Repository Overview: This repository is a platform for organizing, archiving, and analyzing data associated with the PNNL/SERC TEMPEST Project and also provides a forum for scientific collaboration, discussion, and methods standardization.
Project Overview: The manipulative TEMPEST (Temperate Ecosystem Manipulation to Probe Effects of Storm Treatments) experiment seeks to understand and predict the impact of changing water inundation frequencies, intensities, and qualities on the forms and flows of carbon in coastal terrestrial-aquatic-interface ecosystems by simulating ecosystem-scale freshwater and seawater storm events. Three 50 x 40 m plots serve as control, freshwater, and seawater soil inundation treatments, respectively (Figure 1a). This unreplicated statistical design will be analyzed as a Before-After-Control-Impact experiment and requires significant forethought to the collection of pre-treatment data. Thus, we have installed a wide array of spatially distributed infrastructure for continuous measurements (Figure 1b and d) and we have been conducting coordinated field campaigns since 2018 to generate baseline data streams of site characteristics across the microbe-soil-plant-atmosphere continuum.
Figure 1. Aerial view of the TEMPEST site and experimental design (a); TEMPEST plot schematic displaying the spatial coverage of scientific infrastructure and measurement locations (b); 2020 storm simulation schedule and (c); molecular sample collection dates (d); overview of measurements intended to capture ecosystem-scale responses to storm events across the soil-plant-atmosphere continuum– measures outlined in purple and black are directly tied to this FICUS request.
The TEMPEST experiment will simulate extreme storm events by delivering 80,000 gallons of saline water from the mouth of the adjacent Rhode river and commercially-sourced freshwater into respective treatment plots (2,000 m2), saturating the rooting zone (upper 30 cm) of the forest for 12 hours which equates to a 6-inch rain event. The separate freshwater and seawater plots will allow us to disentangle the impact of salinity versus moisture on belowground ecosystem dynamics. The average monthly precipitation for the area ranges from 2.95 - 4.76 inches and the average yearly precipitation is 47.3 inches (data from Baltimore airport). Thus, a single TEMPEST storm simulation delivers approximately 40-times more water than the daily average precipitation during the wettest month and represents 8% of the average yearly total precipitation. The first two simulation events are scheduled for late summer and fall 2020 (Figure 1c), with subsequent events spaced throughout the growing season and fall of the next decade.
To observe ecosystem- and site-scale characteristics, we are continuously monitoring soil and groundwater geochemistry (dissolved O2, conductivity, volumetric water content) at various depths, hydraulic traits of trees (sap flux, specific leaf area, leaf water potential, stem growth), and gaseous C fluxes (soil respiration, tree stem CH4 and CO2 emissions). Our spatially distributed sensor network (Figure 1b) provides information on soil temperature, moisture content, and conductivity at 15-minute intervals and multiple depths (5, 15, and 30 cm below the soil surface), allowing us to track the progress of storm simulations in real-time and providing insight on molecular sampling locations of ecological interest (for example, during seawater inundation). This systematic characterization of geochemistry and microbial reactivity will enable explicit elucidation of various interactions among biotic and abiotic processes and provide ciritcal information for the development of metabolic process-informed biogeochemical models (Figure 2).
Figure 2. Conceptual image of the overarching experimental goals of the field manipulation experiments.