MEMORANDUM OPINION

GEORGE C. YOUNG, Chief Judge.

The United States brought this action for injunctive relief to compel a riparian owner to remove a dike allegedly constructed in violation of Section 10 of the Rivers and Harbors Act of 1899, 33 U.S.C. § 403. The riparian owner, Joder Cameron, erected the dike on the low lying portion of his lakefront property at Lake Harney, a rather shallow, two and one-half mile wide body of water connected with the St. Johns River.

Section 10 of the Rivers and Harbors Act makes it unlawful to:

(1) Create an obstruction "to the navigable capacity of any of the waters of the United States . . ."

(2) Build any "wharf, pier, dolphin, boom, weir breakwater, bulkhead, jetty, or other structures in any port, roadstead haven, harbor, canal, navigable river, or other water of the United States, outside of established harbor lines . . .; or

(3) Excavate or fill, or in any manner to alter or modify the course, location, condition, or capacity of any port, haven, harbor, canal, or of the channel of any navigable water of the United States,

unless the work has been approved by the Army Corps of Engineers.

The theory of the Government is that Cameron is in violation of Section 10 for two reasons: (1) the dike constitutes an obstruction to the navigable capacity of Lake Harney and (2) in building the dike Cameron illegally filled, excavated, altered, and modified the course, condition or capacity of a navigable waterway.

The parties agree that Lake Harney is a navigable water of the United States within the meaning of the Act and that Cameron never obtained the authorization of the Corps of Engineers for the construction of the dike. The principal question of fact raised by the government's suit is therefore whether the Cameron dike is subject to the regulatory jurisdiction of the United States. In a non-tidal lake such as Lake Harney this question is resolved by locating the line of ordinary high water, for it is this line which marks the limits of the government's navigational servitude. E.g., United States v. Rands, 389 U.S. 121, 123, 88 S.Ct. 265, 19 L.Ed.2d 329 (1967); United States v. Virginia Electric Power Co., 365 U.S. 624, 629, 81 S.Ct. 784, 5 L.Ed.2d 838 (1961); United States v. Kansas City Life Insurance Co., 339 U.S. 799, 804-05, 70 S.Ct. 885, 94 L.Ed. 1288 (1950). If the dike is located below the lake's ordinary water mark then it is subject to the regulatory jurisdiction of the Corps and Cameron must account for his failure to secure a permit. If on the other hand, the dike stands on fast lands, the Corps has no jurisdiction and Cameron is exempt from the requirements of the Rivers and Harbors Act.

In addition to contending that his dike was erected on fast land free from the Corps' regulatory control, Cameron raises the defenses of selective enforcement and estoppel. Because resolution of these questions depends upon a preliminary finding that Cameron's dike was constructed in violation of the Act, the Court elected to hold a bifurcated trial; hearing the issue of liability first, and then, if necessary, contending with the question of the appropriate remedy and Cameron's affirmative defenses. Accordingly, the Court held an extensive non-jury trial on the issue of liability. On two occasions the Court viewed the property at issue. This case was taken under advisement in order to consider the difficult questions presented.

I. THE EVIDENCE

The St. Johns River has its source in Lake Washington in south central Florida. From there it flows north more than 250 miles to the deep water port of Jacksonville and then turns east to the Atlantic Ocean. (See Plaintiff's Exhibit 68). At various places along its course the St. John's widens into broad, shallow lakes. (See Defendant's Exhibit 49). Lake Harney, one such lake, is located about 180 miles upstream from the mouth of the river. The lake is said to be 2 1/2 miles wide and 4 miles long. (See Defendant's Exhibits 47, 48 and 49). Lake Harney reaches its maximum depth of approximately 6 to 8 feet in a navigational channel dredged by the Corps of Engineers (See Plaintiff's Exhibit 68).

Joder Cameron owns a stretch of lakefront property on the northeast end of Lake Harney just as it narrows back into the river. The property has been in the Cameron family for more than thirty-five years and has been used almost exclusively for cattle grazing. In the past, particularly during periods of heavy rain and flooding, at least a portion of the Cameron property would become inundated and thus unsuitable for cattle grazing. The frequency with which this occurred is a fact which is disputed by the parties, but it is at least clear that during the wet portion of the year the lake would rise and water would collect in the lower portions of the property and as a result a fairly large stretch of the property would be saturated. At other times during the average year, when the weather was relatively dry, the waters of Lake Harney would recede, leaving the Cameron property substantially dry.

The periodic inundation of Lake Harney property became a major problem for the Camerons. Each time that the property would become inundated the Camerons were forced to move their cattle to higher ground and they lost the use of the Lake Harney tract for cattle grazing. When Joder Cameron inherited the property from his father in 1968, he decided to put an end to the periodic flooding problem by erecting a dike near the shore of the lake. As Cameron's testimony revealed, he concluded that such a dike would be worth the expense because he could use the property continually for grazing and would not have to worry about providing for his cattle during particularly wet weather.

Cameron set his plan in progress by retaining Malichi Hagan, an experienced land surveyor, to undertake a preliminary survey of the low lying property bordering the lake's shore and to prepare a levee plan. Hagan performed the survey, made a rough sketch of the dike (See Defendant's Exhibit 43) and staked out the area for construction. Under Hagan's plan the dike would completely encompass about 203 acres of the Cameron property. After reviewing the plan, however, Cameron decided to move the dike landward, reducing the area enclosed to approximately 156 acres.

Cameron took the revised levee plan to Albert Clark, a contractor and cattle rancher. Clark agreed to build the dike and to perform other work for Cameron in exchange for a tract of real property (See Defendant's Exhibit 41) valued at approximately $40,000. Construction of the dike began in the summer of 1969; Clark started in the northwest corner of the property, the most landward point, scooping up dirt from both sides of the dragline and piling it before the machine. The work was halted a number of times, once because the dragline broke down. A second dragline was eventually added to the job; and the dike was finally completed in the spring of 1970.

