Doi:10.1016/j.tree.2004.09.009

Red leaves, insects and coevolution: a red herring? H. Martin Schaefer1 and David M. Wilkinson2 1Institute for Biology 1, Albert Ludwigs-Universita¨t, Hauptstr. 1, 79104 Freiburg, Germany2School of Biological and Earth Sciences, Liverpool John Moores University, Byrom St, Liverpool, UK L3 3AF W.D. (Bill) Hamilton proposed that coevolution between defensive strength enables well defended individuals to plants and herbivorous insects explains the bright reduce their herbivore load and, at the same time, insects autumnal colouration of leaves. Accordingly, plants to locate suitable hosts more efficiently. Thus, the basic invest in bright signals to reduce their herbivore load, prediction of the theory is that bright colours are honest whereas insects use these bright signals to identify less- signals that enable insects to select less well-defended defended hosts more efficiently. Archetti and Brown trees; therefore, trees with bright leaves will suffer a lower have recently revisited this theory by explaining its basic level of insect attack than will those that are dull coloured.
predictions and providing new research perspectives.
In this proposed signalling system, less-protected plants Their work presents an important basis to our under- are the losers. Such a signalling system requires costs standing of non-green leaf colouration, provided that either in the production of the signal or increasing alternative adaptive explanations on the photoprotec- marginal costs to prevent cheating by low-quality indi- tive and antioxidant role of leaf pigments, or their viduals. The authors are cautious about how the signal possible function in crypsis to herbivores are incorpor- relates to individual quality postulating that the intensity of colouration indicates a plant’s defensive commitment.
Such a relationship between secondary compounds acting The late Bill Hamilton, one of the most important as feeding deterrents to insects and colouration has been evolutionary theorists of the 20th century, suggested documented in some leaves and fruits Archetti shortly before his death in 2000 a strikingly novel theory and Brown propose further that honesty-enforcing costs to explain the leaf colours of autumn trees as a defence are found (i) in the timing of the signal (i.e. plants that against insect attack This theory was labelled as change colour early lose primary production owing to an ‘the coevolution theory’, although here we term it the early cessation of photosynthesis); or (ii) in the intensity of ‘leaf signal theory’ to distinguish it from the many other the signal (i.e. it is the colouration itself that costly).
hypotheses related to coevolution. A new paper by two of In support of leaf signalling theory, an analysis of the his collaborators further develops these ideas and literature documented that species with brighter autumn addresses some of the recent criticisms of the theory .
colouration harboured more species of specialist aphids Before the introduction of the leaf signal theory, most than did duller-coloured species, suggesting that poten- ecologists regarded leaf colouration as a consequence of tially vulnerable plant species evolved brighter leaf leaf senescence preceding abscission and, as such, it colouration . Moreover, an early onset of autumn attracted little attention. By proposing a new adaptive colouration in mountain birch Betula pubescens correlated explanation, Hamilton and co-workers laid out innovative negatively with aphid damage in the following season, research perspectives that triggered experiments probably because fewer aphids laid their eggs on individ- as well as controversy The key attraction for a uals that changed colour early in the season In this wide audience is that the hypothesis touches on several species, individuals with bright leaves also had a lower disparate fields: coevolution and signal theory, insect and degree of fluctuating asymmetry, supporting the idea that plant ecology, biochemistry and plant physiology. This is there is a link between plant vigour and the intensity of also its main challenge: the predictions of leaf signalling theory have to be tested against other adaptive expla- Currently, this theory is limited to leaf senescence in nations of leaf pigments before the theory can gain wide spite of the fact that coevolution between insects and acceptance. Therefore, experiments to test the leaf plants has also been suggested as an explanation of the signalling theory must be designed such that other, not occurrence of anthocyanins in young unfolding leaves necessarily mutually exclusive, hypotheses, such as the Here, red leaves were assumed to be cryptic to insects photoprotection theory suggested first nearly a rather than conspicuous, as postulated by leaf signalling theory. That both hypotheses assume different functions of Leaf signalling theory posits that bright leaf colour- red leaf colouration is attributable to the variation in the ation in autumn serves as a signal to herbivorous insects occurrence of a red-light sensitive receptor type in insects.
