epidemiology and development of forecasting models for white rust of brassica juncea ...

This article was downloaded by: [Northeastern University]On: 29 October 2014, At: 20:29Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registeredoffice: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Archives Of Phytopathology And PlantProtectionPublication details, including instructions for authors andsubscription information:http://www.tandfonline.com/loi/gapp20

Epidemiology and development offorecasting models for White rust ofBrassica juncea in IndiaDr. Chirantan Chattopadhyay a , Ranjana Agrawal b , AmrenderKumar b , R. L. Meena a , Karuna Faujdar a , N. V.K. Chakravarthy c

, Ashok Kumar d , Poonam Goyal a , Dr P. D. Meena a & Dr ChanderShekhar aa National Research Centre on Rapeseed–Mustard (ICAR) , Sewar,Bharatpur, 321303, Indiab Division of Forecasting , Indian Agricultural Statistics ResearchInstitute (ICAR) , New Delhi, 110012, Indiac Division of Agricultural Physics , Indian Agricultural ResearchInstitute (ICAR) , New Delhi, 110012, Indiad Shiwalik Agricultural Research and Extension Centre, CSKHPKV ,Kangra, 176001, IndiaPublished online: 17 May 2011.

To cite this article: Dr. Chirantan Chattopadhyay , Ranjana Agrawal , Amrender Kumar , R. L.Meena , Karuna Faujdar , N. V.K. Chakravarthy , Ashok Kumar , Poonam Goyal , Dr P. D. Meena &Dr Chander Shekhar (2011) Epidemiology and development of forecasting models for White rustof Brassica juncea in India, Archives Of Phytopathology And Plant Protection, 44:8, 751-763, DOI:10.1080/03235400903458571

To link to this article: http://dx.doi.org/10.1080/03235400903458571

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the“Content”) contained in the publications on our platform. However, Taylor & Francis,our agents, and our licensors make no representations or warranties whatsoever as tothe accuracy, completeness, or suitability for any purpose of the Content. Any opinionsand views expressed in this publication are the opinions and views of the authors,and are not the views of or endorsed by Taylor & Francis. The accuracy of the Contentshould not be relied upon and should be independently verified with primary sourcesof information. Taylor and Francis shall not be liable for any losses, actions, claims,

http://www.tandfonline.com/loi/gapp20

http://www.tandfonline.com/action/showCitFormats?doi=10.1080/03235400903458571

http://dx.doi.org/10.1080/03235400903458571

proceedings, demands, costs, expenses, damages, and other liabilities whatsoever orhowsoever caused arising directly or indirectly in connection with, in relation to or arisingout of the use of the Content.

This article may be used for research, teaching, and private study purposes. Anysubstantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

Dow

nloa

ded

by [

Nor

thea

ster

n U

nive

rsity

] at

20:

29 2

9 O

ctob

er 2

014

http://www.tandfonline.com/page/terms-and-conditions

http://www.tandfonline.com/page/terms-and-conditions

Epidemiology and development of forecasting models for White rust of

Brassica juncea in India

C. Chattopadhyaya*, Ranjana Agrawalb, Amrender Kumarb, R.L. Meenaa,Karuna Faujdara, N.V.K. Chakravarthyc, Ashok Kumard, Poonam Goyala,P.D. Meenaa and Chander Shekhara

aNational Research Centre on Rapeseed–Mustard (ICAR), Sewar, Bharatpur 321303, India;bDivision of Forecasting, Indian Agricultural Statistics Research Institute (ICAR), New Delhi110012, India; cDivision of Agricultural Physics, Indian Agricultural Research Institute (ICAR),New Delhi 110012, India; dShiwalik Agricultural Research and Extension Centre, CSKHPKV,Kangra 176001, India

(Received 25 August 2009; final version received 1 October 2009)

Experiments were laid out at Bharatpur, New Delhi and Kangra with Indianmustard (Brassica juncea) cvs ‘Varuna’ and an important one in the locality sownon 10 dates at weekly intervals. First appearance of white rust disease (Albugocandida) on leaves and pods (staghead formation) of mustard occurred between36 and 131 days after sowing (d.a.s.), 60 and 123 d.a.s., respectively. Severity ofwhite rust disease on leaves was favoured by 440% afternoon (minimum)relative humidity (RH), 497% morning (maximum) RH and 16–248C maximumdaily temperature. Staghead formation was significantly and positively influencedby 20–298C maximum daily temperature and further aided by 4128C minimumdaily temperature and 497% morning (maximum) RH. Regional and cultivarspecific models devised could predict, at a few weeks after sowing, the crop age atwhich white rust first appeared on the leaves, as staghead, the highest rust severityon leaves, staghead numbers and the crop age at peak rust severity on leaf, higheststaghead numbers at least 1 week ahead of first appearance of the disease on thecrop.

