1 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Kalevi Salonen1, Pia Högmander1, Victor Langenberg2, Hannu Mölsä3, Jouko Sarvala4, Anne Tarvainen1 & Marja Tiirola1 Jellyfish Limnocnida tanganyicae - A semiautonomous microcosm in the food web of Lake Tanganyika 1 Department of Biological and Environmental Sciences, PO Box 35 (YAC), 40014 University of Jyväskylä, Finland 2 DELTARES, Department of Water Quality and Ecology, PO Box 177, 2600 MH Delft, The Netherlands 3 Fish Innovation Centre, Lohitie 701, 72210 Tervo, Finland 4 Department of Biology, University of Turku, 20014 Turku, Finland kalevi.salonen@jyu.fi, phone 358407592829 Keywords: Genetics, stable isotopes, symbiosis, UV light 1 2 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 Abstract Jellyfish are important members of marine food webs, but are rare in lakes. In one of the largest lakes in the world, Lake Tanganyika, a small Limnocnida tanganyicae jellyfish is a prominent component of zooplankton, but its role in the ecosystem has remained obscure. In this study, we addressed the role of jellyfish in Lake Tanganyika using several approaches. These jellyfish occasionally reached high densities locally. They often inhabited the whole epilimnetic water column. In particular, the largest individuals showed distinct, low amplitude, diel vertical migration, which seemed to be crucial to avoid harmful UV radiation. Vertical migration and consequent adjustment to light intensity also might be important for picocyanobacteria that were regularly present in variable quantities in Tanganyika jellyfish. In different individuals, endosymbiotic picocyanobacteria were morphologically variable and dominated by a particular Lake Biwa type Cyanobium species, which typically are abundant in the Tanganyika water column. Under light, some jellyfish even proved to be net primary producers. Nitrogen stable isotopic ratios indicated that while the free-living cyanobacteria were nitrogen-fixers, the internal picocyanobacteria in jellyfish obtained their nitrogen predominantly from their host. Stable isotopic ratios of carbon and nitrogen further suggested copepod zooplankton as the most likely prey for the jellyfish. Lake Tanganyika jellyfish apparently base their metabolism both on carnivorous and herbivorous diets, with possible internal cycling of nutrients; however, the role of picocyanobacteria gardening for the ecosystem of Lake Tanganyika and its jellyfish requires quantification. 2 3 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 Introduction Lake Tanganyika is the second deepest and oldest lake on the earth. Its numerous endemic biota reflect the history of about ten million years under rather stable conditions prevailing near the equator (Tiercelin & Mondeguer, 1990; Cohen et al., 1993). The total biodiversity of the lake, one of the highest in the world (Coulter, 1994), is largely confined to the littoral zone. Its pelagic biodiversity, by contrast, is low and leads to a rather simple food web. In Lake Tanganyika, a hydromedusa, Limnocnida tanganyicae Böhm, 1883 (Hydrozoa, Limnomedusae), is a prominent component in zooplankton (Sarvala et al., 1999; Langenberg et al., 2008). Of the two most common freshwater jellyfish genera, Craspedacusta, has colonized all continents, while Limnocnida is restricted to Asian and African tropics and subtropics (Dumont, 1994a; Jankowski, 2001). In Africa, L. tanganyicae seems to be the only species (Goy, 1977). In Lake Tanganyika, the medusa stage is predominant and in fact, the tiny (< 0.5 mm) hydroid stage of the hydromedusa was discovered later because of its small size and cryptic life style (Bouillon, 1954). The ecology of freshwater jellyfish is poorly known and their taxonomy is still debated. All non-parasitic Cnidaria are predators, but due to the absence of knowledge of their food and feeding, the trophic position of freshwater jellyfish remains obscure (Rayner & Appleton, 1989; Dumont, 1994b). In Lake Tanganyika, jellyfish up to 25 mm diameter are abundant (Kurki et al., 1999), and their biomass is of the same order as that of predatory crustacean zooplankton (Sarvala et al., 1999). The predators of freshwater jellyfish are unknown (Dumont, 1994a), but it has been hypothesized, albeit contradicted by the observations of Viherluoto (1999), that they might be consumed by benthic decapods. There is no evidence that pelagic fish feed on them (Coulter, 1991). Consequently, Limnocnida may be considered as a dead end in the food web. Looking L. tanganyicae on a high resolution epifluorescence microscope during a cruise in 1995?, surprisingly showed the abundance of picocyanobacteria. This led us to hypothesize that these animals might be able to garden picocyanobacteria and partly base their metabolism on that. To better understand the role of possible gardening by L. tanganyicae in Lake Tanganyika, we studied several aspects of its ecology during several expeditions covering the whole lake, utilizing field abundance data, laboratory experiments, as well as genetic and stable isotope analyses. Materials and methods This study was performed during 1994-2001. To estimate jellyfish biomass, a regression was established between the umbrella diameter, dry mass (DM), and ash free dry mass (AFDM) of medusae. Individuals were selected over the range of sizes. After measurement of the umbrella diameters, individuals were dried on pre-weighed aluminium foil cups. These samples were taken to Finland, dried again at 60oC, and weighed on a Cahn Electrobalance. AFDM was obtained by difference from DM after re-weighing following combustion at 500°C. In the pelagic waters off Kigoma, Tanzania (4°51.00’S, 29°35.00’E), quantitative samples of L. tanganyicae medusae were taken at 10 m vertical hauls from 120 m to the surface using a 500-µm mesh closing net. In the vicinity of Mpulungu, Zambia (08°43.98’S, 31°02.43’E), jellyfish bloom samples were taken with a 7-l tube sampler (Limnos Ltd, Finland) and a PVC 10-l collector was used for surface sampling. Qualitative epifluorescence microscopic observations were made of medusae collected throughout the lake. To avoid possible damage to jellyfish from excessive light, individuals for the experiments were collected at dusk when they ascended to the surface. Generally, single large animals were caught by hand-scooping into 0.5 liter beaker from the surface or with a tube water sampler from slightly deeper layers. In 1998, jellyfish also were collected at night or in dim light by divers in < 2 m water depth using 50-mm-diameter acrylic tubes with 300µm-mesh plankton netting covering one end. When a medusa was captured in the tube, the open end was closed with a stopper and immediately taken to the laboratory aboard the research vessel. In some cases, the umbrella diameters of animals were measured with the aid of a dissecting microscope. The presence of protozoans inside the animals also was recorded. The occurrence of 3 4 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 internal algae was checked with an epifluorescence microscope (Nikon Optiphot) at 1250x magnification. Eukaryotic algae were observed with blue excitation and prokaryotic picocyanobacteria with green excitation. To investigate the effect of UV light on jellyfish survival, an experiment was conducted on board the R/V Tanganyika Explorer. Twelve or 13 animals were placed in each of three 2-l quartz bottles, which were put in a water bath in an open, white polystyrene box. One bottle was exposed to direct sunlight, another was kept under a UV-protected polyethylene film, and one wrapped in aluminium foil was used as the dark control. Water temperature was kept similar to ambient by pumping lake water through the box. During the experiment (beginning at 11:45), spectral sunlight radiation was measured every 15 min with a Macam SR 991 spectroradiometer (planar cosine light collector). Spectral penetration of light into water