and reveals the defensive commitment of the individual Although red-light sensitive receptors have evolved inde- plant. The fundamental conjecture is that signalling pendently in some species of four insect orders (Odonata,Hymenoptera, Lepidoptera and Coleoptera) , most Corresponding author: H. Martin Schaefer (martin.schaefer@biologie.uni- insects are not particularly sensitive to red light, and hence leaf crypsis was proposed. By contrast, leaf photosynthesis in red leaves at low temperatures Anthocyanin accumulation might also result in increased, water-soluble red, rarely blue, pigments found in the cell rather than decreased, levels of photosynthesis in red vacuoles of both juvenile and senescing leaves of many plant species.
compared with green leaves at variable light conditions prevents or slows the breakdown of a substance by oxygen, Thus, variable illumination combined with low Carotenoids: highly unsaturated lipid-soluble yellow–red pigments produced temperatures in autumn alone might explain the syn- Photoinhibition: a light-induced stress reaction embracing all reversible andirreversible phenomena that lower the efficiency of photosynthesis.
Reactive oxygen species (ROS): includes all molecules containing oxygen with an unpaired electron (i.e. a free radical); generally reactive with other The intensity of the signal is not a reliable indicator of substances and often induce damage to tissue.
honesty-enforcing costs: leaf senescence is a tightlycontrolled process for transferring nutrients, particularly signalling theory has focussed on aphids, in which the red- nitrogen and phosphorus, from leaves to perennial tissue.
light sensitive receptors have not yet been found. There is Species with a brighter autumnal colouration appear to a clear need for more comparative data on the colour retain more nutrients from the leaves than do duller vision of herbivorous insects before refuting either of the coloured species potentially explaining the inter- specific variation in leaf colouration observed by Hamiltonand Brown Only the photosynthetic tissue that is well protected against frequent photoinhibition can supply In their paper, Archetti and Brown deal only with sufficient energy for nutrient translocation. This trans- insect and plant ecology and assume that plant physiology location accounts for most of the nitrogen and phosphorus might not explain the observed variation in leaf colour- reservoir that influences growth and reproduction in the ation. It is unfortunate that alternative adaptive hypo- following year Thus, the positive correlation between theses are neglected, because plant pigments can the intensity of autumnal leaf coloration and plant vigour signal to animals and simultaneously serve physiological in mountain birch might be attributable to the extent functions within the leaf. How then can we disentangle of nutrient recovery rather than to signalling defensive these different functions? Here, we summarize how commitment alone. Consequently, the costs of signal pro- photoprotection might affect leaf colouration.
duction (i.e. forming anthocyanins and carotenoids) have Tissue involved in photosynthesis is susceptible to to be balanced against the benefits of nutrient recovery damage by high illumination because it is unlikely to release as heat excess energy that cannot be used for In addition, anthocyanins act as antioxidants scaveng- photosynthesis. Excess light quanta might instead result ing free radicals in leaves In autumnal leaves, this in the formation of REACTIVE OXYGEN SPECIES (ROS) antioxidant role is particularly important because the (see Glossary). To avoid damage to tissues by these and process of nutrient recovery requires the breakdown of other substances, many plants rely on a strategy called leaf material, causing an increased risk of forming ROS.
PHOTOINHIBITION which can be induced by excess light, The oxidative stress increases further owing to low tem- ROS and low temperatures Because photoinhibition peratures combined with a high light level, which might might lower productivity and growth, it is crucial for plants ultimately lead to the destruction of the photosynthetic to reduce the frequency of photoinhibiting reactions, apparatus . Interestingly, and owing to this antioxi- especially those that are irreversible. This is achieved by dant role, leaf colouration might be correlated with insect the photoprotective and ANTIOXIDANT role of the CAROTENOIDS herbivory, not as a signal to divert insects, but as a result and ANTHOCYANINS pigments, which are responsible for the of mechanical injury by herbivorous insects preventing yellow–red hues in leaves These pigments intercept further oxidative damage from the uncontrolled break- excess light quanta that are otherwise absorbed by chloro- down of substances in already injured tissue . By shap- phyll b, and therefore provide a ‘sunscreen’ . Moreover, ing the elemental processes of photosynthesis, nutrient both pigments also act as scavengers of ROS, preventing recovery and protection of tissue against oxidative stress, damage to the tissue involved in photosynthesis .
the physiological role of anthocyanins and carotenoids Several recent studies highlighting these functions chal- contributes fundamentally to plant fitness. Consequently, lenge the basic assumptions of leaf signalling theory.
these adaptive explanations, rather than leaf signalling,might explain the variation in leaf colouration.
