Models For Growth Characteristics And Their Applications In Yield Studies For Pinus Caribaea Morelet 1851 In Southwestern Nigeria

ABSTRACT

Development of empirical models provides ample prospects of exploring established

mathematical theories and relationships among tree growth variables for sustainable

forest management options. However, there is inadequate information on modelling

growth characteristics and productivity potentials of Pinus caribaea, a pulpwood species

which is now being proposed for timber production in Nigeria. Hence, models for growth

characteristics and their applications in yield studies for Pinus caribaea in southwestern

Nigeria were investigated.

Sixty Temporary Sample Plots (TSPs) of size 20m x 20m each and of different Age (A)

series based on stocking density were sampled from Omo Forest Reserve (n=16 TSPs -15

and 21 years), Oluwa Forest Reserve (n=36 TSPs-18, 20, 35, 36 and 37 years) and

Shasha Forest Reserve (n=8TSPs-27years) in Ogun, Ondo and Osun States respectively

using stratified random sampling method. Growth variables: Diameter at Breast Height

(DBH), Stem Quality (SQ) and Total Height (THT) were measured in each plot and

Basal Area (BA), Stem Volume (SV), Tree Slenderness Coefficient (TSC), Crown Ratio

(CR), Crown Projection Area (CPA) and Site Index (SI) estimated. A total of 1,592 trees

were enumerated and each tree per plot classified into four canopy layers as dominant,

co-dominant, intermediate and suppressed. Data obtained were used for modelling and

categorized into individual tree level, size class level and whole stand level with

Maximum Likelihood Estimation (MLE) technique adopted in 3-parameter Weibull

Probability Distribution Functions (WPDF). Data were analysed using descriptive

statistics, ANOVA, correlation and regression analyses at α0.05.

Growth variables exhibited significant variations among the canopy layers with the

dominant canopy layer having the highest mean THT of 14.9±0.2m and intermediate

canopy layer lowest with 7.8±0.1m. Stem volume (1.4±0.04m3) and BA (0.1±0.002m2)

were also highest in the dominant canopy layer. Slenderness coefficient however was

highest (80.6±1.1) within the suppressed canopy layer while crown ratio was highest

(0.27±0.001) in intermediate canopy layer. There were positive relationships between the

growth variables across the stands with coefficients of correlation (r) ranging from 0.01-

0.98 for individual tree level and 0.41-0.91 for whole stand level. Similar results of

UNIVERSITY OF IBADAN LIBRARY

iii

correlation between Weibull parameters and other growth variables at the size class level

were significant with r ranging between -0.72 and 0.92 with location parameter „a‟ of the

WPDF having the highest positive association of r = 0.92 with DBH. The growth-yield

models were of the forms lnCR =-3.5507+0.6263lnTHT+0.1558CPA+10.2339A1, R2 =0.

98 for dominant layer, lnCR = -3.1922+0.6363CPA+7.6876A-1, R2 = 0.994 for

intermediate layer, lnTHT= 0.4547+0.0032lnSQ+0.6757CPA+0.1052A-1, R2=0.989 for

suppressed layer and lnTHT=0.5872+0.5348CPA+0.0567A-1 with R2= 0.996 for codominant

layer respectively. The overall best model among individual, size and whole

stand categories was found within the whole stand level with R2=0.999 and of

exponential form: lnSV=-0.7759+0.0001SI+0.0005A-1 +0.9532lnBA.

The best adjudged growth-yield model among the canopy layers was found within the codominant

layer. The selected whole stand growth-yield model was suitable for both

current and future prediction of major growth characteristics and productivity potentials

of Pinus caribaea in southwestern Nigeria.

Keywords: Pinus caribaea, Growth variables, Yield models, Tree canopy layers