SCOPE 54 - Phosphorus in the Global Environment - Transfers, Cycles and Management

Executive Summary
H. Tiessen

SCOPE 54 Phosphorus in the Global Environment - Transfers, Cycles and Management, H. Tiessen ed., 1995, 480 pp

Introduction And Synthesis

One result of synthesising the information gathered during the SCOPE project on "Phosphorus cycles in terrestrial and aquatic environments" is the realisation that the term "P cycle" is not appropriate in the global context. It is estimated that currently 33.106 Mg y-1 of P are discharged into the oceans (Howarth et al. Ch. 19), while in 1990 16.106 Mg y-1 were applied as mineral fertilizer (FAO, 1992b). The transfer to the oceans is largely the result of natural processes, which are accelerated by arable agriculture, concentrated animal husbandry and direct anthropogenic discharges. This volume therefore addresses P cycling and transformations in specific ecosystems, and also the less well understood transfers across landscapes and between ecosystems. A crucial problem in understanding the transfers from terrestrial to aquatic ecosystems is the great sensitivity of aquatic ecosystem function to P: the quantity of P sufficient to substantially alter the trophic status of a water body is negligible, and often not measurable, in the P budgets of terrestrial ecosystems in the watershed.

What are the processes that supply the essential nutrient P to the environment; and what is responsible for the oversupply which is threatening environmental integrity of aquatic ecosystems in some regions? In the natural environment, P is supplied through the weathering and dissolution of rocks and minerals with very low solubility. Therefore, P is usually the critical limiting element for plant and animal production, and throughout the history of natural production and human agriculture, TP has been largely in short supply.

Many natural ecosystems and low-input farming systems have adapted to low P supply by recycling P from litter and soil organic materials. Increases in productivity or exports of commodities require external nutrient inputs if they are not to cause a decline in fertility. External P inputs have become available on a large scale with the mining of phosphate deposits and the wide-spread availability of the commodity "phosphate". This has decoupled patterns of supply, consumption and waste production from natural nutrient cycles, and has made them dependent on economics. The critical question, when phosphate ore reserves will be exhausted, is hard to answer, with estimates ranging from <100 y to many hundreds (Runge-Metzger, Ch. 3). Although all continents show a positive P balance in their trade (excluding mined P deposits, which do not form part of the land ecosystems) (Beaton et al, Ch. 2), P remains in short supply over large parts of the globe, where economic and political constraints mean that naturally low P availabilities are only insufficiently supplemented by imported fertilisers (RungeMetzger, Ch. 3). In many regions, where the use of N fertilizer is favoured, this may cause a nutrient imbalance and inefficient N use.

Meanwhile, elaborate Systems of fertility management and a long history of P fertilizer use have built up soil P levels in some developed countries to such an extent that additional P is currently not needed to produce food for the resident population. When such countries or regions continue to import both large amounts of animal feeds and P fertilisers, and where populations are dense enough to produce significant amounts of waste phosphate, the P over-supply becomes a pollutant. In Europe, intensive animal production on farms with little land produces manure in excess of the nutrient requirements of crop and pasture lands (Sibbesen and Runge-Metzger, Ch. 4). The excess P from the animal manure, if it is not directly dumped into waste waters, accumulates in the surface soils which leads to increased leaching and erosion losses of P to aquatic environments. Current legislation may not address this problem effectively, although legislation to reduce the land application of P exists and is being tightened with planned substantial reductions in allowable P application to soils, and stricter controls on timing and management. This trend has considerable impact on agricultural land use, and farmers are no longer solely regarded as producers of food but are also held responsible for the ecological functioning of the landscape (Hedley et al, Ch. 5). The stewardship required for intensive agricultural production that does not jeopardise the surrounding environment or the finite resources of rock phosphate (90 y of supply are predicted at current rock exploration and increasing crop production levels), demands sophisticated methods of soil and landscape conservation, and nutrient management and monitoring.

