(dissenting) — It is notoriously difficult to prove a negative. However, under the guise of “best available science,” the majority requires Ferry County to do precisely that — to prove the absence in Ferry County of several species of concern to the “environmental” groups attacking the county’s local planning. The majority, and the hearings board opinion it affirms, arrives at this unrealistic burden by substituting undocumented species concerns for county decisions delegated to local discretion under the Growth Management Act (GMA), chapter 36.70A RCW.
¶35 The result is analogous to requiring designation of critical habitat for the sasquatch, a species which the county and its biologist expert could not prove is not present. Repeated reports of sasquatch sightings do not, in my view, constitute “best available science” nor require counties to hire expensive experts to disprove its presence. As further noted below, this case is more extreme. Unlike the sasquatch, several of these supposed endangered species have no reported sightings in Ferry County.
¶36 This unrealistic burden will frustrate GMA compliance for sparsely populated and financially distressed coun*845ties like Ferry County, at least so long as state agencies such as Department of Ecology (DOE) and Department of Fish and Wildlife (DFW) do not provide good faith input or assistance. I dissent.
I. FACTS
¶37 As part of GMA compliance proceedings, the Eastern Washington Growth Management Hearings Board (Board), by earlier order, required Ferry County to amend its comprehensive plan to comply with RCW 36.70A.172 by including “best available science” in the designation and protection of fish and wildlife habitat conservation areas. See Concerned Friends of Ferry County v. Ferry County, No. 97-1-0018, 1998 WL 498783, at *2-3, 1998 GMHB LEXIS 279, at *2-3 (E. Wash. Growth Mgmt. Hr’gs Bd. July 31, 1998). Because of the county’s very limited resources, it was fortunate to have a resident biologist expert, Dr. Donald E. McKnight, who agreed to assist the county with its good faith GMA compliance efforts.
¶38 Dr. McKnight has extensive credentials. He received a bachelor’s degree in wildlife management and a master’s degree in wildlife ecology from Washington State University, a prominent institution of higher learning geographically close to Ferry County. Administrative Record (AR) at 227. He also received a PhD in wildlife ecology from Utah State University. Id.
¶39 In addition to his impressive education, Dr. McKnight gained 20 years of work experience with the Alaska Department of Fish and Game, which researches, manages, and protects fish and wildlife resources in our largest state notable for its abundant and varied wildlife resources. Id. Dr. McKnight held progressively responsible positions, including waterfowl biologist, wildlife research chief, regional wildlife supervisor, and chief wildlife planner. Id. He knows his wildlife.
¶40 Dr. McKnight was asked to give his expert opinion “regarding the current status and distribution in Ferry *846County of a number of bird and mammal species, with special attention to the potential adverse effects of development in [Ferry] county upon these animals and their habitats.” AR at 225. The species were identified by DFW officials, by letter. That agency chose not to appear in the proceedings to directly present its concerns and expertise (which would have subjected its opinions to the test of cross-examination).
¶41 Dr. McKnight relied on his extensive background and expertise in wildlife, supplemented with research. In his research, Dr. McKnight also used secondary sources for some bird species,12 as well as field guides and big game texts. He concluded that most of the species of expressed concern were not even present in the county and thus were not in need of protection under the comprehensive plan.
¶42 The following summarizes Dr. McKnight’s several analyses and rebuttal of the challenger’s claims by species:
• Sandhill crane (Grus canadensis) — This species does not nest in Ferry County and there are no historical records of its doing so. Id.
• Upland sandpiper (Bartramia longicauda) — This species is known to breed only in eastern Spokane County. Although they may occasionally be seen during migration in other counties, there are no recorded observations of the species in Ferry County. Id.
• American white pelican (Pelecanus erythrorhynchos) — This species breeds almost exclusively on Crescent Island in Walla Walla County. Although there have been summer sightings in several large lakes in eastern Washington, such sightings occur only at elevations below the ponderosa pine zone. Although Ferry County has several lakes large enough to support nonbreeding pelicans, they occur only within the ponderosa pine zone. AR at 225-26.
• Pygmy rabbit (Sylvilagus idahoensis) — This species requires relatively undisturbed tall sagebrush habitat. Although there *847is approximately one-half mile of such habitat on the Colville Indian Reservation, there have been no documented sightings. It is likely that the Columbia River and Lake Roosevelt formed a natural barrier limiting the distribution of this species to the south and east of these bodies of water. AR at 226. Dr. McKnight consulted with a local DFW wildlife biologist with respect to this species. AR at 225.
• Caribou (Rangifer tarandus caribou) — This species exists only in the Selkirk Mountains of extreme northeastern Washington, northern Idaho, and southern British Columbia. The Selkirk herd numbers fewer than 100 and does not migrate as far west as Ferry County. AR at 226.
• Gray or Timber wolf (Canis lupus) — There are no known sightings of this species in Ferry County. It would require a more substantial prey base of large ungulates such as elk, moose, or caribou for survival in this area. Id.
• Ferruginous hawk (Buteo regalis) — This species lives only in steppe vegetation zones in the southern and central Columbia River Basin. There are no published records or sightings of this species in Ferry County. AR at 223.
• Peregrine falcon (Falco peregrinus) — This species was decimated throughout its historical range by the use of the chemical DDT (dichloro-diphenyl-trichloro-ethane). Following a ban on DDT, a nesting pair was sighted in Klickitat County. To date, no nesting of falcons has been sighted in Ferry County. Id.
• Grizzly Bear (Ursus arctos) — This species was decimated throughout its historical range in Washington, Oregon, and California in the late 1800’s and early 1900’s. The nearest viable populations are in British Columbia and in and around Yellowstone National Park. AR at 224.
• Lynx (Lynx canadensis) — Although this species is known to exist in Ferry County, it likely only exists in the highest elevations of the Kettle Range of Ferry County, areas exclusively controlled by the United States Forest Service. The forest service is currently surveying potential habitat, and Ferry County will use the information gleaned from this research for future potential designation. Id.
¶43 Ferry County adopted a second amended comprehensive plan based on and reflected in the scientific re*848search provided by Dr. McKnight. Not surprisingly, that comprehensive plan did not provide for the designation and protection of the habitat for the above animals where Dr. McKnight’s research conclusively demonstrated that they were not present in the county or were not in need of protection.
¶44 On March 27, 2000, the Board held a compliance hearing to determine if the county’s new comprehensive plan amendments complied with the GMA. In an order dated May 23, 2000, the Board held that “Ferry County is found in non-compliance with the Growth Management [Act] . . . for providing insufficient evidence that its failure to list all species that are endangered, threatened or sensitive is based on best available science as required by RCW 36.70A.172.” AR at 256.
II. ANALYSIS
A. Substantial Evidence of Species Listing
¶45 The GMA was enacted in 1990 and 1991 in “response to public concerns about rapid population growth and increasing development pressures in the state, especially in the Puget Sound region.” Alan D. Copsey, Including Best Available Science in the Designation and Protection of Critical Areas Under the Growth Management Act, 23 Seattle U. L. Rev. 97, 97 (1999). The GMA employs concepts that are designed for the densely populated and wealthy Puget Sound counties and imposes them on sparsely populated and financially distressed counties like Ferry County.13
¶46 The GMA requires that all counties and their cities comply with all its mandates if the county has (1) a *849population of over 50,000 and a growth rate of 10 percent during the previous 10 years or (2) a growth rate of over 20 percent in the previous 10 years regardless of population. RCW 36.70A.040(1). Counties not meeting the above criteria are permitted to “opt in” to the GMA and are courted with promised state financial assistance in meeting their compliance obligations.14 Cumulatively, the majority of the funding went to King County (21%), Kitsap County (9.65%), Pierce County (9.4%), Snohomish County (7.2%), and Spokane County (7.15%); these five counties received 55% of the funding. Ferry County received approximately 1%, and most of Washington’s smaller, poorer counties received under 2%.
