In a boundary dispute as treacherous as old man river itself, the appellants challenge the district court's conclusion that accretions to the west bank of the Mississippi River are within the territorial boundaries of Mississippi. Finding that the district court erred in its initial placement of the boundary thalweg between Louisiana and Mississippi, we reverse and render judgment for the appellants.

Meandering Through the Courts: The Proceedings Upstream

At issue in phase one of this bifurcated case was the sovereign ownership of a mass of accretions lying along the west bank of the Mississippi river near Lake Providence, Louisiana.2 The action was originally instituted by Mississippi citizens in federal district court as a suit to quiet title against Louisianians asserting ownership of the accreted lands. One year later, the State of Louisiana and the Lake Providence Port Commission intervened, praying that the accreted portion be adjudged Louisiana property. The intervenors filed a petition in the United States Supreme Court, requesting that the court exercise its original jurisdiction in this dispute between states. Inexplicably, the High Court declined the invitation.

The case proceeded to trial on the question of state boundary only, where the district judge waded through the testimony of experts and would-be landowners, as well as maps, surveys, and charts dating back to the late nineteenth century. The court ruled that the interstate boundary, frozen by an avulsive shift in the river, placed the disputed lands within Mississippi. Alternatively, the court found that Mississippi had exercised sovereign authority over the accretions, and that Louisiana had acquiesced in Mississippi's assertion of ownership. From that judgment, Louisiana takes this appeal.

We begin our voyage down the river with a review of the factual bases for each party's ownership claim. Their tales are so divergent that each will be separately recounted.

Louisiana contends that at the time the land grant patent was issued to Stephen Blackwell in 1881, a land mass identified on the General Land Office survey as "Island Number 94," or "Stack Island," was subject to the divided flow of the Mississippi River. Although relatively narrow, the channel flowing to the east of the island comprised the main navigable thread, or thalweg, of the river.3 Accordingly, since federal common law fixes the interstate boundary at the thalweg, Louisiana argues that the island was incorrectly attributed to Mississippi on the 1881 survey plat.

Louisiana notes that a sudden and perceptible (avulsive) change in the main course of the river occurred in 1882, when dikes constructed by the Mississippi River Commission induced a change in the flow pattern, diverting river traffic into the west channel. Because that change was avulsive, however, the boundary remained legally fixed in the east channel. Flooding enlarged the east channel in 1912, filling the west channel impassibly with silt, and restoring the east channel to dominance. Meanwhile, Stack Island was gradually eroding away, and its fragments were accreting downstream on the west bank of the river near Lake Providence. Eventually, Louisiana suggests it was replaced by a new island formed in approximately the same location. Louisiana further claims that at all relevant times, it exercised sovereign authority over the disputed portion.

In contrast, Mississippi argues that in 1881, shoreline surveys and government lights indicated that the thalweg of the river was located to the west of Stack Island. Because the middle of the west channel formed the interstate boundary, the thalweg, Stack Island was properly attributed to Mississippi in the land grant survey. Mississippi acknowledges that the east channel gradually enlarged in the early 1900's, and the west channel was abandoned by navigation when it filled with silt and alluvium. However, because it characterizes that gradual shift as "avulsive" in nature, it suggests the interstate boundary remained fixed in the west channel.

Though it concedes the island has undergone substantial changes due to the processes of erosion and accretion, it maintains that the "original" Stack Island never eroded away. Finally, Mississippi maintains that it has consistently claimed, taxed, and exercised dominion over the accreted portion.

Navigating a Legal Course: The Rules of the River

Under the "Rule of the Thalweg," when a navigable river flows between states, the middle thread of the main channel of the river constitutes the interstate boundary. Iowa v. Illinois, 147 U.S. 1, 13 S.Ct. 239, 37 L.Ed. 55 (1893). In early cases, courts defined the "main channel" as the deepest and most navigable branch of a waterway. However, in Louisiana v. Mississippi, 466 U.S. 96, 104 S.Ct. 1645, 80 L.Ed.2d 74 (1984), noting that the descriptions "deepest" and "most navigable" are frequently not synonymous, the Court refined the rule. Accordingly, current law dictates that the channel used as the "ordinary course of traffic on the river" is the river's thalweg. Id. at 101, 104 S.Ct. at 1648.