Predicting the transformations - and therefore the potential for transfers from production sites to other ecosystems - of P requires detailed knowledge, not only of chemical equilibria but also of the rates of transformations of organic and inorganic forms of P in the soil (Frossard et al, Ch. 7), and of the ways in which P availability and turnover are affected by plant adaptations (Lajtha and Harrison, Ch. 8). Integrated farming systems, making use of biological processes of nutrient transformation and conservation, and relying on long-term balancing of nutrient budgets, challenge the current knowledge base and the understanding of soil processes and management options. Knowledge has been integrated and extrapolated with computer models which evaluate soil, crop, management and climate information to optimise fertilizer application and permit the use of low-cost rock phosphates or recycled wastes (Redley et al, Ch. 5).

A side effect of sustained heavy P fertilizer additions can be the accumulation and introduction into the human food chain of heavy metal contaminants contained in fertilizers (Mortvedt and Beaton, Ch. 6). Phosphate fertilizer use has caused small but significant increases soil cadmium levels. These inputs are of a magnitude similar to those from the atmosphere in industrialised countries. There are still no agreed safe Cd limits for soils, but it appears that concern about Cd build-up in soils may be warranted only where several critical factors combine, i.e. on acid soils with low cation exchange and low P fertility, to which significant P fertilisation is applied as low grade fertilizer or rock phosphate, particularly to Cd accumulators like leafy vegetable crops, and where the produce is the main source of local food consumption. This may be the case in truck farming belts around population centres in the tropics.

In developing countries, where 95% of the population increase over the next decades will occur, soil fertility is often low, and it will be difficult to meet the increased demand for food and fibre. The limited purchasing power of developing nations has resulted in nutrient mining of some of the better soils with the result that current productivity trends are downward. The limited use of fertilizer is compounded by the nearly complete use of high quality lands, which forces additional agricultural production into infertile acid upland soils. Too little fertilizer P is actually used in those parts of the world where P input could have a major effect on food production. Phosphorus use may have to double in Asia and quadruple in Africa. Fertilizer subsidies have been used in attempts to remedy this situation, but they can cause market distortions, black markets, environmental problems and the misallocation of resources. However, for reasons of long-term resource conservation and for the reversal of land degradation, fertiliser subsidies should be considered on P-deficient soils, where environmental costs of the soil resource depletion may justify the cost of subsidies. Careful monitoring of social and economic side effects would be required (Runge-Metzger, Ch. 3).

The development of acid upland soils, common in developing countries, requires the use of low-grade rock phosphate, which is less susceptible to leaching and fixation in the acid environment (Hedley et al., Ch. 5; von Uexktill and Mutert, Ch. 9). As a result of logging and accelerated shifting cultivation accompanied by a rapid deterioration of soil fertility, an estimated 65-80 million ha worldwide have been degraded into anthropic savannah and scrub (von Uexktill and Mutert, Ch. 9). Reclamation of such degraded lands can be initiated by rock phosphate application and the establishment of leguminous covers. Subsequent management will have to be fine-tuned to release P from fertilizer in synchrony with crop demands.

Natural ecosystems on such acid soils usually conserve P in and recycle P from organic matter and plant residues. The rates of cycling of P, N and organic C are therefore closely inter-dependent, and must be viewed together if nutrient cycling and production dynamics are to be understood. Agricultural management can successfully mimic natural ecosystems, with - for instance - the use of mulches, green manure crops, legume covers, alley cropping and appropriate crop rotations (Hands et al, Ch. 10). Much research and education are required to replace or adapt the originally successful regimes of shifting cultivation with such new technologies appropriate for increased food and fibre demands on a shrinking land base. Traditional farm communities passed on an understanding of nutrient dynamics, as expressed for example in the Norfolk rotation or the intercropping patterns of the tropics. Extensive use of manufactured fertilizer has increased yields but has probably dulled this intuitive skill. Future generations of farmers will need to rediscover this knowledge to manage limited soil and fertilizer resources, aided by computer models and scientific knowledge that reduce the risks in farming diverse soils under variable and changing climatic conditions. Many research results have not yet reached the farm level (e.g. use of mycorrhiza, the thorough understanding of nutrient cycles). This indicates that the linkage between scientists and farmers is not well established, training of farmers is insufficient, and feedback from farmers' problems to research is limited.

The lessons learned in managing a limited P resource under the economic constraints of developing countries may also benefit developed regions where P "leaking" from production systems poses environmental threats. Fertilizer use to overcome P deficiency has permitted intensive sedentary agriculture which feeds large urban populations that now have waste P problems. Negative effects of P use are mostly associated with redistribution of P into aquatic environments where P limits bioproductivity and excess P increases the cycling of carbon, nitrogen and sulphur.

