Jonnul of Plankton Resemch Vo1.19 no.7 pp.805-817, 1997

Nutrient recycling and shifts in N P ratio by different zooplankton structures in a South Andes lake

Esteban G.Balseiro, Beatriz E.Modenutti and Claudia P.Queimaliños Centro Regional Universitario Bariloche, Unidad Postal Universidad, 8400 Bariloche, Argentina

A h & . In South Andes lakes, zooplankton succession is characterized by a change in the domin- ance from the calanoid copepod Boeckellu grucilipes in winter and spring, to the cladoceran Bosrninu longirostris in midsummer, and the rotifer Polyurthra vulguris in late summer. We performed three series of field experiments at different times (late spring, summer and late summer) to examine the role of zooplankton constitution in the released nitrogen:phosphorus (NP) ratio. We observed that changes in the zooplankton constitution over the annual cycle may change the nutrient supply ratio. In South Andes lakes, Boeckellu grucilipes would decrease the P limitation, lowering the N:P ratio, whereas Bosrninu longirostris tends to increase the N:P ratio and therefore increase the P limitation during summer.

L

- *

Introduction

In pelagic systems, species interactions are a key factor determining the food web structure (Carpenter et al., 1987). Changes in composition and biomass at a par- ticular trophic level can cause marked changes at other levels via the metazoan plankton (Vanni, 1988). Zooplankton affect the phytoplankton through grazing and nutrient release, but the net effect of grazing and recycling combined is not straightforward (Sommer, 1988).

Cladocera release phosphorus (P) mainly as dissolved compounds (Peters, 1975), and although calanoid copepods produce faecal pellets, the major portion of their P release is also dissolved (Peters, 1975; Bamstedt, 1985). Ammonium and urea are the major forms of released nitrogen (N) for both groups (Lehman, 1984; Den Oude and Gulati, 1988). These fractions of released N and P are avail- able for algal uptake.

The elemental composition of the food relative to that of the grazer may affect nutrient release (Hessen and Andersen, 1992). Therefore, zooplankton may have different P or N demands and thereby will cause changes in the N:P ratio. For example, cladocerans will retain more P compared with calanoid copepods, and thus the daphniids would increase the N P ratio and so favour the competitive ability of calanoids over the season (Hessen and Andersen, 1992). Consequently, the zooplankton can promote P limitation for phytoplankton by k ing a larger fraction of the P in the system or by recycling a substantial amount of the N to a dissolved form (Urabe et al., 1995). Although the connections between trophic interactions and nutrients in planktonic food webs are recognized conceptually (Carpenter et al., 1985), the net impact of trophic interactions on resource supply is just beginning to be appreciated (Miller et al., 1995). It seems clear that when diatoms dominate phytoplankton, zooplankton excretion can change the nutri- ent ratio available for algae by strongly lowering the Si/P ratio (PCrez-Martinez and Cruz-Pizarro, 1995) because zooplankton excrete Si in an unavailable form

0 Oxford University Press 805

-

E.G.B.lseiro, JLE.Modeootti and C.P.Qoeimdi5os

for the algae (Sterner, 1989). If there are large differences in accumulation efficiencies for different groups of herbivores there will be an interdependence between nutrient cycling and community structure (Lehman, 1984; Elser et al., 1988; Sterner et ul., 1992).

In South Andes lakes, the calanoid copepod Boeckella gracilipes Daday and the cladoceran Bosmina longirostris (0.F.Muller) are the dominant crustacean zoo- plankton in many lakes (Balseiro and Modenutti, 1990; Modenutti and Balseiro, 1991). Remarkably, Duphnia is absent from these lakes, therefore, nutrient recy- cling would depend mainly on soluble forms of P and N excreted by Bosmina and Boeckellu. During annual zooplankton succession, zooplankton structure changes from copepod dominance in winter and spring to Bosmina in summer (Balseiro and Modenutti, 1990; Balseiro, 1991; Modenutti et al., 1993). In this study, we examine the role of zooplankton constitution in the N and P supply available for phytoplankton growth. We performed a series of field experiments with no zoo- plankton and with increasing zooplankton biomass. We attempt to compare in

composition the N:P ratio resulting from the changes in community structure.

