Transmittance Properties of Healthy and Infected Coffee Robusta Leaves with Coffee Leaf Miner (CLM) Pests ^†

Abstract

Coffee Robusta (Coffea canephora) increased its total production by 73.5% during the first quarter of 2023. In this study, twenty (20) samples each of healthy and infected coffee leaves were measured for their transmittance properties in the UV-Vis and NIR regions. Coffee Leaf Miner (CLM)-infected leaves were identified based on translucent patches on the plant foliage. The results showed that a healthy coffee leaf has a mean transmittance of 41.53 μW for the NIR region, while for the infected leaves, the mean transmittance is 47.06 μW. Healthy coffee Robusta leaves showed significant differences in their transmittance properties compared to infected coffee Robusta leaves in the UV (r = −0.15, p = 0.021, F = 5.8, t = −0.286), visible (r = −0.15, p = 0.018, F = 6.11, t = −2.88), and NIR (r = −0.14, p = 0.027, F = 5.28, t = −2.99) regions. A CLM index was introduced based on the intensity ratio of green and red wavelengths. I535/575 showed positive correlation with the estimated chlorophyll-a concentration for healthy (r = 0.94, p = 0.227) and infected (r = 0.56, p = 0.622) leaves. This method leads to the development of portable sensors for the early detection of CLM pests in plants.

1. Introduction

Coffee Robusta (Coffea canephora) is one of the four varieties of coffee cultivated in the Philippines. The others are arabica, liberica, and excelsa [1]. Based on the Philippine Statistics Authority (PSA) 2015 data, there are about 276,000 coffee farms in the country, which comprise 77.4 million trees [2]. Among the three coffee varieties in the Philippines, Robusta is the most commonly planted, which accounted for 73.5% percent of the total production in 2023. Efforts have been made to revive coffee farming and improve the quality of Philippine coffee [1,3].

Over the years, the coffee industry in the Philippines has faced several challenges, including Coffee Leaf Miner (CLM) (Leucoptera coffeella) and other diseases periodically affecting coffee crops, leading to reduced yields and quality, and outdated farming practices, lack of access to modern technology, and inadequate infrastructure, which have contributed to low coffee yields [4,5,6]. CLM moth larvae are the primary causal agent of crop damage to coffee plantations [5]. The damage results from its larvae that feed on coffee leaves, which reduces fruit production [6]. The larvae feed on the mesophyll of the coffee tree leaves and create mines [7,8,9].

Having a method for the detection of plant diseases is very important to lessen the possible impact on its production, and this can be achieved using various methods. The analysis of optical properties of plants can be carried out in two ways, destructive and non-destructive [10,11,12,13]. Destructive processes include microscopy and spectrophotometry that require the actual sample plant to undergo different processes inside the lab [13]. On the other hand, the non-destructive approach incorporates the use of a portable sensor to determine the optical properties, which is easier to carry out and less time consuming [10,11,12]. Such methods stated above have been successfully used for the early detection and diagnosis for plant health [14,15].

In this study, we aim to analyze the optical properties of healthy and infected coffee Robusta leaves using a Thorlabs PM400 Optical Power Meter. Specifically, we identify the transmittance in the UV-Vis-NIR spectrum to quantitatively determine the signature responses and characterize healthy and infected coffee Robusta leaves. Statistical analyses such as t-tests and Pearson’s r correlation were used to determine differences in the healthy and infected coffee Robusta leaves. Lastly, the results from this study provide preliminary data for the development of non-destructive portable sensors in CLM detection.

2. Materials and Methods

2.1. Plant Identification

Coffee Robusta plants were verified and collected from the Department of Agriculture-Bureau of Plant Industry (DA-BPI), Manila, Philippines. The plants were placed inside the laboratory to acclimatize to ambient conditions. Twenty (20) leaf samples of each of the 6-month-old healthy and CLM-infected plants were identified based on translucent patches on the plant foliage. CLM pests reduce the photosynthetic leaf surface, where brown spots are visible on the infected leaves. All leaf samples investigated were marked with regions of interest (ROIs), as shown in Figure 1a. The measurements were performed in a dark room to remove light noise.

