Predator–prey systems as models for integrative research in biology: the value of a non-consumptive effects framework Free

Predation is a major selective force that influences multiple levels of biological organization, from genes, populations and community dynamics to ecosystem structure and functioning. Predator–prey interaction (PPI) involves both direct (consumptive) and indirect (non-consumptive) effects of predators on prey. When we consider predation, the first impression that comes to mind is an organism feeding on another. However, the notion that predation is restricted to direct effects – affecting prey density through consumption – is one aspect of the story; predation also involves non-lethal or non-consumptive effects (NCEs; see Glossary). NCEs are also known as ‘trait-mediated interactions’ – indirect effects of predation risk that induce shifts in prey behavior, physiology, development, cognition and morphology, among other traits (Werner and Peacor, 2003; Peckarsky et al., 2008; Peacor et al., 2022). Predator-induced phenotypic plasticity, where organisms alter phenotype in response to predation risk, often comes at a significant cost (Tollrian and Harvell, 1999; Whitman et al., 2009). Both consumptive and non-consumptive processes are linked to fitness and affect prey populations (Preisser et al., 2005; Preisser and Bolnick, 2008). Experiments testing for NCEs and prey responses simulate predation risk without causing actual lethality using predator-associated cues. Plastic prey responses have recently been found to be transgenerational, affecting multiple generations (Liberman et al., 2019).

In this Commentary, I discuss topics related to cognition, adaptation and aging because these can be studied through both proximate and ultimate perspectives that provide a comprehensive understanding of the mechanisms and evolutionary significance of these phenomena. An all-encompassing approach in cognitive sciences, from the origin and adaptive evolution of cognition to neural and physiological mechanisms underlying cognitive processes such as fear, anxiety and learning, can be tested using the PPI model, and the advantages of an integrated approach cannot be underestimated. Similarly, insights can be gained into physiological adaptability and aging by understanding the underlying genetic and physiological mechanisms of anti-predator trait development, cellular and molecular responses to predator-induced stress, accelerated predator-induced development and its links to stress-induced aging using a PPI model. This is because the underlying basis of these traits are rooted in PPIs throughout evolutionary history, making a PPI approach highly valuable.

I begin this Commentary by providing a brief overview of PPIs and NCEs. I then highlight ongoing research across various fields in biology that can benefit from using an NCE framework to investigate important questions. This Commentary is not a detailed review, but provides a broad overview to showcase the connection among seemingly distinct biological research themes and proposes to integrate them using an overarching eco-evolutionary framework (Fig. 1).

Predation risk influencing prey behavior and physiology is well established (Abrahams and Dill, 1989; MacNamara and Houston, 1986, 1992; Brown et al., 1999; Barbosa and Castellanos, 2005; Davies et al., 2012; Chivers et al., 2012). These NCEs have been formalized through several conceptual frameworks – such as the ‘ecology of fear’ (Brown et al., 1999), ‘decision-making under predation risk’ (Lima and Dill, 1990;,Lima, 1998) and ‘trait-mediated indirect interactions (TMII) (Abrams, 2000; Peacor and Werner, 1997) – that predict how predators drive ecological processes via fear-induced prey responses. Prey organisms are constantly making decisions and face trade-offs between predation risk and fitness-enhancing survival and reproductive activities. Thus, NCEs cause changes in prey population structure, and act in parallel with direct consumptive effects (see Glossary) to shape ecosystems (Kotler & Holt, 1989; Peacor & Werner, 2001; Schmitz et al., 2004; Preisser et al., 2007). Studies highlighting the importance of perceived predation risk on prey behavior and physiology (Boonstra et al., 1998a,b; Denver, 2009; Gaynor et al., 2019; Schmitz and Trussell, 2016; Verdolin, 2006; Peacor and Werner, 2001) are widespread, but it has been challenging to develop a more precise understanding of the multiple steps involved in the manifestation of NCEs, from risk perception to the response and the consequences of that response (Sheriff et al., 2020; Wirsing et al., 2021; Zanette and Clinchy, 2019). Below, I discuss how the NCE framework is used to better understand the biological mechanisms underlying important aspects of behavior, cognition and life history (Fig. 1).

