Fundamentals of Ecosystem Restoration: Practice
Sobre esta edición / About this edition: This is the complete written text of Fundamentals of Ecosystem Restoration. The photographs, diagrams, and field figures from the printed edition are not reproduced here. If the illustrations would help your learning, or if you would like to support this work, the fully illustrated manual is available at planethealers.org.
Section III: What Is Ecosystem Restoration?
Section IV: Ecological Restoration Practices
SECTION III: WHAT IS ECOSYSTEM RESTORATION?
Chapter 45: What Is Ecosystem Restoration?
We spent the first half of this book learning to read ecosystems — species, interactions, energy flow, succession. All of that was groundwork. Now we put it to use.
Ecosystem restoration is the process of halting or reversing the degradation of an ecosystem so that its biodiversity recovers and its services improve. That definition comes from the United Nations, which declared 2021–2030 the Decade on Ecosystem Restoration. The definition matters because it sets the bar: you are not restoring an ecosystem unless biodiversity is actually increasing and the system is functioning better than when you started.
[Figure in the printed edition: degraded landscape transitioning to restored habitat — before and after comparison]
There is a distinction worth understanding early. Ecosystem restoration is the broader umbrella — it covers everything from greening a city block to returning a clearcut watershed to closed-canopy forest. Ecological restoration is a specific discipline within that umbrella, focused on assisting the recovery of ecosystems that have been degraded, damaged, or destroyed. I would have told you a few years ago that the two terms were interchangeable. Then the UN Decade came along and the political language shifted. Ecosystem restoration became the big-tent term — inclusive of urban greening, agricultural improvement, remediation of contaminated sites. Ecological restoration stayed closer to its scientific roots: returning a damaged ecosystem to a self-sustaining state that resembles the original.
For the work we do in this course, you need both. A contaminated industrial site in Hamilton needs remediation before anything else can happen. A degraded pasture in the Yucatán needs ecological restoration — removing cattle, planting native species, letting succession do its work. A stormwater pond in Mississauga needs rehabilitation to function as actual habitat instead of just a concrete basin. These are all points along what we call the ecosystem restoration continuum.
The continuum is the single most useful concept in this field. Picture a spectrum. On one end sits the most degraded condition imaginable — a paved parking lot, a strip mine, a monoculture cornfield drenched in glyphosate. On the other end sits a fully recovered, self-sustaining ecosystem with its original biodiversity intact. Everything we do in restoration is about moving a site from wherever it sits on that spectrum toward the recovered end.
The continuum breaks into stages. Reducing impacts comes first — stop doing the thing that is causing the damage. Then remediation — clean up contamination, remove debris, deal with the legacy of what happened. Then rehabilitation — get basic ecological functions working again. Then ecological restoration — assist the ecosystem toward full recovery. And the target you are aiming for, always, is the reference ecosystem. That is the benchmark. The intact or least-disturbed version of what your site should look like if it were healthy.
Not every project reaches full recovery. A hydro corridor in the middle of Toronto is never going to become old-growth forest. But it can become a functioning prairie meadow ecosystem with native grasses, wildflowers, deep root systems, and pollinators — and that is a massive improvement over mown turf. The continuum gives you permission to work at whatever scale and intensity the site allows, while keeping your eyes on the highest achievable outcome.
The other thing the continuum does is connect your work to the bigger picture. When you restore 200 hectares of meadow in an urban corridor, you are not just improving one strip of land. You are connecting river systems, providing movement corridors for wildlife, sequestering carbon in deep-rooted perennial grasses, reducing the urban heat island, and filtering stormwater. Every project along the continuum contributes to something larger than itself.
That is the frame for everything that follows.
Chapter 46: The Ten Principles of Ecosystem Restoration
The UN Decade gave us ten principles for ecosystem restoration. They are not abstract policy statements — they are practical guides for how to plan and run a project. I am going to walk through each one using a real project as the case study: the Gatineau Hydro Corridor restoration in Toronto, which I led for two years.
[Figure in the printed edition: urban hydro corridor showing mown turf transitioning to native meadow planting]
Principle 1: Ecosystem restoration contributes to the UN Sustainable Development Goals and to national and regional goals.
This sounds bureaucratic until you need to write a grant application or get a municipal permit. When we started the corridor project, we had to demonstrate that it aligned with the city's biodiversity strategy, the province's pollinator health plan, and broader climate targets. That alignment is what opens doors. If your project can show it contributes to recognised goals at multiple levels, you get access to funding, partnerships, and political support that isolated projects never see.
Principle 2: Ecosystem restoration promotes inclusive and participatory governance.
The corridor crossed 13 neighbourhoods, ran past schools and universities and industrial parks. We could not just show up and start ripping out turf. Every community along that 16-kilometre stretch needed to understand what was happening and why. Schools got involved in planting days. Corporate offices along the corridor contributed to clean-up events. The bike trail component brought in the cycling community. Governance means everyone who is affected has a voice in the process — not as a box-ticking exercise, but because their buy-in determines whether the project survives after you leave.
Principle 3: Ecosystem restoration includes a continuum of activities and must produce a net benefit.
You cannot call something restoration if it does not actually improve conditions. This principle forces you to define what "better" looks like before you start. For us, better meant measurable increases in plant biodiversity, pollinator abundance, carbon sequestration, and water infiltration — along with reductions in mowing frequency, fuel use, and heat island effect. If your project cannot demonstrate a net benefit, go back to the planning stage.
Principle 4: Ecosystem restoration aims for the highest level of recovery.
You always push toward the best achievable outcome. Under those power lines, old-growth forest was off the table. But a fully functioning tallgrass prairie ecosystem — with 80+ native species, deep root systems, and a self-sustaining disturbance cycle? That was achievable. We identified two reference prairie ecosystems near Toronto and used them as our targets. You set your sights high and work toward it. If constraints limit you, fine — but you do not settle for less than the site can support.
Principle 5: Ecosystem restoration addresses both the direct and indirect causes of ecosystem degradation.
The direct cause was simple: regular mowing. Six times a year, 200 hectares of potential habitat were being flattened into lawn because that was what a "good neighbour" looked like. The indirect causes were deeper — decades of a landscaping culture that equated short grass with responsible land management. Addressing both meant changing the physical management (stop mowing every few weeks, shift to a four-to-six-year burn/mow cycle) and the cultural expectations (teach communities that tall native plants are not "weeds" but functional habitat).
[Figure in the printed edition: before-and-after of mown hydro corridor turf vs established native meadow with wildflowers]
Principle 6: Ecosystem restoration integrates all types of knowledge.
Two-eyed seeing. Scientific data on one side — species lists, soil chemistry, hydrology. Indigenous and local knowledge on the other — which species belong here, how fire was used historically, what the land looked like before European settlement. In Ontario, Indigenous peoples used controlled burns to maintain prairie openings, which attracted game. Understanding that history changed our approach to managing the restored meadows. You incorporate every knowledge system that can inform the work.
Principle 7: Ecosystem restoration is based on well-defined, short, medium, and long-term ecological, social, and economic goals and objectives.
Before we planted a single seed, we had goals at every scale. Short-term: establish 30 native species in the first growing season. Medium-term: self-sustaining meadow ecosystem within five years. Long-term: functioning wildlife corridor connecting seven river and ravine systems across the city. Without those targets, every decision becomes an argument. With them, you always have something to come back to — is this action moving us toward the goal?
Principle 8: Ecosystem restoration is tailored to the local ecological, cultural, and socio-economic context, as part of the larger landscape.
Our corridor crossed seven rivers and ravine systems, linked 15 existing parks, and affected the movement of over 1,000 species of plants and animals. You cannot plan a project like that by looking at your site in isolation. You have to understand the watershed — where is the water coming from, what is it carrying, where are the forest fragments that will supply seed and wildlife to your site? In the Yucatán, the same applies. Tsunul sits in a landscape of cattle ranches and milpa agriculture. Our restoration works because there are enough forest fragments nearby to provide seed sources and wildlife corridors. If those fragments disappeared, our job would be ten times harder.
Principle 9: Ecosystem restoration includes monitoring, evaluation, and adaptive management.
We set up permanent monitoring plots along the entire corridor — quadrat sampling with a randomised block design so we could track changes in species composition, abundance, and soil health over time. We tracked flowering phenology for every species we found, noting when each one flowered and set seed. That data fed directly into our management decisions. When we saw dog-strangling vine flowering in a section, we knew exactly how much time we had before seed set — and we controlled it in that window.
The phenology tracking served a second purpose: seed collection. When you know that wild bergamot flowers in July and sets seed in August, and that prairie smoke flowers earlier and sets seed in late June, you can schedule collection efficiently. You build these associations over years of observation. They are the foundation of practical restoration work.
Principle 10: Ecosystem restoration is enabled by policies and measures that encourage long-term progress, replication, and scaling up.
This one is about legacy. I created seed mix guidelines for the entire Greater Toronto Area — 6.3 million people. A dozen different mixes for different site conditions: erosion control, site recovery, various soil types and moisture regimes. Those guidelines became policy. Every new development in the GTA now has to use native seed mixes. Every damaged site gets restored with appropriate species rather than whatever the contractor had in the truck.
[Figure in the printed edition: native seed mix being applied to a restoration site, or close-up of diverse native seed collection]
We also documented everything we did across two years into a manual that could be handed to the next project. The cover crop and nurse crop protocols, the seeding rates, the monitoring methods, the community engagement process — all of it. When the next organisation wants to restore a hydro corridor or a right-of-way, they do not start from zero. That is how you scale restoration: not by doing every project yourself, but by making sure the knowledge transfers.
Chapter 47: Measuring Success and Community Engagement
A restoration project that nobody monitors is a project that nobody learns from. And a project that the surrounding community does not understand is a project that gets mowed down the first time a new parks manager takes over.
These two things — measuring what is happening and keeping people connected to it — are not extras. They are as much a part of restoration as planting native species.
[Figure in the printed edition: monitoring plot with quadrat frame in native meadow, or volunteers collecting data]
Monitoring That Means Something
When I set up monitoring for the Toronto corridor project, we used permanent plots marked with fibreglass poles — non-conductive, because we were working in a hydro corridor. Each plot had four corners staked and oriented to a compass direction so we could relocate them year after year. Inside each plot, we recorded every plant species present, estimated its cover, noted whether it was flowering or in seed, and assessed soil conditions.
We also ran transect surveys for butterflies, birds, and other invertebrates. Pick a marked starting point, walk 50 metres along a compass bearing, record everything you see. Do it again the following year along the same line. Over time, you build a picture of how the community is changing.
The monitoring gave us three things. First, a baseline — what the site looked like before we started. Second, a measure of progress — are native species increasing? Are invasive species declining? Is soil organic matter building? Third, and maybe most important, the information we needed for adaptive management. When monitoring showed dog-strangling vine spreading in a particular section, we could respond in the current season rather than discovering the problem a year later.
Phenology charts were one of the most useful tools we developed. My field crew recorded every species they identified — when it first appeared, when it flowered, when it set seed. Over a growing season, those charts gave us a visual timeline of the entire community. You could see at a glance that several invasive species set seed in the same two-week window, which meant you could time a single management intervention to catch them all.
Those charts also showed us associations we would have missed otherwise. When wild strawberry was flowering in May, we knew that several other early-season species were at the same stage. When we spotted one, we could go looking for the others. That kind of pattern recognition is what separates restoration from gardening — you are learning to read the whole community, not just individual plants.
Getting the Community In
The corridor ran through 13 neighbourhoods. There were schools along the entire length. Corporate offices. Residential streets where people had been looking at mowed turf their whole lives and understood that as "maintained."
We planted with school groups. Kids got their hands in the soil, put in plugs, came back months later and found their plants blooming. That kind of direct experience changes how a young person sees the landscape permanently. They are not reading about biodiversity in a textbook — they are watching it happen in the field behind their school.
[Figure in the printed edition: school group or community volunteers planting native species in an urban corridor]
We organised clean-up days with neighbourhood associations. Garbage dumping was a constant issue — people see tall vegetation and assume it is abandoned. Getting the community physically present in the space, picking up garbage, seeing the butterflies on the milkweed, understanding that this is managed habitat and not neglect — that changes the relationship.
Corporate offices along the corridor got involved too. It gave their employees something tangible to connect with. Not a logo on a poster. Actual soil under their fingernails. That kind of engagement builds the social licence that keeps a project alive long after the restoration ecologist has moved on.
