Most plant species per square metre (1 Viewer)

[Arbu]

Where to Watch Birds in Africa states (page 316) that the fynbos of the mountains of the South African Cape support more plant species per square metre than any other habitat on earth. Is this true? Clearly a typical square metre of fertile meadow will hold more species than a typical square metre of rainforest, because the plants are much smaller. But I would have thought a lush lowland meadow would trump South African fynbos in this respect. Looking at pictures, the plants in the fynbos look to be of reasonable size. Does anyone know what WWB's statistic is based on, and has anyone really considered this question?

 

[Butty]

plant species per square metre

a typical square metre of fertile meadow will hold more species than..

These are not the same thing. The words are important.

 

[Arbu]

These are not the same thing. The words are important.

So you're saying the figure relates to all habitats on earth, calculating a figure by dividing the number of species in them by the area. Which means it just comes down to how narrowly you define a habitat. The fynbos have an area of 90,000km2 and about 9,000 species of plant. So that means one species per 10km2. But if we define a more narrow habitat, say Devon hedgerows we can do much better. 50,000 km of hedges in Devon, not all Devon hedgerows, but anyway, probably 2m wide makes 100km2. 600 species of flowering plants gives 6 species just of flowering plants per km2. Way better. So is there a standardised list of habitat types on earth which includes fynbos but not Devon hedgerows? Or is the claim nonsense?

 

[Butty]

So you're saying the figure relates to all habitats on earth, calculating a figure by dividing the number of species in them by the area.

No, I've said what I've said - and I haven't said what you say I've said at all. The figures you quote are essentially (intended to be) floristic diversity. Fynbos is well-known to be highly floristically diverse (and if you spend more than 10 minutes in it, and know anything about plants, you will very quickly see it to be so). Tropical forest is also highly diverse - but far less obviously so to the casual observer. So the actual means of measuring diversity is clearly important - and my guess (I've no idea) is that, in order to be meaningful, it's done in terms of the number of species in a reasonably large area (something of the order of 10-100 ha??). Doing it that way would put tropical forest well ahead of temperate European chalk downland - whereas counting the number of species in a random one-metre square would produce exactly the opposite result.

 

[THE_FERN]

It's commonly held to be true. It's unlikely to certain, though, given that many rich areas of the world remain +/- unsurveyed.. Obviously you have huge species densities in places like the Amazon.

The explanation is that the climate of that tiny corner has remained stable for a very long time. Coupled with the topographic complexity of table mountain &c, and you have many opportunities for narrow specialisation

 

[Arbu]

There are only two possible meanings of "per square metre" here. Either it's a typical square metre, as I suggested in my first post, or it's just meant as an arbitrary reference to a unit of area, referring to the whole existing area of the habitat. So I took it that the claim must refer to the latter when you said it wasn't the first. But we've seen that the latter doesn't make sense either. Now you are saying that "per square metre" probably means "per 10-100 hectares". So the truth seems to be perhaps that the fynbos have the most floral diversity of any habitat on earth if you pick a carefully chosen area, big enough to put them ahead of chalk downland, and small enough to put them ahead of tropical rainforest.

 

[THE_FERN]

Should have said, it's obvious, really, but everything in ecology has an attached scale. Change scale and you're comparing apples and pears. So in the Amazon alpha Diversity (that in one place) is very high but beta Diversity (how Diversity changes over wide areas) is much lower as many species have (under current taxonomy) have vast ranges.

In the cape floristic province, you have both high alpha and beta diversity I think: big changes in species composition over short distances

Edit: should have said scale and location dependent (the definition of the area) of course

 

[dantheman]

Well, I recall that on south facing grazed(?) limestone UK sward 30 - 60 species per square metre ie a count can be made inside a physical square of 1 metre by 1 metre that many plant species ...

 

[THE_FERN]

There are only two possible meanings of "per square metre" here. Either it's a typical square metre, as I suggested in my first post, or it's just meant as an arbitrary reference to a unit of area, referring to the whole existing area of the habitat. So I took it that the claim must refer to the latter when you said it wasn't the first. But we've seen that the latter doesn't make sense either. Now you are saying that "per square metre" probably means "per 10-100 hectares". So the truth seems to be perhaps that the fynbos have the most floral diversity of any habitat on earth if you pick a carefully chosen area, big enough to put them ahead of chalk downland, and small enough to put them ahead of tropical rainforest.

