**1. Introduction**

Bacterial panicle blight (BPB), caused primarily by *Burkholderia glumae*, has become a threat to rice production globally. BPB has the potential to cause significant losses in grain yield and milling quality in epidemic years. The disease causes several types of damage, including seedling blight, sheath rot, floret sterility, grains not filling or aborted, and milling quality reduction, resulting in a reduction of yield by up to 75% [1–4]. In Japan, BPB has become one of the major rice diseases. Severe outbreaks of this disease occurred on more than 69,000 ha in 2013 and 30,000 ha in 2015 [5, 6]. In the USA, BPB has recently become as one of the most important diseases in rice in terms of economic importance. A survey found that the disease was present in approximately 60% of Louisiana rice fields [7]. In the Southern USA, significant yield losses from BPB were reported in 1995, 1996, 1998, 2000, 2010, and 2011 [1, 8–11]. In Louisiana, yield losses for severely infected fields reached 40% in

1995 and 1998 [1, 8]. In Arkansas, BPB was so severe in 2010 that yield losses were estimated at 50% in susceptible cultivars [9]. In Texas, the outbreaks of BPB resulted in an estimate of 10–20% yield loss in the Texas Rice Belt in 2010 [10, 11]. Outbreaks of this disease also occurred in rice under organic production systems in 2010 in Texas [11]. In the disease-yield loss field study, we found BPB was highly destructive and could cause yield losses ranging from 1 to 59% (83–4883 kg/ha), with yield loss increasing approximately 5% (455 kg/ha) for every unit increase in BPB severity on the rating scale of 0–9 [12]. Based on annual rice production in the Mid-South USA in 2003–2013, it is estimated that BPB caused \$61 million USD of damage that would feed 1.1 million people annually (Aaron Shew, personal communication).

Effective management of BPB is critical to minimizing the damage caused by the disease and maximizing production returns. However, limited options for management of the disease are available currently. No single genes or quantitative trait loci (QTLs) for complete resistance to BPB have been found so far [13, 14]. Only a few rice cultivars with partial resistance are available for commercial use. No chemical control options are available in the USA although oxolinic acid has been used as a major control measure for BPB in Japan for more than two decades [15]. Resistant populations of *B. glumae* to oxolinic acid have been found [16–19], which limits increasing use of this antibiotic compound for management of BPB. Oxolinic acid is not labeled for use on rice in the USA and many other countries. Compared to extensive research and significant advances made on management of sheath blight caused by *Rhizoctonia solani* and rice blast caused by *Magnaporthe oryzae*, very limited research has been conducted on the development of effective and sustainable management options for control of BPB.

In this article, we focus on the review of recent advances on the development of management strategies for BPB, including exclusion, genetic resistance, chemical control, biological control, and cultural practice. In addition, world distribution of the pathogen, characteristic symptoms of BPB, and current understanding of epidemics of BPB are also included. Two review articles covering the pathogenesis of *B. glumae* and the detection of BPB have been published previously [20, 21]. The terms "BPB" and "grain rot" have been used interchangeably in the literature. However, BPB has been commonly used in the USA and Latin America, while grain rot in Japan and other countries [20]. The term BPB is used in this review article.

### **2. Pathogens**

Since the first description of *Burkholderia glumae* (formerly *Pseudomonas glumae* Kurita and Tabei) as the bacterial pathogen causing rice seedling rot and grain rot in Japan in 1955 [22], BPB has been reported in more than 18 countries distributed in Africa, Asia, Latin America, and North America (**Table 1**). The total rice production from these countries accounted for more than 65% of total world rice production in 2018 [23]. BPB has become an increasingly important global disease in rice. In addition to *B. glumae*, *B. gladioli* has also been identified as another bacterial pathogen causing the BPB disease. Infection with *B. gladioli* produces the same symptoms as infection with *B. glumae*. The disease caused by *B. gladioli* has been reported in Arkansas (USA), China Japan, Louisiana (USA), Panama, and the Philippines, where *B. glumae* is also co-present (**Table 1**). In the USA, the cause of the BPB was not known at the time when epidemics of BPB occurred in 1995. In 1996–1997, however, when evaluating bacterial isolates from rice tissue for their ability to control the rice sheath blight fungus *R. solani*, investigators in Louisiana accidentally found that some of the *B. glumae* isolates caused panicle blighting symptoms when greenhouse grown rice plants were spay inoculated [44]. This led to the discovery of *B. glumae* as the causal agent of the BPB disease.

