Coastal and Plant Ecology Research
Coastal systems like barrier islands are good model systems for testing drivers of ecological change, specifically community structure and function. Barrier islands are dynamic in nature and can change morphologically over short temporal scales depending what plant communities are present and the intensity of press and pulse disturbance events they experience. Physical processes of barrier islands tend to attract higher focus but understanding interactions with plant communities are critical for predictions of how barrier island resilience will be affected by disturbances associated with changing climate and sea-level rise.
Connections between topography, disturbance, and vegetation communities
Barrier island plant communities are tightly coupled with physical abiotic processes across multiple spatial scales. Previous research has highlighted the importance of understanding topography-disturbance feedbacks, however knowledge gaps remain with regard to vegetation community structure and function on barrier islands that differ in topographic complexity and disturbance frequency. With this project I have been able to show that two geographically proximal barrier islands have distinct topographic profiles that influence trait-based community dissimilarity of adjacent habitats. As a result, islands that experience less disturbance have distinct dune and swale plant communities that significantly differ in ecosystem function (i.e., productivity) and community development drivers (i.e., competition vs. stress tolerance). Conversely, more disturbance was coupled with reduced topographic heterogeneity, resulting in high similarity of adjacent habitats and reduced ecosystem functioning.
​
Read the full story: Brown and Zinnert, Ecosphere 2020
Trait-based perspective of coastal grassland response to nutrient enrichment
Demystifying the impacts of nutrient enrichment is long standing in the scientific literature and has been one of the most pressing topics in plant ecology for decades. The goal of understanding how nutrient enrichment impacts biodiversity, community assembly, community synchrony, trophic stability, and ecosystem functioning is being approached across spatio-temporal scales and is shedding light on the ecological impacts of increased nutrient deposition. Nutrient enrichment research on coastal barrier island grasslands is lacking and previous dune fertilization studies have only focused on species composition and productivity metrics. For this project I shift that focus to understand how community-level functional traits inform community function in response to novel nutrient depositions (N, P, and NP). I found that nutrients affect productivity similar to global patterns (i.e., N and NP significantly increase plant productivity). Increased productivity was significantly correlated with functional trait alpha-diversity, highlighting a possible coexistence mechanism in highly productive communities. Importantly, I show species composition, trait-based composition, and lifeform abundance all produce conflicting results in response to nutrient additions. Such findings highlight complexities of community-level nutrient enrichment studies. A push for continued research is critical to resolve the context dependent nature of nutrient enrichment on plant community structure and function.
​
Read the full story: Brown and Zinnert, Diversity 2021
Nutrient enrichment as a driver of altered community composition, structure, and organization
In light of the important complexities I found in Brown and Zinnert (2021), I designed this project as a detailed examination of community response to nutrients from a species perspective. Nutrient deposition is an abiotic condition that significantly contributes to biotic interactions, as well as community structure and function, and is increasing with climate change and anthropogenic influence. Seitzinger et al. (2002) suggest that rates of nutrient inputs are expected to significantly increase in coastal systems by 2050, making it critical to better understand the affects these additions have on community structure, composition, organization, and function on barrier islands. In this project I will show how increased nutrient deposition on a coastal mesic grassland impacts graminoid and forb productivity differently. Furthermore, I will show limited effects on traditional community structure metrics (species diversity, richness, and evenness). However, despite the limited impacts of nutrients on community structure, nutrients can still significantly alter community composition. Such community differences are driven by increased species dominance, significant community re-organization, and species loss when N enrichment is present.
​
Read the full story: Brown, Moulton, and Zinnert, PLOS One 2022
Synthesizing vegetation differences in scale dependent stability domain states
Understanding large-scale dynamics between biodiversity and ecosystem function is a critical next step in ecology. For barrier islands, one of the most important ecosystem functions is disturbance response through resistance and/or resilience. Barrier islands have been proposed to exist in different stability domain states that are described by elevation and topographic complexity, which is logically coupled with disturbance response. Disturbance-resisting domains rely on ecological resistance of barrier island dune-swale complexes, while disturbance-reinforcing domains rely on resilience after disturbance. Stability domain states are inherently scale dependent and thus large-scale comparisons can create difficulties in unifying barrier island disturbance response. Furthermore, there is a gap in connecting these stability domain states with vegetation communities. My aims for this project are to highlight scale dependency and detail the propensity islands have to exist in multiple domain states. I also aim to synthesize plant diversity and community composition among barrier islands that vary in broad-scale climatic patterns as well as a priori domain state classification.
​
Full story currently in prep
Climate and Data Science Research
Climate models help scientists explore how the Earth system responds to fluxes in greenhouse gas emissions and predict future climate changes driven by human decisions. Some models are highly detailed and require significant computing power and time to run, while simpler models allow for faster analysis due to reduced complexity. These reduced complexity models are important for testing global scale experiments, estimating parametric uncertainty, and understanding how different factors shape future climate outcomes. By using tools like probabilistic simple climate models researchers can better assess the range of possible warming trajectories and their potential impacts.
Matilda: An R package for probabilistic climate projections using a reduced complexity climate model
Matilda is an open-source R package built to streamline probabilistic climate projections using the Hector simple climate model. Many Earth system models are computationally expensive, limiting their use in large-scale uncertainty analysis. Reduced complexity climate models like Hector operate at a global scale and annual time step, making them well-suited for rapidly exploring climate uncertainty. However, existing workflows often require significant programming expertise to generate large ensembles of model runs, evaluate their realism, and extract meaningful insights. Matilda was designed with the intent to simplify this process in an open-source and flexible way. Matilda generates ensembles of climate simulations, systematically by varying key parameters to explore a range of plausible outcomes. These ensembles are then evaluated against historical data to identify parameter sets that produce more realistic projections. As a result, Matilda provides a fast and accessible framework for estimating future temperature changes and exploring how different assumptions, uncertainties, and scenarios shape climate projections.​
​
Read the full story: Brown et al., PLOS Climate 2024​
Code availability here.
​
The effect of different climate sensitivity priors on projected climate: A probabilistic analysis
Understanding how much Earth's temperature may change in response to increasing atmospheric CO2 concentrations is essential for understanding outputs of earth system models. A key component of this process is equilibrium climate sensitivity (ECS), which estimates how temperature changes respond to a doubling of atmospheric CO2 levels. Recent earth system models have shown a wider range of ECS values than before, increasing uncertainty about the future. In this study, we used a simple climate model to investigate how different constraints on ECS uncertainty shapes the uncertainty in end-of-century temperature projections. Our findings show that excluding certain lines of evidence (process and paleoclimate evidence) results in more uncertain temperature projections. This occurs because leaving out critical lines of evidence increases the likelihood of ECS values falling at either ends of the distribution. Therefore, including this evidence is critical for constraining ECS distributions and reducing the uncertainty of temperature projections. Our results also show by how much inducing all information at our disposal to constrain ECS can narrow the uncertainty of future temperature projections. Simple climate models with a probabilistic framework offer a fast, interpretable way to test how ECS distributions impact climate projections.​
​
Full story currently in revision​
​
