119 N.W.2d 135 (1963)
The STATE of Iowa, Appellee,
v.
Frank H. RAYMOND et al., Appellants.
No. 50812.
Supreme Court of Iowa.
January 15, 1963.
Welch & Welch, Logan, and K. C. Acrea, Missouri Valley, for appellants.
Evan Hultman, Atty. Gen., John M. Creger, Asst. Atty. Gen., and Michael Murray, Logan, for appellee.
MOORE, Justice.
Stabilization of the Missouri River channel between Sioux City and Omaha by the United States Corps of Engineers has caused many legal problems to be brought to our attention. This is another such case. Plaintiff, State of Iowa, brought this action claiming absolute and unqualified legal ownership to the real property in Harrison County, described in its petition and known generally as Deer Island. Plaintiff claimed the land in controversy developed in the form of an island that arose in the bed of the Missouri River commencing in 1938. Defendants contended the property formed as a direct accretion to their riparian lands commencing in October 1939 and remained joined to their land until the April 1943 flood. They also claimed they had for many years paid taxes on the land and asserted an estoppel against plaintiff. The trial court quieted title in the state. Most of defendants have appealed.
Defendants argue two propositions for reversal. First, the court erred in finding the land was an island at all times and second, in finding against their claim of estoppel.
I. The parties do not disagree on the rules of law applicable to questions of accretion and title to land so formed. We have often said, to constitute accretion there must be a gradual and imperceptible addition to the shore line by action of the water to which the land is contiguous. Payne v. Hall, 192 Iowa 780, 185 N.W. 912; Meeker v. Kautz, 213 Iowa 370, 239 N.W. 27; Solomon v. Sioux City, 243 Iowa 634, 51 N.W.2d 472; Wilcox v. Pinney, 250 Iowa 1378, 98 N.W.2d 720. This rule applies to land added to an island as well as to the mainland. Bigelow v. Hoover, 85 Iowa 161, 52 N.W. 124; Meeker v. Kautz, supra. See also 45 Iowa L.Rev. 945, 42 Iowa L.Rev. 58.
The title to accreted lands does not vest in the riparian owner of the land bordering a navigable stream until the surface of such accretion arises above the ordinary high-water mark. Payne v. Hall, 192 Iowa 780, 185 N.W. 912; City of Cedar Rapids v. Marshall, 199 Iowa 1262, 203 N.W. 932; Meeker v. Kautz, supra; Solomon v. Sioux City, 243 Iowa 634, 51 N.W.2d 472.
In Meeker v. Kautz, 213 Iowa 370, 372, 239 N.W. 27, 28, it is said: "`High-water mark' means what its language imports a (high) water mark. It is co-ordinate with the limit of the bed of the water; and that only is to be considered the bed which the water occupies sufficiently long and continuously to wrest it from vegetation, and destroy its value for agricultural purposes."
It is conceded by all parties that the Missouri River was and is a navigable stream and the land in question is in Iowa. The State of Iowa, in the absence of a conveyance thereof, is the owner of the bed or channel of the river from the center or thread thereof to the high-water mark of the stream. Payne v. Hall, supra; City of Sioux City v. Betz, 232 Iowa 84, 4 N.W.2d 872; Solomon v. Sioux City, supra, and citations. If an island emerges from the river it becomes the property of the State of Iowa. Holman v. Hodges, 112 Iowa 714, 84 N.W. 950, 58 L.R.A. 673; East Omaha Land Co. v. Hanson, 117 Iowa 96, 90 N.W. 705; Solomon v. Sioux City, supra; Tyson v. State of Iowa, 8 Cir., 283 F.2d 802. See also 45 Iowa L.Rev. 945 and Underhill, Determination of Rights to Real Property along the Missouri River in connection with River Stabilization, 42 Iowa L.Rev. 58.
In the light of the above stated general propositions it becomes apparent the first proposition argued must be determined from the facts as established by the evidence.
This case, being in equity, is triable de novo in this court. Many exhibits in the form of plats, charts, and ground and aerial photographs, covering various periods of time were introduced. Several "river men" testified as to their observations and experiences as hunters and fishermen before, during and after the formation of the land involved. All parties agree it formed as accretion. These and other witnesses used the exhibits as aids to their testimony in pointing out spots indicated. Thus in addition to observing the appearance and conduct of the witnesses, the trial court was better able to determine their credibility. The evidence was also clarified when, by agreement of the parties, the experienced trial court was permitted to visit and observe the area involved. We are, therefore, justified in giving more than casual attention to its findings, especially where, as here, it is clearly borne out by convincing evidence.