The single biggest threat to urban restoration is not invasive species. It is a change in management. A new parks manager comes in, does not know the history of the project, sees tall grass where there used to be lawn, and orders it mowed. It has happened. The defence against that is documentation and community connection. If the manual exists and the community knows what the site is, they become its advocates. They push back when someone proposes mowing it.
That is why community engagement is not a feel-good add-on. It is a survival strategy for the ecosystem you just spent years building.
Chapter 48: The Restoration Continuum
I introduced the continuum in Chapter 45. Now we need to go deeper, because this is the structure you will use to plan every restoration project you ever work on.
[Figure in the printed edition: diagram or illustration of the ecosystem restoration continuum from degraded to fully recovered]
The continuum runs from the most degraded condition to a fully recovered, self-sustaining ecosystem. Between those endpoints sit a series of stages, each with its own set of tools and approaches. Where your site falls on the continuum determines what you do first.
Starting Point: How Degraded Is It?
A paved parking lot is at one extreme. The soil is sealed under asphalt, the hydrology is completely altered, there is zero biological community. An abandoned agricultural field is somewhere in the middle — the soil structure is damaged, the seed bank is depleted of native species and loaded with weeds, but some biological function remains. A selectively logged forest with its canopy thinned is closer to the recovered end — the structure is compromised but the community is mostly intact.
Your first job is to locate the site honestly on the continuum. Not where you wish it were. Where it actually is. That assessment drives every decision that follows.
Stage 1: Reducing Impacts
Before you can move a site toward recovery, you have to stop the thing that is degrading it. If it is an agricultural field being sprayed with herbicides, the spraying stops. If it is a stream receiving untreated stormwater runoff, you address the runoff. If it is a forest being grazed by cattle, you remove the cattle.
This sounds obvious but it is where many projects fail. People want to jump straight to planting trees without addressing the underlying cause of degradation. At Tsunul, the first thing we did was fence the cattle out. Nothing else would have mattered if we had not done that. The land could not begin recovering while it was still being grazed to bare rock.
You also need to think beyond the local. What is happening upstream? What is happening on adjacent properties? If your site is downhill from an industrial operation dumping contaminated runoff, your restoration will struggle until that input is addressed. Reducing impacts means reducing all the impacts — on-site and off-site, direct and indirect.
Stage 2: Remediation
Some sites carry the legacy of what was done to them in the form of contamination. Old gas stations with hydrocarbon-soaked soil. Former industrial sites with heavy metals. Mining operations with acid drainage. These sites need remediation — the contamination has to be dealt with before biological recovery can begin.
Remediation techniques range from simple to complex. Soil removal and replacement. Capping contaminated material with clean soil. Phytoremediation — using plants that accumulate specific contaminants in their tissues. Bioremediation — using microorganisms to break down pollutants. The right approach depends on the type and extent of contamination, the intended future use of the site, and the budget.
Remediation does not restore an ecosystem. It creates the conditions under which restoration can begin. It is the clearing of the obstacle, not the recovery itself.
Stage 3: Rehabilitation
Rehabilitation means getting basic ecological functions running again on a site that has been heavily altered. The site may never return to its original state, but it can support more biodiversity and provide more ecosystem services than it does currently.
The Leslie Street Spit in Toronto is the best example I know. We will cover it in detail in the next chapter, but the short version: a five-kilometre spit of land built entirely from construction rubble and dredged material, sticking out into Lake Ontario. Not a natural landform. Never was. But over decades, with deliberate intervention and a lot of natural succession, it has become one of the most important wildlife habitats in the Greater Toronto Area. That is rehabilitation — taking something that was never a functioning ecosystem and making it one.
Stage 4: Ecological Restoration
This is where you are actively assisting a degraded ecosystem toward full recovery. You have a reference ecosystem as your target. You are planting native species, managing invasive species, restoring hydrology, rebuilding soil biology. The goal is a self-sustaining system that does not need you anymore.
At Tsunul, this is the stage we have been working in for over a decade. The cattle are gone, the land is fenced, and we are assisting the recovery of tropical dry forest on 39 hectares of former pasture. We plant native species from the surrounding forest fragments, manage invasive grasses, and let succession do most of the heavy lifting. The site now holds 514+ documented species and a closed canopy in many areas. The forest is building itself. We are its assistants.
The Reference Ecosystem
Everything on the continuum points toward one thing: the reference ecosystem. This is the intact or least-disturbed version of what your site should become. It is your target, your benchmark, your measure of success.
A good reference ecosystem sits nearby, shares similar soil, hydrology, and climate conditions with your site, and represents the ecological community that would occupy your site in the absence of degradation. You use it to generate species lists, understand community structure, set targets for canopy cover and species richness, and evaluate your progress.
We will spend an entire chapter on reference ecosystems later. For now, understand that the continuum without a reference is a road without a destination.
Chapter 49: Reducing Impacts and Risk Assessment
Every restoration project starts with the same question: what is causing the damage, and how do we stop it?
This is not the exciting part of restoration. Nobody photographs the moment you turn off an irrigation pump or fence out livestock. But it is the most consequential step. Nothing you do afterward matters if the source of degradation is still active.
[Figure in the printed edition: fence line contrast showing grazed vs ungrazed land, or industrial runoff entering a waterway]
Risk Assessment
Before you decide what to address first, you need to understand the full picture of threats to your site. This is called a risk assessment, and it works by evaluating two dimensions for each potential threat: severity and likelihood.
Severity asks how bad the impact would be if it occurred. A chemical spill into a creek is high severity — it could wipe out the aquatic community in hours. Soil compaction from occasional foot traffic is lower severity — it degrades conditions slowly and can be reversed.
Likelihood asks how probable the impact is. If your site sits downhill from an active quarry, the likelihood of dust, runoff, and vibration is high. If the quarry closed twenty years ago, the likelihood of ongoing impacts is low — though the legacy impacts may still be severe.
Map every potential impact onto these two axes and you get a prioritised list. High severity, high likelihood threats get addressed first. Low severity, low likelihood threats can wait or may resolve on their own as the ecosystem recovers.
Local and Global Impacts
You need to think at two scales. Locally, what is directly degrading this site? Mowing, grazing, dumping, chemical application, altered hydrology, invasive species — these are the immediate threats you can often address directly.
Globally, what systemic pressures affect the ecosystem? Climate change is shifting temperature and precipitation patterns. Nitrogen deposition from agricultural and industrial sources is altering soil chemistry across entire regions. These you cannot fix at the site level, but you need to account for them in your planning. If your region is trending hotter and drier, the species mix you plant should reflect that trajectory, not the historical conditions that no longer apply.
On-Site Impact Reduction
The practical work of reducing impacts looks different on every site. At Tsunul, it meant fencing — 39 hectares of perimeter fence to keep cattle from neighboring ranches off the recovering land. In the Toronto corridor, it meant convincing a utility company to change their mowing regime from six times per year to once every four to six years. On a contaminated brownfield, it might mean capping exposed soil or diverting surface water away from a contamination plume.
The common thread is this: identify the active agents of degradation, and either remove them or reduce them to a level the ecosystem can tolerate. Do that first. Then the recovery work can begin.
Off-Site Impact Reduction
Your site does not exist in isolation. Water flows downhill. Wind carries seeds and pollutants. Wildlife moves across property boundaries.
If the farm upstream is dumping nutrient-laden runoff into the creek that feeds your wetland restoration, your project has a problem that no amount of planting will solve. If the neighbouring property is a source of invasive species seed, your invasive management costs will never drop to zero.
Off-site impact reduction is harder because you often do not control the land. It requires negotiation, partnership, sometimes policy change. But ignoring it guarantees that your on-site work will underperform. The best restoration project in the world cannot overcome a constant external supply of the thing that degraded the site in the first place.
Chapter 50: Remediation
Some sites are too damaged for ecological processes to restart on their own. The soil is contaminated. The water is toxic. The substrate is asphalt or concrete or mining tailings. Before biology can get a foothold, the contamination has to be dealt with.
That is what remediation does. It is not restoration — it is the step that makes restoration possible.
[Figure in the printed edition: brownfield or contaminated site being remediated — soil removal, phytoremediation planting, or capping]
Brownfield Sites
Brownfields are abandoned or underused properties where the previous industrial or commercial use left contamination behind. Old gas stations with petroleum hydrocarbons in the soil. Former dry cleaners with chlorinated solvents. Smelter sites with heavy metals. Paint factories, rail yards, chemical plants — the industrial history of every city created a landscape of contaminated sites that nobody wanted to deal with.
Remediating a brownfield starts with a site investigation. What contaminants are present? How deep do they go? Are they migrating — moving through groundwater, volatilising into the air, leaching into adjacent soils? The investigation determines everything that follows.
Remediation Approaches
The simplest approach is dig and dump — excavate the contaminated soil and haul it to a licensed disposal facility. It is fast and definitive, but expensive and it just moves the problem somewhere else. It is appropriate when contamination is localised and severe.
Capping involves covering the contaminated material with a barrier — typically a layer of clean soil, sometimes with a geomembrane or clay liner underneath. The contamination stays in place but is isolated from the surface ecosystem. Many urban parks sit on capped former landfills. The trees are growing in clean imported soil above whatever lies buried underneath.
Phytoremediation uses specific plant species to extract contaminants from the soil. Certain species of willow and poplar accumulate heavy metals in their tissues. Indian mustard (Brassica juncea) is a hyperaccumulator of lead. Sunflowers can extract uranium from soil. The plants are harvested and disposed of as contaminated material, but over multiple growing seasons they can significantly reduce contaminant concentrations.
Bioremediation uses microorganisms — bacteria and fungi — to break down organic contaminants like petroleum hydrocarbons. Some species of fungi can degrade PCBs and other persistent organic pollutants. You can sometimes enhance bioremediation by adding nutrients or adjusting moisture and oxygen levels to favour the microbial communities that do the work.
The Limits of Remediation
Remediation reduces contamination. It does not create an ecosystem. A remediated site still has degraded soil structure, depleted organic matter, an absent seed bank, and no biological community to speak of. It is a blank canvas — cleaner than it was, but still far from functional.
That is where the continuum picks up. Remediation gets the site to a point where rehabilitation or restoration can begin. The two steps are sequential, not interchangeable. Skipping remediation on a contaminated site and jumping straight to planting is a recipe for dead plants and wasted money.
Chapter 51: Rehabilitation — Leslie Street Spit
If you want to understand what rehabilitation looks like at its most dramatic, go to the Leslie Street Spit in Toronto.
[Figure in the printed edition: aerial view of Leslie Street Spit extending into Lake Ontario]
Five kilometres of land extending into Lake Ontario. Not a natural formation — every grain of it was placed there by humans. Starting in the 1950s, the city began dumping construction rubble, dredged sediment from the shipping channel, and excavation spoils from building sites onto an existing sandbar. The plan was to build a massive harbour facility. The harbour never materialised. What happened instead was ecology.
How Rubble Becomes Habitat
While the engineers debated the spit's future, nature moved in. Seeds arrived on wind and water. Birds landed and deposited seeds in their droppings. The concrete rubble provided shelter and structure — crevices for nesting, elevation changes that created microclimates. Organic matter accumulated in pockets between the debris.
Today, the Leslie Street Spit is home to one of the largest ring-billed gull colonies in the Great Lakes, plus colonies of common terns, black-crowned night herons, and double-crested cormorants. It supports breeding populations of species that have disappeared from most of the urbanised lakeshore. It is one of the most biodiverse sites in the GTA.
None of this was planned. The birds found it on their own. But what happened next was deliberate.
Wetland Cells
Along the outer edge of the spit, concrete rubble was arranged into barriers that deflected wave action. Inside those barriers, dredged material from the shipping channel was deposited to create shallow water zones. These became wetland cells — artificial in origin, but providing the conditions that wetland species need.
The cells developed the way any new wetland does, just on an accelerated timeline. Sediment accumulated. Aquatic plants colonised the shallows. Invertebrates moved in, followed by fish and amphibians. The constructed barriers provided wave protection that allowed cattails and bulrushes to establish along the edges. Within a few years, the cells were functionally indistinguishable from natural lakeside wetlands.
[Figure in the printed edition: wetland cell at Leslie Spit with cattails, open water, and waterbirds]
In the shallower areas — the bays created by the rubble breakwaters — you could see ducks feeding on aquatic invertebrates. The elevation changes created by the irregular dumping of rubble meant that you had deep zones and shallow zones right next to each other, which is exactly the kind of structural diversity that wetland organisms need.
