Melinda D. Smith of Colorado State spoke this week at Davis’s Ecology and Evolution seminar.

These are my rough notes from the talk. Any errors or misrepresentations are my own.

Also, a shorter version of the talk is available as a PDF here

  • Melinda’s general framework: Ecological implications of global change. How human caused global changes (climate, biochemical modification, land use), ultimately change community structure and function.
  • Grasslands (inc. steppe, tundra, savanna, and scrublands) cover 40% of earth’s surface. Sequester large amounts of C, economically important for grazing, house most remaining megafauna
  • Shaped by climate variability, both directly and indirectly through grazing and fire. All these regimes are being modified by humans.
    • Increased fire frequency
    • Timing, location, and intensity of grazing.
    • Big question: can we generalize findings about these effects across grasslands?
  • Focus of this talk is savannah grasslands: mesic savannah, where rain is >500mm/yr, and C4 grasses dominate
  • Some argue fire and grazing affect ecosystem structure/function in fundamentally different ways across different grasslands
  • How are S. African grasslands different than North American?
    • Scholes et al. (2003) argued that the systems are different
    • SA grasslands have longer evolutionary history due to no glaciation events
    • They have lower soil fertility
    • Greater large herbivore and plant diversity
  • Big questions:
    • Can we forecast impacts?
    • Are differences between knowledge of systems based on different approaches by scientists?
    • Are impacts differences between the systems?
  • Take a comparative approach
  • Manipulate fire and grazing
  • Study sights
    • Two long-term experiments: Konza Prairie LTER, Kruger Nat’l Park
  • Comparison
    • Similar biomass response to precipitation
    • Similar tree/shrub/grass balance
    • Elephants in S. African site, no equivalent in N. America
    • Fire key control on woody plants
    • Question: What about the herbaceous plant community?
    • Differences: Soil nutrients, temperature, herbivore richness, but similar in when rainfall comes.
    • Not comparing apples to oranges, orchards to orchards
  • Overarching hypothesis: Fire and grazing will be broadly convergent in their effects on ecosystem processes, but the plant community responses will diverge
    • Why? C4 grasses dominate both, but in SA there is much more diversity due to older system (@Scholes2003a)
  • Experiments
    • Removed large herbivores using 40m2 exclosures
    • Used existing 1, 3-4, and never-burn treated areas
    • n=63
  • Results
    • Divergence in plant community response. In SA, community is richer in unburned than in 1-yr communities. No effect in N. America. Similar overall plant diversity. But on 2-m2 scale, SA plots are less diverse due to lower spatial turnover
    • Response to fire and grazing together.
      • Rapid decline in richness in exclosures in N. America
      • In SA, decline was lagged an occurred inly in the absence of fire
    • In N. America, only grazing affected plant group composition. In SA, only fire affected it.
    • For NA, compositional differences driven by reduced abundance of a few C3 species.
    • For SA, it was due to turnover of dominant species and (C4 and legumes)
  • No difference between the systems in terms of NPP. Both continued to be driven mostly by annual precipitation.
  • Unexpected findings
  • Differences between fire/herbivore interactions:
    • In NA, the fire-grazing interactions arises become one species dominate irrespective of fire regime Bison graze these out. See Collins (1998).
    • Magnitude and speed of plant community response was much greater in NA than in SA.
  • How to think about this?
  • Hierarchy of mechanisms expected to differentially drive community and ecosystem change with removal of grazing:
    • individual response –> species re-ordering –> species migration
  • Differences in grazing intensity does not appear to be driving different responses. Similar levels found in both systems.
  • Future directions
  • Phylogenetic analysis (work by colleague)
    • So far, found significant clade overlap but no species overlap
    • Only in SA do we see differences among clades in fire response
  • This comparison suggests generalzation may be easier to do at the ecosystem level, but not at the community level. If we understand the relationship between those two scales, we’ll have a better handle on prediction.

Q & A

  • What relevant traits should we be measuring?
    • Still figuring this out, but currently looking at nutrient content, tensile strength, root/shoot ratio, traits thought to be related to fire and grazing
  • What are management implications?
    • Aiming to provide managers with info to explain how they might modify composition/diversity and still maintain functions they want. Also, understanding fire frequencies that maintain diversity
  • How does the historical temporal/spatial distribution of fire compare to managed fires?
    • Definite differences. Fire was much more patchy in the past, and at much larger scales. We’ve simplified the system. Return interval is thought to be 3-4 years in both NA and SA.
  • Would NA behave more like SA if we had the same herbivore assemblage?
    • Probably not, because of the differences in plant community and diversity


Collins, S. L. 1998. Modulation of Diversity by Grazing and Mowing in Native Tallgrass Prairie. Science 280:745–747.

Scholes, R. J., W. J. Bond, and H. C. Eckhardt. 2003. Vegetation dynamics in the Kruger ecosystem. in H. C. Biggs, J. T. du Toit, and K. H. Rogers, editors & translators. The Kruger Experience: Ecology And Management Of Savanna Heterogeneity. . Island Press.

← An R Users' Group in Davis | All posts | Multiple Goals in Floodplain Restoration - Workshop Notes →