What Do Red and Yellow Autumn Leaves Signal?

The Botanical Review 73(4): 279Ð289

Simcha Lev-Yadun Department of Biology Faculty of Science and Science Education University of Haifa - Oranim Tivon 36006, Israel

and

Kevin S. Gould Department of Botany University of Otago, Dunedin, New Zealand

Abstract ...... 279 Introduction ...... 280 Previous Hypotheses for the Signaling Function of Red and Yellow Autumn Leaves ...... 280 Autumn Leaves and the Nature of Signals ...... 282 We Propose That Autumn Leaves Signal That They Are About to Be Shed ...... 282 The Risk of Herbivore Attacks and the Strength of Defense ...... 282 Autumn Leaf Colors Undermine Herbivorous Insect Camouflage ...... 283 Size of Color Patch and Efficiency of Insect Camouflage ...... 284 Red and Yellow Autumn Leaves Are Aposematic ...... 284 Leaf Color Variability ...... 285 Conclusions ...... 285 Acknowledgments ...... 286 Literature Cited ...... 286

Abstract

The widespread phenomenon of red and yellow autumn leaves has recently attracted considerable scientific attention. The fact that this phenomenon is so prominent in the cooler, temperate regions and less common in warmer climates is a good indication of a climate-specific effect. In addition to the putative multifarious physiological benefits, such as protection from photoinhibition and photo-oxidation, several plant/animal inter- action functions for such coloration have been proposed. These include (1) that the bright leaf colors may signal frugivores about ripe fruits (fruit flags) to enhance seed dispersal; (2) that they signal aphids that the trees are well defended (a case of Zahavi’s handicap principle operating in plants); (3) that the coloration undermines herbivore insect camouflage; (4) that they function according to the “defense indication hypothesis,” which states that red leaves are chemically defended because anthocyanins correlate with various defensive compounds; or (5) that because sexual reproduction advances the

Copies of this issue [73(4)] may be purchased from the NYBG Press, The New York Botanical Garden, Bronx, NY 10458-5125, U.S.A. Please in- quire as to prices.

Issued 28 December 2007 © 2007 The New York Botanical Garden 280 THE BOTANICAL REVIEW onset of leaf senescence, the pigments might indicate to sucking herbivores that the leaves have low amounts of resources. Although the authors of hypotheses 3, 4, and 5 did not say that bright autumn leaves are aposematic, since such leaves are chemically defended, unpalatable, or both, we suggest that they are indeed aposematic. We propose that in addition to the above-mentioned hypotheses, autumn colors signal to herbivorous insects about another defensive plant property: the reliable, honest, and critical information that the leaves are about to be shed and may thus cause their mortality. We empha- size that all types of defensive and physiological functions of autumn leaves may operate simultaneously.

Introduction

Bright red and yellow autumn leaves are a widespread phenomenon, particularly in temperate regions (Hoch et al., 2001). For many decades, most people believed that these colors simply appeared after the degradation of chlorophyll that masked these pigments, and that they served no function. Recently, however, it has been shown that in many plants, anthocyanins are not simply unmasked but rather synthe- sized de novo by leaves in mid-senescence (Lee et al., 2003). Thus, the question of the possible physiological and ecological benefits of this coloration has attracted considerable scientific attention. There is very good evidence for physiological benefits of autumn leaf coloration, such as an enhanced recovery of foliar nitrogen owing to the protection by anthocyanins from photoinhibition and photo-oxidation (Ya- masaki, 1997; Chalker-Scott, 1999; Matile, 2000; Feild et al., 2001; Gould et al., 2002a; Lee & Gould, 2002a; Hoch et al., 2001, 2003; Close & Beadle, 2003; Schae- fer & Wilkinson, 2004; Gould, 2004; Ougham et al., 2005). These physiological ad- vantages notwithstanding, certain hypotheses regarding nonphysiological functions of autumn leaf coloration also merit consideration. Gould et al. (2002b), Lev-Yadun et al. (2002, 2004), Lev-Yadun (2006), and Schaefer and Wilkinson (2004) have already argued that the nonphotosynthetic plant pigments have the potential to serve more than one function concurrently. Thus, various hypotheses concerning coloration of leaves and other plant parts need not contrast with or exclude any other functional explanation of specific types of plant coloration, and those traits such as coloration that might have more than one type of benefit may be selected for by several agents. Consistent with Grubb’s (1992) view that defense systems are not simple, and with Diamond’s (2005) view that single-factor explanations can fail when complex environmental issues are being discussed, we consider that the evolution of autumn leaf coloration reflects an adaptation both to physiological pressures and to other organisms. Such synergistic gains may cause evolution of the red leaf color trait to be quicker and more frequent.

