HUFFMAN, J.

In September 1994, the superior court issued two related judgments granting petitions for writs of mandate that were brought by certain opponents of a landfill project, including the National Parks and Conservation Association and several other individuals and organizations (collectively the opponents). [1] The landfill project proponents include real parties in interest Kaiser Steel Resources, Inc., Kaiser Eagle Mountain, Inc., and Mine Reclamation Corporation, Inc. (MRC) (collectively appellants). The landfill project will be located within one and one-half miles of the southeastern boundary of 794,000-acre Joshua Tree National Park (the Park). The petitions challenged the actions of the Board of Supervisors (the Board) of the County of Riverside (the County) in certifying the environmental impact report (EIR) prepared to analyze that project pursuant to the California Environmental Quality Act (CEQA). (Pub. Resources Code, [2] § 21000 et seq.; Code Civ. Proc, §§ 1085, 1094.5.)

In granting the opponents' petitions in 1994, the trial court ruled that a number of the opponents' challenges to the EIR had merit, but two particular challenges (not involved here) were unmeritorious. The opponents appealed those determinations, and in a published opinion this court affirmed the judgments in both cases. ( National Parks & Conservation Assn. v. County of Riverside (1996) 42 Cal.App.4th 1505 , 50 Cal.Rptr.2d 339 (National Parks) . ) The project proponents went back to the drawing board and completed three additional years' worth of study and analysis of the project's effect upon the surrounding environment, and produced a new EIR, which was then submitted to the County. Extensive public hearings were held and ultimately the Board issued a decision the project as mitigated would have no significant effect upon the environment, with the exception that a statement of overriding considerations was made as to the wilderness experience component (i.e., the effect of the project upon the experience of those visiting the Park).

The appellants then submitted the matter to the trial court as their return on the October 1994 writ. The opponents filed numerous objections to the return on the writ, and after a hearing and a reconsideration motion, the superior court found in its final order that there had been no adequate compliance with the requirements of the writ of mandate as to' two particular items of study: The impacts of the project on the wilderness experience at the Park and on the desert tortoise population.

Appellants now assert in this court that contrary to the findings of the superior court, with respect to the impacts on the Park, substantial evidence supports the decision to approve the project and the EIR is more than sufficient in its analysis and as an informational document. ( Laurel Heights Improvement Assn. v. Regents of University of California (1988) 47 Cal.3d 376, 392-393 , 253 Cal.Rptr. 426 , 764 P.2d 278 ( Laurel Heights I ).) Appellants also assert that substantial evidence supports the County's decision to approve the project in light of the studies made with respect to the desert tortoise population; they argue the trial court incorrectly substituted its judgment for those of the experts in the field and/or misinterpreted applicable law. We find appellants' arguments to be meritorious and reverse the final order with directions that the trial court overrule the objections and discharge the writ previously issued.

We reserve a detailed account of the facts for the discussion portion of this opinion. However, a basic outline is necessary to frame the issues. Attached as appendix A is a 1996 draft map of the project area with relation to the Park. (See fn. 6, post. ) In our prior opinion in this case, the trial court's statement of decision describing the history of the project was quoted in pertinent part:

The landfill will fill in areas left by the huge pits of the mining operations (farthest from the Park, to begin with) and will also fill in nearby canyons and hillsides which already contain the waste material from the mining operations. A six-inch layer of dirt and mine debris will be placed daily upon the fill material. In addition to the landfill, the project will include the operation of a 52-mile railroad line and the upgrading of a county road, both for purposes of bringing in the trash for processing. Also, an existing townsite in the area, an outgrowth of a previous company town run by the mining company, will be expanded to serve the workers at the landfill. Currently, the townsite has a few hundred residents and a privately run prison facility operates there, housing 500 prisoners.

The site of the landfill project is about one and one-half miles from the nearest Park boundary, as established in 1994 when the Park was converted from a national monument to a larger national park through the federal California Desert Protection Act, which expanded the Park boundaries. (16 U.S.C. § 410aaa-21 et seq.) The areas between the site and this portion of the Park boundary include infrastructure such as an aqueduct, a pump station, utility and communication lines, roads, and another employee townsite. The proponents entered into an agreement with the federal Bureau of Land Management (BLM) for a land exchange, in which BLM will acquire wildlife habitat acreage and the proponents will acquire other disturbed mining lands in the nearby Eagle Mountains.

As set forth in our prior opinion, National Parks, supra, 42 Cal.App.4th 1505 , 50 Cal. Rptr.2d 339 , once the initial EIR for the project was prepared, two sets of petitioners sought administrative mandamus, challenging the adequacy of the EIR. (See fn. 1, ante. ) The trial court's 1994 ruling was in favor of the project proponents (appellants) on a number of the opponents' contentions. However, the trial court issued the writ, concluding the evidence was insufficient to support the conclusions drawn by the EIR in several other areas, including the topics involved here (desert tortoise habitat and the effects of the project on the wilderness experience at the Park). Further proceedings under CEQA were required and that judgment was affirmed on appeal, over the opponents' objections that more should have been required. ( National Parks, supra, 42 Cal. App.4th at pp. 1520-1523, 50 Cal.Rptr.2d 339 .)

