A brief history

Assessing the sustainability of a cropping system is complex. In field crops, the MASC tool is implemented with the help of the DEXi qualitative decision support software, which allows any complex decision problem to be broken down into sub-problems that are easier to solve. MASC has proven its worth but its use remains to be optimized (significant calculation time). At the same time, more user-friendly tools intended for research are being developed in various fields: the DEXiPM. These tools allow to design)^[1] innovative and sustainable SoC. They exist for field crops, vineyards, vegetables and apples (DEXiPM_Pomefruit®). But in arboriculture, professionals wanted to benefit from this type of tool, adapted to the field, to be able to evaluate the performance of their SoC already in place or for which changes are being considered.

It is in response to this request that the DEXiFruit tool was developed. It derives from the DEXiPM_Pomefruit tool® developed for apple and pear.

DEXiFruits: a hierarchical decision tree

DEXiFruits is a tool to assess the sustainability of orchards.

It is presented in the form of a hierarchical decision tree (Figure 1). This structure allows the decomposition of sustainability according to its 3 pillars (complex approach), then into indicators (aggregated criteria) and finally into data that can be easily retrieved in the field (input criteria).

DEXiFruits has 81 aggregate criteria and 57 input criteria.

Entry criteria

The input criteria correspond either to the description of the context of the evaluated orchard (23 criteria), or to surveys allowing the qualification of certain agronomic practices and certain results of the system (34 criteria).

Each criterion is defined by qualitative classes, such as "low, medium, high".

However, some criteria are quantitative (e.g. yield). For DEXiFruit, the classes of these quantitative criteria are defined using threshold values (Figure 2). These threshold values are parameterized (or to be parameterized) for each species.

Aggregate criteria

The aggregated criteria, located upstream of the input criteria, combine step by step the information included in the lower level criteria on which they depend according to a weighting reflecting their importance. The aggregations are carried out for each criterion thanks to "utility functions" which are materialized by tables filled in by experts according to a qualitative reasoning of the "if-then" type such as :

 IF "criterion 1 is very low" AND IF "criterion 2 is low to medium

THEN "aggregate criterion is very low"

In DEXiPM type tools and contrary to MASC, each criterion (aggregate or input) can be reused several times in the tree. We say that they are "linked". In DEXiFruits, there are 254 links between criteria.

The referent system

In the case where DEXiFruit is parameterized for a species, a reference system (or typical system) is created (Figure 3). A reference system is a fictitious system representative of the national average of cropping systems for a given species.

It is used to compare its performance with that of the system studied. It can also be used to complete a missing input criteria during the evaluation of the cropping system, to decontextualize the evaluation (use the context of the reference system for several systems and compare only their practices). It can also be a basis for discussion.

Output in the form of synthetic graphs

Thanks to the use of the IZIEval interface, the results of the evaluation with the DEXiFruits tool are represented as:

• synthetic graphs that can be completely adjusted according to your evaluation objective
• dashboard that allows you to make a global diagnosis of the production system and to identify its strengths and weaknesses

The graphs

They allow to represent any criterion (input or aggregated) and to evaluate one or more orchards at the same time:

The dashboard

This second type of representation of the results makes it possible to visualize all or part of the tree. Each criterion is characterized by a color corresponding to its evaluation in terms of sustainability (green = good, red = poor). This dashboard allows a detailed evaluation of the orchard and a self-diagnosis of its strengths and weaknesses.
This representation (Figure 5) also makes it possible to identify the levers (cultural practices or contextual elements) that have the greatest influence on the performance of the orchard. It is also a good support for group discussions on the sustainability of orchards.

________
^[1] Messéan et al, 2010, Outils d’évaluation et d’aide à la conception de stratégies innovantes de protection des grandes cultures, Innovations Agronomiques, 8, 69-81