Simulation of species turnover along a gradient

Download files

Notices

This simulation is written as VisualBasic-Macro for Microsoft Excel 2003. You can use it for free (see agreements). Once you have unzipped the file you will find an excel document and a text file with the instructions. Open the excel file and read the instructions. I included a short species list from my study area in Bolivia as example. I hope there are no problems (see warning) and you enjoy it!

Short Instruction (see zip-file for the long version)

1. Copy your observed species list on sheet "list". This list contains in
column 1: name of the species (without any cryptic symbols/ letters!)
column 2: notice (whatever you want)
column 3: ID of the interval along the gradient where the specie was found
(depending on your data this interval-ID may be elevational distance, i.e. every 100 m. Notice that you have only one record per specie per interval!)

2. Insert the original distribution of species on sheet "model" in the second line by typing for each interval (column) the number of species you have observed. You can use more than 20 columns. I never tried the maximum. But it should be limited by 250. Don't use the Excel-maximum (256) because the macro calculation needs a column beside this table!
The line "max number of sp." will be calculated automatically by the macro. Don't write anything in this line. Once started the macro, you can observe the progress with the counter in the cell below 'Permutation'. The macro will insert randomly all species in the cells below the 4th line. Each line symbolize the niche for one specie in an interval. The presence of a specie will be marked by "1" - "B" marks the beginning and "E" the end of the distribution. "x" signifies "no more niches in this interval available".

3. Don't change anything on the sheets "rangetab", "count" and "turnover"! When the simulation is terminated you can study this sheets. If you want to run the macro again on the same data, make now a copy. Never start the macro again when there are already changes on the sheet "rangetab".

4. You are now ready to start the macro!
(Notice that you have to activate the macros when you open the document! If Excel doesn't ask you when you open the document you have to change the security level in the Excel options.)
Choose extras\macro\macros... and start the item "turnover" from the list of all macros available (by double-clicking on it or clicking on "execute"). The macro will ask you some properties:

• name of the file
• model without singletons ?
  (This is the default. You can use also a list with singletons (type 'no') but you should know if this make sense!)
• the last line of your species list on sheet "list"
• multiplicator for the distribution patterns
  (The simulation almost never can simulate a distribution for all species along the gradient when you try to use identically the same number of species per interval than observed. I propose to use 150 % of the observed patterns (a multiplicator of 1.5). But depending on the data this may be changed. You should try different factors with only few permutations in the beginning.)
• number of intervals along the gradient
• interval with the highest species richness (count from the left!)
  (If the 'highest species richness' is meet in more than one interval take the first from left)
• the rangesize of species to start with
• the number of permutations of the simulation
  (depending on the properties of your computer this will strongly influence the time of the calculation)

5. Check the result.

• On sheet "model" you will find the distribution pattern of the last simulation.
• On sheet "count" the distribution for each permutation is listed ("G" = number of species per interval, "B"/"E" = number of species beginning/ending their distribution in this interval, "S" number of singletons in this interval if you have chosen a model with singletons). The column "error" indicates how many species couldn't enter into the simulation model. This means that other species occupied the niches and their rangesize was to large to find a place for it. If you have too many errors you should enlarge the multiplicator for the distribution patterns.
• On sheet "rangetab" you will find a list of all species with their rangesize calculated from your list on sheet "list". If their are species which couldn't enter into the simulation model they will be marked with "error" in the last column.
• On sheet "turnover" the WILSON-SHMIDA-index is calculated for each permutation in relation to the results of sheet "count". The column "error" means the same than on sheet "count".

6. I recommend to sort the tables by "error" and to consider only those with "error = 0". But you may have other objectives! Then calculate the means and standard deviation as usual.

Warning

I'm not a professional programmer! You can use this VB-macro for free but without any warranty! I checked all problems I meet with my datasets but there may be more problems I don't know!

Agreement

The use of this macro is free. You will need a licence for Microsoft Excel of course. By downloading and using the macro, you must agree not to distribute it in any commercial form. You are most welcome to use this simulation in any way you like for your own research, as long as such use is acknowledged. Please cite the method as: Bach, K., M. Kessler & S.R. Gradstein (2007): A simulation approach to determine statistical significance of species turnover peaks in a species-rich tropical cloud forest. Diversity and Distributions 13: 863–870. The code of the macro is not encrypted and you may improve it. Please inform me if so.