16. Addressing the scientific critique honestly
A credible case acknowledges its strongest critic. Washington State University Extension has argued that hügelkultur lacks a firm scientific basis, that beds subside and must be rebuilt, and that simple soil mounds with mulch achieve similar benefits with less labor. Two responses are warranted. First, the criticism is fair about the thinness of peer-reviewed studies on hügelkultur specifically, which is exactly why this case rests primarily on the large, established literatures it draws upon: coarse-woody-debris ecology, buried-wood carbon storage, riparian buffers, and soil-organic-matter science. Second, the subsidence concern is answered above: it is front-loaded, self-limiting, and managed by ordinary top-dressing, and the buried wood delivers water-storage and carbon benefits that surface mulch cannot match. The honest conclusion is not that hügelkultur is proven beyond question, but that its mechanism is well understood, its risks are manageable, and the weight of related evidence supports it.
A fair related question is why bury wood at all when chipped or shredded woody biomass is already commonly returned to soil as surface mulch. The soil-carbon literature has begun to answer this directly, and on three measurable points the comparison favors burial. First, less carbon survives surface decomposition than buried decomposition: a study in Soil Biology and Biochemistry (2017) reported that surface wood treatments produced roughly twice the wood-derived soil CO₂ efflux of buried treatments. Second, the carbon that does enter the soil after burial is more stable: a 25–48-year study of deep-ploughed soils published in Scientific Reports (Nature, 2017) found that deep-ploughed subsoils contained significantly more soil organic carbon than reference subsoils, about 48% more in forest soil and 67% more in cropland, and that buried SOC was on average 32% more stable, with older apparent radiocarbon ages indicating it had been largely isolated from atmospheric exchange. Third, surface residues, including mulches, "make only a small contribution to longer-term soil C stocks" compared with belowground inputs (Washington State Department of Ecology SOC guide, citing Kirkby et al., 2006); composting itself, while a clear improvement over landfilling biomass, still loses most of the original carbon as CO₂, with roughly 15–35% of the carbon used to make compost ending up as stable humus. The microbial-necromass mechanism described in Section 6 explains why this ordering holds. Mulching and composting remain valuable and should continue, but if the goal is to place more carbon underground for longer and to provide a water-storing reservoir beneath the root zone, hügelkultur does measurably more.
A note on methane, with the farm’s own measurements
Composting can emit methane when piles go anaerobic, waterlogged or poorly turned, because methanogens thrive without oxygen; well-aerated surface mulch, by contrast, stays oxygenated and typically does not produce methane. A buried hügelkultur bed contains zones, the deepest, wettest parts, where conditions are likewise less aerobic, so methane is not exclusively a composting problem, and the case does not run only one way.
This farm has therefore measured it directly. A combustible-gas detector (TopTes PT-990, 0–10,000 ppm, eleven-inch probe) was used at the vent of a newly constructed bed on four occasions spanning a full year and roughly a hundred degrees Fahrenheit of ambient range. Every reading was zero. The instrument was verified responsive on the same occasions: against a propane source (0–130 ppm), against a wood-chip pile (110 ppm), and against the headspace of a closed compost-tea drum (245 ppm). The zero is therefore a verified null rather than an unresponsive sensor.
The working interpretation is that the bed, being newly constructed, is loose and vents freely, so oxygen reaches the decomposing wood throughout and the aerobic fungal pathway dominates, leaving no anaerobic pockets to generate methane. This is consistent with the visible fungal succession on the bed surface, turkey tail and wine cap, both aerobic decomposers. It is also, in the terms of Section 5, a direct statement of which register the bed is operating in: this is a fertility bed, not an anaerobic storage vault.
A prediction follows from that interpretation, and it is worth stating because it is falsifiable: as a bed settles and compacts over years, localized low-oxygen zones may form, and methane may appear where it previously did not. The farm has therefore begun measuring a seven-year-old, fully settled bed for comparison. If the older bed shows methane while the new one does not, the aeration-by-looseness explanation is supported and the practice implication is that older beds warrant continued monitoring. If the older bed also reads zero, the aerobic character is more durable than settling alone would predict. Either outcome is informative, and the result will be reported as found.
Best practice for hügelkultur remains: keep the upper layers aerated, ensure drainage, build with graded material, and use slower-rotting hardwoods at the base, measures that limit anaerobic methane production while preserving the long-term carbon-storage and water-retention benefits the burial provides.
