Genotype identification and genetic diversity of Sugarcane yellow leaf virus in China


  • M. Q. Wang,

    1. Laboratory of Plant Virology, South China Agricultural University, Guangzhou 510642
    2. Sugarcane Research Institute, Yunnan Academy of Agricultural Sciences, Key Laboratory for Sugarcane Genetic Improvement of Yunnan Province, Kaiyuan 661600, China
    Search for more papers by this author
  • D. L. Xu,

    1. United States Department of Agriculture-Agricultural Research Service (USDA-ARS), National Germplasm Resources Laboratory, Beltsville, MD 20705, USA
    Search for more papers by this author
  • R. Li,

    1. United States Department of Agriculture-Agricultural Research Service (USDA-ARS), National Germplasm Resources Laboratory, Beltsville, MD 20705, USA
    Search for more papers by this author
  • G. H. Zhou

    Corresponding author
    1. Laboratory of Plant Virology, South China Agricultural University, Guangzhou 510642
    Search for more papers by this author



A survey of Sugarcane yellow leaf virus (SCYLV) in China showed infection rates of 1·1–42·4% in six different sugarcane growing provinces, with an average rate of 20·3% (129/634), and no infection in one province. The majority of the infected plants showed yellowing of the midrib on the abaxial surface of leaves. Genotypic diversity of the 129 SCYLV isolates was analysed by a reverse transcription–polymerase chain reaction–restriction fragment length polymorphism (RT-PCR-RFLP) assay and nucleotide sequencing. Sixty-six isolates from Guangxi, Yunnan, Hainan, Fujian and Guizhou provinces were of the genotype BRA, whereas another 63 isolates from Guangdong province consisted of five genotypes, BRA, CHN1, CUB, PER and a novel genotype, CHN2. Among these genotypes, BRA was the predominant one identified in 43 samples (68·3%). Mixed infection of BRA and PER was also found in two samples. Phylogenetic analysis based on a 1·3-kb nucleotide sequence (including ORFs 3–4 and partial ORF 5) supported the identification of the novel CHN2 genotype. CHN2 shared 91·9–97·8% identity with five other known genotypes and was most closely related to the CHN1 genotype (97·6–97·8% identities). The nearly complete genomic sequences of seven isolates representing genotypes BRA, CUB and PER were determined, and they shared nucleotide sequence identities of 97·2–97·5%, 98·4–98·9% and 98·4%, respectively, with previously reported sequences of the corresponding genotypes.


Sugarcane yellow leaf disease or sugarcane yellow leaf syndrome was first reported in Hawaii (Schenck, 1990) and has since been found in many sugarcane-producing countries (ElSayed et al., 2011). The disease is caused by Sugarcane yellow leaf virus (SCYLV), a member of the genus Polerovirus in the family Luteoviridae (Mayo, 2005). The virus is transmitted by vegetative propagation materials between areas and by aphids in fields, and its incidence has been shown to be high in some regions around the world (Anonymous, 1995; Scagliusi & Lockhart, 2000; Abu Ahmad et al., 2006b; Li et al., 2008; Behary-Paray et al., 2011). The virus can cause significant yield loss in susceptible cultivars or in commercial cultivars under drought stress (Vega et al., 1997; Viswanathan, 2002; Lehrer & Komor, 2008; Yan et al., 2009).

Sugarcane yellow leaf virus possesses a positive-sense single-stranded RNA genome containing six open reading frames (ORFs), ORF0–ORF5. The protein encoded by ORF0 is a suppressor of RNA silencing and is involved in pathogenesis (Mangwende et al., 2009). ORF1 and 2 are translated together and encode a multifunctional protein and RNA-dependent RNA polymerase (RdRp), respectively. ORF3 encodes a viral coat protein, and ORF4, harboured inside ORF3, encodes a viral movement protein. ORF5 is translated with ORF3 as a read-through protein via a translational read-through mechanism (Smith et al., 2000; Moonan & Mirkov, 2002). The read-through protein is a component of the virus particle and it might also be involved in the aphid transmission. ORF3 and 4 are highly conserved among different isolates (Abu Ahmad et al., 2006a; Wang & Zhou, 2010).

