LOOKING BACKWARDS INTO A MEDITERRANEAN EDGE ENVIRONMENT: LANDSCAPE CHANGES IN EL CONGOST VALLEY (CATALONIA), 1850–2005 ENRIC TELLO*

We examine the evolution from the mid-nineteenth century to the present of the cultural landscape in two Catalan townships that jointly constitute an edge environment between Mediterranean plain and mountain. This environmental edge has changed over time, from a border between an agricultural plain confronting an area of forestry and livestock-raising in the mid-nineteenth century, to a limit of the Barcelona Metropolitan Region confronting two ‘natural’ protected sites in the present. Over a century and a half land use changes in this area have brought about the two main, but superficially ‘opposed’ dynamics that characterise landscape evolution throughout Catalonia and great parts of Europe today: intensification and abandonment. The most salient ecological impact of both processes has been the loss of landscape mosaics. In the plain those mosaics interwove cereal cropping with vineyards and olive orchards together with patches of woodland, scrubland and pastureland into an agricultural land matrix. In the mountains, a woodland matrix of holm oaks and pines became cleared through timber and firewood extraction, and the pruning of branches for charcoal making, creating a mixture of open forest articulated by grazing areas, cropland and scrub. Both sides of the edge interlinked, and jointly played a key role as ecological connectors to maintain biodiversity. From the 1960s onwards the ecological dynamics attendant upon the abandonment of the prevailing integrated management of forests, livestock breeding and cropland has led to a significant loss of these landscape mosaics. In spite of the transformation of much of this area into natural protected sites, the growth of a uniform and continuous reforested woodland canopy, directly confronted with urban sprawl, is endangering its richness in terms of species – such as Mediterranean orchids, whose habitats are disappearing.

Since these landscape mosaics come from human intervention in ecosystems, while at the same time retaining functioning natural processes within them, they become a kind of nature transformed.In order effectively to manage and preserve them we need a historical perspective to understand the magnitude, causes and consequences of these interlinked social and natural processes.Even more, recovering the historical dynamics behind our current land-use patterns becomes a necessity for more encompassing, complex and multidimensional approaches to sustainable landscapes.This landscape history research has to be addressed on different geographical scales and time spans by interdisciplinary teams like the one assembled here -which includes two historians, a biologist, an architect, a geographer and two environmental scientists. 3Given the dual nature of cultural landscapes, both human-made and natural patterns, the loss of agro-forest mosaics also implied a decline of knowledge about managing natural resources in an integrated manner.The increasing apparent homogeneity of these abandoned landscapes is making the relicts of previous ecologies harder to discern (and more easy to degrade) in a society where the local inhabitants have less and less information on, or interaction with, them. 4hind these trends are three major driving forces summarised by Marc Antrop for the whole of Europe: accessibility, urbanisation and globalisation.
Accessibility has to be understood not only as physical transport or cultural connection, but as the increasing purchasing power of goods and services as well.
At the beginning of the twenty-first century, easy market access to natural resources at almost any place in a globalised world has led to some levelling of the standard of living across the metropolitan hinterlands of developed countries, far from the former sharp divides between cities, towns and rural villages.In our case study area the average taxable household income in the two townships that we study, La Garriga and Figaró-Montmany, is near to the average GDP per capita in the fifteen countries of the Eurozone: some €28,400 at 2008 prices (Table 1). 5wever, the situation was very different just fifty years ago, both in comparing the townships with each other, and with the wider European experience -as discussed below.Sources: Taken from IDESCAT (http://www.idescat.cat/en/)and EUROSTAT (http://epp.eurostat.ec.europa.eu/portal/page/portal/eurostat/home/).
In the following sections we will describe how accessibility, the rural-urban divide, metropolitan networks and globalisation combined in different proportions as historical driving forces for land-use changes in the study area from the midnineteenth century to the beginning of the twenty-first century.Throughout this environmental history these local landscapes experienced some long-term continuities, but also several deep discontinuities and big accelerations at specific turning points.The most important of these was, no doubt, the land-use changes fostered by the urbanisation and globalisation experienced during the second half of the twentieth century.Nevertheless, a longer-term perspective helps to identify the agro-ecological roots of the traditional cultural landscape and its environmental services that have been undermined by the fast-growing industrial economy during the era of cheap oil. 6is environmental history of land-use changes has been as complex and multidimensional as cultural landscapes are themselves.The changes range from the substitution of former land cover with newer types, to human perceptions of these, or the senses of identity adopted by the communities who have altered their land usages, consumption baskets and ecological imprints over time.In the rest of this article we will focus on how land cover changes between 1854 and 2005 have affected ecological patterns in these landscapes, and their capacity to host biodiversity. 7IN FEATURES OF OUR CATALAN STUDY AREA

Location
Edge environments show ongoing environmental processes in a particularly dramatic light, partly because they help highlight the existence of earlier, more diverse and integrated landscapes as being characterised by the very prevalence of smaller 'edges' within them. 8The two Catalan townships studied are currently 6 Ch.Pfister, 'The "1950s Syndrome" and the Transition from a Slow-Going to a Rapid Loss of Global Sustainability', in F. Uekoetter (ed.)The Turning Points of Environmental History (Pittsburgh: University of Pittsburgh Press, 2010) pp.90-118. 7 On the complexity of landscapes see Z.For the linkages between changing living standards and their historical ecological imprints, see G. Billen, J. Garnier and S. Barles, 'History of the urban environmental imprint: introduction to a multidisciplinary approach to the long-term relationships between Western cities and their hinterland', Regional Environmental Change 12 (2012): 249-253; and K.P. Donaghy, 'Urban environmental imprints after globalisation', Regional Environmental Change 12 (2012): 395-405. 8We are using in the title and in this introductory section a geographical-wide meaning of 'edge environment', a notion that needs to be distinguished from the ecological 'edge effect' in ecotones seen on a smaller land-cover scale -as discussed below.In the same vein, William Cronon spoke in Changes in the Land: Indians, Colonists and the ecology of New England about the patchwork pattern of the New England pre-colonial landscape as mainly determined by topography (slopes, exposition or aspect, etc.) seen in a geographical scale, and then also about the ecological 'edge effect' looked at on a smaller scale, as in the clearing of forest patches made by the Indians through controlled fires (New York: Hill and Wang, 2003, pp.25-33 and 51).
one of the few North-South axes that for many centuries have connected the coastal area of Barcelona to the Pyrenees.Following this axis along the township of Figaró-Montmany, the river Congost runs North to South through a small gorge between the massif of El Montseny and the cliffs and crags of the Cingles de Bertí.
The southern limit of the gorge ends where the Congost River reaches the neighbouring township of La Garriga, opening to the Vallès plain located between the pre-littoral and littoral mountain ranges of Catalonia.There, it drains its tributary flow to the Besòs River.(Figure 1).

