NATURAL RESOURCES DEFENSE COUNCIL, INCORPORATED; Environmental Defense Fund, Incorporated; Audubon Naturalist Society of the Central Atlantic States; Maryland Conservation Council; Conservation Federation of Maryland; John Gottschalk; Mark Kovach; Ken Penrod; Glen Peacock; C.L. Fitchett; Louis W. Powers, Plaintiffs-Appellants, v. UNITED STATES ENVIRONMENTAL PROTECTION AGENCY; Edwin B. Erickson, Region III Administrator; William Reilly, Administrator, United States Environmental Protection Agency; Westvaco Corporation; The American Paper Institute, Incorporated; Chesapeake Corporation; Union Camp Corporation; The State of Maryland, Department of the Environment, Defendants-Appellees, 16 F.3d 1395

OPINION BRITT, District Judge: This appeal arises out of consolidated suits brought by the Natural Resources Defense Council (“NRDC”) and Environmental Defense Fund (“EDF”) to challenge the approval by the United States Environmental Protection Agency (“EPA”) of state water quality standards implemented by Maryland and Virginia. Specifically, NRDC and EDF contest the approval of these state standards as they relate to dioxin. The district court below issued two published opinions regarding this action; Natural Resources Defense Council v.

EPA, 770 F.Supp. 1093 (E.D.Va.1991) (“NRDC I”), and Natural Resources Defense Council v. EPA, 806 F.Supp. 1263 (E.D.Va.1992) (“NRDC II”). In NRDC I, the district court dismissed the original Count One of the Complaint filed in the Maryland action and held that EPA had discretion under the Clean Water Act (“CWA” or “Act”), 33 U.S.C. §§ 1251 et seq., whether to include numerical criteria for all identifiable effects of dioxin and to revise criteria when the latest available scientific knowledge demanded it.

However, the court allowed plaintiff NRDC an opportunity to amend Count One of the Maryland complaint to assert a claim solely under the Administrative Procedure Act (“APA”), 5 U.S.C. §§ 500 et seq. NRDC I, 770 F.Supp. at 1108-10. In NRDC II, the district court granted EPA’s motions to dismiss and for partial

summary judgment. The court held that EPA sufficiently reviewed the Maryland and Virginia dioxin standards in accordance with the CWA and that EPA did not abuse its discretion in determining that Maryland and Virginia relied on scientifically defensible assumptions in setting dioxin standards. The district court also dismissed amended Count One of the Maryland complaint on grounds that NRDC failed to exhaust administrative remedies. NRDC II, 806 F.Supp. at 1277-78.

NRDC and EDF appeal the district court’s decisions and make the following assignments of error: (1) that the district court applied an incorrect legal standard in deciding whether EPA properly approved the state water quality standards; (2) that the district court erred in affirming EPA’s approval of the state dioxin standards; and (3) that the district court erred in dismissing both the original and amended Count One of the Maryland complaint. Finding no error, we affirm. I. FACTS A full account of the facts can be found in NRDC I, 770 F.Supp. at 1094-96, and NRDC II, 806 F.Supp. at 1266-72.

For ease of reference, this court summarizes the facts as follows: On 11 September 1989, the Maryland Department of the Environment (“MDE”) sought to revise Maryland’s water quality standards to allow its waters to contain dioxin in the amount of 1.2 parts per quadrillion (“ppq”), an amount indisputably less protective than EPA’s own guidance criterion of .0013 ppq.

However, MDE chose this 1.2 ppq criterion because it had been based on the Food and Drug Administration’s (“FDA”) less conservative cancer potency factor and because MDE felt that EPA’s cancer potency factor overestimated the carcinogenic potential of dioxin. After public hearings were held on the matter, Maryland adopted the 1.2 ppq standard and submitted it to EPA for review and approval. Similar events took place in Virginia. On 11 December 1989, the Virginia State Water Control Board (“VSWCB”) proposed to revise its water quality standards to include the 1.2 ppq dioxin standard.

After public hearings were held, VSWCB submitted its proposal to EPA for review and approval on 27 September 1990. EPA approved the Maryland standard on 12 September 1990, and approved the Virginia standard on 25 February 1991. Accompanying each approval, a Technical Support Document (“TSD”) was issued by EPA and set out in detail EPA’s scientific review of MDE’s and VSWCB’s analysis in deriving the 1.2 ppq standard. EPA concluded that Maryland’s and Virginia’s use of the 1.2 ppq standard for dioxin was scientifically defensible, protective of human health, and in full compliance with the CWA.

