ORDER
The opinion filed May 31, 2001, slip op. 6689 and appearing at 253 F.3d 1137 (9th Cir.2001) is amended as follows: At slip op. 6707 and 253 F.3d at 1147, delete the last paragraph and replace with:
The district court held that the NMFS acted “arbitrarily and capriciously” when it concluded that three sales of timber that grew in or partly in riparian reserves, Salvage II, Sugar Pine Density Management, and Little River, were consistent with ACS objectives. Little River was a small sale to be permitted under a research exception. The NMFS concluded that the other two sales were “not likely to adversely affect” listed salmonids. We find nothing in the rec-oi'd to call into question NMFS’ opinions with respect to these sales. Accordingly, we vacate the order appealed from insofar as it prohibited those three sales. With the exceptions noted, the district court order was free from error, and is affirmed. The appellees are entitled to costs on appeal.
VACATED IN PART AND AFFIRMED IN PART.
With the opinion thus amended, the panel has unanimously voted to deny National Marine Fisheries Service’s petition for rehearing and Douglas Timber Operators’ and Northwest Forestry Association’s petition for rehearing.
OPINION
GOODWIN, Circuit Judge:
Six environmental organizations sued the National Marine Fisheries Service [1031]*1031(“NMFS”) for declaratory and injunctive relief to challenge four biological opinions which had the effect of clearing the way for 28 proposed timber sales in the Umpqua River watershed in southwestern Oregon. The district court granted substantial relief and the defendant agency, together with intervening timber operators, appeal.
The Pacific Coast Federation of Fishermen’s Associations, Inc. and five other organizations representing fishermen and environmental concerns are collectively referred to as “Pacific Coast.” Their principal claim is that the “no jeopardy” opinions issued by NMFS filed in Seattle, where the agency has its regional headquarters, were arbitrary and inadequately supported by the “best available science” as required by the Endangered Species Act (“ESA”). At the heart of the controversy is the impact of proposed timber sales on the Umpqua River cutthroat trout and the Oregon Coast coho salmon.1 Douglas Timber Operators (“DTO”) and the Northwest Forestry Association were allowed to enter the cases as defendant-intervenors. The cases have been consolidated for this appeal.
Pacific Coast alleged that NMFS acted arbitrarily and capriciously in reaching the conclusion that the proposed timber sales are not likely to jeopardize the continued existence of the listed species. The district court found that NMFS had acted arbitrarily and capriciously by assessing Aquatic Conservation Strategy (“ACS”) compliance only at the watershed level, by failing to evaluate short-term degradations, and by failing to fully and sufficiently incorporate the watershed analysis consistently with the “best available science” requirements set by the ESA. The district court granted summary judgment in favor of Pacific Coast. Both NMFS and DTO filed timely appeals.
The DTO assert that the publication of the challenged biological opinions by NMFS is not a final agency action within the meaning of the Administrative Procedures Act, 5 U.S.C. § 704, and, therefore, that the district court did not have-jurisdiction. The DTO also challenge the venue in the Western District of Washington, asserting that the appropriate defendants are the Bureau of Land Management (“BLM”) and Forest Service, whose proposed timber sales prompted this litigation, and whose headquarters are in Portland, in the District of Oregon.
JURISDICTION
The NMFS issued four biological opinions stating that 23 timber sales in the Umpqua River Basin were not likely to jeopardize the continued existence of the Umpqua cutthroat trout and the Oregon Coast coho salmon. The proposed sales are within the range of the northern spotted owl, and therefore fall within the region covered by the Northwest Forest Plan (“NFP”). The United States Forest Service (“USFS”) and the BLM adopted the NFP in 1994. The plan was designed to provide a comprehensive management program for 24.5 million acres of federal forest lands throughout the range of the spotted owl. See Seattle Audubon Society v. Lyons, 871 F.Supp. 1291, 1304 (W.D.Wash.1994), aff'd 80 F.3d 1401 (9th Cir.1996). One of the key components of [1032]*1032the NFP is the ACS, a comprehensive plan designed to maintain and restore the ecological health of the waterways in the federal forests.
There are four components to the ACS: (1) key watersheds (the best aquatic habitat, or hydrologically important areas), (2) riparian reserves (buffer zones along streams, lakes, wetlands and mudslide risks), (3) watershed analysis (to document existing and desired watershed conditions), and (4) watershed restoration (a long-term program to restore aquatic ecosystems and watershed health). The ACS also has binding standards and guidelines that restrict certain activities within areas designated as riparian reserves or key watersheds. Additionally, ACS has nine objectives designed to maintain or restore properly functioning aquatic habitats.
When a timber sale or other project is proposed for the NFP region, it is initially subject to an internal planning process by the action agency, either the IJSFS or the BLM. The action agency then creates a team of biologists and other resource management specialists to incorporate the NFP requirements, including ACS standards and guidelines. A biologist on the team uses a Matrix of Pathways and Indicators (the “MPI”) and a checklist developed by NMFS to assess the project’s effect on listed species. The MPI and checklist help the biologist to analyze 18 different habitat indicators and determine whether they are properly functioning, at risk, or not properly functioning. The biologists also determine whether the proposed action is likely to restore, maintain, or degrade the indicator. Projects that receive either zero or only one degrade checkmark are considered “not likely to adversely affect” listed species.
Those projects determined “likely [to] adversely affect” listed species, i.e., those that received one or more degrade check-marks, are referred to a Level 1 Team. This team is made up of biologists from various agencies. It reviews the proposed project for ACS consistency. The team can suggest changes in the plan to bring it into ACS compliance.
If the Level 1 Team agrees that the project complies with ACS, it then forwards the project to NMFS for formal consultation. Otherwise, the team elevates the review to a Level 2 Team, and the project undergoes the same review process. Failure to reach a consensus elevates the project to a Level 3 Team. Once one of these three teams approves the project, it goes to NMFS for ESA consultation.
The NMFS must review the project pursuant to Section 7 of ESA, which requires federal agencies to “insure that any action authorized, funded, or carried out by such agency ... is not likely to jeopardize the continued existence of’ any species listed as threatened or endangered under the ESA. 16 U.S.C. § 1536(a)(2). Then, NMFS must issue a Biological Opinion.
Pacific Coast sued earlier to challenge the first NMFS opinions with regard to several of the same proposed timber sales in Pacific Coast Federation of Fishermen’s Associations, Inc. v. National Marine Fisheries Service, No. C97-775R (W-D.Wash., May 29, 1998) (“PCCFA I”). Pacific Coast challenged in the district court NMFS’s Programmatic Biological Opinion and three other site-specific biological opinions.
Reviewing the Programmatic Biological Opinion in PCFFA I, the district court held that NMFS may assume that projects that are consistent with ACS are unlikely to jeopardize the continued existence of a listed species. Jurisdiction in that litigation was not challenged, and there was no appeal.
[1033]*1033The court invalidated the site-specific biological opinions in the earlier case because the opinions lacked a basis on which NMFS could conclude that the degrade checkrnarks indicated on MPI would have only minor and transitory effects. The agency reinitiated the consultation process after clarifying the documentation required to show ACS consistency and articulating guidance on the "proper" use of MPI in the analysis at the various scales. Using these new procedures, NMFS issued the four biological opinions challenged in this case.
Pacific Coast brought this action under EISA, 16 U.S.C. § 1536. The district court found jurisdiction to adjudicate the claim under 28 U.S.C. § 1331 and under the Administrative Procedure Act, 5 U.S.C. § 702. This court reviews questions of jurisdiction de novo. See Ecology Center, Inc. v. USFS, 192 F.3d 922, 924 (9th Cir.1999).
The DTO assert that the proper defendants are USFS and BLM and that claims against those entities can be brought only in the District of Oregon. They also assert that USFS and BLM are indispensable parties that should have been joined, and that in their absence the district court acted without a complete administrative record.