Two paths of population growth
Both municipalities reacted to these market stimuli in very different ways.La Garriga evolved in a similar way to the adjacent Vallès plain.Vineyards spread and mixed with cereal cropland and small woodlands in an agro-forest mosaic that extended from the irrigated orchards surrounding the Congost River up to the forests and pastures in the nearby mountains.In 1858 vineyards occupied a great proportion of cultivated land in Figaró-Montmany too, in fact it was the northern limit of this crop at that time within the province of Barcelona.Nevertheless most of the municipality's steep and elevated land was devoted to woodland, scrubs and pasture. 10 9 The idea of interconnected micro-regional environments is taken from P. Horden and N. Pourcell, The corrupting sea.A study of Mediterranean history (London: Blackwell, 2008).They express the same idea with the term 'microecologies', which is a rather confusing term from a natural science standpoint (as microecology is the study of the interactions between microorganisms and their environment). 10We have studied the vineyard specialization in this Catalan region in X. Cussó, R. Manufacturing also evolved in opposite ways in the two villages.Traditional as well as modern textile industries, and furniture making were developed in La Garriga.In Figaró-Montmany manufacture remained small, hand-powered and tightly linked to the raw materials locally available such as sawmills, cutting clogs and wooden planks, stone slabs for floors and fireplaces, or charcoal making.In 1852 La Garriga had one taxpayer in industrial or commercial activities for every five agricultural ratepayers, compared to one for every nine in Figaró-Montmany.La Garriga's payment of commercial and industrial taxes was ten times higher than its mountain neighbour's, five times on a per capita basis.12Consequently, population and income growth followed diverging paths.In 2010 the number of inhabitants of Figaró-Montmany was only 52 per cent higher than in 1860, whereas in La Garriga increased tenfold (Figure 2).

A long-lasting wealth divide
While La Garriga became a wealthier town that attracted affluent newcomers as residents and investors, Figaró-Montmany remained a poorer rural provider of inhabitants/km 2 ) than in Figaró-Montmany (42.7).Cultivated land occupied 56 per cent of the useful surface of La Garriga, and woodlands, scrublands or pasturelands only 44 per cent.The opposite was true in Figaró-Montmany: in 1858 only 28 per cent was devoted to crops and 72 per cent was woods, scrubs and pastures.In 1852 the total net taxable income coming from agricultural, forestry and livestock breeding was more than four times higher in La Garriga than in Figaró-Montmany, and nearly two times higher in per capita terms.Inequality in the wealth distribution of land, houses and livestock was also higher in La Garriga (with a Gini Index of 0.743) than in Figaró-Montmany (0.601). 13In both places the land was mainly owned by a group of small or medium-size family farms (called masies in Catalan) that leased small plots and hired labour from landless peasants. 1413 The Gini index measures the inequality among values of a frequency distribution of levels of income or wealth, moving between zero (when everyone has an exactly equal income or wealth) and one (that expresses maximal inequality).The 1852 tax lists referred in the previous footnote evaluated the land, houses and livestock assets of every taxpayer through its net taxable income in an average year, excluding other family incomes coming from wages or domestic manufacturing activities.Therefore, these Gini indices reflect agricultural wealth distribution, as explained in E. Tello and M. Badia-Miró, Land-use profiles of agrarian income and land ownership inequality in the province of Barcelona in mid-nineteenth century, Working Paper DT-SEHA n. 11-01.It is worth noting that the lower rural wealth inequality found in Figaró-Montmany compared with La Garriga contradicted the opposite pattern prevailing in the whole province of Barcelona, as shown in the aforementioned working paper. 14As many places in the North-East of Catalonia, the distribution of land ownership that arose from liberal reforms during the nineteenth century consolidated a previously existing structure of medium-size poly-cultural family farms ranging from some five to fifty hectares each -mainly depending on the more intensive or extensive uses of the land inside them.As population grew and commercial ties with the Atlantic economy developed, these medium-sized landowners started to lease small plots of poor sloping soils to a growing stratum of winegrower tenants.This process of market specialization and land-use intensification polarized rural society, which saw lasting social conflict up to the Spanish Civil War.After the repression and impoverishment suffered during the first decades of Franco's dictatorship, rapid economic upswing started in the sixties putting an end to the old dispute about land entitlements and agricultural income distribution.Again a pattern of small and medium-size family farms prevailed, increasingly managed part-time as agriculture became a more marginal economic activity.'Història Agrària 20 (2007): 185, 196.Thus in the mid-nineteenth century the township border between La Garriga and Figaró-Montmany was already an edge environment between two agro-ecological worlds: the intensive and poly-cultural agriculture of the plain of Vallès County; and on the other hand the multiple and integrated use of woodlands, scrub and pastures of the slopes, hills and small valleys along the Catalan pre-littoral mountains.According to the statistics from 1858-1867, the average share of cereal consumption covered by local production was only 32.3 per cent in La Garriga and 54.7 per cent in Figaró-Montmany -while in the whole province of Barcelona the average was 38.6 per cent.The rest was supplied by wheat imports from Barcelona's seaport, in return for exports of wine, hemp, vegetables, cattle, wood, charcoal or snow.Therefore, these two small rural communities were far from an isolated and autarkic world. 15vestock densities and manure flows in a still organic agricultural system Another salient feature of these two contrasting worlds was the livestock densities on each side of the border.With only 6 livestock units of 500 kg of standardized live weight per square kilometre of cropland (LU500/km 2 ), the intensive agricultural system of La Garriga lacked sufficient beasts to obtain through manure all the nutrients needed for the replenishment of cultivated land, as was then often the case on the Vallès Plain.In contrast, thanks to the small proportion of cropped land and the larger herds grazing in pasturelands and forests, in 1858 Figaró-Montmany had a livestock density per unit of cropland as high as 21 LU500/km 2 (Table 2) -although some of these were horses used for hauling carts along the road. 16 15 The cultivation of hemp and vegetables was pointed out in the descriptions of La Garriga township given in V. de Frigola, Relación de los pueblos de que consta el Principado de Cataluña, and P. Madoz, Diccionario geográfico-estadístico-histórico de España y sus posesiones de ultramar, (Madrid: Est.Tipográfico-Literario Universal, Tomo VIII, 1847), 322.In Figaró-Montmany the same Diccionario geográfico-estadístico stressed the forestry and livestock breeding activities in the mountains and the inns in the gorge (Tomo XI, 1848), 506. 16Most of the horses recorded in Figaró-Montmany were probably employed in the post service and in hauling carts along the route between Barcelona, Granollers and Vic towards the Pyrenees.During the preindustrial 'solar' or 'organic' economy the three inns located in the gorge of this township were strategically placed at one-day's journey from Barcelona, and half a day to Vic (see Figure 1) according to V. Frigola, Relación de los pueblos de que consta el Principado de Cataluña, and Anonymous, Itinerario de la mayor parte de caminos y veredas de las cuatro provincias de Barcelona, Tarragona, Lérida y Gerona, con los pueblos y posadas situadas en sus carreras, y las horas que á paso de tropa distan unos de otros (Barcelona: Imprenta de Manuel Texero, 1838).When in 1854 the train reached the city of Granollers, the railways soon started carrying the loaded carts themselves from Barcelona to Granollers in nearly one hour thus allowing them to attain the city of Vic in one day.The resulting relocation of the need for fresh horses may explain the sharp reduction in the subsequent livestock censuses of Figaró-Montmany.The construction of Catalan railways and their impact is explained in P. Pascual, Los caminos de la era industrial: la construcción y financiación de la Red Ferroviaria Catalana (1843-1989) (Barcelona: Edicions Universitat de Barcelona, 1999).
For the relevance of this things in an 'organic' or 'solar' economy, see R.P. Sieferle, The Subterranean forest: energy systems and the Industrial Revolution (Cambridge: The White Horse Press, 2001).

