How wood gets wet, and why carpenter ants follow

Moisture and the insects that follow it · reviewed August 2026

What does “wet wood” actually mean, in numbers?

It means a moisture content, and the numbers are far more useful than the adjective.

Wood holds water in two ways: bound within the cell walls, and free in the cell cavities. The cell walls fill first, and the point at which they are saturated but the cavities are still empty is the fiber saturation point, which the Forest Products Laboratory puts at an average of about 30 percent moisture content. That number is the threshold that matters here, because the laboratory’s own statement is that serious decay occurs only when the moisture content of the wood is above it.

Underneath that threshold sits a second one that belongs to the insects rather than the fungi. Penn State’s extension guidance on carpenter ants notes that they seldom tunnel into dry, sound wood and typically avoid wood with a moisture content below 15 percent. Nebraska’s guide puts it slightly differently and usefully: carpenter ants prefer to nest in moist wood, but wood that was previously saturated may be soft enough for them to hollow out as well.

So there are three bands, and a building is somewhere in each of them at every point on its frame.

Below about 15 percent the timber is doing its job and nothing much wants it. This is roughly where framing sits inside a conditioned building; the laboratory’s tables put wood at 9.2 to 9.5 percent moisture content at 68 to 70 degrees Fahrenheit and 50 percent relative humidity, and around 11 percent as humidity rises to 60.

Between roughly 15 and 30 percent the wood is damp, not yet decaying seriously, and increasingly interesting to an insect looking for something soft. This is the band nobody notices, because the member looks fine, feels firm to a knuckle and has paint on it.

Above about 30 percent the fungi go to work, and everything accelerates. Once decay has softened a member it becomes trivially excavatable, which is when a satellite colony moves in and the problem becomes visible to the owner.

The whole of the rest of this piece is about how a member gets from the first band to the third, how long it takes, and what that timeline tells you when the ants finally show up.

What does decay do to wood before anyone can see it?

Most of the damage, is the short answer, and this is the fact that should change how a soft-looking joist is treated.

The Forest Products Laboratory reports that losses in toughness ranged from 6 percent to more than 50 percent by the time just 1 percent of the wood’s weight had been lost. One percent weight loss is not visible. It is not a stain, not a soft spot, not a change in color that anyone would notice on a joist in a crawlspace with a flashlight. By the point at which a member looks obviously rotten, the structural conversation happened some time ago.

What the fungi are doing during that invisible phase depends on which of them arrived, and the distinction is worth being able to read on site.

Brown rot removes the cellulose extensively, leaving a browner material that cracks across the grain, shrinks and can collapse. That cross-grain cracking into cubes is the pattern people recognize and misname as dry rot. Brown-rotted wood is exactly the substrate a carpenter ant colony wants, because the structural fiber has already been taken out of it and the remaining material comes away in crumbs.

White rot removes both lignin and cellulose, so the wood loses color and appears whiter than normal, and it does not crack across the grain. It tends to feel spongy and fibrous rather than crumbly.

The conditions the fungi need are unglamorous: mild temperature, moisture and air. The laboratory notes that decay is relatively slow below 50 degrees Fahrenheit and above 95, and essentially ceases at around 35 degrees or 100. A Long Island crawlspace spends a large part of the year comfortably inside the productive band, which is why the season here is long even though the winter is real.

And then the sentence that makes all of it actionable: wood will not decay if it is kept air dry, and decay already present from a prior infection will not progress. That is a genuinely hopeful statement, and it is the entire argument for treating a pest call as a water call.

Why do carpenter ants want decayed wood specifically?

Because they are excavating rather than eating, and excavation is a labor cost.

Penn State’s guidance is unambiguous that carpenter ants do not consume wood as food, which is why the excavated particles get dumped outside the nest instead of passing through the animal. The galleries they leave are smooth-sided and contain no soil particles or fecal pellets, which is the single cleanest way to distinguish their work from a termite’s. Nebraska describes the debris as a coarse, sawdust-like material pushed out of the nest, often forming a cone-shaped pile just below the entrance hole, and Penn State adds that the pile contains dead insect parts and other refuse alongside the wood.

