James G. THOMAS; Kerry Thomas; Gregory Norlin; Nancy Norlin; Sierra Club, Plaintiffs-Appellants, v. Lisa P. JACKSON, in her capacity as Administrator of the Environmental Protection Agency; United States Environmental Protection Agency; William W. Rice, in his capacity as Acting Regional Administrator of Region VII of the Environmental Protection Agency, Defendants-Appellees, 581 F.3d 658

MELLOY, Circuit Judge. Plaintiffs filed a complaint in district court under the Administrative Procedure Act (the “APA”), arguing that approval by the Environmental Protection Agency (the “EPA”) of the State of Iowa’s 2004 “ § 303(

d) lists” violated several aspects of the Clean Water Act (the “CWA”). The district court dismissed the complaint. We affirm. I. A. Statutory and Regulatory Background The CWA requires each state to establish water quality standards for bodies of water within the state’s boundaries. 33 U.S.C. § 1313(a)-(c). To do so, a state first designates the use or uses of a particular body of water (e.g., water supply, recreation), see 40 C.F.R. § 131.10, and then designates the water quality criteria necessary to protect that designated use, id. § 131.11.

The water quality standards comprise: (1) designated uses; (2) water quality criteria defining the amounts of pollutants that the water can contain without impairment of the designated uses; and (3) anti-degradation requirements, which apply to bodies of water whose quality is better than required. 33 U.S.C. § 1313(c)(2)(A); 40 C.F.R. §§ 131.6, 131.10-12.

Section 303(

d) of the CWA requires that each state, after establishing its water quality standards, compile a list of waters, a Ҥ 303(

d) list,” that do not meet those standards. 33 U.S.C. § 1313(d). In creating its § 303(

d) list, a state must “assemble and evaluate all existing and readily available water quality-related data and information.” 40 C.F.R. § 130.7(b)(5). The relevant data and information include the state’s “ § 305(

b) report” and its “§ 319 report.” Id. § 130.7(b)(5)®, (iv). A § 305(

b) report is a water quality assessment report regarding all navigable waters within the state that each state must submit to the EPA pursuant to CWA § 305(b). 33 U.S.C. § 1315(b). The EPA compiles, analyzes, and transmits these § 305(

b) reports to Congress. Id. § 1315(b)(2). A § 319 report is a water quality assessment report regarding the navigable waters within the state that will fail to meet water quality standards without limitation of nonpoint sources of pollution. Id. § 1329(a). Along with the § 303(

d) list, the state must submit, inter alia: (1) “[a] description of the methodology used to develop the list”; (2) “[a] description of the data and information used to identify waters”; (3) “[a] rationale for any decision not to use any existing and readily available data and information” for certain categories of water; and (4) “[a]ny other reasonable information requested by the Regional Administrator.” 40 C.F.R. § 130.7(b)(6). Since 1992, states have been required to submit a § 303(

d) list to the EPA every two years; states were not required, however, to submit lists in 2000. See Revision to the Water Quality Planning and Management Regulation Listing Requirements, 65 Fed. Reg. 17,170 (Mar. 31, 2000). The EPA must approve or disapprove the state’s § 303(

d) list within thirty days of its submission. Id. § 130.7(d)(2). If the EPA disapproves a § 303(

d) list, the EPA must establish its own list for the state within thirty days of the date of disapproval. Id. For each water on the § 303(

d) list, the state must establish total maximum daily loads (each a “TMDL”) of certain “pollutants” that the water can sustain without exceeding water quality standards. 33 U.S.C. § 1313(d)(1)(C); 33 U.S.C. § 1362(6) (defining “pollutant”). The TMDL includes pollutants from both “point sources” and “nonpoint sources,” as well as a margin of safety. 40 C.F.R § 130.2(g)-(i). The EPA must review the state’s TMDLs, and if the EPA disapproves any particular TMDL, the EPA itself must establish the TMDL. 33 U.S.C. § 1313(d)(2). Starting with the states’ 2002 reporting requirements, the EPA recommended that each state submit an “integrated report” that comprised the state’s § 305(

b) report and § 303(

d) list. Robert H. Wayland III, 2002 Integrated Water Quality Monitory and Assessment Report Guidance (Nov. 19, 2001) (“2002 Report Guidance ”). EPA guidance recommended that the integrated report designate bodies of water as belonging to one of five categories, which the EPA specified as: 1. Attaining the water quality standard and no use is threatened. 2. Attaining some of the designated uses; no use is threatened; and insufficient or no data and information is available to determine if the remaining uses are attained or threatened. 3.

