Automated analysis of temperature variance to determine inundation state of wetlands

摘要: Monitoring the inundation state (wet or dry) of wetlands is critical to understanding aquatic community structure but can be costly and labor-intensive. We tested ability temperature data from cost-effective iButton loggers reflect in central Missouri, based on our hypothesis that dry ponds would show greater daily variance than remained inundated with water. evaluated this method two experiments large outdoor mesocosms, existing natural which we had deployed iButtons. True pond visits was compared predicted over different thresholds expected delineate wet ponds. confirmed variances iButtons were higher under water, as expected, influenced by factors such canopy cover. also describe an automated procedure used determine whether a 80 % accuracy. Using approach, changes state, number days dry, drying filling events calculated. Several caveats are provided should considered prior using maximize accuracy assessing state.