Mixed column depth / m

Figure 6.1

Biomass-specific sinking loss rates of phytoplankters from mixed columns as a function of their depth and intrinsic settling rates (ws). Three instances are inserted to show the greater tolerance of shallow mixing by slow-sinking algae. Redrawn with permission from Reynolds (1997a).

Whence, the rate of change in the standing population that is attributable to sedimentation (rs in Eq. 6.1) is:

The equation expresses the sinking loss rate sustained by a population dispersed in a mixed layer.

It may be deduced that the continued residence of non-motile particles in the pelagic is dependent not only on having maximum entrain-ability (low ws) but also on the settling velocity being small in relation to the mixed depth, hm. As discussed in Section 2.6, the depth of mixing is an extremely variable quantity. Disentrainment is not a disadvantage for a swimming organism, especially not a large one, but non-motile organisms are highly vulnerable to variations in mixed depth (see Fig. 6.1). The growth rate of an alga with an intrinsic sinking rate of 3.5 ^m s-1 (or ~0.3 m d-1) may be able to exceed the leakage of sinking cells across the base of a 10-m mixed layer (rs ~ -0.03 d-1) but, in just 2 m (rs ~ -0.15 d-1), the sinking loss rate may become unsustainable. Species with greater settling rates experience proportionately more severe loss rates from any given mixed layer. Thus, they require yet deeper mixed layers to sustain positive net growth. On the other hand, greater mixing depths quickly begin to impose constraints of inadequate photoperiod-aggregation (see Sections 5.4.1, 5.5.3), and the difficulties of sus taining net intracellular carbon accumulation and its deployment in growth are increased accordingly (Section 3.3.3; see also Sverdrup, 1953; Smetacek and Passow, 1990; Huisman et al., 1999). It must be recognised that, in qualitative terms, larger non-motile plankters experience mixing that is 'too shallow' for growth to overcome their sinking velocity (because hm is relatively small in relation to ws in Eq. 6.10) or is 'too deep' (because hm is relatively much larger than hp in Eq. 5.10; see also Section 3.5.3 and Fig. 3.18) (O'Brien et al., 2003; see also Huisman et al., 2002).

6.3.2 Mixed depth and the population dynamics of diatoms

Included among the larger non-motile plankters are some of the larger freshwater desmids and, especially, the diatoms of the seas and of inland waters. The additional ballast that is represented by the complement of skeletal polymerised silica merely compounds the density difference component, (pc - pw), in Eq. (2.16). Accepting that most species of phytoplankton are required to be either small or motile or to minimise excess density if they are to counter the inevitability of mixed-layer sinking losses, it is striking how poorly the diatoms represent all three attributes. Yet more perplexing is the fact that the planktic diatoms of freshwaters are relatively more silicified that their marine cousins (effectively raising pc; Section 1.5.2 and Fig. 1.9), while the density of many non-saline inland waters (pw) is less than that of the sea. Thus, the density difference of some freshwater diatoms may exceed 100-200 kg m-3 (cf. Table 2.3). How are we to explain how this group of organisms, so relatively young in evolutionary terms, became so conspicuously successful as a component of the phytoplankton of both marine and fresh waters, when it has not only failed to comply to Stokes' rules but has actually gone against them by placing protoplasts inside a non-living box of polymerised silica? There is no simple or direct answer to this question, although, as has been recognised, sinking does have positive benefits, provided that subsequent generations experience frequent opportunities to be reintroduced into the upper water column (Section 2.5). In general, however, many of the ecological advantages of a siliceous exoskele-ton were experienced first among non-planktic diatoms. As the diatoms radiated into the plankton, morphologies had to adapt rapidly: siliceous structures mutated into devices for enhancing form resistance and entrainability within turbulent eddies (Section 2.6). As was demonstrated in the case of Asterionella in the experiments of Jaworski et al. (1988) (see also Section 2.5.3), the configuration of the structures is overriding. Despite order-of-magnitude variations in colony volume and dry mass, as well as an approximate twofold variation in cell density, the sinking behaviour of Asterionella remains under the predominating influence of colony morphology. The corollary must be that the advantage of increased form resistance, and its benefits to entrainability, is greater than the disadvantage of increased sinking speed incumbent upon coeno-bial formation. The counter constraint, however, is that these diatoms are continuously dependent upon turbulence to disperse and to randomise them within the structure of the surface-mixed layer. As predicted by Eq. (6.10), positive population recruitment is always likely to be sensitive to the absolute mixed-layer depth (Reynolds, 1983a; Reynolds et al., 1983b; Huisman and Sommeijer, 2002).

