Soil preparation and amendment field guide for planting beds and lawns

Soil prep is the job, not the plant

Soil preparation is everything you do to the ground before a plant or a blade of grass goes into it: test it, correct the pH, work in organic matter, break up the compaction, and get enough good soil over a loosened subgrade that roots can grow down. That is the job. The plant is the easy part.

Here is the part owners and a lot of installers never quite believe. The nursery already grew you a healthy plant. Whether it lives is decided almost entirely by the soil you set it in, and that decision is made before the plant is ever in your hands. Money spent on the prep returns more than money spent on bigger plant material, every time. A small plant in good soil passes a big plant in bad soil inside two seasons, and the big plant in bad soil is often dead by then.

This guide covers the soil work. The planting itself, the depth, the hole, the root flare, the first-season water, lives in the tree and shrub planting guide, and the grading that decides where water goes lives in the drainage and grading guide. Read those alongside this one. The soil prep, the planting, and the grade are one job split three ways, and the seam between them is where plants die.

What does a soil test tell you, and why start there?

A soil test is a lab analysis of a sample from your site that reports pH, the plant-available nutrients, the organic matter, and usually the texture, along with recommendations for what to add. Start here. Amending soil without a test is guessing with a checkbook, and the most expensive guesses are the ones that look like progress.

The standard panel reads pH, phosphorus, potassium, calcium, magnesium, and the micronutrients, plus organic matter as a percentage. A healthy mineral soil sits at 2 percent organic matter or better. The pH number is the one that controls the rest, because it decides how much of the nutrients already in the soil the plant can actually take up. Phosphorus, for example, locks up with iron and aluminum below about 6.0 and with calcium above about 7.3, so a soil can be full of phosphorus the plant cannot reach. Fertilizer does not fix that. pH does.

Get the test from the local cooperative extension lab or a reputable soil lab. It is cheap, often under thirty dollars, and it comes back with a lime or sulfur recommendation calibrated to your soil, not a generic rate off a bag. Sample correctly: pull a dozen plugs from across the area to the depth you will work, six to eight inches for beds and lawns, mix them in a clean bucket, and send the composite. One scoop from one spot tells you about one spot. Test every 3 to 5 years, and always before a major install.

Soil texture and structure

Texture is the ratio of sand, silt, and clay in the soil, and it drives nearly everything downstream: how the soil drains, how it holds water and nutrients, how it compacts, and what it takes to amend it. You do not change texture in any practical way on a jobsite. You work with what you have and you fix structure, which is a different thing.

Sand is the big gritty particle that drains fast and holds little. Clay is the microscopic flat particle that holds water and nutrients but packs tight and drains slowly. Silt is in between. Loam is the blend everyone wants, enough sand to drain and enough clay to hold, and it is rare in the ground and common in a topsoil order. You can read texture in the field two ways. The jar test: a soil sample in a jar of water, shaken and left to settle, layers out as sand on the bottom, silt in the middle, clay on top, and you read the proportions. Faster is the ribbon test: wet a pinch and squeeze it out between thumb and finger. Sandy soil will not ribbon and feels gritty. A long, smooth ribbon that holds together is high clay. A short ribbon that breaks is loam.

Structure is how those particles clump into crumbs and aggregates, and unlike texture you can wreck it or build it. Good structure means pore space, the gaps that hold air and let water and roots move. Compaction crushes the pores. Working soil wet smears them shut. Organic matter and roots and soil life build them back. The whole point of the prep is structure, not texture.

What pH should soil be, and how do you change it?

Most plants do best between about 6.0 and 7.0, where the major nutrients stay available and the soil biology works. Vegetables and most cool-season lawns lean toward 6.5. The exceptions are the acid lovers, blueberries, azaleas, rhododendrons, mountain laurel, hollies, which want 4.5 to 5.5 and will yellow and sulk at neutral. Match the pH to what you are planting, not to a single number.

To raise pH, you add lime. To lower it, you add elemental sulfur, which soil microbes convert to acid over time. The rate is where people go wrong, and it is exactly why the soil test comes first. pH alone does not tell you how much lime a soil needs, because the soil's buffering, how hard it resists a pH change, varies with texture and organic matter. A clay soil needs far more lime to move the same amount as a sandy one. The lab runs a buffer test and gives you the actual rate. Guess it and you either waste the material or overshoot.

