Forewarning Rice Blast in India

Forewarning Rice Blast in India 

C. S. Reddy, K. Susheela, A. S. Kapoor*, R. Kaundal*, 

N. V. Krishnaiah, B. Mishra, Y. S. Ramakrishna**, 

Y. G. Prasad**, D. Yella Reddy** and M. Prabhakar** 

DIRECTORATE OF RICE RESEARCH 

Rajendranagar, Hyderabad –500 030, A.P., India. 

* CSK Himachal Pradesh Krishi Viswavidyalaya, Palampur – 176 062 

** Central Research Institute for Dryland Agriculture, Hyderabad – 500 059

DRR Technical Bulletin No. 9, 2004-2005 

Correct Citation 

Reddy, C. S., K. Susheela, A. S. Kapoor, R. Kaundal, N. V. Krishnaiah, 

B. Mishra, Y. S. Ramakrishna, Y. G. Prasad, D. Yella Reddy and 

M. Prabhakar, 2004. Forewarning Rice Blast in India, Technical Bulletin No. 9, 

2004-2005, Directorate of Rice Research, Rajendranagar, Hyderabad – 500030, 

A. P., India, 46 pp. 

Published by 

Dr. B. Mishra 

Project Director 

Directorate of Rice Research 

Rajendranagar 

Hyderabad - 500 030, India. 

Tel : +91-40-2401 5120, 2401 5036-39 

Fax : +91-40-2401 5308 

Web Site : http://www.drrindia.org 

E-mail : mishra.b@eudoramail.com 

Designed by 

Dr. K. Susheela 

Photographs by 

Dr. C.S. Reddy 

Printed by 

Suneetha Offset Printers 

# 4-5-716/3, Kuthbiguda, Koti, 

Hyderabad - 500 027. A.P., India. 

Ph : +91-40-24657269, 9391034092.

PREFACE 

Rice, the staple food crop in India, holds key to our country’s food security. It is 

grown in 44 million hectares with annual production of 90 million tons. To meet 

the growing demand and sustain our self sufficiency in food, we need to grow an 

additional 2 million tons of rice every year. To achieve this goal, the losses due to 

biotic and abiotic stresses have to be minimised. Among the biotic stresses, blast is 

one of the most destructive diseases and is widely prevalent in India. It is a major 

limiting factor in realizing the full yield potential of rice cultivars in plains and hilly 

areas of the country. 

Blast is endemic to several rice growing areas in the country due to favourable 

environment during the crop season. Weather has very important role to play in the 

appearance, multiplication and spread of blast pathogen. Though a lot of fragmented 

information is available on the blast disease and influence of weather on its 

development, no consolidated effort has been made so far, to bring these together 

to develop an ideal model for rice blast forewarning. Hence, the present study was 

aimed to develop database on climate and blast disease in different agro-ecological 

regions of India, to identify significant weather factors conducive for incidence, 

intensification and spread of pathogen and disease, and to generate / validate 

weather based forecasting models and operational forewarning systems for blast 

control. 

The results reported in this publication ‘Forewarning Rice Blast in India’, are 

generated through the National Agricultural Technology Project on ‘Development of 

Weather based Forewarning Systems for Crop Pests and Diseases’ undertaken at the 

Directorate of Rice Research, Hyderabad, CSK Himachal Pradesh Krishiviswavidyala, 

Palampur, and Central Research Institute for Dryland Agriculture, Hyderabad. It is 

expected that this publication will be useful to researchers, extension workers and 

farmers in understanding the blast disease and its relationship with weather, and 

adopting suitable disease management practices based on the forewarning models 

developed in the project. 

(B. MISHRA) 

Project Director

ACKNOWLEDGEMENTS 

The Authors are grateful to the authorities of National Agricultural 

Technological Project (NATP) for providing financial support for executing 

these studies and bringing out this publication 'Forewarning Rice Blast in 

India' under the sub-project on "Development of Weather based Forewarning 

Systems for Crop Pests and Diseases" {F. No. 27 (27)/99/NATP/MM-III- 

17}. They appriciate the efforts of Dr. T. Madhusudan, Principal Scientist 

(Plant Pathology), Rice Section, Agricultural Research Institute, ANGR 

Agricultural University for thorough scrutiny of the manuscript.

CONTENTS 

CONTENTS 

CONTENTS 

I. Introduction 1 

II. Importance of Blast and need for Forewarning 2 

III. Blast Hotspot areas and extent of Damage 5 

IV. Work carried out already on Blast Forewarning 9 

V. Approaches for developing Forewarning Models 14 

1. Historical data on Blast and Weather in Himachal Pradesh 15 

2. Models developed by DRR, Hyderabad and CSK HPKV, Palampur 17 

3. Validation of Models developed at DRR 29 

VI. Application of the Forewarning Model of DRR 32 

VII. Strategies for the control of Blast 35 

VIII. Future work 37 

Executive Summary 39 

References 42

Back Cover Page (Clock wise): 

l Blast experimental field at DRR 

l Due formation and leaf blast infection in the nursery 

l Growth of blast pathogen in a petri plate 

l Examination of micro-weather station at the experimental field 

l Transfer of weather data from Data Logger to PC

I. I. INTRODUCTION 

INTRODUCTION 


“At best, the world food outlook for the decades ahead is grave; at worst, it is 

frightening.” Thus stated Forrest F. Hill, Member of Ford Foundation, before the Trustees 

of the Foundation in 1959, such was the situation at that time. Precisely twenty years 

later, while commenting on the rice crop, Robert F. Chandler Jr., the first Director 

General of International Rice Research Institute, Philippines, mentioned, “So dependent 

upon rice are the Asian countries that throughout history a failure of that crop has 

caused widespread famine and death”. Again in 1982, Dr. Chandler expressed 

concern that most rice scientists were not perturbed by the low yields they obtained in 

their experiments. He wrote: “It is disturbing to read paper after paper, from various 

research and educational organizations experimenting with rice, in which yield data 

ranging from 1,500 to 3,000 kilograms per hectare are reported and yet no reasons 

are given for the low yields.” Since then, there has been a revival of interest among 

the researchers in finding the reasons for the failure of that crop and low yields, and 

solutions to save the crop on which many of the rice based countries depend. 

"Rice is Life", the theme of International Year of Rice, 2004 reflects the 

importance of rice (Oryza sativa L.), that holds the key to our country’s ability to 

produce enough food for our people. It is grown in 44.62 million hectares with 

annual production of 93.08 million tons. To sustain and exist as a nation, the demand 

for rice is expected to be 100 million tons during 2010 and 140 million tons by 2025 

(Mishra, 2002). Therefore, the major concern in coming years is to increase the 

productivity from the present level of 2.066 t/ha to more than 3 t/ha. To achieve this 

goal, the losses due to biotic and abiotic stresses have to be tackled. Among the 

biotic stresses, it is often mentioned that diseases caused by plethora of 

microorganisms, take a heavy toll of the crop in the humid tropical rice growing 

environment. While an average yield loss span from 5 to 15 % over large areas, total 

crop failure due to pests and disease epidemic is regularly encountered in some or 

the other pockets of the country. 

Of the various diseases of rice, blast caused by Pyricularia grisea Sacc. 

{Magnaporthe grisea (Hebert) Barr.} is one of the most destructive diseases and is 

widely prevalent in India. Further, it is a major limiting factor in stepping up rice 

yields in plains and hills of the country. It continues to be the enigmatic problem in 

several rice growing ecosystems of both tropical and temperate regions of the world 

and is a serious constraint in realizing the full yield potential of rice cultivars. 

1


II. II. IMPORT IMPORT IMPORTANCE IMPORT ANCE OF OF BLAST BLAST AND AND AND NEED NEED FOR FOR FOREW FOREWARNING 

FOREW FOREWARNING 

ARNING 

Blast has been the most important disease occurring in all rice growing areas 

causing heavy losses in yield. In India, it was first recorded in 1913, however, a 

devastating epidemic was recorded in Tanjore delta of Tamil Nadu in 1919 

(Padmanabhan, 1965). Since then, time to time, its occurrence in epidemic form has 

been reported from different parts of the country. The ability of the blast pathogen to 

infect rice at different stages of growth and its adaptation to both upland and lowland 

rice ecosystems, are indications of the plasticity of P. grisea to changing environment. 

The blast disease remains a threat to rice production because of its apparently 

unpredictable outbreaks and the resulting economic losses depending on weather. It 

continues to be the most destructive disease of rice despite decades of research 

towards its control. The disease is aptly named after the damage and consequent 

yield losses it causes. The damage results in complete drying and withering of the 

affected rice plants, and in a severe form the harvest may not be equal to the seed 

sown. 

Blast disease is explosive in nature and the pathogen produces astronomical 

number of spores from an infected field that are wind disseminated over vast areas 

to cause an epidemic through polycyclic infections. Actually spores of this fungus are 

always present in the atmosphere all the year round, and strike the crop when 

environmental conditions favour. Hence, the disease is regarded as one of the greatest 

pathological threats to the rice crop and it is weather and host nutrition sensitive. The 

disease progress is driven by the low night temperature (22 to 28° C), high relative 

humidity (> 95%), dew deposit, extended leaf wetness period (> 10 hrs.), cloudy 

and drizzling weather, soil fertility (high nitrogen), degree of host susceptibility and 

the straw of the previously infected crop heaped nearby. Leaf wetness induced by 

dew fall or incessant drizzle will facilitate in brining down the air borne spores in 

contact with plant parts. These spores can germinate only if free water is available 

and invade host cells within 6 to 8 hours. The symptoms appear on leaves 4 to 5 

days later with concomitant spore production. Infection followed by spore production 

occurs repeatedly culminating in the total destruction of the crop. The low night 

temperatures not only cause heavy condensation of water vapour in to dew but also 

favour rapid disease progress. Free water on leaf surface or high relative humidity 

will facilitate rapid release of conidia from infected leaf surface. 

These airborne conidia infect all the aerial parts of the plant at any stage of 

the crop from seedling to maturity. The name of the disease is suffixed to the plant 

part infected – leaf blast, node blast, neck or panicle blast, the last one being the 

most destructive phase of the disease affecting grain formation causing drastic 

reduction in grain quality and yield. On leaf, symptoms appear as elliptical with 

more or less pointed ends resembling a spindle. Initially, they appear as small greyish 

dots of pin-head size that finally enlarge into a spindle-shaped spot of about 1 cm 

long and 0.5 cm broad. Such spots have brown margin with grey centre. When 

numerous spots occur on leaves it results in the death and drying up of the plant. The 

node blast symptoms appear as black patches on infected nodes and all parts above 

2

the infected node die. If infection occurs at milky stage, the panicle at the infected 

node breaks and hangs down. Early neck blast infection at flowering stage results in 

chaffy ear heads with total yield loss. Neck blast is most serious phase of the blast 

disease as it occurs late in the plant’s development - after the farmer invested all his 

production inputs. 

Blast is epidemic to most rice growing areas of India due to favourable 

environment during the crop season. Weather plays important role in the appearance, 

multiplication and spread of blast fungus. Therefore, understanding the biology of 

the blast pathogen under field conditions is required for rice blast prediction. It is well 

documented that the blast pathogen can infect rice only when free water is present at 

the infection court influencing the major events of disease cycle, viz., spore germination, 

appressorium formation, polycyclic infections, and spore release (Fig. 1). 

Fig. 1. Blast Disease Cycle 


It is hypothesized that blast epidemics depend on the length of the time that 

free water remains on the rice plants in the field; longer the period of wetness, 

greater the number of lesions and faster the epidemic progress. The stage of the 

crop caught in prolonged dew usually gets severe infections. It is possible, therefore, 

to have a high percentage of neck blast incidences in a field that has shown little leaf 

blast, and in such instances the yield losses could be total if adequate timely preventive 

measures are not taken. Considerable efforts have been directed towards developing 

blast resistant cultivars but due to high variability in the pathogen most of the resistant 

3


varieties frequently succumb to the disease. Therefore, the most practical way to 

control blast epidemics is to use fungicides, which would be highly economical if 

used judiciously for which forewarning of blast is very important. So, if sound 

forewarning system is developed, the explosive nature of the disease could be 

prevented by timely application of the control measures. As the management of this 

disease primarily depends upon our ability to anticipate epidemic outbreaks and 

schedule appropriate fungicidal sprays for disease control, an accurate forewarning 

system based on data generated through field experimentation is necessary to combat 

this explosive disease. 

Though a lot of fragmented information is available on the blast disease and 

influence of weather on its development, no consolidated efforts were made so far, 

to bring these inputs together which could lead to the development of ideal model 

for rice blast forewarning. Hence, the present study was aimed to develop database 

on climate and blast disease affecting the production of rice crop in different 

agro-ecological regions of India, to identify significant weather factors conducive for 

incidence, intensification and spread of pathogen and disease, and establish 

crop-pest-weather relationships, to generate / validate weather based forecasting 

models for rice blast and operational forewarning systems for blast control, and for 

use in agro-met decisions that safeguard the farmers’ interest in increasing the rice 

production of the country in general, and of blast epidemic areas in particular. 

4 

Leaf blast with sporulating lesions Node blast Neck blast 

Neck blast infected field Cloudy and rainy weather conditions

III. BLAST HOT SPOT AREAS AND EXTENT OF DAMAGE 

Recurrent blast epidemics are reported from the sub-Himalayan regions of 

Jammu & Kashmir, Himachal Pradesh, hill districts of Uttaranchal and West Bengal. 

It is one of the most destructive diseases in upland rice growing areas of Arunachal 

Pradesh, Manipur, Mizoram, Meghalaya, Assam, Chotanagpur region of South Bihar, 

Chattisgarh and Bastar regions, and Jeypore tract of Orissa. In peninsular India, 

blast epidemics are reported from Andhra Pradesh, Tamil Nadu and Coorg region 

of Karnataka. In Western India, it is of considerable importance in Konkan region of 

Maharashtra and in Gujarat (Fig. 2). 

