Zaytuna Farm: 20 Years of Earthworks Numbers on 66 Hectares

The Baseline

Zaytuna Farm was purchased in 2001 as 66 hectares of degraded cattle pasture at The Channon in subtropical New South Wales. Starting topsoil carbon: 1.2 percent in the top 30 cm. Creek flow: seasonal only, typically 5-7 months per year. This is the standard subtropical Australian pasture degradation profile after decades of clearing and set-stocking. Nothing unusual about the baseline. The property was not selected for exceptional soils or rainfall. It was selected as a teaching site for the Permaculture Research Institute and documented from the start (source: vault_atom_TBD, Geoff Lawton public documentation; PRI Zaytuna site records).

The Interventions

Between 2001 and 2003, the Permaculture Research Institute installed: contour swales on the main slopes, keyline ploughing across primary water-shedding surfaces, 7 farm ponds sited on ridgelines, and check dams in gullies. Progressive food-forest vegetation was established in the swales from 2002 onward. No supplemental irrigation import. No piped water from external sources. The entire system ran on rainfall and gravity throughout (source: vault_atom_TBD, Geoff Lawton public documentation; PRI Zaytuna site records). The intervention phase lasted roughly two growing seasons. The monitoring ran for 20 years.

The Numbers That Moved

By year 7 (2008), creek flow on the main watercourse shifted from seasonal to year-round perennial flow, correlating with completion of the primary earthworks phase (source: vault_atom_TBD, Lawton disclosures; PRI Zaytuna site monitoring). By year 15 (2016), topsoil carbon in the top 30 cm had risen from 1.2 percent to 4.8 percent: a 4x increase over 15 years with no change in annual rainfall averages. Recorded biodiversity increased by factors of 2-5x depending on taxon across the same period.

The correct caveat is that Zaytuna is a subtropical site. The absolute magnitudes of soil carbon gain and creek flow recovery will differ in temperate grain-belt climates. The mechanism is the same: infiltration, residence time, vegetation re-establishment. Temperate calibration data comes from the Mulloon Creek catchment (NSW) and comparable European operations. See the watershed baseflow dispatch for the catchment-scale numbers. For the full mechanism and earthworks economics, the water harvesting pillar essay carries the detail. The Zaytuna numbers are the mechanism proof, not the universal scaling constant. They answer the question of whether earthworks work on degraded land with no external water input. The answer is yes, with a 7-year and 15-year documented record to support it, per the pillar essay on water harvesting.