Building with nature is an important concept in our coastal defence. After all, by hitching a ride on natural processes, we are more resistant to sea level rise. A good understanding of these processes is therefore indispensable. Ecologist Paul Berghuis, affiliated with Utrecht University and the Royal Netherlands Institute for Sea Research (NIOZ), delved into the world of dune formation and discovered that within young dune landscapes, dunes and grasses do not position themselves randomly but on the basis of a fixed pattern.
Berghuis' specialism is spatial ecology: he is particularly interested in spatial patterns in the interaction between organisms and their environment. For his research into dune formation, he studied the relationship between marram grass and sand. It has been known for years that marram grass plays an important role in dune formation, but Berghuis took this phenomenon to a higher level: by studying aerial photographs through time, he looked for fixed patterns in the landscape.
That may sound a bit abstract, but, as Berghuis notes: "Quite a lot is already known about individual plants and processes, but how does such a complex system of sand and vegetation interlock as a whole? That question influenced the scale on which we researched the dunes, and what that taught us." Berghuis therefore did not study the dune landscape as a collection of individually operating dune grasses, but as a landscape that organizes itself at the system level. "Then you suddenly see characteristics with regard to growth and resilience that are separate from how we can imagine that for individual dunes and dune grasses."
In previously studied ecosystems, such as mussel beds or the vegetation in savannas, a regular pattern emerges of spots of approximately the same size and distance between them. But something strange is going on with dune formation, Berghuis discovered. "There is no fixed size. Sand is moved by the wind again and again before it stays somewhere and where marram grass plants are growing closely together, the sand is best retained. At these high densities of marram grass, the dune continues to grow, often at the expense of its neighboring dunes. Small dunes are left behind or even disappear, and the larger dunes melt together."
This pattern is no fluke. Co-author Koen Siteur and other colleagues have previously argued that the spatial organization of some ecosystems resembles the physics of phase separation. This is the process we see, for example, when we make salad dressing, where oil floats on vinegar: after shaking, the oil droplets form larger and larger droplets, with the large ones growing at the expense of the small ones.
"Siteur's theory describes how these types of systems can be recognized by a very specific fingerprint, and we found that fingerprint in the dunes," says Berghuis. "The distribution of dune sizes, the way in which dunes merge, even the speed at which that slows down: it is all in line with what the process of phase separation predicts. This means that we now know not only how dunes behave, but also why they do this."
The fact that dunes have this fingerprint is more than a nice quirk. Siteur's theory also makes predictions about the response of such systems to disturbances. These predictions can now also be applied to dunes. "According to the theory, damage to the pattern does not heal itself once it exceeds a critical threshold value." In dunes you could see this when marram grass is disturbed by treading or when the wind causes a small pit. The disturbances are not repaired. Instead, the sand disappears to denser vegetated dunes elsewhere. "Our measurements show exactly such a threshold value in the amount of marram grass." At the same time, the theory predicts that the system as a whole will remain resilient: the sand will not be lost, but will be redistributed. For coastal management, this means: local damage can be permanent, even if the dune area as a whole remains healthy. For those who want to build with nature and not against it, these kinds of insights are crucial.
Proceedings of the National Academy of Sciences
Computational simulation/modeling
Not applicable
Density-dependent aggregation underlies spatial self-organization in coastal dunes
5-Oct-2026
n/a