Establishment of continuous critical levels for indices of plant and presidedress soil nitrogen status in the potato crop

Critical levels for six plant and presidedress soil nitrogen (N) indices were established for the growing season, and their relative accuracy to diagnose the need for supplemental sidedress N was also compared. Field trials were conducted from 1996 to 1998 with irrigated potato in northeastern Portugal. Fertilizer treatments included several preplant and sidedress N rates. Petiole nitrate concentrations (determined by a standard laboratory method and with the portable RQflex reflectometer), leaf N content, leaf greenness (SPAD-502 chlorophyll meter), presidedress soil NO3-N and presidedress soil inorganic-N (NO3-N + NH4-N) were selected as N indicators. The Cate-Nelson graphical method and an analytical procedure using Mitscherlich type curves were used to determine critical levels. In both cases, a yield reduction of 10% was accepted. The accuracy of the diagnostics was estimated from the Cate-Nelson graphical method, quantifying the point percentage that appears in negative quadrants (the error rate). The graphical method yielded lower critical levels, appearing as the most conservative sidedress N recommendation basis. The critical levels decreased linearly between 15 and 45 days after emergence (DAE). The linear equations achieved provide continuous critical levels for the growing season and are shown below: Petiole NO3-N (g kg-1, dry wt basis) = -0.737 DAE + 36.879(r2 = 0.92); Petiole NO3 (g kg-1, from fresh tissue) = -0.182 DAE + 9.417 (r2 = 0.69); Leaf N (g kg-1, dry wt basis) = -0.453 DAE + 61.028 (r2 = 0.91); Chlorophyll-SPAD (SPAD units) = - 0.463 DAE + 64.400 (r2 = 0.93); Soil NO3-N (mg kg-1) = -1.096 DAE + 49.279 (r2 = 0.92); and Soil inorganic-N (mg kg-1) = -1.245 DAE + 56.599(r2 = 0.92). The N indicators with lower error rate were the presidedress soil NO3-N and presidedress soil inorganic-N (both showing an error rate of 8.3%), followed by petiole nitrate concentration [determined in laboratory (12.0%) and with RQflex reflectometer (12.5%), leaf N content (13.0%), and leaf greenness (14.6%)]. Error rates were similar throughout the growing season, meaning that it is possible to get information about the need for supplementary N in the very early growth stages.