Short communicationA simple allometric model for estimating blueberry fruit weight from diameter measurements

- Data of four genotype and three growth season were obtained for estimating fruit weight from diameter measurements
- From data analysis, a common model for four cultivars was constructed, which followed a power fit.
- The model was validated with the independent data, giving errors ranging from 4.7% to 6.9%.
- Residuals showed a random dispersion pattern indicating homoscedasticity
- The model was adequate for providing a reliable means of estimating blueberry fruit weight of different cultivars.

Monitoring of fruit growth is a measurement widely used in physiology and agronomy studies. This is normally done from detached fruits, which can lead to erroneous results when fruits grow asynchronously in clusters as occurs in blueberry plants (Vaccinium spp.). The aim of this communication is to develop a simple allometric model for estimating blueberry weight from diameter measurements. In three growth seasons, weight and diameter of a total of 416 fruits were measured from four field-grown blueberry cultivars. The obtained data set was then randomly divided resulting in 200 data for calibrating model and 216 for validating. Prior to calibration procedure, the use of cultivar-specific models or a common model for four cultivars was evaluated by Analysis of Covariance between logarithm transformed diameter and cultivar/season variables. From analysis, no interactive effect was observed, so a common allometric model was constructed, giving a power functional relationship between berry diameter and weight. The allometric model was validated with the independent data, giving errors ranging from 4.7% to 6.9%. Residuals showed a random dispersion pattern. Our results showed that a common model was adequate for providing a reliable means of estimating blueberry fruit weight of different cultivars.