Wildlife Population Estimation

Wildlife population estimation uses field observations and statistical models to infer how many animals occupy an area. For elusive carnivores, a simple count is unreliable because many animals remain undetected; methods must therefore estimate detection probability and distinguish abundance or density from mere occurrence.

Survey design and evidence

Researchers divide the landscape into sampling units and collect standardized detections through camera traps, tracks, scats, hair, direct sightings or telemetry. Sampling must cover the target habitat, operate long enough to obtain repeat detections and satisfy the assumptions of the chosen model.

• Repeated surveys help separate true absence from non-detection.
• Closed-population models generally assume no births, deaths, immigration or emigration during the short sampling period.

Methods based on individual identification

When individuals can be recognized from coat patterns, scars, tags or DNA, their repeated detections provide capture histories.

• Camera-trap capture-recapture estimates abundance from the proportions of individuals detected once and repeatedly.
• Spatially explicit capture-recapture combines individual capture histories with detector locations to estimate density while accounting for animal movement and spatial detection.
• Non-invasive genetic sampling identifies individuals from scat or hair and applies capture-recapture models without physically capturing them.
• Physical capture, marking and subsequent recapture or resighting can also be used, subject to ethical safeguards and adequate recaptures.

Methods for unmarked or rarely detected species

Where reliable individual identification is impossible, researchers use models based on repeated presence, encounter rates or direct observations.

• Occupancy models estimate the probability that sampling sites are used while correcting for imperfect detection; occupancy is not a population count.
• Distance sampling can estimate density where sufficient direct sightings are available and detection declines with distance from a transect or point.
• Track, scat and camera encounter rates provide relative-abundance indices, but they should not be treated as absolute density unless independently calibrated.
• Camera-based random-encounter models may estimate density of unmarked animals when movement, camera coverage and detection assumptions are adequately measured.
• Radio or GPS telemetry supplies information on movement and home-range size and can strengthen survey design or density estimation.

Choosing and interpreting estimates

Method choice depends on whether individuals are identifiable, expected detection rates, population closure, terrain, cost and survey objective. Combining camera traps, genetic evidence and sign surveys can improve confidence, while uncertainty intervals and model assumptions should always accompany estimates.