Animal signals often evolve to exploit pre-existing sensory biases in receivers (reviewed in Ryan 1990; Endler & Basolo 1998). For instance, novel signals may resemble stimuli to which a receiver is already attracted, such as food-mimicking orange spots on male guppies and appendages on male tetras (Rodd et al. 2002; Arnqvist & Rowe 2005). Sensory bias can also select for signals that exploit receivers’ aversion to certain stimuli, such as harmless Batesian mimics resembling a dangerous model (Bates 1862). Sickness behaviours (Hart 1988) and physical symptoms of infection are often used to identify infected conspecifics (Zylberberg et al. 2013). Symptoms and sickness behaviours have also recently been framed in terms of signal evolution, with the possibility of dishonesty in these signals being conjectured (Shakhar & Shakhar 2015; Steinkopf 2015; Tiokhin 2016). However, to our knowledge, risk-factors for parasitic infection as a template for sensory exploitation in the context of competition and sexual selection have not been demonstrated in any species. We propose that signs of infection could be promising and flexible sources of dishonest signal evolution.

First, infection-mimicking signals have the benefit of their being weighed and interpreted differently by different receivers. For instance, often some members of a population are more susceptible to infection than others or may incur greater costs upon infection (Zuk 2009; Hawley et al. 2011). We would predict these highly susceptible hosts would avoid infected conspecifics more so than less susceptible individuals. This allows dishonest signals of infection to potentially target their effects towards specific subsets of a population, or at least differ in the strength of their effects on different receivers. Facultatively expressed dishonest signals of infection are more powerful still because, even in the absence of differential susceptibilities of receivers, they could be expressed or even directed towards (e.g., sneezes) certain individuals whilst being concealed from others. Additionally, in extreme cases, infection mimicry has the potential to attract some receivers but repel others. These possibilities are discussed in detail later.

Second, dishonest signals of infection have the benefit that they are unlikely to interfere with species recognition systems (e.g., identification by potential mates). Contrast this to a hypothetical scenario where individuals evolve to resemble their predators. Such predator-mimics run the risk of being misidentified as heterospecifics, which could hamper their ability to engage in beneficial social interactions, e.g. mating.

In the following sections, we lay out more specific scenarios in which dishonest signals of infection might evolve and be maintained, or in which aversion to infection cues could be an initially exploited sensory bias resulting in subsequent signal elaboration.