One further consideration belongs honestly in this comparison, because compost is widely praised and its limits are rarely named. Composting effectively breaks down a great many trace organic contaminants, most pesticide residues at typical levels, many pharmaceuticals during the thermophilic phase, food-scrap organics, and pathogens when the process reaches and holds proper temperature, and that is real and important. But peer-reviewed reviews and U.S. EPA issue papers now make plain that several classes of compound resist composting or do not break down at all. A small family of "persistent herbicides" (clopyralid, aminopyralid, picloram, aminocyclopyrachlor) was formulated specifically to resist biological degradation, and the U.S. Composting Council states plainly that most of the chemical passes through composting into the finished product; documented cases of contaminated municipal and farm compost damaging gardens have appeared across the U.S. since 1999, including in New Jersey. Per- and polyfluoroalkyl substances, the "forever chemicals", do not biodegrade naturally, and the EPA has detected them in food waste, food-contact materials, and the composts and digestates produced from them. Microplastics and nanoplastics persist; an EPA issue paper documents food-waste streams collected for composting at up to 2.8% plastic by weight, and grocery food waste at up to 300,000 microplastic pieces per kilogram. Antibiotics behave inconsistently, some (β-lactams, florfenicol) remain bioactive after composting, while others (ciprofloxacin, neomycin, tetracycline) are largely neutralized, and antibiotic-resistance genes can persist even when the parent antibiotic degrades (Patureau and Pinelli, Journal of Hazardous Materials, 2018). Heavy metals, being elements, are not destroyed at all; their concentration in finished compost can in fact rise as the bulk mass decreases.
A related question, more upstream, is what arrives in the compost pile in the first place. Most commercial and farm-scale compost is a mix of feedstocks, animal manure, food waste and crop residues, sometimes animal byproducts, and tree biomass, and the quality of each input shapes the quality of what comes out. Livestock production accounts for the great majority of antibiotic use in the United States, and the drugs are largely poorly metabolized: a survey of seventy concentrated animal feeding operations found that 98.6% of manure samples contained at least one antibiotic, with concentrations ranging up to roughly half a gram per kilogram in extreme cases (Environmental Pollution, 2020). Heavy metals follow the same path, about 72–80% of the copper added to pig feed as a growth promoter is excreted in the feces, along with 92–96% of the zinc, and direct application of untreated pig manure can raise soil copper and zinc to ten to forty times background levels. Hormones and parasiticides such as ivermectin pass through similarly; ivermectin in particular is known to be lethal or sub-lethal to beneficial soil and dung-decomposing organisms at field concentrations. Grain-fed cattle, kept on diets their digestive systems did not evolve to handle, develop unusually acidic guts in which acid-tolerant strains of E. coli O157:H7 proliferate. Plant feedstocks grown with synthetic agrochemicals carry their own residues: glyphosate does dissipate during composting, but it produces the metabolite aminomethylphosphonic acid (AMPA), which is recalcitrant in soil and phytotoxic even to glyphosate-resistant crops. Animal byproducts, blood, bone, rendered material, add their own categories of considerations involving fats, residual pharmaceuticals, and the rare but consequential question of prion contamination from ruminant tissue. None of this makes compost unsafe by itself; properly managed thermophilic composting reduces many of these substantially. The point is that the input list matters, and that compost made from unknown or low-traceability feedstocks carries more unknowns than compost made from known sources.
