Sensors for Monitoring Harmful Algae, Cyanobacteria and Their Toxins

19 (i.e. different people), highlighting the need for comprehensive metadata. Modeling output errors need to be factored in as well. Overall, the community needs to agree on a consensus for standardization procedures for all platforms and assays . There was a recommendation to approach an organization like IOOS for funding assistance to produce standards where none exist and/or provide them as an open resource. Universal QA/QC is not likely feasible – there will never be a ‘one-size fits all’ solution, but there can be common guidelines for the approach: 1) data prioritization beyond IOOS core variables and resultant data products need to be regional in scope, and 2) introduction of new analytical tools need to be supported by in-field comparisons to existing methods. Further, the research community needs to communicate variability (uncertainty in analytical technique versus inherent biology based variation) in assays / platforms without confusing or discouraging stakeholders as to their efficacy, and work on building trust between all groups. • Q4: What contextual data is required for interpretation of HAB detection patterns and alerts? Complementary ‘contexual sensors’ can be used alongside other technologies to devise measurements for temperature, pH, dissolved oxygen, salinity, nutrients, etc. in order to provide a more holistic picture of the environment in which species and toxins are detected. This approach is routinely being used to trigger ESP sampling (e.g. chlorophyll concentrations; increased salinity + decrease in temperature) since the platform has a limited number of possible sampling events per deployment. At the larger systems level, there is a need for multiple observing nodes coupled to circulation models so we better understand water masses (especially estuarine waters). When developing a model there is a desire to garner as much contextual information as possible but this needs to balance with monitoring efforts. Multi-faceted networks generate large amounts of metadata, and this requires consistency so that data can be pooled. The community needs to continue to work towards generating comparable data across these large scales. • Q5: What end products are needed/desired? The community needs to be able to deliver data products that are usable, particularly for management and modeling applications. This includes being able to equate cell counts with ug toxin/kg mussel tissue. There was a suggestion to work through IOOS Regional Associations as a pilot to develop this process. Also, encourage IOOS to ingest/disseminate core in situ observations for remote sensing. To that end - Can we make enough measurements? How many are desired versus how many are feasible? At some point, the number of measurements and amount of ground- truthing can support a shift in methodologies used. For example, there have been ample places to establish confidence in in situ toxin testing to not need direct assessment of shellfish, an example of regulation based on more low-cost in situ measurements. To arrive at the balance between monitoring needs and cost feasibility, we need to determine which stakeholders care about a given technology. The community can facilitate this by being poised to respond to events in the news such as the HAB bloom in Ohio or the melamine