On the morning of September 19, 2017, at about 11:20, a diver in Lake Jackson, Florida, fed a hose roughly 6.7 meters down to the throat of a lake-bottom sinkhole called Porter Hole Sink and, over about six and a half minutes, pushed 45.4 kilograms of 20 percent Rhodamine WT into the hole. Thirty-five days later, instruments 30.6 kilometers — about 19 miles — to the south registered that same dye rising out of the vent at Wakulla Spring. The trace, published in the Journal of Cave and Karst Studies by Seán E. McGlynn and Alan W. Niedoroda, is one of the longer confirmed conduit connections documented in the Upper Floridan aquifer, and its most interesting number may not be the distance at all. It may be how little of the pink dye ever came back out.
A lake with a drain in the bottom
Lake Jackson sits north of Tallahassee in a landscape that behaves less like a watershed and more like a plumbing diagram. The limestone underneath is riddled with dissolution features, and the lake periodically demonstrates the point by emptying itself. Porter Hole Sink is one of the openings responsible: a sinkhole in the lake bed that, when the plug of sediment over it loosens, accepts water directly into the aquifer beneath. There is no creek carrying that water away, no surface channel to follow with a gauge. The water simply leaves through the floor.
That is the practical problem karst hydrogeology exists to solve. In a normal drainage basin you can stand at a confluence and watch the answer go by. In carbonate terrain the flow paths are buried, irregular, and frequently counterintuitive — conduits can cross beneath surface divides, split, rejoin, and deliver water to springs that no topographic map would nominate. The only way to establish where a sink goes is to put something distinctive into it and wait somewhere downgradient with an instrument sensitive enough to notice.
Rhodamine WT is the usual candidate. It is a fluorescent tracer dye, intensely pink in the drum and effectively invisible once the aquifer has had its way with it, detectable by fluorometry at concentrations measured in micrograms per liter. It moves with the water rather than sticking to rock, which is what makes it useful, and it is used routinely enough that state agencies post public notices when a trace is underway so nobody panics at a tinted ditch. Injecting it into a lake-bottom sinkhole rather than a well is the harder version of the exercise: the dye enters at depth, in a lake, and whatever happens next happens in the dark.
Thirty-five days, and a sonde that noticed
Wakulla Spring is a first-magnitude spring — the highest discharge class — and the centerpiece of Edward Ball Wakulla Springs State Park, where the water arrives through one of the most extensively mapped underwater cave systems in the world. It is also the obvious place to listen for anything injected into the Tallahassee karst, and in 2017 it was instrumented for exactly that.
The signal arrived on October 24, 2017, roughly 35 days after injection: a coherent dye peak at Wakulla Spring recorded by sondes rather than by eye. Nothing about the spring looked different to a visitor. The dye had been diluted through tens of kilometers of limestone into the microgram-per-liter range, which is a concentration a fluorometer can resolve and a human retina cannot. This is worth stating plainly because the alternative image is more fun and entirely wrong: there was no pink spring, no tinted glass-bottom boat ride. There was a curve on a chart, and the curve had a shape consistent with a single slug of tracer that had been traveling for a month.
Sally Ward Spring, nearby, produced its own coherent peak at about the same 35-day mark, with the first faint pulses showing up around day 31. Charcoal packs set at other springs in the area picked up dye as well, which points to a conduit network that braids and branches rather than running as a single pipe from sink to spring. Averaged across the interval, the plume moved on the order of 0.8 to 0.87 kilometers per day — brisk for groundwater, glacial for anything with a current.
The dye that never came back
The finding the authors treat most carefully is the accounting. Very little of the 45.4 kilograms injected at Porter Hole Sink was recovered at Wakulla Spring. The overwhelming majority of the tracer appears to have dispersed into the limestone matrix — the fine porosity and small fractures flanking the conduits — rather than staying in the fast-flow channel all the way to the vent. The paper is explicit that this mass balance is exploratory and to a degree conjectural; you cannot instrument every outlet in a karst system, and dye that goes into the matrix may resurface slowly, elsewhere, or in concentrations below the detection floor.
But the asymmetry matters for how the result gets read. A confirmed connection between a sinkhole and a spring is not a statement that the sinkhole feeds the spring in any bulk sense, and nothing here supports the idea that Lake Jackson’s water is Wakulla’s water. What the trace establishes is that a hydraulic path exists, that it is continuous enough to carry a discrete pulse across the distance, and that whatever enters at the top can arrive at the bottom on a timescale of weeks. For nutrients, spills, and stormwater in a rapidly developing part of North Florida, that is the operationally useful fact — the aquifer offers less filtration and less delay than intuition suggests, even while most of what enters gets absorbed into the rock along the way.
Which leaves the number the experiment was designed to produce: dye into Porter Hole Sink on September 19, dye at Wakulla Spring 35 days later, 30.6 kilometers south.