Bee Cluster 3D: A system to monitor the temperature in a hive over time
Olivier Besson, Hervé Mugnier, Aurélien Neveux, Gaëlle Rey, Sully Vitry
BBee Cluster 3D: A system to monitor the temperature in a hiveover time
Olivier Besson , Hervé Mugnier , Aurélien Neveux , Gaëlle Rey , Sully Vitry . University of Neuchâtel, CH-2000 Neuchâtel, Switzerland Mind Technology, F-74166 Saint-Julien-en-Genevois, FranceAll authors contributed equally to this work.* Corresponding author: Olivier BessonInstitute of mathematics, University of Neuchâtel11, Rue Emile ArgandCH-2000 Neuchâtel, SwitzerlandE-mail address: [email protected]
Abstract
A new system, Bee Cluster 3D, allowing the study of the time evolution of the 3D temperaturedistribution in a bee hive is presented. This system can be used to evaluate the cluster size and thelocation of the queen during winter. In summer, the device can be used to quantify the size of thebrood nest and the breeding activity of the queen. The system does not disturb the activity of thecolony and can be used on any hive. This electronic system was developed to be non-intrusive,miniaturized, and energy autonomous.
Keywords
3D temperature distribution time evolution in a hive, winter loss of colonies, brood nest timeevolution, bee cluster time evolution .
Beekeepers know that the temperature distribution in a bee hive is an important parameter forcolony development, see also (Abou-Shaara et al., 2017; Barter, 2016; Basile et al., 2008; Lensky,1964; Dietman et al., 2014; Tautz et al., 2007) and included references.Various temperature measurement and observation methods have been developed, see, e.g.,(Dunham,1931). A detailed history of these observations is presented in (Meikle et al., 2015) and the includedreferences. Recently, (Barter, 2016; Niewiatowski et al., 2016; Stalidzans et al., 2013; Becher et1 a r X i v : . [ q - b i o . Q M ] J a n l., 2009) developed new data-logging systems. These systems have few sensors, so they cannotprovide a 3D representation of the hive temperature.A new methodology to measure the real-time temperature distribution in a bee hive is presentedin this paper. The sensor locations provide a 3D representation of this distribution and enableinvestigation of:• the density, the morphology and the size of the bee cluster in the hive;• the quality, the size, and the distribution of the brood nest;• the location of the queen in winter.Moreover, the system can be used to evaluate the colony health status and:• to know the temperature distribution in the hive as a function of time;• to have a primary indicator of winter loss of colonies, with a study of risk factors;• to study the influence of the materials used (e.g., wood, polyester) on heat loss.Owing to the fine mesh on each frame, the Bee Cluster 3D system can measure thermoregulationin the colony, providing a 3D thermal image with the following advantages.• The system is non-intrusive in the sense that it is not necessary to open the hive to performmeasurements.• The system is non-invasive because the sensors merge into the the wax frame.• The information is transmitted in real time.• The system is energy self-sufficient and is designed for field use.In general, the presented innovation covers the development of a multi-sensor modular systemfeaturing real-time information about hive development. This system is integrated into a "plug andplay" sensor architecture that feeds data to a network database.This database can be connected to mathematics and vision software on a computer, smartphone, ortablet PC or to a dedicated interface.From a technical perspective, the main advantage of the Bee Cluster 3D system is the integrationof the sensors into the wax frame, enabling natural development of the colony. This integrationsallows the beekeeper to work in the hive and avoid disconnection and connection of the lines andmanipulation of the frames. 2 Materials and Methods
An overview of the data acquisition system is presented in figure 1. The system consists of theequipped hive, the recording frames, a power supply system, a data transmitting/receiving compo-nent, and a computer for the database.Figure 1:
Data collecting system.
The main data collecting components are the frames equipped with sensors. A frame without waxis shown in figure 2. It is equipped with• 64 heat sensors with a spacing of 3 cm and an accuracy of 0 . ◦ C .• 2 humidity sensors with an accuracy of 0 . Frame with electronic components.
Frame with a wax comb foundation.
Figure 4:
Built frame.
