Gotelli: Chapters 2 & 3
Hilborn & Mangel: Chapter 3
There are three primary (scientific) goals for this lab exercise. First, you will learn some of the common plants and animals that live on Prospect Hill. Second, you will learn some basic techniques for sampling species distribution and diversity. Third, you will learn how to compare species distributions from different habitats. In addition, you will learn something about probability distributions, randomization tests, using the primary literature, and hypothesis testing.
We will sample three wooded plots on top of Prospect Hill, on just to the north of the water tower, and one just to the south. These square plots, each 625 square meters (25 meters on a side), were established by the Biology 331 class in the spring of 1991, and have been used since then for studies of forest ecology and succession (data from past years are available by clicking here). Each plot constitutes a sample of the total vegetation that occurs on Prospect Hill. The entire area was abandoned from agricultural cultivation in the early 1900s, and the current forest is approximately 40-50 years old. The three plots were relatively similar in vegetation composition when we first sampled them in 1991, but most of the canopy trees in the south quadrat were damaged or killed during a severe windstorm in the summer of 1991. Since 1992, we have been observing the regrowth in the southern plot, and comparing it with the more established (but still changing) northern plot, and a new plot set up in 1992. All trees > 5 cm DBH (Diameter at Breast Height, approximately 1.3 m above ground) are marked with permanent, numbered, aluminum tags.
During this week, we will:
Identifying and counting plants: Because we live in the well-studied north temperate zone, identification of the plants in these plots is not particularly difficult. Basic field guides in the lab (such as Peterson's Field Guide to Trees and Shrubs, Peterson's Field Guide to Wildflowers, and Newcomb's Field Guide to Wildflowers) are more than adequate to identify to species virtually every plant in these plots. Because many of the plants, especially the herbs, are clonal, it is sometimes difficult to decide what exactly is an 'individual'. For this lab, we will consider each rooted stem as a distinct individual (even if many of them belong to a single genetic individual). In other words, we will be counting ramets as opposed to genets (see Lovett-Doust 1981, or White 1984 for an introduction to some of the unique demographic properties of clonal plants). While time-consuming to obtain, counts of individuals within a community are the necessary data needed to compute measures of species diversity.
Measuring trees: Because of the world-wide economic importance of wood-products, foresters have developed a set of standard methods for sizing trees. One of the best predictors of tree volume or biomass is the diameter of the tree trunk (or bole) measured 1.3 m above the ground (the breast height of the average forester), denoted as DBH. Predictive regression equations (with coefficients of determination, or r-square values usually > 0.90) for the volume and biomass of northeastern trees based on their DBH are given in Tritton & Hornbeck (1982). Because it is difficult to measure tree diameter directly without killing the tree, foresters measure tree diameter by measuring tree circumference with a special tape measure (a DBH tape) scaled as 2(pi):1. Convince yourself that you can measure tree diameter this way. What assumptions go into the construction and use of a DBH tape?
While we will work with the raw DBH values in our explorations of size distributions, ecologists more commonly work with measures of biomass (dry mass) when assessing size structures of plant or animal populations. Why do you think measures of dry mass are more informative or useful than linear measures of organism size?
Why mark new trees? These permanent plots are a valuable resource developed by students who take this class. They have been used for studies of plant competition, forest succession, and species diversity. The long-term value of these plots is enhanced through their regular maintenance. By noting when new trees recruit into the 5-cm-DBH size-class, we are able to develop a more accurate assessment of the structure and dynamics of these stands.
Data management: One of the most important parts of scientific research is the organization and maintenance of accurate data and associated documentation (metadata). Data are maintained most commonly in a written data notebook or in computer files (spreadsheets, digital images, etc.). In this class, you are expected to maintain your data both a hand-written data notebook and in computer files. The data collected in this lab can be tabulated in several ways; the most useful way to organize them for subsequent analysis will be in a spreadsheet (we use Microsoft Excel for spreadsheet operations). An example of such spreadsheet entry would be:
Plot Tag number Species name DBH North 101 Acer saccharum 13.5 South 102 Acer rubrum 22.1 North 103 Quercus alba 18.6 ... ... ... ... Each individual tree is identified by its plot, tag number, its species name, and its DBH. Another file might contain a record of all species abundances in each plot, where the three colums would be: Plot, Species name, and abundance (number of individuals).
Metadata contains information and descriptions of the dataset so that others can use it. The Ecological Society of America (ESA) is in the process of developing metadata standards for ecological data management. These standards are based on those presented in the paper by Michener et al. (1997); an example can be found within the ESA's Data Repository. You should make it your standard practice to fully document all of your datasets as you enter them into the computer. Metadata documentation is normally stored in a word-processed file that is logically linked to the data file itself.
