Species exist in complex webs of interactions (Joern and Laws, 2013) and foraging choices by generalist consumers are often found to have a range of indirect effects on the patterns of coexistence of their multiple food resources (Ohgushi, 2005). Indirect effects occur when the effect of one species influences the strength of interaction between other species within interaction webs (Werner and Peacor, 2003) leading to alternative dynamical outcomes (Golubski and Abrams, 2011; Terry et al., 2017). Many ecological processes are affected by within-species behavioral syndromes (Sih et al., 2012) and it is known that these sets of behavioral syndromes could operate in trophic interactions such as indirect interactions. As such, these behaviors (e.g., changes in survival strategies, foraging decisions and/or the occurrence of associational defenses), mediated through functional responses, are mechanisms of trophic interaction modifications (Bolker et al., 2003; Terry et al., 2017).

Modifications of trophic interactions involve behavioral responses of consumers to the availability of alternative resources—the outcome of these interactions are often identified as associational effects (Barbosa et al., 2009; Underwood et al., 2014). The concept of associational effects is used to describe indirect interactions (similar to apparent competition) where the risk of herbivory to a focal plant species depends on the composition of the neighboring plant community (Sato, 2018). Associational effects can be divided into associational resistance and associational susceptibility. Associational resistance refers to the situation where the focal plant receives less damage, whereas associational susceptibility occurs when the focal plant receives more damage, depending on the plant species in the neighborhood (Sato, 2018). Based on a review of the literature, Sato (2018) concludes that the outcomes of these associational effects depends on herbivore behavior and plant trait changes. Furthermore, studies have investigated how neighborhood effects through resource traits, ie. density and frequency of focal resource and neighboring resources, affect consumer load (Hamback et al., 2014; Underwood et al., 2014; Kim, 2017). More recently the effect of resource neighborhood on consumer traits has come under scrutiny, the focus has shifted to consumer foraging behavior as a key-driver of associational effects (Hahn and Orrock, 2016; Verschut et al., 2016). The spatial scale of resource selection (Hamback et al., 2014; Verschut et al., 2018), responses to resource cues (Verschut et al., 2017b) and the physiological state of the consumer (Verschut et al., 2017a) all contribute to the likelihood of the occurrence of associational effects. Hence, mechanisms of associational effects are thought to act through neighborhood effects on resource traits and consumer traits as well as predator load, microclimate and relevant feedbacks (Kim, 2017).

In the context of long-term dynamics, the occurrence of these mechanisms will give rise to indirect interactions between the focal and the alternative resource. The outcome of the indirect interaction is then dependent on the conditions for the establishment of associational effects together with the state and behavior of the consumer and focal resource (Figure 1; Underwood et al., 2014). An anticipated long-term outcome could be either positive or negative apparent competitive interactions; the presence of one resource species could affect the consumer traits altering the interaction of the consumer with the other (alternative) resource (Werner and Peacor, 2003).

Underpinning the outcomes of positive and negative apparent interactions are foraging and resource switching decisions by the consumer. These behavioral decisions are also critically important for the occurrence of associational effects (Hahn and Orrock, 2016; Verschut et al., 2017a). Switching resource preferences, where the proportion of resource consumed varies from less to greater than expected, can influence the dynamics of competing (or non-competing) resources (e.g., Oaten and Murdoch, 1975; Hassell, 2000). Although switching might not be an unequivocal mechanism promoting the stability of resource-consumer systems (May, 1977), Holt (1977) suggested that strong switching behavior by generalists may relax apparent competition between resources. van Baalen et al. (2001) explore the conditions for stability and population persistence and conclude that the presence of an alternative resource is not necessarily a determinant for population stability but will, under broader conditions, enable population persistence. Resource switching that relaxes apparent competition will thus influence the associational effect of alternative resource on the focal resource. Within this paradigm, the aim of our work is to link the concept of associational effects to the theory of resource switching by generalist consumers and investigate how this modifies the indirect interactions between the resources (Figure 1).

