The persistence of drought influencing and being influenced in California have had a profound impact on the natural ecosystems. Recent drought conditions have been particularly intense and appear to be increasing in frequency (Diffenbaugh et al. 2015, Griffin and Anchukaitis 2014). In fact, the megadrought experienced in the southwestern United States for the past 22 years have shown that central and southern California experienced the most severe drought over twenty years than any in the past 1200 (Griffin and Anchukaitis 2014, Williams et al. 2022). Because lack of water can put profound stress on the ecosystems, other sources of water such as fog can therefore be crucial in studying the dynamics of water and energy in plant communities (Griffith et al. 2016).

The timing and duration of fog conditions shifting could potentially have negative ramifications as not only does fog lessen the short-wave radiation during daylight, but it also moderates the temperature in a greenhouse effect. In California, the two dominant fog types, coastal and tule influence community structure; I evaluated their frequency and contributions in several different California ecoregions. Fog was shown to peak in the early morning around 08:00 and to be lowest in mid-afternoon around 14:00. The Californian coast had the most incidents of fog during the summer while inland saw tule fog during the winter. Current fog frequencies are highly variable and dependent on locations, yet, with the exception of the northern coast, they have shown to be decreasing in nearly all of California.

As fog can ameliorate water stress on plants, it was unsurprising that plant species can take advantage of fog occurrence. I focused on water partitioning with fog and comparison of biomass accumulation in different-aged stands of two common, co-occurring species in semi-arid shrubland of California, chamise (Adenostoma fasciculatum) and redshank (A. sparsifolium). When fog was present, I found that chamise, the species with the greater topsoil fine root biomass would take advantage, especially in winter and fall. Also, even in some droughts, these species can be a significant sink of carbon, sequestering it in aboveground biomass. The older the stand, the more potential for sequestration of carbon in the shoots and litter and in both the dead and live material. The results from these studies can reframe the hydrological impact on ecoregions of changing fog frequencies and inform decision-makers to prepare for a more drought-prevalent California.

Terrestrial carbon exchange is a critical and variable sink for atmospheric carbon dioxide, mitigating approximately 30% of anthropogenic CO¬2 emissions annually. The interannual dynamics of these exchanges are driven by meteorological and disturbance factors that could be exacerbated by changing climate. Semi-arid ecosystems in particular demonstrate the largest contribution to global terrestrial exchange variability. In Southern California, chaparral is an evergreen, semi-arid shrubland community that may have substantial interannual carbon exchange variability. This study addresses gaps in our understanding of chaparral carbon cycling, including the changing roles of various meteorological conditions, the accuracy of current modeling efforts, and the long-term source-sink dynamics of interannual exchange. Experimental design included multiple scales of carbon exchange measurements (from leaf-level to stand-level exchange) and remote sensing (from 3cm to 1km resolutions) to assess these changes over multiple time scales. Overall, vapor pressure deficit (VPD) was found to drive photosynthetic rate of sampled resprouting species, contrasting the historical prioritization of soil moisture. VPD was also found to drive errors in the MOD17 carbon exchange model for chaparral systems, resulting in a 30% underestimation of gross primary productivity (GPP) during the sampling period. Over the past two decades, Net Ecosystem exchange (NEE) in Southern California shrublands strongly correlated with precipitation, mediated by GPP. The average compensation point, where source/sink dynamics shifted was 128 mm precipitation year-1, ranging between 70 – 240 mm year-1 depending on community composition, previous disturbance, and stand age. This study works towards accurate projections and estimates of carbon exchange in chaparral communities, which is cited as a major component of net-zero strategies at global, state, and local levels.

