The National Association of Home Builders (Home Builders) appeals the district court's summary judgment order dismissing its suit against the United States Department of the Interior (Interior) and its Fish and Wildlife Service (FWS).1 Nat'l Ass'n of Home Builders v. Norton, 298 F. Supp. 2d 68 (D.D.C.2003) (NAHB). The lawsuit revolves around the FWS's promulgation of a survey protocol, first in 1999 and again in revised form in 2000 (together, Protocols), that provides a methodology for the detection of an endangered subspecies of butterfly in certain areas of southern California. Before the district court, Home Builders asserted violations of the Administrative Procedure Act (APA), 5 U.S.C. §§ 701 et seq., and the Endangered Species Act (ESA), 16 U.S.C. §§ 1531 et seq., arguing that the FWS failed to comply with those statutes' notice and comment provisions. The district court dismissed the claims for lack of jurisdiction. NAHB, 298 F.Supp.2d at 80. We now affirm.

The ESA provides "a means whereby the ecosystems upon which endangered species and threatened species depend may be conserved." 16 U.S.C. § 1531(b). Under section four of the ESA, the Secretary of the Interior (Secretary) must promulgate regulations that list species deemed "endangered" or "threatened" due to, inter alia, the "present or threatened destruction, modification, or curtailment of its habitat or range." Id. § 1533(a)(1)(A), ©}(1); see also Bennett v. Spear, 520 U.S. 154, 157-58, 117 S. Ct. 1154, 137 L. Ed. 2d 281 (1997).2 The Secretary is further charged with developing and implementing a "recovery plan" for the "conservation and survival of endangered species and threatened species." Id. § 1533(f).

Once a species is designated "endangered" or "threatened," the ESA provides a variety of protections, including a prohibition on "take" of the species. Id. § 1538(a)(1). "Take" is a term uniquely defined by the ESA to mean: "to harass, harm, pursue, hunt, shoot, wound, kill, trap, capture, or collect" the listed species. Id. § 1532(19). Following the statutory labyrinth one step deeper, regulations passed pursuant to the ESA define "harm" as used in the definition of "take" to include "significant habitat modification or degradation where it actually kills or injures wildlife by significantly impairing essential behavioral patterns, including breeding, feeding or sheltering." 50 C.F.R. § 17.3; see also, generally, Endangered and Threatened Wildlife and Plants; Final Redefinition of "Harm," 46 Fed.Reg. 54,748 (1981). Thus, a landowner can effect a take of an endangered species, subjecting himself to liability under the ESA, if he alters the habitat of an endangered species in a manner that causes death or injury to a member of the species. The ESA establishes civil and criminal penalties for any person who unlawfully takes an endangered species. 16 U.S.C. § 1540. In addition to more traditional enforcement mechanisms using federal and state personnel, id. § 1540(e), the ESA contains a "citizen suit" provision that permits a private party to seek injunctive relief against any landowner "alleged to be in violation" of the ESA, id. § 1540(g)(1)(A).

Section 10 of the ESA does permit landowners and other non-federal entities to obtain a permit to "take" a listed species "if such taking is incidental to, and not the purpose of, the carrying out of an otherwise lawful activity." Id. § 1539(a)(1)(B). To obtain such a permit, the landowner must demonstrate to the Secretary through a documented conservation plan that the owner will, inter alia, minimize the impact of the taking and that the "likelihood of the survival and recovery of the species" will not be diminished by the taking. Id. § 1539(a)(2)(B)(iv). In addition, the ESA authorizes the grant of a "recovery" permit, which enables a researcher to engage in actions "for scientific purposes" that could result in a taking. Id. § 1539(a)(1)(A).

