COS 127-1 - Accelerated shrub encroachment across the US Great Plains: Using model-data fusion to assess causes and consequences

Abstract: Shrubs are encroaching upon and eradicating many grasslands across Earth, and we currently have limited understanding about (1) why this is occurring, and (2) the resulting consequences for critical ecosystem services, such as carbon sequestration. This is particularly important as global changes are likely to increase rates of encroachment. In this talk, we will use an extensively studied US tallgrass prairie as a model ecosystem to address both of these knowledge gaps using a combination of observational, experimental, and process-based modeling approaches. More specifically, we have measured a wide range of above and belowground morphological and physiological characteristics of shrubs and grasses in a tallgrass prairie in eastern Kansas (Konza Prairie LTER), as well as soil carbon turnover rates within the grass-shrub ecotone. We have used these data along with four decades of shrub encroachment observations to parameterize and fine tune two state-of-the-art process-based vegetation demographic models, BiomeE and FATES. Using these models, we were able to (1) conduct sensitivity analyses to assess mechanisms driving shrub encroachment, and (2) make predictions of future shrub encroachment rates under various global change scenarios.

We found that shrubs had lower photosynthetic rates than grasses, but photosynthesis was more stable under drought, likely due to smaller yet more numerous conduits in roots. This was an important driver of shrub encroachment and caused acceleration of shrub encroachment under future global change scenarios containing more frequent droughts. Additionally, energetically cheap clonal reproduction associated with shrubs led to fast acceleration rates under elevated CO₂ and muted the ability of fire to control shrub abundance. Instead of the expected increases in carbon storage by shrubs in deeper soils, we found less carbon across all soil depths under shrubs leading to reductions in long-term carbon sequestration in these systems. Altogether, our results suggest that shrub encroachment may continue to accelerate in the future, and this may have substantial negative effects on ecosystem services provided by grasslands in the US Great Plains. We suggest that land management decisions, at least in tallgrass prairie, should reflect the need to limit shrub encroachment in these ecosystems.