**2.1 Spectral reflectance**

218 Soybean – Genetics and Novel Techniques for Yield Enhancement

direction. Stratospheric ozone depletion has led to elevated levels of ultraviolet-B (UV-B) radiation (280-320 nm) on the surface of the Earth. Increased UV-B levels have negative effects on human health (Norval et al., 2006) as well as on the plant development, morphology, and physiology (Jia Gio & Wang, 2008). Low influence UV-B radiation stimulates distinct responses, such as the accumulation of UV-absorbing pigments. Low influence of UV-B was also found to stimulate the transcript levels of a robust set of genes involved in stress responses (Rock, 2000). Although the effects of UV-B on plants are well characterized at the physiological level, little is known about the effects of UV-B on underground (root) physiology, particularly in interaction with other environmental factors. An increasing number of studies have been designed to test the interactions of environmental factors on plants, such as the interaction between UV-B and water stress (Cechin et al., 2008), interaction between salinity and Fe deficiency (Zancan et al., 2006), and

The aim of this chapter is to show some aspects of the recent applications of non-destructive remote sensing techniques, hyperspectral leaf reflectance and chlorophyll fluorescence, for detection and discrimination of the effects of some environmental stresses (salinity and enhanced UV-B radiation) on young soybean plants, as well as the influence of the biological nitrogen fixation on the spectral responses of the plants to stress. To evaluate the effects of a given stress a comparative analysis was performed between the changes of the leaf spectral reflectance and fluorescence data and the stress markers such as phenols, malondialdehyde, thiol groups, proline and hydrogen peroxide, and chlorophyll content

Generally, remote sensing refers to the activities of recording, observing, and perceiving (sensing) objects or events at far away (remote) places. Remote sensing is defined as a science and technology by which the characteristics of objects or events of interest can be identified, measured or analyzed without direct contact with the sensors. The spectral information relies on the properties of the light after multiple interactions, i.e., reflections, transmissions, and absorptions with the object. The information needs a physical carrier to travel from the objects/events to the sensors through an intervening medium. The electromagnetic radiation which is reflected or emitted from an object is the usual source of remote sensing data. However any media such as gravity or magnetic fields can be utilized. Remote sensing is a technology to identify and understand the object or the environmental condition through the uniqueness of their spectral responses. This technology offers advantages such as viewing parts of the Earth at different scales (synoptic view), monitoring of regions that are very remote or with restricted access, ability to obtain imagery of an area of the Earth at regular intervals over many years and to evaluate changes in the landscape as

A basic assumption made in remote sensing is that specific targets (soils of differed types, water with varying degrees of impurities, rocks of differing lithologies, or vegetation of various species) have an individual and characteristic manner of interacting with incident radiation that is described by the spectral response of that target. Different materials reflect and absorb visible (VIS) and infrared light differently at different wavelengths. They have different colours and brightness when seen under the sun. Thus, the targets can be differentiated by their 'spectral reflectance signatures', a term used to describe the spectral

interaction between UV-B radiation and Fe deficiency (Zancan et al., 2008).

that were estimated by biochemical methods.

well as capability to distinguish anthropogenic effects.

**2. Remote sensing methods** 

Methods based on reflectance makes use of VIS, near infrared (NIR), and short-wave infrared (SWIR) sensors to form images of the earth's surface by detecting the solar radiation reflected from targets on the ground. These methods rely on making measurements simultaneously in one or more wavebands. Spectrophotometers offer the simplest solution for spectral reflectance measurements. They measure spectrum of light reflected from the whole (mostly circular) field of view of the instrument but not provide any spatial information on the pattern of reflection (West et al., 2003). Earlier studies utilized multispectral sensors with low spatial (60 m to 80 m) and spectral resolution commonly collected in four to seven spectral bands in the VIS and NIR regions. Spectral resolution refers to the number and width of the portions of the electromagnetic spectrum measured by the sensor. A sensor may be sensitive to a large portion of the electromagnetic spectrum but have poor spectral resolution if it captures a small number of wide bands. Spatial resolution defines the level of spatial detail depicted in an image and it is directly related to image pixel size. The spatial property of an image is a function of the design of the sensor in terms of its field of view and the altitude at which it operates above the surface (Smith, 2001a). Early airborne systems included a multispectral camera mounted on board a light aircraft. Spectrometers at this time were bulky, heavy instruments which were not easily transportable in the field and most measurements were taken in laboratories.

