Sensors for Monitoring Harmful Algae, Cyanobacteria and Their Toxins

8 current limitations in detection capabilities, unknown factors that promote toxin production, and how values translate to public health threats. a. Sensors Myriad challenges were raised concerning sensor technology, many of which translate across platforms. One key aspect, which has been developed into a workshop recommendation, is the storage and transmission of large data files . While eleven agencies within the IOOS framework serve to fill a data hosting need, researchers still must rely on in-house hosting of certain complex data sets such as Flow Cytobot images. This ties into the question of how much data can/should be generated? When is generating species/toxin data more desirable than measurements of biomass (capturing both harmful and non-harmful species)? What type of data are needed for an early warning of an impending event? The desired data type(s) may not match available funding levels. Does the application require Tesla-level technology, or will a ‘Honda Civic’ (i.e. at a more moderate cost) do the job? As several participants pointed out, effective solutions to these overarching questions can best be addressed by engaging end-users of the data products early in the development cycle. Additional gaps for implementation were identified: • Regarding hand-held devices, researchers/managers desire more sensitivity/specificity. • Development of specific molecular assays are needed for each system, for example the Sandwich Hybridization Assay (SHA) can be run on the ESP or a benchtop system, however both involve variations to the protocol and reagents used. • For many sensors, the processes of QCing and intercalibration are intricate, and standardized protocols may or may not be available. Related to this is the long-term challenge of developing ‘gold standard’ reference materials (species and toxin) and performing rigorous testing using local isolates. • Reagents can be proprietary, and often available from only one vendor or research group, which keeps operation costs high and the technology unattainable for many laboratories and/or resource managers. • Balancing the current/future need for a platform conducive to routine monitoring versus HAB event sampling. This affects the type of sensor needed as well as the associated costs and assays available. Related to this is the desire in some cases to validate observations with multiple analytical techniques. • Platforms / deployments have challenges associated with availability of supportive infrastructure, power supply, and readiness level for modularity. • Given costs and complexity, redundancy in deployed instrumentation within a sentinel monitoring structure is often not feasible. For example, if there is one IFCB deployed and it malfunctions, there will be gaps in an otherwise continuous data set unless another instrument is available.