GROWTH MODELS AND SITE QUALITY ASSESSMENT OF PRIVATELY GROWN Tectona grandis LINN.F STANDS IN EKITI STATE, NIGERIA.

Abstract

This study was carried out to develop volume, diameter, site index models and to determine the nutrient status of the soils for sustainable management of privately grown stands of Tectona grandis Linn.f stands in two selected locations (Ikere and Aisegba) in Ekiti State, Nigeria. A field inventory was carried out to obtain tree growth variables in the two selected locations for models fitting and site quality assessment while ages of the plantation were obtained from the plantation records. The data for the study were collected from three (3) age series (10, 13 and 15 years) in each of the locations. Five (5) temporary sample plots (25 x 25m) were randomly located per age series and 70% of the data were used for calibration, while the remaining 30% were used for validation purpose. Soil samples were collected diagonally from three soil depths of 0 -15 cm, 15 – 30 cm and 30 - 45 cm in each of the sample plots. The soil samples were analysed to provide information on the physicochemical properties. The data were subjected to both linear and non-linear regression analyses. Models were generated for each location separately as well as for both locations. Four model forms (Simple linear regression, binomial, polynomial and Logarithm transformed) were adopted. All the models were assessed using several statistical criteria. But only models with good fit were validated using a paired t-test and percentage bias. The Sigmond models were generated for the non-linear regression model using Curve Expert professional software and the models were assessed using Akaike Information Criterion (AIC) and standard error and validated using the Bias percentage, student t-test and residual plots. Five Density Probability Functions were adopted for diameter distribution models with Easy fit software. The goodness of fit were tested with Kolmogorov-Smirnov, Anderson Darling and Chi-square statistics and the graph of observed and the estimated probability functions. The Site quality models were developed using a reference age of ten years. The proportion curve method adopted in developing site index curve was divided into five site quality classes. The result indicated a total of 1085 trees were enumerated in the study area, though the density of trees was higher in Aisegba (3235/ha) than Ikere(2544/ha). The mean DBH of trees ranged from 15.58 cm to 19.06 cm in both locations while mean tree height varied from 6.83 m to 10.49 m. The stand volume varies from 72.81 m3/ha to 192.45 m3/ha. The R2 for the linear models ranged from 59% to 87%. The recommended models for Ikere is the logarithm transformed model LNV=-9.691+0.961LND2H, the binomial model V=0.008+(0.003D)+0.001D2 for Aisegba and logarithm model LNV=-9.610+0.948LND2H for the pooled data. The results also revealed that Logistic, Gompertz Relations, and Logistic models gave the better output when compared with the observed volume in Ikere, Aisegba and combined data for the non-linear regression models. For the diameter distribution models the Weibull, Beta and Johnson SB were selected as the most flexible distribution models for Ikere, Aisegba and pooled data respectively. The site index equation that was found suitable for site quality evaluation for the plantation using 10 years as index age is 𝑆𝐼=exp[ln(𝐻𝑑)− 10.86.(0.1− 𝐴−1)] for Ikere. For Aisegba, the best site index models is 𝑆𝐼=exp[ln(𝐻𝑑)− 5.27 (0.1− 𝐴−1)] and the model for the two locations is given as 𝑆𝐼=exp[ln(𝐻𝑑)− 5.27 (0.1− 𝐴−1)] . The results of the soil physical and chemical properties generally decrease with an increase in soil depth. The models are recommended for the management of privately grown teak in the study area. The active participation of individual in plantation development is also recommended by this study.