The Greenhouse Gas Balance of Italy

In this chapter a comprehensive assessment of the greenhouse gases budget of the Italian terrestrial ecosystems is provided, with particular attention to forest, cropland and grassland ecosystems and some case studies focusing on Italian shrublands and lands naturally or artificially converted to forests. Different methods have been applied and compared, such as regional measurements, use of flux networks and data-driven models within specific sectoral approaches in order to characterize the greenhouse gases budget of terrestrial ecosystems. The results presented respond also to the growing interest of the recent years in the role of the carbon cycle of terrestrial ecosystems and its relevance for national policies on mitigation and adaptation to climate changes.

In this chapter the Eddy Covariance network of Italy is presented, with a short introduction to each of the 29 sites that were active during the CarboItaly project. These sites provided a unique dataset for a better study and understanding of the carbon cycle of terrestrial ecosystems and the links between carbon sink capacity and the main environmental factors. After a number of examples of Eddy Covariance time series where it is possible to see the effect of interannual climate variability and disturbances and managements practices, an analysis of the role of the Carbon Uptake Period in the total Net Ecosystem Exchange (NEE) definition and a study of the effect of temperature and precipitation on the interannual variability of NEE are presented in order to show the way these data can contribute to a better understanding of the role and response of ecosystems to climate change.

A GIS-based model was developed so to predict Biogenic Volatile Organic Compounds (BVOCs) emissions from the Italian forest ecosystems in order to estimate the fraction of the Net Primary Production lost as reduced carbon and to assess the impact of BVOCs in the formation of ozone and secondary organic aerosols. The performance of the model was verified by comparing the predictions with BVOC fluxes measured in a CarboItaly site using the gradient method with tethered balloon profiles, but also with BVOC fluxes measured in previous years. The agreement between observations and predictions indicated a rather accurate estimation of the model and confirmed the importance in the Italian peninsula of monoterpene emissions, expecially of the fast reacting sabinene in areas dominated by Fagus sylvatica L. and Castanea sativa L.

The use of statistical sampling coupled with periodic re-measurements in permanent sample units provides the basis for accurate measuring changes in forest extent and conditions, and constructing reliable models to estimate trends. This chapter explores this topic in order to understand how to integrate land use and forest inventory approaches so to estimate CO₂ emissions and removals in managed forest ecosystems. The approach adopted by the Italian National Registry of carbon sinks, targeted to report the activities elected by Italy under the articles 3.3 and 3.4 of the Kyoto Protocol, is based upon the integration of a robust land use inventory, driven by ortho corrected airborne images, and field data from the permanent plots of the national forest inventory. On the whole, the amount of forest land remaining forest in the period 1990–2008 is estimated to exceed 9,000,000 ha; since 1990, over 350,000 ha of croplands and grasslands have been converted to forest, while nearly 127,000 ha have been deforested. The annual change in aboveground living tree biomass in forest land is estimated as high as +14.7 Mt, with an uncertainty (95 % confidence interval) of ±2.8 Mt, which corresponds to an annual carbon gain estimated between 1.35 t C ha⁻¹y⁻¹ and 1.45 t C ha⁻¹y⁻¹.

Regional approaches to estimate the carbon budget of Italian forest ecosystems using Process-Based Models (PBMs), have been applied by several national institutions and researchers. Gross and net primary productivity (GPP and NPP) have been estimated through the PBMs simulations of carbon, water, and elemental cycles driven by remotely sensed data set and ancillary data. In particular the results of the GPP and NPP estimations provided by the implementation of two hybrid models are presented. The first modeling approach, based on the integration of two widely used models (C-fix and BIOME-BGC), has been applied to simulate monthly GPP and NPP values of all Italian forests for the decade 1999–2008. The approach, driven by remotely sensed SPOT-VEGETATION ten-day Normalized Difference Vegetation Index (NDVI) images and meteorological data, provided a NPP map of Italian forests reaching maximum values of about 900 g C m⁻² year⁻¹. The second modeling approach is based on the implementation of a modified version of the 3-PG model running on a daily time step to produce daily estimates of GPP and NPP. The model is driven by MODIS remotely sensed vegetation indexes and meteorological data, and parameterized for specific soil and land cover characteristics. Average annual GPP and NPP maps of Italian forests and average annual values for different forest types according to Corine Land Cover 2000 classification are reported.

Emissions from forest fires are recognized to be an important health and environment issue. Fire emissions (FE) include a wide range of gaseous compounds and particles significantly contributing to the atmospheric budgets at local, regional and even global scale. In the last decades, several experimental and modelling studies were carried out to improve knowledge of the atmospheric impact of vegetation fires. FE estimates are affected by several errors and uncertainties; improvements were made possible through new advances in remote sensing, experimental measurements of emission factors and fuel consumption models. In this context, the aim of this chapter is to summarize the state of the research concerning atmospheric FE, highlighting the main methodologies and related uncertainties. In addition, this work presents an overview of historical trends and future scenarios of FE in Italy, starting from the most recent inventories.