Insufficient or no data and information to determine if any designated use is attained. 4. Impaired or threatened for one or more designated uses but does not require the development of a TMDL. A. TMDL has been completed. B. Other pollution control requirements are reasonably expected to result in the attainment of the water quality standard in the near future. C. Impairment is not caused by a pollutant. 5. The water quality standard is not attained. The [water] is impaired or threatened for one or more designated uses by a pollutant(s), and requires a TMDL. 2002 Report Guidance.

The EPA slightly revised and repeated this recommendation for 2004 reports. See EPA, Guidance for 200Jp Assessment and Reporting Requirements Pursuant to Sections § 303(

d) and 305(

b) of the CWA (July 21, 2003) (“200Jp Report Guidance ”). Because Category 4C would not be considered a part of the § 303(

d) list, EPA guidance advised that waters that were impaired, but not impaired by a pollutant, should be omitted from the states’ 303(

d) lists. See id. As discussed below, Plaintiffs contend that the CWA does not allow such a distinction and that waters that are impaired, but not impaired by a pollutant, must be included on Iowa’s § 303(

d) list. B. Factual and Procedural Background In 1999, the Sierra Club filed a complaint alleging that the EPA had not carried out its duty to approve or disapprove Iowa’s 1998 § 303(

d) list. In 2001, a consent decree arising from that complaint provided that the EPA had approved part of the § 303(

d) list, disapproved part, and created a final § 303(

d) list. The consent decree required that TMDLs be established over a ten-year period pursuant to a

schedule agreed by the parties. In evaluating compliance with the TMDL schedule, the EPA could consider TMDLs established by Iowa and approved by the EPA, TMDLs established by the EPA, and impaired waters and pollutants the EPA determined do not need TMDLs. The consent decree, which is not directly at issue here, contemplated that water bodies might be removed from future § 303(

d) lists as provided by the CWA and EPA regulations. In 2000, Iowa passed its Credible Data Law, Iowa Code § 455B.171, .193-.195, which limited the data Iowa considers when creating its draft § 303(

d) list to data Iowa considers credible. Id. § 455B.194.1.C. The Credible Data Law provides, inter alia, that Iowa will presume data over five years old not to be credible. Id. § 455B.171. The EPA informed Iowa that the restrictions of the new law were inconsistent with the CWA’s mandate to evaluate all data and that the EPA would be reviewing Iowa’s list without applying the Credible Data Law. Iowa, nevertheless, followed its Credible Data Law when preparing its § 303(

d) lists. On December 17, 2002, Iowa submitted its draft 2002 list to the EPA, along with a

summary of comments it had received on the draft list and a description of the methodology used to compile the list. Iowa’s draft 2002 list omitted a large number of waters that had been on the final 1998 § 303(

d) list. Iowa’s rationale for omitting many of the waters was the insufficiency of data, a rationale Iowa considered a flaw in the original, 1998 analysis. Iowa omitted other waters because new data indicated the cause of the impairment was flow or habitat alteration, which were not considered pollutants. After reviewing Iowa’s 2002 § 303(

d) list, the EPA approved the list in part and disapproved the list in part. The EPA approved the omissions of seventy-one waters, having concluded that Iowa provided sufficient rationale for the exclusions. The EPA disapproved, however, the omissions of twenty waters. After a public comment period, the EPA added eighteen of the twenty waters to the final 2002 § 303(

d) list. On May 26, 2005, Iowa submitted its draft 2004 list. Waters were removed for reasons including new data showing lack of impairment and the establishment of a TMDL. The EPA again approved the list in part and disapproved the list in part. The EPA determined that fourteen new waters should be added to the 2004 list and that six waters included on the 2002 list but not included on the draft 2004 list should not have been removed. After a public comment period, the EPA added seventeen of the twenty waters to the 2004 list. As a result, Iowa’s 2004 § 303(

d) list does not include a number of waters that were included on Iowa’s 1998 § 303(

d) list. Because the rationales supporting omission of numerous waters from the 2004 list were originally forwarded as reasons for delisting in 2002, we refer to the final 2002 and 2004 lists together as the § 303(