Barcelona.
It is likely that some manure collected in Figaró-Montmany was sold to fertilize the more intensive crops grown in irrigated lands in La Garriga, such as hemp and vegetables.At the same time pasturelands existing in La Garriga were probably leased in exchange for manure, either to livestock coming from neighbouring townships or transhumant sheep.These are good examples of the socio-metabolic flows linking both sides of this edge environment, while both townships participated at the same time in other biophysical fluxes moving along broader commercial routes.

Back-to-nature vacationers at the beginning of the 20 th century
These different but complementary land-use patterns began to change at the end of the nineteenth century when well-to-do urban elites started to look for cooler places to spend their leisure time during summer vacations, such as the foothills of the Montseny Mountains.La Garriga was favoured as a location for second homes by some of these wealthy families because it had thermal waters and several old spa-hotels that also allowed a longer stay in winter.Others preferred the cooler gorge of the Congost River in Figaró-Montmany. 17 These summer vacations also led to a Romantic and nationalistic rediscovery of 'wild nature' in the landscapes that already existed behind the . 19The map of the first Catalan Regional Planning passed in 1932 can be seen in http://cartotecadigital.icc.cat/cdm/singleitem/collection/catalunya/id/1112/rec/5.happened when these landscapes were already experiencing one of the most fundamental shifts in their long environmental history.
The new era of cheap oil in the 1960s saw the diffusion of gas butane cylinders, putting a sudden end to local charcoal making.Lorry transportation allowed the delinking of the small cluster of furniture makers located in La Garriga and L'Ametlla de Vallès from their former supply of planks in Figaró-Montmany.
Industrialization of agriculture during the so-called Green Revolution led to the abandonment of steep slopes where mechanization was most difficult.As happened in many other European regions, new intensive monocultures and the breeding of livestock in industrial feed-lots were adopted in the plains, leading to the abandonment of the former integrated management of cropland with extensive cattle-raising in upland pastures and woods.The agro-ecological complementarities among different land usages, that had created the traditional agro-forest mosaics, were suddenly ended.The mosaics themselves started to disappear with the rural society that had kept them alive. 20king advantage of the easy connections by railway and highway to Barcelona, in the last quarter of the twentieth century many former second residences were turned into permanent homes occupied by commuters.First La Garriga and then Figaró-Montmany became new edge residential areas of the metropolitan region.At the same time the socio-ecological pathologies of rural abandonment started to show their face: wildfires -there was a dangerous one between La Garriga and Figaró-Montmany in 1994 -, the loss of many old springs and paths used by former woodcutters, shepherds and charcoal-makers, and the proliferation of wild boar (often crossed with fast-reproducing pigs) which increasingly became pests. 21

