Download Full Text (499 KB)
Hawaiian forests have suffered damages from climate change, introduced plants, and introduced animals that trample or graze on the native vegetation. The 'Alalā, a native Hawaiian bird that is being reintroduced from captivity, is a disperser of native plants and can facilitate the restoration of native vegetation in Hawaiian forests. If the seed retention time and the flight paths of the 'Alalā are known, we can estimate the distribution of the seeds they disperse and see the effect the species has on forest restoration. In this study, the seed retention time of the 'Alalā is predicted from body mass using a generalized linear model (GLM). The model was based on body mass and corresponding seed retention times for multiple bird species that were collected from scientific literature. A random sample of 47 birds was selected from a data set of 55 bird species to create 1000 GLMs, then the estimated body mass of the 'Alalā was used to predict 1000 'Alalā seed retention times. To assess the fit of the 1000 GLMs, the body masses of the remaining, unselected species were used to predict corresponding seed retention times. The root mean square error values between the predicted and observed seed retention times were found for all 1000 GLMs. The estimated 'Alalā seed retention time ranged from 84.54-140.36 min, while the root mean square error values ranged from 4.97-65.98 min. These ranges illustrate the variance in a prediction depending on the chosen sample set. Replicating the experiment allowed us to see the distribution of predicted values, which give a big picture estimate of the 'Alalā's seed retention time. Further research can be done to estimate the flight paths of the 'Alalā, so their affect on forest restoration can be visualized.
Utah State University
Moses, Camilla, "The Importance of Experiment Replication in Understanding Distribution of 'Alalā Seed Retention Time" (2020). Fall Student Research Symposium 2020. 58.