Soil
Reading the site

Soil: the texture test, the lab test, and what construction did to it

Texture judged by feel, a composite sample sent to the state lab, compaction and hardpan found, and the contamination question asked where it should be.

Start here Essentials

Two tests, one free and one cheap, and one hole. For a single bed this is an hour of work plus a week or two of waiting for the lab. Everything marked Go deeper is optional on a first pass.

  1. Take 8 to 10 cores across the area and combine them into one sample. One composite sample describes one area, not one spade of soil (section 3).
  2. Feel a handful of it. Gritty, smooth or powdery tells you most of what you need about drainage (section 2).
  3. Send the composite to your state's lab through the county extension office, and ask what a contaminant screen costs while you are there (section 3, section 6).
  4. Dig one hole and look for a hard layer. A spade that stops at eight inches is telling you something the lab report will not (section 5).

Minimum time: about an hour outdoors, plus a week or two waiting for the report. Done when: the survey's soil sheet carries a texture, a submission date and a note about depth for each area.

1. Texture, and why it decides drainage Understand

Soil texture is the proportion of sand, silt and clay particles in the soil. It is not the same as structure, which is how those particles are clumped together, and it is not the same as fertility. Texture matters first because it governs water: how fast water moves through, how much the soil holds, and how long a root sits in it.

Oklahoma State Extension puts the relationship simply: soil type greatly affects the drainage of an area, with sandy soils draining very quickly and clay soils slow to drain. Physical principle Sand particles are large, so the spaces between them are large and water runs through; clay particles are tiny and flat, so the spaces are tiny, water moves slowly and the soil holds a great deal of it. Silt sits between the two. Most real soils are mixtures, with organic matter on top of that.

The practical consequences, for design rather than for agriculture:

TextureFeelsWaterWhat it means for planting
SandyGritty between the fingersDrains fast; holds littleDries out between rains; plants for dry sites do well; establishment watering matters more
SiltyPowdery, like flourHolds well; can cap and crustOften good growing soil; compacts easily when walked on wet
ClayeySmooth and sticky when wetDrains slowly; holds a lotRich but easily waterlogged; working it wet does lasting damage
LoamSomewhere between all threeModerateWhat most plant tags assume without saying so

Texture is one of the few things about a site you genuinely cannot change. You can add organic matter to a clay soil and improve its structure, which changes how it behaves, and you cannot turn clay into sand by adding sand; in small quantities that produces something closer to concrete than to loam. Convention The useful move is to read the texture and choose plants and bed heights that suit it, which is what the rest of this chapter and Choosing plants that fit the place are for.

2. Judging texture by feel

This costs nothing and takes two minutes. Oklahoma State Extension's description is the whole method: sandy soils feel gritty rubbed between the fingers; clay soils feel smooth; silt feels powdery.

  1. Take a small handful from a few inches down, not from the surface where mulch and leaf litter confuse things.
  2. Moisten it until it is damp, not wet, and work it between finger and thumb.
  3. Rub it. Grit that you can feel individual particles of is sand. Smooth and slightly sticky is clay. Soft and floury is silt.
  4. Try to roll it into a ball, then a sausage. Sand will not hold a shape. Loam holds a ball and breaks when bent. Clay rolls into a thin sausage and can be bent into a ring without cracking.
  5. Write what you found on the soil sheet for that area, and do it again somewhere else in the yard.

Do this in several places. NC State Extension's point is that soil varies over short distances, and it changes most where the elevation changes or where something has been dug. The patch beside the drive, the patch under the old lawn and the patch where the sewer was repaired can be three different soils in a yard you could throw a ball across.

Go deeper

The jar test, and what it does not settle

The jar test is the home version of a particle-size analysis: a few inches of soil in a straight-sided jar, filled with water, a drop of dish soap, shaken hard and left to settle. Sand drops in a minute or so, silt over a few hours, clay over a day or more, and the layers can be measured with a ruler and read as rough percentages against a soil texture triangle.

