CANONICAL CORRESPONDANCE ANALYSIS
Dear friends, I would hereby like to inform you that i have a two year data set (monthly observations as data points) consisting of abundance (log transformed) 40 species of phytoplankton and 15 environmental variables.
My objective is to identify the seasonal patterns in environmental parms and species abundance.
I have tried the canonical correspondence analysis for analyzing this data to identify the linear combination of species scores with environmental variables. But, the results are bit complex and i need your suggestions for analysis (better ordination techniques if any) and the modifications in the same analysis (data as well as analytical steps).
I found CANCORSP, REDUNDANT, PCA, CORRESP analysis as the best methods for ordination. Please help me in this regard
I used detrended correspondence analysis with PC-ORD (https://www.pcord.com) for Multivariate Analysis of Ecological Data.
Multivariate analysis is the place to end, not the place to start. If you want to understand your data, then you first have to do some exploratory data analysis... and given the number of variables you have measured and the amount of data that you have, this will take a considerable amount of time, effort and creativity.
my suggestion is:
the fact the quality data is the main selected way.
-recognize normality and homogeneity with the analyses method such as: Shapiro & Levant test
- separated the hydrological condition in two years
- select the multivariate type( linear method,weighted averaging)( leps and smilauer,..multivariate analysis of ecological data using CANOCO). and consider the station and hydrological condition as factor in analyses.
best regards
my suggestion is non multidimensional scaling nMDS. You could use the package PRIMER 6.
- Badges
- Science method
- Similar topics
- Biological Science
- Ecology