Áreas prioritarias de conservación para la flora leñosa del estado de Colima, México

The diversity of the woody flora (trees and lianas) of the state of Colima was quantified and prioritized using principles of systematic conservation planning (gradsects, indicators groups and four heuristic algorithms for weighting the species richness, the rarity value, and besides the number and density of species endemic to Mexico). A total of 53 environmental units (EU) were defined with information of geomorphology, altitude and vegetation types. Four gradsects embraced most of the EU diversity (91.6%). Within gradsects 15 EU were chosen with the largest area and accessible roads to locate 24 sampling sites of 0.1 ha, where all trees and lianas ≥ 2.5 cm dap were identified. We counted 4419 individuals, including 417 species and 71 families. Species richness among sites contrasted markedly (between 82 and 5 species). Site species richness significantly correlated with the species endemic to Mexico (r = 0.85, P < 0.001). To preserve at least 75% of the species, five sites were always selected in the top six using the four iterative methods. When this analysis was repeated considering the species recorded in two reserves (Biosphere Reserve of Manantlan and Forest Protection and Wildlife Refuge Area El Jabali), both areas played a key role, allowing the protection of 67.4% of the total species and 72.5% of the endemics. Four sites were the most complementary if compared with the flora of the reserves. Since the EU where these sites are located are not connected, we suggest that three nearby units be linked through a system of vegetation corridors, which could expand toward the southern portion of the Manantlan reserve. It is recommended that future conservation proposals consider types of vegetation that are valued for possesing other growth forms (e. g., the herbaceous understory of  Pinus-Quercus  forest) or for the ecosystematic functions they provide (e. g., mangroves). To reach the goal of a representative network of biodiversity areas in Colima, similar data for other groups of plants and animals must be generated.