Sensitivity of plant growth to steady changes in rainfall amounts (i.e., as lands get steadily wetter or dryer in the future) and year-to-year variability (i.e., the magnitude of responses in a wet year versus a dry year) across different geographic regions is important to understand if we are to prepare for future climate scenarios, which impact us in a variety of ways, such as altering sustainable cattle stocking practices and the carbon dioxide buffering ability of landscapes. First, as environmental conditions due to global change drivers change, the magnitude of response of a particular landscape will depend on its (1) ecosystem sensitivity. One factor which determines the sensitivity of a particular area are the (2) soil properties present -- these include the type and size of soil particles, the type of microbes and soil animals, as well as the quantity of nutrients available for plant uptake. (3) Individual species responses to changes in resources eventually result in changes in (4) community composition which can affect landscape responses directly through changes in the identity of dominant species or biodiversity or indirectly through effects on system sensitivity. My research focuses on clarifying various portions of this framework showing how ecosystems can respond to global change drivers.
Grassland sensitivity to precipitation
Great Plains Triangle Experiment: Climate models forecast alterations to both the amount of precipitation falling onto landscapes as well as shifts in the pattern in which this rain falls – coming in larger storms bounded by longer drought periods. Ecosystem sensitivity is the responsiveness of a system to alterations in the environment (i.e., how much biomass is lost due to a severe drought or grows during a wet year); patterns of sensitivity are important to understand if we are to predict responses to future climatic changes. We examined aboveground and belowground (roots) growth sensitivities to experimentally altered precipitation regimes in three different US grassland types. We found that more northern sites dominated by cool season grasses were insensitive to alterations in precipitation amount and pattern (many small vs. a few large events) in both years of the experiment, likely due to early growth patterns of the existing species when soil moisture is saturated from snow-melt. In both of our more southerly grasslands dominated by warm season grasses, we found different patterns of sensitivity depending on whether we looked at above- or belowground growth. This is important information as many past and current studies assume that above and belowground sensitivities are similar when this may not be the case. See Wilcox et al., 2015 for more information.
Plant community controls on ecosystem sensitivity: Central to understanding of earth’s C cycle is the functional relationship between precipitation and net primary production (NPP). Theory and empirical models predict that chronic changes in ecosystem water balance will alter aboveground net primary productivity (ANPP) with the magnitude increasing over time as plant communities shift. However, ecosystem sensitivity (ANPP responses) to interannual precipitation variability is predicted to vary inversely with water availability. We increased water availability for two decades in native grassland and found that ANPP increased and ecosystem sensitivity decreased as predicted, but only initially. After 10-years, plant communities changed but ecosystem sensitivity unexpectedly returned to previous levels. In grassland sites with lower water availability and ANPP, ecosystem sensitivity never exceeded that in more mesic sites over 26-years, despite altered plant communities, showing that shifts in abundance of key plant species can stabilize precipitation-productivity relationships thus providing surprising functional resistance to climate change.
Interactions between fire and altered water availability on long term carbon and nitrogen pools: Chronic changes in climate and disturbance regimes may substantially impact ecosystem function and services worldwide with a forecast increase in periods of extremity pushing systems beyond critical thresholds. In productive grasslands, alterations in the water balance (wetter or drier soils) and more frequent fires are expected in the future. With more frequent fire, ecosystem models predict soil carbon (C) and nitrogen (N) to be reduced through the repeated combustion of aboveground biomass reducing C and N inputs. This loss may be exacerbated if increases in water availability lead to plants allocating more biomass aboveground. In a native mesic grassland, to quantify how extreme climatic and disturbance regimes will impact plant allocation patterns and soil biogeochemical properties in a native mesic grassland, we pushed this ecosystem well beyond past environmental conditions by increasing precipitation inputs (by 32%) for over two decades (1991-2013) while imposing the highest fire frequency possible (annual spring burning). We predicted (1) soil C and N would decline over this 23 year period with annual fire, and (2) this reduction would be much greater in sites with chronically wetter soils as shifts in biomass and production allocation patterns favored greater aboveground growth. Contrary to our prediction, annual fire led to no change in total soil N over time and a significant increase in soil C (52%). More surprisingly, annual fire combined with two decades of extreme growing season precipitation also resulted in a moderate increase in soil C (19%) although total soil N did decline by 14% under the combination of extreme climate and disturbance. The latter increase in soil C occurred despite greater annual losses of aboveground net primary productivity (ANPP) to fire in the extreme precipitation treatment (42%). No significant changes in tissue C:N were detected with precipitation additions and belowground net primary production (BNPP), root turnover rates and root:shoot ratios all were reduced, with each of these potentially reducing soil C. The results from this long-term imposition of extreme disturbance and climatic regimes suggest that soil C pools in this grassland are not likely to decrease as forecast from past modelling exercises but that N limitation may increase when such extremes are combined.
