Patterns in the top-down versus bottom-up regulation of heterotrophic nanoflagellates in temperate lakes

The importance of resources versus that of predators in determiningthe abundance of organisms is modulated by several factors whichvary over time and space. Here, we evaluate the abundance ofheterotrophic nanoflagellates (HNF), the main predators of planktonicbacteria, in the plankton of 16 Quebec lakes as a function ofthe abundance of resources and zooplankton (predators). We analyzethe data at two different scales: among lakes varying in resourcesupply and within lakes. Resource factors (total phosphorus,chlorophyll, bacterial concentration) best explained the abundancesof HNF among lakes differing in resource availability. Whenthe effect of resource availability was removed, the biomassof cladocerans became an important determinant of HNF abundance,with abundance declining and the ratio of bacteria per flagellateincreasing, as the cladocerans increased. Within lakes, HNFabundance was not a direct function of resource concentration.Multiple regression models based on data for many lakes showedthat both bottom-up (food) and top-down (predation) variablespredicted HNF abundance quite well. An examination of the seasonalpattern among individual lakes showed each group of variablesdiffering in importance: bottom-up variables were more importantin spring and top-down variables in midsummer. An examinationof the effects of different assemblages of zooplankton on HNFshowed cladocerans to exert the highest negative impact on HNFand to be responsible for keeping low HNF numbers during mostof the summer with that probably affecting the pathways of energyflux in those lakes dominated by cladocerans versus those dominatedby other zooplankton. The absolute, as well as the relative,importance of the different biotic and abiotic factors examinedin predicting the abundance of HNF is a function of the scaleof the investigation. Among lakes varying greatly in resourcesupply, resources are found to better correlate with HNF abundance.When the analysis is performed at another scale, within lakes,the relative importance of resources and predation in determiningHNF abundances changes depending on the variability superimposedby other factors (seasonality and lake food web structure).     

Water discharge-regulated bacteria/heterotrophic nanoflagellate (HNF) interactions in the water column of the river Rhine

Heterotrophic nanoflagellates (HNF) make up a large fraction of the zooplankton biomass of rivers. Their abundance can be strongly affected by water discharge, but the consequences of this highly dynamic factor for their main prey, the bacteria, is still unknown. The focus of this study was on bacterial/HNF interactions in the Lower River Rhine (Germany) with respect to the discharge-dependent dynamics. The bacterial and HNF abundances and biomasses were determined over the course of 17 months. The potential consumption of bacteria by HNF was calculated based on the biomass data and on data on the HNF production. The mean bacterial abundance in the Rhine at Cologne ranged from 0.3 x 10(6) to 3.5 x 10(6) cells mL(-1), with lowest abundances in winter and highest in late spring. No significant changes in abundance during the downstream passage were found. Neither could a significant correlation be found between bacterial and HNF abundance. The ratio of bacterial to HNF abundance showed high variations which lay between 166 and 19,055 and was negatively dependent on water discharge. Monthly routine calculations on the potential bacterial consumption by HNF revealed a clearance of between 2 and 82% of the bacterial standing stock d(-1). The values increased greatly with water discharge and could exceed 100% d(-1) at times of high water flow. The presented data suggests a change in the top-down control of the planktonic bacteria due to the water discharge: The importance of benthic predation at low water flow (high contact probability to benthic predators) gives way to an increased importance in predation by planktonic HNF at high water flow.     

Top–down control of benthic heterotrophic nanoflagellates by oligochaetes and microcrustaceans in a littoral freshwater habitat

1. We investigated trophic interactions between benthic heterotrophic nanoflagellates (HNF) and oligochaetes and microcrustaceans (cladocerans and copepods) transferred from a silty, littoral freshwater habitat to laboratory microcosms. With a newly adapted experimental design we were able to compute (i) predation rates on benthic HNF by the tested metazoan organisms and (ii) growth rates of the natural benthic HNF population when losses because of this predation were excluded. 2. The experiments covered a temperature range of 4–27 °C and a fivefold variation of predator densities (September 2000–February 2002). For 60% of these experiments, significant predation of oligochaetes and microcrustaceans on benthic HNF was revealed. Predation rates on HNF ranged from 0 to 0.256 day^?1. Growth rates of the benthic HNF assemblage varied from ?0.098 to 0.353 day^?1; they were used to estimate the significance of the measured losses in comparison with possible other loss factors. 3. The data suggested that during the major part of the year a high percentage of the HNF production was consumed in the surficial sediment of the investigated system, resulting in a relatively constant and low HNF standing stock and an uncoupling of benthic bacteria and their protistan grazers. Top–down control by microcrustaceans and oligochaetes was identified as one significant, frequently prevailing regulatory factor, while other parameters responsible for the control of benthic HNF densities remain to be examined.     