Fig. 2. Blast Distribution in India 


Blast occurs under divergent agro-meteorological conditions in India. In high 

rainfall zones (Rainfall: > 1500 mm, Temperature: 20 – 24 0 C) of North and North 

Eastern India, rice crop suffers due to this disease during June – September. In Western 

and Central India (Rainfall: 1000 mm, Temperature: 24 – 30 0 C) the disease occurs 

during August to October. Whereas, blast incidence is primarily associated with dry 

periods and cooler nights (18 – 22 0 C) that are prevalent during November – February 

in Andhra Pradesh, Karnataka, Tamil Nadu and Kerala States. 

The severity and damage caused by rice blast fluctuate year by year and 

from place to place. However, the information collected on the blast endemic districts 

/ areas in the country are given in the table 1, which indicates the prevalence of the 

disease in almost all the rice growing areas. Early appearance of the disease was 

observed during April-July in Arunachal Pradesh, followed by West Khasi hills of 

Meghalaya during June-October, Cuttack, Ganjam and Koraput in Orissa during 

5


July-August, Hill zones of Jammu & Kashmir during July-September, Manipur Central 

valley during July-October, most of the northern parts of the country during August- 

October, and the southern parts in general during the months of September-October 

and continues up to February. 

Table 1. Distribution of Blast in India 

6 

State Endemic Districts/Areas Favourable Period 

Andhra Pradesh Srikakulam, Vishakapatnam, 

Guntur, Nellore, Chittoor, 

Nizamabad, Medak, Ranga Reddy, 

Mahboobnagar, E & W Godavari 

September – February 

Arunachal Pradesh Arunachal Pradesh April -July 

Assam Karimganj, Tinsukia, Nowgong, 

Kamrup, Goalpara, N. Lakhimpur 

August – October 

Bihar Ranchi, Hazaribagh August – October 

Chattisgarh Northern hill regions September – October 

Gujarat Kheda September – October 

Haryana Hissar, Karnal August - October 

Himachal Pradesh Kangra valley (Malan, Palampur), 

Kulu, Mandi 

August – October 

Jammu & Kashmir Hill zones of Anantnag , Rajouri, 

Jammu, Udampur, Larnoo 

July - September 

Karnataka Mandya, Kodagu, Shimoga, Dharwad September - October 

Kerala Palghat, Kuttanad September – February 

Madhya Pradesh Bastar region, Rewa, Bilaspur September – October 

Maharashtra Pune, Ratnagiri, Kolaba, 

Parbhani, Kolhapur 

September – October 

Manipur Manipur Central valley July – October 

Meghalaya West Khasi hills June – October 

Mizoram Mizoram August – October 

Orissa Cuttack, Ganjam, July – August 

Koraput September - December 

Punjab Amritsar, Bhatinda, Patiala, 

Ferozpur, Ropar, Hoshiarpur 

August - October 

Tamil Nadu Tanjavur, Coimbatore, Chengalput, 

S & N Arcot, Periyar, Madurai, 

Pudukkotai, Thirunalvelli 

October – February 

Tripura West & South Tripura July – October 

Uttaranchal Almora, Nainital and other hill areas August – October 

Uttar Pradesh Faizabad, Balia August – October 

West Bengal Darjeeling, Cooch Behar August – September

Table 2. Blast locations, ecosystem, varieties grown, initial blast appearance and 

its severity (%) 

Locations (State) 

Almora (Uttaranchal) Irrigated (1250) K 39 20 th 28 th 7 th 50-60 70-80 

June July Sept. 

Amberpet (A.P.) Irrigated (542) HR 12 12 th 30 th 7 th 30-40 20-30 

July Sept. Nov. (40-50*) 

Arundhutinagar Upland (12.6) Sambha 20 th 5 th - 30-35 30-40 

(Tripura) Mahsuri July Sept. 

Chiplima (Orissa) Irrigated (178) Jaya 20 th 7 th 5 th 10-20 10-15 

July Oct. Nov. 

Cuttack (Orissa) Upland (23) IR 50 16 th 25 th - 25-30 5-10 

June Sept. 

Dungara (RS Pura) Irrigated (1067) K 343 2 nd 4 th - 30-40 - 

May July 

Ghaghraghat Rainfed lowland Jalpriya 16 th 10 th 1 st 10-15 10-15 

(Uttar Pradesh) (112) June July Sept. 

Jagadalpur Upland (553) Mahamaya 18 th 10 th 27 th 20-30 20-25 

(Chattisgharh) June Aug. Aug. 

Kaul(Haryana) Irrigated (241) Taraori 22 nd 21 st 18 th 10-15 30-50 

Basmati June Aug. Oct. 

Khudwani (J&K) Irrigated (1560) K 448 3 rd 20 th 4 th 20-30 10-15 

May June Aug. 

Lonavla (Maharashtra) Rainfed lowland EK 70 4 th 29 th 22 nd 40-50 50-60 

(622) (Kolpi) June June Aug. 

Malan (H.P.) Irrigated (960) T 23 30 th 25 th 25 th 50-60 50-60 

May July Sept. 

Mandya (Karnataka) Irrigated (900) Mandya 8 th - 5 th 10-15 60-70 

Vijaya Aug. Nov. 

Nawagam (Gujarat) Irrigated (10) Pankhari- 30 th 10 th 5 th 30-45 10-20 

203 June Sept. Nov. 

Pattambi (Kerala) Upland (25) Jyothi 25 th 29 th 6 th 20-30 30-40 

June Aug. Sept. 

Pondicherry Irrigated (5) IR 50 5 th 28 th - 30-40 5-10 

(Pondicherry) Oct. Nov. 

Ponnampet Upland (867) Intan 18 th 19 th 15 th 40-70 50-60 

(Karnataka) June Sept. Nov. 

Rajendranagar Irrigated (523) HR 12 21 st 23 rd 14 th 40-50 20-25 

(Andhra Pradesh) June Aug. Oct. 

Rewa Irrigated (360) Basmati 28 th 15 th - 40-50 5-10 

(Madhya Pradesh) June Sept. 

Thoubal, Wangbal Rainfed lowland Punshi 2 nd 7 th - 50-60 20-25 

(Manipur) (781) July Sept. 

Tuljapur (Karnataka) Irrigated (667) Tuljapur - I 3 rd 15 th - 30-50 5-10 

July Sept. 

* Node blast severity (%) 

Ecosystem 

(M above MSL) 

Local 

Variety 

Date of Sowing / Initial 

occurrence of Blast 

(In general) 


Blast severity (%), 

in general 

Sowing Leaf Neck Leaf Neck 

7


Blast is one disease which is most feared by the farmers as it occurs usually 

over a wide area with remarkable destructiveness under favourable conditions. Rice 

seedlings or plants at tillering stage are often completely killed. Severe incidence of 

panicle or neck infection results in drastic reduction in yields. The severity of blast at 

different locations in the country under different ecosystems indicates that the disease 

is a major problem at all the ecosystems, viz., upland, irrigated and rainfed lowland 

conditions (Table 2), and the local commonly grown varieties were found highly 

susceptible to the disease. The disease occurrence was observed as below as at 5 

MSL at Pondicherry to as high as 1560 MSL at Khudwani in J & K. 

Leaf blast phase was found the common occurrence, though neck infection 

was also moderate to high wherever the leaf infection was found. Node infection 

was found not common and insignificant in reducing the grain yield. Initial leaf blast 

symptoms were observed early at Khudwani around 3 rd week of June, then at Lonavla 

around the last week of June, and very late symptoms were observed at Chiplima 

around the 1 st week of October and in the last week of November at Pondicherry. 

Neck infection was very early around the 1 st week of August at Khudwani and late at 

Ponnampet at the end of 1 st fortnight of November (Table 2). Both leaf and neck 

phases of blast were found severe at Almora, Lonavla, Malan and Ponnampet. 

However, leaf blast was also severe at Rajendranagar, Rewa, Thoubal (Wangbal) 

and Tuljapur, while neck blast was severe at Mandya. 

8

IV. WORK CARRIED OUT ALREADY ON BLAST FOREWARNING 

The work carried out in India varied from correlating the weather parameters 

with blast disease to development of step wise multiple regression models. However, 

often the attempts were made to validate the developed forewarning systems under 

field conditions or incorporate them in the package of practices of concerned crops. 

For the prediction of leaf and panicle blast, a number of models have been developed 

in different countries including India. However, majority of them have not been 

validated in field and therefore, their practical utility is almost negligible. Nevertheless, 

based on these studies, further work was carried out in Japan and South Korea; and 

the rice prediction models are now being used as a strategy to manage rice blast. 

Rice blast disease and its relationship with weather 


The intensity of the epidemic outbreak is mainly determined by the influence 

of environmental conditions on the blast fungus and the host plants. Severe epiphytotics 

of blast occurred in Florida during July and August when the average night 

temperatures were approximately 70 o F and the dew periods 11-13 hours (Kahn and 

Libby, 1958). Surin et al (1991) reported that the optimum humidity and temperature 

for blast development in Thailand was 90% and 25 - 28 o C, respectively, while high 

rainfall and high humidity with low temperature resulted in more disease damage. 

The effect of temperature on lesion enlargement and sporulation of Pyricularia oryzae 

in rice leaves was reported by Kato and Kozaka (1974). They stated that, blast lesions 

on rice leaves expanded faster but reached a smaller final size at high temperature 

regimes of 32 o C continuously, 32/25 o C day/night or 32/20 o C day/night in a 12-h 

thermoperiod than at 25 o C or 25-16 o C day/night and sporulation proceeded for 

more than 20 days at each thermal treatment. Huang (1980) reported the disease 

distribution and dynamic influence of climate and weather, effect of fertilizers on 

disease severity epidemics and forecasting. 

The intensity of blast infection is greatly influenced by the environment and 

the variety of the rice grown. Extensive studies on the role of temperature and humidity 

have been made (Padmanabhan, 1953; Chakrabarti, 1971) and concomitant 

occurrence of temperature of 20 to 24oC and a relative humidity of 90 % was 

considered favourable for the blast development. The role of temperature and humidity 

in blast incidence has been stressed by Sadasivan et al (1965) and Subramanian 

(1967). According to them resistance to blast is governed not only by genetic factors 

but also to a large extent by a set of very critical environmental factors including 

night temperature (20oC) which influence the metabolic pattern of the host. 

Manibhushanrao and Day (1972) also reported that low temperatures result in partial 

breakdown of resistance. These findings were found to show a good correlation with 

field incidences of blast (Govindaswamy, 1964; Padmanabhan, 1965). The occurrence 

of a minimum temperature ranging from 20 to 25oC along with humidity of 95 % 

and above lasting for a week or more during any of the susceptible growth phases of 

the crop was found associated with the blast epidemic (Padmanabhan et al., 1971). 

Venkat Rao and Muralidharan (1982) observed rapid development of blast at tillering 

9


and heading stages coinciding with low temperature (20 o C or less) and high relative 

humidity (90 % and above) at these two vulnerable stages of the crop. Further, Kaur 

et al (1977) also studied the influence of temperature on blast development and 

stated that temperature influenced both penetration and establishment phases, and 

particularly appeared to be more critical in case of susceptible variety at 25 o C. 

Data recorded in the upland conditions indicated that at least 5–7 rainy days 

are also essential for the blast development even if the congenial temperature and 

humidity are attained. However, a temperature humidity index (THI) of 74 and above 

lead to the extent of epiphytotic (Chaudhary and Vishwadhar, 1988). They reported 

end of June for the peak foliage blast severity under upland conditions of Arunachal 

Pradesh. Bhatt (1992) identified minimum temperature between 15-20 o C with an 

average of 22-25 o C and having a temperature difference between minimum and 

maximum of 10-15 o C ; more days with 90 % relative humidity or above with an 

average of more than 50 %, higher rainfall and more number of rainy days as the 

congenial and important factors for development of foliar blast in hills. Krishnan 

et al (1992) reported maximum spore releasing capacity of blast fungus when the 

minimum temperature rose from 16 to 25 o C and RH above 90 %. According to 

Sharma et al (1993) maximum blast incidence was observed when the minimum 

temperature was 20.75 o C to 22.29 o C during July and August with the maximum 

frequency of rainy days (19.3 days) in Nagaland. Of the 7 meteorological factors 

studied, the frequency of rainy days had the greatest influence on disease development. 

Further, studies conducted by Prasad Rao et al (1999) revealed that bright sunshine 

hours were positively correlated, while rainfall and number of rainy days (RD) were 

negatively correlated with disease severity. Recently, while studying the influence of 

weather factors in Jharkhand, Dubey (2003) stated that the mean temperatures of 

22-30.7 o C, relative humidity of 85.5%, rainfall of 6.3 to 9.1 mm and 6-8 rainy days 

were favourable for the maximum blast intensity. Sharma and Kapoor (2003) reported 

temperature of 20 to 30 o C and relative humidity >90 % optimum for rice blast 

infection. 

Methodologies for model development and prediction/thumb rules for blast 

Many studies of methods for forecasting blast disease have been made in 

Japan, based on information on the fungus, the host plant and the environment 

(Ono, 1965; Yamaguchi, 1970; Suzuki, 1975; Kato, 1976). Using mathematical 

equations, Kim et al (1985) estimated the number of blast lesions by trapped spores 

and the wetting period of the leaves. Sasaki and Kato (1972) tried to predict panicle 

blast by counting the number of diseased spikelets, plotting it against time and 

extrapolating the curve for the next 6 days. Kiyosawa (1972) proposed an equation 

for forecasting the number of lesions or cumulative spore numbers. Tsai and Su 

(1984) used step-wise regression analysis to analyze relationship between 

meteorological variables and incidence of blast on two cultivars. Combined data of 

various years resulted in equations, which gave better agreement between observed 

and predicted values than equations derived from data of individual years. 

