TO THE east of Amsterdam lies a tract of reclaimed marshland, the site of an epic rewilding project called the Oostvaardersplassen. It is sometimes nicknamed the Dutch Serengeti because of the profusion of large herbivores that graze there. But during the bitterly cold winter of 2017-18, deeply shocking images began to emerge. Thousands of deer, cattle and horses lay dead or dying of starvation. Desperate onlookers threw bales of hay over fences in an effort to help – clearly something had gone badly wrong.
Theoretical ecologist André de Roos was neither shocked nor surprised. His research had . Without the herbivores’ natural predators, he reported, overpopulation was unavoidable – leading to mass death when food ran out. The cold weather may have accelerated the die-off, but it would have happened anyway. “There were only ever two options: to allow mass starvation or to introduce culling,” says de Roos. One way or another, nature has what he calls a “requirement for mortality”.
This requirement takes centre stage in de Roos’s work. But it is often unrecognised by other ecologists, whose models fail to account for the complexity within any population – in particular, the fact that individuals may vary hugely depending on their stage in life, which can result in intergenerational conflict. As well as highlighting the benefits of death, de Roos’s thinking can explain some of the toughest brain-teasers in ecology. It also suggests novel ways of tackling economically important problems, such as the collapse of fisheries and the impact of noise pollution on marine mammals. Perhaps it is time we took it seriously.
Advertisement
Ecology’s classical models of how species interact – the Lotka-Volterra equations devised in the early 20th century – treat every individual in a population as identical. De Roos, who is based at the University of Amsterdam and at the Santa Fe Institute in New Mexico, thinks this simplification isn’t justified – and potentially invalidates any predictions the models make. He points out that it is antithetical to Charles Darwin’s teaching that there is variability among individuals and that natural selection operates on that variability.
“The biggest source of variation is that individuals go through a life cycle,” he says. Between a fertilised egg and an adult, an individual’s weight might increase as much as a billion times, depending on the organism. That requires resources. In most species, maturation isn’t dependent on age but on resources. Reproduction is also resource-dependent, and only a lucky few pull it off. Just 1 per cent of butterflies’ eggs become butterflies, for example. Juveniles and adults have different predation risks too. Essentially, they occupy different ecologies, but the field’s classical models take no account of that. The question de Roos asked himself nearly 30 years ago was: would it change our understanding of ecosystem dynamics if they did?
The answer from his first forays into research was yes. Ever since, he has pursued a “life history” approach, building developmental stage into his ecological models. This means that he accounts for how the available resources affect a population’s structure – the ratio of developing juveniles to reproductive adults, say – as well as how the population’s changing structure affects those resources. “He closes the loop,” says at Umeå University, Sweden, one of de Roos’s former collaborators.
It doesn’t sound too radical, but this approach often generates counter-intuitive results. Take the brown trout in Lake Takvatn, northern Norway. Following a collapse in trout numbers as a result of overfishing, de Roos and at Umeå University carried out a five-year cull of its prey, Arctic char. and the recovery persisted for years after the cull ended. To make sense of this, consider what happened to char. With fewer trout, their numbers grew, but this also increased the competition for food between char. As a result, their size and condition declined, so it took them longer to mature. This meant there were fewer newborns, which is what brown trout eat. Removing adult char relaxed that competition, leading to a proliferation of the trout’s favourite food – and eventually of the trout.
Ecological nudges
The fact that the recovery was maintained long after the cull ended shows that it is possible to nudge an ecosystem into a different stable state. In fact, one of the more profound messages of de Roos’s work is that the different life stages hold each other in tension and can settle into different equilibria depending on how the environment changes, says at Princeton University. “If an organism lived forever, its genes would be doomed because it doesn’t have the capacity to adapt to a changing environment,” he says. “Selection pressures therefore operate in multiple ways to break up gene combinations that are good in the short run, and to replace them with combinations that are better in the new environments.” Mutation and recombination are two ways in which gene combinations are broken up; population restructuring is a third.
The most extreme example of the intergenerational conflict that underpins population restructuring is cannibalism. It is widespread in the animal kingdom and was even practised by humans until it became a taboo. If adults and juveniles of the same species share an environment, and it becomes impoverished – water levels drop, say – they start to compete for a dwindling food supply. The adults may take to eating the juveniles, which makes evolutionary sense, says de Roos, because it increases the likelihood that at least some individuals of the species will survive.
Intergenerational conflict exists even when individuals of different life stages occupy different environments. Think of tadpoles in a small pond, which become frogs in a large forest. The forest is bountiful, so many frogs successfully reproduce and deposit lots of eggs back into the pond. But then the pond becomes overcrowded and food becomes scarce, so the tadpoles develop more slowly and are vulnerable to predators for longer. As a result, fewer adults emerge into the forest. “The outcome is counter-intuitive in the sense that the resource-rich environment, the forest, contains few adults, while the poorly resourced one contains many tadpoles,” says Diehl.
Another surprising result is that two predators preying on the same species may . If the first preys on juveniles, it relaxes competition between these young animals for food, meaning that more of them mature, producing more food for a second predator that prefers eating adults. The second predator, in turn, reduces the adult population so that more of them reproduce, generating more food for the first. “In classic theory, in which population structure is ignored, even the co-existence of two predators foraging on the same prey is impossible – let alone that they would facilitate each other,” says de Roos.
