You can observed hight corelation of prawns species Heterocarpus sp. and Aristeus o St: 8-4

There are some general rules to interpret ordination plots. Please check out the chapter 10 or 11 of canaco 5. Their explain pretty well the "rules". However major information is required on ecological atributes from your samples and from the species, that is up to you.  

PCA deal with the variance of the data, therefore the arrows. You should look on the direction of the arrows (species). That give the relation with 1) axis, 2) other species and 3) with samples. When the angle of the arrow with the 1,2 or 3 is less than 90° the asociaiton is positive, if parallel asociation is very strong. When is more than 90° asociation is negative. The closest the arrowhead with sample x(points), the stronger the relation. You must repeat this exercices for every arrow and point. Axis are the gradient that explain the distribution of your data. So PCA1 eplain x% of the ditrbution, while PCA2 explain x%, and so on.

to easier interpretation you might centralized (and standarized) the data. Some species with the high number of indivials tend to hide the patterns of the other species absorving most of the variability of the datas.

But again you need to know your samples and your species to interpret the plot.

Please go through these papers.

Here we used a PCoA -qualitative and quantitative traits were used together-. In the MM it is showed how the PCoA was conducte. 

In the second paper a RDA was obtained with CWM's trait data. See MM and results to check how was analyzed and explained 

Article Using the response-effect trait framework to disentangle the...

Article Management as a driver of functional patterns and alien spec...