At least one exception to the Rule of the Thalweg is recognized: where an avulsive shift in the course of the river occurs, the boundary remains frozen in the former thalweg. Thus, "where a stream, which is the boundary, for any cause suddenly abandons its old and seeks a new bed, such change of channel works no change of boundary...." Nebraska v. Iowa, 143 U.S. 359, 361, 12 S.Ct. 396, 397, 36 L.Ed. 186 (1892).

Although "true" avulsive action is typically described as sudden and perceptible, our court has modified that characterization, opting to apply the avulsion exception where a change in the identity of the thalweg was slow and gradual. In Hogue v. Stricker Land & Timber Co., 69 F.2d 167 (5th Cir.), cert. denied, 293 U.S. 591, 55 S.Ct. 106, 79 L.Ed. 686 (1934), the ownership of Glasscock Island, a land mass in the Mississippi River, was the subject of a dispute between a Mississippi resident and a Louisiana corporation. The evidence revealed that although the east channel was once the thalweg, the west channel had gradually become dominant through the natural, dynamic forces of the river. While not contesting the original designation of the east channel as the thalweg, the Mississippi plaintiff argued that absent a sudden, perceptible shift in the river's course, the interstate boundary should follow the main channel of the river as it shifted westward.

Our court relying on Missouri v. Kentucky, 11 Wall. 395, 20 L.Ed. 116 (1870), disagreed. Though noting that "strictly speaking there was no avulsion," Id. at 168, it concluded:

... where the main channel changes from one side of an island to the other, it seems that the same rules as to boundary govern as are applied in cases of avulsion.... The old channel remains the boundary in the case of an island as well as in that of an avulsion.

Id. Thus, in the case of an island, even a gradual change in the orientation of the thalweg effects no change in the boundary, or in the sovereign ownership of the island.

Distilling these legal principles, we apply a two-step analysis in resolving this factual dispute. First, we determine which channel constituted the boundary thalweg in 1881. Second, we consider whether, under the Hogue exception, that boundary remained fixed notwithstanding a shift in the identity of the original thalweg. We conduct this inquiry against the backdrop of the clearly erroneous standard, giving due regard to the trial judge's assessments of the witnesses' credibility. See Fed. Rule of Civ. Pro. 52(a).

The appellants submit that the district court clearly erred in concluding that the boundary thalweg lay west of Stack Island in 1881, the year the land was patented to Stephen Blackwell. We agree.

In reaching its conclusion, the district court relied primarily on the interpretation given by Austin Smith, Mississippi's only expert witness, of information contained in an 1881 shoreline survey. The district court considered the surveyor's notation of a "good deep channel" to the west of the island, and Smith's testimony that depth was the determinative factor, as persuasive evidence of the dominance of that channel. It also noted the presence of a government navigation light on the west bank of the river near the island, concluding that light served to guide river traffic from the head of Stack Island down the west channel. Without elaboration, the court pronounced an 1881 survey accompanying a Mississippi River Commission (MRC) study inconclusive, and declined to consider it in formulating its findings. Although we acknowledge that the district court's findings are entitled to deference, after our review of the evidence, we are "left with the clear impression that an error has been made." Stauffer Chemical Co. v. Brunson, 380 F.2d 174, 181 (5th Cir.1967).

We recognize that it is not within the province of the reviewing court to second-guess the district court's assessment of Smith's credibility as a witness. However, it is apparent that Smith disregarded the only conclusive pieces of evidence in formulating his opinion. The first such piece of evidence is the hydrographic survey completed in December of 1881 pursuant to a major Congressionally-funded MRC improvement plan. That plan, for which Congress appropriated nearly one million dollars, was designed in part to improve navigation in the vicinity of the Lake Providence reach. The report issued in 1883 at the conclusion of the project clearly designates the east channel as the "main channel" at the time of the December 1881 survey. For example, the report lists as one of the "general effect[s] of the work" the "closing of the main channel of the river ... and bringing it back to the [downstream] right of Stack Island by a system of deflecting dikes."

