Prospects for the creation and use of paired and multiple correlation and regression models in beekeeping

Abstract

In animal husbandry, including beekeeping, there are a growing number of independent consultancy services to analyse the performance of the industry in relation to disease monitoring status and preventive measures to maintain proper bee family health. In order to provide expert advice, these services must always be backed up by quality data and accurate statistical analysis. It would give clear instructions on how to interpret the results obtained when processing them, and show directions for improving disease prevention. Currently, there are problems related to improving the control of infectious diseases in bees, as various natural and anthropogenic factors have a multidirectional effect on the economic performance of beekeeping. There are also concerns about the control of infectious animal and insect diseases, which is a multifaceted series of causes due to natural and anthropogenic factors that have a polyvector effect on the economic performance of beekeeping. Therefore, the experimental application of different types of correlation and regression analysis in this industry by constructing pairwise and multivariate dependencies and their statistical interpretation was the aim of the paper. The correlation and regression model under study contains four sets of characteristics: result variable (y) - the amount of honey from 20 different apiaries in one season and factor variables: x1 - air temperature in the apiaries; x2 - amount of probiotic "Enteronormin Iodis + Se" to stimulate the immune system as one of the preventive methods; x3 - number of beehives in each apiary. Linear proportional relationships between apiary productivity and the factors included in the regression model are obtained. According to the results of the correlation-regression analysis, paired correlation coefficients showed that the relationship between air temperature in the apiary and produced honey is medium connection (r1 = 0,666), the relationship between the amount of probiotic applied per frame and produced honey is tight (close) connection (r2 = 0,813), the relationship between the number of beehives and produced honey is medium connection (r3 = 0,633). The regression coefficients show how the amount of honey produced in an apiary changes when each factor changes by one, with the other factors in the equation fixed. So, raising the temperature by 1 °C increases the honey production by 216 kg in each apiary, while increasing the concentration of "Enteronormin Iodis + Se" by 1 cm3 per beehive frame increases the nectar production by 1,12 kg for one hive. The coefficient of multiple determination (R2 = 0,954163) identifies a close relationship in the model created (95% of the factors investigated determine apiary performance). Therefore, modelling in the form of linear and multiple correlation and regression analysis is feasible in beekeeping. Key words: beekeeping, modeling, system analysis, factor and result characteristics.