r

three periods (late spring, summer and late summer) with different zooplankton \

Method

The study was carried out in Lake El Tr6bol (No Negro, Argentina) situated at 41"2'S and 71"4'W. The lake lies at 764 m a. s. 1. and has a surface area of 0.3 km2 and a maximum depth of 12 m. The climate of the region is classified as temper- ate (mean annual temperature 8.7"C), with dominant rainfall in winter (1200 mm per year). The lake tends to be dimictic as it freezes during hard winters; other- wise, its behaviour is warm monomictic. Direct stratification occurs during spring with a thermocline at 5 4 m; during summer, the mixed layer occupies the whole lake. The littoral zone is colonized by the aquatic weeds Schoenoplectus cali- fornicus (Meyer) Soj Ak and Poturnogeton linguutus Hangstrom. A pelagic station was fixed in the middle of the lake at 11 m depth.

Dissolved nutrient release rates were examined by incubating different zoo- plankton densities in polycarbonate bottles in the mixed layer of the lake. Three series of experiments were run on 5-8 December 1995 (Experiment l), 16-19 January 1996 (Experiment 2) and 12-15 February 1996 (Experiment 3).

The lake was sampled 1 or 2 days before and on the initial day of each experi- ment. At each sampling opportunity, vertical profiles of water temperature, dis- solved oxygen and conductivity were obtained by means of a thermistor, an oxymeter and a conductimeter. Water transparency was measured with a Secchi disk. Water, phytoplankton and zooplankton samples were collected in duplicates with a Ruttner bottle and a Schindler-Patalas trap.

Experiments

The experimental design consisted of four treatments with three replicates each. One of the treatments had no zooplankton and in the other three increasing zooplankton concentrations (in biomass units, from 100 to 4000 pg 1-1 dry weight)

806

Nutrient recycling and NP ratio

were added. Lake water used for zooplankton experiments was collected at 5 m depth with a Schindler-Patalas trap and filtered through a 55 pm mesh net. The filtered water was placed in four 30 1 isolated tanks. Living zooplankters were col- lected by a vertical tow from 5 to 0 m with a 40 cm diameter conical plankton net with 55 pm mesh. Immediately after sampling, the zooplankton were rinsed with filtered lake water and acclimatized in 1 1 beakers containing filtered lake water. The zooplankton were placed in the experimental bottles within half an hour.

Acid-washed polycarbonate bottles of 2 1 volume were used as experimental units. The bottles were filled with filtered water and the zooplankton were trans- ferred to them. A total of 48 experimental units were prepared, 12 of which (four treatments, three replicates each) were carried to the laboratory to quantify the initial conditions. The remaining 36 bottles were incubated in a frame hanging at 5 m depth in the pelagic lake station (11 m depth). The experimental units were placed in the lake at 11:OO a.m. and this moment was considered as Day 0. lkrelve bottles (four treatments, three replicates each) were removed after 24 h, another 12 after 48 h, and the others after 72 h of incubation (Day 1, Day 2 and Day 3, respectively).

Laboratory methods

In all cases, the experimental units were carried to the laboratory within half an hour of removal from the frame, in darkness and in an isolated container to prevent temperature changes. In the laboratory, 150 ml of each bottle were sep- arated and fixed with acid Lugol solution for phytoplankton counting. The remaining water was filtered through a 55 pm mesh to remove zooplankton, and then through a GF/F filter at 220 mmHg. The filtered water was used to deter- mine soluble reactive phosphorus (SRP), total dissolved phosphorus (TDP) and N&+ concentrations. The filter was used for chlorophyll a (Chl a) determination. From the 12 initial bottles (Day 0), 150 ml of each bottle were separated for total phosphorus (TP) determination previous to any filtration. The concentration of N-N&+ was measured with the indophenol blue method. For TP and TDP, samples were digested with potassium persulphate at 125°C at 1.5 atm for 1 h. SRP, TP and TDP concentrations were measured by the ascorbate-reduced molybdenum blue method. Chlorophyll a concentration was measured by extrac- tion with 90% ethanol following Nusch (1980).

Zooplankton were counted and at least 30 individuals were measured under a dissecting microscope for crustaceans and a direct microscope for rotifers. Phyto- plankton were quantified in an inverted microscope in 50 ml Utermohl chambers. Zooplankton biomass was estimated on the basis of the length-mass regressions of Bottrell et al. (1976).

