Publication Date

Spring 2021

Advisor(s) - Committee Chair

Dr. M. Stone (Director), Dr. E. Gray, and P. McKillip

Degree Program

Department of Agriculture and Food Science

Degree Type

Master of Science


Only limited data had been published on the survivability of specimens within public gardens. This may be due to not enough data collected or how vast the plant selections are in the gardens. The Baker Arboretum has collected data by accessioning plant collections over a period of 30 years. The Baker Arboretum has its specimens GIS mapped on the 115-acre property for easy location and detection of the the woody ornamental plants. However, little research is available to understand which coniferous specimens have the best success in the garden. In this study, six separate binomial logistic regressions were run to determine the odds of success. The dependent variable used in the regressions to measure survival, 0 being dead and 1 being alive, were of the 'quantity now' in each data set. The predictor variables in each regression were nativity, season planted and container size. The specimens that were analyzed were Cupressaceae, Pinaceae, Cephalotaxaceae, Ginkgoaceae families and also the Picea and Chamacyparis genus'. The odds ratio was used to determine what likelihood each significant predictive variable has in accordance to survival. Of the six taxa groups analyzed, only three of the regressions were found to be significant by using predictor variables to determine to the odds of survival. Chamaecyparis, Cupressaceae and Pinaceae were the three specimen groups that showed significance in survival with the predictor variables. The other three, Cephalotaxaceae, Ginkgoaceae and Picea groups were reported non-significant. The Cupressaceae family specifically had shown parallel predictions of the expected survival with biological predictions. Completion of this study provides more knowledge on how to track and analyze survivability odds in public gardens, helping further ex-situ conservation.


Agriculture | Biodiversity | Forest Sciences