Search

A very strong, indirect evidence for the operation of aposematic coloration in thorny, spiny and prickly plants and its convergent evolution is the fact that conspicuous thorn, spine and prickle coloration is found in Angiosperm taxa that have mutually exclusive biochemical pathways of coloration. For instance, taxa belonging to the Caryophyllales (e.g., Cactaceae, Caryophyllaceae, Amaranthaceae) produce yellow, orange, red, purple and black pigments via the betalain pathway (Stafford 1994). Most other Angiosperm families use anthocyanins or carotenoids for similar patterns of coloration. The fact that spines of cacti are usually conspicuous because of their coloration (Lev-Yadun 2001), commonly including yellow, orange, red and black coloration resulting from betalain derivatives, indicates that this group of pigments is involved, among their various important physiological functions (e.g., Jain and Gould 2015), in aposematic coloration. By contrast, in Rosaceae, Asteraceae and Fabaceae as well as in many other Angiosperm families that use anthocyanins for orange, pink, red, blue, purple and black coloration of thorns, spines and prickles, the chemical origin of the aposematic coloration (Lev-Yadun 2001, 2006a; Lev-Yadun and Gould 2009) is different. It seems therefore that the aposematic coloration of thorny, spiny and prickly plants is a good case of convergent evolution.

Sticky plant surfaces made of trichomes were shown experimentally not only to defend plants directly by killing, damaging, or slowing down insects (sensu Levin 1973) but also to enhance indirect defense by attracting predaceous arthropods to such carrion-carrying plants because it provides these predators with food (Krimmel and Pearse 2013; LoPresti et al. 2015). Concerning a possible role of carrion odor in plant defense from herbivory, Lev-Yadun et al. (2009b) proposed on theoretical grounds that plants’ carrion or dung odors produced by flowers of certain species may simultaneously serve both the long-known attraction of pollinators and also act as an olfactory aposematism towards mammalian herbivores. The anti-herbivory (defensive) carrion odor hypothesis was supported by field data showing that cattle repeatedly refrained for many years from grazing in very productive paddocks used for the disposal of cattle carcasses (Lev-Yadun and Gutman 2013).

The genus Pisum includes according to different views either two well-defined species (P. fulvum and P. sativum), or three species (P. fulvum, P. elatius, P. humile) and a domesticated derivative (P. sativum) of its wild progenitor P. humile (Ben-Ze’ev and Zohary 1973). Even for those that think that only two species comprise the genus, the two common wild types of the domesticated Pisum sativum are known as P. elatius and P. humile. In Israel, these two closely related Pisum taxa seem in the field to be two very different biological entities, i.e., seem to be valid species and the genus Pisum seems to include three species.

My attraction to the question of autumn leaf colors started in the autumn/early-winter of 1994, in parallel to my decision to understand why Silybum marianum plants look like green zebras. More specifically, several field days in the autumn of 1994 at elevations of 1500–1900 m on Mount Hermon, the only place in Israel where the phenomenon of red autumn leaves is very conspicuous (Figs. 45.1, 45.2 and 45.3), triggered my attention and curiosity to understand this phenomenon. After I understood in the early spring of 1996 that visual aposematism exists in spiny plants, I became even more aware to the non-physiological aspects of leaf colors and wished to understand the biology of red autumn leaf coloration.

Spiders are very common voracious predators of various herbivorous arthropods on and around terrestrial plants, they play an important role in regulating arthropod populations (Wise 1993; Nyffeler et al. 1994; Nyffeler 2000) and their activity results in positive effects on plant growth in various ecologies (e.g., Louda 1982; Wise 1993; Nyffeler et al. 1994). In the various geographical regions occupied by dangerous spiders such as black widows (e.g., Ori and Ikeda 1998), which can significantly damage and even kill mammals, there is a possibility that even herbivorous mammals do not approach spider-web-like structures, as suggested by human arachnophobia (e.g., Peira et al. 2010; Vetter 2013). Accordingly, spiders are considered to satisfy the requirements for being the models of mimicry, and certain arthropods belonging to diverse taxa indeed mimic spiders in order to prevent predation (Eisner 1985; Mather and Roitberg 1987; Zolnerowich 1992; Rota and Wagner 2006). Hence, there is no theoretical reason to dismiss the possibility that plants may also mimic spiders or their webs.

