Variable selection in near infrared spectra for the biological characterization of soil and earthworm casts

被引:64
作者
Cecillon, Lauric [1 ,2 ]
Cassagne, Nathalie [3 ]
Czarnes, Sonia [4 ]
Gros, Raphael [5 ]
Brun, Jean-Jacques [1 ]
机构
[1] Cemagref Grenoble, Mt Ecosyst Res Unit, F-38402 St Martin Dheres, France
[2] French Agcy Environm & Energy Management, ADEME, F-49004 Angers 01, France
[3] INRA, URFM, UR 629, F-84914 Avignon 9, France
[4] Univ Lyon 1, CNRS, INRA, UMR5557,USC Ecol Microbenne 1193, F-69622 Villeurbanne, France
[5] Fac Sci & Tech St Jerome, IMEP, F-13397 Marseille 20, France
关键词
NIBS; PLSR; wavelength selection; soil biogenic structures; soil macrofauna; microbiological activity;
D O I
10.1016/j.soilbio.2008.03.016
中图分类号
S15 [土壤学];
学科分类号
0903 ; 090301 ;
摘要
Near infrared reflectance spectroscopy (NIRS) was used to predict six biological properties of soil and earthworm casts including extracellular soil enzymes, microbial carbon, potential nitrification and clenitrification. Partial least squares regression (PLSR) models were developed with a selection of the most important near infrared wavelengths. They reached coefficients of determination ranging from 0.81 to 0.91 and ratios of performance-to-cleviation above 2.3. Variable selection with the variable importance in the projection (VIP) method increased dramatically the prediction performance of all models with an important contribution from the 1750-2500 nm region. We discuss whether selected wavelengths can be attributed to macronutrient availability or to microbial biomass. Wavelength selection in NIR spectra is recommended for improving PLSR models in soil research. (c) 2008 Elsevier Ltd. All rights reserved.
引用
收藏
页码:1975 / 1979
页数:5
相关论文
共 34 条
[1]   Development of a sensitive and rapid method for the measurement of total microbial activity using fluorescein diacetate (FDA) in a range of soils [J].
Adam, G ;
Duncan, H .
SOIL BIOLOGY & BIOCHEMISTRY, 2001, 33 (7-8) :943-951
[2]   PHYSIOLOGICAL METHOD FOR QUANTITATIVE MEASUREMENT OF MICROBIAL BIOMASS IN SOILS [J].
ANDERSON, JPE ;
DOMSCH, KH .
SOIL BIOLOGY & BIOCHEMISTRY, 1978, 10 (03) :215-221
[3]   A SUBSTRATE-INDUCED RESPIRATION (SIR) METHOD FOR MEASUREMENT OF FUNGAL AND BACTERIAL BIOMASS ON PLANT RESIDUES [J].
BEARE, MH ;
NEELY, CL ;
COLEMAN, DC ;
HARGROVE, WL .
SOIL BIOLOGY & BIOCHEMISTRY, 1990, 22 (05) :585-594
[4]   The reflectance spectra of organic matter in the visible near-infrared and short wave infrared region (400-2500 nm) during a controlled decomposition process [J].
BenDor, E ;
Inbar, Y ;
Chen, Y .
REMOTE SENSING OF ENVIRONMENT, 1997, 61 (01) :1-15
[5]  
CECILLON L, 2007, COST ACTION 639 GREE, P103
[6]  
Chessel D., 2004, R NEWS, V4, P6, DOI DOI 10.2307/3780087
[7]   Performance of some variable selection methods when multicollinearity is present [J].
Chong, IG ;
Jun, CH .
CHEMOMETRICS AND INTELLIGENT LABORATORY SYSTEMS, 2005, 78 (1-2) :103-112
[8]   Visible-near infrared reflectance spectroscopy for rapid, nondestructive assessment of wetland soil quality [J].
Cohen, MJ ;
Prenger, JP ;
DeBusk, WF .
JOURNAL OF ENVIRONMENTAL QUALITY, 2005, 34 (04) :1422-1434
[9]   CELLULASE ACTIVITY OF SOILS [J].
DENG, SP ;
TABATABAI, MA .
SOIL BIOLOGY & BIOCHEMISTRY, 1994, 26 (10) :1347-1354
[10]  
FAO, 2006, WORLD REFERENCE BASE, DOI DOI 10.1017/S0014479706394902