How a closed hedge spreads water, sorts sediment, and hands nitrogen and phosphorus to the crop below — and why a leaky one does the reverse
I have described a properly functioning vetiver hedge as a poor man’s irrigation system for a long time. It was meant as a way of explaining to a farmer what the hedge does with rainfall: it takes water that would have run off in a few destructive channels and instead gives it back to the field, spread out, where the crop can use it. I have always thought the description was sound. I am no longer sure it goes far enough, because the hedge is not only redistributing water. It is redistributing what the water is carrying — and it is doing so through a piece of hydraulic engineering that I do not think we have ever properly described.
What actually arrives at a hedge
On a cultivated slope, runoff does not arrive as a uniform sheet. Within a few meters of leaving the point where it fell, rainfall that exceeds the infiltration rate collects into rills and depressions, and from then on it travels as concentrated flow. That is why erosion is patchy, why gullies start where they do, and why two parts of the same field can yield quite differently in the same season.
That concentrated flow carries nutrients in three fractions, and they behave completely differently. The distinction matters more than it looks.
The coarse fraction is sand and the larger soil aggregates. It is heavy, it settles quickly, and it carries relatively little plant food per unit weight.
The fine fraction is clay and fine silt, with the organic colloids that travel with them. This is where the fertility is. Phosphorus is adsorbed onto clay surfaces and onto the iron and aluminum oxide coatings on them, and most organic nitrogen travels attached to the same material. Eroded sediment is routinely two to five times richer in nitrogen and phosphorus than the soil it came from, and this is why. The fines are the nutrients.
The dissolved fraction, on cultivated land, is largely nitrate. Nitrate carries a negative charge, and so does clay, so it is not held by the soil in the way ammonium or phosphate are. It simply moves with the water.
What the hedge does with each
A closed vetiver hedge slows the flow, ponds it briefly behind the stems, and releases it below. This is the porous weir behavior we have described for thirty years. What is less often said is that the hedge does not simply trap sediment. It sorts and spreads it.
The coarse fraction drops out above the hedge, because settling velocity rises steeply with particle size and sand needs only seconds of slow water to fall out. That is the sediment wedge, and over time it becomes the natural terrace. Everyone who works with vetiver knows this part.
The fine fraction does not drop out there. A clay particle a couple of microns across settles so slowly that the few seconds or minutes of ponding behind a hedge are nowhere near enough to bring it down. It stays in suspension and passes through the stems — and it takes the phosphorus and the organic nitrogen with it.
The dissolved fraction passes through as well, for a different reason: the hedge cannot hold nitrate any more than the soil can.
So two of the three fractions go through the hedge, and between them they carry most of the plant food. What happens to them next is the whole question, and the answer is hydraulic. Water that reached the hedge in two or three concentrated channels is ponded, spread laterally along the hedge line, and released across the full width of the field below as a broad, shallow front. Spreading the same discharge across thirty metres instead of two collapses the velocity. Fines that stayed in suspension in a fast rill now settle out, not in a wedge at the hedge, but diffusely across the whole alley below it — where the crop is.
The hedge sorts what is passing through it. The coarse material builds the terrace. The fine material, which is where the nitrogen and phosphorus are, is spread across the field below and dropped where the crop can reach it.
That is why I think the layman’s word for it is not far off. Nothing is being applied, so it is not fertigation in the strict sense, and any agronomist will say so. But what the farmer sees — water and dissolved plant food arriving evenly across the field instead of rushing off it — is close enough to the idea that the word communicates something true. For technical writing I would call it lateral redistribution of dissolved nutrients. For a farmer standing in a field, fertigation will do.
A known failure, read the other way round
There is a useful irony here. In the riparian buffer strip literature, the pass-through of dissolved nitrate is treated as a well-documented failure. Buffer strips trap sediment and particulate phosphorus efficiently, and nitrate goes straight through them into the watercourse. Papers have been written about the limitation for forty years.
Inside a field, the same physics is a benefit. A buffer strip is trying to keep nutrients out of a river. A contour hedge is trying to keep nutrients in a field and give them to a crop. The property that makes vegetative barriers a disappointment at the river bank is precisely what makes them useful on the slope. The mechanism was never in doubt. Only the sign was.
Why a hedge is not a bund
An earth bund is impermeable. Water arriving along its length cannot pass through it, so it travels sideways along the upslope face until it finds the lowest point, and everything the bund has intercepted leaves through that one place. The discharge at the breach or at the end of a graded bund is far greater than anything the slope would have produced on its own. That is why bunds fail catastrophically, why the failure is always at a point, and why a breached bund so often leaves a gully where it stood.
