Statistics for Ecologists – Example Data

Click a name to go to a description of the data. The data are contained in two archives, one for the R data, S4E2e.RData and one for spreadsheet data, S4E2e Archive Excel.zip. Note that some items may be listed in more than one section. Graphics are not mentioned as a topic because all the data can be shown graphically one way or another! Similarly, many of the data can be used for practice at manipulating data in Excel and/or R in some way (see Miscellaneous).

Summary Statistics | Data Distribution | Differences (2 samples) | Correlations (2 variables)

Associations | Differences (>2 samples) | Regression | Diversity | Similarity | Miscellaneous

Archives – Instructions (& download)

• Beetle sizes
• Butterflies and Habitat
• Flour beetles
• Hoglouse abundance
• Leaf sizes
• Mayfly (correlation)
• Mayfly (regression)
• Pearson correlation data
• Ridge & Furrow
• Sward height
• Plant species & watering

• Beetle sizes
• Butterflies and Habitat
• Flour beetles
• Hoglouse abundance
• Leaf sizes
• Mayfly (correlation)
• Mayfly (regression)
• Pearson correlation data
• Ridge & Furrow
• Sward height
• Plant species & watering

• Beetle sizes
• Flour beetles
• Hoglouse abundance
• Leaf sizes
• Ridge & Furrow
• Sward height
• Whitefly – matched pairs
• Wilcoxon online exercise – matched pairs

• Butterfly Food – there are three variables but you can practice using two at a time
• Bluebell and Light – polynomial regression
• Freshwater (correlation)
• Growth (plant growth) – logarithmic correlation/regression
• Mayfly (correlation)
• Mayfly (regression)
• Pearson correlation data

• Birds and Habitat
• Heather species
• Invertebrates and Habitat
• Pea genetics – Goodness of Fit test

• Butterflies and Habitat
• Hoglouse abundance
• Leaf sizes
• Sward height
• Plant species & watering – (2-way ANOVA)

• Beach hoppers – logistic regression
• Butterfly Food – multiple linear regression
• Bluebell and Light – polynomial regression (curvilinear regression)
• Growth (plant growth) – logarithmic regression (curvilinear regression)
• Mayfly (regression) – multiple linear regression
• Newt presence-absence – multiple logistic regression

• Ants and Fire
• Butterflies and year
• Diversity Simpson D.xls – calculates the Simpson’s D index
• Diversity Shannon.xls – calculates the Shannon index
• Freshwater invertebrates.xlsx – a simple dataset with taxonomic information and quantity
• Hornbill diet
• Mosses and trees
• Plant species lists
• Plant species abundance

• Ants and Fire
• Butterflies and year
• Hornbill diet
• Mosses and trees
• Plant species lists
• Plant species abundance

• Butterflies and Habitat – Pivot Tables and rearranging/managing data
• Birds and Habitat – Pivot Tables and rearranging/managing data
• Dominance/Abundance scales – Using lookup tables (=VLOOKUP in Excel)
• Plant species lists – Pivot Tables and rearranging/managing data
• Seashore seaweed – Using lookup tables (=VLOOKUP in Excel)

• RData – this file contains all the data in R-format.
• S4E2e Archive Excel.zip – this file contains all the Excel format files (including CSV).

You can use the RData file in several ways:

• Open R then use load(file.choose()) and select the RData file (in Linux you should use the filename (in quotes) explicitly, including the path).
• Double-click the file. If R is already open it will add the data to your workspace, if R is not open it will open and the workspace will contain only these data (and the working directory will be set to wherever the RData file was stored.
• Drag the file to the R icon. The behaviour is the same as above.

