Relating cell-level swimming behaviors to vertical population distributions in Heterosigma akashiwo (Raphidophyceae), a harmful alga

Cell motility may facilitate the formation of harmful algal blooms (HABs) by enabling algal cells to swim to favorable microenvironments that support explosive growth. Motility also augments the formation of algal cell aggregations that are often associated with ecological and economic consequence. In this study, we used computerized video analysis to quantify cell-level swimming characteristics by reconstructing cell trajectories in the motile raphidophyte Heterosigma akashiwo, a unicellular alga that forms toxic surface slicks in temperate coastal waters worldwide. Heterosigma cells are capable of rapid changes between at least two active swimming modes, distinguishable by the magnitude of the oscillatory component of motion. Swimming direction varied during a diurnal photoperiod, with swimming direction changing from random to upward directed shortly after the start of the light phase. Motility assays performed 6–8 h into the light phase showed that two Heterosigma strains from geographically distant locations differed significantly in gross swimming speeds, with mean values of 49–66 mm s21 for strain CCMP452 (West Atlantic, USA), and 88–119 mm s21 for strain CCAP934-1 (North Sea, Norway). A spatially explicit model of vertical distribution of Heterosigma cells based on strain-specific motility data suggests that cells of the two strains may diverge in water-column position within a few hours and that CCAP9341 develops dense surface aggregations more rapidly and more robustly than CCMP452. Propensity to form toxic surface slicks, and therefore frequency and severity of HAB impacts, may vary substantially among Heterosigma strains, mediated by differences in cell-level motility. Many harmful algal bloom (HAB)-forming species swim. For example, 29 of the 33 HAB-forming species identified from the west coast of North America are motile (Horner et al. 1997; Horner pers. comm.). For these species, cell-level swimming may constitute an important mechanism that regulates cell abundance and distribution, either directly by concentrating algal cells via interactions of swimming behaviors with ambient flows (Kessler 1985; Franks 1997) or indirectly by enabling cells to locate favorable microenvironments that enhance algal growth rates (Watanabe et al. 1988; Liu et al. 2001). These consequences of motility suggest that quantitative characteristics of swimming behaviors may strongly influence where and when HABs occur. Unfortunately, quantitative swimming characteristics for most HAB-forming algae remain unidentified; thus, the extent to which swimming characteristics vary across relevant ranges of physiological and environmental conditions and across distinct strains within species also remains unknown. This lack of information makes it difficult to develop and test motility-based predictive models of HAB formation or to understand fundamental biological dynamics such as coevolution of algal swimming behavior with other physiological and ecological traits. This study quantifies the swimming characteristics of the raphidophycean biflagellate Heterosigma akashiwo, which forms toxic surface slicks in temperate and subtropical coastal regions worldwide (Smayda 1998). Previous studies show that Heterosigma cells are highly motile and suggest high variability in individual-level swimming characteristics with markedly differing implications for population distributions and HAB formation (Throndsen 1973; Bauerfeind et al. 1986). However, it remains ambiguous whether discrepancies between these studies reflect differences in culture conditions and observation methods or whether different strains of Heterosigma exhibit intrinsically different motility characteristics. Our study sought to quantify swimming behaviors of Heterosigma cells from distinct geographical origins, to establish whether these strains intrinsically differ in quantitative cell-level swimming characteristics under identical conditions, and to assess whether such differences are substantial enough to cause significant interstrain differences in the timing and location of HAB formation. Methods—Heterosigma akashiwo cells, strains CCMP452 (Provasoli Guillard Center for Culture of Marine Phytoplankton) and CCAP934-1 (Culture Collection of Algae and Protozoa), were cultured in artificial seawater medium, O-3 (McIntosh and Cattolico 1978) at 208C and synchronized via a 12 : 12 light : dark (LD) photoperiod. Exponentially growing cells were seen swimming freely in a 30-cm tall by 10cm diameter Plexiglas observation chamber, which matched culture conditions (208C, 12 : 12 LD) except that fluid motions were suppressed with a weak linear salinity gradient in the O-3 medium (salinity 20‰ at the base, 18‰ at the top). Cells were observed with dark field illumination from an infrared light source. Video was captured to a computer at 10 Hz and analyzed to produce a data file of two-dimensional cell trajectories. Each experiment began between the 6th and 8th hours of the light cycle (L6 and L8) (Cattolico et al. 1976), after which cell trajectories were observed for 1 min at 15-min intervals in the first hour, and then hourly for 24 h to assay variations over a diurnal cycle. Like many flagellates, Heterosigma cells swim in a helical pattern, with an overall swimming direction that can be represented as the instantaneous axis of the helix (Crenshaw et al. 2000). Speed and direction along the axis typically vary more slowly in time than the other swimming components,