Keywords: Brassica juncea; Albugo candida; epidemiology; forecasting; weather

Introduction

Indian mustard [Brassica juncea (L.) Czern. and Coss.] is one of the major oilseedcrops cultivated in India and around the world. Out of 50,577,000 tonnes ofrapeseed–mustard seed produced over 30,805,000 ha in the world, India produces7,438,000 tonnes from 6,790,000 ha (FAO 2009). White rust caused by Albugocandida (Pers. Ex Fr.) Kuntz., an oomycete can result in yield loss up to 47% (Kolte1985) with each percent of disease severity and staghead formation (malformation ofthe inflorescence due to systemic infection by A. candida) causing reduction in seedyield of about 82 kg ha71 and 22 kg ha71, respectively (Meena et al. 2002). Whiterust is considered an important constraint in husbandry of Indian mustard inIndia; its severity differs between seasons, regions and individual crops within a

*Corresponding author. Email: [email protected]

Archives of Phytopathology and Plant Protection

Vol. 44, No. 8, May 2011, 751–763

ISSN 0323-5408 print/ISSN 1477-2906 online

� 2011 Taylor & Francis

DOI: 10.1080/03235400903458571

http://www.informaworld.com

Dow

nloa

ded

by [

Nor

thea

ster

n U

nive

rsity

] at

20:

29 2

9 O

ctob

er 2

014

region. In the absence of stable, desirable and diverse source of resistance to themustard white rust, fungicides remain the only effective means to manage the disease(Mehta et al. 1996). Although consumption of fungicides on the mustard crop inIndia is high (IASRI 2006), timing their application has been erratic. Crops requiringtreatment have been left unsprayed and others sprayed unnecessarily. Knowledge onthe epidemiology of the disease on mustard is limited. Survival of the pathogen ondiseased seed or affected plant debris in tropical or sub-tropical India has been ruledout (Verma and Bhowmick 1988), unlike the situation in temperate conditions(Humpherson-Jones 1992). In India, Indian mustard is sown from September toNovember, depending on the prevailing temperature and availability of soil moisturefor seed germination. Harvest occurs from February to May. Off-season crops aregrown in non-traditional areas from May to September and this, coupled withharbouring of the fungal pathogen by oilseed, vegetable Brassica crops andalternative hosts could cause carryover of A. candida from one crop-season toanother (Kolte 1985). Thus, air-borne spores of A. candida form the primary sourceof inoculum of this polycyclic disease (Kolte 1985). Efficient, economical andenvironment friendly control of the rust may be obtained through knowledge of itstiming of attack in relation to weather factors, which may enable prediction of itsoccurrence so as to allow growers to take timely fungicidal sprays for an efficientcrop management. Weather is an important factor in the severity of white rust ofIndian mustard, which governs the variability in epidemic onset of the disease.Empirical models have been developed to relate occurrence of white rust on Indianmustard to temperatures, relative humidity (RH) and sunshine hours (Bains andJhooty 1979; Kolte et al. 1986; Saharan et al. 1988; Lakra and Saharan 1988a; Hegdeand Anahosur 1994; Mehta and Saharan 1998; Sangeetha and Siddaramaiah 2007).However, the available information provides no insight into quantitative predictionof the disease on the Indian mustard in different parts of India. Accurate forecast ofthe crop age at first appearance of the rust and the risk of a rust epidemic wouldenable farmers to decide on optimum timing of fungicide sprays and to avoidunnecessary pesticide application. Hence, the present study was undertaken todevelop forecasts for crop age at time of first appearance of white rust, highestseverity of the disease on the crop in the season and crop age at peak severity of therust.

Materials and methods

Epidemiological study

The selection of centres for the study was based on the area of crop and importanceof white rust as a disease problem in the region. Field trials were sown on 10 dates atweekly intervals (1, 8, 15, 22, 29 October, 5, 12, 19, 26 November and 3 December) atBharatpur (278120N; 778270E), New Delhi (288390N; 778130E) and Kangra (32840N;768160E) in nine (1999–2000 to 2007–2008) post-rainy (rabi) crop seasons with cvs‘Varuna’ and an important one in the locality (cv. ‘Rohini’ at Bharatpur; cv. ‘BIO-902’ at New Delhi; cv. ‘RCC-4’ at Kangra). Each plot measured 1.5 m 6 5 m withspacing of 30 cm 6 10 cm. The recommended doses of N and P fertilisers(NRCRM 1999) were applied. No K fertiliser was applied. Insect–pest protectionpractices were undertaken as seed treatment with Imidachloprid @ 7 g kg71 andspray of oxydemeton methyl @ 0.025% a.i. at 15-day intervals. No protection wastaken against any disease.

752 C. Chattopadhyay et al.

Dow

nloa

ded

by [

Nor

thea

ster

n U

nive

rsity

] at

20:

29 2

9 O

ctob

er 2

014

Weather data recording

Weather data for maximum and minimum daily temperatures, morning [07:00 hLocal Apparent Time or LAT calculated on the basis of longitude of a locationas per standard norms of the World Meteorological Organisation or WMO(Doorenbos 1976; Ghadekar 2002)] and afternoon (14:00 h LAT) RH, sunshinehours and wind speed were recorded from standard meteorological observatoriesat New Delhi and Kangra. Meteorological observatories at these two locations were110–170 m from the experimental site, and the data recording instruments wereinstalled as per standard specifications of the WMO (Doorenbos 1976; Ghadekar2002). Weather data were recorded at Bharatpur using automatic weather station,located 35 m from the site of experimentation, where apart from recording theweather variables mentioned earlier, data for leaf wetness were also recorded. Thesensors were installed as per standard specifications (Doorenbos 1976; Ghadekar2002).

