Opinion by Judge SKOPIL; Concurrence by Judge O’SCANNLAIN.
SKOPIL, Senior Circuit Judge:
I.
The gray wolf, or Canis Lupus, is listed as endangered under the Endangered Species Act (ESA) throughout the coterminous United States, except in Minnesota, where it is listed as threatened. 50 C.F.R. § 17.11(h). Gray wolf populations in Canada, however, are plentiful. Pursuant to ESA section 10(j), the Fish and Wildlife Service (FWS) captured Canadian gray wolves and released them in Yellowstone National Park as an “experimental population” designed to replenish wolves in Wyoming and parts of Montana and Idaho, where they had been all but eradicated by about 1930. See 59 Fed. Reg. 60,251, 60,252 (Nov. 22, 1994). One of these wolves migrated from Yellowstone to the Red Lodge, Montana area, where it had a fatal encounter with Chad McKittrick. After shooting and killing the wolf, McKittrick skinned and decapitated it, taking the hide and head to his home.
The government charged McKittrick with three counts: one, taking the wolf in violation of 16 U.S.C. §§ 1538(a)(1)(G), 1540(b)(1), and 50 C.F.R. § 17.84(i)(3); two, possessing the wolf in violation of 16 U.S.C. §§ 1538(a)(1)(G), 1540(b)(1), and 50 C.F.R. § 17.84(i)(5); and three, transporting the wolf in violation of the Lacey Act, 16 U.S.C. [1173]*1173§§ 3372(a)(1), 3373(d)(2).1 Magistrate Judge Anderson conducted a trial and then sentenced MeKittriek to six months’ imprisonment after a jury convicted him on all counts. District Judge Shanstrom affirmed the conviction and sentence.
On appeal, MeKittriek argues that the wolf he killed was not protected by the ESA, that his separate counts for taking and for possessing the wolf were multiplicitous, that his taking of the wolf was not “knowing” because he did not realize what he was shooting, and that the court erred in instructing the jury about the “incidental take exception.” We reject each of these challenges and affirm the conviction. MeKittriek also contends, however, that the sentencing magistrate judge should have reduced his offense level by two points for acceptance of responsibility. Because the magistrate judge may have disallowed the reduction on impermissible grounds, we remand for a redetermination of whether MeKittriek accepted responsibility under U.S.S.G. § 3E1.1.
II.
A. Validity of the Regulations
MeKittriek challenges his conviction by alleging four defects in FWS’s designation of the gray wolf experimental population in Yellowstone. Specifically, he maintains that (1) FWS may not draw members of an experimental population from an unlisted population, such as Canadian gray wolves; (2) the experimental population is invalid because it is not “wholly separate geographically” from naturally occurring wolves in the release area; (3) the experimental population regulations are invalid because the Secretary did not make a finding required by ESA section 4(d); and (4) the regulations are invalid because the Secretary did not comply with ESA section 4(f).
McKittrick’s challenges raise questions of law that we review de novo. See Torres-Lopez v. May, 111 F.3d 633, 638 (9th Cir.1997). Because these questions involve FWS’s interpretation of the ESA and the agency’s own regulations, however, our review is subject to deference. See Chevron, U.S.A, Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837, 843-44, 104 S.Ct. 2778, 2782, 81 L.Ed.2d 694 (1984) (agency’s reasonable interpretation of ambiguous statutory provision entitled to deference); Rainsong Co. v. FERC, 106 F.3d 269, 272 (9th Cir.1997) (agency’s interpretation of a statute it administers entitled to considerable deference); Thomas Jefferson Univ. v. Shalala, 512 U.S. 504, 510-12, 114 S.Ct. 2381, 2385-87, 129 L.Ed.2d 405 (1994) (agency’s interpretation of its own regulations entitled to “broad deference”).
1. The Experimental Population Designation
MeKittriek contends that FWS acted improperly in creating an experimental population from members of an unlisted population — that is, Canadian gray wolves. Authority to create experimental populations arises under section 10(j) of the ESA, which provides that “[t]he Secretary may authorize the release ... of any population ... of an endangered species or a threatened species outside the current range of such species if the Secretary determines that such release will further the conservation of such species.” 16 U.S.C. § 1539(j)(2)(A). According to MeKittriek, the language of this provision restricts the Secretary to drawing members of experimental populations from populations already listed under the ESA. We disagree.