As alluded to above, from 1994 to 1997 further studies and analyses of the project's effect upon the surrounding environment were conducted on the designated topics. Those resulted in the preparation of a new EIR designed to address the court's concerns as expressed in the writ of mandate. The areas of analysis included not only the items which are the subject of this appeal (desert tortoise habitat and the effects of the project on the wilderness experience at the Park), but also cumulative impacts of the project and a related hydroelectric project, the proposed land exchange between the proponents and the BLM, landfill liner design, seismic issues and leachate protection.

As part of the preparation for the project, in 1996 the proponents entered into an agreement with the National Park Service (NPS) that provides for an environmental mitigation trust fund to purchase additional wildlife habitat over the life of the project. The proponents agreed to implement as necessary certain mitigation measures designed to protect the Park from unforeseen environmental impacts (such as air pollution and groundwater monitoring, and a program to monitor raven populations at the landfill). [3]

After extensive public hearings from 1996 to 1997, the County approved the project, finding that, as mitigated, it would have no significant effect upon the environment, with the exception that a statement of overriding considerations was issued as to the wilderness experience component. Appellants returned to the superior court, filing their supplemental return to the writ of mandate, which detailed the efforts made in preparing the EIR to comply with the court's directions to bring the project into compliance with CEQA.

The opponents filed numerous objections to the return on the writ, and the matters (now consolidated) went to hearing and to a reconsideration motion. The trial court made certain findings in its order after hearing, sustaining the opponents' objections about the degree of compliance with the requirements of the previously issued writ (discussed, post ). In its final order, the court found, first, a lack of substantial evidence to support the EIR's finding that the impacts of the landfill project on the desert tortoise had been mitigated to a level of insignificance: "Specifically, tortoise-proof fences are not required along the railroad track, and the Desert Tortoise Recovery Plan (DTRP) [4] clearly recommends that no new landfills be constructed in desert tortoise habitat."

Second, the court in its final order sustained the objections to the finding of the EIR regarding impacts of the landfill project on the neighboring Park. The court found insufficient evidence to support the EIR's conclusion that impacts to the Park will be less than significant: "Specifically, the EIR fails to adequately analyze the impacts on the wilderness experience, noise impacts, light from the townsite, and the impact of the landfill on the biological resources of the [Park] as a complex and interrelated system, which the [NPS] describes as eutrophication." (All remaining areas of objection were overruled.)

Accordingly, the court ordered that the County set aside the certification of the EIR and other project approvals, with a subsequent EIR required to be prepared if the County further considered the project. All physical activity related to the development of the landfill project was suspended until compliance with CEQA was obtained. The trial court reserved jurisdiction over any further return on the writ.

Appellants filed timely notices of appeal of both the order after hearing and the final order as to both consolidated cases. The appeals were also consolidated.

On review "`In applying the substantial evidence standard, "the reviewing court must resolve reasonable doubts in favor of the administrative finding and decision."' [Citation.] The appellate court's role `is precisely the same as the trial court's,' and lower courts' findings are not `conclusive on appeal.' [Citation.]" ( San Joaquin Raptor/Wildlife Rescue Center v. County of Stanislaus (1994) 27 Cal.App.4th 713, 722 , 32 Cal.Rptr.2d 704 .) Although the opponents here make a feeble argument that we should defer to the lower court's discretion in ruling that the new EIR, which it had ordered prepared, remained inadequate, they can muster no authority to support such a view; that view is in any case contrary to the weight of authority on the proper standard of review, which is de novo. ( Laurel Heights I, supra, 47 Cal.3d at pp. 392-393, 253 Cal. Rptr. 426 , 764 P.2d 278 .)

The trial court's task in this case was to determine whether there had been adequate compliance with the previously issued writ. This amounted to a decision whether the County had prejudicially abused its discretion in approving the updated EIR and in issuing the related entitlements to proceed with the project. "Abuse of discretion is established if the agency has not proceeded in a manner required by law or if the determination or decision is not supported by substantial evidence." (§ 21168.5.) In Western States Petroleum Assn. v. Superior Court (1995) 9 Cal.4th 559, 570-573 , 38 Cal.Rptr.2d 139 , 888 P.2d 1268 ( Western States ), the Supreme Court approved the view that the substantiality of the evidence supporting an administrative decision (there, a quasi-legislative one) is a question of law that is governed by the same evidentiary rules used to decide the substantiality of the evidence supporting findings of fact made in a trial court.

The same rules apply when the courts evaluate an administrative body's approval of an EIR. We decide this question of law de novo:

Here, as frequently occurs, the major disputes are over whether relevant information was omitted from the EIR. In Barthelemy v. Chino Basin Mun. Water Dist. (1995) 38 Cal.App.4th 1609, 1620-1621 , 45 Cal.Rptr.2d 688 ( Barthelemy ), the court analyzed recent case law and came up with these conclusions on how to approach such arguments:

Thus, our inquiry should be whether the County had substantial evidence to support its decision to approve the project and issue a statement of overriding considerations on the wilderness experience component. We disagree with appellants' argument in their reply brief that under section 21005, subdivision (c), [5] our review cannot address any deficiencies in the EIR other than those identified by the superior court in its final order (based on appellants' theory that all other issues were impliedly decided in their favor, and the Legislature intended that all project problems be identified at the trial level). Such an approach would contravene the rules for conducting de novo review of the record that was before the County when the approval was issued. [6] Also, the issues are sufficiently interrelated here that the record must be considered as a whole. ( Western States, supra, 9 Cal.4th at pp. 571-572, 38 Cal.Rptr.2d 139 , 888 P.2d 1268 ; Barthelemy, supra, 38 Cal.App.4th at pp. 1619-1620, 45 Cal.Rptr.2d 688 ; Cal.Code Regs., tit. 14, § 15384, subd. (a) [all references to the Guidelines shall refer to title 14 of the California Code of Regulations].)