The carbon stock in wood and paper products is increasing in Italy, and the same trend is expected in the coming decades. In forest ecosystems firewood and forest harvesting represent a net carbon loss but the use of wood, a carbon-neutral renewable resource for generating energy also has a strong substitution effect as it avoids the use of fossil fuels which are highly CO₂ emitting. The use of wood for construction purposes, substituting traditional materials, tends to increase carbon sequestration and contributes to climate change mitigation. The application of the GHGs accounting methods (IPCC 2003) suffers in Italy both for the lack of accuracy of wood harvesting official statistics and for the high level of uncertainty in the definition of wood products lifespan. Many authors have demonstrated a large underestimation of clear cuts areas in Italy leading to an underestimation of carbon loss due to a harvesting of about 2 Mt annually. However, it has been recently proved that multitemporal high resolution remotely sensed images may be operatively used with a probabilistic sampling procedure to obtain a more reliable estimation of annual wood harvesting extents. In any case an increase of the use of wood products for energy, building and furniture purposes may contribute to the reduction of GHG emissions and to a more sustainable development.

The aim of this chapter was to quantify the soil organic carbon (SOC) stock in the top 30 cm of mineral soil for the whole Italian territory, according to the different land use types of the Intergovernmental Panel on Climate Change (IPCC) cropland category (arable land, agroforestry, vineyards, olive groves, orchards and rice fields), as a basis for future land use scenarios and to address mitigation policy at country level. Besides, two independent studies addressing the current status and the future trends of SOC for the whole cropland category at regional level were reported. The subdivision of the cropland category into classes is functional to assess the impacts on the SOC stock due to land use changes from and to agricultural uses, providing the starting or ending point scenario. The differences emphasized for the soils of a subcategory under the different types of climate can be possibly used for future-oriented agricultural practices. The comparison of the total mean values of the different cropland subcategories shows significant differences in the SOC stock. Considering the year 2000 and applying to each subcategory area the specific average SOC stock value found in this study, the total amount of C stored in the upper 30 cm of the whole cropland category results to be 516.3 ± 156.9 Tg C. This amount represents about 17 % of the total SOC estimated for the top 50 cm of soils of total surface of Italy, which reports about 2,900 Tg C, thus indicating the importance to preserve the large amount of SOC stored in cropland category. In conclusion, given the few estimates available at European level, repeated SOC inventories aimed to define the SOC content in cropland soils are important for future stock change evaluation.

Nitrous oxide (N₂O) emissions from several Italian croplands along a latitudinal gradient were analyzed and the fertilizer induced emission (FIE) factor, for each single fertilization event, was calculated. Data show that the average emission factor was between 0.7 and 0.3 %, hence much lower than the IPCC EF used for temperate croplands. The relationship between N₂O production and applied N fertilization rate was exponential and not linear, although the rate of exponential increase was lower than previously reported. Maximum N₂O emission rate was correlated with magnitude of the total FIE, whereas it was inversely related to the length of FIE, which varied from a minimum of 8–56 days. Overall data suggest that the internationally applied emission factors for temperate crops, which are empirically derived from sites with cooler and wetter climates than the Mediterranean, would overestimate N₂O emissions for Italian crops, in particular those developing between spring and summer.

According to the latest National Inventory, the Italian agricultural sector is a source of GHGs with 34.5 Mt of CO₂ eq in 2009, corresponding to 7 % of the total emissions (excluding LULUCF). In particular, more than half (19.1 Mt of CO₂ eq) are N₂O emissions from soils. Although the national methodology is in accordance with Tier 1 and 2 approaches proposed by the IPCC (2006), still empirical emission factors are used to assess the emission from fertilizer (e.g. 0.0125 kg N₂O–N kg⁻¹ N from synthetic fertilizers). Disaggregated data at sub-national level, including models and inventory measurement systems required by higher order methods (i.e. Tier 3), are not available in Italy so far and comparisons with the other two approaches cannot be performed at the moment. Despite the large soil organic carbon pool in the agricultural soils and the recent institutionalization of the ‘National Registry for Carbon sinks’ by a Ministerial Decree on 1st April 2008, the last Italian greenhouse gas Inventory did not report CO₂ emissions from the agricultural sector. In this context, this chapter wants to summarize the main outcomes coming from the main long-term experiments present in Italy by integrating experimental and modeling approaches, which can provide national emission rates and a solid base to test and calibrate simulation models to estimate greenhouse gases emissions from Italian agricultural soils. What emerges clearly from the analysis is that the agro-ecosystems may sequester large amount of SOC if appropriate management practices are adopted. Moreover, the use of simulation models calibrated at local level and spatially applied, as done for the Carboitaly project, may certainly reduce the uncertainty of these estimations.

