I do fail to understand, however, why the plurality would not similarly apply this logic to dredged and fill material. The EPA’s authority over pollutants (other than dredged and fill materials) stems from the identical statutory language that gives rise to the Corps’ §404 jurisdiction. The plurality claims that there is a practical difference, asserting that dredged and fill material “does not normally wash downstream.” Ante, at 26. While more of this material will probably stay put than is true of soluble pollutants, the very existence of words like “alluvium” and “silt” in our language, see Webster’s Third 59, 2119, suggests that at least some fill makes its way downstream. See also, e.g., United States v. Deaton, 332 F. 3d 698, 707 (CA4 2003) (“Any pollutant or fill material that degrades water quality in a tributary has the potential to move downstream and degrade the quality of the navigable waters themselves”). Moreover, such fill can harm the biological integrity of downstream waters even if it largely stays put upstream. The Act’s purpose of protecting fish, see 33 U. S. C. §1251(a)(2); S. D. Warren Co., 547 U. S., at __ (slip op., at 13–14), could be seriously impaired by sediment in upstream waters where fish spawn, since excessive sediment can “smother bottom-dwelling invertebrates and impair fish spawning,” OTA 48. See also, e.g., Erman & Hawthorne, The Quantitative Importance of an Intermittent Stream in the Spawning of Rainbow Trout, 105 Transactions of the American Fisheries Society 675– 681 (1976); Brief for American Rivers et al. as Amici Curiae 14 (observing that anadromous salmon often spawn in small, intermittent streams).

IV

While I generally agree with Parts I and II–A of Justice KENNEDY’s opinion, I do not share his view that we should replace regulatory standards that have been in place for over 30 years with a judicially crafted rule dis- tilled from the term “significant nexus” as used in SWANCC. To the extent that our passing use of this term has become a statutory requirement, it is categorically satisfied as to wetlands adjacent to navigable waters or their tributaries. Riverside Bayview and SWANCC together make this clear. SWANCC’ s only use of the term comes in the sentence: “It was the significant nexus between the wetlands and ‘navigable waters’ that informed our reading of the [Clean Water Act] in Riverside Bay-view .” 531 U. S., at 167. Because Riverside Bayview was written to encompass “wetlands adjacent to navigable waters and their tributaries,” 474 U. S., at 123, and reserved only the question of isolated waters, see id., at 131– 132, n. 8; see also n. 3, supra, its determination of the Corps’ jurisdiction applies to the wetlands at issue in these cases.

Even setting aside the apparent applicability of Riverside Bayview . I think it clear that wetlands adjacent to tributaries of navigable waters generally have a “significant nexus” with the traditionally navigable waters downstream. Unlike the “nonnavigable, isolated, intrastate waters” in SWANCC, 531 U. S., at 171, these wetlands can obviously have a cumulative effect on downstream water flow by releasing waters at times of low flow or by keeping waters back at times of high flow. This logical connection alone gives the wetlands the “limited” connection to traditionally navigable waters that is all the statute requires, see id., at 172; 474 U. S., at 133—and disproves Justice KENNEDY’s claim that my approach gives no meaning to the word “‘navigable,’” ante, at 21 (opinion concurring in judgment). Similarly, these wetlands can preserve downstream water quality by trapping sediment, filtering toxic pollutants, protecting fish-spawning grounds, and so forth. While there may exist categories of wetlands adjacent to tributaries of traditionally navigable waters that, taken cumulatively, have no plausibly discernable relationship to any aspect of downstream water quality, I am skeptical. And even given Justice KENNEDY’s “significant nexus” test, in the absence of compelling evidence that many such categories do exist I see no reason to conclude that the Corps’ longstanding regulations are overbroad.

Justice KENNEDY’s “significant nexus” test will probably not do much to diminish the number of wetlands covered by the Act in the long run. Justice Kennedy himself recognizes that the records in both cases contain evidence that “should permit the establishment of a significant nexus,” ante, at 27, see also ante, at 26, and it seems likely that evidence would support similar findings as to most (if not all) wetlands adjacent to tributaries of navigable waters. But Justice KENNEDY’s approach will have the effect of creating additional work for all concerned parties. Developers wishing to fill wetlands adjacent to ephemeral or intermittent tributaries of traditionally navigable waters will have no certain way of knowing whether they need to get §404 permits or not. And the Corps will have to make case-by-case (or category-by-category) jurisdictional determinations, which will inevitably increase the time and resources spent processing permit applications. These problems are precisely the ones that Riverside Bayview ’s deferential approach avoided. See 474 U. S., at 135, n. 9 (noting that it “is of little moment” if the Corps’ jurisdiction encompasses some wetlands “not significantly intertwined” with other waters of the United States). Unlike Justice Kennedy, I see no reason to change Riverside Bayview ’s approach—and every reason to continue to defer to the Executive’s sensible, bright-line rule.