With these standards in mind, we discuss appellants' specific arguments regarding the impacts of the project upon the Park and (last but not least) upon the local desert tortoise population.

In the 1994 writ, appellants were directed to fully analyze in a manner supported by substantial evidence the impact of the landfill project on the neighboring park "including, but not limited to, any negative impacts upon the natural peace and solitude, the clean air, the pristine desert (i.e., the `wilderness experience')" offered by the Park. In its 1998 order after hearing, the court sustained the opponents' objections to the adequacy of the EIR to support its findings and those of the County. The court found insufficient evidence on these points to support the EIR's conclusion that impacts to the Park will be less than significant: "Specifically, the EIR fails to adequately analyze the impacts on the wilderness experience, noise impacts, light from the townsite, and the impact of the landfill on the biological resources of the [Park] as a complex and interrelated system, which the National Park Service describes as eutrophication."

The new EIR prepared in response to the 1994 writ analyzed several components of the wilderness experience by distinguishing between wilderness and nonwilderness, or human-impacted, areas. Thus, the parties now debate on appeal whether there is substantial evidence support in the record for such a distinction, with respect to the "significance" of several categories: noise impacts, lighting, and the objective and subjective aspects of experiencing the wilderness. Both parties discuss the current version of Guidelines section 15064, "Determining the Significance of the Environmental Effects Caused by a Project," which reads in pertinent part as follows:

The opponents interpret this guideline as defining all applicable standards of significance of environmental effects, i.e., that a standard of significance must be derived from a statute or a public agency's standard adopted for environmental protection purposes. The opponents thus rely on the federal Wilderness Act ( 16 U.S.C. § 1131 et seq.), the 1916 National Park Service Organic Act ( 16 U.S.C. § 1 et seq.), the California Desert Protection Act (16 U.S.C. § 410aaa-21 et seq.), and other environmental protection legislation. The opponents argue, "Together, these laws provide a standard of the significance against which impacts to Joshua Tree's wilderness experience should have been measured."

Appellants respond that the Guidelines do not define how standards of significance for environmental effects must be determined, but instead merely recognize that regulatory standards for a particular impact category, if complied with by a project, may support a finding by the public agency issuing the permit for the project that there are no significant impacts in that category. (See Sundstrom v. County of Mendocino (1988) 202 Cal.App.3d 296, 306-309 , 248 Cal.Rptr. 352 .) Thus, appellants contend that the Guidelines section 15064, subdivision (h) does not require that every threshold of significance be based upon an adopted statutory performance standard. Substantively, they further argue that under Guidelines section 15064, subdivision (h)(2), the lead agency has the freedom under the Guidelines to develop a threshold of significance for a particular impact category at issue.

On a procedural note, appellants point out that as of the time of the County's decision (August-September 1997), subdivision (h) of Guidelines section 15064 did not yet list these sources of significance standards, but merely provided that in marginal cases where experts disagree as to whether a project will have a significant effect on the environment, an EIR shall be prepared. The current version with the standards relied on by opponents does not appear in the sources until after September 1998, after the order on appeal was issued. "[C]hanges to the guidelines act prospectively only." ( Laurel Heights Improvement Assn. v. Regents of University of California (1993) 6 Cal.4th 1112, 1126, fn. 10 , 26 Cal.Rptr.2d 231 , 864 P.2d 502 ( Laurel Heights II ).) Accordingly, this subdivision of the Guidelines listing the sources of significance standards does not require a specific statutory basis for the standards applied here.

Further, appellants point to several items in the record that support the EIR's treatment, and the County's acceptance, of different thresholds of significance for impacts upon different areas of the Park, depending on its level of wilderness. Specifically, between the landfill site and the nearest Park boundary are found a major aqueduct, a pump station, utility and communication lines, roads, and another employee townsite. Certain areas within the Park nearest to the landfill site contain miles of electrical transmission lines, an aqueduct, and roads. The EIR commented that this evidence of human development was visible from within the nearby Eagle Mountain range. Most of the Park's visitor activity takes place miles away at the other end of the Park. The Park allows various nonwilderness uses in backcountry transition zones (an NPS term for nonwilderness) of the Park, such as rock climbing, horseback riding, bicycling, and four-wheel drive vehicle use. Appellants thus contend the opponents have no basis to seek a "zeroimpact" standard for the Park, if that is what they are advocating.