of Lake Tanganyika was also measured using a 4m quartz light cable. Without the UV-protected polyethylene film, measurable radiation was observed down to 300 nm wavelength; with the film, the limit was about 350 nm. At wavelengths longer than 400 nm, the film absorbed ca. 1/3 of the radiation. Medusae pulsing their swimming umbrella were considered alive and were counted every 10 min. In the dark bottle, animals were counted only at the end of the experiment. To avoid even short exposure to high UV radiation, the bottle under the UV-screen was counted only once before the end of the experiment; the bottle was placed in a black cotton bag and transferred to the laboratory of the ship for counting. The experiment was terminated when all animals had died in the bottle kept under direct sunlight. We hypothesized that UV radiation affects L. tanganyicae and therefore they avoid surface water in bright light.Fluorescent beads (Polysciences Inc.) were used to qualitatively study the ingestion of picoplankton-sized organisms. To remove possible bead aggregations, a small volume of stock suspension of beads was filtered through a 5-µm Nuclepore filter. The final concentration of beads in lake water offered to jellyfish in a 50-ml water bottle was adjusted roughly to the same density as picoplankton abundance in lake water (105 cells ml-1). Bacterial composition of jellyfish was studied from 11 large (> 10 mm) individuals sampled during December 2001 from the lake surface (0-2 m) off Kigoma harbor at dusk and then stored in 70% ethanol. Water samples (1 liter) were taken with a Limnos sampler at the same time at 10-m depth intervals from the surface to 60 m. For DNA analyses, 0.5 l of water was screened through 50-µm-mesh plankton netting and then filtered onto Filtropur acetylacetate filter units (0.2- m pore size). Bacterial DNA extractions were performed using the combined enzymatic and bead-beating method, and the length heterogeneity-PCR (LH-PCR) targeted V1-V3 variable regions of the 16S rRNA (area 8-534, Escherichia coli numbering), as described by Tiirola et al. (2002a). Direct sequencing of the heterogeneous PCR products was performed bi-directionally using the ABI BigDye kit and ABI 3100 DNA sequencer (Applied Biosystems). Sequences were compared against the EMBL database using the BLAST algorithm (Altschul et al., 1997). A bootstrapped neighbor-joining tree was calculated using Jukes-Cantor correction with the MEGA 4 software. Reference sequences for inferring the tree were the following (from top to bottom in the tree): AF330249, DQ463712 (Lake Tanganyika clone), AF330250, AF448063, AF216955, AF317074, AF330247, AB015058, AF330252, AF001477, AY151249, AF098370, AF330251, AY172819, AY172811, AY172810, AY172801 AF001479, AY172833, AF245618, S000388727, AF053398, S000628344, and AY946243. To measure oxygen consumption, jellyfish were transferred individually into 50-ml bottles filled with lake water and sealed with round, glass stoppers. Bottles with the same water, but without a medusa, served as controls. After filling, oxygen concentration was measured and the bottles were placed in an incubator. Some of the bottles were darkened with aluminium foil. In the incubator, the bottles were kept under an illumination of 511 µ mol m-2 s-1 supplied by 6 daylight type fluorescent tubes. Water temperature was maintained within 1°C of the lake temperature (around 27°C) by pumping water from the lake through the incubator. Temperature was monitored with a thermometer during the oxygen measurements. Oxygen concentration was measured with an 4 5 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 YSI BOD bottle probe with a stirrer at the upper part of the bottle. It was placed in the bottle and mixing was kept on for 30 s before reading the value. After the measurement, the bottle was restoppered and returned to the incubator. Oxygen consumption by jellyfish was calculated as the difference between the initial and final concentrations taking into account the incubation time and the bottle volume. Differences in control bottles were subtracted from the results of bottles with medusae. In one oxygen production/consumption experiment, lake water was amended with autoclaved stock solutions of KH2PO and NH4Cl to final concentrations of 0.8 µmol P l-1 and 12.5 µ mol N l-1, roughly in accordance with the highest concentrations (phosphate 0.1-0.6 µmol, nitrate 1.6-3.7 µmol) reported by De Wever et al. (2008a) for the epilimnion of Lake Tanganyika. Ammonium nitrogen was used because algae preferentially take up ammonium and jellyfish excrete ammonium. NH4-N is typically very low (0-0.05 g m-3) down to roughly 100 m depth in the lake (Plisnier et al., 1999). We hypothesized that nutrients released from the digestion of invertebrate food of jellyfish are sufficient for endosymbiotic picocyanobacteria so that external nutrients do not affect their photosynthesis. To clarify the trophic position of the jellyfish, samples for stable carbon and nitrogen isotope determinations were collected off Kigoma (4°51’S, 29°35’E), in late November to early December 2001. Jellyfish were sampled on 4-10 December 2001 either with vertical net hauls (100or 250-µm mesh) from 120 m to the surface (or from 50 m after sunset), or by scooping individual jellyfish from surface water. In the latter case, the abundance of associated picoplankton was estimated visually by color, and the jellyfish were sorted into five groups accordingly (colorless = no or very few picoplankton; slight pink hue = few picoplankton; medium pink hue = picoplankton moderately abundant; entire jellyfish intensely pink = picoplankton very abundant overall; and pink color only around the marginal ring of the jellyfish). The correlation between jellyfish color and abundance of internal picoplankton was confirmed with the epifluorescent microscope. Jellyfish were stored in carbon- and nitrogen-free alkaline Lugol’s iodine. Later, medusae were rinsed with deionized water and placed in tin cups as groups of small, similar individuals or as pieces of large individuals (total sample dry mass 1-4 mg). The cups were sealed, dried at 60ºC, and sent for analysis by an Europa Scientific Hydra 20/20 isotope ratio mass spectrometer at the Stable Isotope Facility, University of California-Davis, California, U.S.A.. The results are given using the notation, where = [(Rsample/Rreference) 1] × 1000, expressed in units per thousand (‰), and where R = 13C/12C or 15N/14N. Reference materials were PeeDee belemnite for carbon and atmospheric N2 for nitrogen. Zooplankton, shrimps, and fish larvae from the same net hauls were fixed with carbon- and nitrogen-free alkaline Lugol’s iodine immediately after sampling and later sorted by species and size groups in the laboratory. After rinsing with deionized water, groups of 1 to approximately 3000 individuals were transferred to tin cups, sealed, dried, weighed, and sent for stable isotope analysis. Water was sampled from different depths (0-100 m) with a tube sampler (Limnos Ltd, Finland) from 19 November-10 December 2001. The samples (4-20 l) were pre-screened through 50-µmmesh netting to remove zooplankton and large phytoplankton, and the filtrates were then filtered through pre-combusted (at 500ºC overnight) glass fibre filters (Whatman) using a low vacuum (< 20 kPa), first through a GF/D filter (median pore size 2.8 µm, which retained mainly eukaryotic nanoand microplankton) and then through a GF/F filter (median pore size 0.7 µm, which retained mainly picocyano- and heterotrophic bacteria). Larger phytoplankton (mainly cyanobacteria) was collected on 6-10 December 2001 in net hauls with 50-µ m mesh from 5-10 m depth to the surface, which then were concentrated on GF/D filters. The filters were put on pre-combusted aluminium foil and dried at 60ºC. In Finland, the dried filters were weighed and 16-18 (GF/D) or 10 (GF/F) 3-mmdiameter subsample discs were punctured from the filters and placed in pre-weighed tin cups, weighed, and sent for stable isotope analysis. 