Sedimentation of phytoplankton: role of ambient conditions and life strategies of algae
- First Online:
- Cite this article as:
- Yacobi, Y.Z. & Ostrovsky, I. Hydrobiologia (2012) 698: 111. doi:10.1007/s10750-012-1215-9
- 410 Downloads
Pigment content in particles accumulated in sediment traps are often not directly correlated with phytoplankton abundance, but are rather indicative of transformations phytoplankton underwent on its downward move and following resuspension. We argue that the variability in temporal and spatial sedimentation patterns of different phytoplankton groups is not only an outcome of pigment persistence, but is also associated with dissimilarity in life strategies and dependent on the physical conditions of the water column. Pigment concentrations were measured on weekly–biweekly basis in the water column and in five sets of traps positioned in Lake Kinneret, Israel. Highly degradable peridinin and chlorophyll c reached the deep traps in minute quantities indicating that dinoflagellates mostly recycled in the epilimnion; these migrating algae dominated plankton community under low turbulence and high light. When fast-sinking diatoms persisted in the water column during holomixis they could reach the bottom intact, and fucoxanthin was found in equal proportions in water and traps, chlorophytes rarely dominated phytoplankton, but lutein and chlorophyll b harbored by this group were often the most abundant signature pigments in traps, reflecting the effect of high accumulation rates of these stable compounds in resuspensed particles from the bottom.
KeywordsSedimentation Photosynthetic pigments Phytoplankton Decomposition Resuspension Material transport
Sinking phytoplankton is a major component in the downward flux of organic particles in aquatic ecosystems. The domination of algae in the sedimented particles is conspicuously manifested when algal blooms collapse (Graf et al., 1982). Organic material originating from different phytoplankton phyla is deposited to the bottom at various rates, which depend on the composition of the settling material: intact cells, cell debris, and fecal pellets produced by phytoplankton grazing. Despite the fairly uniform horizontal distributions of phytoplankton biomass in small and medium sized lakes, spatial variability of algal related organic matter in the topmost layer of bottom sediments can be high (Ostrovsky & Yacobi, 1999; Yacobi & Ostrovsky, 2000; Bloesch, 2004), indicating that the deposited material is translocated laterally by resuspension-transportation processes. During holomixis, particles suspended in the water column are entrained by currents and large-scale water circulations (e.g., gyres) and deposited onto the bottom in inverse proportion to the near-bottom shear stress (Bloesh, 1995). When the water body is stratified and the water layers below the thermocline are largely detached from the direct impact of wind stress on the water surface, the internal seiching and water motions in the benthic boundary layer (BBL) produce highly heterogeneous fields of turbulence, near-bottom shear stress, and a complex system of currents in the deep layers. These contribute to the relocation of particles in the BBL and the uppermost sediment layer (Frechette et al., 1989). Although the dynamics of BBLs have been intensively studied during the last decade (e.g., Wüest & Lorke, 2003), the effect of boundary processes in basin-scale fluxes of particulate material is still not fully understood. The fate of deposited and resuspended phytoplankton may be traced using various cellular components, such as diatom frustules (Batterbee et al., 2003), chrysophyte scales (Zeeb & Smol, 2003) or biochemical markers, including sterols (Volkman, 1986) and pigments (Leavitt & Hodgson, 2003). It has been shown that the quantitative relationships between various photosynthetic pigments in the sedimented material are often different from the relationships found in the water column due to their different rate of decomposition/preservation (Louda et al., 2002; Yacobi & Ostrovsky, 2008); therefore, these pigments can be useful biomarkers for tracing diagenetic processing of algal detritus. In our previous study on Lake Kinneret (Ostrovsky & Yacobi, 2010), we showed that the dynamics of chlorophyll a (Chl a) and β-carotene (proxies of phytoplankton biomass) collected in hypolimnetic traps reflected the composition and abundance of phytoplankton in the upper mixed stratum and that traps deployed at the lake periphery and in the BBL notably overestimated the export flux of newly produced particulate organic material. This study is focused on the analysis of photopigments specific for different algal phyla. Here, we examine the dynamics and composition of various phyla-specific pigments in the water column and sedimentation traps to quantify the lateral changes of their sedimentation rates and to better understand the fate of different algal groups in a deep lake.
