Urban Rewilding Experiments Signals Worth Tracking in 2026

We planted a mini-prairie on a vacant lot in 2019. Six weeks later, half the seedlings were dead, and the survivors were pale. The soil was the problem—packed hard as concrete, with a pH that made no sense. That's the thing nobody tells you about urban rewilding: the ground is never neutral.

This article is a field guide to the soil myths that quietly sink projects. You'll see where they show up, why they persist, and what actually works when you're staring at a barren lot with a bag of seed mix.

Where the Soil Myths Bite: Real Projects, Real Costs

A community garden that failed because of fill dirt

Just off a busy arterial road, the site looked perfect. Sunny, flat, fenced. The city owned it. A grant covered raised beds and drip line. Volunteers signed up in droves. The soil was the last thing anyone checked. That was the first mistake. The topsoil had been stripped years before, and the contractor backfilled with whatever was cheap. What the volunteers got was a subsoil mix heavy with clay and construction rubble. Carrots forked into little claws. Lettuce bolted as if it were August in Phoenix. The project limped through one season, then the grant ended. Frustrating part? A simple soil test before ordering that first truckload of compost would have cost about fifty bucks. Instead the whole project went sideways. We fixed a similar mess once by installing wicking beds above the contaminated ground, but that was after two failed seasons and a lot of volunteer burnout.

Wrong dirt is a silent killer.

A rain garden that flooded because of compaction

Rain gardens are supposed to drain, so the logic goes. But the actual ground underneath a site behaves nothing like the soil texture you read about online. One homeowner project I consulted on dug a lovely basin, filled it with a sandy loam mix, planted sedges and rushes, and waited. First heavy storm hit. The basin stayed filled for four days. Mosquitoes moved in. The problem wasn't the planting mix—it was the native soil beneath, compacted by years of foot traffic and construction. The sandy loam acted like a bathtub liner sitting on a clay pan. The fix involved ripping out the pretty layer, breaking up the subsoil with a pickaxe, and adding a French drain. That took labor, time, and materials nobody budgeted for. The homeowner's bill ran well north of the original estimate.

The catch is that most people test the soil they *want* to have, not the soil they actually have.

The $40,000 mistake: testing soil too late

And then there are the projects where the money moves before anyone checks the ground. I have seen a municipal planting along a new street median—thirty thousand dollars of native shrubs and trees planted into soil that was mostly pulverized asphalt and old fill. The contractor had the test results in hand the whole time. Nobody read them until the trees started dying in year two. Then the replacements, plus the re-excavation, plus the new soil amendment, doubled the original cost. That's the forty-thousand-dollar mistake in a nutshell. Not that the test was missing. Not that the soil was bad. But that the sequencing was wrong. They tested after they committed. They should have tested before they even wrote the grant application.

Order matters more than volume.

Every project has a moment where you can still turn back without losing everything. That moment is usually before you dig. The soil test is cheap. The compaction test is cheap. The percolation test is cheap. They all feel like a waste of time when you have momentum and a crew standing around. But they're the difference between a site that thrives in year three and a site that gets flipped back to turf by the city maintenance crew. I would rather eat the cost of a lab test that says "all clear" than explain to a funder why the rain garden is a pond and the meadow is a dust bowl.

You can't see bad soil from the surface. You can only see its bills come due, one season later.

— a restoration contractor, after pulling out dead stock from a "site-ready" parcel

What counts as "real cost" anyway

The headline numbers matter, but they're only part of the story. There is also the lost labor. The contractor hours spent replanting. The volunteers who never come back because they watched their work die. The neighbor who posts photos of your flooded basin on the community board and calls it a breeding hazard. None of that shows up on an invoice, but it all drains the same account. The projects that fail on soil don't fail with a bang. They fail with a slow, humiliating wither that makes rewilding look like a boondoggle. And that taints the next project, and the next person who wants to try.

What New Rewilders Get Wrong About Dirt

Topsoil Isn't a Single Thing

Most newcomers treat topsoil like a generic ingredient — grab a yard, spread it, done. That assumption wrecks more urban rewilding plots than any weed invasion. The stuff you get from a bulk supplier is often stripped subsoil blended with compost fines, or worse, screened fill that looks dark but holds no living structure. I have watched a project collapse because the delivered "topsoil" dried into a crust that repelled water like a thatched roof. The real thing is a layered horizon, built by decades of roots, fungi, and invertebrates. A truckload can't replicate that. The catch is that your site's original soil — even if it looks dead — already carries the local microbial bank. Don't bury it under mystery fill. Test what you have before you import anything.

