How floodplain trees survive high water

A flood does not drown a tree. It suffocates the roots, and trees of river bottoms survive by getting air down to theirs. Water fills the soil pores, the microbes living in it use up the last of the dissolved oxygen within a few hours, and roots that cannot respire stop taking up water and minerals. They carry equipment upland trees do not. They grow roots out of the stem into the water, the breathing pores in their bark swell, and gas-filled tissue carries air down the root. Their seed is timed to land when the water falls.

Why standing water kills roots

Roots burn sugar with oxygen to pay for everything they do, and the energy released by anaerobic respiration is not enough to keep them working. Oxygen moves through water roughly ten thousand times more slowly than it moves through air, so once water fills the pore space the soil is cut off from the atmosphere. What oxygen remains dissolved in the water and the soil is consumed by microorganisms within a few hours. Below the top few millimeters, bacteria that do not need free oxygen take over, and their metabolism produces the reduced iron and manganese that are themselves toxic to roots.

Timing decides most of the outcome. A tree flooded in January is asleep, with roots respiring slowly and no leaves to supply, while the same flood in July hits roots that are growing and demanding oxygen at their highest rate.

“Flooding during the dormant season is much less harmful than flooding during the growing season.” (T.T. Kozlowski, Soil Aeration, Flooding, and Tree Growth, 1985)

Three other variables matter. Moving water injures trees far less than stagnant water does, and growth is reduced even in bald cypress when the water sits still. Older trees tolerate flooding better than seedlings and saplings. And duration is the blunt one: sensitive species escape injury if the water is gone within seven days, and suffer severe injury or die when it persists over their roots for a month or less.

Tolerance is a class, not a yes or no, and the classes are set by how long roots can stay submerged in the growing season.

How a flooded tree gets air to its roots

The first answer is new roots. Once the original root system begins to die back, a flood-tolerant tree grows fresh roots into the water itself and into the mud around the stem. They come from the submerged part of the stem and from the surviving major roots. These are adventitious roots, roots that sprout from the stem itself rather than from other roots. The capacity to make them tracks flood tolerance closely: willows, tupelos and sycamore produce them in quantity when flooded. Paper birch and Aleppo pine produce few or none. In green ash, flooded seedlings with adventitious roots absorbed up to ninety percent more water than flooded seedlings without them. The stomata, shut for two weeks by the flood, reopened as the new roots appeared. Eastern cottonwood is typical: the roots die after more than a month under water, and adventitious roots grow from dormant buds in the main trunk.

The second answer is plumbing. Bark lenticels, the breathing pores of the stem, swell on a flooded trunk into enlarged openings that aerate the stem and the roots and let toxic compounds escape. Inside the root, the tree builds aerenchyma, a spongy, air-filled tissue that lets oxygen diffuse down from the shoot. Tolerant species also seal the sides of the root, a barrier to radial oxygen loss, so the oxygen reaches the growing tip instead of leaking out on the way. The same barrier blocks soil toxins on the way in. Bark does other work as well, including photosynthesis without leaves.

Some of that oxygen leaves the root on purpose. Flood-tolerant trees move oxygen down from aerial tissue and release it into the mud around the fine roots, where it oxidizes the reduced iron and manganese before they reach the root surface. Adventitious roots on flooded black gum seedlings did exactly that, and unflooded roots of the same species did not.

The third answer is shape. Old bald cypress trunks are massive and tapering, and in swamps they are buttressed at the base, a flare that widens where the trunk meets ground whose moisture is abundant and fairly permanent.

What cypress knees actually do

Cypress knees are woody protrusions that rise out of a horizontal root and stand clear of the mud and water. They are the most famous structure on a floodplain tree and the least settled.

The Forest Service describes them as structural:

“They start out as small swellings on the upper surface of a horizontal root and then protrude above the mud and water providing extra needed support.” (USDA Forest Service, Fire Effects Information System, Taxodium ascendens, T. distichum)

The textbook explanation most people have heard instead is aeration: that knees are pneumatophores, roots specialized for gas exchange in wet soil. An Ohio State extension review of the question counted seven competing explanations for what the knees do. It found the aeration case weak: knees lack the anatomical features that define a pneumatophore, and sealed-knee experiments measured lower oxygen uptake than other plants manage. Knees are also a poor index of wetness, turning up on trees in dry soil and missing from trees in deep water. Vegetative reproduction, methane venting, mechanical support, nutrient capture and carbohydrate storage were each examined and each left unproven. A week after publishing, the same author added a correction: a paper in the International Journal of Plant Sciences reported that knees do function as pneumatophores after all. The honest summary is that knee height tracks the water level, that support and aeration both have evidence behind them, and that nobody has closed the case.

Silt around the trunk, and why fill kills an upland tree

Every flood leaves sediment, and on a river bottom it mounds against trunks. This looks like the fill that kills yard trees, and physically it is the same thing. The threshold is three inches: silt or sand deeper than that can impede the movement of oxygen to the roots, and small or newly planted trees suffer first.

Piling soil against a trunk buries the root flare and is one of the planting mistakes that kills an upland tree slowly. The measurements are stark: under clay fill placed around tree stems, soil oxygen fell to one percent and carbon dioxide rose above twenty percent, and the deeper the fill the worse both numbers got. Fill around stems is one of the commonest causes of oxygen deficiency in developed areas.

What separates a floodplain species is not that burial does it no harm. What these species have is a reply: a stem that goes under grows new roots from dormant buds in the trunk. An oak or a sugar maple has no such reply and simply loses the roots it had. How far that carries against a deep burial is not something the sources cited here settle, and the practical rule does not change. On a yard tree the sediment should come off where it can. On a river bottom the tree is standing on ground that was laid down the same way.