Plaintiffs then initiated this suit in the district court to challenge EPA’s 1984 dioxin criteria document and EPA’s approval of the Maryland and Virginia water quality standards. As noted above, the district court dismissed original Count One of the Maryland complaint on grounds that § 804(

a) of the CWA does not impose a mandatory duty on EPA to develop numeric criteria for dioxin or to update its 1984 dioxin criteria document. NRDC I, 770 F.Supp. at 1107. After giving NRDC an opportunity to amend Count One, the district court dismissed the amended count for lack of finality and failure to exhaust administrative remedies. NRDC II, 806 F.Supp. at 1278. The district court also granted

summary judgment to EPA on the remaining claims, holding that EPA had not acted arbitrarily or capriciously in approving the state water quality standards. Id. at 1277. This appeal followed. II. STATUTORY SCHEME The main purpose of the CWA is to “restore and maintain the chemical, physical, and biological integrity of the Nation’s waters” by reducing, and eventually eliminating, the discharge of pollutants into these waters. 33 U.S.C. § 1251(a) (Supp.1993).

While the states and EPA share duties in achieving this goal, primary responsibility for establishing appropriate water quality standards is left to the states. See id. §§ 1251(b) (1982); Chevron U.S.A., Inc. v. Hammond, 726 F.2d 483 (9th Cir.1984), cert. denied sub nom. Chevron U.S.A. Inc. v. Sheffield, 471 U.S. 1140, 105 S.Ct. 2686, 86 L.Ed.2d 703 (1985); District of Columbia v. Schramm, 631 F.2d 854 (D.C.Cir.1980). EPA sits in a reviewing capacity of the state-implemented standards, with approval and rejection powers only. 33 U.S.C. § 1313(c) (1982 & Supp.1993).

Water quality standards are a critical component of the CWA regulatory scheme because such standards serve as a guideline for setting applicable limitations in individual discharge permits. In an effort to meet the CWA’s primary goal,

section 402 of the Act (33 U.S.C. § 1342) establishes the National Pollutant Discharge Elimination System (“NPDES”) permit program. Under this program, permits are issued by either the EPA or by states that have been allocated NPDES permitting authority. Id. § 1342 (1982 & Supp.1993). However, a state’s exercise of NPDES permitting authority is subject to EPA approval. Id. §§ 1342(c), (d) (1982 & Supp.1993).

All NPDES permits must take into account technology-based effluent limitations that reflect the pollution reduction achievable based on specific equipment or process changes, without reference to the effect on the receiving water, and, where necessary, more stringent limitations representing the level of control necessary to ensure that the receiving waters attain and maintain state water quality standards. Id. §§ 1311(b) (1982), 1313(c) (1982 & Supp.1993). Additionally, the CWA requires each state to adopt water quality standards for all waters of that state and to review them at least every three years.

Id. §§ 1313(a), (b), (c)(1) (1982 & Supp.1993). To adopt these standards, states must first classify the uses for which the water is to be protected, such as fishing and swimming, and then each state must determine the level of water quality necessary to protect those uses.

Thus, the following three factors are considered when adopting or evaluating a water quality standard: (1) one or more designated uses of the state waters involved; (2) certain water quality criteria, expressed as numeric pollutant concentration levels or narrative statements representing a quality of water that supports a particular designated use; and (3) an anti-degradation policy to protect existing uses and high quality waters. Id. § 1313(c)(2)(A) (Supp.1993); 40 C.F.R. § 131.

States are directed to adopt numerical water quality criteria for specific toxic pollutants, such as dioxin, for which EPA has published numerical criteria guidance under 33 U.S.C. § 1314(a), if that pollutant can reasonably be expected to interfere with the designated uses of the states’ waters. Id. § 1313(c)(2)(B) (Supp.1993). As mentioned previously, states must submit their new or revised water quality standards to EPA for review. Id. § 1313(c)(2)(A) (Supp.1993).

On review, each submission must contain at least six elements: (1) use designations consistent with the CWA; (2) a description of methods used and analyses conducted to support revisions of water quality standards; (3) water quality criteria sufficient to protect the designated uses; (4) an antidegradation policy; (5) certification of compliance with state law; and (6) general information to assist EPA in determining the adequacy of the scientific basis for standards that do not include the “fishable/swimmable” uses as set forth in 33 U.S.C. § 1251(a)(2). 40 C.F.R. § 131.6.

EPA regulations also provide that states should develop numerical criteria based on EPA’s criteria guidance under § 304(

a) of the CWA, EPA’s criteria guidance modified to reflect site-specific conditions, or other scientifically defensible methods. 40 C.F.R. § 131.11(b)(1). Alternatively, states should establish narrative criteria or criteria based on biomonitoring methods if numerical criteria cannot be ascertained, or to supplement numerical criteria. Id. § 131.11(b)(2). III. DISCUSSION A.

Standard of Review on Appeal As to the first and third issues on appeal, that is, whether the district court properly applied the correct legal standard under the CWA in reviewing EPA’s approval of the state water quality standards and whether the district court erred in dismissing both original and amended Count One of the Maryland complaint, this court will apply a de novo standard of review. Schatz v. Rosenberg, 943 F.2d 485, 489 (4th Cir.1991), cert. denied sub nom. Schatz v. Weinberg & Green, — U.S. -, 112 S.Ct. 1475, 117 L.Ed.2d 619 (1992); see also L.K. Comstock & Co. v.

