浅水湖泊浮水植物芡实收割对浮游植物群落的影响

浅水湖泊生态系统正遭受广泛而强烈的人为干扰,但是对收割水生植物干扰的研究甚少。于2019年8月对芡实过度生长的陈瑶湖进行通道式分区收割工程,分析了收割芡实（Euryale ferox）前后不同处理组浮游植物群落的变化。研究期间共鉴定出浮游植物6门47属72种,其中收割前63种,收割后71种。收割后浮游植物的细胞密度和生物量均高于收割前,分别增加了39.78%和5.09%。收割芡实导致陈瑶湖浮游植物群落为由蓝藻-绿藻-硅藻-隐藻群落转变为蓝藻-绿藻-硅藻群落。其中蓝藻细胞密度和生物量显著高于收割前（P<0.05）,归因于有害蓝藻（铜绿微囊藻Microcystis aeruginosa、水华束丝藻Aphanizomenon flos-aquae、小颤藻Oscillatoria tenuis、卷曲鱼腥藻Dolicospermum circinale、小席藻Phormidium tenu）的增加。收割还导致了硅藻群落由附生型向浮游型硅藻的转变,表现为尖针杆藻（Ulnaria acus）减少,而颗粒直链藻极狭变种（Aulacoseira granulata var.angustissima）、梅尼小环藻（Cyclotella meneghiniana）增加。在芡实收割过程中,未收割组和河道的浮游植物群落结构在时空分布上无显著性差异（P>0.05）,但收割组在收割后的不同阶段内差异较为明显,其细胞密度和生物量随着收割实验的结束逐渐降低。浮游植物与环境因子的相关性分析表明,水生植被覆盖度、总磷、总氮、溶解氧和叶绿素a浓度是影响浮游植物细胞密度和生物量变化的主要环境因子。综合陈瑶湖水质状态,本研究认为收割芡实并不能缓解浅水湖泊富营养化状况,研究结果为浅水湖泊水生植被的管理提供理论依据。     

Primary production of phytoplankton and periphyton in two humic lakes of a South American wetland

Seasonal primary productivities of periphyton and phytoplankton were compared in Grande Lake (GL) and a relict oxbow lake (ROL) in winter 2006 and summer 2007. GL was free of floating plants on the sampling dates and covered over 80 and 100% of the ROL surface in winter and summer, respectively. The ¹⁴C assimilation technique was used to obtain the P–E curves of phytoplankton and periphyton on artificial substrata. The periphytic maximum photosynthetic rate (P _max) was higher in the ROL in winter and summer, being better adapted to low irradiances than those in the GL. Phytoplankton and periphytic algae were light-limited in the ROL in summer due to complete coverage by floating macrophytes. In summer, P _max and α values for periphyton in the ROL were higher than those for phytoplankton, and were even higher than in GL. In turn, P _max and α values for phytoplankton in Grande Lake were higher than those for periphyton due to improved light conditions and the presence of algae that were adapted to movement through the water column. These results suggest that the complete coverage by floating macrophytes restricted phytoplankton productivity and allowed the development of a periphytic community that was better adapted to low-light conditions.     

Driving forces shaping phytoplankton assemblages in two subtropical plateau lakes with contrasting trophic status

1. We studied driving forces shaping phytoplankton assemblages in two subtropical plateau lakes with contrasting trophic status, the oligotrophic deep Lake Fuxian and the eutrophic shallow Lake Xingyun. 2. Phytoplankton samples were taken monthly for a year and phytoplankton species were sorted into the main taxonomic groups and associations proposed by Reynolds. A canonical correspondence analysis (CCA) was used to test the occurrence of these classification schemes and to determine their discriminatory power. 3. The results suggest that the major driving forces in Lake Fuxian were physical variables, and particularly the underwater light climate, whereas, nutrients were the important driving force in Lake Xingyun. 4. Top–down control through zooplankton grazing in Lake Fuxian was hardly ever a significant determinant itself, because of the scarcity of zooplankton and their low grazing efficiency of predation while a dominance of inedible cyanobacteria throughout the year rendered top–down controls ineffective failing in Lake Xingyun. Hence phytoplankton communities in both lakes appear to be regulated primarily by bottom–up controls.     

