Wetenschap - 17 januari 2017

Shifts in nature cannot be predicted

Albert Sikkema

Turning points in nature, such as the shift from a clear lake with aquatic plants to a turbid lake with algae, cannot be predicted easily. The "early warning indicators" developed by ecologists fail to provide reliable and constant signals of an imminent shift in lakes. That is the conclusion of researchers at the Netherlands Institute for Ecology, NIOO.

<photo: Pixabay>

For years ecologists, including Marten Scheffer's group in Wageningen, have been studying how complex ecosystems can suddenly shift because of seemingly small changes. Scheffer was studying the underlying patterns and warning systems of these turning points in nature. He demonstrated those early warning indicators for such shifts in climate, human depression and ecosystems in a laboratory setting. His colleague Stephen Carpenter and Chinese researchers also found such signals for lakes.

Alarm bells
The NIOO researchers collected the data from five other fresh-water systems that had been studied at length and that had undergone an ecological shift. Working retrospectively, they investigated whether alarm bells had started to ring before the shifts to these ecosystems occurred. They concluded that none of the four possible indicators had good predictive value.

This conclusion seems to put an end to Scheffer's attempts to find general robust signals warning about an impending collapse of an ecosystem. But the professor of Aquatic Ecology had already reached the same conclusion. 'A shift can never be predicted because it also involves a coincidental effect,' Scheffer stated. 'But we can measure whether or not a system is losing vitality - or perhaps gaining some. To avoid misunderstandings, we no longer speak of early warning signals but of indicators of resilience.’

In some of the cases studied, the researchers Alena Gsell and Annette Janssen at NIOO found indicators of a forthcoming ecological shift, but they were not unambiguous. But Gsel thinks that we will be able to predict ecological shifts in the future. 'In our study, we look at the past and use ecological data that was collected once a week or biweekly. That makes it difficult to find robust indicators. But new equipment enables us to collect ecological data every day or even every hour. Perhaps we'll be able to make progress a bit at a time.'

Critical slowing down
In 2009 Scheffer wrote a prominent article in Nature about warning systems for critical shifts in ecosystems. The most important was that ecosystems have a balance that is constantly being disrupted. The more slowly an ecosystem recovers from such a disturbance, the weaker the system. This process of critical slowing down is thus a good indicator of an impending ecological calamity.

But in 2015 Scheffer reported in Ecology, Evolution and Systematics that there are a number of fundamental limitations to catching these warning systems. Environmental conditions can cause a recovering ecosystem to suddenly shift, against all expectations. Moreover, the process of critical slowing down is also caused by differences in temperature, but this is not an indicator of ecological weakness. At best the indicators can provide a qualitative judgment of whether the vitality of an ecosystem is increasing or decreasing, Scheffer stated.