Scientific publication Paper AgroforesTreeAdvice
Tree species selection
The selection of tree species is a pivotal step in designing sustainable and effective agroforestry systems, and farmers and advisors have expressed a significant need for decision support systems (DSS) to aid in this process. In a survey conducted across six European countries to assess stakeholder needs, the most highly rated topic in terms of perceived usefulness was the development of a tool for selecting appropriate tree species or varieties for agroforestry systems (Tranchina et al. 2024). However, 84% of respondents (among which 57.61% were farmers, landowners or farm advisors), were not aware of any existing tool. Indeed, farmers face several challenges when designing agroforestry systems, starting with a need for knowledge about tree characteristics and performance in a specific context and a changing climate, in order to make appropriate choices on the right species, varieties and combinations. However, this knowledge is largely missing for agroforestry (Ellis et al. 2000). Additionally, the myriad potential combinations of tree and crop species, coupled with the scarcity of scientific data and demonstration plots for many of these combinations, pose a significant obstacle (Wolz and DeLucia 2018). Furthermore, while agroforestry systems can provide a range of ecosystem services (Jose 2009), the effectiveness of these services is contingent upon the specific service, the agroforestry system under study, and the local conditions (Torralba et al. 2016). Lastly, in line with agroecological principles, agroforestry systems should be tailored to local conditions to minimize reliance on external inputs like irrigation, fertilization or plant protection, and should leverage farmers’ knowledge (Wezel et al. 2015) to develop sustainable and resilient agroforestry systems.
These challenges are not new, and over the years, various approaches have been developed to support tree species selection. These approaches can be broadly categorized into three types: participatory, trait-based, and model-based (including ecological niche modeling) approaches. Participatory approaches engage stakeholders directly, utilizing methods such as surveys with semi-structured questionnaires to gather farmers’ preferences (Sebuliba et al. 2022). Other participatory techniques include species ranking exercises, where stakeholders evaluate and prioritize different species based on specific criteria (Van Der Wolf et al. 2016; Kheiri et al. 2024). An example of a participatory technique is the pebble distribution method, which involves stakeholders distributing pebbles to indicate their preferences and priorities for various species (Notaro et al. 2022). Trait-based approaches focus on the characteristics of tree species, such as aerial or root traits, to determine their suitability for specific agroforestry systems (Isaac et al. 2024). These approaches rely on detailed biological and ecological data to match tree species with the requirements and constraints of particular agroforestry settings. Model-based approaches, including ecological niche modeling, use computational tools based on GIS methods to combine soil and climate maps, to predict the suitability of different tree species for various environments. These models take into account a range of environmental and ecological variables to simulate potential outcomes and guide species selection (Ellis et al. 2005; Ranjitkar et al. 2016; Borucke et al. 2020; Tyndall 2022; Shea and Wolz 2024). They can be complemented with information on ecosystem services provided by trees, in order to further refine the choice of tree species.
Once the knowledge is available, it needs to be delivered to users (farmers, landowners, extensionists) as an easy-to-use DSS. The survey by Tranchina et al. (2024) provided the characteristics that stakeholders expect for a DSS to be usable and useful, tailored to their needs and preferences. In particular, it was important to stakeholders that the tools were (i) simple and clear, (ii) intuitive, (iii) accessible on-site, (iv) easy to use and reliable, and (v) compatible and flexible. Another feature that was not mentioned in the survey (because it is so obvious that it stayed implicit), is the support of the user’s language. Existing tools, even when they are available on the web, are often not usable in other countries due to language barriers. Furthermore, they all have a different interface (or no interface if they are in the form of a simple table of tree characteristics), which means that users have to learn how to use each tool. Last but not least, they are not interoperable, meaning that it is not possible to query the content of another tool from a given tool. Integrating several tree selection tools would therefore increase their accessibility by a range of stakeholders across Europe, as well as allowing sharing knowledge across tools.
Our objective was to develop a framework for agroforestry tree selection that would (i) gather and organize the knowledge contained in existing tree selection tools and (ii) provide an intuitive, user-friendly interface to identify tree species, varieties, and rootstocks when needed, adapted to local conditions and effective in fulfilling the farmer’s objectives. We created a unified framework that could integrate various agroforestry tree selection tools, enhancing their (i) findability by consolidating all tools in a single location, (ii) accessibility through a common interface with a consistent look and feel, (iii) interoperability by enabling API (Application Programming Interface) requests to query all tools, and (iv) reusability for other decision support systems (DSS). The following section presents the framework we developed, detailing its agronomical and ecological rationale. We demonstrated its versatility by integrating eight previously developed tree selection DSS and ensured its accessibility via a graphical user interface, called AgroforesTreeAdvice. Finally, we discuss initial user feedback from its deployment in three countries, the benefits derived from this initiative and future work based on further unification of the databases.
Do you want to read the paper? Click here: AgroforesTreeAdvice: a decision support tool combining heterogeneous knowledge resources for tree species selection in agroforestry systems | Agroforestry Systems
Do you want to visit the AgroforesTreeAdvice Tool? Click here: AgroforesTreeAdvice
This text was copied from AgroforesTreeAdvice: a decision support tool combining heterogeneous knowledge resources for tree species selection in agroforestry systems | Agroforestry Systems