He thinks his models could make a significant impact in the real world, if only other ecologists would take them seriously. He has, for example, where overfishing of cod has been accompanied by huge increases in numbers of its prey, sprat, along with reductions in the sprats’ size and condition. These models might also help explain why some Atlantic cod stocks have failed to recover from a collapse despite a three-decade ban on fishing (see “Mystery of the missing cod”, below).
De Roos’s latest research concerns the effect of noise pollution on whales. The US Navy conducts submarine war games off the country’s coast that involve sonar, and it is legally obliged to gauge the environmental impact of those exercises. Until recently, the prevailing view among researchers was that the noise confused beaked whales, causing them to retreat from their habitual foraging grounds. Pregnant females were thought to be at risk of dying or losing a calf as a result.
Working with conservation biologist at the University of St Andrews in the UK, de Roos has found that, in fact, only a small subset of pregnant females – youngsters nursing their first or second calf, that are often still growing themselves – is adversely affected by the sonar. Moreover, this has a positive effect on the population as a whole. “All the others do better because there are more resources and less competition,” says de Roos, and the average number of calves per female goes up. When he and Harwood produce their final report for the US Navy, it is likely to cause much head-scratching over what constitutes a harmful environment.
De Roos has latterly turned his mind to biodiversity in general. Conventional wisdom holds that having a wide range of species present begets stable ecosystems. But in the 1970s, population biologist Robert May used classical models to show that the more interacting species there are, the more instability that should, in theory, generate. Ever since, ecologists have been struggling to explain why diverse ecosystems aren’t wildly unstable. “What this body of work entirely assumed was that species are homogeneous – if you’ve seen one individual of species X, you have seen them all,” says at the University of Florida. De Roos, unsurprisingly, took issue with that. By injecting life history into the models used by May, last year he showed that .
The hydra effect
Given such successes, you might expect de Roos’s ideas to be mainstream by now. Asked why ecology has snubbed him, he replies that he doesn’t think it has. “Colleagues are impressed by the work and the results,” he says. “But at the same time, they are intimidated by the complexity of the models, by the complexity of the methods you might need to investigate them, by the complexity and unfamiliarity of the results that emerge.” This is particularly problematic for ecologists working at the sharp end, managing real-world problems. “They often lack the support, resources or network to stay on the cutting edge,” says at the University of British Columbia in Vancouver, Canada.
Not everyone is entirely convinced by the emphasis on life history, though. Theoretical ecologist at the University of Toronto, Canada, says that is just one of many factors that classical ecology overlooks. In 2005, he and at Yokohama National University in Japan coined the term , after the many-headed serpent of Greek myth that sprouted two heads for each one Hercules sliced off, to describe any situation in which a higher death rate causes a species’ population to rise. Their description of the phenomenon doesn’t take life stage into account, and although Abrams agrees that life history might shape some systems, he suspects it is irrelevant in others. The that clog certain lakes each summer, for example, are best understood not in terms of life history, but in terms of the interactions of the microbial culprits with other microbes.
But de Roos does have a handful of loyal supporters, including Diehl. “The problems that we deal with in ecology are typically hard, and that means the predictive power of models is often limited,” he says. Ecology is complex, making it hard to untangle the various processes and come to simple conclusions. It can be frustrating, Diehl adds. “But if we don’t develop [life history] theory further, we may be looking for the wrong things in our data.”
After all, most ecologists failed to see that the booming grazers of the Dutch Serengeti were heading for bust. They learned it the hard way and the reputation of rewilding took an avoidable hit in the process. It is time this mindset changed, says de Roos. When it comes to nature, we need to see the upside of death.
Mystery of the missing cod
In 1992, the Canadian fisheries minister declared a moratorium on cod fishing off Newfoundland, ending a way of life for local people that had persisted for 500 years. Four years later, ecologists predicted that cod stocks, which had collapsed as a result of overfishing, . They still haven’t. Everyone agrees that the reasons for this are multiple – and include illegal trawling. But one explanation, suggested by André de Roos at the University of Amsterdam in the Netherlands, is especially surprising.
He feels that not enough attention has been paid to the Atlantic cod’s main prey, capelin. Since the cod collapsed, capelin have increased in number but decreased in average size. A concomitant drop in the abundance of their preferred food, zooplankton, suggests the reason might be increased competition among them for that food. Culling capelin seems like a bizarre way to promote the cod’s recovery, but de Roos thinks . His models are based on a radical revision of classical ecological theory, and although he hasn’t studied Atlantic cod, he has found a similar effect elsewhere (see main story).
Unsurprisingly, not everyone is convinced. Indeed, some Scandinavian ecologists have been calling for the opposite approach in the Baltic Sea, where overfishing has also reduced cod numbers. They want a for sprat and herring, two fish that cod prey on there. Meanwhile, Peter Abrams at the University of Toronto, Canada, thinks the problem may lie elsewhere in the complex food web to which Atlantic cod belong. He says that de Roos and other ecologists have failed to consider other species in that web, including ones that prey on cod, such as seals. “To me, that’s a more important omission,” says Abrams.
