Limnol. Oceanogr., 38(6), 1993, I 162-1178 Society of Limnology 0 1993, by the American

and Oceanography,

Inc.

Uptake, exchange, and excretion of orthophosphate in phosphate-starved Scenedesmus quadricauda and Pseudomonas K7 Mats Jansson Department

of Physical Geography,

University

of Umea, S-90 1 87 Urn&,

Sweden

Abstract Phosphate uptake was investigated in batch cultures of P-starved green algae (Scenedesmus quadricauda) and bacteria of pelagic origin (Pseudomonas K7). Phosphate transport in Scenedesmus was characterized by initial rapid surge uptake enhanced by P starvation, followed by subsequent slower uptake. Two systems, defined as high-afftnity (HA) and low-affinity (LA) systems, were found in Scenedesmus. The HA system had high affinity for low phosphate concentrations and bound phosphate in exchangeable form in the cell. Exchangeable phosphate was transformed to nonexchangeable phosphate by the LA system. The phosphate uptake in Pseudomonas followed Michaclis-Mcnten kinetics and no exchangeable phosphate could be detected. Pseudomonas outcompeted Scenedesmus for low concentrations of phosphate and also for high concentrations if glucose was present. Scenedesmus took up phosphate at about the same rate as Pseudomonas at moderately low concentrations in mixed cultures without glucose. The phosphate transport by the HA system in Scenedesmus was not affected by competition with Pseudomonas at very low phosphate concentrations. It is suggested that the HA system makes the algae competitive with bacteria for low concentrations of phosphate. Both Scenedesmus and Pseudomonas excreted considerable amounts of orthophosphate, which release is suggested to be connected with growth.

The trophic state of aquatic ecosystems depends on the supply and concentration of limiting nutrients and P is generally the primary factor limiting growth in freshwater. The principal external P input to the aquatic food web of pelagic ecosystems comes from assimilation of dissolved orthophosphate by algae and bacteria (see Jansson 1988). Thus, the mechanisms for phosphate uptake should be crucial for lake plankton. Studies of phosphate uptake and turnover in planktonic populations have usually been carried out by adding minute amounts of phosphate in the form of 32P04. With that approach it has been possible to demonstrate very rapid assimilation of phosphate by plankton (e.g. Rigler 1956; Lean 1973a,b). Moreover, it appears that initial uptake of P takes place mainly by bacteria (Rigler 1956; Currie and Xalff 1984a, b; Chrost and Overbeck 1987) and that the turnover of phosphate can be very rapid due to excretion of orthophosphate (Lean 1973a,b). Few attempts have been made to analyze how microalgae obtain substantial amounts of Acknowledgments I thank Cecilia Palmborg for technical assistance throughout this investigation. Test algae and bacteria were obtained with the assistance of Goran Samuelsson and Agneta Nordstrom. This study was financially supported by the National Swedish Science Research Council.

P from the extremely low concentrations of orthophosphate in P-limited systems in the presence of superior competitors such as bacteria. It has been suggested that organic-P compounds are essential sources of phosphate for algae (Currie and Kalff 1984~; Chrost and Overbeck 1987; Cotner and Wetzel 1992). However, organically bound phosphates must be transformed to orthophosphate (e.g. by the action of extracellular hydrolytic enzymes before being available for algal uptake, Jansson 1988). Thus, the potential availability of regenerated phosphate still depends on the phosphate uptake efficiency of a given organism relative to competing species. Tn two review articles (Cembella et al. 1984a,b) it was reported that nearly all conclusions concerning phosphate uptake in microalgae are drawn from studies of uptake kinetics in continuous or batch cultures, while investigations of the biochemical nature of phosphate transport in microalgae are lacking. Phosphate uptake can bc explained mostly by a single transport mechanism that depends on light-generated energy and whose transport capacity is enhanced by P starvation. Short-term phosphate uptake in P-starved algae can be up to 100 times higher than the maximal growth rate (Harrison et al. 1989). However, it is also obvious that different algal species have widely different affinities for low concentrations of

1162

P uptake in algae and bacteria phosphate and highly variable rates of specific uptake at different phosphate concentrations (Cembella et al. 1984a; Olsen 1988). Phosphate uptake by aquatic bacteria is even less well investigated than microalgal uptake, but it is generally accepted that bacteria are more efficient than algae in taking up phosphate from low concentrations in the medium (Jansson 1988). Phosphate transport in procaryotes has been defined by extensive studies with Escherichia coli (Rosenberg 1987). In this organism, phosphate uptake is regulated by two separate transport systems, one of which is induced by P starvation while the other is constitutive. A similar uptake has also been reported from studies with the eucaryotic fungus Neurospora crassa (Burns and Beever 1979). From the great variability in phosphate uptake efficiency among different species and groups of organisms it is obvious that existing models of phosphate turnover in pelagic planktonic populations are oversimplified. For example, modern textbooks cite the turnover model of Lean (1973a, b) where phosphate-assimilating organisms are characterized as one entity, i.e. particulate P. It is also obvious that before such models can be developed and refined, more basic knowledge on the regulation of phosphate fluxes in and out of planktonic organisms is needed. The present study reports results on phosphate uptake by the chlorophyte algae Scenedesmus quadricauda and the bacterium Pseudomonas sp. The original purpose of the study was to examine the uptake of orthophosphate in algae with methods similar to those used to characterize the phosphate transport in E. coli. The major hypothesis was that, by using a combination of radioactive and nonradioactive phosphate in the algal medium, it should be possible to determine whether phosphate uptake by the algae followed simple MichaelisMenten kinetics as is generally reported from algal uptake studies or if it involved several transport mechanisms. When it became evident that phosphate uptake by Scenedesmus involved a transport mechanism with a high affinity for low concentrations of phosphate, the purpose was extended to include studies on how the high-affinity system could help the algae to compete with bacteria for low concentrations of phosphate. Characterization of the phosphate uptake mechanism in Pseudom-

1163

onus was carried out in less detail than with Scenedesmus because the major reason for including Pseudomonas was to obtain a tool for investigating the phosphate uptake efficiency of the algae Scenedesmus when forced to compete with a bacterium for P. However, efforts to develop procedures for coincubation of algae and bacteria included explicit studies of the phosphate uptake in Pseudomonas.