For a prey animal to express NCEs in a predator–prey system, it must first perceive the predation threat; this perception can be used as an experimental tool in studies with implications for psychology, artificial intelligence and medicine (Britton et al., 2011; Rizzi et al., 2017). For instance, fear is implicated in a multitude of cognitive disorders such as anxiety, depression and post-traumatic stress disorder (Dukas, 2004). Cognitive and psychosomatic disorders are some of the most prevalent disorders that affect personal and societal well-being (McWhirter et al., 2020), and human mental wellness has become a top priority in the fields of psychiatry, psychology, neuroscience, public health and social sciences. The fear/anxiety mechanism of the vertebrate brain has evolved to evoke fear of predators and to mitigate predation risk (Britton et al., 2011); therefore, incorporating mechanisms of predator perception and predator-induced fear into studies on mental health will greatly help in improving our understanding of fear as well as with the treatment of cognitive disorders. In their perspective paper, Clinchy et al. (2011a) discuss how fear of predators can be utilized as a model to study post-traumatic stress disorder in animals. In mammals and birds, the fear or stress caused by predator exposure leads to long-term changes in neurological functions (Jöngren et al., 2010; Mitra and Sapolsky, 2008), anxiety-like behaviors (Cohen et al., 2006; Zanette et al., 2019 for post-traumatic stress disorder), shifts in glucocorticoid levels (Roseboom et al., 2007; Clinchy et al., 2011b; Sheriff et al., 2009), epigenetic modifications (McGowan et al., 2008, 2009; Yehuda and Bierer, 2009) and reproductive effects (Zanette et al., 2011). The brain physiology of most vertebrates is similar to that of humans and is leveraged as a translational model in pharmacology. However, emerging invertebrate models can also be used to understand similar phenomena of fear-induced modifications in different phenotypic traits such as cognitive processes, and conserved molecular pathways (i.e. neuropeptides and monoamines: dopamine and serotonin) highlight the opportunities for comparative studies (Rivi et al., 2023; Pribadi and Chalasani, 2022; Curran and Chalasani, 2012). The understanding of how neuromodulators (serotonin, dopamine, neuropeptides, insulin-like peptides) impact several cognitive disorders via the regulation of behavior and neuroplasticity remains limited. The molecular pathways of neuromodulators are highly conserved across the animal kingdom, thus opening opportunities to use invertebrates as models in fear and anxiety research for future studies (Van Damme et al., 2021; Anderson and Adolphs, 2014).

Across animals, responses to predators broadly can be classified into two types – innate and learned. Learned predator recognition is more prevalent, possibly because its flexible nature accommodates the natural variability in predation risk (Ferrari et al., 2007). However, the phenotypic traits that selection acts on (for example, sensory modalities of prey and hunting modes of predators) and ecological drivers (environmental factors) favoring innate versus learned fear responses are largely unknown. Identifying such factors will help us to understand the evolution of instinct and whether instinct is influenced by learning.

Evidence from neuroscience suggests that similar neural circuits are involved in innate and learned olfactory responses in bees and flies (Galizia, 2014). Furthermore, in rodents, innate and learned fear circuits exhibit overlap (Isosaka et al., 2015). Tierney (1986) first proposed that instinct can develop from learning. Tiernay suggested that behavioral flexibility – a requirement for learning – is a phylogenetically primitive state and canalized/innate behaviors are derived from this plasticity, which is an inherent property of all nervous systems. However, the link between instinct and learning has remained largely overlooked, perhaps due to a lack of a mechanistic understanding of the neuronal, molecular and genetic pathways responsible for learning and instinct. The transition of a trait from induced to innate does occur within a PPI framework, as predator-induced traits can be genetically assimilated; the canalization (see Glossary) of the trait depends upon the level of intensity of the threat, its predictability and the costs associated with genetic assimilation (Nishikawa and Kinjo, 2018; see Glossary). Some examples of such canalization in behavioral traits are reported in frogs and snails that respond to their respective crayfish and crab predators by lowering activity levels under chronic presence of predators across ecological and evolutionary time scales (Edgell et al., 2009; Nunes et al., 2014).

