This blog adds further detail to a well-received Soil Health, Plant Health, Human Health: How Living Soils Create Abundance. When I gave presentations, several listeners had asked me for dietary advice: How can we maximize our intake of PSMs? As gardeners, how can we maximize our growing methods such that our plants express most PSMs?  In this blog, I am exploring an answer to these questions.

Plant secondary metabolites (PSMs) are the chemical language of living landscapes. They are not the sugars, starches, or proteins that simply build biomass, or which would be mentioned on an ingredient list of your food in a grocery store. PSMs are the polyphenols, glucosinolates, terpenes, carotenoids, alkaloids, and sulfur compounds that plants produce to communicate, defend, adapt, and regulate. In healthy soils, microbes supply precursors, cofactors, and signals that enable this chemistry. In plants, PSMs shape immunity, stress resilience, and ecological interactions. In humans, many of these same compounds modulate inflammation, redox balance, detoxification pathways, and cellular signaling. The soil–plant–human connection is therefore not merely about calories, minerals or macro nutrients —  it is about whether living systems are producing and transferring biologically active information, balancing our metabolism, and enable our human immune systems to thrive.

If you want more nutrient-dense food, here is some straightforward advice: In your food choices,

• favor diversity across plant families.
• eat seasonally.
• prioritize freshness and short supply chains.
• prefer food grown in biologically active soils.
• accept moderate ecological stress (not sterile perfection, not collapse).

This advice is enough for most eaters to make good dietary choices. It is, in fact, supported by a growing body of plant physiology and ecology research.

But the advice is also incomplete, and research has revealed much more detail that can help farmers and eaters alike. Nutrient density—especially when we are talking about plant secondary metabolites (PSMs)—is not a fixed property of a crop. It is the outcome of a multi-layered expression process that unfolds in time.

PSMs are compounds such as polyphenols, glucosinolates, terpenes, carotenoids, alkaloids, and many others. They are not required for plant survival in the same way that sugars or proteins are. Instead, they mediate defense, signaling, stress adaptation, and ecological interactions. Many are also biologically active in humans.

What follows is a technical but accessible exploration of why predicting their levels in food is far more difficult than the simplified soil → plant → human narrative suggests.

I. What the Simplified Soil Narrative Gets Right

The contemporary soil-health movement, including the work of Dan Kittredge and the Bionutrient Institute, mainly emphasizes that soil biology matters. Their reports document wide variability in crop nutrient proxies across farms and regions and repeatedly highlight seasonality and management as important drivers.¹

The foundational claim—that microbial life in soil influences plant nutrient availability and metabolic potential—is well supported. Research on rhizosphere signaling shows that plant roots actively communicate with soil microbes, recruiting and responding to bacterial and fungal consortia.² These microbes, in turn, influence nutrient availability for plants, immune priming, and metabolic pathways.³ Plants also respond to bacterial quorum sensing molecules — chemical signals that microbes use to coordinate their behavior, and also alter the plants’ regulation of their physiology and defense apparatus.⁴

Soil biological activity reflects capacity: access to chelated micronutrients, sulfur cycling, redox buffering, and microbial cofactors. Without that capacity, sophisticated secondary metabolism is unlikely. But capacity is not expression. Plants do not produce all secondary metabolites at all times. This blog will discuss the complexity of how capacity turns into expression.

II. Expression Requires More Than Capacity

1. Developmental and Phenological Gating

First, plant developmental stage matters.

In Brassica studies, glucosinolate concentrations vary several-fold across developmental stages and tissues. Studies on broccoli and related species show that sprouts, vegetative leaves, and mature florets differ markedly in both total levels and composition of glucosinolates.⁵ This means that “same crop” harvested on the same field, but at slightly different developmental stages, may significantly differ chemically.  Researchers call this phenological gating: certain metabolic pathways are only fully active during specific growth windows.

2. Circadian and Diurnal Regulation

Plants have internal clocks. These circadian systems regulate gene expression, redox state, and metabolic flux on a roughly 24-hour cycle.

A comprehensive review by Liebelt, Jordan, and Doherty demonstrates that phytochemical levels can oscillate daily and seasonally, at times enough to confound experimental comparisons.⁶ In some systems, defense-related metabolites peak at particular times of day.

