Calcolo dei gradi-giorno per la previsione delle fasi fenologiche nell'actinidia

In the last years many studies were devoted to accurate predict crop developments. It was found that air temperature is a dominant factor controlling crop growth. Prediction of crop development is widely made using growing degree-days. Critical aspects of degree-days calculation are related to the selection of an appropriate base temperature, the methodology used to calculate degree-days, and the first day from which degree-days can be accumulated. While most research concerned the development of bioclimatic indices that can be used to predict the timing of the key development processes, such as phenological stages of various crops, there is limited information on Actinidia deliciosa. Some researchers found that the number of days between bud break and blooming diminished with increasing of temperature (Warringhton e Stanley, 1986; McPherson et al., 1992) and that a linear degree-days model accounted for 94% of the variance in the seasons examined (Morley-Bunker et al., 1987). Some climatic indices are studied to define the suitability of different areas to kiwifruit culture (Eynard, 1986; Spano et al., 1995). It is not clear the role of low temperature on bud differentiation in this species (Brundell, 1976; Guerriero et al., 1991). The goals of this study are (1) to verify the best methodology for calculating degree-days from daily maximum and minimum temperature, (2) to choose the base temperature for Actinidia deliciosa, and (3) to establish the most appropriate date from which degree-days can be accumulated. The experiment was made over a seven-year period (1987-1994) on an orchard located near Tempio, Sardinia, Italy (40° 50' N, 500 m altitude). Weekly phenological observations (bud break, flowering, fruit set, and commercial harvesting stages) were made on ten vines of kiwi fruit (cv. Harward) previously selected. Degree-days were calculated using three methods (daily averaging, hourly averaging, and single sine curve) and two base temperatures (8 °C and 10 °C). For three years (1992-1994), solar thermal units (STU) were also calculated (Caprio, 1974, 1991). Degree-days were accumulated from January 1 of each year and from the date in which chill unit (C.U.) requirement had been satisfied (550, 600, 650, and 700 C.U.). The standard deviation and the standard error in degree-days, and the coefficient of variation and the standard deviation in days were calculated in order to determine the best method to calculate degree-days and to choose the base temperature for Actinidia deliciosa. Analysis of the results showed no variation of the entire cycle duration for each of the seven years (? = ? 5 days). The number of accumulated degree days (2472) was in accordance with previous results showed by several authors (Eynard, 1986; Monastra and Insero, 1989). The smallest standard deviation in days was obtained using single sine method and 8 °C for all stages. Regression of the dates of phenological stage occurrence (Julian days) and number of accumulated degree-days showed that degree-days model accounted for about 97% of the variance among the years considered. No improvement of the method was found using solar thermal units (STU). Bud break was the stage that showed the highest variability in date occurrence. The use of chill units (C.U.) did not decrease the value of the standard deviation in days. To sum up, the results confirm the validity of degree-days model in predicting phenological stages for Actinidia deliciosa. The single sine method performed well for all stage but bud break.

In questo lavoro sono stati confrontati differenti metodi statistici utilizzati per la previsione della comparsa delle fasi fenologiche delle specie vegetali allo scopo di individuare la migliore metodologia di calcolo per l'Actinidia in relazione alla procedura usata per ricostruire la curva della temperatura giornaliera, alla scelta della migliore temperatura di soglia (Ts) e alla data di inizio per il conteggio dei gradi-giorno(°G). La prova è stata condotta per sette anni, durante i quali sono state registrate le date di comparsa delle fasi di germogliamento, fioritura, allegagione e maturazione. Per i sette anni sono stati calcolati i °G (utilizzando tre diversi metodi) con due differenti soglie termiche inferiori 8°C e 10°C e, per i tre anni per i quali erano disponibili dati di radiazione solare, anche le unità termiche solari (U.T.S.). Il confronto ha permesso di individuare il metodo del seno con soglia 8°C come quello con le migliori caratteristiche predittive per le fasi di fioritura, allegagione e maturazione presentando deviazioni standard in giorni rispettivamente di 1.1, 0.1 e 0.3. Il metodo non sembra invece sufficientemente affidabile per la previsione della comparsa della fase di germogliamento. Il tentativo di definire una più corretta data iniziale per il conteggio dei °G attraverso il calcolo delle unità in freddo non ha apportato peraltro alcun miglioramento.