Growing Collard Greens in a Greenhouse: PAR, DLI, CO₂ and VPD Guide

Collard greens (Brassica oleracea var. viridis) are productive leafy Brassicas capable of growing under substantial light, but there is no scientifically established PAR, CO₂ and VPD recipe that applies to every greenhouse and every growth stage.

Published collard research gives us useful evidence about:

  • Daily Light Integral (DLI)
  • photoperiod
  • CO₂ enrichment
  • light spectrum
  • irrigation
  • salinity
  • photosynthetic acclimation

But the evidence does not support simple rules such as:

high PAR creates tough leaves

or:

high VPD directly causes fiber accumulation.

Leaf texture is influenced by a much larger system involving:

  • cultivar
  • leaf age
  • temperature
  • water status
  • mineral nutrition
  • light
  • harvest maturity

For greenhouse collards, the stronger strategy is therefore to measure the actual environment and connect it with crop response.

Quick Reference

VariableWhat It Tells YouWhat Collard Research Supports
PPFD / PARPhotosynthetic light reaching the crop nowCollards can use substantial light, but PPFD must be interpreted together with photoperiod
DLITotal photosynthetic light accumulated through the dayBaby-leaf collard research directly tested 6–30 mol/m²/day; yield gains began diminishing above about 24 DLI
CO₂Carbon available for photosynthesisDirect collard research compared approximately 350 vs 700 ppm and showed strong growth responses plus photosynthetic acclimation
VPDAtmospheric evaporative demandUseful for monitoring water demand, but no validated collard stage-specific optimum is established
Root-zone waterWater available to support growthDirect greenhouse research shows irrigation level substantially affects biomass and leaf area
SpectrumDistribution of wavelengthsRed-blue light can increase collard leaf dimensions, while blue light can alter glucosinolate profiles
Harvest stageDetermines leaf size and maturityBaby leaf, microgreen and mature collard results should not be treated as interchangeable

These are research references rather than universal crop specifications.

1. Collard Greens Are Not Simply “Tougher Kale”

Collards and kale are closely related members of Brassica oleracea, but they are different crop types.

Collards have been selected for characteristics such as:

  • large leaves
  • broad leaf blades
  • relatively smooth leaf surfaces
  • repeated leaf harvest

Their morphology, commercial harvest stage and quality objectives differ from many kale cultivars.

Therefore, kale lighting data can provide context, but collard-specific research should be used whenever possible.

2. PAR and DLI Answer Different Questions

PPFD tells you how much photosynthetically active light reaches the crop at one moment.

It is measured in:

µmol/m²/s

It answers:

How much photosynthetic light is reaching the collard canopy right now?

DLI measures the amount of PAR accumulated during the entire day.

It is expressed in:

mol/m²/day

It answers:

How much photosynthetic light did the collards receive today?

For constant lighting:

DLI = PPFD × photoperiod × 0.0036

For example:

200 µmol/m²/s × 16 h
11.5 mol/m²/day

300 µmol/m²/s × 16 h
17.3 mol/m²/day

400 µmol/m²/s × 16 h
23.0 mol/m²/day

In a greenhouse, natural PPFD is constantly changing, so continuous PAR logging is preferable when the real question concerns daily light.

3. Collards Have Direct DLI Research

One useful controlled-environment experiment studied baby-leaf:

Collard ‘Flash’

at five DLIs:

6

12

18

24

and:

30 mol/m²/day.

The photoperiod was fixed at:

20 hours.

Collards were harvested at the baby-leaf stage rather than as mature full-size plants.

This distinction is important.

4. Increasing DLI Strongly Increased Baby-Leaf Yield

At Day 12, estimated collard fresh yield increased approximately as follows:

6 DLI → 0.22 lb/ft²

12 DLI → 0.35 lb/ft²

18 DLI → 0.53 lb/ft²

24 DLI → 0.71 lb/ft²

30 DLI → 0.80 lb/ft²

This is strong direct evidence that daily light can substantially affect collard productivity.

But the response was not perfectly linear.

5. Yield Gains Began to Diminish Above About 24 DLI

Collard yield continued to increase at:

30 DLI

but the increase from:

24 → 30 mol/m²/day

was relatively small compared with earlier increases.

The researchers specifically noted that the economic benefit of lighting above approximately:

24 mol/m²/day

might be limited for baby-leaf collards.

That is a much more useful conclusion than:

“More light always means better collards.”

6. Higher DLI Also Shortened Time to Harvestable Leaf Height

The same experiment measured how quickly collard leaves reached approximately:

4 inches / 10 cm

in height.

