The 27-metre Sanford trawler Tengawai steams away from Auckland and out through the Coville Channel, the narrow stretch between Aotea/Great Barrier and the tip of the Coromandel. On this early September day, the ocean is the calmest it’s been all winter, and beyond the gap, the sea is a silken pond, with the islands—Aotea, Cuvier, Mercury, Channel, Hauturu—strung like buoys on the horizon’s rope.
In the wheelhouse, skipper Bert Aitken is scanning his instruments. He’s looking for bottom-feeding snapper on the sounder, choosing his line. Tengawai has two crews who alternate every two weeks, and it’s Aitken’s first day back at work—just another trip in a 40-year fishing career. The deckhands, Mike Jones and Matangaro Ben, are out preparing the trawl gear in their white gumboots and fluoro hard hats. Like Aitken, they’re share-fishermen, only getting paid for what they catch.
The cod-end lands in the water first—a beige, multimillion-dollar, trade-secret composite-material sack dubbed the Precision Seafood Harvester (PSH). Part of it, the “selectivity section”, has small rectangular holes, and it floats at the surface while the rest of the net—made of traditional mesh—unspools with a hydraulic hiss. Next come the bridles and the sweeps: green ropes as thick as my wrist, designed to keep the net open and to scare the fish into its gaping mouth. Finally, Jones and Ben release the trawl doors, which drag the net to the bottom and stretch it wide. They sink out of sight with a dull clank.
Three hours later, we’re halfway up the eastern side of Aotea and still towing. Long ocean swells lift the boat and crash on the island’s distant beaches, the spray drifting like yellow smoke. The sea is oily and impenetrable, pale blue and gold over black. Beneath me, and a way behind, the weighted net and those heavy metal doors scrape the seabed. I picture the fish inside, herded together, still swimming.
The sun has dipped behind the cobalt bulk of Aotea when the crew start reeling in. First the wires, then the doors, the sweeps, the net. A dozen mollymawks are wheeling with interest. The dangling orange bird-scarers seem to keep them away from the wires, though one steals a fish from the rising net. Tails and fins and eyes peek out from some of the little windows in the selectivity section as it’s wound onto the net roller.
Finally the fat, featureless bag of the cod-end rises out of the sea. Ben positions it over three stainless-steel baths and tugs on a rope to release the opening. A foaming tonne of sea life and water tumbles out, drenching him and leaving many hundreds of fish leaping and squirming in the troughs.
Pulling on my own hard hat, lifejacket and oversized gumboots, I clamber down for a closer look. Jones and Ben sort through the catch as fast as they can. There are heaps of snapper, jack mackerel with iridescent pale bellies, maomao in Disney-princess pink. A long, thin frost fish, unnaturally shiny, like polished silver. Tarakihi, trevally, a few big kingfish.
They’re not all alive. But most of them are. “That’s the thing about this cod-end, they look like they’ve been caught on the rod,” Jones says—“though it’s not as easy to work with!” he adds as a snapper flaps in his bearded face. Traditional cod-ends, he explains, jam the fish together. They get tired, the lactic acid in their bodies soars sky-high, and they’re generally dead by the time they’re brought on deck. Theoretically, the design of this new net allows undersized fish to escape at depth, and the rest to keep on swimming until the last possible moment. Fillets from fish caught this way are better quality, due to the lack of stress—firmer in texture, and translucent.
The men measure the fish against marks on the aluminium. Most fish go down the chute into the ice-filled hold. Too small, and they biff them back into the ocean; not all of the little snapper evaded the net. Some scarper, while others swim drunkenly and are snapped up by the waiting mollymawks. Bycatch species like skates and starfish go over the side too. I lose count of how many puffed-up, grumpy-faced porcupine fish the men toss into the sunset like spiky rugby balls.
But the deckhands are keeping a mental tally. When they return to the wheelhouse, they enter the catch on the government forms. There are 25 bins of snapper—about 575 kilograms. Three bins of trevally. Some other mixed fish. And an estimated 175 individual undersized and bycatch creatures chucked over the side.