Wrong order.

People order soil before they dig a single test pit. Then they discover the existing ground is compacted clay or rubble-filled sand, and the new soil just sits on top like a hat. The layers never integrate. Water pools at the seam, roots hit the boundary and circle, and the whole planting staggers into chlorosis. That hurts. Spend the fifty bucks on a texture test and a pH probe first. The results change everything downstream.

pH and Texture Aren't Optional Reading

I get it — pH feels like high school chemistry, not rewilding. But it decides which nutrients actually dissolve for plant uptake. A pH of 7.5 on urban fill locks up iron and phosphorus, no matter how much compost you add. Texture, meanwhile, controls drainage and root penetration. Sandy loam drains fast; clay holds water and heaves in winter. Most urban sites are neither — they're weird hybrids of demolition dust, old gravel, and decades of foot compaction. You can't guess this by looking. One project I consulted on failed with a wildflower meadow that looked great in spring, then browned out by July. The soil was a thin layer over crushed stone, and the roots had nowhere to go. Not a plant problem. A texture problem.

The fix is not romantic, but it's simple.

Adjust your species list to match the soil, not the other way around. Or amend the texture physically — with grit for clay, or organic matter for sand — before planting. Skipping this step is the single most common mistake I see.

The Lie of 'Just Add Compost'

Compost is not a cure-all. It's a soil amendment, and too much of it creates a nutrient spike that favors aggressive weeds like dock and nettle, not the delicate natives you actually want. Worse, a thick compost blanket on compacted ground can smother roots and create anaerobic pockets. The mantra on every forum — "just add compost and nature will sort it out" — is a half-truth. Nature sorts it out eventually, but often by growing what you didn't plant. If you already have decent texture, a thin half-inch layer of compost is plenty for most meadow and prairie mixes. You're feeding the soil food web, not planting a vegetable garden.

“The soil is not a thing you add to. It’s a relationship you join, mid-conversation.”

— paraphrased from a restoration ecologist’s offhand remark at a site visit

Most new rewilders focus on the visible canopy — the grasses, the blooms, the birds. The invisible layer decides the outcome. That said, nobody checks the drainage hole or the soil biology until something fails. We fixed one neglected plot by simply stopping the compost addiction, re-seeding with a pH-appropriate mix, and letting the existing soil breathe. It took two seasons to look right. Not fast, but real.

Before you schedule a single planting day, do this: dig three test pits across your site. Feel the texture, check the pH, and look at the color change with depth. That's your starting point. Ignore the sales pitch for “premium garden soil” and work with what the land already has.

Soil Prep Patterns That Actually Work

Test first, then plan

Dig a hole before you buy a single seed. That sounds obvious, but I have watched three separate projects order prairie mixes for soil that turned out to be demolition rubble capped with six inches of imported topsoil. The seed germinated fine. Then the roots hit the brick layer and the whole meadow browned out by August. A $400 soil test — or even a shovel and a jar of water — would have told them what they needed to know. pH, texture, drainage, and the presence of old building debris change everything downstream. The catch is that most municipal crews treat soil testing as a bureaucratic checkbox rather than a design input. Wrong order. Test first, then pick plants. The gap between those two steps is where rewilding projects quietly die.

Send samples to a lab that tests for heavy metals too. Urban lots accumulate lead, zinc, and petroleum residue in ways that don't show up in standard agronomy panels. You can grow a gorgeous wildflower patch on contaminated ground — but you can't eat anything from it, and kids playing there create exposure pathways nobody planned for. Soil that fails the metal screen changes your plant palette entirely. It also changes your public messaging. Nobody wants to put up a sign that reads "this meadow is pretty but toxic," yet that's precisely the conversation you need to have with funders before they commit.

Odd bit about activities: the dull step fails first.

Odd bit about activities: the dull step fails first.

Cheap tests beat expensive guesses.