For nutrient concentration-zooplankton biomass regressions, geometric mean regressions were applied (Sokal and Rohlf, 1981).

Results During the period investigated, vertical profiles of water temperature presented a tendency to a direct stratification in December and almost isothermal

807

E.G.B.lsei B.E.Modenutti and C.P.Queimaliiios

Dissolved Oxygen (mg I-') 6 7 8 9 10 7 8 9 I 0 6 7 8 9 1 0

0

2 9 4 3 6-

3 8- 10

10 12 14 16 18 20 12 14 16 18 20 12 14 16 18 20

Temperature "C Fig. 1. Vertical profiles of temperature (solid line) and dissolved oxygen concentrations (dotted line) in Lake El TrCbol. (a) 5 December 1995; (b) 16 January 1996; (c) 12 February 1996.

conditions in January and February, as a consequence of strong winds (Figure 1). Dissolved oxygen concentration showed an almost homogeneous distribution in the water column (Figure 1). Total dissolved solids did not vary either within the water column or between experiments (conductivity 65 pS cm-l). Transparency, measured by a Secchi disk, varied from 5 m in December, to 7 m in January and 8 m in February, indicating an extended euphotic zone down to the bottom.

In Lake El Trkbol during the studied period, phytoplankton were largely made up of the nanoflagellate Chrysochromulina parva Lackey as dominant (>90% of total phytoplankton cell abundance). The net phytoplankton fraction (>20 pm GALD) was dominated by the diatom Aulacoseira granulata (Ehr.) Simonsen and by the dinoflagellate Gymnodinium mirabilis Stein. At the initial conditions, all bottles of the different treatments were filled with 55 pm mesh filtered lake water resulting in a cell concentration of 2800,2700 and 2000 cells ml-' of C.parva in Experiment 1, 2 and 3, respectively. Therefore, in our three experimental series, the quantity and quality of food resources were similar.

Zooplankton constitution changed in the three series of experiments. In December, zooplankton were mainly composed of two crustacean species: a calanoid copepod Boeckella gracilipes Daday and the cladoceran Bosmina longirosfris (0.F.Miiller). In January, the latter was the dominant species (up to 99% of the total zooplankton biomass), while in February the rotifer Polyarthra

~

6000 600

450 7 4500 W 1 150 3

3000 E 450 5 0

1500 300 - 150

0 0 Filtered l x 2 x 3 x Filtered l x 2 x 3 x Filtered I x 2 x 3 x

Fig. 2. Zooplankton biomass composition of the different treatments. (a) Experiment 1,5-8 Decem- ber 1995; (b) Experiment 2,16-19 January 1996, (c) Experiment 3,12 February 1996, initial condition; (a) Experiment 3,15 February 1996, final condition. Key: Filtered = without zooplankton; IX, 2X and 3X = 1-, 2- and 3-fold zooplankton biomass.

808

Nutrient recycling and N P ratio

vulgaris Carlin was dominant and the cladoceran Bosmina longirostris was present only at low density (representing 5% of the total abundance, but 50% of total zooplankton biomass). The described situations gave us the opportunity to test nutrient recycling under different zooplankton constitution.

Experiment 1: Zooplankton: Boeckella gracilipes (>go% total zooplankton biomass) + Bosmina longirostris (4'0% total zooplankton biomass) (Figure 2a)

Manipulation of zooplankton biomass as treatments in this experiment was suc- cessful (Figure 2a) (ANOVA: P < 0.001). As expected, zooplankton treatments had significant effects on nutrient and Chl a concentrations during the 3 days of experiments (ANOVA: P < 0.001 for nutrients and P c 0.05 for Chl a). The results obtained indicated higher values of the three forms of nutrients in the bottles with increasing zooplankton biomass (Figure 3).

0 a 20 ,-* 15 v

f 10 = 5

n

" I Chlorophyll a A

9 - 1.0

5 0.5

0.0

m -

4 2 2 1

0

U Fitted

r r l 2 x El2 1 x

Day0 Day1 Day2 Day3

Fig. 3. Nutrient and chlorophyll a concentration during Experiment 1, 5-8 December 1995. The dotted line in the SRP graph indicates the detection limit. Error bars indicate standard errors. Key as in Figure 2.