An interesting approach to understanding the transition of traits from induced to innate would be to compare laboratory and wild populations of model organisms. Several species have lost the ability to recognize predators, especially in island populations (Carthey and Blumstein, 2018; Cliff et al., 2022; Jolly et al., 2018). Similarly, laboratory populations – by virtue of living under artificial conditions for several generations or having manipulated genetic backgrounds – may lose their innate ability to recognize and respond to their natural predators (Owings et al., 2001; Kelley and Magurran, 2003). In some instances, however, laboratory populations retain their innate predator recognition, as there is no selection for its loss (Batabyal et al., 2022). Therefore, comparisons between laboratory and wild populations can reveal important aspects of predator-detection mechanisms, learning and the genetic basis of innate fear (Batabyal and Lukowiak, 2023). This may prove particularly useful when considering the roles nature and nurture play in the evolution of complex traits.

Research focusing on the learning of novel predators (Griffin et al., 2001; Steindler et al., 2020) has implications for biodiversity loss due to invasive predators, which have become widespread across ecosystems. NCEs provide naïve prey with valuable information about the presence or risk of predation through multiple cues that predators emit. These can include visual (e.g. predator body shape or coloration), olfactory (e.g. predator scent or pheromones), auditory (e.g. predator vocalizations) or indirect cues (e.g. alarm cues from injured conspecifics or alarm calls of other prey species). These cues can trigger a series of behavioral, physiological and cognitive responses that aid in recognizing and responding to a novel predator, as prey associate some standard threat cues with the novel predator presence. However, learning about novel predators is a complex process, as it includes multiple modalities from cognition, behavioral signaling (displays), physiology, gene interactions and epigenetic modifications (Abrams, 2000; Batabyal et al., 2014; Wennersten and Forsman, 2012; Whitman et al., 2009; Wirsing et al., 2021). Comparative phylogenetic studies may help us to understand the evolution and complexity of a predator-recognition template in prey, and studies on predator-detection mechanisms in the laboratory will help us to determine the mechanisms by which such a template may be activated or deactivated using environmental cues (Carthey and Blumstein, 2018).

Predator-induced fear can be leveraged to understand another fascinating topic with broad implications – memory formation. Many species exhibit differences in neuronal firing during the formation of memories associated with fear (Orr et al., 2007; Han et al., 2007). Such fear-induced memory formation involves epigenetic modifications such as DNA methylation (Forest et al., 2016; Poon et al., 2020). Thus, tissue- and region-specific epigenetic modifications can be linked to the pattern of neuronal activity during both fear-memory formation and the anti-predatory behavior of the animal (see Glossary). Using PPI systems will therefore help us to understand the molecular and cellular mechanisms underlying fear-memory formation, including changes in synaptic plasticity, gene expression and neurotransmitter systems that alter brain regions involved in fear processing, such as the amygdala, hippocampus and prefrontal cortex (Lubin et al., 2008; Bredy et al., 2007). Investigations on the ecological and environmental triggers and long-term changes in cognitive and behavioral processes would elucidate how such changes become fixed in populations and whether these changes are transgenerational.

To decode cognitive function, understanding the structure and function of individual brain areas is essential, but understanding how the different areas interact in making decisions and responding to stimuli is also crucial. The emergent field of brain connectomics attempts to link brain functions or the resulting behavior with the structural network of neurons. Numerous studies have indicated that a single or a small number of neurons interact in a tight network, allowing processing of detection-response pathway during anti-predatory behavior (Gazzola et al., 2015; Herberholz et al., 2004; Orr et al., 2007; Edwards et al., 1999; Allen et al., 2006). However, anti-predatory behavior involves multiple brain areas (hypothalamus, amygdala, hippocampus and prefrontal cortex among others; Lubin et al., 2008; Bredy et al., 2007; Trogrlic et al., 2011), and dynamic interactions between the predator and the prey inform decision-making during the entire event, with input from past experiences. Whole-brain recordings in smaller animals, such as Caenorhabditis elegans, Drosophila melanogaster and larval Danio rerio, have provided opportunities to capture the neural activity in sensory, motor and decision-making pathways (Barsotti et al., 2021). By combining NCEs with brain connectomics we can now understand the influence of environmental stress on neural function, development and dynamics. This will allow us to test whether structure and function of neuronal circuits are correlated, and understand how stress shapes brain connectivity and vice versa. Additionally, combining NCEs with connectomics will help elucidate the level of conservation of the neuronal circuits involved in eliciting fear responses, and we can ask whether the neuronal circuitry eliciting fear responses changes with chronic exposure to fear stimuli through synaptic plasticity and, if so, how? Only our imagination seems to limit the questions such research can answer, and the use of PPI model systems promises to reveal unprecedented information on brain function.