**69**

**3. Symptoms**

**Table 1.**

*and B. gladioli in rice as of January 2019.*

*Sustainable Strategies for Managing Bacterial Panicle Blight in Rice*

**Country Year BPB pathogen Reference** Japan 1955 *B. glumae* [22, 24] Taiwan (China) 1983 *B. glume* [25] Columbia 1989 *B. glumae* [26] Latin America 1989 B. glumae [26] Vietnam 1993 *B. glumae* [27] Japan 1996 *B. gladioli* [28, 29] The Philippines 1996 *B. glume* and *B. gladioli* [30–32] Louisiana (USA) 2001 *B. glume* and B. gladioli [1, 33] Korea 2003 *B. glumae* [34] China 2007 *B. glumae* [35] Panama 2007 *B. glume* and *B. gladioli* [36] Nicaragua 2008 *B. glumae* [37] Arkansas (USA) 2009 *B. glume* and *B. gladioli* [1, 9] Mississippi (USA) 2009 *B. glumae* [1] Texas (USA) 2009 B. glumae [1, 10]

BPB of rice can be caused by either *B. glumae* or *B. gladioli*. However, the former

Honduras 2011 *B. glumae* Lex Ceamer, personal communication

Mississippi (USA) 2012 *B. gladioli* [38] Costa Rica 2014 *B. glumae* [39] Ecuador 2014 *B. glumae* [40] South Africa 2014 *B. glumae* [41] India 2015 *B. glumae* [42] China 2018 *B. gladioli* [43]

The symptoms of BPB include seedling blight, sheath rot, and panicle blighting [1–4]. These symptoms can be induced by either *B. glume* or *B. gladioli*. Virulent bacterial strains produce the yellow-pigmented toxin toxoflavin on King's B agar medium (**Figure 1**), while avirulent strains do not produce this toxin [1]. Production of toxoflavin is an essential factor to induce the development of the

is the primary cause of the disease. The study of Nandakumar et al. [1] found that 76 and 5% of the bacterial strains collected were *B. glumae* and *B. gladioli*, respectively. In a field survey conducted in Mississippi using PCR analysis, it was found that 84% of rice panicle samples collected were positive for *B. glumae* and 12% of the samples positive for *B. gladioli* [38]. In a recent survey conducted in nine rice-producing counties of Arkansas, all 45 virulent bacterial isolates studied were *B. glumae*, and no *B. gladioli* isolates were identified [9]. In addition, the *B. glumae* pathogen tends to be more virulent and causes more damage to rice plants when

*Countries reported with the presence of bacterial panicle blight (BPB) caused by Burkholderia glumae* 

compared to the *B. gladioli* pathogen [20, 33].

symptoms on rice seedlings and grains [34, 45, 46].

*DOI: http://dx.doi.org/10.5772/intechopen.84882*


*Sustainable Strategies for Managing Bacterial Panicle Blight in Rice DOI: http://dx.doi.org/10.5772/intechopen.84882*

### **Table 1.**

*Protecting Rice Grains in the Post-Genomic Era*

1995 and 1998 [1, 8]. In Arkansas, BPB was so severe in 2010 that yield losses were estimated at 50% in susceptible cultivars [9]. In Texas, the outbreaks of BPB resulted in an estimate of 10–20% yield loss in the Texas Rice Belt in 2010 [10, 11]. Outbreaks of this disease also occurred in rice under organic production systems in 2010 in Texas [11]. In the disease-yield loss field study, we found BPB was highly destructive and could cause yield losses ranging from 1 to 59% (83–4883 kg/ha), with yield loss increasing approximately 5% (455 kg/ha) for every unit increase in BPB severity on the rating scale of 0–9 [12]. Based on annual rice production in the Mid-South USA in 2003–2013, it is estimated that BPB caused \$61 million USD of damage that would

feed 1.1 million people annually (Aaron Shew, personal communication).

