Wetlands self-organize through reciprocal controls between vegetation and hydrology, but external disturbance may disrupt these feedbacks with consequent changes to ecosystem state. Imminent and widespread emerald ash borer (EAB) infestation throughout North America has raised concern over possible ecosystem state shifts in forested wetlands (i.e., to wetter, more herbaceous systems) and loss of forest function, calling for informed landscape-scale management strategies. In this dissertation, I use black ash wetlands as a model system to understand complex ecohydrological dynamics, and I use these dynamics to explain the self-organization of observed patterns in vegetation, hydrology, and microtopographic structure. The combined inferences from the three research chapters strongly implicate black ash trees as autogenic ecosystem engineers, who, through the process of improving their local growing conditions, cause a cascade of environmental changes that result in a unique ecosystem structure. This unique ecosystem structure is under existential threat from the invasive EAB. Through experiment, I show that loss of black ash trees to EAB induces persistent shifts in hydrology that result from reduced evapotranspiration and subsequent changes to water table regime (Chapter 2). These results suggest the potential for catastrophic shifts of black ash wetlands from forested to non-forested, marsh-like states under a do-nothing EAB management scenario. However, research presented here suggests that preemptive management of black ash wetlands can potentially mitigate loss of desirable forested conditions. Forest management to replace black ash with other wetland canopy species may be a slow and steady path towards forest maintenance, and harvesting may facilitate establishment of alternative species. In the case of preemptive harvesting of black ash, I posit that maintenance of microtopographic structure, either through leaving downed woody debris or through physical creation, is paramount to forest recovery. Microtopography in these ecosystems provides crucial relief from anaerobic stress generated by higher water tables, allowing woody species to persist on elevated microsites (e.g., 30 cm above base soil elevation). Moreover, I show that microtopography in black ash wetlands has clear structure and pattern and that its presence arises from self-organizing processes, driven by feedbacks among hydrology, biota, and soils (Chapter 3). I further show that this structured and non-random microtopography has profound influence on biogeochemical processes in black ash wetlands, controlling plant richness and biomass, and soil chemistry gradients (Chapter 4). Based on this work, I propose that structured wetland microtopography is a diagnostic feature of strongly coupled plant-water interactions, and these interactions may be important for ecosystem resilience to disturbance. / Doctor of Philosophy / Plants need water, but not too much nor too little. In wetland ecosystems, plants influence water levels through both water use and their effect on soil surfaces. When wetland plants use water, they take it from the soil, which leads to lowering of water levels and drier soil conditions. In many wetlands, the amount of water that plants take from the soil is a fine-tuned process. Therefore, when disturbances happen to wetland ecosystems, like large-scale tree mortality, major changes can occur to the amount of water in the soil and soils typically become wetter. This change to a wetter ecosystem can persist for long periods, and can affect the types of plants that can live in the wetland. However, plants also affect wetland water levels by engineering the soil around them, essentially lifting themselves to drier conditions. Through this engineering, plants create a mosaic of different habitat types that are important for many organisms and ecological processes. Exactly how plants engineer their environment is still not well understood, but we know that ecosystem engineering by plants is critical to the structure and function of wetlands around the world. Understanding how plants create and maintain their own environmental structures provides a deeper insight into the development of vegetated landscapes and their response to change. This dissertation aims to improve our understanding of ecosystem engineering by plants in forested wetlands so that we may more effectively manage these important natural resources and in turn more accurately predict their response to global change.

Little research has been conducted on the sensitivity of earlywood vessel in ring-porous tree species in response to flooding. The impact of flooding and climate on vessel characteristics in black ash (Fraxinus nigra Marsh.) was studied in the boreal region of Lake Duparquet, northwestern Quebec. In addition to standard tree-ring measurements, numbers and cross-sectional area of earlywood vessels were examined and measured using an image analyzing program. Interestingly, among all Spearman rank correlations between chronologies and hydrologic/climatic variables, the strongest associations were found between earlywood vessel chronologies from floodplain trees and spring river discharge data. High water discharge in the spring was negatively correlated to earlywood vessel area and inversely correlated to number of vessels. The mean earlywood vessel area chronology developed from floodplain trees was found to be the best proxy for high magnitude flood events and it has potential to be used for flood reconstructions.

Little research has been conducted on the sensitivity of earlywood vessel in ring-porous tree species in response to flooding. The impact of flooding and climate on vessel characteristics in black ash (Fraxinus nigra Marsh.) was studied in the boreal region of Lake Duparquet, northwestern Quebec. In addition to standard tree-ring measurements, numbers and cross-sectional area of earlywood vessels were examined and measured using an image analyzing program. Interestingly, among all Spearman rank correlations between chronologies and hydrologic/climatic variables, the strongest associations were found between earlywood vessel chronologies from floodplain trees and spring river discharge data. High water discharge in the spring was negatively correlated to earlywood vessel area and inversely correlated to number of vessels. The mean earlywood vessel area chronology developed from floodplain trees was found to be the best proxy for high magnitude flood events and it has potential to be used for flood reconstructions.