “a person determined to have converted wetlands may become ineligible to receive farm program payments” from the federal government. Clark, 537 F.3d at 935.

This appeal concerns a wetland determination made by the USDA affecting just under an acre (0.8 acres) of the Fosters' farm, a prairie pothole which the parties call Site 1. On June 3, 2002, Arlen Foster initially sought a wetlands determination from the Natural Resource Conservation Service (NRCS), an agency within the USDA, for a larger tract of land which included Site 1. After a number of intermittent agency proceedings not relevant to this appeal, the NRCS ultimately certified Site 1 as a wetland on June 23, 2011. The Fosters appealed the June 2011 determination to the USDA National Appeals Division (NAD), a separate agency within the USDA established to address certain claims and disputes, including wetland determinations.

In the first step of the NAD appeal, the Fosters bore the burden of proving the NRCS's determination “was erroneous by a preponderance of the evidence.” 7 C.F.R. § 11.8(e). Both the Fosters and the NRCS were permitted to present evidence and conduct cross-examination at a hearing held in October 2011. On January 10, 2012, the NAD hearing officer issued a detailed fourteen-page decision determining the NRCS followed the proper procedures and had appropriately found that Site 1 was a wetland, and that the Fosters had not met their burden of proving the NRCS's determination was erroneous. Appellant's App. at 4–17.

In the second step of the NAD appeal, the Fosters sought review of the hearing officer's decision by the NAD director's office pursuant to 7 C.F.R. § 11.9. On July 16, 2012, the NAD director's office issued a decision upholding the hearing officer's decision, which in relevant part held the NRCS proved the presence of the three controlling criteria for a wetland determination by showing that Site 1:

(a) had a predominance of hydric soils, (b) was inundated or saturated by surface or groundwater at a frequency and duration sufficient to support a prevalence of hydrophytic vegetation typically adapted for life in saturated soil conditions, and (c) under normal circumstances would support a prevalence of hydrophytic vegetation.

Appellant's App. at 33; see also 16 U.S.C. § 3801(a)(27) ; 7 C.F.R. § 12.2 (setting forth the three criteria used to determine whether a specific area of land qualifies as a wetland under federal law). The NAD director's office also held the Fosters “did not prove by a preponderance of the evidence that the [NRCS] decision was erroneous.” Appellant's App. at 34. The decision from the NAD director's office constituted the USDA's final agency decision on the matter.

In May 2013, the Fosters filed a complaint in federal district court seeking judicial review of the USDA's final agency decision. Both parties filed motions for summary judgment. In the summary judgment proceedings, the Fosters did not dispute that Site 1 contains a predominance of hydric soils (the first of the three relevant criteria), but challenged the final agency decision with regard to whether Site 1 had the requisite hydrology to qualify as a wetland and whether its soil would support a prevalence of hydrophytic vegetation under normal circumstances (the last two of the three relevant criteria).

The district court granted USDA's motion for summary judgment and denied the Fosters' motion for summary judgment. The district court concluded the NAD's factual findings were supported by substantial (at times, uncontroverted) evidence and the record supported the NAD's legal conclusions. The district court therefore determined the Fosters had failed to show the USDA's final agency decision was arbitrary, capricious, or contrary to the law. This timely appeal followed.

II

We review the district court's grant of summary judgment de novo. Doud v. Toy Box Dev. Co., 798 F.3d 709, 712 (8th Cir.2015). The issue before the district court was whether the USDA's final agency decision was proper under the Administrative Procedures Act (APA), 5 U.S.C. § 706. Under the APA, judicial review of an agency decision is limited to determining whether the agency action is “arbitrary, capricious, [ ] an abuse of discretion, or otherwise not in accordance with law.” Id. § 706(2)(A). If the agency's decision is supportable on any rational basis, the court must uphold it. Voyageurs Nat'l Park Ass'n v. Norton, 381 F.3d 759, 763 (8th Cir.2004).

A

The Fosters first contend the USDA erred in determining Site 1 had the requisite hydrology to qualify as a wetland. More specifically, the Fosters challenge the methodology the NRCS used to determine the presence of wetland hydrology at Site 1's pothole. In this case, the methodology used by the NRCS included viewing aerial photographs of the pothole when it was under normal environmental conditions. The NRCS chose to view aerial photographs because Site 1 was drier than it would have been under normal conditions at the time of the agency's on-site visit in November 2010, and the Fosters had tilled the pothole so it was not in its natural condition.

Although the Fosters generally acknowledge the legitimacy of using aerial photographs to determine whether a site has the requisite hydrology to qualify as a wetland, they contend the NRCS improperly relied upon “color tone” differences in the aerial photographs as an authorized “signature” of a wetland. The Fosters argue that relying on color tone differences as a signature of hydrology is contrary to the prescribed methodology used to make a wetland determination.

