Selecting species for restoration in foundational assemblages

2021 
Humans have long sought to restore species but little attention has been directed at how best to do so for rich assemblages of foundation species that support ecosystems, like rainforests and coral reefs that are increasingly threatened by environmental change. We developed a two-part triage process for selecting optimized sets of species for restoration. We demonstrated this process using phenotypic traits and ecological characteristics for reef building corals found along the east coast of Australia. Without clear linkages between phenotypic traits and ecosystem functions, the first part of the triage hedges against function loss by ensuring an even spread of life history traits. The second part hedges against future species losses by weighting species based on characteristics that are known to increase their ecological persistence to current environmental pressures--abundance, species range and thermal bleaching tolerance--as well as their amenability to restoration methods. We identified sets of ecologically persistent and restorable species most likely to protect against functional loss by examining marginal returns in occupancy of phenotypic trait space per restored species. We also compared sets of species with those from the southern-most accretional reef as well as a coral restoration program to demonstrate how trait space occupancy is likely to protect against local loss of ecosystem function. Synthesis and applications. A quantitative approach to selecting sets of foundational species for restoration can inform decisions about ecosystem protection to guide and optimize future restoration efforts. The approach addresses the need to insure against unpredictable losses of ecosystem functions by investing in a wide range of phenotypes. Furthermore, the flexibility of the approach enables the functional goals of restoration to vary depending on environmental context, stakeholder values, and the spatial and temporal scales at which meaningful impacts can be achieved.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    64
    References
    0
    Citations
    NaN
    KQI
    []