The relative importance of prey-borne and predator-borne chemical cues for inducible antipredator responses in tadpoles

Abstract

Chemical cues that evoke anti-predator developmental changes have received considerable attention, but it is not known to what extent prey use information from the smell of predators and from cues released through digestion. We conducted an experiment to determine the importance of various types of cues for the adjustment of anti-predator defences. We exposed tadpoles (common frog, Rana temporaria) to water originating from predators (caged dragonfly larvae, Aeshna cyanea) that were fed different types and quantities of prey outside of tadpole-rearing containers. Variation among treatments in the magnitude of morphological and behavioural responses was highly consistent. Our results demonstrate that tadpoles can assess the threat posed by predators through digestion-released, prey-borne cues and continually released predator-borne cues. These cues may play an important role in the fine-tuning of anti-predator responses and significantly affect the outcome of interactions between predators and prey in aquatic ecosystems. There has been much confusion regards terminology used in the literature, and therefore we also propose a more precise and consistent binomial nomenclature based on the timing of chemical cue release (stress-, attack-, capture-, digestion- or continually released cues) and the origin of cues (prey-borne or predator-borne cues). We hope that this new nomenclature will improve comparisons among studies on this topic.

Box 1: Clarifying terminology and classifying mechanisms for chemosensory-mediated predator detection

Many studies report the use of chemical cues to detect predators, but they employ widely different definitions and classifications of types of cues. The same terms are used sometimes as synonyms, at other times they refer to different phenomena, and definitions are often missing. For example, ‘diet-released cues’ can refer to those that originate from digested prey (e.g., Ferrari et al. 2007; Ferland-Raymond et al. 2010), but sometimes they also include cues that are released by prey upon attack (e.g., Laurila et al. 1997; El-Balaa and Blouin-Demers 2013). Many authors use the term kairomone in reference to cues from a predator that are independent of its recent feeding history (e.g., Brönmark and Hansson 2000; Hettyey et al. 2010), others state that kairomones include digestion-released cues (e.g., Kats and Dill 1998; Schoeppner and Relyea 2005, 2009), and still others use the term kairomone whenever the receiver is a heterospecific (e.g., Chivers and Smith 1998).

A second example of inconsistent terminology is the classification of cues as indirect or direct. Indirect cues originate from prey and have evolved to alert other prey to predation threat. They include several kinds of chemicals: general prey metabolites that are excreted actively upon stress (‘no-cost disturbance signals’; Wisenden et al. 1995; Kiesecker et al. 1999), special disturbance cues that are costly to produce and are released by prey actively upon attack (‘alarm pheromones’; Fraker et al. 2009), cues that are passively released from injured prey tissue (‘damage-released cues’; Chivers and Smith 1998), and cues that are released from prey by digestion (‘digestion-released cues’, also referred to as ‘predator-labelling’; Mathis and Smith 1993; Chivers and Smith 1998; Ferrari et al. 2007). Direct cues, on the other hand, originate directly from the predator and represent the smell of the predator itself that is independent from its recent feeding history. These cues are released ‘unintentionally’, alerting potential prey to predation threat and lowering the predator’s chance of successful attack. Direct cues include chemicals and tissue fragments that are released more or less continually from the integument of the predator (‘kairomones’; Petranka and Hayes 1998; Brönmark and Hansson 2000), saliva released during capture and consumption of prey (we know of no study demonstrating this), and digestive body fluids of the predator, tissue fragments of the predators’ digestive tract and samples of the predators’ gut flora released during excretion (‘digestion-released cues’; Mathis and Smith 1993; Ferrari et al. 2007). As can be seen from the above list, excrements of predators may contain both indirect and direct cues. Furthermore, kairomones may be released not only continually from the integument of predators, but also during defecation (fractions of ‘digestion-released cues’). This further confuses functional and physiological/mechanistic classification. Finally, some of the current nomenclature is based on functionality and some on the timing of release, while cue origin is only implicitly understood.

To improve clarity, help avoid misunderstandings and facilitate comparability of results, we propose a new terminology for the cues involved in chemosensory-mediated predator detection. We suggest using a binomial nomenclature and classification based on the timing of cue release (stress-, attack-, capture-, digestion- or continually-released cues) in combination with cue origin (prey-borne versus predator-borne cues) (Table 1).