The pattern of colored spots on the pods of P. humile and P. elatius proposed to function as aposematic caterpillar mimicry (Lev-Yadun and Inbar 2002; Aviezer and Lev-Yadun 2015) could also mimic eyes, a character that was only recently suggested to operate as defense in plants (Aviezer and Lev-Yadun 2015) (Figs. 56.1, 56.3, 56.7, 56.8 and 56.15) but is well-known in animals (Cott 1940; Edmunds 1974; Ruxton et al. 2004; Stevens et al. 2007, 2008a, b; De Bona et al. 2015). In animals, the “eyespots” are often not more than circular features (Stevens et al. 2008b), but the resemblance to an eye, to a face with two eyes, or even to more than two “eyes” or more than a single face, can intimidate or startle potential predators (herbivores), since the “eyespots” indicate the probable existence of a bigger and maybe even extremely dangerous creature (see Wickler 1968; Edmunds 1974; Ruxton et al. 2004; Stevens et al. 2007; Janzen et al. 2010). However, it seems that two “eyespots” intimidate better than other numbers (Mukherjee and Kodandaramaiah 2015). It seems that concerning “eyespots”, and concerning the possible snake mimicry by certain P. elatius pods discussed below, both proposed mimicry types do not necessarily have to be exact (see Stevens et al. 2008b; Penney et al. 2012). There is no need to mimic eyes exactly, since the price of a mistake in identifying a very dangerous predator can be lethal (see Janzen et al. 2010) and it is better not to spend more time in close proximity to such a potential danger. The so-called “eyespots” that certainly defend in many cases from attacks, may operate not only as eye mimicry, but also by just being a salient stimulus, through neophobia (Stevens and Ruxton 2014).

Because of the complexity of situations in nature, there is an understandable and even unavoidable need to isolate factors in experimental studies. This is part of the highly fruitful minimization approach that revolutionized many areas in science, including biology. However, the critical need to isolate and examine particular functions, molecules, or aspects, may lead to results that are significant for the experimental conditions, but may sometimes be of limited or of no ecological/evolutionary relevance, or may even be misleading concerning nature’s complex reality.

Visual plant defenses can be based on being cryptic, but can also be based on the opposite strategy, i.e., on being conspicuous and even by signaling. To make things more complicated, certain coloration patterns may serve for crypsis at long distances, but signal at close range, an issue that will be discussed in some detail in the chapter about aposematism. Maynard Smith and Harper (2003) defined signals “as any act or structure which alters the behaviour of other organisms, which evolved because of that effect, and which is effective because the receiver’s response has also evolved”. The recent increasing interests and debates on the signaling functions of red and yellow autumn leaves (e.g., Archetti 2000, 2007a, b, 2009a, b; Hamilton and Brown 2001; Holopainen and Peltonen 2002; Wilkinson et al. 2002; Hagen et al. 2003, 2004; Archetti and Brown 2004; Schaefer and Wilkinson 2004; Lev-Yadun et al. 2004a, 2012; Archetti and Leather 2005; Brown 2005; Schaefer and Rolshausen 2006a, b, 2007a, b; Sinkkonen 2006a, b, 2008; Chittka and Doring 2007; Lev-Yadun and Gould 2007; Schaefer and Gould 2007; Hatier and Gould 2008; Karageorgou et al. 2008; Ougham et al. 2008; Ramirez et al. 2008; Yamazaki 2008a, b; Archetti et al. 2009a; Lev-Yadun and Holopainen 2009, 2011; Lev-Yadun 2010; Sinkkonen et al. 2012) that will be discussed in detail in a later chapter are an excellent demonstration of the complicated experimental, physiological and interpretative issues that have to be dealt with when defensive visual plant signals are involved.

As a child that used to walk barefoot in the summer and step on various sharp objects including thorns, spines and prickles, and because of the wounds inflicted on me, several times I got emergency anti-tettanus injections. After I understood that thorns, spines and prickles are commonly visually aposematic, I wondered how common are bacterial pathogens found on these sharp pointed structures and with our enthusiastic departmental microbiologist Malka Halpern I began to examine this. Four publications (Halpern et al. 2007a, b, 2011; Lev-Yadun and Halpern 2008) showed that thorns, spines and prickles regularly harbor an array of pathogenic bacteria and fungi and discussed its potential significance in the evolution and functionality of plant aposematism. Spines from date palm (Phoenix dactylifera) trees, thorns from common hawthorn (Crataegus aronia) trees and two thorny shrub species, thorny burnet (Sarcopoterium spinosum) and manna tree (Alhagi graecorum), were sampled for aerobic and anaerobic bacteria in Israel. Every typical mature individual of these trees and shrubs carries hundreds or even thousands of conspicuous and therefore potentially aposematic spines or thorns. The common knowledge about the severity and frequency of aggressive bacterial infections among date-palm orchard workers in Israel following spine wounding, has necessitated the costly practice of removal of all the millions of spines from many of the orchards by mechanical saws. This was a very good indication that an important aspect of defensive ecology was hidden there.

Insect damage mimicry, especially of tunneling (Fig. 26.2), was the first specific ecological hypothesis that tried to explain the function, evolution and ecology of white leaf variegation. The first to discuss in detail the function of white leaf variegation in the context of possible aposematism was Smith (1986), and while he rejected the aposematic hypothesis for the species (Byttneria aculeata) he studied, he gave a clear and detailed formulation of the aposematic hypothesis for poisonous plants: “The benefits to the plant of chemical defense against herbivores would be greater if herbivores avoided such plants altogether, rather than testing leaves for palatability, and so causing some damage. A distinct leaf color pattern linked with chemical defense might function in this way. Polymorphism for leaf color should then coincide with polymorphisms for chemical defense. Mullerian and Batesian mimicry could result in evolution of similar patterns of variegation, with or without associated toxicity, among other species which have herbivore species in common with the model species”. Smith (1986) found that the variegated morph of B. aculeata was attacked less by herbivorous insects than the non-variegated one.

