Multivariate Modelling for the Unexplained Relationships of tree Variables for Combretum hartmannianum

Tree crown information contributes to the investigation of several key forest ecosystem attributes: biodiversity, productivity, sustainability, aesthetics, forest environment and wildlife. Recent remote sensing applications utilize estimation of either canopy cover or individual tree canopy area as a first stage towards timber volume. Measurement of CR size for each tree can be time-consuming and difficult to obtain in very dense stands and for very tall trees where the base of live crown is obscured. Studies have often shown that abiotic factors such as soil and topographic characteristics play a major role for species distribution. Moreover, space has been shown to be more important than environment in determining woody plant species distribution. However, trees show considerable variation and flexibility in their size of crowns, height and stem diameters. The tree stem size has its own adaptive significance to a tree. It must be strong enough to withstand the forces that act on it (weight of the tree and the drag exerted on it by the wind). The sizes, shapes and relative locations of crowns both determine and respond to the shading and constriction effects that characterize aboveground interactions between trees. The more important issue is that, whether stem diameter-at-breast height relationships with tree height, crown height, and crown diameter are the only factors that can be used to model growth and yield of trees, especially in case of natural forests and or mixed plantations with different species and sizes. This study investigates the potential for structural indices and other spatial measures to improve the prediction of crown radius and crown length for trees in natural woodlands. Field data were collected for Cambretum hartmannianum tree species from Elgarri natural forest reserve in Blue Nile State, Sudan, to test the performance of crown radius and crown length (height) models on the basis of allometric relationships with stem dimensions. A number of spatial and non-spatial variables were incorporated into modified crown dimension models. Coefficient of determination (R2) and relative bias were used to test the performance of the inclusion of spatial or non-spatial variables into the original models (using tree dimensions alone). Predictions of crown length and radius were found to be better after the incorporation of the spatial and non-spatial. Our results also confirm that not all estimation models are suitable for all forest, but subject to the local forest condition and composition. In this regard and for the studied tree species in the selected forest, Mi index is recommended for the estimation of crown radius, and DCL index for crown length.