Wheat Mingxian 169, which is highly susceptible to P. striiformis f. sp. tritici, were grown in plastic pots (7× 7× 8 cm) in a rust-free growth chamber with temperature of (16± 3) ℃ and photoperiod of 16 h of light and 8 h of darkness^[15] in the State Key Laboratory of Crop Stress Biology for Arid Areas and College of Plant Protection, Northwest A& F University, China. Previous investigations using infrared thermography have demonstrated that environmental changes (such as temperature, humidity, illumination, and airflow speed) greatly affect leaf temperature^[16]. The daily inoculation method was selected to reduce the effect of environmental changes on thermography. Four pots (15~20 wheat plants per pot) were inoculated every day for 16 days in the following way: Firstly, wheat leaves were gently rubbed with a clean moistened finger to remove the waxy layer from the leaf surface at 12 days from sowing. Next, the urediniospores of the pathogen on naturally infected leaves (from living tissues) were collected and used to artificially infect the first leaves of seedlings. Finally, the inoculated plants were covered with wet polyethylene bags to maintain 100% relative humidity and kept for 24 h at 10 ℃ in a dark chamber. Immediately after incubation, plants were transferred to a clean growth chamber set to a diurnal cycle of 16 ℃ for the 16-h light period and 13 ℃ for the 8-h dark period. Plants were assessed daily for stripe rust development by visual rating. On the 16th day, a total of 64 pots of wheat samples, with different incubation days and disease progression, were obtained. Meanwhile, 4 pots of healthy wheat seedlings without inoculation were selected as control.

Thermal images of wheat plants were captured using an un-cooled focal plane array infra-red camera (A300, FLIR, USA) with a spectral sensitivity of 7.5 to 13.5 μ m, a geometric resolution of 1.36 mrad (320× 240 pixels) and a 25° × 18.8° field of view lens. The thermal resolution is 0.05 ℃, and the accuracy of absolute temperature measurement is less than ± 2 ℃. Throughout the entire experiment, the emissivity was set to 0.99, which is within a range for plant leaves quoted by Lindenthal M, et al.^[17]. The assorted BM-IR analysis software was provided with an English interface and windows platform with a temperature analysis function for point, line and region. Colour images of the leaves were taken with commercial digital cameras (Nikon, Tokyo, Japan). At the same time, the images of cell microstructure of the wheat samples were obtained with a laboratory electron microscope (Olympus-BX53, Olympus Corp., Tokyo, Japan).

On the 16th day after inoculation, a total of 64 pots of wheat samples with different incubation days, and 4 pots of healthy wheat samples without inoculation, were removed from the artificial climate chamber and placed in the Key Laboratory of Agricultural Internet of things, Northwest Agriculture & Forestry University for an hour (room temperature of (28± 1) ℃ and relative humidity of 50%) to balance the wheat plant temperature and reduce the experimental error caused by external factors. Firstly, a digital camera was used to capture the visible light pictures of both healthy and infected leaves. Then, a thermal imager was installed at a distance of 40~50 cm from the samples in the horizontal direction to obtain the infrared thermal images under natural light. The dates of each incubation day were repeated 4 times in order to ensure the stability of data. The data were recorded in form of radiometric images and analyzed using BM-IR analysis software. To obtain the stable leag temperature values, eight irregular areas of interest (ROI) with 2 000~2 500 pixelsfor each imagewere selected (Figure 1). Since each pixel reveals individual temperature signals, the temperature of wheat leaf in each day is an average value of more than 8 000 pixel. Finally, the data were conducted statistically using SPSS 22 (SPSS Inc., Chicago, IL, USA) and analyzed by standard analysis of variance. Where a significant treatment effect was demonstrated by the ANOVA (p≤ 0.05), significant differences between means were determined with Tukey’ s test.

Fig.1

	Figure Option ViewDownloadNew Window
	Fig.1 Screenshot of BM-IR analysis software with eight areas of interest (ROI). The air temperature was recorded from a ROI measurement cursor (4 000 pixels) in the blank area

The development of inoculated wheat plants was visually investigated every day under the experimental conditions. During the first 9 days of the experiment, there was no statistically significant difference between the treated and control plants except for changes in shape and size of plants (Figure 2: A1, B1, C1). Irregular yellowish green spots on the leaf surface appeared as the first characteristic symptom of stripe rust at 10 dpi, which then developed into small spore piles at 12 dpi (Figure 2, D1) and extended to the full inoculated blades that appeared curled and atrophied at 15 dpi (Figure 2, E1).