It is a satisfying exercise and a useful teaching one, and it is worth knowing its limits. Test it It measures the mineral fraction of one sample, which is the thing you could already feel; it does not measure organic matter, structure, pH, nutrients or compaction; the boundaries between layers are often indistinct, especially in a soil with much organic matter; and nothing about the result changes a decision that the feel test has not already made. Use it if you enjoy it. Do not use it instead of the lab test, which answers questions the jar cannot.

3. The lab test: how to take a sample worth paying for

A soil test from a state lab, routed through your county extension office, is the cheapest piece of real data about your site that exists. It tells you pH, nutrient levels and usually a lime recommendation, and the whole value of it depends on the sample being representative, which is a sampling problem rather than a laboratory one.

Clemson Cooperative Extension's method is the one this manual follows:

  • Collect 8 to 10 or more cores from across the area, and combine them into one composite sample. A single spadeful is one spot and tells you about one spot.
  • Sample to 6 inches deep for beds and planting areas; 2 to 4 inches for lawns.
  • Send at least 2 cups per composite sample.
  • One composite per area that behaves differently. The front lawn and the shaded back bed are two samples, not one.

Mississippi State Extension adds a pattern worth knowing for anything planted against a house: take a soil test every 20 feet along the sides of the house and 2 feet out from the walls, and take the samples carefully so that mortar mix or concrete chips are not included, because they will ruin the results. Its reasoning is the reason: modern construction turns up subsoil and spreads it near the foundation, and those subsoils are often tight clays without the structure or texture for good plant growth. Convention The strip where most people put their first shrubs is frequently the worst soil on the property, and it is worth sampling separately rather than averaging it into the rest of the yard.

Mixing cores from areas that differ gives you an average of two soils, which describes neither. That is the single commonest way a soil test is wasted.

How to send it is local: the county extension office tells you which lab, which form, what it costs and how long it takes. Build the waiting into the plan. A report that arrives in May does not help a bed you dug in April.

4. Reading the report: pH and lime

pH measures how acid or alkaline the soil is. Clemson Cooperative Extension describes the scale as running from 0 to 14 with 7 as neutral; below 7 is acid, above is alkaline. It matters to gardeners not because plants care about the number directly but because pH governs how far nutrients already in the soil are available to roots.

Two published bands you will meet, both from Clemson:

PlantspHSource
Vegetables, most turfgrasses and most ornamentals6.0 to 6.5Clemson, Soil Testing
Most ornamental plants5.8 to 6.5Clemson, Planting Trees Correctly
Azaleas, rhododendrons, blueberries and conifers5.0 to 5.5Clemson, Soil Testing

The report usually comes with a lime recommendation, which Clemson notes is given per 1,000 square feet. That unit is worth noticing, because a small bed is a fraction of it: a 12-by-4-foot strip is 48 square feet, which is under 5 percent of 1,000, and scaling the recommendation down is arithmetic you do rather than a figure you estimate.

Convention The cheapest route through a pH problem is usually not to fix the pH but to pick plants that want the pH you have. Moving soil pH takes material, time and repeating, and it drifts back. An acid soil is an opportunity to grow the acid-loving plants that other gardeners struggle with, not a defect. Where the report says the pH is a long way from what you want for a specific planting, that is a case for a raised bed with imported soil, or for a different plant, and both are cheaper than a campaign against the ground.

Go deeper

Two other lines on the report

Organic matter is usually given as a percentage by weight. It is worth reading because it tells you something the texture does not: how much the soil has been worked, mulched or stripped. A new-build lot scraped of topsoil and sown with grass often reads low.

Cation exchange capacity, or CEC, is a measure of the soil's ability to hold onto nutrients. Clay and organic matter raise it; sand lowers it. A low CEC is part of why sandy soils need feeding more often and in smaller doses: the nutrients wash through rather than being held. Neither figure changes a design decision on its own; both are context for the recommendations above them.

