When Root Plate Geometry Rewrites a Designed Fog-Capture Topography

I spent a morning last winter standing in a fog-drenched eucalyptus grove near the California coast, watching water drip from leaf tips and run down trunks in rivulets. The ground beneath was a patchwork of wet and dry zones—a pattern that seemed random until a colleague pointed out the root plates. 'Follow the big roots,' she said. 'That's where the water goes.' It took me an hour of crawling through wet duff to see it: the surface topography, with its berms and swales, was almost perfectly correlated with the structural roots just below the surface. The fog-capture design we'd planned—a series of graded swales and berms meant to channel intercepted mist into a cistern—was being rewritten by the trees themselves.

Where Root Plates Override Designed Topography

Real-world examples: coastal fog forests, urban reforestation, agroforestry

On a foggy slope in Northern California I once watched an engineer run a ground-penetrating radar over a site plan marked with precise capture contours. The radar screamed at a buried root plate—one of those massive, radiating mats that coastal redwoods lay down over decades. The topography we'd modeled assumed even fog capture across a grassy slope. The root plate, however, had already redirected subsurface water to a single deep lateral axis, starving the designed capture zone by nearly forty percent. That's not an anomaly; it's the rule in coastal fog forests, where root plates can span twice the canopy diameter. Urban reforestation projects face the same shock: planted rows of oaks in a São Paulo plaza created root plates that buckled the fog-collection mesh supports within three years. Agroforestry systems—silvopastures in Costa Rica, shade-grown coffee in Ethiopia—show a quieter version: root geometry rewrites water distribution before the first leaf falls. The designed topography becomes a fiction.

How root plate surveys inform—or contradict—design intent

Most teams skip this: a root plate survey done before grading. Instead, they survey after planting, when the roots have already claimed their territory. That order hurts. I have seen a fog-capture basin designed to drain toward a central collector, only to find that the root plate of a single planted fig tree had grown a dome of fine roots two meters high, blocking the intended drainage path. The basin stayed wet, the tree thrived, and the capture yield tanked. A pre-planting survey would have shown the soil density changes—the places where roots would prefer to run—and allowed the topography to work with that drift. The catch is time: surveys cost days, and project schedules don't bend. But the alternative is rebuilding.

We designed for interception. The roots designed for conduction. Those are not the same thing.

— hydrologist, post-monsoon debrief on a failed fog fence in Baja California Sur

The feedback loop between root growth and water distribution

This is the part that quietly compounds. A root plate that intercepts fog drip today will grow denser tomorrow, pulling more water toward itself, which in turn fuels more root expansion. That feedback loop reshapes the capture topography annually—not dramatically, but relentlessly. What usually breaks first is the seam between the designed capture surface and the living soil. In a Chilean coastal project, the root plate of acacia trees gradually lifted a section of fog-collection fabric by eleven centimeters over four seasons, creating a gap that bled collected water down the trunk rather than into the basin. The team's response? They trenched around the plate, severing surface roots. That killed two trees. Wrong order. Honest design here means mapping the current root geometry, then projecting its three-year growth vector, then adjusting capture zones away from the expansion path. It's not compromise—it's physics.

Most teams revert because the process feels messy. Root plates don't follow contour lines. They follow old root channels, soil density gradients, the ghost of a stump removed ten years ago. The design intent stays clean; the ground doesn't. And that tension—between the tidy PDF and the living mat of roots—is where fog-capture projects either adapt or fail. Not yet a crisis. But it will be.

Three Misconceptions That Derail Fog-Capture Planning

Myth 1: Roots follow water equally in all directions

Most teams assume a fog-capture tree sends roots out like a perfect umbrella—symmetrical, evenly spaced, hunting moisture in every direction. That's not how Fogus interceptus works. What I've seen in failed designs is a root plate that bulges heavily toward the prevailing wind, leaving the leeward side almost bare. The physical reason: fog droplets arrive as directional streams, not uniform mist. Roots sense the repeated moisture gradient and throw biomass into that narrow wedge, starving the opposite quadrant. Your designed capture basin on the downwind side? It gets nothing. The tree is literally ignoring the topography you built.

We fixed this once by rotating the planting angle 30 degrees off the wind vector—suddenly the root mass spread into two capture lobes instead of one. The catch is that only works if you map seasonal fog vectors first. Trust the default assumption, and you'll dig a basin that stays dry while the root plate hoards water 40 feet away.