Given that the wood is not food, the colony’s interest in it is purely as a place to live, and the calculation is simple. A gallery in sound timber costs a great deal of worker effort. A gallery in wood that brown rot has already reduced to something friable costs almost nothing. So the colony goes where the work has been done for it, and that is why Cornell’s field writing describes carpenter ants as an indicator species that tends to nest in wood damaged by moisture, with their presence suggestive of a roof leak, clogged gutters, poor drainage or another structural fault producing water-damaged wood.

There is a second consequence that follows from the same logic and that surprises people: the ants are not attacking the building. They are occupying a part of it that already failed. Killing them changes the occupancy and not the vacancy, so the site remains available, and on a wet north elevation the next colony to find it usually does so within a year or two.

That is the sense in which an ant job on this kind of structure is a survey with an insect attached, and it is why the ant control page describes the first visit as mostly a moisture inspection.

What is a carpenter ant colony a dated record of?

Years, and this is the part of the argument that generally lands hardest with an owner.

Colonies do not produce winged reproductives on demand. Nebraska’s guide states that a colony may produce swarmers when it gets very large — around six to ten years old with more than 2,000 workers. Penn State puts the same threshold slightly differently: a colony usually does not produce winged males and queens until it is several years old and has about 2,000 to 3,000 workers.

Now read that backwards from a kitchen window in April with a dozen large winged ants gathering at the light. Those insects are evidence of a colony that reached reproductive size, and reaching reproductive size takes years, and the colony has been sitting in wood that was soft enough to excavate for the whole of that period. The wet member did not become wet in March. It has been in the top moisture band, on and off, for most of a decade.

Two practical consequences follow.

The first is that the repair scope is nearly always larger than the sighting suggests, and an owner who has budgeted for an insect problem is budgeting for the wrong thing. A decade of moisture in a band joist is a decade of decay in the sheathing behind it, the sill under it and the subfloor above it, and none of that recovered when the leak was quietly fixed by a roofer four years ago.

The second is that the timing of the sighting is itself diagnostic. Penn State notes that reproductives leave the nest in the spring or early summer after overwintering, and puts the useful window for watching foraging workers at between ten at night and two in the morning, from May through July. Winged ants indoors in the dead of winter are not a swarm arriving from outside; they are a colony inside a heated envelope that has been persuaded it is spring, which localises the nest to the structure immediately.

Why does a wet member so rarely dry out on its own?

Because drying is a function of the air the wood is exposed to, and the assemblies where this problem lives are specifically designed to have very little air movement in them.

Wood comes to equilibrium with its surroundings. The Forest Products Laboratory’s figures for that equilibrium — about 9 percent at 70 degrees and half humidity, about 11 percent at 60 percent humidity — describe a member in a room. They do not describe a band joist behind fiberglass batts in a rim cavity, a sill plate under a floor deck in an unvented crawlspace over bare soil, a porch beam boxed in with trim, or the underside of a deck ledger bolted flat against sheathing. Those are enclosed spaces adjacent to a moisture source with no path for the water to leave, and they can sit for years at a humidity that the conditioned rooms three feet away never see.

Three specific situations dominate the calls.

Enclosure. A member that is wet and covered dries an order of magnitude more slowly than one that is wet and exposed. Insulation is not the villain here, but insulation packed against a cold, damp surface holds the moisture at the surface rather than letting it evaporate, and Penn State’s own list of common nest sites includes the space under insulation between ceiling joists and wall studs for exactly this reason.

Ground contact and near-ground contact. Wood in contact with soil, or separated from it by an inch of mulch that has crept up over the decades, is wicking continuously. There is no seasonal drying phase in that arrangement; there is a wet season and a slightly less wet season.

