Australasian Science: Australia's authority on science since 1938

The Tree That Waters Itself

The stilt roots of pandanus (top) have a network of aqueducts that channel water to a dedicated organ of spongy tissue at the root tip, which is sometimes suspended in mid-air (bottom).

The stilt roots of pandanus (top) have a network of aqueducts that channel water to a dedicated organ of spongy tissue at the root tip, which is sometimes suspended in mid-air (bottom).

By Matt Biddick

Island gigantism might have driven the evolution of one of the most bizarre adaptations in the plant kingdom.

Imagine War of the Worlds meets Day of the Triffids. It’s unmistakable. Stretching 20 metres into the rainforest canopy and perched upon a fortress of stilt-roots, this staunch island giant is the king of the jungle. And the king is thirsty...

Pandanus forsteri comes from a long line of “screw pines” that call nearly every square inch of the Old World Tropics home. While most trees grow thicker to support their ever-increasing size (think of tree rings), screw pines cannot. Instead, they produce stilt-roots that descend from the trunk to the ground for support as they mature. Sticking out from the trunk, they can sometimes take years to reach the ground.

Some 750 species are used by oceanic peoples for everything from weaving, fishing and cooking to medicines, cultural decorations and psychedelic religious practices. There’s even a pandanus language!

On Lord Howe Island, a UNESCO World Heritage Site located 600 km from the east coast of Australia, something truly remarkable has transpired: the evolution of an island giant. Most pandans seldom exceed 4–5 metres. At this height, few P. forsteri trees are even mature enough to produce fruit.

What drove them to the skies? To answer that, we need to look further afield.

Size changes are actually one of the most repeatedly observed and predictable trends in the evolution of life on islands. In animals this is known as the “island rule”. Animals on islands are thought to be freed from the burdens of continental life, and converge on a more optimal ancestral form. Large mainland animals become comparatively dwarfed while smaller critters become comparatively gigantic.

In New Zealand, small continental grasshoppers have taken the place of rodents and evolved into some of the largest insects found anywhere on Earth (e.g. flightless weta). Meanwhile, on the island for Flores, the infamous “hobbit” fossils echo the existence of our pygmy hominid cousins.

Evidence suggests that island plants follow suit. But life as a giant has its costs, and these costs are hefty for P. forsteri.

To support the weight of a colossal pandan, you need colossal stilt-roots. But it’s not that easy. Plants are like natural straws. Roots draw water up from the soil and into their tissues by harnessing the negative pressure produced when water evaporates from their leaves. However, this is an impossible feat for roots that are suspended metres in the air, and yet to reach the ground. It’s also extremely dangerous.

Plant vasculature is not so different from our own. Proper functioning depends on an unbroken line of pressure from point A to point B. Exposed to the air, pandan roots can form air bubbles. An air bubble in a plant is bad in the same way as an air bubble in your artery. It can kill.

To avoid this, selection has favoured the evolution of one of the most extra­ordinary innovations in the plant kingdom: a rainwater-harvesting system. This operates in a remarkably similar way to rainwater-harvesting systems found on modern buildings. It comprises three parts.

Gutter-like leaves capture rainwater and channel it to the trunk, where it descends toward the roots. Water is then couriered along a network of aqueducts formed by the root surface. Descending water completes its journey at a dedicated organ of spongy tissue covering the growing root tip. The end result: an ingenious use of plant architecture that enables P. forsteri to water its own roots before they reach the ground.

Furthermore, the root aqueducts are strategically positioned to maximise the flow of rainwater to the roots. Taking a walk through a P. forsteri grove is like ducking and weaving through the legs of titans. Some roots come down like pillars from above. Others stick out like tent poles and trip you up.

Aqueducts are of no use with near-vertical roots, as rainwater will passively descend to the root regardless. At shallower inclinations, however, water runs to the underside of the root and drips off without ever reaching its destination. As if it knows it consciously, P. forsteri concentrates the production of aqueducts here, in roots that benefit the most from their presence.

This dramatically differs to how we traditionally conceptualise plants. The past century-and-a-half of plant water relations research has seldom left the domain of sap. Think back to your high-school biology class. Plants are comprised of xylem and phloem – the conduits through which water is transported. Water is pulled up the xylem while the products of photosynthesis – the sugars and minerals that fuel growth – are distributed down to plant organs through the phloem. The flow of sap is an entirely internal process.

What P. forsteri so elegantly demonstrates is that, under certain circumstances, plants can evolve to manipulate the flow of water external to the plant. Other plants are known to capitalise on intercepted rainfall. For instance, some epiphytes (plants that live exclusively in the canopy of other plants) preserve a sheath layer of dead tissue around their roots that rapidly soaks up water and charged ions when it rains. Tank bromeliads accumulate rainwater in cup-like structures formed by their leaves. However, P. forsteri is the first species known to possess adaptations that operate jointly to capture, transport and store rainwater in this way.

This humble giant has opened our eyes to an entirely new field of scientific inquiry. It forces us to reconsider the functionality of organs like leaves and roots outside of the context of photosynthesis and the sequestering of water and minerals.

Do other plants harvest rainwater in this way? How many times has it evolved? Have we overlooked an entire aspect of the adaptive utility of plant architecture?

This adaptation may be relatively simple, in that it is comprised of modified organs that already perform other functions. Perhaps that is the reason we are only now discovering such evolutionary innovations.

In science we are reminded all too often how nature engineers elegantly simple solutions to notoriously challenging problems.


Matt Biddick is a PhD candidate at the Victoria University of Wellington’s School of Biological Sciences.