The Loom

Autumn Leaves: The Search for Purpose

i-14979b3b8a337a054a25a8f350a5879c-leaf.jpgAs the autumn leaves turn handsomely, I’ve been wondering, why do trees bother? It’s a question scientists have been asking for the past few years, and for the first time, they’ve carried out an experiment to find out.

The color of an autumn leaf can actually take a lot of work. In the fall, the green chlorophyll in a tree fades away, while the tree actively builds new pigments to turn it red or yellow. It’s generally agreed that these colors must serve some function for trees. Otherwise, natural selection would favor drab trees that dropped their leaves without such bother. They could use the energy they didn’t waste on autumn colors to fight diseases, capture more sunlight, or some other essential task.

In 2001, the late biologist William Hamilton made a provocative proposal: he argued that leaves turned colors in the fall to warn off insects. Healthy trees could produce anti-insect toxins and have energy to spare for building red and yellow pigments. The stronger the tree, the brighter the signal. As I described here, Hamilton and Sam Brown of the University of Texas found support for his hypothesis by comparing different species of trees. Species with bright leaves tend to be attacked by more species of aphids in the fall than species with drab leaves. This correlation was consistent the idea that the evolution of bright leaves was driven by feasting bugs.

But skeptics have argued that autumn colors actually serve a different function. They propose that the colors act as a sunscreen. An autumn leaf actually bustles with molecular preparations for the winter. Leaves pump their nutrients into a tree’s branches, where the tree can use them to survive till spring. But the cold, short days of fall play havoc with leaf chemistry. The sun’s UV rays cause more damage to the leaf tissue, as to harmful charged molecules released by photosynthesis. Some scientists argue that autumn pigments absorb these dangerous rays and molecules, allowing the leaf to do its autumnal business. (For more on the debate, see my blog post and the article I wrote for the New York Times.)

Scientists have continued to come up with new ideas about autumn leaves and to make new observations to test them. But no one has carried out a full-blown experiment. It’s easier said than done. Trees, after all, will not be rushed. Scientists must wait patiently for their leaves to turn in the fall. If they haven’t gathered enough observation by winter, they have to wait another year for another chance. It’s also difficult to reduce the glory of an autumn leaf to a precisely controlled variable, in order to measure the strength of its connection to the behavior of insects.

H. Martin Schaefer and Gregor Rolshausen, two scientists at the University of Frieburg, recently put autumn leaves to the test by painting trees. They found a stand of ash trees, and painted the leaves of some of the trees red and some green. If the colors served as a signal to the insects, the paint should make a difference.

It didn’t.

Aphids did not alter their attacks on trees after they were painted. Nor did painted trees get attacked at a different rate as unpainted ones. In a paper published today in the journal Biology Letters, Schaefer and Rolshausen argue that their experiment doesn’t support Hamilton’s signal hypothesis.

Interestingly, the scientists found that the aphids did prefer to land on some trees over others. The most heavily attacked trees turned out to be the ones producing the most seeds. It’s possible that the trees that produce lots of seeds have to sacrifice some of their defenses against insects, making them good targets for aphids. The aphids would somehow have to detect the seed-heavy ash trees–either by seeing the seeds themselves or perhaps by smelling some compound given off by the trees. In either case, the aphids don’t care about the color of the leaves.

Schaefer and Rolshausen also tested the various theories for autumn leaves by comparing different populations of ash trees near Frieburg. The scientists found an association with cold weather and both bright colors and few aphids. Their findings, which are in a paper in press at Plant Ecology, may favor the sunscreen hypothesis over the signal hypothesis. But there’s a catch. Some of the trees grew on a mountainside at 1100 meters, and others grew at 360. If leaves need sunscreen to withstand harsh environments, the high-altitude trees might have been brighter. But they weren’t.

So for now, the fall foliage remains a mystery. The slow life of trees will mean we have a long time to wait for a resolution. At least we can enjoy the colors in the meantime.


