MARINE ECOLOGY - PROGRESS SERIES Vol. 47: 249-258. 1988 Published August 31 Mar. Ecol. Prog. Ser. Impact of grazing by microzooplankton in the Northwest Arm of Halifax Harbour, Nova Scotia D. J. Gifford* Department of Oceanography, Dalhousie University. Halifax, Nova Scotia, Canada B3H 451 ABSTRACT: Impact of grazing by natural assemblages of microzooplankton was measured in 5 in situ experiments in Halifax Harbour, Nova Scotia (Canada) using a seawater dilution method. The mi- crozooplankton assemblages, dominated numerically by oligotrich chates, exerted a seasonally vari- able grazing impact; 38 % of the initial standing stock of chlorophyll a d-' was consumed (= 47 % of potential chlorophyll production d-l) in June when flagellates < 12 pm dominated the phytoplankton. No significant grazing occurred in November during a bloom of large diatoms. In March, when grazing and phytoplankton growth were in balance, 100 % of the daily chlorophyll production was grazed. Assumptions of the dilution method that threshold feeding does not occur and that phytoplankton nutrients are not limiting were examined, and although probably violated in some cases, were found not to affect the results of the experiments. INTRODUCTION consuming nano- and microphytoplankton (e.g. Beers & Stewart 1970, 1971, Beers et al. 1975, 1980, Hein- The nlicrozooplankton size category (<200 pm) is bokel 1978a, b, Smetairek 1981, Stoecker et al. 1981). composed of a diverse taxonomic assemblage, includ- Indirect estimates suggest that microzooplankton con- ing planktonic Protozoa and larval and naupliar stages sume a substantial fraction of the phytoplankton pro- of Metazoa. Two suborders of ciliate protozoans, the duction in pelagic food webs (e.g. Rley 1956, Beers & Tintinnina (tintinnids) and the Oligotrichina (oligo- Stewart 1970, 1971, Takahashi & Hoskins 1978). How- trichs) are ubiquitous and abundant in pelagic ecosys- ever, the feeding biology of most microzooplankton tems, usually dominating the microzooplankton fauna. taxa has not been examined directly and is not well Within the ciliate microzooplankton, aloricate forms, known, causing uncertainty in appraisal of their trophic primarily but not exclusively oligotrichs, inevitably out- importance. number the loricate tintinnids (e.g. Beers & Stewart While several studies document feeding rates and 1969a, 1970, Beers et al. 1975, 1980, SmetaEek 1981). prey preferences of tintinnids in laboratory culture Some planktonic ciliates are capable of ingesting bac- (Heinbokel 1978a, Stoecker et al. 1981, Verity 1985), teria in certain environments (Sherr & Sherr 1987), the generally more abundant oligotrichs have not been although bacterial abundances are insufficiently high examined in similar detail. Methods to assess grazing in most pelagic environments to constitute a primary rates of herbivorous ciliates in the laboratory involve nutritional source (Fenchel 1980). Some forms, such as monitoring ingestion of inert tracers, such as cornstarch the haptorid ciliate Mynonecta (ex Mesodinium) rubra, particles (Spittler 1973, Heinbokel 1978b), or following are obligate autotrophs (Lindholm 1985), and some changes in concentration of the organisms' algal food oligotrich ciliates retain functional chloroplasts and are with time, either as cell numbers (Heinbokel 1978a, mixotrophs (McManus & Fuhrman 1986, Stoecker et al. Stoecker et al. 1981, Verity 1985) or plant pigments 1987).However, it is generally agreed that most plank- (Gifford 1985a). Laboratory studies have the advan- tonic oligotrichines (= tintinnids & oligotrichs), includ- tages that conditions of food level, media, food type ing the mixotrophic forms, are primarily herbivorous, and quality, and densities of consumers are specified by the experimenter. They have the disadvantage of Present address: Louisiana Universities Marine Consort- uncertainty that the physiological rates measured rep- ium, Chauvin, Louisiana 70344, USA resent the norm in nature. Rates measured in the O Inter-Research/Pnnted in F R. Germany Mar. Ecol. Prog. Ser. 47: 249-258, 1988 laboratory may be overestimates of mean conditions in natural microplankton assemblage, and incubation. All the field if experimental conditions are unrealistically have the advantage of being empirical. Methods which advantageous to the consumers, or underestimates if use tracers of ingestion, whether inert particles or conditions are stressful. The latter point is of concern radioisotopes, yield per capita rates of consumption when studying the delicate aloricate ciliates. and are usually of short duration in order to avoid Field studies of microzooplankton grazing are con- internal recycling of the tracer. Because of the requisite founded by the problem of manipulating small consum- short