FINAL AGENCY ACTION
The DTO argue that the challenged biological opinions are not final agency actions. See 5 U.S.C. § 704. Only final agency decisions are subject to review under the APA. See Ohio Forestry Ass'n, Inc. v. Sierra Club, 523 U.S. 726, 732, 118 S.Ct. 1665, 140 L.Ed.2d 921 (1998), and Ecology Center, Inc., 192 F.3d at 924-26. The NMFS has not joined in the jurisdictional challenge.
The DTO argue that Pacific Coast has chosen the wrong target in an effort to stop all logging in a large part of Western Oregon by seeking to overturn the opinions of NMFS which are only interlocutory in the decision making process of the Forest Service and BLM, whose respective plans to approve the timber sales in the Umpqua River watershed are the real target of this suit. We do not accept that characterization.
The DTO attempt to distinguish Bennett v. Spear, 520 U.S. 154, 117 S.Ct. 1154, 137 L.Ed.2d 281 (1997), in attacking jurisdiction in these cases. The Supreme Court held in Bennett that a jeopardy opinion was final agency action because it effectively stopped further proceedings by the action agency. The Court reasoned that a jeopardy opinion has "direct and appreciable legal consequences," id. at 178, 117 S.Ct. 1154, because it "alters the legal regime to which the action agency is subject," id. at 169, 117 S.Ct. 1154. In the case before us, NMFS issued a "no jeopardy" opinion, which became this agency's final action. We have found no authority for the proposition that while a "jeopardy" opinion is reviewable as a final agency action, a "no jeopardy" opinion is not final and reviewable.
This court, following Bennett v. Spear, applied the two-part test for ascertaining finality of agency action in Ecology Center, Inc. v. United States Forest Service, 192 F.3d at 925-26. We held that for an administrative agency action to be considered final, "(1) the action should mark the consummation of the agency's decision making process; and (2) the action should be one by which rights or obligations have been determined or from which legal consequences flow." See id. at 925.
This no-jeopardy opinion satisfies the first part of our test because the issuance of a biological opinion marks the "consum- [1034]*1034mation” of NMFS’s consultation process. See id. The opinion meets the second part of the test because it “alters the legal regime” and has direct and appreciable legal consequences. As a practical matter, the opinion and its accompanying Incidental Take Statement grant immunity to the proposed actions of other agencies required to obtain an NMFS opinion before proceeding with their own actions, which these plaintiffs seek to block.
We are satisfied that the trial court had jurisdiction, and that BLM and the Forest Service were not necessary parties. Venue, accordingly, was properly placed in the Western District of Washington.
THE MERITS
Agency decisions under ESA are governed by the Administrative Procedure Act, which requires an agency action to be upheld unless it is found to be “arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law.” 5 U.S.C. § 706(2)(A); Friends of the Earth v. Hintz, 800 F.2d 822, 830-31 (9th Cir.1986). This deferential standard is designed to “ensure that the agency considered all of the relevant factors and that its decision contained no ‘clear error of judgment.’ ” Arizona v. Thomas, 824 F.2d 745, 748 (9th Cir.1987) (quoting Citizens to Preserve Overton Park, Inc. v. Volpe, 401 U.S. 402, 416, 91 S.Ct. 814, 28 L.Ed.2d 136 (1971)). Agency action should be overturned only when the agency has “relied on factors which Congress has not intended it to consider, entirely failed to consider an important aspect of the problem, offered an explanation for its decision that runs counter to the evidence before the agency, or is so implausible that it could not be ascribed to a difference in view or the product of agency expertise.” Motor Vehicle Mfrs. Ass’n v. State Farm Mutual Auto. Ins. Co., 463 U.S. 29, 43, 103 S.Ct. 2856, 77 L.Ed.2d 443 (1983). Essentially, we must ask “whether the agency ‘considered the relevant factors and articulated a rational connection between the facts found and the, choice made.’ ” Natural Resources Defense Council v. United States Dep’t of the Interior, 113 F.3d 1121, 1124 (9th Cir.1997) (quoting Resources, Ltd. v. Robertson, 35 F.3d 1300, 1304 (9th Cir.1993), in turn quoting Pyramid Lake Paiute Tribe of Indians v. United States Dep’t of the Navy, 898 F.2d 1410, 1414 (9th Cir.1990)). A biological opinion may also be invalid if it fails to use the best available scientific information as required by 16 U.S.C. § 1536(a)(2). See Greenpeace Action v. Franklin, 14 F.3d 1324, 1336 (9th Cir. 1992).
