Phosphorus uptake by Microcystis during passage through fish guts

Herbivorous fish feed on cyanobacteria. Digestability differs, however, between cyanobacteria species without mucous cover and mucilaginous genera such as Microcystis. The latter can pass fish guts almost undamaged, and it has been hypothesized that they can take up nutrients during gut passage. Here we tested whether live Microcystis, as food for juvenile roach labeled with 33P, indeed showed higher radioactivity after gut passage as compared to gut contents in control experiments with fish fed heated Microcystis. Microcystis showed high viability after passage through roach guts, and live colonies had a significantly higher radioactivity than dead ones. We conclude that Microcystis is protected against digestion in roach guts and can directly use the phosphorus supplied in the fish guts during passage. Cyanobacteria (e.g., Microcystis spp. and Aphanizomenon spp.) are sometimes dominant planktonic primary producers in eutrophic lakes. While a combination of several biotic and abiotic factors (e.g., resistance against grazing, low light climate, high water temperature, low TN/TP ratios, or hydrodynamic effects) influence blooms of cyanobacteria, their function as food for other aquatic animals is less clear. Although some cyanobacteria species are directly consumed by the zooplankton (DeBernardi et al. 1981), others seem to be protected from grazing by producing toxins or ‘bad taste’ compounds (Carmichael 1992). Feeding on cyanobacteria has been reported from herbivorous and omnivorous fish species as well (Prejs 1984; Datta and Jana 1998). Omnivorous roach (Rutilus rutilus), an abundant cyprinid in many European lakes, include cyanobacteria in their food, especially during the midsummer decline of large zooplankton and increasing roach densities. In contrast to more specialized phytoplankton feeders (e.g., tilapia), roach, like other stomachless fish, lack pepsinand acid-secreting cells in the intestine. While digestion is facilitated only through mechanical disruption by pharyngeal teeth, assimilation efficiency for cyanobacteria is low (Prejs 1984). However, the energy gain for fish while they are feeding on cyanobacteria depends on biochemical properties of cyanobacterial colonies. Whereas species without mucous cover, such as Aphanizomenon, can be well digested, and thus, their nutrients can be assimilated by fish (Vörös et al. 1997; Kamjunke et al. 2002a,b), mucilaginous species like Microcystis seem to pass the fish gut rather unaffected (Vörös et al. 1997; Datta and Jana 1998), and only the nutrients from attached bacteria were definitely assimilated (Kamjunke and Mehner 2001). On the other hand, cyanobacteria may even benefit by the gut passage through fish. Miura and Wang (1985) observed a doubling of the photosynthetic activity of cyanobacteria excreted by bighead (Aristichthys nobilis). From investigations of snails fed cyanobacteria, it had been supposed that undigested cyanobacterial cells can take up nutrients from the alimentary canal during passage (Cuker 1983). Phosphorus uptake was demonstrated by microautoradiography for the green algae Sphaerocystis after passage through Daphnia (Porter 1976) and by uptake of radiolabeled P for the green algae Scenedesmus in Daphnia (Boersma pers. comm.). A similar P-uptake of cyanobacteria during passage through fish guts has only been supposed so far (Miura and Wang 1985; Vörös et al. 1997). Therefore, we tested whether the mucous-possessing Microcystis can directly use the high phosphorus concentrations in guts without being digested if they are consumed by omnivorous fish. To do this, we adapted previous experimental procedures with radioactive tracers toward a protocol that labeled juvenile roach with 33P, and we fed them with live and dead Microcystis colonies. Methods—A natural cyanobacteria suspension (mainly Microcystis aeruginosa) was collected, using a 150-mm plankton net, from the surface layer of Lake Templin (80 km north of Berlin, Germany) during August 2002. The sample was subsequently resuspended in aerated tap water. Subsamples for analysis of species composition were preserved with Lugol’s solution. Cells were counted by the sedimentation technique, after having disintegrated the colonies, using a high-speed blender (Ultra-Turrax, 20,000 3 g). Attached bacteria were detached by homogenizing of subsamples with the blender. Biomass of attached bacteria was calculated as the difference between homogenized and natural samples (Kamjunke and Mehner 2001) and was estimated by staining with DAPI (496-diamidino-2-phenylindole) using an epifluorescence microscope (Axiovert, Zeiss). 33P uptake by and adsorption on live versus dead cyanobacteria were compared. To kill cyanobacteria, 400 ml of the suspension was heated to 758C for 45 min. Most of the cyanobacteria remained in colonies after heating. Two hundred milliliters of both live and heated cyanobacteria suspension were incubated with 1 ml PO4 (110 TBq mmol21, Amersham) at 228C for 3 h during daylight. To remove bacteria attached to colonies, 1-ml subsamples were homogenized by an Ultra-turrax (T 25, 10,000 3 g) and filtered onto 0.85mm cellulose nitrate filters (Sartorius). Filters were dried and placed in scintillation vials. After adding 15 ml Filtercount (Packard) to each vial, radioactivity on the filters was measured using a Liquid Scintillation Analyzer (TRICARB 1900; Packard). The external standard ratio method was used to correct for quenching. Viability of cyanobacterial cells was inspected separately