13.1 Introduction

Monitoring national forest resources requires reliable and comparable information, derived from repeated large-scale observations. In order to effectively compare the results of different surveys, it is necessary that they share the definition of the study domain, for example the definition of forest, and that the respective domains are defined according to common criteria, for example the crown coverage or the extension of the forest. It is also necessary that variable definitions and the classification and measurement criteria of the variables be consistent with each other.

In a chapter dedicated to illustrating some of the main changes that occurred in the period between the first Italian national forest inventory (IFNI85) and the second (INFC2005), De Natale and Gasparini (2011) highlighted the importance of having multiple surveys over time to make comparisons between the inventory statistics. This was useful in identifying trends in the Italian forests. Furthermore, the actions to harmonise the two forest inventories, which in their design diverge for some defining and classification aspects and for some measurement standards, were briefly mentioned. Those actions were undertaken to both allow for the desired comparisons and the need to respond with greater accuracy to requests from important supranational statistical processes (cf. Chap. 1).

Contrasted to past comparisons, the differences between INFC2005 and INFC2015 are definitely smaller and limited to the improvement of some operational procedures in the collection and recording of information in the field. The inventory domain, the classification system and the stratification criteria, by region and forest type, have remained unchanged (cf. Chap. 2). Similarly, the measurement thresholds of the quantitative variables and the classes used for the qualitative variables are the same in the two inventories (cf. Chap. 4), and the calculation and estimation procedures remained unchanged (cf. Chaps. 5 and 6). This favourable condition allows to compare the various homologous quantities detected in the two inventories more easily.

Another important aspect for the comparison between the INFC2005 and INFC2015 results concerns the complexity of inventory design and the methods of selecting the sampling units. The first phase sampling points remained the same in the two inventories, while the second and third phase samples of INFC2015 included additional sampling points selected from those affected by changes in land use and land cover (cf. Chaps. 2 and 5). The samples of the two surveys are therefore not independent. The literature devoted to comparisons between repeated multi-phase inventories and to the analysis of the distribution of the estimators used in such complex designs is unfortunately incomplete. However, for the purposes of this chapter, an analysis was carried out to capture the extent and the sign of the changes between the two inventories, taking into account the variability of the estimates, without statistical hypothesis testing.

In general, the analysis of the variations between quantities estimated by sampling must consider not only the values of the estimators, but also the corresponding standard errors (SE), which indicate the variability of the estimates for different samples of the same size, with the same sampling design (Gregoire & Valentine, 2008). If the extent of the variation between the two sample estimates is large and much greater than the relative variability, it is difficult to hypothesize that this variation is not really different from zero. For this reason, despite the limitations mentioned above, we conventionally compared the estimates considering the difference between two estimates “considerable” when their estimation intervals (equal to twice the SE) did not overlap or otherwise “negligible”. In the comments, the percent difference between two estimates is shown together with their percent standard error, in order to immediately compare the size of the difference with the variability of the two estimates.

The results on the extent and signs of variations between the two inventories described in this chapter concern some inventory attributes considered more relevant for the analysis of the current dynamics of the Italian forests. They are the extension of wooded areas (Forest, Other wooded land and Total wooded area) and for the Tall trees forest, the growing stock, the aboveground biomass, the annual volume increment and removals, and the number of trees. The comparison was made by considering the estimates of the total values of each variable and the same estimates disaggregated by region and forest type. This was done to bring out any geographical differentiations in the variations that occurred in the period between the two surveys. Finally, the estimates of the new wooded areas and those of the areas affected by the change from wooded areas to other land uses resulting from the comparison of the sampling point classification in the two inventories are presented.

The results of the comparisons are sufficiently clear at the national level, while the estimates disaggregated by region or forest category sometimes show very limited variations and are associated with high standard errors of estimates.