Defensive plant coloration, morphology and movement, i.e., visual defenses, although variable as will be described and discussed in this book, are only a small part of the complicated and sophisticated anti-herbivory defense system employed by plants. In order to give the reader the right perspective, and to add to the previous chapter, below I describe the basics of their defensive system against other organisms. I stress that there is not a single species that has all these defenses together, but that this is the arsenal that evolution has allowed plants to possess. Since a full discussion of the defense system of plants will need a monograph larger than this book, I will present the data almost in the shape of an abstract.

Concerning defensive signaling by white marking, two evolutionary issues should be specifically mentioned: first, the significance of white marking in contrast to colorful marking and second, the probable convergence of this character as expressed by geography and taxonomy. When a certain herbivore is color-blind to a certain sector of the spectrum or is color-blind altogether; or when illumination is not strong or it has an altered color temperature for some meteorological, ecological, or daytime related reason, then white marking increases the possibility that a signal will still be visible. As will be discussed below in several chapters, defensive white marking seems to be common in many plant species originating in various taxa, continents, ecologies, life forms and in many plant organs (see Lev-Yadun 2001, 2003a, 2014a, b, c, d).

The strongest advocate of the defensive role of red and yellow autumn leaf coloration as a case of Zahavi’s handicap principle was Dr. Marco Archetti. After the unexpected death of Professor Bill Hamilton in the year 2000, Marco Archetti was practically left by himself to carry the red/yellow autumn leaf handicap flag. As a Post Doc at Oxford University, he arranged a small conference on the autumn leaf color issue in Oxford’s St. John’s College in March 2008. In that conference, attended by plant physiologists, plant and animal ecologists, and several scientists that specialised in the sensory systems of herbivorous insects, the various and even contrasting views concerning the biology of autumn leaf coloration were presented. Two days of focused discussions vividly illuminated the need for a well-balanced multidisciplinary understanding, and that such a review had to be written and published by the group, and indeed, a year later it was published in Trends in Ecology and Evolution (Archetti et al. 2009a). Among other things, it became clear to all the scientists involved that yellow and red autumn leaves are in a way two different strategies (although not always and not concerning all defensive aspects) and that at least concerning the co-evolutionary hypothesis, they should be treated separately. Moreover, it also became clear that both physiology and ecolgy are involved and that the gains from autumn coloration do not stem from only one of these very different function types, and also that aposematic signaling is involved as well. This understanding was an important step to lower the flames that blazed in the arena of investigating the evolution and functions of colorful autumn leaves, but this critical understanding and even later progress did not eliminate the need for an even better understanding of the complicated scientific question of the evolution of autumn leaf colors.

Another theoretical and practical issue related to autumn leaf coloration is the issue of leaf color variability. While several authors proposed that such variability in timing and intensity of autumn leaf coloration has been selected for (Archetti 2000; Hamilton and Brown 2001; Hagen et al. 2003; Ougham et al. 2005; Sinkkonen 2006a; Sinkkonen et al. 2012), theoretically, this variability may in certain cases be a non-defensive adaptation, or not adaptive at all. Variation in the timing of color change may reflect positional constraints, such as the position of leaves within the tree, the level of exposure to sunlight or wind action, the status of the tree as dominant, sub-dominant or suppressed individual within the population, or being an understory or canopy species, or else reflect physiological pressures such as fluctuations in water supply or mineral nutrition, factors that may significantly influence a tree’s performance irrespective of its genotype (Kozlowski et al. 1991; Kozlowski and Pallardy 1997). The results of Hagen et al. (2003) that in Betula pubescens ssp. czerepanovii trees that turn yellow early in autumn (September) was negatively correlated with insect damage the following season, did not match the results of Isaksen and Folstad (2014) for the same subspecies but in different years. Isaksen and Folstad (2014) found no difference in the attacks of the most common caterpillar species Epirrita autumnata on leaves, or in the oviposition by mature females on twigs senessing early or late. Thus, it seems that at least for that taxon, the difference in succeptability to herbivory is not only related to variability in phenology, but is also year and habitat depended. Isaksen and Folstad (2014) rejected the aposematic function in that case. However, at least for various Finnish Betula pendula (silver birch) genotypes, Sinkkonen et al. (2012) showed that autumn leaf coloration varies significantly among genotypes, and that this influences fitness and evolution (Lev-Yadun and Keasar 2012). In that species, genotypes that expressed the strongest leaf reflectance in early autumn harbored more egg-laying females of the specialist aphid Euceraphis betulae than those expressing strong leaf reflectance several weeks later. Silfver et al. (2015) studied the phenology of shoot growth termination in various Finnish Betula pendula genotypes that differ in that character. Genotypes that terminated shoot growth earlier suffered from higher aphid egg loads after a short, but not after a long autumn. The Sinkkonen et al. (2012) and Silfver et al. (2015) studies illuminated the need to document the variability in autumn leaf coloration and the corresponding aphid attacks in many other taxa and under various environmental conditions.

