Taxonomy & Evolution

Scientists suggest bringing a fresh set of eyes to the problem, ‘When did the first flowers evolve?’

A disagreement on when flowers evolved could be solved by studying the real experts on pollination, the pollinators.

If you want to learn when the first angiosperms (flowering plants) evolved, you should look beyond the plants to their pollinators, say Casper van der Kooi and Jeff Ollerton in a new Perspective in Science this week. Looking at the interactions with pollinators, and gymnosperms could solve a very peculiar puzzle.

The problem in dating the origin of flowering plants is not that it’s difficult. It’s that it appears to be relatively obvious. Looking at the fossil record, the earliest unambiguous flowering plants date from the early Cretaceous period. So it would appear that they date from the late Jurassic period at the earliest, about 135 million years ago.

To be more sure that the date is accurate, you would test with an independent method. Scientists have done this by comparing genetic differences. If you compare gene sequences, and work out the rate of mutation, you can calculate how long ago the most recent common ancestor existed. The problem is, if you do that, you get a date in the late Triassic period, roughly 200 million years ago.

So which date is it? Rather than picking a date they like, van der Kooi and Ollerton suggest looking at the creatures that angiosperms co-evolved with, their pollinators.

Image: Canva.

This looks like an effective tie-breaker. Van der Kooi and Ollerton point to research that shows butterflies and moths having a burst of evolution in the late Triassic – the same time that the genes suggest flowers evolved. But it’s not that simple. If genes are an effective way to date the origins of angiosperms, then they must also be an effective way of dating butterflies and moths. When you do that, it looks like the lepidoptera evolved in the Carboniferous period.

If there were pollinators or pollinator-like insects before flowers, what were they flitting between? Here, van der Kooi and Ollerton point to research on insect interactions with gymnosperms. Researchers have often made the assumption that gymnosperms were wind-pollinated, and that insect pollination was an angiosperm innovation. Jeff Ollerton says on his blog, this is not the case, and there is insect pollination of gymnosperms too. There’s evidence of thrips pollinating cycads, and similar for beetles and cycads. This would suggest that there was a pool of pollinators ready for angiosperms to exploit when they evolved.

A solution to the discrepancy in the dates has been put forward independently by Ofir Katz and Barba-Montoya and colleagues, who have both suggested a ‘cryptic’ period when angiosperms had evolved, but don’t obviously look different.

“The point about potentially long periods where taxa existed but were not dominant is an important one. This will have consequences for the fossil record (see Signor-Lipps effect). This then turns the question from when the first angiosperm arose to when they became dominant. But then again, what is ‘dominant’? And how does that relate to local (pollinator) fauna?” van der Kooi told Botany One

The Signor-Lipps effect is a comment on the rarity of fossilisation. Because it is unusual for an organism to be fossilised, it’s going to be very rare for the first, or last, example of an organism to appear in the fossil record. This would be consistent with molecular dates being earlier than fossil dates, but maybe not as much as the current gap.

While the flowers may not have fossilised, it’s possible their pollinators may have. Van der Kooi and Ollerton point to the research showing that flowers and their pollinators co-evolve. A change in pollinator assembly could reflect a hidden change in pollinated plants.

In modern gymnosperms that use pollinators, it’s scent that attracts the insect rather than sight. Van der Kooi and Ollerton note that scent does not fossilise, but as angiosperms use visual displays to attract pollinators, it’s possible they directed some insect evolution towards development of vision and visually-triggered behaviour.

However, while you can’t bury a scent, there may be some other ways to track an ancient odour. There may be genetic sequences associated with scent production. If there is a common ancestor, there might be some suggestion as to what plants were producing to attract the earliest insects.

Jeff Ollerton told Botany One, “As far as I’m aware no one has done a comparative study of gymnosperm and angiosperm scents. I think it would make for an interesting project.”

“It seems plausible though that pollinator attraction via visual cues has contributed to Angiosperm success.” said van der Kooi. The interaction of insects and flowers for signalling is the subject of a lot of current research.

While dating the origin of flowers might seem a purely botanical puzzle, van der Kooi and Ollerton’s article make clear that it’s an interdisciplinary issue. Given the job of most flowers is to attract pollinators, it’s going to be necessary to study those pollinators to understand how and when flowers evolved.