While the in situ transformations of P, and their effects on C and N cycling are fairly well understood (Frossard et al, Ch. 7; Hedley et al, Ch. 5; Hands et al, Ch. 10), the workshops have highlighted the complexity of P transfers in the landscape, and the lack of quantitative knowledge on landscape relationships. The P and N status of the world's major rivers reflects land use in their drainage basins (Caraco, Ch. 14), but the intricate relationships between land use patterns, surface and subsurface transport of P, and aquatic P loads are only now being explored. Because of their great sensitivity to P inputs, water bodies are a sensitive diagnostic for monitoring land use practices in the watershed. Based on the impact of land use on water quality, and using process studies of P transport, some general principles on nutrient management in landscapes emerge from the chapters by Sharpley et al (Ch. 11) and Hillbricht-Ilkowska et al (Ch. 12). Point sources of P pollution from human population and industry add important P loads to surface waters, but these point Sources are more easily controlled than diffuse sources. Reclamation and recycling of the P contained in industrial and urban waste waters is an important measure for pollution control. Although recycling is limited to some extent by heavy metal contamination, the problem is one of policy and engineering. Point sources of P pollution are therefore not addressed in this volume.

In many aquatic ecosystems, non-point sources of P dominate the eutrophication process. Total P export rates from a watershed are related to arable land cover, slope, stream network density and rainfall, but the relationships are not strong enough to predict P exports to aquatic systems. Leaching and groundwater transport of P usually contribute less than 10% of the total P transport, the remainder being transported over the surface with runoff and erosion. Leached P, though, consists of soluble compounds with greater bio-availability and greater effects on eutrophication. Amounts of leached P are positively related to amounts of organic P and organic matter in the soil, and to P saturation resulting from over-fertilisation. P losses from soil erosion and surface runoff can range from 0.1 to 10 kg ha-ly-1, or more on steep slopes. Mixed land-use patterns of arable plots with forest fragments, meadows and pastures can halve total P export rates from watersheds. Afforestation belts, meadow strips and mid-field marshes, both natural and planted, are effective land-use modifications effecting P retention.

Land margins of undisturbed vegetation belts or riparian zones bordering aquatic ecosystems act as filters for P and other nutrients, and also as catches for surface sediment runoff. Retention, particularly of soluble P is greatly influenced by the water balance of the ecosystem (Kedziora et al, Ch. 13), and retained P may be rereleased under reducing conditions or in organic forms. In situations of high nutrient loads, it may therefore be necessary to harvest P scavenger plants in riparian strips, or excavate P-saturated sediments and transport them back to the agricultural source site.

Watersheds cannot be managed only to control P transfers; different management goals such as nutrient conservation, the accretion and/or conservation of soil organic matter and the maintenance of the biological diversity of ecosystems must be balanced. Environmental management should be based on a network of small-scale, "closed" nutrient cycles associated with a variety of patches, spatially and temporally connected in the landscape by water and nutrient flows. Control and prediction of nutrient movement may depend on the number, size, shape, sequence and configuration of patches in a landscape. This is a level of complexity well beyond current landscape-, erosion- and nutrient cycling- models.

In addition to sophisticated models of P transport in landscapes, a simplified approach has therefore been developed, based on a susceptibility index using soil and site properties and management information (Sharpley et al, Ch. 11).

Runoff and, to a lesser extent, groundwater move P to surface waters, which thus receive a mix of dissolved and particulate P. This mix varies greatly with the amount of sediment in runoff, and with storm intensity and duration. On average, surface waters contain 10 to 1-1 of soluble P, and suspended sediments carry 1000 g g-1 (Melak, Ch. 15). The average suspended sediment is therefore P enriched relative to the source soils, and the solution P concentration is similar to that of average soil solutions, but higher than rain water (Salcedo and Medeiros, Ch. 20). The resulting P exchanges between the solution and suspended or bed sediments greatly modify P bioavailability. Uptake of P by algae reduces solution concentrations, activating further exchanges. Eventually algae die, sediment out and/or re-release P for further uptake cycles. Algal growth due to P is the main cause for eutrophication of inland waters, although cells or populations may exhibit N or P limitation, or switch between them (Grobbelaar and House, Ch. 16). Cellular storage and population changes shift productivity limitations from one element to another, and introduce micro-scale and temporal variability. Therefore, the composition of surface waters rarely reaches equilibrium, and is largely determined by the kinetics of processes (Melak, Ch. 15).