The Shoreline Effect
Along the edges where wave action eroded through the rubble, something interesting happened. You could see the layers — rock and concrete at the base, then a developing organic layer on top. Woody debris from the lake washed up and accumulated in drift lines. That debris breaks down over time and creates soil. Given enough time and enough wave-deposited organic material, those rubble shorelines start looking and functioning like natural beach berms.
You could go in at that stage and accelerate the process — plant beach grass and other shoreline species to stabilise the developing substrate. Not planting a finished ecosystem, just giving the natural process a push at the right moment.
Beaver Engineering
Balsam poplars established on parts of the spit early in the process. They send out root suckers freely — one tree becomes a grove over a few years. Then the beavers arrived.
A beaver does not see a poplar grove the way we do. It sees building material and food. It cuts the main stem, and the tree responds by sending up multiple suckers from the base. Each of those gets cut in turn, and each one resprouts. The beaver is doing exactly what a restoration ecologist would do if they wanted to create a dense thicket of young woody growth — repeated coppicing.
The result is a low, dense poplar stand with far more stems per square metre than would occur naturally. That density provides cover for nesting birds, shade for the forest floor, and structure that catches wind-blown seeds and organic material. The beaver is an ecosystem engineer here in the fullest sense — it is physically restructuring the plant community in ways that benefit other species.
Managing the Challenges
The spit is not self-maintaining. Without active management, Phragmites would have overtaken most of the wetland habitat — it was spreading aggressively through the cells before the conservation team intervened with targeted control. When they removed the Phragmites, the seed bank of native species they had planted years earlier was still viable in the soil. The wetland rebounded. That is the value of planting a diverse species mix at the start — even if an invasive species temporarily dominates, the native community is waiting underground for its chance.
They also installed perch poles — vertical stakes placed around the wetland cells for birds to land on. Birds perch, eat seeds from surrounding vegetation, defecate, and deposit seeds of species that were never deliberately planted. It is free seed dispersal, courtesy of the local bird population. Around every perch pole, you can see a ring of vegetation that arrived that way.
[Figure in the printed edition: perch poles in Leslie Spit wetland cell with bird perched, or close-up of vegetation around pole base]
The Lesson
The Leslie Street Spit was never a natural ecosystem. It was built from the waste stream of a construction boom. But given time, protection from intensive human use, and some strategic interventions — the wetland cells, the wave barriers, the decision to leave most of the site undeveloped — it became one of the most ecologically productive sites in a city of six million people.
That is rehabilitation. You are not returning the site to a historical condition, because there is no historical condition to return to. You are creating the conditions under which ecological processes can operate, and then stepping back and letting them run. The spit proves that even on the most artificial substrate, given the right conditions, life will find its way in and build something worth protecting.
Chapter 52: Stream and Aquatic Restoration
Water reshapes a landscape faster than any other force. When we alter how water moves — paving over soil, straightening streams, building stormwater ponds — we change everything downstream. Restoring aquatic systems means understanding what we broke and working with water's own physics to fix it.
[Figure in the printed edition: restored stream channel with natural meander, or stream daylighting project]
Stormwater Ponds
Most urban stormwater ponds were never designed as ecosystems. They were designed as engineering structures — catch the runoff, hold it long enough to settle out sediment, release it slowly. The water comes in warm from sun-heated pavement, loaded with road salt and hydrocarbons and lawn fertiliser. It sits in a shallow basin that heats up further in summer, then gets released into a receiving stream.
The problem is that many receiving streams are cold-water systems. Brook trout, for example, need water temperatures below about 20°C. A stormwater pond that releases 25°C water in July is functionally poisoning the trout habitat downstream.
The fix is in the outflow design. Two types of water control structures exist. A top draw is a weir that skims water off the surface — the warmest layer of the pond. That warm water goes straight into the creek. A bottom draw pulls water from the deepest part of the pond, where it is coolest due to thermal stratification. The difference in temperature between the two can be several degrees, which is the difference between a viable trout stream and a dead one.
If you are working on a stormwater project, the first question is: what kind of draw does this pond have? If it is top draw releasing into a cold-water stream, that is a problem you can potentially fix by retrofitting the outflow structure. It is a relatively simple engineering change that can shift the ecological condition of everything downstream.
Planting a Stormwater Pond
If you are planting a stormwater pond, you need to understand the water level fluctuations before you place a single plant. Look at the outlet structure. Look at the line around the edge of the pond where the water interacts with the shoreline — that shows you the typical water level and how much it moves.
Your planting zones work from the edge inward. Above the high-water line, wet meadow species — sedges, grasses, wildflowers. From the water's edge down to about a metre deep, emergent species — cattails, bulrushes, blue flag iris. Past a metre, submerged and floating-leaved species — water lilies, pondweeds — down to two or three metres depending on water clarity.
One thing nobody warns you about: geese. Canada geese love freshly planted ponds. They will follow your planting crew and pull up every plug you install. Solutions I have used: fishing line strung between posts across the pond surface — geese cannot see it, hit it when they try to land, and avoid the area. Bird-scare flash tape — a ribbon with a red and silver side that twists in the wind and mimics fire. Snow fencing along the pond edge to block the easy transition between water and grass that geese need. In Hamilton, Cootes Paradise had such a severe carp problem that they installed barriers in the water to let vegetation re-establish in protected zones before removing the barriers and moving on.
[Figure in the printed edition: newly planted stormwater pond with protective fencing or bird deterrents]
And a note on trees around ponds: everybody wants instant forest. They plant large ball-and-burlap trees — three or four metres tall, roots chopped off at the edge of the root ball. Those trees sit in stasis for five years while the root system regrows. A one-litre container tree, a metre and a half tall with its root system intact, catches up to the big tree within that same five-year window. I have seen it repeatedly. Plant small. Plant with intact roots. The trees do the rest.
Stream Daylighting
Across Ontario and most of urban North America, we spent the last century burying streams. They were put into culverts, encased in concrete, routed underground through storm sewers. An entire generation grew up not knowing there was a creek under their street.
Stream daylighting means bringing those buried watercourses back to the surface. You dig up the pipe, create an open channel, design in the meanders and pool-riffle sequences that a natural stream would have, and plant the riparian zone. The stream comes back to life.
It is one of the most rewarding forms of restoration because the change is so visible and so fast. A buried culvert produces nothing — it is a pipe. An open stream with native plantings on its banks produces habitat for fish, amphibians, invertebrates, and riparian birds within a few growing seasons. It infiltrates water into the ground, reducing downstream flooding. It cools the surrounding area through evapotranspiration. And it gives the community access to something beautiful that they never knew was there.
[Figure in the printed edition: before/after of stream daylighting — concrete channel vs naturalized stream with riparian planting]
Fish Habitat and Riparian Zones
Every stream in a healthy landscape has a riparian zone — the vegetated buffer along its banks. That zone does more than prevent erosion. Tree canopy over the stream provides shade that keeps water cool. Leaf litter falling into the water provides food for aquatic invertebrates, which feed the fish. Root systems along the bank create undercut habitats where fish shelter. Fallen logs and branches in the stream create pools and riffles — the structural diversity that supports different life stages of different species.
When we remove the riparian zone — clearing trees for agriculture, mowing to the edge of the bank, paving right up to the water — we destroy all of those functions simultaneously. The water gets hot, the food web collapses, the banks erode, the channel straightens, and the stream becomes a ditch.
Restoring riparian zones is one of the highest-value interventions available. A 30-metre buffer of native trees and shrubs along a degraded stream can measurably improve water temperature, reduce sediment loading, and increase fish populations within five to ten years. The planting itself is straightforward. The hard part is usually convincing the landowner that the 30-metre strip of land along the creek is worth more as forest than as cropland.
Chapter 53: Site Assessment
Before you make a single decision about what to plant, how to manage, or what outcome to aim for, you need to understand what you are working with. A site assessment is the foundation of every restoration project. Skip it, and you are guessing. Guess wrong, and you waste years.
[Figure in the printed edition: restoration ecologist taking soil samples or conducting vegetation survey in the field]
What to Assess
A site assessment covers four domains: soils, hydrology, vegetation, and landscape context.
For soils, you need to know the type, texture, depth, and chemistry. Dig a test pit or use a soil auger. What colour is the soil? Dark brown indicates organic matter; grey or blue indicates waterlogged conditions; red indicates iron oxidation. What texture is it — sandy, silty, clay? How deep does the topsoil go before you hit subsoil or bedrock? Get a soil test done: pH, organic matter content, nitrogen, phosphorus, potassium. In contaminated sites, test for heavy metals and hydrocarbons.
For hydrology, you need to understand how water moves across and through the site. Where does surface water collect? Where does it drain? How deep is the water table? Does the site flood seasonally? Is there a stream, ditch, or tile drain? The moisture regime of the soil — how wet or dry it stays through the year — determines which plant communities can establish. Get this wrong and your planting fails.
For vegetation, survey what is already there. What species are present? Are they native or introduced? What stage of succession is the site in — bare ground, annual weeds, perennial grasses, shrubs, young trees? The existing vegetation tells you a story about the site's recent history and its current trajectory. A field dominated by goldenrod and asters is in a different place than one dominated by buckthorn and dog-strangling vine.
For landscape context, step back and look at the bigger picture. What surrounds the site? Is there intact forest nearby that could supply seed and wildlife? Or is it isolated in a matrix of agriculture or development? How large is the site — can it support viable populations of the species you want to establish, or is it too small and will need ongoing management to maintain biodiversity? Are there corridors connecting it to other habitat patches?
Reading the Site
The assessment is not just a checklist. It is a skill — the ability to look at a piece of land and read what has happened there, what is happening now, and what it is capable of becoming.
A patch of exposed limestone karst with scattered Acacia trees and overgrazed grass stubble tells a specific story. The cattle removed the regenerating seedlings. The grass cover was reduced to the point where soil began eroding off the rock. The remaining trees are the ones the cattle could not reach or did not eat. The seed bank is likely depleted of native forest species but may still contain some hardy pioneers.
That is what Tsunul looked like when we started. Reading that correctly — understanding that the forest wanted to come back but was being prevented by ongoing grazing pressure — led directly to the decision to fence first and let natural succession begin before supplementing with planted nursery stock.
You develop this skill by spending time in both degraded and intact ecosystems. Visit your reference sites. Walk them repeatedly across different seasons. Learn what a healthy version of your target ecosystem looks like, sounds like, smells like. Then visit your degraded site with that reference in your mind. The gap between the two is your work plan.
Chapter 54: Reference Ecosystems
Every restoration project needs a destination. The reference ecosystem is that destination — not a place you will duplicate exactly, but a model of what your site should look like when the work is done.
[Figure in the printed edition: intact reference ecosystem — old-growth forest, pristine wetland, or undisturbed prairie]
What Makes a Good Reference
A good reference ecosystem shares key characteristics with your project site: similar soils, similar climate, similar hydrology, similar position in the landscape. It represents the ecological community that would occupy your site in the absence of degradation. And it is either currently intact or has recovered enough to demonstrate what a healthy version of that community looks like.
In practice, finding a perfect reference is rare. Most landscapes have been altered to some degree. You are often working with the best available — a forest remnant that has been selectively logged but retains its native species complement, a wetland that receives some agricultural runoff but still supports a characteristic plant and animal community, a prairie fragment that has been grazed but never plowed.
For the Toronto corridor project, we identified two remnant prairie ecosystems within the city limits. Neither was pristine, but both contained the native species assemblages we were targeting. We used their species lists as our planting guides and their community structure as our benchmarks for success.
At Tsunul, our references are the semi-deciduous tropical dry forest fragments scattered across the Yucatán. Many have been selectively harvested for centuries, but they retain the species diversity and structural complexity that we are working toward. We walk those fragments regularly, collecting seed and learning which species occupy which positions in the successional sequence.
Using the Reference
A reference ecosystem gives you three things.
First, a species list. The species present in your reference are the species that belong on your site. You do not plant species that do not occur in your reference unless you have a strong ecological rationale — for example, anticipating climate change by including species from slightly warmer reference sites to the south.
Second, community structure. How tall is the canopy? How many layers does the forest have? What percentage of ground is covered by herbaceous plants versus leaf litter? What is the ratio of grasses to forbs in the prairie? These structural attributes tell you what your site should look like at maturity, and you can track your progress toward them.
Third, ecological processes. How often does the reference ecosystem experience disturbance — fire, flood, windthrow? What role do herbivores play? How does the community regenerate after disturbance? Understanding the processes that maintain the reference helps you plan the management regime your site will need.
