[R] How to conduct a PERMANOVA using a dissimilarity matrix

Wed Dec 23 11:59:44 CET 2015

Hi!

R indeed cannot distinguish between your Site and Species columns. You 
either need to state "row.names=1" while loading your csv-files into 
workspace or you can rearrange your dataframe like this:

|bats2 <-bats[,-1]rownames(bats2)<-bats[,1] Alternatively, you can tell R 
which columns of your dataframes should be excluded in the adonis call. 
About your other questions: You may want to look into ?vegdist to see 
the different methods adonis() accepts. I think Soerensen is the 
presence/absence version of Bray-Curtis. Jaccard is another simple index 
for occurence data, and can be computed by vegdist. Since you are doing 
post hoc hypothesis testing, you might want to state a number of 
permutations within the adonis call. I think (you should update yourself 
on this) that the permutations also removes possible autocorrelation 
from the dataset, which would be possibly convenient in your study 
question. For starters, i suggest reading the many vegan tutorials 
available or - if you can get your hands on it - the Numerical Ecology 
textbook by Legendre & Legendre. I did so, and it served me quite well. 
Also, tutorials on basic data handling and navigation in R would come in 
quite handily, i guess. Cheers|,

On 23.12.2015 10:07, michael.eisenring at agroscope.admin.ch wrote:
> Dear Erica and Hazel,
> Thank you very much for your help.
> Based on Ericas feedback I use now a the raw data giving me information on abundance of bat species (I changed the data set) that were caught at 6 different sites (file named bats)
>
> Thanks to Hazels input I use now an additional file (named forest_type) assigning a forest type (O or Y) to each site.
> Using Adonis I would like to convert my bats in a first step to a soerenson (dis)similarity matrix and then use the Adonis function to see how bat communities differ between forest types.
>
> With the help of you I was now able to write some code
> The code seems to work. However, I have 2 more (basic) questions.
>
> 1) Do I have to convert my data sets in a specific way (I structured it as in the R help example using the dune data)? I have the problem that R thinks the first row (the very left one) with the header "Site" is just another species of bat ( if you open the bat file in R and scroll to the top you see what I mean, "Site" is next to "Kerivoula...." and it is treated just like another bat species). How can I solve the problem?
>
> 2)Is the code correct? I am especially unsure if my conversion to a dissimilarity matrix worked, since the results are exactly the same , regardless if I add method="bray" or not to my code.
>
> Below is the code and the 2 data sets (dput)
>
>
>
> #load data sets
> #file with abundance information of each caught bat at each site
> bats<-read.csv("bats_abundance.csv",sep=";",header=T)
> #file with forest type (O,Y) information
> forest_type<-read.csv("forest_type_6.csv",sep=";")
> attach(bats)
> attach(forest_type)
> bats
> forest_type
>
> #Aim: compare how bat communities differ between O and Y forests
> #How can I convert bats first to soerenson dissimilarity matrix ()
> adonis(bats~forest, data=forest_type,method="bray")
>
>
>> dput(bats)
> structure(c(1L, 2L, 3L, 4L, 5L, 6L, 1L, 1L, 1L, 1L, 1L, 1L, 0L,
> 0L, 1L, 0L, 0L, 1L, 0L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L,
> 1L, 1L, 1L, 1L, 1L, 1L, 1L, 0L, 0L, 1L, 1L, 1L, 0L, 0L, 0L, 0L,
> 0L, 1L, 0L, 1L, 1L, 0L, 1L, 1L, 1L, 0L, 0L, 0L, 0L, 0L, 1L, 0L,
> 0L, 1L, 1L, 0L, 1L, 0L, 0L, 0L, 0L, 1L, 0L, 0L, 0L, 0L, 1L, 0L,
> 0L, 0L, 1L, 0L, 0L, 0L, 0L), .Dim = c(6L, 14L), .Dimnames = list(
>      NULL, c("Site", "Kerivoula_smithii", "Scotophilus_nigrita",
>      "Minipoterus_natalensis", "Miniopterus_fraterculus", "Neoromicia_nana",
>      "Otomops_martiensseni", "Chaerephon_pumilus", "Rhinolophus_clivosus",
>      "Hipposideros_caffer", "Lissonycteris_angolensis", "Pipistrellus_hesperidus",
>      "Pipistrellus_nanulus", "Rhinolophus_fumigatus")))
>
>
>> dput(forest_type)
> structure(list(Site = 1:6, forest = structure(c(1L, 2L, 2L, 2L,
> 1L, 1L), .Label = c("O", "Y"), class = "factor")), .Names = c("Site",
> "forest"), class = "data.frame", row.names = c(NA, -6L))
>
>
>
>
>
> Thanks a lot,
> Mike
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
> Eisenring Michael, Msc.
> PhD Student
>