United Eng’rs & Constructors Inc., 880 F.2d 219, 221 (9th Cir.1989) (explaining that principles of law applied to facts are reviewed de novo). Regarding the second issue on appeal, which is whether the district court erred in affirming EPA’s approval of the state dioxin standards, it is undisputed that the correct standard of review is whether the agency action was arbitrary or capricious.

The applicable statute provides that an agency’s action, such as the EPA action at issue here, must be upheld unless that action was “arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law.” 5 U.S.C. § 706(2)(A) (1977). This is, of course, a highly deferential standard which presumes the validity of the agency’s action. Ethyl Corp. v. EPA 541 F.2d 1, 34 (D.C.Cir.) (en banc), cert. denied, 426 U.S. 941, 96 S.Ct. 2663, 49 L.Ed.2d 394 (1976).

Thus, this court’s task is to scrutinize the EPA’s activity to determine whether the record reveals that a rational basis exists for its decision. Id This court also is mindful that the CWA is a lengthy and complex statute and that its mandate and policy often require the evaluation of sophisticated data. Reynolds Metal Co. v. EPA, 760 F.2d 649, 558 (4th Cir.1985). Of course, in reviewing EPA’s actions here, this court does not sit as a scientific body, meticulously reviewing all data under a laboratory microscope. Id. at 559.

Nonetheless, EPA must fully and ably explain its course of inquiry, its analysis, and its reasoning, and show that a rational connection exists between its decision-making process and its ultimate decision. Id.; Tanners’ Council of Am., Inc. v. Train, 540 F.2d 1188, 1191 (4th Cir.1976). B. The District Court’s Statutory Review of EPA’s Approval of the State Water Quality Standards NRDC argues that the district court applied an incorrect legal standard in reviewing EPA’s approval of the state dioxin standards.

To support this argument, NRDC contends that EPA’s approval of the state dioxin standards is governed principally by § 303(

c) of the CWA, 33 U.S.C. § 1313(c). Error was committed, so the argument goes, when the district court focused not on § 303(c), but on a policy statement contained in § 101(

b) of the CWA, 33 U.S.C. § 1251(b), when it reviewed EPA’s decision. NRDC argues further that the court’s analysis was flawed because it extended excessive deference to EPA’s decision and, accordingly, did not uphold the purposes of the CWA with respect to state-implemented water quality standards as § 101(

a) and § 303(

c) direct it should do. Specifically, NRDC suggests that under sections 101(

a) and 303(c), EPA has an independent duty to objectively ensure that state water quality standards meet the requirements of the CWA, and that nothing in the CWA allows EPA to defer to states on this issue. Stated differently, NRDC contends that the district court misunderstood its function in reviewing EPA’s approval by according undue deference to that decision. In sum, NRDC maintains that EPA as well as the district court, had a duty under the CWA to assert a more dominant role in the review process. The court is unpersuaded.

At the outset it is important to note, as the district court correctly found, that states have the primary role, under § 303 of the CWA (33 U.S.C. § 1313), in establishing water quality standards. Chevron U.S.A., Inc., 726 F.2d at 489. EPA’s sole function, in this respect, is to review those standards for approval. 33 U.S.C. § 1313(c) (1982 & Supp.1993). Appellants question the intensity of that review, arguing that EPA should not accord an overextended deference to the states’ decisions with regard to its water quality standards.

EPA however, asserts that its duty under the CWA is not to determine whether the states used EPA’s recommended criterion but instead to review state water quality standards and determine whether the states’ decision is scientifically defensible and protective of designated uses. See 40 C.F.R. §§ 131.5(a), 131.6(c), 131.11(a) & (b). While the CWA admittedly is less than crystal clear on this precise issue, the court realizes that it must give due weight to EPA’s

interpretation and administration of this highly complex statute, particularly when its determination appears to be reasonable and is supported by substantial evidence in the administrative record. Shanty Town Assocs. Ltd. v. EPA, 843 F.2d 782, 790 (4th Cir.1988). In each Technical Support Document (“TSD”) issued by the EPA the agency conducted an analysis regarding every assumption used by Maryland and Virginia in deriving their respective water quality standards. EPA independently found that each factor and assumption was scientifically defensible.

In reviewing the criteria as a whole, EPA also found that they protected the uses that they were designed to protect. In light of this extensive agency review, the court reiterates that it does not sit as a scientific body and is not called on to meticulously inspect each and every bit of technical evidence. Reynolds Metal Co., 760 F.2d at 559. Rather, the court’s function is to determine whether proper legal standards were applied.

The court agrees with EPA that its duty, under the CWA and the accompanying regulations, is to ensure that the underlying criteria, which are used as the basis of a particular state’s water quality standard, are scientifically defensible and are protective of designated uses. EPA and the district court abided by that standard, and appellants fail to cite persuasive authority to the contrary.