For example in Arabidopsis, clock components such as CCA1 influence resistance to aphids by modulating indole glucosinolate production.⁷ This means that the same plant sampled in the morning and afternoon may not have the same defensive chemistry.

This variability throughout the day is not statistical noise. It is biology running on a schedule.

Grazing animals can sense these diurnal cycles. For example, sheep and other ruminants preferentially graze certain forages in the late afternoon, when photosynthesis has elevated non‑structural carbohydrate (NSC) concentrations and the ratio of energy to some defensive compounds (e.g., tannins) is more favorable. Controlled feeding studies show that ruminants adjust intake across the day in response to shifting nutrient-to-toxin balances, integrating postingestive feedback with learned preferences (Provenza et al. 2003).¹⁵ Indeed, animals display measurable behavior that track within‑day changes in plant chemistry. If herbivores detect and respond to diurnal variation in metabolites, then sampling crops at arbitrary times risks missing biologically meaningful chemical states.—

3. Stress as a Conditional Trigger

Secondary metabolites are often inducible. Moderate herbivory by insects or mammals, UV exposure, cold stress, or pathogen pressure all may trigger additional layers of defensive chemistry.

In broccoli, pre-harvest treatment with methyl jasmonate—a signaling molecule associated with herbivory—significantly increases indole glucosinolate levels when applied several days before harvest.⁸⁻⁹ Again, timing matters: the induction window only lasts days before the plant return to normal levels!

I am not implying that “stress is good.” Severe or chronic stress can suppress both growth and chemistry. But moderate, ecologically normal stress can amplify certain PSM classes—if the plant has the metabolic capacity to respond. However, induction responses frequently involve allocation trade-offs, and chronic or poorly timed stress can reduce both yield and overall plant performance. Farming for stress induction is therefore not a simple optimization strategy.

Operationalizing this in farming in a targeted manner is difficult. Real fields contain heterogeneous stress intensities, multiple attackers, and complex interactions. Dose–response curves for edible tissues under realistic pest scenarios remain limited in the literature.

III. From Allocation to Ecology: Downstream Filters on Expression

A. Allocation Trade-Offs: Growth Versus Defense

Secondary metabolism is energetically expensive. Carbon, nitrogen, and sulfur must be diverted from growth into defense.

Kroymann’s synthesis on natural diversity in plant secondary metabolism emphasizes that trade-offs are central to chemical diversity.¹⁰ In high-resource contexts, plants may allocate more to growth; under moderate stress, more to defense; under severe stress, metabolism may collapse.

This explains why “moderate stress” cannot be reduced to a slogan. The threshold between beneficial induction and metabolic impairment is species- and context-specific.

B. Harvest Timing: Capturing or Missing the Peak

Even if soil health, phenology, and stress align to induce high PSM expression, harvest timing determines whether that expression is captured.

The broccoli methyl jasmonate studies applied induction treatments approximately four days prior to harvest to maximize glucosinolate levels.⁹ If harvest occurs too early or too late relative to induction, measured concentrations differ.

This suggests a critical but under-researched question: what are the optimal harvest windows for specific PSM targets in commercial systems?

At present, these windows are poorly defined for most crops.

C. Post-Harvest Biology: The Hidden Filter

Modern storage systems are designed to preserve appearance, firmness, and logistics—not ecological signals. Contemporary supply chains select crops for visual stability, transportability, and shelf-life consistency. They are not designed to preserve dynamic metabolic expression. As a result, post-harvest systems favor predictability over biochemical diversity.

Controlled atmosphere (CA) storage and 1-methylcyclopropene (1-MCP), which inhibits ethylene signaling, significantly alter apple biochemical composition and volatile profiles after long storage.¹¹⁻¹² Ethylene is not only a ripening hormone; it regulates phenolic and volatile metabolism.

When fruits are stored for months under low oxygen and ethylene inhibition, they remain visually fresh—but their biochemical trajectories change.

For vegetables, cutting and bruising activate endogenous enzymes (e.g., myrosinase in Brassicas), converting stored glucosinolates into other compounds. Oxygen and light promote oxidative degradation of phenolics and carotenoids. Quantitative loss curves vary by compound and condition, but the directionality is well established in post-harvest physiology.

Storage does not necessarily eliminate all PSMs. It flattens profiles, suppresses induction, and reduces diversity.

