Take a clod out of a field that has been worked well and one out of a field that has not, and hold them side by side. The first breaks into crumb, threaded with fine white roots and, on the underside, the faint silver lacework of fungal hyphae. The second comes up as a dense block the colour of old cement, with no visible biology and no smell at all.
The difference between them is mostly invisible, and it is measurable. Microbial biomass carbon in productive temperate cropland typically runs 500 to 900 mg of carbon per kilogram of dry soil. In cultivated arid land it is often below 200 (Bardgett and van der Putten 2014, reviewed in our literature review). Soil organic carbon shows the same gap: 2 to 4 percent by mass in temperate systems against 0.3 to 1.2 percent in cultivated drylands (Plaza et al. 2018; Lal 2004). The plant on top registers that difference long before the agronomist does, and the yield ledger registers it eventually.
This post is about the workforce those numbers describe: what it does, what a synthetic-only programme does to it, and what is and is not known about putting biology back.
What lives in a teaspoon of healthy soil
The cast of characters in a healthy soil community is much larger than most farmers were taught. The list, in rough order of size:
- Bacteria and archaea. The smallest and the most numerous, billions per gram. They drive the early stages of decomposition, fix atmospheric nitrogen in the case of certain rhizobia and free-living strains, and convert organic forms of nutrients into mineral forms the plant can absorb. Without them, organic matter just sits.
- Fungi, especially mycorrhizae. Fine threads, sometimes kilometers per gram of soil. Mycorrhizal fungi form a partnership with most crop roots. They extend the root surface area dramatically and trade phosphorus, zinc, and water for sugars from the plant. A plant with healthy mycorrhizal colonization is functionally a plant with a much bigger root system than the one you can see.
- Protozoa and nematodes. Single-celled animals and microscopic worms that eat bacteria and fungi. Their digestion releases the nutrients those microbes had immobilized, the second link in the soil nutrient cycle.
- Microarthropods. Springtails, mites, tiny insect-like creatures. They shred residue and aggregate the soil into the crumb structure that holds water and air.
- Earthworms. The ones you can see. Each worm passes its own body weight in soil through its gut every day, breaking apart compaction and laying down a coating of mucus and casts that is one of the most fertile substances in agriculture.
The relationship between these groups is what most fertilizer programmes ignore. Bacteria mineralize. Protozoa eat bacteria and release the minerals. Mycorrhizae move them to the root. Earthworms and arthropods build the structure that lets it all happen. It is a workforce. The plant pays it in sugars, sent down through the roots as exudates, and the workforce delivers a much wider panel of nutrients than any bag does.
When the workforce is intact, the plant gets nitrogen, phosphorus, potassium, plus calcium, magnesium, sulphur, plus the eight micronutrients you do not see on a label, plus access to water held in soil pores that compacted soil simply does not have. When the workforce is gone, the plant only gets what you put in the bag.
What synthetic mineral fertilizer does down there
Soluble mineral fertilizer feeds the plant directly, in a form it can take up immediately. That is the design of the product. It is also, over time, the problem.
Three things happen to the soil community when synthetic fertilizer is the dominant input, and research published over the last thirty years across temperate and arid soils has documented all three:
The first is salt shock. Soluble synthetic fertilizers are technically salts, and at the concentrations they reach near the dissolution point, they strip out the bacteria and fungi in the immediate vicinity. The effect is local and recoverable in a single application, but the recovery happens slower in soil that is already low on organic matter, and the cumulative effect across a season is a quiet thinning of the bacterial population.
The second is suppression of mycorrhizal colonization. When phosphorus and nitrogen are spoon-fed to the root in mineral form, the plant stops sending out the chemical signals that recruit mycorrhizal fungi. The partnership is, from the plant’s point of view, no longer worth paying for. Within a few seasons, the mycorrhizal network thins and breaks down. The plant loses access to the broader nutrient panel and the water-foraging capacity those fungi provided. It does not show up as a problem until conditions get hard: a drought week, a heat spike, a season when irrigation is constrained.
The third is the slow loss of soil organic carbon. Without the input of root exudates that feed the broader soil community, organic matter decomposes faster than it accumulates. Carbon drops. Soil structure collapses. Water infiltrates less, runs off more, and pools where it should not. After a decade, you are working a field that physically holds less of what you put on it.
Each of these is small in any single season. The compounding effect is what catches farmers by surprise.
