Wetenschap - 24 november 2016

The ideal chrysanthemum

Roelof Kleis

How can you develop a chrysanthemum that not only looks good but also remains fresh for a long time in the vase? Until recently this was a question of cross-breeding and waiting to see what you got. But PhD candidate Geert van Geest is taking a more targeted approach with genetic markers. ‘A revolution in flower breeding.’

Text and photo Roelof Kleis

The Korte Kruisweg in Maasdijk is the site of a huge greenhouse complex belonging to Deliflor, the world’s largest supplier of chrysanthemum cuttings and chrysanthemum breeding company. Around 700 million cuttings roll off its conveyor belts every year, to be sent to cultivators all over the world. In addition to these production operations, the company also breeds new varieties. The scent of flowers fills the greenhouses used for research. The temperature is pleasant. And chrysanthemums are everywhere you look, in an inconceivable range of shapes, sizes and colours.

According to PhD candidate Geert van Geest, the son of a market gardener who was born and brought up in the area, Deliflor introduces a new variety once every two weeks on average. The company has over 300 varieties in its assortment. ‘80 percent of the new varieties will be dropped within a year,’ he explains. The market is unpredictable. The aim of all this innovation is to create a better flower. Better in the sense of more beautiful, more resistant to diseases and with better keeping quality. But what is ‘beautiful'? Van Geest: ‘That depends on who you ask. A big size is important for the Russian market. They value chrysanthemums more highly than roses. But in Japan, which is also an important market, the flowers have to be small and neat.’

Keeping quality

So tastes differ — but not when it comes to keeping quality. A long post-harvest shelf life is one of the key breeding criteria in the cut flower business. Van Geest has spent the last four years working on this. ‘The keeping quality of flowers in a vase is a difficult subject because so many processes are involved. On top of that, these processes are regulated by many different genes,’ he explains. ‘For example, leaves turn yellow when chlorophyll breaks down. Flowers become limp when their water regulation is disrupted and the water in the vase turns cloudy due to bacterial growth.’

Then there is the discoloration of the heart of the flower, the issue that Van Geest has been tackling. There was a good reason for his choice of subject as the heart turning brown is often the first sign of decay. Van Geest: ‘That’s when the trouble starts. If you can delay that discoloration, you’ll have made a big improvement in the post-harvest quality.’

‘Russians value chrysanthemums more highly than roses’

First, Van Geest developed a test to give a simple, reproducible quantification of the discoloration of the heart. To do this, he reduces the flower to a flower head on a short stalk, then places it in a bottle filled with water and stores it in cool, dark conditions. He then takes daily photos. ‘That lets you track the discoloration nicely. I’ve written some code to determine the r/g value — the ratio between the red and green pixels.’ The test is being validated at Deliflor as a way of assessing keeping quality in a vase.


But why do some flower heads turn brown faster than others? Van Geest guessed sugar might play a role. ‘Sugar water stops flowers from wilting so quickly. Take the Chrysal sachet that florists include with a bouquet of flowers. That consists mainly of sugar. Our hypothesis was that differences between varieties in discoloration could be explained by the amount of sugar being sent to the flower. Many chrysanthemums are transported over long distances and it can take a couple of weeks before they end up in a vase. They are transported in cardboard boxes at 4 degrees Celsius. There is no photosynthesis due to the lack of light, but respiration still continues. There comes a point when the flower hearts start to turn brown because they are no longer getting any sugar.’

That assumption turned out to be correct. Sugar content measurements showed a close relationship between sugar deficit and brown discoloration. What is more, discoloration was kept at bay when the flowers were put in sugar water. ‘So there is genetic variation in post-harvest sugar content,’ concludes Van Geest. ‘That content can differ by a factor of four to five. You can breed to optimize that.’


In addition to this work, Van Geest developed a gene map for the chrysanthemum. Not by sequencing the whole genome but by looking for markers that flag up the location of genes. Markers are bits of DNA with a known location on the chromosome. Van Geest has staked out the chrysanthemum genome with no less than 35,000 such markers. The trick then is to link desirable traits to these markers, he says. That will let you select for traits without knowing the precise location on the genome or how the gene involved functions. Van Geest found three markers that had a significant relationship with the flower head’s susceptibility to discoloration. Together, they explain 20 percent of that susceptibility.


Van Geest shows what that means in the greenhouse. Seedlings are currently only screened for traits such as ‘brown flower heart’ once there are enough plants. ‘That’s only after several months of vegetative propagation. Screening with markers can take place about six months earlier. What is more, we can screen the seedling for multiple traits at the same time, for example a brown heart and resistance to disease. Based on that genetic analysis, I can discard 80 percent of the genotypes straight away. That preselection means we can plant the entire greenhouse with genotypes that we know to have a set of positive traits. That is a huge improvement in the efficiency of plant breeding.'


The method Van Geest uses is known as ‘marker assisted breeding’. The trick itself is not new. ‘They’ve been doing this for ages in vegetable breeding. But it’s a revolution in chrysanthemum breeding. The method lets you work towards a particular result more quickly and directly through plant breeding.’

One reason why the flower industry is only just starting to work with markers is the fact that many flowers have very complex genomes. Chrysanthemums are hexaploid: they have six copies of each of the nine chromosomes. That makes calculations incredibly difficult in cross-breeding trials.

The application of genetic analysis in plant breeding will give Deliflor a huge advantage, says Van Geest. Other companies will probably follow suit. ‘That’s possible as the method will be made available. But Deliflor has the genetic map and the marker codes.’ But first he has to get his doctorate. Only then will Van Geest be putting his findings to use in Maasdijk. The ideal chrysanthemum is on the way.