Furthermore, EPA adequately documented and explained its reasons for approving the states’ water quality standards in the Technical Support Documents attached to its final decisions, and did not merely rubber-stamp each state’s proposed standard. We hold that the district court applied the correct legal standard under the CWA in reviewing EPA’s approval of the state water quality standards at issue. C.

The District Court’s Affirmance of EPA’s Approval of the Maryland and Virginia Dioxin Standards Appellants argue that the district court’s affirmance of EPA’s approval of the Maryland and Virginia water standards should be reversed primarily for two reasons. First, they assert that EPA’s approval was arbitrary and capricious because it was not based on all relevant factors, ignored key aspects of the record before it, and failed to show a rational connection between the facts found and the choices made. Second, they maintain that EPA’s action was contrary to law because it did not ensure, as required by § 303(

c) of the CWA (33 U.S.C. § 1313(c)), that state standards were consistent with the CWA; that is, that the standard protected all designated water uses. Specifically, NRDC attacks EPA’s assessment of the Maryland and Virginia standards regarding the first four factors used in the numeric dioxin criteria determination, namely: (1) cancer potency, (2) risk level, (3) fish consumption, and (4) bioconcentration factor (“BCF”). Of these four, NRDC emphasizes its challenge with respect to the latter two factors, fish consumption and BCF.

NRDC contends that these two factors, when considered together, are important because they determine the ultimate “exposure” of an individual to dioxin, while the remaining factors only involve choices about risk or toxicity. 1. Fish Consumption EPA estimates, on a national average, that an individual eats 6.5 grams of fish per day. Maryland and Virginia used this estimate, inter alia, in calculating the 1.2 ppq water quality standard.

Appellants argue that by affirming EPA’s approval of the states’ use of this estimate, the district court failed to require EPA to protect subpopula-tions with higher than average fish consumption, particularly recreational and subsistence fishers. Specifically, appellants contend that EPA’s 6.5 grams per day fish consumption factor underestimates the actual fish consumption of subpopulations in Maryland and Virginia, and therefore is not protective of a designated use.

Appellants further contend that EPA’s use of the 6.5 grams per day fish consumption factor is unsupported by the record and violates EPA’s own policy and regulations. They emphasize that Maryland and Virginia are coastal states and, as such, are entitled — according to EPA recommendations — to higher than average values for fish consumption. EPA points out that the 6.5 grams per day value is not intended to represent total fish consumption but, rather, that subset of fish containing the maximum residues of dioxin permissible under state law.

In setting this value, EPA was establishing a national standard and was well aware that subpopulations might very well consume more than 6.5 total grams of fish per day. No evidence was presented that the subpopulations referred to are consuming more than 6.5 grams per day of maximum residue fish. Appellants argue that the risk is especially high for the Mattaponi and Pamunkey Native American peoples who live near a major paper mill in Virginia and who, it is argued, consume higher-than-average amounts of fish.

EPA counters that the fish consumption of these subpopulations is speculative at best, that it is based on anecdotal evidence, and that there is no evidence that the fish that actually are consumed are maximum residue fish. In fact, EPA argues that the Native Americans fish in the streams primarily for shad and herring, both of which are anadromous fish that spend a large part of their lives in the oceans and migrate to the rivers only at certain stages during their lives.

The District Court concluded that the EPA, in exercising its judgment, “relied on scientifically defensible means to reach reasoned judgments regarding fish consumption levels.” NRDC II, at 1276. We agree. 2. Bioconcentration Factor (BCF) Based on EPA laboratory studies, dioxin is more soluble in fat tissues than it is in water. As a result, it tends to accumulate in fish fat tissues at concentrations higher than those present in the water.

By averaging the fat content of fish likely to be eaten by an exposed population, a generic BCF can be calculated that reflects dioxin’s presence in fish as some multiple of its concentration in ambient water. In its 1984 dioxin criteria document, EPA calculates a dioxin BCF of 5000 for fish of average (3%) lipid content. Maryland and Virginia used this BCF figure, inter alia, to derive their numeric water quality criteria. Appellants challenge EPA’s use and approval of a 5000 BCF.

They essentially contend that the 5000 BCF figure is outdated because the latest scientific research suggests that a higher BCF should be used. Citing the administrative record, appellants emphasize that:

(1) EPA admits that scientific literature and research has changed significantly since preparation of the 1984 dioxin criteria document;

(2) EPA further admits that BCF factors now range from 26,000 to 150,000, depending on test species;

(3) Virginia conducted a state-specific study which revealed a BCF calculation of 22,000; and

(4) Maryland refused to conduct such a study. Appellants contend that, taking all of these factors into account, EPA ignored all the current scientific data and simply “defaulted” to its old BCF assumption. Appellants argue that EPA acted arbitrarily and improperly in not requiring a higher BCF, especially when Virginia and Maryland chose less stringent factors for cancer potency and risk. We disagree. Once again, we are confronted with an area dominated by complex scientific inquiry and judgment.