The right way to loosen compacted ground (it's not a rototiller)

Rototillers pulverize soil structure. They break up surface compaction while simultaneously destroying the fungal networks and aggregate stability that make soil behave like soil. I have seen tilled rewilding plots turn into crusted pans within two seasons — the exact condition the project was trying to fix. The better tool is a broadfork or a tractor-mounted subsoiler that fractures compacted layers without inverting the profile. Decompaction should leave the soil's vertical architecture intact. Roots need channels to follow, not a blenderized slurry of dead organic matter.

For most urban sites, the practical pattern is this: assess the depth of compaction with a penetrometer, mark the hardpan layer, then rip to just below that depth. Aerate in two directions if the grid is tight. Then walk away. Don't amend the surface with compost unless the soil test shows organic matter below 3 percent. Compost on top of compacted clay creates a perched water table that drowns taproots and breeds shallow rooting. The plants look green for a month, then collapse.

The most reliable approach I have used is building raised micro-topography instead. Mounding soil into low swales and berms gives you drainage and aeration for free. It's slower to install than mechanical ripping, but it creates permanent variation in moisture and exposure — which means more niche habitats and better species diversity.

Matching seed mix to soil, not the other way around

Everyone wants the picture-perfect seed blend from the catalog. That blend was designed for a loamy farm field with consistent moisture. Your site is probably a weird patch of urban clay, sand, or fill. The plants that thrive on compacted, alkaline, drought-stressed city dirt are not the same plants that win awards at native plant society shows. Choose species that occur naturally on your soil type, not the showiest ones. This means fewer forbs and more tough grasses in the first year. It looks less impressive. It survives.

We fixed one site by swapping the standard pollinator mix for a prairie dropseed and little bluestem base, then seeding only 15 percent wildflowers that matched the local calcareous clay. The first spring looked bare. The second spring looked like a grassland with some color. The third spring was stunning. Projects that push a 70 percent forb mix onto heavy urban soil spend their whole first year fighting seedling death, then blame the seed company. The seed was fine. The match was wrong.

Soil wants its own plants. Force a mismatch and you will spend every season weeding, watering, and replacing what should have grown on its own.

— Field observation, multiple failed high-forb plots

The other pattern that works is hydrology-led planting. Map where water pools after a storm and place moisture-loving species there; put drought-tolerant ones on the high spots. Most mixed urban sites contain three or four distinct soil moisture zones within a single acre. Treat them as separate planting areas, not one flat canvas. It sounds like more work, but it cuts mortality dramatically and produces a patchwork that looks intentional rather than sad.

What usually breaks first is the assumption that one seed mix can serve the whole site. It can't. And the second assumption — that you can fix soil chemistry with fertilizer — is equally fragile. Urban soil needs structure first, nutrition later.

Start with a wedge of your site: dig, test, decompact one small plot, plant a test strip with a mix chosen for your actual dirt. Watch it through a full season. Then scale up with evidence, not hope. That pattern has never failed me. It just takes longer — and the patience pays in the third year, when everyone else's meadow is reverting to weeds and yours is still standing.

Why Projects Revert to Lawns and Weeds

The 'native seed, no care' fantasy

Most rewilding projects die in the first eighteen months, and the killer is usually a seed packet. Teams scatter native mixes on bare dirt, walk away, and expect prairie. What they get is a thin layer of annual weeds that burn out by August. Native perennials need time to establish root systems, but nobody budgets for that patience.

The catch is practical: bare soil is a vacuum. Weeds find it first.

I have watched a well-funded restoration site on a corporate campus turn into foxtail and thistle within one season. The crew had done everything right on paper—proper seed ratios, fall planting, even a straw mulch. What they lacked was a single weeding pass in June. By July, the invasives had shaded out the tiny coneflower rosettes. That project never recovered, and the grounds crew mowed it flat the next spring.

Rewilding is not a planting event. It's a five-year relationship with a piece of ground.

When weeds outcompete your planting

Weeds don't cheat. They simply arrive earlier, grow faster, and drop seeds that outlast your native inventory. The real problem is that most project budgets treat weed pressure as an afterthought, not the primary ecological force it actually is.

Think about the timing. You plant in April. By May, the pigweed is knee-high. By June, it has stolen the light, the moisture, and the nutrients your natives were counting on. Your planting is still alive—just starving.

What usually breaks first is the weeding schedule. The volunteer crew that promised to come back in three weeks doesn't materialize. The maintenance line gets cut in a budget review. The site manager gets pulled into another crisis. Nobody is deliberately abandoning the project; it just decays through neglect.