Materials and methods Experiments with S. quadricauda-S. quadricauda for the phosphate uptake study was obtained from the Institute of Physiological Botany in Umea. Stock cultures of S. quadricauda were kept in a growth medium prepared according to Kuhl (1962). Stock algae were regularly transferred to fresh medium to keep them in the logarithmic growth phase before use in different experiments. Algal growth in the stock culture was checked by measuring the light transmission of the culture at 750 nm every second day using a spectrophotometer. Stock cultures were incubated in 500-ml Ehrlcnmeyer flasks on a shaking table at 20°C. Irradiance was altered every 12 h between an intensity of 70 and 80 PEinst m-2 s-l and darkness. All experiments were conducted with algae starved for P for different lengths of time. Starvation was initiated by centrifuging the algae, whereafter the supernatant was discarded and the algae were resuspended in algal medium without phosphate. During starvation the algae were incubated at the same temperature and light conditions as the stock cultures. At the start of uptake experiments of P-starved cultures, 20-30 mg of algal C per liter in a total volume of 100 ml were transferred to plastic beakers. Phosphate uptake and turnover experiments with Scenedesmus were conducted according to two different experimental strategies. In the first, P-starved algae were allowed to take up phosphate from a medium containing nonradioactive phosphate plus trace amounts of radioactive phosphate (carrier-free 32P, Amersham). The total radioactivity in these experiments was l-5 x lo4 cpm x ml-l (Cerenkov radiation in water solution). The uptake and turnover of phosphate was then studied by following 32Pincorporation into the algae under different experimental conditions. Up-

1164

Jansson

take was studied at concentrations between trast to the first technique, this method permits 0.003 and 13 PM to determine the kinetic charthe study of phosphate uptake and the fate of acteristics of phosphate uptake in Scenedes- assimilated P at phosphate exhaustion, i.e. unmus. der conditions simulating those of P-limited Working with phosphate concentrations natural aquatic systems. The turnover of phoswhich correspond to V,,, of the algae, i.e. the phate was followed by measuring the variation concentration at which the algae express maxof radioactivity in cell-free filtrates (Millipore, imum uptake rate, this technique permits study 3 pm, nitrocellulose) of the experimental culof phosphate uptake in a situation where the tures. Samples of 8 ml were removed from the rates of different processes are not limited by culture with a syringe and filtered. The first 3 low phosphate concentrations in the surroundml of filtrate were discarded and the remaining ing medium. The uptake and metabolism of 5 ml were collected in 20-ml plastic scintilla32P was assumed to be identical to that of nontion vials. radioactive phosphate which permitted calThe uptake rates and phosphate exchange culation of P fluxes in relation to standardized (see below) were determined at different times cell parameters (e.g. cell C). after the algae were transferred to phosphateIn all experiments with this technique, phosfree medium. Experiments were conducted phate uptake and efllux in the algae were cal- with algae starved for 0, 2, 5, 7, 9, 12, 14, and culated by measuring the radioactivity (Ceren16 d, respectively. Since this test showed that kov radiation) of algae collected on membrane the algae achieved maximum uptake rates after filters (Millipore, 3-pm nitrocellulose). Sam- 5-9 d of P starvation, all of the following exples of 5 ml were taken from the experimental periments with P-starved Scenedesmus were beakers with a syringe. The sample was then carried out with algae preincubated for 6-8 d in phosphate-free medium. filtered with a syringe-mounted filter. The filExchange of P, i.e. the substitution of celltering equipment was rinsed with 5 ml of fresh bound P for phosphate in the medium, was phosphate-free medium, after which the filter studied with both techniques. This exchange was removed and placed in a 20-ml plastic turned out to be a useful means for characterscintillation vial containing 5 ml of distilled izing one specific transport system in Scenewater. desmus, so the determination of exchange beSince the interpretation of these experiments depends on accurate determination of came a “standard procedure” that was used in most of the uptake experiments. The algae were the amount of phosphate retained together with the algae on the filters, the adsorption of phos- allowed to assimilate radioactive phosphate, after which a large excess of nonradioactive phate by the filters was carefully investigated. Phosphate adsorption of the filters in the con- orthophosphate (usually 10 mM) was supplied centration range used in this study was con- to the experimental culture. Under these conditions, a small amount of the assimilated rasistently < 1% of the amounts of dissolved dioactive phosphate was more or less immephosphate in the filtered solution. Filter blanks were included in each separate experiment and diately returned to the medium. The amounts subtracted from the total P retention of the and rates of this type of P release from algal cells under different conditions could thereby filters. In the second technique, starved algae were be examined. This type of “pulse-chase” experiment was incubated with minute amounts of 32P04 and carried out at different stages during the uptake without nonradioactive phosphate. The total phase in experiments with P-saturated algae radioactivity in these experiments was also l5 x 104 cpm x ml-*. The use of 32P alone (first technique) and algae incubated with 32P does not allow relevant estimates of the or- only (second technique). In total, 46 uptakethophosphate concentrations used in the ex- exchange tests were carried out with the first periments. However, the 32P activities used in technique, and 18 experiments were made with this technique were well below the lowest con- the second technique. In addition to rapid P exchange, the algae centration (0.003 PM) used in the experiments also excreted P at a slow rate. This excretion at defined phosphate concentrations according to the first technique (cf. Rigler 1956). In con- was examined by labeling algae with 32P fol-

P uptake in algae and bacteria lowed by subsequent incubation in high concentrations of nonradioactive orthophosphate. 32P release of the algae was then followed for 17 d. Phosphate additions (10 mM) were maintained at the start of the experiment and on day 3. Algal biomass was determined on days 1, 4, 7, 9, 14, and 17. The experiment was run on three parallel cultures. Dissolved radioactive P was fractionated by gel filtration with Sephadex G-25 to see if exchanged or excreted P occurred in more complex structures than orthophosphate. NaCl (0.1%) plus NaN, (0.02%) were used as eluant. The separation was performed before and after exchange and excretion experiments began. The column volume (void volume) was 60 ml and the flow rate was 0.5 ml min- l. Fractions of 5 ml were collected for determination of 32P activity. Separations were done on four occasions in connection with uptake-exchange experiments with both techniques and on days 6, 9, and 17 in the excretion experiment. The dependence of P uptake and exchange on irradiance was examined in a set of experiments comparing uptake and exchange of P in darkness and light. Three dark treatments were used. Separate cultures were kept in the dark for 24 h, 5 h, and 10 min before and during the uptake experiments. A fourth culture (control) was kept in light 1 d before and during the entire exchange-uptake experiment. Experiments with Pseudomonas - Pseudomonas strain K7 was obtained from the Institute of Microbiology, University of Umea. The bacterium was isolated from the pelagic zone of the brackish water (salinity 3-47~) of the Bothnian Sea in northern Sweden. Pseudomonas was grown in MOPS-medium (Neidhart et al. 1974) with glucose as the C source. P starvation was induced by centrifuging the bacteria followed by discarding the supernatant and resuspending the pellet in MOPS medium without phosphate. As for Scenedesmus, the stock cultures and the starved bacteria were grown in Ehrlcnmeyer flasks on a shaking table at 20°C. Stock and starvation cultures were kept in the dark. P-starved bacteria were transferred to plastic beakers immediately before the start of phosphate uptake experiments. The concentration of bacteria in the experiments was between 20 and 30 mg of bacterial C liter-’ in a total volume of 100 ml. The uptake of phosphate was compared af-

1165

tcr 3, 5, 18, 24, and 48 h of incubation in phosphate-free medium plus glucose. Maximum specific uptake capacity was obtained after 3 h of starvation and it remained the same for at least the following 2 d. In all experiments with starved Pseudomonas, the bacteria had been P starved for 24 h. It was found that the presence of glucose was a prerequisite for efficient uptake. Therefore, phosphate uptake at different concentrations was compared for bacteria that had been P starved for 5 h and glucose starved for 1, 3, and 5 h. Exchange and excretion of phosphate by P-starved bacteria were tested with the same methods used for Scenedesmus,with and without glucose in the test medium. Characterization of released phosphate was done with gel filtration with Sephadex G-25, according to the same procedure described for Scenedesmus.All experiments were carried out by incubating phosphate-starved organisms together with defined concentrations of nonradioactive phosphate plus trace amounts of radioactive phosphate (the first technique described for the experiments with Scenedesmus). Filters used to separate bacteria from medium were nitrocellulose 0.45-pm membrane filters (Millipore) which completely retained the experimental bacterium. Filter blanks were checked as for