Arranged more or less alphabetically

These data are adapted from Hoffmann, B.D. 2003. Austral Ecol. 28, p.182 and show the abundance of 91 species of ant in 10 samples. The samples are from two types of soil (red and black) and from 5 fire regimes. The data are arranged with the samples as columns, the column names indicate the soil and regime as follows:

• r = red soil
• b = black soil
• E2 = burnt every 3yr with grazing early (May)
• E3 = burnt, spelled & burnt in 2 successive yr
• L2 = burnt every 3yr with late grazing (Oct)
• L3 = burnt, spelled , burnt in 2 successive yr
• U = unburnt control

The data are used for dissimilarity calculations (including visualisation of dissimilarity with a dendrogram) but you can also use them to explore diversity. The data are in the S4E2e.RData archive and are named ant.

These data give the sizes (in mm) of a species of water beetle. The main sample can be used for data summary; a second sample is available (in R) to use for comparisons.

• The beetles.xls file in the S4E2e Archive Excel.zip archive contains the main sample as well as examples of histograms.
• The RData file contains the main sample as an object called bd, there is a second sample Mar (there are also copies sunny, shady).

The data can be used for data summary, such as mean, median, standard deviation and so on. You can also practice drawing histograms. The two samples can also be compared with the t-test or U-test.

These data show the allele frequencies at the mannose-6-phosphate isomerase (Mpi) locus in the amphipod crustacean Megalorchestia californiana, Californian beach hopper. Data from McDonald, J.H. 1985 (Heredity 54: 359–366).

• Beach hopper allele.csv – this file is in the S4E2e Archive Excel.zip archive and will open in a spreadsheet. You can use this for practice at importing data into R but see below…
• The RData file contains the data as an object called cbh.

These data are used to demonstrate logistic regression. Each row of the data gives the latitude and the number of specimens that had each form of the allele. There are two forms, so the data are binary, which is why a logistic regression is the appropriate method of analysis. Logistic regression is a form of generalized linear modelling (GLM).

These data show the abundance (as a count) of six butterfly species over five years at a site in Scotland. The data are arranged with the columns giving the year of the sample, each row gives the abundance of a species.

• Butterfly table – the data are in the S4E2e Archive Excel.zip archive as an XLSX file and a CSV. You can read the CSV file into R, in which case you need to add names = FALSE to the read.csv() command.
• The RData file contains the data as two objects: bf is a matrix and butterfly is a data.frame.

You can use the data for graphical summary, showing line plots of abundance and time for example, as well as bar charts and pie charts. You can also look at the diversity of the samples as a bit of practice with diversity indices, such as Shannon and Simpson’s. You could also explore similarity between years.

Although perhaps not the most sensible types of analysis you might also use the data for comparison of differences or changes with time (as a correlation).

These data show the abundance of butterflies in three habitats. Each habitat was sampled several times. The datafile has three columns, for the abundance, habitat and an index variable (the replicate).

• xlsx – the data are in the S4E2e Archive Excel.zip archive as an Excel format file.

These data are used as an example of rearranging and managing data using a Pivot Table in Excel. You can also use the data to look at data summary, graphics and differences (there are 3 samples). To do that using R you’ll need to save a copy as a CSV file then import into R.

These data show the abundance of butterflies and the availability of food plants and nectar resources.

• csv – the data are in the S4E2e Archive Excel.zip archive and will open in Excel.
• The RData file contains the data as an object (data.frame) called bff.

These data can be used to look at (multiple) regression (or correlation), and associated statistics (such as beta coefficients). You can also use the data for some graphical summary, such as scatter plots.

These data show the abundance of some common UK bird species in various habitats.

• xlsx – the data are in the S4E2e Archive Excel.zip archive as an Excel format file. These data are in recording format and you can use these as practice at using a Pivot Table.
• csv – these data are also in the archive and contain the data in the form of a contingency table.
• xlsx – this datafile is in Excel format and shows the data in a contingency table. There is also a completed association analysis in another worksheet.
• The RData file contains the data (contingency table layout) as two objects: birds is a matrix and bird is a data.frame.

The main purpose of these data is to look at tests of association (the Chi squared test). You can also use them for graphical summary, using bar charts and pie charts. The birds.xlsx file can also be used for translating data from recording layout into a contingency table (in Excel using a Pivot table, in R you can use the xtabs() command).