LAND-USE CHANGES IN LA GARRIGA: FROM A POLY-CULTURAL MOSAIC TO A METROPOLITAN EDGE
We have three excellent sources that study landscape changes in La Garriga from the mid-nineteenth century to the present: 1) a cadastral land-use map at  22 The making of these historical cadastral maps in Catalonia has been studied and inventoried by F, Nadal, L. Urteaga and J.I. Muro, El Territori dels geòmetres.The cadastral map of 1854 has been processed with GIS software by Natàlia Valldeperas and Anna Ollés with the funds provided by the Batista i Roca Grant 2010PBR00002.The digital land cover maps of 1956 and 2005 have been provided by the Centre of Forest Ecology Research and Applications (CREAF) at the Autonomous University of Barcelona (http://www.creaf.uab.es/eng/index.htm).To make them comparable, the different categories of land uses adopted in each map have been homogenized in a single common legend following criteria consistent with the subsequent application of landscape ecology metrics (as will be explained in greater detail in a forthcoming article by J. Marull, E. Tello, P. Warde, F. Coll, N. Valldeperas and A. Ollés, 'Recovering the history behind the ecological connectivity between two Unlike the typical distribution provided by other land-use statistics, these matrices not only allow us to know the land cover profiles of the same place at two different points in time.They also let us discover the historical dynamics behind any given land-use changes, as the former land usages that lie behind every type of land cover become apparent.Besides showing which land cover types have increased or decreased from a specific historical period to the next, a matrix of land-use changes reveals how shrinking land cover types were transformed, and from which older types of cover new ones were created.This allows the true complexity of land use change to emerge with unusual exactitude, rather than just relying on descriptions of general trends.Notice that along the second and third rows (marked in grey) of the two following matrixes shown in Table 4, you have all values (in percentage or acreage) of the prevailing land-uses in the former period (1856 in the first matrix or 1956 in the second).Then, in the last two columns on the right side you have the same information for the following time cut (1956 in the first matrix or 2005 in the second).The other cells show the surfaces that changed from the land use described on top of their own column, which corresponds to the start date, to the new land use that appears on the left of their own row corresponding to the end date.Therefore, as in any other matrix, these flows of land-use change are added up from the latter date along the rows, while they are accounted from the former date along the columns.The surfaces that remained unchanged appear in the diagonal (also marked in grey in Table 4).
natural protected sites and a metropolitan edge: The Congost Valley in Catalonia, 1850-2005').The matrices of land-use change have been calculated using these homogenized land-use categories.Another interesting point revealed by these matrices is that land-use changes seem have to be much less unidirectional than expected: reallocations appear in both directions across land uses for almost all possible categories, even where there is a net outcome with a clear direction. 23 seen in Figure 4, the starting point in 1856 was the cultural landscape created by an intensive organic and polycultural agricultural system, combining a significant amount of irrigated orchards on 11.8 per cent of the territory along the Congost River, with a mosaic made out of rain-fed cereals (4.7 per cent), vineyards (26.6 per cent), other tree crops such as olive orchards, almonds or hazelnuts (3 per cent), and pastures (19.5 per cent) that spread over the flattest parts of twothirds of the municipal area.An interesting feature of this case study is that pasturelands were tightly interwoven with arable land, instead of being found with  The biggest landscape turnaround occurred between 1956 and 2005.What stands out most when comparing the second and third maps in Figure 3 is the explosion of built-up areas, whether urban land or infrastructural facilities.During this period +303 ha was paved, or covered with concrete and asphalt: 16 per cent of the whole municipal area and 38 per cent of all land cover changes undertaken in this territory during this period.It is also striking that most of this has probably led to irreversible soil degradation on the most fertile irrigated lands of the township.This draws attention towards the particular dynamics of another type of edge, the one linking built-up land with the countryside.Following the marketdriven logic pointed out by Von Thünen, the village border created around it a ring of irrigated orchards where fresh fruits were grown.A question was going to be posed when the village border needed to be extended due to population and economic growth: should new land developments be extended over the market gardening zone or should that be preserved by building on a more distant location?
Once the riparian vegetation along both sides of the river, long ago transformed into orchards, was built over it could never be reproduced elsewhere.
The choice of building over the gardening zone of a village like La Garriga gave no option of return.This is not an extemporaneous question: it was already raised in Catalonia in the early twentieth century by Cebrià de Montoliu (1873-1923) and other land-use planners who advocated garden-city movement ideas leading to the Regional Planning of Catalonia Act passed in 1932.The failure of these alternative approaches epitomises the lack of an adequate land-use zoning and regional planning in Catalonia both during the long dictatorship of General Franco, and after. 26he second biggest land cover change during the last fifty years is spontaneous reforestation following rural abandonment.Together with urban expansion this means that the traditional agro-forest mosaics only survive in La Garriga as relicts mainly located in the South-East side of the township -perhaps because a small water reservoir allows the irrigation of the area.People owning these remaining fields face the dilemma of whether to continue as peri-urban farmers or take offers from urban land developers to give up their land.As always, decisions will be shaped by public policy and the implementation of regional planning. 27e cultural landscapes of La Garriga experienced a deep socio-ecological transition from a poly-cultural mosaic of 1854, created by highly intensive organic agriculture in a mainly rural but already industrializing village, to a residential town located in 2005 inside the farthest edge of the BMR -now placed face to face with the Montseny Natural Park and the protected site of the Cingles de Bertí.In 1956 its landscape was still a combination of agro-forest mosaics with a growing residential town and a blooming resort area joined to a small industrial cluster of textile manufacturers and furniture makers.Now the future of this community placed in such an edge environment will depend again on how Catalan and European society decides to value environmental services and care for the ecological functioning of the land matrix on which they live.

