Introduction

Many factors that result from global influences indirectly affect land management at a local level (Raskin 2005). These influences include resource production and consumption patterns, economic globalization, cultural differentiation, international governance, geopolitical conflicts, migration of populations, and global warming. Economic influences, demographic factors, and government policies act as driving forces that define the most appropriate land-use strategies for any region based on the particular interests of those responsible for the change (Braimoh and Vlek 2005), without taking into account the opinion or needs of local populations. However, in practice, nearly all environmental management is carried out at a local level, and measures adopted at this level influence the impacts at broader scales; for example, they play a significant role in biodiversity conservation and in establishing effective procedures to achieve sustainable development (Gascó and others 2005; Maestre 2001).

Environmental management strategies affect landscapes because land-use and land-cover change (LUCC) is one of the most important human modifications of the Earth’s surface (Lambin and others 1999). The rate of land-cover alteration is increasing dramatically worldwide due to increasing and intensifying human use of the land (Antrop 2004; Conway and Lathrop 2005; Dale and others 2000; Richards 1990). The effects of such changes have attracted the attention of ecologists, particularly with regard to the impacts of modifications of landscape structure, environmental quality, and biodiversity (Alés and others 1992; Holling 1992; Krummel and others 1987; Levin 1992; Romero-Calcerrada and Perry 2004).

These changes have also been analyzed in semi-arid and dry Mediterranean areas, concentrating especially on the consequences of farmland abandonment (Bonet 2004; Bonet and others 2001, 2004; Martinez-Fernández and others 1995) and reforestation carried out at different scales (Chirino 2003; Chirino and others 2001, 2006; Maestre and Cortina 2004). In these environments, where water scarcity is the main factor that limits ecosystem functioning, any change in land use will modify the relationships between the different components of the hydrological cycle, thereby limiting the availability of water as a resource for human development due to increased runoff and decreased recharge of aquifers (Bellot and others 1999, 2001; Cerdá 1995; Chirino and others 2003, 2006). Cause–effect relationships that describe the impact of landscape changes on the conservation of water resources have been determined (Martinez-Fernández and others 1995; Omerick 1977).

For decades, agriculture and tourism have been the main economic sectors in such Mediterranean environments, and this is also the case in Marina Baixa County (Alicante province, southeastern Spain), where noticeable changes have occurred in both the landscape and the regional water balance (Bellot and others 2005; Sánchez and others 2004). However, the environmental impacts linked to these developments have rarely been appraised (Oxley and others 2004). Recently, some authors have focused on the effects of woodland transformation into other land uses and of agricultural land-cover changes that have resulted in an increase in the area covered by an impermeable surface that affects the catchment’s hydrology (Wissmar and others 2004). Rural areas where woodlands have been replaced by sealed soil surfaces (e.g., urban developments, roads, parking lots, rooftops) or by irrigation agriculture experience changes in the hydrologic fluxes, with increased surface runoff during storms, variable water retention times, increased water consumption, decreased soil water infiltration, and decreased recharge of aquifers. All of these changes affect the reversibility of the land-cover changes and the sustainability of uses of the water resource.

The work described in the present article aimed to evaluate human impacts on the water balance of Marina Baixa County that resulted from changes in land use and land cover over the last half-century. The different management strategies implemented in several municipalities of the county are compared during three time periods (predevelopment, until 1956; a period of agricultural development, from 1956 to 1978; and a period of tourism expansion, from 1978 to 2000) to evaluate the human impacts on land cover and water use. Water balance is used as a key indicator of the study area’s environmental evolution and is analyzed to assess the sustainability of these management strategies at local, county, and regional scales.

Study Area

Marina Baixa County is a 671-km2 watershed located in the northeastern part of Alicante province (Figure 1). The county lies between 0°00′ and 0°24′ west longitude, and 38°28′ and 38°43′ north latitude. The topography varies greatly from the coast to the inland mountains and is characterized by intensive land use, with irrigated crops such as medlars (Eriobotrya japonica Lindl.), citrus (Citrus spp.), and other fruits, coexisting with dryland crops such as carob trees (Ceratonia siliqua L.), olive trees (Olea europaea L.), and almond trees (Prunus dulcis (Miller) D.A. Webb), and interspersed with scattered settlements and a rough area of Mediterranean woodland that is frequently affected by wildfire. The climate is characterized by an average temperature of 18°C and an average annual precipitation of 300 mm along the coast, in contrast with an average temperature of 9°C and an average rainfall of between 700 and 900 mm farther inland. The region has undergone one of the greatest socioeconomic changes in Spain because it hosts more than 60% of the tourist activity in Spain’s Valencia region. Marina Baixa County comprises 18 municipalities, which, for the study period (1956–2000), present a mosaic of different trends due to the influence of their proximity to the Mediterranean coast (services and tourism) and the availability of water in aquifer reserves (for intensive irrigation of crops).

