Attenuated Total Reflectance Spectroscopy of Plant Leaves

8/13/2019 Attenuated Total Reflectance Spectroscopy of Plant Leaves

1/14

www.newphytologist.org 305

Research

BlackwellPublishingLtd

Attenuated total reflectance spectroscopy of plant leaves:a tool for ecological and botanical studies

Beatriz Ribeiro da Luz

Department of Ecology, Institute of Biosciences, University of So Paulo. R. do Mato, Travessa 14, 321, 05508-900, So Paulo, SP, Brazil and US Geological

Survey, Mail Stop 954, 12201 Sunrise Valley Drive, Reston, VA 20192, USA

Summary

Attenuated total reflectance (ATR) spectra of plant leaves display complex absorption

features related to organic constituents of leaf surfaces. The spectra can be recorded

rapidly, both in the field and in the laboratory, without special sample preparation.

This paper explores sources of ATR spectral variation in leaves, including com-

positional, positional and temporal variations. Interspecific variations are also

examined, including the use of ATR spectra as a tool for species identification.

Positional spectral variations generally reflected the abundance of cutin and the

epicuticular wax thickness and composition. For example, leaves exposed to full

sunlight commonly showed more prominent cutin- and wax-related absorption

features compared with shaded leaves. Adaxial vs. abaxial leaf surfaces displayed

spectral variations reflecting differences in trichome abundance and wax composition.

Mature vs. young leaves showed changes in absorption band position and intensity

related to cutin, polysaccharide, and possibly amorphous silica development on and

near the leaf surfaces.

Provided that similar samples are compared (e.g. adaxial surfaces of mature,

sun-exposed leaves) same-species individuals display practically identical ATR

spectra. Using spectral matching procedures to analyze an ATR database containing

117 individuals, including 32 different tree species, 83% of the individuals were

correctly identified.

Key words:

cuticle, identification of plants, chemistry of leaf surface, thermal infrared,

Fourier transform infrared (FTIR) spectroscopy, attenuated total reflectance (ATR).

New Phytologist

(2006) 172

: 305318

The Authors (2006). Journal compilation New Phytologist

(2006)

doi

: 10.1111/j.1469-8137.2006.01823.x

Author for correspondence:

Beatriz Ribeiro da Luz

Tel: +1 703 6486374

Fax: +1 703 6486383

E-mail: [email protected]

Received: 22 March 2006

Accepted: 19 May 2006

Introduction

The need to characterize floristic composition is a key aspectof many ecological studies; however, the identification ofplant species is typically a slow process that depends on mor-phological and anatomical observations of plant structures.For example, many plant species are distinguished by theirfloral structures and thus the flowers must be present at thetime that botanical surveys are being performed. In some areas

with high species diversity determining floristic compositionis further complicated by the fact that many species remainundescribed. This lack of knowledge, coupled with the rapid

destruction of natural vegetation in some areas, points to theneed for more efficient techniques for identifying species and

for recognizing distinctive physical and chemical attributes.Leaves are complex assemblages of organic compounds and

it might be expected that they would display distinctive spectralfeatures in the thermal infrared energy range (TIR; 4000400 cm

1

). Fundamental vibration modes of various molecularfunctional groups produce characteristic spectral absorptionfeatures that can serve to fingerprint many compounds(Silverstein & Webster, 1998). Such functional groups andrelated spectral features include hydroxyl (OH) in alcoholsand acids, carbonyl (C

=

O) in esters, ketones, aldehydes and


2/14

New Phytologist

(2006) 172

: 305318

www.newphytologist.org


(2006)

Research306

acids, and methyl (CH

3

) and methylene (CH

2

) in alkanes.Libraries of TIR spectra currently have a wide range of appli-cations in such diverse fields as chemistry, geology, industrialprocess control and forensics.

In a preliminary exploration of leaf TIR spectral properties,Salisbury (1986) and Salisbury & Milton (1987, 1988)determined that 13 deciduous tree species displayed reflectance

features that were unique for each species. Such spectralvariability between different plant species parallels the findingsof Holloway (1982a) who determined that cuticularstructures observed for particular species also were generallyunique. The cuticle is the most superficial layer of the aerialparts of terrestrial plants, and consists of a matrix of polymerizedlipid, cutin, and/or polymethylene chains, cutan, permeatedby intracuticular waxes and covered by epicuticular waxes(Holloway, 1982b; Jeffree, 1996; Heredia, 2003). Cutin iscomposed mainly of esterified monomers of hydroxyl- andepoxy-fatty acids (Holloway, 1982b; Kolattukudy, 1996). Theepicuticular and intracuticular waxes are composed mainly

of long-chain aliphatic hydrocarbons, esters, primary andsecondary alcohols, ketones, aldehydes and fatty acids. Aromaticand cyclic compounds such as flavonoids and terpenoids mayalso be present in smaller amounts (Tulloch, 1976; Baker,1982b; Bianchi, 1995). The unique characteristics of plantcuticles derive both from the numerous organic compoundsinvolved, and the diverse structural arrangements of thecomponents (Holloway, 1982a).

