EUGENE A. WRIGHT, Circuit Judge:

The Makah Tribe sought a determination of its usual and accustomed fishing places under an 1855 treaty. It argues that the district court applied a too stringent standard of proof and improperly substituted its judgment for that of the Special Master. We affirm.

The Makahs are primarily an ocean fishing tribe that has traditionally depended on its ocean catch for both sustenance and trade. See United States v. Washington, 384 F. Supp. 312, 363 (W.D.Wash. 1974), aff'd, 520 F.2d 676 (9th Cir. 1975), cert. denied, 423 U.S. 1086, 96 S.Ct. 877, 47 L.Ed.2d 97 (1976) (the Boldt Decision). In 1855, the Makahs signed a treaty with the United States which guaranteed the Tribe's right to fish in its "usual and accustomed grounds and stations." Treaty of Neah Bay, Jan. 31, 1855, 12 Stat. 939. The Tribe asserts that, at the time of the treaty, its usual and accustomed fishing grounds extended almost 100 miles out to sea.

Evidence was presented on the Tribe's dependence on ocean fishing and on its ability to navigate great distances from the Washington coast. A report and brief testimony by Dr. Barbara Lane, an expert witness for the United States, were given substantial credence in this and related proceedings. See, e.g., the Boldt Decision, 384 F. Supp. at 350 (Dr. Lane's opinions, inferences, and conclusions were "well taken, sound and reasonable"). In addition, testimony of some elderly members of the Tribe described Makah fishing practices of almost a century ago.

Dr. Lane testified that documentation of the outer limits of the Tribe's fishery at treaty times does not exist. The only evidence from the time of the treaty was the Makah chief's statement in the official record of the treaty proceedings: "TseKaw-Wooth — he wanted the sea — that was his country."

Dr. Lane's report established that the Makahs were expert seamen who could navigate at night or for several days, even in fog or beyond sight of land. They navigated by the stars, winds, and ocean swells. This Tribe had extraordinary ability to handle canoes which were seaworthy, sturdy, and fast, designed for ocean fishing, whaling, and seal hunting. The Makahs were equipped to fish far offshore, using kelp lines up to 100 fathoms long.

Dr. Lane's report cited an 1897 account of frequent Makah whaling expeditions out to 100 miles. She reasoned: "If the Makah were pursuing offshore fisheries at those distances in 1897, there is no reason to suppose that they did not have the same capability in 1855."

The three elderly tribal members, born around the turn of the century, testified as to post-treaty fishing or navigational ability. Only one of the three, Oliver Ides, testified about trips "about a hundred miles" out. When he was young, "the halibut started disappearing" so they began trolling for salmon. He also hunted seal in dugout canoes, beyond the sight of land.

The government does not dispute that the Makahs regularly fished 30-40 miles offshore, or that they were able to navigate up to 100 miles out to sea. On the contrary, the government relies on the same evidence as do the Makahs, especially Dr. Lane's conclusion that the Makahs generally fished within 30-40 miles of the shore. She stated:

Treaty time accounts report that the Makah were accustomed to fish thirty or so miles offshore and that they were known to fish in waters eighty to one hundred fathoms deep. It is reliably reported by a number of observers that they were often away for days at a time. It seems reasonable to conclude that the extent of Makah offshore fishing at treaty times was limited primarily by Makah fishing effort rather than by capability. When stocks were abundant within thirty or so miles of shore, there was little reason for the Makah to fish at greater distances.

In post-treaty years, when altered circumstances made it necessary to go further offshore, as for example in the seal fishery, the Makah demonstrated that they had the capability to do so.

This case is a part of the Northwest Indian fishing rights litigation, which is within the continuing jurisdiction of the district court. See the Boldt Decision, 384 F. Supp. at 408, 419. In 1977, the Makah Tribe filed a Request for Determination of its ocean fishing areas. Pursuant to Federal Rule of Civil Procedure 53(b), the request was referred to a Special Master for an evidentiary hearing, report, and recommendation. The Master issued a recommended order, proposed findings of fact, and conclusions of law.

His recommendation set the western boundary of the Makahs' fishing area at longitude 127 ° W, almost 100 miles offshore. The district court modified the proposed order and set the boundary at longitude 125 ° 42' W, some 40 miles offshore.

Both the district judge and the Special Master limited the Tribe's present fishing right to the area within the fisheries management jurisdiction of the United States. The Magnuson Fishery Conservation and Management Act of 1976, 16 U.S.C. § 1801-1882, established a Fishery Conservation Zone (FCZ), which extends 200 miles from the American shore and in which the United States has exclusive fishery management authority. Id. §§ 1811, 1812. Because of the Canadian border, the FCZ does not extend 200 miles due west from Washington's Cape Flattery, but angles southwestward from the Strait of Juan de Fuca.

The immediately affected area in this dispute is only the triangle bounded on the east by longitude 125 ° 44' W, on the south by latitude 48 ° 2' 15" N, and on the north-west by the boundary of the FCZ. However, the parties dispute the entire area out to longitude 127 ° W because of its relevance in the event of a change in the international border.