Connectome information has implications for human health, as it will greatly help the understanding of phobias, psychological disorders, stress and mental well-being. Some promising models that might be considered in this regard are the invertebrate nematode C. elegans and the vertebrate zebrafish D. rerio, as their brain connectomes are already mapped (DiLoreto et al., 2019; Hildebrand et al., 2017; Jarrell et al., 2012; White et al., 1986; Witvliet et al., 2021) and their predation ecology is well studied (Gerlai, 2010, 2013; Oliveira et al., 2017; Parker et al., 2013; Pirri and Alkema, 2012). Experiments with these organisms making use of predator exposure at different developmental stages could reveal how the interaction between genes and the environment shape brain connectivity. However, the maximum potential of brain connectomes lies in the transformative role they can play in our understanding of brain development as a function of early-life experience/stress. In humans, chronic stress or stress experienced early in life is a major risk factor for developing depressive disorders and emotional dysregulation (Kessler, 1997; Kessler et al., 2003). Studies performed in humans and rodents have recently begun to use MRI-based assessment tools to characterize whole-brain structural changes caused by stress (Singh et al., 2013; Young et al., 2017). The aim of this work is to map synaptic changes as well as brain network reorganization patterns that can signal potential risk markers for depression. Advances in connectomics will allow complex structural and functional brain dynamics to be mapped in vertebrate species using the non-consumptive PPI model – perception and response to threat involve the corticolimbic circuit, comprising the hippocampus, amygdala and prefrontal cortex (Hariri and Holmes, 2015; Herman, 2020), all of which are also major centers in predicting stress-related psychopathology in humans (Satterthwaite et al., 2016; Yang et al., 2010). Finally, it is exciting that the development of the brain can now be tracked, allowing us to understand how functional attributes emerge from structural complexity, providing a glimpse into what has thus far been a neurological black box.

The world is undergoing unprecedented change in terms of habitat modification and climate change, exposing species to varied stress (Malcolm et al., 2006). A species' capacity to adapt to such changes depends upon existing genetic variation, the age of the lineage and the ecological conditions that the lineage has experienced in its evolutionary history (Gaitonde et al., 2018; Ghalambor et al., 2007; Lande, 2009; Sentis et al., 2018). Adaptation is the essence of evolution, producing the remarkable diversity of life that we observe today, and predation is one of the strongest selective pressures driving adaptation and diversification. When prey animals perceive risk of predation through NCEs they exhibit adaptive behaviors that reduce their vulnerability or increase their survival. For example, prey may alter their foraging strategies to avoid risky areas and adjust their reproductive behavior by altering mating preferences or timing, reducing predation risk for themselves or their offspring. Additionally, prey may develop physiological adaptations, such as heightened sensory acuity or enhanced anti-predator defenses, in response to the presence of predators. Therefore, to understand adaptation it is important to understand the roles played by NCE and plastic responses to stress.

