confused by randomForest "%IncMSE" importance output in R

I'm running a random forest model as follows: rf_ranfor Then I want to inspect the importance of the features using: randomForest::importance(rf_ranfor, scale=FALSE) or randomForest::importance(rf_ranfor, scale=TRUE) That gives output: > randomForest::importance(rf_ranfor, scale=FALSE) %IncMSE IncNodePurity LONG 0.055181651 331.3672 LAT 0.175562843 1128.5319 EC 0.134742912 868.0401 OC 0.256949550 874.3882 CaCO3 0.960136877 5922.6065 P 0.013398794 198.9498 N 0.061491399 294.8151 K 0.105034730 911.2521 coarse 0.005298175 136.0192 clay 0.098223706 1388.9736 PTotal 0.009028657 192.3452 WRB_LV1 0.065588390 560.0991 ParMat1 0.051942279 1209.5763 myLU 0.454419466 4514.0025 elevation 0.020700342 199.9043 slope 0.009892190 184.1195 tavg 0.135494664 1021.0937 precp 0.029931459 292.8479 > randomForest::importance(rf_ranfor, scale=TRUE) %IncMSE IncNodePurity LONG 51.01640 331.3672 LAT 27.53817 1128.5319 EC 53.21809 868.0401 OC 47.40027 874.3882 CaCO3 123.05523 5922.6065 P 32.71670 198.9498 N 30.72981 294.8151 K 54.70646 911.2521 coarse 17.88074 136.0192 clay 52.89429 1388.9736 PTotal 21.60447 192.3452 WRB_LV1 33.67086 560.0991 ParMat1 25.67040 1209.5763 myLU 103.56293 4514.0025 elevation 34.68472 199.9043 slope 26.20564 184.1195 tavg 24.01220 1021.0937 precp 41.90694 292.8479 I'm trying to understand what the %IncMSE column now actually means and what its 'unit' is. In the vignette of function importance , the following is specified: The first measure [(=%IncMSE)] is computed from permu…