The impact of this interplay between settlement and population dynamics of diatoms on their distribution in space and time is elegantly expressed in the study of Lund et al. (1963) of the seasonal variations in the vertical distribution of Asterionella formosa in the North Basin of Windermere (Fig. 2.21). The build-up in numbers during the month of April and, especially, towards the maximum in May reflect the general decline in vertical diffusivity. In the end, a near-surface concentration maximum is reached, followed by a rapid decline. In this instance, recruitment through growth was impaired by nutrient deficiencies (Lund et al. cited critical silicon levels but phosphorus is now seen likely to have been the more decisive; see Section 5.5.2). However, it is quite clear from the isopleths that the decline in concentration in the upper 10 m or so is extremely rapid. It is compensated, to an extent, by a temporary accumulation in the region of the developing pycnocline. This behaviour is entirely consistent with elimination through sedimentation from the mixed layer and slow settlement through the weak diffusivity of the metalimnion, revealed in the case of nonliving Lycopodium spores (Fig. 2.20). Particles continue to settle through the subsurface layers for many weeks after the population maximum and, indeed, after the surface layer has become effectively devoid of cells.

Heavy sinking losses are not exclusive to nutrient-limited diatom populations. The sensitivity of the population dynamics of diatoms to the onset and stability of thermal stratification in Crose Mere, a small, enriched lake in the English north-west Midlands, rather than to nutrient limitation, was shown by Reynolds (1973a). Diatoms such as Asterionella, Stephanodiscus and Fragilaria were lost from suspension soon after the lake stratified in late spring, even though the concentrations of dissolved silicon and phosphorus remained at growth-saturating levels. Lund (1966) had already argued for the positive role of turbulent mixing in the temporal periodicity of Aulacoseira populations. He showed, in a field-enclosure experiment in Blelham Tarn (incidentally, the one that inspired the construction of the renowned tubular enclosures in the same lake) that the periodicity could be altered readily by superimposing episodes of mechanical mixing by aeration (Lund, 1971).

A little later, the eventual Blelham enclosures (Fig. 5.11) were the site of numerous quantitative

I Instances in the loss from suspension of Asterionella cells in Blelham Enclosure A during 1980, in response to intensifying thermal stratification (shown by the temperature plots). Algal concentration is sampled in 1-m integrating sampler (Irish, 1980) and counted as an average for a 1- or 2-m depth band. The vertical arrows represent the depth of Secchi-disk extinction on each occasion. Redrawn from Reynolds (1984a).

studies of the fate of phytoplankton populations. One early illustration, cited in Reynolds (1984a), shows the depletion by settlement of a thitherto-active Asterionella population, following the onset of warm, sunny weather and the induction of a stable, near-surface stratification, and despite the availability of inorganic nutrients added to the enclosure each week (Fig. 6.2). In a further season-long comparison of loss processes in these enclosures (Reynolds et al., 1982a), during which changes in extant numbers, vertical distribution, growth (as a function of silicon uptake) and sedimentary accumulation rates into sediment traps and at the enclosure bottom were all independently monitored, a steady 'leakage' of Asterionella cells was demonstrated over the entire cycle of net growth and attrition (see Table 5.5 and Reynolds and Wiseman, 1982). In the Blelham Enclosure B, Asterionella increased at a rate of 0.147 d-1 during its main phase of growth, net of sinking losses calculated to have been ~0.007 d-1. The sedimenting cells intercepted by the traps were calculated to have been sinking at an average rate of (ws =) 0.08 m d-1 (just under 1 ^m s-1) through a water column (hm = ) 11.7 m. As the population reached its maximum, the net rate of increase slowed (to 0.065 d-1) but the sinking loss rate remained steady (-0.007 d-1). However, shortening of the mixing depth led to an accelerated rate of sinking loss (to -0.044 d-1; from a mixed depth of now only 7.5 m, a faster sinking rate of ws = 0.33 m d-1 is also implied). More remarkably, as the epilimnion shrank to 4 m, the loss rate then rose to -0.242 d-1, sustained by a sinking rate of 1.02

m d-1 (equivalent to 11.8 ^m s-1). The data are plotted in Fig. 6.3.

The accelerated sinking loss was contributed, in part, by an accelerated sinking rate. This was not unexpected. Reynolds and Wiseman (1982) had noted the altered physiological condition of the cells at the time, both in the plankton and in the sediment traps, drawing attention to the contracted plastids and 'oily' appearance of the contents. They attributed the changes to the sudden increase in insolation of cells caught in a stagnating and clarifying (cf. Fig. 6.2) epil-imnion at the same time that temperature and light intensity were increasing. They suggested that the changes were symptomatic of photoinhibition. When Neale et al. (1991b) made similar observations on diatom populations in other lakes, they made the similar deduction. A positive feedback implied by the sequence of more insolation ^ more stratification ^ more inhibition ^ faster sinking rates ^ faster sinking loss rates has a satisfying ring of truth. However, a modified interpretation would see the accelerated sinking rate as a withdrawal of the vital mechanism of reducing sinking rate (see Section 2.5.4) for the very positive purpose of escaping the high levels of near-surface insolation.

The bulk 'production-and-loss budgets' compiled by Reynolds et al. (1982a) for phytoplankton populations in the Blelham Enclosures and exemplified in Table 5.6 offer a clear account of the the fate of the total production. In the example given, 81% (confidence interval, 70-95%) of the Asterionella formosa produced in Enclosure B in the spring of 1978 constituted the observed

Figure 6.3

Net increase and attrition of an Asterionella population cells in Blelham Enclosure B during spring 1978. (a) Changes in the instantaneous areal cell concentrations in the water to 3, 5 and 11 m; (b) changes in the silicon-specific replication rate (r'$i) and the net rate of population change (rn); the hatched areas correspond to the rate of population loss, almost wholly to sinking. Redrawn with permission from Reynolds (1997a).

Figure 6.3

Net increase and attrition of an Asterionella population cells in Blelham Enclosure B during spring 1978. (a) Changes in the instantaneous areal cell concentrations in the water to 3, 5 and 11 m; (b) changes in the silicon-specific replication rate (r'$i) and the net rate of population change (rn); the hatched areas correspond to the rate of population loss, almost wholly to sinking. Redrawn with permission from Reynolds (1997a).

maximum. Around 95% (confidence interval, 72-123%) of the production was recruited intact to the sediment. The proportion of the cells produced that was estimated to have been lost to herbivores was probably <6% (see Sections 6.4.2, 6.7). The use of the adjective 'intact' is taken to include cells that may well have been dead by the time they reached the sediment surface. Judged from weekly recoveries from sediment traps placed about ~1 m above the sediment (and to which preservative was added), Reynolds and Wiseman (1982) observed that the proportion of live cells was always greater than 89% throughout the course of the population rise and decline. The proportion of live cells in the superficial sediment (supposedly dominated by the most recently recruited material) was 92% at the beginning of April. By the end of the month, it had fallen to 67%, to <2% by the end of July and to zero by the first week in September.

As part of the same investigation, Reynolds and Wiseman (1982) compared the rates of production, sedimentary fluxes and sediment recruitment of several other species forming major populations in the same enclosures. Of the estimated summer production of another diatom, Fragilaria crotonensis, at least 49% (statistically, possibly all) of the production was recruited to the sediments. In contrast, sedimentation could explain the fate of no more than 4% of the observed population maxima of Ankyra, Chromulina or Cryptomonas. Intermediate between the extremes of heavy diatoms and nanoplanktic unicells, sediment and trap recoveries of Eudo-rina accounted for 55 ± 15% of the maximum standing crops. For Microcystis, the sedimentary behaviour was strongly seasonal, increasing from 8% to 100% through the autumn.

From the measurements of the production and eventual fate of phytoplankton in confined, flat-bottomed Blelham enclosures, at least, the assertion that most of the larger diatoms are destined to be lost to sedimentation is strongly supportable. Scaling up to larger and deeper systems, subject to significant horizontal diffusive transport, the deduction requires some caution. In a 2-year study of sedimentary fluxes in the South Basin of Windermere (maximum depth 42 m), Reynolds et al. (1982b) found good, order-of-magnitude agreement between the annual fluxes into deep sediment traps and the maximal standing crops of five species of planktic diatom (Asterionella formosa, Aulacoseira subarctica, Cyclotella praetermissa, Fragilaria crotonensis, Tabellaria floccu-losa var. asterionelloides) and two of desmid (Cosmar-ium abbreviatum, Staurastrum cf. cingulum). Interestingly, the magnitude of the fluxes (in numbers of cells m-2 d-1) varied with the size of extant poulations but measurable fluxes to depth persisted through most of the year. This is presumed to reflect the relative proportion of the particle settling rates to the vertical distance to be traversed; this also fits with the observations of Lund et al. (1963) for the North Basin and the distribution of population isopleths plotted in

Fig. 2.21. Incidentally, the proportion of live Asterionella cells trapped fell from around 95% at the time of the May population maximum to just 3% in August and September. In the 100-or-so days that it takes some diatoms to settle through 40 m, many must perish, leaving only the empty frustules to continue downwards.

In contrast to the diatoms, the sedimentary fluxes of three colonial chlorophyte species (Coenochloris fotti, Pseudosphaerocystis lacus-tris, Radiococcus planctonicus), three Cyanobacte-ria (Anabaena flos-aquae, Woronichinia naegeliana, Pseudanabaena limnetica) and the dinoflagellate Ceratium hirundinella were 1-3 orders of magnitude smaller than the potential of the maximum standing crop. All these species either sink very slowly or they have sufficient motility to avoid being sedimented for long periods. Cryptomon-ads and nanoplankters were virtually unrecorded in any trap catches; they are presumed to have been subject to loss processes other than settlement.

These various findings supported the earlier deductions of Knoechel and Kalff (1978), who had applied a dynamic model to compare the effects of measured rates of growth, increase and settlement in order to calculate sinking loss rates of planktic populations in Lac Hertel, Canada. Their calculations showed that the rates of sinking loss were sufficient to explain most of the discrepancy between growth and the contemporaneous rate of population change, be it up or down. They were also able to provide quantified support for the idea that, whatever fate may befall them (nutrient, especially silicon, exhaustion, grazing, parasitism), planktic diatoms remain crucially sensitive to the intensity and extent of vertical mixing. Other workers who espoused this explanation for the seasonal fluctuations in diatom development and abundance in limno-plankton include Lewis (1978a, 1986), Viner and Kemp (1983), Ashton (1985) and Sommer (1988a).

There is now also ample evidence to support the qualitative contention of Knoechel and Kalff (1975) that sedimentation is a key trigger to the seasonal replacement of dominant diatoms by other algae. It is also plain that sedimentation is the principal loss to which lim netic diatoms are subject. In most other phyto-plankton, greater proportions are either eaten or decompose long before they reach the sediment. The deduction concurs with the studies of losses made by Crumpton and Wetzel (1982) and with that of Hillbricht-Ilkowska et al. (1979) in Jezioro Mikolajske, Poland, on the seaonal variations in the main sinks of limnetic primary products.

The sensitivity of marine diatom dynamics to mixed-layer depth is not so clearly defined. On the one hand, net population increase is dependent upon an enhancement in insolation above thresholds which may be lower than for many other marine species (Smetacek and Pas-sow, 1990) but the diminution of the mixed layer in the sea to the 1-3 m that may be critical to net diatom increase is inconclusively documented. Nevertheless, oceanic diatom populations experience considerable sinking losses that may be sustained only at or above certain levels of productivity. It is inferred that these are dependent upon adequate physical and chemical support (Legendre and LeFevre, 1989; Legendre and Rassoulzadegan, 1996; see also Karl et al., 2002).

As to the question posed by Huisman et al. (2002) about the long-term persistence of sinking phytoplankton, we have shown that there are obvious short-term benefits in being able to escape surface stagnation and resultant damaging levels of insolation in the near-surface waters (Reynolds et al., 1986). Provided there is an opportunity for surviving propagules to be reestablished within the photosynthetic range, the sooner may the longer-term benefit of population re-establishment be realised. Particle aggregation and, especially, the formation of 'marine snow' (Alldredge and Silver, 1988) may contribute effectively to accelerated sinking and to the escape from high-insolated surface layers. Aggregation may also serve to provide microenvironments that slow down the rate of respirational consumption and resist frustular dissolution of silicon (Passow et al., 2003). The mechanisms of accelerated sinking may also add to the longevity of clone survival and facilitate the improved prospect of population re-establishment when more suitable growth conditions are encountered.

6.3.3 Accumulation and resuspension of deposited material

As has already been discussed, settling is not exclusively a loss process in the population dynamics of phytoplankton: the recruitment of resting propagules to the bottom deposits is recognised to constitute a 'seed bank' from which later extant populations of phytoplankters may arise (see Section 5.4.6). For this to be an effective means of stock perennation and mid- to long-term persistence in a given system, however, there has to be a finite probability of settled material both surviving on the sediments and, thence, of re-entering the plankton. The species-specific regenerative strategies of phyto-plankton -- roughly their ability to survive at the bottom of the water column and the means of 'escape' to the overlying water column -- are extremely varied, ranging from the conspicuous production of morphological and/or physiological resting stages, with an independent capacity for germination, regrowth and reinfection of the water column (as in the case of Microcystis or Ceratium), through a range of resting cysts and stages whose re-establishment in the water depends upon still-suspended or resuspended propagules encountering tolerable environmental conditions (as is true for akinetes of nosto-calean Cyanobacteria, certain species of volvo-calean and chrysophyte resting cysts and the distinctive resting stages of centric diatom), to those that seem to make virtually no such provision at all (see Section 5.4.6).

In most instances, the settlement of vegetative crops should be regarded as terminal. Vegetative cells sinking onto deep, uninsolated sediments have little prospect but to slowly respire away their labile carbohydrates, pending depth. Resting cysts may remain viable for many years (64 a is a well-authenticated claim of viability of Anabaena akinetes: Livingstone and Jaworski, 1980) but without the mechanical resuspension of the resting spores in insolated, nutrient-replete water, the reinfective potential remains unrealised. Once settled to the bottom of a water column, the most likely prospect is progressive burial by the subsequent sedimentary recruitment of further particulate material, including fine, catchment-derived silts and par-

ticulate organic matter, the exuviae and excreta of aquatic animals and a rain of sedi-menting phytoplankters, especially of non-motile diatoms.

Several studies have attempted to focus on the nature of the freshly sedimented material in lakes and its immediate fates. For a time, the newest recruited material remains substantially uncompacted and floccular, like a fluff, on the immediate surface. It comprises live or moribund vegetative cells, often bacterised or beset with saprophytic fungal hyphae, and resembles on a smaller scale, the structure of 'marine snow' (Alldredge and Silver, 1988; see Section 6.3.2). As its substance diminishes, however, it does become slowly compressed by later-arriving material. At the base of the semifluid layer, the same materials are progressively lost to the permanent sediment (Guinasso and Schink, 1975): compacting, losing water, perhaps leaching biominerals, the first stages of sediment diagenesis and formation are engaged.

Accordingly, the manner in which strictly ordered, laminated sediments might flow from the sequenced deposition of specific phytoplank-ton populations seems obvious. However, direct sampling of the semifluid layer from intact cores of the sediment water interface (Reynolds and Wiseman, 1982, used a syringe inserted into pre-drilled plastic tubes fitted to a Jenkin surface-mud sampler, as described in Ohnstad and Jones, 1982) reveals that sedimenting material undergoes a kind of sorting process. Once recruitment to the semifluid layer from the water column is effectively complete, its presence in the semifluid layer is found to decay exponentially. Moreover, the rates of dilution from the semifluid layer are not uniform but vary interspecifically, according to size and shape (Haworth, 1976; Reynolds, 1996b): long cells of Asterionella, filaments of Aulacoseira and chains of Fragilaria are diluted less rapidly from the semifluid layer than centric uni-cells of Cyclotella or Stephanodiscus.

Relative persistence in the surface layer improves the prospect of live specimens being restored to suspension in the water column, supposing that the physical penetration of adequate resuspending energy obtains. In general, friction in the region of the solid sediments creates a velocity gradient and a boundary layer of reduced water velocities, in which freshly settled plankton can accumulate (see Section 2.7.1). Resuspension is thus dependent upon the application of sufficient turbulent shear force to compress the boundary layer to the dimensions of the settled particles or even beyond the resistance of the unconsolidated sediment to penetration by a shear force, by then competent to entrain and resuspend it (Nixon, 1988). Quantitative observations confirm the intuition that shallow sediments are rather more liable to resuspension than sediments beneath a substantial column of water, although the actual depth limits vary with sediment type and the energy of forcing (Hilton, 1985). In many small lakes, sediments at a depth greater than 5 m beneath the water surface are protected from wave action and from most windgenerated shear. In the short to mid term, resuspension may require physical forcing of seismic proportions, or depend upon disturbance by burrowing invertebrates or foraging behaviour of fish or diving animal (Davis, 1974; Petr, 1977). In contrast, shallow sediments (substantially <5 m) may be rather more routinely exposed to resuspension of sediment and, incidentally, the redispersion of sediment interstitial water that may be relatively enriched, with respect to the open water, with nutrients released in decomposition (see also Section 8.3.4). In the Blelham enclosures (see Fig. 5.11), very little resuspension of live phytoplankton, resting spores or even empty diatom frustules was observed from the universal deep sediments of Enclosures A or B but it was observed on numerous occasions in the graded Enclosure C (Reynolds, 1996b). Moreover, disturbance or removal of the semifluid sediment from the shallow-water station, CS (Fig. 5.11, depth ~4.5 m), occurred at such times, whereas, the deeper station, CD (depth ~12.5 m) was exempt from this. In the wake of such resuspension events, material was perceived to resettle uniformly at both stations. Over a series of resuspensions, a net transport of once-settled material from shallow areas to deep sites was deduced.

So far as the accumulation of sediment-ing phytoplankton is concerned, near-permanent deposition follows analogous patterns to nonliving particulate matter. At depths typically greater than 5 m, sedimenting material accumulates and builds up in layers, undergoing diagenesis under substantially anoxic conditions. Neither live vegetative cells nor most resting spores enjoy much prospect of return to suspension and regeneration. In contrast, similar materials settling onto shallow sediments are liable to resuspension. The viable fractions (vegetative cells, resting spores) may well fulfil their infective potential and contribute directly to the establishment of extant, vegetative populations. This has been many times observed in the case of Aulacoseira populations (Lund, 1954, 1966, 1971) and is inferred on other occasions involving other species (Carrick et al., 1993; Reynolds et al., 1993a). For the non-viable detritus, including empty diatom frustules, redeposition is the most likely consequence but with a finite proportion settling into deeper water. This is precisely the mechanism of the process of 'sediment focusing' (Lehman, 1975) whereby fine particulate material is moved progressively away from lake margins and towards greater basin depths (Hutchinson, 1941; Likens and Davis, 1975; Hilton, 1985)

0 0

Post a comment