Aphids do not attend to leaf colour as visual signal, but to the handicap of reproductive investment. H. Martin Schaefer and Gregor Rolshausen. Biology Letters, 2006. DOI:10.1098/rsbl.2006.0548

Do aphids paint the tree red (or yellow)–can herbivore resistance or photoprotection explain colourful leaves in autumn? Gregor Rolshausen H. Martin Schaefer. Plant Ecology in press. DOI: 10.1007/s11258-006-9215-3


  1. #1 Nick (Matzke)
    October 17, 2006

    Interestingly, the scientists found that the aphids did prefer to land on some trees over others. The most heavily attacked trees turned out to be the ones producing the most seeds. It’s possible that the trees that produce lots of seeds have to sacrifice some of their defenses against insects, making them good targets for aphids.

    It is also the case (at least in my experience) that sick/dying trees often put their last gasp into a bigger seed crop — the idea being, “reproduce before you croak.” At least, this is what some forester told me once and I have seen this apparently occuring with fir trees on the West coast many times. I’m sure someone has researched this somewhere…

  2. #2 TAW
    October 18, 2006

    Maybe the aphids can smell the pigments?

  3. #3 Tristram Brelstaff
    October 18, 2006

    Nick (Matzke) wrote:

    It is also the case (at least in my experience) that sick/dying trees often put their last gasp into a bigger seed crop — the idea being, “reproduce before you croak.”

    This can also be seen as the genes abandoning the sinking ship.

  4. #4 Michael
    October 18, 2006

    It’s the University of Freiburg. 🙂

  5. #5 Lab Cat
    October 18, 2006

    Is it possible that the red, yellow etc. pigments allow the leaves to more effectively use the sunlight to photosynthesize?

    I think I read this hypothesis somewhere.

  6. #6 CCP
    October 18, 2006

    I’m not a botanist, but I play one in my intro bio course…
    My understanding is that the yellow carotenoids do act as “antenna pigments” to absorb photons that chlorophyll does not, that these photons can be used to run photosynthesis, and that much of the yellow in fall leaves results from the withdrawal of chlorophyll from leaves (to be saved and recycled), leaving (ha!) the carotenoids behind to be revealed. However, the orange and red autumn pigments appear to be synthesized anew at this time, when photosynthesis is shutting down. This is what needs explaining.
    Is there a cogent counterargument to the sunscreen idea? it makes sense to me. For one thing, trees tend to turn red from the top down, with lower branches only reddening as higher leaves are lost, i.e. when they are exposed to more radiation.

  7. #7 Adrian Paterson
    October 18, 2006

    Here in New Zealand we have an added spin on this as virtually none of the tree species here change their leaf colour and drop their leaves in Autumn. And yet we are a cool temperate country with plenty of lineages that do drop their leaves in other parts of the world. We also have a diverse range of insects. Given that we also live at the edge of the ozone hole one would think that sunscreen would be even more important here than in the northern hemisphere!

  8. #8 Ross
    October 18, 2006

    The New Zealand situation is different from that of the northern hemisphere because our soil quality is low. Trees don’t usually lose their leaves because the nutrients are not easily available ot replace them. The ozone hole is a feature that has developed in just the last few decades – nowhere near long enough for evolution to act. Refer this article on the RSNZ site:

  9. #9 Jeff Chamberlain
    October 18, 2006

    “Species with bright leaves tend to be attacked by more species of aphids in the fall than species with drab leaves.” Why is this considered support for the Hamilton warning signal hypothesis? I’d have thought the opposite, that if the signal hypothesis was true we’d expect fewer insect attacks on brighter leafed trees (i.e., that the signal worked).

  10. #10 Fred
    October 18, 2006

    Not all trees drop their leaves, so one might expect from an evolutionary point of view, leaf dropping in Autumn is a relatively recent phenomenon (once those pesky dinosaurs were out of the picture). Flowering plants are a relatively recent evolutionary trend, associated with a shift in insect lifestyle. Does anyone know if the period of evolution of leaf dropping correlates with the evolution of flowering plants? Perhaps the two phenomenon are a response to a shift in insect ecology at around the same time?

  11. #11 Ben Chase
    October 19, 2006

    One of the more ancient deciduous trees, the gingko tree, turns yellow in Autumn.
    Perhaps turning yellow came first, and only later did a deciduous tree gain the trick of turning orange or red or purplish?