Behavioral coping styles are adaptive strategies exhibited by individuals in response to various environmental challenges or stressors. These coping styles represent consistent, individual differences in behavior, reflecting a suite of behavioral, physiological and cognitive responses to stress (Roche et al., 2016; Abbey-Lee et al., 2019). Personality traits (boldness and shyness) often co-vary with each other (e.g. boldness is positively correlated with aggressiveness), and consistent covariance of traits across contexts is termed a ‘behavioral syndrome’ or ‘coping style’ (Bell, 2007; Koolhaas et al., 1999; Sih et al., 2004). Different coping styles may be favored depending on the ecological context, resource availability, predation pressure or social dynamics. By adopting a particular coping style, individuals can increase their chances of survival and reproductive success within their specific ecological niche (Ferrari et al., 2016; Navas González et al., 2018; Øverli et al., 2007; Wong et al., 2019). The NCE framework approach encompasses the effects of predator cues on prey animals, triggering behavioral and physiological responses. Establishing a relationship between physiology and the molecular processes underlying behaviors (such as boldness, exploration, learning, problem-solving and anxiety) will help predict which phenotypes may be adaptive under a given environment. This may be accomplished by investigating the underlying gene expression or epigenetic changes that determine tolerance ranges under different environments. For example, research on multiple ecotypes of certain species, such as the three-spine stickleback (Gasterosteus aculeatus) (Reid et al., 2021) or guppies (Endler, 1995; Smith and Blumstein, 2010), are being used to investigate complex traits. A recent study on sticklebacks reported that gene expression patterns of dopaminergic, serotonergic and stress pathways can predict boldness and exploration behaviors (Abbey-Lee et al., 2019), but other studies indicate that a certain degree of ‘flexibility’ exists for all individuals under risk. Therefore, studies profiling genomic markers of behavioral types are needed to significantly predict how the genetics of these underlying behavioral types contribute to specific traits (Bensky et al., 2017; Fürtbauer et al., 2015). An annotated genome, linkage maps, allelic variation and gene networks are well characterized in three-spine stickleback and other organisms, and can be used to study individual and population-level variation in behavior. For example, differences in risk-taking behavior were observed in bold and shy European sea bass, and the personality type had molecular signatures wherein bold individuals showed greater gene expression for social and exploratory behaviors and memory encoding genes in the pituitary gland (Sadoul et al., 2022). Researchers can employ the NCE approach by simulating risk of predation to investigate how NCEs shape the development, prevalence, consistency and plasticity of coping styles within populations.

Another interesting aspect of the NCEs of predators lies at the intersection of sexual and natural selection. Often, chronic or acute predation risk leads to shifts in secondary sexual traits such as behavior or morphological signals (Frommen et al., 2022; Cummings et al., 2008). Life-history theory suggests that the evolution of different reproductive strategies depends on predation risk (among other environmental factors; Candolin, 1998; Wolf et al., 2007). For example, guppies (Ruell et al., 2013), cichlids (Meuthen et al., 2018) and sticklebacks (Candolin, 1998) show shifts in ornamentation (coloration and visual signals) under varying predation risk. Therefore, predation risk can alter the strength and direction of sexual selection by causing shifts in mate choice. The influence of chronic predator exposure during development on (1) the expression of secondary sexual traits, (2) mate choice and (3) generation of distinct mating patterns needs to be studied (Frommen et al., 2022; Kunte, 2008; Ruell et al., 2013; Torsekar and Balakrishnan, 2020; Winandy and Denoël, 2015).

To test the influence of predators on the development of sexual traits, zebra finches (Taeniopygia guttata) could be used, along with several piscine models (sticklebacks, guppies and cichlids, among others). In the zebra finch, mate choice is linked to exploratory behavior and novel personality attributes (Schuett et al., 2011). These behaviors are also influenced by predation risk, enabling investigations into the interplay between sexual and natural selection in the evolution of novel traits. Song learning is another behavior that is well studied in the zebra finch; however, the effects of environmental stressors such as predation, energy state of the individual and personality type on the neurology of song learning need be explored (Rosenthal, 2017). Inducing predator presence and using the NCEs paradigm, effects of environmental stressors on the evolution of female choice, social dynamics, mate guarding behaviors and mating systems can be understood and individual and group processes can be linked.

Eliminating or delaying aging is one of the most cherished ambitions of humanity since time immemorial, and the use of NCE systems could be of value in enhancing our understanding of aging. There is a large body of research on the causes and mechanisms of aging (Rose et al., 2008; Tobler et al., 2022). However, if any novel insights are to be gained, these causes and mechanisms need to be combined into a unified framework. The prominent mechanisms of aging are somatic wear and tear, oxidative stress, caloric intake, mutation load, decreased telomere length and other genetic, cellular and physiological processes (Bwiza et al., 2019; Weinert and Timiras, 2003). Studies on causes of aging have emphasized life history evolution (such as lifespan and age-dependent mortality), body size and phylogenetic background as factors that determine the onset and duration of aging (Bwiza et al., 2019). It is evident that both the parameters of aging and the mechanisms can be selected by the environment, with predation as an important factor.