Scenedesmus. Phosphate uptake in mixed cultures of Scenedesmus and Pseudomonas- In one set of experiments, bacteria were starved of phosphate for 2 d in a bacterial medium with glucose. Algae were starved for phosphate in the filtrate from a bacterial growth medium for 6-8 d. Immediately before the uptake experiment, the algae and bacteria were centrifuged and resuspended in bacterial medium plus glucose and the phosphate uptake experiment was initiated according to the first technique The concentrations of both algae and bacteria were 20-30 mg C liter-l in a total volume of 200 ml. These tests were designed to test competition between bacteria and algae when both organisms had an ample supply of energy. Each uptake experiment included controls where the P uptake of the bacteria and algae per se were compared with the uptake of each type of organism in the mixed cultures. In order to separate algae and bacteria in these experiments, I first collected the algae on 3-r.LrnNuclepore filters, after

1166

Jansson

which I collected the bacteria in the 3-pm filtrate on 0.45-pm membrane filters. Retention of algae and bacteria on filters was checked by microscopic counting. The test algae were completely removed by the Nuclepore filter, whereas 95-100% of the bacteria passed the filter. The 0.45-pm filter retained 100% of the bacteria. To better simulate conditions in pelagic systems, where bacteria are dependent on energy in the form of organic compounds supplied by the planktonic community, I carried out another series of experiments. The algae were starved for 6-8 d in medium without phosphate, whereafter 100 ml of algal suspension was transferred to plastic beakers. At the same time, the bacteria were -starved for 1 d in a mixture (1 : 1) of bacterial medium without glucose and phosphate-free filtrate from phosphate-starved algal cultures. After 2 d, the bacteria were centrifuged and resuspended in bacterial medium plus phosphate, whereafter 100 ml of this suspension was immediately mixed with the starved algae. The concentrations of bacterial and algal biomass in these experiments were 20-30 mg C liter-’ for both types of organisms. The bacteria were starved in a mixture of bacterial medium and algal medium filtrate because it was considered necessary to avoid glucose in the uptake experiments to better simulate conditions in natural populations of plankton where the main source of energy for bacteria is organic algal exudates. The need for this precaution was stressed by the fact that glucose increased phosphate uptake efficiency of the bacteria. The uptake experiments were then carried out by the first technique at different phosphate concentrations (0.03-l 3 PM) as well as the second technique (incubation with 32P only). Each uptake experiment included controls where the uptake of the bacteria and algae per se was compared with the uptake of each type of organism in the mixed cultures. Separation of algae and bacteria was done with 3-pm Nuclepore filters and 0.45~pm membrane filters, respectively. Analyses -The C concentrations of algal and bacterial suspensions were determined in a CHN analyzer (Carlo-Erba model 1106). Suspensions of 0.5 ml were pipctted into tin vials used for subsequent combustion in the C analyzer. Cell-free filtrates of algal and bacterial

suspensions were also analyzed to determine the contribution from dissolved C. The contribution of dissolved compounds to the total C concentration of the cultures was negligible. 32P activity (Cerenkov radiation in water solution) was analyzed in a liquid scintillation counter (Beckman Instr.).

Results P transport in S. quadricauda-

A typical curve of P incorporation vs. time for P-starved algae is shown in Fig. 1. In this experiment the cell C concentration was 22 mg C ml-l and the P concentration in the medium was 0.03 PM. Figure 1a shows incorporation apparently following Michaelis-Menten kinetics which is common in algal P-uptake experiments (e.g. see Cembella et al. 1984a,b). However, using tracer amounts of radioactive P together with nonradioactive P made it possible to follow the P uptake in detail (Fig. lb) and clearly demonstrated that uptake during the initial phase proceeds at two different velocities-a rapid initial rate followed by a slower one. This pattern was very reproducible (see Fig. 3) and occurred in all types of experiments where P-starved Scenedesmus was transferred to P-rich medium (46 comparable uptake tests in total). It is difficult to estimate accurately the kinetic characteristics for this initial rapid uptake because it was generally represented by only one to three points on the uptake curve. For an approximate calculation it was assumed that this uptake proceeded at a linear rate until the “break” in the uptake vs. time curve which occurred l-2 min after uptake began. With this assumption, uptake vs. concentration could be plotted separately for the initial rapid rate “O2 min” and the subsequently slower uptake “2-l 5 min” (Fig. 2a,b). The use of superscripts, which denote the time intervals over which determinations are made, is according to the recommendations of Harrison et al. (1989) for characterizing nonlinear uptake. V,,, (pg P mg C- l min- l) and K, (pg P liter - ‘) for both types of uptake were calculated from an Eadie-Hofstee plot (uptake rate vs. uptake rate/P concentration) of the data in Fig. 2a and b (Table 1). The kinetic constants were used in the Michaelis-Menten equation to construct the curves in Fig. 2 showing that both rapid and slow uptake followed Michaelis-Menten

1167

P uptake in algae and bacteria

Incorporated P h!p mg w

Incorporated P (ug p mg 0

b

0

20

40

60

80 100 120

0

(mm Fig. 1.

Phosphate incorporation

2

4

6

8

I

10 12 14

I -

16

bw in phosphate-starved

Scenedesmus quadricuudu.

Note different

scales.

in Fig. 3. In these experiments P-starved algae kinetics. Obviously, the rapid initial uptake were allowed to take up nonradioactive phosreflects a phosphate-transport system with high phate together with trace amounts of 32P. The affinity for low concentrations of phosphate. effect of further addition of nonradioactive In Table 1 this system is therefore called the HA system (high-affinity system) while the sys- phosphate was then followed by studying the tem which maintains the slower transport is 32P dynamics. The introduction of a large excess of phoscalled the LA system (low-affinity system). The phate caused rapid release of a well-defined values for the HA system could be underestimated because uptake of phosphate by the amount of P from the algal cells. Two interHA system was very fast and the uptake rate esting features of this release are demonstrated was characterized by only a few observations in each experiment. It is thus possible that KS for the HA component is lower and V,,, higher Table 1. Maximum specific uptake (V,,,) and halfthan the values in Table 1. That in turn means saturation constants (K,) calculated for HA and LA phosphate uptake in Scenedesmus quadricauda. that the difference in affinity for low concentrations of phosphate between the two systems K (ugvEng can be greater than could be demonstrated in kg p C I min ‘) liter ‘) this investigation. HA uptake (O-2 min) 0.15 0.18 The effect of adding a large excess of orthoLA uptake (2-15 min) 0.31 68 phosphate during phosphate uptake is shown

Jansson

1168 P uptake @g P mg C’ min”)

a

lime (mill) 0.02

0

0.04

0.06

0.08

0.10

0.17

P concentration (MM)

Fig. 3. Phosphate exchange in Scenedesmus quadricaudu. Control culture-a; parallel cultures subjected to excess phosphate added at different times after onset of uptake-O, 0, A.

P uptake @g P mg C’ mid) 4 .