Sugarcane yellow leaf virus is genetically variable and has become one of the most important threats to sugarcane production worldwide (Anonymous, 1995; Izaguirre-Mayoral et al., 2002; Viswanathan, 2002; Rassaby et al., 2003; Xu et al., 2006; Lehrer et al., 2008, 2009; Yan et al., 2009). Six SCYLV genotypes of different geographic origins, BRA (Brazil), CHN1 (China), CUB (Cuba), IND (India), PER (Peru) and REU (Reunion) were identified based on analyses of genomic nucleotide sequences (Abu Ahmad et al., 2006a, 2007b; Viswanathan et al., 2008; Wang & Zhou, 2010).

Distribution of the SCYLV genotypes varies from area to area. At least three genotypes (BRA, PER and REU) were identified in Guadeloupe, Mauritius and Reunion Island, with the REU genotype predominating (Abu Ahmad et al., 2006a,b; Joomun & Dookun-Saumtally, 2010). The CUB genotype was predominant among four genotypes (BRA, CUB, IND and PER) found in India (Viswanathan et al., 2008). The Hawaiian isolates were closely related to the PER genotype (ElSayed et al., 2011).

China is one of the largest sugarcane-producing countries in the world, and sugarcane accounts for approximately 90% of sugar production in China. SCYLV infection rates of 0·5–10% were found in many sugarcane growing areas in southern China, and in some cultivars the rate was as high as 80% (Xu et al., 2006; Zhou et al., 2006). Lack of quarantine regulations for sugarcane pathogens may have resulted in widespread distribution of SCYLV in China as the virus was imported along with infected sugarcane materials from other sugarcane-growing countries (Li et al., 2008). The present study aimed to investigate the SCYLV genotypes and their geographical distribution in China. A reverse transcription–polymerase chain reaction–restriction fragment length polymorphism (RT-PCR-RFLP) assay was developed to differentiate the SCYLV genotypes.

Materials and methods

Sample collection, RNA extraction and RT-PCR detection

Leaves or setts of 201 samples with apparent symptoms of mild to intense yellowing of the midrib on the back of leaves and 433 symptomless samples were collected from hybrid sugarcane (Saccharum interspecific hybrid) and noble sugarcane (S. officinarum) plants in 20 representative locations of seven sugarcane production provinces (Guangxi, Yunnan, Guangdong, Hainan, Fujian, Jiangxi and Guizhou) in China in the summers of 2006–2009 (Table 1). The setts were planted in an insect-proof greenhouse, and new leaves were harvested 1 month after planting for sample preparation. A sugarcane plant generated by meristem tip culture and cultivated in an insect-proof greenhouse was used as a negative control. Total RNA was extracted from leaf tissues of both field and greenhouse samples using the HQ&Q RNA Reagent Kit (Anhui U-Gene Biotechnology) according to the manufacturer’s instructions. Presence of SCYLV was detected by RT-PCR using a pair of specific primers SCYLV-P1 (5′-AATCAGTGCACACATCCGAG-3′) and SCYLV-P2 (5′-GGAGCGTCGCCTACCTATT-3′) (Xu et al., 2006). RT-PCR was performed using the PrimeScript One Step RT-PCR Kit (Version 2; TaKaRa) according to the manufacturer’s instructions. The RT-PCR cycling conditions were: one cycle of 50°C for 30 min and 94°C for 2 min, and 35 cycles of 94°C for 30 s, 55°C for 30 s and 72°C for 1 min, followed by a final extension at 72°C for 5 min. The RT-PCR products were analysed by electrophoresis on a 1·2% agarose gel and observed under UV illumination after staining with ethidium bromide.

Table 1.   Detection of Sugarcane yellow leaf virus and identification of genotypes in field samples from China
ProvinceLocationNumberPositiveInfection rate (%)Genotype
  1. aMixed infection of genotypes BRA and PER.

  2. bMixed infection of genotypes BRA, CHN1 and CHN2.

YunnanKaiyuan241145·8 110000
Mile10330·0  30000
Yuxi12541·7  50000
Xishuangbanna20945·0  90000
GuizhouXingyi22627·3  60000
JiangxiGanzhou3800  00000
GuangxiLiuzhou2400  00000
Beihai15640·0  60000
Yizhou33100  30000
Nanning301240·0 120000
Laibin18844·4  80000
HainanYacheng9011·1  10000
FujianFuzhou5240·0  20000
GuangdongZhanjiang741723·0 11 + 2a04 + 2a00
Wengyuan1152521·7 180520
Huizhou2827·1  20000
Jiangmen13215·4  10100
Zengcheng22100  20000
Baiyun21523·8  50000
Guangzhou701014·3 4 + 2b3 + 2b102b
Total 63412920·311351322