Pacific Coast argued, and the district court agreed, that NMFS acted arbitrarily and capriciously by (1) ignoring site-specific project effects and limiting its ACS compliance analysis to the watershed scale, (2) focusing on a long-term evaluation of ACS compliance that effectively masks all short-term impacts that may have adverse effects on listed species, (3) failing to consider activities on federal lands that might adversely affect salmonid species, (4) “tier-ing” to BLM or USFS determinations of ACS consistency for Projects in Riparian Reserves where no aquatic benefits have been identified, and (5) failing to adequately consider, fully incorporate, or adequately explain deviations from the watershed analysis recommendations, which are designed to accomplish ACS objectives.
One preliminary matter must be addressed to avoid confusion. The NMFS argues that Pacific Coast and the district court inappropriately have required NMFS to serve as a review board or oversight committee for BLM and USFS determinations of ACS consistency. This argument appears significant, but in fact lacks substance. The NMFS is required [1035]*1035under NFP to determine whether or not a project is likely to adversely affect a listed species. The NMFS is not required by NFP to determine ACS consistency. However, in PCFFA I, the district court held that NMFS was permitted to assume that implementation of projects under USFS’s Land and Resource Management Plan (“LRMP”) or BLM’s Resource Management Plan (“RMP”) would result in “no jeopardy” to the listed fish species if those projects were conducted in accordance with ACS. Therefore, because NMFS is allowed to equate ACS consistency with a no jeopardy finding, NMFS chooses to inquire into ACS consistency. Presumably, other methods of reaching a jeopardy determination are available to NMFS. The coincidence of ACS consistency inquiries is immaterial. The NMFS’s primary obligation is to determine a project’s effect on listed fish species. The action agencies, as part of their analyses, must also determine ACS consistency. That they are able to discharge dissimilar duties by the same means does not allow either party to fail to undertake its responsibilities.
WATERSHED SCALE ACS CONSISTENCY
In determining ACS consistency for the 23 timber projects challenged in this case, NMFS analyzed the projects’ consistency with ACS at the watershed level. A watershed, or fifth field, generally covers between 20 to 200 square miles of land. This equates to between 12,800 and 128,000 acres. The largest watershed considered with reference to projects at issue here is 350 square miles, or 224,000 acres. By contrast, a project site generally covers only a few sections (square miles) or fractions of sections. The NMFS conducts its analysis of the program by assessing the effects of any project level degradation on the entire watershed. Any degradation that cannot be measured at the watershed level is considered to be consistent with both ACS standards and objectives and therefore warrants a “no jeopardy” finding.
Pacific Coast contends that the watershed measure effectively masks all project level degradation. This argument raises two questions: (1) whether, because a 128 acre project represents only 1% to 0.1% of a watershed, any degradation would be perceptible at the watershed level; and (2) whether any effect was given to the cumulative degradation in an ACS. In PCFFA I, the court held that NMFS cannot reach a no jeopardy determination without analyzing whether the site-specific projects are in fact complying with ACS. See PCFFA I at 30. The court found that evidence of site specific degradation and the lack of mitigation showed that NMFS rationally could not find the “proposed actions ...' consistent with ACS’s mandate that agencies maintain and restore aquatic systems within the range of the northern spotted owl.” Id. It is clear from the court’s order that application of ACS at the project level explained how NMFS could assume, for that project, that a proposed action would not jeopardize listed fish. The emphasis on site-specific evaluation is evident in the district court’s opinion in PCFFA I, at 24.
