Dear Anatoliy

Your question is rather more complex that appears firstly. And brings new questions, as Nancy noticed. What kind of habitat do you study? Forest, Grassland, Peatbog? What do you expect from measuring environmental conditions? Understanding plant distribution? What is for you environment? Soil parameters, climate, landscape, all of these? And more important, do you expect the present communities to be close to equilibrium to the environmental conditions, or are your habitats likely to change dramatically, due to climate , land use change(s) etc. The questions you put in the research is of course of great importance to determine what kind of data you need...

Hi Anatoliy:

Q: What is the location(s) of your study?

Dear Anatoliy

Your question is rather more complex that appears firstly. And brings new questions, as Nancy noticed. What kind of habitat do you study? Forest, Grassland, Peatbog? What do you expect from measuring environmental conditions? Understanding plant distribution? What is for you environment? Soil parameters, climate, landscape, all of these? And more important, do you expect the present communities to be close to equilibrium to the environmental conditions, or are your habitats likely to change dramatically, due to climate , land use change(s) etc. The questions you put in the research is of course of great importance to determine what kind of data you need...

Good day! Thank you very much for the help and advice.

I have interest in investigation as Grassland as well as forest habitats. More helpful conditions for me are light, moisture and temperature. This parameters need for investigation of environmental conditions for population-based studies.

As said by Didier the first step is using a contingency table of past and present, where it is possible to have archives. Unless sites are undisturbed a main antropic factor would be present. Land use and frequency of species depends on activity AND on environment. Soil loss and pollution can create sensible concern. ADE-4 package of Lyon-1 has got a collection of procedures and references using a free R platform, but it is just one of several possible start up

Guido -  Would the R package you mentioned include multiple regression?  -  What techniques are you suggesting?  - Thanks - Jim 

Under the key R project multilinear or multiple and your field of activity you will find packages for specific apps. The basic R has got lm and glm funtions for multiple regressions. Please see 

http://web.stanford.edu/group/ssds/cgi-bin/drupal/files/Guides/Guide%20to%20R%20Packages_2.pdf  for a start up: in my experience I did use mainly lme4 of D. Bates but this choice may be unfit to general questions

Thank you very much for help. R package proved to be very helpful to me.