Although EPA is aware that some recent BCF studies suggested a higher BCF than 5000, EPA maintains that such results are inconclusive and that no compelling scientific evidence indicates that a 5000 BCF is no longer within the range of scientific defensibility. We simply are not in a position to second-guess this technical decision by administrative experts.

A review of the record does indicate that several more recent BCF studies have been conducted and that some have suggested a higher BCF; however, the court concludes that the best course of action is to leave this debate to the world of science to ultimately be resolved by those with specialized training in this field. Upon a careful review of the administrative record, we find no clear evidence showing that the 5000 BCF figure is not supported by sound scientific rationale.

Accordingly, we hold that EPA did not act arbitrarily in approving the BCF figure used by Maryland and Virginia, and that EPA has made a rational connection between the facts found in the administrative record and its choice to approve the BCF figure. EPA’s approval of the 5000 BCF will not be disturbed. 8. Protection of All Stream Uses Appellants next contend that the district court ratified EPA’s approval of the state dioxin standards without ensuring protection of all stream uses.

Appellants suggest that when EPA adopted the 1.2 ppq standard, it was required to demonstrate that other stream uses were protected. They maintain that EPA ignored record evidence revealing that the 1.2 ppq standard could cause serious, direct, toxic effects to aquatic life and other wildlife that consume fish tainted with dioxin. Appellants thus argue that EPA did not follow the CWA, its regulations, or its own guidelines by asserting that the water quality criteria were intended to address only one of the minimum statutory uses, human health protection.

Essentially, appellants claim that states must adopt a single criterion for dioxin that protects against all identifiable effects on human health, aquatic life, and wildlife. We disagree.

Section 303(c)(2)(

A) of the CWA (33 U.S.C. § 1313(c)(2)(A) (Supp.1993)) requires that new or revised water quality standards “consist of designated uses of the navigable waters involved and the water quality criteria for such waters based upon such uses.” That

section also provides: Such standards shall be such as to protect the public health or welfare, enhance the quality of water and serve the purposes of this chapter. Such standards shall be established taking into consideration their use and value for public water supplies, propagation of fish and wildlife, recreational purposes, and agricultural, industrial, and other purposes, and also taking into consideration their use and value for navigation. Id.

Reference to the regulations also is instructive: “A water quality standard ... defines the water quality goals of a water body, or portion thereof, by designating the use or uses to be made of the water and by setting criteria necessary to protect the uses.” 40 C.F.R. §§ 130.3, 131.2. The regulations define “criteria” as “elements of State water quality standards, expressed as constituent concentrations, levels, or narrative statements, representing a quality of water that supports a particular use. When criteria are met, water quality will generally protect the designated use.” Id. § 131.3(b).

Section 131.11(

a) further provides that “[s]tates must adopt those water quality criteria that protect the designated use. Such criteria must be based on sound scientific rationale and must contain sufficient parameters or constituents to protect the designated use. For waters with multiple use designations, the criteria shall support the most sensitive use.” Id. § 131.11(a). As previously indicated, states should develop either numerical criteria based upon CWA guidance (or other scientific methods), or narrative criteria, if numerical criteria cannot be established.

Narrative criteria might also be developed to supplement numerical criteria. Id. § 131.11(b). Clearly, the form of a particular state’s water criteria may be either numeric or narrative, depending upon the designated use, as the district court correctly recognized. NRDC II, 806 F.Supp. at 1277. In view of the above, we find that use of the term “criteria” in CWA § 303(c)(2)(

A) and the regulations means that states may adopt multiple criteria for the same pollutant. Thus, where multiple uses are designated for a body of water, there may be multiple criteria applicable to it, as long as the criteria support the most sensitive use of that particular body of water. States have exclusive responsibility to designate water uses. See 40 C.F.R. § 131.10.

However, in determining these use designations, states must take into account whether the body of water serves as a public water supply, its role in the protection and propagation of fish, shellfish and wildlife, recreation in and on the water, and agricultural, industrial, and other uses, including navigation. Id. EPA avers that its review of the Maryland and Virginia standards was limited exclusively to protection of human health against any potential adverse effects (both cancerous and non-cancerous) caused by dioxin. The TSDs reflect this position.

In reviewing the Virginia water quality standard, EPA stated: The Virginia criterion for [dioxin] is designed to protect human health. Accordingly, EPA has limited its review to assessing the adequacy of the numeric criterion for that purpose. Virginia did not submit a criterion for [dioxin] for the protection of aquatic life. Depending on the circumstances, greater protection than is afforded by Virginia’s 1.2 ppq criterion may be required for this purpose.

In the absence of a numeric criterion for [dioxin] to protect aquatic life, Virginia’s narrative criterion must, consistent with 40 C.F.R. § 122.-44(d), be interpreted in individual permitting actions to prevent harm to aquatic life. J.A. at 280-81 (footnote omitted). EPA’s comments in the Maryland TSD are similar: The Maryland criterion for [dioxin] is designed to protect human health. Accordingly, EPA’s review is limited to assessing the adequacy of the numeric criterion for that purpose.