Limitations
A reference is not a blueprint. It is a guide. Your site may never look exactly like the reference, because your site has a different history, different soil modifications, different seed inputs. Climate is shifting, which means the community your site supports in 30 years may not be identical to the community the reference supports today.
Historical change also matters. In many parts of Ontario, Indigenous peoples maintained prairie and savannah ecosystems through fire for thousands of years before European settlement. The "natural" state of those landscapes was not the forest that grew up after fire suppression — it was the open, fire-maintained community that preceded it. Choosing which historical period to reference is a decision that shapes your entire project.
And some reference conditions are simply lost. In parts of the Yucatán, the only remaining tropical dry forest fragments are small and isolated, visited by fewer seed dispersers than they would have hosted historically. They may be missing species that were present a century ago. The reference gives you the best available picture, not a perfect one.
One thing you notice walking an old-growth forest that you will not find in a restored one: epiphytes. Mosses, lichens, and liverworts covering every trunk and branch. That whole community disappears when a forest is cut and takes decades to centuries to return. It is a component you need to be aware of — your restored forest may have the right tree species, the right structure, even the right understory herbs, but it will be missing that epiphytic layer for a long time. Patience.
The other thing old growth teaches you is gap dynamics. When a large tree falls, it creates an opening in the canopy. Light hits the forest floor for the first time in decades. Seedlings of shade-tolerant species that were sitting in suppressed growth suddenly bolt upward. Pioneer species germinate from the seed bank. The gap fills, and the mosaic of age classes that characterises old growth gets another patch added. In restoration, we can mimic this by thinning — selectively removing trees to create openings that stimulate regeneration. Some conservation managers do it deliberately to accelerate the development of mixed-age forest structure.
Use the reference as a compass, not a GPS coordinate. It tells you the direction. The exact path will be yours to figure out on the ground.
Chapter 55: Rural Landscapes and Agricultural Restoration
The most widespread form of ecosystem degradation on the planet is agriculture. Not because growing food is wrong — people need to eat. But the way industrial agriculture operates in most of the world actively degrades the land it depends on.
[Figure in the printed edition: industrial monoculture field with bare soil between rows, or eroded agricultural landscape]
The Industrial Model
A farmer operating at industrial scale in Ontario or the American Midwest is locked in a system that demands degradation. To pay for the half-million-dollar combine, they need a thousand hectares in production. To grow corn or soybeans on that scale, they till repeatedly, exposing bare soil to erosion. They apply synthetic fertilisers because the soil organic matter — the natural nutrient reservoir — has been depleted by decades of tillage. They spray herbicides to manage weeds because they are growing a monoculture that cannot compete with them on its own.
The result is predictable. Soil loss. Declining organic matter. Compaction from heavy equipment. Water contamination from fertiliser and pesticide runoff. Biodiversity collapse — the fields support almost nothing except the crop and the weeds that survive the herbicides. Groundwater depletion from irrigation. And a farmer who is often barely breaking even despite growing food on an industrial scale.
This is not a moral failing on the farmer's part. It is a system failure. The economics push toward scale, the scale demands simplification, and the simplification destroys the ecological foundations the farming depends on.
The Transition
Moving an industrial farm toward a more ecologically functional state follows the same continuum we have been discussing. You start by reducing impacts. Then you rebuild function. Then, if conditions allow, you move toward something that more closely resembles a natural ecosystem.
Reducing impacts looks like reducing tillage, reducing synthetic inputs, and diversifying the crop rotation. I filmed a series of fields in Ontario that showed the spectrum. One field had been tilled to bare soil — carbon releasing into the atmosphere, no protection for the microbiome underneath. The next field had a single-species cover crop — oats, with weeds coming up between the rows. Better than bare soil, but still a monoculture.
The step after that is no-till. You harvest the crop, leave the stubble standing, and the following season a drill cuts a line through the stubble, drops seed, and closes it up. The stubble becomes a mulch layer — it holds moisture, protects the soil surface, and keeps the microbial community alive. The difference in soil moisture between a tilled field and a no-till field after a dry week is dramatic. The tilled field cracks. The no-till field stays moist under its residue layer.
And the cover crop mix matters. A single-species cover crop supports the microbiome for that one species. A diverse mix — nitrogen fixers, grasses, broadleaf herbs — supports a diverse microbial community. If you are planting corn next season, you want a cover crop that builds the microbiome corn needs. If you are planting soybeans, you want one that supports legume-associated fungi and bacteria. The conventional corn-soybean rotation works partly because it alternates monocots and dicots, which disrupts pest cycles. A diverse cover crop does the same thing while also feeding the soil.
The next step is integrating ecological elements into the agricultural landscape. Hedgerows between fields provide habitat corridors for pollinators and pest predators. Riparian buffers along watercourses filter runoff and stabilise banks. Windbreaks reduce soil loss from wind erosion and create microclimates that benefit crops and wildlife alike.
Agroforestry takes it further — integrating trees into the farming system. Alley cropping places rows of trees within crop fields. Silvopasture combines trees with livestock grazing. Forest farming grows shade-tolerant crops under a forest canopy. Each of these approaches increases the structural and biological complexity of the farm while maintaining productive capacity.
[Figure in the printed edition: agroforestry system with tree rows and crop alleys, or established hedgerow between fields]
The Yucatán Version
In the Yucatán, the traditional milpa system is a form of agroforestry that has sustained Maya communities for millennia. A patch of forest is cleared, crops are grown for a few years, then the plot is left fallow and the forest regenerates while a new plot is opened. It is a rotational system that works with succession rather than against it.
Industrial agriculture in the Yucatán looks different — large-scale cattle ranching that converts forest to pasture permanently, or soy and sorghum monocultures that replace the milpa rotation with continuous cropping. The impacts are the same as anywhere: soil degradation, biodiversity loss, water contamination.
The restoration path in the Yucatán often starts with removing cattle and allowing natural regeneration to begin. The tropical dry forest is remarkably resilient if you let it recover. Within a few years of removing grazing pressure, pioneer species establish from the surrounding seed rain. Within a decade, you have a young forest with a developing canopy. The milpa model, adapted and improved with restoration ecology principles, offers a pathway that feeds people while building rather than degrading the landscape.
Chapter 56: Goals, Objectives, and Cultural Values
A restoration project without clear goals is a project without direction. You need to know what you are trying to achieve, at what scale, and over what time frame. And you need to know whose values are shaping those goals — because restoration is never only a scientific exercise.
[Figure in the printed edition: planning meeting with community members, or restoration team discussing site goals in the field]
Setting Goals
Goals are broad statements of intent. "Restore native prairie ecosystem on 200 hectares of hydro corridor." "Return degraded tropical dry forest to closed-canopy condition." "Create functioning wetland habitat on a former brownfield site."
Goals give you direction. Objectives give you specifics. An objective is measurable, time-bound, and concrete. "Establish 80+ native species within the first three growing seasons." "Achieve 75% native plant cover by year five." "Reduce invasive species cover to below 10% within three years."
You need both. Goals keep you oriented. Objectives tell you whether you are getting there. Without objectives, your goal is just a wish. Without a goal, your objectives are just a checklist with no purpose.
When setting goals for the corridor project, we worked at three time scales. Short-term goals covered the first two growing seasons — species establishment, community engagement events, baseline monitoring completion. Medium-term goals covered years three through five — self-sustaining meadow, measurable biodiversity increases, corridor connectivity between two river systems. Long-term goals covered the full vision — a functioning ecological corridor connecting seven river systems across the city, with documented species movement, stable management regime, and community stewardship in place.
Cultural Values
Here is where restoration gets complicated in the best possible way.
An ecologist might set a goal based purely on biodiversity targets and ecosystem function. But the community living next to the site has its own relationship with that land. Indigenous peoples may have specific knowledge about what species belong there and how the land was managed historically. Recreational users want access. Neighbours want it to look "maintained." School groups want it to be educational. Funding bodies want measurable outcomes.
All of these values are legitimate. The job of the restoration planner is not to pick the "right" values — it is to create a plan that serves as many of them as possible without compromising the ecological integrity of the project.
In many parts of Ontario, Indigenous fire management maintained open ecosystems for thousands of years before European contact. Restoring those ecosystems means acknowledging that the "natural" condition includes human intervention — that people are part of the system, not separate from it. This is consistent with everything we have discussed in this book. No separation. Relationship. Reciprocity.
In the Yucatán, restoration on Maya land requires understanding the milpa cycle, the ceremonial relationship to specific species, and the community governance structures that determine land use. You do not walk in with a planting plan and tell people what to grow. You listen, learn what matters, and find where the ecological and cultural goals overlap. That overlap is where the strongest projects happen.
Chapter 57: Governance, Policy, and Scaling Up
A single restoration project, no matter how well executed, does not change the trajectory of a landscape. Scaling up requires governance structures that support restoration, policies that incentivise it, and funding mechanisms that sustain it.
[Figure in the printed edition: landscape-level view showing connected habitat patches and restoration corridors]
Why Governance Matters
Every project we have discussed in this section operated within a governance framework — municipal permits, utility company agreements, community consultations, funding applications. The Toronto corridor crossed municipal boundaries, utility easements, and parks department jurisdictions. Getting permission to stop mowing 200 hectares of hydro corridor required buy-in from the utility company, the city, the surrounding neighbourhoods, and the schools.
Governance determines who makes decisions, how conflicts get resolved, and who is accountable for outcomes. Poor governance kills projects. A project with strong community support can be undone by a single municipal council decision. A project with solid funding can stall if the permitting process takes three years. You plan for this the same way you plan for invasive species management — by identifying the risks early and building structures to mitigate them.
Policy as a Force Multiplier
The seed mix guidelines I created for the Greater Toronto Area were a policy intervention. They changed the default. Before the guidelines, a developer who disturbed a site would re-seed with whatever was cheapest — usually a mix of non-native grasses. After the guidelines, the same developer had to use regionally appropriate native seed mixes.
One policy document, applied across 6.3 million people, changed the trajectory of hundreds of individual sites. That is the power of policy. It takes the knowledge generated by individual projects and embeds it in the rules that govern future projects.
The same applies at larger scales. Provincial or national policies that require ecological offsetting — compensating for habitat loss in one location by restoring habitat elsewhere — can drive restoration investment at scales that individual projects cannot reach. International agreements like the UN Decade on Ecosystem Restoration create political pressure and funding streams that trickle down to local projects.
Scaling Up
Scaling restoration from site to landscape is the central challenge of the field. A single restored wetland is good. A network of restored wetlands connected by riparian corridors, embedded in a matrix of sustainable agriculture, protected by policy, and managed by local communities — that is what changes a landscape.
The corridor project in Toronto was a scaling exercise. Sixteen kilometres. Seven river systems. Fifteen parks. Two hundred hectares. The goal was never just to plant some wildflowers — it was to create a connected ecological infrastructure across an entire metropolitan area.
In the Yucatán, scaling means connecting the forest fragments that currently exist as isolated patches in a sea of ranches and farms. Every fragment that is protected, every degraded site that is restored, every corridor that links two fragments — these are the building blocks of a landscape-scale recovery. Tsunul is one node in what needs to become a network.
The opportunity is right next door. Our neighbour's ranch — about 13 hectares — recently went out of cattle production. Standing at the fence line, you can see the difference between the two properties. Our side: filling in with trees, getting greener every year, birds calling constantly. His side: denuded, quiet, the thorny pioneer shrubs just starting to recover. It looks like our property did when we started.
That is a 13-hectare opportunity. If we can work with the neighbour — create a conversation about restoring his land, maybe through ecotourism or community programming at Earth Connection Centre — we expand the reserve. Then the next neighbour. Then the next. Each property that connects to the network increases its value to wildlife. We started getting 50+ bird species per day at our wildlife ponds. The neighbour gets almost none — because he has no water available. One pond on his property, connected to our habitat, changes everything for the wildlife in that area.
The city of Mérida expanded from one million to two million people in eight years. The development pressure on the surrounding forest is enormous. Modelling what ecological restoration looks like at the site level — and then showing neighbours how to replicate it — is how you scale in a landscape like this. Not by buying thousands of hectares, but by building a network of restored properties, one conversation at a time.
Funding remains the biggest barrier. Restoration is chronically underfunded compared to its benefits. The ecosystem services generated by a single restored wetland — flood control, water filtration, carbon sequestration, biodiversity support — are worth far more than the cost of the restoration. But those benefits accrue to the public, and the costs fall on whoever does the work. Until the economics are restructured so that restoration pays for itself, scaling will remain slow.