Once the construction was done, Cameron erected a barbed wire fence adjacent to the lakeward side of the dike from the northwest corner of the property to the water's edge (See Plaintiff's Exhibit 30). For more than two years thereafter Cameron used the property for cattle grazing and for mowing and baling hay. Sometime late in 1972 it became apparent to Cameron that the dike had not alleviated all of the water problems on the Lake Harney property; heavy summer rains inundated much of the land despite the presence of the dike and large pools of water collected in various places (See, for example, Plaintiff's Exhibits 14, 15 and 16). Cameron decided that he needed a pump installed to keep his low lying property dry and so he entered into an agreement with another cattle rancher, Elmer Coen. Cameron agreed to give Coen exclusive grazing rights on the property for a period of four years in exchange for the installation by Coen of a water pump. The pump was installed in early 1973 and Coen moved his cattle on the property shortly thereafter.

The Government first learned of the Cameron levee in 1971. Horace B. Parker, Chief of Permits for the Corps in the Palatka district, was investigating a complaint of turbid river conditions when he noticed the dike. Parker returned to the dike the same year after discovering that Cameron had no permit and took a number of photographs of the property. Parker submitted the photographs and a report suggesting that the Rivers and Harbors Act had been violated. In September of 1972 Parker made a third visit to the Cameron property and took additional photographs (See Plaintiff's Exhibit 30). Further investigation by the Corps led to the institution of this action on September 4, 1973.

The evidence offered at trial was quite extensive. Because of the court's structure of the case, the principal question of fact was the location of the ordinary high water mark of Lake Harney. The testimony relating to this question may be classified for convenience into four main types: (1) eyewitness testimony; (2) numerous photographs; (3) expert testimony pertaining to soil and vegetation pattern analyses; and (4) expert testimony pertaining to survey data collected by the Government from two water gauge stations located on the St. Johns River.

A. Eyewitness Testimony

A number of witnesses testified about their personal observations, at various times in the past, of the level of the lake in relation to the Cameron property. Andrew Larson was the principal eyewitness for the Government. Larson, a retired letter carrier, testified that he had been fishing on Lake Harney at least once a week from about 1960 to 1971. At various times during this period, Larson testified, he had seen the lake as high as is shown in Plaintiff's Exhibits 14, 15, and 16, which are aerial photographs of the Cameron property taken in October of 1974. The photographs show that for at least that period in October 1974 the vast majority of the Cameron property enclosed by the dike was completely inundated by water. Larson's observations thus placed the level of the lake well beyond the Cameron dike; indeed, judging by his account, almost the entire 156 acres falls within the reach of Lake Harney.

Parker's testimony for the Government provides a slightly different perspective. Parker, it will be recalled, was the Corps of Engineers permit officer who set the Government's case in motion. Parker testified about his several trips to the Cameron property during the years 1971 through 1974. On each occasion, Parker maintained, he had been able to drive his boat right up to the Cameron dike and set foot on the shore. The waters of Lake Harney, he stated, touched the dike and the area closed by the dike was itself partially inundated on almost every occasion that he had seen it.

On cross examination Parker recalled being on Lake Harney in 1960, long before the Cameron dike was erected, on unrelated Corps business. On that occasion, he stated, the Cameron property looked to him like a prairie-type lake shore.

According to Parker's testimony the water level of the lake was particularly high during his September 1972 inspection; the property enclosed within the dike was also quite wet. The photographs which Parker took at the time, contained in Plaintiff's Exhibit 30, confirm his testimony. It is most apparent from these photographs that the water of the lake extended up to the dike on all three of the lakefront sides. The borrow ditch which runs along the landward side of the dike was also completely filled with water; and at numerous places in the enclosed area there were large pools of water. The vast majority of the Cameron property was obviously too saturated for cattle grazing.

A subsequent examination of the Cameron property by Parker, on January 1974, revealed that the lake water had receded considerably since his September 1972 inspection. The enclosed portion of the dike had largely dried up. However, Parker still discovered places along the dike at which the lake water extended completely to the dike's edge. (See Plaintiff's Exhibit No. 31). Parker also noticed a patch of dead hyacinths extending from the dike to the water's edge (See Plaintiff's Exhibit 34). This indicated that the water had recently extended all the way to the dike but had then receded a considerable distance (See Plaintiff's Exhibit 32). The photographs Parker took during this examination tend to corroborate his testimony (See Plaintiff's Exhibits 31-35).

Joseph T. Lancaster, Chief of the Corps Permit Section, accompanied Parker on his January 1974 visit to the Cameron's site and subsequently took aerial photographs of the property. Lancaster's testimony about his January examination of the dike confirmed Parker's observation; the lake seemed to be at a relatively low stage and it appeared that much of the property lakeward of the dike would ordinarily be under water. Lancaster's photographs, taken during the same month, show that the lake had receded from the dike at nearly all lakefront points. The borrow ditches on either side of the dike, from which Clark had obtained the fill, were however filled with water. The land enclosed by the dike appears, on these photographs, quite dry (See Plaintiff's Exhibits 36-46).

Thomas Talley, a biologist with the Department of the Interior, also investigated the Cameron dike on a number of occasions. His first visit to the scene occurred on May 18, 1972, when Talley and two others landed at the foot of the dike in a boat. Talley was there to determine what effect the Cameron dike might have on the Lake Harney eco-system; he suspected violations of federal law. During his investigation he noticed, as had Parker and Lancaster, that the majority of the area enclosed by the dike was inundated — and he observed ducks and aquatic vegetation in the interior waters. Lake Harney itself, Talley noted, came almost all the way up to the edge of the dike at all lakeward points. The photographs which he took at the time — Plaintiff's Exhibits 51 through 56 — corroborate this testimony.

