Genetic variation is the basis for selection and thus for evolution. Demonstrating genetic variation in a wild population of long-lived non-model species, simultaneously with the responses of a biological selective agent under field conditions, is a difficult feat. In this issue of the New Phytologist, Sinkkonen et al. (pp. 461–469) describe the relationships between genetic variation in yellow autumn leaf colour of 19 Finnish Betula pendula genotypes and the abundance of phloem-sucking autumn migrating aphids. They show that autumn leaf colouration varies significantly among genotypes of silver birch trees. Genotypes that expressed the strongest leaf reflectance in early autumn harboured more egg-laying females of the specialist aphid Euceraphis betulae than those expressing strong leaf reflectance several weeks later. Moreover, the aphid community size in the beginning of the autumn colouration season explained nitrogen (N) loss by the litter of the different genotypes. This study demonstrates the potential for evolution through plant–herbivore interactions in trees by selecting for genotypes that begin leaf senescence later in the short autumn season. However, insects are not the sole selective agent.
‘… it became clear to all involved scientists that red and yellow autumn leaves should be treated separately concerning plant–herbivore interactions.’
The spectacular phenomenon of yellow autumn leaf colouration (Fig. 1), most common in temperate and boreal regions of northern Europe, has become a hot topic in recent years. For a long time, the majority of plant scientists considered yellow and red autumn leaves to be an unavoidable product of unmasking of these colours when chlorophylls are degraded during senescence. The cornerstone of the whole process of autumn colouration is to re-translocate from the leaves as much of their valuable resources as possible before they are shed, especially amino acids (nitrogen), phosphorous and magnesium. For that, the photosynthetic system and other cellular components are degraded, turning their nutrients into easily transportable compounds and therefore also into easily accessible resources for sucking insects such as aphids.
Plant physiologists discussed the anti-oxidative and photoprotective roles of anthocyanins in red autumn leaves a very long time ago. However, these functions were not widely considered to have much biological significance, and thus were not emphasized in textbooks and teaching. The first decade of the twenty-first century was the arena for an explosion of interest, theoretical discussions (including some very hot debates), modelling and some field and laboratory work on the adaptive physiological and anti-herbivory functions of red and yellow autumn leaves. This wave started with physiological studies showing that the red pigment (anthocyanin) in autumn leaves is produced de novo (Matile, 2000; Hoch et al., 2001). Simultaneously, the co-evolutionary hypothesis was proposed by the school of the late evolutionary biologist Hamilton (Archetti, 2000; Hamilton & Brown, 2001; Archetti & Brown, 2004). In the beginning, the co-evolutionary hypothesis did not distinguish between red and yellow autumn leaves, but rather treated both as ‘bright colours’ that signal insects, especially aphids, about the chemical defence of the trees in accordance with Zahavi’s handicap theory (Zahavi, 1975). Moreover, Archetti (2000) posited that the co-evolutionary hypothesis does not equal aposematism. Very soon, a strong opposition to the co-evolutionary hypothesis emerged, especially as it was well known that aphids are attracted to yellow (Holopainen & Peltonen, 2002; Wilkinson et al., 2002); hence, yellow colouration was unlikely to have evolved as an aphid deterrent signal. A number of additional objections on various theoretical and experimental grounds were also raised (e.g. Schaefer & Wilkinson, 2004).
The issue of colourful autumn leaves is indeed very complicated, including several physiological aspects as well as various simultaneously operating anti-herbivory mechanisms. Colourful autumn leaves are formed by thousands of tree and shrub species, involving an enormous number of herbivorous insects, their predators and parasitoids, fungi, bacteria and viruses, and no single explanation can cover all of this. After several years of adding new physiological and anti-herbivory hypotheses, and of hot debates, a conference was arranged at Oxford University, UK (see Ougham et al., 2008), and a review aiming to moderate between views stressing physiological vs ecological (anti-herbivory) functions of autumn leaf colours followed shortly afterwards (Archetti et al., 2009). At that time, it became clear to all involved scientists that red and yellow autumn leaves should be treated separately concerning plant–herbivore interactions. This separation is supported on the one hand by the attraction of aphids to yellowing leaves. On the other hand, the anthocyanins of red leaves are strongly associated with the biosynthesis of toxic substances and protease inhibitors, and also act as anti-oxidants (e.g. Gould, 2004). Both red and yellow autumn leaves may function similarly, though, in undermining insect camouflage (Lev-Yadun et al., 2004), signalling that the leaves are going to be shed soon (Lev-Yadun & Gould, 2007), or in aposematism (e.g. Lev-Yadun & Gould, 2007; Archetti et al., 2009; Lev-Yadun & Holopainen, 2009). The increasing understanding of overlapping functions of autumn leaf colouration, and of the complicated biological interactions involved, relaxed the tensions between authors of contrasting hypotheses (e.g. Lev-Yadun & Holopainen, 2011; Cooney et al., 2012). It became clear to all that solid field and experimental data concerning the many aspects of autumn leaf colouration are needed. Basically, Sinkkonen et al. provide such desired critical genetic, N content and herbivore abundance data from the autumn leaf colouration biology of a wild tree species with a very broad geographical distribution. Somewhat similar, but