Figure 1. Changes in the position of the centroid of latitudinal distribution (i.e., the mean latitude of the occupied cells) across time and RCP scenarios (i.e., RCP-2.6, -4.5, -6.0, and -8.5; CMIP5) for (A) Pseudo-nitzschia australis, (B) P. seriata, and (C) P. fraudulenta. Positions calculated for 2050 (orange circles) and 2100 (green circles). The dark blue, horizontal dashed line represents the centroid position for the present day (i.e., 2000–2014, based on monthly averages).

Figure 1. Changes in the position of the centroid of latitudinal distribution (i.e., the mean latitude of the occupied cells) across time and RCP scenarios (i.e., RCP-2.6, -4.5, -6.0, and -8.5; CMIP5) for (A) Pseudo-nitzschia australis, (B) P. seriata, and (C) P. fraudulenta. Positions calculated for 2050 (orange circles) and 2100 (green circles). The dark blue, horizontal dashed line represents the centroid position for the present day (i.e., 2000–2014, based on monthly averages).

Figure 2. Mean latitudinal habitat suitability temporal changes (i.e., gains in red and losses in blue) between the present day and 2050 (left—(A,C,E,G)) and 2100 (right—(B,D,F,H)) for Pseudo-nitzschia australis across the four representative concentration pathway scenarios (RCP-2.6, -4.5, -6.0, and -8.5; CMIP5). The vertical dashed line at ‘0′ represents the present-day baseline.

Figure 2. Mean latitudinal habitat suitability temporal changes (i.e., gains in red and losses in blue) between the present day and 2050 (left—(A,C,E,G)) and 2100 (right—(B,D,F,H)) for Pseudo-nitzschia australis across the four representative concentration pathway scenarios (RCP-2.6, -4.5, -6.0, and -8.5; CMIP5). The vertical dashed line at ‘0′ represents the present-day baseline.

Figure 3. Projected spatiotemporal changes in distribution for Pseudo-nitzschia australis between the present day, 2050, and 2100 across four representative concentration pathway scenarios: (A) RCP-2.6, (B) RCP-4.5, (C) RCP-6.0, and (D) RCP-8.5. Projected occurrence distribution changes are represented: unidirectional range shifts (i.e., projected expansions in red and orange and projected contractions in dark and light blue) and transitory fluctuations (i.e., range contraction followed by expansion in pink and vice versa in purple).

Figure 3. Projected spatiotemporal changes in distribution for Pseudo-nitzschia australis between the present day, 2050, and 2100 across four representative concentration pathway scenarios: (A) RCP-2.6, (B) RCP-4.5, (C) RCP-6.0, and (D) RCP-8.5. Projected occurrence distribution changes are represented: unidirectional range shifts (i.e., projected expansions in red and orange and projected contractions in dark and light blue) and transitory fluctuations (i.e., range contraction followed by expansion in pink and vice versa in purple).

Figure 4. Mean latitudinal habitat suitability temporal changes (i.e., gains in red and losses in blue) between the present day and 2050 (left—(A,C,E,G)) and 2100 (right—(B,D,F,H)) for Pseudo-nitzschia seriata across the four representative concentration pathway scenarios (RCP-2.6, -4.5, -6.0, and -8.5; CMIP5). The vertical dashed line at ‘0′ represents the present-day baseline.

Figure 4. Mean latitudinal habitat suitability temporal changes (i.e., gains in red and losses in blue) between the present day and 2050 (left—(A,C,E,G)) and 2100 (right—(B,D,F,H)) for Pseudo-nitzschia seriata across the four representative concentration pathway scenarios (RCP-2.6, -4.5, -6.0, and -8.5; CMIP5). The vertical dashed line at ‘0′ represents the present-day baseline.

Figure 5. Projected spatiotemporal changes in distribution for Pseudo-nitzschia seriata between the present day, 2050, and 2100 across four representative concentration pathway scenarios (A) RCP-2.6, (B) RCP-4.5, (C) RCP-6.0, and (D) RCP-8.5. Projected occurrence distribution changes are represented: unidirectional range shifts (i.e., projected expansion in red and orange and projected contractions in dark and light blue) and transitory fluctuations (i.e., range contractions followed by expansions in pink and vice versa in purple).

Figure 5. Projected spatiotemporal changes in distribution for Pseudo-nitzschia seriata between the present day, 2050, and 2100 across four representative concentration pathway scenarios (A) RCP-2.6, (B) RCP-4.5, (C) RCP-6.0, and (D) RCP-8.5. Projected occurrence distribution changes are represented: unidirectional range shifts (i.e., projected expansion in red and orange and projected contractions in dark and light blue) and transitory fluctuations (i.e., range contractions followed by expansions in pink and vice versa in purple).