The scale of generalist foraging (Hassell and Southwood, 1978) related to resource density (Krivan and Eisner, 2003) and quality (Eubanks and Denno, 2000), and their variability (van Baalen et al., 2001), is a key condition for the appearance of associational effects (Figure 1; Hamback et al., 2014; Underwood et al., 2014). For instance, when the consumer has no strict preference for a particular resource the attack rate is mostly determined by the abundance of the alternative resource (Chesson, 1989) and the generalist consumer growth rates are equally influenced by each resource (Holt, 1977). In systems where a consumer interacts with multiple potential resources, resource preference might be determined by other factors than resource abundance. The effect of encounter rate on switching and subsequent coexistence has been investigated under optimal foraging and/or functional response of the generalist (van Baalen et al., 2001; Krivan and Eisner, 2003, 2006; Abrams and Matsuda, 2004). Deviations in handling time and resource quality between focal and alternative resource will lead to persistence of the focal resource given that the quality of the alternative resource is not too low (ie. not eaten) or too high (i.e., deregulate the consumer dynamics; van Baalen et al., 2001). Under optimal foraging decisions by generalist consumers, Krivan and Eisner (2006) showed how top-down population regulation or the combination of both bottom-up and top-down population regulation affects patterns of resource coexistence. If resources are top-down regulated and consumers have inflexible foraging strategies this can result in competitive exclusion of one of the resources. However, if the consumers forage so that they optimize their per capita growth rate, coexistence of both resources is promoted, suggesting that underlying resource dynamics can be important for the outcome of consumer-mediated indirect interactions between resources. Furthermore, the scale of the interaction between alternative resources and foraging generalists has complex consequences for resource-consumer interactions. Specifically, generalist foraging can be modified by neighborhoods and the density of alternative resources (Hahn and Orrock, 2016). The presence of an alternative resource thus has profound implications for the outcomes of coexistence and indirect interactions, and can therefore effect the long-term outcome of dynamics related to associational effects (Kim et al., 2013). When a generalist consumer feeds on multiple resources, apparent competition between resources might arise (Holt, 1977). Empirically, apparent competition is a potent force in the structuring of multi-species assemblages (Holt, 1977; Holt and Lawton, 1993; Chaneton and Bonsall, 2000; Holt and Bonsall, 2017) and has been shown to occur in many systems including host-parasitoid interactions (e.g., Bonsall and Hassell, 1997), marine intertidal ecosystems (e.g., Menge, 1995), host-parasite interactions (Zhang and Buckling, 2016) and plant-herbivore systems (e.g., Hamback and Ekerholm, 1997).

In our study, we approach apparent competition and associational effects as two sides of the same phenomenon. We analyse consumer switching behavior and density dependence, as these are known to affect the outcome of apparent interactions between two resources and a shared consumer, using drivers of associational effects as the mechanism giving rise to apparent interactions. We explore the following drivers of associational effects: the relative density of focal and alternative resource, effects of generalist density on the quality of focal and alternative resource and the effect of the alternative resource density on the quality of the focal resource. We start with the introduction of the model, and proceed to analyse the stability conditions and how these relate to the relative densities of the focal and alternative resource. Then we use the model to explore the effects of consumer switching and consumer density dependence as underlying mechanisms of apparent competition. In the discussion, we combine the mechanisms and drivers arguing that short-term consumer choices can have long-term on outcomes of trophic interactions.

To explore the indirect effects of associations between resources, we adapted a basic consumer—resource dynamics model (Rosenzweig and Macarthur, 1963),

This model framework is used to investigate the effect of resource switching on the indirect interaction between a focal and alternative resource. The generalist (consumer) dynamics are described by:

where dGdt describes the change in density of the generalist (G) with respect to time (t), R(t) is the abundance of the focal resource at time t, U is the abundance of alternative resource at time t and m is a parameter describing the shape of the switching effect within the functional response. α is the resource consumption rate and μ_G the density-independent mortality rate (see below). The focal resource dynamics are described by:

where λ_R is the constant (allochthonous) rate of resource supply, μ_R is the density-independent resource loss rate &the combination of α and the resource preference response R(t)mR(t)m+Um can be seen as scaled foraging rate (α(R(t)mR(t)m+Um)) of the generalist on the focal resource (functional response). The formulation of the switching response has been applied previously by Hutson (1984) and Teramoto et al. (1979) in studies of predator-prey interactions.

We start with the analysis of the density-independent consumer—resource model (Equation 1) to understand the behavior of the model when the focal and alternative resource are equally preferred. The density of the focal and alternative resource will determine the outcome of the interaction. It has been proposed that this interaction will lead to apparent competition (Holt and Bonsall, 2017), analyzing the model will guide understanding whether this is the case for the dynamics described by the our model. In addition, the same approach will be used to analyze the model under different forms of density-dependence and provide a basis for the initial conditions for further numerical analyses.

Here, we have investigated the behavioral effects associated with resource preference (associational effects) and our results show that these effects can lead to either apparent competition or mutualism between resources. The direction and strength of the apparent interaction between resources depends on switching behavior and life history of the shared consumer. We show that short time-scale choices by the consumer can have long-term implications for resource coexistence patterns. Our results indicate that consumer traits, like switching behavior, density-dependent mortality and intraspecific competition, can have a strong effect on the outcome of long-term consumer mediated interactions between resources, leading to facilitation of coexistence. Negative feedbacks on conversion rates (α) change equilibrium densities but not the apparent interactions (Figure 5) and could therefore be of less importance for resource dynamics.

Our approach to modeling the effect of resource switching on the stability of consumer-resource interactions differs from that in previous studies in the way that our model aims to capture different life histories of the consumer. Previous approaches have focused on the effects of consumer density, resource quality, handling time and/or consumer foraging behavior (van Baalen et al., 2001; Krivan and Eisner, 2003, 2006). As associational effects are often hypothesized to result from focal resource neighborhood (Underwood et al., 2014), our model describes the interactions between two resources and a shared consumer. In our model the switching parameter is determined by the relative density of the focal resource and the alternative resource instead of either a threshold consumer density (van Baalen et al., 2001) or a more profitable resource (Krivan and Eisner, 2003). Using this approach, the mathematical analysis of the model shows the resource-consumer dynamics exhibit donor-controlled stability under equal preference and density independent conditions, which is analogous to findings of van Baalen et al. (2001). Moreover, we find apparent and mutualistic effects on the focal resource within the stability range of the consumer-resource interaction, which provides a different perspective to the effects of consumer switching and allows us to connect these interaction to the long-term impacts of associational effects.

