When you picture a nuclear plant, you probably see a pair of giant concrete cooling towers with steam rolling out of the top. St. Lucie, a two-reactor plant on Florida’s Hutchinson Island, doesn’t go that route. It cools itself with the Atlantic.<\/p>\n
That comes with a slightly odd side job. The seawater rides in through pipes that open on a nearshore reef, right next to where green and loggerhead turtles find their food. Some of those turtles get pulled in with it, and somebody has to fish them back out. Biologists have been doing exactly that since 1977, and they’ve been writing the captures down the whole time.<\/p>\n
That logbook’s now been turned into a study. Ryan Welsh, of Inwater Research Group in Jensen Beach and the University of Central Florida, and UCF sea turtle researcher Katherine Mansfield ran 19,987 captures through a statistical model and published the results in Scientific Reports<\/a> on March 16. Six weeks later, the NRC issued new licenses that let both reactors run for 80 years, so the canal isn’t going anywhere for a while.<\/p>\n It starts with heat. The NRC licenses each St. Lucie reactor at 3,020 megawatts, and those are thermal megawatts, the raw heat coming off the core. A big water-cooled reactor turns only around a third of that heat into electricity, and the rest has to go somewhere, which is why you’ll usually find plants like this one sitting next to a lot of water.<\/p>\n St. Lucie’s seawater comes in through three concrete velocity caps that sit about 1,500 feet offshore, according to an NRC environmental assessment<\/a>. A velocity cap is basically a lid over the mouth of an intake pipe, so the water gets drawn in sideways instead of straight down. Fish tend to feel that sideways pull and swim away from it. Sea turtles, from the looks of things, are a lot less picky.<\/p>\n From the caps, three buried pipes carry the water to two headwalls, where it pours into the plant’s intake canal. That canal’s about 4,900 feet long, so nearly a mile, and the plant moves all that water with 12 independently operated pumps. A turtle that rides the pipes in can swim around the canal and come up to breathe. It can’t get back out to sea on its own, though.<\/p>\n Biologists from Inwater Research Group, mostly with nets. The paper lists dip nets, tangle nets and plain old hand capture. Each turtle gets its shell measured and gets tagged twice, with metal flipper tags and a PIT tag, which is basically the same kind of microchip your vet puts in a dog. Then it’s released offshore. On average, a turtle spends about a week in the canal between getting pulled in and getting caught.<\/p>\n Between 1977 and 2023, that work added up to 19,987 captures of green and loggerhead turtles. Greens accounted for 8,631 of them and loggerheads for 11,356. Greens ranged from about 9 to 44 inches and loggerheads from about 15 to 43 inches, measured in a straight line along the top shell, front to back, so the head and flippers don’t count.<\/p>\n So is that 19,987 different turtles?<\/p>\n Not quite. It’s a count of captures, and some turtles get caught more than once. We’ll come back to those repeat visitors, because they matter. I couldn’t find a count of individual animals in the published paper, and the raw records belong to FPL and aren’t public, so I can’t work it out myself.<\/p>\n The 19,987 figure only covers greens and loggerheads, too. Hawksbills and Kemp’s ridleys have turned up in the canal over the years, and FPL once reported pulling out an olive ridley, a species the company said doesn’t usually show up in the North Atlantic.<\/p>\n Sea turtle biologists have traditionally sorted turtles into life stages mostly by size, from small juveniles up to adults, and a lot of conservation planning leans on those stages. Welsh and Mansfield wanted to see whether the St. Lucie record backed that up.<\/p>\n They used what’s called a hidden Markov model. In plain terms, it assumes each capture belongs to one of a few hidden modes you can’t observe directly, then works out the most likely mode from things you can measure. Here, that meant shell length, the time of year (tracked through hours of daylight), the year itself and how many intake pumps were running. The model didn’t get to see the recapture records. The best fit for both species came out at three modes, and the researchers only gave them names after the model was done sorting: Nomadic, Resident and Transient.<\/p>\n Nomads were generally the smallest turtles, turned up mostly in winter and early spring, and rarely got caught twice. Residents sat in the middle of the size range, stuck around for much of the year and kept getting recaptured. Transients were the big ones, and they showed up seasonally in summer, which the authors read as migrating juveniles or breeding adults passing through.<\/p>\n Here’s where it gets interesting. Once the model had sorted everything, the researchers checked it against the recapture records it never saw. The turtles it had flagged as Residents really were the ones that kept coming back, and the one-and-done captures leaned Nomadic or Transient. That’s a pretty good sign the model was tracking how turtles actually use the spot.<\/p>\n Green turtles mostly did what the textbooks say they’d do. Their three modes lined up with fairly distinct size bands, even if those bands overlapped. Loggerheads weren’t nearly as tidy. For them, shell length was only a weak predictor of which mode a turtle was in, so two loggerheads of the same size could be living very different lives off that stretch of coast, one sticking around and one just passing through.<\/p>\n I’d argue that’s the most useful thing in the paper. Protections for sea turtles often hinge on assumptions about when and where a turtle of a certain size spends its time, and the authors’ point is that size alone won’t tell you that, especially for loggerheads. They’re also careful to describe these modes as patterns a turtle can drift in and out of as it grows.<\/p>\n Why are loggerheads so flexible? The authors float diet as one possible reason, since greens graze on plants and algae that stay put while loggerheads go after prey that moves around. They’re upfront that they didn’t test it, though.<\/p>\n US power reactors start out with a 40-year license, can renew once to reach 60 years and can go back for a second renewal that takes them to 80. That’s the round St. Lucie got through this spring, and it’s why the intake pipes and the canal behind them should be in service for decades.<\/p>\n Florida’s other nuclear plant, Turkey Point, has its own canal wildlife story, with American crocodiles nesting on the banks of its cooling canals<\/a>. And coastal intakes catch more than turtles. In France, operators at Gravelines have shut reactors down by hand before jellyfish could pack the seawater screens<\/a>.<\/p>\nSo how does a sea turtle end up in a nuclear plant’s canal?<\/h2>\n
So who fishes them out?<\/h2>\n
The model sorted the captures three ways<\/h2>\n
Loggerheads are the messy ones<\/h2>\n
St. Lucie is licensed into the 2060s<\/h2>\n