Symmetry has been proposed to increase the efficiency of visual aposematic displays in animals (Forsman and Merilaita 1999; Forsman and Herrstrom 2004), although the role of symmetry in aposematism is not global (e.g., Langridge 2006; Stevens et al. 2008a). This hypothesis was contested with the hypothesis that asymmetry impairs the efficiency of visual aposematism while increasing the defensive value of cryptic color patterns (e.g., Forsman and Merilaita 2003). Forsman and Merilaita (2003) compared intra-individual asymmetry in color patterns of cryptic patterns on forewings and aposematic patterns on hindwings of three moth species. They found somewhat larger asymmetry in cryptic patterns but it was not significant, and therefore they concluded that there were probably developmental or genetic constraints on the evolution of pronounced asymmetry in the system they studied. Langridge (2006) found in the cuttlefish Sepia officinalis that color changes that enhanced crypsis were symmetrical while those that enhanced anti-predator signaling were asymmetrical, contrary to the theoretical predictions. These results indicate that understanding the role of symmetry in both crypsis and aposematism is not straightforward and may also be context related. Later, Stevens et al. (2009) showed that asymmetry in warning signals may not be costly to prey compared with other features of the aposematic signal, such as color and overall size. However, there is an innate preference for symmetry in the visual system of animals, probably because of the need to recognize objects (Enquist and Arak 1994), a well known fact from pollination biology (Neal et al. 1998).

An obvious question concerning visual aposematism is that of the existence of Mullerian and Batesian mimics. When in a given habitat the proportion of aposematic thorny and spiny plants increases for a period long enough for an evolutionary change, Mullerian mimicry may lead to the establishment of defense guilds (see Waldbauer 1988). This question was examined at both a specific flora (of Israel) (Lev-Yadun 2003a, 2009b; Ronel et al. 2009; Ronel and Lev-Yadun 2012) and with a global perspective (Lev-Yadun 2009c, 2014d) and indeed, Mullerian mimicry seems to be common within the group of thorny, spiny and prickly plants.

A major obstacle to studying defensive or functional plant coloration is the fragmentary and inconsistent descriptions of plant coloration, especially of vegetative organs and fruits. Taxonomists have usually refered only to flower colors and even this character has not always been described in full. Thorn, spine and prickle color, unripe and ripe fruit color, leaf colors (adaxial, abaxial, leaf margins, veins, petioles, when young and when old), bark color and color changes in all these organs were usually not systematically described if at all. The significant progress in understanding the defensive role of coloration in zoology and the basics of the involved genetic mechanisms in a few selected species took over a century (e.g., Majerus 1998; Ruxton et al. 2004; Caro 2005; Hoekstra 2006; Stevens and Merilaita 2011; Zhan et al. 2014), and the effort needed to reach the same level of progress in botany is certainly not smaller, and is probably even greater, because many plant pigments also serve various physiological functions and non-defensive signaling functions.

Herbivores use visual cues and signals as part of their complicated, multifactor way of selecting food plants (Arnold 1966; Krueger et al. 1974; Cahn and Harper 1976; Rausher 1978; Prokopy and Owens 1983; Reeves 2011). When discussing defensive plant coloration or any visual plant defense, it should be clear that the relevant herbivorous animals can see such visual cues, signals, deceptive coloration of various types, and distinguish (or not in crypsis) between these and backgrounds. It has been hotly debated whether mammalian or insect herbivores see colors or not, and if they do, which colors and to what degree. Reviews of this issue have provided conclusive proof that they do see colors (Jacobs 1993; Dafni et al. 1997; Kelber 2001; Kelber et al. 2003). While most large mammalian herbivores don’t see colors the way trichromatic humans do, there are still good reasons to conclude that even with the dichromatic vision characterizing many large mammalian herbivores, colorful plant parts look different from regular green tissues because of their hue, saturation, contrast, or brightness (see Sumner and Mollon 2000a, b; Kelber et al. 2003). The herbivores may also use reflectance cues of the colorful plant parts (see Schaefer et al. 2006; Motoyoshi et al. 2007). This view is supported by the case of “color-blind” cuttlefish that successfully camouflage themselves in various backgrounds (Marshall and Messenger 1996) and by the importance of the luminance contrast component of aposematic and other types of defensive plant coloration even for color-blind animals (Prudic et al. 2007). In any case, colors can be seen by animals both as chromatic (wavelength related) signals and achromatic (intensity related) signals (e.g., Schaefer et al. 2006) and this issue should also be studied on a larger scale in the context of visual plant defenses. Herbivore’s character of not seeing part of the spectrum may also be used by plants for defensive coloration. For instance, red appears black or very dark to certain insects. In such cases, “black” red coloration can be used to look dead, for mimicry, camouflage etc., while that red coloration can be used for signaling to animals that see red, or for physiological functions.