10


Leaf blast epidemics start from infection foci. Possible origins of the foci are 

seedling mats left in rice fields. Such foci are, however, scarce and only blast experts 

can find them. In Fukushima Prefecture, Japan, the normal density in 1980 to 1989 

was 0.1 focus / ha. The onset of a leaf blast epidemic is usually recognized a few 

weeks after the identification of foci (Ishiguro and Hashimoto, 1991). Kim (1982) 

used the iodine – potassium iodide methods to detect infection sites of the fungus 

under field conditions within 30 minutes. The detection of infection sites was four 

days earlier than direct observation of the leaf blast lesion. 

Kobayashi (1984) pointed out that the beginning of general epidemics of 

leaf blast, not the beginning of focal epidemics, should be considered for the onset 

of a leaf blast epidemic in a district. He proposed the following criteria to predict the 

beginning of general epidemics of leaf blast. (1) Each day starts at noon, (2) Weather 

is categorized as cloudy or rainy and calm or breezy, (3) Leaves have been wetted by 

dew or light rain during the night, and (4) Minimum night air temperature is above 

16 o C. When these criteria are satisfied, the environment is favourable for blast 

infection. New lesions will appear on leaves about 10 days after the favourable 

period, when they will be visible in ordinary field inspections. These criteria have 

been widely accepted. Kim et al (1987) have written a computer programme to 

predict blast occurrence based on microclimatic events and was tested as an onsite 

microcomputer in upland and flooded field plots in 1984 and 1985. Lee et al (1989) 

used primary meteorological factors related to outbreaks of blast for forecasting rice 

leaf blast. 

Padmanabhan (1965) studied the forecasting methods based on disease 

development linked to weather conditions, and found low temperatures and high 

humidity favoured disease progress. According to him, forecasting of the disease 

can be attempted on the basis of minimum night temperature of 20 - 26 o C in 

association with a high relative humidity range of 90 % and above lasting for a 

period of a week or more during any of the susceptible phases of growth, viz., seedling 

stage, post-transplanting tillering stage, and at neck emergence. The timings for 

spraying the crop with fungicides for direct control of the disease could be fixed on 

the basis of these data. In areas where meteorological data are not available it has 

been suggested by him that the crop should be sown in small test plots under heavy 

fertilization and watched regularly. As soon as first lesions appear on the test plants 

the information can be relayed to cultivators for necessary action. 

El Rafaei (1977) developed an equation to forecast blast incidence five days 

ahead by using dew period, number of lesions and number of air-borne spores. A 

simple method was devised for forecasting the epidemic outbreak of blast in the 

plains. With the help of “trap” plots of susceptible varieties, successful warnings were 

given at least 10 to 15 days in advance of the outbreak in farmers’ fields (Muralidharan 

and Venkatrao, 1980). Prasad Rao et al (1999) stated that the prediction equations 

derived through multiple regression analysis would be useful in forecasting blast 

severity and minimizing fungicidal spray as a component in integrated disease 

11


management programme in the northern region of Kerala. Kapoor et al (2000) 

developed rice leaf blast rules for Kangra district of Himachal Pradesh based on 

weather parameters. 

Models, validation and operational use for rice blast 

Several computer simulation models have been developed (Ota, 1982; 

Hashimoto et al., 1984; Takai et al., 1985; Gunther, 1986; Ishiguro, 1986). Surin et 

al (1991) stated that leaf blast incidence can be estimated by disease severity or by 

incidence on the top four leaves. However, disease severity on leaf 3 was found to be 

most representative of average severity at all leaf positions and therefore disease 

severity on a single leaf rather than on the whole plant is easier to do and saves a lot 

of time. Hence these models cannot sufficiently quantify the dispersion and deposition 

of the blast fungus spores in the rice canopy; quantitative data on the distribution of 

susceptible / non-susceptible tissue and inoculum are still lacking (Gunther, 1986). 

The first leaf blast model, BLASTL, a systemic analytic model that simulates 

the epidemic was developed in Japan (Hashimoto et al., 1982, 1984). This model 

has been verified under field experiments for disease progress and used to predict 

exact time for the fungicide application. A polycyclic model PYRICULARIA, was 

developed by Gunther (1986) using information from the literature. PYRICULARIA 

was modified by Tastra et al (1987) for upland rice farming in Indonesia and was 

subsequently called PYRNEW. The BLASTAM system (Koshimizu, 1983, 1988; Hayashi 

and Koshimizu, 1988) predicted weather conditions favourable for infection using 

weather data of the ‘automated meteorological data acquisition system’ (A Me DAS) 

via telephone modem in real time. BLASTAM can be used to predict the onset of leaf 

blast epidemic and the time for the first fungicidal application. 

Later, LEAFBLAST was developed (Choi et al., 1988) using data from the 

growth chamber experiments and the literature. This model consists of modules that 

compute spore germination, infection, latent period, lesion growth and spore 

production, dispersal and deposition, as affected by weather factors. Another model 

YYJM, was prepared to simulate a leaf blast (Magnoporthe grisea) epidemic under 

the conditions of Jilin Province in China, in order to forecast out breaks and 

requirements for chemical control (Anonymous, 1990). Ishiguro and Hashimoto (1991) 

developed computer based forecasting model of rice blast epidemics in Japan. The 

disease severity of leaf blast on photosynthesis and crop growth of rice crop was 

modeled by Basitaans (1991). Teng et al (1991) analysed blast pathosystem viz. 

BLASTL, BLASTCAST, simulation model PYRICULARIA and panicle blast pathosystem 

simulation model PBLAST to guide modeling and forecasting rice blast. Later, Kim 

and Kim (1993) developed a dynamic simulation model EPIBLAST for quantitative 

forecasting of the incidence of the leaf blast disease. 

Two other simulation models, CERES-RICE (a rice growth simulation model) 

and BLASTSIM (a rice leaf blast epidemic simulation model), were coupled by 

considering the effects of leaf blast on rice leaf photosynthesis and biomass production 

12


(Luo and Teng, 1994 and Luo et al., 1997). Regression equations used as empirical 

models to predict rice blast caused by Pyricularia grisea on cv. Jinheung at Icheon, 

South Korea, and on cvs. IR50 and C22 at Cavinti, Philippines, were generated, 

using weather factors identified by the WINDOW PANE program which showed that 

several weather factors were highly correlated with the disease variables (Calvero 

et al., 1996, 1997). Huang et al (1999) studied a decision model for the integrated 

control of rice blast based on the data of its occurrence, climatic factors, yield loss, 

and the economic benefit of fungicide in China. 

The first developed model in Japan for panicle blast was from Takahashi 

(1958). He treated a spore deposition and penetration as a stochastic process and 

subdivided each panicle into small infection site units. This model does not account 

for secondary infections. Hori (1963) and Kim (1982) attempted the forecasting of 

neck or panicle blast based on the number of diseased leaves per hill. The most 

comprehensive stochastic simulation model-panicle blast (PBLAST) was developed 

and verified by Ishiguro (1986), and Ishiguro and Hashimoto (1988, 1990). 

A Me DAS weather data, data on host development, time of cultivation practices, 

and number of spores formed on leaf lesion were used as input for PBLAST. It provides 

useful information for understanding patho-systems of panicle blast and improving 

the method of fungicidal application. 

A simple method was devised for forecasting the epidemic outbreak of blast 

in plains. With the help of ‘trap’ plots of susceptible varieties, successful warnings 

were given at least 10 to 15 days in advance of the outbreak in farmers’ fields 

(Muralidharan and Venkatrao, 1980). Manibhushan Rao and Krishnan (1991) 

developed a simulation model with computer based forewarning system (EPIBLA) to 

simulate the incidence and progress of rice leaf blast in the fields. EPIBLA (EPIdemiology 

of BLAst) is designed to forecast leaf blast progress over a period of seven days in 

advance. 

13


V. APPROACHES FOR DEVELOPING FOREWARNING MODELS 

The study was intended to develop prediction models and to forewarn the 

rice farming community about the conditions, which were sufficiently favourable for 

blast disease development and application of control measures at proper time. Once 

the models with reference to the different phases of blast are validated, the operational 

forewarning systems would be generated for use in agro-met decisions, which will 

have direct effect on the farmers in judicious use of fungicides for blast control and 

finally result in the economic gain to the farming community. 

For this purpose, the historical data related to blast severity in the country was 

compiled and its analysis helped in studying the long-term effects of interaction of 

climate and crop-pest-disease development. Climatic data was also helpful in finetuning 

the historical information with current situations. 

Blast severity in relation to weather was monitored in the experimental fields 

at Directorate of Rice Research, Hyderabad, and at CSK Himachal Pradesh Krishi 

Viswa Vidyalay, Palampur, and farmer’s fields both in Andhra Pradesh and Himachal 

Pradesh, with different dates of sowing on blast susceptible varieties. Observations 

both on leaf and neck blast were recorded at regular intervals, starting from the date 

of initial observation of the symptoms till the maximum disease and / or at the 

decline of the disease severity. The percentage of the disease severity over time was 

computed to arrive at disease progress curve. The meteorological data was recorded 

that included daily temperatures, relative humidity, sunshine hours, rainfall and intensity 

of the rainfall and also other factors like duration and amount of the leaf wetness 

caused by dew / drizzle were recorded. Analysis of both disease and weather data 

was taken up to identify the significant weather factors conducive for spread, incidence 

and intensification of the different phases of the blast disease. All the data collected 

including meteorological data were computerized and subjected to step-wise 

regression analysis for the development of rice blast predicting model(s). 

14

1. HISTORICAL DATA ON BLAST AND WEATHER IN HIMACHAL PRADESH 

From the historical data on blast and meteorological data in Himachal 

Pradesh, where the blast is endemic, the pattern of progress of blast showed that all 

the disease progress curves were more or less sigmoid. It revealed that perception 

threshold of leaf blast was less than 5 %. The analysis of average daily weather 

variables of a week earlier to these threshold levels of years 1991 to 2000 (Table 3) 

showed that maximum temperature of 23 to 28 o C, minimum of 17.6 to 24 o C, RH of 

more than 80 % with exception during 1997 and 1998, rainfall of more than 3 mm 

per day and more than 4 rainy days per week were critical in the progress of rice leaf 

blast. Similar information has been generated from the studies of rice blast during 

1996 to 1999 under the project funded by DST and resulted in the development of 

rice blast rules. These historical databases of rice blast and weather parameters also 

confirm rice blast rules. The available neck blast data of 1997–2000 (Table 4) revealed 

that maximum temperature of 25 to 26.4 o C, minimum of 15.6 to 18.4 o C, RH > 59 

%, rainfall more than 1 mm per day and > 3 rainy days per week were critical for 

neck blast. The favourable weather for rice leaf blast development was invariably 

available in the first fortnight of July, and maximum disease (Y max ) reached from 2 nd 

fortnight of August to first fortnight of September. 

Table 3. Average daily weather conditions of a week earlier to the 

threshold level of rice leaf blast 

Weather factors 

Year 

1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 

Temperature(Max) ( o C) 26.3 24.4 25.0 25.7 25.7 27.4 26.4 27.7 25.1 23.5 

Temperature(Min) ( o C) 20.4 17.6 19.9 20.1 20.1 20.2 20.9 19.5 19.9 19.4 

Relative Humidity (%) 81 82 83 82 82 76 85 76 86 89 

Rainfall (mm) 31.5 19.3 27.7 13.1 27.8 3.2 8.8 9.0 39.3 30.4 

Rainy days / week 6 5 6 7 6 4 5 7 5 7 

Table 4. Average daily weather conditions of a week earlier to the 

threshold level of rice neck blast 


Year 


1997 1998 1999 2000 

Temperature (Max) ( o C) 25.9 26.4 26.0 26.3 

Temperature (Min) ( o C) 16.1 16.5 18.4 15.6 

Relative Humidity (%) 67 78 82 59 

Rainfall (mm) 0.66 7.29 12.77 0.00 

Rainy days / week 3 5 4 0 

15


Pooled regression models for leaf blast at Palampur during different years 

were developed from the historical data for the years 1991-1996, 1997-2000, 1991- 

2000 and 2001 (Table 5). Out of all these equations, the regression equation of the 

duration 1997-2000 (Y = -20.53+3.16X 1 –7.28X 2 +0.60X 3 +0.54X 4 -0.36X 5 +1.60X 6 ) 

was observed to be the best fit equation. Validation of this model was made on the 

data of 2001 and found more realistic on farmers’ field data (Fig.3). 

Table 5. Pooled regression models for leaf blast at Palampur during different 

years developed from the historical data 

Year Equation R R 2 

1991 – 1996 Y = 5.66+3.39X -9.75X +1.09X +0.47X - 0.25X -1.13X 1 2 3 4 5 6 0.49 0.24 

1997 – 2000 Y = -20.53+3.16X -7.28X +0.60X +0.54X - 0.36X +1.60X 1 2 3 4 5 6 0.64 0.41 

1991 – 2000 Y=11.543+1.749X -5.950X +0.211X +0.696X - 0.233X +1.760X 1 2 3 4 5 6 0.60 0.36 

2001 * Y = -2.41 + 0.26 X1 0.85 0.72 

Y = Disease variable, X = T maximum, X = T minimum, X = RH maximum, X = RH minimum, 

1 2 3 4 

* X = Rainfall, X = Rainy days/week, X = Leaf wetness, R = Multiple correlation coefficient, 

5 6 1 

R2 = Coefficient of determination 

16 

Fig. 3. Validation of leaf blast model for Palampur and 

Farmers' field (Pharer) data 

Y = -20.53+3.16X1-7.28X2+0.60X3+0.54X4-0.36X5+1.60X6 

Year 2001, Palampur 

Y = -20.53+3.16X1-7.28X2+0.60X3+0.54X4-0.36X5+1.60X6 

Year 2001, Pharer

2. MODELS DEVELOPED BY DRR, HYDERABAD AND CSK HPKV, PALAMPUR 

Progress of rice blast in Kharif, 2001 to 2004 was continuously monitored in 

relation to weather, by laying the experiments in the experimental fields and taking 

observations in the farmers’ fields, both in Andhra Pradesh and Himachal Pradesh. 