A common belief is that agricultural fields cannot be net carbon sinks, but perennial tree crops, growing a permanent woody structure with a life cycle of decades could act as carbon sink. Vineyards are good candidates to test this hypothesis, because they are often grown with limited soil cultivation and produce plenty of woody pruning material that can be left on the ground. Three Eddy Covariance sites were established in different vineyards, along a north-south transect, in Italy, to study the role of vine cultivation in the carbon balance of the Italian penisula. The year 2009 was chosen as a reference year for the three sites, in order to compare carbon budget estimates in areas characterized by different meteorological, pedological and geomorphological conditions. In the three sites a carbon sink ranging between 814 (Negrisia site) and 89 (Serdiana site) g C m⁻² y⁻¹ was measured. Both climate (water availability and PAR) and management (in particular the presence of permanent grass cover) have a strong impact on the carbon balance of the ecosystems. Even if it can be argued that this sink may be only temporary and the built-up can be substantially disrupted at the end of the vineyard life cycle, these results show that there is a concrete possibility of storing carbon in agricultural soils. Proper practices can be defined to preserve this storage at best, greatly contributing to the global carbon budget.

In some regions of the world such as the Northern Hemisphere, the abandonment of agricultural land is one of the most widespread forms of land use change. In general, abandonment is followed by colonization by herbaceous and woody plants. Since the 1950s, wide areas of Southern Italy have been afforested for soil conservation improvement. In order to quantify the effects of agricultural abandonment and artificial afforestation on soil organic carbon (SOC), a dataset of 48 Sicilian sites has been analyzed. Because of their high environmental variability, these sites can be considered as representative of Southern Italy and in general of the Mediterranean basin. Soil samples were taken throughout all bioclimates in different successional stages (cultivated areas: orchards, cereal crops, herb-dominated plant communities, grasslands dominated by perennial grasses, garrigues and low shrublands, maquis, natural forests and in nearby artificially afforested sites (Pine plantations)). The study confirmed that SOC accumulation after agricultural abandonment depends on bioclimate: the highest SOC accumulation was recorded in the meso-mediterranean bioclimate, intermediate in the thermo-mediterranean, and the lowest in the supra-mediterranean bioclimate. Data showed that for C sequestration in the soil, artificial afforestation is not convenient in comparison to natural afforestation by spontaneous secondary succession processes.

In the last decades, the European Union has favored the conversion of marginal croplands to forest plantations also to mitigate climate change by increasing carbon storage in the biosphere. In Italy, recent estimates report that forest plantations cover a total area of 122,252 ha. The aim of the present paper was to quantify carbon stock and annual carbon flux in mixed broadleaf and poplar plantation in Friuli Venezia Giulia plain. Overall, 2,592 ha of mixed broadleaf plantation (1992–2006) and 3,113 ha of poplar plantations (2002–2006) have been established in the Region with a standing carbon stock of 223,400 and 397,753 Mg C, respectively. Soil is the largest carbon pool in both plantation types (60 and 63 % of total ecosystem carbon stock). The total annual carbon sequestration is respectively 8,855 and 40,286 Mg C y⁻¹ whereas soil accounts for 13 % in broadleaf plantations and 9 % in poplar plantations.

Although their carbon stock is relevant in assessing the baseline for the negotiation of future agreements with respect to carbon balance, there still are few available studies concerning the biomass and the net ecosystem exchange capacity of Mediterranean shrublands. In this chapter a preliminary overview on the biomass values concerning Italian shrubland communities and/or their dominant/characteristic woody species is provided. Many useful data on above- and belowground biomass issued from investigations carried out in other Mediterranean countries and concerning plant communities, which share the same ecological, floristic and structural traits of Italian shrublands. A preliminary finding of this research is the uneven degree of knowledge concerning the different non-forest woody communities. For example, there is still no literature on the biomass of some 2/3 of all the considered phytosociological units. Besides, both the above and the below-ground biomass of many Mediterranean shrubs show a very wide range of variation as they are strongly influenced by progressive succession processes and by the nature, the intensity and the frequency of disturbance factors. Thus, direct measuring of these values for each vegetation unit and dominant woody species should be encouraged and intensified. Monitoring activities concerning biomass increase are recommended as well: as a matter of fact, at present reference data on this topic are so limited and variable that it is not possible to confidently estimate the annual growth of shrubland communities.