After it became clear to me in February 1996, following compelling evidence in the field that aposematic coloration characterizes many thorny, spiny and prickly plants, I conducted a thorough BIOSIS database search of “aposematism in plants” that yielded nothing. In those pre-Google and Google Scholar days it appeared to be a totally new issue. There were, however, several papers published since the 1960s mentioning toxic plant aposematism via visual and olfactory signals, but they were data-base-wise transparent and I have found them with time, following my very extensive library work. Concerning visual aposematism of thorny, spiny and prickly plants is was indeed a new hypothesis. Twenty years of thorough library study resulted in a very short pre-year-2000 list of authors who discussed visual aposematism (usually very briefly) and only concerning poisonous plants (Cook et al. 1971; Hinton 1973; Harper 1977; Wiens 1978; Rothschild 1972, 1980, 1986; Harborne 1982; Williamson 1982; Knight and Siegfried 1983; Smith 1986; Lee et al. 1987; Givnish 1990; Tuomi and Augner 1993). Moreover, several of these references (Knight and Siegfried 1983; Smith 1986; Lee et al. 1987), while discussing the theoretical possibility of aposematic signaling by toxic plants, dismissed the existence/operation of aposematic coloration in the plants they studied. While these few early mentions of visual aposematism in plants referred to poisonous ones, in papers published since 2001 more attention was given to visual aposematism of thorny, spiny and prickly ones (Lev-Yadun 2001, 2003a, b, 2006a, 2009a, b, c, 2011, 2014a; Midgley et al. 2001; Midgley 2004; Lev-Yadun and Ne’eman 2004, 2006; Rubino and McCarthy 2004; Ruxton et al. 2004; Speed and Ruxton 2005; Halpern et al. 2007a, b, 2011; Lev-Yadun and Halpern 2008; Fadzly et al. 2009; Lee and Gould 2009; Lev-Yadun and Gould 2009; Lev-Yadun et al. 2009a; Nassar and Lev-Yadun 2009; Ronel et al. 2009, 2010; Schaefer and Ruxton 2011; Ronel and Lev-Yadun 2012; Kavanagh et al. 2016; Midgley et al. 2016) and less attention to poisonous ones (Gould 2004; Lev-Yadun and Ne’eman 2004; Hill 2006; Lev-Yadun 2006a, 2009a, 2014a, d; Lev-Yadun and Gould 2007, 2009; Archetti 2009b; Archetti et al. 2009a; Lev-Yadun and Holopainen 2009; Cooney et al. 2012; Gerchman et al. 2012; Maskato et al. 2014; Hughes and Lev-Yadun 2015).

A variety of arthropods (e.g., flies, spiders, butterflies) are known to mimic bees or wasps as defense from predators (Cott 1940; Wickler 1968; Edmunds 1974; Plowright and Owen 1980; Rothschild 1984; Howarth and Edmunds 2000; Bain et al. 2007; Chittka and Osorio 2007; Penney et al. 2012) (Fig. 61.1). Bees and wasps may be aggressive and sting, and by this deter many animals including herbivores (Breed et al. 2004).

