Plant intelligence dux: a comprehensive rebuttal of Kingsland and Taiz

Intelligence is a fundamental property for all life enabling an increased probability of survival and reproduction under wild circumstances. Kingsland and Taiz (2024) think that plants are not intelligent but seem unaware of the extensive literature about intelligence, memory, learning and chromatin topology in plants. Their views are consequently rejected. Their claim of fake quotations is shown to result from faulty reasoning and lack of understanding of practical biology. Their use of social media as scholarly evidence is unacceptable. Darwin’s views on intelligence are described, and their pertinence to the adaptive responses of plants is discussed. Justifications for comments I have made concerning McClintock and her “thoughtful” cell, von Sachs writings as indicating purpose (teleonomy) to plant behaviour, Went and Thimann’s allusions to plant intelligence and Bose legacy as the father of plant electrophysiology are described. These scientists were usually first in their field of knowledge, and their understanding was consequently deeper. The article finishes with a brief critical analysis of the 36 scientists who were used to condemn plant neurobiology as of no use. It is concluded that participants signed up to a false prospectus because contrary evidence was omitted.

My contributions to the topic of plant intelligence began in 1999 when I published an invited article entitled “How plants learn” (Trewavas 1999). In 2002, whilst in conversation with one of the editors of Nature, I mentioned I was writing an article on plant intelligence and was invited to provide a very short note, on this issue that duly emerged (Trewavas 2002). In 2003, I published a longer invited article entitled “Aspects of Plant Intelligence” (Trewavas (2003) which elicited some critical reaction (Trewavas 2004; Firn 2004). Currently, Trewavas (2003) has over 670 citations indicating interest.

The primary problem in recognising plant intelligence arises because we ourselves are animals and see the natural world through animal-tinted glasses. Animals usually have to move to find food. The animal timescale is very uniform and makes it easy to casually identify intelligent behaviour. Due to the need to obtain water from the soil, however, evolutionary pressures for the individual plant to move around were absent, giving rise to the impression of plants as still life. Experimental investigations indicated the falsity of that view but it still remains the public assessment. In 1990, my research group developed the aequorin method for detecting changes in cytoplasmic calcium (Knight et al. 1991). The results were a shock; simple signals such as touch, blue light and others instituted changes in less than an estimated tenth of a second. The following transient lasted only some 30 s and was entirely similar, if not identical, to those observed in animals. Animals and plants were not so different after all. The thinking that followed gave rise to the three papers indicated above and to the slow realisation that intelligence was one consequence. In Trewavas (2003), I characterised plant intelligence as adaptively variable behaviour during the lifetime of the individual (Stenhouse 1974). It was relatively easy to show that changes in cytoplasmic calcium were an essential first step in subsequent phenotypic changes. Scherzer et al. (2022) is a recent nice example. Adaptive responses were identified by Dobzhansky (1956) as behaviour increasing the probability of survival. Whatever the subsequent processes were, they had survived from the origin of plants several billion years back in which an amoebic ancestor captured a blue-green alga that became symbiotic and with time changed into a chloroplast.

The identification of plants as intelligent organisms is now widespread and reflects as much a growing awareness of how mankind has destroyed so much of the natural environment. Dignifying plants with intelligence capabilities is just one arm in attempts to increase conservation and return unnecessarily damaged environments to natural circumstances. It was used for the formulation of Swiss law on the respect to be shown to plants and to their dignity. We depend on plants for so many services that this is fast becoming essential. Inevitably negative reactions occurred, and plant intelligence has received its share of criticism, but continued research is indicating the errors in these claims.

According to Kingsland and Taiz (2024), “Proponents of the concepts of plant intelligence…depend heavily on the rhetorical use of historical sources as evidence to argue that eminent past scientists have supported ideas of plant intelligence, memory, learning, decision-making and agency”. Historical sources include writings by Charles Darwin, von Sachs, Went and Thimann, Mcclintock and apparently Lamarck. Kingsland and Taiz (2024) conclude there is a consistent pattern of distortion, including the use of fake quotations and alterations of quotations (apparently leaving out apostrophes with good reason), “a consistent pattern of distortion which suggests confirmation bias”. I reject the whole of their paper and over-dramatised claims without substance. I cannot find the science in it because the overwhelming evidence for plant intelligence has simply been ignored; the above scientists were quoted for other reasons and I do not depend on historical figures. It is easy to make such charges if you do not know the current literature or fail to read the sources claimed to better understand what they do meant Frankly, historians study the past; when dealing with the present, they are often out of their depth. This group centred around Taiz and Robinson never offers constructive criticism, as discussed later, but entirely negative attitudes.