“Not bad for three hours,” says Aitken. He points to the map on one of the computer screens. A tangled rainbow of overlapping lines marks the trawls this same boat has towed over the past few years. There must be a hundred of them, at least. In total, he says, “that tow has probably been done thousands of times. How can we have wrecked everything, if we’re still catching fish?”
Globally, a quarter of all fish and shellfish landed are caught by bottom trawling or dredging (dragging a small net with a toothed bar along the seabed). In New Zealand, these methods are even more prevalent. In the 2020-2021 fishing year, more than two-thirds of the total catch by volume came from within one metre of the sea bed and was hoovered up by bottom trawlers.
For many inshore species, like trevally, tarakihi, gurnard and scallops, bottom trawling and dredging are by far the cheapest and most efficient options. They’re the only way to catch economically important deepwater fish such as orange roughy and hoki (except during the three months of the year when hoki spawn in midwater).
But bottom trawling takes more than fish.
It’s not completely indiscriminate. Skippers can tweak what they catch by choosing where to tow and changing aspects of the gear setup, such as the height of the net’s opening, the length of the ropes, or the width of the doors. But every tow catches animals the fishers can’t sell or would rather not catch. (Once, Jones tells me, they even caught a boat—a sunken aluminium dinghy hauled up out of the murk.)
In some fisheries, unwanted species are harder to avoid. For every kilo of scampi caught by trawling, 3.8 kilograms of bycatch is dragged aboard as well. In contrast, fishing for southern blue whiting, oreo, and jack mackerel is much more targeted—for each kilo caught, there are just 10 grams of bycatch.
Sonar devices now fitted to the nets scare off whales and dolphins, most of the time; in 2019-2020, fisheries observers—present on around one in five New Zealand trawlers—recorded just two dolphins accidentally caught by trawling. (However, when observers are on board, they consistently record more bycatch than is self-reported by fishers.)
Bottom trawling and dredging also harm sensitive seafloor ecosystems, especially the first time. “It’s a case of, the first cut is the deepest,” says Simon Thrush, a marine ecologist who spent three decades studying the impact of fishing on the seafloor and now heads the University of Auckland’s Institute of Marine Research.
Most vulnerable are areas populated by living structures such as corals, sponges, or shellfish, which provide crucial sanctuaries for young fish to hide from predators and rest out of the current. Between 2018 and 2021, New Zealand commercial deepwater fishers reported catching 275 tonnes of sponges—slow-growing, water-filtering animals which can live for hundreds, even thousands of years.
Offshore, these habitats are often found on seamounts. NIWA scientists using underwater cameras found that as few as 10 deepwater trawls on a small seamount can reduce coral cover from 20 per cent to zero.
And the harm is long-lasting. Dredging in the Firth of Thames in the mid-20th century completely destroyed its rich mussel beds, and they have never recovered. Offshore, on the Chatham Rise (between the South Island and the Chatham Islands), two decades of studies have shown that 20 years after intensive trawling stopped on the gothically named seamount Morgue, coral ecosystems are only just beginning to recover.Trawling damage isn’t always visible, says Thrush. “We can drastically change the nature of seafloor ecology in any habitat type, from gravel through to fine mud.”
International studies—it seems no New Zealand ones have yet been funded—have shown that trawling releases chemicals from sediments, altering the nitrogen cycle, the carbon cycle, and other global processes in ways that scientists are still trying to understand.
But all forms of food production have some kind of environmental impact. And the annual trawl footprint is reducing, with just two to three per cent of our nation’s exclusive economic zone (EEZ) trawled each year (compared with the 40 per cent of our land used for farming.) Still, we don’t trawl the exact same areas each year, and because the impacts are long-lasting, the effects can add up. Research by Fisheries New Zealand indicates that at least one-third of the country’s fishable area—unprotected waters shallower than 1600 metres—was contacted by bottom trawling between 1989 and 2019 (and even more was probably trawled before 1989, when data collection was poor).
In 1367, fishermen on the Thames estuary petitioned England’s King Edward III to ban the sail-powered bottom trawling they feared was catching too many small fish and causing “great damage of the commons’ realm and the destruction of the fisheries”.