809

E . G B M i B.E.Modenutti and C.P.QueimaIi5os

Table I. Nutrient and chlorophyll a concentrations (pg l-')-total zooplankton biomass (mg 1-I) relationships for each experimental day of Experiment 1 (December 1995)

Day 1 Day 2 Day 3

N - W

SRP

TDP

chl n

b = 4.35 b = 8.53 r = 0.9389* r = 0.8822* ND b = 1.09

r = 0.9004* b = 0.94 b = 2.06 r = 0.%31* r = 0.9174* b = 0.42 b = -0.65 r = 0.2712 NS r = 0.8275*

b = 8.41 r = 0.8856* b = 0.86 r = 0.8824* b = 1.17 r = 0.9226* b = -0.20 r = 0.2163 NS

*P < 0.05; ND, not detectable; NS, not significant.

Ammonium and TDP showed similar responses to increasing biomass, and in all cases we found significant and positive relationships (Table I). Obtained values of SRP were very low and for 24 h of incubation the SRP concentration must be considered undetectable (4 pg 1-l) (Figure 3, dotted line). However, after 48 and 72 h of incubation, we found positive and significant relationships with increas- ing biomass, although slopes remained very low (Table I and Figure 3).

Algal biomass (Chl a) did not show any general tendency to increase or decrease with zooplankton manipulations (Figure 3). Nevertheless, after 48 h of incubation, Chl a decreased with increasing crustacean biomass, and was the only significant negative relationship observed (Table I).

-

''

Experiment 2: Zooplankton: Bosmina longirostris (99% of total zooplankton biomass) (Figure 2b)

Manipulations of Bosmina longirostris resulted in biomass concentrations ranging from 0 (filtered water) to 4.3 mg 1-' (ANOVA P c 0.001). Nutrient concentrations, but not Chl a, were an order of magnitude higher than for the other experiments. We observed significant responses of nutrients and Chl a

Table II. Nutrient and chlorophyll u concentrations (pg l-l)-total zooplankton biomass (mg I-') relationships for each experimental day of Experiment 2 (January 1996)

Day 1 Day 2 Day 3

N - W

TDP

SRP

chl a

b = 9.90 r = 0.9811" b = 1.94 r = 0.9594* b = 1.52 r = 0.9535* b = -0.08 r = 0.6277"

b = 26.54 r = 0.9740* b = 3.84 r = 0.9863* b = 3.16 r = 0.9781" b = -0.14 r = 0.5337 NS

b = 51.83 r = 0.9780* b = 9.02 r = 0.9727* b = 7.04 r = 0.9528* b = -0.11 r = 0.1153 NS

*P < 0.05; NS, not significant.

Nutrient recycliig and NP ratio

250m ,- 200

9 150

5 50

0

-i 30 a

n

- v

f 100

0)

n' 20

2 10

0

-f 1.5

- 1.0

6 0.5

0.0 -- 20

9 15

9 (0 -

0 Filtered 1 X

- v

9 10 ? n 5

0

Day0 Day1 Day2 Day3

Fig. A Nutrient and chlorophyll a concentration during Experiment 2,1619 January 1996. Error bars indicate standard errors. Key as in Figure 2.

concentrations to increasing zooplankton biomass in all 3 days of experiments (ANOVA P < 0.001 in all cases).

increment in nutrient concentrations on all days of experiments (Figure 4). Ammonium, TDP and SRP showed positive significant relationships in all cases (P < 0.01) (Table 11). In the filtered treatment, a decrease in SRP was observed and became undetectable during the 72 h of incubation (4 pg 1-I) (Figure 4).

The highest values of Chl a were observed in the treatments without zoo- plankton (Figure 4). At Day 1 (24 h of incubation), the lowest algal biomass values were obtained and only on this day was a significant negative relationship with zooplankton biomass obtained (Table 11). Nevertheless, Chl a did not decrease monotonically with increasing zooplankton biomass, so this significant value must be taken with care. After 48 and 72 h of incubation, Chl a did not show any tendency to increase or decrease with zooplankton manipulations (Figure 4 and Table 11).

