Same Species, Two Soils: Dose Curves for Sand and Loam

A shelterbelt of the same hybrid poplar cuttings, planted the same week across two fields: one a deep sandy loam, one a heavy alluvial loam with a bit of clay at 30 cm. The planting spec said 90 g of water retaining gel per tree on both sides of the fence, because the contract was written once and copied twice. By mid-July the sand field was the healthier of the two, and the loam field had lost 11 percent of its trees to something that looked, confusingly, like drought.

It was the opposite of drought. The loam already holds 25-30 percent of its volume in plant-available water, and a 90 g dose of gel in the root zone pushed that figure high enough that a wet July left the planting holes effectively ponded for days at a time. Poplar tolerates wet better than most, and it still sulked. The sand, with its 8-10 percent available water capacity, used every gram of the same dose and asked for more.

Reading the curve, not the bag

• Sandy soil: 80-120 g per tree for whips and cuttings, or 1.5-2 kg per cubic metre of worked backfill; the gel is doing structural work the soil cannot.
• Loam: 30-50 g per tree, enough to stretch the interval between rains, not to replace the soil’s own storage.
• Clay loam: 20-30 g, and only on drought-prone sites; on moisture-retentive ground the honest dose is often zero.
• Depth: 15-30 cm for all three, around the future root plate, never a single lump under the plant.

The counter-intuitive part of the curve is its shape. Going from 30 g to 60 g on loam usually buys almost nothing measurable in growth, while on sand the same doubling is the difference between a decent establishment season and a write-off. Dose response is not a straight line, and it is not even the same line on different soils, which is why per-hectare rules of thumb travel so badly between farms.

The cost arithmetic follows the curve. On the sand field, gel at 100 g per tree across 2,500 trees per kilometre of shelterbelt is 250 kg, a few hundred euros at trade prices, against a replant bill that had run to four figures on the previous un-gelled planting. On the loam field the same money buys maybe 60 g per tree with no measured benefit and a measurable wet-weather risk. Spending the gel budget where the soil is poor is not generosity to the crop; it is simple arithmetic.

Setting the curve on your own ground

Two things calibrate it faster than any datasheet. First, texture by feel: rub a moist sample between thumb and forefinger; gritty and it will not hold a ball, which means the high end of the range. Second, a paired trial: ten trees at half rate, ten at full, on each soil type, scored at the end of the first summer. That afternoon of planting and one hour of scoring in autumn beats every generic recommendation table, including the ones printed on the sack.

The textures between the extremes

Most ground is neither beach nor brick. A loamy sand sits about a third of the way from sand toward loam and takes roughly 60-80 g per tree where the pure sand figure was 100; a sandy loam takes 50-70; a true loam takes the 30-50 already given; and silty soils behave like loams with less structure, so err toward the lower figure until a trial says otherwise. Interpolate by feel first and by laboratory texture later, never by contract habit.

Where a single contract crosses several textures, dose by zone rather than by average. Mark the texture breaks on the planting plan the day the trial pits are dug, hand each crew a scoop calibrated for its zone, and keep the zone boundaries in the records. An averaged dose is wrong everywhere at once, and it is the most common way a well-intentioned spec ends up proving nothing about any soil it touched.

The loam field’s lesson stuck with the contractor: after the wet July, the affected rows were re-planted at 30 g, and the survivors on the original high dose took until September to look normal. When the next tender went out, the spec called for gel on the sandy blocks only, with the loam blocks watered instead, and the details behind that split are tabulated in the agri-forestry water retention guide for anyone writing their first texture-based spec.