Field Report - Gaulossen Acoustic Monitoring

Executive Summary

Deployed passive acoustic monitoring equipment at Gaulossen Nature Reserve during autumn migration period to assess avian biodiversity using automated deep learning classification. Despite challenging weather conditions (80%+ rain/fog coverage throughout deployment), successfully recorded 48.8 hours of continuous acoustic data, yielding 74 verified bird species from 4,023 detections. Deployment confirms Gaulossen's importance as wetland stopover habitat along East Atlantic Flyway.

Observer: George Redpath

Affiliation: Norwegian University of Science and Technology (NTNU), Department of Acoustics

Location: Gaulossen Nature Reserve (63.341°N, 10.215°E)

Deployment Duration: 48.8 hours continuous recording

Equipment: AudioMoth v1.2 autonomous acoustic recorder

Equipment Configuration

Parameter	Specification
Recorder Model	AudioMoth v1.2 (Open Acoustic Devices)
Dimensions	35 × 58 × 23 mm
Weight	55g (including batteries)
Microphone	Knowles SPU0410LR5H-QB-7 MEMS
Sampling Rate	48 kHz (Nyquist: 24 kHz)
Bit Depth	16-bit (96 dB dynamic range)
Recording Mode	Continuous (no duty cycling)
Storage	32 GB microSD card
Power Source	3× AA Alkaline batteries (Energizer)
Housing	Waterproof green case (IP67 rated)
Mounting Height	1.5 m above ground level
GPS Coordinates	63.341°N, 10.215°E (±10m accuracy)
Distance to Wetland	~100m to primary congregation area

Day 1: October 13, 2025

Deployment (11:37 Local Time)

Time	Temp (°C)	Conditions	Wind	Visibility	Notes
11:37	7	Heavy rain, fog	Light W	<500m	Deployment start
14:00	8	Rain, overcast	Moderate W	~800m	Geese highly active
16:00	9	Rain continuing	Light W	<1km	Peak flock event begins
19:00	7	Rain, darkening	Light	<300m	Crepuscular activity peak
22:00	6	Rain, fog	Calm	<200m	Nocturnal period begins

Gaulossen Nature Reserve wetland area - shallow marshes, reedbeds, and open water with mountain backdrop. Deployment site selected for unobstructed acoustic sight lines.

Gaulossen artsliste for fugler (bird species list) at reserve entrance, documenting over 200 historically recorded species.

Key Observations - Afternoon (11:37-17:30)

Time	Species	Activity	Visual Count	Behavior Notes
12:30	Graylag Goose	Flock calling	~200	High vocal activity
14:00	Hooded Crow	Perching/calling	3-5	Near goose flock
16:00	Graylag Goose	PEAK EVENT	200+	91-min continuous calling

Peak Flock Event (16:00-17:26)

Graylag Goose (Anser anser): Most intensive vocal activity occurred during 91-minute flock event. Visual observation estimated 200+ individuals with continuous calling. Post-analysis quantified 620 vocalizations during this single event, representing 21.6% of all Graylag detections in 1.9% of recording time (extreme concentration).

Corvid Activity: Hooded Crow and Carrion Crow (5-10 individuals) frequently calling in proximity to goose flocks, consistent with sentinel mutualism behavior pattern.

AudioMoth deployment location at wetland edge. Equipment mounted on fence post structure approximately 1.5m above ground.

Evening Crepuscular Activity (19:00-22:00)

Great Snipe (Gallinago media): Dusk period yielded peak detection (82 calls at 20:00), confirming autumn migration stopover use of wetland. These are contact/foraging calls, NOT spring lek displays (which occur April-May). No visual confirmation obtained due to cryptic species behavior and poor visibility.

Day 2: October 14, 2025

Morning Observations (06:00-12:00)

Time	Temp (°C)	Conditions	Wind	Visibility	Notes
06:00	8	Heavy fog	Calm	<200m	Pre-dawn period
08:00	9	Fog, rain	Light	~500m	Dawn chorus peak
10:00	10	Rain, fog clearing	Light NW	~1km	Improved visibility
12:00	11	Overcast, light rain	Moderate	~1.5km	Midday conditions

Dawn Chorus (06:00-09:00)

Time	Species	Activity	Audible Calls	Notes
07:00	Eurasian Woodcock	Roding flights	Frequent	Distinctive calls
08:00	Grasshopper-Warbler	DAWN PEAK	Intense	Continuous reeling
08:00	Waterfowl	Foraging	Moderate	Flock activity

Dawn Chorus Highlights

Common Grasshopper-Warbler (Locustella naevia): Intense reeling trill at 08:00 (51 calls in single hour - 86% of total daily detections, 98% of calls at dawn). Classic dawn chorus territorial singing. Mechanical, insect-like continuous trill.