Every week of missed weeding in the first season costs two months of recovery later. That's the arithmetic nobody wants to hear.

— paraphrased from a restoration foreman in the Pacific Northwest

Hire someone for the second summer. Not the planting—that part is romantic. The weed pulling is what saves the project.

The maintenance budget that vanished

Here is the ugly truth about organizational behavior: rewilding gets funded as a capital project, then starved as an operating expense. The initial grant covers soil prep, plants, and installation. The recurring costs—weed control, spot replanting, watering during drought—have no champion in the next fiscal year.

I have seen this pattern repeat in municipal parks, corporate campuses, and HOA common areas. The project launches with a ribbon cutting and a nice sign. By year two, the sign is faded and the weeds have won. The maintenance crew, stretched thin already, treats the rewilding area as an unkempt mess, not a deliberate habitat.

The revert is not slow. It snaps back.

That's the part that surprises people. You don't lose a native planting gradually—you lose it all at once, usually after a wet spring that gives weeds a head start. The decision to mow it down comes from someone who never believed in the project in the first place, and they now have photographic evidence that the area looks "neglected."

Odd bit about activities: the dull step fails first.

Odd bit about activities: the dull step fails first.

Does that sound like a technical failure? It's an organizational one.

Fix this before you dig. Write the maintenance plan into the approval documents. Get a three-year commitment in writing. Confirm the budget line exists before you order a single seed.

The Long-Term Drift: When Soil Goes Bad Again

Compaction from Foot Traffic and Equipment

Two years in, the meadow looks settled. Then someone walks a skid steer across it to fix a fence, and the soil remembers. Pore space collapses, water pools on top, and the deep-rooted perennials that carried the project start yellowing from the roots up. I have seen entire rewilding strips reduced to plantain and knotweed because one repair crew parked on the wrong spot for an afternoon. The fix is not dramatic—it's boring, constant, and easy to skip.

Foot traffic does the same damage, just slower. A path worn by volunteers checking the site weekly becomes a hardpan ribbon. That hurts.

Keep the heavy machinery off the root zone. If you must drive on it, lay down wide-track mats or wait for frozen ground. And build paths people will actually use—defined, mulched, slightly inconvenient—so they don't improvise shortcuts through the planting beds.

Nutrient Depletion and Salt Buildup

Rewilded soil is not a closed loop. Every harvest of seed heads, every autumn leaf you rake off the path, every weed you pull and toss—all of that exports carbon and minerals. What comes back? Rain, mostly, which leaches what little soluble fertility remains. The soil leans progressively poorer, favoring tough, low-nutrient weeds over the species you wanted. The drift is quiet; the meadow just dulls, losing color and insect activity before you ever see a bare patch.

Salt is the sneakier culprit. De-icing runoff, dog urine near the entrance, even certain compost teas with high sodium—they accumulate in the top few centimeters, drying out soil structure long before they show as white crusts. You can't see it until the drought hits.

The mitigation is deliberate, but it's not heavy. Test the soil every second spring—DIY probes, cheap lab kits, whatever you can afford. Reapply a thin, low-salt compost (leaf mold, not manure) each fall. Mulch pathways with bark instead of gravel that sheds salts downhill. That's maintenance, not transformation—and it's enough.

How to Build a Monitoring Routine

Most rewilding projects fail the long game because monitoring feels like paperwork. It doesn't have to be. Set a phone reminder for the first Saturday of each month. Walk the same transect—ten minutes, no clipboard needed. Photograph the same four points from the same angles. That alone catches compaction early: grass will flatten, moss will spread, water will glint where it didn't before.

We didn't lose the site to a single disaster. We lost it to eleven small decisions we never logged.

— site steward, after three years of neglect

The routine should have one quantitative check: a simple penetrometer poke, or even a sharpened steel rod dropped from waist height. If it stops dead at the same depth twice in a row, you have a compaction layer forming. Dig one test pit per season—yes, it disturbs a square foot—and look at the root depth. Shallow, tangled roots mean the soil is sealing off.

Act when you see the first signal, not the third. Spread a half-inch of compost. Loosen aeration cores with a fork, but only in the affected spots—don't churn the whole site. Clear one drainage channel or silt trap. These are ten-minute responses that prevent the two-year redo.

