Summary We should know the effects of soil use and management on the contents and forms of soil phosphorus (P) and the resulting potential for leaching losses of P to prevent eutrophication of surface water. We determined P test values, amounts of sequentially extracted forms of P, P sorption capacities and degrees of P saturation in 20 differently treated soils and compared these data with leaching losses in lysimeters. One‐way analyses of variance indicated that most fractions of P were significantly influenced by soil texture, land use (grassland, arable or fallow or reafforestation), mineral fertilization and intensity of soil management. Generally, sandy soils under grass and given large amounts of P fertilizer contained the most labile P and showed the largest P test values. Fallow and reafforestation led to smallest labile P fractions and relative increases of P extractable by H2SO4 and residual P. Arable soils with organic and mineral P fertilization given to crop rotations had the largest amounts of total P, labile P fractions and P test values. The mean annual concentrations of P in the lysimeter leachates varied from 0 to 0.81 mg l–1 (mean 0.16 mg l–1) and the corresponding leaching losses of P from < 0.01 to 3.2 kg ha–1 year–1 (mean 0.3 kg P ha–1 year–1). These two sets of data were correlated and a significant exponential function (R2 = 0.676) described this relation. Different soil textures, land uses and management practices resulted in similar values for P leaching losses as those for the amounts of labile P fractions. Surprisingly, larger rates of mineral P fertilizer did not necessarily result in greater leaching losses. The contents of P extracted by NaHCO3 and acid oxalate and the degrees of P saturation were positively correlated with the concentrations of P in leachates and leaching losses. As the P sorption capacity and degree of P saturation predicted leaching losses of P better than did routinely determined soil P tests, they possibly can be developed as novel P tests that meet the requirements of plant nutrition and of water protection.

Sulfonamide antibiotics can enter agricultural soils by fertilisation with contaminated manure. While only rough estimations on the extent of such applications exist, this pathway results in trace level contamination of groundwater. Therefore, we studied the transport of three sulfonamides in leachates from field lysimeters after application of a sulfonamide-contaminated liquid manure. In a 3-year period, the sulfonamides were determined in 64% to 70% of all leachate samples at concentrations between 0.08 to 56.7 µg L −1 . Furthermore, sulfonamides were determined in leachates up to 23 months after application, which indicated a mediumto long-term leaching risk. Extreme dry weather conditions resulted in highest dislocated amounts of sulfonamides in two of the three treatments. Furthermore, soil management such as tillage and cropping affected the time between application and breakthrough of sulfonamides and the intra-annual distribution of sulfonamide loads in leachates. Although the total sulfonamide leaching loads were low, the concentrations exceeded the limit value of the European Commission of 0.1 µg biocide L −1 in drinking water in more than 50% of all samples. Furthermore, the medium-term mean concentration of the sulfonamides ranged from 0.08 and 4.00 µg L −1 , which was above the limit value of the European Commission in 91 out of 158 samples. Therefore, sulfonamides applied to soils in liquid manure under common agricultural practice may cause environmental and health risks which call for a setting up of more long-term studies on the fate of antibiotics.

Exact information about soil water flow is needed to quantify solute transfer within the unsaturated zone. Water flux densities are often measured indirectly, e.g., with water-balance, water content-change, or tracer methods, and, therefore, often predicted with notable uncertainties. Over the last years, direct lysimetry methods have been increasingly used to study water and solute migration in soil profiles. A large weighable lysimeter is the best method to obtain reliable drainage data, but it requires relatively high investment and maintenance expenses. To reduce cost and improve comparability with undisturbed sites, a new technology to collect large monolithic soil columns with a surface area of 0.5-2 m 2 and a depth of 1-3 m as well as a containerized polyethylene (PE-HD) lysimeter station were developed. In addition, the station was fitted with a new high-precision weighing technique. In this paper, the latter is demonstrated with data from a newly constructed gravitation lysimeter. Besides recording rainfall and seepage, its weighing precision makes it possible to register mass input by dew, fog, or rime. It also permits a very accurate calculation of actual evapotranspiration. Because this new type of lysimeter allows a very high temporal resolution, it is ideally suited to develop and test models for soil hydrologic processes.

To test existing hypotheses about the plankton size distribution (PSD) in lakes, we intensively sampled the plankton of a eutrophic, stratified lake (Arendsee) and constructed size spectra for particles ranging in size from bacteria to herbivorous cladocerans. The Arendsee Lake annual mean spectrum extended over nearly 10 decades of body size and 2.1 decades of biomass per size class (minimum, autotrophic picoplankton; maximum, colonial cyanobacteria). The spectrum was continuous (no size class without associated biomass). These data indicate that continuous PSDs are not restricted to the plankton in large pelagic zones (e.g., central oceanic systems or large freshwater lakes). We conclude that differences in the biomass of the various size classes are a consequence of eutrophic conditions in Arendsee Lake, not of littoral or benthic influences. In addition, we evaluated the impact of daphnids on the shape of the PSD with a data set comprising single samples from each of 28 lakes in northern Germany plus seasonal data from Arendsee Lake. In lakes or situations with a large Daphnia biomass (90.5 mg fresh weight liter-'), the slopes of the straight lines fitted to the normalized PSDs were less negative. In the 28 northern lakes, but not in Arendsee Lake, there was greater variation about the fitted line when Daphnia biomass was high, indicating a less homogenous PSD. Discontinuities in the spectra were found only in samples from the 28 northern lakes and for higher Daphnia biomass. Our interpretation is that the feeding mode of large filter-feeders results in a higher transfer efficiency from small to larger individuals (less negative slopes), which produces changes to a nonhomogenous PSD.Plankton size spectra describe the distribution of planktonic organisms among biomass classes. Sheldon et al. (1972) found a "tendency for roughly similar amounts of particulate material to occur in logarithmically equal size ranges," and this observation inspired a considerable number of studies in both marine and freshwater systems. Quantitative comparisons of the size distribution of plankton (PSD) in different systems are usually based on normalized spectra, which show a frequency distribution based on the size of individuals. The PSD is characterized by two parameters of the straight line fitted to the normalized spectrum. The slope (s) characterizes the number of smaller individuals in relation to larger ones. Larger values of s indicate larger numbers of individuals in the higher size classes. The coefficient of determination (R2) of the regression is a measure of the smoothness (homogeneity) of the PSD and the overall variability of the biomass per size class. Higher values of R2 characterize a more uniform PSD.Marine studies in low-productivity offshore regions (equatorial and subtropical regions) have provided spectra with relatively uniform size distributions. The spectra become less homogeneous in shallow, coastal waters and in offshore wa- AcknowledgmentsWe thank B. Franz, U. Link, and E. Ruschak for techni...

No abstract