Thesis (Ph. D.)--Oregon State University, 2007. / Printout. Includes bibliographical references (leaves 195-206). Also available on the World Wide Web.

Agricultural census data have been identified to possess the potential to provide constraints on carbon uptake by croplands at the regional scale. In this study, we build on previous efforts and further assess this potential quantitatively by comparing 1) fractional cropland coverage in southwestern Ontario, Canada derived from agricultural statistics against three different remotely sensed land cover maps; and 2) carbon uptakes determined from agricultural data with simulations generated by a satellite data-driven biospheric model. In addition, we assimilated the census data-derived carbon uptakes with modeled estimates in a Bayesian inverse approach to determine if and by how much constrain the crop data can provide, as exhibited by uncertainty reductions. Uncertainties in census data-derived gross primary production (GPP) estimates are carefully quantified using a Monte Carlo simulation. In general, results from the fractional cropland coverage comparison indicate significant value of the agricultural census data by revealing biases in the spatial distribution of croplands, as found in all three of the satellite land cover products. However, we find that the carbon uptake values derived from crop harvested records are still subject to significant uncertainties that have been underestimated or neglected altogether in past studies. The Monte Carlo simulation suggests that the largest source of uncertainty can be traced to errors in the growth efficiency, followed by harvest production records, and then the harvest index. As result, attention must be paid to such errors when using the agricultural census data for carbon accounting purposes or to provide constraints to simulations of crop carbon uptake.

Thesis (Ph. D.)--Oregon State University, 2010. / Printout. Includes bibliographical references (leaves 116-127). Also available on the World Wide Web.

Thesis (M.S.)--Oregon State University, 2002. / Typescript (photocopy). Includes bibliographical references. Also available on the World Wide Web.

In boreal and northern temperate forests, the moss layer and ectomycorrhizal (ECM) fungi play important roles in carbon and nutrient cycling. ECM mycelium is present in the lower parts of the moss layer, but little is currently known about the interaction between these two key components of northern forest ecosystems. This thesis aims to address this knowledge gap and to improve our understanding of the mechanisms through which the moss layer and ECM fungi influence carbon and nutrient cycling. Nutrient transfer between litter and Scots pine seedlings in symbiosis with the ECM fungus Paxillus involutus (Batsch) Fr. was investigated in highly controlled microcosm experiments using, for the first time, intact moss and pine litter. In addition, moss removal plots were established in a Scots pine forest which allowed measurement of processes involved in carbon (C) and nutrient turnover and related variables. There was a close, reciprocal exchange of carbon and nutrients between the host plant and ECM mycelium colonising moss and pine litter (Chapter 2). This was greatly enhanced by intensive colonisation of moss litter, suggesting that mosses provide a key source of nutrients for ECM fungi and may facilitate transfer of photosynthetic C belowground. During almost 2 years of decomposition, moss tissue released more nitrogen (N) but retained more C than pine litter (Chapter 3), further highlighting the importance of the moss layer in providing nutrients for overstorey trees, and in the accumulation of recalcitrant C in soil. In addition to contributing directly to C cycling through inputs of recalcitrant C in litter, the moss layer can influence C cycling indirectly, by increasing soil microbial activity; CO2 efflux was on average 1.4 times greater from soil under the moss layer than from soil covered only in pine litter (Chapter 3). The results suggest that an indirect influence can occur via two pathways: through an insulating effect of the moss layer on soil temperature, and through inputs of dissolved organic carbon (DOC) leached from moss (Chapter 4), both of which may stimulate activity of soil microbes. These findings demonstrate the importance of both the moss layer and ECM fungi in carbon and nutrient cycling in boreal and northern temperate forests, and indicate that mosses provide a key pathway through which nutrients may bypass sequestration in saprotrophic microbial biomass and be transferred directly from plant tissue to ECM fungi and overstorey trees.