Previous Hypotheses for the Signaling Function of Red and Yellow Autumn Leaves

One of the earliest proposals that autumn leaves function as a signal to animals held that the pigments serve as a fruit flag for frugivores (Stiles, 1982; Willson & Hoppes, 1986; Facelli, 1993). Several additional roles of this coloration in defense against insect herbivory have been proposed. The first proposed, that the colors of autumn leaves signal that the trees are well defended, representing a case of Zahavi’s handicap principle WHAT DO RED AND YELLOW AUTUMN LEAVES SIGNAL? 281 operating in plants (Archetti, 2000, 2007a, 2007b; Hamilton & Brown, 2001; Hagen et al., 2003, 2004; Archetti & Brown, 2004; Archetti & Leather, 2005; Brown, 2005), is an idea that some accept only partly or not at all (Holopainen & Peltonen, 2002; Wilkinson et al., 2002; White, 2003; Schaefer & Wilkinson, 2004; Ougham et al., 2005; Sinkko- nen, 2006a, 2006b; Schaefer & Rolshausen, 2006, 2007; Chitka & Döring, 2007; Rol- shausen & Schaefer, 2007; Schaefer & Gould, 2007). The handicap principle (Zahavi, 1975) states that signaling is costly and therefore reliable. It has been proposed that those features that constitute the handicap evolved as a measure of the quality of the sig- naler (Zahavi, 1975, 1977, 1987; Grafen, 1990; Zahavi & Zahavi, 1997). Organisms that operate under the handicap principle send honest (and usually, but not always, costly) signals (Lachmann et al., 2001), which the receiver can evaluate in the process of decid- ing whether or not to respond. Archetti (2000) specifically rejected the possibility that autumn leaf coloration is aposematic, and other studies that favor the signaling hypothesis (Hamilton & Brown, 2001; Hagen et al., 2003, 2004; Archetti & Brown, 2004; Ar- chetti & Leather, 2005) do not discuss aposematism. Lee and Gould (2002b), Lee (2002), Gould (2004), and Sherratt et al. (2005) interpreted the hypothesis of handicap- related coloration as described by Archetti (2000) and Hamilton and Brown (2001) as a case of aposematism, notwithstanding the different view of the authors. As for the ob- jections, Holopainen and Peltonen (2002) proposed that leaves that had just turned yellow would be good indicators to aphids that nitrogen in the form of amino acids was available in these leaves. Wilkinson et al. (2002) proposed that rather than signaling aphids about their defensive qualities, especially since yellow leaves attract aphids, the yellow coloration serves as a sunscreen, and that red colors both warm leaves and function as antioxidants. Ougham et al. (2005) stressed that the physiological role of autumn leaf coloration is both important and well documented. They argued that the signal is not costly, yet plants often show within-canopy variation in leaf color; thus, studies of individual leaves, rather than of canopies as a whole, might provide a better understanding of the possible role in defense. Since honest signals are not always costly (Lach- mann et al., 2001), there is no need for them to be so in the case of bright autumn leaves. Schaefer and Rolshausen (2006) proposed an upgraded and combined new hypothesis for defensive plant coloration, focusing on anthocyanins. Elaborating on a previous idea by Fineblum and Rausher (1997) concerning the common biochemical pathways for flower color and defensive molecules, Schaefer and Rolshausen (2006) formulated the “defense indication hypothesis.” This hypothesis predicts that fewer herbivorous insects (and we propose any sensitive herbivore) will feed on plants that have strong anthocyanin coloration because it correlates with the strength of a chemical defense. The biochemical basis for this correlation is that anthocyanins and a number of defense chemicals such as tannins stem from the same biosynthetic pathways. A new defensive hypothesis about the function of autumn leaf colors has recently been proposed following the observation that in mountain birch, leaves turn yellow ear- lier if the trees produce an abundance of female catkins (Sinkkonen, 2006a, 2006b). Sinkkonen proposed that leaf coloration might signal to sucking herbivores that the leaves hold low amounts of resources. Although not explicitly stated as such, signaling about low-quality food yet again implies aposematism. This new hypothesis warrants further study. However, since young and even mature nonflowering individuals also ex- press the autumn coloration, the role of reproduction in the evolution of this coloration is probably minor. 282 THE BOTANICAL REVIEW

