How to grow microgreens

45 of the 62 crops here are described as sharp, peppery or bitter in their own tasting notes. And the one study linking light to bitter chemistry was run at 30 to 90 PPFD, entirely below the 234 to 358 this site recommends.

Take the blackout days from your crop, not from a general rule. Across the 55 crops here that publish a figure the range runs 0 to 14 days, and 29 of them top out at 4 or 5.

Published chlorine figures for food contact run 50 to 200 ppm depending on the job, and the required wet contact time is measured in seconds. Grower pages recommend soaking for 10 to 20 minutes, which is 20 to 170 times longer and traceable to nobody.

Roots do not make a new shoot; a surviving node does. The one documented regrowth path takes 3 to 4 weeks, which is about twice as long as simply sowing a fresh tray.

In one peat trial, taking the feed away cost cabbage 7.9 percent of its yield and arugula 47.4 percent, while cutting nitrate in the cut crop by 98 to 227 fold. Feeding is the largest lever anyone has measured on either number.

An on-farm study measured a 10 by 20 tray at 7.31 oz, within 4 percent of this directory's median. It also tested whether fodder could cut feed costs for laying hens: it did not, and egg production fell.

The largest yield gain in the harvest-stage literature is 80.1 percent. Divide it by the five extra days it cost and it is worth 5.1 percent. Nutrient concentration does not peak; some compounds rise, some fall, and nobody reports the total per tray.

A heated home runs about 68F and a mat adds 9 to 20 above ambient, which lands at 77 to 88. Spinach stops germinating at 75. Nothing shows a mat produces an earlier harvest or more crop in a room already in range.

Read the harvest field on all 62 crop entries here and only 15 of the 61 that publish one fit entirely inside 7 to 14 days. The real spread is 4 to 35 days, and the shelf cost between the fastest and slowest crop is 36.6-fold.

Cut height is the one harvest decision shown to change the microbiology: cutting higher above the medium measurably lowers bacterial counts. Here is where to put the blade, what to cut with, and how to keep the cut clean.

In one replicated trial, moving from peat-perlite to coir cost pea 32.7 percent of its yield and gained radish 4.5 percent. Two extension services cannot even agree whether a mat counts as hydroponic, and the word is doing two jobs at once.

Checked against this directory's own fields, one commercial brassica blend's five components overlap on exactly one day, and a three-crop spicy blend keeps to its shortest member at 2 to 3 days while another component lasts 14.

A high tunnel has been recorded at 120 to 140F in March and April and near 0F overnight in January, against a recommended 64 to 75F band. Of seven pages on this subject, none publishes a loss rate or a saving.

For a tray that is already affected the answer is isolate and dispose, then clean the system before the next crop. Spraying is not an option on a seven-day crop, and prevention is where all the leverage actually is.

Every page lists the same seven causes and none tells you which one happened. Where the bare patch sits, and what an excavated seed looks like, rules most of them out in about two minutes.

Every source says reuse loose coir and throw the mat away. Priced from this directory, that permits reuse on the cheapest material and forbids it on the dearest. The one experiment that measured anything covers 4 of the 12 materials an extension service names.

Can I eat this, can I sell this, and will it recover are three separate decisions, and not one of the six pages checked kept them apart. The rescue everybody recommends is the one the evidence argues against.

A windowsill and a shallow tray will grow microgreens; a rinsed jar grows something else. Jars produce sprouts, eaten whole, which FDA treats as a distinct risk category, so the method decides what you end up with.

The standard advice is to soak overnight, 6 to 12 hours. Counted against this site's own per-crop figures, 41 of 62 crops soak for zero hours and only 12 have a soak time that starts above zero. And the one ten-species trial found soaking moved emergence forward without moving the harvest.

Every medium comparison argues about yield, where the honest spread between reasonable media is 13 to 16 percent. The measurement nobody quotes is nitrate, which moved about twofold across three crops in two independent trials, and not always in the same direction.

The published seed rates in this directory span 409-fold, from 1.1 g to 450 g a tray, so sow thick is not an instruction. Density has been measured properly about three times, the optima are six times apart, and it turns out to be a food-safety variable as well as a yield one.

Six pages, six techniques, five causal explanations and not one study between them. The only controlled result says weight during blackout reduces attached hulls, and that more of it reduces the harvest.

Check the height against the crop before you treat it as a fault, because several crops simply finish short. If the tray is genuinely stunted, low light is the one cause it cannot be: too little light makes a tray taller.

About 70°F sits inside the published germination range for 39 of the 44 crops here that have one, so a single setting runs most of a mixed bench. Airflow is the instruction everyone gives and nobody quantifies.

Two 1020 trays on sale today have inside floors of 187.6 and 192.6 square inches against the nominal 200, so every per-tray rate is quoted in a unit that varies. The one depth trial found yield still rising at 2.3 inches, which is above almost all published advice.

In the one dose-response study, broccoli tolerated 37 mg/L of residual chlorine before declining and vegetable seed germination was unaffected at 77. The federal drinking water ceiling is 4.0. Of eight ranking pages audited, none cited a source.

Overwatering is blamed for more failed trays than anything else in this category and has no numerical definition anywhere. The published volumes disagree by four times, and the reason is the medium: hemp holds nearly three times as much water per gram as coir does.

Pick from the 5 crops that reliably finish inside a two-week unit, and plan for the weekend, because every published protocol assumes daily care and a school week is five days in seven.

At the same outdoor temperature, measured indoor temperatures across 327 homes spanned 8 degrees Celsius in summer and more than 10 in winter. Three of six year-round guides never mention room temperature at all.

The 48 by 18 inch rack recommended for four 1020s per shelf is 2 to 3 inches too shallow to hold one without overhang. A 16-position rack in 8 square feet returns about 8 ounces a day at this directory's median rate.

Nobody routinely tests microgreen seed for pathogens, food grade is not a certification, and organic makes no measured difference to the microbes. What matters is a traceable, explicitly untreated lot. Plus what 57 seed searches on this site actually returned.

Long thin stems and pale leaves are a light problem, not a water problem. The stems will not shorten once stretched, and the internet's usual answer is not supported by any source that says so.

Intensity has a measured saturation point, hours change yield more than spectrum does, and the same five-day blackout raised sunflower yield 13 percent while cutting arugula's by 24. Everything published about lighting this crop is about which fixture to buy.

The white fuzz on a microgreen tray is usually root hair, and there is no test at the tray that proves it. What the mist test can and cannot tell you, and the one word that separates the two.

The cue is the leaf, not the calendar. Published day ranges run from 6 to 23 depending on the crop, and the one trial that actually tested harvest stage found the best result at about 75 percent true-leaf emergence, for two crops.

Four different faults end with a tray lying down, and their remedies are opposites. The collar is what tells you which one you have, and a wilted tray that stands back up was thirsty, not diseased.