RT-PCR-RFLP analysis

To analyse the genetic diversity among the SCYLV isolates, a pair of specific primers NF3596 (5′-GCGGCTTACCTGATATTCCTCA-3′) and NR4921 (5′-GCTCCGTCATTYACCTCGTC-3′, Y = C/T) were designed based on a conserved region of available SCYLV sequences for amplification of an approximately 1·3-kb DNA fragment containing ORF3, ORF4 and partial ORF5. The amplicons were then analysed by a triple digestion RFLP using BamHI, HindIII and XhoI for differentiation of the SCYLV genotypes. The 10-μL triplex digestion system contained 0·4 μL of each of BamHI (15 U μL−1), XhoI (10 U μL−1) and HindIII (15 U μL−1), 1·0 μL 10× K buffer, 5·0 μL RT-PCR product and 2·8 μL double-distilled water. After 4 h of incubation at 37°C, digested products were analysed on a 2.0% agarose gel. The genotype of each isolate was identified according to digest patterns determined by different genotype nucleotide sequences available in the GenBank databases. To confirm the identified genotypes, some RT-PCR products were directly sequenced. When a new pattern was observed in a sample, the RT-PCR was conducted again for the RFLP analysis. The RT-PCR product was also cloned into the pMD 18-T vector (TaKaRa) and sequenced (Invitrogen). Sequences were obtained in both directions from at least five clones of each ligation reaction. Seven isolates with undefined RT-PCR-RFLP patterns were selected for genomic sequence analysis according to a method previously described (Wang & Zhou, 2010).

Sequences obtained from different isolates were compared with those of the previously reported genotypes using DNAStar (version 5·01; DNASTAR Inc). Neighbour-joining phylogenetic trees were constructed based on sequences of two different regions on the SCYLV genome from the selected Chinese isolates and previously reported genotypes using mega 4·1 with 1000 bootstrap replicates. The first sequence of 1109 nucleotides (nt) contained ORF1 and partial ORF2, and another sequence of 1284 nt covered ORF3, ORF4 and partial ORF5.


SCYLV distribution in China

Products of the expected size (1·3 kb, representing ORF3, ORF4 and partial ORF5) were amplified from 129 of 634 sugarcane samples (20·3%) by RT-PCR. Ninety-six of the RT-PCR-positive samples were from the plants with symptoms, indicating that infection with the virus only has a certain degree of association with yellowing symptoms in fields. SCLYV was detected in six provinces, but not in Jiangxi (Fig. 1). Infection rates were 32·2% (29/90) for Guangxi, 42·4% (28/66) for Yunnan, 19·5% (63/323) for Guangdong, 1·1% (1/90) for Hainan, 40·0% (2/5) for Fujian and 27·3% (6/22) for Guizhou (Table 1). The highest infection rates were found in Yunnan and Guangxi, the two provinces that produce the most sugarcane. The high infection rate in Fujian was not representative because of a low sample number. Surprisingly, the infection rate was low in Hainan, the fourth-largest province in terms of sugarcane production in China. The virus was not found in Jiangxi province (central China).

Figure 1.

 Sugarcane-growing provinces (inline image) and occurrence of Sugarcane yellow leaf virus (shading) in China.