A hedge does not behave that way. It is a porous weir, and it releases water through its whole length continuously. The ponding behind it is shallow, a few centimeters at most, so the lateral gradient driving water sideways toward a gap is weak. Most of the flow simply goes through the grass where it arrives. A hedge with a few gaps is still spreading water along the rest of its length.
This is one of vetiver’s real advantages over the structures it competes with, and we do not say it often enough. A bund fails at a point. A hedge loses a little performance where the gap is.
A bund does not redistribute anything. Because it is impermeable, water stands behind it until it soaks in or goes over the top, and standing water settles everything. The coarse fraction drops out, and given long enough so do the fines, with the phosphorus and organic nitrogen they carry. All three fractions end up in the same narrow strip at the foot of the structure. The soil immediately above the bund becomes rich. The rest of the interval below it gets nothing at all.
A bund retains fertility. A hedge distributes it. Only one of those is of any use to a crop growing thirty meters further down the slope.
That is the striping familiar in bunded fields: a good band of crop immediately above the structure and an indifferent one through the rest of the interval, made worse where a borrow ditch has exposed subsoil just below. It is the reverse of what Babalola recorded under vetiver, where the best maize was in the lower portions of the slope.
To be fair to the bund, a well-built one that does not breach probably holds more total nutrient in the field than a hedge does, because it lets almost nothing through. The difference is not retention but availability. Nutrients concentrated in a strip a meter wide feed the plants in that strip. Nutrients spread across the alley feed the crop. And the retention lasts only as long as the bund does — when it finally breaches it releases in a single event everything it has accumulated, together with its own material.
And that is the level bund, which at least tries to keep the water. A great many bunds are graded, and there the argument is sharper still.
And that is the level bund, which at least tries to keep the water. A great many bunds are graded, and there the argument is sharper still.
A graded bund is not built to hold water. It is built to move it. It runs across the slope at a designed gradient so that intercepted runoff travels along the channel to a waterway or a drain and leaves the field in a controlled way. The design velocity is chosen to be non-scouring but also non-silting — which means, in plain terms, fast enough that the fine fraction never gets the chance to settle.
A graded bund is engineered to collect the enriched fraction of a farmer’s soil and deliver it to a watercourse. It does the job well.
That is the drainage tradition applied to farmland. It is coherent engineering if the objective is to protect the field from water. It is the wrong objective if what you want is to keep the water and the nutrients on the field. And the material it conveys is precisely the material that does the damage at the other end: phosphorus attached to clay is the main eutrophication pathway into lakes and reservoirs. A landscape of graded bunds is an efficient collection system pointed at the nearest water body.
In fairness, grade is often chosen for a reason. On soils that take water slowly, or in rainfall heavy enough that a level structure would pond, overtop and fail, safe disposal is the prudent choice. But that reasoning assumes the runoff volume is fixed. It is not. A hedge raises infiltration, so there is less water to dispose of in the first place, and the calculation that justified the grade may no longer hold.
A contour hedge has no grade and nothing to convey. It disposes of water vertically, into the soil, rather than laterally, into a drain. That is the difference between the two traditions in a sentence, and the nutrient consequences follow from it.
The same asymmetry applies to the water. A bund concentrates infiltration along one line, which can waterlog that line while the rest of the interval stays dry. A hedge spreads infiltration across the whole width below it. In both cases the impermeable structure concentrates and the porous one distributes, and it is distribution that the crop actually experiences.
The nutrient argument follows the same logic, and with the same proportionality. Water leaving through a gap has not been spread, so its fines and its nitrate go out of the field instead of being handed to the crop below. The loss is local to the gap rather than catastrophic, but it is a loss, and it is a loss of exactly the fraction that matters most.
So the practical points stand, on more modest grounds:
- A gap is a local loss of the spreading function. Water leaving through it has not been slowed or spread, so the fines stay in suspension, the moisture is not distributed, and a streak of poorer crop appears below it.
- The damage appears below the gap, not at it. In a mature hedge the gap is above ground only. The roots of the neighboring clumps have long since met beneath it, so the soil under the gap is reinforced and does not scour. The flow passes through faster than it should, lands on unreinforced ground a little way downslope, and a rill can start there.
- That protection belongs to maturity, not to hedges in general. A hedge that never closed, or one with a long run of dead or missing plants, has no root coverage to fall back on. The first season decides this.
