Leaves from over 1000 Brazilian native plants growing in the cerrado and neighbouring regions were sampled for C and N content. Half of these were analysed for N and further samples forC and ash content. Nodulated legumes from all three sub-families were included, together with two types of reference plant, non-nodulated legumes and non-legumes. Particular emphasis was placed on the large caesalpinioid genus Chamaecrista which is here for the first time reported to fix nitrogen in its native habitats. Woody and herbaceous species of this and other nodulated genera, with the exception of the mimosoid tree Stryphnodendron, showed evidence of nitrogen fixation. Amounts fixed were site-specific as was the N signature of reference plants. There was no evidence that nodulated legumes had higher leaf N than non-nodulated legumes: both were higher than non-legumes. Several species of Chamaecrista from section absus and species of Stryphnodendron had carbon contents of 50-55%, higher than previously reported for leaves. This was coupled with low (1-3%) ash contents. TheC values of plants with ≥49% C were significantly more negative than those with <49% C: most species in the former group were woody and most in the latter group herbaceous. Mimosa pudica was unusual in having a wide range of percent C, percent ash and C values; these parameters were significantly correlated. It is concluded that Brazilian native legumes can fix significant amounts of nitrogen in the nutrient-poor cerrado soils. Consideration of mineral and lipid nutrition will be necessary in order fully to understand relations betweenC, carbon content and other physiological parameters.

2001. Effects of soil decomposer invertebrates (protozoa and earthworms) on an above-ground phytophagous insect (cereal aphid) mediated through changes in the host plant. -Oikos 95: 441-450.We investigated if the activity of soil invertebrates (protozoa and earthworms) affected the performance of barley and if effects propagated higher up the aboveground food chain into herbivores (cereal aphid, Sitobion a6enae). Barley plants were grown individually in microcosms containing defaunated soil and grass residues. Plants were grown in soil containing: a) no added fauna, b) protozoa, c) earthworms, or d) protozoa and earthworms. After 7 weeks growth at 20°C three adult cereal aphids were added to each plant on separate leaves. The aphids were allowed to grow and reproduce for another 2 weeks before the experiment was destructively sampled. Amounts of mineral N in the soil and leached from the microcosms were significantly reduced by the presence of soil animals. Correspondingly plant biomass and total plant N content were increased significantly by soil animals, protozoa in particular. The different mechanisms responsible for changes in nutrient turnover in presence of protozoa and earthworms are discussed. Aphid performance was strongly influenced by the presence of protozoa, but not by earthworms. In the presence of protozoa the numbers and biomass of adult and juvenile aphids were significantly increased. These effects are likely due to an increased N content in barley plants and consequently increased nitrogen availability to aphids. The results underline that the detritivore and herbivore systems are intimately linked. M. Bonkowski, Technische Uni6ersität Darmstadt, Fachbereich

Summary1. Two lectins, concanavalin A (Con A) and Galanthus nivalis agglutinin (GNA), have anti-feedant properties that suggest a potential for the control of invertebrate pests of plants. We tested potato plants genetically engineered to produce each of these lectins constitutively, as well as the puri®ed lectins, for possible non-target eects.2. Laboratory studies with soil bacterial communities and a ciliate protozoan could detect no direct eect of either lectin over a range of concentrations. There was a signi®cant inhibition in the host-®nding response of a bacterial-feeding nematode when Con A or GNA was present in the medium at 0´5±50 mg ml3. A number of GNA-and Con A-producing potato lines had no detectable eects on the rhizosphere microbial and microfaunal populations when examined in pot trials. The incorporation of leaves from transgenic plants into soil reduced protozoan populations signi®cantly, but there was no subsequent eect on the decomposition of added cotton strips. 4. Controlled ®eld-release experiments demonstrated that, although GNA-producing potato lines consistently altered the physiological pro®le of the rhizosphere microbial community at harvest, the eect did not persist from one season to the next over a trial period of two ®eld seasons. There was no signi®cant eect of the best performing GNA line on the development of a subsequent barley crop. 5. A single Con A-producing line was tested in a controlled ®eld-release. The only signi®cant eects were transient reductions of c. 40% in soil protozoan populations and of c. 10% in potential microbial activity.

No abstract

Foliage of potato cv. Désirée was harvested from glasshouse-cultivated plants of five experimental transgenic lines expressing three different insecticidal proteins (snowdrop lectin, Galanthus nivalis agglutinin (GNA); jackbean lectin, Concanavalin A (Con A), cowpea trypsin inhibitor; (CpTi)), tissuecultured control plants and standard control (non-tissue cultured) plants. The foliage was subdivided into stems, upper, middle and lower leaves and analysed separately by HPLC for the solanidine-based glycoalkaloids a-solanine and a-chaconine. The results demonstrate that one or more stages in the plant transformation process (i.e. insecticidal-and marker-gene insertions, gene expression and tissue culture) resulted in a lower level of leaf glycoalkaloids than that found in either the tissue-cultured controls or standard controls, based on the selected potato lines transformed for insecticidal protein expression. However, the distribution of glycoalkaloids throughout the plant foliage was unaffected by genetic transformation and tissue culture, with the highest glycoalkaloid levels being observed in the top third of the plant. The importance of investigating unexpected effects of genetic engineering on plant secondary metabolism is discussed from an ecological viewpoint.