So how could NCE systems play a major role in increasing our understanding of aging? It is known that chronic predation stress can lead to cessation of growth and reproductive activities, and suppression of immune function (McEwen and Wingfield, 2003; Wingfield, 2003). When these processes are turned off for an extended period, the consequences are suppression of cell growth and tissue repair, increased cell death, greater susceptibility to infection and accelerated aging (Noguera and Velando, 2019; Sapolsky, 2000). Thus, PPI systems that involve NCEs could be used as a model of the aging process, allowing aging-type effects to be reliably induced so they can be studied in greater detail.

For example, one area of aging research in which NCEs could provide an important information is the concept of ‘epigenetic clocks’ (see Glossary). Such clocks are observed in multiple species and the DNA methylation states or markers of the epigenetic clock are positively correlated with chronological age and lifespan in most cases (Horvath and Raj, 2018; Simpson and Chandra, 2021). In order to better understand the role of epigenetic clocks, it will be advantageous to link mortality, caloric consumption, somatic maintenance and other physiological parameters to the accumulation of epigenetic markers in the genome that occurs as individuals age (Omholt and Kirkwood, 2021). NCE systems offer a great opportunity to induce and study the consequences of environmental fluctuations, demography and the underlying epigenetic changes within a single unified framework.

Telomere shortening is a prominent mechanism of aging (Shay et al., 2019), and this is another research area to which NCE systems could contribute. Telomeres are repetitive non-coding regions that act as end caps in chromosomes (Blackburn, 2005). There are several stress-related factors that are found to be associated with telomere shortening in some species (Epel et al., 2004; Monaghan, 2014), and there are multiple hypotheses that address energy trade-offs as a cause for telomere shortening (Eisenberg, 2011; Casagrande and Hau, 2019; Young, 2018). These hypotheses are based on the idea that when organisms are faced with situations that demand high levels of energy expenditure (i.e. high-stress events), then the resulting elevation in physiological markers of stress such as glucocorticoids and reactive oxygen species leads to a trade-off with telomere maintenance to support immediate survival; this important assumption can be tested using the NCE framework. For example, using a non-lethal predation risk model, Noguera and Velando (2019) found that embryos of yellow-legged gulls had shorter telomeres after hatching as a result of prenatal predation risk (i.e. eggs exposed to predation risk). However, in pied flycatchers, it was observed that parents experienced shortening of telomeres when exposed to predation threat but this effect was not present in the offspring (Kärkkäinen et al., 2019). Thus, the PPI model can be used to study direct mechanisms of aging, and establish links between early or chronic stressors and the onset and rate of aging (Epel et al., 2004).

Perceived predation also induces plastic effects on the onset of aging in Daphnia sp. (Pietrzak et al., 2015). Such systems can be utilized to test how aging onset varies across populations that differ in their genotypic and phenotypic traits and are under varying selection pressure. In addition, literature exists on epigenetic clocks and molecular and physiological parameters of aging in Drosophila sp., but without an ecological perspective; thus, future research on these topics should aim to couple NCEs with the biology of aging (Piper and Partridge, 2018; Tsurumi and Li, 2020). Several other insect taxa (grasshoppers, butterflies, field crickets, flour beetles, honey bees and others), with a greater ecological diversity in the wild, could also be used to ask questions on how environmental factors (such as NCEs) influencing aging (Promislow et al., 2022). Thus, using an integrative framework with a combination of organismal and molecular biology tools and under the umbrella of the NCE framework will provide great advances in understanding and management of aging.

It is crucial to conduct studies that reflect the ecological context of PPI. Researchers should carefully design experiments that incorporate ecologically relevant predator cues, environmental conditions and prey behaviors to accurately capture the NCEs that shape adaptation, fear responses and aging. Investigating the effects of NCEs on aging requires long-term studies, so longitudinal studies over extended periods are optimal. Recognizing and accounting for individual variation is another critical aspect when investigating adaptation (e.g. coping styles) and fear responses as individual differences in behavior, genetics or physiological traits can influence the responses to NCEs and subsequent adaptive outcomes. Combining molecular techniques, such as gene expression analyses or neurophysiological measures, with behavioral observations and ecological outcomes can provide a comprehensive understanding of the molecular pathways involved in all processes discussed in this Commentary. Thus, NCEs provide a promising opportunity to usher in a new experimental paradigm to integrate important and connected biological processes and make future advances in many fields.

I thank Drs Nikhil Gaitonde, Amod Zambre and Veronica Rivi, and all the reviewers and editors for their valuable suggestions that improved the paper.