What is the cost of skipping this? You re-ground the entire planting. You haul off the dead sod. You buy the seed again. Or you let it go and call it “succession,” which is a fancy word for giving up. Choose the boring route instead—weekly eyes, annual soil tests, and one afternoon of hand-work per season. That's the actual craft of urban rewilding. The planting was the fun part. The drift is where the discipline lives.

When Rewilding Is the Wrong Tool

Sites with Heavy Metal Contamination

Some soils are not gardens waiting to happen. They're hazards wearing a topsoil disguise. Old industrial lots, railyard edges, former auto-body shops—these places often carry lead, arsenic, or zinc at levels that make deep rooting a genuine health risk. Rewilding here means plants pulling contaminants up into stems and leaves that birds then eat. That’s not restoration. That’s a food chain delivery system for poison. I have walked sites where the city wanted wildflowers so badly, and the soil test came back with lead at four times the residential threshold. We walked away. Not every brownfield wants to be a meadow.

The right move is often containment, not cultivation. Cap the soil. Plant shallow-rooted natives in clean imported beds. Or skip plants entirely and let the ground rest under a geotextile membrane. That feels like failure to funders. It's not. The failure is pretending a contaminated lot can be healed by scattering seed.

Steep Slopes Where Erosion Will Win

Gravity has a head start on every slope over 30 degrees. Newly exposed soil, even with a native seed mix, will wash downhill before roots establish. The first hard rain proves it. You get rills, then gullies, then a bare scar that looks worse than the lawn you removed. The painful truth: some slopes need terracing, coir logs, or deep-rooted grasses installed as plugs—not seed—before any real diversity can hold. That's infrastructure work, not rewilding. Budget accordingly.

The catch is that terrace systems need maintenance for years. If your team can't commit to that, leave the slope alone. Mow it. Keep the turf. A boring green hillside beats a muddy collapse.

Wrong order. Many projects dig first and think about drainage second.

Spaces That Need to Stay Usable for People

Rewilding and human use can coexist, but not everywhere. A schoolyard edge where kids chase balls needs open sightlines, not a dense thicket of blackberry and dogwood. A community garden path needs to stay dry and passable, not become a self-seeded bramble tunnel. I get it—everyone wants the ecological win. But when the space hosts weekly farmers markets or doubles as a fire access route, you're designing for conflict.

That sounds fine until someone twists an ankle in a hidden root hole. Then the project gets ripped out within one season and the site goes back to turf. The compromise is a designed edge: tall grasses and forbs near the perimeter, short mown paths, and thornless pollinator shrubs in deliberate clusters. It's less “wild” than the vision. It survives contact with the public.

You don't get to choose what the soil wants. You only get to choose whether you listen before or after it fails.

— paraphrase of a restoration contractor I hired, who now refuses slope work without a drainage plan

Not every plot of land is waiting to be wild. Some are waiting for heavy equipment, clean fill, or just more time. The tactical question is not “can we rewild this?” It's “what does this soil need before it can hold life without risking the neighborhood?” Skip the diagnosis and you burn a budget, lose volunteers, and hand the development lobby a photo of a failed project. Run the test. Check the slope. Count the users. Then decide. I have walked off more projects than I have started, and every one of those exits saved years of regret.

Open Questions & What to Ask Before You Dig

How much soil remediation is ‘enough’?

That question almost always gets answered with a budget number, not a biological one. I have watched teams pour money into pH correction while the real problem sat two inches below—compaction that no amendment can fix. Soil testing gives you numbers, but numbers never tell you whether the site needed intervention in the first place.

Field note: friendly plans crack at handoff.

Field note: friendly plans crack at handoff.

Ask your team: what is the actual constraint? Are we fixing a past use, or are we just anxious about clay? The answer changes everything. A former parking lot with petroleum smells requires action. A yard with heavy loam mostly needs patience. Most rewilding failures start because someone treated the symptoms they could measure while ignoring the ones they couldn’t.

The practical threshold is usually visible. Roots that stop at a certain depth. Water that pools for two days after rain. Those signs matter more than a single lab reading.

We fixed one site by doing almost nothing—just broke up the top four inches and let the existing seed bank do the rest. The soil test said “poor.” The plants disagreed.

Can you rewild on a roof?

Technically yes. Realistically maybe. Roofs impose harsh conditions: shallow substrate, intense wind, no groundwater connection. What grows there won't match the meadow you imagined; it looks more like a gravel bar community, which can still be ecologically rich.