The Decade on Ecosystem Restoration is an opportunity. Political attention creates funding. Funding creates projects. Projects create knowledge. Knowledge creates policy. Policy creates the next generation of projects. The wheel is turning — our job is to keep pushing it.
That is the end of Section III. You now understand what ecosystem restoration is, how the continuum works, and what it takes to plan and govern a restoration project. In Section IV, we get into the practical work — the seeds, the soil, the planting, the monitoring, and the case studies that show how all of this comes together on the ground.
SECTION IV: ECOLOGICAL RESTORATION PRACTICES
Chapter 58: Defining Ecological Restoration
In Section III, we covered ecosystem restoration — the broad umbrella that includes everything from urban greening to brownfield remediation. Now we narrow the focus. Ecological restoration is a specific discipline: the process of assisting the recovery of an ecosystem that has been degraded, damaged, or destroyed.
That word — assisting — is the most important in the definition.
[Figure in the printed edition: degraded site beginning natural recovery with pioneer species emerging]
We are not building ecosystems. We are not engineering nature. We are assisting a process that already knows how to work. The forest wants to come back. The wetland wants to refill. The prairie wants to seed itself in. Our job is to remove the obstacles, provide the missing pieces, and get out of the way.
Restoration ecology is the science behind this work. It is what I do professionally — take practices that people are using in the field, test them in different conditions, figure out which ones work best for which situations, and feed that knowledge back into practice. It is applied science in the truest sense. Every restored site is an experiment. Every monitoring plot is data.
When the UN launched the Decade on Ecosystem Restoration, they chose the broader term deliberately. It was political — ecosystem restoration includes rehabilitating urban environments, improving agricultural practices, and managing degraded landscapes, which brings more countries and more funding into the conversation. I understand the strategy. But I was a little disappointed, because the term ecological restoration carries a precision that matters. It means recovery. It means native species. It means self-sustaining. It means measured against a reference ecosystem. Those standards keep the work honest.
In this section, we use both terms. But when we say ecological restoration, we mean the real thing — assisting an ecosystem toward full recovery of its biodiversity, structure, and function. Not just making a degraded site look better. Making it work again.
Chapter 59: The Eight Principles of Ecological Restoration
The Society for Ecological Restoration established eight principles that define what ecological restoration is and how it should be practiced. These are not the same as the ten principles of ecosystem restoration we covered in Chapter 46 — those were about planning and governance. These eight principles are about the work itself.
[Figure in the printed edition: restoration team planting native species in a degraded landscape]
1. Ecological restoration engages stakeholders and the broader community.
The people who live on, near, or depend on the land have to be part of the process. Not consulted after the fact — involved from the beginning. When I started work in the Yucatán, I did not arrive with a plan and tell the neighbours what was happening. I talked to the farmers, the ejido governance, the local biologists. Their knowledge of the landscape — which areas flood, where the old forest used to be, what species the cattle will not eat — informed every decision.
2. Ecological restoration draws on many types of knowledge and practice.
Scientific literature gives you the ecological theory. Indigenous knowledge tells you what species belong where and how the land was managed before colonisation. Local knowledge from farmers and land managers tells you what has worked and failed on similar sites. Practical experience from other restoration practitioners tells you what the textbooks leave out. You draw on all of it.
In the Yucatán, understanding the mastodon changed how I looked at certain tree species. Many trees have enormous thorns and seeds with incredibly tough coats. Those traits evolved to deter — and survive passage through — large herbivores that went extinct only about 10,000 years ago. The trees have not had time to lose those adaptations. Understanding that evolutionary history changes how you approach seed processing and germination.
3. Ecological restoration is informed by native reference ecosystems.
We covered this in Chapter 54. Your reference is your target, your species list, your structural benchmark. Without it, you are guessing.
When I arrived in the Yucatán, there was no perfect reference ecosystem within easy reach of Tsunul. The surrounding landscape had been modified for so long that finding an intact example of what our site should become required driving a couple of hours. I rebuilt the picture from fragments — conversations with ecologists, visits to remnant forests scattered across the peninsula, species lists from published surveys. The reference became a composite, built from the best available information. It is not perfect. It is what we have.
4. Ecological restoration supports ecosystem recovery processes.
This is the "assisting" principle. You are not forcing the ecosystem into a shape — you are supporting the processes that drive recovery. Succession, seed dispersal, soil building, nutrient cycling, hydrological function. Every intervention you make should align with the direction the ecosystem would go on its own if the obstacles were removed.
At Tsunul, the primary intervention for the first several years was removing cattle and managing invasive grasses. Once those pressures were gone, the tropical dry forest began recovering on its own from seed dispersed by birds, bats, and wind from nearby forest fragments. We supplemented with nursery-grown native species to accelerate the process and fill gaps, but the heavy lifting was done by natural succession. We assisted. The forest did the work.
5. Ecological restoration is assessed against clear goals and objectives.
You must be able to answer the question: is it working? That requires measurable goals set before you start, and monitoring data collected systematically as the project progresses. "Plant native species" is not a goal. "Establish 60 native woody species within five years, achieving 40% canopy cover by year seven" is a goal.
6. Ecological restoration seeks the highest level of recovery attainable.
Do not settle for less than the site can support. If full ecological recovery is achievable, that is the target. If constraints limit the outcome — urban context, contamination legacy, climate change — push for the best possible result within those constraints. The corridor project could not produce forest, but it could produce a fully functioning native prairie ecosystem. That was the highest achievable level for that site, and that is what we aimed for.
7. Ecological restoration gains cumulative value when applied at large scales.
A single restored hectare helps. A thousand connected hectares changes a landscape. The value of restoration is not linear — it compounds. Connected habitat patches support larger populations, allow gene flow between populations, provide movement corridors for species tracking climate change, and generate ecosystem services at a scale that isolated patches cannot.
8. Ecological restoration is part of a continuum of restorative activities.
This connects back to the continuum from Chapter 48. Ecological restoration is not separate from remediation, rehabilitation, or impact reduction. It is the high end of the same spectrum. A remediated brownfield may eventually be ready for ecological restoration. A rehabilitated wetland may continue progressing toward full recovery without further intervention. The stages flow into each other.
Chapter 60: Standards and the Recovery Wheel
How do you measure whether a restoration project is succeeding? Not in the feel-good sense — in the concrete, defensible, this-is-the-data sense? The Society for Ecological Restoration developed standards of practice and a recovery wheel that give you a structured way to track progress.
[Figure in the printed edition: ecosystem recovery wheel diagram showing the six attributes]
The Recovery Wheel
The recovery wheel assesses restoration progress across six attributes, each measured against your reference ecosystem:
Absence of threats. Are the factors that degraded the site still active? If cattle are still getting through the fence, or stormwater is still delivering contaminants, you score low on this attribute regardless of how many native species you have planted.
Physical conditions. Is the soil structure recovering? Is the hydrology functioning? Are the abiotic conditions — temperature, moisture, light — moving toward what the reference ecosystem shows? At Tsunul, we track soil organic matter accumulation as a key physical indicator. When we started, the overgrazed areas had thin, degraded soil on exposed limestone. A decade in, the recovering areas have measurably deeper, darker, more organic soil.
Species composition. Are the right species present? Not just native species in general — the specific species that characterise your reference ecosystem. Are they reproducing? Are all the life stages present — seedlings, juveniles, mature individuals? A forest with only adult trees and no seedlings is a forest on its way out, no matter how impressive it looks.
Structural diversity. Does the site have the physical complexity of the reference? Canopy layers, understory shrubs, ground cover, leaf litter, dead wood, root networks. Structure creates habitat. A structurally simple ecosystem — even one with the right species — supports fewer organisms than a structurally complex one.
Ecosystem function. Are the ecological processes working? Nutrient cycling, decomposition, pollination, seed dispersal, water infiltration. These are harder to measure than species lists, but they are what make the ecosystem self-sustaining. A site where you have to keep planting because nothing is reproducing on its own has a function problem.
External exchanges. Is the restored site connected to the surrounding landscape? Are seeds arriving from adjacent habitats? Are animals moving through? Is water flowing in and out in patterns consistent with the reference? Isolation reduces resilience. A well-connected site recovers faster and maintains itself more reliably than an isolated one.
Using the Wheel
You score each attribute on a scale, compare it to the reference, and plot the results. The shape of the wheel tells you where you are strong and where the gaps are. If your species composition is on track but your structural diversity is lagging, you know you need to focus on creating physical complexity — adding woody debris, leaving dead standing trees, planting in clusters rather than uniform rows.
The wheel gets assessed periodically — annually or every few years depending on the project timeline. Over time, the shape should expand outward as the site converges on the reference. If an attribute stalls or regresses, that tells you something has changed — a new threat has appeared, a management intervention is needed, or your goals need reassessing.
Leading Practices
The standards identify several leading practices for ecological restoration. Developing reference models. Incorporating historical change — understanding how the landscape has been modified over time, including by Indigenous management. Identifying species of concern, both threatened species that need protection and invasive species that need management. Setting clear goals and monitoring them. Engaging the community. Documenting everything and sharing the knowledge.
These are not bureaucratic extras. They are what separates restoration from landscaping. Landscaping makes a site look nice. Restoration makes it function as an ecosystem. The standards are how you prove the difference.
Chapter 61: Site Preparation and Restoration Approaches
Before you put a single plant in the ground, the site has to be ready to receive it. Site preparation is the work that happens between "we have a plan" and "we start planting." Get it right and your plants establish. Get it wrong and you spend years fighting problems that could have been prevented.
[Figure in the printed edition: site being prepared for restoration — invasive removal, soil work, or controlled burn]
Clearing Invasive Species
If the site is dominated by invasive species, those have to go before native species have any chance. The approach depends on the species and the scale.
Mechanical removal — cutting, pulling, mowing — works for some species on smaller sites. Buckthorn can be cut and the stumps treated to prevent resprouting. Phragmites can be mowed repeatedly to exhaust the root reserves, though it usually requires multiple seasons. Dog-strangling vine has to be pulled before seed set or you are spreading the problem.
Solarisation is a technique that works well on smaller areas. You lay clear or black plastic sheeting over the ground and leave it for a growing season. The sun heats the soil underneath, killing vegetation and germinating weed seeds that then die without light and water. When you remove the plastic, you have a relatively clean seedbed.
Seed bank depletion is the agricultural version of the same idea. You till the soil lightly to bring weed seeds to the surface and stimulate germination. When the seedlings reach 10 to 15 centimetres, you knock them back — tilling again, or mowing, or flame weeding. Then you let the next flush germinate and repeat. Each cycle depletes the weed seed bank further. Three or four cycles over a growing season can dramatically reduce weed pressure for your native planting.
Fire, where permitted and appropriate, is one of the most effective preparation tools. It clears accumulated dead vegetation, recycles nutrients, stimulates germination of fire-adapted native species, and sets back invasive species that are not adapted to fire. In prairie restoration, prescribed fire is not just preparation — it is ongoing management, mimicking the natural disturbance regime that maintained prairies for millennia.
Soil Preparation
If the soil has been compacted by heavy equipment or livestock, it may need decompaction before roots can penetrate effectively. This might mean ripping with a subsoiler, or it might mean adding organic matter to improve structure over time.
If the topsoil was removed during previous land use, you may need to import soil or build it in place using nurse crops and organic amendments. This is slow work. It took nature thousands of years to build the topsoil that humans can strip in an afternoon.
Soil pH may need adjustment. Most native plant communities are adapted to specific pH ranges, and a site where soil pH has been altered by industrial use, agricultural liming, or acid deposition may need amendment before native species will thrive.
Hydrology Restoration
Altered hydrology is one of the most common problems on restoration sites. Tile drains in former agricultural fields have lowered the water table, drying out areas that were historically wet. Channelised streams move water off the landscape too quickly, missing the slow infiltration that recharges groundwater. Stormwater infrastructure directs runoff into ponds and pipes instead of letting it soak into the ground.
Restoring hydrology sometimes means removing tile drains and letting the water table recover. Sometimes it means plugging drainage ditches or installing berms to slow water movement across the site. On stream restoration projects, it may mean reintroducing meanders, pools, and riffles to a straightened channel.
The point is that plants need water in the right amounts at the right times. If the hydrology is wrong, the planting will fail no matter how good your species selection is. Fix the water first.
Planting Design
How you arrange your plantings matters more than most people think. Nature does not plant in rows. Trees grow in clusters, with gaps between them where light reaches the forest floor and different species fill different niches. A meadow is not a uniform blanket of identical grasses — it is a mosaic of patches, each with its own mix of species responding to micro-variations in soil, moisture, and light.
