Well, had another paper published today thanks to some diligent students, so thought I’d write a bit about it. This work addresses the question of whether life’s propensity to diversify is slowing down.

The simplest model of the growth in the number of species assumes that speciation happens at a constant rate, hence the number of species grows exponentially. This is what you’d expect to see if there weren’t yet forces, such as competition amongst species for places to live, limiting the number of species.

The next most complex model, you guessed it, assumes such limits exist and are progressively reached such that the rate of increase in species number slows down.

People have tested these models against both data from the fossil record and evolutionary trees, normally of living species. The fossil tests tend to use groups of species, like genera or families, because they are supposedly more robustly identified and have more complete records. Both models have been supported, and in fossil data the second models tends to be more supported in marine groups, the first in terrestrial groups. This makes intuitive sense because the marine environment has been inhabited for longer and tends to support fewer species.

However, one problem with fossil studies is that the fossil record is incomplete, hence groups that appear to diversify exponentially may actually have done so differently because the true origins of groups will be earlier than their current first fossil.

In our study we tested whether groups can appear to diversify exponentially when they actually conform to the second model, in dragonflies and their relatives. The idea behind using dragonflies is that they are insects, hence represent a dominant group of organisms, whose fossil record roughly conforms to the standard terrestrial model of exponential growth. They are also very old, hence we have more data to play with for analysis and there is more opportunity for rates of growth to have slowed.

Above: some fossil “Protodonata” from the Carboniferous and Permian periods, representing some of the earliest representatives of the group to which modern dragonflies belong.

We compiled their fossil record at family level from published descriptions. We then estimated where the gaps in the fossil record are likely to be by comparing the age of the first fossils of sister families. Sister families share a common ancestor, hence have the same date of origin. If however the dates of their first fossils differ, and our sister relationships are correct, the younger family must actually be as old as it’s sister. That means there is a gap in the fossil record which we can hypothetically “fill in”. In this way we can use the sister relationships to correct the origination dates of the families described by fossils alone.

Above: a fossil Odonate preserved in a Jurassic lagoon deposit.

So we did this for dragonflies. We compiled a summary of our best guess of sister relationships by two so-called “supertree” methods (supertree methods are just ways of summarizing published evolutionary trees in a single large tree). We identified the fossil gaps, and then updated the family origination dates accordingly. Then we used these new dates to redraw the growth of the group through time to test if there was a sign of a slow down.

In dragonflies and their kin, the raw fossil data show little sign of a slow-down, as may be typical for terrestrial groups. However, the updated diversity curves do show a slowdown. Hence, our initial speculation turns out to be confirmed by the data. This suggests that other terrestrial groups may also prove to have diversity limits once the imperfections of the fossil record are taken into account, hence perhaps marine and terrestrial both actually conform to the second model, despite what the fossil record might suggest when taken at face value.

We were also able to compare the species richness of the different dragonfly groups to tentatively identify those which had radiated or contracted in diversity at rates different from their ancestors. We identified several such groups; generally supporting previous suggestions that colour is important in generating dragonfly diversity; the more colourful the group, the greater the tendency to diversify. Colour is probably important because it indicates a tendency to be fussy about choosing mates, and that can lead to reproductive isolation between populations, and eventually more species.

Above: a colourful dragonfly belonging to a diverse group.

You can read the full paper here.