Using high-resolution multitemporal imagery to highlight severe land management changes in Mediterranean vineyards
Introduction
Landscape organization, in terms of land uses and management, is the result of centuries-old interaction between man and the environment (Agnoletti et al., 2011, Lieskovský et al., 2014). By definition, the land use encompasses the intentional role of people to adapt the natural land cover to their benefit (Verheye, 2004) whereas land management defines the way in which a given land use is administered by people (Foley et al., 2005). Among the variety of landscapes on Earth, vineyards are excellent case studies, as they represent a strong cultural legacy and support a crucial socioeconomic sector in European countries (Salome et al., 2014). Historical maps of vineyards were widely studied because they have faced several crises (Lieskovský et al., 2013, Lieskovský et al., 2015, Lieskovský et al., 2017, Pazúr et al., 2014, Incze and Novák, 2016). In Europe, the phylloxera aphid caused a total collapse of vineyard production in 1860s, until the grafting onto American vines finally succeeded. In France, the introduction of the Common Market in 1970, which allowed the Italians and Spanish to invest in the French interior market, caused a decrease in the growing of table grapes (Galet, 2008) in conjunction with a decrease in pesticides used in agriculture. In 1990, the arrival of vines “du nouveau monde” from the United States, Australia and Chile influenced vine production (Touzard & Laporte, 1998). As the global drivers affecting evolution of vineyards are well known, their monitoring at landscape scale to understand their driving forces is an essential step still under investigation (Houet, Verburg, & Loveland, 2010).
The driving forces shaping landscapes are both socio-economical and physical. Landscape consolidation is driven by political forces because it helps making large fields more suitable for industrial agriculture (Lieskovský et al., 2014). Landscape composition in land uses is also affected by social forces and particularly the land abandonment (Incze & Novák, 2016). Physical forces drived also landscape architecture and composition in land uses. For instance, geomorphological variables drived land use changes such as slope that affects field accessibility with machinery (Sluiter and Jong, 2006, Lieskovský et al., 2017). Altitude gradient can also affect the loss of natural areas in favour of agriculture expansion (Tovar, Seijmonsbergen, & Duivenvoorden, 2013). The pedological factors are also important to understand land settlement on fertile soils for example (Lieskovský et al., 2015, Lieskovský et al., 2014). As explained below, land management is also affected by socio-economical and physical forces.
Historical evolution of vineyards in regards to other land uses have been largely analyzed in Europe, but a change in land management of the vineyards is still lacking. Tracking subtle changes in land management is sometimes more difficult than for land uses because it necessitates a resolution of the historical maps higher or equal to the metre to distinguish between vineyard types (Delenne et al., 2008). In the same way, changes in land management acted at a smaller scale than for changes in land uses. For example, weeding practices in vineyards were mainly governed by local government areas (Biarnes, Bailly, & Boissieux, 2009). As scales and stakeholders are often correlated (Hein, van Koppen, de Groot, & van Ierland, 2006), monitoring the evolution of vineyard management help us explain the role of local stakeholders in a given area. There is now a large number of available images (Wulder, Masek, Cohen, Loveland, & Woodcock, 2012) to create diachronic maps but they have not always a sufficient resolution to detect changes in land management at field scale whereas the field scale is the main unit on which most decisions on the land management operate in agricultural landscapes (Thenail et al., 2009).
A large amount of research has been conducted on the study of land cover in various fields from LANDSAT images: urban expansion (Chen et al., 2006, Qin et al., 2006, Yuan et al., 2005), urbanization of agricultural zones (Liu et al., 2003, Stefanov et al., 2001, Yang and Lo, 2002), forest disturbance (Gibbs et al., 2010, Huang et al., 2010, Mertens and Lambin, 2000, Tovar et al., 2013), desertification monitoring (Dawelbait and Morari, 2012, Symeonakis et al., 2007) and several ecological applications (Cohen & Goward, 2004). However, image classification for determining land cover categories is limited by satellite resolution (higher than 30 m) for these series, and changes in land use or land management operation at the field scale cannot be evaluated by LANDSAT images.
