Small-scale African farmers typically have few resources for the management of weeds and soil fertility. Weed composition varies with soil properties and crops compete with weeds for available nutrients. We conducted field surveys in four districts to study the effects of cropping systems, fallow, landscape position, and soil characteristics on weed composition and evaluated the value of weed species density in predicting soil fertility conditions. Weed density was higher with annual than for banana-based cropping systems with higher relative densities of hairy beggarticks and garden spurge and lower relative densities for Cyperus spp. and johnsongrass in the banana-based systems. The relative densities of weed species varied with the length of the postfallow period, landscape position, drainage, and soil properties. Long-term rotation of annual and perennial cropping systems may be more effective than fallow in reducing total weed numbers, but fallow may be effectively used to reduce the relative densities of tropical spiderwort, Cyperus spp., and goosegrass. Relative densities of blue couch, goosegrass, garden spurge, Cyperus spp., oxalis, and johnsongrass varied more with nutrient supply than for other species. Soil properties had less effect on the distributions of tropic ageratum, hairy beggarticks, tropical spiderwort, and smallflower galinsoga. High relative densities of blue couch and garden spurge were generally associated with low soil nutrient levels. Goosegrass, johnsongrass, and oxalis were associated with higher nutrient levels. Information on the relationship of soil nutrient levels with weed flora was used to develop a decision guide to aid farmers in the identification of areas in their fields with severe nutrient deficiencies.

The vegetation of the treeline ecotone of the southern declivity of arid High Asia (Hindu Kush, northern areas of Pakistan; Himalaya, northern central Nepal) is dominated by hedgehog-like open dwarf shrublands of thorny cushions. Since climatically sensitive ecotones are always also sensitive to human impact, the question arises whether the current lack of forests is a result of the Subboreal climate decline or of human impact. Due to inadequate knowledge of the pollen flora and of ecological indicator values of the plants, pollen analyses in High Asia have mainly been limited to the regional verification of globally known climatic impulses. However, the role of human impact on regional vegetation patterns has been widely neglected. We postulate that today's open dwarf shrublands replace woodlands and forests. Isolated vigorous juniper trees and successful reforestation appear to confirm our hypothesis. An abrupt decline of Pinus forests before 5700 and 5400 ka cal yr BP can be demonstrated. As the first indicator pollen of human impact appeared at both sites synchronous with the forest pollen decline, we infer human impact to be a more decisive cause for this environment change superimposing the effects of a climatic deterioration. The forests were displaced by open dwarf shrublands.

The relative ozone sensitivities of 25 German native herbaceous plant species, representative of arable field margins or disturbed habitats, were examined over three consecutive growing seasons. Plants were grown from seed, potted into natural soils and exposed in open-top chambers for the entire season to different ozone-exposure regimes covering a range of concentrations from <5 to 48 ppb (seasonal 8 h daily mean). The assessment of ozone effects was carried out by recording the first day of visible symptom appearance and the percentage of injured leaves at the end of vegetative growth. Species exhibited contrasting patterns of symptom expression under ozone stress, with either ozone-specific symptoms or ozone-enhanced foliar pigmentation and senescence. Classifications of species according to their ozone susceptibility varied depending on whether measurement was of the total extent of visible injury, ozone threshold doses for the incidence of symptoms, or modelled exposure–response relationships. The most sensitive species exhibiting ozone-specific symptoms were Cirsium arvense and Sonchus asper, which responded to accumulated ozone exposures <1500 ppb.h (AOT40). For these and three other species, an AOT40 peak of a single day was found to be responsible for the incidence of ozone-specific symptoms, i.e. injury occurred rapidly within a few days of the day with the highest AOT40, while other species responded only to longer-term ozone exposures. The relative ozone sensitivity of the species was calculated by combining the different sensitivity criteria, and possible systematic trends (taxonomic or evolutionary features) are pointed out. The results suggest it may be possible to use a particular group of native herbaceous plant species with contrasting patterns of ozone sensitivity as a biomonitoring system in the field. This allows plant responses to be related either to peak values or to prolonged ozone exposure, making it possible to distinguish between short- and long-term effects of ozone.