f47 Ferry County decided in good faith to “opt in” (a decision they have no doubt learned to regret). Some funding was provided early on but diminished to $29,915 for the 97-99 biennium, the relevant time here, and no money was received in fiscal year 2000. Tim Trohimovich, 1000 Friends of Washington, The Growth Management Act (GMA) After More Than 10 Years: Another Look & A Response to Criticisms (Apr. 2002).
f48 In addition to Ferry County, several other counties with limited population opted in, including Benton, Columbia, Douglas, Franklin, Garfield, Kittitas, Pacific, Pend Oreille, Stevens, and Walla Walla. Richard L. Settle & Charles G. Gavigan, The Growth Management Revolution in Washington: Past, Present, and Future, 16 U. Puget Sound L. Rev. 867, 901 n.217 (1993). Once under the GMA’s mandates, by either law or choice, the decision of a county is irrevocable RCW 36.70A.040(1).
¶49 Under the GMA, all counties and cities in Washington must adopt development regulations that designate and protect critical areas, including fish and wildlife habitat conservation areas. The GMA does not define the term fish and wildlife habitat conservation areas, and the definition *850in WAC 365-190-080(5) was not adopted until April 15, 1991.
¶50 In 1995, approximately four years after enactment, the GMA was amended to clarify the standards under which counties were to designate and protect critical areas. Engrossed Substitute House Bill 1724, § 105, 54th Leg., Reg. Sess. (Wash. 1995); Laws of 1995, ch. 347, § 105, at 1561 (codified at RCW 36.70A.172). “In designating and protecting critical areas . . . counties and cities shall include the best available science in developing policies and development regulations to protect the functions and values of critical areas.” RCW 36.70A.172(1) (emphasis added). The section passed both houses of the legislature unanimously.15 See 1995 Final Legislative Report, 54th Leg., Reg. Sess. at 199. The GMA does not define the term best available science (BAS), and WAC 365-195-900 through -925, which further develop this concept, was not adopted until August 27, 2000. Unfortunately, the BAS requirement does not apply to state agencies.
¶51 In the instant case, the issue for our consideration is “ ‘whether substantial evidence supports the Board’s finding that the County did not base its species listing on the best available science.’ ” Majority at 831-32 (quoting Wash. State Supreme Court Order). In order for the majority to reach the conclusion that the record is supported by substantial evidence, it must take the following untenable position:
Ferry County’s real quarrel with the Board seems to be based on the 12 ETS species recommended by DFW — Ferry County argues that there is no evidence of species other than the lynx and bald eagle being present in Ferry County. The validity of the DFW list is largely irrelevant here; however; Ferry County could have listed whatever species it deemed appropriate if it supported its decision by BAS. . . . We review only the Board’s determination that Ferry County’s listing did *851not include BAS, not whether listing 12 species is supported by BAS.
¶52 A careful review of the Board’s decision demonstrates that the DFW list of priority species is not only relevant but was the impetus for the Board’s decision that the county did not include BAS in the development of its comprehensive plan.
¶53 The Administrative Procedure Act, chapter 34.05 RCW, is the exclusive means of judicial review of an agency action, Diehl v. W. Wash. Growth Mgmt. Hearings Bd., 153 Wn.2d 207, 103 P.3d 193 (2004), including the growth management hearings boards. In such an appeal, we review exclusively16 the hearing board’s order and the administrative record upon which it is based. See City of Burien v. Cent. Puget Sound Growth Mgmt. Hearings Bd., 113 Wn. App. 375, 382, 53 P.3d 1028 (2002). A review of the Board’s order and the record contradicts the majority’s statement that the DFW priority habitat species listing was irrelevant.
