Natural forest regeneration and ecological restoration in human-modified tropical landscapes

2016 
In human-modified tropical landscapes (HMLs) the conservation of biodiversity, functions and services of forest ecosystems depends on persistence of old growth forest remnants, forest regeneration in abandoned agricultural fields, and restoration of degraded lands. Understanding the impacts of agricultural land uses (ALUs) on forest regeneration is critical for biodiversity conservation in HMLs. Here, we develop a conceptual framework that considers the availability of propagules and the environment prevailing after field abandonment as two major determinants of forest regeneration in HMLs. The framework proposes that regeneration potential decreases with size, duration and severity of agricultural disturbance, reducing propagule availability and creating ill-suited environmental conditions for regeneration. We used studies from Southern Mexico to assess this framework. First, we identify regeneration bottlenecks that trees face during transit from seed to follow-up life stages, using demographic analysis of dominant pioneer species in recently abandoned fields. Then, we explore effects of ALUs on forest regeneration at the field and landscape scales, addressing major legacies. Finally, we integrate agricultural disturbance with landscape composition to predict attributes of successful second growth forests in HMLs, and provide indicators useful to select tree native species for active restoration. An indicator of disturbance inflicted by ALUs, based on farmers’ information, predicted better regeneration potential than measurements of soil and microclimate conditions at time of abandonment. Cover of cattle pastures in the landscape was a stronger indicator of forest regenerating attributes than cover of old growth forest remnants. To conclude, we offer recommendations to promote forest regeneration and biodiversity conservation in HMLs.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    97
    References
    57
    Citations
    NaN
    KQI
    []