In the absence of a numeric criterion for [dioxin] to protect aquatic life, Maryland’s narrative criteria must be interpreted in individual permitting actions to prevent harm to aquatic life. See CO-MAR § 26.10.01.03.B.(5)(b). Depending on the circumstances, greater protection than is afforded by Maryland’s 1.2 ppq criterion may be required for this purpose. Id. at 314 (footnote omitted). Thus, EPA duly acknowledged that dioxin may have adverse effects on aquatic life.

However, EPA also noted that application of existing, separate narrative criteria protecting such aquatic life and wildlife could require more stringent controls in some cases than would be required through use of the human health criteria alone. EPA conducted an extensive review of the adequacy of the states’ criteria to protect human health, aquatic life and wildlife.

Appellants have failed to cite any convincing authority showing that states have an obligation under the CWA or its accompanying regulations to adopt a single numeric criterion for dioxin that protects against all identifiable effects to human health, aquatic life, and wildlife. D.

Summary We find that EPA’s review of the Maryland and Virginia water quality standards was neither arbitrary nor capricious. Each review conducted by EPA was supported by lengthy, highly scientific, technical support documents explaining in detail EPA’s rationale in approving the 1.2 ppq standards. EPA has satisfied this court that substantial evidence exists in the administrative record to support its decision, and that it acted rationally and in accordance with the CWA and its regulations.

We therefore refuse to upset either EPA’s decision to approve Maryland’s and Virginia’s adoption of the 1.2 ppq standard or the district court decision affirming the same. E. The District Court’s Dismissal of the Original and Amended Count One of the Maryland Complaint In original and amended Count One of the Maryland complaint, NRDC challenges EPA’s water quality criteria as a whole, alleging that EPA failed to issue and revise complete water quality criteria for dioxin. 1. Original Count One Original Count One of NRDC’s complaint alleged that EPA violated a nondiscretionary duty, assigned to it by § 304(

a) of the CWA (33 U.S.C. § 1314(a)), to issue water quality criteria for dioxin that reflect the latest scientific information and that address all identifiable effects on health and welfare. The district court found that EPA’s duty was discretionary, see NRDC I, 770 F.Supp. at 1107, and dismissed original Count One, ruling that the citizen’s suit provision, § 505(a)(2) of the CWA (33 U.S.C. § 1365(a)(2)) (1982 & Supp.1993), did not confer jurisdiction over NRDC’s claim. Id. at 1110.

On appeal, NRDC maintains that EPA has a mandatory duty to issue complete, numerical water quality criteria for dioxin, and to revise such criteria based on the latest scientific information. The district court conducted a plenary, exhaustive examination as to whether EPA had a mandatory duty to issue or revise numerical water quality criteria for dioxin and concluded that it did not. We agree and therefore affirm dismissal of original Count One for the reasons expressed by the district court in its thorough and well-reasoned opinion, Natural Resources Defense Council v.

EPA, 770 F.Supp. 1093 (E.D.Va.1991) (“NRDC I”). 2. Amended Count One As part of its ruling in NRDC I, the district court permitted NRDC to amend Count One to challenge EPA’s actions solely under the APA Id. at 1110 n. 14. NRDC’s amended complaint asserted such a claim under 5 U.S.C. § 706 of the APA. EPA moved to dismiss amended Count One on grounds that the applicable statute of limitations had run and that NRDC had faded to exhaust its administrative remedies.

M-though the district court found that the applicable statute of limitations had not expired, the court granted EPA’s motion, holding that because EPA was in the process of reviewing its dioxin criteria, it had made a reviewable final administrative decision. NRDC II, 806 F.Supp. at 1278. NRDC claims that the district court’s dismissal of amended Count One improperly closed the door to further judicial review of EPA’s dioxin criteria. NRDC argues that the district court’s analysis of this issue focused only on NRDC’s claim that EPA failed to revise its dioxin criteria based on new scientific evidence.

In light of this, NRDC contends that the district court erroneously failed to explain how NRDC was barred from challenging EPA’s 1984 dioxin criteria document under the APA. Having reviewed relevant provisions of the APA and corresponding case law, we are not persuaded.

Section 704 of the APA provides, in pertinent part, that an “[ajgency action made reviewable by statute and [a] final agency action for which there is no other adequate remedy in a court is subject to judicial review. A preliminary, procedural, or intermediate agency action or ruling not directly reviewable is subject to review on the review of final agency action.” 5 U.S.C. § 704. Thus, the key focus here is whether EPA’s actions with respect to its 1984 dioxin criteria document are “final” such that federal courts may exercise their powers of review under the APA. In FTC v.

Standard Oil Co., 449 U.S. 232, 239—40, 101 S.Ct. 488, 492-93, 66 L.Ed.2d 416 (1980), the Supreme Court outlined several factors to consider in determining whether an agency action is “final”: (1) whether the action is a definitive statement of the agency’s position; (2) whether the action had the status of law and immediate compliance with its terms was expected; (3) whether the action had a direct impact on the day-to-day business of plaintiff; and (4) whether pre-enforcement challenge was calculated to speed enforcement and prevent piecemeal litigation.