V

As I explained in SWANCC, Congress passed the Clean Water Act in response to wide-spread recognition—based on events like the 1969 burning of the Cuyahoga River in Cleveland—that our waters had become appallingly polluted. 531 U. S., at 174–175 (dissenting opinion). The Act has largely succeeded in restoring the quality of our Nation’s waters. Where the Cuyahoga River was once coated with industrial waste, “[t]oday, that location is lined with restaurants and pleasure boat slips.” EPA, A Benefits Assessment of the Water Pollution Control Programs Since 1972, p. 1–2 (Jan. 2000), http://www.epa.gov/ost/economics/assessment.pdf. By curtailing the Corps’ jurisdiction of more than 30 years, the plurality needlessly jeopardizes the quality of our waters. In doing so, the plurality disregards the deference it owes the Executive, the congressional acquiescence in the Executive’s position that we recognized in Riverside Bayview, and its own obligation to interpret laws rather than to make them. While Justice KENNEDY’s approach has far fewer faults, nonetheless it also fails to give proper deference to the agencies entrusted by Congress to implement the Clean Water Act.

I would affirm the judgments in both cases, and respectfully dissent from the decision of five Members of this Court to vacate and remand. I close, however, by noting an unusual feature of the Court’s judgments in these cases. It has been our practice in a case coming to us from a lower federal court to enter a judgment commanding that court to conduct any further proceedings pursuant to a specific mandate. That prior practice has, on occasion, made it necessary for Justices to join a judgment that did not conform to their own views.[13] In these cases, however, while both the plurality and Justice Kennedy agree that there must be a remand for further proceedings, their respective opinions define different tests to be applied on remand. Given that all four Justices who have joined this opinion would uphold the Corps’ jurisdiction in both of these cases—and in all other cases in which either the plurality’s or Justice KENNEDY’s test is satisfied—on remand each of the judgments should be reinstated if either of those tests is met.[14]

Justice Breyer, dissenting.

In my view, the authority of the Army Corps of Engineers under the Clean Water Act extends to the limits of congressional power to regulate interstate commerce. See Solid Waste Agency of Northern Cook Cty. v. Army Corps of Engineers, 531 U. S. 159, 181–182 (2001) (SWANCC) (Stevens, J., dissenting). I therefore have no difficulty finding that the wetlands at issue in these cases are within the Corps’ jurisdiction, and I join Justice Stevens’ dissenting opinion.

My view of the statute rests in part upon the nature of the problem. The statute seeks to “restore and maintain the chemical, physical, and biological integrity of the Nation’s waters.” 33 U. S. C. §1251(a). Those waters are so various and so intricately interconnected that Congress might well have decided the only way to achieve this goal is to write a statute that defines “waters” broadly and to leave the enforcing agency with the task of restricting the scope of that definition, either wholesale through regulation or retail through development permissions. That is why I believe that Congress, in using the term “waters of the United States,” §1362(7), intended fully to exercise its relevant Commerce Clause powers.

I mention this because the Court, contrary to my view, has written a “nexus” requirement into the statute. SWANCC, supra, at 167; ante, at 22 (opinion of Kennedy, J.) (“[T]he Corps’ jurisdiction over wetlands depends upon the existence of a significant nexus between the wetlands in question and navigable waters in the traditional sense”). But it has left the administrative powers of the Army Corps of Engineers untouched. That agency may write regulations defining the term—something that it has not yet done. And the courts must give those regulations appropriate deference. Chevron U. S. A. Inc. v. Natural Resources Defense Council, Inc., 467 U. S. 837 (1984).

If one thing is clear, it is that Congress intended the Army Corps of Engineers to make the complex technical judgments that lie at the heart of the present cases (subject to deferential judicial review). In the absence of updated regulations, courts will have to make ad hoc determinations that run the risk of transforming scientific questions into matters of law. That is not the system Congress intended. Hence I believe that today’s opinions, taken together, call for the Army Corps of Engineers to write new regulations, and speedily so.

--------

[*]Together with No. 04–1384, Carabell et al. v. United States Army Corps of Engineers et al., also on certiorari to the same court.

