Current Biology
Magazine Feature
Could plants have cognitive abilities? Vegetation is traditionally regarded as passive, doing nothing but what is essential to grow and survive. Evidence is accumulating, however, in support of formerly esoteric notions that plants can communicate, remember, even count — features that one would call cognitive if they were observed in animals. Michael Gross reports. In the course of her adventures in Wonderland, which were first published just over 150 years ago, Alice encounters a blue caterpillar smoking a hookah and responding to her questions with very unhelpful answers and counterquestions. The caterpillar’s conversation style is said to have mocked Lewis Carroll’s colleagues at Oxford University, but its pipe smoking is more mysterious. Why would an insect smoke and how would it even be able to survive this activity, given that plants produce nicotine precisely because it is a potent toxin for insects? As it happens, Carroll’s vague description of the blue caterpillar fits the tobacco hornworm, the larval stage of Manduca sexta, which feeds on tobacco plants and is remarkably resistant to nicotine. When kept in the laboratory and fed with wheat germ, the caterpillar turns blue due to the lack of plant pigments in this diet, which would have been known to naturalists among Carroll’s contemporaries. While Carroll describes the caterpillar as sitting on a mushroom cap and enjoying his tobacco via his hookah, it is more interesting to consider what happens when it gets its tobacco fix directly from the plant. Nicotiana has the ability to respond to herbivore attack by increasing the nicotine production. If the attacker is Manduca sexta, however, the plant recognises the chemical composition of its saliva and appears to ‘know’ that more nicotine doesn’t help — it refrains from expending more energy on producing the toxin. On the population level, the nicotine-resistant pest isn’t all bad news, as the adult stage of this species is a pollinator of the tobacco plant. For the individual tobacco plant, however, the attack of a hungry caterpillar can be devastating, so it is not surprising that it has another defence mechanism in reserve: the chemicals released by damaged leaves may attract predatory mites which attack the caterpillar — a
case of tritrophic interaction, where the prey attracts the enemy of its predator. But does it make sense to use cognitive vocabulary in the context of plants? Do they know which insect is nibbling on their leaves, remember past threats, communicate them to conspecifics, or call for the help of a third party? Until a decade ago, such ideas were the domain of esoteric philosophies and considered to be close to pure fantasy, like Carroll’s famous tale and its talking and smoking caterpillar. A deeper understanding of chemical signalling of plants above and below ground is beginning to make the idea of plant cognition more respectable. Volatile signals Plants cannot speak and it is unlikely that they can listen when overenthusiastic gardeners speak to them, but over the last three decades chemical communication channels have been discovered at an increasing rate, with a growing number of recipient species found to tune in, as Martin Heil from Investav at Irapuato, Mexico, has outlined in a recent review (New Phytol. (2014) 204, 291–306). In the beginning, in 1983, David Rhoades from the University of Washington at Seattle reported that willow trees can gain resistance in the neighbourhood of conspecifics damaged by herbivores and speculated that an airborne signal molecule acted as a warning — a hypothesis that was confirmed later in the same year by other workers with poplar. Although he is now regarded as a pioneer, Rhoades was reportedly driven out of science by the hostility that his ideas faced at the time. When a plant sends out volatile molecules indicating that it is being attacked by herbivorous insects, this is a piece of public information that could be received by various interested parties, including not just other plants but also other herbivores, and indeed carnivores that may want to attack the
feeding herbivores. Thus, additional recipients of the signals were identified over the years, including predatory mites (1988), parasitoid wasps (1990), predatory bugs (1995), ladybirds and moths (2001), nematode worms (2005) parasitic plants and anatomically remote parts of the emitting plant (2006), and birds (2008). Which of these species is the ‘intended’ recipient of the signal remains to be established. While there might conceivably be an evolutionary benefit to be gained if closely related conspecifics are efficiently warned of danger, or if predators are attracted to the grazing herbivores, the relatively short reach of the chemical signal supports the hypothesis that Heil favours, namely that the signalling evolved as a communication between leaves of the same plant. These may be nearby in space and thus subject to a shared risk from herbivore attack, but still distant in the branching structure and thus difficult or impossible to reach via the internal fluid channels of the plant.