Fig. 1.
figure 1

The location of Spain’s Marina Baixa County and the municipalities chosen as study sites

Full size image

In order to illustrate human impacts on the landscape, three patterns of land-cover and land-use change were analyzed in two municipalities with different populations but similar areas (Benidorm, with 51,873 inhabitants and 3860 ha, and Callosa d’en Sarrià, with 7057 inhabitants and 3430 ha) and two municipalities with comparable surface area but much smaller populations that we have treated as a single municipality in our analysis (Guadalest and Beniardà, with a total of 341 inhabitants and covering 3159 ha). Benidorm is one of Europe’s major Mediterranean coastal resorts, attracting more than 3 million tourists each year. Hotels, residential complexes, housing developments, and scattered guest houses and other accommodations provide nearly 20 million overnight stays per year during the summer months, with peak occupancy periods having more than 300,000 visitors/day (Ivars and Juan 1998). Guadalest and Beniardà, located inland, have followed a completely different land-use strategy in which natural vegetation and croplands have been the main land uses affected by anthropic factors and natural hazards (e.g., wildfires), whose impacts have been magnified by human interventions (e.g., cropland abandonment). The Guadalest-Beniardà area welcomes more than 2 million visitors each year, most of whom come from coastal settlements on day trips to visit the village’s castle and the surrounding countryside. In contrast, the agricultural policy adopted by Callosa has been to develop a highly specialized cropland irrigation system devoted to the production of citrus and medlar fruits for export to European markets.

Methods

Land cover and land use in the Marina Baixa county in 1956, 1978, and 2000 were mapped using aerial photographs and image-processing techniques. Stereo-pairs of photographs of the land conditions in 1956 were provided by the Ministry of Defense at a scale of 1:33,000; IRIDA (Spain’s Ministry of Agriculture) provided the 1978 images at a scale of 1:18,000 and the Diputación de Alicante provided the 2000 images at a scale of 1:25,000. Through the use of advanced scanning technologies, the aerial images were digitized and imported into the ArcGIS 8.3 software to obtain land-cover maps in polygonal vector format. All images were processed at a scale of 1:10,000 to facilitate comparisons (Peña and others 2005).

A classification system for the main land-cover and land-use categories was developed based on physiognomy (determined by means supervised classification and photointerpretation), using the nomenclature of the CORINE (Bossard and others 2000) and the LUCC projects (Lambin and Geist 2001). This produced the 30 level-3 categories shown in Table 1. These categories were grouped into 12 level-2 categories and 5 level-1 categories to represent the changes in land use and land cover because this approach was found to be far more appropriate for analyzing the main changes that had taken place at local and regional levels (Peña and others 2005). The resulting 1956, 1978, and 2000 land-use and land-cover maps were used to study the spatiotemporal dynamics of land use and land cover.

Table 1. Classification of the main land-cover and land-use categories based on the nomenclature of the CORINE (Bossard and others 2000) and LUCC (Lambin and Geist 2001) projects, which distinguish 30 categories at level 3, 12 composite categories at level 2, and 5 composite categories at level 1
Full size table

Population increases (Figure 2), increased tourism, and the settlement of expatriates during the last half of the 20th century brought important changes in land use, particularly in Benidorm and Callosa, where much of the land was developed. New settlements appeared throughout the region, large areas of woodland were cleared, and new crops were introduced. The impacts of these changes on the area’s environmental quality were evaluated in the present study using a transition matrix approach in which structural complexity and environmental stability (Heylighen 1999) as well as the system’s water sustainability were taken into account. A simple, nonspatially explicit model analogous to the first-order Markov chain transition method was generated for each combination of map pairs (1956 and 1978, 1978 and 2000, and 1956 and 2000) based on the methods of Usher (1992) and Dale and others (2002).

