Not far from the urban hubbub of Cape Town, South Africa, lies one of the world’s most biologically diverse yet little-known landscapes. Most of the action in the Cape Floristic Region, as biologists call it, happens below the knees—and it’s easy to overlook. To really appreciate these gray-green heathlands and scrublands, you must slow down. Bend low. Look carefully: Only then does this place reveal its subtle marvels, like the mountain rose (Protea nana), a yeasty-scented, crimson flower about the size of a child’s teacup that’s pollinated not by insects but by rodents. Or, the thick-leaved sour fig (Carpobrotus edulis) whose unfurled emerald mats protect the soil.

“You can literally walk 3 meters and encounter 30 or 40 species,” says Dewidine van der Colff, a botanist who tracks the health of South African flora on the IUCN’s Red List for the South African National Biodiversity Institute. She specializes in the plants of the Fynbos biome, which dominates the Cape Floristic Region. “If you just take the time and sit down, you see all these amazing things. I think it’s the most beautiful place in South Africa.”

The Cape Floristic Region—which evolved in a stable sliver of Mediterranean climate over millions of years—is the smallest of the world’s six floral kingdoms, yet it is also the richest, and, perhaps the most under-appreciated. The region, which stretches across the Western Cape and parts of the Eastern Cape provinces at the country’s southern end, is home to an astonishing 9,000 species, nearly 70 percent of which are found nowhere else.

This uniqueness extends to the region’s shallow, cascading streams, which begin in the Cape Fold Mountains north and east of Cape Town and empty into the Atlantic and Indian oceans. Merlot-tinted redfin minnows (Pseudobarbus spp.), Clanwilliam yellowfish (Labeobarbus seeberi) and other fish dart among the riffles and pools, shimmering in the amber sun.

“The redfin minnows have been swimming these streams at least since the late Oligocene some 20 million years ago,” says biologist and photographer Jeremy Shelton, who co-authored several studies on the region’s native fish. “Lying on the streambed surrounded by swarms of lively redfin minnows, coming face to face with curious Cape kurpers or witnessing the glint of the scales of a spawning sawfin—this is where I find my inspiration.”

Some of the species Shelton photographs have never been captured on film before. “Snorkeling in the last pockets of river untouched by invasive species or other human impacts offers a window into how things were in the past—and fuel hope for what could be in the future.”

For all its ecological significance, this oasis of biodiversity has endured tremendous losses. And Shelton’s beloved fish have suffered the most. Of the 20 native fish found only in the Western Cape province, 14 are on the “red list” of vulnerable and endangered species of the International Union for Conservation of Nature. This is part of an unfortunate global trend: While freshwater fish comprise more than half of all known fish species on Earth, freshwater biodiversity is declining twice as fast as forest or ocean biodiversity, according to a 2021 World Wildlife Fund report.

“There hasn’t been much attention at all on the biodiversity crisis that’s going on beneath the surface,” Shelton says. One reason is that some of the most imperiled populations of Fynbos fish live in the headwaters of the region’s waterways, far from many South Africans, who primarily live in the lowlands. “It’s just so easily overlooked, I think because it’s kind of hidden,” says Shelton.

The fish of the Fynbos are in a quiet struggle against myriad human pressures, including agricultural expansion, drought, and climate change. The most troubling danger, though, may be non-native fish. Introduced decades ago to boost angling opportunities, they now dominate many streams and are pushing native fish out of existence.

A controversial campaign that has restored native fish to the Rondegat River, a mountain stream in the Cape Floristic Region, has raised hopes that the same can be done to help native fish across their range here. But new and ongoing threats, along with an ever-hotter, drier climate, will only make the job harder.

By the mid-2000s, Impson and a small group of biologists and ecologists had come up with a plan to tame the intruders. They would remove invasive fish from a 4-kilometer stretch of the Rondegat River, a tributary of the Olifants River, which drains into the Atlantic. They planned to use rotenone, a plant-derived toxin that suffocates gilled animals but loses its toxicity in the water quickly. Rotenone has been used to control non-native fish populations around the world for years—sometimes despite local protests—but it would be the first time the chemical would touch a South African river.

