Whole seawater was collected from 3 to 5 m depth on research cruises by Niskin bottle or scientific seawater supply in 1998 (described in Trainer et al., 2000); Niskin bottle (NOAA Ship Bell M. Shimada, 2012; 2016; 2018) and via the scientific seawater system aboard the NOAA Ship Miller Freeman (2009) and NOAA Ship Bell M. Shimada (2015, described in McCabe et al. (2016); also 2017; 2019).

Analysis of satellite-derived sea surface temperature for the region of interest utilized the Multi-Scale Ultra-high Resolution Sea Surface Temperature (MURSST) product MUR-JPL-L4-GLOB-v4.1 (JPL MUR MEaSUREs Project, 2015). This is a global, gap-free, gridded, daily 1 km Sea Surface Temperature (SST) dataset created by merging multiple Level-2 satellite SST datasets. For this analysis monthly averages were created by averaging the months September-November 2002–2013 and 2014–2019, and the difference was calculated by subtracting the 2014–2019 average from the 2002–2013 average.

Particulate (cellular) DA was analyzed by filtering 1L seawater collected from surface waters onboard cruises using a bucket or Niskin bottle and analyzed using the [³H]kainic acid-based receptor binding assay for the 1998 cruise (Trainer et al., 2000), or the Biosense or PNW enzyme-linked immunosorbent assay for all other cruises (Eberhart et al., 2012).

DA in Dungeness crab (Metacarcinus magister) was measured using standard high-performance liquid chromatography (Wekell et al., 1994). These data were kindly provided by the OR Department of Agriculture and the California Department of Public Health.

Transition plans for the conversion of HAB Bulletins from research to operations have been developed for many regions of the US (https://coastalscience.noaa.gov/research/stressor-impacts-mitigation/hab-monitoring-system/). Some forecasts are fully operational (e.g., Lake Erie); others are in the process of transition. Successful transition will depend on collaborative approaches to ensure full operational support, including partnerships with the Integrated Ocean Observing System (Anderson et al., 2019), state monitoring organizations such as the Olympic Region Harmful Algal Bloom partnership and health departments (Trainer and Suddleson, 2005) as well as private companies (e.g., https://www.oceanaero.com/) to ensure the availability of advanced technologies to improve forecasts.

Toxigenic Pseudo-nitzschia blooms are known to initiate in highly retentive regions, designated as HAB “hotspots” along the west coast, including the Juan de Fuca eddy, Heceta Bank, Monterey Bay, and Point Conception (Trainer et al., 2001, 2012). However, the 2015 bloom was different. The warm water anomaly was found to be a source of Pseudo-nitzschia that were delivered to the coast after a series of spring storms, where these cells grew rapidly to bloom proportions when fueled by nutrients from upwelled waters close to shore (McCabe et al., 2016; Trainer et al., 2020). Among the locations that showed record levels of cellular (particulate) DA was a site near the California and Oregon border (McCabe et al., 2016). The emergence of this new toxin source region is also evident from both a compilation of historical observations collected from ships of opportunity and shore stations (Figure 1), as well as from the California Harmful Algae Risk Mapping (C-HARM) predictive statistical model (Anderson et al., 2016). Both the observations and the model identified the Humboldt (Trinidad Head) region as particularly toxic during 2015, with maximum total DA (particulate plus dissolved) exceeding 15,000 ng/L and DA in razor clams at a record level of 340 ppm (McClatchie et al., 2016). The C-HARM model continued to predict elevated probabilities of HABs in the region through 2018 (Wells et al., 2017; Thompson et al., 2018).

The coastwide 2015 bloom appears to have “seeded” Pseudo-nitzschia cells to a site near Trinidad, California where high concentrations of DA were measured (Figure 2, red circled area). Cruises of opportunity prior to 2015, including summer cruises in 1998, 2009, and 2012, showed low concentrations of particulate DA near the California/Oregon border (Figure 2), suggesting that either lower abundances or less toxic species of Pseudo-nitzschia were found at this site prior to 2015. During a cruise from 7 to 10 August 1997 (data not shown), low concentrations of particulate DA, ranging from 200 to 500 ng/L, were measured at several stations between the Cape Mendocino and the CA/OR border. However, during the cruise in summer 2015 (McCabe et al., 2016), a more toxic DA “hotspot” was observed between Cape Mendocino, CA and Cape Blanco, OR that remained evident over 3–4 subsequent summers (Figures 2, 3). These data suggest that this site was freshly seeded with the more toxigenic species, P. australis, promoting higher DA toxicity in filter-feeding organisms in the region. In fact, Pseudo-nitzschia measured at this physically-retentive site have impacted Dungeness crab harvest in both CA and OR where frequent DA closures or harvest delays were first observed in 2015 and continued through early 2019 (Table 1), with Humboldt County (Figure 4) the last county in California to open for harvesting in 2015.

