Gravel and the Changing Holes

On a spring float written up in the April 20131 issue of the state’s conservation magazine, a writer put in at Buck Hollow and passed Blue Spring, where “the spring feeds an excellent swimming hole.” Aerial photographs of the bend tell the same story for the years around his float. In 1995 a broad, dark pool lay along the outer bank below the spring, beside a big rock that swimmers used for jumping3, and the bars on the inside of the bend were small. The pool had much the same outline in 2009, 2012 and 2014. In the photographs of 2020 the inside of the bend is a sheet of sand and gravel, and the pool has become a narrow channel2 squeezed against the rock and the bluff (Fig. 22).

Nearly three miles4 upstream, at the Highway 17 bridge, two people who swam there in the 1970s and 1980s remember the deepest water at the bridge and along the bluff, with a gravel bar to go in from and places to jump from the bluff5. The photographs there show a narrow river with thin bars through 2012, a new bridge built in the middle 2000s, and bars spreading at the access and beside the bridge by 20206 (Fig. 21).

A swimming hole is a temporary arrangement of gravel, and the Jacks Fork has rearranged its holes for as long as it has carried gravel. What happened at these two holes can be set against what is known of the river’s gravel, how much of it there is, where it came from, how it moves and what people have done to it.

Shoals and pot-holes

Any mile of the Jacks Fork repeats a pattern. A short stretch of quick, shallow water runs over gravel. Below it the river slows and deepens into a long pool, often under a bluff, and beside the pool, usually on the side away from the bluff, lies a bank of bare gravel. Then the river shallows and quickens again (Fig. 15.1). Geologists call the three parts the riffle, the pool and the bar. A geographer who studied the Ozarks a century ago7 described its streams as “a rapid succession of riffles and pools, with the pools flanked by wide white ‘gravel bars.’” On the upper Jacks Fork in a drought summer, probably 1934, people told a writer floating the river that its course was “shoals and pot-holes, jest shoals and pot-holes,” the pot-hole being the still pool under a bluff8.

Each part has its own work. The riffle is a step in the river’s long profile, where the water drops over a sill of gravel. Riffles tend to form at the crossings9, where the main current swings from the outside of one bend over toward the outside of the next and spreads thin across the gravel between them. The pool forms where the current is thrown against something it cannot easily move, most often the bluff on the outside of a bend10, and scours a hole at its foot (Chapter 2). These are the holes people swim in and fish in, and they hold the deepest water in a dry August. The bar builds on the inside of the bend, where the current slackens and drops the gravel it was carrying, and over the years it grows out toward the pool.

On an ordinary summer day the river hardly moves its bed. A stone on the bottom stays where it is until the water pushes on it hard enough to roll it or lift it, and on the Jacks Fork that takes a rise. In a flood the bed itself goes into motion, stones rolling and skipping along the bottom by the thousands, a load the river carries on its floor rather than in its water11, and when the water falls they drop where the current slackens. Gravel comes off one bar, rolls downstream and lodges on another, and a pool fills a little in one flood and may be dug out in the next. The pattern of riffle, pool and bar lasts while the gravel in it is replaced, which is why a ford at a riffle could stay in one place for generations12 (Chapter 8) and a particular hole could not.

Where the gravel came from

The gravel of the Jacks Fork is chert. As the dolomite of the uplands dissolved over millions of years, the chert in it was left behind as a mantle of broken rock in red clay, and that mantle is the river’s supply (Chapters 1 and 2). Chert does not dissolve13 and wears down only slowly, so the pieces last through mile after mile of rolling. The geographer who described the Ozark riffles and pools judged it, apart perhaps from porphyry, “the most durable material in the Ozarks.” Most of it breaks into flattened pieces with curved, shell-like faces, a shape the current moves with difficulty, and on the floor of a typical Ozark stream he found the fragments “little smaller and little less angular than the chert on the hillsides.” A gravelly bed of that kind resists the water better than the banks beside it do, so the stream works sideways, and he saw beds spread to as much as twenty times the width of the water at ordinary stages. Water also moves freely between the stones. A creek that looks dry may have water running under its bed, and “shallow pits dug in the bed of a creek usually fill with water in a short time” (Chapter 14). The river has always carried gravel. The question is how much, and since when.

