Introduction

Open Analytics often receives requests to visualize and summarize data from multiple data sources in a user-friendly way. Most of these projects result in in-house applications, while projects for governmental institutions typically result in publicly available web applications. A nice example is faunabeheer.inbo.be, developed for the Research Institute for Nature and Forest (INBO) and the Agency for Nature and Forests (ANB) responding to questions from (local) authorities, land managers, hunters, researchers, journalists, citizens and farmers.

This website summarizes data from e-loket fauna en flora by ANB, enriched with measurements of INBO on collected samples (e.g. mandibles and uteri), waarnemingen.be by Natuurpunt and Wilder, an application from Hubertus Vereniging Vlaanderen. It visualizes results for all big game species and other species relevant for nature management in the Flemish region. The visualizations are grouped in categories and subcategories for easy navigation. Though the user can also navigate directly to a specific visualization and share the route to that visualization using the URL. Once a species has been selected, the visualization choices are restricted to the ones with data available.

As an example, the map for management of wild boars in the Flemish region is shown below (https://faunabeheer.inbo.be/app/01_faunabeheer/#Wild%20zwijn/beheer/beheer-regio/mapFlandersUI)

It visualizes the number of wild boars that have been shot per year and per selected region (e.g. municipality, province or 5×5 UTM squares). When selecting a specific region, the reported number of animals shot is shown as popup in the map and in a trend plot for the chosen period.

Other visualizations involve – Summary statistics on (the management of) populations of wild animals such as wild boar, roe deer, red deer, fallow deer and other species – Interactive maps with current and future spread data – Graphs to indicate the population status (e.g. no. of embryos, net weight) – Figures on the number and locations of reported damage cases – Data on the public support for a species based on surveys

Each visualization is set up in a modular way sharing the same functionalities: relevant data filtering and/or visualization options, a short description and the option to download the plot & raw data behind the graph to guarantee full transparency for the end user.

The online application is developed in line with our best practices for R/Shiny and hosted by a ShinyProxy server. The application is automatically deployed using GitHub Actions workflows. All code from the R-package reportingGrofwild is available on GitHub, including technical documentation and integrated automated testing. The data are retrieved from Amazon S3 buckets to separate data updates from application development. Two parallel streams are set up: a development and production environment both for the data and source code, which facilitates easy and thorough testing before each software release. Switching between environments is made easy using the R-package config in combination with a container-env for “R_CONFIG_ACTIVE” in ShinyProxy.

References

INBO post on LinkedIn: https://www.linkedin.com/posts/inbo-research-institute-for-nature-and-forest_faunabeheer-grofwild-wildbeheer-activity-7343910931208257540-6EHI.