Milpa Gardens

Rotation belongs to the ground, not to the bed

Most garden apps attach crop history to a bed. Rearrange the beds and the history is gone. Here is why this one attaches it to the soil instead, and what that buys you.

The bug you only meet in year three

Draw your garden in almost any planner. Four raised beds, tidy rectangles, each one a record with a name. Plant tomatoes in Bed 2. Next winter you decide the layout was wrong, pull Bed 2 out, widen the path, and put two smaller beds where it was.

Where did the tomatoes go?

In a tool that hangs history on the bed, they went nowhere. Bed 2 no longer exists, and the fact that a particular patch of soil grew nightshades last year has been deleted along with the rectangle you drew. The new beds start clean, the rotation warning never fires, and you plant tomatoes back into the same ground and find out the hard way.

This is why so many garden planners are single-season toys. They are perfectly good until the first redesign, and rotation is the one feature that requires surviving redesigns.

Three things, not one

This planner separates what most tools collapse into a single object:

Occupancy attaches to the plot's coordinates, not to the bed. When you delete a bed, you delete a rectangle; you do not delete what grew there. When you ask "can I plant tomatoes here?", the planner intersects the patch you are pointing at against everything that has ever grown on that ground, geometrically, and answers from the soil's history rather than from a label.

Rearrange the yard as often as you like. The history is in the dirt, which is also where the actual problem is.

What the history is for

Rotation intervals exist because pathogens and pests persist in soil. The interval is family-based, because that is the level at which the persistence works, and the corpus keeps one rule per family with the interval its literature supports:

Do not plant nightshades in ground that held nightshades within 3 years.

Promising · costly if ignored → warn

Mechanism: Alternaria solani (early blight) and Septoria lycopersici overwinter on tomato debris and solanaceous weeds; a 2-3 year break plus debris removal starves this inoculum. Fusarium has a very narrow host range, so rotating away from the host is effective. Verticillium is the exception: microsclerotia persist ~10 years and it infects 300+ host families, so a garden-scale rotation does not control it.

Evidence: confirmed against 5 sources in the research-based literature.

the 5 sources

Clemson HGIC, 'Tomato Diseases & Disorders', hgic.clemson.edu/factsheet/tomato-diseases-disorders (read 2026-07-11): 'tomatoes and related vegetables, such as potatoes, peppers, and eggplants, should not be planted on the same land more than once in three years.'

NC State Extension, 'Early Blight of Tomato', content.ces.ncsu.edu/early-blight-of-tomato (read 2026-07-11): 'Rotate every 2-3 years.'

NC State Extension, 'Septoria Leaf Spot of Tomato', content.ces.ncsu.edu/septoria-leaf-spot-of-tomato (read 2026-07-11): 'The fungus is known to overwinter on infected tomato debris and solanaceous weed hosts'; 'Rotate away from tomato for 2 or more years.'

NC State Extension, 'Verticillium Wilt of Tomato and Eggplant', content.ces.ncsu.edu/verticillium-wilt-of-tomato-and-eggplant (read 2026-07-11): 'The pathogen can survive in the soil without a host for up to 10 years in the form of structures called microsclerotia'; 'affects more than 300 host families.'

UMass Extension, 'Solanaceous, Verticillium Wilt', umass.edu/agriculture-food-environment/vegetable/fact-sheets/solanaceous-verticillium-wilt (read 2026-07-11): 'Verticillium wilt tends to affect a wide range of hosts and can infect many unrelated plants. Fusarium spp., on the other hand, have very narrow host ranges.'

remedy

Plant nightshades in different ground, or wait until 3 years have passed since this ground last held them. Clear and remove the old debris either way — the inoculum overwinters on it.

See this rule in the planner

Do not plant cucurbits in ground that held cucurbits within 2-3 years.

Promising · costly if ignored → warn

Mechanism: Melittia cucurbitae overwinters as a pupa in the soil where its host grew and emerges the next summer; Anasa tristis overwinters as an adult in sheltered sites (plant debris, under boards or rocks) and flies to cucurbits in spring. A host-free break cuts local carryover, but both are strong fliers that re-invade from a distance, so rotation is only partial and works best paired with fall sanitation (destroy spent vines).

Evidence: confirmed against 3 sources in the research-based literature.

the 3 sources

UMN Extension, 'Squash vine borers', extension.umn.edu/yard-and-garden-insects/squash-vine-borers (read 2026-07-11): 'squash vine borers spend the winter in the soil near their host plants ... Practice rotation ... planting cucurbits in different areas of your garden if possible, or alternate seasons'; 'rotate away from the area and destroy crop residue.'