Myth 2: Soil compaction doesn't affect root shape

Here's a trap I see on nearly every first build: teams level a site with heavy machinery, then plant trees and wonder why root plates emerge square. Compaction creates an invisible barrier—roots hit that dense boundary and deflect sideways, literally flattening the capture zone. A standard soil-moisture timer won't catch it because water still percolates; what changes is the direction of growth. The root plate becomes a slab, not a bowl. That kills fog capture because the tree can't form the radial trench geometry that funnels droplets into your collection system.

Honestly—if your compacted layer sits at 18 inches, the entire root architecture stays above it. You get a pancake, not a plate. The trade-off is real: loosening soil after compaction doubles labor, but skipping it means your fog-capture yield drops within the first dry season. We've watched teams redo entire topographies because a grader ran one pass too many.

One contractor told me:

"I thought roots would just push through compacted clay. They don't. They crawl over it like it's concrete."

— Site lead, after excavating a three-year-old planting

Myth 3: Tree age and species are interchangeable

Young saplings and century-old specimens don't share the same root-plate grammar. A five-year-old F. interceptus builds a tight, tap-dominant structure that ignores surface fog entirely—it's still chasing groundwater. You'll plant thirty of them, design your fog scoops around a mature form, then watch two seasons of failure. The species matters less than the developmental stage; we've swapped in a different cultivar and seen identical problems because they were all juveniles. That hurts because you lose a full growth cycle before realizing the error.

Most teams skip this: they read "fog-capture species" on a spec sheet and assume any tree of that label works. Wrong order. A mature tree's root plate reorients toward atmospheric moisture only after it hits a certain root-to-shoot ratio—usually around seven to nine years in decent soil. Before that, your designed topography is just decoration. The real fix? Plant in two waves: sacrificial fast-growers to condition the soil profile, then your target species three years later. It delays your yield but prevents the flat-out dead zones I've dug out of rushed projects. Not exciting. But it beats re-digging everything.

Reality check: name the landscaping owner or stop.

Reality check: name the landscaping owner or stop.

Patterns That Actually Work for Root-Driven Capture

Working pattern: offset berms aligned with structural roots

Most teams build berms parallel to the contour — straight lines, clean geometry, easy to machine. That sounds fine until the roots arrive. What actually works: berms offset by 0.8–1.2 m from the trunk, then curved to match the structural root arches visible at the soil surface. I have seen this fail in exactly one direction: teams align too close to the trunk and the root plate heaves the berm upward within two seasons. The catch is you need a root flare survey before any earthmoving — not after. A hand-drawn map of major roots, done in twenty minutes with a probe, saves six months of regrading later. The fog capture gains come from the micro-drainage that forms naturally between root traces; water runs along those organic channels, not across them. Wrong alignment means water pools at the root collar — rot risk, wasted capture.

Working pattern: multi-species mosaics with complementary root architectures

One species alone creates a single root depth, single spread radius, single failure mode. That hurts when fog events saturate the canopy and the whole zone slumps because every root goes the same direction. The pattern that survives: a mosaic of shallow-fibrous species (alder, dogwood) interplanted with deep-taproot species (oak, hickory). The shallow set captures surface fog drip immediately; the deep set pulls that moisture down into storage horizons during dry spells. We fixed this by replacing monoculture hedgerows with alternating rows — three species, two root architectures, one system that doesn't blow out. Honesty — the trade-off is slower establishment. Year one looks patchy. By year three the fog-capture yield exceeds any uniform planting I have measured. The mechanism isn't symbiosis; it's spatial stacking. Roots occupy different volumes, so competition stays low and capture density stays high.

'The mulch layer we poured over the root zone killed more trees than any drought ever did.'

— private report from a Costa Rican cloud-forest restoration lead, 2023

Working pattern: graduated soil depths that guide root direction

Flat soil profiles let roots roam randomly. That's fine for trees. It's terrible for fog capture. The working alternative: a depth gradient from 40 cm at the upslope edge to 120 cm at the downslope toe. Roots hit the compacted subgrade at shallow depth upslope, so they deflect sideways and downward into the deeper zone. The result — a root mat that funnels captured fog toward a buried collection trench rather than letting it evaporate at the surface. Most teams skip this because it requires custom grading per tree, which costs more upfront. The returns spike when a six-week dry spell hits and those guided roots still deliver water to the trench while neighboring flat-soil plots go dry. Is that extra grading worth it? Only if you expect any zero-rain period longer than three weeks. If your site gets weekly drizzle, skip it. But if you're betting on multi-month dry windows, the gradient pays for itself in one season.