The direct experimental evidence for the operation of aposematic coloration in plants is meager; however, the direct evidence that it operates is slowly accumulating. Cook et al. (1971) showed that poisonous gray seeds of Eremocarpus setigerus are rejected by the mountain dove. Cahn and Harper (1976) showed that sheep avoid Trifolium repens plants with leaf marks, but did not discuss aposematism. Lev-Yadun and Ne’eman (2004) showed that sheep, goats, camels, donkeys and cattle reject conspicuous green plants in the yellow/gray desert in the summer. Numata et al. (2004) found that leaves with delayed greening suffer lower levels of insect damage when they are still young. Hill (2006) showed that the Florida scrub jay (Aphelocoma coerulescens) rejects poisonous red fruits. Karageorgou and Manetas (2006) showed that red young leaves of the evergreen oak Quercus coccifera are attacked less by insects than green ones, but rejected the aposematic coloration hypothesis and adopted other visual defensive explanations (red masks the green color, it undermines insect camouflage). Similar results were found for other species growing in Greece (Karageorgou et al. 2008). Recently, additional data about the defensive operation of white variegation that mimics insect damage in leaves was published (Campitelli et al. 2008; Soltau et al. 2009). Gerchman et al. (2012) tested the question of whether colorful visual signaling for pollinators can simultaneously serve for aposematism. The conspicuous purple tufts of leaves (“flags”), which often terminate vertical inflorescences in the Mediterranean annual Salvia viridis and attract insect pollinators to the flowering patch, have an aposematic potential towards herbivores. Cooney et al. (2012) demonstrated that the width of red margins in leaves of Pseudowintera colorata correlated positively with concentrations of the defense compound polygodial, and negatively with the extent of leaf herbivory observed in the field. Furthermore, when these authors conducted feeding trials using larvae of the generalist herbivore Ctenopseustis obliquana (brown-headed leafroller moth), which are known to naturally feed on P. colorata leaves, a significant preference for leaves lacking red margins was observed under white light, but not under red light or darkness that cancelled the visual signaling, consistent with a visual, herbivory-deterrent function (Cooney et al. 2012). Field data from Finland for ergot (Claviceps purpurea) mutualism in Festuca rubra plants in grazed areas, showed that mammalian grazers indeed refrain from eating infected plants (Wali et al. 2013), thus supporting the hypothesis posited by Lev-Yadun and Halpern (2007) about ergot aposematism. Rostas et al. (2013) showed that red galls that emit various repeling volatiles are rejected by goats in choice feeding experiments. Maskato et al. (2014) studied in more detail the find by Gerchman et al. (2012) that locusts preferred green cabbage over anthocyanin-rich red cabbage. They found that female Pieris butterflies avoid laying eggs on anthocyanin-rich red foliage, suggesting that red color is at least a visual cue affecting oviposition behavior. They reared P. brassicae larvae on green and red cabbage leaves, to determine whether foliage color reliably indicates host plant quality. They did not find a difference in survival rates or maximal larval body mass in the two food treatments. However, larvae feeding on red cabbage leaves exhibited significantly lower growth rates and longer durations of larval development. Interestingly, this longer development was coupled with a higher consumption rate of dry food matter. The lower ratio of body mass gain to food consumption in larvae feeding on red cabbage leaves was coupled with significantly higher (ca. 10 %) larval metabolic rates. This suggested that insect development on red foliage may incur an increased metabolic load associated with detoxification of various secondary plant metabolites. Energy and oxygen allocation to detoxification could come at the expense of growth and thus compromise larval fitness as a result of extended development. Maskato et al. (2014) proposed that from an evolutionary perspective, red foliage color may serve as an honest defensive cue, and I say probably a signal, as it reliably indicates the plant’s low quality as a substrate for larval development. The possibility of simultaneous olfactory aposematism has not been tested in any of these cases.

Colors play an important role in plant/animal communication. The best-known visual communication system is color per se and color patterns (including in the UV) serving as an advertisement to attract various animals, which serve as their pollinators for their flowers, hence as gene-dispersing agents as they transfer pollen among flowers (Darwin 1877; Faegri and van der Pijl 1979; Clegg and Durbin 2003; Schaefer and Ruxton 2011; Renoult et al. 2014). The next best known system is fruit colors, which alert seed-dispersing frugivores to the ripening stage of the fruits (van der Pijl 1982; Willson and Whelan 1990; Schaefer et al. 2007, 2014; Schaefer and Ruxton 2011; Stournaras et al. 2013; Renoult et al. 2014; Karban 2015). The third system is the phenomenon of fruit flags, namely, bright autumn leaves, which may signal to frugivores the existence of fruits (Stiles 1982; Facelli 1993), a debated hypothesis (e.g., Willson and Hoppes 1986; Burns and Dalen 2002). The fourth system is found in carnivorous plants whose colorful patterns, in both visible and UV light, attract insects to their deadly traps (Joel et al. 1985; Moran et al. 1999; Schaefer and Ruxton 2008). The fifth system is the conspicuous colors of autumn leaves, found in many autumn deciduous trees and shrubs, which have been proposed as a signal of a tree’s vigor to its parasitic insects, especially aphids, and thus a case of Zahavi’s handicap principle operating in plants (Archetti 2000; Hamilton and Brown 2001; Archetti and Brown 2004; Archetti et al. 2009a), or aposematism and other types of defensive signaling (Lev-Yadun 2006a; Lev-Yadun and Gould 2007; Archetti et al. 2009a; Lev-Yadun and Holopainen 2009). The sixth system is a variety of color patterns enabling plants to mimic insects so as to reduce herbivory: mimicry of butterfly eggs (Benson et al. 1975; Shapiro 1981a, b; Williams and Gilbert 1981), or of ants, aphids, beetles, spider webs, eyes, snakes, and aposematic caterpillars (Lev-Yadun and Inbar 2002; Lev-Yadun 2009d, 2015a; Yamazaki and Lev-Yadun 2014, 2015; Aviezer and Lev-Yadun 2015). The seventh system is red leaves mimicking dead or old ones or just looking different from young ones (Stone 1979; Lee et al. 1987; Juniper 1994; Richards 1996; Dominy et al. 2002). The eighth system is young colorful leaves that attract herbivores and divert them from the more costly older ones (Luttge 1997). The ninth system is delayed greening of young leaves in the tropics (Kursar and Coley 1992; Coley and Barone 1996; Numata et al. 2004). The tenth system is based on leaves that have different coloration on their adaxial and abaxial sides, or other kinds of non-variegated color patterns in other plant parts, which undermine herbivorous insect camouflage and was proposed to deter invertebrate herbivores (Lev-Yadun et al. 2004a; Lev-Yadun 2006a, b, 2009a; Lev-Yadun and Gould 2007, 2009). The eleventh system is of dazzle and trickery coloration (Lev-Yadun 2014c). Altogether there is a wealth of known coloration types by which plants signal to animals true or false information (e.g., Faegri and van der Pijl 1979; van der Pijl 1982; Lee 2007; Lev-Yadun 2009a; Schaefer and Ruxton 2011; Karban 2015) including some that are not listed here. However, it should be remembered that there are very many aspects of these complicated color systems that are not even well described, not well tested (if at all), (see Schaefer and Ruxton 2009, 2011), and that even after more than 150 years of systematic studies of parallel aspects in zoology, we are still very far from a comprehensive understanding of such aspects in zoology. Concerning plant coloration we are still at the tittering pain stage, very very far behind even the imperfect understanding in zoology.

