This study, using a hybrid ecosystem approach, was designed to test whether hybrid ecosystems can 1) maintain themselves with relatively little input; 2) are capable of sequestering substantial amounts of carbon; 3) sustain a broad range of native biological diversity; and 4) remain open enough at ground level to allow human movement through them. The results of this study are expected to directly benefit the military mission in the Pacific. Currently the prevalence and dominance of invasive species in DoD lands in Hawai‘i and the Pacific has precluded the ability to effectively use the landscape for necessary training maneuvers. Hybrid ecosystems are an approach to allow training while still protecting endangered species and their associated environments.

Technical Approach

Project objectives were addressed and tested in the lowland wet forest at the Keaukaha Military Reservation on the Island of Hawaii. The application of functional trait theory in restoration and management is an exciting new approach that can be used to understand the persistence of species and ecosystems – and to build model communities with desired ecosystem functions. In this project a functional trait based restoration approach was used to select the native and non-native species for the hybrid communities planted. Principal components analysis was implemented to design communities that foster slow and moderate carbon turnover rates and also test ecological theory concerning complementary and redundant trait space within plant communities. The project team hypothesized that this higher functional diversity will be advantageous in the goals of higher carbon sequestration and higher resistance to weed invasion, which should lead to a lower understory cover that fosters native regeneration and allows for the greater human mobility required for military training. Surveys of abiotic (i.e., leaf area index, canopy openness, soil nutrient availability) and biotic (i.e., tree basal area and density) parameters were measured across the twenty plots prior to clearing of invasives and the planting of the experimental treatment communities. Monitoring of abiotic and biotic (i.e., tree growth and survival, native seedling regeneration, litterfall inputs, litter decomposition rates, seed rain, reproductive phenology, carbon sequestration and resistance to weed invasion) parameters were continued post-planting, along with mechanical methods of plot maintenance. The REST computer program was developed to provide a user friendly tool for those wanting to implement a functional trait based restoration strategy to degraded ecosystems.


The early results of this experiment show that the treatments have a drastically different environment than the invaded reference condition, with large increases in light availability and recruitment of new individuals. Specifically, natives and non-natives were found to occupy separate trait space when evaluated with the principal components analysis; natives tended to have higher values of foliar C:N and leaf mass per area, and smaller values for foliar N, seed mass, leaf area, and maximum height. Invasives overlapped trait space of both natives and non-natives. The twenty plots were similar in the abiotic and biotic parameters measured prior to applying the experimental treatments. Yet, pre-removal native species density, LAI, canopy openness, soil carbon and soil sodium differed significantly between the plots. To evaluate the impact of this restoration approach on ecosystem services we projected values of carbon and biodiversity using a return on investment approach. When accounting for estimated expenditures over longer terms, project return on investment varied based on carbon storage income as well as including or excluding biodiversity from income. With higher carbon market rates and a favorable labor decrease (25% of current rates), stored carbon alone presents an investment return of approximately 56 years. Including biodiversity results in an economic recovery period of approximately 45 years, supporting our objectives within a 50-year management time frame.


This study’s results are applicable throughout LWF in Hawai‘i. Most importantly, these results will directly help the military meet the stewardship responsibilities of Army National Guard land by providing guidance on species choice in restoration. The approach could be applied to other heavily-invaded DoD sites to guide these areas toward lower intensity, more sustainable, and cost-effective management in the long-term. With REST, there is the ability to test in other ecosystems the four restoration objectives currently described within the program: successional facilitation, fire tolerance, drought tolerance, and carbon storage.