In the early 1990s the federal hydrologist and the oral historian whose interviews appear in Chapter 14 chose the Jacks Fork as one of two Ozark basins in which to trace the history of land use and the river’s gravel. The Eminence gauge, with its record from 1921, was one reason. They compared land surveys of the 1830s, census returns, old photographs and aerial photographs from 1938 on, and they interviewed older residents in their homes14.

Early descriptions of Ozark streams say little about gravel bars. An 1877 description15 called the Current, the Jacks Fork and the Barren Fork “rapid-flowing and clean, with narrow fertile valleys.” Early travelers rarely described gravel bars, while geologists of the later 1800s found large quantities of gravel in the banks and beds. In 1855 the state’s first geological survey18 reported that “many of our streams abound in water-worn pebbles, which constitute their beds, and form bars along their margins and channels,” and counted the pebbles a resource, valuable “for roads and streets.” The streams had probably always held gravel, then, but had not always been so unstable. They had gone through unstable spells before settlement as well. Old layers of gravel and soil in Ozark stream banks, dated by radiocarbon, record earlier episodes in which channels filled with sediment and shifted, but the episode that followed settlement moved much more gravel and sand19 than any of them. Before settlement they carried a mixed load16, gravel rolling along the bed and silt spread over the bottoms in floods, and the deposits left on old floodplains suggest that the channels were generally narrower and deeper17 than they are now.

The change came in stages tied to land use. Settlers cleared the narrow valley floors for fields and pasture. The timber boom of about 1880 to 1920 (Chapter 7) brought roads along the creeks and across the bars and more clearing in the bottoms. Regional floods in 1895, 1897, 1904 and 1915 struck valleys that had lost much of their streamside timber (Chapter 14). Ozark floods of that time22, as the geographer described them, undercut the banks, uprooted trees, tore down fences, carried off haystacks and shocks of grain, and left the fields “overspread with gravel and driftwood.” From about 1920 to 1960 the woods were burned each spring and grazed by stock on open range (Chapter 5), and the cattle and hogs gathered along the streams and ate the young trees off the bars. By models of how land use governs runoff and erosion, that period should have been the most damaging20, and grazing along the streams was probably the most direct and damaging link21 between land use and unstable streams. Floating the Black River23, the next big river to the east, in the mid-1930s, a writer found the number of hogs at large “incredible”: “At every bend in the stream there stands haunch-deep a band of shoats, grubbing for shell-fish.”

By the mid-1940s it was popularly accepted24 that the filling and instability of the streams came from changes in how the uplands were used, and in 1958 a writer and conservationist25 who knew the Current country well (Chapter 10) put the common view in print: “The thin topsoil washed away down the hollows until there was little left but rocky chert, and the mountain streams became choked with gravel.” Fewer game fish, he wrote, “grew in gravel-choked streams.”

Most of the new gravel probably came out of the small valleys and the valley bottoms, where gravel stored26 under soil and trees was cut loose once clearing and floods opened new channels into it. With the timber gone from their banks, the channels of the small hollows could cut headward, working upstream into ground that had been stable, and the cutting would have released stored gravel27 downstream. A man who had roamed the Big Piney and Jacks Fork basins28 all his life put it plainly: “The gravel was there but it had dirt on it. But the dirt gets washed off of it and then it washed the gravel out.” A man born on an upper-river farm in 190429 said the gravel “was already in the river. It just washed in the deep holes.” The upper Jacks Fork was never used for floating ties, yet its channel was disturbed anyway, so tie drives were not needed30 to fill a river with gravel.

A wave of gravel

Gravel set loose in the side valleys reaches the river and moves down it in pulses, flood by flood, piling up in some reaches while others drain. Geologists call such a pulse a wave. A gauge keeps a record of the riverbed as well as the water, because every time a crew measures the flow it measures the depth across the channel, and from those soundings the average height of the bed31 can be worked out year by year. At the Eminence gauge the average level of the riverbed rose quickly around 1940 and then fell steadily32 for the next half century, probably the trailing edge of a wave of land-use gravel that had started down the river before the gauge was built33. The building of a bridge at the gauge site34 disturbed the bed there as well. Read over the longer span from 1921 to 1990, beside three other Ozark gauges, the bed measurements put the start of the wearing down at Eminence nearer 1930, as at Little Piney Creek, the other smaller basin, while at two gauges on larger rivers the bed built up and wore down by turns, roughly in step with runs of wet and dry years. On that evidence a run of wet or dry years could unsettle an Ozark river even without any change in land use, and the effects of climate and of the wave of land-use gravel are tangled together37 in the record. Long lags35 between changes on the land and the recovery of the streams are to be expected, because gravel moves only in floods and travels a limited distance in each36 before it comes to rest again in a bar.