The texture of the field entity could provide information about the type of land use (Trias-Sanz, 2006) and even the land management of vineyards (Delenne et al., 2008, Wassenaar et al., 2002). Furthermore, spatial segmentation is driven by the raster dataset, which is composed of regular squared cells from one to several decametres, and is different from the irregular shapes that form the field entities, the latter being studied using object-based image analysis (OBIA).
By extracting valuable information from various satellite images at different spatial resolutions, from 50 cm to several metres, OBIA is well suited for performing land change studies at the field scale (Benz et al., 2004, Blaschke, 2010). A large number of studies has been devoted to ecological applications (Xie et al., 2008, Yu et al., 2006 for example) or urban features (for a review, see Blaschke (2010)), but only few OBIA studies have been conducted at the field level (Karakizi and Karantzalos, 2015, Qiu et al., 2014, Zhou and Troy, 2008), and never during long time periods.
The difficulty in studying land use changes at the field scale for large spatial and temporal extents is mainly driven by dataset resolution. The availability of remotely sensed data over long time periods (Wulder et al., 2012) and the constant improvements of photogrammetric and digitalization tools for such spatial database analyses make these studies possible (Grekousis, Mountrakis, & Kavouras, 2016). Arnaud et al. (2015) realized historical geomorphic analysis over 80 years in France thanks to the release of aerial photographs from the Institut Geographique National (IGN), for instance.
The study reported land use changes at the field scale over a time period of 50 years using aerial images of Mediterranean vineyards. We hypothesize that a high-resolution database could provide information about changes in land management of vineyards, providing that the time period encompasses the wine crises periods with sufficient temporal resolution. We aim to determine changes in land use categories at the field level using both field number and field area.
Section snippets
Composition and configuration of the dataset at the field scale
The study area corresponds to the Peyne watershed, which covers approximately 76 km2 and is located in Languedoc-Roussillon in southern France (43° 35′N, 3° 19′E). The smaller enclosed watershed is called the Bourdic subwatershed and covers 7 km2. Both the Peyne and Bourdic watersheds were used for further analyses depending on the temporal richness of the dataset. The area of the Peyne watershed is mostly covered by perennial crops (mainly vineyards), and five towns are present in the zone (
Evolution of landscape architecture at field scale
The number of field entities in the entire area started at 15000 in 1962 but decreased to 8400 in 2012. The median area of field entities increased from 2600 m2 in 1962–4500 m2 in 2012. The highest evolution of field numbers was found during the period of 1970–1990 (data not shown). When considering cultivated areas, other land areas, goblet vines and arboriculture exhibited a median area that was approximately half the size of the median area of trellised, transition vines fields (5500 m2).
Analysis of transitions between land use categories
Transitions between land use categories at the field scale are explained by regional history
The transition between land use categories over time, as determined using Sankey and chord diagrams, revealed the history of the zone during the past century, driven by a profound transformation of vineyards and an expansion of urbanized areas, with the introduction of new land uses, such as arboriculture and annual crops (Galet, 2008, Sluiter and Jong, 2006).
Land reparcelling from 1970 to 1990 strongly influenced the evolution of the landscape, resulting in a decrease of the number of land
Conclusion
The methodology presented in this paper is devoted to the analysis of high resolution dataset in time and space, with a particular attention to the transformation of vineyards management due to local policies. It necessitates a dataset with a spatial resolution sufficient to detect subtle differences between land managements of a given land use. It is best suited to perennial crops rather than annual ones because it is primarly devoted to the study of global trends rather than cyclic variations
Acknowledgements
The authors thank Matthieu Bourmaud, Florent Levavasseur, Marina Chauvet, François-Xavier Weber, and Maxime Jourde for their help in manual digitalization of land use categories. This research was supported by the French National Research Agency (ANR) through the ALMIRA project (ANR-12-TMED-0003). This work was also supported by the metaprogramme Adaptation of Agriculture and Forests to Climate Change (AAFCC) of the French National Institute for Agricultural Research (INRA) through the LACCAVE
References (57)
- et al.
Historical geomorphic analysis (1932–2011) of a by-passed river reach in process-based restoration perspectives: The Old Rhine downstream of the Kembs diversion dam (France, Germany)
Geomorphology
(2015) - et al.
