No species lives in isolation. Partnerships, rivalries and food webs determine which species persist, how ecosystems function, and how resilient they are to change.
Our lab works across these interactions, from cooperation between plants and insects to competition among rivals and the cascading effects of predators and large herbivores. We favor long-term, large-scale field experiments, because many of the most important ecological effects unfold slowly and over whole landscapes. Today, a recurring question in our work is how human-driven change, including biological invasions and the loss of large animals, reshapes the interactions that hold communities together.
Ecosystem resilience under invasion
- The questions
- Once an invader is removed, does a broken partnership come back, or has the ecosystem shifted into a new state? And does the answer depend on how long the invader was there, or on how much rain falls?
- What we've learned so far
- Our removal experiment began in 2025. Earlier work shows what is at stake: invaded trees lose their ant defenders and their protection from elephants (opens in a new tab), capture less carbon (opens in a new tab), and become poor homes for the native ants that might return (opens in a new tab).
- Why it matters
- Removing invasive species is one of conservation's most common tools, yet we rarely know whether ecosystems recover afterward. Our results will help show when restoration can rebuild what was lost, and when it cannot.
Interested? Read our 10 papers on invasion and mutualism disruption
Consequences of mutualism disruption
- The questions
- What happens to an ecosystem when an invasive species breaks a partnership at its foundation? And how far do the effects spread?
- What we've learned so far
- The invasive big-headed ant wipes out the whistling-thorn's most protective ant partners (opens in a new tab). Without their bodyguards, trees are battered by elephants (opens in a new tab), capture less carbon (opens in a new tab) and decline (opens in a new tab), and the savanna opens up. These changes ripple outward through the community, reaching even the interactions between predators and their prey (opens in a new tab).
- Why it matters
- Conservation usually focuses on preventing extinctions. But breaking a single partnership can transform an ecosystem, so protecting healthy ecosystems also means identifying and preserving the most important interactions among species.
Interested? Read our 10 papers on invasion and mutualism disruption
Carbon allocation in savanna trees
- The questions
- How do ant partners shape the way a tree captures carbon, uses water and builds its roots? And what happens to that hidden economy when the partnership breaks down?
- What we've learned so far
- Different ant partners leave distinct signatures on their host's carbon and water budgets (opens in a new tab). Trees with active defenders grow more roots (opens in a new tab), while invasion cuts carbon capture (opens in a new tab) and reshapes the root system (opens in a new tab). Even shade shifts how young trees divide carbon between roots and shoots (opens in a new tab).
- Why it matters
- Mutualisms are everywhere, yet they are rarely considered in how plants spend their carbon. Knowing how partners steer a tree's carbon budget helps predict how invasions and the loss of herbivores will change carbon cycling across whole landscapes.
Interested? Read our 8 papers on carbon allocation in savanna trees
The ontogeny of mutualism
- The questions
- How can cooperation persist when some partners give little in return, or even harm their hosts? And how do the costs and benefits of a partnership change over a long life?
- What we've learned so far
- The protection an ant provides is a poor predictor of its overall value to the tree (opens in a new tab). Trees do best by hosting a sequence of different partners, including cheaters (opens in a new tab), and the partnership shifts with temperature (opens in a new tab), competing grasses (opens in a new tab) and a third partner, scale insects (opens in a new tab).
- Why it matters
- Most mutualisms involve several partners that come and go over a host's life, yet theory usually considers one pair at a time. A lifetime view helps explain why cooperation in nature persists despite widespread cheating.
Species coexistence
- The questions
- Four ant species compete for the same trees, yet none wins outright. What keeps them all in the game?
- What we've learned so far
- Battles are wars of attrition decided by colony size (opens in a new tab). Weaker species persist through trade-offs between fighting and colonizing (opens in a new tab), patchy resources around termite mounds (opens in a new tab), remodeling their host to resist takeover (opens in a new tab), and indirect help from the dominant species (opens in a new tab).
- Why it matters
- How competing species coexist is one of ecology's oldest questions. The acacia ants show how several mechanisms can work together to maintain diversity, offering a model for communities everywhere.
Large mammal decline
- The questions
- What happens to a savanna when elephants, giraffes and other large mammals disappear? We answer this with long-running experiments that fence large animals out (opens in a new tab).
- What we've learned so far
- Losing large herbivores causes the ant-acacia partnership to break down (opens in a new tab), transforms plant-pollinator networks (opens in a new tab) and depletes dung beetle communities (opens in a new tab). Predators matter too: fear of leopards and wild dogs shapes which trees thrive (opens in a new tab).
- Why it matters
- Large mammals are declining worldwide. Our experiments show that they hold together far more of the ecosystem than expected, and that their loss can unravel interactions far removed from the animals themselves.
Termites as ecosystem engineers
- The questions
- How do termites, by building regularly spaced mounds, shape the plants, animals and productivity of a savanna?
- What we've learned so far
- Mounds are hotspots of animal life, and their even spacing boosts productivity across the landscape (opens in a new tab). Trees near mounds rely less on nitrogen fixation (opens in a new tab), are more likely to fruit (opens in a new tab) and host different outcomes of competition among their ants (opens in a new tab).
- Why it matters
- Termites show how small organisms can engineer entire ecosystems, and how spatial pattern in nature, not just which species are present, governs how ecosystems function.
Interested? Read our 4 papers on termites as ecosystem engineers
Island food webs
- The questions
- Do top predators protect biodiversity, as ecologists have long assumed? And how quickly does evolution respond when food webs change?
- What we've learned so far
- On experimental islands, fear of an introduced predator broke down the niche differences that let two lizard species coexist (opens in a new tab). The lizards also evolved in ways that proved hard to predict (opens in a new tab), and colonizing new islands left a rapid epigenetic signature (opens in a new tab).
- Why it matters
- Predators are being lost and introduced around the world. Island experiments let us test how these changes rebuild food webs, and they show that predators do not always promote diversity.
Our research is supported by the National Science Foundation. See the publications page for papers from each project.