Autumn Leaves and the Nature of Signals

Schaefer and Rolshausen (2006) argued that since autumn leaf coloration probably evolved primarily because of physiological pressures rather than as a defense against herbivores, it cannot be considered a signal (i.e., it is not aposematic), but instead it is used only as a cue by the insects. Our view is that the evolutionary history of adapta- tions does not necessarily determine their current status; indeed, several functions for a single character are quite common in biology. Therefore, there is no theoretical reason to dismiss simultaneous roles of pigments in both physiological functions and signaling. There is probably a gradient between signals that serve solely as signals and those that serve mostly other functions with only some contribution to signaling. There is also no reason to ignore the possibility that signals that currently serve only for defense might have evolved primarily for other functions, or vice versa. The assumption that signals serve only as costly ones and never contribute to other functions, implies either that their evolution has ended or that such traits evolved at once, in a single step. This view seems to be an oversimpliﬁcation of a more complicated evolutionary and current functional situation. We are too far from a full or even a good understanding of signal evolution according to any type of deﬁnition to adopt such criteria.

We Propose That Autumn Leaves Signal That They Are About to Be Shed

Autumn leaf signaling may beneﬁt both trees and insects. At a time when the phloem sap is rich with the nutrients remobilized from senescing leaves, aposematic signaling may well beneﬁt the tree by deterring phloem sap feeders such as aphids (Dixon, 1997). We propose that autumn colors may also serve as a warning that the leaves are going to be shed soon, vital information for insects that need leaves, to use as a habitat, that will stay on the tree for a longer time than several days or weeks. Leaf fall is a considerable agent of insect mortality (Faeth et al., 1981), and insects may refrain from occupying leaves that are soon to abscise (Glinwood & Pettersson, 2000; Karban, 2007). Thus, the signaling tree will incur less herbivory, and the insects less mortality. We already know that certain aphids lay eggs on trees whose leaves have just started to change color and are yellow-orange, but that they usually refrain from occupying trees with red leaves (Furuta, 1986). Thus, the potential for such plant-insect communication exists.

The Risk of Herbivore Attacks and the Strength of Defense

Since red autumn leaves seem not to be as severely attacked by insects as are the younger, green leaves, the question arises as to whether there is really any potential danger from which autumn leaves need to be defended. The issue of autumn red leaves is only a special case of a more general question: why are some organisms attacked more (or less) than others? Although some consider that no attack equals no risk (Holopainen & Peltonen, 2002; Wilkinson et al., 2002; Schaefer & Wilkinson, 2004), the fact that certain organisms are not attacked is not a priori an indication that there is no such risk. We argue that the risk for the trees exists, but that it might well be the case that the low level of attacks simply indicates that defenses are strong, and that the defended organisms have successfully deterred possible enemies. For instance, in the deserts of the Near East some plants form green islands in the dry summer when all surrounding plants have turned yellow and have been grazed to their roots. Even under WHAT DO RED AND YELLOW AUTUMN LEAVES SIGNAL? 283 such extreme grazing pressure, those green plants are virtually ignored by the ﬂocks of sheep and goats that pass them daily. These summer-green plants are characterized by being poisonous or thorny as protection against herbivory. Under dry summer condi- tions, green is conspicuous and contrasts with the background colors, as do yellow, red, and black in “greener” ecosystems (Lev-Yadun & Ne'eman, 2004). The lack of attacks on these green plants is a clear indication of their very good defensive qualities rather than of a low level of risk. Thus, the lack of common severe herbivore attacks on autumn leaves, unlike the common attacks we witness on young spring or mature summer green leaves of the same trees, does not prove (1) whether or not the colors are a signal to herbivores or (2) serve only for physiological issues or (3) that the trees have no risk. A well-defended organism may appear to have no enemies because the potential enemies have long evolutionary experience with its defensive qualities and refrain from attack. This interpretation is especially valid if there are other more palatable plants in the vicinity.

