Emerging harmful algal blooms caused by distinct seasonal assemblages of a toxic diatom

Abstract

Diatoms in the Pseudo-nitzschia genus produce the neurotoxin domoic acid. Domoic acid bioaccumulates in shellfish, causing illness in humans and marine animals upon ingestion. In 2017, high domoic acid levels in shellfish meat closed shellfish harvest in Narragansett Bay, Rhode Island for the first and only time in history, although abundant Pseudo-nitzschia have been observed for over 60 years. To investigate whether an environmental factor altered endemic Pseudo-nitzschia physiology or new domoic acid-producing strain(s) were introduced to Narragansett Bay, we conducted weekly sampling from 2017 to 2019 and compared closure samples. Plankton-associated domoic acid was quantified by LC-MS/MS and Pseudo-nitzschia spp. were identified using a taxonomically improved high-throughput rDNA sequencing approach. Comparison with environmental data revealed a detailed understanding of domoic acid dynamics and seasonal multi-species assemblages. Plankton-associated domoic acid was low throughout 2017-2019, but recurred in fall and early summer maxima. Fall domoic acid maxima contained known toxic species as well as a novel Pseudo-nitzschia genotype. Summer domoic acid maxima included fewer species but also known toxin producers. Most 2017 closure samples contained the particularly concerning toxic species, P. australis, which also appeared infrequently during 2017-2019. Recurring Pseudo-nitzschia assemblages were driven by seasonal temperature changes, and plankton-associated domoic acid correlated with low dissolved inorganic nitrogen. Thus, the Narragansett Bay closures were likely caused by both resident assemblages that become toxic depending on nutrient status as well as the episodic introductions of toxic species from oceanographic and climatic shifts.

Fig. 1

Sites within Narragansett Bay, RI, USA and offshore Northeast U.S. Shelf Long‐Term Ecological Research (NES LTER) cruise stations sampled from 2016 to 2019. (A) Sample locations (n = 18) within Narragansett Bay including the Long‐Term Plankton Time Series (NBPTS) at the green point, University of RI Graduate School of Oceanography dock (GD), Whale Rock (WR) at the purple point, East Passage (EP) as the blue point, Castle Hill beach (CHB), and Fort Wetherill (FW), and sites sampled by RI harmful algal bloom (HAB) monitoring in March 2017 (D1–D5) and sites sampled with vertical net tows (A1–A7). (B) Inset shows overview of sampling site locations in North America. The location of the within Narragansett Bay sites are outlined by the black box. Select stations were sampled along the routine NES LTER transect (L1, L3, L4, L7). The figure was made in Ocean Data View (Schlitzer 2002).

Fig. 2

Twenty‐year record of the presence of Pseudo‐nitzschia spp. in Long‐Term Plankton Time Series (NBPTS) samples in Narragansett Bay, RI (A) Live Pseudo‐nitzschia spp. cell abundance observed at the NBPTS site from approximately weekly sampling from January 1999 to November 2019 (n = 1003). Values are log₁₀‐transformed, with +1 cell to all values to create non‐zero values in the transformation. Surface water samples were used in 2000, 2001, and 2008–2019, and mixed surface and depth (~ 7 m) samples were used in 1999 and 2002–2007. No sampling occurred from March 2012 to December 2012. The horizontal dashed line is the log‐transformed RI HAB action threshold of 20,000 cells L⁻¹. The line annotated “this study” is the timeframe in which samples were taken for DNA sequencing and particulate domoic acid analysis. Arrows indicate the 2016 precautionary closure and 2017 closure due to domoic acid (DA), respectively. The color of the points indicates the percentage of total plankton community counts that Pseudo‐nitzschia spp. represent. (B) Data from 20 μm vertical net tow samples (~ 7 m bottom depth) corresponding to count data presented in a; each line represents whether Pseudo‐nitzschia spp. were observed or not (n = 920). Net tow data unavailable from September 2011 to March 2012.

