Crop biotech
Allelopathic crops
Benzoxazinoids and sorgoleone are genuine root-released herbicides. What limits them is not potency but dose at the target, soil persistence and the crop's tendency to poison itself.
Some plants release compounds that suppress the germination and growth of their neighbours. This is not folklore: the molecules are identified, their modes of action are known, and the genes that make them have been mapped. The difficulty is that a compound which inhibits a seedling in a Petri dish has to survive a soil, reach a weed root at a sufficient concentration, and not damage the crop that follows — three filters that remove most of the effect measured in the laboratory.
The two best-characterised chemistries
Rye, wheat and maize make benzoxazinoids. The plant stores them in the vacuole as harmless glucosides; when tissue is damaged, a β-glucosidase meets the substrate and releases the aglycone DIMBOA or DIBOA, which is the toxic form. This compartmentalised arrangement is the same logic as a glucosinolate or a cyanogenic glycoside: the plant carries a weapon it does not have to survive itself. In soil the aglycones degrade further, to aminophenoxazinones, and these breakdown products are more persistent and in several assays more phytotoxic than the parent — so the residue matters more than the exudate.
Sorghum root hairs exude sorgoleone, a lipid benzoquinone, in droplets at the root surface. It inhibits photosystem II at the same binding site as the triazine herbicides, and also interferes with mitochondrial electron transport. It is strongly hydrophobic, which cuts both ways: it accumulates in the rhizosphere rather than washing out, but it also binds tightly to soil organic matter and moves very little from the root that made it. Rice contributes momilactone B, a diterpenoid that doubles as a defence compound against pathogens.
Why the field result is smaller than the bioassay
Three things attenuate it. Sorption to clay and organic matter removes free compound from solution. Soil microorganisms metabolise these molecules readily — they are carbon, and the rhizosphere is carbon-hungry. And the weed seed bank is not uniformly exposed; suppression is a function of distance from a root, so it is patchy at exactly the scale weeds exploit.
There is also a persistent attribution problem. A rye cover crop suppresses weeds by shading, by physical mulch, by nitrogen immobilisation and by chemistry at once. Separating the chemical contribution requires activated-carbon controls or near-isogenic lines differing only in the biosynthetic pathway, and studies that do this generally find the chemical share to be real but smaller than the total suppression.
Autotoxicity is the structural limit
A compound selective enough to be useful is rare. Benzoxazinoid and sorghum residues can suppress the following crop as well as the weeds, and the effect is worst in the situation growers most want — high residue, no tillage, short interval to planting. Species with well-documented autotoxicity, alfalfa and asparagus among them, show the general shape of the problem: the trait cannot be maximised, only tuned against rotation. Breeding for higher allelopathy therefore means breeding for a specific ratio between suppression of a target weed and tolerance in the next crop, which is a much narrower target than “more chemistry”.