5 6 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 The samples of the most important pelagic planktivore, the clupeid fish, Stolothrissa tanganicae Regan, were obtained on 22 and 26 November 2001 directly from the fishermen as they came ashore. The fish were measured for length and a tissue sample of the dorsal white muscle was cut from behind the dorsal fin. The tissue samples were put on aluminium foil and dried in an oven at 60ºC. In Finland, the tissue samples were ground to a fine powder and ca. 0.8 mg from each sample was transferred into pre-weighed tin cups and sent for stable isotope determinations. Linear mixing models were applied to the isotope signatures to quantify the contributions of potential food sources to the diet of the jellyfish (program IsoSource; Phillips & Gregg, 2003). Isotope signatures were adjusted for the stepwise enrichment in the heavier isotopes from one trophic level to the next, using steps of 0.5 and 1‰ for 13C (France & Peters, 1997), and 2 and 3.4 ‰ for 15N (McCutchan et al., 2003). Results Although Limnocnida tanganyicae are so large that they easily attract visual attention, their individual biomasses were low, with AFDM ca. 2 mg for a 12-mm-diameter medusa (Fig. 1.). During several cruises, L. tanganyicae medusae were found at the surface in locally high densities. In early September 1995, we examined their detailed horizontal and vertical distributions by sampling a jellyfish bloom covering nine locations near Mpulungu in the southern basin of Lake Tanganyika (Fig. 2)., Jellyfish density was roughly 3000 ind m-3 nearest to the coast, but it was an order of magnitude lower at more than 15 km offshore. Vertical day and night distributions of L. tanganyicae were studied in April 1998 off Kigoma, when the epilimnion was less than 15 m thick (Fig. 3). Jellyfish were present throughout the oxygenated water column to approximately 100 m deep. Vertical distributions of medusa abundances near noon and midnight both were maximum within the 10-20 m zone; however, at night their median depth of occurrence was 13 m, while at noon it was deeper at 21 m. The difference between day and night vertical distributions was highly significant (Smirnov test, D = 0.219, p < 0.001, n1 = 261, n2 = 367). Small medusae were evenly distributed but large (> 10 mm diameter) ones more frequently occurred in the upper 30 m. Visual observations also showed that, at sunset, jellyfish appeared near the surface when photosynthetically active radiation (PAR) above water reached ca. 200 µmol m-2 s-1 (roughly 20% of the daytime average; Sarvala et al. 1999) and large individuals arrived first. During daytime, medusae were observed only occasionally near the surface. Although they were often alive, in agreement with Dumont (1994), when used in the experiments they did not survive long, suggesting that they were somehow damaged. Because UV radiation can be damaging in clear water lakes, such as Tanganyika, we studied whether UV might explain the absence of or damage to L. tanganyicae . The most harmful UV-B radiation was restricted to the top 5 m of the water column (Fig. 4). In a survival experiment performed near solar noon on the deck of the research vessel, the accumulation of UV radiation developed linearly (Fig. 4). Under a UV-screen, UV-B radiation (290-320 nm) was virtually eliminated. UV-A radiation was detectable at wavelengths of 350-400 nm, but the experiment was too short to cause significant mortality of jellyfish. By contrast, jellyfish exposed to natural solar radiation died within one hour (Fig. 5). Observations with an epifluorescent microscope of living L. tanganyicae without any staining only occasionally showed cells that fluoresced under blue excitation, and only in the gut region. By contrast, under green excitation, the field of view was often full of generally < 1-µmdiameter orange fluorescing cells, which indicated that they contained phycoerythrine pigment of picocyanobacteria. Some of the cells moved freely in the body fluid of the jellyfish and some were stationary. Often it was difficult to judge whether stationary cells were on surface or inside the jellyfish. Sometimes picocyanobacteria were in different types of colonies (Fig. 6). Generally, the cells were globular or very short rods. When the colonies were felt-like, the picocyanobacteria were rods arranged in filaments and the jellyfish had a pink color visible at distance in the lake. Although 6 7 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 picocyanobacteria were nearly always present in jellyfish caught during all expeditions, their abundance differed remarkably among individuals. A sample collected by net from the entire 100-m water column off Kigoma contained about 20% of 123 medusae with so many picocyanobacteria that their enumeration at 1250x magnification was impossible; however, 56% of the medusae had only scattered single cells or small groups of picocyanobacteria. In addition to picocyanobacteria, around the gut of 21% of 705 jellyfish we found wheellike Trichodina type ciliates, similar to those found for Craspedacusta sowerbyi (Lankester) by Green (1998). When fluorescent beads 0.2-2 µm-diameter were offered, ciliates ingested large numbers of all bead sizes, indicating that they can consume both bacteria and picocyanobacteria; however, we found no uptake of beads by jellyfish. Dozens of individually-caught jellyfish observed immediately by microscopy never had zooplankton in their guts. In contrast, jellyfish inspected immediately after collection by plankton net had immobilized copepods in the corners at the bottom of the gut, which we believe were caught in the concentrated sample. LH-PCR analysis of the 16S rRNA genes showed that a single LH-PCR fragment size (470 bp) was present in all jellyfish samples (Fig. 7). This fragment was highly dominant, covering > 6090% of the total fragment intensity in 7 out of the 11 studied jellyfish. Picocyanobacterial biomarker (470 bp) of jellyfish also was very abundant in the epilimnion of Lake Tanganyika, contributing up to 55% of the bacterial abundance at 0-10 m sampling depth (Fig. 8). Another fragment size (472 bp), probably belonging to another picocyanobacterial group, was > 15% of the bacterial abundance in the water (0-50 m), but was completely absent from jellyfish. This indicates that the microbial diversity of the jellyfish did not mirror the diversity of the environment. Five of the 16S rRNA gene PCR products of the jellyfish (numbers 1-5) were subjected to bi-directional sequencing, which revealed that the consensus sequences were 100% identical (387 bp overlap). The BLAST search of all available nucleotide sequences revealed that this sequence was identical to the cyanobacterial sequence TK-SE6 (EMBL accession number DQ463712, length 1432 bp) dominating the oxic epilimnion of Lake Tanganyika (De Wever et al. 2008b) and to sequences obtained from the Chinese MiYu reservoir water (GU305743) and freshwater lakes (GU323646, GU323613, and GU323608). Similar sequences (99% identity) were obtained from Synechococcus sp. strains isolated from tropical and boreal freshwater environments, such as lakes Biwa (strain BS2, HM346183), Taihu (strain LBG2, AF330249), and Tuusulanjärvi (strain 0tu30s01, AM259220), as well as from marine environments (strain CCMP839, AY946244; strain KORDI-28, FJ497720). In the cluster analysis, we used the longer Lake Tanganyika sequence (DQ463712) to reveal the detailed phylogenetic affiliation of the endosymbiont in comparison to well-established picocyanobacterial clusters (Fig. 9). Lake Tanganyika type sequence clustered with the Lake Biwa strains, which form the freshwater picocyanobacteria group E designated by Crosbie et al. (2003). This cluster contains strains recently re-classified from Synechococcus to Cyanobium, and