Sensors 2020, 20, 117; doi:10.3390/s20010117 www.mdpi.com/journal/sensors
Article
Gas Sensor Array and Classifiers as a Means of
Varroosis Detection
Andrzej Szczurek
1
, Monika Maciejewska
1,
*, Beata Bąk
2
, Jakub Wilk
2
, Jerzy Wilde
2
and Maciej Siuda
2
1
Faculty of Environmental Engineering, Wrocław University of Science and Technology, Wybrzeże
Wyspiańskiego 27, 50-370 Wrocław, Poland; andrzej.szczurek@pwr.edu.pl (A.S.)
2
Apiculture Department, Warmia and Mazury University in Olsztyn, Sloneczna 48, 10-957 Olsztyn, Poland;
beata.bak@uwm.edu.pl (B.B.); teofil.wilk@uwm.edu.pl (Ja.W.); jerzy.wilde@uwm.edu.pl (Je.W.);
maciej.siuda@uwm.edu.pl (M.S.)
* Correspondence: monika.maciejewska@pwr.edu.pl
Received: 12 November 2019; Accepted: 20 December 2019; Published: 23 December 2019
Abstract: The study focused on a method of detection for bee colony infestation with the Varroa
destructor mite, based on the measurements of the chemical properties of beehive air. The efficient
detection of varroosis was demonstrated. This method of detection is based on a semiconductor gas
sensor array and classification module. The efficiency of detection was characterized by the true
positive rate (TPR) and true negative rate (TNR). Several factors influencing the performance of the
method were determined. They were: (1) the number and kind of sensors, (2) the classifier, (3) the
group of bee colonies, and (4) the balance of the classification data set. Gas sensor array
outperformed single sensors. It should include at least four sensors. Better results of detection were
attained with a support vector machine (SVM) as compared with the k-nearest neighbors (k-NN)
algorithm. The selection of bee colonies was important. TPR and TNR differed by several percent
for the two examined groups of colonies. The balance of the classification data was crucial. The
average classification results were, for the balanced data set: TPR = 0.93 and TNR = 0.95, and for the
imbalanced data set: TP = 0.95 and FP = 0.53. The selection of bee colonies and the balance of
classification data set have to be controlled in order to attain high performance of the proposed
detection method.
Keywords: honey bee; gas sensor; Varroa destructor; classification
1. Introduction
Honey bees have an essential role in nature that goes beyond the production of honey and
beeswax. They play a vital role in the environment by pollinating both wild flowers and many
agricultural crops. The honey bee population has strongly declined in recent years [1] due to a
combination of multiple stresses, including diseases, pathogens and pesticides. Bees, like all animals
including humans, are susceptible to viruses (e.g., deformed wing virus, Israeli acute paralysis virus,
Kashmir bee virus, Black Queen cell virus and Sacbrood); bacteria (e.g., American foulbrood and
European foulbrood); fungi (e.g., Chalkbrood and Stonebrood); microsporidian parasite (e.g., Nosema
apis and Nosema cearanae); parasitic mites (e.g., Tracheal mites (Acarapis woodi), Varroa mites, and
Varroa mites resistant to fluvalinate and coumaphos); and insect pests (e.g., wax moth, and small
hive beetle) [2–5].
A bee colony has defense mechanisms, but sometimes they are too weak. In this situation, insects
need help from a beekeeper. An essential part of beekeeping is the inspection of colonies for bee
diseases. Early signs of infection normally go unnoticed; they only become apparent when infection
