Intermediatepopulations

The Mark-Recapture Method

Overview

Counting every individual in a wild population is usually impossible - animals hide, move, and avoid observers. The mark-recapture method solves this with elegant statistics: capture and mark a sample, release them, then capture a second sample and count how many are marked. The proportion of marked animals in the second sample estimates the proportion marked in the whole population, yielding the Lincoln-Petersen estimator N = (M x C) / R. This lesson explains the method, its assumptions, its extensions, and its wide applications from fisheries to epidemiology.

Learning objectives

  • Describe the capture, mark, release, and recapture procedure.
  • Apply the Lincoln-Petersen estimator N = (M x C) / R.
  • List the method's assumptions and how violations bias estimates.
  • Name applications beyond wildlife, including fisheries and disease studies.

The procedure

First, capture a sample of animals (M), mark each harmlessly - ear tags, leg bands, fin clips, dye spots, or photo-identification of natural markings - and release them back into the population. After allowing time for the marked animals to mix randomly, capture a second sample (C) and count how many carry marks (R, the recaptures). If 50 fish are marked, and a later catch of 40 contains 10 marked fish, the marked fraction in the sample (10/40 = 25%) estimates the marked fraction in the population (50/N), giving N = 200.

Lincoln-Petersen estimator

N = (M x C) / R: population size equals marked animals times second-sample size divided by recaptures. It assumes the marked proportion in the sample matches the population.

Why the math works

The estimator rests on a simple proportion: R/C = M/N. The fraction of marked animals in your second sample estimates the fraction of marked animals in the whole population, because the marked individuals have mixed randomly. Rearranging gives N = (M x C) / R. Intuitively, if marked animals are rare in your recapture sample, the population must be large to have 'diluted' them; if recaptures are common, the population is small. The method is a beautiful example of using a known subpopulation to measure an unknown whole.

Assumptions and their violations

The method assumes: (1) marks are not lost and do not affect survival or catchability; (2) marked animals mix randomly with unmarked ones; (3) the population is closed - no births, deaths, immigration, or emigration between samples; and (4) all individuals are equally catchable. Violations bias results predictably: trap-shy marked animals inflate the estimate, trap-happy ones deflate it, and an open population with births inflates it. Ecologists test assumptions with multiple recapture sessions and use extended models (Jolly-Seber for open populations) when closure fails.

Beyond two samples: modern extensions

Two-sample Lincoln-Petersen is the foundation, but modern studies use many capture sessions analyzed with sophisticated models. The Schnabel method extends the estimator across multiple recaptures; the Jolly-Seber model handles open populations, estimating survival and recruitment alongside abundance; and spatially explicit capture-recapture (SECR) uses trap locations to estimate density and home ranges. Non-invasive 'marks' now include DNA from hair snares and feces, camera-trap photo ID (tiger stripes, whale flukes), and even eDNA - letting researchers census elusive species without ever handling them.

Applications across biology

Fisheries managers use mark-recapture to set quotas for salmon, trout, and marine fish. Conservationists census tigers by camera-trap stripe patterns, whales by fluke photos, and amphibians by toe-clip or photo ID. Epidemiologists adapt the logic to estimate hidden populations - such as the number of unreported disease cases - from overlapping lists. The method even estimates things like the number of software bugs from two independent testers' overlapping finds. Anywhere a complete count is impossible, mark-recapture turns partial sightings into population estimates.

Key points

  • Mark M animals, release, recapture C animals containing R marks; estimate N = (M x C) / R.
  • The logic: the marked fraction in the sample estimates the marked fraction in the population.
  • Key assumptions: marks persist and are neutral, random mixing, closed population, equal catchability.
  • Violations bias estimates predictably - trap-shy marks inflate N, open populations with births inflate N.
  • Extensions (Schnabel, Jolly-Seber, SECR) handle multiple sessions, open populations, and spatial data; non-invasive marks include DNA and photo ID.

Common misconceptions

✗ Myth: Mark-recapture counts every individual.

✓ Truth: It estimates the total from proportions - you never see most individuals. The power of the method is precisely that a complete count is unnecessary.

✗ Myth: More recaptures always mean a bigger population.

✓ Truth: The opposite: many recaptures mean marked animals are a large fraction of the population, so the population is small. Few recaptures mean the marks were diluted in a large population.

✗ Myth: The method only works for animals you can trap.

✓ Truth: Modern 'marks' include DNA, photographs, and acoustic signatures - camera traps census tigers and microphones census whales without any physical capture.

Practice questions

Q1. Researchers mark 60 trout, release them, then catch 50 trout of which 15 are marked. What is the estimated population?
  1. 110
  2. 200
  3. 125
  4. 90
Answer: B.

N = (M x C) / R = (60 x 50) / 15 = 3000/15 = 200 trout. The marked fraction in the sample (15/50 = 30%) implies 60 marked fish are 30% of the population.

Q2. If marked animals become trap-shy after first capture, how is the population estimate biased?
  1. Underestimated
  2. Overestimated
  3. Unaffected
  4. Exactly correct
Answer: B.

Trap-shy marks are recaptured less often, so R is artificially small. Since N = (M x C)/R, a smaller R inflates the estimate - the population appears larger than it is because the marks seem 'diluted'.

Q3. Which situation violates the closed-population assumption?
  1. Marks fall off some animals
  2. A month passes with many births between the two capture sessions
  3. Some animals are trap-happy
  4. The study area is small
Answer: B.

Births between sessions add unmarked individuals, diluting the marked fraction beyond what the original population would produce - the population is not closed. Mark loss violates a different assumption (mark neutrality/persistence), and trap-happy animals violate equal catchability.

Click a question to reveal the answer and explanation.

Frequently asked questions

Do the marks harm the animals?

Modern marking is designed to be harmless: lightweight bands, small fin clips that regrow, temporary dye, or completely non-invasive photo-identification of natural patterns. Ethical review boards approve marking protocols, and researchers prefer methods with no measurable effect on survival or behavior.

What if marked animals learn to avoid traps?

Trap-shy marked animals are underrepresented in the second sample, so R is too small and N = (M x C)/R overestimates the population. Researchers detect this with multiple sessions (recapture rates should stay constant) and switch models or trap types to correct it.

How long should you wait between capture sessions?

Long enough for marked animals to mix randomly through the population, but short enough that the population stays effectively closed - no significant births, deaths, or migration. For small mammals this may be days; for fish, weeks. Longer studies use open-population models instead.

Can mark-recapture estimate density, not just total number?

Yes. Dividing the abundance estimate N by the effectively sampled area gives density. Spatially explicit capture-recapture (SECR) models do this rigorously using the locations of captures to estimate each animal's home range center.