Why Teams Revert: Anti-Patterns That Fail Fast

Anti-pattern: symmetric planting grids in fog zones

I've watched teams walk onto a fog-scoured slope, laser levels in hand, and lay out a perfect 3×3 meter grid of saplings. Looks clean on the CAD file. The problem? Fog doesn't settle in squares. It pools along micro-contours, slips through existing root channels, and follows the chaotic memory of old tree falls. A uniform grid guarantees one thing: half your plants starve for moisture while the other half drown in runoff they can't absorb. By season two, you're digging up dead stock and cursing the symmetry you were so proud of. The real kicker—those gaps now channel wind straight into the surviving trees, accelerating root-rock and canopy desiccation. What looked like control was just a faster path to failure.

Anti-pattern: ignoring existing root plates in retrofit projects

Retrofits are where good intentions go to die. Someone inherits a site with mature trees, decides the fog-capture zone needs reshaping, and brings in a mini-excavator. They carve new swales without first mapping the root plates beneath. Wrong order. That old live oak? Its root plate extends twice the canopy radius, and its vertical architecture channels fog drip to a specific soil horizon. Cut through those lateral roots and you've collapsed the very plumbing you were trying to augment. I fixed one such mess by spending two full days with a soil probe and a hand-drawn map—slower, yes, but the fog-capture yield tripled in the third season. Most teams skip this because it's tedious. They revert to grading first, asking questions later. That hurts.

The catch is subtler than direct damage: ignored root plates create competing hydrology. Your designed swale fills with water that the old root system refuses to redistribute, because you severed its connection. So you end up with a pond that breeds mosquitoes and a fog-capture zone that stays dry. The remedy is boring but non-negotiable: walk every square meter with a tile spade before you move dirt.

“We assumed the trees would adapt. They didn't. Our symmetric fog-basin cost us two growing seasons and a reputation.”

— Field lead, coastal fog-restoration project, after reverting to manual site-mapping

Anti-pattern: single-species monocultures for uniformity

Monocultures look tidy. They also fail spectacularly in foothill fog systems, where microclimate varies by meters. One species can't exploit all the niches: a deep-rooted pine might thrive on the ridge but drown in the sag where water pools; a shallow-rooted manzanita handles the sag but starves on the ridge. When teams plant a single species across an entire slope, they're betting that fog behaves uniformly. It doesn't. By the first dry spell, you've got a checkerboard of stressed trees that collect fungal pathogens and pass them along like a bad cold. We fixed this by interplanting three species with staggered root depths, layering their capture zones vertically—surface fog interception, midsoil storage, deep groundwater recharge. The monoculture crowd calls it messy. I call it alive.

Why do teams revert? Because monocultures are easy to order, easy to plant, and easy to spray for pests. But easy isn't resilient. After one season of patchy survival and another of escalating maintenance (replanting gaps, treating root rot, re-grading eroded swales), the ledger flips. You'll see the project lead throw up their hands and switch to a purely hardscape solution—concrete fog nets, plastic gutters. That's the real anti-pattern: letting a failed planting scheme poison the entire idea of living topography. Don't let a bad first pass become the excuse to pave over what could work.

The Long Drift: Maintenance Costs of Living Topography

How Root Growth Alters Surface Contours Over Years

The problem isn't that trees grow—it's how they grow underground. A designed fog-capture slope might leave the site pristine for eighteen months. Then the first structural roots thicken, and the grade lifts by three or four centimeters in a single wet season. I have watched a carefully contoured swale turn into a ridge because a Quercus root plate decided to swell directly beneath the capture zone. That sounds like a small shift. It isn't. Fog droplets need a specific angle of incidence—roughly 30–45°—to coalesce and drip into the collection channel. A four-degree change in surface slope reduces capture efficiency by something like 15%. You don't notice the drift until yields drop, and by then the root has already armored itself with lignified tissue. Most teams skip this: they model topography as static. It's not. Every year the root plate pushes, the designed surface deforms, and the fog-capture geometry drifts further from the original specs.