The same fairness that applies to compost must apply to wood. Trees grown in contaminated soil do take up some of what surrounds them, this is, in fact, the basis of phytoremediation. But two features of tree biology limit the load, and they follow directly from the uptake mechanism described in Section 11. First, root membranes act as a selective barrier; the U.S. EPA’s phytoremediation guidance states plainly that organic contaminants are not taken up at the same concentration as in the soil or groundwater because membranes at the root surface reduce the uptake, trees are more selective than the leafy annual crops that dominate the food supply. Second, what does enter is unevenly distributed: heavy metals concentrate disproportionately in roots and bark rather than the woody stem, and sapwood carries recent uptake while heartwood generally contains lower concentrations of recently encountered compounds. A field study of poplar grown on cadmium-contaminated soil found that poplar leaves did not accumulate significant cadmium (Pierzynski et al., 1994); forensic dendrochemistry studies show that uptake is recorded in narrow bands of the sapwood rather than uniformly throughout the trunk (Balouet et al., Environmental Forensics, 2007). Then a piece of good news genuinely specific to a hügel bed: the very fungi that decompose wood, particularly white-rot fungi such as Phanerochaete chrysosporium, Trametes versicolor, and Ganoderma species, produce a family of enzymes (lignin peroxidase, manganese peroxidase, and laccase) that evolved to break down lignin. These enzymes are nonspecific. The same biochemistry that opens lignin attacks the bonds of many persistent organic pollutants: DDT, PCBs, polycyclic aromatic hydrocarbons including benzo(a)pyrene, pentachlorophenol, TNT, pharmaceuticals, a wide range of pesticides, dyes, and explosives (Bumpus and Aust, Applied and Environmental Microbiology, 1987; many later reviews). This is the basis of mycoremediation, and a hügel bed is exactly such an environment. The decay process itself is, to a significant degree, remediative rather than only releasing what was inside.
The honest conclusion is that wood biomass is not innocent, but it is structurally favored. Compared with compost made from CAFO manure or unknown food-waste streams, wood from one’s own land arrives with no antibiotic load, no growth-hormone load, no animal-pharmaceutical residue, no plastic contamination, and no animal-pathogen burden. Its uptake of soil-applied pesticides is moderated by root selectivity; what it does carry is concentrated in bark and outer sapwood rather than the heartwood that sits at the deepest, slowest-decaying layer of the bed; and the fungi that eventually break it down actively degrade many of the persistent organic compounds it may contain. Heavy metals, being elements, cannot be destroyed by enzymes and may release back to the soil as the wood mineralizes, though usually in less bioavailable forms than they entered, and bound to lignin-derived humic substances that further slow their movement. That is a real limit, not an erased one. Wood from a known clean source, one’s own farm, an unsprayed woodlot, a neighbor’s storm-fallen tree, downed limbs from a known orchard, carries the shortest unknown ledger of any common amendment feedstock available to a grower. Pressure-treated lumber, railroad ties, painted wood, plywood with adhesives, and any visibly treated or stained material are always excluded from a hügel bed. So, per the rule in Section 11, is any wood that has performed filtration service. Hardwoods of known provenance belong at the deepest, longest-lasting layers. What the fairness argument asks of the practitioner is the same on each side of the ledger: know your inputs.
Two further notes deserve to be said honestly. The first is on sourcing ethic. The wood used in this farm’s hügelkultur is not from felled living trees. The carbon argument for buried wood only holds if the wood would otherwise have decomposed, burned, or been hauled away, in other words, if the bed is built from material already coming down for other reasons. Storm-fallen trees, orchard prunings, ash killed by emerald ash borer, dead limbs cleared from property edges, neighbors’ yard cleanup, and trees taken down for other land-use reasons constitute the rightful biomass source. Felling living healthy trees to fill a hügel bed would invert the carbon argument and disqualify the practice. The second note is on species diversity. A diverse mix of species, oak, maple, ash, hickory, fruit-tree prunings, locust, and others, supports a more diverse decomposer community than a monoculture pile, and a small proportion of slower-decaying species (oak heartwood, black locust, a touch of cedar) contributes structural longevity to the bed. The one species worth keeping low or excluding entirely is black walnut: juglone is unusually potent and persistent, and a heavy proportion of walnut wood is documented to suppress tomato, pepper, blueberry, apple, and azalea growth for years. Walnut at modest proportions is acceptable; walnut as a dominant ingredient is not.
What ultimately breaks the most persistent of these compounds down is either time on scales no farm can wait through, or industrial-energy interventions, hyperthermophilic composting, supercritical water oxidation, plasma reforming, or high-temperature incineration around 1,500°C for PFAS, all of which require significant electricity or fossil-fueled heat. None of this disqualifies composting. Properly managed, it remains among the better things human societies can do with organic waste, and is unambiguously better than landfilling the same material. But it does mean that imported compost of unknown provenance arrives carrying an unseen ledger, and that local biomass, one’s own prunings, downed wood, and leaves, sidesteps a large class of these risks while doing the same soil-building work. That is not an argument against compost; it is an argument that the two practices, used with informed care about feedstocks, belong together rather than in competition.
