Sonar or Magnetism: exploring bat migration using a data science approach

Overview

Migration is inherent to a substantial proportion of Earth’s wildlife. It has been extensively studied in marine species (fish, reptiles, sea mammals) and birds, but less is known about bat migration. Migratory bats travel closer to the surface than birds and are threatened by man-made structures, such as wind turbines. Knowledge about their migration and movement behaviour is of further concern, given the role that bats play as reservoirs for emerging diseases such as the recent COVID-19 [1].

Migratory animals use various strategies to find their way, including different types of compasses (Sun, polarized light, stars and geomagnetic compass) and cognitive maps [2]. Most bats are nocturnal and so they migrate and hunt at night. This requires both an advanced obstacle avoidance ability and ability for long-distance navigation. For obstacle avoidance and short foraging trips over several hundreds of meters, bats use a natural sonar: they emit ultrasonic waves and sense their reflections from the environment [1]. However, their migration flights are fast and straight and can extend to hundreds of kilometres, suggesting use of other navigational strategies. They have recently been shown to use the Sun’s position at dusk to calibrate their direction [3], but this does not explain how they are able to navigate during the long night flights. One possibility is geomagnetic navigation, since Earth’s magnetic field provides a global source of information [2]. Bats have magnetite tissue structures and may therefore be able to detect changes in geomagnetic field [4], but how they use this information for navigation is unknown, which is what this project will explore.

Animal migration research has benefited from advances in tracking technologies that allow collection of high-resolution locational data (e.g. GPS loggers). However, bats are small (some weight less than 10g), which means that even the smallest currently available GPS trackers are still too heavy to place on individuals. One of the main ways to track migratory bats therefore remains radio-tagging (with tags of <1g in weight), where migrating bats are tracked either as they pass through a stationary array of antennas (Motus.org) or are followed on the ground or from an airplane [5,6]. Unlike high resolution GPS tracking however, radio-tagging data are sparse, temporally irregular and often only cover parts of the entire migration path.

A further complication making bat geomagnetic navigation difficult to study is that, to date, no method exists to connect radio-tagged data on bats – movements to real-time data on geomagnetic field. Earth’s magnetic field is a very dynamic system which responds to the influence of the solar wind (a constantly emitted stream of particles from the Sun), that can distort the field during geomagnetic storms. Such distortions may affect animal migratory navigation and have been linked to whale strandings. Knowing what geomagnetic conditions migratory animals experience on their journeys is therefore crucial to improve our understanding of how they use the field for navigation. While we developed a method that fuses real-time satellite geomagnetic data with high-resolution GPS tracking data [7], fusing geomagnetic data with sparse, irregular and sporadically missing radio-tagging bat data poses a significant data science challenge.

Methodology

This interdisciplinary project will employ data science to explore how migratory bats navigate with geomagnetic field. The objectives are:

Objective 1: To develop a new data fusion method of satellite geomagnetic data with radio-tagging bat migration data.
We will develop a new method to link these two data sources, using principles from spatio-temporal data fusion and indicative data science (which builds information from missing data). We will create reasonable missing movement information from sparse and irregular radio-tagging data using statistical properties of known movements and then link these new data to satellite geomagnetic information.

Objective 2: To explore how bats use geomagnetic information in a data-driven approach, without a priori supporting any theory. We will develop agent-based models (ABM) to simulate different navigational strategies (compasses and cognitive maps) and evaluate results against actual bat migration data to identify which of the migratory strategies are most likely to have occurred in reality.

Objective 3: To explore bat migration across several species through use of open data.
We will re-use migratory bat tracking data from open data portals (Movebank.org, Motus.org) for two European species, the common noctule bat (Nyctalus noctula, [5,6,8]), the Nathusius’ bat (Pipistrellus nathusii [9]); and the American silver-haired bat (Lasionycteris noctivagans, [10]). These are temperate bats that migrate over 1000km.

Scientific reproducibility: this project is based on open science principles. We will use open data (tracking and geomagnetic) and produce new methods as open tools (as Free and Open Source Software). Besides exploring bat migration, this project has a potential for wide impact in ecology through applicability of our methods to any other migratory species where radio-tagging remains the main source of tracking (e.g. songbirds).

Project Timeline

Year 1

1) Advanced training in software/data science skills (e.g. advanced Python skills), 2) literature review on bat migration, 3) start developing the data fusion tool.

Year 2

1) Finish data fusion tool, including writing a journal paper on the topic, 2) start the ABM study for geomagnetic bad navigation for one bat species as proof of concept.

Year 3

1) Finish ABM single-species study with a paper and 2) start working on multiple-species ABM study including all available data.

Year 3.5

Finish the multiple-species study with a journal paper and write PhD thesis.

Training
& Skills

Student will identify where he/she needs skills in the Training Needs Analysis. These will be acquired in GradSkills courses by The Centre for Educational Enhancement and Development (CEED) in St Andrews, IAPETUS2 training events and in external courses, such as AniMove (http://animove.org/).

Training will be through active supervision, consisting of regular meetings (live/virtual) of the supervisory team with the student. Student will attend research meetings and seminars of the Bell-Edwards Geographic Data Institute (https://begin.wp.st-andrews.ac.uk/), on research and career progression topics (publishing, time management, proposal writing). Student’s progress will be formally evaluated through annual review at the University of St Andrews

References & further reading

1. https://doi.org/10.1016/B978-0-12-809633-8.20764-4
2. https://doi.org/10.1038/s41586-018-0176-1
3. https://doi.org/10.1016/j.cub.2019.03.002
4. https://10.1371/journal.pone.0001676
5. https://doi.org/10.1002/ecy.2762
6. https://doi.org/10.1371/journal.pone.0114810
7. Benitez-Paez et al. 2020. Fusion of wildlife tracking and satellite geomagnetic data for the study of animal migration, under review.
8. Migration data, common noctulus: https://motus.org/data/project?id=113
9. Migration data, Nathusius bats: https://motus.org/data/project?id=311
10. Migration data, Silver-haired bat: https://motus.org/data/project?id=71, https://motus.org/data/project?id=79

Further Information

Dr Urska Demsar is Senior Lecturer in Geoinformatics and an expert in movement analytics. She will be the main supervisor and will support the student in research, career planning, publishing and all other aspects. Contact: urska.demsar@st-andrews.ac.uk

Prof Ana Basiri is Professor of Geographic Data Science and an expert in Indicative Data Science. She will support the student with advice on indicative data analysis (missingness) of radio-tagged data

Dr Ciaran Beggan is Geophysicist and an expert in geomagnetism. He will support the project with advice on geomagnetic data fusion and geomagnetic data.

Dr Jed Long is Assistant Professor in GIScience and has experience in spatial ecology. He will provide expertise for Motus data and support the student in ABM modelling

Dr Kamran Safi is a Research Scientist and Principal Investigator of the Computational Ecology lab. He will support the project in the biological feasibility aspects and with animal migration expertise.

Apply Now