Thermal infrared transmission spectra previously havebeen used to help understand the composition and the struc-ture of leaf surfaces (Hallam & Chambers, 1970; Holloway,1982b; Villena et al

., 2000), but traditional transmission

methods involving the preparation of KBr sample pellets arenot well-suited for the study of fresh, water-bearing, plantmaterials. A relatively recent technique, attenuated totalreflectance (ATR), enables the rapid collection of a transmission-like spectrum of a leaf-surface simply by placing a sample incontact with a special, high index of refraction, crystal (Merk

et al

., 1998; Dubis et al

., 1999; Dubis et al

., 2001). This paperdiscusses the general origins of spectral features seen in

ATR spectra of leaves, examines sources of variabilitybetween samples, and explores the potential use of ATRmeasurements in the laboratory and the field as a tool for speciesidentification.

Materials and Methods

Sample collection

The study was conducted in the Washington, DC, area, USA,between July 2001 and September 2004. Leaves, mostly fromnative trees, were collected at the US National Arboretum,

Washington, DC, and in areas surrounding the US Geo-logical Survey, Oatlands plantation and the town of Lovettsvillein northern Virginia. Samples of tropical species were also

obtained from the US Botanic Garden, which has greenhousefacilities in the Washington, DC, area. Lycopersicon esculentum

(tomato) and Beta vulgaris

(red beet) were purchased from theorganic grocery Whole Foods (Reston, VA, USA). In totalthere were samples from 268 individuals belonging to 133species, 89 genera and 59 families.

Leaves from deciduous trees were collected every month

between May and September, 2002, to analyse the spectraldifferences at different stages of the growing season, and in

July 2003 and September 2004 to determine if there wereany interannual spectral differences. The tree species usedfor this analysis of temporal spectral variations were Acerrubrum

, Aesculus hippocastanum

, Aesculus octandra

, Carpinuscaroliniana

, Carya ovata

, Cornus florida

, Fagus grandifolia

,

Ginkgo biloba

, Liquidambar styraciflua

, Liriodendron tulipifera

,

Maclura pomifera

, Magnolia grandiflora

, Prunus serotina

, Quercusalba

, Quercus rubra

and Tilia cordata.

Positional variations of leaf samples were also studied,including spectral variations between the adaxial (upper) and

abaxial (lower) leaf surfaces, and spectral variations related tothe degree of sun exposure of the leaf on the tree. Sun leaves

were collected from the south aspect and upper external partsof tree canopies, and shade leaves from the north aspect andlower internal parts. The tree species used for the sun andshade analysis wereA. hippocastanum

, F. grandifolia

, G. biloba

,

M. grandiflora

, L. tulipifera

, Q. robur

and Q. rubra

.In all cases, leaves were collected in batches of 10 20 samples,

bagged in plastic, and placed in an ice chest for transport tothe laboratory. Damp cotton balls were placed in the bags toavoid desiccation of the leaves, and the spectral measurements

were completed within 15 d from collection.

Laboratory and field attenuated total reflectancemeasurements

The ATR measurements were made with Fourier transforminfrared spectrometers equipped with accessory optics thatinclude flat crystalline plates having a high refractive index (inthis study the crystalline plates were composed of ZnSe). An

ATR accessory is designed so that the infrared beam impingeson the plate at an angle greater than the critical angle causingtotal internal reflection. Under these conditions, the beamintensity is attenuated by a surface evanescent wave thatpenetrates a short distance into any absorbing sample placed

in contact with the crystalline plate. The depth of penetrationvaries with the angle of incidence, the wavelength, and theindices of refraction of both the plate and the sample (seeformula in Spragg, 2000). Plant cuticles have a refractiveindex of c

. 1.5 (Holloway, 1982a, p. 7) and ZnSe crystals havean index of 2.43; this establishes the ATR penetration depthin leaves from c

. 0.3 m at 4000 cm

1

to c

. 1.7 m at 700 cm

1

.The resulting spectrum is similar to a transmission spectrum,

with some differences in peak intensities because of thevariable penetration.


3/14


(2006)

www.newphytologist.org

New Phytologist

(2006) 172

: 305318

Research 307

The laboratory measurements were made with a Thermo-Electron Corp. Nexus 670 FTIR (Fourier transform infrared;Thermo Electron Corp., Waltham, MA, USA) spectrometer thatis continuously purged with dry air. A deuterated triglycinesulfate detector was used to cover the energy range from4000 cm

1

and 650 cm

1

. Leaves were placed in directcontact with the ZnSe crystal, and the average of 100 scans

was recorded for each leaf surface.A field spectrometer equipped with a prototype field ATR

accessory was used to simulate an ecological study requiring insitu

species identification. The simulated study was made inthe State Tree Grove at the National Arboretum, Washington,DC, where there are multiple individuals of numerous treespecies, all of which have been identified and labeled. Thespectrometer was a Model 102F FTIR manufactured byDesigns and Prototypes Ltd, (Simsbury, CT, USA). Thisspectrometer uses a sandwich detector to cover the full spectralrange: an InSb detector spanned the range from 4000 cm

1

to1818 cm

1

, and an HgCdTe detector covered the range from

1818 cm

1

to 714 cm

1

. The detector assembly was cooledwith liquid nitrogen, and a low-power (

Documents

Attenuated Total Reflectance Spectroscopy of Plant Leaves