Talley investigated the Cameron property again in mid April of 1973. At that time the area inside the dike was dry and the water level of the lake was much lower. Talley detected the remains of aquatic vegetation inside the dike. Defendant's Exhibits 2 and 3 and Plaintiff's Exhibits 58 through 66, photographs taken by Talley at the time, confirm his observation that the interior of the dike was dry, indeed, one can detect cattle grazing on the land in Defendant's Exhibit 2. These photographs also show that the interior borrow ditch along the dike was filled with water and the water of the lake comes very close to the dike at all lakeward points.

Seemingly in conflict with the testimony of Larson, Lancaster, Talley and Parker was that of a number of defense witnesses. Joder Cameron testified that with the exception of periods of heavy rain or storms he had never seen the level of the lake rise to such an extent that it touched the area enclosed by the dike.

Cameron's wife, June, stated that she first became acquainted with the Cameron property in the fall of 1968 and that on that occasion the lake was out beyond the point at which the dike is now located. Mrs. Cameron also testified that she has been on the property on numerous occasions since 1968 and that each time the water level was about the same as it was on her first visit — some distance away from the dike line. Moreover, she could recall several times in which cattle were grazing and hay was moved well beyond the point at which the dike was constructed, that is, between the area of the dike and the edge of the lake.

Charles Taber similarly recalled seeing the lake's water well beyond the dike line on a number of occasions. Taber also stated that when Cameron's fence was strung along the southwest side of the dike the land was dry past the dike's end.

Albert Clark's testimony was quite similar. He could recall seeing during the construction of the dike a number of cattle grazing past the dike line. Clark recalled observing, just as had Mrs. Cameron, hay cut on both sides of the Cameron levee. Throughout the construction of the dike, Clark stated, the land was dry and he was never forced to use mats or other wet ground equipment.

Malichi Hagan also recalled observing that the dike line was dry and that cattle were grazing beyond its furthest reach. Two years after the dike had been constructed he returned to the Cameron property and noticed the same conditions. On February 15, 1974, shortly before the trial of this cause, Hagan prepared a survey of the Lake Harney property at Cameron's request (See Defendant's Exhibit 43). Hagan found, among other things, that the water was a considerable distance from the dike at all points along its three lakefront sides.

Two of the defendants expert witnesses, Herbert Gee and Dr. James Lasater, also testified about their personal observations of the Cameron property. Gee, an engineer, noted that when he examined the property on December 11, 1973, he found no evidence of erosion on the lakeward slopes of the dike. This was significant, he felt, because if the lake water had stood against an earthen dike of this nature for any appreciable period of time it would surely have caused some erosion. His conclusion was that except during periods of seasonal flooding the dike did not come in contact with the lake.

Dr. Lasater similarly noted the absence of erosion along the lakeward slopes of the dike. Lasater testified that the only evidence of erosion visible during his March 1974 inspection was along the then existing lake shore at the southwest side of the property — well away from the dike (See Defendant's Exhibits 83-84). This led him to conclude that the high waters of the lake had fallen short of the dike in recent years.

B. Aerial Photographs and Government Surveys

The Government introduced a number of aerial photographs (Plaintiff's Exhibits 1-13) of the Cameron property and Lake Harney, taken at various times from 1948 to 1973. These photographs provide some indication of the degree of fluctuation in the lake's water stages.

Exhibit 1, a photo taken on November 25, 1948, indicates that the Cameron property was at that time completely inundated up to the tree line along the far west end (Compare this with Defendant's Exhibit 43). Exhibit 2, a photograph taken on April 18, 1957, shows that Lake Harney had by that time receded to a considerable extent, perhaps even beyond the Cameron dike line.

Lake Harney was again at low ebb, but not as much as indicated in Exhibit 2, on March 11, 1957, when the photograph in Exhibit 3 was taken. Ten years later, on November 11, 1967, the lake had swelled again to equal the high level obtained in 1948 (Exhibit 4).

The water had receded again by February 21, 1968, when Government's Exhibit 5 was taken. It would appear from this photograph that the Cameron dike would not at that time come in contact with Lake Harney. However, by October 26 the same year the lake had risen once more to its 1948 level, inundating the Cameron property up to the tree line (Exhibit 6). The lake was at about this same level on November 10, 1969, when plaintiff's Exhibit 7 was taken.

By January 17, 1971, a few months after the levee was completed, the waters of Lake Harney had clearly receded again. Exhibit 8 shows that on that day the lake was well away from the dike at all lakefront points. By November of the same year the lake had once again risen, but not to the extent shown in some of the previous photographs (Exhibit 10). The interior of the Cameron property was under water — presumably because of heavy rains — almost back to the tree line. The lake was at about the same level on February 27, 1972 when Exhibit 11 was taken and the Cameron property was even more flooded than before.

By the time Exhibit 12 was taken, on February 4, 1973, the interior of the property was much dryer. The lake, however, remained at a relatively high level and it clearly appears from this photograph that the lake would have engulfed a large portion of the diked area of the Cameron property but for the dike. The condition was essentially the same on February 16, 1973, when the last aerial photograph, Plaintiff's Exhibit 13, was taken.

Defendant's Exhibits 47, 48 and 49 provide additional visual evidence of the lake's former water levels. These exhibits are Department of Commerce Coast and Geodetic Survey Maps of the St. Johns River system. Each map depicts Lake Harney at a different point in time: Exhibit 49 was printed in 1964, Exhibit 47 in 1968 and Exhibit 48 in 1972. Among other things, these survey maps purport to accurately identify the then-existing dimensions of Lake Harney, its "firm shoreline" — which is defined as the "line of mean high water" — and the extent of adjoining marshland.

