RENR 110

Biodiversity Lab Exercise 1

I. Background: As discussed in lecture biodiversity is used increasingly as an integrative

measure to 1) evaluate impacts on ecosystems that result from anthropogenic exploitation of
natural resources, and 2) investigate whether management activities intended to mitigate these
impacts actually achieve their goals. Essentially, we measure biodiversity in disturbed and
‘restored’ systems and ask if it is in the same general domain as the baseline condition in
undisturbed reference systems. In this lab exercise we will illustrate some of the basic analytical
procedures common in work about biodiversity. We will analyse a simple data set
representative of forest biodiversity on the boreal plain of northern Alberta. The raw data
available to you are the numbers of individuals of each of 20 ground-dwelling spider species
(columns) captured in pitfall traps run in two forest cover-types represented by replicated
broadleaf and coniferous stands. Data from the traps are captures of spiders from three
‘treatments’ (rows): 1) uncut mature forest, 2) clear-cut forest and 3) stands harvested with 10%
retention of standing green trees, as collected over a full summer season 5 years after stands in
treatments 2 and 3 were harvested. The data are presented as three replicates (each as a
separate row) of each cover-type x treatment combination.

II. Focal Taxon: It is a vast undertaking to develop data sets representing all biodiversity. The
EMEND project is one of the few examples worldwide that attempts to do this in the context of
natural resource management issues. The data for today’s exercise are a small subset from
EMEND, chosen for illustrative and pedagogical purposes. In most practical work on biodiversity,
we choose focal taxa. The focal taxon for today’s exercise is the Order Araneae, or the spiders.
They are a most useful exemplar for biodiversity.
The Alberta fauna includes c. 600 species of spiders, and about 300 of them may be
found on the EMEND landscape as inhabitants of the boreal mixedwood forest. The data for
today’s exercise include 20 species that represent reasonably well the ecological breadth of
spider species living on the ground in boreal forest (see Table 1). The abbreviations correspond
to those used in the EXCEL spreadsheet. You will note that these 20 species are distributed over
8 different families, all of which are characterized by different body plans and natural history
characteristics. Within families, the main differences among species relate to their use of
habitats and, for this reason, their diversity is related to the diversity of microhabitats provided
by the treatment x cover-type combinations. You will note that species names have two parts;

1
In what follows words in italics (except for species names) are a subset of technical words that you
should know and be able to use after RENR 110. Boldface is used to ensure that the item receives your
attention.
the entire binomen must always be used for formal designation of a species, although the
generic name may be abbreviated after first use in a document. The first part of the binomen is
the Genus (plural=genera) which is always capitalized, and it is followed by the species name (or
epithet) which is always in lower case. This is somewhat equivalent to your personal name,
although in the western world it is the reverse as our family names come last. In China,
however, the parallel is exact, as the more general last name comes first!! You will note that
the name of the species (including both Genus and species epithets) are always italicized or
underlined.

Table 1. Species included in the dataset for today’s lab exercise.

Abbreviation Species Family
Agelutah Agelenopsis utahana Agelenidae
Agroorna Agroeca ornata Liocranidae
Allodent Allomengea dentisetis Linyphiidae
Cybaeuop Cybaeopsis euopla Amaurobiidae
Gnapbore Gnaphosa borea Gnaphosidae
Gnapparv Gnaphosa parvula Gnaphosidae
Imprcomp Improphantes complicatus Linyphiidae
Leptalpi Lepthyphantes alpinus Linyphiidae
Maroampl Maro amplus Linyphiidae
Ozypsinc Ozyptila sincera canadensis Thomisidae
Pardfusc Pardosa fuscula Lycosidae
Pardhype Pardosa hyperborea Lycosidae
Pardmack Pardosa mackenziana Lycosidae
Pardmoes Pardosa moesta Lycosidae
Pardxera Pardosa xerampelina Lycosidae
Sciatrun Sciastes truncatus Linyphiidae
Sisimont Sisicotus montanus Linyphiidae
Walcdire Walckenaeria directa Linyphiidae
Xystcana Xysticus canadensis Thomisidae
Xystemer Xysticus emertoni Thomisidae

In order to do this sort of work, one must first be able to accurately name the species of
one’s focal taxon. Some believe that adequate biodiversity work may be done at higher
taxonomic levels, i.e., by identifying things only to genus or family. However, in natural
resource management this is risky business. Can you think of reasons why this should be so?
Making accurate identifications of invertebrate taxa is no simple matter, but with practice
individuals can become quite competent with a limited focal group of invertebrates or plants.
Of course, this is another reason why biodiversity studies tend to focus on particular groups.
See Appendix I for pictures of some of the spiders represented in this data set, and consider
what else you might want to know about them for full and complete analyses of biodiversity.
Under the CCFM criteria for sustainable forest management discussed in the block of
lectures about Forestry, natural resource extraction activities that leave a biota more or less
intact are judged to be acceptable, and efforts to improve forest management may be
evaluated partly in relation to whether the outcomes for spiders are more like those seen on
undisturbed landscapes with patterns driven by natural disturbance.

III. The Question: Does retention harvest better protect and conserve biodiversity? A
significant aspect of this exercise is to have you think through how the data at your disposal
could be analyzed and best interpreted to answer this question. During the lab part of this
exercise, we will analyze the data. Performance on this lab is assessed in the normal way
through the quiz available in eclass. To complete the quiz, you must first complete the lab
excerise.