Primary evidence for bacterial intelligence is to be found in Trewavas (2014. p201-210), and there have subsequently been more recent additions confirming bacterial intelligence and intelligent behavior (Pinto and Mascher 2016; Majumdar and Pal 2017; Westerhoff et al. 2014; Lyon 2015; Koshland 1980). Allmann (1999. p3-8) lists out the brain-like functions in unicellular organisms. Learning, memory, intelligence, decision-making, quorum sensing and self-awareness are all indicated to be present in these and previous articles on bacterial behaviour.

Westerhoff et al. (2014), who argue intelligence from a systems perspective, also include protozoan intelligence. Frances Darwin, son of Charles, stated in 1908 that his father was fascinated by the similarities between the behaviour of the lower animals and plants. Apart from quorum sensing, similar behaviours to bacterial intelligence were identified in protozoa (Amoeba, Paramecium, Stentor (Jennings 1904, 1906). These organisms, as do plants, respond adaptively to challenging environmental circumstances. Stentor uses trial and error learning in constructing intentional behaviour (Dexter et al. 2019; Tang and Marshall 2018; Jennings 1904, 1906). Physarum, a single coenocytic cell, has recently been characterised as behaving intelligently in a variety of circumstances (summarised in Trewavas 2017). Binet (1897. p3), who went on later to construct the first intelligence test to identify children with special educational needs, authored a textbook on protozoan behaviour and concluded that they were intelligent organisms. Jennings (1906 p334) arrived at an identical conclusion. Research on single cell plants, like Chlamydomonas, confirms their adaptive abilities to environmental stresses (de Carpentier et al. 2019; Shetty et al. 2019; Sasso et al. 2018) and behavioural changes in response to environmental stress. The plant/animal divide is assumed to have occurred when an amoebic ancestor engulfed a blue-green bacterium which eventually evolved into a chloroplast 2 billion years ago (McFadden et al. 1994). Intelligent behaviour was thus present in the amoebic primordial plant cell. Kingsland and Taiz (2024) have to assume that such behaviour was lost some time later by the evolving plant despite the environmental challenges which needed intelligent responses, particularly in the subsequent advances onto the land.

Amoeba (de la Fuente et al. 2019), Paramecium (Gershman et al. 2021) and brainless jellyfish (Botton-Amior et al. 2023) also learn from experience and have also demonstrated the abilities, like Pavlov’s dogs, to learn by association. Such abilities indicate that single cells can remember a previous stimulus and then flexibly join it together with another environmentally remote signal. Learning by association, although rarely looked for, is present in plants (Goh et al. 2003; Meyer et al. 2014), and adaptive behaviours are found throughout the plant kingdom. The reason that there are not more examples of learning by association may be that the authors receive unwarranted negative attacks when they do attempt to publish such information. The final stage of learning by association is present in plants; two remote signals, now joined together, has already been referenced (Trewavas 2021).

Neural networks, a term that originated in artificial neural networks, are now recognised to be present in cells arising from the formation of complexes of numerous interacting proteins whose equivalents are present in all organisms (Bray 2009). Bray (2009) quotes McClintock’s 1984 “thoughtful cell” on his frontpiece (confirmation bias or respect?). Neural networks are discussed in greater detail in Trewavas (2014: chapter 23) and arise more particularly from protein kinase networks. Golden (2004) asks participants in an EMBO conference on neural networks in bacteria to “think like a bacterium”. See the conference summary (Armitage et al. 2005) and Sarkar et al. (2021) to see its sophistication.

One key here may be the almost ubiquitous use of cytoplasmic calcium in plant and single cell learning. Chlamydomonas, a single cell plant, for example, has 37 calcium-dependent protein kinases similar in number to the 34 in Arabidopsis (Merchant et al. (2007). These can feed into a potential and complex assessment network of over 700 protein kinases in Chlamydomonas (Merchant et al. 2007) and over a thousand in Arabidopsis (Wang et al. 2020).