Effective management of BPB is critical to minimizing the damage caused by the disease and maximizing production returns. However, limited options for management of the disease are available currently. No single genes or quantitative trait loci (QTLs) for complete resistance to BPB have been found so far [13, 14]. Only a few rice cultivars with partial resistance are available for commercial use. No chemical control options are available in the USA although oxolinic acid has been used as a major control measure for BPB in Japan for more than two decades [15]. Resistant populations of *B. glumae* to oxolinic acid have been found [16–19], which limits increasing use of this antibiotic compound for management of BPB. Oxolinic acid is not labeled for use on rice in the USA and many other countries. Compared to extensive research and significant advances made on management of sheath blight caused by *Rhizoctonia solani* and rice blast caused by *Magnaporthe oryzae*, very limited research has been conducted on the development of effective and sustainable management options for control of BPB. In this article, we focus on the review of recent advances on the development of management strategies for BPB, including exclusion, genetic resistance, chemical control, biological control, and cultural practice. In addition, world distribution of the pathogen, characteristic symptoms of BPB, and current understanding of epidemics of BPB are also included. Two review articles covering the pathogenesis of *B. glumae* and the detection of BPB have been published previously [20, 21]. The terms "BPB" and "grain rot" have been used interchangeably in the literature. However, BPB has been commonly used in the USA and Latin America, while grain rot in Japan and other countries [20]. The term BPB is used in this review article.

Since the first description of *Burkholderia glumae* (formerly *Pseudomonas glumae* Kurita and Tabei) as the bacterial pathogen causing rice seedling rot and grain rot in Japan in 1955 [22], BPB has been reported in more than 18 countries distributed in Africa, Asia, Latin America, and North America (**Table 1**). The total rice production from these countries accounted for more than 65% of total world rice production in 2018 [23]. BPB has become an increasingly important global disease in rice. In addition to *B. glumae*, *B. gladioli* has also been identified as another bacterial pathogen causing the BPB disease. Infection with *B. gladioli* produces the same symptoms as infection with *B. glumae*. The disease caused by *B. gladioli* has been reported in Arkansas (USA), China Japan, Louisiana (USA), Panama, and the Philippines, where *B. glumae* is also co-present (**Table 1**). In the USA, the cause of the BPB was not known at the time when epidemics of BPB occurred in 1995. In 1996–1997, however, when evaluating bacterial isolates from rice tissue for their ability to control the rice sheath blight fungus *R. solani*, investigators in Louisiana accidentally found that some of the *B. glumae* isolates caused panicle blighting symptoms when greenhouse grown rice plants were spay inoculated [44]. This led

to the discovery of *B. glumae* as the causal agent of the BPB disease.

**68**

**2. Pathogens**

*Countries reported with the presence of bacterial panicle blight (BPB) caused by Burkholderia glumae and B. gladioli in rice as of January 2019.*

BPB of rice can be caused by either *B. glumae* or *B. gladioli*. However, the former is the primary cause of the disease. The study of Nandakumar et al. [1] found that 76 and 5% of the bacterial strains collected were *B. glumae* and *B. gladioli*, respectively. In a field survey conducted in Mississippi using PCR analysis, it was found that 84% of rice panicle samples collected were positive for *B. glumae* and 12% of the samples positive for *B. gladioli* [38]. In a recent survey conducted in nine rice-producing counties of Arkansas, all 45 virulent bacterial isolates studied were *B. glumae*, and no *B. gladioli* isolates were identified [9]. In addition, the *B. glumae* pathogen tends to be more virulent and causes more damage to rice plants when compared to the *B. gladioli* pathogen [20, 33].

## **3. Symptoms**

The symptoms of BPB include seedling blight, sheath rot, and panicle blighting [1–4]. These symptoms can be induced by either *B. glume* or *B. gladioli*. Virulent bacterial strains produce the yellow-pigmented toxin toxoflavin on King's B agar medium (**Figure 1**), while avirulent strains do not produce this toxin [1]. Production of toxoflavin is an essential factor to induce the development of the symptoms on rice seedlings and grains [34, 45, 46].

