An international project is proposed to create a map showing the world subdivided into macroecosystem regions within each of which ecological conditions are relatively uniform but which show certain natural potentials and limitations. The map should tend to supplement the Dasmann-Udvardy system of biogeographical provinces, being of higher resolution and greater ecological relevance. The primary purpose of the map will be to serve as a reporting structure for information about global resources and environment, though it will be based largely on published information.

Maps based on classification of climatic types, vegetation formations, and soil groups, will be synthesized and generalized to delineate the areas to be shown on the ecoregion map. Its delineations will be refined through consultations with local experts, and through the interpretation of low-resolution remote-sensing imagery. The usefulness of the map is considered favourably in relation to national-level policy analysis, environmental monitoring, transfer of agricultural technology, compatibility with remote-sensing systems for monitoring environmental conditions, and agricultural activities, biomass estimation, macroreserve selection, and land management.

Attempts to model ecosystems have increased in recent years through the application of systems theory and the improvement in computer capacity and speed. A major problem with these models is providing data for input or validation. A potential source of data is information collected by remote-sensing techniques. Remotely-sensed data can be used in natural resource simulation models to provide spatial and temporal measurements, data for model calibration or validation, and independent feedback to keep the model simulation on track with reality. Remote sensing can provide spatial and temporal measurements of many landscape parameters that could improve our ability to understand and model the spatial and temporal characteristics of landscapes.

The challenge for remote-sensing scientists, landscape ecologists, and natural resource modellers, is to determine the most effective way to interpret and use the data from remote sensors in natural resource management. Natural resource models that can fully utilize the spatial data which remote-sensing techniques can provide, will almost certainly improve our understanding of landscapes and our ability to simulate and manage them wisely.

In the context of ‘ecology’ and ‘development’, both these terms are usually ill-defined. Development, a finite process, is nevertheless conceived as sustainable, while ecology is only used in a qualitative way in this context. Development concerns resources—renewable and nonrenewable—both of which are finite in their exploitability. Ecology in land-use concerns the interference of Man as a biological species in ecological processes.

Integration of ‘ecology’ in ‘development’ requires the following steps:

1. Quantify the resources, both renewable and non-renewable, to be developed in terms of carrying capacity and inputs available.

2. Determine the regulating ecological factors in both the natural ecosystem and the land-use system.

3. Assess the interventions needed to counter limiting factors, their effectiveness, and their (side-)effects.

4. Define measures to be taken in such a way that they can be integrated with cultural, social, and economic, data, in order to arrive at a balanced approach to development.

Studies have illustrated a sensitivity of the climate of the Arctic to natural and human-induced perturbations which greater than that in mid-latitudes or the tropics. This may be due, in part, to positive feedback linkages between the atmosphere and the cryosphere. We are only beginning understand the nature of the process linkages within the atmosphere-cryosphere-ocean-Biosphere system, however, and this is reflected in the discrepancies between various model simulations. This uncertainty increases our vulnerability to climatic change and variability. In the development of policy regarding management and development of our Arctic Heritage, we must attempt to reduce our vulnerability: we must be able to simulate the physical consequences of a perturbation with some veracity, and translate the results into the language of the policy analysts.

Three areas of research are highlighted as being important contributions towards this objective. The first is the derivation of climatically significant parameters from remotely-sensed data. Our climatologies and model validations are based upon spatially-biased data which include undetermined errors. With remotely-sensed imagery, we have access to previously inaccessible regions as well as spatial consistency in the data. The second area of research the detailed study of micro- and mesoscale processes at climatic ‘oases’. We must understand processes at these scales, and their linkage to the macroscale, if we are to design realistic General Circulation Models to simulate new climate states. The third area is effective communication of our knowledge in a form amenable to policy analysis. Current research is directed towards evaluating the impact of climatic variation in terms of risk assessment, and consideration of the policy implications of possible future climatic regimes through scenario analysis.

More than 485,000 hectares of the Ponderosa Pine forests in the southern Sierra Nevada of California are diseased as a result of air pollution, while approximately 25% of the area of Sequoia-Kings Canyon National Park receives enough oxidant air-pollution to elicit symptoms of smog injury on Ponderosa Pine, the dominant mid-elevation tree.

In 1983 we re-examined four permanent study plots that had been established in 1974–75. Smog symptoms on current needles in 1983 occurred on 23.8% of the trees compared with 14.5% only in 1975. Second-year needles had 60.7% showing symptoms compared with 44.2% in 1975. Less than 0.5% of the trees retained 6-years-old needles, and 21% had only 2 years of needle retention. Mortality attributed to air pollution was observed.

The paper starts with a report on the growing predominance of aquatic weeds of the genus Salvinia in the waterbodies of the world. The various ways in which Salvinia molesta, particularly, causes pollution and depletion water resources are then reviewed, as are the attempts made so far to control and eradicate Salvinia through chemical, biological, and mechanical, means. It is shown not only that these attempts have failed to achieve their purpose but also that there are possibilities of environmental backlashes being caused by the introduction of chemicals or bioagents into the environment. The importance developing environmentally safe methods of utilization Salvinia has been stressed, so that the cost of mechanical removal of the weed can be offset by the profits earned through its utilization.

Salvinia molesta, an aquatic weed native of Brazil, entered the floodplain of the Sepik River in Papua New Guinea in the early 1970s. It caused great distress to the local people by increasing exponentially to form thick mats covering many lakes and channels. A beetle from Brazil, Cyrtobagous salviniae, was established in one lagoon in 1982 and over the next 2 years 900,000 adults were redistributed manually among another 130 lagoons and lakes. Rates of beetle population increase (r) were close to 0.042 adult per adult per day in most lagoons. Local dispersal occurred when population densities reached 15 adults per kg of the Salvinia, damaging between 70% and 90% of Salvinia buds. Feeding by adults and larvae of the beetle caused entire mats of Salvinia to turn brown, die, and sink, but no deleterious changes in water quality were observed.

The time taken for beetles to spread throughout a lagoon and for the weed to be controlled was generally 12–18 months, depending on the area of the Salvinia mat, its thickness, and its mobility when blown by the wind. Herbicide was used to aid establishment of the beetle at one site by thinning a thick mat which had been colonized by sudd vegetation. An equilibrium appeared to be reached with the Salvinia present at < 1% of its former populationdensity.

The Republic of Niger is establishing a vast national nature reserve in the Aïr Massif and Ténéré Desert. With its 77,360 km2, the reserve will cover a wide variety of arid-land habitats. Both rainfall and vegetation are sparse, but the area harbours a varied fauna that includes a number of threatened species. The Twareg inhabitants of the reserve subsist by raising livestock, by irrigating small gardens, and by caravanning. Their day-to-day existence relies heavily on natural resources, and their continued existence will depend on the conservation of those resources. Drought, desertification, and abusive use, are all responsible for the deterioration of those resources.

Although highly necessary, the rational management of the area's natural resources will conflict with current landuse practices. Examples of where this happens are presented, and some proposals are put forward for their solution. Ideally, a flexible management system needs to be elaborated that will satisfy both the aspirations of the zone's managers and the immediate needs of its residents. The Authors hope that work in Niger will find its application elsewhere in the Sahel, and meanwhile emphasize the need to reconcile conservation with development.