Connectivity modelling in conservation science: a comparative evaluation
Landscape connectivity, the extent to which a landscape facilitates the fow of ecological processes such as organism movement, has grown to become a central focus of applied ecology and conservation science. Several computational algorithms have been developed to understand and map connectivity, and...
| Main Authors: | , |
|---|---|
| Format: | Journal article |
| Language: | English |
| Published: |
Springer Nature
2022
|
| _version_ | 1797108420100227072 |
|---|---|
| author | Unnithan Kumar, S Cushman, SA |
| author_facet | Unnithan Kumar, S Cushman, SA |
| author_sort | Unnithan Kumar, S |
| collection | OXFORD |
| description | Landscape connectivity, the extent to which a landscape facilitates the fow of ecological processes such as organism movement, has grown to become a central focus of applied ecology and conservation science. Several computational algorithms have been developed to understand and map connectivity, and many studies have validated their predictions using empirical data. Yet at present, there is no published comparative analysis which uses a comprehensive simulation framework to measure the accuracy and performance of the dominant methods in connectivity modelling. Given the widespread usage of such models in spatial ecology and conservation science, a thorough evaluation of their predictive abilities using simulation techniques is essential for guiding their appropriate and efective application across diferent contexts. In this paper, we address this by using the individualbased movement model Pathwalker to simulate diferent connectivity scenarios generated from a wide range of possible movement behaviours and spatial complexities. With this simulated data, we test the predictive abilities of three major connectivity models: factorial least-cost paths, resistant kernels, and Circuitscape. Our study shows the latter two of these three models to consistently perform most accurately in nearly all cases, with their abilities varying substantially in diferent contexts. For the majority of conservation applications, we infer resistant kernels to be the most appropriate model, except for when the movement is strongly directed towards a known location. We conclude this paper with a review and interdisciplinary discussion of the current limitations and possible future developments of connectivity modelling. |
| first_indexed | 2024-03-07T07:27:29Z |
| format | Journal article |
| id | oxford-uuid:7de46185-75f2-4768-a7d5-e624dfe92e0e |
| institution | University of Oxford |
| language | English |
| last_indexed | 2024-03-07T07:27:29Z |
| publishDate | 2022 |
| publisher | Springer Nature |
| record_format | dspace |
| spelling | oxford-uuid:7de46185-75f2-4768-a7d5-e624dfe92e0e2022-12-06T13:29:30ZConnectivity modelling in conservation science: a comparative evaluationJournal articlehttp://purl.org/coar/resource_type/c_dcae04bcuuid:7de46185-75f2-4768-a7d5-e624dfe92e0eEnglishSymplectic ElementsSpringer Nature2022Unnithan Kumar, SCushman, SALandscape connectivity, the extent to which a landscape facilitates the fow of ecological processes such as organism movement, has grown to become a central focus of applied ecology and conservation science. Several computational algorithms have been developed to understand and map connectivity, and many studies have validated their predictions using empirical data. Yet at present, there is no published comparative analysis which uses a comprehensive simulation framework to measure the accuracy and performance of the dominant methods in connectivity modelling. Given the widespread usage of such models in spatial ecology and conservation science, a thorough evaluation of their predictive abilities using simulation techniques is essential for guiding their appropriate and efective application across diferent contexts. In this paper, we address this by using the individualbased movement model Pathwalker to simulate diferent connectivity scenarios generated from a wide range of possible movement behaviours and spatial complexities. With this simulated data, we test the predictive abilities of three major connectivity models: factorial least-cost paths, resistant kernels, and Circuitscape. Our study shows the latter two of these three models to consistently perform most accurately in nearly all cases, with their abilities varying substantially in diferent contexts. For the majority of conservation applications, we infer resistant kernels to be the most appropriate model, except for when the movement is strongly directed towards a known location. We conclude this paper with a review and interdisciplinary discussion of the current limitations and possible future developments of connectivity modelling. |
| spellingShingle | Unnithan Kumar, S Cushman, SA Connectivity modelling in conservation science: a comparative evaluation |
| title | Connectivity modelling in conservation science: a comparative evaluation |
| title_full | Connectivity modelling in conservation science: a comparative evaluation |
| title_fullStr | Connectivity modelling in conservation science: a comparative evaluation |
| title_full_unstemmed | Connectivity modelling in conservation science: a comparative evaluation |
| title_short | Connectivity modelling in conservation science: a comparative evaluation |
| title_sort | connectivity modelling in conservation science a comparative evaluation |
| work_keys_str_mv | AT unnithankumars connectivitymodellinginconservationscienceacomparativeevaluation AT cushmansa connectivitymodellinginconservationscienceacomparativeevaluation |