Ecohydrological implications of clonal shrub encroachment in tallgrass prairie

Abstract

Plants are an integral part of the water cycle in nearly every terrestrial ecosystem. Acting as ‘pipelines’, they link the atmosphere to subsurface water pools through the process of transpiration and impact belowground movement of water by altering soil structure and infiltration pathways. As a result, plants are highly responsive to the availability of water and play a substantial role in shaping pathways of water-movement through ecosystems. Different plant functional types (e.g., grasses, trees, and forbs) vary widely in growth forms, life history traits, and water-use strategies, and therefore have different relationships with the ecosystems they inhabit. As a result, transitions from one dominant vegetation type to another can create unexpected shifts in water cycling. Many grassland ecosystems are experiencing land-cover change in the form of woody encroachment – the spread of woody vegetation in historically open, grass-dominated ecosystems. Woody shrubs and trees typically have higher rates of water-use than the grasses they replace. Therefore, as woody vegetation spreads across a landscape, water-loss through the vegetation ‘pipeline’ increases. This has the potential to reduce deep soil water availability, stream / river flow, and groundwater recharge. My dissertation research has focused on the relationship between water and vegetation in tallgrass prairie and how woody encroachment alters those relationships. Using long-term records of precipitation and woody cover, as well as a multi-year drought x fire experiment at Konza Prairie Biological Station (KPBS; northeastern KS), I (1) assessed the impacts of shrub encroachment on water yield at KPBS over the last 4 decades and (2) determined how shrub encroachment has altered grassland responses to drought. I found that stream flow at KPBS has declined over the past 30-40 years at KPBS, despite an increase in annual precipitation over the same time period. There has been a slight shift toward larger rainfall events over the last century, but no changes in seasonality or number of rainfall events that would explain this decline in stream flow. Instead, we found that increases in woody cover over the past 3-4 decades have been highly correlated with declining streamflow. A ‘breakdown’ in the relationship between precipitation (supply) and stream discharge (output) occurred shortly after a period of rapid woody expansion at KPBS around the year 2000. This suggests that increasing woody cover has altered the ecohydrology of this tallgrass prairie ecosystem. In a separate study, I explored the primary mechanism by which woody encroachment impacts water yield – increased evapotranspiration. I quantified the increase in woody cover from 1978-2020 in one watershed at KPBS and combined it with known rates of water-use by dominant grasses and shrubs to estimate watershed-scale changes in vegetation water-use through time. I found that a 20% increase in woody cover from 1978-2020 led to an estimated 25% increase in water-use. This represents a substantial shift in the water budget of this ecosystem and has likely contributed heavily to observed declines in stream flow and the weakening of the precipitation-discharge relationship. In addition to assessing how woody encroachment has impacted water cycling, I also explored how increased woody cover impacts grassland responses to water availability. In a long-term drought x fire experiment at KPBS, I altered water availability (50% precipitation reduction vs. ambient precipitation) in watersheds with a history of contrasting fire frequency (1-year vs. 4-year fire frequency) to determine how drought and burn history interact to impact the growth and survival of encroaching shrubs and co-existing C₄ grasses. I also characterized the water-use traits and strategies of a rapidly encroaching clonal shrub (Cornus drummondii) and a dominant C₄ grass (Andropogon gerardii) to better understand their responses to changes in water availability. I found that C. drummondii was highly resistant to reductions in water availability – aboveground biomass and stem density were not impacted by five consecutive years of drought treatment. This resistance was facilitated by the unique water-use strategy of C. drummondii compared to co-existing grasses. Seasonal access to deeper soil water allowed C. drummondii to maintain consistent rates of carbon fixation and transpiration even during drought conditions. In fact, access to deeper soil water facilitated a ‘wasteful’ water-use strategy in these shrubs, where stomata remained open even when there was no additional increase in carbon fixation. This led to low water use efficiency (carbon gain per unit of water lost) and sustained high rates of water-loss even when conditions were relatively dry. Together, these results suggest that (1) shrub growth and survival will likely not be affected by future droughts unless they are severe and/or long enough to impact deep soil moisture and (2) continued transpiration by dogwood during years with low precipitation will likely lead to faster depletion of soil moisture pools and further reductions in stream flow in this system. Taken together, these studies indicate that large-scale increases in shrub and tree cover in mesic grasslands in the Central United States are likely to have widespread negative consequences for local and regional water yield.