Site x species matrix and enviromental Variables. Which tests should I use

#1
Hello Talkstats Forum,

I am currently doing my study with the theme: Effects of farming type and landscape density on carabid diversity and community structure in vineyards.

My question is which tests should I use.

In my research I have 18 vineyards.
The vineyards are spreaded over three areas. Each area is classified in three groups of landscape density (low, medium, high). In each landscape density there is one organic and one conventional vineyard.
In every vineyard I collected carabids with pitfall traps.

From my university I got the CANOCO 5 program, I also have the R program on my computer.

How should I start with the statistic?

First I did a simple comparsion about the area, landscape density and the difference between organic and conventional farming.

Next step was a correaltion between species richness and landscape density as well correlation about abundance and landscape density.

As next I did a DCA and CCA. But I am not sure if I did it right, maybe somewone can help me with the data preperation.

Which other tests should I use, to get information out of my data.

I am thankfull to any help. :)

Thank you in advance.

Best regards

Max
 
#2
Hi,
what is your outcome variable? A diversity index like the Shannon-Index? The most straight foward approach would be probably a mixed regression model in R (e.g., using the function lme() from the package nlme), where your carabid density is the outcome variable, the vineyard-ID is a random factor and farming type and landscape density are fixed effect predictors. The regression coefficients and corresponding p-values tell you how much faring type and landscape density influence carabid density