Of importance here are two common characteristics of these survey maps. First, all three maps characterize the majority of the Cameron property as marshland. Second, in each of these exhibits the "firm shoreline" falls short of the Cameron dike line. This is most apparent from Exhibit 48, which actually identifies and outlines the dike as a landmark on Lake Harney's shore. The dike is clearly shown to be landward of the mean high water line.

Defendant's Exhibit 45, which is a copy of the original United States survey of Lake Harney and part of its surrounding shoreline, provides some additional evidence of the lake's past water levels. The Cameron tract is shown on the Government's map. It is difficult however, to determine from this whether the dike line was at that time within the lake's reach because the survey is of questionable accuracy. Indeed, the surveyor's notes and the map itself caution that the surveyor was unable to traverse the sections "owing to the boggy, muddy, grassy state of the region, etc.".

C. Soil and Vegetation Analysis

Both parties produced expert witnesses who had performed vegetation pattern analyses of the Cameron property. The Government's expert witness was Thomas Talley, who performed his analysis during an April 19, 1973 inspection of the property.

Talley apparently limited his investigation to detecting and analyzing patterns of aquatic vegetation; he did not determine the extent to which the Cameron property was populated by terrestrial vegetation. Talley's findings demonstrate that, at least in April 1973, several types of aquatic or marshland vegetation inhabited the area enclosed by the dike, the property between the dike and the lakeshore, and the land of adjoining riparian owners. His most significant finding, Talley maintained, was the discovery that Eleocharis ("spikebrush") was the dominant species of aquatic plant life for a significant portion of the property enclosed by the dike and the property on the dike's lakeward side. Within the dike, the pattern of Eleocharis ran from immediately behind the borrow ditch to a distance of up to several hundred feet toward the treeline in the far west end of the property. This vegetation pattern is depicted in plaintiff's Exhibit 57 and in a number of photographs taken during the analysis (See Plaintiff's Exhibits 62-64).

The discovery of Eleocharis in such a dense pattern landward of the dike is significant, Talley testified, because this particular species thrives in areas that are inundated over 50% of the time. In addition to finding Eleocharis, Talley encountered a number of other aquatic plants which populate areas of frequent inundation: Scirpus ("bullrush"), Pontaderia ("pickerel weed") Paspalum and Vaginatum.

Talley charted the vegetation patterns of these various aquatic plants on Plaintiff's Exhibit 57. This sketch indicates that a large portion of the property enclosed by the dike is populated by aquatic plants which require significant periods of inundation in order to survive — at least on the order of 50% of the time. To be more specific, by charting the aquatic vegetation patterns in Plaintiff's Exhibit 57, which is a topographical map of the dike, Talley concluded that elevations below two feet would be inundated between 50-75% of the time. And close scrutiny of the elevation patterns on the topographical map reveals that more than a third of the Cameron property is below the two foot elevation — primarily along the north and southwest section of the dike.

Dr. James Lasater was the expert witness for the defense. Lasater, a professor of oceanography and physics, examined the Cameron property on three occasions in March 1974 — slightly over a year after Talley performed his analysis — and performed a number of soil and vegetation tests. At the time of Lasater's visits to the property the water level of the lake had reached well beyond the dike line. (See Defendant's Exhibits 55-95).

Lasater's vegetation analysis focused primarily on measuring the existence of terrestrial plant life on the Cameron property. He found that the property was populated primarily by terrestrial vegetation — both landward and lakeward of the dike. The dominant species, according to Lasater's findings, was bermuda grass. As both Talley and Lasater agreed, bermuda grass is a "transition zone" vegetation; that is, that it is capable of existing in area of periodic inundation but it will not withstand prolonged inundation. Lasater offered the opinion that bermuda grass would have a life span of about two weeks when totally inundated, and Talley placed this estimate closer to three months. Another species which was prevalent on both sides of the dike was bahia grass, which is also incapable of surviving under conditions of prolonged inundation.

The predominance of these terrestrial plants led Lasater to conclude that the area on either side of the dike had not been submerged a significant amount of time in recent years. Another observation which in Lasater's opinion confirmed this conclusion was the sighting of a number of cypress trees on the lakeward side of the north dike. The fact that new knees were not visible on these trees suggested the areas in which they were located had not been inundated a significant portion of the recent time.

Lasater's soil analysis disclosed measurable changes in soil content with each successive examination landward from the then existing lakeshore. The further landward Lasater went, the higher the organic content of the soil. This was significant, Lasater concluded, because it suggested that terrestrial plants populated the soil along the lakeward slope of the dike. When terrestrial plants have been growing in an area, Lasater explained, an organic layer will gradually build up in the soil as a result of plant decay. By contrast, areas populated by aquatic plants will contain subground layers of organic soil.

D. The Survey Evidence

As an additional means of proving that the dike falls within the ordinary high water mark of Lake Harney the Government introduced, through a number of witnesses, the results of a survey of the water stages of a portion of the St. Johns River and a topographical survey of the Cameron property. The Government hoped to prove through a comparison of these surveys that two-thirds of the property between the dike and the treeline at the west end of the Cameron tract was inundated approximately half the time during the periods studied.

The water survey data was collected and analyzed by William Baker, a hydrolics engineer with the Corps. Baker obtained daily elevation readings for a thirty year period — 1943 through 1973 — from each of two United States Geological Survey water gauge stations. One station is located at the State Highway 46 bridge over the St. Johns River, about 5.9 miles south of the Cameron property and the other is located at the U.S. Highway 17-92 bridge over Lake Monroe about 26.1 miles north. These water gauge stations measure and record the daily mean sea level elevations of the river at their respective points.