Tobacco consumer: The caterpillar of the moth Manduca sexta is remarkably resistant to nicotine. If under attack from this species, the tobacco plant appears to know this and refrains from increasing its nicotine production. (Image: British Entomology Volume 5, John Curtis.)
Current Biology 26, R181–R191, March 7, 2016 ©2016 Elsevier Ltd All rights reserved
R181
Current Biology
Magazine
Wooded web: The root space of plants is a highly complex and insufficiently understood system described as the rhizosphere. It enables communication both among plants and between plants and other species. (Photo: Nerys Hamutal Groß, https://www.flickr.com/photos/ nerysgross/)
Richard Karban from the University of California at Davis, USA, and colleagues have recently reported a detailed investigation of the specific volatiles profiles of individual plants of sagebrush (Artemisia tridentata). These authors found that there are characteristic and heritable differences in the chemotype of the plants’ warning signals. Warnings received from plants with the same or similar chemotype offer more efficient protection, which suggests that these chemotypes represent a kind of kin recognition in plants (New Phytol. (2014) 204, 380–385). Even though hundreds of herbivoreinduced volatiles have now been identified, precise information on their concentration in the air surrounding the emitting plant is still scarce, and their further fate in the environment remains incompletely understood. Most importantly, the receptor mechanism, the ‘nose’ of the plant being warned of the danger, has remained elusive. There are indications that accumulation of volatiles in the plant membranes may play a role and that epigenetic mechanisms may enable this information to be stored and passed on to the next generation, but just how R182
a plant sniffs out danger remains to be discovered. An improved understanding of these mechanisms might lead to new approaches in sustainable pest control — one of the reasons why this area is under intense investigation at the moment. Recent progress was assessed at a Gordon Research Conference held at Ventura, California, in February. Pollinator deception One highly specialised and relatively well-studied area of chemical communication by plants is the pollination strategy of sexual deception. Orchids on several continents have independently evolved the ability to mimic pheromones of female insects to attract the males. This deception is so successful that copulation attempts are frequent and the insect’s misdirected energy secures the critical step of pollination without yielding any reward for the pollinator. Researchers are still uncertain as to how this mischievous streak evolved, and why it has evolved repeatedly. A preferred hypothesis in many cases is that of pre-adaptation whereby the
Current Biology 26, R181–R191, March 7, 2016 ©2016 Elsevier Ltd All rights reserved
chemicals involved were co-opted from other functions and re-assigned, with modifications, to the task of fooling the pollinators. However, ongoing investigations into the orchid semiochemicals used by Australian sexually deceptive orchids by Rod Peakall’s group at the Australian National University in Canberra are regularly uncovering novel compounds. For example, in Drakaea glyptodon, the signalling molecules that the orchid uses to trick its pollinator turned out to be alkylpyrazines and a novel hydroxymethylpyrazine (New Phytol. (2014) 203, 939–952). The flowers have also evolved a labellum (lip-shaped petal) that mimics the shape of the flightless female wasps. In closely related Chiloglottis orchids, specific blends of unique compounds called chiloglottones are used to attract the males of only one species of thynnine wasp per orchid. Many more cases of chemical deception are yet to be analysed in detail. “Beyond these published cases we have discovered other new semiochemicals,” Peakall comments. “In fact, every Australian sexually deceptive pollination system that we explore reveals unexpected and exciting new chemical communication systems.”
Roots connection: Microscopy image from a study into the cellular mechanism of the symbiosis between plant roots and Glomus versiforme mycorrhizae. (Photo: reproduced from Curr. Biol. (2015) 25, 2189–2195.)