0.3 -I

b

oil, O

1

2

3

4

5

6

7

8

I 9

I 10

I I 11 12

I13

P concentration @M)

Fig. 2. Phosphate uptake vs. phosphate concentration in Scenedesmus quadricaudu. [a.] HA system (O-2 min). [b.] LA system (2-15 min). Data were fitted to the Michaelis-Menten equation with the kinetic constants in Table 1.

in Fig. 3. First, the amount of P released is seen to be very similar to the amount of P fixed by the HA system (- 1 pg P per mg cell C). Second, it appears that the amount of exchangeable P is about the same at different times during the linear phase of the slow phosphate incorporation maintained by the LA system. The close connection between the amount of P fixed by the HA system and the amount of exchangeable P is emphasized by the fact that the mean ratio between the amount of phosphate fixed by the HA system and ex-

changed phosphate was calculated to be 0.86 (SD: kO.37) based on data from all 46 uptakeexchange experiments. Figure 4 describes P uptake by the HA system and the LA system as a function of the degree of P starvation prior to the uptake test. Both systems have a similar dependence on P starvation with the maximum uptake capacity occurring 5-l 0 d after the onset of starvation. However, the HA system increases its capacity much more during phosphate starvation than the LA system. The maximum rates of phosphate uptake in Scenedesmus were obtained with algae starved of P for 5-l 0 d. It should be stressed that uptake of phosphate was tested with a very low phosphate concentration (0.0 15 PM) in the medium; therefore the HA system was saturated (at V,,,) while the LA system was not. If both systems had been saturated, the difference in uptake rate between the HA and the LA systems should have been much smaller (cf. Table l), which explains the difference in uptake between the two systems in Fig. 4 and stresses the important role of the HA system at low concentrations in the medium. The uptake of P by the HA system and the LA system after incubation of the algae in different light-dark conditions is shown in Table 2. Uptake in the light is compared with uptake where the algae were incubated in the dark for different lengths of time before the uptake experiments. This test showed that the absence

1169

P uptake in algae and bacteria Puptake

0.16

(UgP

mg C’

Table 2. Phosphate uptake by the HA and LA systems in Scenedesmus quadricauda incubated in the dark. The algae were incubated in the dark for different lengths of time before phosphate uptake started. The algae were kept in the dark during the uptake experiment except for the control. The phosphate concentration was 10 PM and the uptake rates of HA and LA uptake of the control were 0.16 and 0.29 bg P mg C-l min-’ respectively. All uptake rates are given as a percentage of that in the control.

miN’)

t

Dark

I I---,

4

6

8

10

. --12

14

16

Days of P starvation

Fig. 4. Phosphate uptake following addition in Scenedesmus quadricauda phosphate starvation time. High-affinity affinity system-o.

a 0.015 PM PO, as a function of system-O; Low-

of light slowed the uptake of both uptake systems and that retardation of phosphate uptake became more pronounced the longer the algae were incubated in the dark. The LA system appears to be affected more rapidly than the HA system by the absence of light. The series of experiments where P-starved algae were incubated with very low concentrations of phosphate in the form of 32P (second technique) showed that under these conditions all phosphate in the medium was rapidly assimilated (Fig. 5). Exchangeable phosphate was also present under these conditions (Fig. 5) and the amount varied as the uptake proceeded. In total, 18 identical exchange tests were carried out at different phases of the phosphate uptake curve. Figure 6 demonstrates how the amount of exchangeable P in these experiments was related to the amount of P taken up by the cell. The quantity of exchangeable P was greatest when - 50% of the 32P043- in the medium had been bound to the cells. When the amount of exchangeable P at any given time during the uptake phase was plotted against total uptake in the algae, it was demonstrated that in early stages of uptake all or nearly all cell-bound P was exchangeable, whereas at the end of uptake

before uptake

0 min

10 min

5h

24 h

100 100

95 35

73 37

54 21

10 14

HA uptake LA uptake

2

period

Control

none, or very little P, was in exchangeable form. When half of the added 32P had been bound to the cells, -50% of the cell-bound P was exchangeable. The results shown in Fig. 6 can be used to illustrate the principle transfer of P from the medium to exchangeable and nonexchangeable cell P. The curves in Fig. 7 are constructed with the data shown in Figs. 5 and 6 and demonstrate the relative distribution of different P compartments during nonsaturated uptake in P-starved algae. Figure 8 shows the slow release of 32P from algal cells grown in phosphate-rich medium. The release of P was almost identical in three parallel cultures. In one of the cultures, the D&solved

“P (cpm x ml-‘)

4omQrr&c.ange----6 ----‘--oj-----o phosphate 3oooo --a-

---

-*-----+-----d

Exchange phosphate

Tie

(min)

Fig. 5. Incorporation and exchange of 32P0,3- in Scenedesmus quadricauda. Incorporation is described by the decrease of dissolved 3zP in the medium. Control culture-0; parallel cultures subjected to excess nonradioactive phosphate added at different times after onset of uptake-0.

Jansson

1170

Exchangeablealgal P

90-

‘1 t

80-

b

Time (relative units)

\’ \

70.

\

60-

3.

\

50-

\ A ‘\;

Ml30-

l

\

20-

l\ .‘.

l \

\

lo01 0

’ 10

I.’ 20 30

’ 40

’ 50

’ 60

’ 70

’ 80

\*

\ ’ 90 11 1

‘*P algal uptake (% of total 32P) Fig. 6. Relationships between the amounts of exchangeable P and the whole-cell uptake of P in Scenedesmus quadricauda. Curves are drawn by inspection.

phosphate concentration was twice as high as in the others. Since it did not affect the outcome of the experiment, the algae were P saturated in all cultures. After the initial rapid release due to exchange (not shown in Fig. 8), a slow release of cell-bound P took place, until all of the cell 32P had been returned to solution after 17 d. The release rate appears to increase toward the end of the experiment. However, after the first addition of phosphate, the algae started to grow at a rate that doubled the biomass (determined as cell carbon) every 5th day throughout the experiment. When the phosphate release rate was divided by the biomass, it can be seen that the specific release rate was more or less constant during this experiment.

Fig. 7. Relative distribution of total algal P, exchangeable algal P, and nonexchangeable algal P during phosphate uptake in Scenedesmus quadricauda. Curves are constructed with data from experiments with very low concentrations of phosphate in the medium (second technique), meaning that phosphate in the medium is more or less completely transferred to the algae during uptake. Total cell P incorporation is according to the typical pattern shown in Fig. 5. Values for exchangeable and nonexchangeable P in relation to total algal incorporation are obtained from the data in Fig. 6. Curves are drawn by inspection.