These elevation readings were then used to develop with a computer "stage duration curves" for the two gauge stations. A stage duration curve is a graphic tool used to illustrate and summarize the elevations of a body of water over a given period of time. The curve is constructed by tabulating the percentage of the total time measured (here 10 years) that given elevations (above mean sea level) are equalled or exceeded.

The stage duration curve Baker developed from the Highway 46 data (Plaintiff's Exhibit 48) indicates that the water stages at that point in the river have equalled or exceeded 2.3 feet above mean sea level approximately 50% of the time. The curve further indicates that about 25% of the time the water surface level has been at an elevation of 4.2 feet or more. The Lake Monroe stage duration curve (Plaintiff's Exhibit 49) registers a lower elevation, the lake reaching an elevation of 1.4 feet or more 50% of the time and 2.7 feet or more 25% of the time.

Once these two stage duration curves had been constructed, Baker attempted to develop a "water stage profile" of the elevation of the river between the Highway 46 and Lake Monroe stations, a distance of some 32 river miles. (See for visual example Defendant's Exhibit 48). This device, which is Plaintiff's Exhibit 50, was designed to graphically illustrate the slope of the river, on the basis of the two stage duration curves, and thus allow the Corps to interpolate the elevation of Lake Harney at the Cameron property.

The stage profile was designed by constructing from the stage duration curve data two "flow lines" or "exceedance frequency curves" which purportedly indicate the elevation by river mile that is obtained either 25% or 50% of the time measured, respectively. These two flow lines appear on Plaintiff's Exhibit 50 and are clearly identified. Baker concluded from an analysis of his stage profile that the variation in the level of the water at the Highway 46 gauge station and the level of the lake at the site of the Cameron dike would not exceed .1 to .2 of a foot. This is a significant finding according to Baker and Angelo Tabita, Chief of the Corps Project Planning Branch, when the stage profile is compared with the Government's topographical survey of the Cameron property.

The Government's survey of the property was conducted by Donald Smith and Joe Locke whose readings appear in Plaintiff's Exhibit 17. With this survey data other Corps personnel — Adolph Hautschlag and Bertha Fredrick — prepared Plaintiff's Exhibit 18, a topographical map of the property with plotted elevations, and Plaintiff's Exhibit 19, a graph of the cross sections of the levee.

The Corps topographic map indicates that about two-thirds of the area enclosed by the Cameron dike occupies an elevation of 2.1 feet or less.

The stage duration curve for the Highway 46 gauge station indicates that the elevation of the river equalled or exceeded 2.3 feet at least 50% of the time between 1943 and 1973. As Baker's testimony reflects, it follows — taking into account the .1 to .2 feet disparity in elevation registered by the stage profile — that the elevation of Lake Harney along the Cameron property must have equalled or exceeded 2.1 feet 50% of the time during the same thirty year period. The alleged significance of this conclusion is this: according to the topographic survey at least two-thirds of the area enclosed by the dike is below the elevation of 2.1 feet; therefore, according to Baker, two-thirds of Cameron's property must have been inundated by the lake 50% of the time over the past thirty years.

This process of reasoning can be carried still further. At least three-quarters of the diked-in area stands at an elevation of 4.0 feet or less. The stage duration curve for the Highway 46 gauge demonstrates that a water elevation of 4.2 feet was equalled or exceeded 25% of the time. Therefore, again taking into account the assumed variance of .1 to .2 feet, a water elevation of 4.0 feet at the Cameron property was equalled or exceeded 25% of the time during the years 1943-1973. It is therefore contended that at least three-quarters of the area has been inundated approximately 25% of the time over this thirty year period. This is particularly significant, according to the Government, because a reasonable measure of the ordinary high water mark of an inland non-tidal waterway would be the point at which the shore is inundated on a frequency level of 25%.

Angelo Tabita testified at length about the methodological validity of Baker's stage profile. It was possible to accurately develop a stage profile from the two distant gauge stations employed, Tabita maintained, because of the relative uniformity in depth. According to Tabita, the slope of the river between these two points is quite flat, falling gradually down river and to the north. The 5.9 miles between the Highway 46 gauge station and the dike are particularly uniform, Tabita testified. He noted that Lake Harney itself is a large, shallow lake almost without a slope from one end to the other and that there is little difference between the slope of the lake and the slope of the river at the Highway 46 station. Tabita conceded on cross examination, however, that for a two week period in May 1968 the slope of the river was actually in the opposite direction from that indicated on the stage profile. That is, for a period of two weeks the elevation of the river was higher at Lake Harney than it was at Highway 46.

Tabita also conceded that Lake Monroe, as opposed to Lake Harney, is a tidal waterway and that Baker did not take into account the fluctuations in the Lake Monroe gauge produced by the tide when he developed the stage profile. He maintained however that this did not affect the accuracy of the stage profile because the effect of the tidal influence on the elevation reading would be almost completely diminished by the volume of the data — since the survey covered a period of thirty years.

Tabita's conclusions were completely disputed by Herbert Gee, the engineer who testified on behalf of the defendant. Gee offered the opinion that Baker's stage profile was substantially inaccurate based upon assumptions that were either incorrect or empirically unverified, methodologically unsound and consequently of very dubious value in identifying with any degree of exactitude the elevation of Lake Harney at the site of the Cameron levee.

According to Gee there have been a number of significant and progressive environmental changes in the upper St. Johns which have cumulatively decreased the amount of water available to Lake Harney and elsewhere on the lower river basin. Because the Government did not make the adjustments in its data required to take these changes into account, Gee concluded, the stage profile is necessarily inaccurate.