Design your planting to mimic this patchiness. Cluster trees of the same species where they would naturally occur together. Leave gaps for meadow species. Create edges between different community types. Include species from every stage of succession, not just the climax species — the pioneers nurse the slower-growing species through the vulnerable early years.
Chapter 62: Managing Succession
Succession is the engine of restoration. We covered how it works in the ecology section — pioneer species colonise disturbed ground, modify conditions, and create opportunities for the next wave of species. In restoration practice, your job is to work with that process, not against it.
[Figure in the printed edition: early successional vegetation giving way to later species — pioneer trees shading out grasses]
Pioneer Species vs Climax Species
Pioneer species produce large quantities of small seeds, dispersed mostly by wind. The seeds stay viable in the soil for years, waiting for disturbance. When it comes — fire, flood, clearing — they germinate fast, grow fast, and die relatively young. They are small at maturity and intolerant of shade. Fireweed colonising an old rail line. Poplars sending root suckers across an abandoned field.
Later successional species — the climax community — are the opposite. Fewer, larger seeds, often dispersed by animals. Shorter seed viability. Slower growth, larger at maturity, shade-tolerant. These are the oaks, the maples, the tropical hardwoods that form the mature forest canopy.
[Figure in the printed edition: fireweed colonising a disturbed site, or poplar suckers spreading across cleared ground]
You need both in a restoration project, and you need them in the right sequence. If you plant only climax species on an open site, they will struggle — too much sun, too much exposure, too much competition from weeds. If you plant only pioneers, you get a pioneer community that never advances.
Nurse Crops and Cover Crops
A nurse crop is a fast-growing species planted specifically to shelter slower-growing target species. In Ontario prairie restoration, we often use oats or annual rye as a nurse crop — they germinate quickly, shade out competing weeds, and die at the end of the season, leaving space for the native perennials to establish. In tropical forest restoration, fast-growing leguminous trees can serve as nurse species, providing shade and fixing nitrogen while the slower hardwoods grow up beneath them.
Cover crops do a similar job but with a focus on soil protection. They hold the soil against erosion, maintain moisture, and add organic matter as they decompose. Crimson clover, winter rye, buckwheat — these keep the soil biology active and the surface protected during the vulnerable period between site preparation and native species establishment.
The key is understanding that these are temporary. Nurse crops and cover crops are not the goal. They are scaffolding. You put them up to support the structure while it builds itself, then they come down — or in this case, they are outcompeted by the native species they were protecting.
Working with the Succession Clock
In a natural system, succession takes decades to centuries. In restoration, we can accelerate it — but only so much. You cannot skip stages. A closed-canopy tropical forest does not establish on an open, sun-blasted former pasture in five years. You need the pioneer canopy first to create the shade conditions that the climax species require.
At Tsunul, the succession clock has been running for over a decade. The pioneer trees — Piscidia piscipula, Bursera simaruba, various acacias — went in first, either planted or self-seeded from surrounding fragments. They created canopy cover within three to five years. Under that canopy, the conditions shifted — more shade, more humidity, more leaf litter, richer soil biology. Now the later successional species are coming in, some from our plantings, some dispersed by birds and bats from nearby forests.
You cannot rush this. But you can help it. You plant the pioneers in year one. You manage the invasive grasses that would otherwise outcompete everything. You add the mid-successional species once the canopy starts closing. And you monitor constantly, adjusting your approach based on what the site is telling you. That is adaptive management — restoration in conversation with the ecosystem.
Chapter 63: Seed Biology and Collection
Seeds are the currency of restoration. Every native plant community you want to establish starts with getting the right seed, from the right source, at the right time. Understanding seed biology is not academic — it is the difference between a successful project and an expensive failure.
[Figure in the printed edition: diverse collection of native seeds — large tree seeds alongside small wildflower seeds]
Seed Structure and Viability
A seed is a plant embryo packaged with a food supply and wrapped in a protective coat. The size of the seed, the thickness of the coat, and the duration of viability vary enormously between species — and those differences determine how you collect, clean, store, and germinate each one.
Pioneer species tend to have small seeds with long viability. They sit in the soil seed bank for years or decades, waiting for the right conditions. That long viability is an adaptation to unpredictable disturbance — the seed does not know when the next fire or flood will open up habitat, so it waits.
Climax species tend to have larger seeds with shorter viability. Many tropical hardwood seeds are recalcitrant — they cannot be dried and stored like grain. They have to be planted fresh, within days or weeks of collection. If you miss the collection window, you wait another year.
Dormancy and Stratification
Many seeds will not germinate even under perfect conditions unless their dormancy has been broken. Dormancy is a survival mechanism — it prevents germination at the wrong time of year or under conditions that would not support the seedling.
Cold stratification mimics winter. You place the seed in a moist medium and refrigerate it for a species-specific period — sometimes a few weeks, sometimes several months. When you bring it out to warmth, the seed interprets the temperature change as spring and germinates.
Scarification breaks down hard seed coats that physically prevent water from reaching the embryo. You can scarify mechanically — nicking the seed coat with a file or tumbling seeds with abrasive material — or chemically, using sulfuric acid for species with very tough coats. In nature, scarification happens through gut passage (the seed passes through an animal's digestive system), fire (heat cracks the coat), or freeze-thaw cycles.
In the Yucatán, many tree species have seeds with incredibly tough coats — thick enough to survive passage through a mastodon's gut. Those mastodons have been gone for 10,000 years, but the seeds still carry the adaptation. We sometimes soak seeds in water for 24 to 48 hours to soften the coat, or scarify mechanically before planting.
Collection Timing
Knowing when to collect is everything. Collect too early and the seed is immature — it will not germinate. Collect too late and the seed has dispersed — it is gone, scattered by wind or eaten by birds.
[Figure in the printed edition: seed collection in the field — person gathering seed from native plants, or phenology chart]
The phenology charts I described in Chapter 47 serve a second purpose here. When you have tracked flowering and seed set times across multiple species over multiple years, you know exactly when to show up for collection. Wild bergamot sets seed in August. Prairie smoke is ready in late June. You build a collection calendar specific to your site and your species list.
In practice, I chart species down the left side of a table and months across the top, divided into four weeks each. When I find a species flowering, I mark it. When it sets seed, I mark that. Over a season, the chart gives me a visual map of the entire community's reproductive timing. After two or three years of data, the patterns are reliable enough to plan collection schedules months in advance.
For grasses like switchgrass (Panicum virgatum), the technique is simple: grab the seed head at the base and run your hand up the stem. The ripe seeds fall into your palm. The unripe ones stay attached. For large-volume collection, use a sickle — grab a handful of stems, cut, drop into a collection bag. The seed needs two to three weeks of drying on a tarp, turned regularly, before storage. Cold moist stratification for about a month in a refrigerator breaks dormancy. Feed bags from a local farm make decent storage containers — they breathe, which prevents mould.
For berries and fleshy fruits, collection is often a matter of colour and firmness. When the berries are fully coloured and come off the stem easily, they are ready. Some species require fermentation to separate the seed from the pulp — you soak the berries in water for a few days, mash them, and the viable seeds sink while the pulp floats.
Collection Ethics and Genetics
Never strip a population. Take no more than 10 to 20 percent of the available seed from any one population. The rest needs to remain for the plants' own reproduction and for the birds and animals that depend on those seeds.
Collect from multiple individuals across the population to capture genetic diversity. If you collect all your seed from one plant, you are planting a monoculture at the genetic level. Collect from 20 or 30 individuals spread across the site, and you are capturing the genetic variation that allows the population to adapt to changing conditions.
Source matters. Seed from a population 500 kilometres away may be genetically adapted to different conditions — different day length, different rainfall patterns, different soil chemistry. Use local seed wherever possible. The general rule is to collect within the same seed zone or ecoregion as your restoration site. There are exceptions — if you are anticipating climate change, you might include some seed from populations slightly south of your site, on the assumption that conditions are shifting in that direction. But this is a deliberate decision, not a convenience.
Chapter 64: Propagation and Growing
Not every restoration project uses nursery-grown plants. Direct seeding is cheaper and works well for prairies, meadows, and some forest types. But for many projects — especially in tropical forests where seeds are recalcitrant, or in wetlands where specific species need to be placed precisely — you need a nursery.
[Figure in the printed edition: restoration nursery with native species in containers, or seedling trays]
Seed Cleaning
Before you can sow seed, you need to clean it. All the chaff, pulp, and debris that came along with collection has to go, because organic matter around the seed feeds bacteria and fungi that cause damping off — a fungal disease that kills seedlings at the soil surface.
Seed screens are the basic tool. I build mine from hardware cloth and a wooden frame. You want at least two mesh sizes: a coarser screen to catch the big chaff, and a finer screen to let dust fall through. The seed lands on the middle screen. Soil sieves from a laboratory supply are more precise but more expensive. Kitchen sieves from a grocery store work surprisingly well for smaller batches.
For some species, the cleaning process involves blending. Yes, a household blender — but with tape on the blades so you are beating the seed free from the pod or fruit without cutting it. Soft-fruited species like berries go into water first. Viable seeds sink. Pulp floats. You skim off the pulp, dry the seeds, and you are ready to sow or store.
Germination
Once cleaned and stratified (if necessary), seed goes into a growing medium. For most native species, I use a mix of peat or coir, perlite, and vermiculite — light, well-draining, low in nutrients. High-nutrient mixes promote the kind of lush, soft growth that does not survive transplanting to a restoration site.
Sow seed at a depth of roughly one to two times the seed diameter. Small seeds go on the surface with a thin covering of vermiculite. Large seeds get planted deeper. Keep the medium consistently moist but not waterlogged. Damping off thrives in wet, stagnant conditions with poor air circulation.
[Figure in the printed edition: seed trays with germinating native species, or close-up of seedlings emerging]
Temperature and light requirements vary by species. Some need light to germinate and must be surface-sown. Some need darkness. Some need warm days and cool nights. Your phenology data and species-specific research guide you here. If you do not have specific information for a species, mimic the conditions under which it germinates in the wild — soil surface in spring for most temperate species, beginning of the rainy season for tropical species.
Vegetative Propagation
Not everything grows from seed. Some species propagate more reliably — or exclusively — from vegetative methods. Cuttings, divisions, layering, and tissue culture all have their place.
Stem cuttings work for many shrub and tree species. Willows are the classic example — cut a stem, stick it in moist soil, and it roots within weeks. But many tropical hardwoods, some wetland species, and various shrubs root reliably from cuttings taken at the right time of year with the right treatment.
Division works for clumping species — grasses, sedges, ferns, many perennials. You dig the parent plant, split it into sections each with roots and shoots, and replant. One mature plant becomes five or ten. This is how we propagated many of the wetland species at Acorus Restoration — we grew stock plants to maturity, then divided them repeatedly to produce the volumes needed for large-scale planting projects.
Tissue culture — micropropagation — is what Sophie specialises in at Tsunul. It produces large quantities of genetically identical plants from small amounts of parent tissue. It is the method of choice for species that are difficult to propagate by other means, or when you need to bulk up a rare species quickly. It requires a lab, sterile technique, and expertise, but it can produce thousands of plants from a single parent.
Hardening Off
Nursery-grown plants live a sheltered life. Consistent moisture, controlled temperature, no wind, no competition. Before they go into the field, they need to be hardened off — gradually exposed to outdoor conditions over a period of weeks.
Move them from the greenhouse to a shadehouse. Then from the shadehouse to partial sun. Then to full exposure. Reduce watering gradually. Let them experience wind. This toughens the stems, thickens the leaves, and prepares the root system for the real conditions they will face on the restoration site.
A plant that goes straight from a greenhouse to a field site in July is a plant that is going to struggle. Take the time to harden off properly. Your survival rates will repay the patience.
Chapter 65: Planting and Implementation
This is where the plan meets the ground. Everything we have covered — the assessment, the reference, the goals, the seed collection, the propagation — converges on the day you start putting plants in.
[Figure in the printed edition: planting day on a restoration site — workers with shovels and trays of native plants]
Timing
In temperate regions, plant in spring or fall. Spring planting gives the growing season for roots to establish before winter. Fall planting allows root growth during the cool moist months, with the plant ready to take off the following spring. Avoid summer planting unless you can irrigate — the combination of heat stress and root disturbance is more than most transplants can survive.
In the tropics, plant at the beginning of the rainy season. At Tsunul, that means June or July. The rains provide the moisture that newly planted trees need to establish, and the cloud cover reduces the heat stress on transplants. Plant at the end of the dry season and you are gambling that the rains will come on schedule. Sometimes they do not.
