Prediction modeling of downy mildew for cut roses in greenhouses
[Objective]Downy mildew of rose can cause severe premature defoliation of rose plants,which poses a great threat to cut rose in greenhouses.The'moderate temperature and high humidity'environmental conditions required for downy mildew occurrence are common in the greenhouse,especially in autumn and winter.The study aimed to identify and model the relationship between the occurrence of downy mil-dew in different common cut rose varieties and greenhouse environmental factors in order to provide insight for ecologically friendly manage-ment of the disease and to build statistical models to predict downy mildew occurrence.[Method]Under indoor conditions,Peronospora spar-sa was inoculated in rose leaves.The incidence of downy mildew and the environmental factors were investigated in four cut rose varieties'Zixiaxianzi''Fenhongxueshan''Luoshen'and'Jiaoao'cultivated in the greenhouses of Baofeng Base,Jinning District,Yunnan Province from 2022 to 2023.Disease prediction models were built using logistic regression,multiple linear regression,and mixed models.[Result]In a temperature range of 5-30 ℃,downy mildew symptoms were observed,and the most suitable disease development temperature range was 18-20 ℃.The lowest humidity of infection of P.sparsa was 75％ ,and the optimal range was greater than 90％ .Data analyses showed that environmental factors had a significant impact on the occurrence of downy mildew in cut roses,and observed disease indices were linearly re-lated to the cumulative infection favorability of environment of the previous 3 days,5 days,7 days,and 15 days.Among them,'Zixiaxianzi'and'Fenhongxueshan'had a significant fitting,while'Luoshen'and'Jiaoao'varieties had a less significant model fitting.However,add variety factor as independent variable,models for the pooled data of all four varieties were highly significant,the optimal model had a Bayesian information criterion(BIC)of 89.189,a standard deviation of 0.023 and a RMSE of 0.041.[Conclusion]The environmental conditions 3 to 15 days before disease occurrence may have important effects,which can be modeled using the infection favorability.The cumulative value of infection favorability 15 days before the onset of disease has the greatest influence on the disease,and the above three models are the most sig-nificant.Therefore,predictive models-based on the favorability changes of environmental factors 7 to 15 days in advance can be used to a-chieve better disease management in greenhouse.
Cut rose downy mildewEnvironmental favorabilityPrediction modelLogistic regression model
万方数据
维普
