Almost every predator problem we are asked to look at has already been diagnosed by the person asking, and about half the time the diagnosis is wrong. That matters more than it sounds, because the countermeasure follows the diagnosis. A fence built against a fox does nothing about a mustelid, and overhead netting does nothing about anything that arrives from below.
Key takeaways
- Evidence at the scene narrows the species; a trail camera confirms it. Guessing costs more than the camera does.
- Half-inch mesh in 19-gauge wire or heavier is the specification the trade converges on for exclusion. Standard hexagonal chicken wire keeps birds in, not predators out.
- Under Directive 2009/147/EC all wild bird species are protected in the EU, so the lawful response to raptors is physical exclusion and cover, never lethal control.
- Wild birds attracted to spilled feed are a disease pathway as well as a predation question, which puts perimeter management inside your biosecurity plan.
Read the scene before you buy anything
The evidence left behind narrows the candidate list faster than any product catalogue. The patterns below are field practice rather than a published identification key, so treat each one as a hypothesis to confirm with a camera rather than a verdict.
| What you find | Points towards | What it changes in the fix |
|---|---|---|
| Birds missing entirely, no remains, scattered feathers outside | A carrying predator, canid or large raptor | Perimeter integrity and overhead cover |
| Several birds killed, little eaten, carcasses left in place | Surplus killing behaviour, often mustelids | Mesh aperture, every gap under 25 mm matters |
| Excavation under the fence line or coop wall | A digger | Buried mesh or a surface apron |
| Damage at latches, doors opened rather than forced | A dexterous mammal such as a raccoon | Two-action latches, not hooks |
| Losses only in daylight, in open ground | Birds of prey | Overhead netting and cover, never lethal control |
A trail camera at the point of entry, left for a week, settles the question. It is the cheapest item on this page and the only one that turns an argument into a fact.
Exclusion: the specification, not the material
Exclusion works on aperture and wire thickness, not on the word written on the packaging. The specification the trade converges on for predator exclusion is half-inch (roughly 13 mm) hardware cloth in 19-gauge wire or heavier, with 16-gauge used at high-pressure points. Note that gauge runs backwards: the lower the number, the thicker the wire.
Standard hexagonal poultry netting, the material most people call chicken wire, is designed to contain birds. Its aperture admits a mustelid’s head and its wire yields to a determined canid. It has a place as a light containment fence and no place as the last barrier between a predator and a flock.
The weak points are predictable and they are almost never the wire itself. Ventilation openings, the junction between two panels, the gap under a door that settled, and the point where mesh is stapled rather than screwed through a batten. Fixings fail before mesh does, which is why washers and battens matter more than an extra gauge of wire. Where a structure is being built rather than patched, these details belong in the design stage, alongside the dimensions and materials covered in our guide to setting up a chicken coop.
Stopping the diggers without trenching the whole perimeter
Digging predators are stopped by geometry, and a trench is only one way to achieve it. Burying mesh 30 cm deep and turning the bottom outwards works, but so does a surface apron: mesh laid flat on the ground, extending 45 to 60 cm outwards from the fence line, pinned down and left to grass over.
The apron exploits how these animals dig. They start at the base of a barrier and work down, so a horizontal skirt puts wire between the nose and the soil at exactly the point the animal chooses to start. On stony or compacted ground where trenching is slow, it is the cheaper of the two and no less effective.
Night closure, because that is when it happens
The majority of losses happen when nobody is present, which makes reliable closure the highest-return item on the list. An automatic pop-hole door on a timer or light sensor removes the human variable, and the human variable is the one that fails on the evening a delivery ran late.
Latches deserve specific attention. A simple hook or a spring-loaded catch is within the capability of a raccoon and of several other dexterous species. Two-action latches, barrel bolts with a pin, or anything requiring a sequence rather than a pull, are the practical standard. Test them by trying to open each one yourself with one hand and no thumbs; anything that yields to that will yield to a paw.
Deterrents: useful, secondary, and prone to habituation
Deterrents change behaviour rather than physically prevent access, which is why they belong after exclusion in the sequence and never instead of it. Motion-activated lights and sound devices work well initially and lose effect as animals habituate, so they earn their keep when they are moved, varied or switched off between episodes rather than left running permanently.
Electric fencing is the exception, because it does not rely on novelty. A correctly earthed wire at the height of an animal’s nose, energised continuously and tested with a meter rather than assumed to be live, teaches avoidance that persists. It also demands maintenance: vegetation growing into the line drains it to the point of uselessness, quietly.
Guardian animals can be effective and are a commitment rather than a purchase. A dog suited to livestock guarding, properly introduced and supervised, deters mammalian predators; an untrained dog in the same yard is itself one of the more common causes of poultry losses. Treat the decision as a husbandry project with its own welfare obligations, not as equipment.
Raptors, and the legal line you cannot cross
Lethal or injurious control of birds of prey is not an option in the EU, and this is not a grey area. Directive 2009/147/EC on the conservation of wild birds requires Member States to protect all wild bird species, and the intentional unlawful killing, destruction, taking or possession of protected specimens constitutes a criminal offence, with Member States required to provide effective and dissuasive penalties. Non-selective methods of capture or killing, including nets, cages and glue, are prohibited outright.
What remains is physical, and it works. Overhead netting on runs, shade cloth or solid cover over part of the range, shrubs and low structures that break the open sightline, and enough covered refuge that birds can reach it in a few seconds from anywhere they are allowed to graze. Raptors hunt open ground; the countermeasure is to have less of it.
The perimeter is also a biosecurity boundary
Spilled feed does two things at once. It attracts the rodents that draw in mustelids and canids, and it attracts wild birds, which is a different and more serious exposure. Wild birds around a poultry unit are a recognised route for avian influenza, so the housekeeping that reduces predation pressure, covered feeders, prompt clean-up of spills, no standing water that attracts waterfowl, is the same housekeeping that reduces disease pressure.
Treating the two as one job is what makes it get done. A perimeter walk that checks mesh, latches, aprons and feed spillage in a single pass takes fifteen minutes and covers both risks at once.
The routine that keeps it working
Every measure described here degrades. Aprons get lifted by frost, staples pull, vegetation shorts the electric line, a door sensor drifts, a latch loosens. The defence is a fixed routine rather than a periodic rebuild.
Daily, confirm the closure happened and count the birds. Weekly, walk the full perimeter, in both directions since you see different things each way, and test the electric line with a meter. Monthly, check fixings and aprons and review any camera footage. After every incident, record what was found and where, because two similar entries three months apart usually reveal a pattern that neither one showed alone.
The other reason wild birds matter
Predation is the visible risk at the perimeter. The one that closes a business is the invisible one arriving on the same wild birds you are already trying to keep off the feed line.
Sources: Directive 2009/147/EC of the European Parliament and of the Council on the conservation of wild birds (codified version); trade practice on hardware cloth specification for predator exclusion. Consulted August 2026. Mesh and fencing specifications are common practice, not a legal standard.
Published previously, fully revised on 6 August 2026. General operational guidance. National and regional rules on protected species, fencing and electric fencing vary; confirm the position that applies to you with your competent authority before acting. Any suspected disease event in the flock is a matter for a qualified veterinarian.