First, gray wolves are protected by the ESA based on where they are found, not where they originate. Canadian gray wolves that migrate into the northern United States, for example, assume protected status when they cross the border. See 59 Fed.Reg. at 60,253 (discussing the probable “southern expansion of the Canadian wolf population” into [1174]*1174Glacier National Park, where they are classified as endangered); see also H.R.Rep. No. 97-567, at 33 (1982), reprinted in 1982 U.S.C.C.A.N. 2807, 2833 (shifting treatment of protected status depending on seasonal movement patterns); Ramsey v. Kantor, 96 F.3d 434, 438 (9th Cir.1996) (shifting “legal regimes” for protected salmon through the course of their migration). Therefore, the wolves transported from Canada were members of “any population ... of an endangered species or a threatened species” as soon as they entered the United States.
Second, McKittrick’s interpretation offends the statute’s essential purpose, which is the conservation of species. See Tennessee Valley Authority v. Hill, 437 U.S. 153, 184, 98 S.Ct. 2279, 2296-97, 57 L.Ed.2d 117 (1978) (noting that ESA’s purpose is to conserve endangered species at any cost). If McKit-trick’s analysis of section 10(j) were correct, FWS would be forced to create an experimental wolf population only by depleting threatened or endangered populations in the United States. Instead, the agency has interpreted section 10(j) to allow gray wolves to be taken from the healthy Canadian population and reintroduced to an area where wolves had been extirpated. See 59 Fed. Reg. at 60,252 (Because gray wolves are common in Canada, “[n]o adverse biological impact is expected from the[ir] removal.”). Neither the ESA nor its implementing regulations expressly permits FWS to obtain experimental populations from unlisted populations, but the agency’s interpretation of the ESA, which it is charged with administering, is entitled to great deference. See Rainsong, 106 F.3d at 272. In addition, FWS’s interpretation comports with the ESA’s purpose, as revealed by legislative history and case law. When section 10(j) was added to the ESA, for example, Congress spoke of the statute’s broad purpose:
In enacting the Endangered Species Act, Congress recognized that individual species should not be viewed in isolation, but must be viewed in terms of their relationship to the ecosystem of which they form a constitutent [sic] element. Although the regulatory mechanisms of the Act focus on species that are formally listed as endangered or threatened, the purposes and policies of the Act are far broader than simply providing for the conservation of individual species or individual members of listed species.
H.R. Conf. Rep. No. 97-835, at 30 (1982), reprinted in 1982 U.S.C.C.A.N. 2860, 2871 (specifying that conservation plans may include unlisted as well as listed species). Congress’ specific purpose in enacting section 10(j) was to “give greater flexibility to the Secretary.” H.R.Rep. No. 97-567, at 33 (1982), reprinted in 1982 U.S.C.C.A.N. 2807, 2833. Thus, each experimental population has its own set of special rules so that the Secretary has more managerial discretion. Id. at 2834; see 50 C.F.R. § 17.80. This flexibility allows the Secretary to better conserve and recover endangered species.
The Supreme Court acknowledged the Secretary’s broad authority when it upheld the regulatory definition of “harm” to include habitat modification, reiterating that “the broad purpose of the ESA supports the Secretary’s decision to extend protection against activities that cause the precise harms Congress enacted the statute to avoid.” Babbitt v. Sweet Home Chapter of Communities for a Great Oregon, 515 U.S. 687, 698, 115 S.Ct. 2407, 2413, 132 L.Ed.2d 597 (1995). Like the regulation at issue in Sweet Home, the agency’s implementation of section 10(j) in creating the experimental wolf population effectuates the ESA’s purpose and is within the Secretary’s authority. FWS’s designation of the experimental population was proper, and the wolf McKittrick shot fell within the ESA’s protection.2
2. The “Wholly Separate Geographically” Requirement
McKittrick claims that the reintroduced wolves were not “wholly separate geographically” from wolves already present in Yellowstone, and that the experimental population designation is therefore invalid. In [1175]*1175order “[t]o protect natural populations and to avoid potentially complicated problems of law enforcement,” H.R.Rep. No. 97-567, at 33 (1982), reprinted in 1982 U.S.C.C.A.N. 2807, 2833, the statute requires that a population qualifies as a section 10© experimental population “only when, and at such times as, the population is wholly separate geographically from nonexperimental populations of the same species,” 16 U.S.C. § 1539©(1). When experimental and nonexperimental populations overlap — even if the overlap occurs seasonally — section 10© populations lose their experimental status. 50 C.F.R. § 17.80(a).