Current Biology
Magazine So far, pyrazines in plants are extremely rare, while chiloglottones are not known elsewhere, and no other functions in plants have been established. Therefore, the authors hypothesise that this may be a case where a chemical deception evolved via an evolutionary novelty, rather than preadaptation. The authors acknowledge, however, that proving the alternative explanation requires a comprehensive knowledge of the molecular and genetic basis of the production of these compounds. Further, an in-depth study across a range of species is required to trace back the molecular phylogeny to the point where the chemical trickery originated. Birgit Oelschlägel from the Technical University Dresden and others have recently uncovered the chemical deception that one flower from the genus Aristolochia uses to attract flies (New Phytol. (2015) 206, 342–353). This is the genus that produces some of the largest flowers known and some that famously smell of rotting flesh — all in the cause of deceiving flies. In the case of the species studied, Aristolochia rotunda, the volatiles are barely perceptible to the human nose, but Oelschlägel and colleagues showed that it has a very specific smell to the pollinator species, the kleptoparasitic fly Chloropidae. These flies feed on secretions from insects being killed and eaten by spiders. Aristolochia rotunda mimics the odour of these secretions to attract the flies. The system described is unusual in that the plant mimics the chemicals emitted by dead insects, sending out the false promise of a meal for the flies, rather than those of live conspecifics, signalling sexual opportunity. In contrast to the situation with bees, the food that the flies hope to find is not to feed their offspring but for themselves. As more cases of chemical communication between plants and pollinators emerge, it becomes clear that the mix of chemicals tends to be highly specific to the species pair. This suggests that the ability to emit these specific chemical signals may have played a role in speciation of the plants involved. Below ground If volatile signalling is difficult to analyse, another communication channel important for a plant’s interaction with
Five strokes: The Venus flytrap appears to count how many times its sensory hairs inside the trap have been touched. Only after five mechanical signals does the digestion of the prey set in. (Photo: © Laura Dottori.)
its environment is literally hidden from view — its root system. The importance of the underground space around the plant roots, populated by a wide range of species attracted by chemicals secreted from the roots, was first recognised by Lorenz Hiltner (1862–1923), who coined the term ‘rhizosphere’ in 1904. More than a century later, the concept has emerged as one of the most significant areas of plant science, as a special issue of Trends in Plant Science published this month demonstrates (www.cell.com/trends/plant-science/ issue?pii=S1360-1385(15)X0004-5). Below ground, as above, plants interact with their own kind and with multiple other species in complex ways. Recent progress in metagenomics has enabled researchers to characterise the rhizosphere microbiome at the genome level, establishing the wide variety of species, including both helpers and pathogens, that thrive in this environment and play a crucial role in the life of the plant. It has been shown that plants expend significant parts of their metabolic energy on substances that serve to feed the rhizosphere microbiome, which in return provides valuable services that may improve the plant’s growth, development,
nutrition, or immunity. This exchange of goods is so rich and complex that economic theories have been applied to characterise it (Proc. Natl. Acad. Sci. USA (2014) 111, 1237–1244). As the success of the interaction with the microbiome depends on the genotype of the plant, there is concern that breeding for agriculture may have reduced the suitability of crop plants for some of the underground interactions whose benefits may not be immediately obvious to breeders, such as immunity. A better understanding of the whole rhizosphere could thus help to restore natural defences of crop plants and enable a more sustainable type of agriculture (Trends Plant Sci. (2016) DOI: http://dx.doi.org/10.1016/j. tplants.2016.01.018). Progress is being made in the analysis of the chemical nature of the communication in the rhizosphere. Even small plants like Arabidopsis thaliana exude more than 100 different compounds into the soil. Sampling has to be well-considered to ensure that analyses are representative of field conditions, and the most sophisticated methods of metabolomic analytics are required to gain quantitative insights into the complex and dynamic situation
Current Biology 26, R181–R191, March 7, 2016 ©2016 Elsevier Ltd All rights reserved
R183
Current Biology
