The pioneer works of EINSELE, MORTIMER, and OHLE on the linking between phosphorus (P) and iron (Fe) cycles seven decades ago created the theoretical basis for a long-standing paradigm among limnologists i.e., 'oxygen controls the P release from sediments'. While many empirical studies as well as strong correlations between oxygen depletion and P release seem to support this paradigm, various field observations, laboratory experiments, and repeated failures of hypolimnetic oxygenation measures cast doubt on its universal validity. The temporal existence of a thin oxidized sediment surface-layer could affect only fluctuations of the temporary P pool at the sediment surface but not the long-term P retention. On longer time scales P release is the imbalance between P sedimentation and P binding capacity of anoxic sediment layers. The P retention of lake sediments strongly depends on sediment characteristics and land use of the catchment. The presence of redox-insensitive P-binding systems such as Al(OH) 3 and unreducible Fe(III) minerals can enhance the P retention and completely prevent P release even in case of anoxic conditions. Alternative release mechanisms such as a dissolution of calcium-bound P and decomposition of organic P under both, aerobic and anaerobic conditions, are often more important than the redox driven Fe-coupled P cycle. Additionally, bacteria affect P cycling not only by altering the redox conditions but also by releasing P during mineralization of organic matter and by accumulation and release of bacterial P. Since microbial processes consume oxygen and liberate P it is difficult to distinguish whether oxygen depletion is the result or the cause of P release. Nowadays, the old paradigm is discarded and a paradigm shift takes place. Sedimentary P exchange ought to be considered as a complex process which is mainly determined by the amount and species of settled P as well as their subsequent diagenetic transformation in the sediment. The classical paradigm is only valid in special cases since reality is much more complex than suggested by that paradigm. Everything should be made simple as possible, but not simpler!ALBERT EINSTEIN

One of the most serious problems caused by eutrophication of shallow lakes is the disappearance of submerged macrophytes and the switch to a turbid, phytoplankton-dominated state. The reduction of external nutrient loads often does not result in a change back to the macrophyte-dominated state because stabilising mechanisms that cause resilience may delay a response. Additional internal lake restoration measures may therefore be needed to decrease the concentration of total phosphorus and increase water clarity. The re-establishment of submerged macrophytes required for a long-term stability of clear water conditions, however, may still fail, or mass developments of tallgrowing species may cause nuisance for recreational use. Both cases are often not taken into account when restoration measures are planned in Germany, and existing schemes to reduce eutrophication consider the topic inadequately.Here we develop a step-by-step guideline to assess the chances of submerged macrophyte re-establishment in shallow lakes. We reviewed and rated the existing literature and case studies with special regard on (1) the impact of different internal lake restoration methods on the development of submerged macrophytes, (2) methods for the assessment of natural re-establishment, (3) requirements and methods for artificial support of submerged macrophyte development and (4) management options of macrophyte species diversity and abundance in Germany. This guideline is intended to help lake managers aiming to restore shallow lakes in Germany to critically asses and predict the potential development of submerged vegetation, taking into account the complex factors and interrelations that determine their occurrence, abundance and diversity.

Sequential P extraction was combined with electron microscop and X-ray spectroscopy to characterise various P species and to study their transformation in settling seston and in recent sediment. During early diagenesis most of the particulate P formed in the water was redissolved. No net transformation into species that would resist dissolution was observed. It was shown that • the phosphorus (P) content and the P flux of settling particles varied seasonally over one order of magnitude • particles became enriched with reductant soluble P (BD-P) while settling through the hypo-.limnion • changes in BD-P were highly significantly correlated with changes in reductant soluble iron (BO-Fe) • bacteria oxidising Fe and Mn seemed to be mainly responsible for this increase in P concentration • other fractions including organic P did not change during sedimentation • most of the organic P and of the Fe bound P and 70% of TP was released from the sediment during early diagenesis • the sediment surface did not act as a trap for P migrating upwards from deeper sediment layers • CaC0 3 sedimentation contributed little to P sedimentation but significantly to the permanent burial of P.

Polyphosphate (poly-P) was detected with the use of 31 P nuclear magnetic resonance (NMR) spectroscopy in sediments from a large variety of lakes with different trophic state and morphometry. In the top 0.5 cm of sediment, poly-P was 1.5 to 11.4% of total P. Nonreactive phosphorus (NRP) in the NaOH fraction (often classified as organically bound phosphate) was up to 46% inorganic poly-P. In some surface sediments, the poly-P content equalled the iron-fixed phosphorus determined by chemical phosphorus fractionation. Sediments were probably supplied with poly-P by sedimentation because there were substantial amounts of poly-P in plankton and settling seston. As demonstrated with sediments of Lake Petersdorf, benthic organisms can also contribute to the formation of poly-P (up to 0.11 mg P [g dry weight]Ϫ1 ) under favorable aerobic conditions. Poly-P is more rapidly transformed into single orthophosphate during diagenesis than other inorganic and organic P species. The transformation of organic P compounds and poly-P can contribute significantly to the release of P during diagenesis and should be considered along with the reductive dissolution of P sorbed to iron oxihydroxides.Microbially induced changes in pH and redox potential affect the ability of lake sediments to retain inorganic phosphorus (Roden and Edmonds 1997; Gächter and Müller 2003). Although these changes are driven by organic matter decomposition, the direct role of sediment bacteria by the release of organically bound phosphorus often has been ignored. Many studies have shown that a substantial proportion of phosphorus in settling seston originates from biomass (e.g., Hupfer et al. 1995a;Pettersson 2001;Kleeberg 2002). The decreasing content of these organic P species in settled particles during early benthic diagenesis indicates that heterotrophic sediment bacteria mineralize sedimentary organic P to inorganic P (Wetzel 1999). However, some of the supplied P is assimilated during microbial growth. Studies with activated sludge and with mixed or pure cultures have shown that several types of microorganisms are able to take up P excessively and form intracellular polyphosphate (poly-P; Wentzel et al. 1991). In waste water treatment plants with biological P elimination, poly-P-accumulating organisms are dominant under oscillating redox conditions. AcknowledgmentsWe thank Christiane Herzog for her help with the analytical work. René Gächter, Roland Psenner, Jörg Lewandowski, and two anonymous reviewers are acknowledged for critical reading and helpful comments on a former version of the manuscript. Sarah Poynton is acknowledged for the linguistic improvements of the text.

Variations in the nonreactive P content of lake sediments as a consequence of changing redox conditions suggest that microorganisms may contribute to uptake and release of P. In this study, 31–50% of the nonreactive P was identified as polyphosphate (poly‐P) in NaOH extracts of sediments from eutrophic Lake Baldegg and oligo‐mesotrophic Lake Lucerne by means of 31P NMR spectroscopy. Poly‐P was present in surface sediments but not in deeper sediment layers. In samples where poly‐P was detected, analysis under a scanning transmission electron microscope equipped with energy‐dispersive X‐ray spectroscopy indicated the presence of sediment bacteria containing P‐rich granules. Although poly‐P is stable in pure NaOH solution, it partially hydrolyzes during and after sediment extraction with NaOH. Cracking of poly‐P molecules to shorter fragments is, however, much slower if the sediment is pre‐extracted with EDTA and NaOH is replaced by a mixture of NaOH/EDTA to extract nonreactive P from sediments.