EU network of mesocosms facilities for research on marine and freshwater ecosystems open for global collaboration.

Task 8.2

Task 8.1
Task 8.2

8.2.1: LAMP Sensor System (LampSS) development (Lead: CNRS-MARBEC, Contributors: SYKE, WCL, RF-SENS, Month 1-48)

A set of commercial sensors was developed to mesocosm experiments (e.g. electronic devices, software, etc.) by partner 10 CNRS-MARBEC (former CNRS ECOSYM) during the FP7 European project MESOAQUA. This set of sensors, installed in the LAMP (Lite aquatic Automated Mesocosm Platform), permits monitoring (minutes) physical, chemical and biological parameters at high temporal resolution without manipulation of the mesocosm water, and transmits the data via RF to a central data hub in real time. Hence key information on pelagic food web metabolism are monitored in a non-invasive approach (Mostajir et al. 2013, L&O: Methods 394-409). A separate module installed on the mesocosm platform captures incident photosynthetically active radiation (PAR: 400-700 nm) and ultraviolet A and B radiation (315-380 and 280- 315 nm, respectively), together with meteorological data.

In the AQUACOSM project, the LAMP Sensor System and associated dataloggers will be further developed. This will result in a more robust system with the possibility to include new sensors that can be employed also in extreme conditions. A radio module is coupled, permitting data download as well as remote control of the system to e.g. change the protocol of measurements. Currently the datalogger is installed outside the mesocosm. The system will be miniaturized allowing installation directly inside the mesocosm. The miniaturization of the datalogger assemblage and its vicinity to sensors allow a significant reduction of cable lengths. The LAMP Sensor System will be developed to handle a large number of digital probes (e.g. nutrient probes, light sensors, laser diffraction sensor, etc.), offering modular opportunities and greater flexibility. This permits to adapt the system to mesocosms with different diameter (for example, the new AQUACOSM in WP7).

The current LAMP Sensor pack and data acquisition/transmission system have been successfully tested in the Mediterranean environments during FP7 MESOAQUA project. However, although most of the probes and electronic cards can nominally operate at a minimum temperature of -5°C, the LAMP Sensors and associated electronics have never been tested at temperatures below 0°C. Adaptation of the system to arctic conditions will be prepared and tested. Adequate datalogger and electronics components will be chosen, and heating options tested, in order to maintain constant temperature for correct operation. Finally, the last improvement of the systemconsists on the choice and test of the best power supply (i. e. batteries, solar energy). At present, all data collected by sensors of each mesocosm (including meteorological station) are saved in an independent datalogger (CR1000). Each data logger then send all data to a hub collector (CR3000) that save and transmit them to a remote PC on the vessel or on land. In task 8.2.3. (below), the existing data hub will be integrated in a general data analysis system.

8.2.2: AquaBox (Lead: SYKE, Participants: WCL, RF-Sense, LMU, GEOMAR, CNRS-MARBEC, Month 1-48)

Based largely on the Ferrybox technology that has emerged from previous projects (among them FP7 Jerico), and is currently being further developed at SYKE (H2020 project Jerico-NEXT), we will design a compact and modular flow-through system (AquaBox) capable for high-precision monitoring of mesocosms. AquaBox combines peristaltic pumps (Fig. 8.1) and multichannel valves, flow cells and sequential flow technology, performing autonomous measurements successively from several mesocosms installed in a joint rig. By employing flow-through sensors/analysers and OA software for system control, data retrieval, data QA/QC, and wireless data transfer, the system can perform multiple measurements on each water sample, allowing for high precision and reproducibility, as well as semi-continuous measurement frequencies hitherto unobtained in mesocosm studies.

During software development, involving sensor integration, data acquisition, data vocabularies, data transfer, data storage and metadata collection, we closely cooperate with WP4 to conform to EU data policies and enable interaction with related H2020 projects like Jerico-Next and NEXOS.

A modular, readily expandable system structure and control software for AquaBox will be first developed and tested with a basic set of sensors for key properties most prone to temporal variation in planktonic systems during mesocosm experiments (temperature, oxygen, in vivo phytoplankton pigment fluorescence at several wavelengths). Robust solutions for environmental tolerance and off-grid power acquisition will be studied, built (outsourcing/subcontracting), and subjected to field tests. At least 2 such versions of AquaBox will be delivered for field tests in joint WP9 campaigns during the latter stages of the project.

A key development task in 8.2.2 will be the assessment of feasibility and further demonstration of expanding the basic AquaBox structure to integrate additional functions, and measurements requiring advanced analysers, into its physical and control architecture, such as: automated water sample retrieval, inorganic nutrient analyses (miniaturized wet chemistry), Fast Repetition Rate Fluorometry (FRRF) and Pulse Amplification Modulated (PAM) fluorometry as primary productivity proxies, in-water carbonate system chemistry (pCO2, pH), dissolved organic matter (CDOM and FDOM), interphase for gas exchange measurements across water surface, flow cytometry, and image analysis of individual cells of phytoplankton and zooplankton species (FlowCAM, CytoSense, FlowCytobot). Also, the possibility to integrate into an AquaBox system miniaturized, highly replicated manipulative biotests like nutrient limitation bioassays will be addressed and tested. The assessment of and decisions on such extensions, to be executed in Task 8.2.2, will be carried out in Task 8.1, and in connection to the Innovation Forum, where leading manufacturers of respective instrumentation are invited for feedback. It is foreseen that a large share of these analysers will be temporarily obtained within the AQUACOSM partnership for basic functional tests and demonstration. The selected advanced features/analysers, chosen on the basis of the development work and laboratory and field tests, will be acquired and incorporated for a high-end AquaBox version.

Subtask 8.2.3: OA data acquisition system (Lead: WCL, Co-lead: RF-Sense, Contributors: SYKE, CNRS, ENS, Month 25-48)

A data storage system compatible with the instruments developed in subtasks 8.2.1 & 8.2.2 will be developed to provide automated data transfer from the instruments to a central IT interface.

The IT interface serves three purposes. (1) Basic QC and warning function, in case sensors/analytical routines produce erratic data. QC algorithms will be based on comparison to nominal parameter ranges and detection of inconsistencies over time. (2) Filtering algorithms that exclude evident outliers from downstream data analysis (based on quantile statistics & meaningful data ranges). (3) A graphical user interface, based on the free statistical software R (building on R-Studio/Shiny), allowing visualisation, summary statistics and basic statistical analyses (comparisons of means, time trends). For this purpose, the programmable user interface allows definition of treatments & replicates among mesocosms. The default export format will be R binary data, with option to export data as excel spreadsheets (POSIXt time format).

The data analysis system will be designed to work with both sensor systems (8.2.1, 8.2.2). For the Lamp Sensor System, effective handling of high frequency data is required (up to 20 measurements per hr possible). The AquaBox, on the other hand, will produce more complex data esp. in terms of photosynthetic- related parameters (optical quenching, FRRF). A modular setup will allow flexibility in terms of handling different types of data. Meteorological data will be coupled by exact time stamps to data obtained from mesocosms.