Carpenter Ant

Camponotus pennsylvanicus

How do you know it is a carpenter ant and not one of the others?

Start with size, then check the shape of the back.

The black carpenter ant is the largest ant most people in this region will ever find in a building. Put one beside the small dark ants that trail along a kitchen counter and there is no comparison — it is a different order of insect. What confuses people is that a single colony produces workers across a wide range of sizes, from small individuals up to substantial ones, so a trail can look like two species at once. That size variation is itself diagnostic. The other common indoor ants here are all essentially uniform.

Then look at the profile from the side, which is the check that professionals actually use. On a carpenter ant, the upper surface of the thorax runs in one smooth, even curve from the neck to the waist, with no dips, humps or spines. That evenly arched back separates it immediately from the pavement ant, which carries a pair of small spines at the rear of the thorax and has a visibly uneven outline, and from the odorous house ant, whose thorax has a distinct dip in the middle of it.

Two further features confirm it. There is a single node — the little raised segment at the waist — rather than two. And at the very tip of the abdomen there is a small circular opening ringed with a neat fringe of hairs, which is where formic acid comes out. You need magnification for that one, but it is the definitive character in the region.

Color on Camponotus pennsylvanicus is uniformly black or nearly so, with a covering of fine pale golden hairs across the abdomen that catch the light. Related species turn up occasionally with reddish or two-tone bodies, and they behave much the same way.

The published measurements are worth carrying in your head, because they are what a specimen gets checked against. Penn State Extension puts workers at a quarter to half an inch, dark shiny brown to black, with winged reproductives running up to three-quarters of an inch. University of Minnesota Extension gives a similar band of three-eighths to half an inch for the common species and names the two structural characters together: one segment at the waist, and a thorax that is evenly rounded seen from the side. That pairing is the whole identification in eight words, and it is the reason the side view matters more than any photograph taken from above.

The winged reproductives cause the most confusion, because they appear in spring at the same windows as termite alates. The ant version has the pinched waist, elbowed antennae and a front pair of wings clearly longer than the back pair. The comparison is set out fully under the eastern subterranean termite profile, and it is worth doing carefully, because the two insects lead to entirely different work.

Why does an insect chew wood it cannot digest?

Because it needs the space, not the material.

Carpenter ants have no capacity to break down cellulose. A termite carries gut microorganisms that turn wood into food; this insect does not, and never has. Its diet is what any omnivorous ant eats — honeydew from aphids and scale insects, nectar, other insects living and dead, and whatever is available in a kitchen. The framing in your wall is not on the menu.

What the framing offers is a cavity. A colony needs a defensible, humid, dark space to hold brood, and hollowing one out of a solid piece of wood produces exactly that. But cutting through sound, dry, dense timber with mandibles is expensive work, and colonies behave like anything else that has to budget effort: they take the path of least resistance. Wood that has been sitting wet is softer, easier to cut, and already holds the humidity the brood requires.

This is the whole reason carpenter ants function as a diagnostic. They are not distributed randomly through a house. They are concentrated wherever water has been getting in and staying in, which means finding the colony is really an exercise in finding the leak. The insect has done the survey for you.

The spoil from that excavation gets carried out and dumped. Because the ants are not eating it, all of it has to go somewhere, and that somewhere is directly below a small opening in the gallery. This is the mechanism behind the single most useful piece of field evidence on the job, covered further down.

One consequence worth naming: the galleries themselves look nothing like termite workings. Carpenter ants keep house. The tunnels are smooth-sided, clean, almost sanded-looking, and follow the grain in broad sweeping chambers with no debris in them. Termite galleries are packed with soil and fecal material. If you break open a joist and the tunnels are clean, you are looking at ants.

Where is the colony, and where is the part you are seeing?

Almost certainly in two different places, and that is the fact that decides whether the job succeeds.

A carpenter ant population is split. The main nest holds the queen, the eggs and the youngest larvae, and it has a hard requirement: consistently high humidity, because eggs and small larvae dry out and die. In practice that puts it in something thoroughly wet — a hollow tree, a rotting stump, a buried root mass, a landscape tie, a fence post at grade, the underside of a woodpile that has been sitting on the earth for two seasons. On a wooded North Shore property in Cold Spring Harbor or Lloyd Harbor, there can be several such sites within sight of the back door and none of them visible from the driveway.

The offshoots are different. Once brood is past its earliest stage it tolerates drier conditions, so the colony establishes secondary sites for older larvae, pupae, workers and developing reproductives — and those sites can be anywhere that is dark and enclosed. A wall cavity above a bathroom. The boxed void over a bay window. The hollow of a porch column. The space behind a dishwasher. Insulation in an attic knee wall. A dropped soffit above kitchen cabinets. Some of these contain almost no wood at all; the ants are occupying an existing cavity rather than cutting one.