their closest described species is Cyanobium gracile. One of the other jellyfish (#10) had a high dominance of the LH-PCR peak 522 bp. The sequence of this peak indicated the dominance of betaproteobacterial heterotroph closest to the type strain of Vogesella indigofera (strain ATCC 19706T, AB021385, 95% identity). The potential role of picocyanobacteria in the metabolism of the jellyfish-picocyanobacteria community was examined by experiments. In the first light incubation experiment, jellyfish > 10 mm in diameter sometimes showed net oxygen production (Fig. 10). In subsequent experiments, we studied how oxygen consumption/production would develop when the same jellyfish were kept sequentially in light and dark. The results corroborated our initial finding, specifically, that under light, oxygen consumption decreased or even switched to production (Fig. 11). The same trend was observed in a similar experiment, although the jellyfish generally were net oxygen consumers (Fig. 12). When inorganic nutrients were added late in the light phase of the experiment, however, three of four individuals had remarkably increased oxygen production. Finally, after the light was 7 8 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 switched off, no jellyfish produced oxygen. These results suggested high variation among individuals, both in consumption and production of oxygen by jellyfish. Oxygen consumption increased in relation to the diameter of jellyfish in the dark (Fig. 13); however, in light there was no clear trend, and oxygen production rates in the smallest size classes were surprisingly high. The stable carbon isotope signatures of L. tanganyicae were similar to those of crustacean zooplankton and fish larvae, but lower than in big shrimps or in the planktivorous clupeid fish Stolothrissa tanganicae, slightly higher than in pico-, nano-, and microphytoplankton, or copepod nauplii, and much higher than in small shrimps or in the net phytoplankton mainly consisting of cyanobacteria (Fig. 14). The nitrogen signatures of L. tanganyicae were similar to those of adult Stolothrissa tanganicae, but higher than those of all other groups except big shrimps. Neither the carbon nor nitrogen isotope signatures of L. tanganyicae were significantly affected by the abundance of internal picocyanobacteria (ANOVA, P = 0.34 for carbon, P = 0.90 for nitrogen, n = 11). Isotope mixing models produced broad distributions of calculated diet proportions, which were sensitive to the choice and grouping of potential food sources, as well as to the trophic step adjustment. The diffuse results were consistent with the fact that the isotope polygon defined by the potential food sources was rather narrow (Fig. 14). No feasible solutions were obtained with a nitrogen step of 3.4‰; however, use of steps of 0.5‰ for carbon and 2.0‰ for nitrogen, achieved isotope mass balance for numerous combinations. When all major groups except adult and larval fish were included as potential food sources for L. tanganyicae , the calculated contributions of all sources remained variable and low (all included zero contributions); the 1-99th percentiles varied from 0-18 to 0-46%, with picocyanobacteria, Tropocyclops tenellus (Sars), Tropodiaptomus simplex (Sars), and copepod nauplii showing slightly higher maximum values than other groups. Net phytoplankton, nano-, and microplankton, as well as small shrimps, always showed low contributions. The results remained qualitatively similar if cyclopoid or all copepod adults and copepodids were one group, but the importance of copepods and picocyanobacteria then became more pronounced (1-99th percentiles 0-58 and 0-46%, respectively). In such cases, the big shrimps appeared as a moderately important diet component (1-99th percentiles 14-38%), provided that phytoplankton with low carbon and nitrogen signatures was simultaneously consumed at a higher extent. Similarly, mixtures of several food sources also could include fish larvae. If copepod nauplii were grouped with adults and copepodids, no feasible solutions were found. The most clear-cut model solution was obtained when all phytoplankton groups, grouped cyclopoid adults and copepodids, Tropodiaptomus, copepod nauplii, and small shrimps were included as potential food sources. The highest contributions then were shown by cyclopoids, Tropodiaptomus, copepod nauplii, and picocyanobacteria (1-99th percentiles 2-62, 24-56, 0-40, and 0-30%, respectively). Discussion Our results verified that Limnocnida tanganyicae medusae have very low dry and ash-free biomass relative to their diameter, and the regression between diameter and dry mass was similar to that for Craspedacusta sowerbii (Jankowski, 2000). Jellyfish are an interesting product of evolution which, during the 600 million years of their existence, may have expanded their prey catching machinery as large as possible with minimum material cost. Kurki et al. (1999) sampled monthly for two years at three sites and found typical abundances of L. tanganyicae between 10 and 100 ind m-3 in the upper 100 m, with the highest abundances (> 500 ind m-3) in the northern part of Lake Tanganyika. By contrast, four lake-wide cruises indicated highest abundances in the southern basin (Kurki, 1998). In our study, the observed maximal abundances were up to 5 to 6 times higher than those observed by Kurki et al. (1999), and even higher concentrations were recorded in December 1994 in surface waters at Kigoma (personal observations). It seems clear that high abundances of jellyfish are temporally and spatially variable (Kurki et al., 1999; Langenberg et al., 2008). The factors behind high density blooms are poorly 8 9 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 known. The large bloom in this study appeared in the southern basin at the end of the windy upwelling season (Coulter & Spigel, 1991), during a transition from a 3-month period of increased availability of nutrients and primary production in cool, well-mixed waters to warmer water, increased epilimnetic stratification, and oligotrophy (Langenberg et al., 2003). Although the blooms generally seem regionally limited, this bloom covered a large area of ca. 400 km2. The jellyfish blooms interfere with the fishery because nearby fishing communities in Zambia and Tanzania must stop fishing for the duration of the bloom (I. Kimirei, Tanzanian Fisheries Research Institute, personal communication). UV-radiation, particularly UV-B, is harmful to zooplankton (e.g., Williamson, 1995; Hylander & Hansson, 2010). Vulnerability to radiation differs among different aquatic organisms (Rhode et al., 2001) and environments, but there is no information about UV-B radiation effects on freshwater jellyfish. Although L. tanganyicae was present throughout the whole water column, their abundance was clearly highest in the epilimnion. Similar to marine jellyfish (Schuyler & Sullivan, 1997), L. tanganyicae actively avoided water layers with high illumination. This avoidance, combined with the results of our light incubation experiment (Fig. 5) and the fact that animals found near the surface at daytime did not survive long, suggests that daytime UV-B radiation is harmful to these highly transparent L. tanganyicae jellyfish. UV-B radiation started to cause mortality of copepods at doses of ca. 0.1 W m-2 (Zagarese et al., 1997), which suggests that in the upper 3-5 m of the water column, UV-B is excessive for zooplankton in Lake Tanganyika. In fact, at high radiation levels prevailing around noon at the surface, our experimental results show clear harmful effects of UV-B radiation on L. tanganyicae jellyfish (Fig. 5).Thus, lower exposures over longer periods of time are needed for more realistic estimates of the effect. It seems probable that vertical migration of L. tanganyicae reflects avoidance of UV radiation in the upper epilimnion, although other factors may be involved as well. The LH-PCR results