Disease assessment

All experiments relied entirely on natural incidence of the disease. Data for initialdate of appearance of white rust, gradual progress of the same on leaves andstaghead formation on Indian mustard were monitored in each location.Observations for percent disease severity (PDS) were recorded twice a week (onTuesday and Friday) until harvest from 10 randomly tagged plants in each plot ofthe crop on leaves and pods following the scale of Conn et al. (1990). For eachassessment date, PDS on leaves and staghead (malformed sterile inflorescence)formation of 10-tagged plants from each plot were averaged to give respective singlevalues.

Forecasting models

Different ranges of weather variables of 1 week preceding the assessment date wereused as independent variables to identify the boundary and favourable conditionsthat influenced the dependent variables or disease severity on leaves and stagheadformation, through regression analysis. Correlations of timing (days after sowing ord.a.s.) of first appearance of the rust on leaves, staghead formation, highest severityof white rust on the crop leaves, staghead numbers and crop age (d.a.s.) at highestseverity of leaf rust, highest number of stagheads with weather variables werestudied.

Linear prediction models based on the weather parameters as independentvariables and crop age (d.a.s.) at time of first appearance of the disease on leaves,staghead on the crop, highest severity of the disease on leaves, highest number ofstagheads in the season and crop age at peak severity of the disease at each weekstarting from week of sowing as dependent variables were fitted by multiple stepwiseregression (Draper and Smith 1981) using data of the initial 8 years separately. Basedon correlation coefficients between dependent variables under study with therespective weather parameter (i) in different weeks, a composite weather variable (zi)was developed as the weighted sum of the weather variable in different weeks startingfrom the pre-sowing week up to the week of prediction (Agrawal et al. 1986; Desaiet al. 2004; Chattopadhyay et al. 2005a, b). Similarly, interaction terms (zij) were

Archives of Phytopathology and Plant Protection 753

Dow

nloa

ded

by [

Nor

thea

ster

n U

nive

rsity

] at

20:

29 2

9 O

ctob

er 2

014

developed as weighted sums of product between two weather variables, weightingsbeing correlation coefficients of the dependent variable under study with products ofweather variables in respective weeks. The important weather indices were selectedthrough stepwise regression.

Models were fitted for prediction of the dependent variables viz., highest diseaseseverity or crop age (d.a.s.) at highest disease severity or crop age (d.a.s.) at firstappearance of the disease on weekly basis starting from the time of sowing, secondweek after sowing and so on (f ¼ 1, 2, . . ..). The dependent variables were relatedwith weather parameters in different weeks. The interactions of weather parameterswere also found to be significant. The models were developed in the followingformat:

Y ¼ a0 þXp

i¼1aizi þ

Xp

i6¼jbijzij þ e ð1aÞ

¼ a0 þ amaxtmpzmaxtmp þ amintmpzmintmp þ � � �þ asshzssh þ bmaxtmp�mintmpzmaxtmp�mintmp þ � � �þ baftRH�sshzaftRH�ssh þ e

ð1bÞ

where

zi ¼Xf

w¼1riwxiw ð2aÞ

zij ¼Xf

w¼1rijwxiwxjw ð2bÞ

with Y being the dependent variable, xiw the value of i-th weather parameter in w-thweek, riw the value of correlation coefficient between Y and i-th weather parameter inw-th week, rijw the correlation coefficient between Y and product of xi and xj in w-thweek, p the number of weather variables, f the week after sowing when predicted ande the error term. Weather indices based on summation of weightings of differentmeteorological factors as per correlation coefficients in different weeks after sowinguntil the forecast was provided, were taken into account. The ninth (2007–2008) cropseason was used to validate the models for forecasting the targeted parameters atdifferent locations based on the models developed in each of the initial 8 years foreach of the parameter, viz., crop age at first appearance of the disease (Y1), crop ageat peak disease severity (Y2) and highest disease severity (Y3); paired t test was usedto assess the difference among predicted and observed values.

Results


The severity of white rust on leaves and the number of stagheads (diseased pods) werehigher in later sown crops (Table 1). First appearance of white rust disease on leavesof mustard occurred between 36 and 131 d.a.s., being highest at 54, 71, 50, 53, 57, 55,58, 53 and 54 d.a.s. in 1999–2000, 2000–2001, 2001–2002, 2002–2003, 2003–2004,2004–2005, 2005–2006, 2006–2007 and 2007–2008, respectively. First appearance ofthe disease on pod (staghead formation) occurred between 60 and 123 d.a.s.


