Powdery mildew is one of the most significant rubber tree diseases, with a substantial impact on the yield of natural rubber. This study aims to establish a detection approach that coupled continuous wavelet transform (CWT) and machine learning for the accurate assessment of powdery mildew severity in rubber trees. In this study, hyperspectral reflectance data (350–2500 nm) of healthy and powdery mildew-infected leaves were measured with a spectroradiometer in a laboratory. Subsequently, three types of wavelet features (WFs) were extracted using CWT. They were as follows: WFs dimensionally reduced by the principal component analysis (PCA) of significant wavelet energy coefficients (PCA-WFs); WFs extracted from the top 1% of the determination coefficient between wavelet energy coefficients and the powdery mildew disease class (1%R²-WFs); and all WFs at a single decomposition scale (SS-WFs). To assess the detection capability of the WFs, the three types of WFs were input into the random forest (RF), support vector machine (SVM), and back propagation neural network (BPNN), respectively. As a control, 13 optimal traditional spectral features (SFs) were extracted and combined with the same classification methods. The results revealed that the WF-based models all performed well and outperformed those based on SFs. The models constructed based on PCA-WFs had a higher accuracy and more stable performance than other models. The model combined PCA-WFs with RF exhibited the optimal performance among all models, with an overall accuracy (OA) of 92.0% and a kappa coefficient of 0.90. This study demonstrates the feasibility of combining CWT with machine learning in rubber tree powdery mildew detection.

Among the most damaging diseases of rubber trees is anthracnose caused by the genus Colletotrichum, which leads to significant economic losses. Nonetheless, the specific Colletotrichum spp. that infect rubber trees in Yunnan Province, an important natural rubber base in China, have not been extensively investigated. Here, we isolated 118 Colletotrichum strains from rubber tree leaves exhibiting anthracnose symptoms in multiple plantations in Yunnan. Based on comparisons of their phenotypic characteristics and internal transcribed spacer ribosomal DNA sequences, 80 representative strains were chosen for additional phylogenetic analysis based on eight loci (act, ApMat, cal, CHS-1, GAPDH, GS, his3, and tub2), and nine species were identified. Colletotrichum fructicola, C. siamense, and C. wanningense were found to be the dominant pathogens causing rubber tree anthracnose in Yunnan. C. karstii was common, whereas C. bannaense, C. brevisporum, C. jinpingense, C. mengdingense, and C. plurivorum were rare. Among these nine species, C. brevisporum and C. plurivorum are reported for the first time in China, and two species are new to the world: C. mengdingense sp. nov. in the C. acutatum species complex and C. jinpingense sp. nov. in the C. gloeosporioides species complex. Their pathogenicity was confirmed with Koch’s postulates by inoculating each species in vivo on rubber tree leaves. This study clarifies the geographic distribution of Colletotrichum spp. associated with anthracnose on rubber trees in representative locations of Yunnan, which is crucial for the implementation of quarantine measures.

Powdery mildew significantly impacts the yield of natural rubber by being one of the predominant diseases that affect rubber trees. Accurate, non-destructive recognition of powdery mildew in the early stage is essential for the cultivation management of rubber trees. The objective of this study is to establish a technique for the early detection of powdery mildew in rubber trees by combining spectral and physicochemical parameter features. At three field experiment sites and in the laboratory, a spectroradiometer and a hand-held optical leaf-clip meter were utilized, respectively, to measure the hyperspectral reflectance data (350–2500 nm) and physicochemical parameter data of both healthy and early-stage powdery-mildew-infected leaves. Initially, vegetation indices were extracted from hyperspectral reflectance data, and wavelet energy coefficients were obtained through continuous wavelet transform (CWT). Subsequently, significant vegetation indices (VIs) were selected using the ReliefF algorithm, and the optimal wavelengths (OWs) were chosen via competitive adaptive reweighted sampling. Principal component analysis was used for the dimensionality reduction of significant wavelet energy coefficients, resulting in wavelet features (WFs). To evaluate the detection capability of the aforementioned features, the three spectral features extracted above, along with their combinations with physicochemical parameter features (PFs) (VIs + PFs, OWs + PFs, WFs + PFs), were used to construct six classes of features. In turn, these features were input into support vector machine (SVM), random forest (RF), and logistic regression (LR), respectively, to build early detection models for powdery mildew in rubber trees. The results revealed that models based on WFs perform well, markedly outperforming those constructed using VIs and OWs as inputs. Moreover, models incorporating combined features surpass those relying on single features, with an overall accuracy (OA) improvement of over 1.9% and an increase in F1-Score of over 0.012. The model that combines WFs and PFs shows superior performance over all the other models, achieving OAs of 94.3%, 90.6%, and 93.4%, and F1-Scores of 0.952, 0.917, and 0.941 on SVM, RF, and LR, respectively. Compared to using WFs alone, the OAs improved by 1.9%, 2.8%, and 1.9%, and the F1-Scores increased by 0.017, 0.017, and 0.016, respectively. This study showcases the viability of early detection of powdery mildew in rubber trees.

