Informed herbivore movement and interplant communication determine the effects of induced resistance in an individual‐based model

Abstract

Summary1. Plant induced resistance to herbivory affects the spatial distribution of herbivores, as well as their performance. In recent years, theories regarding the benefit to plants of induced resistance have shifted from ideas of optimal resource allocation towards a more eclectic set of theories that consider spatial and temporal plant variability and the spatial distribution of herbivores among plants. However, consensus is lacking on whether induced resistance causes increased herbivore aggregation or increased evenness, as both trends have been experimentally documented. 2. We created a spatial individual-based model that can describe many plant-herbivore systems with induced resistance, in order to analyse how different aspects of induced resistance might affect herbivore distribution, and the total damage to a plant population, during a growing season. 3. We analyse the specific effects on herbivore aggregation of informed herbivore movement (preferential movement to less-damaged plants) and of information transfer between plants about herbivore attacks, in order to identify mechanisms driving both aggregation and evenness. We also investigate how the resulting herbivore distributions affect the total damage to plants and aggregation of damage. 4. Even, random and aggregated herbivore distributions can all occur in our model with induced resistance. Highest levels of aggregation occurred in the models with informed herbivore movement, and the most even distributions occurred when the average number of herbivores per plant was low. With constitutive resistance, only random distributions occur. Damage to plants was spatially correlated, unless plants recover very quickly from damage; herbivore spatial autocorrelation was always weak. 5. Our model and results provide a simple explanation for the apparent conflict between experimental results, indicating that both increased aggregation and increased evenness of herbivores can result from induced resistance. We demonstrate that information transfer from plants to herbivores, and from plants to neighbouring plants, can both be major factors in determining non-random herbivore distributions.

The volatile emission of a specialist herbivore alters patterns of plant defence, growth and flower production in a field population of goldenrod

“…Previous modelling suggested that induced resistance in S. altissima could influence spatial distributions of the goldenrod specialist T. virgata (Rubin et al . ). The model predicted how movement from induced to uninduced plants altered herbivore distributions, and the ability of plants to anticipate attack resulted in more even distributions, suggesting that a gradation of primed defences over distances of a meter are likely to have profound effects on the distribution of herbivores, with possible ramifications for the entire goldenrod community (Rubin et al .…”

Plant communication in a widespread goldenrod: keeping herbivores on the move

“…This was predicted for an herbivore like T. virgata because mobile, selectively feeding herbivores (Rubin et al . ) move more frequently within and between herbivore‐induced plants (Fernandes et al . ) and benefit by maximizing their time spent feeding on uninduced tissues (Paschold, Halitschke & Baldwin ; Karban ).…”

“…This conclusion is supported by a recent model (Rubin et al . ) that found that the transfer of chemical information between neighbouring plants is more important for a more even distribution of beetle herbivory within the plant population (risk spreading) than the transfer of information to the herbivore, with a combined effect (induced resistance plus plant communication) being the strongest. This, in combination with our data, may explain the earlier observations that herbivore loads on individual plants are relatively low more than 75% of the time (except during outbreaks) and that at these low loads, herbivory has no measurable effect on S. altissima 's fitness (Root ).…”

Predator–prey interactions in a ladybeetle–aphid system depend on spatial scale

“…By implementing informed emigration and random immigration, species in our model performed informed, but not optimal habitat selection. In theory, less‐than‐optimal dispersal can decrease the effectiveness of predators in suppressing prey at the landscape scale (Holt, 1985); while informed dispersal would result in higher aggregation of predators (Rubin, Ellner, Kessler, & Morrell, 2015). …”