The results presented in this section were obtained during the winter of 2014-2015 and the springof 2015. The observed hive was formed during autumn of 2014 as a 4-frame store. During thewinter, the hive had 4 frames, and new frames were introduced during the spring of 2015.Our measurements enabled the following observations5 During the winter period, the maximum temperature remained greater than 30 ◦ C in a small,slowly moving region. This region corresponds to the location of the queen.• During spring, the activity of the queen resulted in an increase of the brood nest.• Our measurements provide an estimate of the evolution of the number brood cells.Table 1 contains the temperature extrema observed from August 22, 2014 to September 29, 2015,both in the hive and in a covered space near the hive.This table shows that the maximum temperature remained above 33 ◦ C throughout the winter. Aswill be seen in section 3.1, a small volume in the hive remained warmer than 30 ◦ C.6 ate Hive temperature ◦ C External temperature ◦ C From To Minimum Maximum Minimum Maximum22/08/2014 31/08/2014 15.90 37.6031/08/2014 05/09/2014 17.80 40.2021/09/2014 25/09/2014 12.60 36.5030/09/2014 08/10/2014 17.70 36.80 13.30 26.3008/10/2014 16/10/2014 14.30 35.90 12.00 25.3008/12/2014 14/12/2014 0.10 33.40 1.10 25.0014/12/2014 21/12/2014 1.30 35.00 2.70 15.5029/12/2014 04/01/2015 -4.00 33.30 -2.60 11.8011/01/2015 18/01/2015 -2.60 33.50 -0.90 13.9018/01/2015 25/01/2015 -0.60 33.20 -4.70 16.1026/01/2015 31/01/2015 -1.60 33.70 0.40 6.3031/01/2015 06/02/2015 -3.10 33.00 -1.60 8.1008/02/2015 14/02/2015 -3.70 33.60 -1.20 11.1014/02/2015 20/02/2015 -1.90 34.50 0.10 9.7002/03/2015 06/03/2015 1.30 35.3006/03/2015 15/03/2015 -1.20 35.8015/03/2015 23/03/2015 3.50 36.7023/03/2015 30/03/2015 5.80 37.70 1.90 18.3030/03/2015 06/04/2015 6.20 37.70 2.10 18.7006/04/2015 13/04/2015 6.20 42.30 2.70 23.8016/04/2015 24/04/2015 14.20 43.00 6.00 24.2024/04/2015 02/05/2015 22.00 42.4002/05/2015 10/05/2015 27.00 37.5017/05/2015 22/05/2015 22.50 38.0022/05/2015 10/06/2015 23.20 37.9010/06/2015 19/06/2015 25.20 37.2019/06/2015 21/06/2015 26.50 37.3029/06/2015 02/07/2015 21.10 37.8023/07/2015 24/07/2015 15.40 37.9028/07/2015 03/08/2015 23.50 36.7003/08/2015 05/08/2015 18.60 37.9026/08/2015 03/09/2015 11.80 37.8003/09/2015 09/09/2015 14.20 39.4026/09/2015 29/09/2015 15.80 38.30
Table 1:
Temperature extrema
Detailed results are presented in the following sections. In part 3.1, the results obtained during thewinter of 2014-2015 are given. Section 3.2 is devoted to the spring of 2015. The study of the broodevolution is analyzed via graphic localization and by estimating the number of brood cells on eachframe in the hive. 7 .1 Winter temperature distribution in a hive
As already mentioned, throughout the winter, the maximum temperature in the hive remains above30 ◦ C in the small region where the queen is located. However, the temperature in other parts ofthe hive can be below 0 ◦ C, as shown in table 1.The three following figures show the temperature distributions in the hive on January 18, 2015at 6 AM. Note that the temperature extrema in the hive are -2.6 ◦ C and 33.1 ◦ C. All the winterobservations are similar to those presented below. As usual during winter, the bee cluster is locatedin the front of the hive.Figure 5 shows the location of the 30 ◦ C isotherm surface in the hive; the queen is located insidethis surface. Figure 5: ◦ C isotherm surface in the hive.
Figure 6 shows the 22 ◦ C isotherm surface. The bee cluster is included in this surface.8igure 6: ◦ C isotherm surface in the hive.
Finally, Figure 7 shows the temperature distribution of a slice in the middle of the hive.9igure 7:
Temperature distribution in the center of the hive.
Brood development in hives occurs in the spring. The Bee Cluster 3D system can be used toprecisely follow this evolution via a graphical representation of the volume of the brood nest andby quantifying the number of brood cells on each frame.In this section, observations made on the hive equipped with the Bee Cluster 3D system are pre-sented for April 17, 2015. During this period, the hive had seven frames, numbered 1 to 7.Figures 8 and 9 show images of the brood nest from above and below, respectively. These repre-sentations are the 35 ◦ C isotherm surface in the hive.10igure 8:
Brood nest from above.
Figure 9:
Brood nest from below.
11s already mentioned, the total number of brood cells in the hive can be estimated in real time.For this process, the brood quantity on each frame is obtained along with an estimate of the area ofthe surface delimited by an T ◦ C-isotherm curve, where T is given by the user. Then, the numberof brood cells is given by the formula N = AA c (1)where N is the number of cells, A is the area occupied by the brood cells, and A c is the area of onecell. A c is given by A c = √ d i = √ d e (2)where d i is the diameter of the inscribed circle, and d e is the circumscribed circle of a cell. Thefactor 2 in formula (1) accounts for the fact that a frame has two faces.In the following examples, we use the following parameters:• T = . ◦ C • d i = . mm The number of brood cells on each frame, calculated with the preceding method, is given in table2. Frame Number of brood cells1 8902 10183 15784 295 5506 07 0Table 2:
Number of cells on the frames
Note that frame 6 has no brood cells because this frame was introduced one day previously. Figure10 shows the activity of the bees building this frame. Note that the maximum temperature on thisframe is 42 ◦ C. 12igure 10:
Frame 6 under construction.
The following figures show the temperature distributions on the frames and the 35.5 ◦ C isothermcurve. 13igure 11:
Isotherms for frame 2.
Figure 12:
Isotherms for frame 3.
Isotherms for frame 4.
Figure 14:
Isotherms for frame 5. Conclusion
A new method to continuously monitor the temperature on the frames in a bee hive has beenpresented. The fine mesh on each frame gives a precise 3D representation of the temperature inthe hive.The main advantages of this Bee Cluster 3D system are the following.• The system is non-intrusive in the sense that it is not necessary to open the hive to performmeasurements.• The system is non-invasive since the sensors merge into the the wax frame.• The information is transmitted in real time.• The system is energy self-sufficient and is designed for field use.The presented innovation covers the development of a multi-sensor modular system integrated intoa ”plug and play” architecture.It can be used by any beekeeper because the equipped frames can be manipulated as normal.Our results show that in a small region of the hive, the temperature remains above 30 ◦ C throughoutthe winter, even if the external temperature is below 0 ◦ C. The queen is located in this region.The Bee Cluster 3D system also provides very useful information concerning the development ofthe brood nest.The authors declare no conflicts of interest. This research did not receive any specific grant fromfunding agencies in the public, commercial, or not-for-profit sectors.