The catch is structural. Load-bearing limits typically cap substrate depth at four to six inches, which restricts what roots can reach. That means you're managing a container garden with a wildness aesthetic, not rewilding in the strict sense.

Ask about drainage aggressively. Flat roofs fail when fine particles wash into channels and clog outlets. One blocked drain turns your experiment into a legal liability faster than any ecological debate.

Sedums barely need soil. Grasses struggle. If your goal is habitat for pollinators, you might get there—if you accept that the species palette will be tiny and tough.

What does a soil test actually cost?

Standard tests run from fifty to two hundred dollars per sample, depending on whether you want heavy metals, organic matter, or nematode counts. That's nothing compared to the money lost on wrong amendments. I’ve seen one project burn six figures on fertilizer that made things worse, because nobody checked the phosphorus levels first.

Budget for multiple samples. One core per site tells you almost nothing; soils shift in patches. We recommend a grid of at least five, spaced to catch the variation your eyes can’t see.

Cheaper option: the jar test. Put a scoop of soil in water, shake it, watch the layers settle. Sand falls fast, silt next, clay last. That gives you texture without any lab—good enough for basic drainage decisions.

Remember the test is a snapshot, not a promise. A dry summer sample differs from a wet spring one. The real soil changes with weather, compaction, and the roots you’re trying to establish.

“You're not fixing the soil; you're entering a relationship with it. And like most relationships, it fails when you assume instead of ask.”

— paraphrase from a restoration ecologist I spoke with, on why she insists clients walk the site before they dig

Your questions should also touch timing. Test in the season when you’ll actually plant, not six months early. The soil biology shifts that much. Nobody warns you about that on the report.

Digging is the cheap part. The expensive mistakes live in what you assume afterward. So ask the awkward question now: how will we know if this worked in three years? If your team can’t answer that, the project probably won’t survive the first excuse to mow it down.

Field Notes: What to Try Next

Start with a one-meter-square test plot

Pick a corner of your site that’s already failing. Not the pretty patch—the bare, compacted strip where nothing survives. Mark out a single square meter and treat it like a laboratory. No heavy machinery, no soil import. Just you, a spade, and a notebook. Loosen the top four inches by hand. Drop in a mix of native seeds and a few plugs. Then leave it alone.

That’s it. The whole experiment costs an afternoon.

Watch what happens when you stop intervening. You’ll see which species self-seed, which get eaten, which collapse in the first dry week. The data you gather here beats any textbook. Most teams skip this step because they think small plots waste time. The catch is—scaling up a flawed design just scales up the failure. A one-meter square gives you permission to be wrong cheaply.

Join a local soil health network

Rewilding is stubbornly local. The soil in your city’s floodplain behaves nothing like the stuff on the ridge two miles away. Find the people who already know this. Local conservation districts, community garden coalitions, even the old-timer who grows vegetables down the road—they’ve watched soils shift over decades.

Bring them your test plot results. Ask what they’d try next. You’ll get conflicting advice, and that’s fine. Conflicting advice means someone’s actually paying attention.

Worth flagging—these networks often have access to free compost, mulch, or native seed that you’d otherwise buy at retail. The relationships pay for themselves. But don’t join just to extract resources. Show up with observations. Offer to dig. The exchange works better when it’s mutual.

“The soil doesn’t care about your timeline. It cares about consistency, disturbance, and time.”

— noted from a restoration ecologist at a working bee, not a formal study

Measure, adjust, repeat

Choose three metrics you can actually track. Maybe it’s bare soil exposure, earthworm counts per shovel, or how long water ponds after rain. Take a baseline before you touch anything. Then measure monthly, no exceptions. That’s the part that keeps projects honest.

After your first season, compare what changed. The numbers might look flat—that’s data too. It tells you the system needs more time or a different approach. Adjust one variable at a time. Add mulch. Change the planting density. Switch to a different species mix.

The trap here is over-measuring. I’ve seen people spend more hours logging spreadsheet rows than tending actual soil. Keep it simple. One notebook page per month per plot.

Finally, schedule a re-check at the two-year mark. That’s when most rewilding attempts either settle into a working ecosystem or quietly revert to weeds. Your notes from the first season will tell you which path you’re on. Trust them over your memory. Then start another plot somewhere harder.