fmsabatini
commented
1 year ago The prolem seems wider, and might involve more plots from SALVIAS. It seems it was created when importing the data received from SALVIAS into TV2. Probably a mismatch between species in plots and the species list
Evan F. brought to my attention that some plots from SALVIAS in Canada are composed by tropical species.
Minimal reproducible example:
library(tidyverse)load("./data/sPlotOpen/sPlotOpen.RData")canada_plots <- filter(header.oa, Country == "Canada" & Dataset == "Salvias")DT2.oa %>% filter(PlotObservationID == canada_plots$PlotObservationID[1])I checked and this is what I found:
1) The weird plots Evan identified in sPlotOpen were already present in sPlot 2.1 2) They are still present in sPlot 3.0 3) My feeling is that these are legit plots, with wrong spatial coordinates. 4) The mistake seems to occur also in the data we originally received from SALVIAS (at least in the excel file edited by the former coordinator JD and forwarded to SH in 2015(?) - see: "DATA_Data\TurboVeg2\Turboveg_sPlot\Data\Salvias\Excel") 5) I'm unable to assess whether other plots from the SALVIAS database suffer from the same problem
These are the plots IDs 56397 56398 57314 57317 57318 57322 58482 58483 58484 58485 58486 58487 58488 58489 58490 58491 58492 58493 58495 58498
While we check with the data contributors, I recommend excluding these plots from analysis having a geographical component