Eurasian Woodcock (Scolopax rusticola): Roding flight displays 07:00-08:00 peak (31 calls detected at 08:00). Crepuscular specialist (75% of detections at dawn/dusk periods).

Cryptically-colored bird species observed in wetland vegetation during deployment. Excellent example of natural camouflage adaptation in wetland habitat.

Waterfowl congregation on Gaulossen wetland during deployment. Overcast conditions and dramatic sky typical of recording period. Large flock visible on water surface.

Day 3: October 15, 2025 - Equipment Recovery

Pre-Dawn Nocturnal Migration (02:00-05:00)

Nocturnal flight calls documenting active East Atlantic Flyway passage:

Pink-footed Goose: 18 nocturnal flight calls (peak 03:00)
Common Crane: 11 calls (peak 04:00)
Greater White-fronted Goose: 18 calls
Total: 47 nocturnal migration calls

Equipment Status at Recovery

Parameter	Status	Notes
AudioMoth Power	Operational	Green LED indicating recording
Battery Voltage	~3.8V	Sufficient for 24+ additional hours
Memory Card	48.8 hours recorded	~35 GB data captured
Housing Integrity	Waterproof seal intact	No moisture ingress
Microphone Condition	No visible damage	Clean, no water in port
Data Integrity	All files playable	No corruption detected

Equipment Performance: Excellent

AudioMoth v1.2 performed excellently throughout 48.8-hour deployment despite continuous rain exposure. All 48.8 hours of audio data (35.2 GB) successfully captured with zero file corruption. Performance exceeded expectations for adverse weather deployment.

Extensive goose flock formations observed during deployment. Visual documentation corresponding to acoustic data showing 2,871 Graylag Goose detections (69.9% of total vocalizations). Multiple flock groups visible in formation.

Post-Deployment Analysis Summary

Automated Classification (BirdNET v2.4)

Initial detections: 90 species, 6,805 total detections
After human verification: 82 species, 4,108 detections (91.1% species pass rate)
After biological screening: 74 species, 4,023 detections (82.2% final pass rate)

Key Behavioral Findings

1. Social Behavior Dominance

86% of detections from flock/social species
Graylag Goose: 98.7% of calls in flocks (59 discrete flock events)
Largest flock: 620 calls over 91 minutes

2. Corvid-Waterfowl Sentinel Mutualism

8,778 geese-crow co-occurrences within 10-minute windows
Pattern consistent with heterospecific eavesdropping
Crows providing early warning, geese benefiting from predator detection

3. Migration Documentation

37 migratory species detected (45% of verified species)
47 nocturnal flight calls documenting active flyway passage
Great Snipe: 189 detections representing autumn stopover

4. Temporal Specialization

Common Grasshopper-Warbler: 98% dawn activity
Great Snipe: 61% crepuscular (69% dusk-only)
Mallard: 59% nocturnal (biologically valid for ducks)

Conservation Implications

                Gaulossen as Critical Stopover Habitat
                Species Diversity: 74 verified species in 48.8 hours demonstrates high biodiversity
Declining Species Present: Great Snipe, Corn Crake, River Warbler (rare vagrant)
Flyway Importance: Nocturnal migration calls confirm active East Atlantic Flyway corridor
Habitat Quality: High social species concentration indicates productive foraging habitat

            

Passive Acoustic Monitoring Value

Weather-independent data collection (visual surveys impossible during fog/rain)
24-hour coverage capturing nocturnal/crepuscular species
Permanent acoustic archive for future reference
Non-invasive methodology suitable for sensitive species

Challenges and Lessons Learned

Weather-Related Issues

Rain noise contamination: Required extensive post-processing (Wiener filtering, HPSS)
Sampling bias: Cannot claim weather correlations - 80%+ recording during rain/fog
Great Bittern false positives: 129 detections from rain impact noise

Methodological Improvements for Future Deployments

Deploy 2-3 units for spatial redundancy
Co-deploy weather station for correlation analysis
Extend to 7+ days for temporal robustness
Repeat spring deployment to capture breeding season vs. migration

Data Availability

Raw Data: 48.8 hours continuous recording archived at NTNU Digital Repository

Processed Datasets:

247 enhanced audio clips (Wiener + HPSS noise reduction)
247 Raven Pro-style spectrograms (2048 FFT, 512 hop, 0-12 kHz)
BirdNET detection outputs (CSV format)
Temporal clustering analysis (flock detection)
Co-occurrence matrices (sentinel mutualism analysis)

Code Repository: https://github.com/Ziforge/gaulossen-study

Interactive Website: https://ziforge.github.io/gaulossen-study/