One such environmental change, according to Gee, is the periodic deepening of the river channel from the upper reaches of the St. Johns to Sanford. The Corps was authorized by Congress to deepen the channel to ten feet in 1945 and was subsequently given the authority to deepen it still further, to twelve feet. The Corps channel work has had the effect, Gee remarked, of decreasing the water levels on the lower basin, including Lake Harney.

Another significant environmental change, identified by Gee, occurred after the construction of a new Highway 46 bridge in 1959. This bridge is significantly larger than the previous bridge and necessarily changed the flow of that portion of the river. The new bridge, according to Gee, had a backwater effect; that is, it restricted the opening of the flow and consequently extended the duration of the high stages to the south. Water levels at Lake Harney were thus lowered and the Highway 46 gauge consequently recorded a higher elevation for a longer period of time.

A third change in the St. Johns noted by Gee resulted from the construction of the Sabastian Diversion Canal in 1964. Constructed in response to major floods in 1947, 1948 and 1953, the Sabastian Canal is designed to divert a major portion of the flood waters away from the upper basin. The canal has a capacity of 6,000 cubic feet per second. Gee testified that the canal has decreased the quantity of water in the lower basin and has consequently reduced the water stages in the vicinity of Lake Harney.

Still another factor which in Gee's opinion, has altered the character of the St. Johns was the construction of a major water plant by the City of Melbourne in 1960. Serving Eau Gallie and nearby communities as well as Melbourne, the water plant withdraws 10 to 12 million gallons per day — a figure which represents 10% of the base flow of the St. Johns. According to Gee, the Melbourne plant has significantly lowered the water stages in the lower basin, including Lake Harney.

The removal of a center pier from the river in 1964 by the Florida East Coast Railway, Gee maintained, also worked a significant change in the river channel. Again, the effect was to lower the water level in Lake Harney.

Finally, the systematic and dramatic development of the upper St. Johns basin, which has continued unabated for more than twenty years, was identified as a major factor in the alteration of the river. Gee singled out in particular the construction of numerous levees and pumping stations along the upper St. Johns, and the installation of irrigation systems at various points on the river. This rapid development, Gee contended, has significantly reduced the water level on the lower basin. Indeed, Gee stated that the rate of development had caused the Corps of Engineers to fear that there may be a serious deficiency of water in the river in the future.

According to Gee's testimony, the inaccuracies resulting from the Government's failure to take into account these numerous environmental changes in the St. Johns are compounded by the fact that the stage profile is based upon erroneous and/or empirically unverified assumptions. Most prominent among these is the assumption that the long channel between Lake Monroe and Lake Harney is uniform in slope. This, Gee testified, is quite erroneous. The Government's data even belies this notion, Gee testified, because it identified a two week period in which the river sloped upstream rather than downstream as the Government assumed. Gee also noted that the stage profile is based upon unverified notions about the contour and depth of the river channel.

Gee was also of the opinion that the Government's methodology in constructing the stage profile was seriously flawed. In the first place, he noted, the two water gauges which supplied the only data upon which the stage profile is based, were far too remote (32 miles) to allow one to "manufacture" the figures the Government obtained for the Cameron property. Furthermore, Gee observed, the Government ignored the fact that Lake Monroe is a tidal lake. This was a mistake, he suggested, because the attempt to interpolate an elevation between tidal and untidal water gauges was unprecedented and not in accord with sound hydrolic engineering principles.

Finally, Gee stated that the data upon which the stage profile is based was skewed toward a higher than typical elevation by the '47, '48 and '53 floods. The Government did not attempt to correct this distortion and as a consequence, Gee concluded, the frequency curves appearing in the stage profile do not represent exceedance frequency curves which are typical for Lake Harney.

Despite the numerous deficiencies he found in the Government's survey, Gee was of the opinion that a methodologically sound way to interpolate the water stages at the Cameron levee could be devised. Gee gave an example: A survey party could place water gauges at a number of temporary benchmarks between the two gauge systems and then take simultaneous readings of the temporary gauges and the gauge stations. This method, Gee felt, would give one some confidence in the river profile produced. Gee stressed, however, the need for precision in making an interpolation of the stages at the Cameron property because a slight difference in estimated elevation could affect a large quantity of land.

II. THE ADMISSIBILITY OF SURVEY EVIDENCE

An initial question which must be resolved prior to reaching the merits is whether the Government can properly prove the presence of regulatory jurisdiction over the Cameron dike by means of the kind of survey evidence offered at trial. The Court admitted this evidence over the vigorous objections of the Defendant on the ground that it seemed relevant to the central question of fact — where lies the ordinary high water line of Lake Harney? The parties have supplemented their authorities and the Defendant continues to urge that the Government's stage profile and elevation evidence is legally irrelevant to a determination of the ordinary high water line.

Cameron's objection is essentially directed to the proper "test" to be employed by the trier in locating the ordinary high water level. He would reject categorically the use of water stage and elevation data for arriving at this determination, and rely instead upon vegetation analysis or physical inspection of the lake's bank.

The term "ordinary high-water line" has no immutable definition. Although the Supreme Court has long regarded it as the exclusive factor in demarking the limits of the Government's navigational servitude in non-tidal waterways, it has never given the term a precise content. In United States v. Chicago M. St. P P Co., 312 U.S. 592, 61 S.Ct. 772, 85 L.Ed. 1064 (1940) the Court noted only that "[t]he dominant power of the federal [g]overnment . . . extends to the entire bed of a stream, which includes the lands below ordinary high water mark." Id. at 597, 61 S.Ct. at 775. Similar language is found in United States v. Kansas City Life Insurance Co., supra. "High water mark bounds the bed of river . . . [l]ands below that level are subject always to a dominant servitude." 339 U.S. at 805, 70 S.Ct. at 889. And in United States v. Rands, supra, the Court observed in a similar vein that "[t]he power to regulate navigation confers upon the United States a `dominant servitude,' . . . which extends to the entire stream and the stream bed below ordinary high-water mark." 389 U.S. at 123, 88 S.Ct. at 267.