When the lime recommendation is large

A report that recommends a great deal of lime is describing the gap between your soil and the crop the recommendation was calculated for. Applying the whole amount at once is generally not the method; split applications over seasons, with a retest, are. Ask the extension office that handled your sample. This is exactly the kind of question the service exists for, and the answer depends on your soil, your material and your region.

5. Compaction, hardpan, and what construction leaves

A soil test tells you about chemistry. It tells you nothing about whether a root can physically get through the ground, and on any lot that has been built on in the last fifty years that is the more important question.

NC State Extension describes what happens: construction strips the topsoil and leaves compacted subsoil, often with building debris in it. Virginia Cooperative Extension's Master Gardener handbook adds the distinction that matters for what you do next. Soil may be uniformly compacted, or it may carry a hardpan, a dense layer typically 6 to 18 inches under the topsoil that restricts both root growth and drainage. Physical principle Those are different problems. Uniform compaction is a whole-profile problem you work with or build above. A hardpan is a layer: there may be perfectly good soil above it and below it, with a barrier in between that water sits on top of.

The test is a spade and ten minutes.

  1. Dig a hole at least 12 to 18 inches deep, in the area you intend to plant, when the soil is moist but not wet.
  2. Notice where the spade meets resistance. If it goes in easily to the full depth, you probably have neither problem. If it is hard from the surface, that is uniform compaction. If it goes in easily and then hits something at a particular depth, you have found a layer.
  3. Look at the face of the hole. A hardpan often shows as a visible change in color or density, and roots above it frequently turn sideways rather than going through, which is the clearest sign there is.
  4. Check whether water sits in it. Fill the hole and watch. That is the drainage test in Water, drainage and slope, and if a layer is holding water, the two chapters are describing the same thing from different sides.
  5. Write the depth on the soil sheet, with what you found, for each area you tested.

What to do about it is a real decision with published options. University of Wisconsin-Madison Extension's recommendation after construction is to aerate the compacted soil and spread at least 4 to 6 inches of topsoil evenly over it. Clemson Cooperative Extension offers a different and often more practical route for beds: beds can frequently be elevated 8 to 12 inches above the existing grade using native soil from the site. Convention The two are not in conflict; they answer different questions. Wisconsin's is about reclaiming an area, typically a lawn, over a whole lot. Clemson's is about getting a planting bed to work without excavating anything, by building up instead of digging down.

Limits

There is a point where loosening compacted ground stops being gardening. Anything that involves machinery near a structure, anything that changes the level against a foundation, and anything that alters where water goes is outside what this manual covers, and the last of those is also a legal question. If a compacted area is holding water against a house wall, that is a drainage contractor's problem and possibly a builder's, not a soil problem, and Water, drainage and slope names the point at which the line is crossed.

6. The contamination question

A routine soil test does not test for contamination unless you ask for it, and paying for the routine test and assuming it covers lead is one of the quietly dangerous mistakes in home gardening.

NC State Extension's chapter raises contamination as a site-history question, and the histories that raise it are not exotic:

  • An older house with painted exterior woodwork. Paint flakes and weathers into the soil immediately around a building, which is also where people put their first beds.
  • A former orchard or farm. Historical pesticide residues can persist in soil.
  • A demolition site, or a lot where a building was taken down. Debris, and whatever was in it.
  • Old plumbing or a former industrial or commercial use.
  • Proximity to a road that carried traffic before the phase-out of leaded gasoline.

Ask the authority Ask for a contaminant screen when two conditions apply: children will play or dig in the soil, or food will be grown in it. That is the threshold this manual uses, and the decision about what to do with a result is not a gardening decision. It belongs with your county extension office, your state's environmental agency and, where children are involved, a doctor. What a result might lead to, from raised beds with imported soil to covering with turf to leaving an area alone, depends on the contaminant, the level and the use.