Andhra Pradesh: 

Observations taken on blast progress at experimental field, Directorate of 

Rice Research (DRR), Rajendranagar, Hyderabad and farmers’ fields, Medchal, 

indicated that the disease was severe in late sown crop, when compared to the early 

sown crop. However, the neck phase of the disease was high in early sown crop at 

farmers’ fields. 

Experimental field, DRR, Hyderabad 


A local susceptible variety, HR 12, was chosen at DRR, with three dates of 

sowings for the conduct of the trial. Observations both on leaf and neck blast were 

recorded at 12 points of 1 m 2 area marked at random. In each 1 m 2 area, 10 

randomly selected hills were labeled and observations were taken at 3 to 4-day 

interval, starting from the date of initial observation of the symptoms till the maximum 

disease and / or at the decline of the disease severity. 

Blast observations being taken at experimental field, DRR, Hyderabad 

Sowings made after June exhibited severe incidence of blast, and its 

development was fast and reached maximum in late sowings, viz., 1 st and 2 nd fortnight 

of July, in the experimental field at DRR, Hyderabad. But, at farmers’ fields, sowings 

made in the last week of June showed maximum incidence of blast compared to the 

early sowings made in the last week of May and 1st week of June. However, the 

results in general, both at the experimental fields and farmers’ fields in Andhra Pradesh 

indicated that blast severity was high in the late sown crop, when compared to the 

early sown crop. Reports of such high incidence of blast in the late sowings were also 

earlier made in Arunachal Pradesh (Chaudhary and Vishwadhar, 1988), West Bengal 

17


(Pramanick et al., 1990), hills of Uttar Pradesh (Bhatt, 1992) and Nagaland (Sharma 

et al., 1993), though the sowing months varied from location to location depending 

up on the elevation, etc., at the respective locations. 

Results of kharif, 2001 indicated that blast severity was high in the late sown 

crop, when compared to the early sown crop. The average leaf blast infection in the 

three different sowings was 58.8, 63.1 and 80.1 %, respectively. At Rajendranagar, 

the average daily weather conditions of a week earlier to the date of observation of 

high blast severity (17 th Oct.2001) indicated that the minimum temperature (T Min ) was 

22.2 o C and maximum (T Max ) was 29.8 o C , with a rainfall (Rf) of 27.4 mm and 4 rainy 

days with in a week (R W ), with morning Rh (Rh M ) of 92 % and evening Rh (Rh E ) of 58 

%, were found congenial for blast disease development. Neck blast infection averages 

in three different sowings were 57.0, 49.9 and 59.7 %, respectively. Before a week 

earlier to the neck blast observation (24 th Oct. 2001), the daily averages of weather 

conditions were: T Min 21 o C , T Max . 30.4 o C , Rf 31.1mm, 2 rainy days with in a week, 

morning Rh of 91 % and evening Rh of 61 %. It is possible that these weather conditions 

might have favoured the increase in the neck infection. 

Analysis of daily weather variables of 2002 during preceding week (Table 6) 

of high disease intensity showed that the following weather conditions were found 

favourable. T Max : 30.2 0 C, T Min : 21.7 0 C. Rh M : 99.9 %, Rh E : 64.6 %, Rf: 0.03 mm and 

R W : 1. All these conditions led to the epidemic progress of leaf blast to the extent of 

10.26 % (Fig. 4), and the relationship was best explained by the following equation 

Y = 43.287 - 2.859** T Min + 0.354** Rh E + 1.635** Lw . The combined effects of T Max : 

28.7 0 C, T Min : 15.0 0 C. Rh M : 98.3 %, Rh E : 40.6 % resulted in significant effect on neck 

blast severity (10.84 %), and this relationship between neck blast incidence and the 

weather parameters resulted in the following equation: Y = 128.885 -1.968** T Min - 

0. Rh M + 0.18* Rh E + 1.499*L w . Observations made by Bhatt (1992) revealed that 

minimum temperature between 15-20 0 C, daily average temperature between 

22-25 0 C and the difference between maximum and minimum temperature between 

10-15 0 C, more days with Rh 90 % or above, higher rainfall and more number of 

rainy days appeared important factors for the development of blast. 

During 2003, weather in general was more favourable for the disease 

development compared to kharif, 2002, which was officially acknowledged as “the 

first-ever all-India drought year” since 1987. Analysis of daily weather variables of a 

week earlier (Table 7) to the high disease intensity in 2003 were T Max : 29.3 0 C, T Min : 

22.3 0 C. Rh M : 100%, Rh E : 80.1 %, Rf: 5.3 mm and R W : 6. These conditions led to the 

high leaf blast severity (77.1 to 92.9 %) at different sowings (Fig. 4), the relationship 

of which was shown in the equation as: Y = -1537.742 + 13.589** T Max + 13.413** 

T Min - 6.197** R W + 9.431** Rh M + 1.758 L W . Neck blast was also maximum (26.3 to 

34.7 %) in kharif, 2003, with favourable weather conditions, viz., T Max : 29.5 0 C, T Min : 

18.2 0 C. Rh M : 100 %, Rh E : 68.3 %, Rf: 1.6 mm and R W : 5. However, the regression 

equation (Y = - 155.37 + 8.493** T Max - 8.061** T Min + 3.82* Rw + 0.924 Rh E ) 

18

Table 6.Average daily weather conditions of a week, preceding to the date of 

observation of blast at the experimental field, DRR, Kharif, 2002 

Date of 

observation 

Fig. 4. Progress of Blast at Experimental Field, DRR 

Temperature ( 0 C) 

T Max 

T Min 

Rainfall 

(mm) 

Rainy 

days/ 

Week 

(No.) 

Relative humidity (%) 

Rh M 


16-Sep. 30.8 22.0 0.0 0 95.1 57.1 0.1 

20-Sep. 31.2 22.1 0.0 0 98.9 61.3 0.3 

24-Sep. 31.6 22.4 0.0 0 98.7 61.7 0.4 

28-Sep. 32.5 22.8 0.0 0 96.3 57.6 1.9 

02-Oct. 32.3 20.8 0.0 0 98.3 51.4 0.5 

06-Oct. 32.7 19.9 0.0 0 100.0 45.1 0.5 

10-Oct. 32.0 20.5 0.0 0 100.0 50.7 2.1 

14-Oct. 30.2 21.7 0.03 1 99.9 64.6 2.1 

18-Oct. 28.8 22.0 0.03 1 100.0 74.4 0.3 

22-Oct. 30.0 20.9 0.0 0 100.0 60.6 0.0 

26-Oct. 30.9 18.3 0.0 0 100.0 44.4 0.0 

30-Oct. 29.7 17.9 0.0 0 100.0 49.7 0.0 

03-Nov. 28.8 16.8 0.0 0 98.7 47.1 0.0 

07-Nov. 28.7 15.0 0.0 0 98.3 40.6 0.0 

Rh E 

Leaf 

Wetness 

19


derived shows that the temperature and rainy days per week were only the most 

significant factors for the high neck incidence. Though, node blast was negligible in 

kharif, 2002, it was to the extent of 17.4 to 22.6 % in 2003, as weather was in 

general significantly favourable during the crop growth, for all the phases of the 

disease, viz., T Max : 27.6 0 C, T Min : 17.2 0 C. Rh M : 100 %, Rh E : 82.7 %, Rf: 13.3 mm and 

R W : 7. 

All these conditions at the experimental field at DRR, led to the epidemic progress of: 

Ø Leaf blast during the last week of September and 1st fortnight of October 

Ø Neck and node blast during the last week of October and 1st week of November. 

Table 7. Average daily weather conditions of a week, preceding to the date 

of observation of blast at the experimental field, DRR, Kharif, 2003 

Date of 

observation 

20 


T Max 



(mm) 


days/ 

Week 

(No.) 


Rh M 

30-Aug 28.0 22.8 0.2 1 93.4 72.6 0.2 

2-Sept 28.6 22.9 0.2 1 92.1 69.4 0.1 

5-Sept 29.0 23.0 0.6 2 92.4 67.9 0.2 

8-Sept 29.1 23.0 1.0 4 94.0 67.4 0.3 

11-Sept 29.4 22.8 0.8 3 94.7 64.3 0.5 

15-Sept 28.8 22.6 0.0 0 92.9 66.3 1.9 

18-Sept 28.9 22.7 0.0 1 92.7 70.0 1.9 

22-Sept 29.7 23.1 0.8 2 96.1 71.1 2.5 

25-Sept 29.1 23.2 1.8 5 99.0 76.6 2.5 

29-Sept 28.8 22.8 5.6 7 100.0 82.7 10.6 

2-Oct 30.1 22.5 4.6 7 100.0 80.9 10.5 

6-Oct 30.8 22.0 3.3 6 100.0 76.3 1.6 

9-Oct 29.3 22.3 5.3 6 100.0 80.1 1.7 

13-Oct 29.6 21.6 2.1 4 100.0 71.0 0.4 

15-Oct 30.3 20.6 2.1 2 100.0 63.4 0.4 

16-Oct 30.6 19.8 0.0 1 100.0 57.0 0.4 

17-Oct 30.7 18.6 0.0 1 100.0 52.9 0.3 

20-Oct 30.5 19.0 1.1 2 100.0 56.3 0.5 

23-Oct 29.3 20.2 12.5 5 100.0 74.4 4.0 

27-Oct 27.6 19.5 13.3 6 100.0 79.4 3.9 

29-Oct 27.6 17.3 8.3 5 100.0 71.1 3.8 

31-Oct 28.5 17.2 1.6 5 100.0 64.9 0.4 

3-Nov 29.5 18.2 1.6 5 100.0 68.3 0.4 

5-Nov 29.9 19.4 1.6 5 100.0 69.9 0.4 

7-Nov 30.0 17.8 0.1 5 100.0 62.0 2.4 

Rh E 

Leaf 

Wetness

Farmers’ fields, Medchal 

Progress of blast was recorded during kharif, 2001 and 2002 at six farmers’ 

fields near Medchal (Dindighul, Railapur and Nagaloor) in Andhra Pradesh, where 

the local susceptible varieties, viz., BPT 5204 and JGL 1798 were sown. Leaf and 

neck blast observations were taken on 10 randomly tagged hills in a marked area of 

1 m 2 at each of the 25 points, in each of the six farmers’ fields. Initial observations 

were taken at the notice of the symptoms up to the maximum disease and / or at the 

decline of the disease, with four-day interval between the observations. 

At farmers’ fields in the year 2001, maximum leaf blast severity was 43.5% 

on 3rd October and the average weather variables preceding a week earlier to this 

high incidence was 28.50C maximum and 21.00C minimum temperature, 71.3 % Rh 

in the morning and 55.0 % Rh in the evening, 33.3 mm rain fall and 7 rainy days / 

Week (Table 8). High leaf blast severity and weather factors like minimum temperature 

and rainfall were best explained by an equation: Y = -19.618 + 1.241* T + Min 

0.348**R . In case of high neck blast incidence (55.9 to 70.4 %) on 30 f th October, the 

weather conditions, more precisely the minimum temperature of 18.20C might have 

favoured, though the daily average of relative humidity of the week earlier to this 

high neck infection was only to the extent of 51.4 %. However, the following regression 

equation, Y = - 259.522 + 8.86** T shows that the maximum temperature was 

Max, 

only the contributing factor for this neck incidence. 

Table 8. Average daily weather parameters of a week, preceding to the date of 

observation of blast severity at the farmers’ fields, Medchal, Kharif, 2001 

Disease/ 

Date of 

observation 


T Max 


Mean 


(mm) 


days/ 

Week 

(No.) 



Rh M Rh E Mean 

Leaf Blast 

28.09.2001 31.0 21.3 26.2 12.6 5 67.0 52.0 59.5 

03.10.2001 28.5 21.0 24.8 33.3 7 71.3 55.0 63.2 

09.10.2001 28.0 21.7 24.9 14.2 6 71.2 64.2 67.7 

Neck Blast 

22.10.2001 30.3 21.4 25.9 7.5 3 66.8 54.0 60.4 

27.10.2001 31.1 19.6 25.4 0 0 52.7 37.3 45.0 

30.10.2001 31.6 18.2 24.9 0 0 51.4 36.2 43.8 

02.11.2001 31.9 17.9 24.9 0 0 51.6 38.0 44.8 

During the year 2002, independent weather variables, viz., 21.30C minimum 

temperature, 90.4 % Rh and 66.3% Rh , 11.6 mm rainfall and 3 rainy days per 

M E 

week were found favourable (Table 9) for high leaf blast occurrence, at different 

sowings on 26th September in farmers’ fields (Fig. 5). The relationship of leaf blast in 

such weather conditions was explained by an equation, Y = - 88.021 + 0.925** RhM +2.908**R , which indicates the influence of morning relative humidity and rainy 

W 

21


days per week on leaf blast development. The combined effects of minimum 

temperature (19.1 0 C), Rh M (90.1 %), Rh E (52.9 %), rainfall (13.2 mm) and rainy day 

/ week (2), resulted in high neck incidence of 27.85 % on 30 th October (Fig. 5). But, 

the regression equation (Y = 34.684 - 1.403 T Min + 0.602**Rf) shows that the 13.2 

mm rainfall was only the significant weather factor for this neck incidence. However, 

when compared to the previous year, the incidence of both leaf and neck blast was 

low. While looking at the influence of weather factors on the development of blast in 

Jharkhand, Dubey (2003) noted that the low temperature and relative humidity were 

the favourable factors for neck infection. 

Fig. 5. Progress of Blast at Farmer's Fields, Medchal, Kharif, 2002 

The average weather conditions in the farmers’ fields at Medchal during 

2001 and 2002, in general, led to the epidemic progress of leaf blast during the last 

week of September and 1 st week of October, and neck blast during the last week of 

October. 

22 

Date of 

sowings at: 

Dindighul-1: 30 th May 

Railapur-1: 4 th June 

Nagaloor-1: 16 th June 

Dindighul-2: 6 th June 

Railapur-2: 5 th June 

Nagaloor-2: 29 th June


of observation of blast at the farmers’ fields, Medchal, Kharif, 2002 

Date of 

observation 


T Max 



(mm) 


days/ 

Week 

(No.) 