According to McKittrick, such a loss of experimental status deprives the imported wolves of all ESA protection. We disagree. Because the wolves’ ESA status depends in part on their location, a wolf that had been introduced into Yellowstone from Canada would either be classified as experimental or, if it lost experimental status because of overlap with natural populations, as endangered. See 50 C.F.R. § 17.84(i)(10). If the wolf McKittrick shot had been a member of the endangered species population protected under the ESA, instead of the experimental population protected by regulation, McKit-trick would simply have been charged with a violation of ESA section 9(a)(1), subsections (B) (taking an endangered species) and (D) (possessing an endangered species), instead of (G) (violating regulations protecting endangered or threatened species).
We need not decide whether McKittrick should have been charged under a different subsection, however, because we hold that FWS’s regulations for the gray wolf experimental population meet the “wholly separate geographically” requirement. In its rule-making process, FWS specifically determined that “the experimental population area does not currently support any reproducing pairs of wolves.” 59 Fed.Reg. at 60,256. Although McKittrick points to sporadic sightings of isolated indigenous wolves in the release area, lone wolves, or “dispersers,” do not constitute a population. See 50 C.F.R. § 17.3 (defining population as “a group of fish or wildlife ... in common spatial arrangement that interbreed when mature”); 59 Fed.Reg. at 60,256 (defining gray wolf population as “at least two breeding pairs of gray wolves that each successfully raise at least two young to December 31 of their birth year for 2 consecutive years”).
The District of Wyoming reached the opposite conclusion in Wyoming Farm Bureau Federation v. Babbitt, 987 F.Supp. 1349 (D.Wyo.1997). Although it deferred to FWS’s definition of “population” as a permissible construction of the ESA within the agency’s expertise, 987 F.Supp. at 1371, the court nevertheless held the FWS regulations invalid under the “wholly separate geographically” requirement, id at 1372. Concluding that it was erroneous for FWS to consider populations rather than individual wolves, the court declared the agency’s determination of geographic separation “insufficient and contrary to law” because the introduced wolves were not geographically separate from “non-introduced specimens of the same species.” Id at 1372.
We do not agree with the Wyoming District Court’s analysis that section 10© must be read to apply to individual specimens as well as populations. The court based its reading on a House Report containing the word “individuals” once and “specimens” twice. H.R.Rep. No. 97-567, at 33 (1982), reprinted in 1982 U.S.C.C.A.N. 2807, 2833. The quoted section of the report, however, uses the word “population” or “populations” sixteen times, id., and section 10© itself refers only to populations,3 16 U.S.C. § 1539©. We must defer to FWS’s reasonable interpretation of section 10©, see Rain-song, 106 F.3d at 272, particularly where the interpretation involves agency expertise, see Mt. Graham Red Squirrel v. Espy, 986 F.2d 1568, 1571 (9th Cir.1993). FWS has interpreted the “wholly separate geographically” requirement only to apply to populations; this interpretation is reasonable and we decline to disturb it.
[1176]*11763. Section 4(d)
McKittrick further contends that the regulations he violated were invalid because the Secretary failed to comply with ESA section 4(d). That section authorizes the Secretary to promulgate regulations applicable to threatened species “as he deems necessary and advisable to provide for the conservation of such species.” 16 U.S.C. § 1533(d). McKittrick reasons that because experimental populations are given the same level of ESA protection as threatened species, the Secretary’s failure to recite the words “neeessary and advisable” in the special rules applicable to the gray wolf experimental population is fatal to those regulations.
This argument is meritless. Section 10(j) requires two specific findings for regulations pertaining to experimental populations: (1) that the establishment of such a population will further the species’ conservation; and (2) that the population is either essential or nonessential to the species’ conservation. 16 U.S.C. § 1539(j)(2)(A),(B). The gray wolf regulations contain both findings. See 50 C.F.R. § 17.84(i)(l),(2).
4. Section 4(f)
McKittrick also challenges the regulations’ validity because the Secretary did not adopt a recovery plan for gray wolves that “give[s] priority to ... species ... most likely to benefit from such plans.” 16 U.S.C. § 1533(f). Because gray wolves are plentiful in Canada and Alaska, McKittrick argues, FWS’s efforts to reintroduce wolves in Yellowstone represent a poor allocation of resources and therefore violate section 4(f).
This argument, too, is without merit. The Secretary has broad discretion to determine what methods to use in species conservation, see Sweet Home, 515 U.S. at 708,115 S.Ct. at 2418; adoption of recovery plans is discretionary, 16 U.S.C. § 1533(f); and the presence of healthy wolf populations in Canada and Alaska does not, in any event, make the recovery of U.S. populations any less crucial.
