Field Survey Methods: The Science of Counting Wildlife in the Wild

How many elephants are there in the Serengeti? How many tigers remain in the wild? How fast are wolf populations recovering in Europe? These seem like simple questions — and they are of fundamental importance for conservation decision-making — but answering them with scientific rigour is among the most challenging methodological problems in ecology. Animals do not sit still to be counted; many are rare, cryptic, or nocturnal; they occur in habitats that are difficult or dangerous to survey; and the spatial and temporal scales over which populations must be estimated often exceed the practical capacity of field survey teams. The development and refinement of wildlife survey methods over the past century represents one of the most practically important advances in conservation science.

Distance Sampling and Transects

Distance sampling — the method of recording the perpendicular distance from a survey transect to each detected animal, and using the resulting detection function to estimate density — is one of the most widely used methods in wildlife population assessment. The fundamental principle is that detection probability decreases with distance from the transect: animals very close to the transect are almost always detected, while animals far away are frequently missed. By modelling this decline in detection probability, distance sampling allows ecologists to estimate the proportion of animals that were missed during the survey and calculate the true density of animals in the survey area. Distance sampling can be conducted on foot, from vehicles, from aircraft, or at sea from ships — each with specific protocol requirements for transect design, animal detection recording, and statistical analysis.

Mark-Recapture Methods

Mark-recapture — the technique of capturing animals, marking them individually, releasing them, and then using the proportion of marked individuals in subsequent captures to estimate population size — was one of the first rigorous statistical methods in population ecology. The fundamental Lincoln-Petersen estimator relates population size to the number of individuals marked, the total caught in the second sample, and the number of those already marked. Modern mark-recapture methods have evolved into sophisticated statistical frameworks (Cormack-Jolly-Seber, POPAN, robust design) that can simultaneously estimate population size, survival probability, recruitment rate, and movement between sampling sites across multiple sampling occasions.

Transect Methods — Standardising Wildlife Counts

Line transect distance sampling — the standard method for estimating wildlife population densities in terrestrial and marine environments — combines field survey effort with statistical modelling to convert observations of animals at known distances from a survey line into density estimates for the wider area. The fundamental assumption is that all animals on the survey line are detected with certainty, while detection probability declines with increasing distance — a function that can be estimated from the distribution of perpendicular distances of detected animals from the survey line. By fitting mathematical models to this distance distribution and estimating the probability of detection at each distance, analysts can correct the observed count for the animals missed at greater distances and calculate an unbiased density estimate. The method has been applied to populations ranging from African elephants (surveyed from aircraft along systematic transects) to Antarctic krill (surveyed acoustically from ships) to tropical forest primates (surveyed by walking and listening teams), providing standardised density estimates that can be compared across surveys separated in space and time to track population trends.

The deployment of systematic monitoring programs — repeating standardised surveys at fixed locations across multiple years or decades — transforms snapshot biodiversity surveys into trend data that can detect population changes, range shifts, and phenological changes with statistical power that improves as the time series lengthens, making long-term monitoring commitment one of the highest-value investments in ecological science.

Distance Sampling — Statistical Rigour in Wildlife Counting

Wildlife population estimates — essential for conservation planning, hunting management, and ecological research — require statistical methods that account for the fundamental challenge of detectability: not all animals in a surveyed area are detected by observers. Distance sampling — the family of methods that use the relationship between the distance of an animal from a transect line and its probability of detection to estimate true density from observed counts — has become the gold standard for estimating wildlife populations in terrestrial and marine systems. The key insight of distance sampling is that animals close to the transect line are nearly always detected, while detection probability falls off with distance in a way that can be modelled statistically. By recording the perpendicular distance of each detected animal from the transect, fitting a detection function to these distances, and applying the fitted function to correct the raw counts for imperfect detection, unbiased estimates of density can be obtained even when a large fraction of animals is missed. Distance sampling has been applied to estimate populations of everything from blue whales in the Pacific to snow leopards in the Himalayas, and its statistical principles — accounting for imperfect detection — underpin most modern wildlife monitoring protocols.