Building a no-till food garden on Central Texas clay and caliche

Do not till. On Central Texas clay and caliche, renting a tiller buys one satisfying weekend and then a bed that sets up harder than the yard you started with. Build up instead. Cardboard over the grass, compost on top of the cardboard, plants into the compost the same day. That is the entire method. Everything below is the local detail that decides whether the bed is still alive in August.

Why bare soil loses here

July makes the case. Normal highs sit in the mid nineties for weeks and push past a hundred often enough, and bare ground in full sun runs hotter than the air above it. Clay shrinks and cracks open. Seed you sow into that surface either bakes or germinates and dies at the crown within days.

Caliche fails differently, and it is not everywhere. Plenty of west-side yards sit on thin soil over limestone or a white, cemented carbonate layer. Plenty of Blackland Prairie yards east of town are deep clay instead, and College Station is a different soil region again. Where caliche is present it holds water above itself, which drowns roots in a wet spring, and blocks them below, which starves them in a dry summer. Do not take my word for what is under your yard. Push a spade or a piece of rebar down in three or four spots before you decide anything.

Then there is the rain, which arrives less as rain than as delivery by fire hose. Two inches in forty minutes, most of it sheeting toward the storm drain, followed by six dry weeks. Clay swells shut when it finally wets through, the top inch crusts as it dries, and that crust sheds the next storm. Bare soil here is not a neutral starting point. It is a slow loss you keep paying on.

Cover changes what happens at the surface. Shade keeps the top layer cooler and the cracks narrower. Mulch takes the impact of raindrops, so less crusting. Water moving slower means more of it soaking in. Swales and basins catch flash rains, so the soil drinks instead of the storm drain. That is what cover does at the surface. What your tomatoes do with it is a separate argument.

Build the bed up, not down

Cardboard first. Plain brown cardboard, tape and staples pulled off, sheets overlapped a hand’s width so no grass finds the seam. Wet each sheet as you lay it. Dry cardboard sheds water and travels.

Growing medium second, six to twelve inches of it, right on top. Deeper if the spade hit rock. This is the bed. You are not fixing the clay underneath, you are setting a growing medium above it and letting roots, worms and weather make the introduction.

Do not order pure compost. A blended raised-bed mix, roughly a third mineral soil or sandy loam and the rest finished compost, drains better, settles less and carries fewer salts than compost alone. Ask the supplier what went into the pile and whether it is finished. Immature compost stalls seed instead of feeding it, and no squeeze test tells you that. If a load smells sour or sharply of ammonia, let it sit before you plant into it.

Then plant, the same weekend. Transplants go straight in: slit the softened cardboard, set the root ball in compost, firm it. Direct-sown seed sits on the compost surface, barely covered; compost is loose enough for a seedling to push through. There is no cure period and nothing to wait for. A bed with something growing in it is a bed you water, watch, and learn from. An empty bed is a pile of dirt you feel bad about by August.

Start with one bed. Two beds you keep beat ten you abandon.

Where chips belong, and where they must not go

Chips on the paths, as deep as you like. Chips around established perennials and out to the drip line of anything woody, but only two to four inches there, and never piled against a trunk. A cone of mulch heaped on the flare of a young fruit tree holds moisture against the bark and rots it. Leave a clear ring you can see soil through. In those places chips shade the ground, hold moisture, and smother the weeds you would otherwise be hoeing at 6am.

Chips never go in the seed bed. This is the mistake I see most often from people who watched a beautiful garden video and then mulched their carrots. A chip layer over a seed row buries seedlings outright. At the thin line where chips meet soil, the organisms breaking down that wood pull nitrogen from the top layer of soil to fuel the work, and that interface is exactly where your germinating seed is sitting. The result is pale, stalled plants and a blank row you blame on the seed company.

So: compost in the beds, chips on the paths. Thin mulch on vegetables, deep everywhere else. Once a plant is up and growing, bring chips in toward it a handful at a time, leaving a few inches of bare soil at the stem. Mulch held against a stem keeps it wet, and wet stems rot.

Getting chips without paying for bags

Call tree services. A truckload of chips is a disposal cost for a crew, and most would rather dump in your driveway than drive to the yard. Ask for fresh arborist chips or ground trimmings. Ask what they have been cutting: cedar and Ashe juniper loads are stringy and slow, hackberry and elm disappear in a season. Ask whether the trees came off a fence line or a roadside, since those are the trees most likely to have been sprayed, and the next section explains why that matters. And ask for the driveway, not the middle of the street. Neighbors remember.

Bagged dyed mulch is a different product. What sits under the colour varies by manufacturer, anything from bark to recycled wood, and it still breaks down and still adds organic matter. The dye is not the objection. Not knowing the feedstock is. Ask what is in it, and skip anything vague about construction debris.