Approximate days required were:

6 DLI → 14.8 days

12 DLI → 14.2 days

18 DLI → 13.0 days

24 DLI → 12.2 days

30 DLI → 12.0 days

So increasing DLI could shorten the production cycle.

But again, most of the benefit had already been achieved by approximately:

24 DLI.

7. This Does Not Mean Mature Collards Require 24 DLI

The experiment used:

very young baby-leaf collards

at very high plant density.

That is not the same as:

  • mature field-size collards
  • long-cycle greenhouse plants
  • repeated outer-leaf harvest

Therefore, approximately 24 DLI should be described as:

a useful baby-leaf research reference

rather than:

the mature collard optimum.

8. Microgreen Collards Give Us Another Useful — but Different — Reference

A separate experiment studied:

Collard Greens ‘Vates’

as microgreens.

Researchers compared:

14 vs 21 mol/m²/day

and:

16-hour vs 24-hour photoperiods.

Fresh weight ranged approximately from:

20.4 g

to:

24.7 g

across treatments.

Both DLI and photoperiod influenced production.

Again:

microgreens are not mature collards.

These data demonstrate collard light responsiveness, but exact values should not be transferred directly to full-size greenhouse plants.

9. Higher DLI Also Increased Dry-Matter Percentage in Microgreens

For Vates collard microgreens, increasing DLI from:

14 → 21 mol/m²/day

increased dry-matter percentage.

For example, under a 16-hour photoperiod:

14 DLI → about 9.9% dry matter

21 DLI → about 12.0% dry matter.

Under continuous light, dry matter increased even further.

This is important because:

fresh weight

and:

dry matter

do not always respond identically.

10. Higher DLI Can Reduce Lighting Resource Efficiency

The same microgreen experiment calculated electrical-use efficiency.

For collard greens, increasing DLI increased electricity cost per gram of fresh biomass.

That means:

maximum biological yield

and:

maximum lighting efficiency

are not necessarily the same production target.

This is especially important for greenhouse supplemental lighting.

11. Do Not Translate Young-Plant Research Into a Mature Leaf Texture Rule

Baby-leaf and microgreen experiments tell us a great deal about:

  • biomass
  • crop timing
  • morphology
  • dry matter

They do not prove that mature collards become fibrous at a specific PPFD or DLI.

Therefore, statements such as:

“High PAR causes collard fiber buildup.”

should not be presented as research-established facts.

Leaf toughness needs to be studied directly if texture is the production question.

12. Light Spectrum Also Matters

A recent study directly included:

B. oleracea var. viridis — collard

alongside kale and cabbage.

Plants were grown under:

  • red LEDs
  • blue LEDs
  • red + blue LEDs
  • a control treatment

The light spectrum affected both:

plant morphology

and:

glucosinolate composition.

13. Red + Blue Increased Collard Leaf Dimensions

For the collard cultivar studied, combined:

red + blue

lighting increased:

  • leaf length
  • leaf width

relative to several other spectral treatments.

This demonstrates that leaf expansion cannot be predicted from PPFD alone.

Two fixtures delivering similar PAR can still produce different leaf morphology.

14. Blue Light Changed Collard Glucosinolate Profiles

Blue light strongly affected several glucosinolates in the collard cultivar.

The study found high concentrations of compounds such as:

  • gluconapin
  • glucobrassicin
  • glucoraphanin
  • gluconasturtiin

under blue-light conditions.

This matters because Brassica quality involves more than leaf size.

Light environment can also affect:

phytochemical composition.

15. Therefore “Largest Leaf” and “Highest Nutritional Quality” May Be Different Targets

Red-blue treatment favored some growth characteristics.

Blue light strongly influenced several glucosinolates.

That creates the same tradeoff seen in many controlled-environment crops:

the lighting treatment producing the largest leaves does not necessarily maximize every nutritional compound.

Before defining “best light,” determine what you are optimizing.

16. CO₂ Has Direct Collard-Specific Research

Collards have also been studied directly under:

350 ppm

and:

700 ppm CO₂.

In controlled-environment experiments, plants received approximately:

1000 µmol/m²/s PAR

for:

12 hours per day.

Day/night temperatures were approximately:

23/17°C.

This corresponds to an extremely high experimental DLI of approximately:

43.2 mol/m²/day.

That is a physiology experiment — not a commercial greenhouse recommendation.

17. Elevated CO₂ Strongly Increased Collard Biomass

When plants were approximately:

23–27 days old

shoot dry mass was roughly:

10 g/plant

under about 350 ppm CO₂

versus approximately:

20 g/plant

under about 700 ppm CO₂

in the controlled experiments.

