 |
| Titel |
Near infrared spectra are more sensitive to land use changes than physical, chemical and biological soil properties |
| VerfasserIn |
C. Guerrero, R. Zornoza, J. Mataix-Solera, J. Mataix-Beneyto, K. Scow |
| Konferenz |
EGU General Assembly 2009
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 11 (2009) |
| Datensatznummer |
250025187
|
|
|
|
|
|
| Zusammenfassung |
| We studied the sensibility of the near infrared spectra (NIR) of soils to the changes caused by
land use, and we compared with the sensibility of different sets of physical, chemical and
biological soil properties. For this purpose, we selected three land uses, constituted by forest,
almond trees orchards, and orchards abandoned between 10 and 15 years previously to
sampling. Sampling was carried out in four different locations from the province of
Alicante (SE Spain). We used discriminant analysis (DA) using different sets of soil
properties. The different sets tested in this study using DA were: (1) physical and
chemical properties (organic carbon, total nitrogen, available phosphorus, pH, electrical
conductivity, cation exchange capacity, aggregate stability, water holding capacity,
and available Ca, Mg, K and Na), (2) biochemical properties (microbial biomass
carbon, basal respiration and urease, phosphatase and β-glucosidase activities), (3)
phospholipids fatty acids (PLFAs), (4) physical, chemical and biochemical properties (all
properties of the previous sets), and (5) the NIR spectra of soils (scores of the principal
components).
In general, all sets of properties were sensible to land use. This was observed in the DAs by
the separation (more or less clear) of samples in groups defined by land use (irrespective of
site). The worst results were obtained using soil physical and chemical properties. The
combination of physical, chemical and biological properties enhanced the separation of
samples in groups, indicating higher sensibility. It is accepted than combination of properties
of different nature is more effective to evaluate the soil quality. The microbial community
structure (PLFAs) was highly sensible to the land use, grouping correctly the 100% of the
samples according with the land use. The NIR spectra were also sensitive to land use.
The scores of the first 5 components, which explained 99.97% of the variance,
grouped correctly the 85% of the soil samples by land use, but were unable to group
correctly the 100% of the samples. Surprisingly, when the scarce variance presents
in components 5 to 40 was also used, the 100% of the samples were grouped by
land use, as it was observed with PLFAs. But PLFAs analysis is expensive and
time-consuming (some weeks). In contrast, only some minutes are needed for the
obtainment of the NIR spectra. Additionally, no chemicals are need, decreasing the
costs.
The NIR spectrum of a soil contains relevant information about physical, chemical and
biochemical properties. NIR spectrum could be considered as an integrated vision
of soil quality, and as consequence offers an integrated vision of perturbations.
Thus, NIR spectroscopy could be used as tool to monitoring soil quality in large
areas.
Acknowledgements: Authors acknowledge to ”Bancaja-UMH´ ´ for the financial support of
the project ”NIRPROS´ ´ . |
|
|
|
|
|
|