Fig. 2.
figure 2

Population changes during the 20th century in the municipalities in the study area. Data were provided by Spain’s National Institute for Statistics (Instituto Nacional de Estadística)

Full size image

Transitions that decreased water use sustainability during land-use and land-cover changes (Kates and Parris 2003; Parris and Kates 2003) and that increased the degree of irreversibility of those changes were classified as “degradative” (D). Changes that increased environmental quality or that were not irreversible were classified as an “improvement” (I) transition. Finally, we created the “stationary” (S) category to cover transitions between land uses within the same category or a lack of change. Table 2 presents the types of change between categories for the five level-1 aggregated categories (Table 1), giving a total of 5 × 5 = 25 possible transitions. In the context of this study, any transformation of woodland or shrubland into any other use was classified as degradative, whereas any transformation of urban land into a more natural environment was classified as an improvement. Similarly, the change from dryland farming to irrigated crops was also considered degradative due to the resulting increase in water consumption and evaporation, whereas a change from irrigation to dryland farming was considered to be an amelioration.

Table 2. Matrix of the transitions between the five level-1 categories of land use (Table 1) and the corresponding consequences for sustainability of water use
Full size table

To calculate the water balances for each type of land use and cover, a Digital Elevation Model of Marina Baixa County and precipitation data were used to estimate the quantity of water flowing into the Marina Baixa catchment during the study years (1956, 1978, and 2000). Actual evapotranspiration was calculated using different methods for each type of land use and averaged weighting each surface. In natural cover types, evapotranspiration was estimated using the “Ventos” hydrologic model (Bellot and others 2001; Chirino 2003). In dryland agriculture and irrigated cropland, evapotranspiration was estimated using the water consumption data (López Bellido 1998) and the water use efficiency for each crop reported by Allen and others (1998) as input data for the PARloc software (IVIA 2006). The extent and buildings density of urban areas and of scattered settlements were calculated using the land-use data in ArcGIS, supplemented by field data, and were divided into three categories (low, middle, and high) according to the number of buildings per hectare. Field surveys using the rapid appraisal technique (Kachondham 1992) and data provided by local water supply companies and agencies were used to estimate urban water consumption and to evaluate the impact of these uses on the water cycle. For gardens and urban parks, the method of Domene and Saurí (2003) was applied to estimate water consumption and evapotranspiration.

Results

Land-Cover Properties

The patterns of change differed greatly among the municipalities (Table 3), although the predevelopment situations were similar (56–59% natural vegetation, 39–42% agriculture, and 0.1–2% urban). The proportion of the area covered by natural vegetation, which comprises different types of woodland cover (pine forest, dense or open shrubland, and abandoned cropland) decreased by 4.2% during the study period (1956–2000) in Callosa and by 3.2% in Benidorm, with the proportion of the area devoted to agriculture increasing by 1% in Callosa and the proportion of urban area increasing by 33.6% in Benidorm. However, in the Guadalest–Beniardà area, the proportion of the area covered by natural vegetation increased by 16.5%, with the exception of pine forest, which decreased from 1978 to 2000 as a result of forest fires. As a result, the dominant natural vegetation in Guadalest–Beniardà became shrubland and scrubland (35.1% in total).

Table 3. Main land uses in 1956, 1978, and 2000 for the municipalities in the present study
Full size table

Since 1978, agriculture has declined steadily in Guadalest-Beniardà (by 10%) and Benidorm (16%), in contrast with the 177-ha (4.9%) increase in irrigation farming, especially under shade and using drip irrigation technology, in Callosa. Dryland farming decreased considerably (by 23%) in Callosa, to a level well below that in Guadalest-Beniardà and comparable to that in Benidorm (Table 3).

There has been almost no change in the proportion of urban area in Guadalest-Beniardà since 1956, whereas the urban area in Callosa has increased to 10 times its 1956 value; in Benidorm, the urban area increased to 18 times its 1956 value (Table 3). Although this category occupies less land area in total than natural vegetation and agriculture, this change has important consequences for the conservation of water resources because urban areas have impermeable surfaces and because the change has a high degree of irreversibility. The majority of the urban development took place between 1978 and 2000 as a result of the emergence of industrial, commercial, and (above all) recreational and leisure areas. Access to these developed areas has been improved by increases in the county road network, which has expanded to 5 and 11 times the 1956 values in 1978 and 2000, respectively (data not shown). In 2000, road networks in Benidorm occupied more than 120 ha of land (Peña and others 2005).