### **Figure 1.**

*Colonies of Burkholderia glumae and production of yellow pigment (toxoflavin) by B. glumae on King's B agar plate (right) vs. no pigment production control plate (left). Photo was taken at 3 days after inoculation at 30°C.*

### **Figure 2.**

*A focal pattern of bacterial panicle blight (BPB) on the Presidio (cv) rice panicles (center) in a research plot inoculated with Burkholderia glumae at Beaumont, Texas.*

Unlike rice sheath blight and blast, BPB is difficult to be diagnosed based on the symptoms on panicles. Similar symptoms on panicles can be caused by many abiotic and biotic factors including heat, insect damage, and secondary microorganisms [3, 4, 47]. However, BPB has the symptoms that can be distinguished from other causes. BPB occurs sporadically on individual plants or in circular or oval patterns in the field (**Figures 2** and **3**). In contrast, common panicle blanking, caused by abiotic stress such as from excessive heat, develops in the field more uniformly and does not form apparent foci. There are three important characteristics of BPB that separate it from other panicle disorders: (1) BPB often does not appear to prevent successful pollination although it can affect individual glumes or whole panicles (**Figure 4**). Thus, seed may be present on the panicle unlike panicle sterility that is caused by heat stress. (2) Infected florets initially have discoloration ranging from light green to light brown on the basal portion of the glumes with a reddish-brown margin separating this area from the rest that becomes straw-colored later (**Figures 4** and **5**). (3) The rachis or branches of the panicle remain green for a while at the base of each floret, even after the glumes desiccate and turn tan (**Figures 4** and **5**). Florets at the latest stages of infection usually appear to be gray or black due to the abundant growth of saprophytic fungi on the surface (**Figure 5**). The disease can cause linear lesions on sheaths with a distinct reddish-brown border and a gray and necrotic center, resulting in sheath rot (**Figure 6A**) and stem rot (**Figure 6B**). On the leaves, lesions are circular to oval with a smooth reddish-brown border and a gray or strawcolored center (**Figure 6C**). If the infected plants are young, this disease can cause seedling blighting (**Figure 6D**) or seeding rot. The symptoms of seedling rot were

**71**

**Figure 4.**

**Figure 3.**

*Sustainable Strategies for Managing Bacterial Panicle Blight in Rice*

*with Burkholderia glumae at the flowering stage at Beaumont, Texas.*

first reported in Japan [22] and frequently occur in young rice plants. However, these symptoms on leaves, sheaths, stems, and seedlings are rarely observed under the field conditions in the Southern USA [4]. This is one of the reasons why no scouting methods have been developed to detect and predict the development of BPB based

*A close look at the symptoms of bacterial panicle blight (BPB) on Presidio (cv) rice panicles. Photo was taken approximately 2 weeks after inoculation with Burkholderia glumae at the flowering stage at Beaumont, Texas.*

*Symptoms of bacterial panicle blight (BPB) on a Presidio (cv) rice panicle head (arrow) in the field inoculated* 

on the symptoms on leaves and sheaths at the early crop growth stages.

*DOI: http://dx.doi.org/10.5772/intechopen.84882*

*Sustainable Strategies for Managing Bacterial Panicle Blight in Rice DOI: http://dx.doi.org/10.5772/intechopen.84882*

**Figure 3.**

*Protecting Rice Grains in the Post-Genomic Era*

Unlike rice sheath blight and blast, BPB is difficult to be diagnosed based on the symptoms on panicles. Similar symptoms on panicles can be caused by many abiotic and biotic factors including heat, insect damage, and secondary microorganisms [3, 4, 47]. However, BPB has the symptoms that can be distinguished from other causes. BPB occurs sporadically on individual plants or in circular or oval patterns in the field (**Figures 2** and **3**). In contrast, common panicle blanking, caused by abiotic stress such as from excessive heat, develops in the field more uniformly and does not form apparent foci. There are three important characteristics of BPB that separate it from other panicle disorders: (1) BPB often does not appear to prevent successful pollination although it can affect individual glumes or whole panicles (**Figure 4**). Thus, seed may be present on the panicle unlike panicle sterility that is caused by heat stress. (2) Infected florets initially have discoloration ranging from light green to light brown on the basal portion of the glumes with a reddish-brown margin separating this area from the rest that becomes straw-colored later (**Figures 4** and **5**). (3) The rachis or branches of the panicle remain green for a while at the base of each floret, even after the glumes desiccate and turn tan (**Figures 4** and **5**). Florets at the latest stages of infection usually appear to be gray or black due to the abundant growth of saprophytic fungi on the surface (**Figure 5**). The disease can cause linear lesions on sheaths with a distinct reddish-brown border and a gray and necrotic center, resulting in sheath rot (**Figure 6A**) and stem rot (**Figure 6B**). On the leaves, lesions are circular to oval with a smooth reddish-brown border and a gray or strawcolored center (**Figure 6C**). If the infected plants are young, this disease can cause seedling blighting (**Figure 6D**) or seeding rot. The symptoms of seedling rot were