The change is slow, and that is the part to plan around. Lime and sulfur both work over months, not days, as they react and as rain carries them down. Sulfur to drop a full pH unit can run on the order of several hundred pounds per acre and take a season or more to fully express. So you test early, apply ahead of planting, and work the amendment into the soil rather than leaving it on top, because lime in particular barely moves on its own. Do not dump a heap of lime in a planting hole expecting a fast fix. It does not work that way.

How much compost should you mix in?

Compost and organic matter are the one amendment that helps almost every soil, and the common rate is 2 to 3 inches spread over the area and worked into the top 6 to 8 inches, which lands the finished blend near 20 to 30 percent compost by volume. That is the workhorse number for a new bed. Do not exceed about 4 inches incorporated at once. A bed that is half compost slumps as it breaks down and stays too rich.

What the organic matter does depends on the soil it goes into, and it cuts both ways, which is the trick of it. In clay, it pries the fine particles apart into crumbs that drain and breathe. In sand, it acts like a sponge that holds water and nutrients the sand would let run straight through. Same amendment, opposite problems, because what it really does is build structure and feed the soil life that holds that structure together. It also feeds the worms, fungi, and bacteria whose castings and glues are what aggregate the soil in the first place.

Quality matters more than people check. Use compost that is finished, dark, crumbly, and smells like earth, not sour or like ammonia, which means it is still cooking and will rob nitrogen and burn roots as it finishes in your bed. On commercial work, the spec often calls for compost meeting the US Composting Council Seal of Testing Assurance, which gives you lab data on maturity, stability, and contaminants instead of a guess. Cheap, raw, weed-seed-loaded compost is not a bargain. You will weed it for two years.

Work it in, do not layer it. Compost raked on top and left there does little for the root zone below. Till or fork it through the full 6 to 8 inches so the structure improvement reaches where the roots actually grow. For upkeep, an inch a year topdressed onto an established bed maintains the organic matter as it burns off, and that is the cheap habit that keeps a good soil good.

Loosening compacted soil

Compaction is the quiet killer on most new sites, and roots do not negotiate with it. Compacted soil has no pore space, so it holds no air, sheds water instead of soaking it, and is physically too dense for roots to push through. Plants in compacted ground sit in what amounts to a pot with no drainage hole, and they decline slowly enough that everyone blames the plant.

Loosen it before you plant, and match the tool to the area. A garden fork or a broadfork works a bed by hand, and a broadfork reaches 12 to 18 inches without flipping the soil over, which loosens deep while leaving the structure layers in place. For larger areas and serious subgrade compaction, you rip or subsoil with a tractor-mounted shank that fractures the hardpan. The point of all of it is the same: open the pore space so roots, water, and air can move.

Two cautions decide whether decompaction helps or hurts. Do not work soil wet. Tilling or ripping wet soil smears the pores closed and, on clay, turns it into clods that dry into bricks, so you have spent money making the compaction worse. Wait for moist, not wet. And do not over-till. Pulverizing soil to a fine powder with repeated tiller passes destroys the very structure you are trying to build, leaves it prone to crusting and re-compacting, and burns through organic matter. A couple of passes to incorporate amendments and break the layer is the job. More than that is damage. One more honest point: decompaction alone is short-lived. Loosened soil with nothing else done slumps and re-compacts within a season or two, so you decompact and add organic matter together, never decompaction by itself.

Construction-damaged soil

Soil on a finished construction site is rarely soil in any useful sense. Grading cut the native topsoil off and hauled it away, equipment tracked the subgrade to the density of a parking lot, and what is left at finish grade is compacted subsoil, fill, and whatever got buried. Then a thin skin of topsoil gets spread for looks and the landscape goes in on top to fail. This is the most common bad soil a landscaper inherits, and treating it like normal ground is the mistake.

You have two honest paths. Amend in place, which means decompacting the damaged profile and working organic matter through it to rebuild structure, or import and rebuild, which means bringing in topsoil over a properly decompacted subgrade. On a heavily compacted site you usually need both: rip the subgrade first, then place the topsoil and blend the interface. The order matters. Topsoil placed over a compacted subgrade you did not loosen is just a layer waiting to perch water.