The overall result of these processes depends on a mosaic of patches along streams and in lakes with different substrates and biota forming highly complex systems at micro- and meso-scales. The net result of increased primary productivity from excess P, though, is relatively simple to gauge: increased biomass and subsequent decomposition, which can lead to oxygen shortages, faunal death and reductions in water quality. Many of the factors discussed are exemplified in the basin study of Lake Balaton, Hungary (Herodek et al, Ch. 17), which receives river inflow, filtered through a constructed marsh, and direct runoff from the surrounding agricultural and urban landscape.

Fluvial transport of biologically active P adds up to about 2 Tg y-1, all of which eventually reaches estuaries and coastal seas (Howarth et al, Ch. 19). where eutrophication is an increasingly serious problem (Fisher et al, Ch. 18). Nitrogen appears to play a more important role in the eutrophication of coastal waters than P, but spatial variability is great. As freshwaters mix with tidal marine waters, both geochemical and biological conditions change rapidly, and complex nutrient limitations involving N, P and Si occur (Fisher et al., Ch. 18). Coastal biota such as corals, sea grasses and algal beds are particularly sensitive to increased nutrient inputs. Effects are aggravated by long residence time of water masses in deep channels, and continuous density stratification of the water caused by salinity gradients. The latter predispose the water mass to oxygen deficiency resulting from excess biological productivity. Because of the greater volumes over which biogeochemical processes are integrated in marine environments, laboratory and microcosm studies become difficult to apply. As a result, the reliability of data on nutrient cycles and on nutrient limitations decreases from fresh water environments to estuaries, coastal seas and finally open oceans (Fisher et al, Ch. 18; Howarth et al, Ch. 19).

Tropical mangroves serve as an example of the full complexity of P cycles and transfers at the terrestrial/aquatic interface (Salcedo and Medeiros Ch. 20). Here, fluvial P transport meets soils or sediments undergoing periodic flooding with both fresh and saline waters. Both terrestrial plants and aquatic organisms use and recycle this P. Mangroves, about whose P cycles relatively little is known, may therefore be an excellent proving ground for methodological studies. Extractions (sequential or single) of soils and sediments, speciation of water P, kinetics of physico-chemical exchange and biological cycling and the accretion in the long-lived biomass of mangrove trees, all play a role in the nutrient balance. The discussion of the mangrove ecosystem therefore also sheds some light on methodology as it applies and is used by the separate groups of aquatic and terrestrial scientists.

At the core of all the science and data presented are some real concerns about food security and environmental integrity. Some of these can be summarised in programmatic statements:

Phosphorus is a major constraint on food and fibre production in many parts of the world. Therefore, an economical supply of P is a necessity for a secure production in agriculture and forestry.

An inefficient way to satisfy this P demand is by the exclusive use of inorganic fertilisers, which normally are expensive imports, and which often have relatively low use efficiencies in the field.

Management and research must aim at nutrient and organic matter cycling in combination with sufficient fertilizer use to avoid "nutrient mining" in agriculture and forestry.

Regions which have a grossly positive P balance through the import of fertilisers and feeds require political and legal mechanisms to (1) control land-use patterns, (2) avoid the decoupling of P discharges of animal production from the P requirements of plant production, (3) manage landscapes in an integrated manner that avoids discharge to ground and surface waters, (4) facilitate recycling of urban wastes through reducing heavy metal contamination.

Eutrophic inland waters and their watersheds should be managed to limit P inputs, because P is easier to control than N, which can be entrained from the atmosphere.

Coastal waters are increasingly suffering from high nutrient loads. Phosphorus is only part of their problem, and management will be complex. The sensitivity of some coastal ecosystems means that prevention should be favoured over damage management.

The differences in the scientific approaches to terrestrial and aquatic ecosystems are ignoring some of the similarities between the Systems. This might be remedied by more active dialogue among the disciplines. Certainly aquatic ecosystems are the recipient of nutrient losses from the land and should serve as indicators for the success (or lack of) of terrestrial P management.