Rh M 

25-Aug. 28.5 20.9 17.0 4 91.9 77.9 

29-Aug. 28.0 20.4 17.4 4 95.4 75.3 

02-Sep. 29.9 21.3 5.0 1 91.9 73.6 

06-Sep. 28.0 21.2 9.9 3 94.4 82.6 

10-Sep. 28.9 20.7 12.1 3 96.1 75.1 

14-Sep. 30.1 20.5 2.6 1 94.0 65.6 

18-Sep. 31.8 21.1 0.0 0 85.3 61.7 

22-Sep. 31.5 21.0 11.6 2 84.0 64.3 

26-Sep. 32.2 21.3 11.6 3 90.4 66.3 

30-Sep. 32.8 21.4 0.0 1 88.7 58.1 

04-Oct. 33.9 21.0 0.7 1 86.0 55.1 

08-Oct. 34.9 21.2 0.7 1 85.7 52.9 

12-Oct. 34.2 21.5 0.4 2 88.3 64.4 

16-Oct. 30.8 21.0 12.9 6 94.4 84.6 

20-Oct. 28.4 20.2 27.1 6 95.6 79.3 

24-Oct. 31.1 19.1 13.2 2 90.1 52.9 

27-Oct. 32.2 18.4 0.0 0 86.7 47.6 

31-Oct. 30.9 17.4 0.0 0 84.7 50.6 

04-Nov. 31.3 16.5 0.0 0 82.0 45.7 

08-Nov. 31.7 16.4 0.0 0 79.6 44.6 

12-Nov. 31.5 16.4 0.0 0 80.8 45.1 

Regression equations 


The influence of different weather parameters viz., maximum and minimum 

temperatures, morning and evening relative humidity, rain fall, rainy days per week 

and leaf wetness on the leaf blast severity and neck blast incidence were worked out 

and the regression equations (Table 10) were generated, by step-down regression. 

The analysis indicated the influence of weather conditions during crop growth 

on leaf blast severity and neck blast incidence. At the experimental field, DRR, minimum 

temperature and rainy days / week had shown a profound influence on both leaf 

and neck phase of the disease, followed by the other weather conditions like, leaf 

wetness, maximum temperature and relative humidity, while the intensity of rain fall 

had no significance. However, at the farmers’ fields, blast development was influenced 

more by rainfall rather than other climatic conditions, though temperature, morning 

relative humidity and rainy days / week had also shown significant influence on the 

disease development. 

Rh E 

23


Table 10. Regression equations based on step-down regression of one week 

preceding weather data (Andhra Pradesh) 


LEAF BLAST 

DRR, 2002 Y = 43.287 - 2.859** T + 0.354** Rh + 1.635** L Min E W 0.854 0.729 

DRR, 2003 Y = -1537.742 + 13.589** T + 13.413** T - 6.197** R Max Min W 

+ 9.431** Rh + 1.758 L M W 

0.858 0.735 

Medchal, 2001 Y = -19.618 + 1.241* T + 0.348** Rf Min 0.678 0.459 

Medchal, 2002 

NECK BLAST 

Y = - 88.021 + 0.925** Rh +2.908** R M W 0.785 0.616 

DRR, 2002 Y = 128.885 -1.968** T - 0.98 Rh + 0.18* Rh + 1.499* L Min M E W 0.945 0.894 

DRR, 2003 Y = - 155.37 + 8.493** T - 8.061** T + 3.82* R + 0.924 Rh Max Min W E 0.929 0.864 

Medchal, 2001 Y = - 259.522 + 8.86** TMax 0.376 0.141 

Medchal, 2002 Y = 34.684 - 1.403 T + 0.602** Rf Min 0.673 0.453 

T Max = Maximum temperature, T Min = Minimum temperature, Rf = Rainfall, R W = Rainy days per week, 

Rh M = Relative Humidity in morning, Rh E = Relative humidity in evening, L W = Leaf wetness, Y = Disease 

variable, R = Multiple correlation coefficient, R 2 = Coefficient of determination. 

In general, the step-down multiple regression analysis shows, the minimum 

temperature, morning relative humidity, rainy days per week, followed by evening 

relative humidity, leaf wetness, maximum temperature and rainfall were found 

significant weather factors for the leaf blast development, while both minimum and 

maximum temperature, rainy days per week followed by the intensity of rainfall and 

leaf wetness were found to have significant effect on neck blast incidence. 

This clearly indicates that minimum temperature, morning relative humidity 

and rainy days / week increased the leaf blast severity, while the minimum and 

maximum temperature and rainy days / week increased the neck infection under 

natural conditions, and so these may be selected as contributing factors for prediction 

of leaf and neck infection in nature. 

Himachal Pradesh: 

Field trials on rice blast progress in relation to weather were laid out at three 

locations in Himachal Pradesh, viz., Palampur, Malan and at the farmers’ fields (Arla 

and Pharer). At Palampur, trials were conducted at different dates of transplanting on 

a susceptible variety Himalaya 2216, at Malan only one date of transplanting was 

done and at farmers’ fields, a local susceptible variety was transplanted. Data on 

leaf blast and neck blast were recorded as per Standard Evaluation System for Rice 

(IRRI, 1996). 

24

CSK HPKV, Palampur, Malan and farmer’s fields 

At Palampur, during 2001 the daily weather variables of week earlier to 

threshold levels revealed average daily maximum temperature was 28 0 C and 

minimum 15.4 to 20.1 0 C, Rh maximum > 80 %, Rh minimum ranged from 47 to 

76%, rainfall from 5.40 to 6.81 mm and rainy days 5 to 7 days / week in case of first 

two dates of transplanting, but one rainy day in third date of transplanting. Similarly, 

hours of Rh >85 % was maximum (14hrs.) in early two dates of transplanting as 

compared to 0.3 hr in third date of transplanting. In spite of some unfavourable 

critical weather variables viz. rainy days per week and hours with Rh more than 85 

%, leaf blast in III rd date of transplanting progressed considerably which may be due 

to availability of host and inoculum. Rice leaf blast severity was more in II nd date of 

transplanting at Palampur. Whereas during 2002 the analysis of daily weather 

variables of a week earlier to threshold levels at all locations showed that maximum 

temperature of 23.4 to 28.8 0 C, minimum 20 to 26 0 C, Rh maximum 78 to 96 %, Rh 

min 71 to 95 %, rainfall 5 to 310 mm, rainy days per week 3 to 7 and number of 

hours having relative humidity more than 85 % was more than 18 hours (Table 11). 

All these conditions except minimum temperature especially at farmers’ fields, which 

was at higher side led to epidemic progress of rice leaf blast during August month. 

Unlike rice leaf blast, weather conditions for neck blast were not favourable for neck 

blast infection and development. 

Table 11: Average daily weather conditions of a week earlier to the threshold 

level of rice leaf blast during kharif 2001 and 2002 


Palampur 

I II III 

Malan Arla 

2001 2002 2001 2002 2001 2001 2002 2001 2002 

T ( Max oC) 27.4 23.8 28.2 23.4 26.9 29.3 24.5 27.8 28.8 

T ( Min oC) 20.1 23.0 18.4 22.7 15.4 20.8 20.1 20.7 26.0 

Rh (%) Max 87 94 88 96 81 86 95 87 78 

Rh (%) Min 75 93 76 95 47 53 88 60 71 

Rf (mm) 5.9 5.0 6.8 11.4 5.4 115 13.9 16.0 310 

RW 5 3 7 7 1 6 6 7 3 

Rh > 85% (Hrs.) 14 22.5 14 24 0.3 11 18.5 12 0 

I, II and III: Three dates of transplanting. 

Development of Regression Models 


At Palampur location during 2001, the first two dates of transplanting resulted 

best fit equations (Table 12). In I st date of transplanting, the combined effects of T Max , 

Rh Min and rainy days / 4 days resulted significant effects on leaf blast whereas in 2 nd 

date of transplanting, T Min , Rh Max , hours with Rh > 85% and rainfall were found good 

25


predictors. At Malan, it was mean RH, rainfall and rainy days per week and at Pharer, 

it was Rh Min and rainy days per week. The regression equations for neck blast revealed 

significant combined effects of Rh Max and Rh Min on neck blast on first date of 

transplanting, T Max and Rh Min in II nd date of transplanting and T Max and T Min in III rd date 

of transplanting at Palampur. At Malan, it was T Max and rainfall which influenced neck 

blast development. 

Rice blast and weather data of kharif 2002 were subjected to step wise 

regression analysis. The regression equations so developed (Table 13) under 

Palampur-I conditions with coefficient of determination (R 2 ) more than 0.60 are given. 

Leaf wetness period, a very critical weather variable resulted in 0.63 correlation with 

leaf blast and R 2 was 0.40 whereas, during 2001, it was 0.85 and 0.72, respectively. 

During this year, leaf wetness period (X 3 ) and minimum temperature (X 10 ), resulted in 

R = 0.78 and R 2 = 0.61. The rice blast data and observatory cum thermo-hygrograph 

weather data resulted in four regression equations with R = 0.95 and R 2 = 0.90. The 

critical weather variables were rainfall, leaf wetness period (L W ), hours with relative 

humidity (Rh) > 85 %, rainy days (RD)/4 days, minimum relative humidity (Rh Min ), 

maximum relative humidity (Rh Max ) and maximum temperature (T Max ). 

Similarly, when analysed with micro weather station data, rainfall (X 1 ) was 

found more critical weather condition followed by average temperature (X 2 ). However, 

the best fit regression equation was found to be Y = 74.981- 

3.468X 2 +0.221X 1 +0.492X 3 with R = 0.90 and R 2 = 0.81. This remained constant 

even after addition of variables X 4 (average relative humidity) and X 6 (RD / 4 days). 

Table 12. Regression models for leaf and neck blast on cultivar Himalayan 

2216 at different locations during 2001 


LEAF BLAST 

Palampur (a) Y = -120.141+5.497X -0.608X +7.479X 1 4 7 0.83 0.70 

(b) Y = -292.204+3.406X +3.576X -2.803X -5.075X 2 3 5 6 0.85 0.72 

(c) Y = -447.384+12.129X -13.649X +4.835X 1 2 3 0.49 0.24 

Malan Y = -127.61+2.642X -0.114X -5.419X 8 6 9 0.99 0.99 

Pharer 

(Local var.) 

Y = -57.148+4.441X -7.117X 1 9 0.99 0.99 

NECK BLAST 

Palampur (a) Y = 24.83-0.29X -1.52X 3 4 0.86 0.74 

(b) Y = 39.07-0.95X -0.68X 1 4 0.89 0.79 

(c) Y = 5.21-0.42X +0.03X 2 3 0.88 0.78 

Malan Y = 9.62-0.28X -0.08X 1 6 0.99 0.99 

a = Ist D.O.T, b = IInd D.O.T, c = IIIrd D.O.T. 

X = T maximum, X = T minimum, X = Rh maximum, X = Rh minimum, X = hours Rh > 85 %, 

1 2 3 4 5 

X = Rainfall, X = Rainy days/4 days, X = average Rh, X = Rainy days/week, Y = Disease variable, 

6 7 8 9 

R = Multiple correlation coefficient, R2 = Coefficient of determination. 

26


In the second date of transplanting at Palampur, Rh Min (X 7 ) and Rh Max (X 8 ) were 

important weather variables in predicting leaf blast. However, the disease pressure 

was very low. At Malan location, minimum temperature (X 10 ) and Rh Max (X 8 ) were 

found more important variables in predicting leaf blast with R 2 = 0.97. At third location 

(Farmers’ fields), minimum temperature (X 10 ) and rainfall (X 1 ) were critical variables 

in predicting leaf blast, though the disease pressure was very low due to prolonged 

drought conditions. 

Due to low neck blast disease pressure, only data of Palampur-I and Malan 

could be analysed. At Palampur, T Min (X 10 ) and Rh Max (X 8 ) were found more critical with 

R 2 as high as 0.99. At Malan location, maximum temperature (X 9 ) and minimum 

temperature (X 10 ) were critical factors in predicting neck blast. 

Fig. 6. Validation of blast models for Malan data, Kharif, 2002 

Validation of model (s) 

Validation of model Y = -20.53+3.16X 1 –7.28X 2 +0.60X 3 +0.54X 4 - 

0.36X 5 +1.60X 6 was made on the data of 2001 and found more realistic on farmers’ 

field data (Fig. 3). Regression model was also developed based on the weather and 

disease data of the year 2001 at Palampur. However, at Palampur location during 

2001, the first two dates of transplanting resulted in best fit equation (Fig. 3). Regression 

models developed from the rice blast and weather data of 2001 were validated in 

the 2002 data (Fig. 6). The observed and predicted data are plotted. Overall validation 

of models was not very encouraging. An attempt was also made to validate the neck 

blast models developed during 2001 at Palampur and Malan locations (Fig. 6). The 

prediction was not encouraging at Palampur, however, some degree of agreement 

was found between observed and predicted neck blast incidence at Malan location. 