ISSUES

1. What standards of proof determine Indian fishing grounds?

2. What standard of review do we apply to the district court's order and the Special Master's recommendation?

3. In de novo application of law to facts, what do we find to be the Makahs' usual and accustomed fishing areas?

ANALYSIS [18] I. Standard of Proof

The Boldt Decision established "basic facts and law," including an explanation of the treaty reference to "all usual and accustomed grounds and stations." 384 F. Supp. at 330-32. "Stations" indicates fixed locations, while "grounds" refers to "larger areas which may contain numerous stations and other unspecified locations which . . . could not then have been determined with specific precision and cannot now be so determined." Id. at 332. "Usual and accustomed" excludes locations used infrequently. Id.

Judge Boldt held that "every fishing location where members of a tribe customarily fished from time to time at and before treaty times, however distant from the then usual habitat of the tribe . . . is a usual and accustomed ground or station . . . ." Id.

In a compilation of major post-trial substantive orders, Judge Boldt dealt with a Request for Determination of the Tulalip Tribes' usual and accustomed fishing places. United States v. Washington, 459 F. Supp. 1020, 1058-60 (W.D.Wash. 1978). He noted the difficulty of the determination:

In determining usual and accustomed fishing places the court cannot follow stringent proof standards because to do so would likely preclude a finding of any such fishing areas. Little documentation of Indian fishing locations in and around 1855 exists today. The anthropological reports of Dr. Barbara Lane, which this court finds highly credible, have been very helpful in determining by direct evidence or reasonable inferences the probable location and extent of usual and accustomed treaty fishing areas.

Id. at 1059. Judge Boldt also considered "the credible testimony of tribal elders who speak from personal experience or data acquired from other sources," bearing in mind the limitations of such testimony regarding 1855 fishing practices. Id.

In the same orders, Judge Boldt made a preliminary determination of the Suquamish Tribe's fishing area. The Suquamish had made a prima facie showing from uncontroverted evidence of usual and accustomed fishing places. The decree did not set forth the evidence on which Judge Boldt relied.

The Makahs claim that the order was based solely on evidence that the Suquamish traveled from one place to another. They point out the absence of documentation that the Suquamish fished the entire distance. See id. at 1049. The Tribe urges this court to establish the Makahs' usual and accustomed fishing area by inferring that, if the Makahs traveled 100 miles offshore, they must have fished along the way.

Even if that assertion is true, the Suquamish decree has no bearing here. It was only a non-final order, based on uncontested evidence and subject to reconsideration on the basis of a full hearing at any party's request. Id. This is not such a case.

In sum, Judge Boldt's determinations as to the Tulalips and the Suquamish provide guidance for the standard of proof in this case only in (1) his refusal to apply stringent standards; (2) his reliance on Dr. Lane; and (3) his suggestion that testimony of tribal elders may not be "the most accurate and authoritative data." Id. at 1059.

II. Standard of Review

The Special Master determined that the Makah Tribe's "usual and accustomed" fishing grounds in 1855 extended some 100 miles from shore. The district court, applying the "clearly erroneous" standard pursuant to Fed.R.Civ.P. 53(e)(2), rejected the Special Master's determination and ruled instead that the fishing grounds extended only about 40 miles from shore. The government contends that this modification is correct but that the "clearly erroneous" standard is too deferential, both as applied by the district court to the Special Master's determination and as to be applied by this court to the district court's modification.

The standard of review depends on the type of question resolved below. Here, the district court agreed with the Special Master with regard to all of the latter's findings of historical fact. The district court disagreed only with the conclusion that those facts were sufficient to establish that the Makah Tribe's "usual and accustomed" fishing grounds extended 100 miles offshore in 1855. This is a mixed question of law and fact.

The "clearly erroneous" standard provided for by Fed.R.Civ.P. 53(e)(2) applies only to a Special Master's findings of fact. Where a Special Master resolves a mixed question of law and fact, the district court should apply the standard of review required by United States v. McConney, 728 F.2d 1195, 1199-1204 (9th Cir. 1984) (en banc). In that case, we held that where the application of law to facts does not require an inquiry that is "essentially factual," a mixed determination of law and fact is subject to de novo review. Id. at 1201-1203.

The application of the Boldt Decision standard to the facts of the present case did not require an "essentially factual" inquiry. The evidence is largely documentary and undisputed, and the Master was in no better position to draw conclusions than is a reviewing court. The district court should have reviewed the Special Master's mixed finding de novo. Likewise, the de novo standard is appropriate for this court's review of the district court's decision to substitute its own finding for that of the Special Master. In short, our de novo review requires us to apply the legal standard of the Boldt Decision to the historical facts established below.

III. De Novo Application of Law to Facts

Ocean fishing was essential to the Makahs at treaty time. The Makahs probably were capable of traveling to 100 miles from shore in 1855. They may have canoed that far for whale and seal or simply to explore. They did go that distance at the turn of the century, although it is not clear how frequently. About 1900, they fished regularly at areas about 40 miles out, and probably did so in the 1850's.

These facts do not show that their usual and accustomed fishing areas went out 100 miles in 1855. There is no basis for an inference that they customarily fished as far as 100 miles from shore at treaty time.

On the contrary, Dr. Lane suggested that the Makahs would travel that distance only when the catch was insufficient closer to shore. The earliest evidence of insufficient catch was Oliver Ides' statement about disappearing halibut when he was young, some 50 years after the treaty. Even under the less stringent standards of proof of this case, we cannot conclude that the Makahs usually and customarily fished 100 miles from shore in 1855.

CONCLUSION

Our de novo review of the evidence leads us to conclude that the 40-mile boundary found by the district court delineates the Makahs' usual and accustomed fishing areas.

AFFIRMED.