13.2 Variations in the Wooded Area

13.2.1 Forest Area

The Forest area in Italy estimated by INFC2005 is equal to 8,759,202 ha, with a SE of 0.4% (Gasparini & Tabacchi, 2011). The analogous estimate from INFC2015 is equal to 9,085,186 ha, with a SE of 0.4%, and the variation between the two surveys is equal to 325,984 ha (+3.7%). Following the criteria described in the previous section, the variation between the estimates of the two surveys is considerable. At the regional level, it is observed that considerable area variations are reported only for Molise (20,686 ha, +15.6%), while the variations between the estimates for the other regions, all positive except for Trentino (−2143 ha, −0.6%) and Puglia (−3540 ha, −2.4%), are quantitatively negligible (Fig. 13.1). At the level of wider territories in the period between the two inventories, the extension of the Forest increased by 68,596 ha (+2.0%) in the Alpine regions, by 96,910 ha (+3.1%) in the regions of the northern and central Apennines and by 160,479 ha (+7.2%) in those of the southern Apennines and in the islands.Footnote 1

Fig. 13.1
figure 1

Estimates of Forest area of INFC2005 and INFC2015, nationally and for individual regions; error bars are equal to twice the standard error; bars for Italy were shortened to improve readability; real values are given in numbers / Stime dell’estensione del Bosco per INFC2005 e INFC2015, a livello nazionale e per le singole regioni; barre di errore pari a due volte l'errore standard; le barre per l’Italia sono state ridotte per migliorare la lettura, i valori reali sono indicati in numero

Observing the area estimates for the Tall trees forest types, negative variations emerge for the Norway spruce (−0.4%), for the Scots pine and mountain pine (−2.3%), for the Black pines (−3.1%) and for the Chestnut (−1.8%); for the other types the area variation is always positive. However, it should be noted that only for the Other deciduous broadleaved (+11.4%) and the Other evergreen broadleaved (+33.6%) is the variation considerable (Fig. 13.2). These are two residual classes, in which newly formed forest stands, such as the Acero-lime groves and the Robinia and Ailanthus forests, converge. For the whole of the forest types of Tall trees forest, which occupy 98.6% of the area of the Forest in Italy according to INFC2015, the variation in extension is positive and considerable (319,837 ha, +3.7%).

Fig. 13.2
figure 2

Estimates of Tall trees forest types area of INFC2005 and INFC2015; error bars are equal to twice the standard error / Stime dell’estensione delle categorie forestali dei Boschi alti per INFC2005 e INFC2015; barre di errore pari a due volte l'errore standard

13.2.2 Other Wooded Land Area

The area of the Other wooded land in Italy was estimated by INFC2005 to be 1,708,333 ha, with a SE of 1.3%; INFC2015 estimated for the same macro-category an area of 1,969,272 ha with a SE of 1.4%. The area variation between the two surveys is quantitatively considerable and equal to 260,939 ha (+15.3%). By disaggregating the area estimates of the Other wooded land by region, it can be observed that positive changes occur for all regions, but they are considerable just for Lazio (25,938 ha, +41,9%), Campania (26,453 ha, +43.5%) and Puglia (16,238 ha, +49,0%) (Fig. 13.3). At the level of wider territories, in the period between the two inventories the extension of the Other wooded land increased by 49,349 ha (+15.3%) in the Alpine regions, by 78,275 ha (+24.0%) in the regions of northern and central Apennines and 133,314 ha (+12.6%) in the regions of the southern Apennines and in the islands.

Fig. 13.3
figure 3

Estimates of Other wooded land area of INFC2005 and INFC2015, nationally and for individual regions; error bars are equal to twice the standard error; bars for Italy and Sardegna were shortened to improve readability; real values are given in numbers / Stime dell’estensione delle Altre terre boscate per INFC2005 e INFC2015, a livello nazionale e per le singole regioni; barre di errore pari a due volte l'errore standard; le barre per Italia e Sardegna sono state ridotte per migliorare la lettura, i valori reali sono indicati in numero

13.2.3 Total Wooded Area

In INFC2005 the Total wooded area (Forest plus Other wooded land) in Italy was estimated to be 10,467,533 ha with a SE of 0.3%; the similar INFC2015 estimate is equal to 11,054,458 ha, with a SE of 0.3%. The area variation between the two surveys is considerable and equal to 586,925 ha (+5.6%). By disaggregating the estimates of Total wooded area at the regional level, it can be observed that positive variations occur for all regions, with the sole exception of Trentino, by a very slight fraction (-0.1%). If we consider the uncertainty of the estimates, these variations appear considerable for many peninsular regions and for the islands: for Toscana (38,183 ha, +3.3%), Umbria (23,701 ha, +6.1%), Lazio (42,289 ha, +7.0%), Abruzzo (36,009 ha, 8.2%), Molise (24,632 ha, +16.6%), Campania (45,985 ha, +10.3%), Basilicata (35,986 ha, +10.1%), Calabria (37,689 ha, +6.1%), Sicilia (49,063 ha, +14.5%) and Sardegna (87,741 ha, +7.2%) (Fig. 13.4). At the level of wider territories, in the period between the two inventories the Total wooded area increased by 117,944 ha (+3.2%) in the Alpine regions, by 175,184 ha (+5.0%) in the regions of the northern and central Apennines and 293,794 ha (+8.9%) in the regions of the southern Apennines and in the islands.