Autumn Leaf Colors Undermine Herbivorous Insect Camouﬂage

Another recent hypothesis about autumn leaf color holds that the pigmentation undermines the camouflage of invertebrate herbivores, particularly of insects and other small invertebrates (Lev-Yadun et al., 2004; Lev-Yadun, 2006). Thus exposed, the invertebrates are vulnerable to predation, and, in addition, they avoid plant organs with un- suitable coloration, to the benefit of the plants. It has been claimed that the considerable variations in the colors of leaves, petioles, and stems, as well as of flowers and fruits, undermine the camouflage of invertebrate herbivores, especially insects (Lev-Yadun et al., 2004). For instance, color differences between the upper and lower sides of leaves, and between the veins or petioles and the leaf blade, are common across diverse plant forms from short annuals to tall trees and across various habitats from deserts to rain forests and from the tropics to temperate regions. When a given leaf has two different colors: green on its upper (adaxial) side and blue, brown, pink, red, white, yellow, or just a different shade of green on its lower (abaxial) side, a green insect (or otherwise- colored one) that is camouflaged on one of the sides will not be camouflaged on the other. The same is true for vein, petiole, branch, stem, flower, or fruit coloration. It has been thus suggested (Lev-Yadun et al., 2004) that green or otherwise-colored herbivores that move, feed, or rest during the day on plant parts that have different colorations from their own immediately become more conspicuous to their predators. Many plants are simply too colorful to enable a universal camouflage of herbivorous insects and other invertebrates to operate successfully, and they force small herbivores to cross areas (“killing zones”) with colors that do not match their camouflage. This is a special case of “the enemy of my enemy is my friend” and a visual parallel of the chemical signals that plants emit to call wasps when attacked by caterpillars (Kessler & Baldwin, 2001). Moreover, this is a common natural parallel to the well-known principle that forms the basis for the phenomenon of industrial melanism (Kettlewell, 1973; Majerus, 1998), which illustrates the importance of plant-based camouflage for herbivorous insect sur- vival and serves as an independent test of the defensive plant coloration hypothesis. Al- though it is generally agreed that flower and fruit colors facilitate communication between plants and their pollinators and seed-dispersers (Ridley, 1930; Faegri & van der Pijl, 1979; Willson & Whelan, 1990; Weiss, 1995; Schaefer et al., 2004), there is no a priori reason to assume that flower, fruit, and leaf colors cannot serve also in the defense 284 THE BOTANICAL REVIEW against herbivory by undermining camouflage, being aposematic, or even by mimicking unpalatable plants. Plants provide habitat and food for many animals. Intuitively, the common optimal camouflage for herbivorous insects should be green, and, indeed, many aphids, caterpillars, and grasshoppers, for example, have evolved green coloration (Cott, 1940; Purser, 2003). The effectiveness of green camouflage is damaged by diverse nongreen background colors, or even by a variety of green shades as plant backgrounds. Since the vari- able coloration is usually either ephemeral (young red leaves or autumn red leaves) or occupies only a small part of the canopy (young leaves, petioles, flowers, and fruits), the selective pressure on insects to evolve to match such coloration is low and of limited adaptive value (Lev-Yadun et al., 2004). The excellent color vision of many predators of insects, in particular insectivorous birds (the most common and significant predators of herbivorous invertebrates) (Van Bael et al., 2003), means that undermining a herbivores’ camouflage is highly rewarding for plants (Lev-Yadun et al., 2004; Lev-Yadun, 2006). The special importance of the hypothesis that colors undermine herbivores’ camouflage (Lev-Yadun et al., 2004) is that it proposes a defensive function not only for the yellow, orange, pink, red, brown, blue, and black colors resulting from carotenoids and anthocyanins but also for light green and white coloration.

Size of Color Patch and Efﬁciency of Insect Camouﬂage

In heterogeneous habitats, optimal camouflage coloration should maximize the de- gree of crypsis in the microhabitats used by the prey (Endler, 1984; Edmunds & Grayson, 1991; Merilaita, 2003). Theoretically, herbivores may enjoy better crypsis in heterogeneous habitats than in monochromatic ones (Merilaita, 2003; Merilaita et al., 1999). However, the ratio between the size of the herbivore and that of the color patch determines whether any one pigmentation pattern assists with crypsis, or undermines it. Since insects are in general smaller than the color patches of leaves, flowers, fruits, or branches, they tend be more exposed to their predators and parasites on most plant sur- faces, and their crypsis is undermined. In contrast, those types of variegation that form small-scale mosaics are unlikely to facilitate the undermining of insect camouflage, a point that has been addressed by Schaefer and Rolshausen (2006).

