Theory’s roles in ecology and competition

12.2.1 The goals of theory

The ultimate goals of theory in ecology should be to understand and predict real- world changes in the abundances of species, or of larger groupings, such as trophic levels.

Ecologyshould also provide the groundwork for understanding and predicting evolutionary change in traits that influence population dynamics. How well practitioners have succeeded in these goals is a matter of debate (Getz et al. 2018). The record may give the appearance of more success than is deserved, as studies have focused on more easily answered questions in a small range of simple systems.

Half a century ago, the indirect interaction of competition seemed to be an ideal case within community ecology for understanding and predicting changes in species abundances. It was, and still is, arguably the simplest, and certainly the most studied, indirect interaction in natural communities, and it is also central to understanding a range of important topics dealing with processes within a species, ranging from evolution to density dependence. MacArthur’s well-known 1970 paper and 1972 book clearly reflected his feeling that competition theory should at least be consistent with dynamic processes in both consumers and resources. However, the initial expansion of resource-based competition models following MacArthur’s treatments stalled at the end of the 1980s, when competition itself became less popular as an object of study.

The theme of exploring consumer-resource mechanisms seems to be poorly represented (although certainly not absent) in the recent revival of theoretical work on competition (see Chapter 4). Attempts to predict the dynamics of sets of real-world competitors using consumer-resource models seems to have been largely restricted to pairs of very small organisms (e.g., phytoplankton) in the laboratory, a line of research that dates back to Gause (1936).

Laboratory research involving phytoplankton as either consumers or resources has also been the subject of some of the more successful attempts to integrate model predictions with experimental results (e.g. Becks et al. 2010; Ellner and Becks 2011). These systems have also been the subjects of some of the relatively few analyses of impacts of non-neutral trait values on species abundances (e.g. Litchman et al. 2010; Klausmeier et al. 2020a), as well as studies of seasonal systems (e.g. Descamps-Julien and Gonzalez 2005).

All models are simplifications, and even overly simplified models often have some value. Before 1970, the 2-species LV model was the main theoretical guide to understanding the interspecific interaction of competition. It provided two important insights that were not widely appreciated before: (1) the importance of the relative amounts of inter- and intraspecific competition in determining the strength of the interaction; and (2) the possibility of alternative exclusion outcomes depending on initial abundance. The multi-species extension showed the possibility of cyclical dynamics (May and Leonard 1975) and of multiple alternative equilibria involving different sets of species (Gilpin and Case 1976). These findings represent the main contributions of the LV model to understanding competition. However, the LV model is too far removed from the mechanisms of competition to provide quantitative predictions regarding population changes in natural systems. More importantly, the LV model is completely inadequate for assessing how other species or resources in a food web should affect the outcome of competition. Chase et al. (2002) illustrate this claim for the question of how predators alter competition between prey species. Letten and Stouffer (2019) is one of the latest in a string of very recent studies to reemphasize the dependence of pairwise competitive effects on other species.

Saavedra et al. (2017) was the one article in the highly cited group reviewed in Chapter 4 that relied entirely on LV models.

They end their article by stating that, ‘...future theory exploring how and whether higher-order interactions in multispecies system influence coexistence would be an important next step’. The problem is that the next step had been taken in the 1970s, before the confusing term ‘higher order’ had been invented.

12.2.2 The relationship between theory and experiment

The development of all ecological theory has been hampered by a belief that the only role of theory is to make one or more predictions that can immediately be verified or contradicted by a short-term experiment in an empirical system. This seems to underlie the opening quotation by Getz (1998). A prominent example of this ‘immediate empirical test’ viewpoint is the work of Peters (1991), who essentially argued for reducing ecology to finding significant regression relationships between variables. Such relationships always allow one to make simple (although not necessarily valid) predictions. A very recent article having many prominent authors opens with the claim that, ‘Scientific inquiry should operate as a feedback loop in which theory that describes the natural world is developed, tested empirically through carefully articulated hypotheses, modified to better represent reality, and then tested again’ (Grainger et al. 2022).

In spite of the views covered in the preceding paragraph, ‘testability’ is not, or at least should not be, the main criterion for judging all theory for ecological systems (and many other complex systems). Large numbers of theoretical models are needed to develop hypotheses about how complex biological systems might behave under a wide range of circumstances. In ecology, most of the currently popular models are so simple that they cannot be meaningfully used in this way. Slightly more complicated models are needed to identify what earlier assumptions might be misleading for particular circumstances, and what aspects of the population structure and dynamics might be needed to make a better prediction (or a range of potential predictions) for any specific system. Because a wide range of functional forms are known to be possible for most if not all model components, different predictions are likely from assuming different components.