Dissolved radioactive P was characterized by gel filtration with Sephadex G25 in shortterm exchange experiments (five occasions) and on three occasions during the long-term excretion experiments (days 6, 9, and 17). In all of these separations, dissolved 32P eluted in the orthophosphate fraction only, which implies that both exchanged and excreted P is in the form of orthophosphate. Other forms of P (e.g. colloidal P which regularly appears in shortterm turnover experiments with 32P)are formed outside the cell (Lean 1973a,b). Their formation may have been prevented in these experiments by the large dilution of radioactive P with nonradioactive P. Phosphate uptake in Pseudomonas-The specific uptake rates in Pseudomonas were dependent on the presence of glucose in the growth medium. The uptake rates for bacteria starved of both P and glucose 2 d prior to uptake was only 30-60% of the rate in bacteria starved for phosphate only (Fig. 9). Phosphate exchange could not be found in Pseudomonas, but substantial excretion of or-

1171

P uptake in algae and bacteria Incorporated P @v3p mg 0

VQin medium (cpm ml-l) Meal

4ooo lolm 24m

0, 0 0

0

20

30

40

60

50

70

Time (min) DAY

Fig. 8. Phosphate excretion in Scenedesmus quadricauda. Phosphate-starved algae were allowed to take up [32P]phosphate at the start of the experiment. An excess of nonradioactive phosphate was supplied after 10 min and on day 3. The three curves represent three parallel cultures showing excretion in cpm ml--l and the bars rcpresent specific excretion expressed as cpm mg C I d-l.

thophosphate

10

similar

to that described

for

Scenedesmus was also a regular phenomenon in Pseudomonas cells. The typical turnover of 32P in Pseudomonas cultures after labeling with 32P and subsequent addition of nonradioactive P is shown in Fig. 9. It can be seen that not only the uptake rates but also the excretion rates were dependent on the presence of glucose in the medium. Initial rapid, followed by slower, uptake of the type observed in Scenedesmus was not found in Pseudomonas. This finding implies that different P-transport systems do not exist or are not activated in this bacterium during the prevalent experimental conditions. It is, however, possible that the bacterium has a multiphasic uptake which was not detected with the methods used. A technique which permits more frequent sampling during the initial phase of uptake is needed to test this possibility. Failure to detect any exchangeable phosphate in Pseudomonas, however, shows that an uptake system similar to the HA system in Scenedesmuswas not expressed in Pseudomonas. Interpretation of the following experimental results does not depend on whether P uptake in Pseudomonas is a result of one or several transport mechanisms.

Fig. 9. Phosphate uptake and excretion in Pseudomonas K7. Control culture grown in the presence of glucose-0; culture grown in the presence of glucose and subjected to an excess of phosphate added 15 min after phosphate uptake began-O; control culture grown without glucose- x; culture without glucose and subjected to excess phosphate added 16 min after phosphate uptake began - 0.

Phosphate uptake in mixed cultures of Sceneand Pseudomonas- When algae and

desmus bacteria ence of min-l)

were incubated together in the presglucose, the uptake rate (pg P mg C-l of Pseudomonas exceeded that of Scenedesmusby - 15-25 times within the tested phosphate concentration interval (Table 3). The uptake rates in Table 3 represent wholecell uptake calculated for the initial 10 min of uptake. Uptake by the algae in Table 3 is lower than that expected for the phosphate concentration in the medium (cf. Fig. 2). A possible Table 3. Phosphate uptake (pg P mg C-l min-l) in Scenedesmus quadricauda and Pseudomonas K7 incubated together at different phosphate concentrations and in the presence of glucose (0.4%). Incubation was in bacterial medium. The controls were algae and bacteria incubated in filtrates of bacterial medium. P concentration

(wM)

1.8

2.0

7.0

10.0

Scenedesmus alone (control) with Pseud.

0.1 0.1

0.09 0.07

0.08 0.06

0.09 0.09

Pseudomonas alone (control) with Seen.

1.3 1.3

1.3 1.2

0.9 1.3

1.8 1.9

Jansson

1172 P uptake (JlgP mge

‘P uptake (cpm mg C-‘)

min-‘)

1 Control ----i--

.

3x10’

Pseudomonas

-‘-

Wx

0

Scenedesmus

Pseudomonas

strain

x

A

K7

quadricauda

x ,

,

1

,

1

2

3

4

,

,

,

5

6

7

P concentration

,

,

,

,

8

9

10

11

,

,b

lx,o’-

fl

12 13

(J&l)

Fig. 10. Phosphate uptake in Scenedesmus quadricauda and Pseudomonas K7 in mixed cultures at different concentrations of phosphate and without glucose. Data were fitted to the Michaelis-Menten equation with the kinetic constants in Table 1 (Scenedesmus) and V,,, = 0.35 pg P mg C-l min-’ and K, = 95 pg P liter-’ (Pseudomonas).

explanation is suppression of algal uptake caused by products released by the bacteria since, for the experiments reported in Table 3, starvation of the algae preceding the experiment was carried out in a filtrate from a bacterial growth medium. Control experiments showed that this treatment slowed P uptake by 50-70% compared with starvation in a pure algal medium or a pure bacterial medium. Suppression of algal uptake was not detected in mixed cultures unless the algae had been preincubated in filtrates from bacterial cultures. The pretreatment used in this experiment was therefore abandoned. However, the results in Table 3 can still be used to illustrate that Pseudomonas in the presence of glucose is more efficient than Scenedesmus in taking up phosphate. In contrast, when uptake experiments were made without glucose, the specific uptake rates of algae and bacteria were more or less identical within the tested concentration interval (Fig. 10). The uptake in control cultures of each organism was the same as in mixed cultures at each tested concentration. Therefore, down

OK

I..L h

-*-m-x-

-+--X-

I 10

, 12

I 14

together I I 16 18

with Pseudomonas ---x--&j I I I I 1 20 22 24 26 28

IW 30

Time (min)

Fig. 11. Phosphate uptake in Scenedesmus quadricauda and Pseudomonas K7 in single and mixed cultures. Phosphate was supplied in the form of 32P only. Total radioactivity was 3.6 x lo6 cpm per mg algal and bactcrium C.

to medium phosphate concentrations of -0.3 pg P mg-l cell C, phosphate uptake by bacteria and algae did not suffer from the presence of the other organism -both organisms had a similar capacity to take up P. Thus, the competitive ability of Pseudomonas at moderately high concentrations of phosphate (Table 3, Fig. 10) appears to be dependent on energy supply for the bacterium. A different picture was obtained when the experiment without glucose was repeated with extremely low concentrations of phosphate in the form of 32P0 3- only (Fig. 11). P incorporation into bac&ia was considerably faster than into algae and obviously was unaffected by the presence of the algae. Pseudomonas took up all the phosphate from the medium in - 10 min. Incorporation of phosphate in Scenedesmus was considerably lower when the algae were incubated with Pseudomonas compared to cultures with only Scenedesmus. Essentially all incorporation of-phosphate in Scenedesmus in mixed cultures took place during the first 2 min after phosphate addition. Both the amount

1173

P uptake in algae and bacteria

Table 4. Incorporation of 32P043- in Scenedesmus in cultures containing only Scenedesmus and in those with Scenedesmus and Pseudomonas. Incorporation is given for the periods O-2 and 2-10 min after addition of 32P. Incorporation is expressed as cpm mg C-I min- I and given as mean values for five parallel experiments. Scenedesmus

Scenedesmus Uptake

Uptake

O-2 min

0.16x lo6 n=5 SD = +0.05x

10”

2-10 min

0.12x 106 n=5 SD = *0.03x

lo6

and rate of phosphate fixation by the algae during this initial uptake phase was the same in Scenedesmus whether the bacteria were present or not. Five parallel experiments were conducted with the same results as shown in Fig. 11. Incorporation of phosphate in Scenedesmus in all five experiments is summarized in Table 4.