¶54 The Board’s order repeatedly compares the actions of Ferry County to the DFW priority habitat species list. The Board made three findings of fact to support its determination that the county failed to include BAS in listing fish and wildlife species for protection. The Board’s second finding of fact directly compares the DFW priority habitat species list with those protected by Ferry County under its amended comprehensive plan. That finding reads in its entirety as follows:
DFW lists the following as endangered, threatened, or sensitive species in Ferry County:
Birds Bald Eagle Threatened
Ferruginous Hawk Threatened
*852Peregrine Falcon Endangered
Sandhill Crane Endangered
Upland Sandpiper Endangered
American White Pelican Endangered
Mammals Lynx Threatened
Pygmy Rabbit Endangered
Gray Wolf Endangered
Grizzly Bear Endangered
Woodland Caribou Endangered
Fish Bull Trout Sensitive
Findings of Fact 2, AR at 246 (emphasis added).
¶55 The majority must concede that a review of the record conclusively demonstrates that not only is the Board’s finding of fact that these 12 species are present in Ferry County not supported by substantial evidence, it is completely false.
¶56 The finding erroneously states that these species are “in Ferry County.” As the majority reluctantly concedes on page 828, footnote 2, the DFW priority habit species list is actually a compilation of species that are present in the DFW’s “region one.” Region one encompasses the 10 easternmost counties of Washington, including Stevens, Pend Oreille, Lincoln, Spokane, Whitman, Walla Walla, Columbia, Garfield, and Asotin, as well as Ferry.
¶57 By assuming that just because a species is located in region one, which includes 10 counties, that it must necessarily be present in Ferry County, the Board and the majority are guilty of the logical error known as “fallacy of division.” This syllogism falsely argues that what is true of the whole must be true of each of its constituent parts. Ruggero J. Aldisert, Logic For Lawyers: A Guide to Clear Legal Thinking 222-23 (3d ed. 1997).
¶58 Dr. McKnight’s knowledge and research clearly demonstrate that the distribution of many of the species is very limited in geographical extent in eastern Washington and certainly does not extend to all 10 counties in region one. *853For one obvious example, Dr. McKnight’s research demonstrates that the upland sandpiper is known to breed only in eastern Spokane County, but not in Ferry County, in which there is no reported sighting.
¶59 The majority is unable to cite any available scientific information in the administrative record that demonstrates that all of the species on the DFW priority habitat species list exist in Ferry County. It is telling that the petitioners— and the DFW — chose not to present this claim or support it by actual evidence. Because the majority seems so willing to make an illogical leap, I conclude that it, like the Board, has substituted concerns for what the record actually demonstrates.
¶60 Further review of the Board’s order demonstrates that the erroneous finding of fact 2, which states that 12 species on the DFW priority habitat species list are present “in Ferry County,” is more than a mere scrivener’s error. Rather, the DFW priority habitat species list formed the sole basis for determining that the county failed to use BAS in designating and protecting fish and wildlife habitat conservation areas.
¶61 According to the order below, the Board was responding to the following argument by Concerned Friends of Ferry County: “Fish & Wildlife Habitat Conservation Areas fails to include best available science since it rejects best available science recommendations that Washington Department of Fish and Wildlife Priority Habitat and Species Program be incorporated into Second Amended Comprehensive Plan . . . .” AR at 251. The Board then noted that the county’s comprehensive plan did not include the DFW priority habitat species plan and the Board references DFW wildlife biologist Kevin Robinette’s written recommendation that 12 species be listed. The Board then concludes that “Ferry County is in non-compliance with the Growth Management [Act] ... for providing insufficient evidence that its failure to list all species that are endangered, threatened or sensitive is based on best available science as required by RCW 36.70A.172(1).” AR at 256.
*854¶62 After carefully examining the Board’s order, it is impossible to draw any conclusion other than that the DFW priority habitat species list formed the basis of the Board’s decision that Ferry County failed to apply BAS. In addition to the references above, the Board stated that the county’s comprehensive plan includes “only two species as Endangered Threatened Species” and referred to this as a “limited listing.” In order to proclaim that the county listed “only two species” and that the listing was “limited,” the Board needed some frame of reference or baseline. AR at 254-55. The baseline was the DFW priority habitat species list, which indicates that 12 such species may exist in a 10 county area in eastern Washington.17
¶63 Under the GMA, the Board is required to apply a “deferential standard of review” to the actions of local jurisdictions such as Ferry County. RCW 36.70A.3201. Moreover, local ordinances are presumed valid upon adoption, RCW 36.70A.320(1), and challenging parties bear the burden of demonstrating GMA noncompliance. RCW 36.70A.320(2).