Fig. 3

Comparison of total chlorophyll a (Chl a) concentration, Pseudo‐nitzschia spp. cell counts, particulate domoic acid (pDA), and dissolved inorganic nitrogen (DIN) at the most frequently sampled sites during the study period. The East Passage site was usually sampled in the summer and fall only. For details on species assemblages associated with domoic acid maxima and at the Long‐Term Plankton Time Series (NBPTS) site, please see Figs. 6, S6, respectively. (A) Trends of Chl a across time at the East Passage site (n = 22), the NBPTS site (n = 131), and Whale Rock site (n = 69). Values are log₁₀‐transformed with + 1 added to all values to create only positive values in the transformation. (B) Pseudo‐nitzschia spp. cell counts across time at East Passage (n = 26), NBPTS (n = 141), and Whale Rock (n = 73). The dashed line is the 20,000 cells L⁻¹ action threshold for RI HAB monitoring. Open circles denote the absence of cells. (C) Patterns of pDA measured across time at East Passage (n = 26), NBPTS (n = 131), and Whale Rock (n = 72). Open circles denote no pDA was detected. (D) DIN, which is the summation of nitrite, nitrate, and ammonium measured, across time at East Passage (n = 23), NBPTS (n = 136), and Whale Rock (n = 71).

Fig. 4

Principal component analysis (PCA) biplot of all surface seawater samples from Narragansett Bay, RI during the study period from any sampling site, excluding NES LTER stations, with full portfolio of physical and chemical parameters log‐transformed and standardized. Parameters include surface seawater salinity (Sal), surface seawater temperature (temp), chlorophyll a concentration (Chl a), dissolved inorganic silicate (DSi), dissolved inorganic phosphorus (DIP), nitrate, nitrite, ammonium, and nutrient ratios related to the summation of those nitrogen sources as dissolved inorganic nitrogen (DIN) of DIN : Si and DIN : P, Pseudo‐nitzschia spp. cell abundance (P. cells), and particulate domoic acid (pDA) concentration. (A) Bar chart of the contribution of each variable to dimension 1 and dimension 2 corresponding to the PCA shown in C. (B) Samples colored by the season sampled, from September 2017 to November 2019, excluding Chl a, P. cells, and pDA (n = 312). (C) Samples colored by whether pDA was detected (red) or not detected (gray) by LC‐MS/MS, and including Chl a, P. cells, and pDA (n = 238).

Fig. 5

Nonmetric multidimensional scaling (NMDS) of Jaccard distance matrix of the presence or absence of Pseudo‐nitzschia spp. ASVs in samples from Narragansett Bay, RI and offshore (n = 192). Both plots are the same NMDS with stress = 0.0924. Solution was not reached by defaults, so iterations raised to 100 and dimensions were raised to three but visualized in two dimensions. (A) Pseudo‐nitzschia spp. assemblages across seasons (color) by sample group (shape), including the Narragansett Bay Long‐Term Plankton Time Series (NBPTS) and Northeast Shelf Long‐Term Ecological Research (NES LTER) offshore cruises, which has different sample types and sample years. (B) Assemblages of Pseudo‐nitzschia taxa overlaid with concentration of particulate domoic acid (pDA) measured in filtered seawater L⁻¹ from LC‐MS/MS measurements. Samples where no measurements were taken for LC‐MS/MS are indicated with an x, including net tow, precautionary closure, closure, and some NES LTER samples.

Fig. 6

Pseudo‐nitzschia spp. detected in samples around the 2016 precautionary closure, the 2017 closure, and from the seven most toxic samples during the September 2017 to November 2019 study period in Narragansett Bay, RI. Corresponding particulate domoic acid (pDA) concentrations from September 2017 to November 2019 are shown in the top row. The 2016 precautionary closure (7 October 2016–29 October 2016) and 2017 closure (26 February 2017–24 March 2017) samples included those collected at the Narragansett Bay Long‐Term Plankton Time Series (NBPTS) site and RI DEM HAB monitoring sites (D1–D5). Samples from the study period shown include those collected at NBPTS, Whale Rock (WR), and East Passage (EP) during fall 2017, fall 2018, and summer (sum) of 2019. Distinct ASVs identified as the same species were kept separate and denoted with numbers. A group of ASVs in the data set which could only be identified to the genus level, are labeled as “P. sp. Group1.” Purple shading indicates ASVs which occurred at > 1% relative abundance per sample as present and gray as those not present at < 1% relative abundance or absent in an individual sample. Species known to produce domoic acid according to Bates et al. (2018) are denoted in bold. Particulate domoic acid values (ng L⁻¹ seawater filtered) were 60 (WR—29 September 2017), 34 (WR—4 October 2017), 66 (EP—4 September 2018), 29 (WR—30 May 2019), 26 (EP—30 May 2019), 22 (NBPTS—3 June 2019), and 35 (WR—3 June 2019).