Now, in New Zealand, government, industry, and conservation organisations are again debating the future of bottom trawling and dredging, with several consultation processes underway. The big questions: Are existing protections enough to allow both conservation and sustainable utilisation? Should bottom trawling and dredging be further restricted to certain areas? Can gear innovations solve some of these intractable problems? Or should the methods be phased out entirely?
In December 2007, scientists from Crop and Food Research (now Plant & Food Research) were in the Cook Strait, trying to solve a hoki-quality problem. They dropped a handline over the side of the research vessel, hundreds of metres down to where the hoki were, and pulled one up onto the deck.
Unlike traditional trawl-caught hoki, which come up bruised and battered with pink flesh, this one was in perfect condition. The fillets cut from its body were clear and translucent. That was the “aha moment” that led to the Precision Seafood Harvester, says Greg Johansson, who at the time was Sanford’s deepwater fleet manager but is now an independent consultant and the interim chair of the PSH programme.
If you could replicate that kind of quality with a trawl net, you could turn a low-value product into a high-value one, maximise the economic return from each tow and each fish—and hopefully pick up other benefits, such as reducing the mortality of unintended catch and persuading consumers that trawling is a sustainable option.
By 2012, the plan was underway. Three seafood companies—Sanford, Sealord Group, and Moana New Zealand—committed $26 million to the project in cash and vessel time, and the government matched it.
The regulators had conditions. The Fisheries Act strictly prescribes what equipment can be used for trawling. Fisheries New Zealand, tasked with drawing up a bespoke approval framework, decided the new gear would need to be “no worse than” traditional trawlers when it came to sustainability. “We thought that would be a no-brainer,” says Johansson.
Reality was more complicated. “No worse”—in what way? The new criteria included impact on protected species and the seafloor, the number of undersized fish caught, the mortality of fish that are landed and discarded, and the amount of target fish caught per tow, affecting how much of the seabed must be contacted for a given catch.
But the ocean is a tricky place to do research, fish behave in complex and species-specific ways, and figuring out how to test the PSH and traditional trawl methods against these criteria proved difficult. In one study, for instance, the new net did let more undersized tarakihi swim free—but it caught fewer of the target species, barracouta.
Snapper posed a particular paradox. The PSH was designed to keep water flowing inside the net at a speed the fish can keep up with, so they stay relaxed and energetic over the hours the net is dragged along. But because of the way snapper school, juveniles tended to swim happily in the centre of the net, surrounded by larger fish and ignoring the escape holes designed to release them.
Testing different prototypes, the PSH team were able to change how fast the water moved in the net, fatiguing the smaller snapper and encouraging more of them to swim out the holes. But those that did stay inside the net got so tired they were more likely to die after capture. Both these measures also proved difficult to quantify.
The final design, approved by the government and now in use by Tengawai and other inshore vessels, was a compromise. It’s thought to land slightly more undersized snapper than conventional trawl nets, but with slightly higher catch rates, and much lower mortality of unwanted fish, at least for the 48 hours after they are returned to the sea.
It’s certainly improved the condition the fish are landed in, increasing commercial returns by 30 per cent for both snapper and hoki. I asked Tengawai skipper Aitken if he thought the new net had reduced the catch of juvenile fish. “Maybe,” he said. “Depends where you tow.”
Johansson maintains the current iteration of the PSH is just the start—a game-changing proof of concept with far more potential. He’d hoped to trial various innovative ideas, such as altering the hole shapes and positions, or using different colours and sounds to scare certain species out of the net. None were progressed because the team were tied up with meeting the government criteria necessary for approval, he says.
“I’d love to be able to say we’ve cracked it, and I can get all the left-handed ones to go out the side and all the spotted ones to go out the top, we haven’t quite got there yet. But the fact that you’ve got them in a controlled environment where they’re viable and they’re swimming and they’re not under stress… that’s gold. It opens the door to all sorts of opportunities.”
The PSH programme always had a deadline of 2019. By then, 22 per cent of the three companies’ hoki quota was being caught with the new gear, and trials had begun in a flatfish fishery in the Netherlands. In September 2022, on the advice of independent consultants, the government released an extra $9.4 million that the team had been allocated, but hadn’t spent. Now they can finish the job.