Remote sensing technologies have advanced significantly over the past 10 to 15 years. With the development of hyperspectral remote sensing technologies, researchers have benefited from significant improvements in the spectral and spatial properties of the data, allowing for more detailed plant and environmental studies (Thenkabail et al., 2004; Blackburn, 2007). These technologies acquire many hundreds of spectral bands across the VIS, NIR, and midinfrared portions of the electromagnetic spectrum from 350 nm to 2500 nm, using satellite, airborne or hand-held devices. Advances in spectrometry and information technologies have resulted in state-of-the-art portable field instruments which allow for the collection of hand-held hyperspectral signatures. There are certain problems in the area of hyperspectral analysis connected with the optimal selection of bandwidth, number of bands and spatial as well as spectral resolutions and some constraints like data storage, communication bandwidth, discrimination/classification accuracy, minimum signal-to-noise ratio, sensor selection, data acquisition procedures and the cost factor.

The spectral reflectance responses are affected by factors such as soil nutrient status, the growth stage of the vegetation, the colour of the soil (which may be affected by recent weather conditions). In some instances, the nature of the interaction between incident

Spectral Remote Sensing of the Responses of Soybean Plants to Environmental Stresses 221

The shape of the reflectance spectrum is used for identification of vegetation type. For the same vegetation type, the reflectance spectrum also depends on other factors such as the leaf moisture content and health of the plants. Fig. 2 shows typical reflectance spectra of some species of green vegetation compared to a spectral signature for senescent leaves (Smith,

In past decade, the research efforts were focused on the elucidation of some aspects of the link between the spectral responses and the physiology of plants under stress. In stressed vegetation, leaf chlorophyll content decreases, thereby changing the proportion of lightabsorbing pigments, leading to a reduction in the overall absorption of light (Zarco-Tejada et al., 2000; Clay et al., 2006; Gang et al., 2010). These changes affect the spectral reflectance signatures of plants through a reduction in green reflection and an increase in red and blue reflections, resulting in changes in the normal spectral reflectance patterns of plants (Zarco-

More recent works have highlighted the importance of more specific narrow-band regions such as the red edge (maximum slope of vegetation reflectance from 680 nm to 720 nm) for predicting plant stress (Fitzgerald et al., 2006; Blackburn, 2007; Steele et al., 2008). The reflectance around red edge is sensitive to wide range of crop chlorophyll content and leaf internal scattering (Dawson & Curran, 1998). Experimental and theoretical studies show that red edge position shifts according to changes of chlorophyll content, N content, biomass and hydro status, age, plant health levels, and seasonal patterns (Filella & Penuelas, 1994; Pu et al., 2003; Hatfield et al., 2008;). These observations on red edge position can effectively be

Mathematical functions of two or more spectral bands are used rather than direct reflectance data to minimize the negative impact of interfering factors, such as the surrounding land cover, bare soil, or climatic/atmospheric conditions (McDonald et al., 1998; Huete et al., 2002). These functions are called vegetation indices (VIs), each designed for optimal correlation with a particular vegetation feature. The capacity of vegetation indices to

used to classify and distinguish different vegetation types and ages in the study.

Fig. 2. Leaf reflectance spectra of different vegetation types.

Tejada et al., 2000; Campbell et al., 2007).

2001b).

radiation and earth's surface materials will vary in time during the year, such as might be expected in the case of vegetation as it develops from the leafing stage, through growth to maturity and, finally to senescence. These responses also depend upon such factors as the orientation of the Sun, the height of the Sun in the sky (solar elevation angle), direction in which the sensor is pointing relative to nadir (the look angle), the topographic position of the target in terms of slope orientation, the state of health of vegetation if that is the target, and the state of the atmosphere.

