The availability of common standardized geospatial information on composition, structure and distribution of forests is essential to support environmental actions, sustainable forest management and planning policies. Forest types maps are suitable tools for supporting both silvicultural and forest planning choices from local to global scale levels. For this reason local authorities may develop forest types maps independently, in which case a standardized/harmonized framework for their comparison and aggregation is essential. At the same time local forest types maps may not be directly related to pan-European forest resources assessments and classification systems. This paper presents results of the harmonization of four forest types maps available for central Italy. The process is based on a bottom-up approach aimed at maintaining the most detailed common nomenclature system across the different Regions. The final results, in terms of forest types area, are compared with several independent sources of information: (i) two forest maps, one developed at national level on the basis of the Corine Land Cover 2006, and one for high resolution forest / non forest classification developed at pan-European level; and (ii) two sample based inventories: the Italian National Forest Inventory (INFC) and the Italian Land Use Inventory (IUTI). The results show that the proposed bottom- up harmonization approach is a suitable tool to guarantee the integrity and homogeneity of local forest types nomenclature systems, and to integrate such local data with European standards.
Information on the distribution and characteristics of forest ecosystems is essential to support decision makers for a correct implementation of Sustainable Forest Management (SFM) and to optimize the provisioning of ecosystem services (
In such a framework the Forest Types (FTs) approach, originally developed in Italy two decades ago (
FTs classification systems are usually organized according to a hierarchical structure: the main level is based on types which are grouped in categories usually on the basis of the main dominant tree species, types may be optionally further divided in subtypes or variants.
FTs are commonly assessed at stand level to be consistent with the scale of forest resource management units (
At a continental level, the European Environmental Agency (
Forest resource assessments can be applied using two principal methodologies: (i) forest inventory based on the investigation of a small part of the interested region in the form of a sample, the result is an aggregated statistic for the whole region or for a subpart of it; and (ii) forest mapping, that is nowadays based on the elaboration of remotely sensed data to investigate the entire region of interest. The result is a geographical depiction of the location of forest and related attributes (
Currently there are several independent sources of information available in Italy for estimating forest area, which are all based on different systems of classification. Following is a brief description of the state of the art in this field.
The second Italian National Forest Inventory (INFC 2005 -
More recently the Italian Land Use Inventory (IUTI -
Both INFC and IUTI projects provided estimates, with confidence intervals, at country level and for NUTS2 administrative units (called “Regions” in Italy).
In central Italy, FTs maps are currently available for many Regions: Tuscany (
In Italy, the Corine Land Cover (CLC) 2000 and 2006 maps were developed refining the original European nomenclature system. The standard forest classes (broadleaved/coniferous/mixed) were detailed in 14 FCs based on dominant species. The CLC maps were developed by manual delineation of multispectral satellite images, with a scale of 1:100 000 and a MMU of 25 ha (
At pan-European level, the European Commission’s Joint Research Centre (JRC) developed FMAP2006 and FTYP2006, a high resolution (pixel of 25 m) forest, non-forest map and a forest type map derived from the IMAGE2006 database, MODIS data and the CLC2006 map as training data (
It is important to note that in Italy the vegetation was recently classified by
Finally, forest habitats can be also classified according to the CORINE habitat nomenclature system that is based on a total of 230 classes (
The topic of contrasting different information sources for forest area assessment is relevant both for the scientific community and for forest managers and decision-makers. Such an interest was clearly demonstrated at a pan-European level by
Since the different monitoring projects available in Italy (INFC, IUTI, CLC and JRC forest / non-forest map, local forest maps) were developed independently of each other, the problem of harmonizing the different nomenclature systems across the different administrative Regions was extremely relevant. In the framework of the approach recently proposed by
The aim of this paper is to first present the development of bridges to create a harmonized high resolution FTs map (scale 1:10 000, MMU of 0.5 ha) based on local maps available in a study area in central Italy that includes the administrative Regions of Marche, Abruzzo, Lazio and Molise. This approach produced on a harmonized nomenclature system with relationships with other existing systems: INFC, CLC06, EFTs, vegetation of Italy (
The study area covers a total of 42 246 km2 and includes four administrative Regions located in central Italy: Marche, Abruzzo, Lazio, and Molise. The elevation ranges from sea level to the top of Gran Sasso (2914 m a.s.l.), the highest peak of the Apennine mountain range. The climate in the study area is mainly Mediterranean and temperate with precipitations concentrated in spring and autumn.
