Modeling seasonal flushing and shoot growth on different citrus scion-rootstock combinations

Abstract

Occurrence, intensity, and growth patterns of new shoots (NS) were evaluated in orchards of ‘Pera’ and ‘Folha Murcha’ sweet oranges (Citrus × aurantium var. sinensis L.), ‘Ponkan’ mandarin (C. reticulata Blanco) and Persian lime (C. × latifolia var. latifolia) grafted on nine rootstocks to determine potential impacts of scion-rootstock combinations on NS growth parameters favorable to Diaphorina citri. NS number and the area under flush shoot dynamics were used to compare NS intensities. NS number (dependent variable) and meteorological elements (independent variables) were used through multiple linear regression to describe NS occurrences. NS growth data were compared through non-linear regression based on degree-days. NS occurrence dynamics were similar for all combinations, but the intensity was significantly higher on ‘Florida’ rough lemon than on ‘Flying Dragon’ trifoliate. Occurrence of NS on ‘Pera’, ‘Folha Murcha’ and Persian varieties associated positively (P < 0.05) with an increase of minimum and average temperatures (avg. adjusted R2 = 0.406; 0.408; 0.403, respectively), whereas on ‘Ponkan’ was related to the accumulated rainfall (avg. adj. R2 = 0.311). NS grew faster and attained longer lengths during spring (102.7 mm during 34 days) than autumn (71.9 mm during 42 days), except for ‘Folha Murcha’ which took longer to grow and reached shorter lengths (78.8 mm during 38 days) during spring, and longer lengths during autumn (118.3 mm during 42 days). Logistic models better fitted NS growth data (avg. adj. R2 ≥ 0.94). The adjusted R2 values of regression models developed to predict NS occurrence during the study were not high enough to estimate optimum timing to start insecticide applications for the control of D. citri. However, the adjusted R2 values attained for NS growth can be used to improve timing, and thereby adjust frequency of insecticide applications.