Management of Brown Spot Disease of Rice and Studies of Growth Rate of Disease on Application of Different Synthetic Fungicides by using Different Statistical Tools

Rice is a staple food to more than half of the world population around 4 billion people. It is a staple food to two third of Indian (Rout and Tiwari, 2012). It is estimated that 3.4 billion people eat rice everyday (irri.org/news-and-event/news/scaling-sustainable-rice-farming-practices-achieve-food-security-asia, 2020). 

In terms of global rice production India remained as single second largest country with 118.00 million metric tons and China being world number one with 146.73 million metric tons. (worldagriculturalproduction.com/crop/rice, aspx, Apr.16, 2020). Although India held a prominent position in global rice areas and production, the productivity per unit area by world standard is still low with average productivity of about 2.39 t/ha, whereas, in case of China it is 6.71 t/ha. One major factor for low productivity of rice in India is due to pest and disease incidence. Several pathogenic and non pathogenic diseases caused an extensive economic loss to rice crops. The losses due to rice diseases have been estimated to be 10-15% in general (Kandhari, 2005). Among the pathogens, fungi alone account for nearly 30 diseases of rice in the country (Rangaswami et al. 2002). The brown spot disease of rice incited by Helminthosporium oryzae is one major fungal diseases that caused yield loss of upto 45% when no coverage of plant protection were given. 

(http://www.knowledgebank.irri.org/training/factsheets/pestmanagement/diseases/item/brown-spot). Brown spot disease of ricehas been reported to occur in all rice growing countries including Japan, China, Burma, Sri Lanka, Bangladesh, Iran, Africa, South America, Russia, North America, Philipines, Saudi Arabia, Australia, Malaysia and Thailand, (Ou, 1985; Khalili, et al. 2012). In India it was known to occur in all rice growing states but was found more severe in dry and direct seeded rice in the state of Bihar, Chhatisgarh, Madhya Pradesh, Orissa, Assam, Jharkhand and West Bengal(Gangopadhyay, 1983; Sunder, et al., 2014). 

At present era of agriculture, predominant means of crop protection is the use of chemicals. However, the efficacy of existing pesticides available in the open market always need to be thoroughly evaluated so as to deal effectively with the target pest without loss of time, energy and capital, since most chemicals are costly and its indiscriminate use has also resulted a serious ecological and adverse effect on the human and animal health which has become a major global issue. A judicious application of pesticides needs to be advocate at the highest level through researched an extension activity for monitoring economic losses as well as copping with the environmental issue. Hence, the present work was undertaken to resolve issue of menace of brown spot disease of rice and indiscriminate use of chemical through proper evaluation of selected chemicals by using statistical tools such as, Logistic growth model and Gompertz model.

Disease progress curves over time have been referred to be as the "signature" of the epidemic and represent an integration of an all host, pathogen and environmental effects occurring during the epidemic (Campbell and Madden, 1990). A disease progress curve shows the epidemic dynamics over time (Agrios, 2005). This mathematical tool can be used to obtain information about the appearance and amount of inoculums, changes in host susceptibility during growing period, weather events and the effectiveness of cultural and control measures. Growth models provide a range of curves that are often similar to disease progress curves (Van Maanen and Xu, 2003) and represent one of the most common mathematical tools to describe temporal disease epidemics (Xu, 2006). The growth models commonly used are: Monomolecular, Exponential, Logistic and Gompertz (Zadok and Schein, 1979; Nutter, 1997; Nutter and Parker, 1997; Xu, 2006). A brief description of each growth model is presented as follows: Equations with linear parameters from each of four models of Richard’s family of growth curves, i.e. monomolecular (ln[1(1-y)]=ln[1/(1-y0)]+rMt, logistic(ln[/(1-y)]= ln[y0/(1-y0)]+rLt,log-logistic(ln[y/(log-logistic logistic(ln[/(1-y)]= ln[y0/(1-y0)]+rLt and Gompertz (-ln[-ln(y)] = -ln[-ln(yo)]+ rGt were employed as predicted equations to statistically compare linearly transformed data (Campbell and Madden, 1990; Nutter and Parker, 1997). Variables were: y=Mean severity of disease(S) as a proportion from 0 to 1 at time t, yo=the initial disease level and r*=rate of disease increase for each model. After the regression analysis, goodness of fit of the models was determined by examining the coefficient of determination R2, which is the proportion of the variation in the data accounted for by the variation in the data error of estimates (SEE), and the plot of the standardized residuals versus the predicted values. An R2 ≥ 80% is the desirable: if R2 ≤ 50%, the model fits the data poorly.

To compare models using different transformation of the dependent variables for goodness-of-fit, predicted transformed y was back-transformed and the co-efficient of determination calculated based on these values (R*2) (Campbell and Madden, 1990). Having selected the most suitable models, regression analysis was performed between observed and back-transformed dependent variables. Analysis of variance (ANOVA) was used to reveal any significant difference between the regions in rated parameters.