The subgrade is the part everyone skips because it is buried and invisible. For real restoration the compacted subsoil gets ripped or subsoiled, commonly to a depth around 18 inches, before any topsoil goes down, so the full profile drains and roots can eventually reach below the topsoil layer. The right move starts before the dirt work: strip the native topsoil at the beginning of the project, stockpile it, and replace it before planting instead of burying it. That alone saves the most expensive part of the restoration. If you are on a site where that was not done, plan and price the decompaction and the imported soil up front, because discovering it after the plants are in means digging them back out.

Imported topsoil, depth, and the interface problem

When you bring in topsoil, two things decide whether it works: how deep you place it and how you join it to the soil below. The common target for a new lawn is 4 to 6 inches of good topsoil over the whole area, with 3 inches a bare working minimum. Beds want more where the roots run deeper. Thin is the trap. A 1 inch skim of topsoil over compaction looks finished and grows almost nothing.

Specify what you order. Good topsoil is a sandy loam or loam, screened, dark, and weed-free, with no construction debris, no herbicide residue, and a known pH. The cheap pile behind the supply yard is often subsoil dyed with the word topsoil, full of nutsedge and bindweed seed that you will fight for years. Ask for a soil test on the imported material on any job that matters, the same as you would test the native soil. Dirt is not a commodity. The quality varies enormously and you cannot tell by looking once.

Now the interface, which is the failure that catches even careful crews. A layer of fine topsoil sitting on a denser, compacted clay subgrade does not drain freely across that boundary. Water moving down through the topsoil stalls where the texture changes abruptly and the topsoil saturates from the bottom up, a perched water table, sometimes called the bathtub effect. Roots hit the same line, find dense wet clay below, and stop. The fix is to blend, not stack. Till the bottom 2 to 3 inches of your new topsoil into the loosened subgrade so the texture changes gradually instead of as a hard line. No abrupt layer, no perched water, no root barrier. Thin topsoil over unworked compacted clay is the single most common reason a brand-new lawn or bed fails, and it fails six months after you are paid and gone.

Finish grade, bed depth, and the drainage tie

Soil prep and grading are the same job. You can build perfect soil and still drown the plants if the finish grade dumps water into the bed or leaves a low spot that ponds. Beds should drain, crowned slightly or tipped so water sheets off rather than collecting, and the whole site has to fall away from the building and off to a legal outlet. The how and how much of that grade lives in the drainage and grading guide. Do not skip it because the soil looks good.

Set the finish grade with settlement in mind. Freshly worked and amended soil is fluffed up and will settle as it consolidates and the organic matter breaks down, so build beds a touch high knowing they will drop. A bed graded dead flat to the surrounding lawn today is a bed that ponds in a year once it settles below grade.

Edging earns its keep on the grade, not just the look. A clean, deep edge, spade-cut or a steel or poly edge set to depth, holds the bed soil and mulch in and keeps lawn and turf rhizomes out, which is the maintenance crew's biggest ongoing fight. Cut the edge to a real depth, on the order of several inches, so grass roots cannot simply bridge it. A shallow scratch line is gone by midsummer.

Should you add sand to clay soil?

No. Adding sand to clay is the classic well-meant mistake, and it makes the soil worse, not better. The clay particles fill the gaps between the sand grains, and the result packs tighter and harder than the clay did on its own. People describe what they get as low-grade concrete, and that is close to literal. It takes a very large, impractical proportion of sand to actually lighten clay, and short of that you are building a denser, more compacted soil.

The right fix for clay is organic matter, mixed in deep, and it is not close. Compost works as a biological binder: it pushes the fine clay particles together into stable crumbs and aggregates, which opens up the pore space between them so the soil drains and breathes. Bark, leaf mold, aged manure, and quality compost all do this. They feed the soil life whose glues hold the crumbs together. It is slower than dumping sand and it is the only thing that actually works.

