Capture-Mark-Recapture (CMR) Studies

Credit: Øystein Wiig

Much of the available scientific information about polar bears comes from capture-mark-recapture (CMR) studies in which wild polar bears are temporarily sedated, after which scientific data are collected and the animals are released. Physically capturing bears allows them to be individually marked so that they can be reidentified later. It also allows scientists to collect biological samples (e.g., blood or fat samples), take physical measurements such as length, weight, and body condition (i.e., fatness), and apply satellite telemetry devices such as GPS radiocollars that can provide information on where the bear goes and what it does for the next year. Perhaps most importantly, capture and marking efforts that are repeated regularly over multiyear periods allow scientists to estimate vital rates such as reproductive success and survival probability, as well as population size. This information is critically important to polar bear management and conservation. Although new methods are constantly being developed (see below), CMR studies have provided most of the population assessments that we currently have for polar bears worldwide. 

How CMR Studies Work

What is the logic behind CMR studies? Imagine that there is a jar full of white marbles (polar bears). You want to know how many marbles are in the jar (population size) but don’t have the time, money, or ability to empty out the jar and count every marble. So—you reach in, grab a handful of marbles (a capture sample), paint the marbles you grabbed black (mark them), and then put them back in the jar and mix everything up. A few minutes later, you grab a second handful (a recapture sample). The key concept is that there is information in the ratio of white to black marbles in this second handful. Imagine that there is just one black marble. Given that the jar is sealed off and the marbles are not reproducing (a closed population), the most likely explanation is that it’s a huge jar with very many white marbles (i.e., a large population), such that any random sample is expected to consist mostly of white marbles with only one or two blacks. Do a little math based on how many marbles were in each handful and the ratio of white to black marbles, and—congratulations—you have completed a closed-population CMR study to estimate population size.

But wait, could the small number of white marbles in the second handful have other explanations? The answer is yes, if we drop the assumption of having a closed population and imagine that, like polar bears, marbles in the jar can live and die and reproduce. Under these conditions, the small number of black marbles in the second handful could mean that most of the white marbles that you painted black and put back in the jar subsequently died. This time, in contrast with the previous example, the small number of black marbles in your second handful was not because the jar is huge. Rather, this time the patterns in the data arose because most of the black marbles died (low survival) and were simply not available to be resampled. Take several more handfuls, record the ratios of white to black marbles in each, do some more math and—congratulations again—you have completed an open-population CMR study to estimate survival.

Application of CMR Studies to Polar Bears

The examples above are meant to introduce CMR studies in their simplest form. Real-world studies are more complicated. Polar bears are individualistic and intelligent animals that do more than live, die, and reproduce. For example, being the most mobile four-legged animals in the world means that polar bears will often immigrate and emigrate from a study population during a multiyear CMR study. And, because the Arctic is big and remote, there are places that the bears can go but humans can’t follow. This can make it difficult to determine whether the “missing bears” in a CMR study (i.e., those that were captured once but never showed up again) died or simply went somewhere else. Polar bears also have extended maternal care, meaning that adult females keep their cubs with them for several years and the fates of family group members are intertwined (e.g., if mom dies, the cubs are likely to die too). These and other challenges have led to rapid advances in wildlife biology, to the point where some scientists spend their entire careers developing and implementing CMR studies to obtain accurate estimates of vital rates and population size. Excellent reviews of the evolution of CMR methods are found in two books, Williams et al. (2002) and Amstrup et al. (2005), which were used to design and implement CMR studies on polar bears for decades.

Challenges and Future Directions

Many of today’s CMR studies look different than they did 25 years ago. A key development has been that new genetic methods allow animals to be individually identified based on a tissue sample that contains its DNA. These “genetic CMR studies” don’t have to physically capture bears at all. Rather, scientists use a special dart that, when fired at a bear’s rump, collects a small piece of fur and skin, and then bounces off. When genetic CMR sampling is performed regularly over a multiyear period, it provides data on when and where individual bears were “captured” (in this case, biopsy darted) that can be used to estimate vital rates and population size, very much like the data from physical CMR studies as discussed above. The downside of genetic CMR studies is that they only provide a small tissue sample and don’t allow scientists to collect physical measurements or apply satellite telemetry devices. The upside, however, is that genetic CMR studies are less invasive. In other words, they are less stressful and uncomfortable because the bears are not chemically immobilized (i.e., sedated, which is done with the same drugs that veterinarians use on dogs and cats) or handled by humans like during physical captures. The details of genetic CMR studies can be found in Atkinson et al. (2021), the first large-scale application of these methods to polar bears that provided estimates of reproduction, survival, and population size for polar bears in the Baffin Bay subpopulation that is shared between Canada and Greenland.