For this week's field work, you will be provided with Field Guides, DBH tapes, and data sheets. The raw data sheets themselves, as well as the electronic, compiled data (spreadsheets and associated metadata) will become a part of the permanent data archive of this class.
During this week, we will:
Computation of diversity indices can be somewhat confusing. Gotelli & Graves (1996) discuss many of the issues surrounding the meaning and measurement of species diversity in chapter 2 of your textbook. I suggest that you compute several measures of diversity, such as species richness (S, or the total number of species in the plot), Shannon-Wiener diversity (H', equal to -Sp[i]*ln(p[i]), where p[i] is the relative proportion of the ith species, and ln is the natural logarithm), Shannon-Wiener eveness (J, equal to H' divided by ln(S)), and Hurlburt's probability of interspecific encounter (PIE, equal to (N/(N - 1))(1 - S((p[i]^2)), where N is the total number of species, p[i] is the relative proportion of the ith species, and ^2 represents "squared" (raised to the power of 2), and investigate their interrelationships (correlations). Based on their relationships (or lack thereof), what can you conclude about species diversity on Prospect Hill?
Rarefaction is a statistical technique used to compare diversity indices from different samples. Because all measures of diversity are dependent on overall species richness, and because species richness is likely to depend on sample size (number of individuals), it is only appropriate to compare diversity indices from different samples if their sample size is the same. In other words, if one plot has 60 trees, and the other 30 trees, you should compare the diversity of the first plot to that of the second plot using a sample size of 30 from each plot. But in the first plot, which 30 trees do you use? Rarefaction is a random sampling technique through which you can appropriately sample your larger sample (in this case, 60) to yield a smaller sample (in this case, 30). You then can compare diversity indices of your rarefied sample with the smaller of the two original samples. See Gotelli & Graves (1996, pp. 24-31) for details. Rarefaction can be accomplished easily using the EcoSim computer package, or you can program it yourself.
Graphical techniques are a standard part of any scientist's toolbox. You should explore different ways of illustrating your data on species diversity and stand size structure (see below). You are strongly discouraged from using spreadsheet graphics (such as those found in Excel), since they are not optimized for scientific presentation. Keep in mind that pencil and paper works as well as (and often better than) computer software. Also be aware that your instructor has published a paper on graphical display of ecological data.
Analysis of size structure is as straight-forward as computing indices of species diversity. Which is to say, there are many ways to accomplish this. At root, the analysis of size structure is related to assessing degree of variability. Standard measures of variability (standard deviation, variance, skewness) are dependent not only on sample size, but also on mean (average) size, so you are better off using a measure, such as the coefficient of variation (standard deviation divided by the mean) that is not dependent on the mean. Weiner & Solbrig (1984) discuss these issues in greater detail. As with your analysis of species diversity, I suggest that you compute several measures of size structure and contrast their behavior with your data.
Generating predictions is the essence of science. What good are all the data you've collected if you can't use them to predict outcomes of similar, but not-yet-done, experiments (think about that!). Based on your data analysis, what would you expect the species diversity and size structure to be in an identically-sized plot located approximately 50 m north of the sampled 'north' plot?
During this week, we will:
The sampling, data management, and analysis of these data should be done using the same methods you used during the first two weeks of this lab.
You should write up the results from this lab in standard format, in no more than 10 pages. The introduction and discussion of your lab report should reflect your research into, and reading of, the primary ecological literature on this topic.
At a minimum, your lab write-up should address the following seven questions:
Ellison, A. M. 1993. Exploratory data analysis and graphic display. Pages 14-45 in S. M. Scheiner and J. Gurevitch, editors. Design and analysis of ecological experiments. Chapman & Hall, New York.
Lovett Doust, L. 1981. Intraclonal variation and competition in Ranunculus repens. New Phytologist 89:495-502.
Michener, W. K., Brunt, J. W., Helly, J. J. , Kirchner, T. B., and S. G. Stafford. 1997. Nongeospatial metadata for the ecological sciences. Ecological Applications 7:330-342.
Tritton, L. M., and J. W. Hornbeck. 1982. Biomass equations for major tree species of the northeast. United States Department of Agriculture, Northeastern Forest Experiment Station, General Technical Report NE-69.
Weiner, J., and O. T. Solbrig. 1984. The meaning and measurement of size hierarchies in plant populations. Oecologia 61:334-336.
White, J. 1984. Plant metamerism. Pages 15-47 in R. Dirzo and J. Sarukhan, editors. Perspectives in plant population ecology. Sinauer Associates, Sunderland, Massachusetts, USA.