Perma_Crops_PT: A geolocated dataset for permanent crops in Portugal

Abstract

Crop landcover datasets are crucial for modern agriculture, aiding farmers, researchers, policymakers, and stakeholders. These databases offer extensive insights into crop distribution, facilitating informed decision-making for sustainable practices, particularly under a changing climate. Moreover, these datasets drive research, fostering collaborations and innovation for resilient agriculture. In Portugal, the COS dataset is vital, offering insights into agrarian landscapes and supporting sustainable practices. However, in recent versions, since 2007, information on permanent crops has been aggregated, necessitating complementary datasets and tools. The current paper addresses this gap by providing an open-source dataset focusing on perennial crops in mainland Portugal. Based on the 2019 agricultural census from the Portuguese Statistical Institute (INE), this dataset contributes to the spatial understanding of permanent crop distribution, being freely available for researchers, farmers and policymakers. The dataset includes a selection of perennial crops commonly cultivated in Portugal, such as Prunus dulcis (Almond), Malus domestica (Apple), Castanea sativa (Chestnut), Ceratonia siliqua (Carob), Prunus avium (Sweet Cherry), Vitis vinifera (Grapevine), Olea europaea (Olive), Citrus limon (Lemon), Citrus sinensis (Sweet Orange), Juglans regia (Walnut), Citrus reticulata (Mandarin), Prunus persica (Peach), Pyrus communis (Pear), and Prunus domestica (Plum). Further information regarding the Administrative Units of each crop is also available. This comprehensive list provides a detailed overview of the types of permanent crops included in the dataset, offering valuable insights into the Portuguese agricultural landscape.

Specifications Table

Subject	Agronomy and Crop Science; Management, Geography; Agricultural Economics
Specific subject area	Characterization of permanent cropland cover in Portugal
Type of data	Table, Processed
Data collection	Extraction of permanent crop distribution in mainland Portugal from the 2019 agricultural census conducted by the Portuguese Statistics Institute (INE). Spatial data on permanent crop locations were converted into the GeoPackage (GPKG) format using Geographic Information Systems (GIS). Information on crop location associated with administrative units was also collected from the CAOP dataset. This process identified NUTS1, NUTS2, and NUTS3 regions, municipalities, districts, and parishes where permanent crops were located.
Data source location	Original data was collected from the official Portuguese Statistics Institute (INE) Agricultural Census 2019 available under the licence Creative Commons Attribution 4.0 (CC BY 4.0) and available at https://www.ine.pt/xportal/xmain?xpid=INE&xpgid=ine_publicacoes&PUBLICACOESpub_boui=437178558&PUBLICACOESmodo=2
Data accessibility	Repository name: Zenodo Data identification number: doi: 10.5281/zenodo.11220640 Direct URL to data: https://zenodo.org/records/11220640

1. Value of the Data

• •The dataset offers detailed insights into the spatial distribution of perennial crops in Portugal, enabling precise land use planning and resource allocation strategies for optimized and sustainable agricultural production.
• •By identifying areas where specific permanent crops thrive, the dataset supports resilience-building efforts against climate change, aiding in the selection of crop varieties and cultivation techniques that are better suited to future climatic conditions.
• •By mapping permanent crop locations, the dataset aids policymakers and stakeholders in promoting sustainable agricultural practices, conserving biodiversity, and mitigating environmental impacts.
• •Detailed information on crop types and their spatial distribution empowers farmers to make informed decisions (along with the rise in artificial intelligence in agriculture) regarding crop selection, irrigation strategies, pest management, and resource allocation, thereby enhancing productivity and profitability.
• •The dataset addresses the lack of detailed information on permanent crops in existing landcover datasets, providing a valuable resource for understanding the agricultural landscape of mainland Portugal and facilitating comprehensive research and planning efforts.

