A drought-year comparison: two orchards, one decision

Two blocks, eleven kilometres apart, same variety, same rootstock, planted eight days apart in the same winter. From October to March the first block recorded 114 mm and the second recorded 41 mm. Both owners used water retaining gel. One ended the season with 96 percent survival and a first light crop; the other lost 14 percent of its trees and replanted 190 of them the following June.

That comparison gets quoted at grower meetings as proof that gel works or does not, depending on who is talking. It proves neither. What it does show, once you walk both blocks, is which decisions were actually different, and almost none of them were about the polymer.

Same nursery, same tree grade, same 100 g per tree, same drip line, same six-hour weekly run schedule on the controller. Both owners bought on the same specification and both applied it in a single pass at planting with a mixer wagon. On paper they ran the same trial.

What differed was the soil, the timing of the planting within the winter, and one irrigation decision in early December that neither owner remembers making.

Where the divergence actually started

Block one sits on a sandy loam with 1.4 percent organic matter and a gravel layer at 700 mm. Block two is a clay loam with 2.6 percent organic matter and a hardpan at 350 mm from previous cultivation. The gel dose was identical. On block two, that dose went into a profile where water already moved slowly, and the first two irrigation events perched above the hardpan rather than draining through. Roots sat in saturated soil for four days in November.

Block two also planted into a dry profile. The paddock had been worked in September and received 9 mm in October. The trees went in with dry backfill, the gel went in dry, and the first irrigation came five days later. Dry granules in dry soil compete with young roots for the first water that arrives. That is not a gel failure, it is a sequencing failure, and it is a very expensive one.

Then December. Block one’s owner cut his run time by 30 percent after a 22 mm rain event, then put it back after ten days when he dug a hole and found dry soil at 150 mm. Block two’s owner cut his run time by 30 percent in the same week and left it there until February, on the assumption that the gel was now carrying the block. It was not, because a third of the material was in the perched zone and the rest had been through four cycles of swell and collapse.

The numbers that mattered at the end

Survival, block one: 96 percent. Block two: 86 percent. Trunk diameter at the end of the season: block one averaged 21 mm, block two 15 mm on the survivors. Water applied per tree across the season: block one 410 litres, block two 290 litres. The block that used less water lost more trees, and the saving of 120 litres per tree was worth far less than the cost of replanting 190 trees plus a lost year of growth.

Set against that, the difference in rainfall, 73 mm, is not trivial. A fair reading is that block two had 73 mm less rain, 120 litres less applied water, a shallower usable profile and a dry-planting start. The gel was present in both and behaved the way it should in both.

What a comparison can and cannot tell you

A paired comparison like this is useful for generating a hypothesis and almost useless for proving one. Two sites, one season, no replication, and at least four variables moving at once. If the block two owner had decided on the strength of it that gel does not pay, he would have been drawing a conclusion about his hardpan and his December irrigation decision.

The honest version of the story, which I have heard him tell since, is that the gel bought him time in the weeks when he could not irrigate, and that he then spent that time twice by cutting water early. That is a management lesson, not a product one.

What both owners are doing next year

Block one is keeping the dose and adding a pre-planting wetting pass so the granules go in hydrated. Block two is keeping the gel, ripping the hardpan to 500 mm before planting, wetting the profile to field capacity a week before the trees arrive, and leaving the controller alone until a soil probe at 200 mm says otherwise.

If you have two blocks heading in opposite directions this season, the useful thing to record is not just survival. It is water applied, rain received, profile depth to any restrictive layer, and the date of every irrigation change. The application guide has a one-page record sheet for exactly this, and it takes about four minutes a week per block. Four minutes a week is what turns next year’s argument into next year’s decision.