Red and Yellow Autumn Leaves Are Aposematic

If the “defense indication hypothesis” proposed by Schaefer and Rolshausen (2006) is accepted (and a similar hypothesis has been proposed for colorful fruits; Willson & Whelan, 1990), we propose that it immediately follows that plant parts rich in anthocyanins may serve in many cases as aposematic (warning) coloration for chemically based unpalatability. Even if red and yellow autumn leaves are not chemically well- defended, but just have a low nutritive value (another case of unpalatability), all autumn leaves (red and yellow) should also be considered aposematic. The issue of aposematism in chemically defended plants has already been discussed by previous studies (Cook et al., 1971; Hinton, 1973; Harborne, 1982; Lev-Yadun & Ne'eman, 2004; Lev- Yadun, 2006), and autumn leaves simply add to the scope of such phenomena. If the colored senescing leaves are well defended by various chemicals, or if they are avoided because of their low nutritive value (two cases of unpalatability), they should be considered aposematic. Aposematic coloration, a well-known phenomenon in animals, WHAT DO RED AND YELLOW AUTUMN LEAVES SIGNAL? 285 has recently been shown to be common also in thorny (Lev-Yadun, 2001, 2003a, 2003b, 2006; Midgley et al., 2001; Lev-Yadun & Ne'eman, 2004; Rubino & McCarthy, 2004; Ruxton et al., 2004; Speed & Ruxton, 2005) and poisonous plants (Cook et al., 1971; Hinton, 1973; Harborne, 1982; Lev-Yadun & Ne'eman, 2004; Lev-Yadun, 2006), and in plants that use several types of insect or thorn mimicry (Lev-Yadun & Inbar, 2002; Lev- Yadun, 2003a, 2006). It is obvious why one should consider as aposematic autumn leaves in species that are chemically well defended. Moreover, as in other cases of aposematism (Cott, 1940; Wickler, 1968; Lev-Yadun, 2003a), it is tempting to postulate that mimics of true aposematic autumn leaves also exist. The widespread phenomenon of red and yellow autumn leaves in some areas may partly be the result of Müllerian and Batesian mimicry.

Leaf Color Variability

Another theoretical issue related to autumn leaf coloration is variability. While several authors have proposed that such variability in timing and intensity of pigmentation has been selected for (Archetti, 2000; Hamilton & Brown, 2001; Ougham et al., 2005; Sinkkonen, 2006a), this variability may be a nondefensive adaptation, or not adaptive at all. Variation in the timing of color change may reflect topographical constraints, such as the position of leaves within the tree, the status of the tree as a dominant, subdomi- nant or suppressed individual within the population, or as an understory or canopy species, or else physiological pressures such as fluctuations in water supply or mineral nutrition; these factors may significantly influence a tree’s performance irrespective of its genotype (Kozlowski et al., 1991; Kozlowski & Pallardy, 1997). We have no long- term measurements (i.e., over at least five sequential years) of tree vigor, coloration intensity, date of color change, and insect attacks in many marked individuals of any tree species, and such data from a single species might not be sufficient to formulate general hypotheses or to provide a firm understanding of the genetic versus the environmental component of the phenomenon.

Conclusions

We conclude that autumn leaf colors have at least five potential visual defensive functions against herbivory in addition to the (better understood) important physiological roles and the nonvisual (chemical) defenses previously proposed. The first proposed defensive function of colors—as a signal that the trees are well defended (a case of Za- havi’s handicap principle) (Archetti, 2000; Hamilton & Brown, 2001; Hagen et al., 2003, 2004; Archetti & Brown, 2004; Archetti & Leather, 2005)—cannot be ignored, despite the arguments of critics of this hypothesis (Holopainen & Peltonen, 2002; Wilkinson et al., 2002; Schaefer & Wilkinson, 2004; Ougham et al., 2005; Schaefer & Rolshausen, 2006; Chittka & Döring, 2007; Schaefer & Gould, 2007). The second and related hypothesis is that red and yellow autumn leaves serve as aposematic coloration. The third is that varied colors undermine herbivorous insect camouflage (Lev-Yadun et al., 2004; Lev-Yadun, 2006). The fourth is that the colors function according to the “defense indication hypothesis,” which states that red leaves are chemically defended because anthocyanins correlate with various defensive compounds (Schaefer & Rol- shausen, 2006). The fifth hypothesis, proposed by Sinkkonen (2006a, 2006b), holds that yellow foliage signals to sucking herbivores that the tree is becoming poor-quality 286 THE BOTANICAL REVIEW food. The last (sixth) hypothesis (proposed here) is that autumn coloration might deter insects from occupying the soon-to-be-shed leaves. That autumn coloration potentially confers several types of benefit would explain why this phenomenon is so common. Bright autumn leaves have been an exciting focus of research in ecology and evolution and, in parallel, in physiology since 2000, and there has been a regular flow of new ideas. We are probably still quite far from the last word concerning this fascinating issue.

Acknowledgments

We thank Moshe Inbar for his valuable comments.