*A focal pattern of bacterial panicle blight (BPB) on the Presidio (cv) rice panicles (center) in a research plot* 

*Colonies of Burkholderia glumae and production of yellow pigment (toxoflavin) by B. glumae on King's B agar plate (right) vs. no pigment production control plate (left). Photo was taken at 3 days after inoculation at 30°C.*

**70**

**Figure 2.**

**Figure 1.**

*inoculated with Burkholderia glumae at Beaumont, Texas.*

*Symptoms of bacterial panicle blight (BPB) on a Presidio (cv) rice panicle head (arrow) in the field inoculated with Burkholderia glumae at the flowering stage at Beaumont, Texas.*

### **Figure 4.**

*A close look at the symptoms of bacterial panicle blight (BPB) on Presidio (cv) rice panicles. Photo was taken approximately 2 weeks after inoculation with Burkholderia glumae at the flowering stage at Beaumont, Texas.*

first reported in Japan [22] and frequently occur in young rice plants. However, these symptoms on leaves, sheaths, stems, and seedlings are rarely observed under the field conditions in the Southern USA [4]. This is one of the reasons why no scouting methods have been developed to detect and predict the development of BPB based on the symptoms on leaves and sheaths at the early crop growth stages.

#### **Figure 5.**

*Comparison of the developmental symptoms of bacterial panicle blight (BPB) on infected kernels of rice (lower row) and healthy kernels (upper row). Photo was taken for rice kernels collected from different Presidio (cv) rice plants inoculated with Burkholderia glumae at the flowering stage in the field. Note the occurrence of secondary fungal infection on the discolored kernel at the late BPB development stage (lower right end).*

### **Figure 6.**

*Symptoms of sheath rot (A), stem rot (B), leaf lesions (C), and seedling blighting (D) caused by Burkholderia glumae in Presidio (cv) rice. Rice seedlings were inoculated with B. glumae and maintained in the greenhouse.*

### **4. Epidemiology**

The disease cycle and epidemiology of BPB of rice are not completely understood. Both *B. glumae* and *B. gladioli* species have been identified as the cause of the BPB disease. However, the former has much wider distribution in the world as shown in **Table 1**. The bacteria of both species were also found to be widely present in rice seed lots in the studies conducted in China, Japan, the Philippines, and the USA [21, 32, 48]. Therefore, infected seeds serve as the primary source of inoculum [1]. In addition, Jeong et al. [34] reported that *B. glumae* could also infect other plant species, including tomato, sesame, perilla (an herb), eggplant, and hot pepper. The bacteria are capable of inhabiting surface plants and soils under a wide range of environments [49, 50]. In a field survey conducted in Mississippi using PCR analysis, it was found that 83% of soil samples were positive for *B. glumae* and 2% of the soil samples positive for *B. gladioli* [38]. This survey also found that 85% of field irrigation water samples collected were positive for *B. glumae* and 2% of the water samples positive for *B. gladioli*. Therefore, soil and irrigation water can also serve as the sources of inoculum for the spread and development of BPB.