Vegetation has a unique spectral signature which enables it to be distinguished readily from other types of land cover in an optical/infrared part of the electromagnetic spectrum. The spectral responses of vegetation are governed primarily by scattering and absorption characteristics of the leaf internal structure and biochemical constituents, such as pigments, water, nitrogen, cellulose and lignin (Sims & Gamon, 2002; West et al., 2003). In recent years, there has been an expanding body of literature concerning the relationship between the spectral reflectance properties of vegetation and the structural characteristics and pigment concentration in leaves (Gitelson et al., 2003; Blackburn, 2007; Sun et al., 2008; Hatfield et al., 2008). Chlorophyll pigment content is a major factor that dictates the amount of energy reflected or emitted and can be good indicator of crop health (Wu et al., 2008).

The function describing the dependence of the ratios of the intensity of reflected light to the illuminated light on wavelength in VIS (400-700 nm), NIR (700-1200 nm), and SWIR (1200- 2400 nm) spectral ranges is the spectral reflectance characteristic (SRC) of the target. Fig.1 presents the typical spectral reflectance characteristics of green vegetation. The labelled arrows indicate the common wavelength bands used in optical remote sensing of vegetation: A - blue band, B - green band, C - red band, D - NIR band, and E - SWIR band. Reflectance is low in both the blue (450 nm) and red (670 nm) regions of the spectrum, due to absorption by chlorophyll for photosynthesis, also known as the chlorophyll absorption bands. It has a peak at the green region (550 nm) which gives rise to the green colour of vegetation. In the NIR region, the reflectance is much higher than that in the VIS band due to the cellular structure in the leaves. Hence, vegetation can be identified by the high NIR but generally low VIS reflectance. The reflectance of vegetation in the SWIR region is more varied, depending on the types of plants and the plant's water content. Water has strong absorption bands around 1.45, 1.95 and 2.50 µm. The SWIR band can be used in detecting plant drought stress and delineating burnt areas and fire-affected vegetation.

Fig. 1. Typical spectral reflectance characteristic of green vegetation in the VIS, NIR and SWIR ranges.

radiation and earth's surface materials will vary in time during the year, such as might be expected in the case of vegetation as it develops from the leafing stage, through growth to maturity and, finally to senescence. These responses also depend upon such factors as the orientation of the Sun, the height of the Sun in the sky (solar elevation angle), direction in which the sensor is pointing relative to nadir (the look angle), the topographic position of the target in terms of slope orientation, the state of health of vegetation if that is the target,

Vegetation has a unique spectral signature which enables it to be distinguished readily from other types of land cover in an optical/infrared part of the electromagnetic spectrum. The spectral responses of vegetation are governed primarily by scattering and absorption characteristics of the leaf internal structure and biochemical constituents, such as pigments, water, nitrogen, cellulose and lignin (Sims & Gamon, 2002; West et al., 2003). In recent years, there has been an expanding body of literature concerning the relationship between the spectral reflectance properties of vegetation and the structural characteristics and pigment concentration in leaves (Gitelson et al., 2003; Blackburn, 2007; Sun et al., 2008; Hatfield et al., 2008). Chlorophyll pigment content is a major factor that dictates the amount of energy

The function describing the dependence of the ratios of the intensity of reflected light to the illuminated light on wavelength in VIS (400-700 nm), NIR (700-1200 nm), and SWIR (1200- 2400 nm) spectral ranges is the spectral reflectance characteristic (SRC) of the target. Fig.1 presents the typical spectral reflectance characteristics of green vegetation. The labelled arrows indicate the common wavelength bands used in optical remote sensing of vegetation: A - blue band, B - green band, C - red band, D - NIR band, and E - SWIR band. Reflectance is low in both the blue (450 nm) and red (670 nm) regions of the spectrum, due to absorption by chlorophyll for photosynthesis, also known as the chlorophyll absorption bands. It has a peak at the green region (550 nm) which gives rise to the green colour of vegetation. In the NIR region, the reflectance is much higher than that in the VIS band due to the cellular structure in the leaves. Hence, vegetation can be identified by the high NIR but generally low VIS reflectance. The reflectance of vegetation in the SWIR region is more varied, depending on the types of plants and the plant's water content. Water has strong absorption bands around 1.45, 1.95 and 2.50 µm. The SWIR band can be used in detecting

reflected or emitted and can be good indicator of crop health (Wu et al., 2008).

plant drought stress and delineating burnt areas and fire-affected vegetation.