Seed timed to the flood

Adaptation does not stop at the adult tree. Bottomland species fit their reproduction to the flood calendar so that seedlings arrive on bare, wet, newly exposed ground rather than under water.

Overcup oak is the clearest case. Its acorn has a spongy shell that makes it buoyant, so floodwater rather than a squirrel is the dispersal agent. Acorns have turned up on a northwestern Florida island beach miles from any apparent seed source. The acorn stays dormant over winter and germinates in spring after the water recedes. The tree itself waits: it leafs out late, sometimes a month behind the trees around it. That late emergence is probably part of its flood tolerance.

Bald cypress works the other way round. Its seed is too heavy to travel far on wind, so floodwater moves it along rivers and streams. Germination needs the water to go down: seed submerged for as long as thirty months may still germinate once it does. On well-drained ground the seed usually fails for want of surface water. Eastern cottonwood is tuned to the same signal, dispersing and germinating from late spring through midsummer as flows fall, onto the freshly deposited alluvium the flood leaves exposed.

Choosing for a wet spot or a stream bank

The class list above was written for Iowa and leaves out most of the southern bottomland flora. But the ranking holds, and it contains a warning about the tree people most often picture standing in a river.

“American sycamore is intolerant of flooding during the growing season and will die if the entire tree is inundated for more than 2 weeks” (USDA Forest Service, Fire Effects Information System, Platanus occidentalis)

Sycamore belongs on wet ground that drains, not in water that stands through July, and the same caution applies to river birch and American elm.

Where our own wet-ground plants fall:

Two of the nine take a summer under water. The rest are plants of wet ground and short or seasonal floods, which describes most wet corners. A spot that drowns every spring and drains by June suits all nine. A spot that holds water through July suits the cypress and the overcup oak.

For the corner that drowns every spring we sell six plants together at Blue Whale Trees as the Wet Spot Pack: bald cypress, sweetbay magnolia, silky dogwood, winterberry, smooth alder and swamp doghobble. Every plant in that set is a species of wet ground, and the cypress takes the wettest of it. For a bottomland oak, the red oaks and white oaks split sorts the candidates.

Three rules follow from the science above. Plant while the tree is dormant, so its first flood arrives before the roots are working hard. Do not judge a spot by one dry summer, because what matters is how long water stands in June and July, not whether the ground is damp in April. And match the species to the depth and the duration rather than to the look of the place. The site check asks for the moisture at the spot and says whether a given species belongs there.

Common questions

How long can a tree stand in floodwater before it dies?

Flood-sensitive species generally escape injury if the water is gone within seven days, the moderately tolerant ones have survived a full growing season submerged, and the tolerant ones have survived two.

Do bald cypress knees help the tree breathe?

That is the traditional explanation and it is still unresolved: knees lack the anatomy of a true pneumatophore and sealed-knee tests measured lower oxygen uptake than other plants manage. But a later paper in the International Journal of Plant Sciences reported that they do work as pneumatophores.

Is silt piled around the trunk after a flood a problem?

On a yard tree, yes: more than three inches of silt or sand can impede oxygen reaching the roots, and it should be taken off where that is possible. A floodplain species handles it better because a stem that goes under grows new roots from trunk buds, but a deep burial is not harmless to it either.

Sources

• Soil Aeration, Flooding, and Tree Growth: T.T. Kozlowski, Arboriculture & Urban Forestry 11 (1985): how flooding strips soil oxygen, why season and stagnation matter, and the adventitious roots and hypertrophied lenticels that answer it.
• Plant Morphological, Physiological and Anatomical Adaption to Flooding Stress and the Underlying Molecular Mechanisms: Jia, Ma, Chen and Wu, International Journal of Molecular Sciences 22 (2021): oxygen diffusion in water, aerenchyma and barriers to radial oxygen loss.
• Taxodium ascendens, T. distichum, Fire Effects Information System: USDA Forest Service: buttressed trunks, how knees form, and the drawdown that cypress seed needs to germinate.
• Quercus lyrata, Fire Effects Information System: USDA Forest Service: the buoyant acorn, spring germination after floods recede, and late leaf-out as flood tolerance.
• Populus deltoides, Fire Effects Information System: USDA Forest Service: adventitious roots from trunk buds, and seed release timed to falling water and fresh alluvium.
• Platanus occidentalis, Fire Effects Information System: USDA Forest Service: sycamore’s flood-tolerance rating and its limit during the growing season.
• Effects of Flooding on Trees & Shrubs: Iowa State University Extension and Outreach: the flood-tolerance classes by species and the depth of sediment that starts to matter.
• We Speak For The Knees?: Ohio State University Extension: the seven hypotheses for cypress knees, the evidence against aeration, and the correction reporting the opposite.
• Overcup Oak: USDA Forest Service, Silvics of North America: overcup oak surviving continuous flooding for at least two growing seasons in tests.
• Quercus michauxii: North Carolina State Extension Plant Toolbox: swamp chestnut oak tolerating wet soils and occasional flooding.
• Nyssa sylvatica, N. biflora, Fire Effects Information System: USDA Forest Service: black gum tolerating brief spring flooding on alluvial sites.
• Magnolia virginiana, Fire Effects Information System: USDA Forest Service: sweetbay on sites flooded less than five months a year, and continued flooding not tolerated.
• Effects of flooding on woody landscape plants: University of Wisconsin Extension, publication A3871: silky dogwood and winterberry able to tolerate more than a week with roots submerged, winterberry with periodic flooding only.