CHANGE BENEATH APPARENT PERMANENCE: FIGARÓ-MONTMANY
The most prominent feature of the landscape evolution of Figaro-Montmany seems at first to be permanence -exactly the opposite of La Garriga.However, we must take a critical distance from this image of immobility.The overall picture of the land-use pattern in this mountain township is blurred by the ambiguities and shifting criteria of assessment of what was considered pasture, wasteland, scrub or woodland.There were two different aspects behind these problems in categorization.When relying on written sources and maps, historians have to deal with the information that ruling elites drew up about rural society.They always found it more difficult to get information on large and inaccessible forestlands and pastures, as well as the people living on them in scattered settlements, compared with the denser agricultural townships located in the plains. 28The complexity and adaptability of the multipurpose uses of forests integrated with other uncultivated areas was also a great ally in local efforts to escape control and taxation.Thus, it was not only their ability to deceive tax collectors; the land-use categories which the surveyors wanted to assess in a fixed and separate way were always overlapping and shifting their boundaries in a manner incomprehensible to people lacking the local knowledge of the villagers who managed those lands. 29oking at the available cadastral figures, the land devoted to woodland and pasture was always over 70 per cent during the nineteenth century and 90 per cent in the twentieth century.Vineyards were by far the main crop in the cultivated land during the former period, but were reduced to almost nothing after the Phylloxera Plague.Giving up winegrowing further strengthened the forest character of this township (Table 5). 28The inaccuracies and ambiguous meanings of land-use statistics in mainly woodland mountain areas is stressed by A.T. Grove and O. Rackham in The Nature of Mediterranean Europe.As explained in the text, the land-uses taken from the official statistics and maps compiled in a densely populated and intensively cultivated plain of La Garriga become much more trustworthy than the same obtained in the mountains of Figaró-Montmany, mainly covered with a complex interwoven pattern of woodlands and pasturelands managed by a low, scattered and rebel population (see also footnote 33).Unfortunately, the cadastral map of this township made in 1854 has been lost.We have however a very detailed map made in 1949. 30Although this reduces the possibilities for calculating a matrix of land-use changes to the second half of the twentieth century, it is enough to test the lesser land cover reallocation which took place in this municipality.While in La Garriga up to 42 per cent of the area changed land cover from 1956 to 2005, in Figaró-Montmany only 28 per cent did so from 1949 to 2005 (Table 6).In addition to the reduction experienced by all sorts of cropland, and the increase in built-up areas or land devoted to infrastructures, it is worth noting the contraction of the pastureland: from nearly 80 hectares in 1949 to less than 20 hectares in 2005.**** As explained in the text, there are some differences in acreage between this GIS metrics applied to the available land cover maps, and the figures given in Table 6 were taken from land-use historical statistics.They come from the unavoidable margins of error of these land accounts and the diverse criteria used in land-use categories by the original maps.To recover some real changes underlying the seeming immobility, we can examine not only the land cover types but also the values of taxable net incomes recorded in the cadastres.Figure 4 shows a transition from a mixed agro-forestrypastoral village economy in 1861, when most taxes were raised from farming (65 per cent), towards specialization in forestry that by 1941 provided the majority of taxable income (62 per cent) from lumbering and charcoal making.It is likely that mid-nineteenth century non-agricultural incomes were overestimated due the inability of tax collectors to assess livestock breeding and forestry activities.
Nevertheless, the shift is too large not to reflect a real transition towards a mainly forestry economy.This also fits very well with recent historical studies which have stressed that during industrialization and urbanization wood and firewood consumption continued to grow, adding to the increase in fossil fuels consumed instead of being replaced by them.Looking at Tables 5 and 6 we can assume an overall stability in the land devoted to pasture -some of which was probably included under the category of abandoned or denuded lands in the 1949-2005 matrix of changes -or perhaps a slight decline.However it is very difficult to make any definitive statement on that issue due to elusive criteria about what was considered as woodland, scrubland or pasture in each cadastral source or map.Another clue is the relative increase in net income attributed by tax sources to livestock breeding, as shown in Figure 4.This also fits with the impression given by the livestock numbers appearing in the census: they diminished only a little and changed composition when the local economy became more specialized in forestry after the abandonment of vinegrowing from the mid-nineteenth century to the first half of the twentieth century.
Then extensive livestock grazing suddenly collapsed during the 1960s, more or less at the same time as charcoal-making was abandoned.This retreat from traditional multiple-use of woodlands, scrubs and pastures entailed a deep caesura in the landscape leading to reforestation after abandonment. 32 order to understand the environmental impact of this event we need to go beyond the general categories and limitations of land-use statistics, to recover how the integrated uses of land were articulated by the inhabitants.Historical images can help us in this task.We can see in Figure 5 a comparison of two photographs taken circa 1908 and 2012 from the same place in Figaró-Montmany, showing a section of the lateral valley of Can Oliveres.We can see an apparent reforestation today contrasting with the lower woodland coverage at the 32 We omit the data taken from the livestock census kept at the municipal archive from 1861 to 1951-an indicator in itself of the maintenance of this activity -due to its low reliability.They repeated the exact same quantities of each type of cattle over several periods, leading us to suspect tax fraud.This is confirmed for 1951, when the comparison between the individual statements signed by each breeder and the final list sent to the authorities reveals that there were far more beats than officially declared.In these official lists livestock had to be ascribed to the municipality where their owner resided.Leasing of pastures or animals through partnership contracts were common verbal agreements which offered many opportunities for tax evasion.
beginning of the twentieth century, particularly along the slopes of the hills on the left side of the images.Just above the column of smoke coming from the locomotive, the effects of large forest clear-cuttings can be observed circa 1908.
The contrast becomes even more apparent when comparing the hilltops in the background of the images.While in 2012 these are totally wooded, in 1908 most of these lands located between 500 and 600 metres in altitude appeared to be open spaces, probably used as pasturelands.After the abandonment of extensive cattle grazing they became spontaneously reforested.Throughout the topography of the massif, from the Congost River upwards, there was a first level of slopes covered with pines mainly exploited for lumber and firewood -like the clear cuts shown above the locomotive in the picture taken in 33 R. Casellas, Els sots feréstechs (Barcelona: Llibrería de Francisco Puig, 1901), 11, 15, 18 and 20.This novel ('Dark Vales', according to the English translation by Alan Yates edited by Eva Bosch and issued in Dedalus Books in 2014) contrasted the 'untamed' and indomitable nature of the inhabitants of those mountains in a lateral valley of Figaró-Montmany, with the Catholic fervour of their parish priest.The author had lived in this small valley when he was a child, and some of the characters or events included in the narrative evoked real people and historical facts (see J. Castellanos, foreword to Raimon Casellas, Els sots ferèstecs.Barcelona: Laia, 1980).In the 1820s those villagers refused to pay tithes, and they even assaulted the rectory obliging the priest to run away to the Montmany Township.
covered by a peculiar mosaic of croplands, pasturelands, woodland-meadows and woodlands of oaks and holms.Woodsmen regularly removed firewood in these sparse forests where charcoal-makers also pruned their branches for their kilns, usually located in bare or rocky lands as near as possible to a spring in order to better control the fire.
Thanks to this multiple use, the oaks and holms of these woodlands probably became taller and more distant from one another.Sunlight could penetrate the canopy allowing the grassy layer of the land to grow fescues or thatching grasses to be eaten by flocks of sheep and goats.These new types of Mediterranean woodland-meadows increased the spatial heterogeneity of forestlands inside a mosaic with cropland and scrubs, as well as the height and age differentials of the remaining trees.Due to the edge effect generated by this combination of open and wooded mosaics, connected with a network of streams, a wide set of microhabitats was offered to many plants (e.g.orchids) and animalsfrom butterflies to partridges, crested larks, rabbits, hares and their predators. 34is also helps to explain why pasturelands tended to fade out in the landuse statistics and maps drawn up by people with a forestry-oriented mind.When traditional livestock breeding was stopped at the same time as charcoal-burning and firewood collection, the whole system of multiple and integrated use of forestlands and pastures suddenly collapsed.This can be seen by comparing the 1949 or 1956 land-use maps with land covers seen in 2005 (Figure 6).Castellar, while the rest tend to be located in margins of roads or trails that go up towards those crags -which had become a kind of relict linear meadows.We should not view such a great richness of orchid species in these areas only in narrow terms, but also as an indicator of the likely presence of high biodiversity and an environment that deserves to be studied and valued.This edge environment also brings into contact the two main opposite but related dynamics that characterize landscape change throughout Catalonia and great parts of Mediterranean Europe: intensification and abandonment.We find some very global trends in a very site-specific context.Studying land cover changes in small places like this, and their ecological and landscape impacts, may provide useful knowledge to make and implement wiser private choices and better public policies aimed at a more sustainable future not only here, but everywhere.The diversity of traditional agro-forest mosaics offered more habitats to different species and created a greater number of ecotones which, in turn, also provided more opportunities to diverse edge species.Therefore, species richness used to be higher in these patchy landscapes while the populations of each species were lower.Provided that the mobility of all these species and their capacity to interact throughout the land matrix were not prevented by insurmountable barriers -that might reduce and isolate their populations up to the point of endangering them with extinction -, the land cover diversity of traditional agroforestry mosaics could host a greater deal of biodiversity than the more uniform land covers found nowadays. 36The Mediterranean orchids testify to the ecological virtue of a world that no longer exists, whose land-use management relied on an intermediate pattern of ecological disturbance through cultural agro-forest mosaics.No doubt, natural protected sites are important achievements for biological conservation, but as long as they remain isolated and unmanaged, this will not be enough to prevent biodiversity loss.Therefore, just relying on the legacy of land covers inherited from the past and waiting for their 're-naturalization' through abandonment is not enough to preserve biodiversity.The orchids' attempt to survive year after year in Figaró-Montmany, while their microhabitats are fading out, becomes like a silent cry that calls upon us to search for a more sustainable land-use management.The open question is whether we will decipher their cries in a fast changing landscape, while we remain on our urban side of the edge.