Land-Cover Transitions

The analysis of land-cover transitions included all three municipal areas shown in Table 3. The changes observed from 1956 to 2000 (Figure 3) confirm that most of the Guadalest-Beniardá landscape remained stable (52% of the transitions) and had the lowest level of degradative transitions (13%). Callosa and Benidorm had roughly the same level of degradative change (35 and 38%, respectively), but Callosa also showed a high degree of stability (43%). Benidorm went through the largest number of transitions that were characterized by a high level of degradative change (38%). Abandonment of farming appears to have contributed most strongly to the improved transitions (31%) that have taken place in Benidorm.

Fig. 3.
figure 3

Distribution of land-use transitions in the municipalities in the study from 1956 to 1978, from 1978 to 2000, and for the study period as a whole (1956–2000)

Full size image

Impact of Land-Cover Changes on Water Use

Equipment and infrastructure for transporting water have become a prominent feature of the study area, adversely affecting its natural appearance. In 1978, the dam constructed to regulate Guadalest’s river flow came into service, and although the same area of bodies of water was maintained, downstream of the dam the river and its associated vegetation has disappeared. However, the sporadic presence of water has permitted the development of ravine (rambla) vegetation, leading to the replacement of a river system with a high ecological value by another (rambla) and increasing the connectivity between land on opposite sides of the dry riverbed.

The main changes in the water-use system can be attributed to the increase in the area of irrigated cropland (e.g., irrigation reservoirs) and the area of urban development for recreational use (swimming pools, golf courses, water parks, etc.). These changes demand a great amount of water from the region’s aquifers or the availability of a water supply from outside of the region. We have estimated the average water flows (m3/ha/ year) for each land use (Figure 4) based on an average annual precipitation of 400 mm (4000 m3/ha) for the whole county. In the natural vegetation and dryland crop land uses, evapotranspiration represents the main output (87% and 79% of inputs, respectively), with infiltration representing the secondary output (10% and 20%, respectively). In areas under irrigation, the additional input of water greatly modifies the water flows, with infiltration rates increasing to 34% of the inputs and evapotranspiration increasing to 63% of the total water input (rainfall and irrigation combined). In urban areas, evapotranspiration decreased to 3% of inputs (rainfall and supplied water), but urban collectors (“treatment plants” in Figure 4) receive 70% of the water, with the remainder (27%) being accounted for by infiltration or lost by the system.

Fig. 4.
figure 4

Average water flow (m3/ha/year) for each of the main land uses in Marina Baixa. For each land use, the inputs are precipitation (Prec.), irrigation, and urban supply; the outputs are evapotranspiration (Eta), infiltration into the aquifer (Infil.), and surface runoff (Runoff). Inputs and outputs are represented by downward-pointing and upward-pointing arrows, respectively. Percentages represent the proportion of total inputs (precipitation, irrigation, and urban supply volumes)

Full size image

We calculated the annual water balance in millions of cubic meters (Mm3) for the hydrological areas occupied by the four municipalities (Table 4). The changes in land use between 1956 and 2000 increased water consumption to levels exceeding the available water resource in Callosa and Benidorm. In both cases, agricultural and urban uses produced the same effect: a net negative balance of more than 3.2 Mm3/year. However, if water collection by the municipalities of Guadalest and Beniardà is included, the overall balance remains near equilibrium in 2000 (Table 4). Considering the four municipalities as a single system, the balance between collection and use of water suggests that the relative sizes of the regions that contribute and consume water affects the overall water balance and thus the system’s sustainability, with some areas acting as sources and others as sinks. By analyzing the contribution of water flows and reserves (aquifers) in these municipalities (Table 4), we observed that only the Guadalest-Beniardà community is still a source of water supply, with positive contributions to the net balance in all three periods under consideration.

Table 4. Annual water flows (in Mm3/year): inputs (rainfall), outputs (evapotranspiration), consumption (irrigation and urban), and total balances at each local level
Full size table

The role of Callosa in the county’s water balance has changed over time, with a positive contribution during the predevelopment period changing to a strong negative contribution from 1978 to the present as a result of intensive development of irrigated agriculture. Benidorm has acted as a water sink since 1978 due to the strategies implemented at that time for the development of tourism. This net negative balance has occurred despite the implementation of new sewage treatment measures in 2000 that now return 38% of the input water to the Benidorm aquifer. In environmental management terms, Benidorm has lost nearly all of its former agricultural character (only 9% of the area is now agricultural; Table 3), and natural vegetation has also declined slightly; as a result, 35% of its area is now occupied by developed land, including infrastructure for water transportation. Part of the water that is used is now treated and restored to a polluted and saline aquifer, whose water is unsuitable for human consumption or agricultural uses (Figure 4).