Fig. 13.4
figure 4

Estimates of Total wooded area of INFC2005 and INFC2015, nationally and for individual regions; error bars are equal to twice the standard error; bars for Italy were shortened to improve readability; real values are given in numbers / Stime dell’estensione della Superficie forestale totale per INFC2005 e INFC2015, a livello nazionale e per le singole regioni; barre di errore pari a due volte l'errore standard; le barre per l’Italia sono state ridotte per migliorare la lettura, i valori reali sono indicati in numero

13.2.4 Land Use and Land Cover Changes

The variations in the Total wooded area and that of the Forest and Other wooded land resulting from the comparison between the relative estimates produced by the two most recent Italian forest inventories have been illustrated in the paragraphs above. Thanks to the repetition of the photointerpretation on all the points of the first phase sample and to the subsequent ground check on the second phase sample with the INFC2015 surveys (cf. Chaps. 2 and 3), it was possible to identify the changes in land use and cover that are relevant for the purposes of the forest inventory, i.e., between forest use and cover and other land use and cover, and also between macro-categories and inventory categories.

Table 13.1 reports the INFC2015 estimates of areas not affected by relevant changes between the two Italian forest inventories, by region and national. It includes the estimates of new wooded areas and areas in other land uses additional to the previous inventory. It is noted that the Total wooded area confirmed by INFC2015 is estimated to be 10,304,862 ha with SE equal to 0.3%, while the area confirmed in Other land use and cover is estimated to be 18,914,517 ha (SE 0.2%). Nationally, the new Total wooded area is estimated at 749,596 ha with SE equal to 2.7% and the additional area in Other land use and cover is equal to 163,870 with SE equal to 4.7%. The difference between the two estimates, which expresses the balance between the wooded area gained and lost compared to the previous INFC2005 inventory, is equal to 585,726 ha, a value very close to that obtained by the difference between the estimates of the Total wooded area of the two inventories reported in Sect. 13.2.3. The two estimates of the Total wooded area balance differ by a negligible and expected amount, due to the variability that characterises the estimates.

Table 13.1 Area estimates of land use and cover changes to and from the Total wooded area occurring between INFC2005 and INFC2015 / Stima dei cambiamenti di uso e copertura del suolo da e verso la Superficie forestale totale osservati tra INFC2005 e INFC2015

Table 13.2 reports the estimates of the areas affected by changes to and from the Forest and Other wooded land inventory categories at the national level. Similar estimates for the Italian regions are available at https://www.inventarioforestale.org/statistiche_INFC. It is observed that the area of the Forest estimated by INFC2015 that belonged to Other wooded land or to Other land use and cover according to INFC2005 is 21,174 ha and 464,818 ha, respectively. The increase in the Forest area due to the transformation of the areas previously under Other land use and cover is 5.3% of the Forest area according to INFC2005, while the loss of Forest area is 1.5%. The balance is positive for the Forest, which increased by 3.8%. For the Other wooded land, the transition to and from Other land use and cover is more marked, with an increase of 16.7%, a decrease of 2.1% and an overall balance of +14.6%. There are also changes, for very small areas, even between the two inventory macro-categories, with a slightly positive but not quantitatively considerable balance in favour of the change from Forest to Other wooded land.

Table 13.2 Area estimates of land use and land cover changes to and from Forest and Other wooded land occurring between INFC2005 and INFC2015 / Stima dei cambiamenti verso e da Bosco e Altre terre boscate osservati tra INFC2005 e INFC2015

13.3 The Variations of Some Characteristics of the Forest Stands

13.3.1 The Growing Stock

INFC2005 has estimated the growing stock of the Forest in Italy to be 1,269,416,499 m3 with a SE of 1.1%; with INFC2015 the same quantity is estimated to be 1,502,807,089 m3 with a sample uncertainty of 1.0%. The variation between the two surveys is quantitatively considerable and equal to 233,390,590 m3 (+18.4%).