If you are not growing food and nobody digs, the question is less pressing and still worth asking once. It costs a phone call to the extension office to find out what a screen costs where you are.

7. Fix it, or plant around it

By the end of the survey you will have a list of soil problems, and each one needs a decision: fix it, build above it, or choose plants that do not mind.

What you foundFix itBuild above itPlant around it
Sandy, fast-drainingNot realistically; adding sand to clay in garden quantities does not fix itRarely neededYes: plants for dry sites, and more water through establishment
Heavy clay, slow-drainingOrganic matter improves structure over years, not weekendsOften the practical answer: a raised bed, 8 to 12 inches, native soil (Clemson)Yes: plants that tolerate wet feet, in the wettest part
Uniform compaction after constructionAerate and add 4 to 6 inches of topsoil (Wisconsin Extension)Yes, for bedsRarely enough on its own; roots need somewhere to go
A hardpan layerBreaking it is possible and is work; a deep hole through it can simply fill with waterYes, and it is usually the cheaper answerShallow-rooted planting, knowing what is beneath
pH far from what a plant wantsLime or sulfur, repeated, with a retestRaised bed with imported soilUsually the cheapest: pick plants for the pH you have
Suspected contaminationNot a gardening decisionRaised beds with imported soil are one common responseAsk first, decide after; the extension office and your state agency

The column that costs least is usually the right-hand one, and it is the one people reach for last.

8. See it: three soils See it

Fictional example

Rosalind's strip, Burlington, Vermont. Known: the strip is fill left by a sewer repair. She took 8 cores along its 12-foot length, 6 inches deep, combined them in a clean bucket, and sent 2 cups to the state lab through the county office. Feel test: distinctly gritty, no stickiness, will not hold a ball. Hole test: the spade went in to 14 inches with no resistance. Report, when it came back: pH 6.8, low organic matter, no lime recommended. Assumed: that the fill is uniform along the whole strip, which is hers and is plausible for a trench backfilled in one go; she noted it on the sheet. Open: whether the fill is deep enough for a shrub's roots or sits on undisturbed subsoil a foot down, which one deeper hole would settle. What it changed: she had been thinking about moisture-loving perennials because the strip is beside a driveway and she expected runoff. Under these assumptions it is a free-draining site that will dry out fast, which is a different list, and the establishment watering in Planting and the first three years matters more here than it would in clay.

Fictional example

The Okonkwo back yard, Columbus, Ohio. Known: three composite samples, because three areas behave differently: the open lawn, the strip along the 2024 sewer dig, and the ground under the maple. Feel test: smooth and sticky everywhere, clearly clayey. Hole test in the lawn: easy to 11 inches, then a distinct change to a dense gray layer, with roots turning sideways on top of it. Hole test in the sewer strip: hard from 4 inches down, no clean layer, a different problem. Assumed: that the layer in the lawn is a construction hardpan from 1958 rather than a natural horizon, which is their assumption and which does not change what they do about it. Open: the lab report, which was sent in March and had not arrived when they dug. What it changed: everything about the planting beds. Under these assumptions digging deep holes into a hardpan would produce sumps that fill with water, so their beds are being raised 10 inches above grade with soil from elsewhere on the lot, which is the Clemson route, and the sewer strip is a candidate for Wisconsin Extension's aerate-and-topsoil treatment or for plants that tolerate wet. Which of those it gets is decided in Water, drainage and slope, because the drainage test answers it and the soil test does not.

Fictional example

Harlan's front yard, Austin, Texas. Known: a 1970s house with painted wood trim, a front yard of St. Augustine lawn he intends to remove and replant. He asked the extension office about a contaminant screen because he was unsure, and was told what it cost and that the painted trim and the house's age were reasonable grounds. Assumed: nothing; the question was asked rather than answered by guessing. Open: the screen result, and what the extension office and the state agency advise if it shows anything. What it changed: the order of his phases. The planting bed closest to the house wall is on hold until the result is in, while the rest of the design proceeds. Nobody in his household digs and no food is being grown, so the threshold in this chapter does not strictly apply to him; he asked anyway, and the cost of asking was a phone call.

