Phytoplankton Collected by a Time-Series Sediment Trap Deployed in the Southeast Bering Sea during 1997 Stacy Smith and Susan Henrichs University of Alaska Fairbanks Alaska, U.S.A. e-mail: [email protected] Introduction Methods The Bering Sea shelf, over 500 km wide, is sepa- We deployed a time-series sediment trap at 35 m rated into three distinct biological regimes (inner, depth from April 1997 through February 1998. The middle and outer shelf areas) by density fronts. Its deployment site was Mooring 2 (Figure 1) on the middle shelf (located between the 50- and 100-m southeast Bering Sea middle shelf region where the isobaths) averages 30–50 g C m–2 year-1 in new pro- bottom depth was 70 m. The trap was outfitted with duction, derived mainly from the annual spring a rotating carousel containing 12 sample tubes for bloom in April and May. Two unusual events oc- sub-sampling. Trap sub-sample intervals lasted ei- curred over the middle shelf in 1997: The first was ther one or two weeks. The number of phytoplank- an intense though short-lived midsummer bloom of ton in each sample was counted (Figure 2). The mainly small diatoms; the second was the first-ever- d15N and d13C of each sub-sample were measured. recorded coccolithophorid bloom. Anomalous sea Also, a scanning electron microscope was used to surface temperatures - possibly due to El Niño and a identify the coccolithophorid, Emiliania huxleyi relatively storm-free summer – could be the cause (Figure 2). of these events. Figure 1. Location of mooring site 2 in the southeast Bering Sea. (a) (b) Figure 2. Phytoplankton derived from the spring, summer and fall blooms over the Bering Sea middle shelf included diatoms, flagellates and coccolithophores. (a) A silicoflagellate is surrounded by scores of 2-mm Emiliania huxleyi coccoliths. (b) A 400-mm diatom Pleurosigma simonsenii overshadows a Chaetoceros spp. – 15 A high relative number of diatoms in the sediment NO3 . As the nutrient is depleted, the d N of the trap sub-samples (Figure 3) corresponds to phyto- sub-samples increases, eventually to a maximum of plankton blooms over the middle shelf. The late 15.5‰ at nitrate depletion. Over the next two April to mid-May sub-samples represent the end of months, fixed nitrogen is returned to the euphotic the spring bloom, comprised mainly of large dia- zone through microbial breakdown of particulate toms (50–600 mm) such as Coscinodiscus spp. and organic matter and wind mixing of the water col- Pleurosigma simonsenii. Our sediment trap re- umn. As the midsummer bloom resumes, nutrients corded an unusual mid-July bloom, which consisted are quickly depleted and the d15N increases until of a greater number of small diatoms (20–50 mm) algae growth slows. such as Odontella aurita. The 9/22–10/13 sub-samples contain the cocco- In Figure 3 the d15N of each sample reflects lithophorid bloom. Although the bloom was ob- bloom dynamics. The initial d15N value of 12.2‰ served over the middle shelf in July (Napp et al., represents plankton utilizing relatively more new 1998), the trap did not collect numerous coccoliths until autumn. The isotopic signal during these inter- vals changes less dramatically. The plankton are between the heavy isotope, 13C, and the lighter one 13 most likely using more new nitrogen supplied to the they prefer during CO2 uptake. The d C increases euphotic zone, as mixing due to wind events in- as diatom biomass increases in the sub-samples and creases in autumn. decreases as the number of diatoms decreases. In Figure 4 the d13C tracks phytoplankton growth rates. As diatoms grow faster they discriminate less 25 diatoms 16 d15N 15 20 14 15 13 N 15 10 12 d d 11 5 10 Relative number 0 9 6/3 4/22 5/20 6/17 7/15 8/12 9/29 10/27 11/24 12/22 Trap Sampling Dates Figure 3. Temporal change in diatom numbers and d15N of sediment trap samples. 25 -20 diatoms -20.5 20 d13C -21 15 C -21.5 13 10 d d -22 5 Relative number -22.5 0 -23 /3 6 /22 /20 /17 /15 /12 /29 /27 /24 /22 4 5 6 7 8 9 10 11 12 Trap Sampling Dates Figure 4. Temporal change in diatom numbers and d13C of sediment trap samples. Conclusions References During 1997, the composition of material col- Napp, J.M., Baier, C.T., Brodeur, R.T., Cullen, J.J., lected by the sediment trap over the middle shelf Davis, R.F., Becker, M.B., Goering, J.J., Mills, C.E., Schumacher, J.D., Smith, S.L., Stabeno, primarily reflected events affecting primary pro- P.J., Vance, T.C. and Whitledge, T.E. 1998. The ducers. These included spring and summer diatom 1997 eastern Bering Sea shelf-wide coccolitho- blooms and the late autumn sinking of the cocco- phorid bloom: Ecosystem observations and hy- lithophorid bloom. The stable C and N isotopic pothesis. AGU/ASLO Ocean Sciences Meeting, composition of trapped organic matter also appar- ently reflected phytoplankton growth rates. Heavier d15N and d13C were associated with nutri- ent depletion. 1.