Several different visual defenses may serve as visual aposematism even though they mostly operate by other defensive, signaling, and physiological mechanisms. I give a short account of some of them here, since an in-depth treatment of each of them is given elsewhere in this book.

Lev-Yadun and Halpern (2008) proposed that many plant species that do not have thorns, spines, or prickles have an alternative of one out of two types of usually internal, but sometimes external, sharp, microscopic defensive structures: silica needles and raphids (which are needles made of calcium oxalate). Plant silica bodies (phytoliths) are formed by an ordered biological deposition of silicon that enters the plant via the roots (Richmond and Sussman 2003). Silica bodies have several known functions including structural, optical, serving as co-factors in the detoxification of heavy metals, and defense from herbivory (e.g., Richmond and Sussman 2003; Wang et al. 2004; Katz 2015; Meharg and Meharg 2015). Lev-Yadun and Halpern (2008) discussed their specific potential biological warfare defensive function: enabling the penetration of microorganisms into the bodies of herbivores.

A theoretical issue related to both the functionality of defensive plant coloration as well as to any other anti-herbivory defense, and to the operation of herbivores under stress, which has been discussed above, is the question of risk evaluation in ecological and evolutionary plant/herbivore studies (e.g., Lev-Yadun 2006b, 2009a; Lev-Yadun and Gould 2007, 2009). There is a general difficulty in understanding this issue, which became clear to me first from reading the literature and later following personal discussions with other scientists and while attending international conferences. In order both to change this common and in many cases erroneous view, and to enhance relevant theoretical and experimental studies, I will elaborate on this issue. This is not just a mere theoretical issue, rather, the understanding of this principle may significantly influence the planning and interpreting of both experiments and actual herbivory data collected in nature. I stress that there are inherent theoretical difficulties in evaluating the quality of defense by experiments, by theoretical modeling, and especially in field observations when the issue of risk evaluation is not considered. A good plant defense may operate so well towards certain herbivores, or the defense of an animal towards predators, that there are almost no attacks on the defended organism, and thus, the experimental or observational noise may be bigger than the signals. A well-defended plant may have so few attacks that the statistical analysis of experiments may be highly problematic if not impossible.

Trichomes, the unicellular and multicellular appendages of the epidermis (Fahn 1990) are well known for their multiple functions. Trichomes may serve in protecting plants from excess sun irradiation of various wavelengths, including UV (Fahn and Cutler 1992; Manetas 2003), secrete toxic ions, especially in saline habitats (Fahn 1988a, b), function in water absorption (Fahn and Cutler 1992), reduce transpiration (Fahn and Cutler 1992; Werker 2000), be involved in temperature regulation (Wagner et al. 2004), defend from insect or other herbivorous invertebrates by reducing accessibility or by the actual trapping of their legs or by various chemical means (Levin 1973; Fahn 1979, 1988a, b; Werker 2000) and defend from large herbivores when they sting for instance in Urtica (Thurston and Lersten 1969; Levin 1973; Fahn 1990; Wheeler 2004; Fu et al. 2006). In addition, in certain carnivorous plants such as Drosera and Dionea, they may take part in the attraction, capture and digestion of insects (Juniper et al. 1989; Fahn 1990).

Another basic defensive aspect I discuss is that no plant defense system is perfect, an issue already stated long ago (Rothschild 1972; Janzen 1979; Crawley 1983). Assuming that a perfect defense system could exist is naive. All types of defense probably have a cost, although the cost can be mitigated by multifunctionality of some of the defenses and following several types of non-defensive gains by these characters. An organism has to pass through an evolutionary filter and make physiological, developmental and ecological decisions on how much defense can be achieved with the existing resources at any point in time and life cycle (Herms and Mattson 1992). This implies that various defenses must be age, cell type, organ and ecology dependent (see Ochoa-Lopez et al. 2015), the defense has to change with plant ontogeny as a whole, and according to organ age and condition (Boege and Marquis 2005; Barton and Koricheva 2010).