Magazine around the roots of plants, as Nicole van Dam and Harro Bouwmeester report in a review of this field (Trends Plant Sci. (2016) DOI: http://dx.doi. org/10.1016/j.tplants.2016.01.008). Direct investigation of signalling in the rhizosphere is currently focused on communication between plants and their microbial symbionts, as Vittorio Venturi and Christopher Keel report, although the scope could be broadened to include all organisms present (Trends Plant Sci. (2016) DOI: http://dx.doi.org/10.1016/j. tplants.2016.01.005). Plants may send signals to the soil microbiome to recruit beneficial species or to activate beneficial traits. In turn, microorganisms may signal to the host plant to activate defences, influence metabolism and development, or induce the stress response. Plants are also known to communicate with their neighbours through the entangled root networks. Trees, for instance, can feed their offspring through the roots and keep stumps alive. They can also trade information about threats like drought or diseases. Some forest experts have taken to the expression of the ‘woodwide web’ to characterise this hidden information exchange. Just how it works in detail remains to be explored. Memory and maths Communication may be the most important information-processing function in plants, but it is by no means the only one. Evidence is accumulating to suggest that plants can remember certain kinds of events and learn to ignore them even if they would normally trigger stress responses. Stefano Mancuso from the University of Florence, for instance, has trained mimosa (Mimosa pudica) plants to tolerate certain kinds of shocks — such as those experienced when researchers drop them on the floor — without activating their widely known leaf-folding response. After a series of 60 drops, the plants accepted this condition as normal, while still retaining their sensitivity to other unexpected events, such as being touched or shaken. The group of Monica Gagliano at the University of Western Australia at Crawley has worked with Mancuso to analyse the plant’s memory using approaches normally reserved for R184
animal behaviour and found that the learning and memory improve in energetically costly environments where these skills matter more (Oecologia (2014) 175, 63–72). In February, Rainer Hedrich from the University of Würzburg, Germany, reported in this journal that the Venus flytrap Dionaea muscipula can count the number of mechanical stimuli received via the sensory hairs inside its trap (Curr. Biol. (2016) 26, 286–295). One stimulus might come from a nontarget source, such as a fallen leaf. Further hits indicate the presence of a moving insect, and after each signal the plant steps up the response, first closing the trap, then releasing its gastric enzymes to digest the prey. By this stepwise response to the signals received, the plant ensures that it doesn’t waste energy on false alarms. If plants can apparently communicate, remember and count, can we consider these abilities as cognitive processes even in the absence of a brain? The concept certainly resonates with the zeitgeist, witness the book by the German forest ranger Peter Wohlleben, Das geheime Leben der Bäume (The Secret Life of Trees), which has spent months at the top of the non-fiction charts in Germany since its publication in 2015 and is set to appear in translation in English as well as more than a dozen other languages this year. Wohlleben popularises what plant science has established so far by unashamedly using anthropomorphisms to describe the cognitive and social interactions of plants. Thus he talks of brood care, friendship and social networks among trees, which appears to resonate with the forest-loving German readership. The risk of such language is that it may revive controversies over highly esoteric claims made in the past regarding plant consciousness, often based on a mixture of scientific insights and pure fancy. The findings made recently with the most advanced techniques that 21st century science can command suggest, however, that the cognitive abilities deserve to be taken seriously and studied further. They are not just fantasy. Michael Gross is a science writer based at Oxford. He can be contacted via his web page at www.michaelgross.co.uk
Current Biology 26, R181–R191, March 7, 2016 ©2016 Elsevier Ltd All rights reserved
Q&A
Gordon Fain Gordon Fain is Distinguished Professor of Integrative Biology and Physiology and a member of the Jules Stein Eye Institute at the University of California, Los Angeles. He did his B.A. at Stanford and his Ph.D. in Biophysics at Johns Hopkins and Harvard. He is a Guggenheim Fellow, Overseas Fellow of Churchill College Cambridge, and author most recently of the second edition of Molecular and Cellular Physiology of Neurons (Harvard, 2014). His principal interest is the physiology of vertebrate photoreceptors. Why did you decide to be a scientist? My grandfather was head of organic chemistry at the Bureau of Standards in Washington D.C. and let me play with his microscope when I visited him as a boy. I then bought my own microscope from my earnings delivering newspapers and spent many happy hours looking at pond water. I did well in my science courses in high school but, like a lot of teenagers, I read a lot and became deeply interested in poetry and philosophy. When I got to University, I hadn’t the slightest idea what I would do. I had a wonderful time at Stanford, studying poetry with Yvor Winters and philosophy with Donald Davidson; but then, at the beginning of my junior year, I took a course in calculus out of curiosity, did well on the first exam, and quickly changed my major to biology. I had decided that, irrespective of my own opinion, no one could tell whether any one poem was better than another. Philosophy, particularly of the analytical school of Wittgenstein, Moore, and Austin, had sharpened my powers of reasoning, but the real advances in human understanding seemed to come from science. As a result of my sudden decision, I never had the first quarter of freshman chemistry or the third quarter of freshman physics. Stanford looked the other way and let me graduate. What got you interested in vision? I wanted to do research in my senior year at Stanford and went the rounds of faculty looking for someone to take me on. I had the stupendous luck of being accepted by Donald Kennedy,