The manuscript has been greatly improved in clarity. However one of my comments has not been comprehended by the authors and ask me for a clear explanation. Perhaps I have not explained well. I adviced to make regression type 2 that is that when we not controled the independent variable and not have a predictive approach. In regrression type II the slope represents the minimum distance between cases and the "recta" without the necessity of a vertical minimum distance and the "a (the intercep)" factor have not sense alsoknown as major axis regression (RMA) that can be done in Package 'lmodel2' - CRAN. This is different from a tipe I where we controled the values of the dependent variable and want a predictive model and the minimum distance between cases and the "recta" is the minimum vertical distance.

In my opinion exceptuating this little constrain the manuscript is almost good enough to merit to be published in Ecol let

Despite this I encourage to make this type II regressions, The end results will be very probably similar but the proceed will be more consistent.