Discussion Algal phosphate uptake is generally characterized as a single Michaelis-Menten uptake system (see Cembella et al. 1984a; Jansson 1988). In a few cases, multiphasic (mostly biphasic) uptake has been reported for different species e.g. Chlorella pyrenoidosa (JeanJean 1976), Euglena gracilis (Chisholm and Stross 1976a, b), Olisthodiscus luteus (Tomas 1979), Pyrocystis notiluca (Rivkin and Swift 1982), and Staurastrum leutkemuellerii (Olsen 1988). However, none of these investigations reported any attempt to characterize P uptake by means other than kinetic parameters, hence decisive information on the existence of separate transport systems is lacking. As pointed out by Cembella et al. (1984a, p. 340): “. . . multiphasic uptake does not necessarily imply the existence of multiple biochemically distinct carrier molecules although this possibility is certainly not precluded.” The phosphate uptake in P-starved Scenedesmus demonstrated here is similar to that obtained in uptake experiments with other algal species, e.g. enhanced uptake after P starvation, light dependence, and apparent Michaelis-Menten kinetics. However, important characteristics of the HA and LA systems found here have not been reported for algae and resemble those of dual phosphate transport systems described for bacteria and fungi. The most striking resemblance is the simultaneous and possibly conjugated operation of an HA and LA system in the same way as in E. coli (Ro-

Uptake

O-2 min

0.15x 10” n=5 SD = kO.03 x lo6

and Pseudomonas Uptake

2-10 min

no incorporation n=5

senberg 1987) and Neurospora (Burns and Beever 1979). Thus, substantial evidence is given here that high- and low-affinity P transport in Scenedesmusis linked and that the role of the HA system is to bind phosphate to the cell, especially at low concentrations of phosphate in the medium, while the LA system maintains the internal transfer of phosphate. The strongest evidence that high- and low-affinity transport in Scenedesmusis linked comes from the experiments for uptake from extremely low concentrations of phosphate (Figs. 5-7). Phosphate is first fixed by the HA system in exchangeable form, whereafter it is transformed to nonexchangeable P by the LA system. Incorporation of nonexchangable P (LA uptake) shows an initial lag and then proceeds at an approximately exponential rate (Fig. 7). The increase in nonexchangeable P per time unit is greatest, i.e. LA uptake is most rapid, when the amount of exchangeable P is highest. This finding indicates that the HA and LA systems are closely linked, and that the P incorporated in nonexchangeable form by the LA system comes from the pool of exchangeable P filled by the HA system. Nonexchangeable P does not occur until a certain amount of exchangeable P has been formed. When the phosphate concentration of the medium approaches zero, the pool of exchangeable P becomes exhausted, as it cannot be replenished from the medium, and it is metabolized within the cell. The kinetic constants K, and Vmax(Table 1) give additional information concerning the different roles of the HA and LA systems. The maximal transport rate (V,,) is about the same for both systems. The difference in V,,, between the HA system and the LA system, 0.32 and 0.15, is small compared to the order of magnitude differences reported among species (Cembella et al. 1984a,b). The low K, value for the HA system shows that uptake from very

1174

Jansson

low concentrations in the medium is very efficient. This, plus the fact that the HA system is activated by P starvation (Fig. 4), indicates that the HA system is an adaptation of P deficiency. The kinetic constants also raise questions of how the phosphate content of the pool’filled by the HA system (exchangeable P) is regulated. If the HA system supplies phosphate to a pool of limited size that is considerably smaller than the total phosphate storage capacity of the cell, then there must be a balance between filling and emptying of this pool. At high phosphate concentrations (> 5 PM in this study) this balance is obtained because transport by both systems apparently occurs at about the same rate (Table 1). At extremely low concentrations, as in the experiment of Figs. 5-7, the problem never occurs because the capacity of the pool exceeds the amount of phosphate in the medium. At intermediate phosphate concentrations, however, the pool is filled much faster than it is emptied. For example, where the phosphate concentration of the medium was 0.015 PM and the algal biomass was 2.3 mg C liter- I, HA uptake was 10 times greater than the LA uptake after phosphate starvation for 6-8 d. The transition pool would be filled 10 times faster than it was emptied. Possible ways to regulate uptake in this situation would be if the HA system was turned off after the pool was filled or if both filling and emptying of the pool were maintained by the LA system. However, this does not seem to be the case because adding excess phosphate more or less immediately substitutes the phosphate in the pool which then is returned to the medium (i.e. the capacity to transport P to the pool is still very high). The close connection between the amounts of phosphate fixed by the HA system and the amounts of phosphate released from the algae by exchange (Fig. 3) strongly indicates a close relationship between the HA system and exchangeable P. Exchange can thus be a possible means for regulation of the net uptake to the pool supplied with phosphate from the HA system. If this is correct, the I&,, of the HA system (Table 1 and Fig. 2a) may represent a steady state situation where phosphate is continuously exchanged between the cell and the medium. Such a mechanism can be advantageous since it enables a continuous supply of

phosphate for the LA system. It should be stressed that this discussion on the possible regulation of the content of exchangeable P concerns P-starved algae with an activated HA system. In algae saturated with P, uptake by the HA system and the LA system could not be separated (Fig. 4, day 0) which could be the result of, for example, the HA system being repressed to a rate corresponding to the needs of P-saturated cells or all uptake being maintained by the LA system. However, studies on the uptake in saturated cells must be undertaken before this can be discussed in detail. The phosphate uptake characteristics in P-starved Scenedesmusare similar to the temporally nonlinear incorporation of NH4+ in N-limited marine algae (Conway et al. 1976; Goldman and Glibert 1982; Parslow et al. 1985) and silicate in Si-limited marine diatoms (Davis et al. 1978; Harrison et al. 1989). The characteristics of this uptake are summarized by Harrison et al. (1989): the initial rapid uptake of the limiting nutrient is termed “surge uptake” and the subsequent slower uptake is regarded as an “internally controlled uptake.” The surge uptake fills an internal pool and is controlled by feedback inhibition, i.e. uptake is shut down when the pool is filled. The internally controlled uptake represents nutrient transport to metabolic functions in the cells and proceeds until the cell quota for a particular nutrient is reached. The only clear deviation from this model is that activated surge uptake for phosphate in Scenedesmus does not seem to be controlled by feedback inhibition. The fact that the whole pool filled by surge uptake is rapidly exchangeable with phosphate in the medium indicates that uptake is not inhibited after the pool is filled, but rather that net uptake by the HA system can be controlled by exchange as discussed above. However, it is obvious that the HA system is also controlled by the nutrient state of the algae since the transport efficiency of the system is dependent on the degree of phosphate starvation (Fig. 4). In spite of the close connection, the HA and LA systems are separate mechanisms that are regulated individually. The correspondence between uptake rate and substrate concentration (Fig. 2) would thus have been different if the uptake rate of the LA system was dependent only on the HA system. The LA system