indicated that bacterial groups in jellyfish did not match the diversity in the surrounding water. Instead, individual jellyfish had their own bacterial communities more or less dominated by a certain biomarker size, which also was at a maximum in the water column at 10 m depth. Because several cyanobacterial and alphaproteobacterial genera have the same LH-PCR size (Tiirola et al., 2003), further sequencing was needed to identify the microbes represented by the biomarker. All the sequenced jellyfish carried certain Cyanobium-type picocyanobacteria. In other than pathogenic situations (e.g., Rantakokko-Jalava et al., 2000; Tiirola et al., 2002b), it is unusual for a PCR fragment from environmental samples amplified by universal bacterial primers to be directly sequenced without a cloning step. This is usually possible only if some rRNA gene template contributes to > 60-70% of all templates (Tiirola et al., unpublished). The prevailing 16S rRNA sequence of the jellyfish Cyanobium belonged to the same taxonomic unit that predominates in Lake Tanganyika (De Wever et al., 2008b). The characteristic Cyanobium sp. sequence belonged to the Lake Biwa-type freshwater cluster, but it also was closely related to sequences obtained from both freshwater and marine samples ranging from tropical to boreal environments, showing the cosmopolitan nature of the cluster (see Crosbie et al., 2003). Among marine invertebrates, including jellyfish (e.g., Hofmann & Kremer, 1981; Hamner et al., 1982; Muscatine & Marian, 1982; Kremer et al., 1990), endosymbiotic algae are rather common including algae from at least five different classes and animal hosts ranging from protozoans to tunicates (e.g., Trench, 1993). By contrast, endosymbiotic cyanobacteria seem to be less common. Nevertheless, they have been found in invertebrates other than L. tanganyicae medusae. Erwin & Thacker (2008) found Synechococcus-type cyanobacteria in marine poriferans, while Lesser et al. (2004) found endosymbiotic cyanobacteria to coexist with the symbiotic dinoflagellates (zooxanthellae) in a coral and to express the nitrogen-fixing enzyme, nitrogenase. The presence of this prokaryotic symbiont in a nitrogen-limited zooxanthellate coral suggests that nitrogen fixation may be an important source of that limiting element for the symbiotic association. 9 10 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 Jellyfish capture photosynthetic organisms from water (Rumpho et al., 2011) and form a symbiotic association with them. Carbohydrates and low molecular weight lipids are produced for the host, which probably return nutrients to the microorganisms. Thus, each generation has to acquire its photosynthetic partner from the surrounding environment. The picocyanobacteria that are predominant in L. tanganyicae medusae, are very abundant in Lake Tanganyika water (Vuorio et al., 2003; De Wever et al., 2008b; Stenuite et al., 2009). Stenuite et al. (2009) estimated that autotrophic picoplankton with ‘Synechococcus-type pigment’ accounted for 41-99% of the total phytoplankton biomass. Boulenger (1911) suggested that the typical function of the stomach, which is reduced in L. tanganyicae , is accomplished by the canal system, and that the medusae live upon unicellular algae and protozoa driven into the radial canals. Although we saw picocyanobacteria flowing in the canals, ingestion of pico-sized particles was not confirmed by uptake of fluorescent beads, which suggests an endosymbiotic association of the microbiota inside the jellyfish. A well-known example of jellyfish with endosymbiotic algae is the scyphozoanMastigias sp., which relies up to 100% on photosynthesis of endosymbiotic algae and could contribute substantially (16%) to primary production of Eil Malk Jellyfish Lake (McCloskey et al., 1994). Mastigias sp. medusae even orient their umbrella to obtain optimum illumination (Hamner et al., 1982). The vertical migration of L. tanganyicae medusae also could provide improved growth conditions for its picocyanobacteria in Lake Tanganyika. If the medusae adjust their vertical position according to the optimum illumination, they could maximize photosynthesis of their endosymbionts. This possibility seems plausible because jellyfish arrive at the surface at dusk too early when light intensity is too high to capture vertically migrating zooplankton. Previous sampling was too coarse to reveal layers of jellyfish at 100-200 µmol m-2 s-2 PAR illumination throughout the day, but the daytime vertical distribution of large individuals supports this speculation. In low-nutrient environments, heterotrophic-autotrophic associations of organisms offer a competitive advantage through the mutual transfer of otherwise limiting resources (Yellowlees et al., 2008). The results of our oxygen consumption/production experiment with jellyfish and additional nutrients suggest that Lake Tanganyika jellyfish can have net nutrient uptake similar to that of other jellyfish (Muscatine & Marian, 1982; Pitt et al., 2005; Todd et al., 2006). Thus, L. tanganyicae jellyfish may acquire key nutrients both from inorganic pool of water and from their prey; however, if they have a high phosphorus level, similar to Craspedacusta sowerbii (Jankowski, 2000), then while fulfilling their phosphorus requirements, the medusae obtain excess nitrogen which has to be excreted. Because many photosynthetic organisms can store nutrients, internal picocyanobacteria might be able to effectively harvest nutrients released by jellyfish so that the organisms form a closed semiautonomous micro-ecosystem where nutrients (at least nitrogen) are recycled internally (Pitt et al., 2009), as was supported by our stable isotope results. The nitrogen signature of the jellyfish was relatively high and independent of the abundance of internal picocyanobacteria. Because the concurrent free-living picocyanobacteria showed low nitrogen signatures indicative of nitrogen fixation (Vuorio et al., 2006) and thus a shortage of inorganic nitrogen in the environment, the internal picocyanobacteria most likely obtained their nitrogen from L. tanganyicae jellyfish. The oxygen production of endosymbiotic algae may be so high that it exceeds the oxygen consumption of its host (Kremer et al., 1990). Factors affecting their oxygen production are similar to those of free-living algae (Muscatine & Marian, 1982). Contrary to our results showing high variation in L. tanganyicae (Fig. 13), in Linuche uniquiculata medusae (Kremer et al., 1990), oxygen production depended linearly on the size of the medusa, on light intensity, and on the behavior of jellyfish containing endosymbiotic algae (Muscatine & Marian, 1982). Although we found that L. tanganyicae jellyfish sometimes can be net producers, the high variation of the abundance of associated picocyanobacteria complicates understanding their role in jellyfish 10 11 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 metabolism, as well as in the open water ecosystem. The observed picocyanobacteria abundance in the medusae may vary according to their nutritional status. When the status is poor, jellyfish may deplete the picocyanobacteria, while in the opposite case, picocyanobacteria flourish. Thus, both feeding and life histories may affect the picocyanobacteria abundance. Variation in picocyanobacteria abundance was high even among large animals collected from the surface. In Eil Malk Jellyfish Lake, Mastigias sp. medusae maximized their exposure to inorganic nutrients by swimming 15 m to the chemocline at night (Muscatine & Marian, 1982). It might be possible that such migration to the darker part of the water column also contributes to variation in the abundance of picocyanobacteria. Although the distance to the chemocline in Lake Tanganyika is roughly 100 m, L. tanganyicae may harvest nutrients there. The results of our nutrient addition experiment suggest that nutrients from digested prey are