Can any one advise me on which ordination methods is most fit for analyzing diversity with the environment. please share few useful paper or links available with you.
I recommend this work that we do with riverside population in an area of environmental protection in the Brazilian Amazon, available for download at the link below.
Chapter Índice de Qualidade Socio Ambiental (IQSA): Parâmetros da pe...
I prefer varimax-factor, multiple-regression, & MANOVA methods for ecohydrologic & other biophysical analyses of watershed ecosystems & their biodiversity (see my papers at https://www.researchgate.net/profile/Robert_Vadas).
It truly depends on the specific question that you are considering. There are many different options depending on the context of your question. Investigating each ordination option, their assumptions, and data requirements is critical and the "fun" part of working with community data. To be brief, if you are interested in how biodiversity is affected by environmental metrics that you have data for, I would suggest Direct Gradient Analysis (DGA). There are methods of DGA that don't require an ordination, but if you do have to do an ordination you should look into Redundancy Analysis (RDA) or Canonical Correspondance Analysis (CCA). CCA's are generally better for ecological datasets.