Literature Cited

Archetti, M. 2000. The origin of autumn colours by coevolution. J. Theor. Biol. 205: 625Ð630. ———. 2007a. Autumn colours and the nutritional translocation hypothesis: a theoretical assessment. J. Theor. Biol. 244: 714Ð721. ———. 2007b. Colour preference as evidence for the theories on the evolution of autumn colours. J. Theor. Biol. 245: 595Ð596. ——— & S. P. Brown. 2004. The coevolution theory of autumn colours. Proc. Roy. Soc. London, Ser. B 271: 1219Ð1223. ——— & S. R. Leather. 2005. A test of the coevolution theory of autumn colours: colour preference of Rhopalosiphum padi on Prunus padus. Oikos 110: 339Ð343. Brown, S. P. 2005. A view from Mars. Pp. 350Ð356 in M. Ridley (ed.), Narrow roads of gene land—the collected papers of W. D. Hamilton volume 3—last words. Oxford University Press, Oxford. Chalker-Scott, L. 1999. Environmental significance of anthocyanins in plant stress responses. Pho- tochem. & Photobiol. 70: 1Ð9. Chittka, L. & T. G. Döring. 2007. Are autumn foliage colors red signals to aphids? PLoS Biol. 5: 1640Ð1644. Close, D. C. & C. L. Beadle. 2003. The ecophysiology of foliar anthocyanin. Bot. Rev. 69: 149Ð161. Cook, A. D., P. R. Atsatt & C. A. Simon. 1971. Doves and dove weed: multiple defenses against avian predation. BioScience 21: 277Ð281. Cott, H. B. 1940. Adaptive coloration in animals. Methuen & Co., London. Diamond. J. 2005. Collapse. How societies choose to fail or succeed. Viking, New York. Dixon, A. F. G. 1997. Aphid ecology. An optimization approach. Chapman and Hall, London. Edmunds, M. & J. Grayson. 1991. Camouflage and selective predation in caterpillars of the poplar and eyed hawkmoths (Laothoe populi and Smerinthus ocellata). Biol. J. Linn. Soc. 42: 467Ð480. Endler, J. A. 1984. Progressive background matching in moths, and a quantitative measure of crypsis. Biol. J. Linn. Soc. 22: 187Ð231. Facelli, J. M. 1993. Experimental evaluation of the foliar flag hypothesis using fruits of Rhus glabra (L.). Oecologia 93: 70Ð72. Faegri, K. & L. van der Pijl. 1979. The principles of pollination ecology. Ed. 3. Pergamon Press, Ox- ford. Faeth, S. H., E. F. Connor & D. Simberloff. 1981. Early leaf abscission: a neglected source of mortality for folivores. Amer. Naturalist 117: 409Ð415. Feild, T. S., D. W. Lee & N. M. Holbrook. 2001. Why leaves turn red in autumn. The role of anthocyanins in senescing leaves of red-osier dogwood. Pl. Physiol. 127: 566Ð574. Fineblum, W. L. & M. D. Rausher. 1997. Do floral pigmentation genes also influence resistance to enemies? The W locus in Ipomoea purpurea. Ecology 78: 1646Ð1654. Furuta, K. 1986. Host preference and population dynamics in an autumnal population of the maple aphid, Periphyllus californiensis Shinji (Homoptera, Aphididae). J. Appl. Entomol. 102: 93Ð 100. WHAT DO RED AND YELLOW AUTUMN LEAVES SIGNAL? 287