not always sufficient. If so, photosynthesis by picocyanobacteria may not be a steady source of food but photosynthesis by picocyanobacteria would extend the resources obtained from prey over a longer duration. As for algae in Mastigias sp. jellyfish (Muscatine et al., 1986), there is no direct evidence of digestion of picocyanobacteria by L. tanganyicae jellyfish. They may obtain photosynthetic products directly through the cell walls of the picocyanobacteria as in the coral-zooxanthellae relationship (Lesser et al., 2004; Woolridge, 2010). The stable isotope results helped to clarify the trophic position of L. tanganyicae in the Lake Tanganyika food web. The isotope composition of an organism reflects its diet, with a stepwise enrichment in the heavier isotopes from one trophic level to the next. Enrichment in 13C is small (0 –1‰), permitting identification of carbon sources (France & Peters, 1997), while the enrichment in 15N is higher (2–4 ‰; empirical average 3.4‰), allowing estimation of the trophic position of consumers (McCutchan et al., 2003). In the pelagic food web of Lake Tanganyika, the 15 N enrichment per trophic step seems to be rather low, around 2‰ (Sarvala et al., 2003), as is common for ammonotelic freshwater organisms (Vanderklift & Ponsard, 2003). The similarity of carbon signatures and the mean difference of 1.8‰ in nitrogen signatures between the jellyfish and the copepodid and adult copepods are consistent with feeding on mixed copepods by the jellyfish, which was supported by the results of the mixing model application. The mixing model results indicated that jellyfish also might feed on picocyanobacteria, but that feeding on small shrimps was unlikely. Although feeding on big shrimps in combination with low-signature phytoplankton could balance the isotope equilibrium, the big shrimps are too scarce in the lake (Bosma et al., 1998) to be a realistic food source for the jellyfish. Thus, the mixing model can be used only to quantify the contributions of known food items in the diet, but not used to identify the true diet components. In general, food items with strong isotopic signatures in opposite directions, they appear important in the calculated mixture whether or not they are actually eaten and may conceal the importance of items with moderate signatures. It was suggested that L. tanganyicae fed on fish eggs (Dumont, 1994b), but on the basis of stable isotope signatures, it is unlikely that fish formed substantial portions of the diet. The nitrogen signature of the jellyfish was only 1.3‰ higher than that of fish larvae and did not differ from the signature in adult Stolothrissa (Fig. 14). The nitrogen signature of fish eggs is normally very close to that of the female body; thus, jellyfish cannot have been feeding on fish eggs to any significant extent. The mixing model did show that some isotopically-feasible diet combinations could include fish larvae. During five lake-wide cruises in Lake Tanganyika in 1995-1998, abundances of fish larvae and eggs were one or two magnitudes lower than jellyfish, shrimps, and copepods (Bosma et al., 1998); therefore, ichthyoplankton probably would have been only occasional prey of the jellyfish that could not be detected by the isotopic signatures. Jellyfish can be powerful modifiers of the zooplankton community. In River Yamuna, India Limnocnida indica medusae removed cladocerans (Moina sp.), as well as the rotifer Keratella sp. from zooplankton; however, some rotifers (Asplanchna sp. and Brachionus sp.) and some copepods were not affected (Sharma & Chakrabarti, 2000). Craspedacusta sowerbii medusae consume 11 12 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 zooplankton up to 2 mm (Dodson & Cooper, 1983; Jankowski et al., 2005; Smith & Alexander Jr., 2008; Stefani et al., 2010). Thus, L. tanganyicae jellyfish also may exert strong influence on zooplankton particularly during blooms in Lake Tanganyika. More information on the feeding of L. tanganyicae on different components of zooplankton is needed to better understand its trophic role in Lake Tanganyika. Limnocnida tanganyicae medusae are a prominent component in the open water ecosystem of Lake Tanganyika with many metabolic and behavioral adaptations. Primarily, the medusae are likely predators of zooplankton and probably also can garden photosynthetic picocyanobacteria, which enables their survival in the oligotrophic conditions of Lake Tanganyika. This study provides novel insights into the ecology of freshwater jellyfish. Acknowledgments This study was a part of the FAO/FINNIDA Lake Tanganyika Research Project GCP/RAF/271/FIN “Research for the Management of the Fisheries on Lake Tanganyika (LTR)”. Additional funding was received from the Academy of Finland (grants 44130, 52271 and 201414), the University of Turku Foundation and the Universities of Turku, Kuopio and Jyväskylä, Finland. We also thank Jitka Jezberova and David Fewer for commenting the picocyanobacterial systematics, and the reviewers and editors for their assistance. 12 13 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 References Altschul, S. F., T. L. Madden, A. A. Schaffer, J. Zhang, Z. Zhang, W. Miller & D. J. Lipman, 1997. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Research 25: 3389-3402. Bosma, E. M., S. Muhoza & I. Zulu, 1998. The gulf net sample results of five cruises with the R/V Tanganyika Explorer. FAO/FINNIDA Research for the Management of the Fisheries of Lake Tanganyika. GCP/RAF/271/FIN—TD/81 (En): 24p. Bouillon, J., 1954. A hydropolyp in the biological cycle of a freshwater jellyfish. Nature 174: 1112. Boulenger, C. L., 1911. On some points in the anatomy and bud-formation of Limnocnida tanganicæ. Quarterly Journal of Microscopic Sciences 57: 83-106. Cohen, A. S., M. J. Soreghan & C. A. Scholz, 1993. Estimating the age of formation of lakes: An example from Lake Tanganyika, East African Rift system. Geology 21: 511-514. Coulter, G. W. & R. H. Spigel, 1991. Hydrodynamics. In Coulter, G. W. (ed.), Lake Tanganyika and its Life. Oxford University Press, London: 49–75. Coulter, G. W., 1991. Lake Tanganyika and its Life. British Museum, Natural History & Oxford University Press, London, 354 pp. Coulter, G. W., 1994. Lake Tanganyika. In Martens, K., B. Goddeeris & G. Coulter (eds), Speciation in Ancient Lakes. Advances in Limnology 44: 13-18. Crosbie, N. D., M. Pockl & T. Weisse, 2003. Dispersal and phylogenetic diversity of nonmarine picocyanobacteria, inferred from 16S rRNA gene and cpcBA-intergenic spacer sequence analyses. Applied and Environmental Microbiology 69: 5716-5721. De Wever, A., K. Muylaert, D. Langlet, L. Alleman, J.-P. Descy, L. Andre, C. Cocquyt & W. Vyverman, 2008a. Differential response of phytoplankton to additions of nitrogen, phosphorus and iron in Lake Tanganyika. Freshwater Biology 53: 264–277. De Wever, A., K.Van der Gucht, K. Muylaert, S. Cousin & W. Vyverman, 2008b. Clone library analysis reveals an unusual composition and strong habitat partitioning of pelagic bacterial communities in Lake Tanganyika. Aquatic Microbial Ecology 50: 113-122. Dodson, S. I. & S. D. Cooper, 1983. Trophic relationships of the freshwater jellyfish Craspedacusta sowerbyi Lankester 1880. Limnology and Oceanography 28: 345-351. Dumont, H. J., 1994a. The distribution and ecology of fresh- and brackish-water medusae of the world. Hydrobiologia 272: 1-12. Dumont, H. J., 1994b. Ancient lakes have simplified pelagic food webs. Archiv für Hydrobiologie Beihefte Ergebnisse der Limnologie 44: 223–234. Ernst, A., S. Becker, U. I. A. Wollenzien & C. Postius, 2003. Ecosystem-dependent adaptive radiations of picocyanobacteria inferred from 16S rRNA and ITS-1 sequence analysis. Microbiology 149: 217-228. Erwin, P. M. & R. W. Thacker, 2008. Phototrophic nutrition and symbiont diversity of two Caribbean sponge–cyanobacteria symbioses. Marine Ecology Progress Series 362: 139-147. France, R. L. & R. H. Peters, 1997. Ecosystem differences in the trophic enrichment of 13C in aquatic food webs. Canadian Journal of Fisheries and Aquatic Science 54: 1255-1258. Goy, J., 1977. Sur les Limnocnida africaines (Cnidaires: Limnoméduses). Bulletin de l’Institut francais d’Afrique Noire, série A 39: 563-582. Green, J., 1998. Plankton associated with medusae of the freshwater jellyfish Craspedacusta sowerbyi (Lankester) in a Thames backwater. Freshwater Forum 11: 69-76. Hamner, W. H., R. W. Gilmer & P. P. Hamner, 1982. The physical, chemical and biological characteristics of a stratified, saline sulfide lake in Palau. Limnology and Oceanography 27: 896-909. Hofmann, D. K. & B. P. Kremer, 1981. Carbon metabolism and stobilation of Cassiopea andromeda (Cnidaria: Scyphozoa): Significance of endosymbiotic dinoflagellates. Marine Biology 65: 25-33. 