Acidification et cycle des éléments nutritifs

Air pollution is one of the main ways human activities modify global environmental conditions. In particular, the acid rain crisis, which roused the spectre of a generalized forest decline in the 1980s, brought to light the effect of long-distance pollutants from burning fossil fuels and from agricultural practices. The crisis pushed the international political community to put in place legislation to reduce the atmospheric emissions causing the damage (sulphur oxides, nitric oxides, ammonia) and to monitor their impacts on the environment.

In France, several complementary structures were put in place to monitor atmospheric deposition. Two networks, MERA (between 8 and 12 sites) and CATAENAT (27 sites included in the RENECOFOR network), whose sites are located away from the main sources of emissions, both measure background pollution for two European programmes (EMEP and ICP Forests). Though the methodologies used in these two networks respond to different objectives and do not provide comparable results in absolute terms, they do make it possible to independently detect relative tendencies and usefully complement each other when we wish to assess the efficiency of air pollution reduction programmes.

Between 1993 and 2015, both networks recorded a significant decrease in the direct acidity of atmospheric deposits, as reflected by an increase in pH (see Figure). Indeed, pH increased by +0.53% per year in the MERA network and by +0.34% per year in the CATAENAT network. This change is mostly due to decreased concentrations of sulphur in the form of sulphates (-3,12 % per year et -2,74 % per year in MERA and CATAENAT) and to a more moderate decline in nitrogen in the form of nitrates (-1,53 % per year et -1,21 % per year in MERA and CATAENAT). A decrease in nitrogen concentrations in the form of ammonium (-1,95 % per year et -1,86 % per year in MERA and CATAENAT) also indirectly contributed to reduced acidity in atmospheric deposits. On the other hand, a simultaneous reduction in calcium concentrations caused a non-negligible drop in nutrient content for the poorest soils. These trends were confirmed not only for open-air sites but also under forest canopy cover at the 14 CATAENAT sites.

The enforcement of new regulations designed to decrease sulphur oxide pollutants had a clear positive effect on sulphates in atmospheric deposits.

The situation for nitrogen deposition, however, is rather different; trends for nitrogen do not seem to have followed the reported decrease in emissions. The strong estimated decrease in nitrogen oxide emissions has not been entirely reflected in the atmospheric deposits of nitrates.

Meanwhile, ammonia emissions have remained stable, yet a decrease in ammonium deposits was detected.

What is more, spatial modelling based on emissions inventories does not reflect field observations for geographical variations in atmospheric nitrogen deposited in the form of ammonium. This discrepancy underlines the need for more research to better understand and more accurately simulate transport and transformation mechanisms of pollutants in the atmosphere. Therefore, continued monitoring of atmospheric deposits is necessary to evaluate the real impact of air pollutants on the environment and to define a strategy to control polluting emissions.

Summary

The soil is a reservoir of mineral elements where trees and vegetation tape the resources they need. Changes in soil mineral content therefore closely reflect the risks of nutrient imbalance that atmospheric pollution and nutrient loss due to tree harvesting can cause or exacerbate.

The acid rain crisis at the beginning of the 1980s drew attention to the impact atmospheric pollutants were having on forests, often hundreds of kilometres from their emission points. In effect, the acidifying effect of sulphur and nitrogen deposits was causing soil impoverishment in terms of exchangeable elements, with real deficiencies appearing in the poorest, most acidic soils. Notably, the tree foliage yellowing widely observed in the Vosges Mountains reflected a magnesium deficiency (Landmann and Bonneau, 1995). Furthermore, an increase in atmospheric nitrogen deposits had a fertilisation effect which accentuated the risk of deficiency by creating and imbalance in tree mineral nutrition and by increasing the trees' need for other nutrient elements.

In the 1990s, the first studies were launched to measure changes in forest soil properties by re-sampling sites in north eastern France which had been previously sampled, up to 20 years before (Dupouey et al., 1998). Despite the fact that polluting sulphur and nitrogen emissions had already been perceptibly reduced, these studies found that there was a trend toward impoverishment in terms of exchangeable elements (particularly magnesium and calcium), that acidification had increased (lower rates of saturation in exchangeable base cations) and that nitrogen enrichment had occurred (a lower C/N ratio).

Since then, atmospheric sulphur and nitrogen deposits have continued to decrease, yet the question remains as to whether or not chemical soil fertility has been able to recover in forested areas.

When RENECOFOR was established in 1992, for the first time, measurements of changes in the physical/chemical properties of forest soils became available at the national scale. The 102 RENECOFOR sites include a wide range of ecological contexts in mainland France and soil analyses have been carried out twice, first between 1993 and 1995, then between 2007 and 2012. Strict methodological precautions were taken to ensure data comparability. Forest floor litter was sampled according to morphological horizon (OL, OF and OH) and the underlying mineral soil was sampled in successive layers (0-10 cm, 10-20 cm, 20-40 cm). Spatial variations in soil type within plots was accounted for with the same sampling design for each field campaign: at each plot, 25 samples divided equally among five permanent sub-plots (or "clusters") were taken, then combined into a composite sample. The physical and chemical properties of these composites were then analysed at the soil analysis laboratory (INRA, Arras) following identical procedures for both sampling campaigns.