When the Park boundaries were expanded by the Desert Protection Act in 1994, the chief sponsor of the bill, Senator Feinstein, wrote the Secretary of the Interior that the act was never intended to create a private de facto buffer zone around the Park for purposes of land use. The federal law creating the Park and the NPS management guidelines distinguish between wilderness and nonwilderness Park areas. The DOI maps reflect these distinctions.

We conclude that for purposes of analyzing the various "wilderness experience" components, there is substantial evidence in the record to support the approach taken by the EIR and the County, to distinguish the significance of impacts according to the scientific and factual data gathered about the nature of the land affected: wilderness or nonwilderness. (Guidelines, § 15064, subd. (b).) Moreover, the record does not support any claim that no impacts from the project, whether significant or not, whether in a wilderness area or not, should be allowable, merely because of the proximity of the project to the Park.

As explained in Guidelines section 15064, subdivision (b), "an activity which may not be significant in an urban area may be significant in a rural area." Because the significance of an activity may vary with the setting, the EIR's noise consultants used different thresholds of noise significance for different areas of the Park. Residential noise standards were used in projecting noise levels in the parklands closest to the landfill site. That residential standard considers levels of 65 decibels or under to be insignificant, while the 3-decibels level is the level below which increases in noise levels are imperceptible to humans (and thus this standard was used for the wilderness areas). [7] The noise consultants for the EIR used County noise standards as the local regulations most immediately applicable to the project, and chose the more stringent standard applicable to family dwellings, as opposed to a looser standard for commercial/industrial areas. There were no available Park or federal noise standards that could be applied.

Based on the use of the residential standard, the EIR concluded noise impacts as to parklands would be insignificant. This included project construction, vehicular traffic and train traffic. The opponents claim the mitigation measures imposed pursuant to this finding were inadequate, based on a stringent view of wilderness protection, and also because they failed to address such issues as effects on wildlife and on areas away from the townsite. The trial court's ruling as to noise impacts found there was no support in the record that nonwilderness areas within the Park should be treated any differently from other areas within the Park.

We respectfully disagree. We have discussed above the basis in the record for distinguishing between the Park wilderness and nonwilderness areas, such as NPS management guidelines and DOI maps that make distinctions between wilderness and nonwilderness Park areas. Within the Park areas nearest to the landfill site are found electrical transmission lines, an aqueduct and roads. The EIR commented that this evidence of human development was visible from within the nearby Eagle Mountain range. There is no authority to create a private de facto buffer zone around the Park for purposes of land use. Absent more closely applicable standards, it appears the County had a substantial basis for accepting the EIR's use of county residential noise standards for assessing noise impacts. There is no requirement that all noise from the project be mitigated to a level of inaudibility, particularly as to nonwilderness parklands because, as we have discussed above, the standards for assessing impacts of a project require careful judgment on the part of the public agency involved, based to the extent possible on scientific and factual data; these standards allow for a finding of an insignificant degree of impact, not necessarily a zero impact. (Guidelines, § 15064, subd. (b).)

The EIR stated that since it was difficult to measure subjective impacts of the project as to a sensitive park visitor, some impact would have to be admitted. The EIR thus divided wilderness considerations into two categories, wilderness experience (some significant impact found) and wilderness resources (no significant impact found). Extensive studies were made of the resources issues that were capable of quantitative measurements (noise, odor, litter, air quality, view and lighting impacts), while the subjective aspects of wilderness experience were mainly taken as assumptions of significant impacts that were difficult to measure. The County's approval of the project included a finding that overriding considerations justified some such impacts upon the wilderness experience Component. The trial court criticized this approach, stating in the order after hearing: "By doing so, the EIR fails to give a complete picture of the effects of this project on the park. That a `wilderness experience' analysis is possible is found in the record at A.R. 64-22412."

To review the trial court's conclusion that there was inadequate analysis of all aspects of the wilderness experience component, we look to the portion of the record on which it relied, in context with the remainder of the record. The court was specifically referring to a June 1995 NPS document identifying issues for the EIR. In that paper, the local NPS unit proposed that a comprehensive visitor use and perception study should be completed, with the results to be included in the EIR. The NPS document stated:

However, a March 1996 letter by NPS staff stated that the Park personnel had concluded that a Park management program, the Visitor Experience and Resource Protection process (VERP), would not be an appropriate process to be used in assessing impacts for the EIR. No other such types of studies were mentioned.

In December 1996, the NPS entered into the mitigation agreement with the landfill proponents to establish trust funds for future mitigation of unforeseen environmental impacts and to impose additional contractual obligations on the landfill operators; the agreement was not intended to supersede or circumvent the EIR. Rather, the agreement recited that it was intended to assure NPS that the project would be operating in consideration of protecting the purposes and values of the Park and would not derogate the values and purposes for which the Park was established by Congress. [8]

From this sequence of events, the most reasonable inference to be drawn is that the NPS's position was changing and evolving, and although the type of study referenced by the trial court might be possible, the most knowledgeable parties concluded that it was not required. [9] The Board dealt with the issue by assuming there would be significant impacts in this respect, and nevertheless made a finding that overriding considerations justified some such impacts upon the wilderness experience component. Thus, we turn to the remainder of the record to see if, as required, the complete picture of the effects of this project on the Park, through a wilderness experience analysis, was adequately studied, thus supporting the Board's action.