Glinwood, R. & J. Pettersson. 2000. Movement by mating females of a host alternating aphid: a re- sponse to leaf fall. Oikos 90: 43Ð49. Gould, K. S. 2004. Nature’s Swiss army knife: the diverse protective roles of anthocyanins in leaves. J. Biomedicine & Biotechnol. 2004(5): 314Ð320. ———, D. W. Lee & J. A. Callow (eds.). 2002a. Anthocyanins in leaves. Advances Bot. Res. 37. ———, S. O. Neill & T. C. Vogelmann. 2002b. A unified explanation for anthocyanins in leaves? Ad- vances Bot. Res. 37: 167Ð192. Grafen, A. 1990. Biological signals as handicaps. J. Theor. Biol. 144: 517Ð546. Grubb, P. J. 1992. A positive distrust in simplicity—lessons from plant defences and from competition among plants and among animals. J. Ecol. 80: 585Ð610. Hagen, S. B., I. Folstad & S. W. Jakobsen. 2003. Autumn colouration and herbivore resistance in mountain birch (Betula pubescens). Ecol. Letters 6: 807Ð811. ———, S. Debeausse, N. G. Yoccoz & I. Folstad. 2004. Autumn coloration as a signal of tree condi- tion. Proc. Roy. Soc. London, Ser. B 271 (Suppl.): S184ÐS185. Hamilton, W. D. & S. P. Brown. 2001. Autumn tree colours as a handicap signal. Proc. Roy. Soc. Lon- don, Ser. B 268: 1489Ð1493. Harborne, J. B. 1982. Introduction to ecological biochemistry. Academic Press, London. Hinton, H. E. 1973. Natural deception. Pp. 97Ð159 in R. L. Gregory & E. H. Gombrich (eds.), Illusion in nature and art. Duckworth, London. Hoch, W. A., E. L. Zeldin & B. H. McCown. 2001. Physiological significance of anthocyanins during autumnal leaf senescence. Tree Physiol. 21: 1Ð8. ———, E. L. Singsaas & B. H. McCown. 2003. Resorption protection. Anthocyanins facilitate nutrient recovery in autumn by shielding leaves from potentially damaging light levels. Pl. Physiol. 133: 1296Ð1305. Holopainen, J. K. & P. Peltonen. 2002. Bright autumn colours of deciduous trees attract aphids: nutrient retranslocation hypothesis. Oikos 99: 184Ð188. Karban, R. 2007. Deciduous leaf drop reduces insect herbivory. Oecologia 153: 81Ð88. Kessler, A. & I. T. Baldwin. 2001. Defensive function of herbivore induced plant volatile emissions in nature. Science 291: 2141Ð2144. Kettlewell, B. 1973. The evolution of melanism. Clarendon Press, Oxford. Kozlowski, T. T. & S. G. Pallardy. 1997. Physiology of woody plants. Ed. 2. Academic Press, San Diego. ———, P. J. Kramer & S. G. Pallardy. 1991. The physiological ecology of woody plants. Academic Press, San Diego. Lachmann, M. L., S. Számadó & C. T. Bergstrom. 2001. Cost and conflict in animal signals and human language. Proc. Natl. Acad. Sci. U.S.A. 98: 13189Ð13194. Lee, D. W. 2002. Anthocyanins in autumn leaf senescence. Advances Bot. Res. 37: 147Ð165. ——— & Gould, K. S. 2002a. Why leaves turn red. Amer. Sci. 90: 524Ð531. ——— & ———. 2002b. Anthocyanins in leaves and other vegetative organs: an introduction. Ad- vances Bot. Res. 37: 1Ð16. ———, J. O’Keefe, N. M. Holbrook & T. S. Feild. 2003. Pigment dynamics and autumn leaf senescence in a New England deciduous forest, eastern USA. Ecol. Res. 18: 677Ð694. Lev-Yadun, S. 2001. Aposematic (warning) coloration associated with thorns in higher plants. J. Theor. Biol. 210: 385Ð388. ———. 2003a. Weapon (thorn) automimicry and mimicry of aposematic colorful thorns in plants. J. Theor. Biol. 244: 183Ð188. ———. 2003b. Why do some thorny plants resemble green zebras? J. Theor. Biol. 244: 483Ð489. ———. 2006. Defensive coloration in plants: a review of current ideas about anti-herbivore coloration strategies. Pp. 292Ð299 in J. A. Teixeira da Silva (ed.), Floriculture, ornamental and plant biotech- nology: advances and topical issues. Vol. IV. Global Science Books, London. ——— & M. Inbar. 2002. Defensive ant, aphid and caterpillar mimicry in plants. Biol. J. Linn. Soc. 77: 393Ð398. ——— & G. Ne'eman. 2004. When may green plants be aposematic? Biol. J. Linn. Soc. 81: 413Ð416. ———, A. Dafni, M. Inbar, I. Izhaki & G. Ne'eman. 2002. Colour patterns in vegetative parts of plants deserve more research attention. Trends Plant Sci. 7: 59Ð60. 288 THE BOTANICAL REVIEW