Meanwhile, the thermal infrared images of the inoculated wheat plants were collected for comparison. The infrared thermal images of the inoculated plants gradually varied throughout the whole incubation period. An initial thermal response of wheat leaf tissue to Puccinia striiformis infection was observed as dark yellow coloration at6 dpi (Figure 2, B2). The color of the infected area deepened, and the area expanded with an increasing number of days after inoculation. The entire blade became purple at 15 dpi (Figure 2, E2). Since the inoculated leaves were only the first leaf of each plant, the newly developed leaves were not infected and remained green. Hence, one can see that the infected wheat plants without symptomatic appearance can be distinguished from healthy ones at 6 dpi using infrared thermal imagery, while detection occurs at 10 dpi using only the naked eye. Stripe rust infection could be detected 4 days earlier by using thermography instead of visual inspection.

Fig.2

	Figure Option ViewDownloadNew Window
	Fig.2 Comparison between photographs and thermographs of attached wheat seedlings from suripe rust past 3, 6, 9, 12 and 15 days of incubation First thermal effects of stripe rust became detectable at 6 dpi; similarly, areas of low leaf temperature preceded visible symptoms in later stages of leaf colonization. Left: RGB reflectance, right: thermo-images

The average temperature changes in healthy and infected wheat blades for 16 consecutive days are shown in Figure 3. It can be seen that the average temperature of healthy blades fluctuated slightly during the entire detection period. In contrast, the average temperature of the inoculated blades decreased noticeably. With the increasing of days after inoculation, the divergence of the average temperature between inoculated blades and healthy ones gradually increased. The average temperature of the inoculated blades was 2.22 ℃ lower than that of healthy ones at 16 dpi.

Fig.3

	Figure Option ViewDownloadNew Window
	Fig.3 Effect of Puccinia striiformis infection on average leaf temperature of wheat plants during 16 days past inoculation. Asterisks indicate a significant difference at p< 0.05 level of average temperature for healthy wheat plants and for that of wheat plants in the incubation period on different day after inoculation with Puccinia striiformis. Error bars indicate standard deviations

The maximum temperature difference (MTD) between the maximum and minimum temperatures of total leaf areas during the whole detected period are shown in Figure 4. The MTD of non-inoculated leaves was almost constant during the experiment and stayed within a range of 0.5 ℃. The temperature distribution was still indistinguishable from that of inoculated leaves at 3 dpi. However, at 4 dpi, the MTD of the infected leaves was significantly higher (+0.3 ℃) than that of healthy leaves. With the development of stripe rust symptoms, MTD increased by more than 2.3 ℃ at 16 dpi, which was approximately 1.8 ℃ higher than that of non-inoculated control leaves. The result of the variance analysis showed that the average temperature and MTD of the inoculated wheat plants and of the healthy plants appeared to have a significant difference at 16 dpi (p< 0.05): the average temperature of the inoculated plants was 2.22 ℃ lower than that of healthy plants, and MTD was 1.8 ℃ larger than that of healthy plants. Therefore, the infection of Puccinia striiformis could be detected at 4 dpi using temperature data analysis, 6 days earlier than the visual rating method.

Fig.4

	Figure Option ViewDownloadNew Window
	Fig.4 Effect of Puccinia striiformis f. sp. tritici infection on maximum temperature difference (MTD) within the entire wheat leaf during the consecutive 16 dayspast inoculation. Significant differences between treatments are indicated by different letters following ANOVA

Three samples were analyzed with the microscopic images in the experiment, which were as follows: the infected blades at 8 dpi, the disease blades at 15 dpi and healthy blades as control. Figure 5 depicted the reaction of wheat tissue during different stages of stripe rust development in microscopic sections of primary infection sites. It could be seen that the healthy ones had a complete structure of crosscut sections, a compact structure and plentiful epidermal cells, and the chloroplasts were attached around the cell wall (Figure 5: A1, A2, A3). When the blades were infected, the Puccinia striiformis would break through the protection of the epidermal cells to destroy the internal structure. Mesophyll cells were modified, and the number of chloroplasts was reduced considerably at 8 dpi (Figure 5: B1, B2, B3). For the diseased blades (15 dpi), whether the epidermis cells, mesophyll cells or chloroplast were seriously damaged. The pathogens invaded each tissue of the blade and accumulated to create a large number of summer spore banks or spore beds (Figure 5: C1, C2, C3).

Fig.5

	Figure Option ViewDownloadNew Window
	Fig.5 Sections of wheat leaf tissue infected with Puccinia striiformis f. sp. Tritici during different stages of pathogen development A: healthy tissue; B: tissue at the infection site at 8 dpi without macroscopic symptoms but with the epidermis layer destructed; C: chlorotic tissue with hyphae of Puccinia striiformis.