The quino checkerspot butterfly (Quino) is a small butterfly native to southwestern California and northwestern Mexico. The Quino was listed as an endangered species on January 16, 1997. Endangered and Threatened Wildlife and Plants; Determination of Endangered Status for the Laguna Mountains Skipper and the Quino Checkerspot Butterfly (Euphydryas editha quino), 62 Fed.Reg. 2313 (1997) (Listing Rule). Once abundant, only seven or eight known colonies of Quino remain in the United States, all in Riverside and San Diego counties in California. Id. at 2315. The primary suspected cause of the loss of the species is the destruction of Quino habitat through development, grazing and fragmentation. Id. at 2317-2319. See also Endangered and Threatened Wildlife and Plants; Designation of Critical Habitat for the Quino Checkerspot Butterfly (Euphydryas editha quino), 67 Fed.Reg. 18,356, 18,359 (2002) (Critical Habitat Designation). Quino require a very particular habitat to survive, owing in part to their reliance on specific host plants during the larval life stage. Listing Rule, 62 Fed.Reg. at 2314. The Quino live as adult butterflies only for a period of roughly four to eight weeks. Id. The wingspan of an adult Quino measures a mere one inch. Id. Their flight season lasts from mid-January until late April, but peaks in March and April. Id. Quino do not fly in adverse weather conditions such as rain or wind, however, which, combined with their short lifespan and small size, can make detection difficult. The FWS issued its first guidance for detecting the Quino several months after listing the butterfly as fully protected by the ESA. See U.S. Fish and Wildlife Service, Interim General Survey Protocols and Mitigation Guidelines for the Endangered Quino Checkerspot Butterfly (November 4, 1997) (Interim Protocol).

Based on information gathered during the 1998 field season, as well as consultation with scientists and species experts, the FWS revised the Interim Protocol and promulgated the "Survey Protocol for the Endangered Quino Checkerspot Butterfly (Euphydryas editha quino) for the 1999 Field Season" (1999 Protocol), reprinted in Joint Appendix (J.A.) at 87-111, on January 25, 1999. The FWS did not, however, engage in formal notice and comment proceedings in drafting the 1999 Protocol. On February 1, 1999, a notice of availability for the 1999 Protocol was published in the Federal Register.3 Notice of Availability of a Recommended Survey Protocol for the Endangered Quino Checkerspot Butterfly (Euphydryas editha quino) for the 1999 Field Season, 64 Fed.Reg. 4890 (1999) (1999 Notice of Availability). The notice of availability referred to the 1999 Protocol in both its title and text as "recommended." Id. It also provided an address where "comments," "data" and "materials concerning the survey protocol" could be sent for the FWS's consideration during the development of a revised protocol for the 2000 field season. Id. The text of the 1999 Protocol identified on a map attached to the 1999 Protocol as Appendix B "areas with no potential for Quino, with potential habitat where adult surveys may be necessary of [sic] suitable habitat occurs on a site, and with Quino habitat where adult surveys should be conducted." 1999 Protocol at 1 & App. B, reprinted in J.A. at 90, 98. It recommended, but did not mandate, habitat assessment in areas designated by the FWS as Potential Habitat Areas and adult surveys in the Adult Focused Survey Areas or if a habitat assessment indicated suitable Quino habitat. Id. The 1999 Protocol stipulated that in order to avoid take of the species, adult surveys "must be conducted by a biologist possessing a recovery permit pursuant to section 10(a)(1)(A) of the [ESA]." Id. at 1, reprinted in J.A. at 90; see 16 U.S.C. § 1539(a)(1)(A).

The FWS published a revised protocol in the year 2000. U.S. Fish and Wildlife Service, Quino Checkerspot Butterfly (Euphydryas editha quino) Year 2000 Survey Protocol (2000 Protocol), reprinted in J.A. at 112-21. Again, a notice of availability regarding the "recommended survey protocol for the 2000 field season" was published in the Federal Register. Notice of Availability of a Recommended Year 2000 Survey Protocol for the Endangered Quino Checkerspot Butterfly (Euphydryas editha quino), 65 Fed.Reg. 8188 (2000) (2000 Notice of Availability). The revisions were based on information derived from public workshops, the input of a new "recovery team," the development of the recovery plan, the 1999 survey reports and public comments. 2000 Protocol at 1, reprinted in J.A. at 113. The 2000 Protocol provides substantially more detail regarding survey methodology than the 1999 Protocol did. Compare 1999 Protocol at 1-4, reprinted in J.A. at 90-93, with 2000 Protocol at 2-6, reprinted in J.A. at 114-118. As with the 1999 Protocol, however, the FWS describes the 2000 Protocol as merely "recommended" except for "requirements for biologists conducting quino butterfly surveys under recovery permits." 2000 Protocol at 1, reprinted in J.A. at 113. Both Protocols also warn that "surveys may not be considered valid if . . . the specific survey methods described above are not followed." 2000 Protocol at 6, reprinted in J.A. at 118; see also 1999 Protocol at 4, reprinted in J.A. at 93.