Dow

nloa

ded

by [

Nor

thea

ster

n U

nive

rsity

] at

20:

29 2

9 O

ctob

er 2

014

Forecasting models

Correlation study of the data from the relevant centres revealed that white rustseverity on leaves was positively correlated to 440% afternoon (minimum) RH (R2:0.92), 497% morning (maximum) RH (R2: 0.89), 472% daily mean RH (R2: 0.8),4108C daily mean temperature (R2: 0.79) and 16–248C maximum daily temperature(R2: 0.83). Staghead formation was significantly and positively correlated to 20–298Cmaximum daily temperature (R2: 0.81) and further aided by 4128C minimum dailytemperature (R2: 0.84), 497% morning (maximum) RH (R2: 0.89) and 472% dailymean RH (R2: 0.85). Empirically, a look at weather data available from theautomatic weather station indicated 410 h of leaf wetness during the preceding 3days and also favoured the progress of rust severity on leaves and formation ofstagheads.

Crop age (d.a.s.) at first appearance of the rust on the crop (Y1), crop age (d.a.s.)at peak severity of the rust (Y2) and highest disease severity (Y3) were related withweather variables in different weeks including pre-sowing week and the interactionswere found significant. Regional and cultivar specific models devised using data ofthe initial 8 years thereby predicted the crop age at which white rust first appears onthe leaves (Table 2), crop age at highest rust severity on leaves (Table 3) and the peakrust severity on leaves, number of stagheads (Table 4). Weather indices based onsummation of weightings of different meteorological factors as per correlationcoefficients in different weeks after sowing until the forecast was provided, weretaken into account. Most of the models saw entry of variable maximum temperaturewith minimum temperature, morning RH, afternoon RH and sunshine hours alsogetting entered in some cases. Proper monitoring of disease progress duringrecording of observations in experiments could enable devise models for providingaccurate forecasts a few weeks after sowing, about crop age at first appearance, cropage at highest severity and highest level of disease severity. The predictions werepossible at least 1 week ahead of first appearance of the disease on the crop, thusallowing growers to undertake timely sprays. The disease was never found to appearbefore 36 d.a.s. or beginning of the sixth week after sowing, while the prediction for

Table 1. Effect of date of sowing on highest white rust severity on leaves of Indian mustardin 2002–2003 at different locations.

Per cent white rust severity on leaves of cultivars of Indian mustard atdifferent locations

Date of sowing

Bharatpur New Delhi Kangra

‘Varuna’ ‘Rohini’ ‘Varuna’ ‘BIO-902’ ‘Varuna’ ‘RCC-4’

1 October 0.0 0.0 12.2 4.3 24.2 23.28 October 0.0 0.0 14.2 5.8 29.6 24.215 October 0.3 0.0 14.9 6.3 30.4 28.822 October 7.6 13.2 17.9 12.4 33.6 35.229 October 11.3 16.3 18.4 12.6 34.4 36.45 November 13.8 17.2 19.0 13.4 37.6 38.412 November 15.3 21.3 21.0 13.7 38.4 39.219 November 18.2 28.5 24.6 14.6 42.2 40.026 November 32.7 33.0 29.9 14.9 44.0 45.63 December 34.7 42.2 31.4 15.9 50.4 47.2


Dow

nloa

ded

by [

Nor

thea

ster

n U

nive

rsity

] at

20:

29 2

9 O

ctob

er 2

014

Table 2. Models to forecast crop age (Y1) at first appearance of white rust on leaves.

Location Cultivar

Crop age(week) ofprediction Model R2

Bharatpur ‘Varuna’ 4 Y1 ¼ 770.35 þ 0.03 zmaxtmp 6 aftRH 0.92Bharatpur ‘Rohini’ 4 Y1 ¼ 7363.26 þ 0.18 zmaxtmp 6 aftRH þ

0.038 zmaxtmp 6 mornRH

0.95

New Delhi ‘Varuna’ 4 Y1 ¼ 718.86788 þ 0.00335897 6zmaxtmp 6 mornRH

0.99

New Delhi ‘BIO-902’ 4 Y1 ¼ 717.30825 þ 0.003228 6zmaxtmp 6 mornRH

0.96

Kangra ‘Varuna’ 2 Y1 ¼ 47.89 þ 0.081 zmintmp 6 aftRH 70.16 zmaxtmp 6 mintmp

0.82

Kangra ‘RCC-4’ 3 Y1 ¼ 789.50 þ 0.97 zmornRH 0.67

Maxtmp, maximum daily temperature; mintmp, minimum daily temperature; morn, morning; aft,afternoon; RH, relative humidity.

Table 3. Models to forecast crop age (Y2) at highest white rust severity on leaves.