P uptake in algae and bacteria is more susceptible to the absence of light because the LA transport was reduced immediately after transfer of algae from light to dark conditions. The HA system was retarded after 10 min of incubation in the dark, but was unaffected immediately after a switch from light to dark. The HA system is also activated by phosphate starvation to a higher degree than the LA system (Fig. 4). Possible ecological implications of enhanced nutrient uptake (surge uptake) during nutrient deficiency have been discussed previously (McCarthy and Goldman 1979; Goldman and Glibert 1982; Parslow et al. 1985). It has been suggested that surge uptake enables microalgae to exploit short lived nutrient micropatches in their environment (Harrison et al. 1989). Surge uptake also increases the utilization of very low concentrations of nutrients (Parslow et al. 1985), and the HA and LA systems in the fungus Neurospora were shown to permit the organism to acquire phosphate at a constant rate regardless of environmental concentration (Bieleski and Ferguson 1983). Such possibilities would be beneficial in the competition for limiting nutrients in natural plankton assemblages. The results obtained from this investigation can be used to support the role of the HA system in Scenedesmus to obtain P in competition with other planktonic species. The superior efficiency of bacteria in assimilating phosphate at low concentrations has been demonstrated in several investigations with natural populations and mixed cultures of algae and bacteria (Rhee 1972; Friebele et al. 1978; Currie and Kalff 1984a,b,c; Berman 1985). The efficiency of bacteria relative to algae has been attributed to a high affinity for low concentrations due to a high ratio of surface area to volume (cf. Gavis 1976). This explanation neglects the possibility that specific uptake mechanisms can influence competitive ability. Phosphate incorporation in mixed cultures of algae and bacteria in the present study suggests that algae can be more competitive with bacteria than previously thought and that this can be attributed to the efficiency of the HA system. The bacterial : algal biomass ratio on a carbon basis was - 1 : 1 in all experiments with mixed bacterial algal cultures. This ratio approximately represents the relationship between bacterial and algal biomass in oligotro-

1175

phic to moderately eutrophic lakes, i.e. in most P-limited freshwater systems (Persson 1985; Istvanovics et al. 1992). When Scenedesmus and Pseudomonas took up phosphate in mixed cultures, the bacteria were superior when grown in the presence of large amounts of energy-yielding substrates in the medium (Table 3). When energy was derived from algal excretion products, bacteria and algae took up phosphate at about identical rates (Fig. 10). The concentration of bacteria and algae in these experiments was 2-3 mg C liter-l for each group of organisms and the phosphate concentration range was 0.5-l 0 PM. At the low end of this concentration range, biomass : P should be similar to that in natural eutrophic ecosystems. Therefore, at moderate to high phosphate supply, it appears that the superiority of Pseudomonas relative to Scenedesmus in obtaining P depends on the presence of external energy sources. In a situation where energy is supplied solely by algal excretion, Scenedesmus and Pseudomonas were equally efficient in taking up phosphate. This result fits well with the frequent observations that the energy supply is critical for the overall growth and competitive capacity of bacteria relative to algae (Cole et al. 1982; Chrost et al. 1986) and the proposal that bacterial utilization of phosphate can be limited by availability of organic C (Cotner and Wetzel 1992). At extremely low concentrations of phosphate, the situation that best simulates conditions in phosphate-limited pelagic systems, Pseudomonas was much more efficient than Scenedesmus-again in the absence of a defined external source of energy for the bacteria (Fig. 11, Table 4). The result is consistent with the many observations where minute amounts of 32P given to P-limited planktonic populations is incorporated into particles of bacterial size (Rigler 1956; Paerl and Lean 1976; Currie and Kalff 1984a; Berman 1985). In the five parallel experiments, nearly all added phosphate was taken up within 10 min; the algal share was consistently 5-l 0%. All algal uptake was recorded during the first 2 min and the uptake rate during this period was equal to that of the mixed culture and the control with algae only, showing that incorporation into the algae during the first few minutes was not affected by competition with the bacteria. It should be emphasized that these experiments were car-

1176

Jansson

ried out at low concentrations of phosphate, so that phosphate exhaustion in the medium affected uptake, especially in the mixed cultures. Phosphate exhaustion due to bacterial uptake is therefore the most probable explanation for the fact that phosphate incorporation in algae was essentially nonexistent after 2 min. However, P incorporation into the algae during the first 2 min of uptake was not affected, even though about two-thirds of the phosphate in the medium was incorporated into bacteria during this period. If the results of these experiments are compared with the outcome of experiments on the uptake kinetics of Scenedesmus in extremely low concentrations of phosphate (Fig. 7) it is obvious that uptake by Scenedesmus in mixed cultures at low concentrations of phosphate should be totally dependent on the HA system. During the first few minutes after addition of 32P, nearly all phosphate uptake in the algae is due to the HA system. In natural ecosystems, in contrast to the situation in batch experiments with additions of 32P, the turnover of phosphate compounds ensures a continuous supply of orthophosphate which is rapidly reassimilated by algae and bacteria. An activated HA system can here guarantee the algae a supply of phosphate even though uptake by the total bacterial biomass is superior to that of the algae. Excretion of orthophosphate from microorganisms is a well-documented phenomenon (e.g. Goldberg 1948 cited by Cembella et al. 1984a; Lean and Nalewajko 1976), and it appears to occur also in phosphate-limited plankton populations (Lean 1973a,b). The reason is not clear, but it has been suggested that excretion is a means of regulating cell concentration of P (Bieleski and Ferguson 1983; Olsen 1988; Jansson 198 8). By labeling algae and bacteria with 32P, followed by incubation in nonradioactive phosphate, it was demonstrated that both types of organisms excreted orthophosphate (Figs. 8 and 9). This flux of phosphate is clearly different from the rapid exchange of P demonstrated in Scenedesmus, as the slow release from algal cells was the result of a constant excretion of P from actively growing algae. It is possible that the excretion was a result of regulation of the cell P content since it was observed only when the organisms

grew in the presence of high concentrations of phosphate in the medium. The algal excretion experiment conducted for a 17-d period showed that 32P043- taken up during 1 h was excreted at a constant rate for a period of more than 2 weeks of growth where excess P was present. The excreted P thus probably results from metabolic processes connected with growth. That bacterial excretion was more rapid in the presence of glucose supports this conclusion. It is interesting that the release rate per algal cell was constant over a 2-week period (Fig. 8), considering that the released P was taken up during 30 min before the start of the excretion experiment. This type of excretion can occur if the P taken up by the cell during these 30 min was incorporated in a form that was returned to the medium at a constant rate. However, this is very unlikely, considering the multifunctional role of phosphate in metabolism. A more probable explanation is that release of 32P results from initially fixed 32Pbeing mixed with nonradioactive P added after the 30-min incubation and thereafter serving as a tracer for cell-bound P. If this explanation is valid, it means that actively growing cells had a turnover time of cell P of about three algal generations. Excretion from bacteria was more rapid than from algae. Thus, if the release rate from bacteria in Fig. 9 is compared to the average C : P ratio in bacteria (N 50 : 1 by weight) it appears that excretion per hour corresponded to 10% of total cell P content in the presence of glucose and to 2-3% without glucose. In algae, excretion per hour corresponded to l2% of total cell P content. Phosphate excretion in combination with sophisticated regulation of the uptake should help to distribute phosphate to different organisms or species, in addition to regeneration due to grazing, autolysis, and the suggested utilization of dissolved organic-P compounds by means of extracellular hydrolytic enzymes (Jansson 1977; Currie and Kalff 1984c; Ammerman and Azam 1985; Chrost and Overbeck 1987; Cotner and Wetzel 1992). For example, excretion from bacteria may serve as a significant source of phosphate to the HA system of algae. Excretion should, independent of its physiological explanation, therefore contribute to the turnover rate of phosphate in planktonic populations.

P uptake in algae and bacteria References AMMERMAN,J. W., AND F. AZAM.

