Modeling Tree Mortality for Eucalyptus camaldulensis (Dehnh) Stands in Afaka Forest Reserve, Kaduna-Nigeria

Abstract

This research was conducted to investigate mortality using logistic regression (Models) for River red-gum (Eucalyptud camaldulensis Dehnh) in Afaka Forest Reserve, Kaduna, Nigeria. Stratified random sampling and proportional sampling allocation were used in selecting plots and individual trees for measurements. Data on live and dead trees were recorded, variables measured are diameter, total tree heights, spacing and age of the stands within the plots. Diameter at the breast height (DBH) was measured using diameter tape while Haga Altimeter was used in taking the total height of the sampled stands. The Data collected were divided into four portions. Three portions (75%) for model calibration adopting logistic regression model, 25% were used for model validation of logistic model generated.  Chi square test of goodness-of-fit, Negelkerke R2, Cox and snell R2 values were used for model evaluation. The positive regression coefficients (16.737, 49.266 and 1.992) showed by Dbh, Height and age (as independent variables) indicated positive future mortality effects on the stands; as the trees ages and increases in size and height, the chances of mortality will increase. At the present age and sizes of the stands the fitted model shows no significant relationship between the current age, height and size of the stands, as such the notion held by the management of the plantation that Eucalyptus cannot grow for many years and in large sizes is not true. There is therefore, need for periodic re-measurement of the stands to evaluate future effect of the variables (Age, Dbh and height) on mortality.