[*]The scope of the proposed rulemaking was not as narrow as Justice Stevens suggests, post, at 10, n. 4 (dissenting opinion). See 68 Fed. Reg. 1994 (2003) (“Additionally, we invite your views as to whether any other revisions are needed to the existing regulations on which waters are jurisdictional under the CWA”); id., at 1992 (“Today’s [notice of proposed rulemaking] seeks public input on what, if any, revisions in light of SWANCC might be appropriate to the regulations that define ‘waters of the U. S.’, and today’s [notice] thus would be of interest to all entities discharging to, or regulating, such waters” (emphases added)). The agencies can decide for themselves whether, as the SWANCC dissenter suggests, it was wise for them to take no action in response to SWANCC.

[1] In issuing permits, the Corps directs that “[a]ll factors which may be relevant to the proposal must be considered including the cumulative effects thereof: among those are conservation, economics, aesthetics, general environmental concerns, wetlands, historic properties, fish and wildlife values, flood hazards, floodplain values, land use, navigation, shore erosion and accretion, recreation, water supply and conservation, water quality, energy needs, safety, food and fiber production, mineral needs, considerations of property ownership and, in general, the needs and welfare of the people.” §320.4(a).

[2] We are indebted to the Sonoran court for a famous exchange, from the movie Casablanca (Warner Bros. 1942), which portrays most vividly the absurdity of finding the desert filled with waters:

“ ‘Captain Renault [Claude Rains]: “What in heaven’s name brought you to Casablanca?”

“ ‘Rick [Humphrey Bogart]: “My health. I came to Casablanca for the waters.”

“ ‘Captain Renault: “The waters? What waters? We’re in the desert.”

“ ‘Rick: “I was misinformed.’ ” 408 F. 3d, at 1117.

[3] One possibility, which we ultimately find unsatisfactory, is that the “other” waters covered by 33 U. S. C. §1344(g)(1) are strictly intrastate waters that are traditionally navigable. But it would be unreasonable to interpret “the waters of the United States” to include all and only traditional navigable waters, both interstate and intrastate. This would preserve the traditional import of the qualifier “navigable” in the defined term “navigable waters,” at the cost of depriving the qualifier “of the United States” in the definition of all meaning. As traditionally understood, the latter qualifier excludes intrastate waters, whether navigable or not. See The Daniel Ball, 10 Wall. 557, 563 (1871). In SWANCC, we held that “navigable” retained something of its traditional import. 531 U. S., at 172. A fortiori, the phrase “of the United States” in the definition retains some of its traditional meaning.

[4] Justice Kennedy observes, post, at 13 (opinion concurring in judgment), that the dictionary approves an alternative, somewhat poetic usage of “waters” as connoting “[a] flood or inundation; as the waters have fallen. ‘The peril of waters, wind, and rocks.’ Shak. ” Webster’s Second 2882. It seems to us wholly unreasonable to interpret the statute as regulating only “floods” and “inundations” rather than traditional waterways—and strange to suppose that Congress had waxed Shakespearean in the definition section of an otherwise prosaic, indeed downright tedious, statute. The duller and more commonplace meaning is obviously intended.

[5]By describing “waters” as “relatively permanent,” we do not necessarily exclude streams, rivers, or lakes that might dry up in extraordinary circumstances, such as drought. We also do not necessarily exclude seasonal rivers, which contain continuous flow during some months of the year but no flow during dry months—such as the 290-day, continuously flowing stream postulated by Justice Stevens’ dissent (hereinafter the dissent), post, at 15. Common sense and common usage distinguish between a wash and seasonal river.

Though scientifically precise distinctions between “perennial” and “intermittent” flows are no doubt available, see, e.g., Dept. of Interior, U. S. Geological Survey, E. Hedman & W. Osterkamp, Streamflow Characteristics Related to Channel Geometry of Streams in Western United States 15 (1982) (Water-Supply Paper 2193), we have no occasion in this litigation to decide exactly when the drying-up of a stream bed is continuous and frequent enough to disqualify the channel as a “wate[r] of the United States.” It suffices for present purposes that channels containing permanent flow are plainly within the definition, and that the dissent’s “intermittent” and “ephemeral” streams, post, at 16 (opinion of Stevens, J.)—that is, streams whose flow is “[c]oming and going at intervals . . . [b]roken, fitful,” Webster’s Second 1296, or “existing only, or no longer than, a day; diurnal . . . short-lived,” id., at 857—are not.