A generation later the bare gravel bars of the Jacks Fork and the Current were mapped from aerial photographs taken in 1996, 2005, 2010 and 2014. On the rivers mapped, gravel was generally decreasing upstream and increasing downstream38, as moving waves would make it. On the Jacks Fork the gravel gathered in the same places year after year, with more of it below Leatherwood Creek, where the valley widens, and over those eighteen years the maps showed no sign of the gravel moving progressively downstream39 or piling up. The lower river’s share grew. Between 2005 and 2014 the stretch from Alley to the mouth went from about a third of the river’s mapped gravel to nearly two-fifths40.

Both swimming holes lie in a different kind of reach. For roughly its first 35 kilometers inside the park, about 22 miles, the Jacks Fork runs in a narrow valley confined by the Roubidoux Formation41, with a narrow channel and comparatively little gravel. Between walls of rock a flood cannot spread out42 and slow down, so it keeps its speed through the narrows and carries much of its gravel on to the wider valley below. On 2010 photographs the channel at the Highway 17 bridge was about 24 meters43 wide, some 80 feet, and just upstream it was 18, among the narrowest places on the river. From Highway 17 to just below Blue Spring the river held only about 1 to 4 percent44 of the Jacks Fork’s mapped gravel in any of the four years, and neither hole lay in one of the places where gravel gathered year after year45.

How a hole fills

The commonest thing the older residents said in 1993 was that the rivers were filling with gravel, especially the deep pools46 that made the swimming holes and held the fish. Some of their memories were exact.

A man whose family came to the upper basin in 193247 had known one hole since he was eight or nine. It had been “6 to probably 8 feet deep in the deepest parts.” At its lower end lily pads had grown in about four feet of water. In 1993, he said, the deepest place was not over his head, most of it at summer flow would be “a foot or two deep, full of gravel,” and where the lily pads had been “there’s nothing there now, except gravel.” Half of another, the Hiker Hole48, with a bluff on one side, had filled, and he described how:

“The gravel bar has just extended and extended until there used to be just a little bitty edge of gravel over there and now it’s fit itself way the heck out in that deep hole.”

It is a common way for a hole to be lost (Fig. 15.2). The bar on the inside of the bend grows a few feet in each flood. The deep water against the bluff narrows, and if the floods do not scour it back out, the pool shrinks from the side. A pool can also fill from the bottom49. At ordinary flows it is the slackest water in the reach, so gravel rolled into it by a moderate rise settles there, and only a larger flood running hard through the bend can scour it out again. Someone standing on the bar notices it first as distance, when the far bank or the far rock seems closer than it used to be.

Others had seen the same thing elsewhere. A logger born in 191650, who had worked at Rymer’s ranch in the 1930s and later guided fishermen, remembered Dark Hollow Hole below Bunker Hill as fifteen feet deep: “now they say you can wade in there.” Deep holes, he said, formed at the curves, and “where it straightens out is where it filled up more.” Measured on the riverbed survey of 2021–23, Dark Hollow Hole lay about 9 feet below the riffle51 that holds it, roughly its depth at low water (Map 12). A woman born in 1935 remembered a swimming hole near her family’s place, on a branch of the upper river, with a solid-rock bottom, waist-deep when she was thirteen or fourteen and “full of gravel” and “knee deep now”52 by 1993. The man born in 1904 remembered riffles “about waist deep53 and swift enough that you couldn’t stand up in it,” and holes “you could swim a horse in.” “There ain’t no hole in Jacks Fork I think where you could swim a horse,” he said. The gravel had “made the holes just one big continuous riffle.” A woman whose family moved to the upper basin in 191154 said that at the swimming places of her childhood “you can wade it most any of the places, its just the deep places against the bluffs where they fish.”

