Modelling the flight activity of APis melligera at the hive entrance

• Published in: Computers and electronics in agriculture Vol. 42; no. 2; pp. 87 - 109
• Main Authors: Devillers, J, Dore, J C, Tisseur, M, Cluzeau, S, Maurin, G
• Format: Journal Article
• Language: English
• Published: 01.02.2004
• ISSN: 0168-1699

Hives equipped with electronic bee counters were installed near sunflower fields located inthe agricultural district of Birieux (France). Counter data were collected during about 50 consecutive days from about mid-July to the beginning of September 2001. Temperature, relative humidity, rainfall, wind, and global radiation were also recorded. A palinological analysis was performed to help in the interpretation of the results. Principal compinent analysis (PCA) was first used to structure the different data matrices. Co-inertia analysis (CIA) was then employed to find co-structures between the outgoing bee data and climatic parameters. Lastly, quantitative models were derived from partical least squares (PLS) regression analysis. The obtained PLS regression results were confirmed by means of artifical neural network models, While promising results were obtained, they clearly showed the necessity to include additional variables in the modelling process.