Part of the motivation to switch to genetic CMR methods for polar bears has been opposition to physical capture and the application of radiocollars by some Inuit groups in the U.S. and Canada, who oppose hands-on study methods for ethical reasons. Scientists also are concerned about animal welfare. All polar bear studies—whether physical or genetic CMR or using other methods entirely—are reviewed by Animal Care and Use Committees to ensure that the proposed study minimizes stress and risk to the animals while maximizing the amount of scientific information collected. This has led to numerous technological advances, such as electronic devices that are attached to radiocollars and programmed to release them after a specified period, for example 12 months. Although several scientific studies have found no evidence for long-term negative effects of capture and handling on polar bears, physical captures inevitably involve a degree of stress and have been shown to influence bears’ behaviour for 48 hours after immobilization, causing them to spend more time sleeping and recovering, and less time hunting. Details from a recent study of polar bear behaviour following capture can be found in Stirling et al. (2022).  

Capture-mark-recapture studies will remain an important tool for collecting scientific information about polar bears to estimate vital rates and population size, and to understand the effects of climate warming on the species. In any given situation, dialog among wildlife managers, scientists, communities, hunters, and Inuit groups will be necessary to identify study methods that achieve the right balance—reducing invasiveness to animals, while still providing the information needed for management and conservation. These topics are discussed thoroughly in Laidre et al. (2022), which reviews the use of satellite telemetry for polar bears and outlines the pros and cons of such methods in the 21st century. This is an important topic because, as discussed above, polar bears are highly mobile and their movements in and out of a study population can lead to inaccurate estimates of survival and population size if satellite telemetry data are not available. This issue is becoming more challenging as sea-ice loss due to climate warming affects the movements and distribution of polar bears, resulting in bears no longer accessing some of their historical habitats and, in some cases, moving into new areas. Another consideration is that there have been major improvements in the statistical models used to analyse data from CMR studies. Instead of relying solely on patterns in where and when individual bears were encountered, today’s CMR models can incorporate satellite telemetry data, information from subsistence-harvested bears, information about how the environment is changing (e.g., trends in sea-ice availability), and even some aspects of Indigenous Knowledge. Combining multiple data types is one of the best ways to improve estimates of vital rates and population size. The advent of “integrated population models” (IPMs)—flexible and customizable CMR models that take advantage of all available information—has made it easier to combine multiple data types, resulting in better estimates of vital rates and population size than were previously available. For more details about IPMs for polar bears, see Regehr et al. (2018), which provided the first estimate of population size for the Chukchi Sea subpopulation that is shared between the US and Russia.

Further Reading

  • Amstrup SC, McDonald TL, and Manly BFJ. (eds). 2005. Handbook of Capture-Recapture Analysis. Princeton University Press, Princeton, New Jersey, 296 pp.
  • Atkinson SN, Laidre KL, Arnold TW, Stapleton S, Regehr EV, Born EW, Wiig Ø, Dyck M, Lunn NJ, Stern HL, and Paetkau D. 2021. A novel mark-recapture-recovery survey using genetic sampling for polar bears Ursus maritimus in Baffin Bay. Endangered Species Research 46: 105-120.
  • Laidre KL, Durner GM, Lunn, NJ, Regehr EV, Atwood TC, Rode KD, Aars J, Routti H, Wiig Ø, Dyck M, Richardson ES, Atkinson S, Belikov S, and Stirling I. 2022. The role of satellite telemetry data in 21st century conservation of polar bears (Ursus maritimus). Frontiers in Marine Science 9.816666,
  • Regehr EV, Hostetter NJ, Wilson RR, Rode KD, Martin MS, and Converse SJ. 2018. Integrated population modeling provides the first empirical estimates of vital rates and abundance for polar bears in the Chukchi Sea. Scientific Reports 8: 16780.
  • Stirling I, Regehr EV, Spencer C, Burns LE, and Laidre KL. 2022. Using visual observations to compare the behavior of previously immobilized and non-immobilized wild polar bears. Arctic 75: 398-412.
  • Williams BK, Conroy MJ, and Nichols JD. (eds). 2002. Analysis and Management of Animal Populations. Academic Press, New York, New York, 817 pp.