Stop Planting Individuals. Start Planting Communities.

Most planting plans, if you're honest about them, are catalogues. A designer walks a site, imagines a mood board, selects twenty species they love, and distributes them across the plan with a vague sense of height and colour. The result looks considered on paper and acceptable in year one. By year three, the gaps have filled with Epilobium hirsutum or couch grass, the more vigorous plants have swallowed their neighbours, and the client is ringing to ask why it all looks so exhausted. The answer is that it was never designed to work. It was designed to look like it would work.

The shift from planting individuals to designing communities is not a stylistic preference. It's a structural one. It changes how you think about space, time, root zones, canopy, and ultimately about what a garden is for. And I'll say this plainly: most practising designers — including good ones — are still operating in catalogue mode without knowing it.

Competition Is the Point, Not the Problem

The first mistake is treating plant competition as something to be managed away. Space plants generously, the thinking goes, give each one room to breathe, and nobody will bother anyone else. The result is bare soil between crowns, and bare soil is an invitation. Something will colonise it. The question is whether you chose what that something would be.

In a functioning plant community, competition is the mechanism that holds the system together. Roots compete for water and nutrients, and that competition — counterintuitive as it sounds — actually suppresses the most thuggish opportunists, because the ecological niches are already occupied. A dense planting of Geranium × oxonianum at the feet of shrubs isn't just pretty. It is, root by root, making it harder for annual weeds to establish. The geranium doesn't know it's doing you a favour. It's simply doing what plants do when they have neighbours: it fills space.

This is where spacing guidance gets genuinely complicated, and I won't pretend there's a clean formula. For a groundcover-dominant community, the kind where you're using a low-growing species as living mulch beneath a shrub layer, I tend to work at densities that feel almost aggressive: seven to nine plants per square metre for something like Pachysandra terminalis or Ajuga reptans, and closer still on slopes where erosion is a concern. The instinct to space more generously is an instinct trained by nursery catalogues showing plants at their eventual size, not at the density a community approach requires. Resist it.

For a mixed herbaceous community with no strong groundcover component, the calculus shifts. Here I'm thinking about root architecture as much as canopy spread — pairing shallow-rooted spreaders like Alchemilla mollis with deep-taproot species like Echinops ritro or Eryngium × zabelii so the plants are genuinely working different layers of the soil, not competing for the same resources at the same depth.