Location Cultivar


Bharatpur ‘Varuna’ 3 Y2 ¼ 798.50 þ 0.0056 zmornRH 6 aftRH þ1.66 zmaxtmp

0.98

Bharatpur ‘Rohini’ 2 Y2 ¼ 757.36 þ 0.01 zmaxtmp 6 mornRH 70.30 zmaxtmp 6 ssh

0.98

New Delhi ‘Varuna’ 4 Y2 ¼ 1.72577 þ 0.11375 zmaxtmp 0.98New Delhi ‘BIO-902’ 5 Y2 ¼ 0.7446 þ 0.10568 zmaxtmp 0.94Kangra ‘Varuna’ 5 Y2 ¼ 467.85 þ 0.02 zmaxtmp 6 mornRH 7

0.02 zmintmp 6 aftRH þ 1.86 zmornRH

0.94

Kangra ‘RCC-4’ 5 Y2 ¼ 657.36 þ 2.48 zmornRH 7 1.54 zmintmp 0.85

Maxtmp, maximum daily temperature; mintmp, minimum daily temperature; morn, morning; aft,afternoon; RH, relative humidity; ssh, sunshine hours.

Table 4. Models to forecast highest white leaf rust severity and staghead numbers (Y3).

Location Cultivar


Bharatpur ‘Varuna’/leaf 4 Y3 ¼ 133.13 þ 19.87 zmaxtmp 0.82Bharatpur ‘Rohini’/leaf 4 Y3 ¼ 96.4 þ 12.03 zmaxtmp 0.87Bharatpur ‘Varuna’/staghead 3 Y3 ¼ 4.17 þ 0.69 zmaxtmp 0.82Bharatpur ‘Rohini’/staghead 3 Y3 ¼ 4.17 þ 0.58 zmaxtmp 0.87New Delhi ‘Varuna’/leaf 4 Y3 ¼ 796.929 þ 0.6222 zmornRH þ

0.0203 zmornRH 6 ssh

0.99

New Delhi ‘BIO-902’/leaf 8 Y3 ¼ 38.6277 þ 0.02713 6zmornRH 6 ssh 7 0.01637 zaftRH þ0.02282 zaftRH 6 ssh

0.99

Kangra ‘Varuna’/leaf 5 Y3 ¼ 77.18 þ 0.27 zaftRH 0.91Kangra ‘RCC-4’/leaf 5 Y3 ¼ 7117.73 þ 0.62 zmornRH 0.80

maxtmp, maximum daily temperature; morn, morning; aft, afternoon; RH, relative humidity; ssh,sunshine hours.


Dow

nloa

ded

by [

Nor

thea

ster

n U

nive

rsity

] at

20:

29 2

9 O

ctob

er 2

014

crop age at first appearance of rust on leaves was possible for most of the locations inthe beginning of the fifth week (29 d.a.s.). Only at New Delhi, the prediction forhighest severity of the disease on cv. ‘BIO-902’ was delayed till beginning of theeighth week after sowing, where the disease on the plant parts reached its peak muchlater than that. Common models for the targeted three parameters with the cultivarof mustard available at all the locations, i.e., ‘Varuna’ were attempted. But they hadlow R2 values and hence were not considered for validation. The models werevalidated in the ninth (2007–2008) year, wherein the observed and predicted valuesmatched closely with low residual values. Out of the models developed in each of the4 initial years for forecasting each of the targeted parameters, only those modelshave been presented in the Tables 2–4, wherein the observed and predicted valuesmatched closely with low or even no residual values (Tables 5 and 6) or differencebetween predicted and observed values in different cases were not significant(P 5 0.05). Models devised for forecast of crop ages at first staghead appearanceand at peak census of staghead for all locations as also highest staghead numbers atNew Delhi, Kangra had low R2 values and hence were not considered for validation.Models to forecast peak staghead number on the crop could also be devised forBharatpur, which enabled early assessment of the risk involved on the crop at thelocation.

Discussion


Occurrence of higher severity of white rust on leaves and greater number ofstagheads in later sown crops matched with earlier findings (Tomar et al. 1992;Meena et al. 2002). Mathur (1993) indicated higher disease severity during 35–70d.a.s., which was further pinpointed in this investigation. Our findings on rangeand correlation of different weather parameters with white rust severity in thefield matched with some earlier findings under laboratory conditions (Lakra andSaharan 1988b; Mathur 1993), others with those of Saharan et al. (1988) and theminor disagreement regarding range of temperatures favouring rust severity couldbe due to difference in requirements of the geographical isolates (Lakra andSaharan 1988c). The information on conditions favouring severity of white ruston leaves and formation of stagheads could be useful while providing predictionsrelated to the disease (Gilles et al. 2000). A delayed sowing results in coincidenceof the vulnerable growth stage of plants as indicated earlier with favourableweather conditions (high morning and afternoon RH, maximum daily tempera-ture: 16–248C; minimum daily temperature: 4128C). The sustenance of suchfavourable conditions decides the longevity of the period of rust attack andfurther build-up on the crop, which consequently affects yield. Thus, the damagecaused to a crop by white rust is likely to be related to sowing date, i.e. latesowing results in higher rust severity (Meena et al. 2002). Thus, under Indianconditions, it would be appropriate to sow the crop at the earliest possible timeto enable escape or non-coincidence of the flowering stage with favourabletemperature and humidity factors leading to higher rust severity on the leaves andstagheads, which affect yield. Thereafter, the later sown crop matures quickerthan the timely sown one due to rising of temperature towards the end of thecrop season, which leads to faster maturity of the former that results in loweryield. However, it may also be noted that early sown crops have escaped white


Dow

nloa

ded

by [

Nor

thea

ster

n U

nive

rsity

] at

20:

29 2

9 O

ctob

er 2

014

rust on leaf but have encountered high number of stagheads due to coincidenceof favourable weather factors with vulnerable crop stage (Kolte et al. 1986).Hence, it would also be unwise to consider a thumb rule of escaping the diseasein early sown crop, whereby importance of forecasting needs to be underlined inthis prominent disease of the crop in India.

