Box 5: Enemy-risk effects and biological control of vectors of
plant disease
One of the most damaging ways that insect herbivores affect their host plants is by acting as vectors of plant pathogens. Biological control agents can clearly slow the spread of vectored pathogens by suppressing vector population densities; as both consumptive and non-consumptive effects can depress population growth rates of insect vector populations, both can contribute to this ecosystem service (Landis & Van der Werf 1997; Moore et al. 2009; Finke 2012; Long & Finke 2015; Clark et al. 2019).
However, it is now widely recognized that enemy-risk effects may also have a somewhat counterintuitive and unhelpful influence on the epidemiology of insect-vectored pathogens: in some cases, anti-enemy behaviors may involve increased movement of insect vectors on both local and regional scales, accelerating disease transmission. Thus, the net effect of biological control on disease prevalence can be negative, neutral, or positive, depending on the relative magnitudes of consumptive effects and enemy-risk effects and the details of the interactions (Finke 2012; Crowder et al. 2019). The empirical record has shown that outcomes can depend on the identity of the biocontrol agents, the herbivore, and the pathogen (Nelson & Rosenheim 2006; Belliure et al. 2011; Dumont et al. 2015; Clark et al. 2019); in particular, predator-prey interactions that result in strong prey dispersal in response to predation risk or actual predator attacks often result in short-term increases in disease transmission any time pathogen acquisition and transmission by the vector is not interrupted by the decision to leave a feeding site.
The empirical literature shows that a widespread response of insect vectors of plant disease to predator presence and, especially actual predator attacks, is to move away from the attack site via local movements (Weber et al. 2006; Belliure et al. 2011; Hodge et al. 2011; Dáder et al. 2012; Long & Finke 2015). Aphids, which vector more than half of all plant viruses, release alarm pheromones when attacked by predators, causing clone-mates to run away or, in some cases, to drop from the host plant (Vandermoten et al. 2012). Especially in cases where disease transmission requires rapid movement between two host plants (common for viruses that are transmitted via transient contamination of aphid mouthparts), this can accelerate disease transmission.
Predators can also shape longer-distance movements via two potentially offsetting processes. First, many herbivores show density-dependent induction of winged morphs or other forms of density-dependent dispersal (Denno & Peterson 1995; Pepi et al. 2016); in this case, suppression of vector population densities via consumptive or non-consumptive effects has the potential to slow disease spread (Michaud & Belliure 2001). Second, however, many herbivores also induce winged forms in response to detection of predator cues, including, for aphids, alarm pheromones (Weisser et al. 1999; Mondor et al. 2005; Vandermoten et al. 2012), potentially leading to substantial increases in potential for disease transmission over larger spatial scales. Although experimental studies have demonstrated the potential for both of these effects, how this plays out in nature is unknown.
The preponderance of evidence from experimental studies supports the hypothesis that natural enemies accelerate disease transmission in crop plant populations (Long and Finke 2015). However, because most published studies are quite short-duration, they can reveal the immediate effects of increased vector movement, but may underestimate the importance of vector population suppression, which often requires multiple generations of predator-herbivore interactions. Also, because most studies have been performed in lab or greenhouse settings, the importance of predators as elicitors of vector movement may be exaggerated relative to its true effect in the field, where many other factors can trigger the same trivial movements (e.g., effects of wind, mechanical disturbances, and contacts with other herbivores; Bailey et al. 1995; Nelson & Rosenheim 2006). Nevertheless, it is clear that biological control can be a double-edged sword when directed against disease vectors.