Estimation of Photosynthetic Rate of Plant
from Hyper-spectral Remote Sensing of Biochemical Content
3. Materials and Methods
3.1. Plant Materials and Measurements of Leaf Reflectance and of Light-Photosynthetic Rate Curve
Cucumber seeds were planted into a plastic pot and maintained for 28days in a greenhouse chamber. These plants were cultivated under two different conditions in order to make leaves with different concentration of chlorophyll, nitrogen, and so on. While conditions for one series was most suitable with regard to light condition, water condition, and nutrient condition, the other series was cultivated under shade that cut off incident light at 50% by a shade curtain. After 28 days, there were some distinct characteristics between the two. In the first series, Height of cucumber plants was higher, and color of leaf was greener.
Leaf reflectance measurement was carried out in a laboratory darkroom and a greenhouse chamber using the Geophysical Environment Research Field Portable Spectrometer (GER-2600). The GER-2600 records radiation in 650 channels over 350-2500 nm wavelength range. Leaf reflectance is a ratio of foliar radiance to white board's radiance. The reference radiance is defined as the radiance of a white-board made from BaSO
4. Light source used in the darkroom case was the standard halide lamp, and in the greenhouse chamber case the sunlight is the incident light.
The fresh leaf's light-photosynthetic rate curve was measured by the portable photosynthesis system (LI-cor, LI-6400). This instrument can control all environmental variables, such as CO
2, H
2O concentration of both input flux and output flux, leaf surface temperature and an incident light. Also the LI-6400 is able to measure stomatal conductance and evapotranspiration. In this experiment, we measured net photosynthetic rate with photon flux density ranging from 0 to 2000 [
mmol m
-2 s
-1].
3.2. Measurements of Chlorophyll and Nitrogen Concentration, Leaf Area, and Dry Mass of Leaf
The leaf area was determined prior to pigment analysis. Leaf area was determined as the result of the following steps. First step was to cut a fresh leaf, second step was to photocopy the fresh leaf by a copying machine, third step was to import monochrome image of leaf into PC, final step was to measure leaf area using the software-NIH Image, version. 1.62, distributed by National Institute of Health, USA. After photocopying, the cut leaf was dried for 48hrs in the drying machine at 80°C, and then the weight of the dried leaf was weighed.
The chlorophylls were extracted in 100 % DMS (N,N'-dimethylformamide) with a leaf disc of 1.0 cm diameter. Extractable efficiency of DMS is stronger than that of acetone
4.. The absorption of the extract at 663.8 nm and 646.8 nm were measured with a HITACHI, U-1000 spectrophotometer. Chlorophyll a, chlorophyll b and chlorophyll a+b concentration (
mg /ml) are calculated by Eq. (2):
where A
663.8 and A
646.8 are the absorbance at 663.8 nm and 646.8 nm wavelength, respectively.
The nitrogen concentration was determined using the NC-90A (Sumika Chemical Analysis Center), that can detect total nitrogen and carbon content in the sample. The calibration curve was obtained with the analytical standard acetanilide as a standard reagent. Total nitrogen content was determined as Eq. (3):
where area is the measured peak area of total nitrogen.
4. Results and Discussion
4.1. Correlation Between Biochemical Content and Model's Parameters
Correlation between the measured saturated Amax and chlorophyll a concentration per unit area was high with a coefficient of determination of R
2=0.91. There was linear correlation between saturated Amax and chlorophyll a concentration per unit area, as saturated Amax value increased in proportion to chlorophyll a concentration. But chlorophyll a concentration per unit mass didn't correlate with saturated Amax. Furthermore, both chlorophyll b concentration per unit area and its per unit mass didn't correlate with saturated Amax. R didn't correlated for chlorophyll a concentration per unit area or chlorophyll a concentration per unit mass. However, R had the best correlation with saturated Amax. Also total nitrogen concentration per unit area very much correlate with saturated Amax, R
2=0.91. But total nitrogen concentration per unit mass didn't correlate with saturated Amax. Figure 1 shows correlation between saturated Amax and chlorophylls per unit area (left diagram) and between saturated Amax and total nitrogen concentration per unit area (right diagram).
Figure 1. Correlation between saturated Amax and chlorophylls concentration per unit area (left) and total nitrogen concentration per unit area (right); g stands for chlorophyll a, = stands for chlorophyll b, 5stands for chlorophyll a+b, 1 stands for total nitrogen
4.2. Relationship between biochemical concentration and leaf reflectance
Figure 2 shows result of spectral reflectance curves of cucumber leaves in greenhouse chamber. The curves had common characteristics. There was the first positive peak called as the green peak driven by chlorophylls and carotenoids at 550 nm. There was a negative peak identified by absorption band of chlorophyll a at 680 nm. Moreover, these curve had red edge region and plateau region depending on its leaf structure over 700 nm. This reflectance curve is a typical reflectance curve of leaf.
Figure 2. A reflectance of cucumber's leaves in the greenhouse chamber
Three methods were examined to estimate chlorophyll a concentration per unit area from spectral reflectance; (1) a correlation between chlorophyll a concentration per unit area and spectral reflectance at one single wavelength; (2) a correlation between chlorophyll a concentration per unit area and ratio of spectral reflectance at two different wavelengths; (3) a correlation between chlorophyll a concentration per unit area and the first derivative spectral reflectance at one single wavelength.