Aesthetic issues are properly a subject of study in an EIR to assess the impacts of a project. (§§ 21060.5, 21100, subd. (d).) Similarly, NPS management guidelines analyze impacts by measuring such elements as view sheds, air quality, noise and odor. There is also a subjective component of aesthetic issues, which the EIR acknowledged by noting that a park visitor's wilderness experience might be significantly impacted by the mere knowledge that a landfill was nearby, even if such were not perceptible by the senses. Appellants interpret the court's order as mandating that a "comprehensive visitor use and perception study" should be completed, and characterize this requirement as demanding a public opinion poll regarding the psychological feeling of the Park visitors vis-à-vis the project's impacts. Appellants refer to authority that has rejected claims that environmental analysis should include "psychological impacts" of a project (e.g., City of Pasadena v. State of California (1993) 14 Cal.App.4th 810, 831-834 , 17 Cal. Rptr.2d 766 (disappr. in other pt. in Western States, supra, 9 Cal.4th at p. 576, fn. 6 , 38 Cal.Rptr.2d 139 , 888 P.2d 1268 ); Metropolitan Edison v. People vs. Nuclear Energy (1983) 460 U.S. 766, 777 , 103 S.Ct. 1556 , 75 L.Ed.2d 534 ). Rather, CEQA analysis is directed toward identifying any substantial adverse changes in physical conditions (§§ 21060.5, 21100, subd. (d)), and any physical changes, not economic or social effects of a project. (Guidelines, § 15131, subd. (a).)

The opponents reply that an EIR can permissibly discuss interrelated physical and social effects of a project. ( Tongass Conservation Soc. v. Cheney (D.C.Cir.1991) 288 App. D.C. 180 , 924 F.2d 1137, 1142-44 .) In Goleta Union School Dist. v. Regents of University of California (1995) 37 Cal.App.4th 1025, 1032 , 44 Cal.Rptr.2d 110 , the Court of Appeal concluded that an EIR was not required to consider potential classroom crowding in a school district, per se, as a significant effect on the environment, requiring mitigating measures by the project proponent. The court stated that although CEQA guidelines recognize that socio-economic effects of a project may cause physical changes that significantly affect the environment, the expansion of university facilities did not necessarily cause such physical effects in the form of drastically increased elementary school enrollment. ( Id. at pp. 1032-1033, 44 Cal. Rptr.2d 110 , citing Guidelines, § 15064, subd. (f).)

Similarly, although the landfill project will admittedly have some measurable effects upon the physical conditions within the Park, it is difficult to see how a "comprehensive visitor use and perception study" would add useful information regarding the key physical conditions at the site. As the Supreme Court has cautioned: "A project opponent or reviewing court can always imagine some additional study or analysis that might provide helpful information. It is not for them to design the EIR." ( Laurel Heights I, supra, 47 Cal.3d at p. 415 , 253 Cal.Rptr. 426 , 764 P.2d 278 .) The EIR acknowledges the potential subjective impacts of the project, even though it does not go much farther in defining them; however, its analysis goes the extra mile in examining every possible sensory impact on a Park visitor, which might then result in the subjective responses that were recognized as reasonably possible or probable. From the general information and hard data given, the County could reasonably draw conclusions about the significance of these impacts. No more was required. (Guidelines, § 15384, subd. (a).)

As explained above, an existing townsite in the area, an outgrowth of the company town run by the mining company, will be expanded to serve the workers at the landfill. The townsite has only a few hundred residents now, and a privately run prison facility for 500 prisoners is located there. For security reasons, the prison is well illuminated at night and that light is visible for 30 to 40 miles around. The townsite and the pump station (aqueduct) also contribute light to the night sky. The project will increase night lighting within a two-mile radius around 10 percent. The nearest Park campground to the townsite is about 20 miles away; backpackers can travel in the immediate vicinity if they carry their own water.

The opponents complain that the information in the EIR about night lighting only covers the townsite buildings and does not deal with such issues as increased car headlights, potential office or store development, or any analysis of cumulative light from the townsite combined with the landfill project (which can operate until 10 p.m.). They also complain that the lighting analysis does not fully account for the recent additions to parkland made by the 1994 Desert Protection Act.

Appellants respond that in light of the existing prison illumination and townsite lighting (also the aqueduct pumping station), a full analysis is made of all major night light sources and any of the suggested additional items are only speculative in impact or create only de minimis impacts. ( Pala Band of Mission Indians v. County of San Diego (1998) 68 Cal.App.4th 556, 577-578 , 80 Cal. Rptr.2d 294 .) They also argue that this issue was not specifically targeted in the administrative process and thus the opponents have waived any deficiency. ( Resource Defense Fund v. Local Agency Formation Com. (1987) 191 Cal.App.3d 886, 894 , 236 Cal.Rptr. 794 .)

Case law provides that an expert can make a judgment on existing evidence, without further study, that a particular condition will have no significant impact. ( Laurel Heights II, supra, 6 Cal.4th at p. 1140 , 26 Cal.Rptr.2d 231 , 864 P.2d 502 .) The EIR sets forth a realistic assessment of existing night lighting in the area and concludes that the projected expansion of the townsite will not make much difference. This was a reasonable conclusion in light of the already developed nature of the townsite, and the County had an adequate basis for finding no significant impact in this particular respect.