This demonstrates a strong growth response to elevated CO₂ under abundant light.

But the physiological response became more complicated over time.

18. Elevated CO₂ Also Caused Photosynthetic Down-Regulation

Collards continuously grown at approximately:

700 ppm CO₂

eventually showed lower photosynthetic capacity per unit leaf area when compared under standardized measurement conditions.

Researchers linked this response to the accumulation of:

  • starch
  • soluble sugars

in leaves.

In other words:

the plant had more carbon than it could immediately use or export.

19. This Is a Critical CO₂ Lesson

Elevated CO₂ can increase growth.

But the leaf’s photosynthetic machinery can also acclimate to prolonged enrichment.

That means:

CO₂ response is dynamic.

It cannot be summarized as:

“700 ppm is always better than 350 ppm.”

The plant’s ability to use additional carbon depends on:

  • light
  • growth rate
  • sink demand
  • developmental stage
  • nutrient status

20. Daily Light Changed the CO₂ Acclimation Response

The USDA collard experiments found that photosynthetic down-regulation under elevated CO₂ was related to the amount of PAR received previously.

After higher-light days, sugar accumulation and photosynthetic feedback inhibition became more apparent.

This directly demonstrates an important greenhouse principle:

PAR and CO₂ interact.

CO₂ should not be interpreted independently from daily irradiance.

21. Does That Mean 700 ppm Is the Best Collard CO₂ Level?

No.

The research shows that:

350 vs 700 ppm produces important physiological and growth differences.

It does not establish:

700 ppm as the universal commercial optimum.

The experiment used unusually strong controlled lighting and was designed to study photosynthetic acclimation.

Use 700 ppm as:

a collard-specific research reference

rather than:

a greenhouse setpoint.

22. CO₂ Monitoring Is Useful Even Without Enrichment

A CO₂ sensor can still answer a useful question:

Does crop-zone CO₂ change when light becomes strong?

Monitor CO₂ during:

  • morning light increase
  • peak sunlight
  • supplemental-light periods
  • greenhouse closure
  • ventilation

Then compare those values with PAR.

A timeline can reveal whether the carbon environment changes when photosynthetic demand rises.

23. What About Water Availability?

Collard-specific greenhouse research strongly supports the importance of irrigation.

One experiment with:

Collard ‘Georgia’

tested irrigation equivalent to:

55%

70%

85%

and:

100%

of pot water capacity

combined with several irrigation-water salinity levels.

The treatments affected:

  • plant height
  • stem diameter
  • leaf number
  • fresh mass
  • dry mass
  • leaf area

24. Full Water Replacement Produced High Biomass in the Irrigation Study

One of the strongest biomass treatments combined:

100% pot water capacity

with one of the tested irrigation-water salinity conditions.

It produced approximately:

206.7 g fresh matter

25.5 g dry matter

and:

3044 cm² leaf area.

Meanwhile, maximum water-use efficiency occurred under another treatment around:

85% pot capacity.

Again, we see two different production targets:

maximum biomass

versus:

maximum water-use efficiency.

25. This Does Not Support “Dry Air Makes Collard Midribs Fibrous”

The irrigation study directly demonstrates that:

water availability affects collard growth.

But it did not establish:

high VPD → fibrous midribs

or:

early dryness permanently determines leaf texture.

Those are much stronger causal claims.

Without direct texture, lignin or fiber measurements, they should not be published as scientific conclusions.

26. Salinity Is Also Part of Root-Zone Stress

The same research showed that irrigation-water salinity interacted with irrigation volume.

Higher salinity affected several collard morphological characteristics.

This means poor crop performance under an apparently adequate PAR environment could originate from:

  • insufficient water
  • excessive EC
  • salinity
  • root-zone stress

rather than light itself.

27. What Does VPD Tell Us?

VPD describes atmospheric evaporative demand.

It depends mainly on:

  • temperature
  • relative humidity

When temperature rises or humidity falls, VPD generally increases.

That can increase:

  • transpiration demand
  • root-zone water demand
  • substrate drying

VPD therefore provides useful environmental context.

28. But There Is No Validated Collard VPD Target Table

The old article gave precise targets such as:

0.5–0.8 kPa

0.7–1.1 kPa

0.9–1.3 kPa

for different stages.

Current collard-specific evidence does not validate these as universal optimum ranges.

For collards, the research base is much stronger for:

  • irrigation
  • salinity
  • CO₂
  • light

than for experimentally identifying an ideal VPD.

