COS 217-1 - Fire, bison reintroduction, and tipping points in a grassland

Abstract: Tipping points are where a small change in an external driver results in disproportionately large ecological response. Most of our experimental understanding of tipping points comes from small-scales and/or systems dominated by short-lived organisms (e.g. algae in lakes). We conducted a long-term experiment that includes three fire frequencies (1, 2, and 3-4 years) factorially crossed with the presence or absence of native megagrazers (Bison Bison bison). We founded that in simplified grasslands (no megafauna present) there is clear tipping point between disturbance (fire frequency) and woody plant expansion (WPE). Areas with a fire return interval (FRI) of one- to three-years maintained shrub areas between 0 and 8% and negligible woody plant biomass, whereas areas with a slightly higher FRI of 3.5 years had a shrub area of 45% average shrub cover with similar increases in shrub density and biomass. To our knowledge, this is the first example of such a tipping point in a terrestrial ecosystem scale experiment. Areas with bison saw increases in shrub dominance with higher fire return intervals (1 to 2 years), but relationship between fire and woody plant expansion was gradual. We attribute this “unbending” of the threshold response to the added complexity of reintroducing bison, which create a negative feedback between frequent fire and fire intensity. This study has three implications: 1) the removal of influential species might ecosystems prone to non-linear dynamics, 2) bison increased WPE moderately at 1 to 2 year FRIs, but decreased WPE slightly at 3 to 4 year FRIs, and 3) we find that not all difficult to reverse state-changes will be preceded by non-linear relationships between drivers and ecosystem state—a major departure from our current search image of alternative states.