Planting Technique
Dig a hole wider than the root ball but only as deep. Planting too deep is one of the most common mistakes — it buries the root crown, promotes rot, and kills trees that should have survived. The root flare — where the trunk meets the roots — should be at or just above the soil surface.
For container-grown plants, check the roots before planting. If they are circling the inside of the pot, tease them out or make vertical cuts through the root ball. Circling roots will continue to circle in the ground, girdling the tree and eventually killing it. Better to damage some roots during planting than to leave the girdle in place.
Backfill with the native soil you removed. Do not amend the planting hole with compost or rich soil — this creates a "bathtub effect" where the roots stay in the rich pocket and never extend into the surrounding soil. The plant needs to adapt to the site conditions, not to an artificial microenvironment you created.
Water thoroughly after planting to settle the soil around the roots and eliminate air pockets. Then mulch around the base — but not against the trunk. A ring of mulch 5 to 10 centimetres deep, kept 10 centimetres away from the stem, suppresses weeds, retains moisture, and protects the soil surface.
Spacing and Arrangement
In a forest restoration, spacing depends on the species and the desired outcome. Closer spacing — 2 to 3 metres between trees — creates canopy closure faster, which shades out weeds and creates forest conditions sooner. Wider spacing — 4 to 6 metres — gives each tree more resources but leaves the ground exposed to weed competition longer.
I tend to plant in clusters rather than even rows. A cluster of five trees of the same species, spaced a metre apart, mimics how trees naturally establish from a seed source. Between clusters, leave gaps where different species, or the meadow layer, can develop. This creates the patchy, heterogeneous structure that characterises real forests and supports more biodiversity than a plantation grid.
[Figure in the printed edition: clustered native tree planting on a restoration site, showing natural spacing patterns]
In prairie and meadow restoration, seeding is more common than transplanting. Broadcast the seed mix across the prepared site, roll or rake lightly to ensure soil contact, and let the rain do the rest. Seeding rates vary — a typical tallgrass prairie seed mix might go down at 500 to 1,000 seeds per square metre. The mix includes grasses, forbs, and sometimes a nurse crop to provide early cover.
Community Planting Events
Planting days are not just about getting plants in the ground. They are about building the constituency that will protect the site for decades.
When we organised planting events in the Toronto corridor, we had school groups, neighbourhood volunteers, corporate teams, university students. Most of them had never planted anything in their lives. That was the point. The person who digs a hole, places a native tree seedling, firms the soil around its roots, and comes back six months later to see it growing — that person has a relationship with that site that no amount of signage or education programming can create.
Keep the instructions simple. Show people how to dig the right size hole, how deep to plant, how to water. Pair experienced volunteers with newcomers. Have species identification guides available so people know what they are planting and why. And follow up — send photos of the site through the seasons. Let people see what their work produced.
Chapter 66: Case Studies — Temperate Ecosystems
Theory becomes real in practice. These case studies show how the principles, standards, and techniques we have covered play out in different temperate ecosystem types.
[Figure in the printed edition: temperate restoration site — prairie, wetland, or woodland in a Canadian landscape]
Drylands
Dryland restoration is some of the hardest work in the field. Low rainfall, extreme temperature swings, fragile soil crusts that take decades to form and seconds to destroy. In Ontario's alvar ecosystems — thin soil over limestone, exposed to wind and temperature extremes — the plant communities are uniquely adapted and extremely slow to recover from disturbance.
The key to dryland restoration is patience and minimal intervention. Heavy machinery destroys soil crusts and compacts the substrate. Intensive planting with irrigation creates dependency. The better approach is to remove the source of degradation — usually overgrazing or off-road vehicle use — protect the site from further disturbance, and supplement with seed of species that belong there. Then wait. Drylands recover slowly, but they do recover if the pressure is removed.
I filmed a site in Ontario where a single ATV track through a sandy area had taken the ecosystem right back to bare sand. Everything around it was intact — mosses forming biocrusts on the surface, wild strawberry and yarrow colonising, ferns establishing, grasses holding the soil with root systems that went deep into the substrate. Pull back the vegetation and all that is underneath is sand. The organic layer, the root networks, the biocrust — that is what holds it all together. One repeated disturbance strips it away.
The recovery sequence starts with biocrust — mosses and lichens that colonise the bare sand surface and bind it. They hold enough moisture and nutrients for grasses to germinate. The grasses slow the wind, trap organic matter, and put roots into the substrate. Shrubs come next, then trees. Even an orchid was growing in that harsh sandy habitat — some species are tougher than you would expect. But the whole process is fragile. It builds slowly and can be undone in an afternoon.
[Figure in the printed edition: biocrust forming on bare sand, or ATV damage through sandy ecosystem showing contrast]
Barrier Beaches
Barrier beach systems are dynamic — they are supposed to move. Sand shifts, dunes build and erode, the beach migrates with prevailing winds and storm patterns. Restoration of barrier beach systems means working with that dynamism, not against it.
At Port Stanley on Lake Erie, the barrier beach system protects the wetlands behind it from wave action. When that barrier is compromised — by development, by hardened shoreline structures that interrupt sediment transport — the wetlands behind it are exposed and degrade.
Restoring the beach means re-establishing the vegetation that holds the dunes in place — beach grasses that trap windblown sand and build dune volume. Ammophila breviligulata — American beachgrass — is the workhorse species. Its root system binds sand, its stems trap wind-carried particles, and the dune builds around it. Behind the primary dune, you can establish shrub species and eventually the coastal forest that characterises mature barrier beach systems.
The lesson from barrier beaches is that structure matters as much as species. The dune is a physical structure that creates the conditions for everything behind it. Get the structure right and the biology follows.
Wetlands
Wetland restoration in Ontario often means reversing drainage. Farmers installed tile drains to dry out productive land. Roads and culverts interrupted natural flow patterns. Streams were channelised. The water table dropped, and the wetlands dried up.
Restoring them means reconnecting the water. Remove the tile drains. Block the drainage ditches. Install control structures that allow water levels to be managed during the establishment phase. Once the hydrology is right, wetland plants colonise rapidly — cattails, bulrushes, sedges, rushes. These species are aggressive colonisers with wind-dispersed seed and the ability to spread vegetatively from fragments.
The more difficult part is getting the full community established — the submerged aquatic plants, the floating-leaved species, the emergent species in the deeper zones. These often need to be planted directly, because their seeds do not disperse as readily and they are less competitive during initial colonisation.
At Acorus, we grew over 350 native species, and the largest portion of our production was wetland plants. We supplied plugs and container-grown stock for wetland restoration projects across Ontario and into the northern United States. Getting wetland plants established is not hard if the hydrology is right. If it is wrong, no amount of planting will fix it.
Meadow Restoration — The Hydro Corridor
The Toronto hydro corridor project taught me more about meadow restoration than anything I had read in a journal. We ran experiments on seeding timing across nine plots using a randomised block design — fall seeding, winter frost seeding, and spring seeding.
The results changed our whole approach. Fall-seeded plots and frost-seeded plots produced more wildflowers. Spring-seeded plots produced more grasses. The reason: native wildflowers need cold moist stratification to break dormancy. Fall-sown seed gets a full winter of that treatment. Frost seeding — broadcasting seed onto snow — works because the dark seed absorbs heat, melts through the snow surface, contacts the soil, and goes through the remaining freeze-thaw cycles. Spring-sown seed misses the cold period entirely, so only the grasses — which do not need stratification — germinate readily.
We wanted a 50/50 to 60/40 ratio of grasses to wildflowers. To get there, we shifted from all-spring seeding to predominantly fall seeding.
[Figure in the printed edition: hydro corridor meadow in bloom showing grass and wildflower diversity]
Management was the other lesson. Initially, we were mowing the entire 200-hectare corridor in one season. I changed that to a rotational system — mow a section here, a section there, leave the rest intact. The reason is wildlife. Many insects overwinter inside the hollow stems of prairie plants. Mow everything at once and you destroy the overwintering habitat for the entire corridor. Leave refugia — unmowed patches — and the insects recolonise the mowed areas the following season. We settled on a four-year rotation, disturbing any given section once every four years and always leaving adjacent sections intact.
We also installed shrub nodes — small clusters of native shrubs scattered through the meadow. Berries for birds, structure for nesting, and a height layer that pure grassland lacks. These could not be tall trees — nothing that would interfere with the power lines — but low shrubs like dogwood, ninebark, and meadowsweet added a structural element that boosted biodiversity measurably.
About 50 of our species went in as broadcast seed. The other 50 — rarer species, less competitive species, species that do not germinate well from broadcast — were planted as container-grown stock, often by school groups and community volunteers. That dual structure gave us both the diversity and the community engagement that made the project self-sustaining.
Converting Agriculture to Woodland
Converting a farm field to a forest is a long game. The field has compacted, depleted soil. The seed bank is full of agricultural weeds and empty of native forest species. The site is open, exposed to sun and wind — conditions that favour weeds and stress young trees.
The approach I use starts with site preparation — seed bank depletion, cover cropping, soil decompaction if needed. Then plant pioneer tree species at tight spacing to get canopy closure as quickly as possible. Once the canopy starts shading the ground, weed pressure drops and conditions shift toward what the understory species need.
After five to ten years of canopy development, you begin introducing mid-successional and climax species underneath — shade-tolerant trees, woodland shrubs, forest floor herbs and ferns. These come in as planted stock, because the seed rain from surrounding forest fragments is often inadequate to populate the understory naturally.
The whole process takes 20 to 30 years to produce something that looks and functions like young forest. It takes longer — 50 to 100 years — to develop the structural complexity of mature forest. This is not fast work. But every year, the system becomes more self-sustaining and requires less intervention.
Chapter 67: Case Studies — Tropical Ecosystems
Tropical restoration operates under different rules. Higher temperatures, higher rainfall (seasonal or year-round), faster growth rates, higher species diversity, and different ecological dynamics. The principles are the same — succession, reference ecosystems, site assessment — but the application looks different.
[Figure in the printed edition: tropical dry forest restoration at Tsunul — young canopy developing over former pasture]
Tropical Dry Forest
Tropical dry forest is one of the most threatened ecosystem types on Earth. Across Central America, more than 80% of it has been converted to agriculture and cattle ranching. It is also one of the most resilient — given the chance, it recovers remarkably fast.
At Tsunul, the trajectory has been visible in real time. When we started in 2015, the site was overgrazed karst limestone with scattered trees and depleted soil. We fenced out cattle, began removing invasive grasses, and planted native species from seed we collected in surrounding forest fragments.
Within three years, pioneer species had established a patchy canopy. Within five years, canopy closure was advancing across the site. At ten years, we had 514+ documented species — plants, birds, mammals, reptiles, amphibians, insects — on 39 hectares that had been nearly barren a decade earlier.
The tropical dry forest's successional drive is relentless. The seed rain from nearby fragments is diverse and constant — birds, bats, and wind bring in seeds from dozens of species every year. The soil seed bank, even in degraded conditions, contains viable seeds of pioneer species waiting for the opportunity to germinate. The main obstacle is usually grazing pressure. Remove that, and the forest begins rebuilding itself.
Our role has been to accelerate the process. We plant species that are not arriving on their own — large-seeded trees whose dispersers (like agoutis or larger mammals) are absent or reduced. We manage the invasive grasses that compete with seedlings during the early establishment phase. We monitor and adjust.
Species Selection in the Tropics
In temperate restoration, your species list might be 30 to 80 species for a prairie or 20 to 50 for a forest project. In the tropics, a single hectare of intact forest can support 200+ plant species. Matching that diversity in a restoration planting is not realistic, but you can build a species mix that provides the functional diversity the ecosystem needs.
[Figure in the printed edition: diverse collection of tropical tree seeds collected for restoration planting]
I work with species from every stage of succession. Fast-growing pioneers like Bursera simaruba and Piscidia piscipula go in first — they tolerate full sun, grow fast, and create canopy within a few years. Mid-successional species like Brosimum alicastrum and various Nectandra species go in once the canopy starts providing shade. Late-successional hardwoods — the Manilkara, Vitex, Cordia species — go in last, planted under the developing canopy where they can grow slowly toward their eventual position in the mature forest.
Sophie's tissue culture work allows us to propagate species that are difficult to grow from seed. Some tropical hardwoods have recalcitrant seeds that cannot be stored. Others germinate erratically even under ideal conditions. Micropropagation gives us a reliable supply of these species — which are often the most ecologically important and the hardest to replace.