The diversity and distribution of heterotrophic nannoflagellates in the eutrophic Lake Nero

The fauna and abundance of heterotrophic nannoflagellates (HNF) and the quantitative distribution of their main food items (bacteria) have been studied in the highly eutrophic Lake Nero (Yaroslavskaya oblast). A total of 70 HNF species and forms were found, with representatives of the order Choanoflagellida dominant. Abundances of HNF and bacteria were high, reaching levels typical for productive waters. The pattern of HNF seasonal dynamics was characterized by two peaks in June and September and one low in August. The minimal levels of HNF coincided with the peak of bacterioplankton abundance.     

Trophic interactions between viruses, bacteria and nanoflagellates under various nutrient conditions and simulated climate change

Population dynamics in the microbial food web are influenced by resource availability and predator/parasitism activities. Climatic changes, such as an increase in temperature and/or UV radiation, can also modify ecological systems in many ways. A series of enclosure experiments was conducted using natural microbial communities from a Mediterranean lagoon to assess the response of microbial communities to top-down control [grazing by heterotrophic nanoflagellates (HNF), viral lysis] and bottom-up control (nutrients) under various simulated climatic conditions (temperature and UV-B radiations). Different biological assemblages were obtained by separating bacteria and viruses from HNF by size fractionation which were then incubated in whirl-Pak bags exposed to an increase of 3°C and 20% UV-B above the control conditions for 96 h. The assemblages were also provided with an inorganic and organic nutrient supply. The data show (i) a clear nutrient limitation of bacterial growth under all simulated climatic conditions in the absence of HNF, (ii) a great impact of HNF grazing on bacteria irrespective of the nutrient conditions and the simulated climatic conditions, (iii) a significant decrease in burst size (BS) (number of intracellular lytic viruses per bacterium) and a significant increase of VBR (virus to bacterium ratio) in the presence of HNF, and (iv) a much larger temperature effect than UV-B radiation effect on the bacterial dynamics. These results show that top-down factors, essentially HNF grazing, control the dynamics of the lagoon bacterioplankton assemblage and that short-term simulated climate changes are only a secondary effect controlling microbial processes.     

Seasonal dynamics of free and attached heterotrophic nanoflagellates in an oligomesotrophic lake

1. The seasonal development of heterotrophic nanoflagellates (HNF), bacteria, rotiferans and crustacean zooplankton was studied in the epilimnion of Lake Pavin, an oligomesotrophic lake in the Massif Central of France.
2. HNF abundance varied from 0.1 to 2.5 × 10³ mL^–1. Free-living HNF reached their highest density in spring when the copepod Acanthodiaptomus denticornis dominated the metazooplankton. They were present in low numbers when rotifers and cladocerans were numerous.
3. Attached HNF, consisting of bicoecids and choanoflagellates, were fixed to large diatoms and to the colonial cyanobacterium Anabaena flos-aquae . The abundance of attached HNF was significantly correlated to bacterial abundance, which fluctuated between 1.1 and 2.7 × 10⁶ mL^–1. Highest abundance of these epiphytic protists was recorded when free-living heterotrophic nanoflagellates declined.
4. The comparison of the dynamics of heterotrophic nanoflagellates, bacteria, and the impact of zooplankton grazing suggested that prey abundance, the presence of suitable attachment sites and limited competition from the free-living forms were the main factors controlling the development of the epiphytic flagellate protists. In contrast, the low abundance of free-living forms during the period of rotiferan and cladoceran development suggests the prevalence of a top-down control by predation of the metazoopankton.     