9. Practice: sample and feel Practice

Minimum: one composite sample from the area you intend to plant, a feel test, and one hole dug to 12 inches. Time: about an hour outdoors; the report takes a week or two. Materials: a spade or a trowel, a clean plastic bucket (not galvanized, which can affect results for some tests), a bag, and the site survey's soil sheet.

  1. Decide your areas before you take anything. Each area that behaves differently gets its own composite sample: the bed by the house, the open lawn, the shaded part, the patch where something was dug.
  2. Take 8 to 10 cores per area, each to 6 inches (or 2 to 4 for lawn), spread across the area, avoiding anything obviously unrepresentative like a bonfire site or the ground under a downspout.
  3. Mix them in the bucket, remove stones and roots, and bag at least 2 cups. Label the bag with the area, not "garden".
  4. Do the feel test on a handful from the bucket before you seal the bag, and write the answer on the sheet.
  5. Dig one hole to 12 inches or more and record where the spade met resistance, and what the face of the hole looks like.
  6. Send it, through the county extension office, and write the submission date on the sheet so you know when to chase.
Go deeper

Topsoil depth, measured in several places. While you have a spade out, measure how deep the darker topsoil goes before the lighter subsoil begins, and do it in three or four places. NC State Extension's point that soil varies over short distances and changes most where elevation changes is easiest to see this way. A yard with 10 inches of topsoil at one end and 2 inches at the other is two sites, and the thin end will show it in dry summers no matter what you plant.

If you are also sampling for a lawn, remember the depth differs: Clemson gives 2 to 4 inches for lawns against 6 for beds. Sampling a lawn to 6 inches dilutes the layer the grass actually roots in.

10. Your garden Your garden

11. Check your work Check your work

Criteria

  • Each area that behaves differently has its own composite sample, not a share of one average.
  • Each composite is 8 to 10 or more cores, to the right depth for what grows there.
  • The soil sheet carries a texture judged by feel for each area.
  • At least one hole has been dug to 12 inches or more, and what the spade met is written down with the depth.
  • The submission date is on the sheet, so a missing report gets chased.
  • The contamination question has been asked, not assumed, where children dig or food will grow.
  • Each soil problem has a decision beside it: fix, build above, or plant around.

Worked example

The Okonkwo sampling (fictional) is the model, and the useful part is the number of samples. Three areas, three composites, because the lawn, the sewer strip and the ground under the maple have different histories; averaging them would have described none of them. The feel test was done on all three and gave the same answer, clayey, which is itself informative. The hole test found two different problems, a clean hardpan in the lawn at 11 inches and uniform hardness in the sewer strip, and the sheet records both with depths. The weakness is one of timing rather than method: they dug before the report arrived, which is normal and means the pH line has to be read afterward and the planting adjusted if it is surprising. The decision column is filled in for the lawn, raise the beds, and left open for the sewer strip, which is honest, because the drainage test has not been done yet and it is the test that decides.

Common mistakes

  • One sample for the whole yard. An average of two soils describes neither.
  • One spadeful instead of a composite. The lab is precise about a sample that is not representative.
  • Sampling the lawn to bed depth, which dilutes the layer the grass roots in.
  • Including mortar or concrete chips in a sample taken against a house wall, which Mississippi State Extension warns will ruin the results.
  • Assuming the routine test covers contamination. It does not unless you ask.
  • Reading the pH and setting out to change it before considering plants that want the pH you have.
  • Digging a deep planting hole through a hardpan and creating a sump that holds water around the roots.
  • Working clay soil when it is wet, which damages structure in a way that takes seasons to undo.