Figure 1 .
Figure 1.Location of the study area in Catalonia, the Barcelona Metropolitan Region and the main transport routes along the rivers -showing the railways built in 1900.
Garrabou and E. Tello, 'Social metabolism in an agrarian region of Catalonia (Spain) in 1860-70: flows, energy balance and land use', Ecological Economics 58 (2006): 49 -65; X. Cussó, R. Garrabou, J.R. Olarieta and E. Tello, 'Balances energéticos y usos del suelo en la agricultura catalana: una comparación entre mediados del siglo XIX y finales del siglo XX', Historia Agraria 40 (2006): 471 -500; E. Tello, R. Garrabou and X. Cussó, 'Energy Balance and Land Use: The Making of and Agrarian Landscape from the Vantage Point of Social Metabolism (the Catalan Vallès County in 1860/70)', in M. Agnoletti (ed), The Conservation of Cultural Landscapes, 42-56; R. Garrabou, E. Tello, X. Cussó and M. Badia-Miró, 'Explaining agrarian specialization in an advanced organic economy: The province of While the rural economy of La Garriga became mainly horticultural and agricultural, that of Figaró-Montmany remained mainly oriented to lumbering, firewood extraction, charcoal making and livestock breeding.The neighbouring northern townships also provided snow or ice collected in the Montseny Mountains and stored in ice wells dug in the land to be sold in Barcelona and other nearby cities in order to cool beverages and food.According to the Land-Use Statistics of the Province of Barcelona compiled in 1858 the taxable value-added of one acre of cropland in La Garriga was double that in Figaró-Montmany -the former was 42 per cent above the average in the East Vallès County, the latter was 31 per cent below. 11The nucleated habitat of the village in La Garriga also contrasted with the scattered farms of Figaró-Montmany, where most inhabitants lived in lateral valleys near the most important resources: woods, pastures, watercourses and snow.Only a few houses and three inns on the old royal road clustered by the Congost River.The traditional name of the township was then Montmany de Puiggraciós and Vallcàrquera, after the two main side valleys inhabited by cattle breeders, lumberjacks and charcoal makers.The official name changed in the last quarter of the nineteenth century with the arrival of the railway, taking Figaró from the name of the main inn.Thereafter most new houses clustered around the railway stop, while many isolated farm dwellings became abandoned.