Nutrient balances and phytoplankton dynamics in two agriculturally loaded shallow lakes

The impact of agriculture was estimated on two shallow, eutrophic lakes, Lake Kotojärvi and Lake Villikkalanjärvi in southern Finland. The main emphasis was on phosphorus and nitrogen budgets and on the phytoplankton dynamics. Special attention was paid to internal P loading and blue-green algal blooms. The mean Tot-P load from agricultural land was 1.2 kg ha^-1 a^-1 in both basins and Tot-N loads were 19 kg ha^-1 a^-1 in L. Villikkalanjärvi and 12 kg ha^-1 a^-1 in L. Kotojärvi. The Tot-P input to L. Kotojärvi was on an average 0.62 g m^-2 a^-1 (per lake surface area), and the Tot-N input 9.1 g m^-2 a^-1. The corresponding inputs to L. Villikkalanjärvi were 3.1 and 57 g m^-2 a^-1, respectively. The annual variation followed the runoff volumes. About half of the Tot-P and one third of the Tot-N load was retained in L. Kotojärvi. In L. Villikkalanjärvi the retention was only 24% for Tot-P and 19% for Tot-N. The difference was very probably due to a longer theoretical retention time in L. Kotojärvi. In L. Villikkalanjärvi the mean concentration of Tot-P was 120 µg 1^-1 and that of Tot-N 1700 µg 1^-1 and the corresponding figures in L. Kotojärvi 67 and 990 µg 1^-1, respectively. The mean chlorophyll a concentration was, however, higher in L. Kotojärvi (26 µg 1^-1) than in L. Villikkalanjärvi (20 µg 1^-1). This was probably due to an internal P load in L. Kotojärvi: in 1988 the internal load of dissolved P was estimated to be as much as twofold the external load. In L. Villikkalanjärvi the internal dissolved P load was only up to 50% of the external input. In L. Kotojärvi the high internal P load coupled with a low DIN:DIP ratio resulted in a strong blue-green algal bloom in the summer of 1988. In L. Villikkalanjärvi blue-green algae were observed only in small amounts. Even in August 1990, when the DIN:DIP ratio was low enough to favor the occurrence of blue-green algae, they contributed only up to 10–15% of the total phytoplankton biomass.     

An experimental investigation of phytoplankton nutrient limitation in two contrasting low arctic lakes

We investigated whether phytoplankton communities in two lakes in SW Greenland were phosphorus or nitrogen limited. The study lakes have contrasting water chemistry (mean conductivities differ ten fold) and are located near Kangerlussuaq, SW Greenland (~67°N, 51°W). A microcosm nutrient enrichment experiment was performed in June 2003 to determine whether nitrate or phosphate addition stimulated phytoplankton growth. Samples were analysed for species composition, biomass, and alkaline phosphatase activity (APA). Initially, both lakes had extremely low total phosphorus but high total nitrogen concentrations and high APA, suggesting that the phytoplankton were phosphorus limited prior to the start of the experiment. The phytoplankton composition and biomass (mainly Ochromonas spp.) responded to phosphate but not to nitrate addition. In both lakes, chlorophyll a increased significantly when phosphate was added. Furthermore, APA was significantly lower in the two lakes when phosphate was added compared to the control and the nitrogen addition treatment. The dominance of mixotrophic phytoplankton and high DOC values suggest that these lakes may be regulated by microbial loop processes.     

Primary production studies in two linked but contrasting Welsh lakes

SUMMARY. Llyn Padarn and Llyn Peris have distinct phytoplankton populations. During 1975–76, the standing crop measured as chlorophyll- a was 5.5 times greater in Padarn than in Peris and the production rate, determined by the ¹⁴CO₂ method, was faster by 3.4 times. These differences were attributed to the higher concentrations of phosphorus in the lower lake caused by treated sewage effluent. Incident light intensity, which was slightly lower in Peris due to mountain shading, and temperature, which was 1–4°C higher in Padarn, made little significant contribution to these differences during the summer. The reduced transparency of Padarn water, compared with that of Peris, resulted from denser phytopiankton crops in Padarn. During the summer, Padarn exhibited carbon dioxide depletion which correlated with the chlorophyll concentration. Light inhibition at the surfaces of both lakes correlated with solar radiation intensity. However, the relationship between pigment content and maximum photosynthetic rate was poor. Extracellular products accounted for about 16% of the total production in the lakes. Uptake of ¹⁴C-labelled acetate was low compared with that of ¹⁴CO₂ uptake.     