Fig. 1. Typical spectral reflectance characteristic of green vegetation in the VIS, NIR and

and the state of the atmosphere.

SWIR ranges.

The shape of the reflectance spectrum is used for identification of vegetation type. For the same vegetation type, the reflectance spectrum also depends on other factors such as the leaf moisture content and health of the plants. Fig. 2 shows typical reflectance spectra of some species of green vegetation compared to a spectral signature for senescent leaves (Smith, 2001b).

Fig. 2. Leaf reflectance spectra of different vegetation types.

In past decade, the research efforts were focused on the elucidation of some aspects of the link between the spectral responses and the physiology of plants under stress. In stressed vegetation, leaf chlorophyll content decreases, thereby changing the proportion of lightabsorbing pigments, leading to a reduction in the overall absorption of light (Zarco-Tejada et al., 2000; Clay et al., 2006; Gang et al., 2010). These changes affect the spectral reflectance signatures of plants through a reduction in green reflection and an increase in red and blue reflections, resulting in changes in the normal spectral reflectance patterns of plants (Zarco-Tejada et al., 2000; Campbell et al., 2007).

More recent works have highlighted the importance of more specific narrow-band regions such as the red edge (maximum slope of vegetation reflectance from 680 nm to 720 nm) for predicting plant stress (Fitzgerald et al., 2006; Blackburn, 2007; Steele et al., 2008). The reflectance around red edge is sensitive to wide range of crop chlorophyll content and leaf internal scattering (Dawson & Curran, 1998). Experimental and theoretical studies show that red edge position shifts according to changes of chlorophyll content, N content, biomass and hydro status, age, plant health levels, and seasonal patterns (Filella & Penuelas, 1994; Pu et al., 2003; Hatfield et al., 2008;). These observations on red edge position can effectively be used to classify and distinguish different vegetation types and ages in the study.

Mathematical functions of two or more spectral bands are used rather than direct reflectance data to minimize the negative impact of interfering factors, such as the surrounding land cover, bare soil, or climatic/atmospheric conditions (McDonald et al., 1998; Huete et al., 2002). These functions are called vegetation indices (VIs), each designed for optimal correlation with a particular vegetation feature. The capacity of vegetation indices to

Spectral Remote Sensing of the Responses of Soybean Plants to Environmental Stresses 223

mostly associated to PSII (Dekker et al., 1995), whereas the longer wavelength emission originates from antenna chlorophyll of both PSI and PSII (Agati et al., 2000; Buschmann, 2007). Several environmental factors, including water, salinity, light and nutrients, affect the process of photosynthesis and may lead to plant stress. Changes in chlorophyll function take place before changes in chlorophyll content, before any physical signs of tissue or chlorophyll deterioration are manifested in the plant, and therefore alterations in the fluorescence signal occur before any visible signs are apparent (Cambpell et al., 2007; Li et al., 2010). Under conditions of stress, some plant mechanisms for disposing of excess energy do not work efficiently, thus causing changes in the competing reactions of photochemistry, heat loss and fluorescence. Although the total amount of chlorophyll fluorescence is very small (only 2 or 3% of total light absorbed), measurement is quite easy. The spectrum of fluorescence is different to that of absorbed light with the peek of fluorescence emission being at longer wavelength than that of absorption. Therefore, fluorescence yield can be quantified by exposing a leaf to light of defined wavelength and measuring the amount of

Various fluorescence intensity ratios, combining the emissions at blue (F440), green (F520), red (F690), and NIR (F740) wavelengths, were proposed for probing the vegetation vitality status and stress responses (Buschmann et al., 2000; Mishra & Gopal, 2008). The red ChlF emission between 684-695 nm is strongly reabsorbed by the Chl pigments in the upper layer leaf cells (Agati et al., 1993; Dau, 1994), while the NIR ChlF between 730–740 nm is reabsorbed to a much smaller extent. Consequently, the ratio between the red and far-red ChlF bands (e.g. F690/F740) decreases with increasing leaf Chl content in a curvilinear relationship, which can be used as a good inverse indicator of Chl content changes due to plant growth or stress events (Buschmann, 2007). Finally, the UV excited blue-to-red/NIR fluorescence intensity ratios (F440/F690 and F440/F740) were proposed as indicators of the leaf physiological development (Stober et al., 1994; Meyer et al., 2003), but also as marker of