* The results obtained indicated that higher biomass of Bosrnina led to an

811

E.G.Balseiro, B.E.Modenatti and CP.Queimaliiios

Experiment 3: Zooplankton: Polyarthra vulgaris + Bosmina longirostris (50% of total Zooplankton biomass each) (Figure 2c)

Total zooplankton biomass manipulations were successful (ANOVA: P < O.OOl), but, if we analyse the data carefully, it is clear that a change in zooplankton struc- ture occurred during the 3 days of experiments. The above-mentioned zoo- plankton constitution represented the starting condition of the experiment (Figure 2c). At the end of the 72 h of incubation, zooplankton structure changed towards a dominance of Bosmina and a clear decrease in Polyarthra density (Figure 2d). This particular situation should be taken into account for the analy- sis of nutrient recycling.

Treatments resulted in significant differences in nutrient and Chl a concen- trations (ANOVA: P < 0.001). Ammonium, TDP and Chl a increased with zooplankton biomass (Figure 5). Undetectable values (<1 pg 1-l) were obtained for SRP. Ammonium and TDP concentrations showed significant positive relationships with zooplankton biomass in the 3 days of experiments (Table 111).

1"

50 I -- -- 40 -1 Ammonium

5- 20

I 10 n

T

,- 4 Chlorophyll a -

Day0 Day1 Day2 Day3

Fig. 5. Nutrient and chlorophyll a concentration during Experiment 3, 12-15 February 1996. The dotted line in the SRP graph indicates the detection limit. Error bars indicate standard errors. Key as in Figure 2.

812

.

Nutrient recycling and N P ratio

Table IJI. Nutrient and chlorophyll a concentrations (pg I-')-total zooplankton biomass (mg 1-I) relationships for each experimental day of Experiment 3 (February 1996)

Day 1 Day 2 Day 3

N-NH,

TDP

SRP Chl a

b = 12.01 b = 70.21 b = 56.11 r = 0.9491* r = 0.8375* r = 0.8132*

b = 4.20 b = 10.78 b = 4.10 r = 0.8598* r = 0.7710* r = 0.8128* ND ND ND

r = 0.8882* r = 0.7872* r = 0.3474 NS b = 1.51 b = 6.65 b = 4.75

*P < 0.05; ND, not detectable; NS, not significant.

Chlorophyll a exhibited unexpected results as it increased with zooplankton biomass (Figure 5) and significant positive relationships with increasing zoo- plankton biomass were obtained on Day 1 and 2 (Table 111). Nevertheless, on Day 3 no significant result was obtained (Table 111). The change in zooplankton structure can be related with these variable results, since the increase in Bosmina (Figure 2d) resulted in a decrease in Chl a concentrations (Figure 5 , Day 3, Treat- ment 3 X). Remarkably, ammonium concentration had a dramatic response to the increment of Bosmina (Figure 5, Day 3, Treatment 3X).

.

N/P ratio

We have analysed the nutrient N P ratio in the experimental bottles in terms of the dissolved compounds released by grazers, N-NH4+:P-TDP. In Experiment 1, the N:P ratio exhibited rather homogeneous values around 2.5. However, these values varied with zooplankton constitution. Although total zooplankton biomass did not differ significantly between days (ANOVA: P < 0.05), the pro- portion of the two dominant species changed slightly. Bosmina biomass to total zooplankton biomass varied from 0.10 to 0.25 in the different experimental bottles. When N:P is plotted against the proportion of Bosmina, it can be observed that the N:P ratio correlated positively with the proportion of Bosmina

Experiment 2, which had only Bosmina and thus represented the whole zoo- plankton biomass, showed very stable N:P ratios, that were around six. However, no correlation may be made as Bosmina biomass had a constant value of one, but the extrapolation of the regression line of the first experiment coincided with the points of the N P ratio of the second experiment (Figure 6, squares).

In Experiment 3, we had major changes in the constitution of zooplankton during the experiment. Although total biomass did not change between days (ANOVA: P > 0.05), the proportion of Bosmina biomass increased with density and days, ranging from 0.42 to 0.98 from total zooplankton biomass. In this experiment, the N P ratio was the most variable of all the experiments (N:P ranged from 2.8 to 10). The data of this experiment fitted well in the general trend when all the data of the three experiments were pooled (Figure 6, triangles), and did not change the slope obtained for the first two.

813

, biomass (Figure 6, circles).

- .

