Fall vegetable garden timing

Fall timing is a deadline run backward

A spring window opens after a last freeze and stays open until heat or the end of the written range. A fall window closes. The crop has to finish, or at least reach a usable size, before the cold that ends it. Texas A&M AgriLife publishes that idea in two forms, and they do not always land on the same day. One form is a week count before “the first freeze date.” The other is a calendar date that depends on which part of Texas the guide’s map put you in. Both are shown when both exist, and they are kept separate.

The week-count form is Table 8 of Joseph Masabni’s Texas Home Vegetable Gardening Guide. Fall planting is “weeks before the first freeze date.” The guide does not name a 10%, 50%, or 90% probability. The week offset is applied to the median first 32°F date from the nearest qualified station, and the page says so. The fall calculator subtracts the days you enter from the 10%, 50%, or 90% first-32°F date you pick. That tool answers a days-to-maturity question. Table 8 remains the week table.

The calendar form is Table 1 of the Fall Vegetable Gardening Guide for Texas, by Larry Stein and Joe Masabni. The copy used on this site calls those cells average fall planting dates. Table 3 of that guide ties regions to USDA zones: region I is zone 6, region II is zone 7, region III is zone 8, region IV is zone 9a, and region V is zone 9b. Zones warmer than 9b are not in that table. A city page matches the 2023 ZIP zone to one of those regions and prints the cell. If Table 1 has no date for that crop and region, the page says the crop is not recommended in the table. Irish potatoes, for example, have dates for regions II, III, and IV in the copy stored here, and none for I or V.

One city, two Texas instructions

Use Austin-Bergstrom, USW00013904, only as a station example. The median first 32°F date there is November 16. The 10% first 32°F date is November 1. The 90% date is December 5. Austin’s city page may name a nearer station. The arithmetic below is what the median method does when November 16 is the anchor. It is not a claim that every Austin garden freezes that week.

Table 8 places tomatoes 12 to 14 weeks before the first freeze date. Fourteen weeks before November 16 is August 10. Twelve weeks before November 16 is August 24. So the week-count method, applied to the median, gives a tomato planting window of about August 10 through August 24.

Table 1, for a garden whose 2023 zone is 9a, uses region IV. The tomato cell stored from that table is July 10. For 9b, region V, the cell is August 1. July 10 is earlier than the week-count window. August 1 is a week earlier than August 10. The two publications do not collapse into one recommended day. The week-count window moves if your station’s median first freeze moves. The Table 1 date does not. It moves only when the zone region changes. A 9a garden in a colder pocket and a 9a garden on a south wall receive the same Table 1 date. They do not receive the same station date if their nearest qualified stations differ. Read both lines. Do not let a calculator hide one of them.

Bush beans show a wider gap. Table 8 says 8 to 10 weeks before the first freeze. From November 16 that is September 7 through September 21. Table 1 says September 10 in region IV and October 1 in region V. Region IV’s September 10 falls inside the week-count window. Region V’s October 1 does not. Both dates belong on the page. Neither publication names a single winner.

Broccoli is the other direction. Table 8 says 10 to 16 weeks before the first freeze, which from November 16 is July 27 through September 7. Table 1 says October 1 in region IV and November 1 in region V. Those Table 1 dates are later than the week-count window built from this station. Possible reasons include a different freeze date inside the fall guide’s own map, a different meaning of “planting date” for a transplant than for a seed, or a regional adjustment the week table does not make. The cells do not say which of those reasons applies. The crop page prints both dates with their citations.

Counting days to maturity back from a freeze

The fall back-calculator is the days-to-maturity tool. You type a number of days. You may type extra days. The default extra is zero. The tool subtracts that total from the first-32°F probability you select. It does not look up AgriLife Table 9, and it does not decide whether your number starts at seeding or at transplanting. Reading days to maturity is about that choice. If you skip it, you will aim the subtraction at the wrong event.

