Predicting honey production based on morphological characteristics of honey bee (Apis mellifera L.) using multiple regression model

摘要

The main objective of the present study was to model the relationship between honey production (HP) and morphological characteristics of Iranian honey bee. 600 colonies of 12 apiaries located in 12 provinces of Iran were selected for sampling to conduct this research. The results showed that HP had positive and significant correlation with forewing length (FWL, r = 0.690**), forewing width (FWW, r = 0.258**), cubital index (CI, r = 0.109**), third and fourth tergite length (TFTL, r = 0.200**) and sternite index (SI, r = 0.419**) and negative significant correlation with hind wing width (HWW, r = -0.157**) and scutellum color (SC, r = -0.081*). Multiple linear regression analysis indicated that the predicting model for HP explained 51.8% of the totalvariation within the measured traits. The residual plots analysis indicated no problem in the model with selected variables. On the other hand, t-test showed that some of the variables are not important to be present in this model. The results of t-testand stepwise multiple linear regression analysis indicated that FWL (R~2 = 47.7 %), SI (R~2 = 1.7%), HWW (R~2 = 1%) and HLL (R2 = 0.8%) were the best morphological characters of honey bee for predicting HP.