E.G.Balseiro, B.E.Mademtti and C.P.Queimaliiios

0

E

c

.- 5 L c 3 z

10

8

6

4

2

0

b4.1497 ?=0.6776

A

0.00 0.25 0.50 0.75 1.00

Proportion of Bosmina biomass

Fig. 6. N:P ratio of soluble released nutrient (N-w:P-TDP) plotted against the proportion of Bosmina longirosfris biomass to total zooplankton biomass. Circles (o), Experiment 1; squares (o), Experiment 2; triangles (A), Experiment 3.

*

Discussion

Zooplankton manipulations had a substantial effect on nutrient concentration. The pronounced increase in N and P in experimental bottles indicates that the absolute concentration of both nutrients is dependent on zooplankton biomass. The ability of phytoplankton for rapid uptake of P-P04 can be observed in the extremely low SRP concentrations, that were undetectable (SRP < 1 pg 1-l) in part of Experiment 1, all filtered treatments of Experiment 2 and all Experiment 3 (Figures 3, 4 and 5). Grazing increases with zooplankton density, and thus a decrease in Chl a concentration can be expected. However, only on Day 2 of Experiment 1 and Day 1 of Experiment 2 did we observe a significant negative relationship (Chl a-zooplankton biomass; Tables 1 and 11). On the other hand, for Experiment 3 on Day 1 and 2, a significant positive relationship was obtained (Table 111). Through grazing and nutrient excretion, zooplankton influence phytoplankton in opposite ways (Sterner, 1989). The results obtained for Chl a concentrations are obscure due to these opposite effects. Because not all the species of phytoplankton are grazed uniformly, and because not all these species take up nutrients and grow at the same rate, it is likely that these opposite effects are quite difficult to understand (Capblancq, 1990). Probably, the role of zoo- plankton as a factor influencing the dynamics of an algal community is not revealed by Chl a concentration and further studies on species-specific effects at the community level may clarify this aspect of food web interactions.

The rate of nutrient increase per biomass unit differed either with the nutrient N or P, and with the zooplankton constitution (Experiments) (Tables 1-111). As Lehman (1980) indicated, any simple interpretation of this kind of experiment is complicated. The nutrient concentrations inside the enclosures represent a balance between nutrient remineralization by the zooplankton and the simul- taneous uptake of the same nutrients by the phytoplankton (Lehman, 1980, 1984). Although this statement is true, and therefore the slope of nutrient increase does not reflect the real release rate by excretion, the nutrient N:P ratio resulting from the algae-grazer interface will change if excretion releases N and

814

~ ..

Nutrient recycling and NP ratio

P in a different ratio from that ingested by the grazer. Then, a measurable shift in this ratio may be expected inside the bottles.

Freshwater planktonic herbivorous crustaceans possess a rather rigid stoichi- ometry, but those of different species may differ greatly (Hessen, 1990; Ander- sen and Hessen, 1991; Hessen and Lyche, 1991). The herbivorous cladocerans, especially Daphnia, have high specific P contents (Sterner and Hessen, 1994), and so will release proportionally more N than P in comparison to food ratio (Sterner, 1990). On the other hand, copepods produce faecal pellets, which do not release their content immediately. The size and consistency of faecal pellets would depend on the quantity and quality of food (Haney and Trout, 1990); therefore, changes in pellet features may affect nutrient release rates from them. If the N:P ratio in the faecal pellets differs from that of the copepod tissue, this will change the N:P ratio supplied to algae (Sterner, 1990). In fact, the copepod Culanus excretes as a soluble form only 35% of total excreted N, but 60% of the total excreted P (Golterman, 1975). If faecal pellets of freshwater copepods have an increased N:P ratio compared with that excreted as a soluble form, then a decrease in the N:P ratio may be expected when copepods dominate the plank- ton. In Experiment 1,80% of zooplankton biomass was made up of the calanoid copepod Boeckella gracilipes and the other 20% of Bosmina longirostris. The nutrient N:P ratio of this experiment was the lowest of the three, and increased with the increase in the proportion of Bosmina biomass (Figure 6, circles). The low proportion of Bosmina and the faecal pellets produced by Boeckella may be responsible for the low nutrient N:P ratio observed in this experiment, when com- pared with the other two. In Experiment 2, when Bosmina was the only herbi- vore, the nutrient N:P ratio was high. Strikingly, this N:P ratio fitted with the regression line obtained for the first experiment (Figure 6, squares). Sterner (1990) states that considering homeostatic herbivores, an increase in the N:P ratio in the zooplankton will lower the ratio supplied for the phytoplankton through excretion. Apparently, Bosmina has a lower internal N:P ratio than calanoid copepods (Sterner and Hessen, 1994) and therefore Bosminu would increase the released N P ratio. Little, if anything, is known about the N P ratio in rotifers (Sterner and Hessen, 1994) and so about the effect that they may have on the N:P ratio through their excretion or via feedback mechanisms in the microbial food web (Arndt, 1993). In our experiments with Bosminu and Polyarthra, we were not able to determine any shift caused specifically by Polyarthru. During this experiment, and regardless of the mechanism involved, Bosmina replaced Polyarthra. Nevertheless, our results seem to indicate that Bosmina effectively tend to increase the nutrient N P ratio supplied to algae, and therefore to increase the P limitation during summer. On the contrary, Boeckella would decrease the P limitation, lowering the nutrient N:Pratio. Changes in food resources will induce differences in the size and consistency of faecal pellets in Boeckella (Haney and Trout, 1990). Therefore, a decrease in food should increase the rate of recycling via copepod egestion due to slower sinking rates of smaller, lower density pellets (Haney and Trout, 1990).