... In order to force the main channel of the river which flowed down the [east channel] ... a main dike ... was driven from a point below the foot of Baleshed Bar to the head of Stack Island, leaving the low water main channel from Longwood through the [east channel] open for the passage of boats. (emphasis added)

The district court also ignored the hydrographic data contained in an MRC survey depicting the topography and hydrography of this portion of the river in 1881-82. While those data indicate the presence of shoals near the northern end of Stack Island in the west channel, the hydrographic soundings depict ample depths, even at low water, for typical river traffic in the east channel. Given these facts, it is illogical that vessels would employ a route that was not only approximately one mile longer, but also marked by treacherous shoals.

The evidence relied upon by Mississippi, Smith, and ultimately the district court simply does not contain the type of data necessary to support a contrary conclusion. The 1881 shoreline survey was not intended to reflect the hydrographic features of this part of the river, nor does it. Unlike the MRC survey, done in anticipation of major dike construction in the river itself, the shoreline survey of Stack Island was intended for patent purposes only, and contains no true hydrographic readings.

The deputy surveyor's vague notation of a "good deep channel" to the west of the island is less than determinative. Not only is the depth of the channel no longer the focus of our inquiry, Louisiana v. Mississippi, 66 U.S. at 101-02, 104 S.Ct. at 1648, but the survey fails to reflect any depth soundings in the east channel.4 Because we endeavor to determine the relative dominance of the channels, the shoreline survey simply does not contain the type of data necessary to determine the identity of the thalweg in 1881.

Equally inconclusive is the placement of the U.S. navigation lights along both banks of the river as reflected on the shoreline survey. In the vicinity of the Lake Providence reach, that survey depicts a navigation light on the east bank of the river at Reserve Plantation near the northern tip of the island, a light on the west bank directly across from the island, and a light on the east bank near the Shipland Landing south of the island. The MRC survey, completed only three months later, depicts two lights in the vicinity of the Reserve Plantation, a light on the west bank beyond the southern tip of the island, and the Shipland Landing light on the east bank. The MRC survey does not depict a navigation light on the west bank near the mid-point of the island as does the shoreline survey.

Regardless of the resolution of that factual dispute, the placement of the lights does little to determine the identity of the low-water channel. As the district court found, plantations located on both sides of the river were serviced by steamboats that were presumably able to use either channel at high water. Since navigation lights were needed to assist these vessels, the depiction of the light on the west bank cannot compensate for the inadequacies in the shoreline survey.

Thus, we find the court clearly erred in relying upon evidence that, regardless of its veracity, could not support its finding. In rejecting this evidence, and Smith's interpretation of it, we note that we are not the first court to do so. In prior litigation between these same states, the Supreme Court rejected Smith's theory regarding the placement of an interstate boundary, criticizing his undue emphasis on which channel was the deepest and swiftest. Louisiana v. Mississippi, 466 U.S. at 103-06, 104 S.Ct. at 1649-1651. Noting that the "Smith [boundary] line did not conform to the data on the surveys ... [and] that it was not conceivable that a mariner would adopt Smith's track of navigation," the court adopted Louisiana's proposed placement of the thalweg instead. For the reasons outlined above, we do likewise.

Freezing the Boundary: An Avulsive Shift?

As the district court noted, both parties generally concede that flooding in 1911 and 1912 shifted the thalweg from the west channel to the east channel. The difference, of course, is that Mississippi suggests the thalweg was originally in the west channel, while Louisiana suggests it was diverted there from the east by dike construction in 1882-83. Because the district court determined, as a threshold matter, that the thalweg originally lay in the west channel in 1881, it found it necessary to consider whether the 1911 shift was avulsive. Concluding that it was, the court ruled that the interstate boundary was frozen in the west channel.

Our reversal of the district court's ruling on the original location of the thalweg obviates further discussion of the avulsion issue in light of the law of this circuit. Both parties agree, and the district court found, that the main channel lay to the west of the island from 1883 to 1911, and then shifted to the east thereafter. However, the question of whether that shift, or any prior one, was "avulsive" is irrelevant as a matter of law in this context. Under our holding in Hogue, even a gradual change in the identity of the "main" channel around an island effects no change in the boundary. Thus, the boundary remained frozen in the east channel, where it lay at the time of the patent in 1881, regardless of the nature of any shift in the river's course. Hogue, 69 F.2d at 168.Acquiescence

Alternatively, the district court concluded that the disputed lands belonged to Mississippi under the Doctrine of Acquiescence. Because we find the district court improperly applied the law to its factual findings, we reverse this ruling as well.