Why dead soil yields less even when you fertilize more
Here is the mechanism that explains the most frustrating pattern in commercial agriculture: the field where you keep adding more fertilizer, and the yield keeps drifting sideways or down.
A plant in living soil eats from two pots. The first pot is what you applied: the bag, the drip, the foliar spray. The second pot is what the biology releases from the mineral fraction of the soil itself, plus what mycorrhizae deliver from outside the root zone, plus what bacteria mineralize from residue and decomposing roots. The second pot, in healthy soil, is often the larger one over the course of a season.
A plant in dead soil eats from one pot. There is nothing else for it to draw on. So the relationship between input and output becomes linear and brittle. Twenty more kilograms of nitrogen produces twenty kilograms more yield, until it does not, because the plant has run into a different limitation, whether water, or zinc, or sulphur, or root pathogen pressure that healthy soil would have suppressed and dead soil cannot. Then the agronomist arrives with the lab results and says the soil is "tired," and the conversation moves to lime, or gypsum, or a foliar micronutrient blend, all of which is treating a symptom of the same underlying loss.
There is a second cost on top of that. A lot of the synthetic fertilizer you apply to dead soil is not actually being absorbed efficiently. Nitrogen leaches. Phosphorus binds to soil colloids and goes nowhere. You are paying for nutrients that wash through the profile or sit in unavailable forms. In live soil, biological cycling captures more of what you apply and delivers it to the root over time. The ledger reads the same on the way in. It reads very differently on the way out.
You can keep buying nutrients. Or you can rebuild the workforce that delivers them.
What a living input does, and what it does not
A live-microbial input is a different category of product from a soluble mineral one, and the difference is worth stating precisely, including the part that undercuts the sales pitch.
Magic Power is the liquid biological residue of a closed-loop recirculating aquaculture system. What a laboratory found in a sample of it is published in full on the evidence page. The short version: a large and diverse living bacterial community, 952 taxa at a Shannon index of 4.49, with a plate count above the culture method’s reporting ceiling, and a small quantity of plant nutrition, 421.5 mg of nitrogen per litre, 216.8 of potassium, 16.5 of phosphorus (NviroTek Wynland Laboratories report S26/3362, page 1 of 2).
A litre therefore carries about 0.42 g of nitrogen against 460 g in a kilogram of urea. The independent interpretation of that panel is blunt about what follows: the product is "unbalanced and low in plant nutrients, and will therefore not replace a complete fertility program". Read it as a biological input applied inside a fertility programme, not as a substitute for one.
The mechanism the programme is betting on is nutrient-use efficiency: microbes colonizing the root zone and cycling nutrients from both the applied input and the soil’s existing mineral fraction, with soil structure improving as aggregates rebuild. In the wider literature that mechanism is well documented (Rillig 2004; Six et al. 2004; Schütz et al. 2018). On this product it has not been measured in the field, because no yield trial has completed. The protocol that would measure it is published in full on the trials page, and the results go up either way.
Where to start if your soil is already tired
If the field you are walking is closer to the second clod than the first, the recovery is gradual and the order of operations matters. Three steps that have worked for the farms we have onboarded:
The first is a baseline soil test that goes beyond N-P-K. Ask for soil organic carbon percentage, microbial biomass, and infiltration rate. These are the numbers that actually track the biology. Most agri-shop soil tests do not measure them. A regional university lab or an independent soil-health service will. Establish where you are before you change anything.
The second is to reduce, not eliminate, your synthetic programme in the first season. Cutting NPK by half and supplementing with a live-microbial input lets the soil community begin to rebuild without putting the current crop at risk. Eliminating synthetic in a single season on dead soil tends to produce a yield drop that masks the recovery underneath. Reduce gradually, measure each quarter, and let the biology earn its place in the programme.
The third is to measure the same indicators every season. Organic carbon climbing by 0.1 to 0.3 percent per year, infiltration rate improving from minutes-per-inch to seconds-per-inch, microbial biomass rebuilding: these are the leading indicators of a soil that is coming back. They lead the yield response by one to two seasons. Farmers who measure them stop second-guessing the programme in month four. Farmers who only watch the yield ledger frequently quit just before the curve turns.
The measurement schedule in detail, with the panel to run at three, six and twelve months and the rules for reading a result, is in the trial protocol. The cost side is at why imported fertilizer is eating your margin, and the calculator that refuses to compute without your own inputs is at the math.
If you would rather start with a soil sample and a conversation, apply. What we will not do is tell you what your field will do, because nobody has measured that yet.