Box 4: Consequences of NCEs vs. CEs for prey and predator
population dynamics
A central difference between CEs and NCEs is their consequences for natural enemy reproduction (Abram et al. 2019). CEs generally lead to an increase in natural enemy birth rates: an immature parasitoid develops to the reproductive adult stage by attacking and killing a host, and a predator survives and reproduces by consuming prey. NCEs, on the other hand, do not result in any increase in the natural enemy’s population, and if they reduce the victim population through increased mortality or decreased fecundity, they actually shrink the resource pool available to the natural enemy. A natural enemy that induced NCEs only would eventually go extinct, as it would never be able to reproduce. It is worth noting that a generalist enemy may impose strictly NCEs on some of its prey taxa, as long as it is able to consume other species of prey or engage in omnivory. If NCEs are not explicitly accounted for, a gap between high pest mortality and low enemy reproduction may be erroneously attributed to other causes, such as poor assimilation efficiency or natural enemy mortality.
CEs and NCEs may also vary in how their overall magnitudes at the population level are influenced by predator density. A large decrease in the number of predators may lead to a large decrease in consumption of prey, but the small number of predators may still be enough to induce significant NCEs (Carpenter et al. 1987). The strength of NCEs can also be linked to CEs, creating potential feedbacks between the two effect pathways (Weissburg & Beauvais 2015). Understanding the perception of risk and thresholds prey use to make decisions can help determine how NCE strength may vary with enemy population compared to CE strength (see Box 3 for a more thorough discussion of prey perception and risk management).
The inclusion of enemy-risk effects in models has varying effects on the resulting dynamics, ranging from increased to decreased stability, the appearance of population cycles, and even the reversal of predicted trophic cascades (Abrams & Matsuda 1997; Abrams 2008; Peckarsky et al. 2008; Larsen 2012). The classic example of predator-prey dynamics involving lynx and snowshoe hares has discrepancies between observed data and CE-only predictions, but the inclusion of enemy-risk effects can help improve the match between prediction and observation (Hik 1995; Boonstra et al. 1998). The relative contributions of NCEs and CEs to population dynamics can vary with environmental factors and the spatiotemporal scales of study, so these interactions must be accounted for if possible. Considering enemy-risk effects in population dynamics is not simply the addition of ecological complexity for its own sake, but a way to improve predictions of population modeling.