Rubber tree (Hevea brasiliensis Muell. Arg.) is an important industrial crop of tropical areas for natural rubber production. In July 2015, a new leaf spot disease was found on Yunyan77-4 from a rubber tree plantation at Jinghong of Yunnan province. The infected leaves were reddish-brown spots surrounded by a yellow halo, and lesions of 0.2 to 3mm in diameter. Lesion tissues removed from the border between symptomatic and healthy tissue were surface sterilized in 75% ethanol and air-dried, plated on PDA plates, and incubated at 28°C alternating day/night cycles of light for 4 days. After 7 days, the colony was gray and produced black pigment in the medium. Microscopic examination showed the conidiophores were slightly flexuous or straight, septate and brown, 5.0 to 7.4um thick. Well-developed conidia were curved, spindle-shaped, or fusiform, septate conidia with 4 to 10 septa, measuring 39.2 to 82.2 × 9.9 to 15.6 μm(100 colonies were measured), with a slightly protuberant hilum, truncated. Morphological characteristics matched the descriptions of Bipolaris setariae(Sawada) (Zhang 2010).DNA of Bs01Jh1a isolate was extracted, and the ITS region of rDNA was amplified with primers ITS1/ITS4, and sequenced . The sequence of the region was deposited in GenBank (Accession No. KT805922). A BLAST search of the GenBank database revealed 100% similarity with Bipolaris setariae isolates （Accession No.KR183790.1）. partial sequences of Glyceraldeyde-3-phosphate dehydrogenase-like (Gpd) gene was also amplified and sequenced. The sequence of the region was deposited in GenBank (Accession No.KT982612 ) . BLASTn analysis revealed 99% sequence similarity with B. setariae isolate (Accession No. EF513206.1). Morphologic characters and sequence analysis of the ITS rDNA and Gdp gene confirmed that the pathogen was B. Setariae. The pathogenicity was tested with the method of field rubber tree inoculation which repeated for three times based on Koch's postulates. Copper-colored leaf stage rubber plants (the clone of rubber tree Yunyan77-4) were inoculated by spraying a conidia suspension (104 conidia/ml sterilized water) of Bs01Jh1a isolate. Rubber plants sprayed sterile distilled water were used as controls. After inoculation, the plants were covered with plastic bags to maintain high relative humidity. The plastic bags were removed 2 days after inoculation (dpi), and the plants were monitored daily for symptom development. Five days after inoculation, inoculated leaves showed typical symptoms similar to those described above. The pathogen was re-isolated from inoculated leaves. Morphological characteristics were identical to the original isolated fungus. In contrast, the control leaves did not show any symptoms. B. setariae is one of the most important pathogens of lawn grass, gramineous crops, and other plants. A disease of rubber tree caused by B. heveae had been reported in Malaysia in 1904 and China in 1951(Huang 2012). However, to our knowledge, no leaf spot disease caused by B. setariae has been recorded previously on Rubber in China.

“Moringa oleifera Lam (Moringaceae) is a highly valued plant, distributed in many countries of the tropics and subtropics. It has an impressive range of medicinal uses with high nutritional value” (Anwar et al. 2007). It was warm and foggy in late October 2014, nearly 60% of the Moringa plants (cv. PKM 1) showing typical symptoms of leaf spot characterized by circular brown spots about 5.5 mm in diameter were observed in an experimental plot at Jinghong, southwestern China. Severely affected leaves gradually turned yellow and defoliated. And light pink, grey fusiform lesions were observed on diseased stems.Symptomatic leaf samples were surface-sterilized in 75% ethanol for 30 s, 0.1% HgCl2 for 1 min, after which the samples were rinsed three times in sterile distilled water, dried on sterilized tissue paper, plated on potato dextrose agar (PDA), and incubated at room temperature with a natural photoperiod (about 11 h light: 13 h dark). Colletotrichum colonies were consistently isolated from affected leaves after 5 days of incubation. On PDA, colonies were initially pale, gradualy turning black after 3 days with smooth margins and no aerial mycelium. Conidia ranged from 15.7 to 22.1 μm long (mean 18.6 μm) × 2.5 to 4.3 wide (mean 3.3 μm) (n = 200). Conidia were hyaline, aseptate, smooth, curved with tapered ends and a truncated base. Chlamydospores were dark brown, thick-walled, globose to subglobose, in chains and clusters, 8–13 µm diam (n = 20). Based on morphological characteristics, Colletotrichum chlorophyti was suspected (Damm et al. 2009). For molecular identification, the internal transcribed spacer (ITS) region of ribosomal DNA and partial β-tubulin genes of a representative isolate LMhb1 were amplified using the ITS1/ITS4 (White et al. 1990) and Bt2a/Bt2b (Glass and Donaldson, 1995) primer pairs respectively. The gene sequences were deposited in GenBank (Accessions KR052084 for the ITS region; KU923670 for Bt2 region) and showed 99 % identity to C. chlorophyti isolate IMI103806 (KC790962 for ITS region and GU228188 for Bt2 region). A phylogenetic tree was generated via the neighbor-Joining method. This created tree places the isolate in the same group as C. chlorophyti. A Pathogenicity test was conducted on two-week-old Moringa plants cv. PKM 1. Ten Moringa plants planted in pots were sprayed with a spore suspension of 5.2 × 105 spores/ml sterile water. Five Moringa plants were sprayed with sterile water to serve as controls. All plants were placed in a high humidity plastic box at room temperature with a natural photoperiod (about 11 h light: 13 h dark). The cover of the plastic box was opened after 2 days post-inoculation. Leaf spot symptoms similar to the original disease were visible on Moringa plants 3 days after inoculation. No symptoms were observed on control plants. Fungal colonies morphologically identified as C. chlorophyti were reisolated from leaf lesions on inoculated Moringa plants, but not from asymptomatic leaves of control plants. To our knowledge, leaf spot caused by C. chlorophyti has not been described on Moringa oleifera (Liu 2013). This is the first report of C. chlorophyti on Moringa oleifera in China. It is also the first report of this fungal species affecting Moringa oleifera throughout the world.