27


Table 13. Regression models for leaf/neck blast on cultivar Himalayan 2216 

at different locations during 2002 

Year Equation R R2 Palampur – I Y = 93.45 - 4.08X 10 + 0.197X 3 0.78 0.61 

(w.r.t. Thermo - 

hygrograph and 

Observatory 

weather data) 

Y = 9.5 - 0.453X 5 + 0.136X 1 0.88 0.77 

Y = 3.47 - 0.36X 5 + 0.174X 1 + 0.335X 3 0.92 0.85 

Y = 4.221 - 0.375X 5 + 0.201X 1 - 0.978X 6 + 0.512X 3 0.93 0.87 

Y = -17.529+0.273X 7 -0.571X 5 +0.213X 1 -1.554X 6 + 0.703X 3 0.95 0.88 

Y = 26.712-2.352X 8 +2.082X 7 -0.19X 5 +0.248X 1 -0.909X 6 0.95 0.90 

+0.523X 3 

Y = 33.607-0.419X 9 -2.374X 8 +2.139X 7 -0.189X 5 + 0.245X 1 0.95 0.90 

-0.941X 6 +0.518X 3 

Y = 28.919-2.585X 9 +2.172X 10 -2.068X 8 +1.904X 7 - 0.228X 5 0.95 0.90 

+0.236X 1 -1.063X 6 +0.561X 3 

Palampur - I Y = 118.763 - 5.078X 2 + 0.114X 1 0.85 0.73 

(w.r.t. Micro - 

weather station 

data) 

Y = -6.573 + 0.297X 1 - 0.046X 4 + 0.999X 3 0.79 0.63 

Y = -14.995 + 0.339X 1 + 1.256X 6 + 0.936X 3 0.80 0.64 

Y = 74.981 - 3.468X 2 + 0.221X 1 + 0.492X 3 0.90 0.81 

Y = -14.679 - 0.003X 4 + 0.339X 1 + 1.25X 6 + 0.938X 3 0.80 0.64 

Y = 84.499 - 3.506X 2 - 0.095X 4 + 0.226X 1 + 0.533X 3 0.90 0.82 

Y = 81.099-3.46X 2 -0.083X 4 +0.238X 1 +0.339X 6 +0.522X 3 0.90 0.82 

Palampur- II Y = 0.679 - 0.04X 8 + 0.328X 7 0.88 0.77 

Y = 0.691 - 0.0542X 8 + 0.047X 7 + 0.002X 1 0.91 0.82 

Y = 1.118 - 0.078X 8 + 0.065X 7 + 0.005X 5 + 0.003X 1 0.93 0.86 

Y = 1.389-0.014X 10 -0.085X 8 +0.073X 7 +0.005X 5 + 0.003X 1 0.94 0.88 

Y = 2.387-0.037X 10 -0.136X 8 + 0.117X 7 + 0.012X 5 + 0.003X 1 0.95 0.90 

+0.026X 6 

Malan Y = 302.399 - 5.245X 10 - 2.063X 8 0.98 0.97 

Y = 464.706 - 2.492X 9 - 4.965X 10 - 3.198X 8 0.99 0.99 

Y = 500.428 - 2.896X 9 - 4.362X 10 - 3.554X 8 - 0.222X 11 0.99 0.99 

Arla Y = 4.984 - 0.161X 10 - 0.003X 1 0.97 0.94 

(Farmer’s field) Y = 2.337 + 0.127X 9 - 0.191X 10 - 0.003X 1 0.99 0.97 

Y = 3.144 + 0.144X 9 - 0.199X 10 - 0.013X 8 - 0.003X 1 0.99 0.98 

Palampur - I 

(Neck blast) 

Y = 46.625 - 0.558X - 0.469X 10 8 0.99 0.99 

Malan 

(Neck blast) 

Y = 7.338 - 0.51X + 0.399X 9 10 

Y = -32.135 + 0.482X - 1.121X 7 12 

0.99 

0.91 

0.99 

0.83 

Palampur-I = I st D.O.T., Palampur-II = II nd D.O.T., X 1 = Rainfall, X 2 = Temperature average, X 3 = Leaf 

wetness, X 4 = Rh average, X 5 = Hours with Rh > 85%, X 6 = Rainy days/4 days, X 7 = Rh minimum, X 8 = 

Rh maximum, X 9 = Temperature maximum, X 10 = Temperature minimum, X 11 = Hours with Rh > 90% 

and X 12 = Rainy days/week, R = Multiple correlation coefficient, R 2 = Coefficient of determination. 

28

3. VALIDATION OF MODELS DEVELOPED AT DRR 

The regression equations developed for the data of 2001, 2002, 2003 

revealed that independent variables influence the rice leaf blast differentially during 

different years, thereby making it difficult to select a good equation. For example, 

during 2002 the combined linear effects of T , Rh and Leaf wetness (L ) contributed 

Min E W 

to the variation in rice blast severity, whereas during 2003, it was T , T , R and Rf. 

Max Min W 

Similarly, other weather variables in different combinations effected rice blast severity 

in different years of study. However, T was found an important predictor in 2002 

Min 

and 2003. Therefore, the models developed were validated to find a best-fit model 

for better forewarning. Out of all the equations, the regression equation developed 

based on the pooled leaf blast and weather data of the duration 2001-2003 at 

Medchal was observed to be the best-fit equation, which was validated on 2002- 

2003 data of DRR. Tsai and Su (1984) also stated that combined data of various 

years resulted in equations, which gave better agreement between observed and 

predicted values than equations derived from data of individual years. 

The best validation model used for forecasting the disease is: 

LB = 14.5 + 0.90 LB + 0.10 T – 1.41 T –0.10 Rf + 1.20 R + 0.29 

N P Max Min W 

Rh –0.18 Rh M E 

R2 = 0.92 

Where: LB = Leaf blast in next week 

N 

LB = Leaf blast during previous week 

P 

T = Maximum temperature during previous week 

Max 

T = Minimum temperature during previous week 


Rf = Rain fall during previous week 

R = Rainy days during previous week 

W 

Rh = Morning relative humidity during previous week 

M 

Rh = Evening relative humidity during previous week 

E 

The validation models were developed for leaf blast at Medchal, Hyderabad, 

during 2001-2003 and at DRR, Rajendranagar during 2002-2003. The observed 

and predicted values were given in Fig. 7. 

The dependent neck blast variable and independent weather variables of the 

years 2002 and 2003 of DRR subjected to the validation on the three years blast / 

weather data (2001-2003) of Medchal resulted in best fit equation. Observed and 

predicted values of neck blast at DRR and Medchal are given in the Fig. 8. 

NB = -25.61 + 0 .79 NB + 1.8 T –2.5 T + .38 Rh N P Max Min E 

R2 = 0.90 

Where: NB = Neck blast in next week 

N 

NB = Neck blast during previous week 

P 

T = Maximum temperature during current week 

Max 

T = Minimum temperature during current week 


= Evening relative humidity during current week 

Rh E 


29


30 

Fig. 7. Observed and Predicted values of Leaf blast at Medchal 

during 2001-'03 and at DRR, Hyderabad during 2002-'03 

Medchal 

Correlation Coefficients 

Date of observations 

Correlation coefficients between leaf blast and weather parameters at Medchel, 

Hyderabad were assessed based on the best-selected validation model. Significant 

correlation was seen between the disease in the previous week and in the next week. 

Further, analysis of weather variables showed, morning and evening relative humidity 

and, rainy days per week were found the significant weather factors for the leaf blast 

development, while both minimum and maximum temperature, evening relative 

humidity were found to have significant effect on neck blast incidence. 

Blast P T Max T Min Rh M Rh E Rf R W 

Blast N 0.97** 0.10 0.29 0.34* 0.40* 0.21 0.62* 

*Significant at 5% level 

** Significant at 1% level 

DRR


This clearly indicated that relative humidity and number of rainy days / week 

increased the leaf blast severity, while the minimum and maximum temperature and 

evening relative humidity increased the neck blast under natural conditions, and so 

these may be selected as contributing factors for prediction of leaf and neck infection 

in nature. The present findings are supported by the results of some earlier workers 

(Govindasamy, 1964, Muralidharan and Venkatarao, 1980, Sharma et al, 1993, 

Dubey, 2003), who observed that the development of blast was favoured by 15- 

22.3 0 C temperature, more number of rainy days, higher rainfall and relative humidity. 

Their effect varied according to the location. 

Fig. 8. Observed and Predicted values of Neck Blast at DRR 

during 2002-'03 and at Medchal, Hyderabad during 2001-'03 

DRR 


Medchal 

31


VI. APPLICATION OF THE FOREWARNING MODEL OF DRR 

The best-fit equation developed for blast severity was used for forecasting the 

disease based on the preceding week weather variables and blast severity. The 

threshold level of predicted disease considered for spraying of fungicides was around 

10 %. Observations on leaf and neck blast were taken at regular intervals and 

predictions were made for each observation (Table 14 and 15). Based on the prediction 

of the disease severity, timely spray of the fungicides was suggested to control the 

disease (Table 16). Forewarning plots are compared with the plots with no forecast 

situation (untreated plots). 

Table 14. Observed and predicted leaf blast severity in treated plots of different sowings 

Date of 

1 

observation 

st Sowing: 21/6/2004 2nd Sowing: 06/7/2004 3rd Sowing: 20/7/2004 

Observed Predicted Observed Predicted Observed Predicted 

23.8.04 1.3 1.9 0.5 1.2 0.0 0.0 

28.8.04 3.6 5.8 2.0 4.4 0.0 0.0 

31.8.04 5.2 9.7 3.6 8.2 0.0 0.0 

03.9.04 9.4 13.7 5.0 9.8 1.4 6.5 

07.9.04 8.8 13.7 10.1 14.9 4.4 9.8 

10.9.04 8.2 13.1 9.5 14.2 11.5 16.0 

13.9.04 8.0 10.9 9.1 11.9 10.9 13.5 

16.9.04 7.8 12.2 8.8 13.1 10.6 14.7 

20.9.04 7.1 14.8 8.6 16.2 10.2 17.6 

23.9.04 6.6 13.8 8.0 15.1 9.5 16.4 

27.9.04 6.2 13.1 7.7 14.4 9.1 15.7 

30.9.04 6.0 12.8 7.4 14.0 8.8 15.3 

04.10.04 5.8 14.9 7.1 16.1 8.0 16.9 

07.10.04 5.0 13.6 6.5 15.0 7.5 15.9 

11.10.04 4.7 9.2 5.8 10.2 7.0 11.3 

14.10.04 4.5 10.3 5.4 11.2 6.5 12.2 

18.10.04 4.2 12.4 5.1 13.2 6.1 14.1 

21.10.04 3.9 13.7 4.8 14.5 5.6 15.2 

25.10.04 3.3 12.4 3.9 13.0 5.1 14.0 

28.10.04 2.5 9.5 3.0 9.9 4.5 11.2 

Table 15. Observed and predicted neck blast incidence in treated plots of different sowings 

Date of 

1 

observation 

st Sowing: 21/6/2004 2nd Sowing: 06/7/2004 3rd Sowing: 20/7/2004 

Observed Predicted Observed Predicted Observed Predicted 

32 

14.10.04 0.1 1.7 0.2 1.8 0.0 0.0 

18.10.04 0.2 4.6 0.5 4.9 0.3 4.7 

21.10.04 0.4 7.4 0.8 7.8 1.1 8.0 

25.10.04 0.8 4.9 1.4 5.4 2.3 6.1 

28.10.04 1.3 2.1 2.0 2.7 3.6 3.9 

1.11.04 1.6 0.8 3.3 2.2 5.3 3.7 

4.11.04 2.2 0.1 4.0 1.3 5.9 2.9

Table 16. Fungicidal spray schedule based on the predicted leaf blast severity 

Sowings 

(Date of 

sowing) 

Predicted 

disease (%) 

1 st Spray 2 nd Spray 3 rd Spray 4 th Spray 

Days after 

initial 

symptoms 

Predicted 

disease (%) 

Days after 

1 st spray 

Predicted 

disease (%) 

Days after 

2 nd spray 

Predicted 

disease (%) 

Days after 

3 rd spray 

1 st Sowing 13.7 10 12.2 13 14.9 18 - - 

(21.6.2004) 

2 nd Sowing 14.9 15 16.2 12 16.1 14 - - 

(06.7.2004) 

3 rd Sowing 16.0 7 17.6 10 15.3 10 15.9 7 

(20.7.2004) 

In the untreated plots, the progress of leaf blast severity (%) had gone up to a 

maximum of 79.5, 90.5 and 95.4 in 1 st , 2 nd and 3 rd sowings, respectively on 11 th 

Oct. Daily weather variables of a week earlier (Table 17) led to this high disease 

intensity at different sowings (Fig. 9). But in the forewarned plots (treated plots), the 

disease severity (%) was comparably very less viz., 4.7, 5.6 and 9.3 in the respective 

sowings (Fig. 9), as the disease was timely predicted and the fungicidal schedule was 

followed accordingly. This clearly shows that forewarning is very much necessary in 

suggesting the timely application of fungicides and also for judicious use of the 

fungicides. 

Fig. 9. Progress of blast at experimental field, DRR in Kharif, 2004 

Treated Plots 

Untreated plots 



33



of observation of blast at the experimental field, DRR, Kharif, 2004 

In general, the combined effects of different weather variables especially T max, 

T min and Rh E influenced the neck blast severity. However, the development of neck 

blast and subsequently the required predicted neck blast incidence did not cross the 

threshold level of 10 %. So, the spray was not suggested for neck blast. 

34 

Date of 

observation 


T Max 



(mm) 


days/ 

Week 

(No.) 


Rh M 

Rh E 

23-Aug 29.1 22.7 0.0 0.0 95.7 69.3 

28-Aug 29.5 22.8 0.2 1.0 96.4 66.0 

31-Aug 30.1 22.6 0.2 2.0 97.7 63.1 

3-Sept 30.4 22.1 0.0 1.0 98.3 60.0 

7-Sept 29.6 22.0 3.9 3.0 98.9 69.4 

10-Sept 29.4 22.4 8.6 5.0 100.0 79.7 

13-Sept 28.5 21.9 12.4 4.0 100.0 85.4 

16-Sept 29.2 21.8 12.2 4.0 100.0 77.9 

20-Sept 30.7 22.4 1.3 5.0 100.0 69.3 

23-Sept 30.2 22.3 4.1 6.0 100.0 78.0 

27-Sept 30.0 22.1 3.1 6.0 100.0 81.9 

30-Sept 31.4 21.9 0.1 4.0 100.0 73.1 

4-Oct 31.5 21.4 1.4 5.0 100.0 70.3 

7-Oct 30.0 21.7 8.7 7.0 100.0 79.3 

11-Oct 30.4 22.1 7.4 4.0 97.4 76.1 

14-Oct 31.4 21.3 0.4 2.0 95.7 63.1 

18-Oct 31.2 19.1 0.3 1.0 93.0 57.0 

21-Oct 30.9 16.8 0.0 0.0 87.3 50.4 

25-Oct 30.9 17.6 0.0 0.0 87.0 48.0 

28-Oct 30.6 19.3 0.0 0.0 90.9 52.6 

1-Nov 29.9 19.9 0.0 0.0 91.0 55.3 

4-Nov 29.4 19.5 0.0 0.0 89.9 51.4

VI. STRATEGIES FOR THE CONTROL OF BLAST 

Although blast can be managed through host resistance, cultural practices or 

fungicidal treatments, the strategy for its management should be through an integrated 

crop management approach, which is the most effective way to manage the disease. 

