Componentes del paisaje como predictores de cubiertas de vegetación: estudio de caso del estado de Michoacán, México

The State of Michoacan, as well as a handful of other hot spot regions worldwide, harbors an outstanding overlap between natural and cultural richness as a result of intermingle climatic geological, geographical and ecological conditions. Presently both, natural and cultural heritages are jeopardized at most hot spots worldwide and policy makers seek urgently for robust base line information to restore and eventually reorient development. Spatially explicit base line data bases have been recognized as critical in order to facilitate design and implementation of public policies. In this sense, dynamics of native land cover/vegetation patterns (natural units) have helped in enormously to provide base line information and predict outcomes. A rather critical issue relies in developing replicable and robust methods to predict natural units, as a direct response of climatic, geological and geomorphological data (physical units). Often natural units are used as spatial criteria to delineate physical units, yet these latter are regarded as the responsible variables for delimiting natural units. This tautological thought has been largely neglected in most scientific literature in Mexico when constructing cartographic outputs.

In this article, we argue that native land cover/vegetation patterns are the response variable of physical attributes at a meso-scale level. Hence, the objective of this investigation was to predict native land cover/vegetation patterns based upon climatic, geological, geomorphological and ecological attributes hierarchically intermingled. The contribution was made in order to developing a robust and replicable method accordingly to current available information worldwide. The research took place in Michoacan state as it is regarded a typical hotspot comprising geo-ecological complex features. To illustrate this further, Michoacan harbors over 800 tree species which overpasses the number of species of all Western Europe. Prediction modeling was with the aid of a geographic information system. Decisions rules were based upon Boolean logic giving special attention to emerging bioclimatic zoning techniques. The later consists in providing gradients of temperature and precipitation along seasonal threshold values so that sound matching between physical and natural units is found. In addition, an innovative aspect regards the cartographic expression of these gradients of temperature and precipitation here referred as termotype and ombrotypes respectively.

Outcomes demonstrated that prediction of native land cover/vegetation patterns was feasible within a geo-ecologically complex region as Michoacan. In addition, dissecting attributes of the landscape hierarchically organized proved to be a robust and replicable method to reconstruct native vegetation patterns at places currently covered by anthropogenic activities. Tropical seasonally dry forests covers most surface still covers most surface (28.52% of the State), whereas, temperate forests cover the second most abundant types (27.71% of the State). Aquatic vegetation (0.22% of the State), and Xerophytic scrubland (0.08% of the State), are currently the least represented types. The prediction modeling indicated that tropical seasonally dry forests has been depleted in 18.68% of the State, while temperate forests in 14.98% of the State.

Ecotones delimiting tropical seasonally dry and temperate forests are under drastic threat because two mayor productive systems are gaining global importance. Avocado and Mango produced at these regions are expanding their ranges as a result of global demands. In consequence our current findings may serve to reorient policy makers in order to find tradeoffs and thresholds to conciliate encroachment of productive systems and maintenance of environmental services provided by native land cover/vegetation patterns.

To conclude, native land cover/vegetation patterns were predicted as a response variable of physical attributes so that tautological thinking was avoided. The method developed may be applied to other hot spots provided that physical information is available. A practical outcome regard the obtained land cover/vegetation map, which may serve as baseline to predict future scenarios in the light of current man-made and climatic changes foreseen.