Quantification of marine snow fragmentation by swimming euphausiids

Sinking of marine snow is a major mechanism of particulate carbon transport from surface waters to the seafloor. Any process altering the abundance or size of marine snow influences carbon flux and food availability to pelagic and benthic organisms. We explored whether zooplankton can alter carbon transport by a new mechanism—physical fragmentation of marine snow. The fluid stress created around the appendages of swimming Euphausia pacifica is capable of fragmenting a single aggregate into multiple, smaller aggregates. The reduced size and slower sinking rate of the daughter aggregates may increase their residence time in the water column, promoting decomposition and decreasing particle flux to depth. To determine the importance of fragmentation, tethered and free-swimming euphausiids were videotaped in the presence of marine snow representing a range of aggregate strengths, sizes, and ages. Image analysis was used to characterize the particles prior to and following fragmentation and to determine the area of influence around a single euphausiid. Tethered euphausiids pulled in particles from an average of 6.7 mm away, and most aggregates were fragmented in either the region of the rapidly beating pleopods or in the highvelocity jet that forms off the tail. Euphausiids were capable of fragmenting all aggregate types and produced an average of 7.3 daughter particles, with 60% of these daughter particles remaining within the marine snow size class (.0.5 mm). Thus, physical fragmentation by swimming euphausiids increases the abundance of marine snow while decreasing overall marine snow mass. This novel process, which alters particle size structure without loss of total particulate organic carbon (POC), will be important in upwelling regions where euphausiids are the dominant macrozooplankton migrators. Macroscopic organic aggregates .0.5 mm diameter, generically categorized as marine snow, have significance in the ocean as chemically and biologically distinct microhabitats and serve as the primary transporter of surface-derived organic matter to the ocean interior and seafloor (Fowler and Knauer 1986; Asper 1987). Marine snow has diverse origins and variable composition, including phytoplankton (particularly diatoms), discarded mucous feeding structures of certain zooplankton, fecal pellets, and unidentifiable detritus (Alldredge and Silver 1988). Marine snow can comprise as much as 60% of water column particulate organic carbon (POC) (Alldredge 1979) and is often the dominant material collected in sediment traps (Turner 2002). Below the mixed layer, the flux of POC decreases exponentially with depth (Martin et al. 1987). This decrease must arise from processes that either remove marine snow or physically fragment large, sinking particles into smaller, more slowly sinking or neutrally buoyant particles. Karl et al. (1988) described four potential mechanisms responsible for the reduction of particle flux with depth: (1) microbial decomposition, (2) microbially mediated solubilization to dissolved organic matter (DOM), (3) consumption Acknowledgments We thank D. Farrar for help with animal collection; M. Doall and M. Caun for laboratory assistance; and R. Ross, C. Carlson, and two anonymous reviewers for comments on the manuscript. This research was supported by NSF grant OCE-0296101. by zooplankton, and (4) abiotic stress-induced fragmentation. Karl et al. (1988) coined the phrase ‘‘particle decomposition paradox’’ to describe the incongruity between the observed rapid decrease in POC flux with depth and the often slow rate of the four processes thought to account for removal of the sinking particle pool. However, more recent studies have found that both microbially mediated POC removal (Ploug and Grossart 1999, 2000) and particle grazing by invertebrates (Kiørboe 2000), individually, could be adequate to account for the loss of sinking particles. In a modeling study, Ruiz (1997) showed that abiotic fragmentation might also play an important role, since he was able to create observed diel patterns of particle concentration using only daily fluctuations in energy dissipation rate. Dilling and Alldredge (2000) recently suggested a fifth potential mechanism for the decrease in POC flux with depth. In a field study they found that increased euphausiid abundance was correlated with an increase in aggregate abundance but a decrease in average aggregate size, after accounting for all other production and loss mechanisms. They demonstrated that the stress produced around the swimming legs (pleopods) of the euphausiid Euphausia pacifica was sufficient to cause single aggregates in an animal’s vicinity to be fragmented into multiple, smaller particles. These smaller daughter particles should sink more slowly, increasing their residence time in surface waters and the potential for water column remineralization. Thus euphausiids 941 Marine snow fragmentation by euphausiids could potentially reduce particle flux through swimming behavior alone. That zooplankton might fragment particles has been suggested previously (Banse 1990; Lampitt et al. 1990, 1993) and is supported by limited field observations (Steinberg et al. 1997). Euphausiids have global distribution and are among the most dominant macrozooplankton migrators (Mauchline 1980). Locally, E. pacifica often dominate vertical migratory biomass in the California Current (Brooks and Mullin 1983) with daytime depths of about 400 m and nighttime depths of 0–100 m (Brinton 1976). Typical E. pacifica abundance can range from ,1 to 10 animals m23 with occasional swarm densities of up to 10,000 animals m23 (Mackie and Mills 1983). Since all euphausiids have similar morphology with five pairs of rapidly beating pleopods, other euphausiids are also likely to be capable of aggregate fragmentation. Thus, this fragmentation process could have oceanwide importance for our understanding of particle flux. Although Dilling and Alldredge (2000) demonstrated the existence of this new mechanism of altering particle size structure, it has not yet been tested experimentally or extensively quantified. In this study we investigated the fragmentation process using E. pacifica as our model euphausiid. The goal of this laboratory-based study was (1) to determine which particle attributes, including size, composition, strength, and age, most affected fragmentation, and (2) to quantify the size and number of daughter particles produced. The impact of euphausiid size and distance from an aggregate on the resulting daughter particle size spectra was also determined.