A woman born in 191755, who farmed on the North Prong from the 1920s, said her parents had kept the stream clear of drift and “it kept in its channel.” On her own farm the trouble began in the middle 1960s, when people above her began cutting trees and felling them into the creek. After that every rise washed gravel and debris down on her, and the creek “just kept changing ’til it spread it out” and “lost its depth, there wasn’t no swimming holes.”

The change did not all run one way. A woman raised on a Jacks Fork farm remembered a swimming hole that filled and emptied: “Sometimes it’s filled up56 and sometimes it’s cleaned out again.” A woman who had farmed on the South Prong since the 1930s57 said the deep holes had disappeared gradually: “It wasn’t anything that happened suddenly.” The man who knew the Hiker Hole had also seen a new hole form, and holes filled one year and flushed the next. On balance, he said, “there just isn’t the kind of water in these holes that there used to be even 15 years ago,” and most of the change had come in the previous twenty or twenty-five years58.

Taken together, the accounts hold both kinds of change. On balance the rivers had aggraded, their beds building up with gravel, so that the channels were shallower and wider, and the change also varied a great deal from place to place. Most people remembered holes that had filled, and a few remembered holes that filled and were dug out again59. Scale decides how such a change looks. At the scale of one pool and riffle, the filling of the pool is a significant event, while over several tens of kilometers of river it may be offset60 by the scouring of another.

People who swim the river today describe the same kind of change. Lynne Norris61 of Winona, who swam at Alley in the 1970s, said of Bluff Hole above the Alley access that it was “about gone last time I was there.”

Gravel taken out

For most of the twentieth century the river’s gravel was a resource. A geologist who mapped the country around Eminence for the state, in a report of 193062, found the valley bottoms floored with boulders, gravel, sand and silt from a few inches to as much as 20 or 30 feet thick, and he called the sand and gravel of the river valleys “an important, but as yet slightly developed resource.” County road crews and farmers took their road metal from it. A man born in 190863 on the upper river remembered the county crew taking about a hundred truckloads a year from his farm in the 1930s: “We got all our gravel for our roads out of the Jacks Fork.” Each spring his family graded a road down to a bar, for gravel and for swimming, “until the Scenic Riverways got it; they wouldn’t let us grade that road. And they said we couldn’t get no gravel.” The man who knew the Hiker Hole remembered the WPA hauling “truckload after truckload from right at the forks of the river,” and the South Prong woman remembered gravel dug from the creeks for State Route 13764 when it was a gravel highway.

Gravel was dug at a bridge after every rise. A man born at Alley Spring in 190765 said that “down here at the bridge” (the interviewers identified it as State Route 17) “they used to go in there every year. The river would get up and put in more gravel and then they would come and set up their gravel plant down there.” He spoke of it in the past tense in 1993 and gave no dates.

In 1998 four gravel operations held permits in the watershed. Digging in a channel can do harm in several ways at once. The pit deepens the channel where it is dug, and the steeper step at its upstream edge can let the river cut back upstream67. Sediment is sent downstream, banks erode faster, the channel grows wider and shallower, the water table under the floodplain drops, and the channel can shift66. By 2001 the whole river was closed to sand and gravel mining from March 15 to June 1568, while fish spawn. In 1993 the most extensive operation was on the South Prong69, where pits ran for a mile above the bridge. The man who knew the Hiker Hole thought it had helped put that stretch “really in bad shape,” though not by itself: “There is no way in the world that you could haul enough gravel out of this river basin to make any effect whatsoever.” Most of the people interviewed saw no harm70 in the gravel taken for roads. At a single pool, digging the bar beside it or the riffle below it can change the hole in a season.

Bridges change rivers too. A bridge narrows the flood flow at its piers and abutments, scouring in some places and dropping gravel in others, and building one alters the stream’s flow both upstream and downstream of it71. At Eminence, abutments, boat ramps and levees have changed the local flow, scouring in some places and leaving gravel in others72. At Alley in 198173 the Park Service found the paved boat ramp near the Highway 106 bridge useless because the river kept dropping gravel at its foot (Chapter 11). The Highway 17 bridge was on the state’s schedule for replacement in 2001 to 200574. The present bridge is dated 200575 in federal bridge records, and its construction shows in photographs of 2006 and 200776 (Chapter 8). At the Prongs a Mountain View resident who grew up on the upper forks said the new Highway Y bridge “changed the river and the swimming hole quite a bit.”

