I'm am using
adehabitatHR to estimate the size of the foraging range of seabirds using telemetry data. Is there anyway to take into account a physical boundary, where it is impossible for the animals to travel? In my case, the seabirds do not fly over land, therefore the tortuous coastline and islands are not 'available habitat'. Obviously, a simple kernel density estimate does not consider this, and I end up with a polygon that overlaps land.
this is my current code for the home range estimate. Is there anyway I can take into account land, using a polygon file? (I have a polygon shapefile called
dfrelocations is a dataframe containing the coordinates of animal relocations
relocations_xy <- as.data.frame(dfrelocations[,c("longitude","latitude")]) relocations_xysp <- SpatialPoints(relocations_xy) proj4string(relocations_xysp) <- CRS("+proj=longlat +datum=WGS84") relocations_xysp <- spTransform(relocations_xysp, CRS("+proj=utm +zone=9 ellps=WGS84")) bird_ud <- kernelUD(relocations_xysp, h=5000) bird_homerange <- getverticeshr(bird_ud, 95)