———, ———, M. A. Flaishman, M. Inbar, I. Izhaki, G. Katzir & G. Ne'eman. 2004. Plant coloration undermines herbivorous insect camouflage. BioEssays 26: 1126Ð1130. Majerus, M. E. N. 1998. Melanism. Evolution in action. Oxford University Press, Oxford. Matile, P. 2000. Biochemistry of Indian summer: physiology of autumnal leaf coloration. Exp. Geron- tol. 35: 145Ð158. Merilaita, S. 2003. Visual background complexity facilitates the evolution of camouflage. Evolution 57: 1248Ð1254. ———, J. Tuomi & V. Jormalainen. 1999. Optimization of cryptic coloration in heterogeneous habitat. Biol. J. Linn. Soc. 67: 151Ð161. Midgley, J. J., M. A. Botha & D. Balfour. 2001. Patterns of thorn length, density, type and colour in African acacias. African J. Range Forage Sci. 18: 59Ð61. Ougham, H. J., P. Morris & H. Thomas. 2005. The colors of autumn leaves as symptoms of cellular recycling and defenses against environmental stresses. Curr. Topics Developm. Biol. 66: 135Ð160. Purser, B. 2003. Jungle bugs: masters of camouflage and mimicry. Firefly Books, Toronto. Ridley, H. N. 1930. The dispersal of plants throughout the world. L. Reeve & Co., Ashford, UK. Rolshausen, G. & H. M. Schaefer. 2007. Do aphids paint the tree red (or yellow) - can herbivore resistance or photoprotection explain colourful leaves in autumn? Pl. Ecol. 191: 77Ð84. Rubino, D. L. & B. C. McCarthy. 2004. Presence of aposematic (warning) coloration in vascular plants of southeastern Ohio. J. Torrey Bot. Soc. 131: 252Ð256. Ruxton, G. D., T. N. Sherratt & M. P. Speed. 2004. Avoiding attack. The evolutionary ecology of crypsis, warning signals & mimicry. Oxford University Press, Oxford. Schaefer, H. M. & K. S. Gould. 2007. Modelling the evolution of leaf colouration with binary assump- tions is barking up the wrong tree. J. Theor. Biol. 249: 638Ð639. ——— & G. Rolshausen. 2006. Plants on red alert: do insects pay attention? BioEssays 28: 65Ð71. ——— & ———. 2007. Aphids do not attend to leaf colour as visual signal, but to the handicap of reproductive investment. Biol. Lett. 3: 1Ð4. ——— & D. M. Wilkinson. 2004. Red leaves, insects and coevolution: a red herring? Trends Ecol. Evol. 19: 616Ð618. ———, V. Schaefer & D. J. Levey. 2004. How plant-animal interactions signal new insights in communication. Trends Ecol. Evol. 19: 577Ð584. Sherratt, T. N., D. M. Wilkinson & R. S. Bain. 2005. Explaining Dioscorides’ “double difference”: why are some mushrooms poisonous, and do they signal their unprofitability? Amer. Naturalist 166: 767Ð775. Sinkkonen, A. 2006a. Sexual reproduction advances autumn leaf colours in mountain birch (Betula pubescens ssp. czerepanovii). J. Evol. Biol. 19: 1722Ð1724. ———. 2006b. Do autumn leaf colours serve as a reproductive insurance against sucking herbivores? Oikos 113: 557Ð562. Speed, M. P. & G. D. Ruxton. 2005. Warning displays in spiny animals: one (more) evolutionary route to aposematism. Evolution 59: 2499Ð2508. Stiles, E. W. 1982. Fruit flags: Two hypotheses. Amer. Naturalist 120: 500Ð509. Van Bael, S. A., J. D. Brawn & S. K. Robinson. 2003. Birds defend trees from herbivores in a Neotropical forest canopy. Proc. Natl. Acad. Sci. U.S.A. 100: 8304Ð8307. Weiss, M. R. 1995. Floral colour change: a widespread functional convergence. Amer. J. Bot. 82: 167Ð 195. White, T. C. R. 2003. Nutrient retranslocation hypothesis: a subset of the flush-feeding/senescence- feeding hypothesis. Oikos 103: 217. Wickler, W. 1968. Mimicry in plants and animals. Weidenfeld and Nicolson, London. Wilkinson, D. M., T. N. Sherratt, D. M. Phillip, S. D. Wratten, A. F. G. Dixon & A. J. Young. 2002. The adaptive significance of autumn leaf colours. Oikos 99: 402Ð407. Willson, M. F. & W. G. Hoppes. 1986. Foliar ‘flags’ for avian frugivores: signal or serendipity? Pp. 55Ð 69 in A. Estrada & T. H. Fleming (eds.), Frugivores and seed dispersal. Dr. W. Junk, Dordrecht. ——— & C. J. Whelan. 1990. The evolution of fruit color in fleshy-fruited plants. Amer. Naturalist 136: 790Ð809. Yamasaki, H. 1997. A function of colour. Trends Pl. Sci. 2: 7Ð8. WHAT DO RED AND YELLOW AUTUMN LEAVES SIGNAL? 289

Zahavi, A. 1975. Mate selection - A selection for a handicap. J. Theor. Biol. 53: 205Ð214. ———. 1977. The cost of honesty (further remarks on the handicap principle). J. Theor. Biol. 67: 603Ð 605. ———. 1987. The theory of signal selection and some of its implications. Pp. 305Ð327 in V. P. Delﬁno (ed.), International Symposium of Biological Evolution. Adriatica Editrica, Bari, Italy. ——— & A. Zahavi. 1997. The handicap principle: a missing piece of Darwin’s puzzle. Oxford Uni- versity Press, New York.