Why every printed planting calendar is wrong for you

Search for a garden planner and you will find thousands of them, most with a planting calendar already filled in. Sow tomatoes in March. Plant out in May. Direct sow beans the third week of April.

Those dates are wrong for almost everyone who buys them, and the reason is easy to demonstrate.

One state, 146 days

North Carolina is not a large state. You can drive across it in a day. NC State Extension publishes average frost dates for 107 weather stations spread across 88 of its counties, and I plotted the growing season each one records.

The Transou station up in Ashe County, in the mountains, records 140 frost-free days. The Hatteras station out on the Outer Banks records 286. That is a spread of 146 days — about twenty-one weeks — inside the borders of a single state.

It gets worse, or better, depending on how you look at it: the spread exists within counties too. Ashe has a second station reading 152 days. Haywood County has three stations, and they range from 159 to 200. If a printed calendar cannot be right for one county, it certainly cannot be right for a region.

A planting calendar that prints “sow in March” is silently assuming you live somewhere specific. If that assumption is wrong by a few hundred miles, you will either lose a tray of seedlings to a frost or waste six weeks of season waiting.

The two numbers that replace the calendar

Every planting date that matters can be expressed relative to one of two dates:

• Your last spring frost. The average date of the last freeze. Tender plants go out after this.

• Your first fall frost. The average date of the first freeze. Everything tender is finished by this.

Once you have those two, every instruction becomes portable. “Start six to eight weeks before last frost” is correct in Ashe County and correct in Dare County and correct in Oregon. “Start in early March” is correct in one place and misleading everywhere else.

Your county Cooperative Extension office publishes both numbers, and every US state has one. It takes about two minutes to find them and you will use them for as long as you garden at that address.

The fall calculation people get wrong

Spring is intuitive: count forward from the last frost. Autumn is where it goes wrong, because you have to count backwards from the first fall frost, and there is an extra term most people forget.

Take your days-to-harvest from the seed packet. Add about two weeks, because growth slows markedly as day length drops in autumn — a crop that takes 60 days in June does not take 60 days in September. Subtract that total from your first fall frost date. That is the last date worth sowing.

It is almost always earlier than people expect, which is why so many autumn crops go in with real enthusiasm in September and then sit there, half-grown, until the cold takes them.

Two honest caveats

A frost date is an average. That means roughly half of all years are worse than it. It is a planning number, not a promise. Treat the last spring frost as the date after which a freeze becomes unlikely rather than impossible, and keep row cover within reach for a fortnight past it.

Your garden has its own microclimate. A bed against a south-facing brick wall can run two weeks ahead of the published date. A low corner where cold air pools on still nights can run two weeks behind, and cold air really does pool — it flows downhill and collects, which is why the bottom of a slope frosts when the top does not.

Which is the argument for keeping notes. The published date gets you into the right fortnight. Three years of your own records get you to the right week.