Home Builders, a non-profit advocacy group that represents individuals and companies in the residential construction industry, filed suit in federal district court, alleging, inter alia, that the 1999 and 2000 Protocols constituted a "rule" subject to the notice and comment provisions of the APA. 5 U.S.C. § 553. Thus, according to Home Builders, in promulgating the Protocols the FWS exceeded its authority under the ESA by failing to comply with section 553 of the APA and section 4(b)(4) of the ESA, 16 U.S.C. § 1533(b)(4) (incorporating APA's notice and comment requirements with respect to "any regulation promulgated to carry out the purposes" of the ESA). Compl. 13-25. In NAHB, the district court dismissed the suit at the summary judgment stage, holding that the Protocols did not constitute "final agency action" and thus the court lacked jurisdiction under sections 702 and 704 of the APA. 298 F. Supp. 2d at 79. Specifically, the court found that the Protocols did not satisfy the finality test established by the United States Supreme Court in Bennett v. Spear, 520 U.S. 154, 178, 117 S. Ct. 1154, 137 L. Ed. 2d 281 (1997), because "the Protocols do not determine rights or obligations of landowners and legal consequences do not flow from them." NAHB, 298 F.Supp.2d at 76.4 Home Builders timely filed this appeal.

We review the district court's grant of summary judgment de novo. Saint Luke's Hosp. v. Thompson, 355 F.3d 690, 693 (D.C.Cir.2004). The APA authorizes judicial review of "[a]gency action made reviewable by statute and final agency action for which there is no other adequate remedy in a court."5 5 U.S.C. § 704 (emphasis added). There exists no statutory review provision in the ESA that authorizes judicial review of agency action beyond that provided for in the APA. See Cabinet Mountains Wilderness/Scotchman's Peak Grizzly Bears v. Peterson, 685 F.2d 678, 685 (D.C.Cir.1982). Thus, an agency action must be final in order to be judicially reviewable. See, e.g., Ctr. for Law and Educ. v. Dep't of Educ., 396 F.3d 1152, 1165 (D.C.Cir.2005) ("APA . . . bars review prior to final agency action."); Indep. Petroleum Ass'n of Am. v. Babbitt, 235 F.3d 588, 594 (D.C.Cir.2001) ("`[T]he requirement of a final agency action has been considered jurisdictional. If the agency action is not final, the court . . . cannot reach the merits of the dispute.'" (quoting DRG Funding Corp. v. Sec'y of Housing & Urban Dev., 76 F.3d 1212 (D.C.Cir.1996))).

The Supreme Court has established a two-part test to determine when an agency action is reviewable as "final." First, the action under review "must mark the `consummation' of the agency's decisionmaking process — it must not be of a merely tentative or interlocutory nature." Bennett v. Spear, 520 U.S. 154, 177-78, 117 S. Ct. 1154, 137 L. Ed. 2d 281 (1997) (citing Chicago & S. Air Lines, Inc. v. Waterman S.S. Corp., 333 U.S. 103, 68 S. Ct. 431, 437, 92 L. Ed. 568 (1948)). Second, the action must "be one by which `rights or obligations have been determined,' or from which `legal consequences will flow.'" Bennett, 520 U.S. at 178, 117 S. Ct. 1154 (quoting Port of Boston Marine Terminal Ass'n v. Rederiaktiebolaget Transatlantic, 400 U.S. 62, 71, 91 S. Ct. 203, 27 L. Ed. 2d 203 (1970)). The Protocols at issue in this case clearly marked the consummation of the decisionmaking process. The Protocols were published after the FWS solicited input from specialists and reviewed data from past field seasons. 1999 Protocol at 1, reprinted in J.A. at 90; 2000 Protocol at 2, reprinted in J.A. at 114. A notice of availability was published each year in the Federal Register. 1999 Notice of Availability, 64 Fed.Reg. at 4890; 2000 Notice of Availability, 65 Fed.Reg. at 8180. Ongoing revisions to the Protocols based on new data and feedback from interested parties do not negate finality. See United States Air Tour Ass'n. v. FAA, 298 F.3d 997, 1013 (D.C.Cir.2002). Thus, the pivotal issue is whether the 1999 and 2000 Protocols published by the FWS are documents "by which rights or obligations have been determined, or from which legal consequences will flow." Bennett, 520 U.S. at 178, 117 S. Ct. 1154 (internal quotation marks omitted).

