Graduation date: 2007Warming of the terrestrial biosphere due to the anthropogenic addition of carbon dioxide to the earth’s atmosphere is becoming a major focus of scientific inquiry. Predictions of the extent of this warming are hampered by uncertainty in the ability of the earth’s ecosystems to counteract this effect by sequestering carbon dioxide by increases in the mass of vegetation, soil storage, or storage in the Earth’s oceans. Measurement of the carbon isotopic composition of respired CO2 (Delta [superscript 13]C[subscript R-eco]) is becoming increasingly important to ecosystem studies because the information contained in this respiration can be an indicator of ecosystem stress and productivity. This study was conducted as part of a larger research project aimed at developing and testing the capacity for measuring Delta [superscript 13]C[subscript R-eco] in a small, steeply-sloped watershed in western Oregon. The goals of this study were: (1) to develop and test an automated system for sampling nocturnal air to be analyzed in a laboratory for isotopic composition; (2) to collect samples of the atmosphere from the nocturnal cold air drainage of a steeply sloped watershed in the HJ Andrews Experimental Forest and bring those samples back to a laboratory to be analyzed for CO2 concentration and carbon isotope composition; and (3) to conduct an initial analysis of the relationship between Delta [superscript 13]C[subscript R-eco] for the first year of deployment and two environmental forcing factors, soil moisture and atmospheric vapor pressure deficit (VPD). The automated system was designed during 2004 and 2005, tested in the laboratory during the month of April 2005, debugged in the field between August 2004 and May 2005, and deployed during the period from late May, 2005 to November 2005. The design objectives for the automated sampling system were: (1) light weight; (2) portability; (3) high reliability; (4) fast dynamic response; (5) unattended operation; and (6) the ability to capture, transport, and store samples over several days with no loss of data integrity. The automated sampling system proved capable of collecting 15 samples per sampling period and utilized a 16 loop stainless steel sample capture valve (Valcon Instruments, 7806 Bobbitt Houston, TX 77055) for sample containment. The system was designed for automated, labor free operation and uses a Campbell CR 10X datalogger (Campbell Scientific, Inc, 815 West 1800 North, Logan Utah) for both data acquisition and system control. This unique architecture was selected to enhance the ability of the system to capture samples which could be analyzed simultaneously for both CO2 concentration and Delta[superscript 13]C. The system was deployed in a steeply sloped watershed and proved to have a combination of light weight (approximately 34 kg total weight) and portability which allowed a wide range of field personnel to deploy the system without undue physical stress. Some initial issues with electrical wiring, plumbing connections, and control program bugs hampered early season performance, but once the system was debugged, it functioned reliably throughout the remainder of the field season with a minimum of operator input. Test results showed that the automated sampler had a maximum sampling frequency of 0.043 hertz, could store air samples for up to 3 days without detectable changes in either isotopic composition or CO2 concentration and displayed greater precision than the hand sampling process it replaced. Using the automated sampling system, sets of nocturnal cold air drainage samples were collected on nine evenings in 2005. The 15 samples acquired for each of these nine evenings were used to generate Keeling plots to determine a single value of the isotopic composition of ecosystem respiration (Delta [superscript 13]C[subscript R-eco]) for each sampling date. The range of Delta [superscript 13]C[subscript R-eco] over the season was 3.9 0/00 and the values of Delta [superscript 13]C[subscript R-eco] varied from -26.2 0/00 on July 13, 2005 to -22.9 0/00 on September 14, 2005. This seasonal pattern of Delta [superscript 13]C[subscript R-eco] was consistent with a forest under drought stress. Trends in ecosystem respiration over the growing season were compared to corresponding measurements of the environmental variables of soil moisture content and VPD. VPD ranged from 2.7 to 1758 Pa, but these patterns of VPD were not significantly correlated to seasonal Delta [superscript 13]C[subscript R-eco] patterns over the study period. Soil moisture content ranged between 7.2% and 44.3% over the study period and soil moisture content temporal patterns were highly correlated with rain events exceeding 4 mm. The soil moisture content pattern for the south-facing slope was significantly correlated with the seasonal Delta [superscript 13]C[subscript R-eco] pattern, although other groupings of soil moisture content were not. The automated system designed for this project met all of its design objectives and functioned adequately throughout the sampling period. The carbon isotope patterns were consistent with a forest under drought stress and the soil moisture content of the plots on the south-facing slopes of the watershed were significantly correlated to these isotope patterns. The future of this system could be enhanced by making adjustments to the supporting hardware, control program and operating procedure to enable larger quantities of samples, more rapid sampling rates, and automated hardware diagnostics.