Yes but what is the area over which you make the measurement. Density per m averaged over Kew gardens is going to be higher than density averaged over the whole of Richmond park. There's one value averaged over the earth as a whole, but an infinity of different values (bounded at 0) over all other scales below this. Carefully cherry pick your m of English flower meadow and you might surpass the average for the cape (unlikely though..?)

 

[THE_FERN]

The other issue is what we mean by "diversity". Species richness might not be (probably isn't) a sensible measure. Phylogenetic diversity likely is.

In that case, cape probably still wins even though there are so many ericas because you've also got cycads, proteas etc. It will certainly beat anywhere in n Europe, and perhaps only n Queensland would come close/surpass it. But there are significant unknowns here: I'm not convinced distributions are well-known even in Queensland, and then there's all the uncertainty around the phylogenies...

 

[Arbu]

Yes but what is the area over which you make the measurement. Density per m averaged over Kew gardens is going to be higher than density averaged over the whole of Richmond park. There's one value averaged over the earth as a whole, but an infinity of different values (bounded at 0) over all other scales below this. Carefully cherry pick your m of English flower meadow and you might surpass the average for the cape (unlikely though..?)

That's what I'm saying. You can cherry-pick the area in order to give each of a number of different habitats the title.

 

[Butty]

the truth seems to be perhaps that the fynbos have the most floral diversity of any habitat on earth if you pick a carefully chosen area, big enough to put them ahead of chalk downland, and small enough to put them ahead of tropical rainforest.

You're suggesting that the data are being deliberately manipulated to prove a chosen result. There's no reason to think that.

 

[THE_FERN]

That's what I'm saying. You can cherry-pick the area in order to give each of a number of different habitats the title.

These things are inevitably somewhat arbitrary. There are few areas or habitats where the edges are well defined. For example, exactly where does any city in the UK start and finish..? Where does an estuary or sand dunes begin and end (precisely)? The fact that geography is fractal adds to these problems ("how long is the coastline of Great Britain?")

Even more the case with assemblages like "floristic kingdoms" or "provinces"—will inevitably be some species which are not restricted to that area. Unless it's an island in water (not a mountain island) there may not be limiting physical barriers which you can use to precisely define boundaries.

(There are exceptions, of course. Wild-born, truly native mute swans are very careful to not cross into Iberia or the UK—only introduced ones there (ebird).)

But this fuzziness is true of every aspect of biology. Despite it, there are patterns. Given 1/5 of all plants in Africa are in the Cape (wikipedia), there's clearly something going on. Ecologists seek to understand these patterns, how they vary with scale and location. To do so you have to control for issues or arbitrariness (cherry-picking), scale and definition. You might use null models, repeated sampling at different scales, repeated sampling over random subsets at the same scale, and different taxonomies* / phylogenies to test the robustness of any patterns.

* Since all modern taxonomies attempt to reflect phylogeny so you can treat them as unresolved phylogenies.

 

[THE_FERN]

You're suggesting that the data are being deliberately manipulated to prove a chosen result. There's no reason to think that.

The statements you find in many guides and older text books do tend to be simplistic. Certainly the issue of unequal relatedness is a relatively new thing (turn of the century). They also often mix apples and pears—species richness, species density or..., may cherry-pick unintentionally ("table mountain") or not make the bases for comparisons clear.

In summary, I'd say the Cape is an outstanding and outstandingly interesting place. This is likely due to its long climatic stability paradoxically coupled with periodic "medium-frequency" disruption (fire). It's impossible to say whether the phylogenetic diversity per unit area in the Cape is the highest in the world. I suspect it's unlikely—my money's on Queensland given all the relict taxa there.

In contrast, UK is of low diversity, mostly due to its having been covered in ice until recently (but also its terrible climate !). Its botanical interest may mostly lie in incipient or microspecies (Sorbus, some orchids etc):—discuss.

 

[Arbu]

You're suggesting that the data are being deliberately manipulated to prove a chosen result. There's no reason to think that.

There's lies, damned lies and statistics. Happens all the time.

 

[Butty]

lies, damned lies and statistics

And, more relevant here, there's lack of a scientific outlook, not having any more-specific background knowledge, and general sloppiness. The attitude typical of journalists and blurb-writers, in fact 👎🏼

 

[Acid John]

Can I chuck dandelions in to this argument.
Britain and Ireland have either one variable species (Lumper) or circa two hundred and fifty species (Splitter).
But where ever you toss your meter square you will probably still only record one species because the microspecies are rare or widespread but scarce.