The parties agree there are ten recognized signatures the NRCS may rely upon when using aerial photographs to determine a site's hydrology, which are set forth in the South Dakota Mapping Conventions. The ten wetland (or wetness) signatures are listed on a form labeled SD–LTP–33 (Form 33) and are as follows: (1) hydrophytic vegetation; (2) surface water; (3) saturated conditions; (4) stressed crops due to wetness; (5) differences in vegetation due to different planting dates; (7) inclusion of wet areas as set-aside or idled; (8) circular or irregular areas of unharvested crops within a harvested field; (9) isolated areas that are not farmed with the rest of the field; and (10) areas of greener vegetation (especially during dry years). Admin. Record at 451. The parties also agree that if any of these signatures appear in over half of the normal rainfall year photographs, the presence of wetland hydrology is established.

In this case, Michelle Burke, an agricultural engineer employed by the NRCS, testified at the October 2011 hearing that she identified some of these signatures at the site of the pothole in seven out of ten years in which the area had normal rainfall. The Fosters did not cross-examine Burke on this testimony. Instead, the Fosters rely upon a form Burke completed to document the presence of wetlands called Form 28. Instead of using the ten authorized signatures identified in Form 33, Form 28 lists four shorthand abbreviations for those ten signatures, one of which is “CT” for “color tone” differences. When filling out Form 28, Burke used this abbreviated shorthand to document the signatures she identified on the aerial photographs of the Fosters' pothole.

The Fosters contend that checking CT on Form 28 amounts to an improper consideration of “color tone” to identify a wetland, even though “color tone” is not an authorized wetland signature. As the district court noted, however, CT is nothing more than an abbreviation used on one of the agency's forms. Burke's unchallenged testimony established that she actually identified some of the ten authorized signatures at Site 1 in the requisite number of normal rainfall years. She did not testify that she merely saw changes in “color tone” in the aerial photographs. Burke's testimony was therefore sufficient to support the agency's final decision, and the Fosters' contention is merely an attack upon an alleged deficiency in an agency form.

B

The Fosters next contend the USDA improperly determined that Site 1 would support a prevalence of hydrophytic vegetation under normal circumstances. More specifically, the Fosters claim the USDA improperly used a comparison site too far away from their farm to make its determination, that is, the agency's comparison site was outside the “local area” required by the governing regulation.

When “the vegetation on [a disputed site] has been altered or removed,” as was the case here because the Fosters had tilled the pothole located at Site 1, the pertinent federal regulation authorizes the NRCS to “determine if a prevalence of hydrophytic vegetation exists in the local area on the same hydric soil map unit under non-altered hydrologic conditions.” 7 C.F.R. § 12.31(b)(2)(ii) (emphasis added). In other words, when a disputed site is not in its natural vegetative state, the NRCS must use a comparison site in the local area which contains the same soil type as the disputed site to determine what vegetation would typically be found if the disputed site had not been altered.

In this case, the NRCS verified that Site 1 was located in a Tetonka hydric soil map unit (one of approximately twenty soil series identified by the USDA within the Southern Black Glaciated Plains MLRA), and the Fosters do not dispute that fact. The NRCS next chose an unaltered comparison site within the same Tetonka hydric soil map unit, a site located about forty miles away from Site 1 in Kingsbury County, South Dakota. The NRCS chose the Kingsbury County site for a number of other reasons, such as its inclusion on an approved list of sites established as comparison sites due to their undisturbed nature, and the fact that it was a prairie pothole similar to Site 1. In choosing the Kingsbury County site, the NRCS considered but rejected two closer alternative sites proposed by the Fosters on their own farm land (described as two grassed pasture/hayland fields that had recently been cropped, hayed, grazed and/or sprayed) neither of which were ever established by the Fosters as meeting the required regulatory criteria (i.e., located on the same hydric soil map unit as Site 1, or undisturbed and thus under non-altered hydrologic conditions).

Despite their failure to establish that their alternative sites satisfied the regulatory criteria, the Fosters contend the NRCS improperly chose its comparison site because their proposed sites were within the “local area” while the Kingsbury County site was not. We reject this contention, which is unsupported by any authority. The unchallenged testimony of Kevin Luebke, an NRCS biologist who testified at the October 2011 hearing, established that the USDA interpreted the “local area” referenced in § 12.31 to mean the same MLRA as the disputed site. Like Site 1, the Kingsbury County comparison site is also located within the Southern Black Glaciated Plains MLRA, and thus meets the regulatory criteria under the agency's reasonable interpretation, to which we owe deference. See Friends of Boundary Waters Wilderness v. Dombeck, 164 F.3d 1115, 1128 (8th Cir.1999) (requiring deference “to the informed discretion of the responsible federal agencies” when an analysis of the relevant information requires a high level of expertise); see also Downer v. United States, 97 F.3d 999, 1003–04 (8th Cir.1996), affirming Downer v. United States, 894 F.Supp. 1348, 1354, 1354 n. 7 (D.S.D.1995) (upholding the USDA's use of a comparison site which satisfied the regulatory criteria set forth in § 12.31 despite the landowner's complaint that the site was too far away (thirty miles) to be considered). We therefore conclude the agency's use of the Kingsbury County site as a comparison site was not arbitrary, capricious, or contrary to the law.

III

We affirm the district court.