Counting and Monitoring Wildlife

“How many are there, and is the number going up or down?” These are the two most common questions in wildlife management. The answers decide whether a species is listed as threatened, whether hunting is allowed, whether a reserve needs more rangers, and whether a conservation project has worked. This module introduces how wildlife biologists count animals and design monitoring that can detect real change.

What you will learn

• The difference between abundance, density, occupancy and indices
• The main survey methods and when to use each
• Why detection probability matters
• How to design a monitoring programme that can detect trends

What are we measuring?

Total counts are rarely possible for wild animals, so most methods estimate numbers from a sample and account for animals that were missed.

Survey methods

Total counts

Possible for highly visible animals in open habitats or at concentrated sites, such as seabird colonies, seal haul-outs, flamingo flocks or waterbird counts at wetlands. The annual International Waterbird Census, running since 1967, counts waterbirds at thousands of sites worldwide.

Aerial surveys

Observers in light aircraft, or increasingly drones, count large animals along strips. Aerial surveys underpin elephant and large herbivore counts across Africa, including the Great Elephant Census of 2014–2016, which counted elephants across 18 countries and found a 30 percent decline in savanna elephants between 2007 and 2014.

Distance sampling

Observers walk or drive transects, recording each animal seen and its distance from the line. Because animals further away are harder to see, a detection function is fitted to estimate how many were missed. Distance sampling is widely used for primates, antelope, birds and marine mammals.

Capture–recapture

Individuals are “captured” and identified, then recaptured later. The proportion of known individuals in later samples indicates population size. Captures do not need to be physical: individuals can be identified from camera-trap photos (tigers, leopards, jaguars, snow leopards), DNA from hair or dung (bears, wolves), photographs of natural markings (whale tail flukes, whale shark spots, giraffe coat patterns) or calls. Modern spatial capture–recapture models use where each individual was detected to estimate density directly.

Occupancy surveys

When individuals cannot be told apart, biologists survey many sites repeatedly and record whether the species is detected. Occupancy models, developed by Darryl MacKenzie and colleagues in 2002, estimate the true proportion of occupied sites while correcting for missed detections. Occupancy is cheaper than abundance and is widely used for amphibians, carnivores, birds and rare species.

Indices and signs

Counts of dung, tracks, nests (for great apes and orangutans), calls or roadkill can indicate relative abundance. They are cheap but must be interpreted carefully, because detection and sign decay rates can change with season and habitat.

Acoustic monitoring and eDNA

Autonomous recorders and eDNA (Ecology Module 15) are increasingly used to detect species presence across large areas.

Why detection matters

Suppose you survey a pond five times and hear frogs on only two visits. Are frogs rare, or just hard to detect in cold weather? Almost every survey misses some animals, and the proportion missed can change with observer, weather, vegetation and animal behaviour. If detection falls over time because vegetation grows taller, a raw count may show a decline that is not real. Modern survey methods therefore estimate detection probability and correct for it.

Designing a monitoring programme

1. Define the question and the target. For example: “Detect a 30 percent decline in leopard density over ten years with 80 percent confidence.”
2. Choose the method that matches the species, habitat and budget.
3. Choose sites at random or systematically across the area of interest, not just where animals are easy to see.
4. Standardise methods, timing, season and effort, and record conditions.
5. Run a power analysis to check whether your sample size can detect the change you care about.
6. Keep going. Long-term monitoring is valuable precisely because it is long. Many of the most important ecological insights come from datasets spanning decades.
7. Store and share data in standard formats, for example through GBIF.

Case study: counting snow leopards

Snow leopards are notoriously hard to count. For years, global estimates were little more than informed guesses. The Population Assessment of the World’s Snow Leopards (PAWS) initiative, launched in 2017 by the 12 range countries and partners, set out to use standardised camera-trap and genetic surveys with spatial capture–recapture analysis. Mongolia completed its first national snow leopard survey in 2021, estimating around 950 animals, and India’s first national assessment, published in 2024, estimated 718. These surveys give baseline figures against which future change can be measured.

Try it yourself

Scatter 100 dried beans in a garden or grass patch, ask a friend to “survey” by collecting as many as they can find in two minutes, then count how many were found. What was their detection probability? Repeat in taller grass. How would this affect a real survey?

Key terms

• Abundance and density: total numbers and numbers per unit area.
• Occupancy: the proportion of sites occupied by a species.
• Detection probability: the chance that an animal present is recorded.
• Distance sampling: estimating density from sighting distances.
• Power analysis: estimating the sample size needed to detect a change.

Quick quiz

Further reading

• Buckland, S. T. et al. Distance Sampling: Methods and Applications.
• MacKenzie, D. I. et al. Occupancy Estimation and Modeling.