Alan G

I second this, very good book

Leigh

The trouble is that the fossil record is biased in favour of organisms with hard parts. The Burgess Shale is a rare example of preservation of animals with soft parts. In fact,in the Burgess Shale, only about 1 organism in 20 is an organism with hard parts. If we extrapolate from this, there must be many soft-bodied organisms that have left no trace in the fossil record.

The fossil record has been compared to a book, with only a few pages surviving, of those pages only a few sentences, of those sentences only a few words, of those words only a few letters. (sorry, can't remember whose quote this is)

So, 99% is an estimate - but probably an underestimate, if anything.

Pop_Quiz

Since there are over a million species of bugs and insects now (400,000 beetles and counting, among them*), imagine what the numbers were in the Carboniferous period when 3-foot wide dragonflies filled the niche of the then-new flying top predator. 400,000 species of plants now would have been easily multiplied several times over during the past half billion years.

*I wish I could remember where I saw the interesting anti-creationist factoid of considering if Noah took only a "kind" of beetle onto the ark and not all half million species of beetle, how was it possible for that many to amazingly "adapt" into the current speciation. It comes out to a new species every few days over 4000 years---even more amazing than evolution itself.

Sure, Pangea may have cut down on the diversity to a point, but every time a new organism evolved, a new parasite was bound to come right along with it. And a potential predator of that parasite. And since we can't even say we have an accurate count in today's ecosystem, the fossil and amber record is sorely lacking in an accurate picture of true species count.

Momma Nature hates vacuums, as the saying goes, and every bit of land and a good part of the water would have had its share of bugs and insects to occupy predator, prey, and scavenger spots (they still do- it's just that eagles have replaced the mega-dragonfly). No birds or lizards to chomp on 'em at the time...nothing but each other (creating at least as much diversity as we see today) and the occasional 'dino' large-mouth bass.

Just because something isn't pretty, all that visually distinct from another,or found nowhere else on Earth but on a few acres in the Amazon rain forest doesn't mean it shouldn't be included in a species count; a billion all-told count of species is easy to consider; some go as high as 2 billion. The 1% figure comes from the fact there are as many as 20 million species of living things, plant and animal, currently on Earth.

BTW, I heard somewhere some scientists consider the Age of Mammals to be a very drab one as far as diversity goes.

crownboy

Figure I'd give this topic a second go around. Thanks for the answers so far, I'll try and get that book.

Nic Tamzek

In no particular order:

It would be more interesting if humans hadn't killed all the cool ones back in the Ice Age.

I have no idea. It all depends on definitions. Species are well-enough defined for extant organisms but it gets very difficult with fossils.

{An aside: I am also increasingly annoyed with taxonomy in general. Are all the descendents of X also an X, or not? E.g., birds are descended from dinosaurs, but does this really mean that birds are dinosaurs?

I would be fine with this if this were how we decided to name things, but then: why are humans in a different family than chimps?

http://www.umanitoba.ca/anthropology.../taxonomy.html

Hylobatidae
Gibbons Simangs

Pongidae
Orangutans
Gorillas
Chimpanzees
Bonobos

Hominidae
Humans

...it's all a big fat mess if you ask me. Which you didn't, but thanks for letting me vent.

Doubting Didymus

What are you, some kind of cladist?

(Just kidding, I'm a cladist myself. And a gene-centrist to boot. I've got half the biological discipline after my blood!)

MrDarwin

I don't know the precise methodology that came up with this figure, but I can guess a couple ways to estimate. One would be to look at two or more fossil assemblages from the same area, but separated by some length of time, estimate the number of species that are fossilized in each of the assemblages, and the number of species that occur in all of them.

Another would be to look at a relatively recent fossil assemblage, and identify just how many of its species are still alive today. I imagine this would be a more reliable method, as you don't have as large a margin of sampling error for one of the samples (i.e., the species that are alive today). For example, the La Brea tar pits have a pretty good representation of southern Californian Pleistocene vertebrates; how many of those species are still around in southern California (or anywhere, for that matter--we're talking about extinction, not geographic distributions)? A pretty small number, if I recall correctly (although most have related species, often in the same genus, still around).

(Edited to add a slightly less serious observation, because if I don't somebody else will: of course those species are all extinct, if they kept falling into tar pits!)

charlie d

A nice, simple but thorough site on extinctions is here. Good list of references too.

RufusAtticus

Nic,

A couple years ago Pongidae was finally merged into Hominidae.

Tharmas

Raup is a good reference and gives a great many methods, such as examining marine environments that have survived intact across major extinction events so you can really count the numbers of clams and oysters and snails (some local six graders have recently done some publishable work in this regards by counting sharks’ teeth on either side of the KT boundary and cataloging the decline in species).

But Raup also gives a very elegant (simple) metric:
1) We know from the fossil record that on average a species lasts about a million years give or take.
2) The fossil record extends back 500 million years.
That makes the 99% figure very conservative.

Of course that's a bare bones extrapolation. It’s the detailed observations that flesh it out and confirm it.