CEA HERB Part 2: A guide to increasing the sowing density of culinary herbs

In grocery stores across America, people are increasingly passing over the dried spices and reaching for fresh produce.

Forget dry basil — fresh-cut herbs are the crowning jewel of the family meal.

But this growing demand puts more pressure on culinary herb growers, specifically the hydroponic growers who produce these fresh-cut herbs. These growers may find it expensive (or impossible) to add more hydroponic growing space, so they may ask: How can I most efficiently use the production space I have to maximize profits?

In this second article of our six-part series on improving quality and yield of culinary herbs, we will focus on seeding density. One low-cost solution to increase yield is to sow more seeds.

Culinary herbs grow well at high densities in the field (as many as 500 basil plants per square meter), and hydroponic basil research has also achieved greater yield per area with more seeds per cell. Plus, herb seeds are relatively inexpensive. Therefore, increasing sowing density seems like a promising, low-cost way to increase yield, right?

Unfortunately, it’s not that simple.

We often think of leaves as efficient little factories: absorbing light to grow. But between 5% and 25% of light is not captured and instead passes through leaf tissue. As light travels through leaves, some is absorbed, reflected or transmitted.

So, when a leaf is shaded by another leaf above it, the plant can sense a greater proportion of far-red light. The molecule that perceives this light change is phytochrome: a small, tangled-looking protein that is fat on one side. When far-red light hits the phytochrome molecule, the phytochrome moves out of the nucleus and into the cytosol of the plant cell.

If there is a lot of far-red light (from a lot of shading), many hundreds of phytochrome molecules move out of the nucleus — triggering the shaded plant to stretch and elongate. This is why shaded plants are more “leggy”— they sense more far-red and are stretching to compete for sunlight.

When seeds are sown at high densities, this same thing happens: the plants shade each other and cause their neighbors to stretch and elongate. More stretching can result in a greater proportion of weak stem growth — and less growth of those delicious leaves that consumers want to eat.

Therefore, it is crucial for growers to sow the right number of seeds: enough to increase yield without compromising quality. But where does this trade-off lie? How many seeds is too many?

Sowing density study

We tested several popular hydroponically grown culinary herbs: sage, cilantro ‘Santo’, parsley ‘Giant of Italy’, sweet basil ‘Italian Large Leaf’ and purple basil ‘Red Rubin’. We sowed 1, 5, 10, 15 or 20 seeds per cell in Oasis phenolic foam substrate and placed them in a greenhouse ebb-and-flow hydroponic system.

After several days of germination (10 for green basil, 14 for purple basil and cilantro, 21 for sage and parsley), we transplanted the seedlings into deep water culture hydroponic systems. They stayed in these hydroponic systems until they were ready for harvest: 16 days for cilantro, 18 to 25 for basil, 25 for sage and 28 for parsley (Figure 1).

At the end of each species’ growth period, we harvested each cell and measured plant weight, height, stem thickness and more. With more seeds per cell, we saw an increase in yield but a decrease in quality.

What surprised us was how quickly the trade-off occurred. Increasing the sowing density from one to five seeds per cell caused a slight (but measurable) decrease in quality.

As we added more seeds per cell, the individual plants became smaller; each individual plant weighed less, and individual basil and parsley plants also became shorter with more seeds per cell.

Additionally, for basil, the plants were made up of proportionally more stem and less leaf. More stem is not ideal: people enjoy herbs for their flavorful leaves, and stems can be tough and undesirable to eat fresh. This issue isn’t black or white — maybe slightly more stem is acceptable (or even unnoticeable) to consumers.

High sowing density could also potentially make plants more difficult to harvest. In our treatments with more seeds per cell, the parsley, cilantro and sage stems were thinner, making the plants more “floppy” (Figures 1 and 3).

We also found that green and purple basil were not equally sensitive to sowing density. Although the general data trends were the same between green and purple basil, the green basil had a much more dramatic yield increase than the purple basil. For example, the per-area dry weight was 3.45 times greater when green basil was sown at high density, but the purple basil dry weight was only 2.2 times greater (Figure 2).

If growers want to maximize yield and are OK with a slight decrease in quality, they should consider sowing more seeds per cell. However, there is still a lot we don’t know. Does sowing density impact herb flavor or aroma? Do consumers even notice these quality differences?

Take-home message

Based on the data from this experiment, we can give some general recommendations.

For the culinary herbs in this study, your choices will depend on your production goals. If you want the highest yield, sow 20 seeds per cell. But if you want to balance plant quality while also growing a decent yield, we recommend 10 (sage) and 15 to 20 (basil, cilantro and parsley) seeds per cell (Figures 2 and 3).

Results will likely vary based on substrate, dibble hole size, production environment (light, temperature, fertility, etc.), production schedule and more. We encourage small-scale production trials before implementing recommendations across your operation.

There is so much left to learn, and here at the Walters Lab at the University of Tennessee, we’re excited to keep exploring ways we can produce culinary herbs (and other food crops) more efficiently.

This article appeared in the March 2025 issue of Greenhouse Management magazine under the headline "Solid starts."

Samson Humphrey is a Ph.D. student and Kellie Walters is assistant professor of controlled environment plant physiology in the Department of Plant Sciences at the University of Tennessee, Knoxville. The authors gratefully acknowledge Akela Martin, Griff Hagen, Jianyu Li, Lauren Carver, Sarah Armstrong and Spencer Givens, as well as JR Peters for fertilizer, Oasis Grower Solutions for substrate, Ball Horticultural Company for seeds and Hydrofarm for hydroponic production supplies. This work was funded by the USDA SCRI award 2022-51181-38331 and USDA NIFA TEN00617.