The red and NIR fluorescence emissions by Chl a are highly dynamic, being modulated by photochemical and non-photochemical quenching. These dynamic phenomena yielded important insights into the molecular processes of photosynthesis that occur within time-scales ranging from femtoseconds to minutes depending on the power of an actively applied actinic light (Govindjee, 1995; Nedbal & Koblizek, 2006; Baker, 2008). Most widely used field observations are active, using devices exciting the photosynthetic machinery with a measuring light and recording the induced fluorescence. Introduction of the pulsed amplitude modulation (PAM) fluorometer allowed non-imaging outdoor measurements in broad daylight (Schreiber et al., 1986). Fluorescence imaging was introduced in the laboratory by Omasa et al. (1987) and modified for field surveys in the mid-1990s by Nedbal et al. (2000). The laser pulses of actinic light, which can be discriminated from static and panchromatic background light, are applied to elicit fluorescent transients when measuring fluorescence from a distance (Cecchi et al., 1994; Corp et al., 2006). The footprint of such a light detection and ranging (LIDAR) laser beam can be expanded from several centimetres up to metres to cover larger observation areas or to decrease the power of the excitation source (Saito et al., 2005). The first field laser-induced vegetation fluorescence was observed by Measures et al., (1973). Lately, an eye-safe outdoor laser-induced fluorescence transient (LIFT) fluorometer has been constructed. This device is able to measure the fluorescence parameters and nonphotochemical quenching or electron transport rate from a distance of about 30-50 m (Ananyev et al., 2005; Kolber et al., 2005). A new generation active field fluorescence

light re-emitted at longer wavelengths (Maxwell & Johnson, 2000).

the nutrition availability and stress occurrence (Heisel et al., 1996).

characterize natural canopies and agricultural crops has been demonstrated in numerous studies aimed at seasonal phenology (Carter, 1998; Qi et al., 2000), biomass prediction (Broge and Leblanc, 2001; Haboudane et al., 2004), mapping chlorophyll content (Haboudane et al., 2002). Numerous studies have documented the use of vegetation indices such as ratio vegetation index (RVI) and normalized difference vegetation index (NDVI) in the detection of crop stress (Kobayashi et al., 2001; Vigier et al., 2004; Yang et al., 2009). Combining individual spectral reflectance bands as simple ratio vegetation indices (SRVI) has been a common approach in remote sensing because it generally reduces the effects of spectral noise and allows for better temporal comparisons due to minimization of atmospheric effects (Carter & Miller, 1994). Commonly, SRVIs have consisted of the ratio of blue to red wavebands in an effort to detect responses due to changes in chlorophyll a and b concentrations. Gitelson et al. (2003, 2006) suggested the use of empirical vegetation indices, calculated from the reflectance of three wavelengths that were highly correlated with chlorophyll (Chl), carotenoid, and anthocyanin concentrations to estimate the content of foliar pigments in single leaves. Furthermore, various statistical and artificial intelligence methods have been used to analyze the remotely sensed data in agricultural crops. Among many, popular approaches include cluster analysis (Holden & LeDrew, 1998), principal component analysis (Zhang et al., 2002; 2003), partial-least square regression (Huang & Apan, 2006), artificial neural networks (Liu et al., 2008).

With the advent of hyperspectral remote sensing technology, more detailed data are potentially available. Therefore the extracting meaningful relationships of the overwhelming quantity of data are necessary. Currently, a variety of techniques have been used including a number of different vegetation indices, band absorption analysis, spectral mixture analysis, "red edge" position, statistical analysis, wavelet transform and neural networks (Thenkabail et al., 2004; Delalieux et al., 2007; Steele et al., 2008b).