Traffic runs constantly between the two. A house can have ants indoors, active galleries in a wall, and spoil on the sill — while the queen sits in a stump beyond the property line. That is why the arithmetic of a contact treatment never works out. Remove the indoor population and the outdoor nest supplies more; remove the foragers and the colony simply produces replacements.

It is also why the material has to be something the ants carry rather than something they die on. Workers pass liquid food to one another mouth to mouth, and a slow-acting bait taken up on a trail travels through that network to places nobody has found. A fast knockdown does the opposite: it stops the courier before delivery. Anyone who has watched an aggressive spray produce three dead ants and a full trail again on Thursday has seen the principle demonstrated.

What does the calendar tell you?

More than most people expect. When you see them narrows down where they are before anyone opens anything.

January and February. Ants walking across a warm floor in the depths of winter is the single most informative sighting of the year, and it is bad news delivered clearly. A colony living outdoors in a stump is shut down in cold weather; it is not sending foragers into your kitchen in February. Ants moving indoors in midwinter are living inside the heated envelope, and the building has convinced them it is April. This finding alone justifies opening an investigation.

March into May. General resumption as the ground warms, and the period when winged reproductives appear. Large winged ants gathering at an indoor window mean a colony old enough to invest in reproduction has been established for years. Outdoors, this is also when foraging trails become easy to follow along fence rails, deck edges, tree limbs touching the roof and the tops of retaining walls.

June through August. Peak foraging, expanding range, and the season when a colony’s dietary preference shifts. This is a practical detail rather than a curiosity: a bait taken enthusiastically in May can be ignored in August, and the correct response is to change what is offered rather than to decide bait does not work.

September into October. Activity contracts. Colonies with satellite sites inside a structure consolidate there, and this is a good window to close the routes at sills, utility penetrations and wherever foundation planting touches siding.

Weather overlays the season. Heavy rain drives foraging indoors quickly, and a stretch of wet weather in a house with a marginal flashing detail will produce a trail within days. Repeated complaints that track the rain rather than the temperature are a reliable pointer to a water fault rather than a food attraction.

What are they actually damaging?

Wood that is already in trouble, which changes how the risk should be read.

The pace is slower than termite attack. A colony expands galleries gradually, and a house is not hollowed out in a season. But the framing they choose is framing that has been wet long enough to soften, which means fungal decay has been running in the same member, often for years. When a rotted band joist is opened up, the loss of section is a combined result and it is rarely worth arguing about which organism did more.

The members involved are predictable on Long Island housing stock. Band joists behind a deck ledger bolted on with no flashing — a very common 1970s and 1980s detail, and one of the most reliable carpenter ant sites in the region. Sill plates on the shaded side of a house where grade has crept up. Window sills and the framing under a leaking exterior casing. The framing around a shower pan that failed a decade ago and has been wetting a stud bay quietly ever since. Porch beams and columns, particularly where a deck surface above drains onto the beam below. Roof framing under a chimney saddle or a valley that has been letting water in during driving rain. Attic sheathing at a ridge or gable where condensation has been forming.

There is one further cost worth naming, and it is not structural. A colony living in a wall void near a kitchen forages across food-preparation surfaces, and for a food business, an assisted-living kitchen or a restaurant in a Brooklyn or Manhattan storefront, that is a compliance problem irrespective of what the ants are doing to the joists. The commercial version of that work sits under commercial pest management.

What evidence should you be looking for?

Spoil, sound, and the trail — in that order of usefulness.

Spoil. The excavated material pushed out of a gallery is the best single piece of evidence available, because unlike a foraging ant, it does not move. It falls straight down from a slit-shaped opening that usually reads as an ordinary construction joint, so a drift of pale fibrous material on a windowsill, along the top edge of a baseboard, on a cellar floor beneath a joist, on the flat of a porch beam or in the corner of a garage is pointing vertically upward at a gallery. Sweep it away and mark the date. If it reappears within days, the gallery is live and being enlarged. If it does not, you may be looking at a historic colony, and establishing that before opening a wall is worth the wait.

Sound. A colony working inside a void makes a faint, dry rustling, most audible in a quiet house late at night with an ear against the surface. Tapping along a member and listening for the pitch to change where the wood behind has been hollowed is the same principle applied by hand.

The trail. Follow it after dark with a flashlight rather than during the day. Carpenter ants run along edges — the top of a baseboard, the underside of a counter, a deck rail, a wire, a branch touching the roof, the lip of a gutter. Note the direction of travel: ants heading one way with nothing and returning the other way carrying something tell you which end is the nest.

Moisture readings. Because the whole distribution follows water, a meter on suspect framing narrows the search faster than any amount of looking. Painted trim can look perfect and be saturated.

Winged ants at a window in spring, and the shed wings that go with them, confirm colony maturity and locate it roughly, since alates emerge close to where they were reared.

