Background Information

Cedar Creek Natural History Area is a 2200 hectare (1 hectare = 2.47 acres) experimental ecological reserve that includes prairie/savanna habitat and prairie plants growing in abandoned agricultural fields. Prairies in Minnesota are dominated by bunchgrasses such as little bluestem, big bluestem, and Indian grass, and also include a large number of species of forbs such as lupine, purple prairie clover, milkweeds, and goldenrods. Savanna consists of scattered trees within a continuous stand of prairie grassland. Fire is common in these systems and the trees are restricted to a few fire-resistant species, primarily bur oaks.

The Biodiversity/Productivity Experiment

This experiment was set up to examine the effects of manipulating species richness on plant productivity and biomass. Human impacts are driving many species locally and globally extinct. These human impacts include habitat loss or fragmentation, habitat modifications such as pollution and introductions of exotic species, and climate change. Fragmentation or fast shifts in climate may limit the number of species that occur in a habitat simply because the seed of that species cannot reach that area. Our experiment is analogous to this situation in that we allow the seed of many species to reach some areas, and restrict the number of species able to reach other areas. We do not focus on the loss of a particular species, but rather ask whether the loss of biodiversity has any general, predictable effects.

In particular, we focus on whether biodiversity affects plant productivity. Productivity is the amount of plant biomass produced on a given area of land, over a given amount of time. If a grassland were being used to produce hay, its annual productivity would be the amount of hay that it could produce each year. We use aboveground plant biomass as a measure of annual productivity. In our system productivity approximates biomass because no aboveground plant biomass from the previous year survives to the current year. Each year, aboveground plant biomass either dies and decomposes or is consumed in the spring fires we set. Productivity is an important ecosystem trait, as all higher trophic levels depend upon it directly or indirectly as a food resource. In addition, maximizing productivity is a goal of many pasture, forestry, and agricultural systems, and it is possible that insights about the effects of biodiversity can be applied to some of these systems.

In 1993 the vegetation and seed bank were removed from an abandoned agricultural field. In spring of 1994, 168 plots, each 9 m x 9 m, were seeded to contain 1, 2, 4, 8, or 16 grassland-savanna species. All plots received, in total, 10 g m-2 of seed in May 1994 and 5 g m-2 in May 1995, with seed mass divided equally among species. Treatments were maintained by weeding 3 or 4 times/year. Plots were sampled in mid-August for aboveground living plant biomass by clipping, drying, and weighing four 0.1 x 3.0 m vegetation strips per plot from 1996 through 1999, and eight strips per plot in 2000. Different areas were sampled each year.

You will use the Excel spreadsheet software and provided data to create the following graphs and answer the following questions.

  1. Biodiversity Dataset 1
    1. The first column shows the number of species planted in a plot, each row represents a different treatment. In this experiment, the treatment is the number of species planted, which ranged from 1 to 16. Each treatment was applied to dozens of plots (ranging from 29 to 39 depending on the treatment). Each column to the right represents a different year, from 1996 to 2002. The first group of columns gives the value for aboveground biomass, averaged across all the plots of a given treatment. The next group of columns gives the standard error. The standard error is a measure of “confidence,” and tells us how much the average value is expected to vary if the experiment were to be run again. Small standard error bars indicate that our confidence in the accuracy of the estimate is high and large standard error bars indicate that our confidence is low.
    2. We will make a scatter plot graph to visualize the relationship between biomass and species richness and to help us interpret our data. As described below, you can use the chart wizard to create this graph. Although there are many types of graphs to choose from in Excel’s chart wizard, the two primarily used by scientists are the “scatter plot” and the “column.” Species richness is the experimentally manipulated variable, referred to as the predictor variable. By convention, we use the horizontal, or x-axis, to indicate our predictor variable, and the vertical or y-axis to represent our response variable. Species richness is a variable that can be represented quantitatively, and we therefore choose a scatter plot to graph our data. If we had categorical data (e.g., plots characterized simply as “low,” “medium,” and “high” diversity) we would choose a column graph.
    3. Make a graph of the average aboveground biomass versus species richness (i.e., the number of species). Start by highlighting the first two columns and clicking on the chart wizard. Select scatter plot on the graph wizard, select the plot sub-type “scatter with data points connected by smoothed lines,” and then select finish. This shows the patterns for the year 1996. We can now add years by selecting “add data” under the chart menu (NOTE: Excel often produces a graph of gibberish if you try to graph all the years at once). Select the new data, including the heading, for the years you would like to add. You can add the years one at a time to see how the changing patterns are revealed. Remember to label your axes (you can edit the axes labels by double clicking on them). When you finish, look at the graph and describe the pattern that you see (to each other if you are working in a group). How would you describe the pattern to another student looking at the graph for the first time?
    4. Add error bars to the figure. Double-click on one of the data points for the year 1996. When the dialog box opens, select the tab y-error bars. Select “custom” and add the column for SE1996 to both the “+” and “-” sections under custom error bars. Data points for which the error bars overlap are not significantly different from each other.
    5. Questions:

