Exploring Tomato Flower Structure

Potato-type leaf

The world is full of gardening lore, some wise, some preposterous. Diane Ott Whealy, co-founder of Seed Savers Exchange, grew up knowing that you “harvest horseradish only in months with an ‘r’ in them,” and that “every day gets a ‘rooster step’ longer after the shortest day of the year” (Gathering 2011). The Farm Journal of 1885 suggests putting your boot alongside a melon and thumping them with your finger. If they sound the same, the melon is ripe.  One common belief among gardeners is that tomato varieties with potato-type leaves are much more likely to cross pollinate than regular leaf tomatoes. What is the basis of this wisdom? Is there any truth in it? I talked to tomato breeders Tom Wagner and Craig LeHoullier to find out.

To understand the origin of this belief, we must delve into the history of tomatoes. Charles M. Rick, known as the father of tomato breeding, studied and collected seeds of wild relatives of tomatoes in their native habitat of western South America. Rick observed many differences between wild and domesticated tomatoes. Besides obvious differences such as fruit size and growth habit, Rick noted subtle but important differences in flower structure. He recognized that in wild tomatoes, the female style (the slender tube that connects the stigma and the ovary) was very long, causing the stigma to protrude much farther from the flower cone when compared to domesticated varieties (Figure 1).

Regular leaf

Figure 1: The flower on the left has a protruding stigma (also called exserted) which sticks out past the flower cone because of the length of the flower style. The flower on the right has an inserted stigma, which is enclosed by the flower cone.

The domestication of tomatoes started in Central or South America and continued with breeding efforts in Europe and the United States, resulting in a gradual shortening of tomato flower styles.  In wild tomatoes, a long style facilitated cross pollination by native insects. Outside of their native region, tomatoes had far fewer insect pollinators, so their ability to self-pollinate was essential to produce fruit.  Fully inserted stigmas, which are enclosed by the flower cone, were inadvertently selected during breeding of mechanically harvested tomatoes because it ensured a reliable, concentrated fruit set (Cole 2007).  Another factor in decreasing style length was breeding for heat tolerance.  A study conducted by Charles M. Rick and Wesley H. Dempsey, from UC Davis, found that varieties with inserted stigmas were more heat tolerant above 35°C. All of these forces worked together to create modern varieties with short styles that were far less likely to cross pollinate than older strains that had long styles. 

At the same time as tomato breeding was shortening the length of the flower style, another genetic shift was taking place: the loss of the potato-leaf trait among varieties bred for commercial production. How did this happen? Neelima Sinha of UC Davis believes that, “potato leaf was just not a trait that was important to the commercial industry so it was probably just left behind” (phone call). Given that the potato-type leaf trait is genetically recessive to regular leaves and it was much less common than regular leaves to begin with, this theory seems plausible.

Figure 2: Normal tomato flower

However, legendary tomato breeder Tom Wagner suggests a number of hypotheses for why the loss of the potato-type leaf may have been deliberate. According to Wagner, the potato-type leaf was an oddity and did not fit the common notion of what a tomato plant looked like. Additionally, potato-type leaves may have been detrimental to plant health. Wagner says that potato-leaf plants are more prone to wind damage, fungal diseases, and uneven ripening, and the leaves obscure the fruit more than regular leaves. Today, the potato-type leaf trait represents an older class of tomatoes that was not subject to the forces working to reduce the length of the style.

Figure 3: Marigold-type tomato flower

Another reason that potato-type leaf varieties typically have exserted stigmas has to do with genetics. Potato-leaf varieties are relatively rare and comprise a narrow genetic pool with fruit that are typically large and convoluted. These so-called beefsteak tomatoes are the result of double-flowers, sometimes called fasciated or “marigold”-type flowers (Figure 3). By nature, these flowers have more exposed, exerted stigmas.  Hence, the potato leaf trait has become a convenient way to classify varieties that are more prone to out-crossing.

At Seed Savers Exchange, we have been collecting evaluation data on our tomato varieties for the last several years, including data on leaf type and style length. Of the 23 potato-leaf varieties that we have recently evaluated, 16 had exserted stigmas; 6 had stigmas level with the anther cone; and one had an inserted stigma (Figure 4). This supports the belief that the potato-leaf trait is correlated with flower structure, but shows that there are exceptions to this rule. Conversely, among the varieties with regular leaves, the stigma positions were inserted, exserted, and at the same level of the anther cone in roughly equal proportions, suggesting that no conclusion can be drawn about flower structure from this leaf type (Figure 5)

Although the potato-leaf trait may be a useful and convenient way to predict flower structure, there are exceptions. Among the potato-leaf varieties we evaluated, ‘Yellow Cherry’ had potato leaves and an inserted stigma, while several others had stigmas level with the anther cone. Another exception to the rule, according to Tom Wagner, is ‘Bloody Butcher,’ a potato-leaf type which produces red globe tomatoes and has inserted stigmas.  Craig LeHoullier, tomato breeder and author, believes that “the age of the strain and the size of the fruit” are better predictors of flower structure than leaf type (e-mail correspondence).  

Ultimately, the best way to determine flower structure is to actually look at the flowers themselves. It is important to note that flower structure is just one of many factors in predicting crossing rate and deciding how much to isolate a variety. Other important factors to consider are natural barriers between varieties, such as woodlands, and, especially, the number of insect pollinators that visit the flowers. Here at Heritage Farm, where varietal purity is of utmost importance, we have found that an isolation distance of 1000ft between tomato varieties almost always prevents crossing.  However, Seed Savers Exchange’s collection has many more old, heirloom varieties than modern varieties and the farm has a healthy insect population with areas of woodlands and varied terrain.  Every seed saver must assess these factors in their own garden or farm when deciding how much to isolate their varieties. Potato-type leaves as an indicator of flower structure is but one useful tool to help guide this decision. A much more reliable method is look at the flowers themselves.