For the study area, four regional high resolution forest maps were used. The maps were originally produced at a scale of 1:10 000, with a MMU of 0.5 ha, and are all based on the manual delineation of digital orthophotos acquired in the period 2001-2007, integrated with field assessment (
The formal accuracy of these maps is unknown but they are all released with a certified thematic accuracy of at least 85%, according to the well-known international standard originally proposed by
The CLC06 forest map and the 2006 JRC high resolution forest / non-forest map, both available on-line, were clipped to the extent of the four administrative Regions. Finally the aggregated statistics of forest area for the different FCs of the National Forest Inventory (INFC -
The methodology adopted to harmonize the FTs nomenclature systems of the different dataset available in the study area follows a bottom-up approach, from local to European level, and from FTs to FCs.
Firstly the four FTs systems from the high resolution forest maps of Marche, Abruzzo, Lazio and Molise were compared and a new reference system, called Harmonized Forest Types Nomenclature System (HFTNS), was created for the study area. The new reference HFTNS was organized in Harmonized Forest Types (HFTs) grouped in Harmonized Forest Categories (HFCs). The HFTNS was built by comparing and assessing the correspondences among the diagnostic characters of the FTs nomenclature systems implemented in each one of the four investigated Regions (
Under a logical point of view, the different nomenclature systems were firstly aggregated with a merge (union) operator. The resulting system was then simplified by aggregating those local forest types that had the same diagnostic characters. Finally the types were aggregated in categories.
For each one of the resulting HFTs a unique correspondence was found with: (i) the FCs of the INFC nomenclature system; (ii) the EFTs
Each polygon of the four local high resolution forest maps was reclassified according to the HFTNS, and the resulting harmonized maps were merged.
The forest area for the four different administrative Regions was calculated on the basis of: (i) IUTI; (ii) INFC; (iii) CLC06; and (iv) the mosaic of regional HFTs maps.
By analyzing the bridges between the HFTNS and the other nomenclature systems it was possible to create a common simplified nomenclature to compare the forest area for the different classes on the basis of: (i) INFC; (ii) CLC06; and (iii) the mosaic of regional HFTs maps.
The first results of this test are the bridges created to reclassify FTs and FCs across the different nomenclature systems: those locally available in the four administrative Regions; the system adopted by the INFC; the Italian implementation of the CLC06; the EEA EFTs
In
The second set of results concerns the comparison of forest areas by FTs.
The HFTNS developed by harmonization of the different FTs nomenclature systems resulted in 42 HFTs grouped in 16 HFCs. In
Merging the four FTs maps based on the new HFTNS in the study area we produced a new harmonized map (
In
The forest area in the study area was calculated on the basis of the five available datasets (
The design based forest area estimations by IUTI and INFC tends to be very similar (1 470 564 ha and 1 432 485 ha, respectively), the difference for both is within their relative confidence interval. The forest area from the HFCs and the JRC forest / non-forest maps appears to be higher (1 487 822 ha and 1 552 179 ha, respectively), while that from CLC06 appears lower (1 303 730 ha).
When the same comparison is carried out separately for the four administrative Regions (
On the basis of the HFTs map (
To make possible the comparison of FCs across the three available sources (HFTs map, CLC06 map, and INFC statistics), we used an aggregated nomenclature system based on nine macro categories (
The results of the comparison are presented in
The remaining four classes show more relevant discrepancies. For other native broadleaved forests (HH-OBL), which represented 27% of the area, the HFCs and CLC06 figures were higher than INFC by 33% and 65%, respectively. For OWL (HSL_MSL_HMSL) (5% of the total) HFCs map was higher than INFC for the 99%, and the CLC06 was higher for a 257%, for riparian forests (RF) (3.5% of the total) the figure from HFCs map was higher than INFC for a 49% but that from CLC06 was lower for a 70% . Finally, for broadleaved plantations (NSP) (1.5% of the total) HFCs and CLC06 figures were lower than INFC for a 75% and for a 90% respectively.