Table 5. Validation of models for different dependent variables and cultivars at Bharatpur.

Cultivar

Dates ofsowing

(2007–2008)

Crop age(days after sowing)at first appearanceof white rust (Y1)

Crop age(days after sowing)at highest severityof white rust (Y2)

Highest % whiterust severity on

leaf/no. ofstagheads per

plot (Y3)

Predicted Observed Predicted Observed Predicted Observed

‘Varuna’/leaf

1 October 86 87 133 133 0.35 0.78 October 81 81 126 131 3.3 3.015 October 75 74 119 124 4.9 5.122 October 71 70 118 118 8.3 6.929 October 66 63 108 107 14.4 16.65 November 47 45 105 106 22.6 28.612 November 35 35 96 96 39.0 38.519 November 33 33 84 83 39.0 38.726 November 33 33 72 72 40.3 41.83 December 30 31 70 70 47.5 47.0

‘Rohini’/leaf


‘Varuna’/stagheadsper plot


‘Rohini’/stagheadsper plot



Dow

nloa

ded

by [

Nor

thea

ster

n U

nive

rsity

] at

20:

29 2

9 O

ctob

er 2

014

Forecasting models

The application of the first fungicidal spray is critical and differs with seasons andregions (Sansford et al. 1996) for different diseases. Hence, accurate region andcultivar specific forecast for the crop age at first appearance of the disease assumes

Table 6. Validation of models for different dependent variables and cultivars at New Delhiand Kangra.

Location/cultivar

Dates ofsowing

(2007–2008)

Crop age(days after sowing)at first appearanceof white rust (Y1)

Crop age(days after sowing)at highest severityof white rust (Y2)

Highest % whiterust severity on

leaf/no. ofstagheads

per plot (Y3)

Predicted Observed Predicted Observed Predicted Observed

New Delhi/‘Varuna’


New Delhi/‘BIO-902’


Kangra/‘Varuna’


Kangra/‘RCC-4’



Dow

nloa

ded

by [

Nor

thea

ster

n U

nive

rsity

] at

20:

29 2

9 O

ctob

er 2

014

importance, keeping in view the fact that sowing date is controlled by several factors,viz., available soil moisture, time of withdrawal of monsoon, ambient temperature,availability of field for sowing, etc. and need not be a choice of the farmer. Thus,based on the predictions of the time of first appearance of the rust on the crop (d.a.s.)available at least 1 week before the actual appearance of the disease and the riskinvolved on the crop as related to the disease, the growers could undertake timelyfungicidal sprays for better efficacy and avoid unnecessary ones.

Though the Indian mustard is grown across a large part of India and the whiterust disease is also found to be a problem at several of the cropping areas, theirconditions for crop culture vary widely along with the specific conditions thatfavour the disease at different locations. There also could be variation in pathotypeof A. candida across sites (Lakra and Saharan 1988c). These could be the reasonsfor different weather parameters getting entered in models for different locations.Most of the models saw the entry of variable maximum daily temperature withminimum daily temperature, morning RH, afternoon RH and sunshine hours alsogetting entered in some cases. Further investigation can pinpoint the importance ofthe different weather factors favouring white rust attack and progress in order oftheir priority. Though in the present study, relationship of A. candida biology withsome weather variables, viz., sunshine or light hours were not considered, themodels based on weather factors and rust severity on Indian mustard crop couldprovide effective prediction about the crop age as related to their time ofappearance, peak disease severity on leaf, staghead number on the crop and cropage at highest severity of the disease on leaves, highest staghead numbers. Further,in all the cases, the models invariably included temperature among the weatherfactors. Weather indices based on accumulated weightings of different meteor-ological factors as per correlation coefficients in different weeks after sowing untilthe forecast was provided, were taken into account. Proper monitoring of diseaseprogress during recording of observations in experiments could enable devisemodels for providing accurate forecasts of crop age at first appearance, crop age athighest severity and highest level of disease severity. We thereby do not intend toundermine the importance of studying the spore biology of A. candida in relationto weather factors. Rather we wish to state that such in-depth studies as done inAustralia, Canada and Europe for diseases of oilseed Brassicas (Gilles et al. 2000;West et al. 2001) could improve the accuracy of the models presented here as thatcould provide the grower with an advance warning of the risk of the disease, whichmay allow a period of several weeks in which to make a decision about fungicideapplications and the risk can be updated later (Gilles et al. 2000). But, in an effortto cater to the need for providing real-time region and cultivar specific forecasts ofthe white rust menace in India, we have made a beginning by-passing the study ofweather and spore biology relationship, which avoids the risk of inaccuracy,especially at the cultivar specific level as it is based on fewer relationships which donot fully describe the dynamics of the biological processes influencing the diseaseepidemics (Gilles et al. 2000).

