DEVELOPMENT OF ALLOMETRIC MODELSFOR TOTAL ABOVE GROUND BIOMASS Estimation OF Adansonia digitata (BAOBAB) IN SUDANO-SAHELIAN REGION OF NIGERIA

• R. B. Shuaibu

  Department of Forestry and Wildlife Management, School of Agriculture and Agricultural Technology, Moddibo Adama University of Technology, Yola, Nigeria

  Author

• D. F. Jatau

  Department of Forestry and Wildlife Management, School of Agriculture and Agricultural Technology, Moddibo Adama University of Technology, Yola, Nigeria

  Author

• M. G. Saka

  Department of Forestry and Wildlife Management, School of Agriculture and Agricultural Technology, Moddibo Adama University of Technology, Yola, Nigeria

  Author

There is the need for an accurate and reliable ways of analyzing and estimating the accumulated tree biomass of multipurpose indigenous tree species without destroying them. Therefore, this study developed allometric models for Total Above Ground Biomass (TAGB) estimation ofAdonsonia digitata (Baobab tree)in Sudano-Sahelian region of Nigeria using the non-destructive sampling method of data collection. Data for fitting and evaluation of the TAGB were collected from 150 trees. Twenty-Five (25) trees that represent the range of sizes (49cm to 127cm)in each of the Six (6) States in the region selected for the study were selected for sampling. Healthy trees, with good growth form were selected, hollow trees and trees with a broken top were excluded to avoid bias. The biomass data of the sampled trees formed the database for fitting the biomass estimation equations. Eighty percent (80%) of the data set were used for the calibration, while Twenty percent (20%) were used for validation. The Statistical Package for Social Scientists (SPSS) version 20 was used to generate the regression models for TAGB estimation. Nine (9) forms of allometric models were developed for TAGB estimation of this species in the Study Area and Four (4) models from literatures were fitted into the data. The models were evaluated using some statistical criteria. The results of the assessment criteria shows that all the Nine (9) models developed by me and the Four (4) models from the literature yielded high multiple correlation coefficient (R), coefficient of determination (R2) and highly significant F-ratio (p < 0.05),These results show that the independent variables reliably predicted the dependent variable. In addition, the Root Mean Squared Error (RMSE) and the Standard Error of Estimate (SEE) were very low for all the Thirteen (13) models, indicating that there is a very strong relationship between the dependent and the predictor variables. It also indicated the good ability of the models to adequately estimate the total above ground biomass of the stands. Four best models from the 9 models developed are as follows: lnM = 1.674 + 1.000 ln(D*Bc); lnM = 1.673 + 0.00001134 ln(V) + 1.000ln(D) + 1.000 ln(Bc); M = -0.00008664 + 5.333 (D*Bc); M = 6.373 + -4.489 (V) + -18.275 (D)+ 6.150 (Bc). The R, R2, SEE, RMSE, and F-value are 1, 100%, 0.0001, 0.00001, and 58038507; 1, 100%, 0.001, 0.00001, and 13364541; 1, 100%, 0.001, 0.00001, and 3070407; 0.99, 99.8%, 1.1, 1.2, and 19595 respectively. The two best models from the literature are ln TAGB = 3.615 + 3.556 ln(dbh2) by Goussanou et al. (2016); and ln TAGB = -.251 + -.427 ln(dbh) + 4.500 ln (WD) by Basuki et al. The R, R2, SEE, RMSE, and F-value are 0.93, 86.7%, 2.2, 5, and 305; and 0.96, 95%, 0.1, 0.003, and 3132 respectively. The residual plots against nine predicted values indicated homogenous variance. The four biomass models from the available literature also fitted into the data also gave the same trend as the estimated data. Based on the assessment and fit statistic results, the application of the models for tropical forests cannot be overemphasized.

1. R. B. Shuaibu, Department of Forestry and Wildlife Management, School of Agriculture and Agricultural Technology, Moddibo Adama University of Technology, Yola, Nigeria

   Department of Forestry and Wildlife Management, School of Agriculture and Agricultural Technology, Moddibo Adama University of Technology, Yola, Nigeria

2. D. F. Jatau, Department of Forestry and Wildlife Management, School of Agriculture and Agricultural Technology, Moddibo Adama University of Technology, Yola, Nigeria

   Department of Forestry and Wildlife Management, School of Agriculture and Agricultural Technology, Moddibo Adama University of Technology, Yola, Nigeria

3. M. G. Saka, Department of Forestry and Wildlife Management, School of Agriculture and Agricultural Technology, Moddibo Adama University of Technology, Yola, Nigeria

   Department of Forestry and Wildlife Management, School of Agriculture and Agricultural Technology, Moddibo Adama University of Technology, Yola, Nigeria

• PDF