D. Biodiversity and Community Context

Plant chemistry is shaped not only by soil and stress, but by community context.

In experimental plant communities, higher species richness has been shown to alter allocation to chemical defense.¹³ Long-term biodiversity experiments such as the Jena Experiment demonstrate that plant diversity influences trophic interactions and microbial community activity over time.¹⁴

While most biodiversity experiments focus on grasslands rather than vegetable systems, they provide proof of principle: plant chemistry is not solely an individual trait; it is shaped by community-level ecology.

Similarly, quorum sensing molecules in the rhizosphere—chemical signals used by bacteria to coordinate behavior—can modulate plant immune and metabolic pathways.⁴ These findings support the idea that soils function as information networks, not just nutrient reservoirs.

IV. Why Prediction Is Hard

Genotype alone can explain substantial variation in PSM profiles. Decades of breeding for yield, uniformity, and reduced bitterness have often selected against certain defensive and aromatic metabolites. Any soil-based framework must therefore be interpreted alongside cultivar history.

Variability is not a failure of soil-health theory. It is a reflection of living systems operating across multiple interacting scales.  Once we combine genotype differences, phenological gating, circadian oscillations, stress type and intensity, soil microbial context, harvest timing, storage conditions, we get a high-noise, multi-factor system. The Bionutrient Institute’s own reports highlight substantial variability across farms and seasons, and identify time-of-year as a confounding variable in nutrient-density comparisons.¹

This level of complexity is consistent with what we know about living systems: metabolism is regulated, conditional, and context-dependent.

V. Implications

For researchers, the central implication is methodological rigor across time and context. Future nutrient-density studies must move beyond static comparisons of “soil type” or “management system” and instead incorporate time-of-day controls, phenological staging, stress history metadata, and explicit documentation of storage conditions. Targeted metabolomics should complement, and in many cases replace, reliance on broad antioxidant proxies. Without these design elements, we risk attributing differences to soil biology that in fact arise from timing, induction windows, or post-harvest handling.

For farmers, the message is not to chase “moderate plant stress”, but to build capacity first. Soil biological function remains foundational. Chronic stress should be avoided, as it suppresses both yield and chemistry. At the same time, moderate ecological exposure—real weather, real microbial communities, some degree of pest interaction—should not automatically be interpreted as failure. Harvest timing deserves more strategic attention, especially when targeting inducible compounds such as indole glucosinolates in Brassicas. Where possible, shortening storage windows and communicating harvest dates to buyers can help preserve biochemical expression.

For eaters, the guidance remains practical but informed by complexity. Freshness often matters more than branding claims. Seasonal produce is more likely to be harvested near physiological maturity and closer to its expression peak. Flavor intensity can serve as a rough heuristic for chemical richness, since many PSMs contribute to bitterness, aroma, and depth. Dietary diversity across plant families increases exposure to a broader spectrum of secondary metabolites, buffering against the inherent variability of any single crop at any single moment.

It is neither necessary nor biologically realistic to consume all potentially beneficial PSMs at all times. Human physiology evolved under seasonal fluctuation, with periods of abundance and relative scarcity. A diverse diet sourced from biologically active soils will, over the course of a year, supply a rotating spectrum of phytochemicals that our detoxification systems, redox pathways, and immune networks are well adapted to handle. The real risk is not short-term variation, but chronic shortage—an overly homogeneous diet grown in capacity-limited soils, stored for long periods, and stripped of ecological diversity. Avoiding systematic depletion matters more than achieving constant biochemical perfection.

Conclusion: Nutrient Density Is a Dynamic Ecological Process

Nutrient density is not a static attribute that can be stamped onto a crop label. It is the emergent outcome of interacting biological processes unfolding across soil, plant, and time. Soil health establishes metabolic capacity; microbial consortia mobilize precursors and signals; plants gate expression through development and circadian regulation; ecological stress modulates defensive chemistry; harvest either intercepts or misses transient peaks; storage reshapes biochemical trajectories. At every step, chemistry is contingent.

This does not weaken the soil–plant–human framework. It deepens it. If anything, the variability documented by the Bionutrient Institute and others is precisely what one would expect in a living, responsive system. A biologically active field is not a factory line; it is a network of feedback loops operating across scales—from microbial quorum sensing in the rhizosphere to herbivore selection behavior above ground.