SCYLV genotype identification by RT-PCR-RFLP and sequencing

There are 10 expected RT-PCR-RFLP patterns of Sugarcane yellow leaf virus by triple digestion: six (BRA I, CHN1, PER I, PER II, PER IV and REU) determined by sequences available in GenBank and four (BRA II, PER III, CHN2 and CUB) determined by sequences obtained in this study. Out of these 10 patterns, two represent genotype BRA, four represent genotype PER, and one each represents the genotypes CHN1, REU, CUB and a new genotype CHN2 (Fig. 2a). All 64 SCYLV isolates from Guangxi, Yunnan, Hainan, Fujian and Guizhou showed the same RFLP pattern as that of BRA I shown in Fig. 2a, indicating that genetic diversity is extremely low among these isolates. Sixty-three samples from Guangdong province, however, had several different RFLP patterns (Fig. 2b–d and not shown). Forty-one of the Guangdong isolates displayed the BRA I pattern (Fig. 2b, lanes 3–4, 6, 8–10; 2c, lanes 2, 4–5; 2d, lanes 1, 3, 6–7; and not shown), three isolates displayed the CHN1 pattern (Fig. 2c, lanes 1, 3; 2d, lane 4), and 19 isolates (Fig. 2b, lanes 1–2, 11–12; 2c, lanes 7–8; 2d, lanes 5, 8; and not shown) displayed patterns distinct from those determined from the known genotypes including BRA I, CHN1, PER I, PER II, PER IV (also known as HAW) and REU (Fig. 2a). In order to identify the genotypes of these uncertain samples, their RT-PCR products were cloned for recombinant PCR-RFLP analysis. A 1326-bp PCR product using recombinant plasmid as the template was then obtained from at least five clones for each ligation reaction for RFLP analysis and subsequent sequencing. The analysis showed that 15 of these samples were infected by a single genotype, whereas co-infection with two (Fig. 3a) and three (Fig. 3b) genotypes was found in each of two samples, respectively. Sequence comparison showed that the 15 isolates of single infection belonged to three known genotypes, six (Fig. 2b, lanes 13–14 and not shown) to BRA (pattern BRA II in Fig. 2a), five (Fig. 2b, lanes 11–12 and not shown) to PER (Fig. 2a, pattern PER III) and four (Fig. 2c, lanes 7–8 and not shown) to CUB (Fig. 2a, pattern CUB), while the two dual-infected samples (Fig. 2b, lanes 1–2) possessed BRA and PER genotypes (Fig. 3a; Fig. 2a, pattern BRA II and pattern PER II). The two triple-infected samples (Fig. 2d, lanes 5 and 8) possessed the two known genotypes BRA (Fig. 2a, pattern BRA I) and CHN1 (Fig. 2a, pattern CHN1) and a distinct novel genotype, CHN2 (Fig. 2a, pattern CHN2) (Fig. 3b). CHN2 was subsequently determined to be 91·9–97·8% identical to the other five known genotypes. Phylogenetic analysis using the 1·3-kb nucleotide sequences of 45 different SCYLV isolates also clearly placed CHN2 into a distinct phylogenetic branch that represents a new genotype most closely related to the CHN1 and CUB genotypes (Fig. 4b).

Figure 2.

 Genotype identification of Sugarcane yellow leaf virus in plant samples by an RT-PCR-RFLP assay. (a) Expected RT-PCR-RFLP patterns of six Sugarcane yellow leaf virus genotypes: BRA, PER, CHN1, REU, CUB and a new genotype, CHN2. The patterns for BRA I, CHN1, PER I, PER II, PER IV and REU were determined by sequences available in the GenBank database, and those for BRA II, PER III, CHN2 and CUB were determined by sequences obtained in this study. (b) Samples from west Guangdong province. Lanes 1 and 2, mixed infection of genotypes BRA II and PER II; lanes 3, 4, 6, 8, 9 and 10, genotype BRA I; lanes 5 and 7, genotype PER II; lanes 11 and 12, genotype PER III (confirmed by sequencing of isolate CHN-GD-ZJ15, HQ245320); lanes 13 and 14, genotype BRA II. (c) Samples from central Guangdong province. Lanes 1 and 3, genotype CHN1; lanes 2, 4 and 5, genotype BRA I; lane 6, genotype PER II; lanes 7 and 8, genotype CUB (confirmed by sequencing of isolate CHN-GD-WY20, HQ245318). (d) Samples from north Guangdong province. Lanes 1, 3, 6 and 7, genotype BRA I; lane 2, genotype PER II; lane 4, genotype CHN1; lanes 5 and 8, mixed infection of genotypes BRA I, CHN1 and a new genotype tentatively named CHN2 (confirmed by sequencing of isolate CHN-GD-GZ10-1, HQ245324).

Figure 3.

 Mixed infections of two to three genotypes of Sugarcane yellow leaf virus in some Guangdong samples. The presence of different genotypes was confirmed by a recombinant plasmid PCR-RFLP assay. (a) Mixed infection with genotypes BRA II and PER II in a sample with a distinct RT-PCR-RFLP pattern (Fig. 2b; lane 1). Lane 1, genotype BRA II; lanes 2–7, genotype PER II. (b) Mixed infection with genotypes BRA, CHN1 and CHN2 in a sample with another distinct pattern (Fig. 2d; lane 5). Lanes 1, 2, 4, 6, and 8, genotype CHN2; lanes 3, 5, 7 and 9, genotype BRA I; lane 10, genotype CHN1.

Figure 4.