Bacterivory of metazooplankton, ciliates and flagellates in a newly flooded reservoir

Bacterial consumption by metazoan zooplankton and phagotrophic protists was measured in situ during the period of thermal stratification in the epilimnion (1 m) and metalimnion (7 m) of a newly flooded reservoir (Sep reservoir, France). The mean bacterial consumption was 2.53 x 10⁶ bacterial l^-1 h^-1 at 1m and 0.97 x 10⁶ bacteria l^-1 h^-1 at 7m. The main consumers over the whole study period were the cladocerans Daphnia longispina and Ceriodaphnia quadrangula, accounting on average for 72% of the potential total predation of bacteria at 1 m and 56% at m, especially during the months of May-June and August. Heterotrophic nanoflagellates (HNF), which accounted for 12% estimated total predation of bacteria at 1 m and 13% at 7m, only exerted a limited predation, mainly by a Monas-type cell. Ciliates, dominated in terms of abundance by Pelagohalteria viridis, accounted for 4% of total predation in the epilimnion (0.00-0.42 x 10⁶ bacteria l^-1 h^-1). In a newly flooded reservoir, metazoan zooplankton appear to be the main consumers of bacteria. Predation of ciliates and HNF by zooplanktonic crustaceans could account for the low contribution of components of the microbial loop to bacterial consumption.      

Bacterioplankton production and protozoan bacterivory in a mesotrophic reservoir

Protozoan bacterivory [via uptake of fluorescently labelledbacteria (FLB)] and production of bacteria ([³H]thymidine assay)were simultaneously measured in the mesotrophic ímovReservoir (Southern Bohemia) from April to November, 1988. Heterotrophicnanoflagellates (HNF) were mostly responsible for a greaterfraction of protozoan bacterivory during the spring period.From 10 to 23% of bacterial production was grazed daily withthe only exception of the spring peak of ciliate abundance (upto 60%). Protozoans decreased significantly during the clearwaterphase (ciliates disappeared), and thus their role in bacterivorywas negligible. Through the summer-fall period ciliates, notHNF, were the most important bacterial micrograzers. Protozoancommunity grazing balanced or even exceeded the daily bacterialproduction in August and September. Alternate fates of bacterialproduction besides protozoan grazing during the spring periodare discussed.     

Changes in cell volume of bacteria and heterotrophic nanoflagellates in a hypereutrophic pond

Changes in cell volume of planktonic bacteria and heterotrophic nanoflagellates (HNF) were examined in a hypereutrophic pond from April to October, 1997. There were marked changes in the abundance of bacteria, HNF and ciliates and in protistan bacterivory during this period. The cell volume of free-living bacteria (0.121 ± 0.031 m³, mean ± SD) was large relative to that reported in the literature. The cell volumes of HNF was 71.1 ± 24.8 m³. Both cell volumes did not follow a seasonal trend. The dominant size class of bacteria was seasonally variable, whereas density of filamentous bacteria was relatively high between August and September. Biomass of filamentous bacteria accounted for up to 33.6% of total bacterial biomass. A correlation analysis for cell volume of bacteria and HNF, density of filamentous bacteria and some microbial variates was performed. The positive correlations detected (p<0.05) were between density of bacteria and cell volume of HNF, and between density of filamentous bacteria and cell volume of HNF.     

Fern-spore-feeder interaction in temperate forests in Japan: Sporing phenology and spore-feeding insect community

Angiosperms are widely appreciated to have flowering schedules, but far less attention has been paid to the timing of spore production by ferns. Although a range of abiotic factors are likely responsible for the timing of fern sporing, spore predation by specialist spore-feeding insects may also exert selective forces on timing. As a step toward understanding ecological factors that affect the evolution of fern sporing phenology, we tracked annual sporing patterns and examined associations with spore-feeding insects in 38 ferns species in two Japanese temperate forests. Most sporing occurred during June through August, the period of highest temperature and precipitation at the study sites, but some species produced spores during a limited period in spring or very late autumn. Over 70% of all species examined were attacked by spore-feeders, which consisted of seven polyphagous and five oligophagous species. Spore feeders occurred predominantly during June through September, and stathmopodid moth larvae consumed up to 70% of the mature sporangia. Thus, although the warm and humid conditions in the summer is likely favored for prothallial growth and fertilization, spring or late-autumn sporing in some species may have evolved as an adaptation to escape spore predation by spore feeders.     

Predation of released pheasants <Emphasis Type="Italic">Phasianus colchicus</Emphasis> on lowland farmland in the UK and the effect of predator control

We present data accumulated over the last 25 years on predation of radio-tracked released pheasants. In studies of birds during the autumn/winter at six pheasant shoots with high-density releases managed by full-time gamekeepers, predation of released pheasants by foxes before the shooting season began (July–September) averaged 19.2?±?4.0% per site, and during the shooting season (October 1st–February 1st), a further 15.9?±?1.9% were predated. The range in 3-year average predation rates between sites before shooting began was 8.6 to 42.4%. At seven different sites during the spring and summer, between 20 and 71% of released or wild hens that survived the shooting season were predated, mainly by foxes, between mid-March and mid-July. Predation was significantly higher at sites with low-level predator control (59?±?4.7%) compared to those with high-level control (30?±?5.3%). At three of the four sites with low predator control between 5 and 22% of nest failures were caused by incubating hens being predated by foxes. Our data quantify for the first time highly variable predation rates of released pheasants before and during the winter shooting season which we suggest was influenced by a range of site and management factors. During the spring and summer, our data provide evidence that predation of adult hen pheasants as well as nest predation can suppress breeding success and that predator control can reduce these losses.     

