Competition within the framework of food webs
In the 1970s, competition and predation were regarded as the primary interactions between species. This classification of interactions was illustrated by several articles in the 1975 book edited by Cody and Diamond titled ‘Ecology and Evolution of Communities’.
This book contained chapters by many of the most prominent ecologists of the time. Several of those chapters discussed the relative importance of competition and predation for the dynamics of communities. Notably, Connell (1975), who had become well known for his field experiments on two competing intertidal barnacle species, argued that predation had a greater effect on community composition than did competition. Of course, exploitative competition for living resources is a consequence of the predation of the consumers on those living resources, so one of the necessary consequences of predation in any community whose predator species have overlapping sets of prey is to produce interspecific competition. A specialist predator experiences intraspecific competition as a direct result of its predator-prey interaction. These cases show that a comparison of the ‘relative importance’ of predation and competition is fundamentally meaningless. However, it may be useful to compare the magnitudes of indirect effects transmitted between two species on a given trophic level via higher vs lower trophic levels.The most common classification of interactions in textbooks from the 1980s onward is that based on the three possible pairs of‘+’ and signs: (+,+) is mutualism; (-,-) is competition; and (+,-) is predation. Mutualism was often ignored in the 1970s but is now included in all textbooks. The two signs are usually used to denote the effect of an increase in abundance of species 1 on the abundance or per capita growth rate of species 2. However, this classification suffers from ambiguity because the time interval is not specified, and both time interval and the magnitude of the initial change in abundance can alter the sign of the effect on the receiver.
This is true of predation as well as competition. In addition, the change in short-term per capita growth rate may have a different sign than the change in equilibrium or long-term average abundance. In a system that does not come to a stable equilibrium, the sign characterizing the change in the equilibrium may differ from the sign that describes the change in the mean abundance. Adding more predator individuals does initially decrease prey abundance, barring the possibility of increased maladaptive fighting among predators. However, it is quite possible that, by the time the system reaches its new equilibrium, increased productivity of the prey's food more than offsets the predator's effect, and the prey actually increases (Abrams 1992b; Peacor 2002).While it is possible to argue for a sign-based classification of interactions, if this is adopted, the nomenclature needs to be changed. Predation is usually defined as a mechanism of interaction, while mutualism is usually defined as a paired set of positive effects, each of which maybe produced by many mechanisms. The inconsistency of using mechanism and effects for different interactions was discussed in Abrams (1987a). Another problem with the classification framework based on a fixed pair of signs is that it favours thinking about interactions purely in terms of species pairs. The interaction between almost any randomly selected pair of species in a natural community is likely to depend on the abundances of many additional species, and the sign of one or both effects may be changed by these additional species. Thus far, I have largely ignored the impacts of trophic levels higher than that of the competing consumers, but it has long been recognized that the presence of a specialist, foodlimited predator on each of the consumers can eliminate the effect of competition between them on each other's equilibrium abundances. In spite of eliminating the impact of a neutral parameter change in one consumer on the population size of the other, the specialist predators in this case do not eliminate the evolutionary effects of competition between the consumers (Abrams and Cortez 2015b; see Chapter 11).
Trophic level(s) lower than the resources (if they exist) are also likely to affect the interaction between consumers. This was implicit in Levine's (1976) model, although he represented the competition between resources using direct effects.An even wider variety of more complicated food web connections between resources is possible in systems with two or more trophic levels below the prey. In predator-herbivore systems with potentially competing predators, the plants can have effects other than simply defining the food available to the herbivores. For example, plant species can alter the structure of the habitat in a manner that makes one or more herbivore species more vulnerable to their predators (for example, making the predators more difficult to detect; e.g., Holt and Barfield 2013; Pearse et al. 2020). This can alter the competition between different herbivore species. Effects of other trophic levels can arise when there are species at higher or lower trophic levels, whose abundance changes the behaviour of one or both species of a focal consumer-resource pair in ways that alter their interaction (Abrams 1984b, 1995; Preisser et al. 2005; Bolnick and Preisser 2005). Theory suggests that such behavioural effects can be transmitted through multiple trophic levels (Abrams 1992a). Examples of adaptive behavioural effects will be provided in the following chapter. In general, these effects broaden the range of potential outcomes for the interaction of any particular pair of species.
The currently proposed definition of competition based only on shared use of limiting resources means that a pair of species can be competitors while having mutually positive effects on each other's abundance (measured using neutral parameter perturbations of small magnitude). Matsuda et al. (1993) discussed a model of two predators sharing a single prey, in which the prey species exhibits adaptive defence with the property that defending against one predator species makes the prey more vulnerable to the other predator.
This always entails short-term positive effects of increased abundance of one predator on the per capita growth rate of the second. However, the long-term effect on the abundance of the second predator may be negative or positive, depending on the magnitude of the change in prey population and the effectiveness and exclusivity of the two defences (Matsuda et al. 1993). Even with mutually positive effects on ultimate population sizes, the definition of competition proposed above would include this as competition because it involves a shared resource. The fact that the outcome of‘mutualisms' may depend on other species or environmental conditions has long been acknowledged in the literature dealing with that interaction (Bronstein 1994). However, this conditionality is seldom mentioned in discussions of competition.The issue of sign-based vs mechanism-based classifications of interactions brings up the more general issue of how to quantify the ‘effect of one species on another'. This seemingly innocuous phrase actually has many potential interpretations, as should be clear from the above. What is changed in the effector species and what is measured in the affected species? Over what time frame is that quantity measured? Bender et al. (1984) raised these and other ambiguities in measuring interspecific effects, and they were explored by Yodzis (1988, 1989). This topic will be examined further in the following chapter. That chapter will also provide a more detailed introduction to the nature of consumer-resource interactions.