Why Wildlife Surveys Miss Animals & How Scientists Estimate Them
The calculator for this page is switched off while it is checked. The formula is below, and the worked example in the article shows how to use it by hand.
Wildlife population surveys rarely count every animal; they sample a portion of the habitat and use statistical models to estimate the unseen portion. Scientists combine field counts with detection probabilities, distance sampling, and mark-recapture methods to correct for animals that were present but not observed. This article explains why animals are missed and how researchers turn incomplete counts into reliable population estimates.
Formula
The basic correction formula is:
N = n / p
Where:
- N = estimated total population in the survey area (animals)
- n = number of animals actually counted during the survey
- p = detection probability (the proportion of animals that are detected, between 0 and 1)
This simple equation assumes that each animal has the same chance of being detected and that the survey covers the entire area. When only part of the area is surveyed, the estimate is scaled up by the ratio of total area to surveyed area.
Worked Example
Using the default values from the calculator: a 10 km² area, 25 animals counted, and a detection probability of 0.6 (60%).
- Estimate the true population: N = 25 / 0.6 = 41.67, which rounds to 42 animals.
- Calculate density: 42 animals / 10 km² = 4.2 animals per km² (rounded to 2 decimals).
Without the correction, you would have reported only 25 animals, underestimating the true population by about 40%.
Why Animals Are Missed
Animals are missed for many reasons, and understanding these helps scientists design better surveys. Key factors include:
- Camouflage and cryptic coloration – many species blend into their environment, making them hard to spot even when close.
- Behavior – animals may freeze, hide, burrow, or be active only at night, reducing their visibility.
- Dense vegetation – forests, grasslands, and wetlands obscure animals from observers.
- Observer error – fatigue, inexperience, or poor eyesight can cause misses.
- Weather and light – rain, fog, or low light reduce visibility.
- Small population size – rare species are simply hard to find.
- Movement – animals may move into or out of the survey area during the count.
How Detection Probability Is Estimated
Detection probability (p) is rarely known in advance. Scientists use several methods to estimate it:
- Mark-recapture – capture and mark a number of animals, release them, then resample. The proportion of marked animals in the second sample gives an estimate of detection probability.
- Distance sampling – record the distance to each animal seen. Because detection decreases with distance, the pattern of distances allows estimation of p.
- Double-observer surveys – two observers independently count the same area; the proportion of animals seen by both helps estimate p.
- Repeated counts – survey the same area multiple times; the number of new animals seen each time can be modeled to estimate p.
Common Survey Methods
Different methods suit different species and habitats:
- Line transects – observers walk or fly along straight lines, recording distances to animals. Used for large mammals and birds.
- Point counts – observers stand at fixed points and count animals seen or heard within a radius. Common for birds.
- Camera traps – motion-activated cameras capture animals without human presence. Good for elusive or nocturnal species.
- Acoustic surveys – record sounds (e.g., bird songs, whale calls) to detect species. Useful in dense habitats or underwater.
- Dung or track counts – count indirect signs like feces or footprints. Often used for elephants and other large mammals.
Choosing the Right Method
The choice depends on the species, habitat, and budget. For example:
- For a small, secretive rodent in a forest, camera traps or live-trapping with mark-recapture may be best.
- For a wide-ranging carnivore, track counts or camera traps along trails might work.
- For a colonial seabird, aerial photographs of nesting colonies can give a near-complete count.
- For a nocturnal amphibian, acoustic surveys at night are effective.
Scientists often combine methods to improve accuracy. For instance, using distance sampling alongside camera traps can provide both density and detection probability.
Sources of Error and Uncertainty
Even with careful methods, estimates have uncertainty. Common problems include:
- Non-random distribution – animals are often clumped, so a simple average may be misleading.
- Misidentification – similar-looking species can be confused.
- Heterogeneity in detection – some individuals are easier to detect than others (e.g., males vs. females, adults vs. juveniles).
- Edge effects – animals near the boundary of the survey area may be double-counted or missed.
Scientists report confidence intervals (e.g., “the population is estimated at 1,200 ± 150”) to convey this uncertainty. They also use models that incorporate covariates like habitat type, weather, and observer experience.
Frequently Asked Questions
Why can’t we just count every animal?
In most habitats, it is physically impossible to see every animal. Dense vegetation, nocturnal behavior, and small body size mean many animals are hidden. Even in open habitats, animals may move or be overlooked. Counting every individual would require an enormous effort and still be incomplete.
How do scientists know the detection probability?
They estimate it using methods like mark-recapture or distance sampling. For example, in mark-recapture, if you capture 50 animals, mark them, and later capture 100 animals of which 25 are marked, you estimate that 25% of the population is marked, so the total population is 50 / 0.25 = 200. The detection probability is the proportion of marked animals in the second sample (0.25).
What is the difference between abundance and density?
Abundance is the total number of animals in an area (e.g., 500 deer). Density is the number per unit area (e.g., 5 deer per km²). Density is useful for comparing different areas or times, while abundance gives the total population size.
Are camera traps more accurate than human observers?
Camera traps can be more accurate for nocturnal or elusive species because they operate continuously without disturbing animals. However, they have their own biases: animals may avoid the camera, or multiple individuals may trigger the same photo. Human observers can cover larger areas and identify species better in some cases. Neither is universally better; the choice depends on the species and question.
Why do wildlife estimates change so much between years?
Populations naturally fluctuate due to births, deaths, migration, and environmental conditions. Additionally, survey methods may change, or the area surveyed may differ. It’s important to compare estimates using the same methods and account for uncertainty before concluding a real change.