1985. Bacterial 5’nucleotidase in aquatic ecosystem: A novel mechanism of phosphorus regeneration. Science 227: 13381340. BERMAN,T. 1985. Uptake of (“P) orthophosphate by algae and bacteria in Lake Kinneret. J. Plankton Res. 7: 71-84. BIELESKI,R. L., AND I. B. FERGUSON. 1983. Physiology and metabolism of phosphate and its compounds. Encycl. Plant Physiol. N.S. 15A: 422-449. BURNS,D. J. W., AND R. E. BEEVER. 1979. Mechanisms controlling the two phosphate uptake systems in Neurosporu crassa. J. Bacterial. 139: 195-204. CEMBELLA,A. D., N. J. ANTIA, AND J. P. HARRISON. 1984a,b. The utilization of inorganic and organic phosphorus compounds as nutrients by eukariotic microalgae: a multidisciplinary perspective. Part 1. Part 2. Crit. Rev. Microbial. 10: 317-391; 11: 13-117. CHISHOLM,S. W., ANDR. G. STROSS.1976a,b. Phosphate uptake kinetics in Euglena gracilis (Z) (Euglenophyceae) grown in light/dark cycles. 1. Synchronized batch cultures. 2. Phased PO,-limited cultures. J. Phycol. 12: 210-216, 217-222. CHROST,R. J., AND J. OVERBECK. 1987. Kinetics of alkaline phosphatase activity and phosphorus availability for phytoplankton and bacterioplankton in Lake PluBsee (north German eutrophic lake). J. Microb. Ecol. 13: 229-248. -, R. WCISLO, AND G. Z. HALEMEIKO. 1986. Enzymatic decomposition of organic matter by bacteria in a eutrophic lake. Arch. Hydrobiol. 107: 145-165. COLE, J. J., G. E. LIKENS, AND D. L. STRAYER. 1982. Photosynthetically produced dissolved organic carbon: An important carbon source for planktonic bacteria. Limnol. Oceanogr. 27: 1080-1090. CONWAY,H.L.,P.J. HARRISON,ANDC.O. DAVIS. 1976. Marine diatoms grown in chemostats under silicate or ammonium limitation. 2. Transient response of Ske/etonema costatum to a single addition of the limiting nutrient. Mar. Biol. 35: 187-199. COTNER, J. B., AND R. G. WETZEL. 1992. Uptake of dissolved inorganic and organic phosphorus compounds by phytoplankton and bacterioplankton. Limnol. Oceanogr. 37: 232-243. CUFUUE,D. J., AND J. KALFF. 1984a. A comparison of the abilities of freshwater algae and bacteria to acquire and retain phosphorus. Limnol. Oceanogr. 29: 298310. -,AND. 19843. The relative importance of bacterioplankton and phytoplankton in phosphorus uptake in freshwater. Limnol. Oceanogr. 29: 3 1 l-32 1. -, AND -, 1984~. Can bacteria outcompete phytoplankton for phosphorus? A chemostat test. Micrab. Ecol. 10: 205-216. DAVIS,C.O. ,N.F.BREITNER,ANDP.J.HARRISON. 1978. Continous culture of marine diatoms under silicon limitation. 3. A model of Si-limited diatom growth. Limnol. Oceanogr. 23: 41-52. FRIEBELE,E. S., D. L. CORREL, AND M. FAUST. 1978. Relationship between phytoplankton cell size and the rate of orthophosphate uptake; in situ observations of an estuarine nonulation. Mar. Biol. 45: 39-52.

1177

GAVIS, J. 1976. Munk and Riley revisited: Nutrient diffusion transnort and rates of phytoplankton growth. J. Mar. Res:34: 161-179. - - GOLDMAN,J. C., AND P. M. GLIBERT. 1982. Comparative rapid ammonium uptake by four species of marine phytoplankton. Limnol. Oceanogr. 27: 8 14-827. HARRISON,P. J., J. S. PARSLOW,AND H. L. CONWAY. 1989. Determination of nutrient uptake kinetic parameters: A comparison of methods. Mar. Ecol. Prog. Ser. 52: 301-312. ISTVANOVICS,V., K. PETERSON, D. PIERSON,AND R. T. BELL. 1992. An evaluation of phosphorus deficiency indicators for summer phytoplankton in Lake Erken. Limnol. Oceanogr. 37: 890-900. JANSSON,M. 1977. Enzymatic release of phosphate in water from subarctic lakes in northern Sweden. Hydrobiologia 56: 175-l 80. . 1988. Phosphate uptake and utilization by bacteria and algae, p. 177-189. In G. Persson and M. Jansson [eds.], Phosphorus in freshwater ecosystems. Kluwer. JEANJEAN, R. 1976. The effect of metabolic poisons on ATP level and on active phosphate uptake in Chfor&a pyrenoidosa. Physiol. Plant. 37: 107-l 10. KUHL, A. 1962. Inorganic phosphorus uptake and metabolism, p. 221-229. In R. A. Lewin [ed.], Physiology and biochemistry of algae. Academic. LEAN, D. R. S. 1973a. Phosphorus dynamics in lake water. Science 179: 678-680. -. 1973b. Phosphorus movement between biologically important forms in lake water. J. Fish. Res. Bd. Can. 30: 1525-1536. AND C. NALEWAJKO. 1976. Phosphate exchange and organic phosphorus excretion by freshwater algae. J. Fish. Res. Bd. Can. 33: 13 12-1323. MCCARTHY, J. J., AND J. C. GOLDMAN. 1979. Nitrogenous nutrition in marine phytoplankton in nutrientdepleted waters. Science 203: 670-67 1. NEIDHART, F.C., P. BLOCH, AND D. F. SMITH. 1974. Culture media for enterobacteria. J. Bacterial. 119: 736742. OLSEN, Y. 1988. Phosphate kinetics and competitive ability of planktonic blooming cyanobacteria under variable phosphate supply. Ph.D. thesis, SINTEF, Trondheim. 58 p. PAEFU, H. W., AND D. R. S. LEAN. 1976. Visual observations of phosphorus movements between algae, bacteria and abiotic particles in lake waters. J. Fish. Res. Bd. Can. 33: 2805-2813. PARSLOW,J. S., P. J. HARRISON,AND P. A. THOMPSON. 1985. Interpreting rapid changes in the uptake kinetics of the marine diatom Thalassiosira pseudonana (Hustedt). J. Exp. Mar. Biol. Ecol. 91: 53-64. PERSSON, G. 1985. Community grazing and the regulation of in situ clearance and feeding rates of planktonic crustaceans in lakes in the Kuokkcl area, northern Sweden. Arch. Hydrobiol. Suppl. 70, p. 197-238. RHEE,G-Y. 1972. Competition between an alga and an aquatic bacterium for phosphate. Limnol. Oceanogr. 17: 505-5 14. RIGLER,F. H. 1956. A tracer study of the phosphorus cycle in lake water. Ecology 37: 550-562.

1178

Jansson

RIVKIN, R. B., AND E. SWIFT. 1982. Phosphate uptake by the oceanic dinoflagellate Pyrocystis notiluca. J. Phycol. 18: 113-121. ROSENBERG, H. 1987. Phosphate transport in prokaryotes, p. 205-246. In B. P. Rosen and S. Silver [eds.], Ion transport in prokaryotes. Academic. TOMAS, C. R. 1979. Olisthodiscus luteus (Chrysophycae).

3. Uptake and utilization J. Phycol. 15: 5-12.

of nitrogen and phosphorus.

Submitted: 21 November 1991 Accepted: 1 March 1993 Revised: 6 April 1993