Two amendments do have a real, narrow place by topic. Gypsum, calcium sulfate, can help a specific kind of clay, a sodic soil where excess sodium has collapsed the structure, by swapping calcium for the sodium and letting the soil flocculate. On ordinary acidic clay that is not sodic, gypsum does little for structure, and the marketing oversells it badly. A soil test, ideally with a sodium or salts reading, tells you whether gypsum is your problem to solve. For nearly every clay a landscaper meets, the answer is organic matter and decompaction, not sand and not a miracle bag.

Starter fertilizer and feeding the new planting

Fertilizer is the last thing on this list on purpose, because it is the one most people reach for first. Fertilizer feeds plants. It does not fix soil. A nutrient problem on the soil test gets corrected to the lab's recommendation, and beyond that, restraint beats enthusiasm. Most new-planting failures are not nutrient deficiencies. They are bad structure, wrong depth, and no water.

Apply what the test calls for and incorporate it into the root zone rather than scattering it on top. If the test shows phosphorus and potassium are already adequate, which is common on established or previously fertilized ground, you do not add more, and dumping a high-phosphorus starter into that soil is wasted money that can run off into waterways. Where a starter is warranted, a balanced or slow-release product worked into the bed gives the new roots a steady, low feed instead of a salt spike.

Over-fertilizing does real harm. Excess soluble nitrogen pushes soft top growth at the expense of roots, exactly backward for a plant trying to establish, and high salt concentrations from heavy synthetic feeding burn fine roots and stress the plant you are trying to settle in. Feed light, feed to the test, and let the organic matter and the soil biology do most of the long-term feeding. The compost you worked in is doing more for the plant than the bag.

Mulch after planting

Mulch goes on after the soil is prepped and the planting is in, and it finishes the job the soil prep started: it holds moisture, moderates soil temperature, suppresses weeds, and as it breaks down it feeds organic matter back into the soil from the top. The common depth is 2 to 3 inches across the bed. More than that on a regular basis starts to cause its own problems, shedding water and staying soggy underneath.

Keep it off the stems and trunks. This is the mulch mistake you see on every neglected commercial property: the mulch volcano piled up against the trunk. Mulch held against bark keeps it wet, invites rot and pests, and on trees it encourages roots to grow up into the mulch and circle. Pull the mulch back a couple of inches from every stem and trunk so the base of the plant sits in air. Think donut, not volcano.

The depth and placement detail by plant type, and the role mulch plays in establishment, sit alongside the planting guide. The short version for the soil side: 2 to 3 inches, off the stems, refreshed not piled, and an organic mulch that breaks down rather than rock, because the breakdown is part of how the soil keeps improving after you leave.

Raised beds and engineered soil

When the existing soil is hopeless, contaminated, buried under pavement, or so compacted that fixing it costs more than going around it, you build up instead of fixing down. A raised bed or an engineered soil lets you set the root zone independent of what is underneath, which is often the practical answer on a tight urban site or over a structural slab.

The blend matters and the same rules apply, just in a mix you control. A common bed blend runs roughly half quality topsoil, a quarter to a third compost, and the rest a coarse material like coarse sand or fine bark for drainage, but the ratio follows what you are growing and the soil test on the components. Do not fill a raised bed with pure compost. It looks rich and it slumps by half in a season and stays waterlogged and too hot for roots.

Watch the same interface problem in reverse. A raised bed of light, well-drained mix sitting on compacted clay can perch water at the bottom just like imported topsoil does, so loosen the soil under the bed where you can, or accept that the bottom few inches stay wet and plant accordingly. Engineered soils for structural settings, soils over slabs, bioretention media, sand-based sports root zones, are specified to a recipe and tested to it, and on those jobs you build and test to the spec, not to feel.

Timing the prep and the wet-soil rule

Do the soil work ahead of the planting, not in the same rushed afternoon. Lime and sulfur need weeks to months to move the pH. Compost worked in wants a couple of weeks to settle and integrate before plants go in. A soil test takes a week or two to come back, and it is the thing that drives every rate. Prep done in advance is prep that works. Prep crammed into the planting day is mostly motion.