The annual cycle of heterotrophic freshwater nanoflagellates: role of bottom-up versus top-down control

The population dynamics of heterotrophic nanoflagellates (HNF)were analyzed in pre-alpine Lake Constance over three consecutiveyears. A recurrent seasonal pattern led to the identificationof five seasonal phases: winter, spring, clear-water, summerand autumn. HNF biomass was lowest in winter and highest m latespring several weeks after the phytoplankton spring bloom. Theaverage biomass of HNF was 5–12% of bacterial biomassand 13–34% of ciliate biomass respectively. The largestHNF cells were recorded during the spring phase, whereas theaverage cell size was reduced to one-third during the subsequentclear-water phase. The pronounced differences in the mean cellsize were attributed mainly to varying grazing impact on HNFThroughout most of the year, HNF production was balanced bygrazing of microzooplankton, namely ciliates, within the microbialloop. During the dear-water phase, however, the grazing impactwas mainly due to rotifers and daphnids. Changing grazing impactwas primarily responsible for the observed 2-fold interannualdifference m the mean biomass of HNF Overall, top-down controlby grazing was more important in governing the population dynamicsof HNF than bottom-up control by bacterial food supply.     

Strong seasonal differences of bacterial polysaccharide utilization in the North Sea over an annual cycle

Marine heterotrophic bacteria contribute considerably to global carbon cycling, in part by utilizing phytoplankton-derived polysaccharides. The patterns and rates of two different polysaccharide utilization modes – extracellular hydrolysis and selfish uptake – have previously been found to change during spring phytoplankton bloom events. Here we investigated seasonal changes in bacterial utilization of three polysaccharides, laminarin, xylan and chondroitin sulfate. Strong seasonal differences were apparent in mode and speed of polysaccharide utilization, as well as in bacterial community compositions. Compared to the winter month of February, during the spring bloom in May, polysaccharide utilization was detected earlier in the incubations and a higher portion of all bacteria took up laminarin selfishly. Highest polysaccharide utilization was measured in June and September, mediated by bacterial communities that were significantly different from spring assemblages. Extensive selfish laminarin uptake, for example, was detectible within a few hours in June, while extracellular hydrolysis of chondroitin was dominant in September. In addition to the well-known Bacteroidota and Gammaproteobacteria clades, the numerically minor verrucomicrobial clade Pedosphaeraceae could be identified as a rapid laminarin utilizer. In summary, polysaccharide utilization proved highly variable over the seasons, both in mode and speed, and also by the bacterial clades involved.     

Breeding of the bush fly,Musca vetustissima Walker,in an over-wintering area during winter and spring

Field experiments in an area of south-western Australia where bush fly (Musca vetustissima) adults occur permanently, showed that it overwintered by continuous breeding; but only two to three generations occurred between June and September. Low survival of eggs and larvae in the food (cattle dung) from May to August was associated with rainfall rather than low temperatures. High egg-adult survival occurred in late August; lower egg and larval survival in September and early October was attributed partly to dung fauna. Egg and larval survival was high in late October until December but predation or parasitization caused low survival of puparia outside the dung. Experimental and wild adult flies were largest and therefore most fecund in early spring; smaller wild flies in late spring appeared to result from larval competition for food. Simultaneous high egg-adult survival, high fecundity and rapidly decreasing generation times in late winter and early spring provide a basis for explaining the major increase in adult bush fly abundance observed in some areas in mid-spring.     