These data show the abundance of bluebell in a wood in England. Data are presented showing the abundance of the plant and the light intensity at the growing site.

• Bluebell polynomial.xlsx – the data are in the S4E2e Archive Excel.zip archive and will open in Excel.
• The RData file contains the data as an object called bbel.

These data show an interesting relationship between abundance and light, an inverted U shape. This lends itself to a regression using a polynomial equation. You can also use the data for graphical summary, such as a scatter plot and line of best fit (a trendline).

This dataset is used in an exercise on using lookup tables. It is useful to be able to convert from an ordinal scale that uses text values to an ordinal scale with “real” numbers.

• xlsx – the data are in the S4E2e Archive Excel.zip archive as an Excel format file.

Use this to help practice using the =LOOKUP function in Excel. This allows you to replace one value with another. In this case an abundance as a text label (D = dominant, A = abundant etc.) can be replaced with a numerical value. This allows you to carry out non-parametric statistics (e.g. the U-test). You are essentially replacing a text-based ordinal scale with a number-based ordinal scale.

These two files show you how to calculate two indices of diversity; both are in the S4E2e Archive Excel.zip archive.

• Diversity Simpson D.xls – as the name suggests, this calculates the Simpson’s D index of diversity.
• Diversity Shannon.xls – this spreadsheet computes the Shannon index (also called Shannon-Wiener or Shannon-Weaver).

These files can be used as the basis for a spreadsheet calculator (you can add extra rows as you need), which you can use to help compute the two commonly used indices of diversity.

These data show the abundance of flour beetles in samples taken from two different (fictitious) farms. The Excel version shows the data in sample layout, with one column for each sample. There are several R versions, with the data in different layouts:

• flour beetles.xls – the data are in theS4E2e Archive Excel.zip One column shows the counts of beetles from Woad Farm, the other column shows the counts from Glebe Farm.
• The RData file contains the data as four objects:
  • flour1 – a data.frame with a column qty and a column site, i.e. in recording layout
  • flour2 – a data.frame with two columns, Woad.Fm and Glebe.Fm, each containing the counts from a separate farm
  • Woad.Fm – a vector of values representing the counts of beetles at Woad farm
  • Glebe.Fm – a vector of values representing counts of beetles at Glebe farm

These data can be used for exploring differences between samples. You can also use them for graphical summary, e.g. bar charts, box-whisker plots, and for data summary (e.g. mean, median, standard error). The R-format data are in several forms so that you can practice carrying out commands on the different type of object.

These data show the abundance of a freshwater invertebrate and the water speed at the point of collection. See also the Mayfly (correlation) data, which are very similar.

• freshwater correlation.xlsx – the data are in the S4E2e Archive Excel.zip There are two columns, Abund and Speed for Abundance and water Speed respectively.
• The RData file contains the data as a data.frame called fw.

These data can be used for exploration of correlation as well as some graphical summary (e.g. scatter plot).

These data give the abundance of some freshwater invertebrates from Goredale Beck in Yorkshire. There is also some taxonomic information for each invertebrate recorded.

• Freshwater invertebrates.xlsx – the data are in the S4E2e Archive Excel.zip There is a column for the count of each taxa. Other columns give the taxonomic information (e.g. phylum, order).

You can use these data for looking at diversity. You can also practice transferring the data from Excel into R.

These data show the growth of a plant species in response to different levels of a nutrient.

• Growth Logarithmic.xlsx – the data are in the S4E2e Archive Excel.zip There are two columns Growth and Nutrient.
• The RData file contains the data as a data.frame called pg.

These data show an interesting relationship. If you plot one variable against the other you’ll see that the points “curve”. In fact the relationship is a logarithmic one. You can use these data to look at curvilinear regression, in this case logarithmic regression. This is ordinary linear regression but with a logarithmic equation.