Chaparral ecosystems (Adenostoma fasciculatum) are the most extensive native plant community in California and are a biodiversity hotspot for native fauna and flora. Undisturbed chaparral ecosystems have been reported to be significant carbon sinks under normal weather conditions (Luo et al., 2007). With climate models predicting an increase in fire and drought frequency in semi-arid regions due to climate change, carbon storage by chaparral ecosystems is likely to be affected by these changes (Storey et al., 2020). While attention to carbon source-sink dynamics in chaparral ecosystems has gained more interest over the recent decade, understanding the controls and patterns of long-term CO2 flux measurements through wildfire and drought events is still poorly understood. Here, I present long-term CO2 flux measurements from three varying-aged chaparral ecosystems to better understand the temporal patterns and controls on CO2 flux and the effects of stand age on carbon sequestration. Through this work, I found that under current climatic conditions, chaparral ecosystems can take up to a decade or more to sequester carbon following a fire and exposure to abnormal rainfall events as the investigated once burned ~20-year-old chaparral ecosystem (US-SO2) acted as a carbon source emitting up to 848 gCm-2yr-1 for eight years post-fire. The stand reverted back to a carbon sink sequestering -69 to -343 gCm-2yr-1 eleven years post-fire. Contrary to the claim that old-growth ecosystems are in a carbon-neutral state, I found that the investigated old-growth 178-year-old chaparral ecosystem (US-SO4) was a source of CO2 to the atmosphere releasing up to 447 gCm-2yr-1 (Odum, 1969; Salati and Vose, 1984; Tan et al., 2011). The meteorological controls of seasonal NEE in a twice burned ~20-year-old chaparral ecosystem (US-SO3) were soil temperature and relative humidity. Monthly NEE was driven by variations in soil temperature, net radiation, and relative humidity. US-SO3 released 45 gCm-2yr-1 to 830 gCm-2yr-1 for nine of the fifteen-year study period and for five years sequestered -14 gCm-2yr-1 to -1003 gCm-2yr-1. Understanding the controls and patterns of long-term CO2 flux in chaparral ecosystems through fire and drought is needed to assess the role of chaparral ecosystems in reducing atmospheric CO2.

The regional warming of high latitude ecosystems, such as the Arctic tundra, is occurring at an accelerated rate relative to the global average due to a considerable number of positive feedbacks. These ecosystems contain one of the largest terrestrial reservoirs of carbon and nitrogen, locked in the soil column by low temperatures and slow decomposition. However, warming temperatures are increasing emissions of primary greenhouse gases (GHG): carbon dioxide (CO2), methane (CH4), and nitrous oxide (N2O) from these immense Arctic reservoirs. Increased release of greenhouse gases from permafrost dominated regions is providing a strong positive feedback on climate warming.While attention to GHG dynamics in permafrost regions has increased over the past decade, understanding patterns and controls on GHG emissions due to seasonality and spatial heterogeneity are still poorly understood. This is particularly true for non-CO2 greenhouse gas fluxes CH4 and N2O which are still understudied. Here, I use a combination of data from a network of eddy covariance towers, soil carbon measurements, and chamber flux measurements on the Arctic Coastal Plain to better understand the seasonal and spatial controls on GHG emissions from the Alaskan Arctic. I show that the timing and magnitude of carbon (CO2 & CH4) fluxes can be highly variable (77-107 g C-CO2-eq m-2 year-1) based on mesoscale (<1 km) hydrological status. While fall CH4 emissions comprise up to 45% of the annual regional CH4 budget, there has been uncertainty as to whether emissions of CH4 during the fall are the product of active methanogenesis or the release of stored methane generated during the prior growing season. Here, I show that fall methane emissions (1100±50 mg C-CH4 m-2) outweigh methane storage within the soil active layer (106.8±7.9 mg C-CH4 m-2 - 35cm depth). These results indicate that the dominant source of CH4 emissions during the cold period is likely microbial activity rather than the release of stored methane. Finally, while emissions of N2O have been thought negligible in Arctic wetlands due to low nitrogen mineralization and high plant-microbe competition for inorganic nitrogen, I show that features of the landscape that are collapsing due to ice wedge and permafrost degradation are important N2O sources, as high as 38.6 mg N m-2 d-1, more than an order of magnitude higher than previously assumed. This understanding of trends, budgets, and drivers of GHG fluxes in the Arctic is intended to help increase accuracy in regional and global climate model projections.