Garden Design

Designing around deer is a layout problem before it is a plant problem

Most advice on this subject is a list of plants. Lists are useful and they are not a design. A border made entirely of catmint, lavender and lamb's ear is reliably ignored by deer and reliably dull, and people abandon it within two years for something prettier that gets eaten.

The better framing is that deer are a constraint like shade or drought. Constraints do not stop you designing. They tell you where the good decisions are, and in this case they are mostly about arrangement: what sits where, what screens what, and which few plants are worth protecting rather than replacing.

Where deer actually go, and why it decides your layout

Deer are edge animals. They feed in the open and shelter in cover, so they travel the boundary between the two and browse hardest within a short distance of an escape route. That is why the damage in most gardens is not evenly spread: the bed against the woodland edge is stripped and the one by the back door is untouched.

Three consequences for layout.

Distance from cover is protection. An open lawn between the woodland edge and a border does measurable work, because a deer is reluctant to feed far from somewhere it can vanish into. If you are planning from scratch, that gap is free and worth more than most plant substitutions.

Deer follow lines. Fence lines, hedgerows, ditches, the mown edge of a field. A bed sitting on one of those lines is on a route; the same bed twenty feet off it often is not.

The house is a deterrent. Proximity to doors, lights and movement genuinely reduces browsing. Put the plants you are least willing to lose closest to the building and the tough ones at the far end, rather than the other way round, which is what most people do because the far end is what they see from the window.

The three-zone approach

The most useful thing you can do with a deer-pressured garden is stop treating it as one place.

Zone 1, next to the house. Highest protection for free. This is where the hostas and the roses go if you want them, because it is the part of the garden deer enter last and leave first. It is also the smallest area, so if you end up fencing anything it is cheap to fence.

Zone 2, the middle. Mixed planting, mostly resistant, with occasional middle-tier plants you are willing to gamble on. This is where the design actually happens, and where a repellent programme is realistic because the area is reachable.

Zone 3, the boundary. Fully resistant only, chosen for structure rather than flower, and expected to take damage in a hard winter. Anything you would be upset to lose does not belong here.

That single split does more for a garden than any plant list, because it puts your money where the risk is lowest.

Build the structure from plants deer ignore

The bones of a garden are the things that hold it together in February: hedges, evergreens, repeated shapes, the edge of a path. Those are also the most expensive things to replace, which makes them the wrong place to gamble.

Structure planting should come almost entirely from the top tier. Deer-resistant shrubs covers the options, and the honest short list for structure is boxwood, the junipers, pieris, mahonia and the hollies. Where you want year-round screening, evergreen shrubs narrows it to the ones that hold their leaves, with the caveat that holding foliage through winter is exactly what makes a plant valuable to a deer in February.

What not to use for structure, whatever the garden centre suggests: arborvitae, which is the default privacy hedge across most of the country and one of the worst-rated plants on this site; and rhododendron, which is the default evergreen shrub and no better.

Then colour, which is where the difficulty is

The reliable deer-resistant palette skews hard towards blue, purple, silver and white, because that is what the aromatic mint-family plants offer. Catmint, lavender, russian sage, salvia and artemisia are all in that range, and a border built only from them looks cold and slightly samey by August.

Three ways out of it, in order of how well they work.

Bulbs for the early colour. Daffodils and alliums are genuinely ignored and carry the garden from March to May, which is the hungriest window for deer and the emptiest one for the border. Never tulips, which are close to bait.

Annuals for the hot colours. The reds, oranges and clear yellows the perennial palette struggles with come most easily from annuals, and being annual is the point: a bad year costs you a tray of plants rather than a decade.

Foliage rather than flower. Silver, chartreuse and deep green do more work in a mixed border than most people credit, and the felted silver plants deer avoid happen to be the best contrast partners for everything else.

Contrast and repetition do the heavy lifting

Two design moves matter more than plant choice once the palette is set.

Contrast of form. Spikes against domes against flat plates. Russian sage and salvia give you vertical, catmint and lamb's ear give you mounds, yarrow and sedum give you horizontal plates. Those three shapes in one bed read as designed even when the colour range is narrow, which is exactly the problem a deer-resistant palette creates.

Repetition. Using the same plant three or five times down a border, rather than one of everything, is what makes a restricted palette look deliberate instead of impoverished. It also happens to be cheaper.

There is more on both in the design principles, and four worked layouts in the free plans, including a mailbox garden for the hardest site most properties have.

Decide what you are protecting rather than protecting everything

Almost nobody needs to fence a property. Perimeter grows much faster than area: a square quarter-acre is about 420 feet of fence, a square acre about 835, so you buy double the fence for four times the land.

The sensible sequence is to cage the specific plants being destroyed, then fence the one bed that matters, and only then consider the boundary. What a deer fence costs has a shopping list you can price yourself, and if you do fence, eight feet is the working height and six is the most common expensive mistake in the subject.

Designing this way has a pleasant side effect. A small, well-fenced bed near the house lets you grow the things you actually love, which means the rest of the garden does not have to carry that burden and can be planted entirely for toughness.

What changes the answer

Winter. Late winter is when a resistant planting gets tested, because the easy food is gone. A garden that looked bulletproof for three years can be browsed hard in February of a bad one.

New plantings. Nursery stock is fed, irrigated and soft, and is more palatable than the same species after two years in your ground. Plant in autumn where you can, and do not over-feed a resistant border: rich soil produces exactly the sappy growth that undoes an aromatic plant's defence.

Which deer you have. These ratings are white-tailed deer, observed in New Jersey. Mule deer browse differently, and the axis deer in Hawaii and parts of Texas are not a North American animal at all. The methodology page says where the scale stops applying.

Scale, and the mistake of planting too small

The most common failure in a deer-pressured garden is not the wrong plant, it is too few of the right one.

A single catmint in a bed reads as an accident. Five reads as a decision, and five is also enough that losing one to a hard winter does not open a hole. In a garden where a proportion of the planting will take damage in a bad year, planting in drifts is not just an aesthetic preference, it is redundancy.

The same applies to hedging. A mixed hedge of three or four species survives a bad browsing year far better than a monoculture, because local deer develop local habits and rarely hit everything equally. It also looks better, which is the rare case where the resilient answer and the handsome one agree.

Buy fewer species and more of each. It is cheaper, it looks more deliberate, and it fails more gracefully.

Planning across the seasons

A deer-pressured garden has two difficult moments and they are not the ones people plan for.

Late winter is when the planting is tested. Browse is gone, evergreens are the only green thing within reach, and a plant untouched for three years can be stripped in February. Anything you are relying on for winter structure should come from the top tier, not the middle.

Spring emergence is the second. New growth is softer, sweeter and lower in the defensive compounds the mature plant will have, so a perennial ignored in July can be eaten to the ground in April. This is also when nursery stock goes in, at its most palatable, which is why so many people conclude a plant is not deer-resistant in its first season.

Summer is the easy season and the one most gardens are designed in. Planning in June for a garden that has to survive February is how most of these borders go wrong.

Where to start

If the garden already exists, start by mapping the damage rather than replanting. Which beds get hit, which do not, and where the deer are entering from. That map usually shows the three-zone split already exists in your garden whether you planned it or not.

If you are starting from nothing, put the protected bed near the house, run the tough planting along the boundary, and leave the open gap between the woodland edge and anything you care about.

Then choose plants: shrubs for the bones, perennials for the middle, bulbs for the spring, and what to avoid for the list of things that will be eaten no matter how well the bed is placed.