Home Builders asserts three arguments to support its view that the Protocols impose legal obligations. It first claims that the Protocols are binding on their face. This argument fails to pass muster. The Protocols are consistently referred to in agency documents as "recommended," rather than mandatory. See, e.g., 1999 Notice of Availability, 64 Fed.Reg. at 4890 (referring to 1999 Protocol as "recommended" in both title and text); 2000 Notice of Availability, 65 Fed.Reg. at 8188 (same); 1999 Protocol at I, reprinted in J.A. at 88 (protocol "recommended"); 2000 Protocol at 3, reprinted in J.A. at 114 (protocol surveys "recommended"). Moreover, in a letter from the FWS Director to several members of the Congress, the agency stated that the "survey protocol does not contain any prohibitions or restrictions on land development, nor should the protocol be interpreted as such." Letter from Clark to Calvert at 1 (Apr. 16, 1999). An agency's past characterization of its own action, while not decisive, is entitled to respect in a finality analysis. See Skidmore v. Swift & Co., 323 U.S. 134, 140, 65 S. Ct. 161, 89 L. Ed. 124 (1944) (agency opinion constitutes "body of experience and informed judgment to which courts and litigants may properly resort for guidance"); see also Christensen v. Harris County, 529 U.S. 576, 587, 120 S. Ct. 1655, 146 L. Ed. 2d 621 (2000) (opinion letters entitled to respect under Skidmore "to the extent those interpretations have the power to persuade" (internal quotation marks omitted)). But cf. Appalachian Power Co. v. Envt'l Prot. Agency, 208 F.3d 1015, 1023 (D.C.Cir.2000) (disregarding "boilerplate" non-final action language in guidance document). Despite some mandatory language in the Protocols regarding how the survey should be conducted in order to maximize accuracy and minimize incidental take of the species ("Butterfly surveys may only be conducted by a biologist possessing a current recovery permit," 2000 Protocol at 2, reprinted in J.A. at 113), neither the 1999 nor the 2000 Protocol contains any language compelling a landowner to conduct a survey at all. "Protocol surveys," according to the FWS, merely "are recommended for all sites partially or completely within . . . survey areas." 2000 Protocol at 2, reprinted in J.A. at 113 (emphasis added).

Given the voluntary nature of the language contained in the Protocols, it is futile for Home Builders to argue that the Protocols are binding on their face. Home Builders goes on to argue, however, that the Protocols constitute final agency action because in practice they have a coercive effect on both landowners and local governments, in effect compelling compliance with the Protocols in order to avoid prosecution for unlawful take of the species. Appellant's Br. at 24-35. Finality resulting from the practical effect of an ostensibly non-binding agency proclamation is a concept we have recognized in the past. See Gen. Elec. Co. v. Envt'l Prot. Agency, 290 F.3d 377, 383 (D.C.Cir.2002) ("if the language of the document is such that private parties can rely on it as a norm or safe harbor by which to shape their actions, it can be binding as a practical matter"); McLouth Steel Prods. Corp. v. Thomas, 838 F.2d 1317, 1321 (D.C.Cir. 1988) (agency action, though facially non-binding, "created a norm with present day binding effect") (internal quotation marks omitted). Nevertheless, if the practical effect of the agency action is not a certain change in the legal obligations of a party, the action is non-final for the purpose of judicial review. See DRG Funding Corp. v. Sec'y of Housing and Urban Dev., 76 F.3d 1212, 1214 (D.C.Cir.1996) (agency order non-final that "does not itself adversely affect complainant but only affects his rights adversely on the contingency of future administrative action") (internal quotation marks omitted). Thus, we have held that "[p]ractical consequences, such as the threat of having to defend itself in an administrative hearing should the agency actually decide to pursue enforcement, are insufficient to bring an agency's conduct under our purview." Indep. Equip. Dealers Ass'n v. Envt'l Prot. Agency, 372 F.3d 420 (2004) (internal quotation marks omitted).

