This week’s episodes of What is The Future for Cities? podcast turn on a quiet, uncomfortable idea: adding more nature, or more green credentials, is not the same as building a living system that actually works. The research debate unpacked a 2017 literature review on biodiversity–ecosystem service relationships in urban areas. The interview with Caroline Howell, CEO and co-founder of Canopy, asked a still more basic question: can this land sustain the people who want to live on it? Together they move the conversation from counting plants and collecting badges toward function, carrying capacity and structural change.

A species count is not a performance test

When engineers build a bridge, they want the tensile strength of the cables. When cities expand green roofs and parks, the working assumption is often that more variety automatically means cooler streets, cleaner air and healthier people. The review looked at 317 publications and 944 mentions of biodiversity improving urban ecosystems. Only 228 of those mentions — about a quarter — were empirically tested. Of the tested relationships, 52% were positive. That is a useful tilt, not a guarantee. Worse, 47% of the claimed benefits were simply asserted, and 97% of the tests that did exist were statistical correlations rather than cause-and-effect models. Taxonomic metrics — counting species — were the most common tool, used in 41% of the tests, because they are accessible. Planners are not all trained ecologists. A headcount can still be a practical starting point. It is not the same as knowing whether the living infrastructure is doing the job.

The job is in the traits, not the spreadsheet

Function keeps leaking out of the species list. In Lahti, Finland, more plant species improved the soil’s capacity to hold nitrogen, not because twenty plants are magically better, but because different root architectures form a physical net. In Chicago, researchers counted 37 bee species in the city. Only five were doing the pollination that actually matters — the ones with the tongue length and body mass to work ornamental flowers and fly the wind tunnels between towers. An English modelling study found that planting just two species, Eucalyptus gunnii and Populus tremula, could outperform a diverse urban tree stock at carbon capture by a factor of 12. The authors still warned against doing it. Those trees have sparse canopies. They do not cast the dense shade asphalt needs, and they do not feed the caterpillars urban birds eat. Optimising for one service can flatten everything else. The missing design language is response and effect traits: how a plant handles a freeze, and how much stormwater its canopy intercepts. Without that, we risk parks that look busy on opening day and fail under climate stress.

Settlement size is a land question first

Caroline Howell starts somewhere even earlier. Before most animals settle, they ask whether the land can sustain them — food, water, and a pack or a tribe. She argues humans have forgotten the question. For her, a city begins somewhere above about 100,000 people; the work she cares about is towns of a thousand to tens of thousands, sized by what the land can hold. She is not declaring cities good or bad. She is asking how we organise ourselves, and at what size, so that food, water and energy stay in a positive relationship with the place. In practice that looks like walking or a short drive to the farms you eat from, knowing the well or the river that supplies your water, and generating energy locally rather than shipping it through extractive pipelines. Los Angeles is her warning picture: high population, vast sprawl, and millions of people living in a food desert the moment the trucks stop. The answers, she says, are almost never global. They are local — tropical dry rainforest on Panama’s Azuero Peninsula is not Iowa — and technology now lets us ask them with satellite rainfall, groundwater mapping and watershed models instead of folklore.

Endurance is not the same as adaptation

Howell draws a clean line that cities often blur. Sustainability asks whether a system can continue over the long term without depleting its foundations. Resilience asks whether it can absorb a shock and recover when conditions change. One is endurance. The other is adaptation. You cannot have one without the other, and they are still distinct. A highly efficient food system that depends on a single crop and global supply chains may look sustainable in a stable year. It is not resilient to drought, disease or a trade disruption. The research episode makes the same point from the other direction: planting one hyper-efficient tree for carbon, or celebrating a species-rich park that was never tested, is another form of brittle optimisation. Technology can help if it is used to understand living systems — soil sensors, water tables, microclimates, solar and hydro at a town scale. It does not help if it is only a better drill for a deeper well.

Humans can be keystone species again

Howell’s third fear is the one that stitches both episodes together. Recycling programmes, carbon offsets and sustainability certificates are behavioural nudges. They are not structural change. A monocrop teak plantation sold as carbon credits is not a restored watershed. A restored watershed is soil that can absorb rainfall, slow it down the hill, and put water back into the system. In Panama, the first act of her real estate project was not a show home. It was 40,000 trees of 80 native species — a food forest that does not need human watering and fertilising, planted for endangered monkeys and for a canopy that will outlive her. That is the keystone-species invitation: wolves returned to a national park and watersheds came back to life; humans could show up as the species whose presence makes more life possible. Remote work loosens the geographic anchor. Technology has finally become small enough to sit inside ecology rather than on top of it. The future she wants measured is not bigger, faster, more efficient and more concentrated. It is more alive — in biodiversity, and in how present we feel inside a living system.

The two episodes (455R and 456I) do not ask us to plant less. They ask us to stop treating a species list, or a certificate, as proof that the system works. Count what the plants actually do. Size the settlement to the land. Hold endurance and adaptation as two different tests. Then look for the structural act — the 80 species, the restored watershed — that would still be doing its job after we are gone.

What would change in your city if the first design question was not how much green we can add, but whether the land can sustain us, and whether we are making it more alive?

---

Next week we are investigating urban digital technologies with Ruthbea Yesner!

---

Share your thoughts – I’m at wtf4cities@gmail.com or @WTF4Cities on Twitter/X. Subscribe to the What is The Future for Cities? podcast for more insights, and let’s keep exploring what’s next for our cities.