Some of the trial court's findings are:
"The land in controversy comprises an area of approximately five hundred acres. The Missouri River in the vicinity flows generally from north to south. The center of the designed channel which is the boundary between Iowa and Nebraska is west of the land involved. Before 1937 the thalweg of the river flowed close to the east high bank. In August, 1937, the Army Engineers Corps began construction of a system of pile dikes designed to divert the river channel toward the west into a `designed channel'. By September, 1939, the work done had partially accomplished its purpose and the main bed of the river had been diverted into a channel three-quarters of a mile west of its old channel. The diversion of the channel was not accomplished suddenly but by gradual shifting over a period of about one year, so that none of the land in controversy is identifiable as having existed before 1938. * *
"The first identifiable land appears to have risen sometime after 1938 between the old and new channels of the river. The land became clearly identifiable and appears to have risen above ordinary high-water mark at about this time or at least by the spring of 1939. The river was thereby separated into two channels. From this time on the work of the engineers was directed mainly to closing the east channel. This work was suspended shortly before 1943 and resumed in 1958. Beginning in June 1952 with the completion of the Randall dam the engineers were able to exercise some control over the volume of flow in the river, hence over the water level or river stages in the lower reaches. Official gauge readings are taken at Decatur, Nebraska which is approximately twenty miles upstream from `Deer Island.' * * * In this case a series of aerial photographs have been received in evidence which show the shape of the river at Deer Island at various times between 1938 and 1959, correlated with the river stage readings at Decatur, corrected as above noted for the level at Deer Island and afford a reasonably accurate continuity during the span of time covered. All of these exhibits with one exception (Exhibit 12) show a channel on each side of the land in question. The east channel as shown can hardly be described as a chute or swale; it is a broad and well defined channel. In the later pictures, the vegetation and height of land, insofar as the latter can be judged from photographs, indicate that the land has assumed a character of greater size and stability than what is commonly described as a sandbar, reef or towhead in the river.
"Exhibit 12 was taken December 16, 1944 when the river stage was at one of the lowest points of the year, 1015 feet, eight feet below summer stage and thirteen feet below normal high water. Notwithstanding, it shows a substantial east channel. Seven of the remaining photographs were all taken when the river was from 1.6 to 6.2 feet below normal, three were taken at stages above normal, and three at times not correlated to the river stages but all of them show the east channel flowing water over its entire length. Testimony of many witnesses, familiar with the river relating their observations at various times during all of this period of time, strongly supports the conclusion indicated from an examination of the exhibits.
"Mr. R. L. Huber, Chief design engineer for the army, testified that in order to close the east channel a rock dike was constructed across the channel. This project was necessary, as he stated it, to prevent the flow of water in that channel at navigation stage. (A "normal" stage). That they could not afford to lose the water from the navigation channel and wanted to assure that all the flow was in the designed channel at normal navigation stage. The closure was finished in the fall of 1958 and in the summer of 1959 there was a head of water above the closure two or three feet higher than below. During the winter of 1958 one of the contractors doing river work built a dirt causeway across the channel below the closure to move machinery onto the island. This causeway had been washed out by the following August as the result of water flowing in the channel over or through the closure. Since the closure, water appears to have been running through the east channel covering an area almost equal to the surface area of the west channel.
"Witnesses who have had frequent occasion to be on the river in the area fully substantiate the foregoing. During the trial a visit to the area was made in company with counsel for both sides. The land was approached from the east over the defendants' property and a crossing to the island made over the rock closure. The defendants' land east of the east channel is marked by seriated ridges and valleys running parallel to the east bank. These are vestiges of the accretion which has taken place many years before by a filling in of the chutes formed along the river in successive stages, and partially but not completely leveled by the erosive action of surface water. The season at the time of the visit had been dryer than usual. Otherwise these valleys would have in part at least, appeared as potholes. The same phenomena is evident on the island but in this case the ridges and valleys run more nearly northwest-southeast generally parallel to the bend of the river at the north end of the island and are less pronounced toward the south. The old east channel is smooth on its bed with little or no vegetation on the lower part of its banks. The water marks on the bank of the island are plainly visible as is the vegetation line.