The NMFS contends that the proper level to evaluate ACS consistency is the watershed, because NFP and ACS are aimed at maintaining and restoring millions of acres of forest lands. Given that overall protection of forest and water resources is the concern of both NFP and ACS, it does not follow that NMFS is free to ignore site degradations because they are too small to affect the accomplishment of that goal at the watershed scale. For some purposes, the watershed scale may be correct, but NFP does not provide support for so limiting NMFS review. The [1036]*1036purpose of ACS is to maintain and restore ecosystem health at watershed and landscape scales to protect habitat for fish and other riparian-dependent species and resources and restore currently degraded habitats. This general mission statement in NFP does not prevent project site degradation and does nothing to restore habitat over broad landscapes if it ignores the cumulative effect of individual projects on small tributaries within watersheds. The agency also must determine “how the proposed project or management action maintains the existing condition or moves it within the range of natural variability.” Record of Decision for Amendments to Forest Service and Bureau of Land Management Planning Documents Within the Range of the Northern Spotted Owl (hereinafter “Record of Decision for the Northern Spotted Owl”), Attachment A, at B-10 (April 13, 1994). The NMFS relies on this requirement to show that consistency will be attained at the watershed level. However, it is unclear whether NMFS performed an analysis of the cumulative effect of small degradations over a whole watershed. Pacific Coast asserts that NMFS did not consider cumulative effect. The NMFS had an opportunity to place in the record evidence demonstrating that it considered cumulative effect. We find nothing to show that it did. Appropriate analysis of ACS compliance is undertaken at both the watershed and project levels.
Pacific Coast argues that the Forest Ecosystem Management Assessment Team (“FEMAT”) scientific team, which developed ACS, believed that ACS was to be implemented “at four spatial scales: regional, province/river basin, watershed, and site.” Pacific Coast also argues that NMFS has indicated that the “accumulation of effects at the landscape level from numerous actions, if not fully arrested at the project scale, would reduce the likelihood of both survival and recovery of the species.” Although the NFP, FEMAT, and ACS do not appear to address the proper scale for implementation of ACS, they explain that spatial levels should be considered and that watershed consistency is a primary goal. See Record of Decision for the Northern Spotted Owl, at B-9 and FEMAT, Forest Ecosystem Management: An Ecological, Economic, and Social Assessment (July 1998), at V-58. However, the record contains no proof that the cumulative effect of site specific degradation was considered in reaching a no jeopardy opinion at the regional watershed level.
The district court’s earlier decision to allow NMFS to assume no jeopardy from an ACS consistency finding appears to be linked to the belief that ACS consistency was to be measured at the project level. This approach seems reasonable as far as it goes. Any project that maintains or restores fish habitat presumably would not jeopardize the survival of the species. However, a project that degrades habitat at the project level must be included in any realistic study at the watershed scale. Its disregard of projects with a relatively small area of impact but that carried a high risk of degradation when multiplied by many projects and continued over a long time period is the major flaw in NMFS study. Without aggregation, the large spatial scale appears to be calculated to ignore the effects of individual sites and projects. Unless the effects of individual projects are aggregated to ensure that their cumulative effects are perceived and measured in future ESA consultations, it is difficult to have any confidence in a wide regional no-jeopardy opinion. Failure to account adequately for the cumulative effects of the various projects undermines the assumptions that the district court authorized NMFS to make in PCFFA I. If the effects of individual projects are diluted to insignificance and not aggregated, [1037]*1037then Pacific Coast is correct in asserting that NMFS’s assessment of ACS consistency at the watershed level is tantamount to assuming that no project will ever lead to jeopardy of a listed species.
Pacific Coast notes that many of these sales are located in areas that are already considered “not properly functioning,” but still NMFS requires MPI to show a “measurable worsening of those conditions across the entire watershed.” Pacific Coast contends that biological opinions are issued for projects in the same watersheds without any mention of each other. If in fact NMFS disregards these effects as “localized” when they can have significant aggregate effects, it acts arbitrarily and capriciously.