This contribution presents the harmonization test carried out to merge high resolution FTs maps available for four administrative Regions in Central Italy. The maps were independently developed referring to the period 2001-2007 and adopted the same FAO forest definition, and were created with the same methodology.
Following the approach proposed by
The proposed bottom-up approach appears to be feasible to classify and harmonize forest resources. Indeed, it guarantees the integrity and homogeneity of local FTs systems, as well as the best possible characterization of forest resources at regional level, maintaining the possibility of upscaling and downscaling information at the different spatial scales.
The HFTNS was aimed at solving current difficulties in exchanging forest data and information between forest managers, land planners and decision-makers at different scales and across different local administrations. For example, on the basis of the HFTs map, the automatic derivation of a FTs map for those protected areas which are across the regional administrative borders is now possible. Harmonized nomenclature systems should be used for multi-scale forest resources monitoring to limit lacks and mismatches in available data and information, also contributing to reduce the delay in the implementation of common rules, laws and guidelines at regional, national, and European levels.
Even if the methodology was successfully applied in the case study presented, some critical points in harmonizing FTs need to be highlighted. It is possible to create bridges between different FTs systems but it is more difficult to find correspondences between forest classification systems developed using different approaches. This was the case when we compared local FTs systems with the forest classes used by the INFC, which are not based on the “forest type” approach but are more simply based on dominant tree species. On the opposite, the identification of bridges between the four local FTs systems and the EFTs system developed by EEA was quite straightforward (
In Italy, the “forest type” definition is based on a mix of physiognomic-structural characteristics and syntaxonomic and floristic information of the current stand conditions. Unlike phytosociological approach, the FTs approach is not based on the concept of climax or potential vegetation. Indeed, FTs represent only the actual forest vegetation resulting from the spatial and temporal scales and patterns of disturbances. Nonetheless, in most cases the correspondences are possible between the two approaches (FTs
The results obtained in forest area estimation by forest macro-categories (
When these results are analyzed separately for each Region (
When moving from FCs to forest / non-forest area estimation, the divergences between the different sources of information, including the pan-European mapping approaches, are reduced.
It is important to note that the agreement with INFC estimates is higher for HFCs maps than for CLC06 (
Local differences in the investigated Regions are probably due to the quality of the different maps, a more in deep analysis of omission and commission errors could be possible only if the location of INFC sampling units would be available.
In the future we hope that the different forest monitoring and forest inventory programs currently active in Italy, as well as in Europe, will potentially converge on a common nomenclature system. The raw data acquired in forest inventories can be used to obtain forest area estimates on the basis of multiple nomenclature systems. Stakeholders and agencies responsible for forest inventories and land use/land cover mapping should join the efforts at national and local level in order to develop harmonized multiscale systems. The EFTs (
An attempt to establish bridging functions between HFCs and Natura2000 Network would be very interesting to highlight the implications of forest type mapping as well as for mapping Natura2000 Network habitats.
Recently for the first time
Several possible and feasible solutions can be foreseen to derive a EFTs (
We would like to thank the colleagues at the
A special thanks to the anonymous reviewers who helped us in ameliorating a previous version of this manuscript, and to Daniel O. McInerney at Coillte (IR) for word revising and English check.
Finally, we thank Abruzzo, Lazio and Molise regional administrations for making available the FTs maps.
This work was partially carried out within the framework contract “Forests: fifth specific contract” between the European Environmental Agency, the European Forest Institute and the Italian Academy of Forest Sciences (contract no. EEA/NSV/10/004) and partially within the project “Development of innovative methods for forest ecosystems monitoring based on remote sensing”, funded by the Italian Ministry of University and Research (PRIN 2012, grant no. 2012EWEY 2S, national scientific coordinator Prof. G. Chirici).