Time series data from 2015, 2017, 2018 demonstrate that northern CA crab were first reported to be toxic (above the regulatory limit of 30 ppm DA), with southern and central OR crab becoming toxic about 1 week later (Figure 3). This is consistent with northward transport of toxigenic cells from the CA hotspot site during periods of relaxations or northward wind reversals associated with storm events near the time of the fall transition (Kosro, 1987; Send et al., 1987). A site of high DA between Cape Blanco and Cape Mendocino was not observed in summer 2019 (Figure 3). However, a single station with elevated particulate DA was noted just south of Cape Mendocino.

The northern CA region has been called an “ideal barometer” (Sanford et al., 2019) for documenting geographic range expansion and shifts in nearshore community composition. This region is an important transition zone where the poleward migration of marine species in response to warming events has been documented (Sagarin et al., 1999; Sanford et al., 2019). During 2014–2016, alongshore currents exhibited anomalously higher poleward flows which were estimated to have transported plankton over maximum distances of ~500 km in a month (Sanford et al., 2019). Plankton and marine larvae were transported northward over such large distances, with fewer reversals to facilitate their return southward (Sanford et al., 2019), allowing them to be established as new populations in more northern ecosystems. This diffusive transport was enhanced by retentive features or offshore mesoscale eddies especially after the spring transition when mean flow in the California Current System typically is equatorward. The intensified upwelling at coastal headlands (Davis, 1985; Kelly, 1985), including Cape Blanco (Barth et al., 2000) and Cape Mendocino (Largier et al., 1993) provides nutrients to fuel phytoplankton blooms, including HABs (Kudela et al., 2005). Areas on the lee side of coastal headlands are often associated with retentive eddies (Largier et al., 1993; Barth et al., 2000). To summarize, the anomalous poleward flows, such as those observed during 2014–2016, resulted in a higher probability of northward transport of particles (Sanford et al., 2019) such as phytoplankton, and their increased retention near coastal headlands. Indeed, P. australis, a highly toxic species typically observed only as far north in the springtime as central California, was detected in northern CA, OR and WA in the spring 2015 (McCabe et al., 2016). This provided an opportunity for this high DA-producing species to be “seeded” to the retentive site near the CA/OR border for future initiation of Pseudo-nitzschia blooms. During the 2015 event, P. australis emerged from a background of multiple Pseudo-nitzschia species, and dominated along almost the entire U.S. west coast, whereas during other years, less toxic species such as P. delicatissima have dominated, resulting in large blooms but with lower toxicity (Bowers et al., 2018; Smith et al., 2018).

Pseudo-nitzschia species are a major component of marine snow due to their ability to form long chains that can become intertwined while they sink to depths. This shower of organic material that cascades from surface waters to depth can transport DA to the benthos where it has been measured in benthic organisms, including bottom dwelling fish, worms, crustaceans (Vigilant and Silver, 2005, 2007; Kvitek et al., 2008) and Dungeness crab (Taylor, 1993). Toxigenic Pseudo-nitzschia and their inherent DA can sink to as deep as 800 m, providing a source of toxin to benthic filter feeders (Schnetzer et al., 2007; Sekula-Wood et al., 2009, 2011). In the Gulf of Mexico, ~50% of Pseudo-nitzschia cells present in the water column sank into sediment traps with ~10% as live cells (Dortch et al., 1997). Pseudo-nitzschia found in sediments were intact, contained pigment and maintained their chain formation (Sekula-Wood et al., 2009). The flux of Pseudo-nitzschia to sediments can be as high as 3.2 × 10⁸ cells/m⁻²/day (Sekula-Wood et al., 2011). More than 80% of DA contained in a surface bloom is estimated to reach sediments >500 m deep (Schnetzer et al., 2017) and previous studies off OR have demonstrated that floc collected from the bottom boundary layer can seed a surface bloom of diatoms, including Pseudo-nitzschia, under simulated upwelling conditions (Wetz et al., 2004). Furthermore, sediment trap data suggests that once Pseudo-nitzschia are seeded into a retentive region, blooms will be frequent and increasing in intensity. A 15-year time series of sediment from the Santa Barbara Channel indicates that this region was seeded with Pseudo-nitzschia in the mid-1990s with increasingly toxic blooms occurring annually (Sekula-Wood et al., 2011). Therefore, the surficial sediment is a source of Pseudo-nitzschia and DA available to initiate the next bloom during upwelling conditions (Seegers et al., 2015).