d) list. On August 10, 2006, Plaintiffs filed a complaint under the APA challenging the EPA’s approval of Iowa’s § 303(

d) list, seeking a declaratory judgment and an order to set aside the EPA’s approval of the § 303(

d) list. The EPA would then be required, Plaintiffs argue, to prepare a § 303(

d) list that complies with the law. After considering the parties’ briefs on the merits, the district court dismissed the complaint. II. “We review de novo a district court’s decision whether an agency’s action violates the APA.” Niobrara River Ranch, L.L.C. v. Huber, 373 F.3d 881, 884 (8th Cir.2004). “Under the APA, our review of an agency decision is limited.” Voyageurs Nat’l Park Ass’n v. Norton, 381 F.3d 759, 763 (8th Cir.2004).

After a searching and careful review of the record, we may set aside the EPA’s action if it is “ ‘arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law.’ ” Friends of Richards-Gebaur Airport v. FAA, 251 F.3d 1178, 1185 (8th Cir.2001) (quoting 5 U.S.C. § 706(2)(A)). “Because Congress has not directly spoken to the precise question” before us, we “look first to the agency regulations, which are entitled to deference if they resolve the ambiguity in a reasonable manner.” Coeur Alaska, Inc. v.

Southeast Alaska Conservation Council, — U.S. -, 129 S.Ct. 2458, 2469, 174 L.Ed.2d 193 (2009) (internal quotations omitted). Where the regulations are also ambiguous, “we next turn to the [agency’s] subsequent

interpretation of those regulations.” Id. Generally, we must “defer to the agency’s

interpretation so long as it is not arbitrary, capricious, an abuse of discretion, or otherwise not supported by law.” Friends of the Boundary Waters Wilderness v. Dombeck, 164 F.3d 1115, 1121 (8th Cir.1999) (internal quotation omitted); see also Chevron USA, Inc. v. Natural Res. Def. Council, 467 U.S. 837, 842-44, 104 S.Ct. 2778, 81 L.Ed.2d 694 (1984). A. Failure to Consider Important Facts and Information Plaintiffs contend that the EPA’s partial approval of Iowa’s § 303(

d) list was arbitrary and capricious. Plaintiffs assert that Iowa failed to consider all relevant data, including its § 305(

b) report and its § 319 list, and that the EPA’s approval of Iowa’s failure to consider the data indicates that the EPA failed to consider important information when approving the § 303(

d) list. The record does not support Plaintiffs’ assertions. Plaintiffs most forcefully argue that waters shown in the § 305(

b) report as not meeting their uses, only partially meeting their uses, or with threatened uses should necessarily be included on the § 303(

d) list. Although, generally, Iowa must “assemble and evaluate” data on each water included in the § 305(

b) report when approving or creating the § 303(

d) list, 40 C.F.R. § 130.7(b)(5)(i), the regulations do not require Iowa to include all § 305(

b) waters on its § 303(

d) list. Instead, Iowa must provide documentation to support its determination not to list a particular water. Id. § 130.7(b)(6). Here, Iowa provided a rationale for each § 305(

b) water that it did not include on its draft § 303(

d) list. The EPA’s ultimate agreement with Iowa’s proposed conclusion as to a particular water is insufficient, without more, to support a conclusion that the EPA failed to consider important facts or information in reaching its conclusion. Plaintiffs have not pointed to any additional evidence, so their argument fails. Plaintiffs also argue that Iowa improperly omitted numerous bodies of water that had been included on the 1998 list because Iowa found the original data insufficient to maintain a listing. Although Iowa is not required to include all § 305(

b) waters on its § 303(

d) list, the EPA may — and here, did — require Iowa to show good cause for not including a water. Id. § 130.7(b)(6)(iv). “Good cause includes, but is not limited to, more recent or accurate data; more sophisticated water quality modeling; flaws in the original analysis that led to the water being listed in the categories in § 130.7(b)(5); or changes in conditions.” Id. Iowa submitted new data for several delisted waters, but the State delisted many others merely because it now found the data supporting the 1998 listing insufficient.