Figure 6. Mean latitudinal habitat suitability temporal changes (i.e., gains in red and losses in blue) between the present day and and 2050 (left—(A,C,E,G)) and 2100 (right—(B,D,F,H)) for Pseudo-nitzschia fraudulenta across the four representative concentration pathway scenarios (RCP-2.6, -4.5, -6.0, and -8.5; CMIP5). The vertical dashed line at ‘0′ represents the present-day baseline.

Figure 6. Mean latitudinal habitat suitability temporal changes (i.e., gains in red and losses in blue) between the present day and and 2050 (left—(A,C,E,G)) and 2100 (right—(B,D,F,H)) for Pseudo-nitzschia fraudulenta across the four representative concentration pathway scenarios (RCP-2.6, -4.5, -6.0, and -8.5; CMIP5). The vertical dashed line at ‘0′ represents the present-day baseline.

Figure 7. Projected spatiotemporal changes in distribution for Pseudo-nitzschia fraudulenta between the present day, 2050, and 2100 across four representative concentration pathway scenarios (A) RCP-2.6, (B) RCP-4.5, (C) RCP-6.0, and (D) RCP-8.5. Projected occurrence distribution changes are represented: unidirectional range shifts (i.e., projected expansions in red and orange and projected contractions in dark and light blue) and transitory fluctuations (i.e., range contractions followed by expansions in pink and vice versa in purple).

Figure 7. Projected spatiotemporal changes in distribution for Pseudo-nitzschia fraudulenta between the present day, 2050, and 2100 across four representative concentration pathway scenarios (A) RCP-2.6, (B) RCP-4.5, (C) RCP-6.0, and (D) RCP-8.5. Projected occurrence distribution changes are represented: unidirectional range shifts (i.e., projected expansions in red and orange and projected contractions in dark and light blue) and transitory fluctuations (i.e., range contractions followed by expansions in pink and vice versa in purple).

Table 1. Top four most important variables in the ensemble model for each species.

Table 1. Top four most important variables in the ensemble model for each species.

Species#1#2#3#4P. australisBathymetry
71.4%Temperature Maximum
9.7%Temperature Range
5.4%Temperature Minimum
3.4%P. seriataBathymetry
81.9%Temperature Minimum
9.5%Temperature Range
4.7%Temperature Mean
0.9%P. fraudulentaBathymetry
61.9%Salinity Minimum
15.2%Temperature Mean
5.4%Temperature Maximum
4.6% Table 2. Summary list of the ensemble MaxEnt model predictions for each AST-producing species. Accurate, over-, and underpredictions of distribution based on Trainer et al. [27] and Bates et al. [35]. Table 2. Summary list of the ensemble MaxEnt model predictions for each AST-producing species. Accurate, over-, and underpredictions of distribution based on Trainer et al. [27] and Bates et al. [35]. SpeciesAccurate predictionOverpredictionUnderpredictionP. australisW Atlantic (Gulf of Mexico, Argentina, Gulf of Maine, South Brazil)Southern Ocean Bay of Fundy
Gulf of Alaska
Bering StraitN Atlantic (Celt Sea, Ireland, Scotland, Galicia, Portugal, Morocco)Arctic OceanSE Atlantic (Benguela, Namibia)IndonesiaEastern Pacific (Mexico, Baja California, USA, Alaska, Canada, Peru, Chile)Indian Ocean (Oman)Bering Strait and Bering SeaNW Pacific (Okhotsk Sea, Sea of Japan, Eastern China Sea)Oceania (Australia, New Zealand)Mediterranean Sea Atlantic (northern South America)P. seriataNW Pacific (Okhotsk Sea, Sea of Japan, Bering Sea)Southern HemisphereSingaporeNW Atlantic (Gulf of St. Lawrence, Greenland)Gulf of MexicoArctic OceanBlack SeaP. fraudulentaNW Atlantic (USA, Canada)Madagascar
Indonesia
SE Pacific
Red Sea
Arctic OceanMarmara Sea (Turkey)NE Atlantic (Morocco, Celt Sea, Scotland)North SeaMediterranean Sea (NW Mediterranean, Adriatic, Morocco)Gulf of MexicoSW Atlantic (South Brazil, Argentina)NE Pacific (Washington, Gulf of California, Baja California, Mexico, Chile)NW Pacific (Okhotsk Sea, Sea of Japan)Oceania (Australia, New Zealand)Indian Ocean (Pakistan)

Table 3. Precuration, postcuration, and post-environmental-filtering numbers of valid occurrences for each AST-producing species included in the present analysis.

Table 3. Precuration, postcuration, and post-environmental-filtering numbers of valid occurrences for each AST-producing species included in the present analysis.

SpeciesPrecurationCuratedPost-Environmental FilteringPseudo-nitzschia australisA573230Pseudo-nitzschia seriataB1997834782Pseudo-nitzschia fraudulentaC162129124