2.2. Transmittance Properties

Transmittance spectra of healthy and infected leaves were measured with an optical power meter (Thorlabs PM400, Newton, NJ, USA) connected with S120VC standard photodiode power sensor. Leaf transmittance was analyzed through the UV-Vis and NIR spectrum wavelengths at the 200–1100 nm wavelength interval. Healthy and infected leaves were placed individually at 90° by illuminating the adaxial surface about 8 inches to the High-Power Xenon Light source (Ocean Insight HPX-2000, Orlando, FL, USA). Light passed through the leaf surface to the power sensor and the transmittance reading was reflected to the power meter, as shown in Figure 1b.

2.3. Statistical Analyses

The ratio between the normalized intensity values from green (535 nm) and red (575 nm) were calculated. Pearson’s r correlation was used to compare the estimated chlorophyll content and the ratio for both healthy and infected leaves. T-tests for independent samples were computed for the transmitted power for the coffee leaves in the UV-Vis-NIR regions.

3. Results and Discussion

3.1. Microscopic Characterization of Coffee Robusta Leaves

Under microscopic observations, CLM pests only invaded the upper dermis (Figure 2). The epidermal cells and the cell walls are destroyed due to attacks from CLM pests [16]. To further analyze the morphological conditions of the coffee Robusta leaves, transmission or scanning electron microscopies are recommended.

In this study, we introduce a CLM index (I535/575) by measuring the ratio between intensities from 535 nm (green) and 575 nm (red), as shown in

Table 1. Normalized intensity of RGB data from coffee Robusta leaves (n = 20, p = 0.05).

RGB Data *	Normalized Intensity (a.u.)		I535/575	Chlorophyll-a Concentration (mg/L)
	Red	Green	Range (Mean ± St.Dev.)
Healthy	0.27–0.96	0.31–0.98	0.81–0.97 (0.92)	0.34
Infected	0.08–0.16	0.30–0.85	0.26–0.39 (0.33)	0.12

* No significant change in the blue wavelength for both healthy and infected leaves.

. The normalized intensities in the red and green region varies for healthy and infected leaves. The CLM index showed positive correlation with the estimated chlorophyll-a concentration for healthy (r = 0.94, p = 0.227) and infected (r = 0.56, p = 0.622) leaves. This study showed a similar high correlation between chlorophyll content and RGB values, as discussed in different articles in the literature [17,18].

3.2. Transmittance Measurements

Table 2. Transmittance measurements (μW) of healthy and infected coffee Robusta plants (n = 20, p = 0.05).

Robusta Leaves	Healthy			Infected
Spectrum	UV	Visible	Near-IR	UV	Visible	Near-IR
Range	71.15–95.44	42.65–75.9	33.3–47.56	75.95–139.65	59.25–109	38.03–68.18
Mean ± St.Dev.	84.35 ± 6.95	65.96 ± 5.34	41.53 ± 3.26	95.06 ± 15.21	74.19 ± 11.91	47.06 ± 7.16
Variance	48.25	28.55	10.62	231.2	141.77	51.21
Skew	0.05	0.07	0.06	1.34	1.34	1.31
Kurtosis	−0.88	−0.78	−0.66	2.58	2.54	2.76

summarizes the transmittance measurements in the healthy and infected coffee Robusta leaves from the UV-Vis-NIR spectrum. The power measured using the transmittance of healthy and infected coffee Robusta leaves varies from the UV, visible, and near infrared spectrum. The spectral response of infected leaves is higher compared to healthy leaves. This is due to the breakdown of the upper dermis and cell walls caused by CLM [19]. These results may affect the phenotype of leaves but also the photosynthetic ability of leaves to produce chlorophyll and other nutrients. The response of the near-infrared spectrum is significantly different from healthy and infected leaves, which can be used to understand the early detection of pests, physiological disorders, and ozone damage [20,21,22].

The relationships between healthy and infected coffee Robusta leaves were statistically measured using t-tests. Healthy coffee Robusta leaves showed significant differences in their transmittance properties compared to infected coffee Robusta leaves in the UV (r = −0.15, p = 0.021, F = 5.8, t = −0.286), visible (r = −0.15, p = 0.018, F = 6.11, t = −2.88), and NIR (r = −0.14, p = 0.027, F = 5.28, t = −2.99) regions. This study is a good reference for the development of non-destructive techniques in agriculture and environmental monitoring.