Plaintiffs argue that “good cause” requires a showing of “specific data” supporting the rationale. Iowa, however, asserted that even without “more recent or accurate data” indicating to the contrary, the insufficiency of the existing data shows “flaws in the original analysis.” The EPA accepted this rationale as good cause. Plaintiffs, for their part, offer no explanation or support as to why additional data are required to conclude that the existing data are insufficient. We must first consider the EPA’s

interpretation of “flaws in the original analysis.” Id. § 130.7(b)(iv). As stated above, we owe significant deference to the EPA’s

interpretation of its own regulation. Moreover, we note that the EPA “may” request a showing of good cause, but the regulations do not require it to do so. Further, because good cause “includes, but is not limited to ... flaws in the original analysis,” id. (emphasis added), the EPA could still determine that insufficiency of the original data was “good cause” even if not strictly a “flaw[ ] in the original analysis.” Id. The optional nature of the good-cause request and the nonrestrictive definition of good cause reinforces the deference owed to the EPA’s allegedly expansive

interpretation of what constitutes such a flaw in the analysis. Accordingly, we find no difficulty accepting the EPA’s reasonable

interpretation of “flaws in the analysis” as including the flaw of basing the analysis on insufficient data. We must next consider the EPA’s application of this

interpretation to the waters at issue in the present case. We recognize Plaintiffs’ concern as to declining standards of analysis and enforcement, but such concerns do not necessarily preclude a determination that a particular collection or type of data is insufficient to require inclusion on a § 303(

d) list. Notwithstanding the EPA’s earlier determinations regarding the same data, Plaintiffs have not shown that the EPA’s more recent determinations are “implausible” or without a “rational connection” to the facts. Whether Iowa has shown “good cause” for its determination not to include § 305(

b) waters on the § 303(

d) list is a question for which deference to the EPA’s judgment requires that we affirm the district court. We defer to both the EPA’s

interpretation of its own regulation and its determination regarding the specific waters at issue. B. Waters Not Impaired by “Pollutants” Plaintiffs also claim that the EPA erred by allowing Iowa to omit from its § 303(

d) list waters that were impaired but not impaired by any “pollutant,” as that term is defined by CWA § 502(6). 33 U.S.C. 1362(6). We will set aside the EPA’s approval of the § 303(

d) list if it is “not in accordance with the law.” 5 U.S.C. § 706(A)(2). Plaintiffs contend that the CWA requires all impaired waters, including those not impaired by a pollutant, to be included on the § 303(

d) list. The EPA disagrees, contending that the CWA does not require it to include waters that are impaired but not impaired by any pollutant. The CWA does not expressly limit the § 303(

d) list to waters impaired by pollutants. Rather, § 303(

d) provides: “Each State shall identify those waters within its boundaries for which the effluent limitations ... are not stringent enough to implement any water quality standard applicable to such waters.” 33 U.S.C. § 1313(d)(1)(A). The Ninth Circuit has addressed the

interpretation of § 303(

d) in several cases, most recently in Pronsolino v. Nastri, 291 F.3d 1123, 1136-37 (9th Cir.2002) (holding that § 303(

d) lists must include waters impaired only by nonpoint source pollution). That circuit framed the “precise statutory question” before it as “whether the phrase ‘are not stringent enough’ triggers the identification requirement both for waters as to which the effluent limitations apply but do not suffice to attain water quality standards and for waters as to which effluent limitations do not apply at all to the pollution sources impairing the water.” Id. at 1126. We consider the same statutory language here. In an earlier case, the Ninth Circuit “read § 303(d)(1)(

A) as applying to all waters in the state, not only to the subset covered by certain kinds of effluent controls. It understood ‘not stringent enough’ to mean ‘not adequate for’ or ‘inapplicable to.’ ” Pronsolino, 291 F.3d at 1137 (discussing Dioxin/Organochlorine Ctr. v. Clarke, 57 F.3d 1517 (9th Cir.1995)). In Dioxin, the Ninth Circuit held that “the [effluent] limitations ... referred to in § 1313(

d) are not applicable to toxic pollutants; thus any limitations required by those provisions of § 1311, as a matter of law, ‘are not stringent enough’ to achieve established water quality standards.” Dioxin, 57 F.3d at 1528. In Pronsolino, the Ninth Circuit relied on its prior holding to conclude that effluent limitations “are ... ‘not stringent enough’ to achieve applicable water quality standards for other waters not subject to those requirements.” Pronsolino, 291 F.3d at 1137.