You can also use the data to look at graphical summary, e.g. a scatter plot with line of best fit (trendline).

This dataset shows the abundance of two species of heather in Cornwall. The data are in the form of a contingency table. In total 137 quadrats were used and the presence of each species noted. The contingency table shows the frequency of occurrence; thus you have four options:

• Both species present together in a quadrat
• Calluna vulgaris present only
• Erica cinerea present only
• Neither species present (i.e. both absent)

The data can be used to explore the association between the two species.

• csv – the data are in the S4E2e Archive Excel.zip archive and will open in a spreadsheet.
• The RData file contains the data as a data.frame called heather.

You can use the data for tests of association (e.g. the Chi squared test), and since this is a 2×2 contingency you can apply Yates’ correction. You can also use the data for graphical summary (e.g. bar chart, pie chart). You can also use the CSV file to practice transferring data from spreadsheet to R.

These data show the abundance of hoglouse (Asellus spp), a freshwater invertebrate, at three sampling locations.

• xlsx – the data are in the S4E2e Archive Excel.zip archive. These data are in sample format; there is a column of abundance data for each of the three sampling locations. The spreadsheet also contains a second worksheet giving the summary statistics and a bar chart with error bars.
• The RData file contains the data as two data.frame objects:
  • hog2 – gives the data in recording layout, there is a column for count and a column for site.
  • hog3 – gives the data in sample layout, there is a column for each sample.

You can use the data for exploring differences between samples. In the book text you use these data for a Kruskal-Wallis test, which is a non-parametric test of differences between more than two samples. You can also use the data for practice at data summary and graphics (these data are used to draw bar charts in the book text). You could also subset the data and look at comparing just two samples.

These data show the presence of different fruits in the diet of three species of hornbill from India (data adapted from Datta, A. & Rawat, G.S. 2003. Biotropica 35, p.208). The data are in the form of presence-absence, so if a fruit species was found in the diet a 1 is recorded, if the fruit was absent it is shown as 0.

• csv – the data are in the S4E2e Archive Excel.zip archive and will open in a spreadsheet. There is a column for the species names of the fruit (as an abbreviated scientific name). Each of the next columns shows the presence-absence of these fruits in the diet of three species of hornbill:
  • GH = great hornbill
  • WH = wreathed hornbill
  • OPH = oriental pied hornbill
• The RData file contains the data as a data.frame, hornbill. The rownames have been set as the fruit species and the main data are the three columns of fruit presence-absence for the hornbill species.

You can use these data to look at similarity (dissimilarity) and to draw a simple dendrogram to show the relationship between the samples in terms of the presence of fruit species. You can also use the data for diversity (species richness).

These data show the frequency of observation of some terrestrial invertebrate taxa on different parts of plants. There are two datasets, which are similar. Both give the frequency of observation in the form of a contingency table.

• invert habitat.csv – the data are in the S4E2e Archive Excel.zip archive and will open in a spreadsheet. The data are in the form of a contingency table. The first column gives five invertebrate taxa and there are three three sites (Upper, Lower, Stem).
• csv – the data are in the S4E2e Archive Excel.zip archive and will open in a spreadsheet. The data are in the form of a contingency table. The first column gives the names of four sites (Upper leaf, Lower leaf, Stem, Bud) and the subsequent columns show frequencies for three invertebrate taxa.
• The RData file contains the data from “invert habitat.csv” as a data.frame called inv.hab.

You can use these data for tests of association (e.g. Chi squared) and for data summary (e.g. bar charts, pie charts). Note that the CSV dataset invert.csv is not duplicated in R, which means you can practice importing CSV data into R.

Leaf sizes

Mayfly (correlation)

Mayfly (regression)

Mosses and trees

Newt presence-absence

Pearson correlation data

Pea genetics

Plant species lists

Plant species & watering

Plant species abundance

Ridge & Furrow meadow

Seashore seaweed

Sward height

Tree sizes and month

Whitefly

Wilcoxon online exercise