Throughout their article, Kingsland and Taiz (2024) use the word intelligence without stating what they mean by it. From my experience with the first introduction of plant intelligence in 2003, most plant scientists have, unsurprisingly, no knowledge of the intelligence literature. Given the wealth of present biology knowledge that must be read, this is no surprise. When faced with the word intelligence, they fall back, incorrectly, on their limited experience of human intelligence and IQ which has no relevance to plants. Chamovitz (2018), despite being a very good molecular plant biologist, exemplifies this extremely common error and was corrected in Calvo et al. (2020). Kingsland and Taiz (2024) subscribe to this common error too by claiming Chamovitz (2018) to be right.

The psychologist, Anna Anastasi (1983), distinguishes biological intelligence from the psychologist’s human intelligence. Biological “intelligence is not an entity within the organism but a quality of behaviour. Intelligent behaviour is adaptive insofar as it represents effective ways of meeting the demands of a changing environment. Such behaviour varies with the species and the context in which the individual lives. Intelligence is a pluralist concept”. This is applicable to both plants and animals. Intelligence involves changes in behaviour and the environmental conditions that elicit it. It has been summarised as (organism ↔ environment) (Gilroy and Trewavas 2023).

Behaviour is what plants do: germination, growth, flowering to seed formation and back to germination. Adaptive behaviour leads to modification of this cycle to enable it to continue when circumstances demand. In Beer (1990.preface) “Intelligence as Adaptive Behaviour”, the same quality is identified in animals. “I wish to emphasise instead the more universal ability of animals to cope with the complex dynamic and unpredictable world in which they live. To me this penchant for adaptive behaviour is the essence of intelligence”. Adaptive behaviours in plants are induced during imbalances in water supplies, light, gravity, minerals, predation, wind, stressful temperatures, predation, amongst others (Trewavas 2014) that result in phenotypic and chemotypic changes to improve the probability of survival in wild circumstances (Gilroy and Trewavas 2023). The literature here is enormous.

Anastasi (1983) also clarified the difference with human intelligence. “In the human species the influence of learning on intelligent behaviour has been immensely enhanced through the intergenerational cultural transmission of a rapidly accumulating mountain of knowledge. This influence has been further strengthened through the organised transmission of knowledge by means of formal schooling. Traditional IQ tests simply measure scholastic aptitude or academic intelligence”. The ultimate driving force of life is survival (Calvo et al. 2020). Behaviour that improves survival and its continuation to reproduce is profitable to species survival. Monod (1972) identifies teleonomy as important to these fundamental properties.

Plants should be considered complex adaptive systems because such behaviour is essential in wilderness conditions. But a common assumption is that plants are simple and are not complex. In my view, this notion arises mainly from laboratory experience in which plants can be compelled to behave by choice of imposed environment. Stating that plants are composed of leaves, roots, buds and sometimes flowers is also simplistic when considering the range of forms of angiosperm species. The difficulties and failure of getting reproducible growth and cell number when using inbred seeds and exacting growth and soil conditions (Massonet et al. 2010) only indicate that in the wilderness, the variations are going to be much larger. Development is a first-order Markov cycle (Gilroy and Trewavas 2023) reflecting continual adjustment of developing form as the environment varies. Bazzaz (1996. p112-114) provides diagrams to indicate the enormous degree of phenotypic variation in Abutilon when grown in the presence of itself or Datura, Polygonum or Setaria.

A prime example of implying plants as simple organisms comes from the use of Mendel as introducing courses in genetics (Bapty 2023). Mendel’s characters subtly suggest a simplicity in genetics that easily and misleadingly encourages belief in genetic determinism. But Mendel’s genes (there is more than one gene specifying wrinkled seeds (Rayner et al. 2017; Radick 2022) are the exceptions compared to the complex realities of 99% of morphological and physiological characters that do not behave simply. Most traits in plants use multiple interacting gene products.