less detailed data have so far been presented only for domesticated apple varieties (Archetti, 2009). Data of the type presented by Sinkkonen et al. allow us to see the potential for natural selection and therefore evolution because they include timing and level of N stress posed by the aphids on several tree genotypes, which are difficult to measure in trees. This significant progress should be only the opening shot in a long list of such studies in many other taxa, which would ascertain that we have a rule rather than a species-specific adaptation. Moreover, when the basics for silver birch are known, the obvious next step should be studying the genomic, proteomic and metabolomic aspects of both the trees and the aphids. The expression of carotenoids that gives leaves their yellow colour is controlled in many cases by epigenetics (Cazzonelli, 2011). It will be very interesting to learn whether repeated aphid attacks can induce an epigenetically-based shift in the timing and rate of autumn leaf-colour change and whether aphids can control leaf epigenetics for their benefit for > 1 yr, like insects’ control over galling host plants. Since fungal attacks that change leaf colour, nutrient values and chemistry are common during yellow autumn colouration in birch, this aspect should also be studied. The aphids’ preference for early colour-changing leaves may bear not only on their nutritional/reproductive biology, but also on the involvement of a higher trophic level, predators of these aphids. In regions where many predacious birds have to migrate sometime in autumn, and predacious insects that may consume aphids also have to prepare for the approaching freezing winter, timing is critical, and follow-up studies of the fate of the trees, aphids and N in relation to predation are needed.
The current significant global warming is inducing phenological changes in both plants and animals (Root et al., 2003). This warming trend is expected to continue, and in turn to further modify phenology. Thus, the whole phenomenon of autumn leaf colouration is expected to change its dates and possibly also its duration. The study by Sinkkonen et al.is therefore also important as a point of reference for future studies that will show the trajectory of selection under warming climates, a situation that occurred several times during the Pleistocene. In this respect, long-term phenological observations are critical for further research of birch–aphid interactions. The aphids’ choice of trees that turn yellow earlier in the autumn may select for genotypes that turn yellow later. This evolutionary trend is limited by the timing of prevalence of low temperatures that block re-translocation of essential nutrients from the leaves. The global warming trend may allow trees to prolong the re-translocation period and aphids to prolong their attacks. In parallel, aphids’ parasitoids and predators may also shift their seasonality to exploit the new opportunities. Many consecutive years of field documentation are needed in order to realize the direction and rate of evolution of such scenarios. This should be done in view of the many alternating cooling and warming episodes that have occurred during the Pleistocene and shaped the current adaptations of the trees, the insects and their parasitoids and predators.
The anti-herbivory aspects of autumn leaf colouration are part of a much broader awakening research field of defensive plant colouration. As the late J. L. Harper posited long ago (1977, p. 416), ‘botanists were reluctant to accept things that are commonplace for zoologists and often seem reluctant to see the animal as a powerful selective force in plant evolution except in the curiously acceptable realm of adaptation to pollination! It may be that much of the fantastic variation in leaf form, variegation, dissection and marking that is known in the plant kingdom is accounted for by the selective advantage to the plant of associating unpalatability with a visual symbol’. Since the year 2000, this absurd situation seems to have come to an end by the continued efforts of various research groups in many parts of the globe. However, it will take a very long time until the level of understanding of defensive plant colouration equals that of animals. The study of defensive plant colouration is certainly an intellectual and experimental challenge, but this is not the whole story. Because many plants that use defensive colouration are backed by defensive chemistry, defensive colouration may be a cue for hidden chemical and molecular bonanzas that can be exploited probably not less (if not more) than plants used in traditional medicine. This possibility has not been exploited regularly yet.
Theoretically, one of the potential solutions of decreasing aphid attacks is to reduce the yellow component of leaf colour, thus making them less attractive to aphids. However, there are probably some physiological constraints on reducing leaf yellowing since the trees may have difficulties in getting rid of all the carotenoids. Two facts support the possibility of reducing carotenoid levels and thus their cuing to aphids. The first is that a considerable amount of carotenoids is degraded and parts are converted to volatiles emitted to the atmosphere (Keskitalo et al., 2005). The second, operating in parallel to carotenoid degradation, is the browning of leaves because of fungal spread within them, which provides a mechanism for partial masking of their yellow colouration. The possibility that these fungi are mutualistic, similar to fungi that occupy green leaves of many plants and defend them (e.g. Clay, 1990) has never been addressed.
In any case, the progress made by Sinkkonen et al. is a step in the right direction, a bright light in the very complicated issue of the functions of strong and conspicuous autumn leaf colours and possibly also concerning other types of defensive plant colouration. We need many more such studies to better understand this fascinating, complicated and broad phenomenon of plant life.