However, among these the most effective and economical control of blast disease is 

the development and use of resistant varieties, though most of the currently cultivated 

rice varieties and Basmati varieties do not have adequate resistance. 

High nitrogen application at critical growth stages is required for high yields, 

but this high nitrogen levels increases susceptibility of host plants. Potash application 

as per the recommendation is critical as it increases host resistance against blast and 

helps in proper grain filling. However, the most practical way to control blast epidemics 

is to use fungicides judiciously. Through fungicidal application, the life span of a 

variety in terms of durability of resistance is prolonged. There are effective blasticides 

under current usage. But, they are not used at initial stages of disease development 

to curb the epidemic. Any chemical control strategies that do not allow the farmers to 

economically maximise the yields have no value even if the disease is controlled. 

Therefore, the time of application of fungicide need to be integrated with cultural 

practices to manage the disease cost-effectively. 

During vegetative stage, the blast spots initially are seen on lower leaves and 

gradually spreading to the top leaves. When the initial spots of pin-head size are 

seen, further N application has to be suspended and if favourable weather conditions 

for the spread of the disease persist, immediate spray with an effective fungicide has 

to be taken, as the records of blast incidence in the country reveal that weather 

parameters are more important than other epidemiological factors. To prevent the 

flare-up of the disease a regular surveillance of the crop is required because if the 

sensitive stage of the disease escapes unnoticed, the later application of fungicides 

will not be cost effective. The number of sprays and intervals between sprays depend 

on disease prevalence and weather conditions. However, a minimum of two sprays 

at fortnight intervals will keep the disease under control. To prevent neck blast, a 

fungicide spray has to be given during late booting, early heading stage followed by 

one more spray after 10 days to protect late emerging panicles from neck blast. 

Efficient management of blast is mostly dependent on the adequate 

technologies for its management, which are available and recommended at present 

for the control of blast. They are given below: 

1. Resistant Varieties 


Ø Avoid planting susceptible varieties. 

Ø Grow resistant varieties, such as: Rasi, IR 36, IR 64, Sasyasree, Srinivas, 

Tikkana, Simhapuri, Parijatha, Salivahana or Gauthami. 

35


2. Fungicidal Control 

Ø Seed Treatment (With any one of the following fungicides): 

Tricyclazole 75 WP @ 2 g / kg of seed 

Carbendazim 50 WP @ 4 g / kg of seed 

Ø Spray (With any one of the following fungicides): 

Tricyclazole 75 WP @ 0.6 g / liter of water 

Isoprothiolane 40 EC @ 1.5 ml / liter of water 

Kasugamycin 3 L @ 2.0 ml / liter of water 

Ediphenphos 50 EC @ 1 ml / liter of water 

Iprobenphos 48 EC @ 2 ml / liter of water 

Propiconazole 25 EC @ 1 ml / liter of water 

Carbendazim 50 WP @ 1g / liter of water 

Thiophanate-methyl 75 WP @ 1g / liter of water 

3. Cultural Practices 

36 

Ø Select healthy seed 

Ø Apply moderate nitrogen levels (80 – 100 kg/ha) in 3 to 4 splits 

Ø Avoid excess nitrogen; skip final nitrogen in blast infected fields 

Ø Destroy stubbles / weeds, etc. 

Ø Apply recommended level of potash fertilizers (40 kg/ha).

VII. FUTURE WORK 

Historical hypothesis states that the primary infection foci in farmers’ fields 

go unnoticed that would ultimately result in flare up of the disease rendering any 

control measure ineffective and un-economical. So, early detection of the disease 

based on forecasting factors (chiefly weather) will certainly result in desirable economic 

gains. 

Ø Use of results for the future work 


As the indiscriminate use of fungicides led to problems like the pathogen 

developing resistance to chemicals, resurgence of minor diseases, risk to human 

and animal health and environmental pollution, the understanding of blast dynamics 

in relation to weather parameters will help in controlling the misuse of fungicides. 

Organized disease and weather monitoring are important for shortening the 

gap between maximum yields and the average yields. A thorough understanding of 

the critical weather conditions and vulnerable growth stages of crops conducive for 

field occurrence of diseases will facilitate forecasting the outbreaks in advance. The 

information on the incidence of blast in relation to meteorological conditions would 

be useful to develop a method of forecasting outbreaks of the disease so that spraying 

operations could be undertaken at the right time and at lower costs. 

The weather forecasts can be utilized effectively for devising appropriate 

forewarning systems for use in agro-advisory services to gear up the crop protection 

activities at the appropriate time. Such information would be useful in linking to the 

Integrated Disease Management strategy, which will further help in reducing the 

consumption of fungicides and minimize environmental pollution and health hazards. 

Fungicidal spraying can be avoided if weather is not going to be fovourable for 

desease progress even though the disease is present at low level in the field. 

The validated rice blast models would be used for the judicious application of 

fungicides in checking the disease in the farmers’ fields. These results also would be 

useful in further continuing linkages with the farmers and State Agricultural Department 

in disseminating the information on forewarning of rice blast and timely application 

of fungicides. 

Ø Gaps in the research done and suggestions for the future work 

In many plant diseases, epidemics can often develop in spite of ‘inferior’ 

weather conditions. This was attributed to compensation phenomena, in which the 

limitations imposed by a given environmental or biotic factor present in a minimally 

favourable state are compensated by another environmental or biotic factor present 

in a more favourable state (Aust et al, 1980). Such compensation widens the 

geographical and climatic boundaries of the disease and more detailed work in this 

area would be rewarding. 

37


Forecasts are made only when significant correlations are found between 

blast occurrence and some factors, which are surveyed. And so meaningful results 

can be obtained only if a suitably large amount of data is made available from more 

number of locations. Forecasts are intended to provide a basic contribution to control 

plans and control work in a particular area, and so should not be limited to one or 

two special locations, but should constitute an integral part of an overall control 

system. 

38

EXECUTIVE SUMMARY 


Rice blast caused by the fungus, Pyricularia grisea (Magnaporthe grisea) has 

been the most important disease, most feared by the farmers as it occurs usually 

over a wide area with remarkable destructiveness under favourable conditions. The 

intensity of blast infection is greatly influenced by the environment and the variety of 

the rice grown. Extensive studies on the role of temperature and humidity have been 

made, and forecasting of the disease was attempted on the basis of minimum night 

temperature of 20 - 26 o C in association with a high relative humidity range of 90 % 

and above lasting for a period of a week or more during any of the susceptible 

phases of growth, viz., seedling stage, post-transplanting tillering stage, and at neck 

emergence. In plains, with the help of ‘trap’ plots of susceptible varieties successful 

warnings were given at least 10 to 15 days in advance of the outbreak in farmer’s 

fields. Several computer simulation models, viz., BLASTL, PYRICULARIA, BLASTAM, 

LEAFBLAST, EPIBLA, BLASTSIM, and PBLAST have been developed. 

Blast severity in relation to weather was monitored in the experimental fields 

at Directorate of Rice Research, Hyderabad, and at CSK Himachal Pradesh Krishi 

Viswa Vidyalay, Palampur, and farmer’s fields both in Andhra Pradesh and Himachal 

Pradesh, with different dates of sowing on blast susceptible varieties. From the historical 

data on blast and meteorological data in Himachal Pradesh, where the blast is 

endemic, the pattern of progress of blast showed that all the disease progress curves 

were more or less sigmoid. It revealed that perception threshold of leaf blast was less 

than 5 %. The analysis of average daily weather variables of a week earlier to these 

threshold levels of years 1991 to 2000 showed that maximum temperature of 23 to 

28 o C, minimum of 17.6 to 24 o C, Rh of more than 80 % with exception during 1997 

and 1998, rainfall of more than 3 mm per day and more than 4 rainy days per week 

were critical in the progress of rice blast. Pooled regression models for leaf blast at 

Palampur during different years were developed from the historical data for the 

years 1991-2001. Out of all these equations, the regression equation of the duration 

1997-2000 (Y = -20.53+3.16X 1 –7.28X 2 +0.60X 3 +0.54X 4 -0.36X 5 +1.60X 6 ) was 

observed to be the best fit equation. 

Blast severity was high in the late sown crop in Andhra Pradesh, when 

compared to the early sown crop. Analysis of daily weather variables of a week 

earlier to the high disease intensity in 2003 were T Max : 29.3 0 C, T Min : 22.3 0 C. Rh M : 

100 %, Rh E : 80.1 %, Rf: 5.3mm and R W : 6. These conditions led to the high leaf blast 

severity (77.1 to 92.9 %) at different sowings. The relationship was best explained by 

the equation: Y = -1537.742 + 13.589** T Max + 13.413** T Min - 6.197** R W + 

9.431** Rh M + 1.758 L w . Neck phase of the disease was also maximum (26.3 to 

34.7 %) in kharif, 2002, with favourable weather conditions, viz., T Max : 29.5 0 C, T Min : 

18.2 0 C. Rh M : 100 %, Rh E : 68.3 %, Rf: 1.6 mm and Rainy days / Week: 5. All these 

conditions at the experimental field, led to the epidemic progress of leaf blast during 

the last week of September and 1 st fortnight of October and, neck and node blast 

39


during the last week of October and 1 st week of November. The relationship between 

neck blast incidence and the weather parameters was best explained by the equation, 

Y = - 155.37 + 8.493** T Max - 8.061** T Min + 3.82* R W + 0.924 Rh E. The average 

weather conditions in the farmers’ fields in general, led to the epidemic progress of 

leaf blast during the last week of September and 1 st week of October, and neck blast 

during the last week of October. 

In Himachal Pradesh, during 2001 at Palampur location in the I st date of 

transplanting, the combined effects of T Max , Rh Min and rainy days/4 days resulted in 

significant effect on leaf blast severity. In the II nd date of transplanting during 2001, 

T Min, Rh Max , hours with Rh > 85 % and rainfall were good predictors. At Malan, during 

2001, the best predictors were mean Rh, rainfall and rainy days/week for leaf blast 

severity whereas at Pharer (Farmers’ fields), it was Rh Min and rainy days/week. The 

best fit equations for neck blast revealed that the combined effects of different weather 

factors especially T Max and Rh Min influenced the neck blast severity. Validation of model 

(Y = -20.53+3.16X 1 –7.28X 2 +0.60X 3 +0.54X 4 -0.36X 5 +1.60X 6 ) of the data of 2001 

was found more realistic on farmers’ fields data. During 2002 at Palampur location, 

the best fit regression equation was found to be Y = 74.981- 

3.468X 2 +0.221X 1 +0.492X 3 with R = 0.90 and R 2 = 0.81. At Malan location, 

minimum temperature and Rh Max were found more important variables in predicting 

leaf blast. At third location (Farmers’ fields), minimum temperature and rainfall were 

good predictors for leaf blast severity, though the disease pressure was very low due 

to prolonged drought conditions. For leaf blast prediction, model developed for Malan 

location (Y = -127.61+2.642X 8 -0.114X 6 -5.419X 9 ) was found more realistic as 

compared to other locations. Similarly, the model developed for Malan location (Y = 

9.62-0.28X 1 -0.08X 6 ) was found more realistic for neck blast prediction as compared 

to other locations. 

Forewarning models were developed at DRR both for leaf blast (LB N = 14.5 

+ 0.90 LB P + 0.10 T Max – 1.41 T Min – 0.10 Rf + 1.20 R W + 0.29 Rh M –0.18 Rh E ; 

R 2 = 0.92) and neck blast (NB N = -25.61 + 0 .79 NB P + 1.8 T Max –2.5 T Min + 0.38 Rh E ; 

R 2 = 0.90) based on the pooled data of 2001-2003. These forewarning models 

were used for forecasting the disease based on the preceding week weather variables 

and blast severity. The threshold level of predicted disease considered for spraying 

the fungicides was around 10 %. Based on prediction of the disease severity, timely 

spray of fungicides was suggested and the disease was controlled, effectively. 

Correlation coefficients assessed between leaf blast severity and weather 

parameters based on the forewarning model, and analysis of weather variables 

showed that morning and evening relative humidity and, rainy days per week were 

significant weather factors for the leaf blast development, while both minimum and 

maximum temperature, evening relative humidity were found to have significant effect 

on neck blast incidence. 

40


These forewarning systems can be utilized in agro-advisory services to gear 

up the crop protection activities at the appropriate time. Such information would be 

useful in linking to the Integrated Disease Management strategy, which will further 

help in reducing the consumption of fungicides and minimize environmental pollution 

and health hazards. The validated rice blast models would be used for the judicious 

application of fungicides in checking the disease in the farmers’ fields. These results 

also would be useful in further continuing linkages with the farmers and State 

Agricultural Department in disseminating the information on forewarning of rice 

blast and timely application of fungicides. 

41


REFERENCES 

Anonymous. 1990. Study on the simulation and forecast model of epidemic of rice 

leaf blast in Jilin Province II. Establishment and examination of the simulation 

model. J. Jilin Agril.Univ. 17(1): 1-4. 

Aust, H.J., Bashi, E. and Rotem, J. 1980. Flexibility of plant pathogens in exploiting 

ecological and biotic conditions in the development of epidemics. In: J.Patil an 

J.Kranj (eds.) Comparative epidemiology, Pudoc. Centre for Agricultural Publishing 

and Documentation, Wageningen, p.46-56. 