Therefore, the most scientifically defensible role for VPD is:

environmental monitoring.

29. High VPD Does Not Automatically Mean Tough Leaves

A high VPD tells you atmospheric water demand is relatively strong.

It does not directly measure:

  • leaf fiber
  • lignin
  • midrib toughness
  • sweetness

Those characteristics depend on multiple variables.

If high VPD coincides with tough leaves, you still need to determine whether other factors changed, such as:

  • temperature
  • leaf age
  • root-zone moisture
  • cultivar
  • harvest maturity

Correlation should not be presented as causation.

30. Very Low VPD Is Not Automatically Better

Very humid air lowers VPD.

But persistently humid greenhouse conditions may also contribute to:

  • condensation
  • slow canopy drying
  • disease-favorable conditions

Therefore, the objective should not be:

minimum VPD.

The better goal is:

stable water relations without prolonged atmospheric extremes.

31. Temperature Must Be Considered

Collards are cool-season Brassicas.

Temperature changes:

  • growth rate
  • respiration
  • transpiration
  • crop timing
  • leaf development

and directly changes VPD.

Therefore:

PPFD + VPD without temperature context is incomplete.

The same 500 µmol/m²/s PPFD can create very different plant conditions at:

18°C

and:

30°C.

32. Seedling and Early Establishment

During early growth, focus on:

  • uniform germination
  • light distribution
  • compact morphology
  • root-zone moisture
  • temperature

Do not automatically apply mature-plant lighting.

Young collards intercept much less total light than a mature canopy.

Baby-leaf studies demonstrate strong responses to DLI, but those values should remain specific to that production stage.

33. Active Leaf Expansion

As collards develop:

  • leaf area increases
  • canopy photosynthetic demand increases
  • self-shading develops
  • water use increases

At this stage, measure multiple points across the crop.

Use:

PPFD

for instantaneous mapping

and:

DLI

for total daily exposure.

Monitor CO₂ during the same periods.

34. Mature Leaf Production

For mature collards, the production target may include:

  • leaf area
  • fresh biomass
  • marketable leaf count
  • leaf tenderness
  • color
  • nutritional quality
  • regrowth after harvest

Current research does not justify one mature-stage combination such as:

350–550 PPFD + 800–1200 ppm CO₂ + 0.9–1.3 kPa VPD.

Each of those parameters needs context.

35. Repeated Harvest Changes the Problem

Collards are often harvested by removing outer leaves while allowing the plant to continue growing.

That means:

  • plant age continues increasing
  • canopy structure changes
  • sink demand changes
  • root system becomes larger
  • new leaves develop under a different canopy environment

A long-cycle plant should therefore be evaluated through:

environmental trends over time

rather than one fixed stage chart.

36. Leaf Texture Should Be Measured Directly

If tenderness is commercially important, record it.

Possible practical comparisons include:

  • leaf age
  • harvest position
  • petiole/midrib thickness
  • sensory tenderness
  • cooked texture

Then compare those observations with:

  • DLI
  • temperature
  • water status
  • harvest age

This is much more defensible than assuming one VPD number determines fiber.

37. A Practical Greenhouse Measurement Workflow

Step 1 — Measure PAR at Canopy Height

Measure where collard leaves actually receive light.

Step 2 — Map Several Locations

Check:

  • greenhouse center
  • edges
  • structural shadows
  • upper canopy
  • inner shaded leaves

Step 3 — Record DLI

Compare:

  • sunny vs. cloudy days
  • seasons
  • greenhouse zones
  • supplemental-light schedules

Step 4 — Monitor CO₂

Compare CO₂ with PAR during active photosynthesis.

Step 5 — Track Temperature and Humidity

Use these to understand VPD.

Step 6 — Track Root-Zone Water

Record:

  • irrigation
  • substrate moisture
  • nutrient-solution EC where relevant

Step 7 — Record the Crop Response

Track:

  • fresh biomass
  • leaf area
  • marketable leaf count
  • harvest interval
  • regrowth

If tenderness matters, measure it directly.

38. Practical Research-Based Reference Points

Baby-Leaf DLI

Direct collard research tested:

6–30 mol/m²/day.

Yield increased strongly through:

24 mol/m²/day

with smaller additional gain from:

24 → 30.

This is useful for baby-leaf systems.

It is not a universal mature-collard requirement.

Microgreen DLI

‘Vates’ collard microgreens were successfully studied at:

14 and 21 mol/m²/day.

Higher DLI increased fresh weight and dry-matter percentage but reduced electrical-use efficiency.