 Neighbour-joining phylogenetic trees of Sugarcane yellow leaf virus based on the nucleotide sequences of different isolates characterized from the present and previous studies. (a) ORF1 and partial ORF2 (1109 nt), and (b) ORF3, ORF4 and partial ORF5 (1284 nt). The isolates from this study were marked with an asterisk (*). The novel genotype, CHN2, is indicated by a solid triangle.

To further investigate the occurrence of the four genotypes found in this study, nearly complete genomic sequences were obtained from seven selected isolates. Sequence analysis confirmed the existence of PER (three isolates), BRA (three isolates) and CUB (one isolate) genotypes in the samples analysed in this study (data not shown). The Chinese isolates of the genotypes PER, BRA and CUB shared nucleotide sequence identities of 97·2–97·5%, 98·4–98·9% and 98·4% with the corresponding genotypes, respectively. A phylogenetic tree based on the 1109-nt sequences of ORF1 and partial ORF2 supported the genotype identification (Fig. 4a).


Since SCYLV was first identified in Hawaii, USA (Schenck, 1990), it has been found worldwide in sugarcane-growing areas and is now an economically important pathogen (ElSayed et al., 2011). In mainland China, the virus was first detected in Guangxi province in 2002 (Wang et al., 2003), and later in germplasm and field samples in Guangdong province (Xu et al., 2005, 2006). Zhou et al. (2006) reported that the virus occurred in two sugarcane growing provinces (Guangdong and Guangxi) in south China. The present study confirms that SCYLV is widely distributed in the major sugarcane production areas in southern China (Fig. 1). The high infection rates of SCYLV in the three provinces that produce the most sugarcane (Guangxi, Yunnan and Guangdong) are representative because the samples were collected from multiple locations in each province, whereas the extremely low infection rate (1/90) found in Hainan might be unrepresentative as a result of there only being one sampling location. The virus is absent from central (Table 1) and western China (data not shown).

Sugarcane yellow leaf virus is not listed as a quarantine pathogen in China and, therefore, it is most likely that primary infection of SCYLV in southern China was introduced from abroad. Some sugarcane materials in southern China were originally acquired as germplasm resources or commercial cultivars from the mainland USA (CP series), Brazil, Cuba, Australia (Q series), France, Thailand and Taiwan (F series and ROC series). Infected setts might serve as the resource of long-distance transmission of SCYLV, and wide distribution of sugarcane woolly aphid (Ceratovacuna lanigera) in these areas then spreads the virus in the fields (Li et al., 2008).

Two reasons may account for the presence of SCYLV only in southern China but not in the western or central parts of the country. First, sugarcane farmers in southern China mainly grow hybrid sugarcanes (Saccharum interspecific hybrids for sugar milling) and frequent exchange of seedlings may have resulted in quick and widespread distribution of the virus. The higher infection rates of SCYLV in these areas might also be related to active introduction and exchange of germplasm resources. In western and central China, noble sugarcane (S. officinarum) is produced mainly for fruit cane (chewing cane), and the crop is usually propagated from cutting setts year after year by farmers themselves, which lowers the possibility of virus transmission between different locations. Secondly, the highly efficient aphid vector, C. lanigera, is present in large numbers in southern China but is absent in western and central China (Li et al., 2008; Lu et al., 2008). However, SCYLV might still pose a great threat to sugarcane in the western and central areas if the virus or vector is introduced there.

Identification of six genotypes, including CHN2 from this study, indicates that genetic diversity of SCYLV is relatively high. The structures of the SCYLV populations vary geographically (Abu Ahmad et al., 2006a,b; Viswanathan et al., 2008; Joomun & Dookun-Saumtally, 2010; ElSayed et al., 2011). The present study found that the genotype BRA accounted for 84·5% of infections, and, with the exception of Guangdong, it was the only genotype found in the other five provinces (Table 1). Among the five genotypes found in Guangdong, BRA was still predominant, being identified in 68·3% of 63 infected samples. These data suggest that BRA was the first genotype which invaded sugarcane areas in China and had good fitness for these ecological environments. A number of studies also revealed that BRA was the most widespread genotype worldwide and was considered the one with the greatest fitness (Abu Ahmad et al., 2006a,b; Viswanathan et al., 2008; ElSayed et al., 2011).

