Do flowers produce scent when they’re not expecting visitors?

When a plant can self pollinate, it shouldn’t need to work to attract pollinators to its flowers. A new study of Arabis alpina, by Hampus Petrén and colleagues in Lund and Uppsala, has revealed it’s not a simple change. They have found that self-pollination does affect scent. However, self-pollination doesn’t lead to a drop in all forms of floral signalling. The findings, published in Annals of Botany, have implications for studies of pollination-shifts in isolated plant populations.

Arabis alpina is Alpine rock-cress, a small plant distantly related to cabbages and mustard. It lives in the mountains of Europe, Africa and Asia, where it grows in rocky habitats by mountain streams and on rock ledges.

It’s a useful plant to study for shifts in fertilisation, as it varies in how fussy it is. Populations in central Italy and Greece are self-incompatible. This incompatibility means to reproduce they need pollen from another plant. They call this in by attracting a wide variety of visitors. These could be Diptera (flies), Hymenoptera (bees or wasps) or Lepidoptera (moths and butterflies).

In contrast, A. alpina in northern Italy, France and Switzerland is self-compatible. If a suitable pollinator doesn’t visit then a flower can use its own pollen to pollinate itself if it must. If you go further north to Scandinavia, then you find fewer pollinators. Here, A. alpina is happier about self-pollination.

It looks like some populations of A. alpina are undergoing a shift in pollination. As they become more capable of self-pollination, the less they rely on insects visiting. This shift should mean they don’t need to put so much effect into attracting visitors. To find if this was true, Hampus Petrén and colleagues set up experiments to see how the floral signals differed between plants.

The team set up a common garden in a greenhouse at Uppsala University. The greenhouse would control the temperature and lighting for all the plants. That way, differences in scent production wouldn’t be due to different conditions in various locations. The plants included self-incompatible plants from Greece and Italy. The authors compared these against self-compatible plants from Scandinavia. Between these two groups, there was a third population using plants from France and Spain. This group was between the Greece/Italy and Scandinavian groups. They could self pollinate but preferred not to. In total, the scientists grew almost six hundred plants.

The botanists measured flower size and scent production.

As they expected, when plants could self pollinate they produced smaller and less scented flowers. But things were very different for the two types of self-compatible plants. Or rather, they weren’t.

The Scandinavian plants produced similar flowers to the France/Spain group. The fact that the Scandinavian plants were more likely to self-pollinate made no noticeable difference. Both plants had similar sized flowers. They also made a similar effort in producing scent. It was a bit of a puzzle. You would typically expect floral scent to be related to the activity of pollinators. If there’s not as much activity, then why would they produce the same scent?

“[F]loral scent composition varied considerably both among and within the three mating system categories,” write Petrén and colleagues. “As scent was collected from plants growing in a common environment, the results indicate that among-population differences in scent emission rate and composition have a genetic basis.”

“Our results therefore suggest that the evolution of floral scent is driven not only by the need for pollinator attraction, but is instead potentially the result of a complex set of factors including selection, population history and genetic drift.”