Similarly, we could conclude that because effluent limitations are not applicable to non-pollutant impairment, those limitations are “not stringent enough” to achieve established water quality standards. Although we might find such a reading to be reasonable, the phrase “not stringent enough” is sufficiently ambiguous to permit competing

interpretations. See id. at 1135 n. 14. (discussing different meanings of “stringent” and “stringent enough”). The Ninth Circuit also noted that a more inclusive § 303(

d) list supported the general purpose of the CWA: [T]here is no such distinction with regard to the basic purpose for which the § 303(

d) list and the TMDLs are compiled, the eventual attainment of state-defined water quality standards. Water quality standards reflect a state’s designated uses for a water body and do not depend in any way upon the source of pollution .... § 303(

d) is structurally part of a set of provisions governing an interrelated goal-setting, information-gathering, and planning process that, unlike many other aspects of the CWA, applies without regard to the source of pollution. Id. at 1137-38. Still, it is undisputed that the primary, direct purpose of § 303(

d) is to address pollutants, and the arguments supporting the Ninth Circuit’s opinion in Pronsolino are less persuasive as applied to the present dispute. In Pronsolino, the Ninth Circuit noted that when determining TMDLs both point sources and nonpoint sources must be considered. Clearly, the amount of nonpoint source pollutant would directly affect the amount of point source pollutant a water could satisfactorily sustain. Here, in contrast, Plaintiffs have made no showing that the impairments at issue would have more than an indirect effect, if any, on the level of pollutant that would be allowed under a TMDL.

While it may be that addressing broader “impaired by pollution” concerns would help address the problem of bringing defined pollutants within set limits, evaluating this indirect relationship requires the EPA’s specialized knowledge. We also note that in both Ninth Circuit cases, the court addressed the issue in the context of what the EPA was authorized to do, not what it was mandated to do. While the EPA may be authorized to require the inclusion of the waters now at issue on Iowa’s § 303(

d) list, the EPA is not mandated to include these waters. See id. at 1141 (holding “that the EPA did not exceed its statutory authority”); cf. Natural Res. Def. Council v. U.S. E.P.A., 915 F.2d 1314, 1322 n. 9 (9th Cir.1990) (“Those waters for which limitation based on the more demanding best available technology ... did not have to be listed [under § 303(d)].”); Dioxin, 57 F.3d at 1528 n. 14 (“Footnote 9 in Natural Resources does, however, suggest that neither the states nor the EPA are obligated under § 1313(d)(1)(

A) to list waters impaired by toxic pollutants despite the proven failure of [Best Available Technology] limitations.” (emphasis added)). Although § 303(

d) may allow the EPA to include all impaired waters on a state’s § 303(

d) list, it does not require the EPA to include impaired waters where the EPA has determined the impairment is due to something other than a pollutant. C. Iowa’s Methodology Plaintiffs also contend that the EPA erred by inappropriately addressing several aspects of Iowa’s listing methodology. Plaintiffs first argue that the implementation of Iowa’s Credible Data Law is, in effect, a modification of water quality standards and that the EPA cannot approve the § 303(

d) list until it has first approved the modified water quality standards. The EPA agrees that the Credible Data Law does not comply with the CWA but maintains that the Iowa law does not constitute a change in water quality standards and that because the EPA did not rely on the Credible Data Law when approving the § 303(

d) list, the approval was permissible. We disagree with Plaintiffs’ contention that when a state uses a non-compliant variation of approved water quality standards in drafting its § 303(

d) list, the ultimate, EPA-approved list must be invalidated in its entirety even if the EPA applied the correct water quality standards rather than the non-compliant variation. See Sierra Club, Inc. v. Leavitt, 488 F.3d 904, 912 n. 14 (11th Cir.2007) (rejecting a similar argument where the court concluded that the record showed the EPA applied the approved standards rather than the challenged variation); see also 40 C.F.R. § 130.7(d)(2) (establishing a time frame for the EPA to develop § 303(

d) list if it disapproves the state’s draft list). However, if the EPA were to itself apply the unapproved state modification, the resulting § 303(