In its ruling, the court found insufficient evidence to support the EIR's conclusion that impacts to the Park will be less than significant regarding "the impact of the landfill on the biological resources of the [P]ark as a complex and interrelated system, which the [NPS] describes as eutrophication." In 1995, the Park staff requested that as part of the EIR process, the involved agencies study the phenomenon of adding nutrients (trash) to the dry, harsh desert landscape, possibly causing insects and rats to proliferate, then starting the food chain going full tilt and upsetting the Park ecosystem. (This phenomenon is termed eutrophication after the similar effects of adding nutrients to lakes and upsetting their ecosystems.) The Park recommended such studies as computer modeling and animal tracking (trapping rats and insects at existing landfills, and inventorying animal feces to analyze whether human-generated trash was being transported) for several years to project the impact of the landfill, if it were to be built.

The EIR approached this problem in several ways, directed toward containment of the refuse: (1) The EIR proposed such measures as keeping incoming refuse in sealed containers until transported to a limited working area, creating litter fences, conducting a storm watch to avoid the scattering of materials by windstorms, and covering the waste with dirt and mining debris to avoid access by ravens. Similar measures have proved effective at other landfills. (2) The landfill will have a state-of-the-art liner and operations design as part of the system for confining the waste. (3) Studies were made of the Los Angeles experience with landfills, in which rats were fitted with radio transmitters and researchers found that daily landfill operations (bulldozing and compacting waste) kill rats. Other studies have shown that insects do not proliferate at landfills if daily cover is properly applied. (4) .All ponds and water sources will be covered and the areas fenced to prevent access by predators such as ravens, coyotes, or kit foxes. (5) The proponents entered into the mitigation agreement with the NPS to provide for additional mitigation measures if necessary. (6) The EIR includes responses to comments, including the Park issues identification paper, stating that the lead agencies determined that existing data were available and sufficient to address impacts to biodiversity and ecosystem function. The EIR addresses such possible impacts from the project as "habitat loss, additional nutrients originating directly or indirectly from landfill material (defined as `eutrophication' by NPS), the introduction of exotic species, nitrate deposition, and global warming." The comments give the example that biodiversity could be affected by "`changes in the relationship between species in the form of increases in predator/scavenger populations in response to increased food availability at the landfill site, and from increased road kills.'" These impacts and appropriate mitigation were discussed in the EIR.

Several qualified consulting biologists rendered opinions about the ecological impacts of the project, reaching differing conclusions. Dr. Kathy Freas found there would be no significant adverse ecosystem impacts to the Park in the vicinity of the project. Dr. Jerry Freilich and Dr. William Schlesinger opposed the project, opining that the proposed efforts to contain the refuse would likely be unsuccessful and pests would likely proliferate. Dr. Schlesinger called it "a big lab experiment." There was no data available on landfills where eutrophication had or had not occurred.

From this and other evidence, the County determined that the project's impacts on the Park would be less than significant regarding biological resources. We believe this conclusion was sufficiently supported by the record. First, the trial court did not have a valid basis to reject the EIR merely because there was a difference of opinion among experts. ( Fort Mojave Indian Tribe v. Department of Health Services (1995) 38 Cal.App.4th 1574, 1600 , 45 Cal.Rptr.2d 822 , citing Laurel Heights I, supra, 47 Cal.3d at pp. 392-393, 253 Cal.Rptr. 426 , 764 P.2d 278 .) Second, there was expert opinion that the type of studies suggested by NPS would not yield useful or essential data regarding nutrient addition to the desert, because the design of the landfill will in all probability contain all refuse sufficiently within the site so that no significant effects will occur in this respect. It is speculative whether the eutrophication effect will occur, but if it does the NPS agreement provides for environmental mitigation measures to deal with the problems. There was a difference of opinion among the experts as to whether additional studies, such as computer modeling, would be accurate or necessary. CEQA does not require evaluation of speculative impacts. (Guidelines, § 15145.)

The Guidelines also provide that the sufficiency of an EIR is to be evaluated in light of what is reasonably feasible. (Guidelines, § 15151.) The alternative of years of further study could reasonably be rejected as infeasible for economic or planning reasons. ( Chaparral Greens v. City of Chula Vista (1996) 50 Cal.App.4th 1134, 1146, fn. 8 , 58 Cal. Rptr.2d 152 .) Even though the opponents complain the 1996 NPS mitigation agreement was made on the information available at that time, which thus was not necessarily complete, the County could reasonably conclude it had sufficient overall information on which to reach its decision. Effectively, the trial court selected among conflicting expert opinion and substituted its own judgment for that of the County. This was incorrect. "`The question for a reviewing court should then be limited to whether the agency's reasons for proceeding as it did are supported by substantial evidence.' [Citation.]" ( Barthelemy, supra, 38 Cal.App.4th at p. 1620, 45 Cal.Rptr.2d 688 .)