B. Multiplicity
McKittrick challenges Counts I and II of the information for multiplicity. Count I charged him with taking the wolf, and Count II, with possessing it. We review de novo the question whether an indictment or information is multiplicitous and thus violates a defendant’s double jeopardy rights. United States v. Wolfsmnkel, 44 F.3d 782, 784 (9th Cir.1995).
The test for multiplicity — charging a single offense in more than one count — is whether each separately violated statutory provision “requires proof of an additional fact which the other does not.” Blockburger v. United States, 284 U.S. 299, 304, 52 S.Ct. 180, 182, 76 L.Ed. 306 (1932); see Wolfswinkel, 44 F.3d at 784-85. McKittrick violated subsection (3) of the regulations by killing the wolf; he violated subsection (5) by possessing it. 50 C.F.R. § 17.84(i)(3),(5). Each violation required proof of a separate element. Count I required proof of a taking by McKittrick, which Count II did not.4 Count II required proof that McKittrick possessed the wolf, an element not required in Count I. The two counts were not multiplicitous. See United States v. Billie, 667 F.Supp. 1485, 1494 (S.D.Fla.1987) (concluding that separate counts for taking under 16 U.S.C. § 1533(a)(1)(B), and possessing, carrying, and transporting under § 1533(a)(1)(D) were not multiplicitous).
C. Degree of Intent
McKittrick argues that a violation of ESA section 11 requires the government to prove that he knew he was shooting a wolf, and that the jury instructions misled the jury about the requisite intent. We review for an abuse of discretion whether the magistrate judge’s “precise formulation” of the intent element was sufficient. United States v. Knapp, 120 F.3d 928, 930 (9th Cir.), cert. denied, — U.S.-, 118 S.Ct. 417, 139 L.Ed.2d 319 (1997). Whether the instructions actually misstated an element of the crime is subject to de novo review. Id.
[1177]*1177The instructions were accurate. McKittrick need not have known he was shooting a wolf to “knowingly violate[ ]” the regulations protecting the experimental population. 16 U.S.C. § 1540(b)(1). In 1978, Congress changed the wording of section 11 to “reduee[ ] the standard for criminal violations from ‘willfully’ to ‘knowingly.’ ” H.R.Rep. No. 95-1625, at 26 (1978), reprinted in 1978 U.S.C.C.A.N. 9453, 9476. It did this to “make[ ] criminal violations of the act a general rather than a specific intent crime.” Id. As the magistrate judge recognized, the District of Montana had already decided the intent issue in the government’s favor, holding on similar facts that “[t]he critical issue is whether the act was done knowingly, not whether the defendant recognized what he was shooting.” United States v. St. Onge, 676 F.Supp. 1044, 1045 (D.Mont.1988); see also Billie, 667 F.Supp. at 1493 (“[T]he Government need prove only that the defendant acted with general intent when he shot the animal in question.”). The Fifth Circuit has reached the same conclusion in related situations. See United States v. Nguyen, 916 F.2d 1016, 1017-18 (5th Cir.1990) (sustaining possession conviction did not require that defendant know animal’s ESA-protected status); United States v. Ivey, 949 F.2d 759, 766 (5th Cir.1991) (citing St. Onge and Billie in holding that, like § 1540(b)(1), § 1538(c) is a general intent statute). The Eleventh Circuit has expressed its agreement with the reasoning of these cases in holding that an analogous provision of the African Elephant Conservation Act requires only general intent. See United States v. Grigsby, 111 F.3d 806, 817 (11th Cir.1997). As these eases and the legislative history indicate, section 11 requires only that McKittrick knew he was shooting an animal, and that the animal turned out to be a protected gray wolf.
D. The “Incidental Take” Exception
McKittrick also claims that the jury instructions misrepresented the “incidental take” exception and improperly shifted the burden of proving the exception to McKit-triek. As with the instruction on intent, the magistrate judge’s “precise formulation” of the exception instruction and whether it was adequate are reviewed for an abuse of discretion. Knapp, 120 F.3d at 930. Whether the instructions actually misstated an element of the crime—for example, by inaccurately stating the burden of proof—is subject to de novo review. Id.
McKittrick first asserts that the instruction inaccurately recited the exception. The instruction was not inaccurate, however; on the contrary, it precisely tracked the language of the regulation: “Any person may take a gray wolf ... [pjrovided that the take is incidental to an otherwise lawful activity, accidental, unavoidable, unintentional, not resulting from negligent conduct lacking reasonable due care, and due care was exercised to avoid taking a gray wolf.” 50 C.F.R. § 17.84(i)(3)(viii).