Over the next three years, with that extra cash in hand, Johansson’s team will test more designs. One early-stage project involves a sensor on the net they hope will be able to detect unusual, large, or warm-blooded animals. “Heaven forbid you had a Māui or Hector’s dolphin in there. You could open the back end and everything could swim away,” Johansson says.
Still, for all its sort-of precision and future potential, a huge amount of public money was spent on a system that does little to address trawling’s benthic impacts—the effects on the seafloor. Raewyn Peart from the Environmental Defence Society points out: “They completely ignored the biggest environmental threat—bottom contact.”
Offshore, the undersea landscape is just as topographically varied as the mainland. Trenches and abysses fall away into the deep, while rocky hills and volcanoes—seamounts—rise from the silty ocean floor. These features generate currents that keep the seawater clean, says NIWA’s Malcolm Clark, who has been studying the impact of fisheries on “seamounty things” for 25 years.
Importantly, they also provide solid structure for plants and animals to settle on while lifting them within reach of migrating plankton, he says. “Fish don’t have to charge up into midwater to chase their prey, they’ve basically got the food coming to them.” Together, that means these underwater hills and mountains can sometimes give rise to whole ecosystems. They also make many seamounts attractive places for fish—and for fishing.
In the past few decades, researchers have begun to explore these places and the life they support. It’s becoming clear that just like mountains on land, the ecosystems on these pinnacles vary dramatically, even from slope to slope on a single seamount. Ashley Rowden, a fisheries scientist who works closely with Clark at NIWA, says some seamounts might play a significant ecological role, others less so—and we don’t yet have the right tools to be totally sure.
Furthermore, what counts as a seamount at all is disputed. “It might sound a really trivial issue, but it’s hugely important when it comes to politics,” says Clark.
Seamounts were first studied by geologists, who defined so-called underwater topographic features above 1000 metres as seamounts. Those between 250 and 999 metres were dubbed knolls, and between 100 and 249 metres, hills. The fishing industry prefers to restrict discussions of seamounts to the kilometre-high behemoths, of which, at last count, there are just 146 within New Zealand waters.
“That definition is just really a hangover from the 1960s,” explains Clark. “It’s arbitrary. It’s not driven by any biological reasons.” If hills and knolls are included—and they should be, argues Clark—there are almost 2000 seamounts. Ecological importance is not determined by height, he says, but by a jumble of factors including depth from the surface, shape, the type of rock they’re made of, and how far they are from other seamounts or the shore.
Some of the most spectacular undersea forests of sponges and corals Clark has observed via NIWA’s underwater camera surveys have been on small features such as Ghoul, which rises just 100 metres or so above the seafloor. He sends me an image taken there—a tangled forest of bristly, branching, white-and-mustard coral, with a delicate orange sea star perched on top. “They’re really quite special spots, even though they’re not classed, by some people’s definitions, as seamounts—and therefore are outside some of the proposed management regimes.”
For instance, in August 2022, Sealord released a white paper suggesting that bottom trawling be forever limited to 15 of the huge, 1000-metre-plus seamounts that have already been towed upon. The gesture, the company admits, will “have a negligible effect on the operations of Sealord”. Arguably, it’s also unlikely to have much effect on conservation.
The Deep Sea Conservation Coalition—a group of New Zealand environmental non-governmental organisations (NGOs)—is calling for bottom trawling to be banned on all seamounts above 100 metres, and cites surveys showing this is popular with the public. The tide is turning internationally, too: Chile closed all its seamounts to bottom trawling in 2015, and Palau has banned it entirely.
“When what you’re doing is fundamentally altering the habitat, our view has always been you should take a precautionary approach,” says Geoff Keey from Forest and Bird. “If you do something that triggers a trophic cascade, you could find you have a massive problem that will take hundreds of years to put right.”
Industry is, predictably, pushing back, pointing out just how rich these fishing grounds are. Aaron Irving from the Deepwater Group—an industry body—says that 27 per cent of the total oreo and orange roughy catch is caught on these features, almost entirely on the smaller, hill-sized seamounts. Closing all of them, he says, “could have a profound effect on the viability of the seafood industry”.