Dungeness crab fishing generates the most income and has the highest vessel participation of all fisheries on the U.S. west coast (Fuller et al., 2017). The harvest yield of the Dungeness crab fishery and the reliability of the timely opening of crab season determines the livelihood of many fishers in rural communities (Ritzman et al., 2018). The Daily Astorian newspaper stated on 23 April 2018, that “HABs have complicated commercial Dungeness crab harvest on the OR coast for the past 3 seasons, threatening the viability of the state's most valuable fishery.” The commercial season in northern CA, OR, and WA typically opens December 1 and ends July 15. Unfortunately, after the delay of the crab opener until 15 January 2019 due to high levels of DA, the opportunity for CA crabbers was further limited by statewide fishery closure on April 15, almost 3 months earlier than normal, as stipulated under the terms of a settlement agreement reached between the California Department of Fish and Wildlife (CDFW) and the Centers for Biological Diversity (CBD) to limit whale entanglements in crab nets. The agreement requires CDFW to monitor areas off the central and northern CA coasts where whales feed along their spring migration route (Dillman, 2019). If more than 20 whales are observed in an area, fishery managers must order an immediate closure. Thus, the compression of the crab season by DA events, exacerbated by anomalous ocean warming, will have long-lasting impacts on the accessibility to this fishery and the resulting economic benefits to many rural communities. Small vessel fishers may trade their fishing effort disproportionately to access other fisheries when the crab season is canceled, causing concern for resource managers (Saez et al., 2020). Climate change and related warming have compressed habitats for many organisms (Santora et al., 2020). Both habitat compression and intensified Pseudo-nitzschia HABs with range expansion are a result of anomalously warm temperatures in the coastal ocean. In the future, the predicted increased intensity of these HABs (McCabe et al., 2016; McKibben et al., 2017) will reduce fishing opportunity directly due to the presence of toxins in seafood or indirectly through ecosystem impacts such as whale entanglement which may shorten the crab harvest season. Newly seeded HAB retentive sites, resulting from anthropogenic climate change, will challenge managers to ensure the safety of our seafood while having cascading impacts on the health of marine animals and economic welfare of coastal communities.

Early warning systems that anticipate HAB risk at the new hotspot site in northern CA as well as the other known hotspot sites on the US west coast include the Pacific Northwest (PNW) HAB Bulletin (Figure 5; http://depts.washington.edu/orhab/pnw-hab-bulletin/) and the California HAB Bulletin (https://sccoos.org/california-hab-bulletin/). Both Bulletins synthesize biological data from coastal and offshore monitoring of HABs and their toxins, a suite of physical environmental data, and output from realistic numerical models, including the LiveOcean model (https://faculty.washington.edu/pmacc/LO/LiveOcean.html) and the California Regional Ocean Modeling System (Anderson et al., 2019), to forecast potential HAB events for coastal managers. The PNW HAB Bulletin is provided to managers prior to razor clam and Dungeness crab openers and can be helpful in pinpointing areas of the coast that are high risk for HABs. When available, data collected from advanced technologies, such as the Environmental Sample Processor (ESP, Doucette et al., 2009; Scholin et al., 2009; Anderson et al., 2019) and autonomous ocean-going vessels (e.g., http://www.nanoos.org/news/index.php?item=Submaran180924) are also implemented in the forecasting Bulletins. While regionally specific information is utilized in each Bulletin, the PNW HAB Bulletin also tracks larger scale indices such as the Pacific Decadal Oscillation and El Niño, which have been shown to correlate with DA events throughout the PNW (McCabe et al., 2016; McKibben et al., 2017). In the case of the PNW HAB Bulletin, a summary narrative and a HAB risk assessment (low, medium, high; shown as “traffic light” symbols) are provided as a synthesis of results. The early warnings provided by the Bulletins enable mitigative actions such as selective opening of safe shellfish harvest zones, increased harvest limits during low risk periods, and early harvest in anticipation of upcoming toxin outbreaks. The timely and spatially-refined identification of toxin risk will allow managers to define areas of targeted closure, making these early warning Bulletins potentially more cost-effective than other mitigation methods (Holland and Leonard, 2020). In cases when predictions indicate an increasing risk for toxins, managers often modify workflows in order to collect additional shellfish samples (Dungeness crab, clams, and mussels) to ensure human and marine animal safety.

