University of Khartoum

Evaluation of Selected Stem Taper Models for Some Hardwood Tree Species Growing In Central Sudan

Evaluation of Selected Stem Taper Models for Some Hardwood Tree Species Growing In Central Sudan

Show full item record

Title: Evaluation of Selected Stem Taper Models for Some Hardwood Tree Species Growing In Central Sudan
Author: Mohammad, Nedal
Abstract: The main objective of this study was to evaluate the performance of selected taper models in describing stem profile of some hardwood tree species growing in Central Sudan. The study was divided into three parts; (1) evaluation of fitting nine stem taper models (representing four taper-modeling approaches) to eight tree species, namely, Acacia nilotica L. (Sunt), Acacia seyal var. seyal Del. (Talih), Combretum hartmannianum Schweinf. (Habil El Gebel), Eucalyptus camaldulensis Dehnh. (Ban or Kafour), Lannea fruticosa Hochst. (Leyun), Sclerocarya birrea Hochst (Humeid), Sterculia setigera Del. (Tartar) and Terminalia laxiflora Engl. (Darout), (2) validation of the three superior models for Sunt trees of different ages and growing on different sites, and (3) evaluation the three superior models for Eucalypts trees of different ages and from different origins. Data used in the first part of this study came from 240 trees (30 trees for each species) and were composed of 2400 data points. For the second part, the data came from 180 trees of 16 and 24 years old representing the three sites of the riverine forests (gerf, maiya and kareb) and were composed of 1800 measurements points. The data for the third part came from 200 trees of two XII ages (3 and 6 years old) representing two stem origins (seedling and coppice); the data was composed of 2000 measurement points. The measurements undertaken were the main bole height, diameter at breast height (1.3 meter) and the diameters at ten percent intervals of the main bole height. The models were fitted by non-linear least squares regression using Gauss-Newton’s iteration method and residual plots were attained. To compare the efficiency of fitting the models to observed data non-parametric statistics (main bias, standard error of estimate and mean absolute difference) and coefficient of determination were computed and the models were ranked using the procedure of Kozak and Smith (1993). Moreover, equivalence tests were conducted to test equality of the predicted and observed diameters along the stem. In the first part, the results indicated that the variable-exponent models of Lee et al. (2003) and Kozak (1988) were preferred to describe the stem taper of the eight species; their mean bias was less than 0.3 cm and they explained >87% of the variation in diameters. These models were followed by the sigmoid-form models of Ormerod (1973) and Biging (1984). The polynomial taper model of Kozak et al. (1969) was the worst for describing the stem taper of the studied species. Although the two models of Muhairwe (1994) were among the first ranking ones for all species, they failed to describe the stem profile at the lower and upper parts of the stem. In the second part, the models of Lee et al. (2003), Kozak (1988) and Ormerod (1973), which were superior to describe the stem profile of Sunt in the first part, were fitted to the data of Sunt trees representing each of three ages (16, 20 and 24 years) and to the trees growing on the gerf, maiya or kareb. The results indicated that the model of Kozak was the best model to describe the stem XIII profile of the Sunt trees for each of the three ages and three sites. The model explained high proportions of the variation in diameters along the stems (R2 ranged between 0.89 and 0.97) and the mean bias was less than 0.2 cm. The model of Lee et al was a competing model to the model of Kozak and can also be adopted as it is easier to compute and apply. The model of Ormerod failed to describe the stem profile of Sunt; it underestimated and overestimated the diameters of different sections of the main tree stem. Judging by Kozak’s function, the stem taper was comparable among sites but it increased with age. In the third part, the superior models of Lee et al. (2003), Kozak (1988) and Biging (1984), which were superior to describe the stem profile of Eucalypts in the first part, were evaluated on the data of Eucalypts trees of two ages (3 or 6 years) and trees of two stem origins (seed or coppice). The results indicated that the evaluated models were competing models. The mean bias of the models was less than 0.08 cm and stem tapering was strongly correlated with DBH and height of the tree stems (R2 ranged between 94% and 98%). The model of Lee et al. was the most superior model in describing the stem profile for Eucalypts trees of the two ages and from the two stem origins. The entire stem taper of the older trees, based on the stem profile using the Lee et al.’s function, was higher than the younger trees, and it was higher in the trees originated from seed than trees from the coppice
Description: 183 page
URI: http://khartoumspace.uofk.edu/handle/123456789/9377
Date: 2015-04-23


Files in this item

Files Size Format View

This item appears in the following Collection(s)

Show full item record

Share

Search DSpace


Browse

My Account