485 McCormick Road
P.O. Box 400328
Charlottesville, VA 22904

  • Facebook - Black Circle
  • Twitter - Black Circle

Deep Dive: A Colorful Story of Mosaic Evolution from the Galloway Lab

Tan (left) and purple (right) pollen in Campanula americana. The flower displays pollen via secondary pollen presentation on the outer surface of the style. Photos courtesy of Laura Galloway.

Written by Sarah McPeek

On top of Salt Pond Mountain, Campanula pollen is vibrant purple. In the New River valley below, Campanula pollen is faint tan. Why is pollen purple on top of the mountain and tan at the bottom?

The Galloway lab’s forthcoming paper in the Journal of Evolutionary Biology authored by Matt Koski, a recent post-doc in the lab and now a faculty member at Clemson University, is the latest in a series of experiments exploring how such a beautifully simple field observation can have a surprisingly complex natural history.

A single flowering field of tall American bellflowers, Campanula americana, present pollen ranging from white to tan to light purple to dark purple. In a previous study, the lab surveyed this colorful variation across the flower’s entire range. They found a striking geographic pattern: populations from the warm, wide plains of the Midwest display darker pollen colors than populations from the cooler, mountainous climates of the East.

To investigate how different pollen colors perform in different climates, Matt exposed the different pollen colors to warmer and cooler temperatures. Sure enough, darker pollen grains germinated better than lighter grains under high temperature stress, suggesting that natural selection against light pollen may explain the abundance of dark pollen in the warm, western parts of the range.

Matt wondered whether these physiological differences could be due to the underlying biochemistry. He engaged Andrea Berardi, a UVA PhD with Doug Taylor and past researcher at Mountain Lake Biological Station, in a study of Campanula’s pigments. Pigments may engage in a variety of plant functions from heat and UV protection to herbivore defense and regulation of reproduction. To assemble their pollen color palette, Matt mated plants with dark purple pollen from Oklahoma to plants with white pollen from Ohio. He then allowed their light purple offspring to self-fertilize in the greenhouse, creating a third generation of plants with pollen colors mirroring the rainbow of naturally occurring variation.

Matt Koski’s Campanula pollen color palette. Photos courtesy of Laura Galloway

Matt and Andrea’s chemical analyses showed that purple pollen is richer in anthocyanin pigments that aid in heat tolerance, further supporting dark pollen’s advantage in warmer climates like sunny open plains. Still, the lab was puzzled by a larger question: if darker pollen has such a strong advantage in warmer climates, why is some pollen light?

Sometimes, variation in a trait persists when the trait shares a genetic basis with other traits that experience direct selection. Many plant pigments develop from the same core biochemical pathways. Selection on traits that affect a plant’s attractiveness to pollinators, like petal color, could drag the synthesis of certain pollen colors along, maintaining both light and dark pollen in warm regions despite dark pollen’s temperature advantage.

Surprisingly, the lab found no discernible links between Campanula pollen color and petal color. “I was totally shocked,” Laura recalls. Such a strong signal of trait disintegration indicates that pollen and petal traits are selected for and inherited independently. What began as a simple biochemical investigation had suddenly bloomed into a much more colorful story of mosaic evolution.

During his extensive trait measurements, Matt observed that dark purple pollen grains were larger than lighter pollen grains. This gave him the idea to return to the greenhouse and measure more pollen and plant traits. Could other patterns explain Campanula’s wide spectrum of pollen colors?

Polymorphism, or trait variation among individuals in a population, can also prevail if there is no overall selective advantage for one morph over the other. That is, darker pollen may be more heat tolerant, but if this ability does not greatly affect the plant’s reproductive output, then both light and dark pollen plants can grow together in the same environment.

Luckily, the lab still had some greenhouse plants left over from their biochemistry experiments. They developed a third generation of experimental plants by mating light pollen mothers and dark pollen mothers with light pollen fathers and dark pollen fathers. This breeding design allowed them to examine the individual and combined effects of maternal and paternal pollen color on offspring traits.

Sure enough, dark and light pollen offspring from each of these crosses all had similar overall performance, measured as the cumulative product of how many seeds germinated, how big plants grew, and how many plants survived to flower.

But this doesn’t mean that pollen color has no effect on fitness – quite the opposite! Instead, plants with different pedigrees achieve their successes through very different means. Light pollen mothers produce more seeds and flowers than dark pollen mothers. At the same time, dark pollen fathers produce pollen grains that are larger and more robust, germinating at faster rates with greater success than light pollen. Hence, light pollen may have a maternal advantage while dark pollen has a paternal advantage. Together, these parental effects balance out for equal overall reproduction, explaining how color variation could be maintained in wild populations.

Now the lab has a new buzzing question: why do light pollen and dark pollen plants need different pathways to reproductive success? Matt’s work confirms that temperature is certainly a contributing factor to color variation across Campanula’s range. Nevertheless, these latest greenhouse experiments reveal that multiple aspects of the environment are likely at play in the evolution of pollen color in nature.

Laura believes that Campanula’s community of insect pollinators may be a key piece of this natural selection puzzle. While many flowers keep their pollen tucked away inside the flower’s center, Campanula displays its pollen on its long bristly style. Although secondary pollen presentation makes pollen more accessible for transfer between flowers, it also carries risks with some pollinator species. Bombus bumblebees efficiently gather pollen and move it between plants, but Megachilid bees nibble more pollen than they transfer. For these ‘pollen robbers,’ the pollen-coated style is “like a buffet,” Laura says.

Recent work with lab collaborators also shows that Megachilids prefer purple pollen. This suggests that these specialist pollinators could differentially affect the success of tan and purple pollen morphs in the wild. Thus, the local pollinator community could be a driving selective agent on the maternal and paternal differences between color morphs. In the next phase of putting the puzzle together, the lab plans to return to Salt Pond Mountain and study Campanula pollination in the wild.

Pollen is purple on top of Salt Pond Mountain and tan at the base. Koski et al. 2020 tells a compelling story of how elevational changes in climate, differences in pollinator species abundances, and other environmental pressures could combine to create this beautiful and vital natural variation. The Galloway lab’s ongoing research into many aspects of variation in Campanula affirms the benefits of incorporating the full richness of a species’ natural history for understanding the evolution of floral traits.

A Megachilid bee gathers purple pollen from the style. Photo courtesy of Laura Galloway.

Links to studies referenced:





Links to Matt and Andrea’s professional websites, as well as the Galloway Lab: