Integrated cloud computing and cost effective modelling to delineate the ecological corridors for Spectacled bears (Tremarctos ornatus) in the rural territories of the Peruvian Amazon

Spectacled bears (SB) (Tremarctos ornatus) are the only bear species native to South America. This particular bear is the single species of its genus, and it is listed as vulnerable according to the IUCN red list. A critical SB conservation habitat is in the rural territories of the Peruvian Amazon,...

Full description

Autores:
Cotrina Sanchez, Alexander
Salazar, Andres
Oviedo Sanabria, Carlos Humberto
Bandopadhyay, Subhajit
Mondaca, Pedro
Valentini, Riccardo
Rojas Briceño, Nilton Beltrán
TORRES GUZMÁN, CRISTÓBAL
Oliva, Manuel
Guzman Valqui, Betty Karina
Meza-Mori, Gerson
Tipo de recurso:
Article of journal
Fecha de publicación:
2022
Institución:
UNIVERSIDAD FRANCISCO DE PAULA SANTANDER
Repositorio:
Repositorio Digital UFPS
Idioma:
eng
OAI Identifier:
oai:repositorio.ufps.edu.co:ufps/6884
Acceso en línea:
https://repositorio.ufps.edu.co/handle/ufps/6884
Palabra clave:
Spectacled bears
Amazon
Ecological corridors
Dijkstra’s algorithm
Cloud computing
Latin America
Rights
openAccess
License
https://creativecommons.org/licenses/by-nc-nd/4.0/
Description
Summary:Spectacled bears (SB) (Tremarctos ornatus) are the only bear species native to South America. This particular bear is the single species of its genus, and it is listed as vulnerable according to the IUCN red list. A critical SB conservation habitat is in the rural territories of the Peruvian Amazon, where anthropogenic land-use changes and landscape fragmentation threaten SB habitats. The following questions arise in this context: How much has land-use changed? How to design the establishment of ecological corridors (ECs) to support the conservation of SB?. We investigated the temporal land use and land cover changes for last 30 years (1990–2020) for a better projection of the ECs and to quantify the temporal landscape metrics. Furthermore, we integrated cloud computing, machine learning models with cost-effective techniques to delineate the ECs for SB within the rural territories. Ensemble Random Forest model associated with Google Earth Engine (GEE) was used to develop four land use and land cover (LULC) maps (for the years 1990, 2000, 2010 and 2020). The least cost path (LCP) model based on Dijkstra’s shortest path algorithm was assembled based on six variables (altitude; slope; distance to roads; distance to population centers; land use map; inventory map of SB). Then, we calculated the ECs based on the multidirectional origin-destination points, we found that forest patches increased by 57%