[6]The principal definition of “stream” likewise includes reference to such permanent, geographically fixed bodies of water: “[a] current or course of water or other fluid, flowing on the earth, as a river, brook, etc. ” Id., at 2493 (emphasis added). The other definitions of “stream” repeatedly emphasize the requirement of continuous flow: “[a] steady flow, as of water, air, gas, or the like”; “[a]nything issuing or moving with continued succession of parts”; “[a] continued current or course; current; drift.” Ibid. (emphases added). The definition of the verb form of “stream” contains a similar emphasis on continuity: “[t]o issue or flow in a stream; to issue freely or move in a continuous flow or course .” Ibid. (emphasis added). On these definitions, therefore, the Corps’ phrases “intermittent streams,” 33 CFR §328.3(a)(3) (2004), and “ephemeral streams,” 65 Fed. Reg. 12823 (2000), are—like Senator Bentsen’s “ ‘flowing gullies,’ ” post, at 16, n. 11 (opinion of Stevens, J.)— useful oxymora. Properly speaking, such entities constitute extant “streams” only while they are “continuous[ly] flow[ing]”; and the usually dry channels that contain them are never “streams.” Justice Kennedy apparently concedes that “an intermittent flow can constitute a stream” only “ while it is flowing,” post, at 13 (emphasis added)— which would mean that the channel is a “water” covered by the Act only during those times when water flow actually occurs. But no one contends that federal jurisdiction appears and evaporates along with the water in such regularly dry channels.

[7] It is of course true, as the dissent and Justice Kennedy both observe, that ditches, channels, conduits and the like “can all hold water permanently as well as intermittently,” post, at 17 (opinion of Stevens, J.); see also post, at 14–15 (opinion of Kennedy, J.). But when they do, we usually refer to them as “rivers,” “creeks,” or “streams.” A permanently flooded ditch around a castle is technically a “ditch,” but (because it is permanently filled with water) we normally describe it as a “moat.” See Webster’s Second 1575. And a permanently flooded man-made ditch used for navigation is normally described, not as a “ditch,” but as a “canal.” See id., at 388. Likewise, an open channel through which water permanently flows is ordinarily described as a “stream,” not as a “channel,” because of the continuous presence of water. This distinction is particularly apt in the context of a statute regulating water quality, rather than (for example) the shape of stream beds. Cf. Jennison v. Kirk, 98 U. S. 453, 454–456 (1879) (referring to man-made channels as “ditches” when the alleged injury arose from physical damage to the banks of the ditch); PUD No. 1 of Jefferson Cty. v. Washington Dept. of Ecology, 511 U. S. 700, 709 (1994) (referring to a water-filled tube as a “tunnel” in order to describe the shape of the conveyance, not the fact that it was water-filled), both cited post, at 17, n. 12 (opinion of Stevens, J.). On its only natural reading, such a statute that treats “waters” separately from “ditch[es], channel[s], tunnel[s], and conduit[s],” thereby distinguishes between continuously flowing “waters” and channels containing only an occasional or intermittent flow.

It is also true that highly artificial, manufactured, enclosed conveyance systems—such as “sewage treatment plants,” post, at 15 (opinion of Kennedy, J.), and the “mains, pipes, hydrants, machinery, buildings, and other appurtenances and incidents” of the city of Knoxville’s “system of waterworks,” Knoxville Water Co. v. Knoxville, 200 U. S. 22, 27 (1906), cited post, at 17, n. 12 (opinion of Stevens, J.)—likely do not qualify as “waters of the United States,” despite the fact that they may contain continuous flows of water. See post, at 15 (opinion of Kennedy, J.); post, at 17, n. 12 (opinion of Stevens, J.). But this does not contradict our interpretation, which asserts that relatively continuous flow is a necessary condition for qualification as a “water,” not an adequate condition. Just as ordinary usage does not treat typically dry beds as “waters,” so also it does not treat such elaborate, man-made, enclosed systems as “waters” on a par with “streams,” “rivers,” and “oceans.”

[8] Justice Kennedy contends that the Corps’ preservation of the “responsibilities and rights” of the States is adequately demonstrated by the fact that “33 States and the District of Columbia have filed an amici brief in this litigation” in favor of the Corps’ interpretation, post, at 20. But it makes no difference to the statute ’s stated purpose of preserving States’ “rights and responsibilities,” §1251(b), that some States wish to unburden themselves of them. Legislative and executive officers of the States may be content to leave “responsibilit[y]” with the Corps because it is attractive to shift to another entity controversial decisions disputed between politically powerful, rival interests. That, however, is not what the statute provides.