Again, this applies to microgreens.

CO₂

Direct collard research at:

350 vs 700 ppm

shows substantial effects on biomass and photosynthetic physiology.

Approximately 700 ppm is therefore a useful experimental reference.

It is not a validated commercial optimum.

VPD

There is currently insufficient collard-specific evidence to publish a universal growth-stage VPD target.

Use VPD to understand atmospheric water demand.

Water

Direct greenhouse research confirms that irrigation amount and water quality can strongly affect:

  • biomass
  • leaf area
  • morphology

Interpret VPD together with actual root-zone water availability.

39. A Better Way to Think About Collard Measurements

Instead of asking:

What PPFD makes collards tender?

ask:

How much light is the crop receiving, and how does it affect biomass and crop timing?

Instead of:

What VPD prevents fiber?

ask:

How strong is atmospheric water demand, and is the root zone supplying sufficient water?

Instead of:

What CO₂ number makes larger leaves?

ask:

How does CO₂ change photosynthesis and biomass under the actual light environment?

Then measure tenderness separately if tenderness is the commercial goal.

Final Takeaway

Greenhouse collard greens do not have one scientifically established PAR, CO₂ and VPD recipe for every stage of growth.

But collards have several strong crop-specific research findings.

DLI strongly affects young collard productivity.

A direct baby-leaf experiment tested:

6, 12, 18, 24 and 30 mol/m²/day.

Estimated fresh yield increased from approximately:

0.22 lb/ft² at 6 DLI

to:

0.71 lb/ft² at 24 DLI

and:

0.80 lb/ft² at 30 DLI.

The smaller gain above 24 DLI suggests diminishing biological and potentially economic returns.

Higher DLI also shortened the time required to reach approximately 4-inch leaf height from:

14.8 days at 6 DLI

to:

12.2 days at 24 DLI.

CO₂ also matters.

Direct collard research comparing approximately:

350 and 700 ppm

showed strong biomass responses under high light.

But long-term exposure to elevated CO₂ also caused photosynthetic down-regulation associated with carbohydrate accumulation.

That demonstrates why:

more CO₂ is not simply a linear photosynthesis switch.

Spectrum matters.

Recent collard research found red + blue lighting increased leaf dimensions, while blue light substantially altered glucosinolate profiles.

Water matters.

Direct greenhouse irrigation research shows that root-zone water availability and water quality can strongly affect collard biomass and leaf area.

But current research does not justify claims that:

high VPD directly causes fibrous leaves

or:

early dryness permanently determines collard texture.

For VPD, the stronger approach is to use it as a measure of atmospheric water demand and interpret it together with:

  • temperature
  • irrigation
  • root-zone moisture
  • crop response

The better greenhouse strategy is therefore:

Measure PAR at the actual canopy.

Use DLI to understand total daily light.

Monitor CO₂ during active photosynthesis.

Track temperature, humidity and VPD together.

Measure root-zone water conditions.

Then compare those measurements with the production trait that actually matters:

yield, crop timing, leaf area, nutritional quality or tenderness.

That provides a much stronger basis for greenhouse collard production than unsupported stage-by-stage environmental targets.

References

Gagne, C. & Mattson, N. Increasing Yields of Baby Leaf Vegetables. Greenhouse Product News, 2019.

Gagne, C. The Effects of Daily Light Integral on the Growth and Development of Baby Leaf Vegetables. Cornell University.

Continuous Lighting Can Improve Yield and Reduce Energy Costs While Increasing or Maintaining Nutritional Contents of Microgreens. Frontiers in Plant Science, 2022.

Bunce, J.A. & Sicher, R.C. Daily Irradiance and Feedback Inhibition of Photosynthesis at Elevated Carbon Dioxide Concentration in Brassica oleracea. Photosynthetica, 2004.

The Nutritional Value, Biochemical Traits, and Growth of Brassica oleracea Grown Under Red, Blue, and Combined Red–Blue LED Lighting. Plants, 2025.

Viana et al. Saline Irrigation Water Indices Affect Morphophysiological Characteristics of Collard. Horticultura Brasileira, 2021.

Related AquaHorti Tools

For instantaneous PAR / PPFD measurements and checking greenhouse light distribution, see AquaHorti AH-Quantuv.

For recording changing greenhouse PAR throughout the day and measuring DLI, see AquaHorti AH-PARDLI.

For greenhouse monitoring where PAR, DLI, CO₂, temperature, humidity and VPD need to be reviewed together over time, see AquaHorti AH-200.