A plant scientist finds beauty in floral arrangements.

On the face of it, flower arranging is a fiddly affair, and its underlying rules are not immediately obvious to the beholder. But a plant's flowers are always arranged in one of three basic architectures, or 'inflorescences'. These take the form of panicles, loosely but highly branched clusters in which each flower has its own stalk (as in the foxglove); racemes, in which flowers are arranged individually along an unbranched, growing stem (the snapdragon); or cymes, typified by a cluster of branches at the end of a stem that each terminate with flower (the forget-me-not). Simple rules must lie behind this, and simple rules are the foodstuff of mathematical models.

That is the logic behind the work of Przemyslaw Prusinkiewicz at the University of Calgary in Alberta, Canada, and his colleagues. Last year, they published a model in which they imagined that meristems grow into shoots or flowers according to the value of a factor that they named 'veg' (P. Prusinkiewicz et al. Science 316, 1452–1456; 2007). When veg is high, a shoot springs forth; when it is low, a blossom flourishes. Thus, if over time veg decreases at the same rate in all of a plant's growing tips, the model grows a panicle. Other simple rules give rise to a raceme or cyme.

Prusinkiewicz et al. found that, in Arabidopsis, a gene called LEAFY influences the value of veg. But how does this concept apply to plants with different architectures? Recently, Erik Souer of Vrije University in Amsterdam and his collaborators showed that modification of LEAFY activity is crucial for floral architecture in petunia, a cyme, just as the model predicts (E. Souer et al. Plant Cell 20, 2033–2048; 2008). They identify a protein that activates LEAFY only in developing flower buds and that is essential for their architecture. I find the tidy simplicity of these findings more beautiful than any bouquet.

Discuss this paper at http://blogs.nature.com/nature/journalclub