IV. Analysis: No serious analysis should start without a general plan. Of course, in any serious
scientific investigation the plan is informed as the analysis unfolds and as the investigator
comes to better understand the data. In contrast, analyses done mainly to reach specified
management decisions through strictly predetermined decision structure. We will treat the
present exercise as a sort of hybrid; i.e., although few specific management alternatives are
associated from such analyses, we will approach this exercise through somewhat formalized
structure, so that students become familiar with the calculations and use of several biodiversity
measures.
Our initial overall objective will be to use information in the dataset provided for this
exercise to fill in a summary table organized as follows:

A. Calculating the Elements of the Summary Table

To begin you must log in with your CCID and password. Then save the
BiodiversityLab folder from eClass on the D drive of the machine you are using.
This folder contains all the files you will need during the lab.

Fill in your summary table (Calculations Sheet in EXCEL) by adding the following elements:

1. S(obs) / Observed Number of Species: This is the number of species captured in each
replicate of each cover-type x treatment combination. For example, by simply counting the
zeros in the first line of the data set (n=5), you know that 20-5=15 is the number of spider
species captured in the samples from this site. You make Excel do this for you quickly and
accurately by using the COUNTIF function to fill in the table for you; i.e., (Data!Cn:Vn,">0"). You
may fill in the correct numbers for S(obs) in your table by pasting the function with proper
arguments into column C in cells 2-19.

2. N / Abundance: This is the total number of individual spiders collected at each site over the
sampling period. It is a simple sum of the data for all 20 species at each site. You should by now
know how to fill in this column of the table using the SUM function in Excel.

3. Next we calculate the three integrative diversity measures that we studied in class: Simpson’s
Index (λ), the Shannon-Weiner Index (H’) and Evenness (J). In order to do this effectively we
must make the following preliminary calculations. It is most efficient to set these up as sub-
tables of the main data set because each calculation will place a value in each cell of a new
spreadsheet table with the same dimensions as the original data set.

a. In preparation …
• First, calculate the relative abundance of each species within each site, that is, the
proportion (pi) of each row accounted for by each species; i.e., the ‘cell value’/ ‘row
total’.
• Second, calculate pi2 for each cell in your matrix.
• And, finally, calculate the natural logarithm of the relative abundances calculated in the
first step, that is, Lnpi for each cell. Because Ln(0) is undefined, Excel returns an error
(#NUM) in each cell that carries a value of 0 previous to this last calculation. There are
several ways to remedy this, but since the data set is relatively small, it will be easiest
for you to simply replace #NUM in this final version of the preparatory calculations with
either 0 or a blank cell.

b. Now make the calculations by through adding formulas to the summary table of your
spreadsheet for each line …
• λ = ∑pi2
 • H’ = -∑ pi*Ln pi
• J = H’/LnS

4. Next, we estimate Sest (estimated species richness) using rarefaction. In order to do this you
will want you to use R. This is an open-source (and free!) statistical platform used widely for
work in biodiversity (and statistics in general). R has contributors from all over the world,
including some from University of Alberta, and is a most interesting development of
modern technology – a sort of worldwide web of statistics. R is becoming a standard for
statistical analyses in science. Becoming familiar with it early in your career is an
opportunity to prepare yourself for an easier time in the future. Proceed as follows:
• Open R (Windows Start Menu  All programs  R (folder)  R (either)); i.e., either
the 32 or 64 bit version will work.
• You must specify the working directory from which R will read the files. To do this go to
File  Change dir…, navigate to the D drive and select the BiodiversityLab folder.
• Next, you need to open the ‘script’ file (a list of commands for calculations in R are
called scripts) with the instructions (File  Open Script… and then, on the window that
opens double click on the file R_Script.R.
• Now you can simply follow the instructions in the R_Script.R file to produce your
rarefaction analysis and plot. You will need to calculate Sest for each treatment x cover-
type replication to fill in the summary table, but we will plot only the overall rarefaction
for each cover-type. Be sure that you understand whether this is an individual or
sample based rarefaction. The calculations are all pre-programmed for you, as our
purpose here is only to show you how to run R. Just follow the directions.

5. At this point you have all the data required to construct your summary table of means and
standard errors of the mean 2 (SEM) [e.g., to calculate the SEM for Aspen-Clear cut, use the
formula sqrt(var(C67:C69)/3)] for each treatment embedded in each of the two cover-types.
The table headings are laid out for you on the data sheet of the Excel file; you need only to
do calculations.

B. Produce a cluster analysis

2
The standard error of the mean is the standard deviation of the sample means over all possible samples (of a
given size) drawn from the focal population, and is a widely used measure of error for the average of a set of
samples. It is estimated as SE = s/√n, or the sample standard deviation divided by the square root of the sample
size used to calculate the mean (remember s=√variance). As explained during the first lecture block, two standard
errors on either side of the mean should include c. 95% of the means expected in repeated estimates from the
same population.
To produce a cluster analysis based on these and your calcualtions data, continue to follow the
instructions provided with the previous script. As we discussed in lecture, remember that this
analysis permits you to compare faunal composition among treatment x cover-type
combinations.

C. Produce bar graphs from your summary table

To finish the lab part of the exercise you must produce bar plots for each of the variables that
you have calculated (i.e., all of those in the summary table of means). Once you have the plots,
you need to add the error bars (SEM) to the bars representing each of the means. This will be
explained in class by your lab instructor.
Appendix I (pictures from www.bugguide.net, unless stated otherwise)

Agelenopsis utahana Agroeca ornata Cybaeopsis euopla

Gnaphosa borea Gnaphosa parvula Ozyptila sincera canadensis

Picture: J. Pinzon

Picture: J. Pinzon
Picture: J. Pinzon
Pardosa fuscula Pardosa hyperborea Pardosa mackenziana

Walckenaeria directa Xysticus canadensis Xysticus emertoni