d) list would be undermined. Florida Public Interest Research Group Citizen Lobby v. EPA 386 F.3d 1070, 1090-91 (11th Cir.2004) (“FPIRGCL ”) (remanding for additional review where the court concluded that the record showed that the EPA applied the challenged variation rather than the standards as approved). Here, the record shows that the EPA repeatedly informed Iowa that the use of the Credible Data Law violated the CWA; that the EPA informed Iowa that the EPA would review Iowa’s § 303(

d) list “in accordance with existing federal regulations” rather than in accordance with the Credible Data Law; that the EPA requested that Iowa supply the EPA with all data excluded from consideration under the Credible Data Law; and that Iowa did provide additional data in response to the EPA’s request. The EPA then added several waters to the § 303(

d) list based on its review of this additional information. The only evidence in the record that the EPA impermissibly relied on the Credible Data Law is an inference resulting from the fact that the EPA decided not to add the majority of the waters back to Iowa’s § 303(

d) list. This case is thus distinguishable from FPIRGCL, where the EPA adopted the state methodology to the extent the EPA deemed the methodology reasonable. See 386 F.3d at 1078-79 & n. 10 (“[T]he EPA found the [state’s] exclusion of data older than 7.5 years to be ‘reasonable,’ and accordingly adopted the same time frame in its own review.”).

On the record before us, the fact that the EPA ultimately agreed with Iowa’s conclusions concerning the majority of waters is insufficient to meet Plaintiffs’ burden to show that the EPA, after requiring the state to submit the excluded data, then adopted Iowa’s methodology of excluding that data from evaluation. Second, Plaintiffs contend that the EPA impermissibly accepted Iowa’s distinction between “evaluated” and “monitored” waters and its position that waters evaluated, but not monitored, need not be placed on the § 303(

d) list. The EPA maintains that it reviewed all “evaluated” waters consistent with federal regulations, ignoring Iowa’s distinction. As with the Credible Data Law, Plaintiffs’ only evidence that the EPA adopted Iowa’s distinction is the fact that the EPA only added four of the “evaluated” waters back to the § 303(

d) list. Plaintiffs argue that “[w]e must assume” that the EPA approved of Iowa’s distinction. The burden of proof lies with Plaintiffs, however, and Plaintiffs cannot meet that burden through unsupported assumptions. Finally, Plaintiffs contend that even relying on the approved water quality standards was in error because Iowa was in the process of revising its water quality standards. The EPA counters that applicable water quality standards remain in effect until the new standards are approved.

See 40 C.F.K.. § 131.21(e) (“A State or authorized Tribe’s applicable water quality standard for purposes of the Act remains the applicable standard until EPA approves a change, deletion, or addition to that water quality standard, or until EPA promulgates a more stringent water quality standard.”). As Plaintiffs have pointed to no authority to the contrary, and any improper delay in the promulgation of the revised standards is not directly at issue in the case before us, we defer to the EPA’s reasonable application of its own regulations.

We also note that Plaintiffs’ suggestion could be counterproductive, as waiting for revisions to the standards would result in continued delays in producing any § 303(

d) list. Concerns that a particular list will be based on imperfect, though approved, standards are mitigated by the periodic nature of the list. III. For the foregoing reasons, we affirm the judgment of district court. . The Honorable Linda R. Reade, United States District Judge for the Northern District of Iowa, adopting the Report and Recommendation of the Honorable Jon Stuart Scoles, United States Magistrate Judge for the Northern District of Iowa. .

A point source is “any discernible, confined and discrete conveyance, including but not limited to any pipe, ditch, channel, tunnel, conduit, well, discrete fissure, container, rolling stock, concentrated animal feeding operation, or vessel or other floating craft, from which pollutants are or may be discharged.” 33 U.S.C. § 1362(14). A nonpoint source is any other source. . Notwithstanding an attack on "[t]he false distinction between pollution and pollutants,” Plaintiffs do not appear to argue that the causes of the pollution in the disputed waters are pollutants for purposes of the CWA.

Accordingly, we assume, without deciding, that the waters at issue are not, in fact, impaired by "pollutants.” . Iowa’s methodology considered only "monitored” waters as eligible for inclusion on the state’s § 303(

d) list. Iowa describes "monitored waters” as those "for which the use support decision is principally based on current site-specific data.” Iowa considers waters to be only "evaluated,” and thus not considered for inclusion on the state’s § 303(

d) list, if the relevant data is "other than current site-specific data.”