This spring warmed up faster than usual, but also was extremely dry (see my other post, Mixed success in 2026 spring drought). This was a great chance to get a longer growing season in for the three sisters, but with the garden filling up I needed to expand into the parched clay bed that has only grow lawn grass for decades. I tried out the following technique, based loosely on what I’ve seen and heard of Native American techniques for arid agriculture:
Dig a hole, maybe 8 inches across and 6 inches deep. Keep the dry surface soil and damp underground soil seperate for re-use.
Pour at least 2 gallons of water into the whole. It’s “full” when the water takes a few minutes before soaking away, indicating that the clay has sponged up a decent amount of water.
Place a few seeds each of corn, beans, and squash. Seeds should all look as diverse as possible when selecting the mix.
Cover seeds with 3 or 4 inches of a mix of compost and the damp soil from earlier, and water it again
Apply a top coat of the dry top soil from before, it takes on a light color when dry and reflects the hot sunlight.
The super saturated soil under the seeds provides them with just enough water to germinate and come to the surface. Here’s a few of the best holes where all three sisters have already germinated:
All credit for these climate resiliant plants I’m growing, especially the three sisters, goes to Native American people, and specifically Cherokee. I’m fairly certain most of the beans and corn I’m growing are Cherokee but I get most of my seeds at informal seed swaps so I can’t pin down a specific heritage for them. And the land I’m on is also theirs, worth pointing out.
That looks like a very effective method, although also a bit more labor intensive than I want to do. I used a Radius weeder, which is like a spade but with a very narrow and sturdy blade, to make holes about 4-5 inches deep and about an inch wide. I then dropped two or three corn kernels into the hole, not covering, just waiting for rain to arrive. The squash seeds I simply scatter-sow, since at this stage I’m saving mass quantities of squash seeds. The beans I will plant similarly to corn with the Radius tool, once the corn comes up.
Squash, in my experience where i live, germinate and grow very well when scatter sowed on the surface. They just need a good rain spell to get going. But that’s fine with me, I’m not in a hurry and I’m not going to irrigate at all.
In a scenario of no rain, your method looks viable if you’re willing to keep on watering. But I’m not willing to water at all.
Layout and density
A tightly packed milpa (polyculture field, grown 2 years then left fallow for 8 years) generates competition that leads to failure. Sowing in stacks, in accordance with the ancestral method, is still the most effective:
Corn: 1 cluster every 40-50 cm, 3 to 5 seeds per cluster, thinned to 2-3 plants.
Beans: 1 to 2 seeds per corn plant cluster
Squash: 1 plant every 1.2 to 1.5 m
Following up on this, the whole area of the garden I planted like this grew very little. The trench from the road causes the soil to drain too well, the soil is still plain clay with only shallow hummus, and the juglone from the black walnut stunted a lot of the plants. Burnweed grew just fine, but not much else.
I’m letting the burnweed improve the soil, and adding a deep layer of straw to try and improve this area for next year. Juglone tolerant three sisters would also be wonderful and I think it does exist out there.
The Three Sisters technique doesn’t really work for me because I am actively breeding and selecting—it would be a total mess to keep track of everything.
The real question is what you want to achieve with this method: is it just drought resilience for the varieties, or do you also need a solid food yield? If it’s the latter, I’ve developed several upgrades. The Native Americans probably didn’t use these, but they work great for me:
Mulching with dry grass or dry weeds: This drastically cuts down water evaporation, stops the clay soil from cracking, and reduces the need for watering.
Mycorrhizal fungi on plant roots: (I use a local brand called Symbivit, not sure what the exact equivalent is in your area). The fungi help the plant in many ways—better transport of water and nutrients from dry soil where the roots can’t reach on their own, drought survival, and steady water release after heavy downpours. The symbiosis lasts for the plant’s entire lifespan, but the fungi hate chemicals—including chlorinated tap water.
Ziz-zag double rows for corn: For better pollination, I use a zig-zag double row spaced at 30 cm (1 ft) with a 70 cm (2.3 ft) path between them. I plant two seeds per spot and then select the more vital seedling.Pruning squash vines: I pinch off the main squash shoots after they reach one meter (3.3 ft) in length.
My custom methods: I came up with the "Dry Mold" (Vyschlá formička) method for gentle transplanting, the "Controlled Drought Stimulation" method supported by mycorrhizae for faster flowering, and the "Gravity Matchmaker" (Gravitační seznamka) method for applying fungi to the roots after planting or direct sowing. This can also be used retroactively.
Biochar sandwich beds: I place nutrient-rich, pre-charged biochar deep inside a “sandwich” bed. It acts as a massive reservoir for both water and nutrients. For beds with clay content, I water heavily at the start—for example, my 12-meter sunken bed gets a baseline of 250 liters of water in early spring. Then I transplant the tomatoes and keep watering them for at least 14 days to properly establish both the plants and the mycorrhizal fungi. Only after this initial period do I switch to full dry farming, pruning the indeterminate varieties down to a single fruit cluster.
The black walnut itself is a valuable food source. A lot of people where i live complain about it, but it’s very easy to gather a lot of high quality, storable food with very little effort. You just need a well designed nut cracker to extract the kernels. Walnuts have protein and fat, both are very valuable for sustenance. Very carefree tree.