      1. Describe the relationship between the number of plant species and plant biomass.
      2. How does the relationship between biomass and species richness change over time? How does it stay the same? Do the standard error measurements change your interpretation of this pattern? What is the advantage of having data from more than one or two years?
      3. What conclusions can you draw, or hypotheses can you make, about the effect of the loss of biodiversity in natural systems? What are the problems that need to be considered when extrapolating the results of this experiment to natural systems?
  2. Biodiversity Dataset 2
    1. The numbers presented here show the percentage of plots with biomass exceeding the highest monoculture biomass. These data can address the sampling effect hypothesis: diverse plots have more biomass because they are more likely to contain a dominant species with high biomass. Under this hypothesis, one or a few high biomass species are good competitors and dominate the plots in which they are present. Consequently, the total biomass of communities containing these species is predicted to be the same as the biomass of these species when grown alone. In this scenario there should be an increase in the average biomass across diversity gradients, but no increase in the maximum amount of biomass across the diversity gradient.
    2. Questions:

      1. Do the data support or reject the hypothesis that there is no increase in the maximum biomass? Does the answer to this question depend upon the year (if so, state how the years differ)?
      2. What is the relationship between species richness (the number of species) and productivity within this savannah grassland?
      3. How does the relationship between species richness/productivity change over time?
      4. Why might a diverse plot contain more biomass than even the highest monoculture plot? Why might two species be better than one when it comes to biomass production?
  3. Bonus Biodiversity Dataset
    1. This dataset contains the “raw” data from the first three years of the experiment. These data were the source of the averages and the standard errors in Dataset 1. Make a graph similar to the one you made for Dataset 1.
    2. Questions:

      1. Can you answer the questions from #4 above using this graph?
      2. What is the advantage of using the summarized data in Dataset 1 versus using the raw data in the Bonus Dataset?
      3. What might be a disadvantage of using the summarized data?
    3. This dataset also contains information on the presence/absence of functional groups in each plot and can be used to illustrate the utility of different statistical approaches. The functional groups of plants in this experiment are: trees, C3 grasses, C4 grasses, legumes, and non-leguminous forbs. C3 grasses have the C3 photosynthetic pathway, while C4 grasses have the C4 photosynthetic pathway. Forbs are broadleaved herbaceous species. Legumes are forbs that have a symbiotic association with a bacterium which enables them to use atmospheric nitrogen (N2) directly.
    4. Questions:

      1. Is the increase in biomass with increasing diversity associated more with the presence of some functional groups than of others? How would you test this hypothesis? While previous data analyses used graphs with one numerical predictor (species richness), this question involves several categorical predictors. Simple linear regression is a statistical technique applicable to situations with one continuous predictor, whereas Analysis of Variance (ANOVA) is applicable to situations with several categorical predictors.