0

How can I test the differences between species richness (table below) using means (S.obs) and standard deviation (se.obs) with vegan or any other package in R?

 "Group.1"                "S.obs"   "se.obs"
 "Cliona celata complex"  499.7143  59.32867 
 "Cliona viridis"         285.5000  51.68736 
 "Dysidea fragilis"       358.6667  61.03096 
 "Phorbas fictitius"      525.9167  24.66763

Thank you in advance! André

André Soares
  • 309
  • 1
  • 13

2 Answers2

0

From the technical point of view, this looks very much like t-test with unequal variances. Check the formula, and plug in your data. The R function t.test() expects raw data, but if you already have means and se's, it is easy to calculate statistics by hand.

I have no idea how you obtained your numbers and therefore I cannot comment on the scientific point of view.

Jari Oksanen
  • 3,287
  • 1
  • 11
  • 15
0

Actually, being that it was outputted by vegan, I fount out it can be ANOVA tested in the following way:

 richness.aov <- aov(specnumber(dataset))
 summary(richness.aov)
 #Tukey multiple comparisons of means
 #95% family-wise confidence level
 TukeyHSD(richness.aov)

Thank you all anyhow!

André Soares
  • 309
  • 1
  • 13
  • You didn't tell us that it was output from vegan -- and it wasn't. No vegan function outputs items `"S.obs"` and `"se.obs"`. The `specnumber()` function certainly produces data you can analyse with `aov()`, but not the data you displayed in your original query. – Jari Oksanen Jun 04 '15 at 09:30
  • Indeed you are right Dr. Oksanen, that was derived from an output of estimateR. Sorry for the lapse, I am a begginer but my intentions are good :) – André Soares Jun 04 '15 at 09:43