The hard rule that overrides the schedule is do not work wet soil. Wet soil, especially clay, smears and compacts under boots, tires, and tiller tines, and it dries into dense clods that ruin the structure you are trying to build. The field test is the squeeze: grab a handful and close your fist. If it ribbons out sticky and holds the shape of your hand, it is too wet, walk away. If it crumbles when you open your hand, it is ready. Moist, not wet.

This is where the schedule and the soil disagree, and the soil wins. A planting date in a wet spring is a date to push, not a date to honor by tearing up saturated ground. You can replant a missed date. You cannot un-smear a clay soil you tilled wet without starting the structure over from scratch.

Sod, beds, and tree pits each need different prep

The prep depth tracks where the roots will live, and it differs by what you are installing. One depth does not fit all three, and treating them the same shorts the deep-rooted plantings and over-works the shallow ones.

Lawn and sod root shallow and wide, so the prep concentrates in the top 4 to 6 inches: decompact, work in compost, correct pH per the test, and grade dead smooth, because every bump and dip shows in sod and ponds water. Sod is not exempt from prep. It arrives as a thin mat of roots that will only grow into the soil you give it, and sod laid on hard, unprepared ground sits shallow, dries out, and never knits in. Beds for perennials and shrubs want that same amendment carried deeper, the full 8 inches or more, because the roots go down further. The bed is the place to invest in depth.

Tree pits are the exception that trips people. The current practice for a tree is a wide, shallow hole in undisturbed native soil and backfill with that same native soil, not a pit of rich amended mix. A bathtub of fluffy compost in a hole surrounded by clay perches water and discourages roots from leaving the amended pocket, the same interface problem in a different shape. The depth, the hole geometry, and the native-backfill reasoning live in the tree and shrub planting guide. The point here: prep the bed, but do not over-amend a single tree hole.

Establishment watering and the soil the owner inherits

Good soil prep buys nothing if the planting dries out before it roots in, and it earns out only if the soil is maintained after you leave. These are the bookends, and both get dropped once the install looks done.

Water through establishment. A new planting lives on a small root system in a small volume of soil until the roots grow out into the prepared ground around them, and during that window it cannot reach water the way an established plant can. The prepared soil holds moisture better, which helps, but it does not replace watering the first season. The specifics, how much, how often, by plant and season, are in the tree and shrub planting guide, and the lawn version follows the same logic at shallower depth. Underwatering in the first season is one of the two most common ways a correctly planted, correctly sited plant still dies.

Hand the soil off with a plan. Organic matter burns off over time, so the bed you built at 3 percent organic matter drifts back down without input. Tell the owner or the maintenance crew the cheap habits that hold it: topdress an inch of compost a year, keep the mulch refreshed and off the stems, and re-test the soil every 3 to 5 years before throwing more amendments or fertilizer at it. The soil is not a one-time install. It is a thing that stays good only if it is fed.

Commercial and spec landscape soil

On commercial work the soil stops being a judgment call and becomes a specification you build and test to. The landscape spec, often in the CSI Division 32 sections, calls out the soil: the topsoil gradation and source, the organic matter percentage, the pH range, the amendment rates, and the acceptance testing that proves you hit them. Read it before you order a single yard, because the cheap topsoil that would pass on a residential job will get rejected on a spec job and you will haul it back out at your cost.

Acceptance is the part that bites contractors who treated the spec as a suggestion. The spec typically requires lab testing of the soil, before placement, after placement, or both, against named criteria, and a soil that fails the test gets reworked or removed regardless of how good it looks. Sample and test early so you find a problem while it is a delivery to refuse, not a placed-and-graded acre to redo. Keep the lab reports. They are your proof of compliance when the landscape architect's rep walks the site.

Large institutional and data center sites push this further, because the landscape soil interacts with the civil and stormwater design. Bioretention and stormwater basins use a specified engineered media tested to a recipe and an infiltration rate, and the planted areas often have to meet a soil-restoration or decompaction requirement tied to the site's stormwater permit, not just a horticultural one. On those jobs the soil prep is documented and inspected like any other part of the build. Treat it that way and the testing protects you. Treat it casually and the testing is what catches you.

What to document

The soil work disappears the moment it is covered, so the record is the only proof of what you did and the only thing that answers the question when a planting struggles a year out. On a spec job it is also your compliance file. On a residential job it is the difference between a defensible callback and an argument.