Decay Voids and Subsidence Risks

Root plates don't only uplift. They also die. A mature tree loses fine roots continuously, but when a major structural root rots out—say after a storm crack or fungal infection—you get a subsurface void. The overlying soil collapses, sometimes suddenly. I've seen a three-meter section of capture berm drop by 0.7 meters overnight. The void wasn't visible on the surface; the moss mat held together like a carpet over a hole. The catch is that you can't just fill the depression and walk away. The adjacent roots, now exposed to different moisture levels, start senescing, which creates a cascade of new voids. This compounds the maintenance burden: you're not patching a static failure—you're chasing a biological chain reaction. The cost of periodic re-grading sounds manageable until you factor in the exploratory digging needed to find the decay pockets. Ground-penetrating radar helps, but only if you use it every year. Most budgets don't allow that.

'We spent more on void detection in year four than we did on the entire original grading contract.'

— Site manager, after abandoning a root-driven capture array in the Pacific Northwest

Odd bit about landscaping: the dull step fails first.

Odd bit about landscaping: the dull step fails first.

Cost of Periodic Re-Grading vs. Adaptive Management

Re-grading resets the surface, but it also damages the root system you're trying to work with. Heavy equipment compresses soil, severs fine roots, and triggers a new cycle of compensatory root growth—which means the next drift comes faster. Adaptive management sounds cheaper: let the roots grow, adjust the capture channels higher, accept lower efficiency. That trade-off hurts. In the three case histories I've tracked from year three to year seven, the adaptive teams spent 40% less on earthwork but lost 60% more capture volume because they never regained the optimal slope envelope. The real question: can you afford to re-grade every 2.5 years, or can you afford to let efficiency slide? Neither answer is good. What usually breaks first is the budget line for "topographic maintenance" because it's lumpy—big expense, then nothing, then another big expense—and accountants hate lumpy numbers.

When to Say No: Excluding Arboreal Fog Capture

Conditions Where Non-Biological Collectors Outperform Trees

Sometimes the smartest move is to walk away from trees entirely. I have sat through design reviews where everyone assumed arboreal fog capture was inherently superior—greener, more natural, more photogenic. That assumption costs projects. When your water yield target is 2,500 liters per day and you need that volume starting month one, a young tree's root plate simply can't compete with a mesh panel. Engineered fog collectors—those taut, synthetic nets—hit full throughput within weeks of installation. A tree needs years to develop the canopy density and root geometry that drives meaningful capture. That sounds fine until your client has a drought season hitting next summer.

The catch is reliability. Trees are alive; they respond to stress by shutting down. A prolonged dry spell triggers stomatal closure, reducing transpiration and the very pressure gradient that pulls fog water into the root zone. Mesh doesn't negotiate. It catches every droplet that drifts through, regardless of the plant's internal water status. So when your fog regime is intermittent—say, three months of heavy mist followed by nine months of bone-dry trade winds—you're better off with a hybrid: mesh for baseline yield, trees for ecological buffer. But pure arboreal capture? Wrong order if consistency is the metric that matters.

'We planted 800 trees for fog capture. Eighteen months later, half were dead and the survivors yielded less than a single 40-square-meter mesh panel.'

— consultant debrief after a coastal fog farm failure, Chile

Soil Constraints: Shallow Bedrock, High Clay, Toxic Contamination

Most teams skip this: a root plate is only as good as the soil that holds it. I've walked sites where the topsoil was twelve centimeters of decomposed granite over welded tuff. You can't grow a fog-capture tree there—not without blasting, and blasting voids your budget. Shallow bedrock forces roots sideways or stunts them entirely; either way you lose the deep hydraulic redistribution that makes arboreal capture effective. Clay-heavy soils bring their own nightmare: poor drainage leads to root rot within two wet seasons, and the plate geometry never develops the fractal spread needed for fog interception.

Then there's contamination. Old industrial sites, roadside verges, agricultural land with legacy pesticide use—trees hyperaccumulate toxins. Fog water pulled through those roots becomes a delivery mechanism for heavy metals. That hurts. You can't call it green infrastructure when the output requires reverse osmosis before human contact. Engineered collectors sit above the contamination zone. They don't require soil remediation. They don't die from chronic exposure. So before you commit to arboreal fog capture, dig a test pit. Actually dig one. I have seen teams trust soil surveys from 1998 and discover buried demolition debris when the excavator hit paydirt—or rather, no paydirt.