13 14 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 Hylander, S. & L.-A. Hansson, 2010. Vertical migration mitigates UV effects on zooplankton community composition. Journal of Plankton Research 32: 971–980. Jankowski, T., 2000. Chemical composition and biomass parameters of a population of Craspedacusta sowerbii Lank 1880 (Cnidaria: Limnomedusa). Journal of Plankton Research 22: 1329–1340. Jankowski, T. & H. T. Ratte, 2001. On the influence of the freshwater jellyfish Craspedacusta sowerbii on the zooplankton community. Verhandlungen Internationale Vereinigung für theoretische und angewandte Limnologie 27: 3287–3290. Jankowski, T., T. Strauss & H. T. Ratte, 2005. Trophic interactions of the freshwater jellyfish Craspedacusta sowerbii. Journal of Plankton Research 27: 811–823. Kremer, P., J. Costello, J. Kremer & M. Canino, 1990. Significance of photosynthetic endosymbionts to the carbon budget of the scyphomedusa Linuche unquiculata. Limnology and Oceanography 35: 609-624. Kurki, H., 1998. Results of plankton net and torpedo sampling during cruises on board R/V Tanganyika Explorer. FAO/FINNIDA Research for the management of the fisheries on Lake Tanganyika, GCP/RAF/271/FIN-TD/85 (En), 33 pp. Kurki, H., P. Mannini, I. Vuorinen, E. Aro, H. Mölsä & O. V. Lindqvist, 1999. Macrozooplankton communities in Lake Tanganyika indicate food chain differences between the northern part and the main basins. Hydrobiologia 407: 123-129. Langenberg, V. T., L. Mwape, K. Thsibangu, J. M. Tumba, A. A. Koelmans, R. Roijackers, K. Salonen, J. Sarvala & H. Mölsä, 2002. Comparison of thermal stratification, light attenuation, and chlorophyll-a dynamic between the ends of Lake Tanganyika. Aquatic Ecosystem Health & Management 5: 255-265. Langenberg, V. T., J. Sarvala & R. Roijackers, 2003. Effect of wind induced water movements on nutrients, chlorophyll-a, and primary production in Lake Tanganyika. Aquatic Ecosystem Health & Management 6: 279–288. Langenberg, V. T., J.-M. Tumba, K. Tshibangu, C. Lukwesa, D. Chitamwebwa, D. Bwebwa, L. Makasa & R. Roijackers, 2008. Heterogeneity in physical, chemical and plankton-community structures in Lake Tanganyika. Aquatic Ecosystem Health & Management 11: 16-28. Lesser, M. P., C. H. Mazel, M. Y. Gorbunov & P. Falkowski, 2004. Discovery of symbiotic nitrogen-fixing cyanobacteria in corals. Science 305: 997-1000. DOI: 10.1126/science.1099128 McCloskey, L. R., L. Muscatine & F. P. Wilkerson, 1994. Daily photosynthesis, respiration, and carbon budgets in a tropical marine jellyfish (Mastigias sp.) Marine Biology 119: 13-22. McCutchan, J. H., W. M. Lewis Jr., C. Kendall & C. C. McGarth, 2003. Variation in trophic shift for stable isotope ratios of carbon, nitrogen, and sulfur. Oikos 102: 378-390. Muscatine, L. & R. E. Marian, 1982. Dissolved inorganic nitrogen flux in symbiotic and nonsymbiotic medusae. Limnology and Oceanography 27: 910-917. Muscatine, L., F. P. Wilkerson & L. R. McCloskey, 1986. Regulation of population density of symbiotic algae in a tropical marine jellyfish (Mastigias sp.). Marine Ecology Progress Series 32: 279-290. Phillips, D. L. & J. W. Gregg, 2003. Source partitioning using stable isotopes: coping with too many sources. Oecologia 136: 261-269. Pitt, K. A., K. Koop & D. Rissik, 2005. Contrasting contributions to inorganic nutrient recycling by the co-occurring jellyfishes, Catostylus mosaicus and Phyllorhiza punctata (Scyphozoa, Rhizostomeae). Journal of Experimental Marine Biology and Ecology 315: 71-86. Pitt, K. A., D. T. Welsh & R. H. Condon, 2009. Influence of jellyfish blooms on carbon, nitrogen and phosphorus cycling and plankton production. Hydrobiologia 616: 133–149. 14 15 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 Plisnier, P. D., D. Chitamwebwa, L. M. Mwape, K. Tshibangu, V. T. Langenberg & E. Coenen, 1999. Limnological annual cycle inferred from physico-chemical fluctuations at three stations of Lake Tanganyika. Hydrobiologia 407: 45–58. Rantakokko-Jalava, K., S. Nikkari, J. Jalava, E. Eerola, M. Skurnik, O. Meurman, O. Ruuskanen, A. Alanen, E. Kotilainen, P. Toivanen & P. Kotilainen, 2000. Direct amplification of rRNA genes in diagnosis of bacterial infections. Journal of Clinical Microbiology 38: 32-39. Rayner, N. A. & C. C. Appleton, 1989. Occurrence of introduced Craspedacusta sowerbii and indigenous Limnocnida tanganjicae (Cnidaria: Limnomedusae) in Southern Africa. Environmental Conservation 16: 267-270. Rhode, S. C., M. Pawlowski & R. Tollrian, 2001. The impact of ultraviolet radiation on the vertical distribution of zooplankton of the genus Daphnia. Nature 412: 69-72. Rumpho, M. E., K. N. Pelletreau, A. Moustafa & D. Bhattacharya, 2011. The making of a photosynthetic animal. Journal of Experimental Biology 214: 303-311. Sarvala, J., K. Salonen, M. Järvinen, E. Aro, T. Huttula, P. Kotilainen, H. Kurki, V. Langenberg, P. Mannini, P.-D. Plisnier, I. Vuorinen, H. Mölsä & O. V. Lindqvist, 1999. Trophic structure of Lake Tanganyika: carbon flows in the pelagic food web. Hydrobiologia 407: 149-173. Sarvala, J., S. Badende, D. Chitamwebwa, P. Juvonen, L. Mwape, H. Mölsä, N. Mulimbwa, K. Salonen, M. Tarvainen & K. Vuorio, 2003. Size-fractionated 15N and 13C isotope ratios elucidate the role of the microbial food web in the pelagial of Lake Tanganyika. Aquatic Ecosystem Health & Management 6: 241–250. Schuyler, Q. & B. K. Sullivan, 1997. Light responses and diel migration of the scyphomedusa Chrysaora quinquecirrha in mesocosms. Journal of Plankton Research 19: 1417-1428. Smith, A. S. & J. E. Alexander Jr., 2008. Potential effects of the freshwater jellyfish Craspedacusta sowerbii on zooplankton community abundance. Journal of Plankton Research 30: 1323– 1327. Sharma, J. G. & R. Chakrabarti, 2000. Seasonal occurrence of freshwater medusa Limnocnida indica annandale (Cnidaria: Limnomedusae) in a lake associated with the river Yamuna, India. Aquatic Ecology 34: 205–207. Stefani, F., B. Leoni, A. Marieni & L. Garibaldi, 2010. A new record of Craspedacusta sowerbii, Lankester 1880 (Cnidaria, Limnomedusae) in Northern Italy. Journal of Limnology 69: 189192. Stenuite, S., A.-L. Tarbe1, H. Sarmento, F. Unrein, S. Pirlot, D. Sinyinza, S. Thill, M. Lecomte, B. Leporcq, J. M. Gasol & J.-P. Descy, 2009. Photosynthetic picoplankton in Lake Tanganyika: biomass distribution patterns with depth, season and basin. Journal of Plankton Research 31: 1531-1544. Tiercelin, J. & A. Mondeguer, 1990. The geology of the Tanganyika Trough. In Coulter, G. (ed.), Lake Tanganyika and its Life . Oxford University Press, London: 7-48. Tiirola, M. A., M. K. Männistö, J. A. Puhakka & M. S. Kulomaa, 2002a. Isolation and characterization of Novosphingobium sp. MT1, a dominant polychlorophenol degrading strain in a groundwater bioremediation system. Applied Environmental Microbiology 68: 173-180. Tiirola, M., E. T. Valtonen, P. Rintamäki-Kinnunen & M. Kulomaa, 2002b. Diagnosis of flavobacteriosis by direct amplification of rRNA genes. Diseases of Aquatic Organisms 51: 93-100. Tiirola, M. A., J. E. Suvilampi, M. S. Kulomaa & J. A. Rintala, 2003. Microbial diversity in a thermophilic aerobic biofilm process: analysis by length heterogeneity PCR (LH-PCR). Water Research 37: 2259–2268. Todd, B. D., D. J. Thornhill & W. K. Fitt, 2006. Patterns of inorganic phosphate uptake in Cassiopea xamachana: A bioindicator species. Marine Pollution Bulletin 52: 515–521. 