What is still unresolved

  • How fast the ground actually drains: the hole and bucket test in Water, drainage and slope, which measures what the texture only suggests.
  • What to do if a contaminant screen finds something: your county extension office and your state environmental agency, and not this manual.
  • Whether to amend a planting hole: the argument, with its named author, is in Planting and the first three years, and it is not the answer most garden centers give.
  • Which plants suit the texture and pH you measured: Choosing plants that fit the place.
  • How much lime, and over how long: the extension office that handled your sample.

12. Knowledge check and scenario

Why 8 to 10 cores rather than one good spadeful?
Because the lab measures what you send with some precision, and the question is whether what you sent represents the area. Soil varies over a few feet. Clemson Cooperative Extension's method is 8 to 10 or more cores combined into one composite, which averages out the variation deliberately, rather than taking one spot and hoping.
Your test comes back at pH 5.2 and you want to grow vegetables. What are the options, and which is cheapest?
Raise the pH with lime, repeated and retested, over seasons; build a raised bed with imported soil; or grow what suits 5.2, which Clemson's bands put in blueberry and azalea territory rather than vegetables. For vegetables specifically the first two are the real options. The general principle still holds for ornamental planting: picking plants for the pH you have is cheaper and more durable than moving the pH, which drifts back.
What is the difference between uniform compaction and a hardpan, and why does it change what you do?
Uniform compaction is hard all the way down; a hardpan is a dense layer, typically 6 to 18 inches under the topsoil on Virginia Cooperative Extension's description, with better soil above and sometimes below it. The difference matters because digging a deep planting hole through a hardpan makes a sump that holds water around the roots, while the same hole in uniformly compacted ground is just hard work. Building beds above the problem answers both, and is often cheaper than either excavation.
Your routine soil test came back clean. Is the soil safe for a vegetable bed beside a 1920s house?
The test does not say. A routine soil test reports pH and nutrients; it does not screen for lead or other contaminants unless that screen is requested and usually paid for separately. An older house with painted exterior woodwork is one of the standard reasons to ask. Where children dig or food will grow, ask for the screen, and take any result to the extension office and your state environmental agency rather than deciding alone.
Why is the strip of ground against a house wall worth its own sample?
Because it is frequently not the same soil as the rest of the yard. Mississippi State Extension notes that modern construction turns up subsoil and spreads it near the foundation, often tight clay without the structure for good growth, and recommends sampling every 20 feet along the sides of the house, 2 feet out from the walls, taking care not to include mortar or concrete chips. It is also where most people put their first shrubs.
(Go deeper) What does the jar test add to the feel test?
A rough numerical split of sand, silt and clay in one sample, which can be read against a texture triangle. What it does not add is organic matter, structure, pH, nutrients or compaction, and in practice it usually confirms what your fingers already told you. It is worth doing for interest and is not a substitute for the lab test or the spade.
Scenario: you dig a hole for a shrub in a clay yard and hit a dense layer at 10 inches. You have the shrub, it is Saturday, and the hole is dug. What do you do?
This is the intuitive move and it creates the classic problem: a hole dug through a layer that water sits on becomes a container with no outlet, and compost-filled backfill makes it worse by holding more water than the surrounding ground. The argument against amending backfill, with its named author and her reasoning, is in Planting and the first three years. Deep ripping of a hardpan is real work for machinery over an area, not something a spade achieves for one plant.
This is the sound path and it costs an hour. The water test tells you which problem you have: a hole that empties is a hard layer you can live with, and a hole that holds water is a sump you have just dug. Clemson Cooperative Extension's route for a bed on difficult ground is to elevate it 8 to 12 inches above grade using native soil from the site, which puts root run above the problem instead of inside it. The shrub keeps for a few days in its pot in the shade, watered.
Planting shallow and mulching is a reasonable instinct and it is half the answer: the right depth is only as deep as the root ball, which Planting and the first three years covers. What it leaves unresolved is where the water goes, and watering more often on a layer that holds water is the wrong direction. Do the water test first; it may show you have less of a problem than you think.

Checklist

Records your learning only; it is not a qualification.