Near infrared spectroscopy analysis of nitrogen and phosphorus in leaves of woody plants
In this study, we used near infrared reflectance spectroscopy (NIRS) to estimate the N and P contents of leáf samples of eighteen woody plant species. Samples of these species were obtained in western Spain and were typical of mountain, riparian, and relatively dry areas. A remarkable variation in the contents of both elements was found in the samples: N ranged from 6.6 to 45.0 g kil, and P from 0.24 to 2.97 g kg'!. Multiple linear regression (MLR) and partial least squares regression (PLSR) were used for developíng NIRS calibrations. The calibratíon statistics were better when using PLSR. Using this procedure, coefficients of multiple determination (R2) of 0.99 and 0.94 were obtaíned for N and P, respectively; the standard errors of cross validation (SECV) wereO.79 g kg-l and 0.15 g kg-!, and root mean· square errors of prediction (RMSEP) were 0.76 g kg-l and 0.11 g kg'!.
| Main Authors: | , , , , , |
|---|---|
| Format: | capítulo de libro biblioteca |
| Published: |
European Grassland Federation
2004
|
| Subjects: | Chemical composition, Woody plant species, NIRS, |
| Online Access: | http://hdl.handle.net/10261/266615 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | In this study, we used near infrared reflectance spectroscopy (NIRS) to estimate the N and P contents of leáf samples of eighteen woody plant species. Samples of these species were obtained in western Spain and were typical of mountain, riparian, and relatively dry areas. A remarkable variation in the contents of both elements was found in the samples: N ranged from 6.6 to 45.0 g kil, and P from 0.24 to 2.97 g kg'!. Multiple linear regression (MLR) and partial least squares regression (PLSR) were used for developíng NIRS calibrations. The calibratíon statistics were better when using PLSR. Using this procedure, coefficients of multiple determination (R2) of 0.99 and 0.94 were obtaíned for N and P, respectively; the standard errors of cross validation (SECV) wereO.79 g kg-l and 0.15 g kg-!, and root mean· square errors of prediction (RMSEP) were 0.76 g kg-l and 0.11 g kg'!. |
|---|