0
$\begingroup$

I want to study the impact of aquaculture on a hard rocky seafloor community which is naturally low in diversity.

4 replicate images (0.25 m2 each) were taken at fixed distances (0, 20, 40, 80, 120, 160 and 200 m) along 8 transects away from the aquaculture site.

The database has 15 different taxa, some of which are continuous (regular abundance data, from 0 to 16) and others are categorical:

0 = 0%
1 = 1-25% cover
2 = 25-50% cover
3 = 50-75% cover
4 = 75-100% cover

There’s also lots of zero in the database.

I want to know if the benthic community changes based on the distance away from the aquaculture site.

My 1st approach was to transform my species database to presence/absence. Results are interesting but presence/absence is a drastic transformation and I would like to know if other options are available.

My questions are:

1. When performing a single multivariate ANOVA (aka PERMANOVA) and dealing with a species database containing categorical and continuous data, is there other transformation possible than presence/absence?

2. Would a standardization be appropriate?

$\endgroup$
1
$\begingroup$

For the record, after discussions with other colleagues, I believe since my species in categories (from 0 = absence to 4 = 75-100% cover) are somewhat already transformed, the best way will be square root transform my species abundances (already fairly low). This way all species (either in category or abundance) fall between 0 and 4 and have the same potential range and weight across the matrix.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.