Disaggregating the volume estimates by region, it can be observed that always positive and considerable variations occur for the following Italian regions: Piemonte (17,048,366 m3, +13.1%), Lombardia (22,090,286 m3, +20.4%), Veneto (17,269,061 m3, +21.3%), Toscana (23,624,901 m3, +17.9%), Umbria (6,470,477 m3, +22.1%), Lazio (15,330,012 m3, +26.7%), Abruzzo (12,947,702 m3, +25.6%), Campania (14,259,671 m3, +33.6%) and Calabria (23,666,353 m3, +26.9%). For the other regions, the changes are still positive, generally more limited and, given the uncertainty of estimates, they are quantitatively negligible (Fig. 13.5).

Fig. 13.5
figure 5

Estimates of the total growing stock of Forest from INFC2005 and INFC2015, nationally and for individual regions; error bars equal to twice the standard error; bars for Italy were shortened to improve readability; real values are given in numbers / Stime del volume totale del Bosco per INFC2005 e INFC2015, a livello nazionale e per le singole regioni; barre di errore pari a due volte l'errore standard; le barre per l’Italia sono state ridotte per migliorare la lettura, i valori reali sono indicati in numero

Disaggregating the volume estimates by forest type, it is observed that, for the types of Tall trees forest, the variation is always positive and considerable for the Other coniferous forest (6,821,805 m3, +54.5%), for the Beech (52,532,401 m3, +21.9%), for the Temperate oaks (16,963,720 m3, +21.9%), for the Mediterranean oaks (25,549,227 m3,+24.6%), for the Hornbeam and Hophornbeam (14,731,288 m3, +21.3%), for Other deciduous broadleaved (33,626,655 m3, +35.7%), and for the Holm oak forest (15,601,302 m3, +34.2%). For the other forest types, the variation is still positive but negligible (Fig. 13.6). For the Tall trees forest as a whole, the overall variation is also quantitatively considerable (236,286,163 m3, +18.8%).

Fig. 13.6
figure 6

Estimates of the total growing stock of Tall trees forest types from INFC2005 and INFC2015; error bars are equal to twice the standard error / Stime del volume totale delle categorie forestali dei Boschi alti secondo INFC2005 e INFC2015; barre di errore pari a due volte l'errore standard

Considering only the regions for which a considerable change in growing stock volume has occurred, the increase is largely due to the positive variation observed for the Apennine regions, representing over 41% of the overall increase. The geographical location of the positive volume variation is supported by the observation that it occurs especially in the Beech and Oaks forests, which are very present in these regions, constituting almost 47% of the overall volume increase.

13.3.2 The Aboveground Tree Biomass

The aboveground tree biomass of the Forest, expressed in terms of dry weight, was estimated at 874,443,096 Mg by INFC2005, with a SE equal to 1.0%. According to INFC2015, the same quantity is estimated at 1,044,276,604 Mg, with a SE of 1.0%. The variation between the two surveys is quantitatively considerable and equal to 169,833,508 Mg (+19.4%).

Considering the biomass estimates for the individual regions, it can be observed that always positive and considerable variations occur in Piemonte (12,341,412 Mg, +14.1%), Lombardia (14,265,600 Mg, +20.8%), Veneto (10,749,400 Mg, +21.1%), Toscana (18,499,369 Mg, +19.0%), Umbria (5,146,881 Mg, +20.8%), Lazio (12,354,594 Mg, +26.9%), Abruzzo (9,828,520 Mg, +24.5%), Campania (11,079,508 Mg, +33.2%) and Calabria (16,766,828 Mg, +28.7%), while in the other regions the variations, still positive, are generally negligible (Fig. 13.7), in line with the results observed in the comparisons made for the total growing stock.

