Resumo

DE LEON MATA, Gerardo Daniel; PINEDO ALVAREZ, Alfredo  e  MARTINEZ GUERRERO, José Hugo. Application of remote sensing in the analysis of landscape fragmentation in Cuchillas de la Zarca, Mexico. Invest. Geog [online]. 2014, n.84, pp.42-53. ISSN 2448-7279.  https://doi.org/10.14350/rig.36568.

We analyzed the spatial dynamics and composition of the vegetation cover of the Priority Area for Grasslands Conservation (PAGC) Cuchillas de la Zarca, determining their involvement in the fragmentation processes. A supervised classification and fragmentation analysis in the period from 2008 to 2011 was generated. Spectral analysis allowed to clearly define six classes, where the ecosystem is represented mostly by natural grassland, 41.6% of the total area. The fragmentation analysis of the generated landscape metrics shows a gradual process of fragmentation. An increase in the total number of fragments versus time confirms the fragmentation process in the area, whereas the average size of the patches shows decreasing, indicating that large patches were fragmented or divided. In addition to this, the Shannon and Simpson indicators are trending upward (0.66 to 0.89), (0.68 to 0.82), indicating the increase in time pieces. The overall accuracy of the classifications evaluated in the period 2008 - 2011 shows a high accuracy that ranged from 91 to 94%, while the index Kappa discrete multivariate presents a variation of 0.90 to 0.93.

Palavras-chave : Fragmentation; plant coverage; supervised classification; Landsat TM; kappa; diversity index.

        · resumo em Espanhol     · texto em Espanhol     · Espanhol ( pdf )