Bottom-up approach adopted to harmonize FTs nomenclature systems. The flowchart shows the bottom-up approach adopted to harmonize FTs nomenclature systems at regional, national and European levels. From bottom to top, the arrows show the corresponding directions followed to harmonize the different FTs nomenclature systems at each step.
HFTs map of Central Italy. Thematic legend lists all HFTs as obtained through the harmonization process.
HFTs area (ha) for each region in central Italy. See
Forest area estimates (ha) from HFCs map, CLC06 map, IUTI, INFC and JRC for all regions in central Italy. Error bars represent SE of IUTI and INFC, respectively.
Forest area estimates (ha) from HFCs map, CLC06 map and INFC for each aggregated FC (see
Comparison of HFCs and CLC06 forest area estimation against INFC, values in 104 ha. Each case is one of the aggregated forest macro-categories from
Main characteristics of the local FTs maps used in the study.
Region | Regional area (ha) | Referenceyear | Forest area(ha) | Percent offorest cover |
---|---|---|---|---|
Abruzzo | 1 083 004 | 2005 | 454 017 | 41.9 |
Lazio | 1 722 620 | 2007 | 619 575 | 35.9 |
Marche | 972 860 | 2001 | 256 620 | 26.3 |
Molise | 446 107 | 2004 | 157 609 | 35.3 |
Aggregated nomenclature system used to compare HFCs, INFC and CLC06. For each one of the nine aggregated classes the correspondence with the different sources is provided. See
Aggregated HFCs | HFCs | INFC | CLC06 Code (National level) | CLC06Classes description |
---|---|---|---|---|
BF | BF | BF | 3.1.1.5(3.1.3.1.5) | Forests dominated by beech |
CF | CF | CF | 3.1.1.4(3.1.3.1.4) | Forests dominated by chestnut |
HO_CO | HO | HO | 3.1.1.1(3.1.3.1.1) | Forests dominated by oaks and other evergreen broadleaved species (holm oak and cork oak) |
CO | CO | |||
DO_SO_LO_TO | DO | DO_SO_LO | 3.1.1.2(3.1.3.1.2) | Forests dominated by other deciduous oaks (Turkey oak and/or downy oak and/or Hungarian oak and/or sessile oak) |
SO | ||||
LO | ||||
TO | TO | |||
HH_OBL | HH | HH | 3.1.1.3(3.1.3.1.3) | Forests dominated by other site-native broadleaves (mesophilous and meso-thermophilous broadleaves as maple-ash, hop-hornbeam and flowering ash) |
OBL | OBL | |||
RF | RF | RF | 3.1.1.6(3.1.3.1.6) | Forests dominated by hygrophytes |
NSP | NSP | NSP | 3.1.1.7 | Forests and/or ex-plantations dominated by self-sown exotic broadleaved species |
SNC | SNC.1 | SNC.1 | 3.1.2.1(3.1.3.2.1) | Forests dominated by Mediterranean pines and cypresses |
SNC.2 | SNC.2 | 3.1.2.3(3.1.3.2.3) | Forests dominated by silver fir and/or spruce | |
SNC.3 | SNC.3 | 3.1.2.2(3.1.3.2.2) | Forests dominated by oro-Mediterranean and Mountainous pines | |
HSL_MSL_HMSL | HSL | HSL_MSL_HMSL | 3.2.2 3.2.33.2.4 | Moorlands and bushes, sclerophyllous species areas, areas characterized by evolving woody and shrubby vegetation |
MSL | ||||
HMSL |
HFCs and HFTs at regional level in Central Italy, and related correspondences to regional FTs nomenclatures in Central Italy.
Correspondences among HFCs, the INFC FCs and Sub-Cs, the EFTs, and the CLC06 nomenclature systems.
Correspondences among HFCs, vegetation classes (at national level,