IAP-25-068

Din in the dark: the effects of anthropogenic noise on hibernation behaviour and physiology in hedgehogs

Rapid environmental change from urbanisation, particularly nocturnal noise and light, is altering microhabitat quality and disrupting the environmental cues animals rely on. These shifts pose particular risks for species with tightly constrained life cycles, such as hibernators, which must compress weight gain and reproduction into a brief active period.

Hibernation is a physiological strategy that dramatically lowers metabolic demand and enables survival when resources are scarce. Successful torpor depends on stable, low‑disturbance conditions and sufficient energy reserves. Increased sensory disturbance from urbanisation could increase arousal frequency, deplete fat stores and produce sublethal physiological costs that reduce survival and fitness.

Despite these predicted impacts, fine-scale empirical evidence linking sensory disturbance – and noise in particular – to hibernation patterns is limited. Most studies investigate hibernation using coarse metrics, such as sighting intervals, and focus on survival outcomes rather than the mechanistic pathways connecting disturbance, arousal dynamics and physiological consequences. The roles of noise, its temporal structure, and its interaction with microclimate in shaping hibernation patterns and selection of hibernation sites across urban–rural gradients, remains unexplored.

The European hedgehog, Erinaceus europaeus, is an excellent model to address this gap. It is widely distributed across urban and rural habitats and is amenable to minimally-invasive biologging (spine-mounted thermistors) that yield continuous temperature and activity records (Crawford et al., 2025). This project will combine continuous biologging, nest-scale environmental monitoring, experimental manipulation, hormone analysis and machine learning to link sensory disturbance with hibernation physiology. The work will generate mechanistic insights and practical guidance for stakeholders to mitigate sensory impacts on overwintering wildlife.

Project Aims

1. To quantify nest-scale disturbance and microclimate throughout the hibernation season across an urban-rural gradient.
Metrics: mean, maximum and variability in sound pressure, illuminance and air temperature.
2. To test for associations between disturbance metrics and
a. hibernation patterns (hibernation duration and timing, number of arousals, days torpid, number of nest changes).
b. faecal cortisol before and after hibernation.
c. body condition before and after hibernation.
3. To experimentally test effects of elevated noise on hibernation patterns, cortisol concentrations and body condition.
4. To develop and apply machine learning tools for signal extraction and prediction from multimodal sensor streams.
5. To translate results into practical management recommendations in collaboration with the People’s Trust for Endangered Species (PTES).

FIELDWORK
Fieldwork will be conducted along an urbanisation gradient between the University of Glasgow’s SCENE field station and Glasgow following methods described in Crawford et al. (2025). In brief, the student will search for hedgehogs at night during the late active period using a handheld spotlight. Hedgehogs will be captured by hand and marked. A temperature sensing VHF transmitter and light sensor will be attached to their spines (together weighing <2% of hedgehog body mass). Faecal samples and body condition indices will be collected before and after the hibernation period. Radiotracking will be used to locate the hibernation nest and monitor nest changes. During hibernation, skin temperature will be recorded every 10 min via an automated receiver. Environmental sensors will record nest scale ambient temperature and disturbance levels continuously through the hibernation season at each nest.

Experimental manipulation: A subset of nests will be assigned to controlled noise treatments or matched controls in a welfare guided, before after/control impact design.

LABORATORY ANALYSIS
Faecal samples will be analysed for cortisol metabolites using enzyme-linked immunosorbent assay (ELISA).

DATA PROCESSING AND MODELLING
Sensor, telemetry and audio data will be processed through a structured pipeline for quality control, synchronisation and feature extraction. Patterns of torpor and arousal will be identified using statistical segmentation and classification approaches, with predictive models trained and validated to detect arousal events. These outputs will be integrated into statistical frameworks to estimate the effects of treatment and environmental conditions on hibernation dynamics, physiological stress and body condition.

Year 1

Induction; literature review; graduate-level courses in key skill development based on student’s interests and needs (e.g. programming, GIS, welfare assessment); training in machine learning and version control; equipment calibration; external lab visit and field training (led by Dr Julia Nowack, Liverpool John Moores University); animal tagging and radio-tracking during hedgehog active period 1; pre-hibernation sampling 1.

Year 2

Hibernation season 1 data collection; PTES Hedgehog Research Meeting; post-hibernation sampling 1; statistical analysis; animal tagging and radio-tracking during hedgehog active period 2; pre-hibernation sampling 2.

Year 3

Hibernation season 2 data collection; machine learning; post-hibernation sampling 2; lab analysis; statistical analysis; international conference; PTES end-user translation meetings; writing thesis and manuscript drafting.

Year 3.5

Finalising thesis and manuscript drafting.

The successful candidate will be supported by an interdisciplinary cross-institution supervisory team to develop field, laboratory and desk-based skills in:
• Biologging, radio-tracking, sample collection and animal handling in the field – deploy and retrieve environmental loggers, fit and check tags, follow welfare‑led capture/handling protocols
• Equipment calibration and quality control
• Laboratory assays (ELISA)
• Ethics, welfare and risk management
• Data engineering and integration
• Machine learning
• Reproducible statistical analysis and modelling
• Programming in R and/or Python
• Research integrity, open science and data governance
• Communication, leadership and impact
• Translating results into practical management recommendations in collaboration with a conservation NGO
The student will gain experience working as part of a small team in the field, and in problem-solving, resilience and time management and will have the opportunity to gain graduate teaching experience. They will also have the opportunity to audit graduate courses according to their interests and background.