Conclusions & Future Analysis

In summary this project has established that analyzing and classifying a species succession classification can be done based on the frequency, clustering value, and the slope of the dbh distribution of the species in 3 successional stands. Both a PCA and CLUSTER analyses illustrated that the variables used in this project show some indication of the successional classification of a species. A discriminate analysis can be used to differentiate successional classifications for unknown species. The analyses suggests that early successional species can be predicted with low error rate, but mid and late successional species predictions are suggested to be less accurate overall. The accuracy of the differentiations needs to be further tested with larger known data sets to see the error rate from the discriminate analysis predictions can be improved. If further testing shows that using a discriminate analysis can differentiate successional classification amongst species with low error rates, then classify species into successional categories could be simplified, and the time necessary to differentiate the species could be shortened.

If further testing shows that this analysis can successfully classify species' successional classifications, then this process could secondarily be used to identify mid and late successional species that act similar to early successional species. If these species were desirable for the Filipino logging industry and could be identified then they could be planted and help increase the success of replanted clear cut stand in the Philippines to deter loggers from illegally harvesting the same species from the remaining Old Growth forest.


HOME
INTRODUCTION
DATA DETAILS
MULTIVARIATE METHODS
MULTIVARIATE RESULTS & DISCUSSION
CONCLUSION
APPENDICES
REFERENCES & ACKNOWLEDGEMENTS

DATA PREPARATION METHODS
DATA PREPARATION RESULTS & DISCUSSION

PRELIMINARY ANALYSIS