The Morning the Frogbit Took Over

The Morning the Frogbit Took Over

The first sign of trouble wasn’t visible until the third week. By then, the ten-gallon tank looked less like an experiment in natural aquarium keeping and more like someone had dumped a bag of salad greens into it and walked away. Frogbit — Limnobium laevigatum — had carpeted the surface so completely that the light barely reached the water below. The floaters weren’t just growing; they were reproducing like something that had been waiting its whole life for exactly these conditions.

The Walstad Method, as Diana Walstad laid it out in her 1999 book Ecology of the Planted Aquarium, is built on a deceptively simple premise: a layer of organic potting soil capped with sand or gravel, heavy planting from day one, no CO2 injection, no filtration beyond a small sponge or powerhead, and minimal water changes. The soil feeds the plants; the plants clean the water; the whole thing finds a balance — in theory — that requires almost nothing from the keeper after the initial setup. It’s a system that sounds like it rewards patience, then rewards it again. What the literature doesn’t always mention is how quickly it can punish enthusiasm.

This particular setup started in late February, in a rented apartment in a district of Singapore that smells like frying garlic and damp concrete most evenings. The tank sat on a low wooden cabinet near a west-facing window, which seemed like a good idea at the time. A small clip-on LED fixture rated at about 20 micromoles added supplemental light on a timer set for eight hours. The soil layer was organic potting mix from a brand called Hunar, sifted to remove the larger wood chips, then capped with a half-inch layer of fine black sand from a local aquarium shop in the Serangoon area. The plants were mostly fast-growing stems — Hygrophila polysperma, Limnophila sessiliflora, a few bunches of Vallisneria — plus a handful of frogbit ordered online from a seller in Johor Bahru.

The frogbit arrived in a small plastic bag with about an inch of water and a dozen leaves. It looked unremarkable, almost weedy. The seller, a man named Ng who runs a small aquatic plant nursery out of his backyard in Taman Mount Austin, had warned against too much light. The warning came through a WhatsApp voice note that crackled with static: the frogbit didn’t need much, and if you gave it too much, it would take everything. The warning was easy to ignore when the frogbit was just a few small rosettes floating in a newly filled tank. It looked like it needed the light.

The Problem with Measuring Light

There is a kind of cognitive dissonance that happens when an aquarium looks fine from across the room but fails under close inspection. By the second week, the Hygrophila was showing signs of something — not quite deficiency, not quite algae, but a certain tiredness in the lower leaves. The Limnophila was stretching toward the surface, internodes too long, leaves too far apart. Meanwhile, the frogbit had started sending out runners with an efficiency that seemed almost aggressive. Each mother plant produced three or four daughters within days, each daughter equally eager to reproduce. The surface went from open water to scattered patches to a solid mat in what felt like a single weekend.

The light fixture was a cheap brand — Nicrew, the kind sold on Shopee for around thirty Singapore dollars — and its intensity was listed in vague terms like “suitable for planted tanks.” There was no PAR reading, no spectral graph, no way to know whether the 20-micromole estimate was accurate or optimistic. The west-facing window added an unpredictable variable: two hours of direct afternoon sun on clear days, none when the sky was overcast, which was often during the northeast monsoon season. The result was a lighting environment that fluctuated wildly, and the frogbit loved it.

“People overthink light,” says Aditya Ramesh, who keeps a dozen Walstad tanks in his flat in the Lavender area and has been experimenting with the method for about four years. “They think they need a specific spectrum or a specific brand. The real variable is not the fixture — it’s the distance to the surface and the duration. A light that’s too close to a low water level will cook the floaters. A light that’s too far will starve the stems. And a window nearby changes everything.” Ramesh keeps his tanks away from windows entirely, relying solely on his LED fixtures, which he runs for six hours a day. His tanks grow dense carpets of Monte Carlo and Glossostigma. He does not have a frogbit problem.

The Floaters’ Advantage

Frogbit, like most floating plants, has a structural advantage over submerged vegetation. It sits directly at the air-water interface, where CO₂ is abundant and light is at its most intense. Submerged plants have to contend with the attenuation of light through water — every inch of depth reduces intensity by roughly half, depending on water clarity. Floaters don’t face that problem. They intercept light before it reaches anything below, and they grow in response to it. Give them a lot of light, and they grow a lot. The feedback loop is simple, and it’s brutal for the plants underneath.