The new northern CA hotspot has been a repeated feature since the west coast wide 2015 DA event (McCabe et al., 2016), often leading to shellfish harvest closures in southern OR. The higher concentrations of DA in crab in southern OR compared to northern OR in late 2015, 2017, and 2018 (Figure 3), suggest that DA originated from northern CA (Figure 3, compare crab DA in CA North to OR South and OR North). By contrast, in late 2016, lower concentrations of DA in northern CA crab compared to OR crab suggest that toxigenic Pseudo-nitzschia originated from a northern source. Measurement of DA in razor clams provided additional insights regarding the geographical source of toxin. In fact, this northern source of DA in 2016 is supported by the cancellation of razor clam digs in southern WA (https://www.dailyastorian.com/news/local/washington-approves-razor-clam-digs-at-two-beaches-long-beach-still-closed/article_ad972fac-0cd4-5059-a845-fb2b25bb5818.html) and all of OR (DA in razor clams was above the regulatory level of 20 ppm in September and continued to increase, reaching a maximum of 360 ppm on the central OR coast in Dec 2016 [OR Department of Agriculture Food Safety data]; https://www.seattletimes.com/sports/entire-oregon-coast-closed-for-razor-clam-digging-after-marine-toxin-levels-skyrocket/ in the fall due to high DA.

The strong influence of the northern CA hotspot site on OR shellfish safety was documented in several PNW HAB Bulletins that prominently featured the fact that southern OR razor clams, mussels, and crabs all contained elevated DA during fall 2018 (e.g., Figure 3). In September 2018, razor clam DA concentrations as high as 180 ppm were measured at Gold Beach, OR, well above the 20 ppm regulatory limit for that species (in contrast to Dungeness crab which have a 30 ppm regulatory limit). Mussels collected off Humboldt, CA, were similarly over the regulatory limit at that same time, suggesting an ongoing toxin event spanning the northern CA/southern OR hotspot. As discussed in the 04-Oct-2018 PNW HAB Bulletin (Figure 5; http://www.nanoos.org/products/habs/forecasts/bulletins.php), scanning electron microscopy of Pseudo-nitzschia samples collected in late September, confirmed the presence of P. australis off both Bandon, OR, and Humboldt, CA. A resurgence of razor clam DA concentrations (from 39 to 140 ppm) at Gold Beach, OR, as well as elevated DA in crab viscera (40 ppm) in early May 2019 indicated the continued risk posed by this new hotspot, even though high particulate DA was not measured in this region in the few samples collected onboard the July 2019 cruise. However, DA poisoning was again a potential risk in October 2019 as seawater particulate DA was documented at an extremely high 5,005 ng/L at Brookings, OR (19-Oct-2019 PNW HAB Bulletin). Crab closures in CA were reported again in 2020 (CDPH report, 2020) demonstrating the persistence of this toxic feature.

PNW HAB Bulletins are emailed to fisheries managers and posted on the ORHAB and NANOOS websites (http://depts.washington.edu/orhab/pnw-hab-bulletin/ and http://www.nanoos.org/products/habs/forecasts/bulletins.php).

There are plans for HAB forecasts across the entire US west coast region to be coordinated through integration with regional Ocean Observing Systems (Anderson et al., 2019) and through the NOAA West Coast Operational Forecast System (WCOFS) effort, currently in development, which eventually will provide consistent ocean modeling across the entire west coast (https://tidesandcurrents.noaa.gov/ofs/dev/wcofs/wcofs_info.html). This dedicated, persistent, and comprehensive monitoring at key transition sites to quantify ocean changes will provide early warning of HABs, allowing management of important resources to promote the health of ocean ecosystems while assisting with safe harvest of seafood.