A spatiotemporal dataset for integrated assessment and modelling of crop-livestock integration with the MAELIA simulation platform

The general purpose of the primary and secondary data available in this article is to support an integrated assessment of scenarios of crop-livestock integration at the territorial level i.e. of exchanges between arable and livestock farms. The data is a result of a research collaboration between the scientist from INRAE, agricultural advisers from Chamber of Agriculture of Pays de la Loire (CRAPL) and a collective of five arable and two livestock farmers located in the district of Pays de Pouzauges (Vendée department, western France). All participants formed part of the DiverIMPACTS project (https://www.diverimpacts.net/) that aims to achieve the full potential of diversification of cropping systems for improved productivity, delivery of ecosystem services and resource-efficient and sustainable value chains in Europe. The first dataset corresponds to the inputs of MAELIA (http://maelia-platform.inra.fr/), a spatial agent-based simulation platform that was used to support an iterative design and assessment of scenarios to redesign cropping systems. The second dataset corresponds to the outputs of MAELIA simulations and the associated indicators at the farm, group and territory level. The data comprise multiple shape and csv files characterizing the edaphic-climatic heterogeneity of the territory and cropping systems, farmers’ crop management rules (IF-THEN rules) and general information about the farms (e.g. crops, agricultural equipment, average crop yields). Data is reported for the baseline situation and three exchange scenarios containing different innovative cropping systems co-designed by scientists, agricultural advisers and the farmers. The data presented here can be found in the Portail Data INRA repository (https://doi.org/10.15454/3ZTCF5) and were used in the research article “Fostering local crop-livestock integration via legume exchanges using an innovative integrated assessment and modelling approach: MAELIA” [1].


a b s t r a c t
The general purpose of the primary and secondary data available in this article is to support an integrated assessment of scenarios of crop-livestock integration at the territorial level i.e. of exchanges between arable and livestock farms. The data is a result of a research collaboration between the scientist from INRAE, agricultural advisers from Chamber of Agriculture of Pays de la Loire (CRAPL) and a collective of five arable and two livestock farmers located in the district of Pays de Pouzauges (Vendée department, western France). All participants formed part of the DiverIMPACTS project ( https://www.diverimpacts.net/ ) that aims to achieve the full potential of diversification of cropping systems for improved productivity, delivery of ecosystem services and resource-efficient and sustainable value chains in Europe.
The first dataset corresponds to the inputs of MAELIA ( http: //maelia-platform.inra.fr/ ), a spatial agent-based simulation platform that was used to support an iterative design and assessment of scenarios to redesign cropping systems. The second dataset corresponds to the outputs of MAELIA simulations and the associated indicators at the farm, group and territory level. The data comprise multiple shape and csv files characterizing the edaphic-climatic heterogeneity of the territory and cropping systems, farmers' crop management rules (IF-THEN rules) and general information about the farms (e.g. crops, agricultural equipment, average crop yields). Data is reported for the baseline situation and three exchange scenarios containing different innovative cropping systems codesigned by scientists, agricultural advisers and the farmers. The data presented here can be found in the Portail Data INRA repository ( https://doi.org/10.15454/3ZTCF5 ) and were used in the research article "Fostering local crop-livestock integration via legume exchanges using an innovative integrated assessment and modelling approach: MAELIA" [1] .

Value of the Data
• This dataset offers a unique set of detailed and spatially explicit data on 7 farms including a description of crop management strategies through decision rules. • This dataset can be used by researchers to perform a multi-criteria assessment of crop diversification scenarios considering socio-ecological and economic dynamics. • As MAELIA is an open-source platform ( http://maelia-platform.inra.fr/ ) this dataset can be used to define and simulate new scenarios. • This data allows evaluating self-sufficiency, sustainability and vulnerability of cropping systems from field and farm to group of farms levels. • This data permits to obtain several socio-economic (e.g. gross margin) and environmental indicators (e.g. nitrogen use and quantity of pesticide active ingredient applied) to evaluate performance at various scale (from field to territory).

Data Description
We provide two datasets in this paper that were used in the research article "Fostering local crop-livestock integration via legume exchanges using an innovative integrated assessment and modelling approach: MAELIA" [1] . Firstly, a complete dataset of the inputs necessary to run MAELIA ( http://maelia-platform.inra.fr/ ), a high-resolution agent-based platform for IAM (Integrated Assessment Modelling) of agricultural landscapes considered as socio-agroecological systems [2] . Secondly, the raw outputs of MAELIA simulations and the respective indicators at the farm, group and territory level. Overall, it corresponds to an integration of generic data and local knowledge, as well as the data for the simulated baseline situation and the three scenarios considered, as described below in detail.