FIGURE 2 .
FIGURE 2. Populations trends in La Garriga and Figaró-Montmany, 1857-2010 Figaró-Montmany photographs taken in 1956 by the United States Army; and 3) a land cover digital map of 2005 created from satellite images.Beyond the unavoidable margins of error incorporated into any land mapping that explain small differences in acreage obtained with GIS metrics, or taken from the historical land-use statistics assessed without maps, the larger picture of land cover obtained from these digital maps are reliable.They are a fundamental basis for any landscape analysis from a historical perspective, as well as the only sources available to perform a GIS cover intersection analysis in order to obtain matrices of land-use changes experienced between two dates in a given territory.22

*
The land cover map of 1949 does not include scrubland.Due to its strong forestry bias, it is likely that areas covered by shrubs and thickets were considered potential woodlands.** This land-use category includes in the land cover map of 1949 a small proportion of flood plains along the Congost river, and its tributary creeks, which in some places were used to plant and grow alders, poplars and planes.*** This land-use category includes in the land cover map of 1949 some amount of 'unproductive areas'.
Valldeperas and Anna Ollés with GIS software out of the digital maps of land-use and land covers of Figaró-Montmany township in 1949 and 2005.Combining Tables 2, 5 and 6 with Figure2we can see that from the midnineteenth to the mid-twentieth century arable land shrunk from some 28 per cent to between 7 or 10 per cent of the territory, and to less than 3 per cent in 2005.It is also notable that the population almost halved from 695 inhabitants in 1860 to only 379 in 1887 -just when the Phylloxera Plague had ravaged all vines.Then, mainly due to the economic impact of incoming vacationers, the local population increased slowly to463 in 1930, 519 in 1940, and 571 in 1960.Under the influence of urban sprawl inside the MBR, population increased again to 624 in 1991 and 1,057 in 2009.As throughout this period around 80 per cent of the land was woodland, and between 8 per cent to 13 per cent scrubs or reforesting lands, it becomes important to disentangle the different non-agricultural land usages.
Figaró-Montmany.It is even likely that charcoal-making reached its peak during the first half of the twentieth century, just before the arrival of butane gas
of Figaró-Montmany drawn in 1949 allows the identification of pasturelands and croplands that extended in a flat corridor that lay between two rows of cliffs in the Cingles de Bertí and Castellar -in spite of its clear bias towards recording forestry uses and its tendency to omit pasturelands, counting them as forests or wastelands.Pastures also appeared scattered in some open and flatter spaces located higher in the Vallcàrquera valley and other parts of the Montseny massif.We can see some of the same land, although in a more blurred way, with the land cover geo-interpretation made from aerial photographs taken by the US army in 1956 -in this case mixing former meadows or grazing areas with abandoned croplands (Figure6).

FIGURE 7 .
FIGURE 7.An example of Mediterranean orchid in the remaining meadows and other open spaces of Figaró-Montmany: the Ophrys catalaunica

FIGURE 8 .
FIGURE 8. Locations of the Mediterranean orchids found in Figaró-Montmany in 2007-2012 The most salient landscape ecological impact of these long-term land use changes has been similar on both sides of the environmental edge: the loss of landscape mosaics.No doubt, these traditional mosaics were different in the plains or in the mountains.In the Vallès plain they interwove cereal cropping with vineyards, olive orchards and other tree crops together with patches of woodland, scrubland and pastureland into an agricultural land matrix.In the mountains, a woodland matrix of holm oaks and pines became cleared and opened up by timber or firewood extraction, as well as by pruning of branches for charcoal making, creating a mixture of sparse forest articulated with open spaces of grazing meadows, croplands and scrubs.Both types of mosaic interlinked, and jointly they played an important role in supporting biodiversity as ecologically functional cultural landscapes.Unfortunately, although a large proportion of it was transformed into natural protected sites, the abandonment of the integrated management of forests, livestock grazing and cropland has led to a significant loss of these cultural mosaics.

Table 1 .
Socioeconomic profiles of the study area at the beginning of the twenty-

TABLE 2 .
Socioeconomic features of the study area in the mid-nineteenth century

figures taken By Enric Tello from the Centre d'Estudis Demogràfics; land uses from the Land-Use Statistics in the Province of Barcelona; taxable agricultural incomes from the Distribution of Personal
backs of Catalan elites and intellectuals.Hiking, climbing and the study of rural life became fashionable activities at this time, not only among some affluent professionals but also proletarian families educated by the working-class athenaeums in the libertarian ideals of emancipation as well -which in Catalonia were increasingly mixed with vegetarian and naturalist undercurrents.Lying only one hour by train from Barcelona, Figaró-Montmany became popular for hiking, climbing and cycling and as a gateway to the Montseny and Cingles de Bertí mountains.18Duringthe first third of the twentieth century investments in second homes opened up new tertiary employment opportunities for local people: as maids, 17 Leisure as such was not a novelty of that time, as pointed out by J.-Ll.Marfany, 'The Invention of Leisure in Early Modern Europe', Past & Present 156 (1997):174 -197.But the family summer vacations in the countryside do became a new urban habit for a wider range of high and middle classes of the time.See J.-Ll.Marfany, La cultura del catalanisme: el nacionalisme català en els seus inicis (Vic: Empúries, 1995), and Marfany, 'El paissatge, el nacionalisme i la Renaixença', Estudi General 13 (1993): 81 -97; and J. Nogué and J. Vicente, 'Landscape and national identity in Catalonia', Political Geography 23 (2004): 113 -132.urbanlaundresses,seamstresses,nannies,gardeners,plumbers and so on.At the same time the cultural mixing with different back-to-nature intellectual and social movements also led to an increasing adoption of the ideas of nature conservation, tightly linked in Catalonia to new proposals for city-gardens, green belts and regional planning.In 1932 the Catalan autonomous government published a Regional Plan of Catalonia which categorised the Montseny massif as an 'area of reserved parks and forests' surrounded on its North-East border by a 'grazing area'.19Alateconversionintonature protected sites18For the connections of the Catalan anarchism with naturism, vegetarianism, nudism and the citygarden movement, see J. Martínez Alier, 'Urban sprawl and ecology in Barcelona', Temes de raco.cat/index.php/Temes/article/view/29437/50910), and E. Masjuan, La ecología humana en el anarquismo ibérico(Barcelona: Icaria 2000)