The warning that earns its own section

Persistent growth-regulator herbicides, aminopyralid and clopyralid among them, get sprayed on pastures and hay fields because they kill broadleaf weeds and leave the grass standing. They pass through the livestock that ate the hay, they survive ordinary composting, and they keep acting in whatever soil they land on. Tomatoes, beans, peas, squash, and potatoes come up with cupped, curled, strappy leaves that look like a virus. People lose a whole season to it and blame their seed.

Hay, straw, and manure are the usual carriers. Aged is not the same as clean. Composted is not automatically clean either, because these compounds hold up through normal pile temperatures. When a neighbor offers free horse manure, the price is not the problem.

Test before it touches the garden. The test costs a packet of bean seed and three weeks:

1. Take small samples from several places in the pile, not one scoop off the top, and mix them together. Fill three to six pots with roughly half that sample and half bagged potting mix. Fill several more pots with the potting mix alone. Those are your controls, and one is not enough to trust.
2. Plant three bean or pea seeds in every pot. Beans are sensitive to these chemicals and show damage early.
3. Water them the same and keep them in the same light. Watch the first set of true leaves.
4. After two to three weeks, compare the new growth. Cupped, curled or strappy leaves in the test pots while the controls look normal is evidence that something in the material injures plants. It does not tell you which compound, and it does not need to. Keep the material out of the garden, and out of the compost pile too, since the pile just banks the problem for later.

Three weeks and a packet of bean seed against a summer of curled tomatoes. Run it on any hay, straw, manure or compost whose spray history you cannot personally vouch for.

One caution on diagnosis. Curled, distorted growth also comes from viruses, mites, nutrient problems and drift from someone else’s spray day. The bioassay is useful precisely because it compares against a control. A photo on its own settles nothing, whatever the internet tells you.

Composted manure is a good input once you have reason to trust it. Raw manure is the one to be careful with, because it can carry human pathogens. That is a food-safety question before it is a fertility question. If you do use it, the USDA organic intervals are the sane benchmark: at least 120 days before harvesting anything that touches the soil, at least 90 days for everything else. Composted is simpler.

Alkaline soil, and the acidify myth

Plenty of yards here run alkaline, and some run the other way. Four cities across three soil regions do not share one pH, so test rather than assume. You can spend every spring fighting your number with sulfur, or you can plant for it.

Texas A&M publishes tolerance ranges worth knowing before you buy seed. Beets are listed as fine up to about pH 8. Squash, okra, peppers and cabbage sit closer to 7.5. Beans are listed only to about 6.8, which surprises people who plant them everywhere. Blueberries are hopeless here, and no handful of sulfur turns Texas clay into blueberry country. Put blueberries in a pot of acid mix and stop arguing with the yard.

Test instead of guessing. A routine soil test through the Texas A&M Soil, Water and Forage Testing Laboratory reports pH, phosphorus, potassium and a fertiliser recommendation. Check their current form for the fee, it changes.

As for the myth: surface wood chips do not meaningfully acidify soil. Chips break down slowly, at the top of the profile, and the small acidity they generate stays local to that thin layer. They will not walk a pH 8 yard down to 6.5. Mulch for water and weeds, not for pH.

Drip under the mulch, inside the water rules

Lay the drip line on the soil, then mulch over the line. Never the reverse, since watering the top of a chip layer is paying to humidify your yard. Run drip long and less often, so water moves down past the top few inches instead of teasing the surface. Put it on a timer, because nobody hand-waters correctly in August.

Know your drought stage. Central Texas cities move between stages through the summer, the allowed days and hours shift with each stage, and the schedule you set in May is probably against the rules by August. Look up the current rules for your city, write the allowed days on the timer, and recheck when the reservoirs drop. The Texas Water Development Board posts statewide drought conditions.

A realistic first year

Build in September or October. Fall is the kind season here, and roots work through a mild winter while the bed settles.

Those crops do not all go in at once, though, which is where people get tripped up. Around Travis County, broccoli transplants and many root crops start in September, most cool-season greens in October, and garlic anywhere from October into December. Spring is not one window either: some tomato transplants go out in February while beans and squash generally wait for March and okra for April. Pull your own county’s planting calendar and work from that rather than from a single date you read once.

A March build works too. You just meet your first July with shallow roots and more hand-watering.

First-year expectations are modest: a thin harvest, a couple of crop failures, and a lot of learning what your clay does. At the end of the first season, go and look rather than assume. How far did the bed settle. Did water drain or sit. Did roots reach the native soil underneath, or stop at the boundary. Top it up and change the watering plan based on what you actually find, not on what a second year is supposed to bring.

What goes wrong

Grass finds the seam you did not overlap. Chips end up in the seed bed anyway because it looked tidy. Untested manure arrives as a favor from a friend with horses. A young bed dries out in its first summer before any roots have reached the clay. Tomatoes go in during June and get asked to fruit through 104 degree afternoons, which is not a reasonable ask. And the big one: eight beds get built in one heroic weekend, and four of them stand empty by October.

Cardboard, compost, chips, water. Build one bed this fall and pay attention to what it does.