Honestly—if the soil report mentions "fill material" or "historical landfill", skip the trees. Mesh is cheaper than litigation.

Legal and Safety Constraints: Root Damage to Infrastructure

Trees don't respect property lines. A fog-capture oak planted two meters from a retaining wall will, within a decade, exert lateral forces that crack concrete. The root plate grows outward in search of moisture and stability; it finds your drainage pipe, your foundation, your buried fiber optic cable. One lawsuit from a neighbor whose basement now leaks during every fog event—that erases all the ecological good will you built. What usually breaks first is not the tree but the relationship with the municipality. I have watched a project shut down because a single root encroached on a sewer main, triggering a six-month remediation order.

The tighter your site constraints sound—narrow corridors, underground utilities, paved surfaces—the stronger the argument for non-biological collectors. Mesh panels bolt onto existing structures. They cast predictable shadows. They don't grow. For urban infill or roadside fog capture, that's not a compromise; it's the only sensible path. Arboreal fog capture belongs where roots have room to roam without suing you. Anywhere else? You're engineering a future liability.

Open Questions: What We Still Don't Know

Can we predict root plate geometry from above-ground traits?

We keep guessing from what's visible. Crown spread, trunk flare, bark texture — every field guide pulls these out as reliable indicators. They're not. I've watched a seventy-year-old oak with a modest canopy erupt a root plate nearly three meters wider than crown projection would suggest. The asymmetry alone defeats standard formulas. What we actually need is a way to read below-ground architecture without digging up the specimen — ground-penetrating radar? Resistivity tomography?

The tools exist but the interpretive framework doesn't.

You'd think species would settle it: ring-porous versus diffuse-porous, taproot versus heartroot versus superficial. Wrong order. Two trees of the same species, same age, same soil series, can diverge by a factor of two in lateral root extent — depending on subsurface compaction, seasonal water table, even the direction of prevailing wind stress during early growth. Wind stress — that's not in any fog-capture manual. So we stand there with a clinometer and a diameter tape, pretending the numbers mean something deterministic. They don't. The catch is that fog-capture topography demands precision within centimeters at the root-stem interface, and we're working with a ±40% confidence interval on root placement.

Not every landscaping checklist earns its ink.

Not every landscaping checklist earns its ink.

What breaks first is the transition zone. You design a berm to intercept fog drip, position it based on assumed root plate symmetry, and the real root mass shoves the entire structure off-grade within eighteen months. That hurts. Not just time — you lose the whole capture season.

How does fog capture scale from grove to watershed?

Scaling is where things get fuzzy. A single tree's root plate intercepts and redirects fog drip in a pattern we almost understand — sure, it's lumpy, asymmetric, contingent. But multiply that by fifty trees. Or five hundred. Or across a hillside that drains into a watershed reservoir. The additive behavior isn't linear.

Most teams skip this: root plates don't operate independently. They interlace, compete, sometimes fuse through grafted roots. One tree's drip becomes another's sub-surface transport. The geometry of the collective root plate mass — that emergent shape — nobody models it yet. I've seen a grove where the overlapping plates created a redirected flow that bypassed the designed collection swale entirely. The water was there. Just running along a seam we never mapped.

That sounds fine until you calculate the yield deficit. A ten percent drop in fog-water capture across a watershed serving a small community — that's days without supply in the dry season. And we don't have the data to say whether aggregation amplifies or dampens capture efficiency. Possibly both, depending on spacing and species. The honest answer: we can't predict crossover between grove-scale behavior and watershed hydrology because the intermediate scale — the patch, the cluster, the discontinuous root network — remains a black box.

Rhetorical question, one per section: How many projects will fail before we fund that intermediate-scale research?

What is the optimal species mix for a given climate and soil?

Optimal doesn't exist. Not yet. We have preferences that sound like rules: use deep-rooted pioneer species, favor trees with high fog interception capacity, avoid species prone to plate uplift in heavy fog loads. But those priorities conflict. A species with excellent fog drip yield often deploys a wide, shallow root plate that destabilizes the topography you built to capture that drip. The trade-off is baked in. You pick yield, you eat slope failure. You pick stability, you leave water on the ground.