Orange roughy, oreos, and especially cardinalfish like to feed on and above seamounts, but when avoiding fishing nets they dive down, meaning bottom-contacting gear is necessary to catch them, according to Sanford’s head of fishing, Colin Williams. “That’s just where the fish are.” (Several NIWA scientists confirm this, saying when they tried to catch orange roughy in midwater, catch rates were unviably low, even when the fish formed dense plumes. Roughy are especially good at shimmying out from beneath the net. “Bottom trawling is a necessary method for some species!” says Clark.)
Still, fishers don’t like trawling on rocky ground or spiky corals, Williams says—it wrecks the gear—so they tend to return to the patches they’ve already trawled smooth, again and again. Fish still seem to hang out in these places, he says. “It’s in our interest to be accurate, because if you fall off the edge of the track, then you’re going to risk damaging expensive gear.”
If not all seamounts are rich, pristine environments, could we be smarter about which ones we fish on?
Some hills are already protected. In 2001, the government closed to fishing 19 individual seamounts from different areas and depths—though the NIWA scientists involved admitted the selection was something of a “stab in the dark” as so little was then known about their biodiversity.
In 2007, the government accepted a fishing industry proposal to establish 17 Benthic Protection Areas totalling 1.2 million square kilometres, which are now closed to bottom-contacting gear. NGOs point out that nearly three-quarters of this territory is far deeper than trawlers can reach anyway, and much of the rest is away from the main fishing hotspots. In exchange for the deal, industry decreased its contribution to benthic research, and the government agreed not to progress further marine protection in the EEZ until 2013. No new areas have been protected there since.
But our data’s improving all the time. We’re learning about seamounts’ ecological value and recovery rates; meanwhile new reporting requirements mean we also have more detail about where trawling is occurring. So are we protecting the right places? Clark, Rowden and other NIWA scientists have just developed a new database which should help fisheries managers to answer this question. Ultimately, Clark hopes it will bring together multiple datasets pertaining to seamounts—things like underwater video footage, NIWA’s trawl survey data, fisheries information, plant and animal specimens in museum collections, and sea surface temperature records.
“It’s not just a case of, there are a number of features protected, so therefore the job’s done,” says Clark. “This dataset enables people to ask, ‘Are these features valid ones to protect, based on what we see about how representative they are?’”
In 2003, Clark and NIWA colleague Richard O’Driscoll examined where bottom trawling for orange roughy was occurring, finding that tows were increasingly happening close to seamounts, and that 80 per cent of seamounts with peaks at the depths where orange roughy live had already been fished.
The new database will make it easier to update this finding, Clark says—and it’s an important question to ask. The industry might be trawling only two per cent of the seafloor each year, but if that two per cent includes a high proportion of these specific, fishy habitats, we might want to protect more of them.
“Society should be interested in the impacts that it has on the natural environment, and that should include the ocean, and it should include the bits that people don’t think about or see,” says Rowden.
The sky above Napier is leaden, the sea a pale lichen green. Logs float in the restless water after the week’s rain, and snow clings to the tops of the inland ranges. Independent fisherman Karl Warr—a giant of a man in a woollen jumper and plaid shirt in size 7XL, with intense blue eyes and a missing tooth—is taking me out on his one-man day-boat Chips to show me how trawling could be done better. “I’m not doing this for you,” he assures me. “I’m doing it for the fish.”
Motivational quotes adorn the walls of the comfortable cabin, inscribed by Warr in permanent marker years ago. Fight for the things you believe in, but do it in a way that will lead others to join you—Ruth Ginsburg. Warr is a fighter, a philosopher, an innovator, and a contradiction—a trawlerman who hates trawling.
We head out into Hawke Bay, and I hang on to the side by the cabin door while Warr prepares the gear. (The components are similar to Tengawai’s but everything is on a much smaller scale.) Warr’s invention is a new kind of rigid cod-end, a cage with a solid bottom and back that slows down the water flow, like the PSH does.
The other panels are comprised of stainless-steel rods in a zigzag structure with a mixture of horizontal and vertical holes for different-shaped fish—and can be swapped out for other panels depending on what he wants to catch. “The zigzag is to increase the number of holes for the surface area,” Warr explains, and so the fish don’t have to turn 90 degrees to swim out. The metal is smooth and shiny, to avoid snagging their scales as they try to escape.