McKittrick’s second contention is that the instruction shifted the burden of proof from the government to him to prove the applicability or inapplicability of the exception. We disagree. Even if the government did have the burden to prove that the incidental take exception did not apply to the taking, which is not clear,5 the instructions in no way suggest otherwise. Viewed as a whole, see United States v. Harrison, 34 F.3d 886, 889 (9th Cir.1994), the magistrate judge’s instructions to the jury clearly placed the burden of proof on the government.
Any error in the instructions would be harmless, nonetheless, because McKittrick cannot qualify for the incidental take exception. He deliberately shot the wolf; he did not kill it unintentionally in the course of some other activity. The incidental take exception does not apply to “deliberate action.” See Sweet Home, 515 U.S. at 700-01, 115 S.Ct. at 2414-15. Furthermore, the regulatory exception has two requirements—first, that the take be incidental; second, that it be “reported within 24 hours” to FWS. 50 [1178]*1178C.F.R. § 17.84(i)(3)(viii). MeKittrick does not claim to have reported the taking, so he would not qualify for the exception even if his take had been incidental.
E. Acceptance of Responsibility
Finally, MeKittrick argues that, because he admitted shooting the wolf and only contested his guilt based on the applicability and validity of the regulations, the sentencing magistrate judge should have allowed a two-point reduction for acceptance of responsibility under U.S.S.G. § 3El.l(a).
1. Standard of Review
The magistrate judge’s findings of fact underlying an application of the Sentencing Guidelines are reviewed for clear error. See United States v. Thompson, 80 F.3d 368, 370 (9th Cir.1996). Because “[t]he sentencing judge is in a unique position to evaluate a defendant’s acceptance of responsibility,” a judge’s determination under section 3E1.1 is “entitled to great deference on review.” U.S.S.G. § 3E1.1 (Application Note 5).
2. Analysis
The Sentencing Guidelines provide for a two-point decrease in offense level “[i]f the defendant clearly demonstrates acceptance of responsibility for his offense.” U.S.S.G. § 3El.l(a). MeKittrick contends that because he “truthfully admitted] the conduct comprising the offense(s) of conviction,” U.S.S.G. § 3E1.1 (Application Note 1(a)), he should have been allowed the reduction. MeKittrick did not plead guilty, however, and the commentary instructs that the “adjustment is not intended to apply to a defendant who puts the government to its burden of proof at trial,” except in “rare situations.” Id. (Application Note 2). One of those rare situations may occur when the defendant “goes to trial to assert and preserve issues that do not relate to factual guilt ... [such as] a challenge to the applicability of a statute to his conduct.” Id. McKittrick’s was such a trial — he contended that the regulations were invalid or inapplicable and that they required a level of intent he did not possess. Therefore, MeKittrick was eligible for a section 3E1.1 reduction.
In establishing the offense level, the magistrate judge adopted the findings of fact in the presentence report. See United States v. Colussi, 22 F.3d 218, 220 (9th Cir.1994) (noting that the sentencing judge is not obligated to explain factual findings and is free to “simply adopt the conclusions in the presen-tence report, or state that the defendant did not meet the requirements”). The report recommended against the reduction because, although MeKittrick admitted that he shot an animal that turned out to be a wolf, he “denies possessing intent to kill the wolf, thereby thwarting culpability for committing the criminal act for which he was found guilty.” MeKittrick did maintain at trial, in the face of contradictory testimony from a friend who accompanied him, that he thought the animal was a wild dog when he shot it. Because the court followed St. Onge in formulating the jury instructions, however, the jury did not have to decide whether to believe MeKittrick.
MeKittrick had the burden of demonstrating that he accepted responsibility for his crime. See United States v. Innie, 7 F.3d 840, 848 (9th Cir.1993). He was entitled, however, to challenge the intent requirement of ESA section 11 without forgoing his eligibility for the reduction. It is unclear from the language in the presentence report whether or not the magistrate judge based his denial of a section 3E1.1 adjustment on an impermissible ground. We therefore remand to give Magistrate Judge Anderson the opportunity to reexamine his determination.
III.
We hold that the regulations protecting the gray wolf experimental population are valid and that there was no error in the information or the jury instructions. Accordingly, we affirm McKittrick’s conviction for taking, possessing, and transporting a gray wolf in violation of the ESA and the Lacey Act. We vacate and remand the sentence, however, for a redetermination of whether MeKittrick satisfied his burden to show acceptance of responsibility under U.S.S.G. § 3E1.1.
[1179]*1179AFFIRMED in part, REVERSED in part, and REMANDED.