An analysis of these factors reveals that EPA’s actions regarding its 1984 criteria document are not “final” within the meaning of the APA. Considering the first and second factors, EPA’s 1984 criteria document is neither a “definitive” statement of its position nor does it have the status of law, compelling immediate compliance with its terms. Although this document does serve as an important reference manual to states as they develop water quality criteria for dioxin, we note that it does not purport to create or establish rights or responsibilities for any party, nor does it mandate legal action. EPA’s

interpretation of the role the criteria play in the CWA regulatory process is illustrative: “Section 304(a)(1) criteria are not rules and they have no regulatory impact. Rather, these criteria present scientific data and guidance on the environmental effect of pollutants which can be useful to derive regulatory requirements based on the considerations of water quality impacts.” 45 Fed.Reg. 79,319 (1980). Also informative is the language contained in the preface to the 1984 criteria document: [Water quality criteria] represent! ] a non-regulatory, scientific assessment of ecological effects.

The criteria presented in this publication are such scientific assessments .... The water quality criteria adopted in the State water quality standards could have the same numerical limits as the criteria developed under

section 304. However, in many situations States may want to adjust water quality criteria developed under

section 304 to reflect local environmental conditions and human exposure patterns before incorporation into water quality standards. It is not until their adoption as part of the State water quality standards that the criteria become regulatory. 1984 Dioxin Criteria Document at iii (J.A. at 1051) (emphasis added). No compulsory language is included in this provision of the 1984 criteria document. It specifically states that these water quality criteria are non-regulatory, and that they become regulatory only when a state adopts them.

Until such time, however, a state may or may not choose to follow EPA’s water quality criteria, and the preface to the 1984 dioxin criteria document expressly acknowledges this. In view of the above, we conclude that the 1984 dioxin criteria document simply serves as a useful guide to assist the states in developing their own respective water quality standards. See American Paper Inst., Inc. v. EPA 882 F.2d 287 (7th Cir.1989) (holding that EPA Region’s “policy statement” does not constitute final agency action because it does not compel action).

Regarding the third factor, the 1984 criteria does not affect the day-to-day business of NRDC for the reason that any practical effects will occur only when and if state-issued standards are incorporated into enforceable NPDES permit limitations. It would therefore be entirely too speculative to presume that an EPA criteria guidance will have any impact when it may or may not serve as the basis for state adoption and subsequent EPA approval of a state water quality standard.

Finally, with respect to the fourth factor, we believe that to review EPA’s actions at this stage, when it currently is engaging in a reassessment of its dioxin criteria, would be premature and would foster unnecessary piecemeal litigation. A waste of judicial resources is almost inevitable if we were to allow an exhaustive review of EPA’s current water quality criteria, only to have EPA drastically overhaul its existing water criteria ■with a completely new and different standard.

We therefore hold that EPA’s action with regard to its water quality criteria for dioxin is not a reviewable “final” agency action for purposes of the APA. Nothing in the record, briefs, or oral argument leads this court to believe that EPA has not been forthright, or has otherwise proceeded in bad faith, with regard to its reassessment of dioxin criteria. We trust and expect that EPA will expedite its ongoing review of dioxin, as it has so conveyed in its briefs. We will not disturb this highly technical administrative process at this point and instead will allow it an opportunity to run its course.

See Hopewell Nursing Home, Inc. v. Heckler, 784 F.2d 554, 557-58 (4th Cir.1986); American Gen. Ins. Co. v. FTC, 496 F.2d 197, 200 (5th Cir.1974). For the foregoing reasons, the judgment of the district court is AFFIRMED. . NRDC sued EPA challenging the Maryland water quality standards, and EDF sued EPA challenging the Virginia standards. . The term "dioxin” generally encompasses a broad range of closely-related toxic organic chemical compounds. The specific dioxin compound at issue on this appeal is 2,3,7,8-Tetra-chlorodibenzo-p-dioxin ("2,3,7,8-TCDD”). It is highly probable that dioxin is a potent carcinogen.

Dioxin is primarily a by-product of the chlorine bleaching of pulp associated with paper manufacturing. . As explained fully by the district court, numeric water criteria, such as the 1.2 ppq and .0013 ppq standards, are based on an assessment of the dose of dioxin that may cause harm and the dose to humans that can be expected as a result of dioxin present in water. Six factors are considered in determining the numeric dioxin criteria: (1) cancer potency; (2) risk level; (3) fish consumption; (4) bioconcentration; (5) water intake; and (6) body weight. See discussion infra.

Of these factors, the first four are primarily at issue on this appeal. The .0013 ppq figure is taken from EPA's dioxin criteria guidance document, Ambient Water Quality Criteria for 2,3,7,8-Tetrachlorodibenzo-p-dioxin, published in 1984 ("1984 dioxin criteria document”). In this document, EPA summarized the scientific information available in 1984 regarding dioxin toxicity and provided useful information for the states to use in adopting then-own water quality standards.