[9] Justice Kennedy objects that our reliance on these two clear-statement rules is inappropriate because “the plurality’s interpretation does not fit the avoidance concerns that it raises,” post, at 19—that is, because our resolution both eliminates some jurisdiction that is clearly constitutional and traditionally federal, and retains some that is questionably constitutional and traditionally local. But a clear-statement rule can carry one only so far as the statutory text permits. Our resolution, unlike Justice KENNEDY’s, keeps both the overinclusion and the underinclusion to the minimum consistent with the statutory text. Justice KENNEDY’s reading—despite disregarding the text—fares no better than ours as a precise “fit” for the “avoidance concerns” that he also acknowledges. He admits, post, at 25, that “the significant nexus requirement may not align perfectly with the traditional extent of federal authority” over navigable waters—an admission that “tests the limits of understatement,” Gonzales v. Oregon, 126 S. Ct. 904, 932 (2005) (Scalia, J., dissenting)—and it aligns even worse with the preservation of traditional state land-use regulation.

[10] Since the wetlands at issue in Riverside Bayview actually abutted waters of the United States, the case could not possibly have held that merely “neighboring” wetlands came within the Corps’ jurisdiction. Obiter approval of that proposition might be inferred, however, from the opinion’s quotation without comment of a statement by the Corps describing covered “adjacent” wetlands as those “ ‘that form the border of or are in reasonable proximity to other waters of the United States.’ ” 474 U. S., at 134 (quoting 42 Fed. Reg. 37128 (1977); emphasis added). The opinion immediately reiterated, however, that adjacent wetlands could be regarded as “the waters of the United States” in view of “the inherent difficulties of defining precise bounds to regulable waters,” 474 U. S., at 134—a rationale that would have no application to physically separated “neighboring” wetlands. Given that the wetlands at issue in Riverside Bayview themselves “actually abut[ted] on a navigable waterway,” id., at 135; given that our opinion recognized that unconnected wetlands could not naturally be characterized as “ ‘waters’ ” at all, id., at 132; and given the repeated reference to the difficulty of determining where waters end and wetlands begin; the most natural reading of the opinion is that a wetlands’ mere “reasonable proximity” to waters of the United States is not enough to confer Corps jurisdiction. In any event, as discussed in our immediately following text, any possible ambiguity has been eliminated by SWANCC, 531 U. S. 159 (2001).

[11] The dissent argues that “the very existence of words like ‘alluvium’ and ‘silt’ in our language suggests that at least some [dredged or fill material] makes its way downstream,” post, at 22 (citation omitted). See also post, at 17 (opinion of Kennedy, J.). By contrast, amici cite multiple empirical analyses that contradict the dissent’s philological approach to sediment erosion—including one which concludes that “[t]he idea that the discharge of dredged or fill material into isolated waters, ephemeral drains or non-tidal ditches will pollute navigable waters located any appreciable distance from them lacks credibility.” R. Pierce, Technical Principles Related to Establishing the Limits of Jurisdiction for Section 404 of the Clean Water Act 34–40 (Apr. 2003), available at www.wetlandtraining.com/tpreljscwa.pdf, cited in Brief for International Council of Shopping Centers et al. as Amici Curiae 26–27; Brief for Pulte Homes, Inc., et al. as Amici Curiae 20–21; Brief for Foundation for Environmental and Economic Progress et al. as Amici Curiae 29, and n. 53 (“Fill material does not migrate”). Such scientific analysis is entirely unnecessary, however, to reach the unremarkable conclusion that the deposit of mobile pollutants into upstream ephemeral channels is naturally described as an “addition . . . to navigable waters,” 33 U. S. C. §1362(12), while the deposit of stationary fill material generally is not.

[12] Nor does the passing reference to “wetlands adjacent thereto” in §1344(g)(1) purport to expand that statutory definition. As the dissent concedes, post, at 20, that reference merely confirms that the statutory definition can be read to include some wetlands—namely, those that directly “abut” covered waters. Riverside Bayview explicitly acknowledged that §1344(g)(1) “does not conclusively determine the construction to be placed on the use of the term ‘waters’ elsewhere in the Act (particularly in [§1362(7)], which contains the relevant definition of ‘navigable waters’); however, . . . it does at least suggest strongly that the term ‘waters’ as used in the Act does not necessarily exclude ‘wetlands.’ ” 474 U. S., at 138, n. 11 (emphases added).