Recurrence. Many of these members are not continuously wet at all. They are wet for three days after every substantial rain, or for a fortnight after every freeze-thaw cycle that produces an ice dam, and dry in between. Since decay resumes whenever the wood is back above the threshold, a member that is above 30 percent for sixty days a year is decaying for sixty days a year, indefinitely. Owners often reject the diagnosis on the grounds that the wood is dry when they check it, and they are checking on the wrong day.

This is also why “we fixed the leak” is a necessary but not sufficient statement. Fixing the source stops the addition of water. Whether the member then dries depends entirely on whether anything opened it up to air.

Which buildings here produce this most reliably?

Older frame housing with a water problem it has had for a long time, which on the North Shore is a large fraction of the stock.

The recurring profile on Long Island is a frame house of some age with a fieldstone or rubble foundation, a crawlspace or a part-cellar over bare soil, and forty years of landscaping that has raised the bed level toward the sill. Add a shoreline that keeps ambient humidity high for months, add ice damming on a north roof that backs water under the shingles at the eaves each winter, add a porch whose framing was built close to grade, and the moisture map of the property predicts where the galleries will be with unnerving accuracy. Huntington, Northport, Cold Spring Harbor and Lloyd Harbor produce this pattern repeatedly, and so does anything built as a summer house and later winterised, because winterising adds heat and insulation to an assembly that was never detailed to dry.

A second, quite different profile shows up on post-war subdivisions across Nassau and central Suffolk: a slab or shallow crawl, a deck added in the 1970s or 1980s with the ledger bolted straight to the band joist and no flashing behind it, and a downspout discharging at the corner. The deck ledger detail alone accounts for a remarkable share of the carpenter ant calls we take, because it is a horizontal joint that collects water, sits behind a board nobody can see, and stays wet for days after rain.

In the city the geometry changes and the principle does not. A brownstone or rowhouse concentrates its moisture problems at the rear extension roof, the parapet and its coping, the party wall where two roof planes and two owners meet, the cellar joist ends buried in masonry, and the areaway that drains slowly. Timber bearing into a wet masonry pocket is the classic city version of the wet band joist, and it is much harder to inspect because the wet end of the member is inside the wall. Where an extension has been added over a yard, the junction between old and new is where to look first.

How do you measure this rather than guess at it?

With three tools and about twenty minutes, and none of it requires a license.

A moisture meter is the only instrument that answers the actual question, and the pin type is more useful here than the pinless, because you want a reading at depth rather than at a painted surface. Read the suspect member and then read a member of the same species and dimension in a part of the building you believe to be dry, so you have a baseline for that structure rather than a number from a table. The bands above give you the interpretation: near 9 to 12 percent is fine, high teens is a member that is being wetted from somewhere, and anything approaching or over 30 percent is a member in active decay conditions.

A probe — an awl, a screwdriver, a stiff blade — tells you what the meter cannot, which is where the soft zone begins and ends. Sound timber resists and splinters; brown-rotted timber gives and crumbles. Probe at intervals along the member and you will map the wet zone, and its shape usually points at the source, because the widest, softest part is closest to where the water arrives.

A flashlight held at a low angle across a surface reveals what a flashlight held square to it hides: the shallow depression where a member has shrunk and collapsed, the shadow of a gallery under thin paint, the fan of frass that has been swept and is coming back.

Confirm before you open anything up. Painted trim can look sound and be saturated, and a member can read wet at one end and normal at the other, which is itself a finding rather than an inconsistency.

Why does foam belong in this conversation as well as the rodent one?

Because carpenter ants excavate it, and because it hides the very evidence this whole method depends on.

Penn State’s guidance lists it directly: carpenter ants may excavate moist, rotting wood and other soft materials, such as foamed plastic insulation board, to make satellite nests. That is not a marginal observation. A board of rigid foam behind a rim joist, at a band joist, under a slab edge or over a sheathed wall is a material with no structural fiber, no gnaw resistance, no insect resistance, and it is warm.