Basitiaans, L. 1991. Quantifying the effect of rice leaf blast on leaf photosynthesis, 

crop growth, and final yield using an ecophysiological model of crop growth. 

“Rice blast modelling and forecasting”: Selected papers from the International 

Rice Research Conferences, 27-31 August 1990, Seoul, Korea Republic, 89-99. 

Bhatt, J.C. 1992. Epidemiology of foliar blast of rice in Uttar Pradesh hills. Oryza 29: 

350-354. 

Calvero, S.B. Jr., Coakley, S.M. and Teng, P.S. 1996. Development of empirical 

forecasting models for rice blast based on weather factors. Plant Pathology 45: 

667-678. 

Calvero, S.B. Jr., Coakley, S.M., Surin, C., Kim, C.K. and Teng, P.S. 1997. New 

procedure for developing weather-based empirical disease forecasting models. 

Dizon, T.O.; Eusebio, J.E.; Mabbayad, M.O.; Duenas, J.N.; Palis, F.V. (eds.). Pest 

management council of the Philippines, Inc., PMCP 27 th Anniversary and Annual 

Scientific Meeting: Proceedings, College, Laguna (Philippines). 

Chakrabarti, N.K. 1971. Relationship between meteorological factors on the 

development of foliar blast. Oryza 8 (2): Suppl. 592-597. 

Chaudhary, R.G. and Vishwadhar. 1988. Epidemiology of rice blast and effect of 

date of sown under upland conditions of Arunachal Pradesh. Indian Phytopath. 

41(4): 552-557. 

Choi, W.J., Park, E.W. and Lee, E.J. 1988. LEAFBLAST: A computer simulation model 

for leaf blast development on rice. Korean J. Plant Pathol. 4(1): 25-32. 

Dubey, S.C. 2003. Influence of weather factors on the development of blast 

(Magnaporthe grisea) of rice in Jharkhand. Pl. Dis. Res. 18 (2): 101-105. 

El Refaei, M.I. 1977. Epidemiology of rice blast disease in tropics with special reference 

to the leaf wetness in relation to disease development. Ph.D. Thesis, IARI, New 

Delhi. 

Govindaswamy, V.V. 1964. Temperature and blast incidence. Madras Agric. J. 51:255 

- 257. 

42


Gunther, H. 1986. Simulation of the epidemiology of Pyricularia oryzae in rice: a 

first attempt. Agric. Univ. Wageningen, Dept. Theor. Prod. Ecol. pp.80. 

Hashimoto, A., Hirano, A.K. and Matsumoto, K. 1984. Studies on the forecasting of 

rice blast development by application of the computer simulation. Spl. Bull. 

Fukushima Pref. Agri. Exp. Stn. 2:1-104. 

Hashimoto, A., Matsumoto, K. and Adachi, T. 1982. Some trials for forecasting rice 

leaf blast epidemics by application of the computer simulation. Ann. Report of the 

Soc. of Plant Protection of North Japan 33: 12-14. 

Hayashi, T. and Koshimizu, Y. 1988. Computer program BLASTAM for forecasting 

occurrence of rice leaf blast [in Japanese, English summary]. Bull. Tohoku Natl. 

Agric. Exp. Stn. 78: 123. 

Hori, M. 1963. Study on the forecast of the occurrence of rice blast with special 

reference to the forecasting method on the basis of experiments. Spec. Bull. Dis. 

Ins. Forecast M.A.F. pp. 69. 

Huang, Y. T. 1980. Some epidemiological studies on rice blast disease in Taiwan. 

Plant Protection Bulletin, Taiwan. 22(2): 153-181. 

Huang, C.Y., Zhu, C.Y., Zhang, Z.M., Shang, S.J., Zhang, Y.H., Guo, M. 1999. Study 

on the decision model for integrated control of rice blast in Heilongjiang province. 

Research progress in plan protection and plan -nutrition 399 - 403. 

IRRI, 1996. Standard evaluation system for rice. IRRI, Manila, Philippines. 

Ishiguro, K. 1986. Composing a simulation model for epidemics of rice panicle 

blast. Nogyo Gijutu 41:491-495. 

Isiguro, K. and Hashimoto, A. 1988. Development of a simulation model for rice 

panicle blast epidemics. (1) Composition of the basic model [in Japanese, English 

summary]. Bull. Fukushima Pref. Agric. Exp. Stn. 27: 1-19. 

Isiguro, K. and Hashimoto, A. 1990. Development of a simulation model for rice 

panicle blast epidemics. (2) Its verification and sensitive analysis [in Japanese, 

English summary]. Bull. Fukushima Pref. Agric. Exp. Stn. 29: 15-70. 

Isiguro, K. and Hashimoto, A. 1991. Computer based forecasting of rice blast 

epidemics in Japan. Rice blast modeling and forecasting. Selected papers from 

the International Rice Research Conference, 27-31 August 1990, Seoul, Korea 

Republic, pp. 39-51. 

Kahn, P. and Libby, J.L. 1958. The effect of environmental factors and plant age on 

the infection of rice by the blast fungus, Pyricularia oryzae. Phytopathology 48: 

25-30. 

43


Kapoor, A.S., Sood, G.K. and Kaushik, R.P. 2000. Integrated management of rice 

blast. Proceedings of Indian Phytopath., Soc.Golden Jubilee International Conf. 

on “Integrated Plant Disease Management for Sustainable Agriculture”. 1: 218- 

219 (1997). 

Kato, H. and Kozaka, T. 1974. Effect of temperature on lesion enlargement and 

sporulation of Pyricularia oryzae in rice leaves. Phytopathology 64(6): 828-830. 

Kato, H. 1976. Some topics in a disease cycle of rice blast and climatic factors. In 

Proc.Symp.Climate and rice. IRRI, Box 933, Manila, Philippines, pp.417-421. 

Kaur, S., Kaur, P. and Padmanabhan, S.Y. 1977. Influence of temperature on the 

development of blast in rice. Indian Phytopath.30: 544-546. 

Kim, C.H., Mackenzie, D.R. and Rush, M.C. 1988. Field testing a computerized 

forecasting system for rice blast disease. Ecology and Epidemiology 78(7): 931- 

933. 

Kim, C.K. 1982. Improved methods for rice blast forecasting. Korean J. Plant Protection 

(Korea R.) 21(1): 19-22. 

Kim, C.H., Mackenzie, D.R. and Rush, M.C. 1987. A model to forecast rice blast 

disease based on weather indexing. Korean J. Plan-Pathology. 3(3): 210-216. 

Kim, C.K. and Kim, C.H. 1993. The rice leaf blast simulation model EPIBLAST. 

International Rice Research Conferences, 27-31 August, 1990, Seoul Korea 

Republic, pp. 53-67. 

Kim, C.M., Rush, M.C. and Mackenzie, D.R. 1985. Effect of water management on 

the epidemic development of rice blast. Phytopathology 75: 501. 

Kiyosawa, S. 1972. Mathematical studies on the curve of disease increase-a technique 

for developing epidmic development. Ann. Phytopathol. Soc. Japan 38:30-40. 

Kobayashi, J. 1984. Studies on epidemic of rice leaf blast, Pyricularia oryzae, in its 

early stage. Bull. Akita Agric. Exp. Stn. 26: 1-84. 

Koshimizu, Y. 1983. A forecasting method for leaf blast outbreak by the use of AmeDAS 

data [in Japanese, English summary]. Plant Prot. 37: 454-457. 

Koshimizu, Y. 1988. A forecasting method for occurrence of rice leaf blast with AmeDAS 

data [in Japanese, English summary]. Bull. Tohoku Natl. Agric. Exp. Stn. 78: 67- 

121. 

Krishnan, P., Muralidhar, T.T., Manibhushanrao, K. and Reddy, B.C. 1992. 

Epidemiological studies of rice blast disease 1. Spore releasing capacity of various 

types of blast lesions and spore release pattern against weather parameters. 

International J. Tropical Plant Diseases. 10: 153-165. 

44

Lee, J.T., Yun, S.H., Kim, C.K., Im, J.N. and Jung, Y.S. 1989. Forecasting model of 

rice leaf blast (Pyricularia oryzae) by meteorological data. Res.Reports of the Rural 

Development Administration Crop Protection. 31(2): 9-16. 

Luo, Y. and Teng, P.S. 1994. Risk analysis of rice leaf blast epidemics on Asian countries 

associated with global climate change by simulation studies. Zeigler, R.S.; Leong, 

S.A.; Teng, P.S. (eds.). IRRI, Los Banos, Laguna (Philippines). Entomology and Plant 

Pathology Divn., CAB International, Wallingford, Oxon OX10 8DE (U.K.). “Rice 

blast disease”, Los Banos, Laguna (Philippines), IRRI, pp. 602-603. 

Luo, Y., Teng, P.S., Fabellar, N.G. and TeBeest, D.O. 1997. A rice leaf blast combined 

model for simulation model for simulation of epidemics and yield loss. Agricultural- 

Systems (UK) 53(1): 27-39. 

Manibhushanrao, K. and Krishnan, P. 1991. Epidemiology of blast (EPIBLA): a 

simulation model and forecasting system for tropical rice in India, In: Rice blast 

modelling and forecasting. IRRI, Box 933, Manila, Philippines, pp. 31-38. 

Manibhushanrao, K. and Day, P.R. 1972. Low night temperature and blast disease 

development on rice. Phytopathology 62: 1005-1007. 

Mishra, B. 2002. Rice in India. Agrolook, 3: 10 – 15. 


Muralidharan, K. and Venkatarao, G.V. 1980. A simple method of for forecasting 

outbreak of rice blast. Indian Phytopath. 33: 560-564. 

Ono, K. 1965. Principles, methods, and organisation of blast disease forecasting. In 

: The rice blast disease. Johns Hopkins press, Baltimore, Maryland, pp. 173-194. 

Ota, K. 1982. The forecasting of rice blast disease. Kongetu no nooyaku. 26: 12-19. 

Padmanabhan, S.Y. 1953. Effect of nitrogenous fertilization upon the susceptibility of 

blast on rice varities. Curr. Sci. 22: 271- 272. 

Padmanabhan, S.Y. 1965. Studies on forecasting outbreaks of blast disease of rice. 

I. Influence of meteorological factors on blast incidence at Cuttack. Proc. Indian 

Acad. Sci., 62: 117-129. 

Padmanabhan, S.Y., Chakrabarti, N.K and Rao, K.V.S.R.K. 1971. Forecasing and 

control of rice disease. Proc.Indian Acad.Sci. 37: 423-429. 

Pramanick, M., Chakraborti, N.K. and Pramanick, M. 1990. Severe incidence of rice 

blast in West Bengal. Indian J. Mycol. and Plant Pathol. 20: 256-261. 

Prasad Rao, G.S.L.H.V., Premanathan, T. and Mahana, N.K. 1999. Forecasting of 

rice blast disease epidemic in northern region of Kerala. Oryza 36(4): 364-368. 

Sadasivan, T.S., Suryanarayanan, S. and Ramakrishna, L. 1965. Influence of 

temperature on rice blast disease. In: The rice blast disease, Johns Hopkins Press, 

Baltimore, Maryland, pp.163-171. 

45


Sasaki, T. and Kato, H. 1972. A statistical method of predicting outbreaks of rice 

panicle blast. Phytopath. 62: 1126-1132. 

Sharma, B.R. and Kapoor, A.S. 2003. Some epidemiological aspects of rice blast. 

Plant Disease Research 18(2): 106-109. 

Sharma, J.P., Kumar, S. and Verma, R.N. 1993. Relationship between rice blast and 

meteorological factors in Nagaland. Indian Phytopath. 46(1): 78-80. 

Subramanian, S. 1967. Rice blast, nyctotemperature and metabolic changes. 

Phytopath. Z. 60: 41-50. 

Surin, A., Arunyanart, P., Rojanahusdin, W., Munkong, S., Dhitikiattipong, R. and 

Disthaporn, S. 1991. Using empirical blast models to establish disease 

management recommendations in Thailand. Rice blast modelling and forecasting, 

IRRI, Los Banos, (Philippines), pp. 39-51. 

Suzuki, H. 1975. Meteorological faactors in the epidemiology of rice blast. Ann. Rev. 

Phytioath. 13: 239-256. 

Takahashi, Y. 1958. A method for forecasting blast disease using the leaf sheath 

inoculation technique. Plant Protection 12: 339-345. 

Takai, A., Kanai, K., Kobayashi, M., Murata, K. and Komori, R. 1985. Studies on the 

forecasting of rice blast by system dynamic model. Forecast for number of lesions 

and infection area of rice leaf blast by application of computer simulation. Bull. 

Ibaraki Agric. Exp. Stn. 25: 97-131. 

Tastra, I.K., Irmansyah, R., Yunizar, Z.H. 1987. Epidemiology of leaf blast (Pyricularia 

oryzae) under various abiotic and biotic conditions. Preliminary report of the SARIF 

team case study. 

Teng, P.S., Klein Gebbinck, H.W. and Pinnschmidt, H. 1991. Analysis of the blast 

Pathosystem to guide modeling and forecasting. IRRI, Los Banos, Laguna 

(Philippines). Rice blast modeling and forecasting. Los Banos, Laguna (Philippines). 

IRRI. 1-30. 

Tsai, W.H. and Su, H.J. 1984. Prediction of rice leaf blast. I. Meteorological variables 

and the development of the number of lesions. Plant Protection Bulletin, Taiwan. 

26(3): 171-180. 

Ventakarao, G. and Muralidharan, K. 1982. Effect of meteorological conditions on 

the incidence and progress of blast disease on rice. Z. Pflkrankh. Pflschutz. 89: 

219-223. 

Yamaguchi, T. 1970. Forecasting techinques of rice blast. J.Agric.Res.Q. 5: 26-30. 

46 

ggggg

Forewarning Rice Blast in India

You also want an ePaper? Increase the reach of your titles

Delete template?

Save as template?