[13] The sole exception is in Justice KENNEDY’s opinion, which argues that Riverside Bayview rejected our physical-connection requirement by accepting as a given that any wetland formed by inundation from covered waters (whether or not continuously connected to them) is covered by the Act: “The Court in Riverside Bayview . . . did not suggest that a flood-based origin would not support jurisdiction; indeed, it presumed the opposite. See 474 U. S., at 134 (noting that the Corps’ view was valid ‘ even for wetlands that are not the result of flooding or permeation’ (emphasis added)).” Post, at 16. Of course Justice Kennedy himself fails to observe this supposed presumption, since his “significant nexus” test makes no exception for wetlands created by inundation. In any event, the language from Riverside Bayview in Justice KENNEDY’s parenthetical is wrenched out of context. The sentence which Justice Kennedy quotes in part immediately followed the Court’s conclusion that “adjacent” wetlands are included because of “the inherent difficulties of defining precise bounds to regulable waters,” 474 U. S., at 134. And the full sentence reads as follows: “This holds true even for wetlands that are not the result of flooding or permeation by water having its source in adjacent bodies of open water,” ibid. (emphasis added). Clearly, the “wetlands” referred to in the sentence are only “adjacent” wetlands—namely, those with the continuous physical connection that the rest of the Riverside Bayview opinion required, see supra, at 21–23. Thus, it is evident that the quoted language was not at all a rejection of the physical-connection requirement, but rather a rejection of the alternative position (which had been adopted by the lower court in that case, see id., at 125) that the only covered wetlands are those created by inundation. As long as the wetland is “adjacent” to covered waters, said Riverside Bayview, its creation vel non by inundation is irrelevant.

[14] The allusion is to a classic story told in different forms and attributed to various authors. See, e.g., Geertz, Thick Description: Toward an Interpretive Theory of Culture, in The Interpretation of Cultures 28–29 (1973). In our favored version, an Eastern guru affirms that the earth is supported on the back of a tiger. When asked what supports the tiger, he says it stands upon an elephant; and when asked what supports the elephant he says it is a giant turtle. When asked, finally, what supports the giant turtle, he is briefly taken aback, but quickly replies “Ah, after that it is turtles all the way down.”

[15] It is unclear how much more moderate the flouting is, since Justice KENNEDY’s “significant nexus” standard is perfectly opaque. When, exactly, does a wetland “significantly affect” covered waters, and when are its effects “in contrast . . . speculative or insubstantial”? Post, at 23. Justice Kennedy does not tell us clearly—except to suggest, post, at 25, that “ ‘ “isolated” is generally a matter of degree’ ” (quoting Lei-bowitz & Nadeau, Isolated Wetlands: State-of-the-Science and Future Directions, 23 Wetlands 663, 669 (2003)). As the dissent hopefully observes, post, at 24, such an unverifiable standard is not likely to constrain an agency whose disregard for the statutory language has been so long manifested. In fact, by stating that “[i]n both the consolidated cases before the Court the record contains evidence suggesting the possible existence of a significant nexus according to the principles outlined above,” post, at 26, Justice Kennedy tips a wink at the agency, inviting it to try its same expansive reading again.

[1] Pursuant to 33 U. S. C. §§1344(g)–(h), Michigan operates its own §404 permitting program, subject to supervision from the Army Corps.

[2] Dr. Willard did not “stud[y] the upstream drainage of these sites . . . well enough to make a statement” about whether they also performed pollutant-trapping functions. 4 Tr. 96.

[3] By contrast, we “d[id] not express any opinion” on the Corps’ additional assertion of jurisdiction over “wetlands that are not adjacent to bodies of open water, see 33 CFR §323.2(a)(2) and (3) (1985).” 474 U. S., at 131–132, n. 8; see also id., at 124, n. 2 (making the same reservation). Contrary to Justice KENNEDY’s reading, ante, at 23–24 (opinion concurring in judgment), we were not reserving the issue of the Corps’ jurisdiction over wetlands adjacent to tributaries, but only reserving the issue of the Corps’ jurisdiction over truly isolated waters. A glance at the cited regulation makes this clear. Section 323.2(a)(2) refers to “[a]ll interstate waters including interstate wetlands” and §323.2(a)(3) covers “[a]ll other waters such as intrastate lakes, rivers, streams (including intermittent streams), mudflats, sandflats, wetlands, sloughs, prairie potholes, wet meadows, playa lakes, or natural ponds, the use, degradation or destruction of which could affect interstate or foreign commerce including any such waters.” See also Solid Waste Agency of Northern Cook Cty. v. Army Corps of Engineers, 531 U. S. 159, 163–164 (2001) (considering the validity of an application of §328.3(a)(3) (1999), which is substantively identical to §323.2(a)(3) (1985) and to §323.2(a)(5) (1978)). Wetlands adjacent to tributaries of traditionally navigable waters were covered in the 1985 regulation by other provisions of the regulation, namely a combination of §§323.2(a)(1) (covering traditionally navigable waters), (4) (covering tributaries of subsection (a)(1) waters), and (7) (covering wetlands adjacent to subsection (a)(4) waters).