15 16 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 Trench, R. K., 1993. Microalgal-invertebrate symbiosis: a review. Endocytosis Cell Research 9: 135-175. Vanderklift, M. A. & S. Ponsard, 2003. Sources of variation in consumer-diet 15N enrichment: a meta-analysis. Oecologia 136: 169-182. Viherluoto, M., 1999. The food utilisation and diel feeding pattern of shrimps (Atyidea and Palaemonidea) in Lake Tanganyika. In Mölsä, H,. K. Salonen & J. Sarvala (eds), Results of the LTR’s 20th Multi-Disciplinary Cruise. FAO/FINNIDA Research for the Management of the Fisheries of Lake Tanganyika. GCP/RAF/271/FIN-TD/93 (En) 96 pp. Vuorio, K., M. Meili & J., Sarvala, 2006. Taxon-specific variation in the stable isotopic signatures 13 C and 15N) of lake phytoplankton. Freshwater Biology 51: 807-822. Vuorio, K., M. Nuottajärvi, K. Salonen & J. Sarvala, 2003. Spatial distribution of phytoplankton and picocyanobacteria in Lake Tanganyika, in March-April 1998. Aquatic Ecosystem Health & Management 6: 268-278. Williamson, C. E., 1995. What role does UV-B radiation play in freshwater ecosystems? Limnology and Oceanography 40: 386-392. Woolridge, S. A., 2010. Is the coral-algae symbiosis really ‘mutually beneficial’ for the partners? Bioessays 32: 615–625. Yellowlees, D., T. A. V. Rees & W. Leggat, 2008. Metabolic interactions between algal symbionts and invertebrate hosts. Plant Cell & Environment 31: 679–694. Zagarese, H. E., Feldman & C. E.Williamson, 1997. UV-B-induced damage and photoreactivation in three species of Boeckella (Copepoda, Calanoida). Journal of Plankton Research19: 357367. 16 17 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 Figure Captions Fig. 1 Correlation of dry and ash free dry mass of Limnocnida tanganyicae medusae with umbrella diameter. Dotted line for dry mass of Craspedacusta sowerbii medusae (Jankowski, 2000) is shown for comparison Fig. 2 Abundance of Limnocnida tanganyicae medusae in the southern basin of Lake Tanganyika near the town Mpulungu, Zambia 2-4 September 1995. Numbers 1-9 represent sampling stations Fig. 3 Abundances (left side lines; day = grey line, night = black line) and mean umbrella diameters (day = white bars; night = grey bars) with ranges (horizontal lines) of Limnocnida tanganyicae medusae in the epi- and metalimnion of Lake Tanganyika off Kigoma Harbour during 7 April 1998. The temperature vs. depth curve is shown at the right Fig. 4 Penetration of UV-B radiation into water of Lake Tanganyika (April 1998) and cumulative UV-B radiation during the UV-exposure experiment (Fig. 5) on ship board in January 2001 Fig. 5 Survival of >10-mm-diameter Limnocnida tanganyicae medusae collected off Kigoma Harbor and kept in darkened, UV-film-covered, and uncovered (sunlight) quartz bottles in a water bath on ship board in December 2001 Fig. 6 A trichocyst battery (A), filamentous (B) and globular colonies (C) as well as picocyanobacteria in a tentacle (D) of a Limnocnida tanganyicae medusa Fig. 7 Diversity of bacteria in 11 Limnocnida tanganyicae jellyfish according to LH-PCR analysis of the 16S rRNA gene. The biomarker size 470 bp was affiliated to the Cyanobium -type picocyanobacteria Fig. 8 Depth distributions of bacteria in Lake Tanganyika water column according to LH-PCR analysis of the 16S rRNA gene. The gray line shows that the biomarker 470 bp had its relative maximum at 10 m depth Fig. 9 A neighbor-joining tree of the 16S rRNA gene sequence of the Cyanobium (represented by clone TK-SE6) dominating in Limnocnida tanganyicae medusae, with reference sequences of nonmarine and marine picocyanobacteria. Terminal branches display strain code and place of isolation (for Lake Biwa cluster), or the affiliated picocyanobacteria cluster (for strains of other non-marine clusters), based on the clusters designated by Ernst et al. (2003) and Crosbie et al. (2003). Numbers at nodes indicate the percent frequency (if >50%) obtained from the bootstrap analysis of 1273 nt positions of the tree Fig. 10 Oxygen consumption or production by individual Limnocnida tanganyicae jellyfish kept for two hours in light and dark bottles at ~27°C during November 1996 off Mpulungu, Zambia. Medusa diameters ranged from 10 to 20 mm Fig. 11 Oxygen consumption or production by 7 Limnocnida tanganyicae medusae kept first in darkness and then in light at ~27°C during November 1996. The bell diameters of medusae ranged from 10 to 20 mm. Each individual is represented by a uniquely shaded bar Fig. 12 Oxygen consumption or production at ~27°C of 4 Limnocnida tanganyicae medusae kept first in darkness, then in light, and finally with additional phosphorus and nitrogen during November 1996. The bell diameter of medusae ranged from 10 to 20 mm. Different individuals are represented by differently-shaded bars Fig. 13 Oxygen consumption or production of Limnocnida tanganyicae medusae in relation to their size (umbrella diameter) in light (white boxes) and darkness (grey boxes) at ~28°C during March 1998. The box represents the middle interquartile range of 50% of the observed values. The whiskers show the highest and lowest values, excluding outliers (black dots 1.5-3 box heights; circles > 3 box heights from the edge of the box). Horizontal lines across the boxes indicate medians. The numbers denote the number of animals in each size class Fig. 14 Mean stable isotopic composition (with standard deviations) of Limnocnida tanganyicae medusae and other major groups of organisms in the plankton of Lake Tanganyika. Numbers of replicate determinations are in parentheses 17 18 722 723 The figures are also on a separate file sent earlier. Fig. 1 10 Ash free dry mass (mg) 10 Dry mass (mg) 1 0.1 0.01 4 724 725 8 6 0.1 0.01 log y = 2.39 * logx - 2.45 r2 = 0.96 2 1 log y = 2.46 * logx - 2.62 r2 = 0.95 12 10 2 4 Fig. 2 Jellyfish abundance (Ind. m-3) 0 100 200 1000 3000 2000 0 4 6 8 Depth (m) 9 7 20 1 40 3 2 60 5 80 1 3 4 2 100 7 5 6 120 89 10 km Mpulungu 140 726 727 Fig. 3 Temperature (oC) Jellyfish (ind. m-3) 40 30 20 10 24 0 26 25 27 28 0-10 10-20 20-30 Depth (m) 30-40 40-50 50-60 60-70 70-80 Day Night 80-90 90-100 16 728 14 12 10 8 6 4 2 6 8 10 Umbrella diameter (mm) Uumbrella diameter (mm) 0 2 4 6 8 10 12 Umbrella diameter (mm) 18 14 16 12 19 729 Fig. 4 0 20 % of surface UV-B radiation 40 60 80 100 0 1 Cumul. UV-B (kJ m-2) Depht (m) 2 3 4 5 6 8 6 4 2 0 730 731 20 40 60 80 Time (min) Fig. 5 Number of alive jellyfish 14 12 10 8 6 in sunlight 4 with UV-film 2 in darkness 0 0 732 733 20 40 60 80 Time (min) Fig. 6 734 19 100 20 735 Fig. 7 100 Proportion of total (%) Other fractions 522 bp 80 518 bp 60 512 bp 505 bp 40 499 bp 472 bp 20 470 bp 0 736 737 1 2 3 4 5 6 7 8 9 Fig. 8 0 20 Proportion in total (%) 40 60 80 100 0 10 Other fractions 522 bp Depth (m) 20 518 bp 512 bp 30 505 bp 499 bp 40 472 bp 470 bp 50 60 738 739 Fig. 9 740 20 10 11 21 741 Fig. 10 Number of individuals 40 In darkness (n=66) 30 20 10 0 -30 -20 -10 0 10 20 10 20 Number of individuals 40 In light (n=48) 30 20 10 0 -30 742 743 -20 -10 0 O 2 (µg h-1 ind.-1) Fig. 11 10 Oxygen evolution (µg h-1 ind.-1) 13 h in darkness 2 h in light 0 LIGHT ON -10 -20 744 21 22 745 Fig. 12 746 747 Fig. 13 20 O2 (µg h -1 ind.-1) 10 4 3 14 7 3 4 0 12 6 6 10 14 13 3 0 1 7 2 -10 1 Umbrella diameter (mm) 748 749 Fig. 14 22 18-20 16-18 14-16 12-14 10-12 8-10 6-8 4-6 2-4 -20 23 6 Zooplankton 1 2 3 4 5 5 4 Fish 3 Shrimps 2 Phytoplankton 10 Pico (10) 11 Nano-micro (10) 12 Net (4) N 15 Limnocnida (15) Tropodiaptomus (5) Tropocyclops (4) Mesocyclops (4) Copepod nauplii (4) 8 1 . 6 2 6 Larvae (4) 7 Stolothrissa (21) 8 Big (1) 9 Small (10) 10 7 4 3 5 9 1 11 0 12 -1 -2 -3 -30 750 -29 -28 -27 -26 -25 13 -24 -23 -22 C 23 -21