Multi-resolution, object-oriented fuzzy analysis of remote sensing data for GIS-ready information. ISPRS Journal of Photogrammetry and Remote Sensing
Integration of Geodata and Imagery for Automated Refinement and Update of Spatial Databases
(2004) - et al.
Identifying indicators of the spatial variation of agricultural practices by a tree partitioning method: The case of weed control practices in a vine growing catchment
Agricultural Systems
(2009) Object based image analysis for remote sensing
ISPRS Journal of Photogrammetry and Remote Sensing
(2010)- et al.
Mapping ecosystem service supply, demand and budgets. Ecological Indicators
Challenges of Sustaining Natural Capital and Ecosystem Services
(2012) - et al.
Remote sensing image-based analysis of the relationship between urban heat island and land use/cover changes. Remote Sensing of Environment
Thermal Remote Sensing of Urban Areas
(2006) Research note: Sankey diagrams for visualizing land cover dynamics
Landscape and Urban Planning
(2015)- et al.
Monitoring desertification in a Savannah region in Sudan using Landsat images and spectral mixture analysis
Journal of Arid Environments
(2012) - et al.
Spatial scales, stakeholders and the valuation of ecosystem services
Ecological Economics
(2006) - et al.
An automated approach for reconstructing recent forest disturbance history using dense Landsat time series stacks
Remote Sensing of Environment
(2010)
The abandonment of traditional agricultural landscape in Slovakia – analysis of extent and driving forces
Journal of Rural Studies
Driving forces behind vineyard abandonment in Slovakia following the move to a market-oriented economy
Land Use Policy
Common features and different trajectories of land cover changes in six Western Mediterranean urban regions
Applied Geography
Reconstruction of former glacier surface topography from archive oblique aerial images
Geomorphology
Spatial determinants of abandonment of large-scale arable lands and managed grasslands in Slovakia during the periods of post-socialist transition and European Union accession
Applied Geography
Detecting important categorical land changes while accounting for persistence
Agriculture, Ecosystems & Environment, From pattern to process: Landscape Fragmentation and the analysis of land use/land cover change
Remote sensing data change detection based on the CI test of Bayesian networks
Computers & Geosciences
Relevance of use-invariant soil properties to assess soil quality of vulnerable ecosystems: The case of Mediterranean vineyards
Ecological Indicators
Monitoring urban land cover change: An expert system approach to land cover classification of semiarid to arid urban centers
Remote Sensing of Environment
The contribution of crop-rotation organization in farms to crop-mosaic patterning at local landscape scales
Agriculture, Ecosystems & Environment
Monitoring land use and land cover change in mountain regions: An example in the Jalca grasslands of the Peruvian Andes
Landscape and Urban Planning
Opening the archive: How free data has enabled the science and monitoring promise of Landsat. Remote Sensing of Environment
Landsat Legacy Special Issue
Object-based target search using remotely sensed data: A case study in detecting invasive exotic Australian pine in south Florida
ISPRS Journal of Photogrammetry and Remote Sensing
Land cover classification and change analysis of the Twin Cities (Minnesota) Metropolitan Area by multitemporal Landsat remote sensing
Remote Sensing of Environment
Traditional landscape and rural development: Comparative study in three terraced areas in northern, central and southern Italy to evaluate the efficacy of GAEC standard 4.4 of cross compliance
Italian Journal of Agronomy
L’arrachage viticole: Un facteur de mutations pour le département de l’Hérault
Personal, physical and socioeconomic factors affecting farmers' adoption of land consolidation
Spanish Journal of Agricultural Research
Measurement and simulation of evaporation from soil in olive orchards
Irrigation Science
Cited by (26)
Assessment of climate impact on grape productivity: A new application for bioclimatic indices in Italy
2023, Science of the Total EnvironmentSolving photogrammetric cold cases using AI-based image matching: New potential for monitoring the past with historical aerial images
2023, ISPRS Journal of Photogrammetry and Remote SensingDemographic challenges and agricultural abandonment: Solutions for semi-arid winegrowing regions
2023, Journal of Rural Studies