The observation that Mendel, in his will, requested all his seeds and plants should be destroyed should at least raise a question, when taught. Why did he insist on this when his results were apparently so clear-cut or did later experiments on other races of Pisum sativum indicate an apparent failure? Whilst Mendel’s genes have been identified, they seem apparently to be no longer in the same race of Pisum. Mendel may simply have been lucky (Rayner et al. 2017). Other plant breeders in the early twentieth century were unable to repeat many of Mendel’s simple observations (Radick 2022). Mendel’s seeds had been domesticated for many years and thus generated a specific race. Enormous numbers of domesticated races can be generated from both plants and animals by animal and plant breeders as indicated by Darwin (1859).

Teaching of Mendel for introductory genetics can quickly lead to the perception that plants are therefore simple too. And if students are taught that all that is needed to explain any of the dozen or so tropic phenomena are marginal redistributions of auxin, then it confirms the apparent simplicity again. Textbooks tend to write about scientific idealisms! Not the reality that composes the natural world.

Kingsland and Taiz (2024) think memories do not exist in plants. Priming was a term initially used to describe the ability of plants to mount enhanced future responses from a first environmental challenge, leading to an adaptive increase in the probability of survival. The stressful situations initially described as priming were induced by chemicals, pathogens and predators. Priming is now recognised to be simply a memory that is more quickly accessed enabling higher responses on subsequent challenge (Kambona et al. 2023). Kambona et al. (2023) list 15 published examples of established memory in plants often the result of what are termed stressful conditions. Another five are referenced in Trewavas (2014. p274). There are more than 60 reported examples of transgenerational inheritance of parental experience of a variety of environmental conditions by siblings (Gilroy and Trewavas 2023). These proactive memories, improving species survival, last anywhere from 1 to 9 generations. Transgenerational inheritance represents another memory (Cao and Chen 2024; Gilroy and Trewavas 2023).

For these transgenerational memories to be formed, it requires, at the minimum, chromatin topological alterations to be applied by the parent cell. Specific chromatin structures are then retained through sexual reproduction. The probable bases of environmental memories are changes in chromatin structure and have been identified as such (Bhadouriya et al. 2021; Barrett and Wood 2008; Pei et al. 2021; Phanstiel and Wang 2022; Harris et al. 2023; Friedrich et al. 2019; Iwasaki and Paszkowski 2014; Gentrya and Hennig 2014). The mechanism of memory is similar between plants and animals (Watson and Tsai 2017). The altered form of chromatin topology takes time to implement but is then already available for subsequent responses to build on. Plants can reprogram their genomes (Cao and Chen 2024; Schapiro 2011), indicating both cells and the parent plant act as agents controlling their own behaviour and proactively those of siblings, something that Kingsland and Taiz do not apparently understand or know about.

Transgenerational inheritance is proactive behaviour on behalf of the species. Plants act as agents (Sultan et al. 2022) able to control their own behaviour and that of siblings. Transgenerational inheritances are analogous to cultural transmission of information in animals.2.Learning is well-established and is electrical in character

Memory cannot be acquired without learning, and a great deal is now known about the learning process (Trewavas 1999). In molecular characteristics, it shares strong similarity between animals and plants (Gilroy and Trewavas 2023; Luan 2011; Kudla et al. 2018; Dodd et al. 2010). Immediate receipt of almost any environmental signal (at least 20 at the minimum have been established in plants) leads to the initiation of cytoplasmic Ca2+ transients via action potentials and other electrical and chemical signals. The kinetic characteristics of the transient are different between different tissues and activated by tissue-specific receptor proteins, again specifically located adjacent to calcium channels. There are 34 genes in Arabidopsis expressing calcium-dependent protein kinases that possess substrate discrimination (Curran et al. 2011). Information is passed on to complex protein kinase networks including neural networks (Wang et al. 2020). Shortly after the initial cytoplasmic Ca2+ signal (usually cytoplasmic in appearance), nuclear Ca2+ increases (Charpentier 2018; van der Luit et al. 1999) and leads to the translocation of some protein kinases into the nucleus (Adachi et al.2000), activating enzymes or modifying their synthesis, that lead to the modification of chromatin structure.

