What is a forest?

Well… it depends. You might think that I’m being facetious here. When I ask this of my students, they immediately suspect that it’s a trick question. After all, I do describe myself as a forest ecologist. But if you pinned me down and forced me to respond then I would struggle to give a clear answer. In this post I’ll try to explain as clearly as I can, and thereby demonstrate that we have a real, serious problem.

Let’s get the etymology out of the way first. In Medieval English the word forest did not refer to a place with trees. Its original meaning was one of the King’s hunting grounds which were managed particularly to maintain populations of deer. Many of these contained trees, but that wasn’t their distinguishing feature. Some were actually heathlands. To this day Exmoor forest hardly contains any trees. Contrary to popular misconception, that’s not because they were cut down — they were never there in the first place.

Here in Nottinghamshire I live only a short drive from Sherwood forest. Myths of outlaws in the wild woods are doubly fictional; neither Robin Hood nor his primeval haunt have any firm basis in historical fact. This has been a managed landscape for many thousands of years. A quick look at everyone’s favourite tree, the Major Oak, gives the game away.


Major Oak in Sherwood Forest in 2006 by Marcin Floryan.

This is not a tree that grew beneath a closed, dark canopy. Oaks in such situations grow tall and slender, forced upwards by strong competition for light. The north-east of the United States is carpeted in young oak forests, none of which contain trees shaped anything like the Major Oak. Its low spreading branches tell of a lifetime spent in a wide open area, surrounded by heathland, which has only in the last century been filled in by the mass of birch trees which now surround it. These birches are themselves a signifier; birch only grows in bright, open habitats when no browsers are around to eat its seedlings. If you wished to restore Sherwood Forest to how it looked at the time of Robin Hood, the first thing to do would be to cut down most of the trees.

The origin of the word ‘forest’ is therefore not much help in deciding what one is. The modern understanding is of an area covered in trees, so that’s a better place to start. But what is a tree, and how many of them does it take to make a forest?

Defining a tree isn’t straightforward either. Botanists, foresters and ecologists will all give different answers. Here’s my rather technical definition: a tree is any plant actually or potentially forming a free-standing monopodial stem of at least 1.3 m in height, with secondary thickening, and reaching a height of at least 5 m.

Let me parse that quickly so you can see where I’m coming from. The opening caveat ‘actually or potentially’ is necessary because it allows for the fact that many plants are capable of growing into trees, but haven’t got there yet (they might be young) or are prevented from doing so (e.g. by regular grazing or coppicing). A bonsai oak is still the same species as a giant oak, it’s just been constrained by a tiny pot. Free-standing is an attempt to keep out (most) lianas, which are structural parasites. Monopodial means having a single, dominant stem; this separates trees from shrubs, which have many stems and tend to branch below 1.3 m. As for the secondary thickening, that implies the existence of bark and other woody tissues, and excludes giant herbs such as bananas. The height of 5 m is arbitrary and intended to filter out any remaining shrubby species.

As an aside, almost any plant species is capable of evolving into a tree. Nowhere is this more obvious than on oceanic islands, where the chances of reaching them are slim, and therefore whichever plant is lucky enough to get there first can evolve into a dominant tree. On the Juan Fernández islands or in Macaronesia the native trees are actually lettuces, while tree sunflowers dominate forests in St Helena and the Galápagos, and tree silverswords in Hawaii show little resemblance to their weedy ancestors.

Scalesia pedunculata. It's a Galapagos tree, and also, basically, a daisy. By Haplochromis CC BY-SA 3.0 via Wikimedia Commons.

Scalesia pedunculata. It’s a Galapagos tree, and also, basically, a daisy. By Haplochromis CC BY-SA 3.0 via Wikimedia Commons.

There will always be some in-between cases but we’ll leave them to the specialists. Some lianas start their lives as small trees; some species can grow as either trees or shrubs; no-one really knows what to call a hemi-epiphyte like a strangler fig. Nature doesn’t easily fit into our boxes. Still, this gives us something to work with. It also ends up including species like oil palm, which are farmed as cash crops, and no-one would want landscapes dominated by them to be defined as forests. But we have to draw the line somewhere and specific chauvinistic exclusions would make a mockery of scientific impartiality.

It’s worth pausing briefly to note that the definition does create some perverse consequences. It’s not too dissimilar to that used by the FAO (Food and Agriculture Organization of the United Nations). A major bone of contention in their assessment of forest cover is the acceptance that any land which could grow trees is still a forest until its land use category changes. Completely burnt to the ground? Still a forest. Clear-felled? Still a forest. So long as trees are expected to grow there again one day, the FAO will call it a forest, even if there aren’t any there right now.

How many trees makes a forest? One common approach is to look at how much of the land surface area is covered by tree crowns, and I would personally draw the line around 40% (based on Sasaki & Putz 2009). Note that this is more restrictive than the definitions used by the UNFCC (United Nations Framework on Climate Change) or FAO.

All this causes further problems. My criterion includes things like orchards and rubber plantations, which are not forests by any folk definition. Likewise it excludes vegetation types with low tree densities (e.g. savannahs or dehesa) and shrublands. Some will get very sensitive about this because of the popular perception that conserving forests is the most crucial issue. I do not mean to devalue these landscapes, but would argue instead that these deserve separate treatment rather than attempting to cram them into an all-inclusive definition of forests.

South African fynbos. Undeniably a conservation priority, but calling it 'forest' doesn't help.

South African fynbos. A conservation priority, but calling it ‘forest’ doesn’t help.

Why does this matter? Well, deciding what a forest is makes a major difference in determining how much forest there is in the world. A new paper by Sexton et al. in Nature Climate Change points out that varying your criterion for tree cover from 10 to 30% changes the amount of forest recognised globally from satellite imagery by 6%. In other words, it’s a difference in area the size of China. In the tropics alone it represents a difference of 45.2 Gt of carbon. That’s a much bigger discrepancy than anything caused by arguments over defining a tree, or whether we should include rubber plantations. All those fine details drift into irrelevance.

While different countries are allowed to decide what they call ‘forest’ there will be political motivations to move the goalposts. If we’re trying to battle climate change by tracking the amount of carbon locked up in forests, or save species, or maintain ecosystem services, it’s pretty crucial to find out how much forest we have.  When NGOs bemoan ‘forest loss’, or governments declare successes in ‘forest protection’, you should maintain a healthy degree of scepticism about what they’re actually recording and where their figures are coming from. If we are to meet the Aichi Biodiversity Targets or use forest stocks to compensate countries for storing rather than releasing carbon through the REDD+ program then what we call a forest matters. It matters politically, financially, socially.



Global forest cover as detected from satellite data depending on whether one uses 10% (top) or 30% (bottom) land cover by trees as the criterion for defining forest (originals from NASA Earth Observatory, http://earthobservatory.nasa.gov/IOTD/view.php?id=86986&src=eoa-iotd).

All this argument leads to one basic conclusion, which is that the word ‘forest’ is a blunt and imprecise tool when it comes to describing the things we care about. Forests include areas that are old-growth, degraded, secondary or plantations, all of which have different sets of values depending on your perspective (see Putz & Redford 2009 for a great discussion). It is possible for serious degradation of habitats to take place without them ever ceasing to be called ‘forests’, and a dichotomy with ‘non-forest’ is unhelpful in documenting or preventing this. Forests are not merely present or absent.

Let’s just agree that many things can be called forests and stop arguing over the definition. More important is to ask why we care about forests, and start measuring the values that matter to us such as species, carbon, or services. Then, at last, we might be able to move past this distraction.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s