This study provides evidence that changes in the zooplankton constitution and density over the annual cycle may change the nutrient supply ratio, causing

815

.

E.G.Bnlsei B.E.Modenutti and C.P.Queimaliiios

alternative nutrient limitation. In Andean lakes, the cladoceran Bosmina longirostris will retain more P compared with the calanoid copepod Boeckella gracilipes, and thus Bosmina increases the nutrient N:P ratio. Unfortunately, the effect of the rotifer Pvulguris as dominant was not clear in our experiments.

Acknowledgements

This work was supported by a UNComahue Grant B/701 to B.E.M.

References Andersen,T. and Hessen,D.O. (1991) Carbon, nitrogen, and phosphorus content of freshwater zoo-

plankton. Limnol. Oceanogr., 36,807-814. Amdt,H. (1993) Rotifers as predators on components of the microbial web (bacteria, heterotrophic

flagellates, ciliates) - a review. Hydrobiologia, 255/256,231-246. Balseir0,E.G. (1991) Dinhmica poblacional de Boeckella gracilipes (Copepoda, Calanoida) en lagos

Andinos. Biol. Acudt., 15, 136-137. Balseir0,E.G. and Modenutti,B.E. (1990) Zooplankton dynamics of Lake Escondido (Rio Negro,

Argentina) with special reference to a population of Boeckella gracilipes (Copepoda, Calanoida). Int. Rev. Ges. Hydrobiol., 75,475491.

Bamstedt,U. (1985) Seasonal excretion rates of macrozooplankton from the Swedish west coast. Limnol. Oceanogr., 30,607-617.

Bottrell,H.H., Duncan,A., Gliwicz,Z.M., Grygierek,E., Herzig,A., Hillbricht-Ilkowska,A., Kura- sawa,H., Larsson,P. and Weglenska,T. (1976) A review of some problems in zooplankton produc- tion studies. Nonu. J . ZooI., 24,419-456.

CapblancqJ. (1990) Nutrient dynamics and pelagic food web interactions in oligotrophic and eutrophic environments: an overview. Hydrobiologia, 207,l-14.

Carpenter,S.R., Kitchel1,J.F. and Hodgson,J.R. (1985) Cascading trophic interactions and lake productivity. Bioscience, 35,634-638.

Carpenter,S.R. et al. (1987) Regulation of lake primary productivity by food web structure. Ecology, 68,1863-1876.

Den Oude,P.J. and Gulati,D. (1988) Phosphorus and nitrogen excretion rates of zooplankton from eutrophic Loosdrecht lakes, with notes on other P sources for phytoplankton requirements. Hydro- biologia, 169,379-390.

ElserJ.J., Elser,M.M., MacKay,N.A. and Carpenter,S.R. (1988) Zooplankton-mediated transitions between N- and P-limited algal growth. Limnol. Oceanogr., 33,l-14.

GoltermanJ3.L. (1975) Physiological Limnology. An Approach to the Physiology of Lake Ecosys- tems. Elsevier Scientific Publishing Company, New York.