General visual bee mimicry and specific chemical mimicry by flowers to solitary female bees or wasps are well known in several orchid genera, for example, the Mediterranean genus Ophrys, the Australian genera Cryptostylis and Chiloglottis, and the South-African Disa. This mimicry has been shown to attract solitary male bees or wasps, which are their species-specific pollinators. The visual and chemical signals are considered to be a type of deceptive pollination mechanism based on mimicry for the exploitation of perceptual biases of animals. We propose that in addition to this unique pollination mechanism, these plants exhibit another, rarely mentioned and practically forgotten, non-exclusive function of bee or wasp mimicry (Batesian mimicry). This mimicry may deter large mammalian herbivores, and possibly also insects from the plants and especially from their flowers by a type of visual and olfactory deceptive aposematism. While visiting the flowers, bees and wasps may add a Mullerian effect to this defense. We extend this hypothesis to many other rewarding flowers that are bee or wasp pollinated and propose that abundance of pollinating bees or wasps may deter herbivorous mammals and insects from the plants during their peak flowering season.

Reconstructing the evolutionary history of crop plants is fundamental for understanding their adaptation profile and the genetic basis of yield-limiting factors, which in turn are critical for future crop improvement. A major topic in this field is the recent claim for a millennia-long ‘protracted’ domestication process. Here we evaluate the evidence for the protracted domestication model in light of published archaeobotanical data, experimental evidence and the biology of the Near Eastern crops and their wild progenitors. The crux of our discussion is the differentiation between events or ‘domestication episodes’ and the later following crop evolutionary processes under domestication (frequently termed ‘crop improvement stage’), which are by definition, still ongoing. We argue that by assuming a protracted millennia-long domestication process, one needlessly opts to operate within an intellectual framework that does not allow differentiating between the decisive (critical) domestication traits and their ...

To examine whether spininess evolved at random or differently in various life forms and plant organs, we characterized spiny, thorny and prickly organs in the entire wild flora of Israel (294 such species). Of the species, 63.3% defended their reproductive organs (the most-defended organ) and 67.0% defended various non-reproductive organs. Ninety-three species defended both their reproductive organs and at least one other part; 48.3% defended their leaves and 36.4% their stems and branches. Spiny wings defended stems and branches only in herbaceous (annual or perennial) species. There were clear differences between the life forms. Annuals and perennial herbs defended mostly their reproductive organs (95.7 and 83.0%, respectively), dwarf shrubs defended mostly their leaves (54.2%) and shrubs and trees mostly their branches (89.7 and 76.2%, respectively). Trees do not defend their reproductive organs by associated sharp appendages. The differences in defence on various organs among different life forms may influence the results of meta-analysis studies of the optimal defence allocation if such differences are not taken into account. We noted spine, thorn and prickle colours for 167 species with yellow, red, orange and white being the dominant, supporting hypotheses about spines being visually aposematic. © 2012 The Linnean Society of London, Botanical Journal of the Linnean Society, 2012, 168, 344–352.

Plant fibers are one of the most important renewable resources, used as raw material in the paper industry, and for various textiles and for composites. Fibers are structural components in timber and an energy-rich component of fuel-wood. For the plant itself, fibers are important in establishing plant architecture, as a source of mechanical support, in defence from herbivory, and in some cases as elements with contractile properties, resembling those of muscles. In addition, fibers may store ergastic carbon resources and water. Here, we review various aspects of fiber development such as initiation, elongation, cell wall formation and multinuclearity, discuss open questions and propose directions for further research. Most of the recent progress in fiber formation biology, especially in cell wall structure and chemistry, emerged from studies of only a fewmodel plants including flax, Populus spp., Eucalyptus spp., Arabidopsis thaliana and hemp. Considering the enormous importance of fibers to humanity, it is surprising how little is known about the biology of fiber formation.

Red colouration is common in young and old leaves of broadleaf woody species. Assuming that leaf colours are adaptive, we examined, by comparing the colouration in young versus old leaves, the possibility that different selection agents may have operated on spring versus autumn leaf colouration. We observed spring versus autumn colouration in three very different woody floras (Finland, Japan and Israel) in order to allow for a broad ecological and evolutionary spectrum. The null hypothesis was that if the same selective agents operated in spring and autumn, it is expected that when spring leaves are red, they should always be red in autumn, and when spring leaves are green, they should be green or yellow in autumn. We found that green spring leaves are almost exclusively associated with yellow leaf colour at senescence in autumn. Species with red autumn leaves almost always have at least some red colouration in their spring leaves. However, about half of the species with red spring leaves have yellow autumn leaves. Brown autumn leaves were not common in the species we studied. As about half of the species with red spring leaves have yellow autumn leaves but not vice versa , we conclude that there are many cases in which the selecting agents for spring versus autumn leaf colour were not the same.

This fascinating, concise book, of which I read one chapter per day, is part of a small wave of monographs illuminating the sophistication of plants in ways previously reserved for animals. Among these books, four [1–4] are dedicated to plants and two [5,6] discuss plants considerably although they mostly discuss animals. Karban’s book, in a way like that of Trewavas [3], is a late response to the molecular revolution in plant biology marked by the foundation of the very successful journal Plant Cell in 1989 and by the emergence of Arabidopsis thaliana as a model plant, two things that significantly influenced Karban’s work.