Minorsky (2024) points out that it is only very recently (2015) that Taiz and Steiger’s textbook actually mentioned action potentials, over a century after they were identified and characterised in great detail by Bose (1926). Minorsky (2024) also points out that in their brief mention of this topic, the authors of this textbook introduce naïve readers to some serious misconceptions. Was there some block that prevented an earlier mention than 2015 in this textbook?

Debates can get heated and I suggest certain requirements would better enable sensible scientific debates and exchange of ideas. Respect between all participants. An appreciation that no one is the ultimate repository of wisdom. That different scientists have different world pictures of their understanding of plants and behaviour derived initially from training, research and extent of reading of different kinds of literature (e.g. ecology). Some textbook writers can fall into this trap and consider themselves more informed than others, not recognising that the material included in their book has gone through the sieve of their own perceptions and world picture. Textbooks reflect the biases of their authors. Dogmatic statements should be avoided.

In my view, Kingsland and Taiz (2024) fail to understand some basic elements of investigative biology. Biology is not physics with defined laws with exacting equations and highly precise circumstances needed for exacting repetition. If there is one critical feature of biology, it is variation, the source material such as a group of individuals of the same species that has to manage with variation between each of them from the other (Watahiki and Trewavas 2019). Statistics can be used to simplify but these can disguise the interesting question; what about the outliers? How do they respond? The mean value also disguises, when it is assumed to have meaning, that nature does not use identical routes; it relies on variation of mechanism and avoids the question of what causes the variation in the first place (Watahiki and Trewavas 2019). The phenomenon is called individuality and is fundamental to biology and the means of explaining it.

When writing a scientific paper (I have written some 300), there is always an introduction that explains the current state of understanding with concise reference to relevant published precedents but in the author’s own words; we do not sit down and seek out the precise form of words used by each preceding publication because it is simply not necessary. The critical point is to convey the idea and if that melds with current understanding that is sufficient.

There must be a hundred ways of explaining the overall process of natural selection; we know of at least two that advanced the notion independently, and Darwin himself, in public or with friends, relatives and other workers, unless he learned it parrot fashion, would have used different means of expression in explanation each time. But each of the hundred will be correct. What Darwin wrote down is just one formulation. None of these are fakes or exaggerations as Kingsland and Taiz (2024) try to make out. Darwin’s written version is just one of them, and he was a scientist open to admitted but important errors (Gilroy and Trewavas 2023). But should I expect a science historian to understand the realities of research life, grants, papers and meetings all of which require explanations of a main idea in words that inevitably will be different? Variation means variation in scientists too, each of whom will explain the same thing in different ways.

There seems to be a group referred to by Kingsland and Taiz (2024) and operating on social media that uses online versions of Darwin books to claim that statements attributed to him either did appear in Darwin’s book or did not. If it does not, it is claimed to be fake, not only meaningless but wrong. To me, they are wasting every scientist’s time. By doing it on social media, they avoid the critical assessment necessary for the worth of their activity. They may say but this is attributed to Darwin by those that use them; who else should they attribute it to? I only quote directly when I cannot think of an easy explanatory alternative.

I have used the statement “Intelligence is based on how efficient a species becomes at doing the things they need to survive”, which, when I saw it, reflected with reasonable accuracy my view of what I had remembered Darwin had described about intelligence. Whether he said it is irrelevant to me. I read the book on earthworms some 25 years ago to which it is applicable and is described later. By mistake, I attributed it to 1871 rather than as it should be 1881. Kingsland and Taiz describe that as fake, and exaggeration, which as indicated above, shows no understanding of practising biology. But I already knew that; two European scientists had already pointed that out to me that the date was wrong. A pity Kingsland and Taiz do not show the same level of scientific courtesy.

I have participated in two “debates” with this small group of highly conservative scientists centred around Taiz and Robinson. I find there are certain characteristics of exaggeration, error and an unnecessary indication of emotion that are revealing about the real reason for the objection to plant neurobiology and intelligence in the first place. Take, for example, the claim by Kingsland and Taiz that arguments for plant intelligence, sentience and consciousness would have been anathema to Julius von Sachs. No one knows, certainly not Taiz and Kingsland, what was anathema to von Sachs, except for the one thing which von Sachs indicated was anathema to him. That will be indicated later.