Trait-Based Clustering As An Indicator Of Species Competition

Competition is one of the major types of interactions between species in nature. It is especially important in biodiverse communities such as tropical forests, where a large number of plant species must compete for a small number of resources, such as water, light, and nutrients. Yet, we still know little about how competitive interactions influence what kinds of species can coexist.

Considering that other forces such as dispersal and chance events are known to affect ecological dynamics, one pressing question in community ecology is whether competition leaves a recognizable and detectable signal in species assemblages. For example, can we say anything general about what characteristics provide species with a competitive advantage over others such that they become abundant in their community or can invade existing communities?

Species phenotypic traits mediate competition. For example, two species with similar traits are likely to have similar resource needs and similar strategies for acquiring them and, therefore, compete more strongly with each other than two species with dissimilar traits (Figure 1A). It follows that communities where competition is a driving force may display specific patterns in the phenotypic composition of their species. Imagine a community with three species that cannot coexist because they are too mutually similar. The species with the intermediate trait will be under the strongest competitive pressure and eventually will be competitively excluded. Intuitively, this should lead to communities where species are more mutually dissimilar than expected by chance. Indeed, this idea of limiting similarity among competing species is a classical principle in community ecology, dating back to the 1950s. However, it has found limited empirical support, and in fact, more recent theoretical research has shown that it will only arise in special circumstances.

Figure 1. Schematic representation of competition models. A. Lotka-Volterra: competition is directly tied to similarity in species traits, here beak size in birds. The curve shows competitive impact by each species on the focal species, increasing with trait similarity. B. Resource partitioning: we assume a continuum of substitutable resources (seeds of different sizes), and species traits (beak size) determine their use of the different resources. C. Habitat partitioning: species with different traits (seed size) are optimally adapted to different environments (for example, soil pH). D. Competition-colonization tradeoff: depending on seed size, species trade-off fecundity (seed output) with the ability to win sites against other species. Image republished with permission from PLOS from https://doi.org/10.1371/journal.pcbi.1006688

In the three-species example above, it turns out that the more similar the intermediate species is to one of the other two, the longer it takes to be excluded. This may sound counterintuitive, and yet it has major consequences for how competitive exclusion plays out in communities with many species. First, a set of species whose traits are spaced enough from one another to stably coexist gets an advantage in the competitive sorting process, either because of initial conditions or because they are favored by the environment. Once these favored traits emerge, all other species are competitively excluded at a rate inversely related to their trait separation from the closest favored trait. This results in a transient pattern of distinct clusters of species with similar ecological strategies. Limiting similarity will only occur once all weaker species have been fully excluded, which may take an indefinitely long time.

In our study, we used simulations of competitive dynamics models to show that trait-based clustering is a robust signal of community assembly under competition, going beyond the simplified scenarios where it was previously known to occur.

We showed that dispersal of individuals into the community via immigration from surrounding communities turns these transient clusters into permanent features, as population losses to competition are counterbalanced by an inflow of new individuals. Further, we found that chance events known as demographic stochasticity will add noise to the pattern, but the clusters will typically still be detectable with appropriate statistical tools (Figure 2). The pattern is strongest when diversity is high – i.e. when there are many more species than clusters – and when immigration is intermediate – neither so low that most species are fully excluded between immigration events nor so high that it overcomes competition as the main process shaping the community.

Figure 2. Trait-based clustering in a species assemblage. A community of 232 species totaling 21,000 individuals, generated via simulation of a competitive dynamics model. Our statistical analysis reveals trait-based clustering among species, with 13 color-coded clusters in this example, where each species is represented by one stem. Image republished with permission from PLOS from https://doi.org/10.1371/journal.pcbi.1006688

To check that clusters emerge generally under competition, we analyzed models describing different types of competitive dynamics. In addition to the phenomenological model described above where competitors directly affect each other via trait similarity (Figure 1A), we considered competition for substitutable resources where species differ by preferred resource (e.g. animals with overlapping but different dietary preferences, Figure 1B), competition for recruitment where species differ by optimal habitat (e.g. plants optimally adapted to low vs high soil pH, Figure 1C), and competition for recruitment under a trade-off between seed production and seedling survival, known as a competition-colonization tradeoff (Figure 1D).

We found that clustering occurred generally across these distinct types of competition. The strength of the pattern varied across different scenarios: for example, under resource partitioning, clustering was stronger when consumers were able to substantially deplete resource populations because resource depletion caused species to cluster around the remaining resources. However, the key result is that clustering is a robust outcome of competitive dynamics, tending to emerge when simple and intuitive connections between competition and trait similarity hold.

