This Hack Get's Mother Nature to Fight Weeds for You: Fortress Plants

Look: this is a picture of a vegetable at war with the lawn. 

Given all the attention around the "food not lawns" movement, you might think I'm being metaphorical. But an experienced gardener will understand that the veggie patch is ALWAYS fighting a war for survival against lawn and "quack grass" greedily trying to take over any bit of land it can get access to. 

"I've lost my garden completely to quack grass!!!"is one of the most common tales of garden woe, and usually the only recourse is to completely start over. 

But LOOK AGAIN: Here, it is the grass that's on the run! This spunky veg, with no help from the gardener, is actually winning the war. More importantly, this hardy specimen is holding the line, protecting the more easily overrun vegetables behind it. This is what some Permaculturists call a "fortress plant." Best of all, it's perennial vegetable, sorrel, meaning that once it is established it will come back each year, working to keep weeds out of the garden for years to come. This is one of several tools derived from the study of ecology and natural succession that clever gardeners can use to keep grasses and other weeds at bay. But you can't just plant sorrel in your garden and expect not to have any weeds. To work well, we need to understand and apply the ecological principle behind how it works and design with that in mind. 

 

(Another combo of fortress plants that has the grass on the run.)

It wasn't hard to convince myself that pouring a glass of home-grown elderberry-wine sangria and plopping a couple ice cubes in it would be the only possible way to visualize this. Now, staring at that ice cube it's still hard for me to imagine a wall of ice a mile high. But 10,000 years ago, that's what we had right here in parts of Michigan. In fact, as recently as 9,000 years ago most - if not all- of Michigan was under ice. When the ice melted and the water cleared what was left was a blank slate, in ecological terms a "disturbance," ready for mother nature to go to work repairing. 

But she couldn't just jump in with the ramps, morels and and solomon's seal that characterize mature woodlands, because they need the rich, deep fungal duff found in the mature forests of Michigan. She had to start with plants that could get a toehold into this "blank slate, with little organic matter or fertility to help out. This gradual process of transformation that occurs after disturbance is called "succession," and that's what we're interested in when it comes to weeds. 

Depending on the soil, the ecological history of the site, and other factors, a typical process of succession starts with the small, quick plants that evolved to cover poor "new" soils with little organic matter or easily digestible nutrition, like mosses and lichens. As these die back they enrich the soil with carbon and other nutrients, essentially adding "compost" to the soil and making it accessible to an increasing diversity of organisms. After a while of this composting, grasses and other "pioneer" weeds come in covering the land and playing their role in repairing the disturbance, creating a grassland. Over time, broadleaf plants like dandelions outcompete the grasses, making room for woody perennials to move in, creating a shrub field, then a savannah, then an open woodland, and finally coming to "climax" at a dark, dense mature forest. 

 

(A forest edge imitates this process, gradually advancing on the grassland.)

So, to nature, "weeds" grasses, and even non-native grasses have a role in ecosystems and a place in the process of ecological succession. In a natural system, their time is fleeting, doing their job, then slowly phasing out of dominance, eventually becoming rare in landscapes they had once dominated as they are replaced by the species that evolved to succeed them. This is why there are rarely many grasses or dandelions in a mature forest, and when they're present they are in balance, not dominating the other plants. 

 

But in a conventional garden we're constantly setting the ecological clock back every time we till or weed the garden, recreating the "disturbance" that is the promised land of the very grasses and pioneer "weeds" we're trying to get rid of! This is a system designed to fail!

From a Permaculture approach, a better way to keep grasses and other "weeds" from taking over the garden is imitate the process of succession. Author Toby Hemenway called this "fast forwarding to a later stage of succession" beyond the stage where these "weeds" necessarily reign. 

As it turns out, this "mid-succession" state, where grasses and early pioneers have started to naturally cede territory to a greater diversity of herbaceous plants, woody shrubs and early pioneer trees, is the ideal natural habitat for most of our favorite food plants. The soil is rich with organic matter and fertility. A variety of sun-shade habitats provides niches for many plants. Most (but not all) of our veggies can easily find a home in such an ecosystem, and most of our fruit trees are mid pioneers, happiest in such a situation. In nature, this situation is found at the forest edge, where the forest is spreading out converting grassland, and in old-fields and savannah systems. 

This is one reason why a well-designed forest garden works so well: it creates a variety of habitats all hovering somewhere around this mid-succession sweet spot. 

 

(An agro-ecosystem that imitates the mid-succession sweet spot.)

And Fortress Plants are just one great hack that food and ornamental gardeners alike can use to tap into this process. 

 The term "Fortress Plant" is not an ecological term, but rather a handy catch phrase for the various mid-succession pioneers that are adapted to reach out into grassland and begin converting it to forest, through competition, shade and allelopathy (chemicals secreted by one plant to poison its competition.)

So one fortress specimen alone will almost certainly be overrun. But if we can plant these perennial pioneers together with enough density to mimic the spreading pattern of succession, we have a chance at making a garden that is fairly resistant to grass and weed pressures, yet provides ideal habitat to our favorite fruits and veggies. 

 

(An edible hedgerow on the march, converting grassland just like a natural forest edge.)

A final caveat is that different soils require different fortress plant communities, so this strategy takes some experimentation. My recommendation is to use high diversity and high density, giving nature the tools she needs to solve the problem on the soils at hand. 

Starter List of Recommended Potential Fortress Plants

Relatively Low Herbaceous Perennials Idea for the Garden Edges

Most woody perennial herbs, depending on soil (better on sandy soil.)

comfrey, pretty much everywhere

sorrel and blood-veined sorrel

turkish rocket

cardoon

most spring ephemeral bulbs including daffodils and alliums (though the effect is stronger at greater densities and with other plants)

over-wintering annual ephemerals like chickweed, dead-nettle, wild mustards and wild cresses

the monardas

fennel (again, this is probably a weeker effect on most sites, but it has good research proving its allelopathic influence.)

Taller Fortress Plants Often Observed at the Edges of a Forest

pokeweed

sunchokes

Maximillian sunflower

lovage

cane fruit

RELATED: The #1 Reason for Weeds in Native Gardens!: 

And this is just one of many tools that gardeners and farmers can use to put the process of ecological succession to work for them. The strategy gets even better when combined with other succession-mimicking hacks, like "nurse logs," another research-based approach to fast-forwarding succession. At Lillie House, these three factors are often seen working together. 

To get a fuller understanding of what these ecological mechanisms are and how they can easily be applied in the garden, we hope you'll join us for this year's Complete Forest Gardening Course, where you can see how these patterns are applied and managed over the course of a growing season, here at lillie House.