By including density-dependence in the dynamics of the generalist the model reflects, amongst other things, either intraspecific competition leading to density-dependent mortality (Royama, 1992, p. 40–41). Further, when the generalist responds to herbivore density in a (spatially) density-dependent manner, i.e., an aggregative response (Turchin, 2003, p. 78), any increases in the mortality of the consumer can result in an increase in the equilibrium level of coexisting plants. The strength of the density-dependence, form of intraspecific competition and/or predation pressure, will determine the coexistence equilibrium for the focal resource (Figures 3B,C; Long et al., 2012). Predators are known to respond also to resource neighborhood (Guyot et al., 2015). Hence, making the number of offspring negatively related to the density of the alternative resource could reflect either the dilution effect for consumers (Civitello et al., 2015) or increased availability of resources for the predator as predicted by the natural enemy hypothesis (Stiling et al., 2003). Both of these mechanisms have been observed empirically to lead to associational resistance (Castagneyrol et al., 2013; Guyot et al., 2015) and will, as our model confirms, lead to long-term apparent mutualistic effects (Figure 5).

We investigated some of the mechanisms proposed to underlie associational effects and some of the expected feedbacks (Figure 1; Underwood et al., 2014; Kim, 2017). Our results suggest that the outcome of generalist consumer—multiple resource interactions are important to understand from the perspective of associational effects (Underwood et al., 2014) and on a wider scale, trophic interaction modifications (Terry et al., 2017). From an applied perspective, increased herbivore pest mortality as a result of apparent competition does not necessarily reduce inflicted damage on individual plants in agricultural crop systems (Jaworski et al., 2015) as positive ecological effects might not always mean positive economical effects. Using generalist natural enemies for biological control might have negative effects on non-target organisms (Elkinton and Boettner, 2012), potentially without reducing the population density of the target organism. Our results become especially relevant when the theory of associational resistance is applied to managed ecosystems, such as agricultural or forestry systems. We highlight that increasing diversity, either at the resource or consumer level, could benefit the target organism if mechanisms acting at the population level of generalist consumers allow for mutualistic interactions rather than apparent competitive outcomes. However, enriching ecosystems is well known to lead to paradoxical outcomes (e.g., Gilpin and Rosenzweig, 1972) and extending this to the breadth of indirect ecological interactions requires more attention (Bonsall and Holt, 2003). Hence, caution and further study is needed in developing these ecological concepts for production systems.

Further empirical work is needed to tease apart the effects of frequency and density on the dynamics resulting from associational effects (Kim and Underwood, 2015). Within season experiments on consumer choice might not be informative about the outcome of long-term interactions as well as the outcome of interactions at larger spatial scales (van Nouhuys and Kraft, 2012; Hamback et al., 2014). At the food web level, sharing consumers will support the occurrence of indirect (positive or negative) effects when the resources exhibit different, individual population dynamics (Tack et al., 2011). By including density-dependence in the generalist consumer dynamics, we show that the effect of alternative resource(s) on the likelihood of mutualistic interactions increases, this is further enhanced if the focal resource is a nutritionally better resource. As all natural populations are subjected to some form of density-dependence, any kind of indirect interaction should be considered in this context.

In addition, empirical research has shown that behavior as well as spatial scale are important determinants of associational resistance or associational susceptibility (Hahn and Orrock, 2016; Verschut et al., 2016). Sensory responses to plant traits will determine occurrence of associational effects at different spatial scales of foraging (Verschut et al., 2017a,b). Our work has investigated the effect of decisions made by the consumer at patch level, and it has been shown that patch level diversity will determine the consumer load on the focal resource, lower numbers of consumers at higher levels of diversity (Hamback et al., 2014). Further theoretical work is necessary to increase the understanding of the importance of spatial scale in the context of the population dynamical outcomes of associational effects at the patch scale.

In conclusion, our work connects the theoretical body of work on indirect interactions to the concepts of associational effects. Adapting the associational effects framework (Figure 1 and Box 1) together with the results presented here suggests that the indirect interactions between multiple resources need more thorough investigation to appreciate the range of associational effects that could result from the dynamical interaction between a generalist consumers and its focal and alternative resources.

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found below: https://osf.io/w7atk/

MB: conceptualization, formal analysis, validation, writing–original draft, and writing–review and editing. MK: conceptualization, numerical analysis, writing–original draft, and writing–review and editing. Both authors contributed to the article and approved the submitted version.

MK was funded by the Mistra-financed project Future Forests and by the FORMAS research grant reg no. 2015-55.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.