While cluster emergence under competition was general, the characteristics of the clusters varied: both the shape of the clusters – i.e. the relationship between species abundance and trait separation from the cluster center – and the trait separation between cluster centers were highly idiosyncratic across competition types and scenarios. This variation in shape and scale suggests that one is more likely to succeed in detecting the pattern by focusing on general aspects of clustering than on the particular shape or distribution of the clusters. Moreover, a single one-size-fits-all tool for detecting clusters across species assemblages may not exist. Indeed, in our study, we observed that different methods were best suited for different competition scenarios.

Clustering as a signature of coexistence under competition is an update to the prevalent paradigm that competing species will display greater-than-chance mutual differences. Our finding that clusters appear under various types of competition and can be easily maintained by immigration, even when confounding stochastic forces are at play, suggests that clusters are a likely feature of nature beyond the instances where it is currently known to occur. For example, in tropical forests, where both competition and dispersal are recognized as key drivers of community assembly, clustering could help explain the observed high diversity and seemingly continuous phenotypic variation. In fact, our preliminary results applying our methods to forest data indicate that tropical forests may be structured into distinct groups of species clustered by maximum height. If confirmed, these results will support the hypothesis that competition for light is a critical force in these forests, and show that competitive dynamics actually shape the community into recognizable phenotypic patterns.

These findings are described in the article entitled Generalizing clusters of similar species as a signature of coexistence under competition, recently published in the journal PLOS Computational Biology.

About The Author

Rafael D'Andrea

Rafael D'Andrea is a community ecologist. He obtained his Ph.D. at the University of Michigan, and is currently a postdoc at the University of Illinois at Urbana-Champaign.

My research investigates how species interactions create order in complex ecosystems. I try to answer questions such as What are the relative roles of deterministic and stochastic forces in shaping communities and influencing which species get to coexist? Conversely, what can specific combinations of species and their traits reveal about the underlying forces moderating ecological dynamics? What are the key mechanisms behind species coexistence in highly biodiverse communities such as tropical forests?

Speak Your Mind!

READ THIS NEXT

Flexible Substrates For DNA Arrays

Array fabrication technologies date back to the early 1990s when Affymetrix first used ultraviolet light to control the stepwise synthesis of biopolymers from a surface.[1] The most commercially successful result of this technology was high-density arrays of DNA. At the time, DNA arrays were one of the most efficient ways to detect and measure nucleic […]

Cluster Headache – One Of The Most Painful Conditions Known To Mankind

Cluster headache is a devastating and debilitating disease that affects approximately 7 million people worldwide. It is characterized by highly intense headache attacks which are usually located around one eye and can last for up to three hours. The attacks often occur in clusters, reappearing up to eight times daily during a period of weeks […]

Cerro Blanco (Central Andes): The Largest Volcanic Eruption Of The Last 5000 Years

One of the biggest mysteries concerning the origin of many recent volcanic ash deposits in NW Argentina has been solved. New data and interpretation about a major eruption — spreading more than 100 km3 of ashes over about 500.000 km2 — occurred around 4200 years ago in the Cerro Blanco Volcanic Complex. This eruption was […]

Quantitative Observation: Definition And Examples

Quantitative observation is the type of observation that deals in quantifiable variables, things that you can use numbers to express. In contrast, qualitative observation focuses on the characteristics of specific events or experiences. Observation is necessary when it comes to scientific research and studies. To gain a better understanding of the difference between qualitative observation […]

Insect Pest Species Rapidly Evolve, And Not Just To Chemical Pesticides

Pest insects and mites attack agricultural crops and forests directly or damage them by spreading diseases. They can also spread diseases to humans and livestock as well as having an impact on our biodiversity by spreading diseases to endangered species. Attempts to control these organisms has tended to focus on the application of chemical pesticides. […]

Potential Application Of Vacuum Insulated Tubing For Deep Borehole Heat Exchangers

One of the ways to extract geothermal heat from the Earth’s crust is through Deep Borehole Heat Exchangers (DBHE). These can be either drilled especially for a heat extraction or, more cost-effectively, reconstructed from already-existing, negative, or abandoned oil or natural gas wells. In many depleted boreholes, exploiting underground geothermal waters is either impossible or […]

Trip To An Asteroid: Hayabusa 2 Is Now Traveling Toward C-type Asteroid Ryugu

Still, in the 21st century, we don’t have the answer to this big question: How our Earth, life, and this solar system were made? The key is given by organic materials and water in space. Asteroids mainly located between Mars and Jupiter are primitive bodies: they have escaped large impacts or destruction and luckily preserve […]