¶64 Here, the statutes require that Ferry County’s comprehensive plan be presumed in compliance with the GMA. The only “evidence” of the county’s alleged failure to include BAS is the DFW priority habitat species list, neither presented by nor accompanied by qualified experts. Ironically, DFW input clearly does not meet the BAS as defined in the statute, in later WACs, or even as defined in the majority opinion.
f 65 This is no science, much less best available science. At most, the list suggests that the 12 species in question may exist in the 10-county area known as “Region 1,” not in Ferry County. Concerned Friends of Ferry County did not meet its burden to overturn the county. The Board’s conclusion is not supported by substantial evidence, and the *855county must be upheld. It is decisions like this that have led to frustration statewide with the power assumed by these unelected and nonexpert boards.18
B. Best Available Science
¶66 The majority overlooks critical errors in this record in order to discuss the definition of BAS. As demonstrated above, such analysis is unnecessary to the disposition of this case. There simply was no science to support the attack on Ferry County.
¶67 It is worth noting, however, what today’s opinion does not do. It does not determine whether the requirement to include BAS is a procedural or substantive mandate. The validity of the WACs’ interpreting the BAS requirement cited by the majority is not properly before us. Nonetheless, a few comments regarding BAS are necessary in order to respond to the majority.
f 68 The plain language employed by the GMAis not “best science” but is instead “best available science.” RCW 36-.70A.172. We must give meaning to all terms in the statute. As is clearly the case with Ferry County, resources may limit what is “available” for one county versus what is available for another richer or more studied county. “ ‘ “Available” means not only that the evidence must be contained within the record, but also that the science must be practically and economically feasible.’ ” Copsey, supra, at 104 (emphasis added) (quoting Clark County Natural Res. Council v. Clark County, No. 96-2-0017, at 2209 (W. Wash. Growth Mgmt. Hr’gs Bd. Dec 6,1996). Ferry County expressed to the Board the difficult task of including BAS in its species listing if given broad reading because the 1995 BAS amendment could “constitute mandates by the State which have not been funded by the state.” AR at 256. The Board’s response was that “[a] decision on that argument is not within the jurisdiction of this Board .... Devising a compliant comprehensive plan need not be an expensive process for Ferry County.” Id. The Board offers no suggestion for funding and *856none of the “amici” state agencies offer support, financial or otherwise.
¶69 Today’s majority criticizes the county for using only “two [secondary] sources” and for not “conducting on-site observations or conferring with other experts in the area.”19 Majority at 837. The state agencies attacking the county did not comply as previously indicated. This seems to suggest that the GMA imposes an impossible burden requiring Ferry County, with its limited resources, to conduct exhaustive on-site surveys to prove a negative, the nonexistence of several species. I would hold that financial resources are a relevant consideration for inquiry in determining what science is “available” under the BAS requirement.
C. The State Agencies’ (Amici) Obligations Were Not Met
¶70 I also find improper the Board’s acceptance of state agency letter comments as best available science and would correct the court’s silence on this practice. The GMA is intended to be local-focused, “bottom up,” rather than a “top down,” planning. Here, the Board stated on several occasions that state agency letters formed its baseline for BAS determinations. “[T]he County has not provided a scientific foundation, evidence of analysis, or a reasoned process to justify their listing, while rejecting the recommendations of endangered, threatened and sensitive species and wildlife habitat conservation areas provided by DFW.” AR at 255.
¶71 A system controlled by state agencies directly conflicts with the GMA’s declaration of a “broad range of discretion that may be exercised by counties and cities consistent with the requirements . . . and goals of [the GMA].” ROW 36.70A.3201.
¶72 If state agency suggestions are per se BAS, such agencies, especially the DFW with its species expertise and *857statutory responsibility to protect threatened and endangered wildlife, should be required to share their site-specific science with the county. As previously noted, the agency input here clearly did not comply with BAS.