HaneyJF. and Trout,M. (1990) Relationships between fecal pellet production and feeding in the calanoid copepod Boeckella. L Plankton Res., 12,701-716.

Hessen,D.O. (1990) Carbon, nitrogen, and phosphorus status in Daphnia at varying food conditions. L Plankton Res., 12, 1239-1249.

Hessen,D.O. and Andersen,T. (1992) The algae-grazer interface: feedback mechanisms linked to ele- mental ratios and nutrient cycling. Arch. Hydrobiol. Beih., 35,111-120.

Hessen,D.O. and Lyche,A. (1991) Inter- and intraspedc variations in zooplankton elemental com- position. Arch. Hydrobiol., 121,343-353.

L,ehman,J.T. (1980) Nutrient recycling as an interface between algae and grazers in freshwater com- munities. In Kerfoot,W.C. (ed.), Evolution and Ecology of Zooplankton Communities. University Press of New England, Hanover, pp. 251-263.

Lehman,J.T. (1984) Grazing, nutrient release, and their impacts on the structure of phytoplankton communities. In Meyers,D.G. and Strick1erJ.R. (eds), Trophic Interactions Within Aquatic Ecosys- tems. American Assoc. Adv. Sci. Select. Symp. 85. Westview Press, Boulder, CO, pp. 49-72.

Mil1er.C.A.. Penry,D.L. and Glibert,P.M. (1995) The impact of trophic interactions on rates of nitro- gen regeneration and grazing in Chesapeake Bay. Limnol. Oceanogr., 40,1005-1011.

Modenutti,B.E. and Balseir0,E.G. (1991) Zooplancton de cuerpos de agua del sistema sur del Nahuel Huapi. Biol. Acurfr., 15, 134-135.

Modenutti,B.E., Balseir0,E.G. and CervelIini,P.M. (1993) Effect of selective feeding of Galaxias

816

Nutrient recycling and NP ratio

maculatus (Salmoniforme, Galaxiidae) on zooplankton of a South Andes lake. Aquut. Sci., 55,

Nusch,E.A. (1980) Comparison of different methods for chlorophyll and phaeopigment determin-

P&ez-Martinez,C. and Cruz-PiZarro,L. (1995) Species-specific phytoplankton responses to nutrients

Peters,R.H. (1975) Phosphorus regeneration by natural populations of limnetic zooplankton. Verh.

Soka1,R.R. and Rohlf,F.J. (1981) Biometry. Freeman and Co., New York, 859 pp. Sommer,U. (1988) Phytoplankton succession in microcosm experiments under simultaneous grazing

pressure and resource limitation. Limnol. Oceanogr., 33,1037-1054. Sterner,R.W. (1989) The role of grazers in phytoplankton succession. In Sommer,U. (ed.), Plankton

Ecology: Succession in Plankton Communities. Springer-Verlag, Berlin, pp. 107-170. Sterner,R.W. (1990) The ratio of nitrogen to phosphorus resupplied by herbivores: zooplankton and

the algal competitive arena. Am. Nut., l36,209-229. Sterner,R.W. and Hessen,D.O. (1994) Algal nutrient limitation and the nutrition of aquatic herbi-

vores. Annu. Rev. Ecol. Syst., 25,l-29. Sterner,R.W., E1serJ.J. and Hessen,D.O. (1992) Stoichiometric relationships among producers, con-

sumers and nutrient cycling in pelagic ecosystems. Biogeochemktry, 17,49-67. UrabeJ., Nakanishi,M. and Kawabata,K. (1995) Contribution of metazoan plankton to the cycling of

nitrogen and phosphorus in lake Biwa. Limnol. Oceanogr., 40,232-241. Vanni,M.J. (1988) Freshwater zooplankton community structure: Introduction of large invertebrate

predators and large herbivores to a small-species community. Can. J. Fish. Aquat. Sci., 45, 1758-1770.

65-75.

ation. Arch. Hydrobiol. Beih. Ergebn. Limnol., 14,1434.

and zooplankton manipulations in enclosures experiments. Freshwater Biol., 33,193-203.

Int. Ver. Limnol,, 19,273-279.

’

- Received on June 12,19%; accepted on February 4,1997

817