Nutrient controls on community dynamics and ecosystem functioning
Alterations of nutrient levels have the potential to drastically affect ecosystem functioning directly through changing of the availability of essential nutrients (e.g., nitrogen and phosphorus) as well as indirectly through restructuring of the plant community which can have its own effect on system productivity. I am involved in a number of collaborative projects examining the effects of nitrogen (N) and phosphorus (P) on various grassland ecosystems. In tallgrass prairie, we found that long-term chronic addition of N and P initially increased primary production on shorter time scales. However, over time these nutrient additions reduced the abundance of species that typically dominate natural prairie and increased abundance of more weedy species. After the community shift, primary productivity went back to normal levels but year-to-year variation increased. See our paper for more information.
comparing strengths of major drivers in north american and south african grasslands
Since 2006, the Konza-Kruger study has experimentally examined the effects of fire and grazing in a North American tallgrass prairie (Konza Prairie Biological Station, Manhattan, KS) and a South African grassland savanna (Kruger National Park, Mpumalanga). Nested within three long term fire regime treatments (1 year, 3-4 year, and 20 year - unburned) we have established and maintain grazing exclosures at both sites. We collect annual measurements on species composition (abundances), annual net primary productivity, herbivore abundances, and canopy density. So far, we have found that intermediate fire frequencies result in higher richness (i.e., the number of species per plot) than annual or very infrequent burning at Konza, but infrequently burned areas showed the highest richness in Kruger. Also, the removal of grazing caused drastic declines in species richness from 2006 - 2012 while we saw no response at Kruger. We think this is due to the subtly differential ways dominant species at these two sites exist in the landscape as well as the properties of these species. In Konza, most areas are dominated by a tall, C4 perennial grass - Andropogon gerardii, common name: Big bluestem - which is preferential grazed by bison thereby releasing more rare species from light and space competition. Contrastingly, there are many species which can dominate the community in Kruger depending on the fire frequency and many of these species are unpalatable resulting in limited change of dominance (i.e. how abundant is the most abundant species) when grazing pressures are removed. We suspect that this lack of change in dominance is behind the limited response of diversity in Kruger. See the Konza-Kruger website for more information: https://wp.natsci.colostate.edu/konzakrugerexperiment/welcome-to-the-konza-kruger-experiment/
Ecological stoichiometry as a predictor of species responses to global change drivers
In ecology, much research is dedicated to the search for general patterns applicable across broad categories (e.g., across space, species, individuals, etc.). Ecological Stoichiometry has been shown to be particularly useful in discovering some of these patterns. The stoichiometry of organisms (i.e., the amount of essential elements present in an organism) combined with information about how much of these elements the organism needs can tell us much about how it will respond in the future (here's a cartoony example: if I haven't eaten since breakfast and its 5pm, it is reasonable to predict that I will eat soon). One interesting characteristic which differs among plant species is stoichiometric homeostasis. This is more or less the "you are what you eat" principle. Species which have high homeostasis control the levels of their internal elements despite changes in the quantity of elements available in their environment; animals are highly homeostatic but plants vary widely in this ability. In a study done in 2011 at the Konza Prairie Biological Station, we found that more homeostatic plant species were more dominant in native tallgrass prairie under ambient nutrient conditions. However, we also found that under constant nutrient (nitrogen) enrichment, this relationship broke down and species which were less homeostatic began to dominate. This is likely due to high nutrient requirements of less homeostatic species but also higher growth rates when these requirements are met. Lastly, we found species more homeostatic for nitrogen were more resistant to long-term changes in precipitation regimes.