Aware of the controversy over rotenone, Impson and his colleagues took pains to educate area residents about the chemical and the data supporting its safety. They commissioned an independent environmental analysis of the project, which recommended the use of the piscicide.

“The science was absolutely vital in showing that the restoration methodology was appropriate, because there was quite a bit of controversy about the project,” Impson says. By 2013, Shelton, filmmaker Otto Whitehead, and others had also taken up the cause. Their striking images and films of native fish species in peril helped convince local communities and landowners to support the effort.

With the public mostly on board, biologists began the grim work of killing alien fish to save native species. First, they surveyed the target stretch of river to find out what still lived there. In the invaded lower reaches, they found only one of three native fishes, the Clanwilliam yellowfish. And the density of the population was just seven fish per 100 square meters. In an upper reach, which was protected from invaders by a waterfall, the biologists found a remnant of the river’s past glory, and inspiration: Fish density was 97 fish per 100 square meters, or more than an order of magnitude higher. Survey complete, the biologists dosed invaded areas with rotenone using “drip” stations spread along the river. Then they waited.

The change was dramatic, and fast. The team returned to find about 470 smallmouth bass (Micropterus dolomieu) belly up. They also found some collateral damage: 139 Clanwilliam yellowish perished along with the bass. In the weeks and months after the dead fish were removed, the team found zero non-native fish. Redfin and other native fishes have since recolonized the streams and their offspring now swim in those waters.

“It was extremely satisfying to see the recovery of pools that never had any of these small species, these little redfin,” Impson says.

The Rondegat River experiment showed that non-native fish could be successfully purged from South African streams. But replicating this success in other streams, including former redfin strongholds—the dream of many a Fynbos biologist—will not be easy.

Foreign fish remain the biggest culprits in the decline of freshwater fishes in the Fynbos. Since many of the region’s native fish only live in one small, restricted range, they’re especially vulnerable to even highly localized threats.

But they also face other, less obvious perils. Some of the most pernicious ones come from the surrounding land. The region’s streams mostly run through privately-owned lands that increasingly are being converted to agriculture. About 26 percent of the Cape Floristic Region—including about half of Fynbos habitat—has been churned into honeybush and rooibos tea plantations, orchards, vineyards, or plots to grow ornamental flowers.

Still more development is on the way, thanks in part to economic incentives designed to increase South Africa’s agricultural exports. In the next 20 years, another 15 to 30 percent of the remaining habitat is predicted to become thirsty agricultural lands, which will further stress streams at a time when precipitation is expected to decrease under climate change.

The Cape Floristic Region enjoys little protection to gird against this creeping encroachment. A baker’s dozen of areas in the region have collectively earned designation as a World Heritage site. But they comprise less than one quarter of the CFR. Meanwhile, the development goes on. “We have drought issues, and on top of that you have farmers making use of the water,” botanist van der Colff says. “They extract water during a time when water levels are low. So now a perennial system is becoming [intermittent]. It’s quite the situation to deal with.” And it is a situation that will only grow worse, she predicts.

The result is dwindling streams and hammered fish that, despite some victories, remain mired in a seemingly endless struggle for survival.

“If you’ve got the right support from agencies, and the public, I think restoration projects can be successful, but it needs a lot of work, and a lot of funding”—funds that South Africa lacks, van der Colff says. “That’s why I feel we need to show how beautiful these things are.”

Shelton, whose deep concern for these fish spurred him to switch careers from scientist to conservationist, finds hope in the community’s new awareness of the plight of these species and of the Greater Fynbos ecosystem. “The urgency’s there, and I think people are starting to get more mobilized and more excited,” Shelton says.

For his part, he plans to keep bringing what’s beneath the surface to light. “You know, you can’t care if you don’t know what the hell is going on underwater. Hence, photography.”