Fig 1.
Wireless biologging system overview.
(A) Animal-borne mobile nodes document animal-animal meetings, which are triggered by mobile-node beacons 24/7 and independently of the ground infrastructure. Each mobile node forwards its meeting data when it receives beacons from a ground node that is dedicated to downloading and storing data. (B) When a tagged animal enters a grid of localization nodes (depicted by an antenna with red/blue gain patterns), a beacon of a tracking-dedicated ground node triggers the transmission of localization packets from the mobile node to the localization nodes. RSSIs of the impinging localization packets are then sent from the localization nodes to a work station via a WLAN. (C) Long-range bursts, which contain encoded sensor data, are received by long-range receivers. Long-range telemetry enables data transmission over distances of several kilometers at a low data rate. RSSI, received signal strength indicator; WLAN, wireless local area network.
Fig 2.
High-resolution association data in wild vampire bats.
(A) Meeting history of a single vampire bat (ID 56; 50 tagged bats in total) with other tagged bats. Red lines show meetings between bat 56 and other tagged bats (right-hand y-axis). The black line shows the degree centrality (number of associated tagged bats, left-hand y-axis) of bat 56 every 2 s. Date and time are on the x-axis. Shaded areas indicate night time. Vertical dashed lines show egocentric social networks at each snapshot of time during roosting (B, C) and foraging (D). Associations with the focal bat are indicated by red lines. Data and software used to create this figure have been archived by GFBio (https://doi.org/10.7479/vd6t-7a92; https://doi.org/10.7479/ytdf-wf05).
Fig 3.
Tracking bat movements in a forest.
(A) Tracking grid in a deciduous forest of Forchheim, Germany, consisting of 17 localization nodes (gray dots) covering an area of approximately 1.5 ha. Dashed black line: known reference path; blue line and blue shading: estimated path and average localization error obtained by the presented wireless biologging system; yellow lines: 4 individual GPS tracks. (B, C) Estimated flight trajectories of a tagged mouse-eared bat during foraging on August 2nd and 5th. Data and code used to create this figure have been archived by GFBio (https://doi.org/10.7479/vd6t-7a92). GPS, global positioning system.
Fig 4.
WBN tracking performance versus GPS tracking.
Localization errors of a reference path of approximately 300 m by the WBN are shown for different numbers of tracking nodes (6–17) in a deciduous forest of approximately 1.5 ha area. The average positioning error of 4 tracks of a heavy-duty commercial wildlife GPS tracker is shown for comparison by a yellow dashed line. Data and code used to create this figure have been archived by GFBio (https://doi.org/10.7479/vd6t-7a92). GPS, global positioning system; WBN, wireless biologging network.
Fig 5.
Energy distribution of software tasks of a mobile node powered by a 22 mAh battery.
Energy demand per software task depends on parameter settings for active/inactive beacon interval (s) and amount of time an animal spends in the localization grid (h). The energy demand is shown for the 7 major software tasks. Zero time in the localization grid (A, B) refers to a pure proximity sensing scenario. Data underlying this figure have been archived by GFBio (https://doi.org/10.7479/vd6t-7a92).
Table 1.
Estimated runtimes of mobile nodes for 2 battery capacities of 12 or 22 mAh inferred by an energy model for mobile-node runtime.
Although the model comprises 7 energy consuming tasks, the shown runtimes are based only on varying beacon intervals of mobile nodes and localization time (i.e., animal is within the localization grid). For mobile node beacon intervals, 2 operation modes are possible, depending on whether an animal is within reception range of a ground node (inactive mode) or not (active mode).
Table 2.
Overview of tracking systems to track locations and/or encounters in small vertebrates (animal-borne tag ≤ 2.5 g).