¶73 If the DFW priority habitat species list had been compiled using “geographic information system (GIS) technology,” as the majority suggests, DFW should have provided the GIS data specific to Ferry County rather than generic information possibly relevant to the 10 counties in region one. Majority at 828 n.l. It surely should have been timely data.
¶74 In addition, if state agency suggestions are argued as BAS, then these state agencies should become parties to hearing board actions in order to defend their BAS.20 Here, the county made good faith efforts to include BAS in its comprehensive plan to protect fish and wildlife habitat conservation areas. Unfortunately, these good faith efforts were not reciprocated by the state agencies. Instead, three state agencies engaged in limited roles during all stages of this litigation. Neither DFW, DOE, nor Washington Department of Community, Trade, and Economic Development (CTED) participated as parties before the county.
¶75 At the hearings board level, state agency input was limited to several letters sent by DFW habitat biologist Kevin Robinette that disparaged the county’s efforts. The letters “expressed concern” regarding Ferry County’s amendments (AR at 304) and called the county’s efforts “discouraging” (AR at 71) and “incomplete” (AR at 605), among others. Other letters were sent by the DOE. AR at 606-12.
¶76 Three assistant attorneys general filed an amici brief in this court on behalf of three different agencies: CTED, DFW, and DOE. Their brief did not advise this court that the Board’s factual predicate — that the 12 species were present in Ferry County — was false. This is not what courts expect from amici, especially state agency amici. Certainly *858the legislature and counties expected greater and good faith efforts by state agencies to assist counties in GMA compliance.
III. CONCLUSION
¶77 I would reverse the Board’s decision as there is no substantial evidence in the record that Ferry County did not use BAS in designating and protecting wildlife habitat conservation areas. If Concerned Friends of Ferry County and the state agencies continue to maintain that the county has not included BAS in its comprehensive plan, they should bring forth such scientific evidence at the time of the county’s next comprehensive plan review.
Sanders, J., concurs with J.M. Johnson, J.
Sources included well known and respected sources, e.g., Michael R. Smith, Philip W. Mattocks, Jr. & Kelly M. Cassidy, Breeding Birds of Washington State (1997).
Ferry County’s population is 7,300. Of Washington State’s 39 counties, Ferry County ranks (1) last in population density at 3.3 persons per square mile, (2) last in per capita annual personal income at $17,437, and (3) first in the receipt of food stamps and medical assistance. See Office of Financial Management, 2003 Washington State Data Book 194-95. In contrast, King County’s population is 1,779,300 (approximately 243 times greater than Ferry County), and contains at least 24 cities with greater populations than all of Ferry County. Id. at 208-09.
Direct funding was administered through amicus Washington Department of Community, Trade, and Economic Development. The record here shows no support of any kind from amici DFW and DOE to Ferry County.
It is highly doubtful that this legislation would have received such unanimous support if it had been known that the best available science requirement would be interpreted to be as onerous as it is rendered by the majority opinion.
The majority opinion expends considerable resources detailing hearings board decisions issued prior to the order that is the subject of the instant appeal, as well as subsequent appeals to the superior court and Court of Appeals. These are entirely irrelevant for purposes of this administrative appeal.
Neither DFW nor any representative presented evidence or testified as to these species before the county or Board, notwithstanding that department’s statutory obligations to threatened and endangered species, e.g., RCW 77-.12.020(5), (6).
This case does not present a constitutional challenge to the boards or their makeup and procedures.
If the use of secondary sources less than two years old is a requirement for GMA compliance, it is doubtful than any county in the state could meet the requirement. King County’s most recent critical areas ordinance, for example, relies on secondary resources that are more than two years old. See King County, Best Available Science References, Executive Report (Feb. 2004), http://www. metrokc.gov/ddes/cao/PDFs04ExecProp/BAS-LiteratureRefs04.pdf.
As previously noted, the DFW list here would not survive cross-examination.