Signal timingThe timing of the signal is not necessarily an indicator of honesty-enforcing costs: an early shedding of leaves might Thanks to Hamilton and co-workers, the inter- and be adaptive in cold climates because a plant often con- intraspecific variation in autumnal leaf colouration has sumes more CO2 than it produces owing to frequent been brought to the attention of both ecologists and temperature-induced photoinhibition . Moreover, and evolutionists. As yet, it is unclear which of the alternative contrary to the assumptions of Archetti and Brown , hypotheses will explain the phenomenon. To understand photosynthesis does not necessarily cease with the onset of the phenomenon from an evolutionary perspective, we colouration. On the contrary, the photoprotective role of must embrace all contexts in which non-green leaf anthocyanins might lower the frequency of cold-induced colouration occurs. Red pigmentation is found not only photoinhibition, resulting in constantly high levels of during senescence, but also in unfolding new leaves and in Box 1. Research perspectives that integrate physiology and 1 Hamilton, W.D. and Brown, S.P. (2001) Autumn tree colours as a handicap signal. Proc. R. Soc. Lond. Ser. B 268, 1489–1493 2 Archetti, M. (2000) The origin of autumn colours by coevolution.
The main challenge to the study of leaf colouration is to control for the different functions of pigments, summarized as: (i) signalling to 3 Archetti, M. and Brown, S.P. (2004) The coevolution theory of autumn insects; (ii) acting as antioxidants; and (iii) protecting leaf organelles colours. Proc. R. Soc. Lond. Ser. B 271, 1219–1223 via the interception of light quanta during adverse environmental 4 Holopainen, J.K. and Peltonen, P. (2002) Bright autumn colours of conditions and nutrient translocation. (ii) and (iii) can be merged intothe photoprotection theory. Here, we outline a few avenues for deciduous trees attract aphids: nutrient retranslocation hypothesis.
separating the leaf signalling and photoprotection theories.
† A promising model system is an intraspecific comparison between 5 Wilkinson, D.M. et al. (2002) The adaptive significance of autumn leaf anthocyanin-deficient mutants and wild-type individuals An easy test of leaf signalling theory is to ask whether the insect load is 6 Hagen, S.B. et al. (2003) Autumn colouration and herbivore resistance greater in mutants than in wild-type individuals. Likewise, determin- in mountain birch (Betula pubescens). Ecol. Lett. 6, 807–811 ing whether the rate of photosynthesis and nutrient recovery are 7 Hagen, S.B. et al. (2004) Autumn coloration as a signal of tree generally higher under adverse conditions in wild-type individuals condition. Proc. R. Soc. Lond. Ser. B 271(Suppl.), S184–S185 tests the photoprotection theory. The comparison between wild 8 Gould, K.S. et al. (1995) Why leaves are sometimes red. Nature 378, types and mutants also serves as a model to assess the costs of foliar pigmentation. In wild types, the metabolic costs during anthocyanin 9 Gould, K.S. et al. (2002) Do anthocyanins function as antioxidants in accumulation can be balanced against the relative benefits of leaves? Imaging of H2O2 in red and green leaves after mechanical nutrient recovery (compared with mutants).