timescales, the organisms examined by tracer ers whose size is the same order of magnitude as their techniques may not be adapted to the experimental food. A variety of approaches have been used, all of conditions, and may not behave or feed normally. Such which have advantages as well as disadvantages methods demonstrate phagotrophic feeding unequi- (Table 1). Indirect methods have the advantage of vocally. However, in the case of tracer particles, selec- being non-manipulative, and they suggest hypotheses tive feeding by the consumers may result in underesti- to be tested directly. However, correlation of con- mation of consumption (e.g. Stoecker et al. 1981). sumer-prey cycles observed in nature reveals only Metabolic inhibitors are most useful for measuring possible qualitative relationships. As noted above, eukaryotic consumption of prokaryotes, e.g. ciliate extrapolation of laboratory results to the field is quan- grazing of bacteria and cyanobacteria (Fuhrman & titative, but may not reflect in situ conditions. The first 4 McManus 1984, Campbell & Carpenter 1986). They direct methods listed in Table 1 are quasi in situ, have not yet been applied successfully to eukaryote- involving manipulation, consequent alteration of the eukaryote interactions, and may have unknown effects Table 1. Summary of methods used to estimate the grazing impact of microzooplankton Method Advantages Disadvantages Source Indirect methods 1 Correlation of natural Non-invasive Qualitative Smetaeek (1981),Sheldon et a1 consumer-prey cycles (1986) 2 Extrapolation of labo- Non-invasive May not represent in situ Beers & Stewart (1970,19711, Ta- ratory rates to the field conditions guchi (1976),Heinbokel (l978b), Rassoulzadegan & Etienne (1981),Burkill(1982). Rassoul- zadegan (1982),Capriulo & Car- penter (1983),Hernroth (1983), Cosper & Stepien (1984),An- dersen & Sorensen (1986),Paran- jape et al. (1985) 3 Extrapolation from Correlations may not reflect Mey (1956),Takahashi & Hos- other field data natural relationships luns (1978) Direct methods 1 Tracers of ingestion (A) Inert particles Quantitative; demonstrates Selective feeding by consum- Heinbokel & Beers (1979),Bsrs- phagocytosis directly ers may affect results heim (1984) (B) Radioisotopes Quantitative, sensitive Alternate pathways of isotope Lessard & Swift (1985) uptake affect cycling of tracer; highly manipulative 2 Metabolic inhibitors Quantitative Non-specificity of inhibitors Campbell& Carpenter (1986) 3 Slze fractionatlon Quantitative; uses natural No true controls; hlghly mani- Capriulo & Carpenter (1980), assemblage pulative; predators &preyare Verity (1986) not unequivocally separated 4 Seawater dilution Quantitative; simultaneous May alter natural assemblage; Landry & Hassett (1982),Burkill estimation of algal growth & unproven assumption that et al. (1987),Paranjape (1987) mortality; minimally rnani- feedmg thresholds do not pulative to natural asserr- occur blage 5 Pigment budget In situ, no manipulation of Uncertainty of conversion effi- SooHoo & Kiefer (1982), natural assemblage ciency of chlorophyll to Welschmeyer & Lorenzen (1985) phaeopigments Gifford: M~crozooplanktongrazing 25 1 on the organisms they are not intended to inhibit (Sherr calculated from changes in phytoplankton density et al. 1986, Taylor & Pace 1987, Tremaine & Mills 1987). following incubations of different dilutions of seawater Studies employing size fractionation or dilution are of containing the natural microzooplankton assemblage. longer duration than studies which use tracers, and The term l/t ln(P,/P,) is the 'apparent phytoplankton yield assemblage, rather than per capita, grazing rates. growth rate'. The y-intercept of this relationship is the Size fractionation techniques assume that control and 'true' phytoplankton growth rate, k, in the absence of experimental treatments are the same with respect to grazers. The negative slope of the line is the grazing prey species composition, size, and abundance. In coefficient, g (Fig. 1A). Because grazers are diluted addition to altering the assemblage, so that the growth with their food, the observed rate of change in environment of prey populations may not be equivalent chlorophyll is linearly related to the dilution factor in the different fractions, these techniques are particu- (fraction of unfiltered seawater). larly destructive to the aloncate ciliates which usually The method requires 3 assumptions, discussed in dominate the microzooplankton (Gifford 198513). One some detail by Landry & Hassett (1982): first, that direct method, the pigment budget, is a true in situ phytoplankton growth rates are not density-depen- method which does not require manipulation of the dent; second, that ingestion is a linear function of microzooplankton