Add the second problem, which is diagnostic rather than physical. Foam is opaque, it fills the void completely, and it is the material a previous contractor most often used to close a gap. So a wall that has been foamed cannot be read. The frass has nowhere to fall, the moisture cannot be probed, the gallery cannot be seen, and the borescope hole gives you a view of yellow plastic. Several of the longest-running ant problems we have inherited from other firms have had that in common.

Graduate does not use foam in any role — not as a barrier, not as backing behind one, not as an air seal over one. In this cluster the reason is slightly different from the rodent one, and it is worth stating separately: in the moisture context, foam is both a nesting substrate and a blindfold.

What does fixing the water actually involve, and in what order?

Upstream first, then drying, then repair, and the order is not negotiable if you would like the repair to last.

Stop the addition. Whatever is putting the water into the assembly gets corrected before anything is replaced. On the properties we survey that is most often gutter and leader discharge landing within a foot of the foundation, grade that has crept up and now falls toward the house, flashing that is missing or lapped the wrong way at a roof-to-wall junction or a deck ledger, or a slow supply-line leak behind an appliance. Replacing a rotted member without correcting the source produces a new member that decays on the same schedule as the old one.

Open it and dry it. The laboratory’s statement that decay does not progress in air-dry wood is the whole basis for this step. Remove the enclosure, get air across the member, and give it time — weeks rather than days for a substantial timber, and longer in a crawlspace than in a rim cavity. Where the space is chronically humid, the drying step includes the ground cover and the ventilation rather than only the member.

Replace to sound material. Because strength is lost long before appearance changes, the cut has to go back past where the probe says the softness stops, not to where it looks acceptable. Where the member is a sill plate, a girder end or a joist bearing, that is a structural decision and belongs with a competent carpenter or engineer rather than with us.

Detail it so it does not recur. Clearance between wood and soil, a capillary break at bearing points, flashing that sheds water outward, and an assembly that can dry in at least one direction. This is the part that turns a repair into a correction.

Then deal with the colony, which by this stage is a comparatively small job and a much more findable one, because the wet member you have just been working on is where the nest was.

When is it a termite question instead, and where does that go?

When you find mud tubes or discarded wings rather than coarse frass, and it goes to another firm.

The two organisms produce similar-looking damage to a non-specialist and completely different galleries to anyone who has seen both. Carpenter ant galleries are smooth-sided, clean and free of soil particles and fecal pellets, with the debris ejected as a coarse sawdust-like pile below the entrance. Subterranean termite workings carry soil, are packed rather than clean, and the animals build shelter tubes across masonry because they cannot tolerate exposure. Discarded wings on a sill in spring belong to a termite swarm far more often than to an ant one.

Graduate does not do termite work. The service was retired, and this is one of the few places on this site where the honest answer costs us the job. What we will do is settle the identification, because that decides which trade you are calling, and the comparison lives on the eastern subterranean termite profile in the pest library, kept there deliberately as identification material with no service behind it. If it turns out to be termites, engage a company that treats them.

What does all of this mean for how an ant job is scoped?

That the proposal will contain items that do not look like pest control, and that this is the point rather than an upsell.

A scope written from this reasoning has a moisture section before it has an insect section. It names the source, the affected members, the drying strategy and the detail change, and it treats the colony as the last item rather than the first. It also says what is not ours: carpentry, roofing, gutter replacement and grading are trades in their own right, and where a scope needs them we say so and coordinate rather than pretend.

If you are reading your own building, the sequence at the top of this piece is the whole method. Find the wet member with a meter and a probe, follow the water upstream rather than assuming it entered where the damage is, work out roughly how long it has been happening from what the colony has managed to become, and fix the water before anyone talks about the ants. The carpenter ant profile covers identification and the carpenter bee profile covers the other timber insect that turns up on the same elevations. The work behind all of it is ant control, and everything else written in this cluster is indexed on the blog.

The work behind this piece is Ant Control, and everything else in this cluster is indexed on the blog hub.