EVALUATION OF MULTIPLE REGRESSION MODELS TO PREDICT CITRUS CANKER EPIDEMICS - PowerPoint Presentation

EVALUATION OF MULTIPLE REGRESSION MODELS TO PREDICT CITRUS CANKER EPIDEMICS Muhammad Mohsin RazaDepartment of Plant PathologyUniversity of Agriculture Faisalabad

INTRODUCTION Production Valuable Fruit 6th Largest Producer 7373.5 Thousand Hectares Export 0.2 MT Value generated US $30 Millions Juices, Dishes Beverages, Medicinal Values World 538 (MT ) Pakistan 21.32 (MT )

CITRUS QUICK DECLINE CTV & CGD CITRUS SLOW DECLINE Tylenchulus semipenetrans CITRUS CANKER Xanthomonas axonopodis pv. citri CITRUS WITHER TIP Colletotrichum gleosporioids CITRUS GUMMOSIS Phytophthora spp CITRUS SUDDEN DEATH Phytophthora &Fusarium spp Citrus PATHOLOGICAL THREATS TO CITRUS IN PAKISTAN

Citrus Canker Xanthomonas axonopodis pv. citri Potential Threat Most serious and destructive disease Numerous Epidemics Symptoms Leaf Lesions Stem Lesions Fruit Lesion

Epidemiology Influence of the environmental variables Disease variability Mathematical models Host resistance Most promising solution Artificial inoculation Evaluation of degree of susceptibility or resistance Disease prediction and timely management

OBJECTIVES Citrus canker d isease prediction for timely management

Collection of citrus Germplasm Nursery Establishment Inoculum Collection Pathogen Identification Artificial Inoculation Data Collection Statistical Analysis Line of Work Diseased Nursery

Experimental Strategy Citrus limonia cv. china lemon Citrus reticulata cv. KinnowCitrus reticulata cv. Feutral’s earlyCitrus limettioidesCitrus jambhiri Citrus Germplasm RXR = 5ʹ PXP = 5ʹ Establishment of citrus NurseryRCBD 3 Replicates

Collection of InoculumDiseased leaves and other plant parts Pathogen IdentificationMorphological characters Biochemical characters Artificial Inoculation Spray method Injection infiltration Smear method Pinprick method

Injection infiltration Pinprick method Spray method Smear method

Data Collection Disease Incidence = ×100   Statistical Analysis AUDPC   Regression Analysis Best model selection R 2 (coefficient of determination ) Adjusted R 2 Mallows Cp Mean square error (MSE)

RESULTS Citrus Cultivars Disease Incidence (%) AUDPC Citrus limonia cv. china lemon 34.58 e 358 Citrus reticulate cv. kinnow 36.66 d 389 Citrus reticulate cv. Feutral’s early 39.50 c 421 Citrus limettioides 44.94 b 505 Citrus jambhiri 47.79 a 535 Disease Incidence (mean) and total AUDPC (% days) on citrus cultivars

Comparison of Inoculation Techniques in Citrus limonia cv. china lemon

Comparison of Inoculation Techniques in Citrus reticulate cv. kinnow

Comparison of Inoculation Techniques in Citrus reticulate cv. Feutral’s early

Comparison of Inoculation Techniques in Citrus limettioide

Comparison of Inoculation Techniques in Citrus jambhiri

Correlation of environmental factors with citrus canker in citrus cultivars Varieties/cultivars Temperature ( o C) Relative Humidity (%) Rainfall (mm) Wind Speed (km/h) Max. Min. Citrus limonia cv. china lemon -0.3281 (0.0228) 0.4815 (0.0005) 0.3362 (0.0195) -0.0781 (0.5977) -0.1566 (0.2878) Citrus reticulate cv. kinnow -0.2414 (0.0983) 0.3970 (0.0052) 0.2959 (0.0412) -0.1163 (0.4313) -0.1762 (0.2308) Citrus jambhiri -0.1667 (0.2576) 0.2559 (0.0792) 0.1844 (0.2095) -0.0529 (0.7212) -0.0912 (0.5375) Citrus reticulate cv. Feutral’s early -0.2733 (0.0602) 0.4087 (0.0039) 0.2925 (0.0437) -0.0640 (0.6655) -0.1197 (0.4177) Citrus limettioides -0.3191 (0.0271) 0.4985 (0.0003) 0.3664 (0.0104) -0.1019 (0.4906) -0.1661 (0.2591)

CHARACTERIZATION OF ENVIRONMENTAL FACTORS Maximum disease was recorded at :- Temperature 22-28 ( o C )Relative humidity 60% Rainfall 3 mmWind speed 4.5 km/ hr

Stepwise RegressionMultiple Regression ModelFour Environmental Variable Model Minimum TemperatureRelative humidityRainfallWind Speed 93% R 2 MODEL DEVELOPMENT

Environmental Parameters R2 Adj. R 2 Cp MSE Prob > F Overall summary of Model Minimum Temperature ( o C) Relative Humidity (%) Rainfall (mm) Wind Speed (km/ hr ) Citrus limonia cv. china lemon Minimum Temperature ( o C) Rainfall (mm) Wind Speed (km/ hr )   0.93    0.96  0.93   0.96  4      5.86   112.54       45.43   <0.0001*       < 0.0001* 0.0441* 0.0088* Summary of Stepwise Regression developed to predict Citrus Canker on Citrus Cultivars. Citrus reticulate cv. kinnow Minimum Temperature ( o C)   0.94   0.93   11.19   86.30   <0.0001* Citrus jambhiri Relative Humidity (%)   0.90   0.90   1.96   241.24   <0.0001*

Environmental Parameters R2 Adj. R 2 Cp MSE Prob > F Citrus reticulate cv.Feutral’s early Minimum Temperature ( o C)   0.94   0.94   8.67   95.40   <0.0001* Citrus limettioides Relative Humidity (%) Minimum Temperature ( o C) Wind Speed (km/hr)Rainfall (mm)  0.98 0.98   4.00   36.70   <0.0001* 0.0318* 0.0318* 0.0038*

Models (y = bo + b 1 x + b 2 x…) Citrus Canker Incidence Observed Predicted R 2 y = -7.31+0.46x 1 +1.35x 2 +4.35x 3 -17.73x 4   (where x 1 indicates minimum temperature, x 2 relative humidity, x 3 rainfall and x 4 wind speed respectively)34.00 33.00 39.00 33.8333.8338.330.93 y1 = 103.05-3.00x 1+2.74x2-7.34x3  (where x1 indicates minimum temperature, x 2 rainfall and x 3 wind speed respectively) 35.00 40.00 35.00 36.58 41.58 32.16   0.96 y 2 = 14.86+2.64 x 1 ( where x 1 indicates minimum temperature)   29.00 35.00 40.00 29.33 34.75 39.08 0.94 y 3 = -3.35+0.63 x 1 (where x 1 indicates relative humidity) 50.00 51.00 43.00 45.97 45.97 48.18 0.90 y 4 = 15.78+2.87 x 1 (where x 1 indicates minimum temperature) 42.00 35.00 43.00 42.80 33.60 42.80 0.94 y 5 = 68.00-0.37 x 1 +4.49 x 2 -6.46x 3 -2.39x 4 (where x 1 indicates relative humidity, x 2 minimum temperature, x 3 wind speed and x 4 rainfall) 50.00 40.00 45.00 51.50 42.58 47.25 0.98

CONCLUSION All citrus cultivars found susceptible Injection infiltration resulted in maximum disease generation F our environmental variable model (Temperature, relative humidity, rainfall and wind speed) explained 93% variability

Any Question?