2. Background

Crop landcover datasets play a pivotal role in modern agriculture, serving as a foundational resource for farmers, researchers, policymakers, and various stakeholders within the agricultural sector [1]. Land Cover databases (e.g. CORINE) provide comprehensive information about the geographical distribution and characteristics of crops, among other land use types, such as temporary and permanent crops [[2], [3], [4]]. Regarding permanent crops, these geolocated datasets allow efficient land use planning. By mapping the location of perennial crops, these databases help identify suitable areas, allow detection of environmentally sensitive areas, and empower farmers and policymakers to make data-driven decisions, particularly with the rise of artificial intelligence systems in agriculture [5], thus promoting sustainable agricultural practices and preserving biodiversity [6]. Permanent crop location databases also play a crucial role in research aimed at improving crop varieties, cultivation techniques, and resilience to environmental stressors, including climate change adaptation [7], such as the case of the Portuguese WaterQB project. In Portugal, the COS dataset [8], maintained by the Portuguese Government, provides information on the distribution, extent, and types of agricultural land use, including the cultivation of various crops. However, the post-2007 COS versions tend to aggregate several permanent crops, except vineyards and olive trees. This missing layer of information can be mitigated by leveraging other sources, such as the agricultural census conducted by the “Instituto Nacional de Estatística” (INE) [9]. It is typically conducted every 10 years and provides comprehensive data on various aspects of agricultural production, including crop types, their spatial distribution, and characteristics of agricultural holdings. The current data paper aims to fill in the gap regarding the location and extent of commonly cultivated permanent crops in Portugal.

3. Data Description

Our dataset comprises the location of 14 permanent crops in Portugal and is freely available for everyone (Fig. 1). This novel open-source dataset focuses on a selection of perennial crops commonly cultivated in Portuguese agricultural landscapes, including Prunus dulcis (Almond), Malus domestica (Apple), Castanea sativa (Chestnut), Ceratonia siliqua (Carob), Prunus avium (Sweet Cherry), Vitis vinifera (Grapevine), Olea europaea (Olive), Citrus limon (Lemon), Citrus sinensis (Sweet Orange), Juglans regia (Walnut), Citrus reticulata (Mandarin), Prunus persica (Peach), Pyrus communis (Pear), and Prunus domestica (Plum). Location data corresponds to point data, each with longitude (X) and latitude (Y) coordinates (WGS84 geographic coordinate system, EPSG:4326). Each point also provides information on the corresponding Administrative Units (e.g. NUTS), merged from the CAOP dataset [10]. Additional information was also included for olive trees and grapevines, which was available in the original INE report, such as planting density (olive trees), and wine geographic indication type (grapevines). The data [11], is freely available through the Zenodo (https://zenodo.org) data repository, and is provided in the form of a GeoPackage file (.GPKG), which is an open, standards-based, platform-independent, portable file format. Information regarding the attribute table is provided in Table 1.

Fig. 1.

Location of the 14 permanent crops over mainland Portugal.

Table 1.

Example of the dataset attribute table.

Attribute	Description
Perma_ID	Unique identifier for the permanent crop
NAME	Scientific name of the crop
NAME2	Common name of the crop
GEO_IND	Geographical indication type (only for grapevines)
DENSITY	Plants per hectare (only for olive trees)
X	X-coordinate (longitude) of the geographical location
Y	Y-coordinate (latitude) of the geographical location
NUTS_1_COD	Nomenclature of Territorial Units for Statistics (NUTS) level 1 code
NUTS_2_COD	Nomenclature of Territorial Units for Statistics (NUTS) level 2 code
NUTS_3_COD	Nomenclature of Territorial Units for Statistics (NUTS) level 3 code
NUTS_1_LAB	Nomenclature of Territorial Units for Statistics (NUTS) level 1 label (name)
NUTS_2_LAB	Nomenclature of Territorial Units for Statistics (NUTS) level 2 label (name)
NUTS_3_LAB	Nomenclature of Territorial Units for Statistics (NUTS) level 3 label (name)
DICOFRE	CODE for the district, municipality, and civil parish
Parish	Name of the civil parish
Municipality	Name of the municipality
District	Name of the district

There are several potential uses for the current dataset, which include:

• •Agricultural Monitoring: Researchers, farmers, decision-makers and policymakers can use this dataset to monitor the distribution and growth of various permanent crops throughout Portugal. This information for perennial crops, not available in order datasets, enables users to assess crop health, predict yields, and manage resources more effectively.
• •Climate Impact Studies: The dataset can be used to study the impact of climate change on crop distribution, allowing access to the optimum zones for each perennial crop under future climates. Researchers can analyse shifts in agroclimatic conditions and evaluate how environmental factors such as temperature, rainfall, and soil quality may affect specific crops.
• •Precision Agriculture: The detailed spatial data can support precision agriculture practices by allowing farmers to optimise inputs (e.g., water, fertilisers, pesticides) based on the specific needs of each crop type and its location.
• •Food Security Analysis: Governments and organisations can use the dataset to model food production trends and forecast potential shortages or surpluses. This data can support policies and interventions aimed at improving agrarian value chains and food security.
• •Land Use Planning: Urban planners and environmental agencies can leverage the dataset to analyse land use and assess agricultural land availability while planning land use to promote sustainability.

4. Experimental Design, Materials and Method

The maps depicting the distribution of permanent crops were obtained from the Portuguese “Instituto Nacional de Estatística” (INE) agricultural census conducted in 2019 [9]. These maps provided comprehensive spatial information regarding the location and extent of various permanent crops across mainland Portugal. Each map, depicting permanent crop distribution, was composed of vector points, each point indicating a 10-hectare orchard. The spatial data representing the location of permanent crops were, georeferenced and converted into the GeoPackage (GPKG) format using QGIS v 3.28. GeoPackage is an open, standards-based platform-independent database format for spatial data storage and exchange, providing efficient data organization and interoperability. Information about the administrative units associated with each crop location was extracted from additional datasets containing administrative boundaries using the CAOP2023 dataset [10]. This extraction process included identifying the administrative units, such as NUTS1, NUTS2, and NUTS3 regions, as well as municipalities, districts, and parishes, within which the permanent crops were located. This information provided context regarding the spatial distribution of crops within different administrative boundaries and facilitated further analysis and interpretation of the data. A validation procedure was done comparing the existing C0S2018 landcover dataset [8], for vineyards and olive orchards with the current dataset. The experimental design is outlined in Fig. 2.

Fig. 2.

Flowchart representation of the experimental design, illustrating the methodology and tools employed in constructing the database.

Limitations

The methodology relies on maps obtained from the INE agricultural census, which may have limitations in terms of scale and resolution. The methodology is based on the assumption that each vector point represents a 10-hectare orchard, which may lead to generalizations in crop distribution. To address this issue, we compared the existing COS2018 for vineyards and olive orchards with those in the current dataset (Fig. 2) [8,10]. While accessing the accuracy of the current dataset against the COS2018 becomes challenging, due different spatiotemporal resolutions, a visual comparison highlights that the current dataset provides a strong spatial agreement, for vineyards and olive orchards (Fig. 2). Nonetheless, the current dataset is limited by the resolution of the original INE report.

The methodology relies on data from the INE agricultural census conducted in 2019 [9], which may not reflect current conditions or recent changes in land use. Updates to the dataset may be necessary to account for changes in crop distribution over time, particularly in dynamic agricultural landscapes. Furthermore, the methodology focuses only on permanent crops that are included in the INE agricultural census, potentially overlooking other types of permanent crops or agricultural practices not captured in the dataset. This limitation could affect the comprehensiveness and representativeness of the dataset for certain regions or crop types.

Ethics Statement

The authors have read and followed the ethical requirements for publication in Data in Brief and confirm that the current work did not involve human subjects, animal experiments, or any data collected from social media platforms. The developed dataset is based on the INE Agrarian Census 2019, which is published under the licence Creative Commons Attribution 4.0 (CC BY 4.0).

CRediT authorship contribution statement

Helder Fraga: Conceptualization, Methodology, Validation, Formal analysis, Data curation, Writing – original draft, Writing – review & editing. Teresa Freitas: Conceptualization, Methodology, Validation, Formal analysis, Data curation, Writing – original draft, Writing – review & editing. Nathalie Guimarães: Validation, Formal analysis, Writing – review & editing. João A. Santos: Conceptualization, Methodology, Validation, Formal analysis, Data curation, Writing – original draft, Writing – review & editing.

Data Availability

• INE, Recenseamento Agrícola - Análise dos principais resultados - 2019. ISBN:978-989-25-0562-6., Instituto Nacional de Estatística, I. P., Lisbon, Portugal, 2021, p. 166. (Original data) (Zenodo).