Thèse sur les matériaux cimentaires en milieux non saturés / Thesis on unsaturated cement materials

abstract: Decay of plant litter represents an enormous pathway for carbon (C) into the atmosphere but our understanding of the mechanisms driving this process is particularly limited in drylands. While microbes are a dominant driver of litter decay in most ecosystems, their significance in drylands is not well understood and abiotic drivers such as photodegradation are commonly perceived to be more important. I assessed the significance of microbes to the decay of plant litter in the Sonoran Desert. I found that the variation in decay among 16 leaf litter types was correlated with microbial respiration rates (i.e. CO2 emission) from litter, and rates were strongly correlated with water-vapor sorption rates of litter. Water-vapor sorption during high-humidity periods activates microbes and subsequent respiration appears to be a significant decay mechanism. I also found that exposure to sunlight accelerated litter decay (i.e. photodegradation) and enhanced subsequent respiration rates of litter. The abundance of bacteria (but not fungi) on the surface of litter exposed to sunlight was strongly correlated with respiration rates, as well as litter decay, implying that exposure to sunlight facilitated activity of surface bacteria which were responsible for faster decay. I also assessed the response of respiration to temperature and moisture content (MC) of litter, as well as the relationship between relative humidity and MC. There was a peak in respiration rates between 35-40oC, and, unexpectedly, rates increased from 55 to 70oC with the highest peak at 70oC, suggesting the presence of thermophilic microbes or heat-tolerant enzymes. Respiration rates increased exponentially with MC, and MC was strongly correlated with relative humidity. I used these relationships, along with litter microclimate and C loss data to estimate the contribution of this pathway to litter C loss over 34 months. Respiration was responsible for 24% of the total C lost from litter – this represents a substantial pathway for C loss, over twice as large as the combination of thermal and photochemical abiotic emission. My findings elucidate two mechanisms that explain why microbial drivers were more significant than commonly assumed: activation of microbes via water-vapor sorption and high respiration rates at high temperatures. / Dissertation/Thesis / Doctoral Dissertation Biology 2020

Drought conditions and wildfires can induce soil water repellency. Precipitation shifts are expected to exacerbate drought and wildfire in regions such as the southeastern United States, making it critical to understand how repellency affects water exchange processes in soil. The objectives of this dissertation were to 1) quantify the water vapor sorption dynamics of two clay minerals in which water repellency was induced; 2) identify if and for how long wildfires in humid hardwood forests induce water repellency, 3) evaluate if organic carbon content and hydrophobic functional groups explain actual and potential soil water repellency; and 4) understand how vertical position (i.e., depth) of water repellent layers affect infiltration processes. To meet these objectives, a laboratory test was first conducted examining water vapor sorption processes in water-repellent clay minerals. Next, a field study occurred in two forests that experienced wildfires in late 2016: Mount Pleasant Wildfire Refuge, Virginia, and Chimney Rock State Park, North Carolina, United States. Measurements include water drop penetration time, soil water content, and tension infiltration. Complimentary laboratory tests quantified potential soil water repellency, soil organic carbon content and hydrophobic functional groups. Results showed that water repellency inhibited water vapor condensation because of altered mineral surface potentials and decreased surface areas. Burned hardwood forest soils presented water repellency for > 1 year, though laboratory measurements presented different trends than in situ measurements. Total organic carbon content and hydrophobic functional groups correlated with soil water repellency measured in the laboratory but not the field. Soil water content was lower in burned than unburned soils, and negatively correlated with water repellency. Water repellency in the surface layers significantly reduced relative water infiltration rates, whereas subsurface water repellency did not, and water repellency persisted longer in sites with surface compared to subsurface water repellency. Finally, while the wildfires increased the occurrence of water repellency, they did not alter the underlying relationship between relative infiltration and surface water repellency. Altogether, this study provided new insight into water repellency effects on water partitioning at soil-atmosphere interfaces, and presented evidence of soil and hydrological changes induced by wildfires in humid hardwood forests. / PHD / Rising temperatures and shifting precipitation patterns that result from global climate change have the potential to induce long-term droughts, which may induce soil water repellency, as can wildfires that become more prevalent and damaging. Water repellency can alter the physical, chemical, and hydraulic properties of soil. These alterations may drive soil erosional processes and increase the mobility of surface-bound pollutants with the potential to reduce water quality and degrade down-gradient aquatic ecosystems. Thus, it is critical to understand how water repellency affects water movement in and through soils. Despite several decades of research towards this topic, some critical questions still remain. For example, how does water repellent soil influence water characteristics in the vapor phase (which is increasingly important under drought conditions)? Do wildfires in humid hardwood forests cause soil water repellency? If so, how long does water repellency persist? Do water repellency measurements using field and laboratory techniques correspond to one another? How does the depth of water repellent soil layer(s) affect water movement? In order to solve this questions, several tests were conducted in both field and laboratory. The field experiments occurred within forested hillslopes that underwent varying degrees of burning during widespread wildfires that affected the Southeastern United States in late 2016. Choosing two forested locations, we measured actual water repellency, soil moisture, and infiltration in burned and unburned sites after wildfire, and took loose samples for laboratory tests. In the lab, we tested potential water repellency on air-dried soil samples, soil organic carbon content and hydrophobic substance percentage. We also conducted water vapor sorption experiments to quantify water vapor exchange in two types of water repellent minerals: kaolinite and montmorillonite. The results showed that water repellency can affect water exchange between the subsurface and the atmosphere, by both limiting water vapor sorption and reducing liquid water infiltration. Soil organic matter and composition correlate well with potential water repellency measured in the laboratory, though less so with actual water repellency measured in the field. Instead, soil water content provided a high and inverse correlation with actual water repellency. Finally, water infiltration rates were influenced by the vertical position (depth) of water repellent layers, with water repellency at the soil surface causing much reduced initial infiltration rates compared to water-repellent layers in the subsurface.