MAELIA input dataset
This dataset includes spatially explicit data, in the format of shapefiles ( * .shp ), concerning the administrative divisions, soil mapping units, meteorological zones (8 × 8 km) and, for each farm, field blocks (herein islets) and fields. To avoid any sort of identification, we have anonymised fields that could directly be linked to the farmer. It also includes local and expert-based data that were gathered through direct collaboration with the parties involved in the study, such as the farmers and advisors of local chamber of agriculture. Lastly, it contains the observed crop sequences within each field, crop management strategies described through decision rules, equipment used, production (yield) and economic information (prices and costs). The description of each variable, the unit of measurement, the nature of data and respective units are presented in Tables 1 , 2 , 3 , 4 , 5 , 6 , 7 and 8 . Below we explain each of different data files present in the MAELIA input dataset, and the nature of data, that is available at Portail Data INRA repository (https://doi.org/10.15454/3ZTCF5): • Administrative divisions: Spatial data containing information concerning the second-(ADM2, department.shp ) and forth (ADM4, communes.shp ) -order French administrative divisions, referent to provinces and communes respectively. These data serve as a basis for delineating the territory. • Water catchment area: General information regarding the characteristics of the water catchment area ( ZH.shp ). • Soil mapping units (SMUs) and detailed quantitative soil data ( soils.shp ). Each SMUs of the 1:1 0 0 0 0 0 0 French soil map [3] were tagged to the dominant Soil Typological Unit (STU). Then pedotransfer rules [4] were used to transform qualitative data of STU into quantitative values describing characteristics and properties of the corresponding soil. Finally, this soil data were improved using soil analyses provided by the farmers.    In MAELIA, the daily spatiotemporal distribution of cultural operations over the farm's fields is subject to the time necessary to perform each cultural operation and the spatial distribution and size of fields. For further information see [2,8,9] . • Crop parameters: The crop species file ( crop_parameters.csv ) contains the parameters for plant growth (see Table 4 for a detailed description). • The irrigation equipment ( irri_equi.csv ) contains the equipment used for irrigation (see Table 5 for a detailed description). • Economic information. The economic_info.csv file contains information regarding the crop prices and premiums, milk prices, feeding costs, water costs, and average variable costs for each crop. These data were provided by agricultural advisors from the CRAPL for 2015, 2016 and 2017. For the remaining simulated period (2005 to 2014) they were extrapolated using the agricultural producer price index for each year for each input and output [10] .
The file Indicators.csv provides information regarding all indicators representing performance (Energy yield, Protein yield, Gross margin, Economic efficiency, Nitrogen use, Quantity of active ingredients applied and Workload). These data are shown for each farm, group of farms (arable and livestock) and territory, for the baseline situation and the three scenarios (Coexistence, Complementarity and Synergetic) over 12 years (2005-2017).

Experimental Design, Materials and Methods
The data presented here is linked to a case study belonging to the DiverIMPACTS project ( https://www.diverimpacts.net/case-studies/case-study-11-fr.html ) that is based on existing and newly developed initiatives related to crop diversification in Europe. The case study is located in the district of Pays de Pouzauges in the Vendée department (western France) and it is formed  The design-assessment method using MAELIA is supported by an explicit fine-scale representation of the agricultural landscape with multiple biophysical characteristics (e.g. soil and weather), populated by the seven farmers with individual behaviour and objectives. Two classes of data are necessary to implement the case study in MAELIA: generic data and local expertbased data collected with the relevant stakeholders (see section above "MAELIA input dataset"). In addition, with the farmers and agricultural advisers, we have finely adjusted the parcel boundaries and the respective crop sequence, including the classification of rainfed and irrigated parcels. Soil data were as well amended through soil analysis provided by farmers. With farm surveys realised from March to the end of April 2019, we have collected information about average crop yields, farming practices (pesticide and fertilizer use, tillage and mechanical weed control) and general information about the farm (number of crops, agricultural equipment, etc).
For each farmer, the crop management decision rules were collected in parallel via a dedicated farm survey (Supplementary methods in [1] show the template and an example of the survey used to collect decision rules).
Simulations were done using the farm-agent model, incorporated within MAELIA. This model simulates the daily dynamics of technical operations in each field considering their respective soil, climate and plant states and farm-level constraints. The crop management strategy is represented using a set of nested IF-THEN-ELSE statements translating the decision rules obtained from a survey of farmers. The crop yield is modelled with a generic cropping system model (AqYield [11] that simulates in each field the daily interactions between the soil-water cycle, climate, farming practices and crop growth. Finally, based on the requirements of farmers and advisers we have selected nine criteria and associated indicators to evaluate this case study (see section 2.6 in [1] for a detailed description).

Ethics Statement
The participants were informed about the purpose of the study and data collection process before the interview started. To all participants was given the power of freedom of choice to decide whether to answer or decline the questions, as well as the possibility of refusing to participate or withdraw from the study while it was in progress. Personal information was handled properly under Directive 95/46/EC on the protection of individuals concerning the processing of personal and on the free movement of such data. Confidentiality of the responses was given assuring that the collected data would be used solely for research purposes. The anonymity of the spatial data of this present dataset is guaranteed via the attribution of a random number to each farmer, farm and field, so no link between these three elements is possible.