In conclusion, as guided by the Supreme Court in Laurel Heights I, supra, 47 Cal.3d at page 422 , 253 Cal.Rptr. 426 , 764 P.2d 278 , we deem the County's approval decision to be supported by the record with regard to effects upon the Park, even if it does not specifically address all the variants on the issues that the opponents are arguing here: "That some items or types of evidence, however, are less than conclusive does not mean the evidence as a whole is not substantial. As we have explained in a related context, `the reviewing court must resolve reasonable doubts in favor of the administrative findings and decision.' [Citation.] To paraphrase the Guidelines, a fair argument can be made to support the [County's] conclusion, even though other conclusions might also be reached. (Guidelines, § 15384, subd. (a).)" ( Laurel Heights I, supra, 47 Cal.3d at p. 422 , 253 Cal.Rptr. 426 , 764 P.2d 278 .)

Desert tortoises are a threatened (not endangered) species under federal and state law. (Endangered Species Act, 16 U.S.C. § 1531 et seq. (ESA); Cal. ESA, Fish & Game Code, § 2050 et seq.) They inhabit the Park and the areas along the access roads and railroad line. In 1994, the USFWS established areas of critical habitat for the tortoises, including portions of the railroad line and the road areas. The landfill site itself is not a critical habitat area. Previously, in 1992, the USFWS evaluated potential impacts of the project, finding it would not put the desert tortoise population in jeopardy by interfering with its critical habitat due to the mitigation measures imposed. [10] This "no jeopardy opinion" was reaffirmed in 1993 and 1996. The California Department of Fish and Game (CDFG) studied the project in 1994 and concluded that the project would not jeopardize the tortoise and the BLM land exchange would help it. We now turn to the two areas of concern identified in the trial court's ruling.

In its final order, the court found a lack of substantial evidence to support the EIR's, and the County's, finding that the impacts of the project on the desert tortoise had been mitigated to a level of insignificance, and referred specifically to the lack of a requirement of tortoise-proof fences along the railroad track. Appellants contend this ruling erroneously disregards the clear weight of expert opinion that such fencing is not required in the best interests of the tortoise population. The issue of fencing the 52-mile railroad line was studied by the USFWS, which concluded that creating such a barrier was not yet indicated because further study was necessary on migration patterns during operation of the project, which could lead to a policy of fencing only where necessary (as determined by the responsible agencies, CDFG, BLM, and USFWS). Such a barrier could also have ill effects on the bighorn sheep in the area, which have certain travel corridors.

The opponents question whether the proposed mitigation measures will really work, without a fully fenced railroad line. That is not the appropriate question. Rather, did the County have a sufficient basis in expert opinion and other evidence to conclude that the potential impacts of the project on the desert tortoise had been mitigated to a level of insignificance? Similarly, could the County reasonably conclude that further study was needed before a complete fencing requirement was imposed? The record contains evidence that there are significant pros and cons of a 52-mile fence along the railroad line. Where such unknowns exist, certain mitigation measures that are related to existing measures may be deferred in implementation where the other measures will mitigate known impacts to insignificance. ( Gentry v. City of Murrieta (1995) 36 Cal. App.4th 1359, 1396-1397 , 43 Cal.Rptr.2d 170 .)

On balance, it appears that the trial court effectively substituted its judgment on the fencing issue for those of the expert agencies and biologist in the field. It had no adequate basis for doing so. ( Laurel Heights I, supra, 47 Cal.3d at 393, 409 , 253 Cal.Rptr. 426 , 764 P.2d 278 ; Western States, supra, 9 Cal.4th at pp. 572-573, 38 Cal. Rptr.2d 139 , 888 P.2d 1268 ; Fort Mojave Indian Tribe v. Department of Health Services, supra, 38 Cal.App.4th at pp. 1603-1604, 45 Cal.Rptr.2d 822 .)

The trial court's ruling on this issue stated, "[T]he [DTRP] clearly recommends that no new landfills be constructed in desert tortoise habitat." The DTRP was issued in 1994 by the USFWS to set forth general recommendations to protect this creature throughout the Southwest. It creates 14 "Desert Wilderness Management Areas" (DWMA's) in which certain designated "recovery action" recommendations are proposed. One such DWMA is this general Park area (not including the landfill site). One of the 13 recovery action recommendations is:

The issue is whether the trial court correctly interpreted this recovery action as applying to desert tortoise "habitat," as opposed to smaller areas within the region, i.e., the identified "DWMA's." If this document interpretation is incorrect, then the trial court did not have a legitimate basis to find the EIR inadequate or the County's decision unsupported by substantial evidence. It is possible that the trial court was referring to "critical habitat" for the desert tortoise, as defined by USFWS regulations. However, the landfill site itself is not within that critical habitat designation, even though portions of the railroad and access road evidently are. Under the DTRP, outside of the DWMA's, the desert tortoise is to be protected by the Endangered Species Act. When the USFWS issued its no jeopardy opinion, it was under the authority of the ESA. This indicates that existing mitigation measures for the landfill area (not within a DWMA) are sufficient in the experts' view, and the access routes that fall within the DWMA's do not bring the entire project within the protected sphere of the DTRP.

