Fig. 8.1. Series of events that must occur before competition between two individuals will influence populations or communities.
ronments - recall that one of the conditions for competition was the existence of limited resources.
Like so many concepts in ecology, the issue is generally one of semantics - in this case it is about what we define as 'resource-rich' and 'resource-poor'. There is no good way to quantify the division between these two. Hence, most studies find evidence for both theories, though a harsh environment (in terms of its local climate, for example) will create resource-poor conditions. Weeds and other plants in these environments may be limited by abiotic conditions rather than biotic interactions like competition. This is because they may not be able to grow in large numbers and densities to create an excess demand on available resources (see Grime, 1979; Goldberg, 1985; Grace, 1990; Tilman, 1990; Aarssen, 1992, for detailed discussions).
The effect of competition on populations and communities
Establishing the relative importance of competition in real world habitats requires evidence that competition affects population size or biomass, or that it affects community composition (Goldberg and Barton, 1992; Goldberg et al., 1995); to do this, a series of events must occur (Fig. 8.1). Even if we can establish that there is an effect on populations and communities, it may be ephemeral and the population or community may return to its original state. Complicating matters further is that even if observed patterns suggest competition is causing them, there can be other explanations, e.g. current interactions such as predation or environmental changes. Patterns also may be the result of unknown past (historical) events or interactions. Attempting to determine the exact cause (i.e. was it caused by competition) is like chasing the ghost of these past events (Connell, 1980). What ultimately makes competition at the population and larger scales difficult to examine is that the pattern we see results from millions of competitive interactions between individuals. Hence, ecologists often focus on individual-
scale experiments and extrapolate to populations and beyond. By examining the variation in the competitive interactions between individuals, it is possible to determine what is likely to happen at larger scales, though we caution again that populations, communities and ecosystems are the result of more than just competition and individual interactions.
What evidence do we have that competition occurs?
When you observe a habitat, how can you tell if individual weeds (for example) are competing or not? Competition involving weeds is subtle and difficult to observe but perhaps easier to test than for whole populations or communities. Still, in a meadow, you do not see weeds throwing punches at each other. Thus, ecologists often use controlled experiments to test whether inferences made in the field about competition are real. The easiest way to detect competition experimentally is to grow plants together (in competition) and apart (without competition), and compare their growth or survival. There are many variations on this basic experimental design (see Chapter 10).
While these types of paired experiments are useful, they are limited in their broad-scale application because they ignore all other species normally found in a habitat. The design of experiments can be more complex. For example, if we wished to look at the competitive effect of weeds on an unmanaged forest ecosystem, we would have to decide what to measure (e.g. survivorship or growth), the number and types of species to measure it on (e.g. one weed species or many species, weeds and otherwise), and what time span to cover (e.g. 1 year or 10).
One example of how competition is assessed is from Wilson and Tilman (1995). They planted three species of native grasses into prairie soil and measured their growth in three competition treatments: competition for light and nutrients by allowing roots and shoots to interact, competition for nutrients by excluding the shoots from interac tions, and no competition by allowing neither roots or shoots to interact. They found that excluding the shoots of neighbours had no effect on plant growth, i.e. they prevented any possible competition for light but competition for nutrients still was possible. Excluding both shoots and roots caused a drastic increase in growth of all three species, i.e. when the possibility of competition for nutrients also was prevented, the plants benefited. Tilman (1988, 1990) also added a variety of soil nutrients and found that plants responded only when nitrogen was no longer limiting; therefore, he concluded that competition was for nitrogen. This study illustrates that:
• some grasses are better than others at 'exploiting' nitrogen while in competition with one another;
• there were several types of resources (including nitrogen) for which grasses might have competed.
The pattern of adult plants reflects the competition for nitrogen between individuals.
Weed competition studies generally use similar approaches and reach similar conclusions. Bergelson (1996) reviewed her experiments on competition between two weeds, annual bluegrass (Poa annua) and common groundsel (Senecio vulgaris) (see also Bergelson, 1990, for example). By com paring the growth of each species alone with situations where groundsel was surrounded by annual bluegrass individuals, the outcome of competition could be studied. The fundamental results reinforce the general patterns that competition involving weeds often exhibit. The main pattern is that even small differences in seed germination and seedling emergence matter - earlier ones often are better competitors. In theory, getting a head start on competitors means that a weed is competing with others of a similar age (i.e. all start out as seedlings). If this is true, then 'neighbourhood competition' is important (Pacala and Silander, 1990). This means that competitive ability is useful in densely vegetated environments but it also means that some individuals avoid competition because of chance. Avoidance is not a strategy - some individuals happen to colonize a 'neighbourhood' area that is relatively free of competitors of similar ages and have higher fitness regardless of their competitive ability. 'Gap colonization' may be equally or more important in influencing whether competition between individuals, i.e. competition is not always influenced by interactions between co-generational individuals but by the gaps left open or still occupied by the previous generation. Competition can be affected by the influence of the previous or older generation. In
Direct Indirect interaction interaction
Indirect interaction Indirect interaction via shared enemy via another plant _species_
Natural enemies E
Key -direct interaction
-• negative effect
Fig. 8.2. Types of traditional competition and apparent competition showing direct and indirect interactions (redrawn from Connell, 1990).
Bergelson's (1996) review, she explains how the main influence on competition's outcome was that the litter from dead bluegrass prevented groundsel seedlings from emerging. Though this is not what Connell (1980) meant by the 'ghost of competition past' (see previous section), the analogy here is similar sounding because competition is between live groundsel and the 'ghost' (the litter) of dead bluegrass.
Was this article helpful?