The importance of bacterial utilization of released phytoplankton photosynthate in two humic forest lakes in southern Finland

Bacterial utilization of photosynthetically fixed dissolved organic carbon (PDOC) released from natural phytoplankton assemblages was studied in two small, extremely humic, forest lakes in southern Finland. Bacterial activity (measured as uptake of ¹⁴C-glucose) and phytoplankton photosynthesis (measured as light uptake of ¹⁴CO₂) could be most effectively separated using Nuclepore filters of pore size 1–2 μm. Released PDOC was 10–67% of total phytoplankton carbon fixation during in situ experiments, and represented about 0.1% of total DOC. Net uptake of PDOC by bacteria was found to be about 20% during 24 hour laboratory incubations, although about 40% of PDOC present at the start of an experiment could be utilized by bacteria during a 24 hour period. PDOC does not provide a quantitatively important substrate supply for bacterial respiration in humic forest lakes.     

Zooplankton, phytoplankton and the microbial food web in two turbid and two clearwater shallow lakes in Belgium

Components of the pelagic food web in four eutrophic shallow lakes in two wetland reserves in Belgium (Blankaart and De Maten) were monitored during the course of 1998–1999. In each wetland reserve, a clearwater and a turbid lake were sampled. The two lakes in each wetland reserve had similar nutrient loadings and occurred in close proximity of each other. In accordance with the alternative stable states theory, food web structure differed strongly between the clearwater and turbid lakes. Phytoplankton biomass was higher in the turbid than the clearwater lakes. Whereas chlorophytes dominated the phytoplankton in the turbid lakes, cryptophytes were the most important phytoplankton group in the clearwater lakes. The biomass of microheterotrophs (bacteria, heterotrophic nanoflagellates and ciliates) was higher in the turbid than the clearwater lakes. Biomass and community composition of micro- and macrozooplankton was not clearly related to water clarity. The ratio of macrozooplankton to phytoplankton biomass – an indicator of zooplankton grazing pressure on phytoplankton – was higher in the clearwater when compared to the turbid lakes. The factors potentially regulating water clarity, phytoplankton, microheterotrophs and macrozooplankton are discussed. Implications for the management of these lakes are discussed.     

Populations and production of bacteria in the plankton of two shallow lakes

The troglobitic amphipod crustacean Crangonyx antennatus populates two distinct habitats in southern Appalachian caves, mud-bottom pools and gravel-bottom streams. An investigation was conducted to determine if stream populations possess behavioral adaptations to lotic conditions. Observations on the thigmotactic behavior of C. antennatus were made seasonally in six Lee Co., Virginia caves, three containing pool habitats and three with stream habitats. Stream-dwelling amphipods were found to exhibit more cryptic behavior than those inhabiting pools under natural conditions and when disturbed. Artificial stream experiments indicate that stream-dwelling C. antennatus are better adapted behaviorally to lotic conditions than pool-dwelling amphipods.     

Nutrient limitation of bacterioplankton and phytoplankton in humic lakes in northern Sweden

1. Two small humic lakes in northern Sweden with concentrations of dissolved organic carbon (DOC) between 15 and 20 mg L^–1 were fertilized with inorganic phosphorus (P) and inorganic nitrogen (N), respectively. A third lake was unfertilized and served as a control. In addition to this lake fertilization experiment, data from different regional surveys were used to assess the role of different limiting factors.
2. The P fertilization had no effects on bacterioplankton or phytoplankton, while phytoplankton were significantly stimulated by N fertilization. Inorganic nutrient limitation of bacterioplankton was a function of DOC concentration in water of the investigated region and nutrient-limited bacteria were found only in lakes with DOC concentrations less than around 15 mg L^–1
3. The fertilization experiments demonstrated that the DOC-rich experimental lakes contained a bioavailable pool of P that was not utilized to its full potential under natural conditions. The overall mobilization of energy (bacterioplankton plus phytoplankton) in the experimental lakes was restricted by lack of inorganic N.     