For those seeking simple guarantees, this complexity can feel unsatisfying. But ecological systems do not promise uniformity; they promise adaptability. The goal is not to engineer chemically identical produce year-round, nor to maximize single compounds in isolation. The goal is to sustain landscapes with sufficient biological capacity that plants can express their full metabolic repertoire when conditions call for it.

In that sense, nutrient density is less a number and more a signature of ecological function. When soils retain microbial diversity, when crops are grown in real seasons, when harvest and storage respect biological timing, food carries the imprint of those processes. The alternative—capacity-limited soils, chronic stress, long storage, and homogenized diets—produces predictability at the cost of biochemical richness.

Living systems are variable because they are alive. If we want food that carries the depth and resilience of living systems, we must accept and work with that variability rather than simplify it away.

References (Chicago Style)

1. Bionutrient Institute. Final Report – Bionutrient Institute Documentation. 2018; Bionutrient Institute. Food Desert Report. 2020; Bionutrient Institute. Grains Report. 2020.
2. Jamil, F., et al. “Rhizosphere Signaling: Insights into Plant–Rhizomicrobiome Interactions.” Microorganisms 10 (2022).
3. Yang, L., et al. “Mechanisms of Rhizosphere Plant–Microbe Interactions.” 2024.
4. Hartmann, A., and A. Schikora. “Plant Responses to Bacterial Quorum Sensing Molecules.” Frontiers in Plant Science 6 (2015).
5. Omondi, J.O., et al. “Glucosinolate Accumulation across Broccoli Development.” Plants 11 (2022).
6. Liebelt, D.J., J.T. Jordan, and C.J. Doherty. “Only a Matter of Time: The Impact of Daily and Seasonal Rhythms on Phytochemicals.” Phytochemistry Reviews 18 (2019): 1409–1433.
7. Lei, J., et al. “CIRCADIAN CLOCK-ASSOCIATED1 Controls Resistance to Aphids by Altering Indole Glucosinolate Production.” Plant Physiology 179 (2019): 1344–1354.
8. Chiu, Yu-Chun, Kristen Matak, and Kang-Mo Ku. “Methyl jasmonate treated broccoli: Impact on the production of glucosinolates and consumer preferences.” Food Chemistry 299 (2019): 125099.
9. Liu, Ann G., John A. Juvik, Elizabeth H. Jeffery, Lisa D. Berman-Booty, Steven K. Clinton, and John W. Erdman Jr. “Enhancement of broccoli indole glucosinolates by methyl jasmonate treatment and effects on prostate carcinogenesis.” Journal of medicinal food 17, no. 11 (2014): 1177-1182.
10. Kroymann, J. “Natural Diversity and Adaptation in Plant Secondary Metabolism.” Current Opinion in Plant Biology 14 (2011): 246–251.
11. Ferreira C, Ribeiro C, Nunes FM. Effect of Storage Conditions on the Volatilome, Biochemical Composition and Quality of Golden Delicious and Red Delicious Apple (Malus domestica) Varieties. Molecules. 2024 Jun 21;29(13):2954.
12. Muche, Bizuayehu M., Michael Jordan, Charles F. Forney, R. Alex Speers, and HP Vasantha Rupasinghe. “Effect of 1-methylcyclopropene (1-MCP) and storage atmosphere on the volatile aroma composition of cloudy and clear apple juices.” Beverages 6, no. 4 (2020): 59.
13. Mraja, A., et al. “Plant Community Diversity Influences Allocation to Direct Chemical Defense.” PLOS ONE 6 (2011).
14. Jochum, Malte, Markus Fischer, Forest Isbell, Christiane Roscher, Fons van der Plas, Steffen Boch, Gerhard Boenisch et al. “The results of biodiversity–ecosystem functioning experiments are realistic.” Nature ecology & evolution 4, no. 11 (2020): 1485-1494.
15. Provenza, Fred. Nourishment: what animals can teach us about rediscovering our nutritional wisdom. Chelsea Green Publishing, 2018; Provenza, Fred D., Juan J. Villalba, L. E. Dziba, Shelton B. Atwood, and Roger E. Banner. “Linking herbivore experience, varied diets, and plant biochemical diversity.” Small ruminant research 49, no. 3 (2003): 257-274.