Grass and trees on the bars

A hole can also be lost as a river settles down. When stock ran on the open range they ate the seedlings off the bars. The man born at Alley Spring described the valley bottom near Alley in the 1930s as “clean,” with cattle paths “all up through the bottom,” and what happened when the grazing ended:

“As sycamore brush grows up . . . first thing you know, you got a big gravel bar there. A bunch of that sycamore grow in there and catches that gravel. And it just keeps building in, building in.”

A man who farmed in the upper basin from 193577 remembered stock “in the creek, of course,” near Blue Spring: “The hollows were full of them.” After the bottomland passed to the Riverways in the 1960s78 and the cattle left, the riverside grew up. “Since the riverways,” the logger said of the stretch from Rymers to Alley, “that river bottom and all has growed up and you can’t hardly walk through a lot of it.”

The process has been measured on the Jacks Fork. One cross-section of the river was surveyed six times between 1991 and 1997. Willows and sycamores had sprouted on a gravel bar in bands parallel to the stream. Stems standing in moving water slow it, the way a fence row of brush slows a sheet of runoff in a field, and slow water lets go of what it carries. At each flood the bands slowed the water and sand and gravel dropped among the stems, building low ridges along the rows. As the ridges grew, the flow was squeezed into a narrower channel, and the water ran deeper and swifter79 along the opposite side, where the cut bank stayed steep. At one Jacks Fork bend about ten miles west of Eminence, aerial photographs show the same sequence. The bare gravel there peaked in 1996, and then grass and shrubs took it, and then forest80.

A bar that grows in this way pushes the deep water against the far bank and narrows it. In aerial bar inventories, which map only bare gravel81, the bar drops out of the count once it greens, so this kind of narrowing appears in them as a loss of gravel.

Channels that move

Gravel also moves the river. When a reach receives more gravel than it can pass along, the bars grow and push the main current against the outer bank. The bank erodes, the bar on the inside builds out to follow it, and the bend swings outward year by year. Streamside trees do not always stop it, because roots strengthen a bank only as deep as they reach, and many Ozark banks stand taller than that. The lower bank, unprotected, washes out in floods, and the undercut bank above it slumps into the stream and supplies it with loose sediment85 that is easily carried off. At the bend west of Eminence the bar on the inside grew from 2.1 to 6.2 hectares, about 5 to 15 acres, between 1955 and 199682, and in the disturbed reaches of the Jacks Fork and the Current the channel commonly moved 4 to 30 meters a year83, some 13 to 100 feet. Excess gravel clogs the channel and turns the flow against the outer banks, and that sideways push is what makes the bends migrate. A single wave of gravel probably took about fifty years84 to pass that bend.

Those rates belong to disturbed reaches. Above Bay Creek the bends have mostly held. Laid over today’s channel, the river line on the 1948 topographic map lies typically within about 65 feet of it, within that map’s own accuracy, so the loops themselves have stayed in place for some 75 years, and the changes have come inside them, in the gravel bars and in how much of the channel is open water (Map 11). Below Bay Creek the only older map to set beside today’s channel is the 1917 topographic sheet. Whether the river has moved there or the 1917 surveyors drew it loosely has not been settled (The Open Ledger), but in four places, near miles 27, 28, 30 and 3186, that map’s river line lies 500 feet or more from today’s channel.

A river can also split. A flood can drop enough gravel to raise an island87, and the water divides around it into two or three threads. Measured across the whole river on 2010 photographs, only 25 of its 341 measuring points crossed three or more separate threads of water, in about six clusters. One of them was the Blue Spring bend, where the active channel was about 53 meters wide, some 175 feet, half again the river’s typical width, in a valley only about 100 meters88 across.

The record from the air

The bar maps cover both holes, and with care they can be read for 2005 to 2014. The four sets of photographs differ in season, method and flow. Water hides gravel94, so a photograph taken at higher flow shows less bar than one taken at low water, and a bar map is only as comparable as the flows behind it. The 1996 pictures were taken in April, before the leaves were out, at four to five times summer flow, and mapped by a different method89, so they cannot fairly be set beside the summer photographs. The three later years can be compared cautiously, with the flow on the photograph dates beside them (Table 15.1).