Beyond these and similar observations of general character, the Supreme Court has left it to the lower courts to give meaning to the concept of "ordinary high water mark". In the few cases in which the lower courts have been called upon to apply the concept, one finds little in the way of comprehensive definition. Instead, the Courts have been satisfied to tailor their definitions to meet the precise set of facts before them. One may thus find varied definitions in the decisions.

The ordinary high water line has, for example, been defined as the line where the water stands sufficiently long to destroy vegetation below it. Goose Creek Hunting Club, Inc. v. United States, 518 F.2d 579, 583, 207 Ct.Cl. 323 (1975); Kelley's Creek and Northwestern R.R. v. United States, 100 Ct.Cl. 396, 406 (1943).

It has also been said to be the line which diverts upland from the river bed, the river bed being the "land upon which the action of the water has been so constant as to destroy vegetation." United States v. Chicago B Q R. Co., 90 F.2d 161, 170 (7th Cir. 1937). Another Court has defined the mark as "a natural physical characteristic placed upon the lands by the action of the river." U.S. v. Claridge, 279 F. Supp. 87 (D.Ariz. 1966); aff'd. 416 F.2d 933, cert. denied 397 U.S. 961, 90 S.Ct. 994, 25 L.Ed.2d 253 (1969). The Court in Harrison v. Fite, 148 F. 781 (8th Cir. 1906) defined the ordinary high water line as the line below which the soil is so usually covered by water that it is wrested from vegetation and its value for agricultural purposes destroyed. The Third Circuit similarly defined the term as the line below which the waters have so visibly asserted their dominion that terrestrial plant life ceases to grow and, therefore, the value for agricultural purposes is destroyed. See Borough of Ford City v. United States, 345 F.2d 645, 648 (3rd Cir. 1965), cert. denied, 382 U.S. 902, 86 S.Ct. 236, 15 L.Ed.2d 156.

One can find in the reports numerous other definitions of the term, most varying in some way from the others and designed to provide a working definition for the facts in the particular case. In short, the courts have tended to define the term to fit the method of proof employed: vegetation analysis, physical inspection, etc., See Goose Creek Hunting Club, Inc. v. United States, supra at 583.

What these cases suggest is that the "ordinary high water line" is not readily susceptible to a uniform and precise definition which will provide guidance for each and every case. Rather, the term is best regarded as a concept which denotes the point at which the bed of a lake or river ceases and the shore or fast lands begins, a point which may be capable of proof by a variety of methods depending upon the facts and circumstances of the particular case. Vegetation analysis as Cameron suggests, is the method most often employed. But the cases make it clear that it is not the only method.

As the Third Circuit observed in Borough of Ford City v. United States, supra, another approach is to look for the existence of a clear, natural line impressed on the bank of the river or lake. "If there is a clear line, as shown by erosion, and other easily recognized characteristics such as shelving, change in the character of the soil, destruction of terrestrial vegetation, and litter, it determines the line of ordinary high-water." 345 F.2d at 648. Still another method, though of less reliable character, is to resort to the meander line of the original federal survey. See, e.g., United States v. Otley, 127 F.2d 988 (9th Cir. 1942).

There is no logical reason why a fourth approach to determining the line of ordinary high water may not consist of comparing reliable water stage and elevation data. Indeed, for a body of water whose levels fluctuate considerably with changes in climate, accurate water stage and elevation data may provide the most suitable method for determining the ordinary high water mark.

The Court has found no case in which survey data of this nature has been excluded on the ground that it afforded a legally improper basis for determining the ordinary high water mark. Indeed, in one recent case the Court expressly relied upon such evidence in making its determination. United States v. Parker, No. 75-34 (N.D.Fla. 1976).

The Court therefore rejects Cameron's contention that the Government survey data was as a matter of law an inappropriate method of determining the ordinary high water mark at Lake Harney. The Court also reaffirms its ruling that the survey evidence was admissible as relevant under the particular facts of this case.

III. DECISION

The crucial fact issue at this stage of the proceeding is whether the Cameron dike stands below the ordinary high water mark of Lake Harney. The burden of proof on that issue lies with the government. As we have seen, much evidence has been received on this issue — first hand observations of eyewitnesses, aerial and land photographs, expert vegetation analysis, topographical surveys, water survey data, soil analysis and more. Much of this evidence is conflicting. Upon careful consideration, the court is of the opinion that the government has failed to sustain its burden of proof. The reasons for the court's conclusion can best be explained by discussing the probative value of the evidence in the order in which it was previously summarized.

A. The Eyewitness Accounts, Photographs and Surveys:

The eyewitness accounts of the defense and government appear at first to be in hopeless conflict. It seems to the court, however, that this conflict is more apparent than real. Significantly, most of the accounts pertain to observations of the Cameron property at different times, often in different years. This fact suggests that the conflict in the evidence is attributable not to distorted perceptions or misconceptions but to genuine changes in the physical environment at the Cameron property. In short, it seems that the conflict in the testimony is largely the consequence of the fact that the dike is located in a dynamic environment. During high stages the lake did, as the government witnesses testified, reach the dike; and the area enclosed was frequently inundated by rainfall. But at other times during the same year, as the defense witnesses testified, the property was dry and the lake some distance away from the dike.