Each trap set consisted of four plastic cylinders (inner diameter 5 cm, height 50 cm). For a detailed description of the traps and their set up see Koren & Klein (2000). Traps were deployed usually for 1–2 weeks. The four sub-samples of collected material in each trap were pooled before further analysis. Particulate material accumulated in the traps was collected onto GF/C filters and stored in the dark at −18°C. Pigment extraction and analysis of the collected material from water column and traps followed the protocol described by Yacobi & Ostrovsky (2008, and citations therein). Particulate material collected on filters was processed within 1 week, following collection; frozen filters were ground in 3 ml of cold 90% acetone; an additional 3 ml of acetone were used to flush leftovers, and the pooled extract was left overnight in the dark at 4°C. Subsequently, the acetone extract was filtered through a GF/F filter and separated by a reverse-phase HPLC. Pigments were identified on the basis of the retention times and the absorption spectrum following isolation and spectrophotometric examination. The chlorophyll c (Chl c) isolated from lake samples showed absorption peaks at 448 and 634 nm (in HPLC eluant) suggesting a mixture of chlorophyll c1 and chlorophyll c2 (Jeffrey et al., 1997). For simplicity sake, we refer to that pigment Chl c. The quantification of the chromatograms was facilitated by injection of standards of known concentrations into the HPLC system, and calculating the response factor based on the area under the peak. In the system used, lutein and zeaxanthin were not separated, and we report this peak as if it was lutein (see “Discussion”).
Water column samples were drawn from a depth of 1 m with a 5 l Aberg-Rodhe sampler at Stn. A (Fig. 1). A subsample of 1–2 l was stored in opaque plastic carboys until processing in the laboratory approximately 1 h after collection, where duplicate subsamples of several hundred milliliters were filtered onto GF/C filters and stored in the dark at −18°C. From 2006 to 2009 pigments in the water samples were examined with HPLC using the same protocol as described above for the analysis of the trap material. Only in 2005 we used the fluorometric method to measure Chl a concentrations in water.
Correlation analyses were used to reveal whether the sedimentation rates in different traps co-vary and to quantify the strength of these relationships. Since normal distribution could not be assumed, the correlation analyses were performed using Spearman rank correlation. For the same reason, we used the Mann–Whitney test to examine whether the calculated monthly averaged TWI were different from the hypothetical 1. Statistical analyses were performed using the statistics module of SigmaPlot 11 software.
Correlation coefficient (r) between concentrations of different pigments in water and in Aup during three periods: holomixis (January–March), stratification (April–September), and thermocline reclining (October–December)
Peridinin vs. Chl c
Fucoxanthin vs. Chl c
Lutein vs. Chl b
Correlation coefficient (r) between pigment accumulation rates in pairs of various sedimentation trapsa
The mean (±SE) of the flux difference, Fpig (%), in various traps relatively to Aup
The documentation of the distribution of pigments in the epilimnion of Lake Kinneret suggests that the spatial distribution of phytoplankton is rather homogenous, with an exception of the north tip of the lake near the Jordan River inflow (Berman & Elias, 1973; Yacobi & Schlichter, 2004; Ostrovsky & Yacobi, 2009). It is therefore, expected that if only the algal biomass is the main determinant of pigment downward flux, the flux measured by sedimentation traps located below the euphotic zone would be spatially homogeneous. However, that assertion does not hold in Lake Kinneret (Ostrovsky & Yacobi, 2010 and the current study) as well as in other lakes (Bloesch, 2004). In addition, if the vertical downward flux of pigments is only determined by sinking algal material, the measured sedimentation rates should decrease on the way of particle migration down to the bottom, due to cell lysis and material decomposition. Nevertheless, in most cases, the measured sedimentation rates of algal pigments were higher in the near-bottom traps, than in traps positioned shallower in the water column.
Phytoplankton cells and their debris have low-specific density (Reynolds, 2006) and are easily resuspended at the lake periphery, entrained by water motions (Schallenberg & Burns, 2004), and transported laterally via the BBL and metalimnion. The outcome of that fact is an increase in proportion of organic particles both in sedimentation traps and in the uppermost layer of bottom sediments from the lake littoral toward the lake center (Ostrovsky & Yacobi, 1999, 2010; Bloesch, 2004). In this study, we documented a large horizontal difference in sedimentation performance between chlorophyte signature pigments (Chl b and lutein) on the one hand, and signature pigments of the dinoflagellates (Chl c and peridinin), on the other hand (Fig. 6). Lutein, a chlorophyte signature pigment, is stored in sediments for hundreds of years (Züllig, 1981) or even more, as chlorophyte organic cellular residuals are known to be preserved for geological periods (Martín-Closas, 2003). This pigment, as well as Chl b, displayed large spatial heterogeneity in sedimentation fluxes because they contribute to the resuspended fraction, which dominates in peripheral areas of the lake. A similar tendency was displayed by diatoms, which are mostly fast-sinking algae (Reynolds, 2006) and apparently decompose slowly in the water column (Yacobi & Ostrovsky, 2008).