TABLE 4 .
Matrix of the main land-use changes from 1856 to 1956 and 2005 in La

Area with no land-use change from 1856 to 1956 596.0 id. land-use changes
Calculated by Natàlia Valldeperas and Anna Ollés with GIS software out of the digital maps of land-use and land covers of La Garriga township in1854,   1956 and 2005.If we examine the extent of the land cover reallocations experienced in La Garriga, by adding up all the hectares recorded in the diagonal of the two matrices shown in Table4, we can see that from 1856 to 1956 only 32 per cent of land covers remained unchanged; 58 per cent from 1956 to 2005; and for the whole period from 1854 to 2005 the corresponding figure is 26 per cent.Even if the land-use changes registered are potentially exaggerated a little due to unavoidable mismatches in the land cover intersection with GIS from the cumulative effects of small errors in geo-referring the perimeters of all landscape units from the original cadastral maps, the numbers presented here are nevertheless still so large that we can perceive the real existence of a major turnover in the landscape during a century and a half.Change was registered in 68 per cent of the territory from1856-1956, 42 per cent from 1956-2005, and 74per cent from 1856 to 2005.
manure, either to cattle coming from Figaró-Montmany or to transhumant sheep coming from the Pyrenees, a common practice that would help to fill the 23Using GIS techniques in historical studies is still a novel area of work, so there are few examples with which to compare these data.In the case of Tuscany, Mauro Agnoletti has found 60% of land cover changes from 1832 to 1954 and 29% from 1954 to 2004 in a sample of thirteen study areasrather similar values to our figures after considering the marginal errors that all of them may contain (our own calculation, from the land-use matrices published in 'The Development of Historical and Cultural Evaluation Approach in Landscape Assessment: the Dynamic of Tuscan Landscape between 1832 and 2004', in The Conservation of Cultural Landscapes, 16 and 19).. for rectores de la transformación del territorio', in R. Garrabou and Naredo, J.M. (eds), El paisaje en perspectiva histórica.Formación y transformación del paisaje en el mundo mediterráneo (Zaragoza: SEHA/Prensas Universitarias de Zaragoza, 2008), 233 -258; and Ll.Parcerisas, J. Marull, J. Pino, E. Tello, F. Coll and C. Basnou, 'Land use changes, landscape ecology and their socioeconomic driving forces in the Spanish Mediterranean coast (El Maresme County, 1850-2005)', Environmental Spanish lack of regional planning in Ll.Parcerisas et al., 'Land use changes, landscape ecology and their socioeconomic driving forces in the Spanish Mediterranean coast'.27Thesetrends are local expressions of the global ones examined, for example, in E. Lambin, M.D.A.
Coomes, E. Moran, F. Achard, A. Angelsen, J. Xu and E. Lambin, 'Forest transitions: towards a global understanding of land use change', Global Environmental Change 15 (2005): 23 -31.These trends are also studied in the BMR by B. Catalán, D. Saurí and P. Serra, 'Urban Sprawl in the Mediterranean?Patterns of growth and change in the Barcelona Metropolitan Region 1993-2000', Landscape and Urban Planning 85 (2008): 174 -184; and J. Marull et al., 'Social metabolism, landscape change and land-use planning in the Barcelona Metropolitan Region'.

TABLE 5 .
Main land uses in Figaró-Montmany according to the available statistics , the traditional mosaic once maintained by multiple uses has left a lasting imprint on this cultural landscape, to which we can add another clue from modern fieldwork: the location of orchids that can be used as a landscape bioindicator of the loss of meadows and open grazing areas.Here we make use of evidence not usually available to historians, but that also highlights very directly the ecological consequences of the processes driving shifts in land use.European orchids cannot grow under a closed canopy, but need sunlight.They cannot live inside a thick forest -except Limodorum abortivum, Cephalanthera rubra and, sometimes, species of Epipactis.On the other hand, to germinate their seeds they need a mycorrhiza which it is not present in ploughed soils.Finally, orchids are perennial plants that cannot grow in arable land.Thus Mediterranean orchids are mainly present in uncultivated open spaces, making them an interesting bio- human-induced factors', Journal of Biogeography 31 (2004): 905 -915; C. Stefanescu, J. Peñuelas and I. Filella, 'Butterflies highlight the conservation value of hay meadows highly threatened by land-use changes in a protected Mediterranean area', Biological Conservation 126 (2005): 234 -246; and I. Otero, M. Boada, A. Badia, E. Pla, J. Vayreda, S. Sabaté, C.A. Gracia and J. Peñuelas, 'Loss of water availability and stream biodiversity under land abandonment and climate change in a Mediterranean catchment (Olzinelles, NE Spain)'.Nevertheless 37 36 In a forthcoming article we are going to use landscape ecology metrics to undertake a deeper evaluation of the ecological impact of this loss of landscape mosaics (see footnote 22).A state-ofthe-art on the ecological 'edge effect' of patchy landscape mosaics can be found in C. Murcia, 'Edge Effects in fragmented forests: implications for conservation', Trends in Ecology and Evolution 10 (1995): 58 -62; and more recently, in K.A. Harper, S.E.MacDonald, Ph.J. Burton, J. Chen, K.D. Brosofsfe, S.C. Saunders, E.S. Euskirchen, D. Robert, M.S. Jaiteh and P.-A.Esseen, 'Edge Influence on Forest Structure and Composition in Fragmented Landscapes', Conservation Biology 19 (2005): 768 -782.From an agricultural approach, see D. Gabriel, I. Roschewitz, T. Tscharntke and C. Thies, 'Beta diversity at different spatial scales: plant communities in organic and conventional agriculture', Ecological Applications 16 (2006): 2011 -2021.In our Catalan context see M. Guirado, J. Pino and F. Rodà, 'Understorey plant species richness and composition in metropolitan forest archipelagos: effects of forest size, adjacent land use and distance to the edge', Global Ecology and Biogeography 15 (2006): 50 -62.