The changes observed in the soil were surprising: they mainly occured in parameters associated with organic matter (organic carbon and total nitrogen). Organic carbon stocks increased significantly between the two sampling campaigns (+0.34 tC/ha/y on average). This increase occurred mostly in the surface layers (litter and mineral soil down to 10 cm in depth) while organic carbon levels remained stable in the deeper soil layers (mineral soil from 10 to 40 cm in depth). On the other hand, though total nitrogen levels increased slightly in the surface layers (litter and mineral soil down to 10 cm in depth), they clearly dropped in the deeper soil layers (mineral soil from 10 to 40 cm in depth). There was, therefore, a global decrease in nitrogen stocks, which, while only slightly significant statistically, was still more than the decline in atmospheric nitrogen deposits alone could explain (-11 kg/ha/y on average). This unexplained decrease in total nitrogen may have been caused by leaching or by greater nitrogen immobilisation in the biomass. Due to these changes in carbon (C) and nitrogen (N) stocks, the C/N ratio increased in a highly significant manner in all the soil layers whatever the ecological context (+2.6 units on average for the litter and mineral soil layers down to 40 cm combined). This indicates a change in the quality of the organic matter in forest soils, and possibly indicates that decomposition slowed and favoured storage stability over time.

As for acidification, the phenomenon continued in the most acidic soils and in those most sensitive to acidification (pH H20 < 4.5). For these soils, pH decreased in each mineral layer, and the base saturation percentage also fell for global stocks of exchangeable cations in the 0 to 40 cm layers. However, this acidification did not bring about an impoverishment in absolute terms of these most acidic soils. Their exchangeable magnesium and potassium reserves actually increased during the same time period. This apparently contradictory trend appears essentially in the 0-to-10-cm mineral layer where the global increase in cation exchange capacity paralleled an increase in stored organic matter.

These results highlight the importance of organic matter dynamics, not only in terms of carbon sequestration but also for trends in forest soil chemical fertility. When associated with the other parameters measured on the same plots, these analyses make up an exceptional data set which can help us improve our understanding of the carbon and nutrient cycles and build more accurate models to reflect them.

Summary

Human activity has been contributing to increasing sulphur and nitrogen levels in the atmosphere, most notably since the end of the 19th century. These compounds, emitted into the atmosphere where they can travel long distances, then fall as atmospheric deposits, often affecting forest ecosystems.

Sulphur emissions have recently been restricted at the European scale, but nitrogen compounds are still emitted in important quantities and are difficult to control due to their variety of sources and forms and to the complexity of the nitrogen cycle and its chemical transformations. However, within the framework of the Geneva Convention on Long-range Trans-boundary Air Pollution (1979), European countries are making an effort to reduce emissions and atmospheric deposits, notably for nitrogen compounds.

Nitrogen is, of course, an important nutrient for forests, but nitrogen deposition has a recognised impact on soil biogeochemistry, the nutrient balance, tree growth, and more generally on forest health and under-story plant biodiversity. These effects depend on the environmental characteristics specific to each forest, and, when combined with the effects of the global climate change now under way, impact and upset forest ecosystem functions, though we do not as yet understand the processes involved.

It is primordial in such a context to be able to predict the effects of atmospheric deposition on the forest ecosystems in France. To do this, models such as ForSAFE-VEG, have been developed, which couple biogeochemical and ecological data. They make it possible to simulate the long-term impacts of atmospheric deposition and climate change combined on the biogeochemical responses of the soil and to predict cascade effects on forest biodiversity, while also integrating local environmental characteristics in order to define critical loads (a tool designed to assess the sensitivity of an ecosystem to nitrogen deposition). Developing and applying such models depends on having robust entry and validation data (quality and quantity of precipitation, soil solution composition, soil characteristics, biomass estimates, vegetation surveys...). These data are indispensable and must be available for each site or ecosystem the models are applied to.

The first research work on this coupled modelling approach has shown that the models must be adapted to the territory under evaluation to be pertinent. In France, the data obtained from the 25 years of RENECOFOR observations is of exceptional value when calibrating and testing site-scale models, which will then be extrapolated to the forest-ecosystem scale, and spatialised at the national scale. This spacialisation will be a robust basis on which to agree on concrete Europe-wide emission-control measures. Around ten CATAENAT plots were selected to validate and calibrate the models, then the 102 RENECOFOR plots, combined with other well-informed networks in France (the EcoPlant data base with 6000 sites, in particular), were used in the extrapolation phase. Parametrising vegetation data makes it possible to elaborate logistic regression models capable of predicting changes in several hundred forest plant species in mainland France as they respond to major alternations in environmental conditions.

Climate change and changes in nitrogen deposition have an important influence on soil solution composition on one hand, and on plant community composition on the other hand. Results underline the combined impact of deposition and climate change scenarios on certain key biogeochemical forest soil parameters; they highlight a predominant effect of climate on base cation saturation percentages, and of deposition on the nitrogen cycle. Furthermore, impacts have been found on long-term species abundance and diversity, depending on the species and its ecological preferences, as illustrated for plot CHS 41 in Fig. a for nitrogen deposition, and in Fig. b for the combined deposition-climate change scenario. Responses for ferns, grasses, forbs, mosses and bushes have been simulated (Fig. a) and show the ambiguous effect of nitrogen: either inhibiting development (grasses) or favouring it (mosses).

RENECOFOR site CHS41 located in a sessile oak stand

Even though climate seems to be the main driver, as shown by its obvious impact on base cations from 2080 onwards and on plant species responses, reducing nitrogen emissions as much as possible (MFR) has a positive effect on biodiversity. However, peaks and significant variations in plant cover coupled with clearly similar trends in the short term (25 years) can be linked to one or more factors, in addition to the role played by climate change and atmospheric nitrogen deposits.

Silvicultural harvesting, which opens up the canopy, or climatic events (storm events) can influence cation concentrations in the soil. In addition, more light available in the understory can also increase the ground cover percentage of light-demanding and semi-light-demanding species, and cause a decrease in the abundance of shade-tolerant species.