Seasonal,spatial and diel partitioning of Acyrthosiphon pisum (Hemiptera: Aphididae) predators and predation in alfalfa fields

Predators are important natural enemies, often responsible for preventing pest population outbreaks of in many crops. Complementarity in resource use involves spatial or temporal segregation of predators, which can result in better biological control when several species of natural enemies share a prey. In this study, we investigated the seasonal, spatial and diel segregation of Acyrthosiphon pisum predators and its predation in alfalfa fields, by setting out cards with sentinel aphids, and making observations every 3 h for a 24 h period. A temporal and spatial segregation of predators was observed. Coccinellids were the most abundant predators, representing 51% of the total observations, followed by syrphid larvae. Coccinellids were also responsible for high levels of predation throughout the year, although the species responsible varied from spring to summer and autumn. On the other hand, syrphids were only found in spring and summer, while spiders only in autumn. Predator species also differed on their preferred sites for predation, with Heteropterans and syrphids found on the foliage, the spider Neomaso articeps only on the ground, and coccinellid and Anyphaenidae species on both sites. The two main predator groups also showed distinct diel patterns, with coccinellids observed only during day and syrphids only during night. This predatory activity corresponded with aphid predation, observing more predation in spring, on the foliage and during the day time. The proportion of predators observed preying on cards in the different seasons did not corresponded tightly with their field abundance, particularly in the case of coccinellids, which maintained high levels of predation in spite of great variations in its field abundance. Our results support the hypothesis of a spatio-temporal segregation of the predators associated with A. pisum in alfalfa, which might be beneficial for the outcome of biological control of this pest.     

Contrasting effects of a cladoceran (Daphnia galeata) and a calanoid copepod (Eodiaptomus japonicus) on algal and microbial plankton in a Japanese lake, Lake Biwa

Macrozooplankton may affect algal and microbial plankton directly through grazing or predation and indirectly through nutrient regeneration. They may also affect potential prey positively by removing alternative predators. Here, we examined the effects of a cladoceran (Daphnia) and a calanoid copepod (Eodiaptomus) on algal and microbial plankton in a Japanese lake using in situ experiments in which we manipulated the nutrient supply and biomass of these macrozooplankton. The response of algal and microbial plankton to macrozooplankton was diverse and varied depending on the level of nutrient supply. Eodiaptomus seemed to feed mainly on large algae (>20 µm) and microzooplankton, while direct grazing by Daphnia on algae, bacteria, heterotrophic nanoflagellates (HNF), and microzooplankton (ciliates, heliozoa, and rotifers) was pronounced. Trophic linkages within these microbial plankton was also suggested; bacteria were grazed by HNF and these in turn were grazed by microzooplankton. When the nutrient supply was high, both HNF and microzooplankton were exposed to higher amounts of algae and lower bacterial abundance. Moreover, nutrient regeneration by daphnids and Eodiaptomus copepods seemed to differentially stimulate the growth of algae and bacteria. The results suggest that the relationship between macrozooplankton and microbial plankton cannot be fully understood without taking into consideration not only the feeding characteristics of the macrozooplankton, but also the food web structure, the subsidized algal resource, and nutrient regeneration from the macrozooplankton.     

Predator exclusion experiments in an intertidal mud flat

The intensity of predation pressure exerted on the macrofauna in muddy sediments was investigated in Königshafen (island of Sylt, eastern North Sea). Tests in aquaria revealed shrimpCrangon crangon, juveniles of the shore crabCarcinus maenas, and gobiid fishPomatoschistus microps as the most important local predators. Their high abundance from July to September led to the hypothesis, that predation pressure on the infauna will be most intense during that period. Cages were set up to protect the infauna against these predators. No protective effect was achieved with cages constructed of 20-mm mesh nylon net; however, cages with screen wire and gauze of 5-mm mesh and smaller resulted in a considerable increase of the infauna. In cages of 1-mm mesh gauze, set up from March to June, the macrofauna reached an abundance of four times the control density. In the period from July to October, the factor of increase was as high as 23. Species density was 28 400 cm^–2, as compared with only 7 in the control. Species which colonized the mud flat during spring and early summer established dense settlements inside the cages, but failed to do so in uncaged mud-flat areas. It is concluded that predation by young crabs, shrimp and gobies determines to a large extent the structure and dynamics in the local intertidal macrofauna.     