Fig. 13.7
figure 7

Estimates of the total aboveground tree biomass of Forest from INFC2005 and INFC2015, nationally and for individual regions; error bars are equal to twice the standard error; bars for Italy were shortened to improve readability; real values are given in numbers / Stime della fitomassa arborea epigea del Bosco secondo INFC2005 e INFC2015, a livello nazionale e per le singole regioni; barre di errore pari a due volte l'errore standard; le barre per l’Italia sono state ridotte per migliorare la lettura, i valori reali sono indicati in numero

Disaggregating the estimates of aboveground tree biomass for the forest types of the Tall trees forest, it is observed that the variation is always positive and considerable for the Other coniferous forest (3,887.323 Mg, +53.7%), for the Beech (39,774,535 Mg, +21.6%), for the Temperate oaks (13,836,869 Mg, +21.3%), for the Mediterranean oaks (20,477,654 Mg, +24.1%), for the Hornbeam and Hophornbeam (11,292,435 Mg, +19.1%), for the Other deciduous broadleaved (24,100,050 Mg, +34.5%) and for the Holm oak forest (15,321,040 Mg, +35.0%). For the Tall trees forest as a whole, the variation is considerable and equal to 19.8%, and this is perfectly in line with what was observed in the comparisons made for the total volumes. For the increase of the aboveground tree biomass, the important contribution provided by the Apennine forests, especially Beech and Oak forests, is confirmed.

13.3.3 The Annual Volume Increment

The annual volume increment of living trees standing in the Forest at the time of the INFC2005 inventory survey was estimated at 35,872,293 m3, with an estimated SE equal to 1.1%. In INFC2015, the same quantity is estimated to be 37,787,784 m3 with a SE of 1.0%. The positive variation between the two estimates is quantitatively considerable and equal to 1,915,491 m3 (+5.3%).

By disaggregating the estimates of the annual volume increment according to the regions, it can be observed that positive and considerable variations occur for Lombardia (673.572 m3, +21.4%), Umbria (208.009 m3, +25.5%), Lazio (372.571 m3, +24.1%), Abruzzo (253.181 m3, +19.2%), Molise (198.589 m3, + 47.0%), Campania (439.695 m3, +28.1%) and Basilicata (330.271 m3, +44.7%). A considerable negative change is observed for Liguria (−413.453 m3, −26.2%) and for Toscana (−603.917 m3, −14.5%). For the other regions the variations, positive and negative, are generally negligible according to the criteria adopted for the comparison (Fig. 13.9).

Disaggregating the same estimates of annual volume increment according to the forest type, it is observed that, for the types relating to the Tall trees forest, the variation is positive and considerable for the Temperate oaks (586,013 m3, +26.5%) and for the Other deciduous broadleaved (618.803 m3, +15.4%), while it is negative and considerable for the Chestnut (−751.900 m3, −15.2%) (Fig. 13.10). For the Tall trees forest as a whole, the variation is positive and considerable (2,219,885 m3, +6.4%).

13.3.4 The Annual Removal of Growing Stock

The growing stock removed through utilisations during the twelve months before the INFC2005 survey was estimated at 13,796,864 m3, with a SE of 12.9%. In INFC2015 the same quantity is estimated to be 9,566,257 m3 with a SE equal to 16.1%. The negative variation between the two estimates is equal to −4,230,607 m3 (−30.7%) and is negligible given the estimates intervals considered. Similarly, observed variations at the regional level and at the level of forest types are quantitatively negligible according to the criteria adopted for the comparison.

13.3.5 The Number of Trees

A further interesting comparison between the estimates of the two national inventories concerns the number of living trees. The total number for the Forest in Italy was estimated by INFC2005 to be 11,949,630,797, with a SE of 1.3%; INFC2015 estimated a total number of 11,481,562,909 trees with a variability of estimate of 1.2%. The variation between the two surveys is therefore equal to - 468,067,888 trees (−3.9%), a negligible variation given the estimate intervals considered.

Positive differences are observed for the regions Valle d’Aosta, Lombardia, Lazio, Abruzzo, Molise and Sardegna, while for the other regions the variations are always negative. In all cases the differences are negligible.

At the level of the forest types of the Tall trees forest, positive variations are observed for the Other coniferous forest (33.2%), for the Temperate oaks (5.4%), for the Other deciduous broadleaved (10.4%), for the Cork oak (3.1%) and for the Other evergreen broadleaved forest, while the differences are always negative for the other forest types. For the type of Chestnut alone there is a considerable variation (−183,976,338 trees, −16.0%), while for all the other forest types the differences are quantitatively negligible.