- On steep ground and at high discharge a large gap can still become a preferential path whatever is happening below ground.
The evidence was already there
Having set out the mechanism, I went back to the older work to see whether anyone had unknowingly measured it. Two Nigerian studies had, and one of them has been sitting in our own ICV3 proceedings since 2003.
Babalola and colleagues at the University of Ibadan ran vetiver strips at 20 m intervals on a 6% slope, in 40 m by 3 m erosion plots, over three seasons, with maize fertilized at 100 kg N per hectare as urea. Four of their results bear directly on this argument, and I do not think anyone has read them together before.
First, the deposition and the yield were in different places. Soil accretion at the first vetiver strip down the slope was 9.4 cm; at the second strip it was 0.17 cm. Yet the tallest maize plants and the significantly higher grain yields were found in the lower portions of the slope, below the strips. If the sediment wedge were driving yield, the best crop would have been at the top of the plot. It was at the bottom.
That fifty-fold difference between the first and second strips is worth pausing on, because sediment supply alone does not explain it. It is what sorting predicts. The first hedge takes out essentially the whole settleable coarse fraction; what reaches the second hedge is the fine material, which does not build a wedge because it does not settle in the time available. It settles further down instead, spread across the alley — which is exactly where the yield was.
Secondt, the soil moisture gain was concentrated below the hedge rather than above it. Measured by neutron probe, moisture at 20 cm depth was 5.2% higher than the control at a point above the first strip, and 50.1% higher between the first and second strips. Nearly ten times the gain, on the downslope side.
Third, and this is the one I did not expect: they measured nitrate in the runoff water, and it was higher leaving the vetiver plots than the untreated ones — 0.59 ppm against 0.43 ppm.
Babalola attributed it to better nutrient enrichment of the lower slope confined by the strips. That is precisely the pass-through of the dissolved fraction. The hedge did not hold the nitrate back; the nitrate moved down the slope with the water and enriched the ground below. And because runoff volume over the measured storms was 0.30 mm on vetiver plots against 3.88 mm on the controls, roughly nine times more dissolved nitrogen actually left the untreated plots. Concentration up, total loss sharply down. The nitrogen stayed in the field because the water stayed in the field.
Fourth, and most useful of all for anyone arguing with a ministry: nitrogen use efficiency rose from 7.7% to 12.6%, an improvement of about 40%. That is a measured fertilizer efficiency gain attributable to a grass hedge.
Babalola’s later work with S.O. Oshunsanya and K. Are (2007), on vetiver strips, vetiver mulch and organomineral fertilizer, took the nutrient loss question further. Between them, these two studies are the closest thing we have to a direct test of what a hedge does with dissolved nutrients under fertilizer.
A spacing gradient from Nigeria
Effiom Oku’s work in southeastern Nigeria approaches the same question from a different direction, and the design is unusually informative. On a 45% slope — a very steep cultivated field — he laid twelve erosion plots of 50 m by 3 m in a randomized complete block design, with vetiver buffer strips at 5 m, 15 m and 25 m intervals and a fourth treatment with no vetiver at all. Cassava and maize on traditional mounds, NPK at 300 kg per hectare to the maize.
That is not a within-field distance gradient. But it is a hedge density gradient with four points, and it says something the distance gradient could not.
Each cell gives year one, then year two. Rainfall was 1,199 mm and 711 mm. Soil loss is expressed as Oku reports it, in multiples of the 12 t/ha/yr limit taken as acceptable for the humid tropics.
| Treatment | Runoff, % of rainfall | Soil loss, multiples of the acceptable limit | Maize yield above the untreated plots |
| No vetiver | 29%, then 21% | 76–98×, then 12–15× | baseline |
| Strips at 5 m | 7%, then 8% | 3–23×, then 0.2–0.6× | +35%, then +47% |
| Strips at 15 m | 12%, then 11% | 18–23×, then 0.6–1× | +27%, then +41% |
| Strips at 25 m | 13%, then 11% | 16–24×, then 0.3× | +25%, then +40% |
Two things stand out, and it is the second that matters here.
The first is the size of the step from no vetiver to any vetiver. Runoff more than halves. Soil loss goes from seventy-six to ninety-eight times the acceptable limit down to a fraction of it by the second year. Nobody disputes that part, and it is the reason vetiver gets planted.