We found that the horizontal changes in sedimentation rates of dinoflagellates were much lower than those of chlorophytes and diatoms (Fig. 5). This is apparently associated with the fact that signature pigments of the dinoflagellates vanish rather fast following their export from the epilimnion (Hurley & Armstrong, 1990; Leavitt, 1993; Steenbergen et al., 1994) and, thus, they are practically absent from the bottom sediments and cannot contribute to resuspended fraction in sedimentation traps. The evidence from Lake Kinneret shows explicitly that the sedimenting cells of the dinoflagellate Peridinium gatunense decompose rapidly (Viner-Mozzini et al., 2003). The later was a reason why the massive Peridinium bloom which lasted from February until May in 2007 did not leave a prominent pigment signature in trapped material (Yacobi, unpublished). It is, therefore, reasonable to conclude that only a slight proportion of the chromophytes reached the traps and if their signature pigments were detected, they originated in the meager left-over of cells that settled intact.
TWI (Fig. 3) indicates that dinoflagellates, which harbor peridinin and Chl c, are mostly recycled in the upper part of the stratified water column and reach the bottom only in minute quantities. The difference in TWI values for peridinin (dinoflagellates) versus fucoxanthin (diatoms) may be attributed to the fact that migrating dinoflagellates are able to compete with other algae under low-turbulence and high-light (strongly stratified water column) conditions, while diatoms usually dominate during holomixis when turbulence is high and cells persist in the entire water column (Margalef, 1997). Dinoflagellates are dominant Chl c-containing algae in Lake Kinneret and that was the reason for the high correlation between Chl c and peridinin in water; whereas the relatively high degradability of peridinin (Hurley & Armstrong, 1990; Yacobi & Ostrovsky, 2008) explains its lower correlation with Chl c in traps (Table 1), where the contribution of diatoms was relatively higher. It is well documented that a number of species of diatoms can reach bottom sediments intact and be preserved there for a long time (e.g., Sicko-Goad, 1986). This mechanism allows the deposited cells to survive unfavorable conditions and regenerate new blooms from the resuspended cells. Such a life strategy was apparently the reason for the appearance of both sedimenting and resuspended particles in traps.
Although chlorophytes dominated phytoplankton biomass quite rarely (with the most prominent case of Mougeotia bloom in April–May 2005), their signature pigment, Chl b, was usually the most abundant pigment in sediment traps. The other chlorophytes’ signature pigment, lutein, was closely correlated with Chl b (r2 = 0.78, n = 642, P < 0.001). Although lutein could not be separated from zeaxanthin, which is harbored by cyanophytes (see above), zeaxanthin apparently had low contribution to lutein assessments. The latter is related to the fact that the dominant summer-blooming cyanophytes Aphanizomenon ovalisporum and/or Cylindrospermopsis raciborskii (Zohary & Shlichter, 2009; Alster et al., 2010) contain minute quantity of zeaxanthin (Yacobi, unpublished data). The scarcity of this pigment was also noted in other Nostocales (Hirschberg & Chamovitz, 1994). While both Chl b and lutein are relatively stable pigments, lutein has higher stability than Chl b (Leavitt & Hodgson, 2003; Yacobi & Ostrovsky, 2008).
The analysis of algal pigments in traps reveals that apparent sedimentation rates of phytoplankton varies notably between locations. The modification depends on several processes, such as the effect of turbulence on trap performance, decomposition of material in various parts of the water column, lateral transportation and focusing of newly produced material in deeper locations and contribution of resuspended particles in peripheral areas (Bloesch, 2004; Ostrovsky & Yacobi, 2010). In most comparisons, Fpig values (Table 3) indicate positive “surplus” of accumulated material in comparison to Aup, which evaluates the net sedimentation rate from the epilimnion. The evident reasons for the “surplus” can be (a) the trapping of the particles that are reintroduced into the water column due to resuspension at peripheral locations (Stns. F and M) and (b) oversampling of the sinking material under turbulent conditions (Alow). It is important to note that higher decomposition of dominant chromophytes before reaching the traps and during the period of material collection results in low “surplus” values, making the respective signature pigments to be non-indicative for the processes of over-trapping and resuspension. In contrast, the amount of less degradable pigments in various traps can well indicate these processes. In any case, the rate of material degradation should be taken into consideration for accurate assessment of the rates of material resuspension or over-trapping.
Using various photosynthetic pigments as biomarkers one can follow the general fate of different algal phyla in the lake and study the sedimentation trap performance under different ambient conditions. For further quantification of the fate of algal material in aquatic ecosystems it is important to specify the main reasons of the persistence (or vulnerability) of specific signature pigments, which vary with the degree of resilience of intact algal cells to breakdown (lysis) and/or the rate of pigment degradation in sinking debris. The quantification of these processes is a matter of further studies.
We thank S. Kaganovsky and N. Koren for field and laboratory assistance and Ms. Miryam Friedman for lingual corrections. This study was partially supported by Lake Kinneret Monitoring Program funded by the Israeli Water and Sewage Authority and by the research grants from the Israeli Science Foundation (ISF Grant No. 932/04 & 627/07). We acknowledge the contribution of two anonymous reviewers whose constructive criticism helped us in improving the clarity and quality of the presentation.