IT News & Events

News about IT at Indiana University and the world

Image showing the numbers of species that persist to the end of Dr. Muthukrishnan's simulations when plasticity is or is not included, and the initial species pool is (a) 2 species, (b) 5 species, (c) 10 species, (d) 100 species or (e) 200 species. The dark line within boxes indicates the mean, and boxes represent the interquartile range.

Thrive or survive? HPC simulations, species plasticity, and plant coexistence

Ranjan Muthukrishnan of the Environmental Resilience Institute continues to work at the leading edge of research on species plasticity using high performance computer (HPC) simulations. His recent work, in collaboration with Lauren L. Sullivan, Allison K. Shaw, and James D. Forrester, takes a new look at how species are able to coexist in ecosystems without excluding each other, and how that can help ecosystems become more diverse. The group evaluated this question by expanding on a well-established ecological theory, the competition-colonization trade-off. This theory has long held that, within an ecosystem species could coexist with each other, if species that were good competitors, they were also slow to disperse to new habitats, while strong colonizers disperse with ease, but are outcompeted when good competitors show up in their habitat.

Ranjan Muthukrishnan, Environmental Resilience Institute
Ranjan Muthukrishnan, Environmental Resilience Institute

As Muthukrishnan notes, this trade-off relies on the notion that the traits that prompt these trade-offs are fixed; parent plants always produce offspring just like themselves, independent of the environment. He decided to see what would happen if they accounted for the high degree of trait plasticity exhibited by many plants, where they can make seeds of different sizes depending on the conditions where they grow. To do so, he created simulations on IU’s Big Red 2 and Big Red 3 supercomputers in order to evaluate the effect of trait plasticity on plant coexistence. 

To make this work, Muthukrishnan used a spatially explicit (40 x 40 cell grids with each cell representing a 100 x 100m habitat patch) simulation model of annual plant competition that allowed species to alter the size of the seeds they produced in response to local resource levels. He also ran replicate simulations under differently arranged landscapes and trade-off strategies and tracked the abundance of all species to quantify the likelihood of coexistence under each set of conditions.

If one were to consider only one set of conditions, or plants that all acted the same, this type of simulation would not be overly complex; playing it out in lots of different circumstances and to let individuals across the landscape have different behaviors that are constantly responding to current environmental conditions, the computational requirements expand immensely. This makes high performance computing resources, like those IU provides through Big Red 2 and Big Red 3, essential. Playing out simulations in a way that includes a whole landscape of possible locations requires a large amount of computer memory. IU’s supercomputers give the large-scale simulations ample time to run through different scenarios and show results even when they depend on important, rare events.

Muthukrishnan’s simulations show that the classic competition-colonization trade-off is sensitive to environment, and that coexistence can only be sustained within certain conditions. Plasticity, when brought into the equation, allows for competitive hierarchies to shift, leading to more coexistence in a broader range of conditions.

The simulations also show that plasticity increases the number of species that persist in simulations of multispecies assemblages. Plasticity may also generally increase the robustness of coexistence mechanisms and be an important component of scaling coexistence theory for communities with greater species diversity.

Read more about Dr. Muthukrishnan’s work here.