COS 33-4 - CANCELLED - Evaluating the role of biophysical conditions and agronomic legacies in vegetation recovery following agricultural land abandonment

Background/Question/Methods

The expansion of agricultural lands has been a primary engine of land cover change, driving losses in species diversity, abundance, and carbon storage. Yet, millions of hectares of agricultural land have been abandoned in recent decades, presenting opportunities for secondary vegetation growth. While abandoned fields have the potential to return to ecological communities with similar species diversity to their pre-agricultural state, they alternatively may transition to novel ecosystems or persist in degraded states that may have alternative functions that impact ecological and human communities. We still lack an understanding of how and why revegetation varies on degraded landscapes. Here we leverage remote sensing data and causal inference frameworks to assess revegetation patterns on former sugarcane land in Hawai‘i abandoned between 5 and 100 years ago. Here we address the question: What abiotic, biotic, and anthropogenic factors drive vegetation succession in terms of structure, composition, and function? A 2015 land cover data set provides reference classes for vegetation structure (grass, shrub, tree) and composition (native versus invasive species). Function is captured using seasonal composites of Sentinel 2 imagery to calculate variation in a suite of vegetation indices (EVI, NDRE, and NDWI) that reflect productivity, chlorophyll, and water content.

Results/Conclusions

We find that secondary vegetation progresses from grasses to shrubs to trees with time since abandonment, paralleling existing succession theory. Native vegetation represents a higher proportion of the secondary vegetation in the decades immediately following abandonment, but introduced species dominate fields more than 80 years after they were abandoned. Preliminary findings indicate that native vegetation succession tended to occur when abandoned sugarcane fields were at higher elevations, where they were also closer to intact forest. We discuss the implications of these results and ongoing analyses of these succession patterns in the context of key landscape variables which aim to characterize more nuanced threshold conditions that determine succession trajectories. Ongoing research aims to refine the assessment of how biophysical properties such as landscape fragmentation and proximity to intact vegetation influence vegetation recovery. Abandonment is increasingly occurring on high value agricultural land globally, including in California’s Central Valley. This work in Hawai‘i could provide insight into land management strategies for environmental scientists, policy makers and land managers concerned with the potential for abandoned land to be managed as a multifunctional resource with ecological, social, and cultural utility.