This study fortunately is spread over several locations in the Indian mustardcropping regions of the country facing the white rust problem. Using these models incombination with crop planting dates and standard meteorological data, it would bepossible to provide necessary forecasts for the time being centrally from the NationalResearch Centre on Rapeseed–Mustard at Bharatpur. While on one hand, themodels could be improved with further detailed study on pathogen inoculum and


Dow

nloa

ded

by [

Nor

thea

ster

n U

nive

rsity

] at

20:

29 2

9 O

ctob

er 2

014

spore biology, there would be need to provide a simple computer package to enableany user to get a robust, accurate forecast on the internet. This is expected to guidegrowers efficiently for making fungicidal sprays more effective. The forecasters needto take into consideration the other findings, viz., boundary and favourableconditions for disease severity on leaves, stagheads reported here along with theoutput of location and cultivar specific models.

The available literature indicates association of several weather factors withwhite rust severity (Bains and Jhooty 1979; Kolte et al. 1986; Saharan et al. 1988;Lakra and Saharan 1988a; Hegde and Anahosur 1994; Mehta and Saharan 1998;Sangeetha and Siddaramaiah 2007) but is silent about providing forecast of thedisease. Hence, as per available literature, this seems to be the first report of devisingprediction models for forecast of the disease of the important oilseed crop. In yearsof appearance of white rust on crop before the decision week, growers may beadvised about the risk expected. Further, the forecasts need to account for themargin of error in order to maintain the confidence of resource poor mustardgrowers of India in the forecast system. A high priority over the next decade shouldbe the collation of accurate disease and weather data and development of models toforecast the effects of climate change on other plant diseases to provide the necessaryforesight for strategic adaptation to climate change (Evans et al. 2008). More studyin this direction to improve the models with additional data from similar and othernew experiments on pathogen inoculum and spore biology for real-time forewarningof outbreaks of the disease based on climatic variables is in progress.

Acknowledgements

The facilities provided by the Directors of research units where the investigation was carriedout, the funding received for the investigation from the Indian Council of AgriculturalResearch under the World Bank Funded National Agricultural Technology Project and theAll Indian Coordinated Research Project on Rapeseed-Mustard are gratefully acknowledged.Further, help received from Dr S.C. Mehta and Sh Praveen Kumar in analysis of data isthankfully acknowledged.

References

Agrawal R, Jain RC, Jha MP. 1986. Models for studying rice crop-weather relationship.Mausam. 37:67–70.

Bains SS, Jhooty JS. 1979. Mixed infections by Albugo candida and Peronospora parasitica onBrassica juncea inflorescence and their control. Indian Phytopathol. 32:268–271.

Chattopadhyay C, Agrawal R, Kumar A, Bhar LM, Meena PD, Meena RL, Khan SA,Chattopadhyay AK, Awasthi RP, Singh SN, et al. 2005a. Epidemiology and forecasting ofAlternaria blight of oilseed Brassica in India – a case study. Zeitschrift furPflanzenkrankheiten und Pflanzenschutz. (J Plant Dis Prot.). 112:351–365.

Chattopadhyay C, Agrawal R, Amrender Kumar, Singh YP, Roy SK, Khan SA, Bhar LM,Chakravarthy NVK, Srivastava A, Patel BS, et al. 2005b. Forecasting of Lipaphis erysimion oilseed Brassicas in India – a case study. Crop Prot. 24:1042–1053.

Conn KL, Tewari JP, Awasthi RP. 1990. A disease assessment key for Alternaria black spot inrapeseed and mustard. Can Plant Dis Surv. 70:19–22.

Desai AG, Chattopadhyay C, Ranjana Agrawal, Kumar A, Meena RL, Meena PD, SharmaKC, Srinivasa Rao M, Prasad YG, Ramakrishna YS. 2004. Brassica juncea powderymildew epidemiology and weather based forecasting models for India – a case study.Zeitschrift fur Pflanzenkrankheiten und Pflanzenschutz. (J. Plant Dis Prot.). 111:429–438.

Doorenbos J. 1976. Agrometeorological field stations. FAO paper no. 27 Irrigation anddrainage. Rome, Italy: Food and Agriculture Organization. 94 pp.


Dow

nloa

ded

by [

Nor

thea

ster

n U

nive

rsity

] at

20:

29 2

9 O

ctob

er 2

014

Draper NR, Smith H. 1981. Applied regression analysis. 3rd ed. New York: John Wiley, Inc.709 pp.

Evans N, Baierl A, Semenov MA, Gladders P, Fitt BDL. 2008. Range and severity of a plantdisease increased by global warming. J R Soc Interface. 5:525–531.

