Fig. 4. Cate-Nelson models of relative light reflectance measurements and relative clipping yield for turfgrass in relation to soil NO3N desorbed from in situ anion exchange membranes. Relative reflectance and yield values are expressed as a fraction of plateau values for reflectance or yield for each sample date. Vertical lines to the x-axes represent critical x values which divide data into two populations, one which has a high probability of response to higher soil NO3N values and one which has a low probability of response. Horizontal lines minimize data in quadrants I and II. Error I indicates the percentage of observations in quadrant I and represents an overprediction of the critical x value. Error II indicates the percentage of observations in quadrant II and represents an underprediction of the critical x value. Data are pooled from 23 dates across two growing seasons. Number of observations for each plot is given by n.