Several things support this conclusion. First, the aerial photographs, which were taken at various times since 1948, provide visual evidence of the fluctuation of the water level of Lake Harney over the years. These photographs demonstrate that Lake Harney's water level is subject to rapid change. Second, the Coast and Geodetic Survey maps characterize the Cameron property as marshland adjoining the firm shoreline of Lake Harney. This too suggests that the environment in which the dike stands is a changing one; periodically inundated by rainfall and by a rise in the lake, periodically dry and thus useful for agricultural purposes. Much of the expert testimony, in addition confirms this. A common strain in the testimony of the experts is that this portion of the St. Johns system has few firm river banks; instead the level of the river constantly fluctuates, swelling over the low lying marshes on either side at high or flood stages and gradually retreating during low or average stages.

The eyewitness testimony and photographs suggest another conclusion: the water level of Lake Harney has, on the whole, declined progressively in recent years. Even the Government's witnesses noted that the water of the lake had declined in 1974 to a point well away from the dike. The Court's two views of the property confirm this observation. On both occasions, the water was well away from the dike, the land was dry, and the property was quite suitable for cattle grazing. In addition, the Court observed, as had two defense witnesses, that the lakeward slopes of the dike showed no signs of erosion, a fact which indicates that they had not been in recent contact with the water. Finally, there is the testimony of Herbert Gee, which identified a number of factors which cumulatively have caused a decline in the water levels in the lower river basin, including Lake Harney.

The fact that the water levels of the lake have declined and indeed, appear destined to remain at lower stages, is of obvious importance to this case. All evidence suggests that the dike has historically been located in a dynamic marshland environment, periodically inundated with water and periodically dry. This environment may now be changing. If the water levels continued to decline or remain at low levels it is unlikely that the Cameron property would continue to be affected by temporarily high water stages. It would then be difficult in the extreme to view the dike as lying within the lake's ordinary high water line.

But even ignoring the recent decline in the lake's water levels, the evidence considered thus far is quite inconclusive in determining the present ordinary high water mark. Proof that the land has historically been inundated during periods of heavy rainfall and high water stages is not proof that it lies within the ordinary high water mark. There must be some evidence of how frequently this occurs, and the duration of the inundation. Both parties attempted to supply this kind of evidence by making vegetation analyses of the property.

B. The Vegetation Analysis:

Talley performed a vegetation study of the Cameron property for the government in April 1973; Lasater followed in March 1974 with both vegetation and soil analyses for the defense. As already noted, the results of the two studies are in partial conflict. And as with the testimony of the eyewitnesses, it seems that the discrepancy in the findings of the two researchers is largely attributable to the different times in which the analyses were done.

Talley found the Cameron property in a period of transition. The water level of the lake had only recently receded from a higher stage. The land enclosed by the dike was much dryer than Talley had seen it in May 1972; in fact, cattle were grazing on the property. Talley detected aquatic vegetation at various places in both sides of the dike, a finding which led him to conclude that much of the property was under water for major portions of the time.

Lasater viewed the property a year later, when the land was much dryer and the lake further away from the dike. By then, it seems, the transition to lower stages and dryer land had become complete. As a result Lasater found very little vegetation of an aquatic nature on either side of the dike. Instead, Lasater found that the dominant species of vegetation was terrestrial — consisting of bermuda and bahia grass. It will be recalled that these are the kinds of grass that many defense witnesses had seen growing and harvested for hay at various times in the past. Lasater's soil analysis provided confirmation of his visual sightings of terrestrial vegetation. He encountered increasingly higher contents of organic matter in the soil the closer he came to the dike, which indicated that terrestrial rather than aquatic vegetation had populated the soil along the lakeward slope of the dike and the property enclosed.

The undisputed evidence of terrestrial vegetation outside the dike — as well as inside — makes it impossible to conclude that such area is so usually covered by water that it is wrested from vegetation and its value for agricultural purposes destroyed. Based on the vegetation test, the government has failed to prove that the disputed area lies below the ordinary high water line.

C. The Survey Evidence:

The government's survey evidence is even less probative of the issues in dispute. As the Court has noted, evidence of this nature is potentially quite useful in locating the ordinary high water line of a body of water whose water levels fluctuate considerably. Accurate survey data provide the trier of the fact with an historical picture of the water levels of a body of water over a long period of time and thus enable it to identify with some degree of certitude the points which are inundated so frequently as to be fairly considered within the ordinary high water line.

However, evidence of this nature, like all statistical evidence, can be misleading. The stage profile and the water stage data from which it was constructed are in the Court's opinion, too methodologically and empirically unsound to be useful. Herbert Gee's analysis of the weaknesses in the government's data and its attempt to interpolate the water elevation at the dike is essentially unanswered.

The use of two water gauge stations 32 miles apart to interpolate ostensibly precise water level measurements is itself questionable. But to attempt to arrive at such an interpolation with the use of tidal and non-tidal gauges seems highly dubious. In the Court's opinion the government has failed to prove the trustworthiness of this kind of methodology. The government's stage profile is rejected as having probative value in determining the ordinary high water mark of Lake Harney.

CONCLUSION

The Court thus finds that there is insufficient evidence to determine whether Cameron's dike is or is not located below the line of ordinary high water. At most the evidence shows that the dike stands on property that has historically served as marshland for Lake Harney — periodically absorbing the lake's high waters and periodically serving as productive agricultural land for its owners. This is not enough to infer that the dike stands below the ordinary high water mark — the point at which the bed of the lake ends and the fast lands begin. There must be reliable evidence which affords a basis for determining the frequency with which the property has been inundated by the lake's high waters.

The government has failed to sustain its burden of proof on the issue of regulatory jurisdiction and judgment must therefore be entered in favor of the defendant, Joder Cameron.