[4] As The Chief Justice observes, the Corps and the EPA initially considered revising their regulations in response to SWANCC . Ante, at 1–2 (concurring opinion). The Chief Justice neglects to mention, however, that almost all of the 43 States to submit comments opposed any significant narrowing of the Corps’ jurisdiction—as did roughly 99% of the 133,000 other comment submitters. See U. S. General Accounting Office, Report to the Chairman, Subcommittee on Energy Policy, Natural Resources and Regulating Affairs, Committee on Government Reform, House of Representatives, Waters and Wetlands: Corps of Engineers Needs to Evaluate Its District Office Practices in Determining Jurisdiction, GAO–04–297, pp. 14–15 (Feb. 2004), http://www.gao.gov/new.items/d04297.pdf (hereinafter GAO Report) (all Internet materials as visited June 14, 2006, and available in Clerk of Court’s case file); Brief for Association of State and Interstate Water Pollution Control Administrators as Amicus Curiae . In any event, the agencies’ decision to abandon their rulemaking is hardly responsible for the cases at hand. The proposed rulemaking focused on isolated waters, which are covered by 33 CFR §328.3(a)(3) (1999) and which were called into question by SWANCC, rather than on wetlands adjacent to tributaries of navigable waters, which are covered by a combination of §§328.3(a)(1), (5), and (7) and which (until now) seemed obviously within the agencies’ jurisdiction in light of Riverside Bayview . See 68 Fed. Reg. 1994 (2003) (“The agencies seek comment on the use of the factors in 33 CFR 328.3(a)(3)(i)–(iii) . . . in determining [Clean Water Act] jurisdiction over isolated, intrastate, non-navigable waters”).

[5] Unsurprisingly, most Courts of Appeals to consider the scope of the Corps’ jurisdiction after SWANCC have unhesitatingly concluded that this jurisdiction covers intermittent tributaries and wetlands adja-cent—in the normal sense of the word—to traditionally navigable waters and their tributaries. E.g., United States v. Deaton, 332 F. 3d 698 (CA4 2003) (upholding the Corps’ jurisdiction over wetlands adjacent to a ditch that might not contain consistently flowing water but did drain into another ditch that drained into a creek that drained into a navigable waterway); Headwaters, Inc. v. Talent Irrigation Dist., 243 F. 3d 526 (CA9 2001) (treating as “waters of the United States” canals that held water intermittently and connected to other tributaries of navigable waters); United States v. Rueth Development Co., 335 F. 3d 598, 604 (CA7 2003) (observing “it is clear that SWANCC did not affect the law regarding . . . adjacency” in upholding the Corps’ jurisdiction over a wetland without finding that this wetland had a continuous surface connection to its adjacent tributary); Baccarat Fremont v. U. S. Army Corps of Engineers, 425 F. 3d 1150, 1156 (CA9 2005) (upholding the Corps’ jurisdiction over wetlands separated by berms from traditionally navigable channels and observing that “ SWANCC simply did not address the issue of jurisdiction over adjacent wetlands”); but see In re Needham, 354 F. 3d 340 (CA5 2003) (reading “waters of the United States” narrowly as used in the Oil Pollution Act of 1990).

[6] Indeed, “[t]he Corps approves virtually all section 404 permit[s],” though often requiring applicants to avoid or mitigate impacts to wetlands and other waters. GAO Report 8.

[7]According to the Sunding and Zilberman article cited by the plurality, ante, at 2, for 80% of permits the mean cost is about $29,000 (with a median cost of about $12,000). The Economics of Environmental Regulation by Licensing: An Assessment of Recent Changes to the Wetland Permitting Process, 42 Natural Resources J. 59, 63, 74 (2002) (hereinafter Sunding & Zilberman). Only for less than 20% of the permits—those for projects with the most significant impacts on wetlands—is the mean cost around $272,000 (and the median cost is $155,000). Ibid.

Of course, not every placement of fill or dredged material into the waters of the United States requires a §404 permit. Only when such fill comes from point sources—“discernible, confined and discrete conveyance[s]”—is a §404 permit needed. 33 U. S. C. §§1362(12), (14). Moreover, permits are not required for discharges from point sources engaged in, among other things, normal farming activities; maintenance of transportation structures; and construction of irrigation ditches, farm roads, forest roads, and temporary mining roads. §1344(f).