    Yellow poplar (Liriodendron spp.) also turn (duh) yellow, though their lineage may be substantially less ancient than gingkos. But perhaps there is something to be gained by trying to discern an overall ecological/physiological difference between trees that turn yellow and trees that turn other colors?

    (And in comparison, what does it mean to be a magnolia or holly, festooned with thick, waxy, green leaves all through the winter?)

  12. #12 Ben Chase
    October 19, 2006

    And another thought, about the yellow poplars, which turn yellow. They turn yellow and drop their leaves a bit sooner than the trees which turn orange and red. Do the orange and red pigments provide some ability for the leaves to be useful longer, perhaps in the face of early freezes? (This is too obvious an explanation; surely somebody has already thought of this explanation and disproved it.)

  13. #13 Tom Kimmerer
    October 19, 2006

    As a tree physiologist, I am always a bit skeptical of adaptive stories about fall colors. They sound a lot like the “Just So Stories” that Stephen J. Gould warned about. Not everything in nature is adaptive, and that is especially true of events at the end of life.

    Trees throw away leaves like Kleenex, but not until the tree translocates the good stuff (nitrogen) into the perennial parts. As chlorophyll degrades and its nitrogen moves out, the fading green reveals other colors. Leaves always contain carotenoids (yellow) as accessory pigments in the chloroplast. The leaf synthesizes some anthocyanins (reds and purples) de novo, but they only become apparent as the vacuole becomes acidic perimortem.

    Trees vary in their genetic capacity for anthocyanin synthesis for reasons that may have little to do with fall colors. Black maple lacks anthocyanins and turns yellow, while sugar maple produces anthocyanins, but vacuole pH only declines in full sun. Hence sugar maples are a mosaic of yellows to orange while black maples are always yellow.

    Aphids are primarily seeking nitrogen, excreting most of the tree’s carbohydrates as honeydew. Hard to imagine that they would selectively seek out particular trees or leaves just when nitrogen content is declining most rapidly.

    The sunscreen argument assumes that there is something to protect. Since most of the nitrogen is out of the leaf by the time colors appear, and the leaves are in advanced senescence by that time, sun protection doesn’t seem very important.

    Trees with an abundance of nitrogen, like black locust, don’t bother to translocate nitrogen, dropping green or greenish leaves. The biochemical state of a leaf when it is finally dropped may have adaptive significance for nutrient cycling. For example, sugar maple leaves with low concentrations of polyphenolics degrade more quickly than the drab brown oak leaves, cycling nutrients more rapidly.

    As for the cost argument, the assumption that there is a net cost to color production is unfounded. Caretenoids are constitutive, and anthocyanins reflect modest changes in vacuolar pH.

    In the aphid and sunscreen hypotheses, color me skeptical.

  14. #14 Tom Kimmerer
    October 19, 2006


    I did not intend to imply that anthocyanins and carotenoids are not ‘sunscreens.’ They most certainly are. But the sunscreen protection is important at other parts of a leaf’s life history, not at the time of death.

  15. #15 Martin Schaefer
    October 20, 2006

    I fully understand that tree physiologists are skeptical of the adaptive stories of fall colouration. However, pigments often serve multiple physiological and ecological roles. There is now evidence from different tree species that aphid prefer to colonise trees with less intense fall colouraton. Obviously, this evidence is based on correlations and thus fails to establish a causal relationship that the co-evolutionary theory assumes. That is why more experimental approaches are needed.
    To try to answer some of the questions that arised:
    The ash trees in the study were all relatively young small trees that were not in a phase of ‘terminal investment’ in fruits. Still, aphids colonised preferentially trees with higher fruit crops, most likely because more nutrients are transported to the fruits which ripen simultaneously to changes in fall colouration. In this case, aphids apparently used no visual cues in their host choice.
    It is unlikely that aphids can smell plant pigments because these are not votile and stored as explained by Tom Kimmerer. At present, it is conceivable that plants emit votiles that correlate with leaf colour or other biochemical changes in the leaf during autumn.
    There are some studies that show that the red anthocyanins protect the photosynthetic apparatus of the plant during cold temperatures. This and other physiologically-oriented hypotheses seem now most likely to explain leaf colour changes although we clearly need more studies to rule out other roles, such as signalling to herbivores.