Take a Texas tomato and Table 9’s 70 to 90 days. The table does not say the count starts at transplanting. University of Maryland Extension says that for tomatoes the packet number usually does start at transplanting. If you accept Maryland’s reading and type 80 days, and you select the median first 32°F date of November 16, the calculator returns August 28. That is a target date to have an 80-day tomato transplanted, not a date to sow the seed, and not a date that already includes a frost buffer. August 28 is later than both the Table 8 window (August 10 to August 24) and the region IV Table 1 date (July 10). Landing later means you are asking an 80-day crop to finish against a tighter margin than AgriLife’s week table allowed. The calculator did not fail. It answered the question you typed. The week table answered a different question, in weeks, without a variety attached.

If you select the 10% first-freeze date instead, November 1, the same 80 days lands on August 13. You have chosen the early-freeze side of the distribution on purpose. Say so when you write the date on a label. Selecting December 5, the 90% date, and subtracting 80 days lands on September 16, which is a more optimistic frost and a later planting. The 90% first-freeze date means a first 32°F on or before that day is the common case, so a crop that needs until December 5 is depending on the late tail. The calculator will perform the subtraction anyway. The probability definition will not change to match the hope.

Extra days are for a margin you are willing to name. Ten extra days on an 80-day tomato is a 90-day subtraction. From November 16 that is August 18. That margin is one you choose. AgriLife’s table does not add it for you.

What has to be true before the date is usable

Use the station named on your page. Bergstrom is the worked example in this guide. A city page uses the nearest qualified station, and the miles are printed. A difference of 15 miles can move a median by days, so the arithmetic should start from the date on that page.

The zone used for Table 1 has to be the 2023 ZIP zone, not a memory of an older map. Region IV is 9a. Region V is 9b. A half-zone change changes the column. The 2023 map guide explains what that half-zone measures. It does not measure the first-freeze date. Using it as the index into Table 3 is following AgriLife’s own crosswalk, not treating the zone as a frost calendar.

The crop has to be in the table. Sweet potatoes are marked “not recommended” for fall in Table 8, so no fall week offset is shown. Garlic is fall-only in that table, 4 to 6 weeks before the first freeze. Cauliflower has no spring recommendation and a fall window of 10 to 16 weeks before. An empty cell is the guide’s answer.

Other states use their own cited guides, not the Texas region map. Calendar dates are shown as printed. Weeks before a frost are applied to the median and labeled. The fall calculator can count days back from the station’s first-32°F date wherever that station is shown. Beside a cited window, that count sits next to the extension date. On its own, it is the station date minus the days you entered.

A sequence that keeps the documents separate

1. Read the median, 10%, and 90% first 32°F dates on the city page, including the station name and the miles. 2. If the state is Texas and the zone is 6a through 9b, read the Table 1 date for your region on the crop page. 3. Read the Table 8 week window, already converted on the page, and notice whether it overlaps Table 1. 4. If you are using days to maturity, type the number and pick the probability in the fall calculator. Decide first whether the number starts at sowing or at transplanting. 5. Plant inside an overlap when you want both documents satisfied. If they do not overlap, follow one document and keep its citation. A midpoint between them is not an AgriLife date.

Heat still ends a fall sowing that was timed only against frost. These sources are about the cold end of the season. A spinach date in December can match the freeze table and still meet soil that is too warm for germination in region V in a hot year. The cited guides do not give a soil-temperature maximum for that case.

Sources

• Joseph Masabni, Texas Home Vegetable Gardening Guide, Table 8 and Table 9: https://agrilifeextension.tamu.edu/library/gardening/texas-home-vegetable-gardening-guide/
• Larry Stein and Joe Masabni, Fall Vegetable Gardening Guide for Texas, Table 1 and Table 3: https://agrilifeextension.tamu.edu/browse/featured-solutions/gardening-landscaping/fall-vegetable-gardening-guide-for-texas/
• University of Maryland Extension, on which crops count days to maturity from transplanting: https://extension.umd.edu/resource/when-plant-vegetables
• NOAA NCEI station USW00013904, 1991–2020 normals, for the November dates in the worked example.