Study Area: Virunga Massif, Rwanda / DRC / Uganda
Species: Gorilla beringei beringei (Mountain Gorilla)
Tools: Python · GeoPandas · SciPy · Matplotlib · Seaborn
Data: SMART Conservation Software GPS Export
This project analyzes GPS movement data from habituated mountain gorilla groups in the Virunga Massif to characterize:
- Daily movement distances and temporal consistency
- Home range estimates (MCP and KDE methods)
- Inter-group spatial separation and avoidance behavior
- Nest-site predictability for field monitoring efficiency
GPS fixes are recorded every 10 minutes by field researchers walking with gorilla groups from morning nest departure (~06:00) to mid-afternoon (~14:00), SMART Conservation Software exports.
The project was built around four core questions:
1. How do mountain gorilla groups move through the Virunga Massif? What are the daily path lengths, movement speeds, and directional patterns; and are they consistent enough to be useful for field operations?
2. How large are the home ranges of each group, and do they vary by season? Using both MCP and KDE methods; what territory does each group occupy, and does rainfall season affect ranging behavior?
3. Do gorilla groups actively maintain spatial separation from one another? This was the central hypothesis test: are observed inter-group distances greater than what would occur if groups moved independently with no awareness of each other?
4. How predictable is a group's daily location from its previous night's nest site? A question with direct conservation and operational value; can field rangers reliably relocate groups each morning?
Finding 1 — Daily movement is short and temporally consistent The mean daily path length was 1.30 ± 0.19 km across all six groups, with a coefficient of variation of just 0.15. This low variability means group movement is highly routine and predictable from day to day. Pablo was the most active group (1.51 km/day), Umubano the least (1.12 km/day), consistent with group size differences.
Finding 2 — Nest sites are highly predictable The mean nest-to-nest distance between consecutive days was only 0.40 km. Critically, 70% of all nightly transitions placed the group within 500 meters of the previous nest site, and 99% were within 1 km. This means field researchers can begin each morning's follow from the previous evening's nest location with very high confidence of finding the group; a direct operational gain for monitoring efficiency and ranger logistics.
Finding 3 — Groups actively maintain spatial separation (H₀ rejected) This was the statistically strongest finding. Using a permutation null model where each group independently drew random positions from its own empirical location cloud (simulating movement with no intergroup awareness), the null expected minimum pairwise distance was 1.54 km. The observed mean minimum distance was 4.54 km ; a difference of +3.00 km. The permutation test returned p < 0.001 with a Cohen's d of 2.05, a large effect by any standard. All three tests (permutation, t-test, Mann-Whitney U) agreed. H₀ was rejected.
Finding 4 — Spatial avoidance reflects behavioral regulation of group stability The biological interpretation of Finding 3 is that gorillas are not simply wandering independently; they are actively tracking and avoiding neighboring groups. Intergroup encounters in mountain gorillas carry high costs: silverback dominance contests risk serious injury, females may switch groups (a reproductive cost to resident males), and the energetic disruption cascades across the group. Spatial avoidance is therefore a behavioral mechanism that stabilizes group composition and reduces conflict.
Finding 5 — Home ranges are moderate and show seasonal variation KDE 95% utilization distributions ranged from roughly 18–34 km² depending on group. Larger groups (Susa, Pablo) used larger ranges, consistent with published literature on group-size range scaling. Home ranges contracted slightly during the long rains season, likely reflecting higher food availability reducing the need to travel.
| Coefficient of Variation (CV) Bullet | Core Finding |
|---|---|
| GPS movement characterization | Mean 1.30 km/day, low Coefficient of Variation — short and consistent |
| Home range + intergroup metrics | 18–34 km² KDE ranges; 4.54 km mean group separation |
| Hypothesis testing | p < 0.001, d = 2.05 — active avoidance confirmed |
| Nest-site predictability | 70% within 500m — groups reliably relocatable |
| Visualizations for reporting | 16 publication-ready figures across all analyses |
gorilla-movement-ecology/
├── README.md
├── .gitignore
├── requirements.txt
├── data/
│ ├── raw/gorilla_gps_data.csv <- GPS tracking data
│ ├── processed/ <- Cleaned datasets
│ └── README_data.txt
├── src/
│ ├── __init__.py
│ ├── movement_metrics.py <- Distance, speed, turning angle calculations
│ ├── home_range.py <- MCP and KDE home range estimation
│ ├── intergroup_analysis.py <- Pairwise distance and overlap metrics
│ └── visualization.py <- Plotting utilities
├── notebooks/
│ ├── 01_data_loading_exploration.ipynb
│ ├── 02_movement_analysis.ipynb
│ ├── 03_home_range_estimation.ipynb
│ ├── 04_intergroup_dynamics.ipynb
│ ├── 05_hypothesis_testing.ipynb
│ └── 06_final_visualizations.ipynb
└── outputs/
├── figures/ <- All saved plots
└── reports/ <- Summary CSVs and tables
- Movebank: https://www.movebank.org
Search: "Mountain gorilla Gorilla beringei" - GBIF: https://www.gbif.org/species/2889338
- Dryad: Grueter et al. (2022) gorilla ranging data
| Group | Size | Center Lat | Center Lon |
|---|---|---|---|
| Susa | 28 | -1.435 | 29.522 |
| Hirwa | 13 | -1.461 | 29.538 |
| Amahoro | 17 | -1.478 | 29.558 |
| Umubano | 12 | -1.452 | 29.571 |
| Pablo | 24 | -1.490 | 29.542 |
| Kwitonda | 18 | -1.505 | 29.525 |
- Robbins, M.M. et al. (2009). Intergroup encounters among mountain gorillas. Am. J. Primatol.
- Grueter, C.C. et al. (2013). Nest site selection in mountain gorillas. Behav. Ecol. Sociobiol.
- Seaman & Powell (1996). Kernel home range estimation. Ecology.
- IUCN SSC Primate Specialist Group (2023). Mountain Gorilla Status Survey.
Data Technician with research experience at the Dian Fossey Gorilla Fund, building end-to-end computational pipelines across four domains: spatial movement ecology (GeoPandas, KDE, permutation testing), population genetics (CERVUS microsatellite LOD scoring, Queller-Goodnight kinship estimation), machine learning survival analysis (Random Forest, temporal cross-validation), and conservation epidemiology (logistic regression, SciPy hypothesis testing, temporal linkage). Technical stack: Python · R · SQL · scikit-learn · SciPy · GeoPandas · Git. All work is grounded in longitudinal biological datasets with direct conservation policy implications across the Virunga Massif : Rwanda, Uganda, and DRC.