The bacterial pathogen invades germinated seeds, inhabits the roots and lower sheaths, and moves up the growing plant as an epiphyte (an organism growing on a plant surface, but not as a parasite) [2, 51, 52]. A recent study, using real-time fluorescence quantitative PCR to monitor the infection process of *B. glumae*, finds

**73**

**Figure 7.**

*http://beaumont.tamu.edu).*

*Sustainable Strategies for Managing Bacterial Panicle Blight in Rice*

that the bacterium also can directly infect the rice plant by colonizing the vascular bundle of lateral roots and then spreading to upper tissues such as leaf sheaths and leaf blades through vascular system [53]. Infection by the bacterium occurs at flowering by invading rice spikelets through stomata or wound in the epidermis of glumes. The bacterium colonizes and multiplies in spikelets quickly after invasion by utilizing intermediate sugars in developing grains [51, 52]. The bacteria are spread primarily by splashing and windblown rain and panicle contact, resulting in the formation of disease foci that are frequently observed in the field [2, 54, 55]. High temperatures in combinations with high humidity or frequent rain are essential for the development of BPB epidemics. The outbreaks of BPB are usually triggered by conditions of high temperatures in combination with simultaneously high relative humidity during the heading-flowering stages. In the observations of Yokoyama and Okuhara [56], the disease developed when minimum daily temperature was ≥23°C and moderate rainfall (<30 mm/day) occurred during heading. Tsushima et al. [57] found BPB commonly occurred when relative humidity was more than 95% for 24 hours during flowering. Lee et al. [58] reported that the disease did not develop when the minimum daily temperature was less than 22°C and when relative humidity was below 80% during the heading stage. Nandakumar et al. [1] found that the optimum temperature for the growth of *B. glumae* and *B. gladioli* ranged from 35 to 40°C. The outbreaks of BPB in the Southern USA in the epidemic years appeared to be related to unusual weather conditions. Weather conditions favorable for the development of the disease were high nighttime temperatures and high humidity or frequent rainfall during heading and flowering [10]. For example, in the 2010 epidemic year, abnormally high minimum (night time) temperatures occurred on June 21 through July 10 (**Figure 7**) when ca. 60% of the Texas rice acreage was near or at heading and flowering. During that period, rainfall was frequent and relative humidity was 95% or above most of the time (**Figure 7**). The combination of favorable weather conditions, high nighttime temperatures and high humidity, occurring at the most susceptible stages of rice plants promoted the infection and development of BPB. Similar weather patterns were observed in 1995 when a severe epidemic of BPB took place in Texas. There were many days with high maximum temperatures 35°C or

*Air temperatures and rainfalls during the 2010 growing season of rice at the Beaumont Center, Jefferson County, Texas. Note the red-dashed rectangle area showing minimum (night) air temperatures (blue curves) higher above the 65-year historical average (the brown curve) and frequent rainfalls (green bars). The dashed rectangle area represents the period of June 21 through July 10 that coincided with the heading and flowering stages (source:* 

*DOI: http://dx.doi.org/10.5772/intechopen.84882*

### *Sustainable Strategies for Managing Bacterial Panicle Blight in Rice DOI: http://dx.doi.org/10.5772/intechopen.84882*

*Protecting Rice Grains in the Post-Genomic Era*

**72**

**Figure 6.**

**Figure 5.**

**4. Epidemiology**

*Symptoms of sheath rot (A), stem rot (B), leaf lesions (C), and seedling blighting (D) caused by Burkholderia glumae* 

*Comparison of the developmental symptoms of bacterial panicle blight (BPB) on infected kernels of rice (lower row) and healthy kernels (upper row). Photo was taken for rice kernels collected from different Presidio (cv) rice plants inoculated with Burkholderia glumae at the flowering stage in the field. Note the occurrence of secondary* 

*fungal infection on the discolored kernel at the late BPB development stage (lower right end).*

The disease cycle and epidemiology of BPB of rice are not completely understood. Both *B. glumae* and *B. gladioli* species have been identified as the cause of the BPB disease. However, the former has much wider distribution in the world as shown in **Table 1**. The bacteria of both species were also found to be widely present in rice seed lots in the studies conducted in China, Japan, the Philippines, and the USA [21, 32, 48]. Therefore, infected seeds serve as the primary source of inoculum [1]. In addition, Jeong et al. [34] reported that *B. glumae* could also infect other plant species, including tomato, sesame, perilla (an herb), eggplant, and hot pepper. The bacteria are capable of inhabiting surface plants and soils under a wide range of environments [49, 50]. In a field survey conducted in Mississippi using PCR analysis, it was found that 83% of soil samples were positive for *B. glumae* and 2% of the soil samples positive for *B. gladioli* [38]. This survey also found that 85% of field irrigation water samples collected were positive for *B. glumae* and 2% of the water samples positive for *B. gladioli*. Therefore, soil and irrigation water can also

*in Presidio (cv) rice. Rice seedlings were inoculated with B. glumae and maintained in the greenhouse.*

serve as the sources of inoculum for the spread and development of BPB.

