Automatic Environmental Monitoring of Honeybee Hive

Monitoring system component

The design of the monitoring system includes various components from the sensing, data logging and processing (Fig. 1). For this, as an initial step, we employed a multiplexer (DEV-09832 Mux) that is a device that selects one of several analog or digital input signals and forwards the selected input into a single line for 27 sensors. .For temperature sensing, TMP36 sensor was used. The TMP36 is a low voltage, precision centigrade temperature sensor. It provides a voltage output that is linearly proportional to the Celsius temperature. It also doesn’t require any external calibration to provide typical accuracies of ±1°C at +25°C and ±2°C ver the -40°C to +125°C temperature range. For tracking real time, the DS1307 real-time clock (RTC) was employed. This is a low-power, full binary-coded decimal (BCD) clock/calendar plus 56 bytes of nonvolatile SRAM. Address and data are transferred serially to the micro controller through an I²C, bidirectional bus. The clock/calendar provides seconds, minutes, hours, day, date, month, and year information. For the data logging, SD (Secure Digital) memory card (INT106D1P) was used. Mainboard for microprocessor was Arduino328.

Fig. 1.

Architecture of the temperature monitoring system prototype.

Programing

Fig. 2 shows the algorithm for our monitoring system. It begins with the presence of the SD cards. If it detects a SD card presence then a log file is created otherwise the program stops, and starts a timer with 5 minutes. If the timer does not expire, we read the current time and the ADC (Analog-to-Digital Converter) values from the 27 temperature sensors. In this case, analogous voltage data are converted into the temperature digits (°C), then recorded into the memory card. After this process, program switches to sleep mode to save the power until the battery remains.

Fig. 2.

Flowchart of the system algorithm for the temperature sensing, recording and storing.

Field verification and validation

For the verification of the process and validation of the data, field implementation of the monitoring system was set-up in the experimental apiary of Andong National University. Three queen-right colonies of Apis mellifera were selected. Each colony in a single story hive consisted of 7 frames full loaded with normal brood rearing. In the beginning, system verification was conducted, and found there was no problem in the sensing and data retrieving. However, the temperature profiles (27.5±2.7°C) were different to the temperature ranges measured by Hobo pro V2 (Onset Computer Cooperation, Pocasset, MA, USA) system (33.9±0.69°C). The verification process by the programing improved the data converting and resulted in the normal range of temperature (33.6±0.6°C). For the validation and monitoring the spatial distribution of the temperature inside the hive, twenty seven monitoring sensors were deployed into the hive to cover the three dimensional space inside the box (Fig. 3). Details were displayed in Fig. 3. Each point of monitoring spots were named as inside, center and outside or front, mid and back from the hive entrance in horizontal view (Fig. 3a), and lower, middle and upper on the vertical view (Fig. 3b). Honeybee colonies were facing south. Monitoring data acquisition was made every 5 minutes for two weeks in July and August without replacement of the battery.

Fig. 3.

Sensor distribution inside the hive; horizontal view (a), vertical view (b) and whole system view (c).

Temperature distribution

The monitoring system provided valid temperature data inside the hive without further provision of external power. Temperature in the central part was more stable than at the inside or outside within the hive (Fig. 4). Total 3880×27 data points were acquired without error. There were significant variations of the temperature profiles. From the inside measure points, temperature fluctuated with the range of 25.7~40.8, and 24.2~40.2 from the outside part of the hive (Fig. 4a, 4c). However, temperature profiles in the central frames were rather stable especially in the middle and lower part compared to the upper part (Fig. 4b). However, overall temperature was maintained as 32.9± 1.03°C Thermoregulation of honeybee has been well demonstrated through diverse mechanisms (Batian and Esch, 1970; Southwick and Mugaas. 1971), and the range seemed the species-specific (Burgett and Titayavan, 1993, Burgett et al., 1997, 1998). The data seemed a bit different to Yi and Jung (2010) where they showed that the brood temperature was rather stable regardless of the outside temperature fluctuation. In their study, they kept the sensor in the middle of the brood comb while the current study sensors were either systematically allocated. However, there still exist some chance of the sensor precision and resolution.

Fig. 4.

Temperature profiles from inside (a), center (b) and outside (c) of the honeybee colony from three vertical layers from July 22 to Aug. 4.

Spatial temperature profiles are displayed in Fig. 5 which was measured in the morning (Fig. 5a) and at noon (Fig. 5b) when the sun shined directly on the front part of the bee hive. In the early morning, front part was cooler than the other sides and temperature gradients were more on the lower part of the bee hive. However at noon, temperature gradient steeped in the upper part where most of the heat produced and gained crowded in the upper part between the top cover. Since the temperature regulation is the integration of the individual honeybee’s decision (Jones et al., 2004), distribution of different individuals could be reflected into the variation of the temperature variation.

Fig. 5.

Exemplified spatial distribution of the temperature profiles from the early morning (left) and noon (right) at each vertical layer in a summer day.

Implementation

Rapid improvement of environmental monitoring and information technology now allow the precision agriculture through the sensor networks. In the field of beekeeping temperature measurement system can be adapted for the individual bee hive monitoring and maintenance, substituting the traditional monitoring methods. Bee hive temperature measurements can provide beekeeper with continuous, real time data about the state of the bee colony. Through the continuous monitoring of the hive temperature, onset of the brood rearing can be detected (Seeley, 1985; Lee et al., 2008; Zacepines, 2012), and may regulate the process in case of need to slow down, because too early start can lead to the death of the whole colony. From the measurement results during the passive wintering period, while bees are kept in the specific wintering building, Zacepines (2012) showed it is possible to precisely find out the date, when the bee colony died and find out the dates of the potential start of the brood rearing process. Also, Ferrari et al. (2008) reported that by monitoring of the hive environmental variables, swarming can be early-warned, so to escape from the loss of bees from swarming. The initiation of a swarming period is characterized as the raise in temperature from 33°C to 35°C until the actual time of swarming when the temperature drops to 32°C. With more activity, ventilation from bee wings causes drop of temperature. The disturbance-induced heat production or sound production can be detected. For example, if the vespid hornets begin to attack, disturbance of the bees inside the hive can be observed and possible results in the increase of the hive temperature (CJ, personal observation). Ferrari et al. (2012) showed that acoustic method can detect. In that; the microphones were placed inside the hives. The sounds were recorded with a sample rate of 2 kHz, and analyzed. During the monitoring, swarming activities occurred and was indicated by an increase in the power spectral density at about 110 Hz; approaching to swarm the sound augmented in amplitude and frequency to 300 Hz, occasionally a rapid change occurred from 150 Hz to 500 Hz (Ferrari et al., 2008).

The prototype of the environmental monitoring system of the hive needs further improvement especially in the precision and power supply. Also the sensor should have enough resolution to catch up the subtle change of the inside hive environment such as the heat from the swarm or nose itself. Also, surprisingly higher variation was noted even in the middle part of the colony which make the further development more complicated. However, sensor network of the hive environmental monitoring eventually could help mitigate the labor cost to inspect the each colony health from the remote area, if the wireless long distance communication be possible, and contribute to the bee health management.