Even a single intermediate-complexity model often reveals a much wider range of dynamics and responses to altered conditions than what arises in the simpler model upon which it is based. The intermediate models should use functional forms that have some justification, either from empirical measurement of these relationships, or from purely logical deductions. For the near future, the latter (logical) type of justification will often be more important than empirical measurements in community ecology, given the relative lack of quantitative study of the components of consumer-resource relationships in natural environments.

A common mistake of empirical biologists is to believe that all theory must be empirically tested. The propositions that need to be tested are claims about the causes of observed phenomena or predictions of future population trajectories. These causes or predictions need to be based on an understanding of the range of possible underlying mechanisms. It is this understanding that can be achieved by having a variety of different models with intermediate complexity.

12.2.3 What will a more comprehensive theory look like?

A fully developed ecological theory of competition (or any other ecological interaction) should have multiple levels of complexity. The minimum level of complexity for competition theory is one that includes simple forms of resource dynamics and consumer functional and numerical responses. These constitute the minimum because of the role of resources as the defining feature of the interaction. Part of this minimum was investigated more than 40 years ago, but many of the relatively basic elements—particularly adaptive foraging and the many potential forms of interference competition—have still received very little attention.

Our current knowledge of simple consumer-resource models will probably prove insufficient to make predictions that are likely to apply, even qualitatively, to most ecological communities. The necessary knowledge required includes a greatly expanded set of consumer-resource models that includes more of the food web context.

However, this alone will likely not be sufficient. In most real systems, the role of size structure and/or stage structure is an important determinant of the nature of interspecific interactions. Knowing how stages differ in resource use is a key determinant of the effect of stage structure on interspecific interactions. Many of the effects of population structure were reviewed by de Roos and Persson (2013). Genetic and phenotypic variation within species is another aspect of population structure that often needs to be included in competition theory (Schreiber et al. 2011). Similarly, spatial structure has important effects on interactions, some of which were covered in Leibold and Chase (2018), and more briefly reviewed here in Chapter 10. However, the bodies of work on stage-structured or spatially structured competition have an overrepresentation of models with either no resources, or with the simplest possible assumptions about resource growth functions and consumer functional and numerical responses. Furthermore, it is unlikely that population and spatial structure will be the only elaborations required for understanding the dynamics of competing species that are embedded in larger communities. The issue of what species/entities at other trophic levels need to be included to construct a predictive model is another important question that has repeatedly been discussed (Schoener 1993; DeAngelis and Mooij 2003; Loreau 2010). These effects of other levels are treated in Section 12.5.

Some promising developments in the analysis of predator-prey models suggest a movement towards examining the consequences of changing non-neutral parameters in the context of consumer-resource models. Uszko et al. (2017) is an example that focuses on the response of predator-prey dynamics to warming temperatures. Both the inclusion of type III functional responses and explicit resource dynamics produced significant changes in the predicted population-level responses in this study.

In general, the push towards ‘trait-based’ theory (McGill et al. 2006; Klausmeier et al. 2020a) has increased the use of competition models with explicit resources.

Despite some very recent moves towards a broader range of models of interspecific interactions, publications over the last several decades suggest that ecological theory is still playing a minimal role in applied issues. These are the areas where accurate prediction is most important. The applied fields of natural resource management (e.g., fisheries regulation) and conservation biology have largely failed to incorporate models from community ecology. A recent literature analysis of articles published on conservation between 2000 and 2014 (Hintzen et al. 2020, p. 721) suggests that ecology and conservation biology are diverging, ‘...possibly from rising scepticism about the relevance of contemporary ecological theory to practical conservation’. Fisheries biologists have increasingly called for ‘ecosystem-based’ approaches, which include interacting species. However, thus far few works in fisheries have incorporated the type of consumer-resource models discussed in this book (a problem discussed in Matsuda and Abrams (2006), Abrams (2014), and Abrams et al. (2016). In their analysis of conservation biology, Hintzen et al. (2020, p. 730) conclude with the hope that ‘mechanistic models... which are still new... may permit transparent, general, and useful ecological forecasts’. While I share that hope, the fact that mechanistic models could be regarded as new reflects an unacceptable rate of forgetting the recent past.

12.3

<< | >>

↑

Source: Abrams Peter A.. Competition Theory in Ecology. Oxford University Press,2022. — 336 p.. 2022

Theory’s roles in ecology and competition

More on the topic Theory’s roles in ecology and competition: