Info

Maximum standing crop Production from Si uptake Estimated consumption by grazers Sedimentary flux Recruitment to sediments

Source: Data of Reynolds and Wiseman (I982) and Reynolds et al. (I982a), as presented in Reynolds (I996b).

population, it is possible to calculate a budget for the entire phase of production. These are entered in Table 5.6, but are calculated for the full water column (mean depth, 11.0 m). The entries are subject to wide margins of error but the agreement is encouraging: the lower ends of the confidence intervals are mutually compatible. From the aggregate loss of silicon from solution (equivalent to 18.788 g Si m-2), the number of Asterionella cells that could have been produced was estimated to have been between 307.3 and 362.7 x 109 cells m-2. Had they all been in suspension simultaneously, the concentration would have been between 27.94-32.97 x 103 mL-1 (cf. observed maximum, 24.78 x 103 mL-1; 272.6 x 109 cells m-2). The numbers that were recruited to the sediment, the settling flux that was intercepted and the range of fatalities to grazing are each inserted in Table 5.6.

filter-feeding by zooplankton (chiefly of Daphnia galeata, though activity was minimal until late April: Thompson et al., 1982).

The various outcomes are tabulated and compared in Table 5.5. At first, almost all the silicon investment is realised in observable additions to the extant population. Moreover the slowing rate of increase is explained almost wholly by a declining rate of replication, for whatever reason this might be. However, the rates of loss mount and rates of recruitment through cell division slow, until the former exceed the latter and the population goes into decline. In this particular

5.5.2 The spring increase in temperate lakes: the case of Windermere

It is appropriate now to consider an example of the role of growth in the development of phytoplankton maximum, taking the case of a familiar and annually recurrent event, observed over a sequence of several consecutive years and responding to a substantial degree of interan-nual environmental variability. The example of the Asterionella-dominated spring increase in Win-dermere is chosen for the length and detail of the observational record and because the year-to-year variations in the dynamics of growth

Figure 5.12

The seasonal change in the standing population of live Asterionella cells in the upper 7 m of the South Basin of Windermere (heavy line), averaged over the period 1946 to 1990 inclusive, together with the envelope of 95% confidence (shown by hatching) and the maximal and minimal values recorded (the lighter lines). Modified from an original figure in Maberly et al. (1994) and redrawn with permission from Reynolds (1997a).

and population increase (and subsequent decline) have been intensively analysed. Windermere is a glacial ribbon lake in the English Lake District, covering 14.76 km2. It is not far short from being two lakes, a shallow morainic infill separating the lake into two distinct but contiguous basins. The North Basin holds nearly two-thirds of the total volume (314.5 x 106 m3) and has a mean depth of 25.1 m (maximum is 64 m); the mean depth of the smaller South Basin is 16.8 m (maximum 42 m) (data of Ramsbottom, 1976). The southward water flow through the lake is generated mainly from the catchment in the central uplands of the Lake District. The mean annual discharge from Windermere (437 x 106 m3 a-1) corresponds to a theoretical mean retention time of 0.72 a. The upper catchment is based upon hard and unyielding volcanic rocks, the lower foothills comprising younger and slightly softer Silurian slates. Both are poor in bases. Thin soils, mostly cleared of the natural woodland covering and replaced with rather poor, leached grazing land, contribute little in the way of bases or nutrients to the lake. Nutrient loads have increased substantially over the last 50 years or so, through the increased use of inorganic fertilisers on the land. The sheep population that the added fer-tilty supports has roughly doubled over the same period. However, it is the increase in the human population, together with the introduction (in 1965) of secondary sewage treatment, which has most affected the phytoplankton-carrying capacity of Windermere (for full details, reconstructed loadings and the effects of the 'restoration' mea sures, commenced in 1991, see Reynolds and Irish, 2000). By the late 1980s, autumn-winter concentrations of MRP increased from a pre-1965 average of 2-2.5 pg P L-1 (0.06-0.08 pM) to ~8 pg P L-1 in the North Basin and to about 30 pg P L-1 in the South Basin. In both basins, autumn-winter DIN levels more or less doubled over the same period, from ~350 to ~700 pg N L-1 (2550 pM), but SRSi levels have remained steady (at 0.9-1.1 mg Si L-1; 32-39 pM; 1.9-2.3 mg SiO2 L-1).

This information permits us to establish the physical-chemical characters of the habitat of the spring bloom. It is a relatively clear (ewmin ~ 0.31 m-1) - and soft(alkalinity <0.26 meq L-1) - water, barely mesotrophic lake, experiencing a mostly cool, temperate, oceanic climate, but incurring, between 1965 and 1991, substantial anthropogenic enrichment. For almost 200 years, Asterionella has been a conspicuous member of the plankton in both basins and, during at least the last 60 of those, the almost unchallenged dominant species of the spring-bloom period.

Maberly et al. (1994) carried out a thorough statistical analysis of the (then) complete run of data collected from (mostly) weekly samplings of the South Basin, initiated by J. W. G. Lund in 1946 (see Lund, 1949) and maintained, with only detailed methodological variation, over the 45 years to 1990. A simplified version of their main figure is reproduced here, as Fig. 5.12, to illustrate the reproducibility of the main features of the development. The heavy line represents the 45-year mean of the standing population (on a logarithmic scale) as a function of the day of the year. The narrowness of the 95% confidence interval (shown by hatching) attests to the strong interannual comparability of the growth, even though the boundaries of the extreme records over the 45 years (delimited by the lines either side of the mean plot) cover 2 or more orders of magnitude of variation. Maberly et al. (1994) diagnosed several cardinal characteristics of the growth curve, including the size of the extant population at the beginning of the year (mean 9.7 (x/^ 3.85) cells mL-1; range 0.6-330 mL-1); the maximum (mean 3940 (x/^ 2.28) cells mL-1 range 330-11500 mL-1), the date of its achive-ment (day 124 ± 16.7 (17 April-21 May)), the start (day 52 ± 24 (28 January-17 March)) and end of the period of rapid exponential increase (day 106 ± 17 (30 March-3 May)); the mean rate of increase achieved (0.0925 (± 0.0357) d-1), as well as the date of the commencement of the steep post-maximum exponential decrease (day 142 ± 16 (6 May-9 June)).

The source of the bloom is essentially the standing stock in the water at the turn of the year (Lund, 1949; Reynolds and Irish, 2000). Inter-annual variations in this survivor stock (mean 6.6 (x/^ 4.83) cells mL-1) are influenced by the size of the late summer maximum of the previous year and the extent of its net dilution by autumnal wash-out. There is a modest increase in the standing crop detectable throughout the first few weeks of the year, so there is, in no sense, any part of the year when growth is not possible. This is close to, or just prior to, the time of greatest nutrient availability in Windermere, so the restriction on biomass increase has long been supposed to be physical. The lowest water temperatures are generally encountered in late January (days 14-28, weeks 3 or 4, of the year) but usually remain <7 °C until week 12 (day 84 and several weeks after the inception of the main exponential increase phase). On the other hand, Talling's (1957c; see Section 3.3.3) extrapolations of column-integrated photosynthetic production in the lake show that light, especially in a fully mixed water column, strongly indicate that light is the main growth-regulating factor. The early growth of Asterionella in Windermere has been observed to be relatively weak in the stormiest winters predominated by south-westerly winds and stronger in anticyclonic winters. However, general trends over the full period are towards smaller overwintering inocula but faster rates of exponential rise in the spring. Of particular interest is the fact that, although the size of the maximum crop seems not to have increased over the 45 years, the date of its attainment has tended to be reached earlier in the year, as a consequence of the trend towards sustained faster growth rates.

The maxima of Asterionella in the North Basin of Windermere have been similar in magnitude to those observed in the South Basin in the corresponding years but, typically, have always been reached one to two weeks later. The greater mean depth of the North Basin appeared to be the most likely reason for this relative delay. However, the maxima here are also now reached significantly earlier (~10 d) in the year. The proximal explanation is, again, that a faster average rate of increase is maintained but why this should have changed when light is alleged to be the rate-regulating factor is not immediately apparent.

Circumstantial evidence and some simple modelling reveal a complex factor interaction at work. Starting either from the premise of sustainable growth rates (Reynolds, 1990) or photo-synthetic behaviour (Neale et al., 1991b), it may be shown that the long-term average of the underwater light integral, I*, sets much lower constraints on Asterionella growth than does water temperature, SRSi or SRP concentration. The time track of I* in Fig. 5.13 is reconstructed from long-term (42-year) records of daily integrals of irra-diance (I0) and wind run, and is hypothesised to express the intergal daylight period experienced by entrained algae. Comparison of the modelled growth rates, r, as determined by each of the constraining factors in turn, shows that, initially, the least are those set by I*. They move from about 0.04 to 0.08 d-1 during the first 50 days (7 weeks) of the year, before accelerating up to 0.21 d-1 by week 15 (days 98-105). Depending upon the size of the starting inoculum, this is sufficient to take the number of formed cells into the range 103 to 104 cells mL- 1 and incipient Si exhaustion. The supportive capacity of the initially available silicon is 14-19 x 103 cells mL-1, if it is assumed that SRSi is drawn down to extinction and shared at 60-65 pg Si cell-1 produced. Only for the purpose of modelling is it assumed that there is no concomitant loss of cells from suspension, or that

Figure 5.13

Features of the environment of the North Basin of Windermere, during the first half of the year, influencing the spring increase of Asterionella. The top panel shows the water temperature (6 °C), averaged for each week of each of 43 consecutive years (1946 to 1988 inclusive, with the 95% confidence interval). The second panel shows similar averages for Io, the insolation at the water surface, and for I*, the average light level in the mixed layer. Typical (not statistically averaged) changes in the content of SRSi (soluble reactive silicon) are shown in the third panel, while the fourth reflects long-term increases in SRP (soluble reactive phosphorus) levels by providing typical trend lines for two separate periods (1945-65 and 1980-85). The lower panel shows plots of modelled growth rate capacities of 6, I*, SRSi and SRP. The sustainable growth rate is that of the limiting component, which, clearly, is I* until week 16 or 17, when diminishing SRP or SRSi levels become critical. Revised from Reynolds (1990) and redrawn with permission from Reynolds (1997a).

there is no demand for silicon from any other agency. In fact, observable populations of 9-10 x 103 Asterionella cells mL-1 will not have formed without using most of the available SRSi, at least down to the critical half-saturation level of 23 |M (see Section 5.4.4).

The key deduction is that, prior to 1965, the initially available phosphorus (say 2.5 |g P L-1) would have had to have been shared among a maximal population of 10 x 103 cells mL-1, each having a mean residual quota of <2.5 pg cell-1. By halving the quota again, the next cell division will submit the growth rate to P-limitation (see Section 5.4.4). It follows that the diatom-supportive capacity of Windermere is, indeed, set by the available silicon. However, the rate of its assimilation and conversion into Asterionella biomass is closely regulated first by the light income availability but, as soon as this begins to be relieved by lengthening days and weaker vertical mixing, phosphorus availability starts to squeeze the attainable growth rate instead. The final twist in this changing factor interaction comes with the recent phosphorus enrichment of the lake. As first noted by Talling and Heaney (1988), enrichment with phosphorus relieves the growth-rate constraint, which now continues to the limits permitted by I*, until the ceiling of silicon exhaustion is reached, now rather earlier in the year. The sketches in Fig. 5.14 summarise the shifting date of maximal attainment. The silicon limit remains inviolable. Interactions among interannualy varying constraints have influenced the extent of capacity attainment. Over a number of years of enrichment, it became the case that the Asterionella maximum failed to exhaust all the phosphorus in solution, instead leaving it available to be exploited by other phytoplankton. As will be argued later (Section 8.3.2), this is a defining stage in the eutrophication of temperate lakes.

There is much more to this story, and to the next one about the rapid post-maximum collapse of the spring bloom. Both concern the magnitude of the loss terms making up rL in Eq. (5.3), discussion of which is developed in Chapter 6. All the time that the population is increasing, the replication rate is sustaining losses of formed cells to mortality - to such physical agents as out-wash and settling beyond the resurrecting limits

Figure 5.14

Reynolds' (1997b) proposed explanation for the advancing date of the Asterionella maximum in Windermere. (a) On the plot of accelerating standing crop, the ceiling capacity of the silicon resource K(Si) occurs only a little short of the point where the supportive capacity of the phosphorus would have been exhausted, although recent phosphorus enrichment has raised this ceiling further. Thus in (b), the likelihood of a phosphorus-limited growth rate is delayed and relatively fast growth rates can be maintained for longer. The silicon ceiling, which has remained stable, is thus reached sooner. Changing nutrient availability may be cited for the advance which may have little or nothing to do with climate change. Redrawn with permission from Reynolds (1997b).

Figure 5.14

Reynolds' (1997b) proposed explanation for the advancing date of the Asterionella maximum in Windermere. (a) On the plot of accelerating standing crop, the ceiling capacity of the silicon resource K(Si) occurs only a little short of the point where the supportive capacity of the phosphorus would have been exhausted, although recent phosphorus enrichment has raised this ceiling further. Thus in (b), the likelihood of a phosphorus-limited growth rate is delayed and relatively fast growth rates can be maintained for longer. The silicon ceiling, which has remained stable, is thus reached sooner. Changing nutrient availability may be cited for the advance which may have little or nothing to do with climate change. Redrawn with permission from Reynolds (1997b).

of entrainment and to the biological demands of grazing and parasitic consumers. The spring bloom in Windermere, as elsewhere, is sustained to within the limits that light and available nutrients can support, net of ongoing sinks, recognising that the latter may be more or less critical in commuting the size of the maximum crop below that of the chemical capacity.

5.5.3 Selection by performance

The example of Asterionella in Windermere may, or may not, have direct analogues to other diatom blooms in other systems or even to the behaviour of other kinds of phytoplankton. The main illustrative point is that species have to make the best of the environments in which they find themselves and, often, they must press their specific traits and adaptations to perform adequately under environmental conditions verging on the hostile. In this way, we might interpret the ability of Asterionella to dominate the vernal plankton of Windermere as being dependent upon certain attributes. The first is so obvious that it is easily overlooked: it is there! The success of a species in a habitat is a statement that the habitat is able to fulfil its fundamental survival requirements and, of the species that have arrived there, this species will be, relatively, the most efficient in exploiting the opportunity offered. Absence of a species does not inform a deduction that the habitat is not suitable; it may just not have had the opportunity to grow there. No species dominates a habitat just because theory argues it to be the most suitable. However, to be able to to outperform other species at a given point in space and in time must suggest a favourable combination of inoculum and a relatively superior exploitative efficiency under the conditions obtaining.

The fossil record shows that Asterionella dominance is a relatively new phenomenon in Winder-mere, having occurred only since the nineteenth century (Pennington, 1943, 1947). Previously, Cyclotella species had dominated a more olig-otrophic period in the lake's history (Haworth, 1976). Asterionella is able to grow faster than other diatoms under the poor vernal underwater light conditions and faster than Planktothrix at low temperatures. It also manages to carry over substantial winter populations from which spring growths can expand. Asterionella does not have matters exclusively to itself - Aulacoseira species (A. subarctica, A. islandica) overwinter well, though they grow less rapidly than Asterionella; flagellates such as Plagioselmis grow relatively well in winter anticyclones (with frosts, sunshine and weak wind-driven vertical mixing) (Reynolds and Irish, 2000).

Small interannual variations in these environmental features may not make decisive intrasystem differences in outcomes but they may assist us to understand the differences in timing, the magnitude of crops and the species dominance of populations elsewhere. We have shown, through the comparison of growth responses and their sensitivities to environmental deficiencies, how the dynamic performances differ among species in experiments. Can we now discern differences of performance in nature that will confirm -or help us to recognise - the traits that select for some species and against others at a given location? If so, how much does this tell us about the ways in which natural communities are put together and shape trophic relationships?

The answers to these questions are clouded by the usual problems of accurate measurement of population dynamics in the field (see Section 5.5.1). Work with captive wild populations of phytoplankton in the Blelham enclosures, growing within a defined space, subject to well-characterised and, in part, artificially controlled conditions, subject to separately quantified loss rates of cell loss and, above all, sampled at high frequencies (3-4 days), does provide some insights. From data collected from numerous growth phases, observed in three enclosures over 6 years, Reynolds (1986b) assembled a series of in-situ replication rates for each of a number of common species. Summaries are shown in Fig. 5.15. Each datum point is calculated from a minimum of three serial measurements on an increasing population and is corrected for the contemporaneous estimates of loss rates to sinking and grazing (details to be highlighted in Chapter 6). These points were then plotted against a common scale of analogous insolation, this being the product of the day length (r, from sunrise to sunset, in hours) and the ambient ratio of Secchi-disk depth to mixed depth (zs/zm, with the proviso that solutions >1 are treated as 1). Finally, the points are grouped according to the approximate contemporaneous water temperatures.

The plot does reveal an encouraging level of intraspecific consistency of performance and significant interspecific differentiation. Taking the observations on Fragilaria, for instance, replication rates in the field, between 13 and 17 °C, reveal a common dependence upon the aggregate-by-analogy photoperiod, with a slope that appears steeper than (two) observations applying to temperatures between 9 and 11 °C, yet less steep that the indicated photoperiod depen dence between 18 and 20 °C. There also seems to be a common threshold light exposure (~4 h d-1 on the analogue scale) applying at all temperatures. The information sits comfortably with our understanding of growth sensitivites in the laboratory. It may also be noted that the replication rates in the field do not differ widely from the maximum resource- and light-saturated rates observed in culture at 20 °C, if the appropriate adjustments for temperature and photope-riod are applied.

Analogous deductions can be drawn from the data for the other algae represented in Fig. 5.15. The ability of another ^-strategist alga, Asterionella, to adapt to function on low average insolation is confirmed by observed growth rates of up to 0.15 d-1 on a low aggregate daily photoperiods (<4 h d-1 on the contrived scale) at temperatures from 5 to 15 °C. Collected data for Cryptomonas spp. (mostly C. ovata) and Planktothrix mougeotii also confirm that growth rates are maintained by photoadaptation to low aggregate insolation (with thresholds of 1-3 h d-1) but they are not so fast as the most rapid performances of the diatoms. The most rapid growth rates observed in the enclosures have been attained by C-strategists such as Ankyra, which, on several occasions, has been observed to self-replicate at >0.8 d-1 (doubling its mass in less than a day!). These shortlived episodes have been possible in warm, clear, usually water that is restratifying and supplied with nutrients well in excess of growth-rate limiting concentrations. Lack of carbon, self-shading or increased vertical mixing contribute to a slowing growth rate in these instances: note the apparent threshold at ~8 h d-1 on the contrived scale of daily photoperiod. The plots for the C-S species Eudorina unicocca seem to point to an even greater photoperiod response, little influenced by the (somewhat narrow range of) water temperatures available. The two strongly S-strategist Cyanobacteria (Anabaena and Microcystis) are generally slow-growing (<0.36 d-1); as a function of photoperiod, there is an intermediate threshold of 4-6 h d-1 on the artificial scale.

Many other observations on the growth performances of phytoplankton have emerged from the studies using the Blelham enclosures. Some relate to the dynamics of loss and the way these

Figure 5.15

Approximations of the daily specific growth rates (r' = rn + re + rs) reconstructed from detailed observations on the dynamics of populations of named phytoplankters in Blelham enclosures, plotted against the contemporaneous products of the length of the daylight period (T, in h) and the ratio of the Secchi-disk depth to the mixed depth (hs/hm), being an analogue of I* (in instances where hs/hm > 1, hs/hm is put equal to 1). Curves are fitted to data blocked according to contemporaneous temperatures, as stated. The algae are: Anaba, Anabaena flos-aquae; Ankyr, Ankyra judayi; Aster, Asterionella formosa; Crypt, Cryptomonas ovata; Eudor, Eudorina unicocca; Fragi, Fragilaria crotonensis; Micro, Microcystis aeruginosa; Plank, Planktothrix mougeotii. Modified and redrawn with permission from Reynolds (1986b).

Figure 5.15

Approximations of the daily specific growth rates (r' = rn + re + rs) reconstructed from detailed observations on the dynamics of populations of named phytoplankters in Blelham enclosures, plotted against the contemporaneous products of the length of the daylight period (T, in h) and the ratio of the Secchi-disk depth to the mixed depth (hs/hm), being an analogue of I* (in instances where hs/hm > 1, hs/hm is put equal to 1). Curves are fitted to data blocked according to contemporaneous temperatures, as stated. The algae are: Anaba, Anabaena flos-aquae; Ankyr, Ankyra judayi; Aster, Asterionella formosa; Crypt, Cryptomonas ovata; Eudor, Eudorina unicocca; Fragi, Fragilaria crotonensis; Micro, Microcystis aeruginosa; Plank, Planktothrix mougeotii. Modified and redrawn with permission from Reynolds (1986b).

interact with differential growth rates in influencing community assembly and succession, to which reference will be made in subsequent chapters. Of particular interest to the question of selection by growth performances is the collective overview of species-specific development in relation to chemical factors. The enclosures have been subject to differing levels of fertilisation, and to variation in the frequency and the scale of nutrient supplied. Against the naturally soft-water, relatively P-deficient water of Blelham Tarn, it is not surprising that manipulations of the phosphorus and the carbon content of the enclosed water should have yielded the most satisfying outcomes. Reynolds (1986b) contrasted the yield of phytoplankton, in terms of biomass and species composition, through six enclosure-seasons

0 0

Post a comment