Instantaneous light measurements predict relative cumulative light levels within an apple canopy

Percent instantaneous incident photosynthetic photon flux density (%INPPFD) was measured Within an apple (Malus domestica Borkh.) canopy for various sky conditions and used to predict the percent cumulative incident photosynthetic photon density (PPD) for the last 10 weeks of the growing season (%CPPDLS) and the total growing season (%CPPDTS). Instantaneous measurements from overcast conditions were superior to measurements from clear or hazy conditions for the prediction of %CPPDLS in 1989 and 1990. A one-to-one relationship between %INPPFD and %CCPPDLS was found for overcast conditions in both years, even though there was an 11% difference in total cumulative PPD between the years. The models had good predictive accuracy, with prediction coefficients of determination (R2Pred) > 0.83 in both years (n = 30). %INPPFD from overcast conditions also yielded accurate predictive models for %CPPDTS (R2Pred > 0.84, n = 30), which differed from the models for %CPPDLS. Predictive models (for both %CPPDLS and %CPPDTS) from %INPPFD made before the canopy was fully developed differed from the models developed after canopy development was complete. The models still had good predictive accuracy, with R2Pred > 0.76 (n = 30). Predictive models developed for cloudless conditions had inferior predictive accuracy (R2Pred = 0.49 to 0.80, n = 30) compared to models for overcast conditions. R2Pred were higher for hazy than for clear conditions. Time of day (1000 to 1400 HR) had no consistent effect on the development of predictive models for any weather condition. The most reliable models resulted from the average of several measurements within a day, particularly for cloudless conditions.