The second is how little the remaining difference buys. A 50 m plot holds about ten hedges at 5 m spacing and two at 25 m — five times the grass in the field. Runoff does respond to that: 7% against 13% in the first year. Yield barely does. Twenty-five percent above the untreated plots at the widest spacing against thirty-five at the tightest, and in the second year forty against forty-seven.
Five times the hedge, for roughly a tenth more yield. Whatever is raising the yield is not confined to a band around each hedge.
If the benefit were a meter or two either side of the grass, ten hedges should have delivered several times the benefited area of two, and the yields should have separated far more than they did. The influence has to extend across the whole alley — which is what a redistribution mechanism predicts, and what a deposition mechanism does not.
One caveat on the soil loss column, which I would not lean on in either direction. Oku reports it as ranges of multiples of a threshold rather than as tonnages, and in the first year the ranges for the three spacings overlap almost completely. In the second year the 25 m plots came out better than the 15 m plots, which cannot be right as a spacing effect. His conclusion that tighter is better is almost certainly correct, and the runoff figures support it plainly, but the soil loss data as published do not establish it.
The second year is the more striking result. Rainfall fell from 1,199 mm to 711 mm. Cassava yields on the untreated plots collapsed by 80%. On the vetiver plots they rose — by 23% at 5 m spacing, 16% at 15 m and 13% at 25 m. Maize fell 10% without vetiver and rose 10% with it. In a poor rainfall year the protected plots did not merely hold their ground; they improved.
Note also that the spacing gradient reappears in the drought year, and more clearly than in the wet one. When water is the binding constraint, how tight the hedges are starts to matter again. That is an argument for close spacing that has nothing to do with erosion.
Oku reaches the irrigation conclusion himself, independently and in almost the same words. He calculates that with vetiver a farmer obtains the same yield from 852 mm and 561 mm of rain in the two years, and writes that vetiver causes the farmer to enjoy the benefit of every raindrop on the farm.
One limitation should be stated plainly, because it is the limitation on this whole argument. Oku measured carbon and nutrient losses in eroded sediment only. He says so himself — nutrient losses in the runoff water were not accounted for. So his data, powerful as they are on soil and water, cannot speak to the dissolved fraction. Babalola’s can, and did.
Bharad’s ranking
One older result deserves a mention for the pattern rather than the magnitude. Bharad, at Akola in Maharashtra, reported productivity increases of 55.4% on very shallow soils on farm, 28.5% on shallow soils on station and 14.49% on medium-deep soils on station, where cultivation and sowing followed vegetative contour lines. The gain is largest where the soil holds least water, which is what a water redistribution mechanism would predict. The on-farm and on-station comparison is not clean, so I would not press it — but the direction is right.
Malawi: a case that is probably about something else
I had intended to offer Tiyeni deep-bed farming in Malawi as a third example. Permanent vetiver marker lines, very large yield gains, Ministry approval, a long record. On reflection I do not think it belongs here at all, and the reason is worth setting out, because it shows how these mechanisms can be told apart.
In the Tiyeni system the water is handled by the earthworks, not by the grass. Contours are pegged and marker ridges built with a ditch on the upslope side that acts as a dam after heavy rain; the beds between are separated by closed-end furrows. The system is designed to impound water where it falls rather than to pass it through. The hardpan is broken in the first year, which by itself transforms rooting depth and infiltration. The water benefit is therefore largely accounted for before vetiver is considered at all. There, the grass is stabilizing a marker ridge — essential, but a structural role.
So what does the vetiver contribute that the earthworks do not? Ridge permanence, certainly. But there is a second candidate, and I now think it is the more likely one. It is a maize story.
Malawi is a maize country, and Tiyeni gardens are maize-dominant and continuously cropped. Vetiver is a dead-end trap crop for the spotted stem borer, Chilo partellus: moths prefer to lay their eggs on it rather than on maize, and larval survival on vetiver is close to zero (Van den Berg, 2006). Tiyeni’s layout gives that mechanism unusually favorable geometry. Push-pull systems normally place the trap crop as a border around a plot. Tiyeni has permanent vetiver lines at regular intervals down the slope, so every maize plant in the garden is within a few meters of a trap row. In effect it has built a grid of them, without setting out to.
There is a sharper point. From the second season onward the method leaves standing maize stalks in place — and crop residue is exactly where stem borer larvae carry over between seasons.
Stalk destruction is the oldest cultural control for stem borers there is. A no-till system that leaves stalks standing should, other things being equal, build borer populations year on year. Tiyeni’s yields do not behave that way. If vetiver trapping is offsetting the carry-over that the residue practice would otherwise create, that would account for both observations at once — and it would mean the two practices are working as a pair rather than independently.