The bacterial pathogen invades germinated seeds, inhabits the roots and lower sheaths, and moves up the growing plant as an epiphyte (an organism growing on a plant surface, but not as a parasite) [2, 51, 52]. A recent study, using real-time fluorescence quantitative PCR to monitor the infection process of *B. glumae*, finds

that the bacterium also can directly infect the rice plant by colonizing the vascular bundle of lateral roots and then spreading to upper tissues such as leaf sheaths and leaf blades through vascular system [53]. Infection by the bacterium occurs at flowering by invading rice spikelets through stomata or wound in the epidermis of glumes. The bacterium colonizes and multiplies in spikelets quickly after invasion by utilizing intermediate sugars in developing grains [51, 52]. The bacteria are spread primarily by splashing and windblown rain and panicle contact, resulting in the formation of disease foci that are frequently observed in the field [2, 54, 55].

High temperatures in combinations with high humidity or frequent rain are essential for the development of BPB epidemics. The outbreaks of BPB are usually triggered by conditions of high temperatures in combination with simultaneously high relative humidity during the heading-flowering stages. In the observations of Yokoyama and Okuhara [56], the disease developed when minimum daily temperature was ≥23°C and moderate rainfall (<30 mm/day) occurred during heading. Tsushima et al. [57] found BPB commonly occurred when relative humidity was more than 95% for 24 hours during flowering. Lee et al. [58] reported that the disease did not develop when the minimum daily temperature was less than 22°C and when relative humidity was below 80% during the heading stage. Nandakumar et al. [1] found that the optimum temperature for the growth of *B. glumae* and *B. gladioli* ranged from 35 to 40°C.

The outbreaks of BPB in the Southern USA in the epidemic years appeared to be related to unusual weather conditions. Weather conditions favorable for the development of the disease were high nighttime temperatures and high humidity or frequent rainfall during heading and flowering [10]. For example, in the 2010 epidemic year, abnormally high minimum (night time) temperatures occurred on June 21 through July 10 (**Figure 7**) when ca. 60% of the Texas rice acreage was near or at heading and flowering. During that period, rainfall was frequent and relative humidity was 95% or above most of the time (**Figure 7**). The combination of favorable weather conditions, high nighttime temperatures and high humidity, occurring at the most susceptible stages of rice plants promoted the infection and development of BPB. Similar weather patterns were observed in 1995 when a severe epidemic of BPB took place in Texas. There were many days with high maximum temperatures 35°C or

### **Figure 7.**

*Air temperatures and rainfalls during the 2010 growing season of rice at the Beaumont Center, Jefferson County, Texas. Note the red-dashed rectangle area showing minimum (night) air temperatures (blue curves) higher above the 65-year historical average (the brown curve) and frequent rainfalls (green bars). The dashed rectangle area represents the period of June 21 through July 10 that coincided with the heading and flowering stages (source: http://beaumont.tamu.edu).*


**Table 2.**

*Summary of rice crops and weather data at Beaumont and Eagle Lake, Texas in 1995.*

### **Figure 8.**

*Yield (left Y-axis) and bacterial panicle blight (BPB) severity (% panicles affected) (right Y-axis) of eight cultivars of rice (X-axis) in naturally infested field at Beaumont, Texas, in 1995 (source: [11]). Error bars are present in columns.*

above, day temperatures above 32°C from 10 am to 12 pm (the flowering time), and precipitation from the last week of June through the first week of August (**Table 2**). Heading and flowering occurred on a large percentage of the Texas rice crop during that period. These conditions were associated with severe outbreaks of BPB and significant yield losses in 1995. **Figure 8** shows an example of the severity of this disease in 1995 and its association with yield loss for different rice cultivars, with the disease severity levels ranging from 1 to 22% of panicles affected.
