Science Journal of Agricultural Research and Management
August 2014, Volume 2014, ISSN: 2276-8572
© Author(s) 2014. This work is distributed under the Creative Commons Attribution 3.0 License.
Regression Models for Tree Volume Prediction in Stands of Tectona grandis (Linn) at Federal College of Forestry, Jericho, Ibadan, Oyo State Nigeria
*Nurudeen T.A1,. Abiola J.K2,. Salami K.D3,. Erinle O.A1 and Olaniyi W.A3
1Forestry Research Institute of Nigeria, P.M.B.5054 Ibadan, Oyo State Nigeria.
2Federal College of Forestry, P.M.B 5087 Jericho, Ibadan, Oyo State Nigeria.
3Dept of Forest Resources Management, University of Ibadan, Oyo state Nigeria.
Accepted 7th June, 2014, 2014; Available Online 2 August, 2014
The suitability of regression models for tree volume predictions in stands of Tectona grandis at Federal college of forestry, Ibadan was established in this study. Three regression models were developed each comprises of two models resulting into six models in all. Tree growth variables such as diameter at breast height (Dbh), diameter at the base (Db), diameter at the middle (Dm), diameter at the top (Dt) and tree height were measured to estimate the tree volume using Newton formula for tree volume estimation. The tree growth variable measured in the study area shows that the total basal area encountered per hectare was 610.64m2 while the total volume was found to be 15135.63m3 per hectare. The models developed showed that model 3a and 3b were found to be more suitable for tree volume prediction in the context of the data used. Models 1a and 1b were found inadequate and this cannot be used in tree volume prediction. Model 2a and 2b were found to be averagely adequate hence 3a and 3b proved best. Based on the evaluation of the models examined in this study model 3a and 3b were found to be more suitable and fit for volume prediction. The frequencies and yield prediction with the models are significantly different from their observed values according to the validation result with analysis of variance (ANOVA) and T-test values. Therefore, all categories of models generated in this study with good fit are recommended for volume prediction in Tectona grandis plantation at federal college of forestry Ibadan, Nigeria.
Keyword:Model, volume prediction, Tectona grandis, residual plots.