Research Article: Dey, S., Delampady, M., Karanth, K. U., & Gopalaswamy, A. M. (2019). A Spatially Explicit Capture-Recapture Model for Partially Identified Individuals When Trap Detection Rate Is Less than One. Calcutta Statistical Association Bulletin.
Blog Author: Kavya Pandey
Key takeaways:
- In a world of ecology and wildlife research, understanding animal populations is similar to solving a fascinating mystery, which encompasses many intriguing questions: How many animals are there? Where do they go? How do we count them?
- To unravel these mysteries, scientists have developed a special analytical method called the “Spatially Explicit Capture–Recapture (SECR)” model to estimate the number of animals in an area and record their movement within their habitat.
- The model uses tools such as camera traps to track animals and their locations, which helps us learn more about their numbers and where they live in an area.
- A study was conducted in southern India’s Nagarahole National Park and Tiger Reserve to test the efficacy of the SECR model over a period of 50 days.
- 162 camera stations were set up, each with two cameras facing each other, and spaced about 1.5 kilometers apart.
- The study revealed that photographic captures were only 49% effective in capturing tigers, which implies that the existing camera traps are imperfect in estimating actual populations, thus justifying the use of new models in real-world applications.
Understanding wildlife populations is extremely important for conservation efforts, but it can be challenging due to difficulties in data collection, particularly when studying elusive animals such as large carnivores or rare ungulates.
To address this challenge, scientists have developed various methods, including the spatially explicit capture–recapture (SECR) model, which estimates the number and distribution of animals in a specific area more accurately by incorporating the locations of where animals were observed. The SECR model thus has the ability to identify ideal habitat preferences for animals, which is particularly useful in studying endangered species.
While the traditional SECR model was good at counting animals, it could not understand the finer details of animal movement with certainty. Cameras sometimes missed taking pictures of tigers. To mitigate the gap in existing models, a new SECR model was developed that focused on two additional critical factors: (a) how animals find camera traps and (b) how effective these traps are at spotting them. To demonstrate the model’s applicability in the real world, scientists conducted a photographic capture survey of tigers in the Nagarahole National Park and Tiger Reserve of southern India, which is home to a significant tiger population.
162 camera stations were set up, each with two cameras facing each other, and spaced about 1.5 kilometers apart. The results of the study, conducted over 50 days, were fascinating. The camera traps were estimated to be approximately 49% effective, indicating that the cameras missed tigers passing by at certain times. However, the new SECR model could account for these imperfections and provide more accurate population estimates by effectively capturing the patterns and stripes of each individual, making it more efficient in identifying the numbers of males and females in the population of tigers as well.
Scientists also suggest that the advancement of the new SECR model has opened up new avenues for research, not only in the field of wildlife conservation but also in population demographics, where the model can be used to estimate the total human population living in an area, similar to estimating animal numbers. The authors also point out that using a complex model to solve such problems can be challenging, and selecting the right model to predict and describe populations is not easy. Fortunately, researchers are actively developing tools to make this process more manageable, promising a brighter future for the practical application of this approach. With these developments, the authors anticipate further growth and wider use of such innovative methods.
Keywords: Capture-recapture survey, detection probability, hierarchical bayes, SECR mode
