diatomdaniel
commented
2 years ago - Yes...for now. Steve will have recommendations/criticisms but this is probably the easiest solution. However, if we are arguing that lakes are functionally different from rivers, using the same trophic classification scheme remains counter intuitive. I am unsure how to best resolve this issue; Uche said he will aggregate land use data for the study sites. Perhaps land use is ultimately a better classification scheme for our analysis. Until then, I recommend applying the same trophic state calculations to lakes and rivers as it is straightforward and we have the data.
- Hmmmm....perhaps only use sites that have TP data? If trophic state varies between sampling dates and trophic state is a control on processes operating in these systems (as we are arguing), we should take this into account. This is especially true as we currently are treating each observation as independent (AFAIK); ultimately we really need to come to a consensus on how to best split up the data and which variables to keep.
- An alternative approach is to go with what the data tells us.... instead of categorising our data based on trophic state, etc. we do a multivariate analysis of DOC, NO3 and PO4 and split the data into low, intermediate, high etc. categories. If we can see patterns in the data, perhaps we should act on them.
LinneaRock
I ran into a couple problems with calculating trophic states and I'm interested in people's thoughts on how to proceed.
I can only get EPA thresholds for lakes/reservoirs (using TP concentration). I believe they do not calculate eutrophication for rivers. So I could just use the same thresholds for rivers as for lakes. Let me know what you think about this.
Many of our sites were sampled multiple dates and the trophic status can change over those dates. However, if TP was not collected, we will not have a trophic state for that date. Would it be better if I used an average or median for each site that has TP data (~90% of sites)?