Figure 21 Steel Bridge, 1995–2026

Figure 22 Blue Spring, 1995–2026

Areas are rounded, with duplicated map polygons counted once. The 2010 bars were mapped by a different method. The 2014 photographs span five weeks, during which the flow also rose briefly to about 54 cubic feet a second.

Close to Blue Spring, bare gravel did not grow between 2005 and 2014. In the wider circle it rose between 2005 and 2010, the years of the March 2008 flood92, mostly about a quarter mile upstream and a third to a half mile downstream. Because 2010 was photographed at the higher flow90, which should hide gravel, flow does not explain the rise, but the 2010 bars were mapped by a different method, and in this stretch of river the mapped share of gravel rises and falls with the method91. Only 2005 and 2014 were mapped the same way. At the Highway 17 bridge there was almost no bare gravel in any of the three years, and the nearest substantial bar, about a quarter mile downstream93, held steady or grew a little.

Through the summer of 2014, then, the bar maps show no bar growing into either hole, which agrees with the magazine writer’s “excellent swimming hole” of 2012 or 2013. They have three blind spots. They see only exposed gravel in plan view, as if from straight overhead, so gravel that fills a pool from the bottom never shows, and the mapping recorded nothing of the bed rising or falling95. Bars under the leaves of a narrow reach are easy to miss in summer photographs, since bluffs and streamside trees shade them96. And a bar that has grown up in willows is not counted as gravel at all.

Blue Spring, 1995 to 2026

Fig. 22 traces the bend from the same frame in eight years of aerial imagery. The rock stands on the outer, bluff side of the bend, with the spring’s cave set in behind it, and the bars build on the inside, the campground side. In 1995 a broad, dark pool lay along the outer bank, past the rock, with small bars on the inside. In 2009, 2012 and 2014 the pool held its length and the bars stayed small. In 2020 sand and gravel filled the inside of the bend, and the pool was a narrow channel against the outer bank. The bars broadened in 2022, side channels cut across them in 2023, and in 202697 the deep water still hugged the outer bank. The rock and the bluff did not move, so the change is measured against fixed points.

The change falls between the 2014 and 2020 photographs, and two great floods fall in that gap. The first came at the end of December 2015, when the gauge at Highway 17 reached 21.9 feet, its fourth-highest mark. The second, on April 30, 201798, set records at Alley and Eminence (Chapter 14). The 2017 flood made pronounced changes in the shape of the channel99, and it came after the last of the bar mapping100. The flood of November 2024, second only to 2002101 at Highway 17, came between the 2023 and 2026 images (Reach D).

The bend has looked different before. In a drought summer, probably 1934, a writer floating the river found it below Blue Spring running “deep and clear and indigo,” with “long, full-lipped ways between the shoals”102 (Chapter 9). It is one writer’s impression of a single drought summer, and it records deep water below the spring ninety years ago.

Steel Bridge, 1995 to 2026

Fig. 21 traces the river at the Highway 17 bridge from the same frame. In 1995 the road and bridge were narrow and the bars thin, and in 2005 the road was still gravel, with a bar at the access. In 2006 and 2007, around the building of the new bridge104 (Chapter 8), a wide bare strip ran along the road on both sides of the river, narrowing by 2007. It showed again in 2010, with the pale deck of the new bridge, and by 2012 the approach was paved and the strip was gone. In 2020 bars had widened at the Buck Hollow access below the bridge, with a gravel lot beside the bridge, and in 2026 bars lay both at the access and beside the bridge. Through all those years the channel itself, under the bridge and around the bend to the access, stayed where it was, while the open water between the bars narrowed103.

Which water the Steel Bridge swimming hole was matters for any measurement. A swimmer who spent every summer day of 1984, 1985 and 1986 at the bridge marked the deepest water right at the crossing105. A Mountain View resident who swam there remembered that “it was all gravel bar,” that people went in at the bottom of the hill on the west side of the road and swam “all up and down the bank by the bluff,” and that “the deepest water was along the bluff in the 70’s and 80’s.” The two marks lie within about forty meters106 of each other. Just below the bridge a pool under the north-shore bluff, Whitbeck Hole107, was the swimming hole of the old resort there (Reach D). The bluff the Mountain View resident remembered may be the same one.