Ohioline (Ohio State Univ.), 'Squash Vine Borer', ohioline.osu.edu/factsheet/ent-0106 (read 2026-07-11): 'Crop rotation away from previously infested fields helps reduce local populations, as overwintering pupae remain in the soil. However ... fields often need to be rotated more than 5 miles away as SVB can fly into fields from long distances.'

UMN Extension, 'Squash bugs', extension.umn.edu/yard-and-garden-insects/squash-bugs (read 2026-07-11): 'Squash bugs can live through the winter as adults in sheltered places, such as under plant debris, around buildings, or under rocks. When adults come out in the spring, they fly to growing cucurbit plants.'

remedy

Plant cucurbits in different ground, or wait 3 years. Destroy the spent vines in autumn and clear debris and boards nearby — the pupae overwinter in the soil where the crop grew, the adults in the shelter around it. Rotation is only partial; both pests fly in from a distance.

See this rule in the planner

Do not plant brassicas in ground that held brassicas within 4 years (extension recommends 4-7, and even that is not a cure).

Promising · costly if ignored → warn

Mechanism: Plasmodiophora brassicae forms thick-walled resting spores that survive in soil more than 10 years. They germinate in wet, acidic soil (pH < 7); liming above ~7.2 suppresses infection. Spores outlast any garden rotation, so a host-free break reduces but does not clear inoculum.

Evidence: confirmed against 3 sources in the research-based literature.

the 3 sources

NC State Extension, 'Clubroot of Brassicas', content.ces.ncsu.edu/clubroot-of-brassicas (read 2026-07-11): 'Plasmodiophora brassicae is capable of surviving in the soil for more than 10 years by producing thick-walled, resting spores'; 'Infested fields should be rotated away from brassica for a minimum of 4-7 years, but this is not guaranteed to prevent disease'; 'Liming to raise soil pH to 7.[2]'.

Cornell Vegetables, 'Clubroot of Crucifers', vegetables.cornell.edu/pest-management/disease-factsheets/clubroot-of-crucifers (read 2026-07-11): 'the pathogen survives for many years even in the absence of host crops'; 'rotation of at least 7 years out of susceptible crucifers may be effective.'

UMass Extension, 'Brassicas, Clubroot', ag.umass.edu/vegetable/fact-sheets/brassicas-clubroot (read 2026-07-11): 'Long (5-7 years) rotations between brassica crops'; 'Maintain a high pH by regular applications of lime.'

remedy

Plant brassicas in different ground, or wait at least 4 years — extension recommends 4 to 7. Rotation alone is not a cure: the resting spores outlast it, and raising soil pH above ~7.2 suppresses infection.

See this rule in the planner

These are graded honestly. They are not "well established" claims with a single crisp number behind them; the sources give ranges, and the brassica interval in particular comes with an extension recommendation the rule quotes and an admission that even the long interval is not a cure. The rule says so rather than rounding to a number that sounds authoritative.

The rotation advice you were given is probably too strong

The strict version of rotation folklore is that everything must move every year. That is not what the literature supports, and following it rigidly in a small garden produces worse plans, not better ones.

You must rotate every crop to new ground every single year.

Contested

You don't have to rotate everything every year — intervals are family- and pathogen-specific. This corpus keeps the ones with real mechanisms (tomatoes ~3 yr, cucurbits 2-3, brassicas 4-7); a yearly shuffle of a low-pressure crop buys nothing, and perennials must never be rotated.

what is actually known

Rotation intervals are pathogen- and family-specific, not a blanket yearly rule. This corpus encodes the ones with named mechanisms: nightshades ~3 years, cucurbits 2-3, brassicas 4-7, with an allium/carrot family break that needs its own rotation. A yearly shuffle of a crop under no soilborne pressure buys nothing, and perennials must not be rotated at all — tillage destroys the crown. Rotate against the specific pathogen, on its interval.

what this rests on - one pointer

rotation intervals are family- and pathogen-specific; see other rules here

A four-bed garden cannot give brassicas a four-year interval by rotating in a cycle. Something has to give, and knowing which rules are the load-bearing ones lets you give in the right place.

The break year, as a first-class thing

When ground has taken a rotation load, the planner does not just refuse and leave you with a warning. It offers the restorative break: a rest year that puts a nitrogen fixer on the ground, optionally with a hardy overwintering crop, so the season is not wasted while the interval runs.

That is the difference between a validator and a planner. Saying "no" is easy; saying "not this, but here is what this ground wants instead" is the part that took the data model.

The trade

Attaching history to ground rather than beds costs more. Every occupancy record carries geometry; every rotation query is a spatial intersection rather than a lookup by id; the model is harder to explain in a sentence. In exchange, the app is still correct in year five, after you have moved everything twice.

That is the whole bet.

All explainers · Plan a bed