Bacterial chitin hydrolysis in two lakes with contrasting trophic statuses

Chitin, which is a biopolymer of the amino sugar glucosamine (GlcN), is highly abundant in aquatic ecosystems, and its degradation is assigned a key role in the recycling of carbon and nitrogen. In order to study the significance of chitin decomposition in two temperate freshwater lakes with contrasting trophic and redox conditions, we measured the turnover rate of the chitin analog methylumbelliferyl-N,N'-diacetylchitobioside (MUF-DC) and the presence of chitinase (chiA) genes in zooplankton, water, and sediment samples. In contrast to the eutrophic and partially anoxic lake, chiA gene fragments were detectable throughout the oligotrophic water column and chiA copy numbers per ml of water were up to 15 times higher than in the eutrophic waters. For both lakes, the highest chiA abundance was found in the euphotic zone--the main habitat of zooplankton, but also the site of production of easily degradable algal chitin. The bulk of chitinase activity was measured in zooplankton samples and the sediments, where recalcitrant chitin is deposited. Both, chiA abundance and chitinase activity correlated well with organic carbon, nitrogen, and concentrations of particulate GlcN. Our findings show that chitin, although its overall contribution to the total organic carbon is small (~0.01 to 0.1%), constitutes an important microbial growth substrate in these temperate freshwater lakes, particularly where other easily degradable carbon sources are scarce.     

Relations between production and biomass of phytoplankton in four Swedish lakes of different trophic status and humic content

Production and biomass values from phytoplankton populations in four different Swedish lakes were analysed. The production in all lakes was directly proportional to biomass during homothermal periods. When the lakes were stratified there was a strong negative relation between specific growth rate and biomass. The data fitted to a logistic density dependent growth equation of the form: dB/ dt = µ_mB(1-B · K^–1) where B is the biomass, µ_m the maximum specific growth rate and K the carrying capacity. The equation was used to derive the parameters µ · µ_m ^–1 and carrying capacity (the maximum possible biomass). These parameters were then discussed in relation to light climate, phosphorus concentration and humic content.     

Live labeling technique reveals contrasting role of crustacean predation on microbial loop in two large shallow lakes

We tested the hypotheses that the ciliate assemblages in moderately eutrophic lake are controlled by the effective crustacean predation, and the high abundances of planktonic ciliates in highly eutrophic and turbid lake are due to insufficient regulation by crustacean zooplankton. A food tracer method coupled with natural assemblage of microciliates labeled with fluorescent microparticles was used to measure the cladoceran and copepod predation rates on planktonic ciliates and to estimate the carbon flow between the ciliate–crustacean trophic links. The results revealed that the microciliates (15–40 μm) were consumed by all dominant cladoceran and copepod species in both the lakes studied, mainly by Chydorus sphaericus and cyclopoid copepods in Lake Võrtsjärv, and by Daphnia spp. and Bosmina spp. in Lake Peipsi. The grazing loss in moderately eutrophic Peipsi indicated strong top-down control of ciliates mainly by cladocerans. The extraordinary abundant population of planktonic ciliates having a predominant role in the food web in highly eutrophic and turbid Võrtsjärv is explained by the measured low crustacean predation rates on ciliates. The estimated carbon flow from the ciliates to crustaceans suggest that in eutrophic lakes majority of the organic matter channeled via metazooplankton to higher trophic levels may originate from the microbial loop.     

Primary production and phytoplankton in two small,polyhumic forest lakes in southern Finland

Primary production and phytoplankton in polyhumic lakes showed a very distinct seasonal succession. A vigorous spring maximum produced by Chlamydomonas green algae at the beginning of the growing season and two summer maxima composed mainly of Mallomonas caudata Iwanoff were typical. The annual primary production was ca. 6 g org. C · m^–2 in both lakes. The mean epilimnetic biomass was 1.1 in the first lake and 2.2 g · m^–2 (ww) in the second one. The maximum phytoplankton biomass, 14 g · m^–2, was observed during the vernal peak in May.     

Structure, biomass, production and depth distribution of periphyton on artificial substratum in shallow lakes with contrasting nutrient concentrations

1. To examine how the vertical distribution of periphytic biomass and primary production in the upper 0–1 m of the water column changes along an inter‐lake eutrophication gradient, artificial substrata (plastic strips) were introduced into the littoral zones of 13 lakes covering a total phosphorus (TP) summer mean range from 11 to 536 μg L^?1. Periphyton was measured in July (after 8 weeks) and September (after 15 weeks) at three water depths (0.1, 0.5 and 0.9 m). 2. Periphyton chlorophyll a concentration and dry weight generally increased with time and the communities became more heterotrophic. Mean periphytic biomass was unimodally related to TP, reaching a peak between 60 and 200 μg L^?1. 3. The proportion of diatoms in the periphyton decreased from July to September. A taxonomic shift occurred from dominance (by biovolume) of diatoms and cyanobacteria at low TP to dominance of chlorophytes at intermediate TP and of diatoms (Epithemia sp.) in the two most TP‐rich lakes. 4. The grazer community in most lakes was dominated by chironomid larvae and the total biomass of grazers increased with periphyton biomass. 5. Community respiration (R), maximum light‐saturated photosynthetic rate (P_max), primary production and the biomass of macrograzers associated with periphyton were more closely related to periphyton biomass than to TP. Biomass‐specific rates of R, P_max and production declined with increasing biomass. 6. Mean net periphyton production (24 h) was positive in most lakes in July and negative in all lakes in September. Net production was not related to the TP gradient in July, but decreased in September with increasing TP. 7. The results indicate that nutrient concentrations alone are poor predictors of the standing biomass and production of periphyton in shallow lakes. However, because periphyton biomass reaches a peak in the range of phosphorus concentration in which alternative states occur in shallow lakes, recolonisation by submerged macrophytes after nutrient reduction may potentially be suppressed by periphyton growth.     

Zooplankton spatial patterns in two lakes with contrasting fish community structure

Horizontal and temporal patterns in crustacean zooplankton communities were analyzed in two small, oligotrophic lakes which were morphologically and chemically similar, but had contrasting fish communities. Ranger Lake was dominated by two bass species and the planktivores numbered < 25 ind. ha^–1. Mouse Lake had no large piscivores and planktivores numbered > 1200 ind. ha^–1. There were significant differences in the distribution of zooplankton taxa and size classes between sampling stations. In Ranger Lake, the smallest size classes were more abundant at the deeper stations and the larger individuals were more abundant at the shallower stations. In Mouse Lake, the smaller individuals were more common at the shallow stations and the larger individuals were more common at the deeper stations. These differences suggest medium scale patterns induced by vectorial forces, but modified by species specific migration patterns. We tested the hypothesis that horizontal heterogeneity should be influenced by planktivore density and found that none of the taxa showed significant between-lake differences in the variance-mean regressions. We also tested the hypothesis that larger taxa should be more heterogeneous and we found that cladocerans were more heterogeneous than copepods and nauplii. In terms of sampling methodology our data suggest that the between-station variability was so high that a single mid-lake sample would certainly lead to completely unacceptable errors in the estimation of population densities and biomasses.     

Cladoceran and rotifer grazing on bacteria and phytoplankton in two shallow eutrophic lakes: in situ measurement with fluorescent microspheres

Metazooplankton grazing on bacteria and on the phytoplanktonof various sizes was estimated in shallow eutrophic lakes Kaiavereand Võrtsjärv (Estonia) by in situ feeding experimentswith fluorescent microspheres (diameters 0.5 µm for bacteriaand 3, 6 and 24 µm for phytoplankton). Zooplankton communitycomposition, abundance and food density were important factorsdetermining grazing rates in these lakes. Cladocerans and rotifersfiltering rates (FR) and ingestion rates (IR) on bacteria andphytoplankton were several times higher in Lake Kaiavere wherebacterivorous rotifers and Daphnia contributed more to zooplanktonassemblage. While cladocerans were generally the main phytoplanktonconsumers, both lakes differed with respect to the groups ofbacterivores. Based on consumption of fluorescent microspheres,the metazooplankton grazing rates were relatively low and hadlow impact on production and standing stock of bacteria andingestible phytoplankton (<30 µm). On average, 0.5and 0.1% of standing stock of bacteria and 2.6 and 1.0% of standingstock of ingestible phytoplankton was grazed daily by metazooplanktonin lakes Kaiavere and Võrtsjärv, respectively. Thatcorresponded to daily grazing of 4.1% of the bacterial productionand 0.43% of the total primary production (PP) by metazooplanktonin Lake Kaiavere compared with 4.3 and 0.06% in Lake Võrtsjärv,respectively. The results suggest that the majority of consumptionof the bacterial and phytoplankton PP is most likely channelledthrough the microbial loop.     

Productivity of clear and humic lakes: nutrients,phytoplankton, bacteria

The relationships between long-term surface average concentrations of humic acids measured as water colour, dissolved organic carbon (DOC) or Secchi disk transparency and trophic state variables were studied with literature data from more than 600 freshwater lakes. The geometric means of summer surface average nutrient (phosphorus and nitrogen) concentration, phytoplankton biomass (chlorophyll concentration), and hypolimnetic anoxia (anoxic factor) were significantly higher in coloured than in clear lakes. The regressions of colour or DOC on these trophic state variables were positive and significant throughout a range of three orders of magnitude. Phytoplankton or primary productivity was higher in coloured lakes, when expressed per volume of epilimnion. Annual integral primary productivity expressed on an areal basis was smaller in coloured lakes, probably a reflection of shallower phototrophic depths in these lakes. There is evidence that annual integral bacteria productivity is much higher in coloured lakes for two reasons: first, epilimnetic bacteria production was ca. four times higher in coloured lakes, second, other studies have shown that hypolimnetic bacteria production is commonly higher than epilimnetic production, especially in anoxic hypolimnia that are frequent in coloured lakes. All volumetrically expressed variables indicated higher productivity in coloured lakes. In addition, high bacteria productivity reflects a different food chain involving mixotrophs, possibly compensating for low light conditions. Our analyses indicate that primary and secondary productivity is as high as or higher than in clear lakes. This revised version was published online in July 2006 with corrections to the Cover Date.     

Vertical distribution of primary production and phytoplankton in two small lakes with different humus concentration in southern Finland

Vertical distribution of primary production and phytoplankton was studied in a polyhumic brownwater lake and in an oligo-mesohumic lake. During summer both lakes were thermally, chemically and biologically stratified. In the brownwater lake primary production was restricted to the uppermost layer of 1–1.5 m of epilimnion. In the oligo-mesohumic lake noticeable primary production was detected down to depths of 2–3 m. The ice-free period primary production was about 20% higher in the oligo-mesohumic lake, though occasionally the surface production was 2–3 times higher in the brownwater lake. Epilimnetic total phosphorus and total nitrogen concentrations were higher in the brownwater lake, while nitrate-nitrite, ammonium and phosphate concentrations were very low in both lakes.
Phytoplankton was confined to the uppermost productive layer in the brownwater lake. In the oligo-mesohumic lake phytoplankton was distributed more evenly, though the mean maximum biomass was at the depth of 3–4 m. Below the oxic water layer biomass decreased abruptly in both lakes. In the oligo-mesohumic lake chlorophyll concentration was extremely high (max. 320 mg chl a m⁻³) in the anoxic hypolimnion, due to green sulphus bacteria.
Flagellated chlorophytes and thrysophytes dominated in the brownwater lake; in spring Chlamydomonas species, followed by Mallomonas caudata . In the oligo-mesohumic lake small coccal green algae, such as Oocystis, Scenedesmus and Westella -like species, dominated in mid-summer, and chrysophytes and cryptomonads in autumn.     

Relationship between epipelon,epiphyton and phytoplankton in two limnological phases in a shallow tropical reservoir with high Nymphaea coverage

Macrophytes and phytoplankton are recognized as having roles in determining alternative stable states in shallow lakes and reservoirs, while the role of periphyton has been poorly investigated. Temporal and spatial variation of phytoplankton, epipelon and epiphyton was examined in a shallow reservoir with high abundance of aquatic macrophytes. The relationships between algae communities and abiotic factors, macrophyte coverage and zooplankton density were also analyzed. Monthly sampling was performed in three zones of the depth gradient of the reservoir. Two phases of algal dominance were found: a phytoplankton phase and epipelon phase. The phase of phytoplankton dominance was characterized by high macrophyte coverage. Rotifera was the dominant zooplankton group in all the zones. Flagellate algae were dominant in phytoplankton, epipelon and epiphyton. Macrophyte coverage was found to be a predictor for algal biomass. Changes in biomass and species composition were associated with macrophyte cover variation, mainly the Nymphaea. In addition to the abiotic factors, the macrophyte coverage was a determining factor for changes to the algal community, contributing to the alternation between dominance phases of phytoplankton and epipelon. The macrophyte–phytoplankton–periphyton relationship needs to be further known in shallow reservoirs, especially the role of epipelon as an alternate stable state.