- Development and Application of a Classification System for Undisturbed and Disturbed Tropical Montane Forests Based on Vegetation Structure (2002)
- Classification of the vegetation was a central part of a DFG-financed analysis of undisturbed and disturbed montane rain forest ecosystems in Southern Ecuador. The aim of the study presented here was the development and application of a classification system based on structural characters for both, undisturbed and disturbed forests. In an altitude between 1570 m a.s.l. and 3100 m a.s.l. 139 plots of 20 m x 20 m were established and the abundance of 151 structural parameters was quantified per tree stratum. Cluster analysis allowed grouping of the 139 plots into 14 groups, called structural forest types. These types were: Secondary growth after road construction, Secondary growth after landslide, Plantation of Pinus patula, Forest on edges of natural gaps, Secondary ravine forest, Ravine forest under human influence, Primary ravine forest at lower altitude, Primary ravine forest at higher altitude, Microphyll ridge forest, Mesophyll ridge forest, Macrophyll ridge forest, and megaphyll ridge forest. Two forest types were investigated outside the closer investigation area and were classified as Secondary forest in Mindo and Primary forest in Cajanuma. Primary forest types were only found south of the Rio San Francisco, whereas on the northern side in a matrix of pastures and farmland only plan-tations and secondary vegetation along ravines could be detected. In the area of primary forest, a primary ravine forest type of lower altitudes turned into a type of higher altitudes around 2100 m a.s.l. in ravines. Both types were characterised by comparably great tree heights and wide stem diameters. On ridges, three types of ridge forest coexisted in the lower part until microphyll ridge forest dominated in altitudes above 2200 m a.s.l.. Timberline was reached at 2650 m a.s.l. on the eastern side of the cordillera, at approximately 3000 m a.s.l. on the western side in Cajanuma. Decreasing canopy height on ridges could be explained by water stress in wind exposed sites and unfavourable N-mineralisation conditions (due to decreasing pH, profile depth, raising C/N-values, and water stagnation with increasing altitude). Secondary forest types were characterised by human influence (felling of emergent trees), pioneer vegetation and high abundance of connecting elements like lianas and bamboo. The structural forest types could be integrated into accepted schemes of altitudinal zonation of the Andean vegetation without contradiction, but allowing a further distinction of the zone under investigation. Frequency analysis and calculation of correlation between variables allowed the reduction of the catalogue from 151 to 104 structural parameters by excluding variables that did not contribute to the clustering. This exclusion did not alter the result of the cluster analysis. On the basis of the reduced catalogue a key to the 14 structural forest types was developed. This key was applied to 100 plots of 10 m x 10 m in a parallel study of other project participants investigating the distribution of hummingbird-visited plants. Distribution patterns of this functional group could be related to the structural forest types. Recent investigations of bird and moth communities will also be tested for functional relationships.