Testing the explanations

Each explanation for the change at the two holes can be tested against the evidence (Table 15.2).

Less water. If the river carried less water, the same channel would look narrower. Low-water measurements in the fall of 1966 and August 2006108 show nearly the same flow at Highway 17 and at Blue Spring (Chapter 12), and the gauge records show more water in the river in recent decades109 (Chapter 14). Part of a gravelly river’s low flow travels out of sight between the stones of its bed, though, and the share can grow. As banks erode, the finer soil in them is carried off and the chert stays behind, so the bed becomes more gravelly over time, and a bed with more gravel in it carries more of the base flow underground110 and leaves a shallower stream on top. A hole’s edge moves with the water level, so a dry summer makes the bar wider and the water narrower. A drop of a few inches cannot turn a pool that had to be swum into one that can be crossed on foot.

Gravel after floods. The timing points here. The record flood at Highway 17 came on May 8, 2002111, and through 2014 the bar maps that followed it show no bar growing into either hole. The change at Blue Spring appears between the 2014 and 2020 photographs, the years of the December 2015 and April 2017 floods, and the bars at Steel Bridge spread between the 2012 and 2020 photographs.

A split channel. Blue Spring was already a three-thread bend in 2010. If more of the river now runs down another thread, the water past the rock can shrink with no gravel dropped in the hole. The 2010 water outlines behind that count were not released112, so which thread passed the rock then cannot be read from the published data. Steel Bridge is a single channel.

Bars growing up in trees. The Blue Spring bars have not been grazed for half a century, and a bar that grows up in willows and sycamores pushes deep water against the far bank. Photographs of the bar opposite the rock, old and new, would show whether it has grown up. In the 2020 image113 the inside of the bend is bare sand and gravel.

The bridge and the road work. At Steel Bridge the most obvious change was the bridge itself, dated 2005 in federal bridge records114, with its construction strip visible from the air in 2006 and 2007115. Long before that, gravel was taken at the bridge116 after floods, and the end of the digging could have let the pool fill. The bars near the bridge stayed small through 2014, so if the bridge work changed the hole, it did so under the water or slowly.

Memory. Every adult who goes back to a childhood place finds it smaller, and a child’s “over my head” is shallower water than an adult’s. At Blue Spring the comparison is anchored to a rock and a bluff that have not moved, and the aerial record shows the change. At Steel Bridge, with no single fixed rock and two candidate pools, memory weighs more heavily.

For the river’s flood maps, the depth along its deepest line, the thalweg, was sounded and the bridge cross-sections surveyed between December 2021 and May 2023. Those measurements were joined to lidar of the ground, a survey made by laser from the air, in a model of the valley from the South Prong access to within half a mile of the mouth117. The model records the riverbed at both holes as it was before the November 2024 flood, a baseline against which a later survey can measure the change at each hole.

Keeping the holes

No one can put a particular hole back. The gravel already in the river will keep moving for decades, and floods will dig some holes and fill others. The evidence points to the conditions under which holes form and re-form at something like the river’s own pace.

The strip of trees along the banks matters most. The condition of the streamside timber probably set the limit of how much disturbance118 an Ozark stream could take before its channel came apart. In 2001 about 29 percent of the Jacks Fork’s length, 14.5 miles, had streamside corridors less than three-quarters forested, and about two-thirds of the length of each of the two prongs119. The state conservation department set a goal, with willing landowners, of a 25 percent increase over twenty-five years in the streams with a wooded corridor more than 100 feet wide, and it ranked the North Prong120 among the places where that would do the most good. Keeping cattle out of the streams, limiting gravel removal and building roads and crossings with care are the other parts of the same work (Chapter 17). Inside the park, vehicles may cross the river only at the authorized fords121 (Chapter 8; Part Six).

Much of the best evidence about the holes came from people who paid attention: the men and women who remembered how deep a hole had been, and the swimmers who marked on a map122 where the deepest water used to be. A photograph of a favorite hole taken from the same spot each summer, with the date written on it and the gauge reading looked up, adds to that record.