It has been commonly recognized that a series of global glaciation events occurred during the late Neoproterozoic Era (800 - 540 million years ago (Ma)). However, the extent of these glaciations continues to be hotly debated, namely whether the whole Earth was ice covered (ie. a “hard snowball”) or only the continents were fully ice covered but the oceans were not (“slushball/soft snowball”). Through a combination of climate modeling and carbon cycle modeling, I have investigated the plausibility of the “soft snowball” Earth hypothesis. It is demonstrated that the flow of land ice is critical to the formation of a “soft snowball”, such that low latitude land ice must be generated through ice transported from high latitudes. In order for a climate state of this kind to be realizable, continental fragments at low latitude must be well connected to those at high latitude, and the high latitude continents must be sufficiently extensive that a large ice sheet may initiate and subsequently flow to low latitude. It is found that these constraints are satisfied by the most accurate available continental reconstruction for both the initial Sturtian glaciation of the late Neoproterozoic and the subsequent Marinoan event. It is furthermore proposed that the alternative “hard snowball” hypothesis would have been prevented by a negative feedback due to the enhanced remineralization of dissolved organic carbon (DOC) in the ocean due to increased oxygen solubility in seawater at lower temperature. This process would release CO2 to the atmosphere, thus counteracting the initial climate cooling. I have also carried out detailed simulations in which an explicit model of the carbon cycle is coupled to the ice-sheet coupled climate model to investigate this feedback quantitatively. It is found that the remineralization of the DOC does indeed provide a strong negative feedback that counteracts climate cooling. The action of this feedback not only prevents the descent of the climate into a hard snowball state, but also enables the model to re-produce the δ13C carbon isotopic anomalies observed to accompany Neoproterozoic glacial events. The resistance of this carbon cycle coupled climate system to descent into a “hard snowball” state is further tested against stochastic perturbations, and shown to be robust in the presence of such influence.

It has been commonly recognized that a series of global glaciation events occurred during the late Neoproterozoic Era (800 - 540 million years ago (Ma)). However, the extent of these glaciations continues to be hotly debated, namely whether the whole Earth was ice covered (ie. a “hard snowball”) or only the continents were fully ice covered but the oceans were not (“slushball/soft snowball”). Through a combination of climate modeling and carbon cycle modeling, I have investigated the plausibility of the “soft snowball” Earth hypothesis. It is demonstrated that the flow of land ice is critical to the formation of a “soft snowball”, such that low latitude land ice must be generated through ice transported from high latitudes. In order for a climate state of this kind to be realizable, continental fragments at low latitude must be well connected to those at high latitude, and the high latitude continents must be sufficiently extensive that a large ice sheet may initiate and subsequently flow to low latitude. It is found that these constraints are satisfied by the most accurate available continental reconstruction for both the initial Sturtian glaciation of the late Neoproterozoic and the subsequent Marinoan event. It is furthermore proposed that the alternative “hard snowball” hypothesis would have been prevented by a negative feedback due to the enhanced remineralization of dissolved organic carbon (DOC) in the ocean due to increased oxygen solubility in seawater at lower temperature. This process would release CO2 to the atmosphere, thus counteracting the initial climate cooling. I have also carried out detailed simulations in which an explicit model of the carbon cycle is coupled to the ice-sheet coupled climate model to investigate this feedback quantitatively. It is found that the remineralization of the DOC does indeed provide a strong negative feedback that counteracts climate cooling. The action of this feedback not only prevents the descent of the climate into a hard snowball state, but also enables the model to re-produce the δ13C carbon isotopic anomalies observed to accompany Neoproterozoic glacial events. The resistance of this carbon cycle coupled climate system to descent into a “hard snowball” state is further tested against stochastic perturbations, and shown to be robust in the presence of such influence.