EPA recommended that where bodies of water are used as a source for both drinking water and edible fish, a dioxin concentration of .0013 ppq is desirable. This .0013 ppq figure means, approximately speaking, that one out of every ten million individuals faces an excess risk of cancer exposure as a result of the water's dioxin content. Thus, a 1.2 ppq standard would mean that, according to EPA’s assessment, roughly one out of every ten thousand individuals would face such exposure. .EPA relates to the court that its cancer potency factor is among the most conservative, i.e. the most protective, in the world.

Other federal agencies, including the FDA and the Center for Disease Control ("CDC”), adopted less conservative dioxin potency standards because these agencies used different assumptions and risk assessment methodologies. Some foreign countries, such as Canada and the Netherlands, have developed estimates of cancer risk even less protective than that used by the FDA or CDC. . For a more thorough explanation of the NPDES framework, see Westvaco Corp. v. EPA, 899 F.2d 1383 (4th Cir.1990). . It is undisputed that Virginia and Maryland have been given such authority. .

Section 303(c)(3) of the CWA (33 U.S.C. § 1313(c)(3)) provides that “[i]f the Administrator, within sixty days after the date of submission of the revised or new standard, determines that such standard meets the requirements of [the Act], such standard shall thereafter be the water quality standard for the applicable waters of that State.” 33 U.S.C. § 1313(c)(3) (1982) (emphasis added). . Appellants contend that the district court committed serious error in its reading and application of Mississippi Comm’n on Natural Resources v. Costle, 625 F.2d 1269 (5th Cir.1980). See NRDC II, 806 F.Supp. at 1273.

We find this argument without merit particularly because the district court cited Costle merely as a reference and only for the proposition that "[t]he CWA does not require uniformity among states, only compliance with its statutory mandate.” Id. While citation solely to Costle for this proposition may be viewed as debatable, the district court did not in any other way rely on Costle to ultimately conclude that EPA properly reviewed the Maryland and Virginia standards. Any perceived error therefore would be harmless. In any event, the district court’s reading of Costle plays no

part in our analysis. . Cancer potency measures the “strength” of dioxin's potential to cause cancer. . Risk level is the projected risk of cancer incidence among an exposed population, ranging from one case in ten million individuals to one case in 100,000. NRDC concedes that risk level generally is a state policy choice, and thus this factor is challenged the least by NRDC. . Fish consumption predicts the amount of dioxin-exposed fish consumed by a given population. .

BCF predicts how many times greater the concentration of a pollutant, such as dioxin, will be in the tissues of living organisms, such as fish, as compared to the concentration of that pollutant in the ambient water in which the fish lives. According to its 1984 dioxin criteria document, EPA calculates a dioxin BCF of 5000. . This 6.5 grams per day estimate is based on a 1973-74 market survey data compiled by the National Purchase Diary and the National Marine Fisheries Service. This rate includes consumption of all freshwater and estuarine fish and shellfish, both pollutant-bearing and non-pollutant bearing.

It does not include marine fish. . Lipid content refers to the level of fat found in a species. Because dioxin is “lipophilic,” it concentrates in fat tissue to a greater degree than in other body parts. . EPA has not established national numeric criteria guidance for dioxin with respect to its effects on aquatic life and wildlife. .

Section 304(a) (33 U.S.C. § 1314(a) (1982)) provides in pertinent part that the EPA Administrator "shall develop and publish, within one year after October 18, 1972 (and from time to time thereafter revise) criteria for water quality accurately reflecting the latest scientific knowledge (

A) on the kind and extent of all identifiable effects on health and welfare including[, but not limited to, various forms of plant life, wildlife, and fish].” . This count was brought pursuant to the citizen’s suit provision, § 505(a)(2) of the CWA (33 U.S.C. § 1365(a)(2) (1982 & Supp.1993)), which permits citizens to bring suit "where there is an alleged failure of the Administrator to perform any act or duty under [the CWA] which is not discretionary.” .

Specifically, NRDC alleged in amended Count One that EPA's failure to issue and revise current numeric criteria for dioxin "was arbitrary, capricious, an abuse of discretion, and otherwise not in accordance with law; and constituted agency action unlawfully withheld or unreasonably delayed; in violation of 5 U.S.C. § 706.” Pls.' Am.Compl. at 21-22 (J.A. at 71-72).

Section 706 of the APA provides in relevant part: To the extent necessary to decision and when presented, the reviewing court shall decide all relevant questions of law, interpret constitutional and statutory provisions, and determine the meaning or applicability of the terms of an agency action. The reviewing court shall— (1) compel agency action unlawfully withheld or unreasonably delayed; and (2) hold unlawful and set aside agency action, findings, and conclusions found to be— (

A) arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law.... In making the foregoing determinations, the court shall review the whole record or those parts of it cited by a party, and due account shall be taken of the rule of prejudicial error. 5 U.S.C. § 706 (1977).