There is nothing in the record to support Home Builders' claim that the Protocols could affect the outcome of an enforcement proceeding. Just as compliance with the Protocols does not provide a "safe harbor" from prosecution, Gen. Elec., 290 F.3d at 383, failure to comply does not change the legal burden placed on the government (or on a private party in a citizen suit) in a suit for injunctive relief: the enforcing party must convince the court that "the alleged activity has actually harmed the species or . . . will actually, as opposed to potentially, cause harm to the species." Am. Bald Eagle v. Bhatti, 9 F.3d 163, 166 (1st Cir.1993). The results of the survey, the decision to conduct a site assessment but not a survey, the failure to perform either, or any other course of action by the landowner would constitute just one piece of the evidence necessary to obtain an injunction. At the time of any enforcement proceeding, the landowner can challenge the soundness of the Protocols' methodology, Pac. Gas & Elec. Co. v. Fed. Power Comm'n, 506 F.2d 33, 39 (D.C.Cir.1974) (methodology "subject to complete attack"), and to demonstrate that no take is likely to result from his actions. AT & T v. Equal Employment Opportunity Comm'n, 270 F.3d 973, 976 (D.C.Cir. 2001) (Agency guidance "has force only to the extent the agency can persuade a court to the same conclusion.").6 Moreover, the ESA has not brought any enforcement action against a landowner for failing to comply with the Protocols. Thus, in the absence of any record evidence to the contrary, it appears that the scope of a landowner's liability under section nine of the ESA remains exactly as it was before the Protocols' publication: a complete prohibition on "take" of any endangered species. 16 U.S.C. § 1538(a)(1)(B); see Indep. Equip. Dealers, 372 F.3d at 428 (agency action "left the world just as it found it, and thus cannot be fairly described as implementing, interpreting, or prescribing law or policy"); Reliable Automatic Sprinkler Co. v. Consumer Prod. Safety Comm'n, 324 F.3d 726, 732 (D.C.Cir.2003) (Agency action not reviewable if action does not "impos[e] any obligation ..., deny [ ] any right ..., or fix[ ] any legal relationship.").

Home Builders' argument that the Protocols exert a coercive effect on local governments is likewise unavailing. Home Builders asserts that local permitting agencies have adopted the Protocols to guard against their own potential liability under section nine of the ESA. See, e.g., Strahan v. Coxe, 127 F.3d 155, 164 (1st Cir.1997) (state regulatory scheme constituted "continuing violation" of section nine of ESA); United States v. Town of Plymouth, 6 F. Supp. 2d 81, 91 (D.Mass.1998) (inadequate shoreline management "harmed" protected species). Drawing on Appalachian Power, Home Builders characterizes this case as one in which the agency action "leads private parties or State permitting authorities to believe that it will declare permits invalid unless they comply with the terms of the document." 208 F.3d at 1021. Yet in Appalachian Power we were addressing a very different statutory scheme from the one at issue here. Appalachian Power involved the Clean Air Act, 42 U.S.C. §§ 7401 et seq., which establishes an intricate permitting process that involves federal review of state operating permits. See Appalachian Power, 208 F.3d at 1017. The FWS does not have comparable authority under the ESA to "declare permits invalid," id. at 1021. Furthermore, while the record demonstrates that Riverside and San Diego counties have adopted the 1999 Protocol as part of their building permit process, there is no evidence that helps to reveal the local officials' reason for doing so, much less evidence that local officials were coerced by the FWS. While the FWS did issue a letter to the City of Thousand Oaks, California in response to a draft environmental impact report for a proposed golf course in which it noted that the location of the golf course was within a potential habitat area for Quino, Letter from Noda to Smith (Feb. 22, 1999), such warning is within the authority of the FWS under section nine of the ESA. Marbled Murrelet v. Babbitt, 83 F.3d 1068, 1074 (9th Cir.1996). Thus, the record is inadequate to support a finding that the Protocols have the practical effect of binding interested parties to their terms. Gen. Elec. Co., 290 F.3d at 383; McLouth Steel, 838 F.2d at 1321.

Home Builders' third argument is that the Protocols constitute final agency action because they cabin the agency's discretion. Appellant's Br. at 36. In Cmty. Nutrition Inst. v. Young, 818 F.2d 943 (D.C.Cir. 1987), this court concluded that an FDA action level for alfatoxins in corn was a substantive rule because the language of the action level resulted in a "cabining of [the] agency's prosecutorial discretion." Id. at 948. The language of the Protocols at issue here is not nearly as severe as that in Community Nutrition. The regulation in Community Nutrition stated that "an action level . . . may be established to define the level of contamination at which food will be deemed to be adulterated." Id. at 947 (emphasis in original). The FDA had also made statements that any shipment exceeding the stated toxin level would "be considered adulterated and subject to condemnation." Id. at 948 (emphasis in original). No such binding language appears in either of the Protocols. Non-conforming surveys may be accepted. 2000 Protocol at 6. Conforming surveys may also be rejected through the "false negative" designation.7 There have been no enforcement actions that indicate whether the FWS considers itself bound by survey results. Thus, there is insufficient evidence in the record to conclude that either of the Protocols binds the agency sufficiently to make it a substantive rule under the reasoning of Community Nutrition.

For the foregoing reasons, the judgment of the district court is affirmed.

So ordered.