No abstract

No abstract

Climate-driven impacts on marine trophic pathways worldwide are compounded by sea-ice loss at northern latitudes. For the Arctic, current information describing food web linkages is fragmented, and there is a need for tools that can describe overarching trophic structure despite limited species-specific data. Here, we tested the ability of a mass-balanced ecosystem model (Ecopath with Ecosim, EwE) to reconstruct the trophic hierarchy of 31 groups, from primary producers to polar bears, in the Canadian Beaufort Sea continental shelf. Trophic level (TL) estimates from EwE were compared with those derived from two nitrogen stable isotope (SI) modelling approaches (SI linear and scaled) to assess EwE accuracy, using a dataset of 642 δ15N observations across 282 taxa. TLs from EwE were strongly, positively related to those from both SI models (R2 > 0.80). EwE performed well (within 0.2 TL) for groups with relatively well-known diets or for taxa characterized by fewer trophic connections (e.g., primary consumers). Performance was worse (> 0.5 TL) for species groups aggregated at coarse taxonomic levels, those with poorly documented diets, and for anadromous fishes. Comparisons with SI models suggested that the scaled approach can overestimate the TL of top predators if ecosystem-specific information is not considered.

In British Columbia (BC), harmful algal blooms (HABs) regularly cause severe economic losses through finfish mortalities and shellfish harvest closures due to toxin accumulation, gill damage, or hypoxia. As there is no routine governmental monitoring of HAB phenomena in BC, HAB variability, and its potential links to environmental drivers are not well understood. Here we present results from a well-managed citizen science program which collected an unprecedented 4 year, high-resolution (∼bi-monthly, ∼80 stations) dataset of harmful algae (HA) concentrations and corresponding physical and chemical properties of seawater throughout the Strait of Georgia (SoG), BC. Analysis of this dataset revealed statistically significant interannual and seasonal relationships between environmental drivers and the most common HA taxa: Rhizosolenia setigera, Dictyocha spp., Alexandrium spp., Heterosigma akashiwo, Chaetoceros convolutus, and C. concavicornis. HABs exhibited significant interannual variations; specifically, no HABs were found during the summer of 2015, blooms of Dictyocha occurred in 2016 and 2017, and dense blooms of Heterosigma and Noctiluca occurred in 2018. In addition, HA prevalence corresponded with negative effects observed in local aquaculture facilities where higher toxins concentrations (causing Paralytic and Diarrhetic Shellfish Poisonings) in shellfish flesh were detected during years with greater abundance of Alexandrium and Dinophysis. Furthermore, salmon mass mortality at fish farms corresponded to years with high concentrations of Heterosigma and Dictyocha. As such, these results highlight the need for long-term data to evaluate the potential role of HA as a stressor on the SoG ecosystem.

The Strait of Georgia, Canada, is an important region for numerous commercially and culturally important species (e.g., herring, salmon, and orcas), yet little is known about the links between lower trophic level (e.g. phytoplankton and zooplankton) phenology due to historical sampling gaps. Here, we present fourteen years (2003-2016) of data linking interannual variability in phytoplankton and zooplankton phenology in the Central Strait of Georgia, BC. Satellite-derived chlorophyll a (Chl a) data were used to calculate spring bloom dynamics (bloom initiation, bloom intensity, and bloom magnitude). Average spring bloom initiation occurred during the last week of March in the Central Strait of Georgia. Bloom initiation occurred in mid-to-late February/early March during “early” Chl a bloom years (2004, 2005, and 2015) whereas initiation did not occur until the end of April during “late” bloom years (2007, 2008). Spring Chl a bloom initiation was significantly correlated with the North Pacific Gyre Oscillation (NPGO; r = 0.75, p < 0.01) and spring sea surface temperature (SST; r = -0.70, p < 0.01); spring blooms occurred earlier during warm years. When all environmental variables were considered together, NPGO best explained variations in spring bloom initiation (Adj R2 = 0.53, p < 0.01) and bloom magnitude (Adj R2 = 0.57, p < 0.01), whereas stratification best explained variations in bloom intensity (Adj R2 = 0.38, p < 0.05). Early Chl a blooms were associated with high crustacean abundance (maximum of > 1000 ind m-3) but low biomass (37.5 mg m-3). Independent of the Chl a data, hierarchical cluster analysis revealed similar groupings of years for crustacean abundance data. Most notably, community composition in cluster Group 2 (2004, 2005, and 2015; early Chl a bloom years), was comprised of a higher proportion of small crustaceans (e.g. non-calanoid copepods) compared to the other cluster groups. To our knowledge, this study provides the first evidence linking early spring Chl a bloom timing to a shift in the crustacean community towards smaller taxa in response to multiple warm events in the Strait of Georgia. Our results show that early Chl a blooms may potentially result in a mismatch between phytoplankton and large energy-rich crustacean zooplankton, with lower abundances of the latter. In contrast, average Chl a bloom years were optimal for large-bodied euphausiids, whereas late Chl a blooms were a match for some crustaceans (e.g., medium calanoid copepods), but not others (e.g., large calanoid copepods and amphipods). We hypothesize that early bloom years may result in poorer feeding conditions for juvenile salmon and other predators in the region.