Capture the soil test results and the date, the pH before and the target, the lime or sulfur applied and the rate, the compost or amendment and the rate worked in, the decompaction done and to what depth, the topsoil source and depth placed, and the lab acceptance results where the spec required them. Record it per area, because the front lawn, the shrub beds, and the rain garden may each have had different prep and different soil. When the next person asks why one bed thrives and another sulks, the record tells them whether the soil was the difference.

Field checklist

Common mistakes

• Amending blind with no soil test, so the lime, sulfur, and fertilizer rates are guesses.
• Skimming thin topsoil over compacted clay, which perches water and stalls roots at the interface.
• Mixing sand into clay to lighten it, which packs harder and sets up like weak concrete.
• Tilling or ripping wet soil, smearing the pores shut and making clods that dry into bricks.
• Over-tilling to a fine powder, which destroys structure and leaves the soil to crust and re-compact.
• Planting into compacted ground with no decompaction, so roots sit in a pot with no drainage.
• Treating sod as exempt from prep and laying it on hard, unworked ground.
• Building beds dead flat without allowing for settlement, so they pond once the soil consolidates.
• Dumping raw, unfinished compost that robs nitrogen and seeds the bed with weeds.

Standards and references

Soil prep is agronomy and craft more than code, so the governing authorities are the testing labs and the recommendations they issue, not a single rulebook. The local cooperative extension service and its soil testing lab are the practical authority for your region: they set the pH targets, the lime and sulfur rates calibrated to your soil, and the nutrient recommendations, all tuned to local soils and the plants you grow. Their numbers govern over any rule of thumb in this guide.

On contract work, the project's landscape specification controls, commonly the soil and planting sections of CSI Division 32, which name the topsoil gradation, organic matter, pH range, amendment rates, and the acceptance testing. Where a number is specified, the spec wins. For compost quality, the US Composting Council's Seal of Testing Assurance program gives lab-backed maturity, stability, and contaminant data, and many specs reference it. ASTM and the horticultural standards cover specific test methods and materials where a job calls them out.

Treat the figures in this guide as common practice, the working numbers a foreman carries, not mandates. The soil test, the local extension, and the project spec are what you build to. Where this guide and the lab report disagree, the lab report is right, because it measured your soil and this guide did not.

Units, terms, and conversions

Soil prep mixes volume, depth, and percentage, and amendment rates show up in a few forms across a spec, a bag, and a lab report. The depths are inches of material or inches of working depth. Amendment rates come as a depth spread over an area, as a percent by volume of the finished blend, or as pounds per acre or per 1000 square feet on a lab report.

A few conversions keep the math straight in the field. One inch of material over 1000 square feet is about 3 cubic yards, so a 2 inch compost application over 1000 square feet is roughly 6 cubic yards. Organic matter is reported as a percent by weight on the soil test but specified as a percent by volume in a blend, and the two are not the same number, so read which one the spec means. Soil pH runs a logarithmic 0 to 14 scale, where each whole number is a tenfold change in acidity, which is why moving a full unit takes real material and real time.

Soil texture

The ratio of sand, silt, and clay, which sets drainage and water holding and does not change with amendment

Soil structure

How particles clump into aggregates and pore space; built by organic matter, wrecked by compaction and wet tillage

Loam

A balanced blend of sand, silt, and clay that drains and holds water well, the target for topsoil

Organic matter (OM)

Decomposed plant and animal material in the soil, reported as a percentage; 2 percent or more is healthy

Decompaction

Loosening compacted soil by forking, broadforking, ripping, or subsoiling to restore pore space

Perched water table

Water trapped above an abrupt texture change, the bathtub effect under thin topsoil on compacted clay

Sodic soil

Clay whose structure has collapsed from excess sodium, the narrow case where gypsum helps

STA

Seal of Testing Assurance, the US Composting Council program certifying compost maturity and quality

Related tools

Calculators and readiness checks for this work

Compare your options

FAQ

People also ask

• How deep do you plant a tree?
• How much should the ground slope away from a house?
• How deep should mulch be?
• What are native plants?
• Why should you do a soil test before fertilizing?