Global patterns in forest and savannah species

So I memtioned in my last post how I'd enjoyed taking some of the conference attendees on a little safari on the weekend following the conference. We had a night in Tarangire and then a day in Arusha National Park. Now I'm really an ornithologist, who plays at being a savannah ecologist. I'm not a botanist at all. So driving around with people who really are is always educational, and the number one insight that I got from the weekend was the extraordinary degree to which Bill (who works in Brazillian savannahs) and William (working in African savannahs) could identify a plant - say a Xanthoxylem and William would turn to Bill and ask if they had the same genus in Brazil. To which, almost always, the answer was yes. Even more remarkable (to me) was the fact that on occassions they even had the very same species.

Botanists getting serious, Tarangire June 2011. I think it's a grass.

 Interestingly, whenever there was a genus match between continents, if we were in a savannah, the same genus was always a savannah plant in Brazil, whilst if we were in the forest on Meru, the South American members of that genera are also forest plants. To me as an ornithologist thinking quickly, I can come up with no more than two or three bird genera that are shared between the continents (there are a few Turdus thrushes in both places, Tyto barn owls, etc.), and that's it.

I was brought up as an ecologist understanding that biogeographical (bio - from biology, geographic, from geography of course - biogeography being the study of distributions of species) similarities between continents could usually be explained by a process known as vicariance. This idea essentially explains the distribution of related species by assuming that a common ancestor of the current species lived on a continent that then moved around through the process of continental drift. A typical example might be the distribution of Ratites  - the group of large flightless birds that includes the ostrich. The closest relatives to the ostrich include the emu in Australia, the rhea of South America and the kiwis of New Zealand. Their distribution in these southern continents is explained by their shared ancestor living on the ancient continent of Gondwana, a single continent that eventually broke up (around 200 million years ago) to form the southern continents (plus India and the Arabian peninsular). Each fragment carried a population of this ancient ratites and today we see a distribution of birds across the southern hemisphere.

Ostrich are ratites, a typical Gondwanan group with a distribution explained by vicariance

This explanation of shared ancestry, each population of which floated off on it's own continent it the one that immediately springs to the minds of ecologists of my generation where confronted with similar species across southern continents. But the break-up of Gondwana took place around 200 million years ago - and that's a very long time for evolution to have been acting. Although most ratites are fairly similar and the shared ancestry immediately obvious (though check the kiwis!), they're actually very different and certainly not in the same genus. Now, it's important to remember that, unlike species, genera are not very well defined groups - rather they are a taxonomists attempt to identify common ancestors and group similar species - but whether we group 50 similar species into five genera in one family, or one large and diverse genus within a family is rather more arbitrary than the similar decisions about species (though even there it's actually surprisingly tricky!).  So I already knew that the best predictor of how long ago the common ancestor of any particular genus lived is nothing to do with the variety within the species, but everything to do with the number of taxonomists that work on the group - the more taxonomists, the more genera, the more recent the common ancestor. So my first question was whether the common ancestor of these plant species really lived more than 200 million years ago and are just kept in the same genus because there's such a shortage of taxonomists. And I learnt that whilst my head has been full of other things, I've missed one of the biggest revolutions in biogeography of the last decade.

Meru's forests were full of genera also found in South America (and Australia!) Erica have interesting distributions, but not in the New World.

Now we can use DNA to provide fairly accurate dates on when individual species shared common ancestors, we've been able to see that, contrary to the vicariance ideas I've been brought up with, that imply aces over 200 million years, most of the shared genera across the southern continents seem to be far more recent that Gondwanan in origin, which implies that they must, time and time again, have managed to disperse from continent to continent. Wow! What's more, it seems that more often than not, Africa has been the source of the movement, rather than the recipient. Amazingly (to me at least) even some plant species that are dioecous - i.e. have male and female plants - have amnaged to generate almost global distributions through regular long-distance colonisation events. (Unfortunately plant names have a habit of slipping my mind and I can't remember the one that impressed me most - and my pencil was broken so I couldn't take notes. Rubish, huh?!) That is pretty extraordinary I think! So next time you wonder how a seed disperses from a tree, and how it could ever move more than a few metres, remember that most of these genera have managed to get from one continent to another, probably several times! Surprisingly, though, despite these multiple movements across continents, whenever a plant does make the jump it has never (or nearly never) colonised a different biome - savannah plants have to find themselves a slot in a savannah, forest plants in a forest. Which probably tells us all sorts of interesting things about how plant communities are put together, but that will have to wait for another post...