Despite the importance of tropical rivers to the global carbon cycle, the nature of carbon cycling within these watersheds has been dealt with by only a handful of studies. The current work attempts to address this lack of information, using stable isotope and concentration measurements to constrain sources and sinks of carbon in two Peninsular Malaysian watersheds. The basins are located on the central-western and northeastern coasts of the Malaysian Peninsula, and are drained by the Langat and Kelantan Rivers, respectively. Water samples were collected from three points along the two rivers twice a month, in addition to the sampling of groundwater in adjacent aquifers. Principal component analyses (PCA) on water chemistry parameters in the Langat and Kelantan Rivers show the dominance of geogenic and anthropogenic influences, grouped in 4 to 6 components that comprise over 50 % of the total dataset variances. The geogenic input is reflected by components showing strong loadings by Ca, Mg, Mn, Si, and Sr, while anthropogenic influences via pollution are indicated via strong loadings by NO3, SO4, K, Zn and Cl. The carbon isotope and concentration data appear unrelated to these groups, suggesting that the riverine carbon cycle in both locations is dominated by other factors. These may include alternative sources of organic pollution, or inputs from the local vegetation and soils. The mean riverine 13CDOC of -27.8 ± 2.9 ‰ and -26.6 ± 2.2 ‰ in the Langat and Kelantan Basins, respectively, are consistent with the dominance of C3-type vegetation in both watersheds. Riverine 13CDIC signatures approach C3-like values at high DIC concentrations, with measurements as low as -19 ‰ in the Kelantan Basin and -20 ‰ observed in the Langat Basin, consistent with a biological origin for riverine DIC. However, the average 13CDIC in river water is 13C-enriched by about 10 ‰ relative to the expected C3 source in both rivers, and this 13C- enrichment appears to be largest with smaller DIC concentrations. Because of the overpressures of CO2 in the rivers, entrainment of isotopically-heavy atmospheric CO2 is not a likely explanation for the observed 13C-enrichment. Theoretically, dissolution of carbonates could be an alternative source of 13C-enriched carbon, but this lithology is scarce, particularly in the Langat watershed. The increase in DIC downstream and generally high pCO2 values in most river sections argues against aquatic photosynthesis as a primary causative factor for the observed isotopic enrichment. This elimination process leaves the speciation of riverine DIC and the evasion of CO2 as the most likely mechanisms for 13C-enrichment in DIC, via isotope fractionation during HCO3- hydration and CO2 diffusion. Potentially, methanogenic activity could also be, at least partially, responsible for the 13C-enrichment in DIC, particularly immediately downstream of the Langat Reservoir, but due to the absence of empirical data, this must remain only a theoretical proposition. The aquatic chemistry and dissolved carbon data suggests that pollution discharge into the Langat and Kelantan Rivers is the major factor that is responsible for the considerable CO2 overpressures and high DIC and DOC concentrations in the river waters, particularly in the downstream sections. This pollution is likely of biological origin, via sewage and palm oil mill effluent (POME) discharge, and therefore isotopically indistinguishable from natural C3 plant sources. Carbon budgets of the Langat and Kelantan River show CO2 degassing to be a significant mechanism of fluvial carbon loss, comprising roughly 50 %, or more, of the total riverine carbon export in both watersheds. The remainder of the river carbon is transported to the ocean in the form of DIC, DOC and POC in broadly comparable proportions. However, the combined riverine carbon export from the Kelantan and Langat Basins amount to 2 % or less of the total carbon sequestration of the watersheds. Thus, most of the sequestered carbon is returned to the atmosphere via respiration, with smaller amounts incorporated into ecosystem biomass . These results highlight the complexity of carbon cycling in tropical rivers, and agree with previous studies in showing riverine systems to be more than simple conduits of carbon from the land to the ocean.

Agricultural census data have been identified to possess the potential to provide constraints on carbon uptake by croplands at the regional scale. In this study, we build on previous efforts and further assess this potential quantitatively by comparing 1) fractional cropland coverage in southwestern Ontario, Canada derived from agricultural statistics against three different remotely sensed land cover maps; and 2) carbon uptakes determined from agricultural data with simulations generated by a satellite data-driven biospheric model. In addition, we assimilated the census data-derived carbon uptakes with modeled estimates in a Bayesian inverse approach to determine if and by how much constrain the crop data can provide, as exhibited by uncertainty reductions. Uncertainties in census data-derived gross primary production (GPP) estimates are carefully quantified using a Monte Carlo simulation. In general, results from the fractional cropland coverage comparison indicate significant value of the agricultural census data by revealing biases in the spatial distribution of croplands, as found in all three of the satellite land cover products. However, we find that the carbon uptake values derived from crop harvested records are still subject to significant uncertainties that have been underestimated or neglected altogether in past studies. The Monte Carlo simulation suggests that the largest source of uncertainty can be traced to errors in the growth efficiency, followed by harvest production records, and then the harvest index. As result, attention must be paid to such errors when using the agricultural census data for carbon accounting purposes or to provide constraints to simulations of crop carbon uptake.

Typescript. / Thesis (Ph. D.)--University of Hawaii at Manoa, 1986. / Bibliography: leaves 119-129. / Photocopy. / viii, 129 leaves, bound ill. 29 cm