By the third week, the Hygrophila had lost most of its lower leaves. The Limnophila was barely hanging on, its stems thin and pale. The Vallisneria had stopped sending out runners. The frogbit, by contrast, was thriving in a way that felt almost personal. Its roots hung down several inches into the water column, creating a kind of underwater curtain. A visitor to the apartment, upon seeing the tank for the first time, asked if the roots were a deliberate aesthetic choice. They were not.

What the Method Actually Requires

Walstad’s own writing is careful on the subject of light. In her book, she recommends low-to-moderate lighting — roughly 1 to 2 watts per gallon if using fluorescent bulbs, which is the equivalent of about 30 to 50 micromoles of PAR for a ten-gallon tank. She also advises keeping the photoperiod between eight and ten hours, with a break in the middle to prevent algae blooms and give CO₂ levels a chance to recover. What she doesn’t emphasize enough, for a beginner reading the book for the first time, is how differently floating plants respond to those same parameters.

The soil layer, too, plays into the issue. Organic potting soil releases nutrients slowly over months, but it releases them. The frogbit was the first to notice. Its leaves grew larger, darker green, more robust. It started producing flowers — small, white, three-petaled things that floated above the surface on short stems. In a different context, these flowers might have been charming. In a tank that was supposed to be a self-sustaining ecosystem, they were a sign that something had tipped too far in one direction.

The solution, when it came, was not elegant. It involved a net, a bucket, and about forty minutes of manual removal on a Thursday evening. The frogbit came out by the handful, root systems tangled together, the surface finally opening up to reveal the state of the plants below. The Limnophila was salvageable but barely — a few stems with decent leaves near the top, the rest a mess of bare nodes. The Hygrophila was a write-off. The Vallisneria had survived, somehow, its leaves long and thin and slightly translucent, as if it had been living in a cave.

Resetting the Balance

About two-thirds of the frogbit went into the bucket. The rest was left as a thin ring around the edges — enough to absorb excess nutrients, not enough to shade the entire water column. The light timer was cut to six hours, with a two-hour break in the middle of the day, following the pattern Ramesh had described. The window blinds were kept half-drawn during the afternoon, reducing the direct sunlight to a diffuse glow. For the first time in weeks, the tank looked like something intentional rather than something that had gotten away from its keeper.

The recovery was slow. The Limnophila grew new leaves from the surviving nodes within a week, though the stems remained leggy. The Hygrophila didn’t come back — the lower leaves had rotted and the stems had softened. It was replaced with a bunch of Hygrophila corymbosa, which has thicker leaves and a reputation for being harder to kill. The Vallisneria waited two weeks, then started sending out runners again. One of them traveled the length of the tank before sending up new leaves. The frogbit stayed where it was supposed to, growing just enough to need occasional thinning — a five-minute task every ten days, not a crisis.

Where the Method Met Its Limit

The assumption that a method, followed precisely, will produce predictable results — that was the problem. The Walstad Method works, but it works within a specific set of conditions that are not always easy to replicate on a first attempt. The soil needs to be the right type. The plants need to be the right species for the light available. The light itself needs to be measured, not guessed. And the floaters, if included at all, need to be managed from day one, not after they’ve already taken over.

Ng, the plant seller from Johor Bahru, was not surprised when he received a follow-up message a few weeks later. “Normally people don’t listen,” he said, in another voice note. “They think they know better. Then they have frogbit everywhere. Next time, you start with less light. Or no window. Or both.” He laughed, a short, dry sound. “Frogbit is like a relative who stays for a visit. If you are not careful, they never leave.”

The tank, as of late April, is stable. The Limnophila has recovered enough to be trimmed and replanted. The Monte Carlo that was added as a test — a small clump tucked into a gap in the sand — has started to spread, slowly, across the substrate. The frogbit still sits in a thin ring near the surface, doing its job, asking nothing more than occasional thinning. The water stays clear. The soil stays capped. The whole thing hums along, most days, without intervention.