The Fence Grows Into a Tree. Whether Anyone Mends It Depends on the Neighbour.

2,117 corrals in northern Tanzania are fences grown into living trees, and a controlled comparison credits them with stopping 99.9 percent of night attacks. In Kenya, the best predictor of whether a fortified boma got repaired turned out to be the distance to the next one.

A traditional boma is a ring of cut thorn branches around a homestead, and on the Maasai Steppe in northern Tanzania it is the only thing standing between a family's cattle and the night. It is built from acacia, which is free and grows everywhere, and it begins failing the day it is finished. The branches dry out and gaps open at the base. A lion that has decided to come through will come through, and in the morning the owner counts what is left of the herd, and then somebody goes out to find the lion. That last part is why ecologists care about fences.

The alternative now standing in more than two thousand Tanzanian homesteads is called a Living Wall. Crews cut limbs from Commiphora, the tree known locally as African myrrh, and plant them in a circle around the homestead inside a shallow furrow. Chain-link fencing goes into the furrow and gets tied to the standing limbs. Then the cuttings do what Commiphora does, which is root and grow. The posts turn into trees. The wire ends up inside them.

By the end of 2025, African People & Wildlife — the organisation Laly Lichtenfeld runs, working across six landscapes in northern Tanzania — counted 2,117 Living Walls in 80 communities, 24,634 people living behind them, and 276,684 trees grown into the fencing. Its field team checks the structures over the years that follow. The owner pays a quarter of the construction cost. Hold on to that quarter; it comes back later.

Where the 99.9 percent comes from

Most coverage of Living Walls leads with the figure 99.9 percent, and it is worth knowing what was measured to get it. Lichtenfeld, Charles Trout and Elvis Kisimir published the comparison in Biodiversity and Conservation in 2015: 62 fortified bomas against 84 unprotected ones, night-time carnivore attack rates, 1,790 boma-months of observation. The fortified ones stopped 99.9 percent of night attacks. After installation, no lions were killed at any homestead that had one.

Strong numbers. The odd thing about that paper is that it exists at all.

Almost nothing in this field has been checked

In 2017 a team led by Ann Eklund at the Swedish University of Agricultural Sciences went through the scientific literature from 1990 to 2016, looking for evidence that any of the things people do to stop large carnivores eating livestock actually reduce the losses. Guarding dogs, enclosures, killing predators, shock collars, sterilisation, noise and light deterrents, switching to a livestock species that gets eaten less — all of this is done at scale, across continents, mostly on public or donor money.

Studies using a case-control design: twenty-one. That was 3.7 percent of the publications they reviewed.

Their conclusion was not that the interventions fail. It was that almost nobody had checked, so nobody could say. Set Eklund's paper beside Lichtenfeld's and what stands out in the Tanzanian one is a conservation organisation running a controlled comparison against its own flagship programme and printing the number it got.

When they break

Fortified bomas do fail, and there is a study about that too. It comes from a different country and a different programme, which matters.

In August 2018, David Owino Manoa and seven colleagues surveyed 88 predator-proof bomas in the Amboseli ecosystem of southern Kenya, sampled from 294 built between 2012 and 2018. Fifty-seven of them — 64.8 percent — were damaged. The results appeared in Oryx in 2022.

Three variables predicted the damage. Wooden posts did far worse than recycled plastic ones: every single wooden-post boma in the sample was damaged, against 55 percent of the plastic-post ones. Bomas holding fewer livestock were in worse shape, which the authors read as damage going unrepaired rather than damage being caused. And the third variable was distance to the nearest other predator-proof boma. The further your neighbour's fortified boma stood from yours, the more likely yours was broken.

The best single thing you could know about whether a structure got mended, in other words, was how close the next one stood. The authors describe it as a social element and are careful not to claim they know the mechanism — seeing it done, knowing who to ask for wire, being the one homestead on the ridge whose fence is visibly sagging. The implication for anyone spending the money is sharp all the same. Take identical materials, an identical budget and an identical number of bomas, scatter them thinly across a district instead of clustering them in one, and five years later you are looking at a different landscape.

Amboseli is not the Maasai Steppe, and those bomas are not Living Walls. Living Walls use trees instead of posts, which removes the variable the Kenyan study found mattered most, and the two programmes are run by different people under different conditions. But maintenance is everybody's problem, and the quarter of the cost that Tanzanian owners pay is one answer to it.

The same instinct, applied to grass

Last October, at the IUCN World Conservation Congress in Abu Dhabi, African People & Wildlife won the first Tech4Nature award in the NatureTech Stewards category, for work with no lions in it. Neovitus Sianga, the organisation's director of community conservation and environment, submitted it.

The programme trains people from the villages as habitat monitors. They walk their own grazing land on a monthly cycle and record grass height, vegetation cover, soil condition and signs of invasive species on phones, feeding dashboards the communities themselves read. Herds then get moved off the stressed patches, and restoration work goes where the data points. The summary circulated with the award puts it at more than 800,000 acres under community management and roughly 20,000 acres restored.

IUCN's own write-up of the award carries no acreage figure and no measured outcome, and no published study yet says what that pasture would have done without the monitoring — precisely the gap Eklund's team was complaining about.

What a fence does not fix

A Living Wall protects livestock at night. Animals are also lost in daylight, out on open range with a herder, and no enclosure touches that. Lions are in trouble for reasons much larger than retaliation: rangeland converted and wild prey shot out, with the surviving populations cut into fragments too small to exchange animals between them. Even the scale of the trouble is contested. The 2015 paper by Hans Bauer and colleagues, projecting that East African lion numbers would halve within two decades, drew a published objection and a published reply in the same journal.

Then there is the money. Owners pay a quarter, which means three quarters arrives from elsewhere, and elsewhere means grants, renewed one budget cycle at a time.

Two thousand structures across eighty communities, a quarter of the cost paid by the people standing behind them, built out of a plant that gets harder to push through with every year it stays in the ground, and a published comparison saying they do the job. In a field where 96 percent of the literature never tested anything against a control, that is an unusual number of things done in the right order.

Sources

• Living Walls programme, African People & Wildlife — https://www.africanpeoplewildlife.org/our-work/programs/living-walls
• L.L. Lichtenfeld, C. Trout, E.L. Kisimir, "Evidence-based conservation: predator-proof bomas protect livestock and lions", Biodiversity and Conservation 24, 483–491, 2015 — https://link.springer.com/article/10.1007/s10531-014-0828-x
• A. Eklund, J.V. López-Bao, M. Tourani, G. Chapron, J. Frank, "Limited evidence on the effectiveness of interventions to reduce livestock predation by large carnivores", Scientific Reports, 2017 — https://www.nature.com/articles/s41598-017-02323-w
• D.O. Manoa et al., "Drivers of predator-proof boma disrepair in the Amboseli Ecosystem, Kenya", Oryx 57(2), 196–204, 2022 — https://doi.org/10.1017/S0030605321001642
• "Tech4Nature award – NatureTech Stewards winner: Community-led technology for resilient rangelands in Tanzania", IUCN — https://iucn.org/news/202602/tech4nature-award-naturetech-stewards-winner-community-led-technology-resilient
• "Winners of the inaugural Tech4Nature Awards announced at IUCN World Conservation Congress in Abu Dhabi", IUCN, October 2025 — https://iucn.org/press-release/202510/winners-inaugural-tech4nature-awards-announced-iucn-world-conservation
• H. Bauer et al., "Lion (Panthera leo) populations are declining rapidly across Africa, except in intensively managed areas", PNAS, 2015 — https://www.pnas.org/doi/10.1073/pnas.1500664112