Inside a kelp holdfast census at Tinderbox: bagging, sieving and counting the hundreds of small animals that tell us whether a restored reef is alive.
Field Notes are Tidewright's own dispatches from our restoration sites. They describe our fictional projects, crews and results — told as we live them. Our restoration outcomes are illustrative and reported under our field monitoring protocols.
The last Tuesday of the census week at the Tinderbox site, I spent three hours bent over a sorting tray the size of a baking sheet, extracting small living things from a tangle of kelp root-mass the size of a cabbage. Final tally for that one holdfast: 217 animals, 31 kinds. An amphipod every few seconds. A brittle star the size of a thumbnail. A worm I had to photograph and send to a taxonomist friend because being a field ecologist for twenty years does not, it turns out, cover everything a kelp root hosts. This is the field note for the holdfast census — the count we run each year at Tinderbox because the smallest residents of a reef give the most honest report on whether a restoration is alive or merely planted.
First, the biology. Kelp have no roots. What looks like one — the gnarled brown mass gripping the rock — is the holdfast: a bouquet of branching, finger-shaped haptera whose only job is mechanical attachment. The full story of how that structure grips surf-battered rock is told in holdfast biology; here the point is the architecture's side effect. Between the interwoven fingers of a mature holdfast runs a labyrinth of gaps and channels, permanently damp, sheltered from surge and light, flushed with drifting food — in ecological terms, prime real estate. Animals move in. Isopods and amphipods colonise the outer fingers; worms, tiny crabs, snails and brittle stars work deeper into the maze; a full-size holdfast on a healthy reef is less an organ than a neighbourhood, hundreds of individuals and dozens of species living in the floorspace of a dinner plate.
That crowding is precisely why ecologists love the holdfast as a sampling unit. A kelp plant's holdfast is a small, countable, self-contained slice of the reef's invertebrate community — lift it, and you take a census of the building's tenants all at once.
The destructive part comes first and we keep it tiny. Most of our Tinderbox monitoring is strictly non-destructive — photo transects, timed counts, the long, patient record described in five years, one transect. The holdfast census is the exception: a small number of plants, taken from designated sacrificial plots at each site depth, because no camera yet reads the inside of a holdfast.
The dive protocol exists because a lifted holdfast is an alarm bell. Bag the plant before it leaves the rock — slide a fine mesh bag over the holdfast, razor it free, seal. Anything bolted for open water goes into the bag instead of the blue. On the boat, each holdfast gets shaken hard in a tub of seawater, then sieved through a stack of graded meshes; two rinses, because amphipods hold grudges and grip. The catch lands in white sorting trays, and the hours of the real work begin — identify to the lowest taxon we can honestly name, count, weigh wet, record. Subsplits go into ethanol for the lab and, newer every year, a filtered water sample goes for eDNA sequencing: the genetic shadow of the community, read alongside the physical count, because the two methods catch different tenants and their disagreement is itself informative.
Now the honest part, with the standing caveat that all Tidewright project figures are illustrative and reported under our monitoring protocols. At the Tinderbox pilot, year-one holdfasts on our seeded lines averaged about a dozen animal taxa apiece — the early colonisers, mostly small crustaceans that arrive with the spores' first recruits. This year's count, year five, averaged 47 taxa per holdfast on restored plants, with counts of individuals up roughly fivefold from year one. The reference reefs nearby — old, untouched forest — run richer still, which is exactly right: the census is not a victory lap, it's a ruler. Watching the restored holdfasts close the gap on the reference ones is the signal we actually want, and it moves at the speed of small animals finding their way home, one drifted larva at a time.
The pattern inside the numbers matters more than the totals. Filter-feeders and detritivores showed up within two years; the pickier residents — the species that eat things that only exist on an established reef — took four. A community assembles in a stricter order than a forest does, and the census is how we read that order off the trays.
A reasonable outsider question: why does a restoration company count amphipods at all? Three answers, in increasing order of ambition. First, diagnosis: a restored forest whose holdfasts stay empty is a plantation, and the census catches it early enough to change what we do. Second, accountability: the biodiversity metrics that feed our monitoring and credit systems are built from counts like these, and the whole chain of custody from sieve to statement is written down — our note on issuing the first biodiversity credit describes why the chain matters. Third, and my own reason for the three-hour trays: a holdfast census is the reef's own answer to the only question that counts. Not did the kelp grow — a drone can tell you that — but did the neighbourhood come back.
At census scale, no. We sample a handful of plants per site per year from designated sacrificial plots; the reef doesn't register it, and the rest of our monitoring is non-destructive. The diagnostic value of reading the holdfasts honestly far outweighs the cost of a cabbage-sized tangle of root-mass.
It's discrete, standardised and rich. Every mature plant has exactly one, roughly comparable in structure, and the community inside reflects years of local conditions — which makes holdfasts unusually good at showing whether a reef community is assembling or not.
Mostly small crustaceans — amphipods and isopods — plus polychaete worms, small crabs, snails, brittle stars and the occasional juvenile fish sheltering deep in the maze. Mature holdfasts on healthy reefs host hundreds of individuals and dozens of species each.
Each method sees different tenants: the sieve catches what stays put, the water sample reads genetic traces of whatever passed through. Where the two agree, the count is solid; where they disagree, we've learned something — usually about who's living in the reef but eluding the tray.
Tr L-7 is a single 30-metre transect off Tasmania surveyed every month since 2021. Five years of unglamorous re-measurement turned out to be our sharpest instrument.
Jack Callum ·
The first biodiversity credit we ever issued took eleven months, three verifiers and one metric we threw away. Field notes from inside the process.
Priya Raman ·
Kelp has no true roots — its holdfast is a branching biological anchor built for storms. Explore holdfast anatomy, how kelp really feeds, and its hidden biodiversity.
Dr. Theo Nakamura ·