Enemy-Risk Effects in Time
Temporal scaling of enemy-risk effects is complex, since pests can respond to enemies on multiple scales, and consequences of those responses can appear at multiple scales as well. Short-term behavioral changes by pests can lead to two main categories of outcomes: there may be a long-term fitness consequence of short-term changes, or there may be compensation for the short-term effect in the long-term. Other pest responses occur only over a longer time scale, such as changes in life history events. The goals of a biocontrol program affect the importance of different enemy-risk effects across time.
Short-term behavioral responses may lead to long-term fitness consequences. The accumulation of small fitness losses, such as reduced feeding, mating opportunities, or increased energy expenditure, can lead to long-term reductions in population growth. Short-term reductions in feeding rate during a vulnerable life stage may also delay development, which may lead to increased pest mortality due to high CEs (Uesugi 2015). Furthermore, if the focus of a study is solely on short-term effects, these long-term changes may not be measured. Similarly, if long-term population growth is studied without looking at short-term mechanisms, NCEs might be missed entirely, and the change in growth rate may be attributed solely to CEs (see Hermann & Landis 2017 for a more in depth discussion of appropriate time-scales).
Short-term enemy-induced trait responses may also be compensated for in the long-term, generating no NCEs and little impact on the pest population as a whole. If enemy risk is variable, pests that suffer losses in feeding or mating during high-risk periods may be able to compensate during periods of low risk (Houston et al. 1993). Compensatory mortality can also occur in biological control systems, as when density-dependent mortality is replaced by enemy-induced mortality, leading to no overall difference in mortality (Cloutier & Bauduin 1995; Suh et al. 2000). While this has been demonstrated in CEs, the same could occur for NCEs, where strong effects during one life stage lead to no difference in later population size.
Short-term behavioral shifts alone may have a significant impact on biocontrol outcomes if they can be aligned with periods of crop vulnerability. Pests are often only damaging during a particular crop or pest growth stage (Hokkanen 1991; Wiedenmann & Smith 1997). The use of temporal asynchrony between crop and pest stages, achieved through precise timing of crop production, can exploit the narrowness of the crop vulnerability window to reduce pest impact (Letourneau & Bruggen 2006). Similarly, if pest pressure can be reduced during that time through enemy-induced behavioral responses, crop damage may be decreased regardless of impacts on pest population growth.
Some trait responses to natural enemies only occur in the long-term, and as such, their consequences only appear in the long-term as well. Pests can shift their life history in response to enemy risk, including increasing developmental rate (Thaler et al. 2012; Elliott et al. 2016; Rendon et al. 2016). Speeding up development of a vulnerable life stage may reduce overall exposure to natural enemies, but incur costs later on. If shorter development means less time in a crop-damaging life stage (e.g. less time spent as a crop-feeding caterpillar), this may be beneficial for the crop, though it may also increase the rate of pest population growth. Different pests, even within the same order, may allocate risk avoidance behavior to different life stages, either exhibiting oviposition site selection or juvenile enemy-avoidance behavior (Stav et al. 2000, 2010; Kiflawi et al. 2003; Brown & Kotler 2004; Blaustein et al. 2005).
It is important to consider the goals of the biocontrol program when addressing temporal components of enemy-risk effects. In a classical biocontrol program, where the goal is the long-term establishment of the natural enemy, some level of CEs is necessary to sustain the enemy population, even if NCEs are initially very high. However, with an augmentative release, high enemy densities are expected to remain for only a short time. In this case, strong short-term behavioral changes, such as temporarily reduced feeding, or short-term behaviors that lead to long-term fitness consequences may be enough to significantly impact the pest, though the enemy does not establish. For example, if an augmentative release of enemies leads to a large reduction in pest feeding during a week of high crop vulnerability, then long-term impacts on pest population may be of little concern since the damaging behavior itself was prevented.