We tow for an hour or so—not as long as normal—then Warr puts the boat into neutral and goes out on the deck to bring up the gear. When the cage comes up, it’s full of fishy life. Gurnard with blue-fringed peacock wings, a John Dory, diamond-shaped sand flounder.
Warr begins the triage. “My first priority is, what can I save?” he says, unhooking a tiny spiny dogfish from the netting and dropping it over the side.
“Then all you other fullas who are freaking out, to get you into the ice and water as fast as I can. My attitude is, someone’s going to come out here and kill you guys, and I’m glad it’s me showing you a bit of mercy and compassion where I can.”
One small sole and one almost-legal flounder didn’t escape, but everything else is good-sized. When Warr fishes without the cage, he typically catches one undersized fish for every large one. Even if they come up alive, many won’t survive the experience—skin damage from the net or handling makes them extremely vulnerable to infection, he says. “We’re killing things that we don’t need to be killing, just for the sake of adapting the cod-end to have apertures that are smooth and hold their shape properly. This is low-hanging-fruit stuff.”
NIWA fisheries scientist Emma Jones tested Warr’s cage against a traditional 100-millimetre-mesh cod-end, comparing the catches, and says the difference was clear. “He’s retaining larger flatfish and gurnard above 30-35 centimetres, and releasing all the other small fish really well.”
Warr’s rigid cod-end would be illegal, if he didn’t have a special permit from Fisheries New Zealand to operate it. After years of trying to get funding, in 2019 Jones got a grant of $1 million to trial a smarter, more ambitious version of Warr’s cage.
Escape holes for small fish is one thing, but what about larger bycatch species? What if you catch a whole lot of snapper, but don’t have snapper quota? Jones, Warr, and an engineer began experimenting with the idea of a real-time underwater video feed and a smart, angled gate at the front of the cage—allowing a skipper to select which fish to retain and which to allow to swim away, and scientists to count everything that shows up.
“It sounds fairly simple, but it’s not,” says Jones. “Putting cameras underwater in a salty, rugged environment—it didn’t surprise me that we hit a lot of hurdles. Things got flooded and submerged and broke.”
Finally, in 2021, they had it working. Warr could watch the video link, see a fish he wanted, and trigger the gate to trap it in the cod-end. “‘I want that one, quick!’” It’s extra work for the skipper, he acknowledges—especially for a one-man-band like himself. “It’s not going to be as easy as just smashing over the whole environment out there, pulling out of it what you want, and brooming the rest over the side—and hence we stick with that.”
The smart cage is still in the early stages of research. Jones and Warr are working on embedding artificial intelligence in the system, so that the camera can recognise desirable fish and automatically retain those, releasing everything else. But to progress it further, they need more money. Compared with the US, and Jones’ native UK, funding for grassroots innovation in fisheries in New Zealand is hard to come by.
“I see that as a real gap,” Jones says. “There isn’t a funding pot that Karl could really go to and say, ‘I’ve got this idea, and I want to test it.’” Science funding is one part of the equation, but legislation is another, adds Geoff Keey from Forest and Bird. “Humans are really good at innovating, but we need a constraint to make that happen. When it comes to the environment, that usually has to be a rule and a deadline—and once you do that people will go and spend the money to do the innovation required.”
Progress is certainly being made on the bycatch problem. But as for the benthic impacts, they’re harder to solve. Trawl doors can be made lighter, Jones says, and overseas, people have experimented with flying most of the gear just off the seafloor, while dropped chains stir up the fish.
“But you can’t really bottom-trawl without being anywhere near the bottom. One day, maybe we’ll invent some way to catch fish without touching the seabed and without causing any damage or catching any bycatch”—laser nets made of light, for instance. “You’ve got to dream.”
These days, Warr fishes only a few days a week, delivering his catch directly to customers. He’s training to become a psychotherapist. But he keeps trawling, even though parts of it pain him, even though he doesn’t really think we should be trawling at all. Out of stubbornness, almost. To be a witness. “If I go, then who’s going to be here—to show you what it is?”