FAO (Food and Agriculture Organization). 2009; [cited 2009 Aug 14]. Available from: http://www.faostat.fao.org/site/339/default.aspx

Ghadekar SR. 2002. Practical meteorology – data acquisition techniques, instruments andmethods. Nagpur, India: Agromet Publisher. 68 pp.

Gilles Tijs, Evans Neal, Bruce Fitt DL, Michael Jeger. 2000. Epidemiology in relation tomethods for forecasting light leaf spot (Pyrenopeziza brassicae) severity on winter oilseedrape (Brassica napus) in the UK. Eur J Plant Pathol. 106:593–605.

Hegde VM, Anahosur KH. 1994. Influence of sowing date of mustard on the epidemiology ofwhite rust. Indian Phytopathol. 47:391–394.

Humpherson-Jones FM. 1992. The development of weather-related disease forecasts forvegetable crops in the UK: problems and prospects. Eur Plant Prot Organ Bull. 21:425–429.

IASRI (Indian Agricultural Statistics Research Institute). 2006. Agricultural ResearchData Book. New Delhi, India: Indian Agricultural Statistics Research Institute(ICAR). 248 pp.

Kolte SJ. 1985. Diseases of annual edible oilseed crops. Vol. II: Rapeseed-mustard and sesamediseases. Boca Raton, Florida, USA: CRC Press Inc. 135 pp.

Kolte SJ, Awasthi RP, Vishwanath. 1986. Effect of planting dates and associated weatherfactors on staghead phase of white rust and downy mildew of rapeseed and mustard.Indian J Mycol Plant Pathol. 16:94–102.

Lakra BS, Saharan GS. 1988a. Progression and management of white rust of mustard inrelation to planting time, host resistance and fungicidal spray. Indian J Mycol PlantPathol. 18:112.

Lakra BS, Saharan GS. 1988b. Factors affecting sporangial germination of Albugo candidacausing white rust of mustard. Indian Phytopathol. 41:287.

Lakra BS, Saharan GS. 1988c. Morphological and pathological variations in Albugo candidaassociated with Brassica species. Indian J Mycol Plant Pathol. 18:149–156.

Mathur S. 1993. Studies on oospore germination and development of a standard inoculationtechnique for white rust of mustard. Final Report of Adhoc ICAR Sponsored ResearchProject, Department of Plant Pathology, Rajasthan Agriculture University, BikanerCampus, Udaipur 313001, India. 49 pp.

Meena PD, Chattopadhyay C, Fateh Singh, Bhoori Singh, Ajit Gupta. 2002. Yield loss inIndian mustard due to white rust and effect of some cultural practices on Alternaria blightand white rust severity. Brassica. 4:18–24.

Mehta N, Saharan GS. 1998. Effect of planting dates on infection and development ofwhite rust and downy mildew disease complex in mustard. J Mycol Plant Pathol. 28:259–265.

Mehta N, Saharan GS, Kaushik CD. 1996. Efficacy and economics of fungicidal managementof white rust and downy mildew complex in mustard. Indian J Mycol Plant Pathol.26:243–247.

NRCRM (National Research Centre on Rapeseed–Mustard). 1999. Package of practices andcontingency plan for enhancing production of rapeseed-mustard. Bharatpur, India:National Research Centre on Rapeseed–Mustard, Sewar. 39 pp.

Saharan GS, Kaushik CD, Kaushik JC. 1988. Sources of resistance and epidemiology of whiterust of mustard. Indian Phytopathol. 41:96–99.

Sangeetha CG, Siddaramaiah AL. 2007. Epidemiological studies of white rust, downy mildewand Alternaria blight of Indian mustard (Brassica juncea (Linn.) Czern And Coss.).African J Agric Res. 2:305–308.

Sansford CE, Fitt BDL, Gladders P, Lockley KD, Sutherland KG. 1996. Oilseed rape:disease development, forecasting and yield relationships. HGCA Project Report no. OS17,London, UK.

Tomar SS, Tiwari AS, Indapurkar YM. 1992. Effect of dates of sowing and varieties on yieldand yield components of Indian mustard [Brassica juncea (L.) Czern. and Coss.]. NarendraDeva J Agric Res. 7:188–191.


Dow

nloa

ded

by [

Nor

thea

ster

n U

nive

rsity

] at

20:

29 2

9 O

ctob

er 2

014

http://www.faostat.fao.org/site/339/default.aspx

http://www.faostat.fao.org/site/339/default.aspx

Verma U, Bhowmick TP. 1988. Oospores of Albugo candida (Pers. Ex Lev.) Kuntze – itsgermination and role as the primary source of inoculum for the white rust disease ofrapeseed and mustard. Int J Tropical Plant Dis. 6:265–269.

West JS, Kharbanda PD, Barbetti MJ, Fitt BDL. 2001. Epidemiology and management ofLeptosphaeria maculans (phoma stem canker) on oilseed rape in Australia, Canada andEurope. Plant Pathol. 50:10–27.


Dow

nloa

ded

by [

Nor

thea

ster

n U

nive

rsity

] at

20:

29 2

9 O

ctob

er 2

014

epidemiology and development of forecasting models for white rust of brassica juncea ...

Documents