Chapter 68: Mangroves and Coastal Restoration
Mangroves are the tropical equivalent of the barrier beach systems we discussed for temperate regions — but with even higher ecological stakes.
[Figure in the printed edition: mangrove forest along the Yucatán coast, with prop roots visible in clear water]
Mangrove forests line tropical coasts across the planet, from Southeast Asia to West Africa to the Caribbean. In the Yucatán, they fringe the entire peninsula. They sit at the interface between land and ocean, and they do everything at that boundary — stabilise shorelines, filter runoff, buffer storm surge, and function as nursery habitat for a staggering proportion of the ocean's commercial fish species.
Young fish move into mangrove root systems where the tangled prop roots provide shelter from predators and a rich food supply. They grow up in that protected environment, then move out to open water as adults. Remove the mangroves and you do not just lose the trees — you lose the fishery.
Why Mangroves Are Degraded
The same pressures hit mangroves everywhere. Coastal development clears them for resorts and housing. Aquaculture — especially shrimp farming — converts mangrove areas to ponds. Timber harvest takes the wood. Sea-level rise and increased storm intensity from climate change stress the systems further.
In the Yucatán, tourism development along the coast has destroyed extensive areas of mangrove. The sandy barrier beaches are prime real estate for hotels and condominiums. When the mangroves behind those beaches are cleared, the coast loses its storm buffer, the lagoons behind the barrier lose their filtration, and the fishery declines.
Restoring Mangroves
Mangrove restoration starts with understanding the hydrology. Mangroves exist in a specific tidal zone — too high and they dry out, too low and they drown. Many failed mangrove restoration projects planted in the wrong zone, wasted millions of propagules, and declared the species unplantable. The species was fine. The site selection was wrong.
[Figure in the printed edition: mangrove propagules or young mangrove seedlings being planted in a coastal restoration project]
The correct approach is to identify areas within the appropriate tidal range where mangroves historically grew and have been removed or degraded. Often, simply removing the barrier — filling in the aquaculture pond, reconnecting the blocked tidal flow — is enough. Mangroves are aggressive colonisers when conditions are right. Their propagules — the distinctive pointed seedlings that drop from the parent tree — float in on the tide and root wherever they land in suitable substrate.
Where natural colonisation is too slow, you can collect propagules and plant them directly in the restoration site. The propagules are those distinctive pointed structures that hang from the parent tree — about a hand's length, buoyant, with a sharp tip. When they drop naturally, that point drives into the mud and the seedling roots immediately. You can harvest them by the handful and push them into the substrate at the restoration site. It is one of the simplest and fastest restoration techniques in the tropics. I have watched propagules that were pushed into the mud start putting out roots within days. Survival rates in well-planned mangrove restoration projects can exceed 80%.
The other approach is restoring hydrology first and letting the mangroves come back on their own. In many degraded mangrove areas, the problem is blocked tidal flow — roads with undersized culverts, filled-in channels, development that interrupts the water movement. Simply reopening the flow — cleaning out culverts, digging channels to restore tidal flushing — can trigger rapid natural recolonisation. Propagules float in on the tide and do the rest.
[Figure in the printed edition: mangrove propagules — on the tree, floating in water, or freshly planted in mud]
The barrier beach dynamics apply here too. The mangrove forest behind the beach protects the beach. The beach protects the mangrove from wave energy. They are a system, not separate elements. Restoring one without the other is incomplete.
Chapter 69: Project Walkthrough — From Assessment to Completion
Let me walk you through a full restoration project from start to finish, connecting all the concepts from the last two sections into a single narrative.
[Figure in the printed edition: panoramic view of a restoration project site at the planning or early implementation stage]
The Site
You have been asked to restore a 20-hectare parcel that was used as agricultural land for 40 years. The farmer has retired. The property backs onto a 100-hectare forest remnant and has a creek running through it. The surrounding landscape is a mix of active agriculture and rural residential. A conservation authority holds an easement on the property and wants to see it returned to native habitat.
Assessment
You walk the site. The soils are clay loam, compacted from decades of equipment use. The creek has been straightened into a ditch with no riparian buffer — bare soil banks eroding into the water. The field itself is dominated by annual weeds — lamb's quarters, ragweed, quack grass. There are no native tree seedlings anywhere, despite the forest being right next door. The seed bank is agricultural.
You dig soil test pits. Topsoil is 15 centimetres deep, depleted of organic matter. Below that is compacted subsoil. The water table is artificially low — you find tile drain outlets along the creek. You test soil chemistry: pH is high from years of liming, nutrients are adequate but soil biology is minimal.
You visit the adjacent forest — your reference ecosystem. It is a mixed deciduous forest: sugar maple, beech, white oak, basswood in the canopy. Understory of ironwood, musclewood, witch hazel. Rich forest floor flora — trilliums, wild ginger, bloodroot, hepatica. The creek through the forest has pools and riffles, gravel substrate, overhanging canopy. You find brook trout. This is your target.
Planning
Your goals: convert the agricultural field to native forest connected to the existing remnant. Restore the creek to natural channel form with a forested riparian buffer. Remove tile drains to restore the natural water table. Achieve 80% native species cover within five years and self-sustaining forest conditions within 20.
Site Preparation (Year 0)
You remove the tile drains. The water table begins to recover. The low-lying areas near the creek become saturated — that is correct, they were historically wetland.
You begin seed bank depletion on the upland portions. Three cycles of shallow tillage and germination over the growing season reduce the weed seed bank. You plant a winter cover crop — crimson clover and winter rye — to hold the soil through the winter and add organic matter.
Planting (Years 1–2)
In spring of year one, you plant pioneer trees at 2.5-metre spacing across the upland areas. White cedar, trembling aspen, white birch — fast-growing species that tolerate the open conditions and will create canopy cover within a few years. You include some scattered oaks and maples as advance plantings of the climax species.
Along the creek, you install native willow and dogwood stakes directly into the eroding banks. Willows root from stakes within weeks and stabilise the bank immediately. Behind the willows, you plant a mix of riparian trees — silver maple, bur oak, black cherry.
In the saturated areas where the water table has risen, you plant wetland species — sedges, rushes, joe-pye weed, swamp milkweed. These areas will develop into the kind of riparian wetland that filters water before it reaches the creek.
[Figure in the printed edition: year-one planting with small tree seedlings in rows or clusters on a former farm field]
Management (Years 2–5)
You mow between the tree rows in summer to keep weed competition manageable during the establishment years. You spot-treat any invasive species — buckthorn, dog-strangling vine — before they get established. You monitor survival rates and replant any areas with high mortality.
By year three, the pioneer trees are 2 to 3 metres tall and starting to shade the ground. Weed pressure is dropping. You begin interplanting mid-successional species — white ash, red oak, sugar maple seedlings — under the developing canopy.
Succession Takes Over (Years 5–20)
By year five, the pioneer canopy is closing. The character of the site has shifted from agricultural field to young woodland. The creek banks are stabilised. The riparian wetlands are functioning. You can reduce active management to annual monitoring and targeted invasive species control.
Over the next 15 years, the forest matures. The pioneer species reach their full height. The climax species grow up underneath them. The forest floor develops a leaf litter layer, and woodland wildflowers begin appearing — either from seed dispersed by the adjacent forest or from deliberate introduction. Soil organic matter increases. The soil food web rebuilds.
By year 20, the site is a young but functioning forest, continuous with the adjacent 100-hectare remnant. The creek supports fish habitat. The riparian zone filters agricultural runoff from upstream properties. The tile-drained areas have become productive wetlands. The site is self-sustaining — it no longer needs you.
That is the arc of a restoration project. Twenty years from agricultural field to functioning forest. It sounds long. It is long. But the forest will stand for centuries.
Chapter 70: Monitoring, Evaluation, and Long-Term Stewardship
Planting is the beginning, not the end. The real work of restoration happens in the years and decades that follow, and monitoring is what keeps you connected to what the site needs.
[Figure in the printed edition: long-term monitoring plot in a maturing restoration site, or ecologist recording data]
What to Monitor
At minimum, you track vegetation composition, structural development, and the status of key threats.
Vegetation composition means recording which species are present, how abundant they are, and how that is changing. Are the native species you planted surviving and reproducing? Are new native species arriving from the surrounding landscape? Are invasive species increasing or decreasing?
Structural development means tracking the physical complexity of the site. In a forest, that means canopy height, canopy closure, understory density, ground cover, dead wood volume. In a prairie, it means plant height, litter depth, and the ratio of grasses to forbs. Structure creates habitat — the more complex the structure, the more organisms the site can support.
Threat status means keeping an eye on the things that could set the project back. Invasive species. Altered hydrology. Encroachment from adjacent land use. Climate-related stress like drought or flooding. A new threat that was not present when the project started.
Monitoring Methods
Permanent plots are the gold standard. Mark them clearly, record their GPS coordinates, and revisit them on the same schedule every year. Consistency matters more than complexity — a simple survey done annually for 20 years produces far more useful data than a detailed survey done once.
Photo monitoring is the cheapest and most revealing tool you have. Take photos from fixed points at fixed intervals — the same spot, the same direction, the same time of year. Over time, those photos tell a visual story of change that no data table can match. They are also the most compelling evidence you can show a funder, a community group, or a sceptic.
For biodiversity, supplement your vegetation surveys with targeted monitoring of indicator species. Butterflies are excellent indicators of meadow health. Breeding birds indicate forest structural quality. Aquatic invertebrate diversity indicates water quality. You do not need to identify every organism on the site — just track the ones that tell you the most about the ecosystem's condition.
Adaptive Management
Monitoring without response is just record-keeping. Adaptive management means using your monitoring data to adjust your approach.
If native species are establishing well in one section but failing in another, ask why. Is the soil different? Is there more competition from invasives? Is the hydrology wrong? The monitoring data points you toward the answer, and the answer determines what you do next.
If an invasive species is spreading despite your management efforts, escalate. Try a different control method. Increase the frequency of intervention. Consult someone with experience managing that species. Do not keep doing the same thing and hoping for different results.
If the site is progressing faster than expected in some areas, learn from those areas. What is different about them? Can you replicate those conditions elsewhere on the site?
Adaptive management is restoration in conversation with the ecosystem. You make an intervention. The ecosystem responds. You observe the response. You adjust. This cycle continues for the life of the project.
[Figure in the printed edition: restoration ecologist assessing a monitoring plot, comparing current conditions to baseline data]
Stewardship
The hardest part of restoration is not the planting. It is the long-term commitment.
A restored forest needs monitoring and occasional management for decades. Invasive species do not stop trying to colonise because you planted native trees. Climate change does not stop because your carbon sequestration numbers are improving. Adjacent land uses keep changing. New threats emerge.
Stewardship means building the systems — the funding, the organisational structure, the community engagement, the documentation — that ensure the site continues to be cared for after the initial project team has moved on. This is why the documentation manual I mentioned earlier is so important. It is why community engagement matters. It is why policy integration matters.
A restoration project that relies entirely on one person is a project that ends when that person leaves. Build the knowledge into systems that outlast individuals.
Chapter 71: The Work Ahead
We have covered a lot of ground. Four sections. Seventy-one chapters. From the chemistry of soil to the governance of landscape-scale restoration. From cells to ecosystems. From degradation to recovery.
[Figure in the printed edition: view from Tsunul — recovering tropical dry forest canopy at sunrise or sunset]
If there is one thing I want you to carry forward, it is this: the land wants to recover.
This is not sentiment. It is an ecological observation that holds across every degraded site I have worked with or studied, in 30 years across two continents and dozens of ecosystem types. Remove the obstacle, provide the missing piece, and the system moves toward complexity, diversity, and resilience. It does this because that is what ecological processes do. Succession is not a human invention. It is a property of living systems.
Your job — our job — is to assist that process. To be good enough ecologists to understand what each site needs, and humble enough to let the ecosystem do the heavy lifting once we have given it the chance.
At Tsunul, I watch the forest rebuild itself every day. Species I planted a decade ago are fruiting and their offspring are establishing on their own. Birds I had not seen in the early years are now resident. The soil under the canopy is darker, deeper, richer than the bare limestone of a decade ago. The system is gaining momentum. It needs less from us every year.
That is what success looks like. Not a project that depends on constant intervention. A system that runs itself.
There is so much work to do. Billions of hectares of degraded land worldwide. Trillions of dollars in lost ecosystem services every year. The scale of the problem dwarfs the scale of the response — for now.
But the knowledge exists. The techniques work. The land is waiting. And every person who learns this material and applies it — on one hectare or a thousand — is pushing the recovery forward.
Go get your hands in the soil.