[8]Rather than defending its own antagonism to environmentalism, the plurality counters by claiming that my dissent is “policy-laden.” Ante, at 28. The policy considerations that have influenced my thinking are Congress’ rather than my own. In considering whether the Corps’ interpretation of its jurisdiction is reasonable, I am admittedly taking into account the congressional purpose of protecting the physical, chemical, and biological integrity of our waters. See 33 U. S. C. §1251(a); see also Chevron U. S. A. Inc. v. Natural Resources Defense Council, Inc., 467 U. S. 863, 837 (1984) (considering whether the agency regulation was consistent with “the policy concerns that motivated the [Clean Air Act’s] enactment”).

[9]Only 3 of the 21 amici briefs filed on petitioners’ behalf come even close to asking for one of the plurality’s two conditions. These briefs half-argue that intermittent streams should fall outside the Corps’ jurisdiction—though not for the reasons given by the plurality. See Brief for National Stone, Sand and Gravel Assn. et al. 20, n. 7; Brief for Foundation for Environmental and Economic Progress et al. 22–23; Brief for Western Coalition of Arid States 10.

[10]The plurality does suggest that “ seasonal rivers” are not “necessarily exclude[d]” from the Corps’ jurisdiction—and then further suggests that “streams” are “rivers.” Ante, at 14, n. 5. I will not explore the semantic issues posed by the latter point. On the former point, I have difficulty understanding how a “seasonal” river could meet the plurality’s test of having water present “relatively permanent[ly].” By failing to explain itself, the plurality leaves litigants without guidance as to where the line it draws between “relatively permanent” and “intermittent” lies.

[11] Indeed, in the 1977 debate over whether to restrict the scope of the Corps’ regulatory power, Senator Bentsen recognized that the Corps’ jurisdiction “cover[s] all waters of the United States, including small streams, ponds, isolated marshes, and intermittently flowing gullies.” 4 Legislative History of the Clean Water Act of 1977 (Committee Print compiled for the Senate Committee on Environment and Public Works by the Library of Congress), Ser. No. 95–14, p. 903 (1978). His proposed amendment to restrict this jurisdiction failed. Id., at 947.

[12] The plurality’s reasoning to the contrary is mystifying. The plurality emphasizes that a ditch around a castle is also called a “moat” and that a navigable manmade channel is called a “canal.” See ante, at 17, n. 7. On their face (and even after much head-scratching), these points have nothing to do with whether we use the word “stream” rather than “ditch” where permanently present water is concerned. Indeed, under the plurality’s reasoning, we would call a “canal” a “stream” or a “river” rather than a “canal.”

Moreover, we do use words like “ditch” without regard to whether water is present relatively permanently. In Jennison v. Kirk, 98 U. S. 453 (1879), for example, Justice Field used the term “ditch”—not “stream”—in describing a manmade structure that carried water year round. See also, e.g., Knoxville Water Co. v. Knoxville, 200 U. S. 22, 27 (1906) (opinion for the Court by Harlan, J.) (describing “pipes” that would continuously carry water); ante, at 20, 24 (plurality opinion) (using “channel” with reference to both intermittent and relatively permanent waters); PUD No. 1 of Jefferson Cty. v. Washington Dept. of Ecology, 511 U. S. 700, 709 (1994) (describing a “tunnel” that would carry water year round); New Orleans Water-Works Co. v. Rivers, 115 U. S. 674, 683 (1885) (opinion for the Court by Harlan, J.) (describing “conduits” that would supply water for a hotel). The plurality’s attempt to achieve its desired outcome by redefining terms does no credit to lexicography—let alone to justice.

[13] See, e.g., Screws v. United States, 325 U. S. 91, 131–134 (1945) (Rutledge, J., concurring in result); Turner Broadcasting System, Inc. v. FCC, 512 U. S. 622, 674 (1994) (Stevens, J., concurring in part and concurring in judgment); Hamdi v. Rumsfeld, 542 U. S. 507, 553–554 (2004) (Souter, J., concurring in part, dissenting in part, and concurring in judgment).

[14] I assume that Justice KENNEDY’s approach will be controlling in most cases because it treats more of the Nation’s waters as within the Corps’ jurisdiction, but in the unlikely event that the plurality’s test is met but Justice KENNEDY’s is not, courts should also uphold the Corps’ jurisdiction. In sum, in these and future cases the United States may elect to prove jurisdiction under either test.