Discussion

A global view of the situation in 1956 reflects a more rational use of the land and a greater sustainability of water consumption. In that year, Benidorm, Callosa, and Guadalest–Beniardà had a similar distribution of land uses, but in recent decades, land use has irreversibly transformed much of the traditional structure of the rural landscape, fragmenting the agricultural landscape and causing overexploitation of the hydrological resources. The fact that there was less irrigated agriculture in Callosa and less tourism in Benidorm during the predevelopment period resulted in a near-equilibrium water balance or a net positive balance. However, when these areas began to flourish economically and demographically (1978 and 2000), water consumption increased greatly as a result of taking advantage of new water sources and increasing the complexity of the systems used for water exploitation (wells, transfers from other regions, dams). This increase has been mitigated in part by the implementation of new technologies aimed at reducing water consumption, such as the implementation of drip irrigation, water treatment plants, and the reuse of treated sewage water.

The land-use change in favor of housing developments and irrigated croplands illustrates the impact of local decisions on land use and hydrology. The case of Marina Baixa County also illustrates how topographical conditions and distance from the coast significantly affected land-use and land-cover types and therefore affected the landscape configuration, as was the case in Wear and Bolstad’s (1998) study, which described a strategy limiting the use of intensive agriculture to areas of no tourism interest. Road construction and improvement have also affected the subsequent evolution of rural landscapes in our study area, which shifted from a matrix of natural vegetation in 1956 to a primarily anthropogenic matrix in 2000. The environmental impacts of these changes must be evaluated further, paying special attention to their degradative characteristics and the degree of irreversibility of the changes, because even reversible changes will persist for a long time, simultaneously affecting the water cycle and the functioning of the region’s ecosystems.

Additional research is needed to evaluate the ecological implications of continuous construction of housing developments, overexploitation of water resources, and other impacts attributable to human actions. These alterations influence local environments by destroying habitats and modifying water fluxes (Alberti 1999), and it will be necessary to further evaluate these effects from an environmental point of view—that is, from the perspective of biodiversity, functioning of ecosystems, risk assessment, and so on (Dale and others 1998a). Currently, other impacts (both negative and positive) have been observed as a result of the increased volumes of water flowing over the surface in 2000, which contrasts with the predevelopment period (1956), when this volume of water normally remained underground in the aquifers. The expansion of housing developments and the presence of scattered settlements throughout the landscape could be considered as a totally anthropogenic disturbance in the sense defined by Dale and others (1998b), and this might eventually result in an urban region with no positive environmental function (Antrop 2004).

At present, the water balance in our study area is roughly in equilibrium at the county scale. The municipalities of Guadalest and Beniardà, which have implemented a different management strategy based on retaining more of the natural and agricultural environments, can supply Benidorm and Callosa with the extra water they have collected from rainfall in order to compensate for the water deficit in these two larger villages. However, the current trends in land use suggest that an increased water deficit will occur at the county level in the future. Published strategic plans for development of the region suggest that the severity of human impacts will increase on this area (increased population, housing development, and development of recreational sites such as water parks and golf courses), and these increased impacts can be compensated for only by new water transfers from outside of the region (Gascó and others 2005). This management policy will, in turn, affect the sustainability of other remote areas, which will similarly need to reassess their development strategies. Alternatively, water conservation measures and inputs of marine water using desalination technologies could be developed.

To minimize the adverse effects of development, planners and decision-makers must balance the current availability of natural resources with proposed new changes in land use before permitting further landscape changes. As Cumming and others (2005) have proposed, it is necessary to consider the present ecological limitations before defining additional scenarios for sustainable development in which land uses will be in equilibrium with the available resources. Conversely, to create a “huge urban development” would require an input of resources from other regions that would remain undeveloped, thereby intensifying the existing regional imbalance in land use and economic development. Planning and designing landscapes for the future requires a profound understanding of the highly dynamic, complex, and multifunctional urbanized landscape (Antrop 2004) and of ecologically appropriate future development. An environmentally correct management policy must recognize that local and regional sustainability depends on ecological issues that arise at the ecosystem level. Unfortunately, reality provides us with many examples where the sustainability of a local area is maintained only by requiring the contribution of resources from other regional, national, or international areas. Our case study clearly illustrates how an inappropriately planned coastal urban development (in the municipality of Benidorm) has led to occupation of more than one-third of its area by houses, roads, and other infrastructure and services, increasing water consumption to unsustainable levels.