  16. #16 Tom Kimmerer
    October 20, 2006

    Martin – Thanks for your comments.

    What do we know about visual acuity in aphids? Japanese beetles have a strong preference for trees with red or purple leaves. Can aphids distinguish foliar colors? I agree that volatile signals related to color are possible (I have published a number of papers on tree volatiles and insect response).

    Secondly, I wonder about timing. Here in Kentucky, most of the peak in fall colors happens when it is too cold for aphids to be dispersing. So are they cuing on something that precedes coloration?

    I still maintain my skepticism, but it seems that understanding the cues to which aphids appear to be responding might help us figure out whether this is a real phenomenon.

  17. #17 Paul Decelles
    October 21, 2006


    Do the trees really actively make the red pigments or are they a spontaneous chemical reaction that doesn’t require any work on the part of the tree? If the former than this would strongly suggest that there really is some adaptive significance to the fall colors.

  18. #18 Martin Schaefer
    October 23, 2006

    Tom -We know little about the visual acuity in aphids. What we know is that they perceive yellow and some species are attracted to yellow surfaces (which is often used for catching them in pest control). From work done by others I know that the only aphid species that has been studied is not sensitive to red light. This would explain why they did not react to colour. But there might also be differences between species.
    Here in Europe the timing seems pretty acurate. Several aphid species migrate during leaf colour changes. However, for the skeptics among us, there are obviously a number of other aphid species which migrate earlier or not migrate at all.

    As for the production of red pigments: the trees produce them actively, but there are a number of adaptive explanations that do not include defence against aphdis (e.g., protection of photosynthetic tissue and reducing oxidative stress).

  19. #19 mark
    November 14, 2006

    I wonder what, if any, relationships exist between fall leaf color and the color of flowers and fruits.

  20. #20 Carolyn
    December 4, 2006

    So how does any of this explain why oaks and beeches hold their dried-up copper and brown leaves for most or all of the winter?

  21. #21 Steviepinhead
    December 5, 2006

    First, let me say, Carolyn, that I have absolutely no idea!

    But I rather doubt that whatever evolutionary–or other–mechanisms explain the change of color in leaves would directly explain the retention of leaves.

    The latter sounds more like it might have something to do with providing/retaining a certain type of nutrition–or conceivably, parasite or disease prevention–around the base of the tree in the spring which might be lost if the leaves were released to the whim of the winds, snow, and weather in the fall or winter. I guess insulation might also be a possibility–I don’t know if leafless oaks and beeches are possibly more cold-susceptible than, say, maples and birches…

    (It’s possible, I suppose, that the reverse is true: that is, that leaves retained over the winter are less likely to contribute to the soil structure or chemistry around the base of the tree, depending again on the affect of wind, weather, etc. This would flip my speculations around, but I would suspect some of the same factors might be involved, but with the “signs” reversed.)

    While the “parasite” notion sounds superficially similar to some of what has been discussed regarding leaf color changes, I suspect that diseases or parasites affecting the base of the tree–where I’m assuming the winter-retained leaves preferentially wind up (or preferentially don’t wind up, see the parenthetical paragraph…), with some sort of differential impact on the soil structure or chemistry–would be different from those attacking the trunk and branches.

    These are, again, the merest speculations. Scientists who actually know something about trees and leaves would have to pursue these and other notions, filter them through what is known about leaf-detachment mechanisms, the species involved, etc., etc., and then decide which speculations might fruitfully advance to the testible hypothesis stage…

  22. #22 Jeremy
    December 12, 2006

    From someone who knows nothing about this subject, I will throw in a couple of unresearched ideas I’ve been pondering since reading about this some time ago. Here are two hypotheses:

    1) Isn’t chlorophyl broken down and reabsorbed by the tree? Maybe this triggers the production of the yellow pigment. And maybe the yellow pigment, has nothing to do directly with photosynthasis and is really a trigger to cause the leaf to detatch from the tree.

    2) Maybe the photosynthesis conducted by the yellow pigment isn’t to support the tree, but maybe some process within the leaf itself. Perhaps it aids the leaf in drying out, therefor minimizing the risk of rotten green follage at the base of the tree.

New comments have been disabled.