Again, the opponents cite to general concerns about the mitigation measures: Will more human presence in the area lead to more damage to tortoise habitat (from off-road vehicles, dogs, cats, etc.)? We need not decide this issue on that basis. Rather, in light of the record as a whole, we disagree with the trial court's interpretation of the DTRP as prohibiting this project because it will operate within the general range of desert tortoise habitat. We conclude that substantial evidence exists in this respect to support the County's decision to approve the project as mitigated. (See fn. 10, ante; Western States, supra, 9 Cal.4th at pp. 572-573, 38 Cal.Rptr.2d 139 , 888 P.2d 1268 .)

The order is reversed with directions to overrule the objections and enter an order discharging the writ previously issued. Each party to bear its own costs on appeal.

WORK, Acting P.J., and HALLER, J., concur.

[*] Kennedy, J., and Chin, J., dissented.

[1] The opponents include the National Parks and Conservation Association, the Eagle Mountain Landfill Opposition Coalition, the City of Coachella, and a number of individuals: Laurence R. Charpied and Donna J. Charpied, Steve W. Clute, Richard M. Marsh, and Daniel S. Roman.

[2] All statutory references are to the Public Resources Code unless otherwise specified.

[3] Ravens are of interest here because they scavenge trash that is not adequately covered and prey upon young desert tortoises.

[4] The DTRP is a document issued in 1994 by the United States Fish and Wildlife Service (USFWS) (an agency of the federal Department of the Interior (DOI)), setting forth general recommendations to protect this form of wildlife. Other agencies which have studied the project's effect on the tortoise include the California Department of Fish and Game (CDFG), the BLM and its technical adviser, the National Biological Service.

[5] In pertinent part, section 21005, subdivision (c) provides: "It is further the intent of the Legislature that any court, which finds, or, in the process of reviewing a previous court finding, finds, that a public agency has taken an action without compliance with this division, shall specifically address each of the alleged grounds for noncompliance."

[6] Generally, evidence outside the administrative record is inadmissible to show that the agency has not proceeded in the manner required by law. ( Western States, supra, 9 Cal.4th at pp. 565, 574-576, 38 Cal.Rptr.2d 139 , 888 P.2d 1268 .) Extra-record evidence may be admissible if the proponent can show the evidence existed before the agency made its decision but it was impossible in the exercise of reasonable diligence to present it to the agency before the decision was made. ( Id. at p. 578, 38 Cal.Rptr.2d 139 , 888 P.2d 1268 .) Extra-record evidence also may be admissible if it pertains to the accuracy of the administrative record. ( Id. at pp. 575-576, fn. 5, 38 Cal.Rptr.2d 139 , 888 P.2d 1268 ; see Barthelemy, supra, 38 Cal.App.4th at pp. 1620-1621, 45 Cal.Rptr.2d 688 .) Appellants brought a motion to take additional evidence on appeal, i.e., a 1998 Park map and explanatory pages, on the ground that the opponents unfairly charged in their respondents' brief that appellants misrepresented the facts about Park boundaries vis-à-vis wilderness and nonwilderness area. Initially, we denied that motion without prejudice to reconsideration by the merits panel. We again deny the motion as we do not consider the additional map material enlightening on the issues on appeal; the County's decisions were made in 1997 on that record. As of November 1996, a Park representative admitted there were no more up-to-date maps issued by the DOI, and the 1996 maps were used in the EIR. Finally, the parties' dispute in their briefs over this issue fell within the scope of fair argument, and the motion need not be granted to clarify the parties' positions on the record, as requested.

[7] The standard used for measurement is the "A" weighting scale, resembling human hearing. For purposes of comparison, 65 decibels is the sound of an electric typewriter from 10 feet away. Ten decibels is the sound of leaves rustling.

[8] The parties acknowledge there was a significant difference of opinion about the Park's position on the project between the local and the Washington D.C. NPS officials. Eventually, the Washington view prevailed, as indicated by the signing of the 1996 NPS agreement regarding future mitigation measures.

[9] See Guidelines section 15384, defining substantial evidence: "(a) `Substantial evidence' as used in these guidelines means enough relevant information and reasonable inferences from this information that a fair argument can be made to support a conclusion, even though other conclusions might also be reached. Whether a fair argument can be made that the project may have a significant effect on the environment is to be determined by examining the whole record before the lead agency. Argument, speculation, unsubstantiated opinion or narrative, evidence which is clearly erroneous or inaccurate, or evidence of social or economic impacts which do not contribute to or are not caused by physical impacts on the environment does not constitute substantial evidence. [¶] (b) Substantial evidence shall include facts, reasonable assumptions predicated upon facts, and expert opinion supported by facts." (Italics added.)

[10] The mitigation measures imposed by the project were consistent with those that have been imposed for other projects in the general Southwest area, as determined from an independent consulting firm's study of those 234 projects. The main elements are a tortoise awareness program a defined work zone, an on-site monitor, and fencing. The measures include having a biologist survey the railroad line before trains arrive and remove any tortoises on the tracks. The biologist will also monitor and relocate tortoises in burrows that are in harm's way. A three-year monitoring study will be performed to determine the best locations for culverts for tortoise crossings or fences that will keep tortoises off the tracks, while still avoiding any unnecessary fragmentation (socially and genetically) of the tortoise and bighorn sheep populations which live on either side of the line. Portions of the access roads will be fenced. The environmental trust fund for purchase of additional tortoise habitat (375 acres) is another such major measure.