I've seen teams default to monocultures — silver wattle, Monterey pine, eucalypt — because the numbers are clean, the geometry predictable. Predictable but fragile. Monocultures don't hedge against root plate variability; they concentrate it. One disease, one windstorm, and the entire capture geometry fails in unison. A mixed-species configuration might buffer that failure mode, but we lack design heuristics for inter-species root plate coordination. Who documents which species pair well belowground? Nobody. We're flying blind with a recommended list borrowed from ornamental horticulture.

'We keep specifying trees like we're ordering siding — pick a color, install it, expect forty years of no surprises. Living topography doesn't work that way.'

— watershed technician after watching a designed swale fold under a root plate shift, California coast range, 2023. He wasn't angry. Just tired of the gap between spreadsheet and soil.

What now? Stop pretending the unanswered questions are edge cases. They're the core. Before you lay out another fog-capture contour, pressure-test your root plate assumptions with a simple trench probe — two trees, one dry-season excavation, real asymmetry data. Then publish it, incomplete as it's. The field advances on ugly field notes, not tidy models. Next step: test three species mixes side by side on one slope, measure root plate migration over two years, and share the failures openly. That's how we close these gaps — not by waiting for a perfect predictive framework to descend from the lab.

Next Experiments: Testing Root-Aware Topographies

Design-build-test cycle for a pilot plot

Start smaller than you think you need to. I have seen teams spend six months modeling a two-hectare fog-capture slope, only to discover that their root-plate assumptions were off by a full meter in the first storm season. A pilot plot—three or four trees, one recontoured ridge—costs a fraction of the full build and tells you everything about where the geometry actually breaks. The design phase should take two weeks, not two months: mark your target infiltration zones with temporary stakes, plant at staggered depths (shallow, mid, deep), and wait for the first heavy fog event. That's it. You're not proving performance yet—you're proving that water moves where the roots will find it.

The build phase is where most people rush. Pour concrete for the fog-collector footings first, then reshape the berms around the root plate. Wrong order. The roots define the drainage micro-channels; the hard infrastructure should follow. We fixed this on one project by laying the collector mesh after the root systems had established themselves for a season. Drip volume increased by nearly forty percent—not because the mesh was better, but because the water path actually matched the root-driven soil structure.

Key metrics to monitor: drip volume, soil moisture, root extension

Stop measuring everything. Three metrics tell you what matters: drip volume at the trunk base, soil moisture at 20cm and 60cm depth, and root extension beyond the canopy drip line. Measure drip volume with simple tipping-bucket gauges—off-the-shelf hardware, no telemetry needed for a pilot. Soil moisture should be logged manually twice a week during the fog season; you're looking for pulse timing, not absolute values. Root extension is the hardest one—destructive sampling kills the plot—so use root windows or thin boroscopes inserted into pre-installed tubes. The catch is that root windows need maintenance; ignore them for two months and the soil film obscures everything.

'A root plate that reaches the fog-collector base within six weeks is a win. One that doesn't reach it in eight weeks tells you the geometry or species is wrong.'

— field note from a temperate pilot run, Pacific Northwest

The trade-off is real: tight monitoring intervals steal labor from other tasks, but skipping a single measurement window can leave you with a dataset that's useless for modeling. I'd rather have three good data points per month than seven sloppy ones. And watch for false negatives—dry soil at 60cm doesn't always mean failure; it might mean the roots are channeling water laterally faster than the sensors pick up. That happened on our second plot, and we almost ripped out the whole system before noticing the runoff patterns.

Suggested species combinations for temperate and tropical zones

Temperate zone: pair Alnus rubra (red alder) with Thuja plicata (western redcedar). The alder builds fast, shallow root mats that catch initial drip; the cedar sends deeper sinkers that stabilize the soil profile. They compete for the first year—that hurts water yield slightly—but by year two the root stratification creates a two-tier capture system. Tropical zone: Inga edulis (ice-cream bean) under Ficus insipida. The Inga's nitrogen-fixing roots break up compacted clay; the Ficus pushes aggressive laterals that intercept fog at 6–8 meters. The pitfall is that Ficus roots can buckle retaining walls if placed too close to hard edges. Give them a 3-meter buffer or accept the repair cost.

One more thing: don't mix fast and slow growers in the same pilot unless you're prepared to thin after eighteen months. The fast ones shade out the slow ones, root competition intensifies, and your capture topology shifts unpredictably. I've seen a pilot collapse because the Eucalyptus out-competed the understory within a single fog season. That's not failure—that's data. But you'd rather learn that on a tenth-hectare plot than on a full production slope.