It is no surprise that the PER and CUB genotypes were found in Guangdong because it is the most economically active area of China and has imported the largest amount of sugarcane materials from other countries. However, it is noteworthy that two new genotypes, CHN1 (Wang & Zhou, 2010) and CHN2, have emerged in Guangdong. The clustering of these genotypes with the CUB isolates found in India (Fig. 4a) and Guangdong (Fig. 4b) indicates a close relationship with the genotype CUB. They might have been introduced with germplasm materials from other countries. However, it is possible that they have evolved from local SCYLV isolates because the C. lanigera aphid in Guangdong is a vector distinct from those in other parts of the world, and insect vectors play an important role in the evolution of plant viruses (Power, 2000).

Abu Ahmad et al. (2007a) observed that viral infectivity and pathogenicity differ among the SCYLV genotypes. Thus, a genotype-differentiating method is needed for sugarcane quarantine and virus-resistance breeding programmes. Although sequence analysis is the most accurate method of identifying virus genotypes, it is too expensive and time-consuming when a large amount of samples are analysed. Abu Ahmad et al. (2006b) designed genotype-specific RT-PCR primers to differentiate the genotypes BRA, PER, REU and CUB. However, the genotypes BRA and PER were detected together as BRA-PER because the two genotypes are very similar at the nucleotide sequence level. Furthermore, these primers were designed according to the sequences of a limited number of isolates, and as many molecular variants might belong to a single genotype, the RT-PCR assays using these primers may be inaccurate. The in silico PCR analysis used here with the primers described by Abu Ahmad et al. (2006b) via a blast GenBank homology search confirmed its inaccuracy (data not shown). The present study established a RT-PCR-RFLP method for differentiation of SCYLV genotypes. This assay is more informative because the method employs virus-specific primers based on a conserved region of the SCYLV genome, allowing the amplification of the target fragment from a broad range of variants, followed by RFLP analysis using three restriction enzymes (BamHI, HindIII and XhoI). It allows the differentiation of BRA and PER. This method has been used to successfully differentiate not only the previously reported genotypes BRA and CHN1 according to their expected digestion patterns, but also the distinct patterns (BRA II, PER II, PER III and CUB) of the known genotypes and a new genotype (CHN2) (Fig. 2). This assay is useful for the identification of known SCYLV genotypes as well as new genotypes when combined with sequence analysis.

Genetic divergences among most of these genotypes are larger than 4%, with the exception of BRA and PER. The genome nucleotide sequences of BRA and PER share high identity (98·5%) (Abu Ahmad et al., 2006a; Wang & Zhou, 2010). Phylogenetic analyses using sequences of different regions on the SCYLV genomes grouped certain isolates into both genotypes (Abu Ahmad et al., 2006a; Viswanathan et al., 2008; Wang & Zhou, 2010; ElSayed et al., 2011; Singh et al., 2011). Therefore, the two genotypes have been considered as BRA-PER (Abu Ahmad et al., 2006b; Viswanathan et al., 2008; Joomun & Dookun-Saumtally, 2010). However, ElSayed et al. (2011) grouped three Hawaiian isolates with PER as one HAW-PER genotype. Therefore, a standard criterion is needed to group the SCYLV genotypes.

To determine the association of SCYLV infection with yellowing symptoms, 201 samples with symptoms were tested. However, only 96 of these samples were found to be infected with the virus, indicating that not all the plants with symptoms were infected by SCYLV. Most of the SCYLV-free plants with yellowing were collected from highlands or fields without irrigation. The infection of SCYLV and expression of symptoms were also related to sugarcane cultivar. This study showed that the infection rate of SCYLV on Yuetang93–159 (hybrid sugarcane) was high, with very obvious symptoms in fields, regardless of irrigation, but virus infection of Yuetang00–236, ROC16, ROC25 (all hybrids) was only associated with mild symptoms. It was also found that SCYLV infected noble cane, such as cv. Badila. These results support the idea that the most prominent yellow leaf symptom (yellowing of the midrib) may be related to other biotic or abiotic factors (Lockhart & Cronjé, 2000). Cronjéet al. (1998) reported that a phytoplasma might cause SCYLD-like symptoms in sugarcane in Africa. The same phytoplasma was also identified in Mauritius, Cuba and India (Aljanabi et al., 2001; Arocha et al., 2005, 2006; Gaur et al., 2008). Further analyses are needed to understand whether the sugarcane phytoplasma occurs in sugarcane found in China, and whether the yellow leaf symptoms are related to the phytoplasma or other stress factors.


We thank Dr W. K. Shen, Guangzhou Sugarcane Industry Research Institute, for help in sample collection. This study was funded by the Guandong Natural Science Foundation (5006669) and the Guangdong Provincial Department of Science and Technology (2008B020100004).