injury. Plant Cell Environ. 25, 1261–1269 † Another useful approach is to manipulate environmental con- 10 Pietrini, F. et al. (2002) Anthocyanin accumulation in the illuminated ditions during senescence and study their influence on the intensity surface of maize leaves enhances protection from photo-inhibitory of leaf colouration. Photoprotection theory expects a strong influ- risks at low temperature, without further limitation to photosyn- ence, especially by shading and temperature, whereas, according to thesis. Plant Cell Environ. 25, 1251–1259 Archetti the putative signalling system between plants and 11 Hoch, W.A. et al. (2003) Resorption protection. Anthocyanins facilitate insects collapses if the environmental effects are too strong. For all nutrient recovery in autumn by shielding leaves from potentially experiments, leaf colouration should be recorded with a spectro- damaging light levels. Plant Physiol. 133, 1296–1305 meter to enable high repeatability and detailed measurements to 12 Feild, T.S. et al. (2001) Why leaves turn red in autumn: the role of be taken. So far, most researchers have scored leaf colouration anthocyanins in senescing leaves of red-osier dogwood. Plant Physiol.
according to categories based on the human eye. Although these indices might be easy to use, they remain subjective, impeding easy 13 Wheldane, M. (1916) The Anthocyanin Pigments of Plants, Cambridge replication of published results. Using a spectrometer also enables fine-tuned correlation of the intraspecific variation in the intensity of 14 Schaefer, H.M. and Schmidt, V. (2004) Detectability and content as foliar pigmentation with the efficiency of nutrient recovery.
† opposing signal characteristics in fruits. Proc. R. Soc. Lond. Ser. B Quantifying the foliar concentrations of insect-deterring secondary compounds, such as tannins in brightly coloured leaves is essential when testing the central assumptions of leaf signalling that 15 Close, D.C. et al. (2001) Temporal variation of tannins (gallolylglu- defensive commitment is indicated by leaf colouration. If spectro- cose), flavonols and anthocyanins in leaves of Eucalyptus nitens metric readings are taken, the concentrations of secondary com- seedlings: implications for light attenuation and antioxidant activi- pounds can be compared with the intensity of the signal.
ties. Austral. J. Pl. Physiol. 28, 1–10 † Experiments of leaf signalling theory (bright leaves during 16 Dominy, N.J. et al. (2002) Why are young leaves red? Oikos 98, senescence) should be extended to encompass all instances of bright leaf colouration (young leaves and understorey plants) to 17 Briscoe, A.D. and Chittka, L. (2001) The evolution of color vision in facilitate our understanding of the evolutionary origin of the phenom- insects. Annu. Rev. Entomol. 46, 471–570 enon. The fact that anthocyanins are present in plants growing under 18 Tyystja¨rvi, E. and Aro, E.M. (1996) The rate constant of photoinhibi- the variable light conditions of the forest floor, where sun flecks briefly tion, measured in lincomycin-treated leaves, is directly proportional to interrupt otherwise dim light conditions, as well as in both transitory light intensity. Proc. Natl. Acad. Sci. U. S. A. 93, 2213–2218 stages of leaves (construction and destruction of organelles) might 19 Dungan, R.J. et al. (2004) Simulated carbon uptake for a canopy of two ndicate a common photoprotective role in all occasions.
broadleaved tree species with contrasting leaf habit. Funct. Ecol. 18,34–42 shade-adapted understorey plants. Whether its function is 20 Wingler, A. et al. (2004) Spatial patterns and metabolic regulation of photosynthetic parameters during leaf senescence. New Phytol. 161, the same in all these instances should be the focus of future work (). All future research must determine 21 Taylor, B.R. (1998) Air-drying depresses rate of leaf litter decompo- which of several possible ecological and physiological sition. Soil Biol. Biochem. 30, 403–412 functions will explain the occurrence of non-green leaves.
22 Lee, D.W. et al. (2003) Pigment dynamics and autumn leaf senescence In doing so, we should not rely on our tendency to want to in a New England deciduous forest, eastern USA. Ecol. Res. 18, believe in simple and attractive explanations to complex 23 Wise, R.R. (1995) Chilling enhanced photoxidation: the production, action and study of reactive oxygen species produced during chilling inthe light. Photosyn. Res. 45, 79–97 AcknowledgementsWe thank Veronika Schaefer, Tom Sherratt, Marco Archetti, and twoanonymous referees for valuable comments about the article. During thewriting of this article DMW was visiting Carleton University, Ottawa 0169-5347/$ - see front matter Q 2004 Elsevier Ltd. All rights reserved.
with funding from The Royal Society.

Source: http://search.livjm.ac.uk/BIE/treeleaf.pdf