"The vegetation on the island, while not as large as that east of the high bank of the channel, except where some clearing has been done, is no less dense. Its size, type, character and density do not suggest a recent origin. It is undoubtedly virgin land. Recent surveys reveal that both high banks along the east channel are generally equal in height and the channel itself, for the most part devoid of vegetation, is fully four hundred to over six hundred feet wide at the top with steep slopes on either side, a comparatively flat profile across the bottom and some thirteen to eighteen feet deep. * * *
"The facts found, as above related, show quite conclusively that the land in controversy began as accretion to the bed of the river. Originally, no doubt as a bar deposited by the action of the river. The development occurred independently from the east bank and had reached an elevation above normal high-water mark while the east channel carried live water at all stages of river flow. The surfaces of the land exhibits the characteristics common to land in place created by a process of accretion. The relative elevation at various points with respect to the east high bank and the comparative size, density and characteristics of the overgrowth together with the present width, depth, course and scouring of the old channel indicate that the island had achieved mature status as land in place long before the ultimate closure or filling of the old channel. That the closure in 1958 did to some extent, in the recent past, and will in the future, dry up part or all of the old channel, or that the action of the river water combined with surface waters naturally draining into it will fill the channel with soil from the river or the littoral may be conceded. But these events and processes are no longer relevant to the central question. The island had long before achieved its character."
No benefit would result in discussing the evidence in greater detail. It is enough to say that the formation of the island was entirely distinct from any accretion to defendants' land. After a careful study of the entire record we agree with the trial court on the first proposition argued.
II. On defendants' affirmative defense that plaintiff was estopped in its claim of ownership to the land the trial court found:
"In 1957 some of the defendants set fire to the vegetation on the land. They cleared a small area and built some fences in 1958 and 1959. The plaintiff `posted' the land against trespassing. Some of this litigation was commenced promptly thereafter.
"A plat of accretion land in the general area was made and recorded in 1905. Again in 1941 the county auditor certified a plat which was recorded August 18, 1941. The evidence is far from clear with respect to the location and identity of lands taxed as `accretion lands,' `accretion lots' or `accretion' to lands or lots owned by them. Much of the land owned by them east of the high bank is accretion land which has been so described and identified since the plat of 1905. There is some suggestion in the record that a small part of the land in controversy may have been taxed to some of the defendants.
"The defendants have not shown that they paid taxes on all the land in controversy or on any substantial part of it. Nor is it shown the amount of taxes paid by any of them on the land claimed. Tax receipts received in evidence show valuations on `accretion lands,' which might be some of the land in question, in no case more than $1.50 per acre and in most instances less. The tax paid must have been minuscule."
The trial court's findings correctly reflect the record made on this issue.
An estoppel is not favored in the law and strict proof of all its elements is demanded. Anfenson v. Banks, 180 Iowa 1066, 163 N.W. 608, L.R.A.1918D, 482; City Bank of Mitchellville v. Alcorn, 188 Iowa 592, 176 N.W. 628; Knapp v. Knapp, 251 Iowa 44, 99 N.W.2d 396.
The burden of proof was on defendants to prove either benefit to plaintiff or substantial detriment to themselves as it is a fundamental principle of estoppel that the party against whom the estoppel is asserted must receive a benefit, or the party asserting the estoppel has changed his position to his detriment. City Bank of Mitchellville v. Alcorn, supra; Wertz v. Shane, 216 Iowa 768, 249 N.W. 661; Knapp v. Knapp, supra.
19 Am.Jur., Estoppel, section 85, pages 735-739, states:
"Injury or prejudice.Estoppel rests largely upon injury or prejudice to the rights of him who asserts it. Since the function and purpose of the doctrine are the prevention of fraud and injustice, there can be no estoppel where there is no loss, injury, damage, or prejudice to the party claiming it. Moreover, the injury or prejudice involved must be actual and substantial, and not merely technical or formal. * * * In other words, to entitle a party to the benefit of an estoppel, he must have been misled and induced to alter his position in such a way that he will suffer injury if the estoppel is denied and the other person is not held to the representation or attitude on which the estoppel is predicated."
Defendants rely heavily on State of Iowa v. Carr, 191 F. 257 (8th Cir. 1911) and several of our cases cited therein. While they hold the State of Iowa may be estopped upon a proper showing of facts, none hold the mere payment of taxes sufficient on which to base an estoppel. In the Carr case the complainants and their grantors had been in possession of the land in question for more than 20 years before the state made any claim to it, the state had levied taxes and complainants paid them. They had made improvements on the land at an expense of many tens of thousands of dollars. We have no such proof in this case. Defendants' attempt to take possession resulted in immediate injunction proceedings by the state which were joined with the case before us.
We hold defendants' evidence insufficient to establish any estoppel against plaintiff.
We find no error. This case is affirmed. The cause is remanded for a supplemental decree correctly describing the land in question which was reserved by the trial court in the decree from which defendants appealed. Affirmed.
All Justices concur, except PETERSON, J., who takes no part.