Hi.
I'm trying to work out if there is spatial autocorrelation in soil physicochemical properties in an avocado orchard. I collected soil from various points across an orchard block and have measured soil properties for each of those points. I don't have Lat-Long co-ords for each point, but have the X-Y co-ords. I have point data and the associated soil properties - upon reading I've found Moran's I is probably what I need to calculate to see if points which are close by share similar properties (i.e. there is some distance-based clustering of my data).
head(Soil_data_O1)
Sample_ID Site Design X Y Distance Ammonium-N Nitrate-N Phosphorus Potassium Organic Carbon Sulfur Conductivity
1 O1-G10 Orchard 1 Grid -11.00 -9 -16.5 6 3 61 22 3.83 61.4 0.495
2 O1-G11 Orchard 1 Grid 0.01 0 0.0 6 3 111 75 3.89 80.6 0.343
3 O1-G2 Orchard 1 Grid 8.00 -3 12.0 8 2 49 52 4.07 86.7 0.496
4 O1-G3 Orchard 1 Grid 6.00 10 10.2 12 29 154 223 4.41 33.7 0.323
5 O1-G4 Orchard 1 Grid 2.00 14 9.2 4 20 156 119 3.48 27.7 0.167
6 O1-G6 Orchard 1 Grid 2.00 -14 -9.2 6 13 90 85 5.22 26.7 0.155
pH (CaCl2) Aluminium Calcium Magnesium Sodium
1 5.9 0.063 15.86 2.95 1.27
2 5.3 0.095 14.61 2.14 1.14
3 4.1 0.961 15.66 2.71 1.72
4 4.3 0.365 31.53 5.74 1.21
5 4.0 1.593 5.95 1.41 0.43
6 3.8 2.144 6.72 1.34 0.49
mypattern_O1 <- ppp(Soil_data_O1[,4], Soil_data_O1[,5], c(-11,9), c(-14,16)) #Spatstat package
Orch1 <- plot(unmark(mypattern_O1), cols = "#009E73", main = 'Orchard 1 layout', border = 'black', pch=19)
marks(mypattern_O1) <- Soil_data_O1[,7:18]
Orch1_props <- plot(mypattern_O1, cols = "#009E73", main = 'Orchard 1', border = 'black')
I am very new at this kind of analysis, so asking what else I need to do to get those answers? Thanks.