> Federal Department of Economic Affairs, Education and Research
> EAER
> Institute of Sustainability Sciences ISS
> Biosafety
>
> Reckenholzstrasse 191, CH-8046 Zürich
> Tel. +41 44 37 77181
> Fax +41 44 37 77201
> michael.eisenring at agroscope.admin.ch
> www.agroscope.ch
>
> -----Ursprüngliche Nachricht-----
> Von: Hazel Knipe [mailto:hazel_knipe at hotmail.co.uk]
> Gesendet: Dienstag, 22. Dezember 2015 18:34
> An: Eisenring Michael Agroscope <michael.eisenring at agroscope.admin.ch>
> Betreff: Re: How to conduct a PERMANOVA using a dissimilarity matrix
>
> Dear Michael,
>
> You need to create another file dividing the different categories/ variables (make sure that they are in the same order for the variable you are testing), then attach it and run the adonis function. For example:
>
> attach(metadata)
>
> adonis(formula = vegdist(matrixfile, method = "bray") ~ Site, data = metadata)
>
> It's also important to think about if you have nestedness in your design, in which case you would need to add strata = factor:
>
> adonis(formula = vegdist(matrixfile, method = "bray") ~ Species, data = metadata,      strata = Site)
>
> I hope this helps for now! Let me know if you need more detail.
>
> Hazel (BSc student).
>
>
>
> ________________________________________
> From: R-help <r-help-bounces at r-project.org> on behalf of michael.eisenring at agroscope.admin.ch <michael.eisenring at agroscope.admin.ch>
> Sent: 22 December 2015 16:18
> To: r-help at r-project.org
> Subject: [R] How to conduct a PERMANOVA using a dissimilarity matrix
>
> Dear R-List members,
>
> I have to compare how similar two types of forest (old growth=O) and (young forest=Y) in terms of moth communities are.
> I sampled moths at 4 O and 4 Y sites.
> I need to analyse the data using a PERMANOVA approach. But I am having a really hard time to do this in R.
>
> I found out that I need to create a dissimilarity matrix and read this matrix then into R to conduct a one-way Permanova with forest type (O or Y) as factor.
> The package vegan with the function "adonis" seems to be able to do a permanova.
>
> I created the matrix (based on Soerenson (dis)similarities) and imported it into R.
>
> Could anyone help me with the next step? How can I conduct a permanova on my dataset? In the end I would need an R value and significance level telling me if community compositions differ significantly between sites.
>
> Below is my code (not too much) and the data for the matrix.
>
> #dput for matrix:
>
> structure(c("", "O", "Y", "Y", "Y", "O", "O", "Y", "O", "O", "0", "0.544", "0.519", "0.533", "0.481", "0.548", "0.518", "0.479", "Y", "0.544", "0", "0.383", "0.416", "0.383", "0.358", "0.434", "0.399", "Y", "0.519", "0.383", "0", "0.398", "0.359", "0.392", "0.401", "0.374", "Y", "0.533", "0.416", "0.398", "0", "0.398", "0.399", "0.358", "0.348", "O", "0.481", "0.383", "0.359", "0.398", "0", "0.37", "0.317", "0.354", "O", "0.548", "0.358", "0.392", "0.399", "0.37", "0", "0.39", "0.365", "Y", "0.518", "0.434", "0.401", "0.358", "0.317", "0.39", "0", "0.371", "O", "0.479", "0.399", "0.374", "0.348", "0.354", "0.365", "0.371", "0"), .Dim = c(9L, 9L), .Dimnames = list(NULL, c("V1", "V2", "V3", "V4", "V5", "V6", "V7", "V8", "V9")))
>
>
> #Code
> #load dissimilarity matrix (based on Soerenson similarity) moth_dta<-read.csv("Geo_sorenson_8.csv",header=T,sep=";")#Creates matrix from imported data
> moth_dta<-as.matrix(moth_dta)
> moth_dta
> library(vegan)
>
>
> Thank you very much,
> Michael
>
> Eisenring Michael, Msc.
> PhD Student
>
> Federal Department of Economic Affairs, Education and Research EAER Institute of Sustainability Sciences ISS Biosafety
>
> Reckenholzstrasse 191, CH-8046 Z rich
> Tel. +41 44 37 77181
> Fax +41 44 37 77201
> michael.eisenring at agroscope.admin.ch<mailto:michael.eisenring at agroscope.admin.ch>
> www.agroscope.ch<http://www.agroscope.ch/>
>
>
>          [[alternative HTML version deleted]]
>
> ______________________________________________
> R-help at r-project.org mailing list -- To UNSUBSCRIBE and more, see
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.

-- 
Tim Richter-Heitmann (M.Sc.)
PhD Candidate

International Max-Planck Research School for Marine Microbiology
University of Bremen
Microbial Ecophysiology Group (AG Friedrich)
FB02 - Biologie/Chemie
Leobener Straße (NW2 A2130)
D-28359 Bremen
Tel.: 0049(0)421 218-63062
Fax: 0049(0)421 218-63069