How far this would reach
The obvious objection is altitude. Busseola fusca, the other main maize borer, showed no oviposition preference for vetiver in the same study, so wherever Busseola dominates the trap should not work. That sounds like a serious limitation until you look at where maize is actually grown.
Malawi’s main maize areas lie at 1,200 m and below. That is Chilo partellus country. The same is true of most of Tanzania, most of Zambia and most of Zimbabwe. The bulk of the southern and eastern African maize crop is grown at elevations where the borer that vetiver traps is the borer that matters.
If the mechanism is real, it is not a curiosity of one Malawian system. Every vetiver contour hedge installed for erosion control across the regional maize belt may also be trapping stem borers — a benefit nobody is counting, in systems already in the ground.
The exceptions are the higher ground: Zimbabwe’s highveld, the Kenyan highlands, the Ethiopian highlands. There Busseola takes over and the effect should fade. That is not an awkward caveat but a useful prediction, and it is testable with plots that already exist at different elevations. It also means that where vetiver raises maize yields in the high Ethiopian highlands, the explanation has to lie elsewhere — in water, in mycorrhizal nutrition, in nematode suppression — which is a sharper way of separating the mechanisms than any single site could give us.
Two other cautions stand. Fall armyworm has changed the regional maize picture since 2017 in ways not yet settled, and it should be recorded separately rather than lumped into a single pest-damage score. And hedge cutting timing would matter: cutting after egg laying removes the eggs with the leaf and strengthens the trap, while cutting before moth flight removes the attractant. Nobody, as far as I know, has ever cut a vetiver hedge with that in mind.
But it is testable, and Tiyeni is close to an ideal site for the test. Uniform maize; vetiver lines at regular known spacing, which gives a natural distance gradient at no cost; many gardens; neighboring non-Tiyeni gardens as controls; and an established research relationship with the University of Worcester. Deadheart counts, stem tunnelling length and exit holes by distance from the vetiver line, together with egg batch counts on the hedge itself and the borer species recorded, would go a long way in a single season.
I set this out in a piece arguing for the water and nutrient mechanism because being able to say that a well-known success is probably not an example of it is part of arguing for it honestly. A hedge does several things at once. Working out which one is doing the work in a particular field is the whole task, and we have too often been content to observe that the yields went up.
All of this is hydraulics
It is worth standing back and naming what is actually doing the work here, because it is easy to reach for the wrong kind of explanation.
The hedge is not binding nutrients chemically. It is not, on the evidence, taking up much of what passes through it. What it is doing is manipulating flow. It reduces velocity, converts concentrated channel flow into shallow sheet flow, widens the wetted front, and buys infiltration time. Everything else in this argument follows from that: which particle sizes drop and where, how far the dissolved fraction travels before it soaks in, how evenly the moisture is distributed, how much of the rainfall is still on the farm an hour after the storm.
Vetiver’s distinctive contribution here is hydraulic, not botanical. The stems are stiff enough to stay upright under flow, so the hedge goes on spreading water at discharges that would flatten a softer grass and let the flow pass straight over the top.
This is why I have become uncomfortable with vetiver being filed under soil conservation. Holding soil is one consequence of managing water. Redistributing rainfall is another. Spreading the fine fraction and its phosphorus across the field below is a third. They are the same property seen from three sides.
It is also why hedge closure is the variable to manage. Closure is what sets the width of the wetted front, and the width of the front is what sets the velocity. A gap is a hole in that front: water passing through it has not been slowed or spread, so its fines stay in suspension and its nitrate leaves with them. Because the hedge is porous, the rest of the line goes on working, which is not the case for an impermeable structure. But the field below a gap does not get what the field below the rest of the hedge gets.
What is still missing
Babalola has already supplied the nitrate in the runoff water, the moisture gradient by position, and the slope-position split in yield. What nobody has done is measure the nutrients in the soil by position across the alley, which is where the mechanism would leave its clearest mark. The test is cheap: measurement by position within the field rather than plot averages, along a transect running from one hedge down across the inter-hedge strip to the next.
The prediction that distinguishes this mechanism from the alternatives is a change in the character of the deposit across the hedge line. The distinction to hold onto is between how much material is laid down and how rich it is, because the two go in opposite directions.