The FEMAT report, which was instrumental in developing ACS, emphasized the importance of curtailing incremental aquatic habitat degradation because the effects of numerous actions can cause significant damage to fish species and their habitat. See FEMAT, Forest Ecosystem Management: An Ecological, Economic, and Social Assessment V-2 (1993). NMFS’s assuming away site-specific degradations that could lead to a jeopardy finding contradicts the purpose of ESA and is arbitrary. Any effect on a particularly important spawning area should show up as a degrade rating for the entire watershed. Confirming that proper aggregation occurs is central to a determination whether the district court’s assumptions under the site-specific ACS consistency regime still hold true under the watershed scale regime.
DISREGARDING SHORT-TERM EFFECTS
Pacific Coast challenged NMFS’s evaluation of ACS consistency over a time frame of 10 to 20 years. The district court agreed. The court found that “NMFS has failed to adequately assess the short term impacts of the timber sales and ... has failed to adequately explain its assumption that passive restoration will adequately mitigate the adverse impacts of logging.” The district court found that the “NMFS could not rationally conclude, based on the evidence before it, that evaluating only long-term impacts of agency activities satisfied its mandate to ensure ACS compliance. Its failure, therefore, to evaluate the short-term impacts, (i.e. impacts that would manifest in less than a ten-year period) was also arbitrary and capricious.” The district court’s order requires NMFS to evaluate ACS consistency immediately after the project action is completed.
We find nothing in the record to authorize NMFS to assume away significant habitat degradation. Each of the biological opinions challenged acknowledges project-scale degradations but then deems that degradation inconsequential. Under the practice adopted by NMFS, only degradations that persist more than a decade and are measurable at the watershed scale will be considered to degrade aquatic habitat. This generous time frame ignores the life cycle and migration cycle of anadro-mous fish. In ten years, a badly degraded habitat will likely result in the total extinction of the subspecies that formerly returned to a particular creek for spawning.
The NMFS predicts that more trees will grow within the watershed during the ensuing decade than are cut in the proposed project and, therefore, concludes that the “short-term” and “localized” effects of the logging will be naturally mitigated by regrowth. This optimism may be justified for the purpose of counting trees, but for the purpose of counting anadromous fish, it is wholly unrealistic. Pacific Coast contends that there is no scientific evidence in the record to support the con-[1038]*1038elusion that natural vegetation regrowth will adequately mitigate the degradation caused by the logging projects and ensure that fish that never hatched could return to the recovered spawning habitat. We agree.
The record contains the expert opinion of a Level 1 Team biologist that such reliance on projected “restoration” is “scientifically unsound.” The NMFS does not and cannot explain adequately its disregard of short-term effects.
The NMFS never disputes that short-term effects have the potential to jeopardize listed fish populations. On the contrary, NMFS believes that the next few generations will be critical to Umpqua River anadromous species. In the Programmatic Biological Opinion, NMFS states that “even a low level of additional impact to any life form, especially the anadromous form which is at critically low levels, may reduce the likelihood of survival and recovery of the ESU as a whole.” Given the importance of the near-term period on listed species survival it is difficult to justify NMFS’s choice not to assess degradation over a time frame that takes into account the actual behavior of the species in danger.
NON-FEDERAL LANDS
The district court properly rejected the PCFFA argument as to the proper treatment of non-federal lands. As the court noted, that issue had been disposed of in PCFFA I.
ACS CONSISTENCY DETERMINATIONS IN RIPARIAN RESERVES
The district court held that the NMFS acted “arbitrarily and capriciously” when it concluded that three sales of timber that grew in or partly in riparian reserves, Salvage II, Sugar Pme Density Management, and Little River, were consistent with ACS objectives. Little River was a small sale to be permitted under a research exception. The NMFS concluded that the other two sales were “not likely to adversely affect” listed salmonids. We find nothing in the record to call into question NMFS’ opinions with respect to these sales. Accordingly, we vacate the order appealed from insofar as it prohibited those three sales. With the exceptions noted, the district court order was free from error, and is affirmed. The appellees are entitled to costs on appeal.
VACATED IN PART AND AFFIRMED IN PART.