Seasonal and vertical distribution of planktonic bacteria and heterotrophic nanoflagellates in the middle Adriatic Sea

Temporal and spatial patterns of bacteria and heterotrophic nanoflagellates (HNF) were studied monthly from January 1997 to December 1998 in the middle Adriatic Sea. Bacterial and HNF relationships with phytoplankton biomass and temperature were analyzed to examine how the relative importance of bottom-up and top-down factors may shift over seasons and locations. For the coastal area, an inconsistent relationship between bacterial abundance and chlorophyll a and a stronger relationship between bacterial abundance and bacterial production suggest that other substrates than those of phytoplankton origin are important for bacteria. The analysis of simultaneous effects of temperature and bacterial production on bacterial abundance showed that the effect of temperature obscured the effects of bacterial production, suggesting that bacterial growth itself is highly temperature-dependent. The relationship between HNF abundance and bacterial abundance was slightly improved by the inclusion of in situ temperature, bacterial production or both parameters, as additional independent variables. About 60% of the variability in HNF abundance can be explained by bacterial abundance, bacterial production and temperature. In the open sea, tight coupling of bacterial abundance with chlorophyll a concentrations implied that phytoplankton-derived substrates have a dominant role in controlling bacterial abundance. During the colder months, bacterial abundance was high enough to support higher HNF abundance than observed, suggesting that predation exerted a minor depressing influence on bacterial abundance during that period. During the spring-summer period, HNF controlled bacterial standing stocks by direct cropping of bacterial production.Communicated by: H.-D. Franke     

Prey selectivity of bacterivorous protists in different size fractions of reservoir water amended with nutrients

An experiment designed to examine food preferences of heterotrophic nanoflagellates (HNF) grazing on bacterioplankton was performed in the freshwater Rímov reservoir (Czech Republic). Water samples were size-fractionated to obtain < 5 microm filtrate containing bacteria and HNF. To manipulate resource availability, < 5 microm treatments were incubated in dialysis bags submerged in the barrels filled with the unfiltered reservoir water amended with either orthophosphate or glucose or combination of both. We employed rRNA-targeted probes to assess HNF prey preferences by analysing bacterial prey in HNF food vacuoles compared with available bacteria. Actinobacteria (the HGC69a probe) were avoided by HNF in all treatments. Cytophaga-Flavobacterium-Bacteroidetes bacteria (the CF319a probe) were positively selected mainly in treatments in which bacteria were heavily grazed, the < 5 microm treatments, but this trend was less pronounced towards the end of the study. The members of a small subcluster of Betaproteobacteria (the R-BT065 probe) were mostly positively selected. The nutrient amendments differentially affected bacterioplankton dynamics in almost all treatments, and together with the size fractionation, altered HNF overall bacterivory as well as prey selection. Analyses of bacterivores in unfiltered treatments allowed to detect the effect of different protists on shifts in HNF selectivity observed in < 5 microm compared with unfiltered treatments.     

Microbial communities in oligotrophic Lake Toya,Japan

 The vertical distributions of planktonic bacteria, chroococcoid cyanobacteria, and heterotrophic nanoflagellates (HNF) were examined in Lake Toya, an oligotrophic lake located in Hokkaido, the northern island of Japan, at monthly intervals from May 1993 to May 1994. The abundances of bacteria, cyanobacteria, and HNF during the study period ranged from 10⁴ to 10⁵,10² to 10⁴, 10 to 10² cells ml⁻¹, respectively. The range of bacterial abundances was among the lowest previously reported from other oligotrophic lakes. The vertical distributions of both bacteria and cyanobacteria were influenced by thermal stratification: they were homogeneous throughout the water column during the circulation period (January to April) and heterogeneous during the period of stratification (June to November). During the period of stratification, the cyanobacterial abundance decreased toward the surface in the euphotic zone while the frequency of diving cells (FDC) increased, suggesting that grazing pressure was high near the surface. This hypothesis was supported by the relatively high abundance of HNF at the surface and the negative correlation between HNF and cyanobacterial abundances in the euphotic zone (r = −0.503, n = 33, P < 0.05). On the other hand, multiple regression analysis revealed that 52% of the variation in bacterial abundance in the stratified period can be explained by chlorophyll a concentration, water temperature, and HNF abundance (df = 3, 45; F = 16.2; P < 0.01), suggesting that both substrate limitation and grazing loss by HNF were important factors controlling bacterial abundance in the lake. Received: June 21, 2002 / Accepted: October 16, 2002 Present address: 5-2-2-18-805 Kikusui-motomachi, Shiroishi, Sapporo 003-0825, Japan Present address: School of Environmental Science, University of Shiga Prefecture, 2500 Hassaka-cho, Hikone 522-8533, Japan Tel. +81-749-28-8307; Fax +81-749-28-8463 e-mail: ban@ses.usp.ac.jp Acknowledgments We thank Dr. H. Ueda for encouraging this study and Mr. Haruna and the members of the Plankton Laboratory, Hokkaido University, for their help in sampling. We also thank two anonymous reviewers for their critical comments. Correspondence to:S. Ban