Numerous cases recognize that acquiescence by one state in the exercise and preservation of an interstate boundary by another is conclusive evidence of the location of the official boundary. See e.g., Rhode Island v. Massachusetts, 4 How. 591, 638-39, 11 L.Ed. 1116 (1846); Virginia v. Tennessee, 148 U.S. 503, 522-25, 13 S.Ct. 728, 735-37, 37 L.Ed. 537 (1893); Louisiana v. Mississippi, 202 U.S. 1, 53-57, 26 S.Ct. 408, 422-425, 50 L.Ed. 913 (1906). The Supreme Court has even noted that the doctrine pre-empts the "Rule of the Thalweg," when "it is established that there has been acquiescence in a long-continued and uninterrupted assertion of dominion and jurisdiction over a given area...." Arkansas v. Tennessee, 310 U.S. 563, 567-72, 60 S.Ct. 1026, 1030, 84 L.Ed. 1362 (1940).

Over strenuous and repeated objection from the appellants, the district court permitted the introduction of the testimony of numerous residents of the disputed lands. A colorful assortment of Mississippi citizens testified that they had occupied the accreted lands as owners, had run cattle there, had hunted and fished, and even engaged in the illegal cultivation of marijuana. There was some evidence that Mississippi assessed taxes against those individuals, although it is disputed whether the property taxed was Stack Island or the accreted lands. Other evidence suggests Mississippi law enforcement agents exercised criminal jurisdiction over the island.

Louisiana citizens testified to similar facts. Some indicated that they had hunted or fished on the lands with Louisiana licenses, and that wildlife agents had enforced Louisiana game laws on the accretions. Furthermore, Louisiana argues residents of the disputed lands have paid taxes on the accreted portions to the state of Louisiana.

Without regard to the correctness of the district court's findings on these disputed factual issues, and assuming without deciding that the evidence was properly admitted, we find much of this evidence irrelevant to the question of acquiescence. Distinct from any state law theory of adverse possession, the federal common law doctrine of acquiescence is premised upon proof of the relationship between sovereigns. Thus, details of discreet activities of individuals such as hunting, fishing, or farming are largely irrelevant. Although evidence of the assessment of taxes by a sovereign is sometimes probative, see e.g., Arkansas v. Tennessee, 310 U.S. at 567-72, 60 S.Ct. at 1029-1031, its persuasiveness is diminished in cases like this one where there is some evidence that both states claimed the disputed lands as a tax base.

Discounting that extraneous testimony, we are left with little evidence of any acts committed by agents of the state of Louisiana that indicate its recognition of the sovereign authority of Mississippi over the accreted lands. There was some testimony that Louisiana law enforcement agents had handed over a suspected law breaker to Mississippi for prosecution for an offense allegedly committed on the accretions. However, a few such isolated incidents do not constitute a "long-continued and uninterrupted assertion of dominion and jurisdiction over an area...." Arkansas v. Tennessee, 310 U.S. at 571, 60 S.Ct. at 1030. Those acts are simply not of sufficient duration and magnitude to justify application of the doctrine. See e.g., California v. Nevada, 447 U.S. 125, 130-32, 100 S.Ct. 2064, 2067-68, 65 L.Ed.2d 1 (1980) (where the evidence established California's recognition of a putative boundary for over 100 years); Arkansas v. Tennessee, 310 U.S. at 567-72, 60 S.Ct. at 1029-1031 (applying the doctrine where the evidence showed unchallenged an unequivocal dominion by Tennessee for 115 years); Louisiana v. Mississippi, 202 U.S. at 53-57, 26 S.Ct. at 422-24 (involving over 90 years of acquiescence by Mississippi). Reviewing the district court's application of the law to its factual findings de novo, we cannot sustain the district court's ruling on the issue of acquiescence.

For the foregoing reasons, we reverse the judgment of the district court adjudging the disputed property within the state of Mississippi, and enter judgment in favor of the appellants.

REVERSED and RENDERED.