Cost-effective methods are needed to detect radioactive-waste migration from disposal areas and other potentially contaminated sites. Such methods should be capable of providing early warnings of contaminant releases, and be sufficiently accurate and practical for assessing long-term performance of isolation facilities and remediation measures. This presentation summarizes recent advances in plant-based approaches for detecting and mapping tritium contamination in arid environments. The objective of this work, performed at the U.S. Geological Survey's Amargosa Desert Research Site, adjacent to a low-level radioactive waste (LLRW) burial facility, is to improve understanding of transport processes (http://nevada.usgs.gov/adrs/). The new tritium-detection method entails solar distillation of plant water from foliage, followed by filtration and adsorption of scintillation-interfering constituents on a graphite-based solid-phase-extraction column prior to direct-scintillation counting (Andraski et al., 2003). Tritium concentrations in plant water (creosote bush, Larrea tridentata) determined with the new method matched those determined with traditional, more laborious, toluene-extraction methods. Plant-water concentrations correlated well (r = 0.98) with concentrations in root-zone soil-water vapor. Plant sampling required one-fifth the time of soil-water vapor sampling. The new method provides a simple and cost-effective way to detect plant and soil contamination. Work to date has focused on one desert plant; however, the approach may be transferable to other species and environments. Plant-based techniques were tested for field-scale evaluation of tritium transport (Andraski et al., 2005). Plant-water samples collected within a 63-ha area had tritium concentrations ranging from 2.5 Bq/L (background) to 4,890 Bq/L. Semivariogram analysis showed plant-water concentrations were spatially correlated to a separation distance of 380 m. Spatial structure was highly significant relative to measurement uncertainty, which accounted for <0.1% of the total variability in the data. Linear regression equations that predicted soil-water-vapor concentrations from measured plant concentrations were combined with kriged plant-water concentrations to efficiently map root-zone and sub-root-zone tritium distributions. Maps of subsurface contamination indicate preferential lateral movement of tritium through a dry, coarse-textured layer beneath the root zone, with concurrent upward movement through the root zone. This work represents the first large-scale subsurface vapor-phase tritium migration from a LLRW facility documented in the scientific literature. Analysis of transport from the sub-root-zone gravel to the root zone indicates that diffusive vapor-phase fluxes exceed all other fluxes (i.e., advective vapor-phase, diffusive and advective liquid-phase) by more than an order of magnitude. Phytoremediation is often assumed to be limited to material that is in direct contact with plant roots. Application of the plant-based method at this site, however, showed that the remedial effects of desert vegetation can extend well beneath the root zone. Plants may therefore play an important role not only in detection and monitoring, but also in remediation of tritium-contaminated desert soils. Work is continuing to assess the magnitude and mechanisms of tritium movement from the subsurface to the atmosphere, as well as to track and model the spatial and temporal evolution of the plume. The approach includes evaluation of evaporation and transpiration to better quantify the total tritium releases to the atmosphere. References: Andraski, B.J., M.W. Sandstrom, R.L. Michel, J.C. Radyk, D.A. Stonestrom, M.J. Johnson, and C.J. Mayers. 2003. Simplified method for detecting tritium contamination in plants and soil. J. Environ. Qual. 32: 988-995. Andraski, B.J., D.A. Stonestrom, R.L. Michel, K.J. Halford, and J.C. Radyk. 2005. Plant-based plume-scale mapping of tritium contamination in desert soils. Vadose Zone J. 4: 819-827.