- Above the hedge: a great deal of material, relatively poor. The wedge is built from sand and coarse aggregates, which carry little plant food per unit weight. The soil there gets deeper without necessarily getting richer.
- Below the hedge: less material, considerably richer. Clay and organic colloids settling across the alley, carrying the phosphorus and most of the particulate organic nitrogen with them.
- Phosphorus is the cleanest discriminator of the two, because it has almost no coarse carrier at all. Concentration per unit weight of deposit should rise below the hedge and be diluted above it.
- Mineral nitrogen — nitrate and ammonium — should peak in the first band below the hedge and decline with distance from it.
- Particle size should change to match: sand-enriched in the wedge, clay-enriched in the alley.
I predicted that total nitrogen and organic carbon would peak in the sediment wedge. That confuses mass with concentration. Nitrogen and carbon travel on the light, fine fraction, and the light fine fraction is exactly what goes through the hedge. What piles up above it is the heavy, poor fraction — together with whatever crop residue and coarse debris the hedge catches, which is a separate matter.
That last component suggests one more cheap test. Organic matter trapped as plant residue has a wide carbon to nitrogen ratio, somewhere between forty and eighty. Organic matter arriving as humified soil colloids has a narrow one, around ten or twelve. So the ratio in the deposit should be wide above the hedge, where the trash line is, and narrow below it, where the colloids settle. As a check on the reasoning, Oku’s eroded sediment ran at about ten to one, which is soil rather than residue — exactly what one would expect of material collected at a plot outlet rather than at a hedge.
Nothing else predicts that pattern. Coarse deposition is entirely an above-hedge phenomenon. Improved infiltration changes the variability of yield rather than its mean. Mycorrhizal effects follow root density and would center on the hedge rather than step across it.
Add yield recorded in bands by distance rather than as a plot mean — which costs nothing beyond harvesting in strips — and the same transect tests several mechanisms at once.
Four caveats
The hedge takes up nitrogen itself. We argue elsewhere, correctly, that vetiver removes nitrogen and phosphorus from wastewater. Both claims can be true, because retention depends on contact time: a treatment wetland is designed for long residence, a contour hedge for very short residence by construction. Babalola’s nitrate figures suggest hedge uptake is not dominant — concentration in the runoff rose rather than fell as the water passed through the system. But a dense hedge with heavy surface rooting will take some share, and if the cuttings return to the field as mulch or manure that nitrogen stays in the system by a slower route. We should state the trade-off rather than make both claims loudly and hope nobody notices.
Timing decides whether it is fertigation or leaching. Nitrate delivered in a heavy storm before the crop has a root system will move below the root zone. Nitrate delivered mid-season into an actively growing crop gets used. The benefit is real but seasonal, and it should be tied to storm timing rather than expressed as an annual figure.
Fines that pass one hedge may leave the field altogether. Vetiver redistributes nutrients; it does not create them. Within a field the fine fraction is handed to the crop below, but at the bottom boundary it goes to the next field, or eventually to a watercourse — and phosphorus attached to clay is the main eutrophication pathway. We should not claim that vetiver solves phosphorus export. There is a related point for anyone reading erosion plot data: a plot outlet counts fines leaving the plot as a loss, when in a real field most of that material would have landed a few meters further down the slope. Small plots systematically overstate field-scale nutrient loss, and Oku’s nutrient figures should be read with that in mind.
Denitrification works against it. The hedge line is exactly where you find periodic ponding, accumulated organic carbon and anaerobic microsites — the conditions for nitrogen loss to the atmosphere, some of it as nitrous oxide. This should be measured rather than assumed away, and it is better raised by us than by a critic.
Why this matters beyond the agronomy
Governments across Africa and Asia spend very large sums on fertilizer subsidy, and the return on that spending is limited by how much of the applied nutrient stays in the field long enough to be taken up. Every ministry of agriculture understands fertilizer use efficiency, and most are under pressure to improve it.
Babalola measured a 40% improvement in nitrogen use efficiency from a grass hedge. If that figure holds up more widely, it is an argument with a budget line attached to it — which the erosion argument, after forty years, still does not have. It would also mean that vetiver and fertilizer are complements rather than alternatives, which is a more comfortable message for an agriculture ministry than anything we usually bring.
But the argument only works if the hedges are closed. A gappy hedge does not redistribute water, does not redistribute nitrate, and concentrates erosion at its weakest point. If we are going to make this case, we have to be able to show that the hedges we promote actually close — which brings us back, as it always does, to standards, supervision and an acceptance test.