Which building details on Long Island and in the city produce this problem?

The ones that let water sit in framing, and there is a short list that recurs constantly.

Deck ledgers without flashing. Bolted straight to the band joist, with the deck boards shedding water into the joint for thirty years. This is the classic finding on 1970s–1990s Long Island construction.

Gutters and grade. Undersized or blocked gutters, downspouts discharging at the wall, a splash block that has tipped, and beds built up over the decades until the soil level sits above the sill. Water that should be leaving the property is instead being delivered to the framing.

Roof-to-wall junctions. Missing or reverse-lapped step flashing where a lower roof meets a wall, chimney saddles that were never properly formed, and the kick-out at the bottom of a roof-wall intersection, which is the detail most often omitted entirely and which sends water straight down inside the wall.

Porches, columns and bay windows. Porch beams under a draining deck, columns bearing on a surface that ponds, and the small boxed roof over a bay window, which is a flat area with a difficult junction and a void immediately behind it.

Older frame construction on the North Shore. Houses in Northport, Centerport and Huntington Bay with fieldstone or rubble foundations, low crawlspaces, and heavy tree cover keeping the north and east walls permanently damp. Add waterfront humidity and you have framing that never fully dries between events.

City buildings. Carpenter ants are less prominent than in the suburbs but far from absent. They turn up in the wooden joists of brownstone rear extensions, in cornice and parapet framing where the roof edge has been leaking, in the timber of a roof deck, and in planter boxes and terrace framing on upper-floor apartments. In an attached house, the wet joist end pocketed into a party wall is the site to look at.

Vegetation in contact with the structure. Overhanging limbs, ivy on a wall, shrubs pressed against siding. These are bridges past every ground-level measure, and they hold moisture against the cladding as well.

How is a carpenter ant colony actually treated?

By locating it, feeding it something it will carry, and fixing the water — with the emphasis in that order.

Identification and inspection first. Species is established from specimens rather than from a description, because a method suited to this insect can make a budding species considerably worse. Then the survey: trails and their direction, spoil and what sits above it, moisture readings on suspect members, exterior water management, and the interior areas where activity has been reported. Where evidence points into a void, the void is investigated before anything goes into it — a small inspection hole and a borescope tells you whether ants are actually there.

Targeted treatment of located galleries and voids. Where a satellite site is precisely identified, direct application into that cavity is quick and effective. The operative word is precisely; treating a wall on a guess is how a job runs for two years.

Bait on active trails. Placed where the ants are already traveling, matched to what the colony is currently taking, and refreshed as that preference changes. Bait is the part of the work that reaches a nest we have never seen, including one outside the property line.

Non-repellent exterior work where it is appropriate, on the routes rather than as a blanket band.

Correction. The flashing, the gutter, the grade, the crawlspace, the rotted section replaced. This is where the recurrence is decided, and it is why an ant proposal from us sometimes reads like a carpentry scope. Where the sealing element is substantial it belongs under structural exclusion in its own right.

Follow-up. Bait-led work needs checking: placements consumed, trails abandoned, spoil no longer reappearing. A colony that has fragmented needs finding rather than assuming.

What actually keeps them out?

Dry framing and no bridges. Everything else is secondary.

Keep water moving away from the building: clean gutters sized for the roof, downspouts extended well clear of the foundation, grade sloping away, and no bed built up above the sill. Repair flashing properly rather than sealing it with a bead of caulk, particularly at deck ledgers, roof-wall junctions and chimneys. Fix plumbing leaks the week you find them, not the season after. Ventilate or seal crawlspaces so the floor system above them is not sitting in humidity, and put a barrier over bare earth.

Then remove the staging areas. Firewood off the ground and away from the wall. Stumps ground out rather than left. Landscape ties and old fence posts replaced. Dead limbs and rotten sections of tree taken down, especially where a branch touches a roof or a gutter. Cut foundation plantings back so there is a visible air gap between shrub and siding.

Then close the routes: gaps where siding terminates, unsealed utility penetrations, the annulus around an air-conditioning line set, the joint at a door threshold, and openings around window and door casings. None of this stops a colony that is already in a wall, and it should not be sold as if it does. What it does is stop the next one from finding the building convenient.

Related work

Treatment for this species sits under ant control, which covers method, materials and the moisture investigation in detail.

The two profiles worth reading alongside it are the eastern subterranean termite, for the spring identification problem and because the two insects exploit the same wet framing from different directions, and the odorous house ant, which is the species most often misidentified as a small carpenter ant and which needs an almost opposite approach